To REMS Or Not To REMS? Erivedge Shows Division At FDA Over Oncology Risk Management

The review division and CDER senior staff supported approval of Genentech’s highly teratogenic, basal cell carcinoma treatment vismodegib without a Risk Evaluation and Mitigation Strategy, but the Division of Risk Management believed a communications plan should be carried out under a REMS for the first-in-class Hedgehog pathway inhibitor.

FDA’s largely hands-off approach to Risk Evaluation and Mitigation Strategies for oncology drugs was called into question during the NDA review of Genentech Inc.’s basal cell carcinoma treatment Erivedge (vismodegib). Although the policy withstood a challenge from the Center for Drug Evaluation and Research’s own internal risk management experts, the experience could spark a re-evaluation of the use, or rather non-use, of REMS for cancer drugs.

The oncology drug review division and CDER senior staff did not believe a REMS was necessary for vismodegib, which is highly teratogenic and embryo-fetal toxic. This view was largely informed by the risk management approach taken with similarly teratogenic cancer drugs and the potentially precedent-setting nature of the decision to require a REMS for Genentech’s agent.

In the cancer setting, REMS generally have been reserved for those drugs with novel toxicities or where risk mitigation strategies are not well understood by oncologists.

As Division of Oncology Products 2 Director Patricia Keegan noted in her Jan. 23 summary review, the Office of Hematology and Oncology Products generally has not required REMS “for common, well-understood toxicities of a variety of products approved for the treatment of cancer; instead the requirement for a REMS have been reserved for mitigation of serious toxicities which are novel or where risk mitigation strategies are not well known or understood in the medical oncology community.”

Examples of the latter, Keegan said, include AstraZeneca PLC’s medullary thyroid cancer drug Caprelsa (vandetanib), which has been associated with QT prolongation, and Bristol-Myers Squibb Co.’s Yervoy (ipilimumab), a melanoma treatment that carries a risk of serious immune-mediated adverse reactions.

“In contrast, the risk of severe birth defects and intrauterine death is common to many cancer treatments across multiple product classes,” Keegan said. “With the exception of products/product classes first approved for non-cancer indications (e.g., thalidomide), REMS have not be [sic] required to mitigate the risks of teratogenicity.”

The review division’s opinion that a REMS was not necessary for vismodegib – a view reinforced by CDER senior staff at a regulatory briefing – prevailed over that of the Office of Surveillance and Epidemiology’s Division of Risk Management, which believed that communication efforts surrounding vismodegib’s risks should be conducted under the umbrella of a REMS.

Ultimately, DRISK accepted the lack of a REMS, albeit with some reluctance. However, the risk management unit questioned whether the agency’s longstanding, hands-off approach to REMS in oncology was appropriate and cited a need to develop a “consistent regulatory approach” for teratogenic drugs used in cancer settings.

Both the review division and DRISK also left the door open to a more restrictive risk management approach if the patient population for vismodegib expands, either through a broadened label or off-label use.

The unique nature of cancer treatments and their distribution systems saved yet another oncology drug sponsor from the burden of a REMS, with FDA deciding that vismodegib’s risk/benefit profile was acceptable for a difficult-to-treat disease like advanced basal cell carcinoma. However, there is no guarantee of a REMS-less future for Erivedge or, if DRISK has its way, other teratogenic cancer treatments, particularly those with the same mechanism of action as the first-in-class vismodegib.

Blazing A New Trail Along The Hedgehog Pathway

FDA approved vismodegib on Jan. 30 for the once-daily, oral treatment of adults with metastatic basal cell carcinoma, or with locally advanced disease that has recurred following surgery or who are not candidates for surgery, and who are not candidates for radiation.

Previously, there was no FDA-approved therapy for patients with metastatic disease and no effective treatment for metastatic or locally advanced disease that is refractory, or not amenable, to surgery or radiation, Jeffery Summers said in a Jan. 13 cross-discipline team leader review. The lack of approved and effective treatments was one of several factors behind the decision to grant full approval based on a lone Phase II, single-arm study (see related story, (Also see "Erivedge Efficacy Overcomes Limitations Of Single-Arm Study" - Pink Sheet, 1 Nov, 2012.)).

