Abstract
Background Routine screening for nontuberculous mycobacterial (NTM) lung disease is dependent on sputum cultures. This is particularly challenging in the cystic fibrosis (CF) population due to reduced sputum production and low culture sensitivity. Biomarkers of infection that do not rely on sputum may lead to earlier diagnosis, but validation trials require a unique prospective design.
Purpose The rationale of this trial is to investigate the utility of urine lipoarabinomannan (LAM) as a test to identify people with CF with a new positive NTM culture. We hypothesize that urine LAM is a sensitive, non-invasive screening test with a high negative predictive value to identify individuals with a relatively low risk of having positive NTM sputum culture.
Study design This is a prospective, single-center, non-randomized observational study in adults with CF, 3 years of negative NTM cultures, and no known history of NTM positive cultures. Patients are followed for two year-long observational periods with the primary endpoint being a positive NTM sputum culture within a year of a positive urine LAM result and a secondary endpoint of a positive NTM sputum culture within 3 years of a positive urine LAM result. Study implementation includes remote consent and sample collection to accommodate changes from the COVID-19 pandemic.
Conclusions This report describes the study design of an observational study aimed at using a urine biomarker to assist in the diagnosis of NTM lung infection in pwCF. If successful, urine LAM could be used as an adjunct to traditional sputum cultures for routine NTM screening.
INTRODUCTION
People with cystic fibrosis (pwCF) are a population at risk for developing pulmonary nontuberculous mycobacteria (NTM) infection, and despite recent advances in the field, diagnosis and treatment remains challenging (1-3). Historically, approximately 20% of cystic fibrosis (CF) patients have a positive NTM culture over a 5 year period (4); thus, annual screening for NTM is recommended (3). M. avium complex (MAC) and M. abscessus (MABSC) are responsible for nearly all NTM infections in pwCF living in the US (5). Both species can cause NTM lung disease, which is associated with significant morbidity and mortality.
The diagnosis of NTM lung disease in pwCF is challenging, as co-infections with bacteria such as P. aeruginosa and S. aureus are common, and may result in identical symptoms and radiographic findings (6). Frequently, NTM may be detected in a culture, but the infection then resolves without treatment or remains indolent for many years. A retrospective review from the Colorado Adult CF Center showed that the majority of participants with a positive NTM culture had transient (23%) or persistent but clinically indolent NTM infection (38.5%) (7). Currently, sputum culture is the “gold standard” for detecting NTM infection and the assay by which all diagnostic and treatment decisions are made (1). However, limitations of sputum culture include slow growth, high cost, and low sensitivity due to the required decontamination steps (3, 8-11). Also, given advancements in CF care and widespread use of cystic fibrosis transmembrane conductance regulator (CFTR) modulator therapy, pwCF are increasingly unable to routinely expectorate sputum (12) and are presenting less often to clinic or with pulmonary exacerbations (13). This combination of factors highlights the need for sputum-independent markers of NTM infection in this population (14).
Lipoarabinomannan (LAM) is a cell wall lipoglycan found in all mycobacteria species and a biomarker of mycobacterial infection (15, 16). During infection, LAM is released from metabolically active or degrading mycobacterial cells into the circulation with subsequent filtration by the kidneys and passed into the urine (17, 18). LAM is a validated biomarker for active tuberculosis (TB) when antigen detection, rather than antibody measurement, is applied (15, 19-23). These antigen detection assays for TB based on LAM are commercially available both as ELISAs and as a point-of-care test. However, the utility of a commercially available TB-specific LAM antigen was tested in a Danish CF clinic and was determined not to be suitable for diagnosis of NTM due to low concentrations of LAM antigen detected in urine from NTM positive pwCF (24). We reported measurable detection of urine LAM utilizing gas chromatography–mass spectrometry (GC-MS) analysis that corresponded with previous NTM culture history in pwCF (n=44) with well-documented NTM status (25). Urine LAM analysis also reflected treatment response in a pwCF receiving bacteriophage therapy for M. abscessus (26).
Herein, we present the design of a prospective, single-center, nonrandomized observational study to investigate the utility of urine LAM as a test to identify pwCF with clinically low suspicion for having a sputum culture positive for NTM. The primary objective of this study is to determine the ability of the urine LAM assay to predict continued negative NTM sputum cultures over a 12-month period. Secondary objectives are to determine the ability of the urine LAM assay to predict a new positive NTM sputum culture within 12 months and establish the time between a positive urine LAM assay and a newly positive NTM sputum culture. We hypothesize that urine LAM is a sensitive, non-invasive screening test to identify NTM infection in individuals with a relatively low risk of having a positive NTM sputum culture.
METHODS
Study Design
PAINLESS is a prospective, single-center, nonrandomized observational study. The study is approved by the Biomedical Research Alliance of New York (BRANY) review board (20-08-402-528) and is registered in ClinicalTrials.gov (NCT04579211). We comply with the Declaration of Helsinki and Good Clinical Practice guidelines. Written or electronic informed consent is obtained from all participants with a target enrollment of 100. As subject enrollment began on October 23 2020 and continued during restricted clinic access due to the Covid-19 pandemic, informed consent remains available by utilizing an innovative, electronic consenting platform (ClinOne, Inc). The use of electronic consent eliminates the need to have prolonged in-person contact and decreases the risk of infection from COVID-19 and other viruses/bacteria present in a healthcare setting. An instructional video is available for review prior to participation to further explain the rationale for the study, sample collection procedures, and sample timelines (27). Subjects are provided a urine collection kit to expedite sample drop-off. Subjects who are consented in-person are provided written informed consent and may collect the urine specimen at the time of the visit or at home. Subjects are followed on protocol for a minimum of two year-long observation periods (24±3 months). The trial is currently ongoing. The schedule of events is shown in Figure 1 and an overview of the PAINLESS study design and timeline is shown in Figure 2.
a Refer to PAINLESS Study Timeline (Figure 2). b On-study clinical follow-up visits will occur as clinically indicated in clinic, the hospital or at home. c End-of-study visit will be the first clinical visit occurring > 9 months after the start of Observational Year 2. d Study personnel will continue to review NTM culture results and clinical events recorded in the CFFPR for up to 5 years following end of study. e Samples may be obtained in clinic, the hospital or at home, either expectorated or sputum induction. If results from oropharyngeal swab or bronchoscopy are available, they will also be recorded. f Urine sample must be collected within 30 days of a sputum sample at the start of Year 1 and Year 2. End-of-study urine sample does not need to be linked to a sputum sample. g Standard clinical assessments will be obtained from CFF patient registry database. h Replacement urine for LAM assay if needed.
Panel A: Screening criteria requires no history of NTM positive cultures combined with at least 3 negative sputum cultures for NTM over the past 3 years. Panel B: At enrollment, the first urine sample for LAM analysis is collected and is within 30 days of a NTM sputum culture. It is anticipated that a positive urine LAM will be detected in a subset of subject, who will comprise a “high risk cohort” that is more likely to have a positive NTM sputum culture. Panel C: Over the course of the study, NTM sputum cultures will be obtained when available in the context of clinical care. Panel D: At the end of Year 1, a second urine will be collected for LAM analysis. It is anticipated that some subjects with negative LAM at enrollment will convert to positive at each follow-up interval, and likewise some subjects with positive LAM at enrollment will convert to negative. Panel E: At the end of the second year of observation a 3rd urine sample will be collected for LAM analysis, as well as sputum for NTM culture (if available). Panel F: Results from future sputum NTM cultures will be monitored from the CFF Patient Registry for up to 5 years following the end of study procedures.
Study population
Adults (≥18 years old) with a confirmed diagnosis of CF followed at the Colorado Adult CF Center who have a minimum of 3 negative NTM cultures with 3 years of available clinical culture data and at least one negative NTM culture within the year prior to enrollment and no known history of previous NTM positive cultures are eligible for enrollment. There are no criteria related to the use of CFTR modulators or other ongoing therapies unrelated to NTM. It is anticipated that up to 90% of participants will be receiving CFTR modulator therapy for the entire duration of the trial, consistent with current usage within the study population. Complete inclusion and exclusion criteria are shown in Table 1.
Subject related data
Demographic information including date of birth, sex, and race are recorded. The diagnosis of CF consistent with the 2017 CFF Guidelines is confirmed (28). Dates and results from all available NTM cultures are recorded. A minimum of 3 NTM cultures over a 3 year prior to enrollment must be available and documented, with at least 1 negative within the year previous to enrollment. Use of CFTR modulators is also recorded. All subjects in the PAINLESS Trial must also be enrolled in the CF Foundation Patient Registry (CFFPR) (29). If needed, data may be extracted from the CFFPR for exploratory analysis and long-term follow-up.
