Search

Background: Pancreatic cancer patients with positive peritoneal cytology (PPC) as a sole metastatic site are poorly characterized. Whether they behave similarly to other stage IV patients is unknown., Methods: Patients with stage IV disease at our institution between 2003 and 2013 were identified. Inclusion criteria for PPC cohort were PPC at laparoscopy and no laparoscopic and/or radiographic evidence of metastasis. Patients with gross metastasis had laparoscopic and/or radiographic evidence of metastasis., Results: Among 308 patients, 43 patients had PPC and 265 had gross metastasis. PPC cohort: 3 (7%) resectable, 8 (19%) borderline resectable, and 32 (74%) unresectable tumor. Disease progression occurred in 37 (86%). Sixteen of 43 (37%) also received local therapy (1 surgery and 15 chemoradiation). PPC vs gross metastasis cohort differed as follows: baseline Ca 19-9 (440 vs 1,904 IU/mL, P < .0001); Eastern Cooperative Oncology Group (ECOG) score ≤1 (98 vs 88%, P = .04); median overall survival (13.9 vs 9.4 months, P = .0001)., Conclusions: Patients with PPC failed to display long-term disease-free survival, although overall survival was superior compared with those with gross metastasis. Patients with PPC may need to be considered a specific subgroup for staging and survival analysis., (Copyright © 2016. Published by Elsevier Inc.)

Background: Since any intraductal papillary mucinous neoplasm (IPMN) is at least premalignant, avoiding conversion to invasion by pancreatic resection should provide a survival advantage-but how much?, Methods: We reviewed 100 cases of IPMN that were resected. Survival was compared between 3 groups: noninvasive IPMN (n = 75), invasive IPMN (n = 25), and invasive ductal adenocarcinoma (n = 24), the latter matched by tumor-node-metastasis (TNM) stage to the IPMN invasive group., Results: The 5-year disease-specific survival was significantly better for the noninvasive IPMN group (100%) than the invasive IPMN group (46%). Tumor recurrence was infrequent with noninvasive IPMNs (1.3% benign IPMN). Recurrence was common in the invasive IPMN group (46%). Even the subgroup with stage 1 disease had a 25% recurrence of malignancy. Survival curves were not different (P = .11) between the cases matched by stage for those with invasive IPMN cases versus cases with ductal adenocarcinoma., Conclusion: Patients with the invasive form of IPMN will have a similarly poor survival as those with ductal adenocarcinoma. In patients thought to have a benign IPMN, these lesions should be removed to avoid conversion to invasive cancer and to preserve the opportunity for the more favorable prognosis observed in this study.

Background: Patients with cancer who undergo pancreaticoduodenectomy (PD) followed by radiation and 5-fluorouracil (5-FU) therapy have experienced median overall survival from 18 to 24 months and an actuarial 2-year overall survival from 34% to 48%. We previously reported an 84% 2-year survival using a novel adjuvant chemoradiation protocol that included alpha interferon. This report describes the continued observations regarding this methodology with longer follow-up and more than twice the number of patients as the original report., Methods: From July 1995 to May 2002, 43 patients with adenocarcinomas in the pancreatic head underwent PD at our institution. The mean age was 62 years (range 29 to 77) and 60% were men. Final pathologic findings were stage I (2%), II (12%), III (72%), and IVa (14%) while 84% had positive lymph nodes (average number of nodes positive was 3.2 nodes, (range 0 to 13). Tumor extended through the capsule of the surgical specimen in 70%. These patients then received our investigational protocol consisting of external-beam irradiation at a dose of 4,500 to 5,400 cGy (25 fractions over 5 weeks) and three-drug chemotherapy: continuous infusion 5-FU (200 mg/m(2) daily, days 1 to 35), weekly intravenous bolus cisplatin (30 mg/m(2) daily, days 1,8,15,22,29), and subcutaneous alpha, interferon (3 x 10(6) units, days 1 to 35). This chemoradiation was followed by continuous infusion 5-FU (200 mg/m(2) daily, weeks 9 to 14 and 17 to 22). Chemoradiation was generally initiated between 6 and 8 weeks after surgery., Results: All patients completed radiation therapy. There were no deaths due to chemoradiation but 42% were hospitalized during chemoradiation, virtually all due to gastrointestinal toxicity. With a mean follow-up time of 31.9 months, 67% of the patients are alive. Therefore, the median survivorship has not been reached. Actuarial overall survival for the 1-, 2-, and 5-year periods was 95% (confidence interval [CI] = 91% to 98%), 64% (CI = 56% to 72%), and 55% (CI = 46% to 65%), respectively., Conclusions: This follow-up report further suggests overall survival may be improved for patients with adenocarcinoma in the pancreatic head using an adjuvant interferon-based chemoradiation protocol. These results are obtained despite a high incidence of node involvement and advanced tumor stage. From this limited patient series, the actuarial 2-year and 5-year overall survival rates suggest a potential for improved long-term survival. Further study of this regimen in a multiinstitutional setting is needed.

