The value of multiparameter MRI of early cervical cancer combined with SCC-Ag in predicting its pelvic lymph node metastasis

Xiaoqian Xu; Fenghai Liu; Xinru Zhao; Chao Wang; Da Li; Liqing Kang; Shikai Liu; Xiaoling Zhang

doi:10.3389/fonc.2024.1417933

The value of multiparameter MRI of early cervical cancer combined with SCC-Ag in predicting its pelvic lymph node metastasis

Front Oncol. 2024 Sep 11:14:1417933. doi: 10.3389/fonc.2024.1417933. eCollection 2024.

Authors

Xiaoqian Xu¹, Fenghai Liu¹, Xinru Zhao¹, Chao Wang¹, Da Li¹, Liqing Kang¹, Shikai Liu², Xiaoling Zhang³

Affiliations

¹ Department of Magnetic Resonance Imaging, Cangzhou Central Hospital, Cangzhou, Hebei, China.
² Department of Gynecology, Cangzhou Central Hospital, Cangzhou, Hebei, China.
³ Department of Pathology, Cangzhou Central Hospital, Cangzhou, Hebei, China.

Abstract

Purpose: To investigate the value of multiparameter MRI of early cervical cancer (ECC) combined with pre-treatment serum squamous cell carcinoma antigen (SCC-Ag) in predicting its pelvic lymph node metastasis (PLNM).

Material and methods: 115 patients with pathologically confirmed FIGO IB1~IIA2 cervical cancer were retrospectively included and divided into the PLNM group and the non-PLNM group according to pathological results. Quantitative parameters of the primary tumor include K^trans, K_ep, V_e from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), ADC_mean, ADC_min, ADC_max, D, D^* and f from intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) were measured. Pre-treatment serum SCC-Ag was obtained. The difference of the above parameters between the two groups were compared using the student t-test or Mann-Whitney U test. Multivariate Logistic regression analysis was performed to determine independent risk factors. Receiver operating characteristic (ROC) curve analysis was used to assess the diagnostic efficacy of individual parameters and their combination in predicting PLNM from ECC.

Results: The PLNM group presented higher SCC-Ag [14.25 (6.74,36.75) ng/ml vs.2.13 (1.32,6.00) ng/ml, P<0.001] and lower K^trans (0.51 ± 0.20 min^-1 vs.0.80 ± 0.33 min^-1, P < 0.001), ADC_mean (0.85 ± 0.09 mm/s² vs.1.06 ± 0.35 mm/s², P<0.001), ADC_min [0.67 (0.61,0.75) mm/s² vs. 0.75 (0.64,0.90) mm/s², P = 0.012] and f (0.91 ± 0.09 vs. 0.27 ± 0.14, P = 0.001) than the non-LNM group. Multivariate analysis showed that SCC-Ag (OR = 1.154, P = 0.007), K^trans (OR=0.003, P < 0.001) and f (OR = 0.001, P=0.036) were independent risk factors of PLNM. The combination of SCC-Ag, K^trans and f possessed the best predicting efficacy for PLNM with an area under curve (AUC) of 0.896, which is higher than any individual parameter: SCC-Ag (0.824), K^trans (0.797), and f (0.703). The sensitivity and specificity of the combination were 79.1% and 94.0%, respectively.

Conclusions: Quantitative parameters K^trans and f derived from DCE-MRI and IVIM-DWI of primary tumor and SCC-Ag have great value in predicting PLNM. The diagnostic efficacy of their combination has been further improved.

Keywords: cervical cancer; diffusion-weighted imaging; dynamic contrast-enhanced; intravoxel incoherent motion; lymph node metastasis; magnetic resonance imaging; squamous cell carcinoma antigen.