Increased risk of avascular necrosis in patients with psoriatic disease: A nationwide population-based matched cohort study

ORIGINAL

ARTICLE

Increased risk of avascular necrosis in patients with psoriatic disease: A nationwide population-based matched cohort study Hsien-Yi Chiu, MD,a,b,c I-Ting Wang, MS,d Weng-Foung Huang, PhD,e Yi-Wen Tsai, PhD,e Ming-Neng Shiu, PhD,f and Tsen-Fang Tsai, MDc Taipei and Hsin-Chu, Taiwan Background: Avascular necrosis (AVN) and psoriasis have some pathogenic mechanisms and associated conditions in common. Objective: To examine the association between psoriasis and AVN. Methods: This study used data from the Taiwan National Health Insurance Research Database for the period 2004e2006 and identified 28,268 patients with psoriasis, who were then matched for age and sex with 113,072 controls without psoriasis from the Taiwan Longitudinal Health Insurance Database 2000. Multivariate Cox proportional hazards models were used for the analysis. Results: The unadjusted risk of AVN was significantly higher for patients with psoriasis than for controls (hazard ratio [HR] 2.29) and remained significant after adjustment for other risk factors (adjusted HR 1.96; 95% confidence interval 1.62-2.38). The risk for AVN increased in relation to psoriasis severity and was higher for patients with psoriasis and arthritis than for patients without arthritis. The adjusted HRs were higher for male patients than for female patients and for patients younger than 30 years compared with older patients. Limitations: We lacked information on daily tobacco use, alcohol consumption, and physical activity. Conclusion: The risk for AVN increased with the disease severity of psoriasis. ( J Am Acad Dermatol http:// dx.doi.org/10.1016/j.jaad.2016.11.001.) Key words: avascular necrosis; comorbidities; inflammation; National Health Insurance Research Database; osteonecrosis; psoriasis.

From the Institute of Biomedical Engineering, College of Medicine and College of Engineering, National Taiwan University, Taipeia; Department of Dermatology, National Taiwan University Hospital Hsin-Chu Branchb; Department of Dermatology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipeic; and the Department of Physical Therapy and Assistive Technology, National Yang-Ming University, Taipei,d Institute of Health and Welfare Policy,e and Faculty of Pharmacy,f National Yang-Ming University, Taipei. Funding source: Supported by a grant from the Ministry of Science and Technology, Taiwan (formerly, the National Science Council; grant numbers, MOST104-2314-B-002-117-MY3), which follows the guidelines on good publication practice. The study authors designed the study, collected and analyzed the data, interpreted the results, and wrote the manuscript independent of the funders. Conflicts of interest: Dr Tsen-Fang Tsai has conducted clinical trials and received honoraria as a consultant for Pfizer Pharmaceuticals,

Serono International SA (now Merck Serono International), UniPharma/Biogen Idec, Galderma, Celgene, Novartis Pharmaceuticals, and Janssen-Cilag Pharmaceutical and has received speaking fees from AbbVie. Dr Chiu has received speaking fees from AbbVie, Janssen-Cilag Pharmaceutical, and Pfizer. The other authors have no conflicts of interest to declare. Dr Chiu and Ms Wang contributed equally. Dr Shiu contributed equally to the corresponding authorship. Accepted for publication November 1, 2016. Reprints not available from the authors. Correspondence to: Tsen-Fang Tsai, MD, Department of Dermatology, National Taiwan University Hospital, 7 Chung-Shan South Rd, Taipei 100, Taiwan. E-mail: [email protected]. Published online December 13, 2016. 0190-9622/$36.00 Ó 2016 by the American Academy of Dermatology, Inc. http://dx.doi.org/10.1016/j.jaad.2016.11.001

