Abstract
Objectives
Patients with primary Sjögren’s disease (SjD) have an increased risk of B cell lymphoma. The aim of this study was to determine serum protein biomarkers for lymphoma development and to advance our understanding of the functional mechanisms underlying lymphomagenesis in SjD.
Methods
Patients with SjD and incident, current lymphoma (n=18) with serum sampled before treatment and at 6, 12 and 24 months of follow-up, and four patients sampled 1–5 years before lymphoma diagnosis (pre-lymphoma) were included. SjD without lymphoma (n=21), SjD with historical lymphoma (n=6) and healthy blood donors (n=39) served as controls. Differentially expressed proteins between groups were analysed using the Olink Target 96 Immuno-Oncology panel applying a false discovery rate (FDR) adjusted p value of 0.05. Protein-derived interferon activation scores (pIFN scores) were calculated.
Results
We determined 18 differentially expressed proteins in SjD with incident lymphoma compared with both SjD without lymphoma and healthy controls. Among the top upregulated proteins were TNFSF14, FGF2, IL8, CD40 and CXCL13, where CXCL13 was the only protein with decreased levels at follow-up. We also observed upregulated expression of CD40 in the SjD pre-lymphoma group compared with SjD without lymphoma and healthy controls. All SjD patient groups presented elevated pIFN scores compared with healthy controls, where SjD sampled pre-lymphoma showed the most distinct IFN activation.
Conclusions
We identified altered protein expression and an increased IFN system activation in SjD with incident lymphoma and pre-lymphoma. This knowledge may contribute to earlier detection of high-risk patients, identification of therapeutic targets and may ultimately improve SjD patient outcomes.
Keywords: Sjogren's Syndrome, Biomarkers, Inflammation
WHAT IS ALREADY KNOWN ON THIS TOPIC
Patients with primary Sjögren’s disease (SjD) have an increased risk of B cell lymphoma.
Clinical and serological risk factors for lymphoma development have been identified, but more specific molecular risk factors are needed for patient stratification.
WHAT THIS STUDY ADDS
The study identifies distinct changes in serum protein expression in SjD with incident, untreated lymphoma compared with SjD without lymphoma and healthy controls.
Protein expression is also aberrant in SjD pre-lymphoma samples several years prior to lymphoma diagnosis.
Patients with SjD with incident lymphoma and pre-lymphoma display a prominently upregulated protein interferon score.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The study identifies several candidate protein biomarkers for identification of patients with SjD at high risk for developing lymphoma.
Some of the differentially expressed proteins are known drug targets, such as CD40, and may be suitable for intervention in precision medicine.
Increased knowledge of the functional mechanisms behind lymphomagenesis, availability of protein biomarkers and targeted therapies can contribute to improved patient care and outcomes.
Introduction
Patients with primary Sjögren’s disease (SjD) have an increased risk of lymphoma, most commonly marginal zone lymphomas of the B cell type, in the mucosa-associated lymphoid tissue (MALT). Diffuse large B cell lymphomas (DLBCL), nodal marginal zone lymphomas or other lymphoma subtypes are less frequent but may also occur.1 2 The MALT lymphomas in SjD arise preferentially in the salivary glands, the target tissue for inflammation in SjD, where infiltrating B cells undergo malignant transformation.3 Clinical and serological risk factors for lymphoma development have been identified, including major salivary gland swelling, purpura, cryoglobulinaemia, low complement protein C4, rheumatoid factor and a focus score of ≥3 or germinal centre formations in the minor salivary glands, reviewed by Chatzis et al.4 Clinical risk scores have been developed, but the majority of patients with SjD with these manifestations will never develop lymphoma, and more specific risk factors are needed.5 Cytokines promoting B cell stimulation have been linked to SjD lymphoma, a missense mutation in TNF alpha induced Protein 3 gene (TNFAIP3/A20) and dysregulated miRNAs in salivary gland biopsies have been described.6 One study, restricted to B cell activating genes, found an upregulation of Bruton’s tyrosine kinase (BTK) in patients with SjD with incident lymphoma.7 However, most studies are limited to historical lymphomas, precluding the investigation of molecular changes occurring at the time of lymphoma diagnosis.
A transcriptional upregulation of type I interferon (IFN) induced genes, a so-called IFN signature, is well established in different peripheral blood mononuclear cell types, most prominent in patients with SjD positive for Sjögren’s syndrome-related antigen A (SSA)/Ro and B (SSB)/La.8 A DNA methylation (DNAm) IFN score, based on hypomethylation of IFN-induced genes, and a protein IFN (pIFN) score, based on upregulation of IFN-induced proteins, have also been developed, both of which correlate with the mRNA IFN signature.9 One study found an upregulated whole blood DNAm IFN score in patients with SjD <70 years with lymphoma compared with patients with SjD without lymphoma.10 No clear association between an upregulated IFN signature in peripheral blood and lymphoma has hitherto been found.
Here, we hypothesise that the expression of serum proteins differs between patients with SjD and incident lymphoma compared with patients with SjD without lymphoma and healthy controls, especially regarding proteins linked to immune system activation and cancer development. Within the current study, targeted proteomics using the Olink Immuno-Oncology panel was employed with the aim of identifying potential serum protein biomarkers of early lymphoma in patients with SjD.
Materials and methods
Detailed methods are provided in online supplemental methods.
Patients and controls
Patients with SjD and incident, newly diagnosed and untreated current lymphoma were enrolled in the Autolymphoma study from several hospitals in Sweden, between 2010 and 2020. After proteomics quality control (QC), 18 patients with SjD and incident lymphoma were included, all with serum samples before treatment start (time point 0). 14 patients were sampled within 3 months of lymphoma diagnosis and four patients, where a ‘watch and wait’ strategy was employed, were sampled 7–13 months after lymphoma diagnosis (online supplemental table 1). Sera from four time points (tp) (0, 6, 12 and 24 months of follow-up) were available from 11 patients. For comparisons, three different groups of patients with SjD were included from the Rheumatology clinics at Uppsala and Linköping University Hospitals, Sweden: (1) patients with SjD without lymphoma (n=21), matched for age at SjD diagnosis, sex and SSA/SSB antibody frequency, with a median follow-up time of 11 years from SjD diagnosis until 2024 or death (range 4–41 years) and with follow-up samples from a second time point (n=15, median time between samplings 5 years, range 1–11). (2) Four patients with SjD (three unique and one also included in the SjD incident lymphoma group), with sera sampled before lymphoma diagnosis (pre-lymphoma) (median 3.5 years, range 1–5) from serum sampling to lymphoma diagnosis. (3) Historical lymphomas, six patients with SjD with a previous lymphoma (median 12.5 years, range 5–19, from lymphoma diagnosis to serum sampling) (online supplemental table 2, online supplemental figure 1). Clinical data were extracted from the medical records. The lymphomas were initially diagnosed by the local haematopathologist according to the WHO classification of Tumours of Haematopoietic and Lymphoid Tissues and, if needed, sent for second opinion to the Department of Pathology in Uppsala11 (table 1 and online supplemental table 3). All patients fulfilled the 2016 American College of Rheumatology/European League Against Rheumatism classification criteria for SjD.12 Serum samples from 39 blood donors (male, n=9 (23%); female, n=30 (77%); median age 70 years, range 26–83) from the Uppsala Bioresource, Uppsala University Hospital served as healthy controls.
