Abstract

AIM To develop arapid and sensitive method for diagnosing Hp infection in human gastric mucosa bydetecting mucosal NH₄ ⁺ with Nessler′s reagent.
METHODS The ammonia contents of gastric mucosa biopsy specimens with orwithout Hp infection were detected by standard Nessler′s method. The result was used as reference in developingthe “RapidUrease-Nessler′s test” (RUN test) method for diagnosingHp infection. A total of 485 patients undergoing gastroendoscopy were diagnosed by RUNtest and standard methods of culture, smear staining and ELISA as control.
RESULTS ① There was a marked difference of ammonia contents between Hppositive and negative biopsy specimens. ② The “RUN test” possessed high sensitivity, specificity, and positive and negativepredictive values (99.7%, 91.5%, 97.3% and 99.1% respectively), and the total detectingtime was only 5 minutes.
CONCLUSIONS The detection principle and properties of RUN test were greatlydifferent from the traditional urease test. It was a rapid and reliable method for Hpinfection diagnosis based on biopsy specimens.

INTRODUCTION

Helicobacter pylori (Hp) possesses an unusualcha-racteristic of rapid urea hydrolysis, which can hydrolyze the urea in gastric juice toform “ammonia cloud” around the organisms. It hasbeen reported that the ammonia concentration in gastric juice of the Hp-infected patientsis higher than that of the Hp negatives^[1] . Butwhether this difference in gastric mucosa is similar to that in gastric juice, and whetherthe difference of the ammonia contents in gastric mucosa between Hp positives andnegatives could be used as indicators for Hp infection still remain unclear. The aim ofthis study was to evaluate the ammonia contents in gastric mucosal biopsy specimens fromboth Hp positive and negative patients by the standard Nessler′s method, a method for detecting blood ammonia in clinicalche-mistry^[2] . The results were used as the basis of deve-loping aconvenient and rapid method for diagnosing Hp infection by qualitatively detecting ammoniacontents in gastric mucosa biopsy specimens with Nessler′s reagent in microtiter wells, a “Rapid Urease-Nessler′s test” (RUN test). To evaluate the usefulness of this method, we alsocompared its test results with other standard techniques of bacterium culture, smearstaining, histological examination and ELISA for diagnosing Hp infection in biopsyspecimens.

MATERIALS AND METHODS
Collection of biopsy specimens

The mucosal biopsy specimens were collected from485 patients subjected to routine gastroscopic examinations. Three antral specimens fromeach patient underwent smear staining, microbiological culture, histological examinationand “RUN test”, respectively. Some otherspecimens were collected for detecting ammonia content and for optimizing the “RUN test”system. Blood samples were collected forELISA detection of anti-Hp antibody.

Quantitative determination of ammoniacontents in biopsy specimens

The specimens for ammonia quantitativedetermination were selected from 56 patients. Mucosal smears were made, stained andexamined by a pathologist in a blind way. After this, the same specimens were used fordetermining ammonia contents immediately. The preparation of Nessler′s reagent, charting of the standard curve and theprocedures of determination all followed the standard method^[2] . Thereacted solutions were determined with a BECKMEN DU 65 spectrophotometer at a wavelengthof 410nm and the results were converted to ammonia contents (μmol/specimen) according to the standard curve.

Preparation of RUN test reagents

To develop a convenient testing technique, weused the materials and the method of ELISA for reference. In brief, the Nessler′s stock solution was prepared bythe described method^[2] and was diluted 1∶4 with 0.67mol/L KOH. Stored at room temperature for oneweek, the supernatants (diluted Nessler′s reagent) were filled into the drop bottles for use. The ureacoated wells were prepared as follows: Each well of the 12 or 8 microtiter well stripeswas added with 50μlunbuffered 0.133mol/L urea, and then dried at 50℃ overnight for coating. A piece of adhesive of plastic tapewas stuck onto each well to prevent deliquescence after drying. The test system wasoptimized with standard (NH₄ )₂ SO₄ solution to obtain thebest performances in the rapid test before the clinical trials.

