Investigator-assessed PFS in both an intention-to-treat (ITT) analysis and an efficacy analysis that excluded patients with BRCA reversion mutations was 7.4 months in the rucaparib arm, compared with 5.7 months in patients who received either platinum-based chemotherapy or weekly paclitaxel.
Among the 23 patients with BRCA reversion mutations, however, investigator-assessed PFS was 2.9 months with rucaparib and 5.5 months with chemotherapy.
Overall survival data were not mature at the time of data cutoff in September 2020.
“Although the numbers are very small, the results suggest that presence of a BRCA reversion mutation may predict a reduced benefit from rucaparib,” said Rebecca Kristeleit, MBChB, PhD, of Guy’s and St. Thomas’ NHS Foundation Trust in London.
She presented the findings from ARIEL4 at the Society of Gynecologic Oncology’s Virtual Annual Meeting on Women’s Cancer (Abstract 11479).
Invited discussant Ursula Matulonis, MD, of the Dana-Farber Cancer Institute in Boston, commented that the “BRCA reversion mutation data from ARIEL4 is intriguing. Strategies to overcome and better understand this type of resistance mechanism are needed.”
Study rationale and details
Rucaparib is approved as monotherapy for patients with BRCA-mutated, relapsed ovarian cancer who have received at least two prior lines of platinum-based chemotherapy. The approval was based on results of two phase 1/2 studies. ARIEL4 is a phase 3 confirmatory study, designed in consultation with both the U.S. Food and Drug Administration and the European Medicines Agency.
Women with relapsed, high-grade epithelial ovarian, fallopian tube, or primary peritoneal cancer with deleterious germline or somatic BRCA mutations were eligible for enrollment in ARIEL4. The patients had to have received at least two lines of chemotherapy, including at least one platinum-based regimen, with no prior PARP inhibitor or single-agent paclitaxel treatment.
Overall, 95% of patients had epithelial ovarian cancer, 3% had fallopian tube cancer, and 2% had primary peritoneal cancer. About 90% of cancers were serous in histology. Most patients (84%) had germline BRCA mutations, 16% had somatic mutations, and the status was unknown in the remaining patients.
Patients were randomized on a 2:1 basis to receive rucaparib at 600 mg twice daily (n = 233) or chemotherapy (n = 116), stratified by platinum sensitivity status. Patients assigned to chemotherapy whose disease was considered platinum resistant or partially platinum sensitive were assigned to weekly paclitaxel. Patients with fully platinum-sensitive disease were assigned to platinum-based single-agent or doublet chemotherapy. Treatment cycles were 28 days.
On radiologically confirmed disease progression or unacceptable toxicity, patients assigned to chemotherapy had the option to cross over to the rucaparib arm. The follow-up portion of the study began 28 days after the last treatment dose, with visits every 8 weeks thereafter.
Baseline characteristics in the ITT population were similar between arms. There were 13 patients in the rucaparib arm and 10 in the chemotherapy arm who had BRCA reversion mutations and were excluded from the efficacy population.
Efficacy and safety
Investigator-assessed PFS in the efficacy population was a median of 7.4 months with rucaparib and 5.7 months with chemotherapy, translating to a hazard ratio (HR) of 0.64 (P = .001). In the ITT population, the respective median PFS intervals were identical, although with a slightly less favorable HR of 0.67 (P = .002). In the 23 patients with BRCA reversion mutations, the median PFS was worse with rucaparib, at 2.9 months, compared with 5.5 months for chemotherapy. This translated to a HR of 2.77, although the 95% confidence interval was wide and crossed 1, likely due to the small sample size.
Among patients who had measurable disease at baseline, the overall response rate in the efficacy population was 40.3% with rucaparib and 32.3% with chemotherapy, a difference that was not statistically significant (P = .13). The overall response data were similar in the ITT population (37.9% and 30.2%, respectively).
In the efficacy population, the duration of response was significantly longer in the rucaparib arm, at a median of 9.4 months versus 7.2 months (HR, 0.59; 95% CI, 0.36-0.98). The respective median response durations were identical in the ITT population, but the HR was 0.56 (95% CI, 0.34-0.93).
In both the efficacy and ITT populations, global health status was virtually identical and unchanged from baseline in both treatment arms through cycle 7.
Treatment-emergent adverse events (TEAEs) were more frequent with rucaparib. The most common TEAEs in the rucaparib and chemotherapy arms, respectively, were anemia/decreased hemoglobin (53.9% and 31.9%), nausea (53.4% and 31.9%), asthenia/fatigue (49.6% and 44.2%), ALT/AST increase (34.5% and 11.5%), and vomiting (34.1% and 16.8%).
In all, 8.2% of patients in the rucaparib arm and 12.4% of those in the chemotherapy arm discontinued therapy due to TEAEs.
Four patients in the rucaparib arm developed myelodysplastic syndrome or acute myeloid leukemia – one during treatment and three during follow-up. There were no cases of myelodysplastic syndrome or acute myeloid leukemia in patients who received chemotherapy.
“Data from ARIEL4 fits the paradigm that single-agent activity of PARP inhibitors in BRCA-mutated, recurrent ovarian cancer may be comparable to chemotherapy, and may, at times, be superior, depending on the study population, trial design, and treatment for control patients,” Dr. Matulonis said.
The study was funded by Clovis Oncology. Dr. Kristeleit disclosed relationships with Clovis, Roche, and Tesaro. Dr. Matulonis disclosed relationships with Novartis, Merck, and Immunogen.