比較非維生素K拮抗劑口服抗凝血劑與warfarin在心房顫動和糖尿病患者中對糖尿病相關併發症及嚴重低血糖風險的影響

Abstract

引言： 心房顫動和糖尿病是全球性的重要公共衛生問題，隨著人口老化，其盛行率和發病率皆不斷上升。患有心房顫動的病人，尤其又同時患有糖尿病的病人，有相當高的中風發生風險，因此往往需要長期使用口服抗凝血藥物來預防中風。先前有研究指出，維生素K可能對於改善胰島素敏感性和葡萄糖耐受性有所助益。亦有研究進一步指出，非維生素K拮抗劑口服抗凝血劑（non-vitamin K antagonist oral anticoagulants, NOAC）和維生素K拮抗劑華法林（warfarin），對維生素K功能抑制與否之不同，可能導致其使用者，在葡萄糖耐受性和糖尿病控制方面存在差異。然而，目前有關NOAC與warfarin的使用，對於心房顫動和糖尿病病患其各種糖尿病相關併發症發生風險的潛在影響，臨床證據仍然不足。此外，先前曾有研究指出，warfarin與某些糖尿病藥物之間可能有潛在的藥物交互作用，併用時可能增加低血糖風險，但關於NOAC與warfarin的使用對於低血糖風險影響之比較，相關證據仍相當缺乏。在心房顫動和糖尿病病患中，選擇適當的口服抗凝血藥物以最小化糖尿病併發症和低血糖的風險，是重要的臨床問題，然而相關研究證據有限，仍需要進一步探討。

研究目的： 我們的第一項研究旨在探討心房顫動和糖尿病病患使用NOAC相較於warfarin，對糖尿病各項相關併發症和死亡風險的影響。我們的第二項研究旨在探討病患在使用各種糖尿病藥物時，併用NOAC相較於併用warfarin，對於嚴重低血糖風險的影響。

方法： 我們使用台灣全人口健保資料庫，進行回顧性世代研究。我們納入了健保資料庫中曾診斷有心房顫動和糖尿病，並有使用NOAC或warfarin的患者。在研究一中，我們使用目標試驗模擬設計，比較NOAC與warfarin使用者在糖尿病相關併發症（包含大血管併發症、小血管併發症和血糖急症）和死亡風險的差異。我們使用穩定治療權重倒數機率（stabilized inverse probability of treatment weighting, IPTW），來建立一個在不同治療組之間具有相似患者特性的假想族群進行分析。我們採用Cox比例風險模型來估計風險比（hazard ratio, HR）。在研究二中，我們使用泊松回歸模型估計嚴重低血糖的發生率比（incidence rate ratio, IRR）。同樣使用穩定治療權重倒數機率來平衡治療組之間的特徵，以進行分析比較。

結果： 在研究一中，共有19847名NOAC使用者和10372名warfarin使用者被納入分析。與使用warfarin的病患相比，使用NOAC的病患在發展大血管併發症（HR=0.84，95% CI：0.78-0.91，p<0.001）、小血管併發症（HR=0.79，95% CI：0.73-0.85，p<0.001）、血糖急症（HR=0.91，95% CI：0.83-0.99，p=0.043）和死亡（HR=0.78，95% CI：0.75-0.82，p<0.001）的風險皆顯著較低。針對個別NOAC（dabigatran, rivaroxaban, apixaban, and edoxaban）的分析顯示，與warfarin相比，各NOAC的使用者在併發症和死亡風險方面，都有一致的減少趨勢。各項敏感性分析和負對照指標的分析，也都支持我們的研究發現。 在研究二中，我們共納入56774名診斷有心房顫動和糖尿病且有併用糖尿病藥物和口服抗凝血劑之患者進行分析。總體而言，與糖尿病藥物併用warfarin相比，併用NOAC其嚴重低血糖風險顯著較低（IRR=0.73，95% CI：0.63–0.85，p<0.001）。在根據不同糖尿病藥物的次族群分析中，發現在服用糖尿病藥物二甲雙胍（metformin）（IRR=0.73, 95% CI: 0.59–0.89, p=0.002）、二肽基肽酶-4抑制劑（dipeptidyl peptidase-4 inhibitor, DPP-4i）（IRR=0.73, 95% CI: 0.59–0.92, p=0.007）或磺脲類藥物（sulfonylurea）（IRR=0.68, 95% CI: 0.56–0.81, p<0.001）的次族群患者中，併用NOAC者其嚴重低血糖風險顯著較併用warfarin者低。

