Saved Results

No results saved yet.

Enter a patient name and hit Save on a result.

NCCN NSCLC v5.2026ESMO Living Guideline v1.3 · Feb 2026Last reviewed: Apr 2026

Stage IV NSCLC: Biomarker-Directed TherapyKnow the Driver Before Starting Treatment

Before you choose a first-line regimen, you need molecular testing results. Do you have all of them?

Non-negotiable molecular tests before ANY systemic therapy

  1. 1EGFR mutationsexons 18–21, incl. exon 20 ins — ~50% Asian, ~15% Western adenocarcinoma
  2. 2ALK rearrangementIHC D5F3 + FISH or NGS — ~5%
  3. 3ROS1 fusion~2%
  4. 4RET fusion~2%
  5. 5MET exon 14 skipping mutation~3–4%
  6. 6BRAF V600E~2–3%
  7. 7KRAS G12C~13% (mostly Western, squamous rare)
  8. 8HER2 mutation/amplification~3%
  9. 9NTRK fusion<1%
  10. 10PD-L1 TPSmandatory for all no-driver patients; 22C3 clone for pembrolizumab decisions

Preferred testing: Comprehensive NGS (tissue first; ctDNA if tissue insufficient). PD-L1 IHC (22C3 clone for pembrolizumab; SP263 interchangeable; SP142 is the outlier — underestimates TC staining).

All non-squamous adenocarcinoma MUST have full molecular workup.

Section 1 — No Driver Mutation

No actionable driver found — PD-L1 result determines your entire 1L strategy.

PD-L1 ≥50%
  • Pembrolizumab monotherapy (KEYNOTE-024: PFS 10.3 vs 6.0 mo; OS 26.3 vs 13.4 mo) — preferred across squamous and non-squamous
  • Cemiplimab monotherapy (EMPOWER-Lung1: PFS 8.2 vs 5.7 mo) — alternative to pembrolizumab
  • Chemo + IO combination is acceptable but IO monotherapy preferred at ≥50%
PD-L1 1–49%
  • Non-squamous: Pembrolizumab + carboplatin + pemetrexed (KEYNOTE-189: PFS 9.0 vs 4.9 mo; OS 22 vs 10.7 mo)
  • Squamous: Pembrolizumab + carboplatin + paclitaxel/nab-paclitaxel (KEYNOTE-407: PFS 6.4 vs 4.8 mo; OS 17.1 vs 11.6 mo)
  • Atezolizumab-based combinations: acceptable alternatives
PD-L1 <1%
  • Non-squamous:Chemo + IO (KEYNOTE-189 showed benefit even in PD-L1 <1% subgroup)
  • Squamous: KEYNOTE-407 showed benefit across all PD-L1 subgroups
  • IO monotherapy NOT recommendedfor PD-L1 <1%

Chemo Foundation — Platinum Doublet Principles

  • Cisplatin vs carboplatin: Cisplatin superior ORR but greater nephrotoxicity, ototoxicity, emesis — carboplatin preferred for elderly, renal impairment, or hearing concerns. Carboplatin AUC 5–6 is standard backbone with pemetrexed or paclitaxel.
  • Pemetrexed for non-squamous: Preferred partner (JMDB: pemetrexed + cisplatin vs gemcitabine + cisplatin — superior OS in non-squamous 12.6 vs 10.9m). Premedication: folic acid 400–1000 µg daily, vitamin B12 1000 µg IM q9 weeks, dexamethasone.
  • Squamous histology: Gemcitabine or paclitaxel/nab-paclitaxel backbone — pemetrexed is NOT effective in squamous NSCLC.
  • Bevacizumab addition (non-squamous): ECOG 4599 — carbo + paclitaxel + bevacizumab: OS 12.3 vs 10.3m (HR 0.79). Add when non-squamous, no haemoptysis, no anticoagulation, no recent haemorrhage.
  • Number of cycles: 4–6 cycles induction for non-IO patients. For IO combinations: 4 cycles platinum then maintenance IO ± pemetrexed (non-squamous).
  • Elderly / PS2: Carboplatin + weekly paclitaxel or single-agent vinorelbine/gemcitabine. Doublet in PS2 disease-related: reassess after 2 cycles.

Additional Options & Special Populations

  • Atezolizumab + bev + carbo + paclitaxel (IMpower150): ORR 64%, PFS 8.3 vs 6.8 mo — particularly useful in EGFR/ALK+ who have progressed on targeted therapy (unique population approval)
  • Nivolumab + ipilimumab ± 2 cycles chemo (CheckMate 9LA/227): OS benefit; durable responses especially in high-TMB patients
  • STK11/KEAP1 mutations: Associated with IO resistance — consider chemo + IO combination rather than IO monotherapy; clinical trial preferred where available

Maintenance (non-squamous):

Pemetrexed + pembrolizumab continuation after 4 cycles induction

Squamous:

No maintenance benefit with pemetrexed; IO continuation only

Second-Line Options (No Driver, Post-IO Era)

  • Post-chemo-IO progression: Docetaxel ± ramucirumab (REVEL: OS 10.5 vs 9.1m, HR 0.86) or docetaxel ± nintedanib (LUME-Lung1: OS 12.6 vs 10.3m non-squamous)
  • IO-naïve 2L: Pembrolizumab (PD-L1 ≥1%, KEYNOTE-010: OS 10.4m); nivolumab (CheckMate 017/057: all-comers, OS HR ~0.70); atezolizumab (OAK: OS 13.8 vs 9.6m, HR 0.73)
  • KRAS G12C (2L): Adagrasib (KRYSTAL-1: ORR 43%, PFS 6.5m) or sotorasib (CodeBreak 200: ORR 28%, PFS 5.6m)
  • ADC 2L: Dato-DXd vs docetaxel (TROPION-Lung01: PFS 4.4 vs 3.4m HR 0.75; ORR 26% vs 13%) — emerging option especially non-squamous

Key Trial Data

KEYNOTE-024 — Pembrolizumab vs Chemo (PD-L1 ≥50%)

Design: Phase III, 305 pts, PD-L1 TPS ≥50%, no EGFR/ALK, Stage IIIB/IV NSCLC

Pembrolizumab 200 mg Q3W vs platinum-based chemotherapy

PFS: 10.3 vs 6.0 mo (HR 0.50, p<0.001)

OS: 26.3 vs 13.4 mo (HR 0.62; 5-year OS 31.9% vs 16.3%)

ORR: 45.5% vs 27.7%

Toxicity: Grade 3–5 TRAEs 26.6% vs 53.3%

Ref: Reck M et al. NEJM 2016; updated OS NEJM 2019.

KEYNOTE-189 — Pembro + Chemo vs Chemo (Non-squamous)

Design: Phase III, 616 pts, non-squamous Stage IV NSCLC, any PD-L1

Pembrolizumab + carboplatin + pemetrexed vs placebo + carboplatin + pemetrexed

PFS: 9.0 vs 4.9 mo (HR 0.48, p<0.001)

OS: 22.0 vs 10.7 mo (HR 0.56, p<0.001)

ORR: 48.3% vs 19.9%

Benefit across all PD-L1 subgroups, including <1% (OS HR 0.59)

Ref: Gandhi L et al. NEJM 2018; updated Garassino M et al. Lancet Oncol 2020.

KEYNOTE-407 — Pembro + Chemo vs Chemo (Squamous)

Design: Phase III, 559 pts, squamous Stage IV NSCLC, any PD-L1

Pembrolizumab + carboplatin + paclitaxel/nab-paclitaxel vs placebo + chemo

PFS: 6.4 vs 4.8 mo (HR 0.56, p<0.001)

OS: 17.1 vs 11.6 mo (HR 0.71, p=0.0008)

ORR: 58.4% vs 35.6%

Benefit across all PD-L1 subgroups

Ref: Paz-Ares L et al. NEJM 2018; updated OS NEJM 2020.

EMPOWER-Lung1 — Cemiplimab vs Chemo (PD-L1 ≥50%)

Design: Phase III, 710 pts, PD-L1 TPS ≥50%, Stage IIIB/IV NSCLC

Cemiplimab 350 mg Q3W vs investigator-choice platinum-based chemo

PFS: 8.2 vs 5.7 mo (HR 0.54, p<0.0001)

OS: 22.1 vs 14.3 mo (HR 0.68, p=0.0022)

ORR: 39.2% vs 20.4%

Ref: Sezer A et al. Lancet 2021.

Section 2 — EGFR Mutation

EGFR-mutated — Ex19del or L858R? That distinction shapes 1L intensity.

1L Options — Full Trial Comparison (FLAURA / FLAURA2 / MARIPOSA)

FactorFLAURAFLAURA2MARIPOSA
PublicationNEJM 2018 / 2020NEJM 2023 / WCLC 2025 (Final OS)NEJM 2024 / WCLC 2025 (Updates)
PhaseIIIIIIIII
RegionGlobalGlobalGlobal
PopulationEGFRm Ex19del / L858REGFRm Ex19del / L858REGFRm Ex19del / L858R
InterventionOsimertinib 80mg ODOsimertinib + Carbo/Cis + PemAmivantamab + Lazertinib
ControlGefitinib / ErlotinibOsimertinib monotherapyOsimertinib monotherapy
Primary EndpointPFSPFSPFS
PFS (HR)18.9 vs 10.2m (HR 0.46)25.5 vs 16.7m (HR 0.62)23.7 vs 16.6m (HR 0.70)
OS (HR)38.6 vs 31.8m (HR 0.80)47.5 vs 37.6m (HR 0.77) ★Interim: HR 0.80 (NR)
ORR80%83%80%
Asian BenefitSignificantSignificant (Consistent HR)Significant (Consistent HR)
Key SubgroupBenefit both Ex19/L858RGreatest benefit: L858R + CNSBenefit both; MET-driven?
ToxicityRash, diarrhea (well-tolerated)Anemia, neutropenia, fatigueIRR, rash, paronychia, VTE (40%)
Discontinuation13%11% (Osi), 48% (chemo)10% (Laz), 10% (Ami)
Route100% oralOral + IV (q3w)Oral + SC (monthly)*
FDA Approval2018Feb 2024Aug 2024
HK HA StatusSamaritan Fund (SF)SF (Osi) + Standard (Chemo)Self-Financed (Expected SF 2026)

★ Longest median OS reported in 1L EGFRm NSCLC to date. *SC formulation of amivantamab reduces infusion reactions vs IV.

Clinical Decision Framework — Which 1L Option?

FactorOsimertinib MonotherapyOsi + Chemo (FLAURA2)Ami + Laz (MARIPOSA)
Best ForLow tumor burden, Ex19del, Elderly/frail, Toxicity-averseL858R, CNS disease, High tumor burden, Fit for chemoL858R, Fit, IV/SC acceptable, VTE manageable
Primary AdvantageQuality of life, 100% oral, Low toxicity, Well-establishedLongest median OS (47.5m), proven CNS controlWidening survival curve, potential delay of resistance
Main DrawbackEarly resistance in high-risk molecular subgroupsMyelosuppression (anemia, neutropenia), IV chemoVTE risk (40%), Skin/nail toxicity, Infusion reactions
Oral vs IV100% oralOral + IV q3wOral + SC monthly
HK FundingSF (Samaritan Fund)SF (Osi) + Standard (Chemo)Self-Financed

Clinical Decision Rules

  • Ex19del, low burden, elderly/frail, oral preference: → Osimertinib monotherapy
  • L858R, CNS disease, high tumour burden, fit for chemo: → Osimertinib + chemo (FLAURA2) — longest median OS 47.5 months
  • L858R, fit, tolerates IV/SC, VTE-acceptable: → Amivantamab + lazertinib (MARIPOSA)
  • IO is contraindicated as 1L in EGFRm: Targeted therapy takes priority regardless of PD-L1 expression (even TPS 100%).

