Emergence of antibiotic resistance in immunocompromised host populations: A case study of emerging antibiotic resistant tuberculosis in AIDS patients

Abstract

Objective: The evolution of antibiotic resistance is far outpacing the development of new antibiotics, causing global public health concern about infections that will increasingly be unresponsive to antimicrobials. This risk of emerging antibiotic resistance may be meaningfully altered in highly AIDS-immunocompromised populations. Such populations fundamentally alter the bacterial evolutionary landscape in two ways, which we seek to model and analyze. First, widespread, population-level immunoincompetence creates a novel host environment with disrupted selective pressures. Second, within AIDS-prevalent populations, the recommendation that antibiotics be taken to treat and prevent opportunistic infection raises the risk of selection for drug-resistant pathogens.

Design: To determine the impact of HIV/AIDS on the emergence of antibiotic resistance-specifically in the developing world where high prevalence and economic challenges complicate disease management.

Methods: We present an SEIR epidemiological model of bacterial infection, and parametrize it to capture HIV/AIDS-attributable emergence of resistance under conditions of both high and low HIV/AIDS prevalence.

Results: We demonstrate that HIV/AIDS-immunocompromised hosts can be responsible for a disproportionately greater contribution to emergence of resistance than would be expected based on population-wide HIV/AIDS prevalence alone.

Conclusions: As such, the AIDS-immunocompromised have the potential become wellsprings of novel, resistant, opportunistic pathogen strains that can propagate into the broader global community. We discuss how public health policies for HIV/AIDS management can shape the evolutionary environment for opportunistic bacterial infections.

Fig 1. SEIR transmission dynamics.

1a. Basic SEIR model (assuming a closed system), wherein health status changes from susceptible to exposed at a rate of β, from exposed to infectious at a rate of ζ, and infectious to recovered at a rate of γ. The super- and subscripts “i” and “a” are used generically to demonstrate that there are many possible host health outcomes, depending on the combination of immune status and antibiotic-taking behavior. 1b. Transmission dynamics specific to our model; we note all possible progressions for an HIV/AIDS- host that contracts a bacterial pathogen.

Fig 2. Immune status-based contributions to emergence, and infectivity with and without HIV/AIDS.

In both Indonesia (2a) and Swaziland (2b), and under conditions of both high (H) and low (L) TB prevalence, AIDS-immunocompromised hosts are individually responsible for considerably more emergence that would be expected given their prevalence in each country. (We have made the simplifying assumption that HIV/AIDS+, HAART+ hosts are functionally immunocompetent except with respect to loss of immune memory.) Under conditions of high (H) and low (L) TB prevalence, we compare health status-based TB incidence in Indonesia (2c) and Swaziland (2d), visualizing conditions of both actual and zero HIV/AIDS prevalence. In both countries, AIDS increases total incidence. In Swaziland, when TB prevalence is low, we observe an increase in incidence of 9.6%, relative to the AIDS-absent condition; when TB prevalence is high, the relative increase in incidence is 9.9%. In Indonesia, the changes in TB incidence are much more pronounced, with an observed 167.9% increase when TB prevalence is low and a 167.12% increase when TB prevalence is high. Immunocompetent. (We assume 20% adherence to antibiotics, except where specifically noted. [25]).

Fig 3. Emergence as a functions of increasing HIV/AIDS prevalence and adherence.

3a. In Using Swaziland’s low TB condition as an example, and assuming 95% antibiotic adherence [25], we found that, as HIV/AIDS prevalence increases from zero to 30%, we observe a corresponding increase in expected population-wide emergence of antibiotic resistance. Our conservative estimate of 20% antibiotic adherence represents a best-case scenario in terms of expected emergence; however, the likelihood of emergence becomes greater as adherence decreases. [26] 3b. Within the developing world, economic, medical and social barriers can limit antibiotic adherence. [26] To measure the impact of antibiotic adherence on emerging resistance within HIV/AIDS prevalent environments, we varied the probability of complete adherence (represented as C₁ in the model) from 0–100%, while maintaining the original adherence ratios reported in the literature [25, 27] (see S1 Appendix). We use Swaziland’s low TB condition for purposes of illustration, however, we note that the results for Swaziland’s high TB, as well as both the low and high TB conditions in Indonesia, mirror the results presented: antibiotic adherence has very little impact on the emergence of resistance in HIV/AIDS-immunocompromised host populations, relative to the impact of HIV/AIDS prevalence.