Comparison of Oral Prednisone and Topical Fluticasone in the Treatment of Eosinophilic Esophagitis: A Randomized Trial in Children

Article Outline

• Abstract
• Materials and Methods
  • Study Design
  • Patient Selection and Eligibility
  • Randomization and Drug Administration
  • Preliminary Data Collection
  • Esophageal Biopsy Specimens and Histologic Analysis
  • Clinical Response and Analysis
  • Study End Points
  • Data Analysis and Statistics
• Results
  • Patient Characteristics and Clinical Features
  • Histologic Response
  • Clinical Response
  • Adverse Effects
• Discussion
• Acknowledgment
• References
• Copyright

Background & Aims: Although eosinophilic esophagitis is recognized increasingly, outcome data guiding therapy are limited. We conducted a prospective randomized trial comparing oral prednisone (P) and swallowed fluticasone (F) for histologic and clinical response. Methods: Patients were randomized to receive P or F for 4 weeks, followed by an 8-week weaning protocol. Esophageal histology was evaluated at baseline and after 4 weeks of therapy. Clinical assessments were performed at weeks 0, 4, 12, 18, and 24. Results: Eighty patients with eosinophilic esophagitis were enrolled: 40 in the P arm and 40 in the F arm. Histologic improvement was seen in 30 of 32 P and 34 of 36 F patients, with a greater degree of histologic improvement in the P group. All P and 35 of 36 F patients were free of presenting symptom(s) at week 4. Symptom relapse was seen in 45% of patients by week 24. Kaplan–Meier analysis showed no difference between P and F with regard to relapse rate (P = .7399). No significant difference in time to relapse was found between groups (P = .2529). Systemic adverse effects were noted in 40% of the P arm, whereas esophageal candidal overgrowth was seen in 15% of the F arm. Conclusions: Systemic and topical corticosteroids were effective in achieving initial histologic and clinical improvement. P resulted in a greater degree of histologic improvement, without evidence of an associated clinical advantage over F in terms of symptom resolution, relapse rates, or time to relapse. Symptom relapse was common to both groups upon therapy discontinuation, highlighting the need for maintenance treatment protocols.

Abbreviations used in this paper: EE, eosinophilic esophagitis, eos/hpf, eosinophils per high-power field, F, fluticasone propionate, GER, gastroesophageal reflux, P, prednisone

Allergic eosinophilic esophagitis (EE) continues to be an increasingly recognized cause of feeding refusal, recurrent vomiting, epigastric discomfort, dysphagia, and food impaction among children and adults.¹, ², ³ In 1982, Winter et al⁴ described the presence of eosinophils in the esophagus as an indicator of esophagitis induced by gastroesophageal reflux (GER). In most cases of GER, the eosinophilic infiltration of the esophageal epithelium is mild with 5 or fewer eosinophils present per high-power field (eos/hpf).⁵, ⁶, ⁷, ⁸ In 1993, Attwood et al⁵ reported high intraepithelial eosinophil counts (>20 eos/hpf) in 12 adult patients with dysphagia. Twenty-four–hour esophageal pH monitoring was normal in 11 of these 12 patients. These findings, in addition to the observation of persistent esophageal symptoms despite medical and surgical therapies for GER, as reported by Kelly et al⁹ in 1995, led to further investigation of the clinical and diagnostic features of patients with EE.

EE has emerged as a separate clinicopathologic entity over the past decade that is distinguished by the absence of pathologic GER on pH probe study and/or failure to respond to treatment with proton pump inhibitors, along with the presence of large numbers of eosinophils in the esophageal epithelium.⁸ Diagnostic criteria of EE continue to evolve. It generally is accepted that one of the pathologic features distinguishing EE from GER is dense esophageal mucosal eosinophilia, defined as 15–20 or more eos/hpf.⁶, ⁷, ⁸, ⁹, ¹⁰, ¹¹, ¹², ¹³, ¹⁴ Basal zone hyperplasia and papillary lengthening are associated findings.¹⁵ EE is characterized further by eosinophilic inflammation confined to the esophagus with normal gastric and duodenal mucosal biopsy specimens.¹²

The current primary treatment options for EE are limited to dietary modifications and corticosteroids (systemic or topical). Initial studies have analyzed therapeutic response to the earlier-described interventions individually. Kelly et al⁹ reported that a 6-week elimination of potential food allergens using an amino acid–based formula in 10 children with intense eosinophilic infiltration of the esophageal epithelium resulted in resolution of symptoms in 8 children and clinical improvement in the remaining 2 children. Relapse was seen in 9 of the 10 patients after systematic food challenge. A larger trial of 51 patients with EE also showed significant clinical response within 10 days and histologic reduction of eosinophil infiltration after 1 month of an elemental diet.¹⁶ Liacouras et al⁷ reported 20 pediatric patients unresponsive to 3 months of GER therapy who responded clinically and histologically to 1 month of oral methylprednisolone (P). Faubion et al¹⁷ successfully treated 4 pediatric EE patients with a topical corticosteroid through a metered dose inhaler. A recent placebo-controlled trial showed that swallowed (topical) fluticasone propionate (F) is effective in inducing histologic remission in EE, especially in the proximal esophagus.¹⁸

