Background: Lung cancer is a major contributor to cancer related morbidity and mortality.Early detection would improve the prognosis by increasing the fraction of patients eligible for curative treatment.Methylated circulating tumour DNA (ctDNA) carries aberrant tumour-specific methylation patterns, which can be quantified in blood samples.We aimed to evaluate the diagnostic accuracy of a novel multiplex ctDNA panel in a multicentre cohort of high-risk patients undergoing diagnostic workup for lung cancer.Methods: Patients undergoing diagnostic workup for suspected lung cancer were recruited from five Danish hospitals from August 2020 to November 2023.Blood samples were collected before any invasive procedures and analysed by multiplex droplet digital polymerase chain reaction targeting aberrantly methylated regions in relation to the following genes: HOXA9, OTX1, MCIDAS, TFAP1B, and SP9.The cutoffs were predefined in an independent study, and 2 positive targets were used to call a sample ctDNA positive.The reference test was histopathological or cytological verification of lung cancer.Results: Of 265 recruited patients, 178 were diagnosed with lung cancer of which 41% had stage I-II.The ctDNA panel had a sensitivity of 83% and specificity of 33%, positive and negative predictive values of 72% and 48%, respectively, and positive and negative likelihood ratios of 1.2 and 0.5, respectively.The sensitivity for stage I+II was 83% with a similar value for stage III+IV at 82%.A logistic regression model based on sex, age, smoking status and performance status had an area under the receiver operating characteristics curve of 0.74.Adding the ctDNA panel resulted in an adjusted odds ratio of 2.2 (95% confidence interval 1.1-4.3,n=250, p=0.019) for detecting lung cancer, but did not significantly improve overall model performance compared to a model with only the clinical variables (p=0.715). Conclusions:Aberrantly methylated ctDNA might have potential as a diagnostic addon for lung cancer.The multiplex ctDNA panel showed a similar diagnostic sensitivity in both early-stage and late-stage disease in the present high-risk multicentre cohort, but did not significantly improve the clinical logistic regression model.
Bonanno et al. (Tue,) studied this question.