Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Receptor Blockers Before Elective Noncardiac Surgery: An Ongoing Dilemma

Sudarshan Setty, MD; Daniela Orza, MD; Kumar G. Belani, MBBS, MS


Anesth Analg. 2018;127(3):598-600. 

Angiotensin-converting enzyme inhibitors (ACE-I) and angiotensin II receptor blockers (ARBs) are potent and ubiquitous antihypertensive medications.[1] Multiple guidelines suggest that ACE-I/ARB drugs be omitted on the morning of surgery, commensurate with a host of clinical experience detailing profound hypotension after the routine induction of anesthesia during noncardiac surgery when ACE-I or ARB drugs are continued. Moreover, the resulting hypotension may be distressingly resistant to routine rescue doses of vasopressors such as phenylephrine, ephedrine, vasopressin, norepinephrine, and other drugs.[2–4] Thus, continuation of these specific antihypertensive drugs in surgical patients often elicits significant angst for surgeons and anesthesiologists alike. Interestingly, Hollmann et al[5] provide clinicians with contrarian results regarding this conundrum based on the conclusions of a meta-analysis that focuses on major cardiovascular and cerebrovascular outcomes—and not simply the magnitude of change for intraoperative blood pressure.

Their analysis examined major cardiovascular outcomes of over 6000 patients aggregated from a combination of 9 randomized clinical trials and cohort studies. Validating many clinicians' experience, Hollmann et al[5] confirmed an overall 30% increase in the relative risk of hypotension (corresponding to an absolute risk increase of 6.5%, from 23.4% to 29.9%) associated with continued therapy. Thus, the "solution" seems clear —to mitigate this risk, simply withhold all ACE-I and ARB drugs the morning of surgery. Indeed, this is the conclusion of the frequently cited VISION study.[6] The Vascular Events in Non-cardiac Surgery Patients Cohort Evaluation (VISION) analyzed data from 14,687 patients undergoing noncardiac surgery of which 4802 were on ACE-I/ARB in a prospective cohort study. Patients who withheld ACE-I/ARBs in the 24 hours before surgery had a lower incidence of all-cause death, stroke, or myocardial injury (12% in the withheld group versus 12.9% in the continued group with an adjusted relative risk of 0.82, 95% CI, 0.70–0.96, and a P value of .01). But is withholding these specific antihypertensive drugs always necessary, or even prudent? Strikingly, meta-analysis by Hollmann et al[5] did not demonstrate an association between continued administration of ACE-I/ARB therapy and patient mortality (odds ratio = 0.97; 95% CI, 0.62–1.52) or major adverse cardiac events (MACEs; odds ratio = 1.12; 95% CI, 0.82–1.52) despite the frequent occurrence of intraoperative hypotension.

So, what is the most prudent approach for clinicians at this time? What are the likely long-term consequences from continuing ACE-I/ARB therapy versus withholding them? Should all patients have their ACE-I/ARB medication withheld on the day scheduled for surgery, or just select subsets (eg, cardiac versus noncardiac surgery)? Is a 24-hour hold sufficient, or is even more time required for those medications to be eliminated?

ACE-I and ARB pharmacology is complex and may provide some insights into the discrepant findings of the meta-analysis by Hollmann et al.[5] Most of the ACE-I are prodrugs, which are metabolized in the liver and kidneys, with the exception of lisinopril, which is excreted unchanged in urine. Factors like congestive heart failure, kidney, or liver dysfunction can affect the half-life of the ACE-I.[7] It is therefore important that the practitioner evaluates the type and time of the last dose of the ACE-I or ARB that has been prescribed, as each drug has a different pharmacokinetic and pharmacodynamic profile (Table).[8,9] For example, captopril has an elimination half-life of 4–6 hours and enalapril, which is longer acting than captopril, is deesterified in the liver and kidneys to its active form enalaprilat. The elimination half-life of enalaprilat is normally about 5 hours but increases in patients with congestive heart failure to 6–8 hours, and with repeated doses the elimination half-life increases to 11 hours. Thus, whether and when you stop ACE-I may well depend on several comorbidities.

But even after accounting for the complexity of ACE-I/ARB drug pharmacology, 2 questions remain:

  1. How can we account for the consistent occurrence of hypotension but the lack of association with increased mortality or morbidity in the meta-analysis by Hollmann et al[5] (unlike prior studies)?

  2. While the VISION trial[6] accounts for 97.4% weightage for the mortality analysis, the meta-analysis by Hollmann et al[5] comes to a discrepant conclusion regarding MACE outcomes. How do we explain this?

First, clinicians should consider key components and implications of the statistical meta-analysis itself. While this unique tool can lend powerful insights to summarizing disparate findings surrounding a specific medical intervention, the conclusions are potentially susceptible to a host of limitations.[10] For instance, critics opine that the technique is encumbered if authors combine different types (ie, "apples and oranges") of studies—indeed, the underlying studies of Hollmann et al[5] combined 5 randomized controlled trials with 4 cohort studies. Potential differences in reliability between these 2 study formats can be difficult to discern, and the analysis by Hollmann et al[5] was not broken out by each study type. Moreover, the meta-analysis did not adjust for key baseline factors, increasing the likelihood of underestimating a treatment outcome. Any meta-analysis may also be influenced by major discrepancies in study size. Here, the results of the single study by Roshanov et al[6] (Table 1 in Hollmann et al[5]) accounted for 80% of the combined patient data—more than all the other studies combined. Finally, an inherent limitation in some of the original studies included in the meta-analysis is that MACEs were unreported. Indeed, the mortality data appear to be drawn from a minority of the studies. Moreover, one could speculate that smaller studies without serial troponin and/or electrocardiogram monitoring missed substantial numbers of myocardial infarctions or at least myocardial injury.

As expected, the 9 selected studies in the meta-analysis by Hollmann et al[5] include a host of different ACE-I/ARB drugs, the interval from last dose to surgery, the definition of hypotension, and the methods used to treat hypotension in the operating room. Only 3 of the 9 studies included in the meta-analysis by Hollmann et al reported which of the ACE-I/ARBs were administered (or withheld),[11–13] and only 1 reported that they included ARB drugs, but did not specify which ones.[14] Depending on the specific study, the ACE-I/ARB drugs might be withheld for >10 hours,[13] 12–24 hours,[11] 24 hours,[12] or were simply held the day before surgery[6,14,15] without specifying the exact number of hours (Table 2 in Hollmann et al). Thus, there is wide variation in time from last dose and the different type of ACE-I/ARB medications in the meta-analysis.

In addition, 9 studies varied widely in the duration for which the blood pressure was reported, the treatment of hypotension, and the very definition of what constitutes hypotension (surprisingly none of the studies reported the duration of hypotension). The definition of hypotension varied widely from a systolic blood pressure <80[14] or 85 mm Hg[9,11,13] lasting longer than a minute, a systolic blood pressure <90 mm Hg that prompted a clinical intervention,[6] or a systolic blood pressure <90 mm Hg for any duration.[11] Only 1 study defined hypotension as mean arterial pressure of <60 mm Hg.[16] Treatment for hypotension also varied from vasopressors[12,14,16] to intravenous fluids[15] or intravenous fluids with vasopressors[11] to a multimodal approach.[6]

Realistically, all clinicians worry about not just the degree of hypotension, but also its duration. For instance, we know that prolonged hypotension with mean arterial pressures of <55 mm Hg for longer than 20 minutes results in increased mortality, and adverse renal and cardiac outcomes.[17] Thus, the absence of this information (the duration of intraoperative hypotension in the analysis) leaves us to wonder whether the continuation of ACE-I and ARB drugs on the day of surgery causes only transient hypotension (insufficient to cause mortality or MACE), or that these clinicians were very adept and proficient at its correction.

Finally, one is left to reconcile why the Hollmann et al meta-analysis reports no change in mortality regardless of whether ACE-I/ARBs were continued on the day of surgery (odds ratio = 0.96), while the largest single study included in the data set (the VISION study[6]) reported a lower risk of mortality, myocardial injury, and stroke associated with withholding ACE-I/ARB drugs. Perhaps the inability of the current meta-analysis to fully adjust for baseline factors and covariates is critical, whereas Roshanov et al[6] adjusted for a host of potential confounders in their study.[6] Additionally, the VISION study reported results from a modified Poisson regression in the context of adjusted relative risks rather than pooled estimates of the odds ratio as done in the Hollmann et al meta-analysis.

Regardless, the meta-analysis by Hollmann et al[5] provides additional insights and uncertainties while addressing the vital decision about preoperative medications during the preparation of adult hypertensive surgical patients. We agree that a large randomized controlled trial is required to finalize the elusive answer regarding the safety of continuing or withholding ACE-I/ARB drugs on the day of surgery. The current literature both supports withholding ACE-I/ARBs the day before surgery (Roshanov et al[6]) and questions it (Hollmann et al[5]). For now, we encourage clinicians to consider all aspects of the literature along with unique characteristics of each patient in reaching optimal care decisions.