Acute Renal Failure After Radiofrequency Liver Ablation of Metastatic Carcinoid Tumor

Laparoscopic radiofrequency ablation is used to treat unresectable metastatic liver tumors. It is minimally invasive and associated with few side effects. Reported complications from this procedure include bleeding (1,2), hyperbilirubinemia (1), pleural effusion (2), cholecystitis (2), intrahepatic abscess (3), and minor skin burn (4). We describe a case of radiofrequency ablation complicated by hemolysis, rhabdomyolysis, and transient acute renal failure. Case Report A 57-yr-old male with metastatic carcinoid tumor of unknown primary origin was scheduled to undergo laparoscopic radiofrequency ablation of liver metastases. Abdominal computerized tomography visualized four large hepatic metastases (up to 6 cm in diameter) and several smaller ones. The patient’s symptoms included severe flushing, diarrhea, shortness of breath, hypertension, right upper quadrant pain, and anxiety, all of which were attributed to carcinoid disease. He underwent a trial of octreotide in 1997 and chemoembolization in 1998, both without significant improvement. Over the previous several months his symptoms had worsened, and he was scheduled for surgery. His surgical history included cryoablation of the hepatic metastases in 1997 and open reduction and internal fixation of his left hip for degenerative arthritis in 1998. Neither surgery was associated with anesthetic complications or complications related to his carcinoid disease. His current medications consisted of the infrequent use of an albuterol inhaler for shortness of breath. He was on no other medications and had no allergies. His physical examination was notable only for a purple skin complexion, particularly over his face and neck. His preoperative vital signs were as follows: blood pressure, 129/75 mm Hg; heart rate, 66 beats per minute; respiratory rate, 16 breaths per minute; weight, 68 kg; and pulse oximetric oxygen saturation, 98%. An echocardiogram (ECG) several weeks before surgery revealed no evidence of right heart disease; the ejection fraction, chamber sizes, and valves were normal, and the pulmonary artery systolic pressure was 29 mm Hg. The patient’s ECG revealed sinus rhythm without abnormality, and laboratory studies revealed a hematocrit of 42%, serum potassium 4.6 mmol/L, blood urea nitrogen 21 mg/dL, creatinine 0.8 mg/dL, prothrombin time of 10.5 s, and partial thromboplastin time of 30.6 s. Preoperatively, the patient was premedicated with IV midazolam 2 mg and a right radial arterial line was placed. After placement of standard ASA monitors, the patient was preoxygenated. Anesthesia was induced with IV propofol 200 mg and fentanyl 200 μg, and muscle relaxation was achieved with rocuronium 50 mg. After placement of a 7.5-mm internal diameter endotracheal tube, anesthesia was maintained with desflurane, N2O/O2, and fentanyl. Octreotide was administered as a 50 μg IV bolus preoperatively followed by 25 μg every 30 min during surgery. The patient underwent radiofrequency ablation of six hepatic lesions. Total ablation time was 238 min with an average probe temperature of 105°C. Between each radiofrequency ablation cycle, there was a 30-s cool down time period with average hepatic tissue temperature between 65–70°C. During the 7-h procedure, the patient’s mean arterial blood pressure ranged between 65 mm Hg and 120 mm Hg and his heart rate varied between 49 bpm and 80 bpm. He received 3700 mL of crystalloid. Urine output was 50 mL during the first hour and 15 mL during the remaining 6 h. The cause of the decrease in urine output was investigated and discussed with the surgeon. His esophageal temperature was 35.4°C 1 h postinduction but increased to 38.8°C at the end of the procedure despite 1 h of forced ambient air cooling. Intraoperatively, 6 h after induction, laboratory values from arterial blood were as follows: pH, 7.35; Pco2, 41; Po2, 173 (Fio2, 0.4); HCO-3, 22.8; sodium, 135 mmol/L; potassium, 6.8 mmol/L. There was no change in the morphology of the electrocardiogram or in the width of the QRS complex. The surgeon was notified of the increased potassium, and treatment was initiated with furosemide 20 mg IV and CaCl2 500 mg IV. At the conclusion of surgery, the patient’s trachea was extubated in the operating room and he was transferred to the recovery room. On arrival in the recovery room, he was alert and oriented. In the recovery room, 20 mL of “black” urine was noted. On urinalysis, specific gravity was 1.015 with large amounts of hemoglobin and protein >300 mg/dL. Ketones were 15 mg/dL, and both bile and glucose were negative. The urine was not examined microscopically. Serum creatine phosphokinase (CPK) was 1750 u/L, potassium 6.4 mmol/L, blood urea nitrogen 30 mg/dL, and creatinine 1.3 mg/dL. The patient’s CPK peaked at 4500 U/L on postoperative day (POD) 1 and his hematocrit decreased to a low of 28.2% on POD 7. His serum creatinine peaked at 9.9 mg/dL on POD 6, decreasing to 1.3 mg/dL on POD 21. On POD 6, a right-sided abdominal hematoma measuring 12” × 5” was noted. The patient was discharged to home on POD 14 in good condition. Discussion Acute renal failure has been documented as a complication of cryoablation (5); however, to our knowledge, the development of acute renal failure associated with radiofrequency liver ablation, has not been previously reported. Our patient developed intraoperative oliguria and then went on to develop acute renal failure. The most likely cause of renal failure was thermal hemolysis, although rhabdomyolysis also may have contributed. Support for hemolysis as the etiology of the renal failure includes the decrease in hematocrit from a preoperative value of 42% to a postoperative value of 35%, the increase in serum potassium from 4.6 mmol/L (preoperative) to 6.8 mmol/L (intraoperative), and the postoperative