Estimation of country-specific tuberculosis resistance antibiograms using pathogen genomics and machine learning

BACKGROUND

Global tuberculosis (TB) drug resistance (DR) surveillance focuses on rifampicin. We examined the potential of public and surveillance Mycobacterium tuberculosis (Mtb) whole-genome sequencing (WGS) data, to generate expanded country-level resistance prevalence estimates (antibiograms) using in silico resistance prediction.

METHODS

We curated and quality-controlled Mtb WGS data. We used a validated random forest model to predict phenotypic resistance to 12 drugs and bias-corrected for model performance, outbreak sampling and rifampicin resistance oversampling. Validation leveraged a national DR survey conducted in South Africa.

RESULTS

Mtb isolates from 29 countries (n=19 149) met sequence quality criteria. Global marginal genotypic resistance among mono-resistant TB estimates overlapped with the South African DR survey, except for isoniazid, ethionamide and second-line injectables, which were underestimated (n=3134). Among multidrug resistant (MDR) TB (n=268), estimates overlapped for the fluoroquinolones but overestimated other drugs. Globally pooled mono-resistance to isoniazid was 10.9% (95% CI: 10.2-11.7%, n=14 012). Mono-levofloxacin resistance rates were highest in South Asia (Pakistan 3.4% (0.1-11%), n=111 and India 2.8% (0.08-9.4%), n=114). Given the recent interest in drugs enhancing ethionamide activity and their expected activity against isolates with resistance discordance between isoniazid and ethionamide, we measured this rate and found it to be high at 74.4% (IQR: 64.5-79.7%) of isoniazid-resistant isolates predicted to be ethionamide susceptible. The global susceptibility rate to pyrazinamide and levofloxacin among MDR was 15.1% (95% CI: 10.2-19.9%, n=3964).

CONCLUSIONS

This is the first attempt at global Mtb antibiogram estimation. DR prevalence in Mtb can be reliably estimated using public WGS and phenotypic resistance prediction for key antibiotics, but public WGS data demonstrates oversampling of isolates with higher resistance levels than MDR. Nevertheless, our results raise concerns about the empiric use of short-course fluoroquinolone regimens for drug-susceptible TB in South Asia and indicate underutilisation of ethionamide in MDR treatment.

Limited Nosocomial Transmission of Drug-Resistant Tuberculosis, Moldova

Applying whole-genome-sequencing, we aimed to detect transmission events of multidrug-resistant/rifampin-resistant strains of Mycobacterium tuberculosis complex at a tuberculosis hospital in Chisinau, Moldova. We recorded ward, room, and bed information for each patient and monitored in-hospital transfers over 1 year. Detailed molecular and patient surveillance revealed only 2 nosocomial transmission events.

Clinical implications of molecular drug resistance testing for Mycobacterium tuberculosis: a 2023 TBnet/RESIST-TB consensus statement

Drug-resistant tuberculosis is a substantial health-care concern worldwide. Despite culture-based methods being considered the gold standard for drug susceptibility testing, molecular methods provide rapid information about the Mycobacterium tuberculosis mutations associated with resistance to anti-tuberculosis drugs. This consensus document was developed on the basis of a comprehensive literature search, by the TBnet and RESIST-TB networks, about reporting standards for the clinical use of molecular drug susceptibility testing. Review and the search for evidence included hand-searching journals and searching electronic databases. The panel identified studies that linked mutations in genomic regions of M tuberculosis with treatment outcome data. Implementation of molecular testing for the prediction of drug resistance in M tuberculosis is key. Detection of mutations in clinical isolates has implications for the clinical management of patients with multidrug-resistant or rifampicin-resistant tuberculosis, especially in situations when phenotypic drug susceptibility testing is not available. A multidisciplinary team including clinicians, microbiologists, and laboratory scientists reached a consensus on key questions relevant to molecular prediction of drug susceptibility or resistance to M tuberculosis, and their implications for clinical practice. This consensus document should help clinicians in the management of patients with tuberculosis, providing guidance for the design of treatment regimens and optimising outcomes.

