Detection of non-metastatic non-small-cell lung cancer in urine by methylation-specific PCR analysis: A feasibility study.

BACKGROUND

Lung cancer has the highest cancer-related mortality worldwide and earlier detection could improve outcomes. Urine circulating tumor DNA (ctDNA) represents a true non-invasive means for ambulant sample collection. In this prospective study, the potential of urine for perioperative detection of non-metastatic non-small cell lung cancer (NSCLC) using ctDNA methylation analysis is evaluated.

METHODS

Preoperative urine samples of 46 surgical NSCLC patients and 50 sex and age-matched controls were analyzed for DNA methylation of NSCLC-associated methylation markers CDO1, SOX17, and TAC1, using quantitative methylation-specific PCR (qMSP). The accuracy for NSCLC detection was determined by univariable and multivariable logistic regression analysis, followed by leave-one-out cross-validation. Fourteen additional urine samples were collected postoperatively to evaluate whether DNA methylation levels alter after surgery with curative intent.

RESULTS

Methylation levels of CDO1 and SOX17 were significantly elevated in patients compared to controls (P =.016 and P <.001, respectively). This marker combination yielded an area under the receiver operating curve (AUC) value of 0.71 upon leave-one-out cross-validation for non-metastatic NSCLC detection in urine. Stage I patients tended to have higher methylation levels of SOX17 as compared to stage III patients. Similar methylation levels were found across the different histological subtypes of NSCLC. In some patients with preoperative elevated methylation levels, reduced methylation levels were found in post-operative urine samples.

CONCLUSIONS

Urine CDO1 and SOX17 showed increased methylation levels in NSCLC patients as compared to sex- and age-matched controls. This demonstrates that urine ctDNA methylation analysis may provide an interesting non-invasive means to detect non-metastatic NSCLC. Further studies are needed to validate the clinical usefulness of this approach and to assess the potential of post-operative monitoring.