What Is Known/What Is New
What Is Known
- The current guidelines to assess disease activity and response to treatment in pediatric eosinophilic esophagitis require repeat endoscopy with esophageal biopsy to confirm resolution of eosinophilia.
- There are no blood-based biomarkers that accurately assess disease activity and distinguish eosinophilic esophagitis from other comorbid atopic diseases.
What Is New
- Blood eosinophil progenitor levels are higher in patients with active compared to inactive eosinophilic esophagitis disease.
- The eosinophil progenitor level shows promise as a biomarker to monitor eosinophilic esophagitis disease activity in patients undergoing dietary therapy and may decrease the need for repeat endoscopy.
See “Towards Better Diagnosis and Monitoring of Eosinophilic Esophagitis: Are We There Yet?” by Baker and Baker on page 410.
Eosinophilic esophagitis (EoE) is a chronic disease of the esophagus caused by an abnormal immunologic response to an allergen (1,2). The diagnosis of EoE is made in a patient with the characteristic clinical presentation with biopsy-proven esophageal eosinophilia (≥15 eos/HPF) (3). Food elimination trials are a common treatment in the pediatric EoE population and response to treatment is determined by repeat endoscopy with biopsies to confirm resolution of esophageal eosinophilia (3–5). Food trials to reintroduce eliminated foods after achieving disease remission are typically performed serially, with 1 food reintroduced at a time and with repeat endoscopy after each trial to histologically assess disease activity. A blood-based biomarker to monitor disease activity would significantly improve the quality of life for patients, reduce the risk of repeat procedures and anesthesia events, and decrease overall health care costs. A recent meta-analysis comments that there are multiple promising minimally invasive EoE biomarkers (6), but most of these biomarkers cannot distinguish between EoE and other, often co-morbid, atopic disease.
Eosinophils are produced in the bone marrow from a CD34+ lineage-committed progenitor (eosinophil progenitor [EoP]) that can mobilize from the bone marrow into the blood (7). We recently reported that EoP levels were significantly elevated in the blood of pediatric patients with active EoE and that EoP levels correlated with esophageal histologic changes seen in active EoE (8,9). Given the significant unmet need for a blood-based biomarker to assess EoE disease activity distinct from comorbid atopic disease, we aimed to determine in a second independent cohort whether blood EoP levels can be used as a biomarker to identify pediatric patients with active EoE.
In a prospective observational study, peripheral blood samples, symptom history, and laboratory data were collected from consented pediatric patients undergoing endoscopy for evaluation of EoE on dietary therapy at Cincinnati Children's Hospital Medical Center between January and October 2018. The study was approved by the Cincinnati Children's Hospital Medical Center Institutional Review Board (IRB2017-7050).
Participants were eligible if they met 2017 consensus EoE diagnostic guidelines (3) (>15 eosinophils per high power field [eos/HPF] and a failed proton-pump inhibitor [PPI] trial) and were managed with dietary therapy without the use of topical steroids. Exclusion criteria included swallowed topical corticosteroid use within the preceding month; diagnosis with an additional inflammatory gastrointestinal disease, including another form of eosinophilic gastrointestinal disease or inflammatory bowel disease; systemic steroid use within the preceding month; immune modulator or biologic agent use within the past 6 months; a history of surgery on the esophagus; severe/unstable asthma; and/or a history of a bleeding disorder. Eosinophilic gastrointestinal disease was defined as >30 eos/HPF in the stomach or duodenum. No patients were diagnosed with eosinophilic colitis.
