This Week in Oncology — Sep 23, 2026

Welcome to This Week in Oncology. This week we're covering 10 notable papers spanning first-line advances and immune biology in thoracic cancers, a run of breast cancer studies that challenge how we operate and how we combine targeted agents, and a broader set on vaccines, biomarker-driven melanoma therapy, and the health-system side of cancer care. Let's dive in.

We'll start with lung cancer. Cancer Discovery reports on three phase III readouts presented at the World Conference on Lung Cancer that reshape first-line decisions in advanced non-small cell lung cancer [1]. Two trials, PAPILLON and REZILIENT3, established amivantamab and zipalertinib combination regimens as effective first-line strategies for tumours carrying EGFR exon 20 insertions, both improving progression-free survival. Importantly, neither demonstrated an overall survival benefit, and the reason given is high crossover — a familiar pattern that should temper how we counsel patients about survival expectations while still supporting these combinations as front-line options in a genotype that has long been under-served. Separately, HARMONi-2 showed that the bispecific antibody ivonescimab improved overall survival compared with pembrolizumab in PD-L1-positive disease, which is enough to establish it as standard of care in China while larger global trials continue. For clinicians outside that setting, the practical message is to watch this space rather than change practice yet.

Staying in the chest, Cancer Cell offers a mechanistic explanation for why small cell lung cancer responds so poorly to checkpoint blockade, and a possible way around it [8]. Ng and colleagues used single-cell profiling of patient specimens and found substantial infiltration by gamma-delta T cells, which recognise tumours independently of MHC class one — the very molecule that is epigenetically silenced in small cell lung cancer. Despite expressing PD-1, these cells kept a cytotoxic transcriptional profile, and in two practice-changing clinical trial cohorts, high gamma-delta T cell infiltration predicted better response to anti-PD-L1 therapy. In preclinical models, those cells were effective mediators of tarlatamab-redirected killing, and zoledronate — a drug already on our formularies — sensitised tumour cells to gamma-delta-mediated killing. This is hypothesis-generating rather than practice-changing, but a repurposing trial of zoledronate alongside a DLL3 bispecific is an obvious next step.

Turning to breast cancer, where the most immediately actionable paper of the week comes from the Journal of Clinical Oncology. Nash and colleagues analysed more than eleven thousand women treated with breast-conserving surgery for ductal carcinoma in situ, using patient-level data from the National Cancer Database Special Study, with a median follow-up of about five and a half years [3]. The current two-millimetre margin recommendation rests on study-level, not patient-level, data. Here, only about four percent of women recurred, and after multivariable adjustment there was no significant difference in ipsilateral recurrence between negative margins under two millimetres and margins of two millimetres or more — and that held whether or not radiotherapy was given. The one exception was women under fifty at diagnosis, where narrow margins did carry a higher recurrence risk. The clinical implication is direct: for an older woman with a close but negative margin, routine re-excision may be adding morbidity without benefit, and this is retrospective data that should feed a real conversation in the multidisciplinary meeting rather than an automatic return to theatre.

Also in breast cancer, Clinical Cancer Research published the VIOLETTE trial, in which Tutt and colleagues randomised 273 patients with previously treated, PARP-inhibitor-naive advanced triple-negative disease to olaparib alone, olaparib plus the ATR inhibitor ceralasertib, or olaparib plus the WEE1 inhibitor adavosertib, stratified by BRCA and homologous recombination repair status [4]. The adavosertib arm was terminated early for unacceptable toxicity. The primary endpoint was negative: adding ceralasertib did not significantly improve progression-free survival in any of the three molecular strata, with median progression-free survival ranging from about seven and a half months in the BRCA-mutant group down to under four months in the others. The one positive signal was in tumours without homologous recombination repair mutations, where the response rate rose from about four percent with olaparib alone to about fifteen percent with the combination. As the authors themselves conclude, that response advantage has limited clinical significance without a progression-free survival gain. For now, ATR inhibition remains investigational in triple-negative disease.

Complementing those data, Annals of Oncology published a review from Licata and colleagues on adjuvant therapy in premenopausal women with hormone receptor-positive, HER2-negative early breast cancer [2]. Their argument is that this group is clinically distinct — more aggressive biology, different survivorship concerns — yet much of our evidence is extrapolated from trials dominated by postmenopausal women. Endocrine therapy with CDK4-6 inhibitors in higher-risk patients remains the backbone, but genuine uncertainty persists around who needs ovarian function suppression, how to interpret genomic assays for chemotherapy decisions, who benefits from extended endocrine therapy, and how to manage adherence, toxicity, and pregnancy. They propose a pragmatic framework, and it is a useful read before your next young-patient clinic.

