- Email: [email protected]
Diagnosis and treatment of choledochocolonic fistula by ERCP
AJG – September, Suppl., 2001
Abstracts
duced with ALT ⬎4x ⫹ Bil ⬎3x), the SP, PPV and ROC increase to 93%, 0.89 and 0.91 respectively, thus having the best predictive value. (see Table) Conclusions: While our study concurs with previous reports that ALT ⬎4x (⬎150 U/L) and Bil ⬎3x are the best single markers, we conclude that rather than using a single criteria, if the combination of 2 biochemical parameters is used, as in our model (ALT ⬎4x and Bil ⬎3x, biliary AP can be predicted better, especially in females, with a very high degree of accuracy. Variable
SS (%)
SP (%)
PPV
ROC
ALT ⬎4x Bil ⬎3x ALT ⬎4x ⫹ Bil ⬎3x ALT ⬎4x ⫹ Bil ⬎3x ⫹ Female
65 25 72 72
88 98 86 93
0.80 0.91 0.80 0.89
0.76 0.61 0.80 0.91
316 Occult hyperlipidemic acute pancreatitis (OHLP) in diabetic ketoacidosis (DKA): a prospective evaluation Dhiraj - Yadav, E P Norkus and C S Pitchumoni*. 1Gastroenterology, Our Lady of Mercy University Medical Center, Bronx, NY, United States; 2Biostatistics, Our Lady of Mercy University Medical Center, Bronx, NY, United States; and 3Gastroenterology, Our Lady of Mercy University Medical Center, Bronx, NY, United States. Purpose: Our early observations on OHLP in DKA have been reported previously (Am J Gastroenterol 2000;95:3123-8). We showed that CT proven Acute Pancreatitis (AP) occurs in 11% cases of DKA and a subset of these patients have transient and profound hypertriglyceridemia (HTG) which could be a pathogenetic mechanism for AP in these patients. The aim of our present study was to evaluate incidence and clinical features of OHLP in DKA. Methods: We prospectively evaluated 180 consecutive episodes of DKA admitted to our institution. Along with careful history and routine laboratory studies, we estimated amylase, lipase and triglyceride (TG) levels in the serum on admission and 48 hrs later. All patients with abdominal pain and/or elevation of pancreatic enzymes (⬎3x normal) or TG levels (⬎500 mg/dl) underwent CT scan of the abdomen. Diagnosis of AP was established only if CT was positive. Serum TG level ⬎500 mg/dl was considered as HTG. Results: HTG (⬎500 mg/dl) was seen in 34 (19%) episodes of DKA. 18 patients (10%) had AP, out of which 9 (50%) had transient and profound HTG (6 had isolated HTG, 2 had HTG ⫹ gallstones and 1 had HTG ⫹ alcohol abuse). Characteristics of patients with OHLP (n ⫽ 9): There were 6 (66.6%) males and 3 (33.3%) females, the mean age was 40.3 ⫾ 7 years. New onset DM was seen in 5 (56%) patients and history of hyperlipidemia was present only in 3 (33.3%) cases. Abdominal pain was present in 5 (56%) patients. The mean levels of serum TG on admission were 3509 ⫾ 3476 mg/dl (660-9300 mg/dl). 8 out of 9 patients had an admission TG level of ⬎1000 mg/dl. Serum TG levels returned to levels ⬍300 mg/dl after resolution of DKA in all patients. Mean values of other laboratory data in OHLP patients were: glucose 741 ⫾ 368 mg/dl, amylase 272 ⫾ 287 IU/L, lipase 325 ⫾ 630 IU/dl, serum osmolality 329 ⫾ 32 mOsm/L, arterial pH 7.2 ⫾ 0.15. The clinical course of OHLP was similar to AP due to other causes. Only 1 patient had a second episode of OHLP with a subsequent episode of DKA. Conclusions: Occult AP occurs in 10% DKA cases and in nearly 50% of them HTG is the cause. All patients with DKA should be screened for serum TG elevations on admission.
