Advertisement
Biology of Human Tumors

Burden and Profile of Somatic Mutation in Duodenal Adenomas from Patients with Familial Adenomatous- and MUTYH-associated Polyposis

Laura E. Thomas, Joanna J. Hurley, Elena Meuser, Sian Jose, Kevin E. Ashelford, Matthew Mort, Shelley Idziaszczyk, Julie Maynard, Helena Leon Brito, Manon Harry, Angharad Walters, Meera Raja, Sarah-Jane Walton, Sunil Dolwani, Geraint T. Williams, Meleri Morgan, Morgan Moorghen, Susan K. Clark and Julian R. Sampson
DOI: 10.1158/1078-0432.CCR-17-1269 Published November 2017

Abstract

Purpose: Duodenal polyposis and cancer are important causes of morbidity and mortality in familial adenomatous polyposis (FAP) and MUTYH-associated polyposis (MAP). This study aimed to comprehensively characterize somatic genetic changes in FAP and MAP duodenal adenomas to better understand duodenal tumorigenesis in these disorders.

Experimental Design: Sixty-nine adenomas were biopsied during endoscopy in 16 FAP and 10 MAP patients with duodenal polyposis. Ten FAP and 10 MAP adenomas and matched blood DNA samples were exome sequenced, 42 further adenomas underwent targeted sequencing, and 47 were studied by array comparative genomic hybridization. Findings in FAP and MAP duodenal adenomas were compared with each other and to the reported mutational landscape in FAP and MAP colorectal adenomas.

Results: MAP duodenal adenomas had significantly more protein-changing somatic mutations (P = 0.018), truncating mutations (P = 0.006), and copy number variants (P = 0.005) than FAP duodenal adenomas, even though MAP patients had lower Spigelman stage duodenal polyposis. Fifteen genes were significantly recurrently mutated. Targeted sequencing of APC, KRAS, PTCHD2, and PLCL1 identified further mutations in each of these genes in additional duodenal adenomas. In contrast to MAP and FAP colorectal adenomas, neither exome nor targeted sequencing identified WTX mutations (P = 0.0017).

Conclusions: The mutational landscapes in FAP and MAP duodenal adenomas overlapped with, but had significant differences to those reported in colorectal adenomas. The significantly higher burden of somatic mutations in MAP than FAP duodenal adenomas despite lower Spigelman stage disease could increase cancer risk in the context of apparently less severe benign disease. Clin Cancer Res; 23(21); 6721–32. ©2017 AACR.

Translational Relevance

Surveillance duodenoscopy is undertaken in patients with familial adenomatous polyposis (FAP) or MUTYH-associated polyposis (MAP) to reduce the risk of duodenal cancer. Current guidelines in the United States and Europe recommend that the screening interval and decisions on interventions are based upon Spigelman staging of duodenal polyposis. In this study we demonstrate a greater mutational burden in MAP than FAP duodenal adenomas despite lower Spigelman stage duodenal polyposis in the MAP patients studied. These findings suggest that the risk of progression to cancer in the context of early-stage duodenal polyposis could be higher in MAP than FAP patients and challenge the assumption that the same surveillance protocols should be applied in MAP and FAP.

Introduction

Familial adenomatous polyposis (FAP) and MUTYH-associated polyposis (MAP) are inherited disorders characterized by colorectal polyposis and cancer. They are also associated with extra-colonic manifestations including polyposis and cancer in the upper gastrointestinal (GI) tract, most notably duodenal disease that has become an important cause of morbidity and mortality as the management of colorectal disease has improved (1). A recent study of FAP estimated the lifetime risk of duodenal polyposis to be 88% and of cancer to be 18% (2). In a multicenter retrospective study of MAP, duodenal polyps were noted in 26 of 150 (17%) patients undergoing duodenoscopy and the lifetime risk of duodenal cancer was estimated at 4% (3). A more recent study in two specialist centers identified duodenal adenomas in 31 of 92 (34%) MAP patients undergoing endoscopy at a median age of 50 years (4).

In patients with FAP or MAP, regular endoscopic surveillance of the duodenum has been advocated from the age of 25–30 years (1). Spigelman staging based upon the number, size, dysplasia, and presence of villous histology of adenomas was developed to better define the severity of duodenal disease in FAP (5) and is recommended to guide the frequency of surveillance, stratify cancer risk, and inform decisions about surgical intervention (6). Duodenal disease in FAP appears to progress slowly through Spigelman stages (0–IV) with an associated increase in cancer risk (7). The natural history of duodenal polyposis in MAP is not well defined but there are reports of duodenal cancer occurring in the context of minimal background polyposis (3, 8). More evidence is required to support or refute current recommendations to apply the same Spigelman stage–based surveillance and intervention for MAP as FAP (1, 6).

Rapid recurrence of duodenal adenomas has been reported following endoscopic polypectomy in patients with FAP (9, 10). Surgical treatments including ampullectomy, duodenectomy, and pancreatico-duodenectomy appear effective for cancer prevention but are associated with significant procedure-associated risks (7, 11). Medical treatment using the cyclooxygenase (COX) inhibitors sulindac and celecoxib has proven less effective in the duodenum than the colorectum (12–15), but a recent trail of combined COX and EGFR inhibition with sulindac and erlotinib demonstrated promising short-term effects on duodenal polyp burden (16). The efficacy of medical and surgical treatment or prevention of duodenal disease in MAP remains unknown.

In colorectal tumorigenesis, the nature and positions of APC mutations appear to determine a critical level of overactivation of β-catenin signaling that leads to a failure in cell growth control without induction of apoptosis (17), a scenario described by the “just right” hypothesis (18). The situation in FAP-associated upper GI tumors appears to be subtly different as somatic APC mutations cluster in a more 3′ region (19). Severe upper intestinal polyposis is also associated with a more 3′ location of inherited APC mutations (19).

