Supplementary Table S2. Representativeness of Study Participants.

Treemap of most common drug mechanisms for patients treated in biomarker-matched studies.

Metaplastic breast cancer (MPBC) is a rare, heterogeneous invasive subtype of breast cancer with predominant triple negative receptor status. MPBC typically has limited treatment options and portends to a poorer prognosis. Here, we focus on understanding the genomic landscape that may aid in discovering effective targeted therapies for MPBC. Using the data from the American Association for Cancer Research Project Genomics Evidence Neoplasia Information Exchange (AACR GENIE) Cohort v16.1 database, we cataloged genetic alterations in metaplastic breast cancer. ​​ We identified 258 cases from 267 samples of metaplastic breast cancer. The median age known at the time of sequencing was 59 (range 25-87). Gender analysis suggested a predominantly female distribution (96.5%, n=249) compared to male (1.6%, n=4). Race/ethnicity included 62% (n=160) White, Others 25% (n=64), 6.6% (n=17) Asian, 5.8% (n=15) African American, 5.0% (n=13) Hispanic, and Native American 0.8% (n=2). Most frequent genomic alterations in the overall cohort were found in TP53 (57%, 152) with 30.3%; 81 missense and 11.2%; 30 nonsense mutations and 9%; 24 frameshift deletions; PIK3CA (34%, 90) with 30%; 80 missense mutations and 3%; 8 in frame deletions; KMT2C (17.1%, 37); KMT2D (10.8%, 31); FAT1 (13%; 29); TERT (12%; 28); KMT2D (10.5%; 25); PTEN (9.4%; 25); SPEN (10%; 22); PIK3R1 (8.3%; 20); RB1 (6.7%; 18); NF1 (7.1%; 17) genes. The most TP53 frequent mutations were seen in R248Q (14.5%; 9), R175H (8.1%; 5), Q192* (4.8%; 3), H193R (4.8%; 3), and R273C (4.8%; 3). The most frequent PIK3CA mutations were seen in H1047R (55.1%; 49), followed by E545K (14.6%; 13), Q546R (4.5%; 4), E542K (2.2%; 2), and C420R (1.1%; 1). The average mutation count was 5.0. TP53 and PIK3CA aberrations were mutually exclusive (p=0.043). Most frequent copy number alterations (deletions 66%; 181 and amplifications 34%; 93) were seen in AGO2 (22.9%, 16) RAD21 (20.4%, 19); RECQL4 (19.8%, 20); MYC (19.4%, 21); EXT1 (19%; 3); CDKN2A (14%; 15) and CDKN2B (13%; 14) genes. Our analysis provides valuable insights into the genetic landscape of this aggressive and rare cancer subtype. Overall, these findings support further investigation into targeted therapies for MPBC that address its unique genetic landscape, potentially improving outcomes for patients facing limited treatment options and poor prognosis. Blessie Elizabeth Nelson, Jacqueline Nkrumah, Mohammed Gouda, Jordi Rodon, Sarina Piha-Paul, Bora Lim. Genomic landscape in metaplastic breast cancer: Results from the AACR GENIE Database [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3746.

Cholangiocarcinoma (CCA) patients (pts) harboring FGFR fusions treated with FGFR inhibitors (FGFRi) have shown clinical benefit, but often have progression in 5- 7 months. Secondary polyclonal mutations in the FGFR2 kinase domain (KD) may be a mechanism for acquired resistance to FGFRi. Tinengotinib, a spectrum-selective multi-kinase inhibitor with unique binding properties to FGFR2, potently inhibited FGFR2 KD mutations in pre-clinical models and exhibited antitumor activity in CCA pts. Here we present two cases of CCA pts treated with tinengotinib in a phase Ib/II trial. Biomarker samples were collected at baseline and C3D1 timepoints as an exploratory endpoint for pts receiving tinengotinib therapy. Efficacy was evaluated by RECIST v1.1. Tissue and blood samples were analyzed by FoundationOne® CDx and FoundationOne® Liquid CDx test, respectively, for FGFR alterations. Two CCA pts received oral tinengotinib 12mg once daily continuously for 28-day cycles and completed biomarker sampling at baseline and C3D1. Both pts were previously treated with chemotherapy and pemigatinib. Pt A is a 28yr female with CCA metastatic to liver and bone. Tinengotinib was initiated in August 2021. The pt achieved a partial response (PR) with tumor reductions of 29.2%, 36% and 41.6% at C3D1, C4D1 and C5D1, respectively. Pt A was treated for 210 days, discontinuing due to radiographic disease progression following a 17-day dose interruption for suspected grade 3 TSH increase. The progression-free survival (PFS) was 6.9 months. Baseline biomarker data showed presence of KD mutations including: E565G/A, N549H, K659M and I548T. At C3D1, E565G had reduced from 1.36% to 0.09%, and there was disappearance of all other FGFR2 KD mutations. Pt B is a 40yr female with CCA metastatic to liver, lung and lymph nodes. Tinengotinib was initiated in August 2021. The pt achieved a PR with tumor reductions of 23.4% and 48.6% at C3D1 and C5D1, respectively. Pt B was on treatment for 181 days and discontinued due to progression in non-target lesions at C6D1, following a 23-day dose interruption for unrelated grade 3 ALT increase. PFS was 5.9 months. Baseline biomarker data showed presence of KD mutations including: E565A/G/K, V564L/I/F, N549K/T An inverse correlation between the presence of KD mutations and the response to tinengotinib treatment was observed in these two FGFRi relapsed/refractory metastatic CCA pts. This suggests tinengotinib may overcome acquired resistance to FGFRi in CCA pts. An ongoing randomized phase III study is investigating tinengotinib vs. chemotherapy in FGFRi refractory/relapsed CCA (NCT05948475). Clinical trial information: NCT04742959 Sarina A. Piha-Paul, Sanjay Goel, Chih-Yi Liao, Nashat Gabrail, Farshid Dayyani, Syed Kazmi, Sayeh Lavasani, Katie Hennessy, Hui Wang, Caixia Sun, Peng Peng, Jean Fan, Milind Javle. Correlation of clinical and biomarker data in FGFR inhibitor failed metastatic cholangiocarcinoma patients with tumor response to tinengotinib [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 825.

