By Michael H. Crawford, MD
Professor of Medicine, Lucy Stern Chair in Cardiology, University of California, San Francisco
SYNOPSIS: When comparing angiotensin-converting enzyme inhibitors and angiotensin receptor blockers (ARBs) to treat hypertension, researchers observed no difference in major cardiovascular events — but a better safety profile for ARBs.
SOURCE: Chen R, Suchard MA, Krumholz HM, et al. Comparative first-line effectiveness and safety of ACE (angiotensin-converting enzyme) inhibitors and angiotensin receptor blockers: A multinational cohort study. Hypertension 2021;78:591-603.
Angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) are class I evidence A first-line agents for treatment initiation in hypertension. However, little is known about their comparative effectiveness and safety. Chen et al compared the effectiveness and safety of ACEIs and ARBs for the first-line treatment of hypertension in a global network of eight large observational databases. They used statistical and informatics approaches to reduce confounding and bias.
They generated more than 6 million effect estimates for 55 outcomes comparing all recommended first-line antihypertensives. At least 2,500 patients treated with each drug class met the inclusion/exclusion criteria from 1996 to 2018. Inclusion meant patients treated with one agent, either an ACEI or an ARB, for one year during a period when both agents were available. The authors excluded patients who were exposed to any other antihypertensive either before or within seven days of starting an ACEI or ARB. Four of the 55 outcomes studied were the primary effectiveness outcomes: acute myocardial infarction (AMI), heart failure (HF), stroke, and a composite of these three (plus sudden cardiac death). The 51 secondary outcomes were safety outcomes or adverse effects based on the product labels, including angioedema, cough, hypotension, syncope, and electrolyte abnormalities. For the main on-treatment analysis, continuous exposure to the drug was required. Propensity score models were used to adjust for comorbidities and other covariates. Also, to further adjust for residual bias, 76 negative control outcomes were analyzed. Finally, sensitivity analyses, including measured blood pressure (BP), were employed. A total of 2,297,881 patients were started on an ACEI and 673,938 were started on an ARB (48% vs. 15%).
After more than 500 days of follow-up, there was no difference between those started on ACEI and ARB in the primary clinical outcomes, which was not changed by sensitivity analysis. The secondary outcomes showed significantly higher incidences of pancreatitis (HR, 1.32), angioedema (HR, 3.31), cough (HR, 1.32), gastrointestinal (GI) bleed (HR, 1.18), and abnormal weight loss (HR, 1.18) on ACEI vs. ARBs. After a Bonferroni correction, only cough and angioedema remained statistically significant. The authors concluded that although the safety profile for ARBs was better for the first-line treatment of hypertension, these drugs were not more effective at preventing major cardiovascular events.
This report takes observational studies to a new level — or, if you prefer, to the big data level. By amassing eight large observational databases, Chen et al studied more than 3 million patients undergoing initial drug therapy for hypertension with ACEI and ARBs. The results support many clinicians’ current approach: start ARBs rather than ACEIs because of the difference in safety. The name of the game in hypertension treatment is adherence, as this is a chronic condition that usually does not cause symptoms. Many clinicians have realized cough associated with ACEI often is enough to sabotage medication compliance. Once ARBs became generic, I abandoned ACEI for the treatment of hypertension for this reason.
Chen et al found another issue with ACEI: angioedema, which is less common than cough but much more serious. Also concerning was the higher incidence of pancreatitis and more GI bleeding cases with ACEI, although these did not survive the Bonferroni correction (P < 0.01 vs. P < 0.05). However, pancreatitis in this instance may be caused by edema of the pancreatic duct, which can occur with excess bradykinin. ACEIs retard the degradation of bradykinin, which is thought to explain cough and angioedema. GI bleeding is a new adverse effect that has not been reported.
There were limitations to this study, besides its observational nature and the potential for residual confounding and bias. It is not possible to evaluate differences between different drugs in each class. However, most patients in the ACEI arm were treated with lisinopril, which also is the most commonly prescribed antihypertensive. Thus, this study reflects the real-world use of drug therapy for hypertension. In addition, specific drug use one week after initiating first-line therapy was not considered, so some patients may have been prescribed other agents later. In fact, this is highly likely, as most hypertensive patients require more than one drug to control their BP. Despite these drawbacks, this is the largest study to compare the two drug classes for first-line therapy of hypertension. The results favor the preferential use of ARBs rather than ACEIs when initiating treatment for hypertension.