This Week in Infectious Disease — Sep 18, 2026

Welcome to This Week in Infectious Disease. This week we're covering 10 notable papers spanning treatment de-escalation and shorter, simpler regimens; the economics of doing the right thing; and new diagnostics for tuberculosis, hepatitis C and invasive fungal disease. Let's dive in.

We start with two trials that ask whether we can give less drug, or give it less often, without losing efficacy. In the New England Journal of Medicine, McNally and colleagues report SOLARIO, a multicentre open-label noninferiority trial in 500 adults undergoing surgery for orthopaedic infection in whom a local antibiotic carrier was implanted at the time of surgery [1]. Patients were randomised to systemic antibiotics for at least four weeks or for no more than seven days. Definite treatment failure at twelve months, adjudicated by a blinded endpoint committee, occurred in about 14 percent of the long-duration group and about 11 percent of the short-duration group — a difference favouring the short course and comfortably within the ten-point noninferiority margin, with sensitivity and per-protocol analyses agreeing. The tolerability signal is arguably the more striking part: by six weeks, treatment-related symptoms had occurred in close to half OF PATIENTS on the long course versus about one in six on the short course. The caveat is that everyone received a local antibiotic carrier, so this is not licence to abbreviate systemic therapy in bone and joint infection generally — but where local delivery is used, seven days appears to be enough.

Staying with simplification, The Lancet published LATA, a 96-week randomised noninferiority trial of long-acting injectable cabotegravir-rilpivirine every eight weeks versus daily oral dolutegravir, lamivudine and tenofovir in 476 virologically suppressed adolescents across Kenya, South Africa, Uganda and Zimbabwe [2]. Nearly all participants had acquired HIV vertically and had been on antiretroviral therapy for more than a decade. Confirmed viraemia at or above 50 copies by week 96 occurred in two participants on injectables versus fifteen on oral therapy — roughly one percent against six percent. Injectables were not just noninferior but statistically superior. Seven participants permanently discontinued the injectable, including two for drug-related adverse events, one for incident tuberculosis and two planning pregnancy. For a population with notoriously fragile adherence, this is the strongest evidence yet that eight-weekly injections belong in paediatric and adolescent HIV programmes in sub-Saharan Africa — the barrier now is delivery and cost, not efficacy.

Which brings us neatly to two health economic analyses, both in Clinical Infectious Diseases, that make the financial case for regimens we already know work clinically. Reed and colleagues analysed resource use inside the DOTS randomised trial of two doses of dalbavancin versus standard therapy in 200 adults with complicated Staphylococcus aureus bacteraemia after initial clearance [4]. Patients given dalbavancin received antibiotics on roughly 24 fewer days and spent about two fewer days in hospital for bacteraemia-related care. Drug acquisition cost more — a little over two thousand dollars per patient — but savings on inpatient care and outpatient parenteral therapy fully offset that, with net savings of around six thousand dollars per patient in the home-based outpatient scenario. Among patients who inject drugs, savings approached twenty-two thousand dollars. So the standard objection to dalbavancin, that it is too expensive, does not survive contact with the full cost picture.

The same logic applies to cryptococcal meningitis. Clark and colleagues performed a retrospective economic evaluation at a large academic public hospital in the United States comparing the AMBITION protocol — single high-dose liposomal amphotericin plus two weeks of flucytosine and fluconazole — against guideline-compliant daily lipid amphotericin and flucytosine, in 60 patients with HIV-associated cryptococcal meningitis [5]. Net savings to the hospital came to just under four hundred thousand dollars overall, roughly twelve thousand dollars per patient, after accounting for lost revenue, and modelling with a nationwide payor mix for advanced HIV still showed savings from both hospital and payor perspectives. Notably, no patients in the cohort had commercial insurance. Given that AMBITION is also better tolerated, this strengthens the case for adopting it in high-income settings, not just resource-limited ones.

