Machine Learning-Based Prediction of Responsiveness to Neoadjuvant Chemoradiotheapy in Locally Advanced Rectal Cancer Patients from Endomicroscopy.

The protocol for treating locally advanced rectal cancer consists of the application of chemoradiotherapy (neoCRT) followed by surgical intervention. One issue for clinical oncologists is predicting the efficacy of neoCRT in order to adjust the dosage and avoid treatment toxicity in cases when surgery should be conducted promptly. Biomarkers may be used for this purpose along with in vivo cell-level images of the colorectal mucosa obtained by probe-based confocal laser endomicroscopy (pCLE) during colonoscopy. The aim of this article is to report our experience with Motiro, a computational framework that we developed for machine learning (ML) based analysis of pCLE videos for predicting neoCRT response in locally advanced rectal cancer patients. pCLE videos were collected from 47 patients who were diagnosed with locally advanced rectal cancer (T3/T4, or N+). The patients received neoCRT. Response to treatment by all patients was assessed by endoscopy along with biopsy and magnetic resonance imaging (MRI). Thirty-seven patients were classified as non-responsive to neoCRT because they presented a visible macroscopic neoplastic lesion, as confirmed by pCLE examination. Ten remaining patients were considered responsive to neoCRT because they presented lesions as a scar or small ulcer with negative biopsy, at post-treatment follow-up. Motiro was used for batch mode analysis of pCLE videos. It automatically characterized the tumoral region and its surroundings. That enabled classifying a patient as responsive or non-responsive to neoCRT based on pre-neoCRT pCLE videos. Motiro classified patients as responsive or non-responsive to neoCRT with an accuracy of ~ 0.62 when using images of the tumor. When using images of regions surrounding the tumor, it reached an accuracy of ~ 0.70. Feature analysis showed that spatial heterogeneity in fluorescence distribution within regions surrounding the tumor was the main contributor to predicting response to neoCRT. We developed a computational framework to predict response to neoCRT by locally advanced rectal cancer patients based on pCLE images acquired pre-neoCRT. We demonstrate that the analysis of the mucosa of the region surrounding the tumor provides stronger predictive power.

Endoscopic ultrasonography-guided gastroenterostomy versus uncovered duodenal metal stenting for unresectable malignant gastric outlet obstruction (DRA-GOO): a multicentre randomised controlled trial.

BACKGROUND: Endoscopic ultrasonography-guided gastroenterostomy (EUS-GE) is a novel endoscopic method to palliate malignant gastric outlet obstruction. We aimed to assess whether the use of EUS-GE with a double balloon occluder for malignant gastric outlet obstruction could reduce the need for reintervention within 6 months compared with conventional duodenal stenting. METHODS: The was an international, multicentre, randomised, controlled trial conducted at seven sites in Hong Kong, Belgium, Brazil, India, Italy, and Spain. Consecutive patients (aged ≥18 years) with malignant gastric outlet obstruction due to unresectable primary gastroduodenal or pancreatobiliary malignancies, a gastric outlet obstruction score (GOOS) of 0 (indicating an inability in intake food or liquids orally), and an Eastern Cooperative Oncology Group performance status score of 3 or lower were included and randomly allocated (1:1) to receive either EUS-GE or duodenal stenting. The primary outcome was the 6-month reintervention rate, defined as the percentage of patients requiring additional endoscopic intervention due to stent dysfunction (ie, restenosis of the stent due to tumour ingrowth, tumour overgrowth, or food residue; stent migration; or stent fracture) within 6 months, analysed in the intention-to-treat population. Prespecified secondary outcomes were technical success (successful placement of a stent), clinical success (1-point improvement in gastric outlet obstruction score [GOOS] within 3 days), adverse events within 30 days, death within 30 days, duration of stent patency, GOOS at 1 month, and quality-of-life scores. This study is registered with ClinicalTrials.gov (NCT03823690) and is completed. FINDINGS: Between Dec 1, 2020, and Feb 28, 2022, 185 patients were screened and 97 (46 men and 51 women) were recruited and randomly allocated (48 to the EUS-GE group and 49 to the duodenal stent group). Mean age was 69·5 years (SD 12·6) in the EUS-GE group and 64·8 years (13·0) in the duodenal stent group. All randomly allocated patients completed follow-up and were analysed. Reintervention within 6 months was required in two (4%) patients in the EUS-GE group and 14 (29%) in the duodenal stent group [p=0·0020; risk ratio 0·15 [95% CI 0·04-0·61]). Stent patency was longer in the EUS-GE group (median not reached in either group; HR 0·13 [95% CI 0·08-0·22], log-rank p<0·0001). 1-month GOOS was significantly better in the EUS-GE group (mean 2·41 [SD 0·7]) than the duodenal stent group (1·91 [0·9], p=0·012). There were no statistically significant differences between the EUS-GE and duodenal stent groups in death within 30 days (ten [21%] vs six [12%] patients, respectively, p=0·286), technical success, clinical success, or quality-of-life scores at 1 month. Adverse events occurred 11 (23%) patients in the EUS-GE group and 12 (24%) in the duodenal stent group within 30 days (p=1·00); three cases of pneumonia (two in the EUS-GE group and one in the duodenal stent group) were considered to be procedure related. INTERPRETATION: In patients with malignant gastric outlet obstruction, EUS-GE can reduce the frequency of reintervention, improve stent patency, and result in better patient-reported eating habits compared with duodenal stenting, and the procedure should be used preferentially over duodenal stenting when expertise and required devices are available. FUNDING: Research Grants Council (Hong Kong Special Administrative Region, China) and Sociedad Española de Endoscopia Digestiva.

Comparing three-dimensional endorectal ultrasound and magnification chromoendoscopy for early rectal neoplasia invasion depth assessment.

INTRODUCTION: Accurate assessment of invasion depth of early rectal neoplasms is essential for optimal therapy. We aimed to compare three-dimensional endorectal ultrasound (3D-ERUS) with magnification chromoendoscopy (MCE) regarding their accuracy in assessing parietal invasion depth (T). METHODS: Patients with middle and distal rectum neoplasms were prospectively included. Two providers blinded to each other's assessment performed 3D-ERUS and MCE, respectively. The T stage assessed through ERUS was compared to the MCE evaluation. The results were compared to the surgical specimen anatomopathological report. Sensitivity, specificity, accuracy, positive (PPV), and negative (NPV) predictive values were calculated for the T stage and for the final therapy (local excision or radical surgery). RESULTS: In 8 years, 70 patients were enrolled, and all underwent both exams. MCE and ERUS showed an accuracy of 94.3% and 85.7%, sensitivity of 83.7 and 93.3%, specificity of 96.4 and 83.6%, PPV of 86.7 and 60.9%, and NPV of 96.4 and 97.9%, respectively. Kappa for T stage assessed through ERUS was 0.64 and 0.83 for MCE. CONCLUSION: MCE and 3D-ERUS had good diagnostic performance, but the endoscopic method had higher accuracy. Both methods reliably assessed lesion extension, circumferential involvement, and distance from the anal verge.

ENDOSCOPIC SUBMUCOSAL DISSECTION WITH VERSUS WITHOUT TRACTION FOR PATHOLOGICALLY STAGED T1B ESOPHAGEAL CANCER: A MULTICENTER RETROSPECTIVE STUDY.

BACKGROUND & AIMS: Positive vertical margins (VM) are common after endoscopic submucosal dissection (ESD) of T1b esophageal cancer (EC) and are associated with an increased risk of recurrence. Traction during ESD provides better exposure of the submucosa and may allow deeper dissection, potentially reducing the risk of positive VMs. We conducted a retrospective multicenter study to compare the proportion of resections with positive VMs in ESD performed with vs. without traction in pathologically staged T1b EC. METHODS: Patients who underwent ESD revealing T1b EC (squamous or adenocarcinoma) at 10 academic tertiary referral centers in the United States (n = 9) and Brazil (n = 1) were included. Demographic and clinical data were abstracted. ESD using either traction techniques (tunneling, pocket) or traction devices (clip line, traction wire) were classified as ESD with traction (Tr-ESD) and those without were classified as ESD without traction (cESD). The primary outcome was a negative VM. Multivariable logistic regression was used to assess associations with negative VM. RESULTS: 166 patients with pathologically staged T1b EC underwent Tr-ESD (n=63, 38%) or cESD (n=103, 62%). Baseline factors were comparable between both groups. On multivariable analysis, Tr-ESD was found to be independently associated with negative VM (OR 2.25, 95% CI 1.06, 4.91, P=0.037) and R0 resection (OR 2.83, 95% CI 1.33, 6.23 P=0.008). CONCLUSION: ESD with traction seems to be associated with higher odds of negative VM than ESD without traction for pathologically staged T1b EC, and future well-conducted prospective studies are warranted to establish the findings of the current study.

