A systematic phylogenetic approach to study the interaction of HIV-1 with coinfections, non-communicable and opportunistic diseases. Journal of Infectious Disease
Kusejko et al. aimed to systematically test whether coinfections spread along the HIV-1 transmission network and whether similarities in HIV-1 genomes predict AIDS-defining illnesses and comorbidities.
A maximum-likelihood phylogenetic tree was built using sequences of 11’915 patients from the genotypic resistance test database of the Swiss HIV Cohort Study (SHCS) and non-Swiss background sequences from the Los Alamos database.
Among the coinfections, hepatitis C virus (HCV), hepatitis B virus (HBV), syphilis, cytomegalovirus CMV, and latent tuberculosis all clustered significantly on the tree, even after adjusting for risk factors. Several opportunistic diseases such as Kaposi sarcoma clustered significantly on the phylogeny. For most noncommunicable diseases analyzed, patients were likewise not distributed randomly on the phylogeny. In most cases, however, the clustering was not significant in the multivariable analysis. Clustering of patients with psychiatric problems and neurocognitive complaints became weaker in the multivariable mixed effects model, but the odds of having psychiatric problems and neurocognitive complaints if the other patient in the cherry had these problems remained significantly increased.
In conclusion, this work for the first time presents a systematic analysis of interrogating the HIV phylogeny at a population level for the syndemic nature of coinfections and noncommunicable diseases for virus traits potentially relevant for certain diseases. The large variety of conditions tested implies that no universal explanation or interpretation of the clustering can be given. There is evidence for 3 different reasons for clustering: shared transmission routes of pathogens, similar social networks of patients close in the phylogeny, and direct viral genetic impact. Overall, the strategy proposed by the authors together with adjustment for numerous known confounding factors, demonstrates the potential for a new type of analysis, extending conventional epidemiological analyses.