This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Introduction
Helicobacter pylori is responsible for gastritis and peptic ulcers; moreover, it is one of the most studied causal agents of gastric cancer (GC) in the last years (Misra et al., 2014), for which in 1994 it was considered as group I carcinogen by the International Agency for Research on Cancer (IARC, 2012). Infection is frequent during childhood and sometimes induces superficial gastritis, which can progress to atrophic gastritis, intestinal metaplasia, dysplasia, and finally GC (Philippe et al., 2016). The gram-negative bacterium adheres and colonizes the gastric mucosa, with the participation of several virulence factors, including cytotoxin-associated gene A antigen (CagA) and vacuolating cytotoxin (VacA), as well as: induced by contact with epithelium (IceA), blood group antigen-binding adhesion (BabA), sialic acid-binding adhesion (SabA), duodenal ulcer-promoting gene (DupA), and outer inflammatory protein (OipA) (Cadamuro, 2014).
In South America, and particularly in Chile, more than 70% of population is positive for H. pylori (Coelho and Coelho, 2014), a rate that has been significantly stable during the last 10 years. Different studies have shown a prevalence of infection ranging from 60% to 79%, according to socio-economic, educational and health conditions of the population studied (Ministerio de Salud, 2013). Chile has one of the highest rates of H. pylori infection in the world (Ferreccio et al., 2007; Porras et al., 2014), making it necessary to develop a fast, reliable and non-invasive method to detect the pathogen, before the infected patient develops any gastric pathology, including cancer.
Currently there are two basic genres of tests to detect infection by H. pylori: invasive and non-invasive. Invasive tests including culture, histology and the rapid urease test (RUT) (Hunt et al., 2011; Lee et al., 2013; Ministerio de Salud, 2013) are inconvenient, costly, and uncomfortable because a patient is required to go to a hospital or clinic for an endoscopic gastric biopsy. Of the non-invasive tests, the Urea Breath Test (C13-UBT) and the fecal antigen analysis stand out as the most valuable (Tamadon et al., 2013). C13-UBT is a fast and simple method that detects the presence of H. pylori in the gastric mucosa through urease activity of the pathogen (Di Rienzo et al., 2013). Moreover, the test shows high sensitivity and specificity, with sensitivity between 81–100%, and specificity between 80–98% (Honar et al., 2016), but the high initial investment of a isotope ratio mass spectrometer for obtaining results from C13-UBT is not feasible in most public health centers in developing countries. Furthermore, the use of antisecretory drugs or antibiotics can influence the results of the test (Di Rienzo et al., 2013).
Tests that detect H. pylori antigens in stool samples show high levels of specificity and sensitivity, similar to those for UBT (Dore et al., 2016), and lately the costs have become more practical for the population. However, the accuracy of these tests decreases when the stools are aqueous because H. pylori antigens become diluted. These methods are also not recommended for patients with gastric ulcers (Shimoyama, 2013).
In this context, the aim of the present study was to show and evaluate the efficacy of a new non-invasive diagnostic method based on nested-qPCR, using ureC as a gene marker to detect H. pylori in stools samples, even in patients with gastric ulcers or watery stools.
Experimental
Materials and Methods
Patients. For this study 143 patients with digestive symptoms were considered: 103 females and 40 males, who were attended to by the Hospital San Pablo, Coquimbo, Chile for a routine gastrointestinal endoscopy. The mean (±SD) age of our patients was 56.5 years (± 14.18). The inclusion criteria of the patients were as follows: adults over 18 years old, with digestive symptomatology, who had been tested for RUT and histology with Giemsa stain. The Bioethical Committee of the Health Service of Coquimbo, Chile, approved the protocol and patients voluntarily signed their consent.
A patent was requested for this protocol, with the N° 2016–01214 in INAPI (National Institute of Industrial Property, Chile).
Endoscopy and biopsy samples. The endoscopic procedure was performed in Hospital San Pablo, Coquimbo, Chile. Gastric biopsy samples were obtained from each patient for RUT and histology by Giemsa stain analyses, which were processed in the Service of Pathological Anatomy of the Hospital San Pablo according to standard protocols. The same pathologist performed the analysis of all biopsy samples.
Stool samples. Stools by normal evacuation were obtained from each patient before the endoscopic procedure. Each patient provided ~ 5 g of stools placed in a flask containing 3 ml of RNA Later® (Ambion), which were stored in a deep freezer (–80°C) until analysis.
