2005 LANDES BIOSCIENCE. DO NOT DISTRIBUTE.

Transcription

1 [Cancer Biology & Therapy 4:5, e11-e18, EPUB Ahead of Print, May 2005]; 2005 Landes Bioscience Research Paper Immortalizing the Complexity of Cancer Metastasis Genetic Features of Lethal Metastatic Pancreatic Cancer Obtained from Rapid Autopsy Erlinda E. Embuscado 1 Daniel Laheru 2 Francesca Ricci 4 Ki Jung Yun 1 Sten de Boom Witzel 1 Allison Seigel 1 Katie Flickinger 1 Manuel Hidalgo 2 G. Steven Bova 1,3 Christine A. Iacobuzio-Donahue 1,2, * Departments of 1 Pathology, 2 Oncology, and 3 Urology; The Johns Hopkins Hospital; Baltimore, Maryland USA 4 University La Sapienza ; Rome, Italy *Correspondence to: Christine A. Iacobuzio-Donahue; Gastrointestinal Cancer Rapid Medical Donation Program; The Johns Hopkins Hospital; Division of Gastrointestinal/Liver Pathology; 720 Rutland Street, Ross Building Rm 632; Baltimore, Maryland USA; Tel.: ; Fax: ; ciacobu@jhmi.edu Received 01/06/05; Accepted 03/12/05 This manuscript has been published online, prior to printing, for Cancer Biology & Therapy Volume 4, Issue 5. Definitive page numbers have not been assigned. The current citation for this manuscript is: Cancer Biol Ther 2005; 4(5): Once the issue is complete and page numbers have been assigned, the citation will change accordingly. KEY WORDS metastasis, pancreatic, autopsy, xenograft, DPC4, TP53 ACKNOWLEDGEMENTS Supported by NIH grant CA (C.I.D.), The Joseph C. Monastra Fund for Pancreatic Cancer Research, The Jeff Zgonina Fund for Pancreatic Cancer Research, the Michael Rolfe Pancreatic Cancer Foundation, The Sol Goldman Pancreatic Cancer Research Center, and a Soongsan Fellowship from Wonkwang University (K.J.Y.). ABSTRACT The virtual lack of well-characterized metastatic pancreatic tissues for study has limited systematic studies of the metastatic process of this deadly disease. To address this important issue, we have instituted a rapid autopsy protocol for the collection of high quality tissues from patients with metastatic pancreatic, called the Gastrointestinal Cancer Rapid Medical Donation Program (GICRMDP). At the time of preparation of this manuscript, 20 patients with metastatic pancreatic and one patient with metastatic colon have undergone a rapid autopsy in association with the GICRMDP. The average time interval achieved for these 21 patients was 8.0 hours, with more than 500 individual samples of matched high quality primary and metastatic pancreatic tissues, peritoneal/pleural fluid and blood obtained so far. For the first four patients in which the autopsy was performed in <6 hours, we have successfully xenografted the primary tumor and/or two to four independent matched metastases from a variety of target organ sites, with a take rate of almost 60% for the first 26 xenografted tumors attempted. In an initial survey of KRAS2, TP53 and DPC4 genetic status in lethal metastatic pancreatic s, activating KRAS2 mutations were detected in 82% of cases and inactivating TP53 mutations in 55% of cases, consistent with rates of genetic alteration of these genes in early stage pancreatic s. However, DPC4 inactivation was found in 75% of patients analyzed, suggesting that genetic inactivation of the DPC4 tumor suppressor gene continues to be selected for with growth at the primary site and metastatic spread to other organs. The invaluable tissue resources generated by the success of the GICRMDP will provide an unparalleled resource for study of metastatic pancreatic and of the metastatic process in general. INTRODUCTION Pancreatic is an almost uniformly lethal disease. In the year 2003 in the United States, an estimated 31,000 patients were diagnosed with pancreatic, and an estimated 31,000 patients died of pancreatic. 1 The dismal prognosis associated with pancreatic is largely due to the fact that most patients are diagnosed at an advanced stage of disease. Eighty percent of patients with pancreatic are not candidates for surgical resection because they have metastatic disease at the time of diagnosis, and 70% of patients who undergo surgical resection are found to have metastases in their resection specimen. 2 It should not be surprising that virtually all patients with pancreatic will die of metastatic disease. 3 Chemotherapy and radiotherapy are of limited benefit to patients with metastatic pancreatic. Gemcitabine is the current standard of care for patients with advanced disease, but has shown clinical responses in a minority of patients. 3-5 Clinical trials to determine the utility of adding a second cytotoxic drug or radiotherapy to this treatment regimen have shown modest improvements in disease-free survival, with reported survival rates reaching 20 40% at one year. 3,6 Clearly, a better understanding of metastatic pancreatic is urgently needed if we are to have an impact on patient survival. Dramatic advances in our understanding of the molecular biology of pancreatic carcinogenesis have occurred within the past decade. Early studies of pancreatic identified chromosome arms 9p, 18q, 17p, 1p and 6q as having high rates of allelic loss in a series of xenografted primary infiltrating pancreatic s. 7 Based in part on these findings, the tumor suppressor genes p16 (CDKN2A) on chromosome 9p, DPC4 (SMAD4, MADH4) on chromosome 18q, MKK4 on chromosome 17p, STK11/LKB1 on chromosome 19p, and BRCA2 on chromosome 13q were identified as inactivated in primary pancreatic s. 8 Most recently, genetic inactivation of members of the Fanconi s Anemia gene family have been identified in primary infiltrating pancreatic s with important 2005 LANDES BIOSCIENCE. DO NOT DISTRIBUTE. e11 Cancer Biology & Therapy 2005; Vol. 4 Issue 5

2 Table 1 Clinicopathologic features of patients Case Age/Sex Primary Pathology Survival Ischemic Cause of Death Therapy Total Number Tumor from Time Metastases b Diagnosis A1 84F Pancreas Adenocarcinoma 6 mo 6 hr Hepatic abscess, NAa ~10 sepsis A2 62M Pancreas Adenocarcinoma 6 mo 5 hr Hepatic failure TP >1000 A3 68F Pancreas with Adenocarcinoma 11 mo 20 hr Bronchopneumonia G, CA, DO, H, CY ~100 mucinous features A4 70M Colon Mucinous 38 mo 4 hr Hepatic and Renal 5-FU, L, CA, O ~10 Adenocarcinoma Failure A5 48F Pancreas Adenocarcinoma 9 mo 2 hr Sepsis G, P, CA, XRT 0 A6 58M Pancreas Adenocarcinoma 10 mo 2 hr Hepatic failure G, TR >1000 A7 70M Pancreas Adenocarcinoma 45 mo 1 hr Complications of 5-FU, XRT, CA ~10 locally advanced disease A8 60F Pancreas Adenocarcinoma 4 mo 1 hr Bronchopneumonia G ~10 A9 60F Pancreas Adenocarcinoma 8 mo 3 hr Bronchopneumonia CA, G ~10 A10 60M Pancreas Adenocarcinoma 2 mo 2 hr Multiorgan failure No treatment >100 A11 63M Pancreas Adenocarcinoma 21 mo 18 hr Complications of G, TX, XRT ~10 locally advanced disease A12 57M Pancreas Adenocarcinoma NA 17 hr Complications of CA 3 locally advanced disease A13 60M Pancreas Adenocarcinoma 8 mo 3 hr Bronchopneumonia No treatment ~100 A14 60M Pancreas Adenocarcinoma 12 mo 24 hr Head Trauma NA ~100 A15 65M Pancreas Adenocarcinoma 0.5 mo 24 hr Metastatic pancreatic No treatment ~50 A16 66M Pancreas with Adenocarcinoma 36 mo 2 hr NA 5-FU, G, DO, CA, I ~10 mucinous features A17 51F Pancreas Adenocarcinoma 5 mo 3 hr Metastatic pancreatic No treatment ~100 A18 71M Pancreas Adenocarcinoma 42 mo 12 hr Metastatic pancreatic 5-FU, G ~50 A19 52M Pancreas Adenocarcinoma 12 mo 4 hr Sepsis 5-FU, XRT 0 A20 55M Pancreas Adenocarcinoma 22 mo 2 hr Metastatic pancreatic G, CI ~25 /Sepsis A21 68M Pancreas Adenocarcinoma 23 mo 5 hr Metastatic pancreatic G, CI, O ~100 a NA, not available; CA, Capecitabine/Xeloda; CI, Cisplatin; CY, Cyclophosphamide/Cytoxan; DO, Docetaxel/Taxotere; 5-FU, 5-flourouracil; G, Gemcitabine; H, Herceptin; I, Iressa/Gefinitib/ZD1839; L, Leukovorin; O, Oxaliplatin; TP, Taxoprexin; TR, Troxactabine/Troxatyl; TX, Taxol/Paclitaxel; XRT, radiation therapy. b Approximate values based on radiographic data and gross impressions at autopsy. implications for chemotherapeutic responses to mitomycin C in a subset of patients with these mutations. 9,10 While it is evident that the above described findings regarding primary pancreatic represent significant advances in our understanding of this disease, these studies have all been performed on surgically resectable primary pancreatic s and thus represent the minority of patients diagnosed with pancreatic. Moreover, the virtual lack of well-characterized metastatic pancreatic tissues for study has limited systematic studies of the metastatic process. To approach these important issues, we have instituted a rapid autopsy protocol for the collection of high quality tissues from patients with metastatic pancreatic and colorectal, called the Gastrointestinal Cancer Rapid Medical Donation Program (GICR- MDP). This program is similar to other rapid autopsy programs for the study of metastatic prostate. 11 Preliminary findings on patient recruitment, clinicopathologic features of metastatic disease, and the rates of xenograft formation are presented, as well as initial data regarding the rates of KRAS2 activation and TP53 and DPC4 genetic inactivation of lethal metastatic pancreatic s. MATERIALS AND METHODS Recruitment of patient volunteers. The Gastrointestinal Cancer Rapid Medical Donation Program (GICRMDP) was established in February 2003 at The Johns Hopkins Hospital. This program was approved by The Johns Hopkins Institutional Review Board and deemed in accordance with the Health Insurance Portability and Accountability Act (HIPAA). Any patient with biopsy-proven or radiographic evidence of metastatic pancreatic Cancer Biology & Therapy e12