PURPOSE:Intratumoral injection of Clostridium novyi-nontoxic (C. novyi-NT), lacking alpha-toxin, germinates and subsequently replicates in the tumor hypoxic regions, causing cell lysis and inflammation. This phase Ib study investigated the safety and synergistic effects of pembrolizumab and C. novyi-NT in advanced solid tumors. PATIENTS AND METHODS:Patients with injectable solid tumors received one intratumoral C. novyi-NT injection (day 8) across four dose cohorts (3 × 104 to 100 × 104 spores) with pembrolizumab (200 mg intravenously every 3 weeks, day 0-24 months). Sixteen patients were enrolled with a median age of 62.5 years, and the most common tumor types were breast cancer and melanoma. RESULTS:The combination therapy demonstrated a manageable safety profile, with one grade 3 dose-limiting toxicity of abscess formation and the remainder of treatment-related adverse events in ≥10% of patients being grade 1 or 2 injection site reaction (25%), pyrexia (19%), pruritus (13%), leukopenia (13%), and anemia (13%). The maximum tolerated dose was determined to be pembrolizumab intravenously at 200 mg every 3 weeks and a single dose of 100 × 104C.novyi-NT spores intratumorally on day 8. The confirmed overall objective response rate of 25% was observed among four patients with nonkeratinizing undifferentiated nasopharyngeal squamous carcinoma, human papilloma virus-positive squamous cell carcinoma of the base of the tongue, vulvar melanoma, and chordoma. Among these patients, we obtained three partial responses and one complete response. The median duration of the response was 10.93 months, and stable disease was observed in 69% of patients. CONCLUSIONS:This study highlights the potential of the C. novyi-NT and pembrolizumab combination as a promising therapeutic strategy for treatment-refractory solid tumors.

3100 Background: SPYK04 is a novel MEK1/2 inhibitor designed to enhance RAF-MEK binding and potentially inhibit feedback activation of MEK1/2. This approach was developed to address the limited efficacy of conventional MEK inhibitors in RAS-mutated cancers, which is thought to be due to feedback activation of the MAPK pathway. Methods: This first-in human study of SPYK04 is conducted in the US and Japan (NCT04511845). Patients with locally advanced or metastatic solid tumors harboring MAPK pathway alterations were eligible in its DE part. The primary endpoints included assessment of pharmacokinetics, adverse events (AEs) and dose-limiting toxicities (DLTs); the secondary endpoint was objective response rate (ORR). SPYK04 was administered orally once daily in continuous 28-day cycles. Results: A total of 23 patients were enrolled in DE. SPYK04 was evaluated at doses ranging from 0.1 to 1.3 mg/day. An accelerated titration design was employed for doses of 0.1, 0.2, and 0.4 mg/day, followed by a transition to a 3+3 design starting from 0.8 mg/day. Over the dose range of 0.1 to 1.3 mg/day, systemic exposure demonstrated a dose-dependent increase. Grade 3 or higher treatment-related adverse events (TRAEs) were observed in 30.4% of patients. TRAEs observed in >= 20% of patients were: dermatitis acneiform and blood creatinine phosphokinase increased (60.9% each); AST increased (30.4%); nausea and stomatitis (26.1%, each). DLTs were observed in one patient each at dose levels of 0.8, 1.0 and 1.3 mg/day. All DLTs were due to receiving <75% of the planned dose. A decision was made to not escalate beyond 1.3 mg/day. The ORR was 8.7% (n = 2 of 23), with both partial responses (PRs) occurring amongst the 5 enrolled ovarian cancer patients. The disease control rate (DCR; PR+SD) was 52.2% (n = 12 of 23). Conclusions: SPYK04 was tolerated at doses up to 1.3 mg/day in patients with advanced solid tumors. PRs were observed in two patients with ovarian cancer. Further evaluation of safety and efficacy will occur in the expansion part of this study. Clinical trial information: NCT04511845 . SPYK04 dose, mg/day 0.1 0.2 0.4 0.8 1 1.3 Patients, n 1 1 1 8 6 6 Safety, n (%) TRAE 0 1 (100.0) 1 (100.0) 8 (100.0) 6 (100.0) 6 (100.0) Grade 3 or higher TRAE 0 0 0 1 (12.5) 4 (66.7) 2 (33.3) TRAE leading to discontinuation of study treatment 0 0 0 0 1 (16.7) 0 Confirmed best response, n (%) Responders 0 0 0 1 (12.5) 1 (16.7) 0 Partial Response 0 0 0 1 (12.5) 1 (16.7) 0 Stable Disease 0 0 1 (100.0) 2 (25.0) 3 (50.0) 4 (66.7)

