Carolyn Ochoa, MD1; Ellen Cahill, BA2; Dalya Chefitz, MD3; Daphne Esho, MD4; Jason Lewis, DO5, and Lakshmi N. Moorthy, MD, MPH6

1Division of Hospital Medicine, Department of Pediatrics, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

2 Rutgers Robert Wood Johnson Medical School, Piscataway, NJ

3 Division of Pediatric Hospital Medicine, Department of Pediatrics, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

4 Division of Pediatric Hospital Medicine, Department of Pediatrics, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

5 Division of Pediatric Hospital Medicine, Department of Pediatrics, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

6Division of Pediatric Rheumatology, Department of Pediatrics, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

ABSTRACT

Background: Infections are a serious complication of juvenile systemic lupus erythematosus (jSLE). Problems with innate immunity, adaptive immunity, and immunosuppression secondary to disease-controlling medications are likely contributing factors.

Objective: The following study was conducted in order
to: (1) determine the number and type of infections over a 5-year period in a cohort of jSLE patients, and (2) determine the relationships between the number of infections and potential risk factors such as complement levels and disease activity.

Methods: We reviewed the charts of 20 jSLE patients who were diagnosed before age 18 and were followed by Pediatric Rheumatology at Rutgers RobertWood Johnson Medical School between 1/2011-1/2016. Data collected included SLE Disease Activity Index (SLEDAI), stage of renal disease, complement levels, and physician-diagnosed-infection-related data over five years. Systemic Lupus International Collaborating Clinics (SLICC) Damage Index (SDI) was collected at the final visit.We compared the means between number of infections, several lab values (complement C3 and C4, serum albumin, absolute neutrophil count (ANC), white blood cell (WBC) count), medication usage, SLEDAI at diagnosis, and SDI score at the final visit.

Results: Twenty-one infections were seen in nine subjects, the majority being pneumonias (29%).While not statistically significant, patients with lower complement C3 at diagnosis had more infections over the course of the 5-year period. Additionally, in patients who used mycophenolate, cyclophosphamide, and/or rituximab, there was a trend towards more infections over the course of the 5-year period. A significant inverse correlation was noted between serum albumin level at diagnosis and infections within the first six months (r=0.5, p=0.04), and independent t-test revealed patients with lower albumin at diagnosis had significantly more infections within the first six months compared to those with higher albumin (p=0.02).

Conclusion: Patients with jSLE must be monitored closely for infection, and larger studies are needed to confirm the risk factors associated with increased infections as well as determine appropriate prevention strategies including antimicrobial prophylaxis and vaccination.

Keywords: juvenile-onset systemic lupus erythematosus, infection, complement, jSLE, pediatric SLE

INTRODUCTION

Juvenile systemic lupus erythematosus (jSLE), which accounts for about 10-20% of all cases of SLE, can cause multi-system inflammation and organ damage.1,2 Compared to adult-onset SLE, jSLE patients seem to have more severe organ involvement, necessitating early and aggressive treatment with immunosuppressive agents.2-4

Infection is one of the leading causes of morbidity and mortality in patients with jSLE.5-8 Intrinsic defects in innate and adaptive immunity and the use of immunosuppressive medications likely contribute to this risk.7,9,10 Increased disease activity at diagnosis, complement C3 less than 90 mg/dL at diagnosis, and renal and neuropsychiatric disease manifestations have been identified as factors associated with infection-related morbidity in patients with jSLE.7,9,10 Moreover, the use of immunosuppressive therapies contributes to the increased risk of infection. Severe infection morbidity was associated with use of cyclophosphamide and prednisone regardless of duration of disease.7 Common infections affecting jSLE patients include skin and soft tissue infections, pneumonia, and UTI.1,7,10

