This Week in Infectious Disease — Aug 1, 2026

Welcome to This Week in Infectious Disease. This week we're covering 10 notable papers spanning antiretroviral therapy and its metabolic consequences, the hard problems of bone, joint and central nervous system infections, drug safety and regimen simplification, and outbreak and post-pandemic epidemiology. Let's dive in.

We start with HIV, where two papers approach the same broad question from opposite ends: how do we keep people virologically suppressed without paying a price we didn't anticipate? In JAMA, the Opti-DOR randomized clinical trial enrolled 600 antiretroviral-naive adults at two South African sites, almost all Black African and about two-thirds assigned female at birth, and compared once-daily doravirine, lamivudine and tenofovir disoproxil fumarate against dolutegravir, emtricitabine and tenofovir alafenamide [1]. At week 48, viral suppression below fifty copies was achieved by 89 percent on the doravirine arm and 90.7 percent on the dolutegravir arm, a difference of under two percentage points that comfortably met the prespecified noninferiority margin of ten points. The headline secondary outcome is the weight signal: median gain was three kilograms with doravirine-based therapy versus five kilograms with dolutegravir and tenofovir alafenamide, a two kilogram difference that was statistically robust. But this is a trade-off, not a free lunch. Bone mineral density fell substantially more in the doravirine and tenofovir disoproxil fumarate arm — roughly two percent at the hip and three percent at the spine, compared with about half a percent and just over one percent with the tenofovir alafenamide regimen. And the resistance picture deserves attention: of nine virologic failures on doravirine, seven developed doravirine resistance, though five of six who were switched to dolutegravir-based therapy resuppressed. So for a clinician counselling a patient who is deeply concerned about weight gain, particularly a younger patient with cardiometabolic risk, there is now randomized evidence that an alternative anchor drug delivers comparable suppression with less weight gain — at the cost of bone density and a lower barrier to resistance.

The other side of the HIV coin comes from the International Journal of Antimicrobial Agents, where a retrospective case series looked at ten adolescents and young adults who developed virologic abnormalities after switching from suppressive oral therapy to long-acting injectable cabotegravir and rilpivirine [7]. Out of 76 patients on the injectable regimen at that centre, 11 had virologic problems and 10 underwent pharmacogenomic testing. All had viral blips, six developed persistent low-level viremia, and one progressed to virologic failure. Every single one of the ten had at least one non-normal metabolizer phenotype — eight had CYP2B6 intermediate or poor metabolizer status, seven had clinically relevant CYP2C19 variants, and nine had CYP3A5 variants, with most showing variability across several pathways at once. This is a small, uncontrolled series with no comparison group of patients who stayed suppressed, so it cannot tell us that these variants caused the rebound; individually these polymorphisms are common. What it does is generate a genuinely testable hypothesis about why long-acting therapy occasionally fails in the real world despite excellent trial data, and it argues for prospective pharmacokinetic and pharmacogenomic study rather than for ordering panels on your clinic population tomorrow.

Our second theme covers the infections where the hardest question is not which drug, but how much surgery and for how long. Clinical Infectious Diseases published a thoughtful critique of the clean margin concept in diabetic foot osteomyelitis [3]. Updated guidelines now endorse tailoring antibiotic duration to resection margins — short courses if the margin is clean, longer if residual infected bone remains — and the authors argue that this appealing idea rests on a construct that is ill-defined and inconsistently measured. Margin assessment is done surgically, histopathologically and microbiologically, and those three domains are not interchangeable and frequently disagree with each other. The recommendations derive in part from limited trial data with uncertain ascertainment of residual infection, and there has been essentially no prospective validation. The practical message is caution: if you are shortening antibiotics on the basis of a surgeon's impression that the margin looked clean, understand that you are acting on a soft data point, and document how the margin was determined so that your reasoning is transparent when the patient relapses.

