📬 Lung/Head and Neck/Thymoma Intelligence Digest — Week 2026-W44

DeLLphi-305: tarlatamab added to durvalumab maintenance nearly doubles overall survival in extensive-stage SCLC, and the fight shifts to access and sequencing

Clinicians stressed that DeLLphi-305 randomized patients after induction without progression, so overall and progression-free survival gains are a maintenance result; CRS occurred in 49.6% (G3+ 0.4%). Critics pointed out that only 12.2% of the ITT durvalumab arm (34/278) accessed tarlatamab at progression, that OS benefit appeared nullified in those crossing over to second-line DLL3 therapy, and that the trial enrolled ±40% Asian patients with weaker results in the Americas. Patients with planned consolidative chest RT were excluded, leaving thoracic consolidation unanswered. The debate is whether every progression-free patient should receive tarlatamab maintenance or whether it can be held for relapse where second-line attrition is low and monitoring capacity is tight.

Key numbers: mOS 21.1 vs 12.7 mo, HR 0.42 · mPFS 7.4 vs 1.9 mo, HR 0.34 · 12-mo OS 79.4% vs 53.3% · ORR 26.4% vs 4.1% · Grade ≥3 AEs 46.8% vs 40.6%

Discussed by: @christinemphmd, @RManochakian, @g_mountzios, @ErulEnes, @DrChoueiri, @ADesaiMD, @HHorinouchi, @drshieldsmd, @EladSharonMD, @BalazsHalmosMD, @jryckman3, @pistoto2017

→ View representative tweet

Sources cited: Tarlatamab–Durvalumab Maintenance for Extensive-Stage Small-Cell Lung Cancer (N Engl J Med)

FLAURA2 ctDNA analysis and PACE-LUNG: early EGFR clearance is prognostic, not predictive, so it cannot be used to skip chemotherapy

In the exploratory FLAURA2 analysis, EGFRm clearance was prognostic but not predictive: osimertinib plus platinum–pemetrexed outperformed osimertinib alone regardless of clearance status at weeks 1, 2, 3 and 6. Clinicians concluded that clearance does not refine treatment for baseline ctDNA-positive patients, for whom osimertinib plus chemotherapy or amivantamab-lazertinib beats osimertinib alone. Single-arm PACE-LUNG intensified with chemotherapy for persistent ctDNA, reporting mPFS 16.3mo from start of chemotherapy and mOS 29.2mo, with 29% of screened patients ctDNA-persistent but no comparator. The open question is whether baseline or early ctDNA has any role in deciding on added chemotherapy, or whether upfront intensification is now the default.

Key numbers: Week 6 clearance 88% vs 89% · Week 6 PFS HR 0.52 (cleared), 0.40 (not cleared) · Week 6 OS HR 0.73 (cleared), 0.50 (not cleared)

Discussed by: @dplanchard, @JordiRemon, @tnewsomdavis, @StephenVLiu, @GlopesMd, @DrRiyazShah, @Dr_ElvinaA, @HHorinouchi

→ View representative tweet

PERSEE: adding chemotherapy to pembrolizumab in PD-L1 ≥50% NSCLC more than doubles PFS, and reopens whether pembrolizumab monotherapy is a fair control arm

In the open-label phase 3 PERSEE (n=321), chemo-immunotherapy more than doubled median PFS over pembrolizumab alone, while the numerical OS gain was not statistically significant and OS was not the primary endpoint. Trialists called it the only study asking whether combination therapy beats pembrolizumab monotherapy in PD-L1-high patients, and clinicians questioned whether monotherapy remains an acceptable control arm for chemotherapy-eligible patients. The debate is which patients justify the substantially higher grade ≥3 toxicity of upfront chemotherapy without a significant OS difference.

Key numbers: Median PFS 16.1 vs 6.9 mo, HR 0.56 · Median OS 50.2 vs 29.9 mo, HR 0.77 · ORR 77.5% vs 50.3% · Grade ≥3 TRAEs 53.8% vs 19.3%

Discussed by: @dr_yakupergun, @dplanchard, @johanpluvy, @HHorinouchi, @DrRiyazShah, @nataliagandur, @ChandrakanthMv, @Erman_Akkus

→ View representative tweet

Sources cited: Annals of Oncology

New COCOON cohort: lazertinib lead-in plus simplified skin care cuts dermatologic toxicity of subcutaneous amivantamab-lazertinib

The new COCOON cohort of first-line subcutaneous amivantamab plus lazertinib combined oral tetracycline, ceramide cream and oral zinc with lazertinib started before amivantamab, and the lead-in roughly halved grade ≥2 dermatologic events in non-randomized exploratory cohorts. Grade ≥2 dermatologic AEs of interest by week 12 occurred in 34%, with discontinuations of 3%. Clinicians noted the simplified prophylaxis looks comparable to, not better than, the original COCOON regimen, and that PFS data are needed to confirm that delaying amivantamab does not compromise efficacy in the FLAURA2 era.

