A recent study finds that treatment-emergent small-cell neuroendocrine prostate cancer occurs in 17% of patients with the metastatic, castration-resistant form of the disease. This small-cell variety of prostate cancer shows a distinct gene expression signature and reduces patients' survival by nearly 20%.

An aggressive variety of small-cell prostate cancer that sometimes emerges after treatment is a distinct subtype and reduces patient survival by nearly 20%, a recent study shows. The shortened survival highlights the need to identify treatments specific to this subtype.

Less than 1% of patients diagnosed with prostate cancer have the highly aggressive small-cell variety. However, treatment-emergent small-cell neuroendocrine prostate cancer (t-SCNC) can develop in patients who've received androgen receptor–blocking drugs. Whether pretreatment and post-treatment small-cell prostate cancers are separate subtypes has remained unclear, in part because researchers have lacked enough biopsy samples to characterize t-SCNC.

Rahul Aggarwal, MD, of the University of California, San Francisco, and colleagues analyzed biopsy specimens from 160 men with metastatic, castration-resistant prostate cancer. Nearly three quarters of the patients had received abiraterone (Zytiga; Janssen), enzalutamide (Xtandi; Astellas Pharma), or both prior to biopsy, and all the men showed histologic evidence of adenocarcinoma at diagnosis.

The researchers determined that 17% of the patients developed t-SCNC. The tumors from 74% of these patients displayed only small-cell histology; the remaining patients had tumors containing t-SCNC and adenocarcinoma cells.

The scientists found that t-SCNC tumors differed from adenocarcinomas in several ways: For example, the researchers could distinguish the two tumor types using the expression levels of 106 genes. In addition, t-SCNC tumors rarely carried mutations in DNA-repair genes such as BRCA1 and RAD51, which are common in adenocarcinomas. Those distinctions may mean that “you are seeing different methods of developing resistance to androgen receptor therapies” in the two tumor types, says Aggarwal. Furthermore, median survival for patients with t-SCNC was shorter than for those with only adenocarcinoma cells—36.6 months versus 44.5 months, respectively.

The team's results also indicate that the two types of small-cell tumors are distinct. Pretreatment small-cell tumors lack the androgen receptor, usually affect patients who have low PSA levels, and frequently metastasize to the viscera. In contrast, t-SCNC tumors show high levels of androgen receptor expression, typically occur in patients with PSA levels greater than 60 ng/mL, and are less likely to metastasize to organs such as the liver.

An earlier study suggested that the incidence of t-SCNC was 1% in patients with advanced prostate cancer, but the new work indicates that “this small-cell neuroendocrine cancer is far more prevalent than we previously thought,” says Aggarwal. He and his colleagues are now testing sequential biopsies from patients to determine if t-SCNC cells are present when the tumor forms or whether they arise following androgen receptor–blocking therapy.

“This is a good and important paper,” says Scott Tagawa, MD, of Weill Cornell Medical College in New York, NY, who wasn't connected to the study. He notes that the authors identify an unmet clinical need by showing that t-SCNC has a significant impact on patient survival.

The researchers deserve credit for obtaining biopsy samples from metastases in a large cohort of patients receiving treatment, rather than relying on samples from tumor banks as previous studies had done, notes Matthew Galsky, MD, of the Icahn School of Medicine at Mt. Sinai in New York, NY, who also wasn't connected to the study. “The investigators did something no one else had done” in studies of small-cell prostate cancer, he says.

What researchers need to do now, Aggarwal says, is to develop noninvasive ways to test for t-SCNC and find ways to treat it. The researchers' gene expression analysis identified several genes that appear to orchestrate progression from adenocarcinoma to t-SCNC that are possible drug targets, including EZH2. The EZH2 inhibitor tazemetostat (Epizyme) is already being tested for other cancers and might be effective against t-SCNC. –Mitch Leslie