Figure S2. TAK228 treatment induces accumulation of Lysotracker in LUSC cell lines.

PURPOSE This first-in-human, dose-escalation and dose-expansion study evaluated the safety, tolerability, and antitumor activity of datopotamab deruxtecan (Dato-DXd), a novel trophoblast cell-surface antigen 2 (TROP2)–directed antibody-drug conjugate in solid tumors, including advanced non–small-cell lung cancer (NSCLC). PATIENTS AND METHODS Adults with locally advanced/metastatic NSCLC received 0.27-10 mg/kg Dato-DXd once every 3 weeks during escalation or 4, 6, or 8 mg/kg Dato-DXd once every 3 weeks during expansion. Primary end points were safety and tolerability. Secondary end points included objective response rate (ORR), survival, and pharmacokinetics. RESULTS Two hundred ten patients received Dato-DXd, including 180 in the 4-8 mg/kg dose-expansion cohorts. This population had a median of three prior lines of therapy. The maximum tolerated dose was 8 mg/kg once every 3 weeks; the recommended dose for further development was 6 mg/kg once every 3 weeks. In patients receiving 6 mg/kg (n = 50), median duration on study, including follow-up, and median exposure were 13.3 and 3.5 months, respectively. The most frequent any-grade treatment-emergent adverse events (TEAEs) were nausea (64%), stomatitis (60%), and alopecia (42%). Grade ≥3 TEAEs and treatment-related AEs occurred in 54% and 26% of patients, respectively. Interstitial lung disease adjudicated as drug-related (two grade 2 and one grade 4) occurred in three of 50 patients (6%). The ORR was 26% (95% CI, 14.6 to 40.3), and median duration of response was 10.5 months; median progression-free survival and overall survival were 6.9 months (95% CI, 2.7 to 8.8 months) and 11.4 months (95% CI, 7.1 to 20.6 months), respectively. Responses occurred regardless of TROP2 expression. CONCLUSION Promising antitumor activity and a manageable safety profile were seen with Dato-DXd in heavily pretreated patients with advanced NSCLC. Further investigation as first-line combination therapy in advanced NSCLC and as monotherapy in the second-line setting and beyond is ongoing.

Abstract Background: Quantifying immune editing in human tumors is challenging because observed mutation patterns reflect both mutational processes and selection. Whether HLA class 1 genotypes uniformly influence which somatic mutations persist remains unclear. Methods: Whole-genome sequencing (WGS) from CPTAC NSCLC tumors with matched blood and adjacent normal lung underwent germline single nucleotide polymorphism (SNP), somatic single nucleotide variant (SNV), HLA typing, neoepitope prediction, and mutational signature decomposition using established pipelines. SNP/SNVs were annotated with codon position, resulting amino acids, and evolutionary similarity metrics (BLOSUM62, PAM250, Miyata, Atchley, Grantham). Four-digit HLA alleles were collapsed into recognized supertypes; neoepitopes with IC50 ≤150 nM were considered high-affinity. Models of expected nonsynonymous mutations were created using the GRCh38 reference, mutational signature, and tumor mutation burden and compared to observed SNP/SNVs using cosine similarity. Statistical significance was assessed using ANOVA and generalized linear models with FDR correction. Results: 219 cases (111 LUAD, 108 LUSC) were processed successfully. Median SNPs were 862.3, SNVs were 103.8 with an associated median of 62.5 predicted neoepitopes (4.9 high-affinity) per case. Amino acid substitutions arising from the third codon position (N=96) had higher similarity than those from position 1 and 2 (N=226), (p<0.001). SBS4 dominated 167 (76.3%) tumors; in blood, SBS5 and SBS58 characterized 152 (69.4%) and 55 (25.1%) cases, respectively. Codon-position biases differed across signatures, with SBS4 showing relative depletion of position-3 substitutions compared to SBS5 and SBS58 (p<0.001). Signature-based modeling demonstrated cosine similarity >0.90 with higher cosine similarity in SBS4 compared to SBS5 and SBS58. ANOVA comparisons of amino acid substitutions revealed multiple supertypes (A02, B27, B44) with relative depletion of anchor substitutions and high-affinity neoepitopes with these substitutions (FDR<0.05). Conclusion: Germline SNPs exhibit strong evolutionary constraint, whereas somatic SNVs show minimal intrinsic codon-position bias, supporting the hypothesis that cancer mutagenesis is a random process. Controlling for mutational signature uncovers HLA-specific immune editing, demonstrating that antigenic mutations are selectively depleted in specific HLA contexts. This signature-conditioned framework provides a clinically actionable platform for improving neoantigen prediction and immuno-oncology translation. Citation Format: Amy Lauren Cummings, Andy Han, Seung J. Park, Sai S. Kollapaneni, Daniel Li, Arjan Gower, Maria Antonia Velez Velez, Aaron Lisberg, Jonathan W. Goldman, Edward B. Garon. Mutational-signature adjusted models reveal HLA-mediated depletion of antigenic mutations in non-small cell lung cancer (NSCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4243.