Because anti-inflammatory therapy has been shown to be effective in treating EE, several questions have been raised regarding preferred formulation (systemic vs topical), duration of therapy, expected outcomes, and potential for adverse effects. To date, outcome data comparing systemic and topical corticosteroids for the treatment of EE have not been reported. Based on available comparative data of oral P and inhaled F in severe pediatric asthma, Schuh et al¹⁹ reported a significant clinical advantage with the use of oral P, with a greater improvement in pulmonary function in the initial 4 hours after treatment and a decreased rate of hospitalization after therapy. Recognizing the potential for increased efficacy of systemic steroids when compared with topical steroids in asthma patients, we questioned if systemic therapy would yield a more favorable response when compared with topical steroids in the treatment of another steroid-responsive disorder such as EE. Based on this question, we conducted a prospective study of pediatric patients with EE comparing oral systemic P and swallowed F for histologic improvement and clinical response. We expanded the analysis to evaluate clinical relapse patterns and to evaluate for adverse effects associated with the 2 corticosteroid formulations. Because allergy has been reported as a commonly associated finding among EE patients,²⁰, ²¹, ²² we sought to characterize the demographic and clinical features of patients with EE.

Back to Article Outline

Materials and Methods 

Study Design 

We conducted a 24-week, open-labeled, randomized, prospective study comparing response to oral P and swallowed F in children with EE (Figure 1). Study patients received 12 weeks of therapy (4-week induction, 8-week weaning protocol) followed by a 12-week period of follow-up off medications. Esophagogastroduodenoscopy was performed at baseline (week 0) and after 4 weeks of drug administration (week 4). Clinical assessment by physicians was performed at week 0, after 4 weeks of full-dose therapy (week 4), after an 8-week tapering dose regimen (week 12), and off medication at weeks 18 and 24 (study end).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Flow diagram of study participants. ^aSee Figure 2A. ^bSee Figure 2B.

Patient Selection and Eligibility 

The study was approved by the local institutional review board and conducted through the outpatient clinic population of the Division of Pediatric Gastroenterology, Hepatology, and Nutrition at the James Whitcomb Riley Hospital for Children, Indiana University, Indianapolis, Indiana. Patients between 1 and 18 years of age diagnosed with EE (defined as esophageal mucosal biopsy specimens showing ≥15 eos/hpf with negative pH probe studies) were eligible. We prospectively recruited eligible patients from February 2000 to November 2004. Exclusion criteria included co-existing esophageal conditions (eg, stricture, Barrett’s esophagus, caustic injury), Helicobacter pylori infection, inflammatory bowel disease, and inability to tolerate corticosteroids. None of the patients were on corticosteroids at the time of initial endoscopy or at study enrollment. Informed consent was obtained on enrollment.

Randomization and Drug Administration 

Random number assignment was used to generate a concealed allocation schedule that was maintained by a research coordinator. No blocking or stratification schemes were used. The allocated treatment was revealed to enrolled subjects after consent was obtained by the treating physician. Patients received either oral P suspension/tablet (1 mg/kg/dose twice a day; maximum 30 mg twice a day) or swallowed F by metered-dose inhaler (110 μg per puff for ages 1–10 years and 220 μg per puff for ages 11 years or older, 2 puffs 4 times/day) for 4 weeks. Patients receiving F were instructed specifically to use the inhaler without a spacer, swallow the medicine, and not to eat or drink for 30 minutes after F administration. The drugs were weaned off over the next 8 weeks (Table 1). Symptoms and compliance with therapy were assessed by the patient and/or the parent/guardian by maintaining a daily symptom diary.

Table 1. Steroid-Weaning Protocol

Preliminary Data Collection 

Before initiating treatment, study patients were evaluated for allergies by radioallergoabsorbent tests (Clarian Health Department of Laboratory Medicine, Indianapolis, IN) or percutaneous skin testing with commercial extracts (Hollister-Stier Laboratories, Spokane, WA). During the entire study, patients continued a regular diet except for foods identified as possible allergens by allergy testing. Baseline complete blood count with differential and serum immunoglobulin E levels were obtained before the initiation of therapy.

