Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
I think they would have worded it different if we hadn't hit 233 events. They told us when they expected it. And that was several months ago. Unlikely that it hasn't been reached. But I won't say impossible because we are blind to the number of events over 231.
I made it clear that I shared my opinion. No where in my opinion to I claim being content. I clearly expressed disappointment. I understand that it's a 3 year trial from the last patient in. I will not be shocked if they go the 3 years. But yes, I do think it would be wonderful to data lock with at least 2 years of post enrollment data on all patients. I just don't have high hopes for that based on this 10Q. The trial can run full term. And they never guaranteed to any of us when they would data lock an ongoing study. What would be horrible is if it 3 year study ends and they still don't data lock. Now that I'm not expecting. That to me would require some serious explanation on their parts. This does not. If they see it as not mature data then yes, I get the rationale for waiting to data lock.
I'm planning on going silent again shortly Turtle. It's honestly how I read it. I really thought 2 years post enrollment would be good. And I thought they would figure out a way to renegotiate the fees associated with stopping the trial early -- if their true intention is to data lock soon; they'll want to do that. However, they discussed on-going program cost. And doing it on an individual scheduling basis. That kind of language to me means they want to get the remaining injections done. If they plan to do more injections, then it doesn't make sense to data lock. No reason to pay the fees, unless that 9th and 10th injection is going to be included in the analysis and not in a post hoc way IMHO.
Happy holiday to you and yours. :)
Flip,
It isn't just "gradually".
I believe they told us they haven't data locked, as events continue to accumulate.
"The Company is continuing to collect data “events” in the Phase III trial, and the data set is gradually maturing. " -- 10Q
And they told us that they are working out arrangements to reduce their on-going trial cost essentially; there is only one running program (DCVax-L) to do such test at the moment.
"While our DCVax programs are ongoing, the Company is required to pay certain fees for dedicated production suites or capacity reserved exclusively for DCVax production, and pay for a certain minimum number of patients, whether or not we fully utilize the dedicated capacity and number of patients. The Company and Cognate are testing an arrangement under which the Company would no longer have capacity dedicated to its programs and instead would have to arrange for advance scheduling in regard to each patient individually. " --10Q
And, to be clear, I think they have a habit of making us think things are closer than they appear; and once they extend or change their plan; they disclose it creatively. Remember Bosch's slide over the summer stated the Company would data lock at a minimum of 233 OS events. If they end up data locking in a year from now and it's 260 events or more, then the "minimum" of 233 OS events disclosure remains accurate. A bit misleading perhaps considering at the time it was written we were not quite at 233 OS events. But it would remain an accurate disclosure nonetheless.
Hello, to all I know.
Nice to see our share price rise. For the most part, I don't pay attention to the day to day moves. But I have been waiting for some news. Today's 10Q counts.
I think they just disclosed to us that they intend to run the trial until the data matures. And so my hope of the trial ending in a data lock, once they have at least 2 years of post enrollment data on all patients (which is a undisclosed day this month), has just been dashed. Really disappointed to read the following below as a part of me just has been waiting patiently for the trial to end.
"The Company’s lead product, DCVax®-L, is designed to treat solid tumor cancers in which the tumor can be surgically removed. This product in an ongoing Phase III trial for newly diagnosed Glioblastome multiforme (GBM). 331 patients have been enrolled in the trial, and enrollment is closed. The Company is continuing to collect data “events” in the Phase III trial, and the data set is gradually maturing. The Company is also working on preparations for Phase II trials of DCVax-L for other indications." -- 10Q
But I do get that having the most mature data --before locking and doing the final analysis -- means that there will be less post hoc analysis needed later on. The less post hoc, the better. Still not happy to share that in my opinion the wait isn't over, it's gradually getting less and less.
Technically the trial is meant to go 3 years after enrollment on all patients, with a 2.5 years injection schedule. Thus far we have gone 2 years. They haven't stopped treatments. But, they are IMHO trying to create the impression that they plan to end the study early. I don't know if folks remember but I discussed the Company ending the overall study at the same time that they data lock. When they data lock, there is no question in my mind, they will want to stop the injection schedule of the study. They will want to stop spending on the trial once they lock IMHO. But will they be doing it before the three years are up? I'm not so sure anymore. Looks to me that they are instead working out arrangements where the Company keeps the trial running at a reduced run rate, paying for the remaining "alive" in treatment patients only. One main reason for the Company to redo such "requirement" fees arrangements is if they are really trying to keep the program (trial) "going" in order to get to the end of the study without having to stop or end it early. From the 10Q:
"While our DCVax programs are "ongoing", the Company is required to pay certain fees for dedicated production suites or capacity reserved exclusively for DCVax production, and pay for a certain minimum number of patients, whether or NOT we fully utilize the dedicated capacity and number of patients. The Company and Cognate are testing an arrangement under which the Company would no longer have capacity dedicated to its programs and instead would have to arrange for advance scheduling in regard to each patient individually." -- 10Q
There is language that they are in discussion with Cognate to rework the shut down fee arrangements. However, with the use of their words in relationship to the data "gradually maturing" (as seen in the earlier posted 10Q passage above), I believe that they don't feel the data is mature enough. And while it may be good news for shareholders if they redo the "suspend" fee arrangements --particularly if they do end the trial early -- it may just end up being written to simple provide shareholders with hope that "news is imminent". Seeing them redo the fee structure may lead us to believe the overall trial will end soon. And maybe it will but my gut is telling me that it will not. And that they creatively just found a way to let us believe the trial end is near, results are close; while at the same time, they have disclosed a delay to data lock -- the gradually maturing data needs to age after all!; so the study will go until all patients have had 2.5 years of injections at minimum; and we'll be paying less and less in trial cost as we wait -- assuming they do rework the payment for services arrangements.
We've all seen this many times:
"Prior to the last dose of the last patient enrolled in the Phase III trial for DCVax®-L or After the last dose of the last patient enrolled in the Phase III clinical trial for DCVax®-L but before any submission for product approval in any jurisdiction or after the submission of any application for market authorization but prior to receiving a marketing authorization approval: in any of these cases, the fee shall be $3 million." -- 10Q
This part below is new news today; and is written in a way to make it clear that the fee has not triggered. It also makes it clear they are renegotiating the shut down/suspend fees. But, that again isn't the only fees they are renegotiating, they are trying to reduce "ongoing" fees. Perhaps they are trying to make us think it will possibly "stop" soon, after all they are trying to determine overall fees owed for 2016 and 2017; and they try to assure us it will be a "reduced" amount. However, the next paragraph clears up that "ongoing" trial requires that the company pays dedicate -- though not necessarily warranted -- utilization fees. And so it is those fees that will reduce the 2016 and 2017 costs, once negotiations are final.
"As of September 30, 2017, none of the above fees were triggered; however, the Company and Cognate are discussing these fee provisions as part of an ongoing negotiation to determine the overall amounts to be paid for 2016 and 2017. The Company anticipates that when a determination of the amounts is reached, it will involve a reduction of the amounts that would otherwise be due under the contracts."
"While our DCVax programs are ongoing, the Company is required to pay certain fees for dedicated production suites or capacity reserved exclusively for DCVax production, and pay for a certain minimum number of patients, whether or not we fully utilize the dedicated capacity and number of patients. The Company and Cognate are testing an arrangement under which the Company would no longer have capacity dedicated to its programs and instead would have to arrange for advance scheduling in regard to each patient individually. In addition, the Company and Cognate are in the process of negotiating an overall settlement of amounts owed to Cognate for 2016 and 2017." --10Q
Anyway, my take on today's 10Q is that we may be waiting a while but they certainly wouldn't mind us thinking that we were close to trial data news. But we probably are not. We likely have at least some or if not all of the last year of the Phase III trial to go if my interpretation is correct.
I hope all is well. Happy holidays.
On the consent form there is a chart, and the 9th shows as a 24 month visit. That is 2 months further out than the 96 week, from day 0 injection. But you are correct that many of these we may never know as they won't be perfect. I agree it will depend on treatment visits. What is even going to tougher to speculate on is the crossover. The protocol does say somewhere about + or - days, so if who knows maybe the main arm will be approaching the 9th injections soon and the full 248 PFS crossover is approaching it's 12 month injection. My hope is the Month of October injections is the last one that we are waiting to occur so that the Month of November completes the 30 day necessary to monitor eCRF for AE. Data lock, etc. Fingers crossed. :)
Flip,
The clinical site registry is wrong. Both the protocol and the consent form have the treatment schedule as follows:
"Study Visits, Procedures and Schedule
At each study visit, which occur every 2 months during the active treatment phase, you will be seen by your health care provider or one of the study doctors. You will have a physical exam, a neurological exam, and a Karnofsky Performance Status form will be completed. A blood sample will be drawn for both routine and research tests, approximately 7 tablespoons of blood will be taken each time. At eight of the study vaccine administration visits (with the exception of study crossover), an additional 7 tablespoons will be taken to monitor your immune response. These visits are Day 0 (first study vaccine administration), Day 20 (third study drug administration), then 4 months, 8 months, 12 months, 18 months, 24 months, and 30 months.
You will have a brain scan (MRI) on all follow-up visits, i.e. every 2 months. Visits where MRIs are taken will last about 1 hour."
http://neurosurgery.ucla.edu/Workfiles/Site-Neurosurgery/Brain_Tumor_Program/11-000686-%20Main%20ICF%2007Nov2012.pdf
The protocol states:
"All randomized patients receive up to 10 immunizations of DCVax-L or autologous MNC (placebo cohort) at days 0, 10, 20, and at months 2, 4, 8, 12, 18, 24 and 30) following recovery from surgery and radiation with concurrent temozolomide."
Maybe the clinical registry is just the injection timeline when they first opened the study back in 2008, and perhaps that is why those analysis are analyzed as a Vanguard last. Not sure why it doesn't align.
I'm like shareholders Happy, Turtle and China in the sense that I would like more disclosures. They have been sued so often that they adopted a less is more PR approach. I certainly do not mind that others push them to share more information. But, I also never expected the trial to truly end by August, regardless that 233 events were expected to occur in July. They did reveal that the data lock process is a multi-Month process, so there certainly is a chance they are in the midst of doing that now and that study news is close. They will share something once they have something material to share -- which is what they have been doing for a while now. I'm exercising patience until then.
Hi DMB, yes, it's true. GBM Mesenchymal tumors do have higher concentrates of TIL. These type of trying to mount a response to their disease, to varying degrees. However, think of it as faulty T-cell wiring, that they have no idea how to recognize the disease cells, so without a vaccine such immune advantage is lost. But once an immunotherapy is added, their chance of survival improves. At the point of RECURRENCE, most GBM tumors pick up more Mesenchymal signature and begin to recognize that something needs to be attacked. But without a clear set of targets, the progression disease typically goes unchallenged and the patients eventually succumb to it.
Gene Expression Profile Correlates with T-Cell Infiltration and Relative Survival in Glioblastoma Patients Vaccinated with Dendritic Cell Immunotherapy
Robert M. Prins1,2,3, Horacio Soto1, Vera Konkankit1, Sylvia K. Odesa1, Ascia Eskin4, William H. Yong5, Stanley F. Nelson2,3,4, and Linda M. Liau1,2,3
Abstract:
Purpose: To assess the feasibility, safety, and toxicity of autologous tumor lysate–pulsed dendritic cell (DC) vaccination and toll-like receptor (TLR) agonists in patients with newly diagnosed and recurrent glioblastoma. Clinical and immune responses were monitored and correlated with tumor gene expression profiles.
Experimental Design: Twenty-three patients with glioblastoma (WHO grade IV) were enrolled in this dose-escalation study and received three biweekly injections of glioma lysate-pulsed DCs followed by booster vaccinations with either imiquimod or poly-ICLC adjuvant every 3 months until tumor progres- sion. Gene expression profiling, immunohistochemistry, FACS, and cytokine bead arrays were performed on patient tumors and peripheral blood mononuclear cells.
Results: DC vaccinations are safe and not associated with any dose-limiting toxicity. The median overall survival from the time of initial surgical diagnosis of glioblastoma was 31.4 months, with a 1-, 2-, and 3- year survival rate of 91%, 55%, and 47%, respectively. Patients whose tumors had mesenchymal gene expression signatures exhibited increased survival following DC vaccination compared with historic controls of the same genetic subtype. Tumor samples with a mesenchymal gene expression signature had a higher number of CD3þ and CD8þ tumor-infiltrating lymphocytes compared with glioblastomas of other gene expression signatures (P 1/4 0.006).
Conclusion: Autologous tumor lysate–pulsed DC vaccination in conjunction with TLR agonists is safe as adjuvant therapy in newly diagnosed and recurrent glioblastoma patients. Our results suggest that the mesenchymal gene expression profile may identify an immunogenic subgroup of glioblastoma that may be more responsive to immune-based therapies. Clin Cancer Res; 17(6); 1603–15. !2010 AACR.
I believe they have a specific Month in mind to data lock. The 233 OS is the minimum needed, but does not necessarily represent their data analysis point. This study has been costly. Better to have all the data included within the analysis than to censor it in order to conduct a final SAP. And so I truly feel they plan to data lock at a sons specific point in mind. See Oct/November injections as possible what they are aiming for. I will not fault them if they will lock with a ton of tertiary data such as when all 331 patients have past their 9th injection point (24 month post enrollment) and when most patient will have past their 10th injection (30 months post enrollment). By November they may have also reached the booster 8-12 months injection point on all crossover patients (if the 248-PFS event hit in Nov 2016). This trial has another year to go before all 331 patients are out 36 months post enrollment. In my view it is unnecessary to continue following the strict trial protocol when most injections are done. They can easily end 6 months or more earlier. Long tail survival does not need to be done on an open study. And so I don't think the company plans to officially data lock until they feel comfortable with simultaneously ending the high cost of the study -- essentially, censoring data, ending the trial at the same time and conducting analysis then.
Senti, you're right about one thing, they've figured it out. Unfortunately you haven't. I hope it's okay, but this is going to be a very long post as I truly want you to understand what I see that you're missing.
First, let's agree that differentiation between true progression and pseudoprogression is not possible on the initial MRI. The only way to distinguish the two is doing a biopsy. That's not always feasible. And so RANO recommendations, is that true progression on initial MRI must be reported ONLY if there is new enhancement outside the radiation field (beyond the high-dose region or 80% isodose line) or if tumor is seen on histopathological sampling (BIOPSY). This monitoring the first scan became necessary once TMZ was brought upfront, because in the words of UCLA "increased vascular permeability from cytotoxic therapies including radiotherapy and anti-neoplastic treatments have been shown to result in increased contrast enhancement in the context of therapeutic benefit, a phenomena known as “pseudoprogression."
Do you ever remember reading this?:
The Surgery Working Group of RANO identified surgically related end-point evaluation problems that were not addressed in the original Macdonald criteria, performed an extensive literature review, and used a consensus-building process to develop recommendations for how to address these issues in the setting of clinical trials.
And the Surgical Working group of RANO determined this:
"Accordingly, an imaging-based response classification scheme must allow a retrospective categorization of a patient as having disease progression based on the results of serial radiographic imaging. When progression is suspected but treatment-related effects remain a valid possibility, the response should be called indeterminate. If subsequent eval- uation proves that the changes reflect true progression, the time of progression should be retrospectively corrected to the first time at which an indeterminate response was noted."
And this:
"determination of a patient’s status to the central review body and will allow retrospective identification of pseudoprogression as described above."
Those quotes come from these paragraphs:
"Retrospective Assessments of Disease Progression Should Be Permitted in Clinical Trials. As discussed above, there are no validated imaging modalities that reliably distinguish treatment effects from tumor progression, and the clinical management often involves serial imaging studies to determine the persistence of the radiographic changes. In many patients, serial imaging evaluation will show that the extent of new enhancement has regressed without clinical intervention, and these patients are considered to exhibit ‘‘pseudoprogression.’’64,65 Others will show persistent progression of imaging changes and will be considered cases of true disease progression. Accordingly, an imaging-based response classification scheme must allow a retrospective categorization of a patient as having disease progression based on the results of serial radiographic imaging. When progression is suspected but treatment-related effects remain a valid possibility, the response should be called indeterminate. If subsequent eval- uation proves that the changes reflect true progression, the time of progression should be retrospectively corrected to the first time at which an indeterminate response was noted." -- from Application of Novel Response/Progression Measures for Surgically Delivered Therapies for Gliomas: Response Assessment in Neuro-Oncology (RANO) Working Group
And this:
"Central Review: Although overall survival is the most objective end point to measure patient outcome, progression-free survival is often used as a surrogate end point for overall survival. With progression-free survival, a meaningful difference in outcome can be detected after a shorter trial duration, and it is not affected by subsequent effective salvage therapies, which may dilute the benefit seen with the earlier therapy. However, whereas overall survival can be assessed objectively, progression-free survival determined by the interpretation of MRI findings is prone to a number of potential biases owing to the subjective nature of imaging assessment.66 Blinded independent central review is one strategy proposed to reduce the potential for this bias. Blinded independent central review should ideally be incorporated into trial design pro- spectively. Because blinded independent central review will be used to guide decision making in individual patients in real time, there will be a requirement for the technological ability of all sites to transfer images to central review without delay. Furthermore, there will need to be acceptance of the idea that participating investigators will cede final determination of a patient’s status to the central review body and will allow retrospective identification of pseudoprogression as described above." -- from Application of Novel Response/Progression Measures for Surgically Delivered Therapies for Gliomas: Response Assessment in Neuro-Oncology (RANO) Working Group
The above is in line with what I have repeatedly told you. You honestly believe that Central Review will have had no choice -- due to when therapy was initiated -- but to review the first MRI scans following baseline of both true and false progression images, due to lack of time, and in an effort to avoid potentially harming GBM patients, they would have no choice but to skip a confirmation scan, and instead rule the both the unconfirmed true and false scans as progression events. Well, that's faulty logic. You should not need me to tell you that suspicion of progression, without confirmation of some sort of progression, does constitute a progression event. Size alone early on can be misleading, and rightly so, as one scan after initiating treatment is rarely ever enough to define treatment failure.
In my opinion it is common sense to 'confirm' what is seen on the first MRI is TTP verses PsP treatment effects -- which incidentally isn't just isolated effects related to Immunotherapy. Radiation makes therapeutics penetrate the blood brain more effectively. Adjuvant Temozolomide also causes early Psuedoprogression (PsP) effects, and as such, RANO criteria mandates that the first 12 weeks of initiating a therapy confirmation of TTP verses PsP be determined retrospectively. You want to argue harm to a patient by not failing to rule progression early on. You should also know that I can argue that an incorrect diagnosis of a true tumor progression (TTP) could result in erroneous termination of an effective treatment, with a potentially negative influence on patient's survival. It should be understood that the positive false-negative PsP response can be a result of any one of the effective treatments in use. And stopping effective treatments without confirmation that the treatment is in fact ineffective can have life threatening consequences. Newly diagnosed GBM that has been surgically resected is a condition that can be beatable for some. GBM odds of survival are far more dismal once the disease recurs. Logic should reason that a newly initiated therapy should not be stopped for early unclear scans, and that patients should not be crossed over until confirmation of treatment failure has been met and is at least confirmable by a storyline of serial MRI scans. One months MRI scans does not tell the full clinical picture. What you imagine in your mind to be truthful and describe to me, if true would be an outcome of poor study design, it is not the outcome of treatment failure. Treatment failure does not exist in your scenario. And in a clinical trial in a disease indication such as GBM in which standard of care alone is known to produce a pseudoprogression responses, it's impossible to suddenly call, what is not ever deemed progression on the basis of one months scans outside of a clinical trial as progression inside a clinical simply because what one is reviewing is blinded scans. Early Pseudoprogression still needs to be ruled out!
Immunotherapy is known to induce nonspecific imaging changes. It's not the only therapy that alters the brain chemistry. If size changes happen outside the field of radiation after initiating therapy, on the very first, scan that would automatically be ruled progression. Anything else within the field of radiation is fine. Size does not matter, instead change over time is what does. It's RANO criteria standard not to define progressive disease without a adjuvant TMZ serial imaging picture. You continue to discard that fact. But that fact is important. RANO criteria standards allow for trial to trial comparisons. It looks to uniform response criteria. What is not progression in one study can not be labeled as progression in another. And I plan to remind you why. But first, skim this:
"First Progression Definition:
Progressive disease < 12 weeks after completion of chemoradiotherapy Progression can only be defined using diagnostic imaging if there is new enhancement outside of the radiation field (beyond the high-dose region or 80% isodose line) or if there is unequivocal evidence of viable tumor on histopathologic sampling (eg, solid tumor areas [ie, > 70% tumor cell nuclei in areas], high or progressive increase in MIB-1 proliferation index compared with prior biopsy, or evidence for histologic progression or increased anaplasia in tumor). Note: Given the difficulty of differentiating true progression from pseudoprogression, clinical decline alone, in the absence of radiographic or histologic confirmation of progression, will not be sufficient for definition of progressive disease in the first 12 weeks after completion of concurrent chemoradiotherapy." -- RANO criteria
By your logic you believe that the reaction by Central Imaging Review is that of no choice but to ignore the logical approach of the 'confirmation scan'. And I'm here to tell you that is not possible. Early Pseudoprogression resulting from RT/TMZ (as you may or may not know) doesn't just only show up in the immediate window following concurrent RT/TMZ; it also shows up within the first 3 months of the adjuvant TMZ treatment phase. The changes seen on the first MRI (in either arm) can still be a result of radiation chemotherapy induced Pseudoprogression. Please tell me you recognize that? I imagine you do. You hopefully also understand that during the first 12 weeks following RT/TMZ, RANO criteria makes it difficult to rule progression (treatment failure) from adjuvant TMZ therapy as it is understood that the induced false-negative effects may be RT/TMZ pseudoprogression.
Follow-up imaging needs to be employed to make a diagnosis of PsP. Central Image Review will have no way of knowing whether they are looking at possible PsP or TTP of adjuvant TMZ therapy or possible PsP or TTP of Immunotherapy treatment. They can't rule early progression on an MRI that fits the description of possible effects of adjuvant approved TMZ therapy. How do you propose that they do that? They can't. To presume they can one needs to rely on faulty logic. Again the Pseudoprogression response can be the result of RT/TMZ or adjuvant TMZ. A biopsy is the gold standard, but it is not a commonly appliable approach in clinical practice, as it involves the risk associated with surgery. There is no way to reliably distinguish what they are seeing isn't a reaction to standard of care without a clinical picture. And yet, essentially you're proposing that they do; you're proposing that if chemoradiation were to induce a Pseudoprogression response that showed up at month 2 MRI scan -- which again is possible given RT induced permeability -- it constituted a Phase III progression event (treatment failure) and resulted in a patient crossover event in the cohort that experience the false progression (vaccine or placebo arm), regardless that the PsP result may originated from RT/TMZ and not from the vaccine or the placebo. Again, TMZ induced Psuedoprogression early on gets a hall pass for those first 12 weeks until a clinical MRI storyline is established But yet, you feel that somehow, any Psuedoprogression found within that same 12 week window automatically constitutes a progression event for a placebo or vaccine cohort patient. Well that just boggles my mind. Please take off the skeptics thinking cap and put on your own. Think it through, early Pseudoprogression is not progression when it's caused by TMZ, so it can't be deemed a progression event when caught in that same window of time if it induced by immunotherapy. If they can't tell what it is then they can't blame the TMZ response on the vaccine. And the vaccine will need to get a hall pass on those first 12 weeks of therapy as well. Progression needs to be back dated to the first date it was detected if the condition does not improve. It doesn't need to called on mere suspicion. Again pseudoprogression decision tree is a must. And just as the Compassionate Use Arm needed to see 'worsening' to confirm early TTP (rapid progression) verse early PsP, and used the category of 'Indeterminate' when the initial condition did not get significantly better or worse, or stabilized, this Phase III main arm Month 4 scan trial will do the same. RANO criteria makes it clear that the only full proof way to determine early TTP verses early PsP is by a pathological confirmation regardless of what the suspected cause is or serial scans. Accept that as fact, as it is.
Continuing on, in an effort to get you to see the facts once and for all back on March 2011, before this Phase III study resumed recruitment and enrollment the RANO working group released the following paper, "Clinical Trial Endpoint for HGG: the Evolving Landscape" which further delved into the new 2010 RANO criteria. In the next section I will point out quotes lifted from that paper to further Illustrate early pseudoprogression is accounted for with serial follow-up.
Neuro Oncol. 2011 Mar; 13(3): 353–361.
doi: 10.1093/neuonc/noq203
"Clinical trial end points for high-grade glioma: the evolving landscape*"
Within it, a consensus recommendation, to which incidentally UCLA members contributed:
"The current article provides a review of the literature as well as consensus recommendations from the RANO working group for commonly employed end points in HGG clinical trials, and it discusses their applicability in the design of clinical trials for newly diagnosed and recurrent patients."
There is also this:
"Unfortunately, no currently available imaging modality reliably distinguishes pseudoprogression from true progression with sufficient sensitivity and specificity for routine use. In such situations, progression can only be reliably defined if it occurs at a distant site, is confirmed histopathologically, or worsens on sequential imaging."
