A novel, minimally invasive diagnostic test for KIT exon 11 internal tandem duplications in canine cutaneous mast cell tumours I: Assay development

A novel, minimally invasive diagnostic test for KIT exon 11 internal tandem duplications in canine cutaneous mast cell tumours I: Assay development | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A novel, minimally invasive diagnostic test for KIT exon 11 internal tandem duplications in canine cutaneous mast cell tumours I: Assay development Jacob O. Evans, Liam S. Hill, Sam Beck, Grace Edmunds, Melanie Dobromylskyj, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9368633/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background In canine cutaneous mast cell tumours (cMCTs), Internal Tandem Duplications (ITDs) of exon 11 of the KIT receptor tyrosine kinase are associated with higher histological grade and greater risk of metastasis. They also predict response to tyrosine kinase inhibitors (TKIs). Assessment of KIT mutation status provides important information when identifying candidate patients for TKI therapy. Current KIT ITD assays based on semi-quantitative PCR and electrophoretic separation of PCR products lack sensitivity and are not easily comparable between samples. They are limited to analysis of biopsy material from tumours and cannot be used for monitoring of residual disease during routine clinical follow-up, for instance using fine needle aspirates or blood tests. Here, we describe a novel, sensitive, quantitative qPCR assay based on melting curve analysis with potential for use in a wide range of samples during initial clinical assessment/staging and also routine follow up of cMCT patients. Results We identified a minimal region (‘min’ region) of KIT exon 11 commonly amplified in cMCT and designed qPCR primers against that region to produce a small product (44bp) from a wild-type (WT) target. However, an ITD including the min region duplicates primer annealing sites, meaning that additional, larger (>90bp) PCR products are generated. The presence of an ITD-derived product can be distinguished from the WT product by examining peaks on the qPCR melt curves (the WT product produces one peak, the ITD product produces a second distinct peak). The assay was tested on fifteen FFPE cMCT samples with known KIT exon 11 status (five WT, ten with an ITD) and on MDCK normal canine epithelial cell DNA. Considering each individual replicate carried out on all samples, the analytical specificity of the assay was 94% and the sensitivity was 100%. Conclusions We have developed a novel, rapid, qPCR-based assay for the presence of KIT exon 11 ITDs in canine cutaneous Mast Cell Tumours. The assay uses PCR-product melt curve analysis to detect the duplications with high sensitivity and specificity. The assay has the potential for use in a wide range of cMCT samples, including non-surgical samples such as blood or fine needle aspirates. Small Animal Medicine Oncology Canine cutaneous mast cell tumour KIT exon 11 Internal tandem duplication qPCR Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Background Mast cell tumours (MCTs) are one of the most common tumours of dogs, with an age-standardised rate of 126/100,000 dogs/year in a survey of a population of insured dogs in the UK, considering both the cutaneous and subcutaneous forms ( 1 ). The tumour can also occur less commonly in the intestine, visceral organs or other sites ( 2 ). The cutaneous form, cMCT, is among the most common cutaneous tumours of dogs, accounting for 7–21% of all canine skin tumours ( 3 – 9 ). The mutation status of the receptor tyrosine kinase KIT is an important parameter to assess within mast cell tumours ( 10 ). Signalling by KIT, and cross-talk between KIT and IgE receptors, is required for proliferation, differentiation and activation of normal mast cells as part of the allergic inflammatory response ( 11 – 14 ) and deregulated activation of KIT is thought to be a key driver of mast cell neoplasia ( 9 ). Mutations in the KIT gene in canine MCTs have been reported in a number of studies, including in the extracellular (exons 8 and 9) and juxta-membrane (exons 11 and 12) domains ( 9 , 15 – 21 ). These include point mutations, deletions and internal tandem duplications (ITDs) of exon 11. The presence of the latter is reported in between 9 and 45% of canine MCTs, depending on the study ( 15 , 16 , 18 , 21 – 23 ). These mutations result in constitutive activation of KIT by autophosphorylation in the absence of its ligand ( 10 ). In cMCTs, KIT exon 11 ITDs are associated with shorter time to progression and increased likelihood of recurrence or development of metastasis ( 15 , 24 ) and with decreased overall survival ( 24 – 26 ). It is likely that mutated KIT is both directly influential towards cell behaviour, and that it correlates with high histologic grade driven by multiple other factors. Therefore, the presence of a KIT exon 11 ITD has not been shown to be an independent prognostic marker ( 10 ), nevertheless, as the presence or absence of a mutation is more objective than assessment of grade or proliferation, mutation status can still provide useful information for clinical decision making ( 10 ). KIT mutation status may be most useful for prognostication in histologically low grade or ‘ambiguous’ tumours ( 10 ). Importantly, KIT exon 11 mutation status is also valuable in predicting response to TKIs. cMCTs with KIT mutations have a better objective response to toceranib ( 27 ), but potentially worse progression-free survival ( 28 ), while overall outcomes, times to progression and overall response rates to masitinib are better in cMCTs with KIT mutations ( 10 , 24 ). Current approaches to exon 11 ITD testing rely on a standard semi-quantitative PCR approach, followed by analysis of PCR product size by electrophoretic separation ( 29 ). This approach uses material from tissue biopsy, which may not available if a tumour is not suitable for surgical sampling. In those tumours which cannot be removed, incisional biopsy may be carried out to obtain tissue for determining KIT mutation status, but this approach is associated with the risk of incisional complications ( 30 ) or systemic inflammatory responses associated with histamine release ( 31 , 32 ). Furthermore, the lack of sensitivity of the standard approach makes currently available KIT assays unsuitable for long-term monitoring of minimal residual disease burden by liquid biopsy of circulating tumour DNA (ctDNA). Therefore, a new approach which defines KIT mutation status without surgical biopsy is needed. Here, we describe the development and implementation of a novel, inexpensive and rapid SYBR green quantitative real-time PCR (qPCR)-based assay for detection of KIT exon 11 ITDs in canine cMCTs with high sensitivity and specificity (> 90%). The assay has the potential for use in a wide range of samples, fulfilling the need for a new approach to testing for KIT exon 11 ITDs. Methods Aims Development of a novel sensitive and specific quantitative real-time PCR-based test for detection of KIT exon 11 internal duplication in canine mast cell tumours, suitable for use in a variety of sample types. Ethical approval This study was approved by the Bristol University/Bristol Veterinary School Animal Welfare and Ethics Review Body (VIN number 22-041). Supply of samples by Finn Pathologists (Harleston, Norfolk; arranged through MD) was approved by the CVS ethical review panel. Sequences A reference sequence for Canis lupus familiaris KIT proto-oncogene (NM_001003181) based on UU_Cfam_GSD_1.0/canFam4 (chr13:47,940,533-47,940,659; amino acids 559 - 590) was downloaded from the UCSC genome browser into the sequence editing and annotation package of DNASTAR Lasergene version 17.5.0.48 (DNASTAR Inc., Madison, Wisconsin USA) and annotated. Single nucleotide polymorphisms (SNPs) in the region of interest were noted at chr13:47940619 (C/T) and chr13:47940672 (T/C) (33). KIT ITD sequences for the C2 canine mastocytoma cell line and the BK, EH, MK and MS canine mast cell tumours from London et al . Figure 3 (18) were manually copied into DNASTAR Lasergene. Multiple sequence alignment was carried out in using DNASTAR Lasergene MegAlign Pro with Clustal Omega using default settings. A consensus reference sequence for Canis lupus familiaris BRCA1 (NM_001013416) was built from the canFam1, 2 and 3 genome builds, downloaded from the UCSC genome browser into DNASTAR Lasergene. The earlier builds were used as they formed the basis of protein splice isoform predictions. To clarify the intron/exon structure and annotation of C. familiaris BRCA1, we aligned the nucleotide sequences of eight C. lupus familiaris BRCA1 protein splice isoforms predicted from build canFam3 (1: XP_013971930.1; 2: XP_013971932.1; 3: XP_005624371.1; 4: XP_022278220.1; 5: XP_022278221.1; 6: XP_022278222.1; 7: XP_013971933.1; 8: XP_013971934.1; downloaded from https://www.ncbi.nlm.nih.gov/protein/) to this consensus sequence . Primers Primers were designed using Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). See Additional File 1 . Primers were ordered from Merck-Sigma-Aldrich (Gillingham, Dorset, UK) as ‘Pure and Simple Primers’ (https://www.sigmaaldrich.com/GB/en/configurators/tube?product= puresimple), cartridge purified and lyophilised. Primers were resuspended in nuclease-free water as 100uM stocks and then diluted to 10uM working aliquots before use. Synthetic Internal Tandem Duplication sequences Type I and Type II ITD synthetic oligos were ordered from Merck-Sigma-Aldrich as ‘Long Oligos’ (https://www.sigmaaldrich.com/GB/en/configurators/tube?product=longoligo). 179 base pair sequences for both the sense and R+C antisense strands were ordered (5 OD yield and cartridge purification) based on the C2 and MS sequences ( Additional File 1 ). The MS Type II anti-sense strand could not be synthesised and so the Type II synthetic oligo had to be added to reactions single stranded. Both the C2 Type I sense and antisense sequence oligos could be synthesised, and these were annealed into a double-stranded target. The lyophilised sense and anti-sense Type I sequences were each resuspended at 100uM in TE (10mM Tris base, 1 mM EDTA, pH8.0) then 20ul of each oligo was added to 60ul of STE buffer (50mM NaCl, 10mM Tris base, 1 mM EDTA, pH8.0) and the mixture placed in a hot block at 100 o C for 5 minutes, after which the block was allowed to cool overnight. The final concentration of the annealed oligo was taken as 20uM. Clinical samples FFPE blocks of fifteen canine mast cell tumours with known KIT exon 11 status (ten positive for an exon 11 ITD, five negative for an ITD) were provided by Finn Pathologists through MD. Exon 11 status was determined by the Michigan State University veterinary diagnostic laboratory (Lansing, Michigan, USA). All FFPE samples were archival tissue which had been previously used for routine diagnosis as part of standard clinical veterinary care. All were available for research use. See Additional File 2 for sample details. Histological stains Haematoxylin and eosin (H&E) stains were carried out on dewaxed and rehydrated FFPE sections using standard protocols (reagents from Atom Scientific, Hyde, Cheshire, UK). Slides were dipped in xylene and then mounted with glass coverslips using DPX (Atom Scientific). Once dry, slides were scanned using an Olympus VS200 whole slide scanner (Evident Europe GmbH, Hamburg, Germany) at 20x. The relative amount of tumour compared to normal cells within each sample was approximated by measuring the total area of tissue and of tumour on the H&E sections at low power using QuPath 0.5.1 (https://qupath.github.io/). For caveats to this approach, see Discussion . DNA extraction from FFPE sections DNA was extracted from 5 um FFPE sections either as cut scrolls in Eppendorf tubes or sections on slides (two to three sections for each extraction). Sections were dewaxed by washing twice in xylene for 5 minutes each and then twice in 100% ethanol for 5 minutes each. If isolating from scrolls, samples were centrifuged between washes to avoid loss of material. Sections on slides were simply transferred between washes. After the last ethanol wash, samples were air-dried. DNA was then extracted using the QIAamp DNA FFPE tissue kit (Qiagen, Manchester, UK), including the RNaseA step. DNA was eluted in 100ul nuclease-free water and quantified on a Nanodrop (Thermofisher, Horsham, Surrey, UK). PCR conditions All PCR reactions were prepared in a designated pre-PCR area in a Class II laminar flow hood, in a room separate from areas where template was prepared. Template was not handled until the prepared PCR reactions were taken out of the pre-PCR room. Template was added to reactions in a designated PCR laminar flow hood in a separate area. Standard semi-quantitative PCR reactions were carried out using the GoTaq Green PCR system (Promega, Chilworth, Hampshire, UK). Reactions volumes were 50ul (10ul of 5X Promega EasyGreen Buffer, 2mM MgCl 2 , 0.2uM of each primer, 1ul of 10 mM PCR nucleotide mix, 0.2ul Promega GoTaq) (see also Additional File 1 ). gDNA template input was typically 50ng (unless otherwise stated). A ‘No Template’ control was included in all experiments. The reactions were cycled on an Applied Biosystems VeritiPro 96-well Thermal Cycler (Thermofisher) with the following program: 94 o C 2 minutes; 35 cycles of 94 o C 15 seconds, 58 o C 30 seconds, 68 o C 30 seconds; hold at 4 o C (see also Additional File 1 ). Amplified products were resolved on 2% agarose TBE (Tris base 90mM, Boric acid 90mM, EDTA 2mM) gel (all reagents from Thermofisher) with a 100bp DNA ladder (Promega). Quantitative real time PCR on genomic DNA was carried out using a Quantstudio 7 system (Thermofisher) using Fast-96 well settings with SYBR Green reagents to determine Ct times and melt curves. Reaction volumes were 20ul consisting of 10ul 2x PowerUp SYBR Green Master Mix (Thermofisher), 1ul of each of 10uM forward and reverse primers, and the remaining 8ul either template, or template plus nuclease-free water (depending on the input material) or nuclease-free water alone for No Template controls. The amount of starting template input into each reaction varied depending on the experiment. Reactions were run as three or four replicates in 96-well plates in fast cycling mode with the following program (as detailed in the manufacturer’s instructions): hold stage: 50 o C 2 minutes, 95 o C 2 minutes; amplification stage: 40 cycles of 95 o C 1 second, 60 o c 30 seconds (ramping speed was 2.63 o C/s up, 2.42 o C down); melt curve stage, 95 o C 15 seconds, 60 o C 1 minute, ramping of 0.05 o C/s to 95 o C, hold at 95 o C 15 seconds; hold at 4 o C (see also Additional File 1 ). This program took 60 minutes to run. Automatic thresholding of Ct values was used. In some cases, after the program had finished, the contents of representative wells were resolved on 2% agarose TBE gels to confirm product sizes. In most cases, Ct values for the amplification curves (the amplification cycle at which the amount of fluorescence, and hence product, crossed the detection threshold; dependent on the amount of starting template) and Tm values for the melt curves (the temperature at which the rate of change of fluorescence as the product of the reaction melted was at its maximum; dependent on the physico-chemical properties of the product at the end of the amplification) were determined automatically by the Quantstudio 7 software. In some cases where the melting curve had two distinct peaks, the Tm for both peaks could be determined automatically. However, in some cases where one peak was small, the Tm for that peak had to be determined by manual inspection of the fluorescence reporter derivative values to identify the temperature with the highest derivative reporter value (i.e. the highest rate of change). For all melt curve plots, the X-axis is the temperature ( o C) and the Y-axis is the fluorescence reporter derivative. The Y-axis may be plotted on different scales between graphs where assays have been run on different plates. Except where otherwise stated, the fluorescence reporter derivative values across the melt curve temperature range were imported into GraphPad Prism 10.2.3 (GraphPad Software LLC, Boston MA, USA) for plotting. For standard curves of Ct values against DNA input using genomic MDCK DNA, the DNA content of an average canine diploid cell was estimated as ~5pg, based on the haploid length of the canine genome as ~2.4 Gb (34) and 978 Mb of DNA as having a mass of 1pg (35). Therefore, template input at 0.5pg, 5pg, 50pg, 500pg, 5ng or 50ng equated to approximately 0.2, 2, 20, 200, 2000 or 20,000 copies of the target per reaction, respectively. For standard curves of Ct values against DNA input using synthetic oligos, template was added to reactions at 3.3ul of 1fM (approx. 2000 molecules), 10fM (approx. 20,000 molecules), 100fM (approx. 200,000 molecules), 1pM (approx. 2,000,000 molecules) or 10pM (approx. 20,000,000 molecules) per reaction. Sequencing KIT Exon 11 genomic DNA sequencing of cMCT clinical samples was carried out by PCR amplification of the whole of exon 11 from FFPE-extracted DNA samples using the whole exon 11 primers ( Additional File 1 ) and GoTaq PCR conditions. Products were analysed by gel electrophoresis on 2% agarose TBE gels. For wild-type tumours without an exon 11 ITD, only a single product was amplified and the PCR reactions could be directly purified using a PCR cleanup kit (Qiagen) and then sent for routine Sanger sequencing at Eurofins (Eurofins GmbH, Ebersburg, Germany) using the whole exon 11 primers. For tumours with an exon 11 ITD, multiple bands were generated, corresponding to both a wild-type and an ITD product. These products were separated by gel electrophoresis, the bands cut out and then extracted using a QIAquick gel extraction kit (Qiagen). Extracted bands were then sent for routine sanger sequencing at Eurofins. Where possible, both the genotype of the wild-type and ITD bands at the chr13:47940619 (C/T) and chr13:47940672 (T/C) SNPs, as well as the overall structure of the ITD in each sample, were determined. Technical difficulties of sequencing across repetitive DNA sequences, resulting in sequence outputs containing misaligned sequences ( Additional File 3A ), meant the quality of sequence from ITD amplicons was variable, with regions where two different sequence traces were overlaid on each other. However, from the inspection of the sequencing chromatograms, and assessment of where the good quality sequence became poor quality sequence, it was possible to reconstruct the boundaries of the ITD regions, and therefore indirectly determine the sequence of the ITDs in most samples (see Additional Files 3 and 4 for a worked-through example of how this was carried out). Statistics All statistical analysis (curve fitting and linear regressions) was carried out using GraphPad Prism 10.2.3. Results Identification of the minimal commonly amplified region in mast cell tumours To develop our assay, we first aligned the duplicated regions from four MCTs and one mastocytoma cell line (C2) positive for KIT exon 11 ITDs published by London et al . (18) against the sequence of wild type KIT exon 11, to identify whether there was a minimal region commonly duplicated across multiple tumours. This identified such a minimal region (UU_Cfam_GSD_1.0/canFam4 chr13:47,940,605 – 47,940,644) close to the 3’ splice donor site in exon 11, found in all five ITDs ( Figure 1A ). We called this the min region. The min region is flanked by variable regions which are included, to different extents, in the different ITDs ( Figure 1B ). We then designed a pair of primers, FminA and RminA, targeting the min region and the 3’ variable region. The primers were designed to be head to head, with no intervening sequence (see Additional File 1 for details of all primer sets). The FminA site was wholly contained within the min region while the RminA site was partly contained in the min region, and partly in the 3’ variable region ( Figure 1B ). In a wild type, non-duplicated exon 11, the primer pair was predicted to create a small, 44bp, product. However, an internal tandem duplication would duplicate the FminA primer site and potentially the RminA site as well, resulting in the generation of new products of larger sizes. We hypothesised that the larger products would be distinguishable from the 44bp product in SYBR green-based qPCR melting curve analysis, and this would indicate the presence of an ITD. Importantly, we predicted that as long as the core min region was included in a tandem duplication, then using this strategy the presence of any duplication in any tumour could be detected by the min primers. The presence of variable flanking regions, the exact sequence of those regions, and the exact structure of the ITD as a whole, would not affect performance of the assay. Modelling this using the sequences from London et al ., we predicted that the ITDs in these tumours would result in minA ITD products with sizes of 92 bp (in the C2 line and MK tumour), 95 bp (in the BK tumour), 101 bp (in the MS tumour) and 112 bp (in the EH tumour) ( Additional File 5 ). We classified those sequences in which only the forward minA site was duplicated as ‘Type I’ duplications (C2 and MK) and those in which both forward and reverse minA sites were duplicated as ‘Type II’ duplications (BK, EH and MS). As an alternative approach, we designed a second set of primers, FminB and RminB, which flanked the minA primers ( Figure 1B ). In the wild type, these were predicted to generate an 88bp product, but ITDs were predicated to increase the distance between the forward and reverse priming sites, rather than duplicating them. Again, this should alter the melt curves in SYBR green-based qPCR. The predicted outcomes for the sequences from London et al . with the minB primers were products of 136 bp (in the C2 line and MK tumour), 139 bp (in the BK tumour), 145 bp (in the MS tumour) and 156 bp (in the EH tumour) ( Additional File 5 ). As we were initially unsure how the minA primers would perform, whether or not all samples would in fact generate the expected 44bp KIT exon 11 product and what factors might affect this, we also designed a pair of control primers (designated ‘Brca’ primers). These targeted the 5’ end of the large central exon of BRCA1 , which codes for the DNA binding site ( Additional File 6 ) (36). Like the minA primers, the Brca primers were a head-to-head primer set predicted to generate a 44bp product against which the performance of the minA primers could be compared. We also obtained previously described primers (37) for standard PCR amplification of the whole of KIT exon 11 for use in electrophoretic sized-based analysis to use as a control for the detection of ITDs in clinical samples and for exon 11 genomic sequencing ( Figure 1B ). The minA, minB and Brca primer sets amplify discrete PCR products of expected sizes To test the minA, minB and Brca primer sets, we first carried out standard semi-quantitative PCR with gel-based analysis of amplified fragments, as well as qPCR, on genomic DNA (gDNA) from the MDCK canine epithelial cell line. With standard PCR and an input of 50ng gDNA template per reaction, faint bands could be observed on a 2% agarose gel at approximately the expected sizes ( Figure 2A ). Purification of the amplified fragments and their use as template in a second round of standard PCR reactions increased the product yield. Product sizes were consistent with those seen after amplification directly from the gDNA template and also consistent with the predicted sizes of 44bp for minA and Brca, and 88bp for minB ( Figure 2B ). Next, we tested qPCR with an input of 50ng gDNA per reaction. All three primers amplified products with a single discrete peak in melting curve analysis, indicative of a single product being produced. The mean peak melting temperatures (mean Tm) for the minA, minB and Brca curves were 73.04±0.06 o C, 78.04±0.06 o C and 71.86±0.06 o C (mean±SD, n=3) respectively, consistent with the minA and Brca products being of similar size and smaller than the minB product ( Figure 2C-G ). Finally, reaction mixes from representative wells were resolved on an agarose gel. The sizes of the products were consistent with the sizes seen in a standard PCR reaction ( Figure 2H ). Therefore, using gDNA from a non-transformed canine epithelial cell line, the minA, minB and Brca primer sets generate discrete PCR products whose sizes are consistent with their predicted values. KIT exon 11 ITDs can be distinguished in qPCR by melt curve profiles Next, the performance of the minA, minB and Brca primer sets was tested on synthetic KIT exon 11 ITDs using two oligonucleotide targets based on the C2 cell line ITD (Type I amplification) and MS tumour sequence ITD (Type II amplification) ( Additional File 5 ). The primer sets were tested on these synthetic targets, as well as MDCK DNA, in both standard gel-based analysis and qPCR. 50ng of MDCK gDNA and 3.3ul of 10pM synthetic oligos were used as template for each standard PCR reaction. 5ng of MDCK gDNA and 3.3ul of 1pM synthetic oligos were used as template for each qPCR reaction. In standard PCR, with the MDCK template, the minA, minB and Brca primers gave products of sizes similar to those previously observed, consistent with the predicted 44bp, 88bp and 44bp sizes, respectively ( Figure 3A ). With the Type I synthetic oligo template, the minA primers generated two products, one identical in size to the MDCK minA product, the other running slightly below the 100 bp size marker, consistent with the predicted 92bp product ( Figure 3A ). The Type II template also generated one band identical in size to the MDCK minA product, consistent with the predicted 44bp product ( Figure 3A ). In addition, it generated a ‘ladder’ of larger products, the smallest of which was greater than 100bp, again consistent with the predicted size of the ITD but suggesting a cycle of self-annealing and amplification events by the PCR products themselves, resulting in the generation of new products with multiple copies of the duplication events (see Discussion ). With the minB primers and the Type I and II templates, bands consistent in size with the predicted 136bp and 145bp products, respectively, were observed ( Figure 3A ). As expected, no product was observed with the Brca primers with either the Type I or Type II synthetic oligo template, while a band consistent in size with the predicted 44bp product was observed with MDCK template. No product was observed with any ‘No Template’ reaction ( Figure 3A ). Therefore, using a standard gel-based PCR approach, results from all primer sets were consistent with predictions apart from the product laddering observed with minA primers, which was unexpected. With the qPCR analysis ( Figure 3B-I ), for the minA primers there was a mean Ct value for the MDCK template reactions of 28.33±0.26 (mean±SD; n=3 replicate wells) ( Figure 3B , H ). As previously, a single peak was observed for the MDCK melting curve in each of the replicates, with a mean Tm of 72.67±0.06 o C (mean±SD; n=3 replicate wells) ( Figure 3E,H ). qPCR of the Type I synthetic ITD using the minA primers resulted in a robust amplification with a mean Ct of 27.18±0.03 (mean±SD; n=3 replicate wells) ( Figure 3B,H ) and the production of two clear peaks on the melting curve. The first peak, with a Tm of 72.22±0.00 o C (mean±SD; n=3 replicate wells), was consistent with the peak seen in the MDCK reactions (hereafter the ‘wild type’ or WT peak) but there was a second peak with a Tm of 78.00±0.06 o C (mean±SD; n=3 replicate wells), consistent with a larger product (hereafter the ITD peak) ( Figure 3E,H ). With the Type II synthetic ITD and minA primers, the amplification had a mean Ct of 17.82±0.14 (mean±SD; n=3 replicate wells) ( Figure 3B,H ), suggesting the amplification reaction on the Type II ITD was much more robust than on wild type DNA or Type I ITD. Only a single melt curve peak was generated, with a Tm of 82.39±0.06 o C (mean±SD; n=3 replicate wells) ( Figure 3E,H ), substantially higher than the WT peak. Therefore, in qPCR on a WT target, minA primers resulted in a single melting curve peak with a Tm in the region 72 – 73 o C, but presence of an ITD resulted in a peak with a Tm >77 o C, as well as potentially a WT peak, depending on the exact nature of the ITD. With the minB primers, a single melt curve peak was generated with MDCK template, the Type I ITD and Type II ITD, but the Tm shifted from 77.92±0.10 o C with MDCK to 79.86±0.06 o C with Type I ITD and 80.48±0.06 o C with Type II ITD (mean±SD; n=3 replicate wells), consistent with the expected larger product sizes ( Figure 3C,F,H ). Again, as expected, no product was observed with the Brca primers in qPCR with either the Type I or Type II synthetic oligo template but the MDCK template gave a product with a single melting curve peak, with Ct and Tm values similar to that for the minA primers ( Figure 3D,G,H ), consistent with the 44bp Brca product size. Finally, to confirm the products were of the expected sizes, 5ul of the reaction mix from one representative well of each reaction with template were analysed by gel electrophoresis after the qPCR data had been collected. Three wells from the minA/No Template reactions were also analysed (see below). The product sizes ( Figure 3I ) were consistent with those seen after standard PCR and gel electrophoresis, although ‘laddering’ in the minA reactions was now visible in the Type I template, and with the Type II template the laddering was so extreme it had become a smear on the gel. In one minA/No Template qPCR reaction, a very low abundance amplification event had occurred, with a product reaching the Ct threshold after 38.73 cycles. This product also had a single peak on the melting curve at a slightly lower Tm to the MDCK reactions of 71.83 o C ( Figure 3E,H ). No such events were observed in the minB- or Brca/No Template reactions. Consistent with this, a small product could be seen on gel analysis of this No Template reaction ( Figure 3I ; boxed region). We reasoned that this was a background amplification event as a result of formation of ‘primer dimers’, low-affinity primer self-annealing events occurring in the absence of the correct high affinity target sequence (for example, ACAAA at positions 14 – 18 of the FminA primer could anneal to TGTTT at positions 8 – 12 of the FminB primer), followed by amplification to create a small product. To provide further evidence that primer self-annealing could occur in No Template controls, and that this was not a result of contamination, a qPCR plate was set up with multiple No Template controls. Twenty-four No Template wells were sealed in the pre-PCR area before MDCK template was added as a positive control. Five No Template wells were sealed after the MDCK template was added to the plate. The results show ( Additional File 7 ) that even in reactions sealed prior to any reasonable possibility of contamination, occasional low abundance amplification events, with Tm values for the melting curve of approximately 1 o C lower than that of the expected peak of the WT exon 11 minA product, could occur, consistent with low affinity self-priming by the minA primers. Importantly, however, these events cannot be confused with detection of an exon 11 ITD and do not, therefore, give a false positive result. Therefore, the minA and minB primer sets are capable of distinguishing KIT exon 11 WT from KIT exon 11 ITD sequences on the basis of the characteristics of melt curve profiles (number and Tm of peaks) in SYBR Green qPCR. Standard curves demonstrate linear amplification of WT and Type II ITD templates but not Type I templates Next, to better characterise the performance of the minA, minB and Brca PCR reactions on WT, Type I ITD and Type II ITD templates, a series of standard curve experiments were carried out ( Figure 4 and Additional File 8 ). The results showed that for the WT target, amplification with all three primer sets was linear when the input was 50pg (approximately 20 target copies) per reaction or more, but below this input level amplification was non-linear and sometimes failed ( Figure 4A-C ). Notably, the slope of the standard curves was similar, showing that the efficiency of the amplification between the three different primer sets was similar. The melt curves for the MDCK WT target all showed a single peak at a melting temperature consistent with those previously observed ( Additional File 8 ). For the Type I ITD, the lower limit of detection with the minA primer set was an input of 3.3ul of 10fM synthetic Type I target but for minB was 3.3ul of 100fM target, suggesting that the minB reaction is less sensitive than the minA. Type I ITD amplification with both minA and minB was non-linear, but better fitted a hyperbolic curve ( Figure 4D ), however, given that with minB only the three highest target concentrations gave amplification, the minB data have to be interpreted with caution. For the Type II ITD, amplification with both the minA and minB primers was linear ( Figure 4E ), and the template could be detected with input levels as low as 3.3ul of 1fM synthetic template for both primer sets. Note that based purely on the number of targets calculated to be in each of the reactions, the assays appeared to be more sensitive on WT ‘native’ genomic DNA than on synthetic ITD oligos. The melt curves for the Type I ITD with both minA and minB, and for Type II with minB, were qualitatively and quantitatively highly similar at all levels of template input as long as this was above the limits of detection ( Figure 4F ) and similar to those seen previously. However, with the minA primers using the Type II ITD synthetic oligo template, the melt curve patterns changed depending on the amount of template input into the reaction ( Figure 4F ). At 3.3ul 1fM template per reaction, there was a low peak at the Tm associated with the WT product (consistent with that seen in an MDCK control reaction), but also a larger, apparently double, peak with a higher Tm, in the range associated with the ITD product. As the amount of input increased, the peak associated with the WT product disappeared, and the double peak associated with the ITD product merged into a single, clean peak, as previously observed for the Type II ITD (compare Figure 4F with Figure 3E ), suggested altered reaction kinetics with different primer:template ratios. The min assay detects KIT exon11 ITDs in FFPE samples To test the assay in clinical samples, canine MCT FFPE samples which had been previously used for clinical diagnosis were supplied as FFPE blocks from the archives of Finn Pathologists (UK). For the purposes of this study, the samples from Finn Pathologists were coded as FP1 – 15. FP1 – 15 had been previously tested for KIT exon 11 ITDs as part of routine diagnosis. FP4, FP7, FP9, FP10 and FP11 (n=5) were all reported as KIT exon 11 WT, while FP1, FP2, FP3, FP5, FP6, FP8, FP12, FP13, FP14 and FP15 (n=10) were all positive for KIT exon 11 ITDs. Sections from the fifteen samples were first H&E stained to confirm the presence of tumour material ( Additional File 9A,B ). The relative proportion of tumour to non-tumour material in each sample was approximated by measuring the total area of tissue in each section and the approximate area of tumour in each section ( Additional File 9C ; see Discussion for limitations). To confirm exon 11 status, genomic DNA was isolated from the FP samples and standard PCR using the whole exon 11 primers was carried out, with the products visualised on a gel. Samples were analysed twice, to assess reproducibility ( Additional File 10 ). A single band of the expected size (~190bp) was generated for the five exon 11 WT samples in the first repeat ( Additional File 10A ), as expected. In the second repeat of the FP sample analysis, FP11 failed to amplify but the others again showed a single band ( Additional File 10B ). In the first repeat, the known exon 11 ITD samples all generated one band at the WT size and one or more larger bands, in the range 220 to 300 bp ( Additional File 10A ), consistent with the previously observed results for synthetic exon 11 ITDs, although the FP12 ITD band was faint. In the second repeat, the results were similar, except the FP1 and FP13 products were faint, and the ITD product could not be distinguished in FP12 ( Additional File 10B ). These results are broadly consistent with what was known of the FP samples a priori , but do show an inherent variability in the results of a standard PCR / gel analysis approach to assessing KIT exon 11 status, at least using the whole exon 11 primer set and PCR conditions described here. DNA extracted from the blocks was then tested with the minA/minB/Brca qPCR assays (details of Ct and Tm values for these assays are provided in Additional File 11 ). The Brca qPCR assay gave a single peak in the melting curve analysis in the expected Tm range (71.47to 71.94 o C) for all samples ( Additional File 12 ). Ct values varied from 25.44±0.13 (FP3) to 34.59±0.60 (FP11) despite nominally equal template input of 25ng/reaction. In the minA assay, Ct values ranged from 23.71±0.15 (FP2) to 37.45±0.29 (FP11). WT samples gave a single melt curve peak between 72.04 and 73.34± o C, comparable to the single peak seen with MDCK gDNA ( Figure 5A,B ). In contrast, samples with a known KIT exon 11 ITD showed both a WT peak and an ITD peak between 77.57 and 79.77 o C ( Figure 5C ). To quantify the strength of the ITD signal, the ratio of the maximum value of the ITD peak (‘ITD peak height’) to the maximum value of the WT peak (‘WT peak height’) was calculated for each sample. This ranged from 1.72±0.02 (mean±SD; n=3) in FP2 (strongest ITD signal) to 0.11±0.01 (mean±SD; n=3) in FP6 (weakest ITD signal). With the minB qPCR assay, Ct values ranged from 24.42±0.06 (FP2) to 30.60±0.16 (FP11). WT samples again gave a single peak comparable to MDCK gDNA (Tm 77.10 to 78.04 o C), ( Figure 6A,B ). With the ITD samples, FP5, FP6, FP13, FP14, showed two peaks to the melt-curve, one corresponding to the WT peak, the other peaking at 2 – 3 o C higher. FP1, FP2, FP3, FP8, FP15 did not show a distinct second peak but did have a ‘shoulder’ to the melt curve ( Figure 6C). minB ITD peaks were not distinct enough from WT peaks to enable relative peak heights to be reliably quantified. Sequencing and genotyping of KIT exon 11 from FP samples The FminA primer annealing site contains a C/T polymorphism ( C.fam 13: 47,940,619 C/T; hereafter, SNP 619), as does the RminB site ( C.fam 13: 47,940,672 C/T; hereafter, SNP 672). It was possible that variation here could affect the performance of the assay (strength of the ITD peak signal) across samples depending on their genotype. To test this, we first sequenced and genotyped KIT exon 11 from the FP samples by Sanger sequencing of fragments generated by whole exon 11 PCR. For the ITD samples, amplicons corresponding to WT and ITD exons were isolated by gel purification and sequenced separately. For all WT samples (both WT amplification products from WT tumours and WT amplification products from tumours with ITDs), the sequence was identical to the reference genomic sequence except for the variations at the SNP 619 and SNP 672 polymorphisms. A consensus sequence for all WT fragments and the genotypes at SNP 619 and SNP 672 are shown in Additional File 13 (FASTA sequences in Additional File 14 ) . Note the sequence of the WT band represents the germline genotype. The structure of the duplication could be reconstructed from the sequencing data (see Materials and Methods ) in nine of the ten ITD tumour samples ( Additional File 15 ). Three tumours were Type I, in which only the minA forward primer region had been duplicated. Five tumours were Type II, in which both the minA forward and reverse primer binding sites were duplicated. In one tumour (FP14) the ITD was a new variant in which the 5’ limit of the duplication region occurred within the forward minA primer site and resulted in its disruption, but at the same time created a new FminA priming site in a 5’ location, as well as duplicating the RminA primer site. Thus, similar product sizes would be generated to the Type I duplication, but from a different configuration. This ITD type was termed Type III. ITD genotypes, and for comparison the germline genotype (genotype of the WT band), are shown in Additional File 15B . It was possible to directly genotype nine of the ten ITD samples at SNP 619 and four of the samples at SNP 672. For one sample, FP12, which was germline homozygous C/C at SNP 619, it could be assumed that the ITD was ‘C’ as well. For three samples in which the WT SNP 672 genotype was C/C, it could be assumed the SNP 672 genotype of the ITD was C as well. In two samples (FP2 and FP12), which were C/T at WT SNP 672, the genotype of the ITD could not be determined (due to sequence quality. FP6 gave ambiguous results at SNP 672, with sequencing from the forward primer suggesting C/A at SNP 672, but the reverse primer sequencing suggesting C (see limitations in Discussion ). Overall, seven out of ten samples with KIT exon 11 ITDs were C/C homozygous at SNP 619 in the WT band, and C in the ITD. Two samples (FP5 and FP8) were T/T homozygous in the WT bands and correspondingly T in the ITD bands. FP6, which was C/T heterozygous at SNP 619 in the WT, underwent Loss-of-Heterozygousity (LOH) to give a T allele in the ITD ( Additional File 16 ). Similarly, at SNP 672, FP1 underwent LOH in the ITD compared to the WT sequence ( Additional File 16 ). Single Nucleotide Polymorphisms in the FminA and RminB primer annealing sites affect minA/minB assay signal strength The FminA primer has a C at the position equivalent to SNP 619. Therefore, any tumour which is T at this position in the ITD, such as FP6, may have suboptimal detection of the ITD by the minA assay as the FminA primer may bind to its target site with reduced affinity. A similar effect may happen with the RminB primer in a tumour with a SNP 672 ‘C’ allele, as the RminB primer has A at the position equivalent to SNP 672. To test whether optimising primer sequence to the genotype of the ITD would improve the assay, the performance of the FminA primer (CTTCCTTA C GATCACAAATGG) was compared to an alternative primer (FminA-ALT) with T at the equivalent of the SNP 619 position (CTTCCTTA T GATCACAAATGG). The two primers were tested on SNP 619 C/T germline heterozygous or T/T germline homozygous WT (FP4, FP9, FP10) or ITD (FP5, FP6, FP8) tumours, as well in MDCK cells (25ng input per reaction). The results ( Figure 7A ) demonstrated that detection of the ITD in FP5, FP6 and FP8, all of which carry a SNP 619 T allele ITD, was substantially improved with the minA-ALT primers (mean±SD ITD/WT peak ratios, n = 3: FP5 minA: 0.543±0.044, minA-ALT: 1.659±0.065; FP6 minA: 0.193±0.049, minA-ALT: 2.980±0.068; FP8 minA: 0.267±0.021, minA-ALT: 1.458±0.050; Additional File 11 ). Qualitatively, there was little difference between the melt curves for minA and minA-ALT in tumours without an ITD. It was also noted that in this assay, the minA assay on FP9 produced a small peak at a Tm similar to that of an ITD in two of three replicates. Given that all previous assays and a priori testing on FP9 had shown that this tumour was WT, and that there was no indication of an ITD peak with the minA-ALT primers (despite being a T/T genotype and therefore predicted to give a stronger ITD signal with minA-ALT), these small peaks must represent false positive ITD peaks (see further discussion on analytical sensitivity and specificity of the assay below in the Discussion ). Next, a direct comparison was carried out between the minB assay using the original primers, in which the reverse primer sequence corresponds to the SNP 672 T allele (CCCCT A TTTCATACTGACCAAAG) and the minB assay using a reverse primer (RminB-ALT) in which the sequence corresponded to the SNP 672 C allele (CCCCT G TTTCATACTGACCAAAG). One WT tumour and six ITD tumours, as well as MDCK DNA, were tested. With the WT tumour and MDCK DNA, there was little difference in the single WT peak with either the minB or minB-ALT primer set. With the ITD tumours, there was also little difference in strength of detection of the ITD peak, irrespective of primer set. However, in one tumour which was germline C/T but C in the ITD (FP1), the RminB-ALT primer gave a slightly stronger detection. In FP2, which was germline C/T but the ITD could not be genotyped, detection was much stronger with RminB-ALT, consistent with the ITD being C at SNP 672 ( Additional File 17 ). Therefore, use of primers sets corresponding to the genotype of a KIT exon 11 ITD substantially improves ITD detection, in particular with the minA/minA-ALT assay. The minA/minA-ALT assay for KIT exon 11 ITDs is quantitative To determine whether there was a correlation between the strength of the ITD signal and the amount of tumour material in each sample, the ITD peak/WT peak ratios ( Additional File 11 ) were correlated with the ratio of tumour to total tissue in each sample, defined from the H&E slides ( Additional File 9 ). First, a simple linear regression analysis was carried out using the ITD peak/WT peak ratios derived from the minA analysis of tumours with known KIT exon 11 duplications in Figure 5 . This showed no correlation between the strength of the ITD signal and amount of tumour in the sample (R 2 =0.002; P=0.898; Figure 7B ). Next, for