Patent Grant
12344898
This application is a U.S. national stage of PCT/IB2017/056875 filed on 3 Nov. 2017 which claims priority to and the benefit of Italian patent application No. 102016000111174 filed on Nov. 4, 2016, the content of which are incorporated herein by reference in their entireties.
Sequence listing ASCII file Untitled ST25.txt, created on Apr. 6, 2020 and of size of 18,840 bytes is incorporated herein by reference.
The present invention refers to a method performed in vitro to determine a Head and Neck Squamous Cell Carcinoma, preferably at early stage, based on the measuring of the hypermethylation and/or hypomethylation level(s) of CpGs/CpG islands of specific set of genes.
Oral and pharyngeal cancer, grouped together, are the sixth most common cancer in the world. The annual estimated incidence is approximately 600,000 per year, two-thirds of these cases occurring in developing countries. The mortality rates of these tumors have remained unchanged (50% within 5 years after diagnosis) and are mostly related to tobacco smoking and alcohol intake. Oral cancer patients are usually diagnosed in an advanced stage, which is associated with worse prognosis and higher radio- and chemotherapy morbidity. Moreover, patient's quality of life is extremely compromised, since surgical therapy usually involves mutilating which often has significant effects on swallowing, speech, and physical appearance.
In particular, Oral Squamous Cell Carcinoma (OSCC) is the most frequent Head and Neck Squamous Cell Carcinoma (HNSCC—the neoplastic diseases affecting head and neck regions) and is usually preceded by oral pre-malignant lesions (OPML). Clinical and histological features of OPML are not able to provide enough information to identify the lesions, leading to a high risk of enduring malignant transformation and develop an OSCC during follow-up. In addition, patients affected by OSCC can developed a second primary OSCC, with a frequency ranging between 17% and 30%.
Although the oral cavity is easily accessible for examination, several factors limit the identification and early treatment of OPMLs. For this purpose, the current gold standard for screening and detecting, is the visual and tactile palpation during an extra- and intraoral examination by the healthcare professional in a routine dental or physical examination.
However, this disease is not easy to identify in its earliest stages and has often eluded medical and dental professionals because it can be “occult,” or hidden from plain view. Indeed, normal-looking tissue may often hide the truth within the cells below the mucosa's surface.
If the disease is identified in earlier stages (Stage I or Stage II)—the ideal time for identification is before the dysplastic cells have been able to break through the basement membrane—the overall five-year survival rate is greater than 80 percent.
Unfortunately, all too often the manifestations of this invasive and devastating disease are detected in the late stages (Stages III-IV), when the lesions have typically advanced so deeply that it is impossible to treat without radical surgical intervention and significant loss of the patient's quality of life.
In particular, OSCC or High Grade Squamous Intraepithelial Lesion (HG-SIL) are usually diagnosed based on an incisional biopsy. Nevertheless, the incisional biopsy requires a minimally invasive surgical approach that can create discomfort and be refused by the patient.
It looks evident from the comments reported above that improved oral cancer prevention, early detection, diagnostic, and clinical management tools are needed to identify high-risk patients, such as those exposed to smoking and alcohol consumption, patients without adequate access to health care, and with high-risk lesions such as leukoplakia, which may progress to carcinoma lesions. In particular, considering how much HNPCC is widespread, there is a huge need of developing a non-invasive method to early detect this kind of tumors, and in particular oral premalignant lesions (OPML) which may develop neoplastic lesions, in order to reduce the burden of OSCC.
Over the last decade, Quantitative Methylation Specific PCR (qMSP) preceded by bisulfite treatment has been proposed as a method to evaluate biomarkers useful in OSCC detection. In this regard, various genes have been previously studied for promoter methylation status in OSCC tissues. Guerrero-Preston et al, 2014 and the related patent application WO2015066170A1 refer to a comprehensive integrated genomic and epigenomic analysis in HNSCC, focusing on identifying genes that have concurrent promoter methylation, mutation and expression downregulation. In particular, these documents evaluated the methylation pattern of PAX1, PAX5, ZIC4 and PLCB1.
Moreover, Morandi et al. developed a non-invasive method to early detect OSCC and HG-SIL by epigenetic modifications analysis in the oral mucosa evaluating GP1BB and ZAP70 methylation status (Morandi, L. 2015).
However, the above-mentioned study is limited for number of cases and for low sensitivity and specificity of the method. Therefore, is still felt the need to identify methods allowing early and accurate detection of HNSCC, in particular HG-SIL/OSCC. This is not easy to obtain also because identifying the key genes and locating the informative and efficient methylation sites on their sequences are not trivial and obvious activities. Indeed, it is well known, for example that the gene network involved in carcinogenesis is really complex and wide and that the promoter region of genes spans more than one thousand base pairs and contains approximately one hundred potential methylation sites.
The present invention solves the needs reported above with a method involving the measurement of the hypermethylation and/or hypomethylation level(s) of informative CpGs/CpG islands in specific genomic regions of a selected number of genes.
The ability to correlate the methylation levels of several CpGs/CpG islands of a selected number of genes grants the method an accuracy beyond any other methods currently available. Moreover, the disclosed method allows the early detection of HNSCC with high specificity and sensitivity starting from a biological sample, such as an oral brushing.
In the context of the present invention, “Head and Neck Squamous Cell Carcinoma (HNSCCs)” are tumors developing in the mucous membranes of mouth, nose, or throat and, generally, classified by their location. Indeed, this kind of carcinoma can occur in the mouth (oral cavity), in the middle part of the throat near the mouth (oropharynx), in the space behind the nose (nasal cavity and paranasal sinuses), in the upper part of the throat near the nasal cavity (nasopharynx), in the larynx, or in the lower part of the throat near the larynx (hypopharynx).
In the context of the present invention, “Oral Squamous Cell Carcinoma (OSCC)” represents 95% of all forms of HNSCC and during the past decade, its incidence has increased by 50%. OSCC, emerge from the accumulation of genetic changes and epigenetic anomalies in the signaling pathways that are associated with cancer, resulting in phenotypes that facilitate OSCC development. It derives from the stratified squamous epithelium of the oral mucosa. The majority of OSCC are diagnosed at a late phase, in stages Ill or IV, which markedly decreases the chances of survival and leads to a significant deterioration in patient quality of life.
“High Grade Squamous Intraepithelial Lesion (HG-SIL)” has been introduced by Gale et al. 2014. The classification provides two grades: 1) low-grade, and 2) high-grade lesions and, specifically for the larynx, an additional grade-carcinoma in situ (CIS) which must be separated from high-grade laryngeal SILs. This study provided clear morphological criteria with which to define prognostic groups. The retrospective follow-up study demonstrated a highly significant difference in the risk of malignant progression between low grade and high-grade lesions improving patient management and clinical decision-making.
In the context of the present invention, “Oral Pre-Malignant Lesion (OPML)” is a morphologically altered tissue in which oral cancer is more likely to occur than in its apparently normal counterpart. Recently, the World Health Organization (WHO) identified the following OPML: leukoplakia, erythroplakia, palatal lesion of reverse cigar smoking, oral lichen planus, oral submucous fibrosis (SMF) and discoid lupus erythematosus.