And I QUOTE with regards to discussion in the paper on PFS:
"Current radiographic methods to define progression are problematic due to confounding factors including pseudoprogression and pseudoresponse, as discussed above. Improved guidelines, SUCH AS THOSE DEFINED BY RANO criteria, are EXPECTED TO LESSEN THE IMPACT of these issues."
And by these 'problematic' issues, it is clear to me they are referring to PSEUDOPROGRESSION! See, RANO criteria lessens the IMPACT -- with its 12 week window rule!
That established the consensus for the uniformed need of a psuedoprogression decision tree, unless progression is truly suspected it can be confirmed by a BIOPSY.
And then there is this:
"PFS measures time from treatment initiation to either progression or death from any cause. PFS end points have several potential strengths. First, the duration of time without tumor progression is usually clinically meaningful and, in general, reliably reflects treatment effect, because HGGs do not typically exhibit prolonged disease inactivity (in contrast to low-grade glioma)."
To you see that Senti? It states PFS as an endpoint "reliable reflects treatment effect". The only way to reliable reflect treatment effect, when using MRIs -- which is unreliably at distinguishes pseudoprogression from true progression due to positive or negative treatment effects responses being unclear -- with sufficient sensitivity and specificity for by a "confirmed histopathologically, or worsens on sequential imaging." And so, clearly to me that is why it is recommended consensus in the same paper to:
"On the basis of laboratory advances that have significantly enhanced our understanding of cancer biology, a large number of therapeutic agents with diverse mechanisms of action are in clinical development. It is therefore now more important than ever to recognize the strengths and potential pitfalls associated with various clinical trial end points to evaluate these agents as efficiently as possible."
And it established:
"3. Metrics of clinical and radiographic evaluation should be standardized and should incorporate the recently defined RANO criteria."
(Incidentally, hopefully you can see, without me having to point out that your concern of the true recurrent patient doom, a solution can be to conduct a biopsy, if subsequent imaging can not wait.)
These enhancements that simulates tumor growth, and look like Apparent Early Tumor Progression (AETP) are most often caused by radiation (whole brain or focal). And so, it is known that growth of existing lesions or appearance of new lesions within 12 weeks of completion of radiation therapy may be the result of treatment effects rather than growth of tumor. Now what does the current RANO standardized imaging approach do about these AETP growths, after enrollment upon initiation of therapy?
Answer: a retrospective review of the very first scan to standardize PFS.
And so in short, what happens during the first 12 weeks when an AETP shows up?:
Continued follow-up imaging can determine whether initial lesion growth was true progression or pseudoprogression.
-- If lesion continues to enlarge, the initial growth is called true progression.
-- If lesion stabilizes or shrinks, the initial growth is confirmed as pseudoprogression.
-- In such cases, the baseline SPD is no longer included when choosing the nadir value for the purposes of determining when progression occurs.
That spelled out above is a pseudoprogression decision tree.
Now you are telling me that there it is not safe to start treatments where DCVax-L did and follow RANO monitor the first 12 weeks after Concurrent RT/TMZ. I already pointe out a biopsy can rule progression prior to the completion of the 12 weeks. However, let's put that fact aside for a second and discuss RANO criteria baseline start of this Phase iii trial. You feel it's risky to follow that rule. If that is true than why would the new recommendation for modifying the current RANO to include both A and B below (note there are others but I only reference these two to make a point)?:
A) the use of DCvax-L's as a standardized use of the post-radiation time point as the baseline for newly diagnosed glioblastoma response assessment"
And:
B) A pseudoprogression decision tree, at that same said baseline:
"and “treatment-agnostic” response assessment rubrics for identifying pseudoprogression, pseudoresponse".
Answer, they wouldn't suggest it if it was unsafe. They understand a biopsy is useful for some cases of true tumor progression (TTP).
Do you really think RANO criteria was not an evolution of lessens learned from MacDonald? I have stated many times that it was. The limitations of the Macdonald Criteria become even more apparent with the increased incidence of pseudoprogression in patients receiving radiotherapy with Temozolomide (TMZ). I can debate/prove it with a Q&A? What did UCLA do in the Phase I/II clinical trial when they came across Brad Silvers' very first (and only) AETP scan? Answer: They suspected psuedoprogression (PsP) and so they monitored his scans. As I have told you many times before, once TMZ was brought upfront, the phenomenon of PSD became understood. Clinical investors can always take a biopsy. But, generally early on they monitor scans. It's a lesson learned from the STUPP landmark. "Many clinicians are reluctant to change therapy based on the first examination due to the relatively high incidence of treatment-related radiographic changes directly after completion of concurrent chemotherapy and radiation, and instead use this first scan as a new baseline in which to interpret subsequent changes in tumor size.
RANO criteria just made it a formal standardized approach. Research paper from earlier this year is below:
Keep in mind when you're reading the proposed modified RANO, with regards to the Pseudoprogression decision tree, that differences between conventional RANO and the proposed modified criteria are: (1) use of the post-radiation time point as the baseline for
response evaluation in newly diagnosed GBM and (2) considering only objectively defined, measurable enhancing disease in the definition of response and progression (i.e. exclusion of qualitatively assessed T2/FLAIR changes).
It is not new approach to continue therapy beyond suspected radiographic progression. Therapy early on continues "if the treating physician believes there may be a therapeutic benefit and to provide criteria for defining progression and early drug failure while also allowing for the possibility of PsP and PsR, a modified response rubric similar to those described recently [97] should be employed."
Modified Criteria for Radiographic Response Assessment in Glioblastoma Clinical Trials
Benjamin M. Ellingson, 1,2,3,5 Patrick Y. Wen,4 and Timothy F. Cloughesy5,6
Abstract
Radiographic endpoints including response and progression are important for the evaluation of new glioblastoma therapies. The current RANO criteria was developed to overcome many of the challenges identified with previous guidelines for response assessment, however, significant challenges and limitations remain. The current recommendations build on the strengths of the current RANO criteria, while addressing many of these limitations. Modifications to the current RANO criteria include suggestions for volumetric response evaluation, use contrast enhanced T1 subtraction maps to increase lesion conspicuity, removal of qualitative non-enhancing tumor assessment requirements, use of the post-radiation time point as the baseline for newly diagnosed glioblastoma response assessment, and “treatment-agnostic” response assessment rubrics for identifying pseudoprogression, pseudoresponse, and a confirmed durable response in newly diagnosed and recurrent glioblastoma trials.
Introduction
Approximately 89,000 new primary brain tumors are diagnosed in the United States each year, for which 27% are gliomas and 32.8% are malignant [1]. Glioblastoma (GBM) occurs in approximately 46% of gliomas [1] and has a poor prognosis of around 14 months median survival [2] and less than 10% of patients live longer than 5 years from diagnosis [3]. The current standard of care for newly diagnosed GBM patients consists of maximum safe surgical resection followed by external beam radiation therapy plus concomitant and adjuvant temozolomide [2], particularly in patients that demonstrate O6-methylguanine-methyltransferase (MGMT) promoter methylation. At recurrence there is no consensus as to the standard of care as no therapeutic options have produced substantial survival benefit [4].
Although overall survival (OS) is the standard for determining GBM treatment efficacy, using OS as an endpoint when studying new therapeutic strategies can be problematic because of potential influence of therapies prior to or subsequently following the therapy being studied. For example, it is difficult to definitively conclude that bevacizumab has no efficacy in GBM when a large percentage of patients in the placebo arms in both III trials studying efficacy of bevacizumab (i.e. AVAglio and RTOG-0825) eventually crossed over and received bevacizumab (31% in AVAglio [5] and 48% in RTOG-0825 [6]). If bevacizumab increased OS when given at any time during treatment, we may expect both treatment arms to have similar median OS since most patients eventually were treated with bevacizumab, disguising any therapeutic effects of the drug. Together, these results suggest OS may not be a suitable endpoint when studying new therapeutics or when there is a high chance of cross over in the control arm.
To overcome the limitations associated with using OS as the primary endpoint in studies involving new therapeutics, progression-free survival (PFS) and objective response rate (ORR) should be considered important end points [7]. However, PFS and ORR also have challenges, as determination of response and progression using anatomic imaging techniques may suffer from issues associated with measurement variability and discordance in interpretation between radiologists [8]. Therefore, it is important to develop both new response guidelines for identifying these issues as well as new imaging tools for better differentiating treatment-related changes from changes associated with non-responsive, growing tumor.
The goal of this modified response criteria is to meaningfully evaluate radiographic response and progression while simultaneously allowing therapies that may have transient effects on contrast enhancement but therapeutic benefit to be treated equally. This is particularly important in the context of platform trials, where many different therapies may be compared against a common control and there is a significant risk of over or under estimating tumor burden with a single evaluation time point. By allowing patients to stay on therapy longer, a more comprehensive and accurate assessment of therapeutic benefit can be performed on retrospective examination. A universal set of principles and guidelines, rather than treatment-specific response criteria, may allow us to fully understand the possible therapeutic benefits and potential limitations of promising new therapies for patients with GBM.
Brief History of Radiologic Response Assessment in GBM
The formation of new blood vessels, or angiogenesis, is critical for the growth of malignant brain tumors [9–11]. Malignant gliomas with high neovascularity or vascular permeability [12–14] are often associated with higher proliferation rates [15] and higher degree of aggressivity. Because of this association, imaging techniques aimed at identifying abnormal vascularity or vascular permeability, including contrast-enhanced computed tomography (CT) and magnetic resonance imaging (MRI) are commonly used for diagnosis and clinical management of brain tumors, as they have been shown to contain the most aggressive portions of the tumor [16, 17].
In 1990, Macdonald et al. [18] introduced the first radiographic response assessment specific to brain tumors by significantly improving upon the Levin criteria [19] and the WHO oncology response criteria [20]. By standardizing the definition of radiographic response using quantitative bidirectional measurements and accounting for corticosteroid use in neurological status, similar to the response evaluation criteria in solid tumors (RECIST) [21], the new “Macdonald criteria” utilized measurements of contrast enhancing tumor size combined with other clinical metrics to determine treatment response and tumor progression by stratifying response into four categories: complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The original Macdonald criteria continues to be the fundamental framework for response assessment and radiographic interpretation of treatment changes in neuro-oncology, having been used for more than 20 years.
Known Limitations for Current Response Criteria
Although contrast enhancement has been used to assess brain tumor response for more than 60 years and contrast enhancement is generally a strong surrogate of brain tumor disease, there are caveats and exceptions that have been discovered as a result of different treatment mechanisms that affect vascular permeability. For example, increased vascular permeability from cytotoxic therapies including radiotherapy and anti-neoplastic treatments have been shown to result in increased contrast enhancement in the context of therapeutic benefit, a phenomena known as “pseudoprogression.” Additionally, clinical studies examining the efficacy of new anti-angiogenic agents have noticed a substantial decrease in contrast enhancement [22–31] resulting in high response rates, ranging from 28 to 63% in bevacizumab [32–34] and 50% in cediranib [31] compared with?<?10% using other chemotherapies [35–38], which translated into prolonged PFS but no difference in OS [31, 32]. It was assumed this high response rate was due to the use of contrast enhancement as the primary tool for evaluation in the Macdonald criteria, which resulted in a “pseudoresponse”[39], where contrast enhancement is falsely reduced due to changes in vascular permeability independent of anti-tumor effect.
In addition to increased response rates, studies examining tumor relapse/progression while on anti-angiogenic agents note a tendency for growth of nonenhancing, infiltrative tumor prior to emergence of contrast enhancement [25]. Approximately 30-40% of patients are estimated to experience non-enhancing tumor progression prior to changes in contrast enhancement [40, 41]. Malignant gliomas are known to contain proportions of both neovascularized and infiltrative tumor [42, 43] and the relative proportions are thought to reflect different biological phenotypes [44–48]. In 2010, expert opinion and examination of these limitations resulted in the creation of a formal Response Assessment in Neuro-Oncology (RANO) criteria [49] to comprehensively reform the Macdonald criteria using previously documented perspectives and approaches [50–52].
Although the RANO criteria corrects for a number of insufficiencies identified in the Macdonald criteria including inclusion of the evaluation of nonenhancing tumor progression and issues associated with pseudoresponse and pseudoprogression, there remain significant limitations to the current standard RANO criteria given recent data. For example, the current RANO criteria requires use of bidirectional measurements of contrast enhancing tumor size, which have been shown to overestimate tumor volume [53] and result in higher reader discordance [8, 54–59], presumably due to differences in head tilt and accurate identification of longest and perpendicular diameter in relatively irregular tumors. Other studies have shown reasonable agreement between bidimensional and volumetric measurements [60, 61], suggesting quick bidimensional assessment of contrast enhancing tumor size may be a practical alternative to more sophisticated volumetric segmentation. Additionally, the thresholds used to define response and progression is relatively arbitrary and not optimized based on scientific data showing the best correlation with survival benefit or time to treatment failure. (Note: The efficacy of these thresholds remains to be sufficiently challenged). Also, the use of thresholds based on “percentage change” with respect to baseline tumor size are significantly biased toward small tumors where relatively low absolute changes in tumor size are interpreted as a large percentage change [61]. This is particularly an issue in newly diagnosed GBM studies, where patients with tiny tumors often progress early due to triggering of progression (PD) when “non-measurable disease”, defined as having the two largest perpendicular diameters of a contrast enhancing target lesion less than 10mm, reaches the subtle threshold of “measurable disease”. Lastly, although changes in non-enhancing disease were added to the RANO criteria in an attempt to identify non-enhancing tumor progression, particularly in the presence of anti-angiogenic therapy, retrospective evaluations in clinical trials have shown it results in PD approximately a month prior to contrast enhancing disease progression [62], does not result in significant differences in prediction of OS [62, 63], and is one of the most controversial aspects of RANO evaluation due to the subjective nature of the interpretation and high adjudication rates. Further, studies have shown that specific aspects of non-enhancing tumor progression (e.g. circumscribed vs. infiltrative T2 changes) result in dramatically different post-progression survival in GBM patients [41], suggesting evaluation of non-enhancing tumor progression using T2 and/or FLAIR may be more complex than once thought and warrant further investigation before it can be properly integrated as an early radiographic endpoint. Further, new immunotherapy agents can also cause inflammation leading to changes in T2 signal intensity that is ambiguous with regard to interpretation of changes in tumor biology.
Updated Strategies for Response Assessment in Neuro-Oncology: Modified RANO Criteria
Based on these various challenges, an update to the current response criteria is necessary in an attempt to establish a general framework for response assessment in neuro-oncology that is agnostic to the mechanism of action of the particular therapy (e.g. anti-angiogenic, anti-neoplastic, immunotherapy, etc.), each of which has its own challenges associated with interpretation of radiographic changes, and is updated based on recent scientific evidence and current clinical convention. In order to advance the RANO criteria and address these challenges we propose the following “modified” RANO criteria for use in evaluating therapeutic efficacy in patients with GBM.
Image Acquisition Requirements
In response to a need for better standardization of image acquisition in GBM clinical trials [64], a recent consensus paper was published outlining an “international brain tumor imaging protocol (BTIP)” (Table1) with recommended sequences and parameters [65]. At the core of this recommended protocol is parameter matched, pre- and post-contrast 3D (volumetric) inversion recovery gradient recalled echo (IR-GRE) images with less than 1.5-mm isotropic resolution, which allows for both bidimensional and volumetric measurements of enhancing tumor. When possible, this protocol should be employed for prospective clinical trials.
table ft1table-wrap mode=article t1
International Standardized Brain Tumor Imaging Protocol (BTIP) minimum image acquisition requirements for 1.5T and 3T MR systems
If volumetric acquisition is not employed, or if retrospective evaluations of existing trial data are performed, then slice thickness plus interslice gap should be less than 5 mm. If the sum of the slice thickness and gap exceeds 5 mm, then slightly modified definitions of measurable disease should be used (e.g. measurable disease?=?largest perpendicular diameters?>?2× slice thickness?+?gap).
Contrast Enhanced T1-Weighted Digital Subtraction Maps for Increased Lesion Conspicuity
Quantification of contrast enhancing tumor size or volume should be performed on contrast-enhanced T1-weighted digital subtraction maps (Fig. 1) in order to increase lesion conspicuity and better predict tumor burden in the presence of reduced vascular permeability as occurs during anti-angiogenic therapy [66] and/or T1 shortening from blood products or calcifications [67, 68]. Further, the American College of Radiology (ACR) recommends this approach for identification and delineation of subtly enhancing bone and soft tissue lesions [69].
Bidimensional and/or Volumetric Measurements
Similar to the current RANO criteria, two-dimensional, perpendicular measurements of contrast enhancing tumor size, excluding the resection cavity along with any cysts or areas of central macroscopic necrosis, should be used for response assessment if volumetric tools are not available. Table 2 outlines suggested volumetric conversions from two- to three-dimensional measurements for consistency in response definitions, as outlined by Chappell et al. [70].
Bidimensional to volumetric definitions [54, 70, 96] of radiographic response and progressionIt is important to note that the field remains conflicted on whether or not enhancing disease should be included in tumor size measurements, or whether it is more appropriate to monitor total enhancing lesion volume, which may include central macroscopic necrosis and any cystic components (but excluding surgically resected tissue). Scientific studies have shown that both approaches for quantifying change in tumor size as a surrogate of treatment response are valuable. Multiple studies utilizing the Macdonald and RANO criteria have shown that change in enhancing disease size using bidimensional measurements, excluding necrosis and cystic components, can be used to predict survival in a variety of therapies. A recent study from the BRAIN trial, a phase II trial of bevacizumab with or without irinotecan in recurrent GBM, confirmed that change in the volume of enhancing disease can be used to predict survival benefit [66]. However, a recent study examining growth rates in treatment naïve presurgical GBMs showed that changes in enhancing disease only may not be reliable, since changes occurring prior to any therapy often showed stable or decreasing tumor enhancing disease volume [61]. Growth rates were universally positive (i.e. growing) when total lesion volume (including central necrosis) were taken into consideration, which appears more realistic given the fast growth trajectory of these tumors during therapeutic intervention. Regardless, future studies are warranted to determine which measurement may be more clinically meaningful or reliable in predicting early response to new therapies.
Definition of Measurable Disease, Non-Measurable Disease, and Target Lesions
Measurable disease should be defined as contrast enhancing lesions with a minimum size of both perpendicular measurements greater than or equal to 10mm (Fig. 2). For example, if the largest diameter is 15 mm but the perpendicular diameter is 8 mm, this would constitute non-measurable disease. Additionally, in the event that the BTIP protocol is not used, if the slice thickness plus interslice gap is greater than 5mm, then the minimum size for both perpendicular measurements should be twice the sum of the slice thickness and interslice gap (e.g. if the slice thickness is 5mm with 1.5mm interslice gap, the minimum tumor size on both perpendicular dimensions should be 13 mm). Up to a total of five target measurable lesions should be defined and ranked from largest to smallest (Fig. 2).
Algorithm for identifying measurable and target lesionsNon-measurable disease should be defined as lesions that are too small to be measured (less than 1 cm in both perpendicular dimensions), lesions that lack contrast enhancement (non-enhancing disease), or lesions that contain a poorly defined margin that cannot be measured or segmented with confidence.
Correction for “Baseline Tumor Volume” in Newly Diagnosed and Recurrent GBM
An abundance of single center, multicenter, and phase I-III trials have confirmed that baseline contrast enhancing tumor size (volume or bidirectional measurements) is a significant prognostic factor contributing to overall survival (OS) in GBM. In newly diagnosed GBM, both extent of resection [3, 71–87] and post-surgical residual volume [83–85, 88–92] have been shown to be prognostic. Similarly, baseline pre-treatment contrast enhanced tumor size has also been shown to be prognostic for OS in recurrent GBM [53, 66, 93]. However, from a clinical trial perspective, post-surgical residual enhancing tumor volume may be a more practical measurement to obtain, as pre-surgical MRI scans are often not available or collected as part of clinical trials because patients are not enrolled until after surgery and diagnosis. Thus, care should be made to make sure baseline tumor size is a stratification factor during randomization (i.e. prospectively balanced across treatment arms) and used as a covariate in statistical models evaluating treatment efficacy.
Post-Radiation MRI Examination as the Reference for Evaluating Radiographic Response in Newly Diagnosed GBM
The current RANO criterion defines the post-surgical MRI scan as the baseline for treatment response evaluation; however, we propose using the post-radiation examination (i.e. the first scan following completion of concurrent radiation therapy and chemotherapies such as temozolomide and/or experimental therapeutics) as the baseline for response assessment because reliability of tumor assessment on the post-surgical MR scans can be problematic for a number of reasons. First, this scan is typically acquired prior to a final pathological diagnosis, thus patients are not yet enrolled in a clinical trial and therefore the imaging protocol may not be consistent with trial recommendations, leading to a mismatch between the baseline and subsequent follow-up time points. Secondly, post-operative MR scans are often contaminated with post-surgical changes including blood products and increased vascular permeability from surgical trauma. Thirdly, steroid dose can be highly variable during this time and may be poorly annotated, as patients are typically not yet enrolled in clinical trials at this point. Additionally, the timing of the post-operative MR scans can be highly variable from patient to patient, depending on the complexity of the surgery and potential intraoperative complications, and institution by institution, as many factors including availability of inpatient MR scanners can lead to different timing of the post-surgical MRI evaluation. This variability inevitably leads to differing degrees of post-surgical artifacts and fluid levels on the resulting images. Together, these factors appear to indicate the post-surgical MRI examination may not be a reliable reference scan for accurately determining radiographic changes, despite post-surgical residual enhancing volume being a significant prognostic factor as outlined above.
Perhaps the most compelling argument for using the post-radiation scan as the baseline for determining response assessment is the highly unpredictable, transient radiographic changes that often accompany the initial chemoradiation phase (i.e. external beam radiation therapy plus concurrent temozolomide) with or without experimental therapeutics. Within 1 month after completion of standard chemoradiation therapy, approximately 50% of patients will experience radiographic changes suggestive of early tumor progression in reference to the post-surgical MRI exam, of which 50% are likely to have pseudoprogression (i.e. 25% of all patients at 1 month post-chemoradiation are estimated to have pseudoprogression) [94]. This proportion of patients with both early progression and pseudoprogression decreases steadily during the subsequent standard adjuvant chemotherapy phase, which forms the basis for current RANO recommendations of excluding patients in recurrent GBM trials who progressed within 3 months after completion of chemoradiation. Many clinicians are reluctant to change therapy based on this examination due to the relatively high incidence of treatment-related radiographic changes directly after completion of concurrent chemotherapy and radiation, and instead use this scan as a new baseline in which to interpret subsequent changes in tumor size. Additionally, experimental therapeutics that significantly alter vascular permeability, including anti-angiogenic and immunotherapies, when used concurrently with radiation therapy and temozolomide often demonstrate dramatic and transient changes in contrast enhancement that quickly stabilize following completion of radiation [95]. Despite the improved lesion conspicuity on T1 subtraction maps in the settings of these therapies, these early changes between the post-surgical, pre-radiation exam and the post-radiation exam may not accurately reflect true changes in tumor burden nor predict long-term survival benefit [95].
Detailed Definitions Used for Modified Radiographic Response Assessment Criteria
Radiographic response should be determined in comparison to the tumor measurements obtained at baseline (post-radiation scan will be baseline for newly diagnosed GBM and pre-treatment scans will be the baseline for recurrent GBM) for determination of response, and the smallest tumor measurement at either pre-treatment baseline or following initiation of therapy for determining progression.
Because novel treatments are likely to result in a higher than normal incidence of treatment-related increase in contrast enhancement (“pseudoprogression”, PsP) or decrease in contrast enhancement (“pseudoresponse”, PsR), patients should continue therapy with close observation (e.g. 4-8 week intervals) if there is a suspicion of PsP or PsR. If subsequent imaging studies and/or clinical observations demonstrate that progression in fact has occurred, the date of confirmed progression should be noted as the scan at which the potential progression was first identified. Definitions for complete response, partial response, progressive disease, and stable disease should be defined as follows for all target lesions.
Complete Response (CR): Requires all of the following:
1. Disappearance of all enhancing measurable and non-measurable disease sustained for at least 4 weeks. The first scan exhibiting disappearance of all enhancing measurable and non-measurable disease is considered “preliminary CR”. If the second scan exhibits measurable enhancing disease with respect to the “preliminary CR” scan, then the response is not sustained, noted as pseudoresponse, PsR, and is now considered “preliminary PD” (note confirmed PD requires at least two sequential increases in tumor volume). If the second scan continues to exhibit disappearance of enhancing disease or emergence of non-measurable disease (less than 10mm bidimensional product), it is considered a durable CR and the patient should continue on therapy until confirmed PD is observed.
2. Patients must be off corticosteroids (or on physiologic replacement doses only).
3. Stable or improved clinical assessments (i.e. neurological examinations).
Note: Patients with non-measurable disease only at baseline cannot have CR; the best response possible is stable disease (SD).
