Utility of Liver Biopsy in the Diagnosis and Management of Possible Drug-Induced Liver Injury in Patients Receiving Antituberculosis Therapy: A Retrospective Study

Background: Drug-induced liver injury (DILI) secondary to ATT treatment (TB-DILI) is reported in 2–28% of patients. We present here a series of clinical cases of suspected DILI arising during antituberculosis treatment, studied with the aid of liver biopsy. Methods: this was a retrospective descriptive study including 10 tuberculosis patients who underwent liver biopsy for suspected TB-DILI at the “Lazzaro Spallanzani” Institute from 2017 to 2022. Results: Ten patients who underwent LB were extracted from the database and included in the retrospective study cohort. According to the clinical classification, eight patients had hepatocellular liver injury, one patient had cholestatic injury, and another had mixed-type injury. Histopathological diagnosis revealed liver damage due to DILI in 5/10 (50%) cases. In one case, liver biopsy showed necrotizing granulomatous hepatitis. Conclusions: Severe and persistent elevation of hepatic transaminases, hepatic cholestasis despite discontinuation of therapy, and other suspected hepatic conditions are indications for liver biopsy, which remains a valuable tool in the evaluation of selected tuberculosis patients with suspected DILI for many reasons. However, the decision to perform a liver biopsy should be based on clinical judgment, considering the benefits and risks of the procedure.


Introduction
Tuberculosis (TB) is a significant public health problem globally, with approximately 10 million cases and 1.4 million deaths per year [1].Antituberculous therapy (ATT) is the basis of tuberculosis treatment.However, among the common side effects of ATT is hepatotoxicity [2,3].Drug-induced liver injury secondary to antituberculosis treatment (TB-DILI) is reported in 2-28% of patients, varying according to definition, study sample and treatment regimen.TB-DILI is ultimately a clinical diagnosis of exclusion [4].Other causes of liver damage, such as acute viral hepatitis, should always be looked for carefully and, in their absence, the diagnosis of DILI appears plausible.In cases of suspected DILI, it is important to assess causality using scoring systems such as the Roussel Uclaf Causality Assessment Method (RUCAM) score [5].To help define DILI, the DILI Expert Working Group requested consensus on threshold criteria for any drug-related DILI, including criteria based on ALT and ALP [6].
Infect.Dis.Rep. 2023, 15 736 Regarding the management of this condition, the best approach is still a matter of debate.Liver biopsy (LB) is an invasive procedure that can help clinicians in the diagnosis and management of ATT-induced liver injury.In clinical practice, it is performed when the diagnosis is uncertain or there is concern about the severity of the lesion.LB is not mandatory in the DILI assessment and is not required for the RUCAM score.Commonly, the biopsy is performed early and can guide subsequent management such as initiation of steroid therapy or referral for liver transplantation [7].We conducted a retrospective analysis of a series of patients treated at a tertiary care tuberculosis hospital center in Italy who underwent LB for suspected TB-DILI, with the aim of describing the utility of biopsy for the differential diagnosis of liver disease and its treatment during anti-TB treatment.

Materials and Methods
We conducted a retrospective descriptive study of a case series including 10 patients who underwent liver biopsy for the diagnosis of TB-DILI at the "Lazzaro Spallanzani" Institute from 2017 to 2022.The patients were treated based on the institutional protocol, which was written following WHO guidelines on tuberculosis.In all cases, an individualized regimen designed based on the results of drug susceptibility tests and adapted to comorbidities was used [8][9][10].Initial treatment was provided on an inpatient basis, until acid-fast bacillus (AFB) sputum conversion was achieved on three consecutive negative specimens collected over the course of one week.Patients received therapy with direct observation during hospitalization.During hospitalization, laboratory tests were repeated almost weekly, or as needed based on clinical conditions, and monthly or as needed in the subsequent period.Baseline liver tests (LTs) were performed, and all patients were offered baseline screening for HIV and hepatitis B and C. Clinical and blood evaluations were repeated at each visit, including actively and systematically looking for symptoms such as nausea, vomiting, abdominal pain, dark urine, itching and loss of appetite.
