Comparative Review of Consensus-Based Clinical Target Volume Definitions for Prostate Radiotherapy

Background: A major obstacle in the delivery of postoperative radiation therapy (RT) for prostate cancer is accurate delineation of the tumor targets and organs at risk. Although postoperative prostate cancer contouring atlases are quite common, there is still no widely accepted contouring guideline. The purpose of this study is to critically review the various postoperative prostate RT treatment planning consensus guidelines or atlases currently available. Methods: A literature search was conducted using various electronic databases with the key terms: prostate, contour, planning tumour volume, clinical target volume, delineation or definition, guidelines or atlas, and radiation oncology. The search was limited to English publications from the years 1985 to 2011. Results: A total of seven publications relating to contouring guidelines for postoperative prostate radiotherapy were identified. There are four distinct consensus guidelines developed by major institutions: Princess Margaret Hospital, the Australian and New Zealand Radiation Oncology Genito-Urinary Group, the European Organization for Research and Treatment of Cancer, and the Radiation Therapy Oncology Group. Conclusions: After reviewing the consensus contouring guidelines for postoperative prostate cancer radiation therapy that were available in the literature, it is clear that there disagreement with regards to what anatomical borders should be used for delineating an appropriate prostate bed CTV. Additional studies comparing the reproducibility of the various guidelines as well as the performance of these guidelines on clinically important outcomes are needed.


Introduction
Postoperative radiation therapy (RT) is indicated in the treatment of post-prostatectomy patients with highrisk of local recurrence [1]. Recent studies have shown RT to be beneficial following radical prostatectomy in both the adjuvant setting, for patients with high-risk pathological features, such as positive surgical margins and seminal vesicle invasion [2][3][4][5][6][7][8], or as salvage for biochemical disease recurrence [9][10]. Two randomised controlled trials suggest adjuvant RT directly following surgery provides improved progression-free survival, biochemical relapse-free survival, and local control over watchful waiting protocols [3,11]. Although a longterm follow-up of one of these trials also suggests that adjuvant RT also improves metastasis-free survival and overall survival compared with observation alone [2], there is insufficient long-term follow-up data available to adequately assess the effect on these treatment outcomes.
A major obstacle in the delivery of postoperative radiation therapy for prostate cancer is accurate delineation of the tumor targets and Organs at Risk (OARs). Inter-observer variability between oncologist drawn contours in prostate radiation therapy is well documented [12][13][14][15] and has been identified as a highly significant contributing factor to uncertainty in radiation therapy treatment planning [16]. Contouring consensus guidelines or atlases for postoperative prostate RT have been created to aid oncologists in the delineation of tumor targets and OARs with the hope of reducing this variability. clinical target volume, delineation or definition, guidelines or atlas, and radiation oncology. The search was limited to English publications from the years 1985 to 2011. Relevant studies found on references lists of identified articles were included as well as any articles found on national cooperative group websites, including the Radiation Therapy Oncology Group (RTOG) and European Organization for Research and Treatment of Cancer (EORTC). Only studies related to medicine or physics (including radiation oncology, imaging, oncology or cancer) were included. The methodology used in these articles including how each consensus was reached and study validity was reviewed, and any similarities and differences found in the contouring guidelines were detailed.

Literature search
A total of seven publications relating to contouring guidelines for postoperative prostate radiotherapy were identified [17][18][19][20][21][22][23]. There are four distinct consensus guidelines developed by major institutions: Princess Margaret Hospital (PMH) [17], the Australian and New Zealand Radiation Oncology Genito-Urinary Group (FROGG) [19], the European Organization for Research and Treatment of Cancer (EORTC) [21], and the Radiation Therapy Oncology Group (RTOG) [22]. A fifth consensus statement was identified from the Radiotherapy and Androgen Deprivation In Combination After Local Surgery (RADICALS) group in their validation study [18], in which they used a modified version of the PMH consensus guideline. No publication outlining the contents of this version was found. There were five studies discussing the methodology used in the creation of the various consensus statements [17,19,[21][22][23]. The information from the EORTC consensus was published in two separate articles [21,23]. There were two validation publications, one from the RADICALS [18] and the other from EORTC [20]. The study by PMH [17] contains both a description of how their consensus was reached as well as a validation component. Although the article from Boehmer, et al. [23] was on guidelines for radiotherapy for intact prostate, it was referred to in the EORTC consensus for postoperative prostate radiotherapy by Poortman, et al. [21] and was included in this review. The studies were published between 2006 and 2011 and represent a large number of medical institutions from a wide range of countries (

