Combination of Xpert® MTB/RIF and DetermineTM TB-LAM Ag improves the diagnosis of extrapulmonary tuberculosis at Jimma University Medical Center, Oromia, Ethiopia.

BACKGROUND: Ethiopia is one of the high burden countries for extrapulmonary tuberculosis (EPTB); however, the prompt diagnosis of EPTB remains challenging. This study is aimed to evaluate the diagnostic performance of Xpert MTB/RIF and DetermineTM TB-LAM Ag (TB-LAM) for the prompt diagnosis of EPTB in Ethiopia.
METHODS: A total of 147 presumptive EPTB patients, including 23 HIV- positive participants were enrolled. Extra-pulmonary samples were collected from all presumptive EPTB cases and tested for Mycobacterium tuberculosis complex (MTBC) using fluorescent microscopy, Xpert MTB/RIF, and culture. Additionally, urine samples were also collected from 126 participants and were tested by DetermineTM TB-LAM Ag (Alere Inc, Waltham, USA). The Sensitivity and specificity of Xpert and TB- LAM tests were calculated by comparing with a composite reference standard (CRS), which comprises smear microscopy, culture and response to empirical anti-TB treatment.
RESULTS: Of 147 patients, 23 (15.6%) were confirmed EPTB cases (culture-positive), 14 (9.5%) were probable EPTB (clinically, radiologically or cytologically positive and received anti-TB treatment with good response), and 110 (74.8%) were classified as "non- TB" cases. Compared to the composite reference standard (CRS), the overall sensitivity and specificity of Xpert MTB/RIF were 43.2% and 100%, respectively with the highest sensitivity for Lymph node aspirate (85.7%) and lower sensitivity for pleural fluid (14.3%) and 100% specificity for all specimen types. The sensitivity and specificity of TB-LAM were 33.3% and 94.4% respectively with the highest sensitivity for HIV co-infected participants (83.3%). The sensitivity of the combination of Xpert MTB/RIF and TB-LAM tests regardless of HIV status was 61.1% whereas the sensitivity was improved to 83.3% for HIV-positive cases.
CONCLUSION: TB-LAM alone has low sensitivity for EPTB diagnosis; however, the combination of TB-LAM and Xpert MTB/RIF improves the diagnosis of EPTB particularly for countries with high EPTB and HIV cases.

Background

Ethiopia is among the 30 highest-burden countries for tuberculosis (TB), TB-HIV coinfection and multidrug-resistant TB (MDR-TB) in the world [1]. In Ethiopia, extrapulmonary TB (EPTB) is very high- accounting for more than 33% of all forms of TB [1, 2]. EPTB contributes significantly to TB-related morbidity and can cause complications, lifelong sequelae, and death. However, the diagnosis of EPTB remains challenging mainly because of the variable non-specific presentations, the paucibacillary nature of the disease, and the difficulty in obtaining appropriate and adequate samples [2-4]. Conventional techniques such as smear microscopy has limited sensitivity for the diagnosis of EPTB, due to the paucibacillary load of the diseases. Mycobacterial culture has long turnaround time and requires bio-safety level III facilities which limits its applicability in resource poor countries [5, 6]. Thus, the diagnosis of EPTB is often made on clinical suspicion alone, resulting in both under- and over-diagnosis and relatively poor outcomes.

Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) is a cartridge-based nucleic acid amplification test for the diagnosis of TB and rifampicin resistance and the test has been endorsed by the WHO for use in resource constrained settings [7]. Steingart et al. [8] in their Cochrane systematic review reported high diagnostic accuracy (sensitivity 89% and specificity 99%) of Xpert MTB/RIF for pulmonary TB detection. However, Tortoli, et al [9] investigated large numbers of consecutive extrapulmonary clinical specimens with Xpert MTB/RIF and obtained a heterogeneous sensitivity across different specimens (sensitivity >85% for CSF, biopsies, urines, pus samples and fine-needle aspirates but very poor sensitivity for the body fluids). More recent study in Ethiopia also reported a heterogeneous sensitivity of Xpert MTB/RIF for the diagnosis of different forms of EPTB (from 30% for pleural TB and up to 90% for lymph node TB) [10]. Thus, more evidence is still needed on the diagnostic performance of Xpert MTB/RIF for the diagnosis of the different forms of EPTB [11].

Lipoarabinomannan (LAM) is an immunogenic lipopolysaccharide which is found in mycobacterial cell walls. It is released from metabolically active or degenerating bacterial cells and present mainly in people with active TB disease and has shown only low cross-reactivity with non-tuberculosis mycobacterial infections [12, 13]. LAM is excreted and detected in urine during advanced immunodeficiency such as HIV coinfection; which results in systemic dissemination of M. tuberculosis and high mycobacterial burden due to podocyte dysfunction and change in glomerular filtration rate that leads higher antigen concentration of LAM in urine [14, 15]. The other possible way for LAM excretion in urine is due to renal involvement of patients with disseminated TB in patients living with HIV and advanced immunodeficiency [16].

Most laboratory diagnostic tests of EPTB involves invasive procedures to obtain specimens, which further complicates the diagnostic protocol [17]. Thus, there is a need for rapid point of care and non-invasive tests for the diagnosis of EPTB. DetermineTM TB LAM Ag (TB-LAM) detects LAM in the patient’s urine for the diagnosis of TB. Urine-based testing would have advantages over other specimen based testing because urine is easy to collect and store, and lacks the infection control risks associated with collection. WHO recommends TB-LAM for people living with HIV with symptoms suggestive of pulmonary and/or extrapulmonary TB, with CD4 ≤100 cells/mm3 or in those seriously ill [13]. However, the yield of Xpert MTB/RIF and TB-LAM tests used in combination for the diagnosis of EPTB is unknown. This study is aimed to evaluate the diagnostic performance of Xpert MTB/RIF and TB-LAM for the diagnosis of EPTB in Ethiopia.

Materials and methods

Study subjects, sample collection and laboratory tests

This study was conducted at Jimma University Medical Center, Southwest Ethiopia from April to October 2019. A total of 147 presumptive EPTB patients were consecutively enrolled in the study. Consecutive patients were included based on clinical suspicion of EPTB. As part of routine clinical practice, body fluids such as cerebrospinal fluid, pleural fluid, peritoneal fluid, pericardial fluid, synovial fluid, and lymph node aspirate were collected according to standard operating procedure (SOP). Gastric aspirate was collected early in the morning through nasogastric tube following an overnight fast. Each sample was divided into two and transferred into falcon tubes. The first sample was processed for Xpert MTB/RIF at Jimma University Medical Center-Microbiology Laboratory and the second sample was transported to Jimma University-Mycobacteriology Research Center where smear microscopy and L-J culture were done. In addition, all presumptive EPTB patients were requested to submit urine specimen (10-30ml) for the research purpose. Immediately after collection, urine specimen was transported to the Mycobacteriology Research Center and stored at -20°c until processing.

Demographic and clinical characteristics were gathered through interview by using a pre-tested questionnaire. Participants’ medical records were reviewed for HIV status, HIV-WHO staging, clinician’s decision to initiate anti-TB treatment (ATT) and response to anti-TB treatment.

Mycobacterial culture

Culture was done on Lowenstein Jensen (L-J) medium. Lymph node aspirate, gastric aspirate and blood-stained specimens were decontaminated by the standard N-acetyl-L-cysteine and sodium hydroxide (NALC/NaOH) method with a final NaOH concentration of 1% [18]. An equal volume of standard NALC/NaOH solution was added to the specimen and incubated for 15–20 minutes. After centrifugation for 15 minutes at 3000g, the sediment was resuspended in 1.0 to 1.5 ml of sterile phosphate buffered saline (pH 6.8). Specimens expected to be sterile (such as cerebrospinal fluid (CSF), pleural fluid, and peritoneal fluid) were directly centrifuged to concentrate the samples. L-J tubes were inoculated with 0.2 mL (2–4 drops) of the processed specimens [19]. For positive L-J culture, a smear was prepared to detect acid-fast bacilli (AFB), and Mycobacterium tuberculosis complex (MTBC) was confirmed by a p-nitro benzoic acid inhibition test [19].