Esophageal Biopsy Specimens and Histologic Analysis 

Patients underwent esophagogastroduodenoscopy before and after receiving 4 weeks of corticosteroid therapy. Esophagogastroduodenoscopy was performed using the Olympus GIF-140 or GIF-160 Videoscope (Olympus America, Inc, Melville, NY). Grasp forceps were used to obtain at least 2 mucosal biopsy specimens from the distal esophagus taken from 3 cm proximal to the esophagogastric junction at diagnostic endoscopy. In the vast majority, 2 biopsy specimens of the midesophagus (at 8 cm proximal to the esophagogastric junction) were obtained but not required at initial endoscopy in those suspected to have EE; all but one patient with confirmed EE had distal and midesophageal biopsy specimens obtained at week 4. Gastric and duodenal biopsy specimens were obtained at baseline. The specimens were oriented on filter paper, formalin-fixed, paraffin-embedded, cut serially, and stained with H&E. All biopsy specimens were reviewed by pediatric pathologists. A histologic grading system was based on a review of available literature.⁹, ²³, ²⁴, ²⁵, ²⁶ For the purposes of this study, points were assigned based on (1) basal cell zone thickness as a percentage of the epithelial thickness, and (2) the maximum number of eos/hpf (at 400× power using an eyepiece grid covering an area of 0.4 mm²) (Table 2). Points were summed and the totals were translated into histologic grades (normal, mild, moderate, and severe). Grades were assigned a numeric value for statistical analysis.

Table 2. Histologic Evaluation of Esophageal Biopsy Specimens

NOTE. Composite biopsy score = points for basal cell zone + points for number of eos/hpf. Histologic grade was based on composite biopsy scores: 0, normal; 1–2, mild; 3–4, moderate; 5–6, severe.

Clinical Response and Analysis 

A daily symptom diary was maintained by the patient/guardian while on corticosteroid therapy. Clinical assessment was performed at weeks 4, 12, 18, and 24 to monitor for the presence or absence of the presenting esophageal symptom(s). Reported symptoms included vomiting, abdominal pain, epigastric pain, heartburn, dysphagia, feeding problems, foreign body/food impaction, and weight loss.

Study End Points 

The primary end point was defined as a histologic response by an improvement in biopsy grade after 4 weeks of corticosteroid therapy. The secondary end point was defined as a clinical response to corticosteroids based on the presence or absence of the presenting symptom by patient/guardian report and by physician assessment at predetermined intervals.

The patient exited the study if any of the following occurred: (1) clinical resolution was not achieved with the allocated therapy, (2) no histologic improvement was found, (3) relapse of symptoms was noted at follow-up evaluation, and/or (4) if the subject was noncompliant with the medication or tests. Clinical response data were analyzed based on an intention-to-treat approach. In keeping with intention-to-treat analysis, complete data from all randomized patients were included, unavailable data were censored, and protocol deviations were reported. This approach was selected with the goal of providing a pragmatic estimate of the benefit of the change in treatment policy rather than of potential benefit in evaluating only the patients who received treatment exactly as planned.

Data Analysis and Statistics 

The primary analysis compared the P and F groups for an improvement in esophageal mucosal histology by a grade of 1 or more after 4 weeks of treatment. A signed rank test was used to assess overall improvement in biopsy scores. A Mann–Whitney test was used to assess for a difference in median histologic improvement between the P and F arms. Statistical power of 87% was achieved with 40 participants allocated to each group to detect a difference in histologic change between the groups of 0.7 SDs. Statistical significance was determined by an α-level P value of less than .05.

Further analysis compared clinical response in each group. Symptoms were assessed at week 4 (after 4 weeks of full-dose therapy), and weaning off therapy at weeks 12, 18, and 24. The groups were compared for differences in the proportion of patients with resolution of the presenting EE symptom(s) at week 4. Recurrence of the presenting EE symptom(s) was evaluated at follow-up intervals at weeks 12, 18, and 24. Survival analysis was used and a log-rank test was performed to examine the difference in resolution of symptoms between the 2 groups.

Statistical analysis was performed using SAS version 9.1.3 (SAS Institute, Inc, Cary, NC) and survival data were generated by using S-PLUS version 7.0 (Insightful Corp, Seattle, WA).