Recently, a comprehensive survey of the mutational landscape of colorectal adenomas from patients with FAP and MAP was made using exome sequencing (20). This confirmed the importance of somatic APC and KRAS mutations as drivers of early colorectal tumorigenesis in both disease settings. It also identified frequent somatic mutations of WTX (also known as FAM123B and AMER1) as had been reported previously in sporadic colorectal cancer (21) and that, like APC mutations, may act through deregulation of β-catenin turnover. Although comprehensive molecular genetic studies of duodenal adenomas or carcinomas in patients with FAP have not been reported, targeted sequencing has confirmed a role for APC and revealed oncogenic mutations of KRAS in 9%–30% of FAP duodenal adenomas (22–25). Comparable studies of MAP-associated duodenal tumors have not been reported.

In this study, we applied whole exome and targeted Sanger sequencing and array comparative genomic hybridization (aCGH) to characterize somatic genetic variation associated with the development of duodenal adenomas in patients with FAP and MAP.

Materials and Methods

Patients and samples

Ethical approval was granted by the UK NHS Research Ethics Committee system (reference 10/MRE093). All patients provided written informed consent. This study was completed in accordance with the ethical guidelines of the Declaration of Helsinki. Their diagnoses of FAP or MAP were confirmed by genetic testing. Biopsies of approximately 3 mm of duodenal polyps were taken during upper GI surveillance endoscopy. Spigelman stage was calculated using the method described by Saurin and colleagues (26). A blood sample was taken for automated DNA extraction. A small section of each biopsy was formalin fixed and histopathologic classification, dysplasia by the Vienna classification (27) and proportion of adenomatous material were determined. For the latter, the percentage of epithelial adenoma nuclei was determined in relation to the total number of nuclei comprising adenoma, non-neoplastic crypts, stroma/lamina propria/muscularis mucosae/submucosa, lymphoid, and inflammatory cells. The remainder of each biopsy was snap frozen with liquid nitrogen and stored at −80°C until DNA was extracted using the phenol/chloroform method. A potential limitation in sample characterization was that we could not confirm whether sections used for histopathology were representative of the rest of each biopsy.

Whole exome sequencing

Whole exome sequencing of adenoma and matched blood DNA was performed to a mean depth of coverage of 100× at the Beijing Genomics Institute, Hong Kong, using the SureSelect Human 50 Mb Capture Kit (Agilent) and Illumina platforms. A potential limitation of the chosen depth of coverage is failure to detect somatic variants occurring at very low frequency due to tumor heterogeneity.

Bioinformatic analysis and identification of somatic single nucleotide variants

Details of variant calling can be found in the Supplementary Methods and refs. 28–31.

Validation of somatic mutations

Putative protein changing somatic mutations were validated by PCR and Sanger sequencing of original adenoma DNA samples. When the sequencing depth in a matched blood sample was 20× or less, PCR and Sanger sequencing was also performed on the blood DNA sample. Primers were purchased from Eurofins and PCR was completed as described in the Supplementary Methods.

Identification and analysis of recurrently mutated genes

Recurrently mutated genes were defined as those with ≥2 validated somatic protein changing mutations in the 20 duodenal adenoma exomes. Data for adenomas 37A1 and 37A4 and for adenomas 24A3 and 24A8 were merged as each of these pairs shared a significant proportion of confirmed somatic mutations indicating that they were not independent tumors. Mutations present in each of these pairs were counted only once. To determine which genes were significantly mutated, all validated variants were analyzed using MutSig v1.0 (http://www.broadinstitute.org/cancer/cga/mutsig). To adjust for multiple testing and reduce the false discovery rate, q values were calculated (32). Genes with P < 0.05 (Fisher exact test) and a Q ≤ 0.1 were reported as significantly mutated (see Supplementary Methods for details).

To gain insight into potential mechanisms of tumorigenesis, pathway enrichment analysis was undertaken on all 941 validated somatic mutations using ConsensusPathDB (ref. 33; Supplementary Methods).

Sanger sequencing in additional adenomas

Sanger sequencing of 42 additional adenoma biopsies was used to extend data on somatic mutations in ERBB3, KRAS, PLCL1, PTCHD2, and WTX and of 49 additional adenomas for APC exon 15 (for details see Supplementary Methods).

Loss of heterozygosity analysis

Loss of heterozygosity (LOH) analysis at the APC locus was performed on adenomas in which somatic APC mutations were not identified by sequencing (details in Supplementary Methods). A 50% or greater reduction in an allele relative to constitutional DNA was reported as allelic loss.

Identification and confirmation of somatic copy number variants

Somatic copy number variants (CNV) were identified by aCGH of 47 duodenal adenomas, 26 from FAP patients, and 21 from MAP patients, and matched blood DNA using the BlueGnome CytoChip ISCA 8 × 60k (v2.0) array (GRCh37; Supplementary Methods). Slides were scanned at 3-μm resolution and data were analyzed using CytoGenomics software (Agilent). Each putative CNV was confirmed by either independent aCGH analysis using the Illumina CytoSNP-850k v1.0 chip and data analysis with BlueFuse Multi v3.3 or by quantitative (qPCR) using the 7500 Real-Time PCR system (Applied Biosystems; Supplementary Methods). CNVs found by aCGH in samples that had been exome sequenced were also validated from exome data using ExomeCNV software (ref. 34; Supplementary Methods).

Published data on somatic APC mutations in MAP and FAP adenomas

We compiled a database of somatic APC mutations reported in FAP or MAP duodenal or colorectal adenomas via a literature search in PubMed and Google using the search terms “duodenum,” “colorectum,” “FAP,” “MAP,” and “adenoma.”