317 Diagnosis and treatment of choledochocolonic fistula by ERCP B. Zamiri, MD, R. Crock, MD, N. Fahmy, MD. Northeastern Ohio Universities College of Medicine, Affiliated Hospitals at Canton, Ohio.
S101
Choledochocolonic fistula is the rarest of the biliary-enteric fistulas with only a few reported cases. The advent of ERCP gives practitioners the ability to visualize and treat this uncommon condition. To our knowledge, we report the first case involving the diagnosis and treatment of choledochocolonic fistula by ERCP. A 57-year-old female with a ten-day history of watery diarrhea, nausea and non-feculent vomiting presented with diffuse abdominal pain radiating to her back. On admission vital signs were stable, with diffuse abdominal tenderness in the right upper quadrant and epigastric area. There was no rebound tenderness, organomegaly or masses. Murphy’s sign was negative. Liver function tests showed sign of obstruction, hepatitis panel and AMA were negative and blood culture revealed gram negative bacteremia. She was started on an intravenous Quinolone. Ultrasound of the gallbladder and CT of the abdomen were normal. However, the HIDA scan showed a reduced ejection fraction. Patient underwent ERCP which revealed a choledochocolonic fistula with a 3– 4 mm fistulous tract, normal common bile duct caliber and no stones. There was no filling defect in the biliary tree or intrahepatic biliary duct. Colonoscopy after stent placement showed a fistula from the common bile duct to the hepatic flexure of the colon. The repeat colonoscopy in six weeks showed a healed fistula. Biliary-enteric fistula is an uncommon complication of chronic cholelithiasis. It is important to make the diagnosis to prevent further complications. Although rare, high index of suspicion is a key in identifying and treating this type of fistula to prevent chronic biliary complications. Barium enema and surgery have been standard treatment. To our knowledge, ERCP has never been reported as a management modality of choledochocolonic fistula. Despite advances in surgical treatment and postoperative care, the morbidity and the mortality of this condition remain higher with surgery. We have shown that ERCP and stent placement can be a successful management modality in the diagnosis and treatment of choledochocolonic fistula, and may prove a successful alternative to surgery in the high-risk patient.
SMALL INTESTINE/UNCLASSIFIED 318 Etiology of symptomatically non-responsive celiac disease (NCD) AS Abdulkarim, LJ Burgart, JA Murray. Mayo Clinic, Rochester, MN Background: Non-response or relapse of symptoms is common in patients with celiac disease treated with gluten free diet (GFD). True refractory sprue (RS) is rare and is defined as the continuation or resurgence of malabsorption with villous atrophy despite GFD. Aims: 1) Identify the causes of persistent symptoms in this referral population 2) Characterize patients with true RS. Methods: Patients seen between 1/1997 and 9/2000 for systematic evaluation of NCD. Patient’s records and small bowel biopsies (SBB) were reviewed. Both polymerase chain reaction and Southern blot analysis were utilized for the detection of T-cell receptor and immunoglobulin gene rearrangement. Results: Thirty-eight patients were referred with presumed NCD of whom only 9 had true RS. Three did not have CD on review of the original biopsies. Weight loss and diarrhea were the most common reason for evaluation, 68% and 65% in NCD and 100% and 66% in RS. Nineteen patients were identified to have gluten contamination (GC), 15 of 19 had EMA and/or AGA positivity (9 AGA only, 4 EMA and AGA, 1 EMA only). Of the 35 patients with CD additional diagnoses accounting for lack of response to GFD included: pancreatic insufficiency (5), irritable bowel syndrome (4), bacterial overgrowth (4), collagenous colitis (2), ulcerative jejunitis (2) lymphocytic colitis (1), T-cell lymphoma (1), pancreatic cancer (1), and fructose intolerance (1). Of these 35, steroids had been used in 10 of whom only 3 proved to have true RS. Of 9 with RS, 4 had RS alone; 5 had additional diagnoses of ulcerative jejunitis (2), collagenous colitis (2), lymphocytic colitis (1), and bacterial overgrowth (1). Five of 9 with RS were tested for HLA typing, (4) DQw2, (1) DQw8. Five of 9 with RS were