Developed under a collaboration with Curis Inc., vismodegib is the first of a new class of drugs known as Hedgehog pathway inhibitors to reach market (Also see "Curis Set For Transformative Year With Erivedge Skin Cancer Approval" - Pink Sheet, 6 Feb, 2012.). Hedgehog is a growth signaling pathway co-opted by cancers to assist in their own growth and survival; it is also active in fetal development. Vismodegib’s approval was viewed as validating decades of research on the pathway (Also see "Erivedge May Be The Gleevec Of Hedgehog Inhibitors" - Pink Sheet, 13 Feb, 2012.).

Key Documents

Want to read FDA reviewers’ opinions on managing Erivedge’s risks? The memos and other documents mentioned in this story are available at our website, www.elsevierbi.com/pharmaceutical-approvals-monthly.

However, there are significant and well-known safety issues associated with Hedgehog inhibitors – teratogenicity and embryo-fetal toxicity in pregnant women.

“There is extensive literature describing the critical role of the Hedgehog pathway in embryonic and fetal development,” Acting Supervisory Pharmacologist Todd Palmby said in a Jan. 13 review. “There is also evidence that exposure of pregnant animals to Hedgehog pathway inhibitors (e.g., cyclopamine) results in severe midline defects in the offspring of multiple species. In an embryo-fetal developmental toxicity study in rats, vismodegib was teratogenic (missing and/or fused digits, open perineum and craniofacial anomalies) at a dose corresponding to an exposure of 20% of the exposure at the recommended human dose, and was embryotoxic and fetotoxic (early resorption of 100% of the fetuses) at exposures comparable to those at the recommended human dose.”

In a Jan. 30 decision memo, Office of Hematology and Oncology Products Director Richard Pazdur said that although there are no human data on vismodegib’s effects on fetal development, “the extent of the nonclinical data, consistency of findings across the class (e.g., cyclopamine), the well-established mechanism for vismodegib and the established role of this pathway in embryo-fetal development are sufficient to establish this risk.”

A REMS Proposed, But Not Requested

Details about Genentech’s original risk management proposal are redacted from FDA review documents. However, the company told Pharmaceutical Approvals Monthly that it originally proposed a REMS and a Category X pregnancy labeling designation.

Category X is reserved for those drugs that have the most unfavorable risk/benefit profile when used during pregnancy. This designation means that studies in animals or humans have demonstrated fetal abnormalities, and the risk of use during pregnancy clearly outweighs any possible benefit. Such drugs carry a contraindication for use during pregnancy.

Pazdur’s memo notes that the review division “did not request nor recommend that a REMS be submitted.” Since Genentech’s original risk management proposal took a more conservative approach than FDA believed was necessary, the agency wasted little time in letting Genentech know this.

On Oct. 20, 2011 – one-and-a-half months into the review period – FDA held a teleconference with Genentech to seek clarity and gather information about the sponsor’s proposed risk management strategy. Addressed during the teleconference were several discussion points that FDA had sent to Genentech a day earlier, one of which related to pregnancy testing and continued use of vismodegib, according to the teleconference minutes(see box).

October 20, 2011 FDA/Genentech Teleconference Discussion Point 6

“Please justify why vismodegib must be stopped in woman [sic] with positive pregnancy tests. Have you considered alternative actions that a positive pregnancy test should trigger – such as counseling with the prescribing physician and referral as appropriate…?”

“Discussion: Genentech stated they had considered vismodegib to be Pregnancy Category [redacted] and, therefore, placed the requirement in labeling to test female patients of child-bearing potential every month. FDA stated that vismodegib may not be considered as Pregnancy Category [redacted]. FDA also requested that Genentech re-consider the requirement for routine urine versus serum pregnancy tests.”

