BE CAREFUL WITH NEW DRUGS

June 29, 2010 (Seattle, Washington) — During the daylong discussion of last year's controversial Food and Drug Administration (FDA) advisory panel on prasugrel (Efient, Lilly/Daiichi Sankyo), one agency official asked: "Does prasugrel cause cancer?" While his answer, as well as the consensus reached by the expert panel, was that prasugrel did not appear to cause cancer, a couple of researchers continue to warn about the possibility that prasugrel might be a promoter of cancer and that physicians should be aware of these risks, especially in patients prescribed the drug long term [1].

In a new report in the June 29, 2010 issue of the Archives of Internal Medicine, Drs James Floyd (University of Washington, Seattle) and Victor Serebruany (Heart Drug Research Laboratories, Towson, MD) present data from an analysis of the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel (TRITON-TIMI 38) and show that use of prasugrel was associated with an increase in cancer risk, particularly a 62% increase in the rate of new and worsening cancers.

"There are data suggesting a strong signal of new and worsening solid cancers in the prasugrel arm, which grow exponentially and begin to occur at four months and even faster in women, independent of how you look at this," Serebruany told heartwire . "If you look at the centrally adjudicated events, if you look at events the investigators reported, or if look you at the FDA review of the data, there is a very straightforward solid increase in cancer-associated death and the existence of new and worsening cancer."

"Nobody Can Say There Is No Risk of Cancer"

The new report is based on the TRITON-TIMI 38 analysis performed by Dr Thomas Marciniak, a medical reviewer for the FDA, first included in the February 3, 2009 briefing document for the meeting of the Cardiovascular and Renal Drugs Advisory Committee. The analysis, summarized briefly by Floyd and Serebruany in Archives, documents all solid cancers, including nonhematologic malignant cancers, and excludes nonmelanoma skin cancer, which carries a benign prognosis and is usually excluded from clinical trials, and brain tumors.

Overall, there were 92 new solid cancers among patients treated with prasugrel and 64 new solid cancers among the patients treated with clopidogrel (Plavix, Bristol-Myers Squibb/Sanofi-Aventis), a significant 44% increase in risk. Excluding nonmelanoma skin cancers and brain tumors, there were 112 treatment-emergent cancers in the prasugrel arm and 69 treatment-emergent cancers in the clopidogrel-treated patients, a difference that translates into a significant 62% increase in risk of these new and worsening solid cancers. In addition to these findings, there was a nonsignificant 57% increase in deaths from cancer among patients treated with prasugrel.

To heartwire , Serebruany said some have suggested ascertainment bias could explain the difference in cancer rates in the two treatment arms. Because prasugrel is associated with a significant increase in bleeding, more cancers might be detected with the bleeding event, although Serebruany said the data do not show this to be true. After cancers detected by bleeding events were excluded, the difference in cancer rates persisted. Moreover, other trials do not support this theory.

"The data show that in the CHARISMA trial, in the combination group, the patients treated with clopidogrel and aspirin, there was less cancer than in the clopidogrel-group only, even though the combination patients had more bleeding," he said. "If the pattern is there, if there is really something going on with thienopyridines, where bleeding will cause the increased detection of cancer signal, this didn't happen with clopidogrel."

In the past, Serebruany suggested that the high level of platelet inhibition with prasugrel might impair defenses against tumor growth, but recent data from the PLATO study with ticagrelor (Brilanta, AstraZeneca) does not support this. Studies assessing the antiplatelet potency of ticagrelor vs prasugrel, however, are still needed, he said.

Concern About the Long-Term Use

Speaking with heartwire , Dr Sanjay Kaul (Cedar Sinai Medical Center, Los Angeles, CA) said that while the risk of cancer can't be ruled out, "there was already some uncertainty about whether the benefit outweighed the risk associated with prasugrel." Adding more uncertainty with the cancer signal adversely alters the risk/benefit profile of the drug and strengthens the idea that it should be used only short term.

Kaul, famously disinvited from the 2009 advisory panel meeting because of an incomplete vetting process that left unanswered questions about his possible "intellectual biases" with regard to prasugrel but who was later exonerated when the FDA apologized and admitted his removal was a mistake, wrote an editorial, along with Dr George Diamond (Cedars Sinai Medical Center), accompanying the Archives report by Floyd and Serebruany [2]. In it, as he stated to heartwire , Kaul runs through numerous possible scenarios: that prasugrel causes cancer, promotes cancer, triggers the detection of cancer, or is completely unrelated to cancer. Little is truly known right now, he said, leaving physicians to make inferences from the existing data.

"We just don't know about the long-term risks of cancer; it's not sufficiently addressed," said Kaul. "We are in the dark ages with regard to the science of tumor promotion."

In response to the new report, Lilly and Daiichi Sankyo emailed a statement to heartwire , pointing out that the FDA has already concluded it is unlikely prasugrel caused cancer and that the drug's label highlights the imbalance of newly diagnosed malignancies, which primarily occurred in the colon and lung. Also, preclinical data are "not indicative of tumor-growth enhancement," while "oncology experts have concluded that there is no biologic plausibility that prasugrel would be a tumor stimulator," according to the company.

Need to Systematically Collect Data

Commenting on the new report on prasugrel and cancer, Dr Magnus Ohman (Duke Clinical Research Institute, Durham, NC), chair of the ongoing TRILOGY-ACS study comparing prasugrel vs clopidogrel in medically treated acute coronary syndrome patients, said his group is prospectively collecting cancer information at baseline and throughout the course of the trial with a detailed case-report form.

"We need to realize that cancer and coronary disease coinhabit the same sphere, and we need to at least get the data correct," he said. "Whether there is an association or not, that's a much harder thing to prove, but at least we can collect the data."

Asked about the possibility of a link between prasugrel and cancer, Ohman said that researchers and clinicians need to be careful balancing what is known and what still needs to be determined. He questioned why some cancers are excluded from the analysis, such as brain tumors, saying this can alter the relative risks observed in statistical analyses. Ohman also noted that the FDA is seemingly of two voices with prasugrel, with Dr Ellis Unger (Office of New Drugs, FDA, Silver Spring, MD), one of the agency's main reviewers, stating that prasugrel is "unlikely [to be] carcinogenic" and does not promote tumor growth [3], while the Marciniak data raise some concerns. Regardless, Ohman said that "debate is always healthy" and that the issue reminds researchers that "cancer should always be addressed in clinical trials and never forgotten.

"We need to be careful not to overstate it one way or to understate it the other way," he told heartwire . "Nobody would ever want to sweep this under the rug and say this isn't a problem. A systematic, careful assessment of clinical trials is the only way we can move toward the truth."

Serebruany, however, does not believe the TRILOGY study, with an expected enrollment of 10 300 patients, of whom approximately half are enrolled, will answer the cancer question. Based on his calculations, if 280 cancer events are needed to detect a 50% increased risk of cancer, and assuming a control event rate of 1% annually, TRILOGY would need to include 22 000 patients to definitively detect a cancer signal. Kaul told heartwire that if researchers wanted to detect a 33% increase in cancer risk, TRILOGY would require nearly 40 000 patients.

"The leadership of the FDA has said that the cancer signal will be watched very carefully in TRILOGY, but TRILOGY is not powered enough to answer this question," Serebruany pointed out. "If the cancer signal were the same in TRILOGY as it was in TRITON, the study would need to be doubled in size. Stating that TRILOGY will clearly answer the question is a joke."