The third theme is carbapenem resistance, where this week brings one sobering trial and one mechanistic study. In the International Journal of Antimicrobial Agents, Teo and colleagues randomised 102 patients with carbapenem-resistant Gram-negative infections to therapy guided by in vitro antibiotic combination testing versus standard therapy [3]. The trial was stopped early — not for futility in the usual sense, but because nearly two thirds OF PATIENTS assigned to the control arm crossed over at physician request. Thirty-day mortality was identical at 33 percent in both arms. An exploratory comparison of patients who actually received combination-testing-matched therapy suggested a large mortality advantage, but with only seven patients in the non-matched group this is hypothesis-generating at best. The honest reading is that the randomised comparison showed no benefit, and that clinician enthusiasm outran the evidence.

Alongside that, the Journal of Infectious Diseases published laboratory work from Baek and colleagues on Enterobacterales resistant to ertapenem alone [6]. Serial exposure to subinhibitory meropenem rapidly drove full carbapenem resistance through porin loss — loop-region ompC mutations in E. coli, disruptive ompK36 mutations in Klebsiella — while Enterobacter cloacae rose in meropenem minimum inhibitory concentration by a non-porin route. Adding colistin to meropenem significantly delayed resistance emergence. It is bench work, but it gives a mechanistic rationale for why ertapenem-mono-resistant isolates deserve respect rather than casual meropenem use.

Finally, diagnostics, where the news is mixed. In Clinical Infectious Diseases, Agudelo and colleagues applied untargeted shotgun metagenomic sequencing to 176 stool samples from Ugandan children being assessed for pulmonary tuberculosis, median age under four years [7]. Against a microbiological reference standard, sensitivity at the most permissive threshold was only about 36 percent, and in head-to-head comparison stool metagenomics performed worse than stool Xpert Ultra. Specificity was high, but this is a negative result for sequencing as a frontline paediatric tuberculosis test — its value here was identifying ribosomal RNA, virulence protein and membrane protein loci not covered by current PCR platforms, which could inform better assays. Contrast that with the Journal of Infectious Diseases report from Shiha and colleagues on SMART-C, a lateral-flow hepatitis C antigen test using a thermo-responsive smart polymer to concentrate antigen [9]. Across more than 1,350 paired samples from Egypt, Pakistan and Japan, sensitivity was just under 95 percent and specificity just over 91 percent against PCR, with a limit of detection below 10 international units per millilitre and antigen decline tracking viral clearance on direct-acting antivirals. For decentralised hepatitis C elimination programmes, a point-of-care test approaching PCR sensitivity is genuinely significant. Rounding out the diagnostics, Open Forum Infectious Diseases published a 19-case series from De La Hoz and colleagues in which plasma microbial cell-free DNA sequencing was used for chronic disseminated candidiasis in haematologic malignancy, influencing antifungal management in most cases [8]; and in the same journal, Birabaharan and Cowell use a case of Mycobacterium bovis meningitis near the United States-Mexico border to argue for early adjunctive central-nervous-system-penetrating agents where M. bovis, intrinsically pyrazinamide-resistant, is endemic [10].

If you only have time for one paper this week, make it SOLARIO in the New England Journal of Medicine [1]. A well-powered noninferiority trial that cuts systemic antibiotic exposure from weeks to days in orthopaedic infection, with a large reduction in treatment-related symptoms, is the kind of result that changes bone and joint infection protocols immediately where local antibiotic carriers are in use.

Here are the key takeaways from this week in Infectious Disease. First, after surgery for orthopaedic infection with a local antibiotic carrier implanted, seven days of systemic antibiotics was noninferior to four weeks or more, with far fewer treatment-related symptoms. Second, eight-weekly injectable cabotegravir-rilpivirine was superior to daily oral therapy for maintaining suppression in African adolescents living with HIV. Third, both dalbavancin for staphylococcal bacteraemia and the AMBITION protocol for cryptococcal meningitis pay for themselves — cost should no longer be the reason you don't use them. Fourth, combination susceptibility testing to guide therapy for carbapenem-resistant Gram-negatives showed no mortality benefit in a randomised comparison, so treat the enthusiasm with caution. And fifth, stool metagenomics did not outperform Xpert Ultra for paediatric tuberculosis, while a new polymer-based lateral-flow hepatitis C antigen test came close to PCR performance in field settings.

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 a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

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