Sputum collection and culture analysis
All sputum cultures are collected in the context of clinical care. Expectorated or induced sputum samples are collected during clinic visits and hospitalizations. Sputum samples collected at home may be brought to a clinic appointment or shipped using a sputum collection kit. The sputum collection kits, including instructions for collecting and shipping sputum specimens, are routinely ordered by providers to facilitate sputum culture collection. Samples obtained by bronchoscopy (if available) will also be recorded when collected for clinical indications. Microbiology and mycobiology results are recorded in the subject’s electronic medical record and the CFF Patient Registry. Sputum samples are cultured for NTM as well as typical CF pathogens. NTM cultures are performed at the Advanced Diagnostic Laboratories (ADx) at National Jewish Health, which is a national reference laboratory for mycobacteriology species identification and drug resistance testing (30). Following digestion and decontamination, specimens are inoculated on a LJ Slant, Middlebrook 7H11 Agar/ Mitchison 7H11 Selective Agar biplate(s) and MGIT for up to eight weeks. Suspected NTM growth is confirmed from either liquid and solid culture media using a Ziehl-Neelsen staining. Upon growth detection, genomic DNA is isolated for subsequent PCR amplification. A targeted segment of the DNA-directed RNA polymerase subunit beta (rpoB) gene is performed to determine molecular identification.
Urine collection for LAM analysis
The initial urine sample is obtained at enrollment and marks the start of observational year 1. If the subject is consented remotely, the subject completes the electronic consent process prior to collecting the initial urine sample. The initial urine sample is required within 30 days of a NTM sputum sample collected in the context of clinical care. Following enrollment, additional urine samples are collected at the start of observational year 2 and at the end of study. An additional urine collection may occur in the event a subject is found to have a positive NTM sputum culture between the planned collection dates. The urine sample may be collected in the clinic or at home with a collection kit and brought to clinic. Subjects are asked to provide between 20cc to 50cc of urine per collection. After collection, urine samples centrifuged, aliquoted, and stored at -80ºC until analysis.
LAM assay methods
All urine samples require a hexane wash (1:1 v/v) to remove exogenous lipids and proteins in order to reduce background and increase the limit of detection. Coded samples are separated by hydrophobic interaction chromatography over Octyl Sepharose (OS)-CL 4B. The 40% & 65% n-propanol in 0.1 M NH4OAc eluents off the HIC column are processed for GC/MS analysis as previously described (31, 32). GC/MS analyses are carried out using a Thermo GC-TSQ8000 Evo Triple Quad GC mass spectrometer. Chromatograms with respective peaks are integrated manually (i.e. peak areas defined manually and integrated areas generated by software) for the estimation of total D-ara and TBSA content. LAM used for standardization and assay validation in this study is purified from Mtb and M. avium (Figure 3) (33).
Representative GC/MS chromatograms showing the absence and/or presence of urinary LAM related to NTM-negativity and/or positivity. A) D-Arabinose (D-ara) MS/MS method monitoring m/z 420.9-m/z 192.9): Four characteristic peaks of Internal Standard (13C5-D-Arabinose; top panel); Sequentially (top to bottom), 309 and 308 NTM negative, D-ara negative; BP201 and 303 are positive, LAM positive, NTM positive. B) Tuberculostearic acid (TBSA) Single ion monitoring (SIM) at m/z 297.3: TBSA standard (C:19; top panel); Sequentially, 309 and 308 NTM negative, TBSA negative; BP201 and 303 are TBSA positive, LAM positive, NTM positive.
DATA COLLECTION, MANAGEMENT, and ANALYSIS
Study data are collected and managed using the REDCap electronic data capture tools hosted at National Jewish Health. REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies (34, 35). The handling and analysis of the data is conducted using good computing practices for clinical trials, including routine data validation checks. As part of data analysis, the lead biostatistician on the study monitors for data outliers and performs data validation. If data entry errors are encountered, the study database is updated in accordance with the resolved queries. All changes to the study database are documented in an audit trail. Database lock will occur once quality assurance procedures have been completed and will be maintained for 5 years.
General analysis plan
Sample size
Planned sample size is 100 subjects. Based on historical data from the Colorado Adult CF Program, we anticipate approximately 13 subjects will be identified with a new positive NTM culture over the course of two year-long intervals.
Primary analysis
The primary analyses will focus on establishing the predictive accuracy of the urine LAM assay compared to both NTM culture results from the day of sampling and aggregated culture results from within a year of the LAM measurement date. As a secondary endpoint, we will focus on how well a positive urine LAM assay predicts a positive NTM culture over a three year period. For the primary endpoints, the classification performance of the LAM assay will be assessed with respect to its ability to predict a positive culture at the time of the test and over the 12-month period prior or following the test. Of primary interest are the test’s sensitivity and specificity, but we will also investigate positive predictive power, negative predictive power, percent-agreement, and Cohen’s Kappa. These values will be computed twice using the binary results of the LAM assay (positive or not-positive) paired with either the binary culture results from the time of the test or with the aggregated results over the study interval (i.e. ≧1 positive culture versus no positive cultures). We will also investigate if the continuous abundance of LAM is a useful measure by calculating area under the receiver-operator curve, again using culture from the time of the test or aggregated over the 12-month period. Additionally, the multiple repeated observations for each individual in the two 12-month follow-up periods will allow us to perform a time-to-event analysis associating the LAM assay to culture results. Here, Cox proportional hazards regression models will be utilized with time to first positive NTM culture as the outcome and LAM assay (using both the continuous and binary versions) as the time-varying covariate of interest.
