Inflammation ( # 2014) DOI: 10.1007/s10753-014-9917-y
The Risk Factors of Avascular Necrosis in Patients with Systemic Lupus Erythematosus: a Meta-analysis Kao-Kao Zhu,1 Wang-Dong Xu,1 Hai-Feng Pan,1 Min Zhang,1 Jing Ni,1 Fu-Yang Ge,2 and Dong-Qing Ye1,3
Abstract—The aim of this study was to determine the risk factors for avascular necrosis (AVN) in patients with systemic lupus erythematosus (SLE). Four electronic databases (PubMed, EMBASE, Ovid, and Science Direct) were searched for. The search was performed to identify the articles as to SLE with AVN before September 2013. The clinical and laboratory data were extracted, and a meta-analysis was performed to identify the risk factors for AVN in patients with SLE. Publication bias was assessed with funnel plot and Egger’s test. A total of 995 papers were found from the four databases; 16 studies were finally included. Pooled analysis showed the following result. The result showed that arthritis (odds ratio (OR)=2.448, 95 % confidence interval (CI)=1.617–3.707), cushingoid (OR=3.890, 95 % CI=1.591– 9.510), gastrointestinal involvement (OR=2.054, 95 % CI=1.283–3.290), hypertension (OR=1.482, 95 % CI=1.093–2.008), oral ulcers (OR=1.877, 95 % CI=1.182–2.979), pleuritis (OR=2.302, 95 % CI=1.325–4.001), renal disease (OR=1.475, 95 % CI=1.124–1.936), and vasculitis (OR=2.591, 95 % CI=1.358–4.944) were relevant with AVN in SLE patients. Cytotoxic drug (OR=1.834, 95 % CI=1.065–3.156, P=0.029), the total cumulative dose (Standard Mean Difference (SMD) = 1.104, 95 % CI = 0.118–2.090, P = 0.028), maximum daily dose (SMD = 0.484, 95 % CI = 0.288–0.678, P < 0.001), and mean daily dose (SMD=1.305, 95 % CI=0.061–2.549, P=0.040) were significantly higher in AVN group. There were no significantly laboratory features that appeared in this pooled analysis. We conclude that arthritic, cushingoid, gastrointestinal involvement, hypertension, oral ulcers, pleuritis, renal disease, vasculitis, cytotoxic drug, and steroid treatment may contribute to AVN in SLE patients. KEY WORDS: avascular necrosis; systemic lupus erythematosus; risk factor; meta-analysis.
INTRODUCTION Systemic lupus erythematosus (SLE) is an inflammatory autoimmune disease, characterized by autoantibody production, impaired immune responses, and multiple organ involvements [1]. There are many complications of SLE, such as avascular necrosis (AVN), lupus nephritis, Parkinsonian syndrome [2], thrombotic thrombocytopenic purpura [3], subarachnoid hemorrhage [4], and swinging 1
Department of Epidemiology and Statistics, School of Public Health, Anhui Medical University, 81 Meishan Road, Hefei, 230032Anhui, People’s Republic of China 2 Anhui Academy of Medical Sciences, 15 Yonghong Road, Hefei, 230032Anhui, People’s Republic of China 3 To whom correspondence should be addressed at Department of Epidemiology and Statistics, School of Public Health, Anhui Medical University, 81 Meishan Road, Hefei, 230032Anhui, People’s Republic of China. E-mail: [email protected]
heart [5]. AVN of the bone is a common complication of SLE, with the reported prevalence of AVN in SLE patients ranging from 2.5 to 44.44 % [6, 7]. AVN is characterized by death of the bone marrow and trabecular bone due to disruption of the blood supply [8]. AVN that occurred in patients with SLE is usually found in the knee, shoulder, wrist, elbow [10], and the small bones of the foot and ankle [10], but the commonest site was the femoral heads [9, 10]. In AVN of the femoral heads, the opposite hip has been found to be involved within 2 years in 55 % of the cases [11–13]. AVN often results in the collapse of the architectural bone structure, leading to joint pain, bone destruction, and loss of function [14]. Most SLE patients with AVN of the femoral heads eventually require joint replacement [15]. Thus, it is important to know the risk factors of AVN in patients with SLE for optimal therapeutic management. Steroid use appears to be an independent risk factor, and many investigators considered the use of steroid to be a
0360-3997/14/0000-0001/0 # 2014 Springer Science+Business Media New York
