Hypotension Resistant to Therapy with alpha Receptor Agonists Complicating Cardiopulmonary Bypass

Abstract

Lithium is the primary drug used for the treatment of bipolar (manic-depressive) disorders [1]; it is also used for the treatment of other neurologic and psychiatric diseases, including aggressive and self-mutilating behavior [2] and cluster headaches [3]. Patients receiving lithium may demonstrate a number of side effects, including tremor, confusion, cardiac conduction disturbances, polyuria, and hypothyroidism. Lithium has several side effects important to the anesthesiologist. It prolongs sleep time induced by barbiturates [4,5], enhances neuromuscular blockade produced by succinylcholine [6] and pancuronium [7], and may produce cardiac conduction abnormalities. Despite its widespread use, very little is known about lithium's exact mode of action. An important landmark in understanding lithium's mechanism of action occurred when Allison and Stewart [8] demonstrated that lithium decreases the inositol concentration in rat cerebral cortex. Decreased inositol concentrations in brain were later associated with accumulation of inositol phosphate, resulting from an inhibition of the enzyme inositol-1-phosphatase [9]. Hence, lithium may exert therapeutic actions by interfering with the regeneration of phosphoinositides, ubiquitous cellular second messengers; lithium may therefore play an important role in synaptic transmission [10,11]. This theory, called the "inositol depletion hypothesis," offers a plausible explanation for lithium action consistent with known therapeutic and pharmacological characteristics [12,13]. The phosphoinositide pathway is a common intracellular second messenger signaling system [14] mediating numerous receptor-stimulated cellular functions. alpha1-Adrenergic receptors (alpha1 ARs) couple via phosphoinositide hydrolysis to physiologic functions such as vasoconstriction; therefore, lithium therapy may theoretically affect alpha1 AR-mediated vascular tone. We now report a case of hypotension with failure to respond to alpha1 AR agonists after the initiation of cardiopulmonary bypass (CPB) in a patient receiving chronic lithium therapy. Hypotension, particularly during CPB, has not previously been reported as a potential side effect of lithium therapy. Case Report A 74-yr-old, 95-kg man with a history of Class IV angina was scheduled for elective coronary artery revascularization. Previous medical history was significant for coronary artery disease, hypertension (treated with diltiazem 120 mg twice daily and lisinopril, an angiotensin-converting enzyme [ACE] inhibitor, 50 mg once daily), and type II (non-insulin requiring) diabetes mellitus with associated chronic renal insufficiency and gastroparesis. In addition, the patient had a history of a bipolar disorder, which was treated with lithium carbonate 300 mg three times daily. Physical examination was unremarkable. Cardiac catheterization revealed an ejection fraction of 74% and severe three-vessel coronary artery disease. Serum chemistry revealed a mildly increased serum creatinine level (2.0 mg/dL) consistent with chronic renal insufficiency. Four days prior to admission, a serum lithium level was slightly subtherapeutic at 0.25 mmol/L (therapeutic: 0.4-0.6 mmol/L); hence, the patient was given an extra dose of 300 mg of lithium on the evening prior to surgery as well as his usual lithium dose the morning of surgery. (Figure 1) On the morning of surgery, the patient received diazepam 15 mg and methadone 15 mg orally two hours before surgery. Prior to induction of anesthesia, a right radial arterial catheter, pulmonary artery catheter with mixed venous oximetry, electrocardiogram with ST segment analysis, and pulse oximeter probe were placed; a cephalosporin antibiotic was administered without hemodynamic changes. Anesthesia was induced with 4.5 mg midazolam and 225 micro gram fentanyl in divided intravenous doses. Muscle relaxation was facilitated by 10 mg vecuronium; anesthesia maintenance included infusions of fentanyl and midazolam at rates of 0.1 and 1.0 micro gram centered dot kg-1 centered dot min-1, respectively. The pre-CPB period was notable for hemodynamic stability (blood pressure 130/50 mm Hg, heart rate 50-60 bpm, cardiac output 5.0 L/min); thus, additional means of blood pressure measurement were not used. Hematocrit (Hct) immediately preceding CPB was 33%. The CPB circuit was primed with normosol, albumin, mannitol, and tris (hydroxymethyl) aminomethane; no antibiotics were included in the pump prime. A mean arterial pressure (MAP) of 20-30 mm Hg (profound hypotension) with low calculated systemic vascular resistance was evident upon initiation of hypothermic CPB. This hypotension was refractory to repeated boluses of 100-500 micro gram of phenylephrine; the patient received a total of 8.5 mg of phenylephrine during the CPB period in an attempt to keep the MAP >or=to40 mm Hg. Total bypass time was 101 min, including an aortic cross clamp time of 50 min. Hct during CPB was 21%. CPB was terminated using small (renal)-dose dopamine (2.2 micro gram centered dot kg-1 centered dot min-1). Cardiac output remained between 5.6 and 6.6 L/min post-CPB. The right atrium was paced at 90 bpm, and pulmonary artery pressures remained approximately 30/15 mm Hg throughout the early postbypass period. Intermittent repeated boluses of phenylephrine of up to 800 micro gram every 15 min (a total of approximately 5 mg in the postbypass period) were administered to maintain an arterial blood pressure of 100/60 mm Hg secondary to low systemic vascular resistance (SVR). Intermittent boluses of calcium (200-400 mg) were given without any effect on SVR. Coronary perfusion and renal perfusion were not monitored via transesophageal echocardiography during the procedure, as there was no preoperative cardiac indication for transesophageal echocardiography in a patient with a normal ventricle and no evidence of valvular heart disease. The patient was transferred to the intensive care unit, where norepinephrine (0.04 micro gram centered dot kg-1 centered dot min-1) and epinephrine (0.03 micro gram centered dot kg-1 centered dot min-1) infusions were both required to maintain SVR. Catecholamines were gradually discontinued over the first postoperative day and the patient's trachea was extubated the morning after surgery. In spite of stable cardiovascular hemodynamics and successful endotracheal extubation, the patient's underlying renal insufficiency worsened and resulted in acute renal failure complicated by pneumonia requiring temporary reintubation. In spite of these difficulties, the patient was discharged home 1 mo after admission.Figure 1: The diagram illustrates the cellular mechanism of action of the alpha (1-adrenergic) receptor (alpha1 AR) agonists as well as the phosphoinositide cycle, and also attempts to demonstrate how lithium can interfere with the phosphoinositide cycle within the cell. An alpha1 AR agent (or other less selective AR agonist) stimulates the alpha1 AR receptor via an intermediary G protein, resulting in the hydrolysis of phosphatidylinositol 4,5-biphosphate (PI(4,5)]P2) by the enzyme phospholipase C (PLC). This generates the second messengers inositol 1,4,5-triphosphate (I(1,4,5)P3 or IP3) and diacylglycerol (DAG). IP3 results in release of calcium from intracellular pools such as the endoplasmic reticulum, resulting in smooth muscle contraction. Recycling of IP3 can proceed via two routes: phosphorylation by a 3-kinase enzyme to generate a possible second messenger, 1,3,4,5-tetraphosphate (I(1,3,4,5)P4), or dephosphorylation by a 5-phosphatase. Two of the following enzymes of the recycling pathway are noncompetitively inhibited by lithium: inositol poly-1-phosphatase, which dephosphorylates I(1,3,4)P3 and I(1,4)P2, and inositol monophosphatase, which dephosphorylates all the inositol phosphate isomers illustrated. PA = phosphatidic acid; CDP = cytidine diphosphate. Adapted from Nahorski et al. [13] with permission.Discussion The patient presented developed hypotension initially during CPB. It persisted into the post-CPB period and responded poorly to large doses of the direct-acting alpha1-agonist phenylephrine, as well as catecholamines epinephrine and norepinephrine. Hypotension during heart surgery with CPB can occur via several mechanisms. It is common immediately at the start of CPB due to hemodilution, but usually resolves spontaneously or responds to alpha1 AR agonists. Diabetic patients with peripheral neuropathy sometimes require large doses of phenylephrine during heart surgery, but there are no case reports of hypotension so profound that approximate 15 mg phenylephrine (as well as dopamine and epinephrine) has been required in this population; therefore, the profound degree of hypotension seen in the case presented cannot be explained solely by the presence of diabetes and its associated autonomic neuropathy. Two pharmacologic agents that may contribute to hypotension during surgery are Ca2+ entry blockers and ACE inhibitors. Hypotension secondary to Ca2+ channel antagonist during CPB responds to phenylephrine, as shown by there being no significant difference in the phenylephrine dose-response slopes between treated and untreated patients in this setting [15]. ACE inhibitor-associated hypotension during CPB remains controversial, with recent reviews and case studies reporting conflicting data; although decreases in SVR seems to occur in patients taking ACE inhibitors, these patients remain responsive to vasoconstrictor agents [16-18]. Other potential causes for low SVR during CPB include excessively low hematocrit and an acute allergic reaction (type-1 hypersensitivity). In this patient, CPB Hct was 21% (which is not considered excessively low during CPB), vancomycin was not administered, and no drugs were infused immediately prior to the onset of hypotension. We therefore suspect that chronic lithium therapy played a role in the failure of this patient to respond to catecholamine and vasoconstrictor agents. To understand how lithium might cause hypotension unresponsive to alpha1 AR agonists during CPB, the mechanisms of action of this drug are reviewed below. Many agonists acting at cell surface receptors alter cellular function by activating