3 A B Figure 1. Establishment of xenografted tumors from metastatic pancreatic tissues obtained at rapid autopsy. (A) Poorly differentiated pancreatic carcinoma metastatic to the lung (1.0 cm diameter) obtained from rapid autopsy of patient A2. (B) Xenografted tumor tissue A2.4 created from the identical metastatic pancreatic carcinoma shown in (A). is eligible to participate. Patients are recruited directly from the medical and radiation oncology clinics at The Johns Hopkins Hospital (D.L., M.H.) or are self-referred through the Johns Hopkins Pancreas Cancer Web Page ( For those patients identified through the oncology clinics, the decision to discuss participation in the study is based on a strong physician-patient relationship and on the clinician s assessment of the patient s emotional status. Once a patient is referred to the GICRMDP, appropriate verbal and written information about the study are provided, and informed consent is obtained to perform a rapid autopsy and to obtain copies of all past medical records and previously removed surgical tissue. When possible, discussion of the study and informed consent takes place with the patient s spouse and/or next of kin present. Final permission for autopsy remains in the hands of the patient s next of kin at the time of death. The patient s next of kin immediately notify the GICRMDP team when the patient died. All members of the team are on call 24 hours a day/ 7 days a week at the time of an impending autopsy. Performance of the rapid autopsy. Performance of the rapid autopsy procedure at The Johns Hopkins Hospital follows an organized team approach. The team includes a board-certified pathologist (C.I.D. or G.S.B.) and physician assistant (K.F.) for harvesting of all normal and tissues and four additional team members (E.E.E., S.dB.W., F.R., K.J.Y.) responsible for tissue processing, labeling and xenografting of selected neoplastic tissues. The tissue harvesting protocol consists of the following; after opening of the body cavity using standard techniques, approximately 50% of the primary neoplasm and each grossly identified metastasis are sampled using a sterile blade and forceps (one set per tumor) with care to avoid areas of grossly evident necrosis or hemorrhage. Tumor tissues are sliced into 1.0 x 1.0 x 0.2 cm sections for overnight fixation in 10% buffered-formalin or 75% 3:1 ethanol:methanol for paraffin-embedding, for freezing in OCT media in liquid nitrogen and storage at -80 C, or for snap-freezing in liquid nitrogen in 1.5 ml cryovials and storage at -80 C. For those patients in which peritoneal ascites and/or pleural effusions are present, 50 mls of each are collected under sterile conditions and stored at -80 C. In all cases, samples of normal liver, lung and spleen are also harvested for processing as described above, as well as 10 mls of whole blood from the right atrium. In the case of a self-referred patient who died in a geographic location not amenable to transport to The Johns Hopkins Hospital, the tissue harvesting and autopsy procedure are performed at a local hospital by a certified pathologist that is arranged in advance in collaboration with the GICRMDP. In these cases a limited harvesting is performed in which each sample of the normal, primary and/or metastatic tumor tissues is bisected for both snap-freezing on dry ice and fixation in 10%-buffered formalin. In all cases, tissue harvesting is followed by a standard complete diagnostic autopsy minus any restrictions specified in the consent form. The unique identifier of each sample assigned at autopsy is entered into an Access relational database including all clinicopathologic information related to the patients medical and surgical history, medications, imaging studies, pathology and autopsy findings. Creation of pancreatic xenografts. For those patients in which the rapid autopsy is performed at The Johns Hopkins Hospital and the ischemic interval is <6 hours, two to three mm 3 samples each of the primary infiltrating carcinoma and 4 8 different metastases are placed into sterile RPMI media (Invitrogen, Carlsbad, CA) for transport to the animal facility for xenografting into CD1 nu/nu mice. In all cases, care is taken to obtain grossly viable tissue. Xenografting is performed as described by Hahn et al. 7 Tumor tissues are minced with a sterile blade into 1 mm 3 pieces in Matrigel (Becton Dickinson, Bedford, MA), and one to two mm 3 of tissue are implanted beneath the skin fold of the shoulder bilaterally while the mice are under isoflurane inhalation anesthesia. The implantation site is closed with a single 6.0 chromic suture. The animals are followed weekly for growth of xenografts, and the xenografts are harvested when they reach cm in size. The harvested tumors were stored at -80 C, and a small piece of the xenografted tumor (1 3 mm 2 ) is implanted into a new mouse to perpetuate the xenograft. Tissue microdissections. For pancreatic s in which xenograftenriched tumors were not available, microdissection of the primary or metastatic carcinoma tissues was performed to enrich for neoplastic cells as previously described. 