Statistical analysis

Statistical analysis was performed using R (version 3.0.2). The Student t test was used to compare the frequencies of single-nucleotide variants (SNV) in FAP and MAP adenomas and Fisher exact test to compare the frequencies of G>T transversions. A P value of less than 0.05 was considered statistically significant. Correlation of adenoma size with number of SNVs and Spigelman stage with number of SNVs was analyzed by Pearson correlation coefficient, where 1 is a perfect positive correlation, 0 is no correlation, and −1 is a perfect negative correlation.

Results

Characterization of patients and adenomas

Biopsies of 72 apparently independent polyps were obtained (1–7 biopsied polyps per patient). Histology confirmed that 69 were adenomas including 42 from 16 patients with FAP and 27 from 10 patients with MAP (Table 1). Two biopsies contained only normal mucosa and one only inflamed ampullary tissue. MAP patients were significantly older than those with FAP (mean 55.0 years vs. 42.9 years, P = 0.006), but had significantly lower Spigelman stage disease (mode stage II vs. stage IV, P = 0.031). Spigelman stage was also lower in MAP than FAP patients from whom adenomas were used for whole exome sequencing (stages II, II, II, II, III vs. III, III, III, IV, respectively). There was no significant difference in the size of biopsied adenomas from FAP and MAP patients (mean 6.93 mm, range 1–30 mm, SD 6.35 mm vs. mean 8.12 mm, range 1.5–25 mm, SD 6.14 mm, P = 0.4255) or in the size of FAP and MAP adenomas used for whole exome sequencing (mean 11.1 mm, range 2–25 mm, SD 7.5 mm vs. mean 11.7 mm, range 3–25 mm, SD 8.26 mm, respectively, P = 0.867). All adenomas showed only low-grade dysplasia and most had tubular morphology with 7 of 42 (17%) of FAP adenomas and 2 of 27 (7%) of MAP adenomas having a villous component (Table 1). The lower Spigelman grade of duodenal disease in MAP than FAP patients reflected smaller adenoma numbers and less frequent villous morphology.

View this table:
Table 1.

Details of patients and adenomas studied

Somatic mutation landscape in FAP and MAP duodenal adenomas

Whole exome sequencing of 20 duodenal adenomas, 10 from 4 patients with FAP and 10 from 5 patients with MAP, together with matched blood DNA identified 1,449 putative protein altering somatic mutations. PCR and Sanger sequencing validated 941 of these (65%, Supplementary Tables S1 and S2) including 28 APC mutations that were identified initially by manual inspection of the exome data and 913 variants in other genes. Eighty-three percent of the validated mutations were nonsynonymous (missense) changes, 13% were stopgains, 2% were splice site mutations, 1% were frameshifts, and one was a stoploss. There were significantly more validated protein-changing somatic mutations in MAP relative to FAP adenomas (mean 71.6, SD 53.56, range 8–167 vs. mean 22.5, SD 13.25, range 1–44, P = 0.0115; t test; Supplementary Fig. S1; Supplementary Table S1). This equated to a mean of 1.43 validated protein changing mutations per Mb in MAP adenoma exomes compared with a mean of 0.44 per Mb in FAP adenoma exomes (Fig. 1). The per-Mb rates of protein changing mutations were broadly comparable with those reported previously in nonhypermutated colorectal cancers (21) with MAP duodenal adenomas being toward the top end of the reported range and FAP duodenal adenomas toward the bottom. However, differences in sequencing and variant calling methods demand caution in such comparisons. The proportion of truncating mutations was also significantly higher in MAP than FAP adenomas (P = 0.006). Of 716 mutations in MAP adenomas 481 (67%) were G>T transversions compared with 28 of 225 (12%) in FAP adenomas (P < 2.2e−16; Fisher exact test), a finding consistent with failure of base excision repair to remove adenine bases mis-incorporated opposite 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG) in MAP adenomas. Pathway enrichment analysis of all validated mutated genes using ConsensusPathDB highlighted over-representation of gene sets involving ECM–receptor interaction networks (q = 0.0125), ERBB (q = 0.0125), BDNF (q = 0.0174), PI3K/AKT (q = 0.0287), EGF, and FGF (q = 0.0414) signaling pathways in FAP adenomas as well as significant enrichment for protein complexes that are part of canonical WNT (q = 0.00516) and MAPK (q = 0.00516) signaling cascades.

Figure 1.
Figure 1.

Box plot showing per megabase (Mb), median, and 25th and 75th percentiles and range of confirmed nonsynonymous SNVs in FAP and MAP duodenal adenomas.

In MAP adenomas, interrogation for protein complex–based sets showed an enrichment for epigenetic transcription regulators (q = 0.00263) as well as molecules important in DNA repair pathways (q = 0.031) and, consequently, over-representation of genes involved in the maintenance of DNA integrity. The number of mutations in different adenomas from the same individual varied greatly (Supplementary Fig. S1).

We also tested for a correlation between adenoma size and the number of confirmed somatic mutations. Although larger adenomas contained more mutations, this did not reach significance for either FAP adenomas (Pearson product–moment correlation, r = 0.62, P = 0.054) or MAP adenomas (r = 0.36, P = 0.303).

Despite appearing to be distinct at endoscopy, MAP adenomas 37A1 and 37A4 shared the same somatic APC mutations and 30 other validated somatic variants. A further 167 validated variants were not shared. MAP adenomas 24A3 and 24A8 also appeared distinct at endoscopy but shared the same somatic APC mutations and 60 other validated variants while 34 validated variants were not shared. The proportions of adenomatous nuclei also differed between adenomas in these pairs (Table 1). Each pair was considered likely to have diverged from a single progenitor lesion and variants in each pair were counted only once in analyses to identify recurrently mutated genes.

