Next-generation sequencing is highly sensitive for the ...

2 downloads 22 Views 189KB Size Report
Mar 12, 2015 - Gardner's syndrome [11]. Next-generation sequencing (NGS) is increasingly being used to support histological tissue diagnoses. Importantly,.
Virchows Arch DOI 10.1007/s00428-015-1765-0

ORIGINAL ARTICLE

Next-generation sequencing is highly sensitive for the detection of beta-catenin mutations in desmoid-type fibromatoses Sarah J. Aitken 1,2,5 & Nadège Presneau 1,6 & Sangeetha Kalimuthu 3 & Palma Dileo 4 & Fitim Berisha 3 & Roberto Tirabosco 3 & M. Fernanda Amary 1,3 & Adrienne M. Flanagan 1,3

Received: 27 January 2015 / Revised: 12 March 2015 / Accepted: 19 March 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract Desmoid-type fibromatoses are locally aggressive and frequently recurrent tumours, and an accurate diagnosis is essential for patient management. The majority of sporadic lesions harbour beta-catenin (CTNNB1) mutations. We used next-generation sequencing to detect CTNNB1 mutations and to compare the sensitivity and specificity of next-generation sequencing with currently employed mutation detection techniques: mutation-specific restriction enzyme digestion and polymerase chain reaction amplification. DNA was extracted from formalin-fixed paraffin-embedded needle biopsy or resection tissue sections from 144 patients with sporadic desmoid-type fibromatoses, four patients with syndromerelated desmoid-type fibromatoses and 11 morphological mimics. Two primer pairs were designed for CTNNB1 mutation hotspots. Using ≥10 ng of DNA, libraries were generated by Fluidigm and sequenced on the Ion Torrent Personal Genome Machine. Next-generation sequencing had a sensitivity of 92.36 % (133/144, 95 % CIs: 86.74 to 96.12 %) and a

specificity of 100 % for the detection of CTNNB1 mutations in desmoid-type fibromatoses-like spindle cell lesions. All mutations detected by mutation-specific restriction enzyme digestion were identified by next-generation sequencing. Next-generation sequencing identified additional mutations in 11 tumours that were not detected by mutation-specific restriction enzyme digestion, two of which have not been previously described. Next-generation sequencing is highly sensitive for the detection of CTNNB1 mutations. This multiplex assay has the advantage of detecting additional mutations compared to those detected by mutation-specific restriction enzyme digestion (sensitivity 82.41 %). The technology requires minimal DNA and is time- and cost-efficient. Keywords Next-generation sequencing . Desmoid-type fibromatosis . Beta-catenin . CTNNB1 . Molecular pathology

Introduction * Adrienne M. Flanagan [email protected] 1

Research Department of Pathology, UCL Cancer Institute, London WC1E 6BT, UK

2

Department of Histopathology, Addenbrooke’s Hospital, Cambridge CB2 0QQ, UK

3

Royal National Orthopaedic Hospital NHS Trust, Stanmore, Middlesex HA7 4LP, UK

4

University College Hospital, London, UK

5

Cancer Research UK Cambridge Institute, University of Cambridge, Cambridge CB2 0RE, UK

6

Department of Biomedical Sciences, University of Westminster, London, UK

Desmoid-type fibromatoses (DTFs) are soft tissue neoplasms sited deep to subcutaneous tissue and fascia. DTFs have an infiltrative growth pattern and are locally aggressive with a propensity to recur, but they do not metastasise or dedifferentiate (WHO classification: intermediate (locally aggressive, non-metastasising [1]). DTFs are rare, account for G p.T41A, c. 134C>T p.S45F and c.133T>C p.S45P), and codons 15–22, which include the next most frequent mutations. Bidirectional tags were added to each primer: forward primers the common sequence (CS) CS1 ACACTGACGACATGGT TCTACA and reverse primers with CS2 TACGGTAGCAGA GACTTGGTCT (Table 1). Polymerase chain reaction (PCR) using control DNA was used to check that the primers amplified a single amplicon of

Virchows Arch Table 1

Primers with tags Product size

CTNNB1 34-45 forward CTNNB1 34-45 reverse CTNNB1 15-22 forward CTNNB1 15-22 reverse

