Acute liver failure due to antitubercular therapy: Strategy for antitubercular treatment before and after liver transplantation


  • Philippe Ichai,

    Corresponding author
    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    3. Unité 785, Institut National de la Santé et de la Recherche Médicale, Villejuif, France
    • Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, 14 Avenue P. V. Couturier, BP 200, Villejuif, France 94804
    Search for more papers by this author
    • Telephone: 0033 (1) 45596916; FAX: 0033 (1) 45593857

  • Faouzi Saliba,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    3. Unité 785, Institut National de la Santé et de la Recherche Médicale, Villejuif, France
    Search for more papers by this author
  • Fadi Antoun,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    Search for more papers by this author
  • Daniel Azoulay,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    Search for more papers by this author
  • Mylène Sebagh,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    3. Unité 785, Institut National de la Santé et de la Recherche Médicale, Villejuif, France
    Search for more papers by this author
  • Teresa Maria Antonini,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    Search for more papers by this author
  • Lélia Escaut,

    1. Département d'Infectiologie, AP-HP Hôpital Paul Brousse, Villejuif, France
    Search for more papers by this author
  • Delvart Valérie,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    Search for more papers by this author
  • Denis Castaing,

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    3. Unité 785, Institut National de la Santé et de la Recherche Médicale, Villejuif, France
    Search for more papers by this author
  • Didier Samuel

    1. Centre Hépato-BiliaireAP-HP Hôpital Paul Brousse, Villejuif, France
    2. Unités Mixtes de Recherche en Santé 785, Université Paris-Sud, Paris, France
    3. Unité 785, Institut National de la Santé et de la Recherche Médicale, Villejuif, France
    Search for more papers by this author


The standard antitubercular treatment (ATT), which consists of isoniazid (INH), rifampicin (RIF), ethambutol, and pyrazinamide (PZA), is the best available treatment for tuberculosis (TB). However, the hepatotoxicity of INH and PZA can be severe, and even after drug withdrawal, patients may require liver transplantation (LT). In these cases, the strategy for the treatment of TB is poorly defined. Between 1986 and 2008, 14 patients presented at our department with severe hepatitis secondary to INH and PZA treatment. Four of these patients were immunosuppressed: 2 after renal transplantation and 2 because of human immunodeficiency virus infection. In seven of the 14 patients an alternative ATT was begun on admission, which was well tolerated. Hepatitis improved spontaneously in 5 patients, and alternative ATT was continued for 9.3 ± 4.2 months; 1 patient deteriorated and underwent LT, and 1 patient died. ATT was stopped definitively in 2 patients. Six patients required urgent LT, and alternative ATT was started after transplantation and was successful. Five patients receiving RIF had an episode of acute rejection. In conclusion, hepatitis secondary to ATT can be successfully treated with alternative anti-TB regimens. The use of RIF in LT patients may lead to acute rejection. RIF should therefore be avoided in these patients. Liver Transpl 16:1136–1146, 2010. © 2010 AASLD.

The number of cases of tuberculosis (TB) has increased considerably over the past decade; this is especially true for cases involving resistant strains. This has led to the use of newer therapeutic agents in multiple-drug combinations. The treatment regimen recommended for patients with TB consists of an initial 2-month course of isoniazid (INH), rifampicin (RIF), pyrazinamide (PZA), and ethambutol (ETH), which is followed by maintenance therapy for either 4 or 7 months.1 The efficacy of this treatment is due to the combination of bactericidal drugs (INH, RIF, and PZA) and a bacteriostatic drug (ETH). Liver function abnormalities are the most commonly reported adverse effects of this standard regimen. Hepatotoxicity, mainly related to INH and PZA, has been widely reported.2-4 This hepatotoxicity ranges from liver function test abnormalities to acute liver failure (ALF). In most cases, ALF associated with INH and RIF occurs within 10 days of the initiation of the anti-TB regimen, whereas liver failure due to PZA develops after a longer period of treatment.2, 5 In patients receiving both INH and PZA, hepatitis may be due to INH or PZA. Early discontinuation of these drugs leads in most cases to a favorable outcome. Indeed, only a few patients progress to ALF and death in the absence of liver transplantation (LT). To date, few cases of LT for antitubercular treatment (ATT)–induced liver failure have been reported in the literature.

The management of drug-induced liver disease and active TB in these patients is difficult, and the choice of ATT early after LT remains controversial.6, 7 In addition, the hepatotoxicity of anti-TB drugs is enhanced in transplant patients, mainly because of the function of the graft and interactions with immunosuppressive drugs.7, 8 Guidelines about the treatment of these patients exist in the literature, and they include information about the time to initiate ATT after ALF, the decision to reintroduce or not reintroduce standard ATT (with potentially hepatotoxic drugs), and the effectiveness of alternative nonhepatotoxic anti-TB drugs such as ofloxacin (OFX), ciprofloxacin (CFX), moxifloxacin (MOX), and amikacin (AMK).9-17 These drugs have been shown to be effective and safe in patients with strains resistant to multiple drugs, but their use in patients with ALF before or after LT remains limited.

