Proton pump inhibitors (PPI) are a major cause of false-negative Helicobacter pylori test results. Detecting PPI use and stopping it 2 weeks before testing is the preferred approach to improve the reliability of H pylori diagnostic tests. Immunoblot and molecular methods may be useful for the detection of H pylori infection in difficult cases. When conventional tests are negative and eradication is strongly indicated, empirical H pylori treatment should be considered. In this article, an updated critical review of the usefulness of the various invasive and noninvasive tests in the context of extensive PPI use is provided.
Key points
- •
Proton pump inhibitors (PPI) are a major cause of false-negative Helicobacter pylori test results.
- •
Detecting PPI use and stopping it 2 weeks before testing is the preferred approach to improve the reliability of H pylori diagnostic tests in this setting.
- •
Immunoblot and molecular methods may be useful for the detection of H pylori infection in difficult cases.
- •
When conventional tests are negative and eradication is strongly indicated, empirical H pylori treatment should be considered.
Introduction
Helicobacter pylori is the major cause of peptic ulcer and gastric cancer. It is estimated that more than 50% of the world population is affected by this chronic gastric infection. Diagnostic methods for the infection have been classically divided into invasive (those that require endoscopy) and noninvasive (those that not). The invasive methods available in clinical practice are histology, rapid urease test (RUT), and culture. In addition, molecular methods (in particular, biopsy polymerase chain reaction [PCR]) have been gaining popularity, although their use is still mainly restricted to research. Practical noninvasive methods include urea breath test (UBT), stool test, and serology.
In the last decade, we have become increasingly aware of the strengths and limitations of diagnostic techniques for H pylori infection. In particular, (1) molecular methods have suggested that an undetermined percentage have low-density infection, that is not detected by the traditional tests; (2) the current widespread use of proton pump inhibitors (PPI) or antibiotics may markedly reduce the sensitivity of diagnostic techniques, and (3) diffuse mucosal damage associated with mucosal-associated lymphoid tissue (MALT) lymphoma or extensive intestinal metaplasia may produce false-negative results in diagnostic tests ( Table 1 ). These drawbacks have led to a search for improved techniques and have highlighted the need for changes in the diagnostic approach to particular situations ( Table 2 ).
| Factor | Potential Mechanisms | Effect |
|---|---|---|
| Extensive atrophy | Low acid secretion, diffuse mucosal damage, overgrow of urease-positive bacteria | Decreased sensitivity Decreased specificity? |
| Extensive intestinal metaplasia | Low acid secretion, overgrow of urease-positive bacteria | Decreased sensitivity Decreased specificity? |
| MALT lymphoma | Diffuse mucosal damage | Decreased sensitivity |
| Use of antibiotics | Direct antibacterial effect, low density of infection | Decreased sensitivity |
| Use of PPI | Decreased acid secretion, low density of infection, direct antibacterial effect? | Decreased sensitivity |
| Use of histamine 2 receptor antagonists | Decreased acid secretion, low density of infection | Decreased sensitivity |
| Measure | Test Involved | Mechanism |
|---|---|---|
| Avoid antibiotics | Histology, culture, UBT, RUT, SAT | Increase gastric H pylori density |
| Avoid acid secretion, especially PPI | Histology, culture, UBT, RUT, SAT | Increase gastric H pylori density |
| Locally validate the test | SAT, serology, Immunoblot | Accurate test selection |
| Antrum and gastric corpus sampling | Histology, culture, RUT | May detect residual, low-density H pylori infection in gastric corpus glands |
| Use of ancillary stains/immunohistochemistry | Histology | May detect residual, low-density H pylori infection in gastric corpus glands |
| Use of citric acid pretest meal | UBT | Increase urease activity, inhibits non– H pylori urease, decreases gastric overgrow |
In this article, an updated critical review of the usefulness of the various invasive and noninvasive tests in the context of extensive PPI use is provided.
Introduction
Helicobacter pylori is the major cause of peptic ulcer and gastric cancer. It is estimated that more than 50% of the world population is affected by this chronic gastric infection. Diagnostic methods for the infection have been classically divided into invasive (those that require endoscopy) and noninvasive (those that not). The invasive methods available in clinical practice are histology, rapid urease test (RUT), and culture. In addition, molecular methods (in particular, biopsy polymerase chain reaction [PCR]) have been gaining popularity, although their use is still mainly restricted to research. Practical noninvasive methods include urea breath test (UBT), stool test, and serology.
In the last decade, we have become increasingly aware of the strengths and limitations of diagnostic techniques for H pylori infection. In particular, (1) molecular methods have suggested that an undetermined percentage have low-density infection, that is not detected by the traditional tests; (2) the current widespread use of proton pump inhibitors (PPI) or antibiotics may markedly reduce the sensitivity of diagnostic techniques, and (3) diffuse mucosal damage associated with mucosal-associated lymphoid tissue (MALT) lymphoma or extensive intestinal metaplasia may produce false-negative results in diagnostic tests ( Table 1 ). These drawbacks have led to a search for improved techniques and have highlighted the need for changes in the diagnostic approach to particular situations ( Table 2 ).
| Factor | Potential Mechanisms | Effect |
|---|---|---|
| Extensive atrophy | Low acid secretion, diffuse mucosal damage, overgrow of urease-positive bacteria | Decreased sensitivity Decreased specificity? |
| Extensive intestinal metaplasia | Low acid secretion, overgrow of urease-positive bacteria | Decreased sensitivity Decreased specificity? |
| MALT lymphoma | Diffuse mucosal damage | Decreased sensitivity |
| Use of antibiotics | Direct antibacterial effect, low density of infection | Decreased sensitivity |
| Use of PPI | Decreased acid secretion, low density of infection, direct antibacterial effect? | Decreased sensitivity |
| Use of histamine 2 receptor antagonists | Decreased acid secretion, low density of infection | Decreased sensitivity |
| Measure | Test Involved | Mechanism |
|---|---|---|
| Avoid antibiotics | Histology, culture, UBT, RUT, SAT | Increase gastric H pylori density |
| Avoid acid secretion, especially PPI | Histology, culture, UBT, RUT, SAT | Increase gastric H pylori density |
| Locally validate the test | SAT, serology, Immunoblot | Accurate test selection |
| Antrum and gastric corpus sampling | Histology, culture, RUT | May detect residual, low-density H pylori infection in gastric corpus glands |
| Use of ancillary stains/immunohistochemistry | Histology | May detect residual, low-density H pylori infection in gastric corpus glands |
| Use of citric acid pretest meal | UBT | Increase urease activity, inhibits non– H pylori urease, decreases gastric overgrow |
In this article, an updated critical review of the usefulness of the various invasive and noninvasive tests in the context of extensive PPI use is provided.
Invasive tests
Histology
Histology has been the mainstay of the invasive diagnosis of H pylori infection. Its global reliability remains high in recent studies, with sensitivity and specificity rates higher than 95%. However, although histology remains one of the most accurate diagnostic tests, its performance in clinical practice may be lower because of generalized use of PPI, inadequate biopsy sampling, and inadequate staining.
The active use of PPI or antibiotics is known to cause false-negative results in all invasive tests. The rate of false-negative results is reported to be at least 30%. Current guidelines recommend stopping PPI at least 2 weeks before testing for H pylori infection and antibiotics at least 4 weeks beforehand. However, a recent study in Canada showed that 47% of the patients who were tested by histology were actively receiving a PPI.
Regarding sampling, the recommended standards for histology are those of the updated Sydney classification system, which requires at least 1 biopsy from 5 different sites: (1) the lesser curvature of the corpus about 4 cm proximal to the angulus; (2) and (3) the lesser and greater curvature of the antrum, both within 2 to 3 cm of the pylorus; (4) the middle portion of the greater curvature of the corpus, approximately 8 cm from the cardia; and (5) the incisura angularis. However, this approach is not widely followed in clinical practice, and fewer than 5% of gastric biopsies in the United States comply with these standards. Additional data from the Canadian study also show that 60% of the biopsies had samples from only 1 gastric site, mainly from the antrum.
Gastric antrum biopsies seem reliable enough for most patients. The sensitivity of histology decreases in patients with extensive inflammation or atrophy, and the antrum-only approach is not recommended in areas in which atrophic changes are common. In this regard, sampling of the corpus greater curvature is recommended, because it increases the likelihood of a positive biopsy in patients with atrophy.
