Large language model-driven sentiment analysis for facilitating fibromyalgia diagnosis.

BACKGROUND: Fibromyalgia (FM) is a complex disorder with widespread pain and emotional distress, posing diagnostic challenges. FM patients show altered cognitive and emotional processing, with a preferential allocation of attention to pain-related information. This attentional bias towards pain cues can impair cognitive functions such as inhibitory control, affecting patients' ability to manage and express emotions. Sentiment analysis using large language models (LLMs) can provide insights by detecting nuances in pain expression. This study investigated whether open-source LLM-driven sentiment analysis could aid FM diagnosis. METHODS: 40 patients with FM, according to the 2016 American College of Rheumatology Criteria and 40 non-FM chronic pain controls referred to rheumatology clinics, were enrolled. Transcribed responses to questions on pain and sleep were machine translated to English and analysed by the LLM Mistral-7B-Instruct-v0.2 using prompt engineering targeting FM-associated language nuances for pain expression ('prompt-engineered') or an approach without this targeting ('ablated'). Accuracy, precision, recall, specificity and area under the receiver operating characteristic curve (AUROC) were calculated using rheumatologist diagnosis as ground truth. RESULTS: The prompt-engineered approach demonstrated accuracy of 0.87, precision of 0.92, recall of 0.84, specificity of 0.82 and AUROC of 0.86 for distinguishing FM. In comparison, the ablated approach had an accuracy of 0.76, precision of 0.75, recall of 0.77, specificity of 0.75 and AUROC of 0.76. The accuracy was superior to the ablated approach (McNemar's test p<0.001). CONCLUSION: This proof-of-concept study suggests LLM-driven sentiment analysis, especially with prompt engineering, may facilitate FM diagnosis by detecting subtle differences in pain expression. Further validation is warranted, particularly the inclusion of secondary FM patients.

Exploring relief for Behçet's disease refractory oral ulcers: a comparison of TNF inhibitors versus apremilast.

OBJECTIVES: Oral and genital ulcers are the hallmark manifestation of Behçet's disease (BD), significantly impacting patients' quality of life. Our study focuses on comparing the effectiveness and safety of TNF inhibitors (TNFis) and apremilast in controlling oral ulcers of BD, aiming to provide evidence-based guidance for physicians in selecting appropriate treatment modalities. METHODS: A retrospective analysis was performed on BD patients treated between December 2016 and December 2021 with TNFis or apremilast for refractory oral ulcers. The study assessed treatment response by the absence of oral ulcers at 3 and 6 months, with additional evaluations for genital ulcers and articular involvement. RESULTS: The study included 78 patients, equally allocated between TNFis and apremilast treatments. Both groups showed significant oral ulcer reduction at 3 (p< 0.001) and 6 months (p= 0.01) with no significant difference between the treatments. Apremilast had a notable corticosteroid-sparing effect by the 3-month follow-up, persisting through 6 months. Both treatments were equally effective in reducing genital ulcers, with TNFis showing greater effectiveness in addressing articular involvement. Apremilast had a higher discontinuation rate due to gastrointestinal side effects. CONCLUSION: TNFis and apremilast are both effective for treating BD refractory oral ulcers. While TNFis may offer broader benefits for other disease manifestations, apremilast is distinguished by its corticosteroid-sparing effect, especially for patients with a milder disease phenotype. Treatment selection should consider individual disease severity and clinical features to ensure a personalized and effective management strategy.

Prompt engineering: The next big skill in rheumatology research.

Large language models (LLMs) like GPT-4 and Claude are catalyzing transformation across medical research including rheumatology. This review examines their applications, highlighting the pivotal role of prompt engineering in effectively guiding LLMs. Key aspects explored include literature synthesis, data analysis, manuscript drafting, coding assistance, privacy considerations, and generative artificial intelligence integrations. While LLMs accelerate workflows, reliance without apt prompting jeopardizes accuracy. By methodically constructing prompts and gauging model outputs, researchers can maximize relevance and utility. Locally run open-source models also offer data privacy protections. As LLMs permeate rheumatology research, developing expertise in strategic prompting and assessing model limitations is critical. With proper oversight, LLMs markedly boost scholarly productivity.