ACACTGACGACATGGTTCTACACCATGGAACCAGACAGAAAAGC TACGGTAGCAGAGACTTGGTCTGGTATCCACATCCTCTTCCTC ACACTGACGACATGGTTCTACAcaatctactaatgctaatactgtttcg TACGGTAGCAGAGACTTGGTCTgtagtggcaccagaatggattcc

the correct size on a 6 % acrylamide gel with a 100–400 bp ladder (200 V constant for 40 min). Pre-amplification Pre-amplification templates were designed to include three negative controls and at least three positive controls, with samples in all other wells of 48 wells of a 96-well plate. Briefly, 1 μl (≥10 ng) of DNA or PCR-grade water was added to each well containing 9 μl of FastStart High Fidelity PCR System master mix (Roche Applied Science, Penzberg, Germany). The following PCR protocol was then used: 95 °C for 10 min, 2 cycles of 95 °C for 15 s followed by 60 °C for 4 min, then 13 cycles of 95 °C for 15 s and 72 °C for 4 min. The PCR products were then cleaned by adding 4 μl ExoSAP-IT (exonuclease I and shrimp alkaline phosphatase) enzyme mix (Affymetrix, Santa Clara, California, USA) to each well and thermocycled at 37 °C for 15 min and then at 80 °C for 15 min. The cleaned products were diluted by adding 5 μl of product to 13 μl of water. Fluidigm A 48× 48 Access Array Integrated Fluidic Circuit (IFC) (Fluidigm, San Francisco, California, USA) was primed according to the manufacturer’s instructions. One microlitre of diluted pre-amplification product was added to 4 μl of FastStart High Fidelity PCR System master mix (Roche Applied Science). The IFC was then loaded with 4 μl of this sample mix solution into each sample inlet and 4 μl of the primer solutions into each primer inlet. The IFC was cycled (48×48 Standard v1). The PCR products were harvested and diluted 1 μl in 99 μl of water.

140 bp 145 bp

One microlitre of each barcoded DNA sample was pooled to form the library, which was purified by adding 52.8 μl (1.1× volume) of AMPure Reagent (Beckman Coulter, Pasadena, California, USA). The mixture was placed on the DynaMag (Life Technologies, Carlsbad, California, USA) magnet and washed twice with 200 μl of 70 % ethanol, and then, the DNA was eluted in 20 μl of PCR-grade water. The cleaned library was analysed using a 6 % acrylamide gel (100 V for 90 min) and compared to previous libraries to establish the dilatation ratio. Emulsion PCR Twenty five microlitres of the diluted library was added to the Ion OneTouch™ 200 Template Kit v2 amplification solution (Life Technologies) according to the manufacturer’s instructions; 100 μl of ion sphere particles (ISPs) was added and then transferred to the OneTouch reaction assembly. Emulsion PCR was carried out overnight using the Ion OneTouch™ 2 System (Life Technologies). The template-positive ISPs were then enriched according to the manufacturer’s instructions. The Ion Sphere™ Quality Control Kit (Life Technologies) was used to assess the quality of the library pre- and postenrichment, which was measured using the Qubit 2.0 Fluorometer (Life Technologies) according to the manufacturer’s instructions. Pre-enrichment samples should have 10– 30 % template-positive ISPs; post-enrichment samples should have >70 % template-positive ISPs, and all values must be ≥200 RFU to ensure adequate depth of sequencing without an excess of polyclonal beads. Sequencing

Barcoding Custom-designed ten-base oligonucleotides were used to identify each sample. One microlitre of the diluted harvest product was added to 15 μl of FastStart High Fidelity PCR System master mix (Roche Applied Science) and 4 μl of barcode solution. The following PCR protocol was carried out: 95 °C for 10 min followed by 15 cycles of 95 °C for 30 s, 60 °C for 30 s, 72 °C for 1 min and then 72 °C for 3 min. The barcoded PCR products were checked on 6 % acrylamide gels with a 100– 400 bp ladder (200 V constant for 40 min).

DNA sequencing was carried out using the Ion Torrent Personal Genome Machine (PGM) (Life Technologies). Following initialisation of the PGM, control ISPs and annealing sequencing primer were added to the enriched ISPs, cycled at 95 °C for 2 min and then 37 °C for 2 min, and then, PGM 200 Sequencing Polymerase was added to produce a total final volume of 30 μl. An Ion 316 Chip (Life Technologies) was washed with 100 μl of 100 % isopropanol and then twice with 100 μl of annealing buffer. The sample was then loaded onto the chip.