Patients with abnormal LTs were managed according to institutional guidelines, which are based on the indications of the ATS [11].Antituberculous therapy was modified or discontinued in asymptomatic patients with ALT 3-5xULN, who were carefully monitored.In cases of abnormal liver biochemistry, risk factors considered included age, sex, baseline LT, HIV, hepatitis B and C, hepatitis E and A, CMV, EBV, HSV, alcohol consumption and chronic liver disease.ANA, ASMA and LKM-1, anti-neutrophil cytoplasmic AMA (ANCA) and anti-double-stranded DNA were determined in all patients with suspected TB-DILI.Laboratory tests considered included aspartate aminotransferase (AST; normal range 5-40 U/L), alanine aminotransferase (ALT; normal range 5-40 U/L), total bilirubin (T-Bil; normal range 0.2-1.0mg/dL), alkaline phosphatase (ALP; normal range 40-116 U/L) and gamma-glutamyltransferase (GGT; normal range < 38 U/L).Patients were classified as DILI cases if the following conditions were met: alanine aminotransferase (ALT) levels ≥ 5 times the upper limit of normal, alkaline phosphatase (ALP) ≥ 2 times the upper limit of normal and total bilirubin levels > 2 times the upper limit of normal, after excluding other potential causes.
Classification of DILI patients' clinical presentation was based on the RUCAM method.The R-ratio value was calculated as follows: ALT value/LT upper limit of the normal range (ULN)/ALP value/ALP ULN.The suspected DILI cases were classified into three patterns: hepatocellular pattern of DILI = R > 5; mixed pattern of DILI = R between 2 and 5; and cholestatic pattern = R < 2 [5].
DILI severity grade was determined as defined by the DILI Expert Working Group, described as follows: Grade 1, mild: elevated alanine aminotransferase/alkaline phosphatase (ALT/ALP) concentration meeting criteria for DILI but bilirubin concentration < 2× the upper limit of normal (ULN).Grade 2, moderate: elevated ALT/ALP concentration meeting criteria for DILI and bilirubin concentration ≥ 2× the ULN, or symptomatic hepatitis.Grade 3, severe: elevated ALT/ALP concentration meeting criteria for DILI, bilirubin concentration ≥ 2× the ULN and one of the following: INR ≥ 1.5; ascites and/or encephalopathy, disease duration < 26 weeks and absence of underlying cirrhosis; and other organ failures considered to be due to DILI.Grade 4, fatal or transplantation: death or transplantation due to DILI [6].
In the event of severe TB, we prescribed a non-hepatotoxic background regimen (according to local protocol) whilst stopping the most hepatotoxic drugs.LTs were checked at intervals until ALT was normal or approaching normal.Treatment was re-introduced using a staggered incremental dose increase regimen with omission of incriminated drugs and consequential extension of treatment.
We considered as "persistent" the cases in which liver enzymes remained elevated for more than four weeks despite discontinuation of therapy.For TB patients, this interval is crucial, because it corresponds to the time after which we must re-initiate the TB treatment in cases of suspension.
Severe and persistent elevation of liver transaminases and or hepatic cholestasis despite discontinuation of therapy, along with other suspected liver-related conditions, were the indications for liver biopsy (LB) in ten patients.
The liver biopsy was performed under ultrasound or CT guidance (core not smaller than 1 cm, taken intercostally from the right lobe of the liver) under local analgesia with xylocaine.Written informed consent was obtained from all patients.Routine use of ultrasound was helpful in determining the best location for the biopsy.The area was prepped and draped in a sterile manner.The biopsy frustule was fixed in formalin and embedded in paraffin.LB patients underwent routine treatment for histopathological examination.Tissue sections were stained with hematoxylin and eosin, Masson's trichrome, Perls' stain, PAS and PAS-Diastase (PAS-D).CK7 immunostaining was performed to show ductular reaction and ductular metaplasia of hepatocytes.For molecular studies, DNA was selectively extracted with a QIAamp Tissue Kit (Qiagen GmbH, Hilden, Germany) from paraffin sections, and PCR detection of mycobacterium DNA was performed in all samples.Two experienced histopathologists evaluated the histological changes in the liver sections.
Data were collected from electronic medical records.Socio-demographic characteristics, including sex, age, ethnicity and smoking status, as well as clinical characteristics such as comorbidities and any ongoing therapy, were recorded.Descriptive analysis was conducted to characterize subjects enrolled in this study.Statistical analyses were conducted using StataCorp.2023 (Stata Statistical Software: Release 13, StataCorp LP, College Station, TX, USA).