the three tumours (FP5, FP6 and FP8) with a SNP 619 T genotype ITD and for which the minA-ALT assay substantially enhanced performance of the assay ( Figure 7A ), the ITD peak/WT peak ratios for the minA-ALT assay shown in Figure 7A were substituted for the minA ratios in the analysis. The correlation between ITD peak/WT peak ratio and amount of tumour in the starting material was now substantially improved although still did not reach significance (R 2 =0.316; P=0.091; Figure 7C ). Inspection of the linear regression results suggested that FP12 was an outlier with a relatively low ITD/WT peak ratio (0.30±0.05) considering the tumour:total tissue ratio calculated for this sample (0.719; second-highest of all ITD samples). To test this, the analysis was repeated, this time using the minA-ALT assay results for FP5, FP6 and FP8, and excluding FP12. This now demonstrated a significant correlation between the ITD/WT peak ratio signal and the amount of tumour material (R 2 =0.727; P=0.004; Figure 7D ). Therefore, providing the primer set corresponding to genotype of the ITD is used, determining the ITD/WT peak ratios from the minA/minA-ALT assay can provide a quantitative assessment of the proportion of ITD positive tumour material in a sample. For further comment on FP12, and the reasons for possible exceptions, see Discussion. Discussion Here we describe two different strategies for detecting KIT exon 11 ITDs in canine cutaneous Mast Cell Tumours, based around identification of a minimally common amplified region, and either detecting duplication of primer sites within that region (the minA/minA-ALT primers sets) or detecting the expansion of that region (the minB/minB-ALT primer sets). We have found that the small difference in size between WT and ITD products for the minB/minB-ALT primer sets make the melt curves for the respective products difficult to distinguish and they cannot be easily separately quantified. In contrast, the large difference in product sizes with the minA/minA-ALT assay (44bp WT; >90bp ITD) makes the products easily distinguished and enables the relative height of the melt curve peaks (equivalent to the greatest rate of change of fluorescence during the melt curve analysis phase) to be used as surrogate for the relative abundance of KIT exon 11 ITD gene copies, and therefore KIT exon 11 ITD positive cells, within the sample. We therefore recommend use of the minA and minA-ALT primer pairs for detection of a KIT exon 11 ITD in a cMCT. Each sample should be tested with at least three replicates of each primer pair (in a 96-well plate, given that one column on the plate is eight wells, using one column per sample with four minA replicates and four minA-ALT replicates is convenient) and ideally each reaction should have 25ng of genomic DNA as template. The presence/absence of a minA/minA-ALT ITD peak determines whether or not a sample is classified as having an ITD. The accuracy of the test in identifying samples with (sensitivity) or without (specificity) a KIT exon 11 ITD on this categorical basis is important and can be considered in two ways. Assuming a sample with unknown exon 11 status presenting for analysis is also of unknown genotype, then that sample would be tested with both minA and minA-ALT primers, with at least three replicate reactions for each primer set. The melt curve profiles for all replicates from both primer sets would then be interpreted holistically to determine whether or not an ITD was present. On this basis, in this study our test correctly identified 5 out of 5 known wild type samples (in addition to MDCK cells) and 10 out of 10 samples known to have exon 11 ITDs. For this small sample number, therefore, analytical specificity and sensitivity were 100%. However, one could also consider sensitivity and specificity on a reaction by reaction basis, as any potential false positive / false negative events are likely to be due to random mis-annealing events occurring in a reaction well, rather than something affecting the whole sample. This is a more stringent assessment of specificity / sensitivity. Taking this approach, and considering any false negative / false positive events occurring in the replicates in the data presented here for (i) only minA reactions (because of the numbers available), (ii) only reactions with a minimum input of 25ng template, (iii) only MDCK or FFPE DNA template reactions, then of 39 individual replicate reactions on known exon 11 ITD samples, 39 produced melt curves which correctly identified the sample as an ITD, an analytical sensitivity of 100%. Of 36 individual replicate reactions on known wild type samples, 34 produced melt curves which correctly identified the sample as WT, an analytical specificity of 94%. Given that the presence of two replicates in a known WT sample with low ITD false positive peaks (FP9 in Figure 7A ) is not consistent either with the third replicate from that sample or the complete lack of any ITD peak in the matched set of minA-ALT reactions, if this was a sample of unknown status a priori then this result should automatically be treated with caution and trigger a retest. It is therefore unlikely that this sample would be reported back as positive. Future studies comparing samples with known KIT exon 11 ITD status will be needed to increase numbers and refine these data and reporting criteria. Separately from the categorical detection of the presence of a KIT exon 11 ITD in a sample, is the ability of the minA/minA-ALT assay to quantify the relative abundance of cells carrying a KIT exon 11 ITD, compared to wild-type cells, in a sample. It is likely that any cMCT sample contains a mixture of non-transformed and transformed cells (38, 39). Non-transformed cells may include regions of epidermis or subcutaneous tissue included as part of a wide excision, or they may by fibroblasts or leukocytes admixed with the neoplastic cells as part of the tumour mass. The neoplastic cells may or may not contain a KIT exon 11 ITD. Even if they do contain an ITD, that duplication will likely only occur on one copy of the gene, not both, as we have demonstrated here by finding LOH in two tumour samples. Therefore, even in a sample of pure neoplastic mast cells from a cMCT with an ITD, if this were possible, 50% of KIT gene copies will contain an ITD and 50% will be wild type. In reality, in a sample with a mixture of non-transformed and transformed cells, the proportion of KIT gene copies with an ITD will be decreased still further. If a sample consists of 50% transformed cells and 50% non-transformed cells, then only 25% of KIT gene copies in the sample will contain an ITD. Furthermore, these arguments assume that the KIT ITD is a founder mutation in the tumour and that all cells in the tumour contain the ITD. It is likely that a tumour could contain multiple clones and that development of a KIT ITD may be a late event in tumour formation, only occurring in a small clonal population within the tumour bulk. In such a situation, the proportion of ITD to WT KIT gene copies will be diluted still further. An additional complication, as demonstrated by amplification from the synthetic ITD oligos ( Figure 3 and Figure 4 ), is that even a ‘pure’ ITD template can give both WT and ITD products, depending on the exact structure of the duplication and amount of template input. Furthermore, the presence of the duplication appeared to lead to self-priming and concatamerisation of PCR products, as shown by ‘laddering’ or ‘smears’ when products were analysed on DNA gels. Overall, therefore, there are too many variables to be confident of using the results to provide exact numbers of neoplastic cells present in a sample. However, it is clear that height of the ITD melt curve peak relative to the height of the WT peak provides relative quantitative information about the proportion of KIT gene copies with ITD in a sample relative to WT copies. Indeed, with the exception of sample FP12, the ITD/WT peak ratio in the FFPE samples was significantly positively correlated with the amount of tumour in that sample as a proportion of the overall amount of tissue ( Figure 7D ). Considering that the FP12 sample, which had been externally validated as carrying an ITD, was composed of >70% tumour mass (although the mass includes transformed and non-transformed cells), the low ITD/WT peak ratio in that sample was surprising ( Figure 7D ). Other samples with a high proportion of tumour to total tissue (e.g. FP2, FP5, FP6) ( Additional File 9C ) had substantially higher ITD/WT peak ratios. Furthermore, amplification of the ITD in FP12 by standard PCR across the whole of KIT exon 11 and subsequent gel electrophoresis was weak in the first repeat and not successful in the second ( Additional File 10 ). The results for FP12 are consistent with a model in which the KIT exon 11 ITD population in this tumour is a small neoplastic subclone within the tumour as a whole. Further work will be needed to fully explore the significance of such populations for therapy and whether using our assay and the ITD/WT peak ratio will be a way of identifying them in tumours in the future. To facilitate this, it would be helpful to interpret the results of future minA/minA-ALT assays alongside H&E images of the samples used for analysis. Given that cMCTs with, or without, an ITD can respond differently to TKIs (24, 27), knowing whether or not an ITD positive population is likely to form a major or a minor clone within a tumour may help to guide clinical management decisions. We chose to develop our KIT exon 11 ITD assay using a SYBR Green qPCR approach. Standard PCR approaches with gel-based analysis are good at identifying size differences in a PCR product (e.g. the difference between a wild-type template and one containing a duplication) but their sensitivity is lower compared to real-time fluorescence-based approaches, they are, at best, semi-quantitative (40, 41). Probe-based fluorescence approaches, such as digital droplet PCR or probe-based qPCR are sensitive and quantitative, but while a ddPCR/qPCR probe may be designed and optimized for one specific KIT exon 11 ITD, there is no guarantee it would be universally applicable to all KIT exon 11 ITDs. However, SYBR green-based qPCR is sensitive and quantitative and does not rely on a specific probe site to be present in all PCR products. Providing the chosen primer site is capable of amplifying the DNA, the exact sequence of the DNA product (other than the primer binding sites) is irrelevant. SYBR green qPCR is thus likely to work with a broader range of KIT exon 11 ITDs than a probe-based approach. Finally, SYBR green-based qPCR can, at the end of the assay, generate a melting curve analysis of the PCR product. The shape of this curve, and the temperature at which it peaks (Tm), are dependent of the physico-chemical properties of the product(s), specifically its length and GC content. Therefore, if there is a sufficient size difference between two products which may be amplified by a pair of primers in SYBR green-based qPCR assay, then the presence of the two products should be detectable by differences in the melt curves and the Tm. This is normally used in gene expression analysis to determine whether or not a set of primers is amplifying solely its intended target. However, we reasoned we could exploit melting curve analysis of qPCR products from genomic DNA template to determine whether or not a sample contained a KIT exon 11 ITD and indeed this proved to be the case. Our SYBR green-based qPCR melting curve assay has the advantages of simplicity and low cost. It uses widely available reagents and the primers require no special modifications. Melting curve analysis can be carried out by many, if not all, real time PCR machines. Its power lies in the specific primer sequence targeting the common minimally amplified region. Our study has limitations. Tumour area measurements relative to total tissue ( Additional File 9C ) for correlation with ITD/WT peak ratios were carried out at low power by manually drawing around the whole tissue section and the region of the tumour within that. We have not accounted for the presence of non-neoplastic cells within the tumour itself. We have not adjusted for potential differences in necrosis between the tumours which could affect the total cellularity. Furthermore, two tumours (FP8 and FP13) had particularly infiltrative, as opposed to expansile, growth patterns, which made accurate tracing of their borders problematic. Given these issues, it is remarkable that there is (once FP12 is excluded) a significant correlation between the calculated tumour to total tissue ratio and the ITD/WT peak ratio. The minA primer set could generate signals from primer dimers in No Template control reactions. The Tm value for the melt curve peak from minA primer dimers was very similar to the melt curve peak for the WT exon 11 product. However, this issue could not lead to a false negative signal if a sample contained an ITD as the template would outcompete any primer dimer formation. An additional limitation of the study was the use of standard Taq polymerase for amplification of DNA fragments for sequencing and genotyping, which could have introduced sequence errors as the Taq was not an error-free proofreading polymerase. However, the size of the amplicons was relatively small, making the insertion of mutations unlikely, although not impossible. Furthermore, in all cases the DNA sequence of the amplicons was identical to the publicly available sequence builds, the only variations being at the sites of known SNPs, and those variations were known alleles, with one exception. The exception was the results of sequencing the FP6 ITD at SNP 672, which gave a genotype of C with the reverse sequencing primer (consistent with the WT germline sequence) but gave C/A heterozygousity with the forward sequencing primer. It is unclear whether this is a PCR amplification error or sequencing error. As the forward and reverse sequencing reactions were carried out on the same amplified fragment, the latter seems more likely. Overall, therefore, we consider the sequencing and genotyping data reliable. Conclusion We have developed a novel, rapid, qPCR-based assay for detecting the presence of KIT exon 11 internal tandem duplications in canine cutaneous Mast Cell Tumours. The assay uses PCR-product melt curve analysis to detect the duplications with high sensitivity and specificity. The ratio of the height of the ITD peak to the WT peak in the melt curve is a measure of the abundance of the ITD in the sample. The assay has potential for use in a wide range of samples, including non-surgical samples of limited material such as fine needle aspirates or blood samples containing cell-free tumour DNA. Abbreviations FFPE, formalin fixed paraffin embedded gDNA, genomic DNA H&E, Haematoxylin and eosin ITD, internal tandem duplication LOH, Loss of Heterozygousity MCT, mast cell tumour MDCK cells, Madin-Darby canine kidney cells qPCR, quantitative real-time PCR TBE, Tris base (90mM) Boric acid (90mM) EDTA (2mM) WT, wild type Declarations Ethics approval and consent to participate This study was approved by the Bristol University/Bristol Veterinary School Animal Welfare and Ethics Review Body (VIN number 22-041). Supply of archival diagnostic samples by Finn Pathologists (Harleston, Norfolk) was also approved by the CVS internal ethical review panel. Competing interests This study is subject to a patent filing in the UK (application reference 2602035.4) and the min assay has been licenced for commercial use to CVS/Axiom. CVS group supplied samples for use in the study but had no control over the direction of the study, the analysis or the decision to publish.. Sam Beck is Director of Independent Anatomic Pathology Ltd. Melanie Dobromylskyj is a full-time employee of The Veterinary Pathology Group. Langford Vets is a business separate to Bristol Veterinary School. Authors' contributions (CRediT Taxonomy) Conceptualization – MJS, GT; Data curation – MJS; Formal analysis – MJS, LSH, JOE, GT; Funding acquisition – MJS; Investigation – MJS, LSH, JOE, GT; Methodology – MJS, JOE, GT; Project administration – MJS; Resources – SB, MD, GE; Supervision – MJS; Visualization – MJS; Writing – original draft – MJS; Writing – review & editing – MJS, JOE, LSH, GT, SB, GE, MD. Funding This study was funded by a Cardiff University Integrated Master’s student project, a BBSRC Impact Acceleration Account award through Cardiff University and Smalley laboratory discretionary funds. Neither Cardiff University, Finn Pathologists, AML or Bristol University/Langford Vets had any role in the design, analysis and reporting of the study (other than the roles of the individual authors). The Cardiff University Technology Transfer Office have been involved in the development of foreground intellectual property arising from the study. Acknowledgements The authors would like to thank the pathology team at Finn Pathologists, Harleston, Norfolk, UK for supply of archival diagnostic tissue. Consent for publication All authors have read the manuscript and consented to its publication. Availability of data and materials Data sharing is not applicable to this article as no datasets were generated or analysed during the current study. References Dobson JM, Samuel S, Milstein H, Rogers K, Wood JL (2002) Canine neoplasia in the UK: estimates of incidence rates from a population of insured dogs. J Small Anim Pract 43(6):240–246 Meuten DJ (2017) Tumors in domestic animals. Fifth edition. ed. Ames, Iowa: Wiley/Blackwell; viii, 989 pages p Bostock DE (1986) Neoplasms of the skin and subcutaneous tissues in dogs and cats. Br Vet J 142(1):1–19 Brodey RS (1970) Canine and feline neoplasia. Adv Vet Sci Comp Med 14:309–354 Kaldrymidou H, Leontides L, Koutinas AF, Saridomichelakis MN, Karayannopoulou M (2002) Prevalence, distribution and factors associated with the presence and the potential for malignancy of cutaneous neoplasms in 174 dogs admitted to a clinic in northern Greece. J Vet Med Physiol Pathol Clin Med 49(2):87–91 Macy DW (1985) Canine mast cell tumors. Vet Clin North Am Small Anim Pract 15(4):783–803 O'Keefe DA (1990) Canine mast cell tumors. Vet Clin North Am Small Anim Pract 20(4):1105–1115 Rothwell TL, Howlett CR, Middleton DJ, Griffiths DA, Duff BC (1987) Skin neoplasms of dogs in Sydney. Aust Vet J 64(6):161–164 Welle MM, Bley CR, Howard J, Rufenacht S (2008) Canine mast cell tumours: a review of the pathogenesis, clinical features, pathology and treatment. Vet Dermatol 19(6):321–339 Thamm DH, Avery AC, Berlato D, Bulman-Fleming J, Clifford CA, Hershey AE et al (2019) Prognostic and predictive significance of KIT protein expression and c-kit gene mutation in canine cutaneous mast cell tumours: A consensus of the Oncology-Pathology Working Group. Vet Comp Oncol 17(4):451–455 Choi YJ, Yoo JS, Jung K, Rice L, Kim D, Zlojutro V et al (2023) Lung-specific MCEMP1 functions as an adaptor for KIT to promote SCF-mediated mast cell proliferation. Nat Commun 14(1):2045 Franke K, Kirchner M, Mertins P, Zuberbier T, Babina M (2022) The SCF/KIT axis in human mast cells: Capicua acts as potent KIT repressor and ERK predominates PI3K. Allergy 77(11):3337–3349 Gilfillan AM, Beaven MA (2011) Regulation of mast cell responses in health and disease. Crit Rev Immunol 31(6):475–529 Gilfillan AM, Tkaczyk C (2006) Integrated signalling pathways for mast-cell activation. Nat Rev Immunol 6(3):218–230 Downing S, Chien MB, Kass PH, Moore PE, London CA (2002) Prevalence and importance of internal tandem duplications in exons 11 and 12 of c-kit in mast cell tumors of dogs. Am J Vet Res 63(12):1718–1723 Jones CL, Grahn RA, Chien MB, Lyons LA, London CA (2004) Detection of c-kit mutations in canine mast cell tumors using fluorescent polyacrylamide gel electrophoresis. J Vet Diagn Invest 16(2):95–100 Letard S, Yang Y, Hanssens K, Palmerini F, Leventhal PS, Guery S et al (2008) Gain-of-function mutations in the extracellular domain of KIT are common in canine mast cell tumors. Mol Cancer Res 6(7):1137–1145 London CA, Galli SJ, Yuuki T, Hu ZQ, Helfand SC, Geissler EN (1999) Spontaneous canine mast cell tumors express tandem duplications in the proto-oncogene c-kit. Exp Hematol 27(4):689–697 Reguera MJ, Ferrer L, Rabanal RM (2002) Evaluation of an intron deletion in the c-kit gene of canine mast cell tumors. Am J Vet Res 63(9):1257–1261 Riva F, Brizzola S, Stefanello D, Crema S, Turin L (2005) A study of mutations in the c-kit gene of 32 dogs with mastocytoma. J Vet Diagn Invest 17(4):385–388 Zemke D, Yamini B, Yuzbasiyan-Gurkan V (2002) Mutations in the juxtamembrane domain of c-KIT are associated with higher grade mast cell tumors in dogs. Vet Pathol 39(5):529–535 Isotani M, Ishida N, Tominaga M, Tamura K, Yagihara H, Ochi S et al (2008) Effect of tyrosine kinase inhibition by imatinib mesylate on mast cell tumors in dogs. J Vet Intern Med 22(4):985–988 Pryer NK, Lee LB, Zadovaskaya R, Yu X, Sukbuntherng J, Cherrington JM et al (2003) Proof of target for SU11654: inhibition of KIT phosphorylation in canine mast cell tumors. Clin Cancer Res 9(15):5729–5734 Hahn KA, Ogilvie G, Rusk T, Devauchelle P, Leblanc A, Legendre A et al (2008) Masitinib is safe and effective for the treatment of canine mast cell tumors. J Vet Intern Med 22(6):1301–1309 Webster JD, Yuzbasiyan-Gurkan V, Thamm DH, Hamilton E, Kiupel M (2008) Evaluation of prognostic markers for canine mast cell tumors treated with vinblastine and prednisone. BMC Vet Res 4:32 Webster JD, Yuzbasiyan-Gurkan V, Kaneene JB, Miller R, Resau JH, Kiupel M (2006) The role of c-KIT in tumorigenesis: evaluation in canine cutaneous mast cell tumors. Neoplasia 8(2):104–111 London CA, Malpas PB, Wood-Follis SL, Boucher JF, Rusk AW, Rosenberg MP et al (2009) Multi-center, placebo-controlled, double-blind, randomized study of oral toceranib phosphate (SU11654), a receptor tyrosine kinase inhibitor, for the treatment of dogs with recurrent (either local or distant) mast cell tumor following surgical excision. Clin Cancer Res 15(11):3856–3865 Weishaar KM, Ehrhart EJ, Avery AC, Charles JB, Elmslie RE, Vail DM et al (2018) c-Kit Mutation and Localization Status as Response Predictors in Mast Cell Tumors in Dogs Treated with Prednisone and Toceranib or Vinblastine. J Vet Intern Med 32(1):394–405 Rout ED, Avery AC (2012) Molecular Diagnostics of Hematologic Malignancies. In: Thrall MA, editor. Veterinary hematology and clinical chemistry. 