• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
Int J Tuberc Lung Dis. Author manuscript; available in PMC Jun 12, 2012.
Published in final edited form as:
PMCID: PMC3373957

Diagnostic accuracy of commercial urinary lipo-arabinomannan detection in African TB suspects and patients



To evaluate a commercially available antigen capture ELISA based on detecting lipoarabinomannan (LAM ELISA) in urine, for diagnosis of tuberculosis.


Consenting TB suspects and registering TB patients prospectively recruited from 3 hospitals were asked for 2 sputum specimens for microscopy and culture, urine for LAM testing, and blood for HIV testing, with radiological and clinical follow up for 2 months.


Of 427 participants, complete data were available from 397 (307 adult and 23 adolescent TB suspects, and 67 registering TB patients). HIV prevalence was 77%. TB was diagnosed in 195 (49%), including 161 culture-positive patients, and confidently excluded in 114 (29%) participants.

LAM ELISA sensitivity was 44% (95% CI 36-52%) for culture-confirmed TB (52% in smear-positive patients). Specificity was 89% (95% CI: 81-94%). Sensitivity was significantly higher in HIV-related TB (52%: 95%CI 43-62%, p <0.001) compared to HIV-negative TB (21%: 95%CI 9-37%), Sensitivity in smear-negative patients was low (28%: 95%CI 13-43% for combined HIV-positive and –negative patients).


Our findings confirm greater sensitivity of urine LAM detection for HIV-related TB. However, both sensitivity and specificity were suboptimal, suggesting that this version cannot confirm or exclude TB in either HIV-infected or uninfected patients.

Keywords: Tuberculosis, diagnostic, culture, screening, Africa, HIV


Effective management of suspected tuberculosis (TB) and the performance of TB control programs in resource poor settings are currently compromised by the lack of a sensitive, specific and timely diagnostic test [1]. Consequential delays in diagnosis contribute to both high TB mortality and ongoing TB transmission [2-4].

Among the most promising of the new TB diagnostics in development for point-of-care use are antigen-detection assays that are based on liporabinomannan, LAM, a carbohydrate cell wall antigen that is excreted in the urine of TB patients [5, 6]. Unlike assays that rely on antibody or T-cell responses, antigen-detection assays are less vulnerable to host immune system variables and show the presence of replicating organisms. As such, they have high potential to distinguish active disease from latent TB infection, and will tend to have increased, rather than decreased, sensitivity in the immunosuppressed. Previous studies have shown LAM detection to have moderate sensitivity and good specificity, with the highest sensitivity being in HIV-related TB and in patients with smear-positive disease [7-9]. A number of different companies and research groups have LAM-based tests under development. One of these was recently launched as commercial test, Clearview™.

We evaluated the diagnostic performance of the pre-commercial product, a urine-based LAM ELISA, in Harare, Zimbabwe, where adult HIV prevalence is approximate 20% (ref to 2007 data: DHS, DETECTB prevalence, MoH). We used TB microscopy and culture, and also clinical and radiological case-definitions for TB, to evaluate test performance in several distinct patient groups, including probable culture-negative TB. The main aims of the study were to estimate sensitivity and specificity according to HIV status and sputum bacteriology.


Study design

A prospective cohort of TB suspects and registering TB patients, followed-up to confirm or exclude TB disease.

Study setting and sample size

The study was conducted at 3 hospitals in Harare, Zimbabwe, between September 2007 and July 2008 and a total of 427 participants were recruited. Sample size calculations were based around estimating an assumed test 60% sensitivity in HIV-infected TB patients with precision of 5%, and an estimated HIV prevalence of 85% in TB patients, true prevalence of TB of 50% (deliberately oversampled through inclusion of known smear-positive TB patients), and 10% loss to follow-up (ref Carley)

Study sites and participants

Participants were recruited from

  • -
    A tertiary referral hospital, Beatrice Road Infectious Disease Hospital in Harare, Zimbabwe: ambulant adult (≥ 18 years) smear-positive TB patients attending for registration (required to obtain a TB treatment card and continue TB treatment), and ambulant primary care TB suspects with a cough of more than 3 weeks duration attending to submit sputum specimens to the microscopy laboratory at BRIDH
  • -
    Two central hospitals in Harare, Parirenyatwa and Harare Central Hospital: adolescents (10 to 18 years) admitted with a febrile or respiratory illness.

