Tuberculosis (TB) associated with diabetes mellitus (DM) is a growing problem, particularly in low- and medium-resource countries. We conducted an open-label, parallel-group, randomized, and controlled trial in a tertiary care center in Mexico City to assess TB preventive treatment (TPT) with isoniazid (INH) or rifampicin (RIF) in people with type 2 DM. Participants were assigned six months of INH 300 mg/day plus pyridoxine 75 mg or three months of RIF 600 mg/day. The primary outcomes were adverse events resulting in permanent treatment cessation and considered possibly or probably related to study drugs. We included 130 subjects, 68 randomized to INH and 62 to RIF. We prematurely halted the study based on recommendations of the Adverse Event Safety Panel. There was no difference between arms in the overall frequency of adverse events. However, the INH group had significantly more permanent treatment interruptions due to grade 2 recurrent or grade 3 or 4 hepatoxicity. In comparison, the RIF arm had more treatment interruptions due to grade 3 or 4 gastrointestinal intolerance. TPT using INH or RIF is not safe enough to be considered a universal indication to patients with type 2 DM and TB infection. These results underline the need to search for alternative TB preventions with better safety profiles for type 2 DM patients.
Keywords: Mexico; adverse events; diabetes mellitus; isoniazid; latent tuberculosis infection; randomized controlled trial; rifampicin; tuberculosis disease; tuberculosis infection; tuberculosis preventive treatment.