Introduction

Lobular endocervical glandular hyperplasia (LEGH) is a benign cervical lesion, and excessive surgical procedure should be avoided in patients of reproductive age to preserve fertility. LEGH has been considered as a precursor lesion of cervical adenocarcinomas, especially of gastric-type adenocarcinoma (GAS), and early diagnosis is directly linked to improved prognosis, thus making imaging diagnosis an important role ¹. GAS including minimal deviation adenocarcinoma (MDA) is a subtype of cervical adenocarcinomas with poor prognosis that occurs more frequently in young women, and is characterized by the absence of human papillomavirus (HPV) infection in its pathogenesis ^2-4. The purpose of this study was to evaluate the capability of MR imaging for the diagnosis of LEGH-associated uterine cervical adenocarcinoma (LEGH-AC).

Methods

MR imaging features of pathologically proven 9 LEGH-ACs (5 GAS; 2 MDA; 2 Adenocarcinomas, NOS) (Table 1) were retrospectively evaluated. All patients were studied by using 3T (4 patients) or 1.5T (5 patients) superconducting MRI units with body coil transmission and phased-array receiver. T2WI, T1WI and DWI (b=800 s/mm²) were obtained in all 9 lesions. The contrast-enhanced study was performed in 7 patients (including dynamic contrast-enhanced (DCE) MRI study in 4 patients). The parameters of each sequence varied because this study was retrospective and involved a long study period (2007 to 2020). The ADC measurement was performed and computed DWI (b=1500 s/mm²) was synthesized on the workstation (Ziostation2, Ziosoft) in all 9 lesions. MR imaging features of pathologically proven 10 benign LEGH (Table 2) were also evaluated as comparisons.

Results and Discussions

On T2WI, all LEGH-ACs appeared as multicystic masses with large or small solid components. The solid components showed slight high signal intensity with/without sponge-like appearance reflecting tiny glandular cysts on T2WI (Fig. 3, 4). The solid components of all 7 lesions with contrast-enhancement study showed contrast-enhancement (weak 5; intense 2, compared with the myometrium), and all 4 lesions with DCE-MRI showed rapid and prolonged contrast-enhancement pattern (Fig. 3, 4). All 9 lesions showed increased signal intensity on DWI (high 5, slight high 4), and 5 of 9 lesions showed significant signal decrease on computed DWI with higher b-value (1500 s/mm²) (Fig. 4). Mean ADC value of all 9 lesions was 1.84 +/- 0.44 x 10^-3 mm²/s. The mean ADC value of 5 lesions with significant signal decrease on computed DWI (2.16 x 10^-3 mm²/s) was higher than that of the other 4 lesions without significant signal decrease (1.44 x 10^-3mm²/s) (p<0.05).
On the other hand, all LEGH appeared as multicystic masses with low to slight high signal intensity sponge-like solid components on T2WI. The solid components of all 10 lesions showed contrast-enhancement (weak 6; intense 4, compared with the myometrium), and 6 lesions with DCE-MRI showed rapid and prolonged contrast-enhancement pattern (2 lesions) or gradual contrast-enhancement pattern (4 lesions) (Fig. 5). LEGH showed increased signal intensity on DWI (slight high: 7 lesions) or iso intensity (3 lesions), and all 10 lesions showed significant signal decrease on computed DWI with higher b-value (1500 s/mm²) (Fig. 5). Mean ADC value of 10 lesions was 2.07 +/- 0.23 x 10^-3 mm²/s. There was no statistically difference between ADC values of 10 LEGH-ACs and 10 LEGHs, however, ADC value of 4 LEGH-ACs which did not show signal decrease on computed DWI (b=1500) was significantly lower than that of LEGH (p<0.01).
In this study, the presence of multicystic area in LEGH-AC may suggest co-existing LEGH, and the solid component exhibiting water diffusion restriction may be suggestive for carcinomatous component. However, this finding was observed in 4 of 9 LEGH-ACs (44%) and the other 5 lesions (56%) showed increased signal intensity on DWI possibly due to T2 shine-through effect reflecting abundant mucinous material-filled tiny cystic components within the tumor. Whereas 7 of 10 LEGHs also showed increased signal intensity on DWI possibly due to T2 shine-through effect reflecting mucinous material-filled aggregated tiny cysts and may be undifferentiable from LEGH-AC. Ohya et al. evaluated DWI of 7 LEGH-ACs and diffusion restriction was observed in 4/7 lesions (57%) ⁵. These results suggest that there might be the limitation in differentiating LEGH-AC and benign LEGH based on DWI. All 4 LEGH-ACs (100%) and 2 of 6 LEGH (33%) with DCE-MRI showed rapid and prolonged contrast-enhancement pattern, whereas no LEGH-ACs and 4 of 6 LEGHs (66%) showed gradual contrast-enhancement pattern which may be suggestive for benign lesions without angiogenesis.

Conclusion

We conclude that uterine cervical multicystic mass with diffusion restricted solid components may be suggestive for LEGH-AC, and computed DWI with higher b-value may be helpful for the diagnosis. Some LEGH may be undistinguishable from LEGH-AC on unenhanced MRI including DWI, gradual contrast-enhancement pattern on DCE-MRI may be suggestive for their benignity.

Acknowledgements

No acknowledgement found.