Simultaneous follow-up of mouse colon lesions by colonoscopy and endoluminal ultrasound biomicroscopy

摘要

AIM: To evaluate the potential use of colonoscopy and endoluminal ultrasonic biomicroscopy (eUBM) to track the progression of mouse colonic lesions. METHODS: Ten mice were treated with a single azoxymethane intraperitoneal injection (week 1) followed by seven days of a dextran sulfate sodium treatment in their drinking water (week 2) to induce inflammationassociated colon tumors. eUBM was performed simultaneously with colonoscopy at weeks 13, 17-20 and 21. A 3.6-F diameter 40 MHz mini-probe catheter was used for eUBM imaging. The ultrasound mini-probe catheter was inserted into the accessory channel of a pediatric flexible bronchofiberscope, allowing simultaneous acquisition of colonoscopic and eUBM images. During image acquisition, the mice were anesthetized with isoflurane and kept in a supine position over a stainless steel heated surgical waterbed at 37 °C. Both eUBM and colonoscopic images were captured and stored when a lesion was detected by colonoscopy or when the eUBM image revealed a modified colon wall anatomy. During the procedure, the colon was irrigated with water that was injected through a flush port on the mini-probe catheter and that acted as the ultrasound coupling medium between the transducer and the colon wall. Once the acquisition of the last eUBM/colonoscopy section for each animal was completed, the colons were fixed, paraffin-embedded, and stained with hematoxylin and eosin. Colon images acquired at the first time-point for each mouse were compared with subsequent eUBM/colonoscopic images of the same sites obtained in the following acquisitions to evaluate lesion progression. RESULTS: All 10 mice had eUBM and colonoscopic images acquired at week 13 (the first time-point). Two animals died immediately after the first imaging acquisition and, consequently, only 8 mice were subjected to the second eUBM/colonoscopy imaging acquisition (at the second time-point). Due to the advanced stage of colonic tumorigenesis, 5 animals died after the second time-point image acquisition, and thus, only three were subjected to the third eUBM/colonoscopy imaging acquisition (the third time-point). eUBM was able to detect the four layers in healthy segments of colon: the mucosa (the first hyperechoic layer moving away from the mini-probe axis), followed by the muscularis mucosae (hypoechoic), the submucosa (the second hyperechoic layer) and the muscularis externa (the second hypoechoic layer). Hypoechoic regions between the mucosa and the muscularis externa layers represented lymphoid infiltrates, as confirmed by the corresponding histological images. Pedunculated tumors were represented by hyperechoic masses in the mucosa layer. Among the lesions that decreased in size between the first and third time-points, one of the lesions changed from a mucosal hyperplasia with ulceration at the top to a mucosal hyperplasia with lymphoid infiltrate and, finally, to small signs of mucosal hyperplasia and lymphoid infiltrate. In this case, while lesion regression and modification were observable in the eUBM images, colonoscopy was only able to detect the lesion at the first and second time-points, without the capacity to demonstrate the presence of lymphoid infiltrate. Regarding the lesions that increased in size, one of them started as a small elevation in the mucosa layer and progressed to a pedunculated tumor. In this case, while eUBM imaging revealed the lesion at the first time-point, colonoscopy was only able to detect it at the second time-point. All colonic lesions (tumors, lymphoid infiltrate and mucosal thickening) were identified by eUBM, while colonoscopy identified just 76% of them. Colonoscopy identified all of the colonic tumors but failed to diagnose lymphoid infiltrates and increased mucosal thickness and failed to differentiate lymphoid infiltrates from small adenomas. During the observation period, most of the lesions (approximately 67%) increased in size, approximately 14% remained unchanged, and 19% regressed