This week in Oncology
Nov 18, 2026
Intismeran Autogene plus Pembrolizumab for Resected Stage IIB to IV Melanoma.
Adding the individualised neoantigen therapy intismeran to adjuvant pembrolizumab raised 24-month recurrence-free survival to about 82 percent from 70 percent in resected high-risk cutaneous melanoma, with similar severe toxicity.
New England Journal of Medicine · 2026 · PubMed
This week’s highlights
Read the weekly briefing
This AudioScholar briefing is generated by artificial intelligence from PubMed abstracts for healthcare professionals and is not medical advice.
This week: a personalised neoantigen therapy added to adjuvant pembrolizumab in melanoma, a bispecific antibody-drug conjugate extends survival in previously treated esophageal cancer, and a weekend break from dasatinib in chronic myeloid leukemia.
Welcome to This Week in Oncology. We're covering 10 papers this week: 3 in depth, 4 in brief, 3 quick signals.
In the New England Journal of Medicine, Long and colleagues report a phase 3, double-blind trial of intismeran autogene, an individualised neoantigen therapy, given with pembrolizumab after complete resection of stage IIB to IV cutaneous melanoma [1]. A total of 1137 patients were randomised two to one to intismeran at 1 milligram or placebo every three weeks for nine doses, and every patient received pembrolizumab 400 milligrams every six weeks. At a median follow-up of about two years, recurrence-free survival at 24 months was about 82 percent with the combination and 70 percent with pembrolizumab alone. Distant-metastasis-free survival also improved, to just under 89 percent against about 80 percent. The benefit was not immediate. In the first six months the two groups did about the same, and after six months the risk of recurrence was cut by more than half. Grade 3 or higher adverse events occurred in about 24 percent of patients with the combination and 20 percent with placebo. Eight patients in the combination group died of adverse events, five of them before receiving intismeran. Follow-up is short and overall survival is not reported, so the evidence rests on recurrence endpoints. The trial adds randomised phase 3 evidence that an individualised neoantigen approach can add to checkpoint blockade after melanoma resection.
JAMA published PANKU-Esophagus01, an open-label phase 3 trial led by Lu across 80 hospitals in China [2]. It enrolled 497 patients with recurrent or metastatic esophageal squamous cell carcinoma that had progressed after a first-line PD-1 or PD-L1 inhibitor with platinum chemotherapy. Patients received izalontamab brengitecan, a bispecific antibody-drug conjugate directed at EGFR and HER3 with a topoisomerase I inhibitor payload, at 2.5 milligrams per kilogram on days 1 and 8 every three weeks, or physician's choice of irinotecan, paclitaxel or docetaxel. Both dual primary endpoints were met at this interim analysis. Median progression-free survival by blinded central review roughly doubled, from about 2 months to just over 4 months. Median overall survival rose from about 7 months to just under 10 months, a reduction in the risk of death of about a third. Toxicity was higher: grade 3 or higher treatment-related adverse events occurred in about 85 percent of patients on the conjugate and about 59 percent on chemotherapy, mostly hematologic. The population was entirely Chinese and about nine in ten patients were men, and the authors frame the result as supporting a new second-line option in China. The trial adds an overall survival gain, a hard endpoint, in a setting the authors describe as lacking effective therapy.
In the Journal of Clinical Oncology, Meduri and colleagues report five-year results of IRMA, a phase 3 noninferiority trial at 35 centres [3]. It compared external-beam accelerated partial-breast irradiation, 38.5 gray in 10 twice-daily fractions, with whole-breast irradiation in 3,309 patients aged 49 or older with invasive tumours up to 3 centimetres, margins of at least 2 millimetres, and node-negative or limited nodal disease. Median follow-up was about eight years. Ipsilateral breast tumour recurrence at five years was about 1 percent with partial-breast treatment and just under 1 percent with whole-breast treatment. Formal noninferiority was not demonstrated, because the upper limit of the estimate crossed the prespecified margin. The absolute difference between arms was under 0.2 percentage points. The design assumed a five-year recurrence rate of 4 percent with whole-breast treatment, and the observed rates were far below that. Neither patients nor investigators were masked. The authors conclude that external-beam partial-breast irradiation is a reasonable alternative to whole-breast treatment for low-risk patients, a conclusion resting on the very low absolute recurrence rates in both arms, with the formal statistical test unmet.
Now, four in brief.
