Sequence artefacts in a prospective series of formalin-fixed tumours tested for mutations in hotspot regions by massively parallel sequencing
- Stephen Q Wong†1,
- Jason Li†2,
- Angela Y-C Tan1,
- Ravikiran Vedururu1,
- Jia-Min B Pang1, 3,
- Hongdo Do4,
- Jason Ellul2,
- Ken Doig2,
- Anthony Bell1,
- Grant A McArthur5, 6,
- Stephen B Fox1, 3, 5, 6,
- David M Thomas5, 6, 7,
- Andrew Fellowes1,
- John P Parisot5, 6,
- Alexander Dobrovic1, 3, 4, 5Email author and
- The CANCER 2015 Cohort
© Wong et al.; licensee BioMed Central Ltd. 2014
Received: 18 January 2014
Accepted: 30 April 2014
Published: 13 May 2014
Clinical specimens undergoing diagnostic molecular pathology testing are fixed in formalin due to the necessity for detailed morphological assessment. However, formalin fixation can cause major issues with molecular testing, as it causes DNA damage such as fragmentation and non-reproducible sequencing artefacts after PCR amplification. In the context of massively parallel sequencing (MPS), distinguishing true low frequency variants from sequencing artefacts remains challenging. The prevalence of formalin-induced DNA damage and its impact on molecular testing and clinical genomics remains poorly understood.
The Cancer 2015 study is a population-based cancer cohort used to assess the feasibility of mutational screening using MPS in cancer patients from Victoria, Australia. While blocks were formalin-fixed and paraffin-embedded in different anatomical pathology laboratories, they were centrally extracted for DNA utilising the same protocol, and run through the same MPS platform (Illumina TruSeq Amplicon Cancer Panel). The sequencing artefacts in the 1-10% and the 10-25% allele frequency ranges were assessed in 488 formalin-fixed tumours from the pilot phase of the Cancer 2015 cohort. All blocks were less than 2.5 years of age (mean 93 days).
Consistent with the signature of DNA damage due to formalin fixation, many formalin-fixed samples displayed disproportionate levels of C>T/G>A changes in the 1-10% allele frequency range. Artefacts were less apparent in the 10-25% allele frequency range. Significantly, changes were inversely correlated with coverage indicating high levels of sequencing artefacts were associated with samples with low amounts of available amplifiable template due to fragmentation. The degree of fragmentation and sequencing artefacts differed between blocks sourced from different anatomical pathology laboratories. In a limited validation of potentially actionable low frequency mutations, a NRAS G12D mutation in a melanoma was shown to be a false positive.
These findings indicate that DNA damage following formalin fixation remains a major challenge in laboratories working with MPS. Methodologies that assess, minimise or remove formalin-induced DNA damaged templates as part of MPS protocols will aid in the interpretation of genomic results leading to better patient outcomes.
Advances in genomic technologies are improving the capability to arrive at a more informed decision on how to treat a patient [1, 2]. In particular, a large number of genes can be screened for actionable changes using massively parallel sequencing (MPS) approaches [3, 4]. Accurate and reliable interpretation of this new type of genomic data is therefore critical in deciding the appropriate course of management for patients.
Most DNA for genetic testing from cancer patients is extracted from formalin-fixed paraffin-embedded tumour biopsies, where the primary intent is to preserve tumour cellular structure for histological examination and diagnosis . However, the formalin fixation process is detrimental to downstream genomic applications, causing issues, such as DNA cross-linking to DNA and proteins, that can stall polymerases and DNA-DNA crosslinks that can inhibit denaturation .
A common form of DNA damage induced by formalin fixation is fragmentation, which can lead to low amounts of amplifiable template for PCR amplification. Fragmentation of DNA is caused by a number of factors during the fixation process, e.g. low pH formalin over time increases the rate of apurinic/apyrimidinic site formation and eventually decomposition and fragmentation . Long-term storage of formalin-fixed blocks can also induce fragmentation due to exposure to environmental conditions [8–10].
Another prominent type of DNA damage that occurs commonly in formalin-fixed tissues is the hydrolytic deamination of cytosine to form uracil (or thymine if the cytosine is methylated). This results in non-reproducible C>T/G>A sequencing artefacts that are observed after PCR amplification when using formalin-fixed and paraffin-embedded (FFPE) DNA [11–13].
Recently, we assessed an amplicon-based MPS technology and showed that C>T and G>A changes were the most prominent sequence errors in three formalin-fixed lung squamous cell carcinoma samples . While such artefacts occur in many formalin-fixed samples, we have found this to be more pronounced in highly fragmented samples. Due to stochastic effects, the low template numbers increase the probability of occurrence of template artefacts .
