The utility of urine lipoarabinomannan as a tuberculosis treatment monitoring tool in people living with HIV

Abstract

Objective

This pilot study aimed to evaluate whether urine lipoarabinomannan (LAM) testing can be used to monitor tuberculosis treatment response in people living with HIV.

Results

Seventy-five participants were enrolled, of whom 51 (68%) were women, and 43 (57.4%) were antiretroviral therapy naïve. CD4 cell counts were available for 49 participants with a median of 58 cells/mm³ (Interquartile range: 22–175). All participants initially tested positive for urine LAM; however, 12 (16%) converted to negative within two days of starting tuberculosis treatment. During the eight-week follow-up, 49 participants (65.3%) reverted from negative to positive at least once, primarily within the first two weeks, with notable peaks on day 4 (14, 28.6%) and day 5 (10, 20.4%). Of the 53 participants with urine LAM results at eight weeks, 26 participants (49.1%) had sustained urine LAM negativity. There was a significant LAM negativity trend from baseline to week 8 (Cochran-Armitage test for trend, p = 0.002). Urine LAM negativity did not significantly correlate with sputum smear or culture negativity (p = 0.521). These findings indicate that urine lipoarabinomannan testing is unreliable for monitoring tuberculosis treatment in people living with HIV, although repeated measurements offer insight into its performance in real-world settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13104-026-07703-y.

Introduction

The goal of tuberculosis (TB) treatment is to achieve cure and prevent further disease transmission [1]. Monitoring of the response to anti-TB treatment is essential for early identification of patients who are at risk of treatment failure and enabling timely interventions such as dose adjustments or intensified clinical follow-up¹. However, many patients, especially those who are severely ill or living with HIV, are unable to produce adequate sputum samples required for routine monitoring, or often have paucibacillary specimens [2] which cannot be reliably used to monitor treatment response.

Urine TB lipoarabinomannan (LAM) is an approved point-of-care diagnostic test for TB in people living with HIV (PLHIV) [3], with the highest sensitivity in cases of advanced immunosuppression and disseminated disease [4]. Because it does not require sputum collection, urine LAM has potential utility for monitoring treatment in patients who cannot expectorate. Despite this, a rapid and accessible tool for tracking TB treatment response in this population remains unavailable.

Previous work by Drain et al. demonstrated that LAM positivity declines after the first 8 weeks of therapy and that persistence of LAM positivity after the intensive phase of TB treatment predicts mortality [5]. However, the timing of urine LAM conversion during TB treatment, optimal sampling intervals for monitoring, and its feasibility as a test of cure remain poorly understood. To address these gaps, we conducted a prospective study to describe the temporal changes in urine LAM during treatment and assess its potential role in monitoring therapeutic response.

Methods

Study design and participants

In this pilot study, conducted from December 2019 to November 2020, we screened and consecutively enrolled 75 adults aged 18 years and above, living with HIV, and diagnosed with TB using urine LAM. Participants were included if they were referred with a positive TB urine LAM result, provided written informed consent, were willing and able to comply with study visits, and were either yet to start or had initiated TB treatment within two days before enrolment. Participants who were unable to provide a urine sample were excluded. Participants were recruited from the Infectious Diseases Institute (IDI), a tertiary HIV urban outpatient clinic, and Kiruddu National Referral Hospital (KNRH), a national referral center serving patients from across the country, including both people living with HIV and those without. Both facilities are located in Kampala, Uganda.

Procedures

At enrolment, participants’ social demographic characteristics were collected and clinical assessments performed with the aid of a structured questionnaire (Supplementary File 1). Samples for urine LAM testing were taken at baseline (at any time point), each morning for the first 7 days, and subsequently at 2, 4, and 8 weeks following TB treatment initiation. The urine samples were transported within 30 min to the IDI Translational Laboratory and stored at -80 °C for a maximum of 16 months for batch analysis. All samples were thawed once, and urine LAM testing was performed by a single trained reader who was blinded to participants’ clinical information. Urine LAM testing was performed using the Determine™ TB LAM Ag test strip (Abbott Diagnostics Scarborough Inc.) as per the manufacturer’s recommendations. The strip was inspected by eye and graded by comparing it with the intensities of the bands on the manufacturer-supplied four-band reference card (grade 1 representing a very low-intensity band to grade 4 representing a high/dark intensity band). A urine LAM test result of grade 1 or higher was considered positive.

