BioProject

DNA was extracted from serum of 204 patients with prostate cancer (Gleason 2 to 10), 207 male controls, benign hyperplasia (n=10) and prostatitis (n=10). DNA was amplified using random primers, tagged with molecular identifiers, sequenced on a SOLiD system and aligned to the human genome. The number of sequence reads of cfDNA was evaluated in sliding 100 kbp intervals for variation from controls. Chromosomal regions with significant variations in alignment hits were used for their ability to segregate patients and matched controls. Using ROC curves to assess diagnostic performance, the number of regions in a first subset (n=177), with variations in alignment hits alone evaluated, provided an AUC of 0.81 (0.7 – 0.9, p<0.001). Using five rounds of 10-fold cross-validation with the full dataset, a final model was established that discriminated prostate cancer from controls with an AUC of 0.92 (0.87 – 0.95), reaching a diagnostic accuracy of 83%. Both benign prostatic hypertrophy and prostatitis could be distinguished from prostate cancer using cfDNA with an accuracy of 90%. Assessment of a limited number of chromosomal structural instabilities using massive parallel sequencing of cfDNA was sufficient to distinguish between prostate cancer and controls. This large cohort demonstrates the utility of cfDNA in prostate cancer recently established in other malignant neoplasms.