Report from a Satellite Symposium held at the 31st Congress of the European Committee for Treatment and Research in Multiple Sclerosis in Barcelona, Spain, 9 October 2015
The role of B cells in the pathogenesis of multiple sclerosis (MS) may not be simply related to their ability to produce antibodies. They are highly efficient antigen-presenting cells, producing cytokines that can change the microenvironment and can mediate negative effects through astrocyte populations. Furthermore, as well as producing antibodies, B cells produce ectopic lymphoid follicle-like aggregates that persist in the brains of MS patients. This improved understanding of the centrality of the B cell in the biology of MS presents greater opportunities for developing effective therapies. The lymphocyte antigen CD20 is not expressed at early stem and pro B cell stages, nor on most short- or long-lived plasma cells. This presents the possibility that anti-CD20 treatment could deplete the intermediate stage of B-cell development while preserving the ability of stem cells to repopulate and protecting pre-existing humoural immunity. Ocrelizumab is a humanised monoclonal antibody that depletes CD20+ B cells via multiple mechanisms. In the OPERA I and OPERA II trials, compared with interferon beta-1a (IFNβ-1a) treatment over 96 months, ocrelizumab significantly reduced: the annualised relapse rate, 12- and 24-week confirmed disease progression, T1 gadolinium-enhancing lesions and new and/or enlarging T2 lesions. Overall, in OPERA I and OPERA II, ocrelizumab had a similar safety profile to that of IFNβ-1a over the study period. The OPERA I and OPERA II studies therefore provide strong support of for the theory that targeting CD20+ B cells as a potential therapeutic approach in relapsing MS.
Much progress has been achieved in the treatment of multiple sclerosis (MS), however, many important unmet needs remain. A large proportion of patients with MS experience disease activity despite treatment with disease-modifying therapies (DMTs), whereas the desired treatments would have the potential to impact neurodegeneration and promote re-myelination. For some treatments there may be a compromise between efficacy and safety, however, ideally treatments would be well tolerated, highly efficacious and have favourable benefit–risk profiles. The attributes of currently available treatments can present adherence challenges. Treatments with mechanisms of action that promote persistence and adherence are therefore needed. There are a number of treatments currently in development that may meet at least some of these needs and have interesting potential to improve outcomes in MS (Table 1).
The role of B cells in multiple sclerosis
Derived from haematopoietic stem cells residing in the bone marrow or liver, pre-B cells can evolve into mature naïve cells that can migrate throughout the body into secondary lymphoid tissue (Figure 1). If they encounter an antigen that can activate or cross-link their B-cell receptors, they are activated and can move into germinal centres, where they receive help from dendritic cells and T-cells to proliferate, clonally expand and undergo further antigen-driven maturation of the B-cell receptor. This then ultimately yields plasmablasts and memory B cells. The plasmablasts can move into other tissues, in particular, the bone marrow, and continue to produce antibodies for years, potentially decades. The memory B cells also circulate and can mediate surveillance of the entire body, in this case, they can enter the brain and, again, if their B-cell receptors encounter appropriate antigens, they can become activated and receive T-cell help to undergo further clonal expansion. The cells that are clonally expanded can become long-lived in the central nervous system (CNS), ultimately developing the follicle-like aggregates that characterise the brain of MS patients with chronic disease. The result is development of plasmablasts and plasma cells, which are believed to be the source of oligoclonal bands seen in the cerebrospinal fluid of patients with MS. In addition, the CNS-educated B cells can recirculate. Research carried out at the University of San Francisco using a deep sequencing technique to identify the lineage of individual B cells indicates that it is possible to identify B cells in the periphery that are derived from the same clonal lineage as those residing chronically in the brain. The findings suggest that the B-cell population is moving rapidly back and forth across the blood–brain barrier and, further, it appears that B cells are undergoing more clonal expansion in the periphery.
There is another evolving concept in the cellular aetiology of MS. The brain is not an ‘immune desert’ and, in fact, 40–60% of cells in the brain are members of the innate immune system, i.e., astrocytes, which express class I and II major histocompatibility complex; and can internalise and process antigens and present them to T-cells (Figure 2).1–7 There is an emerging model suggests that this T-cell–Bcell interaction may be more complicated than formerly believed. This is based on two key processes. Firstly, B cells that internalize antigen that bind to their B-cell receptor are highly efficient antigenpresenting cells, being up to 10,000 times more efficient than dendritic cells. B cells may consequently be the primary antigen-presenting cell in an MS plaque that activate T cells, resulting in the inflammatory cascade. Secondly, there appears to be involvement of astrocytes, the innate immune cells Interleukin-1β and interferon-γ activate astrocytes to become type II astrocytes that upregulate inducible nitric oxide synthase (iNOS) and produce tumour necrosis factor, leading to axonal injury and oligodendrocyte and neuronal death. This means that there may be two steps with two different targets in the immune cascade, i.e., in both the innate and adaptive immune system. Whether purified B cells from MS patients could themselves damage the brain has been the subject of much exploration. CD19+ B lymphocytes from treatment-naive patients with MS-induced demyelination of cerebellar slices derived from mice, suggesting that B-cells have a direct effect on the brain.8 Although we currently consider the approved therapies as working via T cells, all therapies have now been reported to have B cell effects.