1

J AM ACAD DERMATOL

2 Chiu et al

n 2016

Avascular necrosis (AVN), also known as METHODS osteonecrosis, is a progressively debilitating Study design musculoskeletal disease caused by the death of This retrospective cohort study investigated cellular elements of bone. AVN leads to the collapse the association between psoriasis and AVN. Using of the bony structure and is accompanied by pain the National Health Insurance Research Database and loss of function.1,2 It significantly impairs quality (NHIRD), we identified 28,268 patients with of life and often leads to substantial disability and psoriasis during the period 2004-2006 and joint replacement.1,2 AVN is then randomly selected estimated to be responsible 113,072 matched controls CAPSULE SUMMARY for 5% to 18% of hip arthrowithout psoriasis from the plasties.1-3 NHIRD Longitudinal Health Avascular necrosis (AVN) is associated Psoriasis is a chronic, Insurance Database. AVN with multiple autoimmune rheumatic immune-mediated disorrisk was compared between diseases. der,4,5 has far-reaching the two cohorts during the Psoriasis was associated with a disease systemic inflammatory efstudy period. This study was severityedependent increase in AVN risk, fects, and is associated with reviewed and approved by particularly among young adults, males, multiple comorbidities.4,6-9 the local investigational and patients with psoriatic arthritis. Several studies have shown research bureau of National that proinflammatory cytoTaiwan University Hospital, AVN should be considered in psoriasis kines also might be involved Hsin-Chu Branch, Hsin-Chu, patients with localized pain in a weightin the pathogenic mechaTaiwan (103-024-E). bearing joint. nisms of AVN and influence its development.10,11 Some Data source chronic inflammatory diseases, such as systemic The NHIRD is an insurance administrative datalupus erythematosus (SLE), rheumatoid arthritis, base that is managed by the Taiwan National Health polymyositis/dermatomyositis, granulomatosis with Research Institutes and widely used in academic polyangiitis, and inflammatory bowel disease,12,13 are research. Because Taiwan’s National Health associated with the development of AVN. The Insurance system covers nearly 100% of the nearly proinflammatory cytokines, tumor necrosis factorea 23 million residents of Taiwan, the NHIRD contains and interleukin-6, which are involved in the abundant data on registration information, demopathogenesis of these autoimmune disorders and graphics, outpatient visits and inpatient services, psoriasis might predispose patients to development diagnostic codes, prescription profiles, and surgeries of AVN.11,14,15 Moreover, patients with psoriasis and other procedures for National Health Insurance might have a variety of comorbidities and associated beneficiaries. conditions, including arthritis, thromboembolic This study used 2 NHIRD datasets to identify events, vascular occlusion, endothelial dysfunction, patients with psoriasis and the matched controls: (1) abnormal lipid metabolism, use of corticosteroids or the NHIRD datasets on special request, which cytotoxic drugs, alcoholism, and hyperhomocysteiincluded claims from the 2,210,612 patients with nemia,1,16-20 which are risk factors for AVN. Previous psoriasis during 2003-2011 and (2) the Longitudinal studies enrolled varying populations and used Health Insurance Database 2000, which included differing case definitions and found that 6% to 41% health care information from a randomly selected of psoriasis patients also have psoriatic arthritis sample of 1 million beneficiaries in 2000. All subject (PsA).4,21,22 Chronic joint inflammation, such as information was anonymized and de-identified to arthritis, was reported to increase bone loss and protect privacy. collapse, thereby possibly compromising blood supply and increasing AVN risk.15,23,24 Although AVN and psoriasis have some pathoStudy population genic mechanisms and associated conditions in We identified 32,249 patients with a new diagnosis common, few studies have investigated AVN in of psoriasis in the NHIRD during 2004-2006 (Fig 1). patients with psoriasis.25-27 We therefore assessed We excluded patients with missing data on sex or AVN risk in a large, nationally representative, date of birth, those with a medical history of population-based cohort of predominantly Chinese rheumatic arthritis, ankylosing spondylitis, SLE, hupatients with psoriasis in Taiwan. man immunodeficiency virus infection, or psoriasis, d

d

d

J AM ACAD DERMATOL

Chiu et al 3

VOLUME jj, NUMBER j

Abbreviations used: aHR: AVN: CI: DDD: HR: ICD-9-CM: NHIRD: PsA: SLE:

adjusted hazard ratio avascular necrosis confidence interval defined daily doses hazard ratio International Classification of Disease, Ninth Revision, Clinical Modification National Health Insurance Research Database psoriatic arthritis systemic lupus erythematosus