In the context of the present invention, “DNA methylation” means epigenetic modifications of nucleic acids that alter their accessibility and structure, preferably chromatin structure, thereby regulating patterns of gene expression. They can be modified by exogenous influences and, as such, can contribute to or be the result of environmental alterations of phenotype or pathophenotype.
In the context of the present invention, “CpG islands” are regions of the human genome with elevated G+C content. These regions have generally about 1 kb length and, usually overlap the promoter region of 60-70% of all human genes. They are also present in repeats as transposable elements and functions to repress transcription. Within these regions, the majority of CpG pairs are chemically modified by the covalent attachment of a methyl group to the C5 position of the cytosine ring. Aberrant DNA methylation of CpGs in the proximity of transcription start sites often leads to alterations in gene function and pathway deregulation in human cancer.
In the context of the present invention, “genomic DNA region” means regions of nucleic acid, preferably DNA, more preferably genomic DNA such as the gene promoter, the 5′UTR, the gene body, the 3′UTR and shores (refers to DNA sequence that occur up to about 2 kb distant from a CpG island of comparatively low CpG density that are located near traditional CpG islands).
A first aspect of the present invention refers to a method for detecting (for determining the presence of) a Head and Neck Squamous Cell Carcinoma (HNSCC) comprising the steps of:
The method of the invention allows determining if the biological sample belongs to an individual affected by a HNSCC (HNSCC positive). Preferably, the HNSCC is detected (determined) at early stage. Therefore, the detection is an early stage detection. Thus, the method can be also used as prognostic tool and/or for the follow up of a therapeutic treatment of a HNSCC. Finally, the method of the invention is also useful in the prevention of these kind of carcinomas.
Therefore, the method of the invention is also for prognosing a Head and Neck Squamous Cell Carcinoma and/or for the follow up of a therapeutic treatment of a HNSCC and/or for the prevention of a HNSCC.
In this context, EPHX3 is acronym of epoxide hydrolase or the following synonyms: FLJ22408, ABHD9, EH3; KIF1A is acronym of kinesin family member 1 or the following synonyms: UNC104, C2orf20, ATSV, HSN2C, SPG30, MRD9; LRRTM1 is acronym of leucine rich repeat transmembrane neuronal 1 or the following synonyms: FLJ32082; FLI1 is acronym of Fli-1 proto-oncogene, ETS transcription factor or the following synonyms: EWSR2, SIC-1; ITGA4 is acronym of integrin subunit alpha 4 or the following synonyms: IA4, CD49D; LINC00599 is acronym of long intergenic non-protein coding RNA 599 or the following synonyms: MIRN124A1, MIRN124-1, MIR124A1, MIR124A, Rncr3, mir-124-1; NTM is acronym of neurotrimin or the following synonyms: NT, NTRI, IGLON2, HNT; PARP15 is acronym of poly(ADP-ribose) polymerase family member 15 or the following synonyms: FLJ40597, pART7, ARTD7, BAL3; ZAP70 is acronym of zeta chain of T cell receptor associated protein kinase 70 or the following synonyms: IMD48, TZK, STD, ZAP-70, SRK, STCD, ADMIO2; miR193 is acronym of microRNA 193a or the following synonyms: MIRN193A, mir-193a, hsa-mir-193a, hsa-mir-193, MIRN193; miR296 is acronym of microRNA 296 or the following synonyms: hsa-mir-296, mir-296, MIRN296, miRNA296; GP1BB is acronym of glycoprotein 1b platelet beta subunit or the following synonyms: GPIBB, GPIbbeta, CD42C, BS, BDPLT1, CD42c; hTERT is acronym of telomerase reverse transcriptase or the following synonyms: TRT, TCS1, DKCA2, EST2, PFBMFT1, CMM9, hEST2, DKCB4, hTRT, TP2.
In the context of the present invention, hypermethylation means that the measured methylation level of the at least one CpG/CpG island is increase in a tested sample (isolated from an individual suspected to be affected by a HNSCC) compared to a negative sample (a biological sample isolated from a healthy individual not affected by a HNSCC), wherein the negative sample is preferably a sample of DNA isolated from normal mucosa of healthy individuals not affected by a HNSCC.
In the context of the present invention, hypomethylation means that the measured methylation level of the at least one CpG/CpG island is reduced in a tested sample (isolated from an individual suspected to be affected by a HNSCC) compared to a negative sample (a biological sample isolated from a healthy individual not affected by a HNSCC), wherein the negative sample is preferably a sample of DNA isolated from normal mucosa of healthy individuals not affected by a HNSCC.
The biological sample of step (i) is isolated from an individual who can be affected or suspected to be affected by a HNSCC, preferably he has been surgically treated for a HNSCC, preferably a OSCC. According to a preferred embodiment, said HNSCC is any tumor developing in oral cavity, nose, and throat. Preferably, said HNSCC is any Oral Squamous Cell Carcinoma or its precursor, preferably High-Grade Squamous Intraepithelial Lesion (HG-SIL), or an Oral Premalignant Lesion or a cancerization field. Cancerization field means an area of mucosal epithelium made up of genetically altered cells eventually clonally related to the carcinoma, and therefore possible cause of local recurrence or secondary tumors.
According to a preferred embodiment, said Oral Premalignant Lesion (OPML) is preferably selected from: leukoplakia, erythroplakia, palatal lesion of reverse cigar smoking, oral lichen planus, oral submucous fibrosis (SMF), and discoid lupus erythematosus wherein an oral cancer is more likely to occur than in its apparently normal counterpart and can be detected by the present invention if the lesion is a real precursor of HNSCC.
Therefore, the method is ex vivo. The biological sample is any biological sample, preferably said sample is selected from: brushing from oral mucosa, plasma, saliva and any sample containing epithelial cells, preferably said cells being from oral cavity.
According to a preferred embodiment, the sample is isolated from the buccal cavity, preferably oral mucosa, more preferably by brushing, preferably by using a cytobrush or DNA buccal swab. This procedure allows obtaining a sample comprising exfoliated cells from buccal cavity, preferably from oral mucosa.
According to a preferred embodiment, the sample collection is preceded by oral rinses, at least 1-3 times, preferably with chlorhexidine 0.12% or any germicidal mouthwash. According to a further preferred embodiment, the sample collection is made by gently brushing a surface of the body region of interest, preferably said region showing a clinical lesion, preferably 1-10 times, more preferably about five times. It is advisable in the collecting procedure avoiding 1) to collect blood, and 2) to do not use any local anaesthetic.
The collected sample is preferably placed in a container containing a preservative solution (i.e. DNA/RNA Shield™), to allow nucleic acid preservation. More preferably, the sample is stored at low temperature, preferably about +4° C.
According to a further preferred embodiment, the sample undergoes to step (ii) involving isolating (extracting-purifying) a nucleic acid from the biological sample. Preferably, the nucleic acid isolated is DNA, preferably the genomic DNA. Any technique used in a laboratory to extract nucleic acids, in particular to extract high pure DNA, can be used, for example the MasterPure™ DNA purification kit.
