記住我
ORIGINAL ARTICLE
Year : 2022 | Volume
: 11
| Issue : 4 | Page : 435-441
Evaluation of xpert Mycobacterium tuberculosis rifampicin for tuberculosis diagnosis in a reference laboratory
Naiara Cristina Ule Belotti1, Naiady Konno Madela1, Susilene Maria Tonelli Nardi1, Daniele Cristina Mariano1, Nilza Gomes de Souza2, Rosangela Siqueira Oliveira3, Heloisa da Silveira Paro Pedro1
1 Biomedical Sciences Center, Adolfo Lutz Institute, São José do Rio Preto, São Paulo, Brazil
2 Tuberculosis Department, Epidemiological Surveillance Group, São José do Rio Preto, São Paulo, Brazil
3 Tuberculosis and Mycobacteriosis Laboratory, Adolfo Lutz Institute, São Paulo, Brazil
Correspondence Address:
Heloisa da Silveira Paro Pedro
Biomedical Sciences Center, Adolfo Lutz Institute, São José do Rio Preto, São Paulo
Brazil
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/ijmy.ijmy_195_22
Background: The Xpert Mycobacterium tuberculosis Rifampicin (MTB-RIF) is a technological innovation that presents precision and speed in the diagnosis of tuberculosis (TB). The study aimed to evaluate the performance of Xpert MTB/RIF in the TB diagnosis and compare its results with those of culture, species identification, Antimicrobial Susceptibility Testing (AST), and rpoB gene sequencing of discordant results in AST, used for the diagnosis of TB in a reference laboratory. Methods: Cross-sectional descriptive study of pulmonary and extrapulmonary samples requesting Xpert MTB/RIF and culture for TB diagnosis from 2015 to 2019 at Adolfo Lutz Institute-São Paulo/Brazil. The analysis was performed with Epi-Info 7.2.1, presenting the distribution of frequencies and, for comparative analyses, Pearson's Chi-square test and Fisher's exact test were used, considering P ≤ 0.05 as statistically significant. For variables agreement, the Kappa method was used. Results: A total of 1575 pulmonary and extrapulmonary samples were analyzed using Xpert MTB/RIF and culture, of which 293 were positive for the MTB Complex in both methodologies with a sensitivity of 94.55%; specificity of 95.97%; accuracy 95.69%; positive predictive value 85.53%; negative predictive value 98.59%, substantial agreement by Kappa 0.87, and detection sensitivity even at lower levels of bacillary load (P < 0.05). The Xpert MTB/RIF and AST showed concordant results (P < 0.05). Conclusion: The study brings forward that the Xpert MTB/RIF shows substantial agreement with the results of culture and AST, contributing to the diagnosis of TB and the rapid detection of resistance. The sequencing of resistant cultures in Xpert MTB/RIF and sensitivity in AST identified the H526N mutation, characterized by a low level of resistance to RIF.
Keywords: Diagnosis, molecular biology, tuberculosis
Tuberculosis (TB) is a communicable disease caused by Mycobacterium TB (MTB) and is considered a serious public health problem. Until the coronavirus (COVID-19) pandemic, TB was the leading cause of death from a single infectious agent, ranking above HIV/AIDS. In 2020, the World Health Organization (WHO) estimated about 9.9 million new cases of TB worldwide, with an estimated 1.3 million deaths among HIV-negative people and a further 214.000 deaths among HIV-positive people.[1]
The disease is overwhelming in countries with high population density and large population in conditions of social vulnerability, as well as presenting a higher rate of infection in individuals from risk groups.[2] Efforts have been made so that TB surveillance is greater and the power, together with health agencies, outline strategies to prevent, combat and develop new sensitive, specific, and fast diagnostic techniques.[3],[4]
The commercial Xpert® MTB/RIF test (Cepheid, Sunnyvale, USA) detects the presence of MTB and resistance to Rifampicin (RIF) by means of a nested-polymerase chain reaction (PCR) in real time, using five probes complementary to a region of 81 bp (base pairs) of the rpoB gene. Sensitivity and specificity in detecting RIF resistance in clinical samples range from 96% to 98%, respectively.[5] RIF is a drug of the first choice in antibiotic therapy[6] and is considered a predictor of multidrug-resistant TB, because, in most of the cases there is also resistance to isoniazid.
