RESEARCH ARTICLE

Genetic imbalances detected by multiplex ligation-dependentprobe amplification in a cohort of patients with oral squamouscell carcinoma—the first step towards clinicalpersonalized medicine

Ilda Patrícia Ribeiro & Francisco Marques & Francisco Caramelo & José Ferrão &

Hugo Prazeres & Maria José Julião & Widad Rifi & Suvi Savola & Joana Barbosa de Melo &

Isabel Poiares Baptista & Isabel Marques Carreira

Received: 10 October 2013 /Accepted: 3 January 2014# International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract Oral tumors are a growing health problem world-wide; thus, it is mandatory to establish genetic markers inorder to improve diagnosis and early detection of tumors,control relapses and, ultimately, delineate individualized ther-apies. This study was the first to evaluate and discuss theclinical applicability of a multiplex ligation-dependent probeamplification (MLPA) probe panel directed to head and neckcancer. Thirty primary oral squamous cell tumors were ana-lyzed using the P428MLPA probe panel. We detected geneticimbalances in 26 patients and observed a consistent pattern ofdistribution of genetic alterations in terms of losses and gainsfor some chromosomes, particularly for chromosomes 3, 8,and 11. Regarding the latter, some specific genes werehighlighted due to frequent losses of genetic material—RARB,FHIT, CSMD1, GATA4, and MTUS1—and others due to

gains—MCCC1,MYC,WISP1, PTK2, CCND1, FGF4, FADD,and CTTN. We also verified that the gains ofMYC andWISP1genes seem to suggest higher propensity of tumors localizedin the floor of the mouth. This study proved the value of thisMLPA probe panel for a first-tier analysis of oral tumors. Theprobemix was developed to include target regions that havebeen already shown to be of diagnostic/prognostic relevancefor oral tumors. Furthermore, this study emphasized several ofthose specific genetic targets, suggesting its importance to oraltumor development, to predict patients’ outcomes, and also toguide the development of novel molecular therapies.

Keywords Genetic imbalances .MLPAtechnique .Gainsandlosses . Oral squamous cell carcinoma

I. P. Ribeiro : J. Ferrão : J. B. de Melo : I. M. CarreiraCytogenetics and Genomics Laboratory, Faculty of Medicine,University of Coimbra, 3000-354 Coimbra, Portugal

I. P. Ribeiro : F. Marques : J. B. de Melo : I. P. Baptista :I. M. Carreira (*)Center of Investigation on Environment Genetics and Oncobiology(CIMAGO). Faculty of Medicine, University of Coimbra,3000-354 Coimbra, Portugale-mail: citogenetica@fmed.uc.pt

I. M. Carreirae-mail: icarreira@fmed.uc.pt

F. Marques : I. P. BaptistaDepartment of Dentistry, Faculty of Medicine, University ofCoimbra, 3000-075 Coimbra, Portugal

F. MarquesStomatology Unit, Coimbra Hospital and University Centre, CHUC,EPE, 3000-075 Coimbra, Portugal

F. CarameloLaboratory of Biostatistics and Medical Informatics, IBILI—Facultyof Medicine, University of Coimbra, 3000-354 Coimbra, Portugal

H. PrazeresMolecular Pathology Laboratory, Portuguese Institute of Oncologyof Coimbra FG, EPE, 3000-075 Coimbra, Portugal

M. J. JuliãoDepartment of Pathology, Coimbra Hospital and University Centre,CHUC, EPE, 3000-075 Coimbra, Portugal

W. Rifi : S. SavolaDepartment of Tumor Diagnostics, MRC-Holland,1057DN Amsterdam, The Netherlands

Tumor Biol.DOI 10.1007/s13277-014-1614-9

Introduction

Oral tumors belong to head and neck cancers, of which morethan 90 % are squamous cell carcinoma [1]. The overall inci-dence of oral cancer seems to be rising worldwide: for 2013,36,000 new cases and 6,850 deaths are estimated, including thelip, oral cavity, and pharynx cancer [2]. This shows that oralsquamous cell carcinoma (OSCC) is an aggressive neoplasmfrequently diagnosed at an advanced stage and associated withhigh mortality and morbidity. Its high incidence in young peoplehas been correlated with tobacco, alcohol consumption and,mostly, with human papilloma virus (HPV) infection [3]. Atthe time of diagnosis, more than half of oral cancer patients havelocal or distant metastases, which means a 5-year survival rate ofonly 33.2 % [2]. Albeit the technological progress and theimprovements in therapeutic modalities, the survival rates of oralcancer have remained the same in recent decades, which couldindicate that the markers of prognosis, such as tumor size, lymphnode involvement, and tumor stage, are not enough to predictclinical outcome, choose the best treatment and, least of all, toimprove the early detection. Genetic classification of oral tumorsidentifies subgroups with a different prognosis [4]. In this re-spect, as oral cancer is a heterogeneous disease, attempts havebeen made to identify the genetic imbalances associated with thehistological progression of oral cancer, but it remains imperativeto establish biomarkers for oral tumors, in order to choose themost appropriate clinical therapy according to each patient’sgenetic profile of tumor. Nowadays, in order to facilitate theearly detection and the management of oral tumors, potentiallyuseful diagnostic tools at clinical and molecular levels have beendeveloped [5]. Multiplex ligation-dependent probe amplification(MLPA) is a molecular technique developed in 2002 bySchouten et al. [6] that allows amplification and semi-quantificative detection of up to 50 sequences of DNA in thesame reaction with a single pair of primers. This technique hasbeen shown to be applicable not only for the diagnostic ofinherited or congenital diseases, but also for tumor diagnosticand prediction of cancer progression [7]. This study is the firstone to perform an evaluation of the genetic profile of oral cavitycarcinomas through copy number variations using a MLPAprobe panel specific for head and neck cancer, in a cohort of30 patients. We also discuss the routine clinical application ofthis MLPA probe panel, highlighting the most frequent imbal-ances detected in these tumors and, additionally, correlating thesegenetic alterations with clinical characteristics.

Material and methods

Patients and DNA isolation

The present study was conducted on 30 primary OSCC sam-ples. These samples were obtained between 2010 and 2012

from the Maxillofacial Surgery and Stomatology Unit, of theCoimbra Hospital and University Centre, CHUC, EPE, Por-tugal. All patients were submitted to surgery and the histo-pathologic diagnosis of the mirror sections of these fragmentswas performed by two different pathologists. Hematoxylinand eosin staining was used to evaluate tumor content in eachspecimen. In our set, all samples contained at least 50% tumorcells. Diagnosis and staging were performed in accordancewith the American Joint Committee on Cancer’s TNM stagingsystem [8] