Evaluation of Quadruple Real-Time PCR Method to Detect Enterococci Carrying Vancomycin-Resistant Genes vanA, vanB, vanM in Rectal Swabs

Purpose: To evaluate the sensitivity and accuracy of the quadruple real-time PCR method for the detection of enterococci carrying vancomycin-resistant genes vanA, vanB, and vanM in rectal swabs. Methods: Choosing PCR-sequenced DNA extracted directly from rectal swabs as the gold standard, the results of the quadruple real-time PCR method and the traditional method (screening culture combined PCR-sequencing method whose DNA extracted from single colony) were compared with the gold standard. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the quadruple real-time PCR method and the traditional method were obtained. The time required for the three methods was calculated. Results: The results between gold standard and the quadruple real-time PCR method were similar. Compared to the traditional method, the quadruple real-time PCR method had a much higher sensitivity, specificity, PPV, NPV, and consistency. Our study found that the quadruple real-time PCR method is beneficial for detection of enterococci carrying vanM with vancomycin heteroresistance. The traditional method had high specificity and NPV, but its sensitivity and PPV were not ideal. The time needed for gold standard is a minimum of 28 h; the quadruple real-time PCR method takes 2–3 h while the traditional method consumes a minimum of 72 h. Conclusion: The quadruple real-time PCR method can provide a rapid and reliable result for the diagnosis of patients with colonized vancomycin-resistant enterococci. This new method is beneficial for the active screening, timely clinical treatment measures, epidemiological research, and hospital monitoring of the enterococci carrying vancomycin-resistant gene, especially for the enterococci with vancomycin heteroresistance carrying vanM.


INTRODUCTION
Vancomycin-resistant enterococci (VRE) infection is associated with increased treatment costs and prolonged hospitalization (1,2). In China, the genotypes of VRE mainly consist of vanA, vanB, and vanM. Active screening, which mainly consists of the traditional culture method and the molecular detection method at present, can reduce the incidents of VRE infections. Recently, a new method was developed to simultaneously detect eight genes including vanA, vanB, and vanM genes in strain analysis (3). At the strain analytic stage, a highly accurate performance of the quadruple real-time PCR method was achieved (4). The reproducibility of the quadruple real-time PCR kit is ideal (variation coefficient <5%) and the limitations of detecting vanA, vanB, and vanM are all 100 CFU/mL (data not published). The objective of this study is to evaluate the performance of this novel multiple real time PCR assay in rectal swabs. To our knowledge, this is the first report to evaluate the quadruple realtime PCR method for enterococci carrying vancomycin-resistant genes vanA, vanB, and vanM simultaneously in clinical samples.

METHODS AND MATERIALS Patient Specimens
A total of 403 clinical rectal swabs transported by MW170 transwab (Medical Wire & Equipment, Corsham, Wales, England) from 403 different patients were submitted to the laboratory for genetic screening of vancomycin-resistant enterococci. They were tested to determine the clinical sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the quadruple real-time PCR method. Participants had to meet the following requirements to take part in this study: patient groups with a high risk of VRE colonization, especially for haemato-oncology and transplant patients and immune suppressed and critically ill patients in the ICU, by virtue of their intrinsic compromised immunity, antibiotic exposure, and their potential exposure to colonized patients (5,6). The Institutional Review Board of Peking University First Hospital approved this study (2020 Research 202). Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

Control Group Setting
ATCC29212, BM4147, ZB212, and ZB17 were all Enterococcus. They are preserved by the Institute of Clinical Pharmacology of Peking University and their species and genotypes have been identified. BM4147, ZB212, and ZB17 carry vanA, vanM, and vanB genes, respectively. ATCC29212 carries none of the three genes. Every strain to be tested was recovered and purified before the experiment to ensure the viability and purity of the bacteria. Their DNA were extracted and used in the three methods as the control group. ATCC29212, BM4147, ZB212, ZB17, and the nuclease-free water were chosen as the negative control, vanA positive control, vanM positive control, vanB positive control, and blank control, respectively.

Traditional Method
The screening culture was performed by streaking a rectal swab in trilinear method onto an esculin agar medium containing vancomycin (6 µg/mL) followed by incubation at 37 • C for 24 or 48 h. Suspicious single colonies were transferred to blood plates containing vancomycin (6 µg/mL) for 24-48 h for purification culture, and was identified by API 32E tests. The extracted DNA was amplified by primers and then sequenced to identify their species and genotypes. After application of agar medium in the traditional method, the rectal swab was placed in an EP tube containing 2 mL saline solution and was mixed by vortexing. One milliliter bacterial suspension was taken for DNA extraction for both the gold standard and investigational assay, the remaining bacterial suspension was stored at 4 • C for further analysis when the results of the three methods were discordant. The experimental procedure is shown in Figure 1.

