Fluid CEA Research Articles

Does Cyst Fluid CEA Level Correlate with Malignancy Potential in Mucinous Cystic Lesions of the Pancreas?

Purpose: To determine if cyst fluid CEA can be used to differentiate between benign and malignant cystic neoplasms of the pancreas. Methods: A retrospective case control study was conducted; all mucinous pancreatic cystic lesions that have been surgically resected at one institution from January 1, 2000, to June 1, 2008, were identified and reviewed. Information on cyst fluid CEA level was retrospectively gathered. Results: A total of 123 mucinous cystic lesions of the pancreas were resected at the Hospital of the University of Pennsylvania within this time frame. Of these, 91 patients underwent endoscopic ultrasound (EUS) evaluation prior to surgical resection. Of these, 30 patients had cyst fluid CEA evaluated. Table 1 describes cyst fluid CEA level by lesion type. There was no significant association identified between CEA level (per 1000 unit increase) and aggressive behavior of the lesion, defined as moderate dysplasia or higher (OR 1.004, 95% CI 0.974-1.035).Table: ResultsConclusion: While cyst CEA level can be used to differentiate between mucinous and non-mucinous cystic lesions of the pancreas, it does not allow the clinician to determine malignancy potential of mucinous cystic lesions of the pancreas.

Pancreatic Cysts - A Retrospective Analysis of Our Experience at UMMHC with Focus on Utility of K-Ras Gene Mutation Analysis

Purpose: Pancreatic cysts, which could represent malignant, pre-malignant, or benign lesions, are being detected more frequently due to improvements in imaging techniques. Although the natural history of cysts is not fully understood, accurate characterization of cysts is paramount to optimize patient care. Cyst fluid cytology, CEA marker, and more recently, K-ras gene analysis, have all been used in an attempt to characterize cysts. In this study, we undertook an analysis of our experience at UMMHC over the past seven years with different methods of predicting pancreatic cyst type, focusing specifically on the utility of K-ras gene analysis in improving clinical decision making. Methods: We requested the medical record numbers of all patients seen at UMMHC over the last seven years who were coded with ICD-9 code 577.2, which represents the diagnosis of pancreatic cyst or pseudocyst. We received 1066 unique cases from October, 2003 through February, 2011. Of these, we included in our study those cases that had K-ras analysis and/or came to a definitive diagnosis (surgical specimen), which included 85 unique cases. Some of these patients had one or more follow-up EUS exams, which was often accompanied by cyst fluid analysis, for an additional 42 cases. This provided us with 127 total cases. Results: K-ras gene mutation was analyzed in 40 cases (31%). Of these, 18 (45%) were positive, 21 (53%) were negative, and 1 (2%) was non-diagnostic. Of the 40 patients who were tested for K-ras mutation, 5 went on to have a definitive diagnosis. All 5 cases had pre-malignant lesions - three of which were in patients with a detected K-ras mutation. The remaining two cases tested negative for the mutation, but had cyst fluid CEA values of 11,000 ng/ml and 2,690 ng/ml. Of the 87 cases where K-ras was not analyzed, 47 (54%) went on to have surgery. Of these, the final pathological diagnoses included 13 (28%) benign lesions and 34 (72%) pre-malignant or malignant lesions. Of the 57 cases where cytology was unavailable or not diagnostic, K-ras analysis was performed 21 times, 6 of which tested positive for the mutation. Conclusion: All five cases that were analyzed for K-ras mutation and came to a definitive diagnosis showed pre-malignant pathology. Although two of these cases did not detect K-ras mutations, their CEA levels were elevated to well over ten times the upper limit of 192 ng/ml accepted for predicting mucinous cysts. On the other hand, 28% of cases that were not tested for K-ras and came to a definitive diagnosis had benign lesions. The data shown here is insufficient to draw any large-scale conclusions, and further studies on a larger scale are needed. Following these set of patients into the future may also elucidate the behavior of different cysts. Disclosure: Dr. Wahid Wassef: Boston Scientific - HonorariumTable: Table. K-Ras gene mutation status and relationship to other clinical data

Large Pancreatic Cystic Neoplasms Are Not Associated with Current Biomarkers of Malignancy