Similarly, in their retrospective study including a large series of biopsies, Lash and Genta reported that the use of at least 2 antral and 2 corpus biopsies increased the diagnostic yield of histology for H pylori infection. However, the increase in the number of positive biopsies was moderate, from 10% to 15%. No cost analyses have been performed to ascertain whether this modest increase in sensitivity justifies routine performance and analysis of 2 or 3 sets of biopsies, although placing several specimens on 1 slide may decrease the pathology costs.
Obtaining adequate sampling and corpus biopsies is especially important in patients receiving PPI therapy. Histologic changes induced by PPI in individuals infected by H pylori were first described by Graham and colleagues. Although the changes vary, patients with Helicobacter infection using PPI tend to have an almost normal antrum with minimal chronic inflammation, no neutrophils, and no detectable bacteria; in contrast, the corpus may show chronic active gastritis. The few remaining bacteria may occasionally be found in deeper portions of the oxyntic glands. Therefore, because infection seems to remain in the corpus, antrum-only sampling may further reduce the sensitivity of histology in PPI users.
With regard to the staining of biopsies, Giemsa is generally preferred to hematoxylin-eosin. Giemsa stain is routinely performed at most centers and is cheap and highly reliable for the diagnosis of H pylori infection. By contrast, it has been suggested that hematoxylin-eosin alone has a lower sensitivity for diagnosing H pylori infection. However, pathologists consider the 2 stains to be equivalent.
Other staining techniques including inmunohistochemical detection of the bacteria or silver staining are routinely not indicated, because they do not clearly increase the diagnostic accuracy. However, ancillary tests may be useful when the result is uncertain, particularly in PPI users, in whom other bacteria are likely to be present. It has been suggested that immunohistochemical staining may help to identify low-density infection in oxyntic glands in patients receiving PPI, especially when minimal inflammation is present. However, it remains unclear whether H pylori –negative active gastritis represents a new entity or is caused by low-density infection not detected by histology or other conventional diagnostic methods.
PPI use reduces the sensitivity of histology for diagnosing H pylori infection. Whether corpus sampling, or the use of immunohistochemical staining, or a combination of the 2 may reduce the false-negative results associated with PPI use is an issue that needs further investigation.
Culture
The feasibility and usefulness of H pylori culture has been a matter of debate among microbiologists and clinicians since the bacterium was first isolated. Culture allows the determination of antibiotic resistances to design an optimal therapy. However, it is time and resource consuming and is routinely performed only at a few centers. Specificity of culture is high, approaching 100%, but sensitivity is lower; although some reference centers claim sensitivity rates more than 90%, in most settings they range between 40% and 80%.
The Maastricht consensus states that culture is important before first-line therapy, should be considered before second-line therapy (statement 5), and is mandatory (whenever possible) before third-line therapy (statement 18). However, the scientific evidence supporting the Maastricht recommendations is weak. Two recent meta-analyses showed that treatment guided by antibiotic susceptibility was superior to empirical treatment of first-line therapy. However, a major limitation of all the studies included was that all patients were randomized after endoscopy had been performed. Because no study has randomized patients to empirical treatment versus endoscopy and culture plus culture-guided treatment, the effectiveness of susceptibility-guided treatment in clinical practice has never been evaluated. In second-line therapy there was also a nonsignificant trend favoring culture-guided treatment.
No comparative studies of empirical versus susceptibility-guided treatment after 2 treatment failures have been performed. Nor do the results of the observational studies reporting culture-guided third-line therapies show especially good results. Overall, then, there is little evidence in favor of routine culture and susceptibility-guided treatment in H pylori treatment.
Although it is thought that PPI use reduces culture efficacy by decreasing bacterial density, few data have been reported. In a recent study, Siavoshi and colleagues evaluated the results of H pylori tests in 530 dyspeptic patients and reported that the rate of positive culture was 30% in patients receiving PPI versus 56% in those not receiving the drugs. It is unclear whether sampling an additional corpus biopsy for culture might increase culture sensitivity, as it has been shown to increase histology sensitivity.
Rapid urease test
RUT is an indirect test based on the activity of the H pylori urease enzyme, which degrades urea to form ammonia. This situation increases the pH level, which can be detected by a pH indicator. It requires a sample of gastric mucosa, which is introduced in the mixture of urea and the pH indicator. There are many different commercial RUT kits, including gel-based tests (CLOtest, Kimberly-Clark Ballard Medical Products, Roswell, GA, HpFast, GI supply, Camp Hill, Pennsylvania) paper-based tests (PyloriTek, Serim Laboratories, Elkhart, IN, ProntoDry HpOne, Medical Instruments Corporation, Solothurn, Switzerland), and liquid-based tests (CPtest, Yamanouchi, Milan, Italy, EndoscHp, Cambridge Life Sciences, UK).
Commercial RUTs have a sensitivity of 80% to 95% and specificity higher than 95% to 100%. These figures are slightly lower than those for histology but are still acceptable for clinical practice. The number of bacteria present in the biopsy is the main cause of the reduction in sensitivity. It is estimated that densities lower than 10 4 to 10 5 organisms may result in false-negative tests. This situation may often occur in bleeding patients or after treatment. In these situations, RUT is not indicated as a sole test. Extensive atrophy or intestinal metaplasia may also be associated with low density and false-negative RUT results. In addition, any treatment that reduces bacterial density, such as the use of antibiotics, bismuth compounds, or PPI, may result in false-negative tests. Reviews recommend obtaining 2 samples (1 from the antrum and 1 from normal-appearing corpus) and avoiding areas of ulceration and obvious intestinal metaplasia to obtain optimal results. Combining the 2 samples in the same tests seems to accelerate positive results. Whether or not the inclusion of a corpus sample is useful for increasing the yield of RUT remains to be determined. False-positive tests are less frequent but may occur if the sample is kept beyond 24 hours.
Overall, the test is cheap and rapid and provides adequate screening. Positive results should prompt H pylori treatment. By contrast, after negative RUT, a second confirmatory test may be considered depending on the importance of treating the infection, the degree of suspicion, and the presence of factors that may cause false-negative results as described earlier.
Molecular tests
Both in situ hybridization and PCR are sensitive tests and can be used to detect H pylori in biopsies. Furthermore, these tests can determine antibiotic resistances (see the article of Francis Megraud in this issue).
There is growing evidence that, even when performed in optimal conditions, the sensitivity of conventional tests for detecting H pylori infection may be suboptimal. If this situation occurs, it is most probably caused by low-density occult infection, and many studies have shown that molecular methods can detect at least a part of these occult or low-density cases. Bik and colleagues characterized the gastric microbiota of H pylori –positive and H pylori –negative individuals by using broad-range PCR and 16S ribosomal DNA sequence analysis in 23 healthy volunteers. H pylori infection was detected in 12 individuals by conventional tests and in 19 of the 23 by molecular tests. In another important article, Raderer and colleagues reported 6 patients with gastric MALT lymphoma with negative results for all H pylori conventional tests who presented with complete resolution of the disease after H pylori treatment, thus suggesting that the patients had an undetectable active infection. Weiss and colleagues evaluated 60 gastric biopsies from patients with dyspepsia, performing immunohistochemistry, RUT, and a multiplex PCR. PCR consistently identified H pylori in more patients, including those positive by the other methods. The same group claimed that multiplex PCR was able to detect the infection in patients on PPI therapy even when RUT was negative.
However, even molecular tests may not be sensitive enough to detect all cases of low-density infection. Guell and colleagues reported that 79% of the patients with peptic ulcer bleeding who tested negative for H pylori during the bleeding episode had active infection, which was detected when the tests were repeated a few weeks after the acute bleeding episode. Real-time PCR performed in the biopsy obtained during the bleeding episode detected 64% of all the infected patients.
Some investigators have suggested that the molecular test should be used as a gold standard for diagnosis of infection, because of its superior sensitivity. However, PCR tests are prone to false-positive results. There are many examples of positive PCR tests (eg, from saliva or oral samples) in patients with no evidence of H pylori using any other test, suggesting that cross-reacting DNA from as yet uncultured organisms may be a significant obstacle to the use of PCR-based tests as a gold standard. Therefore, a single PCR should not be used as a gold standard. Reliable demonstration of H pylori infection by PCR may require the positivity of more than 1 gene.
PCR has not been adopted as a routine test because of its many drawbacks. It is technically demanding and expensive compared with conventional tests; it requires special laboratory conditions, and it is prone to false-positive results as a result of contamination. In addition, these tests may be inappropriate for determining cure after treatment (it has been suggested that residual death organisms of DNA fragments might cause false-positive results) and have not been evaluated in this setting.
