Measures in assessMent of pediatric systeMic lupus erythematosus : an experience of retrospective observational study

Systemic lupus erythematosus in children (juvenile-onset SLE, jSLE) is a multisystemic disease with an unpredictable course and a more severe phenotype compared to adults. The patterns of jSLE are extremely heterogeneous, so an enrollment to controlled studies may be rather complicated. Due to this problem and some additional ones, there are no standards for treatment of jSLE yet. The attending physician is fully responsible for the induction and maintenance therapeutical options including durability and aggressiveness. Objectives: finding of jSLE individual course’s features prognostically connected with the disease outcome. Methods: 45 children admitted to the SPbGPMU hospital with the systemic lupus erythematosus diagnosed at the age of 4-17 years were enrolled in this retrospective study. Primary SLE manifestations, the activity of disease according to SELENA-SLEDAI and ECLAM scales during initial treatment period and flares after it, the fact of remission achievement in 6 months were evaluated in each patient. Results: a few organ involvements were considered to be connected with outcome’s characteristics, for example lupus nephritis and early disease oncet are unfavorable predictive factors. The positive connection of favorable outcome with cyclophosphamide, intravenous methylprednisolone and mycophenolate mofetil was found; the negative connection between initial disease activity and flares after induction treatment was also noticed. Conclusion: the patient with initially high disease activity treated aggressively with high cumulative doses of cyclophosphamide, intravenous methylprednisolone and mycophenolate mofetil has more chances of the favorable outcome (the achievement of remission without further flares).


INTRODUC TION
Approximately 10-20% of systemic lupus erythematosus (SLE) cases occur during childhood.Despite differences in design, cohort studies performed in the past few decades suggested that the course of SLE initiated in childhood is associated with greater severity, higher rate of disease activity and degree of organ damage compared with the course of SLE manifested in older age [1][2][3].
There is a significant difference in the frequency of organ involvement and clinical manifestations between adults and children of different age groups.Moreover, there are differences in the use of various drugs suppressing the disease activity.In general, pediatric patients receive more aggressive treatment [6].Currently, there is a lack of established recommndations for the treatment of SLE in children.
The number of randomized trials with qualitative design, investigating SLE in children, is limited.Hence, currently the main source of data on SLE is research conducted on adult populations.In addition, the wide variability of the course of SLE makes it difficult to study homogeneous patient groups.Therefore, the results obtained and the treatment recommendations derived from those findings are specific to the involvement of a particular organ (e. g., treatment of lupus nephritis or central nervous system lesion) or complications of the disease (macrophage activation syndrome) [3].The choice of therapy takes into account a variety of individual factors such as organ involvement, age, heredity, aggressiveness of the disease, as well as the prognosis and risk of exacerbation in remission, and remains the prerogative of the treating physician.
The objective of this study was to determine the characteristics of an individual scenario of SLE in a deliberately selected heterogeneous group of pediatric patients.

M ATERIA L S A ND ME THODS
A retrospective study including 45 patients with various forms of SLE and onset of disease ≤18 years of age.The patients were hospitalized in the Pediatric Department No. 3 of the Federal State Budgetary Educational Institution of Higher Education St. Petersburg State Pediatric Medical University, Ministry of Health of the Russian Federation.Diagnosis of SLE was es-tablished based on the American College of Rheumatology (ACR) 1997 criteria.For each patient, organ involvement at the time of disease onset, and laboratory (including immunological) characteristics at the time of therapy initiation were determined.The concept of initial therapy (remission induction), traditionally recommended for the treatment of lupus nephritis, was introduced to limit the time frame.The average duration of induction therapy was 6 months [5].
Indices such as the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) as modified in the Safety of Estrogens in Lupus National Assessment (SELENA) SLEDAI study and the European Consensus Lupus Activity Measurement (ECLAM) were used to assess the activity of the disease.Evaluation was conducted at the time of initiation and termination of induction therapy.
The therapy involved cumulative doses of cyclophosphamide and intravenous methylprednisolone (pulse therapy) per 1 m 2 of body surface and 1 kg of body weight, respectively; doses of intravenous immunoglobulin and rituximab; and use of other agents controlling disease activity such as mycophenolate mofetil, oral glucocorticosteroids, azathioprine, and hydroxychloroquine.
Achievement of remission based on the SLEDAI (clinical and laboratory), clinical remission based on the ECLAM index, minimal clinically significant improvement in activity scores (-7 points for SELENA SLEDAI and -3 points for ECLAM) [9], and the presence of exacerbations of the disease after the initial treatment period, according to the SELENA flare index (SFI) during the observation period, were used as indicators of treatment efficacy.

STATISTIC AL METHODS
Statistical analysis was performed using the statistical software packages Statistica 10.0 and Microsoft Excel.The Spearman rank correlation coefficient was used to assess correlations between clinical and laboratory findings as well as with the indices of treatment efficacy.The reliability of the differences between independent groups formed according to the presence of correlations was verified using the non-parametric Mann-Whitney U-test.For categorical quantities, the two-tailed exact Fisher's test was used.