Consensus methodology
The methodology for creating the consensus atlases by the RTOG, FROGG and PMH began with delineation of preliminary contours by a small group of clinicians, involving at least one oncologist or urologist experienced with postoperative prostate radiotherapy or prostatectomy [17,19,22]. In these studies, the clinicians responsible for delineating the initial contours were given patients with specific clinical scenarios and were asked to delineate an appropriate clinical target volume (CTV) based on their expert knowledge of anatomy, tumor physiology and patterns of spread [17,19,22]. Consensus guidelines from the RTOG, FROGG and PMH were finalized following a presentation and discussion of their preliminary contours at a conference or consensus workshop [17,19,22]. Inter-professional collaboration between diverse multidisciplinary groups of health care professionals, specifically individuals with expertise in the treatment of prostate cancer, was utilized in the creation of the consensus atlases [17,19,[21][22][23]. The study by PMH focused on postoperative prostate patients with or without seminal vesicle invasion [17], while the RTOG study focused on the clinical scenarios of seminal vesicle invasion and positive apex margins [22]. A review of relevant literature was present in the development of each of the consensus guidelines [17,19,[21][22][23].
The initial contours generated by PMH were presented at the Australian Faculty of Radiation Oncology Genito-Urinary Oncology Group Consensus Workshop on Post-Prostatectomy Radiotherapy in June 2006 [17,19]. The consensus guidelines created at this conference were further modified by the FROGG and were used in the creation of their own consensus atlas [19].
The RTOG used an imputation method of the expected maximum (EM) algorithms for simultaneous truth and performance level estimation (STAPLE) [24] to create their preliminary contours. The STAPLE algorithm has been previously identified as a useful tool in analyzing expert radiation oncologist consensus contours [25]. The RTOG STAPLE contours were presented at a RTOG conference where they were discussed and a consensus was finalized following a teleconference [22].
There was no information given by the EORTC regarding the exact process in which their consensus guideline was reached. Instead, the EORTC presents detailed manuscripts reviewing published works relevant to prostate cancer, specifically studies on surgery, anatomy and local recurrence [21,23].
A detailed description of the methods involved in the creation of the four contouring guidelines is shown in Table 2.  There were three publications validating consensus guidelines for postoperative prostate radiotherapy found in the literature [17][18]20]. Consensus statements from the EORTC, PMH as well as a modified version of the PMH consensus (RADICALS) were evaluated in these studies [17][18]20]. There were no validation studies found from the RTOG or the FROGG. A summary of the validation studies is presented in Table 3.  The publication from PMH validates their consensus in three separate studies [17].

Consensus How Consensus was Reached
The first validation study assessed the coverage of the consensus Clinical Target Volume (CTV) relative to prostate bed surgical clips placed during routine clinical practice at their institution. There were 25 postoperative prostate patients with a total of 339 surgical clips in the entire patient cohort. A single observer was responsible for delineating an appropriate CTV for each patient using the PMH consensus guidelines, while a second observer was responsible for identifying the surgical clip co-ordinates on CT scans for each patient [17]. To visualize how well the surgical clips were contained within the consensus contours deformable registration and a finite element modeling (FEM) method, (Mofeus), [26] was used to generate a single CTV representing the entire cohort to evaluate its coverage of the surgical clips [17]. Results showed all surgical clips were contained within the consensus CTV, minus one outlier, and the distribution throughout the CTV was uniformly observed, with a lower clip density appreciated at the anterior-superior and posterior-most aspects [17].
The second validation looked at the intra-and inter-observer reproducibility of the consensus contours. Two observers individually contoured 16 patients on MRI and CT scans and repeated this at least one week later [17]. The reproducibility of the contours was assessed using the finite element modeling (FEM) method, Morfeus, statistical tool used previously in their first validation study [26]. The variation in observer contours were said to be "mainly random" with small systematic inter-observer errors observed in the Anterior-Posterior (AP) dimension [17].
The third PMH validation study involved a retrospective analysis of the effects of the consensus guidelines on clinical practice. Twenty patients previously treated with various techniques of postoperative radiotherapy (including one or two phase four-field box or Intensity Modulated Radiation Therapy) were replanned using the PMH consensus guidelines. The newly created plans were compared both volumetrically and dosimetrically to the original plans used for treatment. An increase in both the contour volumes and field sizes was observed in the consensus plans and only 50% of the patients would have received the prescribed microscopic dose (V 100 >95%) to the CTV in these plans [17].
The validation study by Ost, et al. assessed whether adopting the EORTC consensus guidelines would reduce inter-observer contouring variability at their institution. Ost, et al. compared the contours of 10 patients, delineated by six observers (five radiation oncologists and one radiologist) using the EORTC guidelines [20]. The variability was assessed using an open-source Matlab-based radiation therapy planning analysis statistical tool [27] for agreement (volume of overlap corrected using kappa statistics) [28]. The specificity and sensitivity of the observer contours with and without the contouring guidelines was evaluated using the STAPLE method [24]. Variation in dimensions of observer contours in the superior, inferior, anterior, posterior, and lateral margins were assed by comparing the distances of the outer margins of each CTV to a pre-determined centre point [20]. The inferior margin variability was assed by comparing the inferior border distance to the penile bulb delineated on MRI. Ost, et al. described the overall inter-observer agreement for the CTVs using the EORTC guidelines as moderate (mean kappa, 0.49), and recommended a refinement of the EORTC consensus atlas [20].
The modified PMH consensus guidelines assessed by Mitchell, et al. on behalf of the RADICALS were not defined in their study or found in the literature. Mitchell, et al. found that using the RADICALS consensus guidelines led to an increase in the volume of the CTV and there was a reduction in inter-observer variability [18].

