Evaluation of Prostate Specific Antigen (PSA) Kinetics in Prediction of Prostate Cancer Progression
| dc.contributor.advisor | Trock, Bruce J. | en_US |
| dc.contributor.author | Zhang, Dongyu | en_US |
| dc.contributor.committeeMember | Wang, Hao | en_US |
| dc.date.accessioned | 2015-02-11T04:15:57Z | |
| dc.date.available | 2015-02-11T04:15:57Z | |
| dc.date.created | 2014-05 | en_US |
| dc.date.issued | 2014-04-25 | en_US |
| dc.date.submitted | May 2014 | en_US |
| dc.description.abstract | Purpose Serum prostate specific antigen (PSA) showed unfavorable accuracy to predict prostate cancer progression. With development of multidisciplinary medical science and impending need, the rate of serum PSA change “PSA kinetics” might be a potential biomarker to help predict. In this study we aim to evaluate the prediction accuracy of prostate-specific antigen velocity (PSAV) and prostate-specific antigen doubling time (PSADT) for low-risk prostate cancer progression among men in Active Surveillance Program from Johns Hopkins Hospital. Methods We evaluated 614 patients in the active surveillance program of Johns Hopkins Hospital from 1994 to 2012 who met the criteria of either low risk or very low risk prostate cancer. During the follow-up, prostate specific antigen (PSA) testing was performed twice every year, and 12-14 core biopsy was performed annually. The demographic and clinic relevant data were analyzed by univariate comparison methods and we used multivariate Cox proportional hazards analysis to calculate the association between PSA kinetics and disease progression and hazard ratio (HR) was the measure of association. Bootstrapping bias-corrected concordance index (c-index) was utilized to measure the ability of discrimination for prediction models. Subgroup analysis was done based on the serum PSA level at diagnosis, and sensitivity analysis was performed when biological relevant endpoints changed. Results In our dataset there were 208 (33%) participants among the 614 developing the progression either in terms of Gleason Score or prostate volume, the median follow-up time of the 614 participants was 2.4 years. Totally 7 prediction models were selected. For all-sample analysis PSAV calculated by averaging arithmetic method showed significance in multivariate prediction model (HR=1.43 P=0.02 95%CI: 1.07, 1.92) if overall progression was treated as the endpoint. For subgroup with diagnostic PSA<4 ng/ml, 3 models were selected; and for subgroup with diagnostic PSA≥4 ng/ml, 2 models were selected. All these selected models had bias-corrected c-index over 0.70. Conclusion PSA kinetics can only show fair discrimination ability in aspect of low-risk prostate cancer progression regardless of biomedical endpoint or subgroup types used in the model. Pathological biopsy inspection should remain as the only reliable method to confirm disease progression. | en_US |
| dc.format.mimetype | application/pdf | en_US |
| dc.identifier.uri | http://jhir.library.jhu.edu/handle/1774.2/37243 | |
| dc.language | en | |
| dc.publisher | Johns Hopkins University | |
| dc.subject | prostate cancer | en_US |
| dc.subject | epidemiology | en_US |
| dc.subject | biomarker | en_US |
| dc.title | Evaluation of Prostate Specific Antigen (PSA) Kinetics in Prediction of Prostate Cancer Progression | en_US |
| dc.type | Thesis | en_US |
| dc.type.material | text | en_US |
| thesis.degree.department | Epidemiology | en_US |
| thesis.degree.discipline | Epidemiology | en_US |
| thesis.degree.grantor | Johns Hopkins University | en_US |
| thesis.degree.grantor | Bloomberg School of Public Health | en_US |
| thesis.degree.level | Masters | en_US |
| thesis.degree.name | Sc.M. | en_US |