Recurrently mutated genes

Sixty-two genes were mutated recurrently in the adenomas subject to whole exome sequencing (Supplementary Table S3) but analysis with MutSig v1.0, which evaluates the number of mutations observed in the context of gene size and the background mutation rate, showed that only 15 were mutated significantly more often than expected (Table 2). Of these, 12 were also mutated significantly in the COSMIC database of somatic mutations in cancer (http://cancer.sanger.ac.uk/cosmic; Table 2). Truncating mutations were observed recurrently in APC, PIGA, TRPM1, and SYNE1 but only APC and PIGA were mutated significantly above the expected background rate. PIGA was not mutated significantly in COSMIC and therefore does not appear to be a driver gene in more extensively studied tumor types.

View this table:
Table 2.

Significantly mutated genes identified by MutSig analysis of mutations in Supplementary Table S3 and COSMIC

Extended analysis of APC, KRAS, PTCHD2, ERBB3, PLCL1, and WTX

To gain further insight into the frequencies and nature of mutations affecting examples of both established and novel candidate driver genes, we extended the analysis of APC (in 49 further duodenal adenomas) and KRAS, PTCHD2, ERBB3, and PLCL1 (in 42 further duodenal adenomas) by Sanger sequencing. PLCL1 was not significantly mutated according to MutSig v1.0, but the four PLCL1 mutations identified during exome sequencing clustered within a region spanning residues 440–547 and this clustering was significant (ref. 35; P = 0.004). Although whole exome sequencing did not identify any mutations in WTX, it was identified recently as a frequently mutated gene in FAP and MAP colorectal adenomas (20) and is also among the most frequently mutated genes in nonhypermutated colorectal cancer (21). We therefore also sequenced WTX in 42 further duodenal adenomas.

Forty further APC mutations were identified by Sanger sequencing (Tables 1 and Supplementary Table S4) and LOH analysis revealed somatic loss affecting three further APC alleles in which sequencing was normal. As aCGH detected no CNVs at the APC locus, the LOH appeared to be copy neutral.

The somatic APC mutations and those reported in previous studies of FAP duodenal adenomas (see Supplementary Table S4) clustered 3′ to the third (last) β-catenin binding 20 amino acid repeat. This nonrandom clustering was highly significant [P = 9.11 × 10−10 by the method of Ye and colleagues (35)] and different to the clustering of somatic APC mutations in FAP-associated colorectal adenomas (Supplementary Table S4) that occurs after the first and second 20 amino acid repeats (P < 3.72 × 10−16 and P < 3.88 × 10−29). In FAP duodenal adenomas, 15 of the 30 APC mutations we identified were insertion of an A in the A6 tract at codons 1554-6 (c.4659dupA; E1554fsX5). This mutation also accounted for 17 of 35 previously reported somatic APC mutations in FAP duodenal tumors but only 1 of 296 in FAP colorectal adenomas (Supplementary Table S4, P < 0.0001; Fisher exact test).

In MAP duodenal adenomas where biallelic APC mutations were identified, significant clustering occurred between codons 1530 and 1576 (P = 1.25 × 10−7) despite the presence of GAA sequences throughout the coding region that could be mutated to stop codons by G>T transversion with only one instance of E1554fsX5 observed (in the adenoma pair 37A1 and 37A4; Supplementary Table S4).

We did not observe any somatic WTX mutations in 60 independent duodenal adenomas (Table 3). This was significantly different (P = 0.0038, Fisher exact test) to the findings reported by Rashid and colleagues (20) in FAP and MAP colorectal adenomas, where 17 truncating mutations were identified in 128 adenomas, making WTX the most frequently mutated gene after APC. WTX forms a complex with APC, Axin, and β-TrCP2 that degrades β-catenin. It is likely that the differences we observed between duodenal and colorectal adenomas in the positions or presence of APC and WTX mutations reflect different requirements for β-catenin signaling for tumorigenesis in these contexts.

View this table:
Table 3.

Summary of somatic analyses completed including exome analysis, ArrayCGH, APC LOH analysis, and targeted sequencing of APC, KRAS, WTX, PTCHD2, ERBB3, and PLCL1

After APC, KRAS was the most frequently mutated gene in duodenal adenomas (12/60, 20%) and KRAS mutations were significantly more frequent in MAP than FAP adenomas (8/22 vs. 4/38, P < 0.023, Fisher exact test). Only 3 of 8 KRAS mutations in MAP duodenal adenomas were the c.34 G:C>T:A (G12C) mutation that has been considered as a potential biomarker of MAP in patients with multiple colorectal adenomas (36). MAP patients whose adenomas harbored KRAS mutations appeared to have lower Spigelman stage polyposis than corresponding FAP patients (stages II, II, II, II, III in MAP vs. II, IV, IV in FAP; Table 1).

Six somatic PTCHD2 mutations were identified in 60 independent adenomas, three by whole exome sequencing and three by sequencing of additional adenomas. Five had CADD scores above 20 (i.e., corresponding to the top 1% of substitutions in terms of predicted deleterious effects). Adenomas 3A2 and 37A1 each contained two PTCHD2 mutations but one of those in 37A1 was unlikely to be of functional significance (Supplementary Table S5). Six independent PLCL1 mutations were also observed: four in whole exomes and two following targeted sequencing. The latter two did not cluster with the others (Supplementary Table S5). All but one of the PLCL1 mutations had CADD scores above 20. No further mutations of ERBB3 were identified by analysis of the 42 additional adenomas but the two mutations identified during exome sequencing had CADD scores of 28.4 and 30 and are very likely to impact function (Supplementary Table S5).

Array CGH

Array CGH revealed eight CNVs (five losses and three gains) in five of 19 MAP duodenal adenomas (Table 4) compared with none in 26 FAP adenomas (P = 0.0052, Fisher exact test). All were confirmed by either quantitative PCR or by using a second array, the Illumina CytoSNP-850k v1.0. Several involved genes in the BMP/TGFβ signaling pathway: the deletion at 18q21.1 in adenoma 44A4 included SMAD4 and that at 9q22 included ENG, whereas the 15q11.1-15q21.1 gains in adenomas 23A3 and 23A4 included GREM1, a BMP antagonist.

