- Email: [email protected]
Steroid-related osteonecrosis in inflammatory bowel disease
GASTROENTEROLOGY
Steroid-Related Osteonecrosis Inflammatory Bowel Disease NIMISH VAKIL and MARSHALL
in
SPARBERG
University of Texas Health Sciences Center, Houston, Texas, and Northwestern School of Medicine, Chicago, Illinois
Osteonecrosis is a serious complication of steroid therapy characterized by death of all the cellular elements of bone. We describe a series of patients with steroid-induced osteonecrosis in inflammatory bowel disease. Seven of 161 patients (4.3%) treated with corticosteroids for inflammatory bowel disease over a lo-yr period developed osteonecrosis. The median age at the onset of inflammatory bowel disease was 20 yr and for osteonecrosis the median age was 28 yr. Patients had received steroids for a mean duration of 42 wk with a mean daily dose of 26 mg/day and a mean cumulative lifetime prednisone dose of 7 g. Osteonecrosis occurred within 6 mo of the last administration of steroid in all patients. It presented with joint pain in the hip or knee and was frequently mistaken for the arthralgia of steroid withdrawal or the arthropathy of inflammatory bowel disease. Multiple joints were involved in 6 of the 7 patients. Surgery for the joint disease was required in 4 of the 7 patients. The median duration of follow-up was 2 yr. Five of the 7 patients continued to have significant joint pain and disability that limited their activity. We conclude that inflammatory bowel disease predisposes to steroid-induced osteonecrosis. The age of patients is younger, and the dose and duration are considerably lower than that reported for steroid-induced osteonecrosis in other disease states. Bone scans or magnetic resonance imaging should be performed in patients with joint pain who are receiving or have recently received corticosteroid therapy. Early detection and treatment may prevent the crippling long-term complications of this disease.
0
steonecrosis [aseptic necrosis or avascular necrosis) is a disorder characterized by death of all the cellular elements of bone: fat cells, osteocytes, and hematopoietic cells. An association with corticosteroid use has long been recognized, but knowledge regarding pathogenesis is incomplete. In recent
1989;96:62-7
University
years the importance of the underlying disease in causation has been emphasized (1). Some disease states are rarely associated with osteonecrosis despite corticosteroid use (rheumatoid arthritis, cerebral edema after head injury), whereas some are more frequently associated with it (lupus and renal transplantation) (1). Steroids are commonly used in inflammatory bowel disease but osteonecrosis has rarely been reported. An adult patient (2) and 3 children (3) have been described with osteonecrosis and in all these patients, the disease was thought to be related to causes other than corticosteroids.
Case Reports Case
1
man with ulcerative colitis of 8-yr A 2%yr-old duration presented with cramping abdominal pain, bloody diarrhea, and weight loss over 6 wk. Colonoscopy showed numerous ulcerations with a mucopurulent exudate throughout the colon. Intravenous hydrocortisone was started and, after a week of therapy, the diarrhea improved. Oral prednisone was begun at 60 mgiday and tapered gradually over 5 mo. Two weeks after commencement of oral prednisone, the patient complained of soreness and pain in both ankles and the left hip. Physical examination revealed no abnormality. The pain was attributed to rapid steroid withdrawal. Nine months later, pain in the left hip had become severe and the patient was limping. Radiographs showed osteonecrosis of both femoral heads and the right talus.
Case
2
A SO-yr-old man presented with signs of intestinal obstruction. Crohn’s disease had been diagnosed I5 yr earlier. The patient had had two previous episodes of intestinal obstruction treated unsuccessfully with steroids, requiring ileal resection. Intravenous hydrocortisone was
0 1989 by the American
Gastroenterological
0016-5085/89/$3.50
Association
January
1989
OSTEONECROSIS IN IBD
started but 2 wk later there was no significant improvement. Contrast radiographs showed narrowing at the ileocolic anastomosis and a further ileal resection was carried out. Steroids were gradually tapered after surgery. Four weeks later the patient developed pain in both hip joints. The pain worsened over the following 2 wk and the patient was unable to walk without crutches. Radiographs showed osteonecrosis of both femoral heads and osteonecrosis of the right humerus.
Cuse
3
A 26-yr-old woman with a history of ulcerative colitis presented with bloody diarrhea. Colonoscopy revealed multiple ulcerations throughout the colon. Stool cultures yielded no pathogens and biopsy specimens were consistent with ulcerative colitis. Her symptoms had previously been controlled with sulfasalazine and occasional use of steroid enemas. Prednisone was begun at 60 mgiday; symptoms resolved and prednisone was tapered gradually over 23 wk. During the last 2 wk of steroid therapy, the patient complained of arthralgias and mild but persistent pain in the right knee. Radiographs showed osteonecrosis of both femoral heads and the right femoral condyle.
difficulty bilateral
Case
Case 5 A 22-yr-old man presented with bloody diarrhea and cramping abdominal pain. A contrast study of the small bowel showed ulcerations in the terminal ileum. Stool cultures yielded no pathogens. Prednisone was begun at 40 mgiday and tapered over 1 mo to 7 mg/day. The patient then developed pain in the left hip, radiating down to the left knee. An orthopedic consultant found no abnormalities on physical examination. Radiographs and tomograms of the joint were normal. The pain was attributed to rapid steroid withdrawal and the dose of prednisone was increased to 15 mg/day. One month later, the patient developed pain in the right hip joint as well and had
Bone heads.
scan
showed
6
A 27-yr-old man developed an acute exacerbation of ulcerative colitis with bloody diarrhea and cramping abdominal pain. Prednisone was begun at 60 mglday. The symptoms worsened and the patient was admitted to the hospital. Hyperalimentation and intravenous hydrocortisone were begun. Sigmoidoscopy showed a friable mucosa with numerous ulcerations. The patient improved slowly over 5 wk; steroids were tapered and discontinued over the next 7 mo. The patient became asymptomatic and colonoscopy showed inactive disease. A year later, the patient developed another exacerbation of colitis. Prednisone was begun at 60 mgiday with some relief. Attempts to taper the steroid dose failed and the patient underwent total colectomy and ileostomy. Steroids were tapered and discontinued 4 wk after surgery. Two months after uneventful recovery from surgery, the patient developed pain in the right knee. Bone scans showed an area of increased uptake in the right medial femoral condyle suggestive of osteonecrosis. Arthroscopic debridement was performed.