BACKGROUND:Intravenous (IV) nemvaleukin alfa (nemvaleukin, ALKS 4230) administered daily on days 1-5 in 21‑day cycles demonstrated antitumor activity and manageable safety in heavily pretreated advanced solid tumors. We present results from cohort 2 of the open-label phase I/II ARTISTRY-3 (NCT04592653) study, which evaluated less frequent IV dosing of nemvaleukin in advanced solid tumors. METHODS:Eligible patients received escalating IV nemvaleukin doses in 21-day cycles on 3 schedules: day 1, days 1 and 8, and days 1 and 4. The primary endpoint was the incidence of dose-limiting toxicities (DLTs). RESULTS:From April 2022 to June 2024, 52 patients received nemvaleukin. No DLTs were reported. Most treatment-related adverse events (TRAEs) were grade 1-2. Six patients (12%) experienced grade 3 TRAEs, the most common being neutropenia. Nemvaleukin exposure increased with escalating doses. Natural killer (NK) cell and CD8+ T-cell expansion in whole blood was observed, with minimal regulatory T-cell expansion. Nemvaleukin at 30 μg/kg on days 1 and 8 was the recommended phase II dose. No objective responses were observed; 16 (31%) patients had stable disease (6 [12%] for ≥3 months). Increased tumor microenvironment infiltration of NK cells and CD8+ T-cells was observed in on-treatment biopsies. CONCLUSIONS:Less frequent IV doses of nemvaleukin demonstrated pharmacodynamic proof of mechanism and were tolerable with some disease stabilization.

PURPOSE:Concurrent use of medications can modulate the effectiveness of immunotherapy. Although this interaction is well documented for immune checkpoint inhibitors, whether this occurs with new experimental compounds has not been evaluated. PATIENTS AND METHODS:A computerized data extraction tool was used to collect clinical data and identify the prescription of a predefined set of medications within 30 days of immunotherapy infusion in the Department of Investigational Cancer Therapeutics at the University of Texas MD Anderson Cancer Center. The primary endpoints were median overall survival (OS) and progression-free survival. Tumor responses were assessed using RECIST. RESULTS:We identified 897 patients. The most prevalent tumor types were colorectal (24.5%), head and neck (10.5%), and pancreatic (9.4%). The immunotherapy administered consisted of monoclonal antibodies and fusion proteins (64.7%), immune modulators (IM; 20.8%), combinations of IMs and antibodies (9.2%), and oncolytic viruses and cancer vaccines (5.3%). The most frequently prescribed drugs were narcotics (70.5%), antiemetics (49.1%), antihistamines (34.6%), antibiotics (31.2%), and proton pump inhibitors (PPI; 28.7%). Patients receiving antihistamines exhibited increased rates of stable disease and partial response (χ2 8.48; P = 0.014) on the IMs and antibodies combination. The benefit of antihistamines was confirmed in a multivariate analysis of OS [HR, 0.752 (95% CI, 0.603-0.938); P = 0.012]. For patients with colorectal cancer, PPI use was associated with shortened survival, with a median OS of 5.2 months with PPI use and 8.6 months without it (P < 0.001). CONCLUSIONS:Our findings highlight the need for strategies to guide concurrent medication choices for patients receiving immunotherapy in early-phase trials. SIGNIFICANCE:Concurrent administration of antihistamines correlates with enhanced survival in patients receiving experimental immunotherapy for cancer. Conversely, PPI use diminishes survival in patients with colorectal cancer. These findings highlight how tumor immunogenicity and drug interactions can modulate response and survival outcomes, offering new insights to optimize investigational immunotherapy.

