While autologous haematopoietic stem cell transplantation (AHSCT) is a successful treatment option for certain haematological malignancies (1–3), including non-Hodgkin’s lymphoma (NHL) (4, 5), it has also been demonstrated to be an effective treatment for autoimmune diseases, such as multiple sclerosis (MS) when compared to other disease modifying therapies (6). A central hypothesis for the observed disease remission is that AHSCT ‘resets’ the patient’s presumably autoreactive immune system to a more tolerogenic state through conditioning chemotherapy and infusion of ‘rescue’ autologous haematopoietic stem cells (7–13). It has been proposed that resetting of the immune balance and remission from CNS inflammation occurs through the removal of proinflammatory cells and the regeneration of immunoregulatory cells that are important in maintaining peripheral tolerance (9). These cells are known to exert a myriad of immunoregulatory mechanisms, including the suppression of proinflammatory cytokine secretion (14–16).

Cytokines are pleiotropic extracellular proteins secreted by immune, endothelial, and stromal cells which predominantly function to coordinate immune responses. Cytokines can be categorised into several classes depending on their functions. For instance, proinflammatory cytokines are those that promote immune responses, such as those involved in T helper (Th) 17-related immune responses, including interleukin (IL)-17 (17), IL-23, IL-21, IL-6, IL-8 (18) and IL-1β (19), and those involved in Th1-related immunity, such as TNFα (20), IFNγ (21) and IL-12p70 (22). In contrast, certain cytokines suppress immune responses, such as IL-10 (23). In the Th1/Th2 differentiation paradigm, IL-4 is known to induce Th2 differentiation, promoting humoral immune responses (24) and suppresses Th1-related responses (25). Because Th1 responses are strongly correlated to MS pathogenesis (26, 27) and Th2 responses are associated with the dampening of MS activity (28), IL-4 is considered an anti-inflammatory cytokine in MS. Additionally, IL-2 is involved in the proliferation and homeostasis of T cells (29) and can also be considered anti-inflammatory due to its important role in the differentiation of immunoregulatory T regulatory cells (Tregs) (30–32).

In MS pathogenesis, the dysregulation of cytokines has been reported to correlate with disease activity (26, 33, 34). A meta-analysis evaluating Th17 cell status and Th17 related cytokines in the peripheral blood of MS patients revealed that the proportion of Th17 cells, and the levels of IL-17 and IL-23 were increased compared to control subjects (34), strongly suggesting that Th17-related responses contributes to MS pathogenesis. Indeed, IL-17-producing mucosal associated invariant T (MAIT) cells were detected around actively demyelinating lesions in post-mortem brain sections of MS patients (35), supporting this hypothesis. Additionally, Th1-related cytokines are also known to contribute to MS pathogenesis (36). In regards to anti-inflammatory cytokine profiles, the majority of studies report significantly lower levels of IL-10-producing regulatory cells in the circulation of relapsing-remitting MS patients (37, 38), whilst one study analysing progressive MS patients reported elevated IL-10 serum levels, along with higher levels of TNFα and IFNγ (39). Altogether these studies denote a dysregulated coordination of the immune system in MS patients.

Thus far, only two studies have investigated the evolution of the cytokine milieu in the blood of MS patients following AHSCT (40, 41), with both employing a lymphoablative cyclophosphamide chemotherapy regimen. Wiberg and colleagues (40) demonstrated significant but transient changes to cytokines predominantly associated with innate immunity following AHSCT that returned to baseline levels by the 3 month timepoint. Jaime-Perez and colleagues (41) reported significant reductions in proinflammatory cytokines IL-21 and IL-22, and increased levels of anti-inflammatory cytokines CCL2 and CCL4 at 14 days following transplantation without longer term follow-up. Overall, these studies give little indication of the influence of AHSCT on adaptive immunity. Furthermore, these studies failed to address the degree to which the cytokine shifts relate to early chemotherapy-induced lymphopenia as opposed to a disease specific consequence of immune reconstitution, which likely occur at later timepoints post-transplantation.

The BEAM conditioning regimen used in AHSCT for MS patients is also widely used in AHSCT for NHL, thus making NHL patients an ideal comparator group to investigate cytokine changes post-AHSCT. As NHL is characterised by a defect in the early development of lymphocytes (42), aberrant cytokine expression is also evident in this disease. Abnormal levels of several cytokines in the patient blood have been associated with NHL pathology, including IL-2, IL-8, IL-10, IFNγ and TNF (43, 44). However, the most consistent finding is significantly elevated IL-6 in NHL patients compared to healthy controls (44–46). In one study, gene expression of IL-10, TNFα, and IL-8 was significantly increased in the blood of NHL patients at day 15-30 post-AHSCT (47), supporting the hypothesis that AHSCT induces cytokine changes that may alter the immune system. Other studies in the allogeneic setting have also reported changes to the cytokine milieu of NHL patients following stem cell transplantation (48–52).