DISCUSSION
Consensus recommendations call for annual NTM screening for pwCF who can produce sputum, as well as screening before and six months after beginning chronic azithromycin therapy and annually thereafter (3). Screening for NTM infection in the CF population is currently reliant on traditional sputum culture techniques, as oropharyngeal swabs are not recommended due to low sensitivity. Sputum sampling has long been challenging in pwCF who do not produce sputum, and reliance on sputum cultures has become even less feasible with widespread adoption of CFTR modulator therapy. The CFTR triple combination therapy of elexacaftor, tezacaftor, and ivacaftor (E/T/I) was approved in 2019 in the U.S. for pwCF who are either homozygous (36) or heterozygous for F508del mutation (37). A reduction in sputum production is a nearly universal benefit of E/T/I (12). The 2022 CFFPR reported an overall decrease in both microbiology and mycobacterial cultures reported for the last three years These decreases can be attributed to not only the benefits of CFTR modulator therapy, but reduced frequency of in-person care as a result of the COVID-19 pandemic (35). However, the ability of a pwCF to produce a sputum sample has remained steady at 70% since the introduction of E/T/I, though it begins to drop after the age of 30 years. As the life expectancy of pwCF continues to rise, it can be expected that fewer and fewer pwCF will be able to spontaneously expectorate sputum and yields from induced sputum will fall as well. Thus, it is widely recognized that sputum-independent, non-culture-based markers for NTM are an unmet need in CF care. (14, 38).
Airway infection with NTM are among the most serious and least understood comorbidities of CF lung disease, with some patients developing rapid progression of disease. These patients may potentially benefit from an early eradication approach to treatment. PAINLESS is a single-center pilot trial created to investigate the utility of urine LAM as a screening test for the majority of pwCF who will likely never have a sputum culture positive for NTM. Results from this study will determine the feasibility of a larger multi-center trial to test this noninvasive urine assay as a strategy to eliminate the need for routine sputum screening for NTM. A positive urine LAM would notify clinicians of the need for airway cultures and close follow-up, while a negative result would assure the patient, family, and care team that in the absence of clinical suspicion, additional costly and invasive testing is not needed. Given the relatively low sensitivity of NTM cultures, the potential exists that when positive, the urine LAM assay would result in earlier detection than annual sputum screening and serve as an important component to an evidence-based eradication protocol, which currently does not exist. This study also serves as a platform for the development of a more accessible ELISA-based assay for the detection of urine LAM to be potentially used widely in clinical practice. This trial is the first to our knowledge for noninvasive screening of NTM infection in the CF population and also pilots aspects of remote consent and sample collection.
The PAINLESS trial has several limitations that are inherent to the identification and validation of biomarkers. Assessment of urine LAM as an assay is linked to sputum cultures, which are insensitive in this setting, and will often not be available. Thus, in some cases positive urine LAM values will be incorrectly assessed as false positives, when in fact they may be a more sensitive marker than the “gold standard” sputum culture. In addition, it is well recognized that NTM may be transient in the CF population (7), and this phenomena may also result in an apparent false positive urine LAM result. Finally, the current trial utilizes GC/MS analysis of urine LAM, which is not yet standardized. Given these limitation, the greatest value of urine LAM may be to identify those who don’t have an NTM infection. Urine LAM collection for the current trial will end in 2024, with a follow-up period needed to assess sputum culture results. Full validation of the method will require a larger, prospective trial including both children and adults over an extended duration.
Data Availability
No datasets were generated or analysed during the current study. All relevant data from this study will be made available upon study completion.
NA
Funding
Funding provided by the Cystic Fibrosis Foundation, grant numbers: NICK17K0, NICK18P0, NICK20Y2-SVC, NICK20Y2-OUT, NICK23A1, and the National Institutes of Health grant number: R01HL146228. The sponsors have no part in study design, collection, management analysis, interpretation of data, writing of the report, and the decision to submit the report for publication.
Conflict of interest statement
KC, EA, KP, SC, VKL, MJ, KCM, BV, EW, NR, JM, IA, DC, JAN: None to disclose
Abbreviations
- CF
- cystic fibrosis
- NTM
- nontuberculous mycobacteria
- pwCF
- people with cystic fibrosis
- CFTR
- cystic fibrosis transmembrane conductance regulator
- LAM
- lipoarabinomannan
- MAC
- Mycobacterium avium complex
- MABSC
- Mycobacterium abscessus
Works Cited
