A national study of the molecular epidemiology of HIV-1 in Australia 2005-2012

Abstract

Introduction: Rates of new HIV-1 diagnoses are increasing in Australia, with evidence of an increasing proportion of non-B HIV-1 subtypes reflecting a growing impact of migration and travel. The present study aims to define HIV-1 subtype diversity patterns and investigate possible HIV-1 transmission networks within Australia.

Methods: The Australian Molecular Epidemiology Network (AMEN) HIV collaborating sites in Western Australia, South Australia, Victoria, Queensland and western Sydney (New South Wales), provided baseline HIV-1 partial pol sequence, age and gender information for 4,873 patients who had genotypes performed during 2005-2012. HIV-1 phylogenetic analyses utilised MEGA V6, with a stringent classification of transmission pairs or clusters (bootstrap ≥98%, genetic distance ≤1.5% from at least one other sequence in the cluster).

Results: HIV-1 subtype B represented 74.5% of the 4,873 sequences (WA 59%, SA 68.4%, w-Syd 73.8%, Vic 75.6%, Qld 82.1%), with similar proportion of transmission pairs and clusters found in the B and non-B cohorts (23% vs 24.5% of sequences, p = 0.3). Significantly more subtype B clusters were comprised of ≥3 sequences compared with non-B clusters (45.0% vs 24.0%, p = 0.021) and significantly more subtype B pairs and clusters were male-only (88% compared to 53% CRF01_AE and 17% subtype C clusters). Factors associated with being in a cluster of any size included; being sequenced in a more recent time period (p<0.001), being younger (p<0.001), being male (p = 0.023) and having a B subtype (p = 0.02). Being in a larger cluster (>3) was associated with being sequenced in a more recent time period (p = 0.05) and being male (p = 0.008).

Conclusion: This nationwide HIV-1 study of 4,873 patient sequences highlights the increased diversity of HIV-1 subtypes within the Australian epidemic, as well as differences in transmission networks associated with these HIV-1 subtypes. These findings provide epidemiological insights not readily available using standard surveillance methods and can inform the development of effective public health strategies in the current paradigm of HIV prevention in Australia.

Fig 1. The number of new HIV-1 diagnoses by year for each Australian jurisdiction outlined in the Australian molecular epidemiology network; NSW = New South Wales, QLD = Queensland, SA = South Australia, VIC = Victoria and WA = Western Australia (Provided by the Kirby Institute [23]).

Fig 2

Australian HIV-1 subtype distribution for 4,873 sequences for B subtype (aqua) and non-B subtypes (grey) over time [2005–2006 (A), 2007–2008 (B), 2009–2010 (C) and 2011–2012 (D)] including the distribution and proportion of HIV-1 subtypes represented in each Australian state (E) from 2005–2012 (blue = B subtype; green = C subtype; tan = CRF01_AE subtype; maroon = CRF02_AG; yellow = D subtype; red = inter-subtype recombination (ISR); light blue = others).

Fig 3

The phylogenetic tree constructions based on HIV-1 pol sequences for (A) HIV-1 B and (B) non-B subtypes assessed in the AMEN study. Each branch colour represents a state (Western Australia (black), South Australia (red), Victoria (blue), western Sydney (green) and Queensland (maroon).

Fig 4

Differences in the distribution and proportions of pairs and larger cluster size for HIV-1 subtypes; HIV-1 C subtype (A), CRF01_AE subtype (B) and B subtype (C) (excludes singletons).

Fig 5

Gender distribution within clusters according to HIV-1 subtypes (A) and for clusters in each state with B subtype (B) or a non-B subtype (C); W Sydney = Western Sydney, QLD = Queensland, SA = South Australia, VIC = Victoria and WA = Western Australia.

Fig 6

Multivariate regression analysis reveals factors associated with a sequence being classified into (A) a cluster regardless of cluster size or as a singleton or (B) only sequences in larger networks where the sequence size equals 2–3 or ≥4. All results are for >18yo (Coefficient and significance are shown); QLD = Queensland, SA = South Australia, VIC = Victoria and WA = Western Australia.

Fig 7. The proportion of interstate and intrastate HIV-1 paired sequences, sequences in a network of 3 or networks > 3 (n = # of pairs or networks).