Intraoperative Hyperkalemia as a Triggering Mechanism or Presenting Sign of Malignant Hyperthermia in Two Patients with Chronic Renal Failure

A 53-yr-old man with a history of hypertension, juvenile-onset diabetes mellitus, and ESRD was scheduled for cadaveric renal transplant. He also reported general fatigue and frequent muscle cramps in the lower extremities, which were treated with quinine. The patient's medications included 22 units of NPH insulin in the morning and afternoon, 0.2 mg clonidine daily, 2 mg doxazosin twice daily, 60 mg nifedipine twice daily, 0.25 micro gram calcitriol daily, and 260 mg quinine twice daily. The patient had had three previous surgeries, all of which were uneventful in regard to anesthesia. One was performed to place a peritoneal dialysis catheter (general anesthesia: sodium thiopental, isoflurane/nitrous oxide, fentanyl, and vecuronium); the catheter was later removed under monitored anesthesia care with midazolam and fentanyl sedation. The third surgery was performed with regional anesthesia (brachial plexus block) to create an arteriovenous fistula. His family history revealed that his sister has "occasional increases in body temperature even after minor procedures, such as dental work" (this information was obtained after the event).

This patient received hemodialysis three times per week and underwent hemodialysis before surgery. Postdialysis electrolyte concentrations were sodium 141 mM/l, potassium 4.7 mM/l, chloride 96 mM/l, bicarbonate 22 mM/l, calcium 9.5 mg/dL, phosphate 5.3 mg/dL; blood urea nitrogen 81 mg/dL, creatinine 13.1 mg/dL, and glucose 80 mg/dL. The electrocardiogram results were normal. Preoperatively, he received his regular oral medications as above, including nifedipine and 12 U of NPH insulin. Standard noninvasive monitors and a central venous catheter were placed. Before anesthesia induction, blood pressure was 175/90 mmHg and heart rate was 72 beats/min. Anesthesia was induced with 325 mg sodium thiopental, 150 micro gram fentanyl, and 40 mg atracurium and maintained with 1.0%-1.5% isoflurane in a mixture of oxygen and nitrogen, 30%/70%. Post-induction blood pressure and heart rate were 110/65 mmHg and 70 beats/min, respectively. The lungs were mechanically ventilated with a tidal volume of 650 ml at a respiratory rate of 10 breaths/min. The end-tidal carbon dioxide was 26 mmHg. Esophageal temperature was 35.2 degrees Celsius. To prevent intraoperative hypothermia, a forced-air patient warming system (Bair-Hugger, Augustine Medical, Eden Prairie, MN) was used.

About 30 min after anesthesia induction, electrolyte concentrations were sodium 137 mM/l, potassium 7.2 mM/l, and glucose 83 mg/dL. To exclude hemolysis as a possible cause of the unexpected hyperkalemia, a second blood sample was analyzed. The repeat potassium concentration was 7.1 mM/l (Figure 1). At the same time, end-tidal carbon dioxide was 27 mmHg, and the body temperature was 35.4 degrees Celsius. To reduce hyperkalemia, we initiated hyperventilation (respiratory rate was increased to 15 breaths/min) followed by a continuous infusion of glucose and insulin. After 30 min the increased respiratory rate had not affected end-tidal carbon dioxide, and the tidal volume was increased to 800 ml. Venous blood gas measurements were pH 7.34, Pa^CO245 mmHg and bicarbonate 24 mM/l; potassium was 6.4 mM/l and glucose 241 mg/dL. Despite increased ventilation, end-tidal carbon dioxide increased to 32 mmHg. The slightly increased body temperature (37.2 degrees Celsius) was attributed to effective heat preservation, and the warming system was turned off. Four hours after anesthesia induction, a surgeon injected 5 mg verapamil into the renal artery. Fifteen minutes later, ventilation difficulties occurred: Peak inspiratory pressure increased from 32 to 48 cmH²O, but there was no audible wheezing. The ventilator was checked to ensure it was functioning properly, and 20 mg atracurium was administered. Despite hyperventilation, the end-tidal carbon dioxide increased to more than 60 mmHg, and the waveform indicated a rebreathing pattern with abrupt change in slope of the ascending limb: the soda-lime also turned purple. Over 45 min, the patient's temperature increased from 37.2 degrees Celsius to 38.6 degrees Celsius. Blood pressure was 140/60 mmHg and heart rate 95 beats/min.

