Fatal lactic acidosis and mimicking Guillain-Barré syndrome in an adolescent with human immunodeficiency virus infection

We report a case of antiretroviral therapy-related fatal lactic acidosis occurring in a vertically infected HIV-positive 17-year-old patient. While receiving antiretroviral therapy with stavudine, didanosine, tenofovir and amprenavir, the patient developed severe acidosis and rapid neuromuscular and respiratory failure mimicking Guillain-Barré syndrome. Highly active antiretroviral therapy (HAART) has increased the duration of survival in HIV-infected children, although short and long term toxicity is concerning. 1, 2 One of the most feared adverse events is severe lactic acidosis, induced by the mitochondrial toxicity of nucleoside analog reverse transcriptase inhibitors (NRTIs). 3 Chronic, compensated, asymptomatic hyperlactatemia is common in patients receiving HAART. 4 Life-threatening hyperlactacidemia is a known complication of antiretroviral therapy; although cases of severe decompensated lactic acidosis are uncommon (1.3 cases per 1000 person-years of NRTI exposure), this syndrome is associated with a high case-fatality rate. 5–9 Recently cases of stavudine (D4T)-associated severe acidosis with rapid development of neuromuscular weakness and respiratory failure mimicking Guillain-Barré syndrome (GBS) have been reported. 10 We report a case of a severely immunocompromised patient who developed fatal lactic acidosis with associated symptoms mimicking GBS during therapy with D4T, didanosine (DDI), tenofovir (TNF) and amprenavir. Case reports. A 17-year-old girl was followed for HIV infection since age 4 years in our Department from 1989. Her HIV infection was detected while she underwent screening for tonsillectomy. She had been vertically infected by her mother. In April 1989 her CD4+ lymphocyte was count of 615/mm3 (30%). In the following months severe HIV-related symptoms (Stage C3 of CDC classification of HIV infection) were observed: oropharyngeal and esophageal candidiasis, persistent fever; chronic diarrhea; and hypopotassemia. In November zidovudine (ZDV) monotherapy was initiated. In the following years she underwent monthly clinical and laboratory assessment and received various antiretroviral regimens. Since 1996 she received prophylaxis with pentamidine because of her allergy to sulfonamides. A complete reduction of the viral load and an increase of the CD4+ cell count were never achieved, although many antiretroviral regimens were used. In 2000 resistance testing revealed multiple resistance to all classes of antiretrovirals. In January 2002 she was clinically stable and no new opportunistic infections had occurred. In February 2002 she was treated with a four drug regimen with ZDV (300 mg twice a daily), TNF (300 mg daily), amprenavir (600 mg twice a daily) and ritonavir (RTV) (100 mg twice a daily). In August 2002, because of the occurrence of neutropenia, ZDV was withdrawn, and DDI (400 mg daily) and D4T (40 mg twice daily) were started. Although DDI and D4T had been used previously, they were used as part of a multidrug rescue therapy. Despite this therapy, the CD4+ count in October 2002 was 12/mm3 (2%), and the viral load was 980 000 copies/ml (AMPLICOR). The HAART regimen was well-tolerated, without signs of liver or pancreas toxicity. Except for the persistence of lopinavir-ritonavir (amprenavir)-related lipodystrophy, therapy was well-tolerated by the patient. In October 2002 an increase of alanine aminotransferase and and aspartate aminotransferase values was observed without other signs of hepatic toxicity; in November RTV was withdrawn because of hepatotoxicity. In December 2002 she was admitted for cough and fever (38°C) for 10 days with vomiting and headache. Four days before admission her physician diagnosed bilateral acute otitis media, and she was given ciprofloxacin 500 mg orally every 12 h. On admission the temperature was 37.5°C and she complained of malaise, otalgia, cough, dyspnea, neuromuscular weakness, numbness in the legs and feet and mild paresthesia of the extremity. On physical examination she was lipodystrophic; there was evidence of oral candidiasis and cervical adenopathy; heart and chest examination were negative. Blood pressure was normal, pulse was 96/min and temperature was 37°C. The abdomen was soft, nontender and nondistended with liver and spleen enlargement (ultrasonography and computed tomography abdominal scans did not show hepatic steatosis). Neurologic assessment showed slight pain, numbness, analgesia and decreased deep tendon reflexes in both legs. Venous blood gas value were: pH 7.237, pCO2 22.1 mm Hg, pO2 66.4 mm Hg, bicarbonate (HCO3−) 11.9 mm Hg and oxygen saturation (SO2) 91.2%. Laboratory values on hospital admission are shown in Table 1. Blood cultures for bacteria and chest roentgenogram for infiltrate were negative.TABLE 1: Laboratory values of the patient at the time of hospitalizationAntiretroviral therapy was stopped. On the second day, 12 h after admission, her physical examination was stable and venous blood gas values were: pH 7.32, pCO2 23 mm Hg, pO2 57.3 mm Hg and HCO3− 9.1 and SO2 91.2%. Blood lactic acid concentration was elevated (13.4 mM; normal value, 0.5 to 2.2). Her neurologic condition was worsening with moderate paresthesia, absent deep tendon reflexes and severe weakness. She was unable to stand. A lumbar puncture was performed to rule out a postinfectious Guillain-Barré syndrome. The cerebrospinal fluid protein, leukocyte and glucose values were normal. An electrocardiogram showed sinus tachycardia at a rate of 150/min without any other signs of cardiac involvement. Her respiratory and hemodynamic status deteriorated. She was unable to walk and complained of dyspnea. The patient was transferred to the intensive care unit to receive mechanically assisted ventilation. The next morning she further decompensated, with worsening oxygenation and refractory hypotension to vasopressors. Cardiopulmonary resuscitation was unsuccessful. The cause of death was believed to be associated with HAART-related lactic acidosis and to a rapidly ascending neuromuscular weakness and respiratory failure, mimicking GBS. Discussion. Although mild lactic acidosis has been described as a side effect of antiretroviral therapy regimens (mainly those containing stavudine), severe lactic acidosis during antiretroviral therapy is rare. 4, 11 Data from the medical literature regarding lactic acidosis in children refer mainly to children exposed perinatally to antiretroviral drugs; however, they did not seem to show increased susceptibility to mitochondrial dysfunction. 12–14 Few data are available on the efficacy and tolerability of antiretrovirals in vertically HIV-infected patients. 15, 16 In contrast to adult studies, large pediatric cohort studies are lacking. All antiretroviral drugs have been associated with toxicities in children, but in general they are relatively well-tolerated. 15 Mitochondrial toxicity has been identified as the common denominator for a group of side effects associated with the use of NRTIs. 17 Severe hyperlactacidemia is extremely rare, but milder and asymptomatic forms in adults with less elevated lactate concentrations are increasingly documented. Presumably homeostatic mechanisms are preserved enough to compensate, at least partially, for lactate excess in most patients. The primary cause of this net rise in lactate (increased production and/or reduced clearance) is not known. A study of exercise physiology has shown that skeletal muscle is probably not a major source of excess lactate production in NRTI-treated HIV-infected patients. 18 In contrast to the common form of mild stable hyperlactatemia, severe lactic acidosis appears to indicate an extreme, decompensated metabolic state in which homeostasis is lost. 19, 20 According to recent studies it seems that in most patients milder or asymptomatic hyperlactacidemia does not progress to more severe clinical forms. Therefore routine measurement of lactate values in NRTI recipients is not required. 21, 22 The clinical presentation of hyperlactacidemia in this patient was insidious. She had several risk factors: obesity; female sex; prolonged drug exposure (above all to NRTIs); and the use of a new antiretroviral drug (TNF) for which data are lacking in the pediatric population. Long term treatment with NRTIs appears to be the predominant risk factor for development of chronic hyperlactatemia. In February 2002 she was receiving treatment with lamivudine, DDI, lopinavir and RTV, and her blood lactate value was normal. Although the lactate value was normal, increased concentrations were observed on screening in 37% of children (without association with any particular antiretroviral). 23 In the cases published to date, neither specific nor predictive signs of outcome were described for fatal cases. Also no effective therapy has been described. In terms of the causative agent, cases of stavudine-associated lactic acidosis syndrome with neuromuscular failure have been reported. 24, 25 In those studies immediate drug discontinuation and appropriate supportive care (e.g. ventilation) at the onset of hyperlactatemia and motor weakness were not effective. 10 We warn of the possible occurrence of severe lactic acidosis in vertically infected children, with clinical manifestations of rapid onset of unexplained general fatigue, weight loss, abdominal pain and digestive signs or dyspnea. In our case muscular weakness was particularly severe, resembling Guillain-Barré syndrome, as has been described in few adults with severe lactic acidosis induced by antiretroviral therapy regimens containing NRTI. 10 The patient had been also treated for 13 years and, at least partially, this toxicity could be explained by the long period of treatment. Besides an increased vulnerability to metabolic role effects of antiretroviral therapy in children compared with adults has been observed, and this vulnerability could be caused by the children’s likely greater cumulative exposure. 26 Physicians should consider monitoring the lactate concentrations of patients treated with antiretrovirals if symptoms such as fatigue, nausea, vomiting, dyspnea and peripheral neuropathy develop. Acknowledgment. We thank Dr. Giorgio Gatti for his useful comments and revision of the paper.