Purpose: Predictors of malignancy in pancreatic cystic neoplasms include symptoms, pancreatic main duct dilation, mural nodularity, and cyst size. Cysts ≥3 cm are believed to correlate with malignancy, and this impacts current management strategies and practice guidelines. Aim: To identify clinical, endosonographic, laboratory, and pathologic factors that are associated with pancreatic cyst size ≥3 cm or <3 cm. Methods: This is a retrospective study performed at a Pancreas Referral Center. All subjects underwent an endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) of pancreatic cystic neoplasms from 6/2006 to 5/2011. Data collected include: demographics (age, gender, race), clinical features (symptoms), EUS findings (size, septations, nodularity), laboratory data (cyst fluid CEA, k-ras, LOH, cytology), and surgical pathology (malignancy defined as high-grade dysplasia, carcinoma in situ, and invasive carcinoma). Univariate and multivariate logistic regression analyses were performed on these variables to measure association and identify predictive variables (SPSS version 16.0). Results: There were 239 patients with pancreatic cystic lesions evaluated with EUS-FNA of which 68/239 (28.5%) were ≥3 cm. Mean age was 63.7 yr with 66% (157/ 239) female and median follow-up 10 months. Evaluation of Pancreas Cyst Size: presence of symptoms (p = 0.001) and absence of k-ras mutation (p = 0.043) were associated with cysts ≥3 cm while malignant pathology was not related to cyst size (p = 0.315). For cysts ≥3 cm, 1.5% (1/68) were malignant while 4.7% (8/171) were in cysts <3 cm. Only symptoms (p = 0.008) remained significantly associated with cysts ≥3 cm in multivariate logistic regression analysis when controlling for age, gender, and race. Conclusion: Pancreatic cystic neoplasms ≥3 cm are not associated with malignancy and k-ras mutation compared to cysts <3 cm. Larger pancreatic cysts are more likely to be symptomatic. Improved biomarkers are necessary to guide management of pancreatic cystic neoplasms regardless of cyst size.

Fluid analysis prior to surgical resection of suspected mucinous pancreatic cysts. A single centre experience.

EUS-FNA cytology and fluid analysis are frequently utilized to evaluate pancreatic cysts. Elevated cyst fluid CEA is usually indicative of a mucinous pancreatic cyst but whether CEA or amylase values can subclassify various mucinous cysts is unknown. The purpose of this study is to determine whether cyst fluid CEA and amylase obtained by EUS-FNA can differentiate between mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms (IPMNs). Using our prospective hospital EUS and surgical databases, we identified all patients who underwent EUS of a pancreatic cyst prior to surgical resection, in the last 10 years. Cysts were pathologically sub-classified as MCNs or IPMNs; all other cysts were considered non-mucinous. Values of cyst fluid CEA and amylase were correlated to corresponding surgical histopathology and compared between the two groups. 134 patients underwent surgery for pancreatic cysts including 82 (63%) that also had preoperative EUS. EUS-FNA was performed in 61/82 (74%) and cyst fluid analysis in 35/61 (57%) including CEA and amylase in 35 and 33 patients, respectively. Histopathology in these 35 cysts demonstrated nonmucinous cysts in 10 and mucinous cysts in 25 including: MCNs (n=9) and IPMNs (n=16). Cyst fluid CEA (p=0.19) and amylase (p=0.64) between all IPMNs and MCNs were similar. Between branched duct IPMNs and MCNs alone, cyst fluid CEA (p=0.34) and amylase (p=0.92) were also similar. In this single center study, pancreatic cyst fluid amylase and CEA levels appeared to be of limited value to influence the differential of mucinous pancreatic cysts. Larger studies are recommended to evaluate this role further.

Utility of cyst fluid CEA cut-off of 192ng/ml in the diagnosis of mucinous cysts of the pancreas

IntroductionThe differentiation of mucinous from non-mucinous pancreatic cysts is important because of the malignant potential of the latter. EUS-FNA allows for high resolution imaging of pancreatic cysts as well as...

Sa1449 Endoscopic Ultrasound Guided Pancreatic Cyst Ablation With Ethanol and Paclitaxel: Will It Also Work in the Western Hemisphere?