Case 4 A 2%yr-old woman with quiescent ulcerative colitis developed bloody diarrhea. Stool cultures yielded no pathogens and colonoscopy showed ulceration, edema, and friability throughout the colon. Biopsy specimens were consistent with an exacerbation of ulcerative colitis. Prednisone was begun at 30 mglday; the patient improved rapidly. Prednisone was gradually tapered over 6 mo. Nine months after steroids had been withdrawn, the patient developed pain in the right hip. Radiographs showed osteonecrosis of the femoral head. The patient underwent surgery and placement of an Austin Moore prosthesis. Five years after surgery, pain recurred in the right hip. Radiographs showed protrusion of the prosthesis through the acetabulum. The patient underwent total hip replacement. During the 5-yr period after initial orthopedic surgery, the patient had received steroids for 13 wk for an exacerbation of colitis.
walking without crutches. osteonecrosis of the femoral
63
Case
7
A 36-yr-old man developed a severe exacerbation of ulcerative colitis and prednisone was begun. Colonoscopy and biopsy specimens obtained during an earlier exacerbation had shown pancolitis. Prednisone was administered at 60 mgiday for 1 mo and then tapered gradually over the next 2 mo. While the patient was still receiving 7.5 mg of prednisone a clay, he developed pain and soreness in the right knee. Antiinflammatory drugs were prescribed with little relief. Two months later, an orthopedic consultant found no physical signs and normal radiographs of the knees. A bone scan showed increased uptake in both medial femoral condyles.
Results The records of 7 patients with inflammatory bowel disease and osteonecrosis seen over a lo-yr period [ 1977-19873, were retrospectively analyzed. They were part of a group of 204 consecutive patients with inflammatory bowel disease seen by one of us in a referral practice based at a university hospital. Of these, 161 patients (79%) had received corticosteroids as part of their management. In the group with osteonecrosis, the median age of presentation with inflammatory bowel disease was 20 yr and the median age at the onset of osteonecrosis was from
the
28 yr (range, diagnosis
the development 0.5-20 disease
22-36
yr).
The
mean
duration
of inflammatory bowel disease to of osteonecrosis was 10 yr (range, severity of the inflammatory bowel
yr). The was variable.
One
patient
developed
osteo-
64
VAKIL AND SPARBERG
Table
1. Prednisone
GASTROENTEROLOGY
Parameters
Cumulative lifetime dose Case No.
(gl
Mean daily dose (nlg/dUY)
1 2 3 4 5 6 7
4.7 13.5 3 2.4 1.8 11.7 12
28 26 19 17 12 34 20
Total duration [lifetime] of therapy (wk) 24 72 23 20 21 49 84
necrosis after treatment of the initial episode of bowel disease. Three patients had a difficult and protracted course requiring bowel resection and prolonged steroid use. Patients with osteonecrosis had a mean of three steroid-treated exacerbations of inflammatory bowel disease since the original diagnosis. Six of the 7 patients developed osteonecrosis during steroid therapy or within 6 mo of the last course of steroid therapy. Two patients had received intravenous steroids and hyperalimentation as part of their treatment. Neither developed hyperlipidemia during hyperalimentation. The patients had received steroid therapy (Table 1) for a mean duration of 42 wk throughout their lives (range, 20-84 wk). The mean daily dose of prednisone was 26 mgi day (range, 12-34 mg/day) and the mean cumulative lifetime dose calculated as prednisone was 7 g [range, 1.8-13.5 g]. The mean daily dose during the month of most intensive therapy was 49 mgiday. All the patients had received prednisone in doses exceeding 20 mg/day for at least 4 wk. Five of the 7 patients had received prednisone in daily doses exceeding 40 mg for more than 1 mo. None of the patients had any of the conditions known to predispose to osteonecrosis, i.e., alcoholism, hemoglobinopathy, hyperlipidemia, hyperuricemia, trauma, pancreatitis, or pregnancy.
Table
2. Clinical
Features
at Diagnosis
All patients presented with pain and stiffness in the hip or knee that persisted for several weeks. Three of the 7 patients developed a visible deformity of the hip caused by bony collapse. In 5 of the 7 patients, the physician initially attributed symptoms to the arthropathy of inflammatory bowel disease or to rapid steroid withdrawal. Bone scans were obtained in 6 of the 7 patients and suggested the diagnosis in all 6. Initial radiographs were abnormal in only 3 of 7 patients and showed extensive collapse of the femoral head in 2 and a crescent sign (fine radiolucent fracture line in subchondral bone that is suggestive of osteonecrosis) in another. Biopsy material was obtained surgically in 4 of the 7 patients and showed necrosis of marrow elements. In the remainder, the diagnosis was established by bone scan. Six of the 7 patients had involvement of multiple joints. Four patients required surgical intervention for their joint disease (Table 2) and 3 were managed conservatively. The median duration of follow-up after the development of osteonecrosis was 2 yr (range, l-10 yr). During the follow-up period, 5 patients continued to have crippling abdominal symptoms; 3 of these patients underwent bowel resections and 2 others needed intermittent prednisone therapy to control their symptoms. Two patients are in remission without steroid therapy. Five of the 7 patients still have significant arthralgia.
Discussion This study demonstrates that osteonecrosis is a significant problem in patients receiving corticosteroids for inflammatory bowel disease. The disease involves multiple joints in most cases and results in significant morbidity. The age of the patients is younger and the dose and duration of steroid therapy are considerably lower than reported in other disease states associated with osteonecrosis. These data
of Osteonecrosis IUD duration
Case No.