DNA purification and PCR amplification. Approximately 200 mg of each stool sample was used to extract DNA, using QIAamp® Fast DNA Stool Mini Kit (QIAGEN) according to the manufacturer’s protocol. Later, 120 ng of the extracted DNA was used to amplify the H. pylori ureC gene by nested-qPCR. DNA concentration was quantified by NanoDropTM One (Thermo ScientificTM). In brief, a first amplification with conventional PCR was performed in a Axygen® MaxyGene Thermal Cycler II, incubating 120 ng of DNA with 5 μl Buffer 5x; 1.5 μl MgCl2 25 mM; 0.5 μl dNTPs 10 mM; 1 μl of each external primer (10 μM each), and 0.2 μl of Platinum Taq® DNA polymerase (5U/μl) (Invitrogen), in a final volume of 25 μl. Amplification conditions were as follows: a pre-denaturation of 95°C for 5 min, 25 cycles of 95°C for 45 s, 57°C for 45 s and 72°C for 45 s, followed by a final extension at 72°C for 10 min. Posteriorly, 2 μl of a 10 × dilution of this first PCR were used for subsequent amplification by qPCR in an Eco Real Time PCR (Illumina®). The qPCR mix contained 5 μl of SYBR Green kit 2x (KAPA SYBR® FAST qPCR) and 0.1 μl of each internal primer (10 μM each), in a final volume of 10 μl. Amplification conditions were: pre-denaturation at 95°C for 5 min, and 30 cycles of 95°C for 10 s and 60°C for 30 s. Sequences of primers are shown in Table I.
Primers used for qPCR assay.
Data analysis. Data was analyzed using the Software Eco v4.1 PCR System and the program XLSTAT Version 2.06 to calculate sensitivity, specificity, and positive and negative ratio probability for each of the three techniques: nested-qPCR, RUT and histology by Giemsa stain for detection of H. pylori infection in symptomatic digestive patients. Cases for each technique were considered to be H. pylori infected according to a combined gold standard of RUT/histology, RUT/qPCR or histology/qPCR, as applicable. P values < 0.05 were considered significant.
Results
A total of 143 patients with digestive symptoms were evaluated by endoscopy, and the presence of H. pylori was evaluated by a novel method of nested-qPCR, and corroborated with RUT and histology by Giemsa stain. We use a combined gold standard for each test evaluated. A patient was considered positive or negative for H. pylori when both of tests used as gold standard gave the same result for the infection. Results of the three tests are shown in Table II.
Results of RUT, Histology and Real Time-PCR, for detect H. pylori in patients.
RUT, the standard method used in medical practice to detect the presence of this bacterium, only detected infection in 71/134 patients (53.0%), with a sensitivity of 67.0% (95% C.I.: 57.2 to 75.8) and a specificity of 92.3% (95% C.I.: 76.5% to 99.1); in turn, PPV was 97.3% (95% C.I.: 90.3 to 99.3) and NPV was 42.6% (95% C.I.: 35.7 to 49.8).
Histology with Giemsa stain classified 71/101 patients (70.3%) as H. pylori positive. Between these patients, 92 had a histologic diagnosis associated with gastrointestinal disease, and the remaining 9 tested normal for histology. Sensitivity for this test was 98.6% (95% C.I.: 92.5 to 100.0) and specificity 89.7% (95% C.I.: 72.7 to 97.8); PPV was 96.0% (95% C.I.: 89.0 to 98.6) and NPV was 96.3% (95% C.I.: 78.7 to 99.5).
Using a nested-qPCR approach, we identified infection of H. pylori in 71/102 patients (69.6%), with a sensitivity to detect the presence of the bacterium in stools of 100% (95% C.I.: 94.9–100.0), and specificity of 83.9% (95% C.I.: 66.3 to 94.6); Finally, PPV and NPV were 93.4% (95% C.I.: 86.4 to 96.9) and 100.0% (95% C.I.: 84.0–100.0), respectively. A 2.5% agarose gel electrophoresis stained with ethidium bromide with the amplification of five positive samples for H. pylori is showed in Figure 1. Complete results are shown in Tables III–V. Comparative results between sensitivity, specificity, PPV and NPV for each test are shown in Table VI.
Fig. 1.
