Bias Due to Correlation Between Times-at-Risk for Infection in Epidemiologic Studies Measuring Biological Interactions Between Sexually Transmitted Infections: A Case Study Using Human Papillomavirus Type Interactions

Reconstructing multi-strain pathogen interactions from cross-sectional survey data via statistical network inference

Infectious diseases often involve multiple pathogen species or multiple strains of the same pathogen. As such, knowledge of how different pathogens interact is key to understand and predict the outcome of interventions targeting only a subset of species or strains involved in disease. Population-level data may be useful to infer pathogen strain interactions, but most previously used inference methods only consider uniform interactions between all strains or focus on marginal pairwise interactions. As such, these methods are prone to bias induced by indirect interactions through other strains. Here, we evaluated statistical network inference for reconstructing heterogeneous interactions from cross-sectional surveys detecting joint presence/absence patterns of pathogen strains within hosts. We applied various network models to simulated survey data, representing endemic infection states of multiple pathogen strains with potential interactions in acquisition or clearance of infection. Satisfactory performance was demonstrated by the estimators converging to the true interactions. Accurate reconstruction of interaction networks was achieved by regularization or penalization for sample size. Although performance deteriorated in the presence of host heterogeneity, this was overcome by correcting for individual-level risk factors. Our work demonstrates how statistical network inference could prove useful for detecting multi-strain pathogen interactions and may have applications beyond epidemiology.

Concurrent infection with multiple human papillomavirus types among unvaccinated and vaccinated 17-year-old Norwegian girls

Background

Whether type-specific human papillomavirus (HPV) infection influences the risk of acquiring infections with other HPV types is unclear. We studied concurrent HPV infections in 17-year-old girls from 2 birth cohorts; the first vaccine-eligible cohort in Norway and a prevaccination cohort.

Methods

Urine samples were collected and tested for 37 HPV genotypes. This study was restricted to unvaccinated girls from the prevaccination cohort (n = 5245) and vaccinated girls from the vaccine-eligible cohort (n = 4904). Risk of HPV infection was modelled using mixed-effect logistic regression. Expected frequencies of concurrent infection with each pairwise combination of the vaccine types and high-risk types (6/11/16/18/31/33/35/39/45/51/52/56/58/59) were compared to observed frequencies.

Results

Infection with multiple HPV types was more common among unvaccinated girls than vaccinated girls (9.2% vs 3.7%). HPV33 and HPV51 was the only HPV pair that was detected together more often than expected among both unvaccinated (P = .002) and vaccinated girls (P < .001). No HPV pairs were observed significantly less often than expected.

Conclusions

HPV33 and HPV51 tended to be involved in coinfection among both unvaccinated and vaccinated girls. The introduction of HPV vaccination does not seem to have had an effect on the tendency of specific HPV types to cluster together.

Capturing multiple-type interactions into practical predictors of type replacement following human papillomavirus vaccination

Current HPV vaccines target a subset of the oncogenic human papillomavirus (HPV) types. If HPV types compete during infection, vaccination may trigger replacement by the non-targeted types. Existing approaches to assess the risk of type replacement have focused on detecting competitive interactions between pairs of vaccine and non-vaccine types. However, methods to translate any inferred pairwise interactions into predictors of replacement have been lacking. In this paper, we develop practical predictors of type replacement in a multi-type setting, readily estimable from pre-vaccination longitudinal or cross-sectional prevalence data. The predictors we propose for replacement by individual non-targeted types take the form of weighted cross-hazard ratios of acquisition versus clearance, or aggregate odds ratios of coinfection with the vaccine types. We elucidate how the hazard-based predictors incorporate potentially heterogeneous direct and indirect type interactions by appropriately weighting type-specific hazards and show when they are equivalent to the odds-based predictors. Additionally, pooling type-specific predictors proves to be useful for predicting increase in the overall non-vaccine-type prevalence. Using simulations, we demonstrate good performance of the predictors under different interaction structures. We discuss potential applications and limitations of the proposed methodology in predicting type replacement, as compared to existing approaches. This article is part of the theme issue 'Silent cancer agents: multi-disciplinary modelling of human DNA oncoviruses'.

