The FDA’s Oncologic Drugs Advisory Committee voted 7 to 5 in favor of olaparib tablets for an indication as a maintenance treatment of adult patients with deleterious or suspected deleterious BRCA-mutant metastatic pancreatic adenocarcinoma whose disease has not progressed on frontline platinum-based chemotherapy.
The FDA’s Oncologic Drugs Advisory Committee (ODAC) voted 7 to 5 in favor of olaparib (Lynparza) tablets for an indication as a maintenance treatment of adult patients with deleterious or suspected deleterious BRCA-mutant metastatic pancreatic adenocarcinoma whose disease has not progressed on frontline platinum-based chemotherapy.
The FDA scheduled an Oncologic Drugs Advisory Committee hearing to discuss a supplemental new drug application (sNDA) for olaparib in this setting, which was based on results from the phase III POLO trial. Findings showed a progression-free survival (PFS) benefit with olaparib compared with placebo in patients with germline BRCA-mutated metastatic pancreatic cancer.
The median PFS with the PARP inhibitor was 7.4 months compared with 3.8 months with placebo (HR, 0.53; 95% CI, 0.35-0.82; P = .0038).1,2 In addition, after 2 years, 22.1% of patients had no disease progression versus 9.6% of those who received placebo.
“To summarize the favorable aspects leading to ‘yes’ votes, there appears to be activity [with olaparib]. It's definitely prolonging progression-free survival; there doesn't seem to be much question about that,” said Philip C. Hoffman, MD, professor of medicine at The University of Chicago Section of Hematology/Oncology, who voted in favor of olaparib. “Elaborate toxicity is not major for most patients—certainly not life-threatening—and so from the standpoint of risks [and] benefits of taking the drug, there seems to be more benefits than detriments.”
“The ‘no’ votes raise the concern about the lack of overall survival data,” Hoffman added. “People would like to have additional data to consider, and this is a relatively modest amount of benefit.”
The committee sought to discuss whether there is compelling evidence of an accurate and reliable treatment effect on olaparib, due to its small trial size and missing data for some patients on disease burden; due to imbalances in known prognostic factors; and due to limitations of imaging technology to accurately measure tumor burden.3
The agency’s panel also highlighted whether the observed PFS effect is clinically meaningful, as a benefit in OS has yet to be determined and olaparib is linked with higher grade 3/5 adverse events than placebo, at 40% versus 23%, and serious AEs (24% vs 15%).
“I would to recognize this [decision] was a challenge because I don’t want our patients to be harmed by a drug, and we don’t want our patients to refrain from platinum-based therapy,” said Diane Reidy Lagunes, MD, a medical oncologist at Memorial Sloan Kettering Cancer Center, who also voted in favor of olaparib. “This drug is not a ‘home run’ for the vast majority of patients, but I think it was done in a way that we can enrich our patient population that will derive benefit, and [these patients] deserve to have that opportunity to have the drug.”
In the randomized, double-blind, placebo-controlled, phase III POLO trial, researchers evaluated the efficacy of olaparib as maintenance therapy in 154 patients who had a germline BRCA1 or BRCA2 mutation and metastatic pancreatic cancer, which had not progressed during first-line platinum-based chemotherapy.
Patients were randomized 3:2 to receive oral olaparib tablets at 300 mg twice daily as maintenance therapy (n = 92) versus placebo, also twice daily (n = 62). Randomization occurred within 6 weeks following last chemotherapy dose and olaparib/placebo treatment began within 4 to 8 weeks of the last chemotherapy dose. Following randomization, patients had weekly clinical visits for the first 4 weeks of treatment, then every 4 weeks while on study treatment.
However, the FDA was uncertain in whether the POLO trial was adequately controlled given its small population, which may not ensure balance for all important prognostic factors.
Some characteristics differed between the two arms, such as ECOG performance status of 0 (71% with olaparib vs 61% with placebo), no evidence of disease at study entry (5% vs 0%, respectively), and ages ≥65 years (21% vs 30%), as well as missing data for baseline disease burden (1% vs 6%). The missing data results in a higher rate of earlier censoring in the placebo arm, the FDA noted.
The median duration of therapy was 6 months for those taking olaparib and 3.7 months for people who received placebo. Treatment continued until objective radiological disease progression. Following progression, patients were followed for second progression every 8 weeks, and then for survival until final analysis.
Patients who were eligible for enrollment were previously treated for metastatic disease and had not progressed following completion of ≥16 weeks of frontline platinum-based chemotherapy. Additionally, patients had to have a known deleterious or suspected deleterious germline BRCA mutation; those who were previously treated with a PARP inhibitor were excluded.
The primary endpoint was PFS by blinded independent central review. Secondary endpoints were overall survival (OS), time from randomization to second progression or death (PFS2), objective response rate (ORR), disease control rate (DCR), safety, and tolerability.
Patients in the treatment arm were a median age of 57 years, 58% were male, and 71% had an ECOG performance status of 0. Two-thirds of patients had BRCA2 mutations, and the remainder had BRCA1 mutations.
Median PFS was consistent irrespective of response to prior platinum-based chemotherapy (complete/partial HR, 0.62; stable disease HR, 0.50). At 6, 12, 18, and 24 months, the percentage of patients who were progression-free in the olaparib arm was more than twice that in the placebo arm (6-month PFS, 53% vs 23%).
The FDA stated, however, that the PFS curves separated after approximately 53% of patients had PFS events or were censored and corresponded to a modest difference in median PFS of 3.6 months.
