Appropriate timing of androgen deprivation treatment (ADT) for prostate cancer is controversial. Our aim was to determine whether immediate ADT extends survival in men with node-positive prostate cancer who have undergone radical prostatectomy and pelvic lymphadenectomy compared with those who received ADT only once disease progressed.

Eligible patients from 36 institutes in the USA were randomly assigned in 1988–93 to receive immediate ADT (n=47) or to be observed (n=51), with ADT to be given on detection of distant metastases or symptomatic recurrences. Patients were followed up every 3 months for the first year and every 6 months thereafter. The primary endpoint was progression-free survival; secondary endpoints were overall and disease-specific survival. Analysis was by intention to treat. To ensure that the treatment groups were comparable, we did a retrospective central pathology review of slides and regraded the Gleason scores for available samples. This trial predates the requirement for clinical trial registration.

At median follow-up of 11·9 years (range 9·7–14·5 for surviving patients), men assigned immediate ADT had a significant improvement in overall survival (hazard ratio 1·84 [95% CI 1·01–3·35], p=0·04), prostate-cancer-specific survival (4·09 [1·76–9·49], p=0·0004), and progression-free survival (3·42 [1·96–5·98], p<0·0001). Of 49 histopathology slides received (19 immediate ADT, 30 observation), 16 were downgraded from the original Gleason score (between groups ≤6, 7, and ≥8) and five were upgraded. We recorded similar proportions of score changes in each group (p=0·68), and no difference in score distribution by treatment (p=0·38). After adjustment for score, associations were still significant between treatment and survival (overall, p=0·02; disease-specific, p=0·002; progression-free survival, p<0·0001).

Early ADT benefits patients with nodal metastases who have undergone prostatectomy and lymphadenectomy, compared with those who receive deferred treatment. The beneficial effects of early ADT, rather than an imbalance in risk factors, are likely to explain the differences in outcomes between treatments.