Sequential consenting adult TB suspects and registering TB patients up to a maximum of 15 per day were recruited from the infectious diseases hospital. Adolescents were recruited with consent of their guardians, as previously described (ref to Rashida). TB suspects were excluded if already on TB treatment. Registering TB patients were only considered for recruitment if they had taken treatment for less than 24 hours.

Investigation, follow-up and case-definitions

All participants had chest radiography, and were asked to provide 4 sputum specimens (2 spot and 2 morning), urine, and blood for anonymous HIV testing. Adolescent in-patients were recruited with consent of their guardians, and had TB blood cultures taken in addition to other investigations for the purposes of another study (ref to Rashida).

Smear-negative TB suspects were treated with antibiotics (amoxicillin and doxycycline) and had repeat smear and culture (1 spot and 1 morning specimen) before starting TB treatment if they had persistent symptoms and a compatible clinical and/or radiological illness.

Repeat TB investigations (smear, culture, radiography) were carried out in all participants who were not already on TB treatment if symptoms persisted by 2 months, or if there were initial radiological changes.

Patients started on TB treatment were followed-up for 2 months to document radiological and clinical response to treatment, unless TB was bacteriologically confirmed in the interim (2 positive smears or 2 cultures growing M. tuberculosis). Response to TB treatment was assessed on the basis of i) weight gain; ii) radiological response (complete or partial), iii) resolution of previously documented temperature iv) reported resolution of previously documented symptoms (cough, haemoptysis, fever, night-sweats and chest pain). On the basis of the above the attending physician categorized the extent to which overall treatment response supported the diagnosis of TB as “highly consistent”, “TB likely”, or “TB unlikely”.

TB case-definitions were as follows:

  • -
    Smear-positive TB: at least one positive smear from the initial (4 specimens) and follow-up (2 specimens) microscopy, plus culture positive (LJ slope positive or blood culture positive) for M. tuberculosis
  • -
    Smear-negative culture-positive TB: ≥1 positive culture of M. tuberculosis with all smears negative
  • -
    Smear-negative culture-negative TB: all samples smear and culture negative, but clinical and radiological evidence of TB, non-response to broad-spectrum antibiotics, response to 2 months of TB treatment.
  • -
    Non-TB: All samples smear and culture negative, and stable recovery at 2 months without TB treatment
  • -
    Indeterminate: not meeting any of the above definitions

Laboratory Methods

Collected urine specimens were kept unprocessed at 2-8°C and run within 3 days of collection. Urine specimens with volumes less than 10mls were not tested. Urine processing and testing was carried out on the same day. Urine was heated for 30 minutes at 100°C and centrifuged at 10000 rpm for 15 minutes, and 100 microlitres of supernatant was used for urine LAM testing following manufacturer’s instructions. Automatic washing was used at all stages. The pre-commercial Chemogen (Chemogen, USA) version of the LAM test was used. All specimens were run in duplicate and ODs read at 450nm after stopping with 1 molar sulfuric acid. The average OD of the negative control plus 0.1 was considered as the cut off, with specimens with ODs less than and more than the cut-off taken as negative and positive respectively. Specimens with one OD value less than the ct-off and one more than the cut-off were recorded as indeterminate. Interpretation of LAM results were done independent of any other clinical or laboratory results.