Here, we sought to evaluate durable changes in the peripheral blood cytokine profile of MS patients following AHSCT with a myeloablative conditioning chemotherapy, BEAM + Horse ATG (20mg/kg/day). Globally, this is the second-most used conditioning regimen in AHSCT for MS. Given MS pathogenesis associates with an adaptive immune response, we focus specifically on pathogenic adaptive immune-related inflammatory cytokines. We hypothesise that proinflammatory cytokines will be suppressed and anti-inflammatory cytokines will be increased post-AHSCT, reflecting the tolerogenic shift as indicated by immune reconstitution data (8, 35, 53). To further determine the degree to which AHSCT re-establishes immune tolerance, a comparator arm of NHL patients has been recruited. This cohort exhibit different immunological abnormalities to MS and received AHSCT as treatment, albeit with a slightly different chemotherapy regimen (BEAM only) (53). Inclusion of the NHL comparator arm enables differentiation between changes induced by chemotherapy versus immune reconstitution in the autoimmune setting, which has not been previously reported but is critical for understanding the mechanisms underlying disease remission post-AHSCT.

This is the first study to report sustained and profound changes to T cell-related proinflammatory cytokines in MS patients up to 24 months post-AHSCT. It is also the first to report these cytokine changes in comparison to a non-autoimmune disease cohort undergoing similar therapy. Few studies have examined the evolution of the cytokine milieu in patients with autoimmune diseases post-AHSCT (40, 41, 61, 62). These studies described significant reductions in the cytokine levels in the blood of both systemic sclerosis (62) and MS (40, 41) patients early post-AHSCT. However, these changes were transient and the overall cytokine concentrations were reported to return to baseline levels by 3-12 months post-AHSCT. Here, we have shown that i) proinflammatory Th17 and Th1 cytokines are elevated in MS patients pre-AHSCT compared to healthy controls; ii) significant and sustained reductions in proinflammatory Th17 and Th1 cytokines up to 24 months post-AHSCT that are specific to MS patients and not NHL patients; and iii) IL-17 was significantly elevated at 12 months post-AHSCT in the relapsed patients compared to non-relapse patients for MS.

Consistent with previous findings (33, 63, 64), the proinflammatory cytokines IL-17 and IFNγ were significantly elevated in the blood of MS patients pre-AHSCT when compared to healthy controls. Other cytokines investigated here, including the anti-inflammatory cytokines, were not significantly different between pre-AHSCT MS patients and healthy controls. A significant proportion of these cytokines were significantly reduced in NHL patients at pre-AHSCT compared to healthy controls and MS patients. These reductions are likely attributed to rituximab pre-treatment within the NHL cohort, which has previously been shown to significantly reduce cytokine levels in patients (65), and was not used in NHL patients. Indeed, only one MS patient in our cohort was pre-treated with rituximab ( Supplementary Table 1 ), and their cytokine concentrations were comparable to the NHL cohort (data not shown), which supports this hypothesis.

Following AHSCT, both Th17-related cytokines (IL-17, IL-1β, IL-21 and IL-23) and Th1-related cytokines (IFNγ and IL-12p70) were significantly reduced in MS patients, and this change was sustained up to 24 months post-AHSCT. In contrast, these proinflammatory cytokines show no-significant changes in NHL patients post-AHSCT. Considering the low starting levels of these cytokines in NHL patients at pre-AHSCT, the post-transplant evolution of these cytokines was generally similar across both disease cohorts. These findings highlight the immunosuppressive action of the conditioning regimen in AHSCT, which was proposed as an important aspect in inducing disease remission for both lymphomas and autoimmune diseases (7). Hence, the suppression of proinflammatory cytokines may simply reflect the ablation of immune cell populations following the conditioning regimen. However, in our MS cohort, total lymphocyte recovery was reached by 12 months post-AHSCT (53), whilst proinflammatory cytokines remain low for a further 12 months after this. This finding supports the hypothesis that AHSCT “recalibrates” the dysregulated immune system in MS patients beyond simple induction of lymphopenia. Long-term suppression of proinflammatory cytokines post-AHSCT may not be observed in previous reports of AHSCT in MS because the patients in these studies received non-myeloablative conditioning (40, 41), unlike this study, which used a higher intensity myeloablative BEAM conditioning.

One point of difference between the MS and NHL group was that ATG was used in the conditioning regimen of the MS group but not in the NHL group (53). ATG is a potent T cell suppressive agent that has been demonstrated to significantly reduce the gene expression of several cytokines (66). This may explain the increase in IL-6 observed at day 8 post-AHSCT in NHL patients that was not observed in MS patients. It is also noteworthy that the NHL patients were older than both the MS cohort and the healthy controls as previously highlighted (53) and this may therefore impact post-AHSCT cytokine evolution.