2nd ed. Ames, Iowa: Wiley-Blackwell; p. xii, 762 p Iodence AE, Wallace ML, Grimes JA, Schmiedt CW (2021) Dogs undergoing surgical excision of mast cell tumors are not at increased risk of incisional complications. J Am Vet Med Assoc 260(S1):S88–S95 Ishiguro T, Kadosawa T, Takagi S, Kim G, Ohsaki T, Bosnakovski D et al (2003) Relationship of disease progression and plasma histamine concentrations in 11 dogs with mast cell tumors. J Vet Intern Med 17(2):194–198 London CA, Seguin B (2003) Mast cell tumors in the dog. Vet Clin North Am Small Anim Pract 33(3):473–489 v Wang C, Wallerman O, Arendt ML, Sundstrom E, Karlsson A, Nordin J et al (2021) A novel canine reference genome resolves genomic architecture and uncovers transcript complexity. Commun Biol 4(1):185 Field MA, Rosen BD, Dudchenko O, Chan EKF, Minoche AE, Edwards RJ et al (2020) Canfam_GSD: De novo chromosome-length genome assembly of the German Shepherd Dog (Canis lupus familiaris) using a combination of long reads, optical mapping, and Hi-C. Gigascience. ;9(4) Dolezel J, Bartos J, Voglmayr H, Greilhuber J (2003) Nuclear DNA content and genome size of trout and human. Cytometry A 51(2):127–128 author reply 9 Evers B, Jonkers J (2006) Mouse models of BRCA1 and BRCA2 deficiency: past lessons, current understanding and future prospects. Oncogene 25(43):5885–5897 Neto RT, Cagnini DQ, Amorim RL (2013) Mutations in C-KIT exon 11 in canine cutaneous mast cell tumors. BMC Proceedings. ;7(2):P56 Sharma A, Merritt E, Hu X, Cruz A, Jiang C, Sarkodie H et al (2019) Non-Genetic Intra-Tumor Heterogeneity Is a Major Predictor of Phenotypic Heterogeneity and Ongoing Evolutionary Dynamics in Lung Tumors. Cell Rep 29(8):2164–74e5 Sun XX, Yu Q (2015) Intra-tumor heterogeneity of cancer cells and its implications for cancer treatment. Acta Pharmacol Sin 36(10):1219–1227 Dagher H, Donninger H, Hutchinson P, Ghildyal R, Bardin P (2004) Rhinovirus detection: comparison of real-time and conventional PCR. J Virol Methods 117(2):113–121 Pestana E, Belak S, Diallo A, Crowther JR, Viljoen GJ (2014) Early, rapid and sensitive veterinary molecular diagnostics - real time PCR applications. Springer Dordrecht, p 310 Additional Declarations The authors declare potential competing interests as follows: This study is subject to a patent filing in the UK (application reference 2602035.4) and the min assay has been licenced for commercial use to CVS/Axiom. CVS group supplied samples for use in the study but had no control over the direction of the study, the analysis or the decision to publish.. Sam Beck is Director of Independent Anatomic Pathology Ltd. Melanie Dobromylskyj is a full-time employee of The Veterinary Pathology Group. Langford Vets is a business separate to Bristol Veterinary School. Supplementary Files 20260409Paper1additionalfilelegends.docx Additional File Legends AdditionalFile1.xlsx Additional File 1: PCR primers, synthetic oligo target sequences and PCR conditions. AdditionalFile2.xlsx Additional File 2: Clinical samples used in the study. AdditionalFile3.tif Additional File 3: Determining ITD structure from boundaries of superimposed sequences. AdditionalFile4.tif Additional File 4: Inferred structure of the FP15 ITD. AdditionalFile5.tif Additional File 5: Sequences of C2 canine mastocytomoa cell line and four canine MCTs (BK, EH, MK, MS) from London et al . (18). AdditionalFile6.tif Additional File 6: Alignment of Brca primers to C. familiaris BRCA1. AdditionalFile7.tif Additional File 7: Primer dimer reactions with minA primers. AdditionalFile8.tif Additional File 8: MDCK standard curves melt curves. AdditionalFile9.tif Additional File 9: H&E stained low-power overviews of FP tumour set with known KIT exon 11 status demonstrating presence of tumour tissue in sections AdditionalFile10.tif Additional File 10: Gel analysis of whole KIT exon 11 standard PCR on FP samples. AdditionalFile11.xlsx Additional File 11: Ct and Tm values for minA, minB and Brca analysis of clinical samples. AdditionalFile12.tif Additional File 12: Brca melting curves for clinical samples. AdditionalFile13.tif Additional File 13: Sequencing of WT KIT exon 11 PCR products from clinical samples. AdditionalFile14.txt Additional File 14: FASTA sequences of KIT exon 11 ITDs shown in Additional File 15. AdditionalFile15.tif Additional File 15: Features of clinical sample ITDs. AdditionalFile16.tif Additional File 16: Loss-Of-Heterozygousity in ITDs, allele frequencies and ITD detection efficiency. AdditionalFile17.tif Additional File 17: The minB-ALT variant primers can enhance detection in tumours with allelic variants at polymorphic loci. AdditionalFile18.pdf Additional File 18: Uncropped gel images from Figure 2, Figure 3 and Additional File 10. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9368633","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":620329684,"identity":"1f29e4df-a578-4725-adc6-e208307755fc","order_by":0,"name":"Jacob O. Evans","email":"","orcid":"https://orcid.org/0009-0007-7816-7694","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Jacob","middleName":"O.","lastName":"Evans","suffix":""},{"id":620329685,"identity":"44cc52fb-b3a1-46d9-ac52-b76fd059a4be","order_by":1,"name":"Liam S. Hill","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Liam","middleName":"S.","lastName":"Hill","suffix":""},{"id":620329686,"identity":"baeb1336-1b0a-4bb5-a75e-37496e167f65","order_by":2,"name":"Sam Beck","email":"","orcid":"","institution":"Independent Anatomic Pathology, Ltd","correspondingAuthor":false,"prefix":"","firstName":"Sam","middleName":"","lastName":"Beck","suffix":""},{"id":620329687,"identity":"47589667-38e9-4db6-b972-2532c4ae71f7","order_by":3,"name":"Grace Edmunds","email":"","orcid":"","institution":"Department of Pathobiology, University of Pennsylvania, School of Veterinary Medicine","correspondingAuthor":false,"prefix":"","firstName":"Grace","middleName":"","lastName":"Edmunds","suffix":""},{"id":620329688,"identity":"9d95a195-3c88-430c-9bfe-7c8359a4a0c5","order_by":4,"name":"Melanie Dobromylskyj","email":"","orcid":"","institution":"The Veterinary Pathology Group","correspondingAuthor":false,"prefix":"","firstName":"Melanie","middleName":"","lastName":"Dobromylskyj","suffix":""},{"id":620329689,"identity":"a5353971-9f79-404a-9beb-ba5cd62889e3","order_by":5,"name":"Giusy Tornillo","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Giusy","middleName":"","lastName":"Tornillo","suffix":""},{"id":620329690,"identity":"9040b71b-5a1e-48d4-99f5-f8db031fd7ce","order_by":6,"name":"Matthew J. Smalley","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDElEQVRIiWNgGAWjYBAC+QYGxgMPGBgS29sbgFw2CX72BrBEAk4tBgcYGA4ApYsbew6AtUiCabxaGCBa6htngBSxMRChRezwgwMJFYdzG2c+fvi5osxCgoeB+eEHxrY03H6ZnWZwIOEMUMvsNGPJM+ckgFrYjCUY23JwamG4nWBwILENpCWHQbKxTaLOnoHBjIGxrQKPlvQPBxL/HU5snHmG+SdQC9AW9m8EtOQAbWk4nDhxBg+bJEQLD8gW3A4zuJ1TcCDhWHribJ40M8sGkF+YeYolEs7h8376xgcfaqwT+9gPP77ZUFYnwcPevvHDh7Jk3A6DgGYkNjMDvliBgzrCSkbBKBgFo2DkAgCh11uIzJH+UgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-9540-1146","institution":"Cardiff University","correspondingAuthor":true,"prefix":"","firstName":"Matthew","middleName":"J.","lastName":"Smalley","suffix":""}],"badges":[],"createdAt":"2026-04-09 12:36:33","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":true,"conflictsOfInterestStatement":true,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":true},"doi":"10.21203/rs.3.rs-9368633/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9368633/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106597380,"identity":"c37b7b20-2850-4df0-b718-0c671ff9bfed","added_by":"auto","created_at":"2026-04-10 09:41:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3530002,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003emin\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e region and primer design.\u003c/strong\u003e \u003cstrong\u003e(A)\u003c/strong\u003e Alignment of regions duplicated in C2 cell line and BK, EH, MK and MS mast cell tumours from London \u003cem\u003eet al\u003c/em\u003e. Figure 3 [18] against KIT genomic sequence showing the consensus overlap corresponding to the minimal commonly amplified region (UU_Cfam_GSD_1.0/canFam4 13: 47,940,605 – 47,940,644). \u003cstrong\u003e(B)\u003c/strong\u003e Structure of wild type \u003cem\u003eC. familiaris\u003c/em\u003e KIT exon 11 (starting at UU_Cfam_GSD_1.0/canFam4 13: 47,940,533) and including the coding sequence for amino acids 549 – 590 of the juxtamembrane region. Parts of introns 10 and 11 are shown. The region also includes two polymorphic loci, 47,940,619 C/T and 47,940,672 T/C, indicated by triangles below the sequence and red boxes. The minimal commonly duplicated region and flanking variable regions (from (A)) have been mapped on to the sequence as have the locations of the minA, minB and whole exon 11 primer binding sites. Note the location of the 47,940,619 C/T polymorphism in the forward minA primer site and the 47,940,672 T/C in the reverse minB primer site.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/956bf1229f9b4407ff5f1194.png"},{"id":106597394,"identity":"df2d739a-6f4f-4fcf-8eb0-5293c167682b","added_by":"auto","created_at":"2026-04-10 09:41:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":4958270,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistinct products are amplified by the minA, minB and Brca primers. (A) \u003c/strong\u003eGel analysis of products (0.1x of reaction volume loaded) amplified by minA, minB and Brca primers using 50ng of MDCK gDNA as template. Products are faintly visible (red boxes) in regions corresponding to \u0026lt;100bp (minA and Brca) or approximately 100bp (minB), consistent with their predicted sizes. \u003cstrong\u003e(B) \u003c/strong\u003eGel analysis of products (0.1x of reaction volume loaded) amplified by minA, minB and Brca primers, using 5ul of purified product from the reactions shown in (A) as template. Products of predicted size are seen more clearly. \u003cstrong\u003e(C, D) \u003c/strong\u003eAmplification curves from qPCR analysis of No Template controls (C) and 50ng MDCK per reaction (D) using the minA, minB and Brca primers (triplicate reactions). \u003cstrong\u003e(E – G)\u003c/strong\u003e SYBR green melt curves of PCR products (minA, (E); minB, (F): Brca, (G)) produced by reactions shown in (D). Note that a single peak is generated by each triplicate reaction in a highly reproducible manner. \u003cstrong\u003e(H) \u003c/strong\u003eGel analysis of product (0.4x of each reaction volume loaded) from one well for each template/primer combination shown in (C) – (G). Product sizes are consistent with the band sizes in (B). Plots in (C) – (G) taken directly for Quantstudio software as examples of the automatically generated data plots. Figure 2A was contrast/brightness enhanced for ease of interpreting this very faint gel. The original and uncropped gel images are available in \u003cstrong\u003eAdditional File 18A\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/2b688f54731c7c12943eb32b.png"},{"id":106597378,"identity":"daae2bdb-7749-4691-a03f-a26a098a06ec","added_by":"auto","created_at":"2026-04-10 09:41:47","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":6084128,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe minA and minB primer sets detect KIT exon 11 ITDs in qPCR assays. (A)\u003c/strong\u003e Gel analysis of amplification products of minA, minB and Brex 11 primer sets on No Template (NT) control, MDCK template, Type I and Type II synthetic ITD. \u003cstrong\u003e(B)\u003c/strong\u003e – \u003cstrong\u003e(D)\u003c/strong\u003e Amplification curves (direct from Quantstudio software) for MDCK template (blue), Type I ITD (green), Type II ITD (yellow) and No Template control (red) with minA (B), minB (C) and Brex 11 (D) primers. \u003cstrong\u003e(E) – (G)\u003c/strong\u003e Melt curves of minA (E), minB (F) and Brca (G)-amplified products with MDCK template, Type I ITD, Type II ITD and No Template control. Triplicates from each reaction set are overlaid. Results from the four different sets of reactions are plotted using Prism and offset on the Y-axis for visibility but are in proportion with respect to each other and therefore comparable. In (E), arrow indicates WT minA peak; Red horizontal bar indicates region where peak indicating an ITD is found (the exact position depending on the ITD). In (F), arrow indicates rightwards shift of minB product peak when ITD is present. Only the MDCK template gives a peak with the Brca primers. \u003cstrong\u003e(H)\u003c/strong\u003e Details of Ct and Tm values and \u003cstrong\u003e(I)\u003c/strong\u003e gel analysis of products from one representative well (and triplicate minA/NT wells) for qPCR amplification reactions shown in (B) – (G). The background band in one well of the minA/NT reactions is indicated by a box. *One minA/NT reaction gave a product after \u0026gt;38 cycles consistent with primer dimer formation in the absence of template. Uncropped gel images are available in \u003cstrong\u003eAdditional File 18B\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/0876006e4026ef1c4ae38d2e.png"},{"id":106597418,"identity":"ef2dd498-a217-4290-aa91-e76f31847156","added_by":"auto","created_at":"2026-04-10 09:42:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1366834,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStandard curve analysis of minA, minB and Brca primer sets. (A) – (C) \u003c/strong\u003eStandard curve for minA (A), minB (B) and Brca (C) primers with MDCK gDNA. Plotted as Ct values against Log10 input values. Simple linear regression plotted with 95% confidence intervals for inputs above 50pg. R\u003csup\u003e2\u003c/sup\u003e values and the slope of the regression are shown. Only a single minA reaction out of four amplified with 5pg input. Three minB reactions out of four amplified with 50pg input. Only two out of four minB reactions amplified with both 5pg and 0.5pg input. There was no amplification from Brca reactions with 5pg or 0.5pg input. See also \u003cstrong\u003eAdditional File 8\u003c/strong\u003e \u003cstrong\u003e(D, E) \u003c/strong\u003eStandard curve for minA and minB primers with Type I (D) or Type II (E) ITD synthetic oligo input. Plotted as Ct values against Log10 input values. For the Type I synthetic oligo (D), the results best fitted a hyperbolic curve (R\u003csup\u003e2\u003c/sup\u003e values shown on the graph). Only two reactions amplified at 3.3ul 10fM input with minA. None amplifed at 3.3ul 1fM input with minA. With minB, no reactions generated product with either 3.3ul 1fM or 3.3ul 10fM input. For the Type II synthetic oligo (E), the results fitted a simple linear regression. \u003cstrong\u003e(F)\u003c/strong\u003e Melting curve analysis for Type I (left) and Type II (right) ITDs with minA (upper) and minB (lower) primers, at inputs ranging from 3.3ul 1fM to 3.3ul 10pM template. The curves from MDCK 50pg reactions are shown as controls. Triplicates from each reaction set are overlaid. Results from the four different sets of reactions are offset on the Y-axis for visibility but are in proportion with respect to each other.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/949f04830cf04e26f2abc5bd.png"},{"id":106597402,"identity":"a40fce17-33f4-4593-80f5-05157ce56282","added_by":"auto","created_at":"2026-04-10 09:42:00","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1335842,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eminA melting curves for clinical samples. (A) \u003c/strong\u003eControl MDCK melt curve.\u003cstrong\u003e (B) \u003c/strong\u003eMelt curves for KIT WT tumour samples.\u003cstrong\u003e (C) \u003c/strong\u003eMelt curves for tumour samples with KIT exon 11 ITDs.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/ef5a8f8ed161be749b359f88.png"},{"id":106597370,"identity":"dd917bc4-7c21-47b2-9b6b-fb2f59b2cb15","added_by":"auto","created_at":"2026-04-10 09:41:44","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1299616,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eminB melting curves for clinical samples. (A) \u003c/strong\u003eControl MDCK melt curve.\u003cstrong\u003e (B) \u003c/strong\u003eMelt curves for KIT WT tumour samples.\u003cstrong\u003e (C) \u003c/strong\u003eMelt curves for tumour samples with KIT exon 11 ITDs.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/273eda40127590ff64a8ff23.png"},{"id":106597414,"identity":"b96595c3-59bf-41b0-bc67-49ea6a6f6264","added_by":"auto","created_at":"2026-04-10 09:42:06","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":3758770,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe minA-ALT variant primers enhance detection in tumours with allelic variants at polymorphic loci.\u003c/strong\u003e \u003cstrong\u003e(A)\u003c/strong\u003e qPCR melt curves for samples analysed using the alternative minA-ALT forward primer\u003cstrong\u003e \u003c/strong\u003ecompared to the standard primers. WT and ITD genotypes are shown on each melt curve and the primer sequences are provided. There are two small false positive peaks in FP9 with the minA primers (black arrow); there are none with minA-ALT. In three ITDs with a SNP 619 T allele, FP5, FP6 and FP8, detection is substantially enhanced by the minA-ALT primer (red arrows). \u003cstrong\u003e(B)\u003c/strong\u003e Linear regression analysis of tumour/total tissue ratio for samples carrying an ITD against the strength of detection of that ITD by the minA assay primers (PdITD/PdWT ratio) for SNP 619 C (blue circles) and SNP 619 T (red squares) FFPE samples. R\u003csup\u003e2\u003c/sup\u003e and P values are indicated. \u003cstrong\u003e(C)\u003c/strong\u003e Linear regression analysis as in (B) using minA or minA-ALT assay PdITD/PdWT ratios appropriate to the ITD genotype. \u003cstrong\u003e(D) \u003c/strong\u003eLinear regression analysis as in (C) but excluding the FP12 sample (green open circle).\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/fc653b65c8091b315ec1ad0d.png"},{"id":106959674,"identity":"fa69f564-a8d1-417b-a957-342934fac35f","added_by":"auto","created_at":"2026-04-15 09:13:24","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":27925586,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/64119bd7-0d11-4229-b5cd-2874309fea8b.pdf"},{"id":106597383,"identity":"99211edc-131a-49f4-9b66-67bc87b10ff8","added_by":"auto","created_at":"2026-04-10 09:41:48","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":53547,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional File Legends\u003c/p\u003e","description":"","filename":"20260409Paper1additionalfilelegends.docx","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/1a7c5d32df89442b1a662f31.docx"},{"id":106597415,"identity":"d5096ad7-ae60-4122-9477-7ea7650b4082","added_by":"auto","created_at":"2026-04-10 09:42:06","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":15509,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 1: PCR primers, synthetic oligo target sequences and PCR conditions.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/fef839e74491177b9f230077.xlsx"},{"id":106597384,"identity":"f6665e81-03eb-4a77-93b9-2f3400a39058","added_by":"auto","created_at":"2026-04-10 09:41:48","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":11812,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 2: Clinical samples used in the study.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/edb85890fce2cbadb9046f57.xlsx"},{"id":106597369,"identity":"156472bf-3242-42be-9267-ed97ac9c0f8b","added_by":"auto","created_at":"2026-04-10 09:41:44","extension":"tif","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":1436172,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 3: Determining ITD structure from boundaries of superimposed sequences.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile3.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/f2d261ee78d4b12a2895638f.tif"},{"id":106597396,"identity":"89ee1ba4-89a8-423a-814d-cebe10f063cf","added_by":"auto","created_at":"2026-04-10 09:41:52","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":329556,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 4: Inferred structure of the FP15 ITD.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile4.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/6b46f086318457c7762a8129.tif"},{"id":106728556,"identity":"52b2c766-8ab8-48f3-bd97-48684b1eb241","added_by":"auto","created_at":"2026-04-12 18:43:14","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1233116,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 5: Sequences of C2 canine mastocytomoa cell line and four canine MCTs (BK, EH, MK, MS) from London \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eet al\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e. (18).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile5.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/fa01e02e0388e7305e4c8fb4.tif"},{"id":106597416,"identity":"94fca4d5-e391-42d6-85ea-7bd82150de58","added_by":"auto","created_at":"2026-04-10 09:42:07","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":778656,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 6: Alignment of Brca primers to \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eC. familiaris\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003eBRCA1.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile6.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/de59dd3c7b95a428c19055ed.tif"},{"id":106597377,"identity":"38148233-b3c7-4fb3-b422-8b5cb4c5ab0f","added_by":"auto","created_at":"2026-04-10 09:41:47","extension":"tif","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":847092,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 7: Primer dimer reactions with minA primers.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile7.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/e4caf39f4832d806d711953b.tif"},{"id":106597372,"identity":"2d45cdac-a68f-48fa-b023-84024664436d","added_by":"auto","created_at":"2026-04-10 09:41:44","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":504484,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 8: MDCK standard curves melt curves.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile8.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/c0e7e2bde34402295e5fe542.tif"},{"id":106597373,"identity":"f190c6e5-8c5f-4d16-b173-0e57681ce227","added_by":"auto","created_at":"2026-04-10 09:41:45","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":5140792,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 9: H\u0026amp;E stained low-power overviews of FP tumour set with known \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eKIT\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e exon 11 status demonstrating presence of tumour tissue in sections\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile9.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/d289636d3ce0161286acba3b.tif"},{"id":106597401,"identity":"bc360291-23ea-46a1-b0c0-21726245a78f","added_by":"auto","created_at":"2026-04-10 09:42:00","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":1469780,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 10: Gel analysis of whole \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eKIT\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e exon 11 standard PCR on FP samples.