189 Dynamics of methylated cell-free DNA in the urine of non-small cell lung cancer patients DISCUSSION Insight into the dynamics of urinary cfDNA is essential to determine whether a preferred collection time and sampling frequency exist, and to correctly interpret molecular analyses. Analysis of the circadian variation of the cfDNA concentration in the urine of NSCLC patients revealed substantial variation between and within subjects, but no clear circadian pattern. Similarly, also for methylation levels of lung cancer markers no clear circadian pattern was found, whereas the biological variation was high. Data of the current study suggests that the moment of urine collection does not significantly affect the urinary cfDNA concentration in NSCLC patients with active disease. Similarly, no day-to-day variation in urinary cfDNA concentration was found. So far, only the dynamics of cfDNA in plasma have been explored. Madsen et al. (24) reported similar results with stable cfDNA amounts in the plasma of lung cancer patients during the day and between days. Contradictory findings have been described for cfDNA concentrations in the plasma of healthy subjects. While constant cfDNA concentrations were observed by Wagner et al. (23), other studies demonstrated a significant decrease during the day in healthy subjects (24, 39). Previous studies also did not find a day-to-day variation of cfDNA in plasma (22, 24, 40), in line with the current findings. The only patient characteristic that influenced urinary cfDNA concentration levels in this study was sex, with a significantly higher concentration found in females, following previous studies (41-43). The proportion of between-subjects variation was expressed using the ICC value, where an ICC value of one indicates a perfectly reproducible test. Although the interpretation of the ICC value differs amongst studies, it has been suggested that ICC values below 0.50 reflect poor reproducibility, values between 0.50 and 0.75 moderate reproducibility, values between 0.75 and 0.90 good reproducibility, and values above 0.90 equal excellent reproducibility (44). The cfDNA concentrations measured in the six urine samples per patient showed between- and within-subject variability of comparable size, approaching a moderate reproducibility (0.49). In other words, both the baseline cfDNA concentration of each patient and the random fluctuation around this baseline contribute equally to the observed variation in cfDNA concentrations. Substantial between- and within-subjects variability has also been reported for serial measurements of cfDNA in plasma of healthy subjects (23, 24). Methylation levels of the CDO1, SOX17, and TAC1 genes were also not affected by the time of urine collection. This is in accordance with the stable detection of EGFR mutations in the plasma of lung cancer patients collected during three time points within one day 7
RkJQdWJsaXNoZXIy MjY0ODMw