Virchows Arch

Data and statistical analysis

The concordance between NGS and MSRED is good

Sequencing data were aligned to hg19 using Torrent Suite software (Life Technologies) and visualised using IGV software (http://www.broadinstitute.org/igv/). CTNNB1 mutation hot spots were analysed manually by two independent observers in order to identify mutations. A mutation was deemed present when there was ≥5 % variant allele fraction (VAF) in a region with ≥500 reads per base. This threshold corresponds to the morphological inclusion criteria of 10 % lesional cells, equating to 5 % somatic VAF. The results were compared to MSRED data, where available. Non-concordant results were repeated with technical and biological repeats from DNA re-extracted from the original FFPE tissue; cases with mutations other than the three most common substitutions (previously detected by MSRED) underwent capillary sequencing at UCL Cancer Institute Core Facility. Categorical data were analysed using the chi-squared test (GraphPad Prism), and a p value of G p.T41A, c.134C>T p.S45F or c.133T>C p.S45P), and in eight cases, a second CTNNB1 mutation was detected by NGS in addition to the expected substitution (Table 5).

Results Patient and tumour demographics The patient and tumour demographics are shown in Table 2. The sporadic DTFs from 144 individuals studied had a demographic profile similar to other published literature [2, 18]: Approximately two thirds of patients were female, and approximately one sixth of tumours were recurrent. Those patients who experienced recurrence of their tumour had an earlier average age of presentation: 29.88 versus 39.61 years for non-recurrent tumours (pT p.S45F substitution alone and 3 cases with a c.133T>C p.S45P substitution alone (Table 4). There were 11 cases in which rarer, nonsynonymous mutations were identified using NGS that had not been detected using MSRED (Table 5): Nine of these cases harboured a second non-synonymous common mutation described above. The mutations in the two cases in which a common mutation was not detected were represented by c.122C>T p.T41I (COSMIC mutation ID COSM5676) and c.134_139delCTCTGA p.S45_S47>C (COSMIC mutation ID COSM28708), neither of which are reported in DTFs in COSMIC. There were eight cases in which mutations were detected using NGS that were not detected using MSRED: c.134C>T p.S45F (n=5), 121A>G p.T41A (n=1), c.122C>T p.T41I (n = 1) and c.134_139delCTCTGA p.S45_S47>C (n = 1). These results were reproducible in repeat, independent experiments using NGS, MSRED and capillary sequencing. All cases in which mutations were detected by MSRED were also detected by NGS. Sensitivity and specificity The sensitivity of NGS for detecting mutations in sporadic DTFs was 92.36 % (133/144, 95 % CIs: 86.74 to 96.12 %). The specificity of NGS was 100 %, since no CTNNB1 mutations were detected in the 11 morphological mimics (superficial fibromatosis, nodular fasciitis and reactive lesions) nor the four DTFs with FAP/Gardener’s syndrome (which are driven by APC mutations). Within this same cohort, the sensitivity of MSRED was 82.41 % (89/108, 95 % CIs: 73.90 to 89.06 %). Recurrent DTFs This study included 25 recurrent tumours in the same number of individuals, all of which had mutations detected by NGS: Ten cases had the p.T41A mutation and 15 cases had the

Virchows Arch Table 2 Patient demographics Number cases Age at first presentation

Gender Anatomical site

All cases

Recurrent tumours

Non-recurrent tumours

Mean SD Range Male

144 37.97 17.27 8 to 81 49

25 29.88 15.43 8 to 72 7

119 39.61 17.13 8 to 81 42

Female Leg Back Abdominal wall Arm Buttock Chest wall Shoulder Hip Intra-abdominal Head and neck Axilla Foot Breast Supraclavicular Hand Intra-thoracic Unknown

95 31 21 16 10 10 10 8 7 6 5 4 4 3 2 1 1 5

18 8 2 0 2 5 2 1 3 0 0 0 2 0 0 0 0 0

77 23 19 16 8 5 8 7 4 6 5 4 2 3 2 1 1 5

144

25

119

Total

p.S45F mutation; the p.S45P mutation was not detected in any recurrent tumours. Detection of a mutation (chi-squared test, p=0.18) or any specific mutation (p=0.32) was not associated with tumour recurrence.

