Conventional Cleaning and Disinfection Techniques Eliminate the Risk of Endoscopic Transmission

Conventional Cleaning and Disinfection Techniques Eliminate the Risk of Endoscopic Transmission of Helicobacter pylori

227

Original contributions

Conventional Cleaning and Disinfection Techniques Eliminate the Risk of Endoscopic Transmission of Helicobacter pylori

George T. Fantry M.D.

Qiao-Xi Zheng Ph.D.

Stephen P. James M.D.

Division of Gastroenterology, Department of Medicine, University of Maryland and the Baltimore VAMC,

Baltimore, Maryland

Objective: To determine whether endoscopes serve as a reservoir for Helicobacter pylori and whether two commonly used cleaning and disinfection methods eliminate the risk of H. pylori transmission. Methods: A prospective study was carried out in 107 patients who were undergoing upper gastrointestinal endoscopy for routine clinical indications. H. pylori DNA was assayed by polymerase chain reaction (PCR) of endoscope washes before and after procedure, in gastric aspirates and in endoscope washes after cleaning and disinfection of endoscopes. Gastric biopsies were assayed by rapid urease test (CLOtest, Tri-Med Specialties Inc., Lenexa, KS) of two antral biopsies. Results: Forty-one of 107 (38%) patients were H. pylori-positive by PCR. Endoscopes were contaminated after 25 of 41 (61%) H. pylori-positive procedures. However, 107 of 107 pre-endoscopy and postcleaning aspirates were negative, indicating that decontamination was 100% effective. The urease test was positive in 25 of 41 H. pylori-positive patients, a sensitivity of 61%. PCR was positive in 41 of 41 H. pylori-positive patients, a sensitivity of 100%. In 5 of 16 PCR-positive/urease-negative patients, the identification of H. pylori was clinically relevant. Conclusion: Endoscopes are frequently contaminated with H. pylori after endoscopy in H. pylori-infected patients, but conventional cleaning and disinfection techniques are highly effective in eliminating H. pylori. When appropriate negative control samples are obtained from the endoscope, PCR of endoscopic gastric aspirates appears to be a sensitive test that can detect clinically relevant H. pylori infection that is missed when only a rapid urease test is used.

Reprint requests and correspondence: Dr. George T. Fantry, University of Maryland Medical System, Division of Gastroenterology, Room N3W62, 22 South Greene Street, Baltimore, MD 21201.

Received Aug. 12, 1994; accepted Nov. 3, 1994.

INTRODUCTION

The potential for transmission of infections by upper gastrointestinal endoscopy is a concern that has led to the promulgation of carefully developed guidelines for disinfection of endoscopes [1] . Greatest attention has been given to high-grade pathogens, such as HIV and hepatitis virus. However, relatively little attention has been given to the most prevalent infectious agent in the stomach, Helicobacter pylori. H. pylori infection is one of the most common infections in humans, and it has been implicated as an important etiological agent in ulcer disease and gastric neoplasia. Epidemiological and clinical data suggest that H. pylori is usually acquired in childhood and often persists for life [2] [3] . Thus the gastric mucosa is an important potential reservoir for transmission of this infection. H. pylori DNA has been identified in dental plaque by PCR [4] and has been identified in feces [5] , suggesting that fecal-oral or oral-oral modes of transmission are possible. The presence of similar strains of H. pylori within families is compatible with person-to-person transmission or common source contact [6] . Nonetheless, the precise mechanism of transmission of this common infection is still unknown. The stomach has nonspecific and specific defense mechanisms that normally keep the gastric environment nearly sterile, with the exception of H. pylori. Therefore, it is not surprising but somewhat paradoxical that nosocomial transmission of H. pylori is in fact the only proved mechanism of transmission [7] [8] [9] [10] .

Although it seems likely that currently recommended disinfection methods for GI endoscopes should kill H. pylori, at present no published data on the efficacy of disinfection for killing H. pylori could be found in a Medline literature search. When appropriate disinfection guidelines are not followed, iatrogenic transmission of H. pylori is clearly possible. Several reports have specifically indicated that contaminated endoscopes or equipment may transmit infection [7] [8] [9] [10] . However, because of the frequency of H. pylori in the general population and its possible production of an asymptomatic or indolent clinical syndrome, it is not expected that nosocomial transmission of H. pylori would be easily recognized or frequently reported. Several reports suggest that H. pylori DNA is frequently present on gastrointestinal endoscopes despite appropriate disinfection [11] [12] [13] [14] . These observations not only suggest that endoscopes might transmit H. pylori, but also indicate that endoscope contamination would cause frequent false-positive results for PCR-based diagnostic tests. Therefore, the aims of this investigation were: to determine the frequency with which H. pylori is present in upper GI endoscopes after procedures in patients with H. pylori infection; and to determine

228

TABLE 1 -- Clinical and Endoscopic Findings in Patient Population

All patients H. pylori-positive

patients Urease-negative,

PCR-positive patients

N 107 41 16

Age (range) 57.7 (24-85) 61.1 (28-83) 67.3 (48-83)

Race

African-American 51 (47.7%) 22 (53.7%) 8 (50%)

Caucasian 52 (48.6%) 16 (39%) 8 (50%)

Other 4 (3.7%) 3 (7.3%) 0 (0%)

Historical findings

Prior antibiotic therapy 6 (5.6%) 2 (4.9%) 1 (6.3%)

Current omeprazole 11 (10.3%) 2 (4.9%) 0 (0%)

Prior history of ulcer 20 (18.7%) 8 (19.5%) 3 (18.8%)

Endoscopic diagnosis

Ulcer 14 (13.1%) 10 (24.4%) 2 (12.5%)

Gastroesophageal reflux disease/Barrett's 13 (12.1%) 4 (9.8%) 0 (0%)

Clinical diagnosis

Nonulcer dyspepsia 23 (21.5%) 12 (29.3%) 5 (31.3%)

whether endoscope contamination with H. pylori DNA could cause false-positive results in PCR diagnostic tests.

METHODS

Patients

Patients who were undergoing upper gastrointestinal endoscopy for routine clinical indications in two endoscopy suites were eligible for inclusion in the study. The only exclusion criterion was a contraindication to antral biopsy. One hundred seven patients were studied prospectively (Table 1) . Informed consent was obtained from each patient. Six patients had received prior antibiotics with activity against H. pylori (four specifically for H. pylori cure), and 11 had been treated with omeprazole at the time of endoscopy.

Endoscopic samples

Endoscope washings and gastric samples were collected by aspirating saline through the suction channel of the endoscope before the procedure, by aspirating gastric juice from the fundic pool upon entering the stomach, and by aspirating saline through the suction channel on completion of the endoscopy as well as after cleaning and decontamination of the endoscope. Five milliliters of each sample was obtained in a sterile suction trap attached directly to the endoscope and stored unrefrigerated until assayed. Two gastric biopsies were obtained from the antrum on all patients during endoscopy for rapid urease test medium (CLOtest). The CLOtest was kept refrigerated until before the procedure, when it was brought to room temperature. Biopsies were immediately transferred into the medium and immersed using a sterile needle. The CLOtest was incubated at 37°C for 4-8 h and subsequently observed at room temperature for a total of 24 h.

Endoscope cleaning and disinfection

Endoscopes were cleaned and disinfected by the nurse or technician who assisted with the procedure according to current recommended methods [1] . No alterations in cleaning procedures were made for the purpose of conducting this study. Upon completion of the endoscopic procedure, the outside of the endoscope was immediately wiped clean with a solution containing a bacteriostatic enzymatic cleaner (Endozyme, Ruhof Corporation, Valley Stream, NY). This solution was subsequently suctioned through the biopsy and therapeutic channels of the endoscope. The endoscope was then completely immersed in this solution, and the exterior of the scope was thoroughly cleaned. Next, the entire suction/biopsy channel system was brushed until no debris was visible on the brush. Cleaning adaptors were attached to the endoscope, and with the endoscope immersed, detergent was flushed through all the channels until no visible material remained. Finally, the endoscope was rinsed thoroughly with water to remove any residual detergent, and all channels were purged with air.

Endoscopes were then disinfected with two conventional methods, one manual method and one automated disinfection method. For the first method, after cleaning the endoscope, the endoscope was disinfected with 2% glutaraldehyde (Cidex, Johnson and Johnson Medical Inc., Arlington, TX). When using glutaraldehyde, all of the channels were filled with glutaraldehyde using the cleaning adapters, and the entire endoscope was then immersed for 20-30 minutes. For the second, automated method (using 0.2% peracetic acid and Steris automated cleaner, Steris Corp., Mentor, OH), the endoscope was placed into a sterilizer, and automated disinfection and rinsing was done in a 20-30 minute cycle. After decontamination was completed, all surfaces and channels of the endoscope were rinsed with water to remove all traces of the disinfectant. All channels were purged with air. Finally, all channels were flushed with

229

70-90% alcohol to facilitate drying and purged with air again. For storage, the endoscopes were hung vertically in a ventilated cabinet.