Fluorescent smear microscopy

Smears were prepared from all processed samples other than urine and examined at Mycobacteriology Research Center of Jimma University (JU-MRC). Approximately 10μl of the pellet was deposited on a clean microscope slide and spread using the side of the pipette tip. Smears were air-dried and heat-fixed and stained with the Auramine-phenol method according to the SOP. The stained smears were examined under the light-emitting diode fluorescent microscopy (Primo Star iLED, Carl Zeiss, Gottingen, Germany) with 200x and 400x magnification for AFB [19].

Xpert MTB/RIF test

The Xpert MTB/RIF assay was performed as previously described by Helb et al. and WHO [20, 21]. Briefly, sample reagent was added in a 2:1 ratio to patient specimen. The mixture was vortexed and incubated at room temperature for 15 minutes. Two ml of the reagent sample mixture was transferred to an Xpert cartridge using a Pasteur pipette. Then the cartridge was loaded onto Xpert machine (GeneXpert- Dx System version 4.4a, Cepheid Company, 904 Caribbean Drive, CA 940889, USA) and results was automatically generated after 1 hour and 50 minutes.

TB-LAM

The stored urine specimen in a deep freeze was de-frozen and processed as previously described [13, 22, 23]. About 10-30m ml of the de-frozen urine sample was centrifuged at 10,000g for 5 minutes, and 60μl of the supernatant was applied to the TB-LAM test strip [DetermineTM TB-LAM Ag test; Abbott, Waltham, MA (formerly Alere)] on the same day with results interpreted using a 4-grade scale, with grade one or above constituting a positive result. The strip was visually inspected by 2 trained study staff members exactly at the end of 25 minutes incubation period. Those individuals who interpreted the TB-LAM were blinded to patients’ clinical data and TB diagnostic status. Any disagreement between interpreters were decided by the decision of a third interpreter. TB-LAM results were not used for patient management or treatment initiation.

Diagnostic classification for analysis

Based on clinical and laboratory findings, study participants were categorized as follows: (i) Confirmed TB: defined as a positive culture of MTBC and /or smear positive; (ii) Probable TB: culture negative but clinical improvement after anti-TB treatment (ATT); (iii) Non TB: patients for whom no microbiological evidence of TB (smear-negative and culture-negative), and/or for whom an alternative diagnosis is available. Composite reference standard (CRS), which comprises smear microscopy, L-J culture and clinical improvement after ATT initiation, was used as gold standard to calculate sensitivity, specificity and predictive values. Any patient that was positive for any one component of the CRS was considered as ‘TB’ cases.

Data management and analysis

Data were entered through Epidata version 3.1 and analyzed using SPSS software package version 20. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of TB-LAM and Xpert MTB/RIF test were calculated in comparison with culture alone and with CRS. The results of the two TB-LAM test readers were compared, and the inter observer agreement was calculated using the Cohen’s kappa coefficient (κ) statistic. Overlapping 95% CI data were considered as showing no significant difference between test results. Comparison of proportion between the methods was done by Chi square test using the MedCalc Software (https://www.medcalc.org/calc/comparison_of_proportions.php) and the P-value < 0.05 showed statistically significance difference between the methods.

Ethics approval and consent to participate

This study was approved by institutional review boards of Jimma University, Ethiopia (Ref. No. IHRPGD/389/2019). Written informed consent was obtained from all participants for use of routine clinical data for research purposes. Laboratory results were reported back to the physicians for treatment initiation or decision as early as available.

Results

Demographic and clinical characteristics

A total of 147 presumptive EPTB patients were included. Participants had a median age of 35 years (IQR, 22–45) and 82 (56%) were males. The majority, 84 (57.1%) of the participants were from rural areas (). Regarding clinical signs and symptoms, 124 (85%) and 122 (83%) of the study participants had tiredness and loss of appetite, respectively. Among the 134 (91.2%) presumptive EPTB cases with documented HIV status, 23 (15.6%) were HIV positive, and most (89.9%) of them were already on antiretroviral therapy (ART) for a mean of 8 years (SD ±1.92).

All participants provided sufficient volume of site specific extrapulmonary specimens including 49 peritoneal fluid, 45 pleural fluids, 28 CSF, 19 lymph node aspirates, 2 pericardial and 4 other fluids (2 gastric aspirates and 2 synovial fluids) (). In addition, all study participants were requested to provide urine specimen but only 126 (86%) of study participants provided urine specimen ().

Diagnosis of EPTB

In total, 147 extrapulmonary samples were successfully analyzed by fluorescent microscopy, Xpert MTB/RIF, and L-J culture. Of 147 presumptive EPTB cases, 49 (33.3%) were suspected for abdominal TB and 45 (30.6%) for pleura TB. Fifty-three (36.1%) of presumptive EPTB cases had various sites affected, including meningitis, lymph node, skin, and pericardial. However, TB disease detection (confirmed cases) rate was highest among patients suspected of having lymph node TB (56.5%) followed by pleura TB (17.4%) and TB meningitis (13%).

Bacteriologically confirmed EPTB was found in 23 of the presumptive EPTB cases, for a detection rate of definite EPTB of 15.6% (95% CI: 10.2%, 22.5%). Of these, 13 had lymph node TB, 4 had TB pleuritis, 3 had TB meningitis and 2 had peritoneal TB (). One TB case was identified from gastric aspirate from whom pulmonary manifestation was excluded (). Among 23 culture-positive EPTB cases, 16 (69.6%) were Xpert MTB/RIF positive, with no rifampicin resistance detected and 8 (34.8%) were positive for AFB by fluorescent microscopy. Furthermore, there were 14 (11.3%) patients with probable TB- cases with clinical improvement after clinician’s decision to start ATT whereas 110 patients were classified as ‘non TB’ cases because no bacteriological or clinical evidence for TB was observed (). The combined detection rate of definite and probable EPTB was 25.2% (95% CI: 18.4–33.0); 37/147. Compared to CRS (definite and probable EPTB cases), L-J culture had a sensitivity of 62.2% (23/37). As shown in , for determination of sensitivity, TB-LAM test was compared against 36 and Xpert MTB/RIF test against 37 positive samples found by CRS.

Diagnostic performance of Xpert MTB/RIF

The diagnostic accuracy of Xpert MTB/RIF test to diagnose EPTB regardless of patients’ HIV status is shown in . Using L-J culture as a comparator, the sensitivity and specificity of Xpert MTB/RIF were 69.6% (95%CI, 47.1–86.8); 16/23 and 100% (95%CI, 97.1–100); 124/124, respectively. However, the sensitivity was reduced to 43.2% (95%CI, 27.1–60.5); 16/37 when compared to CRS without affecting the specificity. When stratified by HIV status, the Xpert MTB/RIF sensitivity was higher in HIV positive patients; 50% (95%CI, 11.8–88.2); 3/6 compared to HIV negatives 35.7% (95%CI, 18.6–56); 10/28 (). The diagnostic accuracy of Xpert MTB/RIF test among different specimens were also determined. The sensitivities among the specimen types differed markedly. Comparing to CRS, the sensitivity of Xpert MTB/RIF was 85.7% (95%CI, 57.2–98.2); 12/14 for lymph node aspirate specimen, 33.3% (95%CI, 4.3–77.7); 2/6 for CSF and only 14.2% (95%CI, 0.4–57.9); 1/7 for pleural fluids (). Xpert MTB/RIF on site specific samples had higher sensitivity 44.4% (95%CI, 28–62); 16/36 than urine TB-LAM 33.3% (95%CI, 18.6–51); 12/36 performed for the same patient (n = 126), p<0.001.

