High performance in risk stratification of intraductal papillary mucinous neoplasms by confocal laser endomicroscopy image analysis with convolutional neural networks (with video).

BACKGROUND AND AIMS: EUS-guided needle-based confocal laser endomicroscopy (EUS-nCLE) can differentiate high-grade dysplasia/adenocarcinoma (HGD-Ca) in intraductal papillary mucinous neoplasms (IPMNs) but requires manual interpretation. We sought to derive predictive computer-aided diagnosis (CAD) and artificial intelligence (AI) algorithms to facilitate accurate diagnosis and risk stratification of IPMNs. METHODS: A post hoc analysis of a single-center prospective study evaluating EUS-nCLE (2015-2019; INDEX study) was conducted using 15,027 video frames from 35 consecutive patients with histopathologically proven IPMNs (18 with HGD-Ca). We designed 2 CAD-convolutional neural network (CNN) algorithms: (1) a guided segmentation-based model (SBM), where the CNN-AI system was trained to detect and measure papillary epithelial thickness and darkness (indicative of cellular and nuclear stratification), and (2) a reasonably agnostic holistic-based model (HBM) where the CNN-AI system automatically extracted nCLE features for risk stratification. For the detection of HGD-Ca in IPMNs, the diagnostic performance of the CNN-CAD algorithms was compared with that of the American Gastroenterological Association (AGA) and revised Fukuoka guidelines. RESULTS: Compared with the guidelines, both n-CLE-guided CNN-CAD algorithms yielded higher sensitivity (HBM, 83.3%; SBM, 83.3%; AGA, 55.6%; Fukuoka, 55.6%) and accuracy (SBM, 82.9%; HBM, 85.7%; AGA, 68.6%; Fukuoka, 74.3%) for diagnosing HGD-Ca, with comparable specificity (SBM, 82.4%; HBM, 88.2%; AGA, 82.4%; Fukuoka, 94.1%). Both CNN-CAD algorithms, the guided (SBM) and agnostic (HBM) models, were comparable in risk stratifying IPMNs. CONCLUSION: EUS-nCLE-based CNN-CAD algorithms can accurately risk stratify IPMNs. Future multicenter validation studies and AI model improvements could enhance the accuracy and fully automatize the process for real-time interpretation.

Total Colonic Hirschsprung's Disease: The Hypermotility and Skin Rash Protocol.

INTRODUCTION: Total colonic Hirschsprung's disease (TCHD) presents a postoperative challenge due to multiple stools and perineal rash. We propose a protocol developed by pediatric surgeons and ostomy nurses to help prevent and treat hypermotility and severe perineal rash, especially in younger children who are not toilet trained. MATERIALS AND METHODS: We retrospectively reviewed our TCHD patients' charts from 2014 to 2017. All patients received a prescribed protocol for the treatment of hypermotility and perineal rash. We describe patients who underwent their pull through before and after the age of urine toilet training, and assessed the number of bowel movements, the perineal skin status, and growth. RESULTS: We treated 25 patients. Out of 25, 9 patients received a straight ileoanal pull through before the age of 18 months. Nine of 25 patients presented for a second opinion and had redo pull through. The remaining seven presented for bowel management after having a pull through at another institution. All these were treated following the hypermotility protocol. In total, 19 of 25 patients were not toilet trained. The mean number of bowel movements in all groups was 4 (3-5). All had a resolution of perineal rash and liquid stools after 3 months. Eleven of the 25 patients presented with failure to thrive. Two older patients experienced severe proctalgia requiring replacement of the ileostomy. CONCLUSION: TCHD patients who underwent definitive pull through had nine high incidence of multiple stool, perineal rash, and low growth. With the implementation of bowel management care to slow the stools and a perineal skin protocol to treat the skin, we believe that these symptoms can be minimized even in patients who are not toilet trained. Since the implementation of this protocol, we have changed our practice to perform the pull through in such patients between the age of 6 and 18 months.

The use of rotational fluoroscopy and 3-D reconstruction in the diagnosis and surgical planning for complex cloacal malformations.

