Activity of Platinum Chemotherapy in Men with Prostate Cancer With and Without DNA Damage Repair Mutations

Abstract

Introduction:

Alterations in homologous recombination repair (HRR) genes occur in 20–30% of men with metastatic castration-resistant prostate cancer (mCRPC) which may increase sensitivity to platinum chemotherapy. Specifically, exceptional responses to platinum chemotherapy have been reported among patients with BRCA mutations. This study aimed to evaluate the efficacy of platinum chemotherapy in patients with mCRPC with and without HRR.

Patient and Methods:

In this retrospective, multi-institution series, we analyzed patients with mCRPC to assess response to platinum-containing chemotherapy based on HRR alteration status. Outcome measures were prostate specific antigen (PSA)50 response rate (percentage of patients achieving at least a 50% decline in PSA from baseline), overall survival (OS) and progression-free survival (PFS).

Results:

From 1999 – 2020, 24 patients with mCRPC who received platinum chemotherapy were included with 7 patients analyzable for PSA outcomes. HRR alterations were found in 19 out of 24 patients (79.2%) with mutations recognized in 11 different HRR genes. Patients with a HRR alteration achieved a PSA50 response rate of 20% (1 out of 5) after platinum chemotherapy compared to 50% (1 out of 2) in patients without a HRR mutation. No difference in OS or PSA PFS was detected among patients with BRCA1/2 mutations compared to HRR alterations other than BRCA1/2 and patients without HRR alterations.

Conclusion:

In patients with mCRPC, we did not find a statistical difference in anti-tumor activity after receiving platinum chemotherapy among patients harboring a pathogenic HRR alterations compared to patients without a HRR alteration. Additionally, we were unable to detect an association between BRCA1/2 mutation status and response to platinum chemotherapy. Platinum chemotherapy, however, had clinically meaningful activity in a subset of patients regardless of HRR alteration status. Additional studies are warranted using genomic data to predict sensitivity to platinum chemotherapy.

Introduction

Twenty to thirty percent of men with metastatic prostate cancer harbor either germline or somatic alterations in genes responsible for homologous recombination repair (HRR).^1,2 HRR is a major mechanism utilized by cells to accurately repair DNA double-strand breaks. Due to this disruption in DNA damage repair, tumors with HRR alterations may be particularly susceptible to DNA damaging agents or drugs that interfere with other DNA repair mechanisms. Capitalizing on this vulnerability, the poly (ADP-ribose) polymerase inhibitors (PARPi) olaparib and rucaparib were approved for use in men with metastatic castration-resistant prostate cancer (mCRPC) who have alterations in one of 14 HRR genes (olaparib) or in BRCA1/2 (rucaparib).³ Additionally, the combination of PARPi and androgen receptor signaling inhibitors (ARSI) including olaparib with abiraterone acetate and niraparib with abiraterone acetate received recent FDA approval for BRCA-mutated mCRPC, and the combination talazoparib with enzalutamide was FDA approved for HRR gene-mutated mCRPC.^4–6

Platinum containing chemotherapies are DNA-damaging cytotoxic agents. Increased platinum activity in BRCA-altered tumors is recognized in other cancer types, including breast, ovarian, and pancreatic cancers.⁷ Platinum agents are also routinely used in unselected populations of patients with a variety of cancer types, including lung, colorectal, gastric, pancreatic, and ovarian. Platinum containing chemotherapy is not included in clinical consensus guidelines for prostate cancer except in the case of small-cell or aggressive variant prostate cancer.⁸ In the only randomized phase III trial of platinum-containing chemotherapy in prostate cancer, the oral agent satraplatin failed to demonstrate an overall survival benefit compared to placebo, although it increased progression free survival and time to pain progression.⁹ In a systematic review of 53 prospective clinical trials evaluating platinum chemotherapy in CRPC, a meta-analysis of five randomized trials found an overall response benefit with the use of platinum-based chemotherapy compared to a non-platinum based chemotherapy but no overall survival benefit.¹⁰ Subgroup analysis for predictive markers such as neuroendocrine differentiation or HRR alteration status could not be performed due to paucity of data.¹⁰