and those who could not be matched with nonpsoriasis controls. Ultimately, data for 28,268 patients with psoriasis and 113,072 controls without psoriasis were analyzed. Patients with psoriasis. We included outpatients and inpatients with new-onset psoriasis diagnosed between January 1, 2004, and December 31, 2006. A patient with psoriasis was defined as a person who had received a dermatologist-diagnosed case of psoriasis at least twice. Psoriasis was identified by International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) codes 696.0, 696.1, and 696.8. The date of the first psoriasis diagnosis was defined as the index date, which was the start of follow-up for these patients. We classified psoriasis severity as mild and severe. Severe psoriasis was defined as disease requiring systemic antipsoriatic therapy or phototherapy at least twice during the first 3 years of follow-up. All other cases were classified as mild psoriasis. We also classified patients with psoriasis by PsA status: those who had received a PsA diagnosis at least twice during outpatient visits or at least once as an inpatient during the follow-up period were classified as PsA-positive. All other patients were classified as PsA-negative.2,28 Matched controls without psoriasis. For each patient with psoriasis, four matched controls without psoriasis were randomly selected from the Longitudinal Health Insurance Database 2000. These subjects were matched for age, sex, and medical history of diabetes mellitus, hip fracture, and alcohol dependencedall of which have been reported to be closely associated with AVN developmentdduring the preceding year.2,29 Outcome All subjects were followed from the index date until a diagnosis of AVN, death, or the end of the study period. AVN incidence was identified by ICD-9-CM code 733.4 for outpatient visits or inpatient services.

Covariates We adjusted for potential confounders by including medical history and medication use in the preceding year. Prior medical history investigated included hip dislocation (ICD-9-CM code 835), kidney transplantation (V420 and 966.81), hyperlipidemia (272), hypertension (401-405), cardiovascular disease (410-429), and chronic kidney disease (250.4, 274.1, 403, 404, 440.1, 442.1, 447.3, 572.4, 580-589, 642.1, 642.2, and 753). Prior medication use investigated included bisphosphonate, corticosteroids, methotrexate, acitretin, cyclosporine, and nonsteroidal antiinflammatory drugs. Statistical analysis The main analysis assessed time to the event of new-onset AVN. Patients without an AVN diagnosis and those who died during the study period were censored. Stratified Cox regression models were used to estimate the risk of developing AVN. The multivariate models were adjusted for medical history (hip dislocation, kidney transplantation, hyperlipidemia, hypertension, cardiovascular disease, and chronic kidney disease) during the preceding year, medication use (bisphosphonate, methotrexate, acitretin, cyclosporine, and nonsteroidal antiinflammatory drugs) during the preceding year, systemic corticosteroids use and mean daily dosage of topical potent and superpotent corticosteroids (total dosage of topical steroids divided by the prescription period) in the preceding year before index date, and follow-up period. In subgroup analysis, patients with psoriasis were classified by age, sex, PsA status, psoriasis severity, and steroid use, and risk of developing AVN was compared between the patients with psoriasis and their matched controls. Because previous studies have suggested the risk of AVN increases with a prolonged systemic corticosteroid therapy (3 months or longer),30-32 we set the cutoff for corticosteroid usage at 90 defined daily doses (DDD). Sensitivity analysis was performed by using different criteria for AVN diagnosis, including AVN diagnosed by different specialists and by imaging findings before AVN, and by excluding patients with osteoporosis or venous thromboembolism. All statistical analyses were performed using the SAS System for Windows, Version 9.3 (SAS Institute Inc, Cary, NC).

RESULTS Table I shows the baseline characteristics of the patients with psoriasis and the matched controls; 59.49% of the subjects were male. Compared with

4 Chiu et al

J AM ACAD DERMATOL

n 2016

Fig 1. Flowchart of study cohort selection. AS, Ankylosing spondylitis; AVN, avascular necrosis; HIV, human immunodeficiency virus; LHID, Longitudinal Health Insurance Database; NHIRD, National Health Insurance Research Database; PSO, psoriatic disease; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus.

controls, the patients with psoriasis were older and had higher rates of hyperlipidemia, kidney transplantation, hypertension, cardiovascular disease, obesity, and chronic kidney disease in the preceding year. In addition, patients with psoriasis had higher rates of medication use, including bisphosphonate, corticosteroid, methotrexate, acitretin, cyclosporine, and nonsteroidal antiinflammatory drug use in the preceding year. The risk of AVN was higher in patients with psoriasis than in controls (adjusted HR [aHR] 1.96, 95% confidence interval [CI] 1.62-2.38) (Fig 2). Subgroup analysis revealed that compared with matched controls, the risk for developing AVN among patients with psoriasis was significantly higher in men of all age subgroups, in those with severe and mild psoriasis, in those with and without

