| Objective:The objective of this study was to examine the effectiveness of previously untreated gestational trophoblastic neoplasia(GTN)and identify risk factors associated with resistance and recurrence under various treatment regimens.Furthermore,our aim was to develop a novel prognostic prediction model to enhance the existing GTN prognostic scoring system.Methods:A total of 553 GTN patients who underwent initial treatment at Peking Union Medical College Hospital between 2013 and 2020 were included.Of these patients,186 received first-line single-agent chemotherapy,and 367 received first-line multiple-agent chemotherapy.20 patients experienced recurrence.We introduced the hCG slope,which represents the linear fit slope of the hCG change trend before chemotherapy,as a novel prognostic factor.We compared the clinical baseline characteristics of GTN patients who were resistant to single-agent and multiple-agent chemotherapy with those who achieved complete remission(CR).Additionally,we compared the clinical characteristics of patients with and without recurrence.Employing logistic regression analysis and nomograms,we developed probability prediction models for resistance to single-agent chemotherapy,resistance to multiple-agent chemotherapy,and recurrence after successful treatment.The reliability of these models was assessed through the utilization of the C-index and calibration curves.Results:1.Among the 186 patients treated with first-line single-agent chemotherapy,the resistance rate was 29.0%,whereas for the 367 patients treated with first-line multiple-agent chemotherapy,the resistance rate was 19.3%.2.The single-agent chemotherapy resistance group exhibited significant differences compared to the CR group in terms of pretreatment hCG levels(median:3143 vs.655 IU/L,p<0.001),interval from index pregnancy(median:52.5 vs.81.0 days,p<0.001),and maximum tumor diameter(median:0.87 vs.0.50 cm,p=0.014).The prediction model for single-agent chemotherapy resistance,which utilized a categorical variable with a pretreatment hCG cutoff of 1160 IU/L,alongside continuous variables for interval from index pregnancy and maximum tumor diameter,demonstrated a C-index of 0.7236 and exhibited a calibration curve that closely approximated the ideal curve.3.Noteworthy differences between the multiple-agent chemotherapy resistance group and the CR group comprised of pretreatment hCG levels(median:22983 vs.67023 IU/L,p<0.001),maximum tumor diameter(median:3.2 vs.3.5 cm,p=0.025),hCG slope(median:1479.3 vs.890.4 IU/(L·d)-1,p=0.002),and the presence of lung metastases(p=0.028).Nevertheless,multivariate logistic regression analysis revealed that solely pretreatment hCG served as a risk factor for resistance to multiple-agent chemotherapy.The resulting nomogram prediction model,incorporating this factor,yielded a C-index of 0.6499 but exhibited a calibration curve that significantly deviated from the ideal curve.4.Patients who experienced recurrence displayed significant differences compared to those without recurrence in terms of antecedent pregnancy(p=0.002),interval from index pregnancy(median:331.5 vs.73 days,p<0.001),the number of metastases(p<0.001),and resistance to first-line chemotherapy(p=0.015).The nomogram model for predicting recurrence,which incorporated these four risk factors,exhibited a substantial degree of concordance between its calibration curve and the ideal curve.However,due to the limited number of recurrence cases,the calculation of its C-index was not feasible.5.The hCG slope,serving as an indicator of pretreatment hCG change trends,demonstrated a significantly lower value in the multiple-agent chemotherapy resistance group than in the CR group.While generally lower in patients with recurrence compared to those without recurrence,the difference did not reach statistical significance(1457.4 vs.1670.5 IU/(L·d)-1,p=0.067).Conclusions:The GTN prognostic scoring system can be subdivided into a new system that assesses resistance and recurrence as independent primary outcomes.Although the nomogram prediction models for resistance to single-agent chemotherapy and recurrence exhibit relatively good predictive power,their effectiveness still requires confirmation through external validation cohort data.Preliminarily,a lower hCG slope appears to be linked to resistance to multiple-agent chemotherapy and recurrence,necessitating further investigation into its underlying mechanism. |
PDF Full Text Request