Performances of TB-LAM, Xpert MTB/RIF and combination of TB-LAM and Xpert MTB/RIF tests comparing to CRS among HIV positive (A) and HIV negative (B) cases.

Diagnostic performance of TB-LAM

Out of 126 presumptive EPTB cases submitted urine sample for the TB-LAM, 17 (13.5%) had a positive TB-LAM result. Among 23 patients with culture-confirmed EPTB, 8 (34.8%) had a positive TB-LAM, and 4 (28.6%) TB-LAM positive cases were among probable EPTB patients (negative-Xpert MTB/RIF and -culture results). There were 5 cases with positive TB-LAM result but no evidence of TB (bacteriologically and clinically negative for TB) ().

When compared to CRS, the TB-LAM had lower sensitivity (33.3%; 95%CI, 18.6–51); 12/36 whereas higher specificity performance (94.4%; 95%CI, 87.5–98.2); 85/90. There is a significant difference in sensitivity of TB-LAM assay in HIV-positive and HIV-negative participants, which was 83.3% (95%CI, 35.9–99.6); 5/6 in HIV-positives and 26% (95%CI, 11.1–46.3); 7/27 in HIV-negatives (P-value = 0.0094) ().

Combination of Xpert MTB/RIF and TB-LAM

The combined use of TB- LAM and Xpert MTB/RIF tests had the sensitivity of 61.1% when compared to CRS, which was significantly higher than the sensitivity of each test alone (). Most importantly, in stratified analysis, combining the two tests (Xpert MTB/RIF and TB-LAM) have significantly improved the sensitivity to 83.3% in HIV-positives compared to HIV-negative patients (52%) (). The detection rate and diagnostic performance of TB-LAM, Xpert MTB/RIF and combination of TB-LAM and Xpert MTB/RIF in HIV-positive and HIV-negative participants was described more in supporting file ().

Discussion

Sensitive and rapid diagnostic tests will increase the number of diagnosed patients; facilitate early treatment, contributing potentially to decreased disease transmission and reduced case fatality, especially in the highly lethal form of TB (such as disseminated TB). Conventional techniques (smear microscopy and culture) have limited sensitivity for the diagnosis of EPTB, due to the paucibacillary load of the diseases. M. tuberculosis needs about 6–8 weeks to grow on Lowenstein Jensen (L-J) culture, which limits the applicability of culture for immediate case management. In most cases, it is difficult to obtain appropriate extrapulmonary specimen and there is a need for invasive procedures. Incorporating tests like TB-LAM which uses urine sample for TB diagnosis is attractive than the other EPTB diagnostic tools which need invasive procedures for sample collection [13, 24]. This study evaluated the performance of two rapid point-of-care tests (Xpert MTB/RIF and TB-LAM) for the diagnosis of EPTB in Ethiopia.

In our study, we found variable sensitivity of Xpert MTB/RIF among different extrapulmonary specimen types. We documented the higher sensitivity of Xpert MTB/RIF in lymph node aspirates and lower sensitivity in CSF and pleural fluid specimens; which is consistent with findings of previous studies [10, 25, 26]. The poor sensitivity in such fluids is probably due to the paucibacillary nature of the disease. Other studies suggested Xpert MTB/RIF Ultra over the conventional Xpert MTB/RIF with an overall improvement of 45.6% for the diagnosis of respiratory and non-respiratory smear-negative cases and paucibacillary diseases [27]. This is mainly because Xpert MTB/RIF Ultra has lower analytical limit of detection than Xpert MTB/RIF (15 CFU/mL versus 100–120 CFU/mL respectively) [28].

Although Xpert MTB/RIF has lower sensitivity than culture, it has some advantages over conventional culture. Xpert MTB/RIF has shorter turnaround time for the diagnosis of TB especially when a rapid diagnosis of TB and early treatment initiation is significantly necessary, particularly in highly lethal form of TB such as TB meningitis. Xpert MTB/RIF can simultaneously detect M. tuberculosis and rifampicin resistance within 2 hours in contrast to L-J culture that could take 2 to 6 weeks for M. tuberculosis to grow and conventional drug resistance tests could take 3 extra weeks. As a result Xpert MTB/RIF provides the information which is used to select treatment alternatives and for early infection control of drug resistant TB [29].

The other interesting finding from the current study is that the high specificity and positive predictive value of Xpert MTB/RIF, which provides significant evidences for the clinicians to rule in EPTB and to treat patients as early as possible following a positive Xpert MTB/RIF result. Whereas the overall low sensitivity of Xpert MTB/RIF (43.2%) compared to CRS shows that negative Xpert MTB/RIF result does not rule out EPTB diagnosis especially from body fluid specimens such as pleural and peritoneal fluids. In the future, the implementation of Xpert MTB/RIF Ultra has a potential to overcome the sensitivity limitations of Xpert MTB/RIF in pleural or peritoneal specimens [27, 30].

Overall, the sensitivity performance of TB-LAM for EPTB is low in our study. The reported performance of the TB-LAM varied greatly among different published studies [22, 31, 32]. A systematic review on TB-LAM for the diagnosis of pulmonary TB using microbiological reference standard showed a pooled sensitivity of 42% and specificity of 91% [33], which has been found higher sensitivity compared with the sensitivity of the current study, with no observed difference of specificity (94.4%). This low sensitivity of TB-LAM indicates that negative TB-LAM result cannot be used to rule out EPTB. However, TB-LAM alone should not be used to diagnose EPTB. Although low sensitivity of urine TB-LAM compromises its utility as a replacement for culture or Xpert MTB/RIF on respiratory samples, the ease of access of urine (easily obtained from children and adults), processing and storage and the low infection risk to health professionals during urine collection makes TB-LAM attractive. Additionally, TB-LAM can be done anywhere, including at the patient’s bedside, and takes only 25 minutes to perform [34].

In our study, there were 5 TB-LAM positive cases for whom TB was bacteriological and clinically ruled out. Absence of bacteriologically confirmed or clinically suggestive-TB strongly supports that these were indeed false positive results. In some cases, factors other than active TB are likely to be responsible for positive urine TB-LAM results. For instance, interference by non-tuberculous mycobacteria or other bacterial species, as well as contamination from environment or stool could be a cause for false positive result [35]. The other possible cause could be variation in the interpretation of the TB-LAM result. In our study, grade 1 cut-off point (the band of lowest intensity) was used as a threshold. Peter et al. found decreased specificity when the grade 1 cut-off point was used as threshold, as compared to using the grade 2 cut-off point [36].

In our subgroup analysis, the sensitivity of TB-LAM in HIV positive patients was much higher than in HIV negative patients. In this study TB-LAM detected TB in additional 3 and 2 HIV infected patients who were missed by smear microscopy and Xpert MTB/RIF respectively. This suggests that TB-LAM could somehow address the challenges for diagnosing TB in HIV co-infected patients through smear microscopy and Xpert MTB RIF test. The higher sensitivity of TB-LAM in HIV positive patients than HIV negative patients might be due to podocyte dysfunction that results change in glomerular filtration which is most of the time secondary to HIV disease that leads higher antigen concentration of LAM in urine [14]. Although not assessed in our study, different studies shows a significant association between the performances of TB-LAM and level of immunosuppression [37-39] and positive TB-LAM is as a predictor for mortality [22, 39]. In their recent study, Kerkhoff et al. [40] reported substantially higher sensitivity of Fujifilm SILVAMP TB-LAM (FujiLAM) over TB-LAM for detecting EPTB in HIV inpatients with moderate sensitivity in pleural fluid and CSF. This suggests a potential use of FujiLAM as a first-line test for the rapid detection of EPTB in HIV patients, with substantial added benefit in paucibacillary diseases such as pleural TB and TB meningitis.