The present study reports consecutive patients with ALF and TB admitted to our unit during a 22-year period for the purpose of assessing ATT and observing the impact of this treatment on the course of liver disease with a particular emphasis on the safety and efficacy of alternative treatment regimens.


ADP, adenopathy; AFB, acid-fast bacillus; ALF, acute liver failure; ALS, anti-lymphocyte serum; ALT, alanine aminotransferase; AMK, amikacin; AMOX, amoxicillin; ARDS, acute respiratory distress syndrome; ATT, antitubercular treatment; AZA, azathioprine; BAS, basiliximab; BTC, belatacept; CFX, ciprofloxacin; CSA, cyclosporine A; CTM, clarithromycin; DDLT, deceased donor liver transplantation; DRESS, drug rash with eosinophilia and systemic symptoms; ETH, ethambutol; F, female; FQ, fluoroquinolone; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HE, hepatic encephalopathy; HIV, human immunodeficiency virus; IMS, immunosuppressive treatment; INH, isoniazid; INR, international normalized ratio; LDLT, living donor liver transplantation; LFX, levofloxacin; LT, liver transplantation; M, male; MMF, mycophenolate mofetil; MOF, multiorgan failure; MOX, moxifloxacin; NA, not applicable; ND, not done; OFX, ofloxacin; P, prednisolone; PT, prothrombin; PZA, pyrazinamide; RFB, rifabutin; RFP, rifapentine; RIF, rifampicin; SMC, streptomycin; TB, tuberculosis; TCR, tacrolimus.


Patient Recruitment

Between 1986 and 2008, 566 patients with ALF were admitted to our liver intensive care unit. Sixteen (11 women and 5 men) presented with ALF associated with TB. Fourteen of these 16 patients (10 women and 4 men; mean age = 37 ± 13 years) had an established diagnosis of anti-TB drug–induced ALF, 1 had miliary TB presenting as liver failure, and 1 had carbamazepine-induced ALF as part of DRESS (drug rash with eosinophilia and systemic symptoms) syndrome combined with TB. The latter 2 patients were excluded from the study, and only ALF due to ATT was considered.

The initial ATT consisted of RIF, INH, PZA, and ETH in 4 patients, INH, PZA, and RIF in 6 patients, INH in 2 patients, RIF, INH, PZA, and streptomycin (SMC) in 1 patient, and INH, RIF, and ETH in 1 patient. An antibiogram was available for 8 patients; in 7 patients, Mycobacterium tuberculosis was sensitive to INH, RIF, PZA, ETH, fluoroquinolone (FQ), SMC, and AMK, and in 1 patient, M. tuberculosis was resistant to SMC and INH but was sensitive to ETH, PZA, ETH, and FQ. In another patient, only the gene sequence rpoB allowed the acid-fast bacilli (AFBs) to be identified as tubercular bacilli with the wild gene, so this strain was probably sensitive to RIF (Table 1). Four patients were immunosuppressed: 2 patients 7 months and 6 years after renal transplantation (patients 2 and 8, respectively) and 2 patients with human immunodeficiency virus (HIV) infection (patients 6 and 9). One of the latter patients also had chronic hepatitis C (patient 6; Table 1).

Table 1. Anti-TB Therapy in the Fourteen Patients Included in the Study Cohort
PatientSexAge (Years)EthnicitySite of TBAnti-TB TherapyInterval Between the Introduction of Anti-TB Therapy and the Onset of Jaundice (Days or Months)
  • *

    TB was suspected before the presence of abnormalities in liver function tests on the basis of histology; the patient was subsequently found to be suffering from sarcoidosis and not TB.

  • RIF was replaced by MOX 1 month later because of intolerance.

  • The patient had recent contact with a person with TB.

  • §

    PZA was withdrawn 2 months later but was reintroduced because of resistance to INH (M. tuberculosis resistant to INH and SMC).

  • INH was withdrawn after 3 months of treatment because of the onset of mild cytolysis.

  • ALF occurred 50 days after PZA therapy.

