FDG-PET/MRI combined with SPECT/CT guided sentinel lymph node mapping for lymph node staging in patients with early stage cervical and endometrial carcinoma: a prospective study.

Aim: Lymph node metastases (LNM) are rare in early stage endometrial and cervical carcinomas. Thus, minimal invasive lymph node staging is desirable to prevent from complications associated with a full lymphadenectomy for diagnostic purpose. Aim of this prospective study was to analyze the impact of a combined protocol including FDG-PET/MRI followed by SPECT/CT guided sentinel lymph node (SLN) biopsy.

Methods: 53 women with initial diagnosis of Figo IA-IIB endometrial and cervical carcinomas were assessed. All patients received preoperative whole body FDG-PET/MRI (Siemens Biograph mMR®) as well as SPECT/CT (GE Discovery 670 Pro®) after intracervical Tc-99m Nanocoll injection performed on the day before surgery. Intraoperative lymph node (LN) mapping was conducted using a laparoscopic gamma probe. LN status was validated by histology.

Results: One patient was excluded from lymphadenectomy due to liver metastases on PET/MRI. 10/52 patients (19%) presented with LNM in 13/104 hemiplevis (13%). Interestingly, there were no LNM in patients with endometrial carcinoma (0/18), whereas the prevalence for LNM in patients with cervical carcinoma was higher than expected (10/34 patients = 29%; literature 14-26%). 4/10 patient with LNM solely had lesions with a diameter ≤5mm. PET/MRI demonstrated an excellent specificity for LNM staging (100% per patient / 100% per hemipelvis), with a good accuracy (90%/92%), but low sensitivity (50%/43%). The size of LNM was found to be the most important factor on PET/MRI sensitivity with a sharp drop in detectability of metastases ≤5mm (LN ≤5mm: 0%; >5mm: 83%). Paraaortal LNM were evaluated accurately by PET/MRI in all 3 cases (LNM in 3/26 patients with paraaortal LN sampling). Detection rate of SLN in SPECT/CT was 82% per patient, 70% for hemipelvis and 32% for paraaortal LN. However, the presence of malignancy in patients with SLN was lower (16% per patient/12% per hemipelvis) than in patients without detected SLN (33%/16%). Only 1/3 patients with paraaortal LNM showed corresponding SLNs in SPECT/CT. The combination of both methods improved the detection rate of LNM significantly to 78% (PET/MRI alone 43%, SLN alone: 64%). Conclusion: FDG-PET/MRI is a highly specific imaging method in lymph node staging and can detect paraaortal and distant lymph node metastases sufficiently but has a limited sensitivity in small metastases ≤5mm. However, the combination of PET/MRI and SPECT/CT guided SLN biopsy increases the sensitivity and might have the potential to replace full lymphadenectomies in early tumor stages.