Introduction: Enhanced B cell activity, particularly memory B cells have gained interest in evaluating response during therapies with biologics. CD27-IgD- double-negative (DN) B cells lacking the conventional memory marker CD27 are reported to be part of the memory compartment, however, only scarce data is available for rheumatoid arthritis (RA). We therefore focused on DN B cells in RA, studied their isotypes and modulation during interleukin-6 receptor (IL-6R) inhibition by tocilizumab (TCZ). Methods: DN B cells were phenotypically analyzed from 40 RA patients during TCZ at baseline week 12, week 24 and 1 year. A single B cell polymerase chain reaction (PCR) approach was used to study Ig receptors, V_H gene rearrangements and specific isotypes. Results: Phenotypic analysis showed a significantly expanded population of DN B cells in RA which contain a heterogeneous mixture of IgG-, IgA- and IgM-expressing cells with a clear dominance of IgG+ cells. DN B cells carry rearranged heavy chain gene sequences with a diversified mutational pattern consistent with memory B cells. In contrast to tumor necrosis factor alpha (TNF-α) inhibition, a significant reduction in mutational frequency of BCR gene rearrangements at week 12, 24 and 1 year (P <0.0001) was observed by in vivo IL-6R inhibition. These changes were observed for all BCR isotypes IgG, IgA and IgM at week 12, 24 and 1 year (P <0.0001). IgA-RF, IgA serum level and IgA+ DN B cells decreased significantly (P <0.05) at week 12 and week 24 during TCZ. Patients with a good European League Against Rheumatism (EULAR) response to TCZ had less DN B cells at baseline as compared to moderate responders (P = 0.006). Univariate logistic regression analysis revealed that the frequency of DN B cells at baseline is inversely correlated to a subsequent good EULAR response (P = 0.024) with an odds ratio of 1.48 (95% confidence interval as 1.05 to 2.06). Conclusions: In RA, the heterogeneous DN B cell compartment is expanded and dominated by IgG isotype. TCZ can modulate the mutational status of DN Ig isotype receptors over 1 year. Interestingly, the frequency of DN B cells in RA may serve as a baseline predictor of subsequent EULAR response to TCZ.

Objective.: Scleroderma renal crisis (SRC) is a severe life-threatening manifestation in patients with systemic sclerosis (SSc). However, the knowledge about risk factors for SRC is limited. We determined here the frequency of SRC and identified risk factors for the prediction of SRC.Methods.: Based on regular followup data from the German Network for Systemic Scleroderma, we used univariate and multivariate generalized estimating equations to analyze the association between clinical variables, SSc subsets, therapy [i.e., angiotensin-converting enzyme inhibitors (ACEi), corticosteroids], and the occurrence of SRC.Results.: Data of 2873 patients with 10,425 visits were available for analysis with a mean number of registry visits of 3.6 ± 2.8 and a mean time of followup of 3.6 ± 3.8 years. In total, 70 patients developed SRC (70/2873, 2.4%). Of these patients, 57.1% (40/70) were diagnosed with diffuse cutaneous SSc, 31.4% (22/70) with limited cutaneous SSc, and 11.4% (8/70) with SSc-overlap syndromes. Predictive independent factors with the highest probability for SRC were positive anti-RNA polymerase antibodies (RNAP), a history of proteinuria prior to SRC onset, diminished DLCO, and a history of hypertension. Interestingly, positive antitopoisomerase autoantibodies did not predict a higher risk for SRC. Further, patients with SRC were significantly more frequently treated with ACEi and corticosteroids without being independently associated with SRC.Conclusion.: In this cohort, SRC has become a rare complication. By far the highest risk for SRC was associated with the detection of anti-RNAP and proteinuria.

Autologous hematopoietic stem cell transplantation (aHSCT) is performed in patients with aggressive forms of systemic sclerosis (SSc). The profile of B cell reconstitution after aHSCT is not fully understood. The aim of this study was to investigate changes of B cell subsets and cytokine production of B cells in patients with SSc after aHSCT. Peripheral blood of six patients with SSc was collected at defined intervals up to 16 months after aHSCT. Immunophenotyping was performed, and B cell function was determined by measuring cytokine secretion in supernatants of stimulated B cell cultures. Within 1 month after aHSCT, a peak in the percentage of CD38++/CD10+/IgD+ transitional B cells and CD38++/CD27++/IgD− plasmablasts was detected. Long-term changes persisted up to 14 months after aHSCT and showed an increased percentage of total B cells; the absolute B cell number did not change significantly. Within the B cell compartment, an increased CD27/IgD+ naïve B cell percentage was found whereas decreased percentages of CD27+/IgD+ pre-switched memory, CD27+/IgD− post-switched memory, and CD27−/IgD− double-negative B cells were seen after aHSCT. Cytokine secretion in B cell cultures showed significantly increased IL-10 concentrations 13 to 16 months after aHSCT. A changed composition of the B cell compartment is present for up to 14 months after aHSCT indicating positive persisting effects of aHSCT on B cell homeostasis. The cytokine secretion profile of B cells changes in the long term and shows an increased production of the immune regulatory cytokine IL-10 after aHSCT. These findings might promote the clinical improvements after aHSCT in SSc patients.

