Primary HIV infection responsible for half of all HIV transmission in Quebec

Primary HIV infection accounts for half of all onward transmission, according to rigorous phylogenetic analysis data from Quebec, Canada published in the April ^1st edition of the Journal of Infectious Diseases. An accompanying editorial suggests that whilst identifying and treating early HIV infections are challenging, they are important prevention strategies that have the potential to serve both individual and public health.

A number of recent studies have suggested that early HIV infection may disproportionately contribute to onward transmission. Notably, a 2005 study of monogamous heterosexual couples in Rakai, Uganda – where one partner was HIV-positive and one HIV-negative at the start of the study – found that the rate of sexual HIV transmission within the first two and a half months was almost twelve-times greater than that observed in chronic HIV infection.

Using phylogenetic analysis to study primary transmission clusters

Although there are limits as to what phylogenetic analysis can ‘prove’ in terms of HIV transmission between two individuals, it can be used somewhat more reliably to estimate transmission events within a larger population.

Investigators in Quebec are particularly well placed to undertake phlyogenetic analysis to ascertain whether HIV transmission is likely due to early or chronic HIV infection, due to two factors:

• In 1997, they established a primary HIV infection cohort, which provides longitudinal data on viral evolution, drug resistance, transmission risk factors, and disease progression.
• They have a large database of HIV gene sequences available (with samples from about 20% of all diagnosed individuals in Quebec), due to the routine genotypic testing, since 2001, of recently infected and newly diagnosed individuals, as well as treatment-experienced individuals for whom antiretroviral therapy (ART) has failed.

Consequently, they analysed HIV gene sequences (from the pol region) in 215 participants in their primary HIV infection cohort, as well as an additional 502 recently infected (i.e. within six months of seroconversion) individuals from their genotypic testing database. A total of 593 unique subtype B infections were identified; a further 65 non-subtype B infections (10% of all recent infections) were excluded from subsequent analysis.

The investigators utilised extremely rigorous phylogenetic analysis methods, with very high bootstrap values (98% or greater), and found that half (293/593) of all primary HIV infections were grouped in 75 different transmission clusters. Half of the clusters involved between two and four individuals; the rest were larger clusters (a median of nine individuals, ranging between five and 17 individuals). The average time from the first to last infection within each cluster was 15 months, although this ranged from one month to 37 months, depending on the number of individuals in the cluster.

These sequences were then compared with 135 treatment-naive, and 660 treatment-experienced individuals with chronic HIV infection, and they rarely clustered with the samples from the individuals with primary HIV infection (1% of treatment-naive and 2.7% of treatment-experienced).

Half of all transmission was due to primary/early infection

After analysing all the data, the investigators concluded that whilst primary/early HIV infection represented just 10% of the total sequenced samples in the genotypic resistance database, they accounted for 49% of all onward transmission events. In contrast, they found that treatment-naive and treatment-experienced chronically infected individuals accounted for 15% and 12% of onward transmission, respectively.

The investigators do not account for the remaining quarter of infections, but note that they “doubtless missed some” primary HIV infection in Quebec during the study period. They add that since many people with early HIV infection are undiagnosed “the role of acute infection may be underestimated.”

They do, however, rule out a substantial onward transmission role for successfully treated, chronically infected individuals in Quebec (who are not included in their resistance database, or in this analysis) because of the much lower average viral load (2.58 log₁₀ copies/ml in this population.

Indeed, average plasma viral loads for each of the groups in this analysis were much higher: the highest viral loads (associated with a higher risk of transmission) were seen in recently-infected and treatment-naive chronically infected individuals (4.6 and 4.7 log₁₀ copies/ml). This compares with 4.1 log₁₀ copies/ml in treatment-experienced individuals.

Behaviour and resistance data

After analysing behavioural data from the 215 individuals in the primary HIV infection cohort no significant differences were seen within or between heterosexual or homosexual sex or injecting drug use in terms of small, large or no transmission clusters.

There was also no difference seen between in terms of numbers of sexual partners and small, large or no transmission clusters. The investigators note that “clustering could not be attributed to differences in behavioural risk factors” and that “this suggests that different infections may vary in transmissability, and further study is necessary.”

About 15% of individuals with primary/early HIV infection had resistance to at least one drug class, although NNRTI resistance was much more likely to be seen than NRTI or PI resistance. “It is noteworthy,” the investigators write, “that approximately half of transmitted resistance can be attributed to clustered infections but that transmission of viruses containing mutations associated with resistance to nucleoside or protease inhibitors was diminished in clustered infections, possibly because of reduced viral fitness.”

Diagnose and treat HIV infection earlier for individual and public health

In an accompanying editorial two British doctors, Deenan Pillay and Martin Fisher, argue that “as clinical management of HIV infection improves and HAART suppression of viraemia is maintained for longer periods of time, the source of further transmission will shift more toward untreated (including undiagnosed) individuals.”

They note that there are many barriers to diagnosing HIV infection during its early stages, and that diagnoses are often missed in GP and hospital settings. But they welcome recent moves to normalise HIV testing in the United States, and stress that in order to diagnose early infection appropriate tests such as combined antibody-antigen tests or pooled HIV-RNA testing should be utilised. “However,” they write, “we argue that the current focus on increasing HIV diagnoses through more widespread testing requires a parallel strategy for minimising ongoing transmission.”

They also argue that despite recent advances in prevention technologies, such as circumcision or treating sexually transmitted infections “these strategies may be of limited effectiveness if individuals at or near to [primary HIV infection] represent a major source of onward transmission, because many such interventions depend on an initial positive diagnosis.”

Even though HAART is “expensive, and relatively toxic”, they argue that “recent improvements in drug formulations (thus enhancing adherence) and reductions in HAART toxicities” make it feasible to treat HIV-positive individuals much earlier than current guidelines suggest.

Consequently, although they concede that this is currently a “contentious” concept, they believe that “changing the paradigm of treatment rationale” and “extending treatment to those with higher CD4 cell counts” for the “benefit of individual and community” is a strategy worth studying.

“It is now time to evaluate application of the most potent intervention to treat this disease – namely, antiretroviral therapy – to its prevention,” they conclude. “Furthermore, strategies to improve recognition of recent infection, in addition to identifying undiagnosed chronic infection, are crucial for effective implementation of prevention strategies.”