Post-Stroke Pneumonia Prevention by Angiotensin-Converting Enzyme Inhibitors Results

Adv Ther (2012) 29(10):900–912.DOI 10.1007/s12325-012-0049-1Post-Stroke Pneumonia Prevention by Angiotensin-Converting Enzyme Inhibitors: Results of a Meta-analysis of Five Studies in AsiansYukito Shinohara ∙ Hideki OrigasaTo view enhanced content go to Received: July 5, 2012 / Published online: September 14, 2012© Springer Healthcare 2012ABSTRACTIntroduction: Angiotensin-converting enzyme inhibitors (ACEIs) are reported to reduce the incidence of aspiration pneumonia in hypertensive patients. In this study, a meta-analysis was conducted to obtain statistically more reliable estimates of outcome. Methods: The MEDLINE and JMEDICINE databases were searched and the following study selection criteria were applied: (1) comparative controlled studies identified with the following keywords: drug therapy, ACEI, hypertension, swallowing function, dysphagia, stroke, and pneumonia; (2) a minimum follow-up period of 6 months; and (3) a minimum number of patients of more than 100. Patients with hypertension and a history of stroke or transient ischemic attack (TIA) in five controlled studies that reported the incidence of pneumonia were included in the analysis. Results: A total of 8,693 post-stroke patients were given ACEIs with another antihypertensive agent or placebo as a control. In all studies, ACEIs, particularly imidapril, exhibited preventive effects equating to a relative risk that ranged from 0.32 to 0.81 compared with controls. In the combined studies the overall relative risk of ACEI-treated patients versus controls was 0.61 (95% confidence intervals [CI] 0.51–0.75; P < 0.001). Among Asian patients, the relative risk was 0.42 (95% CI 0.32–0.56; P < 0.001). Among Japanese patients, an even greater preventive effect was found for ACEIs versus other antihypertensives (relative risk: 0.38 [95% CI 0.27–0.54; P < 0.001]).Conclusion: ACEIs appear to be more effective than other antihypertensive agents or placebo in reducing pneumonia risk in post-stroke patients, especially in Asian populations.Y. Shinohara (*)Department of Neurology, Federation of National Personnel Mutual Aid Associations Tachikawa Hospital, 4-2-22 Nishikicho, Tachikawa, Tokyo, 190-8531, Japan e-mail: yshinoha@tachikawa-hosp.gr.jpH. OrigasaBiostatistics and Clinical Epidemiology, Universityof Toyama Graduate School of Medicine and Pharmaceutical Sciences, Toyama, Toyama, JapanEnhanced content for Advances in Therapyarticles is available on the journal web site:(ACEIs) [10–13]. H owever, these reports all focused on Japanese patients. ACEIs (imidapril, enalapril, and captopril) have been shown to reduce pneumonia risk in patients with previous stroke by a factor of approximately three [11]. In the Perindopril pROtection aGainst REcurrent Stroke Study (PROGRESS), pneumonia occurred in 261 of 6,105 patients and treatment with the ACEI perindopril reduced pneumonia risk by 19% compared with placebo [12]. This preventive effect was significant in the restricted cohort of Asian origin (Chinese and Japanese), but not in the non-Asian cohort. Although these findings are consistent with observational studies and are not reported with angiotensin receptor blockers (ARBs), the evidence is still insufficient to conclude that ACEIs should be routinely used for pneumonia prevention in stroke patients.The authors conducted a meta-analysis on the preventive effects of ACEIs on post-stroke pneumonia, defined as newly developed pneumonia following stroke onset caused by either apparent aspiration or dysphagia-associated micro-aspiration [14], in patients with a history of stroke to obtain statistically more reliable estimates of outcome, by avoiding overestimation of results from individual studies while minimizing variability due to chance fluctuations in the data.MATERIALS AND METHODSData SelectionUsing drug therapy, ACEI, hypertension, swallowing function, dysphagia, stroke, and pneumonia as keywords, the authors identified 21 clinical studies from the MEDLINE and JMEDICINE databases between January 1990 and December 2011. Studies with an unknown follow-up period were excluded. A further selection criterion, that the follow-up periodKeywords: Angiotensin-converting enzyme inhibitor; Aspiration pneumonia; Imidapril; Prevention; Post-stroke; Meta-analysisINTRODUCTIONPneumonia is one of the most common causes of death among the elderly in the US, Europe, and Japan. The dominant form of pneumonia in the elderly is aspiration pneumonia (AP), which involves inhalation of oropharyngeal secretions colonized by pathogenic bacteria in the lower respiratory tract [1], and dysphagia with aspiration of gastric contents due to gastro-esophageal reflux [2]. Individuals who have suffered from stroke, and who have Parkinson’s disease or dementia, are at a particularly high risk of AP [3], with stroke being the most common underlying disease [4]. AP is considered to be the leading cause of death in chronic-phase stroke patients. For example, in a retrospective study, pneumonia occurred in 5.6% of stroke patients (n = 11,289) from 29 hospitals in Cleveland, Ohio, USA. Mortality within 30 days of the onset of stroke was significantly higher in patients with pneumonia (26.9%) than in those without (4.4%); complicated pneumonia resulted in a threefold increase in deaths within 30 days of onset of stroke [5]. In a study that investigated post-discharge mortality and cause of death in 10,981 patients with ischemic stroke and transient ischemic attack (TIA), there was a 1-year cumulative mortality of 7.0% in 10,234 ischemic stroke patients and 3.5% in 747 TIA patients, with pneumonia listed as the cause of death in 22.6% of cases [6].Many strategies have been proposed to prevent AP , including dietary interventions, compensatory strategy/positioning changes, oral hygiene, tube feeding, and pharmacological therapies [3] such as amantadine [7], cilostazol [8, 9], or angiotensin-converting enzyme inhibitorsusing a meta-analytic method versus other antihypertensive agents or placebo as controls. The magnitude of the effect was expressed as the relative risk of the occurrence of pneumonia during the study follow-up. Although these criteria of diagnosis for pneumonia were not uniform in all studies, a meta-analysis gives sufficient information because each study usedmust be equal to or longer than 6 months, reduced the number of studies to 11. An additional limitation to comparative controlled studies left seven studies for the final analysis: Sekizawa et al. (1998) [11], Arai et al. (1998) [13], Arai et al. (2000) [15], Arai et al. (2001) [16], Akutagawa et al. (2002) [17], Ohkubo et al. (2004) [12] and Arai et al. (2005) [18], all of which were conducted in Japan. Of these studies, five [11, 12, 15, 16, 18] dealt exclusively with more than 100 post-stroke patients and they were used for the meta-analysis. The selection process for the analysis is illustrated in Fig. 1.Data ExtractionData collected from each article included the first author’s name, year of publication, journal name, total sample size, treatment arms (ACEI and control), duration of follow-up, average age, gender ratio, and whether the study was prospective or retrospective; it also included whether the population was elderly (≥65 years old) or bedridden, and information regarding concomitant drugs. The numbers of incident cases, and controls for pneumonia were extracted from appropriate tables in each publication and arranged into data and analysis tables for each study.EndpointThe study endpoint was defined as the incidence of pneumonia, without regard to confirmation of swallowing reflex during the study follow-up. The diagnosis of pneumonia was based on clinical symptoms, such as fever and cough, followed by C-reactive protein (CRP) elevation, and/or analysis of x-ray film images. Any preventive effect by ACEIs on pneumonia potentially induced by attenuation of the swallowing reflex was evaluatedFig. 1 Literature search process. ACEIs angiotensin-converting enzyme inhibitorspossible and optimal diagnostic criteria adopted at each time point. The number needed to treat (NNT) was calculated by the reciprocal of the absolute risk reduction (ARR) and the 95% confidence interval (CI) was calculated from the interval estimation of ARR using the Wilson score method [19]. Regarding the 1-year follow-up of the studies, the event rate was calculated after the number of events was divided by the duration of the study follow-up.Statistical AnalysisStatistical analyses were performed according to the types of data extracted from each report. Relative risks for preventive effects in each study were computed with the Mantel-Haenszel method. Although average follow-up periods varied from 2 to 4 years, the use of relative risk enabled the magnitude of the effect to be unified for these different studies. The risk ratio was computed as an index of the relative risk of the particular ACEI versus other antihypertensive agents or placebo in the prevention of pneumonia. A fixed effects model was used to estimate the pooled relative risk and its 95% CI, in accordance with the Mantel-Haenszel model.H eterogeneity between trials was examined using chi-square test. Three subgroup analyses were conducted: the first involved the restricted inclusion of an Asian population, the second included all Japanese patients in studies that used active treatment as a control, and the third included Japanese patients in whom mainly calcium channel blockers (CCBs) were used as antihypertensive agents for a control group. A two-sided P-value of less than 0.05 was considered to be statistically significant. Statistical analyses were performed using Review Manager software version 5.1.0 (Nordic Cochrane Centre, Cochrane Collaboration 2011, Copenhagen, Denmark).RESULTSAmong the total of 8,693 patients from the five cohort studies included in the meta-analysis, 4,940 were Asians and 3,753 were non-Asians (Australian and European), and a total of 473 pneumonia cases were diagnosed. As shown in Table 1 [11, 12, 15, 16, 18], the sample size varied from 394 to 6,105 post-stroke hypertensive patients, with an average of 1,739 patients per study. In the control arms, other antihypertensive agents, such as CCBs, ARBs, beta-blockers (BBs), and others, were included. The study by Ohkubo et al. [12] was an exception, with placebo as the control. Study follow-up varied between 2 and 4 years. Two studies were retrospective [15, 16], and the others were prospective [11, 12, 18].Although the description of patients’ characteristics was limited in all articles, three studies reported average ages between 76 and 82 years, the gender ratio was equally distributed, and the majority of patients were not bedridden (except in the Arai [2000] [15] study involving 91 bedridden patients). Information regarding concomitant drugs other than ACEIs and comparator drugs was not complete in most papers.Figure 2a [11, 12, 15, 16, 18] shows that the incidence of pneumonia in the control groups ranged from 4.7% to 17.9%, with an average of 6.8%. All cohort studies revealed preventive effects of ACEIs compared with the control (relative risk range: 0.32–0.81). The overall combined relative risk following administration of ACEIs compared with controls in these five studies combined was estimated as 0.61 (95% CI 0.51–0.75; P < 0.001), based on 8,693 patients (Fig. 2a). When the analyses were restricted to the Asian population in the five studies, the relative risk was estimated as 0.42 (95% CI 0.32–0.56; P < 0.001; Fig. 2b) [11, 12, 15, 16, 18].T a b l e 1 S u m m a r y o f b a c k g r o u n d c h a r a c t e r i s t i c s o f t h e fi v e r e l e v a n t c o h o r t s t u d i e sF i r s t a u t h o rP a t i e n tN o . o f p a t i e n t s N o . o f p a t i e n t s i n e a c h a r mA v e r a g e f o l l o w -u pP /R A g e (y e a r s )G e n d e r m a l e (%)E l d e r l y (≥65 y e a r s )P o s t -s t r o k e B e d r i d d e n C o n c o m i t a n t d r u g sA C E IC o n t r o lS e k i z a w a (1998)[11]H y p e r t e n s i o n w / p a s t h i s t o r y o f s t r o k e440A C E I : 127I m i d a p r i l E n a r a p r i l C a p t o p r i lC C B o r B B :3132 y e a r s PA C E I g r o u p : m e a n 76C CB g r o u p : m e a n 77N A Y e s Y e s N o a N AA r a i (2000)[15]H y p e r t e n s i o n w / p a s t h i s t o r y o f s t r o k e394A C E I : 208N A C C B : 1864 y e a r s R≥65N A Y e s Y e s Y e s (91 p t s )N AA r a i (2001)[16]H y p e r t e n s i o n w /p a s t h i s t o r y o f s t r o k e404A C E I : 209N A A R B : 1952 y e a r s R N A 48Y e s Y e sN o N AO h k u b o (2004)[12]H i s t o r y o f s t r o k e a n d /o r T I AA C E I 4 m g /d a y :P e r i n d o p r i lP l a c e b o :A l l (A s i a n )6,1052,3523,0511,1763,0541,1763.9 y e a r s 3.9 y e a r s P P M e a n 64N A70N AN o N oY e s Y e sN o N oI n d a p a m i d e I n d a p a m i d eA r a i (2005)[18]H y p e r t e n s i o n w / p a s t h i s t o r y o f s t r o k e1,350A C E I : 430I m i d a p r i l C C B + o t h e r s :92035 m o n t h s PA c t i v e : m e a n 75C o n t r o l : m e a n 76N A Y e sY e sN oN AA C E I a n g i o t e n s i n -c o n v e r t i n g e n z y m e i n h i b i t o r ,B B b e t a -b l o c k e r ,C C B c a l c i u m c h a n n e l b l o c k e r , P /R p r o s p e c t i v e /r e t r o s p e c t i v e c o h o r t s t u d i e s , P t s p a t i e n t s , T I A t r a n s i e n t i s c h e m i c a t t a c k , N A d a t a n o t a v a i l a b l e a B e d r i d d e n p a t i e n t s a l s o i n c l u d e dNo significant heterogeneity, with respect to the preventive effect of ACEIs, was observed among these studies (P = 0.75). In four studies, which enrolled Japanese patients only, the preventive effect of ACEIs compared with other hypertensive drugs was even greater, with a relative risk of 0.38 (95% CI 0.27–0.54; P < 0.001; Fig. 2c) [11, 15, 16, 18]. No significant heterogeneity, with respect to the preventive effect of ACEIs, was observed among these studies (P = 0.90). In three studies, which enrolled Japanese patients to whom CCBs were used as antihypertensive agents, the preventive effect of ACEIs compared with CCBs was even greater, with a relative risk of 0.37 (95% CI 0.25–0.55; P < 0.001; Fig. 2d) [11, 15, 18].Study or subgroup Events T otal Events T otal Weight (%)M-H, fixed, 95% CI Y ear ACEIControlRelative risk M-H, fixed, 95% CIRelative risk a)Study or subgroup Events T otal Events T otal Weight (%)M-H, fixed, 95% CI Y ear ACEIControlRelative risk M-H, fixed, 95% CIRelative risk b)Fig. 2 Relative risk of angiotensin-converting enzyme inhibitors (ACEIs) compared with other treatments or placebo, (a ) based on a total of all 8,693 post-stroke hypertensive patients in the five studies, (b ) based on a total of 4,940 post-stroke hypertensive Asian patients in the five studies, (c ) based on a total of 2,588 Japanese patients in four studies. (d ) Relative risk of ACEIs compared with CCBs, based on a total of 2,184 Japanese patients in three studies in whom mainly CCBs were used as a control group. ACEI angiotensin-converting enzyme inhibitor; CCB calcium channel blocker; CI confidence interval; M-H Mantel-Haenszel modelJapanese patients with CCBs as the antihypertensive agents (D), the estimated NNT was 14.7 (95% CI 11.3–21.8). When evaluating these NNT values, consideration should be applied because the NNT value fluctuates according to the study follow-up year related to the event rate; so, NNT values based on 1-year follow-up were calculated to compare each study.Regarding the 1-year follow-up, the estimated NNT per year of the total number of patients (8,693) from the five cohort studies (A in Table 2b) was 87.2 (95% CI 59.1–164.6), ranging from 18.5 to 443.0 in individual studies. Restricted to the Asian population only in these fiveNo significant heterogeneity, with respect to the preventive effect of ACEIs, is observed among these studies (P = 0.75).Adopting the data directly from Fig. 2, the estimated NNT of the total number of patients (8,693) from the five cohort studies (A in Table 2a) [11, 12, 15, 16, 18] was 34.3 (95% CI 25.9–50.6), ranging from 9.3 to 113.6 in the studies. When restricting the study population to the Asian population only in these five studies (B), the estimated NNT was 20.4 (95% CI 16.3–27.4). In the four studies that enrolled Japanese patients only (C), the estimated NNT was 14.9 (95% CI 11.6–21.2). In three studies that enrolledStudy or subgroup Events T otal Events T otal Weight (%)M-H, fixed, 95% CI Y ear ACEIControlRelative risk M-H, fixed, 95% CIRelative risk c)Study or subgroup Events T otal Events T otal Weight (%)M-H, fixed, 95% CI Y ear ACEIControlRelative risk M-H, fixed, 95% CIRelative risk d)Fig. 2 continuedTable 2a Calculated number needed to treat (NNT) of the ACEI (see Fig. 2)Figure Study or subgroup Data used for calculation Calculated value Y earACEI Control ER ARR NNT (95% CI a)Event T otal Event T otal ACEI Control9127563130.07090.17890.10809.3(6.1–26.2)2a1: Sekizawa(1998) [11]2: Arai (2000) [15]8208151860.03850.08060.042223.7(10.7–195.3)4 3: Arai (2001) [16]10209231950.04780.11790.070114.3(7.9–61.2) 2.41173,0511443,0540.03830.04720.0088113.6(52.5–729.7) 3.9 4: Ohkubo (2004)all [12]5: Arai (2005) [18]12430799200.02790.08590.058017.3(12.3–31.3) 2.9 T otal (1, 2, 3, 4, 5)1564,0253174,6680.03880.06790.029234.3(25.9–50.6)(2–4)261,176481,1760.02210.04080.018753.5(30.0–218.5) 3.9 b4: Ohkubo (2004)Asian [12]T otal (1, 2, 3, 4, 5)652,1502212,7900.03020.07920.049020.4(16.3–27.4)(2–4)c T otal (1, 2, 3, 5)399741731,6140.04000.10720.067114.9(11.6–21.2)(2–4)d T otal (1, 2, 5)297651501,4190.03790.10570.067814.7(11.3–21.8)(2–4)Table 2b Calculated number needed to treat (NNT) of the ACEI follow-up data (see Fig. 2)Figure Study or subgroup Data used for calculation Calculated valueACEI Control Y ear Event rate (ER)ARR NNT (95% CI a)Event T otal Event T otal ACEI Control91275631320.03540.08950.054018.5(10.4–922.6) a1: Sekizawa (1998)[11]2: Arai (2000) [15]82081518640.00960.02020.010594.8(23.1–58.1) 3: Arai (2001) [16]1020923195 2.40.01990.04910.029234.2(14.1–128.0) 4: Ohkubo (2004)1173,0511443,054 3.90.00980.01210.0023443.0(130.6–326.2) all [12]5: Arai (2005) [18]1243079920 2.90.00960.02960.020050.0(29.2–352.2) T otal (1, 2, 3, 4, 5)1564,0253174,668-0.01110.02260.011587.2(59.1–164.6)261,176481,176 3.90.00570.01050.0048208.5(77.1–355.8) b4: Ohkubo (2004)Asian [12]T otal (1, 2, 3, 4, 5)652,1502212,790-0.01000.02900.019052.6(37.4–87.9)c T otal (1, 2, 3, 5)399741731,614-0.01520.04250.027336.6(25.1–70.9)d T otal (1, 2, 5)297651501,419-0.01390.04160.027736.1(24.5–75.3) ACEI angiotensin-converting enzyme inhibitor, ARR absolute risk reduction, CI confidence interval, ER event rate, Year study follow-up year from each referencea Underlined NNT 95% CI calculated as NNH (number needed to harm)studies (B), the estimated NNT was 52.6 (95% CI 37.4–87.9). In the four studies that enrolled Japanese patients only (C), the estimated NNT was 36.6 (95% CI 25.1–70.9). In three studies that enrolled Japanese patients with CCBs as the antihypertensive agent (D), the estimated NNT was 36.1 (95% CI 24.5–75.3).DISCUSSIONPneumonia, which is common in elderly patients and those with severe stroke [20], usually results from oropharyngeal dysphagia with secondary aspiration. It has been reported that dysphagia is present in 64–90% of conscious stroke patients in the acute phase, with aspiration confirmed in 22–42% of cases [21]. Individual observational studies involving elderly Japanese populations indicated that there was a reduced risk of pneumonia in stroke patients given ACEIs compared with those given other antihypertensive agents [11, 13, 15, 18]. However, uncontrolled, randomized studies are susceptible to bias due to confounding factors, and may result in unreliable outcomes [22]. The results from the controlled clinical trial PROGRESS showed a tendency toward a reduced risk of pneumonia in patients receiving ACEIs compared with those receiving placebo, but the difference was significant only in the Asian (Chinese and Japanese) subpopulation, and not in non-Asian participants [12]. In the present meta-analysis of five studies, there was a significant decrease in the overall relative risk of pneumonia in patients taking ACEIs compared with those taking other antihypertensive agents or placebo. When the analysis was limited to the Asian population, the effect was greater, and it was even further enhanced in Japanese patients. As the majority of patients in the overall meta-analysis populations were of Asian ethnicity, no conclusions can be drawn regarding the prevention of pneumonia by ACEIs in non-Asian populations. H owever, in a retrospective study conducted in the UK, a significant reduction in pneumonia risk was reported by ACEIs in patients with diabetes, supposedly for the first time in non-Asian population [23], which suggests that the results obtained in the present study is potentially applicable to non-Asian populations.The molecular mechanisms by which ACEIs reduce the risk of pneumonia in stroke patients require further clarification. It is well-established that patients with pneumonia, particularly AP have reduced cough and swallowing reflexes, which are the primary defense mechanisms against aspiration. The established induction of cough [24] and swallowing [25] reflexes by ACEIs may provide protection against infection. Mechanisms associated with the swallowing and cough reflexes are complicated, but it is known that the final pathway involves production of substance P (SP) by dopamine stimulation of the nigrostriatum. When brain function is reduced by stroke or other causes, reduced dopamine metabolism [26] may decrease SP in the glossopharyngeal and vagal sensory nerves [27], as evidenced by depletion of SP in patients with AP [28]. The enzyme ACE, as well as catalyzing the conversion of angiotensin I to angiotensin II, is also involved in the degradation and inactivation of SP. ACEIs potentiate this action [29, 30] resulting in an accumulation of SP in the upper respiratory tract, thereby enhancing the sensitivity of the cough [31] and swallowing reflexes [25]. For example, administration of the ACEI, imidapril, resulted in a significant increase in serum SP concentrations and eliminated silent aspiration in elderly patients with stroke [10]. Thus, the effects of ACEIs on the mechanisms of cough and swallowing reflexes may play a role in the prevention of AP.In addition to the effect on swallowing reflexes, ACEIs are recently reported as the first and most successful class of mitigators of lung injury [32], suggesting the association of ACEIs with the mitigation of acute respiratory distress syndrome (ARDS) type diffuse lung damage. However, the effect of ACEIs on pneumonia itself has not been proven to require further examination.It is not clear why ACEIs appear to be more effective in reducing pneumonia in Asian compared with non-Asian populations. Some clinical observations indicate that there is a particularly high prevalence of dry coughs associated with ACEIs in Chinese subjects [33–35]. In contrast, another report states that treatment with ACEIs does not necessarily increase the frequency of coughs in some of the Asian population [36].Since onset of AP involves not only a reduced cough reflex, but also a reduced swallowing reflex, the reason for the apparent increased effectiveness of ACEIs in reducing the onset of pneumonia in stroke patients in Asian populations compared with non-Asian populations requires further examination.An epidemiologic survey assessing the importance of ACEI in the prevention of pneumonia in the elderly is currently being conducted [37]. In this survey the risk of developing pneumonia was assessed according to ACE insertion/deletion (I/D) polymorphisms.A significantly higher risk was observed in a group of Japanese inpatients aged 65 or over with the DD polymorphism, with high ACE activity, compared with the group with the II polymorphism, with low enzyme activity. This suggests that the ACE D allele is an independent risk factor for pneumonia in elderly patients. Knowledge of the distribution of these alleles in Asian and non-Asian populations could help explain the apparent difference in responses to ACEIs between these two populations. The results of the current meta-analysis support the clinical usefulness of ACEIs in the prevention of pneumonia in stroke patients, particularly those of Asian ethnicity. In addition, NNT analysis also supported the findings that ACEIs might be more effective than other traditional antihypertensive agents or placebo in reducing pneumonia risk in post-stroke patients, especially in Asian populations. Even in cases of useful intervention, large NNT values can be found, where few patients demonstrate the effect in a large population. In the present study, the estimated NNT values were 15–35 at event rates for control patients, 0.07–0.11 at event rates for whole study cases (Table 2a), and 36–87 at event rates for control patients during the follow-up period, and around 0.02–0.04 at event event rates for study cases during the 1-year follow-up period (Table 2a). Taking these event rates into consideration, these NNT values were considered to be beneficial. Nevertheless, problems can arise with ACEI usage. It is important to ascertain whether risk reduction can be obtained at ACEI dose ranges that do not cause decreases in blood pressure. Additionally, cough, frequently associated with the use of ACEIs, is often resistant to usual treatments and can be sufficiently troublesome to necessitate discontinuation of ACEI therapy. In the observational studies included in the present meta-analysis, the most commonly used ACEI was imidapril. In a double-blind comparative study, this drug was associated with an extremely low frequency of cough compared with the control, enalapril (0.9% vs. 7.0%); a much lower level than with other ACEIs [38]. Thus, it would seem that imidapril has an advantage over other ACEIs in clinical usefulness for reduction of pneumonia risk.Other agents with different mechanisms of action from ACEIs, not associated with cough, have also been shown to reduce the incidence of pneumonia in stroke or to improve theswallowing reflex. For example, amantadine, a drug that acts by releasing dopamine from dopaminergic nerve terminals, reduced the relative risk of pneumonia in patients with cerebral infarction [8]. Amantadine may suppress gastroesophageal reflex, thereby reducing pneumonia risk. It has been reported that other agents, such as capsaicin [39] and cilostazol [8, 9],stimulate the swallowing reflex. Cilostazol, an antiplatelet agent, has also been shown to suppress the onset of pneumonia in stroke patients [9]. These findings suggest that improvement of the swallowing reflex, rather than the cough reflex, may play an important role in reduction of pneumonia. Further comparative studies on ACEIs and other agents are required to establish the extent of their effects on the reduction of pneumonia in stroke patients in both the Asian and non-Asian communities, as well as to understand associated adverse reactions.In conclusion, it is of prime importance to avoid the onset of stroke, which leads to the attenuation or breakdown of protective mechanisms involving cough and swallowing reflexes, thereby triggering pneumonia. Antihypertensive agents have principally attracted attention for their efficacy in reducing cardiovascular events, but ACEIs are also considered to be useful drugs for reducing the risk of pneumonia following stroke. This effect of ACEIs is more marked in Asian populations. Future studies on risk reduction of pneumonia in stroke patients should involve intervention with various drugs known to affect cough and swallowing reflexes, and a comprehensive assessment of their mechanisms of action.ACKNOWLEDGMENTSThe authors thank the late Prof. K. Sekizawa for his valuable advice and careful check-up of the cases to avoid overlap of the subjects analyzed in this study. Dr. Shinohara is the guarantor for this article, and takes responsibility for the integrity of the work as a whole.Conflict of Interest. The authors declare no conflict of interest.REFERENCES1. Marik PE. Aspiration pneumonitis and aspirationpneumonia. N Engl J Med. 2001;344:665–71.2. Pellegrini CA, DeMeester TR, Johnson LF, SkinnerDB. Gastroesophageal reflux and pulmonary aspiration: incidence, functional abnormality andresults of surgical therapy. Surgery.1979;86:110 –9. 3. Loeb MB, Becker M, Eady A, Walker-Dilks C.Interventions to prevent aspiration pneumonia inolder adults: a systematic review.J Am Geriatr Soc.2003;51:1018–22.4. Doggett DL, Tappe KA, Mitchell MD, Chapell R,Coates V, Turkelson CM. Prevention of pneumoniain elderly stroke patients by systematic diagnosisand treatment of dysphagia: an evidence-based comprehensive analysis of the literature. Dysphagia.2001;16:279–95.5. Katzan IL, Cebul RD, Husak SH, Dawson NV, BakerDW. The effect of pneumonia on mortality amongpatients hospitalized for acute stroke. Neurology.2003;60:620–5.6. Kimura K, Minematsu K, Kazui S, YamaguchiT; the Japan Multicenter Stroke Investigators’ Collaboration (J-MUSIC). Mortality and cause ofdeath after hospital discharge in 10,981 patientswith ischemic stroke and transient ischemic attack.Cerebrovasc Dis. 2005;19:171–8.7. Nakagawa T, Wada H, Sekizawa K, Arai H,Sasaki H. Amantadine and pneumonia. Lancet.1999;353:1157.8. Yamaya M, Yanai M, Ohrui T, Arai H, Sekizawa K,Sasaki H. Antithrombotic therapy for prevention ofpneumonia. J Am Geriatr Soc. 2001;49:687–8.9. Shinohara Y; the CSPS Group. Antiplateletcilostazol is effective in the prevention of pneumonia in ischemic stroke patients in thechronic stage. Cerebrovasc Dis. 2006;22:57–60.10. Arai T, Yoshimi N, Fujiwara H, Fujiwara H,Sekizawa K. Serum substance P concentrations andsilent aspiration in elderly patients with stroke.Neurology. 2003;61:1625–6.。