抗憂鬱藥品與心臟瓣膜疾病風險：嵌入型病例對照研究

Antidepressants and valvular heart disease: A nested case-control study in Taiwan

林佳慧(Chia-Hui Lin)

指導教授 ： 沈麗娟

共同指導教授 ： 蕭斐元(Fei-Yuan Hsiao)

國立臺灣大學/醫學院/臨床藥學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.03154

摘要

研究背景
已知藥品引發的心臟瓣膜疾病之致病機轉可能與血清素(serotonin)濃度、受體(receptor)、運輸體(transporter)相關，如serotonin、fenfluramine及pergolide等，而抗憂鬱藥品其藥理機轉也與serotonin有關，然而之前文獻並沒有發現抗憂鬱藥品與心臟瓣膜疾病(valvular heart disease; VHD)之間的相關性。除研究方法及族群不同外，目前缺乏心臟瓣膜閉鎖不全(cardiac-valve regurgitation；CVR)之International Classification of Diseases-9th edition-Clinical Modification codes (ICD-9-CM code)編碼確效研究，因此，利用臺灣健康保險資料庫(National Health Insurance Research Database；NHIRD)來進行VHD之研究。
研究目的 本研究首先進行住院與門診病人之ICD-9-CM code編碼與CVR之確效研究，以確認VHD病例設定條件。其次，利用臺灣健康保險資料庫探討華人族群抗憂鬱藥品之使用與VHD的相關性。

研究方法
本研究分兩大部分，第一部分，利用2007至2011年臺大醫院住院及門診病人之病歷資料，篩選符合病例條件之住院或門診病人。進一步確認心臟超音波報告中CVR之嚴重程度，將至少有一瓣膜CVR大於等於中度之病人視為確認病例，並計算陽性預測值(positive predictive value；PPV)。第二部分，利用1999至2011年健保資料庫2000、2005、2010之百萬承保歸人檔(約300萬人)建立嵌入型病例對照研究。先排除在1999至2001曾使用抗憂鬱藥品之病人後，收錄2002至2010年大於等於20歲至少使用三次抗憂鬱藥品之病人，並且將第一次使用抗憂鬱藥品日期設定為cohort entry date。排除在cohort entry date前一年曾有VHD、相關病因與可疑藥品之病人後，剩下的人進入研究世代。在研究世代中，篩選出在2002至2011年符合VHD條件之住院病人定義為病例，並且將第一次符合VHD的住院日期定義為index date。排除開始用藥後90天內發生VHD之病例病人，剩下之病例病人以1:4的比例與未發生心臟瓣膜疾病之抗憂鬱藥品使用者，進行年齡、性別、cohort entry date之配對，得到病例組與對照組。分析事件前三年內兩組抗憂鬱藥品的使用狀況，主要分析為將所有抗憂鬱藥品視為一個類別，暴露變項包含使用時序、累積時間、累積劑量以及最後一劑每日劑量；次要分析則依照藥理特性、對serotonin transporter親和力，以及單獨抗憂鬱藥品進行分類。利用
conditional logistic regression進行odds ratio的預測並校正干擾因子，以評估抗憂鬱藥品與VHD之相關性。

研究結果
住院與門診病人之ICD-9-CM code編碼確效研究結果不同，住院病人之PPV高達84.5%，相較之下，門診病人只有33.7%，因此只納入住院病人為病例組。收錄研究族群154,416人，排除14,824人，最後有139,592人進入到研究世代。研究世代中有1,792人符合病例條件，進一步排除從第一次使用抗憂鬱藥品到發生心臟瓣膜疾病小於90天之133名病人，剩下病人經過配對得病例組1,618人與對照組6,742人。結果顯示，主要分析中整體抗憂鬱藥品之使用者，相較於三年內未使用者，與心臟瓣膜疾病之發生不存在顯著相關性(adjusted odds ratio[aOR] 1.05; 95% confidence interval[CI] 0.89-1.23)。另外，累積時間、累積劑量與最後一劑每日劑量亦皆無發現兩者之間的關係性。但是，次要分析則發現在trazodone(aOR 1.26 [1.00-1.59])、imipramine(aOR 1.29 [1.03-1.62])、citalopram(aOR 1.70 [1.06-2.72])、duloxetine(aOR 2.41 [1.11-5.25])之current user與bupropion(aOR 2.87 [1.09-7.56])之recent user有看到顯著增加心臟瓣膜疾病之發生風險。

結論
整體而言，本研究並未發現整體抗憂鬱藥品與心臟瓣膜疾病之相關性，但是值得注意的是，半年內曾經暴露過citalopram、imipramine、duloxetine的患者，以及在半年到一年內曾經暴露過bupropion的患者，相較於三年內未使用抗憂鬱藥品者，可能會增加1.26至2.87倍心臟瓣膜疾病之發生機率。

