Outcome of pancreaticoduodenectomy in octogenarians: Single institution's experience and review of the literature.

INTRODUCTION: Pancreatic and perampullary neoplasms in patients aged 80 or older trouble the surgeons because of the risk of surgical treatment. We have reviewed our experience and literature's reports of pancreaticoduodenectomy in octogenarians, evaluating early results and long-term survival in pancreatic cancer group. METHODS: Three hundred eighty-five patients who underwent pancreaticoduodenectomy for neoplasms from 1998 to 2011 were included in the study, and were divided in two groups: group 1, patients younger than 80 years of age, and group 2, patients 80 years of age and older. Operative morbidity, mortality, disease-free and long-term survival were analysed. English literature was systematically searched for pancreatic resection's outcome in octogenarians. RESULTS: There were 385 pancreaticoduodenectomies: 362 patients were in group 1 and 23 patients in group 2. There was no significant difference regarding gender, and pathologic findings between the two groups. Complications' rate (40 vs. 43%), mortality rate (4% vs. 0%), and overall median survival for pancreatic cancer patients were not statistically different in the two groups (median 21 vs. 19 months). Literature's review showed 14 reports of pancreatic resection in octogenarians. Most of the studies (particularly in centres with high-volume pancreatic surgery) showed that outcome after pancreatectomy was not different in octogenarians or in younger patients. CONCLUSION: Pancreaticoduodenectomy is an acceptable option for elderly patients. Age alone should not be considered a contraindication to major pancreatic resection, but a careful preoperative evaluation and an accurate postoperative management are mandatory.

Heat-induced transcription of diphtheria toxin A or its variants, CRM176 and CRM197: implications for pancreatic cancer gene therapy.

Vectors combining the heat shock proteins (HSPs) promoter with the catalytic subunit A of the diphtheria toxin (DTA) or its variants, cross-reacting material (CRM) 176 and 197, were engineered to investigate the effect of bacterial toxins on pancreatic cancer (PC) cells. Three heat-inducible enhanced green fluorescent protein (eGFP)-expression vectors were obtained: V1 (91% homology to HSPA6), V2 (five heat shock elements upstream the minimal HSPA6 promoter) and V3 (V1 and V2 combined). The highest eGFP transcription and translation levels were found in V3 transfected PC cells. The V3 promoter was used to control DTA, CRM176 and CRM197 expression, treatment response being investigated in four PC cell lines. DTAwt or CRM176 transfected cell growth was completely arrested after heat shock. CRM197 toxin presumed to be inactive, caused mild distress at 37 degrees C and induced a 25-50% reduction in cell growth after heat shock. Preliminary in vivo findings showed that heat treatment arrests tumor growth in DTA197 stably transfected PSN1 cells. In conclusion, the efficient HSP promoter identified in this study may be extremely useful in controlling the transcription of toxins such as CRM197, which have lethal dose-related effects, and may thus be a promising tool in PC gene therapy in vivo.

The management of pancreatic cancer. Current expert opinion and recommendations derived from the 8th World Congress on Gastrointestinal Cancer, Barcelona, 2006.

This article summarizes the expert discussion on the management of pancreatic cancer, which took place during the 8th World Congress on Gastrointestinal Cancer in June 2006 in Barcelona. A multidisciplinary approach to a patient with pancreatic cancer is essential, in order to guarantee an optimal staging, surgery, selection of the appropriate (neo-)adjuvant strategy and chemotherapeutic choice management. Moreover, optimal symptomatic management requires a dedicated team of health care professionals. Quality control of surgery and pathology is especially important in this disease with a high locoregional failure rate. There is now solid evidence in favour of chemotherapy in both the adjuvant and palliative setting, and gemcitabine combined with erlotinib, capecitabine or platinum compounds seems to be slightly more active than gemcitabine alone in advanced pancreatic cancer. There is a place for chemoradiotherapy in selected patients with locally advanced disease, while the role in the adjuvant setting remains controversial. Those involved in the care for patients with pancreatic cancer should be encouraged to participate in well-designed clinical trials, in order to increase the evidence-based knowledge and to make further progress.

Extent of lymphadenectomy in the resection of pancreatic cancer. Analysis of the existing evidence.

