Activity of Platinum-Based Chemotherapy in Patients With Advanced Prostate Cancer With and Without DNA Repair Gene Aberrations

Key Points Question Is there a role for platinum-based treatment in molecularly selected patients with advanced prostate cancer? Findings In a case series of 508 patients, platinum-based therapy was associated with antitumor activity, especially among patients with known DNA repair gene aberrations. In patients with DNA repair gene aberrations, nearly half had a decrease in prostate-specific antigen levels of at least 50% and experienced soft tissue responses. Meaning In patients with prostate cancer and DNA repair gene aberrations, platinum-based therapy may be considered a treatment option.


Introduction
Despite significant progress in drug development for advanced prostate cancer patients in the past decade, new active compounds, ideally for molecularly selected patients, are urgently needed. 1,2 Platinum-based compounds have been evaluated in clinical trials and used clinically as monotherapy or in combination with other chemotherapy agents mainly in the setting of castration-resistant disease but also in hormone-sensitive disease. 3,4 A phase 3 clinical trial of satraplatin compared with prednisone 5 demonstrated an improvement in progression-free survival and pain control, but no overall survival benefit was seen in unselected patients.
Platinum chemotherapy activity has been associated with its ability to crosslink with purine bases in DNA, interfering with DNA repair mechanisms and causing DNA damage and apoptosis. In various cancer types, responses are enhanced in the presence of underlying double-strand DNA repair alterations in the tumor, resulting in synthetic lethality. 6 In triple-negative breast cancer (TNBC), carboplatin is highly effective in patients with known BRCA1 (OMIM 113705) and BRCA2 (OMIM 600185) tumors, 7 and 2 phase 3 trials have shown benefit of poly(ADP-ribose) polymerase (PARP) inhibitors in patients with TNBC and germline BRCA variations. 8 The role of these agents in an unselected TBNC population is controversial; however, there is evidence that some patients with variations in homologous recombination genes other than BRCA1 and BRCA1 (germline or somatic) can derive benefit from platinum-based treatment. 9 On the other hand, alterations in other non-homologous recombination DNA damage response genes, such as PTEN (OMIM 601728), are not associated with response to the same extent. 9 Genomic aberrations that impair DNA repair genes occur at a frequency of up to 20% to 30% in advanced prostate cancer. [10][11][12] Some of these aberrations, which can either be found as germline or somatic alterations in homologous recombination DNA repair genes or DNA damage checkpoints, have been associated with sensitivity to platinum compounds and/or PARP inhibition in preclinical studies and in clinical trials. [13][14][15][16][17] Three published case series, [18][19][20] which included a total of 14 patients with metastatic prostate cancer with DNA repair gene aberrations treated with platinum-based chemotherapy, reported encouraging antitumor activity. The recently presented prospective phase 3 PROFOUND study evaluating the PARP-inhibitor olaparib in molecularly selected patients with advanced prostate cancer harboring DNA repair gene aberrations has shown a significant benefit in radiographic progression-free survival and overall response rate for olaparib compared with the sequential use of abiraterone or enzalutamide, thereby strengthening the previously reported findings of the TOPARP study. [21][22][23] Data from these trials suggest higher antitumor activity in patients with BRCA2 alterations. 22 Through the collaborative efforts of an international consortium, we identified a large series of patients with castration-resistant prostate cancer (CRPC) who were treated with platinum-based chemotherapy. Using these data, we performed a retrospective analysis to characterize the antitumor activity (ie, decrease in prostate-specific antigen [PSA] level, radiographic response, and time receiving treatment) of platinum-based therapies (monotherapy and/or combination) in men with CRPC with or without DNA repair gene alterations.

Patients
Patients with biochemically or histologically confirmed advanced prostate cancer (metastatic or locally advanced and not amenable to locoregional treatment with curative intent) treated with a platinum compound (cisplatin or carboplatin) either as monotherapy or as part of combination chemotherapy were eligible for this analysis. Patients with primary pure small cell carcinoma of the prostate or insufficient data for analysis of the primary end point were excluded. Clinical data from 25 cancer centers from 12 countries worldwide were collected by local investigators. Investigators were encouraged to include all patients eligible at their site for the analysis. Approval of the local ethics committee was obtained before data collection, and informed consent was obtained depending on local regulations. Clinical data were collected locally at each center and assembled in an electronic master database at the coordinating center. Quality control was assured by queries in cases of nonplausible data and inconsistencies; however, no review of the source documentation was performed. The the reporting guideline for case series was followed. 24 The primary outcome measure was evaluation of antitumor activity (decrease in PSA level and soft tissue response) to platinum-based therapy (monotherapy and/or combination therapies) and

