Skip to content

Lung squamous cell carcinoma — failed and frontier targets

TL;DR — LUSC target development has repeatedly converted frequent alterations into weak clinical selectors. FGFR1 amplification produced response rates of 11.5% with dovitinib and 7% with AZD4547; PIK3CA alteration produced 5% response with taselisib; cell-cycle amplification produced 6% with palbociclib (Lim 2016, PMID 27315356; Aggarwal 2019, PMID 31195180; Langer 2019, PMID 31158500; Edelman 2019, PMID 31302234). Across Lung-MAP S1400, only 10 of 143 biomarker-matched patients (7.0%) responded to targeted therapy, versus 53 of 315 (16.8%) receiving checkpoint therapy while immunotherapy-naïve (Redman 2020, PMID 33125909). The recurring problem is biomarker imprecision: copy gain, pathway membership, and a nominal mutation often fail to establish clonal functional dependency. A stronger frontier is mechanism-defined state targeting—FGFR1 rearrangement architecture, NRF2-driven metabolism, synthetic lethality, and antigen-directed antibody-drug conjugates (ADCs)—but none yet constitutes an approved LUSC-specific genotype therapy (Malchers 2023, PMID 37606995; Paik 2023, PMID 36240971). The drug-development lesson is constructive: rapid futility decisions, centralized sequencing, non-match arms, and biomarker refinement are outputs even when drugs fail.

The target paradox in numbers

Target/strategy Biomarker Trial result Development inference
Dovitinib FGFR1 >5 copies by FISH ORR 11.5% (95% CI 0.8–23.8); median PFS 2.9 mo Amplification weakly enriches dependency; Lim 2016, PMID 27315356
AZD4547 (S1400D) FGFR1/3 amplification, FGFR3 mutation/fusion ORR 7% (0–17); PFS 2.7 mo; OS 7.5 mo One response each in FGFR3 S249C and FGFR1 amplification; Aggarwal 2019, PMID 31195180
Taselisib (S1400B) PIK3CA alteration ORR 5% (0–24); PFS 2.9 mo; OS 5.9 mo Closed for futility; two possibly treatment-related deaths; Langer 2019, PMID 31158500
Palbociclib (S1400C) CCND1/2/3 or CDK4 amplification ORR 6% (0–15); PFS 1.7 mo; OS 7.1 mo Failed phase-III advancement criterion; Edelman 2019, PMID 31302234
All S1400 targeted arms Multiple matched alterations 10/143 responses (7.0%); PFS 2.5 mo; OS 5.9 mo Master protocol efficient; biomarkers/drugs ineffective as deployed; Redman 2020, PMID 33125909
TAK-228 NFE2L2-mutant LUSC ORR 25%; median PFS 8.9 mo Small phase-II signal for metabolic-state targeting; Paik 2023, PMID 36240971

These single-arm results cannot be compared naively across eras and cohorts. Their consistent low response under pre-specified futility rules nevertheless falsifies the proposition that broad amplification/pathway labels are sufficient selectors.

FGFR1: from 22% prevalence to single-digit response

The original study found FGFR1 amplification in 22% of an independent LUSC cohort and showed selective cell-line and xenograft sensitivity (Weiss 2010, PMID 21160078). Translation then exposed four gaps:

Gap Evidence Consequence
Assay definition FISH cutoffs, copy-number gain, amplification ratio, and broad 8p amplicon differ “FGFR1-positive” populations are heterogeneous; Jiang 2015, PMID 25433983
Expression FGFR1 mRNA predicted PDX response better than amplification DNA copy number may not produce functional receptor; Weeden 2017, PMID 28611104
Amplicon architecture Tail-to-tail FGFR1 rearrangements/ectodomain deletions mark dependency Structural context may outperform copy count; Malchers 2023, PMID 37606995
Co-dependency/resistance Alternate receptor tyrosine kinases bypass FGFR blockade Combination or narrower selection required; Adachi 2017, PMID 28968756

Dovitinib treated 26 amplified tumors and produced three partial responses; ORR was 11.5%, median PFS 2.9 months, and median OS 5.0 months. Grade ≥3 fatigue occurred in 19.2%, anorexia 11.5%, and hyponatremia 11.5% (Lim 2016, PMID 27315356).

