Region-specific deep learning models for accurate segmentation of rectal structures on post-chemoradiation T2w MRI: a multi-institutional, multi-reader study.

Introduction: For locally advanced rectal cancers, in vivo radiological evaluation of tumor extent and regression after neoadjuvant therapy involves implicit visual identification of rectal structures on magnetic resonance imaging (MRI). Additionally, newer image-based, computational approaches (e.g., radiomics) require more detailed and precise annotations of regions such as the outer rectal wall, lumen, and perirectal fat. Manual annotations of these regions, however, are highly laborious and time-consuming as well as subject to inter-reader variability due to tissue boundaries being obscured by treatment-related changes (e.g., fibrosis, edema). Methods: This study presents the application of U-Net deep learning models that have been uniquely developed with region-specific context to automatically segment each of the outer rectal wall, lumen, and perirectal fat regions on post-treatment, T2-weighted MRI scans. Results: In multi-institutional evaluation, region-specific U-Nets (wall Dice = 0.920, lumen Dice = 0.895) were found to perform comparably to multiple readers (wall inter-reader Dice = 0.946, lumen inter-reader Dice = 0.873). Additionally, when compared to a multi-class U-Net, region-specific U-Nets yielded an average 20% improvement in Dice scores for segmenting each of the wall, lumen, and fat; even when tested on T2-weighted MRI scans that exhibited poorer image quality, or from a different plane, or were accrued from an external institution. Discussion: Developing deep learning segmentation models with region-specific context may thus enable highly accurate, detailed annotations for multiple rectal structures on post-chemoradiation T2-weighted MRI scans, which is critical for improving evaluation of tumor extent in vivo and building accurate image-based analytic tools for rectal cancers.

Radiomic Texture and Shape Descriptors of the Rectal Environment on Post-Chemoradiation T2-Weighted MRI are Associated with Pathologic Tumor Stage Regression in Rectal Cancers: A Retrospective, Multi-Institution Study.

(1) Background: The relatively poor expert restaging accuracy of MRI in rectal cancer after neoadjuvant chemoradiation may be due to the difficulties in visual assessment of residual tumor on post-treatment MRI. In order to capture underlying tissue alterations and morphologic changes in rectal structures occurring due to the treatment, we hypothesized that radiomics texture and shape descriptors of the rectal environment (e.g., wall, lumen) on post-chemoradiation T2-weighted (T2w) MRI may be associated with tumor regression after neoadjuvant chemoradiation therapy (nCRT). (2) Methods: A total of 94 rectal cancer patients were retrospectively identified from three collaborating institutions, for whom a 1.5 or 3T T2w MRI was available after nCRT and prior to surgical resection. The rectal wall and the lumen were annotated by an expert radiologist on all MRIs, based on which 191 texture descriptors and 198 shape descriptors were extracted for each patient. (3) Results: Top-ranked features associated with pathologic tumor-stage regression were identified via cross-validation on a discovery set (n = 52, 1 institution) and evaluated via discriminant analysis in hold-out validation (n = 42, 2 institutions). The best performing features for distinguishing low (ypT0-2) and high (ypT3-4) pathologic tumor stages after nCRT comprised directional gradient texture expression and morphologic shape differences in the entire rectal wall and lumen. Not only were these radiomic features found to be resilient to variations in magnetic field strength and expert segmentations, a quadratic discriminant model combining them yielded consistent performance across multiple institutions (hold-out AUC of 0.73). (4) Conclusions: Radiomic texture and shape descriptors of the rectal wall from post-treatment T2w MRIs may be associated with low and high pathologic tumor stage after neoadjuvant chemoradiation therapy and generalized across variations between scanners and institutions.

Radiomic Features of Primary Rectal Cancers on Baseline T2 -Weighted MRI Are Associated With Pathologic Complete Response to Neoadjuvant Chemoradiation: A Multisite Study.