Safety profile of recombinant adjuvanted anti-herpes zoster vaccine (RZV) in high-risk groups: Data from active surveillance program. Puglia (Italy), 2021-23.

BACKGROUND: Since 2021 a recombinant adjuvanted anti-Herpes Zoster vaccine(Recombinant Zoster Vaccine, RZV) is offered in Italy to high-risk patients. Few real-life data about RZV safety are available in target populations. OBJECTIVES: This study investigates Adverse Events Following Immunization(AEFIs), baseline disease flare-ups, and Herpes Zoster (HZ) episodes occurring after RZV administration in a heterogeneous population of fragile patients to design its safety profile. METHODS: This is a retrospective population-based study. RZV-vaccinated patients at Bari Policlinico General Hospital vaccination clinic from October 1st, 2021, to March 31st, 2023, were enrolled. Subjects were screened for reason of RZV eligibility and baseline chronic pathologies. AEFIs occurred in the first 7-days post-vaccination period were collected, and baseline disease flare-ups and post-vaccination HZ episodes were assessed via a 3-month follow-up. RESULTS: Five-hundred-thirty-eight patients were included and total of 1,031 doses were administered. Most patients were vaccinated due to ongoing immunosuppressive therapy(54.65 %); onco-hematological and cardiovascular conditions were the most common chronic baseline pathologies. Out of 1,031 follow-ups, 441 AEFI cases were reported(42.7/100). The most common symptoms were injection site pain/itching(35.60/100), asthenia/malaise(11.44/100), and fever (10.09/100). Four serious AEFIs occurred(0.38/100). Older age, male sex, and history of cardiovascular diseases(OR:0.71; 95CI:0.52-0.98; p-value <0.05) were found to decrease AEFIs risk, while endocrine-metabolic illnesses(OR:1.61; 95CI:1.15-2.26; p-value <0.05) increased it. Twelve patients(2.23 %) reported a flare-up/worsening of their baseline chronic condition within the first three months after vaccination(mean interval 31.75 days, range 0-68 days). Patients with rheumatological illnesses had a higher risk of relapse(OR:16.56; 95CI:3.58-76.56; p-value <0.001), while male sex behaved as a protective factor. Twelve patients who completed the vaccination cycle(2.43%) had at least one HZ episode by the long-term follow-up. CONCLUSIONS: The study demonstrates RZV safety in a significant number of high-risk patients. Hence, RZV should be actively offered as part of tailored vaccination programs to decrease the burden of HZ in fragile populations.

Explainable artificial intelligence for microbiome data analysis in colorectal cancer biomarker identification.

Background: Colorectal cancer (CRC) is a type of tumor caused by the uncontrolled growth of cells in the mucosa lining the last part of the intestine. Emerging evidence underscores an association between CRC and gut microbiome dysbiosis. The high mortality rate of this cancer has made it necessary to develop new early diagnostic methods. Machine learning (ML) techniques can represent a solution to evaluate the interaction between intestinal microbiota and host physiology. Through explained artificial intelligence (XAI) it is possible to evaluate the individual contributions of microbial taxonomic markers for each subject. Our work also implements the Shapley Method Additive Explanations (SHAP) algorithm to identify for each subject which parameters are important in the context of CRC. Results: The proposed study aimed to implement an explainable artificial intelligence framework using both gut microbiota data and demographic information from subjects to classify a cohort of control subjects from those with CRC. Our analysis revealed an association between gut microbiota and this disease. We compared three machine learning algorithms, and the Random Forest (RF) algorithm emerged as the best classifier, with a precision of 0.729 ± 0.038 and an area under the Precision-Recall curve of 0.668 ± 0.016. Additionally, SHAP analysis highlighted the most crucial variables in the model's decision-making, facilitating the identification of specific bacteria linked to CRC. Our results confirmed the role of certain bacteria, such as Fusobacterium, Peptostreptococcus, and Parvimonas, whose abundance appears notably associated with the disease, as well as bacteria whose presence is linked to a non-diseased state. Discussion: These findings emphasizes the potential of leveraging gut microbiota data within an explainable AI framework for CRC classification. The significant association observed aligns with existing knowledge. The precision exhibited by the RF algorithm reinforces its suitability for such classification tasks. The SHAP analysis not only enhanced interpretability but identified specific bacteria crucial in CRC determination. This approach opens avenues for targeted interventions based on microbial signatures. Further exploration is warranted to deepen our understanding of the intricate interplay between microbiota and health, providing insights for refined diagnostic and therapeutic strategies.