Age (yr)
Sex
Disease
1 2 3 4 5 6 7
29 30 26 29 22 27 36
M M F F M M M
UC Crohn’s UC UC Crohn’s UC UC
Vol. 96. No. 1
Involvement PC TI PC PC Ti PC PC
(yrl 8 15 20 6.5 0.5 10 12
Exacerbations of IBD 2 3 2 5 1 4 3
AD, arthroscopic debridement; BHR, bilateral hip replacement; CD, core decompression; Conserv. stimulation; F, femoral head; FC, femoral condyle; H, humerus: HR, hip replacement; IBD, inflammatory acetabuli; PC, pancolitis; T, talus; Tl. terminal ileum; 1JC, ulcerative colitis.
Bones
involved
Treatment
2F + 2F + 2F + 1F + 2F 1FC 2FC
2FC + 1T 1H 1FC PA
ES BHR Conserv HR Conserv AD CD
conservative; bowel disease;
ES, electrical PA, protrusio
0STEONE:CKOSIS
suggest that inflammatory bowel disease predisposes to corticosteroid-induced osteonecrosis. The mechanism of steroid-induced osteonecrosis is unclear. Numerous theories of causation have been proposed, including intramedullary lipocyte hypertrophy resulting in decreased circulation by vascular compression (4,5), fat embolism from a fatty liver (6,7), and steroid-induced suppression of osteoblastic function. Osteoblasts are necessary for the repair of small microfractures that occur during normal weight bearing (7). Kenzora and Glimcher (1) have recently suggested that an underlying multisystern illness impairs osteoblast function, thereby increasing susceptibility to osteonecrosis caused by another agent (e.g.. corticosteroids). This hypothetical mechanism has been called accumulative cell stress and may explain the variable incidence in different disease states treated with corticosteroids. However, this hypothesis does not explain why osteoblasts in certain sites (e.g., head of the femur) are particularly susceptible to osteonecrosis. Corticosteroid therapy appears to be necessary for the development of osteonecrosis in inflammatory bowel disease: osteonecrosis has not been described without it. The frequency of osteonecrosis in steroid-treated were drawn from patients was 4.3%. As the patients a referral population, the occurrence in the community may be smaller and deserves prospective study. The median age at which osteonecrosis developed in other was 28 yr. This is lower than reported disease states associated with a high incidence of corticosteroid-induced osteonecrosis. Boskey et al. (8) reported a series of 15 patients (mean age, 56 yr) and Zizic et al. (9) reported a series of 28 patients with lupus and osteonecrosis (mean age, 34.9 yr). Hip replacements have a limited durability in young patients, due to their tendency to overuse the prosthesis (10). Therefore age at the onset of osteonecrosis has important bearings on the outcome of surgical management. The duration of bowel disease was variable. In this series, patients developed osteonecrosis during the initial episode of inflammatory bowel disease or 20 yr later, suggesting that duration of disease may not be a factor in the pathogenesis. Most of the patients in this study had moderate to severe bowel disease. More than half required intestinal resections and some required corticosteroids to control the bowel disease even after the development of osteonecrosis. These findings suggest that patients with severe disease may be more susceptible; this is consistent with the accumulative cell stress theory (1). There is no clear-cut relationship between either the dose or the duration of corticosteroid therapy and the development of osteonecrosis. Whereas
IN IBU
6.5
some studies have reported associations with a high initial prednisone dose (1 l), high c:umulative prednisone dose (12), and high peak prednisone dose (9), other studies have shown no association with duration of therapy (9,13), peak dose (13), or cumulative dose (9). There are no data regarding the incidence of osteonecrosis or its relationship to corticosteroid dose in inflammatory bowel disease. Brom et al. (2) reported a case of Crohn’s disease with periostitis, arthritis, and aseptic necrosis and attributed the aseptic necrosis to Crohn’s disease. Shapiro et al. (3) reported 3 children with Crohn’s disease who developed hyperlipidemia on hyperalimentation and steroid therapy and subsequently developed osteonecrosis. Hyperlipidemia, a known cause of osteonecrosis, may have played a contributory role in the development. In a series of 110 patients with steroid-induced osteonecrosis in diverse diseases, Cruess (14) found that osteonecrosis caused by steroids manifested 6-8 mo of beginning steitself frequently within roids, whereas the osteopenic complications occurred 2-3 yr later. In our series, 6 of the 7 patients developed symptoms within 6 mo of their last steroid treatment. However, corticosteroid therapy had been instituted several years earlier in many of them. Cruess (5) also reported a group of patients developing osteonecrosis after renal transplantation. They had received a mean cumulative prednisone dose of 42 g. In a prospective study of patients developing steroid-induced osteonecrosis in lupus, Zizic et al. (9) found that the mean cumulative prednisone dose of the 28 patients developing osteonecrosis was 45 g. The patients in our series had a considerably lower mean cumulative dose (7 g). In the series of Zizic et al. (9), the duration of therapy was considerably longer (260 wk) than in ours (42 wk). However, our patients resemble theirs in that they had all received prednisone in daily doses exceeding 20 mg during the month of most intensive therapy. Our data suggest that osteonecrosis in inflammatory bowel disease is associated with considerably lower cumulative lifetime doses of corticosteroids and shorter durations of therapy than in the other disease states associated with osteonecrosis. Prospective data in patients with lupus (9) suggest that high peak prednisone doses are significantly associated with the development of osteonecrosis. Therefore, osteonecrosis may be another reason to consider steroid-sparing drugs in inflammatory bowel disease. Osteonecrosis should be contrasted with the arthropathy of inflammatory bowel disease, which is mild, migratory, nondeforming, frequently asymmetrical, and self-limited (15). Osteonecrosis caused by corticosteroids is bilateral in most cases (80%) and usually presents with persistent, localized pain or
66
GASTROENTEROLOGY
VAKIL AND SPARBERG
deformity, although these symptoms are not always severe. Patients with osteonecrosis of the femoral head often complain of groin pain that may travel to the thigh or to the knee. Pain is frequently aggravated by motion and may be present at rest. The bones most frequently involved in osteonecrosis are the femoral head, the femoral condyles, and the proximal humerus and talus. The capitellum, scaphoid, lunate, and metatarsal heads are less frequently involved (16).The bone scan is a noninvasive test that can establish the diagnosis before roentgenographic changes have appeared. In the earliest stage, the bone scan shows an area of decreased uptake by the necrotic bone cells; as reparative processes begin, an area of increased uptake is seen. In this series, bone scans were diagnostic in all 6 patients in whom they were performed, whereas initial radiographs showed abnormalities in only 3 of the patients with positive bone scans. Magnetic resonance imaging has recently been described as a sensitive method of detecting early disease (10). The treatment of osteonecrosis is unsatisfactory and numerous modalities are being evaluated. Treatment is most likely to be successful in the early stages of the disease. In the precollapse stages, protected weight bearing has been the traditional treatment. Core decompression is a recently described procedure that is both diagnostic and therapeutic; it is used in patients with early disease. One-centimeter cores of bony tissue are obtained surgically from the involved bone. Histologic examination helps establish the diagnosis and core decompression may help relieve pressure within the swollen head (10,17). Alternatively, core decompression may create a channel extending from healthy bone tissue into the necrotic area through which living cells and blood vessels grow. This procedure may be combined with a cortical tibia1 or fibular bone graft. Grafts provide a biomechanical benefit and encourage repair by poorly understood mechanisms. Electrical stimulation is an experimental technique using external electrical fields to stimulate new bone formation (10,18).In late stages, after collapse of the affected area has occurred, hemiarthroplasties have been the most popular method of treating symptomatic disease. The results have been variable and a poor outcome is most frequently due to femoral subsidence or protrusion through the acetabulum (17). Total hip replacement is required if the acetabulum is osteopenic or if degenerative arthritis is present. The outcome of treatment after collapse has occurred is much poorer than in the precollapse stages, underscoring the need for early diagnosis. Our patients have had several established and experimental treatment methods for osteonecrosis but arthralgia and disability have been a lasting problem.