Global and regional estimates of the contribution of herpes simplex virus type 2 infection to HIV incidence: a population attributable fraction analysis using published epidemiological data

BACKGROUND

A 2017 systematic review and meta-analysis of 55 prospective studies found the adjusted risk of HIV acquisition to be at least tripled in individuals with herpes simplex virus type 2 (HSV-2) infection. We aimed to assess the potential contribution of HSV-2 infection to HIV incidence, given an effect of HSV-2 on HIV acquisition.

METHODS

We used a classic epidemiological formula to estimate the global and regional (WHO regional) population attributable fraction (PAF) and number of incident HIV infections attributable to HSV-2 infection by age (15-24 years, 25-49 years, and 15-49 years), sex, and timing of HSV-2 infection (established vs recently acquired). Estimates were calculated by incorporating HSV-2 and HIV infection data with pooled relative risk (RR) estimates for the effect of HSV-2 infection on HIV acquisition from a systematic review and meta-analysis. Because HSV-2 and HIV have shared sexual and other risk factors, in addition to HSV-related biological factors that increase HIV risk, we only used RR estimates that were adjusted for potential confounders.

FINDINGS

An estimated 420 000 (95% uncertainty interval 317 000-546 000; PAF 29·6% [22·9-37·1]) of 1·4 million sexually acquired incident HIV infections in individuals aged 15-49 years in 2016 were attributable to HSV-2 infection. The contribution of HSV-2 to HIV was largest for the WHO African region (PAF 37·1% [28·7-46·3]), women (34·8% [23·5-45·0]), individuals aged 25-49 years (32·4% [25·4-40·2]), and established HSV-2 infection (26·8% [19·7-34·5]).

INTERPRETATION

A large burden of HIV is likely to be attributable to HSV-2 infection, even if the effect of HSV-2 infection on HIV had been imperfectly measured in studies providing adjusted RR estimates, potentially because of residual confounding. The contribution is likely to be greatest in areas where HSV-2 is highly prevalent, particularly Africa. New preventive interventions against HSV-2 infection could not only improve the quality of life of millions of people by reducing the prevalence of herpetic genital ulcer disease, but could also have an additional, indirect effect on HIV transmission.

FUNDING

WHO.

Coinfections by noninteracting pathogens are not independent and require new tests of interaction

If pathogen species, strains, or clones do not interact, intuition suggests the proportion of coinfected hosts should be the product of the individual prevalences. Independence consequently underpins the wide range of methods for detecting pathogen interactions from cross-sectional survey data. However, the very simplest of epidemiological models challenge the underlying assumption of statistical independence. Even if pathogens do not interact, death of coinfected hosts causes net prevalences of individual pathogens to decrease simultaneously. The induced positive correlation between prevalences means the proportion of coinfected hosts is expected to be higher than multiplication would suggest. By modelling the dynamics of multiple noninteracting pathogens causing chronic infections, we develop a pair of novel tests of interaction that properly account for nonindependence between pathogens causing lifelong infection. Our tests allow us to reinterpret data from previous studies including pathogens of humans, plants, and animals. Our work demonstrates how methods to identify interactions between pathogens can be updated using simple epidemic models.

If pathogen species, strains, or clones do not interact, intuition suggests the proportion of coinfected hosts can be obtained by simply multiplying the individual prevalences. However, even simple epidemiological models show this to be untrue. This study develops new tests for interaction between pathogens that account for this surprising lack of statistical independence.

Detecting within-host interactions from genotype combination prevalence data

Parasite genetic diversity can provide information on disease transmission dynamics but most mathematical and statistical frameworks ignore the exact combinations of genotypes in infections. We introduce and validate a new method that combines explicit epidemiological modelling of coinfections and regression-Approximate Bayesian Computing (ABC) to detect within-host interactions. Using a susceptible-infected-susceptible (SIS) model, we show that, if sufficiently strong, within-host parasite interactions can be detected from epidemiological data. We also show that, in this simple setting, this detection is robust even in the face of some level of host heterogeneity in behaviour. These simulations results offer promising applications to analyse large datasets of multiple infection prevalence data, such as those collected for genital infections by Human Papillomaviruses (HPVs).

