AIDS | 2021
Predictive variables for peripheral neuropathy in treated HIV-1 infection revealed by machine learning.
Abstract
OBJECTIVE\nPeripheral neuropathies in HIV-infected patients are highly debilitating because of neuropathic pain and physical disabilities. We defined prevalence and associated predictive variables for peripheral neuropathy subtypes in a cohort of persons living with HIV (PWH).\n\n\nDESIGN\nAdult PWH in clinical care were recruited to a longitudinal study examining neurological complications.\n\n\nMETHODS\nEach subject was assessed for symptoms and signs of peripheral neuropathy and demographic, laboratory and clinical variables. Univariate, multiple logistic regression and machine learning analyses were performed by comparing patients with and without peripheral neuropathy.\n\n\nRESULTS\nThree patient groups were identified: those with peripheral neuropathies (PNP, n\u200a=\u200a111) that included HIV-associated distal sensory polyneuropathy (DSP, n\u200a=\u200a90) or mononeuropathy (MNP, n\u200a=\u200a21), and those without neuropathy (NNP, n\u200a=\u200a408). Univariate analyses showed multiple variables differed significantly between the NNP and PNP groups including age, estimated HIV-1 duration, education, employment, neuropathic pain, peak viral load, polypharmacy, diabetes, cardiovascular disorders, AIDS, and prior neurotoxic nucleoside antiretroviral drug exposure. Classification algorithms distinguished those with PNP, all with area under the receiver operating characteristic curve (AUROC) values of >0.80. Random forest models showed greater accuracy and AUROC values compared with the multiple logistic regression analysis. Relative importance plots showed that the foremost predictive variables of PNP were HIV-1 duration, peak plasma viral load, age, and low CD4 T-cell levels.\n\n\nCONCLUSIONS\nPNP in HIV-1 infection remains common affecting 21.4% of patients in care. Machine learning models uncovered variables related to PNP that were undetected by conventional analyses, emphasizing the importance of statistical algorithmic approaches to understanding complex neurological syndromes.