Use of Postprostatectomy Radiation Therapy at an NCI-Designated Comprehensive Cancer Center (2024)

Methods

We queried our prospective prostate cancer database for patients who received RT to the prostate bed after prostatectomy from 2003 to 2011. Patients receiving a prescription dose ranging from 60 to 68 Gy were selected for study inclusion, and those with metastatic disease were excluded.

RT was defined as adjuvant if patients started RT within 1 year of surgery,¹⁰ and if they were referred based on a high-risk factor identified on pathologic assessment of the surgical specimen (T3 disease [extracapsular extension or seminal vesicle invasion], a positive margin, or Gleason score 8–10). A detectable prostate-specific antigen (PSA) level was allowed in the adjuvant definition as long as only one postoperative PSA result was obtained in the setting of the preceding high-risk factors and was less than 0.2 ng/mL; patients with a detectable PSA level less than 0.2 ng/mL are not accounted for in the American Urological Association (AUA) guidelines.¹¹ RT was defined as salvage if given in the setting of an increasing PSA level, a single postoperative PSA level of 0.2 ng/mL or greater, or a documented clinical recurrence via imaging or digital rectal examination. To assess for a potential change in referral patterns of patients receiving salvage RT, a PSA threshold of 0.5 ng/mL or less was used to determine which patients were being referred for early salvage.⁹

The number of patients with an intact prostate treated with definitive RT was recorded by year as a control for changing volume among patients with total prostate in the department. Additionally, the billing records of Fox Chase Cancer Center (FCCC) were queried for the number of prostatectomies performed at the institution in a given year. Because the prostatectomy billing records were available only between 2003 and 2011, we limited our RT cohort to the same period. Clinical and pathologic data were not available for patients undergoing prostatectomy alone and did not receive PPRT.

Chi-square analysis was used to assess differences between patient cohorts receiving adjuvant and salvage RT. Spearman correlation was used to assess yearly trends in PSA level at time of referral for RT. Linear regression models were used to test the trend of number of PPRT cases, number of prostatectomies, and number of patients with an intact prostate receiving radiation across years respectively. Because of the limited number of years for which complete data were available, only univariate analyses were performed.

Results

From 2003 to 2011, 475 patients received PPRT at FCCC (83 adjuvant and 392 salvage). The patient characteristics and descriptives are listed in Table 1. Patients were more likely to receive adjuvant RT if they were younger (P=.0013), had a positive surgical margin (P<.0001), had a higher pathologic Gleason score (P=.0071), and had a higher pathologic T stage (P<.0001). The postprostatectomy PSA level was also significantly lower (P<.0001), as was expected given the definition of adjuvant RT applied.

During the period of this study, an increasing proportion of patients treated with RT for localized prostate cancer at FCCC were treated in the postprostatectomy setting. The proportion of patients with localized prostate cancer treated with PPRT steadily increased from 9.4% in 2003 to a high of 26.6% in 2011 (Figure 1). During that time, a significant increase was seen in the number of prostatectomies performed at FCCC (P=.02), with a corresponding increase in PPRT use. The increase in the number of patients treated with RT to an intact prostate remained relatively stable over the same period (Figure 2). Regression model analysis found a significant correlation between year of treatment and number of PPRT cases (P=.002), and between number of surgical cases and number of PPRT cases (P=.02). No significant correlation was seen between the number of patients with an intact prostate treated with radiation and the number of patients treated with PPRT (P=.709).

The proportion of patients receiving adjuvant RT did not increase compared with those receiving salvage RT (P=.5; Figure 3). Pre-RT PSA values were inversely correlated with year of treatment (P=.005). A trend was seen toward an increasing proportion of patients referred with a pathologic Gleason score of 8 or more (P=.1; Figure 4). When considering only patients receiving salvage RT with a PSA level of 0.5 ng/mL or less, a significant increase was seen in the number of patients referred for early salvage RT versus those referred with higher PSA levels (P=.0018; Figure 4). Examination of the proportion of patients with high-risk pathologic features (positive margin and/or ≥pT3 disease) revealed no difference in annual proportion receiving adjuvant or salvage treatment over time (P=.58).

Discussion

This review of the institutional practice pattern of PPRT has demonstrated an increased use of RT in the postprostatectomy setting for men with localized prostate cancer. PPRT now accounts for approximately 25% of the patients with localized prostate cancer being treated with definitive RT at FCCC. This increase has occurred despite the number of patients remaining stable who have an intact prostate cancer treated with definitive external-beam RT at FCCC. Potential explanations for the change in patient distribution include the increased number of prostatectomies performed at our institution and in the region,¹² or the possible increased awareness among urologists of the benefit of referring patients for PPRT given the findings of recent randomized studies.^5,6,13

The increase in prostatectomies at FCCC is likely related to the introduction of the robotic prostatectomy in 2006, with more than 97% of the prostatectomies performed robotically since, and the addition of 2 urologic oncologists between 2007 and 2011. An increased number of prostatectomies is also consistent with what has been shown nationally,¹⁴ which has been attributed to the popular use of robot-assisted prostatectomies.² The increase in patients who have undergone PPRT logically corresponds with the escalation in surgical volume. Although we did not have clinical surgical data available for this analysis, we did observe a trend toward

a greater number of patients being referred for PPRT with a Gleason score 8 or more. Because referrals for adjuvant PPRT did not increase, this trend likely reflects an increased number of high-risk patients undergoing prostatectomies.

