https://orcid.org/0000-0001-7199-6851
Abstract
Avascular necrosis (AVN) is a known adverse event associated with corticosteroid (CS) usage. Inflammatory bowel disease (IBD) is often treated with a CS for induction of remission. We sought to describe clinical features and outcomes of IBD patients with AVN.
In this retrospective, single-center, case-control study, patients with IBD who had a diagnosis of osteonecrosis, aseptic necrosis, or AVN from 1976 to 2009 were included, and each was matched with up to 2 controls (IBD but no AVN) on age, sex, IBD subtype, geographic area of residence, and date of IBD diagnosis. We abstracted risk factor data from the medical records. Conditional logistic regression was performed accounting for minor differences in age and date of first IBD visit to assess the relationship between putative risk factors and AVN, expressed as odds ratio and 95% confidence interval.
Eighty-five patients were diagnosed with IBD-AVN and were matched with 163 controls. The mean age at AVN diagnosis was 47.5 years. AVN was diagnosed a median of 12.2 years after IBD diagnosis, and the control group was followed for a median of 15 years after IBD diagnosis to ensure that they did not have AVN. Ten percent of patients with AVN did not have any CS exposure. History of arthropathy or estrogen use in Crohn’s disease and use of CS, osteoporosis, and history of arthropathy in ulcerative colitis were significantly associated with AVN.
Most patients with IBD-AVN had multifocal involvement. Most had received CS, but many patients had other risk factors including arthropathy.
Lay Summary
This single-center, case-control study of inflammatory bowel disease patients with osteonecrosis showed that while corticosteroid use was likely a risk factor, especially among ulcerative colitis patients, other risk factors included estrogen use among Crohn’s disease patients, arthropathy, and osteoporosis.
Although it is understood that most inflammatory bowel disease patients who develop osteonecrosis have used corticosteroids, other risk factors for this complication are not well studied in the inflammatory bowel disease population.
In this case-control study, we found that about 10% of inflammatory bowel disease patients with avascular necrosis had no history of corticosteroid use; in addition, severe/fulminant disease activity, arthropathy, osteoporosis, and obesity appeared to be risk factors.
Recognition of other risk factors for osteonecrosis may help identify inflammatory bowel disease patients who are at higher risk for this complication.
Introduction
Osteonecrosis, also known as avascular necrosis (AVN), aseptic necrosis, or ischemic necrosis, is a severe destruction of the architectural structure of articular bone that often requires joint replacement. Trauma, sickle cell anemia, diabetes mellitus, obesity, alcohol consumption, cigarette smoking, and use of corticosteroids (CSs) have been associated with AVN.1,2 Recently, a population-based cohort study showed that high-dose CS use could double the risk of AVN in adult patients with chronic inflammatory diseases such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and inflammatory bowel disease (IBD), but CSs were not associated with AVN at lower doses.3 Furthermore, the association between CSs and AVN that has been reported in cross-sectional studies was much stronger than what has been found in longitudinal prospective studies.4 In addition to inflammatory disease, AVN was observed in patients who had renal transplantation, and malignant lymphoma, so it is not clear whether AVN is the result of CS therapy alone or the underlying disease process itself.4-7
The prevalence of AVN has been reported to be 2% to 5% in patients in IBD.8,9 Initially, it was believed that AVN in IBD was exclusively due to CS therapy in these patients, but later studies failed to show the causative link between IBD, AVN, and CS therapy, or any dose–response relationship between CS and AVN.10-12 Furthermore, there is no clear time window between the time of initiation of CS and development of AVN, and the association between AVN and the duration of CS therapy is not well understood.13,14 In comparison with other inflammatory diseases such as SLE, AVN could develop in younger IBD patients, even with a lower threshold of cumulative CS dose, or after a shorter duration of CS therapy.15,16 In addition, some studies have suggested that AVN is a rare extraintestinal manifestation of IBD, rather than a complication of CS.17 The patients who were reported to have AVN with no history of exposure to CS would support this latter hypothesis.18
Aside from CS therapy, other risk factors of AVN have not been well studied in the IBD population. In this study, we aimed to analyze the possible risk factors of acquiring AVN in patients with IBD during the 30-year study period.
Methods
Case Ascertainment
This study was a retrospective case-control study using a centralized diagnostic index. Following approval from the Institutional Review Board of the Mayo Clinic, the diagnostic index was used to identify all patients with IBD who had a diagnosis of osteonecrosis, aseptic necrosis, or AVN from January 1, 1976, to January 1, 2009. Medical records for patients who provided research authorization were reviewed to confirm the diagnosis of IBD and AVN. The diagnosis of IBD was made on the basis of 1 of the following criteria: (1) pathology (either resection specimen or mucosal biopsy) consistent with IBD (eg, chronic colitis or presence of granulomas), (2) endoscopy consistent with IBD (eg, loss of vascular pattern, erythema, granularity, friability, erosions, ulcerations in colon or ileum), or (3) radiology consistent with IBD (eg, ileal thickening/narrowing on small bowel follow-through or computed tomography enterography or magnetic resonance enterography, colonic mucosal ulceration on double-contrast barium enema). The diagnosis of AVN was determined by 1 of the following criteria: (1) plain films consistent with osteonecrosis (eg, sclerosis or cystic degeneration of joint surface, subchondral radiolucency, loss of bone surface contour), (2) magnetic resonance imaging consistent with osteonecrosis (eg, early bone ischemia on T1-weighted images, sclerosis on T2-weighted images), or (3) pathology consistent with osteonecrosis (eg, interstitial edema/plasmostasis of marrow, medullary or trabecular necrosis, medullary fibrosis with new bone formation). The Ficat and Arlet staging system was used to classify osteonecrosis (Supplementary Table 1).19-21 All of the diagnostic studies must have been performed or reviewed at Mayo Clinic. Patients without a diagnosis of IBD or AVN confirmed or reviewed at Mayo Clinic Rochester were excluded. Patients who had AVN before the diagnosis of IBD were also excluded.
To analyze whether AVN directly resulted from CS therapy or could potentially be a rare extraintestinal manifestation of IBD, each IBD patient with AVN (IBD-AVN) was matched with up to 2 control subjects, patients with IBD without AVN (IBD-control). The controls were matched to cases by sex, IBD subtype, geographic area of residence, date of IBD diagnosis at Mayo Clinic, and age at Mayo IBD diagnosis. The follow-up of the controls was matched to the interval between the first Mayo IBD diagnosis and AVN diagnosis in cases, too.
Data Collection
Details pertaining to patients’ basic demographic data, IBD subtype (Crohn’s disease [CD] or ulcerative colitis [UC]), duration of IBD, intestinal and extraintestinal extent of disease (arthropathy, aphthous stomatitis, erythema nodosum, pyoderma gangrenosum, primary sclerosing cholangitis, ankylosing spondylitis, and eye disease), severity of the disease, IBD medication, need for total parenteral nutrition, surgery, and the date of the last follow-up were obtained. Disease activity of IBD was adapted from the American College of Gastroenterology clinical guidelines for management of UC and CD.22,23 For CD, the definition of severe/fulminant disease included patients with severe symptoms despite corticosteroids or biologics for outpatients, or inpatients with intestinal obstruction, peritonitis, or abscess. For UC, this included patients with high stool frequency, significant anemia, elevated C-reactive protein, and Mayo endoscopic subscore of 3 (ulcerations, spontaneous bleeding). Peak activity, from time of IBD diagnosis to time of index date, was recorded. More detailed information on previous CS therapy, including stop and start dates and dosages and route of administration, were abstracted from the medical records. Hematological malignancies including leukemia and lymphoma, other hematological disorders including factor V Leiden mutation, vitamin B12 deficiency, hereditary spherocytosis, and other malignancies were captured. Medical records were also reviewed for rheumatological disorders, including RA, psoriatic arthritis, vasculitis, polymyalgia rheumatica, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, sarcoidosis, and SLE. The medical records of IBD-AVN patients were reviewed for the location of AVN and its different treatment including arthroplasty, core decompression, bone graft, osteotomy, resurfacing, arthrodesis, CS injection, use of opioid analgesics, and conservative treatments. We also abstracted charts for other causes of AVN, including estrogen usage, obesity (body mass index >30 kg/m2), fractures, alcoholism, and venous thromboembolism. The findings from bone mass density scan were used to define osteoporosis.