Post-Osimertinib Resistance — Re-biopsy Mandatory (Tissue + ctDNA complementary)

Resistance MechanismFrequencyDetectionTreatment StrategyKey Data
MET amplification~25%NGS / FISHOsimertinib + savolitinib (SACHI) or tepotinib (INSIGHT-2)SACHI: PFS 8.2 vs 4.5m, HR 0.34; ORR 58%
C797S mutation~10–15%ctDNA / NGSTrans-config: 1st-gen TKI + osimertinib; Cis-config: no approved strategy — clinical trialLimited clinical data; depends on cis/trans with T790M
HER2 amplification~5–10%NGS / FISHT-DXd (DESTINY-Lung02) or osimertinib + tucatinib (ORCHARD)T-DXd: ORR 49–56%, PFS 9.9m
BRAF V600E (acquired)~2–3%NGSOsimertinib + dabrafenib + trametinibESMO 17P: ORR 42.5%, PFS 8.5m HR 0.38
RET fusion (acquired)~1–2%NGSOsimertinib + selpercatinibORR ~50%, DCR 83%
SCLC transformation~5–10%Tissue re-biopsy mandatoryPlatinum + etoposide; tarlatamab 2LEP regimen: ORR 40–70%; tarlatamab mOS 14.3m
Squamous transformation~3%Tissue re-biopsyImmunotherapy + chemo per squamous protocolLimited data; treat as primary squamous
Unknown / polyclonal~30–40%Amivantamab + chemo (MARIPOSA-2); or ADC (Sac-TMT, HER3-DXd, Dato-DXd)MARIPOSA-2: PFS 6.3 vs 4.2m, HR 0.48; OptiTROP-04: PFS 8.3m

Post-Osimertinib ADC / Chemo-Targeted Options

TrialDrug/RegimenPhasePopulationPFS (HR)ORRMain ToxicityFDA StatusHK HA
MARIPOSA-2Amivantamab + Carbo + PemIIIPost-Osi EGFRm6.3 vs 4.2m (HR 0.48)64%Hematologic, Rash, VTEApproved Oct 2023SFI (PAP/Capping)
HERTHENA-Lung02HER3-DXd (Patritumab-DXd) 5.6mg/kgIIIPost-TKI EGFRm6.4 vs 3.7m (HR 0.77)~30–40%Nausea, ILD (5%)Breakthrough / PrioritySFI (1+ mechanism)
TROPION-Lung05Dato-DXd (Datopotamab-DXd) 6mg/kgII (Pivotal)Pretreated actionable NSCLC5.8m (single arm)43.6% (EGFR sub)Stomatitis (60%), RashFiled / Fast TrackSFI
OptiTROP-Lung03Sac-TMT 5mg/kg Q2WIIIEGFRm post-TKI + Chemo6.9 vs 2.8m (HR 0.30)45% vs 16%Neutropenia, StomatitisNMPA Approved (China)SFI (1+ mechanism)
OptiTROP-Lung04Sac-TMT 5mg/kg Q2WIIIEGFRm post-TKI8.3 vs 4.3m (HR 0.49)60% vs 43%Neutropenia, StomatitisFDA/EMA (Pending/Priority)SFI (1+ mechanism)

Sac-TMT = sacituzumab tirumotecan; ILD rate notably low/0% in OptiTROP studies. OS OptiTROP-Lung03: 20.0 vs 11.2m (HR 0.45); OptiTROP-Lung04: NR vs 17.4m (HR 0.60).

MET Amplification as Osimertinib Resistance — Trial Data

TrialPhaseRegionInterventionControlPFS (HR)ORROS (Months)Molecular MarkerApproval
SACHIIIIChina (Asian focus)Savolitinib + OsiPlatinum-Pem8.2 vs 4.5m (HR 0.34)58% vs 34%22.9 vs 17.7mMET Amp (FISH/IHC)NMPA Approved 2025
SAFFRONIIIGlobalSavolitinib + OsiPlatinum-Pem9.2 vs 4.8m (HR 0.38)61% vs 35%HR 0.72 (immature)MET Amp (FISH 10+)FDA Filed / HK '1+'
SAVANNAHII (Registrational)GlobalSavolitinib + OsiN/A (single arm)7.2m49% (High MET)~20.0mIHC 90+ or FISH 10+Breakthrough
INSIGHT 2IIGlobalTepotinib + OsiN/A (single arm)5.4m50% (FISH+)NRFISH (GCN ≥5)Phase II Data

SCLC Transformation After Osimertinib — Treatment Options

RegimenStrategyTrialORRmPFSmOSMain ToxicityFDA Status
Platinum + Etoposide (EP)Current SOCExtrapolated from 1L SCLC40–70% (short-lived)4–6m9–11mMyelosuppression, alopeciaApproved
Tarlatamab (DLL3-BiTE)Emerging 2L targeted (DLL3)DeLLphi-301 (Ph II/III)40% (DLL3+)4.9m14.3mCytokine release (CRS)Approved May 2024
Ifinatamab-DXd (I-DXd)Emerging B7-H3 ADCIDYLLIC-01 (Ph II)52%5.6m12.2mNeutropenia, ILD (low)Breakthrough

Tissue re-biopsy is mandatory to diagnose SCLC transformation. Tarlatamab: DLL3 expression prevalent in Asian populations. I-DXd: B7-H3 broadly expressed in SCLC transformation.

EGFR Exon 20 Insertion — Separate Algorithm

Standard 3rd-gen TKIs (osimertinib, gefitinib, erlotinib) largely ineffective against Exon 20 insertions

Biology: exon 20 insertions shift the C-helix inward, creating a compact binding pocket — sterically excludes most TKIs. Represents ~5–10% of all EGFR mutations.

  • 1L — Preferred: Amivantamab + platinum + pemetrexed (PAPILLON: PFS 11.4 vs 6.7m, HR 0.40; ORR 73% vs 47%; FDA approved August 2024). Anti-EGFR/MET bispecific antibody uniquely active against exon 20 steric conformation.
  • 2L — Post-chemo: Amivantamab monotherapy (ORR 40%, FDA approved 2021 2L); mobocertinib (ORR 28%, PFS 7.3m — FDA withdrew, still used HK/global)
  • Emerging: DXd-based regimens (T-DXd exon 20 cohort, HER3-DXd) under investigation in post-amivantamab setting
  • Note: Near-loop insertions (before A767) vs far-loop insertions (after A767) may have different sensitivity profiles. Sanger sequencing alone may miss some insertions — use NGS.

Uncommon EGFR Mutations (G719X, L861Q, S768I)

  • Afatinib: LUX-Lung 2/3/6 pooled ORR 71.1% — preferred for these uncommon mutations; 2nd-gen broader kinase coverage
  • Osimertinib: Limited data; reasonable alternative given improved toxicity profile vs afatinib
  • T790M positivity: Uncommon in de novo — but these mutations can co-exist; re-biopsy at progression to identify T790M

Key Trial Data (Expandable)

FLAURA — Osimertinib vs Standard EGFR-TKI (1L)

Design: Phase III, 556 pts, EGFR Ex19del/L858R, Stage IIIB/IV, treatment-naïve

Osimertinib 80mg QD vs erlotinib/gefitinib

PFS: 18.9 vs 10.2 mo (HR 0.46, p<0.001)

OS: 38.6 vs 31.8 mo (HR 0.80, p=0.046)

CNS PFS: NR vs 11.2 mo (HR 0.48)

Intracranial ORR: 91% vs 68%

Ref: Soria JC et al. NEJM 2018; Ramalingam SS et al. NEJM 2020 (OS).

FLAURA2 — Osimertinib + Chemo vs Osimertinib Mono (1L)

Design: Phase III, 557 pts, EGFR Ex19del/L858R, Stage IV (incl. CNS mets)

Osimertinib 80mg + carboplatin/cisplatin + pemetrexed vs osimertinib mono

PFS: 25.5 vs 16.7 mo (HR 0.62, p<0.001)

OS (WCLC 2025 final): 47.5 vs 37.6 mo (HR 0.77, p=0.0044)

L858R subgroup OS: 45.7 vs 32.6 mo (HR 0.74) — greatest benefit

Ex19del subgroup OS: HR 0.80 (less pronounced)

CNS PFS HR: 0.47 — particular benefit with brain metastases

Toxicity: Grade ≥3 AEs 64% vs 27%; discontinuation rate: 11% (Osi) and 48% (chemo component)

Ref: Planchard D et al. NEJM 2023; OS update WCLC 2025.

MARIPOSA — Amivantamab + Lazertinib vs Osimertinib (1L)

Design: Phase III, 1,074 pts, EGFR Ex19del/L858R, treatment-naïve

Amivantamab (IV or SC) + lazertinib 240mg QD vs osimertinib 80mg QD

PFS: 23.7 vs 16.6 mo (HR 0.70, p<0.001)

OS (interim, WCLC 2025): HR 0.80 — not yet mature

ORR: 86% vs 85%

VTE rate: ~37–40% (anticoagulation prophylaxis often used; MARIPOSA VTE sub-study ongoing)

Infusion reactions: ~67% first IV infusion (SC formulation substantially reduces rate)

Ref: Cho BC et al. NEJM 2024; updates WCLC 2025.

MARIPOSA-2 — Amivantamab + Chemo ± Lazertinib vs Chemo (Post-Osi)

Design: Phase III, 657 pts, EGFR+ progressed on osimertinib

Amivantamab + carboplatin + pemetrexed ± lazertinib vs carbo + pem

PFS (amivantamab + chemo arm): 6.3 vs 4.2 mo (HR 0.48, p<0.001)

ORR: 64% vs 36%

OS (interim): 16.0m; HR 0.77 trend

DCR: 86%; DoR: 6.9m

Note: Triple combination (amivantamab + lazertinib + chemo) NOT recommended — excess toxicity without additional PFS benefit

HK HA: SFI with PAP/Capping mechanism

Ref: Passaro A et al. NEJM 2023; Ann Oncol 2024 updates.

OptiTROP-Lung04 — Sac-TMT vs Platinum + Pemetrexed (Post-EGFR-TKI)

Design: Phase III, EGFRm (Ex19/L858R) post-TKI, Sac-TMT 5mg/kg Q2W vs carbo + pem

PFS: 8.3 vs 4.3 mo (HR 0.49, p<0.001)

OS: NR vs 17.4 mo (HR 0.60)

ORR: 60% vs 43%

18-month OS rate: 65%

ILD rate: 0% — notably absent (important safety advantage over T-DXd)

Asian population: High benefit; major study site China/Asia

Approval: FDA/EMA Pending/Priority Review

Ref: Zhou C et al. NEJM/ESMO 2025.

OptiTROP-Lung03 — Sac-TMT vs Docetaxel (Post-TKI + Chemo)

Design: Phase III, EGFRm post-TKI and prior platinum-chemo, Sac-TMT vs docetaxel

PFS: 6.9 vs 2.8 mo (HR 0.30)

OS: 20.0 vs 11.2 mo (HR 0.45)

ORR: 45% vs 16%

18-month OS rate: 68%

ILD rate: 0% — consistent with Lung04

Approval: NMPA Approved (China 2025); HK SFI 1+ mechanism

Ref: BMJ / ELCC 2026.