View this table:
Table 4.

Summary of CNVs detected by array CGH

Discussion

Duodenal polyposis and cancer present a major challenge in the clinical management of FAP and MAP, but remain understudied and poorly understood. This study is the first to characterize comprehensively the burden and pattern of somatic mutations in duodenal adenomas from patients with FAP or MAP. We found that MAP duodenal adenomas carried a significantly higher burden of somatic protein-changing mutations, truncating mutations, and CNVs than FAP duodenal adenomas even though MAP patients had lower Spigelman stage duodenal polyposis than FAP patients. The greater mutation burden in MAP adenomas appears to reflect defective base excision repair. Although longitudinal or prospective studies of duodenal polyposis in MAP have not been reported, case reports have highlighted the occurrence of duodenal cancer in MAP patients in the absence of advanced duodenal polyposis (3, 8). These observations and our data suggest that current recommendations to manage MAP duodenal polyposis using Spigelman staging in the same way as for FAP (1, 6) may not be appropriate. A low polyp count in a patient with MAP may be falsely reassuring and, in addition, we did not find a significant correlation between adenoma size and mutation burden. Mutation burdens in some small MAP adenomas were among the highest we observed. Large, prospective clinical studies could provide a better evidence base for duodenal surveillance recommendations and intervention in MAP.

Our data confirm the importance of APC and KRAS mutations as drivers of duodenal tumorigenesis in FAP and MAP but show that in contrast with the colorectum (20, 21, 37, 38) WTX is not a significant driver gene in early duodenal tumorigenesis. Neither did we identify by exome sequencing any mutations in a number of known driver genes including NRAS, CTNNB1, FBXW7, and TP53 that were mutated recurrently in previous studies of sporadic or FAP-associated colorectal adenomas (37, 39) and that are also mutated in sporadic duodenal adenocarcinomas (40, 41). They may be mutated later in duodenal tumorigenesis.

The somatic APC mutations we identified in FAP and MAP duodenal adenomas clustered 3′ to the mutation cluster region observed in FAP-associated and sporadic colorectal adenomas and cancers. Groves and colleagues (19) and Miyaki and colleagues (42) have reported similar findings. These more 3′ mutations are predicted to lead to truncated APC proteins that retain three β-catenin binding 20 AA repeats in the majority of duodenal tumors rather than either one or two repeats as occurs in colorectal tumors. In FAP duodenal adenomas, we found that 14 of 25 (56%) somatic APC mutations were ins A mutations at codons 1554-6 (4661 G>GA c.4659dupA; E1554fxs4). This is consistent with data we compiled from previous reports in which this mutation accounted for 17/35 mutations (49%). Although very uncommon in FAP colorectal adenomas (1/296 mutations in the reports we identified; Supplementary Table S4), this mutation has been seen recurrently in colorectal adenomas from patients with attenuated FAP (43–45) where it appears to occur as a “third hit” further reducing the activity of the attenuated germline mutant allele. We did not find any evidence for third hits affecting APC in duodenal adenomas. Instead, this change and the others clustering after the third 20 AA repeat are likely to be selected for as second hits in duodenal tumorigenesis because they determine a specific level of β-catenin signaling that is lower than that selected for in colorectal tumorigenesis. A subtly different β-catenin signaling requirement in duodenal adenomas may also explain the absence of WTX mutations.

In addition to APC and KRAS, 10 of the 13 other genes that were mutated significantly upon whole exome sequencing of duodenal adenomas are also mutated significantly in the COSMIC database of somatic mutations in cancer (Table 2). These genes are likely to be drivers in FAP and MAP duodenal tumors as well as in other tumor types. Following whole exome sequencing, we investigated the recurrently mutated genes PTCHD2, ERBB3, and PLCL1 in a set of 42 additional duodenal adenomas. We identified further mutations in PTCHD2 (N = 3) and PLCL1 (N = 2), supporting a role for these genes as drivers in duodenal tumorigenesis. PLCL1 encodes a multivalent adaptor protein (46). Four of six mutations identified in this study were missense changes clustered around the X-Box region of the PLC core domain. A truncating mutation of PLCL1 (S931X) was also identified in 1 of 14 colorectal adenoma exomes in the study of Rashid and colleagues (20). PTCHD2 (DISP3) has been assigned to the family of Patched-domain containing receptors based on in silico characterization and is likely involved in Hedgehog signaling (47).

A number of genes such as MLL3 and ATRNL1 in which we identified only single truncating mutations were also mutated recurrently in FAP and/or MAP colorectal adenomas in other recent studies (20). They represent candidate driver genes in duodenal as well as colorectal tumorigenesis. aCGH identified CNVs exclusively in MAP duodenal adenomas and several included genes (SMAD4, ENG, and GREM1) that regulate BMP signaling and have established roles in GI cancer. aCGH lacks sensitivity in the context of heterogeneous tumor samples that comprise a mixture of neoplastic and nonneoplastic cells and we are likely to have underestimated the true frequency of CNVs. Pathway enrichment analysis of all validated mutations provided an approach to evaluate the potential roles of multiple genes with related functions. It highlighted involvement of Wnt, ERBB, PI3K/AKT, EGF, FGF, and ECM-receptor signaling in FAP adenomas and of DNA repair pathways and epigenetic transcription regulators in MAP adenomas. Dysregulation of these pathways is well established in tumorigenesis and they are targets for drugs in clinical use or under development. So far, only EGF signaling has been targeted in clinical trials for duodenal polyposis (16). Our data point to additional and novel opportunities for intervention but they also highlight the molecular genetic heterogeneity of duodenal adenomas. Only genes that regulate the Wnt pathway were mutated consistently. The highly specific and restricted pattern of APC mutation and the absence of WTX mutation that we observed in duodenal adenomas suggest that a narrow range of β-catenin activity may be required for duodenal tumorigenesis. Therapeutic manipulation of this activity may hold particular promise for prevention and treatment.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Authors' Contributions

