Antral brushings and biopsies for gastric Campylobacter pylori : A comparative study of a rapid urease test , culture and histology

A prospective study was conducted to assess the sensitivity and 
specificity of diagnostic tests for Campylobacter pylori in 60 unselected patients 
referred for gastroscopy. Urease testing was performed on samples of gastric mucus 
obtained by a disposable cytology brush and on gastric mucosal biopsies. The 
combination of the brush urease test and biopsy urease test had a sensitivity of 
78% and a specificity of 96%. The combination of the brush urease test and examination 
of gastric biopsies for curved bacilli had a sensitivity of 96% and a specificity 
of 96%. The combination of the brush urease and biopsy urease tests will 
rapidly identify the majority of patients with C pylori. This test is rapid, easy, 
sensitive and inexpensive. If a pathologist is available, the addition of histologic 
review for C pylori in combination with the brush urease test, will identify virtually 
all patients with C pylori without the need for culturing this fastidious organism.

I N 1982, WARR EN AND MARSHALL (1 -3)   successfully cultured the Gram-negative bacterium Campylobacter pylori from human gastric mucosa and fi rst established an association between this organism and gastroduodenal disease.Since then, a number of investigators have confirmed the association of C pylori with gastritis, gastric ulce rs and duodenal ulcers (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15).Recent studies have shown that standard treatment for peptic ulcer disease does not eradicate C pylori from gastric mucosa (16)(17)(18).Whether more specific therapy for C /Jylori will change the long term prognosis of peptic ulcer disease is presently under investigation.
Multiple modalities have been used to d etect the presence of C pylori.A n inoculum of thi s fastidiou s organism from gastric mucosa!biopsies wi ll yield visible colonies on agar plates in five to seven days ( 19).Curved bacilli can be seen on gastric mucosa stained with hernatoxylin and eosin , Warth in-Starry sil-

C pylori
The purpose of thi s stud y was to assess the use of urea brorh tests as a simple inexpensive technique available to all endoscopy laboratories ro detect the presence of C J,ylori.Gastric bio psies as well as a disposable cytology brush were used to obtain samples of gastric mucus for culture and ureasc testing in Stuart's modified urea broth.These tests were compared to p reviously established methods for the detection of C pylori.Bacteria recovered at e nd oscopy we re identified as C pylori based o n colony morphology, Gram strain, and the production of urease, catalase and oxidase.Endoscopy: One investigator performed all upper gastrointestinal endoscopies using either the Olympus XQIO or O lympus Q endoscope.Each quadrant of the gastric antrum was randomly brushed with a si ngle di sposable cytology brush until the mucosa became e rythematous.The brush was then swished in a tube containing l ml of urease test broth.Fou r random antral biopsies were obtained.Two were placed in a tube containing I ml of urease test broth and two were placed in buffered formalin and sent for histologic examination.Endoscopic fi ndings were recorded prior to checking the urease rest broth results.Urease test broth: The Stuart's urease test broth (23) was modified by the addition o f 200 g glucose to each litre ( l9).
Urease test broth containing the antral brushings and the tubes containing the antral biopsies we re each observed at room temperatu re for up to 6 h.A minimum of 1000 organisms will ca use a change in colo ur from yellow to pink, defined as a posi tive urease tesr.Incubation at 3 7°C o r 42°C di<l not hasten the reaction.Test resul ts were recorded by a microbiologist who was unawa re o( the endoscopic findings.Microbiology: All sa mples were kept at room temperature fo r 2 to 6 h before culture and the presence o r absence of colour change was recorded .Gastric antral biopsies, first gro und with a sterile mortar and pestle in 0 .4ml o f brainheart infusion broth (BBL, Cockeysvillc, MD) and antral brush ings were inoculated into Columbia blood agar plates (Difeo Labs, Detroit, Ml).The plates were then incubated in 5% oxygen and 10% ca rbon dioxid e at 3 7°C for up to IO days.Colon ies were identified as C 1 ,ylori on the basis of Gram stain and chemical testing for the presence of ureasc (23).catalase and oxidase (Difeo oxidase differentiation disk).The microbiologist recorded the results without knowl edge of the endoscopic o r histologic findi ngs.Histologic studies: Two anrral biopsies were routinely processed a nd stained with hematoxylin and eosin.Hematoxylin and eosi n have been shown to be as effective as the G iemsa stain for detecting C pylori (32).One investigator reviewed each biopsy for the presence of C pylori and gastritis.The d egree of gastritis was graded as mild , moderate or seve re by the following criteria: amo unt of inflammatory infiltrate; relative number of acute inflammatory cells; presence or absence of u Ice ration or other changes involving the epithelial lining.The degree of polymorphonuclear infiltration was characterized as being abse nt, rare, mild, moderate o r seve re .The pathologist recorded the results without kno wledge of the endoscopic, urease test broth or microbiologic studies.Statistical analysis: Sensitivity, specificity, positive predictive value and negative predictive value were calculated for each of the various diagnostic tests.The association between histologic gastritis, polymorphonuclear cell infiltrate and the presence of C /1ylori was compared using x 2 with the Yacc continuity correctio n.