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile10.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/815ae896e3c864be1b1a0be6.tif"},{"id":106597386,"identity":"7bfbf4f7-cd8e-48b5-b4c3-7472e0d81b4a","added_by":"auto","created_at":"2026-04-10 09:41:49","extension":"xlsx","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":26714,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 11: Ct and Tm values for minA, minB and Brca analysis of clinical samples.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile11.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/6744ac45fd9fd01f72b903b7.xlsx"},{"id":106597420,"identity":"3b9869ae-8304-4bc3-ac6a-c0bfab899e87","added_by":"auto","created_at":"2026-04-10 09:42:07","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":666780,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 12: Brca melting curves for clinical samples.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile12.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/6b3f4c3e7cfb5cb483a92e30.tif"},{"id":106597395,"identity":"926ee17d-9bc7-4d07-83dd-f431bd9bc81a","added_by":"auto","created_at":"2026-04-10 09:41:52","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":696972,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 13: Sequencing of WT KIT exon 11 PCR products from clinical samples.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile13.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/064b467781734b4bea70e821.tif"},{"id":106597419,"identity":"d6b5c716-c7be-4902-99b4-5fd403224397","added_by":"auto","created_at":"2026-04-10 09:42:07","extension":"txt","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":2519,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 14: FASTA sequences of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eKIT\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e exon 11 ITDs shown in Additional File 15.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile14.txt","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/5f304c2a6091fcb8f5622a6e.txt"},{"id":106597385,"identity":"c57bbdf4-5800-412b-93ac-6ed8ccd1e729","added_by":"auto","created_at":"2026-04-10 09:41:48","extension":"tif","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":2452812,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 15: Features of clinical sample ITDs.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile15.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/ad52f904073666962850a576.tif"},{"id":106597382,"identity":"985682ec-9265-445a-b74a-9b9a933a7405","added_by":"auto","created_at":"2026-04-10 09:41:48","extension":"tif","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":627532,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 16: Loss-Of-Heterozygousity in ITDs, allele frequencies and ITD detection efficiency.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile16.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/a526495b0254c59550a3c041.tif"},{"id":106597397,"identity":"fe626ee9-3ad7-48a2-bedb-1099ecbab857","added_by":"auto","created_at":"2026-04-10 09:41:54","extension":"tif","order_by":18,"title":"","display":"","copyAsset":false,"role":"supplement","size":855020,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 17: The minB-ALT variant primers can enhance detection in tumours with allelic variants at polymorphic loci.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile17.tif","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/e1f790cc48d810ab5ff471dc.tif"},{"id":106597374,"identity":"cd0fafb2-36b5-475c-b179-8ae480ef4525","added_by":"auto","created_at":"2026-04-10 09:41:45","extension":"pdf","order_by":19,"title":"","display":"","copyAsset":false,"role":"supplement","size":17792973,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional File 18: Uncropped gel images from Figure 2, Figure 3 and Additional File 10.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"AdditionalFile18.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9368633/v1/236c20eb20e2d1a9dd903781.pdf"}],"financialInterests":"The authors declare potential competing interests as follows: This study is subject to a patent filing in the UK (application reference 2602035.4) and the min assay has been licenced for commercial use to CVS/Axiom. CVS group supplied samples for use in the study but had no control over the direction of the study, the analysis or the decision to publish.. Sam Beck is Director of Independent Anatomic Pathology Ltd. Melanie Dobromylskyj is a full-time employee of The Veterinary Pathology Group. Langford Vets is a business separate to Bristol Veterinary School.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eA novel, minimally invasive diagnostic test for \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eKIT\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e exon 11 internal tandem duplications in canine cutaneous mast cell tumours I: Assay development\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Background","content":"\u003cp\u003eMast cell tumours (MCTs) are one of the most common tumours of dogs, with an age-standardised rate of 126/100,000 dogs/year in a survey of a population of insured dogs in the UK, considering both the cutaneous and subcutaneous forms (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). The tumour can also occur less commonly in the intestine, visceral organs or other sites (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). The cutaneous form, cMCT, is among the most common cutaneous tumours of dogs, accounting for 7\u0026ndash;21% of all canine skin tumours (\u003cspan additionalcitationids=\"CR4 CR5 CR6 CR7 CR8\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe mutation status of the receptor tyrosine kinase KIT is an important parameter to assess within mast cell tumours (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Signalling by KIT, and cross-talk between KIT and IgE receptors, is required for proliferation, differentiation and activation of normal mast cells as part of the allergic inflammatory response (\u003cspan additionalcitationids=\"CR12 CR13\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e) and deregulated activation of KIT is thought to be a key driver of mast cell neoplasia (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). Mutations in the \u003cem\u003eKIT\u003c/em\u003e gene in canine MCTs have been reported in a number of studies, including in the extracellular (exons 8 and 9) and juxta-membrane (exons 11 and 12) domains (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan additionalcitationids=\"CR16 CR17 CR18 CR19 CR20\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). These include point mutations, deletions and internal tandem duplications (ITDs) of exon 11. The presence of the latter is reported in between 9 and 45% of canine MCTs, depending on the study (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). These mutations result in constitutive activation of KIT by autophosphorylation in the absence of its ligand (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn cMCTs, \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs are associated with shorter time to progression and increased likelihood of recurrence or development of metastasis (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e) and with decreased overall survival (\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). It is likely that mutated KIT is both directly influential towards cell behaviour, and that it correlates with high histologic grade driven by multiple other factors. Therefore, the presence of a KIT exon 11 ITD has not been shown to be an independent prognostic marker (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e), nevertheless, as the presence or absence of a mutation is more objective than assessment of grade or proliferation, mutation status can still provide useful information for clinical decision making (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). \u003cem\u003eKIT\u003c/em\u003e mutation status may be most useful for prognostication in histologically low grade or \u0026lsquo;ambiguous\u0026rsquo; tumours (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Importantly, \u003cem\u003eKIT\u003c/em\u003e exon 11 mutation status is also valuable in predicting response to TKIs. cMCTs with \u003cem\u003eKIT\u003c/em\u003e mutations have a better objective response to toceranib (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e), but potentially worse progression-free survival (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e), while overall outcomes, times to progression and overall response rates to masitinib are better in cMCTs with KIT mutations (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eCurrent approaches to exon 11 ITD testing rely on a standard semi-quantitative PCR approach, followed by analysis of PCR product size by electrophoretic separation (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). This approach uses material from tissue biopsy, which may not available if a tumour is not suitable for surgical sampling. In those tumours which cannot be removed, incisional biopsy may be carried out to obtain tissue for determining KIT mutation status, but this approach is associated with the risk of incisional complications (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e) or systemic inflammatory responses associated with histamine release (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). Furthermore, the lack of sensitivity of the standard approach makes currently available KIT assays unsuitable for long-term monitoring of minimal residual disease burden by liquid biopsy of circulating tumour DNA (ctDNA). Therefore, a new approach which defines KIT mutation status without surgical biopsy is needed.\u003c/p\u003e \u003cp\u003eHere, we describe the development and implementation of a novel, inexpensive and rapid SYBR green quantitative real-time PCR (qPCR)-based assay for detection of \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs in canine cMCTs with high sensitivity and specificity (\u0026gt;\u0026thinsp;90%). The assay has the potential for use in a wide range of samples, fulfilling the need for a new approach to testing for \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eAims\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDevelopment of a novel sensitive and specific quantitative real-time PCR-based test for detection of \u003cem\u003eKIT\u003c/em\u003e exon 11 internal duplication in canine mast cell tumours, suitable for use in a variety of sample types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Bristol University/Bristol Veterinary School Animal Welfare and Ethics Review Body (VIN number 22-041). Supply of samples by Finn Pathologists (Harleston, Norfolk; arranged through MD) was approved by the CVS ethical review panel.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSequences\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA reference sequence for \u003cem\u003eCanis lupus familiaris\u003c/em\u003e\u003cem\u003eKIT\u003c/em\u003e proto-oncogene (NM_001003181) based on UU_Cfam_GSD_1.0/canFam4 (chr13:47,940,533-47,940,659; amino acids 559 - 590) was downloaded from the UCSC genome browser into the sequence editing and annotation package of DNASTAR Lasergene version 17.5.0.48 (DNASTAR Inc., Madison, Wisconsin USA) and annotated. Single nucleotide polymorphisms (SNPs) in the region of interest were noted at chr13:47940619 (C/T) and chr13:47940672\u003c/p\u003e\n\u003cp\u003e(T/C) (33).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eKIT\u003c/em\u003e ITD sequences for the C2 canine mastocytoma cell line and the BK, EH, MK and MS canine mast cell tumours from London \u003cem\u003eet al\u003c/em\u003e. Figure 3 (18) were manually copied into DNASTAR Lasergene. Multiple sequence alignment was carried out in using DNASTAR Lasergene MegAlign Pro with Clustal Omega using default settings.\u003c/p\u003e\n\u003cp\u003eA consensus reference sequence for \u003cem\u003eCanis lupus familiaris\u003c/em\u003e\u003cem\u003eBRCA1\u003c/em\u003e (NM_001013416) was built from the canFam1, 2 and 3 genome builds, downloaded from the UCSC genome browser into DNASTAR Lasergene. The earlier builds were used as they formed the basis of protein splice isoform predictions. To clarify the intron/exon structure and annotation of \u003cem\u003eC. familiaris BRCA1, \u003c/em\u003ewe aligned the nucleotide sequences of eight \u003cem\u003eC. lupus familiaris\u003c/em\u003e BRCA1 protein splice isoforms predicted from build canFam3 (1: XP_013971930.1; 2: XP_013971932.1; 3: XP_005624371.1; 4: XP_022278220.1; 5: XP_022278221.1; 6: XP_022278222.1; 7: XP_013971933.1; 8: XP_013971934.1; downloaded from https://www.ncbi.nlm.nih.gov/protein/) to this consensus sequence\u003cem\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrimers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePrimers were designed using Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). See \u003cstrong\u003eAdditional File 1\u003c/strong\u003e. Primers were ordered from Merck-Sigma-Aldrich (Gillingham, Dorset, UK) as \u0026lsquo;Pure and Simple Primers\u0026rsquo; (https://www.sigmaaldrich.com/GB/en/configurators/tube?product= puresimple), cartridge purified and lyophilised. Primers were resuspended in nuclease-free water as 100uM stocks and then diluted to 10uM working aliquots before use.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSynthetic Internal Tandem Duplication sequences\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eType I and Type II ITD synthetic oligos were ordered from Merck-Sigma-Aldrich as \u0026lsquo;Long Oligos\u0026rsquo; (https://www.sigmaaldrich.com/GB/en/configurators/tube?product=longoligo). 179 base pair sequences for both the sense and R+C antisense strands were ordered (5 OD yield and cartridge purification) based on the C2 and MS sequences (\u003cstrong\u003eAdditional File 1\u003c/strong\u003e). The MS Type II anti-sense strand could not be synthesised and so the Type II synthetic oligo had to be added to reactions single stranded. Both the C2 Type I sense and antisense sequence oligos could be synthesised, and these were annealed into a double-stranded target. The lyophilised sense and anti-sense Type I sequences were each resuspended at 100uM in TE (10mM Tris base, 1 mM EDTA, pH8.0) then 20ul of each oligo was added to 60ul of STE buffer (50mM NaCl, 10mM Tris base, 1 mM EDTA, pH8.0) and the mixture placed in a hot block at 100\u003csup\u003eo\u003c/sup\u003eC for 5 minutes, after which the block was allowed to cool overnight. The final concentration of the annealed oligo was taken as 20uM.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFFPE blocks of fifteen canine mast cell tumours with known \u003cem\u003eKIT\u003c/em\u003e exon 11 status (ten positive for an exon 11 ITD, five negative for an ITD) were provided by Finn Pathologists through MD. Exon 11 status was determined by the Michigan State University veterinary diagnostic laboratory (Lansing, Michigan, USA). All FFPE samples were archival tissue which had been previously used for routine diagnosis as part of standard clinical veterinary care. All were available for research use. See \u003cstrong\u003eAdditional File 2\u003c/strong\u003e for sample details.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHistological stains\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHaematoxylin and eosin (H\u0026amp;E) stains were carried out on dewaxed and rehydrated FFPE sections using standard protocols (reagents from Atom Scientific, Hyde, Cheshire, UK). Slides were dipped in xylene and then mounted with glass coverslips using DPX (Atom Scientific). Once dry, slides were scanned using an Olympus VS200 whole slide scanner (Evident Europe GmbH, Hamburg, Germany) at 20x.\u003c/p\u003e\n\u003cp\u003eThe relative amount of tumour compared to normal cells within each sample was approximated by measuring the total area of tissue and of tumour on the H\u0026amp;E sections at low power using QuPath 0.5.1 (https://qupath.github.io/). For caveats to this approach, see \u003cstrong\u003eDiscussion\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA extraction from FFPE sections\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDNA was extracted from 5 um FFPE sections either as cut scrolls in Eppendorf tubes or sections on slides (two to three sections for each extraction). Sections were dewaxed by washing twice in xylene for 5 minutes each and then twice in 100% ethanol for 5 minutes each. If isolating from scrolls, samples were centrifuged between washes to avoid loss of material. Sections on slides were simply transferred between washes. After the last ethanol wash, samples were air-dried. DNA was then extracted using the QIAamp DNA FFPE tissue kit (Qiagen, Manchester, UK), including the RNaseA step. DNA was eluted in 100ul nuclease-free water and quantified on a Nanodrop (Thermofisher, Horsham, Surrey, UK).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePCR conditions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll PCR reactions were prepared in a designated pre-PCR area in a Class II laminar flow hood, in a room separate from areas where template was prepared. Template was not handled until the prepared PCR reactions were taken out of the pre-PCR room. Template was added to reactions in a designated PCR laminar flow hood in a separate area.\u003c/p\u003e\n\u003cp\u003eStandard semi-quantitative PCR reactions were carried out using the GoTaq Green PCR system (Promega, Chilworth, Hampshire, UK). Reactions volumes were 50ul (10ul of 5X Promega EasyGreen Buffer, 2mM MgCl\u003csub\u003e2\u003c/sub\u003e, 0.2uM of each primer, 1ul of 10 mM PCR nucleotide mix, 0.2ul Promega GoTaq) (see also \u003cstrong\u003eAdditional File 1\u003c/strong\u003e). gDNA template input was typically 50ng (unless otherwise stated). A \u0026lsquo;No Template\u0026rsquo; control was included in all experiments. The reactions were cycled on an Applied Biosystems VeritiPro 96-well Thermal Cycler (Thermofisher) with the following program: 94\u003csup\u003eo\u003c/sup\u003eC 2 minutes; 35 cycles of 94\u003csup\u003eo\u003c/sup\u003eC 15 seconds, 58\u003csup\u003eo\u003c/sup\u003eC 30 seconds, 68\u003csup\u003eo\u003c/sup\u003eC 30 seconds; hold at 4\u003csup\u003eo\u003c/sup\u003eC (see also \u003cstrong\u003eAdditional File 1\u003c/strong\u003e). Amplified products were resolved on 2% agarose TBE (Tris base 90mM, Boric acid 90mM, EDTA 2mM) gel (all reagents from Thermofisher) with a 100bp DNA ladder (Promega).\u003c/p\u003e\n\u003cp\u003eQuantitative real time PCR on genomic DNA was carried out using a Quantstudio 7 system (Thermofisher) using Fast-96 well settings with SYBR Green reagents to determine Ct times and melt curves. Reaction volumes were 20ul consisting of 10ul 2x PowerUp SYBR Green Master Mix (Thermofisher), 1ul of each of 10uM forward and reverse primers, and the remaining 8ul either template, or template plus nuclease-free water (depending on the input material) or nuclease-free water alone for No Template controls. The amount of starting template input into each reaction varied depending on the experiment. Reactions were run as three or four replicates in 96-well plates in fast cycling mode with the following program (as detailed in the manufacturer\u0026rsquo;s instructions): hold stage: 50\u003csup\u003eo\u003c/sup\u003eC 2 minutes, 95\u003csup\u003eo\u003c/sup\u003eC 2 minutes; amplification stage: 40 cycles of 95\u003csup\u003eo\u003c/sup\u003eC 1 second, 60\u003csup\u003eo\u003c/sup\u003ec 30 seconds (ramping speed was 2.63\u003csup\u003eo\u003c/sup\u003eC/s up, 2.42\u003csup\u003eo\u003c/sup\u003eC down); melt curve stage, 95\u003csup\u003eo\u003c/sup\u003eC 15 seconds, 60\u003csup\u003eo\u003c/sup\u003eC 1 minute, ramping of 0.05\u003csup\u003eo\u003c/sup\u003eC/s to 95\u003csup\u003eo\u003c/sup\u003eC, hold at 95\u003csup\u003eo\u003c/sup\u003eC 15 seconds; hold at 4\u003csup\u003eo\u003c/sup\u003eC (see also \u003cstrong\u003eAdditional File 1\u003c/strong\u003e). This program took 60 minutes to run. Automatic thresholding of Ct values was used. In some cases, after the program had finished, the contents of representative wells were resolved on 2% agarose TBE gels to confirm product sizes.\u003c/p\u003e\n\u003cp\u003eIn most cases, Ct values for the amplification curves (the amplification cycle at which the amount of fluorescence, and hence product, crossed the detection threshold; dependent on the amount of starting template) and Tm values for the melt curves (the temperature at which the rate of change of fluorescence as the product of the reaction melted was at its maximum; dependent on the physico-chemical properties of the product at the end of the amplification) were determined automatically by the Quantstudio 7 software. In some cases where the melting curve had two distinct peaks, the Tm for both peaks could be determined automatically. However, in some cases where one peak was small, the Tm for that peak had to be determined by manual inspection of the fluorescence reporter derivative values to identify the temperature with the highest derivative reporter value (i.e. the highest rate of change).\u003c/p\u003e\n\u003cp\u003eFor all melt curve plots, the X-axis is the temperature (\u003csup\u003eo\u003c/sup\u003eC) and the Y-axis is the fluorescence reporter derivative. The Y-axis may be plotted on different scales between graphs where assays have been run on different plates. Except where otherwise stated, the fluorescence reporter derivative values across the melt curve temperature range were imported into GraphPad Prism 10.2.3 (GraphPad Software LLC, Boston MA, USA) for plotting.\u003c/p\u003e\n\u003cp\u003eFor standard curves of Ct values against DNA input using genomic MDCK DNA, the DNA content of an average canine diploid cell was estimated as ~5pg, based on the haploid length of the canine genome as ~2.4 Gb (34) and 978 Mb of DNA as having a mass of 1pg (35). Therefore, template input at 0.5pg, 5pg, 50pg, 500pg, 5ng or 50ng equated to approximately 0.2, 2, 20, 200, 2000 or 20,000 copies of the target per reaction, respectively.\u003c/p\u003e\n\u003cp\u003eFor standard curves of Ct values against DNA input using synthetic oligos, template was added to reactions at 3.3ul of 1fM (approx. 2000 molecules), 10fM (approx. 20,000 molecules), 100fM (approx. 200,000 molecules), 1pM (approx. 2,000,000 molecules) or 10pM (approx. 20,000,000 molecules) per reaction.