One of the recommended treatments for latent tuberculosis infection (LTBI) is a 9-month regimen of isoniazid (INH); a 2-month regimen of rifampin (RIF) and pyrazinamide (PZA) is an alternative in some instances. In September 2000, a man in New York died of hepatitis after 5 weeks of RIF-PZA, and in December, a woman in Georgia was admitted to a hospital because of hepatitis after 7 weeks of this regimen. This report summarizes the findings of the investigations of these incidents, which underscore the need for clinical monitoring for adverse effects in all patients receiving treatment for LTBI.
A 53-year-old incarcerated man received 600 mg (6.7 mg/Kg) RIF and 1750 mg (19 mg/Kg) PZA daily after screening revealed a tuberculin skin test (TST) with 20 mm induration and no radiologic or clinical findings of active tuberculosis (TB). His risk factors for TB included previous work as a medical orderly, homelessness, and multiple incarcerations. He had a history of hypertensive heart disease and alcoholism without evidence of chronic liver disease. He was not known to inject drugs.
RIF-PZA was standard treatment for LTBI at the jail. Baseline and 1-month serum aminotransferase and bilirubin levels were measured routinely. The patient's baseline aminotransferase levels were slightly higher than the upper-normal limits. He was instructed to stop taking RIF-PZA if he developed symptoms suggestive of hepatitis. He also received 325 mg enteric-coated aspirin daily, 90 mg extended-release nifedipine, and 50 mg hydrochlorothiazide. Nurses supervised the administration of all medication to assure compliance.
Blood specimens tested on day 33 of treatment revealed alanine aminotransferase (ALT) 1734 U/L (normal range: 0-41 U/L), aspartate aminotransferase (AST) 1449 U/L (normal range: 0-38 U/L), and total bilirubin 4.2 mg/dL (normal range: 0-1.0 mg/dL). Blood cell counts showed leukocytosis. On day 35, RIF-PZA was discontinued when the test results were received. On the same day, a correctional officer urged the patient to visit the infirmary because of poor appetite and lassitude that had developed over several days; he declined. Five days after the cessation of RIF-PZA, the patient was evaluated in the infirmary for jaundice and altered mental status and was admitted to a hospital. Serum total bilirubin peaked at 17.8 mg/dL and blood ammonia at 378 µmol/L (normal range: 17-47 µmol/L). He died 3 days after admission.
On postmortem histology, the liver had bridging necrosis, lymphocytic infiltration, focal cholestasis, increased fibrosis, and micronodular cirrhosis. Results were negative for serum anti-A IgM, antibody to hepatitis B core antigen (anti-HBc), antibody to hepatitis B surface antigen (anti-HBs), and antibody to hepatitis C virus (anti-HCV). Antinuclear antibody (ANA) was undetectable. Hepatitis B and C were undetectable by polymerase chain reaction assays. The reported cause of death was liver necrosis and failure as a result of hepatitis following LTBI treatment.
A 59-year-old woman received 600 mg (7.2 mg/Kg) RIF and 2000 mg (24 mg/Kg) PZA daily after testing revealed a TST with 27 mm induration and no findings for active TB. She chose this regimen because of suspected exposure to drug-resistant TB and concern about liver injury from INH. In addition to RIF-PZA, she received beclomethasone dipropionate nasal spray, budesonide inhalation powder, and albuterol inhalation aerosol for nasal allergies and asthma. She had no history of liver disease, rarely drank alcohol, and did not inject drugs. She was vaccinated against hepatitis A but not B. She had a history of anaphylactic reactions to penicillin and an estrogen sulfates blend. Baseline ALT and AST, bilirubin levels, and blood cell counts were normal. She was instructed to contact her health-care provider about adverse effects during treatment. On day 2 of treatment, she reported queasiness. On day 17, her blood tests were repeated: serum aminotransferase and bilirubin levels were normal, and her eosinophil count, which had been 157 cells/µL, was 510 cells/µL (normal range: 50-550 cells/µL).
She subsequently experienced malaise, anorexia, and feverishness, and she occasionally took one bismuth subsalicylate chewable tablet. On the 49th and last day of treatment, she returned to her health-care provider and was admitted to a hospital because of jaundice and altered mental status. AST was 986 U/L (normal range: 7-40 U/L), ALT 1735 U/L (normal range: 17-63 U/L), and total bilirubin 11.4 mg/dL (normal range: 0.1-1.1 mg/dL). The bilirubin peaked at 27.5 mg/dL after 14 days. Peak eosinophil count was 2580 cells/µL. No ova or protozoa were detected by stool examinations. Serum ANA was 1:640 (speckled pattern). Antibody (not IgM) to hepatitis A virus was detected. Test results were negative for hepatitis B surface antigen (HBsAg), anti-HBs, and anti-HCV. After receiving 40 mg prednisone daily, the symptoms and laboratory abnormalities slowly abated, and she was released after 25 days in the hospital.
M DeMartino, MD, Nassau County Office of the Medical Examiner; J Maniscalco, A Greenberg, MD, Nassau County Dept of Health, Mineola; J Grabau, PhD, M Oxtoby, MD, E Foster, MS, Bur of Tuberculosis Control, P Smith, MD, State Epidemiologist, New York State Dept of Health. P Kozarsky, MD, C Pox, MD, Emory Univ School of Medicine, Atlanta, Georgia. National Institute of Diabetes and Diseases of the Digestive System and Kidneys, National Institutes of Health, Bethesda, Maryland. Occupational Health Clinic, Office of Health and Safety, Office of the Director, Hepatitis Br, Div of Viral and Rickettsial Diseases, National Center for Infectious Diseases; Div of Tuberculosis Elimination, National Center for HIV, STD, and TB Prevention, CDC.
Case 1 is the first report to CDC of fatal hepatitis associated with the RIF-PZA regimen for LTBI, although sporadic cases of liver injury have been attributed to PZA used in treatment regimens for TB disease.1 Both cases illustrate that the usually well-tolerated regimens for LTBI occasionally can result in severe adverse effects and that clinical monitoring is crucial during treatment. In these cases, biochemical monitoring did not help to avoid severe liver injury and does not substitute for clinical monitoring.2 Idiosyncratic liver injury can be caused by hypersensitivity, as suspected for case 2, or by toxic drug metabolites. Other cases have implicated various medicines and alcohol as potential co-factors for INH liver injury.3,4 A similar association has not been assessed for RIF and PZA because of small case numbers.
Patients with LTBI and risk factors for active TB should be offered treatment.1,5 Health-care providers should instruct and frequently remind patients about the initial symptoms of hepatitis (e.g., fatigue, nausea, abdominal pain, and anorexia) and the importance of stopping medication if symptoms develop.2 In this report, both patients continued taking their medicines while symptoms were developing, a phenomenon also reported for INH-associated hepatitis.4
CDC's Division of Tuberculosis Elimination is interested in receiving reports of severe hepatitis in patients being treated for LTBI. To report possible cases, telephone (404) 639-8125.
Fatal and Severe Hepatitis Associated With Rifampin and Pyrazinamide for the Treatment of Latent Tuberculosis Infection—New York and Georgia, 2000. JAMA. 2001;285(20):2572-2573. doi:10.1001/jama.285.20.2572