PsA, and in those who did and did not use oral or parenteral steroids S90 DDD. Among patients with psoriasis, the risk for AVN was higher for men (aHR 2.20, 95% CI 1.77-2.72) than for women (aHR 1.30, 95% CI 0.84-1.99), for patients younger than 30 years (aHR 2.64, 95% CI 1.24-5.61) than for those in the other age groups, for patients with severe psoriasis (aHR 3.21, 95% CI 2.09-4.92) than for those with mild psoriasis (aHR 1.75, 95% CI 1.42-2.17), for patients with PsA (aHR 4.34, 95% CI 1.80-10.45) than for those without PsA (aHR 1.88, 95% CI 1.55-2.29), and for patients who used oral or parenteral steroids S90 DDD (aHR 4.92, 95% CI 1.49-16.25) than for those who used systemic corticosteroids \90 DDD (aHR 1.94, 95% CI 1.59-2.37) (Fig 2). In sensitivity analyses that restricted AVN diagnoses to (1) cases diagnosed by imaging findings

J AM ACAD DERMATOL

Chiu et al 5

VOLUME jj, NUMBER j

Table I. Characteristics of psoriasis patients and matched controls without psoriatic disease during the year before the index date Psoriasis patients (n = 28,268) n

Demographics Age* Sex Male Female Comorbidities in preceding year Hip dislocation Hyperlipidemia Kidney transplantation Organ transplantation Hypertension Cardiovascular disease Obesity Chronic kidney disease Medication use in preceding year Bisphosphonate Systemic corticosteroids Methotrexate Acitretin Cyclosporine Nonsteroidal antiinflammatory drugs

%

46.12 6 17.99

Nonpsoriasis controls (n = 113,072) n

%

45.96 6 18.09

P value

.001

16,818 11,450

59.49 40.51

67,272 45,800

59.49 40.51

4 1758 9 14 5736 2996 135 1192

0.01 6.22 0.03 0.05 20.29 10.60 0.48 4.22

14 4821 13 40 16,508 8491 257 3276

0.01 4.26 0.01 0.04 14.60 7.51 0.23 2.90

.855 \.001 .001 .125 \.001 \.001 \.001 \.001

268 543 156 88 24 19,740

0.95 1.92 0.55 0.31 0.08 69.83

596 609 24 1 18 63,556

0.53 0.54 0.02 0 0.02 56.21

\.001 \.001 \.001 \.001 \.001 \.001

*Mean 6 SD.

obtained during the previous 180 days, (2) cases diagnosed by an orthopedic surgeon, and (3) cases that excluded diagnoses of osteoporosis or venous thromboembolism, the results were similar. Compared with controls, patients with psoriasis still had a significantly higher risk for AVN after excluding those with osteoporosis (aHR 2.11, 95% CI 1.70-2.62) and venous thromboembolism (aHR 1.98, 95% CI 1.63-2.40) (Supplemental Table I, available at http://www.jaad.org).

DISCUSSION Although the underlying mechanisms responsible for the pathogenesis of AVN are not well understood, glucocorticoids are known to predispose patients to AVN.1,2 Glucocorticoids increase the lifespan of osteoclasts by suppressing osteoblasts and osteocytes.33 Furthermore, they induce vasoconstriction in intraosseous femoral head arteries, which results in bone ischemia due to modulation of vasoactive agents, such as endothelin-1 and bradykinin.33 Previous case reports described AVN in patients with psoriasis, and long-term use of topical clobetasol propionate or betamethasone dipropionate had been proposed to be responsible for the development of AVN.25-27 However, the presence of AVN in individuals who have never received corticosteroids and the absence of AVN in most patients who receive

