This Week in General Medicine — Jun 16, 2026

Welcome to This Week in General Medicine. This week we're covering 10 notable papers spanning advances in oncology, new frontiers in diabetes care, and a critical look at when more technology or treatment isn't better. Let's dive in.

We begin this week in oncology, with two phase 3 trials from The Lancet offering new hope for difficult-to-treat cancers. First, in pancreatic cancer, the CASSANDRA trial evaluated a new preoperative chemotherapy regimen for patients with resectable or borderline resectable disease [9]. Patients were randomized to receive either PAXG—a combination of cisplatin, nab-paclitaxel, capecitabine, and gemcitabine—or the standard mFOLFIRINOX. The results of this first randomization showed that PAXG significantly prolonged median event-free survival to 16.0 months, compared with 10.2 months for mFOLFIRINOX. While grade 3 or worse adverse events were common in both groups, occurring in about two-thirds of patients receiving PAXG, the significant improvement in survival suggests that preoperative PAXG could be considered a new standard of care and the comparator for future trials in this setting. Shifting to hematologic malignancy, the SUCCESSOR-2 trial addressed a growing challenge in relapsed or refractory multiple myeloma: patients who have already been exposed to both anti-CD38 antibodies and lenalidomide [8]. This study tested the addition of mezigdomide, a potent cereblon E3 ligase modulator, to a regimen of carfilzomib and dexamethasone. At a median follow-up of 10.6 months, adding mezigdomide more than doubled the median progression-free survival, from 8.3 months in the control group to 18.0 months. This benefit came at the cost of increased toxicity; grade 3 or 4 adverse events occurred in 84% of the mezigdomide group versus 56% of the control group, with notable increases in neutropenia and infections. Still, for this heavily pretreated population, the results represent a major step forward.

Turning to diabetes, we have a trio of papers covering a new oral agent for type 2, an AI-powered screening tool for complications, and population screening for type 1. In The Lancet, the ATTAIN-2 trial reports on orforglipron, a once-daily, oral, non-peptide GLP-1 receptor agonist for adults with obesity and type 2 diabetes [2]. This 72-week trial demonstrated dose-dependent weight loss, with the highest dose achieving a mean bodyweight reduction of 9.6%, compared to just 2.5% with placebo. All cardiometabolic measures, including HbA1c, also significantly improved. The safety profile was consistent with the GLP-1 agonist class, with mild-to-moderate gastrointestinal events being most common, particularly during dose escalation, leading to treatment discontinuation in up to 9.9% of patients on the highest dose. For managing complications, a study in JAMA explored an AI-based optical coherence tomography system to improve screening for diabetic macular edema [3]. The current standard, fundus photography, generates a high rate of false-positive referrals. This randomized trial compared standard referral versus a two-step process where referrals from fundus photos were secondarily screened with the AI-OCT system. The intervention was a clear success: it slashed the false-positive referral rate from 69% in the control group to just 24%, without missing a single case of diabetic macular edema. This demonstrates how AI can be integrated into clinical pathways to substantially reduce unnecessary specialist referrals and healthcare burdens. Broadening the scope to early detection, another large study in JAMA reports on population-based screening for early-stage type 1 diabetes in over 220,000 children in Germany [5]. Using islet autoantibody testing, the study identified an adjusted population frequency of 0.3% for presymptomatic, early-stage disease. During a median follow-up of 5.7 years, the five-year progression rate to clinical, or stage 3, diabetes was 36.2%. Critically, this progression risk was not significantly different between children with and without a first-degree family history of the disease. These findings support the idea that screening for type 1 diabetes could be considered for the general pediatric population, not just those with known genetic risk.

Next, we have two important trials that challenge current or proposed practices, reminding us that more intervention is not always better. First, from The New England Journal of Medicine, the OPTIMAL trial investigated whether using intravascular ultrasound, or IVUS, to guide percutaneous coronary intervention in patients with unprotected left main coronary artery disease improves outcomes compared to standard angiography guidance alone [7]. In this multicenter trial of over 800 patients, at a median follow-up of 2.9 years, there was no additional benefit from IVUS guidance. The primary patient-oriented composite endpoint of any stroke, myocardial infarction, revascularization, or death occurred in 33.7% of the IVUS group and 30.9% of the angiography group, a non-significant difference. This result suggests that for this high-risk procedure, the routine addition of IVUS does not improve major clinical outcomes. In critical care, the VICTORY randomized trial, published in JAMA, provides a definitive and cautionary result regarding the use of high-dose intravenous vitamin C in patients with severe burn injuries [6]. The trial was designed to see if vitamin C could mitigate systemic inflammation and organ dysfunction. However, it was stopped early at a prespecified interim analysis for futility and potential harm. The primary composite outcome of 28-day mortality and persistent organ dysfunction occurred more frequently in the vitamin C group than the placebo group. More alarmingly, 28-day mortality was significantly higher with vitamin C, at 15.0% versus 7.6% with placebo, a risk that was nearly doubled. This trial provides strong evidence that high-dose intravenous vitamin C should not be used in this population and may be harmful.

Finally, we look to the future with a first-in-class gene editing therapy and a head-to-head comparison of AI models in medicine. A landmark phase 3 trial in The New England Journal of Medicine details the use of lonvoguran ziclumeran, an in vivo CRISPR-based gene-editing treatment for hereditary angioedema [1]. Patients received a single intravenous infusion of the treatment or placebo. The results were striking: the monthly rate of debilitating angioedema attacks was reduced by 87% in the treatment group compared to placebo. While infusion-related reactions were common, no serious or grade 3 or higher adverse events were reported in the lonvo-z group, offering the potential for a single-dose, durable treatment for this rare genetic disease. And in a fascinating commentary on the state of artificial intelligence in medicine, a study in Nature Medicine directly compared the performance of specialized clinical AI tools against general-purpose frontier large language models like GPT-5.2 [4]. The evaluation was rigorous, using medical knowledge questions, alignment with clinician judgment, and a benchmark of 100 real clinical queries from physicians. Across all three evaluations, the general-purpose LLMs consistently outperformed the specialized clinical AI tools. These findings underscore the critical need for independent, real-world evaluation of all AI tools before they are integrated into clinical settings.

If you only have time for one paper this week, make it the VICTORY trial on high-dose vitamin C for severe burns, published in JAMA [6]. This trial was stopped early for harm, showing a significant increase in mortality and providing clear, definitive evidence against a proposed therapy.

Here are the key takeaways from this week in General Medicine. First, in resectable pancreatic cancer, preoperative PAXG improves event-free survival over mFOLFIRINOX and should be considered a new standard of care [9]. Second, for patients with obesity and type 2 diabetes, the oral GLP-1 agonist orforglipron offers significant weight loss and glycemic control, expanding our oral treatment options [2]. Third, stop using high-dose intravenous vitamin C for severe burn injuries; the VICTORY trial shows it does not improve outcomes and is associated with increased mortality [6]. Fourth, in diabetic retinopathy screening, adding an AI-OCT system can dramatically and safely reduce unnecessary referrals for diabetic macular edema without missing cases [3]. And finally, a crucial reminder from a review in The Lancet on postpartum hemorrhage: successful management is a race against time. The key is avoiding delays in diagnosis through objective blood loss measurement and in initiating bundled treatments [10].

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 for audio briefings on your own clinical questions and papers, visit audioscholar dot C C.

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