12,13 Only samples in which at least 50% neoplastic cellularity could be obtained were used for genetic studies. PCR and sequencing of KRAS2, p53 and DPC4. Genomic DNA was extracted from xenograft enriched or microdissected pancreatic s using DNeasy kits (Qiagen, Valencia CA). PCR amplification of exon 1 of KRAS, exons 5 9 of TP53 and exons 1-11 of DPC4 from genomic DNA was performed as described. 12,14,15 PCR-amplified products were purified using QIAquick columns (Qiagen) and studied by automated sequencing using nested primers and an ABI Prism model 3700 (Applied Biosystems, Foster City, CA). Sequence analysis employed Sequencher version software (Gene Codes, Ann Arbor, MI). Verification of the mutation was accomplished by sequencing of a second PCR product derived independently from the original template. Immunohistochemistry for Dpc4 protein. Formalin-fixed tissues from each patient who participated in the GICRMDP were paraffin-embedded following standard techniques. Archival samples of metastatic pancreatic from 19 patients were also obtained from the Autopsy Pathology Files of The Johns Hopkins Hospital as previously described. 16 For all cases, unstained 5-Μm sections were cut from the paraffin blocks and deparaffinized by routine techniques followed by steaming in sodium citrate buffer ph 6.0 for 20 minutes at 80 C. Slides were cooled 15 minutes and incubated e13 Cancer Biology & Therapy 2005; Vol. 4 Issue 5

4 with a 1:100 dilution of monoclonal antibody to Madh4 protein (clone B8, Santa Cruz Biotechnology) using the DAKO automated stainer (DAKO, Carpinteria, CA). Primary antibody was detected by secondary antibody, followed by avidin-biotin complex and 3,3'-diaminobenzidine chromagen. Sections were counterstained with hematoxylin and evaluated by two of the authors (E.E.E. and C.I.D). Statistics. Summary data are expressed as the mean ± S.D. unless otherwise indicated. The frequencies of Dpc4 immunolabeling among localized primary s versus advanced unresectable s were analyzed by the Chi-squared test. Probability values of 0.05 were considered significant. RESULTS Table 2 Clinicopathologic characteristics of patients. A summary of the first 21 patients to undergo an autopsy as part of the GICRMDP is shown in Table 1. The mean age is 65.4 ± 8.1 years, sixteen patients were male and five were female. All 21 patients were Caucasian. Seventeen patients had clinically evident metastatic disease at the time of diagnosis (clinical stage IV), and four patients underwent pancreaticoduodenectomy at the time of diagnosis (clinical stage III, patients A16, A19, A20, A21) but later developed local recurrence and metastatic disease. The mean time from diagnosis to death was 16.0 months (range months). The mean time interval from death to autopsy for all cases was 8.0 hours (median 3 hours, range 1 24 hours). Delays in performance of the autopsy (>6 hours) were in all cases due to unavoidable limitations in transportation to the facility where the procedure was performed. In all cases, the gross and histologic examination of the autopsy tissues confirmed the diagnosis of primary infiltrating pancreatic or colonic adenocarcinoma. Among the 20 patients with pancreatic adenocarcinoma, 16 primary s were localized to the head of the pancreas, 3 to the tail, and 1 diffusely involved the entire gland. Eighteen carcinomas were poorly differentiated, 1 carcinoma was poorly differentiated with foci of mucinous differentiation, and 1 carcinoma was moderate to well differentiated with mucinous features. Patterns of advanced disease in patients with terminal pancreatic. The patterns of metastatic spread of pancreatic have been reported in several large retrospective studies. 17,18 Our findings of the metastatic pan- months. creatic disease burden in these 20 patients are consistent with those reports. Gross evidence of metastatic disease was present in 17 of 20 patients, with a mean of 2.9 ± 2.8 target organs involved. As expected, the organ most commonly involved by metastatic disease was the liver (16 of 17 patients, 94%), followed by the lungs (8 of 17 patients, 47%), abdominal lymph nodes (7 of 17 patients, 41%) and peritoneum/omentum (7 of 17 patients, 41%). The thoracic lymph nodes (5 of 17 patients, 29%) and adrenal glands (4 of 17 patients, 21%) were involved in a subset of patients, whereas the cardiovascular, reproductive and genitourinary systems were the least frequently involved sites in this set of patients. In one patient (A8), the lungs and the para-aortic lymph node chain were the predominant site of metastasis, with no gross evidence of disease noted in the liver. Creation of xenografted tumors from autopsy tissues. In addition to our goal of obtaining well-characterized high quality tissue from patients with metastatic pancreatic for research purposes, we have successfully created xenografted tumor tissues for use in in-depth studies of the molecular genetics associated with pancreatic metastasis (Table 2 and Fig. 1). For five patients in which the rapid autopsy was initiated <6 hours (A2, A5, A6, A10 and A13), samples of the primary and/or metastatic tumor tissues Growth characteristics of Xenografted tumors created from rapid autopsy material Case Pathology Location Status of Xenograft a Time to Harvest A2.1 Metastasis Liver Grown 3 mo A2.2 Metastasis Liver Grown 3 mo A2.3 Metastasis Liver Grown 2.5 mo A2.4 Metastasis Lung Grown 4 mo A2.5 Primary Pancreas Mouse died A2.6 Primary Pancreas Grown 7.5 mo A5.13 Primary Pancreas No growth A5.14 Primary Pancreas No growth A6.1 Metastasis Liver No growth A6.3 Metastasis Omentum Mouse died A6.5 Metastasis Liver Grown 4 mo A6.7 Primary Pancreas Grown 4 mo A6.9 Metastasis Lymph node No growth A6.12 Metastasis Liver Grown 4 mo A10.1 Metastasis Liver Grown 3.5 mo A10.2 Metastasis Liver No growth A10.3 Metastasis Liver No growth A10.4 Metastasis Liver Grown 5 mo A10.5 Metastasis Liver Grown 5 mo A10.6 Metastasis Liver No growth A10.7 Metastasis Liver Grown 3.5 mo A10.8 Primary Pancreas No growth A13.A Primary Pancreas Grown 5 mo A13.B Primary Pancreas Grown 5 mo A13.C Metastasis Lymph node No growth A13.D Metastasis Lung Grown 5 mo A13.E Metastasis Liver No growth A13.F Metastasis Liver No growth A13.G Metastasis Lymph node No growth A13.H Metastasis Lymph node Grown 5 mo a Designations are Grown, xenografted tumor reached at least 1.0 cm size limit and was harvested; No growth, no growth at 12 were xenografted into nude mice. A total of 30 individual tumor tissues were xenografted from these patients representing the five primary carcinomas and 24 different metastases. Two mice died shortly after the xenografting procedure, leaving 28 mice for study. Using one mouse per tumor and two insertion sites per mouse, 16 of these 28 tumors (57%) were successfully grown as xenografts in nude mice, representing three different primary tumors and twelve different metastases. The mean time from implantation to harvesting for these 16 xenografted tumors was 4.4 ± 1.2 months. Mutations of KRAS2, TP53 and DPC4 in metastatic pancreatic s. The ultimate goal of the GICRMDP is to determine the molecular or genetic events associated with lethal metastasis. Towards this end, we performed a survey of KRAS2, TP53 and DPC4 genetic alterations among a subset of our cases for which high quality genomic DNA with at least 50% neoplastic cellularity could be achieved (Table 3). Twenty-six samples were analyzed from 11 patients representing six primary carcinomas and twenty different pancreatic metastases. For seven of these 11 patients, two to four different metastases were analyzed. Activating mutations of KRAS2 and inactivation of TP53 are among the most common genetic events in pancreatic carcinogenesis. 14,19,20 Activating Cancer Biology & Therapy e14

5 Table 3 Genetic features of lethal metastatic pancreatic TP53 Sample Pathology KRAS2 Exon Effect Predicted Mutation Type Genetics IHC Effect Predicted A1.L2 Liver Metastasis G 1 2D 5 R175H missense no mutation negative phdb A1.L3 Liver Metastasis nda 5 R175H missense no mutation negative phd A2_1 Liver Metastasis G 1 2D 5 T155P missense HD negative HD A2_2 Liver Metastasis G 1 2D 5 T155P missense HD negative HD A2_3 Liver Metastasis G 1 2D 5 T155P missense HD negative HD A2_4 Lung Metastasis G 1 2D 5 T155P missense HD negative HD A2_6 Primary Carcinoma G 1 2D 5 T155P missense HD negative HD A3.5 Liver Metastasis nd 5 C176F missense no mutation negative phd A3.8 Liver Metastasis G 1 2V 6 C176F missense no mutation negative phd A6_E Liver Metastasis WT 7 L257P missense HD negative HD A6_G Primary Carcinoma WT 7 L257P missense HD negative HD A6_L Liver Metastasis WT 7 L257P missense HD negative HD A10_1 Liver Metastasis G12D WT HD negative HD A10_4 Liver Metastasis G12D WT HD negative HD A10_5 Liver Metastasis G 1 2D WT HD negative HD A10_7 Liver Metastasis G 1 2D WT HD negative HD A11.14 Liver Metastasis G 1 2D WT no mutation negative phd A12.46 Liver Metastasis G 1 2D 7 R249I tandem no mutation negative phd A12.72 Lymph Node Metastasis G 1 2D 7 R249I tandem no mutation negative phd A13_A Primary Carcinoma G 1 2V WT WT positive WT A13_B Primary Carcinoma G 1 2V WT WT positive WT A13_D Lung Metastasis G 1 2V WT WT positive WT A13_H Lymph Node Metastasis G 1 2V WT WT positive WT A14.41 Primary Carcinoma WT WT no mutation negative phd A15.63 Liver Metastasis WT WT no mutation negative phd A17.34 Primary Carcinoma G 1 2D 7 R249I tandem no mutation negative phd a nd, not determined. b phd, presumed homozygous deletion. DPC4 mutations of KRAS2 were found in nine of 11 (82%) patients tumors, in all cases at codon 12, while inactivating mutations of TP53 were found in six of 11 (55%) patients s within the evolutionarily conserved domain. 