Partial Response (PR): Requires all of the following:
1. ≥50% decrease in sum of products of perpendicular diameters or ≥65% decrease in total volume [54, 70, 96] of all measurable enhancing lesions compared with baseline, sustained for at least 4 weeks. The first scan exhibiting ≥50% decrease in sum of products of perpendicular diameters or ≥65% decrease in total volume [54, 70, 96] of all measurable enhancing lesions compared with baseline is considered “preliminary PR”. If the second scan exhibits PD with respect to the “preliminary PR” scan, then the response is not sustained, noted as pseudoresponse, PsR, and is now considered “preliminary PD” (note confirmed PD requires at least two sequential increases in tumor volume). If the second scan exhibits SD, PR, or CR, it is considered a durable PR and the patient should continue on therapy until confirmed PD is observed.
2. Steroid dose should be the same or lower compared with baseline scan.
3. Stable or improved clinical assessments.
Note: Patients with non-measurable disease only at baseline cannot have PR; the best response possible is stable disease (SD).
Progressive Disease (PD): Defined by any of the following:
1. At least two sequential scans separated by at ≥4 weeks both exhibiting ≥25% increase in sum of products of perpendicular diameters or ≥40% increase in total volume [54, 70, 96] of enhancing lesions. The first scan exhibiting ≥25% increase in sum of products of perpendicular diameters or ≥40% increase in total volume [54, 70, 96] of enhancing lesions should be compared to the smallest tumor measurement obtained either at baseline (if no decrease) or best response (on stable or increasing steroid dose) and is noted as “preliminary PD.” If the second scan at least 4 weeks later exhibits a subsequent ≥25% increase in sum of products of perpendicular diameters or ≥40% increase in total volume of enhancing lesions relative to the “preliminary PD” scan, it is considered “confirmed PD” and the patient should discontinue therapy. If the second scan at least 4 weeks later exhibits SD or PR/CR, this scan showing “preliminary PD” is noted as “pseudoprogression”, PsP, and the patient should continue on therapy until a second increase in tumor size relative to the PsP scan is observed. Note that any new measurable (>10mm x 10mm) enhancing lesions should not be immediately considered PD, but instead should be added to the sum of bidimensional products or total volume representing the entire enhancing tumor burden.
2. In the case where the baseline or best response demonstrates no measurable enhancing disease (visible or not visible), then any new measurable (>10mm x 10mm) enhancing lesions are considered PD after confirmed by a subsequent scan ≥4 weeks exhibiting ≥25% increase in sum of products of perpendicular diameters or ≥40% increase in total volume of enhancing lesions [54, 70, 96] relative to the scan first illustrating new measurable disease. The first scan exhibiting new measurable disease is noted as “preliminary PD.” If the second scan at least 4 weeks later exhibits a subsequent ≥25% increase in sum of products of perpendicular diameters or ≥40% increase in total volume [54, 70, 96] of enhancing lesions relative to the “preliminary PD” scan it is considered “confirmed PD” and the patient should discontinue therapy. If the second scan at least 4 weeks later exhibits SD, CR, PR, or becomes non-measurable, this scan showing “preliminary PD” is noted as “pseudoprogression”, PsP, and the patient should continue on therapy until a second increase in tumor size relative to the “preliminary PD”, or PsP, scan is observed. Note that any new measurable (>10mm x 10mm) enhancing lesions on the subsequent scan following the preliminary PD scan should not be immediately considered confirmed PD, but instead should be added to the sum of bidimensional products or total volume representing the entire enhancing tumor burden.
3. Clear clinical deterioration not attributable to other causes apart from tumor (e.g. seizures, medication adverse effects, therapy complications, stroke, infection) or attributable to changes in steroid dose.
4. Failure to return for evaluation as a result of death or deteriorating condition.
Stable Disease (SD): Requires all of the following:
1. Does not qualify for CR, PR, or PD as defined above. Note this also applies to patients that demonstrate PsR when the confirmation scan does not show PD or PsP when the confirmation scan does not show PR/CR.
2. In the event that corticosteroid dose was increased (for new symptoms/signs) without confirmation of disease progression on neuroimaging, and subsequent follow-up imaging shows that the steroid increase was required because of disease progression, the last scan considered to show stable disease will be the scan obtained when the corticosteroid dose was equivalent to the baseline dose.
Symptomatic Deterioration & Reporting Clinical Status
Patients with global deterioration of health status requiring discontinuation of treatment without objective evidence of disease progression at that time, and not either related to study treatment or other medical conditions, should be reported as PD due to “symptomatic deterioration.” Every effort should be made to document the objective progression even after discontinuation of treatment due to symptomatic deterioration. Neurological exam data should be provided to the independent radiologic facility as “stable, better, worse” in case report forms or from study sponsor. Clinical status should be recorded as “worse” if the neurological exam is worse, otherwise the clinical status should be set to “not worse.” In the event that necessary clinical data is not available, clinical status should be recorded as “not available” and that particular time point can only be reviewed for PD (otherwise “non-evaluable”). Neurological data must be within ±7 days of the time-point response date, otherwise the data is considered “not available”.
Steroid Use and Dose
Steroid use should be derived from the concomitant medications on the case report forms and recorded as “Yes”, “No”, or “not available”. A value of “No” should be assigned if, at the time-point, the subject is not on steroids or on physiologic replacement doses only (<1.5 mg dexamethasone or equivalent per day).
Steroid dose should be derived from the concomitant medications on the case report forms. Average steroid dose no greater than 2 mg change from baseline should be abstracted to “stable”. If outside this range the steroid dose should be abstracted to “increased” or “decreased” accordingly. Steroid data should be within ±5 days of the time-point response date, otherwise the data is considered “not available”.
Overall Objective Status
The overall objective status for an evaluation should be determined by combining the patient’s radiographic response on target lesions, new disease, neurological status, and steroid dose/usage as defined in Table
Table3
3 for patients with measurable (>10mm x 10mm) disease. Note that patients with possible PsP or pseudoresponse should be given the Objective Status of “Preliminary Progression” or “Preliminary Response”, respectively. Once PsP, pseudoresponse, or true progression/response are confirmed, the Objective Status can be changed accordingly.
Detailed Modified Radiographic Response Assessment Rubric
In order to provide both clinical guidelines for continuing therapy beyond suspected radiographic progression if the treating physician believes there may be a therapeutic benefit and to provide criteria for defining progression and early drug failure while also allowing for the possibility of PsP and PsR, a modified response rubric similar to those described recently [97] should be employed. Two different rubrics should be used depending on whether the patient is newly diagnosed or enrolled in a trial for recurrent disease.
It is important to note that the primary differences between conventional RANO and the proposed modified criteria are: (1) use of the post-radiation time point as the baseline for response evaluation in newly diagnosed GBM and (2) considering only objectively defined, measurable enhancing disease in the definition of response and progression (i.e. exclusion of qualitatively assessed T2/FLAIR changes).
Newly Diagnosed GBM (Fig. 3)
Fig. 3
Modified radiographic response assessment rubric for management of both pseudoprogression and pseudoresponse in newly diagnosed glioblastomaNewly diagnosed GBM patients will initially undergo a pre-entry MRI scan for initial diagnosis prior to entry in the study and prior to therapy. The post-operative scan [MRI(0)] is desired in order to assess residual enhancing disease volume for use as a covariate in survival analyses, as described previously. Patients will then start on standard or experimental therapy with concurrent radiation therapy (RT). The Post-RT scan [MRI(1)] will be required and used as the baseline scan for which response will be determined.1 Following the first cycles of adjuvant therapy, patients will receive additional required MRI scans [MRI(N)].
Recurrent GBM (Fig. 4)
Fig. 4
Modified radiographic response assessment rubric for recurrent glioblastomaRecurrent GBM patients will undergo a pre-entry MRI scan [MRI(0)] at the time of recurrence. At the time of study entry, two scans to confirm progression should be submitted consisting of at least one scan at the time of progression and one scan at Nadir or baseline. If the patient undergoes surgery (optional), then the post-surgical, pre-treatment MRI can be used as the baseline [MRI(1)], assuming it is obtained?<?72 hours from surgery to reduce post-operative reactive enhancement [91, 98]. (Note: If the post-operative MRI scan is used as the baseline reference, the standardized MRI protocols must be used.) If the patient does not go to surgery or if the start of treatment is?>?21 days from the start of therapy, the patient will undergo a pre-treatment MRI [MRI(1)] scan as the baseline scan for which response will be determined. Following the first cycles of therapy, patients will receive additional MRI scans [MRI(N)].
Details Common to Both Newly Diagnosed and Recurrent GBM
Preliminary Radiographic Progression
If the lesion size has increased ≥25% sum of bidirectional product or ≥40% in volume between MRI Scan 1 and N, these patients should be categorized as “preliminary radiographic progression”. If the investigator believes the patient can safely continue on therapy, then they should continue to treat and acquire a follow-up confirmatory scan [MRI(N?+?1)] at the next scan interval (8 weeks?±?4 weeks from MRI Scan (N) or no less than 4 weeks minimum duration between preliminary PD and confirmed PD scans) to verify tumor growth and progression. For patients with gross-total resection (GTR) and no measurable enhancing disease, preliminary radiographic progression is defined as a transition from no measurable disease to non-measureable (but present) disease (<10mm x 10mm) or measurable disease (>10mm x 10mm). If the investigator feels it is safe to keep the patient on, a confirmatory scan at MRI(N?+?1) should be obtained to verify tumor progression.
Confirmed Progression
If the patient has an increase ≥25% sum of bidirectional product or ≥40% in volume between MRI Scan N and N?+?1, this is “Confirmed Progression”, the patient should stop therapy and the date of radiographic progression is the date of suspected progression, MRI(N). If the patient has SD/PR/CR on MRI(N?+?1) with respect to MRI(N), PsP is confirmed and the patient should continue on therapy. Patients will then continue on therapy and receive additional follow-up MRI scans [MRI(M)]. If the lesion size has increased ≥25% sum of bidirectional product or ≥40% in volume on MRI(M) relative to the smaller of Nadir or MRI(N?+?1), then the patient has “Confirmed Progression”, the patient should stop therapy and the date of radiographic progression is the new date, MRI(M). For patients with no measurable disease at the Post-RT baseline, “Confirmed Progression” will be defined as a transition from non-measurable (but present) disease (<10mm x <10mm) on MRI(N) to measurable disease (>10mm x 10mm) on MRI(N?+?1). For patients with confirmed PsP and no measurable disease at Nadir, “Confirmed Progression” should be defined as a transition from no measurable disease to measurable disease (>10mm x 10mm). In all cases, patients with confirmed progression should stop therapy.
Preliminary & Confirmed Radiographic Response
If a measurable lesion has decreased ≥50% sum of bidirectional product or ≥65% in volume between MRI(1) and MRI(N), these patients should be categorized as “preliminary radiographic responders” and will be monitored for an additional time point and/or treatment cycle. After an additional cycle of therapy (8 weeks?±?4 weeks from MRI(N)), patients will receive a confirmatory MRI(N?+?1). If the lesion(s) have increased ≥25% sum of bidirectional product or ≥40% in volume from MRI(N) (indicating radiographic progression from MRI(N)), this is considered an “unsustained radiographic response” or “pseudoresponse”. The date of radiographic progression for these patients will be MRI(N?+?1) and the patient should stop therapy. Alternatively, if the lesion has not increased from MRI(N), this is considered a “durable radiographic response,” the patient will continue on therapy, and the date of preliminary radiographic progression is the time point of an increase ≥25% sum of bidirectional product or ≥40% in volume (from Nadir) during the remainder of the study. The investigator can then decide whether to continue safely on therapy until progression has been confirmed and at the subsequent time point stop therapy if they feel the patient cannot safely continue therapy.
Stable Disease
If the lesion size has not increased or decreased beyond the set thresholds between Scan 1 and N, the patient is considered “stable.” Such patients will continue on therapy, and the date of preliminary progression is the time point of an increase ≥25% sum of bidirectional product or ≥40% in volume (from Nadir) during the remainder of the study. Upon preliminary progression the investigator can choose to either continue therapy and confirm progression or discontinue therapy. For cases with significant neurologic decline at the time of imaging progression as determined from MRI(N), a confirmatory scan at time point MRI(N?+?1) may not be possible or necessary. For these cases, it is appropriate to define MRI(N) as the progression time point.
Conclusions
Although radiographic response assessment is imperfect and many nuances exist, changes in contrast enhancing tumor are both clinically meaningful and appropriate for evaluating efficacy of new treatments in GBM. The outlined modifications in this report are meant to both build on the strengths of the current RANO criteria while providing potential solutions for many of the common challenges.
It certainly is possible that PFS misses the mark. The trial was initially designed to remove as many patients with possible disease progression prior to enrollment as the rationale was that patients with residual and potentially BULKY tumors following surgery could end up depressing the results. There is a correlation between TREGS and residual disease. The more BULKY residual that remain, the less time the immune systems has time to build an adequate immune response. It's actually sound rationale. However, this trial as you well know has a blinded Central Image Review. On one hand blinded Central Review of all brain scans is a great idea as using an image team that is blind to patient data removes selection bias as all MRI scans are going to be reviewed using the same perimeters, which should lead to uniformed analysis. However, the issue is that there is no imaging technique in use that is able to distinguish if on the very first scan after surgery and concurrent chemoradiation that the patients who are excluded from study entry do in fact have residual or BULKY disease -- patients that might have not as most responsive to standard of care (surgery, radiation and chemo given at that point concurrently). Radiologist needs to take serial MRI pictures from different points in time to determine if in fact the patients has evidence of true early disease progression. Again, no one and I mean no one can truly tell from one set of MRI scans if the patient with the inflammatory scan is suffering from rapid progression. And add to that, patients are generally on various degrees of steroids, recovering from surgery and so some patients who may have progression of disease might not have shown at that point. Then add to that the GBM is a disease that has varying degrees of TIL. And TIL is an inflammatory response and that can sometimes look like progression of disease. It is the Mesenchymal gene group that has the highest TIL within the tumor residual bad. And so it is quite possible that in the attempt for the study design to have blinded central review to naturally exclude some of the potentially worst immune compromised patients prior to enrollment, that the same blinded central review may have no choice but to remove some of its most immune privileges Mesenchymal responsive patients. And so you have probably heard me argue that I was very concerned that by default that exclusion criteria could mean that blinded Central Review be enrolling patients who might not have the best TIL before injections. Once the designing investigators realized the potentially study design flaw to remove some vaccine pseudoprogression patients they created a separate arm of the study to randomize those patients 2:1.
And while TIL can improve for main arm patients with no residual disease, it is a concern that the study might not have as many Mesenchymal patients with high TIL in main arm enrollment, as they may have been accidentally removed for their tumors tipping the scale of what might be confused as disease progression. There are some Mesenchymal tumors that will have reacted because they are MGMT methylation status -- meaning they are chemo responsive -- and the higher TIL within the tumor caused a false positive progression sign. However, because the disease is treated with varying degrees of steroids early on, which is used to bring down surgical swelling that there is just as much of a chance that a good number of higher TIL patients made it into the main arm. Keep in mind that the patients who have no residual disease still are able to mount an immune response. However it has been reported that the patients who exhibit the higher TIL before vaccination are the patients with the most impressive survival results for immunotherapy. If those patients only received standard of care, the TIL is not the survival deciding factor, instead if is MGMT promoter status. And so net net the results of true long tail PFS survivors may in part really boil down whether all the Brad Silvers of the study (patients who have higher TIL prior to vaccination but have not progressed) were accidentally removed by Central Review due to TIL passing the "no evidence of disease" threshold of EXCLUSION clause or not. I don't think that all were removed. I don't think Brad's Post RT scan would have tipped the bucket. But I will admit being concerned about it years ago, and have mentioned it on this and other boards well before Linda Liau mentioned it in one of her speeches.
As I do believe that even without some of the Brad's enrolled in the main arm, that the study should be able to get between 3-5 months between cohorts. There is much to be said for improved cellular constitute and when combined with improved surgical procedures which is known to reduce residual bed tumor RECURRENCE , it is the T cells that are circulating other areas of the brain that have the potential to reduce the chance of distance metastatic disease and not the T-cell that make up TIL inside the tumor. And so I'm honestly not as concerned as I once was about falling short on PFS. But again yes, it's certainly possible, which is another reason why I believe they opted to make sure the study has any much PFS long tail data to bring forth to regulators.
Here is a post I wrote Flip many months ago on this EXCLUSION criteria topic. Within it, you'll find the passage from Linda Liau and more discussion on the topic above, in case you're interested.
------
Flip,
When I wrote this:
"The main arm will also have INDETERMINATE patients IMHO. I do not perceive that all the pseudos will be found at the exclusion criteria Post RT MRI scan. You and I respectively disagree there. If I am correct, then Pseudo patient who did not cross the >25% progression threshold on that exclusion scan will be randomized into the main arm. I believe the study image review document includes an algorithm for evaluating and following indeterminate lesions, if seen on Month 2 scans, for assignment of progression status. If progression is confirmed at follow-up Month 4, the progression date is assigned to earliest time that any progression was initially detected. "---RK
I was speaking predominately about RT/TMZ induced pseudoprogression. This trial randomized patients MGMT promoter status. And this trial will need to account for psuedoprogression decision tree, as not all psuedoprogression is caught at 2 weeks post chemoradiation.
Per the protocol: Time to progression, death, or censoring will follow the definitions in the FDA Guidance document” Guidance for
Industry: Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics, May 2007”, including assignment of the progression date to the earliest time that
any progression is detected. https://www.fda.gov/downloads/Drugs/Guidances/ucm071590.pdf
And per the the FDA Guidance document” Guidance for Industry: Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics, May 2007”, it states this:
Suspicious lesions. An algorithm should be provided for evaluating and following INDETERMINATE lesions for assignment of progression status at the time of analysis.
And we know that the Phase III study uses this INDETERMINATE assignment of suspicious lesions, as the Compassionate Use Arm (CUA) has an “INDETERMINATE” cohort.
incidentally, as I have pointed out before, this INDETERMINATE classification is not only from FDA guidance on biological, it is also was a 2011 RANO working group recommendation (before the Phase III resumed) for how to rate suspicious but suspected progression lesions, when treatment-related effects remain a valid possibility for the change early on. And by treatment related affect, in the CUA, and in the main arm, it could still be RT/TMZ related early psuedoprgression.
Application of Novel Response/Progression Measures for Surgically Delivered Therapies for Gliomas: Response Assessment in Neuro-Oncology (RANO) Working Group
Retrospective Assessments of Disease Progression Should Be Permitted in Clinical Trials
As discussed above, there are no validated imaging modalities that reliably distinguish treatment effects from tumor progression, and the clinical management often involves serial imaging studies to determine the persistence of the radiographic changes. In many patients, serial imaging evaluation will show that the extent of new enhancement has regressed without clinical intervention, and these patients are considered to exhibit “pseudoprogression.” 64,65 Others will show persistent progression of imaging changes and will be considered cases of true disease progression. Accordingly, an imaging-based response classification scheme must allow a retrospective categorization of a patient as having disease progression based on the results of serial radiographic imaging. When progression is suspected but treatment-related effects remain a valid possibility, the response should be called INDETERMINATE. If subsequent evaluation proves that the changes reflect true progression, the time of progression should be retrospectively corrected to the first time at which an INDETERMINATE response was noted[/color].[/quote]
https://academic.oup.com/neurosurgery/article-lookup/doi/10.1227/NEU.0b013e318223f5a7
Moving on, at the end of my statement about the CUA, I wrote this:
But understandable since we both know that the vaccine could cause inflammation, which it might have done, particularly if the EXCLUDED patients had high TIL before vaccination. We don't have this information. — RK
you wrote this:
“But experts from Seattle, including an investigator in the trial, seemed to suggest they are just not seeing psPD induced by DCVax-L, (which is probably good since there is not a lot of room for inflammation in the brain) and Dr. Bosch only talks about psPD being induced by DCVax-Direct, because it is a bit more aggressive” — FLIP
To be clear, I meant outside the main arm, in the CUA data. It is the patients with the highest TIL before vaccination that are at risk to be removed due to suspicion of possible progression (then later possible pseudoprogression). In different enrollment periods exclusions of patients from the main arm were just called different thing, but they were essential the same “exclusion perimeters”. For instance, if Brad Silver was going through the screening process during the CUA 2011-2012 period, it is quite possible, that his earliest scan, Post RT MRI, which exhibited high TIL fit the “progression” size criteria as possible “progression”. If it were to trip the scale, he would have been EXCLUDED from the main arm of the Phase trial, and enrolled into the CUA outside the study. AND that his “psuedoprogression” next scan would have been in the CUA INDETERMINATE category. However, if Brad Silver was going through the screening process during the Pseudo Arm 2012 - 2014, it is quite possible, that his earliest scan, Post RT MRI, which exhibited high TIL fit the “progression” size criteria as possible “pseudoprogression”. If it were to trip the scale, he would have been EXCLUDED from the main arm of the Phase trial, and would have been asked to come back in 2 weeks to see about whether he qualified to enroll into the Psuedoprogression arm outside the main arm study, but in the Tertiary data. Again, his high TIL was seen BEFORE DCVax-L was initiated, and got worse after vaccination.
When Linda Liau spoke about DCVax-L trial design, she spoke about a couple things that could affect the final outcome of the study. Namely:
-- post radiation MRI exclusion criteria. Removal of psuedoprogression due to Central Imaging Review not having a third scan to determine/confirm psuedoprogression verses rapid progression.
-- crossover creates an early verses late comparison. Makes looking at progression somewhat murky. That is the risk that Linda Liau spoke about as far as trial design that could change the Phase III results.
" One ISSUE was that at this POST RADIATION MRI SCAN, if the patients had progressive tumor they were actually EXCLUDED from the trial. They were put into the separate informational arm. ……. Unfortunately what we learned over the years is that determination of the early progression is difficult because of the issue of pseudo progression so there may be some patients that were EXCLUDED that probably SHOULD NOT HAVE BEEN EXCLUDED or vice versa." -- LLiau
In the “tail of two tumors”, Linda Liau also did state that those Phase I/II patients who had more TIL in the tumor before vaccination tended to do better after vaccination. TIL correlation was a causation; the patients who have more pseudoprogression due to chemoradiation early-on tend to do better in terms of both progression free and overall survival in the Phase I/II. They are the long tail of both PFS and OS KM curves. I do want to acknowledge that you realize this, and that is why the trial has a psuedoprogression arm. But I do think that you sometimes struggle to recognize that there are possibly over 25% of the CUA that has early psuedoprogressor within it.
AS for your comment, as it relates to the main arm, I happen to agree with you on the “inducing” factor.
"But experts from Seattle, including an investigator in the trial, seemed to suggest they are just not seeing psPD induced by DCVax-L, (which is probably good since there is not a lot of room for inflammation in the brain) and Dr. Bosch only talks about psPD being induced by DCVax-Direct, because it is a bit more aggressive” — FLIP
BUT, I really thought that you understood that I agree with you as it relates to the vaccine “inducing” excess psuedoprogression in the main arm. If it occurs, it will be a small percentage of patients. And like I have stated many times, if elimination of a patient occurs because of “induced” psuedoprogression, it will be due to aboscoal effect, T cell trafficking outside the field of radiation, recognize disseminating tumor cells as they wake up and begin to cause an initially unmeasurable lesion. But even that abscopal phenomena is more likely to be seen in treatments that had metastatic disease to start with. In the GBM setting, the greater the rates of GTR, the more likely the disease reoccurs as a distance mass. And so I don’t see the T-cell trafficking to created a detectable new lesion as possible way to destroy PFS results.
You know that I still think there will be patients who exhibit possibly early PsPD — from RT/TMZ mechanism of action. These remnant early PsPD (who were not excluded on the POST RT MRI) will possibly have lower TIL (then the excluded high TIL pseudos) at baseline. The vaccine should not make their lower TIL tumors WORSE upon initiation. It didn’t make the patients who were excluded with possibly higher TIL worse (Indeterminate, who may have had psPD inside the CUA) when initiated. And I have stated many times that pseudo-progression from vaccine is going to depend on the amount of RESIDUAL disease, and DIRECT is more likely to exhibit PSEUDOPROGRESSION, as patients with bigger bulky tumors. I agree with you and Bosch on it, and incidentally, you may find this interesting, we are not alone. Here is an excerpt from FDA-AACR Immuno-Oncology Drug Development Workshop on October 14, 2016:
My name is [inaudible 02:27:09] Aribas, I'm a medical oncologist from UCLA. I want to first thank the panelist for the great presentations, but I want to challenge the concept of slow progression. We heard it be talked about, slow progression and tumor flare. I think there's very little evidence that it's mechanistically based to slow the progression. I cordially disagree with my friend Axel who's on the images that you showed showing that that's an inflamed tumor. When I've seen it, I've seen true progression followed by a delayed response. To me is a delayed immune response that goes through a period of large tumor. The images I've seen from others seem to suggest to me that that's the case. Whenever we've done a biopsy we've seen that it's mostly tumor with very little infiltrate. Eventually it's infiltrated and the tumor responds. The majority of tumors who respond to these immunotherapies have a distal infiltrate with very few inflammatory cells and you don't see inflammation, you see the tumor get smaller over time.
