Abstract
目的
建立并验证乙肝肝硬化患者行脾切除术后门静脉血栓形成的列线图预测模型,评价其临床应用价值。
方法
回顾性收集重庆医科大学附属第二医院2014年1月~2020年1月期间收治的共180例乙肝肝硬化并行脾切除术患者的临床资料,本研究中患者行脾切除术后均常规予以低分子肝素或口服抗凝剂抗凝治疗,必要时加用双嘧达莫抗血小板治疗。通过随机数字表法按7:3比例分为建模人群(120例)和验证人群(60例),并比较两人群的临床资料有无可比性。在建模人群中根据术后1月内有无门静脉血栓形成分为血栓组(49例)与无血栓组(71例),通过单因素及多因素Logistic回归分析筛选出乙肝肝硬化患者行脾切除术后门静脉血栓形成的独立危险因素,将这些因素纳入并建立列线图预测模型。分别在建模人群及验证人群中对预测模型进行内部及外部验证:采用AUROC(C指数)验证模型区分度;GiViTI校准带及Hosmer-Lemeshow检验验证模型校准度;DCA曲线验证模型临床有效性。
结果
建模人群与验证人群间临床资料差异均无统计学意义(P > 0.05),具有可比性。单因素及多因素Logistic回归分析显示:血栓组与无血栓组在消化道出血史、门静脉直径、脾静脉直径、脾脏体积、静脉曲张程度、术后D-二聚体、血小板变化值等方面差异有显著统计学意义(P < 0.05),其中门静脉直径、脾静脉直径、术后血小板变化值是乙肝肝硬化脾切除术后门静脉血栓形成的独立危险因素(P < 0.05),将上述独立危险因素纳入并成功建立个体化列线图预测模型。分别在建模人群及验证人群中对列线图预测模型进行内部及外部验证:AUROC(C指数)分别为0.880(95%CI:0.818~0.942)和0.873(95%CI:0.785~0.960),预测模型区分度良好;GiViTI校准曲线带的80%及95%CI区域均未穿过45°角平分线,P值分别为0.965和0.632,Hosmer-Lemeshow检验中P值分别为0.624和0.911,预测模型的校准度较高;DCA曲线中阈概率值设定为30.5%,两人群临床净获益分别为30%和34%,表明预测模型具有临床有效性。
结论
通过建立乙肝肝硬化脾切除术后门静脉血栓形成的列线图预测模型,有助于临床早期筛选并识别乙肝肝硬化脾切除术后门静脉血栓形成的高危患者。
Keywords: 乙型肝炎后肝硬化, 门静脉血栓, 脾切除术, 列线图
Abstract
Objective
To construct and validate an individualized nomogram to predict the probability of occurrence of portal vein thrombosis (PVT) after splenectomy in patients with hepatitis B cirrhosis.
Methods
We retrospectively collected the clinical data from 180 patients with hepatitis B cirrhosis undergoing splenectomy with postoperative anticoagulation therapy during the period from January, 2014 to January, 2020 in our hospital. The patients were randomized into modeling group (n= 120) and validation group (n=60), and the former group was further divided into PVT group (n=49) and non-PVT group (n=71) according to the occurrence of PVT occurred within 1 month after splenectomy. The independent risk factors of PVT after splenectomy were screened in the modeling group using univariate and multivariate binary logistic regression analyses and were used for construction of the nomogram prediction model. The area under the receiver-operating characteristic (AUROC) curve (C-index), GiViTI calibration belt and Hosmer-Lemeshow test, and the DCA curve were used to estimate the discrimination power, calibration and clinical efficiency of the prediction model in both the model construction group and validation group.
Results
Univariate and multivariate logistic regression analyses showed that a history of hemorrhage, portal vein diameter, spleen vein diameter, spleen volume, varicose, postoperative platelet change, and postoperative D-dimer differed significantly between PVT group and non-PVT group (P < 0.05), and portal vein diameter, spleen vein diameter, and postoperative platelet change were independent risk factors of PVT after splenectomy (P < 0.05). The prediction model had a good discrimination power with AUROC (C-index) of 0.880 (95% CI: 0.818-0.942) in the modeling group and 0.873 (95% CI: 0.785-0.960) in the validation group. The 80% and 95%CI region of GiViTI calibration belt did not cover the 45-degree diagonal bisector line (P=0.965 and 0.632, respectively), and the P-values of the Hosmer-Lemeshow test were 0.624 and 0.911, respectively, suggesting a high reliability of the predicted probability by the model. DCA curve analysis showed a threshold probability of 30.5%, with a net benefit of 30% in the modeling group and 34% in the validation group, indicating a good clinical efficiency of the model.
Conclusions
The model for predicting the risk of PVT after splenectomy in patients with hepatitis B cirrhosis can help in early identification of patients having high risks of PVT.
