Bendifallah Sofiane, Canlorbe Geoffroy, Huguet Florence, Coutant Charles, Hudry Delphine, Graesslin Olivier, Raimond Emilie, Touboul Cyril, Collinet Pierre, Bleu Géraldine, Daraï Emile, Ballester Marcos
Department of Obstetrics and Gynaecology, Tenon University Hospital, CHU Tenon, Assistance Publique des Hôpitaux de Paris (AP-HP), University Pierre and Marie Curie, Paris 6, Institut Universitaire de Cancérologie (IUC), Paris, France,
Ann Surg Oncol. 2014 Dec;21(13):4239-45. doi: 10.1245/s10434-014-3864-6. Epub 2014 Jun 27.
To develop a risk scoring system (RSS) to determine recurrence in women with early-stage type 1 endometrial cancer (EC).
Data of 396 women with early-stage type 1 EC who received primary surgical treatment between January 2001 and December 2012 were abstracted from multicentre database (training set). A risk model for predicting recurrence was developed and internally validated with the bootstrap technique. The RSS was externally validated using data from an independent population.
Overall, the recurrence rate was 12.1 %. The median follow-up and initial time to recurrence were 34 (range 1-152) and 26 (range 1-151) months, respectively. Recurrence was associated with five variables: age ≥60 years, histological grade III, primary tumor diameter >2 cm, depth of myometrial invasion ≥50 %, and the positive lymphovascular space involvement status. These variables were included in the RSS and assigned scores. A total score of 6.5 points corresponded to the optimal threshold of the RSS. For women with a score <6.5 or ≥6.5, the recurrence rates were 8.4 % (30/357) and 48.7 % (19/39) in the training set, respectively. At this threshold, the diagnostic accuracy of the RSS was 87 %. Areas under the curve of the receiver-operating characteristics for predicting recurrence at internal and external validation were 0.74 [95 % confidence interval (CI) 0.71-0.77] and 0.82 (95 % CI 79-85), respectively.
This RSS identified two subsets of women with low and high risk of recurrence among women with early-stage type 1 EC. It could be helpful to better define indications for nodal staging and adjuvant therapy.
开发一种风险评分系统(RSS)以确定早期1型子宫内膜癌(EC)女性的复发情况。
从多中心数据库(训练集)中提取2001年1月至2012年12月期间接受初次手术治疗的396例早期1型EC女性的数据。开发了一个预测复发的风险模型,并使用自助法进行内部验证。使用来自独立人群的数据对RSS进行外部验证。
总体而言,复发率为12.1%。中位随访时间和首次复发时间分别为34(范围1 - 152)个月和26(范围1 - 151)个月。复发与五个变量相关:年龄≥60岁、组织学分级III级、原发肿瘤直径>2 cm、肌层浸润深度≥50%以及淋巴管间隙受累阳性状态。这些变量被纳入RSS并赋予分数。总分6.5分对应于RSS的最佳阈值。在训练集中,得分<6.5或≥6.5的女性复发率分别为8.4%(30/357)和48.7%(19/39)。在此阈值下,RSS的诊断准确性为87%。内部和外部验证时预测复发的受试者工作特征曲线下面积分别为0.74 [95%置信区间(CI)0.71 - 0.77]和0.82(95% CI 79 - 85)。
该RSS在早期1型EC女性中识别出复发风险低和高的两个亚组。这可能有助于更好地确定淋巴结分期和辅助治疗的指征。