Data on Off-Label Targeted Cancer Drugs More Plentiful Than Helpful

[Hopeful]

The Pink Sheet Daily. 2009 Nov 9, G Twachtman

Evidence within compendia to support the off-label use of targeted cancer therapies varies widely and is rapidly evolving, making it difficult to draw solid conclusions regarding safety and efficacy, according to a draft technology assessment report released for comment by the Agency for Healthcare Research and Quality. Researchers from the Duke University Evidence-based Practice Center prepared the systematic review of eight targeted therapies and 19 respective off-label indications. The indications were derived from a list of all targeted therapy drugs available through the end of 2006 and for which compendia in the 2006 versions provided clear recommendations.

The report was prepared at the request of CMS. The draft notes that CMS "requested a technology assessment of efficacy and safety of selected targeted therapies when prescribed for off-label indications" and "will consider this information as background to its further discussion of coverage for and policies regarding targeted therapies." Comments on the draft are due to AHRQ by Nov. 16.

A CMS spokesman said this technology assessment will help inform work the agency is doing with the Medicare Evidence Development and Coverage Advisory Committee during a January 2010 meeting on pharmacogenomics testing in cancer. In the meeting announcement, CMS said it convened the panel "to consider and make recommendations to CMS regarding the evidence that supports the use of pharmacogenomic testing in the diagnosis and treatment of cancer." There currently is not a national coverage determination for this kind of testing.

The agency is seeking "guidance from the panel to inform future coverage determinations," the notice added. "We want to ensure that Medicare beneficiaries have access to any demonstrated improved health outcomes of pharmacogenomic testing, and are protected from inaccurate or inappropriate pharmacogenomic testing that could compromise therapy or increase the risks of adverse events during therapy."

FDA Commissioner Margaret Hamburg recently said the agency is looking into pharmacogenomics regulatory policies, but the process could take a while.

Varied Quality Of Data

Among the key findings noted within the draft report was the disparity in the quality of data used in off-label listings.

"Each discrete systematic review yielded interesting information on the off-label use(s) of a specific targeted therapy," report authors said. "However, the quality of studies varied tremendously, and the low quality of many published studies made it difficult to draw conclusions regarding the agent's effectiveness or safety."

Report authors, led by Amy Abernathy, noted that study quality varied both across and within drugs and because of the "high paucity of high-quality evidence, the data available - though voluminous - may have little meaning of value for informing clinical practice. Similarly, our horizon scan yielded little useful evidence; review of this level of evidence may serve better as an early warning system, alerting clinicians to potential adverse events (e.g., CMV in alemtuzumab) than as a viable source of data on efficacy."

Alemtuzumab Illustrates Issues

Report authors highlighted alemtuzumab (Genzyme's Campath ) and its supporting evidence for off-label treatment of cancer as an example of some of the challenges that were faced when doing the analysis.

Early in its development, Campath was seen as a possible first-line monotherapy for non-Hodgkin's lymphoma, but data indicated there might be increased risk of opportunistic infection and that efficacy was insufficient for first-line monotherapy. It is indicated as a single agent for the treatment of B-cell chronic lymphocytic leukemia.

The report states that "clinical indication for use of alemtuzumab varied; specifically, we found the agent used at diverse stages of disease presentation. Four of the 11 published reports included alemtuzumab as a part of stem cell transplant conditioning regimen, to reduce toxicity from donor lymphocyte infusion, or in conjunction with some other aspect of transplant therapy."

In the remainder of studies, "participants generally had relapsed or refractory NHL, or minimal residual disease after other standard therapies."
Another pair of studies had alemtuzumab being used as a first-line therapy, with patients with peripheral T cell lymphoma as their NHL subtype.
"In this wide variety of contexts, it is difficult - though necessary for the practicing clinician - to try to interpret efficacy outcomes and tolerability from the available evidence," report authors stated.

But alemtuzumab was hardly alone in data issues. "The quantity, as well as quality, of data varied widely across the included indications," the report says.

Researchers acknowledged that there are certain practical limitations to available data, including the rarity of the disease the targeted therapy is being used to treat and the limited data associated with certain drug/disease combinations.

"A published case series and case reports suggest that bevacizumab [Roche/Genentech's Avastin ] may contribute to clinical response in [epithelial ovarian cancer], including reduction in CA125 levels," the draft report states. "However, these studies were not designed to directly evaluate the efficacy of bevacizumab, and there is inherent bias in relying on case reports or retrospective case series to assess either efficacy or safety of interventions, especially those typically used in combination with other treatments."

Researchers found that for mature drugs, there was typically more data, and in some cases the level of data was "remarkable" and likely "reflective of public enthusiasm for those drugs."

Defining "Evidence"

In preparing the technology assessment, report authors noted that "the array of data presented in randomized control trials, controlled clinical trials, case series, case reports, abstracts, review articles, and other reports raised the issue of what constitutes 'evidence.' Disagreements are likely to surround this question, making the process of evidence review susceptible to differences of opinion and subjectivity."

The report goes so far as to state that much of the evidence it reviewed would not, generally, be considered "good science." It stated that the four recognized compendia presented results from trials that "were inadequately controlled, had potential conflicts of interest due to funding scenarios, involved double-counting across studies, or did not conform to good clinical practices."

The authors continued: "Sometimes the remarkably low quality of the reporting obscured any ability to determine the strength of the evidence. The pace of research, combined with clinical urgency, may at times threaten the integrity of the science underlying clinical decisions."

However, authors do point to a flip side of this coin - that cancer presents its own unique circumstances to those who have it particularly because of the extremely limited treatment options, which can create a different perspective when it comes to evaluating evidence related to the targeted therapies.

"In the cancer setting, the definition of what constitutes 'evidence' may thus be substantially looser than in other disciplines, and providers reviewing the evidence may be more inclined to entertain data derived from lower-quality studies or the 'gray' literature," the draft report states, adding that there are estimates that less than half of all medical care in the U.S. is supported by adequate evidence regarding effectiveness."This may be particularly true in cancer (especially as we question the adequacy of the evidence identified in this review)."

Report authors concluded that "a different model of evidence generation and evaluation is warranted - but is it possible? ... Developing consensus holds that the new system must be designed for rapid learning and expedite translation of research into clinical practice improvement, and must include the evaluation of the comparative effectiveness of available treatments in real-world clinical populations."

Hopeful