Back to Article Outline

Results 

Patient Characteristics and Clinical Features 

Eighty patients with EE were enrolled: 40 patients were randomized to the P arm and 40 patients were randomized to the F arm (Figure 2). Demographic characteristics at enrollment were similar between treatment arms (Table 3). A male predominance (74%) was found, consistent with previous studies.³, ²¹, ²⁷, ²⁸ Sixty-one percent of patients presented with vomiting, 29% had abdominal pain (10% specifically complained of epigastric pain or heartburn symptoms), and 19% presented with feeding problems or dysphagia. Approximately 44% of patients had a history of atopy and 44% had a first-degree relative with a history of allergic symptoms. Food allergy testing was positive in 58% of patients. An increased serum immunoglobulin E level was found in 35% (26 of 74; mean, 177.7 kU/L; median, 75.3 kU/L; range, 2–1036 kU/L), and peripheral eosinophilia (defined as >350 cells/mm³)²⁹ was present in 67% (51 of 76; mean, 579 cells/mm³; median, 500 cells/mm³; range, 0–4000 cells/mm³) of EE patients.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  (A) Flow diagram of subjects randomized to P. ^aNoncompliance with medication, n = 4; patient/family decision, n = 2; lost to follow-up evaluation, n = 1; adverse effect, n = 1 (increased appetite and abdominal pain). ^bBoth exited because of adverse effects, n = 2 (cushingoid, acne, behavioral changes). (B) Flow diagram of subjects randomized to F. ^aNoncompliance with medication, n = 2; patient/family decision, n = 2. ^bPersistence of symptoms, n = 1; asymptomatic, but no histologic improvement, n = 1.

Table 3. Patient Demographics and Clinical Characteristics

	P group N = 40	F group N = 40	P value
Mean age, y	7.0	7.2	NSa
Age range, y	1.1–15.8	1.1–16.1
Male	31(78%)	28(70%)	NS
Presenting symptomsb
Vomiting/regurgitation	23(58%)	26(65%)	NS
Abdominal pain	11(28%)	12(30%)	NS
Epigastric pain/heartburn	4(10%)	4(10%)	NS
Dysphagia/feeding problem	9(23%)	6(15%)	NS
Food/foreign-body impaction	3(8%)	1(3%)	NS
Weight loss	1(3%)	3(8%)	NS
Clinical data at presentationc
Increased serum immunoglobulin E level	12/38(32%)	14/36(39%)	NS
Peripheral blood eosinophilad	21/37(57%)	30/39(77%)	.0394
Positive allergy history(overall)	18/40(45%)	17/40(43%)	NS
Eczema	6/40(15%)	3/40(8%)	NS
Allergic rhinitis	7/40(18%)	7/40(18%)	NS
Wheezing, asthma	6/40(15%)	12/40(30%)	NS
Positive food allergy testing	25/40(63%)	21/40(52.5%)	NS
Immediate relative with history of allergic symptoms	21/40(53%)	14/40(35%)	NS

NS, not significant.

Histologic Response 

Distal esophageal biopsy results were available for all patients at week 0, 68 patients at week 4, and midesophageal biopsies were available for 70 patients at week 0 and 67 patients at week 4 (Table 4). Gastric and duodenal biopsy specimens of study patients were normal or with minimal changes at baseline except for 4 patients who had gastric eosinophilia (>20 eos/hpf). Complete biopsy data for comparison between weeks 0 and 4 were available for 32 P and 36 F patients (Table 5). No difference in histologic grade at baseline was found between P and F groups (P = .2908). Histologic improvement by a grade of 1 or more was seen in 30 of 32 P and 34 of 36 F patients, with complete histologic resolution (defined as normal biopsy specimens at week 4) in 26 of 32 P and 18 of 36 F patients. A signed rank test showed histologic improvement in the biopsy score by a grade of 1 or more after 4 weeks of steroid therapy in both the P and F arms (P < .0001). Mann–Whitney analysis revealed a statistically significant difference in histologic response with the P group showing a greater degree of improvement compared with the F group between weeks 0 and 4 (P = .0440). Figure 3 illustrates an example of pretreatment and posttreatment histologic findings.

Table 4. Individual Patient Symptom and Pretreatment/Posttreatment Esophageal Biopsy Data

AP, abdominal pain; BCH, basal cell hyperplasia; Distal, distal esophagus; Dy, dysphagia; EP, epigastric pain; FB, foreign body; FP, feeding problems; HB, heartburn; Mid, midesophagus; NA, data not available; R, regurgitation; V, vomiting; Wt loss, weight loss.

Table 5. Histologic Grade Results at Weeks 0 and 4 in the 2 Treatment Groups

• 
  • View Large Image
  • Download to PowerPoint

• Figure 3. 

  Histology images of the distal esophagus at the (A) time of diagnosis and (B) after 4 weeks of F demonstrate basal cell hyperplasia comprising 75% of the epithelial thickness and eosinophil counts numbering 30 eos/hpf (40× objective, total 400× magnification), and normalization of epithelium after treatment (20× objective, total 200× magnification), respectively.

Clinical Response 

After 4 weeks of therapy, resolution of the presenting symptom(s) was documented in 100% of P patients and 97.2% of F patients. We examined the proportion of symptom-free patients at follow-up visits. Kaplan–Meier analysis was performed including all 80 patients based on intention-to-treat analysis. A log-rank test was used to compare survival curves between P and F. No statistically significant difference in the proportion of symptom-free patients was found between the 2 groups (Figure 4, log rank P value = .7399).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 4. 

  Kaplan–Meier curve showing proportion of symptom-free patients at each visit.

Accounting for all patients enrolled, including patients who may have failed therapy or were lost to follow-up evaluation, the mean time patients remained in the study was 15.3 weeks (SD = 9.7) for the P group and 15.6 weeks (SD = 9.6) for the F group (P = .8895). Among the patients who relapsed after weaning P or F, the mean number of weeks from discontinuing therapy to symptom relapse was 5.5 weeks (SD = 4.0; n = 12) for the P group and 8 weeks (SD = 6.1; n = 11) for the F group (P = .2529).

Chi-square analysis performed at each follow-up interval showed no significant difference between P and F groups in the proportion of patients with symptom recurrence (Table 6). Overall, 45% (23 of 51) of the patients with documented follow-up data experienced symptom relapse by week 24.

Table 6. Proportion of Patients With Persistence/Recurrence of EE Symptoms by Visit Week

	Week 4	Week 12	Week 18	Week 24
Prednisone	0.0%(0/32)	11.1%(3/27)	35.7%(10/28)	44.4%(12/27)
Fluticasone	2.8%(1/36)	3.6%(1/28)	21.7%(5/23)	45.8%(11/24)
Chi-square P value	.3422	.2817	.2758	.9207

Adverse Effects 

Systemic adverse effects (hyperphagia, weight gain, and/or cushingoid features) were noted in 16 (40%) patients in the P arm and none in the F arm. Of these 16 P patients, 3 exited the study before week 4 as a result of severity of adverse effects and were transitioned to the F group (outside the protocol). Esophageal candidal overgrowth was seen in 6 (15%) patients in the F arm and in none in the P arm. All of the F patients with candidiasis were free of the presenting symptom(s) by week 4.

Back to Article Outline

Discussion 

The limited understanding of the pathogenesis and disease course has prompted therapeutic decisions based on expert opinion and data limited to small case series and trials. This study represents an open-label prospective trial designed to compare systemic versus topical corticosteroid therapy in pediatric patients with EE.

The vast majority of patients in this study showed both symptomatic and histologic improvement with P and F at week 4. This study compared the 2 drugs and our results are in keeping with earlier studies that examined each drug individually. Oral P was the first medical treatment shown to be effective in treating EE.⁷ Swallowed F was shown to be effective in the initial observations by Faubion et al,¹⁷ followed by a prospective study by Teitelbaum et al¹³ showing that F is effective in treating EE. A recently published placebo-controlled trial of F in the treatment of children with EE showed histologic remission after 3 months of therapy.¹⁸ Our results show rapid initial clinical and histologic response to corticosteroid therapy in systemic and topical forms within 4 weeks of therapy.

Analysis of histologic specimens showed a significant improvement in biopsy grade at week 4 for both the P and F arms. Although better histologic improvement was noted in the P group as compared with the F group, this did not translate into long-term clinical remission. Kaplan–Meier estimates of the 24-week period after initiation of therapy show the probability of being symptom free at each follow-up interval, with no difference found between the P and F arms. No difference was detected between groups in symptom resolution, relapse rates, or time to relapse. The therapeutic challenge of EE is compounded by disease relapse on discontinuation of therapy.¹, ⁷, ⁹, ¹⁰, ³⁰, ³¹ In this trial, symptom relapse was common to both groups and occurred within a mean of 5.5 to 8 weeks from discontinuing therapy. Ultimately, by week 24 of the study (12 weeks after corticosteroid discontinuation), nearly half of the patients in the study experienced symptom relapse. The significant percentage of symptom relapse makes long-term treatment decisions difficult. Although absolute risks of nonintervention currently are unknown, observational data suggest that persistent esophageal eosinophilia may be associated with esophageal thickening or stricture.²⁹, ³⁰, ³¹ The response rate in this study may be attributed to the higher corticosteroid doses used, which were based on published data available at the time of the initial protocol development (Liacouras et al⁷ and Faubion et al¹⁷). It is reasonable to consider that improved topical F coverage could contribute to the successful initial response rate.

In our experience, symptomatic improvement coincided with histologic remission. In 4 patients, however, histology showed minimal or no improvement at week 4, although clinical improvement was reported in all 4 instances (2 P and 2 F patients). One asymptomatic P patient with unchanged severe histologic classification later admitted to incomplete medication compliance; the other asymptomatic P patient had improvement of maximal number of eosinophils from 40 eos/hpf to a focal area of 6 eos/hpf, but the grade classification remained moderate because of unimproved basal cell zone thickening. One patient in the F group showed partial histologic improvement in the number of eosinophils (from a maximal count of 53 eos/hpf at diagnosis to 30 eos/hpf at week 4). It was discovered that the patient had been inhaling instead of swallowing F, which could account for partial improvement. The other asymptomatic F patient showed persistent severe histology, but was administering F at a suboptimal dose.

