ESTRO ACROP guideline on prostate bed delineation for postoperative radiotherapy in prostate cancer

Abstract

Purpose/objective: Radiotherapy to the prostate bed is a potentially curative salvage option after radical prostatectomy. Although prostate bed contouring guidelines are available in the literature, important variabilities exist. The objective of this work is to provide a contemporary consensus guideline for prostate bed delineation for postoperative radiotherapy.

Methods: An ESTRO-ACROP contouring consensus panel consisting of 11 radiation oncologists and one radiologist, all with known subspecialty expertise in prostate cancer, was established. Participants were asked to delineate the prostate bed clinical target volumes (CTVs) in 3 separate clinically relevant scenarios: adjuvant radiation, salvage radiation with PSA progression, and salvage radiation with persistently elevated PSA. These cases focused on the presence of positive surgical margin, extracapsular extension, and seminal vesicles involvement. None of the cases had radiographic evidence of local recurrence on imaging. A single computed tomography (CT) dataset was shared via FALCON platform and contours were performed using EduCaseTM software. Contours were analyzed qualitatively using heatmaps which provided a visual assessment of controversial regions and quantitatively analyzed using Sorensen-Dice similarity coefficients. Participants also answered case-specific questionnaires addressing detailed recommendations on target delineation. Discussions via electronic mails and videoconferences for final editing and consensus were performed.

Results: The mean CTV for the adjuvant case was 76 cc (SD = 26.6), salvage radiation with PSA progression was 51.80 cc (SD = 22.7), and salvage radiation with persistently elevated PSA 57.63 cc (SD = 25.2). Compared to the median, the mean Sorensen-Dice similarity coefficient for the adjuvant case was 0.60 (SD 0.10), salvage radiation with PSA progression was 0.58 (SD = 0.12), and salvage radiation with persistently elevated PSA 0.60 (SD = 0.11). A heatmap for each clinical scenario was generated. The group agreed to proceed with a uniform recommendation for all cases, independent of the radiotherapy timing. Several controversial areas of the prostate bed CTV were identified based on both heatmaps and questionnaires. This formed the basis for discussions via videoconferences where the panel achieved consensus on the prostate bed CTV to be used as a novel guideline for postoperative prostate cancer radiotherapy.

Conclusion: Variability was observed in a group formed by experienced genitourinary radiation oncologists and a radiologist. A single contemporary ESTRO-ACROP consensus guideline was developed to address areas of dissonance and improve consistency in prostate bed delineation, independent of the indication.There is important variability in existing contouring guidelines for postoperative prostate bed (PB) radiotherapy (RT) after radical prostatectomy. This work aimed at providing a contemporary consensus guideline for PB delineation. An ESTRO ACROP consensus panel including radiation oncologists and a radiologist, all with known subspecialty expertise in prostate cancer, delineated the PB CTV in 3 scenarios: adjuvant RT, salvage RT with PSA progression, and salvage RT with persistently elevated PSA. None of the cases had evidence of local recurrence. Contours were analysed qualitatively using heatmaps for visual assessment of controversial regions and quantitatively using Sorensen-Dice coefficient. Case-specific questionnaires were also discussed via e-mails and videoconferences for consensus. Several controversial areas of the PB CTV were identified based on both heatmaps and questionnaires. This formed the basis for discussions via videoconferences. Finally, a contemporary ESTRO-ACROP consensus guideline was developed to address areas of dissonance and improve consistency in PB delineation, independent of the indication.

Keywords: Adjuvant radiotherapy; Postoperative radiotherapy; Prostate cancer; Prostate cancer guidelines; Salvage radiotherapy; Target volume delineation.

Fig. 1

Heatmap with areas of greatest variability in Postoperative Prostate Bed Clinical Target Delineation.

Fig. 2

Inferior Border. a. CT-based delineation of the inferior border of the CTV; b. MRI-based delineation of the inferior border of the CTV. Abbreviation: CTV = clinical target volumes; CT = computed tomography; MRI = magnetic resonance imaging; VUA = vesico-urethral anastomosis.

Fig. 3

Anterior Border. a. Cranially, CT-based delineation of the CTV covering 1–2 cm of the posterior bladder wall; b. Cranially, CT-based delineation of the CTV up to the posterior bladder wall; c. MRI-based delineation depicting the anterior border of the CTV at the posterior margin of the pubic bone up to half to two thirds of the symphysis pubis. Abbreviation: CT = computed tomography; CTV = clinical target volumes; MRI = magnetic resonance imaging.

Fig. 4

Posterior Border. a. Cranially, CT-based delineation of the CTV up to the anterior wall of the rectum; b. Caudally, CT-based delineation of the CTV up to the anterior wall of the rectum including the existing surgical clips and the antero-lateral angles of the rectum. Arrow shows antero-lateral angles of the rectum. Abbreviation: CT = computed tomography; CTV = clinical target volumes; MRI = magnetic resonance imaging.

Fig. 5

Lateral Border. a. Cranially, CT-based delineation of the CTV up to the internal margins of the internal obturator muscles; b. Caudally, CT-based delineation of the CTV up to the internal margins of the internal obturator muscles. Abbreviation: CT = computed tomography; CTV = clinical target volumes.

Fig. 6

Superior Border. a. Cranially, CT-based delineation of the CTV superiorly on axial view; b. CT-based delineation of the CTV on sagittal view. c. MRI-based delineation of the CTV on sagittal view. Abbreviation: CT = computed tomography; MRI = magnetic resonance image. Abbreviation: CT = computed tomography; CTV = clinical target volumes; MRI = magnetic resonance imaging.