Preventing organ damage in systemic lupus erythematosus: the impact of early biological treatment

A path to Glucocorticoid Stewardship: a critical review of clinical recommendations for the treatment of systemic lupus erythematosus

Glucocorticoids (GCs) have revolutionized the management of SLE, providing patients with rapid symptomatic relief and preventing flares when maintained at low dosages. However, there are increasing concerns over GC-associated adverse effects and organ damage, which decrease patients' quality of life (QOL) and increase healthcare costs. This highlights the need to balance effective GC use and minimize toxicity in patients with SLE. Herein, we provide an overview of the theoretical considerations and clinical evidence, in addition to the variations and similarities across nine national and eight international recommendations regarding the use of GCs across SLE manifestations and how these compare with real-world usage. In line with this, we propose possible actions toward the goal of GC Stewardship to improve the QOL for patients with lupus while managing the disease burden.

Adherence patterns to medications and their association with subsequent hospitalization in adult Egyptian patients with systemic lupus erythematosus

OBJECTIVE

The aim of this study was to evaluate the adherence pattern to medications and the influence of non-adherence on hospitalization rate among Egyptian patients who were diagnosed with systemic lupus erythematosus (SLE).

METHODS

This cross-sectional interview-based study was conducted in an Egyptian rheumatology outpatient clinic. Individuals with SLE who were prescribed oral medications were included in the study. The 19-item Compliance Questionnaire of Rheumatology (CQR) was employed to evaluate drug adherence. Patients were classified as non-adherent and adherent group based on their CQR score Sociodemographic variables, clinical and therapeutic data, and the frequency of outpatient visits and hospitalization were collected and compared between the two groups.

RESULTS

In total, there were 161 SLE patients who participated in the study with mean age 30.70 ± 9.62 years, 129 females and 32 males (F:M 5.4:1). Of them, 93 patients were assigned to the non-adherent group, whereas 68 patients were assigned to the adherent group. There was a statistically significant difference in terms of marital status (p = 0.009), employment (p = 0.032), and the frequency of physical activity (p = 0.045) between the two groups. The adherents had a higher rate of administration of antimalarials (88.2% vs 75.3%, p = 0.039). The frequency of recurrent fever was lower in adherents (23.5% vs 38.7%, p = 0.042), lymphadenopathy (11.8% vs 31.2%, p = 0.004), recurring oral ulcers (25% vs 50.5%, p = 0.001), and substantial hair loss (87.1% vs 98.5%, p = 0.009) in comparison to non-adherents. The number of yearly outpatient visits was considerably greater in adherents (p = 0.007), while the number of prior hospitalizations was higher in non-adherents (p < 0.001). Non-adherents believed that taking medicines disrupt patients' daily routine (p < 0.001) and were more afraid of the occurrence of an adverse drug reaction (p < 0.001).

CONCLUSIONS

Egyptian SLE patients have an alarmingly high percentage of medication non-adherence. There is a strong association between medication non-adherence and frequency of hospitalization and cumulative clinical manifestations.

Lupus nephritis or not? A simple and clinically friendly machine learning pipeline to help diagnosis of lupus nephritis

OBJECTIVE

Diagnosis of lupus nephritis (LN) is a complex process, which usually requires renal biopsy. We aim to establish a machine learning pipeline to help diagnosis of LN.

METHODS

A cohort of 681 systemic lupus erythematosus (SLE) patients without LN and 786 SLE patients with LN was established, and a total of 95 clinical, laboratory data and 17 meteorological indicators were collected. After tenfold cross-validation, the patients were divided into training set and test set. The features selected by collective feature selection method of mutual information (MI) and multisurf were used to construct the models of logistic regression, decision tree, random forest, naive Bayes, support vector machine (SVM), light gradient boosting (LGB), extreme gradient boosting (XGB), and artificial neural network (ANN), the models were compared and verified in post-analysis.

RESULTS

Collective feature selection method screens out antistreptolysin (ASO), retinol binding protein (RBP), lupus anticoagulant 1 (LA1), LA2, proteinuria and other features, and the hyperparameter optimized XGB (ROC: AUC = 0.995; PRC: AUC = 1.000, APS = 1.000; balance accuracy: 0.990) has the best performance, followed by LGB (ROC: AUC = 0.992; PRC: AUC = 0.997, APS = 0.977; balance accuracy: 0.957). The worst performance is naive Bayes model (ROC: AUC = 0.799; PRC: AUC = 0.822, APS = 0.823; balance accuracy: 0.693). In the composite feature importance bar plots, ASO, RF, Up/Ucr, and other features play important roles in LN.

CONCLUSION

We developed and validated a new and simple machine learning pathway for diagnosis of LN, especially the XGB model based on ASO, LA1, LA2, proteinuria, and other features screened out by collective feature selection.

Comparison of Contributors to Mortality Differences in SLE Patients with Different Initial Disease Activity: A Larger Multicenter Cohort Study

To explore the etiology of risk factors and quantify the mortality differences in systemic lupus erythematosus (SLE) patients with different initial disease activity. The Jiangsu Lupus database was established by collecting medical records from first-hospitalized SLE patients during 1999-2009 from 26 centers in Jiangsu province, China, and their survival status every five years. The initial SLEDAI scores [high (>12) vs. low-moderate (≤12)] differences in mortality attributable to risk factors were quantified using population attributable fraction (PAF), relative attributable risk (RAR) and adjusted relative risk (ARR). Among 2446 SLE patients, 83 and 176 deaths were observed in the low-moderate and high activity groups, with mortality rates of 7.7 and 14.0 per 1000 person years, respectively. Anemia was the leading contributor to mortality, with PAFs of 40.4 and 37.5 in the low-moderate and high activity groups, respectively, and explained 23.2% of the mortality differences with an ARR of 1.66 between the two groups. Cardiopulmonary involvement caused the highest PAFs in the low-moderate (20.5%) and high activity (13.6%) groups, explaining 18.3% of the mortality differences. The combination of anemia and cardiopulmonary involvement had the highest RAR, causing 39.8% of the mortality differences (ARR = 1.52) between the two groups. In addition, hypoalbuminemia and a decrease in the creatinine clearance rate accounted for 20-30% of deaths and explained 10-20% of the mortality differences between the two groups, while antimalarial drug nonuse accounted for about 35% of deaths and explained 3.6% of the mortality differences. Anemia, cardiopulmonary involvement and hypoalbuminemia may cause substantial mortality differences across disease activity states, suggesting additional strategies beyond disease activity assessment to monitor SLE outcomes.