Journal of Pediatric Gastroenterology and Nutrition
`33:S27-S35 © September 200 l Lippincott Williams & Wilkins, Inc., Philadelphia
`
`Approach to Steroid-Dependent and Steroid-Refractory
`Crohn' s Disease
`
`Gary R. Lichtenstein
`
`Department of Medicine, Center of Inflammatory Bowel Diseases, Hospital of the University of Pennsylvania, University of
`Pennsylvania School of Medicine, Philadelphia, Pennsylvania
`
`Crohn's disease (CD) is a chronic inflammatory bowel
`disease (IBD) that requires treatment not only to control
`symptoms of the active form of the disease but also to
`maintain disease remission after it is achieved. Cortico(cid:173)
`steroids have traditionally been the treatment of choice
`for the induction of clinical remission in adult patients
`with moderately to severely active CD, largely based on
`consistent evidence of high response rates in controlled
`clinical trials (1-4). Although these agents have not been
`studied extensively in the pediatric population, they have
`proved to be very effective in inducing remission in chil(cid:173)
`dren and adolescents in clinical practice, explaining their
`widespread use.
`However, despite their well-recognized efficacy in ac(cid:173)
`tive CD, corticosteroids have several limitations that can
`be obstacles to their use, particularly in pediatric pa(cid:173)
`tients. Among patients receiving corticosteroid therapy
`for induction of remission, 20% have corticosteroid(cid:173)
`refractory disease and 36% of those with an initial re(cid:173)
`sponse develop corticosteroid dependence within 1 year
`(4). Chronic corticosteroid exposure in patients who are
`steroid-dependent increases the risk for serious drug(cid:173)
`related adverse effects and provides little or no protec(cid:173)
`tion against relapse (1,2,5,6). Side effects such as moon
`face and acne are common in all individuals treated with
`steroids, regardless of their age, but can have an espe(cid:173)
`cially profound impact on the quality of life of children
`and adolescents. Prolonged use of corticosteroids can
`also have more far-reaching effects, such as exacerbation
`of growth failure, osteoporosis, and ocular abnormalities.
`In light of such limitations, withdrawal or reduction of
`corticosteroids has become an important goal of CD
`treatment. Several therapies, including budesonide, im(cid:173)
`munomodulators, biologic agents, and enteral feeding,
`have been evaluated for their corticosteroid-sparing ef(cid:173)
`fects. This review article briefly outlines the strengths
`and limitations of corticosteroid therapy and presents
`available data on the steroid-sparing benefits of alterna(cid:173)
`tive therapies.
`
`Address correspondence and reprint requests to Dr. Gary R.
`Lichtenstein, Associate Professor of Medicine, Department of Medi(cid:173)
`cine, Hospital of the University of Pennsylvania, University of Penn(cid:173)
`sylvania School of Medicine, 3 Ravdin Building, 3400 Spruce Street,
`Philadelphia, PA 19104-4283.
`
`CLINICAL RATIONALE FOR STEROID
`SPARING IN CROHN'S DISEASE
`As mentioned, conventional corticosteroids are very
`effective in relieving the acute symptoms of CD but are
`also associated with high rates of dependence, resistance,
`and toxicity. Although these agents have not been well
`studied in pediatric patients, the problems identified in
`clinical investigations in adult populations may be of
`even greater concern in younger patients.
`
`Clinical Efficacy
`In controlled clinical trials in adults with active CD,
`investigators have reported a clinical response in 61-
`92% of patients treated with corticosteroids (1-4). How(cid:173)
`ever, the benefits of steroids are frequently short-lived.
`In a prospective study of 109 patients with CD who
`received an initial course of corticosteroid therapy, 20%
`of patients were refractory and 80% responsive to corti(cid:173)
`costeroid therapy at 30 days. By 1 year, 44% of those
`with an initial response remained responsive, 36% be(cid:173)
`came steroid-dependent, and 20% had refractory disease
`(Fig. 1) (4).
`The inability of corticosteroids to maintain long-term
`remission in CD patients has been established in several
`clinical trials (1,2,5,6). It was first documented in a
`placebo-controlled trial of 59 patients by Smith et al. (5),
`in which clinical relapse, recurrence, or extension of CD
`occurred in similar percentages of patients receiving
`low-dose prednisone (7.5 mg/day) and those receiving
`placebo (45% and 42%, respectively) over a 3-year pe(cid:173)
`riod. Similarly, no benefit was demonstrated with corti(cid:173)
`costeroid therapy versus placebo in preventing flare-up
`or recurrence of CD over a 2-year period in either the
`National Cooperative Crohn's Disease Study (NCCDS)
`(1) or the European Cooperative Crohn's Disease Study
`(ECCDS) (2). A recent meta-analysis of pooled data
`from three double-blind, placebo-controlled trials re(cid:173)
`vealed that oral corticosteroids fail to reduce relapse of
`CD at 6, 12, or 24 months (Fig. 2) (6).
