This week in General Medicine

This AudioScholar briefing is generated by artificial intelligence for healthcare professionals and trainees. It is not medical advice.

Welcome to This Week in General Medicine. This week we're covering 10 notable papers spanning acute vascular care from the stroke unit to the emergency department, diagnosis and assessment in everyday clinic medicine, and treatment frontiers in cancer, infection, and fertility. Let's dive in.

We begin with stroke, where two randomised trials in the New England Journal of Medicine, both from China, push endovascular therapy into groups where its value has been uncertain. The first, led by Liu, tackled mild stroke, meaning patients with a National Institutes of Health Stroke Scale score under 6, but with an anterior circulation large-vessel occlusion, a small infarct core, and a large perfusion mismatch, treated within 24 hours [1]. Among 300 patients in the main analysis, an excellent outcome at 90 days, meaning little or no disability, was reached by about 69 percent of patients given thrombectomy plus medical management, against exactly half of patients given medical management alone, roughly doubling the odds of a near-normal recovery. Symptomatic bleeding was identical, three patients in each arm, but serious procedural complications affected about 13 percent of treated patients, a figure the authors themselves flag as high. The second trial, led by Yang, stretched the time window to 24 to 72 hours, with a median of about 38 hours from last known well, in patients with imaging-identified salvageable tissue [2]. Thrombectomy shifted the whole disability distribution toward better function, with odds of a better outcome raised by roughly three quarters, but this came at a real cost: symptomatic haemorrhage in about 5 percent of treated patients versus none of the controls, and 90-day mortality of 6 percent versus just under 2 percent. Read together, the two trials suggest that tissue-based selection may matter more than the clock or the severity score, but each is a single open-label trial from one country with imaging-heavy selection, so how far either result generalises, and whether the harms in the late window are acceptable, remains open.

Staying with acute cardiovascular care, two papers question how much a newer approach actually delivers. In The Lancet, Boeddinghaus and colleagues pooled individual patient data from all five randomised trials comparing a 0/1 hour troponin pathway, which current guidelines recommend, with the traditional 0/3 hour pathway [3]. Across more than 110,000 presentations, death or new type 1 myocardial infarction at 30 days was rare with both, under one percent versus about one percent, and the pooled safety estimate was imprecise. More strikingly, the faster pathway did not shorten emergency department stays, which sat at a little over five hours in both arms, nor did it clearly raise direct discharge, and results varied widely between trials. The authors conclude that the benefit of switching may depend heavily on the health system, a reminder that a faster test does not automatically mean faster flow. On the pacing front, the LEAP-Block trial in JAMA, led by Li, randomised about 500 patients with atrioventricular block and preserved ejection fraction expected to need frequent ventricular pacing [4]. Left bundle-branch area pacing cut the composite of death, heart failure events, or upgrade to biventricular pacing at two years from about 13 percent to about 2 percent, driven mainly by fewer heart failure events, with no significant difference in death, complications, or pacing-induced cardiomyopathy on their own. This is a single Chinese trial with modest event numbers, but it is randomised evidence supporting physiological pacing as the default in this group.

Our second theme turns to assessment in general clinic practice, where diagnosis increasingly relies on non-invasive tools and structured reasoning. The LITMUS study, also in The Lancet and led by Vali, tested a broad range of biomarkers against centrally read liver biopsy in about 1,600 adults with metabolic dysfunction-associated steatotic liver disease across Europe and the United States [5]. Single markers were generally disappointing for detecting at-risk steatohepatitis, and only a proteomic panel, SomaSignal, cleared the prespecified accuracy bar for that target. For advanced fibrosis and cirrhosis the picture was better, with several panels, including FibroMeter combined with elastography and Agile 3 plus, performing well, and a two-step approach starting with gamma-glutamyl transferase also met the threshold. With drug therapies now approved, this is the strongest head-to-head diagnostic dataset yet, though the newest proteomic tests are not widely available and the work does not itself establish a referral pathway. Alongside it, JAMA Internal Medicine offers a review by Irizarry-Alvarado and colleagues on preoperative medicine for internists, reframing the visit as more than clearance: a structured review of conditions and medications, attention to undifferentiated symptoms such as exertional breathlessness or signs of volume overload, and shared decision-making about individual risk [6]. And a Lancet seminar led by Black on irritable bowel syndrome and related disorders of gut-brain interaction notes these conditions affect around a fifth of the global population [7]. The authors emphasise a positive diagnosis built on history and the absence of alarm features, and an empathetic approach that they report improves symptoms and reduces health-care use, while acknowledging that most current treatments offer only modest benefit.

Our final theme covers therapeutic frontiers where the stakes and trade-offs are high. In the New England Journal of Medicine, the phase 3 DeLLphi-305 trial led by Pérol added tarlatamab, a bispecific T-cell engager targeting DLL3, to durvalumab maintenance in extensive-stage small-cell lung cancer that had not progressed after first-line chemoimmunotherapy [8]. At this interim analysis, median overall survival was about 21 months with the combination versus about 13 months with durvalumab alone, with the risk of death cut by more than half, and progression-free survival also improved substantially. Severe adverse events were somewhat more frequent, and cytokine release syndrome, mostly low grade, occurred in nearly half of the patients on the combination. For a disease with historically dismal survival, this is a major result, though the trial was open label and industry funded, and longer follow-up will clarify durability. Also in the New England Journal, Tamma reviews gram-negative resistance, describing how mobile genetic elements spread resistance across ESBL-producing and carbapenem-resistant Enterobacterales, difficult-to-treat Pseudomonas, and carbapenem-resistant Acinetobacter, and highlighting rapid diagnostics, targeted use of newer beta-lactam agents, source control, and careful treatment duration as the pillars of management [9]. Finally, a New Zealand trial in the same journal led by Prentice challenges guidance on unexplained infertility [10]. Four cycles of intrauterine insemination with ovarian stimulation proved not noninferior to one complete cycle of IVF, with cumulative live births at seven months of about 31 percent versus 55 percent. Yet once couples could move on to further IVF, cumulative live-birth rates converged at about 70 percent in both groups, and serious adverse events were more than twice as common in the IVF-first group. So IVF first brings pregnancy sooner, but the eventual chance of a baby looked the same, and the trade-off involves time, cost, and procedural risk, a nuance current guidelines do not yet capture.

If you only have time for one paper this week, make it the Lancet individual patient data meta-analysis of 0/1 hour versus 0/3 hour troponin pathways [3]. It reopens a question many hospitals considered settled, showing that the guideline-endorsed faster pathway has not, on average, delivered the efficiency gains that justified it.

Here is what this week's evidence adds up to in General Medicine. First, two randomised trials suggest thrombectomy can improve function in mild large-vessel stroke and in late-presenting patients with salvageable tissue, but procedural complications and, in the late window, bleeding and death are real costs, and both trials come from a single country. Second, the 0/1 hour troponin pathway appears about as safe as the 0/3 hour approach, though the safety estimate is imprecise, and it has not been shown to shorten emergency department stays. Third, single-trial evidence now favours left bundle-branch area pacing over right ventricular pacing in atrioventricular block with preserved function, mainly through fewer heart failure events. Fourth, multimarker panels outperform single tests for staging steatotic liver disease, though which pathway best uses them is unsettled. And fifth, in unexplained infertility, IVF first speeds pregnancy without improving eventual live-birth rates, while tarlatamab maintenance substantially extends survival in small-cell lung cancer at interim analysis.

That's your roundup for This Week in General Medicine. The full transcript and references are available on the episode page. 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.