S1400D assigned 92 patients, enrolled 43, and treated 27 evaluable patients. Twenty-three (85%) had FGFR1 amplification; only two responses occurred, with durations 1.5 and 2.9 months (Aggarwal 2019, PMID 31195180).

Deep analysis of 52 8p11–12-amplified tumors found ectodomain-deficient FGFR1 variants caused by rearrangements. FGFR1-centered tail-to-tail architecture correlated with dependency, offering a testable refinement of the failed “any amplification” biomarker (Malchers 2023, PMID 37606995).

PI3K pathway: frequent, redundant, toxic

TCGA found PI3K-pathway alterations in 47%, but that category combines PIK3CA mutation/amplification with PTEN and other pathway events (TCGA 2012, PMID 22960745). Functional redundancy among PI3K isoforms and parallel MAPK/GLI signaling can buffer single-agent inhibition (Stamatkin 2015, PMID 26176612; Kasiri 2017, PMID 28652248).

S1400B enrolled PIK3CA-altered tumors; one of 21 primary-analysis patients responded (5%, 95% CI 0–24). Median PFS was 2.9 months, and two deaths were considered possibly treatment-related (Langer 2019, PMID 31158500). Frequency therefore did not overcome modest therapeutic index.

Preclinical combination strategies include:

Combination Model finding Translation barrier
PI3K antagonist + GLI1 blockade Reduced growth in PIK3CA-amplified LUSC models GLI1 activation was noncanonical; no validated clinical selector
Dual mTORC1/2–DNA-PK inhibitor + platinum/taxane Synergy in 6/14 NSCLC lines, enriched in PIK3CA-mutant LUSC; PDX survival gain PI3K/DNA-PK toxicity and complex biomarker
Isoform-specific PI3K inhibition Functional compensation among class-IA isoforms Single-isoform blockade may be insufficient

Cell-cycle matching

TCGA found CDKN2A/RB1-pathway alteration in 72%, a pattern suggesting frequent cell-cycle dysregulation but not necessarily CDK4/6 dependence (TCGA 2012, PMID 22960745). RB loss can itself make CDK4/6 inhibition irrelevant. S1400C instead selected CCND1/2/3 or CDK4 amplification; 26 of 32 palbociclib-treated tumors had CCND1 amplification, yet only two responded and median PFS was 1.7 months (Edelman 2019, PMID 31302234).

The biomarker did not verify intact downstream RB, clonal amplification, cyclin-D protein dependence, or absence of bypass. A future selector would need functional pathway competence rather than an upstream copy event.

DDR2 and dasatinib

DDR2 mutations occurred in approximately 3.8% of discovery specimens; mutant models responded to knockdown/dasatinib, and an index patient responded to dasatinib plus erlotinib (Hammerman 2011, PMID 22328973).

Acquired resistance emerged through a DDR2 gatekeeper mutation and NF1 loss in models, showing both altered target binding and MAPK bypass (Beauchamp 2014, PMID 24296828). Clinically, low prevalence, variant heterogeneity, and dasatinib toxicity prevented rapid preclinical promise from becoming a standard.

EGFR/ERBB-directed therapy: statistically positive, biologically diffuse

Necitumumab is a squamous-specific regulatory precedent but not genotype-matched precision therapy. In SQUIRE, adding EGFR antibody necitumumab to cisplatin/gemcitabine produced a modest OS benefit in stage-IV LUSC (Thatcher 2015, PMID 26045340). EGFR expression did not become a robust high-resolution selector (Paz-Ares 2016, PMID 27207107).

