RESEARCH AREAS OF FOCUS

Immune checkpoint inhibitors have revolutionized cancer care, but are associated with morbid and potentially life-threatening toxicities known as immune-related adverse events. Of these, skin irAEs are the most common and among the earliest toxicities to occur. Our lab has published the seminal studies on the epidemiology, risk factors, and downstream therapeutic implications of skin irAEs as well as their relationship to irAEs of other organ systems and established these toxicities as early biomarkers of immunotherapy treatment response. We have proposed strategies for the algorithmic phenotyping of toxicities in electronic health record data and developed definitions to enable clinicians in diagnosing these events. Our ongoing work combines clinical, genetic, and histopathologic imaging data to develop multi-modal approaches for predicting development of immunotherapy toxicities across different organ systems and correlating these toxicities with ultimately immunotherapy treatment outcomes. 

• Association of Cutaneous Immune-Related Adverse Events With Increased Survival in Patients Treated With Anti-Programmed Cell Death 1 and Anti-Programmed Cell Death Ligand 1 Therapy; JAMA Dermatology 2022
• Multi-organ immune-related adverse events from immune checkpoint inhibitors and their downstream implications: a retrospective multicohort study; The Lancet Oncology 2024 
• Cutaneous immune-related adverse events are associated with longer overall survival in advanced cancer patients on immune checkpoint inhibitors: A multi-institutional cohort study; JAAD 2023

Melanoma is one of the deadliest forms of skin cancer, with more than 300,000 new cases diagnosed annually worldwide. Despite the significant improvement in survival among immunotherapy recipients with metastatic disease, the application of immunotherapy to earlier-stage melanoma has been slow given concerns of balancing risk of treatment-related toxicity with risk of melanoma recurrence. Our laboratory has developed multiple machine learning approaches to improving the prognostication of melanoma recurrence and melanoma-specific survival using a combination of clinical and digital histopathologic features. We aim to continue refining these approaches and ultimately enable their clinical deployment through external validation of these approaches across multiple institutions and treatment settings.   

• Prediction of early-stage melanoma recurrence using clinical and histopathologic features; NPJ Precision Oncology 2022
• Development and validation of time-to-event models to predict metastatic recurrence of localized cutaneous melanoma; JAAD 2024

Recent advances in spatial biology have enabled robust visualization of the tumor microenvironment. However, these tools has also created the need for more sophisticated computational approaches that are able to rapidly analyze and integrate inferences from complementary experimental techniques, often drawing from hundreds of thousands to millions of cells per single tumor section. Our laboratory has been developing automated profiling approaches that combine the power of multiple imaging, spatial transcriptional profiling, and digital histopathology to improve our our understanding of cancer initiation, progression, and treatment response. This multi-modal approach also assists in providing mechanistic explainability to the typically “black-boxed” deep learning image analysis approaches used in conventional computational pathology. 

• SpatialCells: automated profiling of tumor microenvironments with spatially resolved multiplexed single-cell data; Briefing in Bioinformatics 2024