Consent was obtained from a parent or legal guardian as all the subjects were younger than 18 years of age. Assent was also obtained from subjects between 11 and 17 years of age. A series of clinical questionnaires were administered. We calculated a rhinitis activity score for each patient on the basis of their response to the validated rhinitis control assessment test to identify individuals with active allergic rhinitis during the week before the blood draw (and endoscopy), with a score of <21 indicating active allergic rhinitis (10). We used a parent-reported outcome measuring atopic dermatitis control in the past week as follows: “controlled,” “somewhat controlled,” or “not controlled.” Institutional guidelines for elective procedures in patient with asthma limit anesthesia to those under good control-given that all patients qualified for anesthesia, asthma control was not further assessed. Antihistamine (histamine 1-receptor antagonists) and steroid (dermatologic, inhaled, and intranasal) use were recorded as a “yes” if the patient had taken the medication within the 7 days before the endoscopy. We did not report on montelukast use as only 2 patients were taking the drug. Whole blood was collected at the time of endoscopy for EoP enumeration, and a portion was sent to the clinical lab for a complete blood count with differential and total Immunoglobulin E (IgE).
Eosinophil Progenitor Level Calculation
A 5- mL blood sample was collected from each participant, and whole blood was stained for EoP level. EoPs were identified within whole blood by flow cytometry as live, CD45RA-, CD34High+, CD125+ (IL-5Rα) events (see Supplemental Digital Content 1, https://links.lww.com/MPG/B751, for detailed methods and flow cytometry gating strategy). Given prior and current experience with rare event flow cytometry in our laboratory, we established a lower limit of 500,000 gated events and 100 CD34High+ events as internal quality control measures for sample adequacy. Samples that did not meet our quality control measures were excluded from analyses. EoP events were normalized to 106 gated events. We use “EoP level” throughout the manuscript in substitution for EoPs per 106 gated events. EoP levels were calculated without prior knowledge of EoE disease activity.
Eosinophilic Esophagitis Disease Activity
Esophageal biopsies obtained during the same encounter as the blood sample were sent to the clinical pathology laboratory. At our institution it is standard to obtain at least 4 esophageal biopsies (2–3 distal and 2–3 mid to proximal) when assessing EoE disease activity, but it is more typical to get 5 to 6 in patients with known or suspected EoE. Distal and proximal esophageal peak eosinophil counts are based on clinical pathology reports, and active EoE disease was defined as peak eosinophil counts ≥15 eosinophils/HPF in esophageal biopsies. To assess EoE disease extent (stage) and severity (grade), esophageal biopsies underwent blinded assessment by a single pathologist (M.H.C.) using a validated standardized metric (EoE Histologic Scoring System [EoEHSS]) (11). In addition to histologic evaluation of disease activity, we retrospectively reviewed endoscopic video to calculate an EoE Endoscopic Reference Score (12,13), which measures severity based on 5 key endoscopic findings including edema, rings, exudate, furrows, and strictures.
Given that the data were not normally distributed, they are expressed as median and interquartile range. A Mann-Whitney test was used for pairwise comparisons and a Kruskal-Wallis test was used for multiple comparisons. To evaluate the predictive ability of EoP levels, we performed logistic regression with EoE disease activity as the dependent variable and EoP levels as the independent variable. The receiver operating characteristic (ROC) curve was used to establish the threshold that optimized both sensitivity and specificity. When evaluating biomarker performance, we used the Fisher exact test to test for association between misclassification of disease activity and patient factors. To test the correlation between the EoP level and multiple histologic features, Spearman correlation was used. For all tests, P < 0.05 was considered significant.
Fifty-six patients were screened to meet study inclusion criteria, and 50 patients were consented. Two patients were excluded due to eosinophilic enteritis found on biopsy. Of the 68 total samples collected, 34 samples were excluded due to not meeting our flow cytometry quality controls (see Methods for details). Thirty-four total samples from 31 individuals were included in the final analysis. There were no differences in multiple clinical characteristics, including sex, atopic history, EoE disease activity, and absolute eosinophil count, between the individuals from whom samples were included or excluded (see Supplemental Digital Content 2, https://links.lww.com/MPG/B751).