According to the minutes, FDA “requested that Genentech propose its approach(es) in communicating the potential risks of embryo-fetal toxicities associated with vismodegib and specific objectives or goals must be clearly described. Genentech acknowledged FDA’s request and agreed to submit revised risk management programs within 3 weeks.”

Genentech submitted a revised risk management plan on Nov. 4 (see timeline, (Also see "Erivedge Development Timeline" - Pink Sheet, 1 Nov, 2012.)). The submission included revised product labeling, with the pregnancy classification changed to Category D, meaning there is positive evidence of human fetal risk but the potential benefits of use during pregnancy may nevertheless be acceptable. The sponsor also submitted a revised Medication Guide, a voluntary communication plan and an enhanced pregnancy pharmacovigilance plan.

In his Jan. 13 memo, Summers explained why the review division believed revisions to Genentech’s original risk mitigation proposal were warranted.

“The genesis for the Applicant’s concern and initial approach likely rests both in the evolving use of REMS across approved drugs at FDA and the exquisite delineation of the Hedgehog gene’s evolutionarily conserved effects on morphogenesis,” Summers said.

“However, most drugs in oncology, whether older cytotoxic drugs or drugs developed as more ‘targeted’ therapies, can cause profound deleterious effects on fetal development. The practice of oncology entails that detailed and frequent discussions occur between a patient and their oncology team regarding the highly toxic drugs used to treat the patient’s disease. In addition, close monitoring of patients for drug toxicity and continued ongoing counseling regarding such issues as pregnancy are a mainstay of oncology practice. FDA requested that the Applicant re-evaluate the teratogenic toxicity of this drug in relation to oncology drugs in general and alternative mechanisms whereby the described goals of the program could be achieved.”

Regulatory Briefing On Need For REMS

Review documents discuss a CDER regulatory briefing held Dec. 9 to solicit input from senior staff on risk management for teratogenic cancer drugs in general, and vismodegib in particular. FDA holds regulatory briefings to gather input from across the agency on emerging issues with broad impact.

The vismodegib regulatory briefing occurred just one month after a similar briefing on the question of whether a REMS was necessary for Corcept Therapeutics Inc.’s Cushing’s disease treatment Korlym (mifepristone). The severity and rarity of Cushing’s, coupled with the requirement for highly specialized care, led the agency to ultimately conclude that a REMS was not necessary for that drug (Also see "FDA’s Decision To Back Off Korlym REMS Shows Context Is King" - Pink Sheet, 1 Sep, 2012.).

CDER senior staff acknowledged that requiring a REMS for vismodegib “could set a precedent for future approvals of other antineoplastic drugs,” cross-discipline team leader Summers said.

At the vismodegib regulatory briefing, the panel discussed the current management through labeling of teratogenicity risks for most oncology drugs and the existence of a de facto restricted distribution system – a point Pazdur has previously highlighted in explaining why relatively few cancer drugs have REMS (Also see "Oncology Has Different Risk Management Standards, FDA’s Pazdur Acknowledges" - Pink Sheet, 13 Jun, 2011.).

The potential precedent that a REMS for vismodegib would set appeared to concern CDER’s senior staff.

“The panel acknowledged that a regulatory decision for vismodegib requiring a REMS could set a precedent for future approvals of other antineoplastic drugs and raise the question if drugs approved prior to vismodegib should be re-evaluated for a REMS program,” Summers said. “These discussions supported the determination that the teratogenicity risk of vismodegib did not warrant a REMS and that labeling including a Medication Guide would be sufficient to communicate the risk of teratogenicity for the majority of drugs used in the practice of oncology where the drug was shown to have a meaningful clinical impact on an endpoint likely to predict effects on irreversible morbidity or mortality.”

Although this general non-binding advice was in agreement with the views of the clinical and pharmacology/toxicology review teams, DRISK believed that a REMS was necessary.

