Autoimmune Sequelae Following Rituximab Therapy

A Review of the Literature and Potential Immunologic Mechanisms

Anna E. Kersh, MD, PhD; Ron J. Feldman, MD, PhD


J Clin Rheumatol. 2018;24(8):427-435. 

In This Article


Rituximab Effects on Regulatory B Cells

The reported cases of rituximab-induced autoimmune disease demonstrate the potential vulnerability of the immune system that results from alterations to the B-lymphocyte pool. The most posited alteration is depletion or inappropriate regulatory B-cell function. As B cells repopulate, there is a predominance of transitional and naive B cells including those that produce interleukin 10 (IL-10), which are believed to have regulatory function.[30,31] It is possible that these Bregs play a role in maintaining tolerance after rituximab treatment. This has been supported by mouse models showing that regulatory B10 cells impair the ability of CD4 T cells to produce interferon γ and TNF-α, while also down-regulating the ability of dendritic cells to act as antigen-presenting cells, thus indirectly affecting CD4 T-cell activation and proliferation.[32] A similar mechanism has been suggested to explain the paradoxical worsening of cutaneous lupus in systemic lupus erythematosus (SLE) patients treated with rituximab.[33] In this study, it was postulated that a lack of IL-10 production allows for increased T-cell homing to the skin via up-regulation of T-cell cutaneous lymphocyte antigen.[33]

Increased circulating IL-10 transitional B cells after rituximab treatment usually portend a favorable response, as has been observed in patients with pemphigus with long-term complete responses following rituximab treatment. Unexpectedly, a patient with pemphigus receiving rituximab therapy experienced remission but developed cutaneous and pulmonary sarcoidosis (confirmed by skin biopsy and ACE levels) 6 to 7 months after rituximab treatment, corresponding to a time of rituximab clearance and early repopulation of peripheral B cells (Table 1).[15] At this time, the repopulating B cells were dominated by naive and transitional B cells with a relative paucity of circulating memory B cells and increased levels of IL-10 and B-cell–activating factor (BAFF), which in another setting would have predicted improved peripheral tolerance.[15] This unique combination of B-cell subsets in the setting of elevated BAFF and IL-10 has been observed in the peripheral blood of patients with sarcoid, suggesting the postrituximab milieu was sufficient to somehow promote autoimmune pathogenesis.[34] These autoimmune events can occur during B-cell repopulation or depletion, as was illustrated by an unusual case of NMO, an antibody-mediated demyelinating disease, developing after rituximab treatment in a patient with pemphigus vulgaris (personal observations, Table 1). The autoimmune disease developed during the time of B-cell depletion and while the pemphigus was in remission. Similar observations have been made in SLE patients who experience flares in disease after rituximab treatment. In these patients, increased BAFF levels after B-cell depletion therapy correlate with elevations in anti–double-stranded DNA antibodies and inversely correlate with circulating B-cell numbers.[35] The elevation in BAFF levels after rituximab therapy has been suggested to allow for preferential stimulation and expansion of autoreactive B-cell populations in the setting of an overall reduced B-cell number.[36]

Further evidence for the impact of depleting IL-10–producing B cells on regulation of T-cell biology in autoimmune disease comes from patients treated with rituximab who subsequently experienced colitis (Table 1).[27,37] A patient with ulcerative colitis was treated with rituximab salvage therapy and developed a severe exacerbation that correlated with complete depletion of mucosal B cells and suppression of IL-10 production in the colon along with a dense infiltration of T cells by CD3 staining (Table 1).[37] This depletion of mucosal B10 cells could also explain the development of ulcerative colitis in a patient with Graves disease 2 weeks after treatment with rituximab and was the proposed mechanism for the development of Crohn disease in 2 elderly patients after rituximab treatment for lymphoma (Table 1).[27,28]

Two patients reportedly developed autoimmune reactions in the CNS after rituximab therapy, one developing demyelination after being treated for immune thrombocytopenia (ITP) and the other developing nonvasculitic meningoencephalitis, after treatment for RA (Table 1). Analysis of the CSF of these patients revealed a CD4 T-cell–dominant infiltrate and absence of CD19+ B lymphocytes.[24,25] Data from mouse models support the theory that depletion of B10 cells in the CNS of these patients fosters the priming of myelin-specific CD4 T cells, which in turn contribute to the demyelination (Table 1).[32]

