This Week in Infectious Disease — Aug 28, 2026

Welcome to This Week in Infectious Disease. This week we're covering 10 notable papers spanning antibiotic duration and stewardship, tuberculosis and HIV therapeutics, and a cluster of emerging-pathogen surveillance reports from three continents. Let's dive in.

We start with the paper most likely to change what you write on a discharge summary. In the New England Journal of Medicine, Bundgaard and colleagues report POET II, an international open-label randomized trial of 508 adults with left-sided infective endocarditis caused by Staphylococcus aureus, Enterococcus faecalis, or streptococci [1]. Everyone received at least the prespecified two to four weeks of therapy and had to meet clinical stabilization criteria before randomization. At that point, half of the patients simply stopped antibiotics, while the others completed a standard total course of four to six weeks. On the primary efficacy endpoint, days alive without antibiotic treatment over six months, the tailored strategy won, with a median of 183 days versus 169 — a difference of about 13 days, and formally superior. The composite safety endpoint of death, unplanned cardiac surgery, or symptomatic embolism occurred in about 8 percent of the tailored group and about 11 percent of the standard group, meeting the noninferiority margin. But there is a catch, and it matters: relapse of bacteremia or endocarditis occurred in 13 tailored-therapy patients versus 4 standard-therapy patients — roughly one in twenty compared with under one in fifty — and that difference was statistically significant. So the honest reading is that a response-tailored stop rule in a stabilized patient buys about two fewer weeks of antibiotics with no measurable excess of hard cardiac events, at the cost of a threefold higher relapse signal. That is a conversation to have with your cardiac surgeons and with the individual patient, not a blanket policy change.

Staying with antimicrobial use, PLOS Medicine carries a commentary from Shchepanik and colleagues on the LAMPA trial, which asked whether simply improving access to antimicrobial guidelines improves prescribing [4]. A smartphone app delivering the guidelines to prescribers in a low- and middle-income setting did not improve adherence. The message is one many of us have learned the hard way: access is not the bottleneck. Prescriber behaviour is driven by diagnostic uncertainty, drug availability, hierarchy, and habit, so stewardship interventions have to be multifaceted, embedded, and sustained rather than delivered as a single digital tool. Two other papers this week frame the resistance problem from opposite ends. In the International Journal of Antimicrobial Agents, Kandpal and colleagues review the mechanisms behind Helicobacter pylori eradication failure — point mutations in the 23S ribosomal RNA gene for clarithromycin, gyrA and gyrB for fluoroquinolones, rdxA for metronidazole — layered on top of efflux pumps, biofilm, and phase-variable methylation [5]. Practically, they emphasise regionally informed or susceptibility-guided therapy, bismuth quadruple regimens, and note that adjunctive probiotics have been reported to raise eradication rates into the high seventies to high eighties percent range. And in Nature Reviews Microbiology, Sinclair and colleagues give a sober appraisal of phage therapy [2]. Their central point is that in vitro phage susceptibility does not reliably predict clinical success, there is still no consensus on dosing, route, or combination strategy, and phage-bacterium-host dynamics plus immune interactions all shape outcome. Production, scalability, and regulation remain unresolved. Phage is a real tool for otherwise untreatable resistant infections, but it cannot yet be prescribed by antibiotic logic.

Turning to tuberculosis, two papers approach prevention from very different angles. In Emerging Infectious Diseases, Ainiwaer and colleagues built a Markov microsimulation of 400,000 people immigrating to Canada in 2025, comparing current post-arrival latent infection screening at roughly half a percent with a scenario of 68 percent screening [7]. The novel element is accounting for post-tuberculosis morbidity — the long-term respiratory impairment, excess mortality, and healthcare costs that follow cured disease. Including those consequences more than doubled the estimated quality-adjusted life years lost to tuberculosis, and it also roughly doubled the benefit attributed to expanded screening, cutting the cost per quality-adjusted life year from about 235,000 Canadian dollars to about 100,000. In other words, conventional analyses that stop the clock at cure systematically undervalue preventive treatment. Alongside that, the Journal of Infectious Diseases publishes a genomic analysis from Wattiau and colleagues of Mycobacterium tuberculosis isolates from the phase 2b trial of the M72 adjuvanted candidate vaccine [8]. One hundred isolates from 50 participants who developed microbiologically confirmed pulmonary disease were sequenced. Lineages were diverse in both vaccine and placebo arms, no multidrug-resistant strains were found, and although the antigen-encoding sequences showed low variation in pepA and high variation in PPE18, there was no preferential distribution of variants between arms. That is a reassuring negative finding — no evidence that the vaccine selects for escape strains — and it supports continued development. The cluster analysis, which identified ten genomic clusters shared across participants, is a reminder of dense ongoing transmission and mixed-strain infection in high-burden settings.

