Sensitivity, Resistance of Molecular Subtypes in Prostate Cancer Identified Through Drug Response Scores

Robert B. Den, MD

Prostate cancer has recently been recognized as a genomically heterogeneous disease with subtypes similar to breast or ovarian cancers. According to results of a large prospective study¹ presented by Robert B. Den, MD, at the 2017 Society of Urologic Oncology Annual Meeting, these subtypes in the localized disease setting have distinct drug response profiles, which suggests that classifying patients into these subtypes may be useful for treatment decisions, selecting patients for clinical trials, and provide translatable clinical tools for personalizing postoperative androgen-deprivation therapy (ADT) for patients with prostate cancer.

“Over the past several years, multiple drugs have demonstrated improvements in overall survival in men with advanced prostate cancer, prompting further investigations of these drugs. However, characterizing drug response in the localized disease setting and in the context of prostate cancer molecular subtypes needs further investigation,” Den and colleagues wrote in their abstract.

The management of prostate cancer has evolved, with results from the LATITUDE and STAMPEDE trials providing evidence that supports the earlier role of systemic therapy, such as abiraterone acetate (Zytiga) and docetaxel (Taxotere). Results from LATITUDE suggest that adding abiraterone plus prednisone to ADT lowered the risk of death by 38% in men with newly diagnosed, high-risk, metastatic prostate cancer.² In the STAMPEDE trial, which consisted of a broader population of high-risk hormone-naïve patients including some participants with nonmetastatic disease, results showed that the addition of abiraterone to standard initial therapy lowered the relative risk of death by 37% and improved the progression-free survival (PFS) by 71%.³

In the current study, the investigators generated patient-specific drug response scores (DRS) across 15,136 prospective patients who underwent radical prostatectomy. DRS were also generated for 89 drugs administered to 954 patients across 36 cell lines with known drug treatment outcomes. Patient sensitivity for the 89 agents was determined using the NCI-60 panel, which uses 60 different human tumor cell lines to identify and characterize novel compounds with growth inhibition or killing of tumor cell lines. Pearson’s chi squared test was used to determine significant differences in drug sensitivity among prostate cancer subtypes.

Patient samples were classified into basal, luminal A, luminal B, and neuroendocrine subtypes using the PAM50 profiling test and small cell gene expression signatures. The testing classified 43% of samples in the radical prostatectomy cohort as basal, 26% as luminal A, 30% as luminal B, and 2% as neuroendocrine. Additionally, the DRS were able to discriminate between sensitive and resistant cell line in bladder cancer and prostate cancer.

The DRS were highly variable across the different subtypes. Basal tumors were more sensitive to kinase inhibitors, mTOR inhibitors, DNA repair inhibitors, antineoplastic agents, and alkylating chemotherapies. Luminal A and B subtype tumors were more sensitive to steroid inhibition such as abiraterone, antiproliferative chemotherapies, and anti-microtubule chemotherapies. Neuroendocrine tumors were most sensitive to antiproliferative agents, such as mitomycin and topotecan. Significant differences in average DRS scores in subtypes were observed for all 89 drugs included in the study (P <.001).

Basal tumors were also predicted to be more sensitive to platinum-based chemotherapies and topoisomerase inhibitors, whereas the luminal A and B subtypes were predicted to be more sensitive to taxane-based therapy.

The investigators declared the DRS a success, as they were validated in several publicly available treatment response datasets. The DRS also corresponded to expected biological processes, such as cell proliferation and DNA repair.

“While further validation is necessary, we believe that this technology has the potential to improve patient treatment decisions with regard to chemotherapy,” Den and colleagues said in their poster.

References

1. Den R, Takhar M, Lehrer J, et al. Drug response variability between luminal and basal prostate cancer tumors. Poster presented at: 2017 SUO Annual Meeting; November 29-December 1, 2017; Washington, DC. Poster #221.
2. Fizaxi K, Tran N, Fein LE, et al; the LATITUDE investigators. LATITUDE: A phase 3 double-blind, randomized trial of androgen deprivation therapy (ADT) with abiraterone acetate (AA) plus prednisone (P) or placebos (PBOs) in newly diagnosed high-risk metastatic hormone-naïve prostate cancer (mHNPC) patients (pts). J Clin Oncol. 2017;35 (suppl; abstr LBA3).
3. James ND, DeBono JS, Spears MR, et al. Adding abiraterone for men with high-risk prostate cancer (PCa) starting long-term androgen deprivation therapy (ADT): Survival results from STAMPEDE (NCT00268476). J Clin Oncol. 2017;35 (suppl; abstr LBA5003).