Keywords: hepatitis B cirrhosis, portal vein thrombosis, splenectomy, nomograms
在我国乙型病毒性肝炎是导致肝硬化的重要原因[1],当肝硬化失代偿期出现的消化道出血、脾功能亢进、肝功能衰竭、肝性脑病等一系列并发症严重威胁到患者的生命[2-3]。脾切除术是外科治疗肝硬化所导致的脾功能亢进、消化道出血等的重要手段之一[4]。但脾切除术后门静脉血栓形成作为其常见且严重的并发症之一,时刻威胁着患者的生命[5-6]。目前对于肝硬化脾切除术后门静脉血栓形成的高危风险因素研究较多[5, 7],但尚未达成共识,也缺乏相关临床指南作为指导。列线图是根据多因素回归分析某疾病的高危因素绘制,通过计算总分的形式可直观展示出患者出现相应疾病的概率[8]。易于临床快速高效的筛选高危患者,并采取及时有效的干预措施。因此,本研究旨在建立乙肝肝硬化脾切除术后门静脉血栓形成风险的列线图预测模型,早期筛查并识别出高危风险患者,为临床早期采取干预措施,进行有效治疗提供有效依据。
1. 资料和方法
1.1. 研究对象
回顾性分析我院于2014年1月~2020年1月收治的共180例乙肝肝硬化并行脾切除术患者的临床资料,通过随机数字表法按7:3的比例将其分为建模人群(n= 120)和验证人群(n=60)。在建模人群中根据术后1月内有无门静脉血栓形成分为血栓组和无血栓组。本研究经过医院伦理委员会批准,符合医学伦理学规定。
1.2. 纳入及排除标准
纳入标准:临床明确肝硬化病因为乙型病毒性肝炎,并有明确门静脉高压;脾切除术手术指证明确(白细胞 < 3.0×109/L和/或血小板计数 < 3.0×109/L;Ⅱ级脾大合并上消化道出血史或脾功能亢进明显;达到Ⅲ级或以上脾肿大)[9];术前经门静脉彩超或腹部CT检查未发现门静脉系统血栓;术后1月内接受门静脉彩超或腹部CT检查了解有无门静脉血栓形成,门静脉血栓诊断明确者符合典型影像学表现。
排除标准:其他病因导致的肝硬化(如酒精性、脂肪性、自身免疫性、其他病毒性肝炎等);合并恶性肿瘤、血液系统或自身免疫系统疾病;合并严重的全身性疾病(如心脑血管疾病、呼吸系统疾病等);临床资料不完整。
1.3. 抗凝方案
对研究中脾切除术患者均采取预防性抗凝:术后3 d内无明显出血倾向,常规予以皮下注射依诺肝素钠0.4 mL、1~2次/d抗凝治疗7~10 d;若患者有血栓形成时治疗周期适当延长,院外则更换为口服抗凝剂(如华法林1.25~2.5 mg、1次/d或利伐沙班10~20 mg、1~2次/d)抗凝治疗;当血小板超过3.0×109/L时予以口服双嘧达莫25~50 mg、3次/d抗血小板治疗,降至正常后停用;治疗期间密切监测引流、血常规、肝肾功、凝血象等情况评估出血风险。
1.4. 评估指标
根据国内外相关研究文献,收集并筛选以下指标:患者一般情况(性别、年龄、BMI、是否合并高血压、是否合并糖尿病、消化道出血史);肝硬化及门静脉高压特征(门静脉直径、脾静脉直径、脾脏体积、Child-Pugh分级、静脉曲张程度、是否合并腹水、HBsAg值);手术情况(手术方式、手术时间、术中失血量、术中是否输血、术中是否断流);术后辅助检查[D-二聚体、血小板变化值(术后1周内血小板计数的算数均值较术前血小板计数的变化值)]。
1.5. 统计学方法
统计学分析采SPSS软件(Version 26.0, USA)、R软件(Version 3.6.2, USA)。计量资料采用均数±标准差或中位数(四分位数间距)表示;计数资料采用频数(百分比)表示。计量资料分析采用t检验或Mann-Whitney U检验,计数资料分析采用χ2检验或Fisher精确检验。对建模人群进行单因素及多因素Logistic回归分析筛选出PVT的独立危险因素,纳入并建立列线图预测模型。分别在两人群中对预测模型进行区分度、校准度和临床有效性的验证。区分度的验证采用临床上常用的受试者工作特征曲线下面积(AUROC)及C指数进行评价。校准度的验证采用GiViTI校准带及HosmerLemeshow检验进行评价。临床有效性的验证采用ROC曲线中cutoff值与DCA曲线结合进行评价。P < 0.05表示差异有统计学意义。
2. 结果
2.1. 研究对象的一般资料
共180例患者纳入此项研究,其中建模人群120例,验证人群60例(表 1)。在建模人群中脾切除术后门静脉血栓患者49例,发生率为40.8%。对建模人群及验证人群的一般情况、肝硬化及门静脉高压特征、手术情况、术后辅助检查等资料进行比较(P > 0.05),差异无统计学意义,两组间具有可比性。
1.
