Optimizing the Safety of Multidrug-resistant Tuberculosis Therapy in Namibia

Abstract

Introduction: Multidrug-resistant tuberculosis (MDR-TB), a growing global menace, is seriously undermining the previous successes made in the elimination of TB. MDR-TB treatment takes a long time, is complex, and is frequently associated with the occurrence of adverse drug reactions, some of which could severely diminish a person’s health-related quality of life (HRQoL). Research objectives were to determine the occurrence, risk factors and clinical management of adverse events (AEs) associated with MDR-TB treatment; to examine the epidemiology of serious AEs of aminoglycosides in the presence or absence of HIV infection - with or without antiretroviral therapy (ART); and to assess the link between the occurrence of AEs and patients’ perception of their HRQoL at the end of MDR-TB treatment. Methods: This study mainly used observational epidemiologic methods, based on retrospective data collected from medical records of patients treated for MDR-TB in Namibia. Electronic and manual MDR-TB treatment records, audiometry and laboratory test results, were used. Data on patients’ demographic variables, medication exposure, HIV status, presence or absence of AEs were extracted. This was complemented with the WHO global pharmacovigilance data in VigiBase® and patient interviews on their HRQoL perceptions at the end of MDR-TB treatment. Various descriptive and analytic methods were used to summarize the data and to assess the association between drug exposure, covariates and AE occurrence. Results: Adverse events of varying severity were predominant during the intensive phase of MDR-TB treatment. These included gastrointestinal tract AEs (64%), tinnitus (45%), joint pain (28%) and decreased hearing (25%). About 10% of patients experienced serious AEs that were potentially debilitating. HIV co-infected patients experienced more moderate-to-severe AEs compared to the HIV uninfected patients (median 3 AEs versus 1 AE, p=0.01). The cumulative hazard ratio (HR) of moderate-to-severe AEs relative to HIV uninfected patients was HR=4.0; 1.5 – 10.5. Moderate-to-severe AEs were the main determinant of a clinician’s decision to reduce the dose or to stop the suspected offending medicine (RR=3.8; 1.2-11.8). Patients tended to rate moderately low their HRQoL upon completing MDR-TB treatment. The occurrence of AEs was unrelated to HRQoL scores. Amikacin had a higher risk of hearing loss compared to kanamycin (adjusted odds ratio (OR) = 4.0; 1.5–10.8). Patients co-infected with HIV (OR = 3.4; 1.1–10.6), males (OR = 4.5; 1.5–13.4) and those with lower baseline body weight (40–59 kg, OR = 2.8; 1.1–6.8) were most at risk. There was an increased risk of renal insufficiency among patients who concomitantly received kanamycin plus ART as compared to those who received kanamycin alone (HR=3.5; 1.4-8.2).