As the earlier-described cases illustrate, underlying reasons for histologic and clinical discrepancies are not always rapidly evident and require detailed history. Problems with medication compliance may explain histologic partial or nonresponse. In addition, the role of positive food allergy testing in partial/nonresponders should be investigated further because all 4 of the patients outlined earlier had positive food allergy testing. A recent study noted that upper-esophageal histologic response to F was better in nonallergic individuals,¹⁸ whereas other data suggested that short- and long-term clinical response to steroids was not shown to be related to the presence or absence of food allergy.³² The implications of extra-esophageal allergic disease and response to EE therapy require further study. We observed 4 cases with gastric eosinophilia at diagnostic endoscopy. Improvement in esophageal eosinophilia was found in all 4 patients, normalization of gastric eosinophilia in 3 of 4 patients, and all 4 patients reported being asymptomatic by week 4. Exclusion of these 4 patients from the analysis did not alter the study results (data not shown). The significance of mild eosinophilia in other portions of the upper gastrointestinal tract in patients with EE is unclear.

In addition to lack of long-term benefit despite greater histologic improvement with P, therapy with P was associated with several side effects and more patient attrition. Systemic adverse effects were noted in 40% of patients in the P arm and none in the F arm; 3 patients exited the study before week 4 owing to the severity of adverse effects and were transitioned to F (outside the protocol). Esophageal candidal overgrowth was seen in 6 (15%) patients in the F arm and none in the P arm. All of the F patients with candidiasis were free of the presenting symptom(s) by week 4. Candidiasis was not accompanied by esophageal mucosal inflammation, indicating an overgrowth rather than true candidal esophagitis. Our experience is consistent with that of Teitelbaum et al,¹³ showing that F is reasonably easy to administer orally with good compliance and has the advantage of less systemic side effects.

Limitations to this study relate to practical matters such as inability to double blind the treatment regimens because one is in oral liquid/tablet form and the other is administered by a metered dose inhaler. Data collection of the symptom assessment was performed by maintaining a patient diary noting the presence or absence of symptoms during the 12-week period on medication, although, in retrospect, maintaining the diary throughout the 24-week study period would have been desirable. Lastly, patient attrition was encountered in both groups. We surmise it is possible that patients lost to follow-up evaluation may have experienced resolution of symptoms and felt no need to return. Symptomatic patients or those with difficulty adhering to the medication regimen ultimately exited the study. For practical purposes the research protocol defined the study end point at week 24, 12 weeks after medication was discontinued. Extended analysis of outcomes beyond week 24 would provide long-term data characterization of clinical and histologic response, lending insight to the natural course of the disease. A 9% to 18% placebo response in children with EE was reported recently and this variable should be considered in the natural history of this disease.¹⁸

We acknowledge the potential for allergy testing and food elimination to act as a confounder, but in the presence of existing data we believed that it was clinically important for patient care to offer allergy testing. Once presented with allergy testing results, it made practical sense to eliminate any identified food allergen(s), although we did not formally assess and monitor these dietary modifications during the study. It has, however, been our experience that dietary modifications can be difficult to implement, the compliance is variable, and results are unsatisfactory. For example, 11 of the study patients reported being on a food elimination diet at week 0 with partial or no clinical response, and all had histologic grades at week 0 that were classified as severe despite dietary modifications. Other investigators too reported a lack of clinical resolution with dietary modifications based on results of percutaneous skin tests or a radioallergoabsorbent test.¹³, ³³ In a report of 51 children with EE treated with an elemental diet, Markowitz etal¹⁶ stated that as yet there is no single test that can best identify food allergens in patients with EE. In our study, patients were advised to continue avoidance of known food allergens for the duration of the entire study (including during and after the corticosteroid weaning period and off corticosteroid). Further, because relapse rates were similar among patients who tested food allergy positive and those who tested negative, we do not believe dietary modifications played a significant confounding role in the study results.³²

The decision to approach histologic analysis with a composite grading score allowed us to incorporate not only absolute esophageal eosinophil count but also basal cell hyperplasia to represent changes reflective of EE and effects on therapy. Specific levels of quantitative eosinophil counts have been presented recently by Konikoff et al,¹⁸ however, at the time of this study’s initial protocol development, in consultation with the pediatric pathologists at our institution, and after an extensive review of the literature, we decided that a composite grading score accounting for the number of eosinophils and basal cell zone hyperplasia (as a percentage of total epithelial thickness) would be used to detect histologic response.⁹, ²³, ²⁴, ²⁵, ²⁶ This was also in deference to Kelly et al,⁹ who also used a grading system to assess esophageal inflammation. Although the statistical analysis was based on the original protocol’s grading score, we provide the reader with detailed individual patient data, including histologic findings (Table 4).