Vol. 96, No. 1
In summary, osteonecrosis is a significant complication of corticosteroid therapy for inflammatory bowel disease. Patients with moderate or severe bowel disease appear to be particularly at risk and the disease may appear during the presenting episode of the bowel disease or years later. Lifetime cumulative doses of prednisone are much lower in patients with inflammatory bowel disease developing osteonecrosis than in other diseases frequently associated with this condition. Multiple joints are usually involved and significant disability results. Bone scan or magnetic resonance imaging should be performed in patients with persistent arthralgia, especially if they have received corticosteroids during 6 mo. the preceding
References 1. Kenzora JE. Glimcher multifactorial etiology
Clin North Am 1985;16:669-79. B, Bank S, Marks IN, Cobb JJ. Periostitis. aseptic necrosis, and arthritis occurring in a patient with Crohn’s disease. Gastroenterology 1971;60:1106-9. Shapiro SC, Rothstein FC. Newman AJ. Fletcher B, Halpin TC Jr, Multifocal osteonecrosis in adolescents with Crohns disease: a complication of therapy? J Pediatr Gastroenterol Nutr 1985:4:502-6. Wang G, lMoga DB, Richemer WC, Sweet DE, Reger SI, Thompson RC. Cortisone induced bone changes and its response to lipid clearing agents. Clin Orthop 1978:130:81-4. Cruess RL. Cortisone-induced avascular necrosis of the femoral head. J Bone Joint Surg [Br] 1977:59:308-17. Wang GJ, Sweet DE, Reger SI, et al. Fat cell changes as a mechanism of avascular necrosis of the femoral head in cortisone-treated rabbits. J Bone Joint Surg [Am] 1977;59:72935. Jones JP, Engleman EP, Najarian JS. Systemic fat embolism after renal homotransplantation and treatment with corticosteroids. N Engl J Med 1965;273:1453-5. Boskey AL, Raggio CL, Bullough PG, Kinnett JG Changes in the bone tissue lipids in persons with steroid and alcoholinduced osteonecrosis. Clin Orthop 1983;172:289-95. Zizic TM, Marcoux C. Hungerford DS. Danserau JV, Stevens MB. Corticosteroid therapy associated with ischemic necrosis of bone in systemic lupus erythematosus. Am 1 Med 1985:79: 596-603. Meyers MH. Osteonecrosis of the femoral head. Pathogenesis and long term treatment. Clin Orthop 1988;231:51-61. Abeles M, Urman JD. Rothfield N. Aseptic necrosis of bone in systemic lupus erythematosis: relationship to corticosteroid therapy. Arch Intern Med 1978;138:750-4. Bergstein JM, Wiens P, k’ish AJ, Vernier RL. Michael A. Avascular necrosis of bone in systemic lupus erythematosus. J Pediatr 1974;85:31-5. Dimant J, Ginzler E, Diamond H. et al. Computer analysis of factors influencing the appearance of aseptic necrosis in patients with SLE. J Rheumatol 1978;5:136-41. Cruess RL. Steroid induced osteonecrosis. J R Co11 Surg Edinb 1981:26:69-77. Mayer L, Janowitz HD. Extra-intestinal manifestations of ulcerative colitis including reference to Crohns disease. In: RN Allan, MRB Keighley. J Alexander-Williams, C Hawkins,
2. Brom
3
4.
5.
6.
7.
8.
9.
IO. 11.
12.
13.
14. 15.
MD. Accumulative cell stress: the of idiopathic osteonecrosis. Orthop
January
1989
OSTEONECROSIS
eds. Iuflammatory stone.
bowel
disease.
London:
Churchill
Living-
1983:121-30.
16. Kenzora necrosis:
JE, Glimcher the ubiquitous
MJ. Pathogenesis crescent
of idiopathic
sign. Orthop
osteo-
Clin North
Am
1985:16:681-96. 17. Kenzora rent
JE. Treatment
philosophy
16:717-25.
of idiopathic
and rationale.
Orthop
osteonecrosis:
the cur-
Clin North
Am 1985;
18. Spadaro ]A. Electrically stimulated bone and man. Clin Orthop 1977:122:325--32.
growth
IN IBD
67
in animals
-_ Received November 25. 1987. Accepted August 1, 1988. Address requests for reprints to: N.B. Vakil. M.D., Division of Gastroenterology, University of Rochester, Strong Memorial Hospital. Box MED, 601 Elmwood Avenue. Rochester. New York 14642.