LUX-Lung 8 compared pan-ERBB afatinib with EGFR inhibitor erlotinib after platinum therapy. Final median OS was 7.8 versus 6.8 months (HR 0.84, 95% CI 0.73–0.97), a one-month gain in a pre-IO setting (Goss 2021, PMID 34195574).

LUX-Lung 8 biomarker observation Estimate Why not practice-defining
Any ERBB mutation in sequenced subset 53/245 (21.6%) Post hoc, enriched sample subset
Afatinib PFS, ERBB-mutant vs wild-type 4.9 vs 3.0 mo; HR 0.62 (0.37–1.02) CI crossed 1; no randomized biomarker interaction
Afatinib OS, ERBB-mutant vs wild-type 10.6 vs 8.1 mo; HR 0.75 (0.47–1.17) Exploratory
Copy number/EGFR expression No apparent outcome association Weakens broad EGFR-expression selection

These data suggest rare ERBB-mutant biology may be different, but do not establish afatinib as a contemporary post-chemo-IO biomarker standard (Goss 2018, PMID 29902295).

Lung-MAP as an experiment in infrastructure

Lung-MAP S1400 enrolled previously treated stage-IV/recurrent LUSC into centralized NGS screening and biomarker/non-match substudies. Of 1,864 enrolled, 1,841 (98.9%) submitted tissue, 1,674 (90.9%) obtained biomarker results, 1,404 (83.9%) received an assignment, and 655 (46.7%) registered to a substudy (Redman 2020, PMID 33125909).

S1400 output Number Interpretation
Targeted-therapy response 10/143 (7.0%) Broad matched strategy unsuccessful
IO-naïve checkpoint response 53/315 (16.8%) Immune therapy outperformed matched targeted set
Docetaxel response 3/56 (5.4%) Historical control activity low
Targeted median PFS/OS 2.5/5.9 mo Rapid futility observable
IO-containing median PFS/OS 3.0/10.8 mo Durable benefit concentrated in minority

The platform succeeded operationally even when its drugs failed. It reduced serial screening, permitted reassignment, created a 1,672-patient genomic resource, and adapted toward all NSCLC histologies and immunotherapy-resistance questions (Redman 2020, PMID 33125909; Kozono 2024, PMID 39111731).

Checkpoint-combination failures are target failures too

Adding CTLA-4 blockade after prior chemotherapy but before routine first-line IO did not improve nivolumab outcomes in S1400I: OS HR 0.87 (95% CI 0.66–1.16), response 18% vs 17%, and grade ≥3 treatment-related toxicity 39.5% vs 33.3% (Gettinger 2021, PMID 34264316).

After acquired or primary PD-(L)1 resistance, durvalumab/tremelimumab produced a 7% response in the primary-resistance cohort and none in acquired resistance; median PFS was approximately two months (Leighl 2021, PMID 34429332). “More checkpoint blockade” is therefore not a generic resistance solution.

NRF2/metabolic state targeting

TAK-228, a TORC1/2 inhibitor, showed its strongest phase-II activity in NFE2L2-mutant LUSC: ORR 25% and median PFS 8.9 months (Paik 2023, PMID 36240971). Preclinical data suggested glutaminase inhibition might overcome metabolic resistance.

This is the most conceptually important positive signal because it selects a downstream metabolic state rather than merely a common copy-number event. It remains a small nonrandomized cohort and requires confirmation before being called effective targeted therapy.

3q driver kinases and synthetic lethality

The 3q amplicon contains TP63, SOX2, PIK3CA, PRKCI, PAK2, MAP3K13, and TNIK. Several are amplified in more than 20% of squamous cancers across organs and may create co-dependencies (Bensen 2021, PMID 33799513).

The challenge is that co-amplification makes it difficult to identify the dominant dependency. A rational program should combine CRISPR/functional screening, protein activity, clonal architecture, and pharmacodynamic proof rather than selecting by locus gain alone.

