Introduction

Although visceral metastases (VM) are widely recognized to portend worse prognoses compared to bone and lymph metastases in prostate cancer, currently little is known about what predicts VM and the extent to which men with VM do worse, particularly at the time of initial diagnosis of metastatic castration-resistant prostate cancer (mCRPC). Our study aimed to determine whether men with VM at initial mCRPC diagnosis have worse overall survival and identify predictors of VM.

Methods

We analyzed 494 men diagnosed with CRPC post-1999 and no known metastases from five Veterans Affairs hospitals of the SEARCH database who later developed metastases. Radiology scans (bone scan, MRI, CT scan, X-ray) within 30 days of initial metastasis diagnosis were reviewed to collect information on bone, visceral, and lymph node metastases. We analyzed the 236 men who had a CT scan performed. Predictors of VM and overall survival were evaluated using logistic regression and Cox models, respectively. Variables included age, year, race, treatment center, biopsy Gleason, primary localized treatment, metastases in lymph nodes, PSA, PSA doubling time, time from androgen deprivation therapy to CRPC, and time from CRPC to metastases.

Results

Of the 236 mCRPC patients, 38 (16%) had VM. Regarding VM, 19 (50%), 8 (21%), and 16 (42%) patients had metastases in the liver, lungs, and other locations, respectively. VM was a predictor of overall survival on crude analysis (HR=1.88, 95% CI 1.30-2.72, p=0.001) and after risk adjustment (HR=1.84; 95% CI 1.24-2.72, p=0.002). Age, year, treatment center, PSA, and time from CRPC to metastases were significant in predicting overall survival (all p<0.05). None of the variables tested were associated with having VM (all p>0.09).

Conclusions

Neither demographic, tumor, nor PSA-kinetic characteristics were predictive of having VM, but VM predicted worse overall survival. There are currently no known clinical predictors of VM; however, since patients with VM have worse overall survival, there remains a need for further research to determine what predicts VM. Efforts targeted at identifying novel biomarkers for VM have the potential to have the greatest impact for those suffering from this particular disease state.

Funding

None