建模人群与验证人群的基线特征资料
Baseline characteristics of the patients in the modeling group and validation group
| Characteristics | Modeling group (n=120) | Validation group (n=60) | t/Z/χ2 | P |
| Gender [n (%)] | 0.193 | 0.661 | ||
| Male | 78 (65.0) | 37 (61.7) | ||
| Female | 42 (35.0) | 23 (38.3) | ||
| Age (year) | 45.97±10.02 | 46.90±9.49 | 0.599 | 0.550 |
| BMI [n (%)] | 0.388 | 0.824 | ||
| < 18.5 kg/m2 | 9 (7.5) | 6 (10.0) | ||
| 18.5-24 kg/m2 | 76 (63.3) | 38 (63.3) | ||
| >24 kg/m2 | 35 (29.2) | 16 (26.7) | ||
| Hypertension [n (%)] | 0.000 | 1.000 | ||
| No | 113 (94.2) | 56 (93.3) | ||
| Yes | 7 (5.8) | 4 (6.7) | ||
| Diabetes [n (%)] | 0.865 | 0.352 | ||
| No | 106 (88.3) | 50 (83.3) | ||
| Yes | 14 (11.7) | 10 (16.7) | ||
| History of hemorrhage [n (%)] | 0.046 | 0.830 | ||
| No | 72 (60.0) | 35 (58.3) | ||
| Yes | 48 (40.0) | 25 (41.7) | ||
| Diameter of portal vein (mm) | 14.98±2.39 | 15.33±2.31 | 0.946 | 0.346 |
| Volume of spleen (cm2) | 144 (120-180) | 142 (115-180) | 0.442 | 0.658 |
| Diameter of spleen vein (mm) | 11.75 (10.00-13.00) | 12.00 (9.75-13.25) | 0.035 | 0.972 |
| Child-Pugh [n (%)] | 0.022 | 0.881 | ||
| A | 103 (85.8) | 51 (85.0) | ||
| B | 17 (14.2) | 9 (15.0) | ||
| Varicose [n (%)] | 0.760 | 0.859 | ||
| No | 12 (10.0) | 6 (10.0) | ||
| Mild | 16 (13.3) | 6 (10.0) | ||
| Moderate | 30 (25.0) | 18 (30.0) | ||
| Severe | 62 (51.7) | 30 (50.0) | ||
| Ascites [n (%)] | 0.058 | 0.810 | ||
| No | 88 (73.7) | 45 (75.0) | ||
| Yes | 32 (26.7) | 15 (25.0) | ||
| HBsAg (U/mL) | 250.00 (182.07-364.60) | 250.00 (98.58-250.00) | 1.578 | 0.114 |
| Surgery [n (%)] | 0.012 | 0.914 | ||
| Laparoscope | 47 (39.2) | 23 (38.3) | ||
| Laparotomy | 73 (60.8) | 37 (61.7) | ||
| Devascularization [n (%)] | 0.257 | 0.612 | ||
| No | 28 (23.3) | 12 (20.0) | ||
| Yes | 92 (76.7) | 48 (80.0) | ||
| Blood transfusion [n (%)] | 1.467 | 0.226 | ||
| No | 81 (67.5) | 35 (58.3) | ||
| Yes | 39 (32.5) | 25 (41.7) | ||
| Blood loss (mL) | 300 (200-500) | 350 (200-800) | 1.536 | 0.125 |
| Operation time (min) | 200 (170-245) | 218 (178-250) | 1.139 | 0.255 |
| D-Dimer (ng/mL) | 1.60 (0.90-2.55) | 1.55 (0.85-2.50) | 0.158 | 0.875 |
| PLT addition (109/L) | 126.00 (54.50-277.00) | 102.5 (54.50-264.50) | 0.337 | 0.736 |
2.2. 单因素及多因素Logistic回归分析
在建模人群中单因素Logistic回归分析显示差异有统计学意义的危险因素分别是:消化道出血史、门静脉直径、脾静脉直径、脾脏体积、静脉曲张程度、术后D-二聚体、血小板变化值(P < 0.05);差异无统计学意义的因素为:性别、年龄、BMI、合并高血压、合并糖尿病、消化道出血史、Child-Pugh分级、合并腹水、HBsAg值、手术方式、手术时间、术中失血量、术中输血、术中行断流(P > 0.05,表 2)。将单因素回归分析中有显著统计学意义的危险因素纳入多因素Logistic回归,分析结果显示门静脉血栓独立危险因素为:门静脉直径、脾静脉直径、术后血小板变化值(P < 0.05,表 3)。在Logistic多因素回归分析的基础上,可得到预测模型方程Logit(P)=-11.674+0.429门静脉直径(mm)+0.295脾静脉直径(mm)+0.007术后血小板变化值(109/L)。
2.