`
`Toxicity
`
`Corticosteroid therapy is associated with a substantial
`risk for a number side effects that can affect virtually
`
`S27
`
`

`

`S28
`
`Immediate
`Outcomes
`(n=109)
`
`Prolonged
`Outcomes
`(n=87)
`
`Summary
`Outcomes
`(n=109)
`
`G. R. LICHTENSTEIN
`
`Remission
`48%
`
`Improved
`32%
`
`No
`Response
`20%
`
`Remission
`54%
`
`Relapse
`46%
`
`Improved
`57%
`
`Relapse
`43%
`
`Steroid Dependent
`36%
`(n=39)
`
`Prolonged Response
`44%
`(n=48)
`
`Steroid Resistant
`20%
`(n=22)
`
`FIG. 1. Conventional corticoste(cid:173)
`roid therapy in patients with CD.
`Outcome in 109 patients at 1 year
`after acute treatment (4).
`
`every organ system (Table I). Common acute effects
`include cosmetic changes, such as moon face and acne,
`which do not present a threat to patients' physical
`health but can reduce self-esteem and social function(cid:173)
`ing, particularly in young patients. An analysis of ad(cid:173)
`verse events that occurred in the NCCDS revealed
`that the incidence of moon face and acne rose steadily
`throughout the 4-month study phase (Table 2) (7). These
`effects have occurred with a high frequency (3I%
`and 20%, respectively) even in patients being tapered
`on a regimen of prednisolone (40 mg/day to 5 mg/day)
`(8).
`Of greater concern is the risk for morbidity that ac(cid:173)
`companies long-term corticosteroid use, including its ef(cid:173)
`fects on bone metabolism that frequently result in bone
`loss and skeletal fractures (9). This consequence of cor(cid:173)
`ticosteroid therapy is particularly disturbing in children
`because bone formation and turnover are fundamentally
`important to childhood growth and development. Corti(cid:173)
`costeroids produce bone loss by altering calcium me(cid:173)
`tabolism, suppressing production of gonadal hormones,
`and inhibiting bone formation. Prolonged steroid use in
`children can retard both epiphyseal maturation and long
`bone growth, although skeletal maturation is impeded
`more than linear growth (10). Dose-related fractures are
`experienced by approximately 30-50% of steroid-treated
`CD patients (II), with fractures occurring frequently in
`patients with a daily corticosteroid dose higher than I5
`mg/day and a total lifetime steroid dose higher than 30 g
`(12).
`Other effects of chronic corticosteroid use include the
`development of hyperglycemia and hypertension
`(13, I4 ). Adrenocortical suppression induced by cortico(cid:173)
`steroid therapy results in disturbances of sex hormones,
`specifically decreased levels of circulating estrogen (15)
`and low levels of testosterone (16,I7). Ocular abnormali(cid:173)
`ties related to glaucoma were observed in a group of
`prednisone-treated pediatric patients with IBD (18), and
`the prevalence of cataracts is increased by 50% in adults
`receiving systemic corticosteroid therapy ( 19).
`
`J Pediatr Gastroentero/ Nutr, Vol. 33, Suppl. I, 2001
`
`CORTICOSTEROID-SPARING TREATMENTS
`IN CROHN'S DISEASE
`
`Because of the risks and limitations of corticosteroids,
`their replacement with alternative agents should be con(cid:173)
`sidered in most patients. Several currently available and
`investigational agents have been evaluated for their
`corticosteroid-sparing effects in patients with
`corticosteroid-dependent CD. These include the second(cid:173)
`generation corticosteroid budesonide, the immunomodu(cid:173)
`lators azathioprine/6-mercaptopurine, methotrexate, and
`cyclosporine, the anti-tumor necrosis factor-ex (anti(cid:173)
`TNF-cx) therapies infliximab, CDP-57I, thalidomide, and
`etanercept, and enteral nutrition.
`
`Budesonide
`
`Although not currently available in the United States
`for use in the management of IBD, controlled ileal(cid:173)
`release budesonide has been studied in adults for its
`steroid-sparing effects in CD. In a recent multicenter,
`double-blind study by Cortot et al. (20), 120 steroid(cid:173)
`dependent patients [Crohn's Disease Activity Index
`(CDAI) <200] were randomized to treatment with budes(cid:173)
`onide 6 mg/day or placebo once daily for 22 weeks.
