Treatment patterns and clinical effectiveness in metastatic castrate resistant prostate cancer after first-line docetaxel.

BACKGROUND Treatment for metastatic castrate-resistant prostate cancer in community settings is not well understood. OBJECTIVE To examine treatment patterns, sequencing, and outcomes in patients receiving second- and third-line treatment after first-line docetaxel. METHODS We used a community oncology database to identify patients who progressed after line 1 docetaxel (D) and received line 2 cabazitaxel (DC), abiraterone (DA), or other therapy (DO). Progression-free survival (PFS) and overall survival (OS) were assessed using Kaplan- Meier and Cox regression models. Line 3 included subsets DCA and DAC. RESULTS Line 2 groups (DC = 60 patients, DA = 71, DO = 153) did not differ significantly on demographic and clinical characteristics or median PFS on docetaxel therapy. Cox regression for OS by line 2 groups showed increased risk for DA compared with DC (HR, 1.69; P = .026) when 24 untreated DO patients were excluded. A similar nonsignificant pattern was observed when the 24 untreated patients were included. Of patients receiving DC in line 2, a nominally greater proportion received A in line 3 (57%, 34 of 60 patients) than did patients who received DA in line 2 followed by C in line 3 (25%, 18 of 71). LIMITATIONS There was a small sample for line 3, and unexamined confounds and selection biases in observational research. Conclusions Treatment patterns in community settings following docetaxel are complex and may involve multiple hormonal agents prior to disease progression. Cabazitaxel may not be optimally used in advanced disease. Although Cox regression showed increased risk of death for DA compared with DC, results need to be validated prospectively. CONCLUSIONS Treatment patterns in community settings following docetaxel are complex and may involve multiple hormonal agents prior to disease progression. Cabazitaxel may not be optimally used in advanced disease. Although Cox regression showed increased risk of death for DA compared with DC, results need to be validated prospectively.