Transcript of FDA Teleconference on Benlysta:
Capt Beth Fritsch:
Good afternoon and welcome to the health professional briefing on Benlysta. We have identified your organization as a
healthcare professional organization representing practitioners and professional leaders who may be interested in this topic.
We hope this briefing will provide you an opportunity to ask questions directly to FDA staff most familiar with this new
product. I am Beth Fritsch from the office of special health issues and I'll be your moderator for this teleconference.
Before we get started I would like to review how we will proceed. Today's discussion will focus on the primary efficacy and
safety findings for Benlysta which will be lead by Dr. Sarah Okada a clinical team leader from the division of pulmonary,
allergy and rheumatology drug products or DPARP. Immediately following the discussion, there will be a question and
answer session. At the conclusion of the question and answer session I will give concluding remarks. Your phones will be
on mute during the presentation. Prior to the start of the question and answer session the operator will provide instructions
on how to ask questions to FDA's experts. Please be aware this teleconference is being recorded for playback so please
state your name prior to joining the conversation. Speak clearly and directly into your telephone at all times.
Today I have several colleagues from the DPARP with me. As I mentioned Dr. Sarah Okada, the clinical team leader for
Benlysta will be providing you with an overview of the efficacy and safety findings. Dr. Badrul Chowdhury is also with us
today. He's the division director of DPARP. We also have Dr. Sally Seymore who is the deputy director for safety and
DPARP. Thank you for joining us today. At this time Dr. Okada will begin with her overview of Benlysta.
Dr. Okada:
Thank you Captain Fritsch. Good morning everyone. So as mentioned I'm going to just start off by giving a brief overview of
some of the background facts of Benlysta as we reviewed them.
As you know or as you probably know Benlysta or belimumuab is a human IgG1 lambda monoclonal antibody with a new
target, a molecule called a B lymphocyte stimulator, or BLYS, also known in the scientific literature as B cell activating factor
or BAFF. This molecule plays a role in B cell selection, survival and immunoglobulin class switching and by binding BLYS,
or BAFF, Benlysta may act by inhibiting survival of B cells including abnormal B cells and by reducing the differentiation of
those B cells into immunoglobulin producing cells.
The approved dose of Benlysta is 10 mg/kg given at 2 week intervals for the first 3 doses and at four week intervals
thereafter given by intravenous infusion over one hour. It's available as a lyophilized powder in vials for reconstitution with
sterile water for injection and further dilution in normal saline. Dextrose solutions are incompatible with Benlysta. The
limitations of use of this product pertain to primarily gaps in the data given to us by the clinical trials. The efficacy of Benlysta
has not been evaluated in patients with severe active lupus nephritis or severe active central nervous system lupus.
Benlysta has also not been studied in combination with other biologics such as rituximab or IV cyclophosphamide. So the
use of Benlysta is not recommended in these situations primarily just because we do not have the data to suggest that it's
safe and efficacious.
In terms of the basic clinical pharmacology and drug-drug interactions for the product, its pharmacokinetics are typical
monoclonal antibody pharmacokinetics. The terminal half like is approximately 19 days. Age, gender and race did not
appear to effect belimumab pharmacokinetics in the study population. However no formal drug-drug interaction studies have
been conducted with belimumab.
To turn to the efficacy of Benlysta. The initial efficacy was evaluated in a randomized controlled trial called LBSL02. This
involved 449 patients who received 1, 4 or 10 mg/kg of Benlysta plus standard of care and also contained a control group of
patients receiving placebo plus standard of care. The duration of that study was 52 weeks. Patients were required to have a
SELENA-SLEDAI score of greater than or equal to 4 at baseline and a history of autoantibodies, specifically ANA or
anti-dsDNA. However 28% of that study population was autoantibody negative at baseline. The co-primary endpoints for
that study were % change in SELENA-SLEDAI at week 24 and time to first flare over 52 weeks. No significant difference
between treatment groups were observed for any of the primary or secondary endpoints in that study. However exploratory
analysis done on the data from this study suggested that a treatment effect may have been present in the subgroup of
patients who were autoantibody positive and it also helped to inform the development of a new endpoint that the sponsor
proposed, an endpoint called the SLE responder index or SRI. The SRI is a composite endpoint where a patient is defined
as a responder if they have at least a four point reduction in SELENA-SLEDAI score compared to baseline and no
Page 3
worsening in physician global assessment and no new BILAG A or two new BILAG B organ domain scores at the time of the
primary endpoint assessment.