RESU LT S
The majority of patients enrolled in this study were females (84.4%).The median age of disease onset was 13 years (range: 4-17 years).The most frequent organ involvement was skin and mucous membrane lesion (71.1%) and arthritis (55.6%), while the most frequent laboratory finding (after antinuclear factor and antibodies to double-stranded DNA) was anemia (51.1%).
Most patients (80%) received oral corticosteroids during initial therapy.The efficacy of treatment was evaluated 6 months (on average) after treatment initiation.One patient expired during the first month of observation due to disease severity.
As a result of treatment, remission was achieved in 12 to 21 patients (evaluated according to two different scales of activity).Disease exacerbations were noted after the initial therapy period (taking into account the entire observation period) in 13 patients (28.9%).The detailed demographic characteristics of the study population are provided in Table 1.
The Spearman rank correlation coefficient was used to assess the relationship between indicators within the study group.All characteristics correlated to the efficacy of therapy are presented in Table 2. Patients were classified into groups based on these parameters and the significance of the differences observed was determined using analysis of variance (Tables 3 and 4).Mutual correlations of various remission indicators were self-evident and were not tested further.
According to the results of the analysis, a significantly unfavorable prognostic factor for achieving two of the three types of remission (clinical remission based on the SLEDAI and clinical remission based on the ECLAM) is renal damage.Other prognostically signifi- cant factors at disease onset are sparse and noted sporadically.The presence of hepato-or splenomegaly during disease onset is associated with exacerbations after the initial therapy period.Lymphopenia, the presence of anti-Sm antibodies, and fever were shown to be prognostically favorable, whereas skin lesions were prognostically unfavorable.
The groups of patients who achieved clinical and laboratory remission (SLEDAI and ECLAM) had an onset at a significantly older age (for clinical remission based on the SLEDAI the correlation does not reach statistical significance).This may be due to both the "nonstandard" course of SLE in young children and the longer observation period of up to 18 years.
The results of the analysis showed a statistically significant negative association between exacerba-tions after induction and a baseline disease activity score on the SELENA SLEDAI scale.Further analysis revealed that a clinically significant improvement on the SELENA SLEDAI and ECLAM scales is also associated with initially higher activity scores but here the calculation technique plays a role as the initially low score does not change as much as the high score.In patients with induced clinical remission based on the ECLAM, exacerbations were significantly less frequent.
Those drugs which efficacy of therapy is associated (in different combinations and with varying degrees of significance) only with mycophenolate mofetil, cyclophosphamide and intravenous methylprednisolone (for the last two drugs, the association can be traced with both the fact of use and the cumulative dose).Other  drugs did not show significant correlations with therapy efficacy indicators.The aim of SLE therapy is to achieve remission or improvement based on the SLEDAI scale.Thus, a cumulative dose of cyclophosphamide plays a crucial role, whereas the number of methylprednisolone pulses and the use of mycophenolate mofetil appear to be less important.A criterion of therapy efficacy is the absence of disease exacerbations.Therefore, a child with initially high disease activity and clinical remission based on the ECLAM scale, receiving aggressive initial therapy with cyclophosphamide, may have a higher chance of a favorable outcome.

DISCUSSION
The heterogeneity of the clinical and laboratory manifestations of SLE limits the application of the "treat-to-target" principle i. e., treatment until the objective is reached.This principle has been useful for the management of other systemic diseases such as rheumatoid arthritis.According to the recommendations of the international working group, the aim of SLE therapy in adults is clinical remission or, if remission is not achievable, the minimum possible activity of the disease (estimated using a validated index or organ-specific markers such as the degree of proteinuria) [10].Currently, a single definition of remission for SLE is not available.According to the recommendations of the Definitions Of Remission In SLE (DORIS) group, remission is a long-term condition (without specific time limits) [11].Validated clinical indices are used to define remission (SLEDAI = 0; BILAG 2004 -only values of D or E; ECLAM = 0).The exact role of positive serology of SLE in the definition of remission (hypocomplementemia and antibodies to double-stranded DNA for the SLEDAI, and hypocomplementemia for ECLAM) has not been elucidated and requires further investigation.
In the present study, two variants of remission based on the SLEDAI were used, such as without serology (clinical remission) and clinical and laboratory remission (SLEDAI = 0).In remission based on the ECLAM,

◆ Information about the authors
◆ Информация об авторах hypocomplementemia was not considered.Since the "lowest possible achievable" value of indices was not determined, minimal clinically significant changes in the index were used [9].The absence of exacerbations is also considered by the international group as the target of SLE therapy.Although the timing and volume of maintenance therapy for lupus nephritis [7] are defined in the recommendations, these parameters remain at the discretion of the treating physician for the involvement of other organs.Processing data on maintenance therapy may be difficult due to concomitant administration of drugs, timing, doses, and frequent compliance disorders).Therefore, the outcomes of initial therapy were considered as predictors of disease exacerbation.
At present, there is a lack of evidence from controlled clinical studies supporting the advantages of the "treat-to-target" approach.The implementation of this strategy is hampered by the variability of the course of disease, as well as by the limited armamentarium of medications available to the treating rheumatologist [8].
The high toxicity of the most effective agents (cytostatic agents and parenteral methylprednisolone) is associated with irreversible organ damage, which is difficult to control in children and adolescents (usually due to the transition to an adult health care institution network).Aggressive therapy without taking into account the individual characteristics of the course of SLE in a child may be attractive in terms of achieving remission and control of exacerbations.However, this approach may have severe and unpredictable consequences in the long term.Therefore, further research into reliable organ-specific markers of disease activity, as well as the development of individualized therapy using modern genetic engineering drugs is warranted.

Table 1
Demographic characteristics of patients included in the study

Table 2
Coefficient of Spearman rank correlation for indices of efficiency of initial therapy (n = 45).The values noted are statistically significant (p < 0.05)

Table 3
Comparisons of independent groups: U-test (for quantitative variables).The values noted are statistically significant (p < 0,05)