Randomized, double-blind, placebo-controlled phase II study of istiratumab (MM-141) plus nab-paclitaxel and gemcitabine versus nab-paclitaxel and gemcitabine in front-line metastatic pancreatic cancer (CARRIE).

BACKGROUND: Preclinical data suggest that dual blockade of the insulin-like growth factor-1 receptor (IGF-1R) and HER3 pathways has superior activity to IGF-1R blockade alone in pancreatic ductal adenocarcinoma (PDAC). We tested whether istiratumab, an IGF-1R- and ErbB3-bispecific antibody, can enhance the efficacy of standard of care (SOC) chemotherapy in patients with metastatic PDAC selected for high IGF-1 serum levels. PATIENTS AND METHODS: CARRIE was an international, randomized, double-blind, placebo-controlled phase II study for patients with previously untreated metastatic PDAC. In part 1, 10 patients were evaluated for pharmacokinetics and safety. In part 2, patients with high free serum IGF-1 levels were randomized 1 : 1 to receive either istiratumab [2.8 g intravenously (i.v.) every 2 weeks] or placebo combined with gemcitabine/nab-paclitaxel at approved dose schedule. The co-primary endpoints were progression-free survival (PFS) in patients with high IGF-1 levels and PFS in patients with both high serum IGF-1 levels and heregulin (HRG)+ tumors. Key secondary endpoints were overall survival (OS), objective response rate (ORR) by RECIST v.1.1, and adverse events (AEs) rate. RESULTS: A total of 317 patients were screened, with 88 patients randomized in part 2 (experimental arm n = 43; control n = 45). In the high IGF-1 cohort, median PFS was 3.6 and 7.3 months in the experimental versus control arms, respectively [hazard ratio (HR) = 1.88, P = 0.027]. In the high IGF-1/HRG+ subgroup (n = 44), median PFS was 4.1 and 7.3 months, respectively (HR = 1.39, P = 0.42). Median OS and ORR for the overall population were similar between two arms. No significant difference in serious or grade ≥3 AEs was observed, although low-grade AEs leading to early discontinuation were higher in the experimental (39.5%) versus control arm (24.4%). CONCLUSIONS: Istiratumab failed to improve the efficacy of SOC chemotherapy in this patient setting. High serum IGF-1 levels did not appear to be an adverse prognostic factor when compared with non-biomarker-selected historic controls. CLINICAL TRIAL REGISTRATION NUMBERS: ClinicalTrials.gov: NCT02399137; EUDRA CT: 2014-004572-34.

Neoadjuvant cetuximab, twice-weekly gemcitabine, and intensity-modulated radiotherapy (IMRT) in patients with pancreatic adenocarcinoma.

BACKGROUND: Neoadjuvant therapy has been investigated for localized and locally advanced pancreatic ductal adenocarcinoma (PDAC) but no standard of care exists. Combination cetuximab/gemcitabine/radiotherapy demonstrates encouraging preclinical activity in PDAC. We investigated cetuximab with twice-weekly gemcitabine and intensity-modulated radiotherapy (IMRT) as neoadjuvant therapy in patients with localized or locally advanced PDAC. EXPERIMENTAL DESIGN: Treatment consisted of cetuximab load at 400 mg/m(2) followed by cetuximab 250 mg/m(2) weekly and gemcitabine 50 mg/m(2) twice-weekly given concurrently with IMRT to 54 Gy. Following therapy, patients were considered for resection. RESULTS: Thirty-seven patients were enrolled with 33 assessable for response. Ten patients (30%) manifested partial response and 20 (61%) manifested stable disease by RECIST. Twenty-five patients (76%) underwent resection, including 18/23 previously borderline and 3/6 previously unresectable tumors. Twenty-three (92%) of these had negative surgical margins. Pathology revealed that 24% of resected tumors had grade III/IV tumor kill, including two pathological complete responses (8%). Median survival was 24.3 months in resected patients. Outcome did not vary by epidermal growth factor receptor status. CONCLUSIONS: Neoadjuvant therapy with cetuximab/gemcitabine/IMRT is tolerable and active in PDAC. Margin-negative resection rates are high and some locally advanced tumors can be downstaged to allow for complete resection with encouraging survival. Pathological complete responses can occur. This combination warrants further investigation.

Isolation of mutants of an animal virus in bacteria.

