Ex,
Instead of getting into the trees too much, and instead of trying to guess too much, I think even you have missed an important point.
It is more than likely the noncrossovers did poorly for the most part, imo. That does not mean they all were below the median. They only make up 35% of the control group.
DCVax-l treatment group also has people that did not respond. Like control, they were predominantly unmethylated.
There are simply non-responders in both groups. Because idh-1 mutant were such a very very small percentage of patients in the trial, non-responders in both arms are typically short lived.
The treatment arm has some disadvantages.
* They had a higher percentage of patients enrolled in the second half of the trial that allowed in people with weaker immune systems.
* Unlike control, when they progressed they were often taken off DCVax-l. They also, on average, stopped receiving DCVax-l at 28 months.
* DCVax-l + SOC creates more early psPD than SOC alone, and so this wreaks havoc by, especially early on, taking people off DCVax-l rather than putting them on DCVax-l, which is what should have helped.
The control arm has certain advantages.
*They had a lower percentage of patients enrolled in the second half of the trial wherein patients with weaker immune systems were enrolled.
* unlike treatment arm, when they progressed or in some cases, pseudoprogressed, most were placed on DCVax-l.
* when patients progress after treatment with SOC, many of the tumor cells are converted by the SOC to a mesenchymal tumor signature, and the tumor cells become hypermutated, and both of the conditions make those tumor cells more vulnerable to DCVax-l.
* Because crossovers don’t start DCVax-l until after recurrence, their treatment with DCVax-l will be used longer than 28 months after enrollment.
While in some cases it is helpful to get DCVax-l early for those who do not progress, the advantages and disadvantages as listed above with both treatment, instead of just one, can, in many cases, cause confoundment. Moreover, in the earliest small group of patients enrolled, also known as the vanguard group, mgmt methylation status was not able to be tested.
Of course the most pervasive confoundment is the crossover itself.
One solution to confoundment was implementing an external control arm. Another solution was to track radiological bio markers that can utilize each patient as their own control, and offer an internal “dimension of efficacy” provided by DCVax-l.
We have seen the ECA comparisons. The other radiological findings have not been presented yet. We have also seen that certain groups known to have many more mesenchymal features, like those over 65 years of age and/or those who had more residual cells then initially treated by SOC, did far better than their ECA comparisons, because DCVax-l works better against these otherwise most aggressive tumors.
With the global definition of GBM now only being defined by two of the original four subtypes of GBM, Mesenchymal now makes up to approximately 2/3 of all GBM subtypes.
