SPINE
GeneXpert polymerase chain reaction for spinal tuberculosis AN ACCURATE AND RAPID DIAGNOSTIC TEST M. Held, M. Laubscher, H. J. Zar, R. N. Dunn From Groote Schuur and Red Cross Children’s Hospitals, Cape Town, South Africa
The lack of an accurate, rapid diagnostic test for mycobacterium tuberculosis (TB) is a major handicap in the management of spinal TB. GeneXpert, a new, rapid molecular diagnostic test is recommended as the first line investigation for suspected pulmonary TB in areas with a high prevalence of HIV or drug resistance, yet it has not been validated for the diagnosis of musculoskeletal TB. The aim of this study was to assess the accuracy of GeneXpert in diagnosing spinal TB. A prospective clinical study of 69 consecutive adults with suspected spinal TB was conducted at a tertiary hospital in an area with the highest incidence and prevalence of TB in the world. GeneXpert was used on tissue samples of the enrolled patients and its diagnostic accuracy compared with a reference standard of tissue in liquid culture. A total of 71 spine samples from 69 patients (two re-biopsies) were included in the study. The GeneXpert test showed a sensitivity of 95.6% and specificity of 96.2% for spinal TB. The results of the GeneXpert test were available within 48 hours compared with a median of 35 days (IQR 15 to 43) for cultures. All cases of multi-drug resistant TB (MDR TB) were diagnosed accurately with the GeneXpert test. The MDR TB rate was 5.8%. Cite this article: Bone Joint J 2014;96-B:1366–9.
M. Held, FCOrtho (SA), Dr Med, Orthopaedic Surgeon M. Laubscher, FCOrtho (SA), MBChB, Orthopaedic Surgeon R. N. Dunn, FCSOrtho (SA), MBChB, Orthopaedic Surgeon, Professor Department of Orthopaedic Surgery, University of Cape Town, H49 Old Main Building, Groote Schuur Hospital, Observatory 7925, South Africa. H. J. Zar, MBBCh, FCPaeds, PhD, Paediatric Pulmonologist, Professor Department of Pediatrics and Child Health, Red Cross War Memorial Children’s Hospital, University of Cape Town, South Africa. Correspondence should be sent to Dr M. Held; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B10. 34048 $2.00 Bone Joint J 2014;96-B:1366–9. Received 27 February 2014; Accepted after revision 8 July 2014
1366
An accurate diagnosis of musculoskeletal tuberculosis (TB) is essential to enable its timely and effective treatment. It is also important to have the results readily available so that treatment can be started. In spinal TB in particular, diagnostic delay can lead to devastating neurological complications and irreversible disability.1 GeneXpert, an automated polymerase chain reaction (PCR) diagnostic test that simultaneously detects TB and rifampicin resistance, has recently become available.2-4 The World Health Organization (WHO) has endorsed GeneXpert for use in patients with suspected pulmonary TB in areas of high HIV or drug resistance.5 it has been quoted as having similar sensitivity and specificity to conventional testing methods, such as TB culture.2-4,6 In a recent meta-analysis of 18 studies with 7816 patients by the Cochrane Infectious Diseases Group, GeneXpert was tested on the sputum of patients with suspected pulmonary TB: a sensitivity of approximately 90% with a false negative rate of 2% was reported, the sensitivity for rifampicin resistance being 94%.7 According to the WHO Global Tuberculosis Report 2013, the GeneXpert MTB/RIF test will soon be universally available.5 There is, however, no data on its accuracy in the diagnosis of musculoskeletal TB.
The aim of this study was to investigate the accuracy of GeneXpert in diagnosing spinal TB.