Mutants of animal viruses can be isolated in bacteria by recombinant DNA methods. Since no viral functions are required for propagation of recombinants in bacteria, viral mutants with lethal changes in cis- or trans-acting elements can be isolated, as well as partially or conditionally defective mutants. In the cases of viruses with small DNA genomes, such as the tumorigenic simian virus 40 (SV40), the entire viral DNA can be inserted into the bacterial plasmid pBR322 and cloned in Escherichia coli. Recombinant plasmids with a single copy of SV40 DNA cause morphological transformation of mouse cells in culture with the same efficiency as SV40 DNA isolated from virus-infected monkey cells, but the recombinant DNA is noninfectious and replicates poorly in permissive cells. However, SV40 DNA excised from the plasmid replicates as well as authentic viral DNA and is fully infectious. SV40 mutants with small deletions or base substitutions have been isolated by in vitro site-specific or random local mutagenesis of recombinant DNA followed by cloning in E. coli. Many of the mutants thus isolated are defective in specific viral functions.

Repression of intestinal drug metabolizing enzymes by the SV40 large T antigen.

Toxic compounds such as carcinogens are removed from the body by the action of a series of detoxifying enzymes and transporters expressed in the liver and the small intestine. We have found that intestinal epithelial cells expressing the SV40 large T antigen (TAg) contain significantly lower levels of mRNAs, encoding several drug metabolizing/detoxifying enzymes and transporters compared to their non-transgenic littermates. In addition, TAg blocks the induction of these mRNAs by xenobiotics. The repression depends on an intact LXCXE motif in TAg, suggesting that inactivation of the retinoblastoma (Rb) family of tumor suppressors plays a role in the process. These results imply that a functional Rb pathway in the intestine is necessary for the expression of the detoxification system used to clear carcinogens, and suggest that loss of this tumor suppressor might alter susceptibility to chemical injury. In addition, the effect of TAg on the detoxification pathway appears to be tissue-specific, as its ectopic expression in the liver failed to suppress the P450 enzymes. The TAg-mediated suppression of drug metabolizing/detoxifying enzymes may have broad implications in the metabolism and mechanism of action of both carcinogens and prescription drugs.

pp60c-src activation in human colon carcinoma.

We measured the in vitro protein-tyrosine kinase activity of pp60c-src from human colon carcinoma cell lines and tumors. The activity of pp60c-src from six of nine carcinoma cell lines was higher (on average, fivefold as measured by enolase phosphorylation, or eightfold as measured by autophosphorylation) than that of pp60c-src from normal colonic mucosal cells, or human or rodent fibroblasts. Similarly, the activity of pp60c-src from 13 of 21 primary colon carcinomas was five- or sevenfold higher than that of pp60c-src from normal colonic mucosa adjacent to the tumor. The increased pp60c-src activity did not result solely from an increase in the level of pp60c-src protein, suggesting the specific activity of the pp60c-src kinase is elevated in the tumor cells. pp60c-src from colon carcinoma cells and normal colonic mucosal cells was phosphorylated at similar sites. We used immunoblotting with antibodies to phosphotyrosine to identify substrates of protein-tyrosine kinases in colonic cells. Three phosphotyrosine-containing proteins were detected at significantly higher levels in most colon carcinoma cell lines than in normal colonic mucosal cells or human or rat fibroblasts. All colon carcinoma cell lines with elevated pp60c-src in vitro kinase activity, showed increased phosphorylation of proteins on tyrosine in vivo, suggesting the presence of an activated protein-tyrosine kinase(s).

The molecular chaperone activity of simian virus 40 large T antigen is required to disrupt Rb-E2F family complexes by an ATP-dependent mechanism.

The simian virus 40 large T antigen (T antigen) inactivates tumor suppressor proteins and therefore has been used in numerous studies to probe the mechanisms that control cellular growth and to generate immortalized cell lines. Binding of T antigen to the Rb family of growth-regulatory proteins is necessary but not sufficient to cause transformation. The molecular mechanism underlying T-antigen inactivation of Rb function is poorly understood. In this study we show that T antigen associates with pRb and p130-E2F complexes in a stable manner. T antigen dissociates from a p130-E2F-4-DP-1 complex, coincident with the release of p130 from E2F-4-DP-1. The dissociation of this complex requires Hsc70, ATP, and a functional T-antigen J domain. We also report that the "released" E2F-DP-1 complex is competent to bind DNA containing an E2F consensus binding site. We propose that T antigen disrupts Rb-E2F family complexes through the action of its J domain and Hsc70. These findings indicate how Hsc70 supports T-antigen action and help to explain the cis requirement for a J domain and Rb binding motif in T-antigen-induced transformation. Furthermore, this is the first demonstration linking Hsc70 ATP hydrolysis to the release of E2F bound by Rb family members.

