ASSOCIATION OF DUODENAL EOSINOPHILIC INFILTRATE WITH HELICOBACTER PYLORI INFECTION, BUT NOT WITH FUNCTIONAL DYSPEPSIA

Associação de infiltrado eosinofílico duodenal com infecção por Helicobacter pylori, mas não com dispepsia funcional

ABSTRACT

BACKGROUND:

The role of Helicobacter pylori infection on eosinophilic infiltration in duodenal mucosa is poorly studied. An increase in the number of eosinophils in duodenum has been associated with functional dyspepsia.

OBJECTIVE:

To evaluate the influence of H. pylori infection on duodenal eosinophil count and the role of eosinophilic infiltrate of duodenum in functional dyspepsia.

METHODS:

Positive and negative H. pylori individuals were included. Both functional dyspeptic patients according to Rome III criteria (cases) and individuals without gastrointestinal symptoms (controls) were enrolled. They were submitted to upper endoscopy and H. pylori infection was verified by gastric histopathology and urease test. Eosinophils in the duodenal mucosa were counted in five high-power fields, randomly selected on slides of endoscopic biopsies.

RESULTS:

Thirty-nine H. pylori positive (mean age 40.5 and 69.2% women) and 24 negative patients (mean age 37.3 and 75% women) were included. The influence of the infection was observed in the duodenal eosinophil count, which was higher in infected individuals: median 13.2 vs 8.1 in non-infected individuals (P=0.005). When we analyzed patients according to symptoms, cases - mean age 39.6; 71.4% women - and controls - mean age 38.7; 71.4% women - had similar duodenal eosinophil count: median 11.9 and 12.6 respectively (P=0.19).

CONCLUSIONS:

We did not demonstrate association of duodenal eosinophil count with functional dyspepsia but found association with H. pylori infection.

HEADINGS:
Dyspepsia; Biopsy; Helicobacter pylori; Duodenum; Eosinophils

RESUMO

CONTEXTO:

O papel de infecção por Helicobacter pylori no infiltrado eosinofílico duodenal ainda é pouco compreendido. Um aumento no número de eosinófilos duodenais tem sido associado a dispepsia funcional.

OBJETIVO:

Avaliar a influência do H. pylori na contagem de eosinófilos duodenais e o papel do infiltrado eosinofílico duodenal na dispepsia funcional.

MÉTODOS:

Indivíduos H. pylori positivo e negativo foram incluídos. Ambos os grupos, compreendendo dispépticos funcionais pelos critérios de Roma III (casos) e indivíduos sem sintomas gastrointestinais (controles), foram submetidos à endoscopia digestiva alta para pesquisa de H. pylori, efetuada por histopatologia e teste de urease. Eosinófilos na mucosa duodenal foram contabilizados em cinco campos de maior aumento, selecionados randomicamente nas lâminas de biópsia endoscópicas.

RESULTADOS:

Trinta e nove indivíduos H. pylori positivo (média de idade 40,5 e 69,2% mulheres) e 24 H. pylori negativos (média de idade 37,3 e 75% mulheres) foram incluídos. A influência da infecção por H. pylori foi observada na contagem de eosinófilos, que foi maior nos positivos: mediana 13,2 vs 8,1 (P=0,005). Quando analisados pacientes de acordo com sintomas, os casos (média de idade 39,6 e 71,4% mulheres) e controles (média de idade 38,7 e 71,4% mulheres), apresentaram semelhante contagem de eosinófilos duodenais: mediana 11,9 e 12,6, respectivamente (P=0,19).

CONCLUSÃO:

Não demonstramos associação da contagem de eosinófilos duodenais com dispepsia duodenal, mas encontramos associação com infecção por H. pylori.

DESCRITORES:
Dispepsia; Biópsia; Helicobacter pylori; Duodeno; Eosinófilos

INTRODUCTION

Dyspeptic symptoms affect around a third¹ of the world’s population. Most of these individuals suffer from functional dyspepsia (FD), defined by symptoms arising from the gastroduodenal region in the absence of any explanatory organic disease².

The pathophysiology of FD is still poorly understood. Hypothesis of its etiology include gastroduodenal motility disorders³, gastric hypersensitivity to distension and acids^4,5, microbiological triggers such as gastroenteritis, H. pylori infection^6,7 and psychosocial distress^8,9.

Eosinophils participate in the interaction between innate and acquired immunity and are important in the inflammatory response of type 2 hypersensitivity reactions. Infections cause their recruitment and degranulation, including H. pylori¹⁰. They may cause symptoms due to release of cytokines, which lead to neural excitation, muscle spasms and pain¹¹. Recent studies have reported a greater number of eosinophils in the duodenal mucosa of patients with functional dyspepsia than in asymptomatic controls, but the issue is controversial^12-14.

H. pylori infection is very common in south Brazil, affecting more than 2 thirds of functional dyspeptic patients⁶. The association of this bacteria and duodenitis has been reported¹⁵. However, there is uncertainty about the role of H. pylori in number of duodenal eosinophils. It was not reported by Zhao in a study that included patients with FD¹⁶. The infection was not related to eosinophilic infiltration in previous studies^12,14,17, with infection rates as low as 16.3%.

So, the findings of previous studies might not be extrapolated to Brazilian patients. This study aims to evaluate further the influence of H. pylori infection on eosinophil count and the role of duodenal eosinophilic infiltrate in FD in our population.

METHODS

The functional dyspeptic patients (cases) in this investigation participated in the randomized double-blind study Heroes Trial (Helicobacter Eradication Relief of Dyspeptic Symptoms) - ClinicalTrials.gov number NCT00404534. In this paper⁶, it was evaluated the effect of H. pylori eradication on the symptoms of functional dyspepsia. The study has been conducted in our institution, Hospital de Clínicas de Porto Alegre (HCPA). Individuals aged 18 or older who fulfilled the Rome III criteria for functional dyspepsia were included. Exclusion criteria are better described elsewhere⁶.

The control group consisted of individuals with no symptoms in the gastrointestinal tract selected among donors for the blood bank at the same hospital. They were subjected to the same exclusion criteria as the cases.

In this study we aim to evaluate the eosinophil infiltrate in duodenal mucosa of functional dyspeptic and asymptomatic individuals. This research was approved by the HCPA Internal Review Board Committee and informed consent was obtained from all cases and controls prior to inclusion. The study is in accordance with Resolution 466/2012 of the National Health Council of the Ministry of Health (Brazil).

Study procedures

Cases and controls were submitted to a medical consultation, during which the demographic data collection and the physical examination, including anthropometric measurements, were performed.

Symptoms were better assessed by the PADYQ (Porto Alegre Dyspeptic Symptoms Questionnaire)¹⁸. It comprises 11 questions about the frequency, duration and intensity of dyspeptic symptoms over the last 30 days, with scores ranging from 0 (no symptom) to 44 (severe symptoms).

Cases patients were classified into two subgroups according to the predominant symptom established by the Rome III criteria: epigastric pain syndrome (EPS) and postprandial distress syndrome (PDS). Both cases and controls included were submitted to laboratory tests and upper endoscopy with gastric (body, antrum and incisura) and duodenal mucosal biopsies (second portion). Endoscopies were performed by two trained endoscopists. In case of disagreement, a third endoscopist was consulted. Those individuals whose findings were not only gastritis, duodenitis or hiatal hernia were excluded.

Histopathology

Fragments from gastric endoscopic biopsies of the incisura, antrum and body were analyzed for generic histological findings (H&E) and H. pylori presence (Giemsa staining). Rapid urease test to verify this bacterium was performed immediately after endoscopy, using one fragment from each portion. Patients were considered H. pylori carriers when it was confirmed by both methods. In case of disagreement a third pathologist was consulted.

