An empirical antigen selection method identifies neoantigens that either elicit broad antitumor t-cell responses or drive tumor growth

Hubert Lam, Lisa K. McNeil, Hanna Starobinets, Victoria L. Devault, Roger B. Cohen, Przemyslaw Twardowski, Melissa L. Johnson, Maura L. Gillison, Mark N. Stein, Ulka N. Vaishampayan, Arthur P. Decillis, James J. Foti, Vijetha Vemulapalli, Emily Tjon, Kyle Ferber, Daniel B. Deoliveira, Wendy Broom, Parul Agnihotri, Elizabeth M. Jaffee, Kwok Kin WongCharles G. Drake, Pamela M. Carroll, Thomas A. Davis, Jessica Baker Flechtner

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Neoantigens are critical targets of antitumor T-cell responses. The ATLAS bioas-say was developed to identify neoantigens empirically by expressing each unique patient-specific tumor mutation individually in Escherichia coli, pulsing autologous dendritic cells in an ordered array, and testing the patient’s T cells for recognition in an overnight assay. Profiling of T cells from patients with lung cancer revealed both stimulatory and inhibitory responses to individual neoantigens. In the murine B16F10 melanoma model, therapeutic immunization with ATLAS-identified stimulatory neoantigens protected animals, whereas immunization with peptides associated with inhibitory ATLAS responses resulted in accelerated tumor growth and abolished efficacy of an otherwise protective vaccine. A planned interim analysis of a clinical study testing a poly-ICLC adjuvanted personalized vaccine containing ATLAS-identified stimulatory neoantigens showed that it is well tolerated. In an adjuvant setting, immunized patients generated both CD4+ and CD8+ T-cell responses, with immune responses to 99% of the vaccinated peptide antigens. SIGNIFICANCE: Predicting neoantigens in silico has progressed, but empirical testing shows that T-cell responses are more nuanced than straightforward MHC antigen recognition. The ATLAS bioassay screens tumor mutations to uncover preexisting, patient-relevant neoantigen T-cell responses and reveals a new class of putatively deleterious responses that could affect cancer immunotherapy design.

Original languageEnglish (US)
Pages (from-to)696-713
Number of pages18
JournalCancer discovery
Issue number3
StatePublished - Mar 2021

ASJC Scopus subject areas

  • Oncology


Dive into the research topics of 'An empirical antigen selection method identifies neoantigens that either elicit broad antitumor t-cell responses or drive tumor growth'. Together they form a unique fingerprint.

Cite this