Introduction
Immediate or delayed androgen deprivation therapy (ADT) with initial surveillance is the preferred treatment strategy for biochemically recurrent prostate cancer (PCa) after radical prostatectomy and/or radiotherapy.
1,2 Analysis of the relapse patterns after primary PCa treatment suggests that most patients relapse with three or fewer metastases,
3-5 often termed oligorecurrences. In 1995, Hellman and Weichselbaum
6 hypothesized that eradicating oligorecurrences with metastasis-directed therapy (MDT) might prevent additional metastatic spread and improve survival.
7-8 Confirming this hypothesis could shift the paradigm for metastatic PCa from a palliative to a potentially curable disease in a subset of patients. Several retrospective, single-arm studies have supported the notion that MDT delays additional clinical progression and the start of subsequent palliative ADT, with minimal adverse events.
9,10 However, such an advantage has yet to be shown in well-controlled randomized studies. To address this, we compared the time to the start of ADT after MDT or surveillance (standard of care) for patients with oligorecurrent PCa with three or fewer metastases.
Results
Between August 1, 2012, and August 31, 2015, 62 patients were randomly assigned (
Fig 1). Patient characteristics are listed in
Table 1. The median follow-up time for the whole cohort was 3 years (interquartile range [IQR], 2.3-3.8 years). The type of MDT used was SBRT (n = 25), surgery in 6 patients (sPLND in 5 and lung metastasectomy in 1).
For the ITT analysis, the median ADT-free survival was 13 months (80% CI, 12 to 17 months) for the surveillance group and 21 months (80% CI, 14 to 29 months) for the MDT group (HR, 0.60 [80% CI, 0.40 to 0.90], log-rank
P = .11;
Fig 2A).
Table 2 lists the reasons for starting ADT, with polymetastatic progression being the main reason in both groups. No symptomatic or local progression was observed in the MDT group, as compared with three and six occurrences, respectively, in the surveillance group. In the MDT group, 11 patients were treated with a repeated course of MDT because of oligometastatic progression at the time of first radiographic progression, and two patients received two additional courses of MDT for oligometastatic progression. In the subgroup analysis (
Fig 3A), no significant interaction was observed between the effect of MDT and PSA doubling time or the location of metastases (
P value of interaction, .35 and .31, respectively). Because of the difference in the number of lesions between the surveillance and the MDT arm (
Table 1), a post hoc analysis was performed, with stratification on the number of metastases. The corresponding HR was 0.64 (80% CI, 0.42 to 0.96); log-rank
P = .16. This is in line with the findings stratified on the location of metastases and PSA doubling time. Eleven patients developed castration-resistant PCa, six in the surveillance arm and five in the MDT arm. In total, four patients died, three from other cancers and one from cardiac failure.
For the PP analysis, the median ADT-free survival was 12 months (80% CI, 7 to 17 months) for the surveillance group and 21 months (80% CI, 16 to 28 months) for the MDT group (HR, 0.55 [80% CI, 0.36 to 0.85]; log-rank
P = .08;
Fig 2B). In the subgroup analysis (
Fig 3B), no significant interaction was observed between the effect of MDT and the location of the metastases (
P value of interaction, .95). We did observe a larger magnitude of association between MDT and improved ADT-free survival in patients with a PSA doubling time of ≤ 3 months as compared with > 3 months (
P value for interaction, .01).
The PSA change is depicted as a waterfall plot at 3 months after random assignment and as the largest change in PSA (
Figs 4A and 4B). In total, 74% of patients treated with MDT had a PSA decline, as compared with 42% in the surveillance arm. The median time until PSA progression for the ITT was 6 months (80% CI, 4 to 7 months) for the surveillance group, as compared with 10 months (80% CI, 8 to 13 months) for the MDT group (HR, 0.53 [80% CI, 0.37 to 077];
P = .03;
Fig 4C). For the PP analysis, the HR was 0.52 (80% CI, 0.36 to 0.76);
P = .02. Seventy-five percent of patients treated with MDT experienced a PSA decline, as compared with 35% in the surveillance arm.
Six (17%) of the 36 patients treated with MDT experienced grade 1 toxicity. No grade 2 or higher toxicity was observed. The toxicities observed after SBRT were temporary looser stools (n = 1) and temporary muscle soreness (n = 1). One patient who underwent a video-assisted thoracoscopic lung metastasectomy developed thoracic wall pain requiring nonopioid analgesics for 6 months. After sPLND, toxicities observed included hypoesthesia of the genitofemoral nerve (n = 1), lymphorrhea (n = 1), and scrotal and penile edema (n = 1).
In total, 60 (97%) of 62 patients completed the baseline QLQs, 55 (89%) completed them at 3 months, and 52 (84%) completed them at 1 year.