Steroid-Related Osteonecrosis Inflammatory Bowel Disease NIMISH VAKIL and MARSHALL
in
SPARBERG
University of Texas Health Sciences Center, Houston, Texas, and Northwestern School of Medicine, Chicago, Illinois
Osteonecrosis is a serious complication of steroid therapy characterized by death of all the cellular elements of bone. We describe a series of patients with steroid-induced osteonecrosis in inflammatory bowel disease. Seven of 161 patients (4.3%) treated with corticosteroids for inflammatory bowel disease over a lo-yr period developed osteonecrosis. The median age at the onset of inflammatory bowel disease was 20 yr and for osteonecrosis the median age was 28 yr. Patients had received steroids for a mean duration of 42 wk with a mean daily dose of 26 mg/day and a mean cumulative lifetime prednisone dose of 7 g. Osteonecrosis occurred within 6 mo of the last administration of steroid in all patients. It presented with joint pain in the hip or knee and was frequently mistaken for the arthralgia of steroid withdrawal or the arthropathy of inflammatory bowel disease. Multiple joints were involved in 6 of the 7 patients. Surgery for the joint disease was required in 4 of the 7 patients. The median duration of follow-up was 2 yr. Five of the 7 patients continued to have significant joint pain and disability that limited their activity. We conclude that inflammatory bowel disease predisposes to steroid-induced osteonecrosis. The age of patients is younger, and the dose and duration are considerably lower than that reported for steroid-induced osteonecrosis in other disease states. Bone scans or magnetic resonance imaging should be performed in patients with joint pain who are receiving or have recently received corticosteroid therapy. Early detection and treatment may prevent the crippling long-term complications of this disease.
0
steonecrosis [aseptic necrosis or avascular necrosis) is a disorder characterized by death of all the cellular elements of bone: fat cells, osteocytes, and hematopoietic cells. An association with corticosteroid use has long been recognized, but knowledge regarding pathogenesis is incomplete. In recent
1989;96:62-7
University
years the importance of the underlying disease in causation has been emphasized (1). Some disease states are rarely associated with osteonecrosis despite corticosteroid use (rheumatoid arthritis, cerebral edema after head injury), whereas some are more frequently associated with it (lupus and renal transplantation) (1). Steroids are commonly used in inflammatory bowel disease but osteonecrosis has rarely been reported. An adult patient (2) and 3 children (3) have been described with osteonecrosis and in all these patients, the disease was thought to be related to causes other than corticosteroids.
Case Reports Case
1
man with ulcerative colitis of 8-yr A 2%yr-old duration presented with cramping abdominal pain, bloody diarrhea, and weight loss over 6 wk. Colonoscopy showed numerous ulcerations with a mucopurulent exudate throughout the colon. Intravenous hydrocortisone was started and, after a week of therapy, the diarrhea improved. Oral prednisone was begun at 60 mgiday and tapered gradually over 5 mo. Two weeks after commencement of oral prednisone, the patient complained of soreness and pain in both ankles and the left hip. Physical examination revealed no abnormality. The pain was attributed to rapid steroid withdrawal. Nine months later, pain in the left hip had become severe and the patient was limping. Radiographs showed osteonecrosis of both femoral heads and the right talus.
Case
2
A SO-yr-old man presented with signs of intestinal obstruction. Crohn’s disease had been diagnosed I5 yr earlier. The patient had had two previous episodes of intestinal obstruction treated unsuccessfully with steroids, requiring ileal resection. Intravenous hydrocortisone was
0 1989 by the American
Gastroenterological
0016-5085/89/$3.50
Association
January
1989
OSTEONECROSIS IN IBD
started but 2 wk later there was no significant improvement. Contrast radiographs showed narrowing at the ileocolic anastomosis and a further ileal resection was carried out. Steroids were gradually tapered after surgery. Four weeks later the patient developed pain in both hip joints. The pain worsened over the following 2 wk and the patient was unable to walk without crutches. Radiographs showed osteonecrosis of both femoral heads and osteonecrosis of the right humerus.
Cuse
3
A 26-yr-old woman with a history of ulcerative colitis presented with bloody diarrhea. Colonoscopy revealed multiple ulcerations throughout the colon. Stool cultures yielded no pathogens and biopsy specimens were consistent with ulcerative colitis. Her symptoms had previously been controlled with sulfasalazine and occasional use of steroid enemas. Prednisone was begun at 60 mgiday; symptoms resolved and prednisone was tapered gradually over 23 wk. During the last 2 wk of steroid therapy, the patient complained of arthralgias and mild but persistent pain in the right knee. Radiographs showed osteonecrosis of both femoral heads and the right femoral condyle.
difficulty bilateral
Case
Case 5 A 22-yr-old man presented with bloody diarrhea and cramping abdominal pain. A contrast study of the small bowel showed ulcerations in the terminal ileum. Stool cultures yielded no pathogens. Prednisone was begun at 40 mgiday and tapered over 1 mo to 7 mg/day. The patient then developed pain in the left hip, radiating down to the left knee. An orthopedic consultant found no abnormalities on physical examination. Radiographs and tomograms of the joint were normal. The pain was attributed to rapid steroid withdrawal and the dose of prednisone was increased to 15 mg/day. One month later, the patient developed pain in the right hip joint as well and had
Bone heads.
scan
showed
6
A 27-yr-old man developed an acute exacerbation of ulcerative colitis with bloody diarrhea and cramping abdominal pain. Prednisone was begun at 60 mglday. The symptoms worsened and the patient was admitted to the hospital. Hyperalimentation and intravenous hydrocortisone were begun. Sigmoidoscopy showed a friable mucosa with numerous ulcerations. The patient improved slowly over 5 wk; steroids were tapered and discontinued over the next 7 mo. The patient became asymptomatic and colonoscopy showed inactive disease. A year later, the patient developed another exacerbation of colitis. Prednisone was begun at 60 mgiday with some relief. Attempts to taper the steroid dose failed and the patient underwent total colectomy and ileostomy. Steroids were tapered and discontinued 4 wk after surgery. Two months after uneventful recovery from surgery, the patient developed pain in the right knee. Bone scans showed an area of increased uptake in the right medial femoral condyle suggestive of osteonecrosis. Arthroscopic debridement was performed.
