Long-term Angiotensin-converting Enzyme Inhibitor Treatment Attenuates Adrenergic Responsiveness without Altering Hemodynamic Control in Patients Undergoing Cardiac Surgery

After obtaining institutional Ethical Committee approval and informed consent, 41 patients scheduled for elective coronary artery bypass graft (CABG) surgery or mitral valve replacement (MVR) were studied during a 14-month period. All had moderate to well-preserved global left ventricular function (ejection fraction of at least 40%). Patients referred with mitral valve incompetence were selected for the study if no episode of congestive heart failure had occurred during the previous 3 months while receiving appropriate therapy and if left ventricular end-diastolic pressure was less than 18 mmHg. Patients with unstable angina, recent myocardial infarct (less than 30 days), diabetes mellitus, and those receiving amiodarone or alpha-adrenergic therapy were excluded from the study. According to medical history, patients were divided into two groups, those not receiving ACE inhibitors (control group, n = 19) and those treated with ACE inhibitors for at least 3 months (ACEI group, n = 22; captopril, n = 9 and enalapril, n = 13); this was based on previous clinical investigations demonstrating evidence for regression of myocardial hypertrophy and improvement in arterial vascular compliance as soon as 3 months after initiation of treatment. [13].

It was determined that patients in each group would either receive a norepinephrine dose-response challenge to assess adrenergic responsiveness (control-norepinephrine group, n = 10 and ACEI-norepinephrine group, n = 13) or receive saline to assess the "spontaneous" hemodynamic changes related to cardiopulmonary bypass (CPB) and the neuroendocrine responses to surgery (control group, n = 9 and ACEI group, n = 9). Hormonal analysis was not performed in patients receiving the norepinephrine dose-response challenge because plasma renin activity (PRA), plasma angiotensin II, and catecholamines concentrations would have been influenced, directly or through counterregulatory mechanisms by the prolonged norepinephrine infusion (10-25 min). Patients who were treated preoperatively with calcium channel blockers were excluded from the norepinephrine dose-response study because this treatment has been associated with reduced alpha¹-adrenergic responsiveness during anesthesia and surgery. [14].

The usual cardiac medications, including nitrates, beta-blockers, calcium channel blockers and ACE inhibitors, were administered until the morning of surgery. After overnight fasting, each patient received 7.5 mg morphine intramuscularly and 5-10 mg diazepam orally as premedication 1 h before surgery. On arrival at the induction room, the electrocardiogram (standard lead II and V⁵) was monitored for heart rate, rhythm, and ST changes. Arterial radial and pulmonary catheters were inserted under local anesthesia for continuous monitoring of mean arterial pressure (MAP) and central venous pressure, as well as for intermittent determination of cardiac output by the thermodilution technique. General anesthesia was induced according to standard practice by an experienced anesthesiologist not involved in the study: for induction, 10-30 micro gram/kg fentanyl and 0.03-0.06 mg/kg midazolam were administered in incremental doses to produce unconsciousness while maintaining hemodynamic stability; pancuronium was given to facilitate ventilation and tracheal intubation; thereafter, anesthesia was maintained with a continuous infusion of midazolam (0.10 mg *symbol* kg sup -1 *symbol* h *symbol* sup -1 during normothermic circulation and 0.05 mg *symbol* kg sup -1 *symbol* h *symbol* sup -1 during hypothermic extracorporeal bypass) and additional intravenous boluses of fentanyl (up to a total dose of 60 micro gram/kg. Mechanical ventilation was adjusted to maintain normocapnia and normoxia with an inspired oxygen concentration of 50% (oxygen-air mixture) before CPB and 100% after separation from bypass.

After aortic and right atrium cannulation, CPB was instituted with a membrane oxygenator primed with 1.5 l crystalloids and body temperature was decreased to 26-28 degrees C. After aortic clamping, a St. Thomas's cardioplegic solution (600-900 ml) was administered through the aortic root and repeated at 30-min intervals. A nonpulsatile pump flow rate (PFR) greater than 1.8 l *symbol* min sup -1 *symbol* m sup -2 was maintained during moderate hypothermia and increased up to 2.4 l *symbol* min sup -1 *symbol* m sup -2 during rewarming. After completion of the surgical procedure and systemic rewarming, patients were weaned from CPB when a rectal temperature of at least 35.5 degrees C had been reached.

