Clinical Case Summary

Introduction: Leiomyosarcoma (LMS) is a rare mesenchymal tumor, affecting the gastrointestinal tract and subcutaneous soft tissues as the most common locations^1,2,3. Here, we report a case of a primary leiomyosarcoma of the liver, initially misdiagnosed as a hepatocellular carcinoma (HCC) and effectively treated with sorafenib.
Case report: A 69 years old Caucasian man with a HCV-related cirrhosis was hospitalized in May 2021 for an upper gastro-intestinal bleeding from esophageal varices. During the hospitalization, he received a radiological (arterial phase wash-in and portal and delayed venous phases washout) and biochemical (AFP 7,073 ng/mL) diagnosis of a monofocal HCC with neoplastic portal vein thrombosis. Treatment with sorafenib was started in August 2021, and at the first control CT scan a partial response according to mRECIST criteria was demonstrated, with a total normalization of AFP levels. Two months later, following the good response to systemic therapy, the patient underwent right hepatectomy. However, at the histologic evaluation a primary liver leiomyosarcoma was described. Since resection was R1, after this diagnosis, the patients started adjuvant treatment with Pazopanib. Unfortunately, the disease progressed despite various unsuccessful chemotherapy regimens and died in September 2022.
Conclusion: Liver LMS is a rare tumor, rarely taken into consideration in the differential diagnosis of a liver mass. In the case we report, the association with HCV-related cirrhosis, the high levels of AFP and the response to Sorafenib treatment would have hardly put the diagnosis of primary liver LMS in any differential diagnosis list. Here we therefore describe the first case, as far as we are aware, of liver primary LMS expressing high levels of AFP and successfully treated with sorafenib.

References

1. Serrano C, George S. Leiomyosarcoma. Hematol Oncol Clin North Am. 2013;27(5):957-974. doi:10.1016/j.hoc.2013.07.002
2. Metta H, Corti M, Trione N, et al. Primary hepatic leiomyosarcoma--a rare neoplasm in an adult patient with AIDS: second case report and literature review. J Gastrointest Cancer. 2014;45 Suppl 1:36-39. doi:10.1007/s12029-013-9525-3
3. Baek I, Kim JH, Lee MS, et al. Taehan Kan Hakhoe Chi. 2002;8(4):481-185.

Introduction

Detective flow imaging (DFI) is a novel technology that enables the evaluation of low-speed blood flow signal that is difficult to visualize with conventional Doppler techniques^1)-3). This technique allows for evaluating microvascular signals within tumors in endoscopic ultrasound (EUS) without using contrast agents. In this study, we aimed to investigate the usefulness of DFI for the diagnosis of biliary tract lesions.

Aims & Methods

A retrospective study was conducted of patients who underwent DFI with EUS between September 2019 and December 2022 at our institution. Patients with biliary tract lesions diagnosed by either endoscopy or pathology were included. Malignant biliary tract lesions were defined as those with histopathological diagnoses of non-invasive or invasive carcinoma. To evaluate the diagnostic ability of DFI for malignant biliary tract lesions, we assessed the presence of tortuous or variable caliber vessels in DFI findings. Inter-observer agreement and the diagnostic ability of DFI in identifying malignant biliary tract lesions were also assessed. Additionally, the DFI findings of intraductal papillary neoplasms (IPNB) and cholangiocarcinoma in patients who underwent surgical resection.

Results

There were 17 cases of cholangiocarcinoma, 9 cases of IPNB, 3 cases of bile duct stone, and 3 other cases. The final diagnosis of biliary tract lesion was made by surgery in 20 cases and by EUS-FNA or biopsy in 9 cases. In the first analysis, the sensitivity and specificity for the presence of tortuous or variable caliber vessels for the diagnosis of malignancy were 80% and 43% for tortuous vessels and 68% and 100% for variable caliber vessels, respectively. Interobserver agreement was k=0.63 for tortuous vessels and k=0.75 for variable caliber vessels. In the second analysis, DFI identified stalk vessels in 7 of 8 cases (87.5%) of IPNB and in only 1 of 10 cases (10%) of invasive cholangiocarcinoma. The rate of concomitant stalk vessels was significantly higher in IPNB than in invasive cholangiocarcinoma (P<0.01).

Conclusion

DFI is a simple and easy-to-use technique that allows for the evaluation of blood flow in biliary tract lesions without the use of contrast agents. It has the potential to be a useful tool for the differential diagnosis of malignant biliary tract lesions.