Diagnosis:
The presence of C pylori was established in 32 of 60 patients (53%) as defined by a positive culture from either brush or biopsy specimens or identification of organisms on histologic examination.
The results of the diagnostic tests arc listed in Table L.The brush urease and biopsy urease tests have a diagnostic sensitivity of 62 and 68%, respectively.
Cpylori histology has a sensitivityof72% and culture has a sensitivity of 75%.The combination brush urease and biopsy urease test has a sensitivity of 78%.By using a combination of C pylori histology with either the biopsy urease or the brush urease test, diagnostic sensitivity increases to 90 and 96%, respectively.The combination of brush urease, biopsy urease and histologic review will not increase the overall detection of C pylori but will increase the number of patients detected by the rapid urease test.Histology: All patients who were positive for C pylori had some degree of histologic gastritis.There was a significant association between moderate and severe gastritis as well as the presence of polymorphonuclear cell infiltration and the diagnosis of C pylor1.

DISCUSSION
The findings of the present study confirm the previously reported relationship of C pylori to peptic ulcers.Six of the nine patients with duodenal ulcers and six of the eight patients with gastric ulcers were positive for the organism.These results compare with previous studies which have reported the presence of C pylori in 75 to 100% of patients with duodenal ulcers a nd 35 to 8 1% of patients with gastric ulcers (6)(7)(8)(9)(10)(11)(12).C pylori was present in 87% of patients with severe gastritis and 75% of patients with moderate gastritis.
The purpose of this study was to assess the use of Stuart's modified urea broth as a medium for detecting the presence of C pylori from gastric biopsies and cytologic brushings.This is the first study to report the use of the 'brush urease test' to aid in Diagnosis of C pylori gastritis the diagnosis of C pylori.The use of a cytology brush increases the sampling area and can increase the diagnostic yield.The authors selected the modified Stuart's test broth (2]) because it is inexpensive and available in a ny microbiology laboratory.This broth will detect as few as 1000 organisms.
ln the authors' experience, careful histologic examination of mucosa I biopsies, stained with hematoxylin and eosin under high magnification ( 100 x ), will identify C pylori as effectively as the Giemsa stain.This test alone yields a diagnostic sensitivity of 72%.
The combination of the brush urease test and the biopsy urease test has a diagnostic sensitivity of 78°'~ and specificity of 96%.The combination urease test is an easy, rapid and specific test that will detect the majority of patients infected with C pylori within hours of completion of gastroscopy.
The au thors found that the most sensitive test was the combination of the brush urease test and histologic review of the mucosa!biopsies for the presence of curved bacilli.T he advantage of this combination is chat it seems to detect almost a ll patients with C /)ylori associated gastroduodenal disease.The brush urease test will diagnose the majority of the patients within hours of endoscopy while the remaining patients will be diagnosed at the time of histologic review.The addition of the biopsy urease test to the combination of brush urease and C pylori histology will increase the number of patients diagnosed within hou rs of gastroscopy, but will not increase the overall diagnostic sensitivity.
In summary, this study confirms the association of C /)ylori with significant gastroduodenal disease.Recently a number of reviews have highlighted the importance of C pylori in the pathogenesis of gastroduodenal disease.Although optimal therapy for C pylorr gastric disease is not yet available, the presence of C pylori may a lter standard medical management.Therefore, a simple sensitive and specific test for the presence of C /)ylori would be a useful addition to the endoscopic evaluation.These data support the use of the combination brush test and biopsy urease test to identify the majority of patients with C pylori associ-EsTRlNeral ated gastroduodenal disease without the necessity for histologic or microbiologic assessment.This test is rapid, easy, sensitive, inexpe nsive and readily available for use in all gastrointestinal labo ratories.When available, histologic identification of C /rylori in antral biopsies, in combination with the brush urcase test , will identify virtually all patie nts with C pylori without the need fo r culturing this fastidious organism.
PATIENTS AND METHODSA prospective study was cond ucted in which each investiga tor was blinded to the results of the other diagnostic tests.The stud y was approved by the Cleveland Veterans Administration Hospital Investigation Review Board .All patients gave written informed conse nt.Patient population: The study group co nsisted of 60 hospitalized a nd outpatient adult males referred for upper 92 gastrointesti nal endoscopy.The patient's ages ranged from 26 to 77 years with a mean age of 61 years.Patients were excluded for: active gastrointestinal bleeding; therapeutic endoscopy (sclerotherapy, electrocautery, dilatation); prothrombin time more than 15 s; platelets less than 75,000/mm 3 • The presence of C J,ylori was established by either a positive culture for C pylori from the biopsy or gastric brushings, or identification of C pylori by microscopic examinatio n .