Table 1. Clinical characteristics of patients with primary Sjögren’s disease (SjD).
| SjD with incident lymphoma | SjD without lymphoma | SjD with historical lymphoma | SjD sampled pre-lymphoma | |
|---|---|---|---|---|
| Individuals, n | 18 | 21 | 6 | 4 |
| Female, n (%) | 15 (83%) | 20 (95%) | 5 (83%) | 4 (100%) |
| Age at SjD diagnosis, years* | 48 (30–80) | 54 (25–85) | 50 (39–61) | 49 (44–64) |
| Age at lymphoma diagnosis, years† | 63 (29–84) | NA | 49 (36–58) | 69 (51–80) |
| Age at first serum sampling in study, years‡ | 64 (29–84) | 57 (32–85) | 61.5 (55–63) | 60 (50–73) |
| Disease duration from SjD diagnosis to first serum sampling in study, years§ | 6 (-1–36) | 1 (0–24) | 10 (2–23) | 7.5 (0–23) |
| Disease duration from SjD diagnosis until 2024 or death¶, years | 16 (0–44) | 11 (4–41) | 27 (19–32) | 22 (10–40) |
| Autoantibody frequency, n/available (%) | ||||
| SSA/Ro | 14/18 (78%) | 17/21 (81%) | 6/6 (100%) | 4/4 (100%) |
| Ro52 | 5/7 (71%) | 15/20 (75%) | 1/2 (50%) | 1/1 (100%) |
| Ro60 | 5/7 (71%) | 15/20 (75%) | 2/2 (100%) | 1/1 (100%) |
| SSB/La | 11/17 (65%) | 9/21 (43%) | 6/6 (100%) | 4/4 (100%) |
| ANA | 14/18 (78%) | 17/21 (81%) | 6/6 (100%) | 4/4 (100%) |
| RF | 7/8 (88%) | 10/17 (59%) | 5/6 (83%) | 1/1 (100%) |
| Clinical manifestations, n/available (%)** | ||||
| Focus score ≥1 | 10/10 (100%) | 16/17 (94%) | 1/1 (100%) | 3/3 (100%) |
| Focus score, median (range) | 3 (1–6) | 2 (1–12) | 10 | 3 (1–5) |
| Major salivary gland swelling†† | 7/18 (39%) | 8/21 (38%) | 5/6 (83%) | 2/4 (50%) |
| Lymphadenopathy†† | 3/18 (17%) | 1/21 (5%) | 4/6 (67%) | 0/4 (0%) |
| Raynaud’s | 6/18 (33%) | 12/21 (57%) | 2/6 (33%) | 1/4 (25%) |
| Arthritis | 5/18 (28%) | 6/21 (29%) | 2/6 (33%) | 2/4 (50%) |
| Purpura | 4/18 (22%) | 4/21 (19%) | 2/6 (33%) | 1/4 (25%) |
| Laboratory findings‡‡ | ||||
| Leucopenia <4.0×109/L | 4/16 (25%) | 11/21 (52%) | 4/6 (67%) | 3/4 (75%) |
| Lymphopenia <1.0×109/L | 3/7 (43%) | 3/10 (30%) | 1/5 (20%) | 1/2 (50%) |
| P-IgG >15 g/L | 5/10 (50%) | 13/19 (68%) | 5/6 (83%) | 4/4 (100%) |
| Monoclonal component | 1/9 (11%) | 0/19 (0%) | 0/6 (0%) | 2/4 (50%) |
| Cryoglobulinemia | 0/2 (0%) | 0/1 (0%) | 0/3 (0%) | 0/2 (0%) |
| Low C4 | 2/3 (67%) | 4/6 (67%) | 1/2 (50%) | 0/1 (0%) |
| Lymphoma subtypes | ||||
| MALT§§ | 11/18 (61%) | NA | 5/6 (83%) | 1/4 (25%) |
| DLBCL | 1/18 (6%) | NA | 0/6 (0%) | 1/4 (25%) |
| Other¶¶ | 6/18 (33%) | NA | 1/6 (17%) | 2/4 (50%) |
Consecutive values are presented as median (range).
P value calculated with Kruskal-Wallis test for comparison between the groups with p=0.85.
P value calculated with Kruskal-Wallis test for comparison between the groups with p=0.12.
P value calculated with Kruskal-Wallis test for comparison between the groups with p=0.70.
P value calculated with Kruskal-Wallis test for comparison between the groups with p=0.07.
Five patients died before 2024: SjD with incident lymphoma (2015, 2016), SjD without lymphoma (2022), SjD with historical lymphoma (2015), SjD pre lymphoma (2011).
Ever during the disease course.
In SjD with incident lymphoma: >1 year before lymphoma diagnosis; others: ever during the disease course.
In SjD with incident lymphoma: >6 months prior to lymphoma diagnosis; others: ever during the disease course.
MALT localisation: parotid gland, n=8; palate, n=1; lung, n=1; ventricle, n=1.
In SjD with incident lymphoma: follicular lymphoma, n=1; Hodgkin’s lymphoma, n=1; nodal marginal zone lymphoma, n=3; splenic marginal zone lymphoma, n=1; in SjD with historical lymphoma: chronic lymphocytic leukemia/small lymphocyte lymphoma, n=1; in SjD with pre-lymphoma: lymphoplasmatic lymphoma, n=1; plasma cell myeloma, n=1.
ANA, antinuclear antibodies; DLBCL, diffuse large B cell lymphomas; MALT, mucosa-associated lymphoid tissue; RF, rheumatoid factor; SSA, Sjögren’s syndrome-related antigen A; SSB, Sjögren’s syndrome-related antigen B.
Proteomic analysis
Expression of 92 proteins in sera was interrogated using the Olink Target Immuno-Oncology panel (Olink Proteomics AB, Uppsala, Sweden) based on proximity extension assay technology at the Clinical Biomarkers Facility in Uppsala, Sweden.13 Results were reported as normalised protein expression (NPX) values. NPX is a relative quantification of protein expression between samples, given on a log2-scale, where a high NPX value corresponds to a high protein concentration in the sample. For each analysed protein, the limit of detection (LOD) was determined based on three times SD beyond the median NPX of the negative controls. Protein assays with >30% of samples below LOD were excluded from further analyses. The post-QC dataset comprised 143 samples from 87 unique individuals and 81 proteins (online supplemental table 4). Differences in relative protein expression between sample groups were expressed in deltaNPX (dNPX).
Protein-derived type I IFN score
NPX values of CXCL9, CXCL10 and PDCD1 were used to calculate protein-derived type I IFN system activation scores (pIFN scores).9 Briefly, the expression (NPX) mean and SD of each protein in the control group were used to achieve a standardised value (Z-score) for the respective protein in each individual sample (Z-score=(valuesample−meancontrols)/SDcontrols). The sum of Z-scores of the three proteins is the pIFN scores for each sample. By definition, control samples had a mean pIFN score of 0. A high pIFN score was classified as >4.4 (mean pIFN scorecontrols+2SDcontrols) and pIFN scores ≤4.4 were defined as low.
Statistical analysis
Data were assumed to be non-parametric unless normality was demonstrated. A two-sided Mann-Whitney test was applied to test differential protein expression between two independent groups. A Kruskal-Wallis test with post hoc Dunn’s test was performed to compare three or more independent groups. A two-sided paired Wilcoxon signed-rank test was used to compare two dependent groups. Unless specified otherwise, the Benjamini-Hochberg method was used to account for multiple testing defining significant differential protein expression at an adjusted p value of <0.05. Statistical analyses were conducted in R (V.4.0.4, r-project.org/) and GraphPad Prism V.9. For the preparation of cluster heat maps, MORPHEUS (Broad Institute, software.broadinstitute.org/morpheus/) was used, while other graphs were prepared in R V.4.04.
Results
The 18 patients with SjD and incident lymphoma consisted of 15 (83%) women and three (17%) men. The median age at SjD diagnosis was 48 years (range 30–80) and the median disease duration until lymphoma diagnosis was 6 years (range −1 to 36). Antibodies against SSA/Ro and SSB/La were present in 78% and 65%, respectively. Clinical manifestations of the study cohort are presented in table 1. MALT was the most common lymphoma subtype (n=11, 61%) in the group of patients with SjD and incident lymphoma (table 1 and online supplemental table 3).
First, we performed a principal component analysis including the first serum sample from all individuals in the study and expression data of all 81 proteins remaining after QC (online supplemental table 4). The blood donor controls showed least variation and clustered largely separate from the different SjD patient groups (online supplemental figure 2). We found that SjD with incident lymphoma clustered separately from SjD without lymphoma and SjD with historical lymphoma, which were highly overlapping. SjD sampled pre-lymphoma located in a separate cluster.
Differential expression between SjD with incident lymphoma and SjD without lymphoma
To define biomarkers for lymphoma, we first asked which proteins were differentially expressed between SjD with incident, current lymphoma at time point 0 before treatment and SjD without lymphoma. 19 differentially expressed proteins (DEPs) were detected, 18 upregulated and one downregulated in SjD with incident lymphoma. TNF receptor superfamily member 14 (TNFSF14, LIGHT, herpes virus entry mediator-ligand (HVEM-L), CD258) and fibroblast growth factor 2 (FGF2) were the most significantly upregulated proteins (padj=1.6×10−3), whereas interleukin 8 (IL-8) displayed the largest difference in expression with a dNPX of 3.55, fold change (FC) 1.53. Among the upregulated DEPs were TNF receptor superfamily member 5 (TNFRSF5, CD40), TNFRSF9 (4-1BB, CD137), cytotoxic and regulatory T cell molecule (CRTAM), CD5, C-X-C motif chemokine 13 (CXCL13), TNFRSF4 (Ox40, CD134) and programmed cell death protein 1 (PDCD1, PD-1), while CXCL12 was the only downregulated DEP (table 2, figure 1A). Pathway analyses with the 81 proteins remaining after QC as background did not find any enriched pathways. When using the whole genome corresponding to the encoded proteins instead as background, we found the top enriched pathways to be cytokine activity and inflammatory response, padj <1×10−5 (online supplemental table 5).
Table 2. Differentially expressed proteins in the analysis between Sjögren’s disease (SjD) with incident lymphoma (n=18) and SjD without lymphoma (n=21).
| Protein | Protein name | Adjusted p value* | NPX SjD incident lymphoma, mean±SD | NPX SjD no lymphoma, mean±SD | NPX FC | dNPX |
|---|---|---|---|---|---|---|
| TNFSF14 | TNF ligand superfamily member 14 | 1.6×10−3 | 8.3±1.0 | 6.7±0.7 | 1.23 | 1.57 |
| FGF2 | Fibroblast growth factor 2 | 1.6×10−3 | 2.4±1.0 | 1.3±0.5 | 1.80 | 1.06 |
| CASP-8 | Caspase 8 | 1.7×10−3 | 6.5±1.2 | 4.8±0.7 | 1.37 | 1.77 |
| IL8, CXCL8 | Interleukin 8 | 1.9×10−3 | 10.3±2.9 | 6.8±0.7 | 1.53 | 3.55 |
| CD40 | TNF receptor superfamily member 5 | 2.0×10−3 | 11.3±0.7 | 10.5±0.3 | 1.08 | 0.81 |
| CCL3 | C-C motif chemokine 3 | 2.2×10−3 | 9.1±1.3 | 7.5±0.8 | 1.21 | 1.55 |
| ADA | Adenosine deaminase | 2.2×10−3 | 6.7±0.7 | 5.9±0.5 | 1.14 | 0.80 |
| EGF | Pro-epidermal growth factor | 5.1×10−3 | 10.5±0.8 | 9.7±0.9 | 1.08 | 0.76 |
| ARG1 | Arginase 1 | 5.2×10−3 | 6.4±1.1 | 4.9±1.1 | 1.31 | 1.50 |
| TNFRSF9 | TNF receptor superfamily member 9 | 5.2×10−3 | 8.8±1.3 | 7.5±0.7 | 1.17 | 1.28 |
| CRTAM | Cytotoxic and regulatory T cell molecule | 1.2×10−2 | 7.8±0.6 | 7.0±0.6 | 1.11 | 0.78 |
| CCL17 | C-C motif chemokine 17 | 1.5×10−2 | 12.4±0.5 | 11.6±0.9 | 1.07 | 0.80 |
| CD5 | T cell surface glycoprotein CD5 | 1.6×10–2 | 7.3±0.5 | 6.8±0.5 | 1.08 | 0.54 |
| MUC-16 | Mucin 16 | 3.1×10–2 | 4.5±0.9 | 3.6±0.9 | 1.23 | 0.85 |
| CXCL13 | C-X-C motif chemokine 13 | 3.4×10–2 | 11.2±0.8 | 10.3±1.0 | 1.09 | 0.95 |
| PDCD1 | Programmed cell death protein 1 | 3.4×10–2 | 7.0±0.9 | 6.2±0.7 | 1.13 | 0.82 |
| CXCL12 | Stromal cell-derived factor 1 | 3.4×10−2 | 2.2±0.3 | 2.6±0.4 | 0.87 | −0.33 |
| TNFRSF4 | TNF receptor superfamily member 4 | 4.1×10−2 | 8.4±1.0 | 7.5±0.7 | 1.11 | 0.82 |
| IL18 | Interleukin 18 | 4.2×10−2 | 11.0±1.0 | 10.1±0.7 | 1.09 | 0.91 |
Benjamini-Hochberg adjusted p value (False discovery rate, FDR 0.05). Differential protein expression was assessed using a two-sided Mann-Whitney U test.
dNPX, deltaNPX referring to the difference in mean NPX between SjD with incident lymphoma and SjD without lymphoma; FC, relative protein expression fold-change; NPX, normalised protein expression referring to relative protein expression on a log2 scale.
Figure 1. Volcano plots showing results of differential protein expression analysis in sera (A) between SjD with incident lymphoma (sampling tp0 at lymphoma diagnosis before start of treatment) and SjD without lymphoma and (B) between SjD with incident lymphoma tp0 and blood donor controls. Mean differential relative protein expression (dNPX) between groups is presented on the x-axis, and p values are given on the −log10 scale on the y-axis. Differentially expressed proteins are located above the dashed grey horizontal line indicating the significance threshold (padjust <0.05), with upregulated expression in SjD with incident lymphoma indicated in magenta and downregulated expression in SjD with incident lymphoma indicated in blue. IFN, interferon; IL-8, interleukin 8; SjD, Sjögren’s disease; tp, time point.
