Predicting systemic lupus erithematosus disease activity index 2000 (Sledai 2k) score after pulse dose metyl prednisolone in severe systemic lupus erithematosus

Study design

This study is a prospective cohort observational analysis that examines data on patients with systemic lupus erythematosus (SLE) who met the criteria of the American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) 2019 or had class III or IV lupus nephritis (LN) according to World Health Organization (WHO) criteria. These patients underwent a follow-up period after receiving a daily dose of 500 mg MEP pulse for three consecutive days, during which their progress and response to the treatment were closely monitored and assessed.

Participants

The study included patients between the ages of 15 and 50 years who were willing to participate and did not have comorbidities such as pneumonia, tuberculosis, viral, fungal, parasitic, or protozoal infections, sepsis, diabetes mellitus, Graves’ disease, cardiovascular disease, or cancer. Additionally, patients were excluded if they had overlap syndrome with systemic sclerosis, rheumatoid arthritis, or dermatomyositis inflammation. Subjects were withdrawn from the study if they died or developed sepsis during the observation period. Patients who met the inclusion and exclusion criteria were informed about the study’s purpose, process, and potential risks and benefits, and asked to sign a consent form to participate. The explanation also highlighted the advantages and disadvantages of participating, allowing patients to make informed decisions about their involvement.

The determination of our study’s sample size was guided by the “rule of thumb” for multivariate analysis, which stipulates that a minimum of 15 samples per predictor is needed to ensure robust statistical analysis. Since we aimed to investigate five predictor variables in this study, the minimum required sample size was calculated as 75. To account for potential dropouts and missing data, which are common occurrences in research studies, we increased this number by an additional 10%. This precautionary measure resulted in a total minimum sample size of 83. By adhering to this method, we ensured that our sample was statistically adequate for the multivariate analysis of the five predictor variables.

Procedure

Before administering the pulse dose of MEP, all information on the SLE Disease Activity Index 2K (SLEDAI 2K) form was completed, and the SLEDAI 2K score was recorded.¹¹ Blood samples, typically ranging from 5 to 10 milliliters, were carefully collected from each patient using a standard venipuncture procedure. The samples were then processed and analyzed for the levels of sBAFF, IP-10, and BAFF-R. Specifically, the concentrations of sBAFF and IP-10 were measured using enzyme-linked immunosorbent assay (ELISA) techniques, while BAFF-R expression was analyzed through flow cytometry. Following the 3-day pulse dosage of MEP, blood samples were collected again using the same procedure, and these parameters were re-analyzed to assess and compare the changes in response to the treatment.

Measurement of sBAFF and IP-10 were measured using ELISA

S BAFF was examined in PRODIA LABORATORIES, CAP certified, using Microplate Reader Biorad model 680Bio-rad Laboratories inc, CA, USA. Reagent kit used in this study was Quantikine^® HS ELISA Human BAFF/BLyS/TNFSF13B (R&D Systems, Inc., Minneapolis, USA) according to standard protocols. The results were measured in pg/ml.

Measurement of R BAFF and B cells (CD19) using flowcytometry

R BAFF and B cells (CD 19+) was examined in Clinical Pathology Laboratories of Gadjah Mada University Faculty of Medicine, Nursing and Public Health, ISO 17025 certified, using BD FACS Canto II 8 color Flowcytometry. Reagent used was PE anti human CD19 antibody (CD 19+) and CD 45 PerCP cy5.5 according to standard protocols. The results were measured in Mean Fluoresence Index (MFI), percentage of B cells CD 19+ to total lymphocyte, percentage of R BAFF to B cells CD 19+.

The Urine Protein Creatinine Ratio (UPCR) was determined at baseline, prior to the administration of MEP, by dividing the level of spot urine protein by the level of spot urine creatinine. Urine protein and urine creatinine were estimated by ERBA CHEM 5 V PLUS. In this study, the cutoff for normal UPCR is less than 150 mg/g and more or equal to or less than 500 mg/gr, consistent with diagnostic criteria for LN according to EULAR/ACR classification criteria.

Statistical analysis

All statistical analyses were performed using R Studio v. 4.1.1.1, employing the tidyverse and gtsummary packages.^12,13 This study employed various statistical methods, including descriptive statistical analysis, univariate and multivariate logistic regression, and visualization using boxplots. Moreover, Pearson correlation analysis was conducted to evaluate the association between continuous variables by computing the Pearson correlation coefficient and its corresponding p-value, which provided insight into the magnitude and significance of any linear relationship.