Transforming growth factor-beta 1 (TGFβ1) plays a crucial role in tumor progression and immunosuppression in cancer. SAR439459 is a second-generation monoclonal antibody that targets all TGFβ isoforms. A first-in-human study (NCT03192345) established the SAR439459 dose at 15 mg/kg every three weeks as the initial recommended phase 2 dose. TACTIC (NCT04729725) is a phase Ib dose-expansion trial to evaluate the combination of SAR439459 (anti-TGF-β) and cemiplimab (anti-PD1) in pts with advanced solid tumors. Pts were mandated to have been treated with PD-1 or PD-L1 inhibitors (PD-1/L1i) as their most recent therapy. The primary objective of the trial was to assess efficacy; secondary objectives included toxicity; and translational studies included multiplex immunofluorescence (mIF) (n=2) and single-cell RNA-sequencing (5’GEX) of tumor (n=1) and blood (n=3). Three pts were enrolled: advanced clear cell renal cell carcinoma (ccRCC), high-grade pancreatic neuroendocrine carcinoma (NEC), and castration-resistant prostate cancer (CRPC) (n=1 each). The ccRCC pt achieved a confirmed RECISTv1.1 partial response (PR), with -55% reduction in target lesions, progression-free survival (PFS) of 12 mths, and overall survival (OS) of 26 mths. The other two pts had RECISTv1.1 progressive disease (PD) Hemorrhagic adverse events occurred in all pts. Two grade 3 treatment-related adverse events were reported: rash and retroperitoneal hemorrhage. Two pts (ccRCC and CRPC pts) developed skin squamous cell carcinomas, attributed to SAR439459 and were treated with resection. Single-cell sequencing revealed higher Treg activation as well as CD8+ T cell (LAG3, TIGIT) and CD4+ T cell (TIGIT, PDCD1, CTLA4) exhaustion profiles in the blood samples collected at C2D1 (post SAR439459, pre-cemiplimab) of both pts who had PD versus the pt with the PR. LAG3 and TIGIT were also present in CD8+ T cells from on-treatment biopsies collected from one of the pts with PD. A unique on-treatment CD4+ T cell signature associated with NK-kB activation in response to TNF was found in the circulation of the pt with PR. Within the tumor microenvironment, longitudinal mIF analysis of one of the pts with PD demonstrated reductions in TGF-β, pSMAD3 and immune infiltrate by C2D1. In this phase 1b study, SAR439459 and cemiplimab demonstrated a deep and durable RECISTv1.1 PR in a post-PD-1/L1i treated pt with advanced ccRCC. While translational data are limited, these findings support the association of systemic immune activation with response to therapy. Despite this, based on data from another study (NCT03192345), all investigations of SAR439459 were discontinued due to lack of sufficient antitumor activity and the associated bleeding risk. Carlos Torrado, Katarzyna Tomczak, Ecaterina E. Dumbrava, Baohua Sun, Natalie Ngoi, Claudio Arrechedera, Edwin R. Parra, Jordi Rodon Anhert, Sarina A. Piha-Paul, Apostolia M. Tsimberidou, Siqing Fu, Aung Naing, Paula R. Pohlmann, David S. Hong, Funda Meric-Bernstam, Pavlos Msaouel, Cara L. Haymaker, Timothy A. Yap. Clinical and translational study of the TGF-β monoclonal antibody SAR439459 in combination with the PD-1 inhibitor cemiplimab for patients (pts) with advanced solid tumors: Results from the phase 1b TACTIC trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr CT046.

We examined baseline diet and its association with outcomes and social determinants of health (SDoH) factors among patients (pts) enrolled in early-phase immune checkpoint blockade-included combination trials in a comprehensive cancer center. Pts (ECOG 0-1) scheduled to begin an early-phase combination trial with immune checkpoint blockade were recruited into a prospective, longitudinal study design, and completed the validated Multifactor Dietary Screener (0-8, higher score = more frequent use of food during the past 30 days) prior to trial start. Clinical data were extracted from the EMR. Statistics included t-tests and Spearman correlations. Among 259 pts (52% male, 85% White, age=25-83y, BMI=14-46, median prior lines of treatment=3, 41% had ≥2 metastatic organ involvement), increased intakes of vegetables and whole grain breads were associated with longer trial duration (P<0.029 for each). Frequent use of nuts/ seeds was linked to prolonged overall survival (P=0.014), and beans to extended progression-free survival (P=0.039). Lower regular sugar desserts and white potatoes were associated with superior treatment response (CR / PR) (P<0.005 each). Males had bacon/ sausage, hot dogs, red meat, milk, and ice-cream more often than females, and fruits, green leafy vegetables, and other vegetables less often (P<0.008 each). White pts had milk and ice-cream more often, and lettuce and leafy vegetables less often than minority pts (P<0.002 each). Hispanic pts had beans (P=0.024) more often, and cheese, and other vegetables less often than non-Hispanic pts (P<0.006 each). Young adults (25y-39y) had pasta more often than older patients (P=0.006); pts >70y had shellfish and milk more often than younger pts (P<0.028 each). Higher BMI was linked to more frequent intakes of regular fat sausage or bacon, hot dogs, fried potatoes (P<0.003 each), and regular fat salad dressing or mayonnaise (P=0.013), less frequent intakes of fruit (P<0.001), and fewer years of education (P=0.037). Less education was associated with more frequent use of bacon/ sausage, doughnuts/ other baked desserts (P<0.002 each), regular fat salad dressing, and hot dogs (P<0.016 each), and less frequent intakes of whole grain breads, fruits, greens, other vegetables, nuts, yogurt (P<0.004 each), and fish (P=0.013). Dietary food choices, increasingly recognized as impacting the hallmarks of cancer, may offer tangible interception of disease outcomes even during treatment in early-phase combination trials incorporating immune checkpoint blockade. Food preferences for pts in trials may be bidirectionally influenced by SDoH factors including educational attainment, gender, race/ ethnicity, and age, and as such, these factors need to be considered when developing wholesome individualized dietary prescriptions and dietary interventions for trial participants. Goldy C. George, Hangil Lee, Sarina A. Piha-Paul, Ecaterina Dumbrava, Apostolia Tsimberidou, Siqing Fu, Grace Appleton, Laila Noor, Jordi Rodon, Timothy A. Yap, Funda Meric-Bernstam, David S. Hong. Food choices of patients with advanced cancer at start of early-phase combination trials with immune checkpoint blockade [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 809.

Supplemental Figure 5. Tumor biopsy gene expression changes in chemokine, cytokine and TLR related signaling pathways after SYNB1891 treatment.

Abstract Purpose: 4-1BB (CD137) is a costimulatory immune receptor expressed on activated T cells, activated B cells, NK cells, and tumor-infiltrating lymphocytes, making it a promising target for cancer immunotherapy. Cinrebafusp alfa, a monoclonal antibody-like bispecific protein targeting HER2 and 4-1BB, aims to localize 4-1BB activation to HER2-positive tumors. This study evaluated the safety, tolerability, and preliminary efficacy of cinrebafusp alfa in patients with previously treated HER2-positive malignancies. Patients and Methods: This was a multicenter dose-escalation study involving patients with HER2-positive malignancies who received prior treatment. The study assessed the safety and efficacy of cinrebafusp alfa across various dose levels. Patients were assigned to different cohorts, and antitumor responses were evaluated. The study aimed to determine the MTD and to observe any clinical activity at different dose levels. Results: Of 40 evaluable patients in the “active dose” efficacy cohorts, five showed an antitumor response, resulting in an overall response rate of 12.5% and a disease-control rate of 52.5%. Clinical activity was observed at the 8 and 18 mg/kg dose levels, with confirmed objective response rates of 28.6% and 25.0%, respectively. Cinrebafusp alfa was safe and tolerable, with grade ≤2 infusion-related reactions being the most frequent treatment-related adverse event. MTD was not reached during the study. Conclusions: Cinrebafusp alfa demonstrates promising activity in patients with HER2-positive malignancies who have progressed on prior HER2-targeting regimens. Its acceptable safety profile suggests it could be a treatment option for patients not responding to existing HER2-directed therapies. See related commentary by Eguren-Santamaría et al., p. 231

PURPOSE:Trastuzumab deruxtecan (T-DXd) is currently approved for treating metastatic breast cancer (MBC) that is HER2 positive [immunohistochemistry (IHC) score of 3+ or in situ hybridization (ISH) positivity] or HER2-low (IHC score of 1+ or IHC 2+/ISH negative), as well as for HER2-positive gastric cancer, HER2-mutant lung cancer, and HER2-overexpressing solid tumors. Given the increasing utilization of T-DXd, we sought to determine how HER2 status might change following T-DXd therapy. EXPERIMENTAL DESIGN:We retrospectively reviewed patients with MBC who received T-DXd at the University of Texas MD Anderson Cancer Center. We included patients with paired pre- and post-treatment biopsies assessed for HER2 status using IHC. RESULTS:We included 41 patients with MBC who received treatment with T-DXd and had paired pre- and post-treatment biopsies assessed for HER2 status using IHC. HER2 loss was observed in 11 patients [32.4% of 34 patients with pre-treatment HER2 expression (1+, 2+, or 3+)] following treatment with T-DXd. In addition to the 11 patients with HER2 loss, another 10 patients (29.4%) had a decrease in HER2 score after treatment with T-DXd. CONCLUSIONS:HER2 loss and decrease in HER2 expression are common in patients with MBC receiving treatment with T-DXd. Reevaluation of HER2 status following T-DXd treatment should be considered prior to alternate HER2-targeted therapies that require HER2 overexpression for efficacy.

PURPOSE:NUT carcinoma (NC) is an underdiagnosed, poorly differentiated squamous cell cancer with a median survival of 6.7 months. Defined by NUTM1 fusions, NC enhances oncogene transcription, including MYC. We investigated the ability of standard next-generation sequencing (NGS) to identify NUTM1 fusions and describe additional molecular features of NC. EXPERIMENTAL DESIGN:This study included 116 patients with NC whose tumors underwent broad-panel NGS (>80 genes) of DNA, ctDNA, and/or RNA fusion sequencing between 2013 and 2024. NGS reports and medical records were manually reviewed. RESULTS:Of 116 patients (median age, 38; 40.5% female), 84.5% had DNA, 12.1% had ctDNA, and 51.7% had RNA fusion testing. In a subset of 100 patients with DNA/ctDNA testing, 92.9% (n = 79/85) had <10 pack-years/never-smoking history, and 58.8% (n = 47/80) had a BRD4::NUTM1 fusion. The median tumor mutational burden was 1.0 mut/Mb (range 0.0-16.0; n = 71 known), and 19.7% (n = 13/66) had PD-L1 expression ≥1%. DNA, ctDNA, RNA fusion, NUT IHC, and NUTM1 FISH detected NC fusions in 21.6%, 21.4%, 83.9%, 100.0%, and 91.9% of tests, respectively. Co-occurring pathogenic mutations included oncogenes PIK3CA, RET, and FGFR3 and tumor suppressors ATM and BRCA1 (n = 1 each). Secondary genes altered in >5% of NCs included LDL receptor-related protein 1B (LRP1B; 10.4%), histone-lysine N-methyltransferase 2D (KMT2D; 8.0%), and FAT atypical cadherin 1 (FAT1; 5.5%); common pathways with mutated genes were epigenetic (57.0%), cell cycle (26.0%), and DNA repair (24.0%). CONCLUSIONS:Standard DNA NGS detects less than a quarter of NCs; RNA-based fusion testing, or NUT IHC/NUTM1 FISH, should be routine for suspected NC. NCs are enriched in co-occurring epigenetic, cell cycle, and DNA repair alterations, warranting further evaluation.

Summary of response outcomes across subgroups

Supplemental Table 1. Representativeness of Study Participants

8635 Background: Lung cancer remains the deadliest solid tumor, with non-small cell lung cancer (NSCLC) accounting for 80–85% of cases. Patients enrolling in Phase I studies are often heavily pretreated and face limited treatment options. Understanding their demographics and therapeutic responses is crucial to improving patient outcomes. This study aimed to analyze therapeutic responses in NSCLC patients enrolled in Phase I studies. Methods: Data on NSCLC patients treated at the Investigational Cancer Therapeutics (ICT) department, a dedicated Phase I unit at The University of Texas MD Anderson Cancer Center (MDACC), were reviewed from January 2016 to December 2024, using MDACC CHIMERA platform. Collected data included age, gender, histologic type, Eastern Cooperative Oncology Group (ECOG) performance status, prior lines of treatment, treatment regimen, trial details and best response to treatment. Results: A total of 555 NSCLC patients were identified, of whom 267 (48.1%) were female. The median age was 64 years. The number of prior chemotherapy lines included: one line (50.3%), two lines (16%), and three lines or more (11.0%). The most frequent histologic type was adenocarcinoma (80.0%), followed by squamous cell carcinoma (15.7%) and NSCLC not otherwise specified (3.1%). The median number of treatment cycles was three, and the median duration of treatment was 2.0 months. The best response was evaluable in 449 cases (80.9%). The overall objective response rate (ORR) was 21.2%, and the disease control rate (DCR) was 71.3%. Clinical trial enrollment were categorized into seven groups: Targeted Monotherapy (TM), 227 cases (40.9%); Targeted Combination (TC), 49 cases (8.8%); Immunotherapy Monotherapy (IM), 81 cases (14.6%); Immunotherapy Combination, 57 cases (IC) (10.3%); Targeted + Immunotherapy (TI), 70 cases (12.6%); Antibody-Drug Conjugates (ADC) Monotherapy, 64 cases (11.5%); and Others (O), 7 cases (1.3%). When the ORR was compared among these groups, the TM group demonstrated the highest ORR at 32.6%, followed by the O group at 28.6% and the TC group at 28.2%. Conclusions: Phase I studies, especially those involving regimens containing targeted therapy, may serve as a promising treatment option for pretreated patients with advanced NSCLC. Comparison of best response and objective response rates between different regimen groups. Targeted Therapy (Monotherapy) Combination with Targeted Agent Immunotherapy (Monotherapy) Combination with Immunotherapy Targeted Therapy Combined with Immunotherapy Antibody-Drug Conjugate (Monotherapy) Others P CR 1.6% 0% 0% 0% 0% 0% 0% <0.001 PR 31.1% 28.2% 4.7% 11.6% 10.5% 12.2% 28.6% SD 43.2% 59.0% 57.8% 44.2% 57.9% 65.3% 0% PD 24.2% 12.8% 37.5% 44.2% 31.6% 22.4% 71.4% ORR 32.6% 28.2% 4.7% 11.6% 10.5% 12.2% 28.6% <0.001 Abbreviations: CR: Complete Response; PR: Partial Response; SD: Stable Disease; PD: Progressive Disease; ORR: Objective Response Rate.

3558 Background: Gene fusions are rare genomic events in colorectal cancer (CRC) and can co-occur with other potentially actionable alterations. Navigating early-phase trial selection in this scenario can be challenging. Objectives: To evaluate the genomic landscape, treatment selection, and clinical outcomes of patients (pts) with CRC harboring gene fusions enrolled in at least one clinical trial in the Department of Investigational Cancer Therapeutics, The University of Texas, MD Anderson Cancer Center. Methods: We used a computerized data extraction tool to review clinical and genomic data of pts with CRC harboring gene fusions between June 2011 and June 2024. The actionability of the genomic alterations was classified by the Precision Oncology Decision Support team. Median overall survival (mOS) was defined as the time from consent to the date of death or last follow up. Kaplan-Meier method was used to estimate survival and the Cox proportional hazards model to evaluate the impact of multiple variables. Results: 56 pts were included (28 female; 28 male) with a median age of 50.5 years (range 25-84). Colon was the most frequent tumor location (n=29, 51.8%) followed by rectal (n=15, 26.8%), small bowel (n=8, 14.3%), and appendiceal (n=3, 5.3%). Most cases had adenocarcinoma histology (n=54, 96.4%). The median time from diagnosis to completion of the molecular test identifying the gene fusion was 3 mo (range 0-36) for pts whose tumors were diagnosed in the metastatic setting. A total of 65 fusions were identified (range 1-3 per pt); 48 of them occurring in actionable or potentially actionable genes (78.8%) and 17 in non-actionable genes (26.2%). RAS alterations were present in 27 cases (48.2%) and BRAF alterations in 10 cases (17.9%). 8 patients had alterations in ERBB2 (14.3%) and 32 in genes related to DNA damage repair (DDR; 57.1%). In total, patients were enrolled in 91 trials (range 1-7 per pt). Target therapy (TT) was offered to 33 pts (58.9%) while the remaining were offered non-TT (n=23, 41.1%). Treatment selection was fusion-driven in 10 cases (17.9%; 4 NTRK, 3 FGFR, 1 RET, 1 ROS1, and 1 ALK). Notably, one pt with a ETV6-NTRK fusion was enrolled sequentially in 4 immunotherapy (IO) and TT trials, surviving on-trial for over 3.5 years. 3 patients received TT for KRAS (5.4%), 4 for BRAF (7.1%), 4 for ERBB2 (7.1%), and 8 for DDR alterations (14.3%). The mOS was 10.0 mo (95% CI, 5-NR) for pts receiving non-TT and 11 mo (95% CI, 3-21) for patients receiving TT (p 0.985). Age group (cutoff 50 years), gender, cancer stage at diagnosis, RAS, BRAF and DDR alterations, and treatment type (TT vs. non-TT or IO) were not related to survival in the multivariate level (p=0.9). Conclusions: A comprehensive genomic evaluation of pts with CRC harboring gene fusions can expand early-phase treatment possibilities. DDR alterations are more frequent in this group than in unselected CRC cohorts and can represent an additional target for TT.

2660 Background: DNA repair deficiency is common among tumors, and emerging data suggest that genomic instability is associated with response to immuno-oncology (IO) therapies (PMID 28630051). We evaluated patients with advanced metastatic cancer across tumor types, who were treated on the IMPACT2 study (NCT02152254) and analyzed their clinical outcomes by treatment type (anti-DNA damage repair [DDR] agents, IO, or other [non-IO, non-anti-DDR]). Methods: Patients had tumor biopsies followed by molecular profiling in a CLIA-certified lab. All cases were discussed at Molecular Tumor Board meetings. Patients were treated on early-phase clinical trials. Progression-free survival (PFS), overall response rate (ORR), and overall survival (OS) were compared by therapy type in patients with or without DDR/MMR mutations (DDR+). Results: Of 829 enrolled patients, 510 had molecular profiling and received anticancer therapy: 85 were DDR+ (PS 1, 75%; med. age, 59 yrs; males 56.5%; med prior therapies 4; PDL1+ 44.9%; TMB-H 18%,) and 425 DDR- (PS 1, 88%; med. age, 60 yrs; men 47%, med. No. prior therapies 3; PDL1+ 46%; TMB-H 6.5%). Results are shown in the Table. In patients with DDR mutations, IO was associated with a higher ORR compared with “other” treatments (p=0.044); and with longer PFS compared with anti-DDR therapies (p = 0.044); no difference was noted in OS by treatment type. In the DDR-neg group, IO was associated with longer PFS (p = 0.017), and longer OS (p = 0.0004) compared with anti-DDR therapy; OS was longer in the IO group compared with “other” treatments (p = 0.029) and in the “other” treatments group compared with anti-DDR agents (p=0.006). Conclusions: Our data indicate the complexity of assessing outcomes in patients with various tumor types and without DDR mutations. Further work is needed to develop predictive biomarkers for IO and anti-DDR agents. Clinical trial information: NCT02152254 . Rx type DDR-pos N Outcome Anti-DDR vs IO IO vs Other Anti-DDR vs Other DDR-neg N Outcome Anti-DDR vs IO IO vs Other Anti-DDR vs Other Overall response (%) IO 34 6 (25%) OR* 0.78; p=0.81 OR* 20.73; p=0.044 OR* 16.09;p=0.1 116 9 (11%) OR* 0.34;p=0.46 OR* 2;p=0.13 OR* 0.67;p=0.79 Anti-DDR 18 1 (16.7%) 52 0 (0%) Other 33 0 (0%) 257 12 (5.9%) PFS, med. (95% CI) IO 34 4.26 (2.3, 7.4) HR 2.01 (1.02,3.96) (p=0.044) HR 0.91 (0.56,1.47)p=0.69 HR 1.82 (0.95,3.49)p=0.07 116 5.42(3.02,6.61) HR 2.02 (1.14,3.61) p= 0.017 HR 0.85 (0.68,1.06)p= 0.15 HR 1.71 (0.98,2.99)p=0.058 Anti-DDR 18 2.58 (1.68, NA) 52 1.64 (1.45, NA) Other 33 4.54 (3.06, 6.54) 257 3.98 (3.45, 4.57) OS, med. (95% CI) IO 34 14.37 (10.22, NA) HR 1.68 (0.82,3.42)p=0.16 HR 0.73 (0.43,1.23)p=0.24 HR 1.23 (0.62,2.42)p=0.56 116 12.46 (8.75, 20.78) HR 2.98 (1.62,5.48) p= 0.0004 HR 0.76 (0.6,0.97) p=0.029 HR 2.27 (1.27,4.07) p= 0.006 Anti-DDR 18 8.98 (3.48, NA) 52 4.31 (2.66, NA) Other 33 10.82 (7, 16.83) 257 9.4 (8.28, 11.21) *OR, Odds Ratio, p values, unadjusted for multiple comparisons.

ABSTRACT Introduction TLR9 agonists are immunomodulators that have been of interest for combined use with cancer immunotherapy. TLR9 agonists, such as lefitolimod (MGN1703), significantly increased Th1 response in preclinical models and have demonstrated efficacy in early clinical trials. This trial assessed the safety and preliminary efficacy of the combination of lefitolimod and ipilimumab in patients with advanced solid tumors. Methods This was a single-center, open-label, investigator-initiated phase I trial conducted at The University of Texas MD Anderson Cancer Center. Patients received leftolimod either subcutaneously (at escalating doses of 15–120 mg) or intratumorally (at the maximum feasible dose) in combination with ipilimumab (3 mg/kg). Paired biopsy samples were collected before the start of treatment and after two treatment cycles and analyzed by flow cytometry. Results We enrolled a total of 28 patients in this study with a median age of 56 years (range 19–75) in the escalation cohort and 60 years (range 34 –92) in the expansion cohort. The median number of prior lines of therapy was 4 (range 0–12). Eleven patients had at least one treatment-related adverse event (TRAE). The most common TRAEs were skin rash (n = 4, 14%), fatigue (n = 3, 11%), and pruritis (n = 2, 7%). No grade 4 or 5 AEs occurred, and no patients required dose reduction or treatment discontinuation due to AEs. The maximum tolerated dose (MTD) was not reached in this study. Of 28 patients, 21 patients had response-evaluable disease. No patients had a complete or partial response; 8 and 13 patients had stable and progressive disease as the best response, respectively. Paired biopsy samples were obtained from five patients. Increases in intratumoral CD8 T-cell frequency, memory CD8 phenotype (CD45RO+), and proliferation (Ki67+) in four of five patients suggested that the combination of lefitolimod and ipilimumab led to proinflammatory immune conditioning of the tumor microenvironment. Conclusions The combination of lefitolimod (administered subcutaneously or intratumorally) and ipilimumab was safe and well tolerated but demonstrated modest antitumor activity in patients with advanced cancers. ClinicalTrials.gov ID: NCT02668770