Conception and design: L.E. Thomas, J.J. Hurley, S. Dolwani, J.R. Sampson

Development of methodology: L.E. Thomas, J.J. Hurley, M. Moorghen

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.J. Hurley, E. Meuser, H.L. Brito, A. Walters, M. Raja, S.-J. Walton, S. Dolwani, G.T. Williams, S.K. Clark, J.R. Sampson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L.E. Thomas, J.J. Hurley, E. Meuser, S. Jose, K.E. Ashelford, M. Mort, H.L. Brito, M. Harry, A. Walters, M. Raja, M. Morgan, M. Moorghen, S.K. Clark, J.R. Sampson

Writing, review, and/or revision of the manuscript: L.E. Thomas, J.J. Hurley, E. Meuser, K.E. Ashelford, H.L. Brito, S.-J. Walton, S. Dolwani, G.T. Williams, M. Morgan, S.K. Clark, J.R. Sampson

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.E. Thomas, E. Meuser, S. Idziaszczyk, J. Maynard, H.L. Brito, A. Walters

Study supervision: L.E. Thomas, S. Dolwani, S.K. Clark, J.R. Sampson

Other (histopathologic analysis of study samples): G.T. Williams

Other (pathology interpretation): M. Morgan

Grant Support

This project has been funded by the Welsh Government through Health and Care Research Wales through a NISCHR Fellowship to L.E. Thomas, and by the Wales Gene Park and Wales Cancer Research Centre.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Acknowledgments

Special thanks to Dr. Fiona Lalloo and Mr. Jim Hill for assistance with communication with participating patients.

Footnotes

  • Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

  • Received May 2, 2017.
  • Revision received June 21, 2017.
  • Accepted July 25, 2017.

