In this study, we assessed the outcome of observation in men who are recommended to have ART offered following RP based on the current ASTRO/AUA guideline . During 50.5 months of median follow-up period, BCR rate was 16.6%. Therefore, more than 80% of the patients at our institution who qualified for ART based on the guideline did not need ART.
To date, three prospective randomized trials have shown that ART after RP for the patients with pT3 and/or PSMs consistently reduced the risk of BCR anywhere from 50 to 60% and improved the outcome of local control [6–8]. More recently, an update of the SWOG 8794 trial showed improved overall survival with ART when compared to observation . Based on these finding, ASTRO and AUA jointly published a guideline recommending that ART be offered to all men who met the inclusion criteria of the three aforementioned studies. Notwithstanding, the SWOG 8794 trial demonstrated no overall survival benefit for the subset of patients with confirmed undetectable PSA post-operatively . That is, patients who underwent SRT after BCR had similar long-term outcome to those who had ART with undetectable PSA level immediately after RP. Therefore, these results suggest that ART in every man with high-risk features post-operatively is not necessary.
To this end, the present study reported the results of observation without ART in patients who met the ASTRO/AUA criteria for ART. An important finding in our study is that the overall BCR after surgery was very low (16.6%). The prior randomized trials showed a BCR rate of 46% to 61.8% in the observation cohort, respectively. This favorable result may be explained in part by the differences in inclusion criteria and baseline characteristics. Our pre-operative mean serum PSA level was signficantly lower when compared to that of previous studies. For example, the cohort of Wiegel et al. had higher pre-operative PSA level (9.4 ng/ml) than our study patients (6.79 ng/ml) . And the frequency of men with high PSA (>10 ng/ml) was only 14 (8.6%) in our study population. In the trial reported by Thompson et al., the rate of PSA > 10 ng/ml was over 40% . Alternatively, the differences in inclusion criteria may be the underlying reason for the observed low rate of BCR in the present cohort. Specifically, our study analyzed patients who had a pathological stage pT2–3 N0 with undetectable PSA level immediately after RP and met one or more of the three following risk factors: ECE, PMSs, or SVI. Two randomized studies (SWOG and EORTC trials) for patients with pT2 (R1) or pT3 (R0 or R1) disease did not require an undetectable PSA level after RP [6, 7]. The third ARO study selected patients with an undetectable PSA level after RP but limited to pT3 tumors .
Several studies had shown that GS, initial PSA level, SVI, and PSMs are independent predictors of biochemical progression [15, 16]. Our present study also showed that BCR rate was significantly associated with initial PSA level and surgical GS. But, there was no significant association with pathologic stage and surgical margin status. Since the BCR group patients had significantly higher pre-operative PSA level and surgical GS, we performed further analysis to identify factors independently correlated with an increased risk of BCR. The multivariate Cox proportional hazard analysis revealed that GS is an independent predictor of BCR. In the subsequent analysis of the favorable group defined as GS ≤ 7 and pre-operative PSA 6.35 ng/mL or less, BCR was only 3.4%. These results suggest that the criteria for ART based on the ASTRO/AUA guideline need to be re-evaluated to avoid significant overtreatment. We recommend that for patients with pre-operative PSA < 6.35, Gleason score <8, and an undetectable PSA immediately after surgery, observation is a reasonable approach.
Lastly, the present study has a significant economic implication. In the U.S., approximately 90 percent of PCa patients will choose definitive treatment , resulting in a projected $12 billion in medical costs in 2010 . At the present time, the most widely used modality for radiation therapy is intensity modulated radiation therapy (IMRT). The calculated reimbursement for IMRT as a primary treatment was $29,356 in patients with low- or intermediate-risk PCa . Other studies reported that the treatment with IMRT costs $15,000–$20,000 more than alternative standard therapies [20, 21]. Showalter et al. reported that the mean incremental cost for ART versus observation was $6,023 per patient . Therefore, the minimum direct economic benefit of observation over ART in the present cohort was $819,128 (136 × $6,023).
Potential limitations of the current study are the retrospective study design, relatively small sample size and a relatively short follow–up period. It should be pointed out though, that all data in this study were recorded prospectively. Another limitation is that our study is a single center single surgeon series. Therefore, the impact of surgical technique and institutional bias cannot be assessed.