Treatment outcomes in patients with FIGO stage IB–IIA cervical cancer and a focally disrupted cervical stromal ring on magnetic resonance imaging: A propensity score matching study

Prediction of Postoperative Pathologic Risk Factors in Cervical Cancer Patients Treated with Radical Hysterectomy by Machine Learning

Pretherapeutic serological parameters play a predictive role in pathologic risk factors (PRF), which correlate with treatment and prognosis in cervical cancer (CC). However, the method of pre-operative prediction to PRF is limited and the clinical availability of machine learning methods remains unknown in CC. Overall, 1260 early-stage CC patients treated with radical hysterectomy (RH) were randomly split into training and test cohorts. Six machine learning classifiers, including Gradient Boosting Machine, Support Vector Machine with Gaussian kernel, Random Forest, Conditional Random Forest, Naive Bayes, and Elastic Net, were used to derive diagnostic information from nine clinical factors and 75 parameters readily available from pretreatment peripheral blood tests. The best results were obtained by RF in deep stromal infiltration prediction with an accuracy of 70.8% and AUC of 0.767. The highest accuracy and AUC for predicting lymphatic metastasis with Cforest were 64.3% and 0.620, respectively. The highest accuracy of prediction for lymphavascular space invasion with EN was 59.7% and the AUC was 0.628. Blood markers, including D-dimer and uric acid, were associated with PRF. Machine learning methods can provide critical diagnostic prediction on PRF in CC before surgical intervention. The use of predictive algorithms may facilitate individualized treatment options through diagnostic stratification.

Definitive Chemoradiotherapy versus Radical Hysterectomy Followed by Tailored Adjuvant Therapy in Women with Early-Stage Cervical Cancer Presenting with Pelvic Lymph Node Metastasis on Pretreatment Evaluation: A Propensity Score Matching Analysis

Simple Summary

Pelvic nodal involvement is frequently present in early-stage cervical cancer patients on pretreatment imaging studies. However, it is unclear whether radical chemoradiotherapy (CRT) or radical hysterectomy RH followed by tailored adjuvant radiotherapy is more appropriate in these patients. We compared oncological outcomes of up-front surgery followed by tailored adjuvant radiotherapy and definitive CRT in these patients. We found no differences in outcomes existed between definitive CRT and hysterectomy with tailored adjuvant radiotherapy. However, after surgery, 88.7% of patients required adjuvant radiotherapy. These findings suggest that definitive CRT can avoid unplanned tri-modality therapy without compromising oncologic outcomes.

Abstract

To compare the oncologic outcomes between chemoradiotherapy (CRT) and radical hysterectomy followed by tailored adjuvant therapy in patients with early cervical cancer presenting with pelvic lymph node metastasis. We retrospectively analyzed the medical records of women with early cervical cancer presenting with positive pelvic nodes identified on pretreatment imaging assessment. Propensity score matching was employed to control for the heterogeneity between two groups according to confounding factors. Overall survival, disease-free survival, and pattern of failure were compared between the two groups. A total of 262 patients were identified; among them, 67 received definitive CRT (group A), and 195 received hysterectomy (group B). Adjuvant therapy was administered to 88.7% of group B. There were no significant differences between group A and group B regarding the 5-year overall survival rates (89.2% vs. 89.0%) as well as disease-free survival rates (80.6% vs. 82.7%), and patterns of failure. Distant metastasis was the major failure pattern identified in both groups. In multivariate analysis, non-squamous histology was significantly associated with poorer overall survival. As there are no significant differences in 5-year OS, DFS, and patterns of failure, definitive CRT could avoid the combined modality therapy without compromising oncologic outcomes.

Primary radical hysterectomy vs chemoradiation for IB2-IIA cervical cancer: A systematic review and meta-analysis

BACKGROUND

To compare the clinical outcomes of radical hysterectomy (RH) with chemoradiotherapy (CRT) in women with stage IB2-IIA cervical cancer.

METHODS

Based on articles published up to December 2017, a literature search of PubMed, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), and Chinese National Knowledge Infrastructure (CNKI) databases was conducted to identify eligible studies. Overall survival (OS), progression-free survival (PFS) with hazard ratios (HRs), and toxicities with odds ratios (ORs) were analyzed.