Biopsies of duodenum were stained with H&E and evaluated by a pathologist DMU blind to the patient’s group (case or control). Samples were analyzed for villus architecture, intraepithelial lymphocyte count, chronic inflammation signs, metaplasia, presence of pathogens and erosions. Five high-power fields (40X) were randomly selected on each fragment, eosinophils were counted in each field, then sum and mean was obtained. A second and independent pathologist JCA evaluated all these biopsies by the same method, in order to verify the histopathological reliability.

Sample size

Twenty-three patients in each group were needed to detect a difference of one standard-deviation of eosinophil count between H. pylori positive and negative patients, with a power of 90% and alpha 0.05%¹².

Statistical analysis

SPSS 21.0 software for Windows (IBM co, NY) was used, applying chi-square tests with and without Yates correction for categorical variables and Student’s t-test for quantitative variables with normal distribution. The distribution curve of the mean eosinophil count/high-power field of individuals was found to be asymmetrical, and the median and interquartile ranges were obtained.

Mann-Whitney test was used to assess the primary outcome, namely the association of duodenal intramucosal eosinophil count with dyspepsia and H. pylori. The intraclass correlation coefficient (ICC) was calculated for comparing values obtained by the independent pathologists. The level of significance was set at 5%.

RESULTS

Thirty-nine H. pylori positive and 24 negative patients were included, with 66% of cases in each group, comprising 42 cases and 21 controls. Basal characteristics were similar between groups (Table 1). The median of score PADYQ was 19.5 for cases and 0 for controls (P<0.001).

TABLE 1
Sample characterization.

Variables	H. pylori positive	H. pylori negative	P value
	group	group
	N=39	N=24
Age (years) - mean ± SD	40.5±10.5	37.3±10.5	0.25
Sex - n (%)
Male	12 (30.8)	6 (25)	0.84
Female	27 (69.2)	18 (75)
Cases - n (%)
Positive	26 (66.6)	16 (66.6)	1
Race - n (%)
White	34 (87.2)	21 (87.5)	0.98
Black	3 (7.7)	2 (8.3)
Mixed	2 (5.1)	1 (4.2)
BMI (kg/m²) - mean ± SD	25.9±4.1	24.6±5.3	0.54

• SD: standard deviation; H. pylori: Helicobacter pylori; BMI: body mass index.

Upper endoscopy was normal in 31 (79.5%) of H. pylori positive individuals. Abnormal findings were duodenitis (three enanthematous and five erosive). Among negative, endoscopy was normal in 21 (87.5%). Alterations were three with enanthematous duodenitis. Values were statistically similar (P=0.16).

Among positive H. pylori individuals, the duodenum was histologically normal in 31 of 39 (79.5%) and among negative was normal in 17 of 24 (70.8%). These values were also statistically similar (P=0.3). Regarding the alterations found among positive, seven had duodenitis, one had duodenal lymphangiectasia. Among negative five had duodenitis and two gastric metaplasia. Gastric biopsies from antrum, body and incisura were also obtained and set in different jars. All specimens were normal (no atrophy or inflammatory activity) in 20.8% of noninfected patients, and in none of infected patients (P<0.001).

Cases and controls had similar duodenal endoscopic findings, with 81% exhibiting normal duodenum in both groups. Among functional dyspeptic patients, the duodenum was histologically normal in 31 of 42 (73.8%) cases and normal in 18 of 21 (85.7%) controls, P=0.64. As for the alterations found among cases, eight had duodenitis, one had duodenal lymphangiectasia, and two showed gastric metaplasia, whereas three controls exhibited duodenitis, P=0.63.

Duodenal eosinophil count

Biopsies were analyzed by a second pathologist. The eosinophil counts obtained from the independent observers were compared using the intraclass correlation coefficient. There was good agreement: the calculated ICC was 0.79 (95% IC 0.62-0.88, P<0.001).

The influence of H. pylori was observed on the median of eosinophil counts. This value was 13.2 for the 39 infected individuals, whereas it was 8.1 among the 24 non-infected. The difference was statistically significant (P=0.005), Figure 1.

FIGURE 1
Eosinophil count in HP infected X non-infected patients. The band inside the box represents the median. The bottom and top of the box represent 25th and 75th percentiles of the sample [interquartile range or IQR]. The whiskers are the minimum and maximum values in a range of 1.5[IQR]. Dot and asterisks represent the outliers that are above this range.

The best cut-off for eosinophil count that correlated to H. pylori infection was 11.1 (Figure 2). Among individuals infected with H. pylori, 71.8% had values above this limit, and among noninfected, only 25% (P<0.001); refer to Figure 3.

FIGURE 2
ROC curve representing accuracy of eosinophil count to detect H. pylori infection. The AUC was 0.71 (CI 0.56-0.86), P=0.005.

FIGURE 3
Frequency of patients according to eosinophil count fashion. The histogram shows that the majority of H. pylori positive and the minority of H. pylori negative patients had eosinophil count higher than the cut-off 11.1.

The median of eosinophil count per high-power field was similar between functional dyspeptic patients and the asymptomatic controls. The median was 11.9 and 12.6 respectively (P=0.194). They were also similar in the analysis of the FD subgroups containing 26 EPS patients and 16 PDS patients (Table 2). There was no relationship between serum eosinophil value and duodenal count (P=0.08).

TABLE 2
Eosinophil count/HPF*- dyspeptic patients x controls.

Group	Eosinophil count/HPF, median	P value vs controls**
	(interquartile range)
Cases	11.9 (6.7-14.7)	0.194
EPS	11.5 (6.7-14.6)	0.181
PDS	12.2 (6.5-16.8)	0.421
Controls	12.6 (8.8-22.6)	-

• * HPF high-power field. ** Mann-Whitney test.

DISCUSSION

Our study demonstrated association of duodenal eosinophil count with Helicobacter pylori infection, but not with FD. The strength of our findings is that the included cases were carefully selected for participation in Heroes, a randomized clinical trial. Our controls were asymptomatic regarding digestive tract. None of the individuals (dyspeptic or asymptomatic) showed any gastroduodenal alteration in the endoscopy other than gastritis or duodenitis. In addition, the baseline characteristics of groups were similar.

The association of H. pylori and duodenal eosinophilic infiltrate both in dyspeptic and controls has not been previously described. Helicobacter infection has been shown to cause inflammatory infiltrates in the stomach, leading to recruitment and degranulation of eosinophils^10,19. Consequently, there is release of toxic molecules. They can generate gastrointestinal symptoms, as we mentioned before. This recruitment and degranulation cascade might also occur in duodenum.

So, our finding is biologically supported. Most studies neglected this hypothesis, probably because of prevalence of H. pylori positive individuals being as low as 10%^12,14,17,20. A larger sample investigation found no relationship between infection and duodenal eosinophils, but did not include asymptomatic individuals as we did¹⁶.

The hypothesis of a relationship between eosinophilic infiltrate and dyspepsia was initially investigated by Toukan et al.²¹, who analyzed gastroduodenal endoscopies and biopsies in dyspeptic and control patients. Later, Talley and Walker¹² confirmed this finding in a Swedish community. The same group conducted another case-control study including EPS and PDS patients in London’s population, but found only an association of duodenal eosinophilia and symptoms in individuals in the second subgroup²². They also conducted a study in Australia, where they observed no association between functional dyspepsia and eosinophilic infiltrate, except for cases of postprandial fullness¹⁷. A study by Bafutto et al. with 36 dyspeptic patients and nine controls in Brazil showed a similar result²³.

As in the present study, other authors reported negative results for association between eosinophilic infiltrate and dyspepsia, such as the researches carried out by Veerapan et al. (Washington)²⁴, Binesh et al. among Iranians¹⁴ and Lijun Du in China²⁵. A recently published study by Song et al. found a positive association, but the status of the H. pylori infection was not evaluated, which, as we reported, may influence the results¹³.

