Restoration of regulatory B cell deficiency following alemtuzumab therapy in patients with relapsing multiple sclerosis

Abstract

Background: Regulatory B cells (Bregs), which protect from autoimmunity, are deficient in multiple sclerosis (MS). Novel regulatory B cell subsets CD19⁺CD24^hiCD38^hi cells and CD19⁺PD-L1^hi cells, with disparate regulatory mechanisms have been defined. Alemtuzumab provides a long-lasting suppression of disease activity in MS. In contrast to its documented efficacy, alemtuzumab's mechanism of action is not fully understood and information about the composition of repopulating B cell pool is scarce.

Aim: To characterize repopulated B cell subsets and elucidate alemtuzumab's mechanism of action in B cell perspective.

Methods: The frequency and the absolute number of Bregs were studied in peripheral blood mononuclear cells (PBMC) of 37 MS patients and 11 healthy controls (HC). Longitudinal analysis of the frequency and the absolute number of Bregs in PBMC of 11 MS patients was evaluated, before and at 6, 9, and 12 months post alemtuzumab.

Results: We found deficiency of CD19⁺CD24^hiCD38^hi cells during relapse compared to remission and HC (relapse vs remission: p = 0.0006, relapse vs HC: p = 0.0004). CD19⁺PD-L1^hi cells were deficient during relapse than remission and HC (relapse vs remission: p = 0.0113, relapse vs HC: p = 0.0007). Following alemtuzumab, the distribution of B cells shifts towards naïve phenotype and Breg deficiency is restored. The frequency of CD19⁺CD24^hiCD38^hi cells was significantly increased at 6 M and 9 M compared to 0 M (6 M vs 0 M: p = 0.0004, 9 M vs 0 M: p = 0.0079). At 9 M, the frequency of CD19⁺CD24^hiCD38^hi cells started to decrease and by 12 M the frequency was reduced compared to 6 M, although it was significantly higher than baseline level (12 M vs 0 M: p = 0.0257). The absolute number was significantly increased at 6 M and 9 M post-alemtuzumab (6 M vs 0 M: p = 0.0063, 9 M vs 0 M: p = 0.02). The frequency of CD19⁺PD-L1^hi cells significantly increased until 12 M (6 M vs 0 M: p = 0.0004, 12 M vs 0 M: p = 0.0036). The frequency of CD19⁺PD-L1^hi cells at 12 M was significantly higher than 9 M (p = 0.0311). We further pinpoint that CD19⁺CD24^hiCD38^hi cells were deficient at severe relapses following alemtuzumab infusion and restored during recovery.

Conclusions: Our results highlight the preferential reconstitution of Bregs as a possible mechanism of action of alemtuzumab and CD19⁺CD24^hiCD38^hi cells as a potential biomarker for disease activity.

Keywords: Alemtuzumab; Multiple sclerosis; Regulatory B cells; Relapse.

Fig. 1

MS patients show deficiency of CD19⁺CD24^hiCD38^hi cells during relapse. The percentage and the absolute number of CD19⁺CD24^hiCD38^hi B cells were measured in MS patients undergoing relapse (n = 15) and patients in remission (n = 15) and healthy controls (n = 8). PBMC was isolated from fresh peripheral blood and surface stained for flow cytometry. a Representative flow cytometry dot plot of CD24 and CD38 expression in total CD19+ B cells. b Scatter plots showing the percentage of CD19⁺CD24^hiCD38^hi B cells in MS-relapse, MS-remission, and HC. A significant reduction in the frequency of CD19⁺CD24^hiCD38^hi B cells was observed in MS-relapse compared to MS-remission and HC (relapse vs remission: p = 0.0006, relapse vs healthy: p = 0.0004). All values show mean ± SEM. Data were analyzed by one-way analysis of variance (ANOVA) with Tukey’s multiple comparison post hoc analysis. ***p < 0.001. c The absolute number of CD19⁺CD24^hiCD38^hi B cells in MS-relapse was significantly reduced compared to MS-remission (p = 0.009). All values show mean ± SEM. Data were analyzed by unpaired t test. **p < 0.01. Ex vivo data were collected from peripheral blood samples taken during the time course of this study