The third theme is immunotherapy — both broadening it and better targeting it. JAMA published a translational science review from Disis and colleagues on cancer vaccines, arguing that the field has genuinely turned a corner after decades of disappointment [5]. More than two thousand immunogenic tumour antigens have now been identified, and delivery technology has matured. They anchor the case in approved agents — sipuleucel-T, which reduced all-cause mortality in metastatic castration-resistant prostate cancer, and talimogene laherparepvec, which produced durable responses in injected melanoma lesions at roughly eight times the rate of the comparator. Most striking is the combination data: adding the personalised messenger RNA vaccine mRNA-4157 to pembrolizumab in resected high-risk stage three to four melanoma improved relapse-free survival to about 77 percent versus 60 percent with pembrolizumab alone, an effect that sat right at the edge of statistical significance in a randomised phase two setting. Toxicity is generally mild — fatigue, fever, injection site reactions — which is what makes integration into standard regimens, and eventually primary prevention, plausible.

On the targeting side, the Journal of Clinical Oncology published work from Shah and colleagues integrating clinical and multi-omic data from IMspire150, BRIM, AVAST-M, and public immunotherapy cohorts in BRAF-mutant melanoma [6]. They identified two distinct checkpoint-resistant transcriptional states. Differentiated tumours did poorly with checkpoint blockade alone but did better when BRAF and MEK inhibition was added — a rationale for escalation in exactly the patients where the trial literature has been mixed. Undifferentiated tumours responded poorly to both, were enriched for angiogenic and stromal programmes, appeared to do better with anti-VEGF therapy in retrospective analysis, and were sensitive to CDK7 inhibition preclinically. This is a biomarker framework, not yet a test you can order, but it explains why the combination trials have looked equivocal when all patients are pooled.

Two papers this week address the system around the tumour. In The Lancet, Rebbeck lays out a framework for precision prevention and early detection [7], noting that only about 13 percent of cancers are currently detected through guideline-based screening while somewhere between a fifth and half of cancers first present in an emergency department. The argument is that de-escalating prevention in genuinely low-risk people matters as much as intensifying it in high-risk people, and that the binding constraint over the next decade is health-system infrastructure and workforce, not science. Alongside that, the Journal of Clinical Oncology published a population-based Ontario study from Hirpara and colleagues linking nearly 1.8 million symptom assessments from about 286,000 patients to health care costs [10]. Higher symptom burden was associated with lower spending on cancer-directed treatment — roughly a tenth less per ten-point rise in the symptom score, because these patients received less therapy — and simultaneously with nearly a fifth more spending on ancillary services such as emergency visits and home care. In other words, uncontrolled symptoms push patients away from the treatment that might help them and toward unplanned care. Proactive symptom management is not just supportive care; it is what keeps patients on therapy.

One more worth flagging: Nature Medicine reports that levetiracetam may be more than an antiseizure drug in diffuse midline glioma [9]. Barron and colleagues found longer overall survival in children with diffuse midline glioma taking levetiracetam, an effect absent in hemispheric high-grade glioma, and reproduced that subtype specificity in mouse models. Mechanistically, levetiracetam dampened GABAergic neuron-to-glioma synaptic transmission in a tumour-specific way, independent of its usual SV2A target. Prospective study is needed, but if a child with diffuse midline glioma needs an antiseizure medication, this informs the choice.

If you only have time for one paper this week, make it the DCIS margin analysis in the Journal of Clinical Oncology [3]. It is the one study here that could change a decision you make in the next multidisciplinary meeting — whether to send a woman with a close but negative margin back to the operating room.

Here are the key takeaways from this week in Oncology. In EGFR exon 20 insertion lung cancer, amivantamab and zipalertinib combinations improve progression-free survival first-line, but crossover means no proven survival benefit — counsel accordingly. In ductal carcinoma in situ, margins under two millimetres did not predict higher recurrence except in women under fifty, so routine re-excision deserves scrutiny. ATR inhibition added to olaparib in triple-negative breast cancer failed its primary endpoint and remains investigational. Cancer vaccines now have credible combination data in melanoma and are moving toward prevention. And symptom burden is a value problem as much as a comfort problem — poorly controlled symptoms mean less cancer treatment and more emergency care.

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 think of one colleague — in any specialty — who never has time to keep up with the literature. Tell them about AudioScholar: a free ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.

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