On the HIV side, the International Journal of Antimicrobial Agents reports the 144-week update of the BICTEL real-world cohort from Bortolani and colleagues, covering 180 people who switched to bictegravir with emtricitabine and tenofovir alafenamide in routine care [3]. At three years, about 91 percent had HIV RNA below 50 copies per millilitre, and fewer than 2 percent were above 200. There were no permanent discontinuations, regimen switches, or tolerability-related interruptions across the whole period. CD4 count and the CD4-to-CD8 ratio rose, total and low-density lipoprotein cholesterol and triglycerides fell, body mass index stayed stable, liver enzymes were unchanged, and estimated glomerular filtration rate declined modestly. Importantly, there was no signal that older participants fared differently over time. This is observational, retrospective, and single-cohort, so it cannot settle questions about weight gain that randomized data have raised — but it is useful durability evidence for the older, clinically complex patients who dominate many clinics.

Finally, three surveillance reports in Emerging Infectious Diseases that together make a case for keeping travel and exposure history at the front of your consultation. Jabet and colleagues used a global mass spectrometry identification network to track Trichophyton indotineae, the frequently terbinafine-resistant dermatophyte, and documented a threefold increase in identifications between 2022 and 2025 [6]. If you see extensive, treatment-refractory tinea, species identification and antifungal susceptibility — not another course of terbinafine — is the right next step. D'Angelo and colleagues describe five human Oropouche virus infections in rural Venezuela in 2025, all belonging to the reassortant lineage driving the 2024 to 2025 South American epidemic, confirming ongoing transmission and the need for laboratory capacity in the region [9]. And Ogawa and colleagues report the reemergence of summer scrub typhus in Japan: a pregnant woman infected after attending a fireworks event in Niigata Prefecture developed scrub typhus that resulted in fetal death, and field work recovered identical Orientia tsutsugamushi sequences from Leptotrombidium akamushi larvae and a field mouse at the site [10]. The genotype was Karp-like and Vietnam-related rather than the Kato type usually linked with that mite — a locally novel strain in an unexpected season, in a patient with no rural occupational exposure.

If you only have time for one paper this week, make it POET II in the New England Journal of Medicine [1]. It is the first randomized evidence that antibiotic duration in left-sided endocarditis can be shortened by clinical response rather than by consensus calendar — and its relapse signal is exactly the nuance you need at the bedside before you shorten a course.

Here are the key takeaways from this week in Infectious Disease. First, in stabilized left-sided endocarditis, stopping antibiotics after two to four weeks based on clinical response was safe by the trial's composite cardiac endpoint but tripled relapse, so individualise rather than generalise. Second, stewardship is behavioural, not informational — giving prescribers easier access to guidelines through an app did not improve adherence. Third, valuing tuberculosis prevention without counting post-tuberculosis disability roughly halves its apparent cost-effectiveness, and screening programmes should be re-appraised on that basis. Fourth, three-year real-world data support the durability and tolerability of bictegravir-based single-tablet therapy, including in older patients. And fifth, terbinafine-resistant Trichophyton indotineae is rising globally, while Oropouche virus in Venezuela and out-of-season scrub typhus in Japan are reminders that exposure history remains a diagnostic instrument.

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 think of one colleague — in any specialty — who never has time to keep up with the literature. Tell them about AudioScholar: a free ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.

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