We then hypothesised that patients with MALT lymphoma may constitute a more homogenous group, potentially with specific protein expression patterns. However, restricting the analysis to SjD with incident MALT lymphoma compared with SjD without lymphoma did not result in any additional DEPs (online supplemental table 6). Further, we compared SjD with incident lymphoma with the group of SjD with historical lymphoma. The results were similar to those in the comparison with the group of SjD without lymphoma (online supplemental table 7).
Differential expression between SjD with incident lymphoma and healthy controls
Next, we aimed to identify the proteins that differed in their relative expression between patients with SjD and incident lymphoma and healthy controls. We identified 59 DEPs, 57 upregulated and 2 downregulated in SjD with incident lymphoma (padjust<0.05), (online supplemental table 8, figure 1B). The most significantly upregulated DEP was CXCL13 (padjust=1.8×10−7, dNPX 2.07, FC 1.23). Other highly significantly upregulated DEP included CD27, PDCD1, CD40, IL-8, TNFSF14 and FGF2 (online supplemental table 8, figure 2A–G). CXCL12 was downregulated in the analyses of SjD with incident lymphoma compared with both SjD without lymphoma and healthy controls (figures1A, B 2H). In total, a set of 18 proteins was differentially expressed when comparing SjD with incident lymphoma with both SjD without lymphoma and healthy controls (table 2, figure 3, online supplemental table 8).
Figure 2. Boxplots with dotplots illustrating results of the analysis of differential protein expression between blood donor controls (blue), SjD without lymphoma (yellow) and SjD with incident lymphoma tp0 at lymphoma diagnosis before start of treatment (orange) for (A) CXCL13, (B) CD27, (C) PDCD1, (D) CD40, (E) IL-8, (F) TNFSF14, (G) FGF2 and (H) CXCL12. Normalised protein expression (NPX) levels referring to relative protein expression on a log2-scale are shown on the y-axis. Differential expression was assessed using a two-sided Mann-Whitney test with Benjamini-Hochberg adjustment. Boxes indicate IQR and median. Whiskers indicate total range. SjD, Sjögren’s disease; tp, time point.
Figure 3. Unsupervised hierarchical cluster heat map based on the expression of the 18 shared differentially expressed proteins in the analysis between SjD with incident lymphoma (tp0 at lymphoma diagnosis before start of treatment) and SjD without lymphoma, and between SjD with incident lymphoma and blood donor controls. Each column represents one sample. Clinical characteristics and protein interferon (pIFN) score status are presented in the upper panel. The heat map in the lower panel illustrates protein expression levels, each row representing one protein. The colour scale in the lower panel is based on z-score distribution of protein expression levels, −2 (blue) to 2 (red). Unsupervised hierarchical clustering was based on Ward’s method with Euclidean distance as metric. MALT, mucosa-associated lymphoid tissue; SSA, Sjögren’s syndrome-related antigen A; SSB, Sjögren’s syndrome-related antigen B; SjD, Sjögren’s disease; tp, time point.
Differential expression between SjD sampled pre-lymphoma and SjD without lymphoma
Changes in expression in the group of patients sampled several years before lymphoma onset (SjD pre-lymphoma) compared with SjD without lymphoma are of particular interest due to their potential value as early biomarkers that may aid in timely identification of lymphomagenesis. At nominal significance, 33 DEPs, all upregulated, were found in sera from SjD sampled pre-lymphoma compared with SjD without lymphoma (online supplemental table 9). The top significant proteins were CD5, CRTAM, CD70, CXCL11, CXCL1 and CD40. CXCL13 was upregulated in the pre-lymphoma patient samples as well. These 33 DEPs were also differentially expressed between the group of SjD sampled pre-lymphoma and healthy controls (online supplemental table 10, figure 4A–F)
Figure 4. Top differentially expressed proteins in SjD sampled pre-lymphoma. Boxplots with dotplots illustrating results of the analysis of differential protein expression between blood donor controls (blue), SjD without lymphoma (yellow), SjD with incident lymphoma (tp0 at lymphoma diagnosis before start of treatment; orange), SjD sampled pre-lymphoma (green) and SjD with historical lymphoma (purple) for (A) CD5, (B) CRTAM, (C) CD70, (D) CXCL11, (E) CXCL1 and (F) CD40. Normalised protein expression (NPX) levels referring to relative protein expression on a log2-scale are shown on the y-axis. Boxes indicate IQR and median. Whiskers indicate total range. Differential expression between the groups was assessed using a Kruskal-Wallis with post hoc Dunn’s test. P values are presented unadjusted with ***p<0.001, **p<0.01, *p<0.05. SjD, Sjögren’s disease; tp, time point.
Differential expression between SjD without lymphoma and healthy controls
We then sought to answer the question which proteins were differentially expressed between patients with SjD in general, that is, SjD without lymphoma compared with healthy controls. Adjusting for multiple testing, this analysis identified 32 DEPs, 30 upregulated and 2 downregulated (online supplemental table 11, online supplemental figure 3). Lymphocyte activating 3 gene protein (LAG3) was the top significant DEP (padjust=3.8×105, dNPX 1.06, FC 1.20). As in the analysis of SjD with incident lymphoma vs healthy controls, PDCD1, CXCL13 and CD27 were among the top upregulated DEP (all padjust<2.2×10−4). Contrary to the analyses of SjD with incident lymphoma, TNFSF14 was downregulated in SjD without lymphoma compared with healthy controls (figure 2F).
Expression across different time points
In pairwise analyses of consecutive serum samples that were taken over a period of 2 years, protein expression in the SjD incident lymphoma patients was compared at tp0–tp6, tp0–tp12 and tp0–tp24 months. The only protein that was differentially expressed at nominal significance in all three comparisons (punadjust<0.05) was CXCL13 with lower expression at the follow-up samplings compared with tp0 (figure 5). The decrease in CXCL13 was most prominent in rituximab and/or chemotherapy treated patients, while the expression did not change if a ‘watch-and-wait’ strategy was employed (online supplemental table 1, online supplemental figure 4). Investigating potential changes over time in patients with SjD without lymphoma, there were no significant changes in protein levels, indicating stability of protein expression (online supplemental table 12).
Figure 5. CXCL13 protein expression across different time points in SjD with incident lymphoma. Boxplot with dotplots illustrating results of the pairwise comparison of CXCL13 protein expression between tp0 (at lymphoma diagnosis before start of treatment) and tp6 (6 months), tp0 and tp12 (12 months) and tp0 and tp24 (24 months). Differential expression was assessed using a two-sided paired Wilcoxon test, where only individuals with samples present at both time points compared were included in the respective analysis. Unadjusted p values are presented. Normalised protein expression (NPX) levels referring to relative protein expression on a log2-scale are shown on the y-axis. Boxes indicate IQR and median. Whiskers indicate total range. SjD, Sjögren’s disease; tp, time point.
