APCCC 2022: Update on Immune Checkpoint Inhibitors

(UroToday.com) In the session of the 2022 Advanced Prostate Cancer Consensus Conference focusing on the treatment of patients with metastatic castration-resistant prostate cancer (mCRPC), Dr. Antonarakis provided an update on the role of immune checkpoint inhibitors in this disease space.

He began by discussing immune checkpoint monotherapy, first with anti-CTLA4 therapies. Anti-CTLA4 therapies, he noted, act primarily on T regulatory cells, rather than CD8 T cells. In the absence of pharmacotherapy, CTLA4 simulation results in these regulatory CD4 T cells (FoxP3+) secreting suppressive factors to act on the CD8 T cell. 

There have been two phase III trials of ipilimumab, an anti-CTLA4 immune checkpoint inhibitor, in mCRPC. The first of these examined radiotherapy with or without ipilimumab among patients who had previously received chemotherapy for mCRPC. Dr. Antonarakis emphasized that this study “almost met” its primary endpoint with a hazard ratio for survival of 0.85 (95% CI 0.72-1.00, p=0.053), describing it as the “most positive negative study”. However, overall, these results were relatively sobering with a PSA50 response rate of only 13%. Thus, perhaps not surprisingly, these data did not support approval of this agent in advanced prostate cancer. In the pre-chemotherapy space, PSA50 response rates were somewhat higher (23%) and progression free survival was improved, though no overall survival benefit was seen. Again, these data did not lead to regulatory approval and ipilimumab monotherapy is not used in mCRPC.

An alternative approach to immune checkpoint inhibition, anti-PD-(L)1 therapy, has also been explored as monotherapy in mCRPC.

Here again, poor responses were seen. The KeyNote-199 study demonstrated objective response rates of 5% and PSA50 response rates of 7%. However, the authors of this study did further genomic profiling of the extreme responders, finding that many have microsatellite instability and would have met criteria for immunotherapy for this reason. When they undertook subset analysis, while there was a somewhat higher response rate (16.7%), this was not clinically meaningful. A similar trend, without clinically meaningful implications, was found when examining differences in response stratified by tumor mutational burden and PD-L1 status.

Therefore, in keeping with other disease sites and general principles of oncology, a combination approach was tested, first using a combination of CTLA-4 and PD-1 blockade. In CheckMate 650, this approach showed, in Dr. Antonarakis’s words” “a bit better” but “still modest” results with an objective response rate of 18% and PSA50 response rate of 14%. However, in this trial, biomarkers including tumor mutational burden and homologous recombination repair mutation status showed some predictive ability. 

Dr. Antonarakis emphasized that this is somewhat consistent with the tumor site-agnostic approval of pembrolizumab in microsatellite instability (MSI)-high tumors. He noted, however, that MSI-high status is seen relatively infrequently in advanced prostate cancer (2-4%). Further, in this disease space, even when MSI-high is found, responses to pembrolizumab are much more muted (and shorter lived) than those observed in other tumor sites such as colorectal cancer or endometrial cancer. He suggested, however, that while tumor mutational burden offers some prognostic value, there is a qualitative difference of the basis of the mutation: frameshift mutations appear to be more prognostic than missense mutations, perhaps on the basis of higher neoantigenicity.

He further noted that bi-allelic loss of DNA damage repair genes (particularly CDK12) leads to unique genomic instability with a focal genomic instability signature. This can result in fusion genes throughout genome with a high volume of antigenic sequences. While this is relatively uncommon (7% of patients with mCRPC), it may be associated with durable responses to immune checkpoint inhibition. 

He further noted that a recent retrospective publication showed a potential ability of tumor mutational burden (dichotomized at 10) to predict response to immunotherapy as compared to taxane based chemotherapy in advanced mCRPC. While TMB > 10% is seen in approximately 6% of patients, notably, the vast majority of these will have mismatch repair deficiency and be eligible for pembrolizumab on this basis. Thus, it is unclear whether there is additional value from the assessment of TMB. Considering all genomic markers of response to immune checkpoint inhibitor therapy in mCRPC, he emphasized that each are individually rare, though dMMR, tumor mutational burden, HRR mutations, and CDK12 mutations are the most established.

Moving forward, he considered other combination approaches, first examining immune checkpoint inhibitors (in this case atezolizumab) and enzalutamide. Dr. Antonarakis described the data from IMbassador250 as “dead negative”. However, post hoc analyses suggested some benefit among those patients with high PDL-1, CD4, TMB, AR wild type, or p53 mutations.

He further emphasized that there may be a rationale for the combination of PARP inhibition and PD-1 inhibition in mCRPC. 

This was tested in a trial of durvalumab and olaparib, which interestingly showed a differential responses on the basis of DDR mutation status, with much better outcomes seen among patients with mutations. The KEYNOTE-365 trial of the combination of pembrolizumab and olaparib, however, included biomarker unselected patients. In this population, partial responses were seen in 2 of 28 patients (7%). This data formed the basis of the phase III KEYLYNK-010 trial comparing this treatment approach to a novel androgen receptor pathway inhibitor switch among patients with progressive mCRPC following docetaxel and one of abiraterone or enzalutamide.

Unfortunately, just recently, a press release indicated that this trial has been stopped due to futility. However, there is much work assessing alternative combination approaches including immune checkpoint inhibitors with tyrosine kinase inhibitors. In the COSMIC-021 trial, cabozantinib was combined with atezolizumab based on a postulated synergistic benefit of the TAM kinase, VEGFR2, MET, and AXL inhibition with anti-PD-1 effects. 

Data from this trial showed disease regression in 77% of 128 evaluable patients and formed the basis for a subsequent phase III trial (CONTACT-02) which is comparing this approach to a switch in novel hormonal therapy (NHT) among patients with mCRPC who have progression following one NHT. 

In conclusion, he emphasized that there are no approved immune checkpoint inhibitors for molecularly unselected patients with prostate cancer, though pembrolizumab is approved with a tumor site-agnostic indication for patients with MMR deficiency. However, there are multiple ongoing phase III trials of immune checkpoint inhibitor based combination therapies in this disease space.