Secondary Logo

Share this article on:

Sodium Nitroprusside: Twenty Years and Counting

Friederich, Jeffrey A. MD; Butterworth, John F. IV, MD

Review Article

Department of Anesthesia, The Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, North Carolina.

Accepted for publication February 7, 1995.

Address correspondence and reprint requests to John F. Butterworth IV, MD, Department of Anesthesia, Bowman Gray School of Medicine, Medical Center Blvd., Winston-Salem, NC 27157-1009.

Almost from its first reported use in humans in 1928 [1], sodium nitroprusside (SNP) has been characterized as either villain or hero, depending on the clinician's assessment of the drug's risk-benefit ratio regarding cyanide toxicity. Not until 1955 was the safety of short-term SNP infusion established for treatment of severe hypertension [2]. Nevertheless, owing to difficulties with chemical preparation, SNP was not released for clinical use in the United States until 1974. Despite its immediate widespread adoption and apparent safety, concern remained about its potential for toxicity. Recently, the routine use of SNP in doses exceeding 2 micro gram centered dot kg-1 centered dot min-1 or at smaller doses when continued for >24 h has been challenged [3-5].

In 1991, the Food and Drug Administration, yielding to outside pressure as outlined by Robin and McCauley [5], approved new labeling for SNP which contained greater detail about the risk of cyanide toxicity, based on knowledge gained from clinical use and limited research since SNP's initial approval 17 yr earlier [6]. In addition to renewed warnings about SNP toxicity, several studies have examined whether prophylactic administration of various drugs to patients requiring SNP infusions would ameliorate or eliminate cyanide toxicity [7-11]. In short, although beset with controversy, SNP remains a commonly used drug, with total sales in the United States in 1993 (the latest year for which complete data are available) exceeding $2 million (Mr. Mark Sebree, Abbott Laboratories, personal communication, 1995). In this review we discuss the current indications and contraindications to SNP, evaluate its toxicity, and provide a plan for its rational use.

Back to Top | Article Outline

Chemical Properties and Mechanism of Action

SNP (disodium pentacyanonitrosylferrate(2-)dihydrate, Na2 Fe(CN) (5) NO centered dot 2H2 O) is comprised of a ferrous ion center complexed with five cyanide moieties and a nitrosyl group. The molecule is 44% cyanide by weight and soluble in water. Once infused, SNP can interact with oxyhemoglobin, dissociating immediately and forming methemoglobin while releasing cyanide and nitric oxide [12,13] Figure 1. Most investigators recognize nitric oxide (NO) as endothelium-derived relaxing factor, the active mediator responsible for the direct, vasodilating effect of SNP, although some would argue for a related nitrosothiol instead [14-16]. In contrast to the organic nitrates (e.g., nitroglycerin) which require the presence of highly specific thiol-containing compounds to generate NO, SNP spontaneously generates this product, thus functioning as a prodrug [14].

Figure 1

Figure 1

Once released, NO activates the enzyme guanylate cyclase found within vascular smooth muscle Figure 2, resulting in increased intracellular concentrations of cyclic guanosine monophosphate, which inhibits calcium entry into vascular smooth muscle cells and may increase calcium uptake by the smooth endoplasmic reticulum to produce vasodilation [14]. Recent, controversial evidence suggests that NO may possess a direct myocardial effect of unknown clinical significance [14,17].

Figure 2

Figure 2

Back to Top | Article Outline

Metabolism and Toxicology

SNP's spontaneous breakdown products, NO sup. and CN-, are rapidly cleared by nonenzymatic means, through interaction with sulfhydryl groups on proteins in surrounding tissue and in erythrocytes [2,12,13,18,19] Figure 1. Five cyanide radicals are released by each SNP molecule, which may immediately react with methemoglobin to produce cyanomethemoglobin [13]. Normal adult methemoglobin concentrations in blood (approximately 0.5% of all hemoglobin species) are capable of binding the cyanide released from 18 mg of SNP [8]. Cyanomethemoglobin remains in a dynamic equilibrium with free cyanide and is considered nontoxic [12,20].

Adverse effects from methemoglobinemia generated by SNP breakdown are rare, even in patients with congenital inability to convert methemoglobin to hemoglobin (i.e., methemoglobin reductase deficiency). The total SNP dose typically required to generate 10% methemoglobinemia exceeds 10 mg/kg (i.e., 10 micro gram centered dot kg-1 centered dot min-1 for more than 16 h) [18]. Patients receiving such doses of SNP who present with evidence of impaired oxygenation despite adequate cardiac output (CO) and PaO2 should have methemoglobinemia included in their differential diagnosis and measurement of methemoglobin via co-oximetry may be advisable [20-22].

The remaining cyanide radicals enter the "cyanide pool" and are converted to thiocyanate via transulfuration within the liver [8,12,23] Figure 1. Rhodanase (sometimes referred to as rhodanese, thiosulfate sulfurtransferase), the free cyanide-specific mitochondrial enzyme catalyzing this process, utilizes thiosulfate ions as sulfur donors [12,20,23]. The transulfuration reaction is theoretically reversible via the enzyme thiocyanate oxidase found in erythrocytes, but the thermodynamics greatly favor thiocyanate production [12,24]. Most normal adults can detoxify approximately 50 mg total of SNP using existing sulfur stores; factors that reduce these stores, e.g., malnutrition, surgery, and diuretics, decrease this capacity [7,11]. When SNP infusions exceed 2 micro gram centered dot kg-1 centered dot min-1, or when sulfur donors and methemoglobin are exhausted, cyanide radicals may accumulate producing clinical cyanide toxicity [7,11]. Assuming normal rhodanase activity, onset of toxicity may be minutes to hours depending on the above-mentioned sulfur stores.

Since any free cyanide radical may bind and inactivate tissue cytochrome oxidase and prevent oxidative phosphorylation, increased cyanide concentrations may precipitate tissue anoxia, anaerobic metabolism, and lactic acidosis [8,11]. Children may be less able to mobilize thiosulfate stores despite increasing cyanide concentrations, leading to accelerated toxicity [7]. The cytochrome oxidase-cyanide complex can interact with methemoglobin to form cyanomethemoglobin thus freeing cytochrome oxidase resulting in an apparent dynamic equilibrium between cyanomethemoglobin and cyanide [12] Figure 1. Cyanide gradually dissociates from methemoglobin, and is converted to thiocyanate [18]. There is considerable controversy regarding the true incidence of clinically significant cyanide toxicity. Thus, impassioned pleas to reduce the use of SNP are found concurrently with studies which demonstrate no evidence of clinical toxicity despite several days of SNP infusion [3-5,10,18].

