Infectious Complications in Lupus Nephritis and Associated Factors: A Multicenter Study ()
1. Introduction
The prevalence of infections in systemic lupus erythematosus (SLE) patients ranges from 26% to 78% [1] [2] [3]. Patients with SLE are particularly susceptible to infections due to a dysfunction of the innate and adaptive immune response associated with the disease, which is exacerbated by the immunosuppressive therapy used to treat the disease [4]. This explains the higher infectious risk of patients with SLE compared to the general population [5]. Infections are one of the main causes of morbidity and mortality during SLE, especially in case of lupus nephritis (LN). They are responsible for 15% to 40% of patient admissions and 30% to 50% of deaths [3] [6] [7]. The prevalence of infections in LN is not sufficiently described in sub-Saharan Africa. We carried out this study with the aim of describing the prevalence of infections in LN and determining the associated risk factors.
2. Patients and Methods
This was a multicenter, observational, retrospective, descriptive and analytical study over a 10-year period from November 1, 2010 to October 31, 2020. The study was carried out in the nephrology departments of 6 hospitals in Dakar (Aristide Le DANTEC hospital, Dalal Jamm hospital, Pikine hospital, Principal hospital, Ouakam military hospital, Roi Baudoin hospital). The study included all records of patients who were followed for LN. The diagnosis of LN was made by a concordant renal biopsy associated with a proteinuria/creatinuria ratio greater than 0.5 g/g or with an active urine sediment [8]. For each selected patient, the epidemiological, clinical, biological, histological, therapeutic and evolutionary data were studied. The data collected were entered from Excel software. The descriptive study was carried out with the calculation of the means ± standard deviation [minimum, maximum] or medians for the quantitative variables and the calculation of the frequencies for the qualitative variables. Data analysis was performed using SPSS software (Statistical Package for Science Social) version 18. The means and percentages were compared using the chi-square test and the exact Fischer test, respectively, according to their conditions of applicability. Any difference less than 0.05 was considered statistically significant. For multivariate analysis, we used a binary logistic regression model and included in this model the variables that were significant in bivariate analysis. The significance level retained for all the statistical analyzes was at p ≤ 0.05.
3. Results
Ninety-eight patients were included. The mean age was 32.32 ± 11.33 years with a female predominance and a sex ratio of 0.21. The average duration of follow-up was 25.40 ± 20.40 months. Forty-eight patients were known to have lupus prior to renal impairment. The mean glomerular filtration rate (GFR) was 89.88 ml/min ± 53.65 and 27.6% of patients had renal failure. C-reactive protein was measured in 68 patients, it was positive in 27 patients (39.7%). The mean serum albumin was 21.25 ± 9.67 mg/L. The mean protidemia was 54.03 ± 11.87 mg/l. The mean proteinuria was 4.11 g/24h ± 3.40. Nephrotic syndrome was found in 58.2% of cases. Isolated class V was the most common, found in 34 cases (34.7%), class III + V in 24 cases (24.5%), class IV in 11 cases (11.2%) and class IV + V in 10 cases (10.2%), class II was found in 9 cases (9.2%), class III in 6 cases (6.1%) and class I in 4 cases (4.1%). Hydroxychloroquine (HCQ) was prescribed in 83.7% of patients, methylprednisolone (MTP) in 59.2% of patients, prednisone was prescribed in 95.9% of patients, cyclophosphamide (CYC) in 27.6% of patients; mycophenolate mofetil (MMF) in 38.8% of patients and azathioprine (AZA) in 27.6% of patients (Table 1). Eighty-eleven infectious episodes have been identified. Fifty-four patients (55.1%) had at least one infectious episode, of which 53.7% had 1 infection, 24.1% had 2 infectious episodes and 22.2% had 3 infectious episodes. The overall incidence was 55.1 infections per 100 patient-years. The median time to onset of infection was 8.90 months ± 13.13% and 57.2% of these episodes occurred within the first six months after the lupus was diagnosed. The main sites of infection were urinary (30.7%), gastrointestinal (22.0%), pleuropulmonary (16.5%) and skin (13.2%) (Figure 1). The germ was identified in 60.43% of cases and it was a bacterium in 78.18% of cases. The most frequently found germs were Escherichia coli (29.09%) and Staphylococcus aureus (14.55%) (Table 2). Hospitalization was necessary in 27.5% of infectious episodes. The evolution was marked by recovery in 93.4% of cases. Deaths occurred in 15 patients (15.3%) of which 5 (33.3%) were related to infections. In bivariate analysis, the factors significantly associated with the onset of infection were elevated serum creatinine level on admission (p = 0.027), renal failure (p = 0.005), proliferative class of LN (p = 0.001), taking MTP (p = 0.001) and oral corticosteroid therapy (p = 0.038) (Table 3). Factors significantly associated with the onset of infection in multivariate analysis was the presence of a proliferative class of LN (p = 0.013) (Table 4).
