Genetic risk score plus lung function questionnaire improves COPD prediction

Key takeaways:

• Researchers evaluated adults without a history of physician-diagnosed COPD.
• Area under the curve went up with use of a lung function questionnaire score plus a COPD genetic risk score.

The ability to predict spirometry-defined COPD was significantly better when a modified lung function questionnaire score was used in combination with a COPD polygenic risk score, according to results published in JAMA.

“Identifying individuals with undiagnosed COPD can improve outcomes, but spirometry is not readily available in many contexts,” Matthew Moll, MD, MPH, assistant professor at Brigham and Women’s Hospital and physician-scientist at Harvard Medical School, told Healio. “We demonstrated that genetics, which can be obtained from a buccal swab or blood test for a low cost, can help identify undiagnosed COPD individuals dwelling in the community.

“These tests are not yet clinically available, but everyday clinicians can take away that the genetic risks of COPD are important and that in the future it may be possible to identify undiagnosed individuals using questionnaires and genetics to prioritize who gets spirometry,” Moll continued.

In an observational study, Moll and colleagues evaluated 3,385 adults (median age, 52 years; 45.9% men) from the Framingham Heart Study and 4,095 adults (median age, 56.8 years; 55.5% men) from the COPDGene study — all without a history of physician-diagnosed COPD — to uncover the impact of the modified lung function questionnaire (mLFQ) score in combination with a COPD polygenic risk score (PRS) when predicting spirometry-defined moderate to severe COPD.

Using the definition of COPD derived from spirometry, researchers observed that a greater proportion of adults in the COPDGene cohort vs. the Framingham Heart Study cohort met the FEV₁/FVC and FEV₁ criteria for moderate to severe COPD (18.9% vs. 4.7%).

In the Framingham Heart Study cohort, the ability to predict spirometry-defined COPD was significantly better once the mLFQ score was used in combination with PRS (area under the curve [AUC], 0.78 to 0.84), according to the study.

Researchers also found this pattern in two subgroups from the COPDGene cohort: non-Hispanic African American adults (AUC, 0.69 to 0.72) and non-Hispanic white adults (AUC, 0.75-0.78).

Notably, the study reported that 13.8% of the Framingham Heart Study adults with spirometry-defined COPD had been “correctly reclassified” after PRS was added to the mLFQ score. The risk threshold for spirometry referral when researchers observed this was 10%.

In the COPDGene cohort, no adults with spirometry-defined COPD were reclassified with the 10% risk threshold.

“We were surprised to observe that adding genetics to a validated case finding questionnaire correctly reclassified individuals as having COPD in community-dwelling individuals [Framingham Heart Study], but in a cohort enriched for smoking [COPDGene], genetics were really less helpful in this regard and more useful for reclassifying individuals as not having COPD,” Moll told Healio. “These findings suggest that genetics might have the potential to address both under- and over-diagnosis.”

Looking ahead, Moll said the role of genetics in COPD prediction should be researched further.

“Future studies will test whether genetics can identify individuals with normal spirometry or preserved ratio and impaired spirometry who will go on to develop moderate to severe COPD,” Moll told Healio. “We can also assess the role of handheld spirometers. In addition, we are working toward developing genetic risk measures that have improved transportability across multi-ancestry populations.

“This study was a collaborative effort across multiple institutions, and we hope it is a proof-of-concept on how genetic risk measures can be used clinically, including potentially in lower resource settings,” Moll added.

Moll thanks Jingzhou Zhang, MD, MPH, clinical instructor of medicine at Boston University Chobanian & Avedisian School of Medicine, George O'Connor, MD, MS, professor of medicine at Boston University Chobanian & Avedisian School of Medicine, Michael Cho, MD, MPH, physician at Brigham and Women’s Hospital and assistant professor of medicine at Harvard Medical School, study co-authors and study participants.

For more information:

Matthew Moll, MD, MPH, can be reached at remol@channing.harvard.edu.