After purification, the nucleic acid, preferably the purified genomic DNA, is treated with bisulfite, preferably sodium bisulfite (step iii).
Bisulfite converts an unmethylated cytosine (C) to uracil (U), such that uracil (U) is read thymine (T) when the treated nucleic acid is sequenced. This conversion does not affect a methylated C, which remains C in the sequence. Any method allowing to address this purpose should be considered an alternative to the bisulfite treatment and, therefore, part of the disclosure.
The bisulfite treatment of the nucleic acid is performed according to any common lab protocols or using commercially available kits, such as EZ Methylation-Lightning™ kit. Preferably, the bisulfite treatment is performed using an amount of nucleic acid, preferably of DNA, ranging from 50 ng to 10 g, preferably from 250 ng to 1 g.
Preferably, the bisulfite treatment involves a first step of denaturation of the nucleic acid, preferably at 98° C. for 10-20 minutes, more preferably about 8 minutes and/or the incubation of the denaturated nucleic acid with sodium bisulfite. Preferably, the bisulfite is used in a concentration of about 2.3-3.6 M, more preferably in the presence of hydroquinone, preferably in a concentration of about 10 mM. The reaction is preferably performed at a temperature ranging preferably from 45 to 60° C., more preferably from 50 to 55° C., still more preferably of about 54° C. for at least one hour.
At the end of the treatment, the nucleic acid is purified preferably by using a column technology and more preferably with a first ethanol based wash, a desulfonation step of 20 minutes at room temperature, two washing steps with buffered ethanol, and the elution step following the instruction of the provider.
After bisulfite treatment, the nucleic acid, preferably the DNA is amplified, preferably by using a PCR (polymerase Chain Reaction) protocol. The amplification is performed preferably by using at least one primer pair said primer pair allowing the amplification of at least one sequence, preferably a portion/sequence of the genomic region of at least one gene of interest said sequence comprising at least one CpG/CpG island. Preferably, the primer pair allows the amplification of a sequence of the genomic DNA region of at least one gene selected from: EPHX3, KIF1A, LRRTM1, FLI1, ITGA4, LINC00599, NTM, PARP15, ZAP70, miR193, miR296, GP1BB, hTERT and any combination thereof.
Preferably, at least one of the following combination of genes are amplified, meaning that at least one sequence of the genomic (DNA) region of at least one of the following combination of genes said region comprising at least one CpG/CpG island, is amplified in the method of the invention:
However, any further combination of genes can be amplified according to the method of the invention.
Preferably, the genomic region of a gene means any regulating region of the gene, more preferably the promoter (generally upstream the ATG start site of a gene) region of the gene, the 5′ and/or 3′ UTRs, and/or any coding region (gene body), preferably exons and/or introns and/or shores.
According to a preferred embodiment of the invention, said region is preferably selected from SEQ ID NO: 53-65 wherein:
In the context of the present invention, a locus of a CpG/CpG island has been mapped according to the human reference genome 19 (hg19), using blat, in which there is indication of chromosome number, position, positive or negative strand and percentage of identity.
According to a preferred embodiment, the at least one sequence of PARP15 amplified for the methylation level analysis according to step (iii) (amplicon is the name of the amplification product) is SEQ ID NO: 53 corresponding to the region of genomic DNA (strand+) mapped on chromosome 3 and having coordinates from 122296564 to 122296723 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 122296564, 122296586, 122296613, 122296617, 122296623, 122296630, 122296637, 122296645, 122296649, 122296656, 122296663, 122296671, 122296677, 122296680, 122296692, 122296708, 122296710, 122296717, 122296723 and combination thereof, preferably is 122296586.
According to a preferred embodiment, the at least one sequence of ITGA4 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 54 corresponding to the region of genomic DNA (strand+) mapped on chromosome 2 and having coordinates from 182322887 to 182323053 with respect to hg19 coordinates and 25 the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 182322890, 182322898, 182322902, 182322908, 182322915, 182322937, 182322956, 182322962, 182322980, 182322985, 182323004, 182323016, 182323024, 182323028; and combination thereof, preferably is 182322902.
According to a preferred embodiment, the at least one sequence of NTM amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 55 corresponding to the region of genomic DNA (strand+) mapped on chromosome 11 and having coordinates from 131781042 to 131781185 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 131781046, 131781058, 131781060, 131781080, 131781082, 131781084, 131781098, 131781110, 131781113, 131781122, 131781144, 131781151, 131781158, 131781167, 131781183; and combination thereof, preferably is 131781082 or 131781167.
According to a preferred embodiment, the at least one sequence of ZAP70 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 56 corresponding to the region of genomic DNA (strand+) mapped on chromosome 12 and having coordinates from 98340750 to 98340885 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 98340751, 98340762, 98340764, 98340766, 98340772, 98340775, 98340794, 98340801, 98340809, 98340813, 98340816, 98340823, 98340828, 98340834, 98340844, 98340854, 98340859, 98340869, 98340877, 98340881 and combination thereof, preferably is 98340751 or 98340854.
According to a preferred embodiment, the at least one sequence of MIR193 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 57 corresponding to the region of genomic DNA (strand+) mapped on chromosome 17 and having coordinates from 29886860 to 29887068 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 29886868, 29886870, 29886875, 29886881, 29886885, 29886890, 29886903, 29886910, 29886913, 29886936, 29886939, 29886944, 29886949, 29886951, 29886964, 29886967, 29886971, 29886983, 29886992, 29886995, 29887001, 29887008, 29887015, 29887045, 29887049, 29887063 and combination thereof, preferably is 29886870 or 29886944.
According to a preferred embodiment, the at least one sequence of EPHX3 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 58 corresponding to the region of genomic DNA (strand −) mapped on chromosome 19 and having coordinates from 15342826 to 15343049 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 15342895, 15342889, 15342885, 15342875, 15342869, 15342866, 15342864, 15342855, 15342851, 15342848, 15342845, 15342840, 15342838, 15342831, 15343002, 15343000, 15342991, 15342988, 15342985, 15342983, 15342976, 15342962, 15342960, 15342934, 15342928, 15342918, 15342916, 15342913, 15342906 and combination thereof, preferably is 15342840 or 15342885.
According to a preferred embodiment, the at least one sequence of KIF1A amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 59 corresponding to the region of genomic DNA (strand −) mapped on chromosome 2 and having coordinates from 241759586 to 241759727 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 241759726, 241759722, 241759719, 241759717, 241759715, 241759702, 241759697, 241759688, 241759686, 241759681, 241759676, 241759671, 241759665, 241759663, 241759656, 241759651, 241759649, 241759646, 241759643, 241759640, 241759632, 241759621, 241759618, 241759614, 241759602, 241759596, 241759588 and combination thereof, preferably is selected from: 241759686, 241759681, 241759651, and 241759621 and combination thereof;
According to a preferred embodiment, the at least one sequence of LINC00599 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 60 corresponding to the region of genomic DNA (strand −) mapped on chromosome 8 and having coordinates from 9760739 to 9760890 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 9760888, 9760881, 9760878, 9760870, 9760856, 9760852, 9760848, 9760839, 9760832, 9760822, 9760820, 9760815, 9760811, 9760809, 9760805, 9760803, 9760777, 9760764, 9760759, 9760752 and combination thereof, preferably is 9760888.
