This Week in Oncology — Jun 18, 2026

Welcome to This Week in Oncology. This week we're covering 9 notable papers spanning major advances in multiple myeloma, the expanding role of artificial intelligence and proteomics in diagnostics, and key updates in solid tumor management. Let's dive in.

We begin this week in multiple myeloma, where two major phase 3 trials published in The New England Journal of Medicine and The Lancet are set to change the treatment landscape for relapsed or refractory disease. First, in The New England Journal of Medicine, the MonumenTAL-3 trial evaluated talquetamab, a bispecific antibody targeting GPRC5D and CD3, in combination with daratumumab, with or without pomalidomide, against the standard daratumumab, pomalidomide, and dexamethasone triplet [8]. In patients who had received at least one prior line of therapy, both talquetamab-containing regimens led to significantly longer progression-free survival. At a median follow-up of about two years, the 24-month estimated progression-free survival was around 81% for the talquetamab-daratumumab-pomalidomide group and 78% for the talquetamab-daratumumab group, compared to just 51% for the standard care arm. This translated into a significant overall survival benefit as well, with hazard ratios for death around 0.5. The rates of complete response or better were more than doubled with the talquetamab regimens. Serious adverse events were common across all groups, occurring in over half the patients.

Complementing this, a study in The Lancet reports on the SUCCESSOR-2 trial, which tested a different mechanism in a similar patient population [9]. This trial evaluated mezigdomide, a potent next-generation cereblon E3 ligase modulator, added to carfilzomib and dexamethasone. The study enrolled patients with relapsed or refractory myeloma who were exposed to both an anti-CD38 antibody and lenalidomide, a challenging-to-treat group. The results were striking: adding mezigdomide more than doubled the median progression-free survival from 8.3 months in the control group to 18.0 months in the investigational arm. This represents a greater than 50% reduction in the risk of progression or death. This benefit came at the cost of increased toxicity, with grade 3 or 4 adverse events, particularly neutropenia and infections, being substantially more common in the mezigdomide group. Together, these two trials provide powerful new options for patients with relapsed myeloma, targeting distinct pathways with impressive efficacy.

Shifting gears, a cluster of papers this week highlights how artificial intelligence and high-throughput proteomics are moving from research tools to potential clinical game-changers. In The Lancet Oncology, a study details the development of ARTIMES, an AI-assisted volumetric response criterion for pleural mesothelioma [2]. Given the tumor's crescent-shaped growth, standard diameter-based criteria like mRECIST are often inadequate. The ARTIMES model, trained on over 10,000 CT scans, demonstrated superior prognostic performance compared to mRECIST and, critically, detected disease progression a median of 5 weeks earlier. Furthermore, progression-free survival as measured by ARTIMES showed a much stronger correlation with overall survival, suggesting it could be a more reliable surrogate endpoint in clinical trials. This work could facilitate more reliable and timely response evaluation for these patients. Taking a much broader view, a paper in Nature introduces MIRA, an autonomous artificial intelligence agent designed to operate within an electronic health record [1]. In a sandboxed simulation environment using real patient cases, MIRA could take a history, order tests, generate differential diagnoses, and formulate treatment plans. In these simulations, it actually outperformed physicians in diagnostic accuracy and made guideline-concordant decisions. While extensive real-world validation is needed, this provides a proof-of-concept for a more integrated and actionable form of clinical decision support.

Beyond imaging and workflow, two papers explore the power of plasma proteomics for prediction. In Science Translational Medicine, researchers developed a predictive model for venous thromboembolism in patients with newly diagnosed lung or gastric cancer [4]. Using over a thousand plasma proteins and five clinical parameters, their machine learning model vastly outperformed the standardly used Khorana score for predicting VTE. The model not only improves prediction but also offered mechanistic insights, identifying the IL-17 pathway as a potential therapeutic target for thrombo-inflammation. In a similar vein, a large-scale study in Nature Medicine used plasma proteomics from over 60,000 individuals to create machine learning models that estimate the biological age of over 40 different cell types [6]. These cellular aging signatures were highly predictive of future disease. For example, extreme astrocyte aging tripled the risk of incident Alzheimer's Disease in APOE4 carriers, while extreme respiratory epithelial cell aging was associated with a 58% higher lung cancer risk in smokers. This provides a novel framework for quantifying health at a cellular level and identifying specific vulnerabilities long before disease manifests.

Finally, we'll look at key updates in solid tumor management. The most significant is the AENEAS2 trial, published in The Lancet Oncology [7]. This phase 3 trial addressed a critical question in first-line therapy for advanced non-small-cell lung cancer with EGFR-sensitive mutations. Patients were randomized to the third-generation TKI aumolertinib alone, or aumolertinib plus platinum-based chemotherapy. The results were definitive: the combination therapy significantly improved progression-free survival, extending the median from 18.9 months with monotherapy to 28.9 months with the combination. This 10-month benefit represents a major step forward, though it was associated with substantially higher rates of grade 3-4 hematologic toxicities like neutropenia. For a broader perspective on modern management, a review in The New England Journal of Medicine provides a clinical framework for decision-making in differentiated thyroid cancer [3]. It emphasizes a risk-adapted, dynamic process that spans the entire clinical course, from active surveillance for low-risk papillary microcarcinomas to minimalist surgical options and systemic therapies for advanced disease, highlighting the importance of shared decision-making. Lastly, a proof-of-concept study in Science Translational Medicine offers a glimpse into the future of immunomodulation [5]. In a small phase 1/2a trial in kidney transplantation, a combination of donor bone marrow and recipient regulatory T cells induced donor chimerism without the need for toxic recipient irradiation, a long-sought goal in transplantation that could have future implications for hematology and oncology.

If you only have time for one paper this week, make it the AENEAS2 trial in The Lancet Oncology [7]. It establishes aumolertinib plus chemotherapy as a new standard of care in first-line EGFR-mutated non-small-cell lung cancer, demonstrating a substantial 10-month improvement in progression-free survival.

Here are the key takeaways from this week in Oncology: First, in first-line treatment for EGFR-mutated non-small-cell lung cancer, adding chemotherapy to the third-generation TKI aumolertinib extends median progression-free survival by 10 months, establishing a new standard of care, as shown in the AENEAS2 trial [7]. Second, two major phase 3 trials are reshaping treatment for relapsed or refractory multiple myeloma. The bispecific antibody talquetamab combined with daratumumab significantly improves progression-free and overall survival [8], while the next-generation CELMoD mezigdomide more than doubles progression-free survival when added to carfilzomib and dexamethasone [9]. Third, AI-driven tools are showing clinical utility. In pleural mesothelioma, an AI-based volumetric response criterion called ARTIMES outperforms mRECIST for predicting survival and detects progression five weeks earlier [2]. And finally, plasma proteomics is emerging as a powerful tool for risk stratification. A new proteomic model substantially outperforms the Khorana score for predicting venous thromboembolism in cancer patients [4], and proteomic signatures of cellular aging can predict incident disease risk, including for lung cancer [6].

That's your roundup for This Week in Oncology. 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.