PAPILLON — Amivantamab + Chemo vs Chemo (EGFR Exon 20 Ins, 1L)

Design: Phase III, 308 pts, EGFR exon 20 insertion, treatment-naïve

Amivantamab + carboplatin + pemetrexed vs carbo + pem

PFS: 11.4 vs 6.7 mo (HR 0.40, p<0.001)

OS: HR 0.67 (interim, not mature)

ORR: 73% vs 47%

FDA approved: August 2024 for EGFR exon 20 insertion, 1L

Ref: Zhou C et al. NEJM 2023.

SACHI — Osimertinib + Savolitinib vs Platinum-Pem (Post-Osi MET amp)

Design: Phase III (China-focused), EGFR+ progressed on osimertinib with MET amplification

Osimertinib 80mg + savolitinib 300–600mg QD vs platinum + pemetrexed

PFS: 8.2 vs 4.5 mo (HR 0.34, p<0.001)

ORR: 58% vs 34%

OS: 22.9 vs 17.7m (trend)

Asian benefit: Highly significant (primary study site)

Approval: NMPA Approved 2025; SAFFRON (global Phase III ongoing)

Ref: Yang JC et al. Lancet 2024.

HERTHENA-Lung02 — HER3-DXd (Patritumab Deruxtecan) vs Platinum-Pem (Post-TKI)

Design: Phase III vs platinum + pemetrexed, post-EGFR TKI (any line)

HER3-DXd 5.6 mg/kg Q3W

PFS: 6.4 vs 3.7 mo (HR 0.77) — significant improvement

ORR: ~30–40% vs 16%

ILD rate: ~26% any grade; ~5% Grade ≥3 — monitoring required; lower rate vs T-DXd 6.4 mg/kg

HK HA: SFI (1+ mechanism)

Ref: Yu HA et al. WCLC 2024 / ELCC 2026.

Viva Cases — EGFR

🎓 Viva Q: A 58-year-old never-smoker with L858R EGFRm NSCLC and 2 brain metastases. PS1. Which 1L regimen and why?
Model Answer: I would choose FLAURA2 (osimertinib + carboplatin + pemetrexed). This patient has L858R mutation — the subgroup where FLAURA2 shows the largest OS benefit (HR 0.74, median 45.7m). Additionally, brain metastases are present — FLAURA2 demonstrated CNS PFS HR 0.47, the most robust CNS control data in this setting. The patient is PS1 and fit for platinum-based chemotherapy. I would ensure adequate antiemetic prophylaxis and CBC monitoring. Local brain treatment (SRS) could be deferred if lesions are small and asymptomatic, as osimertinib has strong CNS penetration.
🎓 Viva Q: Patient on osimertinib 1L progresses after 18 months. ctDNA shows MET amplification (GCN 12 by FISH). What next?
Model Answer: This is MET amplification-driven acquired resistance to osimertinib. The standard approach is to add a MET inhibitor while continuing osimertinib. Based on SACHI (Phase III, HR 0.34, PFS 8.2 vs 4.5m), savolitinib + osimertinib is the evidence-based choice. Tepotinib + osimertinib (INSIGHT-2) is an alternative. I would continue osimertinib 80mg QD and add savolitinib 300mg QD (weight-based). Monitor for peripheral edema, LFT elevation (MET inhibitor class effects). If savolitinib unavailable, amivantamab + chemotherapy (MARIPOSA-2) is the next evidence-based option.
🎓 Viva Q: Post-osimertinib progression — no identified resistance mechanism on NGS. Which ADC to choose and why?
Model Answer: With no identifiable resistance mechanism, this is likely polyclonal or non-canonical resistance. I have three ADC options: (1) Sac-TMT (OptiTROP-Lung04: PFS 8.3m HR 0.49, ORR 60%) — targets TROP2; notable advantage is 0% ILD rate, critical for EGFRm patients already sensitized from prior TKI use. (2) HER3-DXd (HERTHENA-Lung02: PFS 6.4m) — targets HER3 which is expressed in ~80% EGFRm tumors; ILD risk ~5% G≥3. (3) Dato-DXd (TROPION-Lung05: ORR 43.6% EGFR subset) — also TROP2 target; stomatitis is main toxicity (60%). I would favor Sac-TMT based on superior PFS data and near-zero ILD risk, particularly if patient has pre-existing lung comorbidities.
🎓 Viva Q: EGFR exon 20 insertion NSCLC — patient asks about 'targeted therapy'. What do you explain?
Model Answer: I explain that EGFR exon 20 insertions are different from the classic EGFR mutations (exon 19 deletion, L858R). The standard EGFR inhibitors (osimertinib, gefitinib) do not work for exon 20 insertions due to the altered shape of the binding pocket. The current standard first-line treatment is amivantamab (a bispecific antibody targeting both EGFR and MET) combined with platinum-based chemotherapy — the PAPILLON trial showed this extends progression-free survival to 11.4 months vs 6.7 months with chemotherapy alone. This is FDA-approved since August 2024. I also ensure the insertion site is characterized (near-loop vs far-loop) as this may influence sensitivity.
🎓 Viva Q: When should SCLC transformation be suspected and how do you confirm it?
Model Answer: SCLC transformation occurs in ~5–10% of EGFR-mutated NSCLCs after osimertinib exposure, typically after 12–24 months. Suspect it when: (1) rapid clinical deterioration out of proportion to CT findings, (2) new neuroendocrine features on imaging, (3) very high LDH, (4) rapid appearance of new brain/bone metastases. Confirmation requires tissue re-biopsy — ctDNA alone is insufficient as SCLC transformation is a histological diagnosis. IHC will show synaptophysin, chromogranin A, and CD56 positivity with loss of TTF-1 (in some cases) and very high Ki-67 (>70%). Note: the EGFR mutation is retained in the transformed SCLC — this distinguishes it from primary SCLC. Treatment: platinum + etoposide; tarlatamab (DLL3-BiTE) in 2L.
Section 3 — ALK Fusion

ALK+ — lorlatinib, alectinib, or brigatinib? The CROWN 5-year data changed the calculus.

TrialDrugPFSOSNotes
CROWN (5yr)LorlatinibNR (HR 0.19)NR (HR 0.81)60-month PFS rate 60%; best CNS; hyperlipidaemia >90%
ALEX (final)Alectinib34.8m (HR 0.43)81.1m (HR 0.78)Longest mature OS data; photosensitivity, bradycardia
ALTA-1LBrigatinib24.0m (HR 0.49)NR (HR 0.81)Early pulmonary events (days 1–7); start 90mg × 7d then 180mg

Clinical Decision Rules

  • Brain metastases at diagnosis: Lorlatinib preferred (highest CNS activity, CROWN intracranial ORR 82%; intracranial CR rate 71%)
  • No brain metastases: Alectinib (excellent efficacy, mature OS data 81.1m in ALEX final, better CNS toxicity profile); lorlatinib also acceptable
  • Fit patient wanting maximum PFS: Lorlatinib (HR 0.19, PFS not reached at 5 years)
  • Lorlatinib CNS effects: Counsel all patients about cognitive slowing, mood changes (~30%); start statin proactively for hyperlipidaemia (virtually all patients)

Toxicity Profiles

Lorlatinib — Toxicity Profile
  • Hyperlipidaemia: ≥90% — statins required in virtually all patients; monitor fasting lipids baseline and q4-8 weeks
  • Weight gain: Common, often significant (median +5–7 kg); counsel patients proactively
  • CNS effects: Mood changes (depression, anxiety), cognitive slowing, somnolence (~30%); dose-dependent; may require dose reduction or psychiatric support
  • Peripheral neuropathy: Sensorimotor; dose reduction often needed if Grade ≥2
  • Oedema: Peripheral and facial; manage with diuretics if needed
Alectinib — Toxicity Profile
  • LFT elevation: Transaminase rises; monitor monthly × 3 months then periodically; usually asymptomatic
  • Bradycardia: Asymptomatic HR slowing; HR monitoring; avoid negative chronotropes if possible
  • Myalgia/CPK elevation: Monitor CK; dose reduce if Grade ≥3 symptomatic
  • Photosensitivity: Counsel sun protection (SPF 50+, clothing); avoid direct sunlight during treatment
  • Constipation, nausea: Common, manageable with dietary measures
Brigatinib — Toxicity Profile
  • Early-onset pulmonary events: Days 1–7; 3% Grade ≥3 ILD/pneumonitis — withhold if suspected; avoid in patients with baseline ILD; start dose: 90mg QD × 7d, then 180mg QD
  • Hypertension: ~21% Grade ≥3; antihypertensives required; monitor BP regularly
  • CPK elevation: Monitor; dose-reduce if symptomatic myopathy
  • Visual disturbance: Rarely; ophthalmology referral if persistent

Post-ALK TKI Progression

  • Post-alectinib or brigatinib: Lorlatinib (ORR 48% in phase II; broad compound ALK mutation coverage including G1202R); re-biopsy to characterise resistance
  • Post-lorlatinib: Very limited options; clinical trial strongly recommended; platinum + pemetrexed backbone is fallback
  • Chemotherapy: Platinum + pemetrexed remains backbone for lorlatinib-resistant disease; ORR ~20–30%
CROWN 5-Year Update — Lorlatinib vs Crizotinib (1L ALK+)

Design: Phase III, 296 pts, ALK+ Stage IIIB/IV, treatment-naïve

Lorlatinib 100mg QD vs crizotinib 250mg BID

PFS (5yr): NR vs 9.3 mo (HR 0.19, p<0.001)

5-year PFS rate: 60% vs 8%

OS (5yr): NR vs NR (HR 0.81) — not yet significant

Intracranial ORR (pts with baseline brain mets): 82% vs 23%

Intracranial CR rate: 71% vs 8%

Time to intracranial progression (5yr): NR vs 16.4m (HR 0.07)

Ref: Shaw AT et al. NEJM 2020; Solomon BJ et al. J Clin Oncol 2024 (5yr update).

ALEX Final Analysis — Alectinib vs Crizotinib (1L ALK+)

Design: Phase III, 303 pts, ALK+ Stage IV, treatment-naïve

Alectinib 600mg BID vs crizotinib 250mg BID

PFS: 34.8 vs 10.9 mo (HR 0.43, p<0.001)

OS (final, ~8yr follow-up): 81.1 vs 61.5 mo (HR 0.78, p=0.02) — most mature ALK OS data to date

Intracranial ORR (CNS mets): 81% vs 50%

Ref: Peters S et al. NEJM 2017; Mok T et al. Lancet Oncol 2021; OS update ESMO 2023.

Section 4 — ROS1 Fusion

ROS1+ — crizotinib was the original, but entrectinib/repotrectinib are now preferred.

Epidemiology: ~2% NSCLC; shares biological overlap with ALK+ (both bind crizotinib); adenocarcinoma histology; younger never-smokers over-represented; CD74-ROS1 and SLC34A2-ROS1 are common fusion partners

Key Biology: ROS1 fusions constitutively activate downstream proliferative pathways (RAS/MAPK, PI3K/AKT). High baseline ORRs (~70–80%) with TKIs. Primary resistance is less common than EGFR but acquired G2032R solvent-front mutation is the dominant resistance mechanism (~40% of progressions).