建模人群中的单因素Logistic回归分析
Results of univariate logistic regression models in the modeling group
| Characteristics | PVT group (n=49) | Non-PVT group (n=71) | t/Z/χ2 | P |
| *P < 0.05, **P < 0.001: PVT group vs non-PVT group. PVT: Portal vein thrombosis. | ||||
| Gender [n (%)] | 1.232 | 0.268 | ||
| Male | 29 (59.2) | 49 (69.0) | ||
| Female | 20 (40.8) | 22 (31.0) | ||
| Age (year) | 47.67±10.52 | 44.79±9.56 | 1.560 | 0.124 |
| BMI [n (%)] | 2.348 | 0.391 | ||
| < 18.5 kg/m2 | 3 (6.1) | 6 (8.5) | ||
| 18.5-24 kg/m2 | 35 (71.4) | 41 (57.7) | ||
| >24 kg/m2 | 11 (22.4) | 24 (33.8) | ||
| Hypertension [n (%)] | 0.000 | 0.911 | ||
| No | 46 (93.9) | 67 (94.4) | ||
| Yes | 3 (6.1) | 4 (5.6) | ||
| Diabetes [n (%)] | 0.172 | 0.679 | ||
| No | 44 (89.8) | 62 (87.3) | ||
| Yes | 5 (10.2) | 9 (12.7) | ||
| History of hemorrhage [n (%)] | 7.870 | 0.006* | ||
| No | 22 (44.9) | 50 (70.4) | ||
| Yes | 27 (55.1) | 21 (29.6) | ||
| Diameter of portal vein (mm) | 17.00 (15.00-18.00) | 14.00 (12.00-15.00) | 5.538 | < 0.001** |
| Volume of spleen (cm2) | 160.00 (131.50-220.00) | 130.00 (108.00-150.00) | 3.885 | < 0.001** |
| Diameter of spleen vein (mm) | 13.00 (11.10-15.00) | 10.00 (9.00-12.00) | 5.207 | < 0.001** |
| Child-Pugh [n (%)] | 2.455 | 0.127 | ||
| A | 45 (91.8) | 58 (81.7) | ||
| B | 4 (8.2) | 13 (18.3) | ||
| Varicose [n (%)] | 4.654 | 0.046* | ||
| No | 2 (4.1) | 10 (14.1) | ||
| Mild | 6 (12.2) | 10 (14.1) | ||
| Moderate | 11 (22.4) | 19 (26.8) | ||
| Severe | 30 (61.2) | 32 (45.1) | ||
| Ascites [n (%)] | 0.201 | 0.654 | ||
| No | 37 (75.5) | 51 (71.8) | ||
| Yes | 12 (24.5) | 20 (28.2) | ||
| HBsAg (U/mL) | 250.00 (94.26-250.00) | 250.00 (250.00-2152.00) | 1.817 | 0.263 |
| Surgery [n (%)] | 0.206 | 0.650 | ||
| Laparoscope | 18 (36.7) | 29 (40.8) | ||
| Laparotomy | 31 (63.3) | 42 (59.2) | ||
| Devascularization [n (%)] | 1.142 | 0.288 | ||
| No | 9 (18.4) | 19 (26.8) | ||
| Yes | 40 (81.6) | 52 (73.2) | ||
| Blood transfusion [n (%)] | 0.677 | 0.411 | ||
| No | 31 (63.3) | 50 (70.4) | ||
| Yes | 18 (36.7) | 21 (29.6) | ||
| Blood lose (mL) | 300.00 (200.00-700.00) | 200.00 (200.00-500.00) | 1.818 | 0.129 |
| Operation time (min) | 215.00 (187.50-250.00) | 190.00 (160.00-235.00) | 2.127 | 0.232 |
| D-Dimer (ng/mL) | 2.10 (1.25-2.70) | 1.20 (0.70-2.10) | 3.099 | 0.011* |
| PLT addition (109/L) | 251.00 (136.00-364.00) | 75.00 (29.00-178.00) | 5.486 | < 0.001** |
3.
建模人群中的多因素Logistic回归分析
Result of multivariate logistic regression analysis in the modeling group
| Variable | B | SE | Wald | OR | 95%CI | P |
| SE: Standard error; OR: Odds ratio; CI: Confidence interval. | ||||||
| Diameter of portal vein (mm) | 0.429 | 0.149 | 8.286 | 1.536 | 1.147-2.057 | 0.004 |
| Diameter of spleen vein (mm) | 0.295 | 0.121 | 5.990 | 1.344 | 1.061-1.703 | 0.014 |
| PLT addition (109/L) | 0.007 | 0.002 | 15.634 | 1.007 | 1.003-1.010 | < 0.001 |
| Constant | -11.674 | 2.424 | 23.187 | 0.000 | < 0.001 | |
2.3. 预测模型的建立
将上述3个独立的危险因素纳入预测模型后建立乙肝肝硬化脾切除术后门静脉血栓形成风险的个体化列线图预测模型(图 1)。列线图的应用如下:根据列线图,我们可以得到每个预测指标对应的分数值,这些分数值之和被记录为总分,与总分相对应的预测概率是乙肝肝硬化脾切除术后门静脉血栓形成的风险。
1.

乙肝肝硬化脾切除术后门静脉血栓形成风险的个体化列线图预测模型
Nomogram to predict the probability of PVT after splenectomy in patients with hepatitis B cirrhosis.
2.4. 预测模型的验证
2.4.1. 区分度
通过绘制两人群ROC曲线(图 2),建模人群AUROC为0.880(95%CI:0.818~0.942),Cutoff值为30.5%,P < 0.001;C指数为0.880。验证人群AUROC为0.873(95% CI:0.785~0.960),P < 0.001;C指数为0.873。预测模型在两人群中的C指数均 > 0.75。
2.

预测模型在建模人群和验证人群中的ROC
ROC analysis of the prediction model in the modeling group and validation group.
2.4.2. 校准度
预测模型在建模人群及验证人群的GiViTI校准曲线带的80%及95%置信区间区域均未穿过45°角平分线(P > 0.05,图 3),差异无统计学意义。对预测模型在建模人,进行Hosmer-Lemeshow检验:χ2= 7.125,P=0.624;在验证人群中χ2=4.009,P=0.911;预测模型在两人群中的Hosmer-Lemeshow检验P值均 > 0.05,差异无统计学意义。
3.

预测模型在建模人群及验证人群中的校准度
Calibration plot of nomogram in the modeling group and validation group.
2.4.3. 临床有效性
预测模型在建模人群及验证人群DCA曲线(图 4)中表现出:当阈概率分别在9%~90%及2%~99%范围内,患者净获益都比另外两条极端曲线高(横的那条表示假定所有患者都无门静脉血栓形成,均不治疗,净获益为0;斜的那条表示所有患者都形成门静脉血栓,均接受治疗,净获益为一条斜率为负值的反斜线),在此范围内模型具有临床有效性。通过建模人群的ROC分析得到的Cutoff值(30.5%)同时在上述两DCA曲线阈概率范围内,因此模型具有临床有效性。进一步分析可知:当设置30.5%作为诊断门静脉血栓形成并采取干预措施的阈概率值;在建模人群及验证人群中,每100例运用此模型诊断脾切除术后门静脉血栓形成风险的患者中,分别有30人和34人能从中获益而不损害其他人的利益。
4.