corticosteroids suggest that additional risk factors are involved in AVN development.16 Moreover, a previous study showed that total duration and total dose of corticosteroids did not significantly differ between patients with and without AVN.34 Our study showed that AVN risk was also elevated in psoriasis patients with no recent history of systemic glucocorticoid use (S90 DDD), after adjustment for potential confounding factors, including topical corticosteroid treatment. This, too, suggests that additional factors intrinsically linked to psoriasis were responsible for AVN development in these patients. Similarly, Fialho et al reported that disease activity, which was associated with underlying inflammation,35 was a predictor of AVN development in patients with SLE.36 One hypothesis is that systemic inflammation in autoimmune diseases, such as SLE and psoriasis, produces oxidized low-density lipoproteins, which limits the development of osteoblasts in favor of adipocytes.15 Elevated tumor necrosis factorea, a crucial mediator of psoriasis pathogenesis, was reported to cause apoptosis of osteoblast-lineage cells and increase osteoclast maturation and activity.15,37 Moreover, previous studies indicate that systemic psoriatic inflammation might cause an increase in homocysteine levels or hypercoagulability.38 This could induce thrombosis of small blood vessels

6 Chiu et al

J AM ACAD DERMATOL

n 2016

Fig 2. Adjusted hazard ratios of the risk for avascular necrosis in patients with psoriasis compared with matched controls. aAdjusted for comorbidities and medication use in the preceding year and follow-up period. bSteroid use was defined as S90 DDD of oral or parenteral glucocorticoid. aHR, Adjusted hazard ratio; CI, confidence interval; DDD, defined daily doses; PsA, psoriatic arthritis.

followed by cellular necrosis of osteocytes, ultimately leading to AVN.38 Results from the aforementioned studies15,36,37 allow us to speculate that inflammatory factors contribute to the increased risk of AVN in patients with psoriasis. Consistent with those studies, the present study also found that the AVN risk was higher in patients with severe psoriasis who had a higher burden of inflammation than in patients with mild psoriasis. This strongly suggests that inflammatory factors have a role in both psoriasis pathogenesis and AVN development. AVN risk was higher for psoriasis patients with PsA than for those without PsA. This finding was

consistent with a previous Canadian study, which found that the presence of arthritis was an independent risk factor for AVN development in patients with SLE.23 The increased risk of AVN in patients who develop arthritis is probably due to the dual burden of chronic inflammation of joints and the skin. The inflammatory process has several deleterious effects on bone tissues: intramedullary pressure is increased, marrow sinuses and capillaries are compressed, and few osteoblast precursors are available to replace demineralized bone. These conditions might result in bone resorption and necrosis.15,39 The current analysis also showed that

J AM ACAD DERMATOL

VOLUME jj, NUMBER j

the risk for AVN was higher in patients with psoriasis younger than 30 years than in older patients. This result was consistent with prior research findings indicating a higher risk for AVN in young patients with SLE. Previous reports suggest that AVN incidence is highest among young patients.34,40 Two possible explanations for this finding are that increased physical activity in young patients with psoriasis might lead to articular collapse34,40 or patients with early-onset psoriasis might have more severe intrinsic or genetic immune dysfunction and higher systemic inflammation, accelerating the development of AVN.41-43 This study has several potential limitations that should be considered. First, AVN diagnoses were based on secondary claims data. Consequently, our analyses might have underestimated the risk for AVN because only symptomatic patients underwent AVN evaluation. However, because this underestimation likely occurred in both the psoriasis and control groups, the effect would most likely have been nondifferential and biased the effect estimates toward the null. Moreover, when we conducted sensitivity analyses using a more rigorously defined outcome (eg, AVN diagnosis on the basis of preceding imaging studies) or redefined an AVN diagnosis as one made by only an orthopedic surgeon or rheumatologist, the results were similar to those of the primary analyses. A second limitation is that data on some potential factors for AVN, including tobacco use, alcohol intake, physical activity, body weight, and body-mass index, were not available in the NHIRD dataset. However, some of these unmeasured confounders might have been controlled for by including alcohol dependence, hip fracture, and other comorbidities as alternative covariates in the analyses. Third, our study used treatment with systemic therapy or phototherapy as a surrogate for severe disease. Thus, some misclassification of disease severity is likely, as is the case in all epidemiologic studies. However, the reliability and validity of these methods of classifying psoriasis severity and PsA status have been previously demonstrated.44,45 Fourth, the potential association of topical steroids with AVN in patients with psoriasis could be a concern. Our statistical models had adjusted the influence of topical and systemic corticosteroids. Nevertheless, the risk that AVN could be at least partially attributed to topical, low-potency corticosteroid use or low dosage of systemic corticosteroids use (\90 DDD) is possible. The results of this large-scale epidemiological study indicate that psoriasis is associated with increased risk of AVN. The development of AVN in patients with psoriasis was multifaceted and many