This is a true phenomenon but it's a rare phenomenon. It's around 1 in 20 patients. That's known from the perspective testing of this. We're trying to repeat the evaluation of large studies based on 1 in 20. Whenever it's been tested side by side the [Pembro 02:28:39] phase 1 trial had the side by side testing of [reces 02:28:42] and IRRC and its published and the response rates when you evaluate both of the perspective was very similar, telling us that the effect is very small. We have the clear examples, but the overall effect is small. I'm trying to say is that the majority, the difference between evaluating immune oncology agents may be less dependent on pseudo-progression and more dependent on other factors. Pseudo- progression, I think it's a misnomer.
Okay so I answer this of course as I started the whole thing. I agree with you actually in principal that it was not meant to say that there's only lymphocytic infiltration in a tumor that can make a tumor look bigger. Of course it can be regular progression that then later is followed by an immune response that can shrink it. I agree that pseudo-progression, call it a misnomer or not, actually entails both phenomena. How they distribute, I don't really know that. We haven't biopsied enough tumors to really know. But clearly both of them exist. As it comes to the frequency now, the [inaudible 02:29:55] data that I have seen and the one that I showed actually had about 16% of patients in that about 500 patient data set ?[02:30:00 - 02:40:04] (NOTE: speaker names may be different in each section)
Patients in that about 500-patient dataset that had a phenomenon that was either a delayed response or stable disease following an initial-
(Russian accent) Stable disease, yeah ... That's what I'm saying. Progress ... Increasing size plus [inaudible 02:30:17] response was around 5 percent, 4 percent.
Yeah, the key thing, though, is that benefit, stable disease falls in the same pocket. If you can't call it stable disease and count it as benefit because there was this initial blip of tumor growth, it still takes a lot of patients away from the analysis. The way we've done this from the beginning is, we called it "immune-related disease control," which includes the stable disease patients. You're right, it's a small number of patients that actually have a true delayed response.
Actually, I want to point one thing, that if it wasn't for your and [Jet's 02:30:54] contribution to these phenomenon, we wouldn't have had a confirmation of progression which has allowed us to study this phenomenon with RECIST evaluation. Now we know that it is true. It's reproducible, but it also happens in the control arm.
I think there was a BMS trial on it. Maybe [Rensa 02:31:13] knows it, or you know it, where IRRC was tested in the ipilimumab arm and in the control arm, and it was significant number of IRRC in, I think, it was chemotherapy arm. I forgot what it was.
I say one last thing. I try to make that point during my talk. I-O has had a very fast and furious evolution to the great benefit of many patients, but it's still at the beginning. At least I think we are still fairly early in terms of understanding the immune responses and putting new agents into the clinic that will produce new patterns of clinical activity.
Having said that, what we're trying to do here is really produce tools that help us evaluate what comes next. For the PD-1 blockers, we needed those tools less so than we needed them for ippy or we need them for cancer vaccines, or we may need them for oncolytic viruses. When you actually look at the approved oncolytic virus, T-VEC, that study was actually a response-rate study.
The approval was given on the basis of a modified response endpoint that included things like treatment beyond progression and growth before shrinkage. You can do that. It was never called "IR response," but the point is the phenomenon has been recognized. You just need to use these tools to understand what's going on. "
HERE is more of that particular panel discussion, in case anyone wants to see it.
FDA-AACR Immuno-Oncology Drug Development Workshop Transcript from October 14, 2016
Oct 14, 2016 Slides: http://www.aacr.org/AdvocacyPolicy/GovernmentAffairs/Documents/I-O%20Day2%20SlideDeck-safe.pdf
Oct 14, 2016 Transcripts:
http://www.aacr.org/AdvocacyPolicy/GovernmentAffairs/Documents/FDA-AACR%20Immuno-Oncology%20Workshop%20OCT%2014%20Transcript.pdf
They are mostly concerned with psuedoprogression in the recurrent setting. They gave those mice BULKY tumors.
I agree with all UCLA's statements. But I do not have to take it out of context to mean that somehow this trial is having an issue within their newly diagnosed GBM trial. Again the use of RANO eliminates such issue. I understand that the only way to distinguish true progression from false progression is time (three scans -- a Pseudoprogression decision tree) or a biopsy. There is no way to distingush otherwise, before they figured out a way -- with this research study. Perhaps now you will finally GET why I know for a fact that a pseudoprogression decision tree for this trial exists. Duh. It has too otherwise by Jan 2016 when all 331 passed their Month 2 scan the 248 primary event would have hit. But we know for fact that it had not, as in Oct 2016 they updated their clinical registry expectations for it to occur in November 2016.
Ps I haven't kept up with the all the board postings for at least a month. Life has been very busy for me. That said, I am glad to hear you found and posted a positive study. Great job in trying to suggest it means something it does not. Again their purpose is to use it to see what's going on without having to do a biopsy. It's great news. Hopefully they can replicate it in a larger setting. Time will tell
Some would like to believe that that line Exwannabe pointed out somehow means that the current Phase III has been having issues with Month 2 scans, and that a psuedoprogresdion tree does not exist -- meaning suspicion of Pseudoprogression will automatically be ruled progression. No where in this study did it state that. I have long argued that there is no way to distinguish true progression from false progression without at least three scans. One of the reason RANO criteria was established and incorporated into Neurocology studies was to address Pseudoprogression; the other was to address pseudoresponse. While again, it is true it is impossible to distinguish whether what shows under an MRI is progression or pseudoprogression at first, early changes that occur due to the introduction of therapy can be monitored. That is why at least three scans are needed. A retrospective review is needed early only. Tumor volume should stabilize over time if it is not early progression. Whereas, if it is true progression, resulting from non-responsiveness to a therapy, then tumor size should worsen. Rapid progression does not stabilize if a therapy is not working. And allowing a first scan in a study to be retrospectively reviewed so that true progression separates from false progression (TIL) is one of the many intents of RANO.
Incorporating a RANO 3 month window to not have anything inside the field of radiation ruled as progression has been helpful to clinical trials . But it is not enough, as pseudoprogression doesn't only show up within a 3 month window. There is no approved imaging way to differentiate true progression from false progression with an MRI at that first scan or even later. None. To date only a biopsy can officially confirm this. And because RANO is used, pseudoprogression verses rapid progression, is not an issue on MONTH TWO scan. Time is used to differentiate the two. Again only a biopsy can do this, until now -- possibly. That is the point of this imaging study -- finding an approach that can distinguish the two without having to wait for time to pass or for a surgical biopsy to be done. This research, if replicated in a larger study; if proven in a closed testing setting --and ultimately accepted by the neurology community, as an added imaging modality to use along with an MRI -- it will go a long way in establishing if an immunotherapy which can cause a positive inflammation immune cell infiltrate response is present or not. No more waiting for RANO time to pass to see if the diseases progresses or not; and no more having to do biopsy to tell what is going on within the brain.
I hope that clears up the context for Exwannabe.
You are adding context to the research paper you did not purchase to read. It tested in RECURRENCE GBM where inflamation is most prevalent. But, that said, I will humor your question, with a answer that you know full well what the answer will be, as it remains consistent with all my posts/viewpoint on the subject:
In the presence of large bulky tumors, such as recurrent GBM, where they tested in this research report, then yes. When there is low or no tumor volume to start, RANO criteria, 3 month window where disease can not be confirmed progression until a next scan is done, then no.
One more:
Keep in mind that Checkpoint Inhibitor can not alter the TIL levels. The level of TIL does not go up with CI. It only has the cells that pre-existed in the tumor to restore response which is said to be low in brain cancers due to the immune privileged environment. It is likely why CI is not having much success as a monotherapy in GBM. It needs an agent like DCVax-L to be more effective in GBM disease, as DCVax-L has been shown that it increases TIL. In other diseases the TIL is higher. That's not the case in GBM. Whereas DCVax-L simply needs less immune suppression to be effective. There are other agents that help with lowering T regulators cells and modulating the immune macro environment, for instance chemotherapy drugs and Bevacizumab. Those recurrent GBM mouse studies do not include those agents. But as single agents, if it is known that a higher TIL that is blocked (DCVax-L) verses a lower TIL which is unblocked (CI) the higher TIL blocked (DCVax-L) still induced a higher immune response and that means adding other salvage therapy agents can help; and the study findings of adding CI may then help to improve the immune response rate in rGBM.
Off a few posts of mine earlier this year on the subject:
I personally believe they are seeing a synergistic combination with Bevacizumab (Avastin) after recurrence. Second surgery is not needed to use Avastin. Note that Avastin was only approved for recurrent GBM in May 2009 and their studies started before then. Hypoxia significantly inhibits the infiltration of immune cells into tumor tissue. Bevacizumab is believed to normalize tumor vasculature and decrease tumor hypoxic area. Mesenchymal patients have varying degrees of Tumor infiltrating lymphocytes (TIL) before therapy. The vaccine is said to increase TIL afterwards. The combinations of these two therapies could be enhance immune response infiltrated into tumor. They already have historical data on how long patients survive with salvage therapies after progression.
Improving the hypoxic microenvironment may improve the therapeutic benefits of DCVax-L.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5045963/
Immune modulating effects of Bevacizumab.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5045963/#!po=74.1071
Improving the hypoxic microenvironment may improve the therapeutic benefits of DCVax-L. I suspect it did in the long tail patients.
Next, second surgeries rate may also be high in this trial as this trial removed rapid progression patients. Patients who experience longer progression free survival tend to be higher candidates for second surgeries. While surgery at recurrence state has only been proven to add a short amount of life for late recurrence patient, when the combined with DCVax-L the treatment approach may mean that crossover patients are receiving a vaccine at while they are at a lower T-regulator cell immune suppression state, which should help with efficacy (scientific abstract is that the combination adds at least 5 months to overall survival). We have to wait for the study outcome comes to see if the time from progression to death is vastly different from those historical norms and from their own prior studies.
Prins and Liau both spoke about the TIL biomarker. That TIL biomarker can be used to determine which patients can benefit in advance of treatment. The Phase I/II patients who have TCR overlap, regardless of whether patients had TIL, deriving survival benefit. But those who have both TIL followed by TCR overlap represent the vaccine's best responders:
"TCR Sequencing Can Identify and Track Glioma-Infiltrating T Cells after DC Vaccination."
Hsu MS1, Sedighim S2, Wang T3, Antonios JP4, Everson RG2, Tucker AM2, Du L5, Emerson R6, Yusko E6, Sanders C6, Robins HS7, Yong WH8, Davidson TB9, Li G10, Liau LM11, Prins RM12.
Abstract: Although immunotherapeutic strategies are emerging as adjunctive treatments for cancer, sensitive methods of monitoring the immune response after treatment remain to be established. We used a novel next-generation sequencing approach to determine whether quantitative assessments of tumor-infiltrating lymphocyte (TIL) content and the degree of overlap of T-cell receptor (TCR) sequences in brain tumors and peripheral blood were predictors of immune response and overall survival in glioblastoma patients treated with autologous tumor lysate-pulsed dendritic cell immunotherapy. A statistically significant correlation was found between a higher estimated TIL content and increased time to progression and overall survival. In addition, we were able to assess the proportion of shared TCR sequences between tumor and peripheral blood at time points before and after therapy, and found the level of TCR overlap to correlate with survival outcomes. Higher degrees of overlap, or the development of an increased overlap following immunotherapy, was correlated with improved clinical outcome, and may provide insights into the successful, antigen-specific immune response. Cancer Immunol Res; 4(5); 412-8. ©2016 AACR.
Neuro Oncol. 2016 Mar;18(3):368-78. doi: 10.1093/neuonc/nov153. Epub 2015 Sep 1."
The prior Phase I/II used a DC constitute that was similar to DCVax-L but not exactly the same. It was manufactured by UCLA in their lab and not privy to any batch closed air approved methods or DC precursor improvements from NW Bio proprietary patents manufactured by Cognate. When looking at older DCVax-L data --and beyond considering those cellular differences we may see in this trial over prior others--note that not all patients received Avastin upon progression event, as again it was only approved for rGBM in May 2009, and it started well before then.
The prior Phase I/II did not remove Baseline early rapid progressor or psuedoprogressors, which this Phase III did do. It is also important to note that a progression event does not necessarily mean treatment failure for Temodar chemotherapy (TMZ) or DCVax-L in the Phase I/II or the Phase III. In the case of TMZ, it was an approved salvage agent in the prior Phase I/II. However, the same can not be said for DCVax-L. Patients who were deemed to have a progression while on DCVax-L had to stop injections. This Phase III will be the first to allow patients to continue DCVax-L treatment upon progression and that can improve the OS. Again, this Phase III trial injection restarts upon progression event, whereas the prior study did not. Crossover is open -- meaning Dr and patient will know patient is officially on vaccine -- but both Dr. and Patient will remain blinded as to what Cohort the patient was in prior to the patient's progression event. Patients in the treatment cohort are eligible to restart their schedule provided vaccine supply exists and that may mean some will receive vaccine both "early" and "late" until the vaccine supply is exhausted.
New abstract, for UCLA purchase article released 9-5-17 using image testing of DCVax-L with or without Keytruda. It's a continuation of their 2016 released research. :)
Detection of immune responses after immunotherapy in glioblastoma using PET and MRI.
Abstract
Contrast-enhanced MRI is typically used to follow treatment response and progression in patients with glioblastoma (GBM). However, differentiating tumor progression from pseudoprogression remains a clinical dilemma largely unmitigated by current advances in imaging techniques. Noninvasive imaging techniques capable of distinguishing these two conditions could play an important role in the clinical management of patients with GBM and other brain malignancies. We hypothesized that PET probes for deoxycytidine kinase (dCK) could be used to differentiate immune inflammatory responses from other sources of contrast-enhancement on MRI. Orthotopic malignant gliomas were established in syngeneic immunocompetent mice and then treated with dendritic cell (DC) vaccination and/or PD-1 mAb blockade. Mice were then imaged with [(18)F]-FAC PET/CT and MRI with i.v. contrast. The ratio of contrast enhancement on MRI to normalized PET probe uptake, which we term the immunotherapeutic response index, delineated specific regions of immune inflammatory activity. On postmortem examination, FACS-based enumeration of intracranial tumor-infiltrating lymphocytes directly correlated with quantitative [(18)F]-FAC PET probe uptake. Three patients with GBM undergoing treatment with tumor lysate-pulsed DC vaccination and PD-1 mAb blockade were also imaged before and after therapy using MRI and a clinical PET probe for dCK. Unlike in mice, [(18)F]-FAC is rapidly catabolized in humans; thus, we used another dCK PET probe, [(18)F]-clofarabine ([(18)F]-CFA), that may be more clinically relevant. Enhanced [(18)F]-CFA PET probe accumulation was identified in tumor and secondary lymphoid organs after immunotherapy. Our findings identify a noninvasive modality capable of imaging the host antitumor immune response against intracranial tumors.
https://www.ncbi.nlm.nih.gov/m/pubmed/28874539/
Again, you will need to purchase the article, to see read the paper. Interestingly, they tested on 3 recurrent patients in an on-going imaging study, and the DCVax-L combination with Bevacizmab in one patient produced a complete objective response. That reminds me very much of the current Ovarian DCVax-L survivor. Oh, and tumor volume remaining stable -- in the presence of an increase in immune cells in recurrent GBM -- is good news too. Enjoy :)
"While [18F]-FAC enables PET imaging of dCK activity in mice (17–19, 21, 24), the utility of this probe in humans is limited by its rapid catabolism mediated by cytidine deaminase. To overcome this problem, our group recently developed purine analog probes that are also selectively phosphorylated by dCK for use in human patients. [18F]-CFA PET imaging of dCK activity is an important candidate purine PET probe for human use (22). In an ongoing imaging study, we performed [18F]-CFA PET imaging on patients with recurrent GBM before and after two immunotherapeutic treatments consisting of autologous tumor lysate-pulsed DC vaccination (DCVax-L; Northwest Biotherapeutics) with or without PD-1 mAb blockade (pembrolizumab; Merck).
The case studies of the first three patients are outlined below.
Patient A received tumor lysate-pulsed DCVax before imaging and then PD-1 antibody blockade during the interval between the first and second [18F]-CFA PET scans. The posttreatment [18F]-CFA PET scan demonstrated elevated uptake in several peripheral lymph nodes and in the tumor (Fig. 5A) compared with the first scan done 3 wk earlier, in agreement with the results of the murine studies. Advanced MRI revealed an in- crease in the tumor’s subtraction map (T1+C – T1). MRI with apparent diffusion coefficient (ADC) maps and cerebral blood volume (CBV) perfusion-weighted MRI suggested an almost 300% increase in immune cells in the tumor microenvironment, with the tumor volume remaining fairly constant (Fig. 6A).
Patient B received DCVax during the interval between the two [18F]-CFA PET scans. Again, lymph node [18F]-CFA tracer up- take was observed on the posttreatment scan (Fig. 5B). In contrast to patient A, patient B’s advanced MRI showed a substantial decrease in both [18F]-CFA PET tracer uptake and the tumor’s contrast-enhancing area encompassed by edema and normal/ immune cells (Fig. 6A). However, just before scanning, this patient had started treatment with bevacizumab (Avastin; Genentech), which has been shown to decrease peritumoral edema and alter contrast enhancement patterns, perfusion, and diffusion characteristics on advanced MRI (27–29).
Patient C (not pictured) had a recurrent frontal tumor that was treated with bevacizumab and DC vaccination, which resulted in a complete objective response. Because of this, there was no evidence of disease to quantify by imaging at the time points studied, and thus PET imaging results are not presented here."
AND another section of the paper refers to the prior research article stressed about the combination, that DCVax-L is able to promote TIL, but combing with PD1 can improve it. While the vaccine appears capable of inducing T cell infiltration into tumors, the effectiveness of active vaccination in progressively growing tumors (AKA: RECURRENCE) was reported in 2016 article as less profound. While 3 patients may not be many, recurrence GBM survival long term is virtually impossible, and now seems that combining with PD-1 and Bevacizmab in this latest article may become significantly profound in time. Hence, why some of us see longer uncensored follow-up as something the company might be interested in pursuing -- after 248 progression events -- before data locking and analyzing the Phase III results. Enjoy. :)
"Over the past decade, experimental immunotherapeutic trials have been hindered by the lack of reliable and systematic measures of immune responses within central nervous system (CNS) tumors. In our preclinical models of glioma, the activated tumor-infiltrating T lymphocyte response is responsible for extended survival and therapeutic benefit (1, 23). DC vaccination alone is able to promote an infiltrating T lymphocytic response, but the addition of PD-1 mAb conveys a significant survival benefit only when combined with DC vaccination (1, 23). We describe an imaging-based modality capable of monitoring the penetrance of the lymphocytic response within preclinical CNS tumors, which provides a framework for continuing evaluation of whether similar methodologies can quantify antitumor immune responses in patients with malignant brain tumors treated with immunotherapy. Advanced MRI, including diffusion and perfusion MRI, is not sensitive to specific cell type (immune vs. tumor vs. normal brain), but can demonstrate whether the localized cell density (diffusion) or vascularity (perfusion) has been altered. With the addition of [18F]-CFA PET, we can enhance imaging specificity of the specific type of cells present. Thus, a combination of advanced MRI and PET may be useful for differentiating tumor progression from immune cell infiltration.
In conclusion, by combining dCK-based PET and MRI modalities, we are able to noninvasively localize and quantify immune responses induced by immunotherapy. The percentage of tumor volume containing an activated lymphocytic infiltrate is positively correlated with survival in animals. ITRI not only allows us to quantify the immune response in CNS tumors using noninvasive means, but also serves to standardize how we quantify this immunity, to potentially serve as a predictive biomarker of survival and antitumor immune response in a reproducible fashion among patient populations. Additional studies in patients are needed to explore the clinical significance of the changes observed with these imaging modalities."
Thanks. Do you think it's possible that if they knew by Sept 1 that 233 events have been reached, they could have applied for a late-breaking abstract release spot?
Hi Evaluate,
I continue to believe they are holding out to data lock with a specific date in mind and then will end the study. In November all 331 patients will have crossed their 24 month injection -- assuming no event.
"Treatments will be given at days 0, 10, 20, and at months 2, 4, 8, 12, 18, 24 and 30." -- protocol
That 9th injection to me seems like the target to head for. And if 248 progression events hit by Nov 2016, that month should also represent the timing of the 12 month crossover injection for most patients (unless some qualified for second surgeries). In terms of getting vaccine approved beyond newly diagnosed it would be good if the results (before censoring) included at least a year of injections. November data lock (December if they need to wait 30 days after last injection to monitor AEs) seems right. It's also a nice target to cross for tertiary endpoint analysis. Once they 'officially' data lock it shouldn't be much time after then to do the analysis. My hope is that they have the results done in time for a late-breaking abstract at SNO Annual Conference.
Thanks.
The protocol sections that I copy and pasted for you are very clear. They do not collect at crossover, though I do get your interest. You shouldn't be alarmed though, as it does not mean they don't loads of information. It does not mean they don't have blood samples of patients as they go further and further out from their adjuvant TMZ cycles. Think it through. Not all patients progress in the first 3 years of their disease.
There is a vast amount a data in the immune samples gathered. This trial protocol is set to run 3 years. And so they have 3 years of immune monitoring data on patients who have NOT progressed, regardless if they are in the treatment arm or not. Odds are is that if this is an effective treatment there will be patients who are alive, who have not had a progression event but are done with the study protocol and are simple being tracked with follow-up calls for on-going survival. It would be those patients who will likely have the highest responses to DCVax-L. And the patients who have progressed who may have been in the vaccine cohort, will probably have responses too but to a lesser degree. What they are not doing is gathering immune samples and analyzing in a Petri dish and monitoring for a response to DCVax-L at crossover or beyond. Most of what they do with regards to immune samples will only be used for exploratory unless results are perceived to correlate with clinical efficacy.
"9.2. IMMUNOLOGICAL STUDIES
The immunological studies that are intended for this clinical trial are initially exploratory in nature. Future expansion of these studies to determine immunological
correlates for clinical efficacy may be undertaken based on initial results.
For all patients, until they crossover, blood will be obtained and MNC harvested and stored for future analysis. The following tests for monitoring the immune response to
DCVax-L may be used for this study. Refer to Appendix A for the timing of these blood draws.
• Mixed lymphocyte-DCVax-L culture
• Cytokine release (e.g. through Elispot)
• Immunohistochemical staining of resected brain tissue upon recurrence may be undertaken in a separate study
Feasibility
Some of the above assays (mixed lymphocyte-tumor cell cultures, cytokine release) are dependent on the availability of DCVax-L material in excess of what is required
for treatment. Excess material will not be available for all patients and it may not be available for any patients. Assays will therefore be limited based on the amount of available material."-- protocol
Besides, they know full well what to expect from the samples they gather until progression should they go to analyze all 331 patients, and 32 psuedoprogression arm patients. :)
Cytokine responsiveness of CD8(+) T cells is a reproducible biomarker for the clinical efficacy of dendritic cell vaccination in glioblastoma patients. Everson RG1, Jin RM1, Wang X2, Safaee M1, Scharnweber R1, Lisiero DN3, Soto H1, Liau LM4, Prins RM5.
Author information
BACKGROUND:
Immunotherapeutic approaches, such as dendritic cell (DC) vaccination, have emerged as promising strategies in the treatment of glioblastoma. Despite their promise, however, the absence of objective biomarkers and/or immunological monitoring techniques to assess the clinical efficacy of immunotherapy still remains a primary limitation. To address this, we sought to identify a functional biomarker for anti-tumor immune responsiveness associated with extended survival in glioblastoma patients undergoing DC vaccination.
METHODS:
28 patients were enrolled and treated in two different Phase 1 DC vaccination clinical trials at UCLA. To assess the anti-tumor immune response elicited by therapy, we studied the functional responsiveness of pre- and post-vaccination peripheral blood lymphocytes (PBLs) to the immunostimulatory cytokines interferon-gamma (IFN-?) and interleukin-2 (IL-2) in 21 of these patients for whom we had adequate material. Immune responsiveness was quantified by measuring downstream phosphorylation events of the transcription factors, STAT-1 and STAT-5, via phospho-specific flow cytometry.
RESULTS:
DC vaccination induced a significant decrease in the half-maximal concentration (EC-50) of IL-2 required to upregulate pSTAT-5 specifically in CD3(+)CD8(+) T lymphocytes (p?<?0.045). Extended survival was also associated with an increased per cell phosphorylation of STAT-5 in cytotoxic T-cells following IL-2 stimulation when the median post/pre pSTAT-5 ratio was used to dichotomize the patients (p?=?0.0015, log-rank survival; hazard ratio?=?0.1834, p?=?0.018). Patients whose survival was longer than two years had a significantly greater pSTAT-5 ratio (p?=?0.015), but, contrary to our expectations, a significantly lower pSTAT-1 ratio (p?=?0.038).
CONCLUSIONS:
Our results suggest that monitoring the pSTAT signaling changes in PBL may provide a functional immune monitoring measure predictive of clinical efficacy in DC-vaccinated patients.
