Analysis of Breath Volatile Organic Compounds Using Mass Spectrometry

This study aims to determine whether metabolic changes associated with cancer produce distinct alterations in exhaled breath compared with those of healthy individuals. Breath samples will be analyzed using machine learning techniques to identify volatile organic compound (VOC) patterns and develop diagnostic algorithms capable of detecting multiple types of cancer. The long-term goal is to establish a noninvasive, breath-based screening tool that can facilitate the early detection of various cancers.

Additionally, patients and healthy participants who consent to this study may opt in to be contacted in the future to provide additional breath samples.

• Proton Transfer Reaction Mass Spectrometry Analysis (Device), also known as PTR-MS: This is a noninvasive intervention. Participants will be asked to provide a breath sample using a disposable mouthpiece equipped with a saliva/moisture trap and a non-rebreathing valve. Breath samples will be collected through normal, steady exhalation. The entire breath collection process is expected to take no more than 30 minutes to complete.

Primary outcomes

• VOC Signature Collection. [2 Years]: The successful collection of breath samples from 1000 cancer patients and 1000 healthy volunteers.
• Assess the sensitivity of Machine Learning (ML) Algorithm In The Test Dataset. [1 Years]: Using the training dataset, qualitative output generated by the PTR-MS instrument will be analyzed using machine learning methods to identify volatile organic compound (VOC) patterns associated with different cancer types, including pancreatic, esophageal, hepatocellular carcinoma, lung, and ovarian cancers. The trained machine learning model will be tested using the dataset.

Secondary outcomes

• Assess The Specificity and Accuracy of ML Analysis In The Test Dataset. [1 year]: To determine the specificity, and overall diagnostic accuracy of the machine learning (ML) algorithm for detecting pancreatic, esophageal, hepatocellular carcinoma, lung, and ovarian cancers within the test dataset.

Sex: All, ages: 18 Years to any, accepts healthy volunteers

Inclusion Criteria:

* Age ≥ 18 at the time of consent. Male and female patients to be tested.
* Capable of understanding written and/or spoken English language.
* Able to provide informed consent.
* Cancer of any type.
* Newly diagnosed cancer and untreated or established diagnosis of cancer. For established cancer patients, no active anti-cancer treatment for more than one month (reasons for no treatment are such as relapse, progression of cancer, refractory, or intolerance to treatment etc.)

Exclusion Criteria:

* Under the age of 18.
* Anticipated inability to complete breath sampling procedure.
* Unable to provide informed consent.
* Pregnant women
* Active respiratory infection symptoms
* Recent use of antibiotics
* Difficulty in performing coached exhalation
* Individuals who are unable to follow the instructions
* Cancer patients who are on active treatment for cancer or have received cancer treatment within one month

• Nirmal Choradia, MD, Principal investigator, University of Oklahoma - Stephenson Cancer Center