Case 4 A 2%yr-old woman with quiescent ulcerative colitis developed bloody diarrhea. Stool cultures yielded no pathogens and colonoscopy showed ulceration, edema, and friability throughout the colon. Biopsy specimens were consistent with an exacerbation of ulcerative colitis. Prednisone was begun at 30 mglday; the patient improved rapidly. Prednisone was gradually tapered over 6 mo. Nine months after steroids had been withdrawn, the patient developed pain in the right hip. Radiographs showed osteonecrosis of the femoral head. The patient underwent surgery and placement of an Austin Moore prosthesis. Five years after surgery, pain recurred in the right hip. Radiographs showed protrusion of the prosthesis through the acetabulum. The patient underwent total hip replacement. During the 5-yr period after initial orthopedic surgery, the patient had received steroids for 13 wk for an exacerbation of colitis.
walking without crutches. osteonecrosis of the femoral
63
Case
7
A 36-yr-old man developed a severe exacerbation of ulcerative colitis and prednisone was begun. Colonoscopy and biopsy specimens obtained during an earlier exacerbation had shown pancolitis. Prednisone was administered at 60 mgiday for 1 mo and then tapered gradually over the next 2 mo. While the patient was still receiving 7.5 mg of prednisone a clay, he developed pain and soreness in the right knee. Antiinflammatory drugs were prescribed with little relief. Two months later, an orthopedic consultant found no physical signs and normal radiographs of the knees. A bone scan showed increased uptake in both medial femoral condyles.
Results The records of 7 patients with inflammatory bowel disease and osteonecrosis seen over a lo-yr period [ 1977-19873, were retrospectively analyzed. They were part of a group of 204 consecutive patients with inflammatory bowel disease seen by one of us in a referral practice based at a university hospital. Of these, 161 patients (79%) had received corticosteroids as part of their management. In the group with osteonecrosis, the median age of presentation with inflammatory bowel disease was 20 yr and the median age at the onset of osteonecrosis was from
the
28 yr (range, diagnosis
the development 0.5-20 disease
22-36
yr).
The
mean
duration
of inflammatory bowel disease to of osteonecrosis was 10 yr (range, severity of the inflammatory bowel
yr). The was variable.
One
patient
developed
osteo-
64
VAKIL AND SPARBERG
Table
1. Prednisone
GASTROENTEROLOGY
Parameters
Cumulative lifetime dose Case No.
(gl
Mean daily dose (nlg/dUY)
1 2 3 4 5 6 7
4.7 13.5 3 2.4 1.8 11.7 12
28 26 19 17 12 34 20
Total duration [lifetime] of therapy (wk) 24 72 23 20 21 49 84
necrosis after treatment of the initial episode of bowel disease. Three patients had a difficult and protracted course requiring bowel resection and prolonged steroid use. Patients with osteonecrosis had a mean of three steroid-treated exacerbations of inflammatory bowel disease since the original diagnosis. Six of the 7 patients developed osteonecrosis during steroid therapy or within 6 mo of the last course of steroid therapy. Two patients had received intravenous steroids and hyperalimentation as part of their treatment. Neither developed hyperlipidemia during hyperalimentation. The patients had received steroid therapy (Table 1) for a mean duration of 42 wk throughout their lives (range, 20-84 wk). The mean daily dose of prednisone was 26 mgi day (range, 12-34 mg/day) and the mean cumulative lifetime dose calculated as prednisone was 7 g [range, 1.8-13.5 g]. The mean daily dose during the month of most intensive therapy was 49 mgiday. All the patients had received prednisone in doses exceeding 20 mg/day for at least 4 wk. Five of the 7 patients had received prednisone in daily doses exceeding 40 mg for more than 1 mo. None of the patients had any of the conditions known to predispose to osteonecrosis, i.e., alcoholism, hemoglobinopathy, hyperlipidemia, hyperuricemia, trauma, pancreatitis, or pregnancy.
Table
2. Clinical
Features
at Diagnosis
All patients presented with pain and stiffness in the hip or knee that persisted for several weeks. Three of the 7 patients developed a visible deformity of the hip caused by bony collapse. In 5 of the 7 patients, the physician initially attributed symptoms to the arthropathy of inflammatory bowel disease or to rapid steroid withdrawal. Bone scans were obtained in 6 of the 7 patients and suggested the diagnosis in all 6. Initial radiographs were abnormal in only 3 of 7 patients and showed extensive collapse of the femoral head in 2 and a crescent sign (fine radiolucent fracture line in subchondral bone that is suggestive of osteonecrosis) in another. Biopsy material was obtained surgically in 4 of the 7 patients and showed necrosis of marrow elements. In the remainder, the diagnosis was established by bone scan. Six of the 7 patients had involvement of multiple joints. Four patients required surgical intervention for their joint disease (Table 2) and 3 were managed conservatively. The median duration of follow-up after the development of osteonecrosis was 2 yr (range, l-10 yr). During the follow-up period, 5 patients continued to have crippling abdominal symptoms; 3 of these patients underwent bowel resections and 2 others needed intermittent prednisone therapy to control their symptoms. Two patients are in remission without steroid therapy. Five of the 7 patients still have significant arthralgia.
Discussion This study demonstrates that osteonecrosis is a significant problem in patients receiving corticosteroids for inflammatory bowel disease. The disease involves multiple joints in most cases and results in significant morbidity. The age of the patients is younger and the dose and duration of steroid therapy are considerably lower than reported in other disease states associated with osteonecrosis. These data
of Osteonecrosis IUD duration
Case No.