In the intensive care unit, repeated boluses (1 mg) of morphine and/or diazepam were administered to keep the patient pain-free and comfortable; weaning from the ventilator was started during emergence from anesthesia and when stable hemodynamics and normothermia had been maintained for at least 1 h.

Cardiac output was averaged from three consecutive determinations (within 10% of each other) and cardiac index (CI) was obtained as cardiac output divided by body surface area. In all patients, systemic hemodynamic variables (MAP, heart rate, CI, or PFR) were recorded at these time periods: before anesthesia (pre-anesthesia), after anesthesia induction and before tracheal intubation, 2 min after sternotomy, 5 min after aortic cross-clamping, 5 min before aortic unclamping, 10 min after separation from CPB, and during skin closure. Systemic vascular resistance index (SVRI) was calculated by the standard formula (MAP-central venous pressure/CI or PFR) *symbol* 80 (dynes *symbol* s *symbol* cm sup -5 *symbol* m sup -2).

Arterial blood was sampled at each hemodynamic measurement period (except after anesthesia induction) as well as 12 and 24 h after the end of surgery. Blood was collected in 5-ml polypropylene tubes containing ethylenediaminetetraacetic acid and after cold centrifugation, the decanted plasma was stored frozen until analysis. Plasma epinephrine and norepinephrine concentrations were determined in duplicate by high-pressure liquid chromatography coupled with electrochemical detection. The lower limit of sensitivity for the assay was 0.3 nM/L and 0.1 nM/L for norepinephrine and epinephrine, respectively; intraassay and interassay precision (or coefficient of variation) were 6.8% and 8.7% for norepinephrine, and 7.0% and 8.9% for epinephrine. Plasma renin activity was estimated by a modification of the method described by Haber using New England reagents. [15]Immunoreactive plasma angiotensin II was determined in duplicate using a standard radioimmunoassay technique after plasma extraction by means of Sep-pak cartridges (Waters, Milford, MA) with methanol/water; the lower detectable angiotensin II concentrations were 0.01 pg/ml, intra- and interassay precision were 8% and 12%, respectively. Hormone concentrations were corrected for the hemodiluting effect of CPB according to the change in hematocrit, that is, hormone concentration x Hct^p/Hct^s, where Hct^pand Hct^sare the hematocrit values prevailing before anesthesia and during CPB, respectively.

Norepinephrine dose-response curves were generated during two study periods: (1) 5 min after aortic cross-clamping during conditions of constant circulatory flow (PFR of 1.8 to 2.21 *symbol* min sup -1 *symbol* m sup -2) and temperature (hypothermic CPB), and (2) after separation from the bypass circuit and heparin neutralization, during stable hemodynamic conditions (normothermic post-CPB). Norepinephrine (1 mg/ml; 0.5 mg free-base of norepinephrine diluted to 50 ml with 0.9% saline) was infused continuously via the side port of the internal jugular vein catheter or directly into the CPB venous reservoir. The rate of norepinephrine administration was increased in a stepwise fashion (0.01, 0.02, 0.04, 0.08, 0.16 micro gram *symbol* kg sup -1 *symbol* min sup -1) at 5-min intervals until MAP exceeded values 25% above baseline or reached 105 mmHg. Hemodynamic measurements (MAP, central venous pressure, cardiac output, or PFR) were recorded during the last minute of each infusion step, and the derived values were calculated (CI, SVRI). The vasopressor and vasoconstrictor responses were defined as the changes in MAP and SVRI induced by the norepinephrine infusion.

To assess the hemodynamic changes related to surgery and to the institution of hypothermic bypass alone, hemodynamic variables were measured in patients receiving a saline infusion rather than norepinephrine (i.e., patients in whom blood was withdrawn for hormonal analysis only); variables were recorded over 20 min at the same times as described earlier, i.e., after aortic cross-clamping (hypothermic CPB) and after discontinuation of CPB (normothermic post-CPB).

Time to extubation, total fluid requirements, and the need for vasopressor (phenylephrine) and inotropic support (dobutamine or dopamine at > 5 micro gram *symbol* kg sup -1 *symbol* min sup -1, norepinephrine at > 0.08 micro gram *symbol* kg sup -1 *symbol* min sup -1) was recorded intraoperatively and postoperatively.

To assess the vasopressor or vasoconstrictive effects of norepinephrine and those related to the institution of hypothermic bypass, norepinephrine dosage and time were used as independent variables and changes in MAP and SVRI as the dependent variables. The dose of norepinephrine required to increase baseline MAP by 20% was calculated and defined as the pressor dose 20 (PD²⁰). To consider the simultaneous changes in MAP and flow, changes in SVR versus log dose norepinephrine or versus time were described. A log linear for best fit was constructed for individual dose-responses and the slopes were compared between treatment groups. Parametric data were expressed as mean+/-SEM and analyzed using two-way analysis of variance with Dunnett's test for within-group changes with respect to baseline. Mann-Whitney U test and Wilcoxon paired tests were used for between- and within-group comparisons of non-Gaussian distributed data that were expressed as median (lower and upper quartile). A chi-square analysis with Yates's correction was used to compare percentages of patients in the two groups. Differences were considered to be significant when P < 0.05. For hormonal analysis, data were log transformed and the concentration equal to the sensitivity of the assay was assigned the value of the statistical limit of significance.

The two treatment groups were not significantly different with regards to clinical and surgical characteristics (Table 1), except for preoperative treatment: in the control group, beta-adrenergic and/or calcium channel blocking medications were prescribed in 16 patients (13 with CABG and 3 with MVR) whereas only 7 patients (5 with CABG and 2 with MVR) in the ACEI group received one of these drugs (P < 0.05); in the ACEI group, enalapril (10-20 mg/day) or captopril (30-100 mg/day) were administered for a median period of 14 months (range, 5-24 months) or 12 months (range, 4-16 months), respectively. There were no complications attributable to the study protocol, in particular, no patients developed ST changes during norepinephrine administration. Perioperative fluid requirements tended to be greater in the ACEI group, although differences were not statistically significant (3,422+/-340 ml in control group vs. 4,238+/-452 ml in ACEI group). In the intensive care unit, the use of pharmacologic circulatory support and mechanical ventilation did not differ between the two groups: 5 patients in the control group and 9 in the ACEI group required inotropic or vasopressive support for at least 2 h (not significantly different) and time to extubation was similar in the two groups (560+/-105 in the control group vs. 605+/-135 min in the ACEI group).

Perioperative hemodynamic variables did not differ between the two types of surgery (CABG vs. MVR) as well as between the two types of ACEI inhibitors used (enalapril vs. captopril); therefore, these hemodynamic data were pooled in the control and ACEI groups.

For anesthesia induction, significantly less fentanyl (-44%) and midazolam (-9%) were required in the ACEI group compared with the control group (Table 2). Phenylephrine was administered in two patients in the ACEI group and in one patient in the control group to correct arterial hypotension after anesthesia induction. There were no hemodynamic differences between the two groups at any time, with the exception of a significantly lower heart rate in the control group compared to the ACEI group during the preanesthesia period (60+/-2 vs. 67+/- 2 beats/min, P < 0.05), a finding that was most probably related to the more frequent use of preoperative beta-blockers in this group. From the start of CPB until aortic unclamping, MAP and CI or PFR and SVRI were all significantly reduced compared with the preanesthesia baseline values, in both groups. After complete rewarming and separation from CPB, CI rose up to preanesthesia levels, although MAP and SVRI remained significantly lower (Table 3). After chest closure, three patients in each group required inotropic or vasopressor support and CI was significantly reduced compared to preanesthesia values in both groups, whereas SVRI returned to baseline values.

In our laboratory setting, preoperative measurements of PRA and plasma epinephrine, norepinephrine, and angiotensin II concentrations obtained in the control group gave a range of values similar to those obtained in a group of outpatients of similar age referred to the cardiologist (Figure 1). There were no differences between the two groups in plasma angiotensin II and catecholamines concentrations at any sampling period. Long-term treatment with ACE inhibitors resulted in significantly higher preoperative PRA in the ACEI group (1.7 ng *symbol* ml sup -1 *symbol* h sup -1, median value) compared to the control group (0.12 ng *symbol* ml *symbol* sup -1 *symbol* h sup -1, median value). Within each group, a significant increase in PRA as well as in plasma catecholamines concentrations occurred during and after CPB; greater PRA were consistently observed in the ACEI group, up to 24 h after surgery. In the two groups, angiotensin II plasma concentration decreased transiently after aortic cross-clamping due to pulmonary exclusion, thereafter angiotensin II increased significantly toward the end of surgery. At 12 and 24 h after surgery, plasma catecholamines and angiotensin II concentrations did not differ from preoperative values in the two groups.

During the first 10 min of hypothermic CPB and under constant blood flow conditions, the spontaneous increase in MAP was attenuated in the ACEI group compared to the control group (11+/-1 mmHg vs. 25 +/-2 mmHg, P < 0.01; Figure 2(A)); the slope of the regression line (SVRI/min) was less steep in the ACEI than in the control group, (31 +/-3 vs. 86+/-11, P < 0.05; Figure 3(A)). After weaning from CPB, MAP and CI remained stable during the 20-min observation period in patients receiving saline infusion; as a result, SVRI did not change (-2+/-3% in the control group and +4+/-2% in the ACEI group).

The hemodynamic response to norepinephrine administration was accentuated during the hypothermic compared to the normothermic CPB period because the PD²⁰value was significantly smaller and the slope of SVRI/norepinephrine log dose was significantly steeper in the former condition (Table 4).

The vasopressor and constrictor responses to norepinephrine were markedly attenuated during hypothermic CPB and during the normothermic post-CPB period in the ACEI group compared to the control group; the pressor dose-response curves were shifted to the right (Figure 2(B and C)), the slopes (SVRI/log norepinephrine dose) were diminished by 73% and by 82% (Figure 3(B and C)) and the dose of norepinephrine required to raise MAP by 20% (PD²⁰) were increased by a factor of 2.5 and 2.6, respectively, during and after CPB (Table 4). In the ACEI group, the norepinephrine-induced hemodynamic response was not significantly different between patients receiving captopril (n = 5) or enalapril (n = 8) During hypothermic CPB however, the pressor effects of norepinephrine were greater in patients undergoing CABG compared to the patients undergoing MVR (median values of PD²⁰: 0.05 micro gram/kg in ACEI-CABG patients [n = 8] vs. 0.19 micro gram/kg in ACEI-MVR patients [n = 5], P = 0.013).

The results of the current study demonstrate that the vasopressor and constrictor responses to norepinephrine infusion are significantly attenuated by more than 50% during and immediately after CPB in adult patients receiving long-term treatment with ACE inhibitors. The discrepancy between appropriate hemodynamic and hormonal profiles that are not affected by ACE inhibitor therapy and the norepinephrine-induced increase in MAP and SVR, which is clearly attenuated with ACE inhibition, indicates that the RAS plays a secondary role in blood pressure regulation during surgical stress but that it yet may markedly influence the response to vasopressor agents such as norepinephrine.