The sponsor conducted, subsequently, two pivotal trials, trials 1056 and 1057 with those learnings in mind. They enrolled
only SLE patients who were autoantibody positive at baseline and they used the proportion of SRI responders at week 52 as
the primary endpoint.
Trial 1056 enrolled 819 patients primarily in North America and Europe and had a longer control period of 76 weeks to
assess for possible delayed treatment effect in case no effect was seen at the week 52 timepoint. Trial 1057 enrolled 810
patients primarily in Asia and Latin America. The trials had predominately at baseline musculoskeletal and mucocutaneous
manifestations, approximately 82% from mucocutaneous, from musculoskeletal it was 73% in trial 1056 and 59% of patients
in 1057.
The more robust efficacy results were observed in trial 1057. A summary of the primary endpoint results at week 52 showed
that a higher proportion of Benlysta patients compared to placebo treated patients achieved a response as defined by the
SRI at week 52. For trial 1056 it was approximately 43% in the Benlysta 10 mg/kg group vs 34% of patients in the placebo
group. And in trial 1057 it was 58% of patients in the Benlysta 10mg/kg group compared to 44% of patients in the placebo
treated group.
So these two trials did provide independent substantiation of a difference in treatment effect with respect to the SRI and the
differences were statistically significant. Looking at the components of the SRI, the primary efficacy results appeared to be
primarily driven by the four point reduction in the SELENA-SLEDAI. However the no worsening in the physician global
assessment and the no new BILAG A or two new BILAG B domain scores were also consistent.
As you know there were a couple of concerns that we raised at the advisory committee in November of 2010. One of these
concerns was a possible lack of effect or reversal of treatment effect in patients of African American or African heritage.
Now this is a post-hoc subgroup analysis so we interpret the results with caution however in both studies the small subgroup
of patients that were in this racial subgroup experienced a reversal of effect. In trial 1056, which had the most African
heritage patients, 39% of patients in the placebo group responded vs 33% in the Benlysta 10 mg/kg group. The African
heritage subgroup in trial 1057 was even smaller, however similarly the placebo patients had a higher rate of response: 64%
vs. 46% for Benlysta 10 mg/kg. Again I caution that none of these results are intended to connote a statistical significance.
Another concern was the potential lack of efficacy at week 76 demonstrated in study 1056. At week 76 approximately 32%
of placebo treated patients had an SRI response vs. 39% of Benlysta 10 mg/kg treated patients. When comparing these
results to the week 52 results what's apparent is that the rate of response, or the loss of response, was approximately
similar across the treatment groups so it was concerning that there weren't fewer patients who were losing their response in
the Benlysta group. However again these results are primarily descriptive and not intended to connote statistical
significance.
Finally the last two efficacy issues I wanted to discuss are the clinically important aspects of prednisone reduction and a
reduction in incidence of flares. In trials 1056 and 1057 there was an endpoint of evaluating the proportion of patients with
prednisone reduction of at least 25% from baseline to less than 7.5 mg/day. This endpoint was accepted as a clinically
meaningful reduction in prednisone and the assessment compared baseline prednisone use to the last 12 weeks of the
study prior to the primary endpoint assessment. In trial 1056, well, I would just say in both trials there was a suggestion of
prednisone reduction. 13% of placebo patients achieved this level of reduction vs. 17-19% of the belimumab treated patients
in both studies. Looking at the flip side of that coin we also evaluated patients who required an increase in prednisone of
that amount and the two studies were not consistent in that regard with more Benlysta treated patients requiring an increase
in prednisone in study 1056 vs. trial 1057 where Benlysta treated patients actually did require less incremental increase in
prednisone.
Finally for the flare results, again there was a trend although not statistically significant of a higher proportion of placebo
treated patients having a severe flare during the course of the study vs the Benlysta treated groups. In trial 1056 that was
24% of placebo treated patients who experienced severe flare vs. 18% of the Benlysta 10 mg/kg group. And in trial 1057
that was 23% of placebo treated patients vs. 14% of the belimumab 10 mg/kg group.
I'm going to turn now to a brief discussion of the safety issues. For most safety issues there was a numeric imbalance going
against Benlysta in the pivotal trials, actually in both trials 1056, 1057 and trial LBSL02. These safety issues were placed in
Page 4
the warning section of the approved product label and I'm just going to describe them briefly.
First is mortality. Out of 2133 patients in the 3 clinical trials, 14 deaths occurred in the controlled period, 11 of those deaths
were in the Benlysta groups and 3 of those were in the placebo group. Etiologies included infection, cardiovascular disease
and suicide. For serious infections similarly, serious infections occurred in approximately 6% of the Benlysta groups vs 5.2%
of placebo treated patients. The most frequent serious infections included pneumonia, urinary tract infection, cellulitis and
bronchitis. Infections resulting in death occurred in 0.3% of Benlysta patients vs. 0.1% of placebo patients.
For hypersensitivity reactions including anaphylaxis, again there was a numeric imbalance going against Benlysta with 13%
of Benlysta treated patients experiencing a hypersensitivity reaction on the day of infusion vs. 11% of patients receiving
placebo. Anaphylaxis was observed in 0.6% of patients receiving Benlysta and 0.4% of patients receiving placebo.
Approximately 13% of study patients received premedication. With respect to infusion reactions, again an imbalance with a
respect to 17% of patients on Benlysta vs 15% of patients on placebo.
Finally for depression and suicidality, I'd like to note that psychiatric events overall occurred in a higher proportion in
Benlysta treated patients.16% vs. 12% of placebo treated patients. And this inbalance was related primarily to depression
related events. Serious psychiatric events were reported in 0.8% of patients on Benlysta vs. 0.4% of patients on placebo.
Serious depression was reported in 0.4% of patients receiving Benlysta and 0.1% of patients receiving placebo. Two
suicides were reported for patients receiving benlysta in the control period vs. none in the placebo groups.
Finally a brief summary of the post marketing requirements that have been enacted. The first post marketing requirement is
a randomized placebo controlled clinical trial in 5,000 patients which is intended to assess the long term safety profile and
adverse events of special interest including mortality, malignancy, serious and opportunistic infections and depression and
suicidality. Second is a pregnancy registry. Third is a trial to assess the effects of Benlysta on vaccinations. Fourth is a
prerequired pediatric SLE study. And fifth is a product requirement to improve immunogenicity assays.
The sponsor has also agreed to conduct several post-marketing commitments that we felt were important because of the
data gaps in the original data of the submission. These PMCs include one randomized controlled trial in patients with lupus
nephritis. A randomized controlled trial in African-American patients with SLE. And final reports from the long term open
label continuation trials that are ongoing. The details of these specific clinical trials have not yet been submitted so we have
not provided any feedback to the sponsor and we can't really comment on the details for those trials but the specific
submission dates were listed in the approval letter. That's all I have for the briefing summary. I guess we can open it up for
questions now.
Q&A:
Captain Fritsch:
I'm going to ask a question in the mean time until the questions start queueing up. It sounds like one of the trials was
conducted in Asia and Latin America so I'm wondering if you saw any other demographic differences in those populations.
Okada:
You mean with respect to proportion of Caucasian?
Fritsch:
I guess in terms of the African Americans. What you had talked about in the African American population, the drug may or
may not work as well in that population
I'm wondering if you noticed that within any of the other populations Asian of Latin American.
Okada:
That's a good question. There were not consistent trends with respect to the other racial subgroups. The real reason why we
became additionally concerned with the African American subgroup was that both trial subgroups appeared to have a
similar effect. With respect to Asians for example we have data on the Alaskan Native or American Indian subgroup for
example. In trial 1056 this subgroup actually followed a similar pattern to the African subgroup in the sense that the placebo
treated patients had a higher response than the Benlysta 10 mg/kg group did, however these results were not corroborated
in trial 1057. Again it may be an artificial grouping of folks but we didn't necessarily feel like that rose to the level of the
subgroup results are different.
Page 5
James Powell:
My question is, with respect to the African Americans in the clinical trial, you mentioned lower numbers, can you tell me
about those lower numbers and also related to that related to the safety profile can you tell me whether or not there was a
differential effect in African Americans with respect to safety observations?
Okada:
So the exact numbers that I have, for trial 1056, there were a total of 39 African Americans or African heritage patients in the
placebo group, 15 of whom responded. 40 patients in the 1 mg/kg group, 12 of whom responded and 39 in the 10mg/kg
group, 13 of whom responded. In trial 1057 that number is a lot smaller, there were 11 patients in the placebo group, 7 of
whom responded, there were 8 patients in the 1 mg/kg group 3 of whom responded and 11 patients in the 10 mg/kg group,
5 of whom responded. So we're talking about relatively small numbers.
James Powell:
And with respect to safety was there a differential effect?
Okada:
With respect to safety there didn't appear to be a differential effect. That went a long way to ultimately not being stronger in
the label against use in that subgroup. We're a little concerned about being too definitive based on subgroup results but we
didn't have any corroborating safety concern.