Staying with orthopaedic infection, the International Journal of Antimicrobial Agents reported a single-centre cohort from Basel of 250 patients with implant-associated bone and joint infection due to Staphylococcus epidermidis, screened from nearly 1,800 patients with intraoperative isolates [5]. Fifty-six percent of infections involved multiresistant S. epidermidis. The strongest predictor of multiresistance was previous antibiotic exposure, roughly doubling the odds. Overall, about a third of patients experienced treatment failure, and multiresistance was associated with a more than fourfold increase in the hazard of failure occurring beyond two weeks after revision surgery. That temporal pattern matters — early failures look similar regardless of susceptibility, but the late failures, the ones that represent inadequate suppression of biofilm, cluster among the resistant isolates. Put alongside the diabetic foot piece, the shared lesson is that in device and bone infection our margin for error is small, our data are thinner than the guidelines imply, and prior antibiotic exposure is a modifiable driver of the very resistance that later defeats us.

The third paper in this group tackles a rarer but equally unforgiving problem. In Open Forum Infectious Diseases, investigators report a patient with fluoroquinolone-resistant, pre-extensively drug-resistant tuberculous meningitis who was cured with a ten-month regimen built on bedaquiline, pretomanid and linezolid plus clofazimine and cycloserine, with paired plasma and cerebrospinal fluid drug concentrations measured by liquid chromatography [4]. The cerebrospinal-fluid-to-plasma unbound exposure ratio was about 1.5 for pretomanid and about 1.0 for linezolid; pretomanid stayed above the critical concentration of 0.5 milligrams per litre for most of the dosing interval, and linezolid achieved an exposure to minimum inhibitory concentration ratio of 171 at that same MIC. Estimated unbound peak bedaquiline concentrations in cerebrospinal fluid were comparable to plasma, suggesting therapeutic central nervous system penetration despite bedaquiline's reputation for heavy protein binding. The caveat the authors themselves flag is that for organisms with MICs above 0.5, cerebrospinal fluid concentrations of pretomanid and linezolid may fall short. This is a single case, so treat it as a proof of concept — but it is meaningful reassurance for anyone facing drug-resistant tuberculous meningitis with no good options.

Our third theme is optimisation — using host factors and shorter regimens to make existing drugs safer and simpler. Also in the International Journal of Antimicrobial Agents, a Japanese multicentre retrospective cohort of 608 patients asked whether the albumin-bilirubin score, a simple measure of hepatic reserve, adds anything to therapeutic drug monitoring for predicting voriconazole hepatotoxicity [8]. It does. Patients with an albumin-bilirubin score at or above minus two had roughly double the risk of hepatotoxicity independent of other factors. In a decision tree, patients with good hepatic reserve and a trough below four micrograms per millilitre had the lowest event rate at about six and a half percent, while those with poor reserve and a trough at or above four had the highest at eighteen percent, with the model achieving nearly ninety percent accuracy. That is an easy, free calculation from two labs you already have, and it lets you decide who needs aggressive early trough targeting and closer liver monitoring.

The simplification theme continues with Helicobacter pylori. A six-centre randomized controlled trial randomized 330 treatment-naive patients to one of three vonoprazan-based regimens: fourteen-day dual therapy with amoxicillin, ten-day dual therapy, or a seven-day quadruple regimen of vonoprazan, tetracycline, amoxicillin and bismuth [6]. The seven-day quadruple regimen was noninferior to the fourteen-day standard, with eradication rates around 90 percent versus 92 percent by intention to treat and above 93 percent per protocol, and no meaningful differences in adverse events or compliance. The ten-day dual regimen, however, did not meet the noninferiority criterion — it achieved acceptable per-protocol eradication but formally failed the test. So the message is nuanced: you can shorten to seven days if you intensify the regimen, but you cannot simply lop four days off dual therapy and expect equivalence.