Many studies have reported the feasibility of using DNA from formalin-fixed material using both conventional PCR-based and MPS technologies [4, 15, 16]. Fragmentation can be a rate-limiting factor in amplicon-based approaches especially those that use longer amplicons. Shorter amplicons permit fragmented DNA from formalin-fixed material to be used more successfully.
Some studies have reported elevated numbers of sequence artefacts [12, 16–19] whereas others reported little evidence of artefacts appearing in FFPE samples [15, 20]. The small number of samples assessed as well as the varying age of biopsies, degree of fixation and sequencing technologies used makes it difficult to know how important sequence artefacts are as a source of error in relatively fresh FFPE samples.
This study assessed the prevalence of DNA fragmentation and sequencing artefacts from a large cohort of FFPE tumours using a uniform approach whereby all blocks were of similar age, were extracted in the same manner and were run through the same MPS platform. The originating anatomical pathology laboratory was also recorded to determine if variation in operating practices could affect sequencing artefacts and DNA fragmentation.
Patients and cell lines
Cancer 2015 is a large-scale, prospective, longitudinal, multi-site cohort study of incident cancers in the Victorian population. The aim of the Cancer 2015 study is to classify cancers molecularly using MPS to promote more targeted treatment of cancer patients and improve patient survival and outcomes. An initial pilot phase was established to determine the feasibility of adopting MPS for the diagnostic mutational profiling of tumours. This study was approved by the Human Research Ethics Committees at the Peter MacCallum Cancer Centre, Royal Melbourne, Cabrini, Geelong and Warrnambool Hospitals, all located within the state of Victoria, Australia. All patients provided informed consent to participate in this study. Formalin-fixed, paraffin-embedded (FFPE) tumour blocks or unstained sections from FFPE tumour blocks were acquired from anatomical pathology laboratories performing the diagnosis and sent to the Peter MacCallum Cancer Centre Pathology department. Epidemiological and clinical variables were collected from each patient. As controls representing good quality DNA, the cell lines NCI-H1975 (H1975) and HL-60 were also sequenced.
All blocks/sections were received at the Peter MacCallum Cancer Centre Pathology department for DNA extraction. The age of the blocks/sections was determined by identifying the duration between the date of fixation (date of surgery) and the date of extraction (arrival date at the Peter MacCallum Cancer Centre). Up to ten unstained tumour sections of 5 microns thickness were cut from each block. DNA from FFPE sections and cell line samples were extracted using the DNeasy blood and tissue kit (Qiagen, Hilden, Germany) as per the manufacturer’s instructions. DNA quantification was performed using the Qubit dsDNA HS Assay kit for the Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA). The Qubit readings were used as a guideline for dilution of the DNA samples.
FTH1 Taqman assay
FFPE derived DNA was checked for quality and concentration using an 180 bp FTH1 TaqMan assay (Life Technologies). PCR was performed using the LightCycler 480 (Roche Diagnostics, Penzberg, Germany). The reaction mixture included 1X TaqMan Gene Expression Master Mix (Life Technologies), 1X FTH1 TaqMan Assay labeled with FAM (Life Technologies, # 4331182), 1 μL of DNA and PCR grade water in a total volume of 11 μL. PCR conditions included an activation step of 10 min at 95°C followed by 40 cycles of 95°C for 15 seconds and annealing for 60 sec at 60°C. Based on a standard curve generated from genomic DNA of known concentrations of 50, 25, 12.5, 6.25, 3.13, 1.56 , 0.78 and 0.39 ng/μL and copy numbers per μL of 16667, 8333, 4167, 2083, 1042, 521, 260, 130 respectively, the copy numbers per μL for each sample were calculated according to the LightCycler 480 software manufacturer’s instructions.
TruSeq amplicon cancer panel
The TruSeq Amplicon - Cancer Panel (TSACP) (Illumina, San Diego, CA) comprises 212 amplicons from 48 genes that are simultaneously amplified in a single-tube reaction. We used 5 μL of each DNA sample (50 ng/μl) for the experiment according to the manufacturer’s instructions. We used the MiSeq system (Illumina) for paired end sequencing with a v2 150-bp kit.
