Patients with prostate cancer who underwent radical prostatectomy, with or without lymph node dissection, were approached for enrolment on a prospective basis. Eligibility criteria included at least one of the following: pre-prostatectomy prostate specific antigen (PSA) ≥20 μg/L, pathological Gleason score ≥8, or pathological T3 disease. Post-operative PSA levels were not an eligibility criterion. Patients were ineligible for any of the following: Karnofsky Performance Score (KPS) <70, age <18, inadequate marrow function (platelets <100,000 cells/mm³, hemoglobin <10.0 g/dL, or AST/ALT ≥2x the upper limit of normal), post-operative PSA ≥5.0 μg/L, pelvic lymphadenopathy ≥1.5 cm or distant metastases on imaging, residual tumor detectable on exam or imaging; prior pelvic radiotherapy, malignancy within the past 5 years, or an active severe co-morbidity.

ADT began 2–10 days after patient enrolment. All patients received 28 days of bicalutamide (50 mg once daily) and luteinizing hormone-releasing hormone (LHRH) agonist began [intramuscular injection of 22.5 mg of Lupron (leuprolide acetate) depot every 3 months] on day 15. Radiotherapy started 2 months following ADT's initiation.

Radiotherapy was delivered with photon energies of at least 6 MV using either a 3D conformal radiotherapy (prior to 2012) or intensity modulated therapy technique thereafter. Treatment consisted of pelvic radiation (phase 1) with 44 Gy in 22 fractions followed by a boost (phase 2) of 22 Gy in 11 fractions to the prostate bed. The clinical target volume of the first phase of radiotherapy included the prostate bed, original location and or remnants of the seminal vesicles, and draining lymphatics (internal iliac, external iliac, obturator, distal common iliac, and presacral lymph nodes) (10). Dose was prescribed to the isocenter for the 3D conformal technique and to the prespecified volume for the intensity modulated radiotherapy approach (100% of the prescribed dose to cover 95% of the volume).

Pre-treatment assessment included history, physical exam, questionnaires [EORTC QLQ-C30 (11), EQ-5D (12), International Index of Erectile Function (IIEF-5) (13), and International Prostate Symptom Score (IPSS) (14)], serological assessments (CBC, electrolytes, urea, liver function tests, testosterone levels, and PSA), chest x-ray, CT or MRI of the abdomen & pelvis, and a total body bone scan. During radiotherapy, patients were seen weekly by their Radiation Oncologist. Observed toxicities were reported using the Common Toxicity Criteria version 3 scale (CTCAEv3) (15). Following radiotherapy, patients were assessed every 3 months until 2 years, every 6 months for 3 years, and annually thereafter. At each follow-up visit, PSA, testosterone, EQ-5D, IPSS, IIEF-5, and toxicity were assessed. EORTC QLQ-C30 questionnaires were completed annually. If there was biochemical progression or symptomatic changes worrisome for progression, bone scan and CT of the abdomen and pelvis were repeated.

The primary endpoint was PFS at 5 years. Progression included either biochemical failure or clinical progression. Prior to analysis, biochemical failure was defined as PSA exceeding nadir plus 0.2 μg/L and found to be elevated on two subsequent assessments. Clinical progression was defined as either a local or distant failure. A post-hoc reporting of the (1) primary endpoint among PSA subgroups (<0.34 and ≥0.34 μg/L) postulated to stratify the benefit PBRT treatment intensification and (2) of the PFS when biochemical failure was defined as nadir plus 2 μg/L were done to allow for an indirect comparison (2). Secondary endpoints included: (1) local failure rate at 5 years, (2) distant failure rate at 5 years, (3) toxicity rates, and (4) QoL. For the evaluation of testosterone recovery, the lower end of the normal range was considered 7 nmol/L.

The 5-year PFS with PBRT alone was estimated to be ~45% based on prior trial data (16). We designed this study with Fleming's one-stage method (17, 18). A 5 year biochemical and clinical progression free survival of 45% was considered ineffective for the combined treatment regimen with LT-ADT (24 months), PBRT, and PLNRT in this population whereas a 5 year progression-free survival of 65% or higher was considered to warrant further subsequent studies. Based on a binomial distribution with a one-sided type I error of 0.05, type II error of 0.15 (power of at least 85%), the study was designed to enroll 46 patients, with at least 26 patients required to be alive and progression free at 5 years to reject the null hypothesis in favor of alternative hypothesis (18).

Any PSA or testosterone level below the detectable threshold—typically <0.1 μg/L and <0.7 nmol/L, respectively—was set to 0 for the purpose of calculation. Castration was defined as a testosterone level <1.7 nmol/L while the patient was under the effect of ADT. Testosterone recovery was defined as the testosterone returning to the normal range (i.e., at least 7 nmol/L) and the end of ADT was considered to be 13 weeks following the final Lupron injection. Median follow-up was calculated from the first administration of Lupron injection. A Kaplan-Meier curve was constructed to obtain the biochemical PFS. We conducted a univariate analysis of baseline characteristics and testosterone levels at the time of treatment failure to investigate whether there was a statistical difference between groups of patients who failed vs. those who did not. Chi-square proportion test was used for binary data, otherwise Wilcoxon signed-rank test was employed.