Case series of exceptional platinum chemotherapy responders whose prostate cancer harbored BRCA2 alterations have been reported.^11,12 In a larger multi-institution series of men with mCRPC who received platinum chemotherapy, there was no difference in prostate-specific antigen (PSA) response rate or overall survival in men with or without HRR alterations. This series found PSA50 response rate (percentage of patients achieving at least a 50% decline in PSA from baseline prior to therapy) in the cohort with HRR mutations of 47.1% compared to a PSA50 of 36.1% in the cohort without HRR mutations. Considering the subset of men with BRCA2 alterations, however, PSA response rate (63.9%) and soft tissue response rates were higher.¹³ Taken together, data regarding the optimal use of platinum chemotherapy in prostate cancer remains unclear.

In this multi-institution series, we examined data from 24 patients with metastatic prostate cancer treated with platinum-based therapy. We hypothesized that patients with prostate cancer and pathogenic BRCA1/2 mutations receiving platinum-containing chemotherapy would have a higher PSA50 response rate than those without BRCA1/2 mutations. Further, we hypothesized that men with prostate cancer and any pathogenic HRR alteration receiving platinum-containing chemotherapy would have a higher PSA50 response rate than men without an HRR alteration.

Methods:

Study Design

We utilized the Oncology Research Information Exchange Network (ORIEN) dataset to identify patients for inclusion in this retrospective series. ORIEN is an alliance of 18 U.S. cancer centers established in 2014. All ORIEN alliance members utilize a standard Total Cancer Care® (TCC) protocol. As part of the TCC study, participants agree to have their clinical data followed over time, to undergo germline and tumor sequencing, and to be contacted in the future by their provider if an appropriate clinical trial or other study becomes available. TCC is a prospective cohort study with a subset of patients enrolled to the ORIEN Avatar program, which includes research use only (RUO) grade whole-exome tumor sequencing, RNA sequencing, germline sequencing, and collection of deep longitudinal clinical data with lifetime follow up. All abstracted clinical data elements and molecular sequencing files are harmonized into a standardized, structured format to enable aggregation of de-identified data for sharing across the Network.

1,168 ORIEN Avatar patients diagnosed with prostate cancer between 1999–2020 and consented to the TCC protocol from 11 participating U.S. cancer centers of ORIEN were identified. From this cohort, we included for this analysis patients that had received platinum chemotherapy. All studied patients underwent genomic sequencing, either somatic, germline or both, to determine pathogenic HRR alteration status. We excluded patients who had received platinum-containing chemotherapy for an indication other than metastatic prostate cancer and patients who had received platinum-containing chemotherapy prior to the diagnosis of mCRPC. The primary outcome measure was PSA50 response rate to platinum-containing chemotherapy for those with and without HRR defects. PSA50 response rate was defined as the percentage of patients achieving at 50% decline in PSA or greater from baseline or prior to initiation of therapy. Secondary endpoints were overall survival (OS) and progression-free survival (PFS)

Statistical Methodology:

Cohort characteristics were summarized by alteration (germline and somatic) status using absolute numbers and percentages for categorical variables and using the mean, standard deviation, median and quartiles for continuous variables. The Kaplan-Meier method was used to calculate OS and PFS. For overall survival/progression-free status, censoring occurred when a subject was missing a deceased date and was set as the last date of follow-up. Time to event was calculated from initiation of platinum-containing chemotherapy. P-values are reported based on a null hypothesis of no effect against a two-sided alternative and significance was set at the 0.05 level. Analyses were performed using SAS 9.4 (SAS Institute; Cary, NC) and R version 4.2.0 (R Core Team).

Results:

Patient Characteristics:

1,168 patients with prostate cancer of any stage were identified in the ORIEN database. Of those, 47 patients had received platinum-containing chemotherapy. Twenty-three patients were excluded due to receiving platinum chemotherapy prior to their diagnosis of mCRPC or due to receiving platinum-containing chemotherapy for an indication other than prostate cancer (i.e. an alternative cancer type). Of the remaining cohort of 24 patients, all were analyzable for survival outcomes. Seven were analyzable for PSA outcomes (Figure 1).