Background: The end result of leakage of pancreatic juice into the peripancreatic space can be sterile necrosis, infected necrosis, or rupture into an adjacent hollow viscus or blood vessel (eg, colon, small bowel, or pseudoaneurysm). If a pancreatic duct (PD) leak is present, should treatment be aimed at minimizing the sequela of the leakage of pancreatic juice and not just supportive observation until a necrosectomy is required?, Methods: In 144 patients with severe pancreatitis we investigated whether the presence of a PD leak was associated with necrosis and also asked if PD leak might predict other outcomes such as a length of stay (LOS), mortality, and need for surgery. Furthermore, we questioned whether the use of endoscopic retrograde cholangiopancreatography (ERCP) to search for a PD leak might worsen the clinical outcome because of the potential for introducing microorganisms into an undrained space or exacerbating pancreatitis., Results: The presence of a demonstrable pancreatic duct leak was observed in 37% of patients and was significantly associated with both a higher incidence of necrosis and prolonged LOS (> or = 20 days). These patients were 3.4 times more likely to have necrosis and 2.6 times more likely to have a prolonged LOS. When treated with a combination of percutaneous drains, pancreatic duct stents, and surgery as necessary, a PD leak (even with its higher necrosis rate) was not significantly correlated with either mortality or the need for necrosectomy. The use of ERCP was not associated with LOS, mortality, or the need for necrosectomy, provided discovered PD leaks were immediately drained., Conclusions: A PD leak is common in patients with pancreatic necrosis but it is also important to locate and decompress in order to impede progression of the disease and keep mortality low.

Background: In a small group of patients with acute pancreatitis, Balthazar and Ranson demonstrated the applicability of computed tomography (CT) criteria to predict mortality. Building upon their work with a larger group of patients with acute pancreatitis, we set out not only to demonstrate that the CT severity index can predict death, but also length of hospital stay and need for necrosectomy., Methods: We reviewed all patients admitted to our hospital in the years 1992 to 1997 with a primary diagnosis of acute pancreatitis. Entrance criteria required that a CT scan had been performed during the hospitalization. The index CT scan was used to determine a CT severity index (the CTSI of Balthazar and Ranson). Outcomes measured were death, length of stay (LOS), and need for necrosectomy (NEC). Statistical analysis was performed using Fisher's exact and chi-square tests where appropriate., Results: Between the years 1992 to 1997, 886 patients had 1,774 admissions for acute pancreatitis, of which 268 had a CT scan performed and were entered into our study. These 268 patients had a mean age of 57 years, a mean LOS of 16 days (1 to 118), and a mean CTSI of 3.9 (0 to 10). Overall mortality was 4% (n = 11). A CTSI >5 significantly correlated with death (P = 0.0005), prolonged hospital stay (P <0.0001), and need for necrosectomy (P <0.0001). Patients with a CTSI >5 were 8 times more likely to die, 17 times more likely to have a prolonged hospital course, and 10 times more likely to undergo necrosectomy than their counterparts with CT scores <5., Conclusions: These data show that the CTSI is an applicable and comparable predictor of outcomes in severe pancreatitis.