According to a preferred embodiment, the at least one sequence of FLI1 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 61 corresponding to the region of genomic DNA (strand+) mapped on chromosome 11 and having coordinates from 128564020 to 128564160 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 128564022, 128564033, 128564051, 128564063, 128564065, 128564068, 128564078, 128564089, 128564106, 128564134, 128564158 and combination thereof, preferably is 128564051 or 128564158.
According to a preferred embodiment, the at least one sequence of GP1BB amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 62 corresponding to the region of genomic DNA (strand+) mapped on chromosome 22 and having coordinates from 19710829 to 19710960 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 19710839, 19710842, 19710847, 19710855, 19710864, 19710880, 19710890, 19710902, 19710909, 19710916, 19710919, 19710937, 19710944, 19710946, 19710948, 19710952, 19710954, 19710956, and combination thereof, preferably is 19710946 or 19710956.
According to a preferred embodiment, the at least one sequence of MIR296 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 63 corresponding to the region of genomic DNA (strand −) mapped on chromosome 20 and having coordinates from 57392355 to 57392545 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 57392543, 57392538, 57392536, 57392517, 57392501, 57392469, 57392452, 57392440, 57392437, 57392430, 57392419, 57392394, 57392386, 57392374, 57392363, and combination thereof, preferably is 57392363 or 57392374.
According to a preferred embodiment, the at least one sequence of LRRTM1 amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 64 corresponding to the region of genomic DNA (strand −) mapped on chromosome 2 and having coordinates from 80531676 to 80531807 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 80531806, 80531803, 80531799, 80531796, 80531789, 80531786, 80531783, 80531780, 80531776, 80531763, 80531759, 80531756, 80531752, 80531749, 80531745, 80531742, 80531738, 80531719, 80531702, 80531697, 80531695, 80531677 and combination thereof, preferably is 80531697 or 80531799.
According to a preferred embodiment, the at least one sequence of Human TERT (hTERT) amplified for the methylation level analysis according to step (iii) (amplicon) is SEQ ID NO: 65 corresponding to the region of genomic DNA (strand −) mapped on chromosome 21 and having coordinates from 1279743 to 1279851 with respect to hg19 coordinates and the at least one CpG/CpG island evaluated on this sequence corresponds to a coordinate selected from: 1279847, 1279838, 1279829, 1279775, 1279758, 1279746 and combination thereof, preferably is 1279775, 1279747 or 1279758.
According to a further preferred embodiment, phase (iii) of measuring the epigenetic modifications (methylation) of set (1)—KIF1A and human TERT—involves measuring 1) the hypermethylation of at least one of the CpG/CpG island of SEQ ID NO: 59 reported above, more preferably the CpG/CpG island having the coordinate 241759686, 241759681, 241759651, 241759621 of SEQ ID NO: 59 and/or 2) the hypomethylation of at least one of the CpG/CpG island of SEQ ID NO: 65 reported above, more preferably the CpG/CpG island having the coordinate 1279775 and/or, 1279747 and/or 1279758 of SEQ ID NO: 65.
According to a preferred embodiment phase (iii) of measuring the epigenetic modifications of set (2)—LRRTM1, PARP15, ZAP70, miR193, miR296, hTERT—involves measuring 1) the hypermethylation of at least one of the CpG/CpG island of SEQ ID NO: 64 for LRRTM1, SEQ ID NO: 53 for PARP15, SEQ ID NO: 56 for ZAP70, SEQ ID NO: 57 for miR193, more preferably the hypermethylation of the CpG/CpG island having the following coordinates: 80531799 and/or 80531697 for LRRTM1. 122296586 for PARP. 98340854 and/or 98340751 for ZAP70. 29886870 and or 29886944 for miR193 respectively and combination thereof; and/or 2) the hypomethylation of at least one of the CpG/CpG island of SEQ ID NO: 63 for miR296, SEQ ID NO: 65 for TERT, preferably the CpG/CpG island having the following coordinates: 57392374 and or 57392363 for miR296, 1279758 and or 1279775 and/or 1279747 for TERT respectively.
According to a further preferred embodiment, phase (iii) of measuring the epigenetic modifications of set (3)—KIF1A, LRRTM1, ZAP70, miR193, miR296, hTERT—involves measuring 1) the hypermethylation of at least one of the CpG/CpG island of SEQ ID NO: 59 for KIF1A, SEQ ID NO: 64 for LRRTM1, SEQ ID NO: 56 for ZAP70, SEQ ID NO: 57 for miR193, more preferably the hypermethylation of the CpG/CpG island having the following coordinates: 241759686, 241759681, 241759651, 241759621 for KIF1A, 80531799 and/or 80531697 for LRRTM1, 98340854 and/or 98340751 for ZAP70, 29886944 and/or 29886870 for miR193 respectively and any combination thereof; and/or 2) the hypomethylation of at least one of the CpG/CpG island of SEQ ID NO: 63 for miR296, SEQ ID NO: 65 for TERT, more preferably the hypomethylation of the CpG/CpG island having the following coordinates: 57392374 and or 57392363 for miR296, 1279758 and or 1279775 and/or 1279747 for TERT respectively.
According to a further preferred embodiment phase (iii) of measuring the epigenetic modifications of set (4)—KIF1A, LRRTM1, FLI1, LINC00599, PARP15, ZAP70, miR193, miR296, GP1BB, hTERT—involves measuring 1) the hypermethylation of at least one of the CpG/CpG island of SEQ ID NO: 59 for KIF1A, SEQ ID NO: 64 for LRRTM1, SEQ ID NO: 61 for FLI1, SEQ ID NO: 60 for LINC00599, SEQ ID NO: 53 for PARP15, SEQ ID NO: 56 for ZAP70, SEQ ID NO: 57 for miR193, more preferably the hypermethylation of the CpG/CpG island having the following coordinates: 241759686, 241759681, 241759651, 241759621 for KIF1A, 80531799 and/or 80531697 for LRRTM1, 128564158 and/or 128564051 for FLI1, 9760888 for LINC0059, 122296586 for PARP15, 98340854 and/or 98340751 for ZAP70 any combination thereof; and/or 2) the hypomethylation of at least one of the CpG/CpG island of SEQ ID NO: 63 for miR296, SEQ ID NO: 62 for GP1BB, SEQ ID NO: 65 for TERT, more preferably the hypomethylation of the CpG/CpG island having the following coordinates: 57392374 and/or 57392363 for miR296, 19710956 and/or 19710946 for GP1 BB, 1279758 and or 1279775 and/or 1279747 for TERT.