Sputum smears were made from both direct and concentrated decontaminated (4% NaOH) sputum and read by fluorescence microscopy (Auramine O). All positive and 10% of negative slides were re-read by a second reader. Auramine positive slides were confirmed with Ziehl-Neelsen (ZN) staining (classified as smear negative if ZN negative). Culture used Lowenstein-Jensen (LJ) slopes, with the residual concentrate stored at -20°C for re-culture in case of contamination. Species identification used MPB64 antigen detection (Capillia™) and visual inspection for cording. Further speciation tests (colony morphology, growth at different temperatures, and growth on PNB containing LJ slopes) were used for isolates that were ZN+ve but MPB64 antigen negative. Blood culture used Mycolytic F™ (Becton Dickinson, Johannesberg, South Africa: adolescent participants only).

Confidential HIV serology used Determine (Abbott Diagnostics, Johannesburg, South Africa), with all positives and 10% of negative specimens confirmed by Unigold (Trinity Biotech, Dublin, Ireland). For participants not willing to provide serum, oral mucosal transudate was collected and tested using Vironostika (Vironostika; BioMérieux, Marcy l’Etoile, France).

Data analysis

Data were entered into Epi-info version 3.4.1 (Centers for Disease Control and Prevention, Atlanta, USA) and analysed using STATA 9.0 (STATA Corporation, College Station, Texas, USA). Between-group differences in patient characteristics and LAM ELISA positivity were tested for statistical significance using the Chi-squared test.

Sensitivity, specificity, positive predictive value and negative predictive value of the LAM ELISA were calculated for a) smear-positive TB, b) culture-positive TB (regardless of smear results), and c) all diagnosed TB patients meeting case-definitions for TB disease.

Patients with indeterminate TB status and, where relevant, patients with TB who did not meet the bacteriological criteria for which sensitivity and specificity were being calculated, were excluded from the analysis.

Ethical approval

The study was approved by the Zimbabwe Medical Research Council, Biomedical Research and Training Institute’s Institutional Review Board and the Ethics Committee of the London School of Hygiene and Tropical Medicine. Diagnostic provider-initiated counseling and testing was offered to all participants, according to national guidelines.


Baseline characteristics

Participation was 97% (refusals: 11 registering TB patients and 1 adolescent). Of 427 participants, complete data were available for 397 (307 adult TB suspects, 23 adolescents and 67 registering TB patients) as 30 participants with no urine samples or urine sample less than 10 mls were excluded (Figure 1). Participant characteristics are shown in Table 1.

Participant flow chart
Baseline participant characteristics for TB and non-TB patients

The overall HIV prevalence among the participants was 77%.

TB status

TB was excluded in 114 (29%) and diagnosed in 195 (49%) participants. Disease was culture confirmation in 161, including 121 (62%) smear- and culture-positive and 40 (21%) smear-negative and culture-positive. Thirty-four (17%) patients had culture-negative TB diagnosis through clinical/radiological criteria (Figure 1). There were no smear-positive culture-negative patients. TB was neither confirmed nor excluded in 88 (22%) patients as shown in figure 1.

Test sensitivity for culture-positive TB

There were major variations in the sensitivity of LAM according to HIV and bacteriological status, as shown in Table 2 and Figure 2. Overall test sensitivity (combined HIV groups) was 44% (95% CI: 36-52%).

Chemogen LAM sensitivity in TB patients by smear and culture status
LAM test sensitivity (culture confirmed TB only), and negative and positive predictive values by HIV status

Sensitivity was 50% (95% CI: 40-59%) in smear-positive culture-positive TB, 28% (95% CI: 13-43%) in smear-negative and culture-positive TB, and 27% (95% CI: 11-43%) in clinically diagnosed smear-negative and culture-negative TB patients.

As previously reported for LAM tests, sensitivity for culture confirmed TB was significantly higher in HIV positive patients (52%: 95% CI 43-62%, p<0.001) compared to HIV negative TB patients (21%: 95% CI 9-37%), with the highest sensitivity being in HIV-positive smear-positive patients (64% [95% CI: 53-76%]).