References

1. Yamashita Y, Yoshikawa T, Yamazaki H et al. A novel endoscopic ultrasonography imaging technique for depicting microcirculation in pancreatobiliary lesions without the need for contrast-enhancement: a prospective exploratory study. Diagnostics (Basel) 2021; 11: 2018.
2. Yamashita Y, Yoshikawa T, Kawaji Y et al. Novel endoscopic ultrasonography imaging technique for visualizing microcirculation without contrast enhancement in subepithelial lesions: prospective study. Dig Endosc 2021; 33: 955–961.
3. Miwa H, Sugimori K, Maeda S. Vessel images of gallbladder polypoid lesions on detective flow imaging endoscopic ultrasonography. Dig Endosc 2023. doi: 10.1111/den.14530 [Online ahead of print].

Introduction

Anastomotic leakage after foregut surgeries are often difficult to manage. The conventional way of endoscopic treatment is endoscopic stenting (ES). However, it carries a risk of stent migration and erosion. Endoscopic vacuum therapy (EVT) has recently been proposed as an alternative way of treating these leaks. EVT has the benefit of providing negative suction and promotion of granulation tissue¹. However, an exchange every 3-7 days has been suggested.

Aims & Methods

The aim of this study was to compare the treatment outcomes between ES, EVT and the combination of both. This was a retrospective multi-centre cohort study from 3 major centres in Hong Kong that performs esophagectomy and gastrectomy. All patients with anastomotic leak after total gastrectomy and esophagectomy and who were treated with ES or EVT between Jul 2008 and Apr 2023 from the Prince of Wales Hospital, Queen Mary Hospital and United Chistian Hospital were included. Patients’ demographics, type of surgery, size of anastomotic defect, treatment success rate, complications, number of procedures, duration to healing were recorded and compared. Continuous data are compared with Mann-Whitney test and categorical data are compared with Chi-square/ Fisher's exact test.

Results

A total of 63 patients were recruited (ES: 49; EVT 14. The outcomes of the study are listed in Table 1. The mean (S.D.) of defect was 15.4 (16.0)mm in the ES group and 17.8 (14.1)mm in the EVT group (p=0.39). in the EVT group, there are 3/14 patients who had failed EVT therapy and converted to ES. The success rate of each arm as a single arm therapy were similar (29/49(59.2%) in the ES group vs 6/14(42.9%) in the EVT group; p=0.28). However, there were less complications in the EVT group (2/14 (14.3%) in ES group vs. 23/49(46.9%) in EVT group; p=0.03). On the other hand, the EVT group required more endoscopic procedures (median (range) 3 1-137) procedures in the ES group vs 6(2-16) in the EVT group ; p 0.002). The days required for the anastomosis to heal were also similar (ES (61.7 (38.4) vs EVT 45.3 (25.8); p=0.37). When comparing patients treated as a single-arm therapy (ES or ET) to combination therapy (ES + EVT), the success rates has a trend towards favouring combination treatment (p=0.07).

Conclusion

Both EVT and ES are feasible and safe in treating anastomotic leaks after foregut surgeries. Although more procedures are required, EVT achieved similar success rate while having less complications when compared to ES a single arm. The use of combination therapy as compared to single-arm endoscopic therapy may achieve a higher treatment success rate.

References

1. Chan SM, Auyeung KKY, Lam SF, Chiu PWY, Teoh AYB. Current status in endoscopic management of upper gastrointestinal perforations, leaks and fistulas. Dig Endosc. 2022 Jan;34(1):43-62. doi: 10.1111/den.14061. Epub 2021 Jul 19. PMID: 34115407.

Disclosure

There is no COI.

Introduction

Collagenous colitis (CC) is a form of microscopic colitis (MC), the others being lymphocytic colitis (LC), incomplete CC and incomplete LC (1). It is an inflammatory disease of the large bowel that causes chronic watery diarrhoea, abdominal pain, faecal incontinence, nightly defecation, and weight loss, resulting in significantly impaired quality of life (2). Incidence of MC has increased significantly during the past decades in some countries and the main reason for this increase is thought to be an enhanced disease recognition (3). The diagnosis of MC is challenging as it can only be diagnosed upon histological examination of colonic biopsies taken from normal or near normal appearing mucosa (1). Histological interpretation of biopsies involves subjective evaluation leading to inter-rater variability discrepancies in diagnosis and treatment plan. Deep learning-based assistance system can objectivise diagnostic key feature selection leading to minimisation of inter-rater variability and improvement of diagnostic accuracy.

