High Propensity for Multidrug-Resistant Pneumococcal Shedding Among Adults Living With HIV on Stable Antiretroviral Therapy in Malawi

Abstract

Background: People living with human immunodeficiency virus (HIV; PLHIV) on antiretroviral therapy (ART) are still at risk of pneumococcal disease and have over 2-fold higher pneumococcal carriage prevalence than HIV-uninfected (HIV-) adults). Carriage is a risk factor for pneumococcal disease, antimicrobial resistance (AMR) emergence, and transmission. Therefore, we tested whether the high prevalence of pneumococcal carriage in PLHIV on ART is associated with increased bacterial density, shedding, and AMR.

Methods: We recruited asymptomatic PLHIV on ART for >1 year (PLHIV-ART>1y) and HIV- adults. Nasopharyngeal swab samples were collected on days 3, 7, 14, 21, and 28, followed by monthly collections for 12 months, while shedding samples were collected on days 3, 21, and 28. Peripheral blood samples were collected on day 3 to measure CD4 cell count and HIV viral load. Pneumococcal carriage density and shedding were assessed using standard bacterial culture, multiple carriage was detected using whole-plate sweep sequencing, and AMR profiling was conducted using disk diffusion and Etest.

Results: PLHIV-ART>1y had a higher propensity for high-density carriage (adjusted odds ratio, 1.67 [95% confidence interval (CI), 1.07-2.60]; P = .02). Moreover, PLHIV-ART>1y are more likely to shed pneumococci than HIV- adults (adjusted odds ratio, 2.52 [95% CI, 1.06-6.00]; P = .04), with carriage density identified as an important risk factor for shedding (3.35 [1.55-7.24]; P = .002). Aerosol shed isolates from PLHIV-ART>1y were mostly multidrug resistant (18 of 29 [ 62%; 95% CI, 48%-77%]).

Conclusions: These findings indicate that PLHIV-ART>1y remain at high risk of pneumococcal disease and could also be an important reservoir for shedding multidrug-resistant pneumococci.

Keywords: AMR; ART; HIV; Streptococcus pneumoniae; carriage.

Figure 1.

Recruitment flow diagram showing the number of adults and pneumococcal samples included in the analysis among people living with human immunodeficiency virus (HIV; PLHIV). Abbreviations: ART, antiretroviral therapy; COVID-19, coronavirus disease 2019.

Figure 2.

Overall phenotypic pneumococcal swab positivity among people living with human immunodeficiency virus (HIV) and on antiretroviral therapy for >1 year (PLHIV-ART>1y) and HIV-uninfected (HIV−) adults, including 13-valent pneumococcal conjugate vaccine (PCV13) serotypes (vaccine type [VT]) and non-PCV13 serotypes (non-VT [NVT]). A, HIV-stratified pneumococcal swab sample positivity. B, HIV- and serogroup-stratified pneumococcal swab sample positivity by nasopharyngeal sampling visit. C, HIV- and serogroup-stratified pneumococcal swab sample positivity. D, HIV-stratified proportion of PCV13 serotypes. The denominator for each serotype is the total number of PCV13 serotypes in each group. Whiskers represent 95% confidence intervals. Data were analyzed using χ² tests (35 HIV− adults; 55 PLHIV-ART>1y). Source data are provided as a Source Data file.

Figure 3.

Pneumococcal carriage density among people living with human immunodeficiency virus (HIV) and on antiretroviral therapy for >1 year (PLHIV-ART>1y) and HIV-uninfected (HIV−) participants. A, Log median carriage density stratified by HIV status for 5 months of study follow-up. B, Aggregated log median carriage density stratified by HIV status for 5 months of study follow-up. For all box plots, box boundaries correspond to 25th and 75th percentiles; whiskers extend to a maximum of 1.5 times the interquartile range, with values outside the box and whiskers being outliers. Data were analyzed using the Wilcoxon test (35 HIV− adults [187 swab samples]; 55 PLHIV-ART>1y [358 swab samples]). Pneumococcal density was not assessed on screening samples or on day 0. *P = .04; **P = .006. Abbreviation: CFUs, colony-forming units.

Figure 4.

Phenotypic pneumococcal shedding among people living with human immunodeficiency virus (HIV) and on antiretroviral therapy for >1 year (PLHIV-ART>1y) and HIV-uninfected (HIV−) adults. Pneumococcal shedding was defined as the presence of viable pneumococci following microbiological culture of samples collected through nose poking (mechanical) and coughing (aerosol) or face mask sampling. Viable pneumococci were isolated in only 3 PLHIV-ART>1y using the face mask approach. A, Overall pneumococcal shedding positivity stratified by HIV status. B, Overall pneumococcal shedding positivity stratified by HIV status and serogroup. C, Overall pneumococcal shedding positivity stratified by HIV status and 13-valent pneumococcal conjugate vaccine (PCV13) serotypes; the denominator for each serotype is the total number of PCV13 serotypes in each group. Bars represent medians and whiskers, confidence intervals. Data were analyzed using χ² tests (22 HIV− adults; 61 PLHIV-ART>1y). Source data are provided as a Source Data file.

Figure 5.

Species and pneumococcal serotype diversity from aerosol shedding (cough) and nasopharyngeal carriage (Posterior Nasopharyngeal Swab, PNS) among people living with human immunodeficiency virus and on antiretroviral therapy for >1 year (PLHIV-ART>1y) based on genomic data. A, Proportion of reads for each species per sample stratified by species and sample type. For all box plots, box boundaries correspond to 25th and 75th percentiles; whiskers extend to a maximum of 1.5 times the interquartile range, with values outside the box and whiskers being outliers. B, Proportions of serotypes stratified by serotype and sample type. These samples are pairs of PNS and cough isolates from the same individual among PLHIV-ART>1y. Data were analyzed using the Wilcoxon test (17 PLHIV-ART>1y; 19 nasopharyngeal swab and 19 aerosol samples). **P = .001; ****P < .001.