Adjuvant PPRT has been shown to reduce the risk of biochemical recurrence, local recurrence, and clinical progression of prostate cancer compared with radical prostatectomy alone for men with high-risk

pathologic features after prostatectomy (positive resection margin, extracapsular extension, or seminal vesicle invasion) in 3 randomized trials.¹⁵ The first of these trials (EORTC 22911¹³) was published in 2005 and demonstrated an improvement in biochemical progression-free survival that has remained significant with 10-year follow-up.⁷ Similarly, ARO 96-02 also demonstrated a biochemical progression-free survival.⁵ Although SWOG 8794 was not controlled for PSA level at the time of study entry, long-term results published in 2009 were the first to exhibit an improvement in the rate of distant metastases and an overall survival benefit associated with adjuvant PPRT.¹⁶ At FCCC, no significant increase was seen in adjuvant PPRT over the course of the study period. This could be attributed to the controversy that exists regarding the appropriate timing of PPRT and the reservations regarding the interpretation of these 3 randomized trials.^8,17

Because of concerns that adjuvant radiation requires delivering radiation to patients who may not require additional treatment after prostatectomy, salvage RT is a strategy that uses radiation only after the patient’s PSA level increases. A retrospective analysis by Trock et al¹⁸ demonstrated that RT within 2 years of biochemical recurrence improved prostate cancer–specific survival if the PSA doubling time was less than 6 months. An analysis of more than 1500 patients by Stephenson et al¹⁹ found that approximately 50% of patients who received salvage RT alone with a PSA level of 0.5 ng/mL or less were free from progression at 6 years compared with only 26% for those with higher PSAs. Given the current lack of randomized evidence comparing adjuvant and salvage PPRT, Briganti et al⁹ performed a multi-institutional match-controlled analysis comparing adjuvant RT with observation followed by early salvage RT (postoperative PSA level ≤0.5 ng/mL at least 6 months after surgery) as needed. Early salvage RT had an equivalent 5-year biochemical control compared with adjuvant RT for patients with pT3pN0 disease, and resulted in only 45% of patients in the observation/early salvage RT group requiring treatment, thus minimizing potential overtreatment. The role of ultrasensitive PSA in the postprostatectomy setting, which has been shown to be capable of predicting which patients are more or less likely to experience recurrence,^20,21 remains in question given the variability at such low levels. This question of adjuvant RT versus early salvage RT will hopefully be answered by the 2 ongoing randomized clinical trials: the Medical Research Council/National Cancer Institute of Canada Intergroup RADICALS (Radiotherapy and Androgen Deprivation in Combination after Local Surgery) trial²² and the Trans Tasman Radiation Oncology Group RAVES (Radiotherapy - Adjuvant Versus Early Salvage) trial (ClinicalTrials. gov identifier: NCT00860652).

Taking into account the data currently available, current NCCN Guidelines for Prostate Cancer recommend that, “evidence supports offering adjuvant/salvage RT in most men with adverse pathologic features or detectable PSA and no evidence of disseminated disease”¹⁰ (to view the most recent version of these guidelines, visit NCCN.org). Given our data showing a stable percentage of patients undergoing adjuvant RT and a significant increase in the number of patients being treated with a PSA level of 0.5 ng/mL or less, the institutional bias seems to be favoring a strategy of observation/early salvage RT over adjuvant RT for most men with high-risk pathologic features.

In a recent survey of urologists and radiation oncologists, practitioners from both specialties reported that large randomized clinical trials were the level of evidence most likely to change treatment recommendations for localized prostate cancer.²³ A SEER analysis by Hoffman et al²⁴ in 2011 found that younger age, Gleason score of 8 or greater, positive margin, and pT3a and pT3b were associated with a higher likelihood of receiving PPRT, consistent with the results of our study. From 2000 to 2007, they did not observe an increase in the use of PPRT in relation to the presentation of the EORTC 22911 and SWOG 8794 results in 2004 and 2005, respectively. Similarly, in a SEER-Medicare analysis of patients treated from 2000 to 2006, Sheets et al²⁵ found little increase in the rate of PPRT since the publication of the randomized trials. In our institutional analysis, we were able to follow patients to a more recent time point (2011), which includes the 2009 publication of a third randomized trial (ARO 96-02) and the overall survival benefit seen in the SWOG trial.^5,16 Although we did not show a significant increase in the rate of adjuvant RT, we did find that PPRT now comprises a larger proportion of patients with localized prostate cancer treated with radiation.

Limitations of this study include those related to the fact that it was a single-institution analysis.

FCCC is located in a region of dense health care supply, with many options for patients to choose from regarding where to receive care, including 2 NCI-designated comprehensive cancer centers within 15 miles of each other. We know that not all patients receiving PPRT also had surgery at FCCC. It is also likely that not all surgical patients remained at our institution if they required PPRT. This precludes us from being able to precisely analyze the trend of whether men are more or less likely to receive PPRT after prostatectomy. Another limitation of this study

is the lack of clinical data for the patients undergoing prostatectomy (specifically pathologic data) that would help to better define changes in the pattern of referrals for PPRT. That information would also help determine the percentage of patients undergoing prostatectomy that met the criteria from the randomized trials for adjuvant RT versus the number of patients actually referred for adjuvant RT.

Conclusions

For radiation oncologists treating prostate cancer at FCCC, a significant trend in the composition of patients being treated with radiation is seen, from greater than 90% of patients being treated with an intact prostate in the early 2000s to more recently approximately 25% of patients being treated after a prostatectomy. This change seems to be largely driven by increased surgical volume, rather than by a data-driven increase in referrals for adjuvant RT based on the results of the phase III trials supporting adjuvant radiation.^5,7,16 At FCCC, referral of patients for salvage RT has trended toward an increase in early salvage in patients with PSA levels less than 0.5 ng/mL, rather than immediate referral for adjuvant RT. Further studies should validate these findings in a larger data set, such as a national oncology database, and conduct comparative effectiveness research to evaluate the utilization trends among available treatment modalities for localized prostate cancer.