Oral Microbiome Diagnostics of Lung Cancer

Lung Cancer Clinical Trial

Official title:

Developing an Artificial Intelligence-based Diagnostic Method Based on Oral Microbiome for Non-invasive Diagnosis of Lung Cancer

Clinical Trial Details — Status: Recruiting

Administrative data

• NCT number: NCT06270992
• Other study ID #: 123R030
• Status: Recruiting
• Start date: November 15, 2023
• Est. completion date: May 15, 2026

Study information

• Verified date: February 2024
• Source: TC Erciyes University
• Contact: Aycan Gundogdu, PhD
• Phone: +90 352 207 6666
• Email: agundogdu[@]erciyes.edu.tr
• Is FDA regulated: No
• Study type: Observational [Patient Registry]

Clinical Trial Summary

The study aims to develop a deep learning-based diagnostic method for lung cancer using the oral microbiome. This innovative approach involves establishing an observational cohort of 576 individuals, including lung cancer patients, non-cancerous benign lung disease patients, and healthy controls, to collect tongue swab samples for 16S rRNA sequencing. Additionally, an international cohort of approximately 1700 individuals will be formed using in silico data. The project will utilize deep learning methods to analyze all data integratively and develop an AI diagnostic algorithm capable of distinguishing lung cancer patients from others. The diagnostic method's performance will be tested in a pilot clinical trial with 96 individuals using a PRoBE design. Led by experts in chest surgery, molecular microbiology, and bioinformatics, the project spans over 30 months and aims to create a non-invasive, easily accessible lung cancer screening method that could lead to significant diagnostic advancements and potential spin-off companies in the field of liquid biopsy/molecular diagnosis.

Description:

Cancer is a global health issue that is on an increasing trend in terms of incidence and mortality rates, hindering the increase in life expectancy. According to the World Health Organization, lung, colorectal, and liver cancers are among the most common causes of cancer-related deaths. In Turkey, the incidence and mortality rates of lung cancer are higher than the world average. are among the risk factors that may increase the risk of lung cancer. In addition to risk factors like family history, smoking, different studies have shown that dysbiotic oral microbiome may contribute to the risk of lung cancer.

The oral microbiome is the second most diverse microenvironment in our body and has been associated with many diseases, including lung cancer. Studies to date on lung cancer-oral microbiome have generally involved designs aimed at resolving cause-and-effect relationships through statistical differences and/or mechanisms involving microbiome units.

However, there is no literature on any study aimed at developing a deep learning-based diagnostic method that focuses on the oral microbiome.Therefore, the proposed study aims to develop a microbiome based deep learning diagnostic method for lung cancer diagnosis. To this end, an observation cohort will be established consisting of 192 lung cancer patients, 192 non-cancerous benign lung disease patients, and 192 healthy controls. Tongue swab samples belonging to the cohort will be collected, and 16S rRNA sequencing will be performed. At the same time, an international observation cohort of approximately 1700 individuals will be created using in silico data. All data will be analyzed integratively, and an artificial intelligence diagnostic algorithm that can differentiate lung cancer patients from other lung diseases and healthy individuals will be developed using deep learning methods. In the final stage, the performance of the diagnostic method developed for a pilot clinical trial cohort of 96 individuals will be tested using a PRoBE (prospective specimen collection before outcome ascertainment and retrospective blinded evaluation) design.

The original aspects of the project are the proposal of a novel design in the literature, the creation of an experimental design/clinical trial and the presentation of a potential solution proposal that may have a high impact on an important diagnostic problem.

If the project is successfully completed, an artificial intelligence-based method that can potentially diagnose lung cancer through non-invasive oral microbiome samples will be developed. In addition to its patentability, if the method is further developed (in the medium to long term) into a product, it will enable lung cancer screening to be easily performed even in primary healthcare institutions with a simple oral swab sample.

Recruitment information / eligibility

• Status: Recruiting
• Enrollment: 676
• Est. completion date: May 15, 2026
• Est. primary completion date: February 15, 2025
• Accepts healthy volunteers: Accepts Healthy Volunteers
• Gender: All
• Age group: 18 Years to 65 Years

Eligibility

Inclusion Criteria:

• To be between the ages of 18 and 65,

• Not to have a diagnosed lung disease or suspicion thereof,

• Not to have complaints related to the lungs and/or respiratory tract,

• Not to have alcohol or severe substance dependency,

• Not having a hospitalization history in the last year,

• Not having used antibiotics in the last six months,

• Not having used products manufactured to support the oral microbiome, such as probiotics (lozenges, sublingual drops) for at least the last six months,

• Not being pregnant or breastfeeding,

• Not having undergone dental procedures such as root canal treatment, implants, prostheses, tooth extraction, fillings in the last 6 months

• Not having dominant immune-origin lesions (such as aphthous ulcers, erythema multiforme, pemphigus), viral-origin lesions (such as herpes, Koplik spots, herpangina), dominant bacterial infections like tonsillitis, and/or thermal or chemical mucosal traumas in the mouth.

Exclusion Criteria:

• Not to satisfy inclusion or declining to participate even though all the inclusion criteria are satisfied.

Related Conditions & MeSH terms

• Lung Cancer
• Lung Neoplasms

Intervention

Diagnostic Test:
• NCCN (National Comprehensive Cancer Network) diagnosis
For diagnostic evaluation, the necessary procedures from the standard protocols consisting of anamnesis, physical examination, laboratory tests, radiological imaging methods, and tissue biopsy will be followed. Computerized Tomography (CT) and Positron Emission Tomography-Computed Tomography (PET-CT) will be used as imaging methods, while fiberoptic bronchoscopy and video-assisted mediastinoscopy will be applied for tissue diagnosis and staging.

Locations

Country	Name	City	State
Turkey	Erciyes University Hospital	Kayseri

Sponsors (2)

Lead Sponsor	Collaborator
TC Erciyes University	THE SCIENTIFIC AND TECHNOLOGICAL RESEARCH COUNCIL OF TÜRKIYE

Country where clinical trial is conducted

Turkey, 

Outcome

Type	Measure	Description	Time frame	Safety issue
Primary	Diagnostics accuracy assessment	Diagnostic technology under investigation will be evaluated using sensitivity, specificity, area under receiver operating curve. Cross validation will be used for testing and NCCN diagnosis and patient follow-ups will be considered as the ground truth.	30 months