FDA Panel Votes for New Breast Cancer Approval

The Oncologic Drugs Advisory Committee (ODAC) recommended the approval of Nerlynx (neratinib) for some patients with breast cancer.

The Food and Drug Administration’s (FDA) Oncologic Drugs Advisory Committee (ODAC) recommended the approval of Nerlynx (neratinib) for the extended adjuvant treatment of patients with early stage, HER2-postivie breast cancer after they had postoperative Herceptin (trastuzumab). The voting ratio for the recommendation was 12 to four.

ODAC based its recommendation on data from the phase 3 ExteNET trial and the phase 2 CONTROL trial. In the primary analysis of the ExteNET trial, the invasive disease-free survival (iDFS) rate at 2 years was 94.2 percent with Nerlynx versus 91.9 percent with placebo.

The results indicated that the benefit may vary based on hormone receptor (HR) status. An exploratory subgroup analysis indicated that Nerlynx lowered the risk of recurrence by 51 percent in HR-positive patients, compared with 7 percent in HR-negative patients.

Diarrhea was the primary safety concern with Nerlynx considered by the panel, as 95 percent of patients in the ExteNET trial who received the tyrosine kinase inhibitor experienced the adverse event (AE), including grade 3 diarrhea in 40 percent of patients. Diarrhea led to study discontinuation for 16.8 percent of patients. However, results from the ongoing phase 2 CONTROL trial suggest that antidiarrheal prophylaxis can control the occurrence and severity of diarrhea among patients receiving Nerlynx.

Beyond safety, another concern addressed by ODAC were the numerous amendments to the ExteNet study protocol, the effects of which included enriching the study population with high-risk patients; reducing the trial follow-up interval from five years to two years—changing the assessment from event-driven to time driven; and implementing a reconsent process to increase follow-up to five years after randomization. However, despite the amendments, data from sensitivity analyses were enough to assure the panel that Nerlynx had a positive effect in this patient population.

The FDA will now make its final decision on Nerlynx. The agency is not required to follow the ODAC recommendation.

Explaining his “yes” vote, panelist Andrew D. Seidman, M.D., said, “The results look durable…[and] I’m reassured by the rigorous statistical analyses that were applied, given the changes in study design along the evolution of the trial.”

“I think that physicians will select patients very carefully when using [neratinib],” added Seidman, an attending physician in the Breast Cancer Medicine Service, Memorial Sloan Kettering Cancer Center, and a professor of medicine at Weill Cornell Medical College.

In the ExteNET study, 2840 patients who remained disease-free following one year of treatment with adjuvant Herceptin and chemotherapy were randomized to Nerlynx (1,420 patients) or placebo (1,420 patients). Nerlynx was administered for 12 months at 240 mg per day. In the final study amendment, the primary endpoint was iDFS at two years and 28 days from randomization.

The median age of patients in the study was 52 years and approximately 24 percent had node negative disease, with 47 percent of patients having 1 to 3 positive nodes and 30 percent having four or more positive nodes. Anthracyclines were administered as adjuvant chemotherapy in the majority (77 percent) of patients. Appropriate endocrine therapy was administered to 94 percent of patients with HR-positive breast cancer.

Beyond the primary analysis data, additional follow-up data from two to five years post randomization were submitted from an exploratory iDFS analyses that occurred after the implementation of a reconsent process.

Overall, 2,117 (74.5 percent) of the 2,840 primary analysis patients reconsented, including 1028 patients in the Nerlynx cohort and 1089 patients on the placebo arm. Baseline characteristics were similar between the reconsented and primary analysis populations, as well as between the two cohorts in the reconsent analysis.

In the updated analysis, the two-year iDFS was 94.3 percent in the Nerlynx arm versus 91.7 percent in the placebo group. The five-year iDFS rates were 90.2 percent and 87.7 percent, respectively.

The safety analysis for the primary ExteNET data review was based on 1,408 patients from the Nerlynx arm and 1,408 patients from the placebo arm. The median duration of exposure to Nerlynx and placebo was 11.6 months and 11.8 months, respectively.

In the Nerlynx arm, grade 3/4 AEs occurred in 50 percent of patients and led to treatment discontinuation in 28 percent of patients. The most common AEs leading to Nerlynx discontinuation were diarrhea (16.8 percent), vomiting (3.8 percent) and nausea (2.8 percent). In the Nerlynx arm, 7.3 percent of patients experienced non-fatal serious AEs, the most frequent being diarrhea in 22 patients versus one patient in the placebo arm.

Although patients in the ExteNET study were not required to receive antidiarrheal prophylaxis, the ongoing, open-label phase 2 CONTROL trial examined the preventative measure in HER2-positive patients who received Nerlynx for one year along with antidiarrheal prophylaxis given during the first two 28-day treatment cycles.

At the Jan. 13, 2017, data cutoff, 137 patients had received prophylaxis with loperamide alone, 64 patients had received loperamide plus budesonide, and 10 patients had received loperamide plus colestipol. The median duration of Nerlynx treatment for the three cohorts was 9.07 months, 2.83 months, and 0.56 months, respectively.

Comparing the loperamide-alone CONTROL cohort to the safety data cohort from the ExteNET trial (1,408 patients), the incidence of all-grade diarrhea was 77 percent versus 95 percent, respectively. The rates of grade 3 diarrhea were 31 percent versus 40 percent, respectively.

The rate of dose reductions (7.3 percent vs 26.4 percent) and holds (13.9 percent vs 33.9 percent) due to diarrhea were lower in the patients who received loperamide. However, the rate of discontinuation due to diarrhea was higher in the loperamide cohort at 20.4 percent versus 16.8 percent with Nerlynx alone.