BACKGROUND: Twenty-five percent of rectal adenocarcinoma patients achieve pathologic complete response (pCR) to neoadjuvant chemoradiation and could avoid proctectomy. However, pretreatment clinical or imaging markers are lacking in predicting response to chemoradiation. Radiomic texture features from MRI have recently been associated with therapeutic response in other cancers. PURPOSE: To construct a radiomics texture model based on pretreatment MRI for identifying patients who will achieve pCR to neoadjuvant chemoradiation in rectal cancer, including validation across multiple scanners and sites. STUDY TYPE: Retrospective. SUBJECTS: In all, 104 rectal cancer patients staged with MRI prior to long-course chemoradiation followed by proctectomy; curated from three institutions. FIELD STRENGTH/SEQUENCE: 1.5T-3.0T, axial higher resolution T2 -weighted turbo spin echo sequence. ASSESSMENT: Pathologic response was graded on postsurgical specimens. In total, 764 radiomic features were extracted from single-slice sections of rectal tumors on processed pretreatment T2 -weighted MRI. STATISTICAL TESTS: Three feature selection schemes were compared for identifying radiomic texture descriptors associated with pCR via a discovery cohort (one site, N = 60, cross-validation). The top-selected radiomic texture features were used to train and validate a random forest classifier model for pretreatment identification of pCR (two external sites, N = 44). Model performance was evaluated via area under the curve (AUC), accuracy, sensitivity, and specificity. RESULTS: Laws kernel responses and gradient organization features were most associated with pCR (P ≤ 0.01); as well as being commonly identified across all feature selection schemes. The radiomics model yielded a discovery AUC of 0.699 ± 0.076 and a hold-out validation AUC of 0.712 with 70.5% accuracy (70.0% sensitivity, 70.6% specificity) in identifying pCR. Radiomic texture features were resilient to variations in magnetic field strength as well as being consistent between two different expert annotations. Univariate analysis revealed no significant associations of baseline clinicopathologic or MRI findings with pCR (P = 0.07-0.96). DATA CONCLUSION: Radiomic texture features from pretreatment MRIs may enable early identification of potential pCR to neoadjuvant chemoradiation, as well as generalize across sites. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2.

High-resolution anoscopy or expectant management for anal intraepithelial neoplasia for the prevention of anal cancer: is there really a difference?

BACKGROUND: High-resolution anoscopy has been shown to improve identification of anal intraepithelial neoplasia but a reduction in progression to anal squamous-cell cancer has not been substantiated when serial high-resolution anoscopy is compared with traditional expectant management. OBJECTIVE: The aim of this study was to compare high-resolution anoscopy versus expectant management for the surveillance of anal intraepithelial neoplasia and the prevention of anal cancer. DESIGN: This is a retrospective review of all patients who presented with anal squamous dysplasia, positive anal Pap smears, or anal squamous-cell cancer from 2007 to 2013. SETTING: This study was performed in the colorectal department of a university-affiliated, tertiary care hospital. PATIENTS: Included patients had biopsy-proven anal intraepithelial neoplasia from 2007 to 2013. INTERVENTIONS: Patients were treated with high-resolution anoscopy with ablation or standard anoscopy with ablation. Both groups were treated with imiquimod and followed every 6 months indefinitely. MAIN OUTCOME MEASURES: The incidence of anal squamous-cell cancer in each group was the primary end point. RESULTS: From 2007 to 2013, 424 patients with anal squamous dysplasia were seen in the clinic (high-resolution anoscopy, 220; expectant management, 204). Three patients (high-resolution anoscopy, 1; expectant management, 2) progressed to anal squamous-cell cancer; 2 were noncompliant with follow-up and with HIV treatment, and the third was allergic to imiquimod and refused to take topical 5-fluorouracil. The 5-year progression rate was 6.0% (95% CI, 1.5-24.6) for expectant management and 4.5% (95% CI, 0.7-30.8) for high-resolution anoscopy (p = 0.37). LIMITATIONS: This was a retrospective review. There is potential for selection and referral bias. Because of the rarity of the outcome, the study may be underpowered. CONCLUSIONS: Patients with squamous-cell dysplasia followed with expectant management or high-resolution anoscopy rarely develop squamous-cell cancer if they are compliant with the protocol. The cost, morbidity, and value of high-resolution anoscopy should be further evaluated in lieu of these findings.

Evaluation of inflammatory markers as predictors of hospital stay and unplanned readmission after colorectal surgery.

