胰島素增敏劑對於第二型糖尿病病人心臟血管副作用之藥物流行病學研究

研究背景心血管疾病是第二型糖尿病患者死亡的主因。而近期國外的綜合分析(meta-analysis)顯示，用來治療第二型糖尿病的藥物胰島素增敏劑-- rosiglitazone可能增加心血管疾病的發生率(例如：心肌梗塞、心衰竭)，甚至增加因心血管事件而導致的死亡率。同類藥物pioglitazone的大型臨床試驗，以心血管事件為其最終分析之主要標的，研究顯示不盡相同的結果。因此釐清此類藥品在心血管的副作用為目前重要之議題。目前國內尚無相關的全國性糖尿病藥物造成心血管副作用之藥物流行病學之研究發表，本研究對於國內病人臨床用藥之評估與醫療政策等觀點能夠提供新的思維與參考。研究目的針對第二型糖尿病病人，探討單獨或併服使用胰島素增敏劑(thiazolidindiones, TZD)之病人，相較於使用其他口服降血糖藥物之病人心血管住院事件發生的風險，同時也針對目前市場上兩種胰島素增敏劑rosiglitazone與pioglitazone進行分析比較，試圖了解兩者在心血管副作用上是否有class effect。研究方法本研究為一回溯性世代研究，以健保資料庫2005年100萬人歸人檔為研究材料，針對2003年7月至2006年12月間，18歲以上，第一次使用糖尿病藥物之病患為研究對象。第一部分以研究期間使用過與未使用過TZD類藥物之病人做分組，進行背景資料的描述性分析，探討兩組病人在人口統計學、臨床用藥傾向上的不同。第二部分利用time dependent Cox proportional hazard regression model來進行存活分析，主要研究終點為急性心肌梗塞以及心臟血管手術之住院事件的發生，次要研究終點則為中風之住院事件以及心衰竭造成之住院事件。探討曾經使用或正在使用此類TZD藥物對於三個研究終點的影響，並將rosiglitazone與pioglitazone個別分析。由於使用TZD之病人在背景上與未使用族群不同，可能對研究結果造成影響，我們針對TZD的使用建立propensity score，並納入回歸模式來進行統計校正。研究結果經過篩選後，資料庫中共有20861位符合研究條件的病人，第一部分研究結果，研究期間有使用過TZD的病人相較未使用者有著年齡較輕，追蹤時間較長，男性居多，在糖尿病用藥及心血管用藥比例較高的特質。而使用rosiglitazone與pioglitazone的病人也在用藥傾向及開方型態上有所不同。第二部分研究結果中，針對心肌梗塞以及心臟血管手術事件分析，並校正過可能之干擾因子後，顯示曾暴露過TZD類藥物之病人相較於未暴露者發生事件的hazard ratio為1.47 (95% CI=1.01-2.13, p=0.044)，若將rosiglitazone與pioglitazone分開討論，使用rosiglitazone存在較高之罹病風險，hazard ratio為1.65-1.68，而Pioglitazone對於心肌梗塞以及心臟手術事件則沒有顯著影響。在回溯性分析使用rosiglitazone的病人以及發生心肌梗塞或者心臟手術事件病人的資料，發現可能增加rosiglitazone使用病人發生事件的風險因素，包括較嚴重的高血脂問題以及缺血性心臟病。中風事件的分析中，由propensity score的結果可知，傾向使用TZD類藥物之病人在背景資料上即存在較高罹患中風之風險，但正在使用TZD類藥物的病人對於中風有一保護的趨勢，hazard ratio=0.51 (95% CI=0.24-1.08, p=0.08)。最後心衰竭事件的分析，不論將TZD類藥物合併討論或將兩個藥物分開討論，是否使用藥物與心衰竭事件的發生並無顯著相關。結論研究期間有使用過TZD的病人相較於未使用者，在人口統計學、用藥及住院的背景資料有許多不同之處，而使用rosiglitazone與pioglitazone之病人也在用藥傾向及開方型態上有所不同。曾經使用TZD類藥物，尤其是其中的rosiglitazone，對於心肌梗塞以及心臟手術事件的發生存在較高風險，而對於中風的住院事件與在心衰竭事件的分析上，TZD類藥物則沒有達到顯著影響，但使用TZD類藥物對於中風事件似乎有一保護的趨勢。本研究成果顯示對於已有缺血性心臟病高風險，高血脂控制不佳之病人，對於rosiglitazone此類的糖尿病藥物選用必須更加謹慎，或選擇替代藥品。並建議在中風事件的研究可以全國資料庫作為樣本，篩選出更多病例做進一步分析，獲得更明確的結果。

關鍵字

胰島素增敏劑；副作用；急性心肌梗塞；中風；心衰竭

並列摘要

Background Cardiovascular disease is one of the main causes of death in patients with type II diabetes. Meta-analysis of several clinical trials have raised concerns about increasing risk of myocardial infarction and death from cardiovascular causes in patients treated with rosiglitazone, a drug for diabetes mellitus in the class of thiazolidinedione (TZD). However, a large-scale clinical trial about pioglitazone using time to the occurrence of macrovascular events or death as the primary endpoint focused on has demonstrated a protective effect of pioglitazone on cardiovascular events. Due to lack of national studies examining the effect of TZD on the cardiovascular risk, we conducted the current study to provide a reference for clinical drug safety evaluation and medical policies decision. Objective The study aimed to assess the risk of cardiovascular events between patients who used and who did not use TZD as mono-therapy or combination therapy for DM; and to compare the cardiovascular side effects between rosiglitazone and pioglitazone in order to identify if there was a class effect in the two drugs. Methods This was a retrospective cohort study. We used the National Health Insurance Research Database (NHIRD) to establish a cohort of patients with type 2 diabetes, aged 18 years and older, who initiated any anti-diabetic agents between July 2003 and December 2006. Only patients who didn’t take any antidiabetic agents before were included. Descriptive analysis. The two comparison groups: (1) TZD group is defined as those patients who had any claims for TZDs during our study period, and (2) non-TZD group. We compared the pattern of antidiabetic medication use,, demographic and clinical characteristics, and risk factor for CVD events between the two groups. We also compare the difference in prescription pattern between rosiglitazone users and pioglitazone users。 Analysis for the relation between TZD use and cardiovascular events. Our primary end point was time to acute myocardial infarction (AMI) and coronary revascularization (CR). Our secondary end points were hospitalizations for stroke or congestive heart failure (CHF). We constructed two time dependent Cox proportional hazards models to evaluate the hazard ratio (HR) of TZD use for each study end points we were interested in. One of the models was constructed to estimate the HR for patients who had ever exposed to TZDs (TZD ever users), and the other one for patients who were current user of TZDs (TZD current user). Because of the different baseline characteristics between the TZD and non-TZD groups, we used the propensity score for the statistic matching to decrease the confounding effects from indication of use. We also conducted statistical analysis to identify the impact of rosiglitazone and pioglitazone on different end points separately. Results A total of 20861 eligible patients were identified from the NHIRD. Compared with the non-TZD group, patients in the TZD group were younger, male dominant, treated with more antidiabetic agents and cardiovascular agents use, and were followed for a longer time. There were different patterns of prescription and clinical characteristics between rosiglitazone users and pioglitazone users. After adjusting several confounding factors, patients exposed to TZDs were more likely than those without to have hospitalization for AMI or CR (HR = 1.47, 95% CI=1.01-2.13, p=0.044). If we analyzed the impact of rosiglitazone and pioglitazone separately, rosiglitazone use was associated with a higher risk for AMI or CR (adjusted HR =1.65, 1.68). There was no significant effect of pioglitazone on risks for AMI or CR. The analyses of baseline characteristics of patients who used rosiglitazone or who suffered from AMI or CR retrospectively revealed that patients treated with rosiglitazone were more likely to be younger, and to use second line oral hypolglycemic agents,and lipid-lowering agents. Cox regression analysis revealed that patients who had higher propensity to use TZDs were more likely to have hospitalization for stroke. TZD current user may be less likely to have stroke event (Adjusted HR = 0.51, 95% CI=0.24-1.08, p=0.08). TZD use had no significant impact on hospitalization for CHF. Conclusions Our findings suggest significant differences in demographic and clinical characteristics between the TZD group and the non-TZD group; and significant different prescription patterns between rosiglitazone users and pioglitazone users. Moreover, patients who had ever exposed to TZD, especially rosiglitazone, were at elevated risk of hospitalization for AMI or CR. We recommend that patients who are already in the high risk for ischemic heart disease, who have poor controlled hyperlipidemia may have an additional risk , and should use the rosiglitazone carefully and consider alternative therapies. The current study failed to provide evidence to support an association between TZD use and hospitalization for stroke or HF. It warrants to exam the prediction of TZD to stroke events in further studies according our findings of a marginally significant relationship between TZD and stroke.

並列關鍵字

thiazolidinedione ； adverse event ； acute myocardial infection ； stroke ； heart failure

參考文獻

1. World Health Organization. 2008. (Accessed at http://www.who.int/topics/diabetes_mellitus/en/.)

3. American Diabetes Association 2008. (Accessed 20080313, at http://www.diabetes.org/diabetes-statistics.jsp.)

4. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352(9131):837-53.

5. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352(9131):854-65.

6. U.S. Food and Drug Administration. 2007. (Accessed at http://www.fda.gov/medwatch/index.html )

國際替代計量

胰島素增敏劑對於第二型糖尿病病人心臟血管副作用之藥物流行病學研究

主題瀏覽