药物性肝损伤患者预后影响因素分析及列线图模型的建立
DOI: 10.12449/JCH240320
Influencing factors for the prognosis of patients with drug-induced liver injury and establishment of a nomogram model
-
摘要:
目的 探讨药物性肝损伤(DILI)患者临床转归的影响因素,构建列线图模型并进行内部验证。 方法 回顾性分析哈尔滨工业大学附属黑龙江省医院2017年1月—2022年12月收治的188例DILI患者的一般资料和实验室数据,根据患者临床转归分为结局良好组(n=146)和不良结局组(n=42)。正态分布计量资料两组间比较采用成组t检验;非正态分布计量资料两组间比较采用Mann-Whitney U检验。计数资料两组间比较采用χ2检验。通过单因素和多因素Logistic回归分析筛选DILI患者临床转归相关的独立影响因素。R Studio 4.1.2软件构建列线图模型,通过校准曲线、受试者工作特征曲线(ROC曲线)和决策曲线分析(DCA)对模型进行内部验证。 结果 单因素Logistic回归分析结果显示,肝活检诊断DILI、PLT、ChE、Alb、PTA、IgM和IgG与DILI患者不良结局相关(P值均<0.05)。多因素Logistic回归分析结果显示,肝活检诊断DILI(OR=0.072,95%CI:0.022~0.213,P<0.001)、临床分型(OR=0.463,95%CI:0.213~0.926,P=0.039)、ALT(OR=0.999,95%CI:0.998~1.000,P=0.025)、PTA(OR=0.973,95%CI:0.952~0.993,P=0.011)和IgM(OR=1.456,95%CI:1.082~2.021,P=0.015)是DILI患者临床转归的独立影响因素。构建列线图,经验证校准曲线接近参考曲线,ROC曲线下面积为0.829,决策曲线分析显示该模型具有良好的临床净收益。 结论 构建的列线图模型对评估DILI患者的临床转归具有较好的临床校准度、鉴别能力和应用价值。 -
关键词:
- 化学性与药物性肝损伤 /
- 预后 /
- 列线图
Abstract:Objective To investigate the influencing factors for the clinical outcome of patients with drug-induced liver injury (DILI), and to establish a nomogram prediction model for validation. Methods A retrospective analysis was performed for the general information and laboratory data of 188 patients with DILI who were admitted to Heilongjiang Provincial Hospital Affiliated to Harbin Institute of Technology from January 2017 to December 2022, and according to their clinical outcome, they were divided into good outcome group with 146 patients and poor outcome group with 42 patients. The independent-samples t test was used for comparison of normally distributed continuous data between two groups, and the Mann-Whitney U test was used for comparison of non-normally distributed continuous data between two groups; the chi-square test was used for comparison of categorical data between two groups. Univariate and multivariate Logistic regression analyses were used to investigate the independent influencing factors for the clinical outcome of DILI patients. R Studio 4.1.2 software was used to establish a nomogram model, and calibration curve, receiver operating characteristic (ROC) curve, and decision curve analysis (DCA) were used to perform internal validation. Results The univariate Logistic regression analysis showed that liver biopsy for the diagnosis of DILI, platelet count, cholinesterase, albumin, prothrombin time activity, IgM, and IgG were associated with adverse outcomes in patients with DILI. The multivariate Logistic regression analysis showed that liver biopsy for the diagnosis of DILI (odds ratio [OR]=0.072, 95% confidence interval [CI]: 0.022 — 0.213, P<0.001), clinical classification (OR=0.463, 95%CI: 0.213 — 0.926, P=0.039), alanine aminotransferase (OR=0.999, 95%CI: 0.998 — 1.000, P=0.025), prothrombin time activity (OR=0.973, 95%CI: 0.952 — 0.993, P=0.011), and IgM (OR=1.456, 95%CI: 1.082 — 2.021, P=0.015) were independent influencing factors for clinical outcome in patients with DILI. The nomogram prediction model was established, and after validation, the calibration curve was close to the reference curve. The area under the ROC curve was 0.829, and the DCA curve showed that the model had good net clinical benefit. Conclusion The nomogram prediction model established in this study has good clinical calibration, discriminative ability, and application value in evaluating the clinical outcome of patients with DILI. -
Key words:
- Chemical and Drug Induced Liver Injury /
- Prognosis /
- Nomograms
-
表 1 188例DILI患者的一般资料
Table 1. General information of 188 patients with DILI
临床资料 总体(n=188) 结局良好组(n=146) 不良结局组(n=42) 统计值 P值 男/女(例) 69/119 55/91 14/28 χ2=0.264 0.607 年龄(岁) 52(43~62) 52(42~62) 56(49~64) Z=1.005 0.316 BMI(kg/m2) 23.34±3.21 23.53±3.29 22.68±2.88 t=1.521 0.168 住院时间(d) 11.0(7.0~15.0) 10.0(7.0~15.0) 12.0(8.2~16.5) Z=1.477 0.140 肝活检诊断DILI[例(%)] χ2=22.968 <0.001 是 28(14.9) 12(8.2) 16(38.1) 否 160(85.1) 134(91.8) 26(61.9) 人工肝治疗[例(%)] χ2=0.017 0.999 是 5(2.7) 4(2.7) 1(2.4) 否 183(97.3) 142(97.3) 41(97.6) RUCAM量表(分) 9.0(8.0~10.0) 9.0(8.0~10.0) 9.0(8.0~11.0) Z=0.687 0.492 临床分型[例(%)] χ2=6.243 0.044 1级 102(54.3) 84(57.5) 18(42.9) 2级 52(27.7) 34(23.3) 18(42.9) 3级 34(18.1) 28(19.2) 6(14.3) 严重程度分级[例(%)] χ2=14.074 0.005 1级(轻度肝损伤) 86(45.7) 67(45.9) 19(45.2) 2级(中度肝损伤) 28(14.9) 24(16.4) 4(9.5) 3级(重度肝损伤) 32(17.0) 28(19.2) 4(9.5) 4级(肝衰竭) 38(20.2) 27(18.5) 11(26.2) 5级(致命) 4(2.1) 0 4(9.5) WBC(×109/L) 5.6(4.3~7.2) 5.6(4.3~7.0) 5.6(4.2~7.5) Z=0.245 0.808 EO(×109/L) 0.07(0.03~0.13) 0.08(0.03~0.14) 0.06(0.02~0.12) Z=0.708 0.480 PLT(×109/L) 202.53±85.06 212.90±80.55 166.48±91.29 t=3.193 0.002 ALT(U/L) 342(113~711) 390(152~786) 159(51~614) Z=2.539 0.011 AST(U/L) 198(82~532) 202(85~551) 193(53~358) Z=1.260 0.208 ALP(U/L) 150(101~227) 155(102~238) 147(95~194) Z=1.004 0.316 GGT(U/L) 177(99~326) 186(113~340) 140(51~296) Z=-1.810 0.071 ChE(U/L) 6 468.72±2 437.08 6 810.22±2 284.95 5 281.62±2 601.15 t=3.702 <0.001 TBA(μmol/L) 35(7~115) 27(6~111) 60(9~129) Z=1.463 0.144 TBil(μmol/L) 44(20~136) 44(19~132) 47(25~139) Z=0.924 0.357 DBil(μmol/L) 22(6~103) 19(6~95) 25(8~116) Z=0.792 0.430 Alb(g/L) 40.0(35.5~43.6) 40.8(36.3~43.7) 36.4(30.9~41.2) Z=3.520 0.001 Glo(g/L) 27(24~31) 27(24~30) 27(26~33) Z=1.205 0.229 PA(mg/dL) 17(10~26) 18(11~27) 12(8~21) Z=2.586 0.010 PT(s) 12.1(11.2~13.5) 11.9(11.0~12.8) 13.3(11.6~17.6) Z=3.520 <0.001 PTA(%) 89(77~100) 91(84~102) 78(52~93) Z=3.729 <0.001 INR 1.07(0.98~1.21) 1.06(0.97~1.14) 1.19(1.03~1.61) Z=4.028 <0.001 总IgE(IU/mL) 61(19~199) 63(19~199) 48(17~167) Z=0.627 0.531 IgA(g/L) 2.65(2.06~3.61) 2.58(1.91~3.49) 2.72(2.31~4.19) Z=1.401 0.162 IgM(g/L) 1.02(0.67~1.51) 1.01(0.66~1.45) 1.16(0.68~2.12) Z=0.986 0.325 IgG(g/L) 13.4(11.5~16.1) 12.9(11.3~15.7) 13.9(12.1~19.2) Z=1.659 0.097 表 2 188例DILI患者可疑药物构成与种类
Table 2. The composition and types of suspicious drugs in 188 patients with DILI
药物分类 例数 构成比(%) 主要药物 TCM-NM-HP-DS 76 40.43 何首乌、土三七、柴胡、保健品、减肥药 抗感染药物(含抗结核药物) 25 13.30 异烟肼、利福平、乙胺丁醇、吡嗪酰胺、头孢类药物、红霉素、克林霉素、更昔洛韦、炎琥宁 非甾体类抗炎药 22 11.70 对乙酰氨基酚、去痛片、依托考昔、安乃近 其他 21 11.17 养血清脑颗粒、参松养心胶囊、五灵胶囊 心血管系统用药 14 7.45 瑞舒伐他汀、阿托伐他汀、硝苯地平、利血平 代谢性疾病用药 11 5.85 二甲双胍、达格列净、甲巯咪唑、丙硫氧嘧啶、别嘌醇 抗肿瘤药物 7 3.72 卡马替尼、氟马替尼、多柔比星、紫杉醇 化学制剂/工业毒物 6 3.19 染发、油烟净、工业毒物 中枢神经系统用药 4 2.13 卡马西平、丙戊酸钠、左乙拉西坦 生物制剂/免疫抑制剂 2 1.06 环孢素、甲氨蝶呤、他莫昔芬、氟维司群、曲妥珠单抗 表 3 188例DILI患者临床转归的影响因素分析
Table 3. Univariate and multivariate Logistic regression analysis of factors influencing clinical outcomes of 188 patients with DILI
临床资料 单因素分析 多因素分析 OR 95%CI P值 OR 95%CI P值 性别 1.209 0.594~2.547 0.608 年龄(岁) 1.012 0.987~1.039 0.347 BMI(kg/m2) 0.915 0.811~1.023 0.131 住院时间(d) 1.020 0.967~1.072 0.450 肝活检诊断DILI 0.146 0.060~0.340 <0.001 0.072 0.022~0.213 <0.001 人工肝治疗 1.155 0.165~22.926 0.899 7.706 0.834~177.800 0.104 RUCAM量表(分) 1.198 0.932~1.536 0.154 临床分型 1.175 0.753~1.812 0.469 0.463 0.213~0.926 0.039 严重程度分级 1.238 0.947~1.619 0.116 WBC(×109/L) 0.998 0.906~1.083 0.964 EO(×109/L) 1.387 0.553~3.340 0.423 PLT(×109/L) 0.993 0.988~0.997 0.002 ALT(U/L) 0.999 0.998~1.000 0.062 0.999 0.998~1.000 0.025 AST(U/L) 0.999 0.998~1.000 0.238 ALP(U/L) 0.999 0.996~1.002 0.550 GGT(U/L) 0.999 0.998~1.000 0.460 ChE(U/L) 1.000 1.000~1.000 <0.001 1.000 1.000~1.000 0.053 TBA(μmol/L) 1.002 0.998~1.007 0.253 TBil(μmol/L) 1.002 0.999~1.004 0.245 0.997 0.993~1.001 0.135 DBil(μmol/L) 1.002 0.998~1.006 0.215 Alb(g/L) 0.922 0.870~0.972 0.004 Glo(g/L) 1.039 0.999~1.083 0.053 PA(mg/dL) 0.980 0.954~0.997 0.073 PT(s) 1.029 0.989~1.085 0.184 PTA(%) 0.973 0.958~0.987 <0.001 0.973 0.952~0.993 0.011 INR 1.255 0.897~2.036 0.219 总IgE(IU/mL) 1.000 0.999~1.001 0.856 IgA(g/L) 1.155 0.970~1.382 0.102 IgM(g/L) 1.328 1.044~1.739 0.025 1.456 1.082~2.021 0.015 IgG(g/L) 1.056 1.008~1.116 0.030 -
[1] KAPLOWITZ N. Idiosyncratic drug hepatotoxicity[J]. Nat Rev Drug Discov, 2005, 4( 6): 489- 499. DOI: 10.1038/nrd1750. [2] NAVARRO VJ, SENIOR JR. Drug-related hepatotoxicity[J]. N Engl J Med, 2006, 354( 7): 731- 739. DOI: 10.1056/NEJMra052270. [3] CHALASANI NP, HAYASHI PH, BONKOVSKY HL, et al. ACG Clinical Guideline: The diagnosis and management of idiosyncratic drug-induced liver injury[J]. Am J Gastroenterol, 2014, 109( 7): 950- 966. DOI: 10.1038/ajg.2014.131. [4] Chinese Medical Association, Chinese Medical Journals Publishing House, Chinese Society of Gastroenterology, et al. Guideline for primary care of drug-induced liver injury(2019)[J]. Chin J Gen Pract, 2020, 19( 10): 868- 875. DOI: 10.3760/cma.j.cn114798-20200812-00900.中华医学会, 中华医学会杂志社, 中华医学会消化病学分会, 等. 药物性肝损伤基层诊疗指南(2019年)[J]. 中华全科医师杂志, 2020, 19( 10): 868- 875. DOI: 10.3760/cma.j.cn114798-20200812-00900. [5] BALACHANDRAN VP, GONEN M, SMITH JJ, et al. Nomograms in oncology: More than meets the eye[J]. Lancet Oncol, 2015, 16( 4): e173- e180. DOI: 10.1016/S1470-2045(14)71116-7. [6] PARK SY. Nomogram: An analogue tool to deliver digital knowledge[J]. J Thorac Cardiovasc Surg, 2018, 155( 4): 1793. DOI: 10.1016/j.jtcvs.2017.12.107. [7] IASONOS A, SCHRAG D, RAJ GV, et al. How to build and interpret a nomogram for cancer prognosis[J]. J Clin Oncol, 2008, 26( 8): 1364- 1370. DOI: 10.1200/JCO.2007.12.9791. [8] CHEN S, YANG CQ. Drug-induced liver injury: Advances and confusions in treatment[J]. J Clin Hepatol, 2021, 37( 11): 2505- 2509. DOI: 10.3969/j.issn.1001-5256.2021.11.001.陈帅, 杨长青. 药物性肝损伤治疗的进展与困惑[J]. 临床肝胆病杂志, 2021, 37( 11): 2505- 2509. DOI: 10.3969/j.issn.1001-5256.2021.11.001. [9] European Association for the Study of the Liver. EASL Clinical Practice Guidelines: Drug-induced liver injury[J]. J Hepatol, 2019, 70( 6): 1222- 1261. DOI: 10.1016/j.jhep.2019.02.014. [10] BONKOVSKY HL, KLEINER DE, GU JZ, et al. Clinical presentations and outcomes of bile duct loss caused by drugs and herbal and dietary supplements[J]. Hepatology, 2017, 65( 4): 1267- 1277. DOI: 10.1002/hep.28967. [11] JEONG R, LEE YS, SOHN C, et al. Model for end-stage liver disease score as a predictor of short-term outcome in patients with drug-induced liver injury[J]. Scand J Gastroenterol, 2015, 50( 4): 439- 446. DOI: 10.3109/00365521.2014.958094.