In the ODAC briefing document, the FDA cited that based on 10,000 bootstrap iterations using a basic nonparametric bootstrap, the estimated median difference in PFS between the 2 arms is 3.2 months (95% CI, 0.3, 7.3). The lower bound of the 95% confidence interval specifies that the true difference in median PFS between the 2 groups could be 0.3 months, or 9 days.
Moreover, the FDA noted that there was a limitation of current imaging technology to accurately measure locoregional tumors. The agency noted a review of a meta-analysis of 333 patients pancreatic cancer that assessed the diagnostic performance of imaging technologies to detect recurrent disease; however, contrast-enhanced CT scans demonstrated “moderate” diagnostic accuracy; it also discussed discordant tumor staging by imaging by CT, MRI, and/or ultrasound. In POLO, serial imaging was required with CT or MRI scans at baseline and for tumor measurement assessments; 18-fluorodeoxy glucose (FDG)—PET scans were allowed to identify new lesions only if a baseline FDG-PET scan was performed and correlation was required by either CT or MRI imaging.
“FDA acknowledges that a difference in PFS favoring the olaparib arm has been demonstrated in the POLO trial,” the agency stated in the briefing document. “However, there is uncertainty that the observed magnitude of the treatment effect on PFS accurately represents the true treatment effect, given potential issues with the reliability and accuracy of imaging technologies in this disease setting, the limited patient experience, and potential for imbalances in prognostic factors affecting tumor response between arms.”
Additional results showed that the ORR was 23.1% with olaparib compared with 11.5% in the placebo arm (odds ratio, 2.30); 11.1% (n = 2) of patients on olaparib achieved a complete response compared with 0 on placebo. The median duration of response was 24.9 months in the olaparib arm versus 3.7 months with placebo.
After 1 year, 33.7% of patients receiving olaparib showed no signs of disease progression compared with 14.5% of those who received a placebo.
An interim analysis of OS at data maturity of 46% demonstrated no difference between arms with a median OS of 18.9 months with olaparib and 18.1 months with placebo (HR, 0.91; 95% CI, 0.56-1.46; P = .68). A final event-driven OS analysis is expected to occur in 2020 once 106 deaths have occurred out of 154 patients. Eighty-four deaths have occurred as of October 27, 2019.4
“Based upon the data observed at the interim analysis and the protocol stated assumptions for the remaining data, the probability of seeing a statistically significant difference in OS at the final analysis is 9%, strongly suggesting a statistically significant result for OS is not expected,” the briefing document stated. As of October 17, 2019, 84 deaths have occurred.
The document also stated that due to the rarity of the germline BRCA-mutant patient population in pancreatic cancer, it is not feasible to power a study to reliably detect a statistically significant different in OS.
Additionally, olaparib extended the chemotherapy-free interval and delayed time to first subsequent therapy (TFST). While TFST data were 68% mature, results showed that the median TFST was 8.6 months with olaparib compared with 5.7 months for placebo, which was determined to be clinically meaningful (HR, 0.50; P = .0013).
At the January 15, 2019 data cutoff date, 30% of patients on the olaparib arm remained on treatment compared with 12.9% who were on placebo and were not eligible for subsequent chemotherapy. A total 48.9% and 74.2% on olaparib and placebo, respectively, received subsequent therapy.
The DCR was 53.3% with olaparib versus 37.1% with placebo, and PFS2, which was at 46% maturity, suggested a trend favoring olaparib at 13.2 months versus 9.2 months, respectively (HR, 0.76; 95% CI, 0.46-1.23; P = .26).
Regarding safety, there were no new safety signals with olaparib. Grade ≥3 adverse events (AEs) occurred in 40% of patients in the olaparib arm compared with 23% of those in the placebo arm (95% CI, −0.02% to 31%). In total, 5.5% and 1.7% of patients, respectively, discontinued treatment due to an AE.
At 1, 2, and 3 years, 25.3%, 9.9%, and 3.3% of patients remained on olaparib therapy, respectively, compared with 6.7%, 3.3%, and 0% of those on placebo. Treatment discontinuation in both groups due to AEs was 5.5% and 1.7% with olaparib and placebo, respectively.
Health-related quality of life (HRQoL) findings of the POLO trial were presented during the 2019 ESMO Congress, which showed that there was no difference between arms for global HRQoL.5 The adjusted mean difference for physical function scale did not reach the threshold that was considered to be clinically meaningful, and the global HRQoL and physical function remained moderately stable over time. The global HRQoL score was 29.2% with olaparib versus 22.4% with placebo.
There was no difference in time to sustained clinically meaningful deterioration (TSCMD) at 21.2 months for olaparib versus 6.0 months for placebo (HR, 0.72; 95% CI 0.41-1.27; P = .25). Overall, researchers noted that olaparib appears to provide benefit to patients by delaying time to worsening of pain, which is a key symptom of metastatic pancreatic cancer.
Anthony D. Sung, MD, assistant professor of medicine at Duke University School of Medicine, and Duke Adult Blood and Marrow Transplant Clinic, was one of the 5 dissenting ODAC votes.
“As stated, I believe PFS is an important endpoint, but it needs to be supported by a difference in overall survival—or at least a trend toward a difference in overall survival, or some sort of hint in that direction, or additional evidence, like healthcare-related quality of life. I think, even [with] a year or 2 with additional data, could I see myself changing my vote? Sure. But with the [data] in front of me, I do not think it is sufficient.”
National Comprehensive Cancer Network guidelines recommend olaparib as a maintenance treatment for patients with germline BRCA-mutant metastatic pancreatic cancer, based on the POLO findings.