According to a further preferred embodiment phase (iii) of measuring the epigenetic modifications of set (5)—EPHX3, KIF1A, LRRTM1, FLI1, ITGA4, LINC00599, NTM, PARP15, ZAP70, miR193, miR296, GP1BB, hTERT—involves measuring 1) the hypermethylation of at least one of the CpG/CpG island of KIF1A, preferably 1-4 CpG/CpG island of KIF1A, more preferably selected from SEQ ID NO: 59, SEQ ID NO: 58 for EPHX3, SEQ ID NO: 64 for LRRTM1, SEQ ID NO: 61 for FLI1, SEQ ID NO: 54 for ITGA4, SEQ ID NO: 60 for LINC00599, SEQ ID NO: 55 for NTM, SEQ ID NO: 53 for PARP15, SEQ ID NO: 56 for ZAP70, SEQ ID NO: 57 for miR193, and/or 2) the hypomethylation of at least one of the CpG/CpG island of SEQ ID NO: 63 for miR296, SEQ ID NO: 62 for GP1 BB, SEQ ID NO: 65 for human TERT.
All the sequences disclosed in the present patent application are listed in Table I and the CpGs/CpG islands for each region are in bold. Sequences showing 80-99.9% identity with these sequences have to be considered part of the disclosure.
In general, for the amplification step, a locus-specific amplicon library is preferably generated, more preferably a locus-specific bisulfite amplicon library. To this end is advisable to use a platform selected from: MiSEQ, NEXT500 or MiniSEQ—Illumina, IonTorrent, Pacific Biosciences and 454 GSJunior platform.
In the context of the present invention, IonTorrent is one of next-generation sequencing machine that utilizes massively parallel sequencing technologies to generate thousands of megabases of sequence information by semiconductor chips. In the context of the present invention, Illumina MiSEQ, NEXT500 or MiniSEQ are next-generation sequencing machines, which utilize sequence by synthesis chemistry to perform parallel sequencing of billions of DNA fragments in parallel with high precision, minimizing incorporation bias and reducing raw error rates compared to other technologies, especially for homopolymers. In any case, any method to obtain the same purpose has to be considered part of the present disclosure.
According to a preferred embodiment, the amplification steps involves a first step of amplification—a template specific amplification—and a second round of amplification allowing the barcoding of the template-specific amplicons obtained from the first amplification step.
The first amplification step uses preferably a multiplex strategy, preferably by using tagged primer pairs and a DNA polymerase, preferably the Phusion™ U DNA polymerase (
Preferably, the amplification conditions for this first amplification step are the following:
A final extension step at about 72° C. for about 5 min is preferably added at the end of the last cycle.
By using multiplex conditions several regions of interest of the genes can be amplified together in order to minimize costs and hands on time. This means that at least one region of interest for at least one gene can be amplified (amplicon) at the same time by using multiplex conditions. For the purpose of the present invention, the multiplex allows to amplify at least one sequence of at least one genomic DNA region comprising at least one CpG/CpG island of at least one gene selected from: PARP15, ITGA4, LRRTM1, human TERT, NTM, MIR193a, KIF1A, LINC00599, ZAP70, EPHX3, MIR296, FLI1N, GP1BB and combination thereof, preferably at least one of the following combinations of genes:
According to a preferred embodiment, for the first amplification step—the template specific amplification step—the primer pair is selected from the following tagged primer pairs:
Any combination of primer pairs can be used for each platform depending on the regions of the genes to amplify.
Each tagged primer pair comprises two primers (a forward and a reverse primer) for amplifying the region of interest. Each primer of the pair is a fusion primer comprising 1) a template specific portion hybridizing the specific sequence to amplify (3′-portion); and 2) a universal sequence (5′-portion) that does not have a complementary region in the genome (does not hybridize with any sequence in the genome). Therefore, the universal sequence is fused to a sequence targeting the template-specific sequence ends—that preferably define the boundaries of the amplicon.
In the context of the present invention, “universal sequence” is a sequence that does not have a complementary region in the genome, in other words it does not hybridize with any sequence in the genome. In particular, this sequence is recognized by a second primer combination carrying Illumina or IonTorrent adaptors and/or indexes.
The two forward primers or the two reverse primers for each region to be amplified for the platforms IonTorrent and Illumina, for example SEQ ID NO: 1 and SEQ ID NO: 27, have preferably the same 3′ end (because this end is the template specific end) and a different 5′ end depending of the used platform.
According to a preferred embodiment, in order to maintain the sequencing directionality, different universal sequences (tails) have to be designed for the two fusion primers of the first amplification step (in
The first amplification products (template specific), preferably the multiplex amplification products named amplicons, and/or the combinations are preferably purified, after the amplification step, preferably by using SPRI-AMPure XT. Moreover, they can be also quantified, after purification, preferably with the FluorometerQuantus™ or Qubit™ or similar.
The amplicons, eventually after purification and or quantification, are used as template, preferably in an amount of about 100 ng, for a second round of amplification. This second round is for barcoding the sample and it involves 5-10, preferably about 6 cycles.
This second amplification step is performed preferably in the presence of a further primer pair (forward and reverse). Each primer of the pair comprises starting from the 5′ end: 1) an adaptor sequence specific for the specific next generation sequencing platform used, such as Illumina and/or IonTorrent, and/or 2) a samples-specific barcode sequences, such as multiplex identifiers (MIDs) or indexes, and/or 3) a universal liker tag complementary to the universal sequence of the primer pair used in the first amplification step (
In the context of the present invention, “samples-specific barcode sequences or indexes” are short sequence that have been introduced artificially to specific target sequence allowing the identification of each sample (in other words to identify the individual) in parallel sequencing when samples from different individuals undergo to the method. Individual “barcode” sequences are added to each sample so they can be distinguished and sorted during data analysis. This approach is cost and time effective because pooling samples exponentially increases the number of samples analyzed in a single run.
In the context of the present invention, “multiplex identifiers or indexes” are inserted during the second round of amplification and include also adaptors specific for the specific next generation sequencing platform is intended to use, allowing the annealing of target sequences to the chip or the flow cell.
In the context of the present invention, “amplicon library” means a pool of target sequences derived from various samples, which are needed to be sequenced in parallel. Any sequencing technique has to be considered part of the present disclosure.
CGACCCCCAGAGAACTTATGCACGGAGTTG
GGGAGGCGGGGAGCGTGCTGCCGGCCGG
GCTCTTCCTCCCGGAGTATGGTGAGGAGCG
CGGGGGACGGGTGCG
GGCCGTCGGGCCTCGAGCCGCAGCCGGGG
CGTGACTACGTGCGCCA
GGATC
CGCCGTCGCGCCTGTCGCTGAAGCTGCTGC
GCGCCTTCATGTGGAGCCTGGTGTTCTCGG
GGGGGCGCGCAGCCCGTCACGCGGCGGCG
CGCAGACCTCCGCGCAGCGGCCGCGGGCG
CGAGGGGAGGGGTCTGGAGCTCCCTCCGG
CGAAGGGGCTGCGAGGTCAGGCTGTAACC
GGGTCAATGTGTGGAATATTGGGGGGCTCG
GCGTCCGCGCGGC
CGCCCGCCGCCCGCAAAGCATGAGTGAGC
Table II summarize several information on the genes analysed in the present method, the mapping information and the size of the amplified genomic region the reference of the sequence (NM reference on NCBI) and the number of the CpGs/CpG islands.