`Daily doses of prednisolone were tapered from a range of
`10-30 mg to 0 mg during the first 4 to 10 weeks. Relapse
`was defined as a CDAI score >200, an increase from
`baseline in CDAI score of 2::60 points, or study discon(cid:173)
`tinuation because of disease exacerbation. Compared
`with placebo, budesonide was associated with signifi(cid:173)
`cantly higher remission rates at I week (83% versus
`59%, respectively; P = 0.004) and at 13 weeks (68%
`versus 35%, respectively; P < 0.00 I) after discontinua(cid:173)
`tion of prednisolone. The frequency of adverse events
`was similar between treatment groups. The results of this
`study suggest that budesonide may be useful as a short(cid:173)
`term replacement for conventional steroids in steroid(cid:173)
`dependent patients with CD.
`
`

`

`APPROACH TO STEROID-DEPENDENT AND STEROID-REFRACTORY CROHN'S DISEASE
`
`S29
`
`Expt
`n/N
`
`Ctrl
`n!N
`
`Peto OR
`(95% Cl Fixed)
`
`Weight
`%
`
`Peto OR
`(95% Cl Fixed)
`
`Study
`
`6 Months
`ECCDS, 1984
`
`14166
`
`19152
`
`NCCDS, 1979
`
`8143
`
`14183
`
`Smith, 1978
`
`1133
`
`0126------~----~·~
`
`Subtotal (95%CI) 23/142 331161
`Chi-square 3.00 (df=2) Z=1.10
`
`18.2
`
`12.7
`
`0.8
`
`31.7
`
`0.47 [0.21, 1.05]
`
`1.13 [0.43, 2.96]
`
`5.98 [0.12, 309.75]
`
`0.71 [0.38, 1.31]
`
`- 1
`
`.1 .2
`Ctrl
`n!N
`
`Expt
`n/N
`
`
`Peto OR
`(95% C/ Fixed)
`
`5 10
`Weight Peto OR
`(95% Cl Fixed)
`%
`
`A
`
`Study
`
`12 Months
`ECCDS, 1984
`
`25 I 66
`
`26 I 52
`
`NCCDS, 1979
`
`9133
`
`17160
`
`Smith, 1978
`
`3132
`
`0/26 ------~-----•
`Subtotal (95%CI) 371131 431138
`Chi-square 3.79 (df=2) Z=0.69
`
`--
`
`22.2
`
`13.4
`
`2.2
`
`37.8
`
`0.61 [0.30, 1.27]
`
`0.95 [0.37, 2.43]
`
`6.54 [0.65, 66.34]
`
`0.82 [0.47, 1.44]
`
`.1 .2
`1
`Peto OR
`Ctrl
`(95% Cl Fixed)
`n!N
`
`5 10
`Weight Peto OR
`%
`(95% Cl Fixed)
`
`Expt
`n!N
`
`22 I 42
`
`25 I 39
`
`8
`
`Study
`
`24 Months
`ECCDS, 1984
`
`NCCDS, 1979
`
`Smith, 1978 ·
`Subtotal (95%CI) 36 I 95
`Chi-square 0.68 (df=2) Z=1.01
`
`9128
`5125
`
`10124
`4124
`
`39 I 87 -
`96/368 115/386 -
`
`Total {95% Cl)
`Chi-square 7.62 (df=8) Z=1.60
`
`15.4
`
`0.62 [0.26, 1.50]
`
`9.4
`
`5.8
`
`0.67 [0.22, 2.05]
`
`1.24 [0.30, 5.20]
`
`30.6
`
`0.72 [0.39, 1.35]
`
`100.0
`
`0.75 [0.54, 1.06]
`
`FIG. 2. Combined data from
`analysis of three studies show
`odds ratios of relapse for active
`treatment with corticosteroids ver(cid:173)
`sus placebo at 6, 12, and 24
`months after study entry. From
`Steinhart AH, et al. (6) Repro(cid:173)
`duced with permission from Up(cid:173)
`date Software Ltd.