Inferring Pathogen Type Interactions Using Cross-sectional Prevalence Data: Opportunities and Pitfalls for Predicting Type Replacement

BACKGROUND

Many multivalent vaccines target only a subset of all pathogenic types. If vaccine and nonvaccine types compete, vaccination may lead to type replacement. The plausibility of type replacement has been assessed using the odds ratio (OR) of co-infections in cross-sectional prevalence data, with OR > 1 being interpreted as low risk of type replacement. The usefulness of the OR as a predictor for type replacement is debated, as it lacks a theoretical justification, and there is no framework explaining under which assumptions the OR predicts type replacement.

METHODS

We investigate the values that the OR can take based on deterministic S usceptible- I infected- S usceptible and S usceptible- Infected- Recovered- S usceptible multitype transmission models. We consider different mechanisms of type interactions and explore parameter values ranging from synergistic to competitive interactions.

RESULTS

We find that OR > 1 might mask competition because of confounding due to unobserved common risk factors and cross-immunity, as indicated by earlier studies. We prove mathematically that unobserved common risk factors lead to an elevation of the OR, and present an intuitive explanation why cross-immunity increases the OR. We find that OR < 1 is predictive for type replacement in the absence of immunity. With immunity, OR < 1 remains predictive under biologically reasonable assumptions of unidirectional interactions during infection, and an absence of immunity-induced synergism.

CONCLUSIONS

Using the OR in cross-sectional data to predict type replacement is justified, but is only unambiguous under strict assumptions. An accurate prediction of type replacement requires pathogen-specific knowledge on common risk factors and cross-immunity.

Evidence of synergistic relationships between HIV and Human Papillomavirus (HPV): systematic reviews and meta-analyses of longitudinal studies of HPV acquisition and clearance by HIV status, and of HIV acquisition by HPV status

INTRODUCTION

Observational studies suggest HIV and human papillomavirus (HPV) infections may have multiple interactions. We reviewed the strength of the evidence for the influence of HIV on HPV acquisition and clearance, and the influence of HPV on HIV acquisition.

METHODS

We performed meta-analytic systematic reviews of longitudinal studies of HPV incidence and clearance rate by HIV status (review 1) and of HIV incidence by HPV status (review 2). We pooled relative risk (RR) estimates across studies using random-effect models. I2 statistics and subgroup analyses were used to quantify heterogeneity across estimates and explore the influence of participant and study characteristics including study quality. Publication bias was examined quantitatively with funnel plots and subgroup analysis, as well as qualitatively.

RESULTS AND DISCUSSION

In review 1, 37 publications (25 independent studies) were included in the meta-analysis. HPV incidence (pooled RR = 1.55, 95% CI: 1.29 to 1.88; heterosexual males: pooled RR = 1.95, 95% CI: 1.62, 2.34; females: pooled RR = 1.63, 95% CI: 1.26 to 2.11; men who have sex with men: pooled RR = 1.36, 95% CI: 1.01 to 1.82) and high-risk HPV incidence (pooled RR = 2.20, 95% CI: 1.90 to 2.54) was approximately doubled among people living with HIV (PLHIV) whereas HPV clearance rate (pooled RR = 0.53, 95% CI: 0.42 to 0.67) was approximately halved. In review 2, 14 publications (11 independent studies) were included in the meta-analysis. HIV incidence was almost doubled (pooled RR = 1.91, 95% CI 1.38 to 2.65) in the presence of prevalent HPV infection. There was more evidence of publication bias in review 2, and somewhat greater risk of confounding in studies included in review 1. There was some evidence that adjustment for key confounders strengthened the associations for review 2. Misclassification bias by HIV/HPV exposure status could also have biased estimates toward the null.

CONCLUSIONS

These results provide evidence for synergistic HIV and HPV interactions of clinical and public health relevance. HPV vaccination may directly benefit PLHIV, and help control both HPV and HIV at the population level in high prevalence settings. Our estimates of association are useful for mathematical modelling. Although observational studies can never perfectly control for residual confounding, the evidence presented here lends further support for the presence of biological interactions between HIV and HPV that have a strong plausibility.

