Loss-of-function mutations in the PBRM1 gene are more common among patients with metastatic clear cell renal cell carcinoma (ccRCC) who have clinical responses to immune checkpoint inhibitors (ICIs), but the mutations fall a little short of serving as clinical biomarkers, investigators say.
Among patients with metastatic ccRCC enrolled in a randomized, phase 3 clinical trial comparing nivolumab (Opdivo) with everolimus (Afinitor) in patients who had previously received antiangiogenic therapy, patients with responses to nivolumab were more than twice as likely as nonresponders to have a truncating loss-of-function mutation in PBRM1, reported, and colleagues from the Dana-Farber Cancer Institute in Boston.
Patients with PBRM1 mutations were also more likely to have a clinical benefit (complete/partial response, or stable disease with tumor shrinkage and progression-free survival [PFS] of at least 6 months) as well as better PFS and overall survival than patients without mutations.
“The association of PBRM1 truncating mutations with response to anti–[programmed death-1] therapy was confirmed in an independent ccRCC cohort. However, key limitations restrict use of PBRM1 mutations as a clinical biomarker,” they wrote in a research letter to JAMA Oncology.
Those limitations include an only modest effect of mutations on response and survival, a lack of evidence for a PBRM1 mutation effect in the first-line setting, and a possible association between mutations and benefit from prior antiangiogenic therapies, they acknowledged.
Dr. Choueiri was a coauthor of a previous study of genomic correlates of response to ICIs in ccRCC, which found that PBRM1 loss may influence response to checkpoint inhibitors by altering global tumor-cell expression.
In the current study, he and his colleagues looked at archival tumor tissue from an independent cohort of 382 patients who were part of a larger phase 3 trial. Of this group, 189 were treated with nivolumab, and 193 were treated with everolimus. PBRM1 mutations were identified in 55 of the nivolumab-treated patients (29%) and in 45 of the everolimus-treated patients (23%).
When they looked at clinical responses, they found that 15 of 38 patients with response to nivolumab (39%) had truncating PBRM1 mutations, compared with 16 of 74 nonresponding patients (22%), which translated into an odds ratio for response of 2.34 (P = .04).
Similarly, PBRM1 mutations among nivolumab-treated patients were significantly associated with clinical benefit (OR, 2.14; P = .0497), PFS (hazard ratio for progression, 0.67; P = .03), and overall survival (HR, 0.65; P = .03).
In contrast, there were no significant associations in everolimus-treated patients between PBRM1 mutations and either response, PFS, or overall survival.
“The concomitant presence of other cellular or molecular features may further influence the findings described herein. Nonetheless, this validated association between PBRM1 alterations and ICI response in a large randomized study represents a further step toward the development of genomic predictors for immunotherapies in advanced RCC,” the investigators concluded.
The study was supported by Department of Defense Congressionally Directed Medical Research Programs and Bristol-Myers Squibb. Dr. Choueiri disclosed personal fees from Bristol-Myers Squibb and fees and grants from other companies. One coauthor is a Bristol-Myers Squibb employee and shareholder.
SOURCE: Choueiri TK et al. JAMA Oncol. 2019 Sep 5.