Prostate cancer Ki-67 (MIB-1) expression, perineural invasion, and gleason score as biopsy-based predictors of prostate cancer mortality: the Mayo model.

OBJECTIVE

To determine the role of cellular proliferation and other biopsy-based features in the prediction of prostate cancer mortality.

PATIENTS AND METHODS

Between 1993 and 2012, our institution has performed quantitation of prostate cancer DNA ploidy and Ki-67 (MIB-1) on most prostate cancer needle biopsy specimens. The outcomes of 451 consecutive patients with biopsy-proven cancer treated by radical prostatectomy between January 24, 1995, and December 29, 1998, without neoadjuvant hormonal therapy were assessed. Clinical and biopsy information obtained before radical prostatectomy was placed in multivariate Cox proportional hazards regression models to predict local or systemic progression and cancer-specific death. Predictive ability was evaluated using a concordance index.

RESULTS

With a median follow-up of 12.9 years, 46 patients experienced local or systemic progression, and 18 patients died of prostate cancer. On multivariate analysis, the biopsy features of Ki-67 expression, perineural invasion, and Gleason score were associated with local or systemic progression. Ki-67 expression, perineural invasion, and Gleason score were associated with cancer-specific death with a concordance index of 0.892. After adjusting for perineural invasion and Gleason score, each 1% increase in Ki-67 expression was associated with a 12% increased risk of cancer-specific death (P<.001). Ki-67 expression alone was a strong predictor of cancer-specific outcomes and improved the predictive ability of currently used algorithms.

CONCLUSION

This study documents that long-term prostate cancer outcomes are best estimated with a combination of Gleason score, perineural invasion, and Ki-67 expression. Given its low cost, rapid assessment, and strong predictive power, we believe that adding Ki-67 expression to perineural invasion and Gleason score at biopsy should be considered a standard by which all new biomarkers are compared before introducing them into clinical practice.