Splenectomy increases the subsequent risk
of systemic lupus
erythematosus
Chao-Yu Hsu1,2,3,4,5 · Hsuan-Ju Chen6,7 · Chung Y. Hsu8 · Chia-Hung Kao8,9
Introduction
The reasons for performing a splenectomy include trauma, lymphoma and blood diseases such as hereditary spherocytosis, idiopathic thrombocytopenic purpura, thrombotic
thrombocytopenia purpura, and autoimmune hemolytic disorders.
Systemic lupus erythematosus (SLE) is a systemic autoimmune disease that affects many internal organs and the
hematopoietic system. SLE is more common in women because females have 2 X chromosomes, and X chromosomes
carry immunologically related genes [1]. In the
USA, the incidence and prevalence of SLE are approximately 2.9–5.5 and 30.5–72.8 per 100,000 persons, respectively [2–4]. The prevalence of SLE among people of
Afro-Caribbean descent was found to be as high as 159 per 100,000 persons [4].
Patients with SLE and thrombocytopenia exhibit a
higher incidence of severe diseases and poor survival compared with the general population [5]. Although thrombocytopenia is not directly associated with organ damage or
mortality, it is highly prevalent among patients with SLE [6, 7]. Splenectomy was recently performed to treat SLE patients with thrombocytopenia, and the outcome indicated that the procedure is safe and effective [8, 9]. Splenectomy may be necessary to treat SLE in patients with thrombocytopenia; however, whether performing a splenectomy on
patients without SLE increases the subsequent risk of SLE remains unknown. Therefore, this study was conducted to determine the association between splenectomy and SLE.
Materials and methods Data source
Taiwan’s National Health Insurance (NHI) program, which was initiated in 1995, currently covers approximately 99 % of the 23 million people living in Taiwan. We conducted a longitudinal cohort study by using data from the Taiwan National Health Insurance Research Database (NHIRD), which contains information on hospital admissions for all health insurance enrollees living in Taiwan; specifically,
health care claims data between 1996 and 2011 were analyzed in this study. The claims data contain information on
sex, date of birth, dates of admission and discharge, and diagnosis and procedure codes based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). To protect patient privacy, the Taiwan
National Health Research Institutes—the authority responsible for maintains the NHIRD—encrypts personal identification information before releasing electronic files for research purposes. Therefore, this study was approved to exempt from obtaining informed consent by the Institutional Review Board of China Medical University
(CMUH104-REC2-115). Study population
We used inpatient claims data to identify patients who first received a splenectomy (ICD-9-CM for procedure code 41.2, 41.43, and 41.5) between 2000 and 2006 and assigned them to the splenectomy cohort. The date of the first splenectomy was defined as the index date. Subsequently, we frequency-matched each patient in the splenectomy
cohort by sex, 5-year age interval, and index year
with four patients among all remaining NHI enrollees who did not received a splenectomy between 1996 and 2011; the matched patients were assigned to the non-splenectomy cohort. For the non-splenectomy cohort, the index
date was randomly assigned by day and month, and the index year was identical to that of the matched cases. In both cohorts, we excluded patients who were <20 years old, those who had previously received a diagnosis SLE,
or those whose records were missing information on sex or age.
The following sociodemographic characteristics were obtained for all patients: sex, age, income, occupational status, and urbanization levels. The insurance premium amount was determined on the basis of the patients’ monthly income and categorized into the following three levels: <NT$15,000, NT$15,000–NT$29,999, and ≥NT$30,000. Occupational status was also classified into three groups: white collar, blue collar, and other. We stratified all communities in Taiwan into four levels according to
a previous study [10], with Level 1 referring to the highest level of urbanization and Level 4 referring the lowest level of urbanization.
All patients were followed until their first diagnosis of SLE (ICD-9-CM 710.0), which was identified according to the catastrophic illness registry records during the followup period. The follow-up period was from the index date to the date of SLE diagnosis, withdrawal from NHI, death, or the end of 2011.