Target volume definitions
The consensus information regarding the margins adopted by each institution is displayed in Table  4. Consensus statements from PMH, FROGG and the RTOG focused on border landmarks in their consensus [17,19,22], while the EORTC included areas at greatest risk for relapse in their CTV guidelines [21,23].

Superior Margin:
The anatomical landmarks used as the superior boundary in the consensus statements were the level of the cut end of the vas deferens [17,19,22] or the bladder neck up to the base of the seminal vesicles [21]. The superiorly located prostate bed surgical clips were included in the superior margin by PMH and FROGG [17,19]. PMH recommends a 5mm margin above the vas deferens as an alternative superior margin in the absence of surgical clips or when the vas deferens is poorly visualized [17]. If the vas deferens retracts posteriorly, the RTOG suggests using a superior margin 3-4 cm above the top of the pubic symphysis [22]. The EORTC recommends the bladder neck up to the base of the seminal vesicles be included in the prostate bed CTV [21].

Inferior Margin:
There were two methods used to delineate the inferior border in the consensus guidelines, involving either a designated distance below the vesicourethral anastomosis (VUA) [17,19,22], or slightly above the penile bulb or at the apex [21]. The measurements chosen for the distance below the VUA ranged from 5-6 mm [19] to 8-12mm [22]. The penile bulb was chosen as an alternative landmark for defining the inferior border by PMH, FROGG and the RTOG in the event that the VUA is poorly visualized [17,19,22].
Anterior Margin: The anterior margin chosen by PMH, FROGG and the RTOG was the posterior edge of the pubic symphysis (caudally) and the posterior bladder wall (cranially) [17,19,22]. The anterior margin from the EORTC includes the anastomosis and the urethral axis [21].

Additional consensus information
Information regarding the imaging technology, patient positioning, vesicourethral anastomosis (VUA) visualization method, organs at risk (OARs), and additional target volume recommendations from the consensus guidelines are displayed in Table 5. 2013   Multi-slice CT scans were used in the creation of the consensus guidelines by PMH and the RTOG [17,22]. The use of MRI with or without Positron Emission Tomography (PET) fusion was recommended by PMH and FROGG for delineating treatment plan contours [17,19]. The importance of effectively defining the vesicourethral anastomosis was identified as an important step in the delineation of postoperative prostate radiotherapy contours in the publications from PMH, RTOG and FROGG [17,19,22]. The FROGG and EORTC consensus statements contained recommendations regarding contouring of the OARs [21,23]. There were additional CTV and Planning Target Volume (PTV) recommendations from PMH, FROGG and the EORTC [17,19,21,23].

Conclusions
After reviewing the consensus contouring guidelines for postoperative prostate cancer radiation therapy that were available in the literature, it is clear that there is some disagreement with regards to what anatomical borders should be used for delineating an appropriate prostate bed CTV. There were many similarities in the methodology used in the creation of the atlases, although some atlases were more detailed than others. The validation studies performed using the consensus atlases were done on a small number of patients and tended to focus on inter-observer reproducibility of the newly adopted guidelines relative to a previous standard. The results of these studies seemed inconclusive as to whether or not the consensus guidelines improved inter-observer reproducibility in the creation of target volumes or any other clinically important outcomes. Additional studies comparing the reproducibility of the various guidelines as well as the performance of these guidelines on clinically important outcomes when used in a variety of clinical sites are needed so that physicians can better decide which atlas, or part of an atlas, to adopt into their practice.

Additional Information Disclosures
Human subjects: All authors have confirmed that this study did not involve human participants or tissue. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.