A news report in Cancer Discovery describes the phase 3 Opti-TROP-Lung06 trial in newly diagnosed, PD-L1-negative, advanced nonsquamous non-small cell lung cancer without targetable alterations [4]. Patients who received the TROP-2-directed antibody-drug conjugate sacituzumab tirumotecan with pembrolizumab had a median progression-free survival of 15 months. Patients given standard chemotherapy plus pembrolizumab had a median of just under 6 months. PD-L1-negative disease is a group in which chemoimmunotherapy has been the standard first-line comparator. The evidence comes from a news summary of a single phase 3 trial, with progression-free survival as the reported endpoint, so its firmness rests on the full trial report and on survival data.
A related paper in Nature Medicine, led by Garassino, asks which lung cancer patients benefit from another TROP2-directed conjugate, datopotamab deruxtecan [5]. Visual TROP2 scoring does not predict response, so the group built a computational pathology score, the normalised membrane ratio, which measures how much TROP2 sits at the cell membrane relative to the whole cell. The score was developed in TROPION-PanTumor01 and validated in TROPION-Lung01. A positive ratio was linked to longer progression-free survival with datopotamab deruxtecan, about 7 versus 4 months, with no such difference on docetaxel. This is a retrospective biomarker analysis with an optimised cut-point, so it remains hypothesis-confirming rather than prospectively tested.
In JAMA Oncology, Liu and colleagues report CAMORAL, a phase 2 randomised trial at four hospitals in China in 125 patients with resectable stage III to IVB head and neck squamous cell carcinoma, mostly oral cavity primaries [6]. Two cycles of neoadjuvant camrelizumab with albumin-bound paclitaxel and carboplatin, then surgery, adjuvant therapy and maintenance camrelizumab, raised two-year event-free survival to about 78 percent from 46 percent with upfront surgery. Two-year overall survival was 92 versus 69 percent. Grade 3 or higher adverse events were more frequent with the perioperative approach. The trial is small and phase 2, and the authors call for further evaluation.
In the Journal of Clinical Oncology, La Rosée and colleagues report DasaHIT, a phase 3 trial in 289 patients with chronic phase chronic myeloid leukemia [7]. Dasatinib 100 milligrams daily was compared with five days on and the weekend off. Major molecular response at 24 months was about 81 percent with continuous and 78 percent with intermittent dosing, which met noninferiority. In patients treated second-line or later, response was 24 percent lower and noninferiority failed. Overall grade 2 to 4 toxicity did not differ. In a post hoc analysis, pleural or pericardial effusion was about half as common with intermittent dosing. The noninferiority finding is firmest in first-line patients, and the effusion signal is exploratory.
Three quick signals.
In Annals of Oncology, Janjigian and colleagues used circulating tumour DNA to select patients with resected mismatch repair-deficient tumours of any site for six months of PD-1 blockade [8]. Among 13 treated patients with residual disease, the circulating tumour DNA cleared in 11, and five patients recurred. This is a small single-arm phase 2 study.
The next paper also takes a tissue-agnostic view. In The Lancet Oncology, Black and colleagues analysed whole genomes from 17,029 tumours in English and Dutch cohorts [9]. Immediately actionable alterations appeared in about a fifth of primary and a quarter of metastatic cancers, and about half of these variants fell in tumour types without United States Food and Drug Administration approval for that biomarker. The analysis is retrospective with no treatment outcomes.
In the Journal of Clinical Oncology, Zanwar and colleagues built the CHARGE index from six clinical and biological variables in 943 bispecific T-cell engager treatments for relapsed or refractory myeloma at 15 centres in the United States [10]. Median progression-free survival ranged from about 22 months in the low-risk group to under 2 months in the high-risk group. The index was externally validated, though follow-up is short and the data retrospective.
Stepping back across this week's papers, much of the strongest randomised evidence still rests on recurrence or progression endpoints at early follow-up, while the hard survival signals come from an interim analysis and a small phase 2 trial. Several tissue-agnostic and biomarker studies are retrospective or single-arm, so they define candidates for prospective trials more than settled indications. A large noninferiority trial also shows how very low event rates can leave a formal test unmet while absolute differences stay small.
If you only have time for one paper this week, make it the intismeran autogene trial in resected melanoma [1]. It reopens the question of whether checkpoint inhibition alone is the ceiling for adjuvant therapy in high-risk melanoma, with distant-metastasis and survival data still maturing.
That's your roundup for This Week in Oncology. The full transcript and references are available on the episode page. 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 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. Use it to choose which papers to read in full.
Get Oncology every week
By email, Spotify, Apple or any podcast app.
Spot something worth flagging?