Abstracts
duced with ALT ⬎4x ⫹ Bil ⬎3x), the SP, PPV and ROC increase to 93%, 0.89 and 0.91 respectively, thus having the best predictive value. (see Table) Conclusions: While our study concurs with previous reports that ALT ⬎4x (⬎150 U/L) and Bil ⬎3x are the best single markers, we conclude that rather than using a single criteria, if the combination of 2 biochemical parameters is used, as in our model (ALT ⬎4x and Bil ⬎3x, biliary AP can be predicted better, especially in females, with a very high degree of accuracy. Variable
SS (%)
SP (%)
PPV
ROC
ALT ⬎4x Bil ⬎3x ALT ⬎4x ⫹ Bil ⬎3x ALT ⬎4x ⫹ Bil ⬎3x ⫹ Female
65 25 72 72
88 98 86 93
0.80 0.91 0.80 0.89
0.76 0.61 0.80 0.91
316 Occult hyperlipidemic acute pancreatitis (OHLP) in diabetic ketoacidosis (DKA): a prospective evaluation Dhiraj - Yadav, E P Norkus and C S Pitchumoni*. 1Gastroenterology, Our Lady of Mercy University Medical Center, Bronx, NY, United States; 2Biostatistics, Our Lady of Mercy University Medical Center, Bronx, NY, United States; and 3Gastroenterology, Our Lady of Mercy University Medical Center, Bronx, NY, United States. Purpose: Our early observations on OHLP in DKA have been reported previously (Am J Gastroenterol 2000;95:3123-8). We showed that CT proven Acute Pancreatitis (AP) occurs in 11% cases of DKA and a subset of these patients have transient and profound hypertriglyceridemia (HTG) which could be a pathogenetic mechanism for AP in these patients. The aim of our present study was to evaluate incidence and clinical features of OHLP in DKA. Methods: We prospectively evaluated 180 consecutive episodes of DKA admitted to our institution. Along with careful history and routine laboratory studies, we estimated amylase, lipase and triglyceride (TG) levels in the serum on admission and 48 hrs later. All patients with abdominal pain and/or elevation of pancreatic enzymes (⬎3x normal) or TG levels (⬎500 mg/dl) underwent CT scan of the abdomen. Diagnosis of AP was established only if CT was positive. Serum TG level ⬎500 mg/dl was considered as HTG. Results: HTG (⬎500 mg/dl) was seen in 34 (19%) episodes of DKA. 18 patients (10%) had AP, out of which 9 (50%) had transient and profound HTG (6 had isolated HTG, 2 had HTG ⫹ gallstones and 1 had HTG ⫹ alcohol abuse). Characteristics of patients with OHLP (n ⫽ 9): There were 6 (66.6%) males and 3 (33.3%) females, the mean age was 40.3 ⫾ 7 years. New onset DM was seen in 5 (56%) patients and history of hyperlipidemia was present only in 3 (33.3%) cases. Abdominal pain was present in 5 (56%) patients. The mean levels of serum TG on admission were 3509 ⫾ 3476 mg/dl (660-9300 mg/dl). 8 out of 9 patients had an admission TG level of ⬎1000 mg/dl. Serum TG levels returned to levels ⬍300 mg/dl after resolution of DKA in all patients. Mean values of other laboratory data in OHLP patients were: glucose 741 ⫾ 368 mg/dl, amylase 272 ⫾ 287 IU/L, lipase 325 ⫾ 630 IU/dl, serum osmolality 329 ⫾ 32 mOsm/L, arterial pH 7.2 ⫾ 0.15. The clinical course of OHLP was similar to AP due to other causes. Only 1 patient had a second episode of OHLP with a subsequent episode of DKA. Conclusions: Occult AP occurs in 10% DKA cases and in nearly 50% of them HTG is the cause. All patients with DKA should be screened for serum TG elevations on admission.