UNLABELLED: Hospital length of stay (LOS) and readmissions continue to be expensive and unexpected events following colorectal surgery (CRS) whether patients follow enhanced recovery pathways or traditional care. Predictors of these adverse events could facilitate identification and optimization of CRS patients. THE AIM OF THE STUDY: To examine the impact of white blood cell count (WBC) and C-reactive protein (CRP) levels as predictors of delayed recovery or hospital readmission following CRS. MATERIAL AND METHODS: Patients undergoing laparoscopic or open abdominal colorectal surgery by a single surgeon were managed using standardized enhanced recovery pathways. Those with postoperative day 2 CRP and white blood cell values were evaluated. Outcomes included 30-day hospital readmission rates and postoperative length of hospital stay. RESULTS: CRP values were available for 193 patients (86 Male, mean age 58.6 years). Ninety-nine patients had surgery for colon cancer, 23 for Crohn's disease, 19 for ulcerative colitis, 31 for diverticulitis and 18 for other reasons. Twenty patients (10.4%) were readmitted to the hospital within 30 days of surgery. POD2 CRP accurately predicted short length of hospital stay (p< 0.01). Average CRP was 6.3 in the LOS of < 3 days or less, and 11.7 in patients with LOS >4 days. The mean CRP of the readmission and non-readmission groups was 11.8 and 9.9, respectively (p=0.29). The average POD 2 WBC of the readmission and non-readmission groups was 10.6 and 9 respectively (p=0.01). CONCLUSION: A low POD2 CRP level was correlated with a shorter LOS, but it did not predict readmission. Conversely, POD2 WBC, and the difference in WBC from baseline were associated with readmission. These markers may be useful indicators to predict suitability of early discharge in an ERP. Further evaluation in prospective trials is warranted.

Tract length predicts successful closure with anal fistula plug in cryptoglandular fistulas.

PURPOSE: Collagen anal fistula plug treatment of transsphincteric fistulas produces variable results. The purpose of our study was to determine whether long-tract fistulas (>4 cm) correlated with successful closure. METHODS: All patients undergoing transsphincteric cryptoglandular fistula repair with anal fistula plugs were enrolled in a prospective database. Patients with Crohn's disease were excluded. Fistula tract length was measured intraoperatively by subtracting the remaining plug length from the original plug size. All procedures used standardized techniques and postoperative care pathways. The primary outcome was complete fistula closure assessed through both postoperative outpatient visits and a follow-up telephone questionnaire. RESULTS: Forty-one patients with 42 fistula tracks were enrolled over a 39-month period. Complete closure was achieved in 18 of 42 (43%) fistulas at a mean follow-up of 25 months. Closure was not associated with gender, age, tract location, duration of seton, or length of follow-up. Successful closure was significantly associated with increased tract length, because fistulas longer than 4 cm were nearly 3 times more likely to heal compared with shorter fistulas ((14/23, 61%) vs (4/19, 21%), P = .004; relative risk = 2.8; 95% CI 1.14-7.03). CONCLUSIONS: Anal fistula plug repair of cryptoglandular anorectal fistulas is more successful for long-tract fistulas. Although the overall success is modest, limiting surgical indications to fistulas exceeding 4 cm may maximize benefits of the plug technique.

Emergency laparoscopic colectomy: does it measure up to open?

BACKGROUND: Laparoscopic colectomy has become the standard of care for elective resections; however, there are few data regarding laparoscopy in the emergency setting. METHODS: By using a database with prospectively collected data, we identified 94 patients who underwent an emergency colectomy between August 2005 and July 2008. Laparoscopic surgeries were performed in 42 patients and were compared with 25 patients who were suitable for laparoscopy but received open colectomy. RESULTS: The groups had similar demographics with no differences in age, sex, or surgical indications. Blood loss was lower (118 vs 205 mL; P < 0.01) and the postoperative stay was shorter (8 vs 11 d; P = 0.02) in the laparoscopic patients, and perioperative mortality rates were similar between the 2 groups (1 vs 3; P = 0.29). CONCLUSIONS: With increasing experience, laparoscopic colectomy is a feasible option in certain emergency situations and is associated with shorter hospital stay, less morbidity, and similar mortality to that of open surgery.

Validation of a novel postoperative quality-of-life scoring system.

BACKGROUND: No specific scoring system exists for the assessment of postoperative quality of life (QOL) after major abdominal surgery. This study prospectively validates PQL, a novel prospective scoring system in patients having laparoscopic or open major abdominal colorectal surgery. METHODS: Six experienced surgeons developed the questionnaire. Twenty patients reviewed and selected the most relevant questions, yielding 14 questions. One hundred patients undergoing a variety of colorectal procedures completed the questionnaire preoperatively, and on postoperative days (POD) 1, 2, 4, 8, 12, 30, and 60. Internal validation was assessed by Cronbach's alpha and factor analysis. RESULTS: Cronbach's alpha revealed excellent internal consistency, ranging from .84 to .94 at all time points, even at POD 1 when Cronbach's alpha was .79, demonstrating that the items in the questionnaire measured the same underlying construct. Factor analysis consistently loaded at each follow-up time on the same 2 factors. CONCLUSIONS: Factor analysis consistently loaded at each follow-up time on the same 2 factors, designated the PQL Symptom Score and the PQL Recover Score.