14 In all seven cases in which more than one sample from a patient was analyzed, the KRAS2 or TP53 mutations were concordant in both the primary carcinoma and/or all matched metastases (Table 3). All mutations originally found in xenografted tumors (patients A2, A6, A10 and A13) were confirmed in genomic DNA obtained from microdissection of the original tissues from which the xenografts were derived. In two cases (A14 and A15), no mutations were found of KRAS2 or TP53 despite PCR amplification of genomic DNA obtained from two or more independent microdissections of high quality tissue. DPC4 is a tumor suppressor gene inactivated in ~55% of surgically resected pancreatic s analyzed. 15 As the predominant mechanism of DPC4 inactivation is by homozygous deletion, 21 screening for genetic deletions by amplification failure is not possible using microdissected tissues as normal cellular DNA within the sample may amplify and obscure findings. To account for this possibility, we performed whole gene sequencing of DPC4 together with immunohistochemical labeling for Dpc4 protein in the same tissues. Immunolabeling for Dpc4 protein in paraffin-embedded tissues has been described as a reliable marker of the genetic status of the DPC4 gene. 22 Among the four xenograft enriched carcinomas homozygous deletion of DPC4 was found in three of four cases. These three s were also negative for Dpc4 labeling (cases A2, A6, and A10), while the one case without a deletion or intragenic mutation showed strong positive labeling for Dpc4 protein (case A13) (Fig. 2). Among the seven patients for which only microdissected material was available, no intragenic mutations were found within the entire coding region. However, all seven carcinomas showed complete loss of immunohistochemical labeling for Dpc4 protein consistent with a deletion of DPC4. When combined with the four xenograft enriched carcinomas, Dpc4 inactivation was found in 10 of 11 patient s primary and/or metastatic carcinomas (91%). To rule out that the higher rate of DPC4 inactivation was not related to the number of patients analyzed, we determined the Dpc4 immunolabeling patterns among an additional six patients from the GICRMDP for which high quality tissues were available, as well as the archival material from an additional 19 patients with widely metastatic pancreatic from the Autopsy Files of The Johns Hopkins Hospital. When combined with the original 11 patients, loss of Dpc4 immunolabeling was noted among 13 of 17 (76%) patients metastatic pancreatic tissues obtained from the GICRMDP, and among 14 of 19 (71%) patients metastatic disease obtained from the archival materials. A comparison of the frequency of Dpc4 loss in surgically resectable pancreatic carcinomas reported by Wilentz et al. 23 (22 of 41 carcinomas negative for Dpc4 immunolabeling, 54%) as compared to the rates we have observed for advanced unresectable disease (27 of 36 patients metastatic carcinomas negative for Dpc4 immunolabeling, 75%) indicated the increased rate of Dpc4 inactivation found in advanced disease was statistically significant (p<0.05). e15 Cancer Biology & Therapy 2005; Vol. 4 Issue 5

6 A B C D Figure 2. Dpc4 immunolabeling in metastatic pancreatic. (A and B) Positive immunolabeling for Dpc4 is seen in both the primary carcinoma (A) and the matched liver metastasis (B) from patient A13. (C and D) In contrast, negative immunolabeling for Dpc4 is observed in the primary carcinoma (C) and the matched liver metastasis from patient A6 (D). DISCUSSION In order to evaluate the importance of cellular changes already described in pancreatic, and to provide a solid foundation for future studies of metastatic pancreatic, it is absolutely critical to study pancreatic s that have already caused significant morbidity and mortality. The primary source of such material is through the autopsy of men and women who have died from metastatic pancreatic. The high-quality tissues obtained through the successful establishment of the GICRMDP, in addition to the ability to create xenografted tumors, will permit a similar quality of analyses of metastatic pancreatic as shown for surgically resected disease, including the genetic alterations associated with pancreatic metastasis or the molecular biologic correlates of adjuvant therapy failure, among others. Although this study represents a preliminary survey of metastatic disease spread, we have noted differences in disease burden that related in part to the cause of mortality, similar to that described by Nakahashi et al. 24 Specifically, four patients (A2, A6, A17 and A21) demonstrated extensive metastatic involvement of the liver so that death resulted directly from hepatic failure. In contrast, for eight patients the cause of death related to marked extension of the primary tumor into adjacent vital structures (ex. A1, A3, A5, A7, A8, A11, A12, A15), despite a wide range of metastatic spread to other sites. Finally, for six patients, no obvious relationship between disease burden and cause of death was found. In these cases, morbidity and mortality appeared related to generalized multiorgan failure and cachexia (ex. A9, A10, A13, A18, A19, A20). For two patients, the ultimate cause of death was unknown or unrelated to their disease (A14 and A16). The availability of the complete treatment history for each of these patient included in the study provides an invaluable opportunity to correlate these observations with chemotherapeutic history. Additional studies of the metastatic process with respect to these patterns of growth will determine the significance, if any, of these observations. The utility of xenografted tumors for the study of primary pancreatic s has been unparalleled. 7,9,21,25-27 Our ability to create xenograft enriched metastatic tumors will allow for similar study of the genetic events associated with advanced disease. The xenografted tumors we have established derived from metastases grew more rapidly than those derived from the matched primary carcinomas of the same patient (mean 4.0 months versus 5.4 months) suggesting the selection of neoplastic cell clones more autonomous in growth characteristics than those of the primary tumor counterparts. However, as xenografted tumors have been successfully grown from a large cohort of localized and surgically resectable pancreatic, 7 the enhanced growth rate of these metastases may in part be reflective of their growth characteristics in vivo. At the very least, further studies to explore these differences are warranted. Cancer Biology & Therapy e16

7 While numerous investigations have focused on the genetic alterations of primary infiltrating pancreatic s, few if any have addressed those events characteristic of advanced disease. Our findings of KRAS2 activating mutations in 82% and TP53 genetic inactivation in 55% of advanced carcinomas are consistent with rates reported previously, 14,19,20 and lend support to the genetic alteration of these two genes early in pancreatic carcinogenesis. In contrast, our data indicate that genetic inactivation of DPC4 is more prevalent in advanced stage metastatic disease than that reported for surgically resectable primary infiltrating pancreatic s, inactivated in 75% of the patients metastatic s analyzed (p < 0.05). The increased rate of DPC4 inactivation in advanced disease lends further support to the prominent role of this tumor suppressor gene in the biology and aggressiveness of pancreatic. Moreover, the data suggests that DPC4 inactivation continues to be selected for with tumor progression, a finding with important implications for anti- therapies aimed at restoration of the TGF-β signal transduction pathway in this tumor type. 28 Of interest, in all cases we examined by immunohistochemical labeling, both the primary carcinoma and all matched metastases showed identical Dpc4 immunolabeling patterns (data not shown). While this finding may be interpreted as support for the hypothesis of an outgrowth of dominant metastatic clones within a primary tumor with successful spread to distant sites, 29 based on this data alone we cannot rule out the competing hypothesis that metastatic ability is inherent within a primary tumor. 30,31 At the very least, our finding of a significant difference in the rate of Dpc4 loss of immunolabeling between primary and metastatic pancreatic supports the use of this resource for the identification of cellular changes of biologic significance in advanced disease. In summary, we provide initial data regarding the collection of high quality tissues from rapid autopsy participants of the GICR- MDP. The invaluable tissue resources generated by this program will provide an unparalleled resource for study of metastatic pancreatic and of the metastatic process in general. References 1. Jemal A, Tiwari RC, Murray T, Ghafoor A, Samuels A, Ward E, Feuer EJ, Thun M. J. Cancer statistics, CA Cancer J Clin 2004; 54: Sohn TA, Yeo CJ, Cameron JL, Koniaris L, Kaushal S, Abrams RA, Sauter PK, Coleman J, Hruban RH, Lillemoe KD. Resected adenocarcinoma of the pancreas-616 patients: Results, outcomes, and prognostic indicators. J Gastrointest Surg 2000; 4: Rocha Lima CM, Green MR, Rotche R, Miller Jr WH, Jeffrey GM, Cisar LA, Morganti A, Orlando N, Gruia G, Miller LL. Irinotecan plus gemcitabine results in no survival advantage compared with gemcitabine monotherapy in patients with locally advanced or metastatic pancreatic despite increased tumor response rate. J Clin Oncol 2004; 22: Berlin JD, Catalano P, Thomas JP, Kugler JW, Haller DG, Benson AB. III Phase III study of gemcitabine in combination with fluorouracil versus gemcitabine alone in patients with advanced pancreatic carcinoma: Eastern Cooperative Oncology Group Trial E2297. J Clin Oncol 2002; 20: Rocha Lima CM, Centeno B. Update on pancreatic. Curr Opin Oncol 2002; 14: Chakravarthy A, Abrams RA, Yeo CJ, Korman LT, Donehower RC, Hruban RH, Zahurek ML, Grochow LB, O Reilly S, Hurwitz H, Jaffee EM, Lillemoe KD, Cameron JL. Intensified adjuvant combined modality therapy for resected periampullary adenocarcinoma: Acceptable toxicity and suggestion of improved 1-year disease-free