Follow-up imaging after EUS-guided pancreatic cyst ablation with ethanol alone leads to resolution in about 1/3 of patients. Experience at a single Korean hospital reports that the addition of paclitaxel to ethanol (EUS-EP) increases resolution to 60-70% but this has not been confirmed elsewhere. To report initial results and complications from EUS-EP in a Western population. In a single center prospective cohort study, consecutive patients with a 1-5 cm pancreatic cyst and ≤5 septations (regardless of main pancreatic duct [PD] communication) were enrolled. Exclusion criteria: active pancreatitis, pancreatic necrosis, portal hypertension, ascites, coagulopathy or dilated PD. Baseline evaluation: CT or MRI, EUS imaging, cyst fluid cytology, CEA and DNA analysis (PathFinderTG®, RedPath IP, Inc; Pittsburgh PA). At initial ablation (EUS-EP1), cysts were lavaged for 3-5 minutes with 99% ethanol, re-aspirated and then injected with an equal volume of 2mg/mL paclitaxel (Bedford Laboratories; Bedford, OH) which was left in place. Three months later, repeat EUS and cyst ablation (EUS-EP2) was performed if cyst size was ≥10mm. Repeat ablation was considered on a case-by-case basis for patients with preceding ablation-induced pancreatitis. Three months after EUS-EP-2, pancreas protocol CT was performed at our hospital. Baseline imaging and post-ablation 3D volume CT images were interpreted by a single radiologist. Complete response (CR), partial response (PR) and a persistent cyst were defined using CT criteria: <5% of the original volume (OV), 5-25% of OV, and >25% of OV respectively. 18 patients (13F, median 65 yrs) with cysts located in the uncinate in 3, head in 5, neck 2, body in 5, and tail 3 were enrolled. Baseline median CT or MRI cyst diameter: 2.7 cm (1.5-4.6) and 3D volume: 6.4 cm3 (1.1-38). Median cyst CEA was 1307 ng/mL (range: 0.2-75508). Clinical diagnoses were: IPMN in 11, MCN in 6 and SCN 1. EUS-EP1 (n=18) was complicated by acute pancreatitis (AP) in 3 and abdominal pain in 2. EUS-EP2 was not done in two who refused follow up EUS (n=1) or had surgery (n=1). The remaining 16 returned for EUS- EP2, but injection was not done in 7 with cysts measuring <10 mm (n=5) or in patients with previous AP (n=2). In the 9 injected, abdominal pain without pancreatitis occurred in 2. Follow up CT (available in 12/16 to date) showed: a CR in 5/12 (42%), PR in 5/12 (42%) and persistent cyst in 2/12 (16%). In our initial ongoing experience, EUS-guided pancreatic cyst ablation with ethanol and paclitaxel led to a CT-defined CR in 5/12 (42%), PR in 5/12 (42%) and a persistent cyst in 2/12 (16%). Complication rates encountered were acute pancreatitis in 3/27 (11%) injections and self-limited abdominal pain in 4/27 (15%).

A matched analysis comparing the epidemiology of intraductal papillary mucinous neoplasms to standard pancreatic adenocarcinoma and healthy controls.

173 Background: The epidemiology of intraductal papillary mucinous neoplasms (IPMNs) of the pancreas is poorly defined. Methods: An epidemiologic questionnaire was administered to patients (pts) with IPMN (n=79), pancreatic adenocarcinoma (PC) (n=689) and healthy controls (n=307). Results were adjusted for age, gender and BMI. IPMN was defined either by surgical pathology (n=62) or characteristic endoscopic ultrasound appearance and cyst fluid CEA&gt;200 ng/ml (n=17). Results: In unadjusted analysis IPMN pts were more likely to be ≥ 70 years of age (OR 5.40 [2.88, 10.46]) when compared with PC pts (OR 2.82) and controls. After adjustment for age, gender and BMI, current tobacco smoking was associated with PC (OR 3.06 [1.78, 5.23]) but not IPMN. Pts with IPMN more often had diabetes mellitus for &gt;3 years compared with controls (OR 3.25 [1.45, 7.00], while pts with PC (OR 1.52 (0.86, 2.67]) did not. IPMN pts were more likely to have a history of hypercholesterolemia compared with controls (OR 1.77 [1.05-2.98]); this was not seen for PC pts (OR 1.16 [0.87-1.55]). A first degree relative with PC was not associated with increased risk for IPMN (OR 0.84 [0.27, 2.62]) or PC (OR 1.48 [0.82, 2.67]). Compared to PC, pts with IPMN were more likely to have a history of an unrelated cancer (OR 1.84 [1.08, 3.14]). Conclusions: Risk factors for IPMN and PC may differ. Compared to PC and control pts, IPMN patients were older; more often had diabetes mellitus and hypercholesterolemia; and did not currently smoke. IPMN was more often associated with a prior history of cancer than PC. No significant financial relationships to disclose.

Comparison of Preoperative Cytopathology, Tumor Markers and Molecular Analysis for the Diagnosis of Mucinous Pancreatic Cysts in Patients Undergoing Surgical Resection: 2010 Presidential Poster

Purpose: To prospectively evaluate the individual or combined accuracy of cytopathology, cyst fluid tumor markers (CEA) and molecular analysis to differentiate mucinous from non-mucinous cysts. Methods: Consecutive patients with pancreatic cysts referred to our institution for EUS-FNA who were determined have suspected MPCs and subsequently underwent surgical resection at our institution. Cyst fluid was aspirated for cytopathology, DNA analysis (RedPath®, Pittsburgh, PA) and CEA when quantity was sufficient. DNA quantity ≥40ng/ul and/or k-rasmutation and/or ≥2 allelic imbalance mutations and/or CEA>192 ng/ml were all criteria for a MPC. Surgical pathology from pancreas resection was considered the gold standard for diagnosis. Results: 27 patients underwent surgery following EUS-FNA. Final surgical pathology revealed mucinous lesions in 22 (all were intraductal papillary mucinous neoplasms (IPMN)), and non-mucinous lesions in 5 (serous cystadenoma in 2, retention cyst in 1, and pancreatic intraepithelial neoplasia (PanIN) in non-mucinous background in 2). Only one patient had invasive malignancy and another had high grade dysplasia on final pathology. Cytopathology was highly specific in MPC but with low sensitivity of 9%. Mean CEA levels in cyst fluid were numerically higher in the mucinous compared to the non-mucinous group but the difference didn't reach statistical significance due to small sample size (11799 ng/ml vs. 59 ng/ml, p=0.49). Molecular analysis (any criterion) had a sensitivity of 46% and specificity of 100% to detect mucinous lesions. K-ras mutation alone was the single most reliable molecular criterion with a specificity of 100%. K-ras mutation identified one additional mucinous cyst not identified as such by CEA. Combining all molecular markers with CEA improved sensitivity to 74% (table 1). Conclusion: In this series, the sensitivity of pancreatic cyst fluid CEA outperforms any of the individual molecular criteria evaluated for the diagnosis of MPCs. Therefore, we advocate initial screening of cyst fluid CEA prior to obtaining molecular studies. Molecular markers combined with cyst fluid CEA marginally increase sensitivity for MPC and could add diagnostic value in suspected malignant cysts; however, the limited number of malignant cysts in this study precludes a definitive conclusion. A larger sample size with more malignant cysts is required.Table

Evaluation of Cyst Fluid CEA Analysis in the Diagnosis of Mucinous Cysts of the Pancreas

BackgroundAlthough cyst fluid carcinoembryonic antigen (CEA; >192 ng/ml) is the preferred test for identifying mucinous pancreatic cysts, the data are more robust for mucinous cystic neoplasms (MCN) than for intraductal papillary mucinous neoplasms (IPMN). The role of cyst fluid CEA as a marker for either malignancy or malignant progression is uncertain. MethodsAll patients with pancreatic cysts who had undergone endoscopic ultrasound with cyst fluid CEA measurement between 2001 and 2009 were identified. Patient outcomes and pathology from operative resections were recorded. ResultsTwo hundred sixty-seven patients were identified; pathological diagnosis was obtained in 97. Mucinous cysts were identified in 66 of 97 (68%): benign IPMN, n = 42; malignant IPMN, n = 10; benign MCN, n = 12; malignant MCN, n = 2. CEA > 192 ng/mL had a sensitivity and specificity of 73% and 65% for identifying mucinous cysts; cyst fluid CEA was not associated with malignancy (p = 0.85). One hundred seventy-eight patients were managed with an initial non-operative strategy. Eight (4%) developed radiographic changes necessitating surgery; pathology demonstrated seven benign mucinous cysts and one retention cyst. CEA was not associated with radiographic progression (p = 0.37). ConclusionsCyst fluid CEA is a useful test for identifying mucinous cysts, including MCN and IPMN. In mucinous cysts, cyst fluid CEA is not associated with malignancy or radiographic progression.