INTRODUCTION: Cloacal malformations, a confluence of the urinary tract, vagina and rectum into a single common channel, has a broad and complex anatomic spectrum requiring an imaging tool for visualization, measurement, and surgical planning for the reconstruction of these structures. We evaluated the role of 3-D fluoroscopy for this purpose, as it offers a combination of spatial correlation with precise anatomic measurements. METHODS: We examined our imaging protocol for patients with a cloacal malformation and report our experience with rotational fluoroscopy and 3-D reconstruction in 16 consecutive patients referred for cloacal reconstruction. The length of the common channel (CC), the length of the urethra from the bladder neck to the common channel, and the height (and existence or absence) of a vagina or vaginas were determinants of the surgical procedures used for the repair. RESULTS: We performed 16 consecutive 3-D cloacagrams (age range 4 months to 9 years) using a new protocol (Figure 1) that provided the following data which helped with surgical planning: Gynecologic: 3 cases with a single vagina, 5 cases with a duplicated Mullerian system (3 of which were asymmetric) and 2 cases with high vaginas requiring vaginal replacement. Colorectal: Four had a high rectum requiring an abdominal approach, and 6 had a rectum reachable via a posterior sagittal approach. Urologic: Two ectopic ureters requiring reimplantation, 3 patients had vesicoureteral reflux (1 bilateral, 2 unilateral), 1 patient had no bladder, and 7 had a normal sized bladder. Common channel length and urethral length were demonstrated in all cases and used to decide between a total urogenital mobilization or a separation of vagina(s) from the common channel, urogenital separation. CONCLUSION: The 3-D cloacagram can help predict the surgical plan for urologic, gynecologic, and colorectal components of the cloacal repair. It can predict the CC length as well as the length of the urethra. It helps with predicting the need for vaginal replacement and whether an abdominal approach is needed for the rectum. Its effectiveness is based on the ability to adequately distend structures and see their distal most extent, an advantage over other modalities such as MRI. Added benefits (particularly from the 3D view) include a better spatial understanding of the defect and the diagnosis of concomitant urological abnormalities such as vesicoureteral reflux and ectopic ureters. Disadvantages to this procedure include the need for general anesthesia and a higher exposure to radiation. LEVEL OF EVIDENCE: 3.

Presacral masses and sacrococcygeal teratomas in patients with and without anorectal malformations: A single institution comparative study.

BACKGROUND: Despite variability at presentation, presacral masses in patients with and without anorectal malformations (ARM) appear histologically similar. The purpose of this study was to identify differences in oncologic outcomes between these two groups. METHODS: A retrospective review was performed utilizing our institutional cancer and colorectal and pelvic reconstruction databases for patients with presacral masses and sacrococcygeal teratomas between 1990 and 2017. Data captured included age at surgical resection, type of ARM, tumor location within the pelvis, tumor histopathology, tumor size, adjuvant chemotherapy, recurrence, and follow-up. RESULTS: Forty-six patients comprised our cohort, of whom 12 had an ARM. The median age was older at resection for those with an ARM (1.4 years; range 1 day to 29.4 years) compared to those without an ARM (9 days; range 0 days to 6.9 years) (p = 0.01). The mean tumor size was 2.5 cm in patients with an ARM compared to 6.0 cm in patients without an ARM (p = 0.036). All patients with ARM had exclusively intrapelvic tumors, and histopathology included mature teratoma (8), yolk sac tumor (1), lipoma (1), and unknown (2). Tumor location for patients with sacral and presacral masses without ARM included exclusively extrapelvic (10), primarily extrapelvic with large intrapelvic component (7), primarily intrapelvic with extrapelvic component (1), exclusively intrapelvic (8), and unknown (8). Histopathology for patients with presacral masses without ARM included mature teratoma (20), immature teratoma (7), yolk sac tumor (3), ganglioneuroma (1), neuroblastoma (1), benign epithelial cyst (1), and unknown (1). Tumor recurrence rate was similar between patients with ARM (n = 3, 25%) and those without an ARM (n = 5, 15%) (p = 0.41). The 5-year event free survival was 65% (95% CI: 25%-87%) in the group with ARM and 81% (95% CI: 60%-92%) in the group without ARM (p = 0.44). CONCLUSION: Sacral and presacral masses in patients with ARM are resected at a later age and are more likely to be intrapelvic. They appear histologically similar and have similar rates of recurrence and malignancy when compared to patients without ARM. LEVEL OF EVIDENCE: III TYPE OF STUDY: Retrospective comparative study.

Urinary Outcomes in Patients with Down's Syndrome and Hirschsprung's Disease.