预测模型在建模人群及验证人群中的决策曲线分析
Decision curve analysis of nomogram in the modeling group and validation group.
2.5. 预测模型的可视化应用
以某乙肝肝硬化患者为例:该患者脾静脉直径13 mm,门径直径17 mm,行脾切除术后血小板升高150×109/L,根据列线图预测模型(图 5)预测得到该患者术后门静脉血栓形成风险为62.2% > 30.5%,此时根据DCA曲线,我们应当采取有效的干预措施,降低患者术后门静脉血栓形成的风险。
5.

列线图预测模型个体化风险预测范例
Example of prediction nomogram for risk of PVT after splenectomy in patients with hepatitis B cirrhosis.
3. 讨论
脾切除术作为外科处理脾功能亢进、食管胃底静脉曲张出血等严重并发症的重要术式之一,因其止血效果明显,术后再出血率低,脾亢及肝功能可得到改善而广泛开展[10]。但脾切除术后的一系列并发症也令人担忧,其中门静脉血栓形成便是发生率高、难以预测、严重影响患者预后的并发症之一[7, 11]。门静脉高压患者在自然状态下门静脉血栓的发生率为0.6%~2.1%[11],行脾切除术后升至18.9%~57.0%[12]。本研究中肝硬化脾切除术后门静脉血栓的总体发生率约43%,与目前文献报道的发生率相近。发生门静脉血栓的患者大部分在早期常无明显临床症状及体征,有时仅表现为轻微腹痛,易被忽视[13]。因此,如何能够早期预测肝硬化脾切除术后门静脉血栓风险,并对高危人群及时采取有效干预措施非常重要。
目前对肝硬化脾切除术后门静脉血栓形成的相关危险因素研究中,报道较多的危险因素包括:脾静脉直径、门静脉直径、脾静脉流速、门静脉流速、脾脏体积、术后血小板计数、血浆D-二聚体、术后血小板平均体积等,尚未达成共识[14-17];在本研究中,通过单因素及多因素Logistic回归分析发现乙肝肝硬化脾切除患者术后出现门静脉血栓的独立危险因素为:门静脉直径、脾静脉直径、术后血小板变化值,而消化道出血史、脾脏体积、术后D-二聚体、性别、年龄、BMI、高血压、糖尿病、消化道出血史、Child-Pugh分级、静脉曲张程度、合并腹水、HBsAg值、术式、是否断流、是否输血、失血量、手术时间并非其独立危险因素。
脾切除术后门静脉血栓形成的三要素包括:(1)血管内皮损伤:术前持续门静脉高压反复刺激血管平滑肌增生,引起静脉内膜损伤以及粥样硬化样改变,内皮细胞的脱落、胶原暴露导致血小板黏附,进而导致血栓形成[18]。此外,脾切除术本身会离断或缝扎脾门周围血管,不可避免地造成血管内皮损伤,若术中联合门-奇断流术,也会造成血管内皮损伤,导致胶原纤维暴露激活内源性凝血途径,而术后的血管内局部炎症刺激损伤血管,诱导血小板聚集和血栓形成[19],因此术中应仔细操作,减少血管内皮损伤面积,避免其他血管损伤;(2)血流动力学改变:门静脉及脾静脉直径则反映了血流动力学的变化,本研究发现门静脉直径及脾静脉直径与门静脉血栓形成呈正相关,其OR值分别为1.536和1.344,提示门静脉及脾静脉直径越大,脾切除术后门静脉血栓形成的风险也越大。由于门静脉及脾静脉直径越大,脾切除术后静脉血管内的血流量及流速减少越明显,血液越易形成涡流,血液中的血小板、纤维蛋白原等与血管内壁的接触时间增加,凝血因子的浓度也在一定程度上增加,最终触发血栓形成[20]。有研究表明肝硬化门静脉高压患者门静脉及脾静脉增宽越明显,越易造成静脉血管壁长期受到管腔内高压力作用导致内皮细胞损伤;同时造成血流速度过缓导致血流中的促凝物质清除缓慢,从而促进门静脉血栓形成[21-23],这与本研究结果相符。因此,对临床医生而言,围手术期需警惕门静脉及脾静脉直径显著增加患者,及时评估其术后门静脉血栓风险并采取预防措施;(3)凝血机制改变:术后血小板变化值反映了凝血机制变化,本研究通过多因素回归分析发现术后血小板变化值与门静脉血栓形成有关,其OR值为1.007,笔者认为术后血小板的变化值或许是门静脉血栓形成的条件之一。肝硬化患者造血功能通常无明显下降,但受到脾功能亢进的影响,血小板破坏增加,行脾切除术后脾脏的破坏减少,大部分患者的血小板数量短期内急剧升高,使得血小板的聚集能力增加,血液呈高凝状态,形成血栓的风险也相应增加,血小板数量的激增与门静脉血栓形成有一定相关性[12, 18, 24-25]。因此我们需高度警惕术后血小板数量急剧上升的患者,根据预测模型评估门静脉血栓风险,及时采取预防措施。但临床中我们也发现,部分患者虽术后血小板急剧升高,却无门静脉血栓形成,部分研究则认为血小板功能的相关指标(P选择素,即血小板膜蛋白CD62P等)在脾切除术后门静脉血栓患者中表达程度高,与门静脉血栓相关[26],笔者认为P选择素作为血小板活化程度的指标,虽然在肝硬化患者中本身表达程度已很高,若术后血小板数量未明显升高,血小板聚集能力也会受一定影响,因此,术后血小板变化值可能是门静脉血栓形成的条件之一,对此还需开展相关临床研究来进一步验证两者的相关性。
目前对于脾切除术后门静脉血栓的预防原则之一包括:排除出血风险后,均应早期予以抗凝治疗,而在临床工作中,抗凝是把双刃剑,在预防血栓形成的同时又伴随着出血甚至死亡的风险,哪些患者需真正的早期临床干预值得我们深思,同时临床上更缺乏对乙肝肝硬化脾切除术后门静脉血栓形成的列线图预测模型的相关研究,用于筛选真正需早期干预的患者。本研究通过建立乙肝肝硬化脾切除术后门静脉血栓形成的列线图预测模型,根据患者术前及术后的影像学检查和实验室检查预测其术后门静脉血栓风险概率,可指导个体化治疗方案。在我们的研究中将ROC曲线的Cutoff值30.5%作为DCA曲线的阈值时,我们可以看到患者的临床净获益率均高于不采取干预以及全部采取干预的两种极端方式,这也提示我们:当模型预测出患者门静脉血栓风险高于30.5%后,若立即采取干预措施,会使患者从临床中获益,低于30.5%时可暂不采取干预措施,继续观察门静脉血栓的动态变化。这有利于帮助乙肝肝硬化脾切除术后患者的临床决策方案的制定。
对模型的补充:列线图预测模型中的独立危险因素均为连续型计量变量,在临床上的简洁性不够,后期可考虑通过对该类连续型计量数据变量进行最优尺度回归法进行分组,变换为分类变量,这样更便于列线图预测模型的临床应用;后期若能够纳入多个中心的患者数据,增大样本量及观察变量范围,同时进行外部验证,能够进一步强化模型。
综上所述,乙肝肝硬化脾切除术后门静脉血栓形成是严重的并发症之一,如不及时治疗,会对患者预后造成严重影响,本研究发现乙肝肝硬化脾切除术后门静脉血栓形成的独立危险因素为门静脉直径、脾静脉直径、术后血小板变化值,并构建了个体化列线图预测模型,使得预测模型具有可视化,易于临床操作;预测模型无论在建模人群还是验证人群中都具有良好的区别度、校准度及临床有效性,提高了对乙肝肝硬化脾切除术后门静脉血栓高危患者的早期预测和识别能力。
Biography
徐伟,硕士,电话:023-62887533,E-mail: 496527575@qq.com