It is well established that perioperative hemodynamic control is principally determined by the interplay between the sympathetic nervous system, the RAS, the hypothalamic release of arginine vasopressin, and, finally, the end-organ responsiveness within the cardiovascular system. [16]Because of redundancy among their mechanisms of action, removal of one of these defense lines usually will not threaten blood pressure control under physiologic conditions. [17]Accordingly, the sympathetic nervous system is considered the organism's main defense against arterial hypotension in humans. In patients with autonomic dysfunction and/or limited cardiovascular reserve such as in heart failure or after myocardial infarct, superimposition of pharmacologic RAS blockade is thought to precipitate unstable hemodynamics. Because most inhaled and intravenous anesthetics are responsible for blunting of sympathetic reflexes and for myocardial depression, patients whose RAS is inhibited, are prone to sudden decreases in blood pressure during anesthesia induction or when acute intraoperative hemorrhage occurs. Some of these conditions were actually encountered in the clinical and anecdotal reports relating hypotensive events in ACE inhibitor-treated patients. [9-12]We studied patients with coronary artery disease and mitral valve regurgitation who were free of severe cardiac insufficiency, peripheral neuropathy, and autonomic dysfunction. A "smooth" anesthesia induction resulted in similar and uneventful hemodynamic responses in the two groups. Interestingly, whereas in the control group, patients tolerated the usual loading doses of fentanyl (25-30 micro gram/kg) and midazolam (0.05-0.07 mg/kg), ACEI-treated patients were given significantly lesser amount of anesthetics to maintain a desired stable blood pressure. Sensitivity to the circulatory depressant effects of anesthetics after chronic RAS blockade also has recently been documented by Coriat et al. who observed up to 100% hypotensive episodes requiring phenylephrine administration when a fixed anesthetic dosage regimen (5 micro gram/kg fentanyl and 0.15 mg/kg midazolam) was given in patients chronically treated with enalapril. [11]The lower incidence of hypotension encountered in the current study was most likely attributed to lower midazolam dosage and to slow titration of anesthesia induction that allowed counterregulatory mechanisms to progressively compensate for the circulatory depressant effects of anesthetics.

Systemic hemodynamics, fluid requirements, inotropic and vasopressor support, as well as neuroendocrine responses did not differ between the two groups intraoperatively and postoperatively. We therefore hypothesize that partial preservation of systemic RAS and intact sympathoadrenal activation both contributed to maintain perioperative cardiocirculatory homeostasis in ACEI-treated patients.

During each study period, higher PRA were found in the ACEI group compared with the control group, that resulted from enalapril- or captopril-induced suppression of angiotensin II generation (mostly at the renal juxta-glomerular site) and hence, from interruption of the negative feedback mechanism controlling renin synthesis. As reported by others, [18-20]circulating angiotensin II concentrations usually are not reduced in patients receiving long-term ACE inhibitors and do not correlate with any antihypertensive effects. Because angiotensin II regulates renin and ACE gene transcription via a negative feedback mechanism, synthesis of pulmonary ACE and prorenin substrates are rapidly induced after initiation of ACE inhibitor treatment, leading to increases in PRA and in serum ACE activity and ultimately to restoration of plasma angiotensin II levels. [21]However, it cannot be excluded that the true biologically active plasma angiotensin II concentrations were overestimated in the current study because we used an antibody that cross-reacted with other angiotensin peptides that could have been generated by some other proteolytic enzymes (cathepsines) released by activated leukocytes. Currently, there is a growing body of experimental evidence supporting a central role for RAS blockade within the brain, the cardiovascular system, and the kidney; in particular, vascular ACE activity has been demonstrated to be suppressed independently of plasma RAS and correlates well with the observed hypotensive effects. [22]In addition, alternative vasodilatory mechanisms involving the kallikrein-kinin and prostaglandin pathways appears to be implicated: ACE is a nonspecific enzyme structurally similar to kininase II, hence, ACE inhibition is expected to result in accumulation of endogenous bradykinin with vasorelaxant properties. [23].

The similar changes in plasma catecholamines within the two groups reflected appropriate sympathoadrenal responses to surgery and hypothermic CPB, that helped to maintain stable hemodynamics. In response to other acute stressful events, such as tracheal intubation, [7]postural changes, [24]and exercise, [25]the blood pressure-heart rate relationship has been demonstrated to be preserved in patients chronically treated with ACE inhibitors; this could be attributed to normal or enhanced sympathetic reflexes, as indicated by a resetting or an increased sensitivity of the baroreflex activity [26]and by an even greater release of catecholamines during postural changes. [27]Furthermore, the direct stimulatory effects of bradykinin on the release of norepinephrine and epinephrine may partly compensate for the loss of the facilitatory influences of angiotensin II on the sympathetic nervous system. [28].

