Cough and Angioedema From Angiotensin-Converting Enzyme Inhibitors: New Insights Into Mechanisms and Management

Mark S. Dykewicz


Curr Opin Allergy Clin Immunol. 2004;4(4) 

In This Article

Cough and Respiratory Reactions

Although estimates vary, 20% or more of patients who receive ACE inhibitors develop a dry cough, sometimes severe enough to require discontinuation of the drug. With few exceptions, ACE inhibitor cough is an apparent class effect that may occur with all ACE inhibitors. Although resolution of the cough after withdrawal of ACE inhibitors may occur within days, it may require weeks. ACE inhibitor cough may be associated with upper respiratory symptoms of rhinitis, and it is probably underappreciated that these drugs may lead to rhinitis. Women, individuals with ACE genotype II, and those of black or Asian ethnicity have been reported to be at increased risk of ACE inhibitor cough.[1,2]

Mechanisms of Cough From Angiotensin-Converting Enzyme Inhibitors

The inhibition of ACE (also known as kininase II and bradykinin dehydrogenase) prevents the conversion of angiotension I to angiotensin II, with consequent salutary benefits via the renin-angiotensin system in pathological states. ACE inhibitor cough is thought to be linked to the suppression of ACE, which is proposed to result in an accumulation of substances normally metabolized by ACE: bradykinin or tachykinins (with the consequent stimulation of vagal afferent nerve fibers) and substance P. However, the development of cough may result from a more complex cascade of events than originally believed. Bradykinin has been shown to induce the production of arachidonic acid metabolites and nitric oxide (NO), and there is some evidence that these products, which are subject to regulation by other pathways, may promote cough through proinflammatory mechanisms.[3]

In a study to determine how cough could be promoted by increased prostaglandin synthesis that might follow stimulation of the arachidonic acid cascade by ACE inhibition, Malini et al.[4] studied nine patients with essential hypertension and a history of cough from enalapril. While continuing enalapril, patients were entered into a double-blind crossover study in which they received placebo or picotamide, a platelet antiaggregant that acts through both the inhibition of thromboxane synthase and thromboxane-receptor antagonism. Because picotamide does not inhibit cyclooxygense, prostaglandin production is not directly altered by the drug. As expected, urinary 11-dehydro-thromboxane-B2 was significantly reduced by picotamide; this was associated with the disappearance of cough in eight patients within 72 h. However, controls and coughers had similar levels of 11-dehydro-thromboxane-B2 excretion at baseline, and a rechallenge of coughers with enalapril did not significantly change their 11-dehydro-thromboxane-B2 excretion. The coughers did differ from controls in that the latter showed significantly lower excretion of 6-keto-prostaglandin F1 alpha, and their ratio of 11-dehydro-thromboxane-B2 to 6-keto-prostaglandin F1 alpha was twice that of the controls. The authors concluded that an imbalance between thromboxane and prostacyclin may represent a marker of patients susceptible to ACE inhibitor cough. Although the study has some limitations, and its findings would need to be confirmed in a larger number of patients, the results do serve as an example of how an interplay of several different pathways may ultimately determine whether an ACE inhibitor causes cough in a patient.

Can Angiotensin II Receptor Blockers be Used in Patients With a History of Angiotensin-Converting Enzyme Inhibitor Cough?

Angiotensin II receptor blockers (ARBs) confer many of the same hemodynamic benefits as ACE inhibitors, but they do not directly inhibit ACE activity or inhibit the breakdown of bradykinin. Theoretically, ARBs would be predicted to be acceptable substitutes for ACE inhibitors in patients who have adverse events such as kinin-mediated cough.[5]

However, almost as soon as ARBs became available, there were reports of patients with a history of ACE inhibitor cough who also developed cough in association with ARBs, calling into question both the proposed mechanism of ACE inhibitor cough, and the validity of recommending ARBs as alternative agents for patients with a history of ACE inhibitor cough. Mechanistically, it has been suggested that the pharmacological actions of ARBs may involve not only the blockade of the angiotensin 1 receptor, but also activation of the angiotensin 2 receptor by increased levels of angiotensin II.[6] Some studies have suggested that the angiotensin 2 receptor may be involved in the activation of the bradykinin-prostaglandin-NO cascade.[7,8] This raises the question as to whether some patients intolerant of ACE inhibitors could be at risk of the same reaction to ARBs.

Controlled studies have provided reassuring data that ARBs are in fact generally well tolerated in patients with a history of ACE inhibitor cough. In a double blind, randomized, parallel-group comparison of the ARB losartan, the ACE inhibitor lisinopril, and hydrochlorothiazide diuretic for 8 weeks in a total of 135 patients,[9] the frequency of cough with losartan was lower than with lisinopril (29 versus 72%, P<0.01), and similar to hydrochlorothiazide (34%).

Similarly, a multicenter study by Paster et al.[10] reported that the incidence, severity, and frequency of dry cough in 100 patients with a history of ACE inhibitor dry cough was significantly lower in those treated with losartan than in those treated with the ACE inhibitor lisinopril, and was similar to the incidence, severity, and frequency of dry cough in those receiving placebo.

Another randomized, double blind, parallel group study of 129 patients with a history of ACE inhibitor-induced cough[11] also found that the occurrence of cough after 3 and 6 weeks of therapy was significantly less with the ARB valsartan (19.5%) and hydrochlorothiazide (19%) than with lisinopril (68.9%).

Treatment of Angiotensin-Converting Enzyme Inhibitor Cough

In cases in which the continuation of an ACE inhibitor is necessary despite cough, cromolyn, baclofen, theophylline, and local anesthetics have been reported to be of benefit, although none has been subjected to large-scale trials.[12,13,14] Although no large, controlled studies have demonstrated the inhibition of ACE inhibitor cough with cyclooxygenase inhibitors, sulindac has been reported to be of benefit, and two studies suggested that intermediate doses of aspirin (500 mg/day) but not low doses (100 mg/day) can suppress ACE inhibitor cough.[15,16,17] In a small series, Lee et al.[18] tested the hypothesis that because NO has proinflammatory effects on bronchial epithelial cells, supplemental iron, an inhibitor of NO synthase, may reduce the cough associated with the use of ACE inhibitors. Patients treated with iron, but not placebo, had significant reductions in cough scores.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: