Genetic susceptibility to aminoglycoside cochleotoxicity in South African multidrug-resistant and rifampicin-resistant tuberculosis patients.

Participants were recruited by convenience sampling at two hospitals, which may have selected a group that is not typical of everyone on these regimens. 45 of the 147 recruited patients (30.6%) were excluded because they did not complete two valid hearing tests; the reasons given include death, withdrawal, early discharge and being too ill, so those who left the study may have been sicker than the 102 who stayed. The study was sized to detect a 10 dB change in hearing thresholds, not to detect links between rare gene variants and hearing loss, so the genetic results come from a study too small to answer that question. DNA was unavailable or sequencing failed for some participants, shrinking the group available for the genetic analysis further, and with only three and four carriers respectively no adjusted analysis was possible. Well-known aminoglycoside-associated variants such as m.1555A>G and m.1494C>T were not tested, so the independent contribution of the two variants studied cannot be separated from those. HIV status, kidney function, previous MDR/RR-TB treatment, baseline hearing, cumulative aminoglycoside exposure and other ototoxic drugs could not be controlled for, so other explanations for the hearing loss remain open. Because cochleotoxicity was so common and so few participants did not develop it, the study had little ability to tell a genetic susceptibility apart from the general ototoxic effect of kanamycin.

The study was supported by a grant from the National Institute of Allergy and Infectious Diseases of the US National Institutes of Health (R01AI116155). The authors state that the content is solely their responsibility and does not necessarily represent the official views of the National Institutes of Health. No competing interests are declared in the text provided.

Do not read the variants as the cause of these patients' hearing loss. It is true that all the carriers developed cochleotoxicity and none of the participants who did not, but there were only three and four carriers, the comparison groups were tiny, the statistical tests showed no association, and the authors themselves say the findings do not justify routine screening for these two variants.