Uracil-DNA glycosylase treatment
For samples treated with uracil-DNA glycosylase (UDG), FFPE DNA was dispensed and reduced to a final volume of 2 μL by vacuum centrifugation. The DNA was incubated with 6.25 U (1 U/20 ng of DNA) UDG (New England Biolabs, Ipswich, MA) in a final volume of 20 μL containing 1 U UDG buffer. After an initial incubation at 37°C for 2 h, the UDG enzyme was inactivated at 95°C for 10 min. UDG-treated FFPE DNAs were stored at 4°C before use in the TSACP reactions. Before use in sequencing, the volume of the reaction was reduced to a final volume of 5 μL by vacuum centrifugation.
NRAS mutation testing
To confirm the NRAS negative result in sample Ca97, a second MPS method using deep sequencing was also used to assess exon 3 of the NRAS gene. Briefly, PCR was performed on a Fluidigm Access Array according to standard protocols [Access Array™ System for Illumina Sequencing Platform User Guide (PN 100–3770)] with the same FFPE sample (50 ng/μl) loaded into one well of the Access Array. The primers were 5’-acactgacgacatggttctacaGAGACAGGATCAGGTCAGCG-3’ and 5’-tacggtagcagagacttggtctGATGTGGCTCGCCAATTAAC-3’, giving an amplicon size of 253 bp (CS1 and CS2 tags are shown in lower case). After barcoding, the sample was run on an Illumina Miseq according to the manufacturer’s instructions (v2 150-bp kit).
CASAVA v1.8.2 was used to perform sample de-multiplexing and to convert BCL files generated from the MiSeq instrument into Fastq files containing short-read data. Using the primer sequences that are present in the data, short reads were first assigned to their respective amplicon. Global alignment based on the Needleman–Wunsch algorithm was then performed between the reads and the amplicon reference sequences to identify sequence variations. Likely true variants were identified by a) VarScan2 and b) a variant frequency of >10% and were not included in the analysis of sequencing artefacts. Variants with a frequency < 1% were assumed to be sequencing errors. The remaining base changes (i.e. all except true biological variants and sequencing errors) were counted and categorised into respective nucleotide groups using custom Python scripts. In addition to the 1-10% allele frequency range, the same pipeline was also used to determine artefacts in the 10-25% allele frequency range.
Spearman tests were used to examine associations between continuous variables. The Kruskal–Wallis one-way analysis of variance followed by the Dunn’s post hoc test was used to test for associations between continuous variables (C>T/G>A changes, FTH1 results) versus anatomical pathology laboratory and tumour type. All analyses were performed using GraphPad Prism software version 6.01.
Cell lines and FFPE samples
Summary of formalin-fixed samples
Number of cases
Head and neck
Bone and soft tissue
Central nervous system
Cancer of unknown primary
Assessment of fragmentation in FFPE samples
Overall total coverage of a sample after sequencing would be expected to be inversely associated with the amount of fragmentation whereby samples with high level of DNA fragmentation would have less amplification and therefore associated with lower coverage. By comparing both the FTH1 and coverage results for all FFPE samples, there was a weak but significant association between these two variables (r = 0.29, p < 0.0001, Spearman Rank correlation). Given a minimum of 200,000 total reads to achieve a mean of ~1000X across all amplicons in a sample, 55/488 samples would not achieved this level of coverage.
There were two outlier samples that displayed high amounts of amplifiable template as measured by the FTH1 assay but resulted in low coverage (samples Ca23 and Ca156). These samples however could have low coverage because of experimental error during the processing of the sample resulting in low coverage. However, for the overwhelming majority of other samples, there was a noticeable trend with the degree of fragmentation as measured by the FTH1 assay and the resulting coverage after sequencing. Importantly, this indicates that significant fragmentation is present even in relatively recently fixed biopsies and is an important factor leading to lower amounts of successful sequencing data.
Assessment of sequencing artefacts in FFPE samples
While fragmentation is a well known form of DNA damage that is present in a large number of FFPE samples, sequencing artefacts as a result of PCR amplification have not been extensively studied, especially in relatively recent samples as would be used diagnostically.
Hence, our next analysis was to determine if sequence artefacts were also present in our cohort of recently fixed samples. We previously published an informatic pipeline to assess the level of artefacts in archival lung squamous cell carcinoma samples . This pipeline identifies counts of nucleotide changes within the 1-10% frequency range and removes counts from likely true mutations that are identified through the Varscan2 caller. Using this pipeline, we measured the degree of artefacts between the 1-10% range for all 488 FFPE samples and cell line samples.