New possibilities with B cell targeted therapy
Along the B cell lineage, starting as stem cells, B cells differentiate into pro B cells and then undergo, a process of diversification that occurs mostly in secondary lymphoid organs, which results in antigen-triggered memory B cells, plasmablasts and short- and longlived plasma cells (Figure 3).9–12 B cells are also characterised by a variety of surface markers, which serve as guides to B-cell
differentiation staging. The CD20 surface marker is a membranespanning protein that is expressed at the intermediate stage of B cell differentiation. CD20 is not expressed at the earlier stem and pro B cell stage or on most short-lived plasma cells or any long-lived plasma cells. This poses the possibility that anti-CD20 treatment could deplete the intermediate stage of B-cell development while preserving the ability of stem cells to repopulate and protecting preexisting humoural immunity.
Ocrelizumab is a humanised monoclonal antibody that depletes CD20+ B cells via multiple mechanisms.13–15 The predominant mechanism is believed to be stimulation of cell-mediated antibody-dependent cellular phagocytosis and antibody-dependent cellular cytotoxicity. Ocrelizumab also depletes CD20+ B cells by complement-dependent cytotoxicity and by direct apoptosis, but it is thought that the cellmediated mechanisms predominate. This may explain some of the characteristics in terms of infusion-related reactions (IRR) associated with ocrelizumab. The humanised, rather than chimeric nature of the antibody is another potential advantage.
There is a new understanding of a very dynamic recirculation of B cells from the blood into the brain out into secondary lymphoid tissues (Figure 1). It is thought that activation and stimulation of clonally specific B cells is occurring on both sides of the blood–brain barrier. These cells are released from the protective cytokine-rich lymphoid niches that are present in most secondary lymphoid organs and it is mostly these motile B cells that are targeted by ocrelizumab. It is believed that with anti-CD20 therapy, there is a fairly rapid depletion of the circulating memory B cells and the egressing CNS-educated B cells which is substantially greater than any effects on secondary lymphoid tissue. In phase II and III trials, ocrelizumab has been studied in over 1500 MS patients with over 4000 patient-years of experience.16 The OPERA I and OPERA II studies were double-blind, double-dummy controlled trials with interferon beta-1a (IFNβ-1a) 44 μg three times per week subcutaneously as a comparator and ocrelizumab given as 4 doses four cycles at 24-week intervals. The initial cycle was administered as two 300-mg infusions on Days 1 and 15 for the first dose, and as a single 600-mg infusion thereafter (Figure 4). An infusion of 100 mg of methylprednisolone was given 30 minutes prior to each ocrelizumab dose. The objective of OPERA I and OPERA II was to evaluate the efficacy and safety of ocrelizumab compared with IFNβ-1a in patients with relapsing MS (RMS). The primary endpoint was the annualised relapse rate (ARR) at 96 weeks and key secondary endpoints included: pooled 12- and 24-week confirmed disability progression (CDP), number of gadolinium (Gd)-enhancing lesions and the number of new and/or emerging T2 lesions at weeks 24, 48 and 96.
Over 85% of patients completed the OPERA I and OPERA II studies. Slightly more patients withdrew in the IFNβ-1a arms compared with the ocrelizumab arms in both studies (OPERA I: 17% versus 10%, respectively and OPERA II: 23% versus 14%, respectively). Nearly all (96%) patients entered the open-label extension phase. Patients who withdrew for any reason were then moved into a safety followup phase. MS disease history and baseline characteristics were balanced between the treatment arms in both OPERA I and II, and were representative of a typical RMS population. Mean time since onset of MS was six or more years and mean age was approximately 37 years old. Gender dimorphism (66% of patients in both studies were female) was as expected for MS, and the mean Expanded Disability Status Scale (EDSS) was 2.8. Over 70% of patients were untreated in the past 24 months. There was a 46% reduction in ARR with ocrelizumab compared with IFNβ-1a (p<0.0001) in OPERA I and a 47% reduction favouring ocrelizumab in OPERA II (p<0.0001 in both studies). The adjusted ARR for ocrelizumab was <0.16 in both trials compared with 0.292 and 0.290, for IFNβ-1a in OPERA I and OPERA II, respectively. Consistency was a hallmark between the OPERA I and OPERA II study results. There was also a significant reduction in CDP in the pre-specified pooled analysis of pooled data from OPERA I and OPERA II.For both time to 12-week CDP and 24-week CDP there was a risk reduction of 40% favouring ocrelizumab, which was consistent across the two studies.