Materials and Methods This prospective study was conducted over a period of nine months between March and December 2013. It included samples taken from patients with suspected spinal TB who presented to the spinal unit of a tertiary care hospital. This hospital services an area which has the highest prevalence and incidence of TB worldwide.5 Tissue biopsies were taken for liquid culture, acid-fast smear and GeneXpert. The patients were followed up every three months until there were signs of clinical improvement and nine months of treatment for TB had been completed. The diagnostic performance of the GeneXpert assay was compared with that of standard tissue microscopy and liquid culture, which is the definitive test for osteoarticular TB. Index test. Spinal tissue was processed for GeneXpert (Xpert MTB/RIF) (Cepheid, Sunnyvale, California) testing. GeneXpert lysis buffer was added in a 1:3 ratio to 2 mls of specimen. The mixture was mixed by use of rapid oscillation, or vortexed, and allowed to stand for ten minutes, then vortexed again. A total of 2 mls were removed and processed automatically with the THE BONE & JOINT JOURNAL
GENEXPERT POLYMERASE CHAIN REACTION FOR SPINAL TUBERCULOSIS
1367
Table I. Clinical and radiological red flags on which the suspected diagnosis of TB was made. Clinical red flags
Radiological red flags
History of chronic pain for more than three months Constitutional symptoms: low grade fever, night sweats, loss of appetite, weight loss Chronic cough Elevated erythrocyte sedimentation rate History of tuberculosis contact Local kyphosis9 Neurological deficit Positive Mantoux test10 Immune compromise/HIV
Loss of anterior vertebral height Paravertebral shadow on radiograph
GeneXpert instrument and the result was read after approximately 90 minutes. Reference standard. The automated BACTEC MGIT 960 by BD Diagnostic Systems, Sparks MD (Franklin Lakes, New Jersey) was used for mycobacterial cultures. Direct microscopy for TB was performed with a Ziehl–Nielson (ZN) stain for microbiology and haematoxylin and eosin (HE) stain for histopathology. Specifications from the Centres for Disease Control and Prevention were used for quantification of acid-fast bacilli. STARD (Standards for the Reporting of Diagnostic Accuracy studies) guidelines were followed.8 Identification of Multi-Drug Resistant TB, (MDR-TB) was performed using the GenoType MTBDRplus or GenoType Mycobacterium CM lineprobe assays (Hain Lifescience, Nehren, Germany). Definitions. ‘Definite TB’ was defined as a positive culture of mycobacterium tuberculosis. In patients with negative cultures, ‘Probable TB’ was assumed if the histology, clinical and radiological findings were suggestive of TB and if the patients improved on anti-tuberculous treatment. The clinical and radiological features that were used as red flags are listed in Table I.9,10 ‘Not TB’ was assumed in patients who tested negative on both culture and histology and who improved without anti-tuberculous treatment. ‘MDR TB’ is defined according to the WHO as resistance to isoniazid and rifampicin, with or without resistance to other first-line drugs.11 GenoType MTBDRplus or GenoType Mycobacterium lineprobe assays (Hain Lifescience) were used to confirm resistance to isoniazid and rifampicin. Technical specifications. The samples were collected under strict aseptic conditions in the operating theatre and were taken from predetermined areas of granuloma and destructive disease of the spine, which had been established by MRI and radiological imaging studies. Percutaneous biopsies were reserved for lesions below the level of T1 (first thoracic vertebra) in patients who did not need drainage or stabilisation. A percutaneous transpedicular biopsy was performed in the thoracic spine and a percutaneous paraspinal biopsy in the lumbar spine. An open biopsy was performed when debridement of granulomata and drainage of paravertebral abscesses or stabilisation of the spinal column were needed. Indications for open VOL. 96-B, No. 10, OCTOBER 2014
Shadow of a psoas abscess Adjacent vertebral endplate changes with preserved disc height Changes on chest radiograph indicate suspected tuberculosis Paravertebral abscess on MRI
biopsy included instability of the spinal column; acute deterioration of the neurological signs; a large paravertebral abscess and airway compromise in lesions of the cervical spine. Abscesses were drained through a costotransversectomy in the thoracic spine and through an anterolateral retroperitoneal approach in the lumbar spine. A thoracotomy was performed when stabilisation of the anterior column of the thoracic spine was needed. The Smith–Robinson approach12 was used for lesions of the cervical spine. Biopsy material included pus swabs and/or tissue biopsies from diseased bone or soft-tissue. All samples were subjected to GeneXpert Assay and histopathological examination, as well as microbiological testing. Approval of the hospital’s ethics committee was acquired and all patients gave informed consent. Statistical analysis. The sensitivity, specificity and predictive values of GeneXpert with 95% confidence intervals (95% CI) were determined using a positive culture for definite TB (mycobacterium tuberculosis) and a positive histology in cases of negative cultures (probable TB) as the reference standard. The data was analysed as per sample using STATA 11 statistical software (STATA Corporation, College Station, Texas). Simple descriptive statistics were used to characterise the study population. Normally distributed continuous data were summarised by mean and 95% CI and non-normally distributed continuous data by median and interquartile range. Categorical data were summarised as proportions with 95% CI. Statistical tests included two-sample t-test of proportions, chi-squared test and Wilcoxon rank-sum test. All statistical tests were twosided at α = 0·05.