Test specificity

Using patients in whom TB had been confidently excluded, overall LAM specificity was 89% (95% CI: 81-94%): this was similar in HIV positive (86%, 95%CI: 77-93%) and HIV negative (93%, 95%CI: 77-99%) TB suspects, as shown in Table 2.

Negative and positive predictive values

The overall negative and positive predictive values were 84% (95%CI: 74-91%) and 54% (95% CI: 46-61%) respectively. Negative and positive predictive values by HIV status were similar to the overall predictive values as shown in Table 2.

Test positivity across culture negative clinical groups

LAM test positivity varied significantly between clinically distinct groups of culture-negative patients, as shown in Table 3. Culture- negative patients lost to follow-up because of death (n=16) or relocation to rural areas (n=33) had similar LAM positivity to patients with culture-confirmed TB, significantly higher (31% p=0.035 and 30% p=0.014) than LAM positivity prevalence in patients in whom TB had been excluded (Table 3).

The prevalence of Chemogen LAM positivity rates in clinically distinct culture negative patient groups


The main findings from this study are that the LAM ELISA evaluated here had low sensitivity (44%) for culture-positive TB and suboptimal specificity (89%), resulting in low positive and negative predictive values of 84% and 54% in this cohort. There is urgent need for better TB diagnostics that perform well in detecting HIV-related disease and in this study, as previously reported, LAM sensitivity was higher in HIV-related than HIV-negative TB [7]. However, the low specificity, and consequently low positive predictive values, do not support the use of this test as a first line diagnostic, even if formatted into a rapid lateral flow test.

Unfortunately, the LAM ELISA also performed poorly in detecting smear-negative TB, a patient group where more effective diagnosis is urgently needed [10, 11], with sensitivity of 28% in smear-negative culture-positive patients and 27% in smear- and culture-negative patients who met the case-definitions for TB disease through response to TB treatment. Smear microscopy, the first-line investigation of TB suspects in resource poor settings, has low sensitivity, ranging from 43% to 74% compared to culture in research studies and perhaps as low as 20% in routine programmes [12-15]. Mycobacterium culture is more sensitive, although still not 100%, but suffers from a long turnaround time and requirements for expertise and infrastructure that preclude its use in most resource poor settings. Smear sensitivity is further reduced by concurrent HIV infection; making smears a very suboptimal first-line test in high HIV prevalence settings [11, 14, 16-21]. Thus the main gaps in the currently available diagnostics are rapid point-of-care tests with sensitivity and specificity at least as good as smear, and also diagnostics with good performance in detecting HIV-related smear-negative TB.

Sensitivity in this study is lower than previously reported using other LAM ELISA assays [7-9, 22]. This may reflect differences in test formats, and/or in study design. Studies comparing test results from confirmed TB patients against those of healthy controls typically report higher test accuracies than found in prospective test evaluations in TB suspects [23, 24]. In this study, sensitivity was 64% for HIV-positive smear-positive TB, whereas in previous reports sensitivity in this group has been as high as 79% [7]. Our use of highly sensitive microscopy (fluorescent and mechanical concentration)[steingradt] may, paradoxically, have contributed to the apparently low performance of LAM ELISA in detecting smear-positive TB in this study.

We also show significant differences in LAM positivity between patients in whom TB was excluded on the basis of negative investigations and complete clinical recovery (12% LAM positive) and other clinically distinct culture-negative patient groups (clinical TB, died during assessment, lost to follow-up: 27% to 31% LAM positive), suggesting that urine LAM may have potential as an epidemiologic tool for investigating the overall performance of TB screening clinical algorithms even if not sufficiently specific to allow individual diagnosis. However, this too would require higher sensitivity than demonstrated here. The relatively high LAM positivity in patients lost to follow up through death, which is similar to that of clinically diagnosed TB, concurs with reports of high rates of death from undiagnosed TB in post mortem studies in Africa [25-28].


