Profiling adverse reactions to novel drug-resistant tuberculosis regimens containing bedaquiline and linezolid

INTRODUCTION

Tuberculosis (TB) is a communicable disease that is a leading cause of global ill health, with significant morbidity and mortality. According to the WHO, it is among the top 10 causes of death and ranks second among the infectious killers after COVID-19 and above HIV/AIDS. In 2022, around 10.6 million people were estimated to have fallen ill with TB worldwide, and around 1.6 million died. The demographic distribution of TB shows its predominance in areas with poverty, malnutrition, overcrowding, and co-infection with HIV. Approximately 86% of new TB cases are reported from the region of Southeast Asia, Africa, and western Pacific.¹

Some of the strains of Mycobacterium tuberculosis are resistant to the anti-TB medications and require various second-line treatment courses under the National Tuberculosis Elimination Programme (NTEP). Various nomenclatures for drug resistance have been defined. Resistance may be to a single-drug or to multiple-drugs and is classified accordingly. Multidrug-resistant TB (MDR-TB) is defined as resistance to rifampicin and isoniazid, while MDR-TB strains with additional resistance to fluoroquinolones and any other group ‘A’ drugs (bedaquiline and linezolid) are defined as extensive drug-resistant TB (XDR-TB).^2,3

Out of all the TB patients, globally in 2021, 161,746 people were enrolled on treatment for MDR/rifampicin-resistant TB (RR-TB), a small increase of 7.5% from 150,469 in 2020, but still below 181,533 in 2019. Global data suggest MDR was seen in 78% of the patients having RR-TB. In a survey by the Indian government conducted between 2014-2016, the estimated incidence of MDR-TB was 2.84% in new cases, while it was 11.6% among previously treated patients. Despite the evidence that TB is slowly declining, the worldwide emergence of MDR and extensive drug resistance in MTB strains and their rapid spread have become a major obstacle to TB control.

MDR-TB treatment includes the use of second-line anti-TB drugs, which are to be administered for a longer duration. The drugs in these treatment regimens are often associated with significant side effects, which can complicate adherence and lead to treatment interruptions. The purpose of this study was to document the incidence of adverse drug reactions (ADRs) that occur during the course of DR-TB treatment. Early and accurate identification of ADRs allows for timely adjustments in therapy, minimizing the risk of severe complications and improving overall treatment efficacy.

MATERIAL AND METHODS

RESULTS

In 2 years, 145 patients with DR-TB were registered at our Nodal DR-TB center, out of which 79 patients reported ADRs during prospective follow-up. The ADR incidence was 54.4% (79/145 patients). The 79 patients developing ADRs constituted the study population for detailed analysis [Flowchart].

Close

Flowchart: Flow diagram showing selection of study population. MDR: Multidrug-resistant: TB: Tuberculosis.

Export to PPT

Adverse drug reaction profile

The incidence of ADR was 54.4% (79/145 patients). A total of 121 ADRs were reported among the 79 patients (some patients experienced multiple ADRs).

Peripheral neuropathy was the most frequent ADR (41.3% of all ADRs i.e 34.4% of total cohort), followed by anemia (20.6%), nausea/vomiting (12.3%), skin discoloration (9.0%), joint pain (4.9%), pruritus (3.3%), pain in heels (3.3%), hepatotoxicity (1.6%), irrelevant talks (1.6%), decreased appetite (0.8%), and hypothyroidism (0.8%) [Figure 1].

Close

Figure 1: Pie chart showing frequency and distribution of adverse drug reactions among drug-resistant-TB patients.

Export to PPT

Regimen-wise ADR distribution

All-oral longer regimen

Out of 121 ADRs, 84 (69.4%) were reported by patients on an all-oral longer regimen. Peripheral neuropathy affected 46 (54.7%) patients, anemia was seen in 25 (31%) patients, and skin discoloration and tendinitis were reported in 4 (4.7%) patients each [Figure 2].

Close

Figure 2: Pie chart showing frequency and distribution of ADR in DR-TB patients on all oral longer regimens. DR-TB: Drug-resistant tuberculosis, ADR: Adverse drug reaction.

Export to PPT

Shorter course regimen

37 (30.5%) ADRs were reported in patients on the shorter course regimen. Nausea and vomiting (seen in 14, 37.8% of these patients) was the most common ADR, followed by skin discoloration (7, 18.9%) and joint pain (5, 13.5%) [Figure 3].

Close

Figure 3: Pie chart showing frequency and distribution of ADR in DR-TB patients on a shorter course regimen. DR-TB: Drug-resistant tuberculosis, ADR: Adverse drug reaction.

Export to PPT

DISCUSSION

MDR-TB treatment has progressively evolved with more effective and potent drugs being added according to the sensitivity of the bacteria. The treatment involves a combination of potent second-line drugs, each carrying the risk for various ADRs. These second-line drugs can cause significant side effects such as nephrotoxicity, neurotoxicity, haematological toxicity, hepatotoxicity, etc., increasing the morbidity of patients. These ADRs may lead to discontinuation of treatment in many patients. Therefore, it is necessary to document and address ADRs through dose modification or regimen changes. Vigilant ADR monitoring ensures early detection and management of these adverse effects, which helps prevent severe complications and treatment interruptions, and ensures patient safety.

In our prospectively followed cohort, 79/145 patients (54.4%) encountered ADRs. Our finding lies within the wide range reported by different studies (29-94%). Similar incidence rates were reported by Patel et al.³ (2015) (54.9%) and Joseph et al.⁴ (2011) (58%) from India. Higher incidence was reported by Rupani et al.⁵(2020) (73%) and Bhatt et al.⁶ (2012) (93.8%), while lower incidence was reported by Singh et al.⁷ (2019) (28.8%). International studies showed higher rates: Zhang et al.⁸ (2017) reported 90.7%, and Ngoc et al.⁹ (2021) reported 71.3%. Variations in ADR rates may be attributed to differences in patient factors, literacy rates, cultural differences, treatment adherence, demographic changes, default rates, nutritional factors, physician factors, regimen type, and ADR reporting practices, etc.

The mean age group of the patients was 38.5 ± 3.3 years. This validates the argument that MDR is more commonly seen among the mid age group as compared to the older age group patients. This may be due to exposure to drug resistance in the past 2 decades. In the study by Datta et al.,¹⁰ (2009) mean age of the patients was 39 ± 4.7 years, while it was 33.6 ± 11.0 years in the study by Bhatt et al.⁶ (2012) International studies report a similar incidence rate. The mean age was 39.6 ± 13.1 years in the study by Buziashivili et al.¹¹ (2021) and 38.8 ± 11.8 years in the study by Kushemererwa et al.¹² (2023) The median age of patients was 36 years in the study by Huerga et al.¹³ (2022) The results are consistent with the epidemiology of MDR as described in most literature.

In our study, there were 53 (67.0%) males and 26 (32.9%) females, in a 2:1 ratio. In studies by Patel et al.,³ (2015) Piparva et al.,¹⁴ (2018) and Singh et al.,⁷ (2019) the male-to-female ratio was also 2:1. Borisov et al.,¹⁵ (2019) in a multi-center study, reported a similar male-to-female ratio of 2:1. This shows that the occurrence of MDR is higher in males as compared to females.

In this study, 62 (78.4%) patients were underweight (BMI <18.5), 15 (18.9%) patients had normal weight (BMI 18.5-24.9), and 2 (2.7%) patients were overweight. There was no statistically significant association between BMI and the occurrence of ADR (p = 0.089). Merid et al.¹⁶ (2019) in Ethiopia reported a similar distribution with 74.8% patients being underweight. Most of the Indian studies show a similar distribution. In Piparva et al.’s¹⁴ (2018) study, 69.4% patients belonged to the underweight category. In another study by Mukherjee et al.,¹⁷ (2015) 59.8% patients were in the undernourished category. Kushemererwa et al.¹² (2023) study from Uganda reported 40.4% of the patients were underweight. Undernutrition is highly prevalent among TB patients due to the disease’s impact on metabolism and appetite, which exacerbates nutritional deficiencies. Additionally, chronic illness often leads to weight loss and muscle wasting, further compounding undernutrition.

Other comorbidities seen in our patients included: 14 (17.7%) had hypertension, 13 (16.4%) had diabetes, 7 (8%) had coronary artery disease, and 1 (1.2%) had hypothyroidism. In the study by Huerga et al.,¹³ (2022) diabetes was seen in 16.1% of patients. A study by Sant´ Anna et al.¹⁸ (2023) reported 9.0% of patients with diabetes and 14.2% with hypertension. In the study by Buziashvili et al.,¹¹ (2021) 8.5% patients had diabetes. No study showed a direct correlation of ADR with any comorbidity. These findings show that diabetes, hypertension, and other comorbidities do not play a direct role in the development of ADRs among DR-TB patients.

In the present study, around 59 (74.6%) patients had a previous history of ATT, while 20 (25.3%) had no history of ATT. Other studies reported a slightly higher incidence of previous history of ATT treatment. In a study by Piparva et al.,¹⁴ (2018) all patients with MDR-TB were retreatment cases. In the study by Huerga et al.,¹³ (2022) 92% of the patients with MDR were previously treated for TB, while 8% were new patients. Paikray et al.¹⁹ (2022) in Cuttack reported 89.7% of the sample population with a history of ATT. Similarly, in the study by Ngoc et al.,⁹ (2021) 84.8% had a previous history of ATT, while 7.6% had no history of ATT treatment, and the status was unknown for the rest 7.6%. In most of the Indian studies, the rate is >90%.²⁰ In a study by Kushemererwa et al.¹² (2023) in Uganda, 53.9% patients had a previous history of ATT treatment. These findings show that a significant proportion of patients in all studies have a high incidence of ATT intake in the past, as previously treated patients themselves are at a higher risk of developing DR-TB. Still, none of the studies, including ours, showed a correlation between the occurrence of ADR and previous history of ATT.

In the present study, out of the total patients with and without ADR, 9 patients were HIV positive. Out of these 9 patients, 7 developed ADR. Ngoc et al.⁹ (2021) reported a similar prevalence of 8.6% for HIV in their study. Borisov et al.¹⁵ (2019) reported it to be 4.4%. Other studies reported a higher prevalence of HIV, with Huerga et al.¹³ (2022) reporting the prevalence at 15.5%, Merid et al.¹⁶ (2019) at 27%, and Sant´ Anna et al.¹⁸ (2023) at 35.1%. Prasad et al.²¹ (2021)in their review article found that patients with HIV experience more frequent ADRs to both anti-TB and non-TB medications for other opportunistic infections. They noted that the typical strategy of stopping all medications and rechallenging them one by one is not possible in these patients, as the risk of emergence of resistance, especially for antiretroviral therapy, is very high. Lazarus et al.²² (2021) in their study also concluded that HIV infection was an independent factor that increased the risk of developing adverse events in DR-TB patients by 12%. Sant´ Anna et al.¹⁸ (2023) reported that the incidence of ADR among HIV positive patients with DR-TB was 78.6%. Weakened immune system, also the intake of multiple drugs leading to drug-drug interactions, could be the possible reasons behind high incidence of ADR among PLHIV with DR-TB.

Peripheral neuropathy (50 patients, 34.4%; 41.3% of ADR patients) was the most common ADR. The most frequent adverse event in our study, peripheral neuropathy, has also been found to be the most frequent ADR in other studies like those conducted by Huerga et al.,¹³ (2022) Mishra et al.,²³ (2024) and Desai et al.²⁴ (2022) The most common drugs attributed to cause peripheral neuropathy include linezolid, isoniazid, fluoroquinolones, cycloserine, ethambutol, and ethionamide.²⁵ Linezolid was implicated in 46 cases (92%) in our study, while isoniazid was considered the cause for four patients. Linezolid causes neuropathy through impaired mitochondrial protein synthesis.²⁶ Patients reported tingling, numbness, burning sensations, inability to feel footwear, and coordination problems. Incidence rate of peripheral neuropathy was similar to the results of a study by Zhang et al.⁸ (2017) who reported an incidence rate of 31% for peripheral neuropathy in a meta-analysis combining 15 studies, including those held in countries like India, China, etc. Singla et al.²⁷ (2012) also reported a similar incidence rate of 30%. Huerga et al.¹³ (2022) also reported the incidence of peripheral neuropathy as 28.5%. Khanam et al.²⁸ (2018) found peripheral neuropathy in 28% patients. Lifan et al.²⁹ (2019) in a meta-analysis found the incidence of peripheral neuropathy to be 26.0%. In other studies, 25% in Madhav et al.,³⁰ (2015) 18.7% in Tiwari et al.,³¹ (2015) and 13% in Shin et al.³² (2003) Some studies reported a higher incidence of peripheral neuropathy i.e., Mishra et al.²³ (2024) reported it to be 93.3%, and Von der lippe et al.³³ (2006) reported it to be 60%.

Anemia incidence was 17.2% (25/145 patients). Linezolid-associated myelosuppression causes anemia and other hematological effects.³⁴ Our findings (17.2%) were comparable to Lee et al.³⁵ (2012) (17%) and Park et al.³⁶ (2006) (12.5%). Lower rates were reported by Schecter et al.³⁷ (2010) 6.6%, Huerga et al.¹³ (2022) 5.1%, and Mishra et al.²³ (2024) Higher rates (27% to 42%) were reported by Lifan et al.²⁹ (2019) 42%, Sotgiu et al.³⁸ (2012) 38%, Migliori et al.³⁹ (2012) 27.0%, and Agyemann⁴⁰ (2016) 32.9%.

In our study, nausea and vomiting were seen in 9.6%. A similar incidence of 10.7% was reported by Massud et al.⁴¹ (2022) Other studies across the world have reported a different range of patients reporting nausea and vomiting, between 20-40%.^{5,7,9,15,18,40,42} The drugs attributed to causing nausea and vomiting included ethionamide, levofloxacin, and bedaquiline. Gastrointestinal ADR`s (nausea, vomiting, and abdominal pain) are more common in patients who are taking shorter MDR regimens, possibly due to the increased number of pills and interactions.

Skin discolouration was also among the common side effects encountered in these patients. Clofazimine was identified as the culprit drug for all 11 cases in our study. Clofazimine (CFZ) is a weakly basic, very lipophilic, phenazine antibiotic. Partitioning of the circulating, free base form of the drug into subcutaneous fat leads to its accumulation, which manifests as skin discoloration.⁴³ In our study, it manifested mainly as a coppery red to brownish discoloration of the skin, mainly in the sun-exposed areas of skin. Facial discoloration was seen in 10 patients, whereas 1 patient showed generalised discoloration of skin mainly over the face, neck, collar area, and arms. The incidence rate of skin discoloration was 7.5% in our study. Piubello et al.⁴⁴ (2014) reported incidence rates of 3.1%, Du et al.⁴⁵ (2020) reported it in 14%, Wang et al.⁴⁶ (2024) in 22.7%, and Dalcolmo et al.⁴⁷ (2017) in 52.1%. These variations are due to some studies being purely based on regimens that include clofazimine. Darker skin pigmentation in western rural Indian regions is a common phenomenon, and this can be one of the reasons, apart from just a cosmetic issue, for less reporting of this ADR, when compared with western studies done in fair skin patients.

Pyrazinamide is converted to pyrazinoic acid, which inhibits the kidneys’ ability to excrete uric acid. This leads to increased uric acid levels in the blood, which can cause monosodium urate crystals to build up in the joints. Serum uric acid was raised in 5 out of 6 patients with athralgia in our study. Arthralgia was reported in 6 (4.1%) of our patients. A similar incidence of athralgia was found in other Indian studies. Rathod et al.⁴⁸ (2015) reported athralgia in 4.1% patients, Singh et al.⁷ (2019) reported in 7.6% patients, whereas Paikray et al.¹⁹(2022) reported athralgia in 10.8%. Other studies have reported arthralgia in a higher number of patients, 12.5% in Nilamsari et al.,⁴² (2021) 14% in Dela et al.,⁴⁹ (2017) 24.7% in Kushemererwa et al.,¹² (2023) 25.5% in Rupani et al.,⁵ (2020) 26% in Sagwa et al.,⁵⁰ (2013) and 34.7% in Ngoc et al.⁹ (2021) The potential anti-TB drugs causing arthralgia include pyrazinamide, bedaquilline, fluroquinolones, streptomycin, and ethambutol.⁵¹ In our patients, joint pain was attributed to pyrazinamide, levofloxacin, bedaquiline, and ethambutol.

This prospectively followed programmatic cohort did not record frequent bedaquiline-related serious toxicities (e.g., QT prolongation). This may be due to: (1) limited bedaquiline uses relative to linezolid; (2) focus on clinically overt events rather than systematic ECG-based research monitoring; (3) possible under-reporting of subclinical QT prolongation. These factors are acknowledged as limitations.

The incidences of other less common side effects in our study were 3.3% for pruritis and tendinitis, 1.6% for hepatotoxicity and psychosis, and 0.8% for decreased appetite and hypothyroidism. Similar side effects have been reported in other studies as well at varied incidence rates. Ngoc et al.⁹ (2021) reported hepatotoxicity in 5.8% patients, and in the same study, the incidence of hypothyroidism was in 1.1% patients. Huerga et al.¹³ (2022) (reported hepatotoxicity in 1.5% patients. Loss of appetite was reported in 12% of the patients in the study by Rupani et al.⁵(2020).

CONCLUSION

This prospective observational study found a high incidence (54.5%) of ADRs among DR-TB patients, particularly peripheral neuropathy (34.48%) and anemia (17.24%), predominantly linked to linezolid exposure. No significant associations were observed between ADR occurrence and demographic or clinical factors (i.e., age, gender, HIV status, BMI, previous ATT). Linezolid emerged as the most toxic drug, requiring vigilant monitoring for dose reduction or discontinuation. ADRs have a direct impact on patients’ adherence to recommended regimens in addition to posing serious risks to the safety and effectiveness of MDR-TB therapies. The findings highlight that the severity and frequency of ADRs can lead to modification of treatment, thereby compromising therapeutic outcomes and exacerbating the risk of treatment failure and resistance. Early detection and management of ADRs through routine pharmacovigilance, patient education, and tailored therapeutic strategies are critical for improving treatment adherence, preventing treatment discontinuation, and achieving better DR-TB treatment outcomes. Larger multi-center prospective cohorts with extended follow-up periods are needed to validate these findings and establish robust risk profiles for ADRs in DR-TB management.