Objective: Insulin-like growth factor-binding protein (IGFBP)-2 is a major prostatic IGFBP and may be involved in regulating prostate growth. Patients with prostate carcinoma (PC) have elevated serum IGFBP-2 levels which correlate with the specific PC marker, prostate-specific antigen (PSA). The aims of this study were (i) to investigate whether elevated serum IGFBP-2 is unique to PC or also occurs in benign prostatic hyperplasia (BPH), (ii) to examine the relationships among age, PSA and IGFBP-2 levels, and (iii) to examine longitudinal changes in serum IGFBP-2 with PSA in PC.
Design and patients: Sixteen patients (61-83 years) with inoperable PC attending the Oncology Unit at a tertiary referral hospital were studied. Some serum samples were obtained retrospectively while the majority were collected prospectively over 13 months of treatment. The patients with PC were compared to eight patients (66-73 years) with histologically-proven BPH and seven male control subjects (61-82 years) with no known prostate abnormality.
Measurements: A new IGFBP-2 RIA was developed. Serum PSA (by EIA), and IGFBP-2, IGFBP-3, IGF-I and IGF-II (by RIA) were measured in all subjects, and serially in patients with PC.
Results: Serum IGFBP-2 was significantly higher in PC with high PSA (560 +/- 66 micrograms/l, n = 12) than PC with normal PSA (292 +/- 65 micrograms/l, n = 4, P = 0.02), BPH (364 +/- 61 micrograms/l, P = 0.03) and controls (367 +/- 44 micrograms/l, P = 0.04). Mean IGFBP-2 in BPH was not different from controls. IGFBP-2 and PSA were significantly correlated with age (r = 0.543 and r = 0.433 respectively) and with each other even when the age effect was removed. Serum IGFBP-2 and PSA levels changed concordantly in all seven PC patients who had serial sampling. Serum IGF-II but not IGF-I or IGFB-3 was higher in PC and BPH than controls (PC 332 +/- 23 micrograms/l, BPH 359 +/- 26 micrograms/l vs. controls 241 +/- 37 micrograms/l; P = 0.03 and 0.02 respectively).
Conclusions: Serum IGFBP-2 levels are uniquely elevated in active prostate carcinoma but not. In benign prostatic hyperplasia. In prostate carcinoma, serum IGFBP-2 levels closely parallel those of prostate-specific antigen and probably reflect tumour burden. The relationship between prostatic-specific antigen and IGFBP-2 is partially independent of their individual relationships with age. Although serum IGFBP-2 is less sensitive than prostate-specific antigen in prostate carcinoma it may have adjunctive value in its management.