Spot, spontaneously expectorated sputum samples (≥ 3 ml) were also collected at baseline, one sample between days 3 and 7 [week 1], and at 2, 4, and 8 weeks following enrolment for smear microscopy and cultures using Mycobacteria Growth Indicator Tube (MGIT) and Lowenstein Jensen (LJ) media. All the urine LAM tests, sputum smear microscopy, and cultures were performed at the Makerere University Mycobacteriology Laboratory, a College of American Pathologists (CAP) ISO15189-certified laboratory. Sputum samples were processed using the N-acetyl-L-cysteine–sodium hydroxide (NALC–NaOH) method with an initial NaOH concentration of 6%, and contaminated cultures were subsequently re-decontaminated to obtain evaluable results. Culture positivity was defined as mycobacterial growth on either or both culture methods (MGIT and/or LJ).

All participants were started on anti-TB treatment, and adherence was monitored through either community or facility-based directly observed therapy in accordance with the national guidelines [6]. Patients who missed doses were counseled and encouraged to take them as per guidelines.

Antiretroviral therapy (ART) was given to participants according to national guidelines [7], and adherence wasn’t monitored during the study period. CD4 cell counts were extracted from participants’ medical records. For newly diagnosed participants, baseline CD4 cell count was used, while for ART-experienced participants, the most recent CD4 cell count within the preceding 12 months was considered. The ART-naïve participants were defined as those who had never received ART, while ART-experienced participants were those who had initiated ART before enrolment, regardless of duration or interruptions.

Outcomes

The primary outcome was conversion from urine LAM positivity to sustained negativity during the first 8 weeks of TB treatment. Sustained urine LAM negativity was defined as conversion from a positive to a negative result after treatment initiation, with negativity maintained throughout follow-up and with at least two consecutive negative results at separate visits. The secondary outcome was the urine LAM conversion in relation to microbiological response, measured by sputum smear and culture during TB treatment.

Statistical analysis

Analysis was performed using STATA Version 17.0 Standard Edition (Stata Corp, College Station, Texas, USA). Participant baseline characteristics were summarized using median and interquartile range (IQR) for continuous variables, and as frequencies and proportions for categorical variables. Stacked bar plots were used to compare graded urine LAM results with sputum smear and culture results, and a spaghetti plot to show within-participant urine LAM variation during follow-up. Sustained LAM negativity was analyzed as a proportion and frequency. A Cochran-Armitage test was done to assess a statistically significant change in the trend of urine LAM negativity across the 2 months of TB treatment. To compare the rate of negativity of urine LAM to sputum smear and culture, we used pairwise correlation and a logit generalized estimating equations (logit-GEE) model with an exchangeable correlation structure to account for correlation due to repeated measures. No covariates were adjusted due to the small sample size. Statistical significance was set at a 5% level. Analysis was performed using complete cases.

Results

All 75 participants screened met the eligibility criteria and were enrolled in the study. By the end of the 8-week follow-up period, 8 (10.7%) participants had died, 2(2.7%) had withdrawn consent, and 9 (12%) were lost to follow-up (LTFU) largely because of COVID-19-related travel restrictions. Of the 75 participants enrolled, the majority were women (n = 51,68%) and were ART naïve at the time of enrollment (n = 43, 57.4%). Only 49 participants had available CD4 cell count results with a median of 58 cells/mm³ (IQR: 22–175), and 38 (77.6%) of them had CD4 cell counts < 200 cells/mm³. In addition to a positive urine LAM at screening, 25 (33.3%) had a positive sputum GeneXpert. Although all participants had a positive LAM at TB treatment initiation, 12 (16%) participants tested negative when a urine LAM test was repeated at enrolment (within 2 days of starting TB treatment). In the first 7 days of follow-up, the proportion of participants with a negative LAM fluctuated (Fig. 1).

Fig. 1.

Urine LAM during the first 2 months of TB treatment for all participants. Numbers (n) indicate the number of samples at each time point that were used to calculate the proportions (denominator)

Forty-nine (65.3%) participants reverted from a negative to positive urine LAM at least once during the 8 weeks of TB treatment. This mainly occurred in the initial 2 weeks of follow-up (n = 46, 93.9%) but more markedly between day 4 (14 participants, 28.6%) and day 5 (10 participants, 20.4%) (Fig. 1). After 8 weeks of treatment, 53 participants had a urine LAM test, of whom 27 (50.9%) had a negative test, and 26 (49.1%) had sustained urine LAM negativity. The trend test showed a significant LAM negativity trend from baseline to week 8 (Cochran-Armitage test for trend, P = 0.002).