Figure S1. Treatment with TAK228 does not increase CD31 density in H1703 and H2170 xenografts.

PURPOSE Datopotamab deruxtecan (Dato-DXd) is a trophoblast cell-surface antigen-2–directed antibody-drug conjugate with a highly potent topoisomerase I inhibitor payload. The TROPION-Lung05 phase II trial (ClinicalTrials.gov identifier: NCT04484142 ) evaluated the safety and clinical activity of Dato-DXd in patients with advanced/metastatic non–small cell lung cancer (NSCLC) with actionable genomic alterations progressing on or after targeted therapy and platinum-based chemotherapy. PATIENTS AND METHODS Patients received Dato-DXd 6 mg/kg once every 3 weeks. The primary end point was objective response rate (ORR) by blinded independent central review. Secondary end points included duration of response (DOR), safety, tolerability, and survival. RESULTS Among 137 patients who received at least 1 dose of Dato-DXd, 71.5% received at least three lines of prior therapies for advanced/metastatic disease. Overall, 56.9% had EGFR mutations and 24.8% had ALK rearrangements. Median treatment duration was 4.4 months (range, 0.7-20.6). The confirmed ORR was 35.8% (95% CI, 27.8 to 44.4) overall, and 43.6% (95% CI, 32.4 to 55.3) and 23.5% (95% CI, 10.7 to 41.2) in those with EGFR mutations and ALK rearrangements, respectively. The median DOR was 7.0 months (95% CI, 4.2 to 9.8), and the overall disease control rate was 78.8% (95% CI, 71.0 to 85.3). Grade ≥3 treatment-related adverse events (TRAEs) occurred in 28.5% of patients. The most common TRAE was stomatitis (preferred term; any grade: 56.2%; grade ≥3: 9.5%). Five (3.6%) patients experienced adjudicated treatment-related interstitial lung disease/pneumonitis, with 1 (0.7%) grade 5 event. CONCLUSION Encouraging and durable antitumor activity was observed with Dato-DXd in this heavily pretreated advanced/metastatic NSCLC population with actionable genomic alterations. The rate of treatment-related grade ≥3 toxicities was comparable with previous observations, and no new safety signals were observed.

