Concordance and discordance of GeneXpert MTB/RIF and conventional culture method for diagnosis of Extra-Pulmonary Tuberculosis at a tertiary care hospital in Pakistan

Objective: To identify concordance and discordance between GeneXpert MTB/RIF assay and gold standard bacteriologic culture for the diagnosis of Mycobacterium tuberculosis (MTB) in Extra-Pulmonary tuberculosis (EPTB) specimens in our region. Methods: This is a retrospective cross-sectional study conducted at the Indus Hospital and Health Network. Data from 1st January, 2020 to 31st December, 2021 was analyzed. A total of 1499 EPTB specimens were included for which GeneXpert was requested along with acid-fast bacteria (AFB) culture from the same specimen. Specimens were processed according to specimen type following standard operating procedures of the laboratory. Fluorescent staining was performed on all specimens along with bacteriologic culture. The GeneXpert MTB/RIF assay was carried out in exact accordance with the manufacturer’s instructions. Results: Out of 1499 EPTB specimens, 1370 (91.39%) specimens exhibited concordance between GeneXpert and conventional culture method, while 129 (8.60%) specimens showed discordance. GeneXpert exhibited sensitivity and specificity of 69.4% and 94.3% respectively in comparison to culture. Conclusion: GeneXpert sensitivity for the diagnosis of EPTB varied with the site involved. Lower sensitivity was observed in ascitic and pleural fluids as compared to higher sensitivity observed among urine samples and pus aspirates. However, given the quick turnaround time and ease of use, it is a helpful tool in the diagnosis of EPTB when utilized in the appropriate clinical context. Caution is advised while interpreting negative GeneXpert results in endemic settings and should be interpreted along with other supporting clinical and diagnostic features.


INTRODUCTION
According to the World Health Organization (WHO) global tuberculosis report of 2021 there were an anticipated 9.9 million diagnosed cases of tuberculosis, or an average 127 cases (114-140 per 100,000 people) reported worldwide in the year 2020.Regions with the highest rates of tuberculosis (TB) cases were South-East Asia (43%), Africa (25%), and Western Pacific region (18%).Eastern Mediterranean (8.3%), the Americas (3.0%), and Europe (2.3%) had the lowest reported cases.Eight of these 30 high TB burden countries accounted for two thirds of the total world population and accounted for 86% of all estimated incident cases across the globe: India (26%), China (8.5%), Indonesia (8.4%), the Philippines (6.0%), Pakistan (5.8%), Nigeria (4.6%), Bangladesh (3.6%), and South Africa (3.3%).Pakistan ranked 5 th amongst the 30 high TB burden countries and 4 th in highest prevalence of multidrug-resistant TB (MDR-TB) with an estimated 510,000 incident cases, which accounts for 61% of the TB burden in the Eastern Mediterranean Region. 1,2An upward trend is observed in extra-pulmonary tuberculosis (EPTB) cases in Pakistan, which accounts for 20% of all TB cases. 35][6][7] It causes tuberculosis (TB), a communicable disease which primarily affects the lungs but can disseminate to various organs such as gastrointestinal, genitourinary, cardiovascular and central nervous system, lymph nodes, joints, bones, and kidneys etc. 5 Clinically TB is divided into Pulmonary tuberculosis (PTB) and Extra-Pulmonary tuberculosis (EPTB).Chief presenting complaints of PTB are chronic cough, and purulent blood-tinged sputum.EPTB a new emerging entity amongst children, and immunocompromised individuals, manifests as a chronic granulomatous disease affecting body tissues, and organs other than lungs.Patients with EPTB occasionally have a positive sputum culture but otherwise normal pulmonary function. 4,5,8Diagnosis of EPTB is challenging for both physicians, and pathologists, as the number of bacilli is often too low, and obtaining specimens from deepseated infections can be difficult. 6,7,9old standard for definitive diagnosis of EPTB is culture on solid and liquid media.EPTB is a paucibacillary disease and culture requires only 10-100 bacilli/ml of concentrated sample.Yield of liquid culture medium is 10% more than solid media. 10The drawback of culture is prolonged turnaround time of result due to slow growth rate of MTB.It requires approximately six to eight weeks on LJ medium and, two to six weeks in liquid media. 5,11,12oreover, it is a laborious process which requires specialized lab equipment of biological safety level III and its availability is limited in resource-constrained settings. 12eneXpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) is a fully automated real-time nested PCR assay, highly recommended by WHO for the diagnosis of PTB since 2010 and EPTB since 2013. 5,9,13It can detect both the MTB genome and rifampin resistance (rpoB gene mutation) in clinical specimens within two hours. 9,14,15Acid-fast bacteria (AFB) microscopy either with Ziehl-Neelsen stain (ZN) or fluorescent stain is time consuming, unable to differentiate MTB from Non-tuberculous Mycobacteria and, has been reported to have poor sensitivity and specificity due to paucibacillary nature of EPTB.Main objective of our study was to identify concordance and discordance between GeneXpert MTB/RIF assay and culture for the diagnosis of MTB in EPTB specimens.Our objective was to identify concordance and discordance between GeneXpert MTB/RIF assay and culture for the diagnosis of MTB in EPTB specimens.