Discussion Until now, we have used MSRED to detect CTNNB1 mutations in DTFs in clinical practice. In this study, we compared NGS results using a Fluidigm and Ion Torrent PGM-based NGS approach with those of our previous MSRED results. Table 3

Concordance between MSRED and NGS results NGS result

Total

Mutation No mutation detected detected MSRED result Mutation detected 89 No mutation detected 8 Not done 36 Total 133

0 11 0 11

89 19 36 144

CTNNB1 mutations were detected in 92.36 % (133/144) of all cases of sporadic DTF compared to only 82.41 % using MSRED. The concordance with MSRED was ‘good’ (92.59 %), and the average read depth (2354 reads) and confidence in mutation detection were high, since genetic alterations were detected in an average of 23.1 % reads per sample and at high read depth. These data demonstrate that the overall sensitivity of NGS is higher than the sensitivity of other techniques previously reported to detect these mutations in DTFs: Lazar et al. [20] reported a mutation rate of 85 % (117/138) by PCR amplification of CTNNB1 exon 3 codons 30–48, and Amary et al. showed that the sensitivity was 86.8 % using MSRED in a cohort of 76 DTFs [9]. However, since not all DTFs are caused by CTNNB1 mutations, this figure is likely to under-represent the true sensitivity of the test, since APC mutations account for up to 20 % of sporadic DTFs [21]. In addition to the possibility that mutation-negative cases harbour an APC mutation, the failure to detect a CTNNB1 mutation may also be explained by the fact that only a limited number of CTNNB1 loci were sequenced by the assay that we used in this study. The CTNNB1 hot spot regions sequenced included the common codon 15–22 and 34–45 exon 3 mutations: Nevertheless, these loci account for only 176/569 of all

Virchows Arch Table 4

Detailed NGS results NGS result

MSRED result

Mutation detected No mutation detected Not done

Total

Total

T41A only

S45F only

S45P only

Additional mutation

No mutation detected

48 1 19 68

30 5 16 51

3 0 0 3

8 2 1 11

0 11 0 11

reported CTNNB1 mutations in COSMIC, equating to 2843/4269 of all reported cases. In eight cases of sporadic DTF analysed in this study, a mutation was detected using NGS but not with MSRED. Five of these cases had a p.S45F mutation and one case a p.T41A mutation, both of which should have been detected by our routine MSRED assay. In contrast, the other two genetic alterations, one a small deletion c.134_139delCTCTGA and the other a substitution c.122C>T, p.T41I, have been reported in COSMIC but not in DTF, despite this mutation having being reported previously in DTF by Le Guellec et al. [4]. This highlights the value not only of sequencing the DNA of large numbers of tumour subtypes but also reporting the findings to the scientific community. Such reference materials or catalogues provide researchers and clinicians with greater confidence when reporting rare genetic variants, particularly when they occur in rare diseases. Discrepant results between NGS and MSRED may be accounted for by the lower sensitivity of MSRED, i.e. type II (false negative) error. Additional mutations detected by NGS but not MSRED can be accounted for by the fact that MSRED only detects very specific (targeted) nucleotides; any substitutions in the same codon or any other chromosomal region would not be detected. In addition, more complex mutations including indels or translocations cannot be detected by MSRED. NGS allows all mutations present within the targeted regions to be identified,

89 19 36 144

although any mutations falling out with these regions would still be missed unless a whole gene or exome approach was adopted. It is noteworthy that three of the 14 mutation-negative tumours were sited in the abdominal cavity, a common site for DTFs harbouring an APC mutation. Hence, in the absence of a history of FAP/Gardner’s syndrome, the possibility of an APC genetic alteration cannot be excluded without additional DNA sequencing. However, as we show that six other intraabdominal cases in our study harboured a CTNNB1 mutation, DFT at this site does not infer an APC genetic alteration. Therefore, we recommend that when a confident diagnosis of DTF is made on morphological features in the absence of a CTNNB1 mutation, the possibility of FAP/Gardner’s syndrome should be raised with the multidisciplinary team to ensure that appropriate measures are taken to exclude a hereditary diagnosis and appropriate referral to a medical genetics practice is made. The multidisciplinary team may also decide to instigate reflex APC mutation testing for all cases for which a confident diagnosis of DTF is made but a CTNNB1 mutation is not detected. One case in our study underscores this point: A patient presented with an intra-abdominal DTF (presumed to be sporadic) but no CTNNB1 mutation was detected. Subsequently, the patient’s 3-year-old daughter was diagnosed with an extra-abdominal CTNNB1 mutation-negative DTF. Although these tumours from mother and child had not been

Table 5 Less common mutations MSRED Negative Negative Positive Positive Positive Positive Positive Positive Positive Positive Not done

