This Week in General Medicine — Sep 7, 2026

Welcome to This Week in General Medicine. This week we're covering 10 notable papers spanning thoracic oncology and lung nodule management, transplant and haematology innovation, and the management of chronic multisystem disease. Let's dive in.

We'll start in the chest, because four of this week's papers touch the lung. The New England Journal of Medicine published a clinical practice review on pulmonary nodules from Callister and Silvestri that is worth reading in full if you order chest imaging [1]. The core framework is the distinction between solid and subsolid nodules, with subsolid further split into part-solid and pure ground-glass, because management diverges sharply. Comparison with any prior imaging is described as essential. A solid nodule stable for two years can be called benign; subsolid nodules need a longer period of stability before you can say the same, because they grow slowly, yet they carry a higher chance of being malignant, particularly if a solid component appears or steadily enlarges. For solid nodules, risk prediction models guide the pathway: computed tomographic surveillance for low-risk lesions, positron-emission tomography with computed tomography and biopsy for intermediate risk, and surgery for selected high-risk lesions, with transthoracic needle biopsy and navigational bronchoscopy as the biopsy options. The framing the authors return to is the balance every generalist recognises: diagnosing the cancers promptly without subjecting people with benign disease to invasive procedures.

Once a nodule turns out to be cancer, two papers this week shape what comes next. In The Lancet, the SWOG/NRG S1914 phase 3 trial led by Daly asked whether adding atezolizumab to stereotactic body radiation therapy improves survival in medically inoperable, high-risk early-stage non-small-cell lung cancer [2]. This is an important negative result. Across 146 United States institutions, just over 400 eligible patients were randomised, and accrual was stopped at the first interim analysis for futility. On updated analysis, there was no overall survival benefit whatsoever, with estimated two-year survival of 82 percent in both groups, and grade 3 or higher adverse events were four times more common with the addition of immunotherapy, at 12 percent versus 3 percent, including two fatal respiratory events. So stereotactic radiation alone remains the standard, and adding checkpoint blockade in this setting adds harm without measurable gain. More encouraging, in the New England Journal of Medicine, Arbour and colleagues reported a phase 1-2 study of daraxonrasib, an oral multiselective RAS inhibitor, in previously treated advanced RAS-mutant non-small-cell lung cancer, a group that accounts for roughly 30 percent of cases [3]. Among 136 patients treated at 300 milligrams daily or less, more than 30 percent had an objective tumour response, but toxicity was substantial: rash, diarrhoea, nausea, vomiting and mucositis were each common, and just over half of patients had grade 3 or higher adverse events, with four deaths attributed to adverse events. This is early-phase data, not a practice change, but it signals that pan-RAS targeting is clinically active. Rounding out the theme, The Lancet also published a Series review from Ross and colleagues on the shifting treatment landscape in small-cell lung cancer, where T-cell engagers, antibody-drug conjugates, radioconjugates and cell therapies directed at differentially expressed surface proteins are moving into a disease that has long been refractory after first relapse [4].

Our second theme is where established treatment paradigms are being displaced. The most immediately consequential is the PARADIGM trial in the New England Journal of Medicine, reported by Fathi, Perl and colleagues, which randomised 172 previously untreated adults with acute myeloid leukaemia who were fit and eligible for intensive induction chemotherapy to either standard induction or azacitidine plus venetoclax [5]. That combination has been the treatment for patients deemed unfit; this trial tested it in the fit population, excluding core binding factor fusions and FLT3 mutations, and NPM1 mutations under age 60. Median age was 64, and nearly three quarters of participants had adverse-risk disease. Median event-free survival was 14.5 months with azacitidine-venetoclax against 6.2 months with induction chemotherapy, a reduction in the risk of an event or death of roughly 40 percent. Toxicity also favoured the oral-plus-injectable regimen, with grade 3 or higher infection in about 28 percent versus 41 percent, and severe haemorrhage in 2 percent versus 12 percent. It's a phase 2 trial and event-free survival is not overall survival, but for physicians co-managing these patients this challenges the assumption that fitness for induction means induction is best. Alongside that, The Lancet published a first-in-human report from Riella and colleagues of a gene-edited porcine kidney used as a bridge to human transplantation [6]. The recipient had end-stage kidney disease, no living donor and a long expected wait. The xenograft worked immediately and sustained dialysis independence for 271 days. There was T-cell-mediated rejection on a day-14 biopsy that resolved with treatment, then, after immunosuppression was reduced during a bacterial infection, microvascular inflammation progressed to thrombotic microangiopathy and the graft was removed, with an infiltrate dominated by macrophages and natural killer cells rather than T cells and a persistently negative donor-specific crossmatch. Crucially, no porcine pathogen transmission was detected, anti-HLA antibodies did not change, and 82 days after explantation the patient received a human kidney with immediate function and no evidence of sensitisation over 231 days.