The study participants included pediatric patients, ages 4 to 17 years, including 18 individuals with inactive EoE and 16 with active EoE at the time of endoscopy (Table 1). Most patients were white boys (70% boys; 97% white), which is consistent with the known disease demographics (4). Patient age, sex, or atopic status was not statistically different between the active and inactive EoE disease groups (Table 1). In addition, there was no statistically significant difference between the active and inactive EoE disease groups in age of diagnosis or current medication use, including inhaled, intranasal, and topical corticosteroids, and antihistamines (Table 1). Although the absolute neutrophil count was slightly lower in the inactive EoE group (Table 1), all patients had an absolute neutrophil count within the normal reference range.
Blood Eosinophil Progenitor Levels Were Higher in Active Eosinophilic Esophagitis and Associated With Eosinophilic Esophagitis Tissue Pathology
The median blood EoP level in participants with active EoE disease was 3-fold higher than in participants with inactive disease (Fig. 1A). In contrast, peripheral absolute eosinophil counts were no different between individuals with active and inactive EoE (Fig. 1B).
In our cohort, EoE disease stage and grade EoEHSS scores were higher for biopsies from participants with active EoE than inactive EoE (Table 1), highlighting the strong association between pathologic changes in esophageal histology and active EoE disease (11). EoP level was associated with both EoEHSS combined (proximal and distal) stage and grade scores and distal stage and grade scores (see Supplemental Digital Content 3, https://links.lww.com/MPG/B751). Specifically, the EoP was associated most significantly with distal basal cell hyperplasia.
Blood Eosinophil Progenitor Levels Were Not Affected by Atopic Disease History or Activity
There was no difference in EoP levels between individuals with a history of any atopic disease, history of atopic dermatitis, history of allergic rhinitis, history of asthma, or history of IgE-mediated food allergy (see Supplemental Digital Content 4, https://links.lww.com/MPG/B751). Blood EoP levels were no different between individuals with active, inactive, or no allergic rhinitis at the time of endoscopy, regardless of their EoE disease activity (Fig. 1C,D). In addition, there was no significant difference in EoP levels between participants with controlled or uncontrolled atopic dermatitis (Fig. 1E,F). Given that there is evidence that aeroallergens have a role in EoE (14,15), we explored seasonal differences in EoP levels. We did not find any significant difference between samples from participants drawn in the spring (March–May), summer (June–August), or fall (September–October) (see Supplemental Digital Content 5, https://links.lww.com/MPG/B751).
Eosinophil Progenitor Levels Were Not Affected by Medication Use
We next investigated the effect of medications on EoP levels, including PPIs, alternative steroid preparations (dermatologic, inhaled, and intranasal), and antihistamines. We found that EoP levels did not vary between individuals with PPI, alternative steroid, or antihistamine use (see Supplementary Digital Content 6, https://links.lww.com/MPG/B751) within our full cohort of patients. In addition, between active and inactive EoE disease cohorts, medication use did not affect EoP levels (see Supplementary Digital Content 6, https://links.lww.com/MPG/B751).
Blood Eosinophil Progenitor Level Identifies Active Eosinophilic Esophagitis Disease
We next investigated the ability of the EoP level to be used as a peripheral blood biomarker to effectively rule out active disease in patients undergoing endoscopies for dietary trials. An ROC curve showed an area under the curve (AUC) of 0.81 (95% CI 0.66–0.95; P < 0.0024; Fig. 2A). A cut-off value for EoP level that is ≥17 accurately detected active disease in 79% of patients with 94.4% specificity and 62.5% sensitivity (positive predictive value 91%; negative predictive value 74%; positive likelihood ratio 11.6; negative likelihood ratio 0.4). With this EoP level threshold (Fig. 2A), there were 7 misclassified samples, including 6 false negatives (low EoP level with active EoE disease) and 1 false positive (high EoP with inactive EoE disease).