Underwater versus conventional endoscopic mucosal resection for colorectal lesions: a systematic review and meta-analysis of randomized clinical trials.

BACKGROUND & AIMS: Conventional endoscopic mucosal resection (CEMR) is the established method for the resection of non-pedunculated colorectal lesions (NPCRL) ≥ 10 mm. In the last decade, underwater endoscopic mucosal resection (UEMR) has been introduced as a potential alternative. The aim of this systematic review with meta-analysis is to compare the recurrence and safety of UEMR and CEMR by analyzing only randomized controlled trials (RCTs). METHODS: We systematically searched PubMed, Cochrane Library and EMBASE until April 2023. Studies met the following inclusion criteria: (1) RCTs, (2) comparing UEMR with CEMR, (3) NPCRL ≥ 10 mm, and (4) reporting the outcomes of interest. Primary outcomes were recurrence and safety. Secondary outcomes were en bloc, R0, complete resection, clipping and adverse events per type. RESULTS: Five RCTs were included. UEMR was associated with a lower recurrence rate (OR: 0.56; 95% CI: 0.32-0.97). Thus, the RR of recurrence was 1.7 times higher in the CEMR group (95% CI, 1.04-2.77). There was no significant difference in the pooled safety analysis. UEMR showed better en bloc resection rates (OR: 1.54; 95% CI: 1.15-2.07), but subgroup analysis showed comparable rates in lesions ≥ 20 mm. R0 resection was higher in UEMR (OR: 1.72; 95% CI: 1.23-2.41). Other outcomes were not different between the 2 groups. CONCLUSIONS: UEMR is as safe as CEMR, with a higher overall R0 rate and a higher en bloc resection rate for lesions < 20 mm, leading to a lower overall recurrence rate. The results of this meta-analysis support the widespread use of UEMR.

Self-expandable metallic stent-induced esophagorespiratory fistulas in patients with advanced esophageal cancer.

BACKGROUND/AIMS: Self-expandable metallic stents (SEMSs) are widely adopted for the palliation of dysphagia in patients with malignant esophageal strictures. An important adverse event is the development of SEMS-induced esophagorespiratory fistulas (SEMS-ERFs). This study aimed to assess the risk factors related to the development of SEMS-ERF after SEMS placement in patients with esophageal cancer. METHODS: This retrospective study was performed at the Instituto do Cancer do Estado de São Paulo. All patients with malignant esophageal strictures who underwent esophageal SEMS placement between 2009 and 2019 were included in the study. RESULTS: Of the 335 patients, 37 (11.0%) developed SEMS-ERF, with a median time of 129 days after SEMS placement. Stent flare of 28 mm (hazard ratio [HR], 2.05; 95% confidence interval [CI], 1.15-5.51; p=0.02) and post-stent chemotherapy (HR, 2.0; 95% CI, 1.01-4.00; p=0.05) were associated with an increased risk of developing SEMS-ERF, while lower-third tumors were a protective factor (HR, 0.5; 95% CI, 0.26-0.85; p=0.01). No difference was observed in overall survival. CONCLUSION: The incidence of SEMS-ERFs was 11%, with a median time of 129 days after SEMS placement. Post-stent chemotherapy and a 28 mm stent flare were associated with a higher risk of SEMS-ERF.