關鍵字

抗憂鬱藥品 ； 心臟瓣膜疾病 ； 台灣全民健康保險資料庫 ； 嵌入型病例對照研究

並列摘要

Background Concentration, receptor, transporter of serotonin may involve in mechanisms of drug induced valvular heart disease such as serotonin, fenfluramine, pergolide and so on. Additionally, pharmacologic mechanism of antidepressants are related to serotonin. Nevertheless, no clinical study shows a potential link between use of antidepressants and risk of valvular heart disease (VHD). In addition to different study method and population, the validation of International Classification of Diseases-9th edition-Clinical Modification codes (ICD-9-CM codes) for cardiac-valve regurgitation (CVR) has never been done to study VHD by Taiwan’s National Health Insurance Research Database (NHIRD).
Objective First, to determine the definition of VHD, we conducted two validations of ICD-9-CM codes for inpatients and outpatients with CVR, respectively. Second, we used NHIRD to evaluate the association between use of antidepressants and VHD among Chinese population.
Methods Our study was divided into two parts. In the first part, we requested inpatients’ and outpatients’ medical information from National Taiwan University Hospital and identified patients who met case definition during 2007-2011. The echocardiographic reports were reviewed to confirm the severity of CVR. Potential cases with at least one cardiac valve of moderate regurgitation were defined as confirmed cases and positive predictive value (PPV) were calculated. In the second part, a nested case-control study using 2000, 2005, 2010 Longitudinal health insurance database during 1999-2011 as study resource (covering about a population of 3 millions) was conducted. After excluding patients who used antidepressants in 1999 to 2001, we identified patients aged over 20 years with at least 3 prescriptions of antidepressants during 2002-2010. The cohort entry date was the date of the first prescription of antidepressants. Nevertheless, patients who were diagnosed with VHD or related etiologies, or who used drugs that potential causing DIVHD within 1 year before the cohort entry date were excluded. Among study population, we identified inpatients who first hospitalized with incident VHD during 2002-2011 as cases and the date was defined as index date. We further excluded cases with VHD 90 days after the first use of antidepressants. The cases were matched with 4 controls by age, sex and cohort entry date. All prescriptions of antidepressants 3 year before index date were included. In primary analysis, different type of antidepressants were considered as one group and exposure assessment included timing of use, cumulative duration, cumulative dose and last daily dose. In secondary analyses, we further classified antidepressants by type of antidepressants, affinity for serotonin transporter and individual antidepressants. Conditional logistic regression models were used to estimate odds ratio of VHD associated with use of antidepressants.
Results The validity of ICD-9-CM code for inpatients with CVR was distinct from outpatients with CVR. The PPV of inpatients was as high as 84.5%. In contrast, the PPV of outpatients was only 33.7%. Therefore, definition of cases was decided to include inpatients solely. In the study period, 154,416 patients met the inclusion criteria. After excluding 14,824 patients, 139,592 patients served as the study population. Among study population, we identified 1,792 inpatients who met case definition and further excluded 133 cases with days between cohort entry date and index date less than 90 days. After matching, 1,618 and 6,742 were the cases and controls, respectively. In primary analysis, users of all antidepressants were not associated with risk of VHD (adjusted odds ratio [aOR] 1.05; 95% confidence interval [CI] 0.89-1.23) when compared with ever users who did not use antidepressants 3 year before index date. In addition, no dose-response was observed by cumulative duration, cumulative dose and last daily dose. In secondary analysis, only current users of trazodone (aOR 1.26 [1.00-1.59]), imipramine (aOR 1.29 [1.03-1.62]), citalopram (aOR 1.70 [1.06-2.72]), duloxetine (aOR 2.41 [1.11-5.25]) and recent users of bupropion (aOR 2.87 [1.09-7.56]) significantly increased the risk of incident VHD.
Conclusions Overall, we found antidepressants was not associated with risk of VHD. Nonetheless, it is rather remarkable that patients used trazodone, imipramine, citalopram, duloxetine or bupropion may increase 1.26 to 2.87-fold risk of incident VHD.

並列關鍵字

antidepressants ； valvular heart disease ； National Health Insurance Research Database (NHIRD) ； Taiwan ； nested case-control study

參考文獻

53. 李佩芬. 多巴胺促效劑之處方型態分析及其對巴金森氏病病人心臟瓣膜病變的影響. 台北: 國立臺灣大學; 2008.

1. Iung B. A prospective survey of patients with valvular heart disease in Europe: The Euro Heart Survey on Valvular Heart Disease. Eur Heart J 2003;24:1231-43.

2. Nkomo VT, Gardin JM, Skelton TN, et al. Burden of valvular heart diseases: a population-based study. The Lancet 2006;368:1005-11.

3. Smith SA, Waggoner AD, de las Fuentes L, et al. Role of serotoninergic pathways in drug-induced valvular heart disease and diagnostic features by echocardiography. J Am Soc Echocardiogr 2009;22:883-9.

4. Roth BL. Drugs and valvular heart disease. N Engl J Med 2007;356:6-9.