`
`c
`
`.1 .2
`
`1
`
`5 10
`
`Although budesonide may provide short-term steroid(cid:173)
`sparing benefits, little or no evidence supports its use as
`long-term maintenance therapy (6,21-25). Compared
`with placebo, budesonide failed to demonstrate signifi(cid:173)
`cant relapse prevention at dosages of 3 mg/day (22-24)
`and 6 mg/day (21,22,25). Corticosteroid-associated side
`effects were relatively uncommon in patients treated
`with budesonide as maintenance therapy, (21-23,25), but
`Lofberg et al. (22) reported a higher incidence of moon
`face and acne with budesonide versus placebo. Although
`no significant difference has been found in maintenance
`studies between budesonide and placebo groups in ef(cid:173)
`fects on the pituitary-adrenal axis, pituitary-adrenal axis
`
`suppression remains a concern. In addition, the risk for
`dependence associated with budesonide and its effects on
`bone metabolism have not been adequately assessed.
`Further investigation is therefore needed to better define
`the potential role for budesonide in the management of
`pediatric patients with CD who are resistant to, depen(cid:173)
`dent on, or intolerant of steroids.
`
`Immunomodulators
`
`Over the past decade, the immunomodulators azathio(cid:173)
`prine and 6-mercaptopurine have become increasingly
`used in patients with moderate to severe CD, particularly
`
`J Pediatr Gastroenterol Nutr, Vol. 33, Suppl. I, 2001
`
`

`

`530
`
`G. R. LICHTENSTEIN
`
`TABLE 1. Adrerse effects of systemic corticosteroid
`therap_v (7,10,13)
`
`Body system
`
`Cutaneous
`
`Cardiovascular
`
`Endocrine
`
`Gastrointestinal
`
`Metabolic
`
`Multiple
`Neuropsychiatric
`
`Ocular
`
`Musculoskeletal
`
`Reproductive
`
`Adverse effects
`
`Acne
`Striae
`Petechiae
`Ecchymoses
`Hirsutism
`Vascular effects, purpura
`Hypertension
`Atherosclerosis
`Growth retardation (children)
`Hypothalamic-pituitary-adrenal axis
`suppression
`Nausea, vomiting
`Peptic ulcer disease
`Pancreatitis
`Esophagitis
`Hyperglycemia
`Hyperlipidemia
`Electrolyte imbalance
`Fluid retention (edema)
`Alteration of fat distribution (cushingoid
`appearance)
`Obesity
`Hypocalcemia
`Infections
`Emotional lability
`Anxiety
`Psychosis
`Peripheral neuropathy
`Glaucoma
`Cataracts
`Exophthalmos
`Hemorrhage
`Osteonecrosis
`Osteoporosis
`Myopathy
`Muscle atrophy
`Amenorrhea
`
`as maintenance therapy. These and other immunomodu(cid:173)
`latory agents have been evaluated as steroid-sparing
`therapy in controlled trials, with results varying by agent.
`
`Azathioprine/6-Mercaptopurine
`
`In a recent meta-analysis, Sandborn et al. (26) pooled
`data from five controlled studies that used azathioprine
`or 6-mercaptopurine for the treatment of active CD (27-
`
`TABLE 2. Corticosteroid therapy for Crohn's disease:
`incidence of acute side effects of prednisone at 17 weeks
`
`Side effect
`
`Moon facies
`Acne
`Infection
`Ecchymoses
`Hypertension
`Hirsutism
`Petechial bleeding
`Striae
`
`From Singleton JW et al. (7).
`
`Incidence (%)
`
`47
`30
`27
`17
`15
`7
`6
`6
`
`J Pediatr Gastroenterol Nutr, Vol. 33, Suppl. 1, 2001
`
`31 ). In addition to their general efficacy in active int1am(cid:173)
`matory and fistulizing disease, azathioprine and
`6-mercaptopurine were also evaluated for their steroid(cid:173)
`sparing effects. Overall, steroid dosages were reduced to
`a dose of <10 mg of prednisone equivalent in 65% of
`patients treated with the immunomodulators compared
`with 36% of patients treated with placebo (Table 3 ). The
`pooled odds ratio for these studies was 3.86, indicating
`that three patients required treatment with azathioprine to
`achieve a corticosteroid-sparing effect (prednisone < l 0
`mg) in one patient.
`In a small uncontrolled study of pediatric patients with
`IBD (12 with CD, nine with ulcerative colitis) (32), aza(cid:173)
`thioprine 2 mg/kg/day was effective as adjunctive
`therapy to conventional steroid treatment. Six of the pa(cid:173)
`tients with CD had a complete response to the combined
`therapy, and four patients in this group were able to
`discontinue corticosteroids within 6 months after starting
`azathioprine.