Mutational analysis of simian virus 40 T antigen: isolation and characterization of mutants with deletions in the T-antigen gene.

A series of mutants of simian virus 40 has been constructed with deletions in the coding sequence for large T antigen. Nucleotide sequence analysis indicates that 4 mutants have in-phase and 11 have out-of-phase deletions. Mutant DNAs were assayed for the following activities: the ability to form plaques, the ability to produce T antigen as scored by indirect immunofluorescence, viral DNA replication, and morphological transformation of rat cells. Two viable mutants were found, and these had deletions confined to the carboxyl terminus of T antigen. Only those mutants coding for polypeptides greater than 40% of the length of wildtype T antigen produced detectable nuclear fluorescence. The two viable mutants with deletions in the carboxyl terminus of the protein retained the ability both to replicate their DNA, although at a reduced level, and to transform nonpermissive cells. Mutants with sequence changes that result in the loss of more than 117 amino acids from the carboxyl terminus were not viable and were also defective in the DNA replication and transformation functions of T antigen, although several produced detectable nuclear fluorescence. These functions were also sensitive to the removal of amino acids near the amino terminus and in the middle of the protein.

Somatic cells efficiently join unrelated DNA segments end-to-end.

Molecular substrates for probing nonhomologous recombination in somatic cells were constructed by inserting pBR322 sequences at selected sites on the simian virus 40 (SV40) genome. The chimeric products are too large to be packaged into an SV40 capsid. Therefore, production of viable progeny requires that most of the pBR322 sequences be deleted without altering any SV40 sequences that are essential for lytic infection. As judged by plaque assay, these recombination events occur at readily detectable frequencies after transfection into CV1 monkey kidney cells. Depending on the site of pBR322 insertion, the infectivities of the full-length circular or linear chimeras ranged from 0.02 to 2% of the infectivity of linear wild-type SV40 DNA. Nucleotide sequence analysis of several recombinant progeny revealed three distinct classes of recombination junction and indicated that the causative recombination events were minimally dependent on sequence homology. Potential mechanisms involving recombination at internal sites or at ends were distinguished by measuring the infectivity of chimeric molecules from which various lengths of pBR322 had been removed. These data support end-to-end joining as the primary mechanism by which DNA segments recombine nonhomologously in somatic cells. This end joining appears to be very efficient, since SV40 genomes with complementary single-stranded tails or with short non-complementary pBR322 tails were comparably infectious. Overall, this study indicates that mammalian somatic cells are quite efficient at the willy-nilly end-to-end joining of unrelated DNA segments.

Biochemical activities of T-antigen proteins encoded by simian virus 40 A gene deletion mutants.

We have analyzed T antigens produced by a set of simian virus 40 (SV40) A gene deletion mutants for ATPase activity and for binding to the SV40 origin of DNA replication. Virus stocks of nonviable SV40 A gene deletion mutants were established in SV40-transformed monkey COS cells. Mutant T antigens were produced in mutant virus-infected CV1 cells. The structures of the mutant T antigens were characterized by immunoprecipitation with monoclonal antibodies directed against distinct regions of the T-antigen molecule. T antigens in crude extracts prepared from cells infected with 10 different mutants were immobilized on polyacrylamide beads with monoclonal antibodies, quantified by Coomassie blue staining, and then assayed directly for T antigen-specific ATPase activity and for binding to the SV40 origin of DNA replication. Our results indicate that the T antigen coding sequences required for origin binding map between 0.54 and 0.35 map units on the SV40 genome. In contrast, sequences closer to the C terminus of T antigen (between 0.24 and 0.20 map units) are required for ATPase activity. The presence of the ATPase activity correlated closely with the ability of the mutant viruses to replicate and to transform nonpermissive cells. The origin binding activity was retained, however, by three mutants that lacked these two functions, indicating that this activity is not sufficient to support either cellular transformation or viral replication. Neither the ATPase activity nor the origin binding activity correlated with the ability of the mutant DNA to activate silent rRNA genes or host cell DNA synthesis.