8060 Background: NSCLC treatments and clinical trials include targeted agents and immunotherapy (IO) across stages, yet dAlts and how they relate to the tumor immune microenvironment (TIME) are incompletely characterized in early NSCLC (eNSCLC; stage I-III) and metastatic NSCLC (mNSCLC; stage IV). Here, we evaluated the NSCLC TIME by dAlt status to inform IO biomarker strategies. Methods: From the Tempus Database, we selected de-identified lung adenocarcinoma samples sequenced by xT DNA assay (eNSCLC n=5,535; mNSCLC n=10,299), a subset with whole transcriptome analysis. Targetable dAlts were defined as classic (c) (L858R and exon 19 del) or non-classic (nc) EGFR , KRAS G12C, other non-G12C variants, other guideline defined dAlts (ALK, ROS1, RET, NTRK1-3 fusions, ERBB2 alt, METex14), or no dAlt. Immune cell proportions were estimated by quanTIseq. Additional markers, PD-L1 TPS (IHC) and TMB (mt/mB; DNAseq) were analyzed. Significance (p<0.05) was assessed using χ 2 or Wilcoxon/Kruskal-Wallis rank sum tests. Results: The dAlt prevalence was similar (|Δ| < 2%) across early and late stage (Overall %: cEGFR=13, ncEGFR=2.9, KRAS G12C=15 and KRASother=22). The prevalence of other dAlt were less than 4% across stages. The CD8 proportion was higher in eNSCLC than mNSCLC (p<0.001). Across stages, CD4 Treg and CD8 proportions in the KRAS G12C cohort were nearly identical to the non-dAlt cohort, while c/nc EGFR tumors exhibited the lowest percentage of CD8 cells and higher Tregs cells compared to non-dAlt tumors (Table). PD-L1 and TMB were similar between KRAS G12C and non-dAlt tumors and lowest among c/nc EGFR (Table). Conclusions: This real-world analysis demonstrated similar dAlt prevalence across eNSCLC and mNSCLC, while the TIME was distinct across stage and dAlts. The TIME of KRAS G12C tumors was similar to non-dAlt tumors, and was least immunogenic in the c/nc EGFR cohort. These findings highlight immunological differences across stages and dAlts that should be considered when developing IO strategies. Group IO marker Overall No dAlt cEGFR ncEGFR KRAS G12C KRAS other Other dAlt eNSCLC % CD8 cells 1 * 1.3 (0.5, 2.4) 1.4 (0.6, 2.8) 0.9 (0.4, 1.7) 1.0 (0.5, 1.9) 1.4 (0.6, 2.6) 1.3 (0.5, 2.4) 1.1 (0.5, 2.1) % Tregs 1 * 6.9 (4.8, 9.2) 6.3 (4.2, 8.8) 7.4 (5.6, 9.4) 8.2 (5.8, 10.3) 7.0 (5.1, 9.3) 7.1 (5.2, 9.3) 6.7 (4.6, 8.6) % PDL1 2 * 22 20 9.3 7.9 31 25 24 TMB* 5.8 (3.2, 9.5) 7.9 (3.7, 13.2) 3.2 (2.1, 4.7 4.2 (2.3, 6.3) 6.8 (4.2, 10.0) 5.8 (3.7, 8.9) 3.2 (1.6, 5.3) mNSCLC % CD8 cells 1 * 0.6 (0.04, 1.6) 0.8 (0.1, 1.9) 0.4 (0.0, 1.3) 0.5 (0.0, 1.5) 0.7 (0.1, 1.7) 0.5 (0.01, 1.5) 0.5 (0.0, 1.4) %Tregs 1 * 4.0 (2.6, 5.9) 3.8 (2.4, 5.6) 4.2 (2.9, 6.1) 4.7 (3.0, 6.8) 4.2 (2.7, 6.0) 4.0 (2.6, 6.0) 3.9 (2.6, 5.6) % PDL1 2 * 28 24 15 17 37 34 34 TMB* 5.8 (3.2, 10.0) 7.9 (4.2, 13.1) 3.7 (2.1, 5.8) 4.2 (2.6, 6.8) 7.4 (5.2, 11.1) 6.3 (4.2, 10.0) 3.2 (1.6, 5.8) 1 Median (IQR); 2 PDL1 >50; *p<0.001, excluding “Overall.”