1F17BlackThoracic ADPINH/RIF/PZA7 days
2F32Black*INH/RIF/PZA/ETH9 days
3F40BlackPulmonaryINH/RIF/PZA50 days
4F64BlackMeningitisINH/RIF/PZA/ETH8.5 months
5M41BlackPulmonary, thoracic ADPINH/RFB/PZA/ETH36 days
6M45CaucasianPulmonaryINH/RIF/PZA35 days
7F35BlackPulmonaryINH/RIF/PZA2 months
8F48BlackNo lesionsINH7 months
9M47BlackAbdominal ADP, pulmonaryINH/RIF/ETH§9 months
10M20CaucasianPulmonaryINH/RIF/PZA/SMC1 month
11F25BlackPulmonaryINH§/RIF/PZA5 months
12F20CaucasianPulmonaryINH/RIF/PZA/ETH1 month
13F40CaucasianPulmonaryINH/RIF/PZA1 month
14F49CaucasianPulmonaryINH and then PZA50 days

Hepatitis was caused by INH treatment in 4 patients and by PZA in 10. The mean time from the initiation of ATT to the development of jaundice was 3.5 ± 3.1 months (range = 0.3-9.2 months). The mean delay from the end of treatment to the development of severe hepatitis was 16.5 ± 29.2 days (range = 0-97 days). On admission, the mean liver function test values were as follows: the bilirubin level was 270 ± 179 μmol/L (range = 29-676 μmol/L), the prothrombin (PT) level was 25.8% ± 16% (range = 8%-63%), the factor V level was 40% ± 31% (range = 6%-103%), and the alanine aminotransferase (ALT) level was 1075 ± 485 IU/L (range = 214-2020 IU/L). The mean creatinine level was 153 ± 140 μmol/L (range = 63-523 μmol/L). Three patients (patients 10, 12, and 13) fulfilled Clichy's criteria for LT. The main characteristics of the patients are shown in Table 2.

Table 2. Clinical and Biochemical Characteristics of the Fourteen Patients
PatientCause of ALFPT/Factor V (%)ALT (IU/L)Total Bilirubin (μmol/L)Creatinine (μmol/L)HE GradeNew ATT After Admission (Days or Months)Delay to New Anti-TB Therapy (Days)Improvement in ALF or LTAnti-TB Drugs Post-LT (Months)IMSRejectionFollow-Up (Months or Years)
  • *

    AMK for 1 week.

  • The plasma concentration of the immunosuppressants was not available.

  • ATT was not reintroduced after the onset of ALF. The total duration of treatment was 7.1 months.

  • §

    AMK for 16 days.

  • INH was gradually introduced 5 days after LT.

  • INH was initiated on day 10.

  • #

    The plasma concentration of CSA was 315 ng/mL (the target therapeutic threshold was 200-400 ng/mL).

  • **

    Half of the RIF dose every 2 days.

  • ††

    The plasma concentration of CSA was 135 ng/mL (the target therapeutic threshold was 350-600 ng/mL).

  • ‡‡

    The plasma concentration of TCR was 10.9 ng/mL (the target therapeutic threshold was 8-20 ng/mL).

1INH23/10345129630RIF/ETH/LFX/AMK* (1 month), RIF/ETH/ MOX/INH (5.9 months), and RIF/INH (2.3 months)1ImprovementAlive (3.8 years)
2INH28/258921342970RIF/ETH/MOX (4.5 months)3ImprovementBTC/PYes, treatedAlive (9 months)
3PZA42/54390159840CFX/ETH/SMC (5.2 months)21ImprovementAlive (4.6 years)
4PZA32/402020180920INH/ETH (9 months)64ImprovementAlive (2.4 years)
5PZA43/8413503532500RFB/ETH/MOX (2 months) and ETH/MOX (9.2 months)1ImprovementAlive (1.3 years)
6PZA21/611146212 0ETH/AMK§/CFX (1.1 months)12No improvement/ no LTDied (2 months): MOF
7PZA34/641114192460RIF/ETH/MOX/AMK (8 days)1LTINH/ETH/MOX/ AMK (1.9) and INH/ETH/MOX (4)TCR/MMF/ BASNoAlive (10 months)
8INH30/19214491480Definitively withdrawnNAImprovementAlive (8 months)
9INH11/151228400762Not treated before LTNALTINH/RIF/ETH/MOX (1.3)TCR/PYes, not treatedDied (1 month): intestinal occlusion/ disseminated TB
10PZA16/172000544ND4Not treated before LTNALTRIF/SMC/ETH (2.8) and RIF/ETH (6.3)CSA/P/AZAYes, treated#Alive (19.2 years)
11PZA8/71420320782Not treated before LTNALTRIF/OFX/ETH (3.8) and RIF/OFX (6.5)CSA/P/AZAYes, treated††Alive (12.1 years)
12PZA8/101177433883Not treated before LTNALTRIF**/ETH/OFX (0.4), RIF/ETH/OFX (0.8), and RIF/ETH (1.2)TCR/PYes, treated‡‡Died (2 months): hemorrhagic shock/sepsis
13PZA8/68686765234Not treated before LTNALTNo treatmentALS/P/AZADied (1 month): sepsis shock/ ARDS
14PZA17/31790380890Definitively withdrawnNAImprovementAlive (9.7 years)

In 1 patient, AFB were isolated on admission from a sputum culture, which was later confirmed to be positive for M. tuberculosis (patient 9).

