CHICAGO, IL USA (UroToday.com) - Several previous studies, as well as clinical practice, suggest that the site of metastatic disease influences overall survival among men with metastatic castrate resistant prostate cancer (CRPC). However, these studies include small numbers of patients, making their estimates of the influence of metastatic site on survival potentially inaccurate. Dr. Susan Halabi and colleagues sought to address this area of uncertainty by performing a pooled analysis of 5 randomized controlled clinical trials of men with metastatic CRPC to determine whether site of metastasis affects overall survival in this population. They hypothesized that liver metastases would be associated with the shortest overall survival, lung metastases would be associated with intermediate overall survival, and bone or lymph node metastases (non-visceral) would have the longest overall survival.
The investigators performed a systematic review of the literature and identified 15 trials including chemotherapy naïve men with metastatic CRPC. Of these, only 9 included overall survival as the primary endpoint and were eligible for inclusion. The investigators were able to gather the primary data from 5 of these trials, and these were the studies included in this report. They are in the process of obtaining the primary data from the remaining 4 trials, and plan to incorporate them into the analysis in the future.
Studies included in the meta-analysis were the following: TAX 327, SWOG 9916, CALGB 90401, ENTHUSE 33, and SWOG 0421. All were large intergroup, randomized, controlled trials evaluating overall survival as a primary endpoint. The studies spanned a treatment period from October 1999 – May 2011. In total, 3 993 men were included in the analysis. Men were identified as having either non-visceral metastases (bone or lymph node), or visceral metastases (liver, lung, or other). The investigators used fixed-effects meta-analysis to estimate the pooled hazard ratios and 95% confidence intervals to compare men with lung metastases versus non-visceral metastases, and liver metastases versus lung metastases. They adjusted for performance status, PSA, and age within the proportional hazard model.
Within the cohort, there were men with visceral and non-visceral metastatic disease. There were 187 (5%) with lymph node-only disease, 3 334 (83%) with bone metastases, 300 (7%) with lung (+/- bone) metastases, 280 (7%) with liver (+/- bone) metastases, and 55 (1%) with other visceral metastases. Median overall survival varied by site of metastasis as follows: 27.0 months for lymph node-only disease, 20.3 months with bone metastases, 16.5 months with lung (+/- bone) metastases, 12.1 months with liver (+/- bone) metastases, and 14.4 months with other visceral metastases. A pooled hazard ratio for death among men with lung metastases compared to patients with non-visceral metastases (lymph node or bone) was 1.3 (p < 0.001), and for men with liver metastases compared with men with lung metastases was 1.4 (p < 0.001).
The authors conclude that this information may be helpful in future clinical trial design as site of metastasis should likely be considered during statistical planning. It is unclear whether this information will be immediately applicable to clinical discussions with patients, as frequently men in practice have more comorbidities and may be older than the robust, healthy participant found in a clinical trial. At the very least, however, the study gives us a place to start when broaching a difficult topic with men in clinic.
Presented by Susan Halabi, PhD at the American Society of Clinical Oncology (ASCO) 50th Annual Meeting - May 30 - June 3, 2014 - Chicago, Illinois USA
Duke Cancer Institute, Durham, NC USA
Written by Alicia K. Morgans, MD, assistant professor of medicine and medical oncologist at Vanderbilt-Ingram Cancer Center, and medical writer for UroToday.com