Classic "outlet" rectal bleeding does not require full colonoscopy to exclude significant pathology.

PURPOSE: Full diagnostic colonoscopy often is performed to exclude significant pathology in patients presenting with rectal bleeding. In patients with classic "outlet" bleeding, defined as bright red blood after or during defecation, with no family history of colorectal neoplasia or change in bowel habits, we hypothesize that the diagnostic yield of complete colonoscopy will be low. The purpose of this study was to determine whether complete colonoscopy is necessary in the evaluation of patients with "outlet" rectal bleeding. METHODS: Information for all patients undergoing colonoscopy by a single endoscopist was prospectively recorded. Before each colonoscopy, a complete history, including indication for the examination, was obtained. Using standard definitions, patients with outlet bleeding, suspicious bleeding, hemorrhage, and occult bleeding were accessed and the findings of their colonoscopies were analyzed. Institutional permission was obtained. RESULTS: A total of 9,098 patients had colonoscopy recorded in the database, and 703 had the indication of outlet bleeding, 251 suspicious bleeding, 204 occult bleeding, and 67 hemorrhage. Of the patients with outlet bleeding, only 47 (6.7 percent) had significant lesions on colonoscopy (adenomas >1 cm, villous adenomas, cancer in situ, or invasive cancer). By contrast a greater number of significant lesions were present in patients with all other types of bleeding (17.2 percent; P<0.001). The incidence of invasive cancer was significantly lower in the outlet bleeding group compared with other types of bleeding (1 vs. 3.6 percent; P<0.01). Patients with outlet bleeding were much less likely than patients with other bleeding to have isolated right-sided colonic pathology. Younger patients with outlet bleeding have a particularly low yield on colonoscopy. In 182 patients younger than aged 50 years with outlet bleeding, only 3 (1.6 percent) had adenomas >1 cm and no invasive cancers were detected. CONCLUSIONS: In patients with classic outlet bleeding, the yield of a complete diagnostic colonoscopy is low. If the history is classic for outlet bleeding and no other indication for colonoscopy exists, flexible sigmoidoscopy is enough to exclude significant pathology.

Management of postoperative ileus: focus on alvimopan.

Postoperative ileus (POI) is a transient loss of coordinated peristalsis precipitated by surgery and exacerbated by opioid pain medication. Ileus causes a variety of symptoms including bloating, pain, nausea, and vomiting, but particularly delays tolerance of oral diet and liquids. Thus POI is a primary determinant of hospital stay after surgery. 'Fast-track' recovery protocols, opioid sparing analgesia, and laparoscopic surgery reduce but do not eliminate postoperative ileus. Alvimopan is a mu opioid receptor antagonist that blocks the effects of opioids on the intestine, while not interfering with their centrally mediated analgesic effect. Several large randomized clinical trials have demonstrated that alvimopan accelerates the return of gastrointestinal function after surgery and subsequent hospital discharge by approximately 20 hours after elective open segmental colectomy. However, it has not been tested in patients undergoing laparoscopic surgery and is less effective in patients receiving nonsteroidal anti-inflammatory agents in a narcotic sparing postoperative pain control regimen. Safety concerns seen with chronic low dose administration of alvimopan for opioid bowel dysfunction have not been noted with its acute use for POI.

Surgical management of rectal prolapse.

This article reviews the pathogenesis, clinical presentation and surgical management of rectal prolapse. Full-thickness prolapse of the rectum causes significant discomfort because of the sensation of the prolapse itself, the mucus that it secretes, and because it tends to stretch the anal sphincters and cause incontinence. Treatment of rectal prolapse is primarily surgical. Perineal surgical repairs are well tolerated, but are generally associated with higher recurrence rates. Abdominal repairs involve fixing the rectum to the sacrum by using either mesh or sutures, and tend to have the lowest recurrence rates. If significant preoperative constipation is present, a sigmoid resection can be performed at the time of rectopexy. For many patients, diarrhea and incontinence improve after surgery. Laparoscopic repair of rectal prolapse has similar morbidity and recurrence rates to open surgery, with attendant benefits of reduced length of hospital stay, postoperative pain and wound complications.

Patient safety: effect of institutional protocols on adverse events related to feeding tube placement in the critically ill.

