Proliferating-cell nuclear antigen (PCNA) as an independent prognostic marker in patients after prostatectomy: a comparison of PCNA and Ki-67.

OBJECTIVE

To investigate the prognostic value of prostatic tumour cell proliferation, as measured by Ki-67 and proliferating cell nuclear antigen (PCNA), and to compare these measures in men at low and high risk for progression of tumour.

PATIENTS AND METHODS

Two groups of patients with prostate cancer, i.e. 'metastatic' (M, 22) who had pT3b-4aN0M0 and pTanyN1M0, and 'nonmetastatic' (NM, 18), who had < or =pT3aN0M0 disease, were selected from a well-examined and mapped group of 114 treated by radical prostatectomy. Patients in the NM group were selected by the criteria of having a Gleason score of < or = 7. To assess proliferation, 1000 cells were counted at x 400 magnification by two observers and the percentage of tumour cells positively stained with Ki-67 and PCNA defined as the Ki-67 and PCNA labelling index (LI), respectively. The two LI were compared in the NM and M groups, and the correlation of the LIs with pathological stage, progression and prostate-specific antigen (PSA)-free survival evaluated. Prognostic values of the LI were analysed using multivariate analysis.

RESULTS

The mean (range) follow-up was 33 (4-78) months. The mean LIs were higher in the M than the NM group for both PCNA and Ki-67 (P = 0.02 and 0.019, respectively). Both LIs were markedly different between the groups when stratified by progression, with both significantly higher in men with progression in the NM group. Both LIs had a significant association with Gleason score, pathological stage, progression and PSA-free survival. In multivariate analysis the PCNA LI, surgical margin status and pathological stage were independent factors for progression.

CONCLUSION

Tumour cell proliferation as assessed by Ki-67 or PCNA correlate significantly with progression. The PCNA LI was an independent predictor of progression, especially in patients with a low risk of progression according to predefined criteria.