Comparison of three linezolid management strategies for peripheral neuropathy in multidrug- or rifampicin-resistant tuberculosis treatment: a target trial emulation.

Abstract

Peripheral neuropathy frequently leads to linezolid dose reductions or interruptions during multidrug- or rifampicin-resistant tuberculosis (MDR/RR-TB) treatment. The implications of these modifications on treatment success are uncertain. We conducted a target trial emulation using the endTB Observational Study among individuals who developed non-severe peripheral neuropathy while receiving linezolid 600 mg daily within 6 months of initiating an individualized MDR/RR-TB regimen. We examined three linezolid management strategies: immediate change (i.e., dose reduction, temporary interruption, discontinuation) within Weeks 1-7 after peripheral neuropathy onset, deferred change within Weeks 8-26, and no change (i.e., continuing linezolid 600 mg daily) during Weeks 1-26. To emulate the per-protocol analysis of a trial, we used an approach involving cloning and censoring participant data, and applying inverse probability of censoring weights. Among 303 eligible participants from 12 countries, peripheral neuropathy occurred a median of 11 weeks (interquartile range: 4-18) after treatment initiation. Weighted, standardized probabilities of treatment success were 84.7% (95% CI: 69.2%, 92.9%) for immediate change, 78.9% (95% CI: 65.9%, 87.1%) for deferred change, and 85.2% (95% CI: 80.5%, 89.1%) for no change. Compared with no change, treatment success ratios were 0.99 (95% CI: 0.83, 1.11) for immediate change and 0.93 (95% CI: 0.78, 1.01) for deferred change. We did not find evidence of a substantial negative impact of immediate modification to linezolid on MDR/RR-TB treatment success. Our results support the clinical practice of cautiously adjusting linezolid when needed to manage non-severe peripheral neuropathy.