BiovaxID and the FDA's Recent Cancer Drug Approval Record (Part 2 of a series)

When the FDA told Biovest it wanted a confirmatory phase III clinical trial on top of its existing phase III trial for the therapeutic lymphoma vaccine BiovaxID, it effectively delayed the arrival and availability of BiovaxID in the United States for at least two years.

When I first read that another trial was required, I thought it was a misprint. They mean confirmatory phase IV, I thought, which would not have held up market approval.

FDA vs Biovest

Why does the FDA want another trial? I spoke to Biovest to get their side of things. Then I contacted the FDA for theirs (I knew better, actually, but I emailed them anyway).

Here is the FDA's response, with certain information redacted.

No, the email did not self-destruct.

Considering the enormous potential of BiovaxID to both extend and improve the quality of life for so many B-cell lymphoma patients, the FDA's request for another lengthy and expensive phase III trial forces some questions:

• Was Biovest's phase III trial design flawed?
• Does the FDA hold cancer therapies to a higher standard than therapies for other diseases?

Clinical Trial Design

The gold standard in clinical research is the randomized, double-blind trial. This kind of trial removes as much subjectivity and bias from the results as possible, leaving only objective data. The opposite of the gold standard in clinical research is the single-arm, open-label trial.

Here are some definitions to help you understand the differences in these types of research:

Randomized

This means patients were randomly assigned to two or more groups (or 'arms'). Randomizing a trial helps to limit the influence of selection bias, and it clarifies the impact of the therapy over the other options, such as best supportive care, placebo, etc. The opposite is a single-arm or non-randomized study.

Double-Blinded

This means that neither the patients nor the investigators know who was receiving the drug under investigation and who was receiving something else. Double-blinding a study promotes the objectivity of its findings. The opposite is open-label, where both parties know.

Biovest's phase III trial of BiovaxID was an 8-year, double-blind, randomized trial¹. In clinical research, you can't do any better.

Journal of Clinical Oncology Review

A 2009 review published in the Journal of Clinical Oncology ² surveyed the 68 cancer drugs approved by the FDA between 1973 and 2006 (excluding hormone therapy and supportive care) and found that 31 of them had been approved without a randomized trial. At the time of review, only one of those 31 drugs had been removed from the market. The authors concluded:

"Nonrandomized clinical trials with definitive end points can yield US Food and Drug Administration approvals, and these drugs have a reassuring record of long-term safety and efficacy."

One might interpret the FDA's increasing willingness to accept non-randomized trials as tacit participation by the agency in saving the drug developer time and money (since these trials can be conducted quicker and cheaper). One might also interpret it as an eagerness on behalf of the agency in getting life-saving cancer drugs to the public.

Were the many approved cancer drugs truly life-saving, I would have no problem with this. Unfortunately they aren't life-saving. At best, many appear to extend life a few months while causing severe adverse events, so I would argue that the best interests of the public are not being served by this trend.

FDA History of Cancer Drug Approval

Normally, when the FDA reviews a drug application, it wants to see that the therapy is safe and effective. It wants to be able to present the therapy to the public in such a way that the patient population is well-defined and that all the necessary information is there to inform the safe and effective use of the therapy.

That very general explanation attempts to cover every drug the FDA sees. However, when you start to break drugs into categories, differing standards bubble up. This is especially true in oncology drugs. Does this mean the FDA holds cancer therapies to a higher standard than therapies for other diseases?

The answer is a resounding "No." In fact, the opposite is true: In addition to the trend of accepting non-randomized trials, the FDA routinely lowers other standards for cancer therapies, often by adjusting those standards on a subjective and case-by-case basis.

Approval

Back in 2009, Dr. Richard Pazdur told the New York Times that as the director of the FDA’s cancer drug office, he is in a can't-win situation. “If you approve a drug, they accuse you of lowering standards," he said. "And if you don’t approve it, you’re the worst thing since the Nazi death camps and should be killed.”

Don't approve? Since the early 1990s the FDA has rubber-stamped so many cancer drugs, Congress had to authorize the purchase of a new rubber stamp.

In 2010 a review ³ in the Journal of the National Cancer Institute (co-authored by Pazdur) showed that between 1 July 2005 and 31 December 2007, his office had reviewed 60 new oncology drug and biological licensing applications and supplementary applications. Two were voluntarily withdrawn. Five were rejected. 53 were approved.

Among the 53, new formulations of existing drugs accounted for 35 of them, while the other 18 were for new entities.

Studies: How Many And What Kind

According to the review, in 44 of the drugs, findings from a single study proved good enough for FDA approval. It is not uncommon in cancer therapy for companies to submit applications based on just one study, even though this is contrary to FDA regulations, which require data from more than one independent study.

On top of that, 37 approvals presented randomized studies, while 16 presented single-arm studies. I'll return to this point shortly.

Primary End Point Inconsistency

One manner by which the FDA has loosened the standards for oncology drugs has been to permit so-called 'surrogate end-points'. Bandolier offers a good explanation of a surrogate endpoint:

"Surrogate endpoints are often physiological or biochemical markers that can be relatively quickly and easily measured, and that are taken as being predictive of important clinical outcomes. They are often used when observation of clinical outcomes requires long follow-up. The main thing to remember is that there has to be a good reason to accept a surrogate endpoint."

Among the cancer drugs given regular approval in the JNCI study, endpoints are all over the place:

• Twelve (12) used progression-free survival (PFS) or time to progression
• Ten (10) used overall survival (OS)
• Sixteen (16) used objective response rate (ORR)
• Four (4) used disease-free survival (DFS)
• Three (3) used OS and PFS
• Seven (7) used novel endpoints, including:
  • "Reduction in hepatic iron"
  • "Red blood cell transfusion independence"
  • "Hemoglobin stabilization"
  • "Occurrence of venous thromboembolism
  • "Reduction in tissue injury from chemotherapy extravasation"

Conclusion

Biovest presented the FDA with a phase III trial the only shortcoming of which was that it was rather small—117 patients. But, along with demonstrating an impressive disease-free survival benefit as consolidation therapy, the trial also demonstrated a stunning safety profile.

In almost every way, Biovest equaled or outdid other drugs approved by the FDA. Of all those drugs, precisely none of them are therapeutic cancer vaccines.

In the next installment, I'll look at the science and the data of BiovaxID, on its own and as it compares to its primary rival, rituximab.

End Notes

1. BiovaxID Phase III Clinical Trial Summary. At biovest.com.
2. Tsimberidou A et al. Ultimate Fate of Oncology Drugs Approved by the US Food and Drug Administration Without a Randomized Trial. JCO December 20, 2009 vol. 27 no. 36 6243-6250.
3. Sridhara R. Review of Oncology and Hematology Drug Approvals at the US Food and Drug Administration Between July 2005 and December 2007. J Natl Cancer Inst 2010;102:doi:10.1093/jnci/djp515.

BiovaxID: On the Trail of the Lymphoma Vaccine

Follow the links below to read the rest of this series:

• Part One
• Part Three
• Part Four