Results
Ten patients who underwent LB in the period from January 2017 to December 2022 were extracted from the database and included in the retrospective study cohort.The main characteristics of the 10 patients are shown in Table 1.Most patients (70%) were male.The mean age of the patients was 38.3 (range 19-66) years.None of the patients were HIV-positive.Two patients tested positive for anti-HCV antibodies (viral load detectable in both patients).All patients had a form of pulmonary tuberculosis.One also had pleural localization and another hepatic (post-biopsy diagnosis).Seven patients (70%) had new cases of tuberculosis.Half of the ten patients (50%) had completely drug-sensitive tuberculosis, while five had rifampicin-resistant/multidrug-resistant (RR/MDR) tuberculosis [12].The most common comorbidities were chronic hepatitis C (2/10) and cardiovascular disease (3/10).Half of the patients (5/10) had a history of alcohol abuse.
With regard to concomitant medications, two patients were on long-term antihypertensive drugs, one patient was on long-term hypoglycemic drugs, and one patient was on long-term oral anticoagulant (warfarin).None of the patients were going through immunosuppressive therapy.The number of drugs included in the ATT regimen in this cohort ranged from four to six (Table 1).The time interval from the onset of antituberculosis therapy to the detection of hepatotoxicity ranged from 30 to 330 days, with a mean time of 104 days.All patients had normal liver function before the diagnosis of TB-DILI.Legends: HBV = hepatitis B virus; HCV = hepatitis B virus; AFB = acid-fast bacilli; Lfx = levofloxacin; Cs = cycloserine; PAS = para-aminosalicylic acid; Lzd = linezolid; E = ethambutol; Amk = amikacin; H = isoniazid; Z = pyrazinamide; Cfz = clofazimine; Bdq = bedaquiline; R = rifampicin; Pulm: pulmonary TB; EPTB: extrapulmonary TB; TD: total days of exposure before AST/ALT elevation.
As for the severity of liver injury, eight patients had grades 1-3 DILI while only two patients had grades 4-5 DILI.In all patients, DILI required temporary suspension of the antituberculosis treatment.All patients were treated after onset of DILI according to local protocol and all completed treatment.The clinical manifestations of anti-TB DILI reported in our study population included nausea and loss of appetite (n = 10), fatigue (n = 10) and rash (n = 1).According to the clinical classification, eight patients were diagnosed with the hepatocellular type of DILI, one patient had the cholestasis type, and one patient reported a mixed-type injury.The laboratory analyses are shown in Table 2.
Histopathological diagnosis revealed liver damage due to DILI (Figure 1) in 5/10 (50%) cases (Table 2).The other cases were diagnosed as chronic viral hepatitis 2 (20%), along with one (10%) case of nonalcoholic fatty liver disease (NASH), one (10%) case of granulomatous hepatitis due to tuberculosis and one (10%) case of minimal and nonspecific hepatitis.Histological examination revealed a mild-to-moderate inflammatory infiltrate of the portal tract with lymphocytes (Figure 1A), macrophages with PAS D pigment mixed with some neutrophils and rare eosinophils (Figure 1B).Intralobular inflammation with confluent necrosis (Figure 1E) and aggregates of macrophages with positive PAS-D pigment (Figure 1C,D) were observed.One case showed large and small droplet steatosis affecting more than 5% of the liver parenchyma (Figure 1F).In one case, liver biopsy showed necrotizing granulomatous hepatitis suggestive of mycobacterial infection and confirmed to be mycobacterium tuberculosis via PCR testing.All other patients had negative results on liver samples.confirmed to be mycobacterium tuberculosis via PCR testing.All other patients had negative results on liver samples.Table 2 shows the four anatomopathological criteria identified by the authors as specific to DILI (with a sensitivity of 91%) in the pathological description of the biopsies of our patients: portal inflammation, fibrosis, portal neutrophils and plasma cells and intracellular (hepatocellular) cholestasis.We found an association between the criteria and the diagnosis of DILI (mean value of five non-DILI patients 1.6; mean value of five patients with DILI 3.6; p = 0.0077, comparison of two means-Med calc).Table 2 shows the four anatomopathological criteria identified by the authors as specific to DILI (with a sensitivity of 91%) in the pathological description of the biopsies of our patients: portal inflammation, fibrosis, portal neutrophils and plasma cells and intracellular (hepatocellular) cholestasis.We found an association between the criteria and the diagnosis of DILI (mean value of five non-DILI patients 1.6; mean value of five patients with DILI 3.6; p = 0.0077, comparison of two means-Med calc).