References

  1. 1.↵
    1. Vasen HFA,
    2. Möslein G,
    3. Alonso A,
    4. Aretz S,
    5. Bernstein I,
    6. Bertario L,
    7. et al.
    Guidelines for the clinical management of familial adenomatous polyposis (FAP). Gut 2008;57:704–13.
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    1. Bülow S,
    2. Christensen IJ,
    3. Højen H,
    4. Björk J,
    5. Elmberg M,
    6. Järvinen H,
    7. et al.
    Duodenal surveillance improves the prognosis after duodenal cancer in familial adenomatous polyposis. Colorectal Dis 2012;14:947–52.
    OpenUrlCrossRefPubMed
  3. 3.↵
    1. Vogt S,
    2. Jones N,
    3. Christian D,
    4. Engel C,
    5. Nielsen M,
    6. Kaufmann A,
    7. et al.
    Expanded extracolonic tumor spectrum in MUTYH-associated polyposis. Gastroenterology 2009;137:1976–1985.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Walton S-J,
    2. Kallenberg FGJ,
    3. Clark SK,
    4. Dekker E,
    5. Latchford A
    . Frequency and features of duodenal adenomas in patients with MUTYH-associated polyposis. Clin Gastroenterol Hepatol 2016;14:986–92.
    OpenUrl
  5. 5.↵
    1. Spigelman AD,
    2. Williams CB,
    3. Talbot IC,
    4. Domizio P,
    5. Phillips RK
    . Upper gastrointestinal cancer in patients with familial adenomatous polyposis. Lancet 1989;2:783–5.
    OpenUrlPubMed
  6. 6.↵
    1. Syngal S,
    2. Brand RE,
    3. Church JM,
    4. Giardiello FM,
    5. Hampel HL,
    6. Burt RW,
    7. et al.
    ACG clinical guideline: genetic testing and management of hereditary gastrointestinal cancer syndromes. Am J Gastroenterol 2015;110:223–62.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Serrano PE,
    2. Grant RC,
    3. Berk TC,
    4. Kim D,
    5. Al-Ali H,
    6. Cohen Z,
    7. et al.
    Progression and management of duodenal neoplasia in familial adenomatous polyposis. Ann Surg 2015;261:1138–44.
    OpenUrl
  8. 8.↵
    1. Nielsen M,
    2. Poley JW,
    3. Verhoef S,
    4. van Puijenbroek M,
    5. Weiss MM,
    6. Burger GT,
    7. et al.
    Duodenal carcinoma in MUTYH-associated polyposis. J Clin Pathol 2006;59:1212–5.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Alderlieste YA,
    2. Bastiaansen BA,
    3. Mathus-Vliegen EMH,
    4. Gouma DJ,
    5. Dekker E
    . High rate of recurrent adenomatosis during endoscopic surveillance after duodenectomy in patients with familial adenomatous polyposis. Fam Cancer 2013;12:699–706.
    OpenUrl
  10. 10.↵
    1. Ma T,
    2. Jang EJ,
    3. Zukerberg LR,
    4. Odze R,
    5. Gala MK,
    6. Kelsey PB,
    7. et al.
    Recurrences are common after endoscopic ampullectomy for adenoma in the familial adenomatous polyposis (FAP) syndrome. Surg Endosc 2014;28:2349–56.
    OpenUrl
  11. 11.↵
    1. Heiskanen I,
    2. Kellokumpu I,
    3. Järvinen H
    . Management of duodenal adenomas in 98 patients with familial adenomatous polyposis. Endoscopy 1999;31:412–6.
    OpenUrlCrossRefPubMed
  12. 12.↵
    1. Steinbach G,
    2. Lynch PM,
    3. Phillips RKS,
    4. Wallace MH,
    5. Hawk E,
    6. Gordon GB,
    7. et al.
    The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–52.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Giardiello FM,
    2. Yang VW,
    3. Hylind LM,
    4. Krush AJ,
    5. Petersen GM,
    6. Trimbath JD,
    7. et al.
    Primary chemoprevention of familial adenomatous polyposis with sulindac. N Engl J Med 2002;346:1054–9.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Phillips RKS,
    2. Wallace MH,
    3. Lynch PM,
    4. Hawk E,
    5. Gordon GB,
    6. Saunders BP,
    7. et al.
    A randomised, double blind, placebo controlled study of celecoxib, a selective cyclooxygenase 2 inhibitor, on duodenal polyposis in familial adenomatous polyposis. Gut 2002;50:857–60.
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    1. Arber N,
    2. Eagle CJ,
    3. Spicak J,
    4. Rácz I,
    5. Dite P,
    6. Hajer J,
    7. et al.
    Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355:885–95.
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Samadder NJ,
    2. Neklason DW,
    3. Boucher KM,
    4. Byrne KR,
    5. Kanth P,
    6. Samowitz W,
    7. et al.
    Effect of sulindac and erlotinib vs. placebo on duodenal neoplasia in familial adenomatous polyposis. JAMA 2016;315:1266.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Lamlum H,
    2. Ilyas M,
    3. Rowan A,
    4. Clark S,
    5. Johnson V,
    6. Bell J,
    7. et al.
    The type of somatic mutation at APC in familial adenomatous polyposis is determined by the site of the germline mutation: a new facet to Knudson's ‘two-hit’ hypothesis. Nat Med 1999;5:1071–5.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Albuquerque C,
    2. Breukel C,
    3. van der Luijt R,
    4. Fidalgo P,
    5. Lage P,
    6. Slors FJM,
    7. et al.
    The ‘just-right’ signaling model: APC somatic mutations are selected based on a specific level of activation of the beta-catenin signaling cascade. Hum Mol Genet 2002;11:1549–60.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Groves C,
    2. Lamlum H,
    3. Crabtree M,
    4. Williamson J,
    5. Taylor C,
    6. Bass S,
    7. et al.
    Mutation cluster region, association between germline and somatic mutations and genotype-phenotype correlation in upper gastrointestinal familial adenomatous polyposis. Am J Pathol 2002;160:2055–61.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Rashid M,
    2. Fischer A,
    3. Wilson CH,
    4. Tiffen J,
    5. Rust AG,
    6. Stevens P,
    7. et al.
    Adenoma development in familial adenomatous polyposis and MUTYH-associated polyposis: somatic landscape and driver genes. J Pathol 2016;238:98–108.
    OpenUrlCrossRefPubMed
  21. 21.↵
    The Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature 2012;487:330–7.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Gallinger S,
    2. Vivona AA,
    3. Odze RD,
    4. Mitri A,
    5. O'Beirne CP,
    6. Berk TC,
    7. et al.
    Somatic APC and K-ras codon 12 mutations in periampullary adenomas and carcinomas from familial adenomatous polyposis patients. Oncogene 1995;10:1875–8.
    OpenUrlPubMed
  23. 23.↵
    1. Kashiwagi H,
    2. Spigelman AD,
    3. Talbot IC,
    4. Debinski HS,
    5. McKie AB,
    6. Lemoine NR,
    7. et al.
    p53 and K-ras status in duodenal adenomas in familial adenomatous polyposis. Br J Surg 1997;84:826–9.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Norheim Andersen S,
    2. Løvig T,
    3. Fausa O,
    4. Rognum TO
    . Germline and somatic mutations in exon 15 of the APC gene and K-ras mutations in duodenal adenomas in patients with familial adenomatous polyposis. Scand J Gastroenterol 1999;34:611–7.
