Objectives
No uniformity exists in the definition of metastatic burden in metastatic hormone-naive prostate cancer (mHNPC) across clinical trials making their comparison challenging. We explored the definition conformity and prognostic significance of bulky mHNPC used in the CHAARTED and LATITUDE trial.
Materials
Since 2014, 69 patients with newly-diagnosed mHNPC were prospectively categorized as having high-volume (versus low-volume) and high-risk (versus no high-risk) disease according to the definition of CHAARTED and LATITUDE trial, respectively. Conformity was tested using Cohen’s kappa coefficient. Kaplan-Meier method was conducted to compare castration-refractory prostate cancer-free survival (CRPC-FS) and overall survival (OS) curves of both definitions. The prognostic significance was analyzed using univariate and multivariate Cox regression models. Other parameters included in these models were Gleason score, symptoms, performance status (ECOG) and PSA at diagnosis.
Results
Gleason score ≥8 was diagnosed in 60/69 (87%) patients. Patients were categorized as having M1a, M1b and M1c disease in 15 (22%), 42 (61%) and 12 (17%) of cases, respectively. In patients with bone metastasis (n=50), median number of bone metastases was 8 (range 1-63). Thirty-four (49%) and 36 (52%) patients were categorized as high-volume (CHAARTED) and high-risk disease (LATITIDE), respectively. Thirty (44%) patients were categorized as both low-volume and low-risk disease, 31 (45%) as both high-volume and high-risk disease. Five (7%) patients were categorized as low-volume but high-risk disease whereas 3 (4%) as high-volume but low-risk disease. Cohen’s kappa coefficient was 0.77 indicating an excellent agreement between both definitions (p<0.001).
Mean follow-up was 18 ±15 months. Median CRPC-FS for low and high-volume disease was respectively 40 and 11 months (p=0.001). Median CRPC-FS for low and high-risk disease was identical (40 and 11 months, respectively; p<0.001). Median OS for low and high-volume disease was not reached versus 23 months (p=0.002), and for low and high-risk disease not reached versus 24 months (p=0.004), respectively. The prognostic significance of both definitions remained significant in the multivariate model (both for CRPC-FS and OS).
Conclusion
There is an excellent conformity between the definitions of bulky mHNPC in the CHAARTED and LATITUDE trial. Both definitions have significant prognostic value for predicting CRPC-FS and OS.