Twenty-nine PwRMS participating in CARE-MS I (3) and II (4) trials in 6 European MS centers were enrolled and evaluated at baseline and for 72 months after alemtuzumab treatment. Inclusion and exclusion criteria were described in the original articles (3, 4). Patients were treated with 12 mg/d IV alemtuzumab in 2 annual courses (5 administrations at month 0 and 3 administrations at month 12). Patients’ demographic and clinical characteristics are reported in Table 1 . Neurologic assessments, performed by blinded investigators, were done at baseline and repeated every month or in case of relapses. Clinical data were collected in clinical research forms (CRF) and sent to the coordinating center located at the University of Torino. Blood samples were taken at baseline (before the first alemtuzumab course) and at months 6, 12 (before the second alemtuzumab course), 18, 24, 36, 48, 60, and 72. Fresh blood was collected in heparin-treated vacutainers and immediately sent to the coordinating center located at the University of Torino for immunologic testing. All samples were received and processed within 48 h from the blood withdrawal. Twelve sex- and age-matched healthy subjects were also enrolled from every center (rate cases/controls: 2:1).

The institutional review board of the participating centers approved the study, and all subjects gave written informed consent (protocol number Bio2009001).

We used the same methodology as the one employed in our previous study (7). Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation from heparinized venous blood. PBMC were stained for Treg cells with anti-CD4, anti-CD25, anti-CD127, anti-CD45RO, and anti-CD45RA monoclonal antibodies (mAb) (BioLegend, San Diego, CA) on the cell surface. For detection of the transcriptional factor FoxP3, cells were fixed with fixation and permeabilization buffers (eBioscience, San Diego, CA). PBMC were cultured in Iscove’s modified Dulbecco’s medium (BioWhittaker, Walkersville, MD) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA) and stimulated for 5 h with phorbol 12-myristate 13-acetate PMA (50 ng/ml) and ionomycin (500 ng/ml) in the presence of brefeldin A (10 mg/ml, Sigma-Aldrich, St. Louis, MO). Cells were first stained for the surface antigen CD4 (BioLegend) and then fixed with 4% paraformaldehyde, permeabilized with 0.5% saponin, followed by intracellular staining with anti-IL-17 and anti-IFN-γ mAbs (BioLegend) (10, 11). Stained PBMC were acquired on a FACSCalibur (BD Biosciences, San Jose, CA) and analyzed with FlowJo software (Ashland, OR). For detection of Treg cells, stained PBMC were first gated on CD4 and CD25. CD4+CD25^high T cells were analyzed for co-expression of FOXP3 and CD127low identifying CD4+CD25^highCD127^lowFOXP3+Tregs ( Supplementary Figure 1A ). Tregs were then analyzed for expression of CD45RA and CD45R0. For detection of Th17 and Th1 cells, stained PBMC were first gated on CD4 and then analyzed for IL-17 or IFN-γ production ( Supplementary Figure 1B ). Absolute values of Treg, Th17, and Th1 cells were calculated normalizing the percentage of CD25^highCD127^lowFOXP3+, IL-17-producing cells, and IFN-γ-producing cells on the CD4 T cell count obtained from complete blood count with formula and expressed as cells/μL of blood.

We used the same methodology as the one employed in our previous study (7). Aliquots (0.5 ml each) of blood were mixed with 1.3 ml RNAlater (Ambion, Life Technologies, Carlsbad, CA) immediately after arrival at the coordinating canter and stored at 280°C. To determine mRNA levels, samples were treated as previously described (12): RNA was extracted with the RiboPure Blood Kit (Ambion, Foster City, CA, USA) and cDNA obtained with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Life Technologies, Monza, Italy) (12).

We assessed the mRNA levels of the following 26 immunologically relevant molecules whose function in MS has been documented by using TaqMan low-density arrays (Applied Biosystems 7900HT Real-Time PCR System) (12): molecules with pro-inflammatory function including IL-1b, IL-2, IL-6, IL-12, IL-17A, IL-17F, IL-21, IL-22, IL-23, IL-26, IFN-γ, T-box expressed in T cells (Tbet), retinoid-related orphan receptor g (RORC), tumor necrosis factor-a (TNF-α), C–C chemokine receptor type 3 (CCR3), CCR4, CCR5, CCR6, C–X–C chemokine receptor type 3 (CXCR3), C–X–C motif ligand 10 (CXCL10), C–C motif ligand 20 (CCL20), and very late antigen 4 (VLA4), and molecules with anti-inflammatory function, including IL-10, IL-27, transforming growth factor–b (TGF-β), and forkhead box P3 (FoxP3). Glyceraldehyde-3 phosphate dehydrogenase served as the housekeeping gene for normalization. The relative expression of each gene was calculated using the comparative threshold cycle method as directed by the manufacturer (User Bulletin No. 2, Applied Biosystems, Foster City, CA, USA) and expressed in arbitrary units as previously described in detail (7).

