Plea to FDA: Set Realistic Bar for New Breast Cancer Trial Endpoint

[Hopeful]

The Pink Sheet Daily. 2012 Aug 27, E Hayes

Stakeholders are concerned that FDA’s vague requirement for a “large difference” in treatment effect in its draft guidance on using pathologic complete response to support accelerated approval in neoadjuvant breast cancer could cause confusion, dissuade development and set up a de facto comparative efficacy requirement.FDA has not yet approved any drugs for neoadjuvant (pre-operative) breast cancer, but chemotherapy regimens are increasingly being used earlier off label in order to improve long-term outcomes. FDA has been eager to upend the traditional oncology development model where drugs start out in later lines of therapy and gradually work their way into earlier-stage disease as evidence mounts.

In May, FDA released a draft guidance establishing a regulatory pathway for accelerated approval of drugs for neoadjuvant breast cancer treatment, potentially based on only one well-powered randomized trial with pCR as a surrogate endpoint, a development hailed by industry and cancer researchers.

Commenting on the guidance, a number of companies and organizations called out the document’s vagueness on issues like evaluating the degree of effect on pCR, disease-free survival and overall survival.

According to the guidance, there should be a “large difference” in the pCR rate between treatment arms to produce a statistically significant difference in DFS or OS in the overall trial population that is also “clinically meaningful.”

Michelle Rohrer, vice president of regulatory affairs at Roche’s Genentech Inc., asked the agency to specify how large the improvement in pCR rate would need to be to support accelerated approval in a neoadjuvant indication.

The guidance currently quotes highly variable rates from a few percent to more than 65%, with emphasis placed on a Neoadjuvant Herceptin trial (NOAH), where Genentech’s trastuzumab doubled pCR compared to the chemo-alone arm (38% vs. 19%).

“As more effective treatments become available in the early-stage BC setting, it may become more difficult to show such substantial outcomes as doubling of pCR rates; nevertheless increasing pCR rates remains very important in order to continue to reduce the number of patients with residual invasive cancer and thus drive toward better clinical outcomes,” Rohrer’s letter states.

It would also be helpful to clarify the definition of “clinically meaningful” differences in DFS. A clinically meaningful improvement in DFS is dependent on the DFS in the control arm. For example, in situations where the standard of care has a DFS of 85% to 90%, an improvement of 3% to 5% may be considered clinically meaningful, Rohrer wrote.

Genentech is expected to be among the first to test the new mechanism, having announced plans for a neoadjuvant trial very soon after the draft guidance was published. The study will test Herceptin, the Herceptin-containing antibody drug conjugate T-DM1 and Perjeta (pertuzumab) in various combinations.

In its comments, the company asked whether a neoadjuvant trial could be used to support initial approval, and whether the DFS confirmatory data could derive from a different study from another approved indication. Genentech also requested more discussion about the adjuvant setting, including how “the approach described in the guidance might be used to support approval of an adjuvant indication once clinical benefit in the neoadjuvant setting is confirmed.”

Organizations responding to the draft guidance, including the American Society of Clinical Oncology and the Biotechnology Industry Organization, echoed some of Genentech’s concerns. ASCO President Sandra Swain further cautioned that the discussion of the relationship between pCR and DFS or OS in the draft guidance “may lead sponsors to avoid using pCR because the guidance implies that a large difference in rates will be required. In situations where pCR rates with existing regimens are already relatively high, requiring large improvements may be unrealistic.”

Similarly, BIO asked FDA to clarify what is considered a “large difference” in pCR and how large an improvement would be needed to support accelerated approval.

“In addition, it would be helpful to clarify the definition of ‘clinically meaningful’ differences in the context of this particular guidance on pCR. Especially since this phrase is used throughout the draft guidance,” wrote Andrew Emmett, managing director of science and regulatory affairs at BIO.

BIO: Consider Safety As Gauge Of Benefit

The draft guidance states that the definition of an appropriate magnitude of benefit depends on the prognosis of the patient population under study and the effectiveness of existing therapy for that patient population.

BIO suggested expanding on that definition. “We suggest adding the safety profile of the combination, in particular with add-on therapy to this list,” Emmett said.

Genentech offered the same suggestion, noting that the appropriate magnitude of benefit “also depends somewhat on the safety profile of the combination, in particular with add-on therapy in the neoadjuvant setting.”

A number of groups commenting on the guidance, including Genentech and BIO, asked for a flexible approach to the definition of high risk, in order to broaden the populations that would be eligible for trials designed on the new regulatory pathway.

The draft guidance specifies that the new pathway is meant for use in very high-risk cases of breast cancer where unmet need is great. Given the risks of testing new drugs in earlier stages of disease with an abbreviated pathway for approval, where there will be less up-front data collection, the agency concluded the risks posed by unknown safety profiles are only acceptable in women with poor prognosis despite available therapies, such as those with high-grade tumors lacking estrogen, progesterone and HER-2 receptors (triple negative). The guidance stipulates that the pathway is not appropriate for HER-2 positive cancer lacking high risk features, yet leaves the door open for testing in the HER-2 positive population.

ASCO says it supports the focus on high-risk populations for early-stage neoadjuvant trials, but asked for “expanded discussion” of other high-risk populations beyond triple negative disease, such as HER2-positive and high-risk ER positive.

Based on a recent FDA webinar, “HER-2 positive breast cancer patients are also considered to have unmet medical need (and it is our understanding that this is regardless of hormone receptor status),” Genentech’s Rohrer commented.

Stakeholders would like to have input on the appropriate definition of high risk, commented American Association for Cancer Research President Frank McCormick. A threshold figure such as 25%-30% risk of recurrence within three to four years might be a starting point for discussion, he suggested.

“We are hopeful that gaining clarity on how the FDA defines ‘high-risk’ will also help alleviate concerns about how the FDA will determine which patient populations are appropriate for trials to support accelerated approval, specifically with respect to ER+ subgroups and newly identified biomarker populations,” McCormick said.

A number of comments called for FDA not only to leave the pathway open for more subgroups of breast cancer, but to expand it to other diseases.
The agency’s acceptance of surrogate endpoints in the draft guidance for neoadjuvant breast cancer trials gives the National Brain Tumor Society “hope that we will be able more rapidly to translate new findings into novel treatments” for brain cancer patients, commented David Arons, the organization’s director of public policy.

Accrual for trials involving life-threatening diseases is exceedingly difficult – this is particularly true for the deadliest forms of brain cancer, where it is “almost impossible: to recruit and retain large numbers of subjects to complete clinical trials focused solely on overall survival, Arons wrote. The NBTS believes it’s important to consider other measures, such as quality of life. Furthermore, he wrote that the organization wants to help FDA develop an endpoint similar to pCR for use in brain cancer studies.

BIO also weighed in: “BIO and the biotechnology industry look forward to supporting and contributing to FDA’s efforts to identify and develop additional surrogate and intermediate clinical endpoints that can expand the Accelerated Approval pathway to encompass a broader array of life-threatening diseases and conditions.”

The recently enacted FDA Safety and Innovation Act directs FDA to increase use of the accelerated approval pathway. The accelerated approval language is part of a slate of measures incorporated into FDASIA, the fifth iteration of the Prescription Drug User Fee Act, to boost innovation.

Hopeful