After determining the sequences of the amplicons, preferably according to the disclosure above, the quantitative methylation level measurement is performed using the common tools used for this purpose, preferably selected from: BISMARK, BS-Seeeker2, BISMA, BSPAT, and QUMA.
The analysis is based on the fact that bisulfite treatments of nuclei acid sequences converts an unmethylated cytosine (C) to uracil (U), such that uracil (U) is read thymine (T) when the treated nucleic acid is sequenced. This conversion does not affect a methylated C which remains C in the sequence. Therefore, sequencing reads showing C after bisulfite treatments mean a methylated cytosine, instead a T in the sequencing reads after bisulfite treatments mean and unmethylated cytosine.
According to a preferred embodiment, after the sequencing step, a FASTQ file comprising all the sequences obtained from a sample/individual (the reads) are qualitatively filtered by using cloud computing known for this purpose. FASTQ files are generally produced by Next Generation Sequencing runs. Preferably, for Illumina FASTQ a filter of Q30 for 90% of bases is applied. Less stringent filtering is applied for IonTorrent (Q scores are defined as the property that is logarithmically related to the base calling error probability (P)2 as Q=−10 log10 P. For example, if Phred assigns a Q score of 30 (Q30) to a base, this is equivalent to the probability of an incorrect base call 1 in 1000 times. In this case, the base call accuracy is 99.9%). This filtering step allows to obtain high quality sequences. After the filtering step the high quality sequences are preferably converted in a FASTA format and eventually trimmed, preferably 25 base pairs, at one and/or both ends in order to eliminate the primer sequence.
The sequences, preferably the high-quality sequences, obtained for each region amplified, are analyzed for the quantitative methylation level measurement by using the common tools used for this purpose, preferably selected from: BISMARK, BS-Seeeker2, BISMA, BSPAT, and QUMA. Preferably, the FASTA for each region amplified is uploaded on this tool.
These tools compare the wild type sequence and the sequence after bisulfite treatment and give a quantitative measurement of the methylation level. Therefore, these tools allow the measurement of the methylation level for the CpGs/CpG islands falling in each amplified region.
Preferably the reference sequences to upload for every sequence alignment needed by these tools are selected from: SEQ ID NO: 53-65.
Preferably, the hypomethylation value obtained for the hypomethylated CpG reported above of the hypomethylated genes, preferably selected from: TERT, MIR296 and GP1 BB, is inverted as follows: r=1−x, wherein x is the obtained hypomethylation value and r is the value to be included in the linear discriminant analysis.
According to a preferred embodiment, the methylation level data can be analyzed by using a Linear Discriminat Analysis, preferably selecting at least one, preferably all, CpG selected from: 19710956 for GP1BB; 98340854 for ZAP70; 241759621 for KIF1A; 1279758 for TERT; 80531799 for LRRTM1; 131781167 for NTM; 29886944 for MIR193; 122296586 for PARP15; 182322902 for ITGA4; 128564158 for FLI1; 9760888 for LINC0059; 15342885 for EPHX3; and 57392374 for MIR296.
Preferably, to quantify the methylation level of more than one sequence in order to obtain a total methylation level a multiclass Linear Discriminat Analysis (LDA) is preferably used. This means that, for example, when the CpG methylation level of at least one genomic region of more than one gene of interest, such as the set of genes disclosed above, is measured it is advisable to use this algorithm.
According to method of the present invention, the Linear Discriminant function that weighs the contribution of each methylated CpG is calculated by using the following formula:
y(ri)=K+Σci*ri
In particular, the discriminant coefficient for each epigenetic modification (methylation) is based on the number and the type of genomic regions used during the analysis.
The discriminant coefficient (ci) is calculated taking into consideration the levels (rates) of each methylation analysed in such way that the difference between positive and negative samples is maximized.
The constant (K) is a fixed number that is added to the discriminant function and is calculated in such way that the difference between positive and negative samples is maximized.
The final value of y is calculated after adding the constant to the summary of the product between the threshold of the methylation level for each at least one CpG/CpG island and the respective coefficient.
For a correct grouping of each sample, a threshold that discriminates between positive and N negative samples is preferably, but not limited to, calculated through a ROC curve analysis.
Values above certain y are to be considered positive.
According to a preferred embodiment, by using the Linear Discriminat Analysis and a ROC, a sample is considered HNSCC positive if:
The linear discriminant analysis (LDA) generates a score that weighted the at least one, preferably all, CpG selected from: 19710956 for GP1 BB; 98340854 for ZAP70; 241759621 for KIF1A; 1279758 for TERT; 80531799 for LRRTM1; 131781167 for NTM; 29886944 for MIR193; 122296586 for PARP15; 182322902 for ITGA4; 128564158 for FLI1; 9760888 for LINC0059; 15342885 for EPHX3; and 57392374 for MIR296, from the gene(s) investigated.
This accuracy meets the clinical requirements and thus the method of the invention is ideal to be used in place of and/or alongside the current reference methods.
According to a further aspect of the present invention, the method as disclosed above allow to detect the presence of field cancerization, preferably in surgically resected patients for HNSCC around the site of intervention. In particular, the invention shows that the presence of field cancerization with altered methylation pattern is related to poor prognosis and high risk of recurrence or secondary tumors.
A further aspect of the present invention refers to detect the presence of field cancerization in surgically resected patients for HNSCC around the site of intervention.
The presence of field cancerization with altered methylation pattern is related to poor prognosis and high risk of recurrence or secondary tumors.
A further aspect of the present invention, refers to a kit to determine the presence and/or the risk for HNSCC, preferably HG-SIL and/or OSCC in a biological test sample, preferably containing exfoliating brush of oral mucosa obtained from and individual, who preferably shows OPML or suspect lesions, said kit comprising reagents for sodium bisulfite conversion and a set of oligonucleotides, enzymes and buffers that specifically amplify at least one of the sequences SEQ ID NO: 53-SEQ ID NO: 65 with at least one primer pairs selected from the group consisting of SEQ ID NOs: 1-52 and from SEQ ID NOs: 66-69.
For the setting up and the validation phases of the present method, all consecutive patients referred to the Department of Oral Sciences, University of Bologna, from January 2015 to July 2016.
Lesions with an obvious aetiology, such as trauma and infective aphthous ulcerations, were excluded.
All patients presenting with oral lesions that required incisional biopsy to diagnostic purposed underwent also oral brushing of the same lesion.
Oral brushing samples were always picked before incisional biopsy for histological diagnosis and staging of each lesion.
Histological diagnoses were performed following the WHO criteria (Thompson, 2006). Distinction between HG-SIL and LG-SIL was made according to Ljubjana classification 2014 (Gale, Blagus et al. 2014).
A total of 150 oral brushing sample series for the training dataset was composed of:
The validation dataset was composed of:
A cytobrush was used to collect exfoliated cells from oral mucosa.