Although the hemodynamic and neuroendocrine time course was indistinguishable between patients with and without chronic ACE inhibitors, the adrenergic responsiveness to exogenously infused norepinephrine differed markedly between the two treatment groups. To characterize the adrenergic responsiveness of the cardiovascular system, we administered norepinephrine, a mixed alpha and beta agonist, because it is often used in clinical practice as a vasopressor or inotropic agent to help separate patients from CPB and in the intensive care unit to keep stable hemodynamics. In contrast to phenylephrine, a pure alpha-agonist that was shown to cause coronary vasoconstriction, [29]administration of norepinephrine after completion of surgery, has been associated with unchanged or increased blood flows through venous and arterial coronary bypass grafts. Mean arterial pressure and SVR were chosen as hemodynamic markers: MAP as an end-point of pressor responsiveness and determination of SVR to integrate the simultaneous changes in cardiac output and vascular pressure that were measured during norepinephrine infusion.

To assess the hemodynamic (and neuroendocrine) effects of hypothermic CPB and surgery, we studied patients in the control and ACEI groups receiving intravenous saline instead of norepinephrine. In both groups, baseline hemodynamics remained stable for at least 20 min after complete rewarming and separation from bypass whereas a gradual increase in SVR occurred shortly after institution of hypothermic CPB. Increased plasma viscosity, elevated plasma catecholamines, and tissue metabolic factors related to hypothermia have all been incriminated in increasing peripheral vascular resistance either directly or through enhancement of the sensitivity to endogenous and exogenous vasoconstrictors. [30]Of note, exclusion of the lung from the circulatory compartment was expected to prolong the biologic half-life of norepinephrine because pulmonary endothelium is the predominant clearance pathway of this catecholamine. [31].

The hemodynamic responses associated with hypothermic CPB and those induced by norepinephrine infusion were both markedly attenuated in patients receiving ACE inhibitors on a long-term basis as indicated by reduced values for the regression slopes of the time-response (SVR/time) and dose-response relationships (SVR/norepinephrine log dose) as well as by increased PD²⁰(amount of norepinephrine required to increase MAP by 20%). In support of our results, Tuman and colleagues observed a twofold increased vasoconstrictor requirement immediately after CPB in the presence versus absence of chronic preoperative ACE inhibitors. [32]Attenuation of the vasoconstricting effects of norepinephrine also has been observed in isolated arteries pretreated with various ACE inhibitors [33,34]as well as in normotensive and hypertensive subjects receiving a single 50-mg dose of captopril. [35]Interestingly, blunting of norepinephrine responses was more pronounced in chronically treated patients and reversal of adrenergic "hyporesponsiveness" could be achieved when low doses of angiotensin II or bradykinin type 2 antagonists were administered. [36]Taken together, the rightward shift and the reduced slopes of the vasoconstrictor and vasopressor dose-response are compatible with ACE inhibitors acting as noncompetitive antagonists that blunt the norepinephrine pressor effects through removal of angiotensin II and enhancement of vasorelaxant mediator(s).

Vascular ACE activity is mostly localized within the endothelial layer and adventitia of large and small arteries where it is involved in angiotensin II synthesis and in bradykinin inactivation. [37]Besides its powerful vasoconstrictor effects, angiotensin II acts as a growth factor that may lead to cardiac and vascular hypertrophy and that ultimately results in increased vascular resistance (arteriolar vessels), decreased arterial compliance (large arteries) and nonspecific enhancement of vasoconstrictor responses. [38]Regression of these structural and functional cardiovascular changes has been documented, independently of any blood pressure lowering effect, after a 3-6-month treatment with ACE inhibitors. [13,22,39]Furthermore, the enhanced bradykinin-induced relaxation associated with ACE inhibition is thought to result from the accumulation of endogenous kinins within the endothelium and the secondary release of potent direct vasodilators such as nitric oxide, prostacyclin, and a putative hyperpolarizing factor. The same vasodilatory factors also are released after activation of endothelial alpha²-adrenergic receptors, [40]hence, we may speculate that norepinephrine-induced alpha¹-constriction was more strongly opposed by alpha²-mediated vasorelaxation in the ACEI group than in the control group during the norepinephrine challenge.

From the above data, in vivo attenuation of pressure responses to norepinephrine in patients receiving ACE inhibitors can be viewed as a "resetting" of adrenergic sensitivity secondary to vascular RAS blockade that is characterized by: (1) depletion of angiotensin II, (2) regression of smooth muscle hypertrophy and "normalization" of arterial medial thickness, and (3) improvement of the endothelial vasodilatory function (increase in endogenous bradykinin stores).