Statistical Analysis
Continuous data were summarized by median, minimum, and maximum values. Categorical data were presented as absolute numbers and percentages. Time to event end points were assessed by the Kaplan-Meier method and are presented as median and interquartile range (IQR). The number of missing data points is given for all analyses. Frequency counts of categorical data in subgroups were statistically compared by Fisher exact tests. Between subgroup comparisons of numerical data were carried out by Kruskal-Wallis tests. Time to event end points were compared between subgroups using the log-rank test. The a priori significance level was P < .05, and all statistical tests were 2 sided.
All analyses were done in R version 3.5.0 (R Project for Statistical Computing).
To simultaneously assess several factors that could be associated with OS, we performed a

Results
A total of 508 men who were diagnosed with prostate cancer between 1986 and 2018 were included in the study (median [

Outcomes in Cohort 3
Patients in cohort 3 (ie, those without genomic profiling) had an overall PSA level decrease of at least

Discussion
In this multicenter retrospective analysis of 508 patients with CRPC, most of whom had received at least 1 prior line of therapy, we found encouraging antitumor activity for treatment with platinumbased therapies in the cohort of patients with tumors harboring DNA repair gene aberrations.  Historically, platinum compounds as monotherapy or combination therapy have been widely studied in prostate cancer. 26 Besides the SPARC trial, 5 the studies were mostly small, recruited a molecularly unselected patient population, and demonstrated only moderate antitumor activity.
However, a subgroup of patients seemed to derive benefit, 3   Response to platinum-based combination treatment in our cohort was more favorable than platinum-based monotherapy, and in most cases, a taxane was chosen as the combination partner, which represents the current standard-of-care chemotherapy in the unselected advanced prostate cancer population. However, it seems that in our data set monotherapy was more often used in patients with known DNA repair gene aberrations, whereas in patients without alterations or unknown molecular status, combination therapy was preferred. Taxanes  presented the results of a prospective cohort study investigating the association of germline DNA repair gene aberrations with CRPC outcomes. In their cohort, response to taxanes was not different between variation carriers and noncarriers; however, duration of response was shorter in carriers, and treatment sequencing seemed to be of importance. 30,31 Therefore, activity of the taxane component has to be taken into some account when interpreting our efficacy data in patients receiving combination treatment.
Our results showed consistently higher response rates of platinum-based treatment in molecularly selected patients, even though these patients often received platinum-based monotherapy, which seems to be less active than combination therapy overall. Subgroups for specific types of DNA repair gene aberrations were too small to draw any definitive conclusions, but the response rate (PSA decrease and objective response rate) in patients with BRCA2 alterations was encouraging. These results are in line with the PROFOUND results, in which treatment benefit was more pronounced in patients with BRCA1, BRCA2, and ATM alterations compared with alterations in other gene alterations. 22 In the exploratory analysis of gene-by-gene radiographic progression-free survival, benefit in patients with BRCA1 and ATM alterations was less pronounced compared with patients with BRCA2 alterations. In the TRITON phase 2 trial, evaluating rucaparib in patients with DNA repair gene aberrations, radiographic and PSA responses were observed in only 2 of 19 patients (10.5%) with exclusively ATM alterations. 32 The antitumor activity of platinum compounds in our small cohort of 12 patients with ATM alterations is noteworthy. Overall, different DNA repair gene aberrations are most likely distinct entities with varying responses to platinum as well as PARP inhibition; therefore, they cannot be collectively addressed. More research with prospective trials in molecular subgroups is needed to better characterize which patients might derive benefit.
Importantly, the activity of platinum-based treatment was also seen in the unselected cohort 3 (ie, patients were not tested for DNA repair gene aberrations). This population is possibly most reflective of the general CRPC population seen in many centers around the world, where testing for DNA repair gene aberrations may not be available in daily clinical practice.

Limitations
Our analysis has several limitations, including the retrospective design and data collection, including missing data in a subset of patients. Furthermore, even though quality control was assured by the coordinating center by queries in the case of nonplausible or inconsistent data, no formal review process of the source documentation of each individual contributing center was performed. Also, there was no control of patients excluded from the database on the center level.

JAMA Network Open | Oncology
Platinum-Based Chemotherapy for Advanced Prostate Cancer With DNA Repair Gene Aberrations aggressive features were included; of note, we excluded patients with de novo pure small cell carcinoma from the analysis. Molecular profiling was performed in only 178 patients, of whom 80 had a defect (cohort 1) and 98 patients did not (cohort 2). However, the remaining 330 patients (cohort 3) most likely represent a mixed population, including perhaps some patients with DNA repair gene aberrations, and therefore, the results of this cohort have to be interpreted with caution. Currently, several prospective phase 2 trials (randomized and nonrandomized) evaluating carboplatin alone or in combination with docetaxel in molecularly selected patients with CRPC are recruiting. 34-37

Conclusions
In this multicenter international case series, platinum-based treatment was associated with promising activity in a biomarker-preselected population of patients with CRPC and DNA repair gene aberrations. These results need prospective validation, which hopefully will be provided from the currently recruiting trials in molecularly selected prostate cancer populations. Based on our analysis, platinum-containing therapy should be considered in patients with DNA repair gene aberrations, especially if access to a PARP inhibitor is not available.