DrugTrialORRmPFSCNS ActivityResistance Notes
RepotrectinibTRIDENT-179% (TKI-naïve), 38% (post-criz)~NR (TKI-naïve)Intracranial ORR 89%PREFERRED — macrocyclic; active vs G2032R, L2026M, D2033N
EntrectinibSTARTRK-267%15.7mIntracranial ORR 79%PREFERRED — excellent CNS; FDA/EMA approved; less active vs G2032R
CrizotinibPROFILE 100172%19.3mLimited CNS penetrationHistorical — avoid if CNS disease; G2032R resistance 40%
LorlatinibOff-label35–36%~9m (post-criz)Good CNS penetrationActive in some post-criz resistance; off-label in ROS1

Sequencing Strategy

  • 1L preferred: Repotrectinib (TRIDENT-1: ORR 79% TKI-naïve, PFS not reached; intracranial ORR 89%) — next-gen macrocyclic TKI; active against dominant G2032R resistance mutation; approved FDA 2023
  • 1L alternative: Entrectinib (STARTRK-2: ORR 67%, mPFS 15.7m; intracranial ORR 79%) — preferred if repotrectinib unavailable; excellent CNS coverage
  • Post-crizotinib: Entrectinib or repotrectinib — both active in crizotinib-pretreated patients
  • Post-entrectinib: Repotrectinib — active against G2032R solvent-front mutation (most common acquired resistance in ROS1); ORR 38% in 1-prior TKI setting
  • Lorlatinib: Activity in ROS1+ post-crizotinib (ORR 35–36%) — off-label but a recognised option; shares compound mutations coverage with ALK
  • Post-repotrectinib: Very limited options; clinical trial strongly recommended; platinum-based chemotherapy as fallback
TRIDENT-1 — Repotrectinib in ROS1+ NSCLC

Design: Phase I/II, basket trial, ROS1+ NSCLC, TKI-naïve and TKI-pretreated cohorts

TKI-naïve cohort: ORR 79%; mPFS not reached; intracranial ORR 89% (pts with measurable CNS lesions)

1-prior TKI cohort: ORR 38% (post-crizotinib); active against G2032R (ORR ~48% in G2032R+)

2-prior TKI cohort: ORR 28%

Dose: 160mg QD × 14d then 160mg BID (intra-patient dose escalation design)

Toxicity: Dizziness, dysgeusia, paresthesia (class effect); generally well-tolerated

FDA approved: November 2023 (1L and post-prior ROS1 TKI)

Ref: Drilon A et al. NEJM 2024.

STARTRK-2 — Entrectinib in ROS1+ NSCLC

Design: Phase II basket trial, ROS1+ NSCLC; pooled analysis with STARTRK-1/ALKA-372-001

ORR: 67%; mPFS 15.7m; mOS 44.2m

CNS disease: Intracranial ORR 79% (patients with baseline CNS mets); mCNS DoR 12.9m

Dose: 600mg QD

Toxicity: Dizziness, fatigue, constipation, nausea, weight gain

FDA approved: August 2019 (ROS1+ NSCLC, any line)

Ref: Drilon A et al. Lancet Oncol 2020.

Section 5 — RET Fusion

RET fusion — selpercatinib is the standard. What does the LIBRETTO-431 data show?

Epidemiology: ~2% NSCLC; adenocarcinoma histology; KIF5B-RET most common fusion partner (~70%); younger patients, never/light smokers over-represented; KIF5B-RET has somewhat lower sensitivity to selpercatinib than CCDC6-RET fusions

Biology: RET kinase constitutive activation drives proliferation via MAPK/STAT3/PI3K pathways. Multikinase inhibitors (vandetanib, cabozantinib) were historical but had poor selectivity and high toxicity. Selective RET inhibitors (selpercatinib, pralsetinib) achieve superior efficacy with a cleaner side-effect profile.

DrugTrialSettingORRmPFSmOSCNS ActivityApproval
Selpercatinib 160mg BIDLIBRETTO-431 (Ph III, 1L)1L vs SOC84% vs 65%24.8 vs 11.2m (HR 0.46)ImmatureIntracranial ORR 82%FDA 2024 (1L); 2020 (2L+)
Selpercatinib 160mg BIDLIBRETTO-001 (Ph I/II)2L+ post-platinum64%22.0m44.3mCNS ORR 91%FDA 2020
Pralsetinib 400mg QDARROW (Ph I/II)2L+ post-platinum70%13.3mMaturingCNS ORR ~79%FDA 2020

Selpercatinib — Key Features

  • 1L preferred (LIBRETTO-431): PFS 24.8 vs 11.2m (HR 0.46); ORR 84%. Significant benefit over SOC (pembro ± chemo or chemo alone based on PD-L1). Excellent CNS penetration — intracranial ORR 82%.
  • Toxicity: Hypertension (20% Grade ≥3), LFT elevation (monitor LFTs monthly × 3 months), prolonged QTc (baseline ECG mandatory), peripheral oedema, dry mouth. Avoid strong CYP3A4 inhibitors/inducers.
  • Drug interactions: Selpercatinib is a CYP3A4 substrate — co-medication review essential. Proton pump inhibitors reduce absorption — separate dosing or switch to H2 blocker.

Pralsetinib — Alternative Option

  • ARROW trial: ORR 70% (treatment-naïve), 61% (post-platinum); PFS 13.3m; mOS maturing
  • Toxicity difference: More constipation, less hypertension vs selpercatinib; ILD in ~9%
  • Fasting requirement: Must be taken on an empty stomach (≥1h before or ≥2h after food)

After Selective RET Inhibitor Progression

  • Clinical trial: Strongly preferred — next-gen RET inhibitors (RET G810 solvent-front resistance), bispecific antibodies in development
  • Cross-over (selpercatinib → pralsetinib): May have some activity depending on resistance mechanism; limited data
  • Chemotherapy: Platinum-based backbone; IO has limited activity in RET-fusion NSCLC (immunosuppressive microenvironment)
  • Multikinase inhibitors (vandetanib, cabozantinib): Historical only — significant toxicity, inferior to selective RET inhibitors; use only if selective agents unavailable
🎓 Viva Q: RET-fusion NSCLC, PD-L1 TPS 80%. Patient asks if pembrolizumab monotherapy is appropriate first-line.
Model Answer: No — in RET fusion-positive NSCLC, targeted therapy with a selective RET inhibitor (selpercatinib, LIBRETTO-431) takes absolute priority over immunotherapy regardless of PD-L1 expression. The LIBRETTO-431 trial demonstrated PFS 24.8 vs 11.2m over SOC (which included pembrolizumab). IO response rates are generally poor in RET-fusion NSCLC due to an immunosuppressive tumor microenvironment and low TMB. Starting pembrolizumab in a driver-positive patient also significantly increases the risk of severe immune-related toxicity (particularly pneumonitis) if the patient subsequently receives a TKI.
Section 6 — MET Exon 14 Skipping

MET ex14 skipping — capmatinib or tepotinib 1L. What's the evidence?

Epidemiology: ~3–4% NSCLC; older patients (median age 70–75); sarcomatoid differentiation common (~20%); adenocarcinoma predominant; generally poorer prognosis than EGFR/ALK; often co-occurs with MET amplification (which is a distinct entity)

Biology: MET exon 14 skipping mutations remove the juxtamembrane domain (encoded by exon 14), preventing normal ubiquitin-mediated MET degradation — leading to receptor accumulation and constitutive signalling. Detection: RNA-based NGS is more sensitive (detects splice variants) than DNA-based NGS alone.

DrugTrialSettingORRmPFSCNS ActivityToxicityApproval
Capmatinib 400mg BIDGEOMETRY mono-11L (TKI-naïve)68%12.4mIntracranial ORR 54%Peripheral edema, nausea, LFTFDA 2020; 1L
Tepotinib 500mg QDVISION1L (TKI-naïve)44–46%11.0mIntracranial ORR 55%Peripheral edema, nauseaFDA 2021; 1L
Savolitinib 600mg QDSAVANNAH subset; Asia1L (Asia)~50–60%~9mModerateEdema, LFTNMPA approved; HK available

Treatment Decision Points

  • Capmatinib (GEOMETRY mono-1): ORR 68% treatment-naïve; PFS 12.4m; intracranial ORR 54% — preferred for CNS disease due to CNS penetration data. Available globally.
  • Tepotinib (VISION): ORR 44–46% (tissue + ctDNA cohort); PFS 11.0m; once-daily dosing may improve compliance in elderly patients.
  • Edema management: Both agents cause peripheral edema (20–30% Grade ≥2); prophylactic compression stockings; low-dose diuretics; dose reduce if Grade ≥3

MET amplification ≠ MET exon 14 skipping mutation

MET amplification (high polysomy or gene amplification by FISH) without exon 14 skipping is a distinct biology. Capmatinib/tepotinib have NO approved indication for MET amplification without exon 14 skip. High-level MET amplification (FISH GCN ≥10 or ratio ≥5) may respond but evidence is limited — clinical trial preferred. MET amplification as an EGFR-TKI resistance mechanism (→ SACHI/INSIGHT-2) is a separate scenario covered in the EGFR section.

Post-MET TKI Progression

  • Platinum-based chemotherapy backbone; clinical trial strongly recommended
  • No approved salvage targeted therapy for post-MET TKI progression
  • MET resistance mutations (D1228N, Y1230C/H/S) identified in ~25% progressions — next-gen bivalent MET inhibitors (BMS-1166, elzovantinib) in early clinical trials
  • Teliso-V (Telisotuzumab vedotin, anti-MET ADC): emerging option — METHigh overexpression cohort in METex14 patients (see ADC section)
Section 7 — BRAF V600E

BRAF V600E in NSCLC — dabrafenib + trametinib is standard. New data from 2025–26?

Epidemiology: ~2–3% NSCLC; associated with tobacco smoking (unlike melanoma); adenocarcinoma predominant; median age ~65; V600E accounts for ~50% of BRAF mutations in NSCLC (Class I — constitutively active monomer)

Key distinction: BRAF V600E (Class I) → responds to RAF inhibitor + MEK inhibitor combination. Class II (K601E, L597Q) and Class III (G466V, D594N) → do NOT respond to V600E-targeted therapy; treat per no-driver pathway or enrol in clinical trial.

Drug/RegimenTrialSettingORRmPFSmOSKey Notes
Dabrafenib 150mg BID + Trametinib 2mg QDBRF113928 (Ph II)1L and 2L64% (1L), 63% (2L)10.9m (1L), 10.2m (2L)24.6m (1L)FDA approved NSCLC-specific + tumour-agnostic (V600E)
Dabrafenib + TrametinibESMO 17P / Real-World (2025–26)Post-Osi EGFRm with BRAF V600E acquired resistance42.5% vs chemo8.5 vs 4.8m (HR 0.38)21.2mTriple: Osi + Dab + Tram for acquired resistance
Encorafenib 450mg QD + Binimetinib 45mg BIDGEOMETRY BRAF cohort (investigational)1L (emerging alternative)~60%~11mMaturingNext-gen RAF + MEK; may have lower pyrexia rate

Treatment Sequencing

  • 1L standard: Dabrafenib + trametinib — ORR 64%, PFS 10.9m, OS 24.6m. FDA approved both NSCLC-specific and tissue-agnostic (BRAF V600E)
  • If BRAF targeted therapy not used 1L: Chemo + IO (per PD-L1 TPS) as 1L, then dabrafenib + trametinib in 2L — remains active in pretreated patients (ORR 63% 2L in BRF113928)
  • Acquired BRAF V600E in EGFRm post-osimertinib: Triple therapy — osimertinib + dabrafenib + trametinib (ESMO 17P: ORR 42.5%, PFS 8.5m HR 0.38 vs chemo). This is an emerging resistance approach, not yet standard everywhere.
  • Post-BRAF-targeted progression: Clinical trial strongly recommended. Possible strategies: PI3K/AKT pathway inhibitors, encorafenib ± MEK inhibitor — all largely investigational

Dabrafenib + Trametinib — Toxicity Management

  • Pyrexia: Most common (~50%); typically low-grade fevers; interrupt if Grade ≥2 fever; restart at reduced dose after defervescence
  • Rash / palmar-plantar erythrodysaesthesia: Grade 1–2 common; topical emollients; dose reduce if Grade ≥3
  • Diarrhea, nausea: Manageable; loperamide for diarrhea
  • Uveitis: Rare but important — ophthalmology referral if visual symptoms; hold treatment
  • Hyperglycaemia: Monitor blood glucose; diabetic patients need closer monitoring

Note on BRAF non-V600E mutations (Class II & III): Dabrafenib + trametinib is NOT effective for Class II (K601E, L597Q) or Class III (G466V, D594N) BRAF mutations. These have distinct signalling mechanisms (RAS-independent dimer or kinase-impaired). Treat per no-driver pathway (chemo + IO) or enrol in clinical trial investigating RAF inhibitors effective against non-V600 alterations.