This methodology was implemented in reference laboratories in Brazil in 2014 and is being faster and more sensitive than bacilloscopy for the diagnosis of TB.[7] The test is an automated system, which is based on the purification, concentration, and amplification of nucleic acids and identification of mutations in the rpoB gene, in the same equipment, with no need for human resources specialized in molecular biology[8] or special biosafety equipment.[9]
Conflicting and discordant results between genotypic and phenotypic methods have made it difficult to interpret drug susceptibility or resistance. These discrepancies can be elucidated by sequencing the genes involved in resistance.[10] To compare the results of the Xpert MTB/RIF with those of the Antimicrobial Susceptibility Testing (AST) and carry out the genetic sequencing of the isolates with discordant results in a strategic region of the State of São Paulo. Paulo allows improving the prevention and control of TB and resistant TB, aiming to better serve the population and the goals of the WHO to reduce the disease burden by 2035.[4],[11]
Thus, this study aims to evaluate the performance of Xpert MTB/RIF in the TB diagnosis and compare its results with those of culture, species identification, and AST used in the diagnosis of TB in a reference laboratory in the State of São Paulo.
MethodsThe research was carried out at the regional reference laboratory for TB from the state public network, which serves the region of São Paulo that comprises two Epidemiological Surveillance Groups (ESG-29 and ESG-30) composed of 102 municipalities that belong to the Regional Department of Health XV (DRS-XV).
The study is cross-sectional based on a diagnostic routine, with information from the database obtained from computerized record systems of the Mycobacteria Laboratory (Hospital Information and Management System-SIGH and Laboratory Environment Manager-GAL) of the Adolfo Lutz Institute-Regional Laboratory of São José do Rio Preto (CLR/IAL/SJRP), and the TB and Mycobacteriosis Nucleus of IAL Central, based on tests carried out between 2015 and 2019.
To verify the contribution of Xpert MTB/RIF in the diagnosis of TB in the Northwest region of the State of São Paulo, the period from 2015 to 2019 was evaluated, compared to the period before the implementation of the methodology, which was from 2010 to 2014.
Pulmonary and extrapulmonary samples with concomitant requests for Xpert MTB/RIF and Culture for MTB from the laboratory routine of mycobacteria were included, following standardized algorithms for a reference laboratory in the State of São Paulo.[12]
For the positive samples in the culture, the identification of the MTB Complex (MTC) was carried out by means of an immunochromatographic test (Ag MPT64) and AST. The sequencing of the rpoB gene fragment was performed in the cultures whose Xpert MTB/RIF results were discordant with the AST.
The variables of interest for the present study were: results of Xpert MTB/RIF laboratory tests detected/not detected and the respective detection levels (very low, low, medium, high, and very high), resistance to RIF detected/not detected, positive/negative MPT64 Ag test for the identification of MTC, AST sensitive/resistant to RIF/Isoniazid and sequencing of the rpoB gene fragment. For detection levels analysis, the categories very low, low, medium, and high were considered.
Samples that presented technical problems that made it impossible to perform the examination in liquid culture, positive cultures that did not undergo AST, and contaminated cultures were excluded from the study. For erroneous, invalid and indeterminate Xpert MTB/RIF results, tests were repeated and included when there was a sufficient clinical sample (minimum 2 mL), or excluded when the sample was insufficient for repeat testing.