317 Diagnosis and treatment of choledochocolonic fistula by ERCP B. Zamiri, MD, R. Crock, MD, N. Fahmy, MD. Northeastern Ohio Universities College of Medicine, Affiliated Hospitals at Canton, Ohio.
S101
Choledochocolonic fistula is the rarest of the biliary-enteric fistulas with only a few reported cases. The advent of ERCP gives practitioners the ability to visualize and treat this uncommon condition. To our knowledge, we report the first case involving the diagnosis and treatment of choledochocolonic fistula by ERCP. A 57-year-old female with a ten-day history of watery diarrhea, nausea and non-feculent vomiting presented with diffuse abdominal pain radiating to her back. On admission vital signs were stable, with diffuse abdominal tenderness in the right upper quadrant and epigastric area. There was no rebound tenderness, organomegaly or masses. Murphy’s sign was negative. Liver function tests showed sign of obstruction, hepatitis panel and AMA were negative and blood culture revealed gram negative bacteremia. She was started on an intravenous Quinolone. Ultrasound of the gallbladder and CT of the abdomen were normal. However, the HIDA scan showed a reduced ejection fraction. Patient underwent ERCP which revealed a choledochocolonic fistula with a 3– 4 mm fistulous tract, normal common bile duct caliber and no stones. There was no filling defect in the biliary tree or intrahepatic biliary duct. Colonoscopy after stent placement showed a fistula from the common bile duct to the hepatic flexure of the colon. The repeat colonoscopy in six weeks showed a healed fistula. Biliary-enteric fistula is an uncommon complication of chronic cholelithiasis. It is important to make the diagnosis to prevent further complications. Although rare, high index of suspicion is a key in identifying and treating this type of fistula to prevent chronic biliary complications. Barium enema and surgery have been standard treatment. To our knowledge, ERCP has never been reported as a management modality of choledochocolonic fistula. Despite advances in surgical treatment and postoperative care, the morbidity and the mortality of this condition remain higher with surgery. We have shown that ERCP and stent placement can be a successful management modality in the diagnosis and treatment of choledochocolonic fistula, and may prove a successful alternative to surgery in the high-risk patient.
SMALL INTESTINE/UNCLASSIFIED 318 Etiology of symptomatically non-responsive celiac disease (NCD) AS Abdulkarim, LJ Burgart, JA Murray. Mayo Clinic, Rochester, MN Background: Non-response or relapse of symptoms is common in patients with celiac disease treated with gluten free diet (GFD). True refractory sprue (RS) is rare and is defined as the continuation or resurgence of malabsorption with villous atrophy despite GFD. Aims: 1) Identify the causes of persistent symptoms in this referral population 2) Characterize patients with true RS. Methods: Patients seen between 1/1997 and 9/2000 for systematic evaluation of NCD. Patient’s records and small bowel biopsies (SBB) were reviewed. Both polymerase chain reaction and Southern blot analysis were utilized for the detection of T-cell receptor and immunoglobulin gene rearrangement. Results: Thirty-eight patients were referred with presumed NCD of whom only 9 had true RS. Three did not have CD on review of the original biopsies. Weight loss and diarrhea were the most common reason for evaluation, 68% and 65% in NCD and 100% and 66% in RS. Nineteen patients were identified to have gluten contamination (GC), 15 of 19 had EMA and/or AGA positivity (9 AGA only, 4 EMA and AGA, 1 EMA only). Of the 35 patients with CD additional diagnoses accounting for lack of response to GFD included: pancreatic insufficiency (5), irritable bowel syndrome (4), bacterial overgrowth (4), collagenous colitis (2), ulcerative jejunitis (2) lymphocytic colitis (1), T-cell lymphoma (1), pancreatic cancer (1), and fructose intolerance (1). Of these 35, steroids had been used in 10 of whom only 3 proved to have true RS. Of 9 with RS, 4 had RS alone; 5 had additional diagnoses of ulcerative jejunitis (2), collagenous colitis (2), lymphocytic colitis (1), and bacterial overgrowth (1). Five of 9 with RS were tested for HLA typing, (4) DQw2, (1) DQw8. Five of 9 with RS were