Antigen-specific IFN-γ and IL-17-producing cells and antigen-specific suppressor activity of Treg cells were assessed by enzyme-linked immunospot (ELISPOT) (eBioscience) at months 0, 12, 24, 36, 48, 60, and 72. We used the same methodology as the one employed in our previous study (7). To remove Treg cells from PBMC (PBMC^CD25-), the CD25⁺ fraction was depleted using immunomagnetic beads (CD4⁺ CD25⁺ Regulatory T Cell Isolation Kit, Miltenyi Biotec, Bergisch Gladbach, Germany). The purity of the depleted fraction was immediately analyzed by fluorescence-activated cell sorting (FACS) staining and ranged from 92 to 95%. 1 × 10⁵ total PBMC or PBMC^CD25- was seeded in 96-well ELISPOT assay plates (Millipore, Darmstadt, Germany) in triplicate and incubated for 48 h at 37°C either with myelin basic protein (MBP, 40 mg/ml, Sigma-Aldrich), with the purified protein derivative of tuberculin (PPD, 40 mg/ml, Sigma-Aldrich) as negative control, or with anti-CD3 and anti-CD28 mAbs (10 and 1 μg/ml, respectively) as an internal test control. IFN-γ-specific or IL-17-specific spots were counted by computer-assisted image analysis (Transtec 1300 ELISpot Reader; AMI Bioline, Buttigliera Alta, Italy). The number of IFN-γ or IL-17 spots produced spontaneously was subtracted from the number of spots produced by antigen-stimulated cells to obtain the number of IFN-γ and IL-17 antigen-specific spots in the PBMC or in the PBMC^CD25-, as previously described (7). Median values for the triplicates, adjusted as the number of IL-17- and IFN-γ-producing cells/10⁶ PBMC, were used.

Statistical analysis was performed using GraphPad Prism 8.0 (La Jolla, CA) software. For the longitudinal follow-up, statistical significance was calculated concerning baseline and to HS by using one-way analysis of variance for repeated measures followed by the Bonferroni multiple-comparison post-test. The Pearson t-test was used to analyze the differences between groups. p values < 0.05 were considered statistically significant.

This study represents the continuation of the one previously published (7). Thirteen patients experienced a total of 16 clinical relapses (at months 1, 9, 10, 12, 20, 25, 28, 29, 30, 36, 41, 70) which brought 5 patients to receive additional courses of the drug. The median of Expanded Disease Status Scale (EDSS) score did not change significantly during the 72-month follow-up (EDSS median score from 2 at baseline to 1, 7 at month 72). Secondary autoimmune thyroiditis occurred in 9 patients at months 24, 30, 36, 42, and 59.

Of the 29 patients recruited in this study, only patients who had received the “classical” two courses of alemtuzumab were analyzed in this new study (24 patients). Two patients, described in a different paper (13), were excluded from the analysis due to their atypical CD4+ T population behavior after alemtuzumab administration: they showed persistent disease activity despite repeated alemtuzumab treatment and, while lymphocyte count decreased and fluctuated according to alemtuzumab administration, their CD4+ cell percentage was not affected or was barely affected and was slightly below the lowest normal limit prior to alemtuzumab (13). The other 3 patients had been excluded as they had received a third course of alemtuzumab at month 30 or 36 due to a relapse.

As previously observed, the percentage of CD4+ T cells in the PBMC rapidly decreased after the first administration (7) course and returned to the lowest normal limit (LNL) only at month 48 (13), and this was maintained over months 60 (40.78 ± 2.97%) and 72 (39.37 ± 2.40%) where the LNL was 34 ( Figure 1A ). Within the CD4+ cell fraction, we did not observe any significant change in the percentage of pro-inflammatory Th17 and Th1 cells (defined by the production of IL-17 and IFN-γ, respectively, in the CD4+ T cell fraction) at any time point of the follow-up, when compared to HS ( Figure 1B ). However, the absolute number of circulating Th17 had significantly decreased after alemtuzumab (i.e., month 0: 1,100 ± 610; month 24: 380 ± 300; month 72: 410 ± 270 cells/ml) compared to the number obtained in HS (270 ± 210 cells/ml; Figure 1C upper panel). The absolute number of circulating Th1 had significantly decreased after alemtuzumab until month 48, compared to baseline (6,341 ± 3,182 at month 0 vs. 3,422 ± 1,581 cells/ml at month 48), but all the values assessed during the follow-up were not different from those of HS (5,292 ± 3,378 cells/ml; Figure 1C lower panel). To better evaluate the immune response involved in RMS, we also evaluated the antigen-specific response directed against MBP. One should note that MBP-specific IL-17-producing cells and IFN-γ-producing cells significantly decreased after alemtuzumab administration and were comparable (not statistically different) to HS starting from months 36 (2.85 ± 3.79 vs. 0.27 ± 0.64 IL-17 spots) and 12 (21.50 ± 23.47 vs. 1.1 ± 2.03 IFN-γ spots), respectively ( Figure 1D ).