Fig. 2

MS patients have deficiency of CD19⁺PD-L1^hi cells during relapse. The percentage and the absolute number of CD19⁺PD-L1^hi cells were measured in MS patients undergoing relapse (n = 20) and patients in remission (n = 17) and healthy controls (n = 11). Fresh PBMC was isolated from peripheral blood and surface stained for flow cytometry. a Representative flow-cytometry dot plot of PD-L1 and CD19 expression in total CD19+ B cells. b Scatter plots showing the percentage of CD19⁺PD-L1^hi cells in MS-relapse, MS-remission, and HC. The frequency of CD19⁺PD-L1^hi cells was significantly reduced in MS-relapse compared to MS-remission and HC (relapse vs remission: p = 0.0113, relapse vs healthy: p = 0.0007). All values show mean ± SEM. Data were analyzed by one-way analysis of variance (ANOVA) with Tukey’s multiple comparison post hoc analysis. *p < 0.05; ***p < 0.001. c Scatter plots showing the absolute number of CD19⁺PD-L1^hi cells in MS-relapse and MS-remission. All values show mean ± SEM. Data were analyzed by unpaired t test. Ex vivo data were collected from peripheral blood samples taken during the time course of this study

Fig. 3

Naive B cells predominate repopulated CD19+ cells following alemtuzumab treatment. In order to evaluate the B cell subset distribution post-alemtuzuamb, thawed PBMCs of alemtuzumab-treated patients (n = 11) were evaluated up to 12 months after induction. a Cumulative data for the frequency and the absolute number of total lymphocytes and CD19⁺ B cells. Successful depletion and reconstitution of lymphocytes and CD19⁺ B cells was confirmed. b Cumulative data for the frequency and the absolute number of CD19⁺CD27⁺ memory B cells, CD19⁺CD27⁻ naïve B cells, and CD19⁺CD27⁺CD38^hi plasmablasts. Following alemtuzumab, significant reduction in the frequency of memory B cells (6 M vs 0 M: p = 0.0278) and plasmablasts (6 M vs 0 M: p = 0.0448) was observed and dominance of naïve B cells was observed (6 M vs 0 M: p = 0.0269). The absolute number of memory B cells was significantly decreased compared to 0 M (6 M vs 0 M: p = 0.0112). All values show mean ± SEM. Data were analyzed by repeated measures ANOVA with Tukey’s multiple comparison post hoc analysis

Fig. 4

Alemtuzumab treatment restores CD19⁺CD24^hiCD38^hi cells. In order to evaluate the B cell subset distribution post-alemtuzuamb, thawed PBMCs of alemtuzumab-treated patients (n = 11) were evaluated up to 12 months after induction. a Representative flow-cytometry dot plot of CD24 and CD38 in total CD19⁺ B cells. b. Cumulative data for the frequency of CD19⁺CD24^hiCD38^hi B cells. The frequency of CD19⁺CD24^hiCD38^hi cells were significantly increased at 6 M and 9 M compared to pre-treatment level (6 M vs 0 M: p = 0.0004, 9 M vs 0 M: p = 0.0079). At 9 M, the frequency of CD19⁺CD24^hiCD38^hi cells started to decrease and by 12 M, the frequency was reduced compared to 6 M, although it was significantly increased than baseline level (12 M vs 0 M: p = 0.0257). c The absolute number was significantly increased at 6 M and 9 M post-alemtuzumab (6 M vs 0 M: p = 0.0063, 9 M vs 0 M: p = 0.02). All values show mean ± SEM. Data were analyzed by repeated measures ANOVA with Tukey’s multiple comparison post-hoc analysis. *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 5

Alemtuzumab treatment restores CD19⁺PD-L1^hi cells. In order to evaluate the B cell subset distribution post-alemtuzuamb, thawed PBMCs of alemtuzumab-treated patients (n = 11) were evaluated up to 12 months after induction. a Representative flow-cytometry dot plot of PD-L1 in total CD19⁺ B cells. b Cumulative data for the frequency of CD19⁺PD-L1^hi cells. The frequency of CD19⁺PD-L1^hi cells increased significantly until 12 M (6 M vs 0 M: p = 0.0004, 12 M vs 0 M: p = 0.0036). c Cumulative data for the absolute number of CD19⁺PD-L1^hi cells. No significant difference was observed in the absolute number of CD19⁺PD-L1^hi, CD19⁺PD-L1^int, and CD19⁺PD-L1^low cells. All values show mean ± SEM. All values show mean ± SEM. Data were analyzed by repeated measures ANOVA with Tukey’s multiple comparison post hoc analysis.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Fig. 6

The frequency of CD19⁺CD24^hiCD38^hi cells is decreased during relapse post alemtuzumab treatment. a Flow-cytometry plot of expression of CD24 and CD38 in total CD19⁺ B cells. b Graph showing the clinical course. The frequency and the absolute number of CD19⁺CD24^hiCD38^hi B cells and infusion of 2 cycles of alemtuzumab are shown. Filled circles indicate the frequency of CD19⁺CD24^hiCD38^hi B cells, and hollow circles indicate the absolute number of CD19⁺CD24^hiCD38^hi B cells. Red arrow indicates relapse