PCR assay for H. pylori

One milliliter of each sample was centrifuged for 1 minute at 14,000 × g, and the supernatant was discarded. The pellet was treated with 40 mul of 0.1% Triton and 10 mul of 0.4 N NaOH. The samples were vortexed for 10 s, centrifuged for 10 s, and heated at 95°C for 5 minutes. Ten microliters of 1 M Tris pH 7.5 was added, and the samples were mixed and centrifuged at 14,000 × g for 15 minutes. PCR amplification was performed in 50-mul reaction volumes containing 5 mul of DNA sample and 45 mul of PCR reaction buffer. The PCR mixture consisted of 30.75 mul of water, 5 mul of 10x PCR buffer [100 mM Tris-HCl pH 8.3, 500 mM KCl, 25 mM MgCL, 1% (w/v) gelatin, 4 mul of a mixture of 1 mM deoxyadenosine triphosphate, deoxythymidine triphosphate, deoxycytidine triphosphate, and deoxyguanosine triphosphate, AmpliTaq DNA polymerase (Promega, Madison, WI) (0.25 mul) and 2.5 mul of 20 H. pylori primers. Two 20-base oligonucleotide primers, designated CAM-2 (5`-CATCTTGTTAGAGGGATTGG-3`) and CAM-4 (5`-TAACAAACCGATAATGGCGC-3`), were selected to amplify a 203-base pair fragment of a previously described 1.9-kb region that is highly conserved and specific for H. pylori DNA [15] . A second set of primers was used to amplify a 411-base pair product from the urease gene A: HPU1 5` GCCAATGGTAAATTAGTT3` and HPU2 5` CTCCTTAATTGTTTTTAC3` [16] . The tubes were then overlaid with one drop of mineral oil to prevent evaporation. The standard PCR consisted of 40 cycles on the thermal cycler (Perkin Elmer Cetus, Norwalk, CT). After a 2-minute melting step at 94°C, each cycle consisted of 20 minutes at 94°C, 15 minutes at 45°C, and 20 minutes at 72°C. An additional extension step consisted of 2 minutes at 72°C. Fifteen microliters of PCR product was combined with 5 mul of loading buffer (glycerol 6 ml, 0.5 M ethylenediamine-tetraacetic acid 1.5 ml, 1 M Tris pH 8.1 ml, ddH2 O 1.8 ml, 1% bromphenol blue 0.3 ml) and electrophoresed for approximately 60 minutes at 90 volts on 1.6% agarose gel (Ultra Pure Agarose-GIBCO BRL, Grand Island, NY) stained with ethidium bromide. As a molecular weight standard, a 100-base pair ladder was used [20] .

Each reaction contained both positive and negative controls. To determine the sensitivity of the reaction, serial dilutions of cultured H. pylori cells that were diluted in normal gastric fluid were tested, and it was found that the standard PCR protocol detected 25 bacteria per mul of gastric fluid.

RESULTS

Detection of H. pylori by PCR

In this prospective study, materials from a total of 107 procedures were tested (Table 1) . In initial studies, PCR amplification of DNA from cultured H. pylori and gastric aspirates using CAM primers resulted in 203-base pair bands, the predicted size for this unique H. pylori-associated genomic sequence. In all cases tested, digestion of the PCR products with the restriction enzyme HinfI resulted in two bands of the correct predicted size. In one case, the PCR products from a gastric aspirate were subcloned and sequenced, and the sequence was 99% identical to the published H. pylori sequence for CAM (data not shown). The PCR results were also confirmed by amplification with H. pylori-specific urease primers. Thirty-one of 34 CAM-positive patients were also positive using urease primers. Finally, in selected cases, urease PCR products were demonstrated to hybridize to urease-specific probes on Southern blots.

In 11 of 107 cases, PCR products amplified from gastric samples were obtained using CAM primers that were not the appropriate size (203 base pairs). In these cases, restriction digestion of the PCR fragments did not result in fragments of the correct predicted size, and in two cases, direct sequencing revealed that there was no sequence homology with the published CAM sequence. Therefore, for the purposes of this analysis, CAM-PCR fragments of incorrect size were classified as a negative result.

One hundred seven patients had both PCR of endoscopic gastric aspirates and rapid urease test (CLOtest) performed to detect H. pylori infection. Forty-one (38%) were positive by PCR, and these cases were defined as H. pylori-positive. The urease test was positive in 25 of 41 H. pylori-positive patients, a sensitivity of 61%. PCR was positive in 41 of 41 H. pylori-positive patients, a sensitivity of 100%. The specificity of both tests was 100%. Of the patients who were urease-negative and PCR-positive, PCR was positive for both CAM and urease in 14 of 16 cases.

Endoscope contamination after H. pylori positive procedures

Preprocedure samples were obtained from endoscopes that had been disinfected and stored before 107 procedures. In every case, the sample was negative for H. pylori. An immediate postprocedure (before cleaning) aspirate through the suction channel was obtained after these 107 procedures, including the 41 patients with H. pylori infection as defined above. Evidence of H. pylori DNA in the endoscope was detected by PCR after 25 of 41 (61%) procedures performed in H. pylori-positive patients.