Inclusion criteria:
The data from all of the patients (inpatients and outpatients departments), where AFB Culture was requested along with GeneXpert MTB/RIF from the same Extra-Pulmonary Tuberculosis sample (EPTB), during the targeted time duration were included.
• All EPTB samples where GeneXpert was not requested with culture from the same sample.Statistical analysis: The data was entered and analyzed using the SPSS version 26.0 for windows.The normality distribution was assessed via Kolmogorov-Smirnov and Shapiro-Wilk tests.As per the distribution of data, the quantitative variables e.g., age was reported as mean ± standard deviation.Frequency and percentage were reported for categorical variables i.e., Gender, GeneXpert result and culture results.Sensitivity, specificity and predictive values were calculated by considering mycobacteriology culture as the gold standard, using 2x2 crosstab method on the SPSS software.Kappa analysis was also performed to see the agreement between GeneXpert and Culture results.P value <0.05 was considered statistically significant.
The overall concordance of GeneXpert was seen in 91.39% (n=1370) and discordance in 8.26% (n=129) cases.In culture positive cases the concordance between GeneXpert and culture was seen in 69.36% (n=120) and discordance in 30.63% (n=53) specimens.In culture negative cases the concordance was seen in 94.26% (n=1250) and discordance was seen in 5.3% (n=76) (Fig. 2).Kappa analysis was run to determine if there was agreement between GeneXpert and Culture.There was moderate agreement between GeneXpert and Culture κ = 0.602 p-value <0.01.Smear negative, Culture and GeneXpert positive were 50% (n=53) where tissue specimens represented 62.26% (n=33).The sensitivity and specificity of GeneXpert MTB/RIF in comparison to culture was 69.4% and 94.3% respectively with positive predictive value of 61.2% and negative predictive value of 95.9%.The GeneXpert was 98.7% specific and 75% sensitive in CSF.The sensitivity and specificity of GeneXpert in various samples is urine 100%, sensitive and 96.6% specific, pus 92.3% sensitivity and 85% specificity, fluids 80% sensitivity and 97.1% specificity, site not specified 90% specificity, gastric aspirate 75% sensitivity and 96.3% specificity, tissues 69.6% sensitivity and 94.2% specificity, and pleural fluid 32.3% sensitivity and 97.4% specificity.The sensitivity and specificity of GeneXpert in different samples is shown in Table-III.