James Powell:
If it's possible for one more question can you tell me if there's any studies with respect to the particular antigen in African
American vs other patients that was used to create the antibody.
Okada:
That's a really good question. I don't think most of the studies looking at the serum levels of BLYS of BAFF, I don't think
there was any attempt to do racial subgrouping in any of those studies so I don't have that data.
Fritsch:
One of the post marketing commitments had to do with the use of Benlysta on vaccines, I'm wondering if for example, could
a person taking Benlysta also get a seasonal influenza vaccine this year, would that be contraindicated or allowed.
Okada:
The main contraindication with respect to vaccines is to not use Benlysta with live vaccines, so the flu shot should be just
fine. What we can't really comment on is does Benlysta impair vaccine responses. That's why we're asking for a post
marketing study to assess that. In theory the Human Genome Sciences suggests that there may be less of an impact on
vaccine responses because they're targeting a very specific subgroup of B cells so there shouldn't be an impaired response
but we'll have to see when the data actually come in.
Ann Santa-Donato:
Thank you. I'm from the association of Women's Health Obstetrics and Neonatal nurses and my question is have any
studies been done, animal studies been done on the efficacy of this drug for use in pregnancy.
Okada:
Benlysta is pregnancy category C. I'm trying to pull up the pregnancy section of the label real quick. Either way we wouldn't
really be able to comment on the efficacy. Animal studies don't always reflect human safety in this respect so we're a little
leery even if we have some the animal data to make the conclusion that it could be safe to use in pregnancy. I should
mention that there were approximately 47 pregnancies in the clinical trials occurring in lupus patients. They were told to use
a double barrier method of contraception but we know how things happen. And in those pregnancies there wasn't a specific
pattern that we could see in terms of potential problems however there was a scattering of spontaneous pregnancy loss. At
the moment we can't really say that it is safe so that's why we encourage anyone that gets pregnant while they're taking
Benlysta to register with the Human Genome Sciences Benlysta pregnancy registry.
Ann Santa-Donato:
So is it the recommendation that if a women becomes pregnant while taking the pregnancy should she discontinue or
consult with the healthcare provider about risk vs. benefit.
Page 6
Okada:
This is an infused medication so the healthcare provider is going to be involved in that decision at each visit but she should
consult with her healthcare provider. A lot of times it does come down to individual response and a willingness to accept the
risks
Lenore Coleman:
Yes this is Dr. Coleman. I have, I guess, a question and then a comment. Since based on the data it appears as though
African Americans are at a 3 fold increased risk of this in terms of higher incidence yet the drug does not appear based
upon the data that you reported, to not only not be efficacious but there is no way for us to be even assured of its safety.
What steps are being taken in terms of the promotion of this drug to physicians nationally regarding to lack of African
American patients that participated in the study and your concerns. My second is a comment. I think that this really points to
the fact that even though your recommendations that suggest that in chronic diseases where there is a preponderance of
patients in a particular ethnic group that has this disease, manufacturers are still not recruiting a proportion of minority
patients enough to assure not only efficacy but safety for use of products in the United States.
Okada:
With respect to the marketing, I can say that we do review the marketing materials. Our division of drug advertising and
marketing does that in conjunction with the review divisions and obviously any attempt to target that population would be
frowned upon by us and likely stopped. However I'm not aware and I sort of doubt that the company is going to try to market
to that subgroup given that we've already included in the label the potential reversal of treatment effect in that subgroup.
With respect to your comments I agree with you. I'm not quite sure why we didn't get more African American patients in trial
1056. It's a little baffling to us, it might be that because African American patients predominantly do experience renal
disease and patients with active renal disease were purposefully excluded from that trial. I think we're trying to address is by
the post-marketing commitment that the company has agreed to conduct a trial in African American patients. I'd like to also
recognize Dr. Rosemary Needer who is the primary medical officer involved in the application. Did you have an additional
comment?
Needer:
No I just want to say in defense of the sponsor, they did try to recruit as many African American patients as Dr. Okada did
allude to; the trial entry criteria prohibited entry of people who had major organ involvement such as lupus nephritis and
cerebritis
As you know African American patients tend to have a worse disease and that therefore a large number of them were not
eligible to participate in the clinical trial as they were designed.
Fritsch:
Thank you. This concludes the question and answer session of today's briefing. Just a reminder a replay of this
teleconference will be available at 866-513-4388 one hour after the call until April 14. No passcode is required. I want to
thank all of the attendees for their participation on such short notice. I hope this exchange was informative and has taken us
one step closer to providing timely information for you to inform y