We investigated differences in age, laboratory data, medication use, and medical history between patients accurately classified and those misclassified. All 7 misclassified samples were from participants currently taking an antihistamine versus 10 out of the 27 accurately classified samples (P < 0.035). With this in mind, we developed 2 different EoP level cut-off values, one for patients taking an antihistamine and 1 for those not currently taking an antihistamine. An ROC curve for those patients currently taking an antihistamine (n = 17) has an AUC of 0.73 (95% CI 0.45–0.97; P > 0.14; Fig. 2B). A cut-off value for EoP level that is <7.4 accurately excluded active disease in 100% of patients taking an antihistamine at the time of the endoscopy, and a value ≥7.4 reliably predicted active disease in 60% of patients (Fig. 2B; sensitivity 100%; specificity 43%; positive predictive value 69%; negative predictive value 100%; positive likelihood ratio 1.75; negative likelihood ratio 0). There were 4 false positives and 0 false negatives. An ROC curve for those patients not currently taking an antihistamine (n = 17) showed an AUC of 1 (P < 0.0009; Fig. 2C). A cut-off value for EoP level that is ≥17 accurately detected active disease 100% of the time, and a value of <17 reliably predicts inactive disease in 100% of patients (Fig. 2C). Using these 2 EoP cut-off values to detect active EoE disease based on antihistamine use leads to a test with a combined 100% sensitivity and 77.8% specificity of detecting active disease (positive predictive value 80% and negative predictive value 100%).
Eosinophil Progenitor Level Identifies Participants With Endoscopic Findings Associated With active eosinophilic esophagitis
EoE disease activity is associated with specific endoscopic findings that can be scored and used to identify pediatric patients with active disease (12,13). Using an EoP level cut-off to identify individuals in our full cohort with active EoE disease, the endoscopic findings, including edema, exudates, and furrows, in the distal esophagus were associated with active EoE disease (Supplemental Digital Content 7, https://links.lww.com/MPG/B751), supporting the association of blood EoP level and active EoE disease.
In this prospective observational study with pediatric EoE participants, we demonstrate that the blood EoP level shows promise as a biomarker to identify pediatric patients with active EoE disease on dietary therapies. Notably, these results with a second independent cohort are fully consistent with our previously published study (8). In this study, we determined an EoP cut-off level to detect active EoE disease, and surprisingly, antihistamine use lowered this EoP threshold in spite of the lack of direct effect of antihistamines on the EoP levels. EoP levels have been shown to be higher in adults with atopic airway disease (16,17), highlighting that mobilization of EoPs into the bloodstream is responsive to allergic disease activity. In our pediatric cohort, we showed that active allergic rhinitis (Fig. 1D) or uncontrolled atopic dermatitis (Fig. 1F) at the time of endoscopy did not increase the EoP level into a false positive range when the participant had inactive EoE. This suggests that EoE disease activity takes precedence over other allergic disorders in EoP blood level regulation, at least in the pediatric population.
Our new observation that antihistamine use, not atopy or active allergic rhinitis as hypothesized, influenced blood EoP levels in patients with active EoE disease (with no effect seen in inactive EoE) needs further investigation. We speculate that the histamine blockade may modulate a Th2 immune response that promotes EoP mobilization into the periphery. Infiltrating eosinophils have been shown to express H1R in EoE (18) and we noted expression of H1R by flow cytometry on the surface of a subset of EoPs in a small sample of patients with EoE (see Supplementary Digital Content 8, https://links.lww.com/MPG/B751), suggesting that histamine may have a direct impact on a portion of EoPs. There was a trend for antihistamine use to lower blood EoP levels in the full cohort, especially within the active disease cohort, but the difference did not meet statistical significance, likely due to the variability in EoP histamine-responsiveness within individuals. Further work needs to be done to determine whether antihistamines block mobilization of EoPs from the bone marrow and lower blood EoP levels.