Accordingly, mRNA levels of pro-inflammatory cytokines produced by Th17 cells, IL-17A (14), IL-17F (14), IL-21 (15), IL-22 (11), IL-26 (16), and by Th1 cells, IFN-γ (17), and their related transcriptional factors, RORC (18) and Tbet (19), significantly declined to the values of HS at each time point of the follow-up when compared to baseline ( Table 2 ). A similar behavior was observed for their chemokines and chemokine receptors, CCL20 (20) and CXCL10 (21), CCR6 (22), CCR5, and CXCR3 (23), involved in T cell recruitment into the CNS ( Table 2 ). Alemtuzumab administration also reduced mRNA levels of cytokines known to guide Th17 [IL-1β (24), IL-6 (25), IL-23 (26)], and Th1 [IL-12 (27)] differentiation, but their levels remained higher compared to HS, suggesting that a pro-inflammatory environment could persist in these 24 patients.

As previously reported, a significant increase in Treg cell percentage (evaluated as CD4+ CD25^high CD127^low FoxP3+ cells) was detected at month 24 when compared to baseline (7). Afterward, the values of Treg cell percentage revert back to their basal levels which are similar to those observed in HS (i.e., month 36: 3.48 ± 1.48 in PwRMS vs. 2.55 ± 1.25% in HS; Figure 2A ). This behavior is reflected by the Treg cell absolute count in the blood that, after an initial decrease at month 12, then reaches levels similar to HS as of month 24. Up to month 36, the majority of Treg cells exhibited a memory phenotype as indicated by the significant increase in the percentage of memory CD45RO+FoxP3+CD4+lymphocytes, accompanied by a relative contraction of the percentage of naive CD45RA+ FoxP3+CD4+ cells ( Figure 2B ). mRNA levels of Treg transcription factor FoxP3 (28), and of anti-inflammatory cytokines related to Treg subset IL-10, TGFβ (29), and IL-27 (30), had increased throughout all time points of the follow-up compared both to baseline and to HS ( Table 2 ).

Treg suppressor function was assessed by measuring the MBP-specific IL-17 and IFN-γ production in the PBMC both before and after CD4+CD25+ depletion. This method allowed us to test the presence of functional Treg cells also in months with high lymphopenia, through the increase in IL-17 and IFN-γ spot production by PBMC^CD25-. Treg suppressor function, expressed as fold change over PBMC ( Figure 2C ), was restored at month 24 (as previously observed in 7) and persisted through to the end of the follow-up, even if their functions remain significantly lower compared to HS ( Figure 2C and Supplementary Figure 2 ). Furthermore, the increase in suppressive capacity of Treg cells was confirmed both by FoxP3 expression in Treg cells ( Figure 2D ) and by the increase of FoxP3 mRNA levels in the blood ( Table 2 ).

As thirteen PwRMS had one or more relapses during the follow-up, we wondered if some of the immunological parameters we had evaluated differed in “responders” (i.e., PwRMS that did not develop any relapses) versus “relapsing PwRMS” before starting alemtuzumab. Clinical data of these subgroups are reported in Table 3 . Interestingly, we observed a significantly higher percentage of Th17 cells (1.76 ± 0.77 vs. 0.93 ± 0.38%), but not of Th1, and a low percentage of Treg (2.40 ± 1.05 vs. 3.70 ± 0.82%) cells in “relapsing PwRMS” compared to “responders” at month 0 ( Figure 3A ). In the same way, the Th17/Treg cell ratio was highest in “relapsing PwRMS” ( Figure 3B ). The same analysis was performed on the patient who developed secondary autoimmunity; a significant increase in the mRNA levels of IL-21 was detected at baseline in the subject who developed thyroiditis compared with subjects who did not develop secondary autoimmunity ( Figure 3C ).