Limitations of our study include we did not investigate the presence of parasitic infection or allergic factors in the individuals included. However, we believe this was no longer relevant as negative results were found, and both cases and controls are from the same population and exposed to the same environmental factors. We obtained a lower eosinophil count than the developed countries. This may show less than expected influence of parasitical infection.

The lack of an association between eosinophils and dyspeptic symptoms in some studies does not disagree with previous positive findings. The populations are different, with local genetic and microbiological influences and allergen exposure, factors known to influence the recruitment and accommodation of eosinophils in the mucosa of the gastrointestinal tract²⁶.

CONCLUSION

We suggest that results from previous studies have little validity to our population. However, functional dyspepsia is a multifactorial disease whose etiology should be tailored to each person. This research contributes with a new finding, influence of H. pylori on eosinophil count. How this can contribute to clinical management is an issue that should drive further studies.

ACKNOWLEDGEMENTS

The authors are very grateful to the contributors of Heroes Trial and to the staff of Experimental Pathology of HCPA.

REFERENCES

• ¹
  Mahadeva S, Goh K. Epidemiology of functional dyspepsia: A global perspective. World J Gastroenterol. 2006;12:2661-6.
• ²
  Tack J, Talley N, Camilleri M, Holtmann G, Hu P, Malagelada J, et al. Functional Gastroduodenal Disorders. Gastroenterology. 2006;130:1466-79.
• ³
  Stanghellini V, Ghidini C, Maccarini MR, Paparo GF, Corinaldesi R, Barbara L. Fasting and postprandial gastrointestinal motility in ulcer and non-ulcer dyspepsia. Gut. 1992;33:184-90.
• ⁴
  Lémann M, Dederding JP, Flourié B, Franchisseur C, Rambaud JC, Jian R. Abnormal perception of visceral pain in response to gastric distension in chronic idiopathic dyspepsia. The irritable stomach syndrome. Dig Dis Sci. 1991;36:1249-54.
• ⁵
  Samsom M, Verhagen M, vanBerge Henegouwen G, Smout A. Abnormal clearance of exogenous acid and increased acid sensitivity of the proximal duodenum in dyspeptic patients. Gastroenterology. 1999;116:515-20.
• ⁶
  Mazzoleni L, Sander GB, Francesconi CF, Mazzoleni F, Uchoa DM, De Bona LR, et al. Helicobacter pylori eradication in functional dyspepsia: HEROES trial. Arch Intern Med. 2011;171:1929-36.
• ⁷
  Thumshirn M. Pathophysiology of Functional Dyspepsia. Gut. 2002;51:63-6.
• ⁸
  Talley N, Boyce P, Jones M. Dyspepsia and health care seeking in a community: How important are psychological factors? Dig Dis Sci . 1998;43:1016-22.
• ⁹
  Filipović B, Randjelovic T, Ille T, Markovic O, Milovanović B, Kovacevic N, et al. Anxiety, personality traits and quality of life in functional dyspepsia-suffering patients. Eur J Intern Med. 2013;24:83-6.
• ¹⁰
  McGovern TW, Talley NJ, Kephart GM, Carpenter HA, Gleich GJ. Eosinophil infiltration and degranulation in Helicobacter pylori-associated chronic gastritis. Dig Dis Sci . 1991;36:435-40.
• ¹¹
  Walker MM, Warwick A, Ung C, Talley NJ. The role of eosinophils and mast cells in intestinal functional disease. Curr Gastroenterol Rep. 2011;13:323-30.
• ¹²
