Simon Grandjean Lapierre

Tuberculosis (TB) remains a leading cause of infectious disease morbidity and mortality worldwide despite substantial advances in diagnostic technologies. Multiple international and national organizations have developed evidence-based recommendations for TB diagnosis; however, differences in guidance can create uncertainty for laboratories, clinicians, and policymakers implementing diagnostic services. Understanding areas of consensus and divergence may facilitate evidence-informed decision-making and improve access to appropriate TB diagnostics. This minireview compares major TB diagnostic guidelines from the World Health Organization and organizations in Canada, the United States, Europe, the United Kingdom, and Australia, highlighting areas of consensus and divergence. While substantial agreement exists regarding the core components of TB diagnosis, recommendations differ in the positioning of molecular diagnostics, specimen selection, biomarker use, sequencing technologies, and treatment monitoring. Understanding the factors underlying these differences, including epidemiology, laboratory capacity, resource availability, and approaches to evidence appraisal, can support context-specific implementation of diagnostic strategies, improve equitable access to high-quality TB diagnostics, and identify priorities for future research and guideline harmonization.

Background: The analytical performance of bacterial targeted sequencing (BTS), fungal targeted/panfungal sequencing (FTS), and metagenomic next-generation sequencing (mNGS) assays has been previously evaluated and their clinical use is increasing. Limited evidence is available on their true clinical impact on infectious disease diagnosis and treatment. Methods: We conducted a 3-year retrospective cohort study including all patients for whom broad-range sequencing assays were performed directly from clinical samples for the detection of bacterial and fungal pathogens. The operational characteristics, diagnostic and therapeutic impacts of the assays were assessed by reviewing patient clinical files and laboratory information system charts. Results: A total of 279 samples from 185 patients were included. The positivity rates for BTS, FTS, and mNGS were respectively 20.5% (47/229), 20% (9/45), and 20% (1/5). Of these 279 samples, 40 (14.3%) had an impact on patient management. The test results helped to establish a diagnosis in 26 (9.3%) cases and led to treatment modifications in 14 (5%). FTS achieved higher impact rates (26.7%) than both BTS (12.2%) and mNGS (0%). Short turnaround times increase impact rates, and the most impactful tests were those performed on bone and intervertebral disc samples, or in patients with negative culture results due to prior antibiotic administration. Conclusions: In this study, the overall diagnostic impact of BTS and FTS was high. Both the diagnostic and treatment impact of those assays can be increased if prescribed in well-selected clinical syndromes and performed on well-selected clinical samples.

The COugh Diagnostic Algorithm for Tuberculosis (CODA TB) DREAM Challenge recently evaluated the performance of artificial intelligence (AI) algorithms for tuberculosis (TB) screening using cough sounds. Eleven AI models were developed using a dataset of 733,756 cough sounds collected from 2143 adults from seven countries. This study evaluates the CODA Challenge AI models with an external independent cough dataset from Peru. Cough recordings from 303 coughing adults were collected from health facilities in Lima, Peru. The AUCs of the models ranged from 0.480 to 0.615, showing a decrease in performance compared to their performance when internally validated using the CODA Challenge, which ranged from 0.689 to 0.743. The best performing model in the CODA Challenge was also the best performing model in this external validation. Sub-group analyses revealed that models performed better in older (≥ 35 years) populations and among people with prior TB. The external validation revealed limitations in the generalizability of the CODA Challenge models to other settings. While some models showed promise, the overall performance decline highlights the need for continued model validation on external datasets. It also underscores the importance of developing context-specific models to account for population-specific factors that influence cough characteristics and TB prevalence.