Key numbers: Grade ≥2 dermatologic AEs 39% vs 18% (standard vs lead-in) · All-grade dermatologic AEs HR 0.31 at 3 months

Discussed by: @tnewsomdavis, @StephenVLiu, @DrRiyazShah, @Tony_Calles, @herbloong, @HHorinouchi

→ View representative tweet

SAFFRON makes osimertinib plus savolitinib a chemotherapy-free standard after osimertinib progression in MET-overexpressed/amplified EGFR-mutant NSCLC

SAFFRON improved PFS, OS and response over platinum–pemetrexed after mandated post-progression re-biopsy, in which 35% of patients evaluable for MET testing had MET overexpression (IHC3+ ≥90%) and/or amplification (FISH 10+). Clinicians called it a new standard and urged MET testing for every EGFR-mutant patient progressing on osimertinib, naming access to the combination as the main barrier; others noted the control arm contained no osimertinib, the MET-amplified subgroup was not formally tested, and peripheral edema reached 49%. The debate is whether every patient should undergo tissue re-biopsy with both IHC and FISH before choosing between osimertinib–savolitinib, an ADC and platinum–pemetrexed.

Key numbers: BICR-PFS 7.6 vs 5.4 mo, HR 0.55 · OS 17.8 vs 12.9 mo, HR 0.68 · ORR 51% vs 27% · mDoR 9.7 vs 5.7 mo · 12-mo OS 70% vs 53%

Discussed by: @RManochakian, @dplanchard, @SuyogCancer, @YGaritaonaindia, @TommyJohn00, @DrMirallas, @ozdogan_md, @EgfrUk

→ View representative tweet

Sources cited: NEJM

Sunvozertinib finally reaches US patients more than a year after FDA approval for EGFR exon 20 insertion NSCLC

Sunvozertinib became available in the US over one year after its FDA approval on July 2, 2025, for previously treated EGFR exon 20 insertion NSCLC based on WU-KONG1B (RR 46%, mDOR 11.1 months), with key AEs of diarrhea (73%) and rash (60%). First-line phase III WU-KONG28 data versus carboplatin/pemetrexed improved PFS, response and response duration but are outside the US label. Clinicians had expected zipalertinib to reach the market first, and patient advocates urged those progressing on amivantamab plus chemotherapy to ask about sunvozertinib and check pharmacy ordering. The open question is where it sits against amivantamab plus chemotherapy and zipalertinib.

Key numbers: WU-KONG28 PFS 10.3 vs 7.5 mo, HR 0.65 · ORR 59% vs 31% · DoR 11.2 vs 7.1 mo

Discussed by: @Latinamd, @StephenVLiu, @lungoncdoc, @drgandara, @BalazsHalmosMD, @RManochakian, @Exon20Group

→ View representative tweet

Sources cited: FDA press release

OptiTROP-Lung04 final OS: sac-TMT extends survival over platinum-pemetrexed after EGFR TKI progression

The final OS analysis of OptiTROP-Lung04 (LBA67) favoured the TROP2 ADC sac-TMT over platinum plus pemetrexed after EGFR-TKI therapy, at 25.9m vs 17.4m (HR 0.69), with greater benefit when first-line therapy had been a third-generation TKI. Commentators said full data on CNS activity, toxicity and quality of life are still needed. The debate is whether a TROP2 ADC now displaces platinum-pemetrexed as the default after third-generation TKI failure, or whether MET-directed therapy comes first when MET is positive.

Discussed by: @StephenVLiu, @DrMirallas, @GlopesMd

→ View representative tweet

KANDLELIT-001 and SUNRAY: next-generation KRAS G12C inhibitors move into first-line combinations with immunotherapy

Calderasib (MK-1084) with pembrolizumab in PD-L1-positive disease and first-line olomorasib with chemoimmunotherapy both showed high response rates, with olomorasib active across PD-L1 levels and reporting no new safety signals. The phase 3 SUNRAY-02 trial of olomorasib plus immunotherapy in resected or unresectable KRAS G12C-mutant NSCLC is now enrolling. Clinicians cautioned that sequencing PD-(L)1 before sotorasib strongly increases grade ≥3 hepatotoxicity with a short washout and that KRAS inhibitor plus immunotherapy toxicity varies markedly by agent, making agent choice and washout the deciding factors for first-line combinations.