Background: Assessing serological inflammation is difficult in tocilizumab (TCZ)-treated rheumatoid arthritis (RA) patients, as standard inflammation parameters, like erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP), are influenced by interleukin-6-receptor inhibition. Calprotectin in the serum, also named S100A8/S100A9, might be a more useful inflammation parameter in TCZ-treated patients. Methods: Sixty-nine RA patients taking TCZ were included. Serum-calprotectin levels were assessed, as well as ESR, CRP, need for a change in disease-modifying anti-rheumatic drugs due to RA activity (= active RA), and the RA clinical disease activity score (CDAI). Forty-five RA patients taking tumor-necrosis factor-inhibitors (TNFi) were investigated for the same parameters. Results: TCZ-treated patients with active RA had higher calprotectin values than not active RA patients (4155.5 [inter quartile range 1865.3–6068.3] vs 1040.0 [676.0–1638.0] ng/ml, P < 0.001). A calprotectin cut-off value of 1916.5 ng/ml resulted in a sensitivity and specificity of 80.0 %, respectively, for the detection of RA disease activity. Calprotectin values correlated with CDAI-scores (r = 0.228; P = 0.011). ESR and CRP were less suitable to detect RA activity in TCZ-treated patients. Also TNFi-treated patients with active RA had higher calprotectin values compared to not active RA (5422.0 [3749.0–8150.8] vs 1845.0 [832.0–2569.0] ng/ml, P < 0.001). The calprotectin value with the best sensitivity and specificity for detecting RA activity was 3690.5 ng/ml among TNFi-treated patients. Conclusion: Calprotectin in the serum can be a useful inflammation parameter despite TCZ-treatment.

Background: Patients with axial spondyloarthritis (axSpA) benefit from regular home-based exercise (HbE). In spite of recommendations, a relevant proportion of German axSpA patients does not adhere to recommended HbE practices. To enhance HbE care, we developed the novel digital therapeutic (DTx) “Axia” compliant with the European medical device regulation (MDR). Axia offers a modern app-based HbE solution with patient educative content and further integrated features. Objective: We aimed to assess Axia’s efficacy, attractiveness, and functionality through a survey among axSpA-patients involved in the first user tests. Methods: A mixed-method online questionnaire with 38 items was administered to 37 axSpA volunteers after using Axia. Numeric rating scales (NRS) and likelihood scales were primarily used. Results: HbE frequency significantly increased from a median of 1 day/week to 6 days/week (p < 0.001) by using Axia. Existing HbE practitioners also increased their frequency (median of 4 days/week before, 6 days/week with Axia, p < 0.05). Axia received a median rating of 5 out of 5 stars. On NRS scales, Axia scored a median of 9 for intuitiveness and design, and a median of 8 for entertainment. 64.9% reported improved range of motion, 43.2% reported reduced pain, and 93.6% enhanced disease-specific knowledge. All users recommended Axia to other patients. Conclusion: Axia increases axSpA patients HbE frequency, possibly due to its good intuitiveness and design, leading to reduction in pain and subjective improvement of range of motion. This warrants further investigation in large randomized controlled interventional trials to establish its efficacy conclusively and patients adherence to HbE.

Background: The complex nature of rheumatic diseases poses considerable challenges for clinicians when developing individualized treatment plans. Large language models (LLMs) such as ChatGPT could enable treatment decision support. Objective: To compare treatment plans generated by ChatGPT-3.5 and GPT-4 to those of a clinical rheumatology board (RB). Design/methods: Fictional patient vignettes were created and GPT-3.5, GPT-4, and the RB were queried to provide respective first- and second-line treatment plans with underlying justifications. Four rheumatologists from different centers, blinded to the origin of treatment plans, selected the overall preferred treatment concept and assessed treatment plans’ safety, EULAR guideline adherence, medical adequacy, overall quality, justification of the treatment plans and their completeness as well as patient vignette difficulty using a 5-point Likert scale. Results: 20 fictional vignettes covering various rheumatic diseases and varying difficulty levels were assembled and a total of 160 ratings were assessed. In 68.8% (110/160) of cases, raters preferred the RB’s treatment plans over those generated by GPT-4 (16.3%; 26/160) and GPT-3.5 (15.0%; 24/160). GPT-4’s plans were chosen more frequently for first-line treatments compared to GPT-3.5. No significant safety differences were observed between RB and GPT-4’s first-line treatment plans. Rheumatologists’ plans received significantly higher ratings in guideline adherence, medical appropriateness, completeness and overall quality. Ratings did not correlate with the vignette difficulty. LLM-generated plans were notably longer and more detailed. Conclusion: GPT-4 and GPT-3.5 generated safe, high-quality treatment plans for rheumatic diseases, demonstrating promise in clinical decision support. Future research should investigate detailed standardized prompts and the impact of LLM usage on clinical decisions.