In this study, we demonstrated that upregulation of the eosinophil count ≧ 3.0% two weeks after treatment might be a predictive biomarker for the onset of irAEs of grade ≧ 2 in patients with RCC treated with ipilimumab and nivolumab. Recent comprehensive studies on onset to irAE induced by ipilimumab and nivolumab have shown that the median-time onset of all-grade irAEs ranged from 2.4 to 13.9 weeks (Tang et al., 2021), indicating the need to predict irAEs before two course of treatment with ipilimumab and nivolumab. Although several reports have suggested an association of blood cell count parameters with the prediction of irAEs in patients under ICI therapy (Hommes, Verheijden, Suijkerbuijk, & Hamann, 2020), no reports exist on effective biomarkers, especially in RCC treated with ipilimumab and nivolumab. Eosinophils in two weeks after treatment might be one of the key determinants of irAEs onset of grade ≧ 2 in patients with RCC and who underwent ipilimumab and nivolumab therapy, as reported for other, different types of cancers. Our approach using eosinophils as an effective prediction biomarker might help to diagnose and treat irAEs.
Eosinophils have a homeostatic role in the immune response. They combat several parasitic, bacterial, and viral infections by interacting with B cells, T cells, macrophages, and neutrophils. Eosinophils are further involved in the regulation of several diseases, including allergic asthma, esophagitis, myopathies, and autoimmune disorders due to directly or indirectly promoting tissue damage or altering the local immune status (Kita, 2011; Rosenberg, Dyer, & Foster, 2013; Wechsler et al., 2021). A recent study showed that upregulation of the eosinophil count in the peripheral blood is linked with a better prognosis and response to ICI therapy. In patients with recurrent or metastatic head and neck squamous cell carcinoma treated with nivolumab, a higher eosinophil count was associated with better survival (Minohara et al., 2021; Nishikawa et al., 2021; Nishikawa et al., 2018). The eosinophil count positively correlated with overall survival in patients with melanoma treated with ipilimumab (Martens et al., 2016). Indeed, eosinophils induced by ICI play an important role in the elimination of tumors. Mechanistically, functional experiments showed that anti-cytotoxic T-lymphocyte–associated protein 4 increased eosinophil infiltration into tumors. Activated tumor-infiltrating eosinophils produced chemokines and recruited CD4 + T cells and CD8 + T cells, which resulted in tumor elimination (Carretero et al., 2015; Jia-Nan Cheng & Zhihua Gong, 2021; Zheng et al., 2020). Furthermore, eosinophils associate with the onset of irAE induced by ICI. A recent case report described how specific tissue-infiltrating eosinophils occurred as an adverse effect, such as in eosinophilic fasciitis and eosinophilic pneumonia triggered by ICI (Chan et al., 2019; Jodai et al., 2019). Therefore, several studies have focused on the eosinophil count as an effective biomarker of the onset of irAEs (Adam Diehl, 2017; Chu et al., 2020; Nakamura et al., 2019). Our data showed that the upregulation of eosinophils is associated with the clinical outcome and onset of irAEs, as reported in different types of cancers. These results collectively indicated that the upregulation of eosinophils in peripheral blood reflects the efficacy and toxicity of ICI by tumor and tissue-infiltrating eosinophils.
Previous studies on eosinophils as a biomarker to predict irAEs have focused on levels before treatment or one month after treatment. For example, a high eosinophil count before or one month after nivolumab or pembrolizumab monotherapy increased the risk of occurrence of irAEs by 1.3 times in solid tumors (Adam Diehl, 2017). Additionally, upregulation of eosinophils before or one month after nivolumab or pembrolizumab monotherapy may be an effective biomarker to predict endocrine irAEs in patients with melanoma (Nakamura et al., 2019). Interestingly, our data revealed that patients in the irAE group who experienced endocrine, gastrointestinal, and skin disorders had a significantly higher eosinophil count in two-week samples than patients in the non-irAE group (P < 0.05 for all; Supplementary Fig. S3b–d). Patients in the irAE group who experienced pulmonary disease tended to show a higher eosinophil count than those in the non-irAE group (P = 0.06; Supplementary Fig. S3e), but not for other disorders (P = 0.42; Supplementary Fig. S3f). Furthermore, the eosinophil count in two-week samples of patients in the irAE group was significantly upregulated at the onset of irAE, not only during early courses (one or two courses) but also during late courses (three or four courses) compared to that of patients in the non-irAE group (P < 0.05; Supplementary Fig. S3g and h), indicating that the eosinophil level two weeks after treatment was upregulated by the onset of any type of irAE, excluding other diseases, regardless of when the adverse event occurred. These results suggest the eosinophil count is regulated by different ICIs and types of cancers. Our data also suggest the necessity of examining the eosinophil count during one course of ipilimumab and nivolumab to predict the onset of irAEs.
In the past, most studies had shown ICI-mediated irAEs might be associated with an improved response to therapy and survival outcome in several types of cancers (Freeman-Keller et al., 2016; Y. Wang et al., 2018). However, especially in RCC, no consensus of evidence exists in the association between the onset of irAEs and clinical outcome. The CheckMate 214 trial showed that OS between patients with and without irAE did not reveal a significant difference in RCC after ipilimumab and nivolumab therapy (R. J. Motzer et al., 2020). In contrast, Ikeda et al. revealed that the onset of irAEs was significantly associated with an improvement of clinical outcome in RCC treated with ipilimumab and nivolumab (Ikeda et al., 2021). Our data revealed the significantly longer mOS of patients showing irAEs, along with the upregulation of eosinophils in peripheral blood, than that in patients without irAEs, as reported in different types of cancers. These controversial results might be reflected in different patient characteristics.
Interestingly, we found that having an irAE of grade ≧ 2 was associated with a poor clinical outcome, compared with having an irAE of grade 1 (mPFS; 16.7 months and 29.0 months, respectively; Fig. 1d). Consistently, in a study of 42-month results of the CheckMate 214 Trial, treatment-free survival of patients who experienced an irAE of grade ≧ 3 tended to be shorter than that of patients who did not experience an irAE of grade ≧ 3 (0.6 months and 6.1 months, respectively; (Regan et al., 2021)). Several studies have shown that the discontinuation of treatment by the onset of irAEs was associated with a poor clinical outcome (Naqash et al., 2020; Russano et al., 2021). Whereas mPFS of grade ≧ 2 irAE was shorter than that of Grade 1 irAE, mOS was not apparently different between Grade 1 and grade ≧ 2 irAEs. These results indicated that the follow-up period of OS may not be sufficient. Although further investigation is required, treatment discontinuation due to the onset of grade ≧ 2 irAEs might be related to a poor clinical outcome. Therefore, we need to predict and suppress the onset of irAEs, particularly irAEs of grade ≧ 2, to improve clinical outcomes. Our observations indicated that the prediction of grade ≧ 2 irAEs using the eosinophil level as an effective biomarker could further improve clinical outcome and prognosis due to treatment continuation.
The present study had several limitations. Only a small number of participants were involved in the study. In addition, because of the retrospective nature of the study, we could not control biases in the selection of patients. An interventional prospective study will be required to confirm our data.
In conclusion, our data provides a novel rationale for measuring the blood eosinophil level, with an elevated eosinophil count likely to be an effective biomarker for predicting grade ≧ 2 irAE onset in patients with RCC under ipilimumab and nivolumab therapy.