Key numbers: Calderasib + pembro ORR 72%, mPFS 28.9 mo · Olomorasib + chemo-IO ORR 61%, mPFS 11.8 mo

Discussed by: @BrunaPellini, @g_mountzios, @HHorinouchi, @weoncologists

→ View representative tweet

Krascendo 1: divarasib beats sotorasib/adagrasib head-to-head on PFS and OS, but the absent crossover and sotorasib-heavy control arm draw fire

Divarasib improved PFS, OS and response over first-generation inhibitors with fewer severe toxicities and discontinuations, and clinicians called it a new second-line standard. However, the control arm was 91% sotorasib, 30% of patients had treated brain metastases, untreated brain metastases were excluded, and Dr. Benjamin Besse called the lack of crossover in a PFS-endpoint trial unjust. The debate is whether the OS advantage reflects a true class advance or partly what control patients could access next.

Key numbers: mPFS 9.8 vs 5.7 mo, HR 0.58 · mOS 17.7 vs 12.0 mo, HR 0.56 · ORR 58.4% vs 32.3% · Grade 3–4 TRAEs 24.6% vs 34.5% · TRAE discontinuation 4.8% vs 11.7%

Discussed by: @StephenVLiu, @CathyEngMD, @YGaritaonaindia, @Latinamd, @NarjustFlorezMD, @AnaVManana, @SuyogCancer, @Dr_Ivanoncologo

→ View representative tweet

OptiTROP-Lung06: sac-TMT plus pembrolizumab triples PFS versus chemo-pembrolizumab in PD-L1-negative nonsquamous NSCLC

OptiTROP-Lung06 roughly tripled PFS with sacituzumab tirumotecan plus pembrolizumab over chemotherapy plus pembrolizumab, with less grade ≥3 toxicity. Clinicians framed it as one of two strategies for the unmet need of first-line PD-L1 <1% NSCLC. The debate is whether a TROP2 ADC plus pembrolizumab is ready to displace platinum-pemetrexed-pembrolizumab while overall survival remains immature.

Key numbers: PFS 15.0 vs 5.7 mo, HR 0.49 · Grade ≥3 AEs 47.6% vs 60.6%

Discussed by: @OncoAlert, @DrMirallas, @GlopesMd, @phsiao4

→ View representative tweet

Sources cited: Sacituzumab Tirumotecan plus Pembrolizumab in PD-L1–Negative Advanced NSCLC (N Engl J Med)

Five-year outcomes published for 30-Gy de-escalated elective nodal irradiation in HPV-associated oropharyngeal cancer

The published cohort reports 5-year outcomes in 276 patients with HPV-related oropharyngeal cancer treated with 30 Gy elective RT within 70 Gy chemoradiotherapy. The approach gives 30 Gy to the elective neck plus a 40 Gy boost to gross disease, omits uninvolved levels 1B and 5 and contralateral RP/RS nodes, and is intended for patients receiving platinum-based chemotherapy, in contrast to traditional 50-60 Gy elective doses that drive much of the toxicity. The debate is whether single-institution five-year durability justifies adoption outside a trial or whether randomized confirmation is needed.

Key numbers: 5-y LRC 97.0% · OS 90.8% · PFS 85.9% · DMFS 91.2%

Discussed by: @safaviaa, @safaviaa, @PierluigiBonomo, @jryckman3

→ View representative tweet

Near-loop versus far-loop EGFR exon 20 insertions differ in TKI sensitivity, raising ultra-precision subtyping

EGFR exon 20 insertions separate into near-loop (A767-P772) and far-loop (H773-R776) groups. Molecular dynamics simulations show near-loop insertions have multiple conformational states and lower transitional energy, producing differential responses and PFS with some TKIs but not all. Clinicians countered that variant-level data remain limited and that sequencing is still driven by patient preference, toxicity and access, leaving open whether insertion position should guide agent choice.

Discussed by: @LeXiuning, @jennifermarksmd

→ View representative tweet

SARON: consolidation radiotherapy in synchronous oligometastatic NSCLC improves PFS and local control but delivers no overall survival benefit

Consolidation radiotherapy added to systemic therapy improved PFS and local control without extending survival, and clinicians concluded that better disease control does not always translate into longer survival in driver-negative oligometastatic NSCLC. Technique drew scrutiny: of 62 patients who received RT, 51 had conventional treatment to the primary and only 11 had primary SABR, and about 53% of the cohort had N2/3 disease. The debate is which patients, and which radiotherapy technique, still justify consolidation.