Funding Statement
重庆市科卫联合医学科研项目(2018MSXM132)
Contributor Information
徐 伟 (Wei XU), Email: 496527575@qq.com.
涂 兵 (Bing TU), Email: 300403@hospital.cqmu.edu.cn.
References
- 1.中华医学会肝病学分会 肝硬化诊治指南. 中华肝脏病杂志. 2019;11(27):846–65. doi: 10.3760/cma.j.issn.1007-3418.2019.11.008. [中华医学会肝病学分会.肝硬化诊治指南[J].中华肝脏病杂志, 2019, 11(27): 846-65.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bandali MF, Mirakhur A, Lee EW, et al. Portal hypertension: Imaging of portosystemic collateral pathways and associated imageguided therapy. World J Gastroenterol. 2017;23(10):1735–46. doi: 10.3748/wjg.v23.i10.1735. [Bandali MF, Mirakhur A, Lee EW, et al. Portal hypertension: Imaging of portosystemic collateral pathways and associated imageguided therapy[J]. World J Gastroenterol, 2017, 23(10): 1735-46.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.张 丽娟. 肝硬化门静脉高压症外科手术后门静脉血栓形成的危险因素分析. 实用临床医药杂志. 2016;20(17):71–3, 77. doi: 10.7619/jcmp.201617022. [张丽娟.肝硬化门静脉高压症外科手术后门静脉血栓形成的危险因素分析[J].实用临床医药杂志, 2016, 20(17): 71-3, 77.] [DOI] [Google Scholar]
- 4.Maruyama H, Okugawa H, Takahashi M, et al. De novo portal vein thrombosis in virus-related cirrhosis: predictive factors and longterm outcomes. Am J Gastroenterol. 2013;108(4):568–74. doi: 10.1038/ajg.2012.452. [Maruyama H, Okugawa H, Takahashi M, et al. De novo portal vein thrombosis in virus-related cirrhosis: predictive factors and longterm outcomes[J].Am J Gastroenterol, 2013, 108(4): 568-74.] [DOI] [PubMed] [Google Scholar]
- 5.Yoshida M, Watanabe Y, Horiuchi A, et al. Portal and splenic venous thrombosis after splenectomy in patients with hypersplenism. Hepatogastroenterology. 2009;56(90):538–41. doi: 10.1109/MWS.2009.22. [Yoshida M, Watanabe Y, Horiuchi A, et al. Portal and splenic venous thrombosis after splenectomy in patients with hypersplenism[J]. Hepatogastroenterology, 2009, 56(90): 538-41.] [DOI] [PubMed] [Google Scholar]
- 6.蒋 国庆, 柏 斗胜, 钱 建军, et al. 华法林防治改良腹腔镜脾切除联合贲门周围血管断流术后门静脉血栓的临床疗效. 中华消化外科杂志. 2016;15(1):71–4. [蒋国庆, 柏斗胜, 钱建军, 等.华法林防治改良腹腔镜脾切除联合贲门周围血管断流术后门静脉血栓的临床疗效[J].中华消化外科杂志, 2016, 15(1): 71-4.] [Google Scholar]
- 7.牟 思玉, 杨 哲, 吴 力群, et al. 肝硬化门静脉高压脾切除术后门静脉血栓形成的预测因素分析. 临床肝胆病杂志. 2018;34(1):106–11. doi: 10.3969/j.issn.1001-5256.2018.01.022. [牟思玉, 杨哲, 吴力群, 等.肝硬化门静脉高压脾切除术后门静脉血栓形成的预测因素分析[J].临床肝胆病杂志, 2018, 34(1): 106-11.] [DOI] [Google Scholar]
- 8.Balachandran VP, Gonen M, Smith JJ, et al. Nomograms in oncology: more than meets the eye. Lancet Oncol. 2015;16(4):e173–80. doi: 10.1016/S1470-2045(14)71116-7. [Balachandran VP, Gonen M, Smith JJ, et al. Nomograms in oncology: more than meets the eye[J]. Lancet Oncol, 2015, 16(4): e173-80.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.李 涛, 杨 镇. 精准医学时代门静脉高压症新断流术的进展. 临床外科杂志. 2017;25(1):30–1, 32. doi: 10.3969/j.issn.1005-6483.2017.01.007. [李涛, 杨镇.精准医学时代门静脉高压症新断流术的进展[J].临床外科杂志, 2017, 25(1): 30-1, 32.] [DOI] [Google Scholar]
