Mul Joeri, Melchior Patrick, Seravalli Enrica, Saunders Daniel, Bolle Stephanie, Cameron Alison L, Gurtner Kristin, Harrabi Semi, Lassen-Ramshad Yasmin, Lavan Naomi, Magelssen Henriette, Mandeville Henry, Boterberg Tom, Kroon Petra S, Kotte Alexis N T J, Hoeben Bianca A W, van Rossum Peter S N, van Grotel Martine, Graf Norbert, van den Heuvel-Eibrink Marry M, Rübe Christian, Janssens Geert O
Princess Máxima Center for Pediatric Oncology, Utrecht, the Netherlands.
Department of Radiation Oncology, University Medical Center Utrecht, Utrecht, the Netherlands.
Clin Transl Radiat Oncol. 2021 Mar 11;28:39-47. doi: 10.1016/j.ctro.2021.03.001. eCollection 2021 May.
Recently, the SIOP-RTSG developed a highly-conformal flank target volume definition for children with renal tumors. The aims of this study were to evaluate the inter-clinician delineation variation of this new target volume definition in an international multicenter setting and to explore the necessity of quality assurance.
Six pediatric renal cancer cases were transferred to ten radiation oncologists from seven European countries ('participants'). These participants delineated the pre- and postoperative Gross Tumor Volume (GTV), and Clinical Target Volume (CTV) during two test phases (case 1-2 and 3-4), followed by guideline refinement and a quality assurance phase (case 5-6). Reference target volumes (TV) were established by three experienced radiation oncologists. The Dice Similarity Coefficient between the reference and participants (DSC) was calculated per case. Delineations of case 5-6 were graded by four independent reviewers as 'per protocol' (0-4 mm), 'minor deviation' (5-9 mm) or 'major deviation' (≥10 mm) from the delineation guideline using 18 standardized criteria. Also, a major deviation resulting in underestimation of the CTV was regarded as an unacceptable variation.
A total of 57/60 delineation sets were completed. The median DSC for the CTV was 0.55 without improvement after sequential cases (case 3-4 vs. case 5-6: p = 0.15). For case 5-6, a major deviation was found for 5/18, 12/17, 18/18 and 4/9 collected delineations of the GTV, GTV, CTV-T and CTV-N, respectively. An unacceptable variation from the CTV was found for 7/9 participants for case 5 and 6/9 participants for case 6.
This international multicenter delineation exercise demonstrates that the new consensus for highly-conformal postoperative flank target volume delineation leads to geometrical variation among participants. Moreover, standardized review showed an unacceptable delineation variation in the majority of the participants. These findings strongly suggest the need for additional training and centralized pre-treatment review when this target volume delineation approach is implemented on a larger scale.
最近，国际小儿肿瘤学会肾肿瘤研究组（SIOP-RTSG）为肾肿瘤患儿制定了一种高度适形的侧腹靶区体积定义。本研究的目的是在国际多中心环境中评估这种新靶区体积定义在临床医生之间的勾画差异，并探讨质量保证的必要性。
将6例小儿肾癌病例转给来自7个欧洲国家的10名放射肿瘤学家（“参与者”）。这些参与者在两个测试阶段（病例1-2和3-4）勾画术前和术后大体肿瘤体积（GTV）以及临床靶区体积（CTV），随后进行指南完善和质量保证阶段（病例5-6）。由三名经验丰富的放射肿瘤学家确定参考靶区体积（TV）。计算每个病例中参考靶区体积与参与者勾画的靶区体积之间的骰子相似系数（DSC）。病例5-6的勾画由四名独立评审员根据18项标准化标准，按照与勾画指南的偏差程度分为“符合方案”（0-4毫米）、“轻微偏差”（5-9毫米）或“严重偏差”（≥10毫米）。此外，导致CTV低估的严重偏差被视为不可接受的差异。
共完成了57/60组勾画。CTV的DSC中位数为0.55，连续病例后无改善（病例3-4与病例5-6比较：p = 0.15）。对于病例5-6，分别在收集的GTV、GTV、CTV-T和CTV-N勾画中，发现5/18、12/17、18/18和4/9存在严重偏差。病例5中有7/9的参与者以及病例6中有6/9的参与者的CTV勾画存在不可接受的差异。
这项国际多中心勾画研究表明，高度适形的术后侧腹靶区体积勾画的新共识导致参与者之间存在几何差异。此外，标准化评审显示大多数参与者的勾画差异不可接受。这些发现强烈表明，当这种靶区体积勾画方法大规模实施时，需要额外的培训和集中的治疗前评审。