INTRODUCTION: Previous research in children with Hirschsprung's disease (HD) and Down's syndrome (DS) has focused on colorectal outcomes. We set out to review urinary outcomes in this patient group. MATERIALS AND METHODS: The medical records of all patients aged five years and older with HD were reviewed, and patients and caregivers filled out the Vancouver Symptom Score at intake, which is designed and validated to diagnose dysfunctional elimination syndrome. RESULTS: A total of 104 patients with HD were included in this study. Of these, 16 (15%) patients had DS. There were no significant differences in the prevalence of enterocolitis or colorectal symptoms between patients with or without DS. Five of 88 (6%) patients without DS and 7 of 16 (44%) (p = 0.00001) with DS reported having urinary accidents. Patients with HD and DS scored higher on the Vancouver score (9 vs. 17.5; p = 0.007), indicating more severe urinary symptoms. Patients who also reported fecal accidents scored significantly higher on the Vancouver (12 vs. 9; n = 61; p = 0.016), indicating more problems. CONCLUSION: Patients with DS appear to be a unique subset of HD patients who have a higher prevalence of urinary symptoms after surgery. In the postoperative care of patients with HD and DS, a strong focus should be placed on postoperative urinary care in addition to their bowel care. This could significantly ease care and contribute to the quality of life of the parents and the patient.

A comparison of Malone appendicostomy and cecostomy for antegrade access as adjuncts to a bowel management program for patients with functional constipation or fecal incontinence.

BACKGROUND: Appendicostomy and cecostomy are two approaches for antegrade enema access for children with severe constipation or fecal incontinence as adjuncts to a mechanical bowel management program. Each technique is associated with a unique set of complications. The purpose of our study was to report the rates of various complications associated with antegrade enema access techniques to help guide which option a clinician offers to their patients. METHODS: We reviewed all patients in our Center who received an appendicostomy or cecostomy from 2014 to 2017 who were participants in our bowel management program. RESULTS: 204 patients underwent an antegrade access procedure (150 appendicostomies and 54 cecostomies). Skin-level leakage (3% vs. 22%) and wound infections (7% vs. 28%) occurred less frequently in patients with appendicostomy compared to cecostomy. Nineteen (13%) appendicostomies required revision for stenosis, 4 (3%) for mucosal prolapse, and 1 (1%) for leakage. The rates of stenosis (33 vs. 12%) and wound infection (13 vs. 6%) were higher in patients who received a neoappendicostomy compared to an in situ appendicostomy. Intervention was needed in 19 (35%) cecostomy patients, 15 (28%) for an inability to flush or a dislodged tube, and 5 for major complications including intraperitoneal spillage in 4 (7%) and 1 (2%) for a tube misplaced in the ileum, all occurring in patients with a percutaneously placed cecostomy. One appendicostomy (1%) patient required laparoscopic revision after the appendicostomy detached from the skin. CONCLUSION: Patients had a lower rate of minor and major complications after appendicostomy compared to cecostomy. The unique complication profile of each technique should be considered for patients needing these procedures as an adjunct to their care for constipation or fecal incontinence. TYPE OF STUDY: Retrospective comparative study. LEVEL OF EVIDENCE: Level III.

A structured bowel management program for patients with severe functional constipation can help decrease emergency department visits, hospital admissions, and healthcare costs.

BACKGROUND: Published health-care costs related to constipation in children in the USA are estimated at $3.9 billion/year. We sought to assess the effect of a bowel management program (BMP) on health-care utilization and costs. METHODS: At two collaborating centers, BMP involves an outpatient week during which a treatment plan is implemented and objective assessment of stool burden is performed with daily radiography. We reviewed all patients with severe functional constipation who participated in the program from March 2011 to June 2015 in center 1 and from April 2014 to April 2016 in center 2. ED visits, hospital admissions, and constipation-related morbidities (abdominal pain, fecal impaction, urinary retention, urinary tract infections) 12 months before and 12 months after completion of the BMP were recorded. RESULTS: One hundred eighty-four patients were included (center 1 = 96, center 2 = 88). Sixty-three (34.2%) patients had at least one unplanned visit to the ED before treatment. ED visits decreased to 23 (12.5%) or by 64% (p < 0.0005). Unplanned hospital admissions decreased from 65 to 28, i.e., a 56.9% reduction (p < 0.0005). CONCLUSION: In children with severe functional constipation, a structured BMP decreases unplanned visits to the ED, hospital admissions, and costs for constipation-related health care. LEVEL OF EVIDENCE: 3.

Can sacral development as a marker for caudal regression help identify associated urologic anomalies in patients with anorectal malformation?