Background: Since 1980 a group of pancreatic tumors have been termed intraductal papillary mucinous tumors (IPMT). Because these tumors occupy an intraductal position they are demonstrated by pancreatography to reside in the main pancreatic duct (MPD) or side branch ducts (SBD). Lesions of IPMT result in abdominal pain or pancreatitis symptoms because mucin production or papillary growth results in ductal obstruction. Only 104 cases had been reported in the literature by 1996 but more are being presented in abstract form. We reviewed our own 33 cases to assist defining operative decision-making criteria., Methods: All cases of IPMT between 1989 and 1997 were reviewed for clinical presentation, anatomy by endoscopic retrograde cholangiopancreatography and computed tomography, histologic findings, and long-term outcomes., Results: Our cases were older (65 years) and presented with disease centered mainly in the head of the gland. Clinical presentation was epigastric pain (82%), pancreatitis (56%), weight loss (36%), diabetes (27%), and jaundice (9%). Operations were pancreatectomy in 31 (Whipple n = 15, total n = 5, distal n = 10, local n = 1), bypass only (n = 1), and no operation (n = 1). Malignancy was found in 14 of 33 (42%). Factors significantly associated (P <0.05 Fisher exact test) with malignancy were history of alcohol abuse or death from disease. Jaundice or presence in both MPD and SBD approached a significant association with malignancy but not abdominal pain, weight loss, diabetes, preoperative serum elevations of amylase, SGOT, CA-19-9, or CEA; diffuse MPD dilation, gland region, gross mucus in ducts or filling defects, cytology, calcifications, or a pancreatic mass. In 31 resected patients after a follow-up of 37 months (1 to 103) death had occurred in 6 of 13 malignant cases and 0 of 18 with benign disease. Three-year actuarial survival was 82% (all) and 56% (malignant). Symptom recurrence after resection was found in 6 of 31 at a mean of 13 months postoperatively and was associated with death from disease (P <0.05) or presence of pain preoperatively., Conclusion: Malignancy is common with IPMT and is more likely to be present with the clinical history of alcohol abuse or jaundice and if the tumor involves both the MPD and the SBD. The prognosis after resection is better than pancreatic cancer but the 19% recurrence of symptoms was equally seen with benign or malignant cases owing to residual disease in pancreatic remnants. The amount of resection should be extensive in patients likely to have malignancy (alcohol, jaundice, MPD+SBD). In those likely to redevelop symptoms, ie, those with preoperative pain, a careful assessment should be made via imaging studies for extent of disease.

Introduction: Fine needle aspirations (FNAs) and endoscopic retrograde cholangiopancreatography (ERCP)-guided brushings (BRUSH) are useful tools in the differentiation between malignant and benign disease of the pancreas. Once the decision to obtain a cytologic confirmation of one's clinical suspicion is made, the interpretation of the findings, especially an equivocal or negative cytology finding, can be unclear. This study seeks to evaluate the utility of cytologic studies in the evaluation of a patient with suspected pancreatic malignancy., Methods: A retrospective review of 224 cytologic reports, including 174 FNAs and 50 BRUSHs, from all pancreatic FNAs and BRUSHs performed between January 1989 and June 1995, was performed. Subsequent confirmation of the cytologic diagnosis was made either by histologic or strict clinical criteria., Results: Forty-three percent of the cytologic reports were read as malignant, all others reported as suspicious, atypical, or negative. All cytology studies read as malignant and all FNAs reported as suspicious were histologically or clinically confirmed to be malignant. Of those reported as atypical or negative, 55% and 49% were confirmed to be malignant. Both FNA and BRUSH were 100% specific, 75% sensitive, and 80% accurate., Conclusions: We conclude that a cytological diagnosis of malignant or suspicious is reliable and useful for further therapy planning in the patient suspected to have a pancreatic malignancy. The reason for the cytologic tests was a strong clinical suspicion, therefore, a high incidence of cancer was found in the patients with atypical or negative readings. A diagnosis of atypical or negative is equivocal and requires further diagnostic maneuvers, frequently including surgery, to make the definitive diagnosis.

Background: Cystic duct leak (CDL) appears to complicate laparoscopic cholecystectomy (LC) more often than open cholecystectomy (OC). No mention of CDL was found in a literature review that covered 48,822 OCs and their complications., Patients and Methods: Fifty-four patients who developed biliary tract injuries following LC were reviewed for: the time from LC to presentation, presenting symptoms, method of diagnosis, treatment, outcome, and follow-up., Results: Seventeen of 54 biliary tract complications (31%) were CDLs. The CDLs presented at a median of 4 days after LC with pain (76%) and nausea and/or vomiting (35%). Endoscopic retrograde cholangiopancreatography (ERCP) defined the diagnosis and the anatomy of the leak in 11 patients (65%). Biliary endoprosthesis placement was employed in 8 patients, with concomitant sphincterotomy in 5 (63%), and resolved CDL in every case. Seven (88%) of these patients were asymptomatic at a median interval of 10 months after stent retrieval. Six patients (35%) underwent reoperation. Five had laparotomy with ligation of the cystic duct stump and 1 underwent laparoscopic examination with reclipping of the cystic duct stump. Five (83%) were asymptomatic at a median follow-up of 26 months. CDLs may result from inaccurate clip placement, perforations proximal to the clips, and stump necrosis, as documented at reoperation., Conclusions: CDLs occur more frequently in LC than in the OCs reported in the literature. Most leaks require intervention. ERCP with stent placement is the diagnostic and therapeutic procedure of choice and has a high success rate of resolving leaks. To forestall CDLs, it is important to place clips accurately and avoid electrocautery in the vicinity of the cystic duct.