In contrast, the FFPE samples showed a large range of C>T/G>A changes in a large proportion of samples (Figure 2). The increase in C>T/G>A changes was not discrete and not confined to only a subset of samples but appeared to be continuous suggesting that this type of artefact is ubiquitous in nature and will occur to some extent in every FFPE sample. The relative proportion of C>T/G>A changes was highly significant compared to other nucleotide changes (p < 0.0001, Kruskal–Wallis one-way analysis of variance) and formed 32% of all nucleotide changes overall, a proportion that was double what was expected due to chance.
By contrast, the same analysis, applied to look at possible artefacts between the 10-25% allele frequency range, indicated that C>T/G>A changes appear at much lower levels compared to the 1-10% allele frequency range (the scale on the y axis of Figure 2B is much smaller than that of 2A) and most likely represent real variants that were not be removed using our algorithm. Importantly, there was no observable bias towards C>T/G>A changes (Figure 2B).
Assuming that cell line DNA represents good quality DNA that would display little or no artefacts, the median C>T/G>A change count in the 1-10% allele frequency range was 272 contrasting to a much more higher median C>T/G>A count in FFPE samples of 1515. These findings not only confirm our original findings that C>T/G>A are the most predominant type of sequencing artefact but also indicate that they occur in all FFPE samples to some degree.
Correlation of sequence artefacts with coverage
An interesting observation in samples with high C>T/G>A changes in the 1-10% range was a parallel increase in the number of other nucleotide changes. This could be due to the stochastic nature of low template samples which increase the probability for not only C>T/G>A artefacts but also other nucleotide artefacts to appear.
Correlation of anatomical laboratory with sequence artefacts and coverage
While DNA was extracted at the same location with the same protocol, the tissue blocks were formalin fixed and paraffin embedded in different anatomical pathology laboratories across Victoria, Australia. Since different anatomical laboratories might have different practices regarding the fixation of tissues, we examined DNA damage in relation to the originating anatomical laboratory. There was a significant difference for both fragmentation and sequence artefacts between anatomical pathology (AP) laboratories (p < 0.0001, Kruskal-Wallis test). Dunn’s Multiple Comparison test indicated significant difference between AP7 versus AP22 and AP26 for fragmentation analysis, and AP7 versus AP22 for sequencing artefacts (Additional file 2: Figure S1). Despite standardization, these results suggest that fixation processes may vary across different labs. The age of the blocks or the tumour type was not significantly correlated with the degree of fragmentation or sequence artefacts.
Case studies: samples with potential false positive clinical mutation calls
In the case of a melanoma with an apparent NRAS c.35 G>A, p.G12D mutation (Ca97), resequencing (after UDG treatment) using the TSACP platform did not confirm the same mutation. To further validate this result, the sample was tested by an orthogonal MPS method (Fluidigm Access Array microfluidic chip system) that covers the NRAS exon 2 region at extremely high coverage (Additional file 2: Figure S2). This method also could not detect the mutation suggesting that the original call for the mutation was a false positive.
While the implementation of MPS into a clinical setting is currently in progress in numerous laboratories world-wide, there are still a number of challenges using DNA from FFPE cancer specimens for this application. The prevalence of DNA damage associated with formalin fixation in clinical material has not been extensively studied, particular in terms of cancer genomics. Most genomic studies dealing with FFPE DNA are relatively small in numbers with the archival status of specimens in most cases remaining unknown [4, 16]. In contrast, this study had a relatively narrow window in which samples were formalin-fixed and had a standardised and centralised point of DNA extraction and sequencing. FFPE samples in this study therefore accurately represent those that would enter a molecular diagnostic clinical workflow.
To our knowledge, this study represents the largest assessment of formalin induced fragmentation and sequencing artefacts using clinical FFPE samples. From our set of analyses, we have confirmed that both fragmentation and sequencing artefacts are common forms of DNA damage from formalin-fixed material even if blocks are relatively new. As a result, this work illustrates these forms of DNA damage can have major consequences for downstream interpretation as not all actionable mutations could be validated.
There are a limited number of publications highlighting the ability of amplicon-based MPS to identify clinically relevant mutations or canonical mutations from DNA obtained from various tumour types [23–26]. However, these studies looked at a relatively small number of cases, usually reported only canonical mutations, and where validation of variants was performed, demonstrated variation in the results obtained. For example, in the study investigating the sensitivity of immunohistochemistry (IHC) versus Sanger sequencing to BRAF V600E mutations in FFPE derived DNA from papillary thyroid carcinoma specimens, Bullock and colleagues utilised the TSACP as an alternative sequencing platform . The TSACP was able to identify the BRAF V600E mutation in three out of 11 IHC positive cases with variant frequencies of 10-32%, which Sanger sequencing failed to detect. When DNA was then macro-dissected from the same tumour blocks, both sequencing and the TSACP analyses could now both detect the V600E mutation in seven cases (7/11). This confirms that the platform is comparatively more sensitive than the “gold standard” of Sanger sequencing for low frequency mutations.