The overall sensitivity of the combination of Xpert MTB/RIF and TB-LAM irrespective of HIV status was higher than either test alone in this study, which was 61.1% and it is comparable to L-J culture 62.2%. Most importantly, the sensitivity was increased to 83.3% when these two tests (Xpert MTB/RIF and TB-LAM) were used together in HIV-positive patients. This finding suggests that it is better to use the combination of Xpert and TB-LAM tests than each test alone or culture alone when diagnosing EPTB. Both Xpert and TB-LAM have short turnaround time than culture which results in early diagnosis and treatment of TB patients, which is one of the main pillars and components of WHO End TB strategies [24]. It is noticeable that using the combination of Xpert and TB-LAM is relatively more expensive than using each test alone, however it provides important contribution for the diagnosis of EPTB especially when there is a need for very sensitive test and short time to treatment initiation as in the case of disseminated TB, TB meningitis or TB in HIV infected patients. However, patients whose ultrasound and radiological or clinical findings strongly suggests EPTB should start ATT despite a negative Xpert MTB/RIF and TB-LAM results.

Our study is not without limitations. First, since some patients were not able to provide urine specimen, they had no TB-LAM results available for analysis. Second, the use of frozen urine samples for TB-LAM would have affect the sensitivity. Though not reproduced by other investigators, Peter et al. has reported that the use of frozen urine has been associated with reduced TB-LAM sensitivity [36].

Conclusions

Xpert MTB/RIF has the highest sensitivity for lymph node aspirates and lowest sensitivity for pleural fluids. The overall sensitivity of TB-LAM was low, but the sensitivity was significantly improved among EPTB patients infected with HIV. The sensitivity of combination of Xpert MTB/RIF and TB-LAM was superior to either test alone or equivalent to culture to diagnose EPTB. The evidence from this study suggests that the combination of these tests can improve the diagnosis of EPTB particularly in patients infected with HIV.

Detection rate and the diagnostic performance of TB-LAM, Xpert MTB/RIF and combination of TB-LAM and Xpert MTB/RIF in HIV-positive and HIV-negative participants.

(RTF)

Click here for additional data file.

The original SPSS dataset used and analyzed in the current study.

(SAV)

Click here for additional data file.

18 Jun 2021

PONE-D-21-10435

Combination of Xpert®MTB/RIF and DetermineTMTB LAM assayimproves the diagnosis of extrapulmonary tuberculosis at Jimma University Medical Center, Oromia, Ethiopia

PLOS ONE

Dear Dr. Tadesse,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 02 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Shampa Anupurba, MD

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified whether consent was informed.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

Additional Editor Comments (if provided):

The manuscript has been reviewed thoroughly. I have very minor additional comments although they might be repetition of the reviewer's comments:

1. Line 188- 'Forty-nine (33.3%) of the presumptive EPTB cases had the classification of colitis'-On what basis was this classification made and what was the sample collected from these patients?

2. Although 4 of the 23 HIV positive patients were bacteriologically confirmed, it is not clear how many were positive by Xpert MTB/RIF. What is meant by 50% in Line 211? Similarly for Determine TB LAM assay

3. Line 320- How was the sensitivity of LJ culture derived to be 62.2%? What was the gold standard?I think that was not within the scope of the study

4. Fig 1- The flow chart could start with inclusion of 147 patients suspected of presumptive TB instead of 160 patients

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In the current study the authors aim to investigate the diagnostic performance of combined use of Xpert MTB/RIF and TB LAM Ag test for detection of M. tuberculosis complex among EPTB patients in a medical center of Ethiopia. Considering the high prevalence of EPTB and association of HIV in Ethiopia, the current diagnostic approach provides the important insight for increased detection of EPTB cases in clinical settings. However, the current manuscript needs to address several issues before approval.

Major comments:

1. According to the manufacturer (Abbot Laboratories, Chicago, US), the test name is ‘Determine™ TB-LAM Ag’. It is suggested to use the exact test name in the Title of the manuscript. Throughout the text, authors should be consistent mentioning the test name. Examples of some discrepancies are: ‘Determine TB LAM test’ in Line-31, ‘TB LAM’ in Line 41, ‘Determine TB-LAM test’ in line 136, ‘TB-LAM assay’ in line 140 and so on. It is suggested to use ‘Determine™ TB-LAM Ag (TB-LAM)’ when appears first in the text, followed by ‘TB-LAM’ all through the text.

2. In the introduction section, please add one paragraph describing the other published data on the performances of the Xpert MTB/RIF assay and Determine TB-LAM assay for detection of EPTB cases.

3. One of the EPTB samples as has been described inconsistently throughout the text. For example, in Table-1 as ‘pus (abscess)’, Table-2 and throughout the text as ‘Abscess’. According to Line-191, it appears that samples were obtained from ‘Lymph node TB’. It would be easier for the readers to understand if the term ‘Lymph node aspirate’ is used all through the manuscript instead of ‘Abscess’.

4. Gastric aspirate is considered as an alternative specimen for detection of pulmonary TB when patients cannot expectorate the sputum. As this study aims to detect the EPTB cases, authors should exclude the two ‘gastric aspirate’ samples from the study and re-analyze the results and correct the statistics accordingly in different relevant Tables, Figures and throughout the text.

5. For Table-2 and Table-3: Authors should check carefully that the statistics are presented correctly. It seems that the value of sensitivity, specificity, PPV and NPV and their corresponding 95% CI are not accurate. As for example, in Table-2 (based on the calculation of Fig-1), for ‘Determine TB LAM test’: the sensitivity would be 34.8% with 95% CI ranging from 16.4-57.3, specificity would be 92.7% (86.7-96.6), PPV would be 47.1% (27.7-67.4), and NPV would be 88.5% (85.0-91.2). Please check both table 2 and 3, and correct accordingly. Please also add the 95% CI for specificity of Xpert MTB/RIF test in both tables. Should use the value of 95% CI in a consistent format like 0.0% (0.0-0.0) instead of 0.0% (0.0%-0.0%) in the tables and text.

6. For clarity and better understanding authors should include the number of samples detected or not detected by the index tests compared to the reference tests along with the value of sensitivity and specificity. As for example, in Table-2, for ‘Determine TB LAM test’ please write the Sensitivity: 34.8% (34.1-35.5); 8/23 instead of 34.8% (34.1-35.5), and for specificity: 92.7% (86.7-96.6); 115/124 instead of 92.7% (86.7-96.6). Please add the numbers in Table-2 and Table-3 and describe accordingly in the text.

7. Figure 2: please describe in the result section (by mentioning the numbers of samples) how the percentages of different tests were obtained between HIV positive and negative patients.

8. Line 215-217: It is not clear how the sensitivities of 44.4% for Xpert and 30.6% for TB LAM were obtained. Please clarify. Please also mention the statistical methods in ‘Methods and Material’ section that was used to obtain the p value.

9. Recently, Xpert MTB/RIF Ultra and Fujifilm SILVAMP TB-LAM have been appeared with higher sensitivities for detection of tuberculosis. In the discussion section authors should add one paragraph about the future applicability of these two assays compared with the current findings of Xpert MTB/RIF assay and Determine TB LAM assay.

10. Line 344: Please clarify the statement ‘equivalent to culture to diagnose EPTB’. It is not clear how the sensitivity of combined use of Xpert MTB/RIF and TB LAM was equivalent to ‘culture’ as the authors did not mention anything about the sensitivity of ‘culture’ in the manuscript.

11. While the text is readable, there are some grammar mistakes. Please correct the text to improve the comprehensibility.

Minor comments:

Line 37-The authors should mention the name of ‘reference’ test that is the ‘CRS’ to whom the results of Xpert MTB/RIF were compared.

Line 42: Please write ‘EPTB cases’ instead of ‘TB cases’.

Line 41-42: Please correct and re-phrase the sentence ‘The combination of Xpert MTB/RIF and TB LAM detected 61.1% of all EPTB participants and 83.3% of HIV co-infected TB cases’ by mentioning the sensitivity instead of detection.

