Distinct Immune Cell Populations Define Response to Anti‑PD‑1 Monotherapy and Anti‑PD‑1/Anti‑CTLA‑4 Combined Therapy

Using T cells directly or indirectly as therapeutic targets for cancer has revolutionized cancer immunotherapy. Removal of immunosuppressive factors by the use of checkpoint inhibitors such as PD-1, PD-L1 and CTLA-4, highlight therapeutic approaches that are showing great promise in the clinic due to good anti-tumour effects in patients with melanoma and other types of cancer. Anti‑PD‑1 antibodies such as nivolumab and pembrolizumab show high anti-tumour efficacy and minimal toxicity in metastatic melanoma. Combinations of anti‑PD‑1 and anti‑CTLA‑4 are even more promising but significantly more toxic than the monotherapy.

Many patients fail to respond to either treatment type and given the range of alternative therapies being evaluated in clinical trials, such as IDO, GITR, TIGIT and LAG3 Gide et al., attempted to discover potential biomarkers of response and resistance in a cohort of melanoma patients.

120 patients either treated with anti-PD-1 monotherapy (63) or combined anti‑CTLA‑4 and anti‑PD‑1 therapy (57) were stratified according to RECIST response and progression free survival. Responders were patients with complete response (CR), partial response (PR), or stable disease (SD) of greater than 6 months with no progression whereas non-responders were classified by progressive disease (PD) or SD for less than or equal to 6 months before disease progression

Differential expression of genes (DEGs) in responders versus non-responders was investigated in biopsies taken at baseline and early during therapy (EDT). 310 DEGs were identified that clustered into immune signalling or cellular signal transduction.

For anti-PD1 monotherapy responders had high expression of IFN-related genes such as TBX21, STAT1, IRF1, TNF, and IFNG. Tumour infiltrating T-cell genes, CD8A, CD8B, ITGAE [CD103], PDCD1 [PD-1], CCL5, CXCL13 and IL2 were also associated with better outcomes. TIGIT, TNFRSF9 (CD137), IDO and LAG3, which are immunosuppressive checkpoint related genes, were also highly expressed in responders compared to non-responders.

Gene set enrichment analysis identified over-representation of immune signalling, JAK/STAT signalling and cytokine and chemokine signalling pathways. Anti‑PD‑1 monotherapy responders have enriched for IFN- and T cell-mediated immunity compared with non-responders.

For anti‑CTLA‑4 and anti‑PD‑1 there were 328 DEGs between responders and non-responders. Pathway analysis revealed T-cell-related genes such as CD8A, CD247, CD5, CD6, and CD69 and genes associated with NK cell-mediated cytotoxicity (CD96). T-cell cytotoxicity genes such as GZMK, CD274 [PD-L1], CD2, and ITGAL were also expressed higher in responders in addition to cytokine signalling genes, CXCL13, CCL4, CCR5, CCL5, and CXCL9.

Analysis of the DEGs in the baseline samples of responders and non-responders identified higher expression of carbonic anhydrase IX, WNT3, LAMB4, ITIH5 and glutamate receptors GRIA2, GRIA4 and GRIK3 in non-responders to anti-PD1 monotherapy. These genes may be involved in resistance to immunotherapies.

For the combined immunotherapies DEGs identified in non-responders clustered into three main pathways, WNT, melanogenesis and oxidative phosphorylation, although these analyses failed to reach significance.

In the combined immunotherapy cohort in the early during therapy (EDT) biopsies CD8 and EOMES expression, at log2 expression >3 and 2 respectively distinguished responders from non-responders.

FFPE melanoma biopsies were assessed for T-cell localisation and known predictive markers (PD-L1) with multiplex immunofluorescence. Higher infiltrates of immune cells were detected in responders to monotherapy or combined immunotherapy pre-treatment and during treatment.

Significantly higher intratumoural staining of CD8 and PD-L1 was also detected in pre-treated samples for both mono and combined therapies. Increased peritumoural expression of PD-L1 and PD-1 was found in pre-treatment responders to monotherapy.

In EDT intratumoural PD-L1 and PD-1 expression was significantly higher in both therapies whereas CD8 was only associated with the combined therapy. CD8 and PD-1 had significantly higher peritumoural expression in responders to both types of therapy indicating increased T-cell tumour infiltration and checkpoint inhibition.

Multiplex immunofluorescence confirmed higher intratumoural expression of CD45RO, EOMES, and TBET in PRE samples of responders to monotherapy. Responders also had significantly higher numbers of activated T-cells in EDT samples with both mono- and combined therapy as indicated by Granzyme B expression.

Using CyTOF on dissociated melanoma tissue with a panel of 43 markers of immune cell subtypes 3 different T-cell clusters were identified. A T-cell population with markers CD45RO+EOMES+ was highly abundant in responders to the combined therapy.

Monotherapy treated patients with high expression of the CD8+/CD4+ EOMES+ CD69+CD45RO+ memory T-cell phenotype had longer PFS times than those with low expression. PFS for combined therapy was also longer but did not reach significance.

Tumour shrinkage was observed in both combined (81%) and monotherapy (71%) in tumours with high levels of CD8/CD4, EOMES, CD69, and CD45RO.

In this study the authors identify biomarkers for response and resistance to mono- and combined immunotherapies. They have also quantified specific T cell populations in melanomas treated with anti‑PD‑1 monotherapy or combined anti‑PD‑1 and anti‑CTLA‑4 immunotherapy that can be used to predict patient response. The authors also noted that non-responders to monotherapy expressed alternate immune checkpoint genes e.g., IDO1, ICOS, and TIGIT, whereas non-responders to the combination did not. This information could be vital for selecting patients for specific therapies and identifying those that would benefit from alternative therapies.

Want to ask a question about


Epistem’s Models and Services for Immune Checkpoint Inhibitors

At Epistem we can offer a validated orthotopic imaging model of leukaemic and solid tumours for drug development studies. We have extensive experience in in vivo imaging to monitor treatment effects in the same animal over time and to detect treatment effects/potentiation and disease regression using single drug or a combination of additional therapeutics, including CAR‑T cell therapies. We also offer validated models to address immune function in tumour bearing mice.

Our detailed flow cytometric expertise can further assess the identity of tumour infiltrating and peripheral cell populations such as CD8+ T cells, MDCSs and regulatory T cells, by surface identification of key molecules, or the intracellular mediators they express (cytokines, transcription factors).

All our models are supported by GCLP-compliant gene expression and IHC services for biomarker discovery and validation. Our laser capture microdissection service is ideal for extracting specific tumour infiltrating cells using “RNA friendly” staining suitable for downstream expression analysis.

Gide et al., 2019. “Distinct Immune Cell Populations Define Response to Anti-PD-1 Monotherapy and Anti-PD-1/Anti-CTLA-4 Combined Therapy”. Cancer Cell, 35, 238–255 DOI: 10.1016/j.ccell.2019.01.003

About Epistem

Epistem's contract research service is committed to providing reliable, innovative and transferable pre-clinical models and services to support decision making throughout the drug discovery and development pipeline.

Tel: +44 (0)161 850 7600  Email: info@epistem.co.uk