Pancreatic ascites, etiologically related to a leaking pseudocyst or ductal disruption, has been treated medically with hyperalimentation, somatostatin analog, and large-volume paracentesis. Surgery is ultimately required in more than 50% of such patients. Mortality figures in patients with pancreatic ascites approximate 15% to 25% with either treatment modality. We describe 4 patients who were found to have ductal disruptions in conjunction with pancreatic ascites who responded to transpapillary pancreatic duct endoprosthesis placement. There has been no recurrence of ascites in these patients at a mean follow-up of 12 months following stent-retrieval. Further evaluation of endoscopic therapy for pancreatic ascites appears warranted.

We collected the records of 81 patients with biliary tract injuries occurring during laparoscopic cholecystectomy (LC) who were referred to 3 referral centers during a 33-month (May 1990 to March 1993) period. All records were reviewed to provide data concerning the anatomy of the lesion induced, method of injury, timing of injury detection, role of intraoperative cholangiography (IOC), methods of treatment, and outcome of these injuries. Injuries were classified by our own method as follows: (1) cystic duct leaks (n = 15), (2) bile leaks and/or ductal strictures (n = 27), and (3) ductal transections or excisions (n = 39). Peak occurrence by quarter of the year was 4th quarter, 1990 (Lahey), and 3rd quarter, 1991 (Mason), and 1st quarter, 1992 (Mayo). The majority (62%) of the injuries were recognized after LC. At the time of LC, 31 of 81 (38%) injuries were recognized and converted to open procedures. Data regarding IOC were available in 63 of 81 (78%) cases. In patients in whom IOC was not performed, 14 of 38 (37%) operations were converted; if an IOC was obtained and interpreted correctly, 13 of 21 (62%) operations were converted. Primary repair was attempted in 11 leaks and/or strictures, but 36% required additional treatment. Primary repair was used in six transections or excisions, and 17% have required further intervention. In patients who had biliary-enteric bypass (BEB) performed outside (17) versus at the referral institution (29), 94% (16 patients) versus 0%, respectively, required additional operative (e.g., revision of a hepaticojejunostomy) or nonoperative (e.g., radiologic or endoscopic stenting or balloon dilation) procedures. When used as initial therapy or after a primary ductal repair, stents (with or without balloon dilation) resolved 100% of simple cystic duct leaks and 91% of leaks and/or strictures. In conclusion, the peak incidence of LC-related biliary injuries appears to have passed. A completed and correctly interpreted IOC increases the chance of detection of biliary injuries intraoperatively and should assist surgeons who use routine IOC. Nonsurgical techniques allow treatment of most simple cystic duct leaks, major ductal leaks and/or strictures, and postoperative BEB strictures, although follow-up is limited. The poor results of pre-referral BEB is not surprising since all of these patients were selected for referral because their treatments had not been successful.

We assessed the use of endoscopic retrograde cholangiopancreatography (ERCP) after laparoscopic cholecystectomy (LC) at our hepatobiliary referral center. This assessment included patients from outside institutions with post-LC problems. Between May 1990 and September 1992, we performed 522 LCs and 1,723 ERCP examinations. There were 78 patients who underwent 143 ERCP examinations after LC, 65% of whom were referred. ERCP findings were categorized as follows: normal results (8%), problems inherent to stone disease (65%), and iatrogenic injury (27%). The types of inherent problems were common bile duct (CBD) stones, pancreatitis, and papillary stenosis/microlithiasis. Within the CBD stone group, 5 of 26 patients also had papillary stenosis, and, within the pancreatitis group, 9 of 11 patients also had papillary stenosis, making papillary stenosis the most frequent observation (55%). Almost all of these patients (96%) required endoscopic papillotomy for successful treatment. The iatrogenic injury group was comprised of 21 patients, 16% of whom had cystic duct leak and 84% of whom had CBD injury. These patients required a variety of endoscopic procedures including endoscopic papillotomy (67%), CBD endoscopic stenting (76%), percutaneous drainage of biloma (29%), and percutaneous transhepatic biliary drainage (24%). Open surgical procedures after endoscopic assessment or treatment were required in only three patients in the iatrogenic group and in none in the inherent group. At this time, long-term follow-up is not possible with regard to biliary stricture. We conclude that the majority of problems after LC are either due to papillary stenosis/microlithiasis with or without CBD stones or to biliary injury. Both can be successfully diagnosed and treated with endoscopic techniques.