Unraveling the microbiome-metabolome nexus: a comprehensive study protocol for personalized management of Behçet's disease using explainable artificial intelligence.

The presented study protocol outlines a comprehensive investigation into the interplay among the human microbiota, volatilome, and disease biomarkers, with a specific focus on Behçet's disease (BD) using methods based on explainable artificial intelligence. The protocol is structured in three phases. During the initial three-month clinical study, participants will be divided into control and experimental groups. The experimental groups will receive a soluble fiber-based dietary supplement alongside standard therapy. Data collection will encompass oral and fecal microbiota, breath samples, clinical characteristics, laboratory parameters, and dietary habits. The subsequent biological data analysis will involve gas chromatography, mass spectrometry, and metagenetic analysis to examine the volatilome and microbiota composition of salivary and fecal samples. Additionally, chemical characterization of breath samples will be performed. The third phase introduces Explainable Artificial Intelligence (XAI) for the analysis of the collected data. This novel approach aims to evaluate eubiosis and dysbiosis conditions, identify markers associated with BD, dietary habits, and the supplement. Primary objectives include establishing correlations between microbiota, volatilome, phenotypic BD characteristics, and identifying patient groups with shared features. The study aims to identify taxonomic units and metabolic markers predicting clinical outcomes, assess the supplement's impact, and investigate the relationship between dietary habits and patient outcomes. This protocol contributes to understanding the microbiome's role in health and disease and pioneers an XAI-driven approach for personalized BD management. With 70 recruited BD patients, XAI algorithms will analyze multi-modal clinical data, potentially revolutionizing BD management and paving the way for improved patient outcomes.

Beyond the horizon: Innovations and future directions in axial-spondyloarthritis.

Axial spondyloarthritis (axSpA) is a chronic inflammatory disease of the spine and sacroiliac joints. This review discusses recent advances across multiple scientific fields that promise to transform axSpA management. Traditionally, axSpA was considered an immune-mediated disease driven by human leukocyte antigen B27 (HLA-B27), interleukin (IL)-23/IL-17 signaling, biomechanics, and dysbiosis. Diagnosis relies on clinical features, laboratory tests, and imaging, particularly magnetic resonance imaging (MRI) nowadays. Management includes exercise, lifestyle changes, non-steroidal anti-inflammatory drugs and if this is not sufficient to achieve disease control also biological and targeted-synthetic disease modifying anti-rheumatic drugs. Beyond long-recognized genetic risks like HLA-B27, high-throughput sequencing has revealed intricate gene-environment interactions influencing dysbiosis, immune dysfunction, and aberrant bone remodeling. Elucidating these mechanisms promises screening approaches to enable early intervention. Advanced imaging is revolutionizing the assessment of axSpA's hallmark: sacroiliac bone-marrow edema indicating inflammation. Novel magnetic resonance imaging (MRI) techniques sensitively quantify disease activity, while machine learning automates complex analysis to improve diagnostic accuracy and monitoring. Hybrid imaging like synthetic MRI/computed tomography (CT) visualizes structural damage with new clarity. Meanwhile, microbiome analysis has uncovered gut ecosystem alterations that may initiate joint inflammation through HLA-B27 misfolding or immune subversion. Correcting dysbiosis represents an enticing treatment target. Moving forward, emerging techniques must augment patient care. Incorporating patient perspectives will be key to ensure innovations like genetics, microbiome, and imaging biomarkers translate into improved mobility, reduced pain, and increased quality of life. By integrating cutting-edge, multidisciplinary science with patients' lived experience, researchers can unlock the full potential of new technologies to deliver transformative outcomes. The future is bright for precision diagnosis, tightly controlled treatment, and even prevention of axSpA.

