An Update on the Use of Exhaled Breath Analysis for the Early Detection of Lung Cancer

Machine learning enabled detection of COVID-19 pneumonia using exhaled breath analysis: a proof-of-concept study

Detection of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) relies on real-time-reverse-transcriptase polymerase chain reaction (RT-PCR) on nasopharyngeal swabs. The false-negative rate of RT-PCR can be high when viral burden and infection is localized distally in the lower airways and lung parenchyma. An alternate safe, simple and accessible method for sampling the lower airways is needed to aid in the early and rapid diagnosis of COVID-19 pneumonia. In a prospective unblinded observational study, patients admitted with a positive RT-PCR and symptoms of SARS-CoV-2 infection were enrolled from three hospitals in Ontario, Canada. Healthy individuals or hospitalized patients with negative RT-PCR and without respiratory symptoms were enrolled into the control group. Breath samples were collected and analyzed by laser absorption spectroscopy (LAS) for volatile organic compounds (VOCs) and classified by machine learning (ML) approaches to identify unique LAS-spectra patterns (breathprints) for SARS-CoV-2. Of the 135 patients enrolled, 115 patients provided analyzable breath samples. Using LAS-breathprints to train ML classifier models resulted in an accuracy of 72.2%-81.7% in differentiating between SARS-CoV2 positive and negative groups. The performance was consistent across subgroups of different age, sex, body mass index, SARS-CoV-2 variants, time of disease onset and oxygen requirement. The overall performance was higher than compared to VOC-trained classifier model, which had an accuracy of 63%-74.7%. This study demonstrates that a ML-based breathprint model using LAS analysis of exhaled breath may be a valuable non-invasive method for studying the lower airways and detecting SARS-CoV-2 and other respiratory pathogens. The technology and the ML approach can be easily deployed in any setting with minimal training. This will greatly improve access and scalability to meet surge capacity; allow early and rapid detection to inform therapy; and offers great versatility in developing new classifier models quickly for future outbreaks.

Health State Utilities Associated with False-Positive Cancer Screening Results

INTRODUCTION

Early cancer detection can significantly improve patient outcomes and reduce mortality rates. Novel cancer screening approaches, including multi-cancer early detection tests, have been developed. Cost-utility analyses will be needed to examine their value, and these models require health state utilities. The purpose of this study was to estimate the disutility (i.e., decrease in health state utility) associated with false-positive cancer screening results.

METHODS

In composite time trade-off interviews using a 1-year time horizon, UK general population participants valued 10 health state vignettes describing cancer screening with true-negative or false-positive results. Each false-positive vignette described a common diagnostic pathway following a false-positive result suggesting lung, colorectal, breast, or pancreatic cancer. Every pathway ended with a negative result (no cancer detected). The disutility of each false positive was calculated as the difference between the true-negative and each false-positive health state, and because of the 1-year time horizon, each disutility can be interpreted as a quality-adjusted life-year decrement associated with each type of false-positive experience.

RESULTS

A total of 203 participants completed interviews (49.8% male; mean age = 42.0 years). The mean (SD) utility for the health state describing a true-negative result was 0.958 (0.065). Utilities for false-positive health states ranged from 0.847 (0.145) to 0.932 (0.059). Disutilities for false positives ranged from - 0.031 to - 0.111 (- 0.041 to - 0.111 for lung cancer; - 0.079 for colorectal cancer; - 0.031 to - 0.067 for breast cancer; - 0.048 to - 0.088 for pancreatic cancer).

CONCLUSION

All false-positive results were associated with a disutility. Greater disutility was associated with more invasive follow-up diagnostic procedures, longer duration of uncertainty regarding the eventual diagnosis, and perceived severity of the suspected cancer type. Utility values estimated in this study would be useful for economic modeling examining the value of cancer screening procedures.

Exploring Circulating Tumor DNA (CtDNA) and Its Role in Early Detection of Cancer: A Systematic Review