In this study we explored two common form of DNA damage caused by formalin, i.e. fragmentation and sequencing artefacts. Assessment of fragmentation was performed using a copy number assessment PCR prior to MPS. The weak association between total coverage counts and the FTH1 results could be explained by the limitations of the FTH1 assay. While the assay can clearly discriminate between samples based on the degree of fragmentation, the assay is designed only around a product of 180 bp and for only one locus. Consequently, the FTH1 assay cannot accurately assess the entire range of product sizes generated by the TSACP protocol. Testing multiple loci across various fragment lengths is therefore recommended for FFPE samples in MPS testing to predict poor quality amplifiable DNA .
Interestingly, the degree of C>T/G>A artefacts forms a continuum suggesting that the source of DNA damage causing these changes occurs to some degree in every sample. Procedurally, although the practice of fixation of tumour specimens is standardised, the observation that samples processed from one anatomical pathology laboratory (AP7) had a significantly higher rate of sequencing artefacts and fragmentation indicates that fixation processes may not actually be uniform between different laboratories.
Assessment of other variables potentially involved in DNA damage such as the duration of transport from patient to the anatomical pathology laboratory, and size of tumour specimens could not be feasibly assessed in this study. In addition, the degree and spectrum of DNA damage from a formalin-fixed sample depends on environmental factors such as exposure to heat, light and the concentration and age of formalin used for fixation. The age of blocks examined is another consideration as long-term storage in suboptimal environments can cause significant DNA damage. Further dissection of these factors in a more controlled setting will provide significant benefits to the preservation of DNA for subsequent testing.
Another consideration is the type of polymerase used in this study. While the type of polymerase used in the TSACP was not disclosed due to commercial reasons it is probably not proofreading since proofreading enzymes are known to stall at uracil lesions and in effect, replicate the effect of UDG treatment. Further examination of polymerases which increase correct sequence amplification but still maintain sufficient amplification for sequencing are required.
There is a possibility that some changes within the 1-10% allele frequency do represent low-level heterogeneity and were not detected by the Varscan2 variant caller and removed from our analysis of sequencing artefacts. Fresh frozen samples counterparts would be suitable for this assessment but were not available for this study to confirm this.
In this study, we confirmed the detection of two low frequency actionable mutations in the PIK3CA gene, in breast carcinoma cases. These mutations remained after the sequencing was repeated (and after UDG treatment). Given the clinical context of these mutations in breast cancer, these mutations are of potential clinical benefit to the patient that may have implications for PIK3CA inhibitors that target the PI3K/AKT/mTOR pathway [21, 22]. It also illustrates that this platform can detect real mutations even at low frequencies.
In contrast, an activating NRAS G12D mutation discovered originally in our first screen but not confirmed in the same DNA specimen by subsequent UDG treatment or MPS has major implications for the patient. Activating mutations in NRAS have been reported in approximately 20% of all melanomas  and are potentially sensitive to therapeutics that target downstream signaling through mitogen-activated protein kinase kinase and phosphatidylinositol 3-OH kinase or AKT [29, 30]. This is a clear demonstration of the dangers that can arise from MPS data without proper validation.
Complete validation of all variants was not possible in this study because of the extremely large number of variants detected and the lack of remaining DNA available for all samples. Undoubtedly a large number of these variants will be false positives, especially in the 1-10% allele frequency range. While validation of every variant is a laborious task costing time, material and expense, we recommend that validation of actionable mutations that will directly affect patient management be performed in this kind of testing. We also propose that quality control measures be adopted prior to sequencing including the implementation of more efficient and accurate quality control assays that evaluate DNA concentration, fragmentation and the presence of uracil lesions.
While we have shown treatment with UDG is effective in reducing artefactual variants, there are a number of other strategies to minimise artefacts occurring, particularly those using non-capture based-technologies . This includes performing duplicate reactions for the same sample or through internal validation by having overlapping amplicons covering the same loci (as is the case in the newer TruSight™ version of the TSACP panel that was released after this study was concluded).