2.5% agarose gel electrophoresis stained with ethidium bromide, showing the amplification of five fecal samples positives for H. pylori.
Comparative evaluation of RUT, histology and nested qPCR for detection of H. pylori.
Discussion
In this work, we evaluated a combined method of nested-qPCR for detection of infection by H. pylori in the stools of 143 gastrointestinal symptomatic patients, and demonstrated that this technique is superior to RUT, the invasive test commonly used in clinical practice today.
One of the pioneering works in this area was conducted in 1994, in which stools of 24 patients diagnosed with H. pylori infection were analyzed by PCR. Half of the patients had gastric ulcers at the time of endoscopy, while the other 12 had only related dyspepsia. Unfortunately, this study was unable to show that PCR technique is helpful to diagnose infection by the pathogen in the stools. Nonetheless, by inoculating each of the samples with 103 bacteria/mg of feces they received a positive result for all samples, a successful advancement in this technique (Mishra et al., 2008). In 1998, a new protocol to detect H. pylori by PCR was tested on 100 patients (63 diagnosed with H. pylori and 37 healthy for the pathogen). This technique identified 59 infected patients (sensitivity 93.7%), while all uninfected patients tested negative by PCR (specificity 100%) (Aktepe et al., 2011). Moving forward, several studies have shown that this technique may be a useful clinical alternative for H. pylori detection in stool samples. In this regard, a study from India in 2008 used nested-PCR of feces to demonstrate the prevalence of infection in the population of that country, finding that of 245 patients evaluated 105 were detected as positive for infection, using a new model for clinical evaluation (Momtaz et al., 2012). Alas, this study did not corroborate their findings with currently accepted techniques, such as RUT or histology. However, this same group later showed that this technique has a high sensitivity, finding 40/52 patients positive for H. pylori according by RUT and biopsy, with a sensitivity of 72.5% (Smith et al., 2012).
Another study applied the stool PCR test to 300 patients, 271 of them positive for H. pylori by RUT, finding 167/300 positive by PCR (61.6%), using the ureC gene as a marker (Uno et al., 2016). Liu and coworkers (2016) evaluated stool samples of 97 digestive symptomatic patients by PCR, with the ureC marker and compared their results with UBT. In this work, sensitivity was 42.6% and specificity was 100% (Liu et al., 2016). The authors claim that despite the observed low sensitivity, this technique could be useful for diagnosis in children, especially in health centers that do not have pediatric endoscopes.
In 2014 Patel et al. presented a review suggesting that PCR could be superior to other diagnostic tests for detection of H. pylori infection, owing to higher sensitivity and specificity, especially with nested and semi-nested approaches (Patel et al., 2014).
In this work, we found that our nested-qPCR is more effective than RUT and similar to histology by Giemsa stain in detecting the presence of infection by H. pylori in the patients with digestive symptoms, with a sensitivity of 100% and a specificity of 83.9%. PPV and NPV values were 93.4% (95% C.I.: 86.4 to 96.9) and 100.0% (95% C.I.: 84.0–100.0), respectively.
It is important to note that specificity is close to 84%, as our technique detected five fecal samples as positive for H. pylori, which were not detected by RUT or histology. We repeated the test on the samples in question three more times with different portions of the fecal samples, and in all of them the result was positive for H. pylori. We hypothesized that our technique had the potential to detect the presence of small quantities of nucleotides from H. pylori beyond the limit of detection of the compared techniques.
Moreover, the nested-qPCR method is non-invasive and the patient needs only to send a stool sample to the laboratory, eliminating the need to go to the hospital. Currently RUT, with its low sensitivity, is the standard test in the medical practice. Indeed, in our patients the sensitivity of RUT was only of 67.0%, a result likely associated with ulcer bleeding or the use of proton pump inhibitors, which can give RUT a false negative (Coelho and Coelho, 2014). This is not an issue for PCR based diagnostics.
The proposal that blood could affect the sensitivity of RUT, by the presence of albumin acting as a buffer on the pH indicator of the reaction, is controversial, because other studies report that blood has no effect on the test (Honar et al., 2016).
Our results are encouraging because this technique could soon become a non-invasive method for detection of H. pylori in stools, providing the population with an inexpensive and sensitive method to observe presence of the bacterium.
We are waiting for the approval of our patent request Nº 2016-01214 for this protocol in INAPI.