That's true. They were not treating this study as a recurrent GBM trial. They will not be analyzing immune responses of either vaccine or placebo crossover patients. This is only a newly diagnosed 'official' study.
Remember immune responses to DCVax-L will be analyzed as a tertiary endpoint, which is meant support the primary and secondary endpoint which is hoping to prove introduction of the vaccine at baseline has a positive effect on vaccine cohorts PFS and OS. Patients who initiate DCVax-L treatment (vaccine cohort) verses immune responses of patients who do not (placebo cohort) at their newly diagnosed GBM state will be all that they are interested in terms of statistical analysis. Gathering immune data will stop at progression free survival events.
"Immune Monitoring
• Immune monitoring samples will not be drawn for patients enrolled in the crossover study arm" -- protocol
But they will still want to see then see if those difference, captured prior to crossover, translate to an improvement in OS between cohorts.
"Other (tertiary) endpoints include time to progression (TTP), decline in Karnofsky Performance Status (KPS), landmark analyses of survival, and immune responses to DCVax-L, as well as PFS and OS in patients with either no progression or pseudoprogression at baseline.
The underlying hypothesis is that DCVax-L, i.e. DC generated from the peripheral blood of a patient with GBM and loaded (combined) with tumor lysate antigen, will safely (1) delay time to disease progression, (2) increase survival and (3) induce anti-tumor immunity when administered as a series of intradermal (i.d.) injections in newly diagnosed patients." -- protocol
And:
"3.1. OBJECTIVES
3.1.1. Primary objective
The primary objective of this study is to compare progression free survival (PFS) between patients in the DCVax-L cohort and patients in the placebo cohort in patients with no evidence of disease progression at baseline.
3.1.2. Secondary objective
The secondary objective of this study is to compare Overall Survival (OS) between patients in the DCVax-L group and patients in the placebo group in patients with no evidence of disease progression at baseline;
3.1.3. Tertiary objectives
• To compare PFS between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized);
• To compare OS between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized);
• To compare immune responses between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized);
• To compare decline in physical functioning, measured by KPS evaluation, between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized);
• To compare time to progression of disease (TTP; includes as events all radiographic evidence of disease progression; all deaths are censored) between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized);
• To compare survival at 6, 12, 18, 24, 36 and 48 months between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized)." -- protocol
Yes. Study would end, so study injection treatments would stop.
But, they could apply to just treat them more cost effectively under expanded Access. If they don't end the study early, then it's still another year plus for the last patients to adhere to the treatment protocol visits, which is more costly on us.
In Germany, they may be able to apply for hospital exemption reimbursement. But they can't have patients in the US pay for their own vaccine without approval.
Thanks.
It's okay if you disagree. You know I don't mind debating the merits of what I see as a possible.
Now, I disagree that it would be a legal and regulatory disaster to end early. Once the data is censored, results that are submitted to the FDA can't get better. The long shareholder assumption is that most of the surviving patients are within the vaccine cohort. At this point they have had to have most of their vaccine supply (not all survivors will be from the last quarter enrollment period). And so as far as supply is concerned it is near exhausted, from the long perspective, for the vaccine patient. It's also possible true that most placebo crossover patient have had their fair share of their own vaccine too. It's my belief that the company has a Month in mind to data lock. And so, let's say that the 248 progression event, likely occurred in November 2016 timeframe, longs also assume that most placebo contrIbuted to these events. And so that means placebo crossover supply is well into booster stage for majority of placebo patients at this point. If the study were stopped, and injections stopped, I don't see the regulatory tragedy that you do. Besides it is quite possible that a solution can be found (i.e. Expanded Access).
The vaccine has a 3 year manufacturing self life. The only way they can get it over the 3 years is if the study is NOT stopped early OR if regulators somehow allow all patients to crossover before the manufacturing expiration date (i.e. Expanded Access). But they don't have the millions needed to fund the study another year at this stage. Certainly they can get the money by more dilutions. But that's at a significant cost to the future viability of them as a company. They are trading at fumes at this stage. And treating the last possible 90 plus patients for the full 3 years according to protocol of follow-up MRI visits every 2 months, simple so a patient can get their booster schedule injection would not come cheap. It could destroy the company. If they do feel compelled to treat patients, which one would hope they would, they can get compassionate use allowance to allow end of study patients access to their vaccine. The company can work with regulators as they have in the past for failed screen patients, and regulators can allow Compassionate Use under the expanded access and that can be done on the same injection schedule, though just not under the same costly protocol adherence. Spending the millions to fund the study is just not necessary from my point of view.
Yes, that's true too. We are close. :)
Exwannabe, Yes, they can go that route. The point is that they don't have the funds to treat the patients post stopping the trial. And so I'm suggesting that they may decide to conduct EOT visit on all patients before locking the data; and so censoring actually coincides with the conclusion of the study (albeit technically a year early), and analysis would occur afterwards. They don't follow the norm, and so it's quite rationale to end the study early as diluting us further and further to fund and support a Phase III study that is on the verge of censoring data in order unblind and conduct the primary and secondary endpoint analysis may not make economic sense. While we both know the study can continue while it is unblinded, the bottom line is millions more in funds are needed to complete the 3 years of the study and approvals do not happen overnight. You know their quarterly run rate is high. Ending the study at final statistical analysis would reduce it.
Ps, I already mentioned in my first post that tracking survival after EOT would be on-going. It's clear on the consent form. But let's be real, follow-up calls should not cost millions per quarter. :)
If your numbers are accurate, and they data lock in October, 322 out of 331 patients should still be enough tertiary data at 24 months. And while I haven't discussed it much, Sept/Oct will also represent some 36 months of tertiary data on the August 2014 WBC change. Truly think we are at the final stretch and they will do the best they can to ensure they capture as much data as possible, and they have a specific Month in mind to end things, without making us wait for the trial to complete the full 3 years for last enrolled patient. Whatever they do, I'm just trying to exhibit patience. :)
They can end the trial any time they want to. It is my view that they won't want to censor the Phase III data until they have at least 2 years of tertiary data on majority if not all patients. It is not necessary that they lock a specific Month as the study has at least 18 months post enrollment on all patients - unless they are concerned results may not separate early on -- (meaning a lagging OS) due to both crossover possible confoundment and under-enrollment (some missing early events that would be there had they enrolled the last 17 patients (a third of which would have been placebo --some of which would not have crossed over)) -- and they were interested in obtaining the most tertiary data to support their primary endpoint.
In order to get 2 years of tertiary data on all 331 patients, the data lock would need be conducted two years after the date of the last enrolled patients, which we were told was in November 2015. But that said, the rest of the 330 patients (majority) could have been enrolled by October 2015 or earlier. (Incidentally, mostly all of the trial enrollment was earlier as by the August 2015 halt we do know now that 305 patients were already enrolled). So to answer your question, it truly up to them on when to end the trial, and assuming they do want cross the 2 years of tertiary data on all 331 patients, then yes, that means locking sometime in November. Informing us that the study has ended would be something that is material. And so we would be told likely before the final analysis is done that it has been requested and anticipated.
Tertiary endpoint I referred to:
"• To compare survival at 6, 12, 18, 24, 36 and 48 months between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized)." -- protocol
Remember, both progression free and survival times will be calculated from time of randomization not from newly diagnosed surgery. And so if they lock the data now, then they will only have 18 month threshold of tertiary data on all 331 patients. And they would have 24 months of tertiary data on some 300+ patients. I don't know about you, but given we are so close to November, I'd rather they not censor data until they have 24 months on all patients (or October if that simply means majority (I.e, 325-330 of 331 patients)).
Note, if they stop the study before all patients undergo 3 years of treatment, it will mostly affect the crossover patients. It may mean less overall survival confounding. Sadly that will come at a cost, as ending the study early to coincide with statistical analysis data lock, will mean crossover progression patients who need their vaccine will not be getting it. The only positive to that is that regulators would get to see if ending treatment results in the death of first progression placebo cohort crossover patients and if immune memory kicked in for those who may have had a sufficient number of injections.
GLTA.
The Company does not have the money to continue the trial another year for those last in after locking and censoring the trial data. The Company has the right to end the trial any time they want to, for any reason. And so, it is quite possible that they are tying data lock with the conclusion of study end and that is why we are not being told what is going on.
If they learned in August that the trial reached the 233 in July, as they anticipated, then I would imagine that that is the point the trial clinical site investigators would be told, by their CROs, to get the last MRI on each remaining patient in the study, and conduct an end of treatment visit. The trial is meant to be 3 years from enrollment for patients who are alive that have not crossed over. The Phase III consent forms makes it clear that each patient must go through an EOT visit at the end of the 3 years (36 months in).
http://neurosurgery.ucla.edu/Workfiles/Site-Neurosurgery/Brain_Tumor_Program/DCVax%20Consent.pdf
"18. STOPPING THE STUDY
The sponsor may decide to stop the study at any point, for any reason." -- protocol
I truly think it is not reasonable to expect a virtually bankrupt company to continue to fund a trial that is about to be censored upon data lock for another year plus. Long tail follow-up is still on-going and can still be done on their study if it is ended early. But if they do decide to end early, further injection treatment schedule would not be followed. And so it's quite reasonable to believe they may be extending their right to end the study a year early for last in patients upon considering both the primary and second Endpoints are crossed.
If NW Bio decided in August to data lock survival, it still may mean the trial can run into November as they attempt to line up data lock with the study ending after 2 years after the last enrolled patients. End of study (EOT) visits require that each patient undergoes an MRI, which are conducted every 2 months, plus or minus 14 days at this point. Ensuring each patient undergoes their final MRI is a process is a multi-month. And then remember, some of those visits will coincide with scheduled injection treatments. The last DCVax-L injections may still take place. And so then one more additional month of 30 days follow-up would be needed to record tracking information on AE.
From the protocol:
"8.5. END OF TREATMENT (EOT) VISIT SCHEDULE AND PROCEDURES (ALL PATIENTS – RANDOMIZED AND CROSSOVER):
• EOT Visits for all patients who discontinue from the study should occur at least 7 days, but ≤ 30 days, after the last immunization and prior to beginning other treatment. Procedures to be performed during the EOT Visit include:
• Physical Exam
• Neurological Exam
• Vital Signs
• KPS
• MRI of brain
• CBC and Differential
• Blood Chemistry - Comprehensive metabolic panel, including electrolyte balance, and hepatic and renal functions
• Serum markers of Autoimmune disease (anti-DNA)
• Urinalysis
• AE Assessment
• Concomitant Medication" -- protocol
Also note this from the protocol, 30 days after the patient exits from participation in the trial or last study treatment, (whichever comes later) tracking information on AE is still necessary:
"Study-Specific Adverse Event Recording
• The time period for AE assessment starts at pre-leukapheresis through 30 days after the patient exits from participation in the trial (off study date) or last study treatment, (whichever comes later). It is recognized that the date the patient is determined to be “off-study” may not coincide with a visit date. " -- protocol
And:
"13.3. RECORDING AND FOLLOW-UP OF ADVERSE AND SERIOUS ADVERSE EVENTS
All AEs that occur (or that worsen from pre-leukapheresis status) from preleukapheresis through 30 days after the patient exits from participation in the trial or last study treatment, (whichever comes later), will be recorded. Adverse events that occur after enrollment but before the first immunization will be considered not related to study drug. Duration, severity, and outcome for each AE will be recorded on the CRF Adverse Event Form, and treatment administered for the event will be recorded on the Concomitant Medications pages; this information must also be recorded in the source documentation. Adverse events will be followed until resolution, until no further improvement is expected, or until the patient is lost to follow-up, whichever comes first." -- protocol
If they were to decide to treat overall survival secondary endpoint crossing as news that would start the process for conclusion of the study (verses going the full 3 years - November 2018), that would mean treatments would stop upon final statistical analysis. In short that would leave Sept and October timeframe to conduct all final visits and then November to undergo the 30 day AE follow-up. Again, considering they have the option to end the study at any time over any reason, it may be necessary that they do so because of funds; and therefore, they are tying up all loose ends, and plan to conduct the final analysis once all patients undergo an end of treatment visit (albeit earlier than the 3 years for the last enrolled patients) and all necessary follow up is complete. And we won't hear about it until it is done.
You may ask why I believe this might be the action they are taking? Well, because the clinical sites are in the process of gathering information for a review paper that can only be released once the study concludes. And it is also quite possibly that the Company gave the okay to begin the process of ending the study, with a specific trial end date in mind, a couple of months ago.
From the protocol on publications:
"The Study is designed as a multicenter study and not powered for analysis and presentation of Study results by individual Study sites. It is anticipated that the final results of this study will be submitted to a peer-reviewed scientific journal. Authorship on such a paper will be acknowledged with customary scientific practice. As such, without the expressed permission of the Sponsor, only clinical Study data relating the Study as a whole will be published. If permission is granted by Sponsor for publication of ancillary data from individual sites, prior to submission for publication of any manuscript or presentation of any poster, presentation, abstract or other written or oral material that describes the results of Study, Institution and/or PI shall provide Sponsor at least 60 days (or as otherwise specified in the sites executed Clinical Trial Agreement) to review any such materials. Such materials shall not divulge any of Sponsor’s Confidential Information, and Institution and/or PI shall promptly remove any Confidential Information as requested by Sponsor. If requested by Sponsor, the PI and Institution shall delay the submission of any publication or presentation up to 60 days from the date of Sponsor’s request for such a delay. In addition, Sponsor has the right to require that any publication or presentation concerning the Study will acknowledge Sponsor’s support." -- protocol
Time will tell if they continue the study after overall survival data lock. But my bet is that they will end the overall Phase III study at the time they conduct the final analysis. And I sincerely doubt we will be waiting until November 2018 (the full 3 years after the last enrolled patient) for them to do so. I think they are in the final stretch, now.
Hi DMB,
First, I did not say that 2 years post enrollment was their target. What I said is that depending when 233 events actual hit/hits, and when the trial reaches the point where final statistical analysis can be 'officially' done, they will make the determination of whether to data lock "on time" or "to delay" in an effort to get more trial data -- in terms of follow-up time (long tail, tertiary, post progression).
Let me make sure it's clear that in my post -- which were written over the summer to Flip -- that I do see that the Company can be in the process of going through the steps to end the trial on time. It is possible that the overall survival data lock process, once the minimum of 233 death events have been recorded, might be at least a 3 month process in this Phase III study-- if all patients are expected to go through and end of study visit and undergo a final MRI for their eCFR (patient visits of remaining enrolled patients are on a 3 month schedule, after all). It is just my view that depending on when the OS threshold of 233 death events are/were crossed that the Company will/would determine if it makes sense to wait a long/short time more in order to get more long tail data -- IF data is ready to be locked before 1 year post last patient enrollment hasn't been crossed. They can easily decide to end the trial on time once the Month-Month process is done. I just stated I understood the rationale to extend until they have 2 years of post enrollment follow-up time.
Now, you asked me why 2 years post surgery wouldn't be sufficient. (Incidentally the date of last surgeries we do not have, so your 24 months of data for all patients may not be accurate -- August screening halt only represents when last first consents were signed, which came before surgery). To answer you question, it is not that that it wouldn't be. It's just that the statistical analysis plan survival data does not start at surgery. It starts at baseline. The time to treatment (from surgery to randomization) is not included in the tertiary survival calculation.
From the protocol:
"In the Phase III trial, PFS and survival times will be calculated from time of randomization, which is expected to occur approximately three months after initial surgery." -- protocol
Tertiary endpoint I referred to:
"• To compare survival at 6, 12, 18, 24, 36 and 48 months between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized)." -- protocol
And so if they lock the data now, then they will only have 18 month threshold of tertiary data on all 331 patients. They will not have 24 months of tertiary data on all 331 patients. Remember, survival is calculated from the 'time of randomization', it is NOT calculated from surgery.
Hopefully you may see now, you're adding the time to treatment (from surgery to randomization) back in, and making the assumption that because 2 year landmark historically data exists -- starting from surgery -- that it can then be used by the Company to determine just how well this study performs in comparison. However, that 2 year landmark analysis is of "all comers", and that considering this Phase III trial excluded "actually or apparent" disease progression at baseline, it is not an "All Comers" trial so it would not be an equal comparison.
In this trial, historical landmark survival data is much more useful when it comes to recurrent GBM -- where most rapids and some pseudos are typically excluded from study entry/landmark analysis, and therefore "all recurrent comers" is a similar patient population.
Also, it isn't just more tertiary data that they would gain by waiting. They would also gain more recurrent GBM data follow-up. We don't know exactly which Month the 248 progression event hit. However, we do know it was estimated to hit in November 2016 -- per their clinicaltrial site updated in October 2015. Having a full year of survival data, post progression, on all those 248 progression patients, before the study is locked, can be useful. Avastin, the last drug approved for rGBM, trial showed a 9-11 month median, if memory serves me. Here, that 12 month post progression data can be quite revealing.
Hope that helps.
(Please see this post instead; left a word out of my prior post response and that makes a difference.)
Hi DMB,
First, I did not say that 2 years post enrollment was their target. What I said is that depending when 233 events actual hit/hits, and when the trial reaches the point where final statistical analysis can be 'officially' done, they will make the determination of whether to data lock "on time" or "to delay" in an effort to get more trial data -- in terms of follow-up time (long tail, tertiary, post progression).
Let me make sure it's clear that in my post -- which were written over the summer to Flip -- that I do see that the Company can be in the process of going through the steps to end the trial on time. It is possible that the overall survival data lock process, once the minimum of 233 death events have been recorded, might be at least a 3 month process in this Phase III study-- if all patients are expected to go through and end of study visit and undergo a final MRI for their eCFR (patient visits of remaining enrolled patients are on a 3 month schedule, after all). It is just my view that depending on when the OS threshold of 233 death events are/were crossed that the Company will/would determine if it makes sense to wait a long/short time more in order to get more long tail data -- IF data is ready to be locked before 1 year post last patient enrollment hasn't been crossed. They can easily decide to end the trial on time once the Month-Month process is done. I just stated I understood the rationale to extend until they have 2 years of post enrollment follow-up time.
Now, you asked me why 2 years post surgery wouldn't be sufficient. (Incidentally the date of last surgeries we do not have, so your 24 months of data for all patients may not be accurate -- August screening halt only represents when last first consents were signed, which came before surgery). To answer you question, it is not that that it wouldn't be. It's just that the statistical analysis plan survival data does not start at surgery. It starts at baseline. The time to treatment (from surgery to randomization) is not included in the tertiary survival calculation.
From the protocol:?
"In the Phase III trial, PFS and survival times will be calculated from time of randomization, which is expected to occur approximately three months after initial surgery." -- protocol
Tertiary endpoint I referred to:
"• To compare survival at 6, 12, 18, 24, 36 and 48 months between patients in the treatment cohort and patients in the placebo cohort for all randomized patients (i.e., patients either randomized at baseline with no evidence of disease progression or patients with pseudoprogression at baseline who are later randomized)." -- protocol
And so if they lock the data now, then they will only have 18 month threshold of tertiary data on all 331 patients. They will not have 24 months of tertiary data on all 331 patients. Again, survival is calculated from the 'time of randomization', it is calculated from surgery.
Hopefully you may see now, you're adding the time to treatment (from surgery to randomization) back in, and making the assumption that because 2 year landmark historically data exists -- starting from surgery -- that it can then be used by the Company to determine just how well this study performs in comparison. However, that 2 year landmark analysis is of "all comers", and that considering this Phase III trial excluded "actually or apparent" disease progression at baseline, it is not an "All Comers" trial so it would not be an equal comparison.
In this trial, historical landmark survival data is much more useful when it comes to recurrent GBM -- where most rapids and some pseudos are typically excluded from study entry/landmark analysis, and therefore "all recurrent comers" is a similar patient population.
Also, it isn't just more tertiary data that they would gain by waiting. They would also gain more recurrent GBM data follow-up. We don't know exactly which Month the 248 progression event hit. However, we do know it was estimated to hit in November 2016 -- per their clinicaltrial site updated in October 2015. Having a full year of survival data, post progression, on all those 248 progression patients, before the study is locked, can be useful. Avastin, the last drug approved for rGBM, trial showed a 9-11 month median, if memory serves me. Here, that 12 month post progression data can be quite revealing.
Hope that helps.
We are talking about the overall study patients who remain alive with or without a first progression event.
Before posting my posts from the past, I just want to send my thoughts and prayers to iHub posters that I know in Florida, and to all in the path of Irma. Please stay safe.
As for data news, I'm still waiting patiently. Life has been so busy for me. I only have had the time to read the board sporadically. Hope all is well. Starting a week break. But, really don't want to get sucked back into posting as often as I once did. :)
If OS has passed 233 events, then yes, I do think that the company being so close to 2 years of survival data (post enrollment, on all enrolled patients) might consider continuing the trial into November in order to not have tertiary data censored. Once the trial ends, patients can continue to live beyond 2 years. However, if the trial has ended it would impossible to know what subsequent therapies all the surviving patients moved onto. At least if the data has not been locked for the 2 years, then all information that contributed to patients survival is recorded within the study. Continues to be my review that the company wants to capture the long tail within the data before locking. Whether they do or not remains to be seen.
Here is my post on July 8 in response to Flip who was upset when I shared my perception for rationale until waiting until later this Fall. This post follows others around that timeframe.
On 7-8-17:
" I'm not implying you didn't know under enrollment would delay events. I acknowledge now, just for clarity sake, that you knew that before my last post. I really didn't mean for you to walk away with that impression. Again, my point for even bringing it up was I perceive that the fact that they under enrolled changed the course of the decisions they made as far as data analysis are concerned.
I believe that Linda Liau being the GBM surgeon who routinely sees patients in practice was correct in her assumptions of pseudos being within the Indeterminate arm data. The Indeterminate arm includes both rapids and PsPD, it isn't all rapids as it is impossible that some pseudos were not removed due to mimicking progression. Yet, it is impossible to tell the rapid to pseudo split in that Indeterminate arm, as the vaccine was given at baseline, and the injections of a biological confounds the Month 2 confirmation scan. The only thing we know for sure is that those Indeterminate patients scans did not get progressively worse. We have no confirmation of they were either rapid or pseudo as nor did the scan get significantly better to clear up perfectly. Only one of 46 patients within the 55 Compassionate Use Arm patients scans cleared up completely to be labeled a pseudoprogression patients. But understandable since we both know that the vaccine could cause inflammation, which it might have done, particularly if the EXCLUDED patients had high TIL before vaccination. We don't have this information.
The main arm will also have Indeterminate patients IMHO. I do not perceive that all the pseudos will be found at the exclusion criteria Post RT MRI scan. You and I respectively disagree there. If I am correct, then Pseudo patient who did not cross the >25% progression threshold on that exclusion scan will be randomized into the main arm. I believe the study image review document includes an algorithm for evaluating and following indeterminate lesions, if seen on Month 2 scans, for assignment of progression status. If progression is confirmed at follow-up Month 4, the progression date is assigned to earliest time that any progression was initially detected. These would be the patients that AVII perceived to be bringing down the vaccine's progression free survival median. In his opinion, these patients would be labeled as progression patients on Month 2 scan. And I wholeheartedly respectively disagree with him.
Again, I don't think anyone said the events to date are not enough. Again, I do think PFS is mature for nGBM. The trial can data lock on time (233 events) or late (a specific follow-up Month). It has options. For you to imply what you think I meant is not correct. I said more follow-up time might be an avenue they pursue merely to get as much TERTIARY data as possible. Just because I think they have enough PFS data does not mean I don't see the benefit of having more follow-up data included within the analysis that goes in front of the regulators for each country. More follow-up data will allow the trial to pursue rGBM with the "late" as well. This is especially important to Germany HE. You're free to continue to disagree with my rationale. You might not see enough reasoning for option "additional follow-up". And that is okay. Lastly, just because I think there is rationale to wait for 2 year survival data on patients, given how close we are in terms of time to obtaining that information, does not mean that I think there is any rationale for waiting until 36 data point. That's more than a year away. We agree there. I can't even entertain the idea. No need to discuss it.
Of course, there are cases where progression statistical findings can be held back and not reported in the three day period. I'm speaking about this Phase III trial. I'm not speaking generically about all clinical trials. There are always exceptions to rules.
But I perceive that you are aware that I was not referring to any one of AVII's posted trial exception examples. He has a plethora of regulatory knowledge that he shares that I do not feel pertains to this Phase III study; those examples are not relevant here IMHO. From all I write about how he is wrong about the "false" events at Month 2 should make my opinion on the subject clear. I do not think they are hiding any PFS bad "false" outcome. His posts about this trial not allowing confirmation scan at Month 4 is speculation IMHO. I do not think this trial has failed PFS due to "false" events. They have not provided us with any notice of PFS data lock. My understanding is they plan to data lock both PFS and OS at the same time, per below:
"The Trial will reach data lock when the threshold events have been reached and the quality control checking has been completed for both the PFS and OS endpoints.
When data lock is reached, external statisticians and experts will make an independent analysis of the Trial data. The Company will remain blinded until data lock has been reached and the external data analysis has been completed." -- February PR
What is your review point? Do you think one of AVII's example for not reporting within 3 day rule applies here? Do you think the company is holding back from reporting unblinded PFS information? If your answer is yes, I wholeheartedly disagree with you.