Endoscope testing after cleaning and decontamination

In 107 cases, a postcleaning and decontamination aspirate was obtained and tested by PCR for detection of H. pylori DNA. A manual glutaradelhyde disinfection method (Cidex) was used in 51 cases, and an automated (Steris) method was used in 56 cases. Eleven different video and fiberoptic endoscopes were tested, including both diagnostic and therapeutic endoscopes. These cases included the 41 H. pylori-positive patients, of which the immediate postprocedure

230

TABLE 2 -- Two Commonly Used Disinfection Methods Eliminate H pylori from Upper GI Endoscopes Disinfection method Total N

procedures N H. pylori-positive patients N H. pylori-positive

endoscopes immediately

after procedure N H. pylori-positive endoscopes

after disinfection

Automated peracetic acid (Steris) 56 24 16 0

Manual, glutaraldehyde (Cidex) 51 17 9 0

Total 107 41 25 0

aspirate was positive in 25 (61%). In all 107 cases (100%), postcleaning and decontamination aspirates were negative by PCR for both the manual and automated method of disinfection (Table 2) .

PCR-positive, urease test-negative patients

Sixteen patients were PCR-positive and urease test-negative. Fourteen of these patients were positive using both CAM and urease primers. In 5 of 16 (31%) of the PCR-positive and urease test-negative patients, the detection of H. pylori infection was clinically relevant (Table 1) . Four patients had a history of peptic ulcer disease, and two patients had active ulcer disease at endoscopy. No patients who were rapid urease test-negative and PCR-positive were receiving omeprazole. One had received prior therapy with tetracycline alone. None had received specific therapy for H. pylori infection.

DISCUSSION

Although accurate statistics are not available, it appears that the number of upper GI endoscopies performed yearly in the United States numbers in the millions [17] . Because the percentage of individuals positive for H. pylori ranges from about 12% for individuals under age 20 to about 50% by age 50 [2] [3] , the potential for contamination of endoscopes with H. pylori and iatrogenic transmission is high. It has been well recognized for many years that because of the complex, fragile construction of endoscopes and because of the difficulty of decontaminating them, endoscopy serves as a potential source of iatrogenic infection. Recently, Spach et al.[18] reviewed the English literature on transmission of infections by gastrointestinal endoscopy and bronchoscopy. Fortunately, the numbers of papers reporting transmission of infection by upper GI endoscopy is very small, and most cases have been associated with improper disinfection. Graham et al.[8] reported transmission of H. pylori to a single patient, possibly due to contaminated biopsy forceps. More recently, Langenberg et al.[9] reported transmission of H. pylori in three cases due to improper decontamination of endoscopes. Other papers have reported transmission of different organisms, most commonly Pseudomonas aeruginosa, associated with contamination of automatic washers and improper disinfection [19] . The low frequency of documented transmission of H. pylori by endoscopy, and the fact that a correctable decontamination error has been identified in the few cases reported, is reassuring. However, no prospective surveys of H. pylori contamination have been carried out, for several possible reasons. First, because of the indolent nature of H. pylori infection and because of its high frequency in the general population, nosocomial transmission would be difficult to recognize. Secondly, the mode(s) of transmission of H. pylori are currently unknown. The facts that H. pylori is found primarily in human gastric mucosa and that it can be identified in feces have led to the hypothesis that transmission is a result of contaminated water, although there has been little direct evidence to support this possibility. Nevertheless, this remains a concern because of the potential contamination of endoscopy washing machines. Third, H. pylori is not easily cultured, and it is thought that environmental forms of the organism outside the stomach, such as those present in water, may be difficult to recognize [20] , thus hindering surveillance studies of contaminated equipment.

In the present study, PCR amplification of DNA was used to assay for the presence of H. pylori. When carried out with proper negative and positive controls, this method has the advantage of speed and very high sensitivity compared to culture methods [15] [16] [21] [22] [23] . A potential shortcoming of the PCR method is that if DNA is detected, it does not necessarily indicate that infectious organisms are present, because fragments of naked (noninfectous) DNA could be detected that would not represent a risk for transmission. In the present study, in no instance was H. pylori DNA detected after decontamination of endoscopes using two commonly used methods. One method used manual cleaning with glutaraldehyde disinfection, and the second method used an automated machine with peracetic acid disinfection. This finding was a relevant negative finding in that we were able to demonstrate that after endoscopy of patients with H. pylori infection, the endoscope was contaminated in the majority of cases before cleaning and disinfection. This study was specifically designed to test for the presence of H. pylori in endoscopes that were contaminated by normal use and in the presence of gastric secretions, rather than to test endoscopes purposefully contaminated with laboratory isolates, which might be more readily recoverable. The study involved two different endoscopy units and multiple different types of upper GI endoscopes, including both video and fiberoptic instruments. In this study, no specific instructions were given to the endoscopy personnel to alter cleaning and decontamination techniques, and personnel were instructed to follow routine procedures for cleaning and disinfection. It