Pancreaticoduodenectomy is considered the standard procedure for the surgical treatment of the pancreatic head cancer. However, the extent of lymph node clearance associated to the procedure is still largely debated. Arguments in favour of an extended lymphadenectomy are the regular progression of lymph node invasion, without skip metastases, and the removal of the extrapancreatic neural plexus that is invaded in 52-72% of patients. Arguments against the extended lymphadenectomy are the failure of extended lymphadenectomy to improve survival in other cancers, and the severe diarrhoea that follows the skeletonisation of the superior mesenteric artery. After Ishikawa's paper, several retrospective studies supported a longer survival after an extended than after a standard lymphadenectomy, but as much retrospective studies failed to demonstrate any difference. Only three prospective randomised controlled trials have been performed so far. Unfortunately all are underpowered, and the substantial differences in the surgical procedures, in the adjuvant treatment, and in the length of follow-up make the comparison impossible. Only one study reports a significantly longer survival for lymph node positive patients who underwent an extended lymphadenectomy, but adjuvant treatment was not performed. Furthermore, the difference was of minimal clinical impact. At least two adequately powered prospective Randomised Controlled Trials including a true extended lymphadenectomy, and a standardised adjuvant treatment, would be required to answer the question. Unfortunately, we have not yet a standardised adjuvant (or neoadjuvant) treatment, and we do not know the impact of such treatment on the expected statistical difference in the survival after a standard or extended lymphadenectomy. The lot of work required to perform such trials probably doesn't worth the expected results.

Killer genes in pancreatic cancer therapy.

This review describes: 1. The main genetic alterations found in pancreatic cancer (EGF-R overexpression, SST-2 somatostatin receptor loss of expression, k-ras, p53 mutations and DPC4 mutations) and the effect of their replacements by gene therapy on tumor growth; 2. The use of suicide genes (HSV-TK and CD) for pancreatic cancer gene therapy in vitro and in vivo; 3. The implications for pancreatic cancer treatment when using cytotoxic bacterial toxins; 4. Viral and non-viral delivery systems for the transfer of therapeutical genes into pancreatic cancer cells. Overall both the correction of pancreatic cancer cells main genetic alterations and the use of suicide genes allow only partial tumor regression in vitro and in vivo. The lack of a 100% effect for any studied strategy considered alone, indicates the need for combined therapies to achieve a satisfactory treatment of this tumor.

Pancreatic cancer cells invasiveness is mainly affected by interleukin-1beta not by transforming growth factor-beta1.

BACKGROUND: We investigated in vitro whether IL-1beta and TGF-beta1 affect pancreatic cancer cell growth, adhesion to the extracellular matrix and Matrigel invasion. MATERIALS AND METHODS: Adhesion to fibronectin, laminin and type I collagen, and Matrigel invasion after stimulation with saline, IL-1beta and TGF-beta1 were evaluated using three primary and three metastatic pancreatic cancer cell lines. RESULTS: Extracellular matrix adhesion of control cells varied independently of the metastatic characteristics of the studied cell lines, whereas Matrigel invasion of control cells was partly correlated with the in vivo metastatic potential. IL-1beta did not influence extracellular matrix adhesion, whereas it significantly enhanced the invasiveness of three of the six cell lines. TGF-beta1 affected the adhesion of one cell line, and exerted contrasting effects on Matrigel invasion of different cell lines. CONCLUSIONS: IL-1beta enhances the invasive capacity of pancreatic cancer cells, whereas TGF-beta1 has paradoxical effects on pancreatic cancer cells; this makes it difficult to interfere with TGF-beta1 signaling in pancreatic cancer treatment.

[Enucleation-resection of pancreatic neuroendocrine tumors: 25 years of experience]. / Enucleoresezione di tumori neuroendocrini pancreatici venticinque anni di esperienza.

From 1980 to 2004, out of 109 patients who underwent surgery for neuroendocrine pancreatic tumor, 33 had a simple tumor excision. Seventy-two percent of cases were insulinomas. Age, sex, site and size of the tumor, associated diseases, hospital stay and complications were retrospectively reviewed by the clinical records. Patients (12 males and 21 females) averaged 56.8 years, range 20-86. Mean size of the tumor was 1.7 cm and 54.5% were in the pancreatic head; 78.8% of cases had medical associated diseases. Hospital stay was 12 days (median; range, 6-81 days) and mean period of gastric suction was 4 days. Forty-eight percent had a uneventful postoperative course. Complications were divided in early (related to pancreatic surgery, related to general open surgery and medical) and late events. Complication related to pancreatic surgery were 6/33 (18%); 5 pancreatic fistulas (4 low output) and 1 acute pancreatitis, while 5/33 had a general surgery complication (2 leacking due to gastric and duodenal associated operations). Medical complications were recorded in 13 cases. Late complications occurred in 4 cases (2 incisional hernias, 1 pseudocyst and 1 keloid). No patient was re-operated for pancreatic complications; 1 was reoperated for evisceration and 1 for hyper-parathyroidism in the early post-operative period. No mortality occurred. Re-evaluation of the clinical records in order to be submitted to laparoscopic surgery excluded 17/33 cases (51%) as candidate to laparoscopic approach.

