Regulatory T cell frequency, but not plasma IL-33 levels, represents potential immunological biomarker to predict clinical response to intravenous immunoglobulin therapy

The overall rate of new incidence of GBS is 0.6–4.0/year/100,000 people. We limited our investigation only to newly diagnosed and treatment-naïve GBS patients to avoid possible modulatory effects of past IVIG therapy or other therapeutic agents on the T_reg cells and other immune parameters. This explicates the difficulties in enrolling treatment-naïve patients for the investigation. As kinetics of IL-33 induction in the plasma of patients with autoimmune diseases following IVIG therapy was not explored earlier, we first analyzed plasma levels of IL-33 in ten GBS patients at various time points (pre-IVIG and weeks 1, 2, and ≥4 after initiation of IVIG therapy). Patients had variable pre-IVIG plasma IL-33 levels that significantly increased at week 1 in all the patients. This data thus confirms our recent observation in IVIG-treated inflammatory myopathy and anti-neutrophil cytoplasmic antibody-associated vasculitis patients [30]. At subsequent time points, IL-33 levels in the plasma of IVIG-treated GBS patients declined progressively but remain significantly higher than the pre-IVIG levels (Fig. 1a) indicating sustained effect of IVIG beyond its half-life on plasma IL-33 levels.

In order to explore the role of IL-33 in IVIG-mediated therapeutic effects, we analyzed whether IL-33 induction correlates with clinical response to IVIG therapy and T_reg cell expansion [29]. Of note, the fold changes in plasma IL-33 level at week 1 showed no correlation with the MRC scores at weeks 1, 2, and ≥4 post-IVIG therapy (Fig. 1b). Also, the r values are consistently negative at all the time points. On the other hand, T_reg cell response following IVIG therapy appears to be linked with clinical recovery from GBS and is consistent with the previous observations on essential role of T_reg cells in IVIG-induced protection in mouse models [18, 31]. Yet, we observed no correlation between the fold-changes in T_reg cells at week 1 and clinical score at weeks 1, 2, and ≥4 (Fig. 1c). These results suggest that T_reg cells, but not IL-33, might predict clinical response to IVIG therapy.

Similar observations are also made with MRS parameters. We found that the fold changes in IL-33 level at week 1 displayed no correlation with MRS scores at weeks 1, 2, and ≥4 following initiation of IVIG therapy indicating that clinical improvement as analyzed by MRS is also not associated with changes in IL-33 levels in the blood. As in the case of MRC scores, changes in T_reg cells were suggestive of clinical response to IVIG therapy by MRS parameters as well. However, the fold changes in T_reg cells at week 1 and MRS scores at weeks 1, 2, and ≥4 post-IVIG therapy did not correlate (Table 2).

As data from the recent mouse study suggested that T_reg cell expansion by IVIG is dependent on IL-33 [20], we analyzed correlation between IL-33 and T_reg cells in these patients at various time points. Contrary to Fiebiger et al., we found no consistent association between the fold changes in IL-33 level at week 1 and T_reg cell frequency at weeks 1, 2, and ≥4 after initiation of IVIG therapy (Fig. 2a), implying that T_reg cell expansion by IVIG in patients with autoimmune diseases is not related to levels of IL-33 in the plasma.

Expansion of T_reg cells in the periphery is mediated by the signals derived from professional antigen presenting cells such as DCs. Unlike other cytokines, IL-33 can be released into the microenvironment only upon injury to the cells and acts as alarmin to signal tissue damage to the immune system [32]. Our recent report shows that DC-SIGN-positive human innate cells derived either from the peripheral blood or from the spleen do not release IL-33 upon IVIG exposure [30]. Therefore, unlike murine models [19, 20], it is likely that damaged non-immune cells like endothelial cells or epithelial cells might have contributed to increased levels of IL-33 observed in the plasma of patients following IVIG therapy. Despite the lack of IL-33 induction, “IVIG-educated” DC-SIGN-positive human DCs induced T_reg cell expansion, a process mediated via cyclooxygenase-2-dependent PGE₂ and independent of Fc-fragments of IVIG [15]. Notably, there was a significant increase in the plasma levels of PGE₂ in IVIG-treated GBS patients [33]. In the current study, although plasma PGE₂ levels are enhanced in IVIG-treated patients, we found no correlation between the fold changes in PGE₂ level at week 1 and T_reg cell frequencies at week 1, 2, and ≥4 post-IVIG therapy (Fig. 2b). Similarly, fold changes in PGE₂ level at week 1 did not correlate with MRC score at weeks 1, 2, and ≥4 post-IVIG therapy (Fig. 3). The reason for the non-significant correlation of values might rests in the low number of patients. A recent report shows that plasma PGE₂ levels correlate with the prevention of IVIG resistance in Kawasaki disease [34]. Taken together, these results are suggestive of a role for PGE₂-mediated expansion of T_reg cells in the clinical recovery of patients following IVIG therapy.