IFN system activation (pIFN score)
The IFN system is known to play a prominent role in the pathogenesis of SjD. Therefore, we aimed to further explore a potential impact of IFN system activation in the pathogenesis of lymphoma by applying a protein-based IFN score. Patients with SjD with incident lymphoma presented a mean pIFN score of 6.7±3.8 compared with 3.9±4.0 in SjD without lymphoma (p=0.069), but clearly elevated compared with healthy controls (mean pINF score=0 by definition, p=4.2×10−8) (figure 6A). Intriguingly, patients with SjD sampled pre-lymphoma showed a particularly high pIFN score (10.2±3.2), significantly increased compared with both SjD without lymphoma (p=0.034) and healthy controls (p=4.6×10−5). Patients with a historical lymphoma presented a mean pIFN score solely increased compared with healthy controls (5.7±2.8, p=9.7×10−4). On the group level, pIFN scores in SjD with incident lymphoma tended to be relatively stable over the 24 months follow-up period (tp0 vs tp6, tp0 vs tp12, tp0 vs tp24, p>0.1 in all comparisons, online supplemental figure 5A). For SjD without lymphoma with expression data available from a consecutive sampling, the pIFN score did not differ between the two time points (online supplemental figure 5B, p = 0.14).
Figure 6. Protein interferon (pIFN) score. (A) Boxplot with dotplot depicting pIFN score levels in blood donor controls (blue), SjD without lymphoma (yellow), SjD with incident lymphoma (tp0 at lymphoma diagnosis before start of treatment; orange), SjD sampled pre-lymphoma (green) and SjD with historical lymphoma (purple). Differences in pIFN score levels between the groups were tested using a Kruskal-Wallis with post hoc Dunn’s test. Nominally significant p values are marked in bold. Boxes indicate IQR and median. Whiskers indicate total range. The horizontal red dashed line represents the threshold to a high pIFN score, defined as meancontrols+2SDcontrols, here pIFN >4.4. (B) Stacked percentage barplot presenting frequencies of pIFN score status (high in orange and low in yellow) between the different groups. Fisher’s exact test was used to assess differences in distribution between the groups. P values compared with controls: SjD no lymphoma 0.0002, SjD incident lymphoma <0.0001, SjD pre-lymphoma <0.0001, SjD historical lymphoma 0.013. SjD, Sjögren’s disease; tp0, time point 0.
When dichotomising individuals into high (>4.4) or low (≤4.4) pIFN score, 72% of patients with SjD with incident lymphoma presented a high pIFN score, compared with 48% in SjD without lymphoma (p=0.22). Notably, all patients with SjD sampled pre-lymphoma presented a high pIFN score, while 50% of patients with SjD with historical lymphoma and 5% of healthy controls were classified as having high pIFN score (figure 6B). In addition, we found expression of type II IFN (IFN-γ) upregulated in all SjD sample groups compared with healthy controls, and comparing the group of SjD sampled pre-lymphoma with SjD without lymphoma (punadjust<0.01 for all comparisons, online supplemental figure 6).
Discussion
Lymphoma development is a major complication in SjD and the only cause of increased mortality compared with the general population.14 Despite efforts to define risk factors, there is still an unmet need for biomarkers that can facilitate timely detection of patients at high risk for lymphoma. Here, we present for the first time proteomic analyses in patients with SjD with incident, untreated lymphoma and in patients sampled one to five years before lymphoma onset, the pre-lymphoma group. Patients with SjD with incident lymphoma presented a set of 18 proteins with altered expression in sera compared with both SjD without lymphoma and healthy controls. Many of the upregulated proteins are part of immune checkpoints or chemokines involved in NFκB activation and ectopic germinal centre formation, features that have been linked to lymphomagenesis in SjD.15 16
Several TNF ligand and receptor superfamily members showed up-regulated expression in SjD with incident lymphoma; TNFSF14/LIGHT, TNFRSF4/Ox40, TNFRSF9/4-1BB and TNFRSF5/CD40. These transmembrane receptors also exist in soluble forms due to enzymatic cleavage or alternative splicing, where the soluble forms may act as decoy receptors or have broader effects on the immune system compared with direct cell-cell interactions.17 TNFSF14/LIGHT, TNFRSF4/Ox40 and TNFRSF9/4-1BB are expressed on activated T cells, which after binding to their cognate ligands on antigen presenting cells, promote T cell and B cell proliferation, activate NFκB and stimulate pro-inflammatory cytokine production.18 TNFSF14/LIGHT is also a ligand to the lymphotoxin-β receptor (LTβR) present on different cell types, promoting tertiary lymphoid organisation and germinal centre formation.19 High serum levels of TNFSF14 have previously been reported in a subgroup of patients with SjD characterised by an IFN signature. Ox40/Ox40L has been implicated in SjD pathogenesis and TNFSF9/4-1BB has been linked to SjD in a Mendelian randomisation study.18 20 21 To the best of our knowledge, this is the first report of upregulated expression of these molecules in sera from patients with SjD with lymphoma.
Of particular interest is the upregulated expression of TNFRSF5/CD40. The CD40-CD40L interaction promotes B cell activation, production of class-switched antibodies, is critically involved in germinal centre formation and has been detected in SjD minor salivary gland epithelial and inflammatory cell infiltrates.22 Recent clinical trials targeting CD40 or CD40L in SjD are promising, with improvement of both patient-reported outcomes and systemic disease activity.23 24 It is noteworthy that patients with SjD with historical lymphoma did not overexpress CD40, indicating that the upregulation in sera may be restricted to active pre-lymphoma and incident lymphoma.
PDCD1 (PD-1) expressed on T cells is an immune checkpoint with inhibitory properties negatively regulating T cell responses to B cells or tumours expressing PD-L1/PD-L2. We found PDCD1 to be consistently up-regulated in both SjD with incident lymphoma and SjD sampled pre-lymphoma, compared with SjD without lymphoma and healthy controls. PDCD1 was also overexpressed in SjD without lymphoma compared with healthy controls, which has previously been described, possibly as a control mechanism of autoimmunity.25 Treatment with immune checkpoint inhibitors of the PD-1/PD-L1 pathway enhances the immune response to various malignant diseases but can induce a SjD-like syndrome.26
Previous studies suggest an association between IFN system activation and lymphoma, although a direct link has not been proven.27 28 To assess type I IFN activation on the protein level, we calculated a pIFN score based on the expression of PDCD1, CXCL9 and CXCL10.9 All SjD sampled pre-lymphoma and nearly three quarters of the SjD with incident lymphoma presented a high pIFN score. A biological mechanism for the potential role of type I IFN in lymphomagenesis is the continuous stimulation of immune complexes consisting of SSA/SSB antibodies and nucleic acids, which through ligation with TLR7 leads to type I IFN protein synthesis. Type I IFN binds to the IFN alpha receptor present on many immune cells and elicits the transcription of type I IFN-induced genes, encoding proteins with broad immune stimulatory properties.29 Type I IFN-receptor blocking treatment with anifrolumab is currently in clinical trials in SjD and could be a treatment of choice in selected patients.30
The PDCD1, CXCL9 and CXCL10 genes are also induced by type II IFN (IFN-γ) and there is considerable overlap between type I and type II IFN signatures.8 31 We found upregulated expression of IFN-γ in all SjD patient groups compared with healthy controls. An increased IFN-γ/IFN-α mRNA ratio in minor salivary glands from patients with SjD and lymphoma has previously been demonstrated.32 Notably, although the majority of lymphomas are of the MALT type and typically localised in the parotid glands, a systemic inflammatory response is evident, reflected by the elevated levels of pro-inflammatory cytokines in sera. The precise role for IFN-γ in SjD lymphoma development is yet to be determined.