Several aspects of the study design deserve further comment because a nonrandomized mixed population comprising patients undergoing CABG or MVR and who were treated preoperatively either with ACEI (captopril or enalapril) or with non-ACEI medications were studied.

First, although systemic hemodynamic and neuroendocrine responses behaved similarly in CABG and MVR patients, we observed greater pressor responses to norepinephrine in patients undergoing coronary artery surgery compared with patients undergoing valvular surgery; this could be related to the higher prevalence of arterial hypertension and diffuse atherosclerotic lesions in patients with coronary artery disease whereas those with mitral regurgitation often are seen with isolated inflammatory, infectious, or degenerative valvular abnormality. There is experimental evidence suggesting that the reactivity, but not the sensitivity, to humoral agents is increased in hypertensive patients because of endothelial dysfunction, reduced arteriolar density, and increased ratio of wall thickness to lumen diameter. [38,41]Second, beta-blockers and calcium-channel blockers were prescribed more often as antihypertensive, antianginal, or antiarrhythmic agents in the control group. Continuation of these medications has been found to be associated with reduced incidence of myocardial ischemia, blunting of the hemodynamic response to surgery but also with increased needs for vasopressor support after CPB. [42]It is conceivable that a greater degree of beta-blockade could account for some attenuation of the sympathetically induced release of renin and for reduced inotropic response to the beta-stimulatory effects of norepinephrine, during and after surgery. [43]However, it is unlikely that an increase in the relative ratio of alpha- to beta-receptor activity due to chronic beta-blockade contributed to relevant differences in vascular alpha¹-responsiveness, because the hemodynamic response to phenylephrine during surgery has been shown to be unaltered after chronic propranolol therapy. [44]Because the vasorelaxant and antiadrenergic effects of ACE inhibition were expected to be amplified or potentiated by calcium channel blockade, we excluded patients pretreated with calcium channel blockers from the norepinephrine dose-response study. In fact, prolonged treatment with these medications also may lead to regression of vascular and cardiac hypertrophy in hypertensive patients as well as to reduced pressure response to phenylephrine during cardiac surgery. [45].

Third, the type of ACE inhibitor does not seem to affect the perioperative hemodynamic changes and norepinephrine-induced pressor responses, because comparable results were observed in patients receiving captopril--a lipophilic compound bearing a sulfhydryl group--or enalapril, which is 10-30 times more potent. Although differential access to tissues and differential inhibition of angiotensin-converting isoenzyme by various ACE inhibitors could account for some discrepant effects, these were probably not relevant in the settings of the current investigation. In a large prospective study, [32]similar vasoconstrictor requirements were found after CPB in patients pretreated with enalapril or captopril, supporting the finding that pooling the data of these two subpopulation of patients has most probably not affected the current results in a significant manner. Yet, it is possible that the variable duration of preoperative ACE inhibitor treatment (424 months) is implicated in time-related regression of the cardiovascular hypertrophic changes that would secondarily influence adrenergic responsiveness.

In summary, long-term treatment with ACE inhibitors in patients with preserved ventricular function is associated with stable hemodynamics and appropriate hormonal responses during cardiac surgery as indicated by activation of the sympathoadrenal and renin angiotensin systems. We hypothesize that the attenuated responsiveness to adrenergic stimulation more likely results from vascular angiotensin II depletion, "normalization" of the vascular structure, and enhanced endothelial vasodilatory function.

Based on these results, it appears that preoperative withdrawal of treatment with ACE inhibitor is not justified. However, the anesthesiologist should be aware of the greater sensitivity of patients receiving ACE inhibitors to anesthetic-induced circulatory depression, caution should also be exercised during surgery involving major fluid losses or when autonomic nervous and endocrine responses are obtunded secondary to disease processes or some drug therapies. Clinicians should keep in mind that continuation of ACE inhibitor treatment is potentially associated with beneficial effects: prevention of hypertension, [46]improved renal function, [6]and attenuation of coronary vasoconstriction. [47]In the future, it might be worthwhile to devote further clinical scrutiny to detect cardiac and renal protective effects associated with the use of ACE inhibitors, in patients undergoing cardiac or noncardiac surgery.