Possible cyanide poisoning has been reported over a wide range of SNP infusion rates and total doses; deaths clearly linked to cyanide toxicity involve excessive total doses and infusion rates, i.e., 30-120 micro gram centered dot kg (-1) centered dot min-1 [18]. Blood cyanide concentrations required for clinical toxicity appear to exceed 40 micro Meter [11,25,26]; deaths have been reported with concentrations exceeding 77 micro Meter [11], 100 micro Meter [22], or 1309 micro Meter [27,28]. Nevertheless, many patients have survived blood cyanide concentrations, not necessarily as a consequence of nitroprusside therapy, which exceed the lethal range [25]. There is controversy as to the definitive method for assaying blood cyanide concentrations and the importance of photodecomposition during the assay or the infusion [11,25,29]. In any case, measurements of cyanide in blood are of limited clinical utility due to the time delay (ranging from several hours to days) before the results will be known [8].

Regardless of the SNP infusion rate or total administered dose, any patient receiving SNP who subsequently exhibits central nervous system dysfunction, cardiovascular instability, and increasing metabolic acidosis should be assessed for cyanide toxicity [8,11,18] Table 1. SNP infusion should be discontinued with further intervention based on the clinical assessment. Some authors believe that metabolic acidosis occurs only as a terminal event in cyanide toxicity and that cerebral dysfunction provides the more valuable diagnostic clue [5]. We are aware of no studies with SNP which substantiate this view; we believe that the literature supports a closer association of cyanide toxicity and metabolic acidosis [25,30,31].

Table 1

Table 1

When SNP infusion overwhelms the cyanide-removing capacity, the arterial base deficit correlates well with increased blood lactate concentration although not necessarily in critically ill patients [30,31]. Thus, base deficit, a readily available diagnostic test, may assist in the decision whether to initiate further therapy. In smoke inhalation, where cyanide toxicity also occurs, plasma lactate concentrations exceeding 10 mM correlate well with blood cyanide concentrations exceeding 40 micro Meter and may also provide an additional guide to interventions [20,25]. When life-threatening tissue hypoxia is present, having the patient breath 100% oxygen, correcting metabolic acidosis with sodium bicarbonate, and administering 3% sodium nitrite (4-6 mg/kg very slowly intravenously [IV]), and sodium thiosulfate (150-200 mg/kg IV over 15 min) is the usual treatment [8,11,12,18]. Sodium nitrite converts hemoglobin to methemoglobin which competes with cytochrome oxidase for cyanide radicals, as previously noted. Some clinicians will withhold sodium nitrite therapy in anemic patients, particularly when oxygen delivery is already compromised while others consider its vasodilating effects to be problematic [9,11,12].

Some practitioners administer hydroxocobalamin (vitamin B12a) to prevent or treat cyanide toxicity. Hydroxocobalamin binds CN- forming cyanocobalamin which acts as a nontoxic reservoir and can be excreted in the urine. The vitamin can be infused at 25 mg/h to a total of 100 mg or more during and after SNP infusion [9,12,18-20,25,27,32]. Infusion of vitamin B12 (cyanocobalamin) is considered ineffective in removing CN- due to poor binding, thus is not a substitute for hydroxocobalamin [9]. The principal toxic effect of hydroxocobalamin is reddish discoloration of the skin and mucous membranes [33]. The evidence that hydroxocobalamin administration is either needed or efficacious is controversial. One study of patients receiving SNP infusion at 2.5 micro gram centered dot kg-1 centered dot min (-1) for an average of 100 h showed no evidence of either cyanide toxicity or significant vitamin B12 deficiency despite average daily doses of 179 mg SNP [10]. Williams et al., using a dog model, showed hydroxocobalamin loading increased the minimum cyanide dose required to produce evidence of cyanide toxicity [34]. Of interest, hydroxocobalamin competitively inhibits SNP's relaxant effects and is quite expensive [35].

Perhaps the most commonly advocated prophylaxis for SNP toxicity is concomitant infusion of sodium thiosulfate to provide a continuous source of sulfur donors. Many studies and reviews advocate this method as the one sure way to prevent the accumulation of cyanide radicals regardless of the SNP dose [7,11,19,36]. Unfortunately, the thiocyanate formed can itself cause toxicity in patients with impaired renal thiocyanate excretion [8,11,12,18]. The normal thiocyanate elimination half-life is 2.7 days, but it may be prolonged up to 9 days in patients with renal insufficiency [11]. Thiocyanate was once evaluated as an antihypertensive drug [2,11,31]. Although thiocyanate was abandoned as an antihypertensive, these pilot studies provided an understanding of its toxic effects. Once formed, thiocyanate is eliminated intact via renal excretion (elimination half-life of 3-7 days). Clinical thiocyanate toxicity is rare, the drug being 100-fold less toxic than cyanide [8,11]. In patients with normal renal function, 7-14 days of SNP infusion in the 2-5 micro gram centered dot kg-1 centered dot min-1 range may be required to generate thiocyanate levels high enough to produce toxicity. SNP infusions as short as 3-6 days may prove toxic to patients with chronic renal failure not maintained on hemodialysis [8,11]. As might be anticipated, with normal renal function, increased water and chloride intake enhances thiocyanate excretion; hyponatremia and administration of thiosulfate favor thiocyanate accumulation [18].

Nonspecific symptoms of thiocyanate toxicity include fatigue, tinnitus, nausea, and vomiting. Clinical signs of thiocyanate neurotoxicity include hyperreflexia, confusion, psychosis, and miosis. Toxicity may progress to seizures and coma when thiocyanate concentrations exceed 60 mg/L. Life-threatening thiocyanate toxicity is of concern when blood thiocyanate concentrations approach 200 micro gram/mL [8,11,12,19,31]. Although serum thiocyanate levels are not helpful in diagnosing cyanide toxicity, they can confirm the diagnosis of thiocyanate toxicity [8,11,12,18]. Increased concentrations of thiocyanate competitively inhibit uptake and binding of iodine in the thyroid gland, sometimes producing clinical hypothyroidism [8,11,12,18]. Thiocyanate removal can be facilitated by hemodialysis or peritoneal dialysis [8,11,18].

Back to Top | Article Outline

Hemodynamic Effects

SNP produces direct venous and arterial vasodilation with preservation of adequate blood flow to all organs provided hypotension is avoided and arterial occlusive disease is absent. It is a potent pulmonary vasodilator and a direct inhibitor of hypoxic pulmonary vasoconstriction. It has an immediate onset and short half-life, dissipating in 1-2 min. In the setting of left ventricular failure, SNP reduces systemic vascular resistance (SVR), pulmonary vascular resistance, and right atrial pressures, while the effect on CO is dependent on initial left ventricular end-diastolic pressure Figure 3. With increased left ventricular end-diastolic pressure, CO is increased whereas with normal filling pressures CO may be unchanged. Control subjects without cardiovascular disease can experience a reduction in CO secondary to reduced preload [37] or an increase in heart rate and CO with adequate volume replacement Figure 4.

Figure 3

Figure 3

Figure 4

Figure 4

Back to Top | Article Outline

Uses and Indications

In 1991, the Food and Drug Administration updated the package insert for SNP, emphasizing the risk of cyanide toxicity and recommending that smaller doses be used [6]. The drug remains indicated for immediate reduction of arterial blood pressure in hypertensive emergencies and for induced hypotension during surgery to reduce operative bleeding [6]. Clearly, clinicians continue to use SNP for a wide range of other medical and surgical indications.

Back to Top | Article Outline

Hypertensive Emergencies

Since the early 1960s, SNP has been considered an option for treating hypertensive emergencies after failure of less potent drugs [31,38]. SNP possesses many favorable characteristics that have made it the mainstay of aggressive management of hypertensive crises [19]. Rapid onset, ease of titration, and rapid dissipation of its effects after discontinuation, and almost universal efficacy (regardless of the etiology of hypertension) contribute to the drug's popularity [18,19]. SNP typically acts within 30 s; its effects persist 2-4 min after cessation of infusion [18,19]. SNP has been used to successfully treat all forms of hypertensive emergencies, including malignant hypertension and hypertensive encephalopathy, whether caused by renovascular disease, pheochromocytoma, cerebral hemorrhage, aortic coarctation, postoperative status, toxemia of pregnancy, or "essential" hypertension [18,19,31,38,39]. Current opinion favors SNP used as a temporary measure with replacement by longer-lasting medications as soon as feasible [5,18].

Back to Top | Article Outline

Acute and Chronic Heart Disease

Use of SNP after acute myocardial infarction was first reported in 1972. These patients were treated with SNP infusions to reduce afterload, decrease the left-ventricular filling pressure, and increase the CO [19,31,40,41]. Subsequent studies have confirmed benefit from SNP infusion in severe mitral or aortic regurgitation, severe congestive heart failure, and after acute myocardial infarction complicated by left-ventricular failure [19,31,40,42-44]. In 1976, a new controversy arose concerning the effects of SNP on ischemic myocardium after acute myocardial infarction [19,40,45]. Increased ST segments on electrocardiogram were noted in patients treated with SNP shortly after the diagnosis of myocardial infarction; other patients treated with nitroglycerin demonstrated reduced ST segment signs of ischemia [45]. Subsequent work comparing the two drugs favored nitroglycerin for myocardial ischemia, due to emerging concern that SNP might produce coronary "steal" [40,46], which may occur when vasodilated small resistance vessels in the myocardium shunt blood away from areas of ischemia [47]. The clinical significance of coronary steal remains unclear; other reports have demonstrated improvement in ST segments after treatment with SNP [48]. In 1982, a study looking at the use of SNP in 812 patients with left-ventricular dysfunction after acute myocardial infarction determined that SNP worsened patient outcome when used within the first 8 h of an acute myocardial infarction. In patients with left-ventricular failure which persisted more than 8 h after myocardial infarction, SNP improved outcome [40,49]. In recent years, clinicians have tended to use nitroglycerin in preference to SNP after acute myocardial infarction; however, in these patients with continuing left-ventricular dysfunction, particularly in association with regurgitant valvular lesions, SNP can rapidly improve cardiac function and clinical stability, provided adequate ventricular filling is maintained [18,40,44].

In many studies of chronic congestive heart failure, regardless of the etiology, SNP has improved CO [18,19,48,50-53]. Enhancement of ventricular function is secondary to reduced impedance to ventricular ejection, permitting the ventricle to eject to a lower end-diastolic volume [18,19,53]. Preload is reduced by blood pooling in venous capacitance vessels, further contributing to decreased ventricular end-diastolic volume [18,19,53]. When filling pressures are increased, SNP will improve CO with only minimum increases in heart rate. If, despite adequate filling pressures, CO fails to improve during SNP therapy, then addition of an inotrope is indicated. The combination of SNP with dopamine, dobutamine, or epinephrine significantly improves cardiac performance [19,43,50,51] Figure 3.

Back to Top | Article Outline

Induced Intraoperative Hypotension

Purposeful hypotension is an anesthetic technique used in a variety of surgical procedures. Mean arterial pressures (MAP) of 50-60 mm Hg can be maintained in healthy patients without apparent complication [54,55]. The ability to rapidly, reversibly, and (somewhat) predictably titrate the infusion to achieve the desired MAP makes SNP nearly an ideal drug Figure 4. Orthopedic surgery, particularly major spinal operations, often involve considerable blood loss. Induced hypotension provides a relatively bloodless field and reduces transfusion requirements [6,18,19,54]. Concern regarding SNP toxicity, particularly in children, has prompted many anesthesiologists to reduce their reliance on SNP infusions by supplementing with beta-adrenergic receptor blockers, calcium channel blockers, nitroglycerin, or volatile anesthetics to maintain MAP within the targeted hypotensive range [19,54]. beta-Adrenergic receptor blockers work well for this purpose, but also decrease cardiac index (CI) and should be avoided in patients with impaired left-ventricular function. Calcium channel blockers, save for nicardipine, may also be inappropriate in reduced left-ventricular function [56,57]. Nitroglycerin may require excessive doses to achieve the target reduction in MAP, but it may be the more appropriate drug for patients with coronary artery disease. Volatile anesthetics, particularly isoflurane, are effective but have a more prolonged onset to achieve the desired MAP [54].

In neurosurgical patients, induced hypotension facilitates exposure and access during transsphenoidal hypophysectomy and intracranial aneurysm repair [18,19,31,54]. Concern arose in the late 1970s when there were reports of increased intracranial pressure during SNP infusion. However, subsequent reports showed that slowing the rate of SNP infusion, high inspired concentrations of oxygen, and hyperventilation appeared to eliminate this potential side effect in patients with intracranial hypertension [58]. Furthermore, SNP has been shown to have to direct effect on cerebral vasculature during cardiopulmonary bypass and indeed preserves cerebral autoregulation [59].

Back to Top | Article Outline

Aortic Surgery

Surgical repair of thoracic aortic aneurysms, dissections, and coarctations involves cross-clamping the aorta resulting in immediate, proximal hypertension, which unless treated is severe enough to produce cerebral edema. SNP, in conjunction with beta-adrenergic receptor blockers has been administered to aneurysm patients preoperatively for reduction of shearing forces within the dissecting lumen and intraoperatively to decrease proximal MAP. SNP has largely replaced trimethaphan as the drug of choice in this setting, due to concern over trimethaphan's side effects and tendency for tachyphylaxis. A possible contribution of SNP to spinal cord ischemia, a devastating complication of aortic surgery, has prompted rigorous assessment of SNP's effects on spinal cord blood flow. When used to control proximal (to the aortic clamp) hypertension, SNP may aggravate hypotension distal to the clamp. This has led many investigators to advocate a target MAP (proximal to the clamp) just below the upper limits of cerebral autoregulation to maintain spinal cord perfusion [60]. Children undergoing aortic surgery and receiving SNP may possibly be more prone to accelerated toxicity as previously stated [7].

Back to Top | Article Outline


Safe resection of pheochromocytomas remains a challenge for surgeons and anesthesiologists, due to the extreme intraoperative lability of blood pressure. Standard preoperative management includes alpha-adrenergic receptor blockade with oral phenoxybenzamine or prazosin over 1-2 wk [61]. Slow initiation of alpha-adrenergic receptor blockade permits restoration of intravascular volume, typically reduced in patients with untreated pheochromocytomas. Occasionally in patients with apparently adequate alpha-blockade prior to surgery, and frequently in patients with untreated pheochromocytomas, excessive hypertension occurs during resection. This hypertension can be controlled with either IV phentolamine, SNP, or both. Some clinicians prefer to rely more heavily on SNP, the effects of which dissipate more rapidly than those of phentolamine, to avoid the postoperative hypotension frequently complicating pheochromocytoma resection [61].

Back to Top | Article Outline

Uses in Cardiac Surgery

Intraoperative Hypertension Systemic hypertension is frequent during cardiac surgery, particularly after coronary artery bypass grafting or replacement of a stenotic aortic valve [51,62,63]. Although characterized by increased SVR, filling pressures, and heart rate, the mechanisms responsible for the hypertension remain unclear. Various theories have proposed that activation of the renin-angiotensin system and increased concentration of circulating catecholamines are major contributors to the increased pressure [52]. These hormonal responses to cardiopulmonary bypass and surgery are reviewed in detail in a recent text [64]. Regardless of the etiology, SNP is often chosen to reduce this hypertension [51,62,63]. Alternative methods of blood pressure control during coronary artery bypass surgery include increasing the depth of anesthesia (with either volatile anesthetics, IV opioids, or IV sedatives), calcium channel blockers, beta-adrenergic receptor blockers, and nitroglycerin [51,62,63,65].

Postoperative Hypertension After successful cardiac surgery, hypertension remains common, with an incidence variably reported to range between 30% and 60% [66-68]. Early after separation from cardiopulmonary bypass, increased SVR and hypertension can depress left-ventricular function, increase myocardial oxygen consumption, predispose patients for postoperative bleeding, and increase the incidence of arrhythmias [51,66,69]. SNP remains a common treatment for postoperative hypertension either alone or in combination with other drugs and its continued use despite newer drugs and ongoing controversy may be a reflection of its efficacy and ease of use [19,52,66,70-74]. Moreover, no other drug has proven to be superior to SNP in this setting. As noted earlier, calcium channel blockers and beta-adrenergic receptor blockers cause more myocardial depression with the exception of nicardipine which may prove equally efficacious when filling pressures are not increased [57]. Longacting vasodilators are difficult to titrate acutely, and nitroglycerin often requires excessive doses to achieve the desired blood pressure reduction and may decrease CO [62,66,68,71-77].

The efficacy of SNP as a pulmonary vasodilator is well known. SNP has been used to treat postoperative pulmonary hypertension after heart valve replacement, and may be the drug of choice for these patients in the setting of systemic hypertension and increased filling pressures [57,78]. Inhaled NO is currently under investigation and may be a promising replacement in this patient population [79,80].

In pediatric patients with low CI (CI < 2.0 L centered dot min-1 centered dot m-2) after cardiopulmonary bypass, despite adequate filling pressures, SNP increases CI, and if combined with epinephrine additional increases in CI are noted without significant increase in systemic vascular resistance [51] Figure 3.

Back to Top | Article Outline

Miscellaneous Uses

Increased intraocular pressure during ophthalmic surgery is a concern for the anesthesiologist, particularly during intubation and extubation. A sudden increase in intraocular pressure can produce permanent loss of vision secondary to hemorrhage, vitreous extrusion (with an "open" eye injury), or prolapse of the iris and lens. SNP effectively controls arterial hypertension without causing or increasing intraocular hypertension [81]. Bolus doses of 1-2 micro gram/kg can be given to decrease blood pressure during laryngoscopy and extubation [82].

Back to Top | Article Outline


Medical contraindications to SNP use are rare and largely theoretical. It is possible that the risk for cyanide toxicity may be increased in individuals with abnormal cyanide-thiocyanate pathways or decreased availability of hepatic rhodanase enzyme. Generations of SNP reviews have stated that patients with congenital Leber's optic atrophy (LOA) or tobacco amblyopia (TA) should be considered candidates for alternative drugs [18,19,31]. However, these previous publications have based their statements on a case report published in 1975 involving the death of a 14-yr-old male who received 400 mg (10 mg/kg) of SNP over 80 min. At autopsy 20 h after death, thiocyanate could not be detected in the blood or urine. The authors surmised that the absence of thiocyanate at autopsy implied an abnormality of cyanide metabolism and, since LOA and TA were the only known pathologic states with this same apparent abnormality, they declared them contraindications to SNP use [83,84]. They based this suggestion on a report published in 1965 in which patients with LOA were shown to have a possible defect in cyanide metabolism based on increased cyanide concentrations from tobacco smoking. Of interest, these LOA patients all had detectable concentrations of both plasma and urine thiocyanate (>50% of control concentrations), as well as adequate rhodanase activity in liver biopsy tissue [85]. Recent publications on LOA point to an apparent multitude of genetic defects causing an overall decrease in cell energy production resulting in the disease, rather than a specific enzyme defect [86-88]. We are aware of no reported cases of adverse effects related to actual SNP use in patients diagnosed with either LOA or TA. Likewise, no data exists to our knowledge showing that SNP use in these patients is safe. Prudence would dictate close monitoring of patients with LOA or TA when SNP is deemed necessary and early consideration of alternative therapy.

We believe that patients with hepatic or renal failure are at greater risk for cyanide toxicity and that this risk is actual, rather than theoretical. Patients with compensatory hypertension from a primary vascular lesion, e.g., untreated coarctation of the aorta, and those patients with known inadequate cerebral perfusion, e.g., dangerously increased intracranial pressure or symptomatic carotid artery stenosis, are also considered to be inappropriate candidates for SNP therapy [18,89].

Caution in the use of SNP in patients undergoing cardiopulmonary bypass was advised due to the theoretical possibility of enhanced bleeding from increased bleeding time and decreased platelet aggregation [90,91]. SNP appears to impair platelet actions via NO [91]. The clinical importance of this effect has been questioned, since no differences in blood product requirements, mediastinal drainage, or postoperative coagulation studies were noted in a retrospective study of cardiac surgical patients with postoperative hypertension treated with SNP [92]. The duration of the SNP platelet effect is apparently 5-25 min [93].

SNP should be used cautiously in pregnant patients due to the potential for fetal cyanide toxicity [18,89,94]. However, this is not a major concern when modest SNP doses (<3 micro gram centered dot kg-1 centered dot min-1) are administered for short durations; SNP continues to be used for acute afterload reduction when indicated in parturients [94].

Back to Top | Article Outline

Dose and Administration

The manufacturers of SNP currently recommend that therapy be initiated by IV infusion at an initial rate of 0.3 micro gram centered dot kg-1 centered dot min-1 titrated to a maximum rate of 10 micro gram centered dot kg-1 centered dot min-1, with the maximum rate not to be infused longer than 10 min [6,89]. An indwelling arterial catheter for continuous blood pressure monitoring is recommended [18]. Delivery is recommended by infusion pump with manual titration or via computerized closed-loop titration system [67]. IV boluses of 1-2 ng/kg have also been recommended for rapid lowering of blood pressure, i.e., aneurysm clipping, direct laryngoscopy [82]. SNP should be mixed only in 5% dextrose in water [89]. It is important that the solution be protected from light continuously due to the rapid conversion of SNP to aquapentacyanoferrate ion, [Fe(CN)sH2 O]2-, which would readily cause release of hydrogen cyanide in the patient [29].

A healthy person can eliminate cyanide hepatically at a rate equivalent to the cyanide production during a SNP infusion of approximately 2 micro gram centered dot kg-1 centered dot min-1 [8,9,11]. SNP administered at rates higher than 2 micro gram centered dot kg-1 centered dot min-1 therefore results in dosedependent accumulation of cyanide radicals. Unless the clinician concurrently infuses sodium thiosulfate or hydroxocobalamin with SNP infusion at rates exceeding 2 micro gram centered dot kg-1 centered dot min-1, the dose-dependent risk of cyanide toxicity must be considered [6,8,9,11,89] Table 1.

One potentially useful way of avoiding toxicity from SNP is to combine SNP with another drug, e.g., trimetaphan or a beta-adrenergic receptor blocker. MacRae et al. [95] induced hypotension with a 10:1 mixture of trimetaphan:SNP. Controlled hypotension was produced easily with considerable reduction in drug doses compared to the two drugs administered separately [95,96]. These same authors also observed that the dose of trimetaphan:SNP required for controlled hypotension could be further reduced by oral administration of a beta-adrenergic receptor blocker prior to induction of anesthesia [97]. These studies illustrate several ways that the clinician might prevent SNP toxicity by coadministering it with other drugs so as to reduce the total SNP dose.

Back to Top | Article Outline


SNP remains an effective, reliable, and commonly used drug for the rapid reduction of significant arterial hypertension regardless of the etiology, for afterload reduction in the face of low CO when blood volume is normal or increased, and for intraoperative induced hypotension. After establishing indwelling arterial monitoring, an initial infusion rate of 0.3-0.5 micro gram centered dot kg-1 centered dot min-1 is begun with titration as needed up to 2.0 micro gram centered dot kg (-1) centered dot min-1. Higher rates for brief periods of time (10 min) are acceptable. The use of alternative drugs to reduce the dose or shorten the duration of infusion should be considered when the 2.0 micro gram centered dot kg-1 centered dot min-1 range is exceeded Table 1.

SNP should not be used by individuals unfamiliar with its potency and metabolic pathways, as the many reports of adverse reactions testify. Careful attention to infusion rates, particularly in patients at risk for depleted thiosulfate stores, is mandatory, and the use of other drugs in conjunction with or instead of SNP should always be considered. As with many therapeutic interventions, SNP requires careful administration to appropriately selected patients by a clinician who knows its inherent hazards. Despite its toxicity, SNP is popular because it is often the most (in some cases, the only) effective drug in some difficult clinical circumstances.