Several proteins involved in T cell responses, including those considered immune checkpoint proteins, were overexpressed already before lymphoma occurrence or at the time of lymphoma diagnosis. These include the T cell receptor CD5, its ligand CD72 and CD70 (a cognate ligand of CD27), as well as proteins discussed above such as PD-1 and its ligands, and the TNF superfamily proteins.33 Although a complex interaction exists, these proteins are all involved in regulation of immune responses, and the findings suggest a dysregulated immune microenvironment that might contribute to lymphomagenesis or to an inadequate immune response to the development of malignant cells.
The B cell attracting chemokine CXCL13 was overexpressed in both SjD with incident lymphoma and SjD sampled pre-lymphoma. Of note, CXCL13 was the only protein found to be downregulated at follow-up, primarily following rituximab and/or chemotherapy. This observation is in line with findings from a recent study of DLBCL in patients with rheumatoid arthritis, where CXCL13 was the only cytokine that showed decreased levels at 12 months follow-up after chemotherapy.34 In SjD salivary glands, CXCL13 is expressed by epithelial cells and its receptor, CXCR5, is expressed on the infiltrating B cells. The CXCL13-CXCR5 interaction is crucial for homing of B cells to lymphoid follicles and implicated in germinal centre formation.35 36 A hypothesis for lymphomagenesis in SjD salivary glands is that the chronic antigen stimulation of autoreactive B cells in germinal centres leads to their monoclonal expansion and malignant transformation.6 Serum levels of CXCL13 have been associated with the degree of salivary gland inflammation and lymphoma in patients with SjD.37,39 Together with our findings, these observations suggest CXCL13 as a potential biomarker for immune activation in SjD, particularly in the context of lymphomagenesis, and suggest it may also be sensitive to therapy.
Our study has several limitations. The cohort size is relatively small, particularly the group of patients with SjD sampled before lymphoma diagnosis, underscoring the need for replication in larger and independent cohorts. Multiple lymphoma subtypes are included, and except for MALT lymphoma, subgroup analyses were not feasible due to limited numbers. We did not compare our results to patients with or without other autoimmune diseases and lymphoma. However, MALT lymphoma in the parotid glands is uncommon in other patient groups apart from SjD.2 All included patients were of Swedish origin, and the results may not be applied directly to other ethnicities.
The use of a targeted Olink panel restricted to 92 proteins is another limitation that constrains the breadth of protein biomarker discovery in SjD lymphoma. Several proteins previously associated with SjD lymphoma, either on the genetic or protein level, such as TNFAIP3/A20, B cell activating factor (BAFF), Fms-like tyrosine kinase (FLT3), CCL11 and BTK, were not represented on the array.1638 40,42 To fully elucidate how the proteins identified in this study align with known SjD lymphoma risk markers, and to enable the identification of novel biomarkers, a large-scale discovery panel such as the Olink Explore HT 5400+ or the SomaLogic aptamer-based SomaScan 11k array, is desired.43 Nevertheless, in a targeted clinical biomarker study, smaller panels may still be relevant. For translation into clinical practice, validation and application of an ELISA is recommended.
The key strengths of our study are the inclusion of patients at the time of lymphoma diagnosis, before the start of lymphoma treatment and the availability of sera from patients with SjD sampled years before their lymphoma diagnosis. This pre-lymphoma group is of particular interest from a clinical perspective where therapeutic prevention might become possible in the future. Most previous studies have analysed patients with a history of lymphoma. In our study, we found that patients with historical lymphoma more than 5 years before serum sampling exhibited a proteomic profile similar to that of SjD without lymphoma. To define biomarkers for lymphoma and elucidate the processes driving lymphomagenesis, the availability of timely samples prior to treatment is crucial. Further integration with multi-omics data will help gain a comprehensive understanding of lymphomagenesis and biomarkers in SjD lymphoma.
In conclusion, we identified altered protein expression and increased IFN system activation in SjD with incident lymphoma as well as in SjD sampled pre-lymphoma. These findings may contribute to earlier detection of high-risk patients, improve our understanding of the mechanisms underlying lymphoma development and, through the identification of therapeutic targets, ultimately improve the clinical care of patients with SjD.
Supplementary material
Acknowledgements
We thank Rezvan Kiani Dehkordi and Marianne Petersson for collecting blood samples of the patients with SjD and Maija-Leena Eloranta for collecting samples from healthy blood donors. Proteomics analyses were performed at the Clinical Biomarkers Facility, Science for Life Laboratory at Uppsala University, part of the National Genomics Infrastructure (NGI) Sweden.
Footnotes
Funding: The project was supported financially by grants from the Swedish Cancer Foundation to EB (CAN #2013/456, #2016/322, #2018/757), the Swedish Research Council to GN (VR #2022-00637), the Swedish Rheumatism Association and the King Gustaf V's 80-year foundation to GN, EB and JI-K, and the Agnes and Mac Rudberg Foundation and Brunnberg Foundation, Uppsala University to CF.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study involves human participants and was approved. The study protocol was approved by the regional ethics board in Uppsala (number 2009/238) and carried out in compliance with the Declaration of Helsinki. All patients and controls gave written informed consent.
Data availability free text: Data are available on a collaborative basis.
Collaborators: The Autolymphoma Study Group: Eva Baecklund, chair (Department of Medical Sciences, Rheumatology, Uppsala University, Uppsala, Sweden). Karolinska Institutet, Stockholm: Johan Askling (Division of Clinical Epidemiology, Department of Medicine Solna), Karin Ekström Smedby (Division of Clinical Epidemiology, Department of Medicine Solna, Department of Hematology, Karolinska University Hospital), Eva Kimby (Department of Hematology), Lars Klareskog (Division of Rheumatology, Department of Medicine Solna), Richard Rosenquist Brandell (Department of Molecular Medicine and Surgery). Uppsala University: Rose-Marie Amini and Christer Sundström (Department of Immunology, Genetics and Pathology), Gunilla Enblad (Department of Immunology, Genetics and Pathology, Cancer Immunotherapy, and Cancer Precision Medicine, Department of Oncology, Uppsala University Hospital). Collaborators: Eskilstuna: Charlott Mörth (Department of Oncology, Västerås Hospital, Västerås, Sweden; previously Mälarsjukhuset, Eskilstuna). Falun: Max Flogegård (Department of Internal Medicine, Hematology, Falun General Hospital), Tomas Husmark and Jörgen Lysholm (Department of Internal Medicine, Rheumatology, Falun General Hospital). Karlstad: Karin Hallén (Department of Oncology), Csaba Balaspiri (Department of Rheumatology), Karlstad Hospital. Linköping: Ingemar Lagerlöf (Division of Drug Research, Department of Medical and Health Sciences, Linköping University; previously University Hospital Linköping, Oncology), Per Eriksson, Christopher Sjöwall (Department of Biomedical and Clinical Sciences, Linköping University). Malmö/Lund: Thomas Relander, Mats Jerkman, Mats Ehinger, Elisabeth Szekely, (Department of Clinical Sciences Lund, Section for Oncology and Pathology, Lund University and Department of Oncology, Skane University Hospital, Lund), Elisabet Lindqvist, Elke Theander, Peter Olsson, Tomas Mandl (Section of Rheumatology, Department of Clinical Sciences Lund University; Skane University Hospital, Lund). Sunderbyn/Luleå: Lena Brandefors (Department of Internal Medicine, Hematology, Sunderby Hospital), Maria Klosinska Linder and Katarina Wänkkö (Department of Internal Medicine, Rheumatology, Sunderby Hospital). Uppsala: Carin Backlin (Biology Education Centre, Uppsala University, previously Department of Medical Sciences, Rheumatology, Uppsala University), Alina Johansson (Department of Medical Sciences, Rheumatology, Uppsala University), Hans Hagberg (Department of Immunology, Genetics and Pathology, Uppsala University; Department of Oncology Uppsala University Hospital), Mattias Mattsson (Department of Immunology, Genetics and Pathology, Uppsala University; Department of Hematology, Uppsala University Hospital. Örebro: Urban Jerlström (Department of Oncology, Örebro University Hospital), Per Salomonsson (Department of Rheumatology, Örebro University Hospital).