PCR in feces may be an attractive alternative for noninvasive evaluation of not only the presence of infection but also the bacterial resistances to clarithromycin. However, some studies have found its sensitivity to be low, mainly because of the presence of PCR inhibitors.
Noninvasive Tests
Urea breath test
H pylori urease splits urea into ammonia and carbon dioxide. In the UBT, patients ingest urea labeled with either 13 C or 14 C. Although the 14 C test is safe, cheap, and reliable, nonradioactive 13 C is generally preferred.
The test mechanism involves urease, an enzyme that is not present in mammals. By contrast, H pylori has a strong urease activity. An individual infected with H pylori presents urease activity in the stomach. Hydrolysis of urea occurs, causing the production of labeled CO 2 , which diffuses into the blood vessels and appears in the subject’s breath within a few minutes. UBT is probably the most reliable noninvasive test, with reported sensitivity and specificity of more than 95%.
However, the reliability of UBT is affected by many factors. As in the case of most of the tests detecting active infection, its sensitivity is markedly decreased by recent antibiotic use. In addition, PPI use reduces the sensitivity of the test by 30% and histamine 2 receptor antagonist use by approximately 10% unless the test used citric acid. Technical characteristics may also affect the reliability of the test; it has been shown that the use of a test meal including citric acid enhances the activity of H pylori urease. This enhancing of urease activity may help to increase sensitivity in patients with low-density infection caused by PPI use.
In addition, false-positive results have been attributed to the activity of urease-producing bacteria other than Helicobacter , such as Proteus mirabilis or Staphylococcus aureus , which may colonize gastric lumen in situations of achlorhydria. In this case, acidification of the media by citric acid might help to bring down the number of false-positive results by reducing the density of the contaminating bacteria.
However, there are many commercial tests that skip the use of the citric acid pretreatment. Many recent reports have reported an unacceptably high rate of false-positive results with these tests. This finding is especially relevant in terms of posttreatment evaluation, in which a positive UBT must be followed by a second expensive and burdensome antibiotic treatment and then another test to confirm the cure of the infection.
Stool antigen tests
The stool test detects bacterial antigens of H pylori that are present in stool in infected patients. It has been shown that these antigens disappear when H pylori infection is cured. Because each commercial stool antigen test (SAT) detects different antigens, the results change from one test to another. Furthermore, because of the marked genetic variability of H pylori , the accuracy of the same test may vary according to population. For this reason, the Maastricht consensus recommends local validation before using a given SAT. SAT may use polyclonal or monoclonal antibodies. Polyclonal antibodies are obtained by immunizing laboratory animals. This method produces extremely variable sets of antibodies, which differ from batch to batch. This finding may explain why the excellent results obtained in the first SAT validation studies have not been reproduced in later research. Meta-analyses have shown that SAT using polyclonal antibodies are consistently inferior to those using monoclonal antibodies. Regarding the test method, in the office, immunochromatographic tests have generally produced slightly worse results than laboratory enzyme-linked immunosorbent assay (ELISA) tests.
A major problem with the use of SAT is that many commercial tests are available but have never been validated; in addition, in some areas, sensitivity and specificity of even the best monoclonal ELISA SAT barely reaches 90%. Many recent tests did not even achieve these minimal rates for sensitivity or specificity.
Although there is no consensus on the minimal values that a diagnostic test should achieve, a sensitivity of less than 90% seems inadequate for the diagnosis of the infection, and a specificity less than 90% may generate many false-positive results, thus rendering the test inadequate for posttreatment evaluation.
PPI use may also cause false-negative results with SAT. In this case, little is known about possible corrective measures. SAT may be preferable in patients who have very recently initiated PPI therapy; Gisbert and colleagues reported a high sensitivity of monoclonal ELISA tests in SAT performed immediately after admission for upper gastrointestinal bleeding.
Related posts:
Is There a Role for Probiotics in Helicobacter pyloriTherapy?
Gastric Cancer Risk in Patients with Helicobacter pylori Infection and Following Its Eradication
Treatment Strategy for Gastric Mucosa-Associated Lymphoid Tissue Lymphoma
Helicobacter pylori
When Is Endoscopic Follow-up Appropriate After Helicobacter pyloriEradication Therapy?
Screening to Identify and Eradicate Helicobacter pyloriInfection in Teenagers in Japan
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