The Cancer 2015 study is supported by the Victorian Government through the Victorian Cancer Agency Translational Research Program. SW is supported by the Melbourne Melanoma Project funded by the Victorian Cancer Agency Translational Research Program and established through support of the Victor Smorgon Charitable Fund. We would like to thank all the cancer patients who participated in the study. We gratefully acknowledge the cooperation of the following Victorian institutions: The Andrew Love Cancer Centre; Geelong Hospital, Barwon Health; Warrnambool Hospital; Southwest Health; The Peter MacCallum Cancer Centre; Ludwig Institute for Cancer Research, Austin Health; Royal Melbourne Hospital, Melbourne Health; Centre for Health Economics, Monash University; Department of Epidemiology and Preventative Medicine, The Alfred Centre, Monash University; Cabrini Health; Department of Pathology, University of Melbourne and Monash Institute of Medical Research. We thank Mark Lucas for assistance with the clinical database for Cancer 2015 patients.
- MacConaill LE, Campbell CD, Kehoe SM, Bass AJ, Hatton C, Niu L, Davis M, Yao K, Hanna M, Mondal C, Luongo L, Emery CM, Baker AC, Philips J, Goff DJ, Fiorentino M, Rubin MA, Polyak K, Chan J, Wang Y, Fletcher JA, Santagata S, Corso G, Roviello F, Shivdasani R, Kieran MW, Ligon KL, Stiles CD, Hahn WC, Meyerson ML, et al: Profiling critical cancer gene mutations in clinical tumor samples. PloS one. 2009, 4: e7887-10.1371/journal.pone.0007887.View ArticlePubMedPubMed CentralGoogle Scholar
- Macconaill LE, Garraway LA: Clinical implications of the cancer genome. J Clin Oncol. 2010, 28 (35): 5219-5228. 10.1200/JCO.2009.27.4944.View ArticlePubMedPubMed CentralGoogle Scholar
- Mar VJ, Wong SQ, Li J, Scolyer RA, McLean C, Papenfuss AT, Tothill RW, Kakavand H, Mann GJ, Thompson JF, Behren A, Cebon JS, Wolfe R, Kelly JW, Dobrovic A, McArthur GA: BRAF/NRAS wild-type melanomas have a high mutation load correlating with histologic and molecular signatures of UV damage. Clin Cancer Res. 2013, 19: 4589-4598. 10.1158/1078-0432.CCR-13-0398.View ArticlePubMedGoogle Scholar
- Wagle N, Berger MF, Davis MJ, Blumenstiel B, Defelice M, Pochanard P, Ducar M, Van Hummelen P, Macconaill LE, Hahn WC, Meyerson M, Gabriel SB, Garraway LA: High-throughput detection of actionable genomic alterations in clinical tumor samples by targeted, massively parallel sequencing. Cancer discovery. 2012, 2 (1): 82-93. 10.1158/2159-8290.CD-11-0184.View ArticlePubMedGoogle Scholar
- Berg D, Malinowsky K, Reischauer B, Wolff C, Becker KF: Use of formalin-fixed and paraffin-embedded tissues for diagnosis and therapy in routine clinical settings. Methods Mol Biol. 2011, 785: 109-122. 10.1007/978-1-61779-286-1_8.View ArticlePubMedGoogle Scholar
- Gilbert MT, Haselkorn T, Bunce M, Sanchez JJ, Lucas SB, Jewell LD, Van Marck E, Worobey M: The isolation of nucleic acids from fixed, paraffin-embedded tissues-which methods are useful when?. PLoS One. 2007, 2 (6): e537-10.1371/journal.pone.0000537.View ArticlePubMedPubMed CentralGoogle Scholar
- Sikorsky JA, Primerano DA, Fenger TW, Denvir J: DNA damage reduces Taq DNA polymerase fidelity and PCR amplification efficiency. Biochem Biophys Res Commun. 2007, 355 (2): 431-437. 10.1016/j.bbrc.2007.01.169.View ArticlePubMedPubMed CentralGoogle Scholar
- Bruskov VI, Malakhova LV, Masalimov ZK, Chernikov AV: Heat-induced formation of reactive oxygen species and 8-oxoguanine, a biomarker of damage to DNA. Nucleic Acids Res. 2002, 30 (6): 1354-1363. 10.1093/nar/30.6.1354.View ArticlePubMedPubMed CentralGoogle Scholar
