The anti Provenge JNCI article written by an ex hedge fund employee was debunked by Ocyan and other knowledgeable people in yesterdays issue of Pharmalot.
Here is Ocyan's take on the so called Immunodepletion theory that is being promoted by one of the hedge funds who are short Dendreon.
"The Immunodepletion theory in Huber’s paper has no merit. The survival analysis for the subgroup division at 65 was likely biased and spurious. Randomization even failed to balance the numbers of patients on the two trial arms of this partition so there was little hope that other crucial diagnostic factors would be balanced. Using this statistically spurious result to launch a hypothesis that Provenge did not work was naive from mathematical point of view.
Nonetheless, that naivete quickly became a pernicious scientific deception when Huber et al proposed an elaborate immunological theory to explain the abnormal results without vetting the truth of some of its obvious predictions.
For example, if about 90% of the immune cells of a patient on the control arm of the trials were indeed depleted by the leukapheresis procedure as Huber et al asserted, those patients should have become rather sick quickly compared to patients on the treatment arm. Yet, the safety database showed no such imbalance. Indeed, the fact is that most immune cells are sequestered in various tissues so that less than 2% of them would be taken in a leukapheresis procedure. The human body produces more than this cell count daily.
Next, consider the conclusion that Provenge was just a placebo treatment so that the survival benefits seen in the trials were merely from harm done to patients on the control arm. First, if that was the case, patients on the control arm should have died en masse early. Yet the survival curves showed that the two populations performed about the same for several months before separating in Provenge’s favor. Then, a recent analysis of the data showed that patients on the control arm who received a weak form of Provenge made from salvaged blood (ie, the colloquial Frovenge) after disease progression (often months after randomization) lived far longer than those who did not get it. If Provenge was a placebo, Frovenge must be even worse. Why did these Frovenge patients on the control arm of the trials perform so well when receiving it only after months of lacking treatment?
When a scientist proposes a theory, it is incumbent upon her/him to at least vet simple predictions to see if they hold up. Huber et al did none of that. Instead, they published a theory indicting Provenge, then ask others to prove them wrong. This is arrogant and pernicious! These authors should be ashamed of themselves.
Huber et al should also think about the harm that they might have done to doctors and patients without the statistical and immunological background to understand the shortcomings of their Immunodepletion theory. If those patients and doctors end up avoiding a treatment that may extend their life, such harm will be attributed directly to their work.
Lastly, I’d like to quickly address this statement: “There has also been debate about effectiveness because Provenge does not shrink tumors.” This is an oft-repeated misconception about statistics. The Provenge trials did not demonstrate statistical significance results on Time to Progression. That is not the same as “Provenge does not shrink tumors”. Unlike a chemotherapy that delivers a toxin to tumors (and other parts of the body too), an immunotherapy often takes months to ramp up the immune system. During that time, some tumors may get larger and some new ones may develop. Too frequent radiographical assessments might catch these cases and showed that disease has progressed while, in fact, the immune system was just getting started to exert its protection. So, the lack of a statistical significant result on Time to Progression is far more likely due to the measurement methodology than because of the efficacy of Provenge. Indeed, with three separate phase-3 trials showing survival benefits, it should be clear that Provenge did arrest the disease at some point.
There is a need to develop new measuring methods to see how Provenge and, indeed, other immunotherapies help stopping or slowing down cancers. That’s what the various Cancer Vaccine Workshops organized by the FDA tried to do. See the references in the Wikipedia page on Provenge (Sipuleucel-t) for details."
Here is Ocyan's take on the so called Immunodepletion theory that is being promoted by one of the hedge funds who are short Dendreon.
"The Immunodepletion theory in Huber’s paper has no merit. The survival analysis for the subgroup division at 65 was likely biased and spurious. Randomization even failed to balance the numbers of patients on the two trial arms of this partition so there was little hope that other crucial diagnostic factors would be balanced. Using this statistically spurious result to launch a hypothesis that Provenge did not work was naive from mathematical point of view.
Nonetheless, that naivete quickly became a pernicious scientific deception when Huber et al proposed an elaborate immunological theory to explain the abnormal results without vetting the truth of some of its obvious predictions.
For example, if about 90% of the immune cells of a patient on the control arm of the trials were indeed depleted by the leukapheresis procedure as Huber et al asserted, those patients should have become rather sick quickly compared to patients on the treatment arm. Yet, the safety database showed no such imbalance. Indeed, the fact is that most immune cells are sequestered in various tissues so that less than 2% of them would be taken in a leukapheresis procedure. The human body produces more than this cell count daily.
Next, consider the conclusion that Provenge was just a placebo treatment so that the survival benefits seen in the trials were merely from harm done to patients on the control arm. First, if that was the case, patients on the control arm should have died en masse early. Yet the survival curves showed that the two populations performed about the same for several months before separating in Provenge’s favor. Then, a recent analysis of the data showed that patients on the control arm who received a weak form of Provenge made from salvaged blood (ie, the colloquial Frovenge) after disease progression (often months after randomization) lived far longer than those who did not get it. If Provenge was a placebo, Frovenge must be even worse. Why did these Frovenge patients on the control arm of the trials perform so well when receiving it only after months of lacking treatment?
When a scientist proposes a theory, it is incumbent upon her/him to at least vet simple predictions to see if they hold up. Huber et al did none of that. Instead, they published a theory indicting Provenge, then ask others to prove them wrong. This is arrogant and pernicious! These authors should be ashamed of themselves.
Huber et al should also think about the harm that they might have done to doctors and patients without the statistical and immunological background to understand the shortcomings of their Immunodepletion theory. If those patients and doctors end up avoiding a treatment that may extend their life, such harm will be attributed directly to their work.
Lastly, I’d like to quickly address this statement: “There has also been debate about effectiveness because Provenge does not shrink tumors.” This is an oft-repeated misconception about statistics. The Provenge trials did not demonstrate statistical significance results on Time to Progression. That is not the same as “Provenge does not shrink tumors”. Unlike a chemotherapy that delivers a toxin to tumors (and other parts of the body too), an immunotherapy often takes months to ramp up the immune system. During that time, some tumors may get larger and some new ones may develop. Too frequent radiographical assessments might catch these cases and showed that disease has progressed while, in fact, the immune system was just getting started to exert its protection. So, the lack of a statistical significant result on Time to Progression is far more likely due to the measurement methodology than because of the efficacy of Provenge. Indeed, with three separate phase-3 trials showing survival benefits, it should be clear that Provenge did arrest the disease at some point.
There is a need to develop new measuring methods to see how Provenge and, indeed, other immunotherapies help stopping or slowing down cancers. That’s what the various Cancer Vaccine Workshops organized by the FDA tried to do. See the references in the Wikipedia page on Provenge (Sipuleucel-t) for details."