Back to Top | Article Outline


1. Johnson CC. Mechanisms of actions and toxicity of nitroprusside. Proc Soc Exp Biol Med 1928;26:102-3.
2. Page IH, Corcoran AC, Dustan HP, Koppanyi T. Cardiovascular actions of sodium nitroprusside in animals and hypertensive patients. Circulation 1955;11:188-98.
3. Patel CB, Laboy V, Venus B, et al. Use of sodium nitroprusside in post-coronary bypass surgery. A plea for conservatism. Chest 1986;89:663-7.
4. Sarvotham SS. Nitroprusside therapy in post-open heart hypertensives. A ritual tryst with cyanide death? [letter]. Chest 1987;91:796.
5. Robin ED, McCauley R. Nitroprusside-related cyanide poisoning. Time (long past due) for urgent, effective interventions [special report]. Chest 1992;102:1842-5.
6. The Food and Drug Administration. New labeling for sodium nitroprusside emphasized risk of cyanide toxicity. JAMA 1991;265:847.
7. Ivankovich AD, Braverman B, Stephens TS, et al. Sodium thiosulfate disposition in humans: relation to sodium nitroprusside toxicity. Anesthesiology 1983;58:11-17.
8. Hall VA, Guest JM. Sodium nitroprusside-induced cyanide intoxication and prevention with sodium thiosulfate prophylaxis. Am J Crit Care 1992;1:19-27.
9. Zerbe NF, Wagner BK. Use of vitamin B12 in the treatment and prevention of nitroprusside-induced cyanide toxicity. Crit Care Med 1993;21:465-7.
10. Lipman J, Hesdorffer C, Fernandes Costa F, et al. Vitamin B12 levels in the prolonged use of sodium nitroprusside. Crit Care Med 1984;12:161-3.
11. Curry SC, Arnold-Capell P. Toxic effects of drugs used in the ICU. Nitroprusside, nitroglycerin, and angiotensin-converting enzyme inhibitors. Crit Care Clin 1991;7:555-81.
12. Ivankovich AD, Miletich DJ, Tinker JH. Sodium nitroprusside: metabolism and general considerations. Int Anesthesiol Clin 1978;16:1-29.
13. Smith RP, Kruszyna H. Nitroprusside produces cyanide poisoning via reaction with hemoglobin. J Pharmacol Exp Ther 1974;191:557-63.
14. Moncada S, Palmer RM, Higgs EA. Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev 1991;43:109-42.
15. Palmer RM, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 1987;327:524-6.
16. Myers PR, Minor RL Jr, Guerra R Jr, et al. Vasorelaxant properties of the endothelium-derived relaxing factor more closely resemble S-nitrosocysteine than nitric oxide. Nature 1990;345:161-3.
17. Paulus WJ, Vantrimpont PJ, Shah AM. Acute effects of nitric oxide on left ventricular relaxation and diastolic distensibility in humans. Circulation 1994;89:2070-8.
18. McEvoy GK, ed. American Hospital Formulary Service, Drug Information 1994;24:08, 1163-6.
19. Cohn JN, Burke LP. Nitroprusside. Ann Intern Med 1979;91:752-7.
20. Kulig K. Cyanide antidotes and fire toxicology. N Engl J Med 1991;325:1801-2.
21. Bower PJ, Peterson JN. Methemoglobinemia after sodium nitroprusside therapy. N Engl J Med 1975;293:865.
22. Langford RM, Armstrong RF. Algorithm for managing injury from smoke inhalation. Br Med J 1989;299:902-5.
23. Himwich WA, Saunders JP. Enzymatic conversion of cyanide to thiocyanate. Am J Physiol 1948;153:348.
24. Goldstein F, Rieders F. Conversion of thiocyanate to cyanide by an erythrocytic enzyme. Am J Physiol 1953;173:287.
25. Baud FJ, Barriot P, Toffis V, et al. Elevated blood cyanide concentrations in victims of smoke inhalation. N Engl J Med 1991;325:1761-6.
26. Schulz V, Gross R, Pasch T, et al. Cyanide toxicity of sodium nitroprusside in therapeutic use with and without sodium thiosulfate. Klin Wochenschr 1982;60:1393-1400.
27. Fahmy NR. Consumption of vitamin B12 during sodium nitroprusside administration in humans. Anesthesiology 1981;54:305-9.
28. Gettler AD, Baine JO. The toxicology of cyanide. Am J Med Sci 1938;195:182-98.
29. Bisset WI, Butler AR, Glidewell C, Reglinski J. Sodium nitroprusside and cyanide release: reasons for re-appraisal. Br J Anaesth 1981;53:1015-8.
30. Simpson PJ, Adams L, Vesey CJ, Cole P. Some physiological and metabolic effects of sodium nitroprusside and cyanide in the dog. Br J Anaesth 1979;51:81-7.
31. Tinker JH, Michenfelder JD. Sodium nitroprusside: pharmacology, toxicology, and therapeutics. Anesthesiology 1976;45:340-54.
32. Cottrell JE, Casthely P, Brodie JD, et al. Prevention of nitroprusside-induced cyanide toxicity with hydroxocobalamin. N Engl J Med 1978;298:809-11.
33. Hall AH, Rumack BH. Hydroxocobalamin/sodium thiosulfate as a cyanide antidote. J Emerg Med 1987;5:115-21.
34. Williams HL, Johnson DJ, McNeil JS, Wright DG. Studies of cobalamin as a vehicle for the renal excretion of cyanide anion. J Lab Clin Med 1990;116:37-44.
35. Li CG, Rand MJ. Effects of hydroxocobalamin and haemoglobin on NO-mediated relaxations in the rat anococcygeus muscle. Clin Exp Pharmacol Physiol 1993;20:633-40.
36. Lundquist P, Rosling H, Tyden H. Cyanide release from sodium nitroprusside during coronary bypass in hypothermia. Acta Anaesthesiol Scand 1989;33:686-8.
37. Pouleur H, Covell JW, Ross J Jr. Effects of nitroprusside on venous return and central blood volume in the absence and presence of acute heart failure. Circulation 1980;61:328-37.
38. Ahearn DJ, Grim CE. Treatment of malignant hypertension with sodium nitroprusside. Arch Intern Med 1974;133:187-91.
