ACE-inhibitor therapy and nosocomial pneumonia

In the case-control study reported by Okaishi and colleagues (AJH 1999;12:778–783), elderly Japanese inpatients who had used angiotensin converting enzyme (ACE) inhibitors had a risk of nosocomial pneumonia that was less than half that of nonusers, after adjustment for known risk factors.1 What are we to make of this finding of a noncardiovascular benefit of ACE-inhibitor therapy? Okaishi and colleagues identified cases of hospital-acquired pneumonia and matched control subjects without pneumonia from among residents of a Japanese inpatient ward that serves both as a hospital and a long-term care facility. Use of antihypertensive medications was measured from a computerized pharmacy database, and risk factors for pneumonia such as dementia, bedridden state, previous stroke, asthma, other lung diseases, and hypoalbuminuria, were assessed by physician examination and from medical records. Patients who were currently using ACE inhibitors had a 62% lower risk of nosocomial pneumonia than nonusers (odds ratio (OR) = 0.38, 95% confidence interval 0.15–0.97), and this association persisted after adjustment for multiple risk factors. This study supports the hypothesis recently proposed by several investigators that the ACE inhibitors, through the same mechanism that produces the well-known dry, chronic cough,2 may prevent silent aspiration of oropharyngeal pathogens into the lower airway in patients with depressed cough reflex, and thus confer protection against pneumonia.3,4,5,6 Two previous epidemiological studies have shown an association between ACE-inhibitor use and reduced risk of pneumonia. Sekizawa and colleagues studied 468 residents of long-term care facilities who had previously suffered a stroke.3 During 2 years of follow-up, the occurrence of pneumonia in users of ACE inhibitors was approximately half that in nonusers (7% v 18%). In a study of 576 men and women aged 65 or over conducted by Arai and colleagues,4 the 3-year incidence of pneumonia was 3.3% among users of ACE inhibitors, compared with 8.9% among calcium channel blocker users, and 8.3% among subjects not using antihypertensive therapy. Unlike Okaishi and colleagues,1 neither of these previous groups of investigators controlled for other risk factors that may have produced a spurious association between the use of ACE inhibitors and risk of pneumonia. Okaishi and colleagues also found a positive association between calcium channel blocker use and risk of nosocomial pneumonia (OR = 1.84, 95% confidence interval 0.89–3.78). While this finding is compatible with chance, it could also reflect an association between treated hypertension and incidence of pneumonia. If so, then the odds ratio of 0.38 for ACE-inhibitor use would underestimate the magnitude of the true association among treated hypertensives. The results for calcium channel blocker users are also reassuring because they suggest that the association between ACE-inhibitor use and reduced risk of pneumonia is not explained by other characteristics that tend to be present in patients who are compliant with chronic drug therapy. While the randomized clinical trial is properly regarded as the gold standard for evaluating medical therapies, observational studies such as this one play an indispensable role as well.7 Observational studies are particularly useful for identifying unexpected drug effects. Until now, even the largest and most well-designed clinical trial of ACE-inhibitor therapy for treatment of hypertension would have been unlikely to include incident pneumonia as a prespecified endpoint. Furthermore, patients like the ones in this study who reside in long-term care facilities are not usually enrolled in clinical trials, and this patient population is precisely the one that might be expected to derive benefit from ACE-inhibitor therapy on risk of pneumonia. This study calls attention to the fact that the outcome of any medical intervention is the sum of all its beneficial and harmful effects. Here is perhaps an example of an unintended benefit of ACE–inhibitor therapy that, in some populations, may be as important as the intended benefit, which is to reduce cardiovascular morbidity and mortality. The results of the Systolic Hypertension in the Elderly Program (SHEP) trial, for instance, suggest that providing antihypertensive treatment to 1,000 patients 60 years or older would prevent about 55 major cardiovascular events over 5 years.8 Over the same period of time, 152 patients in a nursing home population of 1,000 might require hospitalization for pneumonia.9 If ACE inhibitors were to prevent 62% of these 152 episodes of hospitalized nosocomial pneumonia, it is possible that in a population treated with ACE inhibitors, as many patients might be spared serious illness or death from pneumonia as from cardiovascular disease. The observational studies conducted to date suggest the possibility that the use of ACE inhibitors to treat hypertension may reduce the risk of pneumonia in older adults. The results of this study should not affect clinical practice at this time, although clearly they have raised an intriguing hypothesis. The implications of this study will not be known until further research not only confirms that ACE inhibitors are associated with a reduced risk of pneumonia, but also clarifies what the magnitude of the benefit may be. Future or ongoing clinical trials of ACE-inhibitor therapy, particularly those such as Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)10 that are conducted in older patients, should perhaps collect data on the occurrence of pneumonia. This provocative study reminds us that we should routinely anticipate that drug therapies will have unintended effects in addition to intended ones, and that the effects that are unintended may be surprisingly large in their potential health impact.