A significant number of samples were excluded on the basis of poor sample quality and sample inadequacy. We learned throughout the course of the study that it is necessary to acquire a sufficient number of gated events as the EoP is a rare event in the peripheral blood on flow cytometry. Despite these limitations, there was not a statistically significant difference in the number of samples excluded from the active and inactive disease groups or any differences in the demographic or clinical characteristics between participants that were included and excluded. Despite the small sample sizes, antihistamine use, rather than active allergic rhinitis or uncontrolled atopic dermatitis, had a larger effect on EoP levels and lowered the EoP cut-off level to detect active disease. We acknowledge that we were unable to assess the effect of asthma activity on the EoP level given that in our cohort only 5 patients had severe enough asthma to require an inhaled corticosteroid and all patients must have had controlled asthma at the time of endoscopy to undergo anesthesia. Although our results need to be validated in a larger cohort and the exact EoP cut-off value may ultimately be modified, the EoP level may potentially meet the goal of decreasing the number of repeat endoscopies for patients undergoing food trials. In addition, the specified EoP thresholds must be interpreted with caution as our sample size was limited especially when the sample was stratified by antihistamine use. Clearly, additional studies are warranted to validate these thresholds. Nonetheless, while based on small numbers, our results may enable effective use of EoPs to minimize endoscopies in individuals with EoE.
We acknowledge that we did not include a control group for this preliminary study since we are not suggesting the EoP to be used as a diagnostic tool, but rather as a biomarker reflecting esophageal disease activity with changes in therapy in previously diagnosed EoE patients. This is an important distinction, as our goal is not to replace the original diagnostic endoscopy, which serves a number of other useful purposes including ruling out other non-EoE causes of the presenting symptoms. Planned future studies will examine the EoP level in individual patients over time in a larger and broader population to address these concerns.
The nature of the patient population on dietary therapy includes some differences from the overall EoE population. Typically, patients and families who choose dietary therapy are younger, as parents have more control over the patient's diet. EoE is a heterogeneous disease, and studies demonstrate varying response rates to therapies. Some patients respond to PPI alone (5), some to swallowed steroids (19–21), others only to elemental diet (22), and a subset require multiple therapies or biologic medications (23–25) to control esophageal eosinophilia. It is also unclear which patients will develop fibrosis with esophageal stenosis regardless of treatment course. There are a growing number of promising genetic and molecular panels including the Eosinophilic Diagnostic Panel, which is a panel of dysregulated genes that reflect EoE disease severity and remission (26) that may aid in determining patient response to specific treatments. We need to consider that patients who are diagnosed at a younger age, and therefore more likely to choose dietary therapy, may have a specific EoE phenotype and endotype. Future studies will look at a broader representation of patients with EoE, including those on nondietary therapy.
Our study suggests that the EoP level in the peripheral blood has promise as a blood-based biomarker to assess EoE disease activity after a change in therapy, specifically food trials. As the disease is currently only assessed by histologic changes in esophageal biopsies, our findings, if further supported in larger cohort studies, would decrease the need for endoscopy and exposure to anesthesia in children with EoE.