Figure 5 depicts the global health score for both arms at the three time points for the ITT group. We did not observe clinically relevant changes in HRQOL scores between the two study arms (across all visits; Data Supplement). Overall, HRQOL scores remained stable for both groups at baseline, 3 months, and 1 year (Data Supplement).
Discussion
To our knowledge, this is the first randomized trial of MDT to all lesions versus surveillance alone for patients with oligorecurrent PCa after primary treatment with curative intent. Patients treated with MDT experienced a longer time to start of palliative ADT than did those who underwent surveillance alone. Time to ADT is a composite end point chosen because of the effect of the start of ADT on health care costs and HRQOL.
23 Recently the Intermediate Clinical Endpoints in Cancer of the Prostate working group identified metastasis-free survival as a surrogate for overall survival.
24 Unfortunately, this end point is valid only for localized PCa and, as such, cannot be extrapolated to patients with metastatic disease.
Many retrospective case series with promising results have been reported. However, because of the absence of comparative or randomized trials, the overall low number and heterogeneity of patients treated, a recent systematic review concluded that MDT should not be considered the standard of care.
9 Nevertheless, approximately two thirds of experts at the 2017 Advanced Prostate Cancer Consensus Conference had already considered MDT a treatment option for patients with oligorecurrent PCa despite only retrospective evidence of its value.
25 A multi-institutional case series tried to overcome some of the shortcomings of the reported retrospective studies by using fixed inclusion and exclusion criteria, resulting in an ADT-free survival of 28 months (95% CI, 16.2 to 69.7 months) after SBRT oligorecurrent PCa.
26Nevertheless, important limitations were still present because a comparator arm was missing and 50% of these patients received a temporary course of adjuvant ADT at the time of SBRT. This temporary course of ADT is probably responsible for the better ADT-free survival than that of the current trial, because the addition of temporary ADT to radiotherapy is known to prolong progression-free survival and overall survival in both high-risk and biochemical recurrent PCa.
2 Consequently, we believe it is worthwhile to investigate the addition of a temporary systemic drug to MDT in future trials. The synergistic approach might improve the therapeutic ratio by eradicating microscopic disease, which is still often missed by choline PET-CT. This is demonstrated clearly by the current trial—30% of patients treated with MDT progressed to polymetastatic disease within the first year. Advances in imaging, such as
68Ga prostate-specific membrane antigen (PSMA) PET-CT, might also improve patient selection for MDT.
27 PSMA-PET, which is widely available and has a better sensitivity and specificity than does choline PET-CT,
27 holds great promise in this field. The Advanced Prostate Cancer Consensus Conference consensus panel agreed, with 78% of the panelists voting for one of the next-generation imaging methods to restage biochemically recurrent PCa (PET-CT and/or whole-body magnetic resonance imaging), with 76% preferring PSMA over fluciclovine (10%) or choline (6%).
25 In the case of nodal recurrences, one might also consider opting for salvage lymph node dissection or whole pelvis radiotherapy to consider microscopic disease.
28,29To our knowledge, this randomized trial provides the first evidence of the effect of MDT compared with surveillance with delayed ADT. We demonstrated that MDT results in a PSA response in three out of four patients. A remarkable observation was the fact that 35% of patients undergoing surveillance experienced spontaneous PSA declines without receiving any therapy (
Figs 4A and 4B). However, these declines were not durable, as demonstrated by the fact that only approximately 20% of men were free from PSA progression at 1-year follow-up and approximately 10% at 2-year follow-up. Nevertheless, these data support the Hellman and Weichselbaum
6 concept of oligometastases—that certain tumors have not fully developed their metastatic potential and show a slow natural history.
Our subgroup analysis showed a comparable benefit with MDT for patients with nodal and non-nodal metastases. Retrospective data suggested that patients with a PSA doubling time < 3 months had a shorter time to progression-free survival with MDT than did patients with a longer doubling time. From our PP data, it seems that the magnitude of benefit with MDT as compared with surveillance is greater for those patients with a shorter PSA doubling time. Consequently, there is no rationale to decline MDT for patients with a short PSA doubling time.
The global health status of men with biochemically recurrent PCa was high in our study and was comparable to that of other studies that included patients with biochemically recurrent PCa.
30 It is not surprising that no benefit of MDT was seen in terms of HRQOL at month 3 and year 1 because the difference in patients who started with ADT in the first year was comparable (
Fig 2A). The safety of MDT was excellent, with no grade 2 or higher events and only 17% grade 1 toxicity.
In conclusion, MDT for patients with oligorecurrent PCa is safe and improves ADT-free survival when compared with surveillance. We recommend testing MDT in larger phase III studies.