39. Paull J. Clinical report of the use of sodium nitroprusside in severe preeclampsia. Anaesth Intensive Care 1975;3:72.
40. Francis GS. Vasodilators in the intensive care unit. Am Heart J 1991;121:1875-8.
41. Franciosa JA, Limas CJ, Guiha NH, et al. Improved left ventricular function during nitroprusside infusion in acute myocardial infarction. Lancet 1972;1:650-4.
42. Warner RA, Bowser M, Zuehlke S, et al. Treatment of acute aortic insufficiency with sodium nitroferricyanide. Documentation of beneficial effect by noninvasive means. Chest 1977;72:375-9.
43. Miller RR, Awan MA, Joye JA, et al. Combined dopamine and nitroprusside therapy in congestive heart failure. Greater augmentation of cardiac performance by addition of inotropic stimulation to afterload reduction. Circulation 1977;55:881-4.
44. Grossman W, Harshaw CW, Munro AB, et al. Lowered aortic impedance as therapy for severe mitral regurgitation. JAMA 1974;230:1011-3.
45. Chiariello M, Gold HK, Leinbach RC, et al. Comparison between the effects of nitroprusside and nitroglycerin on ischemic injury during acute myocardial infarction. Circulation 1976;54:766-73.
46. Fremes SE, Weisel RD, Mickle DA, et al. A comparison of nitroglycerin and nitroprusside. I. Treatment of postoperative hypertension. Ann Thorac Surg 1985;39:53-60.
47. Becker LC. Conditions for vasodilator-induced coronary steal in experimental myocardial ischemia. Circulation 1978;57:1103-10.
48. Yeh BK, Gosselin AJ, Swaye PS, et al. Sodium nitroprusside as a coronary vasodilator in man. Am Heart J 1977;93:610-6.
49. Cohn JN, Franciosa JA, Francis GS, et al. Effect of short-term infusion of sodium nitroprusside on mortality rate in acute myocardial infarction complicated by left ventricular failure: results of a Veterans Administration cooperative study. N Engl J Med 1982;306:1129-35.
50. Uretsky BF, Hua J. Combined intravenous pharmacotherapy in the treatment of patients with decompensated congestive heart failure. Am Heart J 1991;121:1879-86.
51. Benzing G III, Helmsworth JA, Schreiber JT, Kaplan S. Nitroprusside and epinephrine for treatment of low output in children after open-heart surgery. Ann Thorac Surg 1979;27:523-8.
52. Roberts AJ, Niarchos AP, Subramanian VA, et al. Systemic hypertension associated with coronary artery bypass surgery. Predisposing factors, hemodynamic characteristics, humoral profile, and treatment. J Thorac Cardiovasc Surg 1977;74:846-59.
53. Guiha NH, Cohn JN, Mikulic E, et al. Treatment of refractory heart failure with infusion of nitroprusside. N Engl J Med 1974;291:587-92.
54. Black S, Cucchiara RF. Tumor surgery. In: Cucchiara RF, Michenfelder JD, eds. Clinical neuroanesthesia. New York: Churchill Livingstone, 1990:285-308.
55. Michenfelder JD, Theye RA. Canine systemic and cerebral effects of hypotension induced by hemorrhage, trimethaphan, halothane or nitroprusside. Anesthesiology 1977;46:188-95.
56. Franklin C, Nightingale S, Mamdani B. A randomized comparison of nifedipine and sodium nitroprusside in severe hypertension. Chest 1986;90:500-3.
57. David D, Dubois C, Loria Y. Comparison of nicardipine and sodium nitroprusside in the treatment of paroxysmal hypertension following aortocoronary bypass surgery. J Cardiothorac Vasc Anesth 1991;5:357-61.
58. Marsh ML, Aidinis SJ, Naughton KV, et al. The technique of nitroprusside administration modifies the intracranial pressure response. Anesthesiology 1979;51:538-41.
59. Rogers AT, Prough DS, Gravlee GP, et al. Sodium nitroprusside infusion does not dilate cerebral resistance vessels during hypothermic cardiopulmonary bypass. Anesthesiology 1991;74:820-6.
60. Gelman S, Reves JG, Fowler K, et al. Regional blood flow during cross-clamping of the thoracic aorta and infusion of sodium nitroprusside. J Thorac Cardiovasc Surg 1983;85:287-91.
61. Shapiro B, Figure LM. Management of pheochromocytoma. Endocrinol Metab Clin North Am 1989;18:443-70.
62. van Wezel HB, Bovill JG, Visser CA, et al. Myocardial metabolism, catecholamine balance, and left ventricular function during coronary artery surgery: effects of nitroprusside and nifedipine. J Cardiothorac Anesth 1987;1:408-17.
63. Hess W, Schulte-Sasse U, Tarnow J. Nifedipine versus nitroprusside for controlling hypertensive episodes during coronary artery bypass surgery. Eur Heart J 1984;5:140-5.
64. Butterworth JF IV, Utley JR, Swain JA. Neuroendocrine and electrolyte responses to cardiopulmonary bypass. In: Gravlee GP, Utley JR, Davis RF, eds. Principles and practices of cardiopulmonary bypass. Baltimore: Williams & Wilkins, 1993:291-322.
65. Tobias MA. Comparison of nitroprusside and nitroglycerin for controlling hypertension during coronary artery surgery. Br J Anaesth 1981;53:891-7.
66. Zall S, Kirno K, Milocco I, Ricksten SE. Vasodilation with adenosine or sodium nitroprusside after coronary artery bypass surgery: a comparative study on myocardial blood flow and metabolism. Anesth Analg 1993;76:498-503.
67. Bednarski P, Siclari F, Voigt A, et al. Use of a computerized closed-loop sodium nitroprusside titration system for antihypertensive treatment after open heart surgery. Crit Care Med 1990;18:1061-5.
68. Flaherty JT, Magee PA, Gardner TL, et al. Comparison of intravenous nitroglycerin and sodium nitroprusside for treatment of acute hypertension developing after coronary artery bypass surgery. Circulation 1982;65:1072-7.