Age (yr)
Sex
Disease
1 2 3 4 5 6 7
29 30 26 29 22 27 36
M M F F M M M
UC Crohn’s UC UC Crohn’s UC UC
Vol. 96. No. 1
Involvement PC TI PC PC Ti PC PC
(yrl 8 15 20 6.5 0.5 10 12
Exacerbations of IBD 2 3 2 5 1 4 3
AD, arthroscopic debridement; BHR, bilateral hip replacement; CD, core decompression; Conserv. stimulation; F, femoral head; FC, femoral condyle; H, humerus: HR, hip replacement; IBD, inflammatory acetabuli; PC, pancolitis; T, talus; Tl. terminal ileum; 1JC, ulcerative colitis.
Bones
involved
Treatment
2F + 2F + 2F + 1F + 2F 1FC 2FC
2FC + 1T 1H 1FC PA
ES BHR Conserv HR Conserv AD CD
conservative; bowel disease;
ES, electrical PA, protrusio
0STEONE:CKOSIS
suggest that inflammatory bowel disease predisposes to corticosteroid-induced osteonecrosis. The mechanism of steroid-induced osteonecrosis is unclear. Numerous theories of causation have been proposed, including intramedullary lipocyte hypertrophy resulting in decreased circulation by vascular compression (4,5), fat embolism from a fatty liver (6,7), and steroid-induced suppression of osteoblastic function. Osteoblasts are necessary for the repair of small microfractures that occur during normal weight bearing (7). Kenzora and Glimcher (1) have recently suggested that an underlying multisystern illness impairs osteoblast function, thereby increasing susceptibility to osteonecrosis caused by another agent (e.g.. corticosteroids). This hypothetical mechanism has been called accumulative cell stress and may explain the variable incidence in different disease states treated with corticosteroids. However, this hypothesis does not explain why osteoblasts in certain sites (e.g., head of the femur) are particularly susceptible to osteonecrosis. Corticosteroid therapy appears to be necessary for the development of osteonecrosis in inflammatory bowel disease: osteonecrosis has not been described without it. The frequency of osteonecrosis in steroid-treated were drawn from patients was 4.3%. As the patients a referral population, the occurrence in the community may be smaller and deserves prospective study. The median age at which osteonecrosis developed in other was 28 yr. This is lower than reported disease states associated with a high incidence of corticosteroid-induced osteonecrosis. Boskey et al. (8) reported a series of 15 patients (mean age, 56 yr) and Zizic et al. (9) reported a series of 28 patients with lupus and osteonecrosis (mean age, 34.9 yr). Hip replacements have a limited durability in young patients, due to their tendency to overuse the prosthesis (10). Therefore age at the onset of osteonecrosis has important bearings on the outcome of surgical management. The duration of bowel disease was variable. In this series, patients developed osteonecrosis during the initial episode of inflammatory bowel disease or 20 yr later, suggesting that duration of disease may not be a factor in the pathogenesis. Most of the patients in this study had moderate to severe bowel disease. More than half required intestinal resections and some required corticosteroids to control the bowel disease even after the development of osteonecrosis. These findings suggest that patients with severe disease may be more susceptible; this is consistent with the accumulative cell stress theory (1). There is no clear-cut relationship between either the dose or the duration of corticosteroid therapy and the development of osteonecrosis. Whereas
IN IBU
6.5
some studies have reported associations with a high initial prednisone dose (1 l), high c:umulative prednisone dose (12), and high peak prednisone dose (9), other studies have shown no association with duration of therapy (9,13), peak dose (13), or cumulative dose (9). There are no data regarding the incidence of osteonecrosis or its relationship to corticosteroid dose in inflammatory bowel disease. Brom et al. (2) reported a case of Crohn’s disease with periostitis, arthritis, and aseptic necrosis and attributed the aseptic necrosis to Crohn’s disease. Shapiro et al. (3) reported 3 children with Crohn’s disease who developed hyperlipidemia on hyperalimentation and steroid therapy and subsequently developed osteonecrosis. Hyperlipidemia, a known cause of osteonecrosis, may have played a contributory role in the development. In a series of 110 patients with steroid-induced osteonecrosis in diverse diseases, Cruess (14) found that osteonecrosis caused by steroids manifested 6-8 mo of beginning steitself frequently within roids, whereas the osteopenic complications occurred 2-3 yr later. In our series, 6 of the 7 patients developed symptoms within 6 mo of their last steroid treatment. However, corticosteroid therapy had been instituted several years earlier in many of them. Cruess (5) also reported a group of patients developing osteonecrosis after renal transplantation. They had received a mean cumulative prednisone dose of 42 g. In a prospective study of patients developing steroid-induced osteonecrosis in lupus, Zizic et al. (9) found that the mean cumulative prednisone dose of the 28 patients developing osteonecrosis was 45 g. The patients in our series had a considerably lower mean cumulative dose (7 g). In the series of Zizic et al. (9), the duration of therapy was considerably longer (260 wk) than in ours (42 wk). However, our patients resemble theirs in that they had all received prednisone in daily doses exceeding 20 mg during the month of most intensive therapy. Our data suggest that osteonecrosis in inflammatory bowel disease is associated with considerably lower cumulative lifetime doses of corticosteroids and shorter durations of therapy than in the other disease states associated with osteonecrosis. Prospective data in patients with lupus (9) suggest that high peak prednisone doses are significantly associated with the development of osteonecrosis. Therefore, osteonecrosis may be another reason to consider steroid-sparing drugs in inflammatory bowel disease. Osteonecrosis should be contrasted with the arthropathy of inflammatory bowel disease, which is mild, migratory, nondeforming, frequently asymmetrical, and self-limited (15). Osteonecrosis caused by corticosteroids is bilateral in most cases (80%) and usually presents with persistent, localized pain or
66
GASTROENTEROLOGY
VAKIL AND SPARBERG
deformity, although these symptoms are not always severe. Patients with osteonecrosis of the femoral head often complain of groin pain that may travel to the thigh or to the knee. Pain is frequently aggravated by motion and may be present at rest. The bones most frequently involved in osteonecrosis are the femoral head, the femoral condyles, and the proximal humerus and talus. The capitellum, scaphoid, lunate, and metatarsal heads are less frequently involved (16).The bone scan is a noninvasive test that can establish the diagnosis before roentgenographic changes have appeared. In the earliest stage, the bone scan shows an area of decreased uptake by the necrotic bone cells; as reparative processes begin, an area of increased uptake is seen. In this series, bone scans were diagnostic in all 6 patients in whom they were performed, whereas initial radiographs showed abnormalities in only 3 of the patients with positive bone scans. Magnetic resonance imaging has recently been described as a sensitive method of detecting early disease (10). The treatment of osteonecrosis is unsatisfactory and numerous modalities are being evaluated. Treatment is most likely to be successful in the early stages of the disease. In the precollapse stages, protected weight bearing has been the traditional treatment. Core decompression is a recently described procedure that is both diagnostic and therapeutic; it is used in patients with early disease. One-centimeter cores of bony tissue are obtained surgically from the involved bone. Histologic examination helps establish the diagnosis and core decompression may help relieve pressure within the swollen head (10,17). Alternatively, core decompression may create a channel extending from healthy bone tissue into the necrotic area through which living cells and blood vessels grow. This procedure may be combined with a cortical tibia1 or fibular bone graft. Grafts provide a biomechanical benefit and encourage repair by poorly understood mechanisms. Electrical stimulation is an experimental technique using external electrical fields to stimulate new bone formation (10,18).In late stages, after collapse of the affected area has occurred, hemiarthroplasties have been the most popular method of treating symptomatic disease. The results have been variable and a poor outcome is most frequently due to femoral subsidence or protrusion through the acetabulum (17). Total hip replacement is required if the acetabulum is osteopenic or if degenerative arthritis is present. The outcome of treatment after collapse has occurred is much poorer than in the precollapse stages, underscoring the need for early diagnosis. Our patients have had several established and experimental treatment methods for osteonecrosis but arthralgia and disability have been a lasting problem.