The Role of Probe-Based Confocal Laser Endomicroscopy (pCLE) in the Diagnosis of Sustained Clinical Complete Response Under Watch-and-Wait Strategy After Neoadjuvant Chemoradiotherapy for Locally Advanced Rectal Adenocarcinoma: a Score Validation.

BACKGROUND: Watch-and-wait strategy has been increasingly accepted for patients with clinical complete response (cCR) after multimodal treatment for locally advanced rectal adenocarcinoma. Close follow-up is essential to the early detection of local regrowth. It was previously demonstrated that probe-based confocal laser endomicroscopy (pCLE) scoring using the combination of epithelial and vascular features might improve the diagnostic accuracy of cCR. AIM: To validate the pCLE scoring system in the assessment of patients with cCR after neoadjuvant chemoradiotherapy (nCRxt) for advanced rectal adenocarcinoma. METHODS: Digital rectal examination, pelvic magnetic resonance imaging (MRI), and pCLE were performed in 43 patients with cCR, who presented either a scar (N = 33; 76.7%) or a small ulcer with no signs of tumor, and/or biopsy negative for malignancy (N = 10; 23.3%). RESULTS: Twenty-five (58.1%) patients were men, and the mean age was 58.4 years. During the follow-up, 12/43 (27.9%) patients presented local regrowth and underwent salvage surgery. There was an association between pCLE diagnostic scoring and final histological report (for patients who underwent surgical resection) or final diagnosis at the latest follow-up (p = 0.0001), while this association was not observed with MRI (p = 0.49). pCLE sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 66.7%, 93.5%, 80%, 88.9%, and 86%, respectively. MRI sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 66.7%, 48.4%, 66.7%, 78.9%, and 53.5%, respectively. CONCLUSIONS: pCLE scoring system based on epithelial and vascular features improved the diagnosis of sustained cCR and might be recommended during follow-up. pCLE might add some valuable contribution for identifying local regrowth. Trial Registration This protocol was registered at the Clinical Trials (ClinicalTrials.gov identifier NCT02284802).

Impact of radiotherapy on adverse events of self-expanding metallic stents in patients with esophageal cancer.

Self-expanding metallic stents (SEMS) are considered the treatment of choice for the palliation of dysphagia and fistulas in inoperable esophageal neoplasms. However, the safety of SEMSs in patients who received or who will be submitted to radiotherapy (RT) is uncertain. The study aimed to evaluate the impact of RT on adverse events (AEs) in patients with esophageal cancer with SEMSs. This is a retrospective study conducted at a tertiary cancer hospital from 2009 to 2018. We collected information regarding RT, the histological type of the tumor, the model of SEMSs and AEs after stent placement. Three hundred twenty-three patients with malignant stenosis or fistula were treated with SEMSs. The predominant histological type was squamous cell carcinoma (79.6%). A total of 282 partially covered and 41 fully covered SEMSs were inserted. Of the 323 patients, 182 did not received RT, 118 received RT before SEMS placement and 23 after. Comparing the group that received RT before stent insertion with the group that did not, the first one presented a higher frequency of severe pain (9/118 7.6% vs. 3/182 1.6%; P = 0.02). The group treated with RT after stent placement had a higher risk of global AEs (13/23 56.5% vs. 63/182 34.6%; P = 0.019), ingrowth/overgrowth (6/23 26.1% vs. 21/182 11.5%; P = 0.045) and gastroesophageal reflux (2/23 8.7% vs. 2/182 1.1%; P = 0.034). Treatment with RT before stent placement in patients with inoperable esophageal neoplasm prolongs survival and is associated with an increased risk of severe chest pain. Treatment with RT of patients with an esophageal stent increases the frequency of minor, not life-threatening AEs.

Results of endoscopic biliary drainage in patients with malignant hilar stricture.

