腦中風與頸動脈粥狀硬化的非傳統危險因子與存活分析研究

Studies on Unconventional Risk Factors and Survival of Stroke and Carotid Atherosclerosis

鄭建興(Jiann-Shing Jeng)

指導教授 ： 陳建仁

共同指導教授 ： 葉炳強(Ping-Keung Yip)

國立臺灣大學/公共衛生學院/流行病學研究所/博士(2005年)

https://doi.org/10.6342/NTU.2005.02719

摘要

腦血管疾病一直為台灣的主要死因，雖然自1981年的死因首位降至2004年的第三位，但因台灣社會人口的快速老化、腦中風的年齡層發生率變化不大、疾病致死率的下降，腦血管疾病將仍是台灣中、老年人的最重要的身心障礙原因。
台灣的腦血管疾病相關流行病學研究雖可回溯至日據時代，但主要為近20年的研究，在本論文的第一部分將廣泛的回顧已正式發表的台灣腦血管疾病流行病學研究報告，將自腦中風的生命統計資料研究、社區為基礎腦中風盛行率與發生率研究、以醫院登錄為基礎的腦中風研究、病例-對照研究、臨床試驗、遺傳流行病學研究、存活分析研究等回顧。過去台灣的腦血管疾病研究主要偏重於傳統的危險因子，本論文的研究將注重於獨特、非傳統的危險因子探討，包括懷孕相關的腦中風研究與頸部放射治療的梗塞性腦中風研究。此外，頸動脈狹窄在台灣已愈來愈常見，因此針對接受不同治療的頸動脈狹窄患者，進行長期存活分析的探討。
本論文的研究架構與方法主要為「台大醫院腦中風與頸動脈粥狀硬化研究（簡稱SCAN）」，SCAN是以台大醫院為基礎的登錄研究，目的為：（1）了解不同型態腦中風的危險因子、病理生理機轉與長期預後；（2）探討特殊型態的腦血管疾病，如與懷孕/妊娠相關的腦中風、腦靜脈栓塞、放射治療後腦血管疾病等；（3）了解頸動脈粥狀硬化的危險因子、致病機轉、不同的治療方式對於頸動脈狹窄的長期預後。登錄庫分為腦中風登錄庫與頸動脈狹窄登錄庫，對於各型態腦中風與危險因子的診斷與頸動脈粥狀硬化的量化等均經過信效度的檢測，此長期、一致、大尺度的登錄庫，可提供腦血管疾病廣泛的臨床與研究資料的收集，特別有助於一些特殊疾病的研究。依據SCAN架構，本論文研究提出三個子計畫，包括子計畫一的懷孕相關的腦中風研究、子計畫二的頸部放射治療的腦梗塞中風研究、子計畫三的頸動脈粥狀硬化狹窄病患的長期預後與存活分析。
懷孕與產褥會增加腦中風發生的危險性，腦中風也是懷孕與產褥時期的重要致病與致死原因。之前懷孕相關的腦中風研究不多，主要為歐美地區的研究，這些研究多未以不同的腦中風型態分析。子計畫一的研究目的為探討懷孕相關的腦中風發生率與危險因子，並比較懷孕相關與懷孕不相關的年輕女性腦中風。於1984-2002年在台大醫院，共有402位年輕女性初發腦中風患者，包括49位懷孕相關的腦中風與353位懷孕不相關的腦中風患者，同時期共有49,796產次（30,232次自然產、18,341次剖腹產、1,223次墮胎）。49位懷孕相關的腦中風包括27位腦梗塞（16位為動脈腦梗塞，11位為腦靜脈栓塞）、19位腦內出血與3位蜘蛛膜下腔出血患者；而353位懷孕不相關的腦中風包括145位腦梗塞（138位為動脈腦梗塞，7位為腦靜脈栓塞）、147位腦內出血與61位蜘蛛膜下腔出血患者。懷孕相關的腦梗塞明顯較懷孕不相關的腦梗塞有較多的腦靜脈栓塞（39%比7%，P<0.001），血液凝固異常佔了67%，特別是S-蛋白缺乏致腦靜脈栓塞較多。懷孕相關的腦內出血明顯比懷孕不相關的腦內出血有較高比例的高血壓（包括子癲前症與子癲症）（37%比11%，P=0.01）。懷孕與產褥的腦中風發生率為十萬分之98.4（95%的信賴區間為十萬人分之74.4至130.2）；扣除中風後轉院病患，校正後的腦中風發生率為十萬分之46.2（95%的信賴區間為十萬分之30.7至69.5）。產褥期的腦中風發生率明顯高於其它懷孕時期（P<0.001）。45%的腦中風發生於產褥期，特別是因腦靜脈栓塞引起的中風（73%）。台灣的懷孕相關腦中風發生率不比歐美低，多數懷孕相關的腦中風發生於懷孕後期與產褥期，子癲前症與子癲症為腦內出血的主因，心因性栓塞為動脈腦梗塞的主因，血液凝固疾病為腦靜脈栓塞中風的主因，對於已有心臟疾病、子癲前症/子癲症病史與血液凝固疾病應特別預防於懷孕及產褥的腦中風發生。
因頸部的腫瘤而接受放射線治療會導致頸動脈的硬化狹窄，台灣為鼻咽癌的高發生區，放射線治療為主要的治療方式，因此以放射線治療鼻咽癌應注意頸動脈狹窄與腦中風的發生。計畫二的研究目的為探討初發腦梗塞中風患者，放射線治療是否為頸動脈狹窄的重要危險因子，那些危險因子會加重放射線治療後的腦梗塞中風患者的頸動脈狹窄。於1995至2004年，共有4,862位的初發腦梗塞中風成人患者，包括70位（佔1.4%，男性為69%，平均年齡為62.2+10.3歲）在腦中風發生之前曾接受過頸部放射治療，頸部腫瘤主要為鼻咽癌（76%），頸部放射治療至腦中風的平均發生時間為12.3+9.0年。曾接受過頸部放射治療的患者有較多的完全前腦循環腦梗塞（27.1%比14.9%）與部份前腦循環腦梗塞（38.6%比31.9%）。曾接受過頸部放射治療的患者有較高比例的頸動脈粥狀硬化（91.4%比78.3%，P=0.009）、嚴重頸動脈狹窄（44.3%比11.1%，P<0.001）、兩側嚴重頸動脈狹窄（14.3%比1.9%，P<0.001）。以多變項回歸分析，曾接受過頸部放射治療為頸動脈狹窄>50%（勝算比為9.16倍，95%的信賴區間為5.51-15.22，P<0.001）與頸動脈粥狀硬化斑指數>6的顯著相關因子（勝算比為11.89倍，95%的信賴區間為6.57-21.51，P<0.001）。分析70位初發腦梗塞中風曾接受過頸部放射線治療的患者，頸動脈狹窄>50%的顯著相關因子為男性（勝算比為3.196，P=0.05）與接受頸部放射治療至發生腦梗塞中風的時間超過10年（勝算比為6.06，P=0.016）。對於接受頭頸部放射線治療的患者，若已確定有嚴重的頸動脈狹窄，應依據現有的頸動脈狹窄處理指引進行。
對於嚴重的顱外內頸動脈狹窄，頸動脈內膜切除術（簡稱CEA）與頸動脈支架置放術（簡稱CAS）能預防腦梗塞的發生。過去頸動脈狹窄治療的臨床試驗均在歐美進行，因此應仔細評估CEA與CAS是否亦能有效預防國人因頸動脈狹窄引起的腦中風。子計畫三的研究目的為探討頸動脈狹窄患者的腦中風再發其影響因子、比較不同頸動脈狹窄治療對於長期存活之影響。1995-2004年於台大醫院的頸動脈狹窄登錄庫，有1,269位頸動脈狹窄>50%的患者（男性71%，平均年齡為69.6+11.0歲，有症狀頸動脈狹窄為58.3%）納入分析。接受CEA治療有76位，CAS治療有292位，內科治療有901位。個案的平均追蹤時間為3.09+2.62年。兩種介入治療的主要術中與術後30日的併發症均低，CEA在有症狀與無症狀頸動脈狹窄分別為7.1%與5%，CAS在有症狀與無症狀頸動脈狹窄分別為7.1%與3%。頸動脈患者接受介入治療後，頸動脈狹窄的每年復發率在CEA與CAS治療分別為2.0%與1.4%。分析不同治療型態的頸動脈狹窄病患的死因，對於有症狀頸動脈狹窄患者，最主要死因為腦中風，次為感染、癌症、其它血管疾病與心肌梗塞。對於無症狀頸動脈狹窄患者，死因主要為感染、癌症、其它血管疾病與心肌梗塞，腦中風反而不是致死的主因。症狀性與無症狀的頸動脈狹窄病患的累積存活分析，CEA與CAS組無明顯差異，但兩種介入治療明顯較內科治療的累積存活率高（P<0.001）。以Cox等比危害模式多因子分析症狀性頸動脈狹窄患者只接受內科治療之長期存活決定因子，年齡>65歲、糖尿病、缺血性心臟疾病、合併有腎臟疾病與癌症等5個為顯著決定因子。對於有症狀頸動脈狹窄患者，接受CEA與CAS的介入治療較只接受內科治療有較好的長期存活，CEA與CAS術後的頸動脈狹窄復發也不高。對於無症狀頸動脈狹窄患者，腦中風不是長期的主要死因，是否需接受CEA與CAS的治療應更嚴謹的判斷。
雖然以單一醫院疾病登錄的研究有先天限制，如病患選擇醫院的偏差、醫學中心有較多的難、重症，也無法如社區研究探討因果相關，但其研究所費較低，且長時期、大尺度的登錄有助廣泛的臨床與研究資料收集，特別有利於一些特殊疾病的研究，也有助於研究假說的形成。「台大醫院腦中風與頸動脈粥狀硬化研究」會持續進行，今後會以國人較多的腦血管疾病類型研究為主，如顱內動脈粥狀硬化狹窄、後腦循環疾病、腦內出血等。而在基因的時代，除了腦血管疾病基因研究的進行，進一步的基因治療（如幹細胞治療）是今後努力的方向。

關鍵字

腦血管疾病 ； 頸動脈狹窄 ； 懷孕相關的腦中風 ； 放射線治療 ； 存活分析 ； 流行病學

並列摘要

Cerebrovascular disease (CVD) was the 1st leading cause of death from 1963-1981, the 2nd leading cause of death from 1982-2003 in Taiwan. Although CVD became the 3rd cause of death in 2004, it still will prevail in the future decades for the ageing population, unchanged age-specific stroke incidence and low case-fatality of stroke in Taiwan.
Past Taiwan epidemiological studies on CVD were reviewed extensively. The review was divided into several parts, including vital statistic data of CVD, population-based incidence and prevalence study of stroke, hospital-based stroke registry studies, risk factor studies of stroke and carotid atherosclerosis, stroke clinical trials, genetic epidemiological stroke studies, and survival analytic studies of CVD. The past studies emphasized on conventional risk factors of CVD. Therefore, the present study focus on unconventional risk factors of CVD, including pregnancy related stroke, and post-irradiation carotid atherosclerotic disease. For increasing prevalence of carotid stenosis in Taiwan, long-term outcome and survival analysis of patients with cervical carotid artery stenosis was studied.
This study’s method mainly stemmed from “The Stroke and Cerebral Atherosclerosis Study in National Taiwan University Hospital” (SCAN) The objectives of SCAN are: (1) to study the risk factors, pathophysiological characteristics and prognosis in different types of stroke; (2) to study specific types of CVD, such as stroke related to pregnancy and puerperium, cerebral venous thrombosis (CVT), and post-irradiation vasculopathy, etc.; (3) to investigate the extent and severity of carotid atherosclerosis and its risk factors and the long-term outcome and survival of patients with cervical carotid artery stenosis. Both stroke registry and carotid stenosis registry were set up. The stroke registry in the SCAN project was closed to the standard of an optimal stroke data bank. There was good validity in stroke subtyping and carotid atherosclerotic quantification. Based on the SCAN structure, three substudies were present in this doctoral thesis: substudy-1, pregnancy-related stroke study; substudy-2, cervical radiation therapy (RT) as a significant risk factor of carotid atherosclerosis study; substudy-3, long-term outcome and survival analysis of patients with cervical carotid artery stenosis.
Pregnancy is a known risk factor for stroke, but relatively few studies have been conducted in Asian populations to document the risk. The substudy-1 aimed to analyze the incidence and etiologies of stroke occurring during pregnancy and puerperium, and to assess risk factors and etiologies of stroke in women of reproductive age based on relationship to pregnancy. From 1984 to 2002, all female patients 15 through 40 years of age with a first-ever stroke at National Taiwan University Hospital were included in this study. Pregnancy-related stroke (PRS) was defined as patients who had stroke occurrence during pregnancy or within 6 weeks postpartum. Stroke was classified as cerebral infarct (CI, including arterial infarct [AI] and CVT), cerebral hemorrhage (CH), and subarachnoid hemorrhage (SAH). Of 402 young female stroke patients, 49 had stroke during pregnancy and puerperium, including 16 with AI, 11 with CVT, 19 with CH and 3 with SAH. After excluding referral patients, the incidences of pregnancy-related stroke were 46.2 (95% confidence interval [CI], 30.7-69.5) per 105 pregnancies. CVT was more common in PRS than stroke unrelated to pregnancy (39% vs. 7%, P<0.001), and 73% of these cases occurred postpartum. Preeclampsia-eclampsia was an important cause of peripartum CH (37%), but not CI (4%). Eclampsia (37%) and arteriovenous malformation (26%) were the most important etiologies of CH. The incidence of stroke occurrence during pregnancy and puerperium in Taiwanese women is higher than that of Caucasian populations; the majority of strokes occurred in the third trimester and puerperium, particularly CVT. Among PRS cases, postpartum cerebral venous thrombosis and preeclampsia-eclampsia were the major causes of CI and CH, respectively.
Cervical carotid artery disease following cervical irradiation has been well documented. In Taiwan, the incidences of head and neck malignancies, especially nasopharyngeal carcinoma, are quite high and accounted for one of the leading causes of cancer. Therapeutic RT to the affected area has successfully increased survival time for afflicted patients and has created a pool of potential RT-induced vasculopathy. The substudy-2 aimed to analyze whether cervical irradiation therapy as an important risk factor for cervical carotid stenosis in patients with ischemic stroke. From 1995 to 2004, there were 4,862 adult patients with first-ever ischemic stroke, including 70 (1.4%, male were 69%, mean age was 62.2+10.3 years). The average duration from patients receiving RT to the onset of stroke was 12.3+9.0 years. Total anterior circulation infarct was higher in RT patientsthan non-RT patients (27.1% vs. 14.9%). RT patients had higher incidences than non-RT patients in carotid atherosclerosis (91.4% vs. 78.3%, P=0.009), carotid stenosis (44.3% vs. 11.1%, P<0.001) and bilateral carotid stenosis (14.3% vs. 1.9%, P<0.001). By multivariate logistic regression analysis, RT was a significant risk factor for carotid stenosis >50% (odds ratio [OR]: 9.16, 95% confidence interval [CI]: 5.51-15.22, P<0.001) and carotid plaque score >6 (OR: 11.89, 95% CI: 6.57-21.51, P<0.001). Of 70 ischemic stroke patient with RT, only male gender (OR: 3.196, P=0.05) and the duration from RT to stroke more than 10 years (OR: 6.06, P=0.016) were significant independent rrisk factors. For patients ever received cervical RT, periodical screening of cervical arteries is required.
There were several large clinical trials documenting the efficacy of interventional therapies (carotid endarterectomy [CEA] and carotid artery stenting [CAS]) in cervical carotid artery stenosis. However, all these studies were undergone in Caucasian countries, and no large studies was ever reported from Asian countries. The substudy-3 aimed to analyze the long-term outcome in patients with cervical carotid artery stenosis receiving different methods of treatment. From 1995 to 2004, a total of 1,269 patients (male, 71%; mean age, 69.6+11.0 years, asymptomatic patients, 58.3%) included. There were 76 patients receiving CEA, 292 receiving CAS, and 901 patients with medical therapy. The average duration of follow-up was 3.3+2.5 years. Major peri-procedural complications (including death, stroke, and acute myocardial infarction within 30days) were acceptable for both CEA (symptomatic, 7.1%, asymptomatic, 5%) and CAS (symptomatic, 7.1%, asymptomatic, 3%). The annual recurrent severe carotid stenosis for both CEA (2.0%) and CAS (1.4%) was low. There were significantly lower death rates for long-term follow-up of patients with interventional therapies than those with medical treatment (p<0.001). For symptomatic patients, the major cause of death during long-term follow-up was stroke, followed by infection, cancer and myocardial infarction. However, stroke was not the major cause of death in asymptomatic patients. By Cox proportional hazard model, age>65 years, diabetes, ischemic heart disease, renal disease and malignancy were unfoavorable long-term survival factors. CEA and CAS had lower stroke recurrence and had better long-term survival than medical therapy for symptomatic carotid stenosis patients. The treatment strategies for different types of carotid stenosis patients with different severity of carotid stenosis should be defined.
Single-hospital based registry study has potential limitations, including patient selction bias, more complicated cases in medical center, and inability to delineate causal- relationship as population-based study. However, it has advantages of low study cost, large-scaled clinical informations and data, particularly helpful for study on specific forms of CVD. The SCAN project will continue and highlight on intracranial atherosclerosis, posterior circulation stroke, and cerebral hemorrhage in the future. In the genetic era, genetic study on CVD, esp. gene therapy (such as stem cell therapy) in stroke will be emphasized.

並列關鍵字

cerebrovascular disease ； carotid stenosis ； epidemiology ； pregnancy related stroke ； radiation therapy ； survival analysis

參考文獻

Akinkugbe OO. (1990) Epidemiology of cardiovascular disease in developing countries. J Hypertension. 8(suppl 7):S233-S238.

Al-Mubarak N, Roubin GS, Iyer SS, Gomez CR, Liu MW, Vitek JJ. (2000) Carotid stenting for severe radiation-induced extracranial carotid artery occlusive disease. J Endovasc Ther.7:36-40.

Atkinson JD, Sundt TM, Dale AJD, Cascino TI, Nichols DA. (1989) Radiation-associated atheromatous disease of the cervical carotid artery: report of seven cases and review of the literature. Neurosurgery. 24:171-178.

Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. (1990) Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet. 337:1521-1526.

Barbour LA, Pickard J. (1995) Controversies in thromboembolic disease during pregnancy: a critical review. Obstet Gynecol. 86:621-633.