- 10.Li MX, Zhang XF, Liu ZW, et al. Risk factors and clinical characteristics of portal vein thrombosis after splenectomy in patients with liver cirrhosis. HBPD INT. 2013;12(5):512–9. doi: 10.1016/S1499-3872(13)60081-8. [Li MX, Zhang XF, Liu ZW, et al. Risk factors and clinical characteristics of portal vein thrombosis after splenectomy in patients with liver cirrhosis[J]. HBPD INT, 2013, 12(5): 512-9.] [DOI] [PubMed] [Google Scholar]
- 11.Fujiyama S, Saitoh S, Kawamura Y, et al. Portal vein thrombosis in liver cirrhosis: incidence, management, and outcome. BMC Gastroenterol. 2017;17(1):112. doi: 10.1186/s12876-017-0668-8. [Fujiyama S, Saitoh S, Kawamura Y, et al. Portal vein thrombosis in liver cirrhosis: incidence, management, and outcome[J]. BMC Gastroenterol, 2017, 17(1): 112.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.He SS, He FP. Predictive model of portal venous system thrombosis in cirrhotic portal hypertensive patients after splenectomy. Int J Clin Exp Med. 2015;8(3):4236–42. [He SS, He FP. Predictive model of portal venous system thrombosis in cirrhotic portal hypertensive patients after splenectomy[J]. Int J Clin Exp Med, 2015, 8(3): 4236-42.] [PMC free article] [PubMed] [Google Scholar]
- 13.李 玲. 从Virchow三要素分析肝硬化患者发生门静脉血栓危险因素. 实用肝脏病杂志. 2015;18(3):325–8. doi: 10.3969/j.issn.1672-5069.2015.02.032. [李玲.从Virchow三要素分析肝硬化患者发生门静脉血栓危险因素[J].实用肝脏病杂志, 2015, 18(3): 325-8.] [DOI] [Google Scholar]
- 14.Amitrano L, Guardascione MA, Brancaccio V, et al. Risk factors and clinical presentation of portal vein thrombosis in patients with liver cirrhosis. J Hepatol. 2004;40(5):736–41. doi: 10.1016/j.jhep.2004.01.001. [Amitrano L, Guardascione MA, Brancaccio V, et al. Risk factors and clinical presentation of portal vein thrombosis in patients with liver cirrhosis[J]. J Hepatol, 2004, 40(5): 736-41.] [DOI] [PubMed] [Google Scholar]
- 15.Kee KM, Hung CH, Wang JH, et al. Serial changes of clinical parameters in a patient with advanced hepatocellular carcinoma with portal vein thrombosis achieving complete response after treatment with sorafenib. Onco Targets Ther. 2014;15(7):829–34. doi: 10.2147/OTT.S61740. [Kee KM, Hung CH, Wang JH, et al. Serial changes of clinical parameters in a patient with advanced hepatocellular carcinoma with portal vein thrombosis achieving complete response after treatment with sorafenib[J]. Onco Targets Ther, 2014, 15(7): 829-34.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.倪 晓凌, 顾 大镛, 胡 国华. 前列地尔预防脾切除和断流术后门静脉系统血栓形成的历史对照研究. 中华肝胆外科杂志. 2014;20(5):351–4. doi: 10.3760/cma.j.issn.1007-8118.2014.05.009. [倪晓凌, 顾大镛, 胡国华.前列地尔预防脾切除和断流术后门静脉系统血栓形成的历史对照研究[J].中华肝胆外科杂志, 2014, 20(5): 351-4.] [DOI] [Google Scholar]
- 17.潘 卫东, 许 瑞云. 肝炎肝硬化脾切除术后门静脉系统血栓形成的相关因素分析和预防. 中华肝胆外科杂志. 2006;9(12):594–8. doi: 10.3760/cma.j.issn.1007-8118.2006.09.011. [潘卫东, 许瑞云.肝炎肝硬化脾切除术后门静脉系统血栓形成的相关因素分析和预防[J].中华肝胆外科杂志, 2006, 9(12): 594-8.] [DOI] [Google Scholar]
- 18.Jiang GQ, Bai DS, Chen P, et al. Risk factors for portal vein system thrombosis after laparoscopic splenectomy in cirrhotic patients with hypersplenism. J Laparoendosc Adv Surg Tech A. 2016;26(6):419–23. doi: 10.1089/lap.2015.0481. [Jiang GQ, Bai DS, Chen P, et al. Risk factors for portal vein system thrombosis after laparoscopic splenectomy in cirrhotic patients with hypersplenism[J]. J Laparoendosc Adv Surg Tech A, 2016, 26(6): 419-23.] [DOI] [PubMed] [Google Scholar]