The third theme is chronic disease that generalists see and often under-treat. In The Lancet, Podda and colleagues report a multicentre German randomised trial of liposuction versus continued conservative decongestive therapy in 410 women with lipoedema and leg pain of at least four out of ten [7]. After twelve months, just over two thirds of the women assigned to liposuction achieved the primary endpoint of at least a two-point reduction in leg pain, compared with fewer than one in ten of those continuing conservative therapy, with quality of life also better. That is a large effect in a condition where standard care offers limited relief, but the trade-off is real: adverse events occurred in nearly half of the surgical patients and 8 percent had a serious adverse event, versus 3 percent with conservative care. Thirty-six month follow-up is ongoing. Turning to the liver, Nature published a review from Gallage and colleagues on metabolic dysfunction-associated steatohepatitis, and the practical message is that pharmacotherapy has arrived, with accelerated approval of the thyroid hormone receptor-beta agonist resmetirom and of semaglutide for non-cirrhotic disease with fibrosis [8]. The authors emphasise that this is a systemic illness driving cardiovascular disease and chronic kidney disease as well as cirrhosis and hepatocellular carcinoma, and that distinct liver-centric and cardiometabolic phenotypes probably require stratified, and eventually combination, therapy on top of lifestyle intervention. In neurology, a multicentre study in The Lancet from Maure-Blesa and colleagues followed 4804 adults with Down syndrome and mapped the emergence of epilepsy [9]. Epilepsy prevalence rose steeply with age and above all with symptomatic Alzheimer's disease: cumulative incidence climbed from about one in ten at the time of Alzheimer's diagnosis to roughly 57 percent nine years later, mostly myoclonic or tonic-clonic seizures. Risk rose with each year since diagnosis, with severe-to-profound intellectual disability, and with APOE epsilon-4 carriage. Late-onset myoclonic epilepsy in Down syndrome roughly doubled mortality and was accompanied by faster cognitive decline, and interictal electroencephalogram showed little diagnostic utility, so this remains a clinical diagnosis that depends on asking carers directly about myoclonic jerks. Finally, Nature Reviews Disease Primers published a comprehensive primer on varicella zoster virus from Bubak and colleagues, a useful refresher given that the virus infects more than 90 percent of people worldwide and that reactivation can produce vasculopathy, cranial neuropathies, myelopathy and cardiovascular or gastrointestinal disease without any rash at all, with valacyclovir the oral drug of choice and both the live attenuated varicella vaccine and the recombinant zoster vaccine central to prevention [10].

If you only have time for one paper this week, make it the New England Journal of Medicine review of pulmonary nodules [1]. Almost every general physician receives incidental nodule reports, and this gives you a defensible, current framework for who needs surveillance, who needs a biopsy, and who can be reassured.

Here are the key takeaways from this week in General Medicine. First, for incidental pulmonary nodules, compare with prior imaging, distinguish solid from subsolid, and remember that two years of stability reassures you only for solid lesions. Second, adding atezolizumab to stereotactic radiation in inoperable early-stage lung cancer did not improve survival and increased serious toxicity, so do not extrapolate immunotherapy benefit into this setting. Third, in fit adults with acute myeloid leukaemia outside favourable genetic subgroups, azacitidine plus venetoclax more than doubled median event-free survival compared with intensive induction, with less severe infection and bleeding. Fourth, liposuction substantially outperformed conservative therapy for lipoedema pain and quality of life, but with a higher rate of serious adverse events that must be part of the consent conversation. And fifth, in adults with Down syndrome, once Alzheimer's disease becomes symptomatic, epilepsy should be actively anticipated, because it emerges in the majority within a decade and signals worse survival and faster decline.

That's your roundup for This Week in General Medicine. 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.