NGS No mutation detected No mutation detected Codon 41 ACC to GCC Codon 41 ACC to GCC Codon 41 ACC to GCC Codon 41 ACC to GCC Codon 41 ACC to GCC Codon 45 TCT to TTT Codon 45 TCT to TTT Codon 45 TCT to TTT

c.122C>T, p.T41I c.134_139delCTCTGA, p.S45_S47>C c.121A>G, p.T41A; c.61G>A, p.A21T c.121A>G, p.T41A; c.65T>C p.V22A c.121A>G, p.T41A; c.115G>A, p.A39T c.121A>G, p.T41A; c.119C>T, p.T40I c.121A>G, p.T41A; c.130C>T, p.P44S c.134C>T, p.S45F; c.107A>C, p.H36P c.134C>T, p.S45F; c.119C>T, p.T40I c.134C>T, p.S45F; c.121A>G, p.T41A c.121A>G, p.T41A; c.134C>T, p.S45F; c.95A>G, p.D32G; c.101G>A, p.G34E; c.109T>C, p.S37P

Virchows Arch

tested for APC mutations, there was a high degree of clinical suspicion for FAP/Gardner’s syndrome, and the mother was referred for genetic counselling. Neither the presence nor absence of a CTNNB1 mutation (p=0.18), nor the identification of any specific mutation in CTNNB1 (p=0.32), were associated with tumour recurrence in this study. Our findings are consistent with those of Mullen et al. [22], who found that there was no significant difference in recurrence of disease associated with either the CTNNB1 mutation status or the presence of a specific CTNNB1 mutation. This is in contrast to the results of Lazar et al. [20] who found that 5-year recurrence-free survival was significantly lower in DTFs with the c.134C>T p.S45F substitution in codon 45 (23 %, pG p.T41A in codon 41 (57 %) or tumours in which no CTNNB1 mutation was detected (65 %). This finding was supported by Colombo et al. [23] who found that DTFs with a p.S45F mutation had a greater tendency for local recurrence. It would be of interest to resolve these conflicting results, since confidence in an association of recurrence with a specific mutation may help to stratify patients for closer follow-up or even more aggressive surgery or radiotherapy. One possible confounding factor is that individuals in our study have a slightly older age of presentation (average 37.97 years) compared with the cohort of Lazar et al. [20] (average 32 years). This is noteworthy, as Lazar et al. established an association between risk of recurrence and younger age, which may explain the negative association between recurrence and the p.S45F mutation in our study. This difference in age of the cohort makes the results difficult to compare. Furthermore, conflicting results may reflect that the patients in the different groups received different treatments (more aggressive surgery, radiotherapy, hormone treatment or a combination of the above) when they first presented. Obtaining an answer to this question may require a prospective study in which patients with similar treatments are compared. In conclusion, we have shown that NGS is a robust technology for detecting CTNNB1 mutations in DTFs and has a sensitivity of 92.36 %, which could be further increased by screening a greater number of loci. In addition, NGS is more efficient than MSRED as it allows screening of multiple alterations in one assay and is as cost-effective as MSRED and other technologies.

References 1.

2. 3.

4.

5.

6.

7.

8.

9.

10. 11.

12.

13.

14. 15. 16.

Acknowledgments This work was supported by Skeletal Cancer Action Trust (SCAT), UK. The material was obtained from the RNOH Musculoskeletal Research Programme and Biobank. Support was provided to AMF (UCL) by the National Institute for Health Research, UCLH Biomedical Research Centre and the UCL Experimental Cancer Centre. SJA is a NIHR-funded Academic Clinical Fellow. We are grateful to the patients for participating in the research and to the clinicians and support staff of the London Sarcoma Service. Conflict of interest No conflicts of interest to declare.

17. 18.

19.

