Radical prostatectomy is potentially curative in patients with high-risk localized or locally advanced prostate cancer; however, relapse occurs within 5 years in up to 50% of patients.
We conducted a phase 3, double-blind, placebo-controlled trial in which patients with newly diagnosed high-risk localized or locally advanced prostate cancer were randomly assigned in a 1:1 ratio to receive androgen-deprivation therapy (ADT) plus apalutamide (240 mg per day) or ADT plus placebo for 6 cycles (28 days each) before and after radical prostatectomy with pelvic lymph-node dissection. The dual primary end points were a composite of pathological complete response or minimal residual disease (defined as a pathological stage of ypT2 or lower, with a tumor size of ≤5 mm in the greatest dimension) and metastasis-free survival, as assessed with conventional imaging or prostate-specific membrane antigen positron-emission tomography. Secondary end points included event-free survival, first subsequent treatment, and distant metastasis (assessed in time-to-event analyses), as well as safety.
A total of 2109 patients underwent randomization: 1057 were assigned to receive ADT plus apalutamide, and 1052 to receive ADT plus placebo. The median follow-up was 61.7 months. The percentage of patients with a pathological complete response or minimal residual disease was significantly higher in the apalutamide group than in the placebo group (8.9% vs. 1.0%; odds ratio, 10.17; 95% confidence interval [CI], 5.27 to 19.64; P<0.001), as was the percentage of patients with metastasis-free survival (probability of metastasis-free survival at 5 years, 78.2% vs. 73.5%; hazard ratio for distant metastasis or death, 0.80; 95% CI, 0.67 to 0.96; P=0.02). Event-free survival, time to the first subsequent treatment, and time to distant metastasis significantly favored ADT plus apalutamide over ADT plus placebo (P<0.001 for all between-group comparisons). Grade 3 or 4 adverse events occurred in 39.6% of the patients in the apalutamide group and in 31.0% of those in the placebo group, with the difference between the groups driven primarily by a higher incidence of rash in the apalutamide group.
Perioperative treatment with ADT plus apalutamide was associated with better oncologic outcomes of radical prostatectomy in patients with high-risk localized or locally advanced prostate cancer than treatment with ADT plus placebo. Adverse events were more common in the apalutamide group than in the placebo group. (Funded by Johnson & Johnson; PROTEUS ClinicalTrials.gov number, NCT03767244.)
PROTEUS: a landmark trial, but not yet a universal perioperative standard
The history of neoadjuvant systemic therapy in patients treated with radical prostatectomy for localised prostate cancer has been one of persistent biological promise and repeated clinical disappointment. For decades, the concept was attractive: high-risk disease may be systemic from the diagnosis, and earlier treatment of micrometastatic disease should, in theory, improve long-term outcomes. Yet previous neoadjuvant studies mostly improved pathological parameters without convincingly changing clinically meaningful endpoints. In this context, the phase 3 PROTEUS trial represents an important moment for the field.
In more than 2100 patients with newly diagnosed high-risk localised or locally advanced prostate cancer undergoing radical prostatectomy and pelvic lymph-node dissection, perioperative androgen deprivation therapy (ADT) plus apalutamide improved both co-primary endpoints compared with perioperative ADT plus placebo. Pathological complete response or minimal residual disease differed – 1.0% vs. 8.9%; the same pattern was observed for 5-year metastasis-free survival 73.5% vs. 78.2% (HR 0.80, 95% CI, 0.67-0.96; p=0.02). Event-free survival, time to first subsequent local or systemic therapy, and time to distant metastasis also favoured apalutamide. These are meaningful results from a large, double-blind, international phase 3 trial with a median follow-up of approximately five years.
The main message is therefore clear: PROTEUS provides the strongest evidence so far that perioperative androgen receptor pathway intensification can modify the natural history of surgically treated high-risk prostate cancer. This alone makes the trial a milestone. However, a milestone is not necessarily the same as an immediate universal standard. The more difficult question is not whether perioperative intensification can work, but which patients truly require it.
Several aspects of the trial deserve careful interpretation. Although pathological complete response or minimal residual disease increased almost ten-fold, the absolute improvement in 5-year metastasis-free survival was less than 5%. True ypT0 responses remained uncommon. This observation is important because it reminds us that prostate cancer does not behave like several other solid tumours in which pathological complete response is a well-established marker of treatment effect. In localised prostate cancer, improved pathological findings may reflect tumour cytoreduction and androgen sensitivity, but they do not necessarily guarantee eradication of occult systemic disease.