Figure 1: CONSORT flow diagram.

The initial ORIEN dataset contained 1,168 patients. Of those, 47 received platinum-containing chemotherapy (Carboplatin, Cisplatin or Oxaliplatin). Additional subjects’ data were excluded due to only receiving therapy prior to their diagnosis or due to receiving platinum containing therapy for something other than metastatic prostate cancer. Remaining cohort included 24 patients with survival data and 7 PSA-analyzable patients.

Cohort characteristics of the 24 included patients are summarized in Table 1. Nineteen (79.2%) had at least one pathogenic HRR alternation compared to 5 (20.8%) patients without a pathogenic HRR alteration. The median age at diagnosis was 61.25 years and the median baseline PSA was 12.5 ng/mL. There were some differences between the cohort with HRR alterations and those without alterations, however, none of these differences were statistically significant. Patients with HRR alterations had numerically higher baseline median PSA levels at diagnosis (20.1 ng/mL vs 6.16 ng/mL, p = 0.275). Compared to patients without HRR alterations, patients with pathogenic HRR alterations had higher rates of prior therapy including prior 2^nd generation hormonal therapy (68.4% vs 60.0%, p = 1), prior PARPi treatment (21.1% vs 0%, p=0.544), and prior taxane chemotherapy (68.4% vs 20.0%, p =0.122). Out of the 14 patients that received prior taxane chemotherapy, 5 patients initiated platinum and taxane chemotherapy concurrently.

Table 1.

Cohort Characteristics of mCRPC Patients that Received Platinum Containing Chemotherapy

		DDRm Status
Variables	Values	Yes (N = 19)	No (N = 5)	P-value	Overall
Clinical Measures
Race	Other	3 (15.79%)	1 (20%)	12	4 (16.67%)
	Caucasian	16 (84.21%)	4 (80%)		20 (83.33%)
Ethnicity	Hispanic; Spanish	3 (16.67%)	0 (0%)	12	3 (13.04%)
	Non-Hispanic; Non-Spanish	15 (83.33%)	5 (100%)		20 (86.96%)
Histology	Adenocarcinoma, NOS	19 (100%)	5 (100%)		24 (100%)
PathTStage	No TNM applicable for this site/histology combination	2 (13.33%)	0 (0%)	12	2 (10.53%)
	T2	1 (6.67%)	0 (0%)		1 (5.26%)
	T3b	4 (26.67%)	1 (25%)		5 (26.32%)
	p3a	2 (13.33%)	0 (0%)		2 (10.53%)
	p3b	4 (26.67%)	2 (50%)		6 (31.58%)
	px	2 (13.33%)	1 (25%)		3 (15.79%)
PathNStage	N0	3 (20%)	1 (25%)	0.29982	4 (21.05%)
	N1	5 (33.33%)	0 (0%)		5 (26.32%)
	No TNM applicable for this site/histology combination	2 (13.33%)	0 (0%)		2 (10.53%)
	p0	1 (6.67%)	2 (50%)		3 (15.79%)
	p1	2 (13.33%)	0 (0%)		2 (10.53%)
	pX	2 (13.33%)	1 (25%)		3 (15.79%)
PathMStage	M0	1 (10%)	1 (33.33%)	0.77972	2 (15.38%)
	MX	1 (10%)	0 (0%)		1 (7.69%)
	No TNM applicable for this site/histology combinat	3 (30%)	0 (0%)		3 (23.08%)
	c0	5 (50%)	2 (66.67%)		7 (53.85%)
PathGroupStage	II	1 (8.33%)	0 (0%)	0.08132	1 (6.67%)
	III	5 (41.67%)	1 (33.33%)		6 (40%)
	IIIB	0 (0%)	1 (33.33%)		1 (6.67%)
	IIIC	0 (0%)	1 (33.33%)		1 (6.67%)
	IV	6 (50%)	0 (0%)		6 (40%)
Prior PARP Inhibitors	No	15 (78.95%)	5 (100%)	0.5442	20 (83.33%)
	Yes	4 (21.05%)	0 (0%)		4 (16.67%)
Prior Taxane Chemotherapy	No	6 (31.58%)	4 (80%)	0.12222	10 (41.67%)
	Yes	13 (68.42%)	1 (20%)		14 (58.33%)
Prior Hormonal Therapy	No	6 (31.58%)	2 (40%)	12	8 (33.33%)
	Yes	13 (68.42%)	3 (60%)		16 (66.67%)
Initial Platinum-Containing Chemotherapy*	Carboplatin	13 (68.42%)	5 (100%)	0.00342	18 (75%)
	Cisplatin	5 (26.32%)	0 (0%)		5 (20.83%)
	Oxaliplatin	1 (5.26%)	0 (0%)		1 (4.17%)
Visceral Metastases (excl. Bone and Lymph Node)	No	9 (47.37%)	2 (40%)	12	11 (45.83%)
	Yes	10 (52.63%)	3 (60%)		13 (54.17%)
Age at Diagnosis		19 63.56 (8.08) 63.21 (55.96, 69.68)	5 55.39 (7.38) 57.78 (55.78, 58)	0.10353	24 61.86 (8.49) 61.25 (55.87, 66.53)
PSA (ng/mL)		16 31.04 (31.62) 20.1 (7.35, 51.5)	5 31.31 (60.87) 6.16 (3, 6.6)	0.27483	21 31.11 (38.61) 12.5 (5.7, 49.4)
Time to next Prostate Therapy (years)		7 0.56 (0.59) 0.46 (0.1, 0.57)	3 0.78 (0.45) 0.75 (0.36, 1.25)	0.51673	10 0.63 (0.54) 0.48 (0.36, 0.75)

