Sjögren’s Syndrome–Associated Lymphoma

1

Introduction

Sjögren’s syndrome (SS) is an autoimmune systemic disease associated with B-cell lymphoproliferation and an increased risk of B-cell non-Hodgkin lymphoma (NHL) development compared with the general population.

NHLs are a highly heterogeneous group of malignancies that originate in lymphatic hematopoietic tissue, classified into B-cell and T-cell lymphomas, which account for about 90% and 10% of lymphomas respectively. The distinct subtypes of NHL show significant differences in epidemiological and geographical distribution and in histological patterns, suggesting a potential interplaying role of many different factors, such as infectious triggers, individual genetic status, and environment influences, in the development of NHL. Despite this heterogeneity, new insights into NHL etiology have focused on possible underlying common mechanisms, considering immune modulation and chronic antigenic stimulation as the main shared basis of NHL pathophysiology.

The association between autoimmune chronic inflammatory diseases and NHL has been demonstrated by several studies. In autoimmune disorders, an antigen-driven chronic stimulation in the context of specific genetic background would play a central role in generating and perpetuating a clonal lymphocytic expansion through a multistep oncogenic process, leading to an overt malignant disease at the end. The sustained risk of lymphoma observed in chronic autoimmune conditions over time, especially in SS, and the time-dependent increase of malignant transformation risk, strongly support this hypothesis.

A recent pooled analysis, of 8692 NHL cases and 9260 controls from 14 studies (1988–2007) within the International Lymphoma Epidemiology Consortium, evaluated the interaction between immune system genetic variants and autoimmune conditions in NHL risk. The authors confirmed that autoimmune conditions mediated by B-cell responses were associated with increased NHL risk, specifically diffuse large B-cell lymphoma (DLBCL) (odds ratio [OR] = 3.11, 95% confidence interval [CI]: 2.25, 4.30) and marginal zone B-cell lymphoma (MZBCL) (OR = 5.80, 95% CI: 3.82, 8.80), whereas those mediated by T-cell responses were associated with peripheral T-cell lymphoma (OR = 2.14, 95% CI: 1.35, 3.38). Moreover, in the presence of the rs1800629 AG/AA genotype of tumor necrosis factor (TNF) gene, the risk of NHL in B cell-mediated autoimmune conditions increased (OR = 3.27, 95% CI: 2.07, 5.16; P-interaction = 0.03).

2

Epidemiology and Histological Subtypes of Lymphoma in Sjögren’s Syndrome

The analysis of NHL risk in different large cohorts of patients with autoimmune disease shows the highest risk in SS and in cryoglobulinemic vasculitis (CV) compared with the other autoimmune diseases. In SS, 6.1- to 44.4-fold increased risk of NHL was observed and followed by much lower risk in systemic lupus erythematosus (SLE) and rheumatoid arthritis. The lifetime risk of NHL in SS is about 5% in most studies. It has become clear that the occurrence of lymphoma is a somewhat later complication of SS, with a median time from diagnosis of 7.5 years in the multicenter study by Voulgarelis et al. In this study the incidence of lymphoma in primary SS (pSS) was estimated to be 4.3%. No gender differences in lymphoma evolution have been reported in a recent meta-analysis.

SS has been linked to an increased risk of specific subtypes of NHL, ie, MZBCL, with the mucosa-associated lymphoid tissue (MALT) lymphoma being much more common than the nodal MZBCL, and the more aggressive DLBCL. The distribution of these subtypes is not homogeneous in different cohorts of SS patients but MALT NHL has been widely recognized as the most common subtype.

In the pooled analysis of autoimmune conditions and the risk of NHL subtypes conducted within the InterLymph Consortium, SS was associated with a ninefold increase in risk of DLBCL (OR 8.92; 95% CI, 3.83–20.7) and with a 30-fold increase in risk of MZBCL (OR 30.6; 95% CI, 12.3–74.6), with a 1000-fold increased risk when the parotid gland MALT lymphoma was considered in detail (OR = 995; 95% CI, 216–4596).

In a Swedish registry-based study, the predominance of the DLBCL subtype has been reported. Trigger stimuli of lymphomagenesis in different geographic regions were speculated. However, in a large multicenter European analysis of SS-associated NHL by Voulgarelis et al., there was no difference in the distribution of different subtypes and grades of SS lymphoma between the north and south of Europe.

Whereas the multistep evolution of lymphoproliferative lesions in SS patients toward an overt MALT lymphoma has been widely studied, the development of DLBCL in the context of SS still remains ill-defined. Of note, it has been recently demonstrated that DLBCL may arise from the same B-cell clone of a previous indolent low-grade lymphoma, especially from a preexisting MALT lymphoma, with the addition of genetic alterations, finally resulting in a high-grade and more aggressive malignant process. Thus the development of DLBCL complicating the course of SS may occur as an evolution of an indolent NHL, especially of the MALT type.

