An Investigation On The Correlation Between Prostate-specific Antigen Kinetics And The Prognosis Of The Patients With Prostate Cancer After The Hormonal Therapy

Objective:To evaluate the usefulness of the long-term surveillance of Prostate-specific antigen (PSA) kinetics in predicting the prognosis of patients with prostate cancer treated by the hormonal therapy through the investigation on the association between PSA kinetics and the progression and survival of the disease, accompanied with the assessment of other risk factors, including body mass index, clinical stage, bone metastasis, Gleason scores, etc.Methods:Our study enrolled the patients’data from the database of patients with prostate cancer admitted into the department of urology in Huashan Hospital affiliated Fudan University, from2002.1.1to2009.12.31.The clinical data we collected from the patients covered the age at the start of the treatment, TNM staging, the chief complaints, body mass index (BMI), PSA, prostate volume (PV) measured by transrectal ultrasound (TRUS), the result of digital rectal examination (DRE), the pathological diagnosis of prostate needle biopsy, Gleason score, and the imaging test results from chest X-ray, the ultrasound of abdominal organs and lymph nodes in posterior peritoneum, TRUS, bone scan, computed tomography (CT) and magnetic resonance imaging (MRI). The results of CT and MRI test were optional. The extent of disease score (EOD) was also calculated by the results of bone scan. The three study end points were respectively overall survival, cancer-specific survival and biochemical failure. All the patients had the completed objective following data on PSA test (at least every3months during the first3years and at last each half year during the following years) and abdominal ultrasound (every2years), bone scan (every3years). Subjective following data was obtained by telephone investigation in order to gain the information of the survival, and general performance. The last date of following-up ranged from2012.1.1to2012.2.28.We used the SPSS19.0as the tool for statistic analysis. The univariate and multivariate analysis were launched to figure out the potentially relevant factors for PSA normalization and the occurrence of undetectable post-treatment level of PSA. COX regression was used for seeking the risk factors of the death for any cause, the death for prostate cancer and the biochemical failure (BF). Receiver operating characteristic (ROC) curve was applied in the evaluation of the proper cut-off of PSA nadir and time to PSA nadir for predicting the death for any cause. Kaplan-meier curves were analyzed to distinguish the effects of PSA nadir and time to PSA nadir on the overall survival (OS), the cancer-specific survival and the biochemical progression-free survival (bPFS). Logistic regression was used to analysis the risk factors related to the different level of PSA nadir and time to PSA nadir.ResultsIn our study, there were332eligible patients enrolled with the median age at diagnosis,75years old (50to90yrs). The initial mean PSA was206.44±416.97ng/ml (median78.8ng/ml) and the average PSA nadir after the launch of the ADT was3.10±12.80ng/ml. There were45cases with the locally advanced stages of T3/T4/Nx-N1(13.6%), while251cases were at the Ml stage, which meant metastatic disease (75.60%).325cases in all of the patients received surgical castration combined with flutamide(97.90%), while7were received medical castration and flutamide. The median follow-up time was48.7months. The results showed that the PSA normalization, the occurrence of the undetectable PSA level and time to PSA nadir seemed to be the best independent risk factors of the prediction of the three types of disease outcomes. In overall death, Gleason score (HR=1.397, P=0.020), PSA normalization (HR=0.108, P=0.004). PSA reaching the undetectable level (HR=0.205, P=0.000) and time to PSA nadir (HR=0.889, P=0.000) were the major risk factors. In cancer specific death, only PSA normalization (HR=0.128, P=0.014), PSA reaching the undetectable level (HR=0.128, P=0.000) and time to PSA nadir (HR=0.889, P=0.000) were the independent risk factors. In biochemical failure, PSA normalization (HR=0.461, P=0.032), PSA reaching the undetectable level (HR=0.388, P=0.000) and time to PSA nadir (HR=0.950, P=0.000) were the distinctive factors. In the mean time, we used respective cut-offs of time to PSA normalization, PSA nadir, time to PSA nadir, adjusted time to PSA nadir and PSA half time to compare the risk of overall survival, cancer specific survival and biochemical failure. Then all the results showed that longer time to PSA nadir predicted the better prognosis, but the shorter time to PSA normalization implied the good outcome. With the help of ROC curves, we identified that, in our study, the best cut-off of PSA nadir was0.2ng/ml (sensitivity65.7%and specificity80.6%) and the best cut-off for TTPN was10months (sensitivity71.6%and specificity63.9%). We used multivariate COX regression to identify the certain predictive factors relevant to PSA normalization and the occurrence of the undetectable level of PSA, and found higher EOD (OR=2.489, P=0.000) was associated with the higher possibility of the failure of PSA normalization. And higher Gleason Score, higher EOD (0R=1.871, P=0.000) and higher initial PSA level were associated with the higher possibility of the failure in reaching the undetectable level of PSA.Conclusions:Our study proved PSA kinetics, including PSA normalization, TTPN, PSA nadir, reaching to the undetectable PSA level, were the risk factors for predicting the prognosis in the following-up for patients with hormonal naive advanced prostate cancer. The best cut-off of PSA nadir is0.2ng/ml and the best cut-off for TTPN is10months. EOD score can influence the PSA kinetics.