In summary, this prospective randomized trial showed that as primary therapy both systemic and topical corticosteroids showed success in achieving initial histologic and clinical improvement. Although P resulted in a greater degree of histologic improvement, there was no evidence for an associated short- or long-term clinical advantage over F. Systemic adverse effects were associated commonly with P. F patients experienced a lower frequency of adverse effects, with esophageal candidiasis as the only reported complaint. This study illustrated that clinical efficacy can be achieved by topical corticosteroids with less risk of systemic adverse effects, which suggests that F provides a favorable treatment option. Our results show a strikingly high relapse rate, suggesting a strong need for the development of maintenance treatment protocols. Long-term management and remission maintenance is a challenge and data on remission maintenance are limited. Longitudinal studies are needed to determine the natural course of EE, effective treatment options, and follow-up recommendations for these patients.

Back to Article Outline

The authors are grateful for the insight provided by Drs Phillip Faught, Mary Davis, and Kevin Kernek of Pediatric Pathology, Dr Fredrick Leickley of Pediatric Allergy and Immunology, and the Indiana University School of Medicine Division of Biostatistics. The authors would like to thank Gail Waltz, CPNP, Elizabeth Mitchell, RN, Miriam Davis, and Sharon McPheeters for their technical support.

Back to Article Outline

References 

1. Liacouras CA. Eosinophilic esophagitis in children and adults. J Pediatr Gastroenterol Nutr. 2003;37:S23–S28
   • View In Article
   • CrossRef
2. Sant’Anna AMGA, Rolland S, Fournet JC, et al. Eosinophilic esophagitis in children: symptoms, histology and pH probe results. J Pediatr Gastroenterol Nutr. 2004;37:373–377
   • View In Article
3. Noel RJ, Putnam PE, Rothenberg ME. Eosinophilic esophagitis. N Engl J Med. 2004;351:S23–S28
   • View In Article
4. Winter HS, Madera JL, Stafford RJ, et al. Intraepithelial eosinophils: a new diagnostic criterion for reflux esophagitis. Gastroenterology. 1982;83:818–823
   • View In Article
   • Abstract
5. Attwood SEA, Smyrk TC, DeMeester TR, et al. Esophageal eosinophila with dysphagia: a distinct clinicopathologic syndrome. Dig Dis Sci. 1993;38:109–116
   • View In Article
   • MEDLINE
   • CrossRef
6. Steiner SJ, Gupta SK, Croffie JM, et al. Correlation between number of eosinophils and reflux index on same day esophageal biopsy and 24 hour esophageal pH monitoring. Am J Gastroenterol. 2004;99:801–805
   • View In Article
   • MEDLINE
   • CrossRef
7. Liacouras CA, Wenner WJ, Brown K, et al. Primary eosinophilic esophagitis in children: successful treatment with oral corticosteroids. J Pediatr Gastroenterol Nutr. 1998;26:380–385
   • View In Article
   • MEDLINE
   • CrossRef
8. Walsh SV, Antonioli DA, Goldman H, et al. Allergic esophagitis in children: a clinicopathological entity. Am J Surg Pathol. 1999;23:390–396
   • View In Article
   • MEDLINE
   • CrossRef
9. Kelly KJ, Lazenby AJ, Rowe PC, et al. Eosinophilic esophagitis attributed to gastroesophageal reflux: improvement with an amino acid-based formula. Gastroenterology. 1995;109:1503–1512
   • View In Article
   • Abstract
   • Full-Text PDF (13151 KB)
   • CrossRef
10. Gupta SK, Fitzgerald JF, Kondratyuk T, et al. Cytokine expression in normal and inflamed esophageal mucosa: a study into the pathogenesis of allergic eosinophilic esophagitis. J Pediatr Gastroenterol Nutr. 2006;42:22–26
    • View In Article
    • MEDLINE
    • CrossRef
11. Wang F, Gupta SK, Fitzgerald JF. Is there a seasonal variation in the incidence and severity of allergic eosinophilic esophagitis (AEE)?. Gastrointest Endosc. 2005;61:AB305
    • View In Article
    • Full Text
    • Full-Text PDF (41 KB)
    • CrossRef
12. Furuta GT, Straumann A. Review article: the pathogenesis and management of eosinophilic oesophagitis. Aliment Pharmacol Ther. 2006;24:173–182
    • View In Article
    • MEDLINE
    • CrossRef
13. Teitelbaum JE, Fox VL, Twarog FJ, et al. Eosinophilic esophagitis in children: immunopathological analysis and response to fluticasone propionate. Gastroenterology. 2002;122:1216–1225
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (291 KB)
    • CrossRef