- Pfeifer GP, You YH, Besaratinia A: Mutations induced by ultraviolet light. Mutat Res. 2005, 571 (1–2): 19-31.View ArticlePubMedGoogle Scholar
- Lindahl T: Instability and decay of the primary structure of DNA. Nature. 1993, 362 (6422): 709-715. 10.1038/362709a0.View ArticlePubMedGoogle Scholar
- Do H, Dobrovic A: Dramatic reduction of sequence artefacts from DNA isolated from formalin-fixed cancer biopsies by treatment with uracil- DNA glycosylase. Oncotarget. 2012, 3 (5): 546-558.View ArticlePubMedPubMed CentralGoogle Scholar
- Do H, Wong SQ, Li J, Dobrovic A: Reducing sequence artifacts in amplicon-based massively parallel sequencing of formalin-fixed paraffin-embedded DNA by enzymatic depletion of uracil-containing templates. Clin Chem. 2013, 59 (9): 1376-1383. 10.1373/clinchem.2012.202390.View ArticlePubMedGoogle Scholar
- Hofreiter M, Jaenicke V, Serre D, von Haeseler A, Paabo S: DNA sequences from multiple amplifications reveal artifacts induced by cytosine deamination in ancient DNA. Nucleic Acids Res. 2001, 29 (23): 4793-4799. 10.1093/nar/29.23.4793.View ArticlePubMedPubMed CentralGoogle Scholar
- Do H, Dobrovic A: Limited copy number-high resolution melting (LCN-HRM) enables the detection and identification by sequencing of low level mutations in cancer biopsies. Mol Canc. 2009, 8: 82-10.1186/1476-4598-8-82.View ArticleGoogle Scholar
- Kerick M, Isau M, Timmermann B, Sultmann H, Herwig R, Krobitsch S, Schaefer G, Verdorfer I, Bartsch G, Klocker H, Lehrach H, Schweiger MR: Targeted high throughput sequencing in clinical cancer settings: formaldehyde fixed-paraffin embedded (FFPE) tumor tissues, input amount and tumor heterogeneity. BMC Med Genomics. 2011, 4: 68-10.1186/1755-8794-4-68.View ArticlePubMedPubMed CentralGoogle Scholar
- Yost SE, Smith EN, Schwab RB, Bao L, Jung H, Wang X, Voest E, Pierce JP, Messer K, Parker BA, Harismendy O, Frazer KA: Identification of high-confidence somatic mutations in whole genome sequence of formalin-fixed breast cancer specimens. Nucleic Acids Res. 2012, 40 (14): e107-10.1093/nar/gks299.View ArticlePubMedPubMed CentralGoogle Scholar
- Schmitt MW, Kennedy SR, Salk JJ, Fox EJ, Hiatt JB, Loeb LA: Detection of ultra-rare mutations by next-generation sequencing. Proc Natl Acad Sci U S A. 2012, 109 (36): 14508-14513. 10.1073/pnas.1208715109.View ArticlePubMedPubMed CentralGoogle Scholar
- Spencer DH, Sehn JK, Abel HJ, Watson MA, Pfeifer JD, Duncavage EJ: Comparison of clinical targeted next-generation sequence data from formalin-fixed and fresh-frozen tissue specimens. J Mol Diagn. 2013, 15 (5): 623-633. 10.1016/j.jmoldx.2013.05.004.View ArticlePubMedPubMed CentralGoogle Scholar
- Anaka M, Hudson C, Lo PH, Do H, Caballero OL, Davis ID, Dobrovic A, Cebon J, Behren A: Intratumoral genetic heterogeneity in metastatic melanoma is accompanied by variation in malignant behaviors. BMC Med Genomics. 2013, 6: 40-10.1186/1755-8794-6-40.View ArticlePubMedPubMed CentralGoogle Scholar
- Schweiger MR, Kerick M, Timmermann B, Albrecht MW, Borodina T, Parkhomchuk D, Zatloukal K, Lehrach H: Genome-wide massively parallel sequencing of formaldehyde fixed-paraffin embedded (FFPE) tumor tissues for copy-number- and mutation-analysis. PLoS One. 2009, 4 (5): e5548-10.1371/journal.pone.0005548.View ArticlePubMedPubMed CentralGoogle Scholar
- Cizkova M, Susini A, Vacher S, Cizeron-Clairac G, Andrieu C, Driouch K, Fourme E, Lidereau R, Bieche I: PIK3CA mutation impact on survival in breast cancer patients and in ERalpha, PR and ERBB2-based subgroups. Breast Cancer Res. 2012, 14 (1): R28-10.1186/bcr3113.View ArticlePubMedPubMed CentralGoogle Scholar