`Promising results were also reported in a study evalu(cid:173)
`ating the corticosteroid-sparing efficacy of 6-mer(cid:173)
`captopurine in adolescent patients with CD (33). All 36
`patients had been treated with corticosteroids, sulfasala(cid:173)
`zine, antibiotics, and nutritional support for 5.0 ± 3.0
`years before receiving 6-mercaptopurine, with persis(cid:173)
`tence of intractable symptoms. Among 30 patients
`treated with 6-mercaptopurine for at least 1 year, 29
`(93%) required prednisone in the year before starting
`treatment with 6-mercaptopurine yet remained highly
`symptomatic. During the first 6 months of 6-mer(cid:173)
`captopurine use, 17 (57%) patients were able to discon(cid:173)
`tinue prednisone. At the end of 1 full year of treatment
`with 6-mercaptopurine, 24 (80%) patients no longer re(cid:173)
`quired corticosteroid therapy.
`More recently, Markowitz et al. (34) conducted a
`placebo-controlled study of children (n = 55) with mod(cid:173)
`erately to severely active CD, which also showed that
`6-mercaptopurine enabled the withdrawal of corticoste(cid:173)
`roids after induction of remission and increased the rate
`of maintenance of remission with less corticosteroid use.
`In this study, 6-mercaptopurine (1.5 mg/kg/day) was
`administered to children with a tapering regimen of cor(cid:173)
`ticosteroids (prednisone 40 mg/day tapered according
`to clinical response to 0 mg). At 12 weeks, 6-mer-
`
`TABLE 3. Steroid withdrawal in active Crohn's disease:
`AZA/6-MP efficacya (26 )
`
`Patients achieving a
`reduced steroid doseb
`
`Medication
`Placebo
`
`n (o/c)
`
`76/117 (65)
`39/109 (36)
`
`AZA, azathioprine: 6-MP, 6-mercaptopurine.
`a Pooled data (27-31).
`h Final prednisone dose :s I 0 mg/day.
`
`95% Confidence
`interval
`
`56-74%
`27-45%
`
`

`

`APPROACH TO STEROID-DEPENDENT AND STEROID-REFRACTORY CROHN'S DISEASE
`
`S3!
`
`captopurine did not demonstrate significant efficacy
`compared with placebo for induction of remission. How(cid:173)
`ever, at 12 months after induction of remission, 23 of 24
`pediatric patients (96%) who continued on 6-mercap(cid:173)
`topurine remained in remission versus 15 of 25 patients
`(60%) who continued on placebo (P < 0.01) (34). More(cid:173)
`over, the mean cumulative use of prednisone at 12
`months was significantly lower (P < 0.01) among
`6-mercaptopurine-treated patients ( 4841 mg) than among
`placebo-treated patients (7930 mg).
`The steroid-sparing benefits of azathioprine and
`6-mercaptopurine may be limited in some patients by the
`delayed onset of response and the rare but serious tox(cid:173)
`icities associated with these agents. The potential for
`adverse effects such as pancreatitis, bone marrow sup(cid:173)
`pression, and opportunistic infection is a concern in pe(cid:173)
`diatric patients with CD, requiring regular monitoring.
`An increased risk for immunosuppressive sequelae has
`not been confirmed with these agents, but evaluation is
`ongoing as the numbers of patients receiving long-term
`therapy increase (35).
`
`Methotrexate
`
`To date, no published trials have evaluated methotrex(cid:173)
`ate use in pediatric CD. However, the corticosteroid(cid:173)
`sparing effects of methotrexate have been recognized in
`adults with the disorder in recent years. In a double-blind
`16-week trial, 141 patients with corticosteroid-refractory
`CD were randomized to treatment with intramuscular
`methotrexate at a weekly dose of 25 mg or placebo (36).
`At the end of 16 weeks, significantly more methotrexate(cid:173)
`treated patients than placebo patients were in remission
`and able to discontinue prednisone (39% versus 19%,
`respectively; P = 0.025). Furthermore, patients treated
`with methotrexate required significantly less prednisone
`than those treated with placebo (P = 0.03). However, 16
`of 94 patients (17%) in the methotrexate group discon(cid:173)
`tinued treatment because of adverse effects (including
`nausea, vomiting, and asymptomatic elevation of serum
`aminotransferase levels) compared with 1 of 47 patients
`(2%) receiving placebo (P = 0.01 ), suggesting that side
`effects may compromise the steroid-sparing benefits of
`methotrexate in a substantial proportion of patients.