DNA Extraction
The DNA used in both the gold method and quadruple real-time PCR method was directly extracted from the rectal swabs using DNA extraction Kit (Bacterial Genome DNA Extraction Kit; TIANGEN, Catalog no. DP 302) according to the manufacturer's guidelines (Bacterial Genome DNA Extraction Handbook DP190814). Briefly, 1 mL of saline solution was taken from the EP tube containing an anal swab; 20 µL of internal control (2010806, IPE, Beijing, China) was added, mixed, and centrifuged at 10,000 rpm (∼11,500 × g) for 1 min and the supernatant was discarded. Liquid amies buffer GA was added and mixed until homogenous suspension was achieved. The proteinase K and the liquid amies buffer GB were subsequently added to the suspension and incubated at 70 • C for 10 min. Subsequently, 100% ethanol was added followed by centrifugation. DNA was then isolated following a series of centrifugations with TIANGEN spin columns and buffer solutions. DNA was eluted in 200 µL elution buffer (Buffer TE); quantity and quality was assessed using an enzymelabeling measuring instrument. Extracted DNA was stored at −20 • C for subsequent PCR and quadruple real-time PCR. The DNA extraction in the traditional method: A single colony was transferred into 200 µL saline solution and was boiled for 10 min at 100 • C. After boiling, the liquid containing bacteria was centrifuged for 30 s with 12,000 r/s. Extracted DNA was stored at −20 • C for subsequent PCR in the traditional method. The DNA extraction for control strains was the same as the traditional method.

The Investigational Assay
The investigational assay procedure consisted of the DNA extraction and the quadruple real-time PCR method as shown in Figure 1. The reaction mixture of quadruple real-time PCR consists of the primers and probes of vanA, vanB, vanM, and the internal control gene. The quadruple real-time PCR method was performed according to the manufacturer's instructions (2010806, IPE, Beijing, China). The DNA in the reaction was extracted directly from rectal swabs. The primers and experimental procedures were described previously (4). The results were analyzed according to the instructions of the manufacturer (quadruple real-time PCR kit).

Discordant Result Analysis
Any specimen showing contradictory results by the traditional and gold standard methods would be re-cultured on the drug free medium and the species and genotypes would be identified by PCR sequencing. Any specimen that showed negative results with the quadruple real-time PCR method while positive in the gold standard would be tested again by the quadruple real-time PCR method and the result obtained was defined as the final result.

Vancomycin Susceptibility
Enterococcus carrying vancomycin resistance genes was tested for vancomycin susceptibility by the agar dilution method and E-test (AB Biodisk, Solna, Sweden) to determine the source of the genes. The Enterococcus strains were identified as VRE with a vancomycin minimal inhibitory concentration (MIC) ≥32 µg/mL. The methods were all according to Clinical and Laboratory Standards Institute (CLSI) guidelines updated in 2018 (7).

The Time Required for the Three Methods
The details of the time for the three methods: For the gold method and quadruple real-time PCR method: the starting point was set to the moment when DNA extraction from the rectal swab began, while the ending point were all set to the moment when the result was obtained. For the traditional method: the starting point was set to the moment when culture was started, while the ending point was set to the moment the result was obtained.

Statistical Analysis
Fisher's exact test was used to calculate the sensitivity, specificity, PPV, NPV, and their 95% confidence intervals by GraphPad Prism 5.0 software (GraphPad Software, Inc., La Jolla, CA). The diagnostic accuracy and its 95% exact confidence intervals were calculated by Stata 9.0 (Stata Corp., College Station, TX, USA). The consistency was analyzed through the Kappa test method by SPSS 20.0 (IBM Corp., Armonk, NY, USA): the consistency showing statistical significance.