Results A total of 71 spine samples from 69 patients (two rebiopsies) were included in the study. These were taken from 31 men (44.9%) and 38 women whose median age was 40 years (IQR 27 to 60). Table II shows the incidence of HIV, the gender and age of the patients along with the results of using GeneXpert. Table III shows the per sample analysis of all patients with either culture-positive (‘definite TB’) or histology-positive (‘probable TB’) samples. There were two false negative and one false positive GeneXpert samples in this analysis. A secondary analysis was
1368
M. HELD, M. LAUBSCHER, H. J. ZAR, R. N. DUNN
Table II. Baseline characteristics of the patient cohort (TB, tuberculosis)
N HIV + (%) Gender (male (%)) Age (median (IQR)) GeneXpert +(%)
Overall
Definite TB
Probable TB
Not TB
p-value
69 22 (31.9) 31 (44.9) 40 (27 to 60) 44/69 (63.8)
36 13 (36.1) 14 (38.9) 31 (27 to 46) 35/36 (97.2)
8 4 (50.0) 6 (75.0) 37.5 (23 to 57.5) 8/8 (100.0)
26 5 (20.0) 11 (44.0) 51 (40 to 63) 1/25 (4.0)
0.28 0.28 0.18 0.06
Table III. Per sample accuracy of GeneXpert for detecting tuberculosis (TB) (PPV: positive predictive value (NPV, negative predictive value) Reference standard
All (N)
Sensitivity % (n)
Specificity % (n)
PPV % (n)
NPV % (n)
TB culture or histology positive TB culture positive only
71 71
95.6 (44/46) 92.6 (25/27)
96.2 (25/26) 79.5 (35/44)
97.7 (44/45) 73.5 (25/34)
92.6 (25/27) 94.6 (35/37)
undertaken to compare GeneXpert with culture positive samples only (definite TB) as a reference standard. Culture detected 37 TB patients and histology a further eight. The rate of TB detection with GeneXpert compared with cultures alone increased by 21.6% (8 of 37 samples). This shows that the eight samples which were only detected histologically (probable TB) were likely to be false negative TB cultures Table III. The results of the GeneXpert test were available within 48 hours compared with a median of 35 days (IQR, 15 to 43) for culture. The GeneXpert test showed rifampicin-resistant TB in four patients (5.8%) and ‘possible MDR TB’ was therefore suspected. All four patients were started on the correct MDR TB treatment the day after their biopsy. In three of the four, MDR TB was confirmed with sensitivity testing for rifampicin and isoniazid once the culture was available. In one of the four GeneXpert samples which showed rifampicin resistance, the TB culture was negative and therefore MDR TB may have been missed without the GeneXpert test. Conversely, in another GeneXpert test sample, isoniazid mono-resistance was evident only on culture as the GeneXpert tests solely for rifampicin resistance.
Discussion With the advent of PCR methods to test for TB, various studies have reported that the sensitivity of nonautomated PCR tests range from 61% to 83%.13-18 The PCR method has the advantage of needing much smaller quantities of M. tuberculosis to identify infection and gives a rapid diagnosis by amplification of the nucleic acid of the pathogen. Its detection limit of 130 colony-forming units (CFU) per millilitre, compared with 10 000 CFU/ml in cultures, increases the probability of diagnosing extrapulmonary tuberculosis even with samples containing only a few bacteria.19,20 In this study, the sensitivity of the GeneXpert test was 95.6%, the specificity 96.2%, the positive predictive value 97.7% and negative predictive
value 92.6%. A total of four patients with MDR TB were accurately diagnosed with the GeneXpert test. One of these had a false-negative culture result which was only reported 43 days after biopsy. These patients would have been discharged into the community without adequate treatment for 43 days, and in one patient without treatment at all, had it not been for the GeneXpert test. Another major benefit of the GeneXpert test is the speed with which the results become available. The automated process of the GeneXpert device takes only about 90 minutes to complete. Also, compared with other PCR testing methods, the automated, cartridge-based test obviates the need for extensive laboratory support. This makes it a tool which can be moved closer to smaller hospitals and clinics, cutting down the long transport times often present in countries with a high burden of TB. However, Monni et al21 have reported that there is no difference in the time taken to produce a report compared with conventional testing methods, when the GeneXpert device is not on site. One of the disadvantages of PCR testing is that, in contrast to TB culture, PCR testing will give a positive result even if the pathogens are not viable.2 In these patients, active TB has to be confirmed clinically and by means of various imaging modalities. Another concern about the GeneXpert method is that it tests drug resistance only for rifampicin, therefore a mono-resistance for isoniazid can only be detected with other PCR testing methods, such as GenoType MTBDRplus or GenoType Mycobacterium CM lineprobe assays (Hain Lifescience, Nehren, Germany). In our series, we discovered one patient with isoniazid monoresistance. According to current treatment guidelines isoniazid mono-resistance does not alter the conventional TB regime for patients without isoniazid mono-resistance.22 In conclusion, the GeneXpert test is more sensitive and gives faster access to results than the gold standard. Consequently, we recommend its use as the initial test in the diagnosis of patients with TB of the spine. THE BONE & JOINT JOURNAL
GENEXPERT POLYMERASE CHAIN REACTION FOR SPINAL TUBERCULOSIS We appreciate the support of C. Bamford with the laboratory setup for this study. N. Beylis has provided us with special insight regarding GeneXpert and TB diagnostics and methods. We thank L. Workman for her assistance with the statistical analysis. No benefits in any form have been received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by S. P. H. Hughes and first proof edited by A. Ross.
References 1. Jain AK. Treatment of tuberculosis of the spine with neurologic complications. Clin Orthop Relat Res 2002;398:75–84. 2. Rachow A, Zumla A, Heinrich N, et al. Rapid and accurate detection of Mycobacterium tuberculosis in sputum samples by Cepheid Xpert MTB/RIF assay--a clinical validation study. PLoS One 2011;6:20458. 3. Hillemann D, Rüsch-Gerdes S, Boehme C, Richter E. Rapid molecular detection of extrapulmonary tuberculosis by the automated GeneXpert MTB/RIF system. J Clin Microbiol 2011;49:1202–1205. 4. Theron G, Peter J, Dheda K. Xpert MTB/RIF test for tuberculosis. Lancet 2011;6:378–481. 5. No authors listed. WHO tuberculosis report 2014. http://www.who.int/mediacentre/factsheets/fs104/en (date last accessed 15 January 2014). 6. Peters D, Theron G, Peter J, Dheda K. Should Xpert® MTB/RIF be rolled out in low-income countries? Int J Tuberc Lung Dis 2012;16:702–703. 7. Steingart KR, Sohn H, Schiller I, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev 2013;1:CD009593. 8. No authors listed. STARD checklist. http://www.stard-statement.org/ checklist_maintext.htm (date last accessed 8 August 2014). 9. Rajasekaran S. The problem of deformity in spinal tuberculosis. Clin Orthop Relat Res 2002;398:85–92.
VOL. 96-B, No. 10, OCTOBER 2014
1369
10. No authors listed. NICE guidelines for the diagnosis of TB: http://www.nice.org.uk/ guidance/cg117/chapter/guidance (dated last accessed 8 August 2014). 11. No authors listed. WHO website on MDR TB: http://www.who.int/tb/challenges/ mdr/tdrfaqs/en/ (date last accessed 8 July 2014). 12. Smith GW, Robinson RA. The treatment of certain cervical-spine disorders by antero removal of the intervertebral disc and interbody fusion. J Bone Joint Surg [Am] 1958;40-A:607–624. 13. Davies A, Newport L, Billigton O, Gillespie S. Length of time to laboratory diagnosis of Mycobacterium tuberculosis infection: comparison of in-house methods with reference laboratory results. J Infect 1999;39:205–208. 14. Negi SS, Gupta S, Khare S, Lal S. Comparison of various microbiological tests including polymerase chain reaction for the diagnosis of osteoarticular tuberculosis. Indian J Med Microbiol 2005;23:245–248. 15. Sun Y, Zhang Y, Lu Z. Clinical study of polymerase chain reaction technique in the diagnosis of bone tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 1997;20:145148.[Article in Chinese]:. 16. Tiwari V, Jain A, Verma RK. Application of enzyme amplified mycobacterial DNA detection in the diagnosis of pulmonary and extrapulmonary tuberculosis. Indian J Med Res 2003;118:224–228. 17. Pandey V, Chawla K, Acharya K, Rao S, Rao S. The role of polymerase chain reaction in the management of osteoarticular tuberculosis. Int Orthop. 2009;33:801–805. 18. Rasit AH, Razak M, Ting FS. The pattern of spinal tuberculosis in Sarawak General Hospital. Med J Malaysia 2001;56:143–150. 19. Singh KK, Muralidhar M, Kumar A, et al. Comparison of in house polymerase chain reaction with conventional techniques for the detection of Mycobacterium tuberculosis DNA in granulomatous lymphadenopathy. J Clin Pathol 200;53:355–361. 20. Kramer N, Rosenstein ED. Rheumatologic manifestations of tuberculosis. Bull Rheum Dis 1997;46:5–8. 21. Monni T, Visser HF, Motsitsi SN. Clinical utility of tissue polymerase chain reaction in the diagnosis of spinal tuberculosis. SA Orthop J 2012;11:23–27. 22. Cattamanchi A, Dantes RB, Metcalfe JZ, et al. Clinical characteristics and treatment outcomes of patients with isoniazid-monoresistant tuberculosis. Clin Infect Dis 2009;48:179–185.