Although initially not sufficiently powered to detect a treatment effect on CDP in each of the individual studies, the OPERA trials individually, but this was investigated in exploratory analysis, which revealed a consistent reduction in 12- and 24-week CDP. On imaging endpoints, a significant reduction was observed in the number of T1 Gd+ lesions with ocrelizumab treatment compared with IFNβ-1a (94% and 95% ARR reduction versus IFNβ-1a in OPERA I and II, respectively; p<0.0001). There was a striking 94% reduction (p<0.0001) in the mean number of T1 GD+ lesions favouring ocrelizumab in OPERA I and in OPERA II, this was 95% (p<0.0001). This remarkable effect was consistent and sustained across the 96-week treatment period. There was also a significant reduction in the number of new and/or enlarging T2 hyperintense lesions compared with IFNβ-1a (77% and 83% for OPERA I and OPERA II, respectively). This result was also highly statistically significant (p<0.0001) and there was significant improvement associated with ocrelizumab versus IFNβ-1a in terms of the exploratory endpoints of rate of brain volume loss from baseline to week 96 (23.5% and 23.8% reduction in OPERA I and II, respectively; p< 0.0001) and no evidence of disease activity (NEDA) (64% and 89% improvement in OPERA I and II, respectively; p< 0.0001).
OPERA I and II – safety outcomes
Adverse events (AEs) over 96 weeks in the OPERA I and II studies are shown in Table 2. The number of patients with one or more AEs was similar the IFNβ-1a and ocrelizumab arms. Similarly, the total number of patients with one or more AEs occurring at a frequency of at least 5% in either arm was virtually identical in the two treatment arms (Table 2). There were more general disorders and administration site reactions in the group treated with IFNβ-1a. Slightly more infections (mainly upper respiratory tract) occurred in the ocrelizumab arms versus IFNβ-1a (58.4% vs. 52.4%), although other AEs were well balanced between treatments. There were slightly fewer patients with one or more serious AEs (SAEs) in the ocrelizumab group compared with IFNβ-1a (Table 3).
During OPERA I and OPERA II, three deaths occurred (in the IFNβ-1a arm, one case of suicide and one of mechanical ileus and in the ocrelizumab arm, one case of suicide). Six malignancies were
reported (in the IFNβ-1a arm: mantle cell lymphoma and squamous cell carcinoma, and in the ocrelizumab arm: renal cancer, melanoma and two breast cancers).
The most common AE associated with ocrelizumab was the IRR, which were mostly mild to moderate in intensity (less than 2% were severe in intensity). Eleven patients (1.3%) withdrew from ocrelizumab treatment due to an IRR during the first infusion.
• The role of B cells in the pathogenesis of MS may not be simply related to their ability to produce antibodies.
• B cells are likely to be important in MS because of their ability to be highly efficient antigen-presenting cells. They also produce cytokines that can change the microenvironment and can mediate negative effects through astrocyte populations, as well as producing antibody that may have a role in addition to producing ectopic lymphoid folliclelike aggregates that persist long term in the brains of MS patients.
• This improved understanding of the role of B cells in MS presents greater opportunities for developing effective therapies.
• Compared with IFNβ-1a, ocrelizumab significantly reduced:
• 12- and 24-week CDP;
• T1 GD+ lesions; and
• new and/or enlarging T2 lesions.
• In exploratory analysis compared with IFNβ-1a, ocrelizumab:
• reduced brain volume loss; and
• increased proportion of patients with NEDA.
• Ocrelizumab is the first investigational treatment for MS to significantly reduce, in two separate phase II studies:
• Both 12- and 24-week CDP against any comparator
• CDP versus active comparator.
• OPERA I and OPERA II showed that targeting CD20+ B cells is a potential therapeutic approach in RMS.
• Overall, in OPERA I and OPERA II, ocrelizumab had a similar safety profile compared with IFNβ-1a over 96 weeks.
• AEs occurred with similar frequency in the ocrelizumab and IFNβ-1a groups, with the following exceptions:
• IRRs associated with ocrelizumab;
• infections and infestations: IFNβ-1a (52.4%), ocrelizumab (58.5%); and
• influenza-like illness and local cutaneous reactions associated with IFNβ-1a.
• Overall, the proportion of patients reporting an adverse event in the controlled treatment period was identical:
• IFNβ-1a (83.3%) ocrelizumab (83.3%).
• The proportion of patients reporting a SAE in the controlled treatment period was low and similar between groups:
• IFNβ-1a (8.7%), ocrelizumab (6.9%).
• Targeting CD20+ B cells may preserve B cell reconstitution and long-term immune memory (Figure 3), which may account for the favourable safety and tolerability profile observed for ocrelizumab.
• In summary, B-cell targeted therapy, such as ocrelizumab, offers exciting possibilities for the treatment of MS.