In a Jan. 9 REMS Options Review memo, Risk Management Analyst Amarilys Vega said DRISK believed vismodegib should be approved with a communications plan and education program under the umbrella of a REMS but not linked to a provider’s ability to prescribe the drug.

Requiring a communication plan under a REMS will “allow FDA to monitor the content of the risk message … enforce the continuity of prescriber training long after initial product approval; and allow for systematic assessments of program effectiveness,” DRISK reviewer Vega said.

“This approach will provide prescribers access to FDA approved, product-specific risk information to enhance their understanding of vismodegib’s high teratogenic potential, facilitate patient counseling, and emphasize the importance of compliance with pregnancy testing,” Vega said. “Requiring this strategy as part of a REMS will also allow FDA to monitor the content of the risk message and communication of specific requirements for safe use; enforce the continuity of prescriber training long after initial product approval; and allow for systematic assessments of program effectiveness.”

Vega said DRISK was aligned, but not in full agreement, with the advice provided by the regulatory briefing panel and the decision to approve without a REMS. “However, DRISK has a low threshold for re-evaluating the need for a REMS for vismodegib, particularly if the treated patient population expands or if new safety data become available indicating that product labeling alone is not effective at managing vismodegib’s risk of teratogenicity.”

An on-label, expanded population is possible for Erivedge down the road: the compound is in Phase II trials for operable basal cell carcinoma and in Gorlin syndrome patients, who have a large burden of basal cell carcinoma, Genentech said. Third-party investigator studies are ongoing or planned in various other cancers.

Summers also envisioned revisiting the decision not to require a REMS in the event of broader-than-expected use of the drug.

“The off-label use of vismodegib in patients with BCC that should be treated with surgery or radiation would entail a negative risk:benefit assessment in this reviewer’s opinion,” Summers said. However, “the economic realities of pharmaceutical pricing and reimbursement practices make the likelihood of off-label use for less advanced BCC highly unlikely.”

Looking Beyond Vismodegib

Moving beyond the confines of vismodegib’s risk/benefit profile, DRISK suggested the agency should rethink how it approaches risk management for teratogenic cancer drugs.

“DRISK acknowledges there is a regulatory precedent for not requiring a REMS for most oncology drugs demonstrating a risk for teratogenicity. While this has been the standard approach for oncology products, it is unclear this is the most appropriate approach for all oncology drugs,” Vega said.

“We urge further discussion regarding the development of a consistent regulatory approach for the management of the risk of teratogenicity for oncology drugs that demonstrate a risk of teratogenicity. Points to consider in this discussion should include whether an oncology drug with a teratogenic risk would ever require a REMS to mitigate the risk, and if so what factors would be most important in making that determination (e.g., nature of the disease, patient population, prescriber population, characteristics of the drug, and expected benefits),” Vega said. “In addition, FDA should engage professional organizations (e.g., ASCO) to develop guidances and educational programs to assist oncology prescribers in counseling females of childbearing potential on how to mitigating [sic] the risk of teratogenicity.”

Post-marketing Requirement Fills The REMS Void

Erivedge was approved with Pregnancy Category D labeling and a boxed warning on the risk of embryo-fetal death and severe birth defects. “Verify pregnancy status prior to the initiation of Erivedge,” the warning states. “Advise male and female patients of these risks. Advise female patients of the need for contraception and advise male patients of the potential risk of Erivedge exposure through semen.”

While the review division did not require a REMS, it did insist on a pregnancy pharmacovigilance program as a post-marketing requirement.

Genentech must collect registry data to evaluate pregnancy and infant outcomes following exposure to vismodegib. “This study will include a mechanism to collect, classify and analyze data on direct exposures (women exposed to vismodegib as treatment) and indirect exposures (women exposed to vismodegib through the seminal fluid of a male partner),” the approval letter states. Genentech must analyze the outcomes data and provide descriptive statistics in a stand-alone report submitted annually for nine years.

Genentech said the pharmacovigilance program also includes a communication plan with educational materials for prescribers and patients that describe the risks to fetal development.