For the Xpert MTB/RIF, the samples were prepared according to the manufacturer's instructions, and at the end of this preparation step, 2 mL of the sample was added to the cartridge, which was inserted into one of the compartments of the equipment and the software's automatic process was initiated for DNA extraction and real-time PCR.[13]
All samples were decontaminated by the Petroff Method. Cultures were performed in liquid medium using the Bactec MGIT 960 method (Becton and Dickinson-BD), or in Lowenstein–jensen solid medium (LJ) or in Ogawa Kudoh medium (OK), according to Brazil (2022). In positive cultures, the presence of acid-fast bacillus resistant was confirmed by observing smears stained by the Ziehl–Neelsen method.[14]
All isolates were sent to IAL Central for confirmation of species identification, the performance of AST, and sequencing of the 81 bp fragment of the rpoB gene in isolates with discordant results between Xpert MTB/RIF and AST, as described below:
The sensitivity test was performed using the Bactec MGIT 960 method (Becton and Dickinson– BD, Baltimore, USA).[14]
DNA extraction from the isolates, the conditions of each amplification reaction and sequencing of the 81 bp fragment of the rpoB gene, were performed.[15]
Data were transported to a Professional spreadsheet for grouping and analysis of study variables. Data were transported to a Professional spreadsheet for grouping and analysis of the study variables. Data analysis was performed using the statistical software Epi-Info 7.2.1 (Atlanta, Georgia, USA). The data were presented by frequency distribution and, for comparative analyses, Pearson's chi-square test or Fisher's exact test were used. A p-value≤0.05 was considered statistically significant. To measure the degree of agreement between variables, the Kappa method was used considering its recommended values.[16]
This study was approved by the Scientific Technical Council– CTC/IAL of IAL Central n CTC 66-J/2017, fully preserving the anonymity and details of the patients, who were excluded during the analyses.
ResultsA total of 1575 (31.8%) samples were analyzed from a total of 4953 from the entire area covered by the ESG-29 and ESG-30, which met the criteria for performing Xpert MTB/RIF and culture [Figure 1].
Figure 1: Flowchart of samples with Xpert MTB/RIF and culture for laboratory diagnosis of TB, period 2015–2019. *AST not performed = there was no growth in AST. MTB: Mycobacterium tuberculosis/Rifampicin, TB: Tuberculosis, AST: Antimicrobial susceptibility testing, MTC: Mycobacterium tuberculosis complex, NTM: Nontuberculous mycobacteriaRegarding the origin nature of the clinical material, 1507 pulmonary samples and 68 extrapulmonary samples were processed. Among the pulmonary ones, the most frequent clinical material was sputum (1452 - 96.4%), followed by bronchoalveolar lavage (45 - 3%) and gastric lavage (10 - 0.7%). In extrapulmonary samples, the frequency of clinical materials was more heterogeneous: cerebrospinal fluid (22 - 1.46%), tissue fragments (13 - 0.86%), and other specimens (33 - 2.2%).
The detection of MTB by Xpert MTB/RIF in pulmonary samples was statistically significant when compared to extrapulmonary samples (P < 0.05).
The association between LJ, OK, or MGIT media and positive and negative cultures did not result in significance, showing that both the liquid and solid media are in agreement in their results (P = 0.862); however, the solid media (LJ or OK) contaminated more than the MGIT liquid medium (P < 0.05) 0.0000814095.
[Table 1] shows the culture results against Xpert MTB/RIF.
Table 1: Culture results against Xpert Mycobacterium tuberculosis rifampicin, 2015-2019Detected had contaminated cultures. Culture contamination frequency = 4.2% (66/1575).
For the calculations of sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV), the 47 samples that were identified as nontuberculous mycobacteria (NTM) were excluded.
The sensitivity of Xpert MTB/RIF in relation to culture was 94.55% (278/294), the specificity was 95.97% (1122/1169), the accuracy was 95.69% (1400/1463), the value PPV was 85.53 (278/325), and the NPV was 98.59 (1122/1138). The analysis showed that the Xpert MTB/RIF and culture tests are in agreement (P < 0.05). Xpert MTB/RIF Kappa Value in relation to culture: κ= 0.87 (κ between 0.81 and 1.00 – Almost perfect agreement), 5% confidence interval: 0.840–0.902.
Culture showed 1169 negative samples and 340 positive samples, of which 47 were identified as NTM.
Among the identified NTM, it was possible to identify the following species: Mycobacterium avium (4 isolates), Mycobacterium intracellulare (12 isolates), Mycobacterium intrecellulare-chimaera (9 isolates), Mycobacterium avium-colombiense (1 isolate), Mycobacterium fortuitum (14 isolates), Mycobacterium abscessus (4 isolates), Mycobacterium gordonae (1 isolate), Mycobacterium kansassii (1 isolate), and Mycobacterium sp. (1 isolated).
The Xpert MTB/RIF specifies the bacillary load in the sample. The results of the TRM-TB/RIF detection levels, and their influence on the culture result, are shown in [Table 2].