T1377 Pancreatic Cyst Fluid CEA Concentration >1000 NG/Ml is Not an Indicator of Malignancy

T1377 Pancreatic Cyst Fluid CEA Concentration >1000 NG/Ml is Not an Indicator of Malignancy

T1368 Serial EUS-Guided Pancreatic Cyst Fluid CEA Measurement Leads to Reclassification in Many Patients With Cysts of Unknown Etiology

T1368 Serial EUS-Guided Pancreatic Cyst Fluid CEA Measurement Leads to Reclassification in Many Patients With Cysts of Unknown Etiology

Abstract 1734: MMP-2 as a putative biomarker in body fluid of gastric cancer

Abstract Carcinoembryonic antigen (CEA) and fluid cytology are widely used in patients with gastrointestinal malignancies for evaluating tumor status, but its sensitivity is low and unsatisfactory. In this study, we evaluated the possibility of MMPs, known to promote cancer progression through increasing cancer cell growth, migration, invasion, metastasis and angiogenesis, as a novel biomarker in the body fluid of advanced gastric cancer. Total 135 body fluids (105 ascites, and 24 pleural fluids) from metastatic or relapsed gastric cancer patients diagnosed with peritoneal carcinomatosis at Severance Hospital, Yonsei University from 2000 to 2009 were collected. To determine the cut-off expression level of MMPs, we performed zymography and ELISA using the conditioned media of HT-1080 at various protein concentrations. Ascitic CEA (aCEA) and pleural CEA (pCEA) was measured by EIA. MMP-2 and MMP-9 levels in patient samples were determined with ELISA. Finally, we compared the expression of MMPs with clinical parameters such as fluid CEA, cytology, and survival. Comparing MMPs activity and expression in HT-1080, positive levels of MMP-2 and MMP-9 were 8.59ng/mL and 0.14ng/mL, respectively. Out of 105 ascitic fluids, MMP-2 had a positivity of 93.7%, MMP-9 34.2%, and aCEA 82.9%. Ascitic fluid cytology of 95 patients showed postitive rate of 55.9%. Combining aCEA and MMP biomarkers, the positivity improved; the positivity of MMP-2 and aCEA was 99.1%, while MMP-9 and aCEA was 86.5%. Interestingly, positivity of MMP-2 in aCEA negative patients was 94.7% and 21.1% in MMP-9. To evaluate the role of MMPs as a prognostic marker, the duration of survival was analyzed from the date when ascites was detected until death. The overall median survival in all patients was 62 days. Comparing survival with MMPs expression level, we found a negative correlation between MMP-2 and survival. We observed that patients with a higher MMP-2 expression, ≥ 22.6ng/mL which showed significant MMP-2 activation in zymography, had a shorter survival with a median of 15 days, while lower expression group had a median of 35 days, P &amp;lt; 0.01. In 24 pleural samples, MMP-2 had a positivity of 66.7%, MMP-9 45.8%, pCEA 70.5%, and cytology was 66.7% (n=18). When we combined these markers, MMP-2 and pCEA together increased the positivity to 91.7%, MMP-9 and pCEA 79.2%, MMP-2, MMP-9, and pCEA 91.7%. In conclusion adding MMPs, especially MMP-2, to fluid CEA increased the positivity allowing us to evaluate clinical status more accurately. Also, higher MMP-2 expression showed poor prognosis in advanced gastric cancer with ascites, suggesting the possibility of MMP-2 as a prognostic marker. With further development, MMPs expression in body fluid can be used to detect residual micrometastasis or predict recurrence in peritoneal washing fluid or pelvic fluid. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1734.

Intraductal papillary mucinous neoplasm and the pancreatic incidentaloma

Asymptomatic pancreatic lesions (APL) are a commonly encountered problem in today's pancreatic surgical practices. Current literature regarding etiologies and incidence of APLs, particularly intraductal papillary mucinous neoplasm (IPMN), is presented. APLs constitute a wide spectrum of pathology (solid/cystic, benign/premalignant/malignant) but, overall, IPMN is now the most common diagnosis. The Sendai Guidelines and their function as a basis for risk stratification in branch duct IPMN are presented. The importance of traditionally analyzed cyst characteristics including size, presence of mucin or nodules and cyst fluid aspirate as indicators of malignancy is emphasized, noting also the potential correlation of main duct dilatation, thickened septae and elevated cyst fluid CEA with increased risk of malignancy. Current complication rates after resection of APLs are reviewed and found to be generally equivalent to those for symptomatic resections. A potential multidisciplinary treatment strategy is offered considering the costs of surgery versus repeated imaging or follow up endoscopy for these lesions. The decision for intervention is ultimately based on the Sendai Guidelines in the context of the individual patient.

Clinical value of tumour markers and serum-ascites albumin gradient in diagnosis of malignant ascites

Objective To investigate the clinical value of tumour markers and serum-ascites albumin gradient (SAAG) in diagnosis of malignant ascites. Methods One hundred and fourteen patients with ascites, who were admitted to the hospital between Jan. 2005 and Jan. 2008, were retrospectively reviewed. The patients were divided into malignant (n = 39) or benign (12 with tuberculosis and 93 with aseptic liver cirrhosis) ascites groups according to the etiology. The distribution of tumor markers (CEA, CA19-9 and CA125) and SAAG in both groups were analyzed and receiver operating characteristic (ROC) was constructed. Results The tumor markers and SAAG were found both in malignant ascites group and benign ascites group. The concentrations of CEA and CA19-9 in serum and ascites were higher in malignant ascites group than in benign ascites group. The SAAG in malignant ascites group was significantly lower than that in patients with liver cirrhosis (P 0. 05). There was no difference in level of CA125 in serum or ascites between malignant ascites group and benign ascites group (P>0.05). The area under the curve of ascitic fluid CEA, CA19-9 and SAAG were 0.79, 0.82 and 0.85, respectively. The cutoff values of ascitic fluid CEA, CA19-9 and SAAG were optimally chosen at 1.45 μ/L, 19.50 μ/L and 13. 50 g/L, respectively. The sensitivity and specificity were 66.7% and 78.1% in CEA, 74.4% and 84.8% in CA19-9, as well as 82.9% and 84.6% in SAAG.The combination of ascitic fluid CA19-9 with SAAG could increase the specificity to 97.14%, but decrease the sensitivity to 61.54%. Conclusion It is feasible to achieve optimum combination of biochemical indicators using ROC in differential diagnosis of malignant ascites from benign ascites. Key words: Ascites; Tumor markers; Receiver operating characteristics curve

Use of cyst fluid CEA, CA19-9, and amylase for evaluation of pancreatic lesions

Objectives The study goals were development of reference intervals and an interpretive algorithm for pancreatic cyst fluid tumor markers. Design and methods 442 pancreatic cyst fluids were tested for CEA, CA19-9, and amylase. Results CEA > 30 ng/mL discriminates mucinous from non-mucinous cysts. After CEA analysis, amylase and CA19-9 segregate non-mucinous and mucinous subtypes, respectively. Conclusions Pancreatic cyst fluid tumor markers supplement other diagnostic measures. This study provides estimated reference intervals and an algorithm for interpretation.

Pancreatic Cystic Neoplasm: The Role of Cyst Morphology, Cyst Fluid Analysis, and Expectant Management