BACKGROUND AND AIM: The sacral ratio (SR) is a well-established tool to quantify sacral development in patients with anorectal malformations (ARM) and can be used as a predictor of fecal continence. We hypothesized that a lower SR correlated with the presence of urologic and renal malformations. METHODS: We retrospectively reviewed the medical records of patients with ARM treated at our center from 2014 to 2016. We measured the lateral SR as a marker for sacral development and assessed the spine for the presence of tethered cord (TC). Urological and renal anomalies, including single kidney, hydronephrosis, hypospadias, vesicoureteral reflux (VUR), ectopic ureter, and penoscrotal transposition were assessed. Analysis of variance (ANOVA), t-tests, and multivariable linear regression were used to test for differences in SR with consideration of associated urologic malformations and tethered cord. RESULTS: 283 patients with ARM were included for analysis (156 females). The median age was 39months (10-90). Among these, 178 (55.6%) had 1 or more urologic malformations, and 81 (25.3%) had a TC. Hydronephrosis, high-grade VUR (3-5), solitary kidney, and tethered cord were significantly associated with lower SR (p<0.01). In multivariable regression models, the presence of urologic abnormalities remained significantly associated with lower a SR despite the presence or absence of TC (p<0.001). CONCLUSION: SR is a potentially useful indicator of certain urologic anomalies including hydronephrosis, high grade VUR, and solitary kidney in patients with ARM. This association is independent of the presence of TC. A sacral ratio as a part of the VACTERL screening can help the surgeon identify which patients need closer urologic follow up. LEVEL OF EVIDENCE: IV.

Are Senna based laxatives safe when used as long term treatment for constipation in children?

BACKGROUND AND AIM: Senna is a stimulant laxative commonly used by pediatricians, pediatric gastroenterologists, and pediatric surgeons. Many clinicians avoid Senna for reasons such as tolerance or side effects but this has little scientific justification. We recently found several patients we were caring for developed perineal blistering during the course of Senna treatment. Because of this we chose to review the literature to identify side effects in children taking this medication as well as to analyze our Center's experience with Senna's secondary effects. METHODS: We performed a literature review (MEDLINE, PUBMED) using the keywords of Senna, sen, sennosides and children, and pediatric and functional (idiopathic) constipation. We looked for articles with information regarding perineal blisters related to Senna as well as other secondary effects of Senna laxatives in children when used on a long-term basis. We also reviewed the charts of our patients who had previously taken Senna or are currently taking Senna, looking for adverse reactions. RESULTS: Eight articles in the literature reported perineal blisters after administration of Senna laxatives in 28 patients. Of those occurrences, 18 patients (64%) had accidental administration of Senna and 10 (36%) had Senna prescribed as a long term treatment. All of the blistering episodes were related to high dose, night-time accidents, or intense diarrhea with a long period of stool to skin contact. At our institution, from 2014 to 2017, we prescribed Senna and have recorded data to 640 patients. During the study period, 17 patients (2.2%) developed blisters during their treatment. Patients who developed blisters had higher doses 60mg/day; 60 [12-100] vs. 17.5 [1.7-150] (p<0.001). All of the blistering episodes were related to night-time accidents, with a long period of stool to skin contact. 83 (13%) patients presented minor side effects such as abdominal cramping, vomiting or diarrhea which resolved once the type of laxatives were changed or enemas were started. The doses of Senna was not significantly different in these patients 15mg/day [4.4-150] vs. 17.5mg/day [1.5-150]. There were no other long-term side effects from Senna found in the pediatric literature for long-term treatment besides abdominal cramping or diarrhea during the first weeks of administration. We found no evidence of tolerance to Senna in our review. CONCLUSION: There is a paucity of information in the literature regarding side effects of sennosides as a long-term therapy, and to our knowledge, this is the first review of Senna side effects in children. Senna induced dermatitis is rare, but may occur when patients need a higher dose. All of the cases described had a long period of exposure of the skin to stool. Besides the perineal rash with blisters, we could find no other described major side effect with Senna administration in the pediatric population or evidence of the frequently mentioned concern of the development of tolerance to Senna. Pediatric caregivers should advise families of the rare side effect of skin blistering and educate them to change the diaper frequently in children who are not toilet- trained to reduce stool to skin exposure. We can conclude from this review that Senna is a safe treatment option for constipation in children. LEVEL OF EVIDENCE: IV.