結論： 在患有心房顫動和糖尿病之患者中，與warfarin相比，使用NOAC與較低的糖尿病相關併發症（例如大血管併發症、小血管併發症、血糖急症）和死亡風險相關。與糖尿病藥物同時使用時，NOAC使用者也比warfarin使用者有顯著較低的嚴重低血糖風險。因此，對於需要使用口服抗凝血劑治療的心房顫動和糖尿病患者而言，NOAC可能是一個比warfarin更好的治療選擇，以降低這些糖尿病相關併發症和死亡風險。

Introduction: Atrial fibrillation (AF) and diabetes mellitus (DM) are significant public health issues, with rising prevalence and incidence in the aging population worldwide. Patients with AF, particularly those with DM, are more likely to experience stroke during their lifetime, necessitating long-term oral anticoagulant use for stroke prevention. Previous evidence suggests that vitamin K may play a beneficial role in improving insulin sensitivity and glucose tolerance. Some studies have further indicated potential differences in glucose tolerance and diabetes control between non-vitamin K antagonist oral anticoagulants (NOACs) use and warfarin use due to their varying effects on vitamin K function. However, current clinical evidence regarding the potential effects of NOACs versus warfarin on the risks of developing various diabetes-related complications in patients with AF and DM remains limited. Furthermore, a few previous studies have suggested potential drug-drug interactions between warfarin and some specific antidiabetic drugs, which may increase the risk of hypoglycemia, but evidence comparing NOACs and warfarin on this issue is scarce. Identifying an optimal oral anticoagulant to minimize the risk of diabetes complications and hypoglycemia in patients with AF and DM is a critical clinical concern, yet relevant evidence is limited and requires further investigation.

Study Aims: Our first study aimed to investigate the risk of diabetes-related complications and mortality in patients with AF and DM taking NOACs compared to those taking warfarin. In our second study (Study 2), we aimed to specifically investigate the risk of serious hypoglycemia when taking various antidiabetic medications with concurrent NOACs compared to those with concurrent warfarin.

Methods: We conducted retrospective cohort studies using nationwide data from Taiwan''s National Health Insurance Research Database (NHIRD). Patients with AF and DM taking either NOACs or warfarin in the NHIRD were identified. In Study 1, we investigated the risks of diabetes-related complications (macrovascular complications, microvascular complications, and glycemic emergencies) and mortality in NOAC versus warfarin users using a target trial emulation framework. Stabilized inverse probability of treatment weighting (IPTW) was used to create a pseudo-population with balance patient characteristics between treatment groups for analyses. Cox proportional hazard models were employed to estimate hazard ratios (HRs). In Study 2, we estimated the incidence rate ratios (IRRs) of serious hypoglycemia using Poisson regression models. Stabilized IPTW was also utilized to balance characteristics between treatment groups for comparisons. Various sensitivity analyses were further performed in both Study 1 and Study 2.

Results: In Study 1, a total of 19847 NOAC users and 10372 warfarin users were included. Patients taking NOACs had significantly lower risks of developing macrovascular complications (HR=0.84, 95% CI: 0.78-0.91, p<0.001), microvascular complications (HR=0.79, 95% CI: 0.73-0.85, p<0.001), glycemic emergency (HR=0.91, 95% CI: 0.83-0.99, p=0.043), and mortality (HR=0.78, 95% CI: 0.75-0.82, p<0.001) than those taking warfarin. The analyses for individual NOACs (dabigatran, rivaroxaban, apixaban, and edoxaban) revealed a consistent trend of decreased complication and mortality risks for NOAC users compared to warfarin users. Several sensitivity analyses and the analyses for negative control outcomes also supported our study findings. In Study 2, a total of 56,774 patients with AF and DM taking antidiabetic drugs with concurrent use of NOACs or warfarin were included for analyses. Overall, concurrent use of NOACs with antidiabetic drugs had a significantly lower risk of serious hypoglycemia than concurrent use of warfarin with antidiabetic drugs (IRR=0.73, 95% CI: 0.63–0.85, p<0.001). In the subgroup analyses according to different antidiabetic drugs, the significantly lower risk of serious hypoglycemia associated with concurrent NOAC use was found in patients taking metformin (IRR=0.73, 95% CI: 0.59–0.89, p=0.002), DPP-4i (IRR=0.73, 95% CI: 0.59–0.92, p=0.007), or sulfonylurea (IRR=0.68, 95% CI: 0.56–0.81, p<0.001).

Conclusions: Among patients with AF and DM, NOAC use was associated with lower risks of diabetes-related complications (e.g., macrovascular complications, microvascular complications, glycemic emergencies) and mortality compared with warfarin use. Patients taking antidiabetic drugs with concurrent use of NOACs also had a significantly lower risk of serious hypoglycemia than those with concurrent use of warfarin. Therefore, NOAC may be a better therapeutic choice than warfarin for decreasing these complications and mortality in diabetic AF patients requiring oral anticoagulant treatment.