In the previous report [30], we analyzed IL-33 in the plasma of patients with rheumatic disorders as early as 48–72 h following initiation of IVIG therapy. With the exception of two, the remaining 14 patients had IL-33 levels below 500 pg/ml of plasma following IVIG therapy. Although the underlying pathologies in that report and current study are different, we found that at 1 week following initiation of IVIG therapy, IL-33 level in the plasma was more than 1000 pg/ml in nine GBS patients. Furthermore, in majority of the patients (seven), IVIG was given daily for five consecutive days. Therefore, lack of consistent association between the fold changes in IL-33 level and T_reg cell frequency following IVIG therapy might not be due to loss of IL-33 subsequent to proteolytic cleavage and binding to ST2 receptor. Also, IL-33 levels observed at first week are maintained in majority of the patients even after 4 weeks post-IVIG. Whether polymorphisms in either ST2 receptor, IL-1RL1, and/or IL-33 itself, skew the plasma kinetics and/or downstream effects of IL-33 remains to be determined [35, 36].

Current literature suggests that IVIG can induce T_reg cell expansion by several mutually non-exclusive mechanisms [9, 14, 16, 17, 31] that may vary among the pathologies and even among the patients with same underlying pathology. This might explain the lack of significant correlation between PGE₂ and T_reg cells. As we restricted our study only to treatment-naïve GBS patients, difficulty to get large number of patients provides an alternative explanation for the lack of significant positive correlation between changes in the IL-33 and PGE₂, and T_reg cells. Role of IL-33 in the microenvironment towards DC-mediated human Treg expansion requires further exploration.

In addition to expansion of T_reg cells, IVIG is also known to enhance their functions and suppress effector T cell proliferation and production of cytokines such as IFN-γ and TNF-α [16, 31]. Interestingly, in IVIG-treated GBS patients, there was a significant negative correlation between the fold changes in T_reg cells at week 1 and T cells expressing TNF-α at week 1 post-IVIG therapy. However, such consistent association was not observed with T cells expressing IFN-γ (Fig. 4a, b). Thus, T_reg cell-mediated clinical response to IVIG therapy is associated with reciprocal regulation of effector T cells in GBS patients.

Because of its role as promoter of type 2 immune responses and regulator of innate and adaptive immune responses, IL-33 has been explored as therapeutic option in pre-clinical models of Alzheimer’s disease [37], stroke [38], cerebral malaria [39], and transplantation [40]. In addition, IL-33 was also reported to be essential to attenuate viral-induced encephalitis [41]. In line with these reports, enhanced expression of IL-33 in mice was suggested to mediate T_reg cell expansion and protection by IVIG [19, 20, 27]. Translation of these results to human is complicated and requires thorough investigation. Indeed, we found that up-regulation of plasma IL-33 does not correlate with the clinical response to IVIG therapy and expansion of T_reg cells in GBS patients. T_reg cells on the other hand are negatively correlated to effector T cells expressing TNF-α. Thus, the mechanism of IL-33 induction by IVIG and its role in T_reg cell expansion observed in mouse models might not be applicable to human. Small size of the patients’ cohort represents limitation of this study. Nevertheless, based on these findings, T_reg cell frequency, but not plasma IL-33 levels, represents potential immunological biomarker to predict clinical response to IVIG therapy.