Model Helps Optimize Who Should Be Screened for Cancer

Detecting cancers early remains an important goal, but current approaches for deciding who should be screened may miss some people with earlier-stage disease.

Researchers from France are developing a model to find biomarkers and clinical risk factors that can better identify individuals at risk for cancer who should be screened.

Using machine learning, the team highlighted more than 30 biomarkers and 2 clinical risk factors among patients with cardiovascular disease who smoked and 13 biomarkers and 8 clinical risk factors among patients with Li-Fraumeni syndrome that could effectively identify those at risk for cancer.

The aim of this research is to optimize and individualize cancer screening and improve cancer prediction, detection, and prevention, according to Marine Fidelle, PharmD, PhD, research pharmacist at the Gustave Roussy Institute, Paris, who presented the findings at the Society for Immunotherapy of Cancer’s (SITC) annual meeting in San Diego.

In the study, Fidelle and colleagues first analyzed data in approximately 500 patients from the PREVALUNG trial, which included current or former smokers with cardiovascular disease who were part of a lung cancer screening program. The team assessed blood and feces samples that had been banked years before cancer detection, as well as low-dose chest CT scans.

To validate the model, the team performed the same assessment of blood and feces samples on more than 100 patients in the LIFSCREEN trial who have Li-Fraumeni syndrome, a rare genetic disorder that increases a person’s risk for cancer.

Using machine learning, the researchers found 33 biomarkers and 2 clinical risk factors among patients with cardiovascular disease that led to a cancer risk estimate with an area under the curve (AUC) of 0.78. In the cohort of patients with Li-Fraumeni syndrome, 13 soluble markers and 8 clinical risk factors predicted cancer risk with an AUC of 0.82.

In the PREVALUNG cohort, 7% of patients were diagnosed with cancer and 3.2% with lung cancer.

Compared with previous studies that delineated screening eligibility based on tobacco screening scores, “what is striking,” Fidelle said, is that about 20%–50% of patients diagnosed with cancer in the PREVALUNG cohort would not have been included in those screening programs.

Christopher J. Manley, MD, who was not involved in the research, noted that successful cancer screening requires broad implementation, low cost, and ease of use.

“This study, combining biomarkers with clinical risk factors, may prove to be a safe and effective way to identify patients at risk for developing malignancy,” said Manley, director, of interventional pulmonology at Fox Chase Cancer Center, Philadelphia. “As the authors report, this may help to optimize screening. If healthcare providers can identify a group at increased risk, then a more concentrated effort can be made for cancer prevention and early detection.”

The study received funding from the European Union’s Horizon Europe research and innovation program. Fidelle received funding from the Seerave Foundation. Manley serves as a consultant and educational speaker for Johnson and Johnson and has received research funding from Mauna Kea, Aquyre Biosciences, and Johnson & Johnson.

Roxanne Nelson is a registered nurse and an award-winning medical writer who has written for many major news outlets and is a regular contributor to Medscape.

Source: Read Full Article