Section 8 — HER2 Alteration

HER2 mutation vs HER2 amplification — different diseases, different treatments.

HER2 Mutation (Exon 20 Insertion — ~2–3% NSCLC)
DrugTrialSettingORRmPFSmOSILD RateFDA Status
T-DXd 5.4 mg/kg Q3WDESTINY-Lung022L+ post-platinum49–56%9.9m19.5m5% (G≥3) ★FDA Accelerated 2022
T-DXd 6.4 mg/kg Q3WDESTINY-Lung012L+55%8.2m17.8m13% (G≥3)Not preferred — higher ILD
Zongertinib (Hernexeos)BEAMION LUNG-1 (Ph Ib)2L+ HER2 exon 20~71%~12mMaturingLowBreakthrough/Priority
Trastuzumab emtansine (T-DM1)Off-label2L+~20–30%~4–5m~10mLowNot approved for NSCLC

★ Use T-DXd 5.4 mg/kg (not 6.4 mg/kg) — superior safety profile with preserved efficacy (DESTINY-Lung02 primary analysis).

ILD monitoring mandatory: Baseline CT chest before starting T-DXd; repeat CT every 6 weeks × 6 months. Hold immediately for any suspected ILD. See FAQ for Grade-based management.

HER2 Amplification / Overexpression (IHC 2+/3+, FISH+)
DrugTrialPopulationORRmPFSmOSNotes
T-DXd 6.4 mg/kgDESTINY-Lung01 Cohort 1HER2 overexp/amp (not mutation)26.5%5.4m12.1mWeaker evidence vs HER2-mutation cohort
Osi + TucatinibORCHARD (HER2 module)EGFRm + HER2 amp post-Osi~20–30%~5.5m13.8mFor EGFR+HER2 acquired resistance

HER2 amplification without HER2 mutation is a distinct biology — lower ORR and shorter PFS. Predictive value of IHC 2+ vs 3+ is still being refined for NSCLC.

Emerging HER2-Targeted Agents in NSCLC

  • Zongertinib (Hernexeos): Selective HER2 TKI (irreversible kinase inhibitor); BEAMION LUNG-1: ORR ~71% in HER2 exon 20 insertion; Breakthrough Designation; potentially more convenient than ADC
  • Sevabertinib: Also selective HER2 TKI targeting exon 20 insertion; early data promising; head-to-head vs zongertinib design under consideration
  • 1L role: No HER2-targeted agent approved 1L yet for NSCLC — standard 1L is chemo ± IO; HER2-targeted therapy introduced at progression
Section 9 — KRAS G12C

KRAS G12C — adagrasib or sotorasib? And should anti-EGFR be added?

Epidemiology: ~13% overall NSCLC (mostly Western populations); adenocarcinoma; strongly associated with tobacco smoking; squamous very rare (<3%); highest frequency in Caucasian populations; ~25% in all KRAS mutations in NSCLC

Biology: KRAS G12C — cysteine-12 substitution creates an allosteric pocket (P2) accessible only in the GDP-bound (inactive) state. Sotorasib and adagrasib covalently bind this pocket, locking KRAS in inactive GDP-bound form. EGFR re-activation is a major bypass mechanism — rationale for adding anti-EGFR agents.

No approved G12C inhibitor in 1L (as of Apr 2026)

Standard 1L treatment = chemo + IO per PD-L1 TPS — identical to no-driver pathway. First-line trials ongoing: KRYSTAL-7 (adagrasib + pembrolizumab vs pembrolizumab in PD-L1 ≥50%), CodeBreak 200 1L arm.

DrugTrialORRmPFS (vs docetaxel)mOSApprovalNotes
Adagrasib 600mg BIDKRYSTAL-1 (Ph I/II)43%6.5m (vs 4.3m, HR 0.66)12.6mFDA 2022 — 2L+CNS active (intracranial ORR 42%)
Sotorasib 960mg QDCodeBreak 200 (Ph III)28%5.6m (HR 0.66 vs docetaxel)10.6mFDA 2021 — 2L+Direct head-to-head vs docetaxel
Adagrasib + cetuximabKRYSTAL-1 Ph Ib46%6.9mMaturingNot approved — investigationalEGFR bypass blockade rationale
Sotorasib + panitumumabCodeBreak 101 Ph Ib30–35%MaturingMaturingNot approved — investigationalPhase III (CAPSTONE) planned

Adagrasib vs Sotorasib — Key Differences

FeatureAdagrasibSotorasib
ORR (2L)43%28%
mPFS6.5m5.6m
CNS penetrationYes — intracranial ORR 42%Limited CNS data
Half-life~23h (BID dosing)~4–5h (QD dosing)
GI toxicityHigh (nausea/diarrhea ~75%)Moderate (~50%)
Hepatotoxicity~15% Grade ≥3 LFT~20% Grade ≥3 LFT
CYP interactionsCYP3A4 inhibitor — many interactionsP-gp substrate
CNS disease preferencePreferredLess data

Co-mutation Impact on Outcomes

  • STK11 co-mutation: Markedly worse outcomes on G12C inhibitors AND IO — median PFS ~3–4m; consider earlier clinical trial enrolment
  • KEAP1 co-mutation: Also predicts poorer IO and G12C inhibitor response; activates NRF2 antioxidant pathway
  • No co-mutations (pure KRAS G12C): Best response rates — prioritise G12C inhibitor after chemo-IO

Emerging — Dual KRAS G12C + SHP2 Inhibition

SHP2 (PTPN11) is a phosphatase upstream of RAS that enables adaptive re-activation of KRAS when G12C inhibitors are used. Adding SHP2 inhibitors (e.g., RMC-4630, JAB-3312) blocks this feedback loop. Early phase I/II data suggest higher ORR with dual inhibition; multiple trials ongoing.

Section 10 — NTRK Fusion

NTRK fusion (<1%) — tissue-agnostic approval. What's the sequencing evidence?

Epidemiology: <1% NSCLC; NTRK1 most common in NSCLC (NTRK2/3 rarer); pan-TRK IHC screening is practical first step — confirm positives by NGS (DNA and/or RNA); adenocarcinoma predominant; younger patients; no strong smoking association

Biology: Neurotrophic receptor tyrosine kinase (NTRK1/2/3) gene fusions → constitutive TRK kinase activation → RAS/MAPK, PI3K/AKT, PLC-γ pathway activation. High ORR with TRK inhibitors — one of the highest biomarker-selected ORRs in oncology. Tissue-agnostic approval reflects remarkably consistent responses regardless of tumour type.

DrugTrialTumour Types (incl NSCLC)ORRmPFSCNS ActivityKey ResistanceApproval
Larotrectinib 100mg BIDNAVIGATE / Phase I/II basketPan-tumour (incl NSCLC)75% (all tumors); ~100% in NTRK+ NSCLC subset28.3mIntracranial ORR 75%NTRK G595R/G639C (kinase domain)FDA 2018 — pan-tumour; NSCLC included
Entrectinib 600mg QDSTARTRK-2 / basketPan-tumour (incl NSCLC, ROS1, ALK)57% (all TRK tumors); 75.8% NSCLC-specific11.2m (TRK); 15.7m (NSCLC)Intracranial ORR 54%NTRK G595R (same as laro)FDA 2019 — pan-tumour; NSCLC included
Repotrectinib 160mgTRIDENT-1 (TRK cohort)Pan-tumour post-TKIORR ~58% post-1-prior TKIMaturingCNS activity — data emergingG595R, G667C active (next-gen)FDA 2023 — 2L+ post-prior TRK TKI
Selitrectinib / CabozantinibPhase I/II (rescue)Post-2 TRK TKI~20–30%~5–6mLimitedSolvent-front mutations G595RNot approved; last-resort

Clinical Pearls

  • Larotrectinib preferred if no CNS disease: Historically highest ORR (~75–100% in NTRK+ NSCLC), durable responses (DoR not reached in many series), well-tolerated
  • Entrectinib preferred if CNS disease: Demonstrated CNS ORR 54% (better CNS penetration than larotrectinib in some series); also covers ROS1 and ALK fusions — useful if diagnosis uncertain
  • Testing strategy: Pan-TRK IHC (clone EPR17341 or A7H6R) as screening — sensitivity ~90%; confirm IHC-positive with RNA-based NGS (detects cryptic fusions DNA-NGS may miss)
  • Post-TRK TKI resistance: Repotrectinib covers G595R and G667C kinase domain mutations — preferred next-gen option
  • Toxicity: Dizziness, fatigue, nausea, constipation — generally mild; dose reductions rarely needed; both agents have paediatric indications (infantile fibrosarcoma) — reflects TRK biology across age groups
Section 11 — ADC Landscape in NSCLC

Antibody-drug conjugates — which target, which payload, which patient?

ADC principles: Antibody–drug conjugates combine a targeting antibody with a cytotoxic payload (linker-payload) for selective cancer cell delivery. The bystander effect — payload diffusion to adjacent cells — is particularly relevant for heterogeneous NSCLC tumors. Dual-payload ADCs (emerging) carry two distinct cytotoxic agents.

Key toxicity classes: ILD/pneumonitis (DXd-payload class effect — monitor carefully); stomatitis (most common with Sac-TMT, ~60%); ocular toxicity (some payloads); haematological toxicity.

Target: TROP2TROP2 expressed in ~60–70% NSCLC; higher in non-squamous
ADCTrialSettingORRmPFSmOSILD RateKey Toxicity
Sacituzumab tirumotecan (Sac-TMT) 5mg/kg Q2WOptiTROP-Lung04 (Ph III)Post-TKI EGFRm vs carbo+pem60% vs 43%8.3 vs 4.3m (HR 0.49)NR vs 17.4m (HR 0.60)0% (notable)Neutropenia, stomatitis
Sacituzumab tirumotecan (Sac-TMT)OptiTROP-Lung03 (Ph III)Post-TKI+chemo EGFRm vs docetaxel45% vs 16%6.9 vs 2.8m (HR 0.30)20.0 vs 11.2m (HR 0.45)0%Neutropenia, stomatitis
Datopotamab deruxtecan (Dato-DXd) 6mg/kg Q3WTROPION-Lung01 (Ph III)2L+ any NSCLC vs docetaxel26.4% vs 12.8%4.4 vs 3.4m (HR 0.75)13.7 vs 11.9m (HR 0.94)12% any grade; ~3% G≥3Stomatitis (60%), nausea
Sacituzumab govitecan (SG) 10mg/kgPhase II (NSCLC)2L+ any NSCLC~15–20%~4.5m~12mLowNeutropenia, diarrhea

Sac-TMT uses a different payload (camptothecin-based CL2A-SN-38 vs DXd in Dato-DXd) — explaining absent ILD signal vs DXd-class agents. TROPION-Lung01 OS benefit marginal; primary PFS endpoint met.

