Overall survival is certainly the outcome of greatest importance for any cancer therapy, incorporating the effect of mortality secondary to cancer, the interventions used, and all other causes. Given the relatively indolent natural history of prostate cancer, the expectation is that lengthy follow-up is necessary to assess differences in overall survival, and the results of this review bear this out. Neither the Southwest Oncology Group (SWOG) nor the European Organization for the Research and Treatment of Cancer (EORTC) trial detected a survival benefit with adjuvant RT; median follow-up times were 10.6 years and five years in each trial, respectively. A meta-analysis of the survival data from these two trials also did not demonstrate a statistically significant result. Follow-up at this time is simply not long enough to accurately determine if adjuvant radiotherapy (RT) is associated with a survival benefit. Updates of this review and meta-analysis are planned as the data mature and as new trial results become available. It should also be borne in mind that for neither trial was overall survival the primary endpoint; as such, neither trial was specifically powered to detect a difference in overall survival between the two arms. This is of particular relevance to the SWOG trial, in which only 431 patients were randomized.
Biochemical progression is a controversial surrogate marker for other prostate cancer outcomes. The meta-analysis performed of these data unequivocally demonstrates that, compared to observation, adjuvant RT confers a major reduction in the rate of biochemical failure. The magnitude of benefit in this endpoint is remarkably similar across the three included trials. As prostate-specific antigen (PSA) progression is often a trigger for initiation of androgen deprivation therapy (ADT), it is not surprising that a reduction in ADT use of similar magnitude was also observed.
While time-to-event analyses for freedom from locoregional recurrence are not available and the sole analysis for freedom from metastasis did not demonstrate a significant benefit from adjuvant RT, two of the trials report an outcome that is a composite of locoregional failure and metastasis - clinical progression-free survival. Both trials reported a significant improvement in this outcome with adjuvant RT. On the basis of these data, adjuvant RT does therefore significantly reduce locoregional and distant recurrences when considered together.
A major shortcoming of the trials included in this review relates to the management of patients in the observation arms of these trials. In short, none of the trials employed a definite protocol as to how and when PSA and clinical failures should be treated. As a consequence, there was considerable variability in patient management in these arms (see Table 1 in the original guideline document). It may be argued that, in many cases, local intervention with RT was delayed until such time as it was unlikely to be effective. The ongoing RADICALS (Radiotherapy and Androgen Deprivation in Combination After Local Surgery) trial addresses this shortcoming. Once completed, it is hoped that this trial will serve to clarify the optimal timing of RT and the role for ADT following radical prostatectomy (RP) in patients with high-risk pathologic features.
Comparing toxicity data across the three trials is made difficult by differences in the reporting of toxicity data. In the SWOG trial, significantly greater rates of urethral stricture, urinary incontinence, and rectal complications were seen in irradiated patients compared to those randomized to observation. Caution must be used, however, in interpreting the toxicity data from the SWOG study. As noted above, toxicity was not recorded using a validated, graded toxicity-scoring instrument. Instead, complications were recorded only if annotated on study flow sheets. Such data are vulnerable to the bias that retrospectively collected unsolicited toxicities are more likely to be reported in the intervention arm. In the EORTC trial, where toxicity was graded prospectively using validated scales, it is clear that there is significantly greater minor (< grade 2) acute toxicity in patients who receive adjuvant RT. However, there was no significant excess grade 3 or higher toxicity observed at five years of follow-up. As the late toxicity evaluations performed in the trials only considered genitourinary and gastrointestinal symptoms, the potential benefits of adjuvant RT in terms of sparing the systemic toxicity of ADT could not be assessed.