14. Steiner SJ, Kernek KM, Fitzgerald JF. Severity of basal cell hyperplasia differs in reflux versus eosinophilic esophagitis. J Pediatr Gastroenterol Nutr. 2006;42:506–509
    • View In Article
    • CrossRef
15. Furuta GT, Straumann A. Eosinophilic esophagitis: an emerging clinicopathologic disease of children and adults. Gastroenterol Hepatol. 2006;2:371–374
    • View In Article
16. Markowitz JE, Spergel JM, Ruchelli E, et al. Elemental diet is an effective treatment for eosinophilic esophagitis in children and adolescents. Am J Gastroenterol. 2003;98:777–782
    • View In Article
    • MEDLINE
    • CrossRef
17. Faubion WA, Perrault J, Burgart LJ, et al. Treatment of eosinophilic esophagitis with inhaled corticosteroids. J Pediatr Gastroenterol Nutr. 1998;27:90–93
    • View In Article
    • MEDLINE
    • CrossRef
18. Konikoff MR, Noel RJ, Blanchard C, et al. A randomized, double-blind, placebo-controlled trial of fluticasone propionate for pediatric eosinophilic esophagitis. Gastroenterology. 2006;131:1381–1391
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (777 KB)
    • CrossRef
19. Schuh S, Reisman J, Alsherhri M, et al. A comparison of inhaled fluticasone and oral prednisone for children with severe acute asthma. N Engl J Med. 2000;343:689–694
    • View In Article
    • MEDLINE
    • CrossRef
20. Orenstein SR, Shalaby TM, Di Lorenzo C, et al. The spectrum of pediatric eosinophilic esophagitis beyond infancy: a clinical series of 30 children. Am J Gastroenterol. 2000;95:1422–1430
    • View In Article
    • MEDLINE
    • CrossRef
21. Cheung KM, Oliver MR, Cameron DJ, et al. Esophageal eosinophilia in children with dysphagia. J Pediatr Gastroenterol Nutr. 2003;37:498–503
    • View In Article
    • MEDLINE
    • CrossRef
22. Gonsalves N, Policarpio-Nicolas M, Zhang Q, et al. Histopathologic variability and endoscopic correlates in adults with eosinophilic esophagitis. Gastrointest Endosc. 2006;64:313–319
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (505 KB)
    • CrossRef
23. Knuff TE, Benjamin SB, Worsham GF, et al. Histologic evaluation of chronic gastroesophageal reflux (An evaluation of biopsy methods and diagnostic criteria). Dig Dis Sci. 1984;29:194–201
    • View In Article
    • MEDLINE
    • CrossRef
24. Hyams JS, Ricci A, Leichtner AM. Clinical and laboratory correlates of esophagitis in young children. J Pediatr Gastroenterol Nutr. 1988;7:52–56
    • View In Article
    • MEDLINE
    • CrossRef
25. Shub MD, Ulshen MH, Hargrove CB, et al. Esophagitis: a frequent consequence of gastroesophageal reflux in infancy. J Pediatr. 1985;107:881–884
    • View In Article
    • Abstract
    • Full-Text PDF (373 KB)
    • CrossRef
26. Goldman H, Antonioli DA. Mucosal biopsy of the esophagus, stomach, and proximal duodenum. Hum Pathol. 1982;13:423–448
    • View In Article
    • MEDLINE
27. Baxi S, Gupta SK, Swigonski N, et al. Clinical presentation of patients with eosinophilic inflammation of the esophagus. Gastrointest Endosc. 2006;44:473–478
    • View In Article
    • Full Text
    • Full-Text PDF (1223 KB)
    • CrossRef
28. Straumann A, Spichtin HP, Grize L, et al. Natural history of primary eosinophilic esophagitis: a follow-up of 30 adult patients for up to 11.5 years. Gastroenterology. 2003;125:1660–1669
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (317 KB)
    • CrossRef
29. Rothenberg ME. Eosinophilia. N Engl J Med. 1998;338:1592–1600
    • View In Article
    • MEDLINE
    • CrossRef
30. Liacouras CA, Spergel JM, Ruchelli E, et al. Eosinophilic esophagitis: a 10-year experience in 381 children. Clin Gastroenterol Hepatol. 2005;3:1198–1206
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (516 KB)
    • CrossRef
31. Khan S, Orenstein SR, Di Lorenzo C, et al. Eosinophilic esophagitis: strictures, impactions, dysphagia. Dig Dis Sci. 2003;48:22–29
    • View In Article
    • MEDLINE
    • CrossRef
32. Schaefer ET, Gupta SK. Short-term and long-term clinical response to corticosteroids in children with eosinophilic esophagitis: do these vary by food allergy testing? (abstract ID 324803). (abstr) Gastroenterology. 2007;132:A608
    • View In Article
33. Noel RJ, Putnam PE, Collins MH, et al. Clinical and immunopathologic effects of swallowed fluticasone for eosinophilic esophagitis. Clin Gastroenterol Hepatol. 2004;2:568–575
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (367 KB)
    • CrossRef