TOPLINE

Prolonged co-trimoxazole prophylaxis in adults with newly diagnosed HIV briefly lowered carriage of Staphylococcus aureus (S aureus) and methicillin-resistant S aureus (MRSA), but these reductions did not sustain by 1 year. Deaths and serious adverse events did not differ between the co-trimoxazole and placebo groups over the 1-year follow-up.

METHODOLOGY

• In a randomized controlled trial conducted across six sites in Dar es Salaam, Tanzania, researchers tested whether daily co-trimoxazole prophylaxis affects the colonization and carriage of resistant bacteria in the gut, nose, and throat of people with HIV.
• The study enrolled 537 adults with newly diagnosed HIV and CD4 counts ≥ 350 cells/µL between 2017 and 2018 (median age, 37 years; 77% women).
• Participants were randomly assigned to receive either daily oral co-trimoxazole (80 mg trimethoprim/400 mg sulfamethoxazole) or an identical placebo for 1 year.
• Swabs from the rectum (to detect gut colonization) and the nasopharynx/nose (to evaluate nasal carriage) were collected at the start, then at 2 weeks, 6 months, and 12 months. The study's primary outcomes were colonization with the following resistant bacteria during follow-up: extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli (ESBL-EC), ESBL-producing Klebsiella species (ESBL-K), MRSA, and penicillin-non-susceptible pneumococci.
• Researchers also tracked deaths and episodes of malaria, pneumonia, and diarrhea over the year.

TAKEAWAY

• At 1 year, nasal colonization rates for both S aureus and MRSA did not differ significantly between the groups, despite earlier temporary reductions in the co-trimoxazole arm. Co-trimoxazole reduced S aureus carriage at 2 weeks (risk difference [RD], -7.3%; 95% CI, -12.0% to -2.6%) and 6 months (RD, -7.6%; 95% CI, -13.9% to -1.3%), whereas MRSA carriage was reduced only at 2 weeks (RD, -4.5%; 95% CI, -8.0% to -1.1%) before rebounding.
• Co-trimoxazole had no significant effect on pneumococcal colonization at 1 year; 74% of all pneumococcal isolates identified were penicillin-non-susceptible, and roughly a quarter of isolates were serotypes included in the 13-valent pneumococcal conjugate vaccine.
• ESBL-EC colonization rose substantially in both groups over 1 year — doubling in the co-trimoxazole group and rising moderately in the placebo group — with no significant between-group difference at 1 year; similarly, ESBL-K colonization tripled overall. The use of non-study antibiotics during the trial was associated with an increased likelihood of ESBL-EC colonization (pooled RD, 12%; 95% CI, 3%-22%).
• Rates of mortality (1.2% vs 2.2%), hospitalization (2.3% vs 3.2%), serious adverse events (4.6% vs 5.4%), and episodes of self-reported malaria did not differ significantly between the co-trimoxazole and placebo groups over 1 year.

IN PRACTICE

"Our findings provide some support for the WHO recommendation for prolonged co‑trimoxazole prophylaxis by showing temporary reductions in S. aureus and MRSA carriage. However, the rebound within 6–12 months raises concerns about its long‑term effect on MRSA colonisation. The absence of measurable impact on survival or severe adverse events, though not a primary endpoint, offers little justification for extended prophylaxis in immunocompetent people with HIV," the authors wrote. 

SOURCE

The study was led by Bjørn Blomberg, MD, PhD, University of Bergen, Bergen, Norway. It was published online on September 28, 2026, in the International Journal of Infectious Diseases.

LIMITATIONS

Patient compliance with the daily regimen was self-reported and fell below optimal levels over the course of trial. High proportion of screened participants declined to enroll. Using one swab for both the nostrils and nasopharynx may have lowered pathogen yield, particularly for pneumococci. 

DISCLOSURES

The study was supported by grants from multiple agencies, including the Western Norway Regional Health Trust, National Centre for Tropical Infectious Diseases at Haukeland University Hospital, and Combatting Anti-Microbial Resistance with Interdisciplinary Approaches. The authors reported no competing interests.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.