This article is more than 5 years old. Information may no longer be current.
Olaparib fails to extend OS in advanced gastric cancer
Click Here to Manage Email Alerts
COPENHAGEN, Denmark — The addition of olaparib to paclitaxel demonstrated a trend toward an OS benefit in patients with advanced gastric cancer independently of ATM protein status, according to results of the phase 3 GOLD trial presented at the European Society for Medical Oncology Congress.
Although olaparib conferred greater benefits in PFS, overall response rate and health-related quality of life appeared among patients with ATM–negative tumors, overall, the study failed to demonstrate a statistically significant improvement in OS based on the prespecified study criteria.
A previously published phase 2 study showed that the addition of the PARP inhibitor olaparib (Lynparza, AstraZeneca) to paclitaxel significantly extended OS in Asian patients with advanced gastric cancer, especially among patients with ATM–negative tumors.
Based on these results, Yung-Jue Bang, MD, PhD, professor of medical oncology at Seoul National University College of Medicine and president of Biomedical Research Institute at Seoul National University Hospital, and colleagues conducted a phase 3 study to evaluate the addition of olaparib to paclitaxel in this patient population.
“ATM is a key sensor of the DNA damage response pathway, which might explain greater olaparib activity in patients whose DNA damage response mechanisms were not being activated by ATM,” Bang said during his presentation.
The analysis included data from 525 patients with advanced gastric cancer who had progressed on first-line treatment. Ninety-four patients (18%) had ATM–negative tumors.
Researchers randomly assigned patients 1:1 to receive 80 mg/m2 IV paclitaxel (days 1, 8, 15 per 28-day cycle) with placebo (n = 262; ATM negative, n = 46) or with 100 mg olaparib twice daily (n = 263, ATM negative, n = 47).
OS for all patients, and in ATM–negative patients, served as the co-primary endpoints. PFS, objective response rate and safety served as secondary endpoints.
In the full analysis set, median OS was 8.8 months among patients assigned olaparib and 6.9 months among patients assigned placebo (HR = 0.79; 97.5% CI, 0.63-1). The .0262 P value did not meet the study’s predefined criteria for significance of P < .025.
Among ATM–negative patients, median OS was 12 months among those assigned olaparib and 10 months among those assigned placebo (HR = 0.73; 97.5% CI, 0.4-1.34).
A subgroup analysis showed olaparib demonstrated a greater OS benefit in patients who underwent full or partial gastrectomy.
“This finding is consistent with what was observed in the previous phase 2 study; however, at this time, we don’t know what makes this difference,” Bang said.
Olaparib tended to demonstrate a benefit in terms of PFS (HR = 0.74; 97.5% CI, 0.45-1.29), ORR (37.5% vs. 16.1%; P = .0309) and time to deterioration of health-related quality of life (HR = 0.63; 97.5%; 0.34-1.16) among ATM–negative patients.
“These endpoints were numerically improved with the addition of olaparib, especially among the ATM–negative population,” Bang said.
More patients assigned olaparib experienced grade 3 or worse adverse events (78% vs. 62%), the most common of which was neutropenia (30% vs. 23%). Further, serious adverse events were more common in the olaparib arm (35% vs. 25%), as were adverse events leading to study discontinuation (16% vs. 10%).
“There was a trend toward a survival improvement in the overall population, and numerically improvements in secondary endpoints, particularly in ATM–negative patient,” Bang said. “Please note that, in combination with paclitaxel, we used low-dose olaparib, at 100 mg twice daily. The monotherapy dose currently under investigation in other olaparib phase 3 studies is 300 mg twice daily.” – by Alexandra Todak
Reference:
Bang Y-J, et al. Abstract LBA25. Presented at: European Society for Medical Oncology Congress; Oct. 7-11, 2016; Copenhagen, Denmark.
Disclosure: Bang reports research funding from and consultant roles with AstraZeneca. Please see the abstract for a list of all other researchers’ relevant financial disclosures.
Perspective
Back to Top
Yelena Y. Janjigian, MD
Preclinical and clinical data have shown that PARP inhibition is important, particularly in ATM–deficient cells. BRCA mutations are relatively rare in gastric cancer, so this subset of ATM–deficient cells were noted to be sensitive to olaparib in gastric cancer. In addition, loss of TP53 function is crucial in PARP inhibitor response in ATM–negative tumors. Beyond ATM immunohistochemistry, development of gene signatures predicting “BRCA-ness” in gastric cancer is important for the future PARP inhibitor studies.
A phase 2 trial in the second-line setting randomly assigned patients to receive paclitaxel with or without olaparib. It’s important to notice that these patients were particularly enriched for ATM–negative cohorts. Patients were preselected and put on the study, and 50% of patients in this study were ATM negative.
Overall, this phase 2 study was negative. The prespecified endpoint was PFS. Although there was some discussion about tumor biology and colony stimulation, there was no real clear biologic explanation about why the median OS in the overall population was 13.1 months. The OS for patients in the ATM–negative cohort, the patients most likely to benefit, was not even reached. With that data showing very promising OS in second-line setting, this phase 3 study was not surprising and certainly logical.
Patients on this phase 3 study, also performed in Asia, were treated with low-dose olaparib — half the standard dose — with paclitaxel, due to concerns of neutropenia and other toxicities.
Comparing the data, even though this was a large phase 3 trial, there were only 17 more ATM–negative patients who were treated with olaparib than in the phase 2 trial. Only 18% of patients were ATM negative on the phase 3 trial, which is the general incidence of ATM negativity in patients who just walk into the clinic and are not enriched.
Unfortunately, the signal that was seen in OS in the overall study population was not seen in this phase 3 trial. Once again, we are disappointed with phase 3 results after promising phase 2 results. Median OS was only 8.8 months. Even though there was a trend toward improvement, it was not statistically significant. The ATM–negative cohort did do better, which could be promising. One could argue this was underpowered based on the small sample size and short follow-up.
The study was not appropriately enriched for the biomarker, and I think that’s the real crux of the problem. This lack of enrichment likely underpowered the ATM–negative arm and limited the power of the study.
Click Here to Manage Email Alerts