Michael Schweizer, MD, on New Research About CDK12 Mutations in Prostate Cancer

CDK12 mutations have been proposed as a potential biomarker for response to immune checkpoint blockade in prostate cancer patients. CDK12 mutations, thought to be likely early events in the disease, also occur across a variety of other malignancies, which raises the possibility that CDK12 alterations could be predictive for immune responses regardless of tumor type.

A recent multi-institution retrospective study in JCO Precision Oncology set out to determine the clinical behavior of CDK12-mutated prostate cancer. The goal was to evaluate the clinical outcomes of patients harboring CDK12 alterations after the use of standard prostate cancer therapies, as well as the outcomes after off-label checkpoint inhibitor (CPI) therapy.

The study, led by Michael Schweizer, MD, of the University of Washington and the Seattle Cancer Care Alliance, and Gavin Ha, PhD, of Fred Hutchinson Cancer Research Center in Seattle, included 52 patients with CDK12-mutated prostate cancer, with 27 patients (52%) with detected biallelic CDK12 alterations. After a median follow-up of 8.2 years, unconfirmed PSA50 response rates (i.e., a decline of at least 50% in prostate-specific antigen levels from baseline) with front-line abiraterone or enzalutamide for castration-resistant prostate cancer were 65% and 75%, respectively. In contrast, among the 19 patients who received CPI therapy, PSA50 responses were 11%.

In the following interview, Schweizer elaborated on the team's findings.

What does your study add to the literature about the role of CPIs in patients with CDK12-mutated prostate cancer?

Schweizer: If you look at the data in total, we only had a small number of patients treated with CPIs and the overall response rate is relatively modest, with only 11% of patients showing a PSA50 decline with CPIs. This is not as encouraging as we had hoped for. Patients did not respond as well to immunotherapy as prostate cancer patients who have tumors with mismatch repair deficiency (MMRd) or microsatellite instability (MSI).

Given the small sample size of CPI-treated patients, there may still be a role for using CPIs in patients with CDK12 mutations; however, the routine use of CPIs in this patient population is not ready for prime time. Instead, I would suggest trying to get these patients into clinical trials to evaluate if CDK12 mutations are truly a biomarker predictive of response to immunotherapy.

What are the highlights of the study?

Schweizer: When you find a CDK12 mutation, this delineates a more aggressive prostate cancer. A series of studies have now shown that CDK12 mutations are associated with a high rate of de novo metastatic disease and higher Gleason scores. Patients appear to respond to standard therapies, but these responses are relatively short lived. This is clearly a high-risk population.

Were there any surprises in the data?

Schweizer: We anticipated that the association with CDK12 mutations would lead to a more aggressive phenotype, similar to MMRd. We were surprised that patients receiving CPIs were not as responsive as predicted from preclinical and translational studies. We had high hopes that CDK12 mutations would be as predictive as MMRd for immunotherapy responses.

What is the significance of a higher progression-free survival in chemotherapy-naive versus chemotherapy-pretreated patients who received CPIs?

Schweizer: Other studies show that heavily pretreated cancers have a more immune-invasive microenvironment, and this may explain why responses were low in our heavily pretreated patient population. In developing future clinical trials, it probably makes more sense to provide CPIs earlier in the treatment paradigm. Most studies have shown that immunotherapy is more effective when used early on.

For which prostate cancer patients would you recommend adding CPIs to therapy?

Schweizer: Ideally, if you have a prostate cancer patient with a CDK12 mutation, try to locate a clinical trial. For now, standard-of-care drugs still make the most sense for these patients. If a patient has no other viable options and is a good candidate for ongoing systemic therapy, then it would be reasonable to consider using a CPI. But these are better used in the context of a clinical trial.

How would you characterize the safety profile of the two regimens?

Schweizer: There was no evidence that CPIs or standard drugs were more toxic in CDK12-mutated prostate cancer patients.

What is the next step in this research?

Schweizer: Some clinical trials have opened to investigate the use of CPIs in CDK12-mutated prostate cancer patients, including at the University of Michigan (the IMPACT study) and the Veterans Administration CHOMP study. This is a relatively rare subset of patients -- 7% of castration-resistant prostate cancer cases. It will likely be a few years before these studies finish accruing and we have concrete results to point to.

What message would you give to practicing oncologists?

Schweizer: CDK12 mutations are not as strong a biomarker as MSI for predicting responses to CPIs. The data show a modest response to CPIs in a subset of patients. Most of these patients should not get CPIs off-label, but should be referred for a clinical trial.

Read the study abstract here and expert commentary about the clinical implications here.