LAUSR

We read with interest the article by Ham et al [1], which investigates whether dividing Gleason score (GS) 8–10 disease into GS 8 and GS 9–10 provides prognostic information regarding mortality. We sought to validate their findings of increased hazards in GS 9–10 compared with GS 8 across other prostate cancer cohorts with longterm follow-up. We leveraged data from men of the Health Professionals Follow-up Study and the Physicians’ Health Study who were treated by radical prostatectomy (RP) over the period 1983– 2009. Of 1395 men, 279 (20%) scored as GS 8–10. In the GS 8 group (n = 108), 18 lethal or metastatic events were observed over a median of 10.2 yr follow-up. In the GS 9–10 group (n = 171), 37 events were observed over a median of 9.2 yr follow-up. The hazard ratio (HR) comparing GS 9–10 with GS 8 was 1.34 (95% confidence interval [CI]: 0.74–2.36; p = 0.31). The association was similar when adjusted for pathologic stage and age, with HR 1.39 (95% CI: 0.71–2.71). Analogous results emerged from 545 men who underwent RP over the years 1987–2001 at the Mayo clinic [2]. Among 211 men with GS 8–10, the odds ratio for lethal or metastatic disease comparing GS 9–10 with GS 8 was 1.21 (95% CI: 1.06–1.39; p = 0.01). Taken together, our analyses suggest that, while the trend of higher mortality in the GS 9–10 group compared with GS 8 persists, the prostate cancer-specific HR may be less than the estimate derived from the Johns Hopkins cohort, 2.38 (95% CI: 1.74–3.28). Discrepancies in our point estimates compared with those of Johns Hopkins may arise because of several reasons. Firstly, approximately 8.9% of all men in the Johns Hopkins cohort (2006–2016) suffered from GS 8–10 [3]. However, in the current study only 1047 men—which represents only 4.3%—were included in the final analysis. Therefore, around 1100 (4.6%) men were excluded because of ‘‘incomplete clinicopathologic or follow-up data’’ or a ‘‘history of neoadjuvant treatment.’’