In my opinion, if upon statistical review the PFS falls short of significant it will be because the main enrolled patients difference between placebo and vaccine legitimately fall below 4 months needed. I sincerely doubt we would find the trial fell short due to excessive "false" events at Month 2. I believe they do a confirmation scan at Month 4, and per the protocol progression events are dated to the event it was first seen. However, the analysis can not be held back if there is no legitimate reason to hold it back. And I obviously see no reason to hold it back. When they unblind and learn the PFS and OS results my view is they have 3 days to tell us and no exception to that rule can be applied in their case.
Now all that said, you do know that I'm of the opinion that PFS endpoint, when it is finally analyzed, might be close (within 3-5 month window). I perceive that the vaccine delays the time from tumor growth rate from initial surgical inventions to death but that extension may not show up all before progression. Depends on so many factors. It's an impossible thing to predict given this trial has an exclusion criteria, that may or may not remove some of their very best high TIL exhibiting patients (before vaccine) before enrollment. And so, yes, I do think it will be helpful to have at least 2 years survival Tertiary data on patients. You're free to continue to disagree.
As for what you wrote about my perception of fairness, this trial has been very fair. It allows placebo patients to crossover upon progression and so the patients that wanted to crossover and get there vaccine. The placebo patients who have not progressed will have the option to crossover if they do progress. To maintain integrity of the OS results, including long-tail follow-up, they can't let other placebo progression patients who passed that 3 month crossover window the opportunity to crossover without FDA approval IMHO.
And, yes I gave a range, as I think it's not an all or nothing additional follow-up decision. Again, I feel they'll make the best call of when to end the study. And I perceive that the decision will be made at the time OS endpoint is reached.
Anyway, we're going to continue to disagree on the merit of more follow-up time. Nothing either of us writes will affect whether they do immediately end the study or not. I'm choosing to wait patiently. "-- RK on 7-8-17
Here are posts that covered so of my viewpoints on the topics discussed.
On 4-26-17, I wrote:
"Flip,
Most PsPD are MGMT methylation. But that does not mean that most MGMT methylation presents PsPD. Radiation causes a blood barrier disruption to allow the drugs to get through better. But many tumors will not respond due to hypoxia. The rates of PsPD went up once TMZ was brought up front. But that doesn't mean all PsPD (mimic progression -- inflammation swelling) will show up at the 2 weeks post radiation. Many will show up later. It's the patients that have residual disease at BASELINE scan that were more likely to have shown signs of psPDs that early on IMHO. And you know my viewpoint afterwards -- there will be PsPD who show up in both arms as not all PsPD show up at the two week post window.
RANO effectively allows for a pseudo decision tree, and this trial will benefit from it. And RANO also captures progression of disease earlier than MacDonald as the non-enhancing disease component of using T-2/FLAIR means that PFS can be caught earlier in patients, before it shows up as enhancement disease in patients. In fact when RANO was used the medians PFS is about 1.3 months sooner than MacDonald. OS didn't change. There has to be something said about capturing progression before neurological signs show in a patient. Had RANO, with the FLAIR non enhancement disease component, been used when my nephew was a patient, then his progression would have been caught before his seizure event. They may have been able to start Bevacizamub and TMZ again or salvage therapy sooner. And he would have been eligible for his vaccine at crossover. And so the way I see it is that the patients who progress in this study will be able to be treated at a time that all therapies can help and be synergistic with the vaccine.
You and I debated long ago whether the trial would remain blinded at PFS endpoint. I was adamant that it would. You were convinced it would not. Now that it has and I was correct, the shorts are having a field day. But the protocol always had the option to treat the 100% PFS events as an interim to OS and they are in fact doing that.
I don't have much time lately to read/post. When I do, it's the same nonsense of allegations and "trial failure" hogwash. Rinse and repeat as they do. Once enrollment closed, the company had no control over PFS or OS. They are waiting just as we are.
I will say that I see that the death event rate at this point in the study will be slower than what most of this iHub expects. I will not be surprised one but if we are still waiting in the Fall. While the study has reached over 248 progression events by now, I suspect it will take a bit of time for those progression events to convert to death. While the median time from progression to death may be 7-9 months (depending on the image assessment -- using RANO is results in a 9 month median) even the event rate should have a long tail, similar to PFS and OS. Therefore in my opinion it isn't a simple 5-7 months after progression endpoint that the OS endpoint will hit. The fact is that the study under enrolled by 17 patients, so there is less of a patient pool to which to derive OS events. It is also fact that the event rate is not a stagnant occurrence. The one the company witnessed may have been based on the median. And I perceive the event rate is now at the longer tail. Patients who evented shortly after enrollment will be patients who were not as responsive to one or more of the standard of care therapies (i.e, Radiation, TMZ or a bit of both). But the patients who reached progression later in the study were responsive to standard of care therapy. And there reason to believe they will be responsive again. Typically the longer the patients go without recurrence, the more likely they will qualify for another surgery. MGMT status does not change at progression. The more time the patient is off TMZ, the more likely they will be responsive once again. I'm long as the next person, but let me be clear, I know we reached 248 progression events. I just honestly believe we still have a way to go to OS endpoint. I was correct on the my assumption that the company wouldn't unblind the trial at progression. I sense I'll be accurate about this too. Time will tell. "-- RK on 4- 26-17
--
On 4-27-17, I wrote:
"No, I do not mean to imply that, at all. It is impossible for me to determine -- without specific enrollment and progression study data— that if OS endpoint were to to trigger in May/June" timeframe it somehow means a OS median bad news. It’s a lot easier to determine whether the study is going well, verses badly based on when endpoints are reached and how much time has passed since then. But that said, I still think it’s unlikely that we reach the OS threshold anytime soon.
The company gave us an estimated timing of when we could expect the 233 OS event to come in, below:
“There are 331 patients enrolled in the Trial. The Trial endpoints involve thresholds of 248 “events” for PFS and 233 “events” for OS. PFS events are primarily tumor progression (i.e., recurrence), although they can occasionally be patient deaths which occur without prior evidence of tumor recurrence. OS events are patient deaths. The PFS and OS events are continuing to accumulate as the Trial continues. The PFS events have surpassed the 248-event threshold, but the OS events have not yet reached the 233-event threshold. Based upon the pace of OS events during the last six to eight months, the Company’s current anticipation is that it will be several months until the Trial reaches 233 OS events.” — NW Bio Feb 2017
https://www.nwbio.com/nw-bio-announces-lifting-clinical-hold-dcvax-l-phase-iii-trial-fda-progression-free-survival-events-reached-overall-survival-events-not-yet-reached/
They told us they anticipated it would take “several months” to cross the OS endpoint. But they used the progression-to-death conversion pace average from starting from either August 2016 (6 months) or June 2016 - (8 months). Think about that 6-to-8 month period that the company referenced for a second. Most of the death events that comprised that period would be coming from patients who possibly had first progression events that came in at least a few months earlier, which dates back progression to Early 2016/Late 2015. And that would mean the pace events would contain a fair amount of patients who may have been screened during the height of enrollment ramp-up period (Fall 2014 - Fall 2015). What do we know about screening in that period? Germany opened enrollment and had begun screening to a lower White Blood Count inclusion criteria. And so it’s reasonable for me to assume that the death events that came during the June 2016 - February 2017 timeframe contained a fair amount of German patients whose immune systems may have been compromised. These immune comprised patients, particularly if on placebo only, would be more likely to be unable to tolerate cytotoxic drugs, which would result in lower PFS, and then subsequently and earlier death. These are the patients that they include in their event pace.
Said another way, we are waiting for the patients with progression events that triggered in late 2016/early 2017 to convert to a death event. We don't know how many patients we are specifically waiting for. But the patient pool that remains alive -- both with and without a of progression event -- are not the same as those who have died during the company "death pace" estimation period -- regardless that all the patients may have come from similar enrollment periods. Understand there are patients who recur before progression medians. And these patients would be more likely to convert into death events that triggered before the OS median. Then there are patients who recur at or after progression medians. And these later recurrences patients would be more likely to convert into death events that trigger after the OS Median. PFS events correlate to OS events. But, the earlier on a patient progresses, the more likely the time between progression to death be brief. The longer progression is delayed, the MEAN time between progression to death begins to rises. But means are averages. Healthier patients at progression, leads to a health delay in death, particularly if the patients have time to mount an immune response. And so we possibly have to wait for the events of the above average patients to come in.
Neither the MEAN pace or the MEDIAN pace accurately captures the distance it could take for "late progression" patient on average to die. The reality is that the company under enrolled. And because they did, there is a smaller group of patients that death events can convert from progression than originally anticipated. Those patients that contributed to progression event that occurred after the 248 patient progression event was reached are least likely to contribute to death events here and now. The longer the patient goes without a first progression event, the more likely they are able to move onto subsequent therapies. The patients tumor MGMT status doesn't change even if it does take on a Mesenchymal signature. And so if they were responsive to TMZ before progression, the odds are they will be again. Those odds go up the greater the distance to their last TMZ adjuvant treatment. The late progression patients may be eligible for repeat surgery and possibly even targeted radiation therapy -- therapies that are synergistic with the vaccine, particularly for the crossover patients. Prior trials did not have patients continue on vaccine therapy after progression. This one does. And there has been no DCVax-L study that treated recurrent patients at the first confirmed sign of progression. This study allows for that, supply permitting. Then there is the fact that the crossover patients do not have to wait for the vaccine to be made for them. And so hopefully for the healthier immune patient the vaccine extends the distance between progression and death. While we may not know how many months out from progression the patients may be, we do know that it could take a long time for patients who are responding to salvage therapy to die. The event pace of deaths that converted from early last year's progression patients shouldn't be the pace that is used to ballpark how long we have to wait now. And so I was simply was pointing out that I think we may have a way to go to reach OS. "--RK 4-27-17
After the company slides were revealed in June, I wrote on 6-6-17:
"I've spent the last month not reading the iHub. Signed off. Time away has been great. Hope everyone is well. Lots going on in life, haven't had time. Nor did I have the patience to correct the posts that construe my words out of context, both long and short.
Thanks Senti for transcribing Bosch's speech.
The event rate is much slower at this point. Just like I said it would be. We are waiting for the late tail patients to pass at this point. And hopefully many of them will benefit once Bevacizamub is added. The last enrolled patient was 19 months ago. That brings us to late Oct/early Nov 2015 timeframe, which is what many longs here agree with. The last screened patient was 22 months ago, another aspect we agree with. My enrollment median was off by a Quarter. And I suspect that means that they stopped enrollment for a couple of quarters around the time they decided to work on changing the statistical analysis plan. In any event, it still means that 331 patient enrollment are approaching 2 years out from surgery. About 30% are alive. And while we don't know how many exactly are approaching their 3 + year point, but we do know there not all those 100 or so patients were enrolled the last months. Being blind, and having the update we received, odd are that the vaccine is delaying death in a majority of patients, IMHO.
I still think we will hear August timeframe at the earliest. But again, I will not be surprised if we are waiting until the Fall. Time will tell. " -- RK on 6-6-17
Then I had questions on my posts, and in response to Hank on 7-7-17, I wrote:
"Hi Hank,
No, I did not mean August 2018. I do not think it is necessary to wait beyond Q4 2017. I just was not clear why I meant that November 2017 date. Sorry about that.
I'll elaborate. I meant 3 years of data on patients recruited from different protocol enrollment periods, but not 3 years of data on all patients.
While the Company reached median enrollment around May 2014. Before May 2014, the first 50% of 331 patients were enrolled. And after May 2014, the other 50% of 331 patients were enrolled. And so naturally some are think that the Company will have 3 years of data on the second 1/2 of the enrollment period at this time, as today is already 3 years later. However, the patients they enrolled -- if there were more than a few, which I sincerely doubt -- those patients would be considered part of the older protocol analysis. That final SAP, the newest one, includes looking at the effect WBC has on OS.
I know that the ASCO enrollment chart shows that enrollment was rising during that May - August 2014 period. What I think they did with that chart is to just plot a few points where enrollment stood and then connected the dots on the graph -- it's not a perfect depiction on enrollment at months in between the plots, IMHO. However, we also know that the newest protocol (which none of us have) aligns with the newest SAP (which none of us have seen).
Screening for the new WBC inclusion criteria (in Germany -- which had not begun enrollment yet) -- didn't begin before August 2014. In order to get a fair amount of new SAP screened patients, the Company would need to wait 3 months. Waiting until November 2017 would capture three months of data in the longer tail of 3 years of data, for the second 1/2 of the enrollment period, which then can be used to compare to historical SoC (only about 1/3 of GBM Patients live over 2 years -- when considering all comers (rapids, pseudo, stable disease).
Plus if they wait until November 2017, they would have 2 years of enrollment data on all patients in the Phase III trial (last patient enrolled in Nov 2015). BUT at that point they would have at least 3 years of enrollment data for some of the Germany WBC inclusion screened patients. And at least 1 year post survival progression information on the primary endpoint patients (assuming it hit at their estimated month November 2016).
Here's the quarters I referenced above. Add 3 years:
Q2 2014 -
-- Expanded access arm added on ClinicalRegister, pseudo arm likely closed.
-- Fraunhofer IZI site clears, to enter the trial.
Q3 2014 -
Germany starts screening, August 2014 Fraunhofer IZI batch manufacturing begins.
Page 34 of the 2014 Fraunhofer IZI annual report confirms Germany official recruitment started after August 2014, production batch manufacturing started then:
http://www.izi.fraunhofer.de/content/dam/izi/en/documents/Publications/Annual_Report_IZI_2015.pdf
-- August 2014: Trial Enhancement Announced: New SAP, lower WBC inclusion
-- 348-patient Phase III trial (now 331-patients)
Q4 2014:
Germany HE patient announcement. Germany enrollment period gearing up, from the inclusion screened patients.
Q1 2015:
-- Additional Germany sites running
-- Canada enters the trial.
By the way, I don't think that the Company screened many patients between May -August 2014, as they were in pursuit of a new WBC protocol. Not that it matters much, but the new protocol would need to eliminate the pseudoprogression arm, as that arm was closed and not running in Germany. There is no lower WBC inclusion criteria for those pseudoprogression cohort patients (enrolled separately outside the main arm).
When one thinks about the SAP data breakdown for the trial, it should be considered as different enrollment periods, for updates to the protocol depending on the the trial. And so this trial can benefit from having at least 3 years of enrollment data for the latest protocol WBC inclusion patients.
1) The Vanguard cohort -- 2008 -2009 -- 140 patient trial - oldest SAP
2) The restart cohort -- 2011 -2012, included a new SAP -- analysis up to 240 patients.
2a) The pseudo period - 2012 - 2014 -- used the restart SAP for the main arm, from the 240-patients trial, but added tertiary, pseudo analysis outside the main arm. A total of 312 patients could be including within the data, but only 240-patients in the main arm.
3)The WBC inclusion period (lower). New SAP for 348-patient trial. This still included Pseudo analysis, outside the study, but recruitment was closed sometime before August 2014 WBC screening opened. (We do not know if this enrollment portion of the trial uses later patents on obtaining higher quality Dendritic cell precursors.)
I imagine that the newest SAP from the WBC cohort will be evaluating all the overall trial. But then similiar to the old SAP, from the restart period, it will then look at the cohorts periods, one by one.
"Overall cohort (i.e., all data from the beginning of the study, followed by the “vanguard” cohort – i.e. those patients originally enrolled prior to the hold and followed up to the current time, and a “restart cohort”, i.e., those follow-up since the restart of the clinical trial (i.e., since August 2011). Although the primary view is towards all patients enrolled (i.e. overall cohort), a consistency of effect across subsets would be the primary intent of these additional cuts.
The DMC review will contain, but not be limited to, the following efficacy and safety analyses: patient enrollment/disposition, patient demographics/baseline characteristics, AEs and of SAEs (both tabular and listings as
appropriate), discontinuations from treatment and/or follow-up, survival (overall and progression-free survival) –displayed in graphic (i.e., Kaplan- Meier curves) and in tabular formats, causes specific mortality, and selected
lab data." -- old restart SAP " -- RK on 7-7-17
--
My post to Flip on 7-8-17:
"Flip,
The company under enrolled the Phase III trial. That is one of the contributing factors of why the data is delayed (less pool of patients to which to derive events; getting 233 events from 331 patients would take longer than getting 233 events from 348 patients). Whether it's delayed significantly above that under enrollment we do not know. I mention this, as under enrollment has played a factor as to whether or not to "unblind" at progression. The Company is taking no chances in unblinding too early, in light of under enrolling, and they want as much aged data in my opinion.
To be clear, waiting has nothing to do with needing more events. Waiting has to do with wanting more follow up time. Waiting provides more perspective on the longer tail data too (inside and outside the main arm of study: pseudoprogression arm). The event count between waiting between August to end of October might not rise much if at all. Instead what rises is the amount of time either post progression that all the original 248 patient might be from their initial progression diagnosis. And what also rises is the length of the the remaining survivors, without events, are alive from their initial diagnosis. We would get the EXACT amount of patients in both arms that are alive 24 months post enrollment. That's a historical known figure (see recent pasted retrospective report below). In general if 33% of standard of care is alive at 2 years (speculated estimated inclusion criteria odds (note the historical for All Comers GBM is roughly 20%: per post-RT-TMZ and post-BEV period (5/2009–12/2013, P3: 19.8 (18.8-20.8)) and this trial has recorded 40% (placebo) and over 50% (vaccine) that would be HUGH.
And I haven't had any communications with the company or with Chris. I'm basing my thoughts on my own observations of what additional follow-up time would reveal within the data.
As for AA, that's off the table while we wait. The data is continuing to accumulate for progression. And it's just as important that it does. This trial is not a recurrent GBM trial. However, the fact the investigators are blind to the initial cohort that the crossover patients were in it essentially makes it a standardized recurrent entry for the placebo cohort. Depending on how the data for survival post progression comes out came for placebo 60/40 yes/no to crossover, we could end up seeing an approval for initiation of vaccine for recurrent GBM as well. And sure, they may have enough data for stopping the trial with less follow-up on the last of the 248 progression patients, but I sincerely believe they are considering holding out as long as they can.
As for "be careful what you wish for", I find that to be an interesting comment if taken at face value reference to regulators. (Unless between the lines you meant that for stock issuances, which depending on how the cash position plays out (if they are able to raise from prior issuance (I.e, stock offerings to insiders) could influence their interest to wait or not, obviously. But, I think the money situation will work itself out in the interim.) Seriously, there is no foul if the either the Company or the FDA ends the trial at 233 events. They would just have less follow-up time data before the lock, and more afterwards. We will just have to see how it plays out. I personally think if UCLA's data comes out in strong favor for one of the three combinations (without or with TLR3 cream or with TLR9 cream) they may fully encourage NWBO about having the trial go until two years of enrollment on all patients. To see the amount of patients who are 27 months out from initial surgery both within the non-progression and the recurrent patient population in the information that they review, would be very useful in making the determine whether or not to provide full approval.
As for NW Bio providing us insight as to when the data is "officially" being analyzed, I'm not sure they will. We saw how that information was manipulated. And we all know that the trial is likely at 233 events as we speak. Once they get the unblinded study news, they only have 3 days to tell us it's being analyzed. They do not have to tell us that they are waiting a couple of months to have more follow-up time of remaining survival. We can read about it in SEC statements from 10Q that the Company remains blinded as the Phase III trial nears completion. We reported that the patient data from all the 80 clinical sites continues to carefully scrubbed and crossed checked by the CROs who will report that final information to the DMC who is monitoring the trial and is also blind to patients initial cohorts. The Phase III data will be locked and analyzed shortly thereafter once that process is complete. Or something like that...
I sincerely do not mind if you disagree with my speculative thoughts on whether they will or will increase the follow-up time. It's just something I'm considering that they might decide to do given how close they are to getting significant data on all enrolled patients. I respect your opinion to respectively disagree with mine, as always. Best,
Historical GBM data paper, I referenced:
Survival benefit of glioblastoma patients after FDA approval of temozolomide concomitant with radiation and bevacizumab: A population-based study Ping Zhu1,2, Xianglin L. Du1, Guangrong Lu2 and Jay-Jiguang Zhu2
Received: November 23, 2016 Accepted: March 20, 2017 Published: April 12, 2017
ABSTRACT Few population-based analyses have investigated survival change in glioblastoma multiforme (GBM) patients treated with concomitant radiotherapy-temozolomide (RT- TMZ) and adjuvant temozolomide (TMZ) and then bevacizumab (BEV) after Food and Drug Administration (FDA) approval, respectively. We aimed to explore the effects on survival with RT-TMZ, adjuvant TMZ and BEV in general GBM population based on the Surveillance, Epidemiology, and End Results (SEER) and Texas Cancer Registry (TCR) databases. A total of 28933 GBM patients from SEER (N = 24578) and TCR (N = 4355) between January 2000 and December 2013 were included. Patients were grouped into three calendar periods based on date of diagnosis: pre-RT-TMZ and pre-BEV (1/2000–2/2005, P1), post-RT-TMZ and pre-BEV (3/2005–4/2009, P2), and post-RT-TMZ and post-BEV (5/2009–12/2013, P3). The association between calendar period of diagnosis and survival was analyzed in SEER and TCR, separately, by the Kaplan-Meier method and Cox proportional hazards model. We found a significant increase in median overall survival (OS) across the three periods in both populations. In multivariate models, the risk of death was significantly reduced during P2 and further decreased in P3, which remained unchanged after stratification. Comparison and validation analysis were performed in the combined dataset, and consistent results were observed. We conclude that the OS of GBM patients in a “real-world” setting has been steadily improved from January 2000 to December 2013, which likely resulted from the administrations of TMZ concomitant with RT and adjuvant TMZ for newly diagnosed GBM and then BEV for recurrent GBM after respective FDA approval.
Quote from the research report:
"Our findings suggested that OS and 1-year survival rate improved significantly across calendar period of diagnosis in SEER, TCR and the combined dataset, except that the improved survival within P3 compared to P2 in TCR was not significant (P = 0.833). There was no survival benefit observed in 1-year or 2-year survival rate between P2 and P3 in TCR, nor 2-year survival rate between P2 and P3 in SEER and in the combined dataset. One explanation for the reduced survival in P3 comparing to P2 in TCR could be related to the lower proportion of patients received radiation during P3. The proportion of patients treated with radiation therapy in TCR decreased from 80.8% to 69.8% and to 60.1% in P1, P2, and P3, successively. However, when we did a sub-analysis of the extended follow-up of the TCR dataset up to May 2015, we found that the difference in survival between P2 and P3 was significant (Supplementary Figure 1 and Supplementary Table 2). This finding indicated that the survival benefit of BEV could be observed when given longer period of follow- up and that may compensate the loss of benefit with lower radiation proportion in P3."
Open PDF to see Table 2
Table 2: Median OS, 1-year and 2-year survival rate by calendar period of diagnosis in SEER, TCR and the combined dataset*
Survival statistics
Calendar Period of Diagnosis
Jan 2000 - Feb 2005 (P1)
Mar 2005 - Apr 2009 (P2)
May 2009 - Dec 2013 (P3)
Total
(P1, P2 and P3)
P1 vs P2
P1 vs P3
P2 vs P3
http://www.impactjournals.com/oncotarget/index.php?journal=oncotarget&page=article&op=download&path%5B%5D=17054&path%5B%5D=59124
Older SEERs data on GBM:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4954825/pdf/fpubh-04-00151.pdf
http://scholarworks.waldenu.edu/cgi/viewcontent.cgi?article=1267&context=dissertations
" --RK on 7-8-17
The 7-8-17 post all the way at the top of this post followed Flips response to this 7-8-17 above.
(Re-posting to hopefully get rid of "?" which are not part of anything I had typed - it's a format added error from paste from iPhone notes).
Hi Turtle, Hope all is well.
Before posting my posts from the past, I just want to send my thoughts and prayers to iHub posters that I know in Florida, and to all in the path of Irma. Please stay safe.
As for my thoughts, I'm still waiting patiently. Life has been so busy for me. I only have had the time to read the board sporadically. Starting a week break. But, really don't want to get sucked back into posting as often as I once did. :)
If OS has passed 233 events, then yes, I do think that the company being so close to 2 years of survival data (post enrollment, on all enrolled patients) might consider continuing the trial into November in order to not have tertiary data censored. Once the trial ends, patients can continue to live beyond 2 years. However, if the trial has ended it would impossible to know what subsequent therapies all the surviving patients moved onto. At least if the data has not been locked for the 2 years, then all information that contributed to patients survival is recorded within the study. Continues to be my review that the company wants to capture the long tail within the data before locking. Whether they do or not remains to be seen.
Here is my post on July 8 in response to Flip who was upset when I shared my perception for rationale until waiting until later this Fall. This post follows others around that timeframe.