[Positron emission tomography with fluorodeoxyglucose in gastro-entero-pancreatic tumors: diagnostic role and prognostic implications]. / La tomografia ad emissione di positroni con fluorodesossiglucosio nei tumori neuroendocrini gastro-entero-pancreatici: ruolo diagnostico e implicazioni prognostiche.

From November 1994 to November 2004, seventy-seven patients with neuroendocrine gastro-entero-pancreatic tumor (71% pancreatic) were investigated with 18-fluorine-deoxi-glucose positron emission tomography (FDG-PET). PET results were compared with CT-scan, MRI and octreoscan scintigraphy and clinico-pathologic features of patients and survival. Overall PET sensitivity was 57%; 78% of malignant tumors, 67% of borderline and 17% of benign tumors were detected by FDG-PET. No duodenal tumor was detected by PET scan. Only 16% of primary less than 2 cm in size was localized. In 16% of cases PET scan provided new information able to change therapeutic management. In PET positive patients the addictive information obtained by PET scan when compared with octreoscan, MRI and CT scan were respectively 50% more, 26% more and 30% more. In malignant neuroendocrine tumors PET positivity was related to short survival. No patient with malignant tumor died for disease progression in the follow-up when PET was negative, while 13/35 PET positive patients died (p <0.003). FDG-PET proved to be a second line technique in neuroendocrine digestive tumors. PET results improve clinical staging of disease and is related to survival in malignant cases; in 16% of cases may change the therapeutic option.

Altered glucose metabolism and proteolysis in pancreatic cancer cell conditioned myoblasts: searching for a gene expression pattern with a microarray analysis of 5000 skeletal muscle genes.

BACKGROUND AND AIMS: We verified whether conditioned media (CM) from pancreatic cancer cell lines (MIAPaCa2, CAPAN-1, PANC-1, BxPC3) alter glucose metabolism and gene expression profiles (microarray experiment with a platform of 5000 skeletal muscle cDNA) in mice myoblasts. METHODS: Myoblasts were incubated with control or pancreatic cancer CM for 24 and 48 hours. RESULTS: Lactate significantly increased in CM compared with non-conditioned myoblasts. No variations in expression levels of the main genes involved in glycolysis were found in CM myoblasts. Propionyl coenzyme A carboxylase and isocitrate dehydrogenase 3 beta genes, which encode enzymes of the tricarboxylic acid cycle, were overexpressed, while IGFIIR and VAMP5 genes were underexpressed in CM myoblasts. PAFAH1B1 and BCL-2 genes (intracellular signal transduction) and the serine protease cathepsin G (proteolysis), were overexpressed in CM myoblasts. Tyrosine accumulation in CM myoblasts suggested that proteolysis overcomes protein synthesis. Sorcin, actin alpha, troponin T1, and filamin A were underexpressed in CM myoblasts. CONCLUSIONS: Our findings demonstrate that pancreatic cancer cell conditioned media enhanced lactate production and induced proteolysis, possibly by altering expression levels of a large number of genes, not only those involved in protein biosynthesis and degradation or glucose metabolism, but also those involved in the tricarboxylic acid cycle and in vesicle traffic.

Suicide gene therapy with HSV-TK in pancreatic cancer has no effect in vivo in a mouse model.

AIM: To study in vivo whether pancreatic cancer tumour growth and metastasis can be modified by a gene construct with HSV-TK suicide gene and IL2 co-expression. METHODS: Seventy-eight female SCID mice were i.p. inoculated with retrovirally transduced or control MIA PaCa 2, CAPAN-1 and PANC-1 cell lines. The animals were then randomly selected for saline or ganciclovir (GCV) treatment from the second week, for a total of two weeks. RESULTS: Most inoculated mice developed tumour nodules and spleen metastases. The liver was colonized by control CAPAN-1 and MIA PaCa 2, but not by PANC-1. Tumours in transduced MIA PaCa 2 cell injected mice were smaller, and in transduced CAPAN-1 injected mice larger, than in control-inoculated mice. There were increased pancreatic and decreased spleen metastases from transduced CAPAN-1, and diminished liver involvement from transduced MIA PaCa 2. No differences were found between mice inoculated with transduced and control PANC-1 cell lines. GCV treatment had no effect on tumour's size or metastases. CONCLUSIONS: The HSV-TK suicide gene does not confer GCV sensitivity to pancreatic cancer in this in vivo model. Different pancreatic cancer cell lines cause different growth and metastasis patterns after inoculation in SCID mice, possibly because of variations in their inherent characteristics. The different effects of our vector on cell growth and metastasis may be attributable to the effects of the immunostimulatory cytokine IL2.

Epidemiology of pancreatic cancer in Northeastern Italy: incidence, resectability rate, hospital stay, costs and survival (1990-1992).

BACKGROUND: The incidence of pancreatic cancer and relative hospital stay and costs are not well known. AIMS: To define the incidence, hospital stay and cost of pancreatic cancer in a well-defined area of Italy. PATIENTS AND METHODS: Each new case of pancreatic cancer diagnosed between 1990 and 1992 among 669,703 inhabitants in the Veneto Region of Northern Italy was recorded and followed until death or for 5 years after diagnosis. Four types of hospital stay were defined. Type 1: undiagnosed pancreatic cancer; type 2: first diagnosis of pancreatic cancer, treatment excluded; type 3: main treatment; and type 4: follow-up and disease-related complications. Data were analysed for hospital stay-related procedures, costs and survival. RESULTS: Pancreatic cancer was diagnosed in 253 patients (12.6/100,000 per year), 43 patients (17.7%) underwent surgical resection, and 93 (36.8%) palliative surgery. The mean duration of type 3 hospital stay was similar for resection, palliative and exploratory surgery. The estimated hospital cost was significantly higher for surgical resection, almost the same for palliative and exploratory surgery, and only slightly lower for medical treatment. Each patient spent a mean of 57.7 days in the hospital. The hospital mortality rate was 4.6% for surgical resection, 22.1% for palliative surgery, and 18.7% for exploratory laparotomy. Overall, the 1-, 2-, 3- and 5-year survival rates were 20.9%, 5.1%, 2.9% and 1.2%, respectively. CONCLUSIONS: Pancreatic cancer is an expensive, almost incurable disease. Integrated treatments in specialized Centres should reduce the mortality rate and costs.

Influence of resection margins on survival for patients with pancreatic cancer treated by adjuvant chemoradiation and/or chemotherapy in the ESPAC-1 randomized controlled trial.

OBJECTIVE: To assess the influence of resection margins on survival for patients with resected pancreatic cancer treated within the context of the adjuvant European Study Group for Pancreatic Cancer-1 (ESPAC-1) study. SUMMARY BACKGROUND DATA: Pancreatic cancer is associated with a poor long-term survival rate of only 10% to 15% after resection. Patients with positive microscopic resection margins (R1) have a worse survival, but it is not known how they fare in adjuvant studies. METHODS: ESPAC-1, the largest randomized adjuvant study of resectable pancreatic cancer ever performed, set out to look at the roles of chemoradiation and chemotherapy. Randomization was stratified prospectively by resection margin status. RESULTS: Of 541 patients with a median follow-up of 10 months, 101 (19%) had R1 resections. Resection margin status was confirmed as an influential prognostic factor, with a median survival of 10.9 months for R1 versus 16.9 months months for patients with R0 margins. Resection margin status remained an independent factor in a Cox proportional hazards model only in the absence of tumor grade and nodal status. There was a survival benefit for chemotherapy but not chemoradiation, irrespective of R0/R1 status. The median survival was 19.7 months with chemotherapy versus 14.0 months without. For patients with R0 margins, chemotherapy produced longer survival compared with to no chemotherapy. This difference was less apparent for the smaller subgroup of R1 patients, but there was no significant heterogeneity between the R0 and R1 groups. CONCLUSIONS: Resection margin-positive pancreatic tumors represent a biologically more aggressive cancer; these patients benefit from resection and adjuvant chemotherapy but not chemoradiation. The magnitude of benefit for chemotherapy treatment is reduced for patients with R1 margins versus those with R0 margins. Patients with R1 tumors should be included in future trials of adjuvant treatments and randomization and analysis should be stratified by this significant prognostic factor.

