Variation in the Use of Postoperative Radiotherapy Among High-risk Patients Following Radical Prostatectomy

TM Morgan; SR Hawken; KR Ghani; DC Miller; FY Feng; SM Linsell; JA Salisz; Y Gao; JE Montie; ML Cher


Prostate Cancer Prostatic Dis. 2016;19(2):216-221. 

In This Article


Using the data from diverse urology practices, we investigated the use of ART and SRT for patients at high risk of local recurrence post RP. Younger patients and those with higher tumor stage and grade were more likely to receive ART, and margin status also appeared to be a driver of ART decisions in patients with pT3 disease. Across the entire high-risk cohort, the adjusted probability of ART administration was relatively low (8.18%) and highly variable across practices. Furthermore, SRT administration among patients with a post-prostatectomy PSA ≥0.1 ng ml−1 also varied significantly across practices and was closely related to a given practice's tendency to utilize any form of postoperative radiation. Notably, many patients reached PSA levels ≥1.0 ng ml−1 without undergoing salvage treatment.

In the context of earlier studies reporting low rates of postoperative radiotherapy administration for patients with adverse pathologic features, ART rates in Michigan are not surprising.[6,12–14] However, a key limitation of these population-based investigations is the inability to clearly differentiate ART from SRT due to a lack of post-prostatectomy PSA data and granular follow-up information. In contrast, our study provides unique insight into the patient characteristics and practice patterns that appear to drive ART and SRT use in high-risk patients.

The wide variation in ART and SRT utilization across the diverse MUSIC practices highlights the lack of consensus on this issue.[14–16] Given the mixed and often conflicting data from prior phase 3 trials of ART,[1–3,17–19] physicians appear to have developed highly divergent practice patterns surrounding its administration. Specifically, these data imply that a patient's likelihood of receiving postoperative radiotherapy is highly dependent on the practice where they receive care and that urologists have not coalesced around a given treatment approach. The positive correlation between ART and SRT use by practice suggests that practices tending not to administer ART also infrequently utilize SRT. Further work is needed to better understand the factors driving this variation in care. In addition, two ongoing phase 3 trials (RADICALS and RAVES) comparing ART with early SRT may help guide decisions and bring more uniformity to the postoperative care of this population.[20,21]

Finally, although there is ongoing debate regarding the relative effectiveness of ART versus early SRT, current guidelines are clear that postoperative radiation should be delivered before a PSA of 0.5 ng ml−1.[4,5,22–24] We found that 25/88 patients (28.4%) that reached a PSA of ≥0.2 ng ml−1 received SRT before progressing to a PSA of ≥0.5 ng ml−1, and an additional 17 patients underwent even earlier SRT. However, 38 patients (43.2%) have progressed to a PSA ≥0.5 ng ml−1 and 20 (22.7%) to a PSA ≥1.0 ng ml−1 without receiving prior SRT. These data suggest an opportunity for quality improvement through earlier utilization of SRT.

Our analysis has several limitations. First, with a 15-month median follow-up, it is possible that SRT rates may increase over time. However, even within this follow-up period, many patients with a rising PSA post RP reached PSA levels >0.5 ng ml−1 without receiving SRT, indicating that early SRT is utilized variably. Second, a small number of patients underwent RP at participating MUSIC practices but were followed after surgery at a non-participating practice, and were excluded. This was necessary to ensure accurate ascertainment of post-RP radiotherapy administration; however, it is unlikely that these patients differed in any systematic way from patients that were followed at one of MUSIC's diverse practices. Third, there are some unmeasured factors, most notably patient preferences factoring in functional status and quality of life, and these undoubtedly affect doctor–patient shared decision-making regarding post-RP radiotherapy. Fourth, MUSIC encompasses a specific geographical region, and it is possible that practice patterns could differ elsewhere. However, given the diversity of practices within Michigan and the inclusion of the majority of practices in the state, these findings are likely to be more widely applicable. Last, our validation of radiotherapy administration was limited to patients with BCBSM health insurance, the major payer for non-Medicare beneficiaries in Michigan. Nevertheless, the data from the MUSIC registry demonstrated excellent concordance with claims.

These limitations notwithstanding, our findings have implications for patients and providers. For patients, the fact that some men received late SRT, and others no salvage therapy at all, suggests opportunities to expand ART and early SRT use for patients with high-risk cancer. For providers, the wide variation across practices demonstrates that there are significant differences in perceptions regarding the relative risks and benefits of post-prostatectomy radiotherapy. A lack of strong evidence for improved prostate cancer-specific survival with ART versus early SRT likely contributes to this finding.[25] Moreover, there may be other factors, such as efforts to minimize morbidity, which differentially influence treatment decisions across the state.

Moving forward, there is a need to better understand why some patients receive ART or early SRT and why others receive neither in the setting of biochemical recurrence.[26] Currently, SRT represents the only potentially curative treatment for patients with a post-RP PSA recurrence, and a number of studies have demonstrated that the effectiveness of SRT is inversely correlated with the PSA level at radiotherapy initiation. Thus, the substantial percentage of patients with early PSA recurrences not receiving SRT suggests an opportunity for quality improvement and increased cure rates. Relevant factors impacting radiotherapy use may include providers' interpretation of the evidence, interpretation of the data surrounding the impact of radiation on functional outcomes and thresholds for treatment, as well as the influence of patient preferences. A better understanding of these factors and the opportunity to engage and learn from practices throughout the statewide collaborative may allow MUSIC to implement targeted efforts to improve the care of men who experience a PSA rise following RP.

Taken as a whole, these data may have implications for developing consensus criteria for identifying patients that would most benefit from ART or early SRT. Given the current level of evidence favoring ART and early SRT use in the appropriate setting, these quality improvement efforts are likely to yield significant gains for this high-risk population. In addition, recently published data suggest a potential role for genomic classifiers in guiding these treatment decisions.[27,28] With improved risk stratification, either through molecular markers or via comparable strategies, it may be possible to diminish the variation in care and improve cancer outcomes without a significant adverse impact on long-term sexual and urinary function.