663 Background: Dato-DXd is a TROP2-directed antibody-drug conjugate under investigation in various solid tumor types. We report updated results in patients (pts) with locally advanced/metastatic urothelial cancer (la/m UC) from the ongoing phase 1 TROPION-PanTumor01 study (NCT03401385). Methods: Pts with unresectable la/m UC (stage III/IV) treated with ≥1 prior line of therapy, including an immune checkpoint inhibitor, received Dato-DXd 6 mg/kg Q3W. Primary study objectives were safety and tolerability. Secondary endpoints were objective response rate (ORR), duration of response (DOR), and progression-free survival (PFS) per RECIST 1.1 by blinded independent central review (BICR). Results: At data cutoff (April 22, 2024), 40 pts had received Dato-DXd; 8 (20%) were receiving ongoing treatment. Median follow-up was 10.0 (range 5–28) months (mo). Pts were heavily pretreated: 24 (60%) had received ≥3 prior regimens in the locally advanced/metastatic setting, 36 (90%) received prior platinum chemotherapy, and 33 (83%) received prior enfortumab vedotin. Confirmed ORR by investigator was 27.5% (95% CI, 14.6–43.9), including 11 partial responses (PR). Confirmed ORR by BICR was 25.0% (95% CI, 12.7–41.2), including 1 complete response (CR) and 9 PR (Table). Median DOR was not reached (95% CI, 2.6–not evaluable [NE]); 76.2% (95% CI, 33.2–93.5) of responders had ongoing responses at 6 mo. Median PFS by BICR was 6.9 mo (95% confidence interval [CI], 2.9–NE). Treatment-emergent adverse events (TEAEs) occurring in >20% of pts (any grade; grade ≥3) were stomatitis (53%; 5%), nausea (38%; 3%), decreased appetite (30%; 3%), and fatigue (28%; 0%). Grade ≥3 TEAEs occurred in 55% of pts. TEAEs associated with treatment discontinuation, dose reduction, and dose interruption occurred in 8%, 20%, and 35% of pts, respectively. No serious treatment-related AEs or treatment-related grade 4 or 5 AEs were reported. Two pts (5%) had adjudicated drug-related interstitial lung disease/pneumonitis (grade 2 and 3). Conclusions: In heavily pretreated pts with la/m UC, Dato-DXd demonstrated encouraging antitumor activity with a manageable safety profile. Dato-DXd is being evaluated in pts with urothelial cancer in the phase 1/2 TROPION-PanTumor02 (NCT05460273) and the phase 2 TROPION-PanTumor03 (NCT05489211) studies. Clinical trial information: NCT03401385 . Efficacy by BICR. Response All patients N=40 Confirmed ORR, n (%) [95% CI] 10 (25.0) [12.7–41.2] CR 1 (2.5) PR 9 (22.5) SD 20 (50.0) Non-CR/non-PD 1 (2.5) PD 5 (12.5) NE 4 (10.0) DOR at 6 mo, % [95% CI] 76.2 [33.2–93.5] Median PFS, mo [95% CI] 6.9 [2.9–NE] BICR, blinded independent central review; CI, confidence interval; CR, complete response; DOR, duration of response; mo, months; NE, not evaluable; ORR, objective response rate; PD, progressive disease; PFS, progression-free survival; PR, partial response; SD, stable disease.

PURPOSE:Immune checkpoint inhibitor (ICI)-induced thyroid dysfunction is a common endocrine immune-related adverse event. Although rates during therapy are well documented, data on post-treatment thyroid dysfunction are limited. We hypothesized that post-ICI thyroid dysfunction is under-recognized because of inadequate surveillance. METHODS:A retrospective analysis of 3,626 patients treated with ICIs for various malignancies within a single health system was conducted. Clinically acted-upon thyroid dysfunction (diagnosis or thyroid-directed medication) was evaluated before, during, and after ICI therapy, alongside rates of thyroid laboratory surveillance. RESULTS:Clinically acted-upon thyroid dysfunction occurred in 8.1% (294/3,626) during treatment and 4.4% (159/3,626) after treatment. Among the 1,170 patients with post-ICI thyroid laboratory results and no prior dysfunction, 11.6% (136/1,170) developed post-ICI thyroid dysfunction. However, 48.6%% (1,764/3,626) had no post-ICI thyroid laboratory results. Thirty percent of patients with abnormal thyroid stimulating hormone (TSH) values and no clinically acted-upon thyroid dysfunction before therapy discontinuation subsequently developed clinically acted-upon thyroid dysfunction, and the rate of post-ICI clinically acted-upon thyroid dysfunction was higher in patients who received 9 or less months of ICI therapy. CONCLUSION:Post-ICI thyroid dysfunction is frequent, with 11.6% of monitored patients being affected, and patients with abnormal TSH before ICI discontinuation and those who received treatment for 9 months or less may benefit from more stringent post-ICI surveillance.