Recurring infection with ecologically distinct HPV types can explain high prevalence and diversity

The high prevalence of human papillomavirus (HPV), the most common sexually transmitted infection, arises from the coexistence of over 200 genetically distinct types. Accurately predicting the impact of vaccines that target multiple types requires understanding the factors that determine HPV diversity. The diversity of many pathogens is driven by type-specific or "homologous" immunity, which promotes the spread of variants to which hosts have little immunity. To test for homologous immunity and to identify mechanisms determining HPV transmission, we fitted nonlinear mechanistic models to longitudinal data on genital infections in unvaccinated men. Our results provide no evidence for homologous immunity, instead showing that infection with one HPV type strongly increases the risk of infection with that type for years afterward. For HPV16, the type responsible for most HPV-related cancers, an initial infection increases the 1-year probability of reinfection by 20-fold, and the probability of reinfection remains 14-fold higher 2 years later. This increased risk occurs in both sexually active and celibate men, suggesting that it arises from autoinoculation, episodic reactivation of latent virus, or both. Overall, our results suggest that high HPV prevalence and diversity can be explained by a combination of a lack of homologous immunity, frequent reinfections, weak competition between types, and variation in type fitness between host subpopulations. Because of the high risk of reinfection, vaccinating boys who have not yet been exposed may be crucial to reduce prevalence, but our results suggest that there may also be large benefits to vaccinating previously infected individuals.

Effect of HSV-2 infection on subsequent HIV acquisition: an updated systematic review and meta-analysis

Background

HIV and herpes simplex virus type 2 (HSV-2) infections cause a substantial global disease burden and are epidemiologically correlated. Two previous systematic reviews of the association between HSV-2 and HIV found evidence that HSV-2 infection increases the risk of HIV acquisition, but these reviews are now more than a decade old.

Methods

For this systematic review and meta-analysis, we searched PubMed, MEDLINE, and Embase (from Jan 1, 2003, to May 25, 2017) to identify studies investigating the risk of HIV acquisition after exposure to HSV-2 infection, either at baseline (prevalent HSV-2 infection) or during follow-up (incident HSV-2 infection). Studies were included if they were a cohort study, controlled trial, or case-control study (including case-control studies nested within a cohort study or clinical trial); if they assessed the effect of pre-existing HSV-2 infection on HIV acquisition; and if they determined the HSV-2 infection status of study participants with a type-specific assay. We calculated pooled random-effect estimates of the association between prevalent or incident HSV-2 infection and HIV seroconversion. We also extended previous investigations through detailed meta-regression and subgroup analyses. In particular, we investigated the effect of sex and risk group (general population vs higher-risk populations) on the relative risk (RR) of HIV acquisition after prevalent or incident HSV-2 infection. Higher-risk populations included female sex workers and their clients, men who have sex with men, serodiscordant couples, and attendees of sexually transmitted infection clinics.

Findings

We identified 57 longitudinal studies exploring the association between HSV-2 and HIV. HIV acquisition was almost tripled in the presence of prevalent HSV-2 infection among general populations (adjusted RR 2·7, 95% CI 2·2–3·4; number of estimates [N_e]=22) and was roughly doubled among higher-risk populations (1·7, 1·4–2·1; N_e=25). Incident HSV-2 infection in general populations was associated with the highest risk of acquisition of HIV (4·7, 2·2–10·1; N_e=6). Adjustment for confounders at the study level was often incomplete but did not significantly affect the results. We found moderate heterogeneity across study estimates, which was explained by risk group, world region, and HSV-2 exposure type (prevalent vs incident).

Interpretation

We found evidence that HSV-2 infection increases the risk of HIV acquisition. This finding has important implications for management of individuals diagnosed with HSV-2 infection, particularly for those who are newly infected. Interventions targeting HSV-2, such as new HSV vaccines, have the potential for additional benefit against HIV, which could be particularly powerful in regions with a high incidence of co-infection.

Funding

World Health Organization.