DeepAMR for predicting co-occurrent resistance of Mycobacterium tuberculosis

MOTIVATION

Resistance co-occurrence within first-line anti-tuberculosis (TB) drugs is a common phenomenon. Existing methods based on genetic data analysis of Mycobacterium tuberculosis (MTB) have been able to predict resistance of MTB to individual drugs, but have not considered the resistance co-occurrence and cannot capture latent structure of genomic data that corresponds to lineages.

RESULTS

We used a large cohort of TB patients from 16 countries across six continents where whole-genome sequences for each isolate and associated phenotype to anti-TB drugs were obtained using drug susceptibility testing recommended by the World Health Organization. We then proposed an end-to-end multi-task model with deep denoising auto-encoder (DeepAMR) for multiple drug classification and developed DeepAMR_cluster, a clustering variant based on DeepAMR, for learning clusters in latent space of the data. The results showed that DeepAMR outperformed baseline model and four machine learning models with mean AUROC from 94.4% to 98.7% for predicting resistance to four first-line drugs [i.e. isoniazid (INH), ethambutol (EMB), rifampicin (RIF), pyrazinamide (PZA)], multi-drug resistant TB (MDR-TB) and pan-susceptible TB (PANS-TB: MTB that is susceptible to all four first-line anti-TB drugs). In the case of INH, EMB, PZA and MDR-TB, DeepAMR achieved its best mean sensitivity of 94.3%, 91.5%, 87.3% and 96.3%, respectively. While in the case of RIF and PANS-TB, it generated 94.2% and 92.2% sensitivity, which were lower than baseline model by 0.7% and 1.9%, respectively. t-SNE visualization shows that DeepAMR_cluster captures lineage-related clusters in the latent space.

AVAILABILITY AND IMPLEMENTATION

The details of source code are provided at http://www.robots.ox.ac.uk/∼davidc/code.php.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Risk Factors for Sputum Positive Pulmonary Tuberculosis Retreatment Cases and Factors Responsible for Treatment Outcome

BACKGROUND

Identification of the characteristics that confer higher risk of relapse, failure, or default and factors associate with treatment outcome in retreatment cases may help in planning country-specific prevention strategies.

OBJECTIVE

To evaluate the risk factors for retreatment failure, default or relapse and factors responsible for the treatment outcome.

METHODS

In this study sputum positive pulmonary TB retreatment cases were included. All patients were treated by eight months Revised National Tuberculosis Control Program (RNTCP) Cat II regimen. Outcome was recorded as Cured, Failure, Death or Defaulted.

RESULTS

Patients having body weight > 45 kgs had higher cure rates (94.74%). Poor outcome was significantly higher in patient with cavitory lesions on Chest X ray (30.43% vs 7.14%) and in patient with bilateral lesions (28.57% vs 4.35%). Patients with initial sputum of 3 + grade was significantly associated with poor outcome than having sputum of scanty to 2 + grade (26.93% vs 8%).

CONCLUSION

Patients presenting for TB retreatment have distinct demographic and clinical characteristics, important difference in retreatment outcomes in relation to different parameters. So, new country specific strategies are required to identify and address risk factor for retreatment cases and factors responsible for poor outcome of these cases.

Tetany in an Extensively Drug Resistant Tuberculosis (XDR-TB) Patient Treated with Capreomycin

Gross electrolytes disturbances including hypokalemia, hypomagnesaemia, and hypocalcaemia have been reported in tuberculosis patients who have been treated with capreomycin.1-3 Capreomycin is recommended in the treatment of M. tuberculosis isolates resistant to kanamycin at baseline in multi drug resistant tuberculosis patients (MDR - TB) and treatment of extensively drug resistant tuberculosis (XDR-TB) under programmatic management of drug resistant tuberculosis (PMDT) in India.4 We report a case of tetany in a extensively drug resistant tuberculosis (XDR-TB) patient treated with capreomycin. She developed hypokalemia after 7 weeks of administration of injection capreomycin intramuscularly daily in dose of 750 mg. Hypokalemia was refractory to intravenous potassium replacement therapy. At 12 weeks during the treatment she developed tetany and hypocalcaemia. Hypomagnesaemia was also associated with hypocalcaemia and hypokalemia. Normal level of serum potassium and calcium were achieved with correction of hypomagnesaemia.

