目的: 总结肛管直肠黏膜黑色素瘤(anorectal mucosal melanoma, ARMM)的临床病理特征, 分析其预后相关因素。方法: 回顾性分析2010—2018年北京大学肿瘤医院收治的原发性ARMM手术切除标本共68例, 复阅病理切片评估病理特征, 分期采用Slingluff分期法。结果: (1) 临床特征: 患者诊断时中位年龄61.5岁, 男女比例1 ∶1.62, 首发临床表现最常见的是便血(49例); 肿瘤部位最常见的是肛管直肠(66.2%), 其次为直肠(20.6%); 诊断时分期为Ⅰ期28例(局限期, 41.2%), Ⅱ期25例(区域淋巴结转移期, 36.8%), Ⅲ期15例(远处转移期, 22.1%); 5例行局部扩大切除术, 其余为腹会阴联合切除术, 48例术后接受了辅助治疗。(2)病理特征: 大体检查88.2%为隆起型肿块, 肿瘤中位最大径3.5 cm, 中位厚度1.25 cm, 黏膜肌以下浸润深度0~5.00 cm(中位1.00 cm); 肿瘤浸润达固有肌层者27例, 达肠周脂肪组织者16例; 67.6%为无色素性或色素不明显; 细胞形态最常见的是上皮样(45例, 66.2%); 89.7%有溃疡形成, 35.3%可见坏死, 55.9%有血管淋巴管侵犯, 30.9%有神经周围侵犯, 核分裂象计数中位值18/mm²; 免疫组织化学检测S100、HMB-45和Melan-A阳性率分别为92.0%、92.6%、98.0%, Ki-67增殖指数中位值50%;53例进行了CKIT、BRAF及NRAS基因突变检测, 突变率分别为17.0%(9例), 3.8%(2例), 9.4%(5例)。(3)预后: 66例患者获得了随访资料, 中位随访17个月, 中位生存时间17.4个月, 1年、2年、5年的总生存率分别为76.8%、36.8%、17.2%;诊断时淋巴结转移率56.3%;49例(84.5%)发生了远隔部位转移, 最常见的转移部位是肝; 单因素分析显示, 肿瘤最大径(>3.5 cm)、黏膜肌以下浸润深度(>1.0 cm)、坏死、血管淋巴管侵犯、BRAF基因突变、术后未行辅助治疗、肿瘤浸润层次深、诊断时分期晚均为总生存期的预后不良因素; 多因素分析显示, 血管淋巴管侵犯、BRAF基因突变为影响总生存期的独立危险因素, 诊断时分期相关的P值接近临界值。结论: ARMM患者的总体预后差, 血管淋巴管侵犯、BRAF基因突变是独立的预后不良因素; Slingluff分期法可以有效提示预后, 诊断时应详尽评估病理特征, 明确分期, 并尽可能做基因检测; 肿瘤黏膜肌以下的浸润深度可能是比肿瘤厚度更好的预后指标。

Objective: To investigate the clinicopathological characteristics of anorectal mucosal melanoma (ARMM), and to evaluate the prognostic factors. Methods: A total of 68 primary ARMM surgical specimens from 2010 to 2018 were retrospectively studied. Slides were reviewed to evaluate pathological features. Slingluff staging method was used for staging. Results: (1) Clinical features: The median age at diagnosis in this group was 61.5 years, with a male-to-female ratio 1 ∶1.62. The most common complaint was blooding (49 cases). For anatomic site, anorectum was the prevalent (66.2%), followed by rectum (20.6%). At the time of diagnosis, 28 cases were stage Ⅰ (localized stage, 41.2%), 25 cases were stage Ⅱ (regional lymph node metastasis, 36.8%), and 15 cases were stage Ⅲ (distant metastasis, 22.1%). Five patients underwent wide local excision, the rest abdominoperineal resection, and 48 patients received adjuvant therapy after surgery. (2) Pathological features: Grossly 88.2% of the tumors were exophytic polypoid masses, with the median tumor size 3.5 cm and the median tumor thickness 1.25 cm. Depth of invasion below lamina muscularis mucosae ranged from 0-5.00 cm (median 1.00 cm). The deepest site of tumor invasion reached muscular layer in 27 cases, and perirectal tissue in 16 cases. Melanin pigmentation was absent or not obvious in 67.6% of the cases. The predominant cytology was epithelioid (45 cases, 66.2%). The rate for ulceration, necrosis, lymphovascular invasion, and perineural invasion was 89.7%, 35.3%, 55.9%, and 30.9%, respectively. The median mitotic count was 18/mm². The positive rate of S100, HMB-45 and Melan-A were 92.0%, 92.6% and 98.0%, respectively. The median of Ki-67 was 50%. The incidences of mutations within CKIT, BRAF and NRAS genes were 17.0% (9 cases), 3.8% (2 cases) and 9.4% (5 cases), respectively. (3) Prognosis: Survival data were available in 66 patients, with a median follow-up of 17 months and a median survival time of 17.4 months. The 1-year, 2-year and 5-year overall survival rate was 76.8%, 36.8% and 17.2%, respectively. The rate of lymphatic metastasis at diagnosis was 56.3%. Forty-nine patients (84.5%) suffered from distant metastasis, and the most frequent metastatic site was liver. Univariate analysis revealed that tumor size (>3.5 cm), depth of invasion below lamina muscularis mucosae (>1.0 cm), necrosis, lymphovascular invasion, BRAF gene mutation, lack of adjuvant therapy after surgery, deep site of tumor invasion, and high stage at diagnosis were all poor prognostic factors for overall survival. Multivariate model showed that lymphovascular invasion and BRAF gene mutation were independent risk factors for lower overall survival, and high stage at diagnosis showed borderline negative correlation with overall survival. Conclusion: The overall prognosis of ARMM is poor, and lymphovascular invasion and BRAF gene mutation are independent factors of poor prognosis. Slingluff staging suggests prognosis effectively, and detailed assessment of pathological features, clear staging and genetic testing should be carried out when possible. Depth of invasion below lamina muscularis mucosae of the tumor might be a better prognostic indicator than tumor thickness.