Serum protein profiles of patients with pancreatic cancer and chronic pancreatitis: searching for a diagnostic protein pattern.

In this study, 13 sera from patients with pancreatic cancer, 9 from chronic pancreatitis and 10 from healthy subjects were analyzed by matrix-assisted laser desorption/ionization (MALDI) mass spectrometry. The MALDI mass spectra revealed the presence of several low molecular weight peptides, among which some were detected only in the sera from both pathological conditions. On the other hand many peptides were observed only in control sera, and were absent in the sera from the two diseases. Therefore, MALDI analysis of the low molecular weight fraction (<10 000 Da) of sera from patients with pancreatic diseases enabled us to identify the presence of some disease-related signals and also some signals characteristic of normal subjects.

Pancreatic head resection for noninflammatory benign lesions of the head of the pancreas.

INTRODUCTION: Duodenum-preserving pancreatic head resection (DPPHR) has been safely performed in patients with chronic pancreatitis. The procedure has rarely been used to remove benign or borderline lesions of the head of the pancreas. AIMS: To review our experience with 13 patients who underwent DPPHR and to review reports in the literature on the same subject. METHODOLOGY: From October 1991 to September 2000, 13 patients underwent DPPHR to resect endocrine pancreatic tumors (n = 4), beta cell hyperplasia (n = 1), pancreatic pseudocysts (n = 2), serous cystadenomas (n = 3), congenital (n = 1) and choledochal (n = 1) cysts, and intraductal papillary mucinous tumor (n = 1). The Kocher maneuver was performed in seven patients (group 1) and avoided in six (group 2). Type 1, 2, and 3 DPPHR were defined depending on the amount of pancreatic tissue left at the inner surface of the duodenum. Ten patients underwent evaluation that included an oral glucose tolerance test and exocrine pancreatic function test. RESULTS: The mortality rate was zero; the complication rate was 69%. Patients in whom the Kocher maneuver was not performed (group 2) experienced fewer complications, shorter stay on nasogastric tube and abdominal drain(s), and earlier water intake and discharge. Type of DPPHR did not influence the postoperative course. One patient died 3 months after surgery of unrelated disease. Twelve patients were alive and well 2 months to 8 years after surgery. CONCLUSION: DPPHR is a low-risk procedure in patients with benign or borderline noninflammatory lesions of the head of the pancreas in whom pylorus-preserving pancreaticoduodenectomy is otherwise indicated. Whenever possible, the Kocher maneuver should be avoided.

Value of 18-fluorodeoxyglucose positron emission tomography in the management of patients with cystic tumors of the pancreas.

OBJECTIVE: To assess the reliability of 18-fluorodeoxyglucose positron emission tomography (18-FDG PET) in distinguishing benign from malignant cystic lesions of the pancreas. SUMMARY BACKGROUND DATA: The preoperative differential diagnosis of cystic lesions of the pancreas remains difficult: the most important point is to identify malignant or premalignant cysts that require resection. 18-FDG PET is a new imaging procedure based on the increased glucose metabolism by tumor cells and has been proposed for the diagnosis and staging of pancreatic cancer. METHODS: During a 4-year period, 56 patients with a suspected cystic tumor of the pancreas underwent 18-FDG PET in addition to computed tomography scanning, serum CA 19-9 assay, and in some instances magnetic resonance imaging or endoscopic retrograde cholangiopancreatography. The 18-FDG PET was analyzed visually and semiquantitatively using the standard uptake value. The accuracy of 18-FDG PET and computed tomography was determined for preoperative diagnosis of a malignant cyst. RESULTS: Seventeen patients had malignant tumors. Sixteen patients (94%) showed 18-FDG uptake with a standard uptake value of 2.6 to 12.0. Twelve patients (70%) were correctly identified as having malignancy by computed tomography, CA 19-9 assay, or both. Thirty-nine patients had benign tumors: only one mucinous cystadenoma showed increased 18-FDG uptake (standard uptake value 2.6). Five patients with benign cysts showed computed tomography findings of malignancy. Sensitivity, specificity, and positive and negative predictive values for 18-FDG PET and computed tomography scanning in detecting malignant tumors were 94%, 97%, 94%, and 97% and 65%, 87%, 69%, and 85%, respectively. CONCLUSIONS: 18-FDG PET is more accurate than computed tomography in identifying malignant pancreatic cystic lesions and should be used, in combination with computed tomography and tumor markers assay, in the preoperative evaluation of patients with pancreatic cystic lesions. A positive result on 18-FDG PET strongly suggests malignancy and, therefore, a need for resection; a negative result shows a benign tumor that may be treated with limited resection or, in selected high-risk patients, with biopsy, follow-up, or both.