Finally, epidemiology and preparedness. The Lancet Infectious Diseases published a review of Bundibugyo virus in the context of the 2026 outbreak in the Democratic Republic of the Congo and Uganda, declared a Public Health Emergency of International Concern by the World Health Organization in May 2026 [2]. Bundibugyo is one of the least studied human-pathogenic orthoebolaviruses — before this event there had been only two recognised outbreaks since its discovery in 2007, which is why we lack species-specific countermeasures. The authors' central argument is that the bottleneck is no longer scientific: pan-filovirus diagnostics, investigational vaccines and therapeutics, and adaptive trial platforms all exist. What has constrained the response is delayed diagnosis, limited access to species-inclusive diagnostics, conflict-related insecurity, population displacement and fragile health systems. For clinicians, the operational point is that Ebola virus disease is not one disease with one test and one vaccine, and species-inclusive diagnostics need to be part of your differential and your laboratory's capability.

Two further papers address post-pandemic and enteric risk. In Open Forum Infectious Diseases, a systematic review and meta-analysis of 43 studies asked whether oral antivirals still work in vaccinated populations in the Omicron era [9]. The randomized evidence is sobering: the two trials conducted in vaccinated populations showed no reduction in death with nirmatrelvir-ritonavir and no reduction in hospitalization or death with molnupiravir — though the authors note those trials were underpowered to detect clinically important differences. Observational studies told a different story for nirmatrelvir-ritonavir, consistently associating it with roughly a 40 percent lower odds of hospitalization and a substantially lower odds of death. Molnupiravir showed no significant reduction in hospitalization and mixed results for mortality. The discordance between randomized and observational data is the real finding here, and confounding by indication cuts in both directions; the pragmatic conclusion is to reserve nirmatrelvir-ritonavir for genuinely high-risk patients rather than treating broadly. Complementing this, a retrospective cohort using United States commercial claims data examined over 1.7 million all-cause acute gastroenteritis episodes and nearly 5,800 coded norovirus episodes between 2022 and 2024 [10]. Eleven percent of all-cause and nearly 45 percent of coded norovirus episodes led to hospitalization within three days. Cardiovascular disease raised hospitalization risk by roughly two-thirds for all-cause gastroenteritis, and having two or more chronic conditions roughly doubled it. Notably, the relative effect of comorbidity was larger in adults aged 18 to 64 than in those 65 and over — a useful reminder as norovirus vaccine candidates advance that risk targeting should not be by age alone.

If you only have time for one paper this week, make it the Opti-DOR trial in JAMA [1]. It is the first randomized evidence directly addressing the weight gain problem that dominates antiretroviral conversations in clinic, and it hands you a concrete, evidence-based alternative — with an equally concrete set of trade-offs to discuss.

Here are the key takeaways from this week in Infectious Disease. First, a doravirine, lamivudine and tenofovir disoproxil fumarate regimen matches dolutegravir-based therapy for 48-week suppression with about two kilograms less weight gain, but costs more bone density and carries a lower resistance barrier. Second, in bone, joint and device infection, prior antibiotic exposure selects for multiresistant Staphylococcus epidermidis and predicts late treatment failure, while the clean margin concept guiding antibiotic duration in diabetic foot osteomyelitis remains unvalidated and inconsistently measured. Third, a simple albumin-bilirubin score combined with voriconazole trough concentration stratifies hepatotoxicity risk with high accuracy — use both, not just the trough. Fourth, for Helicobacter pylori, a seven-day vonoprazan-based quadruple regimen was noninferior to fourteen-day dual therapy, but shortening dual therapy to ten days was not. And fifth, randomized trials in vaccinated populations have not demonstrated that oral antivirals reduce COVID-19 hospitalization or death, even though observational data continue to support nirmatrelvir-ritonavir — target treatment to the highest-risk patients.

That's your roundup for This Week in Infectious Disease. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And if you have a paper you have been meaning to read, upload the PDF, or paste any link, at audioscholar dot C C. We will turn it into audio like this one, in any of thirty-one languages.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.