**1. Introduction**

The treatment landscape of metastatic prostate cancer (mPCa) has completely changed over the last decades. In 2004, docetaxel was the first drug to demonstrate an overall survival (OS) benefit of 2.4 months in mPCa, compared with mitoxantrone, and was approved for the treatment of men with metastatic castration-resistant prostate cancer (mCRPC) [1]. Cabazitaxel showed a similar OS increase compared with mitoxantrone and became a second-line treatment option for mCRPC in 2010 [2]. Subsequently, abiraterone acetate and enzalutamide were approved in both post-docetaxel [3,4] (2011–2012) and pre-docetaxel mCRPC [5,6] (2013–2014), reporting OS advantages between 4.0 and 4.8 months compared with placebo (Figure 1). Docetaxel was also introduced for the hormone-sensitive phase of mPCa (mHSPC) in 2015 [7]. Several androgen-receptor signaling inhibitors (ARSi)—abiraterone, enzalutamide, and apalutamide—were then approved for the treatment of mHSPC [8].

**Figure 1.** Regulatory timeline of approvals in advanced prostate cancer therapies. DDR+: DNA damage response genes mutated; mHSPC: metastatic hormone-sensitive prostate cancer; mCRPC: metastatic castration-resistant prostate cancer; nmCRPC: nonmetastatic castration-resistant prostate cancer; post-doc: post-docetaxel; pre-doc: pre-docetaxel.

Although the aforementioned randomized trials showed significant survival improvements in the first- and second-line of mCRPC, the real-world survival benefit in the population of patients outside of clinical trials is largely unexplored. The ideal population of patients enrolled in clinical trials might overestimate the true benefit induced by approved drugs in the general population of patients with newly diagnosed mPCa. For example, not all patients can receive chemotherapy. Although no specific advice is included in the U.S. National Comprehensive Cancer Network guidelines, the European Association of Urology guidelines recommend that docetaxel should be only offered to mHSPC patients who are fit enough for chemotherapy [9]. Of note, the STAMPEDE trial of docetaxel in mHSPC only included patients fit for chemotherapy and without significant cardiovascular history. Many patients with mPCa in the real-world are elderly with many comorbidities, and they cannot receive chemotherapy [10]. In addition, patients with poor general conditions or poor performance status are often not suitable for aggressive anticancer therapies. Moreover, although some retrospective data have been reported [11], no randomized trial has ever assessed the long-term, cumulative benefit on survival that can derive from the temporal sequence of different treatment strategies. Finally, the U.S. insurance policies or limited access to healthcare services could contribute to producing a discrepancy between the expected survival gain and the real-world data [12].

Here, we investigated the survival trends and prognostic variables in patients with de novo mPCa included in the U.S. Surveillance, Epidemiology, and End Results (SEER) database. Given the introduction of chemotherapy in 2004 and of ARSi in 2011, we hypothesized that a significant difference in OS and cancer-specific survival (CSS) was detectable in patients diagnosed in three time periods: 2000–2003 (Cohort A), 2004–2010 (Cohort B), and 2011–2014 (Cohort C). Of note, our study should not be intended to provide data on the efficacy of the newer treatments, but to provide epidemiological results about the survival trends in patients with de novo mPCa diagnosed in the United States in the last two decades.

#### **2. Results**

## *2.1. Study Cohort*

Our selection criteria identified 26,434 patients with de novo mPCa diagnosed between 2000 and 2014. Of these, 6047 were diagnosed between 2000 and 2003 (Cohort A), 11,815 between 2004 and 2010 (Cohort B), and 8572 between 2011 and 2014 (Cohort C). The main characteristics of the study population are summarized in Table 1. Overall, 68.3% of patients were ≥65 years. The percentage of patients younger than 75 years was higher in Cohort C compared to Cohorts B and A (64.8% vs. 61.1% vs. 58.3%, respectively). The majority of patients were white (62.7%), followed by black (19.4%) and Hispanic (11.6%). Metastatic classification (American Joint Committee on Cancer (AJCC), 6th edition) was available for Cohorts B and C. The majority of patients were M1b (72.7%), with a significant difference between Cohorts B (70.1%) and C (76.4%). The full contingency table with the comparison of baseline characteristics among the cohorts is available in Table S1. The median follow-up was 25, 26, and 29 months in Cohorts A, B, and C, respectively, with a median follow-up of censored patients of 60, 60, and 51 months.