Rethinking global digital health and AI-for-health innovation challenges

Digital health technologies can help tackle challenges in global public health. Digital and AI-for-Health Challenges, controlled events whose goal is to generate solutions to a given problem in a defined period of time, are one way of catalysing innovation. This article proposes an expanded investment framework for Global Health AI and digitalhealth Innovation that goes beyond traditional factors such as return on investment. Instead, we propose non monetary and non GDP metrics, such as Disability Adjusted Life Years or achievement of universal health coverage. Furthermore, we suggest a venture building approach around global health, which includes filtering of participants to reduce opportunity cost, close integration of implementation scientists and an incubator for the long-term development of ideas resulting from the challenge. Finally, we emphasize the need to strengthen human capital across a range of areas in local innovation, implementation-science, and in health services.

Piloting a Survey-Based Assessment of Transparency and Trustworthiness with Three Medical AI Tools

Artificial intelligence (AI) offers the potential to support healthcare delivery, but poorly trained or validated algorithms bear risks of harm. Ethical guidelines stated transparency about model development and validation as a requirement for trustworthy AI. Abundant guidance exists to provide transparency through reporting, but poorly reported medical AI tools are common. To close this transparency gap, we developed and piloted a framework to quantify the transparency of medical AI tools with three use cases. Our framework comprises a survey to report on the intended use, training and validation data and processes, ethical considerations, and deployment recommendations. The transparency of each response was scored with either 0, 0.5, or 1 to reflect if the requested information was not, partially, or fully provided. Additionally, we assessed on an analogous three-point scale if the provided responses fulfilled the transparency requirement for a set of trustworthiness criteria from ethical guidelines. The degree of transparency and trustworthiness was calculated on a scale from 0% to 100%. Our assessment of three medical AI use cases pin-pointed reporting gaps and resulted in transparency scores of 67% for two use cases and one with 59%. We report anecdotal evidence that business constraints and limited information from external datasets were major obstacles to providing transparency for the three use cases. The observed transparency gaps also lowered the degree of trustworthiness, indicating compliance gaps with ethical guidelines. All three pilot use cases faced challenges to provide transparency about medical AI tools, but more studies are needed to investigate those in the wider medical AI sector. Applying this framework for an external assessment of transparency may be infeasible if business constraints prevent the disclosure of information. New strategies may be necessary to enable audits of medical AI tools while preserving business secrets.

A convolutional neural network highlights mutations relevant to antimicrobial resistance in Mycobacterium tuberculosis

Long diagnostic wait times hinder international efforts to address antibiotic resistance in M. tuberculosis. Pathogen whole genome sequencing, coupled with statistical and machine learning models, offers a promising solution. However, generalizability and clinical adoption have been limited by a lack of interpretability, especially in deep learning methods. Here, we present two deep convolutional neural networks that predict antibiotic resistance phenotypes of M. tuberculosis isolates: a multi-drug CNN (MD-CNN), that predicts resistance to 13 antibiotics based on 18 genomic loci, with AUCs 82.6-99.5% and higher sensitivity than state-of-the-art methods; and a set of 13 single-drug CNNs (SD-CNN) with AUCs 80.1-97.1% and higher specificity than the previous state-of-the-art. Using saliency methods to evaluate the contribution of input sequence features to the SD-CNN predictions, we identify 18 sites in the genome not previously associated with resistance. The CNN models permit functional variant discovery, biologically meaningful interpretation, and clinical applicability.

Management of childhood MDR-TB in Europe and Central Asia: report of a Regional WHO meeting

BACKGROUND: As the WHO European Region has the highest proportion of multidrug-resistant TB (MDR-TB) among total incident TB cases, many children and adolescents are at risk of MDR-TB infection and disease.METHODS: We performed an electronic survey of clinicians and TB programme personnel who attended the 2020 Regional Consultation on child and adolescent TB organised by the WHO Regional Office. We characterised access to diagnostics and drugs, and practices in the prevention and management of child and adolescent MDR-TB.RESULTS: Children and adolescents are inconsistently represented in national guidelines and budgets; child-friendly drug formulations for MDR-TB treatment are insufficiently available in 57% of countries, and 32% of countries reported paediatric drug stock-outs. The novel drugs, bedaquiline and delamanid, are accessible by respectively 80% and 60% of respondent countries. Respondents were asked how many children were diagnosed with MDR-TB in 2019, and a comparison of this number to modelled estimates of incidence (to identify the case detection gap) and WHO notifications (to identify the case reporting gap) showed substantial differences in both comparisons.CONCLUSIONS: Better representation of this patient group in guidelines and budgets, greater access to drugs and improved reporting are essential to reach TB elimination in this Region.

