Immune checkpoint blockade (ICB) has two main limitations: responses vary between subgroups of patients and immune-related adverse events (irAEs) can be severe. To assess the mechanism of action behind the induction and variability of such responses, Hailemichael, Johnson, and Abdel-Wahab et al. performed a deep immune analysis of tumor and immune-related enterocolitis tissues from ICB-treated patients and animal models. Their results were recently published in Cancer Cell.

The researchers profiled immune-relevant gene expression in matched immune-related enterocolitis (irEC) and normal colon tissue from patients treated with ICB (anti-PD-1 alone, anti-CTLA-4 alone, or the combination). This revealed 52 upregulated genes in the irEC tissue, with the highest expression fold change for IL6. Other upregulated genes included acute-phase reactants induced by IL6, such as IL11, and genes encoding neutrophil- and monocyte-attracting chemokines. Cell scores calculated based on the gene expression profiles revealed that the cell expression gene scores for Th17, Th1, and neutrophils were upregulated in irEC. Multiplex IHC confirmed higher percentages of Th17, Th1, and hybrid Th1/Th17 cells in irEC tissues.

Comparing patients treated with anti-CTLA-4 to those treated with anti-PD-1, similar upregulation of IL6 and genes encoding chemokines for neutrophils and monocytes were found. However, the Th17 cell score was only higher in irEC of the anti-CTLA-4 group, not the anti-PD-1 group. Multiplex IHC confirmed more Th17 memory cells in anti-CTLA-4-treated irEC samples.

Immune signatures from a cohort of patients with melanoma who had tumor biopsies taken at baseline and at 6 weeks after initiation of anti-CTLA-4 were compared with the same gene expression panel used above between responders and non-responders and between the tumor tissue,  irEC tissue, and normal tissue. There were 343/770 genes differentially expressed between irEC tissue and responder tumors, while only 70 genes were differentially expressed between irEC and non-responding tumors, indicating more similarity between irEC and non-responding tumor tissue. Among the 343 genes, the largest differences were found in IL6 and genes encoding for neutrophil and monocyte chemokines, while there was no significant difference for these genes between irEC and non-responding tumor tissue. Comparing gene expression of irEC with responding tumors, multiple Th17- and Th1-differentiating cytokines were found to be differentially expressed. The Th17 cell score was higher than Th1, and the neutrophil score was higher than CD8⁺ T cell scores in the irEC analysis. These differences were not found between irEC and non-responding tumors.

The researchers then moved to animal models to study the mechanisms behind these findings. To assess whether increased IL-6 expression in irEC and non-responding tumors is mechanistically related to tumor resistance to ICB,  anti-CTLA-4 was tested in the aggressive B16.BL6 melanoma model. Mice received GVAX and anti-CTLA-4 or IgG control. In tumors of mice receiving anti-CTLA-4, there was an increase in IL-6, IL-1β, and TNFɑ expression. Next, the impact of adding IL-6 blockade was tested. Mice that received IL-6 blockade alone had reduced tumor burden, but none of the mice were cured. Anti-CTLA-4 therapy resulted in 20% of mice alive at day 60, while combined blockade of CTLA-4 and IL-6 resulted in 58% of mice alive. In Balb/c mice bearing immunogenic CT26 tumors, adding IL-6 blockade to anti-CTLA-4 resulted in improved tumor reduction, with 48% of mice cured, while anti-CTLA-4 alone resulted in 32% cures. No mice were cured with IL-6 blockade alone, but it did reduce the tumor load and improve survival.

The researchers then assessed the mechanism behind the synergistic antitumor response by assessing tumoral immune infiltration. The addition of IL-6 blockade resulted in an increase of CD44^hiCD11a^hiCD8⁺ and CD4⁺ effector T cells (T_eff) induced by anti-CTLA-4. There was also an increase in the number of CD8⁺ T_eff specific to B16 melanoma antigens (p15 and TRP-2). Polyclonal CD4⁺ and CD8⁺ T_eff had increased expression of activation (PD-1, Tim-3, Lag-3, ICOS), survival (Bcl-2), and cytotoxicity markers (granzyme B, IFNγ, TNFα, T-bet). There was also an increase in Th1 cytokines and IFNγ-induced chemokines, while Th17 cytokine levels decreased. CyTOF analysis revealed 10 CD8⁺ T cell clusters, which could be characterized as cytotoxic and/or exhausted, and most of which were associated with antitumor response. Meanwhile, among 8 CD4⁺ T cell clusters, three were Th1/Th17-like clusters, which were not associated with antitumor responses. There were also two CD4⁺ regulatory T cell (T_reg) clusters, of which one (ICOS^-CD62L⁺) had an inverse correlation with antitumor responses. Therefore, IL-6 blockade increased the T_eff/T_reg ratio in both the CD8 and CD4 compartments.

IL-6 blockade reduced the levels of IL-17⁺ Th17-like CD4⁺ T cells and the expression of CXCL1. This lower level of chemotactic CXCL1 might be involved in the lower levels of granulocytic and monocytic myeloid-derived suppressor cells (MDSCs) in tumors treated with IL-6 blockade. Treatment also reduced the number of tumor-associated macrophages (TAMs) of both the M1 and M2 phenotypes, which may be related to lower levels of CCL2. 