Target: HER2HER2 mutation (~2–3%), overexpression/amplification (~15–20%)
ADCTrialPopulationORRmPFSILD RateStatus
T-DXd 5.4 mg/kg Q3WDESTINY-Lung02HER2 mutation, 2L+49–56%9.9m5% G≥3 — PREFERRED DOSEFDA Accelerated 2022
T-DXd 6.4 mg/kg Q3WDESTINY-Lung01HER2 overexp/amp, 2L+26.5%5.4m13% G≥3Not preferred — higher ILD
Target: METMET overexpression (METHigh by IHC H-score ≥225) in ~25% NSCLC
ADCTrialPopulationORRmPFSKey BiomarkerStatus
Telisotuzumab vedotin (Teliso-V) 1.9mg/kg Q2WLUMINOSITY (Ph II)METHigh non-squamous NSCLC, 2L+35–37%5.4mMET IHC H-score ≥225Breakthrough; FDA review ongoing
Teliso-V + OsimertinibLUMINOSITY Combo cohortEGFRm + METHigh post-Osi (acquired MET OE)~50%~7mMET OE (IHC)Phase II data; emerging option

Teliso-V targets MET overexpression — distinct from MET exon 14 skipping (targeted by capmatinib/tepotinib) or MET amplification (targeted by savolitinib+Osi). Payload = MMAE (tubulin inhibitor); neutropenia and peripheral neuropathy are class effects.

Target: B7-H3B7-H3 highly expressed in NSCLC (~85%) and SCLC transformation
ADCTrialPopulationORRmPFSmOSStatus
Ifinatamab deruxtecan (I-DXd/DS-7300) 12mg/kg Q3WIDYLLIC-01 (Ph II)SCLC (incl EGFR SCLC transformation), heavily pretreated52%5.6m12.2mBreakthrough Designation

I-DXd particularly relevant for SCLC transformation post-EGFR-TKI — SCLC transformation tumors have high B7-H3 expression. See EGFR resistance section.

Target: HER3HER3 expressed in ~80–85% NSCLC; HER3 expression maintained post-TKI
ADCTrialPopulationORRmPFSILD RateStatus
Patritumab deruxtecan (HER3-DXd/MK-1022) 5.6mg/kg Q3WHERTHENA-Lung01 (Ph II)Post-EGFR-TKI EGFRm29.8%5.5m~12% any grade; 2.6% G≥3FDA Breakthrough; Phase III (HERTHENA-Lung02) completed
Patritumab deruxtecan 5.6mg/kgHERTHENA-Lung02 (Ph III)Post-TKI EGFRm vs carbo+pem~30–40%6.4 vs 3.7m (HR 0.77)~5% G≥3Priority Review; SFI HK

ADC Selection Guide — Post-osimertinib EGFRm Context

Clinical ScenarioPreferred ADCAlternativeAvoidReason
Pre-existing ILD / lung fibrosisSac-TMT (0% ILD)HER3-DXd (2.6% G≥3)Dato-DXd, T-DXdDXd payload ILD risk
Unknown resistance / polyclonalSac-TMT (best PFS data)HER3-DXdBroad TROP2 expression; no resistance mutation testing needed
HER2 mutationT-DXd 5.4 mg/kgT-DXd 6.4 mg/kgLower ILD at 5.4 mg/kg; non-inferior efficacy
MET overexpression (METHigh)Teliso-V + OsiTeliso-V monotherapyMET OE as resistance mechanism; add Osi for EGFR coverage
SCLC transformationPlatinum + etoposide (1L)Tarlatamab (2L); I-DXd (2L+)Standard NSCLC regimensHistological transformation — SCLC biology applies
High stomatitis risk / poor oral mucosaHER3-DXd or T-DXdSac-TMT, Dato-DXdStomatitis >60% with TROP2-targeting TROP2 ADCs
Section 12 — Brain Metastases in NSCLC

Brain mets — SRS vs WBRT vs TKI-first? The sequencing question depends on driver status.

Epidemiology: ~20–40% of NSCLC patients develop brain metastases during their disease course; ~10–15% at presentation. Risk higher in adenocarcinoma, EGFR+, ALK+ histologies. Median OS without treatment: 1–3 months (historic); with modern systemic therapy: 10–18 months (driver+), 6–12 months (no driver).

Decision Framework — Initial Management

Driver-positive (EGFR, ALK, ROS1, RET)

  • Osimertinib (EGFR): CNS penetration data superior — FLAURA CNS PFS NR vs 11.2m (HR 0.48); intracranial ORR 91% — deferring upfront cranial RT is supported by evidence for asymptomatic brain mets
  • Lorlatinib (ALK): CROWN intracranial ORR 82%, intracranial CR rate 71%, time to intracranial progression HR 0.07 — preferred ALK TKI for brain mets
  • SRS for 1–3 large/symptomatic lesions alongside TKI; re-assess systemic control before committing to WBRT
  • Defer WBRT in driver+ patients: Neurocognitive toxicity outweighs benefit when TKI controls CNS disease; WBRT reserved for leptomeningeal disease or TKI-refractory CNS

No driver / non-targetable driver

  • 1–4 lesions, each <3 cm: SRS (RTOG 9508; EORTC 22952) — add WBRT only for salvage or >4 lesions; SRS preserves neurocognition
  • 5–10 lesions: SRS increasingly used (Memorial SRS experience); WBRT remains option if SRS technically unfeasible
  • >10 lesions or leptomeningeal disease: WBRT or systemic therapy (IO ± chemo); hippocampal-avoidance WBRT + memantine reduces cognitive decline

SRS vs WBRT — Evidence Summary

TrialPopulationInterventionKey FindingRecommendation
RTOG 95081–3 brain mets, any histologySRS alone vs SRS + WBRTWBRT improved local control but not OS; QoL worse with WBRTSRS alone preferred for 1–3 mets
EORTC 229521–3 brain mets post-resection or SRSObservation vs WBRTWBRT reduced intracranial failure but did NOT improve OS; neurocognitive toxicity significantOmit adjuvant WBRT after SRS/surgery
NCCTG N107C / CEC.3Post-resection 1 metSRS to cavity vs WBRTWorse cognitive decline with WBRT (4.5 vs 3.7m to deterioration); OS identicalSRS to resection cavity preferred
QUARTZ (UK)Multiple brain mets, PS variableWBRT vs optimal supportive careNo OS benefit in poor-PS patients; QoL not improvedWBRT not appropriate for poor-PS non-driver patients
Radiation Necrosis vs Tumour Recurrence — Differentiation

Clinical challenge: Both appear as enhancing lesions on contrast MRI — clinically indistinguishable in ~20% cases

Radiation necrosis clues: 6–24 months post-SRS; "soap bubble" or stippled enhancement on MRI; high ADC value on diffusion-weighted imaging; low perfusion on DSC-MRI; stable or decreasing over time on clinical observation

Recurrence clues: Any time post-RT; progressive enhancement; high perfusion (rCBV >1.5 on DSC-MRI); spectroscopy: elevated choline/NAA ratio; FET-PET: high uptake ratio

Advanced imaging: MRI perfusion (DSC/DCE) + MRS spectroscopy + FET-PET as a panel — sensitivity/specificity for recurrence ~85% with multimodal approach

Management: Grade 1–2 radiation necrosis: bevacizumab 7.5mg/kg Q3W (± dexamethasone) — resolves in ~60% cases. Grade 3+: surgical resection ± laser interstitial thermal therapy (LITT).

Re-SRS criteria: Confirmed recurrence (not necrosis); lesion <3 cm; ≥3–6 months since prior SRS; cumulative dose constraints met; no evidence of leptomeningeal spread. Each centre uses different dose-volume constraints (V12Gy, Dmax limits).

Progression on Osimertinib with Brain Metastases — Management

  • CNS-only progression (extracranial controlled): Continue osimertinib + local CNS treatment (SRS or neurosurgery) — systemic control maintained; re-biopsy not urgently needed
  • CNS + systemic progression: Re-biopsy (tissue or ctDNA) to identify resistance mechanism → guide next systemic treatment (see EGFR resistance section)
  • Leptomeningeal disease (LMD): High-dose osimertinib 160mg QD (off-label; case series support); intrathecal chemotherapy (methotrexate); WBRT; enrolled clinical trial. Prognosis poor — median OS ~3–6m; careful goals-of-care discussion essential
  • Re-SRS for progressive CNS mets on TKI: Can be given for 1–4 new/progressive lesions meeting size/dose criteria. Multiple SRS courses are feasible; cumulative dose-volume constraints must be reviewed with radiation oncology.
🎓 Viva Q: EGFR L858R patient with 3 asymptomatic brain metastases (each 1–1.5 cm). No neurological symptoms. PS1. What is your approach?
Model Answer: In an EGFR-mutated patient with asymptomatic, small brain metastases, I would start systemic osimertinib first and defer upfront cranial radiation. Rationale: FLAURA demonstrated CNS PFS NR vs 11.2m (HR 0.48) with osimertinib, and intracranial ORR of 91% — the vast majority of small asymptomatic lesions respond to osimertinib alone. Upfront radiation adds neurocognitive toxicity without proven survival benefit in this setting. I would perform a baseline brain MRI and repeat at 8–12 weeks to confirm CNS response. If the patient were symptomatic (e.g., oedema, deficits), I would expedite SRS before starting systemic therapy.
🎓 Viva Q: How do you decide between SRS and WBRT for a patient with 5 brain metastases from KRAS G12C NSCLC?
Model Answer: For 5 brain metastases in a KRAS G12C patient (no CNS-active targeted therapy available), the decision depends on lesion size, location, and patient functional status. I would favour SRS for up to 5 lesions each ≤3 cm — large institutional series (including Memorial Sloan Kettering) demonstrate feasibility and preserved neurocognition with SRS for up to 10 lesions. WBRT would be reserved for: lesions too numerous for SRS planning, leptomeningeal involvement, poor PS making SRS sessions impractical, or prior SRS precluding further local treatment. Hippocampal-avoidance WBRT + memantine is the preferred WBRT technique if WBRT is unavoidable — demonstrated reduction in cognitive decline (NRG Oncology CC001).
Section 13 — Oligometastatic / OligoPD NSCLC

Oligometastatic NSCLC — when and how to use local ablative therapy (LAT)?