Multiple endocrine neoplasia type 1 and adrenal lesions.

PURPOSE: We investigated the relationship of long-term pancreatic hormone hypersecretion with adrenal lesions in patients with multiple endocrine neoplasia type 1 and in those with sporadic pancreatic endocrine tumors. MATERIALS AND METHODS: We assessed the prevalence of adrenal lesions in 20 patients with multiple endocrine neoplasia type 1 and in a control group of 12 with sporadic pancreatic endocrine tumors. We also performed genetic testing for germline mutations of MEN1, the multiple endocrine neoplasia type 1 gene. RESULTS: Adrenal lesions were common in multiple endocrine neoplasia type 1, accounting for 35% of cases. All adrenal lesions were nonfunctioning and benign. The relative risk of adrenal tumors was higher in patients with multiple endocrine neoplasia type 1 than in controls (p <0.05). No apparent relationship was observed of hormonal pattern or genotype with adrenal disease. CONCLUSIONS: Hormone hypersecretion by pancreatic endocrine tumors is not the primary cause of the development of adrenal lesions and the role of the MEN1 gene in adrenal tumorigenesis remains unclear. Adrenal lesions follow a benign course in most multiple endocrine neoplasia type 1 cases but careful morphological and functional followup is advisable.

ME-PCR for the identification of mutated K-ras in serum and bile of pancreatic cancer patients: an unsatisfactory technique for clinical applications.

Our aim was to assess the clinical reliability of mutated K-ras detection in serum or bile for the diagnosis of pancreatic cancer using ME-PCR. DNA was extracted from 1 ml serum obtained from 29 patients with pancreatic cancer and 12 control subjects. ME-PCR was optimized using a mixture of normal DNA added with different amounts of mutated DNA. The analysis of sera obtained from the 29 patients and of bile obtained from 11 pancreatic cancer patients demonstrated the presence of mutated K-ras in two (6.9%) and four cases (36%). By contrast K-ras was not amplifiable in any of the 12 serum samples obtained from healthy controls. In conclusion the DNA obtained from pancreatic cancer patients' sera is suitable for K-ras amplification and for the identification of codon 12 point mutations. However ME-PCR alone has an unsatisfactory sensitivity for the detection of pancreatic cancer using serum DNA as starting template.

CEA mRNA identification in peripheral blood is feasible for colorectal, but not for gastric or pancreatic cancer staging.

OBJECTIVE: It has been suggested that the molecular identification of cancer cells in the circulation may be useful in predicting the presence of micrometastasis in several cancer types. The aim of the present study was therefore to assess the feasibility of CEA mRNA identification in blood for diagnosing and staging colorectal, gastric and pancreatic cancer. METHODS: We studied 16 control subjects, 69 patients with colorectal (CRC), 30 with gastric (GC), 27 with pancreatic cancer (PC) and 8 with benign diseases of the pancreatobiliary tree. At diagnosis CEA mRNA was identified in peripheral blood by means of a RT-PCR procedure. RESULTS: The specificity of this test in control subjects was 94%, and its sensitivity in identifying CRC, GC and PC were 34, 37 and 41%, respectively. False-positive findings were recorded in 25% patients with benign diseases. No association was found between CEA mRNA and stage in patients with GC or PC. In CRC patients, positive CEA mRNA findings were correlated with local spread (chi(2) = 14.6, p<0.01), lymph node (chi(2) = 18.95, p<0.001) and distant metastasis (chi(2) = 11.3, p<0.001). In these cases, CEA mRNA, but not CEA, was entered in stepwise discriminant analysis to classify the presence of lymph node metastasis. CONCLUSIONS: The molecular detection of micrometastasis in the blood by means of CEA mRNA identification is feasible for colorectal, but not for gastric or pancreatic cancer staging. Further studies are needed in order to define the clinical utility of this marker also in follow-up protocols.