GeneXpert polymerase chain reaction for spinal tuberculosis AN ACCURATE AND RAPID DIAGNOSTIC TEST M. Held, M. Laubscher, H. J. Zar, R. N. Dunn From Groote Schuur and Red Cross Children’s Hospitals, Cape Town, South Africa
The lack of an accurate, rapid diagnostic test for mycobacterium tuberculosis (TB) is a major handicap in the management of spinal TB. GeneXpert, a new, rapid molecular diagnostic test is recommended as the first line investigation for suspected pulmonary TB in areas with a high prevalence of HIV or drug resistance, yet it has not been validated for the diagnosis of musculoskeletal TB. The aim of this study was to assess the accuracy of GeneXpert in diagnosing spinal TB. A prospective clinical study of 69 consecutive adults with suspected spinal TB was conducted at a tertiary hospital in an area with the highest incidence and prevalence of TB in the world. GeneXpert was used on tissue samples of the enrolled patients and its diagnostic accuracy compared with a reference standard of tissue in liquid culture. A total of 71 spine samples from 69 patients (two re-biopsies) were included in the study. The GeneXpert test showed a sensitivity of 95.6% and specificity of 96.2% for spinal TB. The results of the GeneXpert test were available within 48 hours compared with a median of 35 days (IQR 15 to 43) for cultures. All cases of multi-drug resistant TB (MDR TB) were diagnosed accurately with the GeneXpert test. The MDR TB rate was 5.8%. Cite this article: Bone Joint J 2014;96-B:1366–9.
M. Held, FCOrtho (SA), Dr Med, Orthopaedic Surgeon M. Laubscher, FCOrtho (SA), MBChB, Orthopaedic Surgeon R. N. Dunn, FCSOrtho (SA), MBChB, Orthopaedic Surgeon, Professor Department of Orthopaedic Surgery, University of Cape Town, H49 Old Main Building, Groote Schuur Hospital, Observatory 7925, South Africa. H. J. Zar, MBBCh, FCPaeds, PhD, Paediatric Pulmonologist, Professor Department of Pediatrics and Child Health, Red Cross War Memorial Children’s Hospital, University of Cape Town, South Africa. Correspondence should be sent to Dr M. Held; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B10. 34048 $2.00 Bone Joint J 2014;96-B:1366–9. Received 27 February 2014; Accepted after revision 8 July 2014
1366
An accurate diagnosis of musculoskeletal tuberculosis (TB) is essential to enable its timely and effective treatment. It is also important to have the results readily available so that treatment can be started. In spinal TB in particular, diagnostic delay can lead to devastating neurological complications and irreversible disability.1 GeneXpert, an automated polymerase chain reaction (PCR) diagnostic test that simultaneously detects TB and rifampicin resistance, has recently become available.2-4 The World Health Organization (WHO) has endorsed GeneXpert for use in patients with suspected pulmonary TB in areas of high HIV or drug resistance.5 it has been quoted as having similar sensitivity and specificity to conventional testing methods, such as TB culture.2-4,6 In a recent meta-analysis of 18 studies with 7816 patients by the Cochrane Infectious Diseases Group, GeneXpert was tested on the sputum of patients with suspected pulmonary TB: a sensitivity of approximately 90% with a false negative rate of 2% was reported, the sensitivity for rifampicin resistance being 94%.7 According to the WHO Global Tuberculosis Report 2013, the GeneXpert MTB/RIF test will soon be universally available.5 There is, however, no data on its accuracy in the diagnosis of musculoskeletal TB.
The aim of this study was to investigate the accuracy of GeneXpert in diagnosing spinal TB.