Zhu, Xu, Pan, Zhang, Ni, Ge and Ye major risk factor for AVN in SLE. However, some patients developed AVN in the absence of steroid treatment [10, 16]; SLE itself is also considered to be a strong risk factor for the development of AVN [17, 18]. Although many risk factors such as vasculitis, Raynaud’s phenomenon, malar rash, cushingoid, and arthritis have been reported for AVN in patients with SLE, these results are still controversial. Vasculitis was significantly more frequently found in the AVN group than in the non-AVN group in some studies [9, 19–22], but a similar vasculitis was frequently observed between two groups as in some other studies [10, 14, 16]. Patients with AVN had a higher incidence of Raynaud’s phenomenon (51 vs 33 %, P=0.040) according to the study of Sayarlioglu et al. [9], however, nine (20.9 %) with Raynaud’s phenomenon in the AVN group (vs 14, 15.1 %; P=0.395) in Fialho et al. study [14]. The main purposes of this study are to identify the risk factors for AVN in patients with SLE based on the existing information and provide a basis for later research. The main taxonomy about risk factors contains clinical features and laboratory features. A single study may be too underpowered to detect a risk factor, especially when the sample size is relatively small. Meta-analysis is means of augmenting the effective sample size by pooling data from individual association studies, thus enhancing the statistical power of the analysis for the estimation. Therefore, we carried out the present meta-analysis of all the eligible studies about risk factors for AVN in patients with SLE in order to derive a better estimation.
MATERIAL AND METHODS Identification of Eligible Studies A comprehensive literature search of the PubMed, EMBASE, Ovid, and Science Direct databases was conducted using the following search terms: osteonecrosis/ bone necrosis/AVN/ischemic necrosis and SLE. Published reports in the period from database inception to September 2013 were identified. These results were further screened by checking the titles and abstracts for articles to the coexistence of AVN and SLE. Studies fulfilling the following inclusion criteria might be included in the analysis: (1) AVN was identified by one or more of the following imaging techniques: plain X-ray, radioisotope bone scan, and magnetic resonance imaging (MRI); (2) compared the clinical features and/or laboratory parameters of SLE concomitant AVN to SLE without AVN; (3) provided numerical data (numbers or percentages) of the incidence
of the clinical features and/or laboratory parameters; (4) case was diagnosed by 1997 (ACR) American College of Rheumatology Criteria/1982 ARA (American Rheumatism Association Preliminary Criteria/1971 ARA); (5) were published in English or Chinese. Even if they met the inclusion criteria, the studies were still excluded if they met the following criteria: (1) Valid data was not available from retrieval; (2) reviews articles, case reports, and meeting abstracts were not available. If more than one identified study used the same study population, only the largest sample size of the studies was included. If publications from the same group reported different features, each publication was used for nonredundant items. Two reviewers independently screened the citations based on the above criteria. If there was disagreement, agreement was achieved by discussion. Data Extraction Two investigators independently extracted the data from all eligible studies that were selected according to the criteria listed above. Data was extracted according to first author, country, year of publication, sample size, clinical features, and laboratory features (expressed as either the number or percentage of case group and control group). Data was entered in a spreadsheet. Quality Assessment This meta-analysis only included case control studies which reported data involving the clinical feature and/or laboratory feature of AVN in patients of SLE. Study quality was assessed using the Newcastle-Ottawa Scale (NOS) checklist for case control studies. The NOS checklist consists of eight items within three domains: selection, comparability, and exposure. The NOS uses a “star” rating system to judge quality. A study can receive one star for meeting each criterion. We interpreted star values as follows: >8 was very good, 7–8 was good, 5–6 was moderate, 3–4 was fair, and ≤2 was poor. The stars rating less than 5 were excluded from this analysis. Analysis Methods The raw numerical data/rates were extracted from the studies, and the odd ratios (ORs) were recalculated. Extracted data were used for combining the studies for clinical features and laboratory features of interest using forest plots. In some instance, several related features were combined into one forest plot.