the enzyme phospholipase C, via the guanine nucleotide protien (G protein) Gq, resulting the hydrolysis of membrane phospholipids and the formation of two important cellular second messengers, inositol 1,4,5-triphosphate (IP3) and diacylglycerol [14]. IP3 is important in controlling intracellular calcium homeostasis by causing the release of Ca2+ from nonmitochondrial intracellular pools; diacylglycerol stimulates protein kinase C, a kinase that regulates a large number of intracellular processes. Increasing evidence suggests that lithium exerts its therapeutic action by interfering with the phosphoinositide pathway in the brain (as well as other tissues), preventing inositol recycling by noncompetitive inhibition of the enzyme inositol monophosphatase. Noncompetitive inhibition arises when an inhibitor binds only to an enzyme-substrate complex. In contrast, competitive antagonists interact with enzymes only. Berridge et al. [10] have suggested that lithium might reduce the supply of inositol required to sustain phosphoinositol synthesis; cells that contain higher concentrations of active enzymes (as is believed to be the case in manic-depressive disorders) are more susceptible to blockade via lithium's noncompetitive inhibitory activities. Another important aspect of the inhibitory action of lithium is that it is time dependent. Cells have a finite reserve of lipid that can be hydrolyzed to maintain signaling before blockade of phosphoinositide resynthesis becomes apparent. Balla et al. [19] have demonstrated that formation of IP3 in response to angiotensin II remains normal for five minutes after lithium therapy, but thereafter lithium markedly suppresses IP3 formation. As intracellular concentrations of IP3 decrease, mobilization of intracellular Ca2+ is impaired. Since IP (3) and calcium mobilization are important in maintaining vessel tone via diverse receptor systems such as the endothelin, adrenergic, and angiotensin II receptor systems, lithium therapy may result in end organ effects such as vasodilation. In addition to experiencing profound vasodilation, this patient was resistant to catecholamine vasoconstrictors, particularly the alpha1 AR selective agonist phenylephrine. alpha1 ARs mediate vasoconstriction via the hydrolysis of membrane phospholipids, forming IP3 and ultimately stimulating mobilization of intracellular calcium [20-22]. In the presence of lithium, very little substrate is available from alpha1 AR-mediated generation of IP3, decreasing alpha1 AR-mediated vasoconstriction, as demonstrated in this patient by ineffectiveness of the selective alpha1 AR agonist phenylephrine. Epinephrine and norepinephrine bind alpha1 and alpha2 ARs; since alpha2 AR-mediated vasoconstriction occurs via a different second messenger pathway (inhibition of adenylyl cyclase and/or ion channel activation), it is likely that increases in MAP with these agents (particularly after failure of alpha1 AR stimulation) occurred via vascular alpha2 AR activation. It is important to note that lithium is handled by the kidney in a manner similar to Na+. Therefore, a method for decreasing serum lithium levels, or total body lithium concentrations, includes diuresis--specifically osmotic diuresis. Since osmotic diuresis is frequently induced by mannitol added to the CPB pump prime, acute reductions in serum lithium concentrations during and after CPB may protect against generalized lithium-induced hypotension in patients with normal renal function. However, this patient had impaired renal function, which may have prevented or attenuated this protective response, particularly in the face of augmented lithium therapy the evening prior to surgery. Although large doses of alpha1 AR agonists can impair renal function, it is likely that, in this patient, smooth muscle in all vascular beds was resistant to alpha1 AR stimulation; therefore, acute postoperative worsening of renal function experienced by this patient probably occurred as a result of prolonged hypotension aggravating already compromised diabetic renal insufficiency and/or possibly the combination of ACE inhibitors and lithium as has recently been reported [23]. In conclusion, in this patient, mildly impaired vasoconstriction resulting from underlying diabetic neuropathy was likely complicated and exaggerated by lithium therapy, particularly in the setting of recent lithium supplementation and underlying impaired renal function. Exaggerated hypotension during and after CPB in patients on lithium therapy (particularly in the setting of impaired renal function) is an important clinical finding that has not been reported. Further studies documenting interactions between lithium and vasoactive drugs in vitro and in vivo will be required before any definitive recommendations can be made regarding the management of patients receiving lithium therapy during CPB. However, in light of this case report, withholding lithium therapy on the morning of surgery may be prudent, particularly in patients with renal insufficiency. In addition, lithium should be added to the list of potential inciting agents for hypotension in the peri-CPB period.