Discussion
The present study describes the histopathological features of liver biopsies from ten tuberculosis patients who developed suspected DILI during treatment.
Effective treatment of tuberculosis requires a combination of bactericidal and/or bacteriostatic drugs.The combination of these drugs is regulated by the recommendations of the World Health Organization (WHO) [8,9].All adverse events reduce the effectiveness of treatment, possibly contributing to treatment failure, relapse or the emergence of drug resistance [13].
During the study period, we diagnosed 1900 new cases of pulmonary and extrapulmonary tuberculosis at the INMI Spallanzani and started their treatment.We observed increased transaminases in 16% of patients.Symptoms and liver abnormalities usually resolved after stopping treatment.The rechallenge of drugs was progressive and chosen according to local institutional protocol.
Hepatic drug reactions can occur at any time during the treatment period, although DILI has been described to generally occur within 8 weeks of initiation [14].In our study, 7 out of 10 patients (70%) developed hepatitis within 2 months (mean time 104 days; range 30-330), consistent with other published studies [15,16].
Well-known risk factors for TB-DILI are a low body mass index (BMI), metabolic isoniazid acetylator (INH) status, age, gender, metabolic factors, drug interactions and consumption of alcohol [17].
Although no patients reported alcohol consumption during treatment, it cannot be conclusively stated that, given the high percentage of referred history of alcohol intake (50%), concomitant unreferred use of alcohol did not influence the severity of the episode [18].Five patients in this case series were RR/MDR patients.Risk factors of TB-DILI or abnormal liver functioning include the type and dosage of drugs, the duration of medication and the combination of medications [19,20].We previously reported that 8.1% of MDR TB patients experienced liver toxicity, demonstrating that the onset of drug-induced hepatitis is also a relevant event in the treatment of MDR/XDR TB [13].
There is no widely agreed-upon definition of TB DILI, and most definitions focus on the value of ALT.The DILI Expert Working Group has defined consensus threshold criteria valid for any drug-related DILI, including ALT-and ALP-based criteria [21].In this context, the lack of standardized diagnostic criteria makes the diagnosis difficult even for experts, so liver biopsy could prove useful in selected cases.While liver function tests are commonly used to monitor liver function during ATT, liver biopsy can be a valuable tool to establish the differential diagnosis, assess the severity and guide the management of DILI [22].
Traditionally, histology has been considered the gold standard in the diagnosis of chronic liver disease.Furthermore, specific liver diseases, including DILI, have been diagnosed with the help of laboratory tests and, possibly, histological data of the liver.Based solely on laboratory tests such as ALT and ALP, DILI and other forms of liver injury are now classified as hepatocellular, cholestatic or mixed liver injury.However, there remain rare cases of diagnostic uncertainty regarding alternative causes in which liver biopsy should be considered as the final diagnostic approach, provided that the patient benefits from this procedure.
When faced with clinical suspicion of DILI, the indications for performing a biopsy could be the presence of multiple possible etiologies; information on the potential mechanism of damage; assessment of the severity of the injury to enable a clinical decision; unrecognized liver disease diagnosis; and exclusion of autoimmune hepatitis [23].
As regards TB-DILI, the criteria based on ALP may miss the diagnosis of hepatic localization of the disease, as we saw in the case of the patient with granulomatous hepatitis [24].
In our study population, we decided to perform liver biopsy in ten selected patients due to the persistence of the damage despite the suspension of drugs.Although the role of liver histology in the management of DILI and TB-DILI is still uncertain, we do believe that it has been helpful.After reviewing the liver histology, we could exclude DILI in 5/10 patients.
In one patient, we found necrotizing granulomatous inflammation with positive ZN.Hepatic TB lacks typical clinical symptoms and imaging diagnosis so can easily be misdiagnosed, which may result in the interruption of treatment.In this case, a biopsy provided additional diagnostic data.Two patients (both RR/MDR) in this case series had chronic HCV infection (Table 1) [25].Hepatitis C virus (HCV) chronic infection is more frequent in people affected by TB than in the general population, and it is a well-known concurrent risk factor for the development of TB-DILI [26].