Simian virus 40 large-T antigen expresses a biological activity complementary to the p300-associated transforming function of the adenovirus E1A gene products.

In this report we present evidence that simian virus 40 T antigen encodes a biological activity that is functionally equivalent to the transforming activity lost by deletion of the E1A p300-binding region. T-antigen constructs from which the pRb-binding region has been deleted are virtually unable to induce foci of transformed cells in a ras cooperation assay in primary baby rat kidney cells. Nevertheless, such a construct can cooperate with an E1A N-terminal deletion mutant, itself devoid of transforming activity, to induce foci in this assay. The heterologous trans-cooperating activity observed between E1A and T-antigen deletion products is as efficient as trans cooperation between mutants expressing individual E1A domains. The cooperating function can be impaired by a deletion near the N terminus of T antigen. Such a deletion impairs neither the p53-binding function nor the activity of the pRb-binding region.

Species-specific elements in the large T-antigen J domain are required for cellular transformation and DNA replication by simian virus 40.

The J domain of simian virus 40 (SV40) large T antigen is required for efficient DNA replication and transformation. Despite previous reports demonstrating the promiscuity of J domains in heterologous systems, results presented here show the requirement for specific J-domain sequences in SV40 large-T-antigen-mediated activities. In particular, chimeric-T-antigen constructs in which the SV40 T-antigen J domain was replaced with that from the yeast Ydj1p or Escherichia coli DnaJ proteins failed to replicate in BSC40 cells and did not transform REF52 cells. However, T antigen containing the JC virus J domain was functional in these assays, although it was less efficient than the wild type. The inability of some large-T-antigen chimeras to promote DNA replication and elicit cellular transformation was not due to a failure to interact with hsc70, since a nonfunctional chimera, containing the DnaJ J domain, bound hsc70. However, this nonfunctional chimeric T antigen was reduced in its ability to stimulate hsc70 ATPase activity and unable to liberate E2F from p130, indicating that transcriptional activation of factors required for cell growth and DNA replication may be compromised. Our data suggest that the T-antigen J domain harbors species-specific elements required for viral activities in vivo.

Use of transgenic mice reveals cell-specific transformation by a simian virus 40 T-antigen amino-terminal mutant.

We have used the multifunctional transforming protein, simian virus 40 T antigen, as a probe to study the mechanisms of cell growth regulation in the intact organism. T antigen appears to perturb cell growth, at least in part, by stably interacting with specific cellular proteins that function to maintain normal cell growth properties. Experiments in cultured cells indicate that at least three distinct regions of simian virus 40 T antigen have roles in transformation. Two regions correlate with the binding of known cellular proteins, p53, pRB, and p107. A third activity, located near the amino terminus, has been defined genetically but not biochemically. By targeting expression of wild-type and mutant forms of T antigen to distinct cell types in transgenic mice, we have begun to systematically determine which activities play a role in tumorigenesis of each cell type. In this study, we sought to determine the role of the amino-terminal transformation function with such an analysis of the T-antigen mutant dl1135. This protein, which lacks amino acids 17 to 27, retains the p53-, pRB-, and p107-binding activities yet fails to transform cells in culture. To direct expression in transgenic mice, we used the lymphotropic papovavirus transcriptional signals that are specific for B and T lymphocytes and the choroid plexus epithelium of the brain. We show here that although defective in cell culture, dl1135 specifically induced the development of thymic lymphomas in the mouse. Expression of the protein was routinely observed in B- and T-lymphoid cells, although B-cell abnormalities were not observed. Choroid plexus tumors were observed only infrequently; however, dl1135 was not consistently expressed in this tissue. Within a given transgenic line, the penetrance of T-cell tumorigenesis was 100% but appeared to require secondary events, as judged from the clonal nature of the tumors. These experiments suggest that the amino-terminal region of T antigen has a role in the transformation of certain cell types (such as fibroblasts in culture and B lymphocytes) but is dispensable for the transformation of T lymphocytes.

Mutational analysis of simian virus 40 T antigen: stimulation of cellular DNA synthesis and activation of rRNA genes by mutants with deletions in the T-antigen gene.

