DiagnosticsDiagnosis of active tuberculosis by e-nose analysis of exhaled air
Introduction
It is estimated that one third of the world's population is infected with Mycobacterium tuberculosis, often leading to active tuberculosis (TB).1 In 2010 there was an estimated incident case count of 8.8 million active TB infections, resulting in 1.5 million deaths. The primary detection technique is the 125-year old Ziehl-Neelsen (ZN) staining2 combined with microscopy. The major drawback of microscopy is that it only allows for detection of pulmonary disease cases in an advanced stage, meaning that often the disease has already been transmitted to close contacts. Chest X-ray and microbiological culture have been added to the diagnostic arsenal in developing countries while other, more advanced techniques such as nucleic acid amplification, serology, and cytokine release assays3 are becoming available in the developed countries while research is still ongoing. In 2010 the World Health Organization (WHO) endorsed the use of the Cepheid Xpert MTB/RIF system for use in endemic areas. This polymerase chain reaction system, however, relies on a constant power supply making it not ideally suited for portable operation.
Using exhaled air as potential diagnostic indicator is an emerging activity.4, 5, 6, 7, 8 In previous studies gas chromatography combined with mass spectrometry (GC/MS) was used to identify volatile biomarkers unique to M. tuberculosis.9, 10, 11, 12 From these studies it appeared that volatile biomarkers may be used to discriminate TB-patients from healthy controls. However, GC/MS is not viable as a diagnostic tool due to the complex settings of the equipment and operation skills needed and using animals7, 8 introduces a whole other set of challenges.
A viable diagnostic tool which uses volatile biomarkers to differentiate between people with and without TB should be based on an easy-to-use method. Electronic noses have already been used for different medical purposes13, 14, 15 including the diagnostics of asthma,4 chronic obstructive pulmonary disease,4, 5 urinary tract infection,16 wound infection17 and even cancer.18, 19 In the past, we used an electronic nose for the laboratory-based identification of bacterial pathogens.20
In this study, we used three independently produced electronic nose units and determined the device-independent diagnostic accuracy in detecting TB. We started with a proof of principle (PoP) study to determine the possibility to discriminate between severely TB-infected and healthy people. Subsequently, we conducted a validation study (VS) to substantiate the PoP results and extend the variance between the examined groups.
Section snippets
Experimental equipment
The DiagNose (C-it BV, Zutphen, The Netherlands) is an electronic nose device incorporating 12 metal-oxide sensors, being 4 different sensor types (AS-MLC; AS-MLN; AS-MLK; AS-MLV, Applies Sensors Gmbh) in triplicate.
The Diagnose is equipped with a pump, where the inlet is controlled by a solenoid switching between 2 different inlets, to create an active airflow across the sensors. One inlet is connected to an active carbon filter to provide a baseline free from environmental influence while the
Results
The studies were conducted in two parts during a time span of 14 months in Dhaka, Bangladesh. For the PoP study all possible permutations of the 4 sensor types where evaluated and a combination of 2 sensors yielded the optimal result. Within the VS study these findings were first validated before conducting the full data-analysis based on the measurements of only two of the 4 sensor types. All possible sensor permutations for the triplicate of these 2 types result in 9 unique combinations per
Discussion and conclusion
In the present study we tested the diagnostic accuracy of DiagNose breath analysis for the detection of TB in Bangladesh. In a PoP study we established that the DiagNose was able to differentiate almost perfectly (sensitivity 95·9% and specificity 98·5%) between healthy controls and TB patients. However, the control group was without health complaints while the group of patients was obviously sick, possibly making the differentiation one of healthy vs. sick individuals. Besides this, there was
Acknowledgments
This study was partially funded by BD.
References (23)
- et al.
The history of the Ziehl-Neelsen stain
Tubercle
(1970) Ability of Cricetomys rats to detect Mycobacterium tuberculosis and discriminate it from other microorganisms
Tuberculosis (Edinb)
(2012 Mar)- et al.
Honeybees Apis mellifera can detect the scent of Mycobacterium tuberculosis
Tuberculosis (Edinb)
(2011 Jul) - et al.
The scent of Mycobacterium tuberculosis
Tuberculosis (Edinb)
(2008) - et al.
The scent of Mycobacterium tuberculosis – part II breath
Tuberculosis (Edinb)
(2009) Volatile biomarkers of pulmonary tuberculosis in the breath
Tuberculosis (Edinb)
(2007)Detection of Mycobacterium tuberculosis (TB) in vitro and in situ using an electronic nose in combination with a neural network system
Biosens Bioelectron
(2004)Use of an electronic nose system for diagnoses of urinary tract infections
Biosens Bioelectron
(2002)Featuring of bacterial contamination of wounds by dynamic response of SnO2 gas sensor array
Sens Actuators B Chem
(2006)An electronic nose in the discrimination of patients with non-small cell lung cancer and COPD
Lung Cancer
(2009)
Cited by (107)
Current perspectives of e-noses
2023, Nanotechnology-Based E-Noses: Fundamentals and Emerging ApplicationsEmerging applications of nanotechnology for e-nose
2023, Nanotechnology-Based E-Noses: Fundamentals and Emerging ApplicationsE-noses: an introduction
2023, Nanotechnology-Based E-Noses: Fundamentals and Emerging ApplicationsEndometrial cytokine expression from clinically suspected genital tuberculosis patients at tertiary care hospitals in Dhaka
2022, Journal of Clinical Tuberculosis and Other Mycobacterial DiseasesClassification techniques of electronic nose: a review
2024, International Journal of Bio-Inspired Computation