1. Liacouras CA, Furuta GT, Hirano I, et al. Eosinophilic esophagitis: updated consensus recommendations for children and adults. J Allergy Clin Immunol
2. Dellon ES, Gonsalves N, Hirano I, et al. ACG clinical guideline: evidenced based approach to the diagnosis and management of esophageal eosinophilia and eosinophilic esophagitis (EoE). Am J Gastroenterol
3. Lucendo AJ, Molina-Infante J, Arias Á, et al. Guidelines on eosinophilic esophagitis: evidence-based statements and recommendations for diagnosis and management in children and adults. United European Gastroenterol J
4. Furuta GT, Liacouras CA, Collins MH, et al. Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment. Gastroenterology
5. Dellon ES, Liacouras CA, Molina-Infante J, et al. Updated international consensus diagnostic criteria for eosinophilic esophagitis: Pro∗ceedings of the AGREE Conference. Gastroenterology
6. Hines BT, Rank MA, Wright BL, et al. Minimally invasive biomarker studies in eosinophilic esophagitis: a systematic review. Ann Allergy Asthma Immunol
7. Kuo HP, Wang CH, Lin HC, et al. Interleukin-5 in growth and differentiation of blood eosinophil progenitors in asthma: effect of glucocorticoids. Br J Pharmacol
8. Morris DW, Stucke EM, Martin LJ, et al. Eosinophil progenitor levels are increased in patients with active pediatric eosinophilic esophagitis. J Allergy Clin Immunol
9. Schwartz JT, Morris DW, Collins MH, et al. Eosinophil progenitor levels correlate with tissue pathology in pediatric eosinophilic esophagitis. J Allergy Clin Immunol
10. Meltzer EO, Schatz M, Nathan R, et al. Reliability, validity, and responsiveness of the Rhinitis Control Assessment Test in patients with rhinitis. J Allergy Clin Immunol
11. Collins MH, Martin LJ, Alexander ES, et al. Newly developed and validated eosinophilic esophagitis histology scoring system and evidence that it outperforms peak eosinophil count for disease diagnosis and monitoring. Dis Esophagus
12. Wechsler JB, Bolton SM, Amsden K, et al. Eosinophilic esophagitis reference score accurately identifies disease activity and treatment effects in children. Clin Gastroenterol Hepatol
13. Bolton SM, Kagalwalla AF, Wechsler JB. Eosinophilic esophagitis in children: endoscopic findings at diagnosis and post-intervention. Curr Gastroenterol Rep
14. Ram G, Lee J, Ott M, et al. Seasonal exacerbation of esophageal eosinophilia in children with eosinophilic esophagitis and allergic rhinitis. Ann Allergy Asthma Immunol
15. Reed CC, Iglesia EGA, Commins SP, et al. Seasonal exacerbation of eosinophilic esophagitis histologic activity in adults and children implicates role of aeroallergens. Ann Allergy Asthma Immunol
16. Sehmi R, Howie K, Sutherland DR, et al. Increased levels of CD34+ hemopoietic progenitor cells in atopic subjects. Am J Respir Cell Mol Biol
17. Makowska JS, Grzegorczyk J, Ciéslak M, et al. Recruitment of CD34+ progenitor cells into peripheral blood and asthma severity. Ann Allergy Asthma Immunol
18. Merves J, Chandramouleeswaran PM, Benitez AJ, et al. Altered esophageal histamine receptor expression in eosinophilic esophagitis (EoE): implications on disease pathogenesis. PLoS One
19. Butz BK, Wen T, Gleich GJ, et al. Efficacy, dose reduction, and resistance to high-dose fluticasone in patients with eosinophilic esophagitis. Gastroenterology
20. Gupta SK, Vitanza JM, Collins MH. Efficacy and safety of oral budesonide suspension in pediatric patients with eosinophilic esophagitis. Clin Gastroenterol Hepatol
21. Noel RJ, Putnam PE, Collins MH, et al. Clinical and immunopathologic effects of swallowed fluticasone for eosinophilic esophagitis. Clin Gastroenterol Hepatol
22. Arias A, González-Cervera J, Tenias JM, et al. Efficacy of dietary interventions for inducing histologic remission in patients with eosinophilic esophagitis: a systematic review and meta-analysis. Gastroenterology
23. Poulakos MN, Cargill SM, Waineo MF, et al. Mepolizumab for the treatment of severe eosinophilic asthma. Am J Health Syst Pharm
24. Assa’ad AH, Gupta SK, Collins MH, et al. An antibody against IL-5 reduces numbers of esophageal intraepithelial eosinophils in children with eosinophilic esophagitis. Gastroenterology
25. Spergel JM, Rothenberg ME, Collins MH, et al. Reslizumab in children and adolescents with eosinophilic esophagitis: results of a double-blind, randomized, placebo-controlled trial. J Allergy Clin Immunol
26. Wen T, Stucke EM, Grotjan TM, et al. Molecular diagnosis of eosinophilic esophagitis by gene expression profiling. Gastroenterology