¹Baseline PSA taken at Date of Diagnosis

²Fishers Exact Test

³Wilcoxon Rank-Sum Test

^*One Patient (DDRm = Yes) started on Carboplatin and changed to Cisplatin

The genomic alterations identified in our cohort of 24 patients are highlighted in Table 2. There were 7 germline and 43 somatic mutations identified in the HRR pathway. Mutations were recognized in 11 different HRR genes, with CHEK1 (N=16, 66.67%) and BRCA2 (N=6, 25%) being the most common somatic mutation identified. FANCL (N=3, 12.5%) and BRCA2 (N=2, 8.33%) were the most frequently identified germline mutations.

Table 2.

Breakdown of HRR Alterations of Interest

Gene	Germline (N = 24)	Somatic (N = 24)
ATM	1 (4.17%)	4 (16.67%)
BRCA1	0 (0%)	4 (16.67%)
BRCA2	2 (8.33%)	6 (25%)
CDK12	1 (4.17%)	4 (16.67%)
CHEK1	0 (0%)	16 (66.67%)
CHEK2	0 (0%)	2 (8.33%)
FANCL	3 (12.5%)	3 (12.5%)
PALB2	0 (0%)	1 (4.17%)
RAD51B	0 (0%)	1 (4.17%)
RAD51C	0 (0%)	1 (4.17%)
RAD54L	0 (0%)	1 (4.17%)

PSA50 Responses:

Seven patients had PSA data available to assess response. Maximum percent change in PSA after initiation of platinum chemotherapy is shown in Figure 2. Among the 5 patients who harbored at least one HRR alteration, 3 (60%) achieved a PSA reduction from baseline, and 1 (20%) achieved a 50% reduction in PSA. In the 2 patients without an HRR mutation, 1 (50%) had a reduction in PSA after platinum chemotherapy and that same patient achieved a PSA50 response.

Figure 2:

Waterfall plot for maximum percent change in PSA (ng/mL)

Out of the 4 patients with a BRCA1/2 mutation, either somatic or germline, and available PSA data, none achieved a PSA50 response. In the 3 patients without a BRCA1/2 mutation, 2 (66.7%) experienced PSA50 response. One patient had an ATM and CHEK1 mutation, whereas one patient had no HRR alteration.

Survival Outcomes

The median OS after initiating platinum chemotherapy in patients harboring a BRCA1/2 alteration was 11.77 months (95% CI: 9.54 to NA) compared to 21.53 months (95% CI: 0.72 to NA) in patients without a HRR alteration and 8.78 months (95% CI: 5.26 to NA) with an HRR alteration other than BRCA1/2 (Figure 3). There were no significant differences between groups. The median OS among the entire cohort of 24 patients was 11.8 months (95% CI: 9.54, NA).