There is a significant increase in the need for an efficient screening method that might identify cancer at an early stage and could improve patients' long-term survival due to the continued rise in cancer incidence and associated mortality. One such effort involved using circulating tumor DNA (ctDNA) as a rescue agent for a non-invasive blood test that may identify many tumors. A tumor marker called ctDNA is created by cells with the same DNA alterations. Due to its shorter half-life, it may be useful for both early cancer detection and real-time monitoring of tumor development, therapeutic response, and tumor outcomes. We obtained 156 papers from PUBMED using the MeSH approach in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) criteria and ten articles from additional online resources. After removing articles with irrelevant titles and screening the abstract and full text of the articles that contained information unrelated to or not specific to the title query using inclusion and exclusion criteria, 18 out of 166 articles were chosen for the quality check. Fourteen medium to high-quality papers were chosen out of the 18 publications to be included in the study design. The reviewed literature showed no significant utility of ctDNA in detecting early-stage tumors of size less than 1 cm diameter. Still, the ideal screening test would require the assay to detect a size <5 mm tumor, which is nearly impossible with the current data. The sensitivity and specificity of the assay ranged from 69% to 98% and 99%, respectively. Furthermore, CancerSEEK achieves tumor origin localization in 83% of cases, while targeted error correction sequencing (TEC-Seq) assays demonstrate a cancer detection rate ranging from 59% to 71%, depending on the type of cancer. However, it could be of great value as a prognostic indicator, and the levels are associated with progression-free survival (PFS) and overall survival (OS) rates, wherein the positive detection of ctDNA is associated with worse OS compared to the tumors detected through standard procedures, with an odds ratio (OS) of 4.83. We conclude that ctDNA could be better applied in cancer patients for prognosis, disease progression monitoring, and treatment outcomes compared to its use in early cancer detection. Due to its specific feature of recognizing the tumor-related mutations, it could be implemented as a supplemental tool to assess the nature of the tumor, grade, and size of the tumor and for predicting the outcomes by pre-operative and post-operative evaluation of the tumor marker, ctDNA, and thereby estimating PFS and OS depending on the level of marker present. A vast amount of research is required in early detection to determine the sensitivity, specificity, false positive rates, and false negative rates in evaluating its true potential as a screening tool. Even if the test could detect the mutations, an extensive workup for the search of tumor is required as the assay could only detect but cannot localize the disease. Establishing the clinical validity and utility of ctDNA is imperative for its implementation in future clinical practice.

Selectivity of Exhaled Breath Biomarkers of Lung Cancer in Relation to Cancer of Other Localizations

Lung cancer is a leading cause of death worldwide, mostly due to diagnostics in the advanced stage. Therefore, the development of a quick, simple, and non-invasive diagnostic tool to identify cancer is essential. However, the creation of a reliable diagnostic tool is possible only in case of selectivity to other diseases, particularly, cancer of other localizations. This paper is devoted to the study of the variability of exhaled breath samples among patients with lung cancer and cancer of other localizations, such as esophageal, breast, colorectal, kidney, stomach, prostate, cervix, and skin. For this, gas chromatography-mass spectrometry (GC-MS) was used. Two classification models were built. The first model separated patients with lung cancer and cancer of other localizations. The second model classified patients with lung, esophageal, breast, colorectal, and kidney cancer. Mann-Whitney U tests and Kruskal-Wallis H tests were applied to identify differences in investigated groups. Discriminant analysis (DA), gradient-boosted decision trees (GBDT), and artificial neural networks (ANN) were applied to create the models. In the case of classifying lung cancer and cancer of other localizations, average sensitivity and specificity were 68% and 69%, respectively. However, the accuracy of classifying groups of patients with lung, esophageal, breast, colorectal, and kidney cancer was poor.

Sensitivity and specificity of in vivo COVID-19 screening by detection dogs: Results of the C19-Screendog multicenter study

Trained dogs can recognize the volatile organic compounds contained in biological samples of patients with COVID-19 infection. We assessed the sensitivity and specificity of in vivo SARS-CoV-2 screening by trained dogs. We recruited five dog-handler dyads. In the operant conditioning phase, the dogs were taught to distinguish between positive and negative sweat samples collected from volunteers' underarms in polymeric tubes. The conditioning was validated by tests involving 16 positive and 48 negative samples held or worn in such a way that the samples were invisible to the dog and handler. In the screening phase the dogs were led by their handlers to a drive-through facility for in vivo screening of volunteers who had just received a nasopharyngeal swab from nursing staff. Each volunteer who had already swabbed was subsequently tested by two dogs, whose responses were recorded as positive, negative, or inconclusive. The dogs' behavior was constantly monitored for attentiveness and wellbeing. All the dogs passed the conditioning phase, their responses showing a sensitivity of 83-100% and a specificity of 94-100%. The in vivo screening phase involved 1251 subjects, of whom 205 had a COVID-19 positive swab and two dogs per each subject to be screened. Screening sensitivity and specificity were respectively 91.6-97.6% and 96.3-100% when only one dog was involved, whereas combined screening by two dogs provided a higher sensitivity. Dog wellbeing was also analyzed: monitoring of stress and fatigue suggested that the screening activity did not adversely impact the dogs' wellbeing. This work, by screening a large number of subjects, strengthen recent findings that trained dogs can discriminate between COVID-19 infected and healthy human subjects and introduce two novel research aspects: i) assessement of signs of fatigue and stress in dogs during training and testing, and ii) combining screening by two dogs to improve detection sensitivity and specificity. Using some precautions to reduce the risk of infection and spillover, in vivo COVID-19 screening by a dog-handler dyad can be suitable to quickly screen large numbers of people: it is rapid, non-invasive and economical, since it does not involve actual sampling, lab resources or waste management, and is suitable to screen large numbers of people.

