Assesing Conversion Rate, Treatment Success and Mortality of Drug Induce Liver Injury from First Line Antituberculosis Regiment: A Systematic Review and Meta -Analysis

Introduction

Tuberculosis (TB) remains a critical global health issue, necessitating the use of The primary anti-tuberculosis drugs, including isoniazid, rifampicin, ethambutol, and pyrazinamide, are commonly used in the treatment of tuberculosis. However, these medications are often linked to tuberculosis-related drug-induced liver injury (TB-DILI), which complicates the effective management of TB. The prevalence of TB-DILI varies widely, with studies reporting rates ranging from 2% to 33% among patients undergoing anti-tuberculosis therapy.^1,2 This variability underscores the urgent need for comprehensive data on the outcomes of TB-DILI, as the current literature is limited and often inconsistent in defining the clinical features and risk factors associated with this adverse effect.

Despite the recognition of DILI as a serious complication of anti-tuberculosis treatment, there is a notable scarcity of large-scale, longitudinal studies that systematically assess the incidence, patterns, and outcomes of TB-DILI across diverse populations.² For instance, while some studies have identified specific risk factors such as age, baseline liver enzyme levels, and co-infections, the overall understanding of how these factors interact to influence treatment outcomes remains fragmented.^2,3 Furthermore, the clinical manifestations of TB-DILI can range from mild, asymptomatic elevations in liver enzymes to severe liver failure, complicating the management of TB and potentially leading to treatment interruptions.⁴

The urgency of addressing TB-DILI is compounded by the increasing prevalence of multidrug-resistant (MDR) TB, which necessitates prolonged and more complex treatment regimens that may further elevate the risk of liver injury.⁵ Given the potential for significant morbidity and mortality associated with TB-DILI, it is imperative that future research prioritizes the collection of robust data on this topic, enabling healthcare providers to make informed decisions regarding the management of TB in patients at risk for liver injury.⁶ A concerted effort to enhance the understanding of TB-DILI outcomes is essential for improving treatment adherence, optimizing therapeutic strategies, and ultimately reducing the burden of tuberculosis globally.

Methods

Results

Systematic review and study characteristics

The initial search yielded 717 articles, of which 672 underwent title and abstract screening. Five studies met the eligibility criteria ( Figure 1). After a thorough review, eight studies were fully assessed. Five studies were excluded for the following reasons: absence of a TB or TB-DILI control group, no relevant outcomes, or other factors. Five studies, totaling 5,798 patients receiving first-line TB treatment, were included in the final analysis.^9–12 These studies, conducted across multiple countries, evaluated treatment durations of at least six months. A detailed description of the included studies is provided in Table 1.

Figure 1. The PRISMA flowchart illustrating the search and selection of studies.

Table 1. Features of the studies included in the analysis.

Author [Reference]	Year	Study design	Population	Sample size	Regiment	Outcomes	NOS
Shang13	2011	Prospective Cohort Study	China	4304	RHZES*	Treatment Prolongation, Conversion in 2 months, Treatment Outcomes	8
Sharma14	2002	Prospective Cohort Study	India	346	RHZES*	Liver Function, DNA HLA Analysis, Mortality	6
Shetty15	2024	Retrospective cohort study	India	42	RHZE*	Treatment duration, 2 Months Sputum Conversion, Treatment Succes, Mortality	6
Song9	2019	Retrospective cohort study	South Korea	168	RHZES*	2 Months Sputum Conversion, Treatment Duration, Time to sputum conversion, Treatment Succes, Mortality, 1 year Recurrence	8
Sun10	2016	Prospective Cohort Study	China	938	RHZES*	Time to sputum Conversion, Time to cavity closure, Treatment Succes, Mortality	8

* Isoniazid (H), Rifampicin (R), Pyrazinamide (Z), Ethambutol (E), and Streptomycin (S) administered alternately every other day for 6 to 9 months.

Table 2. Heterohgeneity and Risk of Bias (RoB).

NOS criteria	Studies (year)
	Shang (2011)	Sharma (2002)	Shetty (2024)	Song (2019)	Sun (2016)
A. Selection (up to four stars)
1. How representative the exposed cohort is	★	★	★	★	★
2. How the non-exposed cohort was selected	★	★	★	★	★
3. Method used to determine exposure	★	☆	★	★	★
4. Confirmation that the outcome of interest was absent at the start of the study	★	★	★	★	★
B. Comparability (up to two stars)
1. Degree of comparability between cohorts based on study design or statistical analysis	★☆	☆	☆	★☆	★☆
C. Outcome (up to three stars)
1. Evaluation of outcome	★	★	★	★	★
2. Was the duration of follow-up sufficient for the outcomes to be observed?	★	★	☆	★	★
3. Sufficiency of cohort follow-up	★	★	★	★	★
Total (up to nine stars)	8	6	6	8	8

* NOS criteria fo Heterohgeneity and Risk of Bias (RoB).