Fibrosing Progressive Interstitial Lung Disease in Rheumatoid Arthritis: A Multicentre Italian Study.

BACKGROUND: The INBUILD study demonstrated the efficacy of nintedanib in the treatment of progressive fibrosing interstitial lung disease different to idiopathic pulmonary fibrosis, including rheumatoid arthritis (RA)-related ILD. Nevertheless, the prevalence of RA-ILD patients that may potentially benefit from nintedanib remains unknown. OBJECTIVES AND METHODS: The aim of the present multicentre study was to investigate the prevalence and possible associated factors of fibrosing progressive patterns in a cross-sectional cohort of RA-ILD patients. RESULTS: One hundred and thirty-four RA-ILD patients with a diagnosis of RA-ILD, who were confirmed at high-resolution computed tomography and with a follow-up of at least 24 months, were enrolled. The patients were defined as having a progressive fibrosing ILD in case of a relative decline in forced vital capacity > 10% predicted and/or an increased extent of fibrotic changes on chest imaging in a 24-month period. Respiratory symptoms were excluded to reduce possible bias due to the retrospective interpretation of cough and dyspnea. According to radiologic features, ILD was classified as usual interstitial pneumonia (UIP) in 50.7% of patients, nonspecific interstitial pneumonia in 19.4%, and other patterns in 29.8%. Globally, a fibrosing progressive pattern was recorded in 36.6% of patients (48.5% of patients with a fibrosing pattern) with a significant association to the UIP pattern. CONCLUSION: We observed that more than a third of RA-ILD patients showed a fibrosing progressive pattern and might benefit from antifibrotic treatment. This study shows some limitations, such as the retrospective design. The exclusion of respiratory symptoms' evaluation might underestimate the prevalence of progressive lung disease but increases the value of results.

COVID-19 Vaccine Knowledge, Attitude, Acceptance and Hesitancy among Pregnancy and Breastfeeding: Systematic Review of Hospital-Based Studies.

The risk of unfavourable outcomes for SARS-CoV-2 infection is significant during pregnancy and breastfeeding. Vaccination is a safe and effective measure to lower this risk. This study aims at reviewing the literature concerning the anti-SARS-CoV-2 vaccine's acceptance/hesitancy among pregnant and breastfeeding women attending hospital facilities. A systematic review of literature was carried out. Hospital-based observational studies related to vaccination acceptance, hesitancy, knowledge and attitude among pregnant and breastfeeding women were included. Determinants of acceptance and hesitancy were investigated in detail. Quality assessment was done via the Johann Briggs Institute quality assessment tools. After literature search, 43 studies were included, 30 of which only focused on pregnant women (total sample 25,862 subjects). Sample size ranged from 109 to 7017 people. Acceptance of the SARS-CoV-2 vaccine ranged from 16% to 78.52%; vaccine hesitancy ranged between 91.4% and 24.5%. Fear of adverse events for either the woman, the child, or both, was the main driver for hesitancy. Other determinants of hesitancy included religious concerns, socioeconomic factors, inadequate information regarding the vaccine and lack of trust towards institutions. SARS-CoV-2 vaccine hesitancy in hospitalized pregnant women appears to be significant, and efforts for a more effective communication to these subjects are required.

Effectiveness of biological targeted therapies may discriminate seronegative from seropositive rheumatoid arthritis.

OBJECTIVE: To assess the real-world effectiveness of targeting biologic drugs (bDMARD) in rheumatoid arthritis (RA) patients negative for rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA). METHODS: We retrospectively selected 81 seronegative and 404 seropositive RA patients receiving treatment with abatacept, anti-tumor necrosis factor (TNF) alpha, or tocilizumab. Effectiveness was evaluated by analyzing drug survival using Kaplan-Meyer analysis over 10-year follow-up. Survival rates were compared by log rank test, and hazard ratios (HRs) of therapy discontinuation were estimated through multivariate Cox-regression. RESULTS: Clinical characteristics were similar between the two groups, except for a significantly higher percentage of inadequate responders to prior bDMARDs in the seronegative RA patients (p= 0.02). Among seronegative RA, tocilizumab demonstrated a survival rate of 73.9% with a mean survival time (MST) of 76.8 months (95% CI 61-92), which was significantly higher than abatacept (37.5%, MST 37.1 months (95% CI 22-51; p= 0.01). Anti-TNF alpha therapy fell in the middle (50.0%, MST 63.5 months (95% CI 47-79) but the difference was not significant. Nevertheless, seropositive RA patients did not show significantly different drug survival rates. Negative predictors of drug discontinuation were RF/ACPA positivity (HR 0.56) and sex male (HR 0.58), but treatment with abatacept (HR 1.88) or anti-TNF alpha (HR 1.79), no co-therapy with cDMARD (HR 1.74), absence of bone erosions (HR 1.41), and higher HAQ (HR 1.58) were positive predictors. CONCLUSIONS: To confirm these preliminary findings and to explore the hypothesis of a distinctive therapeutic algorithm in seronegative RA, prospective studies on larger cohorts are needed.

AI am a rheumatologist: a practical primer to large language models for rheumatologists.

Natural language processing (NLP), a subclass of artificial intelligence, large language models (LLMs), and its latest applications, such as Generative Pre-trained Transformers (GPT), ChatGPT, or LLAMA, have recently become one of the most discussed topics. Up to now, artificial intelligence and NLP ultimately impacted several areas, such as finance, economics and diagnostic/scoring systems in healthcare. Another area that artificial intelligence has affected and will continue to affect increasingly is academic life. This narrative review will define NLP, LLMs and their applications, discuss the opportunities and challenges that components of academic society will experience in rheumatology, and discuss the impact of NLP and LLMs in rheumatology healthcare.

Rheumatoid Arthritis from Easy to Complex Disease: From the "2022 GISEA International Symposium".

Rheumatoid Arthritis (RA) is a systemic disease with many different clinical phenotypes. RA could be classified according to disease duration, seropositivity for rheumatoid factor (RF) and/or anti-citrullinated protein antibodies (ACPA), joint subtype, clinical behaviourbehavior and many other subgroups. In this review, we summarize and discuss the multifaceted aspects of RA, focusing on the relationship between autoimmunity status and clinical outcome, achievement of remission and influence on treatment response, from the 2022 International GISEA/OEG Symposium.

Uncovering the Underworld of Axial Spondyloarthritis.

Axial spondyloarthritis (axial-SpA) is a multifactorial disease characterized by inflammation in sacroiliac joints and spine, bone reabsorption, and aberrant bone deposition, which may lead to ankylosis. Disease pathogenesis depends on genetic, immunological, mechanical, and bioenvironmental factors. HLA-B27 represents the most important genetic factor, although the disease may also develop in its absence. This MHC class I molecule has been deeply studied from a molecular point of view. Different theories, including the arthritogenic peptide, the unfolded protein response, and HLA-B27 homodimers formation, have been proposed to explain its role. From an immunological point of view, a complex interplay between the innate and adaptive immune system is involved in disease onset. Unlike other systemic autoimmune diseases, the innate immune system in axial-SpA has a crucial role marked by abnormal activity of innate immune cells, including γδ T cells, type 3 innate lymphoid cells, neutrophils, and mucosal-associated invariant T cells, at tissue-specific sites prone to the disease. On the other hand, a T cell adaptive response would seem involved in axial-SpA pathogenesis as emphasized by several studies focusing on TCR low clonal heterogeneity and clonal expansions as well as an interindividual sharing of CD4/8 T cell receptors. As a result of this immune dysregulation, several proinflammatory molecules are produced following the activation of tangled intracellular pathways involved in pathomechanisms of axial-SpA. This review aims to expand the current understanding of axial-SpA pathogenesis, pointing out novel molecular mechanisms leading to disease development and to further investigate potential therapeutic targets.