Table 2: Relationship between culture and Xpert Mycobacterium tuberculosis/rifampicin detection levels, 2015-2019Statistical analysis showed a strong correlation between Xpert MTB/RIF semiquantitative results and culture positivity results (P < 0.05).
Regarding the performance of Xpert MTB/RIF against AST in the determination of resistance and sensitivity to RIF among positive samples, 274 samples showed concordant sensitivity and four samples showed concordant resistance in both tests. Two samples were resistant in the AST to isoniazid and were sensitive to RIF in the Xpert MTB/RIF, which does not constitute disagreement, since the test does not include the Isoniazid detection.
Regarding the 13 results of RIF Resistant by Xpert MTB/RIF, the following was observed: 4 confirmed resistance to RIF also by the AST method; 4 had contaminated cultures, making it impossible to perform the AST, and 5 were RIF sensitive by the AST method, a result that is divergent from the molecular technique. These isolates were submitted to Sanger sequencing of the 81 bp region of the rpoB gene, and in three, the mutation was H526N, and in two isolates, it was not possible to analyze the sequencing.
Furthermore, five RIF results were indeterminate by Xpert MTB/RIF: 4 of these had contaminated cultures, and one had a sensitive AST result [Table 3].
Table 3: Comparison of rifampicin resistance in Xpert Mycobacterium tuberculosis/rifampicin and antibiotic sensitivity test, 2015-2019For the calculations of RIF, Xpert MTB/RIF, and AST, listed below, the two samples resistant to isoniazid were excluded: Sensitivity S = 100 (4/4) and Specificity E = 98.2 (274/279). The accuracy, positive, and NPVs were A = 98.2 (278/283), PPV = 44.4 (4/9), and NPV = 100 (274/274), respectively.
Laboratory confirmation data on TBWEB include smear and culture. The Xpert MTB/RIF has been inserted into the TBWEB as a note preventing data from being exported through reports in the system. There was no difference in the increase in the TB diagnosis (smear and culture) in the periods before and after the introduction of Xpert MTB/RIF in the study region (51.45%/51.45%) [Table 4].
Table 4: Total tuberculosis cases diagnosed in 2010/2014 and 2015/2019 in Epidemiological Surveillance Groups 29 and 30 Discussion In view of the epidemiological picture of TB, of disease outbreaks by resistant and multidrug-resistant strains, early diagnosis is a constant challenge for the TB control program. The recommendation of the Xpert MTB/RIF by the WHO in 2010 for use as the main tool for the fast diagnosis of TB was an innovative step.
The diagnostic agility of the Xpert MTB/RIF test is extremely important since the faster the patient is diagnosed with TB, and the treatment can be started earlier, aiming to reduce the risk of infecting other people, increasing the success rate in the management of treatment,[17] especially for patients who are carriers of immunodeficiencies, such as HIV.[9]
Lung samples were significantly more detectable by Xpert MTB/RIF than extrapulmonary samples. The better performance of the Xpert MTB/RIF method in pulmonary samples was also observed by some authors.[18],[19],[20] However, Xpert contributes to early positivity in extrapulmonary samples that are paucibacillary and difficult to be detected by smear microscopy.
In the present study, the analysis of the data found in the tests carried out in parallel, allowed us to assess that the Xpert MTB/RIF had a good performance against culture in negative samples; however, some samples were classified as not detected by the equipment, showed positive culture. Subsequently, about 75% of these cultures were identified as NTM, which explains why they were not detected by Xpert MTB/RIF since the equipment is programmed to amplify and detect only the DNA of MTC. This result highlights the importance of carrying out the culture to not harm the diagnosis of NTM. The fact that the culture was still positive for 25% of the samples that were negative in the Xpert MTB/RIF can be explained by the occurrence of probe failures.[21]
When analyzing the reverse situation, the detection of Xpert MTB/RIF in relation to the positivity of culture, the Xpert MTB/RIF presented a better performance in relation to culture. Some factors may have contributed to a sample being detectable in Xpert MTB/RIF and negative in culture, concomitantly. The Xpert MTB/RIF is a test for amplification and detection of bacillus DNA, and it does not need to be viable since in culture, it is necessary that the bacilli are alive so that there is multiplication and there has still been a drastic decontamination of the clinical material to the elimination of the associated microbiota, which may have caused bacillary death.[22]
When we analyze the detection levels of Xpert MTB/RIF, and its influence on the culture result, we can observe that among the samples processed with a negative result in the culture, most of them presented very low and low detection levels, highlighting the effectiveness of the Xpert MTB/RIF in the accurate and fast diagnosis of TB, even when the bacillary load is reduced.