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eKIT\u003c/em\u003e Exon 11 genomic DNA sequencing of cMCT clinical samples was carried out by PCR amplification of the whole of exon 11 from FFPE-extracted DNA samples using the whole exon 11 primers (\u003cstrong\u003eAdditional File 1\u003c/strong\u003e) and GoTaq PCR conditions. Products were analysed by gel electrophoresis on 2% agarose TBE gels. For wild-type tumours without an exon 11 ITD, only a single product was amplified and the PCR reactions could be directly purified using a PCR cleanup kit (Qiagen) and then sent for routine Sanger sequencing at Eurofins (Eurofins GmbH, Ebersburg, Germany) using the whole exon 11 primers.\u003c/p\u003e\n\u003cp\u003eFor tumours with an exon 11 ITD, multiple bands were generated, corresponding to both a wild-type and an ITD product. These products were separated by gel electrophoresis, the bands cut out and then extracted using a QIAquick gel extraction kit (Qiagen). Extracted bands were then sent for routine sanger sequencing at Eurofins.\u003c/p\u003e\n\u003cp\u003eWhere possible, both the genotype of the wild-type and ITD bands at the chr13:47940619 (C/T) and chr13:47940672 (T/C) SNPs, as well as the overall structure of the ITD in each sample, were determined. Technical difficulties of sequencing across repetitive DNA sequences, resulting in sequence outputs containing misaligned sequences (\u003cstrong\u003eAdditional File 3A\u003c/strong\u003e), meant the quality of sequence from ITD amplicons was variable, with regions where two different sequence traces were overlaid on each other. However, from the inspection of the sequencing chromatograms, and assessment of where the good quality sequence became poor quality sequence, it was possible to reconstruct the boundaries of the ITD regions, and therefore indirectly determine the sequence of the ITDs in most samples (see\u003cstrong\u003e Additional Files 3 \u003c/strong\u003eand\u003cstrong\u003e 4 \u003c/strong\u003efor a worked-through example of how this was carried out).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analysis (curve fitting and linear regressions) was carried out using GraphPad Prism 10.2.3.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eIdentification of the minimal commonly amplified region in mast cell tumours\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo develop our assay, we first aligned the duplicated regions from four MCTs and one mastocytoma cell line (C2) positive for \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs published by London \u003cem\u003eet al\u003c/em\u003e. (18) against the sequence of wild type \u003cem\u003eKIT\u003c/em\u003e exon 11, to identify whether there was a minimal region commonly duplicated across multiple tumours. This identified such a minimal region (UU_Cfam_GSD_1.0/canFam4 chr13:47,940,605 – 47,940,644) close to the 3’ splice donor site in exon 11, found in all five ITDs (\u003cstrong\u003eFigure 1A\u003c/strong\u003e). We called this the \u003cem\u003emin\u003c/em\u003e region. The \u003cem\u003emin\u003c/em\u003e region is flanked by variable regions which are included, to different extents, in the different ITDs (\u003cstrong\u003eFigure 1B\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eWe then designed a pair of primers, FminA and RminA, targeting the \u003cem\u003emin\u003c/em\u003e region and the 3’ variable region. The primers were designed to be head to head, with no intervening sequence (see \u003cstrong\u003eAdditional File 1\u003c/strong\u003e for details of all primer sets). The FminA site was wholly contained within the \u003cem\u003emin\u003c/em\u003e region while the RminA site was partly contained in the \u003cem\u003emin\u003c/em\u003e region, and partly in the 3’ variable region (\u003cstrong\u003eFigure 1B\u003c/strong\u003e). In a wild type, non-duplicated exon 11, the primer pair was predicted to create a small, 44bp, product. However, an internal tandem duplication would duplicate the FminA primer site and potentially the RminA site as well, resulting in the generation of new products of larger sizes. We hypothesised that the larger products would be distinguishable from the 44bp product in SYBR green-based qPCR melting curve analysis, and this would indicate the presence of an ITD. Importantly, we predicted that as long as the core \u003cem\u003emin\u003c/em\u003e region was included in a tandem duplication, then using this strategy the presence of any duplication in any tumour could be detected by the min primers. The presence of variable flanking regions, the exact sequence of those regions, and the exact structure of the ITD as a whole, would not affect performance of the assay.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eModelling this using the sequences from London \u003cem\u003eet al\u003c/em\u003e., we predicted that the ITDs in these tumours would result in minA ITD products with sizes of 92 bp (in the C2 line and MK tumour), 95 bp (in the BK tumour), 101 bp (in the MS tumour) and 112 bp (in the EH tumour) (\u003cstrong\u003eAdditional File 5\u003c/strong\u003e). We classified those sequences in which only the forward minA site was duplicated as ‘Type I’ duplications (C2 and MK) and those in which both forward and reverse minA sites were duplicated as ‘Type II’ duplications (BK, EH and MS).\u003c/p\u003e\n\u003cp\u003eAs an alternative approach, we designed a second set of primers, FminB and RminB, which flanked the minA primers (\u003cstrong\u003eFigure 1B\u003c/strong\u003e). In the wild type, these were predicted to generate an 88bp product, but ITDs were predicated to increase the distance between the forward and reverse priming sites, rather than duplicating them. Again, this should alter the melt curves in SYBR green-based qPCR. The predicted outcomes for the sequences from London \u003cem\u003eet al\u003c/em\u003e. with the minB primers were products of 136 bp (in the C2 line and MK tumour), 139 bp (in the BK tumour), 145 bp (in the MS tumour) and 156 bp (in the EH tumour) (\u003cstrong\u003eAdditional File 5\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eAs we were initially unsure how the minA primers would perform, whether or not all samples would in fact generate the expected 44bp \u003cem\u003eKIT\u003c/em\u003e exon 11 product and what factors might affect this, we also designed a pair of control primers (designated ‘Brca’ primers). These targeted the 5’ end of the large central exon of \u003cem\u003eBRCA1\u003c/em\u003e, which codes for the DNA binding site (\u003cstrong\u003eAdditional File 6\u003c/strong\u003e) (36). Like the minA primers, the Brca primers were a head-to-head primer set predicted to generate a 44bp product against which the performance of the minA primers could be compared. We also obtained previously described primers (37) for standard PCR amplification of the whole of \u003cem\u003eKIT\u003c/em\u003e exon 11 for use in electrophoretic sized-based analysis to use as a control for the detection of ITDs in clinical samples and for exon 11 genomic sequencing (\u003cstrong\u003eFigure 1B\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe minA, minB and Brca primer sets amplify discrete PCR products of expected sizes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test the minA, minB and Brca primer sets, we first carried out standard semi-quantitative PCR with gel-based analysis of amplified fragments, as well as qPCR, on genomic DNA (gDNA) from the MDCK canine epithelial cell line. With standard PCR and an input of 50ng gDNA template per reaction, faint bands could be observed on a 2% agarose gel at approximately the expected sizes (\u003cstrong\u003eFigure 2A\u003c/strong\u003e). Purification of the amplified fragments and their use as template in a second round of standard PCR reactions increased the product yield. Product sizes were consistent with those seen after amplification directly from the gDNA template and also consistent with the predicted sizes of 44bp for minA and Brca, and 88bp for minB (\u003cstrong\u003eFigure 2B\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eNext, we tested qPCR with an input of 50ng gDNA per reaction. All three primers amplified products with a single discrete peak in melting curve analysis, indicative of a single product being produced. The mean peak melting temperatures (mean Tm) for the minA, minB and Brca curves were 73.04±0.06\u003csup\u003eo\u003c/sup\u003eC, 78.04±0.06\u003csup\u003eo\u003c/sup\u003eC and 71.86±0.06\u003csup\u003eo\u003c/sup\u003eC (mean±SD, n=3) respectively, consistent with the minA and Brca products being of similar size and smaller than the minB product (\u003cstrong\u003eFigure 2C-G\u003c/strong\u003e). Finally, reaction mixes from representative wells were resolved on an agarose gel. The sizes of the products were consistent with the sizes seen in a standard PCR reaction (\u003cstrong\u003eFigure 2H\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eTherefore, using gDNA from a non-transformed canine epithelial cell line, the minA, minB and Brca primer sets generate discrete PCR products whose sizes are consistent with their predicted values.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs can be distinguished in qPCR by melt curve profiles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNext, the performance of the minA, minB and Brca primer sets was tested on synthetic \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs using two oligonucleotide targets based on the C2 cell line ITD (Type I amplification) and MS tumour sequence ITD (Type II amplification) (\u003cstrong\u003eAdditional File 5\u003c/strong\u003e). The primer sets were tested on these synthetic targets, as well as MDCK DNA, in both standard gel-based analysis and qPCR. 50ng of MDCK gDNA and 3.3ul of 10pM synthetic oligos were used as template for each standard PCR reaction. 5ng of MDCK gDNA and 3.3ul of 1pM synthetic oligos were used as template for each qPCR reaction.\u003c/p\u003e\n\u003cp\u003eIn standard PCR, with the MDCK template, the minA, minB and Brca primers gave products of sizes similar to those previously observed, consistent with the predicted 44bp, 88bp and 44bp sizes, respectively (\u003cstrong\u003eFigure 3A\u003c/strong\u003e). With the Type I synthetic oligo template, the minA primers generated two products, one identical in size to the MDCK minA product, the other running slightly below the 100 bp size marker, consistent with the predicted 92bp product (\u003cstrong\u003eFigure 3A\u003c/strong\u003e). The Type II template also generated one band identical in size to the MDCK minA product, consistent with the predicted 44bp product (\u003cstrong\u003eFigure 3A\u003c/strong\u003e). In addition, it generated a ‘ladder’ of larger products, the smallest of which was greater than 100bp, again consistent with the predicted size of the ITD but suggesting a cycle of self-annealing and amplification events by the PCR products themselves, resulting in the generation of new products with multiple copies of the duplication events (see\u0026nbsp;\u003cstrong\u003eDiscussion\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eWith the minB primers and the Type I and II templates, bands consistent in size with the predicted 136bp and 145bp products, respectively, were observed (\u003cstrong\u003eFigure 3A\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eAs expected, no product was observed with the Brca primers with either the Type I or Type II synthetic oligo template, while a band consistent in size with the predicted 44bp product was observed with MDCK template. No product was observed with any ‘No Template’ reaction (\u003cstrong\u003eFigure 3A\u003c/strong\u003e). Therefore, using a standard gel-based PCR approach, results from all primer sets were consistent with predictions apart from the product laddering observed with minA primers, which was unexpected.\u003c/p\u003e\n\u003cp\u003eWith the qPCR analysis (\u003cstrong\u003eFigure 3B-I\u003c/strong\u003e), for the minA primers there was a mean Ct value for the MDCK template reactions of 28.33±0.26 (mean±SD; n=3 replicate wells) (\u003cstrong\u003eFigure 3B\u003c/strong\u003e,\u003cstrong\u003eH\u003c/strong\u003e). As previously, a single peak was observed for the MDCK melting curve in each of the replicates, with a mean Tm of 72.67±0.06\u003csup\u003eo\u003c/sup\u003eC (mean±SD; n=3 replicate wells) (\u003cstrong\u003eFigure 3E,H\u003c/strong\u003e). qPCR of the Type I synthetic ITD using the minA primers resulted in a robust amplification with a mean Ct of 27.18±0.03 (mean±SD; n=3 replicate wells) (\u003cstrong\u003eFigure 3B,H\u003c/strong\u003e) and the production of two clear peaks on the melting curve. The first peak, with a Tm of 72.22±0.00\u003csup\u003eo\u003c/sup\u003eC (mean±SD; n=3 replicate wells), was consistent with the peak seen in the MDCK reactions (hereafter the ‘wild type’ or WT peak) but there was a second peak with a Tm of 78.00±0.06\u003csup\u003eo\u003c/sup\u003eC (mean±SD; n=3 replicate wells), consistent with a larger product (hereafter the ITD peak) (\u003cstrong\u003eFigure 3E,H\u003c/strong\u003e). With the Type II synthetic ITD and minA primers, the amplification had a mean Ct of 17.82±0.14 (mean±SD; n=3 replicate wells) (\u003cstrong\u003eFigure 3B,H\u003c/strong\u003e), suggesting the amplification reaction on the Type II ITD was much more robust than on wild type DNA or Type I ITD. Only a single melt curve peak was generated, with a Tm of 82.39±0.06\u003csup\u003eo\u003c/sup\u003eC (mean±SD; n=3 replicate wells) (\u003cstrong\u003eFigure 3E,H\u003c/strong\u003e), substantially higher than the WT peak. Therefore, in qPCR on a WT target, minA primers resulted in a single melting curve peak with a Tm in the region 72 – 73\u003csup\u003eo\u003c/sup\u003eC, but presence of an ITD resulted in a peak with a Tm \u0026gt;77\u003csup\u003eo\u003c/sup\u003eC, as well as potentially a WT peak, depending on the exact nature of the ITD.\u003c/p\u003e\n\u003cp\u003eWith the minB primers, a single melt curve peak was generated with MDCK template, the Type I ITD and Type II ITD, but the Tm shifted from 77.92±0.10\u003csup\u003eo\u003c/sup\u003eC with MDCK to 79.86±0.06\u003csup\u003eo\u003c/sup\u003eC with Type I ITD and 80.48±0.06\u003csup\u003eo\u003c/sup\u003eC with Type II ITD (mean±SD; n=3 replicate wells), consistent with the expected larger product sizes (\u003cstrong\u003eFigure 3C,F,H\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eAgain, as expected, no product was observed with the Brca primers in qPCR with either the Type I or Type II synthetic oligo template but the MDCK template gave a product with a single melting curve peak, with Ct and Tm values similar to that for the minA primers (\u003cstrong\u003eFigure 3D,G,H\u003c/strong\u003e), consistent with the 44bp Brca product size.\u003c/p\u003e\n\u003cp\u003eFinally, to confirm the products were of the expected sizes, 5ul of the reaction mix from one representative well of each reaction with template were analysed by gel electrophoresis after the qPCR data had been collected. Three wells from the minA/No Template reactions were also analysed (see below). The product sizes (\u003cstrong\u003eFigure 3I\u003c/strong\u003e) were consistent with those seen after standard PCR and gel electrophoresis, although ‘laddering’ in the minA reactions was now visible in the Type I template, and with the Type II template the laddering was so extreme it had become a smear on the gel.\u003c/p\u003e\n\u003cp\u003eIn one minA/No Template qPCR reaction, a very low abundance amplification event had occurred, with a product reaching the Ct threshold after 38.73 cycles. This product also had a single peak on the melting curve at a slightly lower Tm to the MDCK reactions of 71.83\u003csup\u003eo\u003c/sup\u003eC (\u003cstrong\u003eFigure 3E,H\u003c/strong\u003e). No such events were observed in the minB- or Brca/No Template reactions. Consistent with this, a small product could be seen on gel analysis of this No Template reaction (\u003cstrong\u003eFigure 3I\u003c/strong\u003e; boxed region). We reasoned that this was a background amplification event as a result of formation of ‘primer dimers’, low-affinity primer self-annealing events occurring in the absence of the correct high affinity target sequence (for example, ACAAA at positions 14 – 18 of the FminA primer could anneal to TGTTT at positions 8 – 12 of the FminB primer), followed by amplification to create a small product.\u003c/p\u003e\n\u003cp\u003eTo provide further evidence that primer self-annealing could occur in No Template controls, and that this was not a result of contamination, a qPCR plate was set up with multiple No Template controls. Twenty-four No Template wells were sealed in the pre-PCR area before MDCK template was added as a positive control. Five No Template wells were sealed after the MDCK template was added to the plate. The results show (\u003cstrong\u003eAdditional File 7\u003c/strong\u003e) that even in reactions sealed prior to any reasonable possibility of contamination, occasional low abundance amplification events, with Tm values for the melting curve of approximately 1\u003csup\u003eo\u003c/sup\u003eC lower than that of the expected peak of the WT exon 11 minA product, could occur, consistent with low affinity self-priming by the minA primers. Importantly, however, these events cannot be confused with detection of an exon 11 ITD and do not, therefore, give a false positive result.\u003c/p\u003e\n\u003cp\u003eTherefore, the minA and minB primer sets are capable of distinguishing \u003cem\u003eKIT\u003c/em\u003e exon 11 WT from \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD sequences on the basis of the characteristics of melt curve profiles (number and Tm of peaks) in SYBR Green qPCR.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStandard curves demonstrate linear amplification of WT and Type II ITD templates but not Type I templates\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNext, to better characterise the performance of the minA, minB and Brca PCR reactions on WT, Type I ITD and Type II ITD templates, a series of standard curve experiments were carried out (\u003cstrong\u003eFigure 4\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Additional File 8\u003c/strong\u003e). The results showed that for the WT target, amplification with all three primer sets was linear when the input was 50pg (approximately 20 target copies) per reaction or more, but below this input level amplification was non-linear and sometimes failed (\u003cstrong\u003eFigure 4A-C\u003c/strong\u003e). Notably, the slope of the standard curves was similar, showing that the efficiency of the amplification between the three different primer sets was similar. The melt curves for the MDCK WT target all showed a single peak at a melting temperature consistent with those previously observed (\u003cstrong\u003eAdditional File 8\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eFor the Type I ITD, the lower limit of detection with the minA primer set was an input of 3.3ul of 10fM synthetic Type I target but for minB was 3.3ul of 100fM target, suggesting that the minB reaction is less sensitive than the minA. Type I ITD amplification with both minA and minB was non-linear, but better fitted a hyperbolic curve (\u003cstrong\u003eFigure 4D\u003c/strong\u003e), however, given that with minB only the three highest target concentrations gave amplification, the minB data have to be interpreted with caution. For the Type II ITD, amplification with both the minA and minB primers was linear (\u003cstrong\u003eFigure 4E\u003c/strong\u003e), and the template could be detected with input levels as low as 3.3ul of 1fM synthetic template for both primer sets. Note that based purely on the number of targets calculated to be in each of the reactions, the assays appeared to be more sensitive on WT ‘native’ genomic DNA than on synthetic ITD oligos.\u003c/p\u003e\n\u003cp\u003eThe melt curves for the Type I ITD with both minA and minB, and for Type II with minB, were qualitatively and quantitatively highly similar at all levels of template input as long as this was above the limits of detection (\u003cstrong\u003eFigure 4F\u003c/strong\u003e) and similar to those seen previously. However, with the minA primers using the Type II ITD synthetic oligo template, the melt curve patterns changed depending on the amount of template input into the reaction (\u003cstrong\u003eFigure 4F\u003c/strong\u003e). At 3.3ul 1fM template per reaction, there was a low peak at the Tm associated with the WT product (consistent with that seen in an MDCK control reaction), but also a larger, apparently double, peak with a higher Tm, in the range associated with the ITD product. As the amount of input increased, the peak associated with the WT product disappeared, and the double peak associated with the ITD product merged into a single, clean peak, as previously observed for the Type II ITD (compare \u003cstrong\u003eFigure 4F\u003c/strong\u003e with \u003cstrong\u003eFigure 3E\u003c/strong\u003e), suggested altered reaction kinetics with different primer:template ratios.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe min assay detects KIT exon11 ITDs in FFPE samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test the assay in clinical samples, canine MCT FFPE samples which had been previously used for clinical diagnosis were supplied as FFPE blocks from the archives of Finn Pathologists (UK). For the purposes of this study, the samples from Finn Pathologists were coded as FP1 – 15. FP1 – 15 had been previously tested for \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs as part of routine diagnosis. FP4, FP7, FP9, FP10 and FP11 (n=5) were all reported as\u0026nbsp;\u003cem\u003eKIT\u003c/em\u003e exon 11 WT, while FP1, FP2, FP3, FP5, FP6, FP8, FP12, FP13, FP14 and FP15 (n=10) were all positive for\u0026nbsp;\u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs.\u003c/p\u003e\n\u003cp\u003eSections from the fifteen samples were first H\u0026amp;E stained to confirm the presence of tumour material (\u003cstrong\u003eAdditional File 9A,B\u003c/strong\u003e). The relative proportion of tumour to non-tumour material in each sample was approximated by measuring the total area of tissue in each section and the approximate area of tumour in each section (\u003cstrong\u003eAdditional File 9C\u003c/strong\u003e; see\u0026nbsp;\u003cstrong\u003eDiscussion\u003c/strong\u003e for limitations). To confirm exon 11 status, genomic DNA was isolated from the FP samples and standard PCR using the whole exon 11 primers was carried out, with the products visualised on a gel. Samples were analysed twice, to assess reproducibility (\u003cstrong\u003eAdditional File 10\u003c/strong\u003e). A single band of the expected size (~190bp) was generated for the five exon 11 WT samples in the first repeat (\u003cstrong\u003eAdditional File 10A\u003c/strong\u003e), as expected. In the second repeat of the FP sample analysis, FP11 failed to amplify but the others again showed a single band (\u003cstrong\u003eAdditional File 10B\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eIn the first repeat, the known exon 11 ITD samples all generated one band at the WT size and one or more larger bands, in the range 220 to 300 bp (\u003cstrong\u003eAdditional File 10A\u003c/strong\u003e), consistent with the previously observed results for synthetic exon 11 ITDs, although the FP12 ITD band was faint. In the second repeat, the results were similar, except the FP1 and FP13 products were faint, and the ITD product could not be distinguished in FP12 (\u003cstrong\u003eAdditional File 10B\u003c/strong\u003e). These results are broadly consistent with what was known of the FP samples \u003cem\u003ea priori\u003c/em\u003e, but do show an inherent variability in the results of a standard PCR / gel analysis approach to assessing \u003cem\u003eKIT\u003c/em\u003e exon 11 status, at least using the whole exon 11 primer set and PCR conditions described here.\u003c/p\u003e\n\u003cp\u003eDNA extracted from the blocks was then tested with the minA/minB/Brca qPCR assays (details of Ct and Tm values for these assays are provided in\u0026nbsp;\u003cstrong\u003eAdditional File 11\u003c/strong\u003e). The Brca qPCR assay gave a single peak in the melting curve analysis in the expected Tm range (71.47to 71.94\u003csup\u003eo\u003c/sup\u003eC) for all samples (\u003cstrong\u003eAdditional File 12\u003c/strong\u003e). Ct values varied from 25.44±0.13 (FP3) to 34.59±0.60 (FP11) despite nominally equal template input of 25ng/reaction.\u003c/p\u003e\n\u003cp\u003eIn the minA assay, Ct values ranged from 23.71±0.15 (FP2) to 37.45±0.29 (FP11). WT samples gave a single melt curve peak between 72.04 and 73.34±\u003csup\u003eo\u003c/sup\u003eC, comparable to the single peak seen with MDCK gDNA (\u003cstrong\u003eFigure 5A,B\u003c/strong\u003e). In contrast, samples with a known \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD showed both a WT peak and an ITD peak between 77.57 and 79.77\u003csup\u003eo\u003c/sup\u003eC (\u003cstrong\u003eFigure 5C\u003c/strong\u003e). To quantify the strength of the ITD signal, the ratio of the maximum value of the ITD peak (‘ITD peak height’) to the maximum value of the WT peak (‘WT peak height’) was calculated for each sample. This ranged from 1.72±0.02 (mean±SD; n=3) in FP2 (strongest ITD signal) to 0.11±0.01 (mean±SD; n=3) in FP6 (weakest ITD signal).\u003c/p\u003e\n\u003cp\u003eWith the minB qPCR assay, Ct values ranged from 24.42±0.06 (FP2) to 30.60±0.16 (FP11). WT samples again gave a single peak comparable to MDCK gDNA (Tm 77.10 to 78.04\u003csup\u003eo\u003c/sup\u003eC), (\u003cstrong\u003eFigure 6A,B\u003c/strong\u003e). With the ITD samples, FP5, FP6, FP13, FP14, showed two peaks to the melt-curve, one corresponding to the WT peak, the other peaking at 2 – 3\u003csup\u003eo\u003c/sup\u003eC higher. FP1, FP2, FP3, FP8, FP15 did not show a distinct second peak but did have a ‘shoulder’ to the melt curve (\u003cstrong\u003eFigure 6C).\u003c/strong\u003e minB ITD peaks were not distinct enough from WT peaks to enable relative peak heights to be reliably quantified.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSequencing and genotyping of \u003cem\u003eKIT\u003c/em\u003e exon 11 from FP samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe FminA primer annealing site contains a C/T polymorphism (\u003cem\u003eC.fam\u003c/em\u003e 13: 47,940,619 C/T; hereafter, SNP 619), as does the RminB site (\u003cem\u003eC.fam\u003c/em\u003e 13: 47,940,672 C/T; hereafter, SNP 672). It was possible that variation here could affect the performance of the assay (strength of the ITD peak signal) across samples depending on their genotype. To test this, we first sequenced and genotyped \u003cem\u003eKIT\u0026nbsp;\u003c/em\u003eexon 11 from the FP samples by Sanger sequencing of fragments generated by whole exon 11 PCR. For the ITD samples, amplicons corresponding to WT and ITD exons were isolated by gel purification and sequenced separately.\u003c/p\u003e\n\u003cp\u003eFor all WT samples (both WT amplification products from WT tumours and WT amplification products from tumours with ITDs), the sequence was identical to the reference genomic sequence except for the variations at the SNP 619 and SNP 672 polymorphisms. A consensus sequence for all WT fragments and the genotypes at SNP 619 and SNP 672 are shown in\u0026nbsp;\u003cstrong\u003eAdditional File 13\u0026nbsp;\u003c/strong\u003e(FASTA sequences in\u0026nbsp;\u003cstrong\u003eAdditional File 14\u003c/strong\u003e)\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eNote the sequence of the WT band represents the germline genotype.\u003c/p\u003e\n\u003cp\u003eThe structure of the duplication could be reconstructed from the sequencing data (see\u0026nbsp;\u003cstrong\u003eMaterials and Methods\u003c/strong\u003e) in nine of the ten ITD tumour samples (\u003cstrong\u003eAdditional File 15\u003c/strong\u003e). Three tumours were Type I, in which only the minA forward primer region had been duplicated. Five tumours were Type II, in which both the minA forward and reverse primer binding sites were duplicated. In one tumour (FP14) the ITD was a new variant in which the 5’ limit of the duplication region occurred within the forward minA primer site and resulted in its disruption, but at the same time created a new FminA priming site in a 5’ location, as well as duplicating the RminA primer site. Thus, similar product sizes would be generated to the Type I duplication, but from a different configuration. This ITD type was termed Type III.\u003c/p\u003e\n\u003cp\u003eITD genotypes, and for comparison the germline genotype (genotype of the WT band), are shown in\u0026nbsp;\u003cstrong\u003eAdditional File 15B\u003c/strong\u003e. It was possible to directly genotype nine of the ten ITD samples at SNP 619 and four of the samples at SNP 672. For one sample, FP12, which was germline homozygous C/C at SNP 619, it could be assumed that the ITD was ‘C’ as well. For three samples in which the WT SNP 672 genotype was C/C, it could be assumed the SNP 672 genotype of the ITD was C as well. In two samples (FP2 and FP12), which were C/T at WT SNP 672, the genotype of the ITD could not be determined (due to sequence quality. FP6 gave ambiguous results at SNP 672, with sequencing from the forward primer suggesting C/A at SNP 672, but the reverse primer sequencing suggesting C (see limitations in\u0026nbsp;\u003cstrong\u003eDiscussion\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eOverall, seven out of ten samples with \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs were C/C homozygous at SNP 619 in the WT band, and C in the ITD. Two samples (FP5 and FP8) were T/T homozygous in the WT bands and correspondingly T in the ITD bands. FP6, which was C/T heterozygous at SNP 619 in the WT, underwent Loss-of-Heterozygousity (LOH) to give a T allele in the ITD (\u003cstrong\u003eAdditional File 16\u003c/strong\u003e). Similarly, at SNP 672, FP1 underwent LOH in the ITD compared to the WT sequence (\u003cstrong\u003eAdditional File 16\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle Nucleotide Polymorphisms in the FminA and RminB primer annealing sites affect minA/minB assay signal strength\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe FminA primer has a C at the position equivalent to SNP 619. Therefore, any tumour which is T at this position in the ITD, such as FP6, may have suboptimal detection of the ITD by the minA assay as the FminA primer may bind to its target site with reduced affinity. A similar effect may happen with the RminB primer in a tumour with a SNP 672 ‘C’ allele, as the RminB primer has A at the position equivalent to SNP 672.\u003c/p\u003e\n\u003cp\u003eTo test whether optimising primer sequence to the genotype of the ITD would improve the assay, the performance of the FminA primer (CTTCCTTA\u003cu\u003eC\u003c/u\u003eGATCACAAATGG) was compared to an alternative primer (FminA-ALT) with T at the equivalent of the SNP 619 position (CTTCCTTA\u003cu\u003eT\u003c/u\u003eGATCACAAATGG). The two primers were tested on SNP 619 C/T germline heterozygous or T/T germline homozygous WT (FP4, FP9, FP10) or ITD (FP5, FP6, FP8) tumours, as well in MDCK cells (25ng input per reaction). The results (\u003cstrong\u003eFigure 7A\u003c/strong\u003e) demonstrated that detection of the ITD in FP5, FP6 and FP8, all of which carry a SNP 619 T allele ITD, was substantially improved with the minA-ALT primers (mean±SD\u0026nbsp;ITD/WT peak ratios, n = 3: FP5 minA: 0.543±0.044, minA-ALT: 1.659±0.065; FP6 minA: 0.193±0.049, minA-ALT: 2.980±0.068; FP8 minA: 0.267±0.021, minA-ALT: 1.458±0.050;\u0026nbsp;\u003cstrong\u003eAdditional File 11\u003c/strong\u003e). Qualitatively, there was little difference between the melt curves for minA and minA-ALT in tumours without an ITD. It was also noted that in this assay, the minA assay on FP9 produced a small peak at a Tm similar to that of an ITD in two of three replicates. Given that all previous assays and \u003cem\u003ea priori\u003c/em\u003e testing on FP9 had shown that this tumour was WT, and that there was no indication of an ITD peak with the minA-ALT primers (despite being a T/T genotype and therefore predicted to give a stronger ITD signal with minA-ALT), these small peaks must represent false positive ITD peaks (see further discussion on analytical sensitivity and specificity of the assay below in the\u0026nbsp;\u003cstrong\u003eDiscussion\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eNext, a direct comparison was carried out between the minB assay using the original primers, in which the reverse primer sequence corresponds to the SNP 672 T allele (CCCCT\u003cu\u003eA\u003c/u\u003eTTTCATACTGACCAAAG) and the minB assay using a reverse primer (RminB-ALT) in which the sequence corresponded to the SNP 672 C allele (CCCCT\u003cu\u003eG\u003c/u\u003eTTTCATACTGACCAAAG).\u0026nbsp;One WT tumour and six ITD tumours, as well as MDCK DNA, were tested.\u003c/p\u003e\n\u003cp\u003eWith the WT tumour and MDCK DNA, there was little difference in the single WT peak with either the minB or minB-ALT primer set. With the ITD tumours, there was also little difference in strength of detection of the ITD peak, irrespective of primer set. However, in one tumour which was germline C/T but C in the ITD (FP1), the RminB-ALT primer gave a slightly stronger detection. In FP2, which was germline C/T but the ITD could not be genotyped, detection was much stronger with RminB-ALT, consistent with the ITD being C at SNP 672 (\u003cstrong\u003eAdditional File 17\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eTherefore, use of primers sets corresponding to the genotype of a \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD substantially improves ITD detection, in particular with the minA/minA-ALT assay.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe minA/minA-ALT assay for \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs is quantitative\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether there was a correlation between the strength of the ITD signal and the amount of tumour material in each sample, the ITD peak/WT peak ratios (\u003cstrong\u003eAdditional File 11\u003c/strong\u003e) were correlated with the ratio of tumour to total tissue in each sample, defined from the H\u0026amp;E slides (\u003cstrong\u003eAdditional File 9\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eFirst, a simple linear regression analysis was carried out using the ITD peak/WT peak ratios derived from the minA analysis of tumours with known \u003cem\u003eKIT\u003c/em\u003e exon 11 duplications in \u003cstrong\u003eFigure 5\u003c/strong\u003e. This showed no correlation between the strength of the ITD signal and amount of tumour in the sample (R\u003csup\u003e2\u003c/sup\u003e=0.002; P=0.898; \u003cstrong\u003eFigure 7B\u003c/strong\u003e). Next, for the three tumours (FP5, FP6 and FP8) with a SNP 619 T genotype ITD and for which the minA-ALT assay substantially enhanced performance of the assay (\u003cstrong\u003eFigure 7A\u003c/strong\u003e), the ITD peak/WT peak ratios for the minA-ALT assay shown in \u003cstrong\u003eFigure 7A\u0026nbsp;\u003c/strong\u003ewere substituted for the minA ratios in the analysis. The correlation between ITD peak/WT peak ratio and amount of tumour in the starting material was now substantially improved although still did not reach significance (R\u003csup\u003e2\u003c/sup\u003e=0.316; P=0.091; \u003cstrong\u003eFigure 7C\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eInspection of the linear regression results suggested that FP12 was an outlier with a relatively low ITD/WT peak ratio (0.30±0.05) considering the tumour:total tissue ratio calculated for this sample (0.719; second-highest of all ITD samples). To test this, the analysis was repeated, this time using the minA-ALT assay results for FP5, FP6 and FP8, and excluding FP12. This now demonstrated a significant correlation between the ITD/WT peak ratio signal and the amount of tumour material (R\u003csup\u003e2\u003c/sup\u003e=0.727; P=0.004; \u003cstrong\u003eFigure 7D\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eTherefore, providing the primer set corresponding to genotype of the ITD is used, determining the ITD/WT peak ratios from the minA/minA-ALT assay can provide a quantitative assessment of the proportion of ITD positive tumour material in a sample. For further comment on FP12, and the reasons for possible exceptions, see\u0026nbsp;\u003cstrong\u003eDiscussion.\u003c/strong\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eHere we describe two different strategies for detecting \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs in canine cutaneous Mast Cell Tumours, based around identification of a minimally common amplified region, and either detecting duplication of primer sites within that region (the minA/minA-ALT primers sets) or detecting the expansion of that region (the minB/minB-ALT primer sets). We have found that the small difference in size between WT and ITD products for the minB/minB-ALT primer sets make the melt curves for the respective products difficult to distinguish and they cannot be easily separately quantified. In contrast, the large difference in product sizes with the minA/minA-ALT assay (44bp WT; \u0026gt;90bp ITD) makes the products easily distinguished and enables the relative height of the melt curve peaks (equivalent to the greatest rate of change of fluorescence during the melt curve analysis phase) to be used as surrogate for the relative abundance of \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD gene copies, and therefore \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD positive cells, within the sample.\u003c/p\u003e\n\u003cp\u003eWe therefore recommend use of the minA and minA-ALT primer pairs for detection of a KIT exon 11 ITD in a cMCT. Each sample should be tested with at least three replicates of each primer pair (in a 96-well plate, given that one column on the plate is eight wells, using one column per sample with four minA replicates and four minA-ALT replicates is convenient) and ideally each reaction should have 25ng of genomic DNA as template.\u003c/p\u003e\n\u003cp\u003eThe presence/absence of a minA/minA-ALT ITD peak determines whether or not a sample is classified as having an ITD. The accuracy of the test in identifying samples with (sensitivity) or without (specificity) a \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD on this categorical basis is important and can be considered in two ways. Assuming a sample with unknown exon 11 status presenting for analysis is also of unknown genotype, then that sample would be tested with both minA and minA-ALT primers, with at least three replicate reactions for each primer set. The melt curve profiles for all replicates from both primer sets would then be interpreted holistically to determine whether or not an ITD was present. On this basis, in this study our test correctly identified 5 out of 5 known wild type samples (in addition to MDCK cells) and 10 out of 10 samples known to have exon 11 ITDs. For this small sample number, therefore, analytical specificity and sensitivity were 100%. However, one could also consider sensitivity and specificity on a reaction by reaction basis, as any potential false positive / false negative events are likely to be due to random mis-annealing events occurring in a reaction well, rather than something affecting the whole sample. This is a more stringent assessment of specificity / sensitivity. Taking this approach, and considering any false negative / false positive events occurring in the replicates in the data presented here for (i) only minA reactions (because of the numbers available), (ii) only reactions with a minimum input of 25ng template, (iii) only MDCK or FFPE DNA template reactions, then of 39 individual replicate reactions on known exon 11 ITD samples, 39 produced melt curves which correctly identified the sample as an ITD, an analytical sensitivity of 100%. Of 36 individual replicate reactions on known wild type samples, 34 produced melt curves which correctly identified the sample as WT, an analytical specificity of 94%. Given that the presence of two replicates in a known WT sample with low ITD false positive peaks (FP9 in \u003cstrong\u003eFigure 7A\u003c/strong\u003e) is not consistent either with the third replicate from that sample or the complete lack of any ITD peak in the matched set of minA-ALT reactions, if this was a sample of unknown status \u003cem\u003ea priori\u003c/em\u003e then this result should automatically be treated with caution and trigger a retest. It is therefore unlikely that this sample would be reported back as positive. Future studies comparing samples with known \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD status will be needed to increase numbers and refine these data and reporting criteria.\u003c/p\u003e\n\u003cp\u003eSeparately from the categorical detection of the presence of a \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD in a sample, is the ability of the minA/minA-ALT assay to quantify the relative abundance of cells carrying a \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD, compared to wild-type cells, in a sample. It is likely that any cMCT sample contains a mixture of non-transformed and transformed cells (38, 39). Non-transformed cells may include regions of epidermis or subcutaneous tissue included as part of a wide excision, or they may by fibroblasts or leukocytes admixed with the neoplastic cells as part of the tumour mass. The neoplastic cells may or may not contain a \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD. Even if they do contain an ITD, that duplication will likely only occur on one copy of the gene, not both, as we have demonstrated here by finding LOH in two tumour samples. Therefore, even in a sample of pure neoplastic mast cells from a cMCT with an ITD, if this were possible, 50% of \u003cem\u003eKIT\u003c/em\u003e gene copies will contain an ITD and 50% will be wild type. In reality, in a sample with a mixture of non-transformed and transformed cells, the proportion of \u003cem\u003eKIT\u003c/em\u003e gene copies with an ITD will be decreased still further. If a sample consists of 50% transformed cells and 50% non-transformed cells, then only 25% of \u003cem\u003eKIT\u003c/em\u003e gene copies in the sample will contain an ITD. Furthermore, these arguments assume that the \u003cem\u003eKIT\u003c/em\u003e ITD is a founder mutation in the tumour and that all cells in the tumour contain the ITD. It is likely that a tumour could contain multiple clones and that development of a KIT ITD may be a late event in tumour formation, only occurring in a small clonal population within the tumour bulk. In such a situation, the proportion of ITD to WT \u003cem\u003eKIT\u003c/em\u003e gene copies will be diluted still further.\u003c/p\u003e\n\u003cp\u003eAn additional complication, as demonstrated by amplification from the synthetic ITD oligos (\u003cstrong\u003eFigure 3\u003c/strong\u003e and \u003cstrong\u003eFigure 4\u003c/strong\u003e), is that even a \u0026lsquo;pure\u0026rsquo; ITD template can give both WT and ITD products, depending on the exact structure of the duplication and amount of template input. Furthermore, the presence of the duplication appeared to lead to self-priming and concatamerisation of PCR products, as shown by \u0026lsquo;laddering\u0026rsquo; or \u0026lsquo;smears\u0026rsquo; when products were analysed on DNA gels. Overall, therefore, there are too many variables to be confident of using the results to provide exact numbers of neoplastic cells present in a sample. However, it is clear that height of the ITD melt curve peak relative to the height of the WT peak provides relative quantitative information about the proportion of \u003cem\u003eKIT\u003c/em\u003e gene copies with ITD in a sample relative to WT copies. Indeed, with the exception of sample FP12, the ITD/WT peak ratio in the FFPE samples was significantly positively correlated with the amount of tumour in that sample as a proportion of the overall amount of tissue (\u003cstrong\u003eFigure 7D\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eConsidering that the FP12 sample, which had been externally validated as carrying an ITD, was composed of \u0026gt;70% tumour mass (although the mass includes transformed and non-transformed cells), the low ITD/WT peak ratio in that sample was surprising (\u003cstrong\u003eFigure 7D\u003c/strong\u003e). Other samples with a high proportion of tumour to total tissue (e.g. FP2, FP5, FP6) (\u003cstrong\u003eAdditional File 9C\u003c/strong\u003e) had substantially higher ITD/WT peak ratios. Furthermore, amplification of the ITD in FP12 by standard PCR across the whole of \u003cem\u003eKIT\u003c/em\u003e exon 11 and subsequent gel electrophoresis was weak in the first repeat and not successful in the second (\u003cstrong\u003eAdditional File 10\u003c/strong\u003e). The results for FP12 are consistent with a model in which the \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD population in this tumour is a small neoplastic subclone within the tumour as a whole. Further work will be needed to fully explore the significance of such populations for therapy and whether using our assay and the ITD/WT peak ratio will be a way of identifying them in tumours in the future. To facilitate this, it would be helpful to interpret the results of future minA/minA-ALT assays alongside H\u0026amp;E images of the samples used for analysis. Given that cMCTs with, or without, an ITD can respond differently to TKIs (24, 27), knowing whether or not an ITD positive population is likely to form a major or a minor clone within a tumour may help to guide clinical management decisions.