類風濕性關節炎患者伺機性感染風險Janus激酶抑制劑與TNF-α抑制劑相關結核病與帶狀皰疹風險之比較隨機對照試驗之系統性文獻回顧與統合分析

背景：Janus kinase抑制劑與Tumor necrosis factor-alpha（TNF-α）抑制劑為治療類風濕性關節炎之常用二線藥物，但其相關機會性感染風險，特別是結核病與帶狀疱疹，在臨床應用上仍具不確定性。目的：本研究進行系統性回顧與統合分析，探討Janus kinase抑制劑相較於TNF-α抑制劑在類風濕性關節炎病患中引發結核病與帶狀疱疹感染之風險差異。方法：本研究依據PRISMA 2020指引進行系統性回顧與統合分析，系統性搜尋Ovid Medline、Embase、Cochrane （CENTRAL）等資料庫至2025年5月18日，額外檢索2012-2024年美國風濕病學會（ACR/ARHP）年會摘要，篩選隨機對照試驗。結局變項為感染事件之odds ratio與risk difference，採隨機效應模式，並執行敏感性分析、Trial Sequential Analysis分析、出版偏誤評估、Risk of Bias風險評估、劑量反應回歸分析、不同藥品次族群分析與GRADE證據等級評比。結果：共納入9篇隨機對照試驗，總樣本數10157人。（一）結核病方面，Janus kinase抑制劑與TNF-α抑制劑之風險無顯著差異（OR = 0.59, 95% CI: 0.26-1.34；RD = -0.00, 95% CI: -0.00-0.00），異質性低（I² = 0%），TSA（RIS = 80233），GRADE證據等級評為「中等」。（二）帶狀疱疹方面，Janus kinase抑制劑之相對風險顯著升高（OR = 2.15, 95% CI: 1.19-3.91；RD = 0.02, 95% CI: -0.00-0.04），異質性高（I² = 96%），TSA（RIS = 47526），GRADE證據等級評為「低」。結論：本研究顯示，Janus kinase抑制劑與TNF-α抑制劑在結核病感染的相對與絕對風險方面無顯著差異；而在帶狀疱疹感染方面，Janus kinase 抑制劑之相對風險顯著升高。臨床上建議應謹慎評估個別病患的疾病嚴重程度、免疫狀況及感染風險，對使用Janus kinase抑制劑之患者，可考慮施打重組型帶狀疱疹疫苗作為預防策略。未來仍需更多具高品質設計之大型隨機對照試驗或真實世界研究，以進一步確認不同藥物間之感染風險差異。關鍵字：類風濕性關節炎、Janus kinase抑制劑、TNF-α抑制劑結核病感染風險、帶狀疱疹感染風險

Background:Janus kinase inhibitors and tumor necrosis factor-alpha (TNF-α) inhibitors are commonly used second-line therapies for rheumatoid arthritis. However, their associated risk of opportunistic infections, particularly tuberculosis and herpes zoster virus infection, remains uncertain in clinical practice.Objective:This study aimed to conduct a systematic review and meta-analysis comparing the risk of tuberculosis and herpes zoster virus infections in rheumatoid arthritis patients treated with Janus kinase inhibitors versus TNF-α inhibitors.Methods:Following the PRISMA 2020 guidelines, we systematically searched Ovid MEDLINE, Embase, and Cochrane CENTRAL databases up to May 18, 2025. Abstracts from the American College of Rheumatology (ACR/ARHP) annual meetings from 2012 to 2024 were also reviewed. We included randomized controlled trials comparing Janus kinase inhibitors with TNF-α inhibitors. The primary outcomes were odds ratios (OR) and risk differences (RD) for infection events, analyzed using a random-effects model. Sensitivity analyses, trial sequential analysis (TSA), publication bias assessment, risk of bias evaluation, dose–response meta-regression, subgroup analyses by individual agents, and GRADE assessments were conducted.Results:A total of 9 randomized controlled trials involving 10157 participants were included.1. For tuberculosis, there was no statistically significant difference in infection risk between Janus kinase inhibitors and TNF-α inhibitors(OR = 0.59, 95% CI: 0.26-1.34; RD = -0.00, 95% CI: -0.00-0.00) , with low heterogeneity (I² = 0%). TSA suggested the required information size (RIS) was 80233, and the GRADE level of evidence was rated as “moderate”.2. For herpes zoster, Janus kinase inhibitors were associated with a significantly increased relative risk (OR = 2.15; 95% CI: 1.19–3.91; RD = 0.02, 95% CI: -0.00-0.04). Heterogeneity was substantial (I² = 96%). TSA indicated an RIS of 47526, and the GRADE level of evidence was rated as “low”.Conclusion and suggestion:This study demonstrates that Janus kinase inhibitors and TNF-α inhibitors carry comparable risks of tuberculosis infection in rheumatoid arthritis patients. However, Janus kinase inhibitors are associated with a significantly increased relative risk of herpes zoster virus infection, though the absolute risk increase is modest, and the overall evidence remains uncertain. Clinicians should carefully assess individual patients' disease severity, immune status, and infection risk before prescribing Janus kinase inhibitors. In high-risk individuals, vaccination with the recombinant zoster vaccine may be considered as a preventive measure. Further high-quality randomized controlled trials and real-world evidence are needed to clarify infection risk differences among targeted therapies.Key words: Rheumatoid arthritis, Janus kinase inhibitors, TNF-α inhibitors, risk of tuberculosis infection, risk of herpes zoster virus infection