King's College London

Research portal

Ki-67 is an independent predictor of prostate cancer death in routine needle biopsy samples: proving utility for routine assessments

Research output: Contribution to journalArticle

Solène-Florence Kammerer-Jacquet, Amar Ahmad, Henrik Møller, Holly Sandu, Peter Scardino, Geraldine Soosay, Luis Beltran, Jack Cuzick, Daniel M Berney

Original languageEnglish
Pages (from-to)1303-1309
Number of pages7
JournalModern Pathology
Issue number9
Early online date11 Apr 2019
Publication statusPublished - Sep 2019

King's Authors


Standard clinical parameters fail to accurately differentiate indolent from aggressive prostate cancer. Our previous studies showed that immunohistochemical testing for Ki-67 improved prediction of prostate cancer death in a previous cohort of conservatively treated clinically localized prostate cancer. However there is a need for validation of usage with whole biopsy sections rather than tissue micro-arrays for use in routine diagnostics. Prostate cancer biopsy cases were identified in the UK, between 1990 and 2003, treated conservatively. Tumor extent and prostate-specific antigen (PSA) serum measurements were available. Biopsy cases were centrally reviewed by three uropathologists and Gleason conformed to contemporary ISUP 2014 criteria. Follow-up was through cancer registries up until 2012. Deaths were divided into those from prostate cancer and those from other causes. The percentage of Ki-67 in tumor cells was evaluated by immunohistochemistry on whole biopsy sections and was available for 756 patients. This percentage was used in analysis of cancer specific survival using a Cox proportional hazards model. In univariate analysis, the interquartile hazard ratio (HR) (95% confidence intervals) for continuous Ki-67 was 1.68 (1.49, 1.89), χ 1 2 = 47.975, P < 0.001. In grade groups 1 and 2, continuous Ki-67 was a statistically significant predictor of time to death from prostate cancer, HR (95% CI) = 1.97 (1.34, 2.88), χ 1 2 = 9.017, p = 0.003. In multivariate analysis, continuous Ki-67 added significant predictive information to that provided by grade groups, extent of disease and serum PSA, HR (95% CI) = 1.34 (1.16, 1.54), Δχ 1 2 = 13.703, P < 0.001. We now advocate the introduction of Ki-67 as a viable and practicable prognostic biomarker in clinical practice. The association of Ki-67 with mortality was highest in grade groups 1 and 2, showing that Ki-67 can be used as a routine biomarker in patients being considered for active surveillance.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454