In OSCC and OPML lesions all surface of lesions was gently brushed repeatedly five times.
Brushing cell collection was always performed before incisional biopsy and without use of any local anaesthetic.
After brushing, each cytobrush sample was placed in a 2 ml tube containing absolute ethanol for cell preservation.
Ethics Statement
All clinical investigations have been conducted according to the principles expressed in the Declaration of Helsinki. The study was approved by local Ethics Committee (number of study: 14092, protocol number: 899/CE). All information regarding the human material used in this study was managed using anonymous numerical codes.
DNA Purification
DNA from exfoliating brush samples was purified using The MasterPure™ Complete DNA extraction kit (Epicentre, Madison, WI).
Bisulfite Treatment of Genomic DNA
Sodium bisulfite treatment of genomic DNA was carried out with the EZ DNA Methylation-Lightning™ Kit (Zymo Research Europe, Freiberg, Germany) according to the manufacturer's protocol.
Amplification Phase
For Illumina Platform (MiSEQ), locus-specific bisulfite amplicon libraries were generated multiplexing with tagged primers (see Table 1) using Phusion U DNA polymerase (ThermoFisher) in 25 μl. Illumina platform needs a different method to evaluate amplicons, the Nextera™ approach which requires adaptors at the 5′ end as described in Table 1. The insertion of one N or two N in the middle guarantees the minimum complexity for clustering. Barcoding was performed by using Nextera™ index kit.
Cycling conditions for the first template specific PCR were: initial incubation at 98° C. for 2 min followed by 1 cycle of 98° C. 10 sec, 62° C. 2 min, 72° C. 1 min, followed by 35 cycles at 98° C. for 10 s, 62° C. for 40 s and 72° C. for 20 sec. A final extension step at 72° C. for 5 min was added at the end of the last cycle.
Multiplex PCR consisted of:
Reaction conditions for multiplex were: ultrapure water: 8.8 μl, Buffer 5×HF: 5 μl, betaine 5M: 5 μl, dNTPs 10 mM: 0.5, primers 0.1 μM: 1.5 μl, MgCl2 50 mM: 0.5 μl, Phusion U: 0.2 μl, DNA bisulfite treated: 3.5 μl.
Amplification products from multiplex 1-2 and 3-4 were combined and purified using SPRI-AMPure XT (Agencourt-Beckman Coulter, Beverly, MA), quantified with the FluorometerQuantus™ (Promega, Madison, WI) and then employed as template (100 ng) for a second round of PCR (6 cycles).
Sample-specific barcode sequences (Indexes) and universal linker tags (P5/P7 adaptor sequences) were added in this second PCR performed with 6 cycles (98° C. 1 min./98° C. 10 sec./63° C. 30 sec./72° C. 3 min.).
The amplicon library was purified using AgencourtAMPure XP beads (Beckman Coulter, Krefeld, Germany), then quantitated with the FluorometerQuantus™ (Promega, Madison, WI).
The libraries were diluted, pooled at 4 nM and loaded on Nano Flow cell into MiSEQ. Alternatively, for amplicon sequencing experimental design for 454/IonTorrent platforms, the ‘Universal Tailed’ design requires two sets of primer pairs, used in two successive rounds of PCR.
The first round uses fusion primers targeting the template-specific sequences (defining the boundaries of the amplicons) fused to a universal sequence that will be the target of the second round primers (
In the second round, the Univ-A sequence is targeted by a fusion primer that is tailed by the 454 Sequencing system's Primer A sequence plus an MID sequence to identify the sample; and same for the Univ-B sequence, with Primer B (
As reported in Table I, the forward Universal A Tailed sequence of choice is GTAATACGACGGTCAGT (SEQ ID NO: 70) and the reverse Universal B Tailed sequence of choice is CAGGAAACAGCTATGAC (SEQ ID NO: 71).
In Silico Prediction of CpG Island and Primer Design
In order to identify putative the CpG islands on promoter region of the following genes: ZAP70, ITGA4, KIF1A, PARP15, EPHX3, NTM, LRRTM1, FLI1, MIR193, LINC00599, MIR296, hTERT and GP1BB, genomic sequence as stored on Ensembl genome browser, including 1000 bp upstream the ATG site were employed as query sequence. MethPrimer design were applied to identify CpGs and the best primers of choice (see Table 1).
In the second round of PCR, Universal A and B are recognized to add the specific Adaptors for 454 as follows including Adaptor a+key+MID+Universal A or B:
Bioinformatic Analysis:
After the end of run, each FASTQ file obtained after the NGS was processed as follows:
The behaviour of each CpG respect to their position along the human genome, is graphically pointed out by methylation plotter tool in
A table including the best CpGs for each gene has been created wherein class 0=normal donors and class 1=OSCC.
Using this table for LDA discriminant analysis by the tool SPSS (Selecting the following options: variable interval: 0-1; Statistics: not standardized values; Classifications: calculate from group dimension; create a summary Table; Save Discriminant Score).
Set (a) involves the following CpGs: 241759621 for KIF1A and 1279758 for TERT. The Linear Discriminant Analysis calculated the following coefficients:
Canonical Discriminant Function Coefficients
y=−0.808+5.338*(KIF1A value)+0.707*(1−hTERT value)
By the ROC curve analysis (
Area Under the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
Therefore, when the y value exceeds the threshold of −0.5306765, the sample should be considered positive and related to the presence of HG-SIL or OSCC.
Set (b) involves the following genes: LRRTM1 (CpG 80531799), PARP15 (CpG 122296586), ZAP70 (CpG 98340854), MIR193 (CpG 29886944), MIR296 (CpG 57392374), TERT (CpG 1279758).
In this case, the Linear Discriminant Analysis calculated the following coefficients:
Canonical Discriminant Function Coefficients
Then the algorithm will be the following: y=−3.470+2.422*(LRRTM1 value)+3.073*(miR193 value)+4.730*(1−miR296 value)+0.165*(1−hTERT value)+1.335*(PARP15 value)+3.571*(ZAP70 value).
By the ROC curve analysis (
Area Under the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
Under the nonparametric assumption b. Null hypothesis: true area=0.5
Coordinates of the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
The smallest cutoff value is the minimum observed test value minus 1, and the largest cutoff value is the maximum observed test value plus 1. All the other cutoff values are the averages of two consecutive ordered observed test values.
Therefore, when the y value exceeds the threshold of 0.1924607, the sample is considered positive and related to the presence of HG-SIL or OSCC.
Set (c) involves the following genes: KIF1A (CpG 241759621), LRRTM1 (CpG 80531799), ZAP70 (CpG 98340854), MIR193 (CpG 29886944), MIR296 (CpG 57392374), TERT (CpG 1279758).
Canonical Discriminant Function Coefficients
By the ROC curve analysis (
Area Under the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
y=−3.474+1.844*(LRRTM1 value)+2.852*(miR193 value)+5.441*(1−miR296 value)+3.656*(ZAP70 value)+1.744*(KIF1A value)+0.228*(1−hTERT value)
Therefore, when the y value exceeds the threshold of −0.2394214 (sensitivity: 0.943; 1-specificity: 0.077), the sample should be considered positive and related to the presence of HG-SIL or OSCC.