12116 Background: Immune checkpoint inhibitor (ICI)-induced thyroid dysfunction is the most common endocrine immune-related adverse event. While ICI-induced thyroid dysfunction rates during therapy are well documented, data on post-treatment dysfunction is limited. Describing these rates is important as ICIs are increasingly used in the curative treatment setting. This study aimed to evaluate the rates of post-ICI thyroid dysfunction, evaluate for predictors of post-ICI thyroid dysfunction and assess adequacy of post-ICI thyroid function surveillance. Methods: A retrospective analysis of 3626 patients treated with ICIs for various malignancies and cancer stages within a single health system from March 2013 to December 2022 was conducted. Rates of clinically acted upon thyroid dysfunction (diagnosis or thyroid-directed medication) were evaluated before, during, and after ICI therapy, alongside rates of thyroid laboratory surveillance in the post treatment setting. A multivariate analysis evaluated the odds of developing clinically acted upon post-ICI thyroid dysfunction based on patient/treatment characteristics. Rates of clinically acted upon thyroid dysfunction were evaluated based on therapy duration. Statistical analyses were carried out using R V4.1.0. Results: Clinically acted upon thyroid dysfunction occurred in 8.1% of patients (294/3626) during treatment and 4.4% (159/3626) after treatment. However, in patients alive two months after ICI cessation, 53.9% (989/1834) had no post-ICI thyroid function tests performed. Among the 1170 patients with post-ICI thyroid labs and no prior dysfunction, 11.6% (136/1170) developed post-ICI thyroid dysfunction. Thirty percent of patients with abnormal TSH values and no clinically acted upon thyroid dysfunction prior to therapy discontinuation subsequently developed clinically acted upon thyroid dysfunction. The rate of post treatment thyroid dysfunction in patients who underwent thyroid test surveillance and received <9 months of therapy was 13.3% compared to 6.5% in those who received therapy >9 months. The odds ratio for developing post-ICI dysfunction were 1.76 (95% CI, 1.03 to 2.95) for patients with urologic malignancies compared to patients with respiratory malignancies as the reference group. Conclusions: Post-ICI thyroid dysfunction is frequent, with 11.6% of patients who undergo thyroid function surveillance being affected. Patients with abnormal TSH before ICI discontinuation and those who received treatment for < 9 months as well as those with urologic malignancies may benefit from more stringent post-ICI surveillance.

TPS2700 Background: Delta-like ligand 3 (DLL3) is an inhibitory Notch ligand that is aberrantly expressed on the surface of tumor cells, in particular on those with neuroendocrine differentiation. Tarlatamab is a bispecific T cell engager that binds to DLL3 and CD3 to promote T cell killing of DLL3-expressing cells. Prior studies of tarlatamab have demonstrated encouraging antitumor activity and manageable toxicity in patients with small cell lung cancer (SCLC; DeLLphi-301) and neuroendocrine prostate cancer (NEPC; DeLLpro-300). Meanwhile, DLL3 has been reported to be highly expressed in multiple tumor types, including in many neuroendocrine neoplasms (NENs) other than SCLC and NEPC. The role of anti-DLL3 therapies in these cancers has not been established. Methods: This is a phase 2, multicenter, open-label, basket study designed to evaluate the efficacy of tarlatamab in patients with DLL3-expressing cancers. Key inclusion criteria include presence of advanced stage disease with progression following ≥1 prior line of therapy and positive tumor DLL3 expression by immunohistochemistry (Ventana SP347 assay). Patients with de novo SCLC or NEPC are excluded, but all other tumor types and NENs are eligible, including large cell neuroendocrine carcinoma and SCLC transformed from previously treated NSCLC. Tarlatamab will be administered at an initial step-up dose (1 mg on D1 and 10 mg on D8 and D15 of cycle 1) followed by 10 mg every 2 weeks. Treatment will continue until unacceptable toxicity, progressive disease, or withdrawal of consent. The study will follow a Simon's two-stage design: in Stage 1, 10 patients with tumor DLL3 expression ≥25% will be enrolled, and the study will be stopped if ≤1 patient achieves an objective response; otherwise, an additional 19 patients with tumor DLL3 expression ≥1% will be enrolled for Stage 2. The primary endpoint is the objective response rate. Secondary endpoints include safety, progression free survival, duration of response, and overall survival. Exploratory studies will evaluate correlation of antitumor activity with tissue and blood-based biomarkers, such as DLL3 expression on tumor and liquid biopsies. This study is currently enrolling patients through the University of California Lung Cancer Consortium (UCLCC). Clinical trial information: NCT06788938 .

