Comparison of HIV-1 A6 dispersal dynamics in Poland before and after the war in Ukraine

Abstract

The war-related migrations from Ukraine to Poland have resulted in an increased prevalence of individuals diagnosed with human immunodeficiency virus type 1 (HIV-1) A6 variant. We examined the impact of the influx of people living with HIV (PLWH) displaced from Ukraine on the emergence of transmission events and evolving patterns in the A6 epidemic in Poland. We created a dataset of 13,696 unique HIV-1 pol gene fragments of sub-subtype A6 including 1,889 sequences from Poland. To evaluate the import of distinct clusters and estimate dispersal dynamics, we performed time calibration of the maximum-likelihood phylogenetic trees and phylogeographic inferences using the software package BEAST with discrete and continuous diffusion models. Our results indicated that A6 infections among males predominated within the domestic population (76.1%, n = 1,437), primarily within large clusters. Among Ukrainian migrants, 69.5% of the cases occurred as singletons or dyads (n = 473; p < 0.0001) with a balanced male-to-female ratio of 1.1. Since the war, the contribution of HIV-acquired individuals born in Ukraine to the virus circulation in Poland has increased to 30.2%, with an additional 334 distinct A6 introductions, inferred as internal nodes and descendant clusters that likely entered Poland from other countries. These migration events were concentrated in the central regions with a higher HIV prevalence. After the war outbreak in 2022, the number and complexity of A6 transmission chains in Poland expanded, driven by male-dominated domestic clusters and war-related migration. Understanding the existence of two distinct transmission dynamics is critical for designing targeted public health interventions. Halting national sub-subtype A6 circulation requires a combined approach that harmonizes the existing strategy focused on the men who have sex with men population with enhanced efforts to link migrants to care.

Fig 1. Alluvial diagram shows from left to right the HIV-1 A6 sequence distribution among people linked to care in Poland by country of origin, gender, and risk factor; PL – Poland/Polish; UA – Ukraine/Ukrainian; MSM – men who have sex with men; HET – heterosexual individuals; PWID – people who inject drugs.

Fig 2. Characteristics of HIV-1 A6 sequence distribution among identified cluster sizes.

(A) Sampling density by cluster size and country of origin. Cluster categories are defined as Large Clusters (n > 14), Networks (3 ≤ n ≤ 13), Singletons (n = 1), and Dyads (n = 2) – S&D; (B) Donut charts showing gender and country of birth distribution for various cluster sizes; HET – heterosexual individuals; MSM - men who have sex with men; PWID - people who inject drugs.

Fig 3. Discrete phylogeographic reconstruction of A6 lineage migration events between regions and national groups (PLWH diagnosed in Poland born in Poland or Ukraine) over time with a bar graph depicting the relative contribution of population groups to the spread of the A6 sub-subtype: (A) all migration events from 1996 to 2023, (B) between 2017 and 2022, (C) after 2022.

Arrow thickness corresponds to the average number of inferred migration events between regions, while transparent circle sizes represent lineage dispersal events inferred within regions. Calculations were averaged over posterior trees sampled from each posterior distribution. Inward movements between specific groups (PLWH born in PL - Locals and PLWH born in UA - Migrants) are depicted in the same color. Only migration events with an adjusted Bayes factor support ≥3 are reported. Abbreviations: HIV, human immunodeficiency virus; PL, Poland/Polish; UA, Ukraine/Ukrainian. Detailed information is found in Table C in S1 Text. Base layer source: GADM database of Global Administrative Areas, version 4.0, available at https://gadm.org.

Fig 4. Continuous phylogeographic reconstruction of migration events between regions in Poland over time: (A) all migration events from 1996 to 2023, (B) between 2017 and 2022, (C) after 2022.

We depict the maximum clade credibility (MCC) tree and the 80% highest posterior density (HPD) regions to illustrate the uncertainty associated with the Bayesian inference. The nodes of the MCC tree are color-coded according to their temporal occurrence, and the 80% HPD regions were calculated for successive time intervals and overlaid using a consistent color scale to reflect time. Base layer source: GADM database of Global Administrative Areas, version 4.0, available at https://gadm.org.

Fig 5. Investigation of spatio-temporal distribution of singletons and dyads.

Phylogenetic clusters identified from the maximum clade credibility (MCC) tree, including n = 1 (singletons) and n = 2 (dyads) sequences, are randomly placed on the map, maintaining regional specificity. Points represent individual sequences colored by origin: Local (born in Poland - blue) and Migrant (born in Ukraine - red). Lines represent phylogenetic branches between dyads. The points and lines in this figure do not represent migration events. Instead, they illustrate the spatial placement of singletons and dyads based on their sampling region. (A) Singletons and dyads before 2022; (B) singletons and dyads after 2022; (C) distribution of singletons and dyads across two population groups, including dyad linkage composition and intra- or inter-regional dispersal before and after 2022; Base layer source: GADM database of Global Administrative Areas, version 4.0, available at https://gadm.org.