Real-World Treatment Patterns and Overall Survival of Patients with Metastatic Castration-Resistant Prostate Cancer in the US Prior to PARP Inhibitors

Abstract

Introduction: Therapeutic options for metastatic castration-resistant prostate cancer (mCRPC) patients are continuously advancing. We described mCRPC treatment patterns in the US from 2013 to 2019.

Methods: Patients with a confirmed mCRPC diagnosis and adenocarcinoma histology were included in the US Flatiron Health Electronic Health Record-derived de-identified database. Treatment patterns [including treatment per lines of therapies (LOTs), LOT sequences, and time on treatment] and overall survival (OS) have been described in mCRPC settings.

Results: Of 5213 patients (mean age: 72.6 years), 4374 (83.9%) were treated with ≥ 1 LOT post-mCRPC diagnosis (among those with ≥ 1 LOT, 55.3%, 29.5%, 14.7%, and 6.7% had ≥ 2, 3, 4, and 5 LOTs, respectively). In first line (1L), the main treatment class was next-generation hormonal agents (NHA; 62.5% of patients with ≥ 1 LOT), while the shortest and longest time on 1L were observed for chemotherapy (median 2.8 months) and NHA (median 5.1 months), respectively. The most common LOT sequences were NHA → NHA (29.4% of patients with ≥ 2 LOTs) and NHA → NHA → chemotherapy (16.7% of patients with ≥ 3 LOTs). In Kaplan-Meier analyses, the median OS was 19.4, 14.6, and 11.1 months post-1L, 2L, and 3L start, respectively. Patients who moved rapidly through LOTs had an increased risk of death.

Conclusions: NHA were widely used as 1L therapy in mCRPC patients from 2013 to 2019, but time on 1L NHA treatment was on average < 6 months. While NHA → NHA was the most observed 1L → 2L LOT sequence, a plethora of other LOT sequences were observed. OS was poor, highlighting an unmet need for life-prolonging treatments.

Keywords: Chemotherapy; Electronic medical records; Lines of therapy; Next-generation hormonal agents; Treatment duration.

Fig. 1

MCRPC treatment patterns: number and proportion of patients by line of therapy and treatment class. 1L first line, 2L second line, 3L third line; 4L fourth line; 5L fifth line; LOT line of therapy, mCRPC metastatic castration-resistant prostate cancer, NHA next-generation hormonal agent, Sip-T sipuleucel-T. 1. Other therapy classes include thalidomide, bcg vaccine, nivolumab, atezolizumab, lenolidomide, durvalumab, ipilimumab, pembrolizumab, targeted therapies, radium-223, and clinical study drugs

Fig. 2

mCRPC treatment sequences: time on and time off treatment. 1L first line, 2L second line, 3L third line; Chemo chemotherapy, Combo combination therapy, LOT line of therapy, mCRPC metastatic castration-resistant prostate cancer, NHA next-generation hormonal agent, Sip-T sipuleucel-t. *For Sip-T only the start date is available in the data, and thus we cannot distinguish between time on and off treatment for this agent. 1. Only the top 10 sequences are shown (of a total of 25 observed 1L → 2L sequences and 97 observed 1L → 2L → 3L sequences; of note, there are a total of 25 possible sequences for 1L → 2L and a total of 125 possible 1L → 2L → 3L sequences by combining 5 treatment classes across 2 and 3 LOTs, respectively)

Fig. 3

Top 5¹ most common regimens (agent level) by lines of therapy. 1L first line, 2L second line, 3L third line, 4L fourth line; 5L fifth line; mCRPC metastatic castration-resistant prostate cancer, Sip-T sipuleucel-T. 1. Top 5 regimens at an agent level out of a total of 100 regimens observed in 1L, 123 regimens observed in 2L, 94 regimens observed in 3L, 78 regimens observed in 4L, and 66 regimens observed in 5L

Fig. 4

Overall survival for patients with mCRPC. A Overall survival after 1L start. B Overall survival after 2L start. C Overall survival after 3L start. Number of patients still observed at the specific point in time: 1040 patients participating in clinical trials or having a diagnosis of another cancer were excluded from the analysis; 466 patients participating in clinical trials or having a diagnosis of another cancer were excluded from the analysis; 300 patients participating in clinical trials or having a diagnosis of another cancer were excluded from the analysis