RESULTS

In total, 7 studies comprising 687 patients were identified for this meta-analysis. RH showed a significant trend toward improved survival outcomes compared with those of CRT, regardless of OS (HR = 0.49, 95% confidence interval [CI] 0.36-0.67, P < .001); or PFS (1.61, 95% CI 1.15-2.26, P = .005) for IB2-IIA cervical cancer. Subgroup analysis revealed that stage IB2 cervical cancer patients obtained better OS (HR = 0.36, 95% CI 0.23-0.56, P < .001; heterogeneity: P = .32, I = 13%). However, a higher incidence of grade 3/4 genitourinary abnormalities was evident with RH (OR = 2.3, 95% CI 1.42-3.87, P = .021).

CONCLUSION

Our study suggested that RH had distinct advantages over CRT for carcinoma of the uterine cervix with FIGO stage IB2-IIA, especially for IB2 cervical cancer.

Evaluation of a New Prognostic Tumor Score in Locally Advanced Cervical Cancer Integrating Clinical Examination and Magnetic Resonance Imaging

PURPOSE

The integral results of clinical examination and magnetic resonance imaging (MRI) of patients with locally advanced cervical cancer may provide prognostic information that cannot readily be placed in current staging systems, such as proximal versus distal parametrial invasion, unilateral versus bilateral involvement, or organ infiltration on MRI. The aim was to develop and investigate the performance of a simple but comprehensive tumor score for reporting and prognostication.

METHODS AND MATERIALS

In the present study, 400 consecutive patients with locally advanced cervical cancer treated 2005 to 2018 with chemoradiation and image guided adaptive brachytherapy (IGABT) were analyzed. The diagnostic workup included clinical examination, positron emission tomography/computed tomography, and MRI. International Federation of Gynecology and Obstetrics 2009 stage distribution was IB to IIA 9%, IIB 61%, and III to IV 30%. Involvement of 8 anatomic locations (cervix, left parametrium, right parametrium, vagina, bladder, ureter, rectum, and uterine corpus) was scored according to a ranked ordinal scale with 0 to 3 points. The total sum of points constituted the tumor score (T-score).

RESULTS

The median T-score was 6 (range, 0-20). Based on the frequency distribution of the T-score, 4 equally sized groups were formed: 0 to 4, 5 to 6, 7 to 9, and >9 points. The T-score grouping was highly significant in both univariate and multivariable analysis and outperformed International Federation of Gynecology and Obstetrics stage for both survival and local control enabling also intrastage prognostication. Used as a linear variable, the T-score was correlated with IGABT target volume (high-risk clinical target volume, CTV_HR), use of interstitial needles, dose (D90 of CTV_HR), and total reference air kerma.

CONCLUSIONS

The T-score is a simple instrument for combining clinical findings and imaging into a powerful prognostic factor for survival and local control with capabilities surpassing traditional staging. In addition, the T-score may already at diagnosis predict essential IGABT parameters and may be used for audit and comparison of results in multicenter settings.

Selection criteria and colpotomic approach for safe minimally invasive radical hysterectomy in early-stage cervical cancer

OBJECTIVE

To evaluate oncologic outcomes of minimally invasive radical hysterectomy (RH) in early cervical cancer before and after the application of parametrial invasion (PMI) criterion on magnetic resonance imaging (MRI) and vaginal colpotomy (VC).

METHODS

A total of 216 International Federation of Gynecology and Obstetrics stage IB-IIA cervical cancer patients who underwent minimally invasive RH was identified between April 2006 and October 2018. Patients were classified into the pre-PMI intracorporeal or VC (IVC) (n=117) and post-PMI VC groups (n=99). In the pre-PMI IVC group, PMI criterion (intact stromal ring) on MRI was not applied and the patients received IVC. In the post-PMI VC group, surgical candidates were selected using the PMI criterion on MRI and all patients received VC only. Oncologic outcomes and prognostic factors associated with disease recurrence were analyzed.

RESULTS

The rate of positive vaginal cuff margins in the pre-PMI IVC group was higher than that in the post-PMI VC group (11.1% vs. 1.0%, p=0.003). Two-year disease-free survival was different between the 2 groups (84.5% in pre-PMI IVC vs. 98.0% in post-PMI VC groups, p=0.005). Disrupted stromal ring on MRI (hazard ratio [HR]=20.321; 95% confidence interval [CI]=4.903-84.218; p<0.001) and intracorporeal colpotomy (HR=3.059; 95% CI=1.176-7.958; p=0.022) were associated with recurrence.

CONCLUSION

The intact cervical stromal ring on MRI might identify the low-risk group of patients in terms of PMI and lymphovascular/stromal invasion in early cervical cancer. Minimally invasive RH should be performed in optimal candidates with an intact stromal ring on MRI, using VC.