positive hemoglobin urine dipstick. We believe the cause of the thermal hemolysis to be prolonged radiofrequency liver tumor ablation. The temperature of the tissue adjacent to the radiofrequency probe is reported to reach temperatures between 50°C and 100°C (1) and blood hemolyzes if warmed to ≥50°C (6). We believe the size and number of the intrahepatic metastases undergoing ablation in our patient may have resulted in greater intrahepatic warming than that experienced in cases of shorter duration and/or in cases with fewer intrahepatic metastases. Previous reports of radiofrequency liver tumor ablation describe using 50–100 W of power for approximately 10 minutes per tumor (1,3). These levels of power deposition produce probe temperatures up to 90°C and liver parenchyma temperatures of approximately 60°C (3). Direct comparison of hepatic tissue heating in the case presented here and previous published reports (1–3) is complicated by these reports not reporting total ablation time, average radiofrequency probe temperature during the ablation period, or average hepatic tissue temperature during the ablation period. Although hemolysis may explain the decrease in hematocrit and increase in serum potassium, the increase in serum CPK requires further explanation. Red blood cells do not contain CPK (7), which is found only in brain and muscle tissue. Most surgery results in modest increases in serum CPK because of muscle trauma, but these increases are generally in the range of 150–800 U/L (8). In our case, the CPK peaked at 4500 U/L. We believe the cause of this level of CPK to be rhabdomyolysis resulting from local trauma to the abdominal wall as evidenced by a postoperative 12” abdominal wall hematoma. Unfortunately, neither serum nor urine myoglobin were measured. Nevertheless, myoglobin causes “black” urine and the urine will test positive for hemoglobin when analyzed using a urine dipstick (9). Theoretically, the rhabdomyolysis may have contributed to the development of acute renal failure. When acute renal failure develops after surgery not involving radiofrequency ablation, the most common causes are renal hypoperfusion from prolonged hypotension (with or without cardiac failure), hypovolemia, or renal toxic drugs, such as aminoglycoside antibiotics. These did not play a role in our case. Clinically, there was no evidence for hypovolemia, and no blood loss was seen at any time during the laparoscopic surgery. Neither hypotension nor tachycardia occurred intraoperatively. No changes in the pulse pressure were observed in the appearance of the radial arterial line tracing. Preoperative cardiac evaluation did not indicate any potential for cardiac compromise and intraoperative and postoperative cardiac function remained normal. Without intraoperative hemodynamic or ECG changes, invasive monitoring of central venous pressure or pulmonary artery pressure was not done. None of the anesthetics used are nephrotoxic. Cephalosporin nephrotoxicity is rare. When it does occur, acute tubular necrosis follows the use of extremely large doses of cephalothin or after standard large-dose therapy in the elderly. Cephalothin can, however, enhance the nephrotoxicity of aminoglycosides (10,11). Cefazolin administered intraoperatively in a 1-g IV dose, as in our case, does not cause acute renal failure. Rapid lysis of large amounts of tumor, usually with chemotherapy, can result in tumor lysis syndrome producing renal failure. When renal failure occurs, it is the result of uric acid and/or calcium phosphate crystals precipitating in the renal tubules. Tumor lysis syndrome is associated with hyperkalemia, hyperuricemia, hyperphosphatemia, hypocalcemia, and increased lactic dehydrogenase (12,13). In our case, laboratory values revealed hyperkalemia, and an increased lactic dehydrogenase, but not the other chemical abnormalities associated with tumor lysis syndrome. Radiofrequency ablation has not been reported as a cause of tumor lysis, and we believe the increase in lactic dehydrogenase and potassium was because of the release of red cell contents. Laparoscopic insufflation results in intraoperative oliguria (14,15), but has not been reported to cause renal failure. The two postulated mechanisms for intraoperative oliguria are decreased renal perfusion and increased renin activity, both because of the increased intraabdominal pressure caused by CO2 insufflation with direct compression of the kidneys and surrounding vessels. In these studies, average abdominal insufflation time was less than two hours. The effect of insufflation time of longer than three hours on renal function has not been studied. In conclusion, we believe the most likely cause of transient acute renal failure in our case was hemolysis from prolonged radiofrequency ablation. The clinical implications from our case relate to surgery where prolonged radiofrequency liver ablation is anticipated. On the basis of our experience, we now recommend using laboratory monitoring during extensive or prolonged radiofrequency liver ablation procedures so that hemolysis may be detected early. Laboratory tests that should be considered include hematocrit, serum potassium, urine hemoglobin, and serum CPK. In addition, if hemolysis is suspected, or urine output decreases significantly, measures to increase urine output should be undertaken and consideration given to alkalinizing the urine if rhabdomyolysis is demonstrated. Invasive central venous pressure or pulmonary artery monitoring is indicated if the intravascular volume status is uncertain. Whenever possible, the number of ablations and the duration of each ablation sequence should be reduced to the minimum necessary. For extensive or multiple tumor sites, or when prolonged radiofrequency ablation is anticipated, several separate operations should be considered to prevent acute renal failure.