At this point, MH was considered. Hyperventilation was continued with a high oxygen flow, isoflurane was discontinued, and a propofol infusion was started, cold intravenous solutions were administered. Arterial blood gas measurements revealed a pH of 7.1, Pa sub CO²97 mmHg, Pa^O277 mmHg, bicarbonate 24 mM/l, and potassium 7.5 mM/l. Dantrolene, 2.5 mg/kg, was given. Over the next 15 min, the body temperature reached 39 degrees Celsius, and at that time, arterial blood gases and electrolyte concentrations were pH 6.88, Pa^CO2146 mmHg, Pa^O298 mmHg, bicarbonate 21 mM/l; potassium was 8.6 mM/l, ionized calcium 1.34 mM/l, and lactate 11.1 mM/l. The same dose of dantrolene was administered again, and although the temperature immediately decreased to 37.6 degrees Celsius, serum potassium increased to 8.8 mM/l despite continuous infusion of glucose and insulin (Figure 1). The electrocardiogram demonstrated wide QRS complexes with peaked T waves that soon converted to asystole. A 30-min course of cardiopulmonary resuscitation was characterized by episodes of severe bradycardia, asystole, ventricular fibrillation, and ventricular tachycardia. The hyperkalemia was treated with continuous hyperventilation, calcium, insulin, and 50% glucose. Resuscitation was successful, and a third dose of dantrolene (2.5 mg/kg) was given. The arterial blood gases after resuscitation were pH 7.33, Pa^CO232 mmHg, Pa^O2218 mmHg, bicarbonate 17 mM/l; potassium was 8.3 mM/l and glucose was 435 mg/dL. After resuscitation, the kidney appeared ischemic, and the transplant attempt was aborted.

The patient was transferred to the intensive care unit, where he underwent hemodialysis twice over the next 12 h that decreased his serum potassium from 8.8 to 6.0 mM/l. The highest postevent CPK concentration was 11,000 IU/l (normal 30-220); no CPK-MB fraction was present. Lactic dehydrogenase was 557 IU/l (normal 100-220), and LD^s(muscle isoenzyme fraction) accounted for the entire increase. The serum myoglobin was 10,320 ng/ml (normal 0-85). The patient continued receiving dantrolene, 1 mg/kg three times daily, for the first 2 days postoperatively, and his serum potassium at this time was between 6.7 and 7.5 mM/l. The patient suffered a stroke with right-sided hemiparesis that was noted the following day and entirely resolved during the first postoperative week. During the first week, the patient's CPK was about 3,000 IU/l, and he was discharged from the hospital with a CPK of 2,227 IU/l. The myoglobin concentrations were 4,370 and 2,940 ng/ml on the 3rd and 4th postoperative days, respectively. Postoperative workup results for thyroid disease were negative. The patient and family members were counseled about MH and advised to undergo muscle biopsy testing to determine MH-susceptibility.

A 25-yr-old man was admitted in December 1982 for eye enucleation, required because of neurovascular glaucoma. He had a history of chronic renal failure induced by congenital obstructive uropathy and was undergoing hemodialysis three times a week. His preoperative medications included 40 mg furosemide and 80 mg propranolol three times per day, which was administered before surgery. His preoperative blood pressure was 130/70 mmHg, and his pulse was 85 beats/min. The electrocardiogram results were normal. Previously he had undergone several uneventful anesthesias for cystoscopy with thiopental and a mixture of oxygen and nitrous oxide as well as an uneventful anesthesia with cyclopropane and succinylcholine for tonsillectomy at age 9 yr. Postdialysis electrolyte concentrations (3 h before surgery) were sodium 140 mM/l, potassium 4.7 mM/l, chloride 101 mM/l, blood urea nitrogen 115 mg/dL, and creatinine 17 mg/dL.