RESULTS AND DISCUSSION
In this study, we developed a new method to detect the colonization of the enterococci carrying vancomycin-resistant gene. We tested 403 rectal swabs from 403 patients by the gold standard, the quadruple real-time PCR method and traditional method.
Before the resolution of discordant results, a total of 19 rectal swabs showed differences among the three methods. These 19 rectal swabs were positive (carrying vancomycin resistance genes) by gold standard but were negative by the traditional method. One of the 19 rectal swabs (rectal swab 1-62) was vanA positive by the gold standard but was negative by the quadruple real-time PCR method. They were all tested again and analyzed for further resolution. Finally, two strains of vancomycin heterogeneous resistant Enterococcus carrying vanM (Figure 2) were obtained.
After resolution of discordant results, a total of 49 strains of Enterococcus containing vancomycin resistance genes were obtained by the traditional method. Forty-seven of those strains were VRE and two were vancomycin heterogeneous resistant Enterococcus carrying vanM. The VRE strains included 39 strains with vanA and 8 strains with vanM. These strains were obtained from the medium containing 6 µg /mL vancomycin, and their vancomycin MIC were all ≥32 µg/mL. The vancomycin heterogeneous resistant Enterococcus were obtained from the drug-free medium and their vancomycin MIC were 256 µg /mL by E-test.
The results of the quadruple real-time PCR method were very similar to those of the gold standard. Except in one case, the rectal swab 1-62 was positive for vanA by the gold standard while negative by the quadruple real-time PCR method (this sample was also negative by the traditional method). None of the three methods detected the Enterococcus containing vanB gene. Partial electrophoresis results of the vanB gene are shown in Figure 3. More details about the number of samples containing vancomycin resistance genes are shown in Table 1.
When we tested whether the samples only contained the vanA gene, 39 positive samples were obtained by the traditional method. While detection by the quadruple real-time PCR method and the gold standard revealed that 25 rectal swabs were positive, while the other 14 rectal swabs all contained both vanA and vanM genes, which were defined as negative.
When the samples were tested for the presence of the vanM gene alone, 10 positive samples were obtained by the traditional method. The same samples when detected by the quadruple realtime PCR method and the gold standard, obtained 4 positive and 6 negative samples, respectively. In the negative samples, 5  samples contained both vanA and vanM, while one sample (rectal swab 1-138) had no vanM detected by both the methods. This may be due to the extremely limited amount of vanM present in that sample (1-138). After it was used in the traditional method, the amount of bacteria had reduced and could not reach the detection limit of the quadruple real-time PCR method and the gold standard.
When tested whether both vanA and vanM simultaneously existed in the same rectal swab, 25 samples were all positive by both the gold standard and the quadruple real-time PCR methods. The results obtained from the same samples by the traditional method revealed that 14 samples contained vanA only, 5 samples contained vanM only, and 6 samples didn't contain any vancomycin-resistant genes. A further resolution of discordant results showed that 6 rectal swabs did not contain the Enterococcus carrying vancomycin resistance genes after the re-culture and purification in the drug-free medium. However, one vancomycin heterogeneous resistant Enterococcus carrying vanM was isolated from rectal swab 2-024 by the same procedure. It is undeniable that the quadruple real-time PCR method has a higher sensitivity and specificity than the culture method. Only part of the samples' target strains could be isolated in the discordant results during resolution steps which might be because of inactivation of enterococci due to the long storage time of the rectal swabs.
Other reasons might be as follows: (i) Some Enterococcus strains containing both vanA and vanM were detected by the gold standard and quadruple real-time PCR method, but they could not be isolated by the traditional method. In the strain analysis step, the quadruple real-time PCR method can accurately detect 20 enterococci strains that contain both vanA and vanM, with a 100% accuracy (4). In this study, although, the Enterococcus containing both vanA and vanM were not isolated by the culture method, it is undeniable that rectal swabs may contain Enterococcus harboring both vanA and vanM when the results were vanA and vanM positive by the gold standard and quadruple real-time PCR methods.
(ii) vanA and vanM genes exist in different Enterococcus in the same rectal swab, respectively. However, the traditional method failed to isolate both genotypes of different Enterococcus in the same rectal swab. In subsequent studies, ZB366 (vanM) & ZB410 (vanA), ZB317 (vanA) & ZB407 (vanM), and ZB329 (vanA) & ZB36 (vanM) were isolated from anal swabs taken from three patients at two different time points in the week (data not included in the study). However, traditional methods have failed to isolate Enterococcus containing different vancomycin resistance genes in the same swab, but the result does not rule out that different Enterococcus containing vanA and vanM genes separately may exist in the same patient at the same time. The reasons for the different vancomycin-resistant Enterococcus genes isolated from the rectal swabs collected at different time points from the same patient are still to be fully explored, which may be due to human factors having a greater impact on the results obtained by traditional methods compared to the other two methods or the different proportion of Enterococcus with different vancomycin resistance genes at different time points.