SS patients with NHL have an increased mortality compared with the general population. A large retrospective Greek study evaluated the outcome of 53 B-cell NHL cases in a cohort of 584 patients with pSS diagnosis. The median age at NHL diagnosis was 54 years (range, 28–90 years) and the median period from diagnosis of SS to NHL was 11 years (range, 1–23 years). Age and sex were not significantly different compared with the SS population without lymphoma. Thirty-one NHLs (59%) were of the MALT type, 8 (15%) DLBCL, 8 (15%) nodal marginal zone lymphomas (NMZLs), and 6 (11%) other lymphomas. The overall survival for the entire NHL cohort was 0.96 (95% CI, 0.83–0.99) at 3 years and 0.92 (95% CI, 0.76–0.97) at 5 years. The actual overall survival at 3 years was 0.97 for patients with MALT lymphoma, 0.80 for those with NMZL, and 1.0 for those with DLBCL (with a marked improvement from 0.37 to 1.0 after the introduction of rituximab in the management of DLBCL).

In this study the age/sex-adjusted standardized mortality ratio (SMR) of the SS-NHL group compared with the general population in Greece was estimated at 3.25 (95% CI, 1.32–6.76), and the SMR of pSS without NHL was 1.08 (95% CI, 0.79–1.45), comparable to the general population. Notably, in another Greek study, Skopouli et al. demonstrated that in SS the SMR increased to 2.07 (95% CI, 1.03–3.71) only in the presence of adverse predictors of NHL, such as purpura, mixed cryoglobulinemia and low C4 levels. When patients with adverse predictors were excluded from the analysis, the mortality rate of pSS was the same as the general population (SMR 1.02).

In a Swedish prospective cohort study, excess mortality in pSS was found only for the subgroup of patients with a lymphoid malignancy (cause-specific SMR 7.89 [95% CI 2.89–17.18]), corresponding to 2.53 excess deaths per 1000 person-years at risk. Overall, no increased mortality has been found in SS patients without lymphoma, compared with the general population.

All this data demonstrate that malignant lymphomas have a very high impact on SS survival and mortality, supporting the efforts to identify the possible adverse predictors.

3

Predictors of Lymphoma in Sjögren’s Syndrome

The clinical manifestations during the course of SS that could predict an evolution to an NHL have been extensively studied since the 1970s. The persistent enlargement of parotid glands and the presence of lymphadenopathy and/or splenomegaly were identified as the most important clinical signs.

The association between mixed monoclonal cryoglobulinemia (linked to a complement decrease) and the risk of lymphoma in SS was first documented by Tzioufas et al., who also reported an increased risk related to the presence of specific monoclonal rheumatoid factor (mRF)–associated cross-reactive idiotypes (CRIs), such as 17,109 and G-6.

Skoupouli and coworkers reported that the development of lymphoproliferative disorders in SS was associated with low levels of C4 complement (relative risk, 7.5; P = 0.0016), the presence of mixed monoclonal cryoglobulins (relative risk, 7.9; P = 0.0012), and purpura (relative risk, 3.9; P = 0.037). Low levels of C4 were additionally indicated as the strongest predictor for mortality after adjusting for age (relative risk, 6.5; P = 0.0041). Of interest, in this study the authors focused on the concept that the initial presentation of SS with these manifestations could determine the outcome and mortality of SS.

The need to stratify SS patients on the basis of their risk of developing lymphoproliferative disorders, including lymphoma, was assessed for the first time by Ioannidis et al. At SS diagnosis, parotid enlargement (hazard ratio [HR] 5.21, 95% CI 1.76–15.4), palpable purpura (HR 4.16, 95% CI 1.65–10.5), and low C4 levels (HR 2.40, 95% CI 0.99–5.83) were independent predictors of lymphoproliferation in pSS. The authors then proposed a classification of pSS distinguishing two distinct categories according to the presence or absence of low C4 levels and/or palpable purpura, respectively: the type I high-risk pSS that accounts for about 20% of pSS at diagnosis and the more common (80%) type II low-risk pSS without a significant increase in risk of lymphoma evolution. Brito-Zeròn and coworkers confirmed this data and identified besides palpable purpura and low C4, severe parotid involvement, demonstrated by scintigraphy, and serum cryoglobulinemia as two additional risk factors of lymphoma development in their study cohort analysis. A survival analysis performed in this study found that patients with at least two adverse factors at diagnosis (parotid involvement by scintigraphy, cryoglobulinemia, purpura, and low C4) had a significantly lower survival rate than patients without risk factors.