Set (d) involves the following genes: KIF1A (CpG 241759621), LRRTM1 (CpG 80531799), FLI1 (CpG 128564158), LINC00599 (CpG 9760888), PARP15 (CPG 122296586), ZAP70 (CpG 98340854), MIR193 (CpG 29886944), MIR296 (CpG 57392374), GP1BB (CpG 19710956), hTERT (CpG 1279758).
Canonical Discriminant Function Coefficients
By the ROC curve analysis (
Area Under the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
y=−3.977−2.630*(FLI1 value)+3.899*(1−GP1BB value)+1.631*(LRRTM1 value)−0.443*(LINC00599 value)+0.625*(miR193 value)+8.743*(1−miR296 value)+1.995*(PARP15 value)+1.996*(ZAP70 value)+1.585*(KIF1A value)−0.083*(1−hTERT value)
Therefore, when the y value exceeds the threshold of 0.6760154 (sensitivity: 0.971; 1-specificity: 0.015), the sample should be considered positive and related to the presence of HG-SIL or OSCC.
Set (e) involves the following genes with their CpGs: 19710956 for GP1 BB; 98340854 for ZAP70; 241759621 for KIF1A; 1279758 for TERT; 80531799 for LRRTM1; 131781167 for NTM; 29886944 for MIR193; 122296586 for PARP15; 182322902 for ITGA4; 128564158 for FLI1; 9760888 for LINC0059; 15342885 for EPHX3; and 57392374 for MIR296.
Canonical Discriminant Function Coefficients
By the ROC curve analysis (
Area Under the Curve
Test Result Variable(s): Discriminant Scores from Function 1 for Analysis 1
y=−3.954−0.415*(EPHX3 value)−3.471*(FLI1 value)+3.801*(1−GP1BB value)+2.276*(ITGA4 value)+1.428*(LRRTM1 value)−0.036*(LINC00599 value)+0.894*(miR193 value)+7.938*(1−miR296 value)−0.637*(NTM value)+2.338*(PARP15 value)+1.983*(ZAP70 value)+1.789*(KIF1A value)−0.265*(1−hTERT value)
Therefore, when the y value exceeds the threshold of 0.8732193 (sensitivity: 0.971; 1-specificity: 0), the sample should be considered positive and related to the presence of HG-SIL or OSCC.
Multiple Range Tests Using the Score from the Combination “e”:
A Duncan's Multiple Range Test evaluating the final score of patients from different groups (65 Normal donors, 29 OSCC, 6 HGSIL, 30 distant mucosa in OSCC patients) has been calculated using the combination “e” as follows:
y=−3.954−0.415*(EPHX3 value)−3.471*(FLI1 value)+3.801*(1−GP1BB value)+2.276*(ITGA4 value)+1.428*(LRRTM1 value)−0.036*(LINC00599 value)+0.894*(miR193 value)+7.938*(1−miR296 value)−0.637*(NTM value)+2.338*(PARP15 value)+1.983*(ZAP70 value)+1.789*(KIF1A value)−0.265*(1−hTERT value)
This test identifies a statistical difference between the OSCC group vs healthy donors and vs contralateral mucosa. Furthermore, multiple range test shows a statistical difference between the HGSIL group and healthy donors and contralateral mucosa:
By these data, Multiple Range Test for LDA-generated scores (combination “e”) identifies a significant difference between the OSCC group and healthy donors and contralateral mucosa. Furthermore, multiple range test showed a significant difference between the HGSIL group and healthy donors and contralateral mucosa, demonstrating that OSCC and HGSIL have a different epigenetic behavior than normal donors and normal contralateral mucosa.
Algorithm Validation in an Independent Cohort
Considering the validation dataset, all normal donors were detected under the threshold value, as well as one oral fibroma and 12 out of 14 OLP; on the contrary the remaining two OLP, all PVL and all OSCC were positives as expected.
Kruskal-Wallis test (
| Number | Date | Country | Kind |
|---|---|---|---|
| 102016000111174 | Nov 2016 | IT | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2017/056875 | 11/3/2017 | WO |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2018/083646 | 5/11/2018 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 20080057590 | Urdea | Mar 2008 | A1 |
| 20110300536 | Li et al. | Dec 2011 | A1 |
| Number | Date | Country |
|---|---|---|
| 2011112880 | Sep 2011 | WO |
| Entry |
|---|
| Hu et al. BMC Bioinformatics 2015; 16: 220; DOI 10.1186/s12859-015-0649-2 (Year: 2015). |
| Guo et al. BMC Genomics 2013; 14: 774; doi: 10.1186/1471-2164-14-774. (Year: 2013). |
| Hernández et al. BioTechniques 2013; 55: 181-197. (Year: 2013). |
| Lee et al. Cancer Letters 2013; 340: 171-178. (Year: 2013). |
| Warnecke et al. Methods 2002; 27: 101-107. (Year: 2002). |
| Ehrich et al. Nucleic Acids Research 2007; 35: e29. (Year: 2007). |
| Cronn et al. American Journal of Botany 2012; 99: 291-311. (Year: 2012). |
| Hadd et al. The Journal of Molecular Diagnostics 2013; 15: 234-247. (Year: 2013). |
| Ergüner et al. Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2015; 6453-6456. (Year: 2015). |
| Morandi et al. Clinical Epigenetics 2017; 9: 85 (Year: 2017). |
| Gissi et al. International Journal of Molecular Sciences 2020; 21: 6691 (Year: 2020). |
| Morandi et al. Journal of Cranio-Maxillo-Facial Surgery 2015; 43: 1494-1500 (Year: 2015). |
| Qureshi et al. International Journal of Surgery 2010; 8: 194-198 (Year: 2010). |
| Gissi et al. 13-gene DNA Methylation Analysis from Oral Brushing: A Promising Non Invasive Tool in the Follow-up of Oral Cancer Patients. Journal of Clinical Medicine 2019; 8: 2107; doi:10.3390/jcm8122107. (Year: 2019). |
| Demokan, S., et al., “KIF1A and EDNRB are differentially methylated in primary HNSCC and salivary rinses”, International Journal of Cancer, vol. 127, No. 10, Nov. 15, 2010, pp. 2351-2359. |
| Jithesh P.V., et al., “The epigenetic landscape of oral squamous cell carcinoma”, British Journal of Cancer, vol. 108, No. 2, Jan. 3, 2013, pp. 370-379. |
| Khor, G.H., et al., “DNA methylation profiling revealed promoter hypermethylation-induced silencing of p16, DDAH2 and DUSP1 in primary oral squamous cell carcinoma”, International Journal of Medical Sciences, vol. 10, No. 12, Jan. 1, 2013, pp. 1727-1739. |
| Search Report and Written Opinion of PCT/IB2017/056875 of May 3, 2018. |
| Number | Date | Country | |
|---|---|---|---|
| 20200299773 A1 | Sep 2020 | US |