Patient Management

On admission, all standard ATTs and other potentially hepatotoxic drugs were stopped. Antiretroviral HIV therapy was discontinued in the 2 HIV-positive patients. Sputum, bronchoalveolar lavage, urine, and ascites were cultured and patients were screened for hepatitis A, B, and C viruses and HIV. According to the clinical condition of the patient, the extent of TB, and the laboratory findings, a new ATT using a combination of 3 or 4 anti-TB drugs was initiated. The histological status of the liver was assessed on the basis of transjugular liver biopsy samples for nontransplant patients (n = 7) and/or native liver specimens (n = 6). All patients with ALF were managed according to a standardized protocol.18, 19 ALF was defined as a PT level <50% with or without hepatic encephalopathy (HE).20 Encephalopathy was graded from 1 (least severe) to 4 (most severe).21 The criterion for LT was the presence of grade 3-4 HE associated with either a factor V level <20% of normal in patients <30 years of age or a factor V level <30% of normal in patients >30 years of age.18, 22, 23 On admission, extracorporeal liver support involving a molecular adsorbent recycling system24, 25 or a bioartificial liver was initiated in 3 patients.25-27

After LT, all patients received immunosuppressive therapy according to the standard practice of our center. In the HIV-positive patient who underwent LT, antiretroviral HIV therapy was reintroduced 2 weeks after transplantation. Plasma concentrations of tacrolimus (TCR) and cyclosporine A (CSA) were monitored daily. Postoperative follow-up included routine clinical surveillance, liver function tests, and cultures from all sites. The efficacy and safety of ATT were monitored by physical examination, chest X-rays, and liver function tests.

Statistical Analysis

Statistical analysis was carried out with Fischer's test and the Student t test.


Clinical Outcomes

Seven of the 14 patients recovered spontaneously with a return to normal hepatic function (patients 1-5, 8, and 14). On admission, the mean liver function tests for these patients were better than those for patients who underwent transplantation or died (total bilirubin level = 183 ± 137 versus 372 ± 177 μmol/L, P = 0.05; PT level = 36% ± 15% versus 15% ± 10%, P = 0.0075). None of these patients had encephalopathy (P = 0.02), and the mean Glasgow, score was 15 versus 11 in the other patients (P = 0.006). Six of the 14 patients deteriorated rapidly and were listed transplantation within 24 hours of admission. The average delay from admission to LT was 4.2 ± 4 days (range = 0-10 days). The final patient deteriorated after admission despite discontinuation of ATT and subsequently died (patient 6).

ATT in Nontransplant Patients

In 2 of these patients, definitive treatment was stopped on admission either because ATT had not been justified at the start (patient 14) or because the patient had completed treatment when severe hepatitis became evident (patient 8).

ATT was reintroduced in 5 patients who showed significant improvements in liver function and in 1 patient who showed partial and transient improvement. At the time of initiation of ATT, the mean values of the total bilirubin, PT, factor V, and ALT levels were 126 ± 138 μmol/L (range = 10-350 μmol/L), 71% ± 26% (range = 28%-100%), 81% ± 29% (range = 32%-119%), and 82 ± 69 IU/L (range = 30-77 IU/L), respectively. The majority of the patients (n = 4) received triple therapy, 1 received dual therapy, and 1 received quadruple therapy. The ATT regimen consisted of a quinolone in 5 patients, ETH in all patients, RIF or rifabutin (RFB) in 3 patients, and AMK or SMC in 3 patients. No patient received PZA, and only 1 patient received INH. Over the course of treatment, liver function continued to improve and reached normal limits in 5 patients (patients 1-5). One patient (patient 6) deteriorated despite treatment and died of sepsis and multiorgan failure (MOF) 52 days after admission. In this patient, despite only partial improvement in liver function, ATT was initiated because of acute worsening of his TB and HIV status. The average treatment duration was 6.5 ± 3.6 months (range = 3.2-11.2 months). Five of the 6 patients were alive 3.5 ± 2.6 years later (range = 0.6-5.8 years) with no evidence of TB.

ATT in Transplant Patients

One patient (patient 7) received ATT before LT because he presented with disseminated TB. At the start of treatment, his total bilirubin level was 192 μmol/L, and his PT level was 34%; he was not encephalopathic. Treatment consisted of RIF, ETH, MOX, and AMK and lasted for 8 days. In another patient (patient 9), an intraperitoneal node was discovered at the time of surgery, and a pathological examination revealed caseating necrosis. A culture of this node was positive for M. tuberculosis.