Chiu et al 7

factors seem to be at play. AVN risk was positively associated with male sex, age younger than 30 years, corticosteroid use, severe psoriasis, and concomitant PsA. Thus, clinicians should maintain a high index of suspicion for patients presenting with localized pain in a weight-bearing joint. To minimize the risk of developing AVN, cautious use of corticosteroids, whether systemic or topical, is recommended. This study is based in part on data from the National Health Insurance Research Database, which is provided by the National Health Insurance Administration Ministry of Health and Welfare and managed by the National Health Research Institutes. The interpretation and conclusions contained herein do not represent those of the National Health Insurance Administration Ministry of Health and Welfare or National Health Research Institutes. REFERENCES 1. Aldridge JM, Urbaniak JR. Avascular necrosis of the femoral head: etiology, pathophysiology, classification, and current treatment guidelines. Am J Orthop (Belle Mead NJ). 2004;33: 327-332. 2. Lafforgue P. Pathophysiology and natural history of avascular necrosis of bone. Joint Bone Spine. 2006;73:500-507. 3. Coventry MB, Beckenbaugh RD, Nolan DR, Ilstrup DM. 2,012 total hip arthroplasties. A study of postoperative course and early complications. J Bone Joint Surg Am. 1974;56:273-284. 4. Nestle FO, Kaplan DH, Barker J. Psoriasis. N Engl J Med. 2009; 361:496-509. 5. Sano S. Psoriasis as a barrier disease. Dermatologica Sinica. 2015;33:64-69. 6. Chiu H-Y, Cheng Y-P, Tsai T-F. T helper type 17 in psoriasis: from basic immunology to clinical practice. Dermatologica Sinica. 2012;30:136-141. 7. Alexandroff AB, Pauriah M, Camp RD, Lang CC, Struthers AD, Armstrong DJ. More than skin deep: atherosclerosis as a systemic manifestation of psoriasis. Br J Dermatol. 2009;161: 1-7. 8. Chiu HY, Huang HL, Li CH, et al. Increased risk of glomerulonephritis and chronic kidney disease in relation to the severity of psoriasis, concomitant medication, and comorbidity: a nationwide population-based cohort study. Br J Dermatol. 2015;173:146-154. 9. Puig L, Kirby B, Mallbris L, Strohal R. Psoriasis beyond the skin: a review of the literature on cardiometabolic and psychological co-morbidities of psoriasis. Eur J Dermatol. 2014;24:305-311. 10. Tektonidou MG, Moutsopoulos HM. Immunologic factors in the pathogenesis of osteonecrosis. Orthop Clin North Am. 2004;35:259-263. vii. 11. Kim TH, Hong JM, Oh B, et al. Association of polymorphisms in the interleukin 23 receptor gene with osteonecrosis of femoral head in Korean population. Exp Mol Med. 2008;40:418-426. 12. Shigemura T, Nakamura J, Kishida S, et al. Incidence of osteonecrosis associated with corticosteroid therapy among different underlying diseases: prospective MRI study. Rheumatology (Oxford). 2011;50:2023-2028. 13. Klippel JH, Gerber LH, Pollak L, Decker JL. Avascular necrosis in systemic lupus erythematosus. Silent symmetric osteonecroses. Am J Med. 1979;67:83-87. 14. Goldring SR, Goldring MB. Cytokines and skeletal physiology. Clin Orthop Relat Res. 1996:13-23.