ADC frontier

ADCs replace intracellular oncogene addiction with surface-antigen delivery. Candidate antigens in NSCLC include TROP2, MET, CEACAM5, B7-H3, and others; expression is continuous and spatially heterogeneous.

In ICARUS-LUNG01, 100 pretreated NSCLC patients receiving datopotamab deruxtecan had ORR 26.0% and median PFS 3.6 months, with greater benefit in nonsquamous tumors. Lack of TROP2 cytoplasmic staining and early DNA-repair activation were candidate resistance markers (Planchard 2026, PMID 41999747). The histology interaction cautions against assuming a pan-NSCLC ADC will solve the LUSC gap.

TROPION-Lung07 specifically studies nonsquamous NSCLC, demonstrating a persistent pipeline asymmetry rather than LUSC evidence (Okamoto 2024, PMID 39469838). LUSC-specific ADC development needs adequate enrollment and antigen validation rather than extrapolation.

A better biomarker hierarchy

Evidence level Example Confidence
Genomic presence FGFR1 amplification Necessary for discovery, insufficient for therapy
Transcription/protein FGFR1 mRNA/protein Closer to function; still not dependency
Structural architecture FGFR1 ectodomain deletion/tail-to-tail amplicon Mechanistically sharper
Functional dependency Knockout/drug rescue in patient-derived model Strong preclinical evidence
Pharmacodynamic engagement On-treatment pathway suppression Shows drug reaches target
Randomized treatment interaction Biomarker-positive benefit vs control and biomarker-negative contrast Required for predictive validation

Most failed LUSC programs entered trials at the first level; the frontier must climb the hierarchy before large expansion cohorts.

Open questions

  • Will FGFR1 rearrangement architecture or expression rescue FGFR inhibition from amplification-based failure (Weeden 2017, PMID 28611104; Malchers 2023, PMID 37606995)?
  • Can the TAK-228 NFE2L2 signal replicate in a randomized cohort and distinguish prediction from poor-prognosis enrichment (Paik 2023, PMID 36240971)?
  • Which surface antigen and payload produce a favorable LUSC therapeutic index after chemo-IO (Planchard 2026, PMID 41999747)?
  • Can functional assays be completed fast enough for master-protocol assignment without excluding rapidly progressing patients (Redman 2020, PMID 33125909)?
  • Should future LUSC trials be organized by cross-organ squamous state rather than lung site alone (Bensen 2021, PMID 33799513)?