`A maintenance trial conducted by the investigators of
`the aforementioned trial evaluated patients in remission
`after 16 to 24 weeks of acute treatment with intramus(cid:173)
`cular methotrexate 25 mg/week who were subsequently
`maintained on intramuscular methotrexate 15 mg/week
`for an additional 40 weeks (37). At 40 weeks, signifi(cid:173)
`cantly more patients treated with methotrexate (26/40;
`65%) were in remission than those treated with placebo
`(14/36; 39%; P = 0.04). This response resulted in sig(cid:173)
`nificantly fewer methotrexate-treated patients than
`placebo-treated patients requiring prednisone for disease
`flares (28% versus 58%, respectively; P = 0.01). It is
`
`noteworthy that a large number of patients in both the
`methotrexate ( 42%) and placebo ( 64%) groups discon(cid:173)
`tinued treatment before the end of the 40-week study.
`
`Cyclosporine
`
`Unlike the clinical findings reported with the previ(cid:173)
`ously mentioned immunomodulators, study data do not
`clearly support the use of cyclosporine in steroid(cid:173)
`dependent CD. In three large (n = 305, n = 182, n =
`147) placebo-controlled studies of patients with chronic
`active corticosteroid-resistant or dependent disease, oral
`cyclosporine therapy at a dose of 5 mg/kg/day failed to
`demonstrate either clinical efficacy or a corticosteroid(cid:173)
`sparing effect compared with placebo (38--40). In con(cid:173)
`trast, among patients with active chronic CD (median
`CDAI score 261) who were resistant to or intolerant of
`corticosteroids, clinical improvement was achieved after
`3 months in a significantly higher proportion of patients
`receiving a median oral cyclosporine daily dose of 7.6
`mg/kg (22/37; 59%) than those receiving placebo (11/34;
`32%; P = 0.032) (41). However, although the mean
`reduction from baseline in CDAI score at the end of the
`3-month treatment period in the cyclosporine group was
`significantly greater than that in the placebo group (P =
`0.00012), the final mean CDAI score (CDAI score 227)
`was not below the score that defines remission (CDAI
`<150), indicating that this cyclosporine regimen was in(cid:173)
`effective in inducing disease remission.
`The high incidence of toxicity associated with cyclo(cid:173)
`sporine is also an obstacle to its use in the treatment of
`CD. Adverse events associated with cyclosporine in con(cid:173)
`trolled clinical trials that required treatment discontinu(cid:173)
`ation included paresthesia, hypertrichosis, headache,
`tremor, hypertension, and increased serum creatinine
`levels (38--40). The risk for infection, neuropathy, and
`nephrotoxicity, albeit relatively low, is also a concern.
`Patients with CD treated chronically with cyclosporine
`can be expected to experience a 20% reduction in glo(cid:173)
`merular filtration rate ( 42), which may be associated with
`irreversible nephropathy at dosages >5 mg/kg/day (43).
`In clinical trials, serious adverse events occurred in ap(cid:173)
`proximately 9-12% of patients receiving cyclosporine
`( 44,45), including 3 deaths.
`
`Clinically Available Anti-TNF -a Therapy
`
`lnfliximab
`
`In controlled clinical trials, infliximab has demon(cid:173)
`strated significant efficacy in adult patients with moder(cid:173)
`ately to severely active CD resistant to conventional
`treatment, patients with fistulizing disease, and patients
`with quiescent disease who responded to initial treatment
`with infliximab ( 46--48). Open-label data in adults with
`CD indicate that infliximab also has utility in patients
`
`J Pediatr Gastroenterol Nutr, Vol. 33, Suppl. I, 2001
`
`

`

`S32
`
`G. R. LICHTENSTEIN
`
`who are refractory to or dependent on corticosteroids.
`Cohen et a!. (49) reported on the corticosteroid-sparing
`effects of infliximab in 81 patients with luminal disease
`and 48 patients with fistulizing disease who received at
`least 1 infusion of infliximab. Corticosteroid tapering
`was achieved in >90% of patients with luminal disease
`after an initial infusion; 54% of patients were able to
`completely discontinue corticosteroids after the second
`infusion of infliximab. A sustained median steroid dose
`of 0 mg was achieved from 4 months after infusion on(cid:173)
`ward (Fig. 3).
`Ricart et a!. (50) reported on the clinical outcome and
`safety of 100 patients at the Mayo Clinic treated with
`infliximab for refractory inflammatory and/or fistulizing
`CD. Forty-four patients in this study cohort were evalu(cid:173)
`ated for steroid tapering. Of these, 40 patients (91%)
`underwent steroid tapering, and 29 of these 40 (73%)
`were able to completely discontinue steroid therapy. Me(cid:173)
`dian time of follow-up was 34 weeks (range 14-48
`weeks).
