{
  "abstract": "Background Recent advances in artificial intelligence (AI) and digital pathology have enabled comprehensive characterization of the tumor microenvironment (TME). Understanding the spatial relationship between immune-response cellular effectors and tumor cells in the TME can support elucidation of complex MOAs with antibody-drug conjugate (ADC) therapies. TV is a tissue factor-directed ADC hypothesized to drive apoptosis via known MOAs, including direct cytotoxicity, antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), and immunogenic cell death (ICD), with ADCC, ADCP, and ICD depending on immune cell presence in the TME. 1 In this exploratory analysis of pretreatment biopsy specimens from the global, open-label, phase 2 innovaTV 207 trial (part C; HNSCC cohort; NCT03485209),2 an AI-trained digital pathology algorithm was used to systematically identify baseline spatial features associated with response to therapy consistent with TV MOA.Methods Biopsy samples from patients with previously treated recurrent/metastatic HNSCC were collected at screening were used to identify spatial histological features from hematoxylin and eosin (H&E)-stained whole-slide images using the PathExplore algorithm (PathAI). In part 1, a hypothesis-driven approach was used to define a set of prespecified features corresponding to known TV MOAs; these were selected prior to examination of numerical data and analyzed for association with response and feature correlation. In part 2, all 414 features included by the PathExplore algorithm were analyzed for association with response, and those with P≤0.20 were analyzed for feature correlation and clustering.Results 19/40 enrolled patients (9 responders, 10 non-responders) were included in the biomarker-evaluable dataset following pathologist-supervised quality control of H&E-stained tissue. In the prespecified approach, 25/414 features (6%) were preselected for potential association with TV MOAs; 10 (40%) had a raw P value <0.05, and 19 (76%) had an adjusted P value <0.20 for association with response. Each TV MOA had ≥1 feature associated with response. Low raw P values were highly enriched in the preselected feature set vs the overall set. Spatial proximity of immune cells (lymphocytes and macrophages) to tumor cells in the tumor area was associated with response to TV. Supervised learning showed that features from the preselected MOA set were enriched as important features for response.Conclusions This proof-of-concept analysis highlights a novel approach using routinely collected H&E staining methods to identify spatial features associated with response to TV and improves understanding of the hypothesized immune-related TV MOAs. Further validation of digital pathology capabilities and translational research may aid in elucidation of complex MOAs.Acknowledgements The authors thank all patients who participated in the innovaTV 207 study, their families and caregivers, and the investigators and research staff at all clinical sites. This study was sponsored by Seagen Inc. (acquired by Pfizer in December 2023) and Genmab, the codevelopers of tisotumab vedotin. Medical writing assistance, funded by Pfizer according to Good Publication Practice guidelines, was provided by Akshaya Srinivasan, PhD, CMPP, of Nucleus Global, an Inizio Company.Trial Registration ClinicalTrials.gov ID, NCT03485209References Gray E, Hensley K, Allred S, et al. Tisotumab vedotin shows immunomodulatory activity through induction of immunogenic cell death. J Immunother Cancer. 2020;8:A371.Sun L, Fayette J, Salas S, et al. Tisotumab vedotin in head and neck squamous cell carcinoma: updated analysis from innovaTV 207 part C. J Clin Oncol. 2024;42:S6012.Ethics Approval All patients provided informed consent before taking part in the study. Independent ethics committee or institutional review board approval was obtained for each study site (IRB number at first enrolling site, 22-312; IRB registration number, 00001230).",
  "authors": [
    {
      "affiliations": [
        "Pfizer, Bothell, WA, USA"
      ],
      "name": "Mark C Bieda"
    },
    {
      "affiliations": [
        "Pfizer Inc., La Jolla, CA, USA"
      ],
      "name": "Diane Fernandez"
    },
    {
      "affiliations": [
        "Pfizer Inc., Cary, NC, USA"
      ],
      "name": "Parimi Vamsi"
    },
    {
      "affiliations": [
        "Pfizer Inc., Tucson, AZ, USA"
      ],
      "name": "Christine Boyiddle"
    },
    {
      "affiliations": [
        "Pfizer, Bothell, WA, USA"
      ],
      "name": "Dana Vento"
    },
    {
      "affiliations": [
        "Pfizer Inc., Houston, TX, USA"
      ],
      "name": "Maura Correas"
    },
    {
      "affiliations": [
        "Genmab US, Inc., Plainsboro, NJ, USA"
      ],
      "name": "Jeffrey R Harris"
    },
    {
      "affiliations": [
        "Genmab US, Inc., Plainsboro, NJ, USA"
      ],
      "name": "Sriram Sridhar"
    },
    {
      "affiliations": [
        "Pfizer Inc., Los Angeles, CA, USA"
      ],
      "name": "Carmen Wright"
    },
    {
      "affiliations": [
        "Pfizer, Bothell, WA, USA"
      ],
      "name": "Yitong J Zhang"
    },
    {
      "affiliations": [
        "Pfizer Inc., New York, NY, USA"
      ],
      "name": "Umberto Conte"
    },
    {
      "affiliations": [
        "Pfizer Inc., South San Francisco, CA, USA"
      ],
      "name": "Khyati Shah"
    }
  ],
  "title": "53 Assessing the mechanism of action (MOA) of tisotumab vedotin (TV) in head and neck squamous cell carcinoma (HNSCC) with digital pathology analysis of H&E images from innovaTV207 part C",
  "uid": "0b62cbbf-1978-5828-8b78-35135adebedc"
}