References

  1. Weiss J, Sos ML, Seidel D, Peifer M, Zander T, Heuckmann JM, et al. Frequent and focal FGFR1 amplification associates with therapeutically tractable FGFR1 dependency in squamous cell lung cancer. Sci Transl Med. 2010;2(62):62ra93. PMID 21160078
  2. Jiang T, Gao G, Fan G, Li M, Zhou C. FGFR1 amplification in lung squamous cell carcinoma: a systematic review with meta-analysis. Lung Cancer. 2015;87(1):1-7. PMID 25433983
  3. Lim SH, Sun JM, Choi YL, Kim HR, Ahn S, Lee JY, et al. Efficacy and safety of dovitinib in pretreated patients with advanced squamous non-small cell lung cancer with FGFR1 amplification: A single-arm, phase 2 study. Cancer. 2016;122(19):3024-31. PMID 27315356
  4. Aggarwal C, Redman MW, Lara PN Jr, Borghaei H, Hoffman P, Bradley JD, et al. SWOG S1400D (NCT02965378), a Phase II Study of the Fibroblast Growth Factor Receptor Inhibitor AZD4547 in Previously Treated Patients With Fibroblast Growth Factor Pathway-Activated Stage IV Squamous Cell Lung Cancer (Lung-MAP Substudy). J Thorac Oncol. 2019;14(10):1847-1852. PMID 31195180
  5. Weeden CE, Holik AZ, Young RJ, Ma SB, Garnier JM, Fox SB, et al. Cisplatin Increases Sensitivity to FGFR Inhibition in Patient-Derived Xenograft Models of Lung Squamous Cell Carcinoma. Mol Cancer Ther. 2017;16(8):1610-1622. PMID 28611104
  6. Malchers F, Nogova L, van Attekum MH, Maas L, Brägelmann J, Bartenhagen C, et al. Somatic rearrangements causing oncogenic ectodomain deletions of FGFR1 in squamous cell lung cancer. J Clin Invest. 2023;133(21). PMID 37606995
  7. Adachi Y, Watanabe K, Kita K, Kitai H, Kotani H, Sato Y, et al. Resistance mediated by alternative receptor tyrosine kinases in FGFR1-amplified lung cancer. Carcinogenesis. 2017;38(11):1063-1072. PMID 28968756
  8. Cancer Genome Atlas Research Network. Comprehensive genomic characterization of squamous cell lung cancers. Nature. 2012;489(7417):519-25. PMID 22960745
  9. Langer CJ, Redman MW, Wade JL 3rd, Aggarwal C, Bradley JD, Crawford J, et al. SWOG S1400B (NCT02785913), a Phase II Study of GDC-0032 (Taselisib) for Previously Treated PI3K-Positive Patients with Stage IV Squamous Cell Lung Cancer (Lung-MAP Sub-Study). J Thorac Oncol. 2019;14(10):1839-1846. PMID 31158500
  10. Stamatkin C, Ratermann KL, Overley CW, Black EP. Inhibition of class IA PI3K enzymes in non-small cell lung cancer cells uncovers functional compensation among isoforms. Cancer Biol Ther. 2015;16(9):1341-52. PMID 26176612
  11. Kasiri S, Shao C, Chen B, Wilson AN, Yenerall P, Timmons BC, et al. GLI1 Blockade Potentiates the Antitumor Activity of PI3K Antagonists in Lung Squamous Cell Carcinoma. Cancer Res. 2017;77(16):4448-4459. PMID 28652248
  12. Castellano GM, Zeeshan S, Garbuzenko OB, Sabaawy HE, Malhotra J, Minko T, et al. Inhibition of Mtorc1/2 and DNA-PK via CC-115 Synergizes with Carboplatin and Paclitaxel in Lung Squamous Cell Carcinoma. Mol Cancer Ther. 2022;21(9):1381-1392. PMID 35732569
  13. Edelman MJ, Redman MW, Albain KS, McGary EC, Rafique NM, Petro D, et al. SWOG S1400C (NCT02154490)-A Phase II Study of Palbociclib for Previously Treated Cell Cycle Gene Alteration-Positive Patients with Stage IV Squamous Cell Lung Cancer (Lung-MAP Substudy). J Thorac Oncol. 2019;14(10):1853-1859. PMID 31302234
  14. Hammerman PS, Sos ML, Ramos AH, Xu C, Dutt A, Zhou W, et al. Mutations in the DDR2 kinase gene identify a novel therapeutic target in squamous cell lung cancer. Cancer Discov. 2011;1(1):78-89. PMID 22328973
  15. Beauchamp EM, Woods BA, Dulak AM, Tan L, Xu C, Gray NS, et al. Acquired resistance to dasatinib in lung cancer cell lines conferred by DDR2 gatekeeper mutation and NF1 loss. Mol Cancer Ther. 2014;13(2):475-82. PMID 24296828
  16. Thatcher N, Hirsch FR, Luft AV, Szczesna A, Ciuleanu TE, Dediu M, et al. Necitumumab plus gemcitabine and cisplatin versus gemcitabine and cisplatin alone as first-line therapy in patients with stage IV squamous non-small-cell lung cancer (SQUIRE): an open-label, randomised, controlled phase 3 trial. Lancet Oncol. 2015;16(7):763-74. PMID 26045340
  17. Paz-Ares L, Socinski MA, Shahidi J, Hozak RR, Soldatenkova V, Kurek R, et al. Correlation of EGFR-expression with safety and efficacy outcomes in SQUIRE: a randomized, multicenter, open-label, phase III study of gemcitabine-cisplatin plus necitumumab versus gemcitabine-cisplatin alone in the first-line treatment of patients with stage IV squamous non-small-cell lung cancer. Ann Oncol. 2016;27(8):1573-9. PMID 27207107
  18. Goss GD, Cobo M, Lu S, Syrigos K, Lee KH, Göker E, et al. Afatinib versus erlotinib as second-line treatment of patients with advanced squamous cell carcinoma of the lung: Final analysis of the randomised phase 3 LUX-Lung 8 trial. EClinicalMedicine. 2021;37:100940. PMID 34195574
  19. Goss GD, Felip E, Cobo M, Lu S, Syrigos K, Lee KH, et al. Association of ERBB Mutations With Clinical Outcomes of Afatinib- or Erlotinib-Treated Patients With Lung Squamous Cell Carcinoma: Secondary Analysis of the LUX-Lung 8 Randomized Clinical Trial. JAMA Oncol. 2018;4(9):1189-1197. PMID 29902295
  20. Redman MW, Papadimitrakopoulou VA, Minichiello K, Hirsch FR, Mack PC, Schwartz LH, et al. Biomarker-driven therapies for previously treated squamous non-small-cell lung cancer (Lung-MAP SWOG S1400): a biomarker-driven master protocol. Lancet Oncol. 2020;21(12):1589-1601. PMID 33125909
  21. Kozono D, Hua X, Wu MC, Tolba KA, Waqar SN, Dragnev KH, et al. Lung-MAP Next-Generation Sequencing Analysis of Advanced Squamous Cell Lung Cancers (SWOG S1400). J Thorac Oncol. 2024;19(12):1618-1629. PMID 39111731
  22. Gettinger SN, Redman MW, Bazhenova L, Hirsch FR, Mack PC, Schwartz LH, et al. Nivolumab Plus Ipilimumab vs Nivolumab for Previously Treated Patients With Stage IV Squamous Cell Lung Cancer: The Lung-MAP S1400I Phase 3 Randomized Clinical Trial. JAMA Oncol. 2021;7(9):1368-1377. PMID 34264316
  23. Leighl NB, Redman MW, Rizvi N, Hirsch FR, Mack PC, Schwartz LH, et al. Phase II study of durvalumab plus tremelimumab as therapy for patients with previously treated anti-PD-1/PD-L1 resistant stage IV squamous cell lung cancer (Lung-MAP substudy S1400F, NCT03373760). J Immunother Cancer. 2021;9(8). PMID 34429332
  24. Paik PK, Fan PD, Qeriqi B, Namakydoust A, Daly B, Ahn L, et al. Targeting NFE2L2/KEAP1 Mutations in Advanced NSCLC With the TORC1/2 Inhibitor TAK-228. J Thorac Oncol. 2023;18(4):516-526. PMID 36240971
  25. Bensen R, Brognard J. New Therapeutic Opportunities for the Treatment of Squamous Cell Carcinomas: A Focus on Novel Driver Kinases. Int J Mol Sci. 2021;22(6). PMID 33799513
  26. Planchard D, Cozic N, Mosele MF, Corcos N, Le Bescond L, Pradat Y, et al. Efficacy, safety, and biomarker analysis of datopotamab deruxtecan in advanced non-small cell lung cancer: ICARUS-LUNG01 phase 2 study. Cancer Cell. 2026;44(6):1147-1159.e6. PMID 41999747
  27. Okamoto I, Kuyama S, Girard N, Lu S, Franke F, Li Z, et al. TROPION-Lung07: Phase III study of Dato-DXd + pembrolizumab ± platinum-based chemotherapy as 1L therapy for advanced non-small-cell lung cancer. Future Oncol. 2024;20(37):2927-2936. PMID 39469838