A nanostructured Al-doped ZnO as an ultra-sensitive room-temperature ammonia gas sensor

Novel chemi-resistive gas sensors with strong detection capabilities operating at room temperature are desirable owing to their extended cycle life, high stability, and low power consumption. The current study focuses on detecting NH3 at room temperature using lower gas concentrations. The co-precipitation technique was employed to produce pure and Al-doped ZnO nanoparticles, which were calcined at 300 °C for three hours. The effect of aluminium (Al) doping on the structural, morphological, optical, and gas-sensing abilities was investigated and reported. The presence of aluminium was confirmed by XRD, EDX, and FTIR spectroscopy. Additionally, to assess the various characteristics of Al-doped ZnO nanoparticles, scanning electron microscopy (SEM), ultraviolet-visible diffuse reflectance spectroscopy (UV-DRS), atomic force microscopy (AFM), and Brunauer-Emmett-Teller (BET) techniques were used. The crystallite size increased from 14.82 to 17.49 nm in the XRD analysis; the SEM pictures showed a flower-like morphology; and the energy gap decreased from 3.240 to 3.210 eV when Al doping was raised from 1 wt% to 4 wt%. AFM studies revealed topographical information with significant roughness in the range of 230-43 nm. BET analysis showed a mesoporous nature with surface areas varying from 25.274 to 14.755 m2/g and pore diameters ranging from 8.34 to 7.00 nm. The sensing capacities of pure and Al-doped ZnO nanoparticles towards methanol (CH3OH), toluene (C7H8), ethanol (C2H5OH), and ammonia (NH3) were investigated at room temperature. The one-wt% Al-doped ZnO sensor demonstrated an ultrafast response and recovery times at one ppm compared to other AZO-based sensors towards NH3.

Application of Artificial Intelligence in Lung Cancer

Simple Summary

Lung cancer is the leading cause of malignancy-related mortality worldwide. AI has the potential to help to treat lung cancer from detection, diagnosis and decision making to prognosis prediction. AI could reduce the labor work of LDCT, CXR, and pathology slides reading. AI as a second reader in LDCT and CXR reading reduces the effort of radiologists and increases the accuracy of nodule detection. Introducing AI to WSI in digital pathology increases the Kappa value of the pathologist and help to predict molecular phenotypes with radiomics and H&E staining. By extracting radiomics from image data and WSI from the histopathology field, clinicians could use AI to predict tumor properties such as gene mutation and PD-L1 expression. Furthermore, AI could help clinicians in decision-making by predicting treatment response, side effects, and prognosis prediction in medical treatment, surgery, and radiotherapy. Integrating AI in the future clinical workflow would be promising.

Abstract

Lung cancer is the leading cause of malignancy-related mortality worldwide due to its heterogeneous features and diagnosis at a late stage. Artificial intelligence (AI) is good at handling a large volume of computational and repeated labor work and is suitable for assisting doctors in analyzing image-dominant diseases like lung cancer. Scientists have shown long-standing efforts to apply AI in lung cancer screening via CXR and chest CT since the 1960s. Several grand challenges were held to find the best AI model. Currently, the FDA have approved several AI programs in CXR and chest CT reading, which enables AI systems to take part in lung cancer detection. Following the success of AI application in the radiology field, AI was applied to digitalized whole slide imaging (WSI) annotation. Integrating with more information, like demographics and clinical data, the AI systems could play a role in decision-making by classifying EGFR mutations and PD-L1 expression. AI systems also help clinicians to estimate the patient’s prognosis by predicting drug response, the tumor recurrence rate after surgery, radiotherapy response, and side effects. Though there are still some obstacles, deploying AI systems in the clinical workflow is vital for the foreseeable future.