Patients with prostate cancer have typically derived very limited benefit from immune checkpoint inhibitors to date, with most clinical trials showing underwhelming results and failing to meet their primary endpoints. Despite this, some patients do respond, and a new phase 2 trial testing CTLA-4 blocker ipilimumab in patients with metastatic, castration-resistant prostate cancer (mCRPC) has attempted to find out why this minority of patients do well.

“We had 2 phase 3 clinical trials with ipilimumab, where overall survival didn’t meet statistical significance for the study,” said Padmanee Sharma, MD, PhD, professor of genitourinary medical oncology and immunology at MD Anderson Cancer Center in Houston, Texas, and senior author of the study. “In the phase 3 clinical trials we originally ran, there were quite a few patients who did very well, their PSA was undetectable, and their tumors regressed significantly,” Dr Sharma added.

The positive responses seen in these trials are much less frequent than in cancers typically known to respond well to immune checkpoint inhibitors such as melanoma and lung cancers — despite this, some prostate cancer patients do appear to respond.


Continue Reading

“We’ve always seen these patients, but we’ve never had them in enough numbers to reach significance targets overall in phase 3 trials,” said Dr Sharma.

The new results published in Science Translational Medicine1 looked at a small cohort of 29 patients with mCRPC enrolled in the trial (ClinicalTrials.gov Identifier: NCT02113657) between January 2015 and May 2018 who received at least 1 dose of ipilimumab. The median follow-up of patients after the first treatment was 45.5 months.

“Not enough data was gained from the original phase 2 trial, in which most patients didn’t respond. We wanted to look more from a research standpoint — how and why CTLA-4 blockade would or would not work in metastatic prostate cancer,” said Dr Sharma.

Median progression-free survival (PFS) was 3 months, with overall survival of 24.3 months, but the researchers identified 2 distinct cohorts with “favorable” and “unfavorable” characteristics. The favorable group consisted of 9 patients with PFS of more than 6 months and overall survival (OS) of more than 1 year, whereas the unfavorable group consisted of 10 patients with PFS of less than 6 months and OS of less than 1 year.

“Because the best objective response in this study was stable disease, it’s reasonable to use PFS and OS as clinical endpoints to divide patients, although many variables (not just clinical activity of ipilimumab) will influence OS,” said Xiao Wei, MD, medical oncologist specializing in genitourinary cancers at the Dana-Farber Cancer Institute in Boston, Massachusetts. “By comparing ‘extremes’ in clinical outcomes, investigators may be able to better identify potential predictive biomarkers. The sacrifice to this approach is losing power in an already small study,” added Dr Wei.

Related Articles

To identify possible biomarkers predictive of response, the researchers looked at pretreatment samples from patients in these 2 cohorts, using immunohistochemistry and RNA sequencing analyses to identify parameters that were different between the groups.

“They found CD8+ T-cell density and interferon gamma signatures were high in the favorable, but not unfavorable, cohort. They suggest that these 2 parameters could be taken into account pre-treatment for selecting patients who are likely to benefit from treatment with ipilimumab,” said Astero Klampatsa, PhD, team leader in thoracic oncology immunotherapy at The Institute of Cancer Research in London, U.K. “One problem is that the cohorts are generally very small — so everything about the paper is beautiful, but it has to be done on 10 times these patients to really prove the findings,” Dr Klampatsa noted.

Perhaps most crucially, the researchers showed that T cells isolated from patients in the favorable — but not unfavorable — cohort were able to recognize and respond to neoantigens in the tumor, despite their tumors having low tumor mutational burdens (TMBs).