`The steroid-sparing effects of infliximab have also
`been observed in pediatric patients. In a preliminary
`open-label study of 15 pediatric patients (aged 6-18
`years) with medically refractory CD [inability to taper
`steroids, lack of response to immunomodulator therapy
`over 4 months, and active disease as measured by the
`Pediatric Crohn's Disease Activity Index (PCDAI)],
`clinical remission and a significant decrease in daily
`corticosteroid use were observed in 10 patients (67%)
`at 4 and 10 weeks after a single infliximab infusion
`(5 mg/kg) (51). A retrospective chart review of 19 chil(cid:173)
`dren and adolescents with corticosteroid-resistant or
`corticosteroid-dependent CD who received one to three
`infusions of infliximab over 12 weeks revealed signifi(cid:173)
`cant improvement within the first 4 weeks of treatment
`(as evidenced by decrease in PCDAI values; P < 0.0001)
`(52). Mean daily prednisone dosages at baseline, 4
`weeks, and 12 weeks were 28 ± 14 mg, 20 ± 12 mg, and
`8 ± 12 mg, respectively (P < 0.01).
`
`The most common side effects reported with inflix(cid:173)
`imab therapy include headache, nausea, upper respirato(cid:173)
`ry tract infection, abdominal pain, fatigue, and fever. In
`placebo-controlled clinical trials and open trials in pa(cid:173)
`tients with CD and rheumatoid arthritis (n = 771 ), in(cid:173)
`fusion reactions were observed in 17% of patients re(cid:173)
`ceiving infliximab and 7% of those receiving placebo:
`Serious reactions were reported with <0.5% of infusions
`and infusion reactions were the cause for discontinuation
`in less than 2% of patients (53). Analysis of long-term
`safety data (including follow-up of 3 years) indicated no
`increased risk for mortality, serious infections, or malig(cid:173)
`nancy among infliximab-treated patients, although the
`numbers of patients included for analysis are relatively
`low and assessment of long-term safety is ongoing.
`
`Investigational Anti-TNF -a Therapy
`
`CDP-571
`
`Steroid-sparing effects have also been demonstrated
`with the investigational anti-TNF-a drug CDP-571. A
`double-blind, placebo-controlled study evaluated CDP-
`571 in 71 corticosteroid-dependent CD patients (CDAI
`<150 and prednisone dose of 15 to 40 mg/day or budes(cid:173)
`onide dose of 9 mg/day for >8 weeks) (54). Patients were
`randomized to treatment with CDP-571 (20 mg/kg at
`week 0 and 10-mg/kg dose at week 8) or placebo. Pred(cid:173)
`nisone and budesonide were tapered to 0 mg over 10
`weeks in all patients. At week 40, 17 of 39 patients
`(44%) in the CDP-571 group remained free of disease
`flare compared with only seven of 32 patients (22%) in
`the placebo group (P = 0.05). Serious adverse effects
`occurred in both treatment groups (18% for CDP-571;
`13% for placebo). Anti-double-stranded DNA was posi(cid:173)
`tive in 6.7% of patients.
`
`Etanercept
`The human IgG 1 TNF-a receptor fusion antibody
`etanercept has been investigated for its efficacy in the
`
`-<>-Mean
`
`...... Median
`
`30
`
`20
`
`10
`
`c; 40
`.§.
`Cll
`II)
`0
`0
`Cll c:
`0
`-~ c:
`'0
`....
`Cll
`c..
`~
`cv
`0
`MonthsAfter 0+---~--~---±--~~~----~--~~--~~~~~~t FIG. 3. Steroid tapering with in(cid:173)
`fliximab since its initial infusion.
`2
`3
`5
`1
`4
`6
`First Infusion:
`7
`8
`9
`10
`11
`Reproduced with permission
`23
`12
`4
`32
`32
`4
`n= 54
`34
`9
`19
`18
`from Cohen RD, et al. (49).