Vol. 96, No. 1
In summary, osteonecrosis is a significant complication of corticosteroid therapy for inflammatory bowel disease. Patients with moderate or severe bowel disease appear to be particularly at risk and the disease may appear during the presenting episode of the bowel disease or years later. Lifetime cumulative doses of prednisone are much lower in patients with inflammatory bowel disease developing osteonecrosis than in other diseases frequently associated with this condition. Multiple joints are usually involved and significant disability results. Bone scan or magnetic resonance imaging should be performed in patients with persistent arthralgia, especially if they have received corticosteroids during 6 mo. the preceding
References 1. Kenzora JE. Glimcher multifactorial etiology
Clin North Am 1985;16:669-79. B, Bank S, Marks IN, Cobb JJ. Periostitis. aseptic necrosis, and arthritis occurring in a patient with Crohn’s disease. Gastroenterology 1971;60:1106-9. Shapiro SC, Rothstein FC. Newman AJ. Fletcher B, Halpin TC Jr, Multifocal osteonecrosis in adolescents with Crohns disease: a complication of therapy? J Pediatr Gastroenterol Nutr 1985:4:502-6. Wang G, lMoga DB, Richemer WC, Sweet DE, Reger SI, Thompson RC. Cortisone induced bone changes and its response to lipid clearing agents. Clin Orthop 1978:130:81-4. Cruess RL. Cortisone-induced avascular necrosis of the femoral head. J Bone Joint Surg [Br] 1977:59:308-17. Wang GJ, Sweet DE, Reger SI, et al. Fat cell changes as a mechanism of avascular necrosis of the femoral head in cortisone-treated rabbits. J Bone Joint Surg [Am] 1977;59:72935. Jones JP, Engleman EP, Najarian JS. Systemic fat embolism after renal homotransplantation and treatment with corticosteroids. N Engl J Med 1965;273:1453-5. Boskey AL, Raggio CL, Bullough PG, Kinnett JG Changes in the bone tissue lipids in persons with steroid and alcoholinduced osteonecrosis. Clin Orthop 1983;172:289-95. Zizic TM, Marcoux C. Hungerford DS. Danserau JV, Stevens MB. Corticosteroid therapy associated with ischemic necrosis of bone in systemic lupus erythematosus. Am 1 Med 1985:79: 596-603. Meyers MH. Osteonecrosis of the femoral head. Pathogenesis and long term treatment. Clin Orthop 1988;231:51-61. Abeles M, Urman JD. Rothfield N. Aseptic necrosis of bone in systemic lupus erythematosis: relationship to corticosteroid therapy. Arch Intern Med 1978;138:750-4. Bergstein JM, Wiens P, k’ish AJ, Vernier RL. Michael A. Avascular necrosis of bone in systemic lupus erythematosus. J Pediatr 1974;85:31-5. Dimant J, Ginzler E, Diamond H. et al. Computer analysis of factors influencing the appearance of aseptic necrosis in patients with SLE. J Rheumatol 1978;5:136-41. Cruess RL. Steroid induced osteonecrosis. J R Co11 Surg Edinb 1981:26:69-77. Mayer L, Janowitz HD. Extra-intestinal manifestations of ulcerative colitis including reference to Crohns disease. In: RN Allan, MRB Keighley. J Alexander-Williams, C Hawkins,
2. Brom
3
4.
5.
6.
7.
8.
9.
IO. 11.
12.
13.
14. 15.
MD. Accumulative cell stress: the of idiopathic osteonecrosis. Orthop
January
1989
OSTEONECROSIS
eds. Iuflammatory stone.
bowel
disease.
London:
Churchill
Living-
1983:121-30.
16. Kenzora necrosis:
JE, Glimcher the ubiquitous
MJ. Pathogenesis crescent
of idiopathic
sign. Orthop
osteo-
Clin North
Am
1985:16:681-96. 17. Kenzora rent
JE. Treatment
philosophy
16:717-25.
of idiopathic
and rationale.
Orthop
osteonecrosis:
the cur-
Clin North
Am 1985;
18. Spadaro ]A. Electrically stimulated bone and man. Clin Orthop 1977:122:325--32.
growth
IN IBD
67
in animals
-_ Received November 25. 1987. Accepted August 1, 1988. Address requests for reprints to: N.B. Vakil. M.D., Division of Gastroenterology, University of Rochester, Strong Memorial Hospital. Box MED, 601 Elmwood Avenue. Rochester. New York 14642.