Outcomes in TB – DILI patients

A meta-analysis was performed, and the findings are presented in four forest plots: (a) the combined results from all four included studies ( Figure 2), (b) studies assessing 2-month conversion ( Figure 3), (c) treatment success ( Figure 4), and (d) mortality outcomes ( Figure 5). To evaluate the homogeneity or heterogeneity among the included studies, the Chi-squared test and the I² statistic were applied. Figure 2 displays the association measures extracted from the five selected studies. The forest plot showed a relative risk (RR) for TB treatment regimen associated with TB-DILI events, with a pooled RR of 9.53 (95% CI: 3.39–26.82). Significant heterogeneity was observed across the studies, with a Tau² value of 1.34, Chi-squared value of 317.27, degrees of freedom (df ) of 4, I² value of 99%, z-value of 4.27, and a p-value < 0.00001. Since the I² value exceeded 50%, a random-effects model was applied. The combined results from the five studies using this model indicated that TB Non-DILI patients had a higher likelihood of treatment success, with an RR of 1.11 (95% CI: 1.05–1.18, p = 0.0004) ( Figure 4), and a greater risk of mortality, with an RR of 2.62 (95% CI: 1.14–6.03, p = 0.02) ( Figure 5). These findings suggest that TB Non-DILI patients had 3.56 times the odds of achieving treatment success and 2.26 times the risk of mortality compared to TB-DILI patients. In this analysis, the studies by Shetty et al. and Sharma et al.¹² reported no cases of mortality. Despite this, all included studies met the established inclusion criteria, thus they were retained in the analysis. The results revealed no significant difference in 2-month conversion rates between the two groups, with an RR of 1.03 (95% CI: 0.97–1.03, p = 0.31) ( Figure 2).

Figure 2. Forest plot of all studies.

Figure 3. Conversion in 2 months.

Figure 4. Treatment success.

Figure 5. Mortality.

Discussion

This study indicates that TB DILI has no significant risk on 2 months conversion compared to TB Non – DILI patients with RR: 1.03; 95%CI: 0.97–1.03, p<0.31. Treatment success (RR: 1.11; 95%CI: 1.05–1.18, p<0.0004) and mortality risk (RR: 2.62; 95%CI: 1.14–6.03, p<0.02) were significant in TB Non – DILI patients. Theres a significant heterogeneity when all the RR pooled as there were only an observational study that available currently. It has been noted among the study there is a significant gap between the number of TB patients that develop TB - DILI and TB Non-DILI. It was notable that all of the outcomes of the pooled studies have zero heterogeneity. This is the first study that analyse the outcomes of Drug Induce Liver Injury from antituberculosis in TB Patients as far as we notice.

Additionally, the timing of the conversion assessment plays a significant role. The two-month mark is often a critical point in TB treatment where initial responses are evaluated. However, the biological variability in how individuals respond to treatment can lead to similar outcomes despite differences in liver health.¹² Factors such as the extent of the TB disease, the patients overall health status, and adherence to the treatment regimen can all influence conversion rates. The presence of liver injury does not necessarily correlate with the efficacy of the treatment in converting sputum cultures from positive to negative. Studies have shown that the conversion rate can remain similar in both groups due to the robust nature of the immune response elicited by the TB infection itself, which can compensate for the hepatic impairment caused by the drugs.^12,16

Studies indicate that patients with DILI often face a higher risk of treatment interruption or modification, which can adversely affect TB treatment outcomes. In contrast, patients with non-DILI, such as those with existing comorbidities might not experience the same level of treatment disruption, allowing for a more straightforward approach to TB management.¹⁷ However, the prognosis for TB patients with DILI tends to be poorer compared to those with non-drug-induced liver injury. This can be attributed to the compounded effects of liver dysfunction towards metabolism of anti-TB drugs, which can lead to suboptimal drug levels and consequently, treatment failure.¹⁸

Although there is a significant gap of TB DILI events between the studies, mortality rates in TB DILI were lower compared to those with non-DILI can be attributed to several factors that are related to the management and underlying conditions of these patients.¹⁹ The hepatotoxicity of anti-TB drugs often necessitates treatment interruptions or modifications, which can lead to treatment failure and increased mortality risk.²⁰ The nature of DILI often allows for a more reversible condition compared to chronic liver diseases. In cases of DILI, once the offending drug is discontinued, liver function can often improve significantly leading to better overall outcomes. This contrasts with non-drug-induced liver injuries, where there might be an underlying comorbidities such as liver disease may be progressive and irreversible resulting in a higher likelihood of complications such as liver failure which directly correlates with increased mortality²¹

One of the main limitations is the absence of clinical studies that directly compare the different outcomes, as no randomized trials were available. Additionally, since the trials included used a 6-month duration for the first-line regimen, the potential for recurrence and liver-related side effects associated with this treatment regimen yet to be explored. We further advise that the regiment can be attributed to more effective management strategies, the potentially reversible nature of DILI, and the ability to maintain TB treatment without significant interruptions. These factors collectively contribute to improved outcomes in patients with DILI.

Conclusion

Our meta-analysis indicates that there is no substantial risk of delayed conversion after two months in patients with TB-DILI. However, treatment success and mortality reveal a significant risk for both outcomes. The findings of this meta-analysis provide additional evidence reinforcing the clinical outcomes in TB-DILI patients.

Ethics approval

Ethical approval and consent were not required.

Author contributions