Two alternatives to paclitaxel for recurrent ovarian cancer failed to slow disease progression in separate randomized trials.
In one trial paclitaxel and cabozantinib (Cometriq) led to similar median progression-free survival (PFS) of 5 to 6 months. Analyses of overall survival, response to therapy, and event-free survival (EFS) suggested that patients treated with cabozantinib might have fared worse than their counterparts in the paclitaxel arm, Ursula A. Matulonis, MD, of Dana-Farber Cancer Institute in Boston, reported here at the Society of Gynecologic Oncology meeting.
“At the dose and schedule examined in this study, cabozantinib was not worthy of further investigation,” said Matulonis.
The second trial showed no PFS benefit with the addition of an oncolytic virus to paclitaxel for recurrent ovarian cancer.
Cabozantinib inhibits both VEGF receptor and hepatocyte growth factor receptor (MET) and has FDA approval for treatment of metastatic medullary thyroid cancer. In a preliminary clinical trial involving women with platinum resistant/refractory ovarian cancer, treatment with cabozantinib resulted in an overall response rate of about 30%, increasing to 40% in patients with platinum-sensitive recurrent disease.
Subsequently, a phase II trial demonstrated response rates of 18% to 28% in recurrent ovarian cancer across a range of platinum-free intervals.
Matulonis reported findings from the Gynecologic Oncology Group (GOG) 186K randomized trial involving patients with persistent or recurrent ovarian, fallopian, or primary peritoneal cancer. Investigators enrolled patients who had received as many as three prior regimens and randomized them to cabozantinib 60 mg BID or dose-dense paclitaxel (80 mg/m2/week).
The primary endpoint was PFS, assessed by CT or MRI prior to cycle five/week 16 and again prior to cycle nine/week 32. Secondary objectives included toxicity, response rate, overall survival, and translational studies to correlate c-MET expression and copy number with outcome (still ongoing).
Data analysis included 111 patients who had a median follow-up duration of about 14 months. The results showed a median PFS of 5.3 months with cabozantinib and 5.5 months with weekly paclitaxel. By conventional Cox proportional hazards analysis, the two regimens were similar (HR 1.11, 90% CI 0.766-1.614).
Analysis of secondary outcomes suggested that cabozantinib did not perform as well as paclitaxel, said Matulonis. The cabozantinib group had a median overall survival of 19.4 months, whereas the median had yet to be reached in the paclitaxel arm. A unstratified analysis yielded a hazard ratio of 1.66 for the comparison of cabozantinib versus paclitaxel (P=0.15).
After stratification for platinum-free interval, measurable versus nonmeasurable disease, and bevacizumab exposure, however, patients in the cabozantinib had significantly worse overall survival (HR 2.27, 90% CI 1.17-4.41, P=0.04).
Response rates were 8.3% with cabozantinib and 28.3% with paclitaxel, which translated into an odds ratio of 0.23 for cabozantinib versus paclitaxel. Analysis of EFS also yielded a significantly increased hazard for cabozantinib (HR 1.81, 90% CI 1.24-2.63).
Toxicity was similar in the two groups, with the exception of gastrointestinal adverse events, which occurred more often with cabozantinib. Fewer patients in the cabozantinib group discontinued treatment because of disease progression (49% versus 70%), but more patients on cabozantinib discontinued because of toxicity (30% versus 6%), Matulonis reported.
Another randomized GOG trial also concluded with a decision not to continue clinical investigation of a therapy in recurrent ovarian cancer. The 186H trial involved a patient population similar to the one enrolled in 186K. More than 70% had received two or three prior regimens for recurrent disease, and more than 40% had prior exposure to bevacizumab, said David E. Cohn, MD, of Ohio State University in Columbus.
The trial evaluated the PFS impact of giving dose-dense paclitaxel with oncolytic reovirus (Reolysin). Oncolytic viruses offer an attractive approach to cancer therapy because of their selectivity in tumor-cell killing and ability to induce systemic antitumor immunity, Cohn said in a preface to the results. Reovirus is a double-stranded RNA virus that is highly prevalent in the community but rarely causes symptoms.
In normal cells, reovirus activates the PR pathway, inhibits protein translation, and inhibits viral production. In Ras-transformed cells, reovirus does not activate PKR and facilitates viral infection and cell lysis, Cohn continued.
Preclinical and clinical studies showed that the oncolytic virus led to viral replication in peritoneal tumors, had synergistic tumoricidal activity in combination with paclitaxel, and was safe in humans when administered along with paclitaxel.
GOG 186H was a phase II randomized trial that had a primary objective of PFS. Investigators enrolled 108 patients and randomized them to paclitaxel plus placebo or reovirus.
The primary analysis showed a median PFS of 4.3 months with paclitaxel alone and 4.4 months with the oncolytic virus. The paclitaxel-placebo group had a median overall survival of 15.0 months compared with 12.9 months with reovirus.
Consistent with previous studies, the reovirus did not appear to add substantially to paclitaxel toxicity. The only exceptions were grade 4 neutropenia (12% with reovirus, 0% without) and grade 3+ respiratory adverse events (25% versus 2%).
“Adding REO to weekly paclitaxel did not significantly change the toxicity profile of paclitaxel alone,” said Cohn. “However, REO plus paclitaxel did not sufficiently reduce the hazard of progression or death compared with paclitaxel alone to warrant further investigation.”
To read this entire article on MedPageToday.com, click here.