After LT, ATT was restarted in 5 patients (patients 7 and 9-12) after an average of 11.2 ± 12.3 days (range = 1-31 days). One patient did not receive ATT because he died suddenly after LT (patient 13). All patients received aggressive ATT consisting of a combination of 3 or 4 anti-TB drugs. The 2 patients treated with quadruple therapy received a combination of INH, ETH, and MOX plus AMK or RIF. Those treated with triple therapy (3 patients) received RIF and ETH plus OFX (2 patients) or SMC (1 patient). Maintenance therapy in the surviving patients consisted of dual therapy (2 patients) or triple therapy (1 patient). ATT was continued for an average of 8.2 ± 2.4 months (range = 5.6-10.2 months). Even though ATT consisted of INH, RIF, ETH, and MOX, 1 HIV-positive patient (patient 9) presented with a bowel obstruction 35 days after LT that was caused by intestinal miliary lesions. A histological examination revealed necrosing granulomata. The HIV viral load was then 4.9-log whereas it was undetectable on admission. In 1 patient (patient 7), AFB were isolated 3 days after transplantation despite 2 months of standard treatment before presentation with ALF and the rapid reintroduction of ATT on admission; this was due to the disseminating nature of the TB.

Histological Findings

None of the 13 patients with available histology had evidence of underlying chronic hepatitis. Confluent coagulative necrosis was observed in 11 of 13 patients, and this was either centrilobular with or without bridging or panlobular with an intensity ranging from 15% to 80%. There was mild to severe inflammatory infiltration consisting of lymphocytes (9/13), macrophages (3/13), polynuclear cells (5/13), and plasmacyte cells (6/13). Granulomas were seen in 4 patients without caseum.


No major side effects (particularly hepatotoxicity) were seen with any of the alternative ATT regimens, regardless of whether the patients underwent transplantation or not. Four of the 5 transplant patients who received ATT after transplantation (patients 9-12) developed acute rejection 25 ± 21 days after the start of treatment. During the days preceding the development of acute rejection, plasma levels of calcineurin inhibitors were below the therapeutic threshold despite daily increases in the doses of TCR and CSA (Table 2). The plasma concentrations of other immunosuppressants such as corticosteroids and mycophenolate mofetil (MMF) were not available. Three of these 4 patients were treated with steroid boluses. Similarly, the patient who had received a renal transplant 6 years earlier also had an episode of acute rejection. These 5 patients had all received RIF as part of their anti-TB regimen after transplantation.

Among the 3 patients who received AMK as alternative ATT (patients 1, 6, and 7), only 1 (patient 6) developed nephrotoxicity, which required cessation of treatment after 16 days. The origin of this nephrotoxicity was probably mixed: drug toxicity linked to AMK and hepatorenal syndrome. Treatment with AMK was well tolerated in 2 other patients (patients 1 and 7), and no additional toxicity was observed in the transplant patient (TCR; patient 7).


The overall survival rate was 71.4% (10/14). The survival rate of patients who improved spontaneously was 87.5% (7/8), and the survival rate of those who underwent transplantation was 50% (3/6).

Three of the 6 patients who underwent transplantation were still alive after an average of 11.4 ± 9.6 years (range = 1-20 years). One patient died as a result of disseminated TB (patient 9). Two other patients died because of either sepsis associated with hemorrhagic shock (patient 12) or acute respiratory distress syndrome (ARDS; patient 13). The average time from LT to death was 42.7 ± 29.7 days (range = 15-74 days).


Acute liver failure secondary to ATT is rare.28 In our series, these cases represented only 2.8% of 566 patients admitted with severe hepatitis between 1986 and 2008.29 According to a retrospective study of patients from the United Network for Organ Sharing registry (1987-2006), 0.07% of LT procedures for ALF were due to ATT (50 of 73,977 transplant patients).30 Forty-eight were caused by INH alone, and 2 were due to the association of PZA and RIF.

Current international recommendations concerning the management of TB in patients with severe hepatitis due to ATT requiring LT are unclear,1 and the anti-TB regimens used at the time of treatment reintroduction vary considerably in the literature (Table 3). The main results of the present series are that alternative ATT is generally well tolerated and effective in transplant and nontransplant patients without signs of TB recurrence, and all patients whose alternative ATT included RIF presented with acute rejection resolved by corticoid treatment.

Table 3. Review of the Literature on Patients with Anti-TB Therapy–Induced ALF
AuthorsNumberSex/Age (Years)TB LocalizationATTTime from the Onset of ATT to ALF (Days, Weeks, or Months)Liver Tests at the Time of ALFATT DiscontinuedNew ATTOutcomeATT After LT
  • *

    Before LT.

  • After the improvement of liver function tests.

  • No manifestations of hepatotoxicity after the reintroduction of ATT.

  • §

    Lost to follow-up (n = 41).

  • The standard regimen was stopped because of resistance to INH/RIF.

  • Associated carbamazepine for generalized epilepsy.