In this study, we minimized potential biases through several strategies. To mitigate selection bias, clear inclusion and exclusion criteria were established, ensuring a representative study group. Measurement bias was addressed by using standardized assessment tools and established laboratory test methods. Multivariate analyses were conducted to control for confounding variables, such as age and sex. Attrition bias was tackled by carefully tracking all dropouts and withdrawals, and their potential impact on the study’s findings was thoroughly assessed. Additionally, all data and code were made openly available to promote transparency and reproducibility.

Ethical considerations

The sampling for this study was conducted from June 2021 to September 2022, after obtaining ethical approval from the Gadjah Mada University Faculty of Medicine, Nursing and Public Health ethics committee for human research (approval number: KE/FK/0359/EC/2021). The study was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent prior to their participation in the study.

In adherence to ethical guidelines, the inclusion of minor participants, aged between 15 to 17 years, in this study necessitated informed consent. In these cases, the consent was obtained through a two-step process: firstly, the minor’s assent was sought, which involved explaining the study’s purpose, procedures, potential risks, and benefits in an age-appropriate manner. Secondly, informed consent was also obtained from a parent or legal guardian. It was ensured that both the minor and their guardian fully understood the study before consent was provided. It was clarified that participation was voluntary, and they could withdraw from the study at any time without any negative consequences.

Correlation between routine laboratory parameters of SLE disease activity

Our study found a strong correlation between the SLEDAI 2K pre-dose MEP and various markers of disease activity, including CRP levels, urine casts, and renal activity parameters. An increase in CRP levels and the presence of urine casts were both indicative of high disease activity. Additionally, we observed a positive correlation between Protein Creatinine Ratio, casts in the urine, hematuria, leukocyturia, and renal involvement. Immunology tests also revealed correlations, such as a positive connection between C3 and C4 levels and a negative connection between anti-dsDNA and C3 and C4 levels. We also found a substantial positive correlation between hemoglobin and leucocyte levels, as well as between thrombocyte and leucocyte levels, regarding hematological involvement. However, we discovered a significant negative correlation between anti-dsDNA and C3, C4 levels, thrombocytes, and leukocytes. Additionally, we observed a significantly negative relationship between ESR and hemoglobin levels.

Our study found a significant positive correlation between the NLR and leucocytes, neutrophils, and CRP levels. This suggests that as the NLR level increases, leucocytes, neutrophils, and CRP levels also increase. On the other hand, we observed a significant negative correlation between the NLR and lymphocyte levels, indicating that the lymphocyte level decreases as the NLR level increases. This is in line with the calculation of the NLR, which is determined by dividing the neutrophil count by the lymphocyte count.

In the presence of systemic inflammation, it is common for leukocyte numbers to decrease and for there to be alterations in circulating leukocytes, with lymphopenia and neutrophilia being typical features. Despite this, the NLR is a more stable indicator of inflammatory processes than leukocytes. This is because leukocyte numbers can be more readily impacted by factors such as dehydration or overhydration, excessively dilute specimens, or the influence of the blood collection procedure.¹⁵

Correlation of S BAFF, MFI BAFF R, BAFF R (%) with parameters of SLE disease activity

The study by Zaki et al. aimed to investigate the association between BAFF R levels and SLE activity markers. The study revealed a negative association between BAFF R levels and SLEDAI, ESR, CRP, and anti-dsDNA levels.¹² Conversely, the study found a positive correlation between BAFF R levels and complement protein C3 and C4 levels (p < 0.001).¹² Additionally, the study observed a rapid reduction in BAFF R levels with the progression of the disease, with the lowest expression frequency observed in the group with the most active disease.¹⁶ Similar to this study, we also found a strong negative correlation between BAFF R levels and disease activity markers such as CRP, sBAFF, and ESR. In our study, the MFI of BAFF R levels showed a significant positive correlation with leucocyte count. At the same time, MFI BAFF R, but not BAFF R (%), had a significantly negative association with anti-dsDNA. This study found that sBAFF levels correlated positively with SLEDAI pre- and post-dose MEP, demonstrating that sBAFF levels increased during severe SLE flares and remained elevated following a pulse dose of MEP. Conversely, the study found a significant negative connection between sBAFF and proteinuria, contrasting with another study that found a positive correlation between sBAFF and proteinuria.¹⁷ The authors noted that the significant chronicity of lupus nephritis (LN) in the study population may have influenced proteinuria results. Moreover, the accuracy of actively proliferative LN in the study is questionable, as the classification of LN was based on World Health Organization criteria and not by biopsy.

