Retrospective study on the benefit of adjuvant radiotherapy in men with intraductal carcinoma of prostate

Intraductal carcinoma of the prostate (IDC-P) is a histologic subtype detected in up to 20% of routine radical prostatectomies (RPs) [1, 2]. IDC-P is mostly found alongside high-grade high volume and/or advanced acinar adenocarcinoma, and is independently associated with poor prognosis and shorter biochemical recurrence (BCR)-free survival [3‐6]. Notably, IDC-P incurs a 20-month reduction in overall survival in metastatic castration-resistant prostate cancer (PC) [7]. IDC-P also incurs a 29% reduced disease-specific survival at 15 years in patients with a Gleason score (GS) ≥7 [8]. Due to the clinical relevance of IDC-P, it is now mandatory to indicate its presence regardless of grade [9].

Recent treatment response studies of patient-derived xenografts have demonstrated that IDC-P can withstand androgen deprivation with an associated rapid emergence of castrate-tolerant cells [10]. It also does not respond to treatment in the same fashion as usual PC without IDC-P. In low-grade organ-confined PC, IDC-P has been associated with a high risk of progression [11]. Yet, to our knowledge, no treatment response studies exist for IDC-P in non-metastatic PC, even though it often progresses towards early distant metastases and reduced survival [12].

Adjuvant radiotherapy (ART) is an option for high-risk localized and locally advanced PC [13, 14]. Its advantages mainly outweigh prominent side effects in PC that are prone to high recurrence rates [15]. Currently, guidelines suggest that ART should be considered for PC exhibiting high-risk features (HRF): positive margins (PMs), extraprostatic extension (EPE), seminal vesicle invasion (SVI), and sometimes Grade Groups (GG) 4 & 5 [16]. However, IDC-P to date has not been factored into the therapeutic decision-making process for patients at high risk of recurrence [15, 17‐19]. We therefore explored whether patients with IDC-P respond to ART and compared its impact against currently acknowledged HRF.

IDC-P is a prevalent finding that negatively and independently affects recurrence-free and disease-specific survivals. However, there is a paucity of data about treatment response for non-metastatic PC with IDC-P [7, 10]. In parallel, ART has been observed to improve survival in patients with PC and is a viable option for aggressive and locally advanced disease [24]. In this study, we therefore suggest that against the currently recognized HRFs, for which ART is often considered, IDC-P confers a strong negative impact on BCR risk, an impact which is comparable to the effect of ≥1 HRF, while ART appears to portend a strong protective effect.

Our results are concordant with multiple studies that show a beneficial effect for ART and that also show the detrimental effect of IDC-P on survival, although they were analyzed separately [15, 17‐20, 25, 26]. Additionally, IDC-P’s impact on BCR risk was estimated to be of 2.39 by Murata et al., as well as of 2.98 by Trudel et al. in a different cohort, on par with our results [20, 27].

Interestingly, as IDC-P showed the strongest deleterious effect on BCR compared to other HRFs, patients with this finding might be the most prone to benefit from ART. Actually, many other recent studies also show that IDC-P has a stronger effect on clinical progression than any other factor, including Gleason grade 5 and SVI [28‐30]. These results highlight the need for the validation of the effect of aggressive treatment, notably, radiation therapy, in the presence of IDC-P.

Altogether, these results complement current literature concerning patient selection for ART. This process is focused on balancing ART’s reduction of recurrence rates, with its significant side effects [14]. Current literature suggests that patients with residual cancer at margins seem to benefit the most from ART [14]. In addition, it can also be considered in men with SVI, EPE and GG4–5, as ART seem to improve their recurrence rates [14, 31]. However, our results suggest that the relative importance of these factors is confounded by IDC-P status. This is particularly true as these clinicopathological factors are sometimes inadequate to guide treatment in seemingly low-grade PC which progresses unexpectedly. Additional factors are needed to distinguish those vulnerable men from truly indolent cancer. IDC-P is an excellent candidate, as it is readily available in any standard laboratory, easy to detect by pathologists, and present in 13% of our patients with seemingly low-grade GG1 and GG2 PC lacking HRF. Altogether, we strongly believe that IDC-P deserves a more prominent role in treatment algorithms.

Since PC can be treated using very different and evolving modalities and can experience a wide spectrum of outcomes, we attempted to mitigate the intrinsic limitations of this retrospective study. We limited our study to a specific set of patients treated by RP (with or without ART) while excluding all men who received other treatments before BCR. Even though RP techniques have varied from 1993 until 2015, progression-related outcomes have not varied between RP approaches [32]. Similarly, the technical aspects of ART have remained largely unchanged between centres and throughout time [33]. Similarly, patients selected for pelvic lymphatic pathway irradiation during ART were not identified in the databases. However, it is worth noting that treatment approaches from both centers lead to similar outcomes [34]. Also, since PSA increases following RP can be unpredictable and prone to many confounders, we limited our cohort to patients who had undetectable PSA following RP. This approach would isolate the effect of postoperative ART. Another limitation was that the number of patients studied in each centre was unbalanced, as it was not feasible to review more than an estimated 1600 eligible patients only treated by RP from center 1. This is why we only extracted patients treated by ART from this much larger cohort. Patients from center 2 were not selected, and recruited prospectively. Moreover, the overall study group, though unbalance, showed comparable results in prevalence of IDC-P and other PC pathological factors. The follow-up time for patients with ART was also shorter, as this treatment was used more extensively in recent years in both centers. However, the second set of analyses of IDC-P as a stand-alone indicator for ART showed consistent results, and did not rely on this subgroup of patients. Importantly, we used BCR as an endpoint, despite the fact that the time to PC specific death after BCR is known to be variable [35]. As our study was aimed to explore whether patients with IDC-P respond to ART and to compare the impact of IDC-P against currently recognized HRFs, the use of BCR allowed us to perform those comparisons in a small study sample and to avoid the effect of treatments subsequent to BCR, but overall definitely calls for validation in larger cohorts.

Our data has demonstrated higher biochemical failure in men with IDC-P who were not treated by ART. Additionally, IDC-P conferred the same level of BCR risk as current HRF of PC, which are often used as indicators for ART. If confirmed in other larger cohorts, current treatment algorithms should be modified to include IDC-P as a key factor leading to ART.