Line 98: Should write Lowenstein Jensen (L-J) as it appears first in the text.

Line 135: Please mention the volume of urine used for centrifugation.

Authors should specify why the LAM test was performed on refrigerated urine samples instead of freshly collected samples. Should discuss the point in the discussion section whether the sensitivity of LAM test varies between refrigerated versus fresh urine sample.

Line172: Please write as 35 years (IQR, 22-45).

Line-187: ‘Forty-nine (33.3%) of the presumptive EPTB cases had the classification of colitis’. Please clarify whether all of the 49 cases from where peritoneal fluid were collected had ‘colitis’.

Line-195 and 202: Authors should mention ‘detection rate’ instead of ‘prevalence’.

Authors have mentioned in Table-1 that one sample among ‘Others’ group (2 gastric aspirates and 2 synovial fluids) was diagnosed as ‘Confirmed TB’. Please, specifically mention the sample type (either the gastric aspirate or synovial fluid) that was positive for TB in the table and add the description in line-196.

Line 228-229: please write as (0.0%; 95% CI, 0.0-0.0) instead of ‘(30.6% (95%CI, 30%-

229 31%))’ and ‘(93.3% (95%CI, 93.2%-93.5%))’.

Line 230 and 238: Authors should include the ‘p value’ of the significant difference, and the test used.

Line 235-236: Please rephrase the sentence ‘Adding Determine TB LAM test to Xpert MTB/RIF test increases the sensitivity of Xpert MTB/RIF test from 43.2% to 61.1%’ as it does not increase the sensitivity of Xpert MTB/RIF test, but increases the sensitivity of combined use TB LAM and Xpert assay.

Line 307-308: It is not clear how the values of 2 (33%) and 1 (17%) were obtained. Please add in the result section to clearly state how these values were obtained.

Line 320: It is not clear how the sensitivity of L-J culture (62.2%) was obtained. Please specify.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

4 Sep 2021

Response to Editor and Reviewers

We appreciate the editor and reviewers for the constructive comments which we have used to improve the quality of the manuscript. We have re-written some portions of the manuscript accordingly. We have carefully addressed the comments line by line as follows.

Response to Editor’s comments:

Comment 1: Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming

�  Response 1: Thank you for your comment. We have accepted your comments and correction has made accordingly.

Comment 2: Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified whether consent was informed.

�  Response 2: Written informed consent was obtained from each participants before enrolling into our study. This was described method parts.

Comment 3: We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail and who has imposed them. Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent. If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers.

�  Response 3: Thank you for your feedback. There is no legal restriction to share our data set. We uploaded our data set as a supporting information file (S2 File).

Comment 4: PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager.

�  Response 4: Thank you for your comment. This is the ORCID iD of the corresponding author https://orcid.org/0000-0003-4751-2225.

Comment 5: We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

�  Response 5: We agree with reviewer’s comment and the phrase referring the missed data was deleted from the manuscript.

Additional editor question 1: Line 188- 'Forty-nine (33.3%) of the presumptive EPTB cases had the classification of colitis'-On what basis was this classification made and what was the sample collected from these patients?

�  Response1: About 49(33.3%) of the presumptive EPTB cases were classified as colitis based on the site of infection and symptoms. The site of infection for these 49 presumptive EPTB cases were abdomen and almost all of them experienced abdominal discomfort and pain during specimen collection. The sample collected from them were peritoneal fluid. So, due to the site of infection, abdominal discomfort and pain symptoms we have classified them as colitis.

Additional editor question 2: Although 4 of the 23 HIV positive patients were bacteriologically confirmed, it is not clear how many were positive by Xpert MTB/RIF. What is meant by 50% in Line 211? Similarly, for Determine TB LAM assay.

�  Response 2: Thank you for your comment and question. Although not described in the document, out of 23 HIV positive cases 3 of them were positive by Xpert MTB/RIF. Out of 23 HIV positive cases only 21 provided urine specimen and 7 were positive by TB LAM. Out of 23 HIV positive cases 6 were positive for EPTB by composite reference standard (CRS) and out of these 6 cases, only 3 were positive by Xpert MTB/RIF. So, the sensitivity of Xpert MTB/RIF for HIV positive cases were 3/6 (50%). The detection rate and diagnostic performance of Xpert MTB/RIF, TB-LAM and their combination in HIV-positive and –negative patients is now described in the supporting information file (S1 File).

Additional editor question 3: Line 320- How was the sensitivity of LJ culture derived to be 62.2%? What was the gold standard? I think that was not within the scope of the study.

�  Response 3: Of course, determining the sensitivity of LJ culture is not our primary objective. Previous studies have shown that LJ culture has suboptimal sensitivity for diagnosis of EPTB and there are a significant number of cases which are negative on LJ culture but clinically diagnosed as TB. Taking note of this, we proposed to determine the sensitivity of LJ culture using CRS (composite reference standard) as a gold standard. In our study, a total of 37 patients were classified as EPTB based on CRS, of whom only 23 were positive by LJ culture. Thus, the sensitivity of LJ culture was 23/37(62.2%). We kept this just as an additional information.

Additional editor comment 4: Fig 1- The flow chart could start with inclusion of 147 patients suspected of presumptive TB instead of 160 patients.

�  Response 4: Thank you for your comment. We have accepted your comment and it was corrected on the revised manuscript.

Response to Reviewers' comments

Reviewers' comment 1: In the current study the authors aim to investigate the diagnostic performance of combined use of Xpert MTB/RIF and TB LAM Ag test for detection of M. tuberculosis complex among EPTB patients in a medical center of Ethiopia. Considering the high prevalence of EPTB and association of HIV in Ethiopia, the current diagnostic approach provides the important insight for increased detection of EPTB cases in clinical settings. However, the current manuscript needs to address several issues before approval.

�  Response 1: We thank the reviewer for the constructive feedback. We tried our best to address comments/questions raised by the editor and reviewers.

Response to Reviewer #1 Major Comments

Comment 1: According to the manufacturer (Abbot Laboratories, Chicago, US), the test name is ‘Determine™ TB-LAM Ag’. It is suggested to use the exact test name in the Title of the manuscript. Throughout the text, authors should be consistent mentioning the test name. Examples of some discrepancies are: ‘Determine TB LAM test’ in Line-31, ‘TB LAM’ in Line 41, ‘Determine TB-LAM test’ in line 136, ‘TB-LAM assay’ in line 140 and so on. It is suggested to use ‘Determine™ TB-LAM Ag (TB-LAM)’ when appears first in the text, followed by ‘TB-LAM’ all through the text.

�  Response 1: Thank you for your comment. In the revised manuscript, this was corrected.

Comment 2: In the introduction section, please add one paragraph describing the other published data on the performances of the Xpert MTB/RIF assay and Determine TB-LAM assay for detection of EPTB cases.

�  Response 2: We agree with reviewer’s comment and more evidence on the performance of Xpert MTB/RIF and TB-LAM was included in the introduction section of revised version.

Comment 3: One of the EPTB samples as has been described inconsistently throughout the text. For example, in Table-1 as ‘pus (abscess)’, Table-2 and throughout the text as ‘Abscess’. According to Line-191, it appears that samples were obtained from ‘Lymph node TB’. It would be easier for the readers to understand if the term ‘Lymph node aspirate’ is used all through the manuscript instead of ‘Abscess’.

�  Response 3: Thank you for your comment. We agree with reviewer’s comment and it was corrected as lymph node aspirate in the revised manuscript.

Comment 4: Gastric aspirate is considered as an alternative specimen for detection of pulmonary TB when patients cannot expectorate the sputum. As this study aims to detect the EPTB cases, authors should exclude the two ‘gastric aspirate’ samples from the study and re-analyze the results and correct the statistics accordingly in different relevant Tables, Figures and throughout the text.