On 7-8-17, I wrote in response to Flip:
" I'm not implying you didn't know under enrollment would delay events. I acknowledge now, just for clarity sake, that you knew that before my last post. I really didn't mean for you to walk away with that impression. Again, my point for even bringing it up was I perceive that the fact that they under enrolled changed the course of the decisions they made as far as data analysis are concerned.
?
I believe that Linda Liau being the GBM surgeon who routinely sees patients in practice was correct in her assumptions of pseudos being within the Indeterminate arm data. The Indeterminate arm includes both rapids and PsPD, it isn't all rapids as it is impossible that some pseudos were not removed due to mimicking progression. Yet, it is impossible to tell the rapid to pseudo split in that Indeterminate arm, as the vaccine was given at baseline, and the injections of a biological confounds the Month 2 confirmation scan. The only thing we know for sure is that those Indeterminate patients scans did not get progressively worse. We have no confirmation of they were either rapid or pseudo as nor did the scan get significantly better to clear up perfectly. Only one of 46 patients within the 55 Compassionate Use Arm patients scans cleared up completely to be labeled a pseudoprogression patients. But understandable since we both know that the vaccine could cause inflammation, which it might have done, particularly if the EXCLUDED patients had high TIL before vaccination. We don't have this information.
?
The main arm will also have Indeterminate patients IMHO. I do not perceive that all the pseudos will be found at the exclusion criteria Post RT MRI scan. You and I respectively disagree there. If I am correct, then Pseudo patient who did not cross the >25% progression threshold on that exclusion scan will be randomized into the main arm. I believe the study image review document includes an algorithm for evaluating and following indeterminate lesions, if seen on Month 2 scans, for assignment of progression status. If progression is confirmed at follow-up Month 4, the progression date is assigned to earliest time that any progression was initially detected. These would be the patients that AVII perceived to be bringing down the vaccine's progression free survival median. In his opinion, these patients would be labeled as progression patients on Month 2 scan. And I wholeheartedly respectively disagree with him.
?
Again, I don't think anyone said the events to date are not enough. Again, I do think PFS is mature for nGBM. The trial can data lock on time (233 events) or late (a specific follow-up Month). It has options. For you to imply what you think I meant is not correct. I said more follow-up time might be an avenue they pursue merely to get as much TERTIARY data as possible. Just because I think they have enough PFS data does not mean I don't see the benefit of having more follow-up data included within the analysis that goes in front of the regulators for each country. More follow-up data will allow the trial to pursue rGBM with the "late" as well. This is especially important to Germany HE. You're free to continue to disagree with my rationale. You might not see enough reasoning for option "additional follow-up". And that is okay. Lastly, just because I think there is rationale to wait for 2 year survival data on patients, given how close we are in terms of time to obtaining that information, does not mean that I think there is any rationale for waiting until 36 data point. That's more than a year away. We agree there. I can't even entertain the idea. No need to discuss it.
?
Of course, there are cases where progression statistical findings can be held back and not reported in the three day period. I'm speaking about this Phase III trial. I'm not speaking generically about all clinical trials. There are always exceptions to rules.
?
But I perceive that you are aware that I was not referring to any one of AVII's posted trial exception examples. He has a plethora of regulatory knowledge that he shares that I do not feel pertains to this Phase III study; those examples are not relevant here IMHO. From all I write about how he is wrong about the "false" events at Month 2 should make my opinion on the subject clear. I do not think they are hiding any PFS bad "false" outcome. His posts about this trial not allowing confirmation scan at Month 4 is speculation IMHO. I do not think this trial has failed PFS due to "false" events. They have not provided us with any notice of PFS data lock. My understanding is they plan to data lock both PFS and OS at the same time, per below:
?
"The Trial will reach data lock when the threshold events have been reached and the quality control checking has been completed for both the PFS and OS endpoints.
When data lock is reached, external statisticians and experts will make an independent analysis of the Trial data. The Company will remain blinded until data lock has been reached and the external data analysis has been completed." -- February PR
?
What is your review point? Do you think one of AVII's example for not reporting within 3 day rule applies here? Do you think the company is holding back from reporting unblinded PFS information? If your answer is yes, I wholeheartedly disagree with you.
?
In my opinion, if upon statistical review the PFS falls short of significant it will be because the main enrolled patients difference between placebo and vaccine legitimately fall below 4 months needed. I sincerely doubt we would find the trial fell short due to excessive "false" events at Month 2. I believe they do a confirmation scan at Month 4, and per the protocol progression events are dated to the event it was first seen. However, the analysis can not be held back if there is no legitimate reason to hold it back. And I obviously see no reason to hold it back. When they unblind and learn the PFS and OS results my view is they have 3 days to tell us and no exception to that rule can be applied in their case.
?
Now all that said, you do know that I'm of the opinion that PFS endpoint, when it is finally analyzed, might be close (within 3-5 month window). I perceive that the vaccine delays the time from tumor growth rate from initial surgical inventions to death but that extension may not show up all before progression. Depends on so many factors. It's an impossible thing to predict given this trial has an exclusion criteria, that may or may not remove some of their very best high TIL exhibiting patients (before vaccine) before enrollment. And so, yes, I do think it will be helpful to have at least 2 years survival Tertiary data on patients. You're free to continue to disagree.
?
As for what you wrote about my perception of fairness, this trial has been very fair. It allows placebo patients to crossover upon progression and so the patients that wanted to crossover and get there vaccine. The placebo patients who have not progressed will have the option to crossover if they do progress. To maintain integrity of the OS results, including long-tail follow-up, they can't let other placebo progression patients who passed that 3 month crossover window the opportunity to crossover without FDA approval IMHO.
?
And, yes I gave a range, as I think it's not an all or nothing additional follow-up decision. Again, I feel they'll make the best call of when to end the study. And I perceive that the decision will be made at the time OS endpoint is reached.
?
Anyway, we're going to continue to disagree on the merit of more follow-up time. Nothing either of us writes will affect whether they do immediately end the study or not. I'm choosing to wait patiently. "-- RK on 7-8-17
?
Here are posts that covered so of my viewpoints on the topics discussed.
On 4-17-17, I wrote:
?
"Flip,
?
Most PsPD are MGMT methylation. But that does not mean that most MGMT methylation presents PsPD. Radiation causes a blood barrier disruption to allow the drugs to get through better. But many tumors will not respond due to hypoxia. The rates of PsPD went up once TMZ was brought up front. But that doesn't mean all PsPD (mimic progression -- inflammation swelling) will show up at the 2 weeks post radiation. Many will show up later. It's the patients that have residual disease at BASELINE scan that were more likely to have shown signs of psPDs that early on IMHO. And you know my viewpoint afterwards -- there will be PsPD who show up in both arms as not all PsPD show up at the two week post window.
?
RANO effectively allows for a pseudo decision tree, and this trial will benefit from it. And RANO also captures progression of disease earlier than MacDonald as the non-enhancing disease component of using T-2/FLAIR means that PFS can be caught earlier in patients, before it shows up as enhancement disease in patients. In fact when RANO was used the medians PFS is about 1.3 months sooner than MacDonald. OS didn't change. There has to be something said about capturing progression before neurological signs show in a patient. Had RANO, with the FLAIR non enhancement disease component, been used when my nephew was a patient, then his progression would have been caught before his seizure event. They may have been able to start Bevacizamub and TMZ again or salvage therapy sooner. And he would have been eligible for his vaccine at crossover. And so the way I see it is that the patients who progress in this study will be able to be treated at a time that all therapies can help and be synergistic with the vaccine.
?
You and I debated long ago whether the trial would remain blinded at PFS endpoint. I was adamant that it would. You were convinced it would not. Now that it has and I was correct, the shorts are having a field day. But the protocol always had the option to treat the 100% PFS events as an interim to OS and they are in fact doing that.
?
I don't have much time lately to read/post. When I do, it's the same nonsense of allegations and "trial failure" hogwash. Rinse and repeat as they do. Once enrollment closed, the company had no control over PFS or OS. They are waiting just as we are.
?
I will say that I see that the death event rate at this point in the study will be slower than what most of this iHub expects. I will not be surprised one but if we are still waiting in the Fall. While the study has reached over 248 progression events by now, I suspect it will take a bit of time for those progression events to convert to death. While the median time from progression to death may be 7-9 months (depending on the image assessment -- using RANO is results in a 9 month median) even the event rate should have a long tail, similar to PFS and OS. Therefore in my opinion it isn't a simple 5-7 months after progression endpoint that the OS endpoint will hit. The fact is that the study under enrolled by 17 patients, so there is less of a patient pool to which to derive OS events. It is also fact that the event rate is not a stagnant occurrence. The one the company witnessed may have been based on the median. And I perceive the event rate is now at the longer tail. Patients who evented shortly after enrollment will be patients who were not as responsive to one or more of the standard of care therapies (i.e, Radiation, TMZ or a bit of both). But the patients who reached progression later in the study were responsive to standard of care therapy. And there reason to believe they will be responsive again. Typically the longer the patients go without recurrence, the more likely they will qualify for another surgery. MGMT status does not change at progression. The more time the patient is off TMZ, the more likely they will be responsive once again. I'm long as the next person, but let me be clear, I know we reached 248 progression events. I just honestly believe we still have a way to go to OS endpoint. I was correct on the my assumption that the company wouldn't unblind the trial at progression. I sense I'll be accurate about this too. Time will tell. "-- RK on 4/26/17
--
On 4/27:
"No, I do not mean to imply that, at all. It is impossible for me to determine -- without specific enrollment and progression study data— that if OS endpoint were to to trigger in May/June" timeframe it somehow means a OS median bad news. It’s a lot easier to determine whether the study is going well, verses badly based on when endpoints are reached and how much time has passed since then. But that said, I still think it’s unlikely that we reach the OS threshold anytime soon.
?
The company gave us an estimated timing of when we could expect the 233 OS event to come in, below:
?
“There are 331 patients enrolled in the Trial. The Trial endpoints involve thresholds of 248 “events” for PFS and 233 “events” for OS. PFS events are primarily tumor progression (i.e., recurrence), although they can occasionally be patient deaths which occur without prior evidence of tumor recurrence. OS events are patient deaths. The PFS and OS events are continuing to accumulate as the Trial continues. The PFS events have surpassed the 248-event threshold, but the OS events have not yet reached the 233-event threshold. Based upon the pace of OS events during the last six to eight months, the Company’s current anticipation is that it will be several months until the Trial reaches 233 OS events.” — NW Bio Feb 2017
?
https://www.nwbio.com/nw-bio-announces-lifting-clinical-hold-dcvax-l-phase-iii-trial-fda-progression-free-survival-events-reached-overall-survival-events-not-yet-reached/
?
They told us they anticipated it would take “several months” to cross the OS endpoint. But they used the progression-to-death conversion pace average from starting from either August 2016 (6 months) or June 2016 - (8 months). Think about that 6-to-8 month period that the company referenced for a second. Most of the death events that comprised that period would be coming from patients who possibly had first progression events that came in at least a few months earlier, which dates back progression to Early 2016/Late 2015. And that would mean the pace events would contain a fair amount of patients who may have been screened during the height of enrollment ramp-up period (Fall 2014 - Fall 2015). What do we know about screening in that period? Germany opened enrollment and had begun screening to a lower White Blood Count inclusion criteria. And so it’s reasonable for me to assume that the death events that came during the June 2016 - February 2017 timeframe contained a fair amount of German patients whose immune systems may have been compromised. These immune comprised patients, particularly if on placebo only, would be more likely to be unable to tolerate cytotoxic drugs, which would result in lower PFS, and then subsequently and earlier death. These are the patients that they include in their event pace.
?
Said another way, we are waiting for the patients with progression events that triggered in late 2016/early 2017 to convert to a death event. We don't know how many patients we are specifically waiting for. But the patient pool that remains alive -- both with and without a of progression event -- are not the same as those who have died during the company "death pace" estimation period -- regardless that all the patients may have come from similar enrollment periods. Understand there are patients who recur before progression medians. And these patients would be more likely to convert into death events that triggered before the OS median. Then there are patients who recur at or after progression medians. And these later recurrences patients would be more likely to convert into death events that trigger after the OS Median. PFS events correlate to OS events. But, the earlier on a patient progresses, the more likely the time between progression to death be brief. The longer progression is delayed, the MEAN time between progression to death begins to rises. But means are averages. Healthier patients at progression, leads to a health delay in death, particularly if the patients have time to mount an immune response. And so we possibly have to wait for the events of the above average patients to come in.
?
Neither the MEAN pace or the MEDIAN pace accurately captures the distance it could take for "late progression" patient on average to die. The reality is that the company under enrolled. And because they did, there is a smaller group of patients that death events can convert from progression than originally anticipated. Those patients that contributed to progression event that occurred after the 248 patient progression event was reached are least likely to contribute to death events here and now. The longer the patient goes without a first progression event, the more likely they are able to move onto subsequent therapies. The patients tumor MGMT status doesn't change even if it does take on a Mesenchymal signature. And so if they were responsive to TMZ before progression, the odds are they will be again. Those odds go up the greater the distance to their last TMZ adjuvant treatment. The late progression patients may be eligible for repeat surgery and possibly even targeted radiation therapy -- therapies that are synergistic with the vaccine, particularly for the crossover patients. Prior trials did not have patients continue on vaccine therapy after progression. This one does. And there has been no DCVax-L study that treated recurrent patients at the first confirmed sign of progression. This study allows for that, supply permitting. Then there is the fact that the crossover patients do not have to wait for the vaccine to be made for them. And so hopefully for the healthier immune patient the vaccine extends the distance between progression and death. While we may not know how many months out from progression the patients may be, we do know that it could take a long time for patients who are responding to salvage therapy to die. The event pace of deaths that converted from early last year's progression patients shouldn't be the pace that is used to ballpark how long we have to wait now. And so I was simply was pointing out that I think we may have a way to go to reach OS. "--RK 4-27-17
?
After the company slides were revealed in June, I wrote on 6-6-17:
?
"I've spent the last month not reading the iHub. Signed off. Time away has been great. Hope everyone is well. Lots going on in life, haven't had time. Nor did I have the patience to correct the posts that construe my words out of context, both long and short.
?
Thanks Senti for transcribing Bosch's speech.
?
The event rate is much slower at this point. Just like I said it would be. We are waiting for the late tail patients to pass at this point. And hopefully many of them will benefit once Bevacizamub is added. The last enrolled patient was 19 months ago. That brings us to late Oct/early Nov 2015 timeframe, which is what many longs here agree with. The last screened patient was 22 months ago, another aspect we agree with. My enrollment median was off by a Quarter. And I suspect that means that they stopped enrollment for a couple of quarters around the time they decided to work on changing the statistical analysis plan. In any event, it still means that 331 patient enrollment are approaching 2 years out from surgery. About 30% are alive. And while we don't know how many exactly are approaching their 3 + year point, but we do know there not all those 100 or so patients were enrolled the last months. Being blind, and having the update we received, odd are that the vaccine is delaying death in a majority of patients, IMHO.
?
I still think we will hear August timeframe at the earliest. But again, I will not be surprised if we are waiting until the Fall. Time will tell. " -- RK on 6-6-17
?
Then I had questions on my posts, and in response to Hank on 7-7-17, I wrote:
?
"Hi Hank,
?
No, I did not mean August 2018. I do not think it is necessary to wait beyond Q4 2017. I just was not clear why I meant that November 2017 date. Sorry about that.
?
I'll elaborate. I meant 3 years of data on patients recruited from different protocol enrollment periods, but not 3 years of data on all patients.
?
While the Company reached median enrollment around May 2014. Before May 2014, the first 50% of 331 patients were enrolled. And after May 2014, the other 50% of 331 patients were enrolled. And so naturally some are think that the Company will have 3 years of data on the second 1/2 of the enrollment period at this time, as today is already 3 years later. However, the patients they enrolled -- if there were more than a few, which I sincerely doubt -- those patients would be considered part of the older protocol analysis. That final SAP, the newest one, includes looking at the effect WBC has on OS.
?
I know that the ASCO enrollment chart shows that enrollment was rising during that May - August 2014 period. What I think they did with that chart is to just plot a few points where enrollment stood and then connected the dots on the graph -- it's not a perfect depiction on enrollment at months in between the plots, IMHO. However, we also know that the newest protocol (which none of us have) aligns with the newest SAP (which none of us have seen).
?
Screening for the new WBC inclusion criteria (in Germany -- which had not begun enrollment yet) -- didn't begin before August 2014. In order to get a fair amount of new SAP screened patients, the Company would need to wait 3 months. Waiting until November 2017 would capture three months of data in the longer tail of 3 years of data, for the second 1/2 of the enrollment period, which then can be used to compare to historical SoC (only about 1/3 of GBM Patients live over 2 years -- when considering all comers (rapids, pseudo, stable disease).
?
Plus if they wait until November 2017, they would have 2 years of enrollment data on all patients in the Phase III trial (last patient enrolled in Nov 2015). BUT at that point they would have at least 3 years of enrollment data for some of the Germany WBC inclusion screened patients. And at least 1 year post survival progression information on the primary endpoint patients (assuming it hit at their estimated month November 2016).
?
Here's the quarters I referenced above. Add 3 years:
?
Q2 2014 -
-- Expanded access arm added on ClinicalRegister, pseudo arm likely closed.
-- Fraunhofer IZI site clears, to enter the trial.
?
Q3 2014
Germany starts screening, August 2014 Fraunhofer IZI batch manufacturing begins.
?
Page 34 of the 2014 Fraunhofer IZI annual report confirms Germany official recruitment started after August 2014, production batch manufacturing started then:
?
http://www.izi.fraunhofer.de/content/dam/izi/en/documents/Publications/Annual_Report_IZI_2015.pdf
?
-- August 2014: Trial Enhancement Announced: New SAP, lower WBC inclusion
-- 348-patient Phase III trial (now 331-patients)
?
Q4 2014
Germany HE patient announcement. Germany enrollment period gearing up, from the inclusion screened patients.
?
Q1 2015
- Additional Germany sites running
-- Canada enters the trial.
?
By the way, I don't think that the Company screened many patients between May -August 2014, as they were in pursuit of a new WBC protocol. Not that it matters much, but the new protocol would need to eliminate the pseudoprogression arm, as that arm was closed and not running in Germany. There is no lower WBC inclusion criteria for those pseudoprogression cohort patients (enrolled separately outside the main arm).
?
When one thinks about the SAP data breakdown for the trial, it should be considered as different enrollment periods, for updates to the protocol depending on the the trial. And so this trial can benefit from having at least 3 years of enrollment data for the latest protocol WBC inclusion patients.
?
1) The Vanguard cohort -- 2008 -2009 -- 140 patient trial - oldest SAP
2) The restart cohort -- 2011 -2012, included a new SAP -- analysis up to 240 patients.
2a) The pseudo period - 2012 - 2014 -- used the restart SAP for the main arm, from the 240-patients trial, but added tertiary, pseudo analysis outside the main arm. A total of 312 patients could be including within the data, but only 240-patients in the main arm.
3)The WBC inclusion period (lower). New SAP for 348-patient trial. This still included Pseudo analysis, outside the study, but recruitment was closed sometime before August 2014 WBC screening opened. (We do not know if this enrollment portion of the trial uses later patents on obtaining higher quality Dendritic cell precursors.)
?
I imagine that the newest SAP from the WBC cohort will be evaluating all the overall trial. But then similiar to the old SAP, from the restart period, it will then look at the cohorts periods, one by one.
?
"Overall cohort (i.e., all data from the beginning of the study, followed by the “vanguard” cohort – i.e. those patients originally enrolled
prior to the hold and followed up to the current time, and a “restart cohort”, i.e., those follow-up since the restart of the clinical trial (i.e., since August 2011). Although the primary view is towards all patients enrolled (i.e. overall
cohort), a consistency of effect across subsets would be the primary intent of these additional cuts.
?
The DMC review will contain, but not be limited to, the following efficacy and safety analyses: patient enrollment/disposition, patient demographics/baseline characteristics, AEs and of SAEs (both tabular and listings as
appropriate), discontinuations from treatment and/or follow-up, survival (overall and progression-free survival) –displayed in graphic (i.e., Kaplan- Meier curves) and in tabular formats, causes specific mortality, and selected
lab data." -- old restart SAP " -- RK on 7-7-17
?
--
?
My post to Flip on 7-8-17:
?
"Flip,
?
The company under enrolled the Phase III trial. That is one of the contributing factors of why the data is delayed (less pool of patients to which to derive events; getting 233 events from 331 patients would take longer than getting 233 events from 348 patients). Whether it's delayed significantly above that under enrollment we do not know. I mention this, as under enrollment has played a factor as to whether or not to "unblind" at progression. The Company is taking no chances in unblinding too early, in light of under enrolling, and they want as much aged data in my opinion.
?
To be clear, waiting has nothing to do with needing more events. Waiting has to do with wanting more follow up time. Waiting provides more perspective on the longer tail data too (inside and outside the main arm of study: pseudoprogression arm). The event count between waiting between August to end of October might not rise much if at all. Instead what rises is the amount of time either post progression that all the original 248 patient might be from their initial progression diagnosis. And what also rises is the length of the the remaining survivors, without events, are alive from their initial diagnosis. We would get the EXACT amount of patients in both arms that are alive 24 months post enrollment. That's a historical known figure (see recent pasted retrospective report below). In general if 33% of standard of care is alive at 2 years (speculated estimated inclusion criteria odds (note the historical for All Comers GBM is roughly 20%: per post-RT-TMZ and post-BEV period (5/2009–12/2013, P3: 19.8 (18.8-20.8)) and this trial has recorded 40% (placebo) and over 50% (vaccine) that would be HUGH.
?
And I haven't had any communications with the company or with Chris. I'm basing my thoughts on my own observations of what additional follow-up time would reveal within the data.
?
As for AA, that's off the table while we wait. The data is continuing to accumulate for progression. And it's just as important that it does. This trial is not a recurrent GBM trial. However, the fact the investigators are blind to the initial cohort that the crossover patients were in it essentially makes it a standardized recurrent entry for the placebo cohort. Depending on how the data for survival post progression comes out came for placebo 60/40 yes/no to crossover, we could end up seeing an approval for initiation of vaccine for recurrent GBM as well. And sure, they may have enough data for stopping the trial with less follow-up on the last of the 248 progression patients, but I sincerely believe they are considering holding out as long as they can.
?
As for "be careful what you wish for", I find that to be an interesting comment if taken at face value reference to regulators. (Unless between the lines you meant that for stock issuances, which depending on how the cash position plays out (if they are able to raise from prior issuance (I.e, stock offerings to insiders) could influence their interest to wait or not, obviously. But, I think the money situation will work itself out in the interim.) Seriously, there is no foul if the either the Company or the FDA ends the trial at 233 events. They would just have less follow-up time data before the lock, and more afterwards. We will just have to see how it plays out. I personally think if UCLA's data comes out in strong favor for one of the three combinations (without or with TLR3 cream or with TLR9 cream) they may fully encourage NWBO about having the trial go until two years of enrollment on all patients. To see the amount of patients who are 27 months out from initial surgery both within the non-progression and the recurrent patient population in the information that they review, would be very useful in making the determine whether or not to provide full approval.
?
As for NW Bio providing us insight as to when the data is "officially" being analyzed, I'm not sure they will. We saw how that information was manipulated. And we all know that the trial is likely at 233 events as we speak. Once they get the unblinded study news, they only have 3 days to tell us it's being analyzed. They do not have to tell us that they are waiting a couple of months to have more follow-up time of remaining survival. We can read about it in SEC statements from 10Q that the Company remains blinded as the Phase III trial nears completion. We reported that the patient data from all the 80 clinical sites continues to carefully scrubbed and crossed checked by the CROs who will report that final information to the DMC who is monitoring the trial and is also blind to patients initial cohorts. The Phase III data will be locked and analyzed shortly thereafter once that process is complete. Or something like that...
?
I sincerely do not mind if you disagree with my speculative thoughts on whether they will or will increase the follow-up time. It's just something I'm considering that they might decide to do given how close they are to getting significant data on all enrolled patients. I respect your opinion to respectively disagree with mine, as always. Best,
?
Historical GBM data paper, I referenced:
?
Survival benefit of glioblastoma patients after FDA approval of temozolomide concomitant with radiation and bevacizumab: A population-based study
Ping Zhu1,2, Xianglin L. Du1, Guangrong Lu2 and Jay-Jiguang Zhu2??Received: November 23, 2016 Accepted: March 20, 2017 Published: April 12, 2017?