In Malignant Hilar Biliary Stricture (MHBS) palliative biliary drainage is a frequent strategy, improving the quality of life, reducing pruritus, loss of appetite and relieving cholangitis. The endoscopic approach is an effective, although challenging procedure. This study aimed to evaluate technical and clinical success rates of biliary drainage by ERCP. This is a retrospective study including all patients with MHBS referred to Instituto do Cancer do Hospital de São Paulo (ICESP) submitted to biliary drainage by ERCP, between January 2010 and December 2017. Multivariable logistic regression was performed to evaluate predictors of clinical failure, as total bilirubin levels, Bismuth classification, number of hepatic sectors drained and presence of cholangitis. In total, 82 patients presenting unresectable MHBS were included in this study. 58.5% female and 41.5% male, with a mean age of 60±13 years. Bismuth classification grades II, IIIA, IIIB and IV were noted in 23.2%, 15.9%, 14.6% and 46.3%, respectively. Technical and clinical success was achieved in 92.7% and 53.7% respectively. At multivariable logistic-regression analyses, Bismuth IV strictures were related to higher clinical failure rates when compared to other strictures levels, with an Odds Ratio of 5.8 (95% CI 1.28‒20.88). In conclusion, endoscopic biliary drainage for malignant hilar biliary stricture had a high technical success but suboptimal clinical success rate. Proximal strictures (Bismuth IV) were associated with poor drainage outcomes.

Impact of a Routine Colorectal Endoscopic Submucosal Dissection in the Surgical Management of Nonmalignant Colorectal Lesions Treated in a Referral Cancer Center.

BACKGROUND: Recent data show an increasing number of abdominal surgeries being performed for the treatment of nonmalignant colorectal polyps in the West but in settings in which colorectal endoscopic submucosal dissection is not routinely performed. This study evaluated the number of nonmalignant colorectal lesions referred to surgical treatment in a tertiary cancer center that incorporated magnification chromoendoscopy and endoscopic submucosal dissection as part of the standard management of complex colorectal polyps. OBJECTIVE: The study aimed to estimate the number of patients with nonmalignant colorectal lesions referred to surgical resection at our institution after the standardization of routine endoscopic submucosal dissection and to describe outcomes for patients undergoing colorectal endoscopic submucosal dissection. DESIGN: Single-center retrospective study from a prospectively collected database of endoscopic submucosal dissections and colorectal surgeries performed between January 2016 and December 2019. SETTING: Reference cancer center. PATIENTS: Consecutive adult patients with complex nonmalignant colorectal polyps were included. INTERVENTIONS: Patients with nonmalignant colorectal polyps were treated by endoscopic submucosal dissection or surgery (elective colectomy, rectosigmoidectomy, low anterior resection, or proctocolectomy). MAIN OUTCOMES MEASURES: The primary outcome measure was the percentage of patients referred to colorectal surgery for nonmalignant lesions. RESULTS: In the study period, 1.1% of 825 colorectal surgeries were performed for nonmalignant lesions, and 97 complex polyps were endoscopically removed by endoscopic submucosal dissection. The en bloc, R0, and curative resection rates of endoscopic submucosal dissection were 91.7%, 83.5%, and 81.4%, respectively. The mean tumor size was 59 (SD 37.8) mm. Perforations during endoscopic submucosal dissection occurred in 3 cases, all treated with clipping. One patient presented with a delayed perforation 2 days after the endoscopic resection and underwent surgery. The mean follow-up period was 3 years, with no tumor recurrence in this cohort. LIMITATIONS: Single-center retrospective study. CONCLUSIONS: A workflow that includes assessment of the lesions with magnification chromoendoscopy and resection through endoscopic submucosal dissection can lead to a very low rate of abdominal surgery for nonmalignant colorectal lesions. See Video Abstract at http://links.lww.com/DCR/C123 . IMPACTO DE LA DISECCIN SUBMUCOSA ENDOSCPICA COLORRECTAL DE RUTINA EN EL MANEJO QUIRRGICO DE LESIONES COLORRECTALES NO MALIGNAS TRATADAS EN UN CENTRO ONCOLGICO DE REFERENCIA: ANTECEDENTES:Datos recientes muestran un número cada vez mayor de cirugías abdominales realizadas para el tratamiento de pólipos colorrectales no malignos en Occidente, pero no en los entornos donde la disección submucosa endoscópica colorrectal se realiza de forma rutinaria. El estudio evaluó el número de lesiones colorrectales no malignas referidas a tratamiento quirúrgico en un centro oncológico terciario, que incorporó cromoendoscopia de aumento y disección submucosa endoscópica como parte del manejo estándar de pólipos colorrectales complejos.OBJETIVO:Estimar el número de pacientes con lesiones colorrectales no malignas referidos para resección quirúrgica en nuestra institución, después de la estandarización de la disección submucosa endoscópica de rutina y describir los resultados para los pacientes sometidos a disección submucosa endoscópica colorrectal.DISEÑO:Estudio retrospectivo de un solo centro, a partir de una base de datos recolectada prospectivamente de disecciones submucosas endoscópicas y cirugías colorrectales realizadas entre enero de 2016 y diciembre de 2019.AJUSTE:Centro oncológico de referencia.PACIENTES:Pacientes adultos consecutivos con pólipos colorrectales no malignos complejos.INTERVENCIONES:Pacientes con pólipos colorrectales no malignos tratados mediante disección submucosa endoscópica o cirugía (colectomía electiva, rectosigmoidectomía, resección anterior baja o proctocolectomía).PRINCIPALES MEDIDAS DE RESULTADO:La medida de resultado primario fue el porcentaje de pacientes remitidos a cirugía colorrectal por lesiones no malignas.RESULTADOS:En el período, 1,1% de 825 cirugías colorrectales fueron realizadas por lesiones no malignas y 97 pólipos complejos fueron extirpados por. disección submucosa endoscópica. Las tasas de resección en bloque, R0 y curativa de disección submucosa endoscópica fueron 91,7%, 83,5% y 81,4%, respectivamente. El tamaño tumoral medio fue de 59 (DE 37,8) mm. Se produjeron perforaciones durante la disección submucosa endoscópica en 3 casos, todos tratados con clipaje. Un paciente presentó una perforación diferida 2 días después de la resección endoscópica y fue intervenido quirúrgicamente. El seguimiento medio fue de 3 años, sin recurrencia tumoral en esta cohorte.LIMITACIONES:Estudio retrospectivo de un solo centro.CONCLUSIONES:Un flujo de trabajo que incluye la evaluación de las lesiones con cromoendoscopia de aumento y resección a través de disección submucosa endoscópica, puede conducir a una tasa muy baja de cirugía abdominal para lesiones colorrectales no malignas. Consulte Video Resumen en http://links.lww.com/DCR/C123 . (Traducción-Dr. Fidel Ruiz Healy ).

