Pre-operative assessment in International Federation of Gynaecology and Obstetrics-2009 stage IB2 to IIB cervical cancer: role of delayed gadolinium-enhanced magnetic resonance imaging after neoadjuvant chemotherapy.

OBJECTIVE

To assess the value of delayed gadolinium-enhanced magnetic resonance imaging (MRI) in addition to standard T2- and diffusion-weighted (T2/diffusion-weighted) MRI for evaluating residual locoregional tumor in patients with International Federation of Gynaecology and Obstetrics (FIGO) stage-2009 IB2 to IIB cervical cancer treated with neoadjuvant chemotherapy.

METHODS

This retrospective study included patients diagnosed with FIGO-2009 stage IB2 to IIB cervical cancer between 2016 and 2020, who were treated with neoadjuvant chemotherapy and radical hysterectomy with pelvic lymphadenectomy, and staged with pelvic T2-, diffusion-weighted, and delayed gadolinium-enhanced MRI after neoadjuvant chemotherapy. Diagnostic accuracies of T2/diffusion-weighted MRI and delayed gadolinium-enhanced MRI, assessed by a sub-specialty radiologist (reader 1) and a general radiologist (reader 2), were compared for identification of residual tumor, parametrial, vaginal invasion, and detection of lymph node metastases. Inter-observer agreement was calculated. Additionally, Bland-Altman plots were used to compare the residual tumor size measured by T2/diffusion-weighted MRI and delayed gadolinium-enhanced MRI. Histopathology after surgery was the standard reference.

RESULTS

Forty-two patients were eligible for study analysis. Delayed gadolinium-enhanced (reader 1/reader2) enabled identification of all 4 complete responders, while T2/diffusion-weighted MRI identified only 2. Delayed gadolinium-enhanced MRI (reader 1/reader 2) significantly improved diagnostic accuracy for detecting lymph node metastases (84.5%/82.2% vs 72.6%/70.2%, p <.0001) compared to T2/diffusion-weighted MRI. The addition of delayed gadolinium-enhanced MRI improved accuracy for detecting parametrial invasion (97.6%/88.1% vs 85.7%/78.6%) and vaginal invasion (92.9%/88.1% vs 80.9%/76.2%, p =.063). Delayed gadolinium-enhanced MRI showed better inter-observer agreement than T2/diffusion-weighted MRI for assessment of vaginal invasion (κ, 0.86 vs 0.33), parametrial invasion (κ, 0.63 vs 0.44), and nodal status (κ, 0.90 vs 0.79).

CONCLUSIONS

Addition of delayed gadolinium-enhanced MRI substantially improved pre-operative identification of extra-uterine primary tumor extent, significantly improved detection of metastatic lymph nodes, and allowed better assessment of residual tumor size compared with T2/diffusion-weighted MRI alone in patients with FIGO-2009 stage IB to IIB cervical cancer treated with neoadjuvant chemotherapy.