In a large retrospective Greek study investigating hematologic manifestations and predictors of lymphoma development in pSS by Baimpa and coworkers, lymphocytopenia was the only independent variable predicting the development of any type of lymphoma other than MZBLC, especially the DLBCL type, whereas neutropenia, low C4 levels, cryoglobulinemia, lymphadenopathy, and splenomegaly were independent predictors for the development of MZBCL type.

Low lymphocyte cell count has also been reported as predictor of NHL in pSS in a cohort study on cancer incidence and lymphoma predictors in pSS by Theander et al., which identified the CD4+/CD8+ T-cell ratio ≤ 0.8 (HR 10.92, 95% CI 2.80–41.83) as the strongest risk factor for developing lymphoma. Other predictors in this study were CD4+ T lymphocytopenia (HR 8.14, 95% CI 2.10–31.53), low C4 (HR 9.49, 95% CI 1.94–46.54), purpura/skin vasculitis (HR 4.64, 95% CI 1.13–16.45), and low complement factor C3 (HR 6.18, 95% CI 1.57–24.22).

Reduced complement levels are closely associated with the two worst adverse outcomes in SS: lymphoma development and death. Many authors have reported that low levels of C3 and C4 were predictors of both lymphoma evolution risk and mortality in SS, mainly as the result of lymphoproliferative disease. Only in Ioannidis’ study were low C3 levels not significantly predictive for increased mortality, possibly because of a lower cut-off point of C3 used in the Greek study (0.50 g/L instead of 0.83 g/L).

Overall, based on these studies, it can be concluded that a heavy MALT involvement, mainly of the parotid glands (clinically leading to their persistent swelling), and cryoglobulinemia (often linked to complement consumption and with possible vasculitic features such as purpura) are the two main predictors of lymphoma in SS.

Two clinical entities of SS, ie, CV and the persistent enlargement of salivary glands (usually parotid swelling), can therefore be considered as prelymphomatous conditions. Focusing on this particular issue, a multicenter Italian study by Quartuccio et al. investigated the association between laboratory biomarkers and lymphoma risk in pSS, differentiating patients with prelymphomatous conditions (CV and salivary glands (SGs)/parotid enlargement) as separate groups in order to better evaluate in specific different subsets the risk of lymphoma development. All the selected 601 pSS patients, fulfilling the American European Criteria for the classification of pSS, were negative for hepatitis C virus (HCV) antibodies and were repeatedly tested for the presence of cryoglobulins. Patients were categorized into four groups: group 1/NHL, patients with lymphoma (including lymphoma patients with concomitant CV and/or SG swelling); group 2/CV, patients with CV and without lymphoma; group 3/SW (swelling), patients with SG swelling without lymphoma, with or without concomitant CV; and group 4/pSS controls, pSS patients without lymphoma and without CV or SG swelling. The study showed that four biomarkers, ie, cryoglobulinemia, low C4, anti-La/SSB antibodies, and leukopenia, were significantly associated to lymphoma in pSS. An interesting finding was that in the group of patients with persistent SG swelling without lymphoma, the presence of two of these four biomarkers identified a 9-fold higher risk of lymphoma, whereas the positivity of only one or no biomarker provided a negative predictive value for lymphoma of about 90% in the same subset.

Because SG swelling is a more common finding in SS (about 30%) compared with lymphoma (about 5%), this study better characterized for the first time the risk of NHL in SS patients with SG swelling, with an increased risk of lymphoma only in those patients with at least two negative predictors.

Additional new data has recently been provided comparing the clinical and laboratory features in pSS patients who are positive or negative for the anti-Ro/SSA and/or anti-La/SSB antibodies. Of note, anti-Ro/SSA–anti-La/SSB-negative pSS showed a lower risk of lymphoma evolution. Moreover, a younger age at SS onset determined an increased risk for lymphoproliferation.

4

Lymphomagenesis in Primary Sjögren’s Syndrome

SS is characterized by both B- and T-cell lymphocytic infiltrates in inflamed SGs and a B-cell hyperactivity that leads to a significant expansion of B-cell clonal populations in different times, tissues, and stages, a process potentially evolving into a malignant B-cell monoclonal lymphoma. Polyclonal autoantibodies such as anti-Ro/SSA, anti-La/SSB, and rheumatoid factor (RF) are produced in the inflamed MALT tissue.

MALT lymphoma arises from chronic inflamed tissues through a multistep process in which a local, chronic antigenic stimulation, together with a predisposing genetic background, allows the emergence of B-cell clonal expansion and the evolution to an overt malignant process.