The pharmaceutic armamentarium for MDR-TB is limited and includes drugs with potential liver toxicity.Reducing the cumulative risk of hepatic side effects in these patients by treating HCV chronic infection is crucial to completing a long and challenging therapeutic regimen.Recent data from a study by Tunesi and colleagues show that concomitant treatment for HCV and MDR/RR-TB was effective and well-tolerated with sustained virologic response and fewer common adverse events in MDR TB patients [27].We observed this for two patients in this case series as well.
In one patient, the LB showed diffuse steatosis and mixed inflammatory infiltrate with foci of perivenular necrosis with hypertrophic Kupffer cells containing PASD+ macrophages, consistent with a diagnosis of nonalcoholic fatty liver disease (NAFLD).DILI in the context of NAFLD is a growing concern for practitioners.There are relatively few studies on the influence of NAFLD on the incidence of ATT-DILI.Many authors have reported a high incidence of ATT DILI in patients with pulmonary tuberculosis complicated by fatty liver [28] because ATT drugs can also exacerbate pre-existing NAFLD and accelerate the transformation from steatosis to steatohepatitis [29].
According to the Drug-Induced Liver Injury Network (DILIN), five distinct patterns of liver injury can be recognized: acute hepatitis, chronic hepatitis, acute cholestatic, chronic cholestatic and cholestatic hepatitis [7].Gourishankar et al. reported moderate-to-severe and lobular lymphocyte and plasmocytic portal inflammation, prominent ductal proliferation, moderate cholestasis (predominantly hepatocellular and canalicular), hepatocellular damage and stage 3 bridging fibrosis in a young patient treated with isoniazid [30].Consistent with these data, all five ATT-DILI patients in this case series exhibited hepatocellular injury [31].
A proportion of patients with drug-induced liver injury (DILI) present with autoantibodies, leading to the current concept of autoimmune-like DILI [32].In our case series, we did not find any patients with this diagnosis.
Given the high diagnostic value reported for the patients described in this case series, we speculate that liver biopsy and histologic examination of the liver are likely to be beneficial in identifying specific liver diseases, as well as in ATT DILI.These can aid in confirming the diagnosis, evaluating disease severity and guiding treatment decisions.It is, however, essential to recognize that liver biopsy carries inherent risks, including bleeding and sampling variability.Therefore, its indication should be balanced with clinical judgment and consideration of individual patient factors.Noninvasive methods, such as imaging and biomarkers, continue to evolve and may provide complementary information, further reducing the need for liver biopsy in the future.
The results of this study should be considered in light of several limitations, since it was a retrospective, single-center case series report study; hence, it is not representative of the whole target population.

Conclusions
Severe and persistent elevations in hepatic transaminases and/or hepatic cholestasis despite discontinuation of therapy and suspicion of other liver damage conditions are indications for liver biopsy, which remains a valuable tool in the evaluation of selected patients with TB-DILI for many reasons.First, it provides more information about the state of the liver than any other single test.Secondly, the analysis of the type of lesion provides a pathological differential diagnosis that can be decisive for the differential diagnosis.Third, it could help establish a diagnosis of extrapulmonary tuberculosis.However, the decision to perform a liver biopsy should be based on clinical judgment, considering the benefits and risks of the procedure.Careful monitoring of liver function tests and timely recognition of liver damage can help prevent serious liver damage and improve patient

Figure 1 .
Figure 1.Histological features of DILI.(A) Interface hepatitis characterized by portal inflammatory infiltrate extending into the adjacent parenchyma; (B) higher magnification shows the presence of occasional eosinophils; (C,D) enlarged sinusoidal Kupffer cells (C) and clusters of macrophages revealed via histochemical staining with PAS-Diastase (D); (E) lobular confluent necrosis affecting groups of hepatocytes; and (F) micro-and macro-vesicular steatosis.

Figure 1 .
Figure 1.Histological features of DILI.(A) Interface hepatitis characterized by portal inflammatory extending into the adjacent parenchyma; (B) higher magnification shows the presence of occasional eosinophils; (C,D) enlarged sinusoidal Kupffer cells (C) and clusters of macrophages revealed via histochemical staining with PAS-Diastase (D); (E) lobular confluent necrosis affecting groups of hepatocytes; and (F) micro-and macro-vesicular steatosis.