(iii) vanA or vanM genes were found in other strains but not from Enterococcus in some samples. However, this is less likely as the vancomycin resistance genes were not detected in 90 non-enterococci strains in the strain analysis step (4).
Compared to the traditional method, the quadruple real-time PCR method is confirmed to have a much higher sensitivity, specificity, PPV, NPV, and accuracy. When the vanA, vanB, and vanM genes were detected separately, the kappa coefficients between the gold standard and quadruple real-time PCR method were all nearly equal to 1. The traditional method has high specificity and NPV, but its sensitivity and PPV are not so ideal; more details are shown in Table 2.
The time needed for the gold standard (PCR-sequencing whose DNA was directly extracted from the swabs) is 28 h   For the three methods, the average extraction time of DNA from a sample was 30 min. The gold standard method includes DNA extraction, PCR, DNA cleavage agarose gel electrophoresis, and sequencing for the last identification. The quadruple realtime PCR method includes DNA extraction and quadruple realtime PCR. The time cost of PCR-sequencing and quadruple real-time PCR were as described previously [3]. The traditional method includes screening the purification of the culture (24 h at least), DNA extraction, and identification of genotype and species by PCR-sequencing.
In some studies, associated with the evaluation of molecular methods to detect VRE in clinical samples, vanB is the most common in their positive results. However, the amount decreased significantly after culture reconfirmation (8)(9)(10). The reasons for this are speculated as follows: (i) The vanB gene may exist in non-Enterococcus (8,11). (ii) The specificity of some methods needs to improve. The VanB gene was not detected by any of the three methods probably due to the time and place of swab collection being relatively limited. Although the evaluation of the quadruple real-time PCR method on vanB was not obtained in the clinical analysis stage, it can accurately distinguish the vanB gene from other genotypes in the strain analysis stage (4). It demonstrated that our quadruple real-time PCR method may have a high specificity in vanB gene detection in clinical samples.
Unlike vanM VRE with high resistance to vancomycin which appeared in previous studies (12)(13)(14), enterococci carrying vanM with vancomycin heteroresistance has rarely been reported all over the world. The vancomycin heteroresistance enterococci carrying vanM is important due to its great clinical value. Conventional antibiotic susceptibility tests may miss detecting them, and subsequently mistake vancomycin sensitive strains leading to vancomycin treatment failures impairing hospital infection prevention and control. This further illustrates the importance of the detection of vanM gene by this quadruple real-time PCR method.
The advantages of the quadruple real-time PCR method are as follows: (i) Compared to the tedious traditional culture method, this new method provides results in only ∼2-3 h. (ii) avoids the misdiagnosis of vancomycin heteroresistance enterococci carrying vanM by traditional culture (15).
Our study has several advantages as follows: (i) Screening culture medium containing vancomycin 6 µg/mL, which is more sensitive compared to one medium with 32 µg/mL vancomycin (16). (ii) Although the swab causes discomfort to patients, it can be obtained immediately. (iii) The usage of rectal swabs can reduce the repetition of experiments compared to fecal samples, as some substances in fecal matter can inhibit the quadruple realtime PCR reaction (10). (iv) The traditional method for screening VRE has a more accurate result when cultured in 48 h than in 24 h (16). (v) The adoption of internal control for screening the extraction of DNA and the whole process of PCR amplification, which is necessary for the judgement of the false negative results caused by the action of the inhibitor (17,18). (vi) The evaluation of these new methods is more objective as we chose two different detection methods as the reference methods. (vii) We re-cultured the clinical samples on the antibiotic-free medium to further analyze inconsistent results. This improves the accuracy of the traditional method.
However, as the rectal swabs had been stored for a long time when we re-cultured on the drug-free medium, some vanM vancomycin heterogeneous resistant enterococci may have been inactivated, reducing the accuracy of the traditional method.

CONCLUSION
In summary, the quadruple real-time PCR method is ideal for obtaining the results quickly and differentiating vanA, vanB, and vanM accurately in rectal swabs. This method is of great clinical significance for the detection of vancomycin heteroresistance Enterococcus carrying vanM, and it is beneficial for active screening, taking timely clinical treatment measures, epidemiological research, and hospital monitoring of enterococci carrying vancomycin-resistant genes (19).

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/supplementary material.

ETHICS STATEMENT
The Institutional Review Board of Peking University First Hospital approved this study (2020 Research 202). Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
YH, GR, HH, BX, and BZ conceived the study design. YH and GR were responsible for the recruitment and collection of samples. YH, GR, HH, and FX were responsible for the laboratory assays. YH and SZ performed the data analysis. YH completed the initial manuscript. YH, GR, and BZ revised the manuscript. All authors read and approved the final version of the manuscript.