Contributor Information
The Autolymphoma Study Group:
Eva Baecklund, Johan Askling, Karin Ekström Smedby, Eva Kimby, Lars Klareskog, Richard Rosenquist Brandell, Rose-Marie Amini, Christer Sundström, Gunilla Enblad, Charlott Mörth, Max Flogegård, Tomas Husmark, Jörgen Lysholm, Karin Hallén, Csaba Balaspiri, Ingemar Lagerlöf, Per Eriksson, Christopher Sjöwall, Thomas Relander, Mats Jerkman, Mats Ehinger, Elisabeth Szekely, Elisabet Lindqvist, Elke Theander, Peter Olsson, Tomas Mandl, Lena Brandefors, Maria Klosinska Linder, Katarina Wänkkö, Carin Backlin, Alina Johansson, Hans Hagberg, Mattias Mattsson, Urban Jerlström, and Per Salomonsson
Data availability statement
Data are available upon reasonable request.
References
- 1.Theander E, Henriksson G, Ljungberg O, et al. Lymphoma and other malignancies in primary Sjögren’s syndrome: a cohort study on cancer incidence and lymphoma predictors. Ann Rheum Dis. 2006;65:796–803. doi: 10.1136/ard.2005.041186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Vasaitis L, Nordmark G, Theander E, et al. Population-based study of patients with primary Sjögren’s syndrome and lymphoma: lymphoma subtypes, clinical characteristics, and gender differences. Scand J Rheumatol. 2020;49:225–32. doi: 10.1080/03009742.2019.1696403. [DOI] [PubMed] [Google Scholar]
- 3.Nocturne G, Mariette X. B cells in the pathogenesis of primary Sjögren syndrome. Nat Rev Rheumatol. 2018;14:133–45. doi: 10.1038/nrrheum.2018.1. [DOI] [PubMed] [Google Scholar]
- 4.Chatzis LG, Stergiou IE, Goules AV, et al. Clinical picture, outcome and predictive factors of lymphoma in primary Sjögren’s syndrome: results from a harmonized dataset (1981-2021) Rheumatology (Oxford) 2022;61:3576–85. doi: 10.1093/rheumatology/keab939. [DOI] [PubMed] [Google Scholar]
- 5.Fragkioudaki S, Mavragani CP, Moutsopoulos HM. Predicting the risk for lymphoma development in Sjogren syndrome: An easy tool for clinical use. Medicine (Baltimore) 2016;95:e3766. doi: 10.1097/MD.0000000000003766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Nocturne G, Pontarini E, Bombardieri M, et al. Lymphomas complicating primary Sjögren’s syndrome: from autoimmunity to lymphoma. Rheumatology (Oxford) 2021;60:3513–21. doi: 10.1093/rheumatology/kez052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Duret P-M, Schleiss C, Kawka L, et al. Association Between Bruton’s Tyrosine Kinase Gene Overexpression and Risk of Lymphoma in Primary Sjögren’s Syndrome. Arthritis Rheumatol . 2023;75:1798–811. doi: 10.1002/art.42550. [DOI] [PubMed] [Google Scholar]
- 8.Bodewes ILA, Al-Ali S, van Helden-Meeuwsen CG, et al. Systemic interferon type I and type II signatures in primary Sjögren’s syndrome reveal differences in biological disease activity. Rheumatology (Oxford) 2018;57:921–30. doi: 10.1093/rheumatology/kex490. [DOI] [PubMed] [Google Scholar]
- 9.Björk A, Richardsdotter Andersson E, Imgenberg-Kreuz J, et al. Protein and DNA methylation-based scores as surrogate markers for interferon system activation in patients with primary Sjögren’s syndrome. RMD Open. 2020;6:e000995. doi: 10.1136/rmdopen-2019-000995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Imgenberg-Kreuz J, Sandling JK, Norheim KB, et al. DNA Methylation-Based Interferon Scores Associate With Sub-Phenotypes in Primary Sjögren’s Syndrome. Front Immunol. 2021;12:702037. doi: 10.3389/fimmu.2021.702037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127:2375–90. doi: 10.1182/blood-2016-01-643569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Shiboski CH, Shiboski SC, Seror R, et al. 2016 American College of Rheumatology/European League Against Rheumatism classification criteria for primary Sjögren’s syndrome: A consensus and data-driven methodology involving three international patient cohorts. Ann Rheum Dis. 2017;76:9–16. doi: 10.1136/annrheumdis-2016-210571. [DOI] [PubMed] [Google Scholar]
- 13.Assarsson E, Lundberg M, Holmquist G, et al. Homogenous 96-plex PEA immunoassay exhibiting high sensitivity, specificity, and excellent scalability. PLoS One. 2014;9:e95192. doi: 10.1371/journal.pone.0095192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Theander E, Manthorpe R, Jacobsson LTH. Mortality and causes of death in primary Sjögren’s syndrome: a prospective cohort study. Arthritis Rheum. 2004;50:1262–9. doi: 10.1002/art.20176. [DOI] [PubMed] [Google Scholar]
- 15.Theander E, Vasaitis L, Baecklund E, et al. Lymphoid organisation in labial salivary gland biopsies is a possible predictor for the development of malignant lymphoma in primary Sjögren’s syndrome. Ann Rheum Dis. 2011;70:1363–8. doi: 10.1136/ard.2010.144782. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nocturne G, Tarn J, Boudaoud S, et al. Germline variation of TNFAIP3 in primary Sjögren’s syndrome-associated lymphoma. Ann Rheum Dis. 2016;75:780–3. doi: 10.1136/annrheumdis-2015-207731. [DOI] [PubMed] [Google Scholar]
- 17.Tang T, Cheng X, Truong B, et al. Molecular basis and therapeutic implications of CD40/CD40L immune checkpoint. Pharmacol Ther. 2021;219:107709. doi: 10.1016/j.pharmthera.2020.107709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Webb GJ, Hirschfield GM, Lane PJL. OX40, OX40L and Autoimmunity: a Comprehensive Review. Clin Rev Allergy Immunol. 2016;50:312–32. doi: 10.1007/s12016-015-8498-3. [DOI] [PubMed] [Google Scholar]
- 19.Shuptrine CW, Perez VM, Selitsky SR, et al. Shining a LIGHT on myeloid cell targeted immunotherapy. Eur J Cancer. 2023;187:147–60. doi: 10.1016/j.ejca.2023.03.040. [DOI] [PubMed] [Google Scholar]
- 20.James JA, Guthridge JM, Chen H, et al. Unique Sjögren’s syndrome patient subsets defined by molecular features. Rheumatology (Oxford) 2020;59:860–8. doi: 10.1093/rheumatology/kez335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shi W, Xu Y, Zhang A, et al. Inflammatory cytokines and their potential role in Sjogren’s syndrome risk: insights from a mendelian randomization study. Adv Rheumatol . 2024;64:14. doi: 10.1186/s42358-024-00354-2. [DOI] [PubMed] [Google Scholar]