Materials and Methods This prospective study was conducted over a period of nine months between March and December 2013. It included samples taken from patients with suspected spinal TB who presented to the spinal unit of a tertiary care hospital. This hospital services an area which has the highest prevalence and incidence of TB worldwide.5 Tissue biopsies were taken for liquid culture, acid-fast smear and GeneXpert. The patients were followed up every three months until there were signs of clinical improvement and nine months of treatment for TB had been completed. The diagnostic performance of the GeneXpert assay was compared with that of standard tissue microscopy and liquid culture, which is the definitive test for osteoarticular TB. Index test. Spinal tissue was processed for GeneXpert (Xpert MTB/RIF) (Cepheid, Sunnyvale, California) testing. GeneXpert lysis buffer was added in a 1:3 ratio to 2 mls of specimen. The mixture was mixed by use of rapid oscillation, or vortexed, and allowed to stand for ten minutes, then vortexed again. A total of 2 mls were removed and processed automatically with the THE BONE & JOINT JOURNAL
GENEXPERT POLYMERASE CHAIN REACTION FOR SPINAL TUBERCULOSIS
1367
Table I. Clinical and radiological red flags on which the suspected diagnosis of TB was made. Clinical red flags
Radiological red flags
History of chronic pain for more than three months Constitutional symptoms: low grade fever, night sweats, loss of appetite, weight loss Chronic cough Elevated erythrocyte sedimentation rate History of tuberculosis contact Local kyphosis9 Neurological deficit Positive Mantoux test10 Immune compromise/HIV
Loss of anterior vertebral height Paravertebral shadow on radiograph
GeneXpert instrument and the result was read after approximately 90 minutes. Reference standard. The automated BACTEC MGIT 960 by BD Diagnostic Systems, Sparks MD (Franklin Lakes, New Jersey) was used for mycobacterial cultures. Direct microscopy for TB was performed with a Ziehl–Nielson (ZN) stain for microbiology and haematoxylin and eosin (HE) stain for histopathology. Specifications from the Centres for Disease Control and Prevention were used for quantification of acid-fast bacilli. STARD (Standards for the Reporting of Diagnostic Accuracy studies) guidelines were followed.8 Identification of Multi-Drug Resistant TB, (MDR-TB) was performed using the GenoType MTBDRplus or GenoType Mycobacterium CM lineprobe assays (Hain Lifescience, Nehren, Germany). Definitions. ‘Definite TB’ was defined as a positive culture of mycobacterium tuberculosis. In patients with negative cultures, ‘Probable TB’ was assumed if the histology, clinical and radiological findings were suggestive of TB and if the patients improved on anti-tuberculous treatment. The clinical and radiological features that were used as red flags are listed in Table I.9,10 ‘Not TB’ was assumed in patients who tested negative on both culture and histology and who improved without anti-tuberculous treatment. ‘MDR TB’ is defined according to the WHO as resistance to isoniazid and rifampicin, with or without resistance to other first-line drugs.11 GenoType MTBDRplus or GenoType Mycobacterium lineprobe assays (Hain Lifescience) were used to confirm resistance to isoniazid and rifampicin. Technical specifications. The samples were collected under strict aseptic conditions in the operating theatre and were taken from predetermined areas of granuloma and destructive disease of the spine, which had been established by MRI and radiological imaging studies. Percutaneous biopsies were reserved for lesions below the level of T1 (first thoracic vertebra) in patients who did not need drainage or stabilisation. A percutaneous transpedicular biopsy was performed in the thoracic spine and a percutaneous paraspinal biopsy in the lumbar spine. An open biopsy was performed when debridement of granulomata and drainage of paravertebral abscesses or stabilisation of the spinal column were needed. Indications for open VOL. 96-B, No. 10, OCTOBER 2014
Shadow of a psoas abscess Adjacent vertebral endplate changes with preserved disc height Changes on chest radiograph indicate suspected tuberculosis Paravertebral abscess on MRI
biopsy included instability of the spinal column; acute deterioration of the neurological signs; a large paravertebral abscess and airway compromise in lesions of the cervical spine. Abscesses were drained through a costotransversectomy in the thoracic spine and through an anterolateral retroperitoneal approach in the lumbar spine. A thoracotomy was performed when stabilisation of the anterior column of the thoracic spine was needed. The Smith–Robinson approach12 was used for lesions of the cervical spine. Biopsy material included pus swabs and/or tissue biopsies from diseased bone or soft-tissue. All samples were subjected to GeneXpert Assay and histopathological examination, as well as microbiological testing. Approval of the hospital’s ethics committee was acquired and all patients gave informed consent. Statistical analysis. The sensitivity, specificity and predictive values of GeneXpert with 95% confidence intervals (95% CI) were determined using a positive culture for definite TB (mycobacterium tuberculosis) and a positive histology in cases of negative cultures (probable TB) as the reference standard. The data was analysed as per sample using STATA 11 statistical software (STATA Corporation, College Station, Texas). Simple descriptive statistics were used to characterise the study population. Normally distributed continuous data were summarised by mean and 95% CI and non-normally distributed continuous data by median and interquartile range. Categorical data were summarised as proportions with 95% CI. Statistical tests included two-sample t-test of proportions, chi-squared test and Wilcoxon rank-sum test. All statistical tests were twosided at α = 0·05.