4. Discussion
The prevalence of infections varies across studies. In our series, 55.1% of patients presented at least with one infectious episode. These results are similar to those of Ibn Abdelouahab [9] and Yin et al. [10] which were 61.6% and 49.3% respectively. In Australia, in the series by Feldman et al. [1], the prevalence was lower (25.7%) (Table 5). This difference is explained by the fact that Feldman et al only considered severe infections. Infections can affect one or more organs at the same time. They affect the lung, skin and urinary tract in more than two thirds of cases [11] [12]. In our study, these three locations represented 60% of cases. The other locations (osteoarticular, central nervous system, endocardium...) were rare. Lupus itself predisposes to infectious complications. Indeed, some intrinsic factors are involved in this immunosuppression: decrease in chemotaxis and phagocytosis; functional asplenia [13]; hypocomplementemia by excessive consumption of the C3 and C4 fractions of the complement or congenital deficit
![]()
Table 1. Demographical and clinical characteristics of participants.
GFR = glomerular filtration rate, ANA = anti-n556.uclear antibodies, anti-Sm = Anti-Smith antibodies CYC = cyclophosphamide, MMF = mycophenolate mofetil, AZA = Azathioprine, RTX = rituximab.
![]()
Figure 1. Distribution of patients according to the infectious site.
![]()
Table 2. Distribution of documented infections according to germs.
![]()
Table 3. Bivariate analysis of risk factors for infection.
ANA = anti-nuclear antibodies, HCQ = hydroxychloroquine, MTP = methylprednisolone, CYC = cyclophosphamide, MMF = mycophenolate mofetil, AZA = Azathioprine, RTX = rituximab.
![]()
Table 4. Multivariate regression analysis of factors associated with infection.
MTP = methylprednisolone; NS = not significant; OR = odds ratio; IC = interval of confidence.
![]()
Table 5. Incidence of infections reported by other studies.
in certain fractions of the complement (C1r, C1s, C3 and C4) [14] [15]; polymorphism of MBL and ostepontin [11] [16] [17]; increased levels of Fc gamma III and GM-CSF levels [16], decreased cytotoxic activity of T lymphocytes (CD8) and the production of several factors having a major anti-infectious role (interleukins 1 and 2, interferons) [11].
Renal involvement also exposes to infectious complications due to immunosuppression caused by chronic renal failure and hypogammaglobulinemia in nephrotic syndrome. Other elements are frequently associated with the onset of infectious episodes. These are lymphopenia [18] and the use of immunosuppressive treatments (corticosteroids, azathioprine, cyclophosphamide, etc.) [14] [19] [20]. In our series, infections occurred in 57.1% of cases during the first six months of follow-up, coinciding with the period of induction therapy. These results agree with those of other studies in China (57.8%) [10] and Morocco (72.9%) [9]. The high prevalence during that period is explained by the fact that firstly the disease is more active during that period and secondly the immunosuppressive treatment is more intense during that period because it corresponds to the induction treatment. In our series, death occurred in 15.3% of cases, of which 33.3% of deaths were related to infectious causes.
This rate matches the one found in the series by Liu et al. [21] and Béji et al. [22] where infection was responsible for 43% and 38.8% of deaths, respectively. In the series by Ibn Abdelouahab [9] and Yap et al. [23], infectious complications were responsible for 66.6% and 50.5% of deaths, respectively, which is to be expected, since infections are one of the leading causes of death in LN. The proliferative class of LN was significantly associated with the onset of infection.
In a Moroccan study, Ibn Abdelouahab found that patients with proliferative LN had a significantly higher infection rate compared to patients with non proliferative LN [9]. This risk is explained by the fact that on one hand the proliferative classes are more severe and more active and on the other hand they require more intense immunosuppressive treatment than the non proliferative classes.
5. Conclusion
Infections were common during LN. The presence of a proliferative class was risk factors for infection.
Funding Sources
This research did not receive a specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contributions
Mansour Mbengue reviewed the literature and wrote the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.
Limitations
The retrospective design of our study represents a limitation.