GenTB: A user-friendly genome-based predictor for tuberculosis resistance powered by machine learning

BACKGROUND

Multidrug-resistant Mycobacterium tuberculosis (Mtb) is a significant global public health threat. Genotypic resistance prediction from Mtb DNA sequences offers an alternative to laboratory-based drug-susceptibility testing. User-friendly and accurate resistance prediction tools are needed to enable public health and clinical practitioners to rapidly diagnose resistance and inform treatment regimens.

RESULTS

We present Translational Genomics platform for Tuberculosis (GenTB), a free and open web-based application to predict antibiotic resistance from next-generation sequence data. The user can choose between two potential predictors, a Random Forest (RF) classifier and a Wide and Deep Neural Network (WDNN) to predict phenotypic resistance to 13 and 10 anti-tuberculosis drugs, respectively. We benchmark GenTB's predictive performance along with leading TB resistance prediction tools (Mykrobe and TB-Profiler) using a ground truth dataset of 20,408 isolates with laboratory-based drug susceptibility data. All four tools reliably predicted resistance to first-line tuberculosis drugs but had varying performance for second-line drugs. The mean sensitivities for GenTB-RF and GenTB-WDNN across the nine shared drugs were 77.6% (95% CI 76.6-78.5%) and 75.4% (95% CI 74.5-76.4%), respectively, and marginally higher than the sensitivities of TB-Profiler at 74.4% (95% CI 73.4-75.3%) and Mykrobe at 71.9% (95% CI 70.9-72.9%). The higher sensitivities were at an expense of ≤ 1.5% lower specificity: Mykrobe 97.6% (95% CI 97.5-97.7%), TB-Profiler 96.9% (95% CI 96.7 to 97.0%), GenTB-WDNN 96.2% (95% CI 96.0 to 96.4%), and GenTB-RF 96.1% (95% CI 96.0 to 96.3%). Averaged across the four tools, genotypic resistance sensitivity was 11% and 9% lower for isoniazid and rifampicin respectively, on isolates sequenced at low depth (< 10× across 95% of the genome) emphasizing the need to quality control input sequence data before prediction. We discuss differences between tools in reporting results to the user including variants underlying the resistance calls and any novel or indeterminate variants CONCLUSIONS: GenTB is an easy-to-use online tool to rapidly and accurately predict resistance to anti-tuberculosis drugs. GenTB can be accessed online at https://gentb.hms.harvard.edu , and the source code is available at https://github.com/farhat-lab/gentb-site .

The phylogenetic landscape and nosocomial spread of the multidrug-resistant opportunist Stenotrophomonas maltophilia

Recent studies portend a rising global spread and adaptation of human- or healthcare-associated pathogens. Here, we analyse an international collection of the emerging, multidrug-resistant, opportunistic pathogen Stenotrophomonas maltophilia from 22 countries to infer population structure and clonality at a global level. We show that the S. maltophilia complex is divided into 23 monophyletic lineages, most of which harbour strains of all degrees of human virulence. Lineage Sm6 comprises the highest rate of human-associated strains, linked to key virulence and resistance genes. Transmission analysis identifies potential outbreak events of genetically closely related strains isolated within days or weeks in the same hospitals.

Multidrug resistance of the opportunistic pathogen Stenotrophomonas maltophilia is an increasing problem. Here, analyzing strains from 22 countries, the authors show that the S. maltophilia complex is divided into 23 monophyletic lineages and find evidence for intra-hospital transmission.