Underwater versus conventional EMR for nonpedunculated colorectal lesions: a randomized clinical trial.

BACKGROUND AND AIMS: Conventional endoscopic mucosal resection (CEMR) is the standard modality for removing nonpedunculated colorectal lesions. Underwater endoscopic mucosal resection (UEMR) has emerged as an alternative method. There are few comparative studies between these techniques, especially evaluating recurrence. Therefore, the purpose of this trial was to compare CEMR and UEMR for the resection of colorectal lesions with respect to efficacy, safety, and recurrence rate. METHODS: This was a randomized controlled trial of UEMR versus CEMR for naïve and nonpedunculated lesions measuring between 10 and 40 mm. The primary outcome was adenoma recurrence at 6 months after the resection. Secondary outcomes were rates of technical success, en bloc resection, and adverse events. Block randomization was used to assign patients. Tattooing was performed to facilitate localization of the scars and eventual recurrences. Endoscopic follow-up was scheduled at 6 months after the procedure. The sites of resections were examined with white-light imaging, narrow-band imaging (NBI), and conventional chromoscopy with indigo carmine followed by biopsies. RESULTS: One hundred five patients with 120 lesions were included, with a mean size of 17.5 ± 7.1 (SD) mm. Sixty-one lesions were resected by UEMR and 59 by CEMR. The groups were similar at baseline regarding age, sex, average size, and histologic type. Lesions in the proximal colon in the CEMR group corresponded to 83% and in the UEMR group to 67.8% (P = .073). There was no difference between groups regarding success rate (1 failure in each group) and en bloc resection rate (60.6% UEMR vs 54.2% CEMR, P = .48). Intraprocedural bleeding was observed in 5 CEMRs (8.5%) and 2 UEMRs (3.3%) (P = .27). There was no perforation or delayed hemorrhage in either groups. Recurrence rate was higher in the CEMR arm (15%) than in the UEMR arm (2%) (P = .031). Therefore, the relative risk of 6-month recurrence rate in the CEMR group was 7.5-fold higher (95% CI, 0.98-58.20), with a number needed to treat of 7.7 (95% CI, 40.33-4.22). The higher recurrence rate in the CEMR group persisted only for lesions measuring 21 to 40 mm (35.7% vs 0%; P = .04). CONCLUSION: This study demonstrated that UEMR was associated with a lower adenoma recurrence rate than was CEMR. Both endoscopic techniques were effective and had similar rates of adverse events for the treatment of nonpedunculated colorectal lesions.