  • #

    Gradual reintroduction of the drugs and dose.

  • **

    Cessation of treatment due to the development of Stevens-Johnson syndrome and INH resistance.

  • ††

    Cessation of CTM due to abnormal liver function tests.

  • ‡‡

    With cirrhosis.

Farrell et al.441M/49Positive tuberculin testINH4 monthsBilirubin, 30.9 mg/dL; factor V, 26%INHAlive with LTNo
1F/60PulmonaryINH/RIF/PZA6 weeksBT, 22.4 mg/dL; PT time, 30 seconds; coma score, 3INH/RIF and then PZAOFX/ETH/SMCAlive with LTETH/SMC/OFX
Barcena et al.381M/39Vertebral TBINH/RIF/PZA1 monthBilirubin, 12.2 mg/dL; ALT, 1993 IU/LBilirubin, 38 mg/dL; HE INR, 8RIF (rash), INH/LFX, and PZAINH/PZA/ETH/LFX, PZA/SMC/ETH/CFX, and ETH/CFX/SMC*Alive with LTETH/CFX/SMC
Pessayre et al.456F (5) and M (1)/46 ± 24 (15-69)Renal (1), peritoneal (2), spondylitis (1), pleural (1), or intestinal TB (1)INH/RIF/SMC (2) or INH/RIF/ETH (4)7.3 ± 1.8 days (6-10 days)Bilirubin, 11.25 ± 2.7 μmol/L; PT, 13.25% ± 2.36%; HE (6)INH/RIF in allRIF/ETH/SMC (1), RIF/ETH (3), or no treatment (2)Alive without LTNA
Singh et al.3172F (35) and M (37) (41±12) INH/RIF/PZA35 ± 32 daysLT anomaliesINH/RIF/PZAINH/RIF/PZA (44§)Alive and successful (93.2%) Recurrence of ATT (14%) Died with FH (9)ND
Mitchell et al.461F/59Cervical TBINH/RIF/PZA28 daysBilirubin, 310 μmol/L; INR, 3; HEAll ATT; reintroduction of INH (1 month later) and later cessation of INH and RIF and then PZA and INHCFX/SMCAlive with LTND
1M/61Urinary TBINH/RIF/PZA2 monthsBilirubin, 656 μmol/L; INR, 7; HEAllDied after LT (sepsis and MOFNA
1F/56Pulmonary TBINH/RIF/PZA6 monthsDied without LTNA
1F/31Pulmonary TBINH/RIF/ETH/PZA5 monthsINH/RIF/ETH/ PZA (rash)NDAlive without LTNA
Marra et al.471F/28Tuberculous lymphadenitisCFX/PZA/ETHND AllNoAlive with LTLFX/SMC/ AMK/ETH
Cillo et al.481M/10Positive tuberculin testINH3 monthsBilirubin, 275 μmol/L; INR, 1.7INHNoAlive with LTRIF/SMC/ CFX/AMOX
Campos- Franco et al.491F/16Intestinal TBINH/RIF/PZA5 daysBilirubin, 3.2 mg/dL; PT, 12%; HEAllRIF/PZAAlive without LT 
Tahaoglu et al.3245M (30) and F (15) (40±16)Pulmonary (37) or pleural TB (8)INH/RIF/PZA/ ETH or SMC17.3 ± 18.7 days (6-102 days)Bilirubin, 1.55 ± 1.58 mg/dLAllETH/SMC plus INH/RIF# (20 patients) or INH/RIF/PZA/ ETH (25 patients)Alive without recurrence of hepatotoxicity or recurrence (6)NA
Idilman et al.81F/19Ovarian peritoneal TBINH/RIF/ ETH/PZA3 daysBilirubin, 10.5 mg/dL; PT time, 100 seconds; HEAllAlive with LTCycloserine/ CFX/SMC/ETH
Nash et al.501M/44Tuberculous cervical lymphadenitisINH/RIF/ETH/PZA,** CFX/CTM,†† CFX, and CFX/SMC/RIF>9 weeksBilirubin, 356 μmol/L; INR, 2.5CFX/SMC/RIFAlive with LTMOX/SMC/ETH
Wang et al.401M/30‡‡PulmonaryINH/RIF/ETH/PZA2 monthsBilirubin, 574 μmol/L; PT time, 60 secondsAllAlive with LTMOX/INH/ ETH/RFP

The risk of recurrence of hepatotoxicity with INH and/or PZA cannot be excluded during the reintroduction of standard ATT. The risk of recurrence was 14% and 24% in the studies of Singh et al.31 and Tahaoglu et al.,32 respectively. The reintroduction of PZA and a full dose from day 1 of the treatment reintroduction are both risk factors.32 However, these results were not confirmed in the randomized study of Sharma et al.,33 who included 175 patients with acute hepatitis due to ATT. In effect, regardless of the protocol for the reintroduction of ATT (the maximum dose of INH, RIF, and PZA from day 1 or the sequential reintroduction of increasing doses of INH, RIF, and PZA), the risk of recurrence of hepatotoxicity was the same. The authors concluded that the reintroduction of these 3 drugs at the full dose and simultaneously was safe. However, in comparison with our patients, hepatic disease was less severe (a mean total bilirubin level of 11.3 μmol/L in their study versus 179.8 μmol/L in our study), and this may explain the difference in the interpretation of the results. In addition, we must ask whether this is a risk worth taking when a number of effective and nonhepatotoxic drugs are currently available.1 Four of the 5 evaluable patients who were treated after admission with alternative ATT (including nonhepatotoxic drugs) did well, recovered normal liver function, and were free from TB over the medium and long term.

Opinions concerning ATT after LT also vary. Several authors have reported that standard ATT consisting of INH, RIF, PZA, and ETH or regimens including 1 or more potentially hepatotoxic drugs are effective and well tolerated if the patients are closely monitored7 (Table 4). However, in a study by Meyers et al.,6 standard ATT (triple or quadruple therapy) had to be stopped in 5 of 6 patients (83.3%) and replaced with an alternative regimen consisting of 2 anti-TB drugs (ETH and FQ). In a review of the literature from 1967 to 1997 that included 511 patients who presented with TB after solid organ transplantation, INH hepatotoxicity, defined as an increase in aminotransferases alone, was noted in 41% of LT recipients.34 INH hepatotoxicity was more prevalent because this drug is often used in conjunction with other anti-TB drugs such as RIF, and the toxic effects are often additive.35, 36 More recently, in a meta-analysis reviewing 139 LT patients, 86 had received a recognized ATT regimen. This induction regimen consisted of some combination of INH (96%), RIF or RFB (76%), ETH (81%), PZA (51%), FQ (31%), and SMC (17%). For maintenance, 70% of patients received INH, 45% received RIF or RFB, 18% received PZA, 73% received ETH, 14% received SMC, and 52% received a quinolone. Thirty of the 86 patients (35%) presented with side effects requiring withdrawal from ATT or ATT modification. Twenty-four of the 86 patients (28%) presented with hepatotoxicity; 22 had received INH, which was associated with RIF or RFB in 18.37 Although the use of standard ATT may be considered when TB occurs after solid organ transplantation and notably after LT, the reintroduction of drugs able to provoke ALF is difficult to justify after transplantation. With the aim of quantifying the risk of toxicity with INH, Barcena et al.38 took biopsy samples from liver grafts to determine the genotype of cytochrome P450-2 E1 and the status of N-acetyl transferase, and they showed that these 2 components play an important role in INH toxicity. However, this is not practicable in routine practice, and the use of an anti-TB regimen that excludes hepatotoxic drugs and for which the efficacy has been demonstrated remains the best alternative.

Table 4. Previous Case Reports of TB in LT
AuthorsNumberLT IndicationSex/Age (Years)TB SiteDuration from LT to TB Therapy (Days, Months, or Years)TB TherapyDuration of TB Therapy (Months or Years)OutcomeTolerance
  • *

    There was a delay between LT and the diagnosis of TB.

  • The diagnosis of TB was made before LT; the patient was treated with INH/ETH/MOX/AMK.

  • ATT was modified after the histological diagnosis of hepatotoxicity.

  • §

    Average age.

Torre- Cisneros et al.512FH undeterminedM/31 ImmediateINH/PZA/ETH and INH/PZA2 and 16 monthsAlive and free of disease (18 months)Acute rejection on days 28-40
HBV FHF/42Pulmonary2 weeksINH/PZA/ETH/ OFX and INH/ OFX2 and 18 monthsAlive and free of disease (22 months)No rejection
Clemente et al.75HCV cirrhosisF/44Disseminated2.7 months*INH/RIF/PZA1 yearAlive and free of disease (27 months)No rejection
α1-Antitrypsin deficiencyF/21Pulmonary7.5 years*INH/RIF/PZA1 yearAlive and free of disease (29 months)No rejection
HCV cirrhosisF/47Disseminated11 months*INH/RIF/PZA1 yearAlive and free of disease (6 months)No rejection
Autoimmune cirrhosisF/25 and 28Pulmonary/ extrapulmonary41 days and 3 years*INH/RIF/PZA and then INH/ RIF/PZA/ETH6 months and 1 yearAlive and free of disease (31 months)No rejection
Primary biliary cirrhosisF/61Extrapulmonary8 months*INH/RIF/PZA, then ETH/OFX/ SMC, and then INH/RIF/PZA15 monthsAlive and free of disease (1 month)Acute rejection
Wang et al.402HBV FHM/48Pulmonary1 dayMOX/INH/RFP/ETH and then MOX/ETH12 monthsAlive with LTATT toxicity, no rejection
HBV cirrhosisF/50PulmonaryNDMOX/AMK/INH/ETH and then INH/ETH12Alive with LTNo rejection
Chan et al.398HBV cirrhosis (7) or HCC (1)M (7) and F (1)/47§Pulmonary (6), urinary bladder (1), or positive tuberculin test (1)Median: 9 months (4-20 months)RIF/INH/ETH/LFX (5: 1 with PZA, 2 with AMK, and 1 with SMC), INH/ ETH/SMC/OFX (2), or RIF/INH/ ETH/PZA (1)11-18 monthsAlive with LDLT or DDLTNo rejection
Torre- Cisneros et al.528Viral cirrhosis (4), alcoholic cirrhosis (1), or HCC (3)M (6) and F (2)/58§Pulmonary (5), extrapulmonary (2), or disseminated (1)Median: 5.9 months (0.9-10.6 months)INH/RIF/PZA (5), INH/RIF/ETH (1), INH/PZA/ETH (1), or INH/RIF/ETH/ PZA (1)Not relatedAlive (6) or dead (2)Not related