Prioritising children and adolescents in the tuberculosis response of the WHO European Region

In 2017, in recognition of the challenges faced by Member States in managing childhood and adolescent tuberculosis (TB) at a country level, the WHO Regional Office for Europe held a Regional Consultation. In total, 35 countries participated in the consultations representing both high- and low-incidence Member States. Here, we provide an overview of the existing World Health Organization (WHO) documents and guidelines on childhood and adolescent TB and describe the outcomes of this regional meeting. National childhood and adolescent TB guidelines are available in 25% of Member States, while 33% reported that no such guidelines are at hand. In the majority of countries (83%), childhood and adolescent TB is part of the National Strategic Plan. The most pressing challenges in managing paediatric TB comprise the lack of adequate drug formulations, the difficult diagnosis, and treatment of presumed latent TB infection. Investments into childhood and adolescent TB need to be further advocated to achieve the End TB goals set by WHO to eliminate TB by 2030.

A regional consultation on child and adolescent TB was held in 2017 by the WHO Regional Office for Europe. It identified common challenges and key priorities, useful in informing and strengthening the regional response to TB.http://ow.ly/Fg8H30nwBRo

Random glucose sampling as screening tool for diabetes among disadvantaged tuberculosis patients residing in urban slums in India

Noncommunicable diseases like diabetes are increasingly recognised as important risk factors for tuberculosis (TB) and poor treatment outcomes [1]. While the link between TB and diabetes was described many decades ago, several recent epidemiological studies and systematic reviews have confirmed the association of diabetes with a three-fold increased risk of developing TB [2]. Since 2011, the World Health Organization has recommended bidirectional screening of all TB patients for diabetes [3]. However, it is currently unclear at which point in treatment one should screen and which diagnostic tools should be used. Following the American Diabetes Association, diabetes is diagnosed by a fasting plasma glucose ≥7 mmol·L⁻¹, a 2-h plasma glucose value ≥11.1 mmol·L⁻¹ during the oral glucose tolerance test, glycated haemoglobin (HbA1C) ≥48 mmol·mol⁻¹ or a random plasma glucose value ≥11.1 mmol·L⁻¹ in patients with classic symptoms of hyperglycaemia [4]. The Concurrent Tuberculosis and Diabetes Mellitus (TANDEM) consortium recently suggested a simplified two-step diagnostic algorithm where all patients with random plasma glucose levels >6.1 mmol·L⁻¹ receive point-of-care HbA1C testing [4]. With laboratory-based HbA1C as the gold standard, this two-step combination resulted in a sensitivity and specificity of >90% to detect diabetes. Here, we evaluate the feasibility of diabetes screening by random glucose sampling among disadvantaged TB patients residing in urban slums in New Delhi, India.

Recently, a two-step diagnostic algorithm to diagnose diabetes among TB patients was proposed comprising random glucose and point-of-care HbA1c. This study evaluates the first part of this algorithm among disadvantaged TB patients.http://ow.ly/UI7d30nK1UN

Perspectives for personalized therapy for patients with multidrug-resistant tuberculosis

According to the World Health Organization (WHO), tuberculosis is the leading cause of death attributed to a single microbial pathogen worldwide. In addition to the large number of patients affected by tuberculosis, the emergence of Mycobacterium tuberculosis drug-resistance is complicating tuberculosis control in many high-burden countries. During the past 5 years, the global number of patients identified with multidrug-resistant tuberculosis (MDR-TB), defined as bacillary resistance at least against rifampicin and isoniazid, the two most active drugs in a treatment regimen, has increased by more than 20% annually. Today we experience a historical peak in the number of patients affected by MDR-TB. The management of MDR-TB is characterized by delayed diagnosis, uncertainty of the extent of bacillary drug-resistance, imprecise standardized drug regimens and dosages, very long duration of therapy and high frequency of adverse events which all translate into a poor prognosis for many of the affected patients. Major scientific and technological advances in recent years provide new perspectives through treatment regimens tailor-made to individual needs. Where available, such personalized treatment has major implications on the treatment outcomes of patients with MDR-TB. The challenge now is to bring these adances to those patients that need them most.

Non-Steroidal Anti-inflammatory Drugs As Host-Directed Therapy for Tuberculosis: A Systematic Review

Lengthy, antimicrobial therapy targeting the pathogen is the mainstay of conventional tuberculosis treatment, complicated by emerging drug resistances. Host-directed therapies, including non-steroidal anti-inflammatory drugs (NSAIDs), in contrast, target host factors to mitigate disease severity. In the present Systematic Review, we investigate whether NSAIDs display any effects as therapy of TB and discuss possible mechanisms of action of NSAIDs as adjunctive therapy of TB. Ten studies, seven preclinical studies in mice and three clinical trials, were included and systematically reviewed. Our results point toward a beneficial effect of NSAIDs as adjunct to current TB therapy regimens, mediated by decreased lung pathology balancing host-immune reaction. The determination of the best timing for their administration in order to obtain the potential beneficial effects needs further investigation. Even if the preclinical evidence requires clinical evaluation, NSAIDs might represent a potential safe, simple, and cheap improvement in therapy of TB.