J AM ACAD DERMATOL

8 Chiu et al

15. Lane NE. Therapy insight: osteoporosis and osteonecrosis in systemic lupus erythematosus. Nat Clin Pract Rheumatol. 2006; 2:562-569. 16. Faezi ST, Hoseinian AS, Paragomi P, et al. Non-corticosteroid risk factors of symptomatic avascular necrosis of bone in systemic lupus erythematosus: a retrospective case-control study. Mod Rheumatol. 2015;25:590-594. 17. Seguin C, Kassis J, Busque L, et al. Non-traumatic necrosis of bone (osteonecrosis) is associated with endothelial cell activation but not thrombophilia. Rheumatology (Oxford). 2008;47:1151-1155. 18. Mok MY, Farewell VT, Isenberg DA. Risk factors for avascular necrosis of bone in patients with systemic lupus erythematosus: is there a role for antiphospholipid antibodies? Ann Rheum Dis. 2000;59:462-467. 19. Mehta NN, Yu Y, Pinnelas R, et al. Attributable risk estimate of severe psoriasis on major cardiovascular events. Am J Med. 2011;124:775.e1-775.e6. 20. Ahlehoff O, Gislason GH, Lindhardsen J, et al. Psoriasis carries an increased risk of venous thromboembolism: a Danish nationwide cohort study. PLoS One. 2011;6:e18125. 21. D’Angelo S, Palazzi C, Gilio M, Leccese P, Padula A, Olivieri I. Improvements in diagnostic tools for early detection of psoriatic arthritis. Expert Rev Clin Immunol. 2016:1-7. 22. Gelfand JM, Gladman DD, Mease PJ, et al. Epidemiology of psoriatic arthritis in the population of the United States. J Am Acad Dermatol. 2005;53:573. 23. Gladman DD, Urowitz MB, Chaudhry-Ahluwalia V, Hallet DC, Cook RJ. Predictive factors for symptomatic osteonecrosis in patients with systemic lupus erythematosus. J Rheumatol. 2001;28:761-765. 24. Goldring SR, Gravallese EM. Mechanisms of bone loss in inflammatory arthritis: diagnosis and therapeutic implications. Arthritis Res. 2000;2:33-37. 25. Reichert-Penetrat S, Trechot P, Barbaud A, Gillet P, Schmutz JL. Bilateral femoral avascular necrosis in a man with psoriasis: responsibility of topical corticosteroids and role of cyclosporine. Dermatology. 2001;203:356-357. 26. Takahashi H, Tsuji H, Honma M, Ishida-Yamamoto A, Iizuka H. Femoral head osteonecrosis after long-term topical corticosteroid treatment in a psoriasis patient. J Dermatol. 2012;39:887-888. 27. Hogan DJ, Sibley JT, Lane PR. Avascular necrosis of the hips following longterm use of clobetasol propionate. J Am Acad Dermatol. 1986;14:515-517. 28. Lee J, Kwok SK, Jung SM, et al. Osteonecrosis of the hip in Korean patients with systemic lupus erythematosus: risk factors and clinical outcome. Lupus. 2014;23:39-45. 29. Zhu KK, Xu WD, Pan HF, et al. The risk factors of avascular necrosis in patients with systemic lupus erythematosus: a meta-analysis. Inflammation. 2014;37:1852-1864. 30. Fisher DE, Bickel WH. Corticosteroid-induced avascular necrosis. A clinical study of seventy-seven patients. J Bone Joint Surg Am. 1971;53:859-873. 31. McKee MD, Waddell JP, Kudo PA, Schemitsch EH, Richards RR. Osteonecrosis of the femoral head in men following short-course