- Janku F, Wheler JJ, Westin SN, Moulder SL, Naing A, Tsimberidou AM, Fu S, Falchook GS, Hong DS, Garrido-Laguna I, Luthra R, Lee JJ, Lu KH, Kurzrock R: PI3K/AKT/mTOR inhibitors in patients with breast and gynecologic malignancies harboring PIK3CA mutations. J Clin Oncol. 2012, 30 (8): 777-782. 10.1200/JCO.2011.36.1196.View ArticlePubMedPubMed CentralGoogle Scholar
- Bullock M, O'Neill C, Chou A, Clarkson A, Dodds T, Toon C, Sywak M, Sidhu SB, Delbridge LW, Robinson BG, Learoyd DL, Capper D, von Deimling A, Clifton-Bligh RJ, Gill AJ: Utilization of a MAB for BRAF(V600E) detection in papillary thyroid carcinoma. Endocr Relat Cancer. 2012, 19: 779-784. 10.1530/ERC-12-0239.View ArticlePubMedGoogle Scholar
- Diaz-Padilla I, Hirte H, Oza AM, Clarke BA, Cohen B, Reedjik M, Zhang T, Kamel-Reid S, Ivy SP, Hotte SJ, Razak AA, Chen EX, Brana I, Wizemann M, Wang L, Siu LL, Bedard PL: A phase Ib combination study of RO4929097, a gamma-secretase inhibitor, and temsirolimus in patients with advanced solid tumors. Invest New Drugs. 2013, 31 (5): 1182-1191. 10.1007/s10637-013-0001-5.View ArticlePubMedPubMed CentralGoogle Scholar
- Hubers AJ, Heideman DA, Yatabe Y, Wood MD, Tull J, Taron M, Molina MA, Mayo C, Bertran-Alamillo J, Herder GJ, Koning R, Sie D: EGFR mutation analysis in sputum of lung cancer patients: a multitechnique study. Lung Cancer. 2013, 82 (1): 38-43. 10.1016/j.lungcan.2013.07.011.View ArticlePubMedGoogle Scholar
- Pemovska T, Kontro M, Yadav B, Edgren H, Eldfors S, Szwajda A, Almusa H, Bespalov MM, Ellonen P, Elonen E, Gjertsen BT, Karjalainen R, Kulesskiy E, Lagstrom S, Lehto A, Lepisto M, Lundan T, Majumder MM, Lopez Marti JM, Mattila P, Murumagi A, Mustjoki S, Palva A, Parsons A, Pirttinen T, Ramet ME, Suvela M, Turunen L, Vastrik I, Wolf M, et al: Individualized Systems Medicine (ISM) strategy to tailor treatments for patients with chemorefractory acute myeloid leukemia. Canc Discov. 2013, 3 (12): 1416-4129. 10.1158/2159-8290.CD-13-0350.View ArticleGoogle Scholar
- Sah S, Chen L, Houghton J, Kemppainen J, Marko AC, Zeigler R, Latham GJ: Functional DNA quantification guides accurate next-generation sequencing mutation detection in formalin-fixed, paraffin-embedded tumor biopsies. Genome Med. 2013, 5 (8): 77-10.1186/gm481.View ArticlePubMedPubMed CentralGoogle Scholar
- Devitt B, Liu W, Salemi R, Wolfe R, Kelly J, Tzen CY, Dobrovic A, McArthur G: Clinical outcome and pathological features associated with NRAS mutation in cutaneous melanoma. Pigment Cell Melanoma Res. 2011, 24 (4): 666-672. 10.1111/j.1755-148X.2011.00873.x.View ArticlePubMedGoogle Scholar
- Ascierto PA, Schadendorf D, Berking C, Agarwala SS, van Herpen CM, Queirolo P, Blank CU, Hauschild A, Beck JT, St-Pierre A, Niazi F, Wandel S, Peters M, Zubel A, Dummer R: MEK162 for patients with advanced melanoma harbouring NRAS or Val600 BRAF mutations: a non-randomised, open-label phase 2 study. Lancet Oncol. 2013, 14: 249-256. 10.1016/S1470-2045(13)70024-X.View ArticlePubMedGoogle Scholar
- Kelleher FC, McArthur GA: Targeting NRAS in melanoma. Cancer J. 2012, 18 (2): 132-136. 10.1097/PPO.0b013e31824ba4df.View ArticlePubMedGoogle Scholar
- Wong SQ, Li J, Salemi R, Sheppard KE, Hongdo D, Tothill RW, McArthur GA, Dobrovic A: Targeted-capture massively-parallel sequencing enables robust detection of clinically informative mutations from formalin-fixed tumours. Sci Rep. 2013, 3: 3494.PubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1755-8794/7/23/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.