Background: The percutaneous endoscopic gastrostomy (PEG) tube in patients with oesophageal cancer is controversial, owing to a perceived risk of tumour seeding at the PEG site, theoretical difficulty in formation of gastric conduit during oesophagectomy and a fear of increased post-operative complications, including anastomotic leak. We aimed to assess the impact of PEG tubes on nutritional status and post-operative complications in patients with oesophageal cancer who underwent PEG tube insertion prior to neo-adjuvant treatment. Methods: We performed a retrospective review of 800 patients with oesophageal or gastro oesophageal junction (GOJ) cancer, who underwent PEG insertion from June, 2010 to May, 2015. Out of these, 168 patients who fulfilled the inclusion criteria were analysed further. All of them were followed up for 3 years after treatment to assess overall survival. Thus, the follow up of the last patient included in the study was completed on May, 31, 2018. Results: The average body mass index (BMI) of patients was maintained following PEG tube, during neoadjuvant treatment (22.34±4.84 before PEG vs. 21.85±3.90 after PEG, P value: 0.1). Out of 168 patients, 33 (19.7%) developed a complication following PEG tube, with PEG site infection as the most common in 24 (14.2%). PEG-related mortality at 1 month was 0%. Ninety out of 168 patients (59%) underwent surgery after neo-adjuvant treatment. Three patients had tumour seeding at the PEG site and thus surgery could not be performed. Gastric conduit formation was possible in all 99 patients. Postoperative complications were seen in 17/99 (17%) patients, including surgical site infections in 7 (7.07%), anastomotic leak in 6 (6.06%) and anastomotic stricture in 4 (4.04%). Overall survival at 3 years was 87%. Conclusions: Pre-operative PEG tube in oesophageal cancer is safe and does not compromise the future anastomosis. Also, it helps in maintaining the nutritional status during neo-adjuvant treatment.

The role of self-expandable metallic stents is gradually evolving for a diverse group of benign and malignant gastrointestinal tract problems, with luminal obstruction being by far the most common. Although its role in refractory variceal bleeding is well established, it has rarely been tried for tumor-related bleeding, with only a few case reports in this regard. We share our experience of successfully controlling esophageal tumor–related bleeding with the use of a fully covered self-expandable metallic stent. A 58-year-old woman with irresectable distal esophageal cancer, presented with hematemesis. Esophago-gastro-duodenoscopy revealed an obstructing esophageal tumor with diffuse oozing of blood. This was unamenable to local injection of adrenaline and hemospray; therefore, a temporary self-expandable metallic stent was parked to create a tamponade effect. This successfully stopped bleeding and the patient remained asymptomatic till discharge. However, she was lost to follow-up, and therefore, the stent was removed after a period of 5 months instead of 2 weeks.

Histopathological image analysis is an examination of tissue under a light microscope for cancerous disease diagnosis. Computer-assisted diagnosis (CAD) systems work well by diagnosing cancer from histopathology images. However, stain variability in histopathology images is inevitable due to the use of different staining processes, operator ability, and scanner specifications. These stain variations present in histopathology images affect the accuracy of the CAD systems. Various stain normalization techniques have been developed to cope with inter-variability issues, allowing standardizing the appearance of images. However, in stain normalization, these methods rely on the single reference image rather than incorporate color distributions of the entire dataset. In this paper, we design a novel machine learning-based model that takes advantage of whole dataset distributions as well as color statistics of a single target image instead of relying only on a single target image. The proposed deep model, called stain acclimation generative adversarial network (SA-GAN), consists of one generator and two discriminators. The generator maps the input images from the source domain to the target domain. Among discriminators, the first discriminator forces the generated images to maintain the color patterns as of target domain. While second discriminator forces the generated images to preserve the structure contents as of source domain. The proposed model is trained using a color attribute metric, extracted from a selected template image. Therefore, the designed model not only learns dataset-specific staining properties but also image-specific textural contents. Evaluated results on four different histopathology datasets show the efficacy of SA-GAN to acclimate stain contents and enhance the quality of normalization by obtaining the highest values of performance metrics. Additionally, the proposed method is also evaluated for multiclass cancer type classification task, showing a 6.9% improvement in accuracy on ICIAR 2018 hidden test data.

No abstract

No abstract