n 2016

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

corticosteroid therapy: a report of 15 cases. CMAJ. 2001;164: 205-206. Lo PC, Li CY, Huang WF, et al. Risk of fractures in vitiligo patients treated with phototherapy-a retrospective population-based cohort study. J Dermatol Sci. 2016;82: 197-203. Kerachian MA, Seguin C, Harvey EJ. Glucocorticoids in osteonecrosis of the femoral head: a new understanding of the mechanisms of action. J Steroid Biochem Mol Biol. 2009; 114:121-128. Smith FE, Sweet DE, Brunner CM, Davis JS. Avascular necrosis in SLE. An apparent predilection for young patients. Ann Rheum Dis. 1976;35:227-232. Lin J, Li N, Chen H, Liu C, Yang B, Ou Q. Serum Cyr61 is associated with clinical disease activity and inflammation in patients with systemic lupus erythematosus. Medicine (Baltimore). 2015;94:e834. Fialho SC, Bonfa E, Vitule LF, et al. Disease activity as a major risk factor for osteonecrosis in early systemic lupus erythematosus. Lupus. 2007;16:239-244. Lee C-H, Hwang ST-Y. Pathophysiology of chemokines and chemokine receptors in dermatological science: a focus on psoriasis and cutaneous T-cell lymphoma. Dermatologica Sinica. 2012;30:128-135. Karabudak O, Ulusoy RE, Erikci AA, Solmazgul E, Dogan B, Harmanyeri Y. Inflammation and hypercoagulable state in adult psoriatic men. Acta Derm Venereol. 2008;88:337-340. Conte Neto N, Bastos AS, Chierici-Marcantonio RA, Marcantonio E Jr. Is rheumatoid arthritis a risk factor for oral bisphosphonate-induced osteonecrosis of the jaws? Med Hypotheses. 2011;77:905-911. Nakamura J, Saisu T, Yamashita K, Suzuki C, Kamegaya M, Takahashi K. Age at time of corticosteroid administration is a risk factor for osteonecrosis in pediatric patients with systemic lupus erythematosus: a prospective magnetic resonance imaging study. Arthritis Rheum. 2010;62:609-615. Ferrandiz C, Pujol RM, Garcia-Patos V, Bordas X, Smandia JA. Psoriasis of early and late onset: a clinical and epidemiologic study from Spain. J Am Acad Dermatol. 2002; 46:867-873. Gudjonsson JE, Karason A, Runarsdottir EH, et al. Distinct clinical differences between HLA-Cw*0602epositive and enegative psoriasis patientsean analysis of 1019 HLA-Ce and HLA-Betyped patients. J Invest Dermatol. 2006;126: 740-745. Golden JB, Wang Y, Fritz Y, et al. Chronic, not acute, skin-specific inflammation promotes thrombosis in psoriasis murine models. J Transl Med. 2015;13:382. Chiu HY, Chang WL, Huang WF, Wen YW, Tsai YW, Tsai TF. Increased risk of arrhythmia in patients with psoriatic disease: a nationwide population-based matched cohort study. J Am Acad Dermatol. 2015;73:429-438. Asgari MM, Wu JJ, Gelfand JM, et al. Validity of diagnostic codes and prevalence of psoriasis and psoriatic arthritis in a managed care population, 1996-2009. Pharmacoepidemiol Drug Saf. 2013;22:842-849.

J AM ACAD DERMATOL

VOLUME jj, NUMBER j

Supplemental Table I. Results of sensitivity analyses No. (%) of events

Total follow-up (PY)

Mean duration of follow-up (PY)

Primary model Psoriasis patients 172 (0.61) 185,028.56 6.55 Nonpsoriasis subjects 315 (0.28) 767,752.62 6.79 Primary model restricted to avascular necrosis diagnosed by orthopedic surgeon Psoriasis patients 154 (0.55) 184,968.04 6.55 Nonpsoriasis subjects 268 (0.24) 767,562.89 6.79 Diagnosis of avascular necrosis with imaging exam during prior 180 daysy Psoriasis patients 172 (0.61) 185,028.56 6.55 Nonpsoriasis subjects 315 (0.28) 767,752.62 6.79 Primary model excluding patients with osteoporosis Psoriasis patients 138 (0.53) 170,128.38 6.56 Nonpsoriasis subjects 233 (0.22) 706,394.65 6.81 Primary model excluding patients with venous thromboembolism Psoriasis patients 167 (0.60) 183,136.41 6.55 Nonpsoriasis subjects 307 (0.27) 761,518.69 6.79

Incidence (1000 PY)

HR

95% CI

aHR*

95% CI

6 1.13 6 0.95

0.93 0.41

2.29 1

1.90-2.76 Reference

1.96 1

1.62-2.38 Reference

6 1.13 6 0.95

0.83 0.35

2.41 1

1.98-2.94 Reference

2.10 1

1.72-2.58 Reference

6 1.13 6 0.95

0.93 0.46

2.29 1

1.90-2.76 Reference

1.96 1

1.62-2.38 Reference

2.50 1

2.02-3.10 Reference

2.11 1

1.70-2.62 Reference

2.29 1

1.90-2.77 Reference

1.98 1

1.63-2.40 Reference

6 1.10 6 0.93 6 1.12 6 0.95

0.91 0.40

aHR, Adjusted hazard ratio; CI, confidence interval; HR, hazard ratio; PY, person-years. *Adjusted for cardiovascular conditions, comorbidities, and medication use in preceding year and follow-up period. y Includes imaging studies such as computed tomography, magnetic resonance imaging, bone scan, or radiography performed within 180 days before diagnosis of avascular necrosis.

Chiu et al 8.e1