Figure 3:

Overall Survival (from date of chemotherapy initiation) for Cohort by BRCA1/2 Mutation Status

Median PFS in those with BRCA1/2 alterations was 11.62 months (95% CI: 11.38 to NA) compared to 21.53 months (95% CI: 0.72 to NA) in patients without an HRR alteration and 8.78 months (95% CI: 5.26 to NA) in patients with an HRR alteration other than BRCA1/2 (Figure 4). Again, there were no significant differences between groups. The median PFS among the cohort was 11.6 months (95% CI: 8.78, NA).

Figure 4:

Progression-Free Survival (from date of chemotherapy initiation) for Cohort by BRCA1/2 Mutation Status

Individual disease courses are described in the Swimmer’s Plot (Figure 5). This figure depicts the time course between diagnosis and last follow up or death.

Figure 5:

Arm plot of individual disease course between diagnosis and last follow up or death. Filled in blocks represents time that the patient remained in treatment. A triangle marks a PSA50 response rate, a white circle represents death of the patient.

Discussion:

In this real-world retrospective study analyzing response to platinum-based chemotherapy and the association between pathogenic HRR gene alterations in 24 patients with metastatic prostate cancer, we did not find a statistically significant difference in antitumor activity in patients harboring HRR alterations compared to patient without HRR alterations. Specifically, we did not detect a statistically significant difference in PSA50 response rates, PFS, or OS. Overall, platinum chemotherapy appeared to have efficacy in a subset of patients, regardless of HRR status.

These treatment responses align with results from PROREPAIR-B study which looked at the effect of germline HRR mutations (specifically, ATM/BRCA1/BRCA2/PALB2) on cause-specific survival (CSS) and responses to currently approved therapies for mCRPC. They found no statistically significant differences in CSS, PSA50 and PFS when comparing germline HRR alteration carriers compared to non-carriers after therapy with androgen signaling inhibitors and taxane chemotherapy.¹⁴

Furthermore, our results are similar to a larger case series by Schmid et al. looking at the activity of platinum chemotherapy in 508 men with CRPC¹³. Patients in this study were grouped into cohorts based on whether they harbored DNA repair gene aberrations or not. They found no statistically significant difference in PSA50 response rate or OS between the two cohorts. Interestingly, when they analyzed the subgroup of patients with BRCA2 mutations, favorable response rates were seen suggesting a role for molecular selection of patients who may gain the most benefit from platinum-based chemotherapy.¹³ In our analysis, although none of the 4 patients with BRCA1/2 alterations with available PSA data had a PSA50 response, there was numerically longer PFS and OS in the BRCA1/2 cohort compared to non-BRCA1/2 HRR alterations.

While we did not detect BRCA1/2 mutation status to be associated with an improved response to platinum chemotherapy, a major limitation to our study was its small sample size. With 24 total patients and 7 PSA analyzable patients, our analysis was underpowered to detect small differences in responses. Other limitations of the study include the retrospective nature and missing datapoints. Historically, HRR mutations have been reported in 20–30% of men with advanced prostate cancer, where we found that over 79% of these platinum-treated patients harbored a HRR mutation. Given data of improved platinum sensitivity among other cancer types with HRR alterations, providers may be more likely to give patients with HRR-altered prostate cancer platinum chemotherapy¹⁵. This likely explains the high proportion of patients with HRR mutations seen in our cohort of patients.

In our retrospective study of patients with mCRPC, patients harboring HRR alterations had similar outcomes in response to platinum chemotherapy compared to patient without HRR alterations. Overall, there are limited safe and effective treatment options in patients with CRPC that progress after initial chemotherapy. This multicenter analysis suggests that platinum therapy may be active in advanced prostate cancer patients, irrespective of HRR alteration status. However, there may be other potentially predictive features that were not assessed in this series, highlighting the need for further research to better identify factors that may influence response to platinum therapy and to define the optimal patient population.