Background: Open-access data challenges can accelerate innovation in artificial intelligence–based tools. In the Cough Diagnostic Algorithm for Tuberculosis (CODA TB) DREAM Challenge, we developed and independently validated cough sound–based artificial intelligence algorithms for tuberculosis screening. Methods: We included data from 2143 adults with ≥2 weeks of cough from outpatient clinics in India, Madagascar, the Philippines, South Africa, Tanzania, Uganda, and Vietnam. A standard tuberculosis evaluation was completed, and ≥3 solicited coughs were recorded using a smartphone. We invited teams to develop models using training data to classify microbiologically confirmed tuberculosis disease using (1) cough sound features only and/or (2) cough sound features with routinely available clinical data. After 4 months, they submitted the algorithms for independent test set validation. Models were ranked by area under the receiver operating characteristic curve (AUROC) and partial AUROC (pAUROC) to achieve at least 80% sensitivity and 60% specificity. Results: Eleven cough models and 6 cough-plus-clinical models were submitted. AUROCs for cough models ranged from 0.69 to 0.74, and the highest performing model achieved 55.5% specificity (95% confidence interval, 47.7%–64.2%) at 80% sensitivity. The addition of clinical data improved AUROCs (range, 0.78–0.83); 5 of the 6 models reached the target pAUROC, and the highest performing model had 73.8% specificity (95% confidence interval, 60.8%–80.0%) at 80% sensitivity. The AUROC varied by country and was higher among male and human immunodeficiency virus–negative individuals. Conclusions: In a short period, an open-access data challenge facilitated the development of new cough-based tuberculosis algorithms and demonstrated potential as a tuberculosis screening tool.

Background: Molecular typing of Mycobacterium tuberculosis complex isolates enhances understanding of tuberculosis (TB) transmission dynamics, supporting public health efforts in outbreak investigations. This study aims to validate TBpore, a novel bioinformatic pipeline for clustering TB transmission isolates using Oxford Nanopore Technology (ONT) data and comparing it against conventional Mycobacterial Interspersed Repetitive-Unit Variable Number (MIRU-VNTR) typing and Illumina sequencing. Methodology/Principal findings: This retrospective case-control study included 58 clinical isolates from two TB outbreaks in Canada, previously characterized by public health investigations and MIRU-VNTR typing. DNA extraction and sequencing were performed on both Illumina and ONT platforms. Illumina data were processed using Clockwork and psdm, while Nanopore data were analyzed with TBpore. SNP distances were used to compare clustering results across methods, with clusters defined by SNP distance thresholds of ≤5 and ≤12. Both sequencing methods showed a high degree of concordance in clustering results. All isolates from the M. africanum outbreak clustered within the defined SNP thresholds, consistent with MIRU-VNTR and epidemiological data. In the M. tuberculosis outbreak, 20 out of 21 isolates clustered similarly across methods, with one exception. Within outbreak pairwise SNP distances were lower with Nanopore. Conclusion/Significance: ONT sequencing and the TBpore pipeline offer an accurate alternative to Illumina technology for TB molecular epidemiology. This study suggests potential increased clustering sensitivity with Nanopore technology, warranting further validation on larger datasets with robust epidemiological metadata.

Idiopathic granulomatous mastitis (IGM) is a rare inflammatory disease of the breast. Various clinical management approaches have been described, but their efficacy and optimal sequential order remain uncertain. We describe the first Canadian cohort of patients with IGM, discuss treatment outcomes and outline a practical management approach. This retrospective study included patients diagnosed with biopsy-confirmed IGM between 2014 and 2023, aged over 18 years. Based on a scoping review of the literature, a diagnostic and management approach was developed, and we present here the disease course and outcomes using this approach. 22 females were included, with a mean age of 40 (24-65) years, mostly presenting with a breast lump (n =22, 100%) and breast pain (n = 15, 68%). Mean rheumatology follow-up was 28.7 months (range 3-79). Mean time from first symptom to diagnosis was 3.5 months (range 1-13). Corynebacterium kroppenstedtii was found in 8 patients. Treatment including lipophilic antibiotics, corticosteroids and disease modifying anti-rheumatic drugs (DMARDs), led to complete remission in 95% of patients, in a mean time of 11.6 months (range 1-36), and relapse in only 1 patient. 11 patients required DMARDs (50%), most commonly methotrexate (n=9). We highlight the variable severity of IGM and the benefits of a severity-based treatment approach. A diligent evaluation and work-up is essential to manage IGM. The proposed severity-based management approach with medical treatment and less aggressive surgical intervention led to complete remission in 95%.