Statistics
The distributions of sociodemographic variables were compared between the splenectomy and non-splenectomy cohorts. The categorical variables are expressed as frequencies and percentages, whereas the continuous variables are
expressed as means and standard deviations. Chi-squared and Student t tests were employed to determine the difference in the categorical and continuous variables between
the two cohorts. The sex- and age-specific incidence density rates of SLE were calculated for the follow-up period
until the end of 2011, the date of the first diagnosis of SLE, death, or loss to follow-up. We assessed the cumulative incidences of SLE between the two cohorts by using
the Kaplan–Meier method and estimated their differences by using a log-rank test. Cox proportional hazards models were employed to estimate the hazard ratios (HRs) and 95 % confidence intervals (CIs) of developing SLE associated with splenectomy compared with patients who did not
receive a splenectomy. All analyses were conducted using
SAS (Version 9.3 for Windows; SAS Institute, Inc., Cary, NC, USA). All statistical tests were two-tailed with the significance
level set at 0.05. Results
We identified 10,298 patients who received a splenectomy between 2000 and 2006 (splenectomy cohort) and 41,192
NHI enrollees who did not receive a splenectomy (nonsplenectomy cohort). Table 1 shows that the patients in
both cohorts were predominantly. Approximately 40.0 % of all patients were older than 60 years. Compared with the patients who did not receive a splenectomy, those who did were more likely to have lower incomes, be employed in white-collar occupations, and live in least-urbanized areas of Taiwan.
A comparison of the cumulative incidence of SLE
between the two cohorts (Fig. 1) revealed that the incidence of SLE (log-rank p < 0.001) was significantly higher
in patients who received a splenectomy. During the average follow-up of 7.12 years, the overall incidence density
rate of SLE was higher in the splenectomy cohort than in the non-splenectomy cohort (5.81 vs. 0.48 per 10,000 person-y), with an adjusted HR of 10.3 (95 % CI 5.51–19.2)
(Table 2). The sex-specific analysis indicated that the incidence density rates of SLE in the women (14.7 per 10,000 person-y) and men (0.31 per 10,000 person-y) who received a splenectomy were higher than that in the NHI enrollees who did not receive a splenectomy (0.93 and 0.20 per 10,000 person-y, respectively). In addition, the women in the splenectomy cohort exhibited a 13.2-fold higher risk of SLE (95 % CI 6.57–26.2) compared with those in the non-splenectomy group.
The age-specific analysis indicated that among the
patients who received a splenectomy, the incidence density rates of SLE in the age groups of 20–44 years (8.11 per 10,000 person-y) and ≥45 years (3.90 per 10,000 persony) were higher than those among the patients who did not receive a splenectomy in the same groups (0.45 and 0.49
per 10,000 person-y, respectively). In all age categories,
the patients who received a splenectomy exhibited a significantly higher risk of SLE compared with those who did
not receive a splenectomy. The adjusted HRs for SLE were 16.4 (95 % CI 6.09–44.0) in the age group of 20–44 years and 6.68 (95 % CI 2.82–15.8) in age group of ≥45 years. In a subsequent analysis, the splenectomy cohort was
divided into four subgroups: trauma group, acquired hemolytic anemias (ICD-9-CM 283), immune thrombocytopenic
purpura (ICD-9-CM 287.31), and other cause (Table 3).
The results showed that compared with the non-splenectomy
cohort, patients who received a splenectomy had a history of trauma were associated with a null risk of SLE
(adjusted HR 0.90; 95 % CI 0.12–6.88). However, acquired hemolytic anemias and other cause in the splenectomy cohort had a significantly higher risk of developing SLE compared to the non-splenectomy cohort (adjusted HR 164; 95 % CI 64.0–419 and adjusted HR 12.1, 95 % CI 6.28–23.5, respectively).
Discussion
This is the first population-based study to identify splenectomy as a risk factor for SLE. We determined that patients
who received a splenectomy were more likely to have an increased risk of SLE compared with those who did not receive a splenectomy. After adjustment for age, index date, and comorbidities, the patients with splenectomy were 10.3-fold more likely to develop SLE, and this was more pronounced among the women who were 13.2-fold more likely to develop SLE.