Results A total of 71 spine samples from 69 patients (two rebiopsies) were included in the study. These were taken from 31 men (44.9%) and 38 women whose median age was 40 years (IQR 27 to 60). Table II shows the incidence of HIV, the gender and age of the patients along with the results of using GeneXpert. Table III shows the per sample analysis of all patients with either culture-positive (‘definite TB’) or histology-positive (‘probable TB’) samples. There were two false negative and one false positive GeneXpert samples in this analysis. A secondary analysis was
1368
M. HELD, M. LAUBSCHER, H. J. ZAR, R. N. DUNN
Table II. Baseline characteristics of the patient cohort (TB, tuberculosis)
N HIV + (%) Gender (male (%)) Age (median (IQR)) GeneXpert +(%)
Overall
Definite TB
Probable TB
Not TB
p-value
69 22 (31.9) 31 (44.9) 40 (27 to 60) 44/69 (63.8)
36 13 (36.1) 14 (38.9) 31 (27 to 46) 35/36 (97.2)
8 4 (50.0) 6 (75.0) 37.5 (23 to 57.5) 8/8 (100.0)
26 5 (20.0) 11 (44.0) 51 (40 to 63) 1/25 (4.0)
0.28 0.28 0.18 0.06
Table III. Per sample accuracy of GeneXpert for detecting tuberculosis (TB) (PPV: positive predictive value (NPV, negative predictive value) Reference standard
All (N)
Sensitivity % (n)
Specificity % (n)
PPV % (n)
NPV % (n)
TB culture or histology positive TB culture positive only
71 71
95.6 (44/46) 92.6 (25/27)
96.2 (25/26) 79.5 (35/44)
97.7 (44/45) 73.5 (25/34)
92.6 (25/27) 94.6 (35/37)
undertaken to compare GeneXpert with culture positive samples only (definite TB) as a reference standard. Culture detected 37 TB patients and histology a further eight. The rate of TB detection with GeneXpert compared with cultures alone increased by 21.6% (8 of 37 samples). This shows that the eight samples which were only detected histologically (probable TB) were likely to be false negative TB cultures Table III. The results of the GeneXpert test were available within 48 hours compared with a median of 35 days (IQR, 15 to 43) for culture. The GeneXpert test showed rifampicin-resistant TB in four patients (5.8%) and ‘possible MDR TB’ was therefore suspected. All four patients were started on the correct MDR TB treatment the day after their biopsy. In three of the four, MDR TB was confirmed with sensitivity testing for rifampicin and isoniazid once the culture was available. In one of the four GeneXpert samples which showed rifampicin resistance, the TB culture was negative and therefore MDR TB may have been missed without the GeneXpert test. Conversely, in another GeneXpert test sample, isoniazid mono-resistance was evident only on culture as the GeneXpert tests solely for rifampicin resistance.
Discussion With the advent of PCR methods to test for TB, various studies have reported that the sensitivity of nonautomated PCR tests range from 61% to 83%.13-18 The PCR method has the advantage of needing much smaller quantities of M. tuberculosis to identify infection and gives a rapid diagnosis by amplification of the nucleic acid of the pathogen. Its detection limit of 130 colony-forming units (CFU) per millilitre, compared with 10 000 CFU/ml in cultures, increases the probability of diagnosing extrapulmonary tuberculosis even with samples containing only a few bacteria.19,20 In this study, the sensitivity of the GeneXpert test was 95.6%, the specificity 96.2%, the positive predictive value 97.7% and negative predictive
value 92.6%. A total of four patients with MDR TB were accurately diagnosed with the GeneXpert test. One of these had a false-negative culture result which was only reported 43 days after biopsy. These patients would have been discharged into the community without adequate treatment for 43 days, and in one patient without treatment at all, had it not been for the GeneXpert test. Another major benefit of the GeneXpert test is the speed with which the results become available. The automated process of the GeneXpert device takes only about 90 minutes to complete. Also, compared with other PCR testing methods, the automated, cartridge-based test obviates the need for extensive laboratory support. This makes it a tool which can be moved closer to smaller hospitals and clinics, cutting down the long transport times often present in countries with a high burden of TB. However, Monni et al21 have reported that there is no difference in the time taken to produce a report compared with conventional testing methods, when the GeneXpert device is not on site. One of the disadvantages of PCR testing is that, in contrast to TB culture, PCR testing will give a positive result even if the pathogens are not viable.2 In these patients, active TB has to be confirmed clinically and by means of various imaging modalities. Another concern about the GeneXpert method is that it tests drug resistance only for rifampicin, therefore a mono-resistance for isoniazid can only be detected with other PCR testing methods, such as GenoType MTBDRplus or GenoType Mycobacterium CM lineprobe assays (Hain Lifescience, Nehren, Germany). In our series, we discovered one patient with isoniazid monoresistance. According to current treatment guidelines isoniazid mono-resistance does not alter the conventional TB regime for patients without isoniazid mono-resistance.22 In conclusion, the GeneXpert test is more sensitive and gives faster access to results than the gold standard. Consequently, we recommend its use as the initial test in the diagnosis of patients with TB of the spine. THE BONE & JOINT JOURNAL
GENEXPERT POLYMERASE CHAIN REACTION FOR SPINAL TUBERCULOSIS We appreciate the support of C. Bamford with the laboratory setup for this study. N. Beylis has provided us with special insight regarding GeneXpert and TB diagnostics and methods. We thank L. Workman for her assistance with the statistical analysis. No benefits in any form have been received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by S. P. H. Hughes and first proof edited by A. Ross.
References 1. Jain AK. Treatment of tuberculosis of the spine with neurologic complications. Clin Orthop Relat Res 2002;398:75–84. 2. Rachow A, Zumla A, Heinrich N, et al. Rapid and accurate detection of Mycobacterium tuberculosis in sputum samples by Cepheid Xpert MTB/RIF assay--a clinical validation study. PLoS One 2011;6:20458. 3. Hillemann D, Rüsch-Gerdes S, Boehme C, Richter E. Rapid molecular detection of extrapulmonary tuberculosis by the automated GeneXpert MTB/RIF system. J Clin Microbiol 2011;49:1202–1205. 4. Theron G, Peter J, Dheda K. Xpert MTB/RIF test for tuberculosis. Lancet 2011;6:378–481. 5. No authors listed. WHO tuberculosis report 2014. http://www.who.int/mediacentre/factsheets/fs104/en (date last accessed 15 January 2014). 6. Peters D, Theron G, Peter J, Dheda K. Should Xpert® MTB/RIF be rolled out in low-income countries? Int J Tuberc Lung Dis 2012;16:702–703. 7. Steingart KR, Sohn H, Schiller I, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev 2013;1:CD009593. 8. No authors listed. STARD checklist. http://www.stard-statement.org/ checklist_maintext.htm (date last accessed 8 August 2014). 9. Rajasekaran S. The problem of deformity in spinal tuberculosis. Clin Orthop Relat Res 2002;398:85–92.
VOL. 96-B, No. 10, OCTOBER 2014
1369
10. No authors listed. NICE guidelines for the diagnosis of TB: http://www.nice.org.uk/ guidance/cg117/chapter/guidance (dated last accessed 8 August 2014). 11. No authors listed. WHO website on MDR TB: http://www.who.int/tb/challenges/ mdr/tdrfaqs/en/ (date last accessed 8 July 2014). 12. Smith GW, Robinson RA. The treatment of certain cervical-spine disorders by antero removal of the intervertebral disc and interbody fusion. J Bone Joint Surg [Am] 1958;40-A:607–624. 13. Davies A, Newport L, Billigton O, Gillespie S. Length of time to laboratory diagnosis of Mycobacterium tuberculosis infection: comparison of in-house methods with reference laboratory results. J Infect 1999;39:205–208. 14. Negi SS, Gupta S, Khare S, Lal S. Comparison of various microbiological tests including polymerase chain reaction for the diagnosis of osteoarticular tuberculosis. Indian J Med Microbiol 2005;23:245–248. 15. Sun Y, Zhang Y, Lu Z. Clinical study of polymerase chain reaction technique in the diagnosis of bone tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 1997;20:145148.[Article in Chinese]:. 16. Tiwari V, Jain A, Verma RK. Application of enzyme amplified mycobacterial DNA detection in the diagnosis of pulmonary and extrapulmonary tuberculosis. Indian J Med Res 2003;118:224–228. 17. Pandey V, Chawla K, Acharya K, Rao S, Rao S. The role of polymerase chain reaction in the management of osteoarticular tuberculosis. Int Orthop. 2009;33:801–805. 18. Rasit AH, Razak M, Ting FS. The pattern of spinal tuberculosis in Sarawak General Hospital. Med J Malaysia 2001;56:143–150. 19. Singh KK, Muralidhar M, Kumar A, et al. Comparison of in house polymerase chain reaction with conventional techniques for the detection of Mycobacterium tuberculosis DNA in granulomatous lymphadenopathy. J Clin Pathol 200;53:355–361. 20. Kramer N, Rosenstein ED. Rheumatologic manifestations of tuberculosis. Bull Rheum Dis 1997;46:5–8. 21. Monni T, Visser HF, Motsitsi SN. Clinical utility of tissue polymerase chain reaction in the diagnosis of spinal tuberculosis. SA Orthop J 2012;11:23–27. 22. Cattamanchi A, Dantes RB, Metcalfe JZ, et al. Clinical characteristics and treatment outcomes of patients with isoniazid-monoresistant tuberculosis. Clin Infect Dis 2009;48:179–185.
