Article Text

Download PDFPDF

1245 Th17 and Type 2 CD8+ cytokine signatures predict irAEs in solid tumors
  1. Chester J Kao1,
  2. Soren Charmsaz1,
  3. Madalena Brancati1,
  4. Stephanie L Alden1,
  5. Howard L Li1,
  6. Aanika Balaji1,
  7. Kabeer Munjal1,
  8. Hua-Ling Tsai1,
  9. Ludmila Danilova1,
  10. Alexei Hernandez1,
  11. Nicole Gross1,
  12. Erin M Coyne1,
  13. Sarah M Shin1,
  14. Jennifer Durham1,
  15. Brian Christmas1,
  16. Christopher Thoburn1,
  17. Maximilian F Konig1,
  18. Evan J Lipson2,
  19. Jarushka Naidoo1,
  20. Laura Cappelli1,
  21. Yasser Ged1,
  22. Marina Barretti1,
  23. Julie Brahmer1,
  24. Jean Hoffman-Censits1,
  25. Tanguy Y Seiwert1,
  26. Rachel Garonce-Hediger3,
  27. Sanjay Bansal4,
  28. Laura Tang5,
  29. Elizabeth M Jaffee1,
  30. G Scott Chandler6,
  31. Rajat Mohindra7,
  32. Won Jin Ho1 and
  33. Mark Yarchoan1
  1. 1Johns Hopkins University School of Medicine, Baltimore, MD, USA
  2. 2The Bloomberg∼Kimmel Institute for Cancer Immunotherapy, Johns Hopkins University, Baltimore, MD, USA
  3. 3F. Hoffmann-La Roche Ltd, Basel, Switzerland
  4. 4Genentech, Pearland, TX, USA
  5. 5Genentech, Hoboken, NJ, USA
  6. 6Genentech/Roche, Basel, Switzerland
  7. 7F.Hoffmann-La Roche Ltd, Basel, Switzerland
  • Journal for ImmunoTherapy of Cancer (JITC) preprint. The copyright holder for this preprint are the authors/funders, who have granted JITC permission to display the preprint. All rights reserved. No reuse allowed without permission.


Background Immune related adverse events (irAEs) and their associated morbidity/mortality are a key concern for patients receiving immune checkpoint inhibitors (ICIs). A prospective evaluation of the drivers of irAEs in a diverse pan-tumor cohort is needed to identify patients at greatest risk and to develop rational interception strategies.

Methods We prospectively collected clinical data and blood samples from patients with solid tumors at a single institution who received ICIs as standard of care. Blood samples were collected at baseline and early on treatment (month 1 or 2). We analyzed 32 circulating cytokines with Luminex multiplex assay and utilized Cytometry by Time-of-Flight (CyTOF) in an enriched cohort to investigate mechanisms of irAEs. Grade 2 or higher irAEs by CTCAE v5.0 were analyzed to enrich for clinically meaningful toxicities. Multi-testing adjustment utilizing false discovery rate (fdr) was performed for the primary cytokine time to irAE Cox analysis at baseline and early treatment.

Results 104 patients were prospectively enrolled. Reported race was 64.4% white, 28.8% black, and 2.9% Asian. 22.1% received combination anti-PD1/CTLA-4 therapy while 77.9% anti-PD1/PDL1 (monotherapy or in combination with other therapies). IrAEs occurred in 37.5% of the total patients with median onset of 2.1 months: 65.2% in anti-PD1/CTLA-4 (median 1.2 months) and 29.6% in anti-PD1/PDL1 groups (median of 2.9 months). Reported race was not associated with irAE development. No individual cytokine at baseline was associated with subsequent development of an irAE. In the Cox model, early fold changes in interleukin (Il)-6, Il-17f, Il-13, and Il-25 were significantly associated with the development of an irAE (table 1, adjusted p<0.05) and stratified patients by optimal cutoffs utilizing max log-rank statistics (figure 1, p<0.05). In the subgroup analysis, Il-6, Il-17f, and Il-25 were also significantly upregulated (figure 2, p<0.05) in patients who developed specific irAEs. To identify the cellular populations underlying these cytokine changes, we performed CyTOF on 15 irAE patients matched with 15 non-irAE patients by regimen and tumor type. Higher baseline proportions of two CD8+ T cell populations (CCR3+ and CCR4+), known to secrete type 2 cytokines,1 2 were associated with irAEs, while on-treatment persistent elevation of CD8+CCR4+ and peripheral expansion of effector memory Th17+ cells were associated with irAEs (figure 3, p<0.05).

Conclusions In a diverse, pan-tumor cohort, Il-6/Il-17f related Th17 and Il-13/Il-25 related CD8+ type 2 cytokine signatures are associated with the development of irAEs, serving as possible targets for monitoring and therapeutic interventions.

Acknowledgements The Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, F. Hoffman-La Roche Ltd., and Genentech are members of the imCORE Network.


  1. Kondo T, Takiguchi M. Human memory CCR4+CD8+ T cell subset has the ability to produce multiple cytokines. International Immunology. 2009;21:523–532.

  2. Fuschiotti P, Larregina A, Ho J, Feghali-Bostwick C, Medsger T. IL-13-producing CD8+ T cells mediate dermal fibrosis in patients with systemic sclerosis. Arthritis Rheum. 2013;65:236–246.

Ethics Approval This study was approved by the Johns Hopkins institution’s Ethics Board #IRB00267960.

Abstract 1245 Table 1

Significant cytokines in the adjusted time to irAE Cox model. HRs are reported per unit increase in fold change and were adjusted for treatment regimen used (combination anti-PD1/CTLA-4 treatment vs. anti-PD-1/PD-L1 groups). Abbreviations: HR = hazard ratio

Abstract 1245 Figure 1

Kaplan-Meier Curves of cumulative irAE stratified by early treatment cytokine fold changes. (A) IL-6 at cutoff 0.81, (B) II-17f at cutoff 1.4, (C) II-13 at cutoff of 1.1, (D) II-25 at cutoff of 1.6, (E) combined cytokine status. High combined cytokine status was defined as ≥2 cytokines were high, an intermediate status was defined when at least one cytokine was high, and a low status was when all four cytokines were low

Abstract 1245 Figure 2

transformed early treatment fold change between patients with no irAEs and specific types of irAEs: (A) II-6, (B) II-17f, (C) II-25

Abstract 1245 Figure 3

Proportion of cell populations. (A) Baseline, (B) On Treatment. Abbreviations: Tc_I = CD8+CCR3+,Tc_II = CD8+CCR4+, Thi7_EM = CD4+CD45RO+RORYT+Ki67+

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.