The Risk Factors of AVN in SLE The ORs, 95 % confidence intervals (CIs), and P (P< 0.05) were used to assess the association between risk factors and AVN in patients with SLE. The Q test and χ2based I2 test were used to examine the between-study variations and heterogeneity. The effect of heterogeneity was quantified using P and I2. I2 values of 25, 50, and 75 % were nominally considered low, moderate, and high estimates, respectively. Based on the result of the heterogeneity test, fixed effect model (where P>0.1 and I2
The Risk Factors of Avascular Necrosis in Patients with Systemic Lupus Erythematosus: a Meta-analysis Kao-Kao Zhu,1 Wang-Dong Xu,1 Hai-Feng Pan,1 Min Zhang,1 Jing Ni,1 Fu-Yang Ge,2 and Dong-Qing Ye1,3
Abstract—The aim of this study was to determine the risk factors for avascular necrosis (AVN) in patients with systemic lupus erythematosus (SLE). Four electronic databases (PubMed, EMBASE, Ovid, and Science Direct) were searched for. The search was performed to identify the articles as to SLE with AVN before September 2013. The clinical and laboratory data were extracted, and a meta-analysis was performed to identify the risk factors for AVN in patients with SLE. Publication bias was assessed with funnel plot and Egger’s test. A total of 995 papers were found from the four databases; 16 studies were finally included. Pooled analysis showed the following result. The result showed that arthritis (odds ratio (OR)=2.448, 95 % confidence interval (CI)=1.617–3.707), cushingoid (OR=3.890, 95 % CI=1.591– 9.510), gastrointestinal involvement (OR=2.054, 95 % CI=1.283–3.290), hypertension (OR=1.482, 95 % CI=1.093–2.008), oral ulcers (OR=1.877, 95 % CI=1.182–2.979), pleuritis (OR=2.302, 95 % CI=1.325–4.001), renal disease (OR=1.475, 95 % CI=1.124–1.936), and vasculitis (OR=2.591, 95 % CI=1.358–4.944) were relevant with AVN in SLE patients. Cytotoxic drug (OR=1.834, 95 % CI=1.065–3.156, P=0.029), the total cumulative dose (Standard Mean Difference (SMD) = 1.104, 95 % CI = 0.118–2.090, P = 0.028), maximum daily dose (SMD = 0.484, 95 % CI = 0.288–0.678, P < 0.001), and mean daily dose (SMD=1.305, 95 % CI=0.061–2.549, P=0.040) were significantly higher in AVN group. There were no significantly laboratory features that appeared in this pooled analysis. We conclude that arthritic, cushingoid, gastrointestinal involvement, hypertension, oral ulcers, pleuritis, renal disease, vasculitis, cytotoxic drug, and steroid treatment may contribute to AVN in SLE patients. KEY WORDS: avascular necrosis; systemic lupus erythematosus; risk factor; meta-analysis.
INTRODUCTION Systemic lupus erythematosus (SLE) is an inflammatory autoimmune disease, characterized by autoantibody production, impaired immune responses, and multiple organ involvements [1]. There are many complications of SLE, such as avascular necrosis (AVN), lupus nephritis, Parkinsonian syndrome [2], thrombotic thrombocytopenic purpura [3], subarachnoid hemorrhage [4], and swinging 1
Department of Epidemiology and Statistics, School of Public Health, Anhui Medical University, 81 Meishan Road, Hefei, 230032Anhui, People’s Republic of China 2 Anhui Academy of Medical Sciences, 15 Yonghong Road, Hefei, 230032Anhui, People’s Republic of China 3 To whom correspondence should be addressed at Department of Epidemiology and Statistics, School of Public Health, Anhui Medical University, 81 Meishan Road, Hefei, 230032Anhui, People’s Republic of China. E-mail: [email protected]
heart [5]. AVN of the bone is a common complication of SLE, with the reported prevalence of AVN in SLE patients ranging from 2.5 to 44.44 % [6, 7]. AVN is characterized by death of the bone marrow and trabecular bone due to disruption of the blood supply [8]. AVN that occurred in patients with SLE is usually found in the knee, shoulder, wrist, elbow [10], and the small bones of the foot and ankle [10], but the commonest site was the femoral heads [9, 10]. In AVN of the femoral heads, the opposite hip has been found to be involved within 2 years in 55 % of the cases [11–13]. AVN often results in the collapse of the architectural bone structure, leading to joint pain, bone destruction, and loss of function [14]. Most SLE patients with AVN of the femoral heads eventually require joint replacement [15]. Thus, it is important to know the risk factors of AVN in patients with SLE for optimal therapeutic management. Steroid use appears to be an independent risk factor, and many investigators considered the use of steroid to be a
0360-3997/14/0000-0001/0 # 2014 Springer Science+Business Media New York