Definition of Oligometastatic NSCLC (2026 Consensus)

  • Classic oligometastatic (OligoMET): ≤5 metastatic lesions, ≤2–3 organ sites (excluding primary and regional nodes); synchronous (at diagnosis) or metachronous (≥6 months after curative-intent primary treatment)
  • Oligoprogressive (OligoPD): ≤3–5 lesions progressing on systemic therapy while majority of disease remains controlled — the "sweet spot" for local therapy while continuing systemic treatment
  • Local ablative therapy (LAT) synonyms: SABR (stereotactic ablative radiotherapy), SBRT (stereotactic body RT), local consolidative therapy (LCT), local radical therapy (LRT)

Key Trials in Oligometastatic NSCLC

TrialDesignPopulationInterventionKey OutcomeNotes
SABR-COMET (Phase II)RandomisedOligo-mets any histology (1–5 lesions)SABR (all sites) vs standard careOS 41 vs 28m (HR 0.57); PFS 12.0 vs 6.0mMulti-tumour SABR; includes NSCLC (most common histology)
NRG-LU002 (Phase II)RandomisedStage IV NSCLC 1–4 mets, PS 0–2Concurrent LCT (within 30d of systemic) vs sequential LCT (after 3 cycles)Concurrent LCT: PFS benefit; 1-yr PFS 36% vs 28%Concurrent > sequential LAT — 'treat early' principle
ETOP-CHESS (ETOP 14-18, Phase III)RandomisedOligoPD on systemic therapy (driver+ or -)LAT to all PD sites vs systemic switchPending (primary endpoint: PFS); interim data suggest LAT arm numerically superiorFirst Phase III in oligoPD; results awaited 2026
EORTC ICARS (Phase II/III)RandomisedOligometastatic NSCLC ≤3 sitesLAT (SABR/surgery) + SOC vs SOC aloneOS benefit in LAT arm (HR 0.65, p=0.04); PFS benefitValidates LAT benefit in prospective phase II/III — strongest prospective evidence
MDACC Phase II (Gomez 2016)RandomisedStage IV NSCLC ≤3 mets, response/stable on 1L systemicLocal consolidation + maintenance vs maintenance alonePFS 11.9 vs 3.9m (HR 0.35); OS 41.2 vs 17.0m (HR 0.40)Landmark trial — established LCT concept in NSCLC
SINDAS (China Phase III)RandomisedStage IV NSCLC ≤5 mets, EGFR-mutatedEGFR-TKI + SABR vs EGFR-TKI alonePFS 20.2 vs 12.5m (HR 0.62); OS 25.5 vs 17.4m (HR 0.68)Driver-positive oligoMET — LAT + TKI prolongs PFS and OS

Timing of Local Ablative Therapy — The "Sweet Spot"

Clinical evidence converges on a 3–6 month window after systemic therapy initiation:

  • 3–6 months of systemic therapy first: Establish systemic disease control; eliminate occult micro-metastases not visible on baseline imaging; select patients with truly oligometastatic biology (vs those who will develop rapid systemic progression)
  • LAT criteria at 3–6 months: ≤5 active lesions (≤3 strongly preferred); all sites technically addressable by SABR/surgery; systemic disease otherwise controlled or responding; PS 0–2
  • Concurrent LAT (NRG-LU002 protocol): Delivering LAT within 30 days of starting systemic therapy — showed superior PFS vs sequential (3 cycles then LAT). Practical challenge: requires rapid multidisciplinary planning
  • Driver-positive disease (EGFR, ALK): SINDAS shows SABR + TKI superior to TKI alone. For EGFR-mutated oligoMET, combining LAT with osimertinib or alectinib is a valid strategy — multidisciplinary discussion essential

Oligoprogression (OligoPD) on Systemic Therapy

When 1–5 lesions progress on IO, targeted therapy, or chemotherapy while the bulk of disease remains controlled:

  • Continue systemic therapy + add LAT to progressing site(s): This avoids premature systemic treatment switch; exploits the concept that oligoPD lesions may represent focally resistant clones while systemic drug still controls the majority
  • CNS oligoprogression on TKI: The paradigm case — if 1–3 brain lesions progress on osimertinib while extracranial disease is controlled, SRS to those lesions + continue osimertinib is strongly supported by retrospective series and is standard practice
  • Bone oligoprogression: SBRT to 1–3 progressing bone lesions + continue IO/targeted — REMARK trial (bone-targeted SABR) supports this approach
  • Systemic PD: If ≥5 sites progress or new distant organ involvement — systemic treatment switch is required, not LAT

Patient Selection for LAT — Practical Checklist

Favourable for LAT:

  • ≤5 lesions (≤3 preferred)
  • Response/stable on systemic therapy
  • ECOG PS 0–1 (PS 2 selectively)
  • Metachronous >6 months disease-free interval
  • Technically addressable (SABR feasibility confirmed)

Unfavourable for LAT:

  • >5 active lesions
  • Systemic disease uncontrolled/progressing
  • ECOG PS 3–4
  • Leptomeningeal disease
  • Pleural/pericardial effusion dominant
SABR-COMET Phase II — Multi-tumour SABR vs Standard Care

Design: Phase II randomised, 99 pts, ≤5 metastatic lesions (any histology including NSCLC ~35%), oligometastatic intent, PS 0–1

SABR to all metastatic sites + standard care vs standard care alone

OS: 41.0 vs 28.0 mo (HR 0.57, p=0.09 — hypothesis generating)

PFS: 12.0 vs 6.0 mo (HR 0.47, p=0.001)

5-year OS rate: 42.3% vs 17.7%

Safety: 4 Grade 4–5 events in SABR arm — patient selection and technique quality critical

Ref: Palma DA et al. Lancet 2019; 5-yr update 2020.

SINDAS — EGFR-TKI + SABR vs EGFR-TKI (Oligometastatic EGFR+ NSCLC)

Design: Phase III randomised (China), EGFR-mutated Stage IV NSCLC with ≤5 mets, all sites treated with SABR before TKI

EGFR-TKI + SABR to all sites vs EGFR-TKI alone (historical 1st gen TKIs used)

PFS: 20.2 vs 12.5 mo (HR 0.62, p<0.001)

OS: 25.5 vs 17.4 mo (HR 0.68, p=0.026)

Caveats: 1st-gen TKIs used (pre-osimertinib era); modern repeat with osimertinib needed; applicable principle likely still valid

Ref: Wang XS et al. JAMA Oncol 2022.

🎓 Viva Q: A 52-year-old EGFR wild-type (no driver), PD-L1 60%, with 3 metastatic sites (adrenal, bone, lung) has a complete response in the adrenal and bone after 4 cycles of pembrolizumab + carboplatin + pemetrexed, but the 2cm lung nodule persists. What do you do?
Model Answer: This is a classic oligopersistent scenario. The patient has had an excellent systemic response with only one residual active lesion after 4 cycles of chemoimmunotherapy. My approach would be: (1) Continue pembrolizumab maintenance (pemetrexed + pembrolizumab) as per KEYNOTE-189; (2) Refer for SABR/SBRT to the persistent lung lesion — this is the 'sweet spot' for local ablative therapy, converting a near-complete systemic response into what may be a durable remission. Evidence base: the GOMEZ 2016 and SABR-COMET trials support this approach. The risk of SABR to a single lung nodule is low (Grade 3 radiation pneumonitis ~3–5%). I would plan SABR after 1–2 further cycles to confirm non-progression, targeting consolidation at ~week 16–20 of overall treatment.
Section 14 — Second-Line & Beyond

Second-line and beyond: platinum-based salvage, ADCs, and re-challenging IO

At progression: Re-biopsy (tissue or ctDNA) strongly recommended

Both modalities are complementary — ctDNA detects resistance mutations in circulating tumour DNA but misses histological transformation (e.g., SCLC) and focal clonal events. Tissue biopsy of the progressing lesion adds information about histological subtype change, tumour microenvironment, and spatial heterogeneity not captured by liquid biopsy.

2L After Chemo + IO (No Driver)

RegimenTrialPFSOSNotes
Docetaxel 75mg/m² Q3WHistorical standard~3.0m~7.0mORR 10–15%; backbone for combinations
Docetaxel + ramucirumabREVEL4.5 vs 3.0m (HR 0.76)10.5 vs 9.1m (HR 0.86)All histologies; anti-VEGFR2; Grade ≥3 neutropenia 55%
Docetaxel + nintedanibLUME-Lung13.4 vs 2.7m (HR 0.79)12.6 vs 10.3m (HR 0.83, non-sq)Non-squamous only OS benefit; adenocarcinoma subgroup best
Dato-DXd (TROP2 ADC)TROPION-Lung014.4 vs 3.4m (HR 0.75)13.7 vs 11.9m (HR 0.94)ORR 26% vs 13%; emerging standard vs docetaxel
Sac-TMT (TROP2 ADC)OptiTROP-Lung01/03PFS superior to chemoMaturingORR ~40–54%; strong EGFR+ data in OptiTROP-Lung04
Pembrolizumab (IO-naïve)KEYNOTE-0103.9 vs 4.0m10.4 vs 8.5m (PD-L1 ≥1%)PD-L1 ≥1%; IO-naïve 2L only; ORR 18%
NivolumabCheckMate 017/057~2.3–3.5mOS HR ~0.73All-comers; approved regardless of PD-L1
AtezolizumabOAK2.8 vs 4.0m13.8 vs 9.6m (HR 0.73)All-comers; approved 2L+; modest PFS but OS benefit

2L After IO Progression — Types of Progression Matter

Slow PD on IO (≥6 months response then slow growth):

Clinical evidence supports IO rechallenge in some patients after a washout period. The KEYNOTE-024 retreatment data showed ~29% ORR to pembrolizumab rechallenge in long-term responders who subsequently progressed. This remains experimental — discuss multidisciplinary.

Oligoprogression on IO (1–5 sites while rest controlled):

Continue IO + add LAT/SABR to the progressing lesion(s) — standard approach. Supported by retrospective series and the emerging Phase III ETOP-CHESS trial framework.

Systemic PD on IO (rapid progression, multiple new sites):

Switch systemic therapy — docetaxel ± ramucirumab/nintedanib; ADCs (Dato-DXd, Sac-TMT); or clinical trial. Hyperprogression (~10–15%): identify rapidly; poor PS and high baseline LDH are risk factors. Discontinue IO immediately and switch to salvage chemotherapy.

IO Resistance Mechanisms & Emerging Strategies

Resistance CategoryMechanismEmerging Strategy
Primary IO resistanceLow TMB, immunologically cold tumour, STK11/KEAP1 mutationsIO + anti-angiogenic; IO + ADC combinations; KRAS G12C inhibitors (if G12C+)
Acquired IO resistanceLoss of MHC-I antigen presentation, T-cell exhaustion, CAF-mediated immune exclusionDual checkpoint blockade (CTLA-4 + PD-1); TIM-3, LAG-3 inhibitors
Adaptive resistance (IO)IFNγ-induced PD-L1 upregulation, regulatory T-cell expansionAnti-TIGIT (vibostolimab) + IO; bispecific antibodies
IO + antiangiogenic (combination)VEGF suppresses T-cell trafficking; dual blockade may synergiseAtezolizumab + bevacizumab (IMpower150 data); ramucirumab + IO (RELAY data)
SCLC transformationHistological switch in EGFR-mutated NSCLC; NEUROD1/ASCL1 activationPlatinum + etoposide; tarlatamab (DLL3) 2L; IO ± DLL3-targeted

Next-Generation IO Strategies — Emerging (2026)

  • Dual checkpoint blockade (nivolumab + ipilimumab): CheckMate 227 OS benefit (HR 0.79) especially high-TMB (≥10 mut/Mb). CheckMate 9LA: nivo + ipi + 2 cycles chemo vs 4 cycles chemo (OS HR 0.66) — approved 1L all-comers.
  • Anti-TIGIT + PD-1 (vibostolimab + pembrolizumab): MK-7684A Phase III; TIGIT expression common in exhausted T cells in NSCLC — potential synergy with PD-1 blockade.
  • Anti-LAG-3 (relatlimab + nivolumab): Established in melanoma (RELATIVITY-047); NSCLC trials ongoing. LAG-3 expression enriched in PD-L1-low/negative NSCLC — complementary to PD-1 axis.
  • Bispecific antibodies: Ivonescimab (anti-PD-1/VEGF bispecific) — AK112-301 Phase III in China: PFS 11.1 vs 8.3m vs pembrolizumab in PD-L1 ≥1% NSCLC (HR 0.51); HARMONi-2 trial underway globally.
  • mRNA cancer vaccines (BNT111, mRNA-4157): Personalised neoantigen vaccines + pembrolizumab — Phase II data in melanoma maturing; NSCLC trials initiated. Not yet practice-changing.

