Preexisting metabolic dysregulation in HIV predicts severe COVID-19, long COVID

The higher risk for severe COVID-19 infection and progression to long COVID observed among patients with HIV infection may be attributable to preexisting metabolic dysregulation, particularly increased tryptophan catabolism, according to results of a study published in The Journal of Infectious Diseases.

The severity of COVID-19 disease and postacute sequelae of SARS-CoV-2 have been linked to metabolic dysregulation. However, it remains unclear whether HIV-specific metabolic characteristics can predict severe COVID-19 outcomes.

Researchers at Adaptive Phage Therapeutics and the Wistar Institute sourced data for this study from the Multicenter AIDS Cohort Study and Women’s Interagency HIV Study combined cohort. The analysis included a subset of patients (N=177) with self-reported COVID-19 infection who had provided plasma samples prior to the pandemic (between November 2017 and October 2019). The researchers aimed to evaluate the relationship between preexisting metabolic dysregulation and severe COVID-19 outcomes on the basis of HIV status.

The final analysis comprised 148 (83.6%) HIV-positive and 29 (16.4%) HIV-negative patients. Of patients with HIV infection, all were on antiretroviral therapy and most (78.4%) were virologically suppressed (HIV viral load, 40 copies/mL). In the overall population, 141 patients reported mild acute COVID-19 and 36 were hospitalized for severe disease.

A total of 106 patients completed a long COVID assessment, 23% of whom met criteria for the disease. Patients with long COVID most commonly reported muscle aches or weakness (67%), headache (67%), fatigue (58%), and brain fog (54%).

Plasma samples collected from HIV-positive patients prior to the pandemic revealed significant differences in 23 metabolite levels and 2 metabolite level ratios between those who later required hospitalization for COVID-19 and those who did not (all P <05). Of the altered metabolite levels and ratios, all but 5 were elevated among patients who were required hospitalization for severe disease. Several of the affected metabolites were downstream products of increased tryptophan catabolism.

In a metabolic enrichment analysis, increased risk for severe COVID-19 was significantly associated with altered amino acid/aminoacyl-tRNA biosynthesis (P =.000489), nicotinamide (P =.00653), and tryptophan (P =.0335) metabolism.

Using a Synthetic Minority Over-sampling Technique approach to account for cohort imbalance and a machine learning algorithm, the researchers developed a predictive model to assess hospitalization risk due to severe COVID-19 based in on pre-pandemic metabolic dysregulation. The model performed well overall, with a sensitivity of 96.97%, specificity of 89.57%, accuracy of 91.22%, and area under the curve of 99%.

Further analysis showed significant differences in the plasma levels of 7 metabolites between HIV-positive patients who developed long COVID and those who recovered after acute infection. Patients with long COVID exhibited increased levels of 2-hydroxyvaleric acid (P =.0068), xylitol (P =.012), hydroxyisocaproic acid (P =.0274), and 7-methylguanine (P =.0439) and decreased levels of 10-hydroxydecanoic acid (P =.008), l-asparagine (P =.0219), and abscisic acid (P =.0221). The researchers also noted an elevated ratio of quinolinic acid to tryptophan in patients who developed long COVID (P =.0221), indicating increased tryptophan catabolism.

Study limitations include the exploratory nature of the analysis, the relatively small sample size, and the lack of data on clinical and demographic factors.

According to the researchers, “These findings underscore the necessity for additional research across diverse populations to better understand the impact of metabolic health on infectious disease outcomes.”

Disclosure: One study author declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.

By Jessica Nye, PhD

References:

Agrawal P, Giron LB, Singh S, et al. Pre-pandemic metabolic correlates of COVID-19 severity and long COVID incidence in people living with HIV. J Infect Dis. Published online July 16, 2024. doi:10.1093/infdis/jiae362