Zhu, Xu, Pan, Zhang, Ni, Ge and Ye major risk factor for AVN in SLE. However, some patients developed AVN in the absence of steroid treatment [10, 16]; SLE itself is also considered to be a strong risk factor for the development of AVN [17, 18]. Although many risk factors such as vasculitis, Raynaud’s phenomenon, malar rash, cushingoid, and arthritis have been reported for AVN in patients with SLE, these results are still controversial. Vasculitis was significantly more frequently found in the AVN group than in the non-AVN group in some studies [9, 19–22], but a similar vasculitis was frequently observed between two groups as in some other studies [10, 14, 16]. Patients with AVN had a higher incidence of Raynaud’s phenomenon (51 vs 33 %, P=0.040) according to the study of Sayarlioglu et al. [9], however, nine (20.9 %) with Raynaud’s phenomenon in the AVN group (vs 14, 15.1 %; P=0.395) in Fialho et al. study [14]. The main purposes of this study are to identify the risk factors for AVN in patients with SLE based on the existing information and provide a basis for later research. The main taxonomy about risk factors contains clinical features and laboratory features. A single study may be too underpowered to detect a risk factor, especially when the sample size is relatively small. Meta-analysis is means of augmenting the effective sample size by pooling data from individual association studies, thus enhancing the statistical power of the analysis for the estimation. Therefore, we carried out the present meta-analysis of all the eligible studies about risk factors for AVN in patients with SLE in order to derive a better estimation.
MATERIAL AND METHODS Identification of Eligible Studies A comprehensive literature search of the PubMed, EMBASE, Ovid, and Science Direct databases was conducted using the following search terms: osteonecrosis/ bone necrosis/AVN/ischemic necrosis and SLE. Published reports in the period from database inception to September 2013 were identified. These results were further screened by checking the titles and abstracts for articles to the coexistence of AVN and SLE. Studies fulfilling the following inclusion criteria might be included in the analysis: (1) AVN was identified by one or more of the following imaging techniques: plain X-ray, radioisotope bone scan, and magnetic resonance imaging (MRI); (2) compared the clinical features and/or laboratory parameters of SLE concomitant AVN to SLE without AVN; (3) provided numerical data (numbers or percentages) of the incidence
of the clinical features and/or laboratory parameters; (4) case was diagnosed by 1997 (ACR) American College of Rheumatology Criteria/1982 ARA (American Rheumatism Association Preliminary Criteria/1971 ARA); (5) were published in English or Chinese. Even if they met the inclusion criteria, the studies were still excluded if they met the following criteria: (1) Valid data was not available from retrieval; (2) reviews articles, case reports, and meeting abstracts were not available. If more than one identified study used the same study population, only the largest sample size of the studies was included. If publications from the same group reported different features, each publication was used for nonredundant items. Two reviewers independently screened the citations based on the above criteria. If there was disagreement, agreement was achieved by discussion. Data Extraction Two investigators independently extracted the data from all eligible studies that were selected according to the criteria listed above. Data was extracted according to first author, country, year of publication, sample size, clinical features, and laboratory features (expressed as either the number or percentage of case group and control group). Data was entered in a spreadsheet. Quality Assessment This meta-analysis only included case control studies which reported data involving the clinical feature and/or laboratory feature of AVN in patients of SLE. Study quality was assessed using the Newcastle-Ottawa Scale (NOS) checklist for case control studies. The NOS checklist consists of eight items within three domains: selection, comparability, and exposure. The NOS uses a “star” rating system to judge quality. A study can receive one star for meeting each criterion. We interpreted star values as follows: >8 was very good, 7–8 was good, 5–6 was moderate, 3–4 was fair, and ≤2 was poor. The stars rating less than 5 were excluded from this analysis. Analysis Methods The raw numerical data/rates were extracted from the studies, and the odd ratios (ORs) were recalculated. Extracted data were used for combining the studies for clinical features and laboratory features of interest using forest plots. In some instance, several related features were combined into one forest plot.
The Risk Factors of AVN in SLE The ORs, 95 % confidence intervals (CIs), and P (P< 0.05) were used to assess the association between risk factors and AVN in patients with SLE. The Q test and χ2based I2 test were used to examine the between-study variations and heterogeneity. The effect of heterogeneity was quantified using P and I2. I2 values of 25, 50, and 75 % were nominally considered low, moderate, and high estimates, respectively. Based on the result of the heterogeneity test, fixed effect model (where P>0.1 and I2
