The accuracy of prostate biopsy to assign patients with low-grade prostate cancer to active surveillance
Ghleilib, Intisar Ali
MetadataShow full item record
PURPOSE: To determine the accuracy of prostate biopsy Gleason score (GS) compared to prostatectomy GS. To determine whether a biopsy is a satisfactory diagnostic procedure to offer active surveillance for patients with low-grade prostate cancer. METHODS: This study was conducted in Tuft Medical Center as retrospective cohort study over the period from 2007-2010. The study included 83 patients for whom biopsy and prostatectomy GS were available. MEASUREMENTS: Gleason scores of 6, 7, and 8-10 were assigned to low, moderate, and high-grades, respectively. The kappa statistic was calculated to assess the degree of agreement between biopsy and prostatectomy. The ROC curve was used to evaluate the sensitivity and specificity of prostate biopsy for different Gleason grades. Also, compared whether the use of specific criteria for active surveillance (Johns Hopkins and UCSF) may decrease the level of up-grading in patient with low-grade prostate cancer using Chi-square test. RESULTS: The distribution of low, moderate, and high-grade cancer in biopsy (52%, 32%, 16%) and prostatectomy specimen (33%, 55%, 12%) showed fair agreement with weighted kappa 0.35. The prostate biopsy accurately predicted GS in 46%, up-graded in 38%, and down-graded in 16%. The patients with low-grade cancer and potentially eligible for active surveillance showed up-grading in 50% of cases. This up-grading reduced to 40% with the use of Johns Hopkins criteria and to 41% with the use of UCSF criteria. CONCLUSIONS: The accuracy of biopsy GS in predicting prostatectomy GS is severely limited and therefore biopsy is not enough diagnostic procedure to offer active surveillance.