`
`J Pediatr Gastroenterol Nutr. Vol. 33. Suppl. I. 2001
`
`16
`
`

`

`APPROACH TO STEROID-DEPENDENT AND STEROID-REFRACTORY CROHN'S DISEASE
`
`S33
`
`treatment of active CD, although specific steroid-sparing
`endpoints have not yet been assessed. D'Haens et aL (55)
`conducted a small pilot study in which I 0 patients with
`active Crohn's disease received 25-mg subcutaneous in(cid:173)
`jections of etanercept twice weekly for 12 weeks. Pa(cid:173)
`tients also received prednisone, 5-aminosalicylic acid,
`and azathioprine during this study. Clinical response, de(cid:173)
`fined as a reduction in CDAI score of ?:.70 points, was
`seen in 60% of patients at week 2 of treatment. Median
`CDAI scores decreased from 305 (range 294-418) to 166
`(range I 07-392). Mucosal healing was not observed in
`these patients. A recent randomized, double-blind,
`placebo-controlled trial demonstrated no efficacy over
`placebo when subcutaneous etanercept was given at 25
`mg twice weekly for treatment of active CD (56). Clini(cid:173)
`cal investigation of etanercept as treatment for CD has
`been discontinued by the manufacturer.
`
`Thalidomide
`
`The results of two recent open-label pilot studies sug(cid:173)
`gest that thalidomide may have a corticosteroid-sparing
`effect (57,58). In the first of these 12-week studies, 22
`patients with refractory CD were treated with a bedtime
`dose of thalidomide 200 mg or 300 mg (57). Three of 14
`patients (21%) were able to discontinue corticosteroids
`by week 12. Of I4 patients who completed 12 weeks of
`treatment, nine patients ( 41%) achieved clinical remis(cid:173)
`sion. The second of these studies enrolled 12 patients
`with active, steroid-dependent CD. Half received tha(cid:173)
`lidomide 50 mg nightly and half received thalidomide
`I 00 mg nightly (58). All patients in this study were able
`to taper corticosteroid therapy by ?:.50%, and by week 12
`four of I 0 ( 40%) of patients discontinued corticosteroid
`therapy completely.
`High incidences of dose-related sedation (58-100%)
`and peripheral neuropathy (14-42%) in both studies, as
`well as discontinuation of medication due to adverse
`events by 14% of patients in the first study, suggest that
`the potential benefit of thalidomide as a steroid-sparing
`agent may be outweighed by its toxicity (57,58).
`
`Enteral Nutrition
`
`Enteral nutrition is prescribed in patients with IBD to
`correct nutrient deficiencies and as primary therapy for
`active disease (59). The benefits of this treatment ap(cid:173)
`proach in adult patients include induction of remission,
`correction of nutrient deficiencies, and possible
`corticosteroid-sparing effects, whereas limitations in(cid:173)
`clude failure to maintain remission, high cost, and poor
`compliance. Pooled results of 12 randomized, controlled
`trials in adult CD patients showed an overall remission
`rate of 71% for those receiving an enteral diet versus
`85% of those receiving corticosteroids (59).
`
`Although enteral nutrition has been evaluated in both
`adult and pediatric patients with CD, outcomes have
`been somewhat more positive in children (60). In a pro(cid:173)
`spective study of 44 children (median age 48 months),
`induction of remission occurred in 25 of 30 patients
`(83%) on enteral therapy compared with 18 of 28 pa(cid:173)
`tients (64%) on prednisolone (61). After 12 months of
`treatment, height-for-age z-scores were significantly bet(cid:173)
`ter in the nutritional therapy group than the prednisolone
`group (1.31 versus 2.15, respectively; P < 0.001 ).
`
`CONCLUSIONS
`
`Clinical experience has established the utility of con(cid:173)
`ventional corticosteroids in inducing remission in adults
`and children with moderately to severely active CD.
`However, children are especially vulnerable to the harm(cid:173)
`ful and even permanent side effects of corticosteroids,
`which are numerous and have the potential to affect al(cid:173)
`most every organ system. The pronounced cosmetic
`changes, growth failure, and bone loss experienced in
`young patients treated with steroids are impediments to
`their psychosocial and physical well-being that may out(cid:173)
`weigh steroid-induced relief from CD symptoms. There(cid:173)
`fore, therapeutic agents with steroid-sparing efficacy
`may be particularly beneficial in the management of pe(cid:173)
`diatric CD that requires ongoing treatment.
`Although few studies have been conducted in pediatric
`populations, clinical data suggest that clinically available
`agents with corticosteroid-sparing efficacy include the
`immunomodulators azathioprine/6-mercaptopurine and
`methotrexate and the anti-TNF-a monoclonal antibody
`infliximab (Table 4). Conventional corticosteroids can be
`replaced by budesonide in corticosteroid-dependent pa(cid:173)
`tients without substantially compromising disease stabil(cid:173)
`ity and with a lower risk for toxicity, but no maintenance
`benefit has been documented with this second-generation
`steroid. Preliminary findings suggest that the experimen(cid:173)
`tal anti