Sánchez-Pérez et al. [11] reported that the incidence of splenectomy in the 1986–1990 and 1995–1999 periods was 0.52 and 0.19 %, respectively, and attributed the reduced incidence to the availability of new treatments for hematologic diseases. Rose et al. [12] also reported a decrease in
the incidence of splenectomy from 1986 to 1995 and attributed the reduction to the advantages of splenic preservation
being recognized, as well as the increased application of angiography and embolization [13].
According to previous studies, splenectomy for idiopathic thrombocytopenic purpura was more common
in women [14], but more common for men when trauma involved [15]. Our study found that splenectomy for any
cause was more common in men. Because motorcycles are primary mode of transportation for most men in Taiwan;
this finding might be attributable to the disproportionate number of road accidents involving motorcycles.
Recently, Ramos-Casals et al. [16] reported that SLE ranked second among unbalanced sex ratio diseases (Sjögren syndrome is the most unbalanced). SLE is more common among women [17] because females have two chromosomes, and X chromosomes carry immunologically related genes [1]. In Europe, approximately sixfold more common in females, with most cases involving middleage women [18, 19]. The reason for this trend might be the effect of estrogen on epigenetic microRNA regulation [20]. Our results showed that splenectomy was more common in men; however, women were more likely to develop
SLE after receive a splenectomy, which can be attributed to women being more vulnerable to developing SLE.
The spleen is believed to regulate the immune system. In patients who received a splenectomy, the hyper activation of its humoral chain coinciding with signs of autoimmune processes was observed due to inhibition of cellular immunity [21]. Because of defects in IgM memory B cells, many patients have severe bacterial and viral infections after receive a splenectomy [22]. Wu et al. [23] reported that splenectomy influenced the production of spleen-supplied innate-like B cells. Innate-like B cells are a subgroup
of B cells that modulate inflammatory responses by producing immunomodulatory mediators. SLE as a systemic
autoimmune disease of B cell hyperactivity, implying that enhanced B cell function is the defining pathogenic event [24]. Potential mechanism underlying B cell autoimmunity in SLE have been comprehensive defined [25], indicating the existence of a mechanism of developing SLE after splenectomy.
Several studies have reported that stress is associated with SLE because stress factors are involved in epigenetic mechanisms and influenced the innate immune system [26]. Adams et al. [27] explicitly stated that stress, depression, anxiety and anger are associated with (and may exacerbate) SLE. Because of the stress involved in receiving a splenectomy, it is reasonable to infer that the procedure influences
the immune system and induces or exacerbates SLE. In Taiwan, most of women with diagnosed SLE are 20–60 years old; however, 20- to 40-year-old and 40- to 60-year-old patients exhibited identical incidence rate [28]. In the present study, we found that young women were at the highest risk of SLE after receiving a splenectomy. Therefore, the stress burden of receiving a splenectomy should be considered.
The strengths of this study are its population-based study of NHIRD records with a large sample size of experimental
and control cohorts. In addition, the NHIRD is a highly representative sample of Taiwan’s general population. Furthermore,
the reimbursement policy is universal and operated
by a single buyer, the government of Taiwan. All insurance claims should be scrutinized by medical reimbursement specialists and by peer review. However, several limitations were encountered while conducting this study. First, information on the “severity of diseases” is not included in the
NHIRD. Although the diagnosis of splenectomy is highly reliable, the severity of SLE could not be measured in this study. Second, the NHIRD does not contain laboratory test results; consequently, the disease progression could not be measured. Third, the NHIRD does not include information on the lifestyle and nutrition status of enrollees; and such characteristics might influence immunology. Despite these limitations, using a population-based database prevented selection bias and ensured adequate statistical power for the data analysis.
Conclusion
Non-traumatic splenectomy increases the subsequent risk of SLE. The risk of SLE should be considered before performing
a splenectomy, particularly in women and younger patients.