Application of machine learning techniques to tuberculosis drug resistance analysis

MOTIVATION

Timely identification of Mycobacterium tuberculosis (MTB) resistance to existing drugs is vital to decrease mortality and prevent the amplification of existing antibiotic resistance. Machine learning methods have been widely applied for timely predicting resistance of MTB given a specific drug and identifying resistance markers. However, they have been not validated on a large cohort of MTB samples from multi-centers across the world in terms of resistance prediction and resistance marker identification. Several machine learning classifiers and linear dimension reduction techniques were developed and compared for a cohort of 13 402 isolates collected from 16 countries across 6 continents and tested 11 drugs.

RESULTS

Compared to conventional molecular diagnostic test, area under curve of the best machine learning classifier increased for all drugs especially by 23.11%, 15.22% and 10.14% for pyrazinamide, ciprofloxacin and ofloxacin, respectively (P < 0.01). Logistic regression and gradient tree boosting found to perform better than other techniques. Moreover, logistic regression/gradient tree boosting with a sparse principal component analysis/non-negative matrix factorization step compared with the classifier alone enhanced the best performance in terms of F1-score by 12.54%, 4.61%, 7.45% and 9.58% for amikacin, moxifloxacin, ofloxacin and capreomycin, respectively, as well increasing area under curve for amikacin and capreomycin. Results provided a comprehensive comparison of various techniques and confirmed the application of machine learning for better prediction of the large diverse tuberculosis data. Furthermore, mutation ranking showed the possibility of finding new resistance/susceptible markers.

AVAILABILITY AND IMPLEMENTATION

The source code can be found at http://www.robots.ox.ac.uk/ davidc/code.php.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

The impact of improved detection and treatment of isoniazid resistant tuberculosis on prevalence of multi-drug resistant tuberculosis: A modelling study

Introduction

Isoniazid-resistant, rifampin susceptible tuberculosis (INHR-TB) is the most common form of drug resistant TB globally. Treatment of INHR-TB with standard first-line therapy is associated with high rates of multidrug resistant TB (MDR-TB). We modelled the potential impact of INHR-TB detection and appropriate treatment on MDR-TB prevalence.

Methods

A decision analysis model was developed to compare three different strategies for the detection of TB (AFB smear, Xpert MTB/RIF, and Line-Probe Assays (LPA)), combined with appropriate treatment. The population evaluated were patients with a globally representative prevalence of newly diagnosed, drug-susceptible (88.6%), isoniazid-resistant (7.3%), and multidrug resistant (4.1%) pulmonary TB. Our primary outcome was the proportion of patients with MDR-TB after initial attempt at diagnosis and treatment within a 2-year period. Secondary outcomes were the proportion of i) individuals with detected TB who acquired MDR-TB ii) individuals who died after initial attempt at diagnosis and treatment.

Results

After initial attempt at diagnosis and treatment, LPA combined with appropriate INHR-TB therapy resulted in a lower proportion of prevalent MDR-TB (1.61%; 95% Uncertainty Range (UR: 2.5^th and 97.5^th percentiles generated from 10 000 Monte Carlo simulation trials) 1.61–1.65), when compared to Xpert (1.84%; 95% UR 1.82–1.85) and AFB smear (3.21%; 95% UR 3.19–3.26). LPA also resulted in fewer cases of acquired MDR-TB in those with detected TB (0.35%; 95% UR 0.34–0.35), when compared to Xpert (0.67%; 95% UR 0.65–0.67) and AFB smear (0.68%; 95% UR 0.67–0.69). The majority of acquired MDR-TB arose from the treatment of INHR-TB in all strategies. Xpert-based strategies resulted in a lower proportion of death (2.89%; 95% UR 2.87–2.90) compared to LPA (2.93%; 95% UR 2.91–2.94) and AFB smear (3.21%; 95% UR 3.19–3.23).

Conclusion

Accurate diagnosis and tailored treatment of INHR-TB with LPA led to an almost 50% relative decrease in acquired MDR-TB when compared with an Xpert MTB/RIF strategy. Continued reliance on diagnostic and treatment protocols that ignore INHR-TB will likely result in further generation of MDR-TB.

Socio-economical and Clinico-Radiological Profile of 474 MDR TB Cases of a Rural Medical College

OBJECTIVE

To study the socio economical and clinico radiological profile of 474 diagnosed MDR TB cases who came for the initiation of MDR TB regimen in DRTB center of R.D.Gardi Medical college, Ujjain.

METHODOLOGY

This is a retrospective and prospective observational study for a total period of three years from October 2013 to September 2016. The patients were evaluated clinically, radiologically and were investigated thoroughly according to PMDT guidelines and then were started on MDR TB treatment. The study was conducted at drug resistance tuberculosis center (DR-TB) managed by department of Pulmonary Medicine.