survival. Int J Radiat Oncol Biol Phys 2000; 48: Hahn SA, Seymour AB, Hoque AT, Schutte M, da Costa LT, Redston MS, Caldas C, Weinstein CL, Fischer A, Yeo CJ. Allelotype of pancreatic adenocarcinoma using xenograft enrichment. Cancer Res 1995; 55: Kern SE, Hruban RH, Hidalgo M, Yeo CJ. An introduction to pancreatic adenocarcinoma genetics, pathology and therapy. Cancer Biol Ther 2002; 1: van der Heijden MS, Yeo CJ, Hruban RH, Kern SE. Fanconi anemia gene mutations in young-onset pancreatic. Cancer Res 2003; 63: van der Heijden MS, Brody JR, Gallmeier E, Cunningham SC, Dezentje DA, Shen D, Hruban RH, Kern SE. Functional defects in the fanconi anemia pathway in pancreatic cells. Am J Pathol 2004; 165: Rubin MA, Putzi M, Mucci N, Smith DC, Wojno K, Korenchuk S, Pienta KJ. Rapid ( warm ) autopsy study for procurement of metastatic prostate. Clin Cancer Res 2000; 6: Caldas C, Hahn SA, Hruban RH, Redston MS, Yeo CJ, Kern SE. Detection of K-ras mutations in the stool of patients with pancreatic adenocarcinoma and pancreatic ductal hyperplasia. Cancer Res 1994; 54: Seymour AB, Hruban RH, Redston M, Caldas C, Powell SM, Kinzler KW, Yeo CJ, Kern SE. Allelotype of pancreatic adenocarcinoma. Cancer Res 1994; 54: Redston MS, Caldas C, Seymour AB, Hruban RH, da Costa L, Yeo CJ, Kern SE. p53 mutations in pancreatic carcinoma and evidence of common involvement of homocopolymer tracts in DNA microdeletions. Cancer Res 1994; 54: Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa LT, Rozenblum E, Weinstein CL, Fischer A, Yeo CJ, Hruban RH, Kern SE. DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science 1996; 271: Xin W, Yun KJ, Ricci F, Zahurak M, Qiu W, Su GH, Yeo CJ, Hruban RH, Kern SE, Iacobuzio-Donahue CA. MAP2K4/MKK4 expression in pancreatic : Genetic validation of immunohistochemistry and relationship to disease course. Clin Cancer Res 2004; 10: Kamisawa T, Isawa T, Koike M, Tsuruta K, Okamoto A. Hematogenous metastases of pancreatic ductal carcinoma. Pancreas 1995; 11: Mao C, Domenico DR, Kim K, Hanson DJ, Howard JM. Observations on the developmental patterns and the consequences of pancreatic exocrine adenocarcinoma. Findings of 154 autopsies. Arch Surg 1995; 130: Hruban RH, van Mansfeld AD, Offerhaus GJ, van Weering DH, Allison DC, Goodman SN, Kensler TW, Bose KK, Cameron JL, Bos JL. K-ras oncogene activation in adenocarcinoma of the human pancreas. A study of 82 carcinomas using a combination of mutant-enriched polymerase chain reaction analysis and allele-specific oligonucleotide hybridization. Am J Pathol 1993; 143: Smit VT, Boot AJ, Smits AM, Fleuren GJ, Cornelisse CJ, Bos JL. KRAS codon 12 mutations occur very frequently in pancreatic adenocarcinomas. Nucleic Acids Res 1988; 16: Iacobuzio-Donahue CA, Song J, Parmiagiani G, Yeo CJ, Hruban RH, Kern SE. Missense mutations of MADH4: Characterization of the mutational hot spot and functional consequences in human tumors. Clin Cancer Res 2004; 10: Wilentz RE, Su GH, Dai JL, Sparks AB, Argani P, Sohn TA, Yeo CJ, Kern SE, Hruban RH. Immunohistochemical labeling for dpc4 mirrors genetic status in pancreatic adenocarcinomas: A new marker of DPC4 inactivation. Am J Pathol 2000; 156: Wilentz RE, Iacobuzio-Donahue CA, Argani P, McCarthy DM, Parsons JL, Yeo CJ, Kern SE, Hruban RH. Loss of expression of Dpc4 in pancreatic intraepithelial neoplasia: Evidence that DPC4 inactivation occurs late in neoplastic progression. Cancer Res 2000; 60: Nakahashi C, Oda T, Kinoshita T, Ueda T, Konishi M, Nakagohri T, Inoue K, Furuse J, Ochiai A, Ohkohchi N. The impact of liver metastasis on mortality in patients initially diagnosed with locally advanced or resectable pancreatic. Int J Gastrointest Cancer 2003; 33: Schutte M, Hruban RH, Geradts J, Maynard R, Hilgers W, Rabindran SK, Moskaluk CA, Hahn SA, Schwarte-Waldhoff I, Schmiegel W, Baylin SB, Kern SE, Herman JG. Abrogation of the Rb/p16 tumor-suppressive pathway in virtually all pancreatic carcinomas. Cancer Res 1997; 57: Su GH, Bansal R, Murphy KM, Montgomery E, Yeo CJ, Hruban RH, Kern SE. ACVR1B (ALK4, activin receptor type 1B) gene mutations in pancreatic carcinoma. Proc Natl Acad Sci USA 2001; 98: Su GH, Hilgers W, Shekher MC, Tang DJ, Yeo CJ, Hruban RH, Kern SE. Alterations in pancreatic, biliary, and breast carcinomas support MKK4 as a genetically targeted tumor suppressor gene. Cancer Res 1998; 58: Sohn TA, Su GH, Ryu B, Yeo CJ, Kern SE. High-throughput drug screening of the DPC4 tumor-suppressor pathway in human pancreatic cells. Ann Surg 2001; 233: Poste G, Fidler IJ. The pathogenesis of metastasis. Nature 1980; 283: Weigelt B, Glas AM, Wessels LF, Witteveen AT, Peterse JL, van t Veer LJ. Gene expression profiles of primary breast tumors maintained in distant metastases. Proc Natl Acad Sci USA 2003; 100: Ramaswamy S, Ross KN, Lander ES, Golub TR. A molecular signature of metastasis in primary solid tumors. Nat Genet 2003; 33: e17 Cancer Biology & Therapy 2005; Vol. 4 Issue 5