Hyperprogression on IO — Recognition & Management

Hyperprogression (≥2× increase in tumour growth rate; >50% new lesion burden) occurs in ~10–15% initiated on IO checkpoint inhibitors. Associated with: MDM2/MDM4 amplification, EGFR mutations, poor ECOG PS, high baseline LDH. Identify early with 4–6-week imaging if clinically suspected. Discontinue IO; switch to docetaxel ± ramucirumab or ADC.

REVEL — Docetaxel + Ramucirumab vs Docetaxel (2L, All-NSCLC)

Design: Phase III, 1,253 pts, Stage IV NSCLC progressed after platinum-based chemotherapy (any PD-L1, any driver)

Docetaxel 75mg/m² + ramucirumab 10mg/kg Q3W vs docetaxel + placebo

PFS: 4.5 vs 3.0 mo (HR 0.76, p<0.001)

OS: 10.5 vs 9.1 mo (HR 0.86, p=0.023)

ORR: 23% vs 14%

Grade ≥3 neutropenia: 49% vs 40% — G-CSF support recommended

Benefit across all histologies including squamous (OS HR 0.86)

Ref: Garon EB et al. Lancet 2014.

TROPION-Lung01 — Dato-DXd vs Docetaxel (2L+, Non-Squamous)

Design: Phase III, 604 pts, Stage IV non-squamous NSCLC, ≥1 prior platinum-based therapy

Datopotamab deruxtecan (Dato-DXd) 6mg/kg Q3W vs docetaxel 75mg/m² Q3W

PFS: 4.4 vs 3.4 mo (HR 0.75, p=0.004)

OS: 13.7 vs 11.9 mo (HR 0.94 — not statistically significant)

ORR: 26.4% vs 12.8%

EGFR-mutated subgroup (post-TKI): PFS HR 0.87 — less benefit than non-EGFR population

ILD rate: 10.2% any grade (2.9% Grade ≥3) — lower than T-DXd

Oral mucositis: 34% (any grade) — unique Dato-DXd toxicity

Ref: Awad MM et al. NEJM 2024.

CheckMate 9LA — Nivolumab + Ipilimumab + 2 Cycles Chemo vs 4 Cycles Chemo (1L)

Design: Phase III, 719 pts, Stage IV NSCLC (any PD-L1, any histology, no driver), treatment-naïve

Nivo 360mg Q3W + Ipi 1mg/kg Q6W + carbo/plat + paclitaxel/pem × 2 cycles vs 4 cycles chemo

OS: 15.6 vs 12.2 mo (HR 0.84, p=0.038) at 2yr follow-up

3-year OS rate: 27% vs 19%

PFS: 6.7 vs 5.0 mo (HR 0.67)

ORR: 38% vs 25%

Benefit independent of PD-L1 TPS — advantage over IO monotherapy approaches

Ref: Paz-Ares L et al. NEJM 2021; 3-yr update J Clin Oncol 2022.

Clinical FAQs

What is the preferred 1L treatment for EGFR L858R vs Ex19del?
For L858R: consider FLAURA2 (osimertinib + carboplatin + pemetrexed; OS 47.5m) or MARIPOSA (amivantamab + lazertinib). The OS benefit from adding chemotherapy is largest in L858R. For Ex19del: osimertinib monotherapy is generally sufficient (FLAURA OS 38.6m) — adding chemo adds toxicity without the same magnitude of OS benefit. CNS disease or high tumour burden favours an intensified regimen regardless of mutation subtype.
Is lorlatinib always better than alectinib for ALK+ NSCLC?
CROWN 5-year data shows superior PFS (NR vs 10.9m, HR 0.19) but no OS advantage over alectinib to date. Lorlatinib carries more CNS toxicity (mood changes, cognitive slowing) and requires statin use in nearly all patients. For CNS-disease-free patients, alectinib is a well-tolerated option with mature OS data (81.1m in ALEX final). Lorlatinib is preferred for baseline brain metastases given dominant intracranial activity (ORR 82%).
What comes after osimertinib at progression?
Re-biopsy (tissue or ctDNA) is mandatory before choosing salvage treatment. MET amplification (~25%) → add savolitinib (SACHI: PFS 8.2 vs 4.5m, HR 0.34) or tepotinib (INSIGHT-2). HER2 amplification → T-DXd (DESTINY-Lung02: ORR 49–56%). SCLC transformation → platinum + etoposide; tarlatamab for 2L SCLC. Unknown mechanism or polyclonal resistance → amivantamab + carboplatin + pemetrexed (MARIPOSA-2: PFS 6.3 vs 4.2m, HR 0.48); or ADCs: OptiTROP Sac-TMT (PFS 8.3m), HER3-DXd (PFS 6.4m), Dato-DXd.
When can IO be used in driver-positive NSCLC?
IO monotherapy is generally not recommended in driver-positive NSCLC (EGFR, ALK, ROS1, RET, MET) — it is inferior to targeted therapy in 1L and significantly increases ILD risk after subsequent TKI use (especially EGFR-TKI after IO). The sole exception is IMpower150 (atezolizumab + bevacizumab + carboplatin + paclitaxel), which showed activity in EGFR/ALK+ patients who had progressed on targeted therapy — this combination avoids direct TKI + IO pairing.
Does PD-L1 matter in KRAS G12C NSCLC?
PD-L1 TPS determines the 1L IO strategy (standard chemo + IO per PD-L1 TPS algorithm — same as no-driver pathway). PD-L1 TPS does not predict KRAS G12C inhibitor response in 2L; adagrasib and sotorasib response rates are similar across PD-L1 subgroups. STK11 and KEAP1 co-mutations predict markedly worse outcomes on both IO-based regimens and G12C inhibitors — these patients should be prioritised for clinical trials.
How do I manage ILD with T-DXd (trastuzumab deruxtecan)?
Grade 1 ILD: hold T-DXd immediately; start systemic corticosteroids (prednisolone 1 mg/kg/day); monitor with CT chest and pulse oximetry; consider restarting at reduced dose (4.4 mg/kg) once resolved to Grade 0. Grade 2: hold; systemic corticosteroids ≥4 weeks tapering; resume only if Grade ≤1 at reassessment. Grade 3–4 or recurrent Grade 2: permanently discontinue T-DXd. Baseline CT chest before starting; repeat CT every 6 weeks × 6 months then per clinical judgement.
Is a KRAS G12C inhibitor approved first-line?
Not yet (as of April 2026). Both adagrasib and sotorasib are approved 2L+ after at least one prior platinum-based chemotherapy. First-line trials are ongoing: KRYSTAL-7 (adagrasib + pembrolizumab vs pembrolizumab monotherapy in PD-L1 ≥50%); CodeBreak 200 1L arm. Current 1L standard for KRAS G12C NSCLC: chemo + IO per PD-L1 TPS — identical to no-driver pathway.
Can IO be combined with targeted therapy in NSCLC?
Generally no — combining EGFR-TKI with checkpoint inhibitors substantially increases hepatotoxicity and ILD risk. The CAURAL trial (osimertinib + durvalumab) was terminated early due to excess ILD (38% vs 15% any grade). Combining ALK-TKI with IO similarly increases pneumonitis risk. The exception is IMpower150 (atezolizumab + bevacizumab + carboplatin + paclitaxel) — this is a chemo-IO-antiangiogenic combination that does not include a TKI, and has demonstrated activity in EGFR/ALK+ post-targeted therapy.
When is EGFR exon 20 insertion treated differently from classic EGFR mutations?
EGFR exon 20 insertion mutations are fundamentally different — standard EGFR TKIs (osimertinib, gefitinib, erlotinib, afatinib) have minimal activity. The first-line standard is amivantamab + platinum + pemetrexed (PAPILLON: PFS 11.4 vs 6.7m, HR 0.40; FDA approved August 2024). Mobocertinib (ORR 28%) and amivantamab monotherapy (ORR 40%) are 2L options. T-DXd-based regimens are also being investigated. Always confirm the specific exon 20 variant — some insertions near the C-helix may have different drug sensitivities.
What is the definition of oligometastatic NSCLC and when should local ablative therapy be given?
Oligometastatic NSCLC is defined as ≤5 metastatic lesions in ≤2–3 organ sites, with controlled primary disease. Local ablative therapy (LAT/SABR) is indicated after systemic therapy demonstrates disease control — the 'sweet spot' is 3–6 months after starting systemic therapy if ≤5 lesions remain active. NRG-LU002 showed PFS benefit with concurrent vs sequential LAT. For EGFR-mutated oligoMET, SINDAS (Phase III) demonstrated SABR + TKI superior to TKI alone (PFS 20.2 vs 12.5m, HR 0.62; OS 25.5 vs 17.4m, HR 0.68).
When should WBRT be preferred over SRS for brain metastases in NSCLC?
SRS is preferred for 1–4 brain metastases each <3 cm — preserves neurocognition (RTOG 9508, EORTC 22952, NCCTG N107C). WBRT may be preferred for >10 lesions, leptomeningeal disease (though intrathecal chemotherapy is often concurrent), or when SRS delivery is technically unfeasible. If WBRT is unavoidable, hippocampal-avoidance WBRT + memantine (NRG CC001) reduces cognitive decline. For driver-positive NSCLC with good CNS penetration (osimertinib, lorlatinib), deferring upfront radiation and starting TKI first is supported — monitor with brain MRI at 8–12 weeks.
How do sotorasib and adagrasib compare for KRAS G12C NSCLC?
CodeBreak 200 (sotorasib vs docetaxel): ORR 28%, PFS 5.6m, OS 10.6m (HR 0.85, not significant for OS). KRYSTAL-1 (adagrasib): ORR 43%, PFS 6.5m, OS 12.6m. Head-to-head comparison is lacking. Adagrasib has numerically higher ORR and slightly longer PFS, potentially explained by greater depth of KRAS G12C target engagement. Adagrasib has CNS activity (ORR ~33% in NSCLC with untreated brain mets). Both have similar mechanism: covalent irreversible binding to GDP-bound KRAS G12C. Toxicity: adagrasib carries QTc prolongation risk and higher GI toxicity rate; sotorasib has more hepatotoxicity (Grade ≥3 ALT elevation ~9%). Combination with anti-EGFR is investigational but mechanistically rational.
Which ADC is preferred for post-EGFR-TKI NSCLC in 2026?
Multiple ADCs are emerging for post-osimertinib NSCLC: (1) Sacituzumab tirumotecan (Sac-TMT, TROP2-targeted): OptiTROP-Lung04 showed PFS 8.3 vs 4.3m vs docetaxel (ORR 54%); regulatory submission underway. (2) Patritumab deruxtecan (HER3-DXd): HERTHENA-Lung02 PFS 6.4 vs 3.7m vs platinum + pem (ORR 30%); registration trial. (3) Datopotamab deruxtecan (Dato-DXd): TROPION-Lung05 (EGFR-mutated subgroup) ORR 37%, PFS 8.2m — single arm. (4) Amivantamab + carboplatin + pemetrexed (MARIPOSA-2): PFS 6.3 vs 4.2m (HR 0.48). Currently, MARIPOSA-2 regimen and OptiTROP Sac-TMT have the strongest randomised data for this setting.

Clinical Calculator Hub

Clinical Disclaimer: These guidelines are intended for qualified healthcare professionals. Content reflects NCCN NSCLC 3.2025 and ESMO Living Guideline v1.3 Feb 2026. Treatment decisions must be individualised; consult local formulary, funding pathways, and the full trial data before prescribing. Last reviewed April 2026.