| Popis: |
INTRODUCTION: The aim of the study was to investigate the interobserver variability of the newly developed Prostate Imaging for Relapse Reporting (PI-RR) to assess recurrence in patients with prostate cancer who received radiotherapy. METHODS: A total of 81 patients who underwent multiparametric magnetic resonance imaging (mpMRI) examination between January 2018 and November 2022 due to the suspicion of local recurrence were retrospectively evaluated. All the mpMRI images were independently reviewed by three radiologists. None of the observers were blinded to clinical and pathological data (including primary tumor location). It was examined whether interobserver agreement increased according to the prostate-specific antigen (PSA) cut-off value being taken as 0,5 or 0,6 (µg/L) and the size of the tissue with tumor suspicion being greater than 1 cm or 1,5 cm. RESULTS: The mean age of the patients was 68,3 +- 6,6 years. The mean PSA was 2,32 +- 2,95 (µg/L), and the mean tumor size was 1,23 +- 0,77 (cm). A very high level of agreement was detected between the gold standard and observers 1, 2, and 3 (κ: 0,921 / 0,873 / 0,825, respectively). The agreement between the observers and the gold standard was also increased in the evaluation of patients with a PSA value above 0,5 (µg/L). Furthermore, patients with a tumor size greater than 1 cm had a very high level of agreement between the gold standard and the observers. DISCUSSION AND CONCLUSION: PI-RR detection rate of relapse tissue and interobserver agreement were correlated with PSA level and size of relapse tissue. |