�  Response 4: Thank you for the interesting feedback. As the reviewer mentioned, in most cases, gastric aspirate is an alternative specimen for detection of pulmonary TB when patients cannot expectorate the sputum. But, in our case, the 2 gastric aspirate specimens were taken from patients who were suspected of having disseminated tuberculosis that is gastro intestinal tuberculosis.

Comment 5: For Table-2 and Table-3: Authors should check carefully that the statistics are presented correctly. It seems that the value of sensitivity, specificity, PPV and NPV and their corresponding 95% CI are not accurate. As for example, in Table-2 (based on the calculation of Fig-1), for ‘Determine TB LAM test’: the sensitivity would be 34.8% with 95% CI ranging from 16.4-57.3, specificity would be 92.7% (86.7-96.6), PPV would be 47.1% (27.7-67.4), and NPV would be 88.5% (85.0- 91.2). Please check both table 2 and 3, and correct accordingly. Please also add the 95% CI for specificity of Xpert MTB/RIF test in both tables. Should use the value of 95% CI in a consistent format like 0.0% (0.0-0.0) instead of 0.0% (0.0%-0.0%) in the tables and text.

�  Response 5: The reviewer’s comments are well taken and were corrected in the revised manuscript. The sensitivity, specificity, positive predictive value and negative predictive value for Xpert MTB/RIF and Smear microscopy were calculated from 147 participants but for TB-LAM the participants were 126.

Comment 6: For clarity and better understanding authors should include the number of samples detected or not detected by the index tests compared to the reference tests along with the value of sensitivity and specificity. As for example, in Table-2, for ‘Determine TB LAM test’ please write the Sensitivity: 34.8% (34.1-35.5); 8/23 instead of 34.8% (34.1-35.5), and for specificity: 92.7% (86.7-96.6); 115/124 instead of 92.7% (86.7-96.6). Please add the numbers in Table-2 and Table-3 and describe accordingly in the text.

�  Response 6: We agree with what the reviewer suggested. In the revised version of the manuscript, the number of samples tested positive or negative by index test compared to the reference standard along with the values of sensitivity, specificity and predictive values are mentioned.

Comment 7: Figure 2: please describe in the result section (by mentioning the numbers of samples) how the percentages of different tests were obtained between HIV positive and negative patients

�  Response 7: Details of the percentages obtained for Xpert MTB/RIF and TB-LAM tests between HIV-positive and –negative cases are well addressed in the revised version of the manuscript. The detection rate and diagnostic performance of TB-LAM, Xpert MTB/RIF and combination of TB-LAM and Xpert MTB/RIF between HIV-positive and -negative participants was also addressed in supporting file (S1 File).

Comment 8: Line 215-217: It is not clear how the sensitivities of 44.4% for Xpert and 30.6% for TB LAM were obtained. Please clarify. Please also mention the statistical methods in ‘Methods and Material’ section that was used to obtain the p value.

�  Response 8: There were 126 patients who provided both site-specific extrapulmonary specimen (for Xpert MTB/RIF) and urine specimen (for TB-LAM). Comparison of Xpert MTB/RIF and TB-LAM was done only for 126 samples. Out of the 126 cases, about 36 were positive for EPTB by composite reference standard (CRS). Xpert MTB/RIF detected only 16 EPTB cases out of the 36 EPTB positive cases. So, the sensitivity of Xpert MTB/RIF test to detect EPTB among 126 cases were 16/36(44.4%). TB-LAM test detects only 12 EPTB cases out of the total EPTB positive cases (36). So, the sensitivity of TB-LAM test to detect EPTB among 126 cases were 12/36(33.3%). The sensitivity of TB-LAM test to detect EPTB among 126 cases was corrected on the revised manuscript from 30.6% to 33.3%.

Comment 9: Recently, Xpert MTB/RIF Ultra and Fujifilm SILVAMP TB-LAM have been appeared with higher sensitivities for detection of tuberculosis. In the discussion section authors should add one paragraph about the future applicability of these two assays compared with the current findings of Xpert MTB/RIF assay and Determine TB LAM assay.

�  Response 9: We appreciate the reviewer for the insightful complement. Regarding the Xpert MTB/RFI Ultra, we have a bit discussed the potential added value of Ultra over the Xpert MTB/RIF particularly in paucibacillary samples in the second paragraph of discussion. As we have mentioned in the discussion, the better performance of Ultra over Xpert MTB/RIF is mainly due to its lower analytical limit of detection (LLD;15 CFU/mL for Ultra versus 100–120 CFU/mL for Xpert MTB/RIF). Somewhere in the discussion (at the end of 4th paragraph), we also recommend the potential use of Xpert Ultra, when Xpert MTB/RIF yielded a negative result in body fluids such as pleural and peritoneal fluids. Regarding, the diagnostic yield of FujiLAM over TB-LAM, we also entertained this in the seventh paragraph of discussion as shown below’

“In their recent study, Kerkhoff et al. reported substantially higher sensitivity of Fujifilm SILVAMP TB-LAM (FujiLAM) over TB-LAM for detecting EPTB in HIV inpatients with moderate sensitivity in pleural fluid and CSF. This suggests a potential use of FujiLAM as a first-line test for the rapid detection of EPTB in HIV patients, with substantial added benefit in paucibacillary diseases such as pleural TB and TB meningitis.’’

Comment 10: Line 344: Please clarify the statement ‘equivalent to culture to diagnose EPTB’. It is not clear how the sensitivity of combined use of Xpert MTB/RIF and TB LAM was equivalent to ‘culture’ as the authors did not mention anything about the sensitivity of ‘culture’ in the manuscript.

�  Response 10: Thank you for your comment. In our study, LJ culture was positive for 23 cases out of the total 147 presumptive EPTB cases. Composite reference standard (CRS) detected 37 EPTB cases out of the total 147 presumptive EPTB cases. So, the sensitivity of L-J culture was 62.2% (23/37) that is equivalent to the previously calculated sensitivity of combined use of Xpert MTB/RIF and TB LAM test (61.1%). In the revised manuscript, we have mentioned the sensitivity of LJ culture.

Comment 11: While the text is readable, there are some grammar mistakes. Please correct the text to improve the comprehensibility.

�  Response 11: Thank you for your comment. We have thoroughly gone through the manuscript and edited it for grammar mistakes.

Response to Reviewer #1 Minor Comments

Line 37:

Comment 1: The authors should mention the name of ‘reference’ test that is the ‘CRS’ to whom the results of Xpert MTB/RIF were compared.

�  Response 1: We have accepted your comment and it was mentioned in the revised manuscript.

Line 42:

Comment 2: Please write ‘EPTB cases’ instead of ‘TB cases’

�  Response 2: Comment accepted and corrected in the revised manuscript.

Line 41-42:

Comment 3: Please correct and re-phrase the sentence ‘The combination of Xpert MTB/RIF and TB LAM detected 61.1%of all EPTB participants and 83.3% of HIV co-infected TB cases’ by mentioning the sensitivity instead of detection.

�  Response 3: The reviewers’ comments are noted and corrected in the revised manuscript.

Line 98:

Comment 4: Should write Lowenstein Jensen (L-J) as it appears first in the text.

�  Response 4: Comment accepted and corrected on the revised manuscript.

Line 135:

Comment 5: Please mention the volume of urine used for centrifugation.

Authors should specify why the LAM test was performed on refrigerated urine samples instead of freshly collected samples. Should discuss the point in the discussion section whether the sensitivity of LAM test varies between refrigerated versus fresh urine sample.