Results of endoscopic biliary drainage in patients with malignant hilar stricture

Abstract In Malignant Hilar Biliary Stricture (MHBS) palliative biliary drainage is a frequent strategy, improving the quality of life, reducing pruritus, loss of appetite and relieving cholangitis. The endoscopic approach is an effective, although challenging procedure. This study aimed to evaluate technical and clinical success rates of biliary drainage by ERCP. This is a retrospective study including all patients with MHBS referred to Instituto do Cancer do Hospital de São Paulo (ICESP) submitted to biliary drainage by ERCP, between January 2010 and December 2017. Multivariable logistic regression was performed to evaluate predictors of clinical failure, as total bilirubin levels, Bismuth classification, number of hepatic sectors drained and presence of cholangitis. In total, 82 patients presenting unresectable MHBS were included in this study. 58.5% female and 41.5% male, with a mean age of 60±13 years. Bismuth classification grades II, IIIA, IIIB and IV were noted in 23.2%, 15.9%, 14.6% and 46.3%, respectively. Technical and clinical success was achieved in 92.7% and 53.7% respectively. At multivariable logistic-regression analyses, Bismuth IV strictures were related to higher clinical failure rates when compared to other strictures levels, with an Odds Ratio of 5.8 (95% CI 1.28‒20.88). In conclusion, endoscopic biliary drainage for malignant hilar biliary stricture had a high technical success but suboptimal clinical success rate. Proximal strictures (Bismuth IV) were associated with poor drainage outcomes.

TC-325 hemostatic powder in the management of upper gastrointestinal malignant bleeding: a randomized controlled trial.

Background and study aims Upper gastrointestinal bleeding (UGIB) from malignancy is associated with high rebleeding and mortality rates. Recently, TC-325 powder has shown promising results in the treatment of UGIB, including malignant bleeding. The aim of this study was to compare the efficacy of TC-325 versus best clinical management. Patients and methods From August 2016 to February 2020, all patients with evidence of UGIB from malignancy were randomized to receive TC-325 therapy or control group, in which endoscopic treatment was not mandatory. Exclusion criteria were hemoglobin drop without overt bleeding and UGIB from non-tumor origin. The primary outcome was 30-day mortality. Secondary outcomes were 30-day rebleeding, blood transfusion and length of hospital stay. Results Sixty-two patients were randomized, three were excluded and 59 were included in the final analysis (TC-325 group = 28; control = 31). Groups were similar at baseline. Active bleeding was observed in 22 patients in the TC-325 group and 19 in the control group ( P  = 0.15). Successful initial hemostasis with TC-325 was achieved in all cases. Additional therapy (radiotherapy, surgery or arterial embolization) was equally performed in both groups (42.9 % vs 58.1 %; P  = 0.243). There were no differences in 30-day mortality (28.6 % vs. 19.4 %, P  = 0.406) or 30-day rebleeding rates (32.1 % vs. 19.4 %, P  = 0.26). Logistic regression identified no significant predictors of rebleeding. Age, Eastern Cooperative Oncology Group (ECOG) score 3 to 4 and AIMS65 score > 1 predicted greater mortality. Conclusions TC-325 was effective in achieving immediate hemostasis in malignant gastrointestinal bleeding but did not reduce 30-day mortality, 30-day rebleeding, blood transfusion or length of hospital stay. Age, ECOG 3-4, and AIMS65 > 1 were predictive factors of mortality.