RESULTS

474 cases were included in the study and we found that patients were in the age range of 10-84 years, maximum patients were in age group of 30 to 39 years, and mean age was 38 yrs. Male to female ratio was 2.73 to 1, most of the patients in the study were from rural area i.e. 61.6%. Illiteracy was found in 339 (71.5%) cases and out of these 339, 165 patients (48.6%) were defaulter, 101(29.8%) are cases of relapse, 39(11.5%) were failure, 34(10.02%) of new cases. Maximum numbers of patient were in lower class accounting 63.7% and upper lower class 31.6%, lower middle class only 4.5%. Study also showed mean BMI was 14.9 kg/m2 (range 5.7-25.4 kg/m2), 88.6% of patients were undernourished with BMI less then 18.5kg/m2 .The most common symptoms was cough seen in 96%, followed by fever 67.5%, Dyspnea 52.7%, Anorexia 26.2%, chest pain in 19.8% and least common was haemoptysis seen in 7.6% of patient. Common co-morbidities with MDR-TB found was anemia in 176 out of 474 (i.e.37.1%), 123(25.9%) COPD. Radiological severity showed 219(46.2%) moderate lesion, 139 (29.3%) mild, 107(22.6%) extensive lesion and 9(1.9%) normal, 312(65.8%) of patient are non-cavitory and 162 (34.2%) are cavitory in which 99 (20.9%) were unilateral and 63(13.3%) are bilateral cavitory lesion. Defaulter are most common accounting of 218(46.0%), relapse139 (29.3%) and failure 68 (14.3%), new 48 (10.2%), most of them had taken more than one episode of ATT (72.8%). Most common source of ATT taken by patient is RNTCP it accounts 424 (89.5%) and 46 (9.7%) from private. 181 out of 474 (38.2%) cases delayed the treatment for 1-7 days, 82 out of 474 (17.3%) cases delayed treatment for 8-10 days, 96 out of 474 (20.3%) cases delayed treatment for 11-19 days and 115 out of 474 delayed the treatment for more than 19 days. 95 out of 474 cases i.e. 20.1% cases come from more than 150 km away from their residing area for the initiation of treatment.

CONCLUSION

The epidemiological picture of TB showed that males were predominant in our study however female were more affected in younger age group compared to male. More than 51% of the cases were in productive age group which affects the socioeconomic condition of family and society. More than 2/3 of patients were from lower socioeconomic group with low BMI. Therefore improving nutrition and immunity can play an important role. 2.3% of the cases were HIV reactive and were on ART. Co-morbidities like COPD and Diabetes were seen in our study which were statistically significant and had impact on the treatment outcome of results. Significant delay in initiation of MDR-TB regimen from date of DST was seen in 24.3% cases which is matter of concern. Most of the patients had taken ATT from RNTCP in which Defaulter and relapse was major contributor of MDR-TB suspect in our study and patient taking ATT privately were less. Large number of cases which resides more than 150 kilometers from DRTB center initiated the drug after a gap of more than 19 days from the date of DST.

The never-ending story of the fight against tuberculosis: from Koch's bacillus to global control programs

Tuberculosis (TB) is one of the oldest diseases known to affect humanity, and is still a major public health problem. It is caused by the bacillus Mycobacterium tuberculosis (MT), isolated in 1882 by Robert Koch. Until the 1950s, X rays were used as a cheap method of diagnostic screening together with the tuberculin skin sensitivity test. In the diagnosis and treatment of TB, an important role was also played by surgery. The late Nineteenth century saw the introduction of the tuberculosis sanatorium, which proved to be one of the first useful measures against TB. Subsequently, Albert Calmette and Camille Guérin used a non-virulent MT strain to produce a live attenuated vaccine. In the 1980s and 1990s, the incidence of tuberculosis surged as a major opportunistic infection in people with HIV infection and AIDS; for this reason, a combined strategy based on improving drug treatment, diagnostic instruments and prevention was needed.

Expression of a recombinant, 4'-Phosphopantetheinylated, active M. tuberculosis fatty acid synthase I in E. coli

Background

Fatty acid synthase 1 (FAS I) from Mycobacterium tuberculosis (Mtb) is an essential protein and a promising drug target. FAS I is a multi-functional, multi-domain protein that is organized as a large (1.9 MDa) homohexameric complex. Acyl intermediates produced during fatty acid elongation are attached covalently to an acyl carrier protein (ACP) domain. This domain is activated by the transfer of a 4'-Phosphopantetheine (4'-PP, also termed P-pant) group from CoA to ACP catalyzed by a 4'-PP transferase, termed acyl carrier protein synthase (AcpS).

