No evidence of FGFR3 mutations in prostate cancer

Abstract

Background: FGFR3 mutations are associated with a good clinical disease course in bladder tumors. Currently, prognostic markers to stratify prostate cancer (PCa) patients for conservative management are lacking. Conflicting results have been found on the presence of FGFR3 mutations in PCa. Our objective was to determine the prevalence of FGFR3 mutations in a subset of prostate tumors. Next, determine the prevalence of FGFR3 mutations in PCa patients with coexistent tumors in other tissues.

Methods: Primary and locally advanced prostate tumors (n =132) were collected at our medical center. From the 132 PCa patients, 28 (21%) were diagnosed with coexistent primary tumors (bladder, skin, pancreas, renal cell, gastric, colon, hepatic, and lung). Tumors were analyzed by FGFR3 mutation analysis on exon 7, 10, and 15, known to harbor the most frequent mutations.

Results: The prevalence of FGFR3 mutations in patients with only PCa was 0%. Most PCa patients presented with coexistent bladder (n=12) and bladder and skin tumors (n =7). Other coexistent tumors in PCa patients included: bladder and pancreatic cancer (n=1); bladder and renal cell carcinoma (n=1); bladder and gastric carcinoma (n=1); skin cancer (n=1); colon cancer (n= 3); hepatic carcinoma (n=1); and lung cancer (n = 1). FGFR3 mutations were detected in 9/15 (60%) analyzed bladder tumors.

Conclusions: FGFR3 mutations were absent in the investigated prostate tumors, suggesting a minor role of these mutations in tumorigenesis. Hence, FGFR3 mutation analysis is not suitable to select patients for conservative management. Interestingly, if a prostate tumor coincided with other tumors these were mostly bladder and skin.