Anesthesia was induced with 375 mg sodium thiopental and 120 mg succinylcholine and maintained with a mixture of oxygen and nitrogen, 66%/34%, and 0.5-1.0% halothane. Postinduction blood pressure was 140/100 mmHg, heart rate 100 beats/min, and rectal temperature 36.8 degrees Celsius. Approximately 80 min after anesthesia induction, supraventricular and ventricular ectopy were noticed. Serum potassium was 6.3 mM/l. Treatment for hyperkalemia was initiated with insulin, 50% dextrose, and hyperventilation, but the arrhythmias did not resolve. Despite treatment, the serum potassium concentration 30 min later was 7.1 mM/l, and arterial blood gas measurements (FI^O21.0) were pH 6.92, Pa^O261 mmHg, Pa^CO277 mmHg, and bicarbonate 12 mM/l; glucose was 351 mg/dL. The patient's temperature increased to 37.4 degrees Celsius, and MH was considered. Hyperventilation with high oxygen flow was continued, halothane was discontinued, and dantrolene was administered. Over the next 15 min, the body temperature increased to 38.8 degrees Celsius, and serum potassium increased to 9.4 mM/l, resulting in a wide-complex bradycardia followed by asystole. The patient died despite cardiopulmonary resuscitation, repeated dantrolene administrations, and cooling. The hyperkalemia was aggressively but unsuccessfully treated during the 60-min resuscitation effort with 50% glucose, 320 U of regular insulin, bicarbonates, and hyperventilation. The patient's serum potassium concentration was 8.8 mM/l, and the CPK was 2.200 IU/l.

The course of intraoperative events in our patients with ESRD was unusual for MH. The most prominent sign was acute hyperkalemia followed much later by hypercarbia and hyperpyrexia. The onset of hyperkalemia was surprising, and initially, poor potassium homeostasis caused by ESRD was thought to be the problem in both cases. Although patient 1 declined muscle biopsy, the entire clinical picture suggests a diagnosis of MH: Massive increases in carbon dioxide production, an acute increase in temperature from 35.8 degrees Celsius to 39 degrees Celsius, and a prompt decrease in temperature (signifying a decrease in metabolism) after dantrolene administration strongly support our diagnosis. [7,8]A valuable nonspecific diagnostic sign of MH, such as an unexplained increase in heart rate, was absent in both patients. In patient 1, marked tachycardia was absent, possibly because of diabetic autonomic neuropathy, premedication with clonidine and nifedipine, or both. In patient 2, tachycardia was absent, possibly because of pre-medication with propranolol. In the absence of diagnosis confirmed by muscle biopsy, an MH clinical grading scale may be used to define the probability that an event is MH. [8]According to this scale, the probability that MH occurred in patient 1 was "almost certain" (Table 1). [8]Both patients had previous uneventful exposures to general anesthesia with a triggering agent, but such uneventful exposure does not discredit our diagnoses, because approximately 21% of all human MH episodes have been preceded by one or more uneventful anesthesias. [9]Also, the late manifestation of MH during anesthesia is not atypical, especially after the use of depressant drugs such as barbiturates, benzodiazepines, and nondepolarizing muscle relaxants [10,11]; thiopental has been found to decrease sarcoplasmic reticulum Calcium²+ release, [12]but this mechanism was not studied in relation to MH.

The initial acute increase in serum potassium from 4.7 to 7.2 mM/l in patient 1 occurred in the absence of acidosis. We postulate that this hyperkalemia may be the result of one of the following two mechanisms. First, the acute hyperkalemia might have been caused by MH, with a large shift of extracellular potassium from skeletal muscles into extracellular space early during the course of anesthesia. The following findings support this mechanism: end-tidal carbon dioxide moderately increased (from 26 to 32 mmHg) despite a 75% increase in minute ventilation; and body temperature slowly increased immediately after induction of anesthesia, although this finding also can be attributed to active warming (Figure 1). Thus, it is possible that the MH episode started with an unusual presentation--with hyperkalemia as a primary event--progressed slowly and was recognized late, after sudden increases in end-tidal carbon dioxide and temperature during the 4th hour of anesthesia (Figure 1).