Second, the metastasis-free survival endpoint was both modern and challenging. PROTEUS is among the first phase 3 trials in localised prostate cancer to incorporate PSMA-PET into a primary endpoint. This is clinically relevant, since PSMA-PET is now widely used in contemporary practice and detects metastatic disease at substantially lower disease burden than conventional imaging. At the same time, most metastatic events in the trial were detected by PSMA-PET, while metastasis-free survival assessed by conventional imaging alone was not significantly improved (HR 0.84; 95% CI 0.67-1.07). Therefore, part of the observed benefit may represent delayed PET-detected progression rather than prevention of conventionally visible metastatic disease. Whether PSMA-PET-defined metastasis-free survival will ultimately translate into overall survival benefit remains uncertain. Unlike conventional imaging-based metastasis-free survival, which has been validated as a surrogate endpoint for overall survival in localised prostate cancer, the surrogacy of PSMA-PET-defined metastasis-free survival has not yet been established. Consequently, caution is warranted when interpreting the clinical significance of relatively small differences driven primarily by molecular imaging findings.
This does not mean that the endpoint is clinically irrelevant. Delaying metastatic progression, further imaging, salvage treatment, and additional systemic therapy may matter to patients, even before an overall survival advantage is demonstrated. The significant delay in first subsequent therapy is therefore an important and patient-centred result. Nevertheless, the reliance on a PSMA-PET-driven endpoint should make us cautious when interpreting the magnitude of benefit and when comparing PROTEUS with older trials based on conventional imaging.
Third, the control arm must be considered. PROTEUS tested apalutamide added to perioperative ADT; it did not compare perioperative intensification with a contemporary surgery-first strategy followed by selective early salvage radiotherapy and/or ADT when indicated. This is not a flaw in the scientific design, which was rigorous and appropriate for isolating the contribution of apalutamide. However, it limits direct translation into daily urological practice. Many men with high-risk disease today are managed with radical prostatectomy first, followed by treatment escalation based on adverse pathology, PSA kinetics, imaging, and patient preference. PROTEUS does not fully answer whether all such patients should receive one year of ADT plus apalutamide around surgery.
Fourth, the potential for overtreatment is real. The trial population was high risk, but not uniformly ultra-high risk. Most patients had Gleason score 8-10 disease, but approximately 60% had PSA levels below 20 ng/ml and only about one third had clinical T3/T4 disease. High-risk prostate cancer defined by clinicopathological criteria is not the same as biologically high-risk prostate cancer. Some tumours remain highly androgen-dependent and may benefit substantially from androgen receptor pathway inhibition. Others may already harbour aggressive, less androgen-dependent biology, be de-differentiated tumours, and derive limited benefit. Treating all patients according to broad high-risk criteria risks exposing many men to toxicity without ensuring proportional benefit. This was seen in recent STAMPEDE secondary analyses, where only patients in the highest 25% digital pathology multimodal artificial intelligence (MMAI) model score, benefited from systemic combination therapy.
This issue is particularly important because the treatment burden was not negligible. Grade 3 or 4 adverse events occurred in 39.6% of patients in the apalutamide group compared with 31.0% in the placebo group. Testosterone recovery to at least 200 ng/dl occurred in 81.6% and 83.0% of patients, respectively, meaning that a relevant minority did not recover to this threshold during follow-up. Overall survival was immature, with overall mortality of 8.5%; numerically, the hazard ratio for death was 1.08 in the apalutamide group compared with placebo. This finding should not be overinterpreted, as it was not considered evidence of excess mortality by the investigators, but it supports the need for longer follow-up before concluding that the benefits clearly outweigh the long-term risks for all eligible patients.
From a practical perspective, PROTEUS should therefore be viewed as proof-of-principle rather than the final answer. It convincingly shows that perioperative apalutamide plus ADT can improve pathological response and delay disease progression in men undergoing radical prostatectomy for high-risk localised or locally advanced prostate cancer. At the same time, the modest absolute improvement in 5-year metastasis-free survival, the immature overall survival data, the PSMA-PET-driven endpoint, the absence of a surgery-first comparator, and the toxicity burden argue against uncritical routine use in every patient meeting high-risk criteria.
The future after PROTEUS should be more selective, not simply more intensive. Emerging tools such as multimodal artificial intelligence models, digital pathology, genomic classifiers, and molecular imaging may help identify the subset of patients in whom perioperative intensification provides meaningful benefit beyond optimal local therapy. In this sense, PROTEUS may be remembered less as the trial that established universal perioperative ARPI therapy, and more as the trial that opened the door to perioperative precision medicine in localised prostate cancer.
For now, the most balanced interpretation is that PROTEUS has changed the evidence landscape but should be implemented cautiously. It provides a new option for selected patients with very high-risk features after multidisciplinary discussion, while reinforcing the need for longer follow-up, patient-reported outcomes, cost-effectiveness analyses, and predictive biomarkers. The challenge ahead is not only to treat high-risk prostate cancer earlier, but to treat the right patients earlier.