Another study by Suso et al. observed high glomerular expression of BAFF in the inflammatory infiltrate of patients with class II mesangial and class III lupus nephritis in their histopathological analysis of LN.¹⁸ The authors noted that the high expression of BAFF was strongly associated with activity scores but not chronicity ratings, making this one of the most intriguing findings of their study.¹⁸

Correlation of IP 10 with parameters of SLE disease activity

There was a significantly positive correlation between IP 10 and CRP, NLR, and s BAFF. Qin et al. found significant positive correlations between IP 10 and CRP, NLR, and sBAFF.¹⁵ Specifically, there was a positive correlation between NLR and CRP (r = 0.471, p < 0.01), ESR (r = 0.610, p < 0.01), and SLEDAI (r = 0.471, p < 0.01). These findings suggest that IP 10 levels may be associated with markers of disease activity in systemic lupus erythematosus (SLE).

This study found that IP-10 levels were significantly negatively correlated with leucocyte, C-reactive protein (PCR), and BAFF R (%). This finding was consistent with the study by Hrycek et al., which showed a negative correlation between IP-10 levels and leucocyte (r = -0.423), polymorphonuclear cells (PMN, r = -0.303), lymphocyte (r = -0.386), and monocyte (r = -0.365) levels.¹⁹ However, the inverse association between IP-10 levels and PCR levels was inconsistent, as active lupus nephritis is expected to have higher PCR levels than inactive lupus nephritis.

A meta-analysis confirmed this finding by Puapatanakul et al., which showed that serum IP-10 levels in active non-renal lupus did not differ significantly from inactive lupus nephritis and non-renal systemic lupus erythematosus (mean difference 22.6 pg/mL, p = 0.83).²⁰ Another study found that IP-10 was a poor biomarker for predicting renal involvement.¹⁴ However, other investigations have reported conflicting findings. This discrepancy may be related to the varying levels of CXCR receptor expression on kidney-infiltrating cells. The current study’s authors hypothesized that the infiltration of IP-10 into renal tissue could increase its correlation with proteinuria in urine relative to serum levels.

Univariate and multivariate analysis correlation of variables with SLEDAI 2k post pulse dose MEP

The results of the univariate correlation test as shown in Table 3 indicated a strong positive association between the SLEDAI 2k, IP 10, and BAFF S variables prior to therapy and the post-therapy SLEDAI 2k score. This suggests that the post-therapy SLEDAI 2k score is proportional to the value of these variables prior to therapy. Additionally, the study found a non-significant negative connection between C3 levels and the post-therapy SLEDAI 2K score. This indicates that a lower C3 level is associated with a higher post-therapy SLEDAI 2K score. The correlation between the pre-and post-therapy SLEDAI 2K levels of CD19, C4, Hb, and platelets was modest and not statistically significant.

The study also analyzed the difference in mean post-SLEDAI 2K scores between clinical manifestation variables that were present and absent, including seizures, psychosis, SOO, visual disturbances, cranial nerve disorders, lupus headaches, CVA, vasculitis, arthritis, myositis, rash, alopecia, mouth ulcers, fever, hematuria, and proteinuria. The results revealed a significant difference in the mean post-therapy SLEDAI 2K score for arthritis.

A prediction model was generated using multivariate analysis with multiple linear regression after eliminating certain variables that were found to be correlated with each other (Table 3). This was done to ensure that the assumption criteria of multivariate linear regression analysis were met. The purpose of developing a formula that can predict SLEDAI 2K scores after MEP pulse dose is because SLE patients sometimes present with complicated conditions, the most common being flare-ups and infections. By predicting the SLEDAI score post-treatment, we can consider whether it will be beneficial or even worsen the infections before administering MEP pulses.

The resulting model was used to predict the response to MEP 500 mg (pulse) therapy for three days based on pre-therapy variables. The prediction formula was derived as follows: 13.41 + (0.0008542 * x1) - (0.1829338 * x2) + (0.0008776 * x3) + (7.1801728 * x4) + (7.5429676 * x5), where x1 is the UPCR (Urine Protein to Creatinine Ratio) prior to therapy, x2 is the CD19 (%) value prior to therapy, x3 is the sBAFF (B-cell activating factor) value prior to therapy, x4 represents the status of vasculitis prior to therapy (presumably coded as a binary value), x5 indicates the presence of a rash prior to therapy (also presumably coded as a binary value). This model is designed to predict the post-therapy SLEDAI 2K (Systemic Lupus Erythematosus Disease Activity Index 2000) score. The SLEDAI 2K score is a commonly used tool for assessing disease activity in patients with systemic lupus erythematosus. It incorporates a wide range of clinical and laboratory findings, and the prediction from the model above will be a crucial aspect of personalized treatment strategies for patients undergoing MEP 500 mg (pulse) therapy.