Lung cancer screening model identifies more at-risk African Americans

BY WALTER ALEXANDER

credit: THEPALMER/Getty Images

A new model for lung cancer screening may ensure that African Americans at risk for lung cancer are no longer undercounted, according to researchers.

Mary M. Pasquinelli, DNP, of the division of pulmonary, critical care, sleep and allergy at the University of Illinois at Chicago, and a team of investigators found that applying United States Preventive Services Task Force and the Center for Medicare & Medicaid Services criteria at a 1.70% risk threshold identified only 50.3% of African American lung cancer cases. However, 71.3% of cases were identified with the Prostate, Lung, Colorectal, and Ovarian Screening Trial–derived PLCOm2012 model.

The criteria for low-dose CT lung cancer screening in current USPSTF and CMS guidelines are age 55-80 years or 55-77 years, respectively; 30 or more pack-years cigarette smoking history; and, in former smokers, having quit smoking within the past 15 years.

The study findings were published in the Journal of Thoracic Oncology.

Previous research has found that African American ever-smokers are at increased adjusted risk for lung cancer. Features relevant to lung cancer risk in African Americans have been shown to be at variance with those of their White counterparts. African Americans’ cancers tend to be more advanced at diagnosis, and their survival rates are lower. They start smoking at a later age but are diagnosed at an earlier one, smoke fewer cigarettes per day, have longer smoking duration, and are less likely to quit.

Mary M. Pasquinelli

Studies suggest that African American ever-smokers at high risk for lung cancer may benefit from screening more than any other racial/ethnic group. However, an analysis of the recent Southern Community Cohort Study of 48,364 ever-smokers found that a smaller proportion of African Americans would have met the current USPSTF criteria, compared with Whites (17% vs. 31%, respectively; P < .001). The current study suggests that USPSTF criteria may be insufficiently sensitive for identifying risk in African Americans and may exclude individuals who are at high risk for lung cancer, denying them the potential benefits of screening. The study’s objective was conducted to compare the sensitivities of the USPSTF model with PLCOm2012, a validated logistic regression lung cancer risk prediction model based on 11 predictors and derived from data collected from the control arm of the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, a randomized, controlled trial studying screening to reduce cancer mortality.

The focus was a large case series (n = 883) of incident smoking-related lung cancers in a predominately African American population from an urban academic medical center. The patients’ average age was 64.8 years, and 55.8% were men.

Race distribution was 56.3% African American, 29.2% White, 7.8% Hispanic, 2.7% Asian, and 4.0% other or missing. Compared with Whites, African Americans had lower pack years (37.3 vs. 48.2; P < .001), were more likely to be current smokers (65.6% vs. 57.0%; P < .001), and had a lower mean body mass index (25.8% vs. 27.1%; P = .004).

PLCOm2012 model more sensitive

The researchers used 6-year risk thresholds of ≥1.51%, ≥1.70%, and ≥2.00%.

In screening for lung cancer at the 1.70% threshold, the PLCOm2012 model was more sensitive than the USPSTF model. While only 44/833 cases (5.0%) were USPSTF-criteria positive and PLCOm2012-criteria negative (USPSTF+/PLCO1.7–), 166 (18.8%) were PLCO1.7+/USPSTF– (McNemar’s odds ratio, 3.77; 95% confidence index, 2.69-5.39; P < .0001).

“This does provide some good evidence by applying risk prediction models to data from real patients, and it could help narrow the gap in lung cancer screening,” Stacey Fedewa, PhD, of the American Cancer Society, said in an interview.

The study also showed that overall, 400 of 883 (45.3%) cases were USPSTF ineligible. Of these 400, 91 were White and 240 were African American. Among them, 145 (36.2%) would be eligible by the PLCOm2012 ≥1.7% criteria.

Using PLCOm2012 ≥1.7% criteria, 24 of the 91 Whites (26.4%) would become eligible, and 109 of 240 African Americans (45.4%) would become eligible.

The analysis further showed that, in contrast to the USPSTF criteria, the proportion of patients found to be eligible by the PLCOm2012 model did not differ significantly between White and African American ever-smokers at any of the three risk thresholds (all P > .10).

“We found that the PLCOm2012 model was significantly more sensitive in selecting lung cancer patients as being eligible for screening,” the authors wrote. “Use of an accurate lung cancer risk prediction model which accounts for the independent elevated risk in African American individuals can help overcome this disparity in lung cancer screening.”

The study findings using the PLCO m2012 model have implications for public and clinical health, the researchers wrote. “The USPSTF lung cancer screening is a grade B recommendation requiring private insurance companies to cover the cost without cost sharing (copayments, deductibles, or coinsurance) under the Affordable Care Act with the aim of eliminating financial barriers. It has been suggested that lowering the pack-year criteria from ≥30 to ≥20 might be a solution to the current disparity in enrollment of African American ever-smokers into screening, however, the PLCOm2012 model generally estimates risk better because of inclusion of additional risk factors and superior modeling of smoking exposures.”

Dr. Fedewa noted that an updated draft of USPSTF guidelines for lung cancer screening has been adjusted to include more African Americans.

The researchers did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.