The risk of interactions between RIF and immunosuppressant drugs after transplantation suggests that RIF should be avoided in anti-TB regimens. RIF is an enzyme inducer acting at the level of cytochrome P450 and could increase the catabolism of immunosuppressive drugs such as TCR, CSA, and corticosteroids with the direct result being acute rejection. In a meta-analysis by Holty et al.,37 39% of the patients who had received RIF or RFB needed an adjustment of their immunosuppressant dosage. All 5 transplant patients (patients 2 and 9-12) whose ATT included RIF had acute rejection treated by a corticoid bolus and readjustment of immunosuppression. Our results may appear to disagree with those of other studies in which the rate of rejection was lower.7, 34, 37, 39, 40 One possible explanation is that in most of these studies, TB became apparent a long time after transplantation (ie, long after the time of maximal risk of acute rejection in patients with a lower level of immunosuppression). RFB has been proposed as an alternative to RIF in the treatment of TB in transplant patients. It is as effective as RIF41 and allows better control of plasma concentrations of immunosuppressants; as a result, it reduces the risk of acute rejection.42 However, although drug interactions are less of a problem in comparison with RIF, they may still occur, and close monitoring of immunosuppressants is necessary.1 Despite this close monitoring, the need for readjustment of the doses of immunosuppressants may lead to some delay before therapeutic thresholds are reached. The nephrotoxicity of AMK, which is associated with that of calcineurin inhibitors, may limit its use. Close monitoring is also required here in order to maintain normal effective therapeutic levels and avoid overdosing. The recommended dose of AMK is 15 mg/kg 5 to 7 times per week and then 2 to 3 times per week for a period of 2 months in order to reduce the risk of nephrotoxicity. In the case of a change in renal function, the dose should be adapted to the renal function.

The duration of ATT in transplant patients remains poorly defined and differs between studies. Meyers et al.6 proposed treatment for 12 to 18 months. Clemente et al.7 estimated that bactericidal ATT (INH, RIF, and PZA) required 1 year.43 In our study, the average duration of treatment after transplantation was 8.4 months (range = 5.9-10.3 months) in patients who survived. The duration of treatment seems to depend on the therapeutic protocol used: quadruple therapy in the initial high-intensity phase for 2 months followed by dual therapy in the continuation phase or triple therapy for 3 months followed by maintenance for 7 months when 1 of the 3 major bactericidal drugs (INH, PZA, and RIF) cannot be used. When the treatment regimen consists of only 1 or 2 of the major anti-TB drugs, the treatment time is generally longer. In addition, the duration of the initial high-intensity treatment that the patient had received before he developed hepatitis should also be included as long as the treatment cessation lasted no longer than 2 weeks. In our study, the average duration of the high-intensity phase in the 3 surviving transplant patients who received alternative ATT was 1.52 ± 0.63 months (range = 5.9-10.3 months). Only 1 patient stopped treatment for less than 2 weeks and would therefore be considered successful (Table 2).

In conclusion, in approximately 50% of cases of severe acute hepatitis secondary to ATT (INH and PZA), the discontinuation of hepatotoxic medication allows for improvement of liver function. After improvement of hepatic function, ATT consisting of second-line drugs can be used to good effect with less worry about further hepatotoxicity. After LT, treatment options remain difficult once the standard ATT can no longer be used. However, it is important for the following points to be considered: (1) RIF should be avoided because it can interfere with immunosuppressants and lead to acute rejection, and (2) PZA should be avoided because of its hepatotoxicity. One possible anti-TB regimen after LT could be INH (as long as it is not implicated in the initial hepatitis), ETH, and FQ (particularly MOX) with or without AMK.