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Wagner PL,
    2. Chen Y-T,
    3. Yantiss RK
    . Immunohistochemical and molecular features of sporadic and FAP-associated duodenal adenomas of the ampullary and nonampullary mucosa. Am J Surg Pathol 2008;32:1388–95.
    OpenUrlCrossRefPubMed
  26. 26.↵
    1. Saurin J-C,
    2. Gutknecht C,
    3. Napoleon B,
    4. Chavaillon A,
    5. Ecochard R,
    6. Scoazec J-Y,
    7. et al.
    Surveillance of duodenal adenomas in familial adenomatous polyposis reveals high cumulative risk of advanced disease. J Clin Oncol 2003;22:493–8.
    OpenUrl
  27. 27.↵
    1. Schlemper RJ,
    2. Riddell RH,
    3. Kato Y,
    4. Borchard F,
    5. Cooper HS,
    6. Dawsey SM,
    7. et al.
    The Vienna classification of gastrointestinal epithelial neoplasia. Gut 2000;47:251–5.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    1. Li H
    . A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 2011;27:2987–93.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. McKenna A,
    2. Hanna M,
    3. Banks E,
    4. Sivachenko A,
    5. Cibulskis K,
    6. Kernytsky A,
    7. et al.
    The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res 2010;20:1297–303.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Koboldt DC,
    2. Zhang Q,
    3. Larson DE,
    4. Shen D,
    5. McLellan MD,
    6. Lin L,
    7. et al.
    VarScan 2: Somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res 2012;22:568–76.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Larson DE,
    2. Harris CC,
    3. Chen K,
    4. Koboldt DC,
    5. Abbott TE,
    6. Dooling DJ,
    7. et al.
    SomaticSniper: identification of somatic point mutations in whole genome sequencing data. Bioinformatics 2012;28:311–7.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Storey JD
    . The positive false discovery rate: a Bayesian interpretation and the q-value. Ann Stat 2003;31:2013–35.
    OpenUrlCrossRef
  33. 33.↵
    1. Herwig R,
    2. Hardt C,
    3. Lienhard M,
    4. Kamburov A
    . Analyzing and interpreting genome data at the network level with ConsensusPathDB. Nat Protoc 2016;11:1889–907.
    OpenUrl
  34. 34.↵
    1. Sathirapongsasuti JF,
    2. Lee H,
    3. Horst BAJ,
    4. Brunner G,
    5. Cochran AJ,
    6. Binder S,
    7. et al.
    Exome sequencing-based copy-number variation and loss of heterozygosity detection: ExomeCNV. Bioinformatics 2011;27:2648–54.
    OpenUrlCrossRefPubMed
  35. 35.↵
    1. Ye J,
    2. Pavlicek A,
    3. Lunney EA,
    4. Rejto PA,
    5. Teng C-H
    . Statistical method on nonrandom clustering with application to somatic mutations in cancer. BMC Bioinformatics 2010;11:11.
    OpenUrlCrossRefPubMed
  36. 36.↵
    1. Nielsen M,
    2. Morreau H,
    3. Vasen HFA,
    4. Hes FJ
    . MUTYH-associated polyposis (MAP). Crit Rev Oncol Hematol 2011;79:1–16.
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Nikolaev SI,
    2. Sotiriou SK,
    3. Pateras IS,
    4. Santoni F,
    5. Sougioultzis S,
    6. Edgren H,
    7. et al.
    A single-nucleotide substitution mutator phenotype revealed by exome sequencing of human colon adenomas. Cancer Res 2012;72:6279–89.
    OpenUrlAbstract/FREE Full Text
  38. 38.↵
    1. Sanz-Pamplona R,
    2. Lopez-Doriga A,
    3. Pare-Brunet L,
    4. Lazaro K,
    5. Bellido F,
    6. Alonso MH,
    7. et al.
    Exome sequencing reveals AMER1 as a frequently mutated gene in colorectal cancer. Clin Cancer Res 2015;21:4709–18.
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    1. Borras E,
    2. San Lucas FA,
    3. Chang K,
    4. Zhou R,
    5. Masand G,
    6. Fowler J,
    7. et al.
    Genomic landscape of colorectal mucosa and adenomas. Cancer Prev Res 2016;9:417–27.
    OpenUrlAbstract/FREE Full Text
  40. 40.↵
    1. Laforest A,
    2. Aparicio T,
    3. Zaanan A,
    4. Silva FP,
    5. Didelot A,
    6. Desbeaux A,
    7. et al.
    ERBB2 gene as a potential therapeutic target in small bowel adenocarcinoma. Eur J Cancer 2014;50:1740–6.
    OpenUrlCrossRefPubMed
  41. 41.↵
    1. Yuan W,
    2. Zhang Z,
    3. Dai B,
    4. Wei Q,
    5. Liu J,
    6. Liu Y,
    7. et al.
    Whole-exome sequencing of duodenal adenocarcinoma identifies recurrent Wnt/beta-catenin signaling pathway mutations. Cancer 2016;122:1689–96.
    OpenUrl
  42. 42.↵
    1. Miyaki M,
    2. Yamaguchi T,
    3. Iijima T,
    4. Takahashi K,
    5. Matsumoto H,
    6. Yasutome M,
    7. et al.
    Difference in characteristics of APC mutations between colonic and extracolonic tumors of FAP patients: variations with phenotype. Int J Cancer 2008;122:2491–7.
    OpenUrlCrossRefPubMed
  43. 43.↵
    1. Spirio LN,
    2. Samowitz W,
    3. Robertson J,
    4. Robertson M,
    5. Burt RW,
    6. Leppert M,
    7. et al.
    Alleles of APC modulate the frequency and classes of mutations that lead to colon polyps. Nat Genet 1998;20:385–8.
    OpenUrlCrossRefPubMed
  44. 44.↵
    1. Su L-K,
    2. Barnes CJ,
    3. Yao W,
    4. Qi Y,
    5. Lynch PM,
    6. Steinbach G
    . Inactivation of germline mutant APC alleles by attenuated somatic mutations: a molecular genetic mechanism for attenuated familial adenomatous polyposis. Am J Hum Genet 2000;67:582–90.
    OpenUrlCrossRefPubMed
  45. 45.↵
    1. Sieber OM,
    2. Segditsas S,
    3. Knudsen AL,
    4. Zhang J,
    5. Luz J,
    6. Rowan AJ,
    7. et al.
    Disease severity and genetic pathways in attenuated familial adenomatous polyposis vary greatly but depend on the site of the germline mutation. Gut 2006;55:1440–8.
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    1. Sugiyama G,
    2. Takeuchi H,
    3. Kanematsu T,
    4. Gao J,
    5. Matsuda M,
    6. Hirata M
    . Phospholipase C-related but catalytically inactive protein, PRIP as a scaffolding protein for phospho-regulation. Adv Biol Regul 2013;53:331–40.
    OpenUrlCrossRef
  47. 47.↵
    1. Katoh Y,
    2. Katoh M
    . Identification and characterization of DISP3 gene insilico. Int J Oncol 2005;26:551–6.
    OpenUrlPubMed
View Abstract
Back to top
View this article with LENS

Open full page PDF
Article Alerts
Email Article
Citation Tools
Burden and Profile of Somatic Mutation in Duodenal Adenomas from Patients with Familial Adenomatous- and MUTYH-associated Polyposis
Laura E. Thomas, Joanna J. Hurley, Elena Meuser, Sian Jose, Kevin E. Ashelford, Matthew Mort, Shelley Idziaszczyk, Julie Maynard, Helena Leon Brito, Manon Harry, Angharad Walters, Meera Raja, Sarah-Jane Walton, Sunil Dolwani, Geraint T. Williams, Meleri Morgan, Morgan Moorghen, Susan K. Clark and Julian R. Sampson
Clin Cancer Res November 1 2017 (23) (21) 6721-6732; DOI: 10.1158/1078-0432.CCR-17-1269
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement