To Stop or Not?

Over the past 20 years as I have traveled around the world to lecture on angiotensin axis blockade (AAB), I have been asked whether angiotensin-converting enzyme inhibitors (ACE-Is) or angiotensin receptor blockers (ARBs) should be stopped preoperatively.1 We still do not have a definitive answer.2–6 To stop or not to stop medication used to maintain hemodynamic and homeostatic stability in patients with systemic disease is a fundamental anesthetic question. This is because anesthesia and surgery upset the hard-won stability gained by therapy with these agents, potentially uncovering the underlying fragility of a patient.7,8 The question of stopping preoperative therapy to allow anesthesia and surgery to occur more safely has its roots in early recommendations to stop β-blocker therapy.9,10 The aim was to minimize myocardial depression. This approach, however, resulted in demonstrable rebound myocardial ischemia, as highlighted in a vascular surgery case report.11 Myocardial ischemia has also been reported after gradual preoperative withdrawal of propranolol in patients receiving general12 and spinal anesthesia.13 The latter 2 randomized, controlled studies demonstrated increased myocardial ischemia when β-blockers were withdrawn. The authors thus suggested that β-blockers should be continued during the perioperative period. Subsequent to these studies and recommendations, an increase in mortality was demonstrated with perioperative β-blocker withdrawal.14,15 Then, in the belief that β-blockers might actually confer perioperative myocardial protection, preoperative β-blocker therapy was instigated with the hope of improving perioperative outcomes.16 It was later shown that, in noncardiac surgical patients, perioperative β-blocker administration was associated with lower rates of 30-day all-cause mortality in patients with ≥2 risk factors.15 As a result of these studies, new guidelines for the management of perioperative β-blockade have recently been published by both the American Heart Association (AHA)17 and the European Society of Cardiology (ESC).18 In essence, both recommend continuing β-blocker therapy perioperatively. For patients not taking β-blockers, they recommend initiating therapy preoperatively in circumstances where there is an intermediate or high risk of myocardial ischemia, as determined by risk stratification using both patient comorbidities and the surgical procedure. They also recommend that, if β-blockers are to be administered, therapy should be initiated ≥24 hours before surgery and not on the day of surgery (AHA guidelines).17 The question of whether to stop a medication preoperatively has arisen with many other agents, including long-acting mono amine oxidase inhibitors (MAOIs),19 (but not the newer short-acting agents),20 tricyclic antidepressants,21 oral hypoglycemics,7 α-antagonists (such as clonidine),22 antiarrhythmics (such as amiodarone),16 statins,23 diuretics,6 calcium channel blockers,24 and, of course, ACE-Is25 and ARBs.26 Of these medications today, only oral hypoglycemics24 and, perhaps, diuretics6,16 are consistently discontinued the day before surgery. This too is controversial because both the potential cardiac-protective effects and lactic acidosis associated with the use of metformin are better understood.27,28 The recurring theme in to stop or not has been an initial call to stop a medication following expert opinion (β-blockers)9,10 (statins)29 or case reports.23 Frequently, there was then a realization that a rebound or withdrawal reaction may occur, as demonstrated with statins,30 clonidine,31 and the aforementioned β-blockers,15 or that the patient becomes destabilized as with MAOIs8 and calcium channel blockers.8 Furthermore, it has been recognized that anesthetic techniques can be modified to manage patients without needing to stop these medications.8,16 For example, a MAOI-safe technique that requires avoiding pethidine/meperidine and indirect sympathomimetic amines such as ephedrine has been described.8 Additional studies have demonstrated the beneficial effects of β-blockers and statins with respect to cardiovascular and mortality outcomes.32,33 In addition, improved renal34 and neurological outcomes32 have been demonstrated in patients receiving perioperative statin therapy. Collectively, these findings have resulted in the recommendation that preoperative therapy with β-blockers and statins be specifically instituted6,33 to improve outcomes, bringing the process of to stop or not full cycle. How is this cycle playing out for ACE-I and ARB therapy and what might the recommendation to stop or not be if history is our guide? The first oral ACE-I (captopril) was introduced into clinical practice in 197735 and the first oral ARB (losartan) in 199136 (Table 1).37–39 Soon after, there were case reports describing significant hypotension.40,41 In a patient receiving chronic captopril therapy, systolic blood pressure decreased from 180 mm Hg preinduction to 80 mm Hg after induction for intubation, with a nadir of 60 mm Hg.40 In another patient, receiving irbisartan, blood pressure decreased to 47/30 mm Hg 5 minutes after induction, despite the administration of 18 mg ephedrine.41 The degree of hypotension reported while on angiotensin-converting enzyme therapy is affected by the coadministration of diuretics42 and is related to both the duration of AAB discontinuation43 and the half-life of the ACE-I administered (Table 1).2 It is important to note that postinduction hypotension is more dramatic with ARB therapy.41,44Table 1: Pharmacokinetic Parameters and Duration of Effect of ACE-I and ARB TherapyThis refractory hypotension resulted in recommendations that AABs should be stopped preoperatively.25,26 Subsequently, individual randomized clinical studies2,26,43,45 (Table 2) and a meta-analysis46 demonstrated that withholding AAB therapy preoperatively resulted in fewer episodes of hypotension. However, withholding AAB therapy preoperatively generated a concern for rebound hypertension,2 worsening cardiac failure,47 or myocardial infarction.46 This proved not to be the case in a group of patients with cardiac failure47,48 and hypertension.49 Despite the hypotension associated with AAB therapy, some have recommended continuing AABs preoperatively, maintaining that the hypotension seen is easily treated.50,51 Suggestions for treating AAB-associated hypotension are based on heart rate.26 With heart rates of <60 beats/min, administration of glycopyrrolate and ephedrine has been suggested; when the heart rate is >60 beats/min, phenylephrine should be used.52 Refractory hypotension may be treated with norepinephrine and/or low-dose vasopressin.26,52Table 2: Clinical Studies Assessing the Effect of the Withdrawal or Continuation of ACE-I or ARB Therapy on Perioperative OutcomeIn the event that chronic ACE-I or ARB therapy is discontinued before surgery, the period of withdrawal for specific agents should be based on an understanding of the differential duration of their blood pressure effects (Table 1). Thus, captopril should be stopped ≥10 hours before surgery, valsartan 24 hours, and irbisartan 36 hours, respectively. There have been just 4 prospective, randomized, controlled trials evaluating the effect of withdrawal of AAB therapy on anesthesia-induced hypotension and perioperative outcomes (Table 2). Three of these studies (top 3, Table 2),2,26,45 comprising 123 patients, demonstrated that withdrawal of AAB therapy may decrease the associated hypotension. None of the studies, however, were adequately powered nor did they address outcomes. The fourth prospective randomized study,49 in 526 patients, evaluated the effect of AAB withdrawal on the potential for developing perioperative hypertension. No significant difference between groups was found. In addition to these 4 prospective, randomized, controlled trials, there are several studies, most propensity score matched, evaluating whether patients receiving an AAB had impaired outcomes after surgery compared with those not-so treated (Table 3).41,42,55–60Table 3: Clinical Studies Comparing Perioperative Outcomes in Patients With ACE-I or ARB Therapy or Without Therapy PreoperativelyBecause of the associated patient complexity and intensity of surgery, these studies can best be grouped into cardiac,55–57 vascular,41,58 or noncardiac surgery.42,59,60 In both the cardiac and vascular groups, there was an increased incidence of hypotension,41,56,57 30-day mortality,57,58 development of renal impairment,5 and acute kidney injury (AKI).56 All of these worsened 90-day mortality. In a meta-analysis of patients undergoing cardiac surgery, there was an increased incidence of AKI and mortality.61 In the noncardiac surgery group, hypotension was demonstrated only if AAB therapy was associated with diuretic therapy.42 There was no statistical difference in renal failure, postoperative myocardial infarction rate,42 30-day mortality, and cardiac or renal morbidity.59 In contrast, in patients undergoing major orthopedic surgery, we recently demonstrated an increased incidence of hypotension and AKI in patients continuing AAB therapy in the perioperative period.60 There is, however, a concern that the previously cited nonrandomized, observational studies may be inadequately powered and may suffer from errors of residual confounding.62,63 Furthermore, even with the most significant associations, observational studies cannot prove causality. There are, therefore, very little reliable data to inform the question: to stop or not?5,17,33,62 This begs the question of whether recommendations to stop AABs preoperatively or not should be based on a one-size-fits-all42,49 approach or on patient characteristics and the intended surgery. This may be especially pertinent in surgery where significant blood loss is anticipated or where renal morbidity has been demonstrated (Table 3). In a hemorrhagic shock model in dogs64 and rats,65 administration of ACE-Is is associated with decreased cortical renal blood flow64 and an inability to recover blood pressure rapidly after moderate blood loss.64,65 In addition, there is a recent prospective but nonrandomized study of cardiac surgical patients in whom AAB therapy was either absent, stopped, or continued on the day of surgery (Table 2).54 Associated with these 3 levels of exposure to AABs was a graded increase in the development of AKI but, importantly, no structural changes in kidney function as assessed by 4 different biomarkers. The authors54 suggest that perioperative AAB might not be deleterious to kidney function but may, in fact, offer, benefit through enhanced renal cortical perfusion.66,67 Notwithstanding, meticulous attention to factors detrimental to kidney function (nephrotoxic agents) and perfusion (hypotension) may serve to decrease the potential for perioperative AKI.68 With regard to the potential beneficial effects of perioperative AAB therapy, there is now evidence that continuing (versus stopping) as well as the postoperative addition of AAB therapy in cardiac surgical patients is beneficial in terms of composite (cardiac, cerebral, and renal) outcomes (Table 2). This despite the greater hypotension seen when AABs were continued.53 Furthermore, with patients in whom ACE-Is are not resumed after cardiac surgery, there was an increased rate of gastrointestinal complications.69 After general surgery, there was an independently associated greater 30-day mortality compared with patients in whom AABs were restarted.70 These findings, in addition to the previous work demonstrating the beneficial effect of randomizing coronary artery bypass grafting (CABG) patients to ACE-I therapy preoperatively,71,72 have resulted in specific recommendations to initiate preoperative or postoperative AAB in CABG patients not previously treated with such agents.73 The potential beneficial effects of perioperative AAB therapy are beginning to bring the debate full cycle as described for β-blocker and statin therapy. We are now at the stage in this cycle where there are contrasting recommendations on whether AABs should be stopped4,16,24,33 preoperatively or not.5,6,33 The tendency toward continuing AAB therapy in the perioperative period is well illustrated by the American College of Physicians’ information and education resource (www.pier.acponline.org) on perioperative medication management. In an editorial in 2008, this resource was quoted as recommending the discontinuation of AAB on the day of surgery.74 Present-day recommendations for preoperative ACE-I therapy: controversial-continue with caution (avoid hypovolemia); and for ARBs: uncertain, although usually are continued.6 UpToDate, the widely used point of care tool, suggests: Although there are insufficient data upon which recommendations can be based, it seems reasonable to continue these drugs in patients who are taking them for the management of hypertension. On the other hand, it is also reasonable to withhold them on the morning of surgery in patients who are taking them for congestive heart failure in whom the baseline blood pressure is low, to avoid significant hypotension during the induction of anesthesia.75 New guidelines have just been introduced by the AHA and the ESC for the management of noncardiac surgery patients. The AHA task force recommends: Continuation of ACE-Is or ARBs perioperatively is reasonable (class IIa recommendation, level of evidence: B). They also recommend that if AAB therapy is withheld, it should be restarted as soon as feasible (level: C)17 referencing work highlighted in this article.26,44,46,59 ESC recommendations go further in differentiating management for hypertensive patients and those with heart failure (HF). In HF patients, AAB therapy is recommended to be continued under close monitoring. For those HF patients not already on AAB therapy, it is now recommended that AAB therapy be started ≥1 week before surgery (both recommendations class IIa, level of evidence: C). For hypertensive patients, the Task Force recommends: transient discontinuation of ACE-I and ARBs before noncardiac surgery should be considered based on hypotension encountered with the continuation of these agents.33 It should be noted that there is very little referenced work76 pertaining to AAB provided in the ESC guidelines to substantiate this approach. Taken together, however, it is likely we will see more patients continuing AAB therapy predicated on specific patient characteristics and the planned surgery. In our own Institution, based on our findings of the development of AKI in patients undergoing major orthopedic surgery,60 we stop AAB therapy on the day of such surgery.77 We predict, however, that if history is our guide (as more evidence accumulates around the protective effects of perioperative AAB therapy), these agents will not only be continued preoperatively, but specifically instituted de novo. This brings us full cycle with regard to the question to stop or not. DISCLOSURES Name: Berend Mets, MB ChB, PhD, FRCA, FFA (SA). Contribution: This author wrote the manuscript. Attestation: Berend Mets approved the final manuscript. This manuscript was handled by: Martin J. London, MD. ACKNOWLEDGEMENTS Dr. Pat McQuillan is thanked for reviewing the manuscript.