Supplementary Figure 1: Graphical Abstract Supplementary Figure 2: Overall Survival in NSCLC patients treated with pembrolizumab in KEYNOTE-001 according to EGFR mutation Supplementary Figure 3: Recursive partitioning identified classical monocytes and NK as the main immune populations whose frequency splits patients into distinct patterns of overall survival Supplementary Figure 4: Classical monocytes, NK cells and ICOS+ CD4+ T cells are associated with improved pembrolizumab efficacy in advanced NSCLC patients. Supplementary Figure 5: Gating strategy of the 3 main immune populations associated with pembrolizumab efficacy. Supplementary Figure 6: Comparison side by side of Maxstat cut-offs with cut-off from Younden index for overall survival in the 3 populations of interest. Supplementary Figure 7: Survival outcomes in NSCLC patients treated with pembrolizumab in KEYNOTE-001 according to blood baseline frequency of classical monocytes. Supplementary Figure 8: Baseline immune peripheral score is also associated with PFS in melanoma patients treated with PD-1 inhibitors.

3000 Background: DB-1310 is a novel ADC comprised of a humanized anti-HER3 IgG1 monoclonal antibody, cleavable peptide linker, and DNA topoisomerase I inhibitor. Here, we report the preliminary results of the FIH trial. Methods: This global, multi-center, open-label Ph 1/2a trial includes dose escalation and expansion. Pts with advanced solid tumors who had failed standard therapy were enrolled. In Ph1, DB-1310 was planned to be administered at doses from 1.5 mg/kg to 6.5 mg/kg, Q3W, iv, using a 3+3 design, with additional pts enrolled to determine the RP2D. Ph 2a will include approximately 30-40 pts per cohort to optimize the RP2D and assess efficacy. Results: As of Jan 17, 2025, 123 pts were enrolled and treated with DB-1310 monotherapy in Ph1 (ECOG PS 1, 80.5%; White, 39.0%, Asian, 52.8%; NSCLC, 65.0%, EGFRm NSCLC, 37.4%; brain metastasis, 17.1%), median prior lines of systemic therapy was 3 (range, 1-11). Of the 42 efficacy-evaluable pts with EGFRm NSCLC, 92.9% had previously received 3 rd generation EGFR TKI, 92.9% had received platinum-based chemotherapy. The unconfirmed ORR was 25.5% (95% CI, 17.63, 34.65) across all tumor types and 35.7% (95% CI, 21.55, 51.97) in EGFRm NSCLC. Median PFS was 5.4 months overall and 7.0 months for EGFRm NSCLC. 38 (30.9%) pts experienced ≥ G3 TRAEs, while 7 (5.7%) had drug-related SAEs. TRAEs led to dose reduction in 14 (11.4%) pts and discontinuation in 5 (4.1%) pts. No TRAE leading to death was reported. Most common TRAE ( > 20%, any grade/≥G3) were nausea (36.6%/0.8%), anemia (35.8%/4.1%), neutrophil count decreased (34.1%/17.9%), platelet count decreased (31.7%/9.8%), white blood cell count decreased (29.3%/8.9%), decreased appetite (23.6%/0.8%), and vomiting (21.1%/0%). Interstitial lung disease occurred in 7 pts (5.7%, 6 G1 and 1 G2). PK exposure was increased through dose escalation, with low systemic payload exposure and no accumulation of DB-1310 upon repeated administration. Conclusions: DB-1310 showed a manageable safety profile and encouraging antitumor activity in pts with heavily pretreated advanced solid tumors, particularly EGFRm NSCLC. Clinical trial information: NCT05785741 . Tumor response by dose (efficacy-evaluable). Dose (mg/kg) 1.5 3 4.5 5.0 5.5 6 Total All tumors, n 3 10 25 53 16 3 110 uORR, n (%) (95% CI) 0 (0)(0.00, 70.76) 1(10.0)(0.25, 44.50) 8 (32.0)(14.95, 53.50) 13 (24.5) (13.76, 38.28) 6 (37.5) (15.20, 64.57) 0 (0)(0.00, 70.76) 28 (25.5) (17.63, 34.65) DCR, n (%) (95% CI) 3 (100.0)(29.24, 100.00) 8 (80.0)(44.39, 97.48) 23 (92.0)(73.97, 99.02) 42 (79.2) (65.89, 89.16) 11 (68.8) (41.34, 88.98) 2 (66.7) (9.43, 99.16) 89 (80.9) (72.31, 87.78%) EGFRm NSCLC, n 0 7 9 16 8 2 42 uORR, n (%) (95% CI) - 1 (14.3)(0.36, 57.87) 4 (44.4) (13.70, 78.80) 5 (31.3) (11.02, 58.66) 5 (62.5) (24.49, 91.48) 0 (0)(0.00, 84.19) 15 (35.7) (21.55, 51.97) DCR, n (%) (95% CI) - 6 (85.7) (42.13, 99.64) 9 (100.0) (66.37, 100.00) 14 (87.5) (61.65, 98.45) 7 (87.5) (47.35, 99.68) 2 (100.0)(15.81, 100.00) 38 (90.5) (77.38, 97.34)