Fletcher CDM, Bridge JA, Hogendoorn P, Mertens F, World Health Organization Classification of Tumours (2013) Pathology and genetics of tumours of soft tissue and bone, 4th edn. WHO Press, Lyon Escobar C, Munker R, Thomas JO, Li BD, Burton GV (2012) Update on desmoid tumors. Ann Oncol 23(3):562–569 Reitamo JJ, Häyry P, Nykyri E, Saxén E (1982) The desmoid tumor. I. Incidence, sex-, age- and anatomical distribution in the Finnish population. Am J Clin Pathol 77(6):665–673 Le Guellec S, Soubeyran I, Rochaix P, Filleron T, Neuville A, Hostein I et al (2012) CTNNB1 mutation analysis is a useful tool for the diagnosis of desmoid tumors: a study of 260 desmoid tumors and 191 potential morphologic mimics. Mod Pathol 25(12):1551– 1558 Stoeckle E, Coindre JM, Longy M, Binh MBN, Kantor G, Kind M et al (2009) A critical analysis of treatment strategies in desmoid tumours: a review of a series of 106 cases. Eur J Surg Oncol 35(2): 129–134 Devata S, Chugh R (2013) Desmoid tumors: a comprehensive review of the evolving biology, unpredictable behavior, and myriad of management options. Hematol Oncol Clin North Am 27(5):989– 1005 Briand S, Barbier O, Biau D, Bertrand-Vasseur A, Larousserie F, Anract P et al (2014) Wait-and-see policy as a first-line management for extra-abdominal desmoid tumors. J Bone Joint Surg Am 96(8):631–638 Miyoshi Y, Iwao K, Nawa G, Yoshikawa H, Ochi T, Nakamura Y (1998) Frequent mutations in the beta-catenin gene in desmoid tumors from patients without familial adenomatous polyposis. Oncol Res 10(11–12):591–594 Amary MFC, Pauwels P, Meulemans E, Roemen GM, Islam L, Idowu B et al (2007) Detection of beta-catenin mutations in paraffin-embedded sporadic desmoid-type fibromatosis by mutation-specific restriction enzyme digestion (MSRED): an ancillary diagnostic tool. Am J Surg Pathol 31(9):1299–1309 Willert K, Jones KA (2006) Wnt signaling: is the party in the nucleus? Genes Dev 20(11):1394–1404 Lips DJ, Barker N, Clevers H, Hennipman A (2009) The role of APC and beta-catenin in the aetiology of aggressive fibromatosis (desmoid tumors). Eur J Surg Oncol 35(1):3–10 Yost SE, Smith EN, Schwab RB, Bao L, Jung H, Wang X et al (2012) Identification of high-confidence somatic mutations in whole genome sequence of formalin-fixed breast cancer specimens. Nucleic Acids Res 40(14):e107 Salto-Tellez M, de Castro DG. Next generation sequencing: a change of paradigm in molecular diagnostic validation. J Pathol. 2014. Desai AN, Jere A (2012) Next-generation sequencing: ready for the clinics? Clin Genet 81(6):503–510 Katsanis SH, Katsanis N (2013) Molecular genetic testing and the future of clinical genomics. Nat Rev Genet 14(6):415–426 Untergrasser A, Cutcutache I, Koressaar T, Ye J, Faircloth B, Remm M et al (2012) Primer3—new capabilities and interfaces. Nucleic Acids Res 40(15):115 Koressaar T, Remm M (2007) Enhancements and modifications of primer design program Primer3. Bioinformatics 23(10):1289–1291 Eastley N, Aujla R, Silk R, Richards CJ, McCulloch TA, Esler CP, et al. Extra-abdominal desmoid fibromatosis—a sarcoma unit review of practice, long term recurrence rates and survival. Eur J Surg Oncol. 2014. Sim J, Wright CC (2005) The kappa statistic in reliability studies: use, interpretation, and sample size requirements. Phys Ther 85(3): 257–268

Virchows Arch 20.

21.

Lazar AJF, Tuvin D, Hajibashi S, Habeeb S, Bolshakov S, Mayordomo-Aranda E et al (2008) Specific mutations in the betacatenin gene (CTNNB1) correlate with local recurrence in sporadic desmoid tumors. Am J Pathol 173(5):1518–1527 Tejpar S, Nollet F, Li C, Wunder JS, Michils G (1999) dal Cin P, et al. Predominance of beta-catenin mutations and beta-catenin dysregulation in sporadic aggressive fibromatosis (desmoid tumor). Oncogene 18(47):6615–6620

22.

23.

Mullen JT, DeLaney TF, Rosenberg AE, Le L, Iafrate AJ, Kobayashi W et al (2013) β-Catenin mutation status and outcomes in sporadic desmoid tumors. Oncologist 18(9):1043–1049 Colombo C, Miceli R, Lazar AJ, Perrone F, Pollock RE, Le Cesne A et al (2013) CTNNB1 45F mutation is a molecular prognosticator of increased postoperative primary desmoid tumor recurrence: an independent, multicenter validation study. Cancer 119(20):3696– 3702