ATT Hepatitis: The Evidence Behind the Pathway

A 52-year-old woman weighing 50 kg started the four first-line drugs for pulmonary tuberculosis three weeks ago: isoniazid, rifampicin, pyrazinamide and ethambutol (HRZE). For four days she has had no appetite, and she has been vomiting. This morning her daughter noticed her eyes were yellow. Her ALT is about 6 times the upper limit of normal (ULN), and her total bilirubin is 3.2 mg/dL. Her INR is normal, and she is alert and oriented. This is an illustrative case, not a real patient.

By the end of this article you will know whether to stop her drugs, and which of them is the likely culprit. You will know how to keep treating her tuberculosis while the liver recovers, and how to bring the drugs back.

1. Not Every Rise Is an Injury

Start with the patient who is well. The tool reminds us that a mild, asymptomatic rise in the transaminases, up to 3 times the ULN, occurs in up to 20% of patients starting ATT. This is hepatic adaptation to isoniazid. It is transient, and it settles on its own while treatment continues.

“Then why not stop the drugs, just to be safe?” Because stopping is not safe. Unnecessary stopping of ATT in these asymptomatic patients is a major cause of treatment failure and acquired resistance. So the tool keeps the drugs going below 3 times the ULN when the patient is well and the bilirubin is normal. Between 3 and 5 times the ULN without symptoms, it still says do not stop yet. Repeat the liver tests in 3 to 5 days, and teach the patient the warning signs.

2. When the Drugs Must Stop

The tool stops the drugs at any one of three points:

• an AST or ALT above 5 times the ULN;
• above 3 times the ULN with symptoms of hepatitis (anorexia, nausea, vomiting, jaundice);
• a total bilirubin of 1.5 mg/dL or above, whatever the enzymes.

Our patient has crossed all three.

One more question comes first: is the liver failing? An INR above 1.5, encephalopathy or ascites is a medical emergency. All ATT stops, and the patient goes to a tertiary intensive care unit or a hepatology centre at once. The tool adds a point that surprises many residents: at that stage the transaminases mean nothing, because massive necrosis leaves too few surviving liver cells to release the enzyme. The numbers fall while the patient gets worse.

And before you blame the ATT, remember where we work. India is highly endemic for viral hepatitis. So the tool asks you to rule out acute hepatitis A, B, C and E, with dengue and malaria, before a severe transaminitis is put down to ATT alone.

3. Which Drug, and How

A bridging regimen is only as safe as the drugs left in it. So you need to know which agent is most likely to have caused the injury, and how. The tool sets this out in one table. In the tool it is now shown as two narrower tables, each keeping the drug class and the agent; here it is as one:

Read the first four rows as a story about our patient. Pyrazinamide carries the highest risk, and its injury is dose-dependent, direct structural damage, with the highest risk of fulminant necrosis. Isoniazid injures the liver in an idiosyncratic way, through its toxic acetylhydrazine metabolites. The risk rises with age, and in slow acetylators, people in whom the NAT2 enzyme works slowly. Rifampicin is different again. It mainly raises the unconjugated bilirubin, by competing for the way bilirubin is excreted into the bile, rather than by killing liver cells. Ethambutol is cleared by the kidney, and that is why it is the backbone of every liver-sparing bridge.

“So which one did it?” At this point you cannot know. The table tells you which drugs are likely, and which are safe to keep. The rechallenge, below, is what names the culprit.

4. Treating the Tuberculosis While the Liver Recovers

Stopping all ATT stops the treatment of the tuberculosis too. So the tool builds a bridge from drugs that are not hepatotoxic. It insists on a rule of 3: the bridge must hold at least three such drugs, to prevent acquired drug resistance and disease progression while the liver recovers. For a drug-sensitive TB, it is ethambutol, levofloxacin and a third agent, cycloserine or amikacin, each dosed by the NTEP weight band. At 50 kg, our patient is in the 46 to 70 kg band: ethambutol 1200 mg and levofloxacin 1000 mg once daily, with cycloserine 750 mg a day as the third drug.

The table explains each choice. Levofloxacin is safe for bridging unless there is documented fluoroquinolone resistance. Cycloserine is not hepatotoxic, but watch for psychosis and seizures. The aminoglycosides are renally cleared, with no hepatic metabolism, so their risks are to the kidney and the ear instead.

In drug-resistant TB the tool refuses to build the standard bridge. A second-line regimen stripped down to its non-hepatotoxic drugs leaves too few active agents against an organism that is already resistant. That is functional monotherapy. The Nodal DR-TB Centre builds an individualised liver-sparing regimen instead.

5. Bringing the Drugs Back

Wait until three things are true: the AST and ALT are below 2 times the ULN, the total bilirubin is below 1.5 mg/dL, and the symptoms have gone completely. Check the liver tests every 3 to 7 days until then. Putting a hepatotoxic drug back too early lands a second injury on a liver still repairing the first.

Then the drugs come back one at a time, starting with rifampicin, the least hepatotoxic and most efficacious. Rifampicin 150 mg on days 1 to 3, then the full dose on days 4 to 6, with a liver test on day 6. If it is normal, isoniazid 100 mg on days 7 to 9, then the full dose on days 10 to 12, with a liver test on day 12. If symptoms or a rise in the enzymes come back at any step, the last drug added is the culprit, and it is stopped for good.