A few proinflammatory cytokines (IL-8, IL-6 and TNFα) were comparable between healthy controls and both disease cohorts prior to AHSCT. Although the long-term profiles of these cytokines were essentially unchanged or minimally changed from baseline, they appear to play a more important role in the early reconstitution process. All three cytokines increased in the early post-transplant period in NHL treated patients, and IL-8 and TNFα, which similarly increased early post-AHSCT in both NHL and MS patients. IL-8 is a proinflammatory cytokine associated with acute phase immune responses that is produced by innate immune cells when stimulated by TNFα. TNFα is a macrophage derived cytokine that is well established to play an important role in the control of viral infection (67), which are common post-transplantation (68, 69). Hence, the significant surges in IL-8 and TNFα a few days post-transplantation may be a generalised immune response prophylactic against viral reactivation in the lymphopenic environment that may occur following AHSCT regardless of the disease.

In general, we found that anti-inflammatory cytokines were similar between MS and healthy controls at pre-AHSCT in contrast to previous reports (38, 39). This may be due to the high proportion of patients in our cohort receiving Natalizumab [as previously reported (53)], which is known to alter immunosuppressive Treg populations in MS (70). These anti-inflammatory cytokines were generally reduced following transplant except for IL-10 at day 14 in MS patients. Because anti-inflammatory cytokines are often secreted as a negative feedback response towards inflammation (23), and can contradictorily act as markers of inflammation themselves (71), their significant reduction post-AHSCT may be an indication that between 3, 6, 12, and 24 months post-AHSCT, the inflammatory responses induced by Th1, Th17 and/or acute phase inflammatory cytokines are less active than they were before transplantation in MS patients. Additionally, the significant reduction of IL-2 early after transplant suggests that previously reported Treg reconstitution (53, 59, 72–75) may not be reliant on this cytokine (76), and may instead occur as a result of lymphopenia-induced homeostatic proliferation (10). These results suggest that MS pathogenesis relates to an over-represented proinflammatory milieu, rather than a deficiency of anti-inflammatory cytokines.

Importantly, IL-17 was significantly increased at 12 months post-AHSCT in the relapse cohort compared to non-relapse patients. Overall, relapse patients did not show significant suppression of the proinflammatory Th17 cytokines IL-17, IL-23, IL-21 and IL-1β post-AHSCT, unlike the non-relapse cohort. Interestingly, the occurrence of clinical relapses ranged from 11-13 months after transplant, approximately when IL-17 was shown to be elevated, to later at 22-23 months post-AHSCT. These results suggest that failure to suppress Th17-mediated responses within the first 12 months can induce immediate to long-term clinical relapses in MS patients following AHSCT. Although not statistically significant, Th17 cytokines consistently display a transiently increasing trend at day 14 post-AHSCT in relapsed patients ( Figure 3 ). However, no conclusions could currently be drawn from this result due to a low sample size at this timepoint within the relapse cohort. Further studies would be required to confirm these the mechanisms behind these findings. For instance, suppression of pathogenic T cells in MS patients may be affected by varying rates of ATG clearance by patients (77). It may be possible to further explore the relationship between post-AHSCT clinical relapses and blood ATG clearance flow cytometrically (78), in future AHSCT MS studies.

The limitations of this study are the relatively small numbers of patient cohorts and controls, and the lack of ATG and use of more intensive immune chemotherapy in the control group. Additionally, this study did not include data for B/T cell immune reconstitution following AHSCT. Nevertheless, this study represents the first significant analysis of cytokine expression in MS patients undergoing AHSCT using BEAM/ATG.

In summary, we demonstrate that AHSCT significantly suppresses Th1- and Th17-related inflammatory cytokines in MS patients for up to 24 months. Additionally, our results support the importance of suppressing these Th17-proinflammatory responses to achieve long-term remission in MS patients but this requires validation with a larger cohort and prospective study. The suppression of Th17 proinflammatory cytokines in MS patients may be conducive to the regeneration of several immunoregulatory cell populations, including Tregs, as these cytokines are known to inhibit this process (79, 80). Future studies are required to investigate the relationship between these post-AHSCT cytokine profiles to the reconstitution of immune cell populations to further elucidate the mechanisms of immunological tolerance regeneration in MS patients following transplantation.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by St Vincent’s Hospital HREC (File number 10/135). The patients/participants provided their written informed consent to participate in this study.

KH, MK, and JCM designed the research methodology. KH and MV performed the experiments. KH, MV, MK, JCM, DM, BW, and JJM analyzed the data. KH wrote the manuscript. DM, IS, and JJM designed the study. JCM and IS reviewed clinical status of MS patients. JJM, DM, and KH provided funding for the study. All authors reviewed and edited the final manuscript. All authors contributed to the article and approved the submitted version.

This study was funded by an NHMRC Postgraduate Scholarship (RG161591), an AMR Translational Research Grant from the St Vincent’s Clinic Foundation, Reset Australia, an Arrow Bone Marrow Transplant Foundation Special Research Grant, the SVH Haematology Research Fund, Maple-Brown Family Foundation, the John Kirkpatrick Family Foundation, NSW Health, and Multiple Sclerosis Research Australia.

DM has received a research grant from Phebra Pty Ltd outside of submitted work. JCM reports personal fees from Biogen, Merck, Sanofi Genzyme and Teva, outside the submitted work.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.