Blood test helps predict risk for lung cancer mortality

August 25, 2023

3 min read

Source/Disclosures

Disclosures:
Hanash reports honoraria or research funding from Abbott Laboratories, Bristol Myers Squibb, Cosmos Wisdom and Dynex. Please see the study for all other authors’ relevant financial disclosures.

We were unable to process your request. Please try again later. If you continue to have this issue please contact [email protected].

Key takeaways:

  • A model showed improved ability to predict lung cancer-specific mortality compared with USPSTF criteria.
  • Efforts are underway to make the blood-based assay commercially available.

Researchers have developed a method to predict which individuals are at higher risk for lung cancer-related mortality.

The approach — which combines results from a blood-based biomarker panel with a personalized risk model — may be particularly impactful for individuals who do not meet U.S. Preventive Services Task Force criteria for annual lung cancer screening with low-dose CT scans, according to the investigators.

“Most individuals who are at high risk for lung cancer are not being screened,” Samir M. Hanash, MD, PhD, professor in the department of clinical cancer prevention at The University of Texas MD Anderson Cancer Center, told Healio. “There is a continuum for the treatment of lung cancer, and that continuum begins with identifying early those who are most at risk.”

Background

Researchers from MD Anderson and Fred Hutchinson Cancer Center previously collaborated to develop a blood-based biomarker panel to assess lung cancer risk.

“We searched high and low to find the right combination of proteins to make our panel the most informative, but also keep it as simple as possible,” Hanash said.

The investigators validated the panel — comprising the precursor form of surfactant protein B, cancer antigen 125, carcinoembryonic antigen and cytokeratin-19 fragment — via a previous blinded study using prediagnostic blood samples obtained during the randomized PLCO Cancer Screening Trial.

That study included heavy smokers at higher risk for lung cancer. The researchers found the combination of the blood-based biomarker panel and personalized risk model more accurately identified individuals at higher risk for dying of lung cancer than current USPSTF criteria, which recommends yearly low-dose CT scan for adults aged 50 to 80 who have a smoking history of at least 20 pack-years and currently smoke or have quit within the past 15 years.

Those criteria often miss individuals at high risk for developing lung cancer because they do not meet the threshold, Hanash said.

“The question became whether this panel plus risk model could identify light smokers who could benefit from screening and whether that would be effective at reducing mortality due to lung cancer,” he said.

Methodology

Hanash and colleagues conducted a retrospective cohort study to evaluate the ability of a blood-based biomarker panel combined with a personalized risk model assessment to identify individuals at increased risk for lung cancer-related mortality.

Researchers used blood samples from a biorepository created to store samples from participants in the U.S.-based PLCO Cancer Screening Trial. The samples included 552 lung cancer cases and 2,193 non-lung cancer cases.

Seventy percent of those with lung cancer in the cohort died of the disease.

Researchers used a previously validated logistic regression model that combined a blood-based, four-marker protein panel with the PLCO risk model. They then used HRs to evaluate the relationship between risk scores generated by the combination model compared with the incidence of lung cancer death in the study cohort.

The analysis used HRs based on combination model risk scores at predefined 1% and 1.7% 6-year risk thresholds to correspond with current (2021) and prior (2013) USPSTF screening criteria.

Key findings

The analysis showed an estimated area under the receiver operation characteristics curve of 0.88 (95% CI, 0.86-0.9) for the blood panel plus model’s risk prediction of lung cancer death.

Patients with combined four-marker panel plus model risk scores more than 1% above the 6-year threshold had a statistically significant higher incidence of lung cancer death (modified chi-squared, 166.27; P < .0001).

The combined model demonstrated superior ability to predict lung cancer-specific mortality among heavy smokers than 2021 USPSTF criteria, including improved sensitivity, specificity and positive predictive values.

Clinical implications

The key takeaway, according to Hanash, is that having a simple blood test can help identify those who may benefit from additional screening but would normally not meet current guidelines, including light smokers and nonsmokers.

“Most people don’t meet the current diagnostic screening criteria — or if they do, many won’t ever develop lung cancer,” he told Healio.

Hanash — who is listed as one of the inventors on a patent application for the blood test — said he hopes the blood-based assay will be made available for clinicians to use for risk assessment in the future.

“This has now been taken out of the research domain and into the business development domain, and MD Anderson is working very intensely on that,” he said.

For more information:

Samir M. Hanash, MD, PhD, can be reached at [email protected].

Leave a Reply

Your email address will not be published. Required fields are marked *