ABSTRACT
Few population-based analyses have investigated survival change in glioblastoma multiforme (GBM) patients treated with concomitant radiotherapy-temozolomide (RT- TMZ) and adjuvant temozolomide (TMZ) and then bevacizumab (BEV) after Food and Drug Administration (FDA) approval, respectively. We aimed to explore the effects on survival with RT-TMZ, adjuvant TMZ and BEV in general GBM population based on the Surveillance, Epidemiology, and End Results (SEER) and Texas Cancer Registry (TCR) databases. A total of 28933 GBM patients from SEER (N = 24578) and TCR (N = 4355) between January 2000 and December 2013 were included. Patients were grouped into three calendar periods based on date of diagnosis: pre-RT-TMZ and pre-BEV (1/2000–2/2005, P1), post-RT-TMZ and pre-BEV (3/2005–4/2009, P2), and post-RT-TMZ and post-BEV (5/2009–12/2013, P3). The association between calendar period of diagnosis and survival was analyzed in SEER and TCR, separately, by the Kaplan-Meier method and Cox proportional hazards model. We found a significant increase in median overall survival (OS) across the three periods in both populations. In multivariate models, the risk of death was significantly reduced during P2 and further decreased in P3, which remained unchanged after stratification. Comparison and validation analysis were performed in the combined dataset, and consistent results were observed. We conclude that the OS of GBM patients in a “real-world” setting has been steadily improved from January 2000 to December 2013, which likely resulted from the administrations of TMZ concomitant with RT and adjuvant TMZ for newly diagnosed GBM and then BEV for recurrent GBM after respective FDA approval.
?
Quote from the research report:
?
"Our findings suggested that OS and 1-year survival rate improved significantly across calendar period of diagnosis in SEER, TCR and the combined dataset, except that the improved survival within P3 compared
to P2 in TCR was not significant (P = 0.833). There was no survival benefit observed in 1-year or 2-year survival rate between P2 and P3 in TCR, nor 2-year survival rate between P2 and P3 in SEER and in the combined dataset. One explanation for the reduced survival in P3 comparing to P2 in TCR could be related to the lower proportion of patients received radiation during P3. The proportion of patients treated with radiation therapy in TCR decreased from 80.8% to 69.8% and to 60.1% in P1, P2, and P3, successively. However, when we did a sub-analysis of the extended follow-up of the TCR dataset up to May 2015, we found that the difference in survival between P2 and P3 was significant (Supplementary Figure 1 and Supplementary Table 2). This finding indicated that the survival benefit of BEV could be observed when given longer period of follow- up and that may compensate the loss of benefit with lower radiation proportion in P3."
?
Open PDF to see Table 2
?
Table 2: Median OS, 1-year and 2-year survival rate by calendar period of diagnosis in SEER, TCR and the combined dataset*
Survival statistics
Calendar Period of Diagnosis
Jan 2000 - Feb 2005 (P1)
Mar 2005 - Apr 2009 (P2)
May 2009 - Dec 2013 (P3)
Total?(P1, P2 and P3)
P1 vs P2
P1 vs P3
P2 vs P3
?
http://www.impactjournals.com/oncotarget/index.php?journal=oncotarget&page=article&op=download&path%5B%5D=17054&path%5B%5D=59124
?
Older SEERs data on GBM:
?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4954825/pdf/fpubh-04-00151.pdf
?
http://scholarworks.waldenu.edu/cgi/viewcontent.cgi?article=1267&context=dissertations
" --RK on 7-8-17
The 7-8-17 post all the way at the top of this post followed Flips response to this 7-8-17 above.
The patients that show lymphopenia at baseline were not the patients that Beach referenced! He is speaking of patients who were not sick yet at baseline. And yes, it can be up to 45% of GBM. A percentage naturally will fit the description of evidence disease progression or illness at baseline. But, those that he referenced being randomized were those that he expected would be spotted after enrollment, sometime during adjuvant chemotherapy. And I paste facts because you keep missing his point. From Ellsworth paper:
"Lymphopenia is common and can be severe, with up to 45% of GBM patients developing grade III-IV lymphopenia 2 mo after completing this multi-modal therapy. Treatment-induced lymphopenia generally persist for at least one year after combined RT/TMZ therapy, and prior evidence suggests that lymphopenia may be detected as long as 10 y after focal external-beam RT.3,4"
Which is off the paper that NW Bio referenced in their August 2014 PR:
Sustained CD4+ T cell-driven lymphopenia without a compensatory IL-7/IL-15 response among high-grade glioma patients treated with radiation and temozolomide
Susannah Ellsworth, Ani Balmanoukian, [...], and Charles G Drake
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4004618/
Agree.
They specifically referenced two outside studies in that PR, (See, for example: Grossman et al., 2011: Clin Cancer Res 17(16):5473-80; Ellsworth et al. 2014: Oncoimmunology 3(1):e27357. Epub 2014 Jan 3) one of which I posted the abstract related to it.
Here is that portion of the PR:
"The new variable involves the level of certain white blood cells in GBM patients when they finish the 6 weeks of daily radiation to the brain which is part of the current standard of care treatment. Important recent research has found that as many as 40% of GBM patients have such severely depressed white blood cell counts following radiation that their level is comparable to the level at which AIDs patients are put on continuous antibiotic treatments, prophylactically. Further, this research has found that these GBM patients’ white blood cell counts do not recover with the passage of time. See, for example: Grossman et al., 2011: Clin Cancer Res 17(16):5473-80; Ellsworth et al. 2014: Oncoimmunology 3(1):e27357. Epub 2014 Jan 3.
Most importantly, the recent research has identified a major impact on these GBM patients’ Overall Survival (OS): the variable relating to severe depression of white blood cell counts can make a difference of 6 months in OS. As a comparison, the standard of care drug for GBM, Temodar, only makes a difference of 10 weeks in OS. If the statistical analysis of GBM trial results does not take account of such a major variable, the overall trial results could be significantly skewed." -- NW Bio PR
What Exwannabe is not comprehending is that it's a condition that shows up in time -- typically 2 months. It is not that is always evident at the Post RT MRI scan, which is where this trial's baseline was. Off Ellsworth 2014 paper:
" Introduction
Glioblastoma multiforme (GBM) is the most common primary brain tumor and is responsible for approximately 70 000 deaths worldwide each year.1 The current standard of care for GBM patients includes maximal debulking surgery followed by 6 wk of concurrent temozolomide (TMZ) and radiation therapy (RT) plus 6 mo of maintenance therapy with TMZ as a single therapeutic agent.2 Patients also frequently require the management of symptomatic cerebral edema with corticosteroids. This comprehensive GBM treatment course is unfortunately associated with significant hematologic toxicity, including thrombocytopenia and lymphopenia. Lymphopenia is common and can be severe, with up to 45% of GBM patients developing grade III-IV lymphopenia 2 mo after completing this multi-modal therapy. Treatment-induced lymphopenia generally persist for at least one year after combined RT/TMZ therapy, and prior evidence suggests that lymphopenia may be detected as long as 10 y after focal external-beam RT.3,4"
Here is the full paper:
Sustained CD4+ T cell-driven lymphopenia without a compensatory IL-7/IL-15 response among high-grade glioma patients treated with radiation and temozolomide
Susannah Ellsworth, Ani Balmanoukian, [...], and Charles G Drake
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4004618/
Seriously, that's a ridiculous question. Severe lymphopenia is associated with infections. Patients that were too sick to enter the trial were not enrolled. Patients whose CD8:CD4 ratio counts fall significantly to "severe" levels below by baseline would automatically be a safety concern. They don't enroll patients who are on the verge of toxic shock syndrome. But Beach never said they enrolled patients who were too sick for entry. Not once. His stance is that radiation affects circulating lymphocytes (mostly T lymphocytes verses B) and indirectly alters the bone marrow function, leads to immunosuppression. T helper cells recovery over time is affected. He understands that the company enrolled patients whose T-helper cells (CD4) that started out fine, got hit badly (not at infection point (CD4 well below 200 would be a trigger for body being unable to fight infection) and that do not recover in time. As such, the company is now statistically monitoring the affect of WBC counts have on OS.
The Etiology of Treatment-related Lymphopenia in Patients with Malignant Gliomas: Modeling Radiation Dose to Circulating Lymphocytes Explains Clinical Observations and Suggests Methods of Modifying the Impact of Radiation on Immune Cells
Susannah Yovino, Lawrence Kleinberg, [...], and Eric Ford
Additional article information
Abstract
Purpose
Severe treatment-related lymphopenia (TRL) occurs in 40% of patients with high grade gliomas (HGG) receiving glucocorticoids, temozolomide, and radiation. This occurs following radiation, persists for months, and is associated with reduced survival. As all three treatment modalities are lymphotoxic, this study was conducted to estimate the radiation dose that lymphocytes receive passing through the radiation field and if this could explain the observed TRL.
Materials and Methods
A typical glioblastoma plan (8-cm tumor, 60 Gy/30 fractions) was constructed using the Pinnacle™ radiation planning system. Radiation doses to circulating cells (DCC) were analyzed using MatLab™. The primary endpoints were mean DCC and percent of circulating cells receiving ≥0.5 Gy. The model was also used to study how changes in target volumes (PTV), dose rates, and delivery techniques affect DCC.
Results
The modeling determined that while a single radiation fraction delivered 0.5 Gy to 5% of circulating cells, after 30 fractions 99% of circulating blood had received ≥0.5 Gy. The mean DCC was 2.2 Gy and was similar for IMRT, 3D-conformal techniques, and different dose rates. Major changes in PTV size affected mean DCC and percent of circulating cells receiving ≥0.5 Gy.
Conclusions
Standard treatment plans for brain tumors deliver potentially lymphotoxic radiation doses to the entire circulating blood pool. Altering dose rates or delivery techniques are unlikely to significantly affect DCC by the end of treatment. Novel approaches are needed to limit radiation to circulating lymphocytes given the association of lymphopenia with poorer survival in patients with HGG.
INTRODUCTION
Marked reductions in the lymphocyte count are common following therapy for malignant glioma and can have significant clinical consequences. Severe lymphopenia in this setting has been associated with serious opportunistic infections (1, 2). Perhaps more importantly, emerging data from patients with malignant gliomas and pancreatic cancer demonstrate that patients with severe treatment-induced lymphopenia have significantly worse survival and die from early tumor progression (3, 4).
Interest in treatment-induced lymphopenia was sparked by a cluster of Pneumocystis jiroveci pneumonia cases, associated with extremely low CD4 counts, in patients treated with radiation therapy and steroids (without chemotherapy) for brain tumors (1). This led to a prospective study of serial total lymphocyte and CD4 counts in brain tumor patients receiving radiation and corticosteroids. This demonstrated that CD4 counts were >450/µL in all patients before starting therapy, but that during treatment, approximately one-fourth developed CD4 counts <200/µL (5). After temozolomide became standard therapy, a second study prospectively evaluated serial lymphocyte counts in high-grade glioma patients receiving radiation and temozolomide. In this study, over 40% of patients developed CD4 lymphocyte counts under 200/µL 2 months after completing treatment and over 70% had CD4 counts under 300/µL. Participants with CD4 counts <200/µL 2 months after initiating radiation and temozolomide had significantly worse overall survival than those with higher CD4 counts (4).
Although patients with malignant glioma receive a triad of lymphotoxic agents (corticosteroids, temozolomide, and radiation therapy), radiation may play an important role in lymphopenia. Lymphopenia following radiation therapy was first described in the early 20th century, only a few years after x-rays were discovered, and has since been documented to occur after either external beam radiotherapy or brachytherapy directed to virtually every part of the body (6). Radiation can induce lymphopenia regardless of whether chemotherapy or steroids are given concurrently or whether bone marrow or lymphatic tissue is included in the field. For example, irradiation of the brain, which contains neither bone marrow nor lymphatic tissue, can cause over a 60% reduction in the lymphocyte count (7). Furthermore, even radiation of extracorporeal blood in patients undergoing renal dialysis can result in profound and durable lymphopenia (8). These observations suggest that irradiation of circulating lymphocytes may contribute to the development of radiation-induced lymphopenia.
In order to further evaluate the etiology of treatment related lymphopenia, we used modern radiation planning software and mathematical modeling to estimate the dose of radiation delivered to circulating blood cells during a typical 60-Gy radiation treatment course for glioblastoma, a nonmarrow, nonlymphatic containing site. We subsequently used our model to investigate changing treatment parameters that could affect the dose to circulating lymphocytes such as target volume, dose rate, and radiation delivery technique.
METHODS AND MATERIALS
Model design
A normal CT scan of the head was loaded into the treatment planning system, and a spherical tumor volume centered in the falx superior to the lateral ventricles was created. Dose to the brain was calculated using the Pinnacle™ radiation treatment planning system (Version 9.0, Philips Inc., Madison, WI). The entire brain was also delineated as a region of interest. A treatment plan was then generated resulting in a calculation for the dose, voxelized in 0.4 mm3 voxels. Using Pinnacle scripts, the three-dimensional dose grid for the brain was extracted and analyzed using an in-house program written in MatLab (version R2011b, Mathworks, Inc., Natick, MA). Several treatment plans were considered. Unless otherwise noted, the results below are for a four-field conformal plan that treats the PTV to a homogeneous dose. For comparison, a 3D-conformal plan using wedges and five fields was calculated, as was an intensity-modulated radiotherapy (IMRT) plan using a 2-cm wide sliding window and an inverse planning approach. In each case, the dose was extracted beam-by-beam (for IMRT, segment-by-segment).
The model uses as its input the dose to the brain during a normal course of RT then calculates the associated dose to blood circulating through this radiation field using the following assumptions: (a) 16% of cardiac output goes to the brain; (b) blood flow is evenly distributed, flowing inferiorly to superiorly through the dose grid; (c) blood flow velocity is 10 mm/s; and (d) the total volume of blood is 5 L (9). Furthermore, we assumed that blood could pass through the beam multiple times but that during the duration of a single beam and/or segment, blood does not reenter the treatment field. This latter assumption breaks down with beam times that are longer than the 30-s heart-to-heart circulation time, but this only affected our calculations at the very lowest dose rates. Between beams and between treatment fractions, the blood was considered to reenter the treatment field, and the cumulative dose was calculated via a convolution of blood pool dose histograms. Within the MatLab code, we could control the dose rate, expressed in monitor units (MU)/min and the dose delivered per fraction by scaling the total number of monitor units delivered.
The mean dose to the total blood pool was calculated, and the total volume of blood receiving greater than 0.5 Gy was quantified. This latter dose cutoff was chosen based on in vitro data on lymphocyte radiosensitivity that showed a D10 (dose required to reduce the surviving lymphocyte population to 10% of initial values) of ~3 Gy, a D50 of ~2 Gy, and a D90 of ~0.5 Gy (10). The model was used to calculate these dose parameters for PTVs of two sizes: 2-cm diameter (4.2 cm3 volume) and 8-cm diameter (268 cm3 volume), plans administered at varying dose rates (300, 600, and 1,200 MU/min), and radiation techniques (IMRT and 3D-conformal).
RESULTS
Using the technique described above, we modeled the radiation dose delivered to circulating blood during a typical partial intracranial field in an effort to determine the role of radiation in the observed lymphopenia. The effects on dose to circulating blood after varying different treatment-related parameters in the model are described below.
Dose to circulating blood pool as a function of number of fractions
The model indicates that a single fraction (2 Gy) delivered ≥0.5 Gy to 4.6% of the total blood pool. After 10 fractions (20 Gy), 61.5% of the blood pool received ≥0.5 Gy, and after 20 fractions (40 Gy), 92.2% of the blood pool received ≥0.5 Gy. By the end of a typical 60-Gy radiation treatment plan to the brain (2 Gy × 30 fractions to an 8-cm diameter field), 98.8% of all circulating blood receives at least 0.5 Gy. Mean dose to the blood pool was 2.2 Gy for a 60-Gy course with a PTV of 268 cm2 at a dose rate of 600 MU/min. According to the model, as the total dose and the number of fractions increase, the percentage of blood receiving ≥0.5 Gy increases rapidly. Figure 1 shows the percentage of the blood pool receiving at least 0.5 Gy in terms of the number of fractions and total dose administered (8-cm diameter/268-cm3 PTV, dose rate 600 MU/min).
Figure 1
Figure 1
Percent of blood receiving >0.5 Gy plotted against total dose/number of fractions administered. Dose rate = 600 MU/min; PTV diameter = 8 cm (volume = 268 cm3); 2 Gy/fraction.
Dose to circulating blood pool as a function of dose rate
Increasing the dose rate effectively shortens the treatment time, thereby reducing the number of lymphocytes passing through the radiation field. An increased dose rate (i.e., a shorter beam time), therefore, increases the percentage of the blood pool receiving a low dose. As shown in Figure 2, the percent of the blood pool receiving ≥0.5 Gy was less with higher dose rates at the beginning of treatment. However, once 30 fractions have been administered, the percentage of blood receiving ≥0.5 Gy approached 100% for all dose rates tested. Only minor dose-rate-dependent differences were observed when using the mean dose as an endpoint; a mean dose of 2.2 Gy was calculated for both 600 MU/min and 1,200 MU/min plans, while a mean dose of 3.1 Gy was calculated for 300 MU/min plans.
Figure 2
Figure 2
Percent of blood receiving >0.5 Gy with varying dose rates (300, 600, and 1,200 MU/min). PTV diameter = 8 cm (volume = 268 cm3); 2 Gy/fraction.
Dose to circulating blood pool as a function of treatment technique
The model was used to compare dose to the blood pool for IMRT and 3D-conformal treatment techniques. By the end of a 60 Gy plan administered in 30 2-Gy fractions to an 8-cm diameter PTV, no differences were observed in the mean dose to the blood pool or in the proportion of blood receiving ≥0.5 Gy. Mean dose to the blood pool was 2.4 Gy for the tested 3D-conformal plan and 2.7 Gy for the tested IMRT plan. In both cases, nearly all of the blood received at least 0.5 Gy after 30 2-Gy fractions.
Dose to circulating blood pool as a function of target volume size
Major differences in target volume size had a significant impact on the mean dose to the blood pool, particularly at higher doses, as shown in Figure 2. For a 60-Gy plan administered in 30 fractions, the mean blood dose was 2.2Gy for an 8-cm diameter PTV but was 0.3 Gy for a 2-cm diameter PTV. These data were also analyzed in terms of the percent of blood receiving at least 0.5 Gy, as shown in Figure 3. Smaller PTV size decreased CBD. For a 2-cm diameter PTV (volume = 4.2 cm3) receiving 60 Gy, 15% of the blood received ≥0.5 Gy; however, for the 8-cm diameter PTV (volume = 268 cm3), 99% of blood received ≥0.5 Gy.
Figure 3
Figure 3
Percent of blood receiving >0.5 Gy with varying PTV sizes of 2-cm diameter (4.2 cm3) and 8-cm diameter (268 cm3). Dose rate = 600 MU/min; 2 Gy/fraction.
DISCUSSION
This study demonstrates that the entire circulating blood pool receives a potentially lymphotoxic dose of radiation during a typical external beam radiation treatment course for glioblastoma. Our model estimates the mean dose to circulating lymphocytes is approximately 2 Gy, and that nearly all the circulating blood receives at least 0.5 Gy. Even though this radiation dose may be relatively low compared to the total dose received by the targeted tumor, it may still be sufficient to eliminate a large number of lymphocytes. These cells are among the most radiosensitive in the body, with a D10 of approximately 3 Gy and a D50 of about 2 Gy (10). The model also demonstrates that neither increased dose rate nor choice of treatment technique (IMRT vs. 3D-CRT) significantly changes the radiation dose to circulating lymphocytes. However, major decreases in target volume size would result in a significant decrease in the dose to circulating lymphocytes.
It is likely that irradiation of the circulating blood plays a significant role in the pathophysiology of radiation induced lymphopenia. Although irradiation of the bone marrow or lymph nodes may contribute to lymphopenia, irradiation of structures such as the brain that contain neither lymphatic tissue nor bone marrow can also cause lymphopenia. Irradiation of the circulating blood alone via a radioactive source mounted within a dialysis unit can also cause lymphopenia, with decreases of over 60% in the circulating lymphocyte count persisting for over 2 years in patients undergoing irradiation of the circulating blood in preparation for renal transplant (8).
Lymphocytes are well known to be radiosensitive, and are the only highly radiosensitive nonproliferative tissue (11). Even low-dose total body radiation exposures cause decreases in the circulating lymphocyte count and detectable cytogenetic abnormalities in lymphocytes that can persist for up to 10 years following radiation exposures (12). The exact mechanisms underlying lymphocyte radiation sensitivity are not well understood. Lymphocytes appear to have a decreased DNA repair capacity, possibly related to the active DNA recombination that aids in the development of an individual’s immune repertoire (12). B-lymphocytes are slightly more radiosensitive than T-lymphocytes, and naïve T lymphocytes appear to be more radiosensitive than memory cells (13). The difference in radiosensitivity between naïve and memory T-cells is a potentially important phenomenon in terms of immune control of cancer. Malignant tumors represent a novel challenge to an individual’s immune system, and depleting the T cell population that responds to novel antigens (i.e., naïve T cells) may contribute to decreased immune surveillance of malignant cells and capacity for tumor control.
Lymphocytes are important mediators of the immune response to cancer, and treatment-induced lymphopenia is associated with worse disease-specific survival in high-grade gliomas and in both resectable and locally advanced/unresectable pancreatic cancer (3, 4, 14). Most current immunotherapy strategies rely on modulation of lymphocyte activity for efficacy, and it is plausible that treatment-induced lymphopenia could impede efforts to enhance the immune response to malignancy. For example, sipuleucel-T, the recently approved prototypic cancer cell vaccine, relies on stimulating a patient’s own T lymphocytes to attack prostate cancer cells using a chimeric antigen consisting of prostatic acid phosphatase (PAP) fused to the immune stimulant G-CSF (15). Ipilimumab, which inhibits the lymphocyte regulatory receptor CTLA-4, has been found to prolong survival in patients with metastatic melanoma and is the subject of considerable clinical interest. Preliminary data suggest that the efficacy of ipilimumab may be dependent on circulating lymphocyte count, with lymphopenic patients being significantly less likely to respond to ipilimumab (16).
In this study, we sought to identify alterations in radiation treatment techniques that could modify the dose to circulating blood. Changing the dose rate seemed to be a promising technique, as higher dose rates decrease treatment time and thus the length of time that circulating blood would be exposed to the radiation beam. However, altering the dose rate had a limited ability to spare circulating lymphocytes. Although higher dose rates were associated with lower dose to the circulating lymphocytes early in treatment, these differences disappeared by the 10th fraction for changes in the dose rate. This study was also designed to investigate whether IMRT would increase the circulating blood dose compared with three-dimensional conformal radiation therapy (3D-CRT). IMRT creates highly conformal plans at the target but usually results in a larger area of tissue receiving low doses of radiation compared with 3D-CRT (17). However, the model showed that doses to circulating blood appeared to be similar regardless of whether IMRT or 3D-CRT was used. Changing the target size appears to have a larger impact on the dose to circulating blood; a 2-cm diameter PTV (volume = 4.2 cm3) resulted in 15% of the blood receiving ≥0.5 Gy, whereas an 8-cm diameter PTV (volume = 268 cm3) resulted in 99% of the blood receiving the potentially lymphotoxic dose of ≥0.5 Gy. However, such small PTVs may not be appropriate for many patients.
The data presented above on field size, dose rates, and delivery techniques suggest that changing the delivery of radiation in an effort to spare circulating blood from exposure to focal radiation will be challenging. However, our data also suggest several strategies to minimize the radiation dose to circulating blood. One of the most promising is hypofractionation, or increasing fraction size while decreasing the total number of fractions given. In the mid-1970s, MacLennan et al. demonstrated that lymphocyte count was inversely proportional to the number of fractions administered when the total radiation dose was held constant (7). This finding is intuitive and is corroborated by our modeling study results. Because of the clear dependence of dose to circulating blood on PTV size and the number of fractions given, combining hypofractionated treatment plans with smaller target volumes (as in stereotactic radiosurgery) is a promising approach to decrease the dose to circulating blood. A detailed model of hypofractionated treatments would be required to confirm the benefit of such treatments because of the long beam-on times, which would require consideration of reentry of blood into the treatment field during therapy. These strategies also may not be suitable for all patients due to large primary tumor sizes or other factors, including the importance of maintaining appropriate chemotherapy dosing schedules. Other strategies could focus on radioprotective agents or immune reconstitution.
Limitations of this model include the fact that estimating the dose to circulating blood is challenging and can be affected by estimates of blood flow velocity, volume of blood within the target, and recirculation. The model makes simplifying assumptions about all of these factors. The analysis was also unable to account for the continuous regeneration of circulating blood cells from the stem cell compartment (which may be impacted by long-term temozolomide therapy) and the fact that lymphocytes are continually redistributed among different organs, including the spleen, bone marrow, circulating blood, and lymph nodes. Although some regeneration of the lymphocyte population may occur between radiation fractions, it is important to note that radiation-induced lymphopenia is prolonged and can last up to 10 years after irradiation. Other factors, such as abnormal cytokine responses to lymphopenia, may deplete an individual’s capacity to regenerate the lymphocyte pool after radiation. Further investigation into the pathophysiology of radiation-induced lymphopenia is clearly needed. Finally, variations in treatment technique (such as the use of a vertex field) may result in an expanded low-dose radiation field and affect lymphocyte counts. Nevertheless, we believe that this model provides an important basis for further investigations. Because standard radiation fields result in lymphotoxic doses of radiation to circulating lymphocytes and because posttreatment lymphopenia is associated with opportunistic infections as well as poorer disease-specific survival in cancer patients, further studies aimed at evaluating new ways to limit the effect of radiotherapy on circulating lymphocytes are critical.