Key numbers: PFS 13.9 vs 8.9 mo, HR 0.64 · 12-mo local failure 13.0% vs 39.5% · OS 31.0 vs 28.0 mo, HR 0.99 · Grade ≥3 AEs 28.6% in both arms

Discussed by: @jryckman3, @Latinamd, @nataliagandur, @HHorinouchi

→ View representative tweet

CROWN ctDNA analysis: no new on-target ALK resistance mutations with first-line lorlatinib, with bypass pathways and TP53 driving early progression

The CROWN ctDNA analyses, including Guardant Infinity data, found no ALK kinase domain mutations causing acquired resistance to lorlatinib, with bypass RTK/RAS/MAPK pathways driving resistance instead. Early progressors had higher baseline genomic complexity including TP53, TP53 and clearance were prognostic, and inflammatory response was the predominant pathway associated with early progression. The question is whether early progressors should be identified upfront by ctDNA complexity and offered combination strategies rather than a sequential TKI.

Discussed by: @StephenVLiu, @DrRiyazShah, @KSamalaMD, @TommyJohn00, @Tony_Calles, @HHorinouchi

→ View representative tweet

Telisotuzumab adizutecan OS data position a c-MET ADC in pretreated nonsquamous NSCLC, with the biomarker question unresolved

In heavily pretreated nonsquamous NSCLC, telisotuzumab adizutecan showed survival in both EGFR wild-type and EGFR-mutant cohorts, with c-MET expression predicting outcomes in EGFR wild-type but not EGFR-mutant disease. Clinicians questioned whether MET positivity is truly a predictive biomarker in EGFR-mutant disease, whether adding osimertinib raises activity, brain activity, and whether the biomarker is easy to test in daily practice. The debate is whether c-MET IHC is predictive or only prognostic, and whether the high-expression cutoff holds up in practice.

Key numbers: EGFR WT mPFS 6.8m, OS 13.8m · EGFR MUT mPFS 10.9m, OS 16.2m · High c-Met expressers (EGFR WT ≥3+/≥10%) OS 25.2mo

Discussed by: @StephenVLiu, @JordiRemon, @g_mountzios, @TejasPatilMD

→ View representative tweet

Sources cited: Annals of Oncology

ESMO head and neck: EGFR-directed bispecifics and ADCs post response rates near 60-77% in first-line recurrent/metastatic disease

OrigAMI-4 tested amivantamab with pembrolizumab and carboplatin in HPV-unrelated disease, and the EGFR×HER3 bispecific ADC iza-bren was combined with pembrolizumab, both with high response rates. The OX-40 agonist INBRX-106 doubled ORR but carried high toxicity and treatment discontinuation in a population eligible for subcutaneous pembrolizumab alone. The program also featured interest in adding neoadjuvant chemotherapy to perioperative immunotherapy, new KEYNOTE-689 analyses and a surge of ADCs, leaving open whether these agents add enough over pembrolizumab-based therapy and which biomarker should select patients.

Key numbers: OrigAMI-4 ORR 77%, CR 14% · Iza-bren + pembro ORR 58–63%, mDoR 13.6–17.8 months · INBRX-106 ORR 48 vs 26%

Discussed by: @JiaJennyLiu, @PaulJiL, @DrMirallas

→ View representative tweet

Final PEARLS/KEYNOTE-091 analysis: adjuvant pembrolizumab keeps its DFS benefit at 87.7 months but shows no overall survival advantage

In 1177 patients with resected stage IB–IIIA NSCLC followed for 7.3 years, adjuvant pembrolizumab maintained its DFS benefit with a favourable lung cancer-specific survival trend, but OS was not significantly improved and the study was not powered for OS. Clinicians read the mature data as showing that adjuvant immunotherapy can reduce recurrence risk without a clear population-level OS gain. The debate is which resected patients derive enough absolute DFS benefit to justify a year of pembrolizumab and its added grade ≥3 toxicity.

Key numbers: DFS HR 0.85, median 58.7 vs 43.3 mo · OS HR 0.98, 6-y OS 68.5% vs 66.4% · LCSS HR 0.81 · Grade ≥3 AEs 34.0% vs 25.6%

Discussed by: @ChandrakanthMv, @nataliagandur, @HHorinouchi

→ View representative tweet