Regarding the discordant results in Xpert MTB/RIF and Culture, in a study carried out for the implementation of Xpert MTB/RIF,[23] a small discrepancy was also observed between the positivity of Xpert MTB/RIF in relation to Culture, corroborating our results.[23]
Studies have shown that the Xpert MTB/RIF has significantly contributed to the early detection of TB cases,[5],[6],[24] but they have observed that there were large variations in the results obtained in several of them, which can probably be explained due to the different locations of these cases studies, and also probably because of the heterogeneous form of the methodologies used for the analyses.[6] In addition, state networks and local realities in health services can also influence the use of technology implementation in different contexts,[25] as observed in this study, perhaps because of the difficulty of health teams to absorb the implementation of new methodologies. Even in the face of any and all difficulties faced, the methodology has provided agility at the beginning of treatment, thus preventing the spread of the disease.
One of the great advances in diagnosis that the Xpert MTB/RIF test makes is to detect, together with the MTC, the existence of resistance to RIF. This study also analyzed the test's performance against the AST. The Brazilian Ministry of Health recommends repeating the Xpert MTB/RIF whenever the first test detects resistance to RIF, and performing culture for isolation of the MTC and AST from the other drugs and also to RIF, to confirm the resistance and suitability of the drug treatment.[10],[26]
The results found in our study showed good accuracy between AST and resistance to RIF by Xpert MTB/RIF. However, Xpert MTB/RIF found more resistance to RIF than AST, a fact also found by other authors. By sequencing the 81 bp region of the cultures which were resistant in Xpert MTB/RIF and sensitive in AST, we observed that the mutation was H526N, which is characterized by a low level of resistance to RIF,[27] that is, they present the minimum inhibitory concentration below of the critical concentration of the AST.[28],[29]
The implementation of the Xpert MTB/RIF is part of the innovation and advancement in the diagnosis of TB. Through this study, it can be inferred that the method meets the demands of high sensitivity and specificity, both in the diagnosis of MTC and in resistance to RIF, according to studies by other authors.[5],[30]
Thus, the indication of nucleic acid amplification tests and the implementation of the Xpert MTB/RIF technology should be seen as a promising global policy to achieve the goals of elimination and resistance to TB in the 21st Century,[9],[31] which will certainly contribute for access to quality diagnosis.[3]
The use of Xpert MTB/RIF proved to be a diagnostic tool of low technical complexity, with low sample contamination and good accuracy.
This study will be able to provide subsidies for the expansion and maintenance of the Xpert MTB/RIF network, which is part of the State Health Plan (2016–2019)- “Prevention and integrated care centered on people with TB, one of the challenges of the National Health Program TB in the country that aim to improve disease care.”[3],[11]
A limitation of the study was the obtaining of secondary data in the information system (TBWEB) that may be incomplete. In Brazil, there was an increase in bacteriological confirmation with the introduction of molecular diagnosis by the Xpert MTB/RIF system.[7] The State of São Paulo has 645 municipalities, and the State represents about 20% of TB cases in Brazil,[17] being the region of this study comprising 102 of these municipalities in the state. However, at the level of GVE-29 and ESG-29, this increase was not confirmed, a fact that may have occurred due to the management difficulty in extracting information from the database or even due to incomplete information in the TBWEB notification bank, where the information on the diagnostic criteria is not available had Xpert MTB/RIF data in many of the reported patients, but the test already produces the benefits of early diagnosis and but the test already brings the benefits of early diagnosis and interruption of the transmission chain.
The implementation of the test already produces the expected benefits, but for a real evaluation of these benefits in the general context of the country, the Program has to overcome the difficulties of the information to be inserted in the notification bank, which is another challenge to be faced in our country.
Ethical clearance
This work was approved at the Adolfo Lutz Institute by Scientific Committee, under number CTC 66-J/2017.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)