Immunotherapy with agents stimulating the immune system to act against cancer are now a new standard of care in various cancers as lung cancer and melanoma, but also bladder cancer, kidney cancer and head & neck cancer. However, even though a subset of patients derives long-term benefit from these agents, depending of cancer type still at least half of patients do not respond to these new drugs. Our understanding of possible factors predicting whether a patient might actually benefit from immunotherapy is poor. Volatile organic compounds (VOCs) are gases exhaled with a person's breath, which are released into the lung from blood and bacteria and therefore can give information about infections as well as inflammation and possibly cancer cells in a person's body. Breath analysis of these VOCs with special devices called electronic noses (eNose) generate a specific electric signals patterns called breathprints. There is early evidence that specific breathprints can actually help to select patients who will be likely to benefit from immunotherapy.

This study is being undertaken in an effort to evaluate breathprint analysis as a potential predicting factor for benefit from immunotherapy, so that treatment selection can further be improved.

This study is designed to help us identify the role of breathprint analysis to better select patients for immunotherapy.

Immune checkpoint blockade with anti-PD-1 and anti-PD-L1 antibodies has become a new standard of care in several cancer types as NSCLC and melanoma. However, in biomarker-unselected patient populations, overall response rate (ORR) depending on type of cancer and whether single or combination treatment is chosen remains still only 20%-60%. Though overall well tolerated approximately 5-10% of patients treated with PD1/PD-L1 targeting agents will experience grade 3 or 4 toxicities, including potentially life-threatening auto-immune toxicities such as colitis, hepatitis, and pneumonitis. Therefore, due to high costs of treatment and its possible complications, improved selection of patients is a crucial goal and an easily available non-invasive, point-of-care test for better patient selection is very much needed.

A promising approach in this regard is the analysis of volatile organic compounds (VOCs) in breath. Breath analysis for the detection of VOCs is increasingly investigated for its utility in diagnosis and management of cancer. Electronic noses (eNoses) are promising as cheap and clinically-practical devices that are designed to detect patterns of VOCs. Recently published prospective observational data showed very promising discriminant function for breathprint analysis for non-response to immunotherapy in NSCLC patients.

The principle goal of this study is to validate a prior study that found that breathomics-based classifiers predicted 12-week early progression vs non-progression in advanced NSCLC patients treated with nivolumab or pembrolizumab. Secondarily, we will expand assessment of breathomic-based classifiers to include other cohorts of advanced tumors treated with ICI, and also consider using response instead of non-progression as an endpoint. Exploratory goals include refinement of the breathomics classifier using alternative machine-learning techniques, and correlate with other biomarkers of immunotherapy outcomes.

INCLUSION CRITERIA

• Patients 18 years of age or older

• Histologically confirmed advanced/metastatic non-small cell lung cancer, melanoma or solid tumor such as urothelial, kidney or head and neck cancer and planned treatment with

  • NSCLC validation cohort: Pembrolizumab or Nivolumab
  • NSCLC Cohort 1: Pembrolizumab-chemotherapy combination therapy 1L
  • Melanoma Cohort 2: Nivolumab/ipilimumab combination treatment 1L, Pembrolizumab or nivolumab monotherapy treatment 1L , Ipilimumab
  • Solid tumors Cohort 3: Any ICI-treatment, any line
  • NSCLC Cohort 4: Chemotherapy-only (either platinum-based combination treatment or docetaxel monotherapy)

• At least one measurable lesion as defined by RECIST 1.1. A lesion at a previously irradiated site may only be counted as a target lesion if there is clear sign of progression since the irradiation.

• Able to provide informed consent.

EXCLUSION CRITERIA

• Patients who are unable to perform the breathing manoeuvres needed for eNose-analysis of exhaled air.
• Patients who are unable to independently consent to participation in the trial.
• Patients with severe, acute, or chronic medical conditions (including uncontrolled diabetes mellitus) or psychiatric conditions or laboratory abnormalities that in the opinion of the Investigator or their physician may cause undue harm or inconvenience to the patient, or that may interfere with the interpretation of study results.

Subjects included in this study are patients over the age of 18 years with a histologically-confirmed diagnosis of non-small cell lung cancer, melanoma or other solid tumor for which ICI-treatment is a standard of care. Patients must have advanced/metastatic disease and planning to receive ICI treatment (categories defined above) or chemotherapy treatment (cohort 4) at the Princess Margaret Cancer Centre/University Health Network. Study subjects will be consented specifically for that study.