DISCUSSION
The overall sensitivity, specificity, positive and negative predictive value of GeneXpert results in our study were 69.4%, 94.3%,61.2%and 95.9%, respectively Table-II.The specificity of GeneXpert is consistent with studies conducted in other regions where specificity ranges from 73-100%. 5 However, there is significant variation in sensitivity which ranges from 52-100%.The study conducted by Osei et al. 7 in 2019 demonstrated 50%, and another study by Elbrolosy et al. 13 in 2021 exhibited 81.6% sensitivity, while in our study it was 69.4% (Table-IV). 7,13his could be due to blood and other PCR inhibitory substances in samples with a paucibacillary disease and smashing of tissues during homogenization. 7,17,18he ranking of Pakistan among the top five high TB burden countries is alarming.A 20% increase in incident TB cases of EPTB further aggravates the situation.Directly observed therapy (DOT) strategy for TB which started in 2011 has been implemented in almost all the public health sectors.The National TB control program (NTP) revived under the Ministry of Health has developed uniform policies and strategies to counter the rising number of both PTB and EPTB cases in response to the declaration of TB as a national emergency in Pakistan. 17,19Despite all these measures, the diagnosis and management of EPTB remains challenging.
To the best of our knowledge, there is very little information on the performance of GeneXpert MTB/RIF assay in diagnosis of EPTB cases in Pakistan.This highlights the need of evaluating the diagnostic capability of GeneXpert in not only Urban Sindh but also in Rural Sindh and other provinces of Pakistan, as the number of EPTB cases may vary according to the burden of illness and its severity.The concordant results of GeneX-pert have a great impact in early identification of TB in comparison with the culture which requires a prolonged time, but the number of discordant cases cannot be overlooked, addressing even a small number of cases is vital for lowering the disease burden and eradication of TB in endemic countries.Hence GeneXpert MTB/RIF cannot eliminate the necessity of conventional culture methods that are required to establish the diagnosis of TB.
Variation between sensitivities and specificities was observed according to the specimen type ranging from 20% sensitivity in ascitic fluids to 92.3% in pus aspirates and, 50% specificity in synovial fluids to 100% in cold abscess, bone marrow, pericardial and peritoneal fluids.Our results showed higher pus sensitivity of 92.3% compared to 56.7% demonstrated in a study by Parkash et al. 15 when compared to a study by Vadwai et al. 8 where CSF material showed a subpar sensitivity of 29%, the CSF specimens in our study demonstrated superior sensitivity of 75%. 8In a study conducted in Lahore, Pakistan by Iram S et al. 12 the sensitivity and specificity of GeneXpert in EPTB specimens was reported as 100% and 86%, respectively. 12his is in contrast with our results where sensitivity was 69.4% and specificity was 94.3%.The stark difference between both studies can be attributed to the difference of sample size, with our sample size being significantly higher than the study in comparison.In another study conducted in Peshawar, Pakistan by Khan AS et al. 20 the overall sensitivity of GeneXpert was 73% and specificity was 100%.In terms of specimens, the tissue samples in our investigation showed a sensitivity of 69.6% compared to 100% by Khan AS et al. 20 75% compared to 83% for CSF samples, and 100% compared to 57% for urine samples.The heterogeneity between sensitivities in different studies can be attributed to the difference in the disease burden, patient populations, type of EPTB, sample size and specimen quality, in countries where TB is endemic.

Limitations:
It is a single center study thus its findings cannot be generalized as the data may vary.Secondly, this was a retrospective, cross-sectional study, hence the possibility of handling error, processing of specimens, observer's bias in microscopy, and technical errors related to GeneXpert could not be repeated for confirmation.For samples labelled as site-not-specified and other fluids, the nature and anatomic location of samples could not be verified.Furthermore, radiologic evidence and histopathological findings could not be compared and sensitivity and specificity of GeneXpert MTB/RIF assay for rifampin resistance was not evaluated.

CONCLUSION
The performance of GeneXpert varied with the site of extrapulmonary involvement with lower sensitivity in ascitic and pleural fluids and a higher sensitivity in pus aspirates, cold abscesses and urine specimens.The rapid turnaround time of GeneXpert can help in timely detection of tuberculosis in these cases and appropriate therapeutic intervention can be started before culture results are available.However, in areas where TB is endemic, caution is advised for interpreting negative GeneXpert results in clinical settings and should be interpreted along with clinical signs and symptoms, positive contact history, radiological findings, histopathological diagnosis and microbiological culture.Further research with larger sample size is needed to evaluate the utility of GeneXpert in EPTB cases in endemic settings, which will aid in development of concrete diagnostic guidelines for effective treatment.

Table - I
: Demographic parameters of culture negative and culture positive cases.
Fig.1: Distribution of specimens included in study according to culture positivity.

Table -
II: Diagnostic performance of GeneXpert in comparison to culture Qurat-ul-Ain Zahid et al.

Table -
III: Diagnostic Performance of GeneXpert by specimen type.Table-IV: Comparison of sensitivity and specificity of GeneXpert in EPTB specimens in different studies.