One large study identified the incidence of major infections (requiring parenteral antimicrobials or enteral antimicrobials for greater than a week) in jSLE patients over a 10-year period to be 169/1000 patient-years of follow-up.7 Complications of infections in jSLE included sepsis, macrophage activating syndrome, end organ damage, and death.7 The Systemic Lupus International Collaborating Clinics (SLICC) Damage Index (SDI) is a measurement of end organ damage caused by jSLE.11 Using the SDI, the investigators found infections were associated with damage (OR 3.2, CI =1.4-7.2; p = 0.004), and this association was even more significant for jSLE patients with infections that progressed to sepsis (OR 17.1, CI 2 – 144.3; p = 0.002).7 Similarly, another study found major infections to be associated with severity of disease damage in jSLE.8 A study conducted in Brazil of hospitalized patients with jSLE found the greatest risk of death to this population was severe sepsis (OR 17.8, CI = 4.5- 70.9).12 They found severe sepsis was associated with a 100-fold increase in mortality compared to the other identified risk factors including increase in disease activity, general infection, bacterial infections, fungal infections, acute thrombocytopenia, and acute renal failure.12

The risk of infection is a significant issue that must be addressed in the treatment of children with jSLE.The objectives for this study were to: (1) determine the number and types of infections over a 5-year period in a cohort of jSLE patients, and (2) determine the relationships between number of infections and potential risk factors including several lab values (complement C3 and C4, albumin, white blood cell (WBC) count, absolute neutrophil count (ANC)), medication use, SLE nephritis class, SLE damage (SDI), and SLE Disease Activity Index (SLEDAI).

METHODS

We retrospectively reviewed the charts of twenty patients diagnosed before age 18 with jSLE and followed by Pediatric Rheumatology at Rutgers RobertWood Johnson Medical
School between January 2011 and January 2016. Patients were diagnosed using the American College of Rheumatology (ACR) revised criteria.13 Patients with neonatal lupus, major congenital malformation/immunodeficiency, or HIV were excluded from
the final analysis. Data collected included SLEDAI [score range 0 – 65], stage of renal disease14, several lab values (complement C3 and C4, albumin, WBC count, ANC), medication usage (corticosteroids, mycophenolate, hydroxychloroquine, cyclophosphamide, rituximab), and infection-related data at the following intervals: time of diagnosis with jSLE, 6 months, 1 year, and yearly intervals up to five years. SDI score was collected at the final visit. An infection was defined as one that was diagnosed by a physician in either the outpatient or inpatient setting. SPSS version 25 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Initial descriptive analyses were performed on all variables. Individual variables and instrument scores (SLEDAI and SDI) were examined for ceiling and floor effects. All variables were examined to ascertain the data distributions and assess for normality. Based on data distribution, we conducted parametric and non-parametric tests.We used Spearman’s rank- order correlation, independent samples t-test, independent samples median test, and the Mann-Whitney U test to examine relationships between several risk factors and the number of infections.

This study was approved by the Rutgers Institutional Review Board.

RESULTS

3.1 Demographics
A total of 20 subjects with jSLE were identified, as shown in Table 1. Seventeen were female. The average age of diagnosis
was 13 ± 3 years.Ten (50%) subjects were Hispanic or Latino, 4 (20%) were Black or African American, 4 (20%) were Asian, and 2 (10%) were Caucasian.

Table 1. Demographics

TOTAL NUMBER OF PATIENTS (N = 20)

N (percentage %)

Gender:

Male

Female

 

3 (15)

17 (85)

Race/ethnicity:

Caucasian

Black or African American

Hispanic or Latino

Asian

 

2 (10)

4 (20)

 

10 (50)

4 (20)

Insurance

Private 8(40)

Medicaid 11(55)

Unknown 1 (5)

Age (years) at Diagnosis

Mean ± SD (range)

13 ± 3 (6-18)

School Grade at Diagnosis

Mean ± SD (range)

9 ± 2 (4-12)

Total number of Infections from

Jan 2011 – Jan 2016

21 infections in 9/20 subjects
Total number of Infections from diagnosis to 6 months 11 (52%)
Total number of Infections from diagnosis to 1 year 16 (76%)

 

The mean SLEDAI at presentation was 14.6 ± 9 (5 – 42) with a median of 12. Mean complement levels at diagnosis were: 50 ± 32 (5 -138, median 43) for C3 and 13.2 ± 8 (5-60, median 8.3) for C4.Thirteen had lupus nephritis and a renal biopsy to determine staging, class II (n=1, 5%), class III (n=5, 25%), class IV (n=5, 25%), and class V (n=2, 10%). The mean SDI at the final visit was 2.4 ± 1.4 (range 1-5, median 3). Mean albumin level
at diagnosis was 3.23 ± 0.9 (1.8-4.5, median 3.1). Mean WBC count at diagnosis was 5.59 ± 2.49 (2.4-12, median 5.0). Mean ANC at diagnosis was 3.91 ± 2.6 (0.97-11).
3.3 Medication Usage:

Medication use of corticosteroids, hydroxychloroquine, methotrexate, cyclophosphamide, and rituximab is described in Table 2. Medication use at diagnosis (n=15) was as follows: corticosteroids (n=11/15, 73%), hydroxychloroquine (n=8/15, 53%), mycophenolate (n=3/15, 20%), rituximab (n=0, 0%), and cyclophosphamide (n= 2/15, 13%). For the remaining 5 patients, diagnosis visits were out of our approved timeline or information was not available. Medication use at the 6-month follow-up (n=14) was as follows: corticosteroids (n=13/14, 93%), mycophenolate (n=11/14, 79%), hydroxychloroquine (n=13/14, 93%), and cyclophosphamide (n= 3/14, 21%).

Table 2: Cumulative medication use over five years

Current or Discontinued Use Never Used
Systematic Steroid Use 19 (95%) 1 (5%)
Hydroxychloroquine 20 (100%) 0
Mycophenolate mofetil 18 (80%) 2 (10%)
Cyclophosphamide 9 (45%) 11 (65%)
Rituximab 8 (40%) 12 (60%)

 

3.4
Infections

As shown in Table 3, a total of 21 infections were noted in 9 of the study subjects over the 5-year period. Between time of diagnosis and 6-month follow-up, there had been 11 infections, and from time of diagnosis to 1-year follow-up, there had been
16 infections. Four infections led to hospital admission from diagnosis to 6-month follow-up, and a total of 6 infections led to hospital admission from diagnosis to 1-year follow-up. Infections were as follows: pneumonia (n=6, 29%), pharyngitis (n=3, 14%), sinusitis (n=2, 9.5%), gastroenteritis (n=2, 9.5%), acute otitis media (n=2, 9.5%), oral thrush (n=1, 4.8%), shingles (n=1, 4.8%), mucositis (n=1, 4.8%), upper respiratory infection (n=1, 4.8%), viral exanthem (n=1, 4.8%), and urinary tract infection (n=1, 4.8%). Fifteen infections were treated with antibiotics, 1 infection was treated with an antifungal (fluconazole), and 2 infections were treated with antivirals (acyclovir, valacyclovir). None of the following infections were identified in this cohort: Mycobacterium tuberculosis, Pneumocystis jirovecii, Salmonella, or other fungal infections except for oral thrush.

Table 3: Infections in patients with jSLE over five years

Type of Injection Number Causative Organism
Pneumonia 6 Mycoplasma (1) Unknown (5)
Sinusitis 2 Unknown (2)
Pharyngitis 3 Group A streptococci (2), Unknown
Oral Thrush 1 Candida albicans
Gastroenteritis 2 Unknown (2)
Acute otitis media 2 Unknown (2)
Shingles 1 HZV
Mucositis 1 HSV
Upper respiratory infection 1 Unknown
Vocal Exanthem 1 Parvovirus
Urinary tract infection 1 Unknown

 

3.5 Complement Levels and Infections

Comparison of mean number of infections between patients with complement C3 at diagnosis greater than 50 versus less than 50 was not significant using independent samples t-test, however a trend towards increased number of infections in those with lower complement C3 was noted. Low C3 level at diagnosis was associated with a higher number of infections within the first six months after diagnosis (r= 0.23, p=NS). Independent samples median test revealed patients without infections had a median complement C3 level of 49.4 at diagnosis, while those with infections had a median complement C3 of 34.1 (p=NS). Mann- Whitney U test was not significant. Comparison of the mean number of infections between patients with lower versus higher complement C4 levels at diagnosis was not significant.