BACKGROUND: Inadvertent passage of a nasoenteric feeding tube into the tracheobronchial tree can result in pneumothorax. Measures requiring feeding tube passage to 35 cm only followed by a radiograph to verify intraesophageal placement and creation of a specialized placement team were implemented to decrease the incidence of procedure-related pneumothorax. This study evaluates the effectiveness of our safety measures. STUDY DESIGN: Radiology reports from January 2000 through July 2003 were searched by computer with an algorithm designed to detect feeding tube placements possibly associated with the complication of intrabronchial placement or pneumothorax. Results were manually examined to eliminate false positives and verify causality. RESULTS: Feeding tubes were placed in 4,190 unique patients during the study period; 87 patients had an intrabronchial malposition, and 9 experienced a pneumothorax caused by their feeding tube. The safety measures resulted in a significant decrease in procedure-related pneumothorax (0.09% versus 0.38%, p < 0.05), and a decrease in pneumothorax among patients with an intrabronchial placement (3% versus 27%, p < 0.05). More than two-thirds of patients with a misplaced tube had an endotracheal tube or tracheostomy, illustrating that such patients are not protected. Repeated malposition in the same patient was surprisingly common; 32% of patients with one intrabronchial misplacement ultimately had multiple misplacements. The risk of pneumothorax increased with misplacement at night (p < 0.05) and increased exponentially with each additional misplacement (p < 0.05). CONCLUSIONS: Creating a specialized placement team, and initiating the safety measure of limiting feeding tube placement to 35 cm and obtaining a radiograph before full advancement reduced the incidence of procedure-related pneumothorax.

A critical role for C/EBPbeta binding to the AABS promoter response element in the human iNOS gene.

The human iNOS (hiNOS) gene is expressed in a tissue-specific manner, but the molecular basis for this regulation has not been elucidated. Here, we show that liver cell-specific hiNOS gene activation involves protein-DNA binding to an A-activator binding site (AABS) located at -192 nucleotides in the hiNOS promoter region. Mutation of this site in the -7.2 kb hiNOS promoter construct inhibited basal hiNOS promoter activity in primary rat hepatocytes (77%), and two human liver cell lines, AKN-1 (63%) and HepG2 (60%), but had no significant effect on basal hiNOS activity in three non-hepatic human cell types. Interestingly, mutation of AABS significantly abrogated cytokine-induced promoter activity in all cell types. C/EBPbeta transcription factor bound to AABS by gel shift assay. Overexpression of C/EBPbeta active form (LAP) increased hiNOS basal promoter activity approximately sixfold in liver cells, but had minimal effect in non-hepatic cells. In contrast, overexpression of the transcriptional inhibitor (LIP) strongly suppressed both basal and cytokine-inducible promoter activity. These data show that the cis-acting AABS DNA element mediates liver-specific basal hiNOS promoter activity through binding of the trans-acting C/EBPbeta factor. Further, C/EBPbeta binding to AABS functions as a "switchpoint" that is necessary for cytokine-inducible hiNOS gene expression in all cell types examined.

Protection of rat hepatocytes from apoptosis by inhibition of c-Jun N-terminal kinase.

BACKGROUND: Apoptotic cell death and c-Jun N-terminal kinase (JNK) activation occur after hepatic ischemia/reperfusion injury. In other cell types, JNK activation was shown to be required for apoptosis. This study tested the hypotheses that JNK contributes to hepatocellular apoptosis, and that inhibition of JNK activity improves cell viability. METHODS: Rat hepatocytes were harvested from Sprague-Dawley rats and pretreated with SP600125, a JNK inhibitor. Subsequently, they were exposed to apoptotic stimuli consisting of either the bile salt glycochenodeoxycholic acid (GCDC) or tumor necrosis factor (TNF)-alpha and actinomycin D. RESULTS: Western blotting demonstrated specific inhibition of JNK by SP600125. Inhibition of JNK resulted in improved viability measured with crystal violet, decreased in situ DNA nick end labeling positivity, and decreased cleavage of poly (ADP-ribose) polymerase and caspase-3. TNF-alpha and actinomycin D induced apoptosis, upregulated p53, and downregulated expression of the anti-apoptotic protein X-linked inhibitor of apoptosis protein. These effects were abrogated by JNK inhibition. CONCLUSIONS: These data show that pharmacologic inhibition of JNK activity reduces bile salt or TNF-alpha-induced apoptosis by maintaining expression of anti-apoptotic proteins. The results indicate that JNK is an important component of the apoptosis signaling cascade and suggest a possible therapeutic strategy in certain liver disorders.