Aims & Methods

The aim of this pilot study was to develop an algorithm for robust segmentation of light microscopy images of histological specimen slides (tissue slides) emphasizing key diagnostic features of CC.
Histological specimens were fixed with formalin, embedded in paraffin, cut into 3 µm sections, and stained with haematoxylin and eosin (H&E) for histological examination. Images were taken using OLYMPUS IX71 light microscope (x20 magnification) equipped with Q IMAGING EXI aqua camera at (1392 x 1040 pix.) resolution and 24-bit pixel encoding. Registered image pixel values were normalized in RGB colour space by histogram alignment using empty tissue-less areas using an algorithm described previously (4). Superpixel technique using Simple Linear Iterative Clustering algorithm (5) was used for initial segmentation of the images. All 1000 superpixels defined in each image were annotated by the expert, indicating ones containing thickened subepithelial collagen layer as the sought class among the others containing the rest of the tissue indicated as normal class. The feed – forward neural network (implemented in MatLab environment) was trained to classify superpixels using their normalized histogram bin-counts of the pixel values as a feature. Due to significant inequality in the data set, where the number of sought class superpixels was significantly smaller than the number of the remaining ones, the training set was formed by equalising it to the lowest class case’s value (18761). Only randomly selected cases of the bigger class were included in the training set.

Results

The proposed algorithm was tested on images from 5 treatment-naive patients with newly diagnosed CC. Patients' histological specimens were rated by an expert pathologist and 50 images were chosen (with changes characteristic of CC and without). The classification training process with extracted data set took 0.13 seconds on personal computer with an AMD Ryzen Threadripper 3970X 32-Core, 3.70 GHz processor and 128GB of RAM memory, epochs for training was set to 30 (after tests the model was not improving after 30 epochs) and data shuffling was set to every-epoch. The algorithm showed 0.807 accuracy, 0.801 sensitivity and 0.813 specificity.

Conclusion

The shown ability of deep learning algorithm to classify histology image segments can assist the diagnostic process of CC emphasizing thickened subepithelial collagen layer – the essential diagnostic feature.

References

1. Miehlke S, Guagnozzi D, Zabana Y, Tontini GE, Fiehn AMK, Wildt S, et al. European guidelines on microscopic colitis: United European Gastroenterology (UEG) and European Microscopic Colitis Group (EMCG) statements and recommendations. United European Gastroenterol J. 2020;
2. Verhaegh BPM, Münch A, Guagnozzi D, Wildt S, Cebula W, Diac AR, et al. Course of Disease in Patients with Microscopic Colitis: A European Prospective Incident Cohort Study. J Crohns Colitis. 2021;15(7):1174–83.
3. Weimers P, Ankersen DV, Lophaven S, Bonderup OK, Münch A, Løkkegaard ECL, et al. Incidence and Prevalence of Microscopic Colitis between 2001 and 2016: A Danish Nationwide Cohort Study. J Crohns Colitis. 2021;14(12):1717–23.
4. Petrolis R., Čižas P., Borutaitė V., Kriščiukaitis A. Method of fluorescence imaging for evaluation of membrane potential in cultured neurons using transmembrane voltage sensitive dye. Biomedical engineering 2011: Proceedings of International Conference. 14:16-19, 2011.

5. Achanta R, Shaji A, Smith K, Lucchi A, Fua P, Susstrunk S. SLIC Superpixels Compared to State-of-the-art Superpixel Methods. IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 34, Issue 11, pp. 2274-2282, May 2012

Clinical Case Summary

A 43-year-old woman with a rectal 90mm granular laterally spreading tumor involving 50% of the circumference (distal limit at 12cm from the anal verge) underwent endoscopic submucosal dissection (ESD). Due to submucosal fibrosis, deep laceration of the muscularis propria occurred at the end of the procedure and two through-the-scope (TTS) clips were applied and broad-spectrum antibiotics were administered. In the following day, the patient developed sepsis and an emergency CT scan revealed free air in mesorectum and densification of the perirectal/presacral adipose tissues with contrast extravasation. Proctosigmoidoscopy showed an orifice of approximately 12mm, near the previous TTS clips, that communicated with a 15mm abscessed cavity, compatible with delayed perforation (Figure 1). Endoluminal vacuum therapy (EVT) was initiated with a 2cm Endo-SPONGE^® (B.Braun Medical B.V., Melsungen, Germany), partially intraluminal, without immediate complications (Figure 2). The patient was kept under parenteral nutritional, antibiotics and twice weekly EVT sponge replacement, with a total of four substitutions. The evolution was favorable with sepsis resolution, decrease of the cavity length and complete tissue granulation (Figure 3). The patient was discharged after 16 days and remained asymptomatic with complete closure at the 1.5 months follow up (Figure 4).

References

(Figure 1). (Figure 2). (Figure 3) (Figure 4).