Rituximab-induced psoriasis was the most common autoimmune reaction overall (Table 1). The mechanism by which B-cell depletion elicits a T-cell–mediated autoimmune skin disease remains unclear. The most common hypothesis as discussed previously is that removal of Bregs by rituximab leads to unbridled T-cell activity and disease induction.[6,8,9] There has been no published evaluation of the phenotype of circulating B cells in rituximab-induced psoriasis patients; however, supportive data from mouse models demonstrate IL-10–producing Bregs harness pathogenic T cells in psoriasis and that their depletion exacerbates disease.[38] An alternative explanation is that the depleted humoral immune system permits an aberrant T-lymphocyte response to infection, which could subsequently cause a psoriasis flare.[6,8,9] Interestingly, of the 10 patients reported here, the time from initiation of rituximab to the development of skin symptoms ranges from 2 weeks to 2 years (Table 1). This patient-to-patient variation in the kinetics of disease development may indicate varying mechanisms of immune pathogenesis, each with a different time course to symptom development. These findings are ironic, as rituximab has demonstrated efficacy as a treatment for psoriasis.[39]

The impact rituximab therapy has on the resident immune system of the skin is illustrated in a retrospective study describing patients on biological agents who paradoxically developed hidradenitis suppurativa (HS).[14] The study describes 2 cases of HS that developed in response to rituximab therapy in addition to several cases of HS that developed in response to anti–TNF-α therapies.[14] The authors suggested the disruption of the cytokine milieu in the skin likely contributed to HS pathogenesis; however, the observation that HS develops in response to inhibition of 2 separate pathways of the immune system highlights the complexity of the regulation of the cutaneous immune system.

Rituximab Effects on T Cells

Given the influence of B cells on T-cell activation, it is also possible that rituximab has a direct effect on T cells. Mechanistically, rituximab may result in decreasing activation markers (CD40L, CD69, and CD25) and cytokine production by CD4 T cells.[40–42] Studies in RA patients have shown decrease in frequency of peripheral naive and memory CD4 T cells after rituximab treatment.[43] This may be the result of rituximab-mediated lysis of a small population of CD20+ IL-17–producing T cells in RA patients, although both B-cell–dependent and B-cell–independent mechanisms have been proposed for the observed T-cell depletion.[43–45]

The reported rituximab-induced impairment of CD4 T-cell activation is hard to reconcile with the many reports of T-cell–mediated autoimmune diseases such as psoriasis observed in patients after rituximab treatment (Table 1).[5–8,10–12,29] The activation of T-cell–mediated autoimmune disease in the skin and lung could be explained by tissue-resident memory T cells, a population of nonrecirculating T lymphocytes that populate the skin and barrier tissues such as the lung and intestine, tissues in which the majority of rituximab-induced autoimmune reactions are observed (Table 1).[46] It remains unclear what effects rituximab has on resident memory T cells or the cellular subsets that regulate them. B-cell depletion has been reported to decrease numbers of regulatory B and T lymphocytes, and the relief of such regulatory processes may explain how T-cell–mediated autoimmune disease could be established after rituximab therapy.[38,47] Evidence that topical steroid therapy alleviates symptoms supports a T-cell cytokine-driven pathogenesis and a role for tissue-resident CD4 T cells.[6–8,11]

While CD4 T cells are generally the subset of focus in autoimmune disease, CD8-mediated autoimmune reactions have been documented in response to rituximab therapy. CD8 lymphocytic interstitial pneumonia was observed in a patient treated for follicular lymphoma, who developed symptoms 7 months after starting rituximab, and was distinct from drug-induced interstitial pneumonia because the patient did not develop symptoms after the first dose (Table 1).[17] It is possible that the priming of a CD8 T-cell response in the context of B-cell depletion can be explained by the phenomenon of "cross-presentation." This involves the phagocytosis of extracellular debris by dendritic cells and subsequent presentation of peptide antigen on major histocompatibility complex (MHC) class I molecules to prime CD8 T cells in addition to presentation on MHC class II resulting in CD4 T-cell activation. This has been demonstrated experimentally by the successful priming of lymphoma-specific cytotoxic CD8 T cells after rituximab treatment of B-cell non-Hodgkin lymphoma.[48] The potential adverse ramifications of clonal CD8 T-cell expansion via cross-presentation were observed in a patient who developed T-cell lymphoma in response to rituximab treatment of a high-grade B-cell lymphoma.[49]