At each follow-up visit, among participants who had both a urine LAM test and either a sputum smear or culture performed, sputum tests yielded higher negative results compared to urine LAM tests; however, there was no significant correlation (p = 0.521, Fig. 2).

Fig. 2.

Comparison of urine LAM results with sputum-based tuberculosis monitoring tests during treatment. A TB LAM results and sputum microscopy from week 1 to week 8 for participants with both urine LAM and sputum microscopy results available. B TB LAM results and LJ colony counts from week 1 to week 8 for participants with both urine LAM and sputum culture results available. Numbers (n) indicatethe number of samples at each time point that were used to calculate theproportions (denominator)

Discussion

Although we have demonstrated that urine LAM cannot be used as a treatment monitoring tool, there was a general trend towards negativity in the urine LAM test results after 2 months of follow-up. Many participants reverted from a negative urine LAM test to a positive one during treatment. Similar findings have been reported [5], but our study uniquely demonstrates this occurrence in a greater number of participants. The reason for the observed reversion in TB urine LAM results is uncertain, and multiple factors could be at play, such as the degeneration of LAM in collected urine over time and variable time of sample collection (morning vs. spot urine) [8, 9]. A key strength of this study was the repeated measurement of urine LAM levels among participants in a real-world clinical setting, allowing for the observation of changes over time and providing insights into how the test performs under routine programmatic conditions. In conclusion, our findings suggest that the urine TB LAM test is unreliable for monitoring TB treatment. Other innovative tools need to be explored for this purpose.

Limitations

The study relied on a single reader to interpret the urine LAM results, which may have introduced observer bias and increased the potential for misclassification of test outcomes. Another limitation was the absence of data on adherence to anti-tuberculosis therapy, which could have influenced treatment response and consequently affected the observed changes in urine LAM results over time. In addition, the study compared a diagnostic tool primarily designed for detecting disseminated TB (urine LAM) against a gold standard optimized for pulmonary TB, which may not have provided a fully appropriate reference for evaluating test performance. Samples were collected at varying times of day, which may have introduced variability in test results. Finally, the relatively small sample size limited the statistical power of the study and reduced the generalizability of the findings to broader populations.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HIV

Human immunodeficiency virus

LAM

Lipoarabinomannan

PLHIV

People living with HIV

TB

Tuberculosis

Author contributions

Conceptualization, C.S.W; methodology; C.S.W; formal analysis, S.B, and J.M; investigation, R.N.A., W.S., and O.B; writing—original draft preparation, O.B., C.S.W., B. C., and S.J.R.; writing—review and editing, O.B., B.C., C.S.W., S.J.R., Y.C.M., W.S., D.J.S., L.N., R.N.A.; visualization, J.M., and S.B.; funding acquisition, C.S.W. All authors have read and agreed to the published version of the manuscript.

Funding

The work was supported by Grant Number D43TW010132 supported by Office Of The Director, National Institutes Of Health (OD/NIH), National Institute Of Dental & Craniofacial Research (NIDCR), National Institute Of Neurological Disorders And Stroke (NINDS), National Heart, Lung, And Blood Institute (NHLBI), Fogarty International Center (FIC), National Institute On Minority Health And Health Disparities (NIMHD) and in part (SJR) bys the Division of Intramural Research, National Institute of Allergy and Infectious Diseases. This publication was produced as part of the SAfety and EFficacy in high-dose RIFampicin trial, which is part of the European and Developing Countries Clinical Trials Partnership-2 (EDCTP2) Programme supported by the European Union (grant TMA2016CDF-1580). Additional support was provided by a Global Challenges Research Fund award from the Scottish Funding Council, administered via the University of St. Andrews.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was approved by the Mulago Hospital Research and Ethics Committee (reference number: MHREC 1603), and all participants signed written informed consent. The study was also conducted in accordance with general principles outlined in the International Ethical Guidelines for Biomedical Research Involving Human Participants and the Declaration of Helsinki.

Consent for publication

Consent for publication of de-identified data was informed and was obtained in written from all participants as part of the consenting process.

Competing interests

The authors declare no competing interests.