Gastropathy on the basis of mesenteric arterial ischemia can be masked in presentation as the typically more benign entities of gastritis, gastric ulceration, or gastric atony. Gastritis and ulceration are commonly associated with stress, hyperacidity, Helicobacter pylori infection, or medication injury. Gastric atony is less commonly seen and usually attributable to diabetes mellitus, vagotomy, or mechanical gastric outlet obstruction. Gastric ischemia as a cause of gastropathy is an underappreciated phenomenon with a particularly poor prognosis in which early diagnosis is essential to potentially successful intervention. Seven patients with ischemic gastropathy are described; all are women, aged 41 to 71 years, smokers, with hypertension. Nausea, vomiting, weight loss, and gastrointestinal bleeding were the common presenting symptoms. All patients had endoscopic or autopsy-proven gastric ulcerations or necrosis, and two patients had proven gastroparesis. Four of five patients with ischemic gastritis died within 3 months of diagnosis despite vascular reconstruction. The two patients with gastroparesis underwent aorto-celiac bypass and are well 9 and 20 months, respectively, after operation. Treatment results were distressingly unsatisfactory, especially in those patients in whom gastritis rather than gastroparesis was the presenting problem. Although the high mortality of mesenteric ischemia is well described, little documentation of gastric ischemia exists in the literature. This entity is generally not considered in the differential diagnosis of gastritis, ulceration, or gastroparesis. Empirically, an early diagnosis and treatment may improve the survival in this select patient group.

The purpose of this study was to validate Congo red (CR) testing as a means of determining the status of vagal innervation of the stomach after gastric surgery by comparing it with the sham feeding test (SF). Twenty-six patients with previous gastric operations were studied prospectively. The operations were vagotomy/antrectomy (n = 14), vagotomy/antrectomy/Roux-en-Y anastomosis (n = 4), proximal gastric vagotomy (n = 3), vagotomy/pyloroplasty (n = 2), and antireflux (n = 3). In each case, the CR test was performed prior to the SF test so that the endoscopist was unbiased. The CR test solution was sprayed on the gastric mucosa through a wash catheter and immediately aspirated. A two-dimensional map of the endoscopist's conception of the red and black mucosa was drawn, and areas were calculated using a computerized scanner. The SF test was performed as described previously, with sham/peak ratios (sham feeding stimulated acid output [SAO]/peak acid output [PAO]) greater than 0.1 indicating a positive test, or incomplete vagotomy. The CR test was negative in 11 patients and positive in 15 patients. SF was negative in 16 patients (SAO/PAO of less than 0.1) and positive in 10 patients. The results correlated in 18 of 26 (69%) patients (p = not significant). Positive sham feeding tests and negative Congo red tests were more reliable, whereas negative sham feeding tests and positive Congo red tests were less reliable. Except patients with Roux-en-Y anastomosis, patients with negative or weakly positive CR tests did not have recurrent ulcers or other symptoms attributable to acid hypersecretion. The CR test was easier to perform and preferred by the patients. We conclude that both tests provide valuable clinical information and that both tests should be available in the evaluation of postgastrectomy patients.

Pancreatic fluid collections (PFC) can be drained surgically or nonsurgically with endoscopic or radiologic techniques. To define subgroups of patients with PFC who would benefit from the new modalities, we reviewed a period (1977 to 1990) during which both surgical and nonsurgical invasive techniques were available. Patients with phlegmon or necrosis at initial diagnosis were excluded. Sixty-five patients (35 male, 30 female) underwent 1 or more drainage procedures. A mean postprocedure follow-up of 10.2 months was available for 59 patients. Initial management was nonsurgical in 80% of patients. Procedures in patients with follow-up comprised invasive nonsurgical drainage (n = 25), invasive nonsurgical drainage plus surgery (n = 22), and surgery only (n = 12). Results for each group, respectively, were: morbidity, 20%, 20%, and 24%; mortality, 8%, 5%, and 0%; and successful drainage, 92%, 82%, and 83%. The choice of management appeared to be based on etiology and radiologic characteristics. Patients with nonalcohol- and nonbiliary-associated pancreatitis without a radiographically defined wall were more common in the invasive nonsurgical group and were successfully treated without surgery. Nonsurgical invasive techniques are efficacious in the treatment of PFC in this subgroup of patients.

The insertion of pancreaticobiliary drains at the time of endoscopic retrograde cholangiopancreatography can provide access to the bile or pancreatic duct, decompress an obstructed biliary tree, and allow infusion of solubilizing agents to effect gallstone dissolution. Twenty-six patients underwent 32 attempts at drain placement with 90 percent success and 6 percent minor complication rates. The use of pancreaticobiliary drains should supplement traditional diagnostic and therapeutic modalities in patients with surgical lesions of the pancreas or biliary tree.