Our findings confirm the considerably greater sensitivity of LAM urine-based antigen detection tests for HIV-positive TB compared to HIV-negative TB. However both sensitivity and specificity were suboptimal in this evaluation. The low specificity undermines the potential of this test to be used in a rapid test kit format as a first line point-of-care test for HIV-related TB, while the very low sensitivity for smear-negative TB patients suggesting little added-value as a diagnostic in this important subgroup. Overall, this test cannot be used to either confidently confirm or exclude TB in either HIV positive or HIV-negative patients.


[1] Perkins MD, Cunningham J. Facing the crisis: improving the diagnosis of tuberculosis in the HIV era. The Journal of infectious diseases. 2007 Aug 15;196(Suppl 1):S15–27. [PubMed]
[2] Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, et al. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med. 2003;163(9):1009–21. [PubMed]
[3] Uys PW, Warren RM, van Helden PD. A threshold value for the time delay to TB diagnosis. PLoSONE. 2007;2(1):e757. [PMC free article] [PubMed]
[4] Young DB, Perkins MD, Duncan K, Barry Ce r. Confronting the scientific obstacles to global control of tuberculosis. J Clin Invest. 2008;118(4):1255–65. [PMC free article] [PubMed]
[5] Hamasur B, Bruchfeld J, Haile M, Pawlowski A, Bjorvatn B, Kallenius G, et al. Rapid diagnosis of tuberculosis by detection of mycobacterial lipoarabinomannan in urine. J Microbiol Methods. 2001;45(1):41–52. [PubMed]
[6] Tessema TA, Bjune G, Assefa G, Svenson S, Hamasur B, Bjorvatn B. Clinical and radiological features in relation to urinary excretion of lipoarabinomannan in Ethiopian tuberculosis patients. Scand J Infect Dis. 2002;34(3):167–71. [PubMed]
[7] Boehme C, Molokova E, Minja F, Geis S, Loscher T, Maboko L, et al. Detection of mycobacterial lipoarabinomannan with an antigen-capture ELISA in unprocessed urine of Tanzanian patients with suspected tuberculosis. Trans R Soc Med Hyg. Dec;99(12):893–900. [PubMed]
[8] Tessema TA, Bjune G, Assefa G, Bjorvat B. An evaluation of the diagnostic value of clinical and radiological manifestations in patients attending the addis ababa tuberculosis centre. Scand J Infect Dis. 2001;33(5):355–61. [PubMed]
[9] Tessema TA, Hamasur B, Bjun G, Svenson S, Bjorvatn B. Diagnostic evaluation of urinary lipoarabinomannan at an Ethiopian tuberculosis centre. Scand J Infect Dis. 2001;33(4):279–84. [PubMed]
[10] Siddiqi K, Walley J, Khan MA, Shah K, Safdar N. Clinical guidelines to diagnose smear-negative pulmonary tuberculosis in Pakistan, a country with low-HIV prevalence. Trop Med Int Health. 2006;11(3):323–31. [PubMed]
[11] Mendelson M. Diagnosing tuberculosis in HIV-infected patients: challenges and future prospects. Br Med Bull. 2007;81-82:149–65. [PubMed]
[12] Apers L, Wijarajah C, Mutsvangwa J, Chigara N, Mason P, van der Stuyft P. Accuracy of routine diagnosis of pulmonary tuberculosis in an area of high HIV prevalence. Int J Tuberc Lung dis. 2004 Aug;8(8):945–51. [PubMed]
[13] Steingart KR, Henry MNV, Hopewell PC, Ramsey A, Cunningham J, Urbanczik RP, Aziz M, Pai M. Sputum processing methods to improve the sensitivity of smear microscopy for tuberculosis: a systematic review. Lancet Infect Dis. 2006;6:664–74. M. [PubMed]