“And pyrazinamide?” If the first hepatitis was severe (grade 3 or 4, or with prolonged jaundice), the tool follows the international consensus and abandons pyrazinamide permanently. Bringing it back after an injury of that severity carries a risk of a recurrent hepatitis that has proved fatal. The patient completes an extended 9-month regimen instead: 2 HRE / 7 HRE.

Back to Our Patient

She has symptoms, an ALT about 6 times the ULN and a bilirubin of 3.2 mg/dL: all ATT stops. Her INR is normal and she is alert, so she is not in liver failure. A viral hepatitis screen goes off, with dengue and malaria if the season fits. She starts the bridge: ethambutol 1200 mg and levofloxacin 1000 mg once daily, and cycloserine 750 mg a day. Her liver tests are repeated every 3 to 7 days. When her ALT is below 2 times the ULN, her bilirubin below 1.5 mg/dL and she is well, rifampicin comes back, then isoniazid. Whether pyrazinamide follows depends on how severe her hepatitis proves to be. If it was grade 3 or 4, or her jaundice was prolonged, pyrazinamide stays out for good, and she completes 2 HRE / 7 HRE.

The Common Mistake

The common mistake is the opposite of what you might expect: stopping ATT too readily. It is easy to see why. An abnormal liver test on a drug chart feels like something must be done. But a mild rise, in a well patient, is usually adaptation. Stop for that, and you trade a liver test for a treatment failure. Stop when the tool’s criteria are met, and not before.

Try This

A 30-year-old man is two weeks into HRZE. He feels well and is eating normally. His ALT is 2.5 times the ULN, and his bilirubin is normal. What do you do? Illustrative case.

Answer. Continue the ATT without interruption. He is asymptomatic, his enzymes are below 3 times the ULN and his bilirubin is normal. The cessation criteria are not met, and a mild rise like this is expected hepatic adaptation. Reassure him, keep to the routine monitoring, and tell him to stop the tablets and come back at once if anorexia, nausea, vomiting or jaundice appear.

Take-Home Points

• A rise up to 3 times the ULN in a well patient is usually hepatic adaptation: do not stop ATT for it.
• Stop ATT above 5 times the ULN, above 3 times with symptoms, or at a bilirubin of 1.5 mg/dL or above.
• Rule out viral hepatitis A, B, C and E, dengue and malaria before blaming the ATT.
• Bridge with at least three non-hepatotoxic drugs: ethambutol, levofloxacin and cycloserine or amikacin; in DR-TB, ask the Nodal DR-TB Centre.
• Bring the drugs back one at a time, rifampicin first; after a severe hepatitis, leave pyrazinamide out for good.

How the first-line drugs are dosed and what else each one does is in Tuberculosis Treatment: The Evidence Behind the Pathway.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried on 5 November 2026, in its Evidence tab and in the plan it prints, each from the NTEP, ICMR and international guidance the tool names. The illustrative cases follow the tool’s rules; check the tool’s result for a real patient.

References

1. Central TB Division, Ministry of Health and Family Welfare, Government of India. Guidelines for Programmatic Management of Drug-Resistant Tuberculosis in India. New Delhi: MoHFW; March 2021, with the 2024 guidance on BPaLM.
2. Indian Council of Medical Research. Standard Treatment Workflow: Anti-Tubercular Therapy (ATT) Related Hepatitis. New Delhi: ICMR; 18 March 2022.
3. Devarbhavi H, et al. Drug-induced liver injury: Asia Pacific Association of Study of Liver consensus guidelines. Hepatol Int. 2021;15(2):258-282.
4. Saukkonen JJ, et al. An official ATS statement: hepatotoxicity of antituberculosis therapy. Am J Respir Crit Care Med. 2006;174(8):935-952.
5. Nahid P, Dorman SE, Alipanah N, et al. Official American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America Clinical Practice Guidelines: Treatment of Drug-Susceptible Tuberculosis. Clin Infect Dis. 2016;63(7):e147-e195.
6. Sharma SK, Singla R, Sarda P, et al. Safety of 3 different reintroduction regimens of antituberculosis drugs after development of antituberculosis treatment-induced hepatotoxicity. Clin Infect Dis. 2010;50(6):833-839.
7. Singh J, Garg PK, Tandon RK. Hepatotoxicity due to antituberculosis therapy. Clinical profile and reintroduction of therapy. J Clin Gastroenterol. 1996;22(3):211-214.
8. Sharma SK, Balamurugan A, Saha PK, Pandey RM, Mehra NK. Evaluation of clinical and immunogenetic risk factors for the development of hepatotoxicity during antituberculosis treatment. Am J Respir Crit Care Med. 2002;166(7):916-919.
9. Tostmann A, Boeree MJ, Aarnoutse RE, de Lange WCM, van der Ven AJAM, Dekhuijzen R. Antituberculosis drug-induced hepatotoxicity: concise up-to-date review. J Gastroenterol Hepatol. 2008;23(2):192-202.
10. World Health Organization. WHO consolidated guidelines on tuberculosis. Module 4: treatment. Drug-susceptible tuberculosis treatment. Geneva: WHO; 2022.