Prostatic involution in men taking finasteride is associated with elevated levels of insulin-like growth factor-binding proteins (IGFBPs)-2, -4, and -5

BACKGROUND: Insulin-like growth factor-binding proteins (IGFBPs)-2, -4, and -5 are associated with upregulation of apoptosis in the ovary. The purpose of this study was to assess the roles of IGF-I and IGFBPs during involution of the prostate. Frozen and fixed tissue was collected by transurethral prostatectomy from Caucasian men, aged 52-82 years, scheduled for prostatectomy for benign prostatic hyperplasia, who took either placebo (n = 7) or the 5alpha-reductase inhibitor finasteride for 6 days to 6 years (n = 15) prior to surgery. METHODS: Intraprostatic androgen levels were measured by radioimmunoassay. Tissues were immunostained for IGF-I and IGFBP-2, -3, -4, and -5, and staining was quantitated by computerized image analysis. Serial sections were stained for markers of apoptosis (TUNEL and tissue transglutaminase) and IGFBP-2, -4, or -5. RESULTS: IGF-I staining was significantly decreased in the medium-term (18-43 days) treatment group and remained so for the duration of the study (P = 0.026). IGFBP-3 staining was unchanged in the early and medium-term treatment groups; however, a transient earlier rise in the level of this proapoptotic protein cannot be ruled out. The percentage of epithelial cell area staining positively for IGFBP-2 increased significantly, from 1.6 +/- 0.5 in the placebo group to 12.0 +/- 2.0 (P < 0.0001), and 7.6 +/- 1.9 (P = 0.003) in the short (6-13 days) and medium-term treatment groups, respectively. IGFBP-4 staining increased from 2.2 +/- 0.6 to 9.8 +/- 1.9 (P < 0.0001) and 7.4 +/- 1.2 (P = 0.004) in the short and medium-term groups, respectively, and IGFBP-5 staining increased from 0.2 +/- 0.1 to 3.8 +/- 2.0 (P = 0.004) in the medium-term group. The results from serial sections showed that IGFBP-2 and -4 costained with markers of apoptosis, while IGFBP-5 did not. CONCLUSIONS: These results indicate that IGFBP-2, -4, and -5 are associated with prostatic involution. Because of the timing and distribution of expression, we hypothesize that IGFBP-2 and -4 have a role as signals for apoptosis, but that IGFBP-5 likely does not.