目的: 副肿瘤性天疱疮合并实体肿瘤的重症患者通常需要在手术后收住重症医学科(intensive care unit，ICU)治疗，这类患者的长期死亡率较高。本研究对这类患者的临床特点及远期预后的影响因素进行分析。方法: 回顾性分析2005年1月至2020年12月间63例副肿瘤性天疱疮合并实体肿瘤、手术后收住ICU重症患者的临床和实验室资料，并对患者的生存状况进行随访。结果: 63例患者中，原发肿瘤为Castleman病的占79.4%，其他病理类型为20.6%；皮损程度在重度-广泛的占69.8%，其他皮损程度占30.2%；合并闭塞性细支气管炎的占44.4%，不合并的占55.6%。23.8%的患者并发术后真菌感染，无真菌感染占76.2%。术后中位随访时间为95个月，25例患者在研究期间死亡，术后1年、3年、5年生存率分别为74.6%(95%CI 63.8%~85.4%)、67.4%(95%CI 55.6%~79.2%)和55.1%(95%CI 47.9%~62.3%)。通过对分类因素采用Log-rank法进行单因素分析结果表明：年龄＞40岁(P=0.042)、发病后体质量下降＞5 kg(P=0.002)、术前白蛋白 < 30 g/L(P < 0.001)、并发闭塞性细支气管炎(P=0.002)、围术期存在真菌感染(P＜0.001)的患者死亡率增加；Cox单因素分析显示术前体质量下降>5 kg(P=0.005)、术前白蛋白 < 30 g/L(P < 0.001)、术前合并闭塞性细支气管炎(P=0.009)、术前肺部细菌感染(P=0.007)、手术时间长(P=0.048)、术后入ICU时氧合指数(P=0.012)和白蛋白(P=0.010)、血红蛋白浓度低(P=0.035)、入ICU后急性生理学及慢性健康状态评分(acute physiology and chronic health evaluation, APACHE Ⅱ, P=0.001)、序贯器官衰竭评分(sequential organ failure assessment, SOFA, P=0.010)，以及术后真菌感染都是影响远期存活的危险因素(P < 0.001)。Cox回归模型进行多因素分析结果表明，术前体质量下降＞5 kg(HR 4.44; 95%CI 1.47~13.38; P=0.008)、术前白蛋白 < 30 g/L(HR 4.38; 95%CI 1.72~11.12; P=0.002)、术前合并闭塞性细支气管炎(HR 2.69; 95%CI 1.12~6.50; P=0.027)及术后并发真菌感染(HR 4.85; 95%CI 2.01-11.72; P＜0.001)是术后死亡的独立危险因素。结论: 因副肿瘤性天疱疮合并实体肿瘤而接受手术治疗的重症患者术后的5年存活率约为55.1%，术前体质量下降＞5 kg、白蛋白 < 30 g/L、合并闭塞性细支气管炎和术后并发真菌感染是术后近远期死亡风险增加的影响因素。

Objective: Critically ill patients with solid tumors complicated with paraneoplastic pemphigus are usually treated in intensive care units (ICU) for perioperative management after surgical treatment. In this study, the clinical characteristics and predictors of long-term prognosis of these critically ill patients were analyzed. Methods: the clinical and laboratory data of 63 patients with solid tumors complicated with paraneoplastic pemphigus admitted to ICU from 2005 to 2020 were retrospectively analyzed, and the survival status of the patients were followed up. Results: Among the 63 patients, 79.4% had Castleman disease as the primary tumor, and 20.6% with other pathological types; 69.8% had severe-extensive skin lesions, and 30.2% had other skin lesions; the patients with bronchiolitis obliterans accounted for 44.4%, and 55.6% were not merged. Postoperative fungal infection occurred in 23.8% of the patients, and 76.2% without fungal infection. The median follow-up time was 95 months, and 25 patients died during the study period. The 1-year, 3-year and 5-year survival rates were 74.6% (95%CI 63.8%-85.4%), 67.4% (95%CI 55.6%-79.2%) and 55.1% (95%CI 47.9%-62.3%), respectively. The log-rank univariate analysis showed that the patients had age>40 years (P=0.042), preoperative weight loss>5 kg (P=0.002), preoperative albumin < 30 g/L (P < 0.001), paraneoplastic pemphigus complicated with bronchiolitis obliterans (P=0.002), and perioperative fungal infection (P < 0.001) had increased mortality. Cox univariate analysis showed that preoperative weight loss >5 kg (P=0.005), preoperative albumin < 30 g/L (P < 0.001), paraneoplastic pemphigus complicated with bronchiolitis obliterans (P=0.009), preoperative bacterial pulmonary infection (P=0.007), prolonged surgical time (P=0.048), postoperative oxygenation index (P=0.012) and low albumin (P=0.010) and hemoglobin concentration (P=0.035) in ICU, acute physiology and chronic health evaluation (APACHE Ⅱ) score (P=0.001); sequential organ failure assessment (SOFA) score (P=0.010), and postoperative fungal infection (P < 0.001) were risk factors for long-term survival. Cox regression model for multivariate analysis showed that preoperative weight loss > 5 kg (HR 4.44; 95%CI 1.47-13.38; P=0.008), and preoperative albumin < 30 g/L (HR 4.38; 95%CI 1.72-11.12; P=0.002), bronchiolitis obliterans (HR 2.69; 95%CI 1.12-6.50; P=0.027), and postoperative fungal infection (HR 4.85; 95%CI 2.01-11.72; P < 0.001) were independent risk factors for postoperative mortality. Conclusion: The 5-year survival rate of critically ill patients undergoing surgery for paraneoplastic pemphigus combined with solid tumors is approximately 55.1%, with preoperative weight loss > 5 kg, albumin < 30 g/L, bronchiolitis obliterans and postoperative fungal infection were associated with an increased risk of near- and long-term postoperative mortality.