The biological activity of several deletion mutants of simian virus 40, cloned in pBR322, was determined. Three functions of the simian virus 40 A gene were studied: (i) the ability to express T antigen; (ii) the ability to induce cell DNA replication; and (iii) the ability to reactivate silent rRNA genes in hybrid cells. Recombinant plasmid DNA was introduced into cells by manual microinjection or by transfection. The results (together with previous reports) indicate that the critical sequences for these three functions are located separately on the simian virus 40 A gene, as follows: (i) the sequences necessary for the detection of the common antigenic determinant of T antigen extend from nucleotide 4147 to nucleotide 4001 (map units 0.45 to 0.42); (ii) the sequences critical for the stimulation of cell DNA synthesis extend from nucleotide 4327 to nucleotide 4001 (map units 0.49 to 0.42); and (iii) those critical for the reactivation of rRNA genes extend approximately from nucleotide 3827 to nucleotide 3526 (map units 0.39 to 0.33).

The amino-terminal transforming region of simian virus 40 large T and small t antigens functions as a J domain.

Simian virus 40 (SV40) encodes two proteins, large T antigen and small t antigen that contribute to virus-induced tumorigenesis. Both proteins act by targeting key cellular regulatory proteins and altering their function. Known targets of the 708-amino-acid large T antigen include the three members of the retinoblastoma protein family (pRb, p107, and p130), members of the CBP family of transcriptional adapter proteins (cap-binding protein [CBP], p300, and p400), and the tumor suppressor p53. Small t antigen alters the activity of phosphatase pp2A and transactivates the cyclin A promoter. The first 82 amino acids of large T antigen and small t antigen are identical, and genetic experiments suggest that an additional target(s) important for transformation interacts with these sequences. This region contains a motif similar to the J domain, a conserved sequence found in the DnaJ family of molecular chaperones. We show here that mutations within the J domain abrogate the ability of large T antigen to transform mammalian cells. To examine whether a purified 136-amino-acid fragment from the T antigen amino terminus acts as a DnaJ-like chaperone, we investigated whether this fragment stimulates the ATPase activity of two hsc70s and discovered that ATP hydrolysis is stimulated four- to ninefold. In addition, ATPase-defective mutants of full-length T antigen, as well as wild-type small t antigen, stimulated the ATPase activity of hsc70. T antigen derivatives were also able to release an unfolded polypeptide substrate from an hsc70, an activity common to DnaJ chaperones. Because the J domain of T antigen plays essential roles in viral DNA replication, transcriptional control, virion assembly, and tumorigenesis, we conclude that this region may chaperone the rearrangement of multiprotein complexes.

Ribosomal protein genes are overexpressed in colorectal cancer: isolation of a cDNA clone encoding the human S3 ribosomal protein.

We have isolated a cDNA clone encoding the human S3 ribosomal protein from a normal human colon cDNA library. The clone was identified as one of many that detected genes whose level of expression was increased in adenocarcinoma of the colon relative to normal colonic mucosa. Increased levels of the S3 transcript were present in the tumors of all eight patients examined. Moreover, the S3 mRNA was also more abundant in 7 of 10 adenomatous polyps, the presumed precursor of carcinoma. Additional studies demonstrated that increased levels of mRNAs encoding several other ribosomal proteins, including S6, S8, S12, L5, and P0, were present in colorectal tumors and polyps. These results suggest that there is increased synthesis of ribosomes in colorectal tumors and that this increase is an early event in colon neoplasia.

Expression of the gastrin gene in the normal human colon and colorectal adenocarcinoma.

Gastrin, produced in the G-cells of the gastric antrum and regulating acid secretion in the stomach, also acts as a trophic factor in the gastrointestinal tract. Because of its possible role in colon cell proliferation and differentiation, evidence for its presence in normal colorectal mucosa and adenocarcinoma was sought. Utilizing tumors and matched normal mucosa from 26 patients, mature gastrin and progastrin were studied by immunohistochemistry. In normal colonic mucosal crypts, occasional cells stained concordantly for gastrin, progastrin, and chromogranin A, suggesting that they are of neuroendocrine origin. Adenomatous polyps stained neither for gastrin nor chromogranin A. In 22 of 23 adenocarcinomas, more than 50% of tumor cells stained for gastrin and progastrin. The expected gastrin transcript was demonstrable by polymerase chain reaction and RNase protection in tumors and by polymerase chain reaction in normal mucosa. Its identity was confirmed by sequencing the polymerase chain reaction product. A larger transcript containing Intron II was present in both cancers and normal mucosa but was barely discernible in the gastric antrum. Aberrant expression of gastrin may contribute to deregulated proliferation of many colorectal carcinomas.