�  Response 5: As it has been mentioned in the Materials and Methods section, under the subheading of study subjects, sample collection and laboratory tests, all participants were requested to submit 10-30ml of urine sample, which was immediately refrigerated upon arrival to Mycobacteriology Research Center. Since there was a delay in the shipment process of TB-LAM kit, in our study TB-LAM test was done on the refrigerated urine samples. Though not reproduced by other researchers, Peter et al. has reported that the use of frozen urine has been associated with reduced TB- LAM sensitivity. We have mentioned the use of frozen urine sample as the possible limitation of our study and we have indicated that the low sensitivity of TB-LAM in our study could be partly because of the fact that the test was done on frozen urine sample.

Line172:

Comment 6: Please write as 35 years (IQR, 22-45).

�  Response 6: Comment noted and corrected on the revised manuscript.

Line-187:

Comment 7: ‘Forty-nine (33.3%) of the presumptive EPTB cases had the classification of colitis’. Please clarify whether all of the 49 cases from where peritoneal fluid were collected had ‘colitis’

�  Response 7: Thank you for your comment. Out of 147, about 49(33.3%) of the presumptive EPTB cases were classified as colitis based on the site of infection and symptoms. Peritoneal fluid was collected from all (49) cases and all of them experienced abdominal discomfort and pain during specimen collection so that all of them were diagnosed as colitis.

Line-195 and 202:

Comment 8: Authors should mention ‘detection rate’ instead of ‘prevalence’. Authors have mentioned in Table-1 that one sample among ‘Others’ group (2 gastric aspirates and 2 synovial fluids) was diagnosed as ‘Confirmed TB’. Please, specifically mention the sample type (either the gastric aspirate or synovial fluid) that was positive for TB in the table and add the description in line-196.

�  Response 8: We also agree with reviewers’ comments and considered in the revised version of the manuscript.

Line 228-229:

Comment 9: please write as (0.0%; 95% CI, 0.0-0.0) instead of ‘(30.6% (95%CI, 30%- 229 31%))’ and ‘(93.3% (95%CI, 93.2%-93.5%))’.

�  Response 9: We agree with reviewer’s suggestion and corrected accordingly.

Line 230 and 238:

Comment 10: Authors should include the ‘p value’ of the significant difference, and the test used.

�  Response 10: Comment well accepted and corrected in the revised manuscript.

Line 235-236:

Comment 11: Please rephrase the sentence ‘Adding Determine TB LAM test to Xpert MTB/RIF test increases the sensitivity of Xpert MTB/RIF test from 43.2% to 61.1%’ as it does not increase the sensitivity of Xpert MTB/RIF test, but increases the sensitivity of combined use TB LAM and Xpert assay.

�  Response 11: Thank you for the interesting comment. We have now rephrased the sentence as “The combined use of TB- LAM and Xpert MTB/RIF tests had the sensitivity of 61.1% when compared to CRS, which was significantly higher than the sensitivity of each test alone”

Line 307-308:

Comment 12: It is not clear how the values of 2 (33%) and 1 (17%) were obtained. Please add in the result section to clearly state how these values were obtained.

�  Response 12: We also agree with reviewer’s feedback. We found that it is better to describe the diagnostic performance of Xpert MTB/RIF, TB-LAM and their combined use among HIV-positive and negatives case, and we added one subheading in the result section. By including this, we assumed we have addressed some of questions/concerns raised from the reviewer.

Line 320:

Comment 13: It is not clear how the sensitivity of L-J culture (62.2%) was obtained. Please specify.

�  Response 13: We have already addressed this question above in Response 10 under Major comments. To clarify it once again, of the total 37 EPTB cases diagnosed by CRS (the reference standard method used in our study), 23 were identified as EPTB positive by L-J culture. Using CRS (definite and probable-EPTB cases) as a gold standard, the sensitivity of L-J culture was 62.2% (23/37).

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

1 Dec 2021

28 Dec 2021

Response to Reviewers

We appreciate the reviewers for the constructive comments which we have used to improve the quality of the manuscript. As we did in the last time, we have re-written some portions of the manuscript accordingly. We have also carefully addressed the reviewers’ comments line by line as follows:

Response to Reviewers' comments

Reviewer 1

Comment 1: It is appreciated that the authors have responded well the comments. However, it would be easier for the reviewer to go through the text if the specific line numbers are mentioned against the comments (where appropriate) in the cleaned version.

�  Response 1: Sorry for not mentioning the specific line numbers against the comments in the last cleaned version. We agree with the reviewer’s comment. Now, in the recent version, the specific line numbers are indicated against the comments in the cleaned version.

Comment 2: From Table 2, it is evident that TB-LAM test was compared against 36 and Xpert MTB/RIF test against 37 positive samples found by CRS. Would be easier for the reader if it is explained in short in the result section how these CRS have been found.

�  Response 2: In general, as indicated in the method section (Line 166-169 ), CRS, which comprises of smear microscopy, L-J culture and clinical improvement after ATT initiation, was used as gold standard to calculate the sensitivity, specificity and predictive values of both TB-LAM and Xpert MTB/RIF. Specifically for calculation of sensitivity, as it was stated by the reviewer, TB-LAM test was compared against 36 CRS-positive cases whereas Xpert MTB/RIF was compared against 37 CRS-positive cases. This is now addressed in the revised version of the manuscript. See Line224-226 of cleaned version of the manuscript.

Comment 3: Line 227: the author should write ‘sensitivity’ instead of ‘pooled sensitivity’

�  Response 3: We have modified it accordingly. Shown in Line 232.

Comment 4: Line 179-180: ‘MedCalc Software Ltd (Comparison of proportions calculator) was used……….’ this sentence can be written as ‘Comparison of proportion between the methods was done by Chi square test using the ‘MedCalc Software’ (please include the weblink).

�  Response 4: We agree with the reviewer’s feedback and we have modified it accordingly as shown Line 177-180 of cleaned version of the manuscript .

Reviewer 2

Comment 1: The referencing style need to be uniform throughout the manuscript in accordance with Journal’s requirement. For example, please note the discrepancy in citing reference no. 1, 7, 13, 20, 24

o Response 1: The reviewer’s comment is accepted and corrected accordingly.

Comment 2: In ABSTRACT;

�  Comment 2.1: Line 29 of the revised manuscript: Replace “examined for tuberculosis (TB)” with “tested for Mycobacterium tuberculosis complex (MTBC)”.

o Response 2.1: Comment accepted and corrected accordingly. Shown in Line 27 of cleaned version.

�  Comment 2.2: Line 40 of the revised manuscript: Write “and 100% specificity” instead of “with 100% specificity”.

o Response 2.2: Comment accepted and corrected accordingly. Shown in Line 38 of cleaned version.

Comment 3: In BACKGROUND

�  Comment 3.1: Line 60 of the revised manuscript: Replace “category III biosafety level” with “bio-safety level III facilities”.

o Response 3: We agree with the reviewer’s suggestion and corrected accordingly. Shown in Line 58 of cleaned version.

Comment 4: In MATERIALS AND METHODS

�  Comment 4.1: Line 106-107 of the revised manuscript: There is no mention about collection of ‘gastric aspirate’ and ‘synovial fluid’.

o Response 4.1: Gastric aspirate and synovial fluid samples were collected as per the standard of care. Information on the collection of synovial fluid and gastric aspirate is shown in Line 104-105 of cleaned version.

�  Comment 4.2: Line 121-122 of the revised manuscript: Was decontamination also carried out for gastric aspirate specimens?

o Response 4.2: Of course, like that of lymph node aspirates and blood stained samples, gastric aspirate samples were decontaminated by the standard N-acetyl-L-cysteine and sodium hydroxide (NALC/NaOH) to reduce overgrowth from contaminating microorganism. This was addressed in the revised cleaned version. See Line 119 of cleaned version

�  Comment 4.3: Line 179-180 of the revised manuscript: Earlier minor comment no. 10 by the reviewer is not addressed adequately. The tool used in determining the statistical difference has been mentioned but the statistical test used for the same has not been mentioned in the revised version of the manuscript.

o Response 4.3: As we have explained in the revised version (Line 177-180), presence or absence of statistically significant difference is determined based on the P-value. The P-value<0.05 showed statistically significant difference whereas the P-value>0.05 showed no difference (similar, not merely the same).

