Changes in serum tumor biomarkers may indicate treatment efficacy. Mathematical modeling is a promising tool for analyzing serum tumor marker declines. Indeed it allows calculation of the mathematical equations describing the longitudinal tumor biomarker time-changes (Almufti & You, et al. Annals of Oncology 25: 41–56, 2014). The model-based population kinetic approach is particularly relevant as it enables determination of individual kinetic profiles parameters based on a few timepoints, with limited impact of inter- and intra-individual variability of timepoints and assays.
We have used these approaches to assess the kinetic profiles of different serum tumor markers, such as prostate specific antigen (PSA), human chorionic gonadotrophin (hCG), alfa fetoprotein (AFP), CA-125, circulating tumor cells,…, during cancer treatments. In these studies, mathematical modeling of tumor marker individual kinetics was feasible. Moreover modeled kinetic parameters harboring strong reproducible predictive values regarding treatment efficacy were extracted (Almufti & You, et al. Annals of Oncology 25: 41–56, 2014). Based on a few timepoints analyzed with these models, it is easily possible for any clinician to calculate modeled kinetic parameters able to inform early on the risk of failure.
In the present site, 3 of these tools (Engines items above) are provided for clinicians/scientists:
Modeled PSA clearance after radical prostatectomy in patients with low-risk prostate cancers (pT2-pT3aN0R0; no lymph node or seminal vesicle involvement, no positive margins). In a prospective validation study, this kinetic parameter calculated
with 3 to 4 values of PSA measured during the first 30 days after positive prostatectomy was shown to be predictive of the risk of biochemical relapse without any adjuvant treatment (for more details). It informs clinicians on the risk of relapse in individual patients after radical prostatectomy, when they enter minimum 3 values of PSA during the post-operative first 30 days. Ideally these PSA values should have been performed on day 0 (surgery day); day 2; day 7; and day 27.
Modeled hCGres in patients with low-risk gestational trophoblastic neoplasias treated with the 8-day Methotrexate regimen. In two large retrospective studies, modeled hCGres parameter calculated with
weekly values of hCG measured during the first 50 treatment days after start of methotrexate was shown to be predictive of the risk of methotrexate resistance (for more details). It informs clinicians on the risk of resistance in individual patients, when they enter minimum 4 values of hCG during the first 50 days after methotrexate start.
Modeled CA-125 KELIM (elimination rate ) in patients with high grade serous ovarian carcinomas treated with first line chemotherapy (carboplatin-paclitaxel +/- bevacizumab). This parameter, calculated with CA-125 value measured at each cycle during the first 100 days of chemotherapy, exhibits strong independent and reproducible prognostic value regarding progression-free survival and overall survival in the retrospective analysis of 3 large phase III trials (AGO-OVAR 7; AGO-OVAR 9, and ICON-7). Based on CA-125 concentrations measured at every 3-week cycle during the first chemotherapy 100 days, patients can be classified with “favorable CA-125 decline” associated with long overall survival > 60 months, or “unfavorable decline” associated with short overall survival < 40 months. Moreover median expected survival associated with patient KELIM value can be simulated using AFT model.
The kinetic parameters are calculated for information purpose only. The authors of this site do not take any responsibility or endorse treatment decisions.