  Talley NJ WM, Aro P, Ronkainen J, Storskrubb T, Hindley LA, Harmsen WS, et al. Non-ulcer dyspepsia and duodenal eosinophilia: an adult endoscopic population-based case-control study. Clin Gastroenterol Hepatol. 2007;5:1175-83.
• ¹³
  Song S, Song Y, Zhang H, Li G, Li X, Wang X, et al. Increased counts and degranulation of duodenal mast cells and eosinophils in functional dyspepsia- a clinical study. Med Glas (Zenica). 2014;11:276-82.
• ¹⁴
  Binesh F, Akhondei M, Pourmirafzali H, Rajabzadeh Y. Determination of relative frequency of eosinophils and mast cells in gastric and duodenal mucosal biopsies in adults with non-ulcer dyspepsia. J Coll Physicians Surg Pak. 2013;23:326-9.
• ¹⁵
  Mirbagheri SA, Khajavirad N, Rakhshani N, Ostovaneh MR, Hoseini SM, Hoseini V. Impact of Helicobacter pylori infection and microscopic duodenal histopathological changes on clinical symptoms of patients with functional dyspepsia. Dig Dis Sci . 2012;57:967-72.
• ¹⁶
  Zhao W, Zhong X, Zhuang X, Ji H, Li X, Li A, et al. Evaluation of Helicobacter pylori eradication and drug therapy in patients with functional dyspepsia. Exp Ther Med. 2013;6:37-44.
• ¹⁷
  Walker MM, Aggarwal KR, Shim LS, Bassan M, Kalantar JS, Weltman MD, et al. Duodenal eosinophilia and early satiety in functional dyspepsia: confirmation of a positive association in an Australian cohort. J Gastroenterol Hepatol. 2014;29:474-9.
• ¹⁸
  Sander GB, Mazzoleni LE, Francesconi CF, Wortmann AC, Ott EA, Theil A, et al. Development and validation of a cross-cultural questionnaire to evaluate nonulcer dyspepsia: the Porto Alegre Dyspeptic Symptoms Questionnaire (PADYQ). Dig Dis Sci . 2004;49:1822-9.
• ¹⁹
  Ko SH, Jeon JI, Kim YJ, Yoon HJ, Kim H, Kim N, et al. Helicobacter pylori Outer Membrane Vesicle Proteins Induce Human Eosinophil Degranulation via a β2 Integrin CD11/CD18- and ICAM-1-Dependent Mechanism. Mediators Inflamm. 2015;2015:301716.
• ²⁰
  Taki M, Oshima T, Li M, Sei H, Tozawa K, Tomita T, et al. Duodenal low-grade inflammation and expression of tight junction proteins in functional dyspepsia. Neurogastroenterol Motil. 2019;31:e13576.
• ²¹
  Toukan AU KM, Amr SS, Arnaout MA, Abu-Romiyeh AS. Gastroduodenal inflammation in patients with non-ulcer dyspepsia. A controlled endoscopic and morphometric study. Dig Dis Sci . 1985;30:313-20.
• ²²
  Walker MM, Salehian SS, Murray CE, Rajendran A, Hoare JM, Negus R, et al. Implications of eosinophilia in the normal duodenal biopsy - an association with allergy and functional dyspepsia. Aliment Pharmacol Ther. 2010;31:1229-36.
• ²³
  Bafutto M OE, Bafutto AA, Bafutto EHF, Bomfim IC, Rezende Filho J. Duodenal Eosinophilia: A Link Between Functional Dyspepsia and Post-Infective Functional Dyspepsia? Gastroenterology. 2012;142(Suppl 1):S-171.
• ²⁴
  Veerappan G, Moawad F, Duncan TJ, Maydonovitch C, Baker TP, Perry JL, et al. Non-Ulcer Dyspepsia Not Associated with Gastric or Duodenal Eosinophilia. Gastroenterology. 2009;136:58-9.
• ²⁵
  Du L, Shen J, Kim JJ, Yu Y, Ma L, Dai N. Increased Duodenal Eosinophil Degranulation in Patients with Functional Dyspepsia: A Prospective Study. Sci Rep. 2016;6:34305.
• ²⁶
  Prussin C. Eosinophilic gastroenteritis and related eosinophilic disorders. Gastroenterol Clin North Am. 2014;43:317-27.

• Disclosure of funding: no funding received

Publication Dates

• Publication in this collection
  27 Feb 2020
• Date of issue
  Jan-Mar 2020

History

• Received
  23 Nov 2019
• Accepted
  06 Jan 2020