Methods

In order to obtain an activated FAS I in E. coli, we transformed E. coli with tagged Mtb fas1 and acpS genes encoded by a separate plasmid. We induced the expression of Mtb FAS I following induction of AcpS expression. FAS I was purified by Strep-Tactin affinity chromatography.

Results

Activation of Mtb FAS I was confirmed by the identification of a bound P-pant group on serine at position 1808 by mass spectrometry. The purified FAS I displayed biochemical activity shown by spectrophotometric analysis of NADPH oxidation and by CoA production, using the Ellman reaction. The purified Mtb FAS I forms a hexameric complex shown by negative staining and cryo-EM.

Conclusion

Purified hexameric and active Mtb FAS I is required for binding and drug inhibition studies and for structure-function analysis of this enzyme. This relatively simple and short procedure for Mtb FAS I production should facilitate studies of this enzyme.

The effect of metformin on culture conversion in tuberculosis patients with diabetes mellitus

BACKGROUND/AIMS

Patients with diabetes mellitus (DM) and tuberculosis (TB) have increased morbidity and a high risk of treatment failure or recurrence. It is important to manage both diseases simultaneously. Among anti-diabetic drugs, metformin inhibits intracellular growth of mycobacteria. Therefore, we examined the effects of metformin on TB treatment, especially in patients with DM.

METHODS

This retrospective cohort study included patients with culture-positive pulmonary TB diagnosed between 2011 and 2012. The primary study outcome was sputum culture conversion after 2 months of treatment.

RESULTS

Of 499 patients diagnosed with culture-positive pulmonary TB, 105 (21%) had DM at diagnosis. Among them, 62 (59.5%) were treated with metformin. Baseline characteristics, except for the presence of chronic renal disease, were not significantly different between the metformin and non-metformin groups. Metformin treatment had no significant effect on sputum culture conversion (p = 0.60) and recurrence within 1 year after TB treatment completion (p = 0.39). However, metformin improved the sputum culture conversion rate in patients with cavitary pulmonary TB, who have higher bacterial loads (odds ratio, 10.8; 95% confidence interval, 1.22 to 95.63).

CONCLUSION

Among cavitary pulmonary TB patients with DM, metformin can be an effective adjunctive anti-TB agent to improve sputum culture conversion after 2 months of treatment.

Male gender and duration of anti-tuberculosis treatment are associated with hypocholesterolemia in adult pulmonary tuberculosis patients in Kampala, Uganda

BACKGROUND

Patients with Pulmonary tuberculosis (PTB) and hypocholesterolemia have an altered immune function, delayed sputum conversion at two months and increased mortality. However, the assessment for dyslipidemias is not often done in our setting.

METHODS

A cross-sectional study was conducted among adults at an urban TB clinic in Kampala, Uganda. We included different participants at diagnosis (0), 2, 5, 6 and 8 months of anti-TB treatment. Data was collected from a complete physical examination, a pre-tested structured questionnaire, six-hour fasting lipid profiles and random blood glucose levels.

RESULTS

Of the 323 included participants, 63.5% (205/323) were males and the median age was 30 years, IQR (23-39). The prevalence of hypocholesterolemia was 43.65% (95% CI 38.3-49.2). The participants at diagnosis had the highest hypocholesterolemia prevalence, 57.3%, 95% CI (46.7-67.2); and lowest amongst those completing treatment at 6/8 months, 32.2%, 95% CI (21.6-45.2). Significant factors associated with hypocholesterolemia were: male gender (PR 1.52, 95% CI: 1.13-2.03), and duration of anti-TB treatment (0.88, 95% CI: 0.80-0.98).

CONCLUSION

Hypocholesterolemia is common among patients with PTB. The risk of hypocholesterolemia increases with being male and reduces with increased duration of treatment. There is a need for further research in lipid abnormalities in TB patients.

TB Anywhere is TB Everywhere

To control and prevent outbreaks, public health programs in all countries, regardless of tuberculosis (TB) incidence, must maintain the capacity to perform core control and prevention activities. These include diagnosing and treating cases, contact investigations, and infection prevention and control activities. Congregate settings and healthcare facilities demand special attention, because of the potential for outbreaks associated with infectious cases in these settings. Since almost one-fourth of the world population is thought to be infected with Mycobacterium tuberculosis, enhanced efforts to diagnose and treat latent TB infection are needed to prevent future cases and accelerate progress towards TB elimination.