Immunologic Characteristics and Predictive Risk Factors of Primary Sjögren's Syndrome Complicated with Malignant Lymphoma: A Multicenter Case-Control Study.
Yuebo Jin,Linqing Hou,Hao Li,Liyun Zhang,Yuanhong Peng,Naidi Wang,Jing He
DOI: https://doi.org/10.1097/cm9.0000000000002912
IF: 6.133
2024-01-01
Chinese Medical Journal
Abstract:To the Editor: Primary Sjögren's syndrome (pSS) is a multi-systemic autoimmune disease characterized by the exocrine gland (mainly salivary and lacrimal glands) involvement, leading to xerostomia and xerophthalmia.[1] The course of pSS is relatively slow, but patients with pSS had higher risk of overall cancer, including malignant lymphoma (ML).[2] It is widely known that all-cause mortality rates were not increased in pSS patients as compared with healthy controls. Nonetheless, secondary lymphoproliferative disease was an important cause of excess mortality. Talal et al[3] first reported the association between pSS and lymphoma in 1964. Subsequent studies reported that the occurrence of lymphoma is closely correlated with various rheumatic diseases. Among rheumatic diseases, lymphoma risk in pSS patients is higher than the risk of systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). The known risk factors of pSS-ML include repeated parotid gland enlargement and anemia, among others. Here, we investigated the correlation between pSS and lymphoma. Furthermore, we assessed overall survival across both cohorts and explored relevant risk factors for lymphoma in pSS to identify early diagnosis and intervention predictors. This was a multicenter case-control study conducted at three participating hospitals in China: Peking University People's Hospital, Linfen Third People's Hospital, and Affiliated Hospital of North Sichuan Medical College. A total of 120 patients diagnosed with pSS were enrolled from January 2010 to June 2022, of which 40 pSS patients with lymphoma were defined as study group for analysis, and 80 pSS patients without lymphoma were matched as controls. All patients fulfilled American-European Consensus Group (AECG) 2002 revised classification criteria for pSS. Lymphoma was diagnosed by pathology biopsy or positron emission tomography-computed tomography (PET-CT). The study was approved by the ethical committee of Peking University People's Hospital (No. 2017PHB167-01). As a retrospective study using data from anonymous patients, the requirement for patient written consent was waived. Statistical analyses were performed with IBM SPSS, version 20.0 (IBM Corp, Armonk, USA). Continuous variables were compared by Student's t test or Mann–Whitney U test, described as mean ± standard deviation (SD) or median (P25, P75). Categorical data were analyzed by Chi-square test or Fisher's exact test, presented as frequencies with percentages. The Kaplan–Meier curve was performed for assessing overall survival and prognosis. Binary logistic regression was conducted to explore associations between pSS and lymphoma, estimating odds ratios (ORs) and 95% confidence intervals (95% CIs). P-value below 0.05 was considered statistically significant. Forty pSS patients with lymphoma were included in the study group, of which 90% were women (n = 36). The mean age was 55.6 ± 13.6 years, and median disease duration of pSS was 9 (1, 30) years. Among them, 9 ML occurred before pSS diagnosis; 18 ML occurred after pSS diagnosis; and 13 ML were diagnosed at the same time as pSS. Pathology data of 28 lymphoma patients were collected [Supplementary Table 1, https://links.lww.com/CM9/B900]. All 28 cases were non-Hodgkin lymphoma (NHL) including B-cell lymphoma (n = 25) and T-cell lymphoma (n = 3). Additionally, the B-cell types mainly covered marginal zone lymphoma (MZL) (16/28, 57%), including malignant mucosa-associated lymphoid tissue (MALT) lymphoma (13/28, 46%), nodal margin zone lymphoma (NMZL) (2/28, 7%), and splenic marginal zone lymphoma (SMZL) (1/28, 4%), followed by diffuse large B-cell lymphoma (DLBCL) (9/28, 32%). Meanwhile, lymphomas were mostly located in extranodal tissues (18/40, 45%) and then lymph nodes (15/40, 38%) [Supplementary Table 2, https://links.lww.com/CM9/B900]. The clinical characteristics and laboratory data of pSS-ML group and control group are shown in Supplementary Table 3, https://links.lww.com/CM9/B900, in which no significant differences were observed in clinical symptoms including dry mouth, dry eye, and rampant caries. There was a higher incidence of lymphocytopenia (P = 0.001) and anemia (P = 0.010) in the pSS-ML group. Interstitial lung disease and renal tubular acidosis incidences were higher in the study group than those in the control group (P = 0.036 and P = 0.027, respectively). pSS-ML patients also showed higher levels of EULAR Sjögren syndrome disease activity index (ESSDAI) than pSS patients without ML (26.85 ± 7.87 vs. 11.58 ± 5.86, P <0.001). Supplement Table 3, https://links.lww.com/CM9/B900 also presents immunological parameters of both groups. Elevated IgA and monoclonal (M) protein were more common in the pSS-ML group (P <0.001 and P <0.001, respectively). However, there were more patients with anti-SSB antibody in the control group than in the pSS-ML group (P = 0.006). Of note, pSS-ML patients showed lower percentages of regulatory T (Treg) cells (6.38 ± 2.15 vs. 8.19 ± 1.87, P = 0.016) but higher percentages of peripheral T follicular helper (pTfh) cells (3.24 ± 1.52 vs. 2.27 ± 0.92, P = 0.032), tumor necrosis factor-alpha-producing T helper 1 cells (Th1-TNF-α) (49.20 ± 13.93 vs. 36.60 ± 12.37, P = 0.006), and T helper 17 (Th17) cells (1.85 ± 0.62 vs. 1.37 ± 0.50, P = 0.021) [Supplementary Table 4, https://links.lww.com/CM9/B900]. Markers and gating strategy of T/B/NK subsets are summarized in Supplementary Figure 1, https://links.lww.com/CM9/B900. Other immunological indexes, such as ANA, elevated IgG, and declined C3 showed no significant differences. Among patients with lymphoma, six cases underwent surgery, 17 participants were treated with chemotherapy, 15 patients only accepted pSS maintenance treatment, 7 individuals were treated with rituximab, 1 case was treated with surgery combined with rituximab, and 1 case received hormone therapy. After therapeutic approach of rituximab, two patients developed thrombocytopenia requiring cyclosporin therapy, after which platelet levels remained within the normal range. In the study group, two untreated patients died at a mean age of 64 years. Therapeutic regimens and prognosis in pSS-ML patients (n = 40) were presented in Supplementary Figure 2A, https://links.lww.com/CM9/B900. For 38 patients who were treated, most patients' disease condition kept stable (31/38, 82%), five relapsed (5/38, 13%), and two died (2/38, 5%). The causes of death were viral encephalitis (n = 1) and recurrence after chemotherapy (n = 1), which were correlated with the primary disease. The control group had three cases of death with the mean age of 71 years. The causes were infection (n = 2) and pulmonary arterial hypertension (n = 1), which were not correlated with primary disease. It is noteworthy that the mean survival time for pSS complicating lymphoma was significantly shorter than that of the control group (24.2 ± 2.6 years vs. 34.5 ± 0.9 years, P = 0.006) [Supplementary Figure 2B, https://links.lww.com/CM9/B900]. For all the 40 pSS-ML patients and 80 pSS individuals without ML, multivariate logistic analysis adjusted for pSS duration was applied to assess risk factors for ML. As shown in Supplementary Figure 2C, https://links.lww.com/CM9/B900, the factors associated with increased risk of ML were higher ESSDAI levels (OR = 1.475, 95% CI = 1.260–1.727), anemia (OR = 2.768, 95% CI = 1.241–6.173), lymphocytopenia (OR = 4.467, 95% CI = 1.958–10.192), elevated IgA (OR = 5.168, 95% CI = 2.087–12.796), and elevated Th1-TNF-α% (OR = 1.066, 95% CI = 1.009–1.127). In the present study, pSS-ML cases were non-Hodgkin's lymphoma, among which MZL accounted for 57% and DLBCL accounted for 32%. Furthermore, MALT has a proportion of 46%, most of which were located in extranodal tissues (45%). The most frequently affected extranodal tissues were parotid glands, followed by thymus and thyroid, again, as previously reported.[4] These results might suggest physicians to emphasize overly enlarged tissue and send specimens for pathological examination in time for early lesion detection. Here, we observed that higher scores of ESSDAI were significantly associated with ML. With respect to laboratory data, anemia and lymphocytopenia had higher odds of ML risk. Additionally, IgA elevation together with Th1-TNF-α% subsets increased risk of lymphoma among pSS. Prior studies have pointed out that reduction of CD4+/CD8+ T cell proportion was a strong predictive agent for lymphoma. In our study, the pSS-ML group had a lower absolute number of CD4+ T cells than the control group. When compared with controls, the study group showed higher levels of Treg%, possibly suggesting pSS patients with lower levels of Treg cells were more likely to suffer from secondary lymphomas. But this tendency did not reach a significant difference with adjustment for SS duration. Due to the imbalance of T subsets between two groups, we suggest pSS patients should be closely monitored for changes in lymphocyte subsets. On the other hand, therapeutic regimens and disease management in the pSS-ML group were made in accordance to the immunopathological and histological classification as well as disease staging by EULAR. These patients were treated with personalized chemotherapy, surgery, maintenance treatment of pSS, or CD20 monoclonal antibody therapy. Currently, there is no specific guideline for treatment of pSS complicating ML, but treatment programs for pSS-ML should involve treatments for both pSS disease and lymphoma condition. Notably, patients with asymptomatic lymphoma and no SS disease activity may not be treated. According to the follow-up cohort, MALT patients reached remission after CD20 monoclonal antibody therapeutic regimen and showed good clinical response of pSS. Rituximab was one of standard first-line medications for several lymphoma types, and some new therapies showed high safety and efficacy.[5] However, for pSS complicated with lymphoma, the effect of rituximab remains unclear. Whether early use of rituximab can improve the prognosis of patients with pSS-ML is worth further investigation. There are several limitations in our study. First, this is a case-control study and therefore prone to selection bias. Besides, we failed to analyze death rate and causes of death in these cohorts. A larger prospective cohort study or controlled trials are required to confirm the characteristics and risk factors of pSS-ML patients. In conclusion, NHL was the most frequent pathologic type, in which MALT was the most common in pSS-ML patients. Higher levels of ESSDAI, IgA and Th1-TNF-α, anemia, and lymphocytopenia significantly increased risk of ML in pSS patients. Complicated lymphoma had a negative influence on survival time and disease prognosis. Funding This study was supported by the grants from the National Natural Science Foundation of China and by the Beijing Sci-Tech Program (No. Z191100006619114), the Natural Science Foundation of China (No. 81701607), and Peking University People's Hospital Scientific Research Development Funds (No. RDZH2022-03). Conflicts of interest None.