3.6 Medication Usage and Infections

Comparison of mean number of infections between patients who took cyclophosphamide and/or rituximab over the 5-year period versus those who took neither medication was not significant using independent samples t-test, however a trend towards increased number of infections in those who took either cyclophosphamide or rituximab was noted. Independent samples median test revealed patients who took either cyclophosphamide or rituximab during the 5-year period (n=11) had a median of 1 infection compared to a median of 0 infections in those who did not take either medication (n=9) (p=NS).
Patients who used mycophenolate over the 5 -year period were noted to have a mean of 1.11 ± 1.4 compared to a mean of 0.5 ± 0.7 in patients who never used mycophenolate (p=NS).

3.7 Disease Status and Infections

Comparison of mean number of infections between patients with SLE Nephritis Class III or higher versus those with class II or lower was not significant using independent samples t-test, however a trend towards increased number of infections in patients with SLE Nephritis Class III or higher was noted. Independent samples median test revealed patients with SLE Nephritis Class III or higher (n=12) had a median of 0.5 infections compared to a median of 0 infections for those with SLE Nephritis Class II or lower (n=8) (p=NS). Mann-Whitney U Test was not significant. Comparison of mean number of infections between patients with SDI score of 3 or higher versus patients with SDI score of 2 or lower was not significant using independent samples t-test, however a trend towards increased number of infections in patients with SDI score of 3 or higher was noted. Independent samples median test revealed patients with an SDI score of 3 or greater (n=14) had a median 0.5 infections compared to a median of 0 infections for those with an SDI score of 2 or less (n=6) (p=NS). Mann-Whitney U Test was not significant.

No significant difference in number of infections was found for patients with low versus high SLEDAI score at diagnosis.

A significant inverse correlation was noted between serum albumin level at diagnosis and number of infections within the first six months of diagnosis (r= 0.5, p=0.04). Independent samples t-test revealed patients with serum albumin at diagnosis below 3.1 had significantly more infections within the first six months (1.14 ± 0.90) compared to 0.3 ± 0.48 infections in those with serum albumin above 3.1 (p=0.02). Independent samples median test revealed patients with infections in the first 6 months after diagnosis (n=10) had a median albumin of 2.8 at diagnosis and patients without infections had a median albumin of 3.4 at diagnosis (n=7) (p=NS). Mann-Whitney U Test was not significant.

No significant difference in number of infections was found for patients with lower versus higher WBC count or ANC.

DISCUSSION

In this single-center study of 20 patients with jSLE, the majority were female, of Hispanic/Latino descent, and had a mean age of 13 years with mild-moderate disease activity, mild- moderate damage level and low complement C3. During the course of the 5-year period, 80% of patients took mycophenolate, 45% took cyclophosphamide, and 40% took rituximab. A total of 21 infections were noted in 9 of the study subjects over a 5-year period. 76% of infections occurred within the first year of diagnosis and the most frequent infection was pneumonia.

Multiple studies have identified that several risk factors may predispose patients with jSLE to infections, including disease activity (SLEDAI), lupus nephritis class, low complement C3, and medications used in disease treatment.5-7,15,16 While this study is limited due to its small sample size, the trends are overall consistent with prior findings of increased number of infections in patients with low complement C3 at diagnosis, higher grade of lupus nephritis at presentation, higher SDI score, and use of certain immunosuppressive medications such as cyclophosphamide, mycophenolate, and rituximab.5-7 This study also found that patients with lower albumin at diagnosis had significantly more infections within the first six months compared to those with higher albumin, which is consistent with prior studies reporting an association between low albumin and increased infections in jSLE patients.17,18

Given the increased risk of infection associated with jSLE, it is important to determine the role of screening, vaccination, and prophylactic treatment in this population. In the present study, of the 21 infections, the majority were pneumonia (29%), indicating the respiratory system may be particularly susceptible to infection. Previous studies have suggested screening for certain infections such as tuberculosis and viral hepatitis at the time of SLE diagnosis, so that those infections can be addressed before starting immunosuppressive treatments.15 Additionally, it has been recommended that SLE patients receive pneumococcal vaccination, yearly influenza vaccination, HPV vaccination, and yearly cervical cytology.15 While there is currently limited evidence to support the use of prophylactic antibiotics in this patient population for the prevention of pneumocystis, this is an ongoing area of study.15

Infection is one of the leading causes of morbidity and mortality in jSLE.5-8 Although aggressive treatment with immunosuppressive agents is essential to control disease activity in jSLE, patients must be carefully monitored to prevent serious infections and their associated morbidity and mortality. Larger multi-center studies with greater geographic and ethnic diversity are required to further examine the risk factors associated
with increased infections in jSLE patients as well as determine appropriate prevention strategies including antimicrobial prophylaxis and vaccination.

