[Risk assessment of subsequent drug resistance in patients with chemoresistant gestational trophoblastic neoplasia].

Abstract

Objective: To investigate risk factors associated with subsequent drug resistance in patients with chemoresistant gestational trophoblastic neoplasia (GTN). Methods: Clinical and pathological data of GTN patients diagnosed and treated at Peking Union Medical College Hospital, Chinese Academy of Medical Science between January 2002 and June 2019 were retrospectively collected. Based on their prior treatment history and drug resistance status, GTN patients were divided into the chemoresistant group (n=248) and the primary treatment group (n=438) for comparison. Logistic regression analysis was used to evaluate the association between the International Federation of Gynecology and Obstetrics (FIGO, 2000) prognostic score and the risk of subsequent resistance in the chemoresistant group. Receiver operating characteristic (ROC) curve analysis was applied to assess the predictive value of the FIGO prognostic score for subsequent resistance; the Youden index of the FIGO score was calculated to determine its optimal cut-off value. Based on the cut-off value, the chemoresistant group was further stratified into two subgroups for prognosis comparison. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for subsequent resistance, and a novel scoring system was developed using these factors to predict the risk of subsequent resistance. Results: (1) The median FIGO prognostic score was significantly higher in the chemoresistant group than that in the primary treatment group (9 vs 7, P<0.001). The resistance rate and disease progression rate after the current treatment were significantly higher in the chemoresistant group [34.7% (86/248) and 11.3% (28/248), respectively] than those in the primary treatment group [14.8% (65/438) and 0.9% (4/438), respectively; both P<0.001]. Stratified analysis by FIGO prognostic score showed that, in each score stratum of FIGO prognostic score≥9, the resistance rate and disease progression rate were significantly higher in the chemoresistant group than in the corresponding stratum of the primary treatment group (all P<0.05). (2) FIGO prognostic score was significantly associated with subsequent resistance risk (OR=1.532, 95%CI: 1.359-1.726, P<0.001). The chemoresistant group was then divided into subgroups with FIGO prognostic score≥10 and <10. Compared with the FIGO prognostic score<10 subgroup, the FIGO prognostic score≥10 subgroup had significantly higher rates of subsequent resistance [13.5% (19/141) vs 62.6% (67/107), P<0.05], disease progression [2.1% (3/141) vs 23.4% (25/107), P<0.05] and recurrence [5.7% (8/141) vs 14.0% (15/107), P<0.05]. (3) Multivariate logistic regression analysis identified the following as independent risk factors for subsequent resistance: interval from index pregnancy≥12 months (OR=4.367, 95%CI: 2.199-8.927, P<0.001), pretreatment human chorionic gonadotrophin-β subunit≥500 U/L (OR=3.969, 95%CI: 2.076-7.833, P<0.001),≥2 previous failed chemotherapy lines (OR=2.458, 95%CI: 1.124-5.399, P=0.024), and ≥7 previous failed chemotherapy courses (OR=2.216, 95%CI: 1.046-4.665, P=0.036). (4) The aforementioned 4 risk factors were used as predictive variables to develop a novel scoring system for subsequent resistance in chemoresistant GTN patients, with a maximum score of 6. ROC curve analysis of the novel scoring system showed the area under the curve of 0.829, the sensitivity of 69.8%, the specificity of 83.3%, and the optimal cut-off value of 3.5. Accordingly, chemoresistant GTN patients with a score≥4 were classified as the high-risk population for subsequent resistance. Conclusions: Chemoresistant GTN exhibit distinct prognostic characteristics. We propose a new FIGO prognostic score cut-off value and a novel scoring system for this population. Chemoresistant GTN patients with a FIGO score≥10 or patients with a novel score≥4 should be regarded as at high risk of subsequent resistance.