BACKGROUND:This exploratory analysis assessed datopotamab deruxtecan (Dato-DXd) in pretreated patients with advanced or metastatic NSCLC and EGFR mutations. METHODS:Data were pooled from the phase II TROPION-Lung05 (NCT04484142) and phase III TROPION-Lung01 (NCT04656652) trials. Patients with EGFR-mutated advanced or metastatic NSCLC, who had received previous targeted therapies and platinum-based chemotherapy, received Dato-DXd 6 mg/kg (TROPION-Lung05) or were randomized to Dato-DXd 6 mg/kg or docetaxel 75 mg/m2 (TROPION-Lung01) once every 3 weeks. End points included objective response rate, duration of response, and progression-free survival, all per blinded independent central review, overall survival, and safety. RESULTS:In total, 117 patients with EGFR mutations who received Dato-DXd were included in the pool. The population was heavily pretreated (median three lines of previous therapies; range, 1-5) and predominantly Asian (69%). The most common mutations were exon 19 deletion (51%), L858R (32%), and T790M (27%); more than one EGFR mutation could be present. The confirmed objective response rate was 43% (95% confidence interval [CI]: 34-52), including five complete responses (4%). Median duration of response was 7.0 months (95% CI: 4.2-9.8). Median progression-free survival and overall survival were 5.8 (95% CI: 5.4-8.2) and 15.6 months (95% CI: 13.1-19.0), respectively. The safety profile of Dato-DXd was consistent with the individual studies. No new safety signals were observed. Rates of grade greater than or equal to 3 treatment-related adverse events and serious adverse events were 23% and 6%, respectively. CONCLUSION:Dato-DXd demonstrated meaningful clinical activity in patients with EGFR-mutated advanced or metastatic NSCLC who had progressed on EGFR-directed therapies and chemotherapy, with an acceptable safety profile.

Supplementary Table 1: List of antibodies used in the mass cytometry panel Supplementary Table 2: Number of immune cells analysed by CYTOF in our cohort Supplementary Table 3: Demographic and clinical characteristics of the CyTOF UCLA cohort (n=27), the Total UCLA cohort (n=97) and the Total KEYNOTE-001 (n=495) of non-small cell lung cancer patients Supplementary Table 4: Cox's proportional hazards model for overall survival in NSCLC patients treated with pembrolizumab in KEYNOTE-001