Immunogenicity and Safety of Adjuvanted Recombinant Zoster Vaccine in Rheumatoid Arthritis Patients on Anti-Cellular Biologic Agents or JAK Inhibitors: A Prospective Observational Study.

Rheumatoid arthritis (RA) patients on JAK inhibitors (JAKi) have an increased HZ risk compared to those on biologic DMARDs (bDMARDs). Recently, the Adjuvanted Recombinant Zoster Vaccine (RZV) became available worldwide, showing good effectiveness in patients with inflammatory arthritis. Nevertheless, direct evidence of the immunogenicity of such a vaccine in those on JAKi or anti-cellular bDMARDs is still lacking. This prospective study aimed to assess RZV immunogenicity and safety in RA patients receiving JAKi or anti-cellular bDMARDs that are known to lead to impaired immune response. Patients with classified RA according to ACR/EULAR 2010 criteria on different JAKi or anti-cellular biologics (namely, abatacept and rituximab) followed at the RA clinic of our tertiary center were prospectively observed. Patients received two shots of the RZV. Treatments were not discontinued. At the first and second shots, and one month after the second shot, from all patients with RA, a sample was collected and RZV immunogenicity was assessed and compared between the treatment groups and healthy controls (HCs) receiving RZV for routine vaccination. We also kept track of disease activity at different follow-up times. Fifty-two consecutive RA patients, 44 females (84.61%), with an average age (±SD) of 57.46 ± 11.64 years and mean disease duration of 80.80 ± 73.06 months, underwent complete RZV vaccination between February and June 2022 at our center. At the time of the second shot (1-month follow-up from baseline), anti-VZV IgG titer increased significantly in both groups with similar magnitude (bDMARDs: 2258.76 ± 897.07 mIU/mL; JAKi: 2059.19 ± 876.62 mIU/mL, p < 0.001 for both from baseline). At one-month follow-up from the second shot, anti-VZV IgG titers remained stable in the bDMARDs group (2347.46 ± 975.47) and increased significantly in the JAKi group (2582.65 ± 821.59 mIU/mL, p = 0.03); still, no difference was observed between groups comparing IgG levels at this follow-up time. No RA flare was recorded. No significant difference was shown among treatment groups and HCs. RZV immunogenicity is not impaired in RA patients on JAKi or anti-cellular bDMARDs. A single shot of RZV can lead to an anti-VZV immune response similar to HCs without discontinuing DMARDs.

Evolution of Rheumatoid-Arthritis-Associated Interstitial Lung Disease in Patients Treated with JAK Inhibitors: A Retrospective Exploratory Study.

BACKGROUND: The aim of this multicenter retrospective study was to investigate the effectiveness and safety of the available JAK-inhibitors (JAKi) in patients with rheumatoid arthritis (RA) and interstitial lung disease (ILD). METHODS: We retrospectively analyzed patients with classified RA and RA-ILD undergoing JAKi in 6 Italian tertiary centers from April 2018 to June 2022. We included patients with at least 6 months of active therapy and one high-resolution chest tomography (HRCT) carried out within 3 months of the start of JAKi treatment. The HRCT was then compared to the most recent one carried out within 3 months before the last available follow-up appointment. We also kept track of the pulmonary function tests. RESULTS: We included 43 patients with RA-ILD and 23 males (53.48%) with a median age (interquartile range, IQR) of 68.87 (61.46-75.78) treated with JAKi. The median follow-up was 19.1 months (11.03-34.43). The forced vital capacity remained stable in 22/28 (78.57%) patients, improved in 3/28 (10.71%) and worsened in 3/28 (10.71%). The diffusing capacity of lung for carbon monoxide showed a similar trend, remaining stable in 18/25 (72%) patients, improving in 2/25 (8%) and worsening in 5/25 (20%). The HRCT remained stable in 37/43 (86.05) cases, worsened in 4/43 (9.30%) and improved in the last 2 (4.65%). DISCUSSION: This study suggests that JAKi therapy might be a safe therapeutic option for patients with RA-ILD in a short-term follow-up.