Comment 5: In RESULTS

�  Comment 5.1: Line 206-107 of the revised manuscript: With respect to addition editor’s question no. 1 and the reply by the authors, will it be better if the authors mention the ‘colitis’ cases as ‘suspected abdominal TB’ cases because abdominal pain and discomfort could be due to causes other than colitis?

o Response 5.1: We agree with the feedback from the reviewer and corrected it. See Line 206-207 of cleaned version.

�  Comment 5.2: Line 215-216 of the revised manuscript: “1 disseminated TB from gastric aspirate”- It remains dubious how detecting MTBC from gastric aspirate only can label the case as ‘disseminated TB’. Clarification is needed.

o Response 5.2: This is an interesting observation from the reviewer. We strongly agree with the reviewer’s concern and one MTBC case from gastric aspirate cannot signify disseminated TB. However, pulmonary manifestation was excluded from this patient and we assumed this patient had extrpulmonary form of TB. Accordingly, we revised the manuscript as shown in Line 215-217 of cleaned version.

�  Comment 5.3: Line 242-243 of the revised manuscript: It is not clear whether culture confirmation (23 patients) could only be done only among 126 presumptive EPTB cases who submitted urine samples for TB-LAM. In other words, was there no culture confirmed cases among the other 21 presumptive EPTB cases who did not submit urine sample for TB-LAM? This needs to be clarified.

o Response 5.3: This is again another interesting comment from the reviewer. The other reviewer also raised similar concern. There was one culture-positive case from presumptive cases who did not submit urine sample. That could be why we excluded this case from CRS when determining the diagnostic accuracy of TB-LAM. As it was shown in Table 2, the TB-LAM test was compared against 36 CRS positive cases whereas Xpert MTB/RIF test was compared against 37 CRS positive cases. For combination of TB-LAM and Xpert MTB/RIF, the 36 cases who have all the three test results were considered. Shown in Line 224-226 of cleaned version.

Comment 6: In DISCUSSION

�  Comment 6.1: Line 316 of the revised manuscript: Replace ‘Additionally’ with ‘However’.

o Response 6.1: We agree. This is corrected in the revised version of the manuscript as shown in Line 315.

�  Comment 6.2: Line 355 of the revised manuscript: Write ‘End TB strategy’ instead of ‘end TB strategy’.

o Response 6.2: It is written as End TB strategy (Line 358) in the revised version of the manuscript.

Comment 7: PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

�  Response 7: We are OK with the option to publish the peer review history.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

14 Jan 2022

Combination of Xpert® MTB/RIF and DetermineTM TB-LAM Ag improves the diagnosis of extrapulmonary tuberculosis at Jimma University Medical Center, Oromia, Ethiopia

PONE-D-21-10435R2

Dear Dr. Tadesse,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Shampa Anupurba, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Arghya Das, MD

25 Jan 2022

PONE-D-21-10435R2

Combination of Xpert MTB/RIF and DetermineTM TB-LAM Ag improves the diagnosis of extrapulmonary tuberculosis at Jimma University Medical Center, Oromia, Ethiopia

Dear Dr. Tadesse:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Shampa Anupurba

Academic Editor

PLOS ONE

Table 1

Socio demographic and clinical characteristics of presumptive EPTB cases visited Jimma University Medical Center from April to October 2019(N = 147).

Characteristics		All,(N = 147)	Confirmed TB(n = 23)	Probable TB (n = 14)	Non TB (n = 110)
Sex	Male	82(55.8%)	10(43.5%)	11(78.6%)	61(55.5%)
	Female	65(44.2%)	13(56.5%)	3(21.4%)	49(44.5%)
Residence	Urban	63(42.9%)	12(52.255)	6(42.9%)	45(40.9%)
	Rural	84(57.1%)	11(47.8%)	8(57.1%)	65(59.1%)
Age (years)	0–15	24(16.3%)	4(17.4%)	2(14.3%)	18(16.4%)
	16–30	35(23.8%)	10(43.5%)	3(21.4%)	22(20%)
	31–45	52(35.3%)	6(26.1%)	4(28.6%)	42(38.2%)
	>45	36(24.5%)	3(13.0%)	5(35.7%)	28(25.5%)
Types of specimen	Peritoneal fluid	49(33.3%)	2(8.7%)	6(42.9%)	41(37.3%)
	Pleural fluid	45(30.6%)	4(17.4%)	3(21.4%)	38(34.5%)
	Cerebrospinal fluid	28(19%)	3(13%)	3(21.4%)	22(20%)
	Lymph node aspirate	19(12.9%)	13(56.5%)	1(7.1%)	5(4.5%)
	Pericardial fluid	2(1.4%)	0	0	2(1.8%)
	Others*	4(2.7%)	1(4.3%)	1(7.1%)	2(1.8%)
HIV status	Positive	23(15.6%)	4(17.4%)	2(14.3%)	17(15.5%)
	Negative	111(75.5%)	16(69.6%)	12(85.7%)	83(75.5%)
	Unknown	13(8.8%)	3(13%)	0	10(9.1%)

*Others include 2 gastric aspirates and 2 synovial fluid specimens. Of these, 1 gastric aspirate specimen was positive on L-J culture.

Table 2

Diagnostic performance of TB-LAM and Xpert MTB/RIF test.

Diagnostic Accuracy	Culture as a reference standard
	TB-LAM test	Xpert MTB/RIF test	Combination of TB-LAM and Xpert MTB/RIF tests
Sensitivity (95%CI)	34.8% (16.4–57.3);8/23	69.6% (47.1–86.8);16/23	82.6% (61.2–95.1);19/23
Specificity (95%CI)	91.3% (84.1–95.9);94/103	100% (97.1–100);124/124	92.2% (85.3–96.6);95/103
PPV (95%CI)	47.1% (27.8–67.3);8/17	100%(79.4–100); 16/16	70.4% (54.3–82.6);19/27
NPV (95%CI)	86.2% (82.2–89.5);94/109	94.7% (90.5–97.1);124/131	96. % (90.7–98.3);95/99
CRS as a reference standard
Diagnostic Accuracy	TB-LAM test	Xpert MTB/RIF test	Combination of TB-LAM and Xpert MTB/RIF tests
Sensitivity (95%CI)	33.3% (18.6–51);12/36	43.2% (27.1–60.5);16/37	61.1% (43.5–76.9);22/36
Specificity (95%CI)	94.4% (87.5–98.2);85/90	100% (96.7–100);110/110	94.4% (87.5–98.2);85/90
PPV (95%CI)	70.6% (47.7–86.4);12/17	100% (79.4–100);16/16	81.5% (64.4–91.5);22/27
NPV (95%CI)	78% (73.7–81.8);85/109	84% (79.8–87.4);110/131	85.9% (80.1–90.2);85/99

Key: PPV; Positive predictive value, NPV; Negative predictive value.

Table 3

Performance of Xpert MTB/RIF test among different non respiratory samples.

		Xpert MTB/RIF		CRS
Sample type	Total	Positive	Negative	Positive	Negative	Sensitivity (95%CI)	Specificity (95%CI)
Lymph node	19	12	7	14	5	85.7% (57.2–98.2);12/14	100% (47.8–100);5/5
CSF	28	2	26	6	22	33.3% (4.3–77.7);2/6	100%(84.6–100);22/22
Pleural fluid	45	1	44	7	38	14.3% (0.4–57.9);1/7	100%(90.6–100);38/38
Other*	4	1	3	2	2	-	-

Note: other* includes (2 gastric aspirate and 2 synovial fluid). All peritoneal fluid and pericardial fluid specimens were negative for Xpert MTB/RIF test.

Key: CRS; Composite reference standard.