REFERENCES

  1. Aggarwal A, Phatak S, Srivastava P, Lawrence A, Agarwal
    V, Misra R. Outcomes in juvenile onset lupus: single center cohort from a developing country. Lupus. 2018;27(11):1867- 1875.
  2. Malattia C, Martini A. Paediatric-onset systemic lupus erythematosus. Best Pract Res Clin Rheumatol. 2013;27(3):351- 362.
  3. Cervera R, Khamashta MA, Hughes GR.The Euro-lupus project: epidemiology of systemic lupus erythematosus in Europe. Lupus. 2009;18(10):869-874.
  4. Smith EMD, Lythgoe H, Midgley A, Beresford MW, Hedrich CM. Juvenile-onset systemic lupus erythematosus: Update on clinical presentation, pathophysiology and treatment options. Clin Immunol. 2019;209:108274.
  5. Blancas-Galicia L, Guevara-Cruz M, Berrón-Pérez R, Berrón- Ruiz L, Gutiérrez-Castrellón P, Espinosa-Rosales FJ. Survival of Mexican patients with paediatric-onset systemic lupus erythematosus and abnormal electroencephalogram. Allergol Immunopathol (Madr). 2013;41(2):108-113.
  6. Cabral M, Escobar C, Conde M, Ramos M, Melo Gomes JA. Juvenile systemic lupus erythematosus in Portugal: clinical and immunological patterns of disease expression in a cohort of 56 patients. Acta Reumatol Port. 2013;38(4):274-285.
  7. Costa-Reis P, Nativ S, Isgro J, et al. Major infections in a cohort of 120 patients with juvenile-onset systemic lupus erythematosus. Clin Immunol. 2013;149(3):442-449.
  8. Lee PP, LeeTL, Ho MH,WongWH, LauYL. Recurrent major infections in juvenile-onset systemic lupus erythematosus–a close link with long-term disease damage. Rheumatology (Oxford). 2007;46(8):1290-1296.
  9. Kang I, Park SH. Infectious complications in SLE after immunosuppressive therapies. Curr Opin Rheumatol. 2003;15(5):528-534.
  10. Torrente-Segarra V, Salman-Monte TC, Rúa-Figueroa Í, et al.Associated factors to serious infections in a large cohort of juvenile-onset systemic lupus erythematosus from Lupus Registry (RELESSER). Semin Arthritis Rheum. 2020;50(4):657- 662.
  11. Gladman D, Ginzler E, Goldsmith C, et al.The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupus erythematosus. Arthritis Rheum. 1996;39(3):363-369.
  12. Faco MM, Leone C, Campos LM, Febrônio MV, Marques HH, Silva CA. Risk factors associated with the death of patients hospitalized for juvenile systemic lupus erythematosus. Braz J Med Biol Res. 2007;40(7):993-1002.
  13. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1997;40(9):1725.
  14. Weening JJ, D’AgatiVD, Schwartz MM, et al.The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int. 2004;65(2):521-530.
  15. Barber C, Gold WL, Fortin PR. Infections in the lupus patient: perspectives on prevention. Curr Opin Rheumatol. 2011;23(4):358-365.
  16. Rozenblyum EV, Levy DM, Allen U, Harvey E, Hebert D, Silverman ED. Cytomegalovirus in pediatric systemic lupus erythematosus: prevalence and clinical manifestations. Lupus. 2015;24(7):730-735.
  17. LinW, Chen P.Tubercular meningitis in patients with systemic lupus erythematosus: clinical characteristics, risk factors, and outcomes of 10 patients. Clin Rheumatol. 2020;39(4):1141- 1145.
  18. Hou C, Jin O, Zhang X. Clinical characteristics and risk factors of infections in patients with systemic lupus erythematosus. Clin Rheumatol. 2018;37(10):2699-2705.