Alternatively, acute hyperkalemia as a primary event triggered MH episodes in MH-susceptible ESRD patients at the beginning of anesthesia, but the MH did not fully develop until 3 h later. Although it is not clear which mechanism caused acute hyperkalemia in patient 1 because hyperglycemia and acidosis were absent, it has been shown that acute hyperkalemia can develop during anesthesia in patients with diabetes who have ESRD. [5]Most often the hyperkalemia that occurs in diabetic patients is related to hyperglycemia, [13]but the concomitant ESRD would substantially increase the tendency toward hyperkalemia. [14,15]The role of potassium in increasing muscle metabolism is less well defined. Van der Kloot showed that, in normal muscle, potassium increases oxygen consumption (metabolism), but only at concentrations above 20 mM/l. [16]Moulds and Denborough showed in vitro that increased potassium stimulates skeletal muscle metabolism and produces the contracture at lower concentrations in MH-susceptible muscle than in normal human muscle. [17]Therefore, it is unlikely that hyperkalemia in our patients triggered an increase in metabolism that caused an MH-like syndrome in non-MH-susceptible individuals. Gronert et al. found that hyperkalemia greater than 6.0 mM/l precipitates MH in MH-susceptible swine, and he postulated that the pathogenetic mechanism may be a "hyperkalemic muscle cell depolarization." [6]Acute and severe hyperkalemia after anesthetic induction occurred in our patients in the absence of other obvious signs of hypermetabolism (Pa^CO2rise, acidosis, or hyperpyrexia), and this clinical picture supports the hypothesis that acute hyperkalemia developed in patients with ESRD who were MH-susceptible as a primary event and then triggered an MH episode.

The acute increase in serum potassium in the patient with diabetes (patient 1) was initially reversed with treatment (from 7.2 to 6.4 mM/l). The second potassium increase in this patient may be attributed to acute development of respiratory acidosis because a Pa^CO2increase of 10 mmHg can increase serum potassium by up to 0.3 mM/l. [18](in our patient, Pa^CO2increased from 45 to 146 mmHg, and serum potassium concentration from 6.4 to 8.6 mM/l). After dantrolene administration, the respiratory acidosis was easily corrected with hyperventilation, suggesting that hypermetabolism was halted; the hyperkalemia, however, resisted all therapeutic interventions.

The acidosis and hyperkalemia also resulted in cardiac arrest, but an interaction between dantrolene and verapamil must be considered as a contributing factor. [19,20]Concomitant administration of verapamil and dantrolene has been shown to induce hyperkalemia and cardiovascular collapse in animals [19]and humans. [20]Salzman postulated that verapamil alters the normal homeostatic mechanism for potassium regulation. [19]Simultaneous increase in serum potassium caused by dantrolene [21]can augment negative dromotropic and inotropic effects of verapamil on the heart. Hall et al. found that, 15 min after dantrolene administration, potassium production increased fourfold in the leg muscle of an MH-susceptible swine, which suggests an increased potassium release from the skeletal muscle. [22]The mechanism of dantrolene-induced hyperkalemia has never been elucidated, but because mannitol is added to dantrolene (3 g mannitol per 20 mg dantrolene), it is possible that dantrolene-induced hyperkalemia is a result of mannitol-induced hyperosmolarity; plasma potassium concentration may rise by as much as 0.4-0.8 mM/l for every 10 mOsm/kg elevation in effective plasma osmolality. [23].

In our second case, we cannot precisely document the temporal sequence of the early MH signs, such as an increase in end-tidal carbon dioxide, because technology for monitoring such signs was unavailable at the time. Nonetheless, with such severe hypercarbia, we would have expected to see spontaneous breathing efforts in an anesthetized patient who was not paralyzed; however, we did not. The first obvious symptom was dysrhythmia caused by hyperkalemia; hyperpyrexia followed. In both patients, the hyperkalemia was resistant to treatment (including postoperative hemodialysis in patient 1), which might be attributable to ESRD, hyperosmolar states induced by mannitol, glucose administration, and in patient 2, also to propranolol. [24].

We believe that these cases suggest an atypical presentation or triggering of MH by hyperkalemia. They also suggest that MH-susceptible patients may be at increased risk for an MH episode if they have a coexisting disease that alters potassium homeostasis. Finally, the cases point out a potential therapeutic dilemma: In the patient with ESRD who develops MH, hyperkalemic cardiovascular complications may be more prominent and difficult to treat because dantrolene can potentiate primary hyperkalemia.

The authors thank Cassandra Talerico. The Department of Scientific Publications, Cleveland Clinic Foundation, for suggestions and help in preparing this manuscript.