Further studies have shown a role for regulation of clonal populations of CD8 T cells by B cells that is dependent on presentation of peptide on MHC class I molecules.[50] These studies demonstrate that antigen presentation by B lymphocytes induces deletion of antigen-specific CD8 T-cell populations and thus tolerance to these antigens.[50] This B-cell–mediated regulation of CD8 T lymphocytes provides another possible mechanistic explanation for postrituximab CD8-mediated autoimmune disease.

Rituximab Effects on Innate Immunity

The inflammasome is an important effector of innate immunity as it responds to metabolic and inflammatory danger signals by activating caspase 1 and subsequent production of IL-1β and IL-18 and can be potentially affected by rituximab therapy. Activation of the Nod-like receptor pyrin domain-containing protein 3 (NLRP3) inflammasome and subsequent production of IL-1β and IL-18 were the proposed mechanism of pathogenesis of interstitial lung disease that developed in 2 patients after rituximab administration for ITP and follicular lymphoma (Table 1).[16] Although both ILs contribute to the inflammatory response, IL-1β is thought to play a significant role in pulmonary fibrotic processes.[51]

Role of Environment

Following rituximab therapy, peripheral CD20+ B cells undergo depletion including CD20+ Bregs, and it is possible that an "immunologic window" exists where conditions favor the development of an unbridled T-cell response to infectious pathogens or gut-resident microbiota.[6] Evidence supportive of a microbiota-driven response comes from sigmoidal biopsies of a rituximab-treated patient who developed ulcerative colitis, which demonstrated repopulation of gut-resident B cells within 3 months of treatment, long before peripheral B-cell reconstitution (Table 1).[27] Upon repopulation, B cells could present antigen and provide costimulation for gut-resident T cells primed with microbiota-derived peptides. The relative depletion of peripheral Bregs would allow for an uninhibited systemic T-cell response to predominate. This illustrates the relatively unexplored differential effects of rituximab treatment on tissue-resident lymphocytes compared with circulating lymphocytes.

Similarly, in a reported case of psoriasis developing 25 months after rituximab therapy for RA, it was suggested that rituximab therapy and subsequent altered immune milieu rendered the patient vulnerable to infection, which could then trigger autoimmune disease by activating a "bystander" autoreactive T cell.[13]

Other environmental factors that warrant consideration are immune-altering treatments given prior to or in conjunction with rituximab treatment and/or the altered immune milieu of the primary disease itself. For example, a patient with follicular lymphoma initially treated with fludarabine, cyclophosphamide, and rituximab was kept on rituximab maintenance therapy and subsequently developed ileocolitis characterized by a dense T-cell infiltrate into the lamina propria of affected colon (Table 1).[26] It is likely that the initial chemotherapy course altered the immune cellular environment in the colon and may have contributed to the development of autoimmune colitis. Additional studies have shown high-dose methylprednisolone therapy in NMO patients reduces the frequency of Bregs and increases the frequency of memory B cells while inhibiting interferon γ production by B cells.[52] As the predominance of memory B cells after rituximab treatment is linked to early relapse or incomplete therapeutic response, it is possible that high-dose methylprednisolone could predispose patients to having a poor therapeutic response. The impact this altered immune milieu has on the development of new autoimmune disease is unclear.

Finally, the local microenvironment of various tissues has variable penetrance with rituximab, particularly in inflamed tissues. Lymph nodes and bone marrow are examples of lymphoid tissues that have been reported to have an inherent resistance to B-cell–depleting therapies and may influence downstream autoimmunity.[53] Poor rituximab infiltration of the salivary glands in patients with Sjögren syndrome correlates with the persistence of clonally related B-cell populations in these glands after treatment, which may influence disease exacerbation.[54,55] In the patient who developed interstitial lung disease following rituximab therapy for ITP, diffuse pulmonary infiltrates of CD19+ B cells were noted 2 weeks after completing a course of rituximab, suggesting pulmonary tissue may exhibit poor penetration of rituximab in specific situations.[16]