Oral Presentation Australasian Cytometry Society 44th Annual Conference and Workshop

High-dimensional panel design for spectral flow cytometric evaluation of T-cell reinvigoration by immune checkpoint blockade in melanoma. (24545)

Jack M. Edwards 1 2 , Miles Andrews 3 , Hayley Burridge 3 , Robin Smith 2 , Carole Owens 2 , Mark Edinger 4 , Kate Pilkington 4 , Juliette Desfrancois 4 , Mark Shackleton 3 , Sasha Senthi 2 , Menno C. van Zelm 1
  1. Department of Immunology and Pathology, Monash University, Melbourne, Victoria, Australia
  2. Alfred Health Radiation Oncology, The Alfred Hospital, Melbourne, Victoria, Australia
  3. Department of Medical Oncology, The Alfred Hospital, Melbourne, Victoria, Australia
  4. Cytek Biosciences, Fremont, California, U.S.A

Rationale

Despite the success of immune checkpoint blockade therapy, most metastatic melanoma patients fail to respond to therapy or experience severe toxicity. Assessment of biomarkers before or early into treatment, in conjunction with longitudinal immunophenotyping, will help to understand favorable responses and improve therapeutic outcomes. We present a high-dimensional approach for T-cell immune profiling of peripheral blood by spectral flow cytometry.

Methods

Three multi-parameter cytometry panels were developed: (1) a conventional TruCount panel for absolute cell counts at time of sampling, (2) a 27-color spectral panel assessing ex vivo T-cell markers, and (3) a 20-color spectral panel evaluating cytokine expression after CD3/CD28 stimulation. Pre-treatment blood mononuclear cells from patients and healthy controls were cryopreserved before staining across 11 batches. Batch effects were tracked with a single-donor control and the suitability of normalization assessed. Data were analyzed using expert gating and high-dimensional strategies.

Results

Batch-to-batch variation was minimal and did not affect analysis, as demonstrated by dimensionality reduction of batch-control samples, and normalization did not improve manual or high-dimensional analysis. At baseline, patients had significantly fewer lymphocytes than controls (1390 vs 1742 cells/μL), due to lower naive CD4 (244 vs 387 cells/μL) and CD8 (34 vs 88 cells/μL) T-cells, and follicular helper T-cells (31 vs 64 cells/μL). Patients had higher portions of Ki67 and IL-2 expressing cells within multiple CD4 and CD8 memory subsets, and increased CD57 (p=0.023), EOMES (p=0.019), TIGIT (p=0.039), CD56 (p=0.047) and Granzyme B (p=0.032) expression within TCRγδ+ T-cells.

Discussion

Our high-parameter spectral flow cytometry panels provide in-depth profiles of blood T-cells, capable of detecting abnormalities in untreated melanoma patients. The robustness of our approach as demonstrated by minimal batch effects makes this highly suited for longitudinal evaluation of treatment effects.  Pre-treatment, patients showed evidence of immune activation as well as reduced naïve T-cell counts.