Patients were recruited from 2010 to 2016 to meet the accrual goal of 46 patients. Three patients were not treated as per protocol: one was found to be ineligible, one opted for surgery, and one refused radiotherapy. No patients had connective tissue disorders. On review of the remaining 43 patients, two did not meet the prospectively stated trial criteria—their post-operative PSA's were 5.7 and 9.8 μg/L at registration.

Significant pathological features included 19% (n = 8) with positive lymph nodes, 40% (n = 17) with Gleason ≥8, 63% (n = 27) with extracapsular extension, and 18 (n = 42%) with seminal vesicle involvement (Table 1). Before radiotherapy, the median PSA value was 0.30 μg/L [interquartile range (IQR)−0.20 to 0.47]. For 5 patients, the PSA was below 0.1 μg/L, 4 had PSAs between 0.1 and 0.19 μg/L, 6 had PSAs of 0.20 μg/L, 21 had PSAs between 0.21 and 0.50 μg/L, 3 had PSAs between 0.51 and 1.00 μg/L, and 4 had PSAs >1.00 μg/L. ADT was initiated within a median of 71 weeks post-surgery.

Seven patients have experienced biochemical failure at a median follow-up of 5.2 years (Figure 1). PFS at 3, 4, and 5 years were 100.0 [95% Confidence Interval (CI) 91.5–100%], 93.9 (95% CI 86.1–100%), and 78.0% (95% CI 63.7–95.5%), respectively, surpassing the predefined 5-year PFS of significance of 65%. There was one new failure after 5 years (at 6.2 years). None of the patients with biochemical failure had detectable disease on either examination or restaging investigations (i.e., no clinical failures). There have been no deaths among the study's population.

When treatment failure was defined as nadir plus 2 μg/L, the first failure was 4.8 years with a 5-year PFS of 92.1% (95% CI 82.2–100%). Among participants with a PSA below 0.34 ug/L, at 5 years there were two failures a PFS of 87.8% (95% CI 73.4–100.0). Among those with PSAs of at least 0.34 ug/L, at 5 years there were four failures and a PFS of 70.0% (49.2–99.7%). A test of interaction was not significant (p = 0.66).

During ADT, 11 patients had testosterone values that did not remain <0.7 nmol/L and 3 had values that did not remain <1.7 nmol/L. Castration levels of testosterone (<1.7 nmol/L) did not predict for progression in a univariate model (p = 0.43, Table 2). There were no biochemical or clinical failures in the 9 patients with baseline PSA values under 0.2 μg/L. Of the 8 node positive patients, 1 had subsequent biochemical failure. The remaining univariate analyses did not demonstrate any significant difference in biochemical failures according to baseline testosterone levels, pathological T stage, extent of prostatic involvement, Gleason score, PSA levels, lymph node involvement, extracapsular extension, seminal vesicle invasion, or perineural invasion.

The acute and long-term adverse effects experienced by the patients are reported in Table 3. Excluding erectile dysfunction, no grade 3 toxicites were observed. Ten (23.3%) patients experienced a grade 2 toxicity. Regarding the overall QoL as assessed by the EQ-5D's visual analog score (12), there were no statistically significant changes observed while on or after ADT. While there was a decrease in QoL while patients were on ADT, this was not statistically significant (p = 0.39) and the mean differences are summarized in Figure 2. The mean minimum QoL experienced at any given time while on ADT by a patient was 7.8 (standard deviation = 2.0), compared to a mean baseline QoL of 8.2 (standard deviation = 1.2).

Of the 9 (20.9%) patients who experienced any grade of gastrointestinal toxicity, 2 (4.7%) patients presented with rectal bleeding, one of which was diagnosed as radiation proctitis. Six patients (13%) experienced 8 instances of grade 2 or greater toxicities that were possibly associated with ADT. The toxicities were hot flashes (n = 2), and single instances of headaches, hypertension, fatigue, breast tenderness, long-term gynecomastia with breast tenderness, or stroke. The single patient experiencing stroke had this 8 months after initiating ADT. Hormonotherapy was thus stopped for this patient after a total of 3 injections.

At baseline, 21 patients (48.8%) reported some degree of urinary incontinence. There were 2 patients in the acute and 2 in the chronic period with grade 2 urinary genitourinary toxicities and there was no new chronic incontinence. The median IPSS scores at baseline, the first appointment following radiotherapy, and 2 years following the start of ADT were 6 (IQR = 2.0–9.3), 7.5 (IQR = 4.0–13.0), and 4.5 (IQR = 2.5–7.5), respectively.

Regarding erectile dysfunction, baseline function was reported as a median IIEF-5 of 5 (IQR = 5.0–9.5) and 28 men (65.1%) had already Grade 3 toxicity (decreased erectile function not improve with intervention). Among the men with a documented baseline IIEF-5 of 10 or greater, the median IIEF-5 at baseline and 3 years or more following the start of ADT was 15 and 8.5, respectively. Grade 3 erectile dysfunction was reported in all patients while ADT was ongoing.

The number of patients who completed 18 and 24 months of ADT were 27 (62.7%) and 19 (44.2%), respectively. Testosterone recovered within normal ranges within a year of ADT completion in 15 (34.9%) patients and 10 patients have not recovered their testosterone to normal levels. For the patients who achieved castration while on ADT, the median time to testosterone recovery was 78.0 weeks (IQR 51.7–104.0) after the last ADT injection.

The authors declare that this study received funding from AbbVie Pharmaceuticals.