Among pancreatic cysts, mucinous cystadenoma, and intraductal papillary mucinous neoplasms have the potential for malignant transformation. Differentiation between benign and potentially malignant/malignant (PMM) cysts remains difficult. The purpose of this study was to: (1) identify the diagnostic value of endoscopic ultrasound findings, serum, and cyst fluid tumor markers (CA19-9 and CEA), (2) determine the rate of subsequent surgical resection in patients initially managed conservatively, and (3) determine the role of cyst fluid viscosity "string sign" in differentiating pancreatic cysts. Patients with cytologic or pathologic diagnosis for pancreatic cystic neoplasms were analyzed. The study included 79 patients. Cyst fluid CEA had a median of 1.0 ng/mL in benign cysts and 471.1 ng/mL in PMM cysts (P < .0001). Cyst fluid CA 19-9 was not statistically significant (P = .22). Neither serum CA 19-9 nor CEA was useful (P = .68 and P = .31). Increased cyst fluid viscosity was associated with PMM cysts (P < .0001). Median string sign was 0 mm in benign cysts and 3.5 mm in PMM cysts. The presence of thick walls (5 of 5, 100%) or intracystic growth (6 of 6, 100%) were associated with PMM cysts. Of the 50 patients with PMM cysts, 19 were treated conservatively. In those patients followed for more than 6 months, 2 of 12 (16.7%) had surgical resection after a median of 29.5 months for worrisome changes on imaging. The presence of a thick cyst wall or intracystic growth, elevated cyst fluid CEA, and a long "string sign" were associated with PMM cysts. 16.7% of patients with a PMM cyst managed conservatively ultimately required surgical resection.

S1305 Does Fluid CEA Level Predict Malignancy in Intraductal Papillary Mucinous Neoplasms (IPMNs)?

S1305 Does Fluid CEA Level Predict Malignancy in Intraductal Papillary Mucinous Neoplasms (IPMNs)?

968 Performance Characteristics of Cyst Fluid CEA Analysis for the Diagnosis of Mucinous Cysts of the Pancreas

968 Performance Characteristics of Cyst Fluid CEA Analysis for the Diagnosis of Mucinous Cysts of the Pancreas

The Role of Molecular Analysis in Pancreatic Cystic Neoplasms

Introduction: Analysis of cyst fluid for DNA quantity, quality, Kras and loss of heterozygosity (LOH) mutations may contribute to the understanding of risk associated with pancreatic cystic neoplasms. Current guidelines identify lesions that are very low risk for neoplastic progression (serous cystadenomas, pseudocysts) and some that are very high risk (main duct IPMN or cystic lesions with associated mural nodule). However, less is known about suspected mucinous lesions that do not fall into either of these categories. Our goal was to examine the role of molecular markers in the diagnosis of these lesions. Methods: We retrospectively and prospectively collected data from patients who underwent EUS with FNA of pancreatic cysts. We excluded all patients with a recent history of pancreatitis, morphologic features of SCA, dilated MPD or mural nodule. Molecular analysis was performed by RedPath IP (Pittsburgh, PA) and lesions were categorized as serous or as one of three types of mucinous cysts (low, intermediate or high risk) based on DNA quantity, quality and Kras mutations and LOHs. Statistical analysis using chi square and Fisher's exact test was performed. Results: 115 cystic lesions, with average size 20.6 +/- 13 mm and CEA of 373 +/- 1079 were identified. By molecular category the majority were low risk (72%), 21% were intermediate and 7% were high risk. Molecular categories showed a positive association with CEA level (p<0.0001) but not with size. Final pathology, when available, confirmed high grade histology in 4/5 lesions with high-risk features but low grade MCA or IPMN in all indolent and in 3/4 intermediate risk lesions. One intermediate risk lesion had IPMN with CIS. In addition, of the indolent and intermediate lesions follow-up of at least 12 months was available for 35% (n=33) and only two lesions showed progression. Importantly, of the lesions with a CEA>800 or size> 3 cm and either indolent or intermediate molecular features none of the lesions that were operated had dysplasia. Conclusions. Our results show that there is a strong correlation between fluid molecular categories and CEA but not size. High risk molecular features are very suggestive of high grade dysplasia or carcinoma. On the contrary, indolent molecular features, even in a patient with elevated CEA, may identify a subset with more benign pathology. Therefore, these markers may be of added value in identifying the highest risk lesions and can also be used when an unexpectedly high CEA is found in a small otherwise morphologically low risk cystic lesion.

Do Cyst Fluid CEA and Amylase Levels Help Differentiate Between Mucinous Pancreatic Cysts?

Do Cyst Fluid CEA and Amylase Levels Help Differentiate Between Mucinous Pancreatic Cysts?