CD4+ T Cells Recognizing PE/PPE Antigens Directly or via Cross Reactivity Are Protective against Pulmonary Mycobacterium tuberculosis Infection

Mycobacterium tuberculosis (Mtb), possesses at least three type VII secretion systems, ESX-1, -3 and -5 that are actively involved in pathogenesis and host-pathogen interaction. We recently showed that an attenuated Mtb vaccine candidate (Mtb Δppe25-pe19), which lacks the characteristic ESX-5-associated pe/ppe genes, but harbors all other components of the ESX-5 system, induces CD4⁺ T-cell immune responses against non-esx-5-associated PE/PPE protein homologs. These T cells strongly cross-recognize the missing esx-5-associated PE/PPE proteins. Here, we characterized the fine composition of the functional cross-reactive Th1 effector subsets specific to the shared PE/PPE epitopes in mice immunized with the Mtb Δppe25-pe19 vaccine candidate. We provide evidence that the Mtb Δppe25-pe19 strain, despite its significant attenuation, is comparable to the WT Mtb strain with regard to: (i) its antigenic repertoire related to the different ESX systems, (ii) the induced Th1 effector subset composition, (iii) the differentiation status of the Th1 cells induced, and (iv) its particular features at stimulating the innate immune response. Indeed, we found significant contribution of PE/PPE-specific Th1 effector cells in the protective immunity against pulmonary Mtb infection. These results offer detailed insights into the immune mechanisms underlying the remarkable protective efficacy of the live attenuated Mtb Δppe25-pe19 vaccine candidate, as well as the specific potential of PE/PPE proteins as protective immunogens.

Mycobacterium tuberculosis (Mtb), the causative agent of human tuberculosis, is one of the most widely spread human pathogens, responsible for more than 9.6 million of new tuberculosis cases and 1.5 million deaths, annually. The resurgence of pulmonary tuberculosis in immuno-compromised patients, including HIV-co-infected populations, and increasing spread of drug-resistant Mtb strains are worrying. Given the estimated 2 billion cases of latent Mtb infections and the only partial efficacy of the unique, currently available tuberculosis-vaccine Mycobacterium bovis BCG (Bacille Calmette-Guerin) it is necessary to develop improved vaccines. Here, we demonstrate that the host cellular immunity, mediated by CD4⁺ T lymphocytes, specific to the “PE/PPE” families of mycobacterial antigens, contribute to the protection against Mtb-induced disease. We revealed the fine composition of the PE/PPE-specific T cells by characterizing their effector functions and differentiation status. We previously described a live attenuated mycobacterial strain as a vaccine candidate that is able to induce such CD4⁺ T cells and which displays particular properties at stimulating the cells of the innate immune system. These responses play a central role in the initiation of the host defense and in the protection against tuberculosis. Our results pave the way for further development of candidates in preclinical models of anti-tuberculosis vaccination.

Multidrug-Resistant Tuberculosis Complicated by Nosocomial Infection with Multidrug-Resistant Enterobacteriaceae

Treatment of mycobacterial diseases such as tuberculosis (TB) entails long and intense antimicrobial therapy. TB patients are at risk of coinfection with other multidrug-resistant bacteria, such as those from Enterobacteriaceae family, because of antimicrobial selection pressure and nosocomial transmission during prolonged hospital admission. Here, we report on two patients treated for multidrug-resistant TB, who developed severe sepsis due to an extended spectrum β-lactamase producing organism. Diagnostic culture identified the venous access port as source, and upon surgical removal and antimicrobial therapy rapid clinical improvement was achieved. Increased awareness and knowledge on the prevalence of multi-resistant Enterobacteriaceae is needed, notably in TB centers, to provide a safe hospital environment to our patients.