PARP with PD-1/PD-L1 inhibition: Is There Any Magic to the Combination in a Molecularly Unselected Patient Population?

Previously, I’ve written Clinical Trials Portal articles about the concept of PARP inhibition in men with prostate cancer, with a strong focus on the biologic selection of patients with homologous recombination deficiency as most likely candidates for response to such agents.1,2  About 1.5 years ago, I wrote an article highlighting ongoing PD-1/PD-L1 antibody combination trials in prostate cancer.3  At that time, the goal was to consider intelligent combinations with immune-oncology agents in what has essentially been considered an immune “cold” prostate tumor microenvironment.  Hence, multiple trial partners in combination with PD-1/PD-L1 antibodies were proposed and referenced at that time.

We now have a couple of trials that have either published or presented early results of PARP inhibition with olaparib in combination with a PD-1 or PD-L1 antibody.  Neither trial mandated molecular enrichment with homologous recombination deficiency or mismatch repair alterations/microsatellite insufficiency as inclusion criteria.  With the understanding that the former would predispose to response to PARP inhibition and the latter predispose to response to an immune-oncology, the hope was that there really would be some magic behind this novel therapeutic combination.

The first report of this combination led to a publication of data on 17 metastatic castration-resistant prostate cancer patients who had received prior enzalutamide and/or abiraterone and were subsequently treated with durvalumab 1500 mg intravenously every 28 days with olaparib 300 mg orally twice daily.4  Of the 17 patients, 9 had a radiographic and/or PSA response.  Genomic analysis of the responding patients found 4 with germline and 2 with bi-allelic somatic alterations in known DNA repair genes.  Two other patients had mono-allelic alterations, one in PMS2, a mismatch repair gene, and the other in BRCA2.  The remaining two responding patients had no predisposing gene alterations.  One interesting point is that a very high percentage of patients in this trial (at least 6/17 or 35%) harbored bi-allelic alteration in known DNA repair genes.

The recent KEYNOTE-365 presentation at the 2019 Genitourinary Multidisciplinary Symposium reported on 42 previously docetaxel-treated metastatic castration-resistant prostate cancer patients who underwent combination pembrolizumab 200 mg intravenously every 21 days with olaparib 400 mg orally twice daily.5  In this trial, Guardant360® was utilized to evaluate circulating tumor DNA for BRCA1, BRCA2 or ATM alterations.  None were definitively found, but we have to recognize potential challenges with sensitivity issues of the assay for these genetic alterations in prostate cancer.  Whole exome sequencing was performed in 17 patients with soft tissue disease available for assessment, yielding only 12 qualifiable results and 1 patient with a BRIP1 alteration.  However, bi-allelic alteration status was not able to be determined in the patient with a BRIP1 alteration.  Hence, no patients were definitively identified to have homologous recombination deficiency.  The confirmed PSA decline rate of at least 50% was found to be 12% and confirmed RECIST v1.1 response rate was 7%.  Yet, 29% of the patients with RECIST evaluable disease had a reduction in tumor burden of at least 30%.

These above studies, along with others, are emphasizing the need for standardization of next-generation sequencing assays for our patients with prostate cancer.  Our confidence in the results and interpretations from next-generation sequencing technologies still yields significantly more to be desired.  Additionally, sequencing of circulating tumor DNA from liquid tumor biopsies is becoming more commonplace, yet these technologies may harbor even greater technology and sensitivity issues.

Regardless, the question of how well the combination of a PARP inhibitor with a PD-1/PD-L1 antibody remains, especially in a population that truly does not harbor a DNA repair alteration.  For those that do harbor a known alteration in a homologous recombination repair gene, such as BRCA1 or 2, there is data to show in breast and ovarian cancer that there is higher tumor mutational burden and a greater number of tumor-infiltrating lymphocytes.6,7  Additional preclinical data supports CTLA4 inhibition as synergistic with PARP inhibition in BRCA1 deficient tumors.8  Yet there is growing literature that supports the use of this novel combination of therapeutic agents even in homologous recombination proficient population.  For example, PARP inhibition induces PD-L1 expression in homologous proficient breast cancer.A recent manuscript also shows that PARP inhibition can promote the accumulation of cytosolic DNA fragments due to unresolved DNA lesions, which in turn activates the DNA sensing cGAS-STING pathway.  This stimulates the production of type I interferons, and in turn, induces antitumor immunity that may be further enhanced by checkpoint inhibitor blockade, independent of BRCAness.10

Ultimately, the future of the use of PARP inhibitors in combination with PD-1/PD-L1 inhibitors in prostate cancer will depend upon additional studies in randomized patient populations.  The above is hypothesis-generating in the fact that it shows some activity in molecularly unselected patients, yet we cannot yet say that we are confident in the homologous recombination status of the treated patients.  Below, I list some ongoing trials that are conceptually similar in combining a PARP inhibitor with a PD-1/PD-L1 antibody for patients with metastatic castration-resistant prostate cancer.  Future efforts will need to convincingly prove that this therapeutic combination is truly efficacious for a population of prostate cancer patients that have not been enriched with molecular selection inclusion criteria.

Relevant trials currently open for accrual
1. Durvalumab with Olaparib  (NCT02484404)
2. KEYNOTE 365 Cohort A – Pembrolizumab with Olaparib (NCT02861573)
3. Nivolumab with Rucaparib (NCT03572478)
4. JAVELIN PARP Medley – Avelumab with Talazoparib (NCT03330405)
5. KEYLYNK 010 – Phase 3 of Pembrolizumab with Olaparib vs. Abiraterone acetate or Enzalutamide (NCT03834519)

Written by: Evan Yu, MD

1. Yu EY.  Urotoday Clinical Trials Portal.  On-line: February 12, 2017.
2. Yu EY.  Urotoday Clinical Trials Portal.  On-line: January 7, 2019.
3. Yu EY.  Urotoday Clinical Trials Portal.  On-line: August 14, 2017.
4. Karzai F et al.  J Immunother Cancer 2018; 6:141.
5. Yu EY et al.  J Clin Oncol 37, 2019 (suppl 7S; abstr 145).
6. Nolan E et al.  Sci Transl Med 2017; 9:pii: eaal4922. doi:10.1126/scitranslmed.aal4922
7. Strickland KC et al.  Oncotarget 2016; 7:13587-98.
8. Higuchi T et al.  Cancer Immunol Res 2015; 3:1257-68.
9. Jiao S et al.  Clin Cancer Res 2017; 23:3711-20.
10. Shen J et al.  Cancer Res 2019; 79:311-9.
email news signup