[14] Steingart KR, Ramsay A, Pai M. Optimizing sputum smear microscopy for the diagnosis of pulmonary tuberculosis. Expert Rev Anti Infect Ther. 2007;5(3):327–31. [PubMed]
[15] Urbanczik R. Present position of microscopy and of culture in diagnostic mycobacteriology. Zentralblatt fur Bakteriologie, Mikrobiologie, und Hygiene. 1985 Aug;260(1):81–7. [PubMed]
[16] Getahun H, Harrington M, O’Brien R, Nunn P. Diagnosis of smear-negative pulmonary tuberculosis in people with HIV infection or AIDS in resource-constrained settings: informing urgent policy changes. Lancet. 2007;369(9578):2042–9. [PubMed]
[17] Colebunders R, Bastian I. A review of the diagnosis and treatment of smear-negative pulmonary tuberculosis. Int J Tuberc Lung dis. 2000 Feb;4(2):97–107. [PubMed]
[18] Corbett EL, Marston B, Churchyard GJ, De Cock KM. Tuberculosis in sub-Saharan Africa: opportunities, challenges, and change in the era of antiretroviral treatment. Lancet. 2006;367(9514):926–37. [PubMed]
[19] Palmieri F, Girardi E, Pellicelli AM, Rianda A, Bordi E, Rizzi EB, et al. Pulmonary tuberculosis in HIV-infected patients presenting with normal chest radiograph and negative sputum smear. Infection: 8/13/2008. 2002;30(2):68–74. [PubMed]
[20] Harries AD, Maher D, Nunn P. An approach to the problems of diagnosing and treating adult smear-negative pulmonary tuberculosis in high-HIV-prevalence settings in sub-Saharan Africa. Bull World Health Organ. 1998;76(6):651–62. [PMC free article] [PubMed]
[21] Harries AD, Nyirenda TE, Banerjee A, Boeree MJ, Salaniponi FM. The diagnosis of smear-negative pulmonary tuberculosis: the practice of sputum smear examination in Malawi. Int J Tuberc Lung dis. 1999 Oct;3(10):896–900. [PubMed]
[22] Chan ED, Reves R, Belisle JT, Brennan PJ, Hahn WE. Diagnosis of tuberculosis by a visually detectable immunoassay for lipoarabinomannan. Am J Respir Crit Care Med. 2000;161(5):1713–9. [PubMed]
[23] Pai M, O’Brien R. Tuberculosis diagnostics trials: do they lack methodological rigor? Expert Rev Mol Diagn. 2006;6(4):509–14. [PubMed]
[24] Small PM, Perkins MD. More rigour needed in trials of new diagnostic agents for tuberculosis. Lancet. 2000 Sep 23;356(9235):1048–9. [PubMed]
[25] Ansari NA, Kombe AH, Kenyon TA, Hone NM, Tappero JW, Nyirenda ST, et al. Pathology and causes of death in a group of 128 predominantly HIV-positive patients in Botswana, 1997-1998. Int J Tuberc Lung dis. 2002 Jan;6(1):55–63. [PubMed]
[26] Lucas SB, Hounnou A, Peacock C, Beaumel A, Djomand G, N’Gbichi JM, et al. The mortality and pathology of HIV infection in a west African city. AIDS (London, England) 1993;7(12):1569–79. [PubMed]
[27] Pronyk PM, Kahn K, Hargreaves JR, Tollman SM, Collinson M, Hausler HP, et al. Undiagnosed pulmonary tuberculosis deaths in rural South Africa. Int J Tuberc Lung dis. 2004 Jun;8(6):796–9. [PubMed]
[28] Rana FS, Hawken MP, Mwachari C, Bhatt SM, Abdullah F, Ng’ang’a LW, et al. Autopsy study of HIV-1-positive and HIV-1-negative adult medical patients in Nairobi, Kenya. J Acquir Immune Defic Syndr. 2000;24(1):23–9. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...