The procedure of RUN test

In clinical test, two drops of distilled waterwere added into each urea coated well. When the dried urea had dissolved, a gastric biopsyspecimen was attached to and one drop of the diluted Nessler′s reagent was dripped into the well. The color of the testsolution was first observed by naked eye on a white background (such as a piece of whitepaper) within 5 minutes. A positive result was recorded if there was a change fromcolorless to yellow, while the negative remained colorless. The observed test wells werethen read with a microtiter reader (MINIREADER Ⅱ, Dynate Inc.) at a wavelength of 410nm after theinstrument was calibrated with well containing distilled water. Specimens that gave an ODat 410nm greater than or equal to 0.15 and the p/n over 3.0 were regarded as Hp positive,and otherwise, negative.

Standard examinations

The smear staining and culture of each specimen,as well as the identification of bacterial colonies, were performed as described before^[3] . The second specimen of each patient was fixed in 10% formalin,stained and examined by a pathologist according to the reported method^[4] and classified histopathologically with the Sydney system^[5] . ELISA of blood samples was performed as the direction of the Kit(purchased from the Shanghai Gastroenterological Diseases Institute). The patient wasregarded as Hp infection positive, provided that two out of the results of the smearstaining, culture and ELISA were positive.

RESULTS
Comparison of ammonia contents between Hp positives and negatives

The 56 biopsy specimens by smear staining showed36 Hp positives and 20 negatives. While these specimens were examined by the standardNessler′s method,obvious difference was observed between the ammonia contents of positive and negativesamples (Table 1). These results not only suggested that the ammonia contents in Hpinfecting mucosa were higher than those in negative mucosa, but also implied that thevalues of ammonia contents in biopsy specimens could be used as indicators for identifyingHp infection in mucosa.

Table 1 Ammonia content in Hp infection

	Hp infection
	Negative	Positive
Specimens	20	36
Ammonia content (×10-2 μmol/specimen)	1.0-2.4	2.6-12.8
Mean±SD (×10-2 μmol/specimen)	1.6±0.37	5.4±1.6

Figure 1 The inhibitiory effect of urea on Nessler′s reaction.

The (NH₄ )₂ SO₄ solution(0.6mmol/L) containing serial concentration of urea were reacted with diluted Nessler′s reagent in microtiter wells for10 minutes and the reacted solution in wells was read with the microtiter reader at awavelength of 410nm. The results showed that the inhibition of urea on Nessler′s reaction depended on theconcentration of urea (Figure 1). Therefore, these results made it possible to eliminatethe “ammoniabackground” effect ofHp negative gastric biopsy specimens by employing 0.133mol/L urea in the RUN test system.

RUN test results

In all 485 patients, 368 patients (75.9%) werediagnosed as being Hp positive through “standard examination”. The RUN test correctly identified 367 out of the 368 Hp positivespecimens by both naked eye observation and microtiter reader determination (Table 2).Judged by naked eye, the statistical ana-lysis indicated the RUN test having asensitivity, specificity and positive and negative predicative values of 99.7%, 91.5%,97.3% and 99.1% respectively; whereas with the microtiter reader, the correspondant valueswere 99.7%, 93.2%, 97.9% and 99.1% respectively. Therefore, there was a goodcorrespondence between subjective observation and instrumental determination. Thediagnostic efficiency^[6] of RUN test, bacterial culture, smear staining and ELISA was alsocompared and the results showed that the RUN test possessed higher diagnostic efficiency(Table 3).

Table 2 Comparison of the results of RUNtest

RUN test		Standardexaminations		Total
		Positive	Negative
Positive	Ⅰ	367	10	377
	Ⅱ	367	8	375
Negative	Ⅰ	1	107	108
	Ⅱ	1	109	110
Total	Ⅰ	368	117	485
	Ⅱ	368	117	485

Ⅰ: Judged by naked eye Ⅱ: Determined with microtiter reader Table 3 Efficiency of diagnostic methods for Hp

	Standard examination	Culture	Smear staining	ELISA	RUN test*
Positive	368	295	348	357	377
Negative	117	190	137	128	108
False positive		0	9	18	10
False negative		73	29	29	1
Sensitivity (%)		80.2	92.1	92.1	99.7
Specificity (%)		100	92.3	84.6	91.5
Positive (%) predictive value		100	97.4	95.0	97.3
Negative (%) predictive value		61.6	78.8	77.3	99.1
* Efficiency (%)		84.9	92.2	90.3	97.7