\u003c/p\u003e\n\u003cp\u003eWe chose to develop our \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD assay using a SYBR Green qPCR approach. Standard PCR approaches with gel-based analysis are good at identifying size differences in a PCR product (e.g. the difference between a wild-type template and one containing a duplication) but their sensitivity is lower compared to real-time fluorescence-based approaches, they are, at best, semi-quantitative (40, 41). Probe-based fluorescence approaches, such as digital droplet PCR or probe-based qPCR are sensitive and quantitative, but while a ddPCR/qPCR probe may be designed and optimized for one specific \u003cem\u003eKIT\u003c/em\u003e exon 11 ITD, there is no guarantee it would be universally applicable to all \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs.\u003c/p\u003e\n\u003cp\u003eHowever, SYBR green-based qPCR is sensitive and quantitative and does not rely on a specific probe site to be present in all PCR products. Providing the chosen primer site is capable of amplifying the DNA, the exact sequence of the DNA product (other than the primer binding sites) is irrelevant. SYBR green qPCR is thus likely to work with a broader range of\u003cem\u003e KIT\u003c/em\u003e exon 11 ITDs than a probe-based approach. Finally, SYBR green-based qPCR can, at the end of the assay, generate a melting curve analysis of the PCR product. The shape of this curve, and the temperature at which it peaks (Tm), are dependent of the physico-chemical properties of the product(s), specifically its length and GC content. Therefore, if there is a sufficient size difference between two products which may be amplified by a pair of primers in SYBR green-based qPCR assay, then the presence of the two products should be detectable by differences in the melt curves and the Tm. This is normally used in gene expression analysis to determine whether or not a set of primers is amplifying solely its intended target. However, we reasoned we could exploit melting curve analysis of qPCR products from genomic DNA template to determine whether or not a sample contained a\u003cem\u003e KIT\u003c/em\u003e exon 11 ITD and indeed this proved to be the case.\u003c/p\u003e\n\u003cp\u003eOur SYBR green-based qPCR melting curve assay has the advantages of simplicity and low cost. It uses widely available reagents and the primers require no special modifications. Melting curve analysis can be carried out by many, if not all, real time PCR machines. Its power lies in the specific primer sequence targeting the common minimally amplified region.\u003c/p\u003e\n\u003cp\u003eOur study has limitations. Tumour area measurements relative to total tissue (\u003cstrong\u003eAdditional File 9C\u003c/strong\u003e) for correlation with ITD/WT peak ratios were carried out at low power by manually drawing around the whole tissue section and the region of the tumour within that. We have not accounted for the presence of non-neoplastic cells within the tumour itself. We have not adjusted for potential differences in necrosis between the tumours which could affect the total cellularity. Furthermore, two tumours (FP8 and FP13) had particularly infiltrative, as opposed to expansile, growth patterns, which made accurate tracing of their borders problematic. Given these issues, it is remarkable that there is (once FP12 is excluded) a significant correlation between the calculated tumour to total tissue ratio and the ITD/WT peak ratio.\u003c/p\u003e\n\u003cp\u003eThe minA primer set could generate signals from primer dimers in No Template control reactions. The Tm value for the melt curve peak from minA primer dimers was very similar to the melt curve peak for the WT exon 11 product. However, this issue could not lead to a false negative signal if a sample contained an ITD as the template would outcompete any primer dimer formation.\u003c/p\u003e\n\u003cp\u003eAn additional limitation of the study was the use of standard Taq polymerase for amplification of DNA fragments for sequencing and genotyping, which could have introduced sequence errors as the Taq was not an error-free proofreading polymerase. However, the size of the amplicons was relatively small, making the insertion of mutations unlikely, although not impossible. Furthermore, in all cases the DNA sequence of the amplicons was identical to the publicly available sequence builds, the only variations being at the sites of known SNPs, and those variations were known alleles, with one exception. The exception was the results of sequencing the FP6 ITD at SNP 672, which gave a genotype of C with the reverse sequencing primer (consistent with the WT germline sequence) but gave C/A heterozygousity with the forward sequencing primer. It is unclear whether this is a PCR amplification error or sequencing error. As the forward and reverse sequencing reactions were carried out on the same amplified fragment, the latter seems more likely. Overall, therefore, we consider the sequencing and genotyping data reliable.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eWe have developed a novel, rapid, qPCR-based assay for detecting the presence of \u003cem\u003eKIT\u003c/em\u003e exon 11 internal tandem duplications in canine cutaneous Mast Cell Tumours. The assay uses PCR-product melt curve analysis to detect the duplications with high sensitivity and specificity. The ratio of the height of the ITD peak to the WT peak in the melt curve is a measure of the abundance of the ITD in the sample. The assay has potential for use in a wide range of samples, including non-surgical samples of limited material such as fine needle aspirates or blood samples containing cell-free tumour DNA.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eFFPE, formalin fixed paraffin embedded\u003c/p\u003e\n\u003cp\u003egDNA, genomic DNA\u003c/p\u003e\n\u003cp\u003eH\u0026amp;E, Haematoxylin and eosin\u003c/p\u003e\n\u003cp\u003eITD, internal tandem duplication\u003c/p\u003e\n\u003cp\u003eLOH, Loss of Heterozygousity\u003c/p\u003e\n\u003cp\u003eMCT, mast cell tumour\u003c/p\u003e\n\u003cp\u003eMDCK cells, Madin-Darby canine kidney cells\u003c/p\u003e\n\u003cp\u003eqPCR, quantitative real-time PCR\u003c/p\u003e\n\u003cp\u003eTBE, Tris base (90mM) Boric acid (90mM) EDTA (2mM)\u003c/p\u003e\n\u003cp\u003eWT, wild type\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Bristol University/Bristol Veterinary School Animal Welfare and Ethics Review Body (VIN number 22-041). Supply of archival diagnostic samples by Finn Pathologists (Harleston, Norfolk) was also approved by the CVS internal ethical review panel.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study is subject to a patent filing in the UK (application reference 2602035.4) and the min assay has been licenced for commercial use to CVS/Axiom. CVS group supplied samples for use in the study but had no control over the direction of the study, the analysis or the decision to publish.. Sam Beck is Director of Independent Anatomic Pathology Ltd. Melanie Dobromylskyj is a full-time employee of The Veterinary Pathology Group. Langford Vets is a business separate to Bristol Veterinary School.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions (CRediT Taxonomy)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization \u0026ndash; MJS, GT; Data curation \u0026ndash; MJS; Formal analysis \u0026ndash; MJS, LSH, JOE, GT; Funding acquisition \u0026ndash; MJS; Investigation \u0026ndash; MJS, LSH, JOE, GT; Methodology \u0026ndash; MJS, JOE, GT; Project administration \u0026ndash; MJS; Resources \u0026ndash; SB, MD, GE; Supervision \u0026ndash; MJS; Visualization \u0026ndash; MJS; Writing \u0026ndash; original draft \u0026ndash; MJS; Writing \u0026ndash; review \u0026amp; editing \u0026ndash; MJS, JOE, LSH, GT, SB, GE, MD.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by a Cardiff University Integrated Master\u0026rsquo;s student project, a BBSRC Impact Acceleration Account award through Cardiff University and Smalley laboratory discretionary funds. Neither Cardiff University, Finn Pathologists, AML or Bristol University/Langford Vets had any role in the design, analysis and reporting of the study (other than the roles of the individual authors). The Cardiff University Technology Transfer Office have been involved in the development of foreground intellectual property arising from the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the pathology team at Finn Pathologists, Harleston, Norfolk, UK for supply of archival diagnostic tissue.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have read the manuscript and consented to its publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData sharing is not applicable to this article as no datasets were generated or analysed during the current study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDobson JM, Samuel S, Milstein H, Rogers K, Wood JL (2002) Canine neoplasia in the UK: estimates of incidence rates from a population of insured dogs. J Small Anim Pract 43(6):240\u0026ndash;246\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeuten DJ (2017) Tumors in domestic animals. Fifth edition. ed. Ames, Iowa: Wiley/Blackwell; viii, 989 pages p\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBostock DE (1986) Neoplasms of the skin and subcutaneous tissues in dogs and cats. Br Vet J 142(1):1\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrodey RS (1970) Canine and feline neoplasia. Adv Vet Sci Comp Med 14:309\u0026ndash;354\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaldrymidou H, Leontides L, Koutinas AF, Saridomichelakis MN, Karayannopoulou M (2002) Prevalence, distribution and factors associated with the presence and the potential for malignancy of cutaneous neoplasms in 174 dogs admitted to a clinic in northern Greece. J Vet Med Physiol Pathol Clin Med 49(2):87\u0026ndash;91\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacy DW (1985) Canine mast cell tumors. Vet Clin North Am Small Anim Pract 15(4):783\u0026ndash;803\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO'Keefe DA (1990) Canine mast cell tumors. Vet Clin North Am Small Anim Pract 20(4):1105\u0026ndash;1115\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRothwell TL, Howlett CR, Middleton DJ, Griffiths DA, Duff BC (1987) Skin neoplasms of dogs in Sydney. Aust Vet J 64(6):161\u0026ndash;164\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWelle MM, Bley CR, Howard J, Rufenacht S (2008) Canine mast cell tumours: a review of the pathogenesis, clinical features, pathology and treatment. Vet Dermatol 19(6):321\u0026ndash;339\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThamm DH, Avery AC, Berlato D, Bulman-Fleming J, Clifford CA, Hershey AE et al (2019) Prognostic and predictive significance of KIT protein expression and c-kit gene mutation in canine cutaneous mast cell tumours: A consensus of the Oncology-Pathology Working Group. Vet Comp Oncol 17(4):451\u0026ndash;455\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChoi YJ, Yoo JS, Jung K, Rice L, Kim D, Zlojutro V et al (2023) Lung-specific MCEMP1 functions as an adaptor for KIT to promote SCF-mediated mast cell proliferation. Nat Commun 14(1):2045\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranke K, Kirchner M, Mertins P, Zuberbier T, Babina M (2022) The SCF/KIT axis in human mast cells: Capicua acts as potent KIT repressor and ERK predominates PI3K. Allergy 77(11):3337\u0026ndash;3349\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGilfillan AM, Beaven MA (2011) Regulation of mast cell responses in health and disease. Crit Rev Immunol 31(6):475\u0026ndash;529\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGilfillan AM, Tkaczyk C (2006) Integrated signalling pathways for mast-cell activation. Nat Rev Immunol 6(3):218\u0026ndash;230\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDowning S, Chien MB, Kass PH, Moore PE, London CA (2002) Prevalence and importance of internal tandem duplications in exons 11 and 12 of c-kit in mast cell tumors of dogs. Am J Vet Res 63(12):1718\u0026ndash;1723\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones CL, Grahn RA, Chien MB, Lyons LA, London CA (2004) Detection of c-kit mutations in canine mast cell tumors using fluorescent polyacrylamide gel electrophoresis. J Vet Diagn Invest 16(2):95\u0026ndash;100\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLetard S, Yang Y, Hanssens K, Palmerini F, Leventhal PS, Guery S et al (2008) Gain-of-function mutations in the extracellular domain of KIT are common in canine mast cell tumors. Mol Cancer Res 6(7):1137\u0026ndash;1145\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLondon CA, Galli SJ, Yuuki T, Hu ZQ, Helfand SC, Geissler EN (1999) Spontaneous canine mast cell tumors express tandem duplications in the proto-oncogene c-kit. Exp Hematol 27(4):689\u0026ndash;697\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReguera MJ, Ferrer L, Rabanal RM (2002) Evaluation of an intron deletion in the c-kit gene of canine mast cell tumors. Am J Vet Res 63(9):1257\u0026ndash;1261\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiva F, Brizzola S, Stefanello D, Crema S, Turin L (2005) A study of mutations in the c-kit gene of 32 dogs with mastocytoma. J Vet Diagn Invest 17(4):385\u0026ndash;388\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZemke D, Yamini B, Yuzbasiyan-Gurkan V (2002) Mutations in the juxtamembrane domain of c-KIT are associated with higher grade mast cell tumors in dogs. Vet Pathol 39(5):529\u0026ndash;535\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIsotani M, Ishida N, Tominaga M, Tamura K, Yagihara H, Ochi S et al (2008) Effect of tyrosine kinase inhibition by imatinib mesylate on mast cell tumors in dogs. J Vet Intern Med 22(4):985\u0026ndash;988\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePryer NK, Lee LB, Zadovaskaya R, Yu X, Sukbuntherng J, Cherrington JM et al (2003) Proof of target for SU11654: inhibition of KIT phosphorylation in canine mast cell tumors. Clin Cancer Res 9(15):5729\u0026ndash;5734\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHahn KA, Ogilvie G, Rusk T, Devauchelle P, Leblanc A, Legendre A et al (2008) Masitinib is safe and effective for the treatment of canine mast cell tumors. J Vet Intern Med 22(6):1301\u0026ndash;1309\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebster JD, Yuzbasiyan-Gurkan V, Thamm DH, Hamilton E, Kiupel M (2008) Evaluation of prognostic markers for canine mast cell tumors treated with vinblastine and prednisone. BMC Vet Res 4:32\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebster JD, Yuzbasiyan-Gurkan V, Kaneene JB, Miller R, Resau JH, Kiupel M (2006) The role of c-KIT in tumorigenesis: evaluation in canine cutaneous mast cell tumors. Neoplasia 8(2):104\u0026ndash;111\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLondon CA, Malpas PB, Wood-Follis SL, Boucher JF, Rusk AW, Rosenberg MP et al (2009) Multi-center, placebo-controlled, double-blind, randomized study of oral toceranib phosphate (SU11654), a receptor tyrosine kinase inhibitor, for the treatment of dogs with recurrent (either local or distant) mast cell tumor following surgical excision. Clin Cancer Res 15(11):3856\u0026ndash;3865\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeishaar KM, Ehrhart EJ, Avery AC, Charles JB, Elmslie RE, Vail DM et al (2018) c-Kit Mutation and Localization Status as Response Predictors in Mast Cell Tumors in Dogs Treated with Prednisone and Toceranib or Vinblastine. J Vet Intern Med 32(1):394\u0026ndash;405\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRout ED, Avery AC (2012) Molecular Diagnostics of Hematologic Malignancies. In: Thrall MA, editor. Veterinary hematology and clinical chemistry. 2nd ed. Ames, Iowa: Wiley-Blackwell; p. xii, 762 p\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIodence AE, Wallace ML, Grimes JA, Schmiedt CW (2021) Dogs undergoing surgical excision of mast cell tumors are not at increased risk of incisional complications. J Am Vet Med Assoc 260(S1):S88\u0026ndash;S95\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshiguro T, Kadosawa T, Takagi S, Kim G, Ohsaki T, Bosnakovski D et al (2003) Relationship of disease progression and plasma histamine concentrations in 11 dogs with mast cell tumors. J Vet Intern Med 17(2):194\u0026ndash;198\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLondon CA, Seguin B (2003) Mast cell tumors in the dog. Vet Clin North Am Small Anim Pract 33(3):473\u0026ndash;489 v\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang C, Wallerman O, Arendt ML, Sundstrom E, Karlsson A, Nordin J et al (2021) A novel canine reference genome resolves genomic architecture and uncovers transcript complexity. Commun Biol 4(1):185\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eField MA, Rosen BD, Dudchenko O, Chan EKF, Minoche AE, Edwards RJ et al (2020) Canfam_GSD: De novo chromosome-length genome assembly of the German Shepherd Dog (Canis lupus familiaris) using a combination of long reads, optical mapping, and Hi-C. Gigascience. ;9(4)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDolezel J, Bartos J, Voglmayr H, Greilhuber J (2003) Nuclear DNA content and genome size of trout and human. Cytometry A 51(2):127\u0026ndash;128 author reply 9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEvers B, Jonkers J (2006) Mouse models of BRCA1 and BRCA2 deficiency: past lessons, current understanding and future prospects. Oncogene 25(43):5885\u0026ndash;5897\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeto RT, Cagnini DQ, Amorim RL (2013) Mutations in C-KIT exon 11 in canine cutaneous mast cell tumors. BMC Proceedings. ;7(2):P56\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma A, Merritt E, Hu X, Cruz A, Jiang C, Sarkodie H et al (2019) Non-Genetic Intra-Tumor Heterogeneity Is a Major Predictor of Phenotypic Heterogeneity and Ongoing Evolutionary Dynamics in Lung Tumors. Cell Rep 29(8):2164\u0026ndash;74e5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun XX, Yu Q (2015) Intra-tumor heterogeneity of cancer cells and its implications for cancer treatment. Acta Pharmacol Sin 36(10):1219\u0026ndash;1227\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDagher H, Donninger H, Hutchinson P, Ghildyal R, Bardin P (2004) Rhinovirus detection: comparison of real-time and conventional PCR. J Virol Methods 117(2):113\u0026ndash;121\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePestana E, Belak S, Diallo A, Crowther JR, Viljoen GJ (2014) Early, rapid and sensitive veterinary molecular diagnostics - real time PCR applications. Springer Dordrecht, p 310\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Cardiff University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Canine cutaneous mast cell tumour, KIT, exon 11 Internal tandem duplication, qPCR","lastPublishedDoi":"10.21203/rs.3.rs-9368633/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9368633/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn canine cutaneous mast cell tumours (cMCTs), Internal Tandem Duplications (ITDs) of exon 11 of the \u003cem\u003eKIT\u003c/em\u003e receptor tyrosine kinase are associated with higher histological grade and greater risk of metastasis. They also predict response to tyrosine kinase inhibitors (TKIs). Assessment of \u003cem\u003eKIT\u003c/em\u003emutation status provides important information when identifying candidate patients for TKI therapy.\u003c/p\u003e\n\u003cp\u003eCurrent \u003cem\u003eKIT\u003c/em\u003e ITD assays based on semi-quantitative PCR and electrophoretic separation of PCR products lack sensitivity and are not easily comparable between samples. They are limited to analysis of biopsy material from tumours and cannot be used for monitoring of residual disease during routine clinical follow-up, for instance using fine needle aspirates or blood tests. Here, we describe a novel, sensitive, quantitative qPCR assay based on melting curve analysis with potential for use in a wide range of samples during initial clinical assessment/staging and also routine follow up of cMCT patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe identified a minimal region (‘min’ region) of \u003cem\u003eKIT\u003c/em\u003e exon 11 commonly amplified in cMCT and designed qPCR primers against that region to produce a small product (44bp) from a wild-type (WT) target. However, an ITD including the min region duplicates primer annealing sites, meaning that additional, larger (\u0026gt;90bp) PCR products are generated. The presence of an ITD-derived product can be distinguished from the WT product by examining peaks on the qPCR melt curves (the WT product produces one peak, the ITD product produces a second distinct peak). The assay was tested on fifteen FFPE cMCT samples with known \u003cem\u003eKIT\u003c/em\u003e exon 11 status (five WT, ten with an ITD) and on MDCK normal canine epithelial cell DNA. Considering each individual replicate carried out on all samples, the analytical specificity of the assay was 94% and the sensitivity was 100%.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe have developed a novel, rapid, qPCR-based assay for the presence of \u003cem\u003eKIT\u003c/em\u003e exon 11 ITDs in canine cutaneous Mast Cell Tumours. The assay uses PCR-product melt curve analysis to detect the duplications with high sensitivity and specificity. The assay has the potential for use in a wide range of cMCT samples, including non-surgical samples such as blood or fine needle aspirates.\u003c/p\u003e","manuscriptTitle":"A novel, minimally invasive diagnostic test for KIT exon 11 internal tandem duplications in canine cutaneous mast cell tumours I: Assay development","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-10 09:39:30","doi":"10.21203/rs.3.rs-9368633/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9fe19466-ebc8-458a-9d90-526b871a0a27","owner":[],"postedDate":"April 10th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66015319,"name":"Small Animal Medicine"},{"id":66015320,"name":"Oncology"}],"tags":[],"updatedAt":"2026-04-10T09:39:36+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-10 09:39:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9368633","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9368633","identity":"rs-9368633","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}