Predictive analysis of Th2 cytokines, eosinophils, and T-lymphocyte subsets for anti-tuberculosis drug-induced idiosyncratic liver injury

Abstract

Abstract: Objective To investigate the abnormal expression of Th2 cytokines, peripheral blood eosinophils (EOS), and T-lymphocyte subsets in patients with anti-tuberculosis drug–induced idiosyncratic liver injury (ATD-ILI) and to evaluate their predictive value for adverse hepatic outcomes. Methods Ninety-three patients who received anti-tuberculosis therapy in Huai′an Fourth People′s Hospital between May 2021 and May 2025 were enrolled. According to whether ATD-ILI occurred during treatment, patients were divided into a liver injury group (n=18) and a non-injury group (n=75). Serum levels of interleukin-4 (IL-4), IL-5, and IL-13 were measured, EOS counts were recorded, and CD4⁺, CD8⁺, and CD4⁺/CD8⁺ ratios were analyzed by flow cytometry. Differences in general clinical data, liver function indices, and immunological parameters between the two groups were compared. Variables with statistical significance in univariate analysis were entered into a multivariate logistic regression model to identify independent risk factors. Receiver operating characteristic (ROC) curves were constructed to assess the predictive performance of individual indicators and their combined application for ATD-ILI. Results In the liver injury group, patients had ALT (156.94 ± 54.78 U/L), AST (121.84 ± 36.52 U/L), TBil (39.61 ± 7.98 μmol/L), and RUCAM score (9.17 ± 3.03), all of which were higher than those in the non-injury group (ALT: 49.73 ± 8.28 U/L, AST: 31.47 ± 7.62 U/L, TBil: 15.84 ± 4.26 μmol/L, RUCAM score: 4.36 ± 1.57), with statistically significant differences (all P<0.05). The liver injury group also exhibited elevated IL-4 levels (23.88 ± 4.82 pg/mL), EOS (0.74 ± 0.24 × 10⁹/L), and CD4⁺/CD8⁺ ratio (2.27 ± 0.57) compared with the non-injury group (IL-4: 16.14 ± 4.17 pg/mL, EOS: 0.45 ± 0.16 × 10⁹/L, CD4⁺/CD8⁺: 1.51 ± 0.42), all differences being statistically significant (all P<0.05). Multivariate logistic regression analysis demonstrated that increased RUCAM score (OR=2.077, 95% CI: 1.070~4.034, P=0.031), elevated IL-4 level (OR=1.478, 95% CI: 1.108~1.972, P=0.008), increased EOS (OR=10.097, 95% CI: 2.158~47.254, P=0.003), and elevated CD4⁺/CD8⁺ ratio (OR=7.504, 95% CI: 1.809~31.132, P=0.005) were independent risk factors for anti-tuberculosis drug-induced idiosyncratic liver injury (all P<0.05). The combination of RUCAM score with IL-4, EOS, and CD4⁺/CD8⁺ ratio exhibited the best predictive performance, with an AUC of 0.976 (95% CI: 0.940~1.000), sensitivity of 88.9%, specificity of 100.0%, and Youden index of 0.889, markedly outperforming individual indicators. Conclusion Immune dysregulation involving Th2 cytokines, eosinophils, and T-lymphocyte subsets is closely associated with the occurrence and severity of anti-tuberculosis drug-induced idiosyncratic liver injury. Their combined assessment provides a valuable biological tool for predicting the risk of idiosyncratic liver injury and for evaluating adverse hepatic outcomes.

Key words: Th2 cytokines, eosinophils, T-lymphocyte subsets, Anti-tuberculosis drugs, Idiosyncratic liver injury

JIANG Hong-mei, FEI Zhong-ting, YANG Jing-yue, MA Xue-jiao. Predictive analysis of Th2 cytokines, eosinophils, and T-lymphocyte subsets for anti-tuberculosis drug-induced idiosyncratic liver injury[J]. Chinese Hepatolgy, 2026, 31(5): 732-736.

References

[1] 阮军,尹恒,寇国先,等. HIV/MTB双重感染合并抗结核药物性肝损伤患者的临床特征[J]. 肝脏,2023,28(4):483-486.
[2] 何虹,吴桂辉,司果. 泼尼松联合标准四联抗结核疗法对结核性胸膜炎临床指标及炎症因子的影响[J]. 中国现代医学杂志,2025,35(11):67-71.
[3] Cheng J, Chen R, Pan H, et al. Construction and validation of a predictive model for the risk of anti-tuberculosis drug-induced liver injury based on machine learning algorithms[J]. J Clin Pharmacol, 2025.
[4] Ramalingam G, Khan J M, Jasmine S, et al. Role of Th1/Th2 cytokine balance in predicting treatment outcomes and disease severity in tuberculosis[J]. J King Saud Univ Sci, 2024, 36(11):103538.
[5] 吴于青,姜国强,刘㑇,等. 抗结核药物超敏反应患者PBMCs中T淋巴细胞亚群及CD69、CD40L的表达意义[J]. 重庆医学,2021,50(5):784-786,790.
[6] Chalasani N P, Maddur H, Russo M W, et al. ACG clinical guideline:diagnosis and management of idiosyncratic drug-induced liver injury[J]. Am J Gastroenterol, 2021, 116(5):878-898.
[7] 中华医学会结核病学分会,《中华结核和呼吸杂志》编辑委员会. 抗结核药所致药物性肝损伤诊断与处理专家建议[J]. 中华结核和呼吸杂志,2013,36(10):732-736.
[8] Huang D, Peng J, Lei L, et al. Time of liver function abnormal identification on prediction of the risk of anti-tuberculosis-induced liver injury[J]. J Clin Transl Hepatol, 2022, 11(2):425.
[9] Liu R, Liu X, Wang Y. Network toxicology and molecular docking analysis of antituberculosis drug-induced hepatotoxicity[J]. J Biochem Mol Toxicol, 2025, 39(11):e70602.
[10] Chen C B, Hung W K, Wang C W, et al. Advances in understanding of the pathogenesis and therapeutic implications of drug reaction with eosinophilia and systemic symptoms:an updated review[J]. Front Med (Lausanne), 2023, 10:1187937.
[11] Devarbhavi H, Kurien S S, Raj S, et al. Idiosyncratic drug-induced liver injury associated with and without drug reaction with eosinophilia and systemic symptoms[J]. Am J Gastroenterol, 2022, 117(10):1709-1713.
[12] Porter M, Smith R, Teixeira N, et al. First-line antituberculosis drug challenge reactions in drug reaction with eosinophilia and systemic symptoms syndrome in an HIV endemic setting[J]. J Allergy Clin Immunol Pract, 2024, 12(10):2798-2808. e12.
[13] He Y, Zheng X, Dang Z, et al. Association of T-cell profiles with disease severity, drug-induced liver injury, and treatment completion in tuberculosis[J]. Clin Respir J, 2025, 19(8):e70114.
[14] 邹艳梅,陈红霞,王剑,等. 白蛋白/碱性磷酸酶比值与初次抗结核治疗诱发药物性肝损伤的关系[J]. 转化医学杂志,2025,14(11):114-118.
[15] 谢琴琴,季欢欢,龚美玲,等. 抗结核药物性肝损伤发生率及其危险因素的Meta分析[J]. 中国药房,2024,35(24):3052-3058.
[16] Wu S Q, Yang Q H, Li H, et al. The feasibility of using liver function indices and FibroScan in combination to predict the occurrence of anti-tuberculosis drug-induced liver injury in patients with liver disease[J]. J Med Biochem, 2025, 44(1):17.
[17] Lim W S, Avery A, Kon O M, et al. Anti-tuberculosis drug-induced liver injury[J]. BMJ, 2023, 383.
[18] Teschke R, Danan G. Idiosyncratic DILI and RUCAM under one hat:the global view[J]. Livers, 2023, 3(3):397-433.
[19] Arora N, Chalasani N, Rahnama-Moghadam S. A review of drug-induced liver injury with rash, eosinophilia, and systemic symptoms (DRESS) syndrome:cutaneous manifestations, clinical features, and management[J]. Clin Liver Dis (Hoboken), 2024, 23(1):e0198.
[20] Dong H, Feng J, Chang X, et al. Predictive value of systemic immune-inflammatory biomarkers for drug-induced liver injury in hepatitis B virus surface antigen positive tuberculosis patients:a retrospective observational study[J]. Medicine (Baltimore), 2024, 103(45):e40349.