Next, the effects of anti-CTLA-4 and anti-IL-6 on the severity and incidence of experimental autoimmune encephalomyelitis (EAE) was assessed by inducing EAE in C57BL/6 mice by immunization with MOG. Mice treated with anti-CTLA-4 experienced increased EAE disease severity compared to mice receiving IL-6 blockade or the combination treatment. EAE-induced paralysis was more common in the anti-CTLA-4 group (100%), while 50% of those treated with the combination, and only 20% in the anti-IL-6 monotherapy group developed paralysis. In the Balb/c model, which is resistant to EAE development by MOG immunization alone, immunized mice treated with anti-CTLA-4 had a 4-fold increase in disease severity compared to untreated mice, while these symptoms were not observed in immunized mice receiving IL-6 blockade or the combination.

To assess whether IL-6 blockade could improve cancer treatment without increasing EAE severity, EAE was induced in C57/BL6 mice three days after B16.BL6 tumor inoculation. Mice treated with anti-CTLA-4 alone had accelerated signs of EAE, while the combination therapy resulted in both milder EAE symptoms and improved tumor control. This was also true when anti-CTLA-4 was replaced with anti-PD-1.

Finally, the researchers reviewed patient datasets for experience with IL-6 blockade in ICB-treated patients. Out of 13,735 patients with melanoma treated with ICB, 31 received IL-6 blockade treatment. These patients experienced a 74% reduction of irAEs after IL-6 blockade initiation. Among this small subset of patients, the best objective response rate to ICB in these patients was 57.7% before IL-6 blockade and 65.4% after treatment, indicating that blockade of IL-6 activity did not negatively impact ICB treatment.

Therefore, gene expression analysis in humans and mechanistic studies in mice identify IL-6 and Th17-like cells as key components in the development of irAE, and show that targeting IL-6 during ICB treatment may not only reduce the severity of irAEs, but improve tumor responses. This may be one strategy to decrease irAEs and increase the response rate to ICB therapy, and clinical studies are in progress to assess this combination.

Written by Maartje Wouters, image by Lauren Hitchings.

Meet the researcher

This week, first co-authors Hailemichael, Johnson, Abdel-Wahab, and senior author Diab answered our questions. 

What prompted you to tackle this research question?
Immunotherapy with checkpoint inhibitors (ICIs) is currently the best therapeutic strategy for treatment of multiple cancer types. However, durable remission rates with ICI monotherapy remain low, and therefore, combination strategies for ICIs are needed to overcome therapy resistance. Nevertheless, the high rate of immune-related adverse events (irAEs) limits the use of combination ICIs and can first and foremost increase the patient’s symptom burden and can potentially, although rarely, be fatal. In addition, these irAEs are associated with a high treatment discontinuation rate. For instance, more than 50% of patients with melanoma receiving combined ipilimumab plus nivolumab can develop high-grade toxicities, and more than a third of the patients require treatment discontinuation (CheckMate 067). Thus, we recognized that the identification of strategies to treat these toxicities without scarifying the antitumor immunity to ICI therapy represents a critical unmet need. It is well established that the interleukin-6 (IL-6)/T helper 17 (Th17) cell pathway plays a major role in pathogenesis of several autoimmune diseases that could manifest clinically within the constellation of irAEs. Therefore, we thought to study its role in ICI-related immune toxicity, and we then evaluated how IL-6 blockade can impact ICI tumor immunity.

What was the most surprising finding of this study for you?
Our multidisciplinary approach using clinical, preclinical, and translational analyses implicated a role for the IL-6/Th17 pathway in both ICI-related autoimmunity and resistance. Briefly, our data showed a higher expression of IL-6, as well as neutrophil and chemotactic markers in colitis vs. non-inflamed intestinal tissue, suggesting IL-6/Th17 pathway is involved in irAE development. We also found the genes upregulated in colitis were not upregulated in responding tumors, suggesting potential mechanistic differences between autoimmunity and tumor immunity. Furthermore, when comparing immune-related colitis in patients treated with anti-CTLA-4 based therapy vs. anti-PD-1 monotherapy, we found that CD4⁺ Th17 cells with memory phenotypes were enriched more in the anti-CTLA-4 group. In addition, these Th17 memory cells in both groups tended to have high expression of CTLA-4. In a mouse model of experimental autoimmune encephalomyelitis (EAE), we demonstrated that the ICI-induced exacerbation of toxicity could be mitigated by adding IL-6 blockade. Interestingly, these mice were also concomitantly transplanted with B16 melanoma or CT26 colon cancer tumors, and we found that the addition of IL-6 blockade enhanced the ICI-induced tumor shrinkage, suggesting that IL-6 inhibition can simultaneously mitigate autoimmunity and promote antitumor immunity. We then validated our findings by retrospectively reviewing all patients with melanoma who had received tocilizumab or sarilumab (IL-6 receptor-blocking antibodies) for treatment of irAEs at the University of Texas MD Anderson Cancer Center, and we found that the best overall response rate to ICI therapy did not differ before vs. after initiation of IL-6 blockade. Overall, our data suggested that IL-6 blockade could be an effective targeted therapy for irAEs without dampening the tumor response to ICI therapy. However, prospective clinical trials are needed to further validate these findings.

What was the coolest thing you’ve learned (about) recently outside of work?  
Man plots and God Laughs!!! The peak of pleasure is not only in achieving your own goals, but more likely in giving to others and helping them to achieve their goals in life!!!