- 22.Dimitriou ID, Kapsogeorgou EK, Moutsopoulos HM, et al. CD40 on salivary gland epithelial cells: high constitutive expression by cultured cells from Sjögren’s syndrome patients indicating their intrinsic activation. Clin Exp Immunol. 2002;127:386–92. doi: 10.1046/j.1365-2249.2002.01752.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fisher BA, Mariette X, Papas A, et al. Safety and efficacy of subcutaneous iscalimab (CFZ533) in two distinct populations of patients with Sjögren’s disease (TWINSS): week 24 results of a randomised, double-blind, placebo-controlled, phase 2b dose-ranging study. Lancet. 2024;404:540–53. doi: 10.1016/S0140-6736(24)01211-X. [DOI] [PubMed] [Google Scholar]
- 24.St. Clair EW, Baer AN, Ng W-F, et al. CD40 ligand antagonist dazodalibep in Sjögren’s disease: a randomized, double-blinded, placebo-controlled, phase 2 trial. Nat Med. 2024;30:1583–92. doi: 10.1038/s41591-024-03009-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hinrichs AC, Kruize AA, Lafeber FPJG, et al. CCR9/CXCR5 Co-Expressing CD4 T Cells Are Increased in Primary Sjögren’s Syndrome and Are Enriched in PD-1/ICOS-Expressing Effector T Cells. Int J Mol Sci. 2023;24:11952. doi: 10.3390/ijms241511952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bitoun S, Rousseau A, Gosset M, et al. Immune Checkpoint Inhibitor-induced Sicca Syndrome. Rheum Dis Clin North Am. 2024;50:291–300. doi: 10.1016/j.rdc.2024.02.004. [DOI] [PubMed] [Google Scholar]
- 27.Brkic Z, Versnel MA. Type I IFN signature in primary Sjögren’s syndrome patients. Expert Rev Clin Immunol. 2014;10:457–67. doi: 10.1586/1744666X.2014.876364. [DOI] [PubMed] [Google Scholar]
- 28.Nguyen Y, Beydon M, Gottenberg J-E, et al. Distinct Pathophysiologic Pathways Support Stratification of Sjögren’s Disease Based on Symptoms, Clinical, and Routine Biological Data. Arthritis Rheumatol . 2025;77:876–83. doi: 10.1002/art.43096. [DOI] [PubMed] [Google Scholar]
- 29.Nordmark G, Eloranta ML, Ronnblom L. Primary Sjögren’s syndrome and the type I interferon system. Curr Pharm Biotechnol. 2012;13:2054–62. doi: 10.2174/138920112802273290. [DOI] [PubMed] [Google Scholar]
- 30.Mathian A, Felten R, Alarcon-Riquelme ME, et al. Type 1 interferons: A target for immune-mediated inflammatory diseases (IMIDs) Joint Bone Spine. 2024;91:105627. doi: 10.1016/j.jbspin.2023.105627. [DOI] [PubMed] [Google Scholar]
- 31.Soret P, Le Dantec C, Desvaux E, et al. A new molecular classification to drive precision treatment strategies in primary Sjögren’s syndrome. Nat Commun. 2021;12:3523. doi: 10.1038/s41467-021-23472-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Nezos A, Gravani F, Tassidou A, et al. Type I and II interferon signatures in Sjogren’s syndrome pathogenesis: Contributions in distinct clinical phenotypes and Sjogren’s related lymphomagenesis. J Autoimmun. 2015;63:47–58. doi: 10.1016/j.jaut.2015.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Dong J, Zhang K, Hong J, et al. The multiple functions of CD5 in diseases related to immune disorders. Ann Med. 2025;57:2519682. doi: 10.1080/07853890.2025.2519682. [DOI] [PubMed] [Google Scholar]
- 34.Euler N, Hellbacher E, Klint EA, et al. Diffuse large B cell lymphoma in rheumatoid arthritis patients is associated with elevated B-cell driving factors including CXCL13. Clin Immunol. 2025;275:110476. doi: 10.1016/j.clim.2025.110476. [DOI] [PubMed] [Google Scholar]
- 35.Amft N, Curnow SJ, Scheel-Toellner D, et al. Ectopic expression of the B cell-attracting chemokine BCA-1 (CXCL13) on endothelial cells and within lymphoid follicles contributes to the establishment of germinal center-like structures in Sjögren’s syndrome. Arthritis Rheum. 2001;44:2633–41. doi: 10.1002/1529-0131(200111)44:11<2633::aid-art443>3.0.co;2-9. [DOI] [PubMed] [Google Scholar]
- 36.Salomonsson S, Larsson P, Tengnér P, et al. Expression of the B cell-attracting chemokine CXCL13 in the target organ and autoantibody production in ectopic lymphoid tissue in the chronic inflammatory disease Sjögren’s syndrome. Scand J Immunol. 2002;55:336–42. doi: 10.1046/j.1365-3083.2002.01058.x. [DOI] [PubMed] [Google Scholar]
- 37.Chatzis L, Goules AV, Stergiou IE, et al. Serum, but Not Saliva, CXCL13 Levels Associate With Infiltrating CXCL13+ Cells in the Minor Salivary Gland Lesions and Other Histologic Parameters in Patients With Sjögren’s Syndrome. Front Immunol. 2021;12:705079. doi: 10.3389/fimmu.2021.705079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nocturne G, Seror R, Fogel O, et al. CXCL13 and CCL11 Serum Levels and Lymphoma and Disease Activity in Primary Sjögren’s Syndrome. Arthritis Rheumatol . 2015;67:3226–33. doi: 10.1002/art.39315. [DOI] [PubMed] [Google Scholar]
- 39.Traianos EY, Locke J, Lendrem D, et al. Serum CXCL13 levels are associated with lymphoma risk and lymphoma occurrence in primary Sjögren’s syndrome. Rheumatol Int. 2020;40:541–8. doi: 10.1007/s00296-020-04524-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Quartuccio L, Salvin S, Fabris M, et al. BLyS upregulation in Sjogren’s syndrome associated with lymphoproliferative disorders, higher ESSDAI score and B-cell clonal expansion in the salivary glands. Rheumatology (Oxford) 2013;52:276–81. doi: 10.1093/rheumatology/kes180. [DOI] [PubMed] [Google Scholar]
- 41.Corneth OBJ, Verstappen GMP, Paulissen SMJ, et al. Enhanced Bruton’s Tyrosine Kinase Activity in Peripheral Blood B Lymphocytes From Patients With Autoimmune Disease. Arthritis Rheumatol . 2017;69:1313–24. doi: 10.1002/art.40059. [DOI] [PubMed] [Google Scholar]
- 42.Tobón GJ, Saraux A, Gottenberg J-E, et al. Role of Fms-like tyrosine kinase 3 ligand as a potential biologic marker of lymphoma in primary Sjögren’s syndrome. Arthritis Rheum. 2013;65:3218–27. doi: 10.1002/art.38129. [DOI] [PubMed] [Google Scholar]
- 43.Rooney MR, Chen J, Ballantyne CM, et al. Correlations Within and Between Highly Multiplexed Proteomic Assays of Human Plasma. Clin Chem. 2025;71:677–87. doi: 10.1093/clinchem/hvaf030. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon reasonable request.