Distribution of cells expressing myc proteins in human colorectal epithelium, polyps, and malignant tumors.

The myc gene family encodes nuclear phosphoproteins that are thought to play a role in the control of cellular proliferation and differentiation. We have undertaken an immunohistochemical study assessing the expression of myc gene family proteins in individual cells of normal colonic mucosa, colorectal polyps, and colorectal adenocarcinomas. We screened a panel of mouse monoclonal antibodies that we raised against recombinant human c-myc and N-myc proteins for recognition of myc proteins in paraffin tissue sections. Two of these antibodies, H120C69 and H8C150, were selected for indirect immunoperoxidase staining of tissue sections from 16 normal mucosas, 24 polyps, and 30 adenocarcinomas. In normal colon, about 25% of the cells in the lower one-third of the crypts of Lieberkühn stain for myc-related protein. This distribution resembles that of proliferating cells in the crypt. Benign hyperplastic polyps resemble normal mucosa in their myc staining pattern, with about 25% of the cells positive. In adenomatous polyps, the putative precursors of adenocarcinomas, from 50 to 100% of the cells stain positively for myc protein. In these cases, stained cells extend to the luminal surface, consistent with the previously reported expansion of the proliferation zone in these lesions. All adenocarcinomas examined had increased levels of myc protein relative to normal mucosa. The tumor cells exhibited markedly heterogeneous myc staining patterns, both among different tumors and, in some cases, within a single tumor. Comparison with Ki-67 monoclonal antibody staining indicates that myc protein expression in many tumors is uncoupled from cellular proliferation. Surprisingly, we observed increased numbers of myc-expressing cells and increased levels of myc protein in histologically normal colon directly adjacent to tumor, suggesting that many colorectal carcinomas secrete growth factors that activate gene expression in neighboring normal mucosa.

T-antigen mutant activities in vivo: roles of p53 and pRB binding in tumorigenesis of the choroid plexus.

To study the mechanism by which SV40 large T antigen transforms cells under physiological conditions, we analysed several mutant forms of T antigen for their ability to induce cell proliferation and tumorigenesis in transgenic mice. These mutant proteins, which differ in their ability to form complexes with the tumor suppressors pRB and p53, were analysed under conditions in which wild-type T antigen induces choroid plexus papillomas as a result of uniform proliferation of the entire choroid plexus epithelium. The results presented here show that binding of T antigen to p53 is not required for induction of choroid plexus tumors. However, tumorigenesis does appear to require the binding of T antigen to pRB/p107. An additional activity, resident in the amino-terminal one-fifth of the protein, may also play a role. These experiments indicate the importance of whole-animal assays in determining the molecular basis of transformation, since each of these mutants possessed similar transformation phenotypes in culture but showed distinct phenotypes in the choroid plexus of the animal.

Expression of the myc gene family in different stages of human colorectal cancer.

Colorectal carcinoma serves as a model system for the study of changes in gene expression and structure relating to tumorigenesis. In this study, the levels of c-myc, N-myc and L-myc mRNAs were assessed in normal human colonic mucosa in 33 cases representing different stages of adenocarcinoma and in 36 adenomatous polyps, the presumed premalignant stage. Consistent with the findings of Erisman et al. (1985), we found that the c-myc gene was overexpressed (3-24-fold) in approximately two-thirds of the tumors examined. Amplification of the gene (3-4-fold) was observed in 2 of 12 tumors examined and did not correlate with the level of expression. Greater amounts of c-myc-specific mRNA than occur in normal tissue was also found in about two-thirds of the polyps examined demonstrating that premalignant lesions also overexpress the gene. N-myc and L-myc specific transcripts can be detected at low abundance in normal colonic mucosa. These genes were found to be frequently overexpressed in tumors and polyps, but in most cases the level of overexpression was modest. A single case of adenocarcinoma showed an approximately 30-fold increase in the level of N-myc mRNA without gene amplification. Adenomatous polyps more frequently overexpressed the L-myc gene than tumors.