- 19.Sun LC, Zhou H, Gu L, et al. Effects of surgical procedures on the occurrence and development of postoperative portal vein thrombosis in patients with cirrhosis complicated by portal hypertension. Int J Surg. 2015;16(PtA):31–5. doi: 10.1016/j.ijsu.2015.02.005. [Sun LC, Zhou H, Gu L, et al. Effects of surgical procedures on the occurrence and development of postoperative portal vein thrombosis in patients with cirrhosis complicated by portal hypertension[J]. Int J Surg, 2015, 16(PtA): 31-5.] [DOI] [PubMed] [Google Scholar]
- 20.叶 勇, 马 金良, 余 继海, et al. 门静脉高压症断流术后血栓形成的多因素分析. 肝胆外科杂志. 2014;22(1):24–6. [叶勇, 马金良, 余继海, 等.门静脉高压症断流术后血栓形成的多因素分析[J].肝胆外科杂志, 2014, 22(1): 24-6.] [Google Scholar]
- 21.Kuroki T, Kitasato A, Tokunaga T, et al. Predictors of portal and splenic vein thrombosis after laparoscopic splenectomy: a retrospective analysis of a single-center experience. Surg Today. 2018;48(8):804–9. doi: 10.1007/s00595-018-1655-6. [Kuroki T, Kitasato A, Tokunaga T, et al. Predictors of portal and splenic vein thrombosis after laparoscopic splenectomy: a retrospective analysis of a single-center experience[J]. Surg Today, 2018, 48(8): 804-9.] [DOI] [PubMed] [Google Scholar]
- 22.吴 晓峰, 伏 志, 刘 召波, et al. 肝硬化门静脉高压患者行脾切除断流术后早期门静脉血栓形成的原因分析. 临床肝胆病杂志. 2015;31(3):400–3. doi: 10.3969/j.issn.1001-5256.2015.03.021. [吴晓峰, 伏志, 刘召波, 等.肝硬化门静脉高压患者行脾切除断流术后早期门静脉血栓形成的原因分析[J].临床肝胆病杂志, 2015, 31 (3): 400-3.] [DOI] [Google Scholar]
- 23.Abdel-Razik A, Mousa N, Elhelaly R, et al. De-novo portal vein thrombosis in liver cirrhosis: risk factors and correlation with the Model for End-stage liver disease scoring system. Eur J Gastroenterol Hepatol. 2015;27(5):585–92. doi: 10.1097/MEG.0000000000000325. [Abdel-Razik A, Mousa N, Elhelaly R, et al. De-novo portal vein thrombosis in liver cirrhosis: risk factors and correlation with the Model for End-stage liver disease scoring system[J]. Eur J Gastroenterol Hepatol, 2015, 27(5): 585-92.] [DOI] [PubMed] [Google Scholar]
- 24.Wu SL, Wu Z, Zhang XG, et al. The incidence and risk factors of portal vein system thrombosis after splenectomy and pericardial devascularization. Turk J Gastroenterol. 2015;26(5):423–8. doi: 10.5152/tjg.2015.0063. [Wu SL, Wu Z, Zhang XG, et al. The incidence and risk factors of portal vein system thrombosis after splenectomy and pericardial devascularization[J]. Turk J Gastroenterol, 2015, 26(5): 423-8.] [DOI] [PubMed] [Google Scholar]
- 25.杨 正茂, 张 岭漪. 肝硬化患者脾切除术后门静脉血栓形成的危险因素. 临床肝胆病杂志. 2015;31(7):1173–5. doi: 10.3969/j.issn.1001-5256.2015.07.044. [杨正茂, 张岭漪.肝硬化患者脾切除术后门静脉血栓形成的危险因素[J].临床肝胆病杂志, 2015, 31(7): 1173-5.] [DOI] [Google Scholar]
- 26.吕 少诚, 顾 万清. 肝硬化患者脾切除术后门静脉血栓的防治研究进展. 中华肝胆外科杂志. 2015;21(1):59–63. doi: 10.3760/cma.j.issn.1007-8118.2015.01.019. [吕少诚, 顾万清.肝硬化患者脾切除术后门静脉血栓的防治研究进展[J].中华肝胆外科杂志, 2015, 21(1): 59-63.] [DOI] [Google Scholar]