69. Arom KV, Angaran DM, Lindsay WG, et al. Effect of sodium nitroprusside during the payback period of cardiopulmonary bypass on the incidence of postoperative arrhythmias. Ann Thorac Surg 1982;34:307-12.
70. Bixler TJ, Gardner TJ, Donahoo JS, et al. Improved myocardial performance in postoperative cardiac surgical patients with sodium nitroprusside. Ann Thorac Surg 1978;25:444-8.
71. Halpern NA, Goldberg M, Neely C, et al. Postoperative hypertension: a multicenter, prospective, randomized comparison between intravenous nicardipine and sodium nitroprusside. Crit Care Med 1992;20:1637-43.
72. Boylan JF, O'Leary G, Weisel RD, et al. A comparison of diltiazem, esmolol, nifedipine and nitroprusside therapy of post-CABG hypertension. Can J Anaesth 1990;37:S156.
73. Nathan HJ, Laganiere S, Dube L, et al. Intravenous nifedipine to treat hypertension after coronary artery revascularization surgery. A comparison with sodium nitroprusside. Anesth Analg 1992;74:809-17.
74. Cruise CJ, Skrobik Y, Webster RE, et al. Intravenous labetalol versus sodium nitroprusside for treatment of hypertension postcoronary bypass surgery. Anesthesiology 1989;71:835-9.
75. Gray RJ, Bateman TM, Czer LS, et al. Comparison of esmolol and nitroprusside for acute post-cardiac surgical hypertension. Am J Cardiol 1987;59:887-91.
76. Mullen JC, Miller DR, Weisel RD, et al. Postoperative hypertension: a comparison of diltiazem, nifedipine, and nitroprusside. J Thorac Cardiovasc Surg 1988;96:122-32.
77. Stinson EB, Holloway EL, Derby G, et al. Comparative hemodynamic responses to chlorpromazine, nitroprusside, nitroglycerin, and trimethaphan immediately after open-heart operation. Circulation 1975;51-52(Suppl 1):I26-33.
78. Goenen MJ, Leenaert L, Petein M, et al. The effects of tolazoline, nitroprusside, nitroglycerin, isoproterenol, and hydralazine on pulmonary circulation early after heart valve replacement. Thorac Cardiovasc Surg 1982;30:253-8.
79. Rossaint R, Gerlach H, Falke KJ. Inhalation of nitric oxide--a new approach to severe ARDS. Eur J Anaesthesiol 1994;11:43-51.
80. Winberg P, Lundell BP, Gustafsson LE. Effect of inhaled nitric oxide on raised pulmonary vascular resistance in children with congenital heart disease. Br Heart J 1994;71:282-6.
81. Elliott WJ, Karnezis TA, Silverman RA, et al. Intraocular pressure increases with fenoldopam, but not nitroprusside, in hypertensive humans. Clin Pharmacol Ther 1991;49:285-93.
82. Stoelting RK. Attenuation of blood pressure response to laryngoscopy and tracheal intubation with sodium nitroprusside. Anesth Analg 1979;58:116-9.
83. Davies DW, Kadar D, Steward DJ, Munro IR. A sudden death associated with the use of sodium nitroprusside for induction of hypotension during anaesthesia. Can Anaesth Soc J 1975;22:547-52.
84. Davies DW, Greiss L, Kadar D, Steward DJ. Sodium nitroprusside in children: observations on metabolism during normal and abnormal responses. Can Anaesth Soc J 1975;22:553-60.
85. Wilson J. Leber's hereditary optic atrophy: a possible defect of cyanide metabolism. Clin Sci 1965;29:505-15.
86. Volpe NJ, Lessell S. Leber's hereditary optic neuropathy. Int Ophthalmol Clin 1993;33:153-68.
87. Johns DR, Neufeld MJ. Cytochrome c oxidase mutations in Leber hereditary optic neuropathy. Biochem Biophys Res Commun 1993;196:810-5.
88. Brown MD, Voljavec AS, Lott MT, et al. Leber's hereditary optic neuropathy: a model for mitochondrial neurodegenerative diseases. FASEB J 1992;6:2791-9.
89. Sodium nitroprusside. Physicians desk reference. 47th ed. Montvale, NJ: Medical Economics Data, 1993:1999-2001.
90. Hines R, Barash PG. Infusion of sodium nitroprusside induces platelet dysfunction in vitro. Anesthesiology 1989;70:611-5.
91. Maurice DH, Haslam RJ. Molecular basis of the synergistic inhibition of platelet function by nitrovasodilators and activators of adenylate cyclase: an inhibition of cyclic AMP breakdown by cyclic GMP. Adv Second Messenger Phosphoprotein Res 1988;22:1-38.
92. Snow N, Lucas A, Gray LA Jr. Effect of sodium nitroprusside on postoperative blood loss in the cardiac surgical patient. Crit Care Med 1981;9:827-8.
93. Rovin JD, Stamler JS, Loscalzo J, Fults JD. Sodium nitroprusside, an endothelium-derived relaxing factor congener, increases platelet cyclic GMP levels and inhibits epinephrine-exacerbated in vivo platelet thrombus formation in stenosed canine coronary arteries. J Cardiovasc Pharmacol 1993;22:626-31.
94. Wright RG, Shnider SM, Fong CJ. Fetal and neonatal effects of maternally administered drugs [appendix]. In: Shnider SM, Levinson G, eds. Anesthesia for obstetrics. 3rd ed. Baltimore: Williams & Wilkins, 1993:714.
95. MacRae WR, Wildsmith JA, Dale BA. Induced hypotension with a mixture of sodium nitroprusside and trimetaphan camsylate. Anaesthesia 1981;36:312-5.
96. Wildsmith JA, Drummond GB, MacRae WR. Metabolic effects of induced hypotension with trimetaphan and sodium nitroprusside. Br J Anaesth 1979;51:875-9.
97. Simpson DL, MacRae WR, Wildsmith JA, Dale BA. Acute beta-adrenoreceptor blockade and induced hypotension. Anaesthesia 1987;42:243-8.
© 1995 International Anesthesia Research Society