Statistical Analysis
Descriptive statistics are reported for baseline characteristics of patients. Categorical data were expressed as number and percentage and compared using the Pearson chi-square test or Fisher exact test. Continuous variables were reported as median (interquartile range [IQR]).
Univariable conditional logistic regression models, adjusted for the matching variables of age at IBD diagnosis and date of first visit to Mayo for IBD which were not exactly matched, were used to assess the odds of IBD status (case vs control) within the 2 IBD subtypes. This was done again as a single assessment including all IBD (not by subtype). Results were reported as odds ratio with 95% confidence interval.
For all IBD as a whole, not by subtype, a multiple variable conditional logistic model was examined. Variables with a univariate P value ≤.1 and a risk factor prevalence of ≥5% were considered. A backward selection method retained those variables in a parsimonious model that were statistically significant (P < .05), again adjusting for age at Mayo IBD diagnosis and date of first visit to Mayo for IBD. Analysis was completed using SAS version 9.4 (SAS Institute, Cary, NC, USA).
Results
Eighty-five patients were diagnosed with IBD-AVN from 1976 to 2009 and were matched to 163 patients with IBD-controls. Among the patients with IBD-AVN, 78 patients matched with 2 and 7 patients matched with 1 control subject. Mean age at AVN diagnosis was 47.5 ± 18.7 years. AVN was diagnosed a median of 12.2 (IQR, 5.7-21.8) years after diagnosis of IBD. The entire control group was followed for a median of 15 (IQR, 9.7-23.7) years after diagnosis of IBD, and for a median of 12.4 years after first Mayo visit for IBD (IQR; 6.7-18.2) to ensure that they did not develop AVN. Fistulizing disease development was not significantly different between the cases and controls (41.3% vs 45.5%; P > 0.1). Perianal fistula was the most common location for fistulizing disease followed by rectovaginal and rectal. A total of 209 (84.3%) patients, 90.6% of cases and 81.6% of controls, were treated with systemic CS, with a median cumulative duration of 4.3 (IQR, 2.1-14.0) months. The median cumulative CS dose was 2400 (IQR, 1200-5650) mg, and the average daily dose was 21.3 ± 13.3 mg. The median maximum daily dose among the patients was 40 (IQR, 20-60) mg. CS was administered in oral and intravenous forms in 206 (98%) and 59 (28.2%) patients, respectively. Prior and current arthropathy, obesity, osteoporosis, having at least 1 extraintestinal manifestation (specifically stomatitis and erythema nodosum), liver transplantation, history of thrombotic events, and use of total parenteral nutrition (TPN) were significantly higher in the IBD-AVN cases compared with controls (Table 1).
Clinical features of IBD-AVN cases and matched IBD controls
| Clinical Feature | Cases (n = 85) | Controls (n = 163) | P Value | ||
|---|---|---|---|---|---|
| CD | UC | CD | UC | ||
| Subtype number | 46 | 39 | 86 | 77 | .84a |
| Sex | |||||
| • Male | 24 | 28 | 48 | 56 | .68a |
| • Female | 22 | 11 | 38 | 21 | |
| Alcohol consumption | 4 (8.7) | 3 (7.7) | 3 (3.5) | 4 (5.2) | .21b |
| Smoking | 24 (52.2) | 13 (33.3) | 41 (47.7) | 25 (32.5) | .64a |
| Obesity | 4 (8.7) | 2 (5.1) | 1 (1.2) | 2 (2.6) | .04b |
| Systemic steroids | 39 (84.8) | 38 (97.4) | 70 (81.4) | 62 (80.5) | .05a |
| Arthropathy | 9 (19.6) | 7 (17.9) | 3 (3.5) | 8 (10.4) | .00a |
| Extraintestinal manifestation | 12 (26.1) | 15 (38.5) | 12 (14.0) | 16 (20.8) | .01a |
| Stomatitis | 4 (8.7) | 2 (5.1) | 2 (2.3) | 0 (0.0) | .01b |
| Erythema nodosum | 3 (6.5) | 1 (2.6) | 0 (0.0) | 1 (1.3) | .03b |
| Pyoderma gangrenosum | 0 (0.0) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .50b |
| PSC | 0 (0.0) | 10 (25.6) | 6 (7.0) | 8 (10.4) | .42a |
| Spondylitis | 2 (4.3) | 0 (0.0) | 4 (4.7) | 0 (0.0) | .10b |
| Eye manifestation | 2 (4.3) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .23b |
| Hematological malignanciesc | 1 (2.2) | 2 (5.1) | 0 (0.0) | 2 (2.6) | .22b |
| Other hematological disordersd | 2 (4.3) | 3 (7.7) | 0 (0.0) | 0 (0.0) | .00b |
| SLE | 0 (0.0) | 1 (2.6) | 0 (0.0) | 0 (0.0) | .17b |
| Other rheumatological disordere | 3 (6.5) | 2 (5.1) | 4 (4.7) | 6 (7.8) | .94a |
| Estrogen consumption | 9 (19.6) | 2 (5.1) | 10 (11.6) | 4 (5.2) | .28a |
| Estrogen in females only | 9 (40.9) | 2 (18.2) | 8 (21.1) | 4 (19.0) | .17a |
| Thrombotic events | 8 (17.4) | 5 (12.8) | 6 (7.0) | 4 (5.2) | .02a |
| Renal transplantation | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Liver transplantation | 1 (2.2) | 10 (25.6) | 3 (3.5) | 3 (3.9) | .01a |
| High lipid profile | 10 (21.7) | 5 (12.8) | 15 (17.4) | 12 (15.6) | .83a |
| Presence prior arthropathy | 15 (32.6) | 14 (35.9) | 9 (10.5) | 17 (22.1) | .00a |
| Presence of osteoporosis | 8 (17.4) | 13 (33.3) | 8 (9.3) | 4 (5.2) | .00a |
| Presence of vasculitis | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Other malignancies | 4 (8.7) | 6 (15.4) | 10 (11.6) | 11 (14.3) | .80a |
| TPN | 6 (13.0) | 4 (10.3) | 3 (3.5) | 3 (3.9) | .01a |
| Clinical Feature | Cases (n = 85) | Controls (n = 163) | P Value | ||
|---|---|---|---|---|---|
| CD | UC | CD | UC | ||
| Subtype number | 46 | 39 | 86 | 77 | .84a |
| Sex | |||||
| • Male | 24 | 28 | 48 | 56 | .68a |
| • Female | 22 | 11 | 38 | 21 | |
| Alcohol consumption | 4 (8.7) | 3 (7.7) | 3 (3.5) | 4 (5.2) | .21b |
| Smoking | 24 (52.2) | 13 (33.3) | 41 (47.7) | 25 (32.5) | .64a |
| Obesity | 4 (8.7) | 2 (5.1) | 1 (1.2) | 2 (2.6) | .04b |
| Systemic steroids | 39 (84.8) | 38 (97.4) | 70 (81.4) | 62 (80.5) | .05a |
| Arthropathy | 9 (19.6) | 7 (17.9) | 3 (3.5) | 8 (10.4) | .00a |
| Extraintestinal manifestation | 12 (26.1) | 15 (38.5) | 12 (14.0) | 16 (20.8) | .01a |
| Stomatitis | 4 (8.7) | 2 (5.1) | 2 (2.3) | 0 (0.0) | .01b |
| Erythema nodosum | 3 (6.5) | 1 (2.6) | 0 (0.0) | 1 (1.3) | .03b |
| Pyoderma gangrenosum | 0 (0.0) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .50b |
| PSC | 0 (0.0) | 10 (25.6) | 6 (7.0) | 8 (10.4) | .42a |
| Spondylitis | 2 (4.3) | 0 (0.0) | 4 (4.7) | 0 (0.0) | .10b |
| Eye manifestation | 2 (4.3) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .23b |
| Hematological malignanciesc | 1 (2.2) | 2 (5.1) | 0 (0.0) | 2 (2.6) | .22b |
| Other hematological disordersd | 2 (4.3) | 3 (7.7) | 0 (0.0) | 0 (0.0) | .00b |
| SLE | 0 (0.0) | 1 (2.6) | 0 (0.0) | 0 (0.0) | .17b |
| Other rheumatological disordere | 3 (6.5) | 2 (5.1) | 4 (4.7) | 6 (7.8) | .94a |
| Estrogen consumption | 9 (19.6) | 2 (5.1) | 10 (11.6) | 4 (5.2) | .28a |
| Estrogen in females only | 9 (40.9) | 2 (18.2) | 8 (21.1) | 4 (19.0) | .17a |
| Thrombotic events | 8 (17.4) | 5 (12.8) | 6 (7.0) | 4 (5.2) | .02a |
| Renal transplantation | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Liver transplantation | 1 (2.2) | 10 (25.6) | 3 (3.5) | 3 (3.9) | .01a |
| High lipid profile | 10 (21.7) | 5 (12.8) | 15 (17.4) | 12 (15.6) | .83a |
| Presence prior arthropathy | 15 (32.6) | 14 (35.9) | 9 (10.5) | 17 (22.1) | .00a |
| Presence of osteoporosis | 8 (17.4) | 13 (33.3) | 8 (9.3) | 4 (5.2) | .00a |
| Presence of vasculitis | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Other malignancies | 4 (8.7) | 6 (15.4) | 10 (11.6) | 11 (14.3) | .80a |
| TPN | 6 (13.0) | 4 (10.3) | 3 (3.5) | 3 (3.9) | .01a |
Values are n or n (%).
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aChi-square test.
bFisher exact test.
cIncludes leukemia and lymphoma.
dIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
eIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
Clinical features of IBD-AVN cases and matched IBD controls
| Clinical Feature | Cases (n = 85) | Controls (n = 163) | P Value | ||
|---|---|---|---|---|---|
| CD | UC | CD | UC | ||
| Subtype number | 46 | 39 | 86 | 77 | .84a |
| Sex | |||||
| • Male | 24 | 28 | 48 | 56 | .68a |
| • Female | 22 | 11 | 38 | 21 | |
| Alcohol consumption | 4 (8.7) | 3 (7.7) | 3 (3.5) | 4 (5.2) | .21b |
| Smoking | 24 (52.2) | 13 (33.3) | 41 (47.7) | 25 (32.5) | .64a |
| Obesity | 4 (8.7) | 2 (5.1) | 1 (1.2) | 2 (2.6) | .04b |
| Systemic steroids | 39 (84.8) | 38 (97.4) | 70 (81.4) | 62 (80.5) | .05a |
| Arthropathy | 9 (19.6) | 7 (17.9) | 3 (3.5) | 8 (10.4) | .00a |
| Extraintestinal manifestation | 12 (26.1) | 15 (38.5) | 12 (14.0) | 16 (20.8) | .01a |
| Stomatitis | 4 (8.7) | 2 (5.1) | 2 (2.3) | 0 (0.0) | .01b |
| Erythema nodosum | 3 (6.5) | 1 (2.6) | 0 (0.0) | 1 (1.3) | .03b |
| Pyoderma gangrenosum | 0 (0.0) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .50b |
| PSC | 0 (0.0) | 10 (25.6) | 6 (7.0) | 8 (10.4) | .42a |
| Spondylitis | 2 (4.3) | 0 (0.0) | 4 (4.7) | 0 (0.0) | .10b |
| Eye manifestation | 2 (4.3) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .23b |
| Hematological malignanciesc | 1 (2.2) | 2 (5.1) | 0 (0.0) | 2 (2.6) | .22b |
| Other hematological disordersd | 2 (4.3) | 3 (7.7) | 0 (0.0) | 0 (0.0) | .00b |
| SLE | 0 (0.0) | 1 (2.6) | 0 (0.0) | 0 (0.0) | .17b |
| Other rheumatological disordere | 3 (6.5) | 2 (5.1) | 4 (4.7) | 6 (7.8) | .94a |
| Estrogen consumption | 9 (19.6) | 2 (5.1) | 10 (11.6) | 4 (5.2) | .28a |
| Estrogen in females only | 9 (40.9) | 2 (18.2) | 8 (21.1) | 4 (19.0) | .17a |
| Thrombotic events | 8 (17.4) | 5 (12.8) | 6 (7.0) | 4 (5.2) | .02a |
| Renal transplantation | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Liver transplantation | 1 (2.2) | 10 (25.6) | 3 (3.5) | 3 (3.9) | .01a |
| High lipid profile | 10 (21.7) | 5 (12.8) | 15 (17.4) | 12 (15.6) | .83a |
| Presence prior arthropathy | 15 (32.6) | 14 (35.9) | 9 (10.5) | 17 (22.1) | .00a |
| Presence of osteoporosis | 8 (17.4) | 13 (33.3) | 8 (9.3) | 4 (5.2) | .00a |
| Presence of vasculitis | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Other malignancies | 4 (8.7) | 6 (15.4) | 10 (11.6) | 11 (14.3) | .80a |
| TPN | 6 (13.0) | 4 (10.3) | 3 (3.5) | 3 (3.9) | .01a |
| Clinical Feature | Cases (n = 85) | Controls (n = 163) | P Value | ||
|---|---|---|---|---|---|
| CD | UC | CD | UC | ||
| Subtype number | 46 | 39 | 86 | 77 | .84a |
| Sex | |||||
| • Male | 24 | 28 | 48 | 56 | .68a |
| • Female | 22 | 11 | 38 | 21 | |
| Alcohol consumption | 4 (8.7) | 3 (7.7) | 3 (3.5) | 4 (5.2) | .21b |
| Smoking | 24 (52.2) | 13 (33.3) | 41 (47.7) | 25 (32.5) | .64a |
| Obesity | 4 (8.7) | 2 (5.1) | 1 (1.2) | 2 (2.6) | .04b |
| Systemic steroids | 39 (84.8) | 38 (97.4) | 70 (81.4) | 62 (80.5) | .05a |
| Arthropathy | 9 (19.6) | 7 (17.9) | 3 (3.5) | 8 (10.4) | .00a |
| Extraintestinal manifestation | 12 (26.1) | 15 (38.5) | 12 (14.0) | 16 (20.8) | .01a |
| Stomatitis | 4 (8.7) | 2 (5.1) | 2 (2.3) | 0 (0.0) | .01b |
| Erythema nodosum | 3 (6.5) | 1 (2.6) | 0 (0.0) | 1 (1.3) | .03b |
| Pyoderma gangrenosum | 0 (0.0) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .50b |
| PSC | 0 (0.0) | 10 (25.6) | 6 (7.0) | 8 (10.4) | .42a |
| Spondylitis | 2 (4.3) | 0 (0.0) | 4 (4.7) | 0 (0.0) | .10b |
| Eye manifestation | 2 (4.3) | 0 (0.0) | 1 (1.2) | 0 (0.0) | .23b |
| Hematological malignanciesc | 1 (2.2) | 2 (5.1) | 0 (0.0) | 2 (2.6) | .22b |
| Other hematological disordersd | 2 (4.3) | 3 (7.7) | 0 (0.0) | 0 (0.0) | .00b |
| SLE | 0 (0.0) | 1 (2.6) | 0 (0.0) | 0 (0.0) | .17b |
| Other rheumatological disordere | 3 (6.5) | 2 (5.1) | 4 (4.7) | 6 (7.8) | .94a |
| Estrogen consumption | 9 (19.6) | 2 (5.1) | 10 (11.6) | 4 (5.2) | .28a |
| Estrogen in females only | 9 (40.9) | 2 (18.2) | 8 (21.1) | 4 (19.0) | .17a |
| Thrombotic events | 8 (17.4) | 5 (12.8) | 6 (7.0) | 4 (5.2) | .02a |
| Renal transplantation | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Liver transplantation | 1 (2.2) | 10 (25.6) | 3 (3.5) | 3 (3.9) | .01a |
| High lipid profile | 10 (21.7) | 5 (12.8) | 15 (17.4) | 12 (15.6) | .83a |
| Presence prior arthropathy | 15 (32.6) | 14 (35.9) | 9 (10.5) | 17 (22.1) | .00a |
| Presence of osteoporosis | 8 (17.4) | 13 (33.3) | 8 (9.3) | 4 (5.2) | .00a |
| Presence of vasculitis | 0 (0.0) | 0 (0.0) | 1 (1.2) | 1 (1.3) | .31b |
| Other malignancies | 4 (8.7) | 6 (15.4) | 10 (11.6) | 11 (14.3) | .80a |
| TPN | 6 (13.0) | 4 (10.3) | 3 (3.5) | 3 (3.9) | .01a |
Values are n or n (%).
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aChi-square test.
bFisher exact test.
cIncludes leukemia and lymphoma.
dIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
eIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
A total of 222 (90%) patients received other medications for IBD, including sulfasalazine, mesalamine, azathioprine, 6-mercaptopurine, and anti-tumor necrosis factor agents. Specifically, 96% of the IBD-AVN cases were treated with other IBD medications. Sulfasalazine/5-aminosalicylate was given to 89%, while 41% of patients were treated with immune modulators, and 16.5% were on biologics at the time of AVN. In the IBD control group, 86% of patients were treated with other IBD medications, including 71% with sulfasalazine/5-aminosalicylate, 43% with immunomodulators, and 25% with biologics.
In total, more than 60% of the patients (62.4% in IBD-AVN and 63.0% in IBD) required at least 1 surgery related to IBD.
The median number of bones involved with AVN was 2 (range, 1-10). The femoral head was the most common site of involvement, followed by the distal femur and humeral head (Figure 1). These patients had an average Ficat and Arlet staging score of 2 (IQR, 2-3). In more than half of the patients with IBD-AVN, hip joints were involved. Knees, followed by shoulders, were the second and third most common joints involved with AVN, respectively.
Distribution of osteonecrosis in patients with inflammatory bowel disease who had avascular necrosis.
There were 46 CD-AVN cases who were matched with 86 CD-controls. CD extent was ileocolonic in 59 (44.7%) patients, ileal in 38 (28.8%), colonic in 29 (22.0%), and upper and lower gastrointestinal in 6 (4.5%). Extent of CD was not associated with increased risk of AVN (P = .3). However, having severe/fulminant CD was strongly associated with AVN (odds ratio, 84.9; 95% confidence interval, 4.1->1000; P = .004). In CD, a history of arthropathy was associated with a 5-fold higher risk of AVN. Nine (40.9%) of 22 female patients with CD-AVN and 8 (21.1%) of 38 female CD-controls were using estrogen, which was associated with a 7-fold increase in the risk of AVN (Table 2).
Univariate assessment of putative risk factors among IBD-AVN by subtype (cases) and their respective IBD controls
| Risk Factor | CD | UC | ||
|---|---|---|---|---|
| Odds Ratio (95%CI) | P Value | Odds Ratio (95%CI) | P Value | |
| Alcohol consumption | 2.7 (0.6-12.5) | .21 | 1.5 (0.3-7.0) | .62 |
| Smoking | 1.3 (0.6-2.7) | .54 | 1.1 (0.5-2.7) | .80 |
| Obesity | 8.4 (0.9-75.4) | .06 | 2.2 (0.3-17.1) | .45 |
| Systemic steroids | 1.3 (0.4-3.6) | .64 | 10.8 (1.4-86.4) | .02 |
| Arthropathy | ∞ (6.5-∞) | <.0001 | 2.2 (0.6-8.2) | .26 |
| Extraintestinal manifestation | 2.6 (0.9-7.4) | .07 | 2.3(0.9-5.8) | .07 |
| Aphthous stomatitis | 4.3 (0.7-25.4) | .10 | ∞ (1.1-∞) | .05 |
| Erythema nodosum | ∞ (2.2-∞) | .01 | 2.3 (0.1-37.5) | .57 |
| Pyoderma gangrenosum | 0 (0-12.4) | .37 | — | — |
| PSC | 0 (0-0.6) | .02 | 3.0 (1.0-9.0) | .06 |
| Spondylitis | 0.8 (0.1-4.6) | .82 | — | — |
| Eye manifestation | ∞ (0.6-∞) | .09 | — | — |
| Hematological malignanciesa | ∞ (1.3-∞) | .13 | 2.8 (0.3-21.8) | .34 |
| Other hematological disordersb | ∞ (1.06-∞) | .04 | ∞ (4.5-∞) | <.01 |
| SLE | — | — | ∞ (0.7-∞) | .07 |
| Other rheumatological disorderc | 1.5 (0.3-7.1) | .58 | 0.6 (0.1-3.4) | .57 |
| Estrogen consumption | 4.9 (0.9-27.4) | .07 | 1.0 (0.1-6.4) | .98 |
| Estrogen consumption in female only | 7.1 (1.1-45.7) | .04 | 1.0 (0.1-6.3) | 1.00 |
| Thrombotic events | 3.3 (0.9-11.4) | .06 | 2.4 (0.6-9.4) | .22 |
| Renal transplantation | 0 (0-11.5) | .36 | 0 (0-8.7) | .31 |
| Liver transplantation | 0.6 (0.1-6.1) | .68 | ∞ (6.4-∞) | <.0001 |
| High lipid profile | 1.4 (0.5-3.7) | .47 | 0.8 (0.3-2.4) | .71 |
| Presence prior arthropathy | 5.1 (1.6-15.8) | .01 | 3.13 (1.1-9.0) | .03 |
| Presence of osteoporosis | 2.3 (0.7-7.5) | .18 | 6.2 (2.0-19.7) | <.01 |
| Presence of vasculitis | 0 (0-12.4) | .37 | 0 (0-18.7) | .45 |
| Other malignancies | 0.8 (0.2-2.7) | .70 | 1.2 (0.4-3.7) | .80 |
| TPN | 3.9 (1.0-16.0) | .06 | 3.2 (0.5-18.7) | .20 |
| Risk Factor | CD | UC | ||
|---|---|---|---|---|
| Odds Ratio (95%CI) | P Value | Odds Ratio (95%CI) | P Value | |
| Alcohol consumption | 2.7 (0.6-12.5) | .21 | 1.5 (0.3-7.0) | .62 |
| Smoking | 1.3 (0.6-2.7) | .54 | 1.1 (0.5-2.7) | .80 |
| Obesity | 8.4 (0.9-75.4) | .06 | 2.2 (0.3-17.1) | .45 |
| Systemic steroids | 1.3 (0.4-3.6) | .64 | 10.8 (1.4-86.4) | .02 |
| Arthropathy | ∞ (6.5-∞) | <.0001 | 2.2 (0.6-8.2) | .26 |
| Extraintestinal manifestation | 2.6 (0.9-7.4) | .07 | 2.3(0.9-5.8) | .07 |
| Aphthous stomatitis | 4.3 (0.7-25.4) | .10 | ∞ (1.1-∞) | .05 |
| Erythema nodosum | ∞ (2.2-∞) | .01 | 2.3 (0.1-37.5) | .57 |
| Pyoderma gangrenosum | 0 (0-12.4) | .37 | — | — |
| PSC | 0 (0-0.6) | .02 | 3.0 (1.0-9.0) | .06 |
| Spondylitis | 0.8 (0.1-4.6) | .82 | — | — |
| Eye manifestation | ∞ (0.6-∞) | .09 | — | — |
| Hematological malignanciesa | ∞ (1.3-∞) | .13 | 2.8 (0.3-21.8) | .34 |
| Other hematological disordersb | ∞ (1.06-∞) | .04 | ∞ (4.5-∞) | <.01 |
| SLE | — | — | ∞ (0.7-∞) | .07 |
| Other rheumatological disorderc | 1.5 (0.3-7.1) | .58 | 0.6 (0.1-3.4) | .57 |
| Estrogen consumption | 4.9 (0.9-27.4) | .07 | 1.0 (0.1-6.4) | .98 |
| Estrogen consumption in female only | 7.1 (1.1-45.7) | .04 | 1.0 (0.1-6.3) | 1.00 |
| Thrombotic events | 3.3 (0.9-11.4) | .06 | 2.4 (0.6-9.4) | .22 |
| Renal transplantation | 0 (0-11.5) | .36 | 0 (0-8.7) | .31 |
| Liver transplantation | 0.6 (0.1-6.1) | .68 | ∞ (6.4-∞) | <.0001 |
| High lipid profile | 1.4 (0.5-3.7) | .47 | 0.8 (0.3-2.4) | .71 |
| Presence prior arthropathy | 5.1 (1.6-15.8) | .01 | 3.13 (1.1-9.0) | .03 |
| Presence of osteoporosis | 2.3 (0.7-7.5) | .18 | 6.2 (2.0-19.7) | <.01 |
| Presence of vasculitis | 0 (0-12.4) | .37 | 0 (0-18.7) | .45 |
| Other malignancies | 0.8 (0.2-2.7) | .70 | 1.2 (0.4-3.7) | .80 |
| TPN | 3.9 (1.0-16.0) | .06 | 3.2 (0.5-18.7) | .20 |
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aIncludes leukemia and lymphoma.
bIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
cIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
Univariate assessment of putative risk factors among IBD-AVN by subtype (cases) and their respective IBD controls
| Risk Factor | CD | UC | ||
|---|---|---|---|---|
| Odds Ratio (95%CI) | P Value | Odds Ratio (95%CI) | P Value | |
| Alcohol consumption | 2.7 (0.6-12.5) | .21 | 1.5 (0.3-7.0) | .62 |
| Smoking | 1.3 (0.6-2.7) | .54 | 1.1 (0.5-2.7) | .80 |
| Obesity | 8.4 (0.9-75.4) | .06 | 2.2 (0.3-17.1) | .45 |
| Systemic steroids | 1.3 (0.4-3.6) | .64 | 10.8 (1.4-86.4) | .02 |
| Arthropathy | ∞ (6.5-∞) | <.0001 | 2.2 (0.6-8.2) | .26 |
| Extraintestinal manifestation | 2.6 (0.9-7.4) | .07 | 2.3(0.9-5.8) | .07 |
| Aphthous stomatitis | 4.3 (0.7-25.4) | .10 | ∞ (1.1-∞) | .05 |
| Erythema nodosum | ∞ (2.2-∞) | .01 | 2.3 (0.1-37.5) | .57 |
| Pyoderma gangrenosum | 0 (0-12.4) | .37 | — | — |
| PSC | 0 (0-0.6) | .02 | 3.0 (1.0-9.0) | .06 |
| Spondylitis | 0.8 (0.1-4.6) | .82 | — | — |
| Eye manifestation | ∞ (0.6-∞) | .09 | — | — |
| Hematological malignanciesa | ∞ (1.3-∞) | .13 | 2.8 (0.3-21.8) | .34 |
| Other hematological disordersb | ∞ (1.06-∞) | .04 | ∞ (4.5-∞) | <.01 |
| SLE | — | — | ∞ (0.7-∞) | .07 |
| Other rheumatological disorderc | 1.5 (0.3-7.1) | .58 | 0.6 (0.1-3.4) | .57 |
| Estrogen consumption | 4.9 (0.9-27.4) | .07 | 1.0 (0.1-6.4) | .98 |
| Estrogen consumption in female only | 7.1 (1.1-45.7) | .04 | 1.0 (0.1-6.3) | 1.00 |
| Thrombotic events | 3.3 (0.9-11.4) | .06 | 2.4 (0.6-9.4) | .22 |
| Renal transplantation | 0 (0-11.5) | .36 | 0 (0-8.7) | .31 |
| Liver transplantation | 0.6 (0.1-6.1) | .68 | ∞ (6.4-∞) | <.0001 |
| High lipid profile | 1.4 (0.5-3.7) | .47 | 0.8 (0.3-2.4) | .71 |
| Presence prior arthropathy | 5.1 (1.6-15.8) | .01 | 3.13 (1.1-9.0) | .03 |
| Presence of osteoporosis | 2.3 (0.7-7.5) | .18 | 6.2 (2.0-19.7) | <.01 |
| Presence of vasculitis | 0 (0-12.4) | .37 | 0 (0-18.7) | .45 |
| Other malignancies | 0.8 (0.2-2.7) | .70 | 1.2 (0.4-3.7) | .80 |
| TPN | 3.9 (1.0-16.0) | .06 | 3.2 (0.5-18.7) | .20 |
| Risk Factor | CD | UC | ||
|---|---|---|---|---|
| Odds Ratio (95%CI) | P Value | Odds Ratio (95%CI) | P Value | |
| Alcohol consumption | 2.7 (0.6-12.5) | .21 | 1.5 (0.3-7.0) | .62 |
| Smoking | 1.3 (0.6-2.7) | .54 | 1.1 (0.5-2.7) | .80 |
| Obesity | 8.4 (0.9-75.4) | .06 | 2.2 (0.3-17.1) | .45 |
| Systemic steroids | 1.3 (0.4-3.6) | .64 | 10.8 (1.4-86.4) | .02 |
| Arthropathy | ∞ (6.5-∞) | <.0001 | 2.2 (0.6-8.2) | .26 |
| Extraintestinal manifestation | 2.6 (0.9-7.4) | .07 | 2.3(0.9-5.8) | .07 |
| Aphthous stomatitis | 4.3 (0.7-25.4) | .10 | ∞ (1.1-∞) | .05 |
| Erythema nodosum | ∞ (2.2-∞) | .01 | 2.3 (0.1-37.5) | .57 |
| Pyoderma gangrenosum | 0 (0-12.4) | .37 | — | — |
| PSC | 0 (0-0.6) | .02 | 3.0 (1.0-9.0) | .06 |
| Spondylitis | 0.8 (0.1-4.6) | .82 | — | — |
| Eye manifestation | ∞ (0.6-∞) | .09 | — | — |
| Hematological malignanciesa | ∞ (1.3-∞) | .13 | 2.8 (0.3-21.8) | .34 |
| Other hematological disordersb | ∞ (1.06-∞) | .04 | ∞ (4.5-∞) | <.01 |
| SLE | — | — | ∞ (0.7-∞) | .07 |
| Other rheumatological disorderc | 1.5 (0.3-7.1) | .58 | 0.6 (0.1-3.4) | .57 |
| Estrogen consumption | 4.9 (0.9-27.4) | .07 | 1.0 (0.1-6.4) | .98 |
| Estrogen consumption in female only | 7.1 (1.1-45.7) | .04 | 1.0 (0.1-6.3) | 1.00 |
| Thrombotic events | 3.3 (0.9-11.4) | .06 | 2.4 (0.6-9.4) | .22 |
| Renal transplantation | 0 (0-11.5) | .36 | 0 (0-8.7) | .31 |
| Liver transplantation | 0.6 (0.1-6.1) | .68 | ∞ (6.4-∞) | <.0001 |
| High lipid profile | 1.4 (0.5-3.7) | .47 | 0.8 (0.3-2.4) | .71 |
| Presence prior arthropathy | 5.1 (1.6-15.8) | .01 | 3.13 (1.1-9.0) | .03 |
| Presence of osteoporosis | 2.3 (0.7-7.5) | .18 | 6.2 (2.0-19.7) | <.01 |
| Presence of vasculitis | 0 (0-12.4) | .37 | 0 (0-18.7) | .45 |
| Other malignancies | 0.8 (0.2-2.7) | .70 | 1.2 (0.4-3.7) | .80 |
| TPN | 3.9 (1.0-16.0) | .06 | 3.2 (0.5-18.7) | .20 |
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aIncludes leukemia and lymphoma.
bIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
cIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
There were 39 UC-AVN cases who were matched with 77 UC controls. Seventy-nine (68.7%) patients had extensive colitis, 23 (20.0%) had left-sided colitis and 13 (11.4%) had proctitis. Extent of UC was not associated with increased risk of AVN (P = .6). In addition, having moderate, severe or fulminant UC was not associated with an increased risk of AVN (P = 0.8). Two (18.2%) of 11 female patients with UC-AVN and 4 (19.0%) of 21 female patients with UC had a history of estrogen use. Use of CS, osteoporosis, and history of arthropathy were significantly associated with AVN in patients with UC, and increased the risk of AVN more than 10, 6, and 3 times, respectively. Primary sclerosing cholangitis, aphthous stomatitis, hematological diseases, and liver transplantation could be considered as significant risk factors, too (Table 2).
Regardless of IBD subtype, having severe/fulminant active disease, extraintestinal manifestations, use of TPN, and liver transplantation were all significantly associated with AVN (Table 3). In our study, smoking was not significantly associated with increased risk of AVN. Furthermore, patients with IBD who had prior arthropathy and osteoporosis had an approximately 4-fold increased risk of having AVN.
Univariate assessment of various putative risk factors for IBD-AVN regardless of IBD subtype
| Variable | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| CD extension | |||
| • Ileal vs colonic | 1 (0.3-3) | .98 | .26 |
| • Ileocolonic vs colonic | 1.6 (0.6-4.4) | .36 | |
| • Upper and lower vs colonic | 7.8 (0.8-79.8) | .08 | |
| UC extension | |||
| • Left-sided vs proctitis | 2.5 (0.4-14.9) | .31 | .60 |
| • Extensive colitis vs proctitis | 1.8 (0.4-8.1) | .43 | |
| Severity of disease | |||
| • Moderate/severe vs mild/moderate | 0.9 (0.5-1.7) | .74 | <.01 |
| • Severe/fulminant vs mild/moderate | 5.6 (2.0-16.0) | .00 | |
| Alcohol consumption | 2.0 (0.7-5.9) | .20 | |
| Smoking | 1.2 (0.7-2.0) | .61 | |
| Obesity | 4.2 (1.0-17.0) | .04 | |
| Systemic steroids | 2.5 (1.0-6.0) | .04 | |
| Arthropathy | 5.5 (1.8-17.1) | .00 | |
| Extraintestinal manifestation | 2.5 (1.2-4.9) | .01 | |
| Stomatitis | 5.6 (1.1-28.5) | .04 | |
| Erythema nodosum | 7.5 (0.8-68.2) | .07 | |
| Pyoderma gangrenosum | 0.0 | .43 | |
| PSC | 1.4 (0.6-3.4) | .47 | |
| Spondylitis | 0.8 (0.1-4.5) | .82 | |
| Eye manifestationa | Infinity | .09 | |
| Hematological malignanciesa | 3.5 (0.7-21.8) | .18 | |
| Other hematological disordersb | Infinity | .00 | |
| SLE | Infinity | .08 | |
| Other rheumatological disorderc | 1.0 (0.3-3.0) | .96 | |
| Estrogen consumption | 2.5 (0.8-8.2) | .13 | |
| Thrombotic events | 2.9 (1.2-7.0) | .02 | |
| Renal transplantation | 0.0 | .17 | |
| Liver transplantation | 5.6 (1.5-20.7) | .01 | |
| High lipid profile | 1.1 (0.5-2.2) | .82 | |
| Presence of prior arthropathy | 3.9 (1.8-8.2) | .00 | |
| Presence of osteoporosis | 4.0 (1.8-8.9) | .00 | |
| Presence of vasculitis | 0.0 | .26 | |
| Other malignancies | 0.9 (0.4-2.2) | .89 | |
| TPN | 3.7 (1.2-11.0) | .02 | |
| Variable | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| CD extension | |||
| • Ileal vs colonic | 1 (0.3-3) | .98 | .26 |
| • Ileocolonic vs colonic | 1.6 (0.6-4.4) | .36 | |
| • Upper and lower vs colonic | 7.8 (0.8-79.8) | .08 | |
| UC extension | |||
| • Left-sided vs proctitis | 2.5 (0.4-14.9) | .31 | .60 |
| • Extensive colitis vs proctitis | 1.8 (0.4-8.1) | .43 | |
| Severity of disease | |||
| • Moderate/severe vs mild/moderate | 0.9 (0.5-1.7) | .74 | <.01 |
| • Severe/fulminant vs mild/moderate | 5.6 (2.0-16.0) | .00 | |
| Alcohol consumption | 2.0 (0.7-5.9) | .20 | |
| Smoking | 1.2 (0.7-2.0) | .61 | |
| Obesity | 4.2 (1.0-17.0) | .04 | |
| Systemic steroids | 2.5 (1.0-6.0) | .04 | |
| Arthropathy | 5.5 (1.8-17.1) | .00 | |
| Extraintestinal manifestation | 2.5 (1.2-4.9) | .01 | |
| Stomatitis | 5.6 (1.1-28.5) | .04 | |
| Erythema nodosum | 7.5 (0.8-68.2) | .07 | |
| Pyoderma gangrenosum | 0.0 | .43 | |
| PSC | 1.4 (0.6-3.4) | .47 | |
| Spondylitis | 0.8 (0.1-4.5) | .82 | |
| Eye manifestationa | Infinity | .09 | |
| Hematological malignanciesa | 3.5 (0.7-21.8) | .18 | |
| Other hematological disordersb | Infinity | .00 | |
| SLE | Infinity | .08 | |
| Other rheumatological disorderc | 1.0 (0.3-3.0) | .96 | |
| Estrogen consumption | 2.5 (0.8-8.2) | .13 | |
| Thrombotic events | 2.9 (1.2-7.0) | .02 | |
| Renal transplantation | 0.0 | .17 | |
| Liver transplantation | 5.6 (1.5-20.7) | .01 | |
| High lipid profile | 1.1 (0.5-2.2) | .82 | |
| Presence of prior arthropathy | 3.9 (1.8-8.2) | .00 | |
| Presence of osteoporosis | 4.0 (1.8-8.9) | .00 | |
| Presence of vasculitis | 0.0 | .26 | |
| Other malignancies | 0.9 (0.4-2.2) | .89 | |
| TPN | 3.7 (1.2-11.0) | .02 | |
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; CI, confidence interval; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aIncludes leukemia and lymphoma.
bIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
cIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
Univariate assessment of various putative risk factors for IBD-AVN regardless of IBD subtype
| Variable | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| CD extension | |||
| • Ileal vs colonic | 1 (0.3-3) | .98 | .26 |
| • Ileocolonic vs colonic | 1.6 (0.6-4.4) | .36 | |
| • Upper and lower vs colonic | 7.8 (0.8-79.8) | .08 | |
| UC extension | |||
| • Left-sided vs proctitis | 2.5 (0.4-14.9) | .31 | .60 |
| • Extensive colitis vs proctitis | 1.8 (0.4-8.1) | .43 | |
| Severity of disease | |||
| • Moderate/severe vs mild/moderate | 0.9 (0.5-1.7) | .74 | <.01 |
| • Severe/fulminant vs mild/moderate | 5.6 (2.0-16.0) | .00 | |
| Alcohol consumption | 2.0 (0.7-5.9) | .20 | |
| Smoking | 1.2 (0.7-2.0) | .61 | |
| Obesity | 4.2 (1.0-17.0) | .04 | |
| Systemic steroids | 2.5 (1.0-6.0) | .04 | |
| Arthropathy | 5.5 (1.8-17.1) | .00 | |
| Extraintestinal manifestation | 2.5 (1.2-4.9) | .01 | |
| Stomatitis | 5.6 (1.1-28.5) | .04 | |
| Erythema nodosum | 7.5 (0.8-68.2) | .07 | |
| Pyoderma gangrenosum | 0.0 | .43 | |
| PSC | 1.4 (0.6-3.4) | .47 | |
| Spondylitis | 0.8 (0.1-4.5) | .82 | |
| Eye manifestationa | Infinity | .09 | |
| Hematological malignanciesa | 3.5 (0.7-21.8) | .18 | |
| Other hematological disordersb | Infinity | .00 | |
| SLE | Infinity | .08 | |
| Other rheumatological disorderc | 1.0 (0.3-3.0) | .96 | |
| Estrogen consumption | 2.5 (0.8-8.2) | .13 | |
| Thrombotic events | 2.9 (1.2-7.0) | .02 | |
| Renal transplantation | 0.0 | .17 | |
| Liver transplantation | 5.6 (1.5-20.7) | .01 | |
| High lipid profile | 1.1 (0.5-2.2) | .82 | |
| Presence of prior arthropathy | 3.9 (1.8-8.2) | .00 | |
| Presence of osteoporosis | 4.0 (1.8-8.9) | .00 | |
| Presence of vasculitis | 0.0 | .26 | |
| Other malignancies | 0.9 (0.4-2.2) | .89 | |
| TPN | 3.7 (1.2-11.0) | .02 | |
| Variable | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| CD extension | |||
| • Ileal vs colonic | 1 (0.3-3) | .98 | .26 |
| • Ileocolonic vs colonic | 1.6 (0.6-4.4) | .36 | |
| • Upper and lower vs colonic | 7.8 (0.8-79.8) | .08 | |
| UC extension | |||
| • Left-sided vs proctitis | 2.5 (0.4-14.9) | .31 | .60 |
| • Extensive colitis vs proctitis | 1.8 (0.4-8.1) | .43 | |
| Severity of disease | |||
| • Moderate/severe vs mild/moderate | 0.9 (0.5-1.7) | .74 | <.01 |
| • Severe/fulminant vs mild/moderate | 5.6 (2.0-16.0) | .00 | |
| Alcohol consumption | 2.0 (0.7-5.9) | .20 | |
| Smoking | 1.2 (0.7-2.0) | .61 | |
| Obesity | 4.2 (1.0-17.0) | .04 | |
| Systemic steroids | 2.5 (1.0-6.0) | .04 | |
| Arthropathy | 5.5 (1.8-17.1) | .00 | |
| Extraintestinal manifestation | 2.5 (1.2-4.9) | .01 | |
| Stomatitis | 5.6 (1.1-28.5) | .04 | |
| Erythema nodosum | 7.5 (0.8-68.2) | .07 | |
| Pyoderma gangrenosum | 0.0 | .43 | |
| PSC | 1.4 (0.6-3.4) | .47 | |
| Spondylitis | 0.8 (0.1-4.5) | .82 | |
| Eye manifestationa | Infinity | .09 | |
| Hematological malignanciesa | 3.5 (0.7-21.8) | .18 | |
| Other hematological disordersb | Infinity | .00 | |
| SLE | Infinity | .08 | |
| Other rheumatological disorderc | 1.0 (0.3-3.0) | .96 | |
| Estrogen consumption | 2.5 (0.8-8.2) | .13 | |
| Thrombotic events | 2.9 (1.2-7.0) | .02 | |
| Renal transplantation | 0.0 | .17 | |
| Liver transplantation | 5.6 (1.5-20.7) | .01 | |
| High lipid profile | 1.1 (0.5-2.2) | .82 | |
| Presence of prior arthropathy | 3.9 (1.8-8.2) | .00 | |
| Presence of osteoporosis | 4.0 (1.8-8.9) | .00 | |
| Presence of vasculitis | 0.0 | .26 | |
| Other malignancies | 0.9 (0.4-2.2) | .89 | |
| TPN | 3.7 (1.2-11.0) | .02 | |
Abbreviations: AVN, avascular necrosis; CD, Crohn’s disease; CI, confidence interval; IBD, inflammatory bowel disease; PSC, primary sclerosing cholangitis; SLE, systemic lupus erythematosus; TPN, total parenteral nutrition; UC, ulcerative colitis.
aIncludes leukemia and lymphoma.
bIncludes factor V Leiden mutation, B12 deficiency, and hereditary spherocytosis.
cIncludes rheumatoid arthritis, psoriatic arthritis, vasculitis, polymyalgia rheumatic, mixed connective tissue disease, CREST (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) syndrome, cryoglobulinemia, rheumatic fever, degenerative joint disease, and sarcoidosis.
Of the 8 IBD patients who had AVN with no CS exposure, 2 (25%) had alcoholism, 3 (37.5%) were on estrogen therapy, 5 (62.5%) had hyperlipidemia, 2 (25%) had prior arthropathy, 5 (62.5%) were smokers, and 3 (37.5%) had obesity as AVN risk factors.
In a multivariable conditional logistic model, after adjusting for other factors, the severity of IBD, prior history of arthropathy, and osteoporosis were significantly associated with AVN (Table 4).
Multivariable conditional logistic model for having osteonecrosis
| Risk factor | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| Severity of disease | .01 | ||
| • Moderate/severe vs mild/moderate | 0.8 (0.4-1.5) | .47 | |
| • Severe/fulminant vs mild/moderate | 4.3 (1.4-13.1) | .01 | |
| Presence prior arthropathy | 3.6 (1.6-8.1) | .00 | |
| Presence of osteoporosis | 3.8 (1.5-9.2) | .00 |
| Risk factor | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| Severity of disease | .01 | ||
| • Moderate/severe vs mild/moderate | 0.8 (0.4-1.5) | .47 | |
| • Severe/fulminant vs mild/moderate | 4.3 (1.4-13.1) | .01 | |
| Presence prior arthropathy | 3.6 (1.6-8.1) | .00 | |
| Presence of osteoporosis | 3.8 (1.5-9.2) | .00 |
The analysis was adjusted for the date of diagnosis of inflammatory bowel disease in the Mayo Clinic and age at inflammatory bowel disease diagnosis by including the variables that had P value of ≤.10 in univariate conditional logistic analysis and ≥5% prevalence in patients.
Abbreviation: CI, confidence interval.
Multivariable conditional logistic model for having osteonecrosis
| Risk factor | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| Severity of disease | .01 | ||
| • Moderate/severe vs mild/moderate | 0.8 (0.4-1.5) | .47 | |
| • Severe/fulminant vs mild/moderate | 4.3 (1.4-13.1) | .01 | |
| Presence prior arthropathy | 3.6 (1.6-8.1) | .00 | |
| Presence of osteoporosis | 3.8 (1.5-9.2) | .00 |
| Risk factor | Odds Ratio (95% CI) | P Value | Overall P Value |
|---|---|---|---|
| Severity of disease | .01 | ||
| • Moderate/severe vs mild/moderate | 0.8 (0.4-1.5) | .47 | |
| • Severe/fulminant vs mild/moderate | 4.3 (1.4-13.1) | .01 | |
| Presence prior arthropathy | 3.6 (1.6-8.1) | .00 | |
| Presence of osteoporosis | 3.8 (1.5-9.2) | .00 |
The analysis was adjusted for the date of diagnosis of inflammatory bowel disease in the Mayo Clinic and age at inflammatory bowel disease diagnosis by including the variables that had P value of ≤.10 in univariate conditional logistic analysis and ≥5% prevalence in patients.
Abbreviation: CI, confidence interval.
A total of 43 (50.6%) patients were treated for AVN. Half of those (n = 22) underwent arthroplasty, 21% required narcotics, 9.3% had CS injection, 7% had core decompression, 7% had vascular-bone graft, and 4.7% had osteotomy. One-fourth of patients (n = 11) had no medication except conservative treatment. In 14% (n = 6) of the patients, symptom worsened after treatments, and 10.6% had new bone involvement after initial diagnosis.
Discussion
In this study of IBD patients with AVN, we found that CS therapy was quite common in IBD patients and was a risk factor for AVN. Only 10% of our patients with AVN were not exposed to CS. Other AVN risk factors appeared to be history of arthropathy, obesity, and osteoporosis. Patients who had severe or fulminant IBD were over 5 times more likely to have AVN. Extraintestinal manifestations, particularly arthropathy, stomatitis, and erythema nodosum, were associated with AVN. Liver transplantation and use of TPN were associated with AVN.
Similar to our study, a recently published cohort study from Los Angeles showed that 93% of patients with AVN had CS exposure, but fewer than half of them were exposed to at least 20 mg/d of prednisone (or equivalent formulation) for 6 months.8 Of note, 7% of patients had no CS exposure, which calls into a question the “conventional wisdom” that CS is the only risk factor for AVN in the IBD population.8 An overall 5-fold elevation in risk associated with severe IBD in the current study supports this hypothesis. It is possible that patients with more severe disease activity have higher levels of acute phase reactants and more hypercoagulability, which is thought be a risk factor for AVN. Similarly, in a recent population-based study, IBD itself was reported as an independent risk factor for AVN, and being exposed to CS was associated with a smaller increase in AVN incidence than having an inflammatory disease such as IBD, arthritis, and SLE.3 Longer lapse of time between CS use and AVN may suggest an extraintestinal manifestation of IBD, rather than CS-related AVN.17 The reasons for differences in the risk profiles for AVN between CD and UC are unknown but may point to the multifactorial pathogenesis of AVN among IBD patients.
Determining the risk period among IBD patients who received long-term CS is critical. The patients in our study received systemic CS for a cumulative median duration of 4.3 months. Similarly, there are several studies that reported AVN in renal transplantation patients, SLE and other autoimmune disease within <13 months after first dose of CS.24-26 Close follow-up and screening for AVN is recommended, especially in the first year of CS therapy, among patients who are supposed to be on long-term CS therapy.27 Although the exact thresholds of duration of CS therapy or maximum dose resulting in AVN still need to be determined, some studies have suggested that the cumulative dose of CS is the most significant variable increasing the risk of AVN.28,29
Unlike other musculoskeletal manifestations in IBD, information about AVN in IBD is scarce.30-36 Most of the previous studies have been case reports or case series of AVN in IBD patients, or included other inflammatory diseases such as SLE and RA.37 One of the strengths of our study is the relatively large sample size of IBD-AVN, with 15-year follow-up after diagnosis.8,28,38-43 Another strength of our study is that it was retrospective, in which patients who had no symptom or had very subtle manifestations of AVN in imaging would not be missed or underdiagnosed.8,27
The major limitation of our study is that it was performed at a tertiary referral center with patients who likely had more severe disease. We tried to address the limitations and matched each case with 2 controls if possible on age, sex, and date of diagnosis.
Conclusions
We found that the majority of patients with IBD-AVN had involvement of one or both femoral heads followed by bilateral distal femoral heads and proximal humerus. Most of them received CS, but many patients had other risk factors, including severe fulminant disease activity, osteoporosis, and prior arthropathy, which should be considered before starting CS therapy. In addition to considering earlier use of immune modulators and biologics for IBD, more studies are needed to fully understand the etiopathogenesis of AVN and selecting the best treating options.
Author Contribution
Z.B. contributed to data collection and interpretation, statistical analysis, and writing the draft of the manuscript; S.Y. contributed to study conception, data collection and interpretation, statistical analysis, and writing the draft of the manuscript; W.S.H. contributed to statistical analysis and interpretation of data, and critical revision of the manuscript; J.E.V. contributed to study conception, data collection and interpretation, statistical analysis, and writing the draft of the manuscript; K.A.K. contributed to data collection and critical revision of the manuscript, W.J.T. contributed to critical revision of the manuscript; E.V.L. contributed to study conception, data collection and interpretation, statistical analysis, and critical revision of the manuscript.
Funding
This work was supported in part by the Mayo Foundation for Medical Education and Research.
Conflicts of Interest
E.V.L. has served as a consultant for AbbVie, Amgen, Arena, Boehringer Ingelheim, Bristol-Myers Squibb, CALIBR, Celltrion Healthcare, Eli Lilly, Fresenius Kabi, Genentech, Gilead, Gossamer Bio, Iterative Scopes, Janssen, Morphic Therapeutic, Ono Pharma, Pfizer, Protagonist Therapeutics, Scipher Medicine, Sun Pharma, Surrozen, Takeda, and UCB Biopharma; and received research support from AbbVie, Bristol-Myers Squibb, Celgene, Genentech, Gilead, Gossamer Bio, Janssen, Pfizer, Receptos, Takeda, Theravance, and UCB Biopharma. None of the other authors have disclosures.