* Judged by naked eye Correlation between Hp infection and gastric diseases Table 4 shows a good correlation between Hp infection rates and gastric diseases. Table 4 Relationship between Hp infection rate and gastric diseases

Gastric diseases*	Total	Hp positive	Hp infection rate (%)
CSG	256	184	72.16
CAG	80	60	75.00
GU	31	25	80.65
DU	70	67	95.71
GM	23	18	78.26
GD	11	9	81.82
GC	6	4	66.67

^* CSG: chronic surperficial gastritis; CAG: chronic atrophic gastritis; GU: gastric ulcer; DU: duodenal ulcer; GM: gastric mucosal metaplasia; GD: gastric mucosal dysplasia; GC: gastric cancer

DISCUSSION

The principle of the RUN test described here is greatly different from the conventional urease tests based on detecting urease activity^[3,6] , but the RUN testis somewhat like the method of detecting ammonia in gastric juice of patients for Hpinfection diagnosis^[1] . Hp possesses a high urealytic activity, and is able to neutralizediffusion of hydrogen ion into the gastric mucosa through generation of ammonia, forming “ammonia cloud” around theorga-nisms to protect themselves from being killed by gastric acid. Therefore, we couldspeculate that the ammonia contents in areas of Hp colonization should be higher thanthose in the areas without Hp. This speculation was proved in our determination of ammoniacontents in biopsy specimens with the Nessler′sreagent (Table 1). The RUN test was thereby designed on the basis of detecting ammoniacontents in biopsy specimens with diluted Nessler′sreagent in microtiter wells. Although the mechanism of the inhibition of urea on Nessler′s reaction (Figure 1) remains unknown, this discovery allowed us toeliminate the “ammonia background” in Hp negative specimens by coating and drying the urea in the testwells previously. Our test system has the advantage that, unlike the complicated andunreliable method of detecting ammonia concentration of gastric juice, the positive testresults can be conveniently judged by the color change within 5 minutes by naked eye, ordetermined by a microtiter reader. Basing on these characteristics, we further deve-lopeda semiquantitative method for diagnosing Hp infection in human gastric mucosa (the resultswill be published in part Ⅱ of this paper). Theother advantage of the RUN test is that the dried urea coating wells and diluted Nessler′s reagent can be stored at room temperature for over six months.

¹ Cancer Research Center of XiamenUniversity, Xiamen 361005, China
² Department of Gastroenterology, Xiamen Zhongshan Hospital, Xiamen 361005,China
CHEN Rui-Chuan, male, born on 1963-05-02, Director of Biochemical Laboratory, having 15papers published.
Correspondence to: Dr. CHEN Rui-Chuan, Cancer Research Center of XiamenUniversity, Xiamen 361005, China
Received 1997-08-21

REFERENCES

1 Neithercut WD,Rowe PA, Elnuhumi AM, Dahill S, Mccoll KEL. Effect of Helicobacter pylori infection onintragastric urea and ammonia concentration in patients with chronic renal failure. J Clin Pathol, 1993;46(6):544-547
2 Henry RJ. Clinical chemistry: principles and technics. New York: Harper & Row Publishers, 1964:258-262
3 Husson MO, Leclerc H. Detection of Helicobacterpylori in stomach tissue by use of a monoclonal antibody. J Clin Microbiol,1991;29(12):2831-2834
4 Rugg M, Dirnanio F, Cassaro M, Baaffa R, FarinatiF, Rubio J et al. Pathology of the gastric antrum and body associated with Helicobacterpylori infection in non-ulcerous patients is the bacterium, a promoter of intestinal metaplasia. Histopathology, 1993;22(1):9-15
5 Price AB. The Sydney system: histologicaldivision. J Gastroenterol Hepatol, 1991;6(2):207-322
6 Thillainpagam AV, Arvind AS, Cook RS, Harrison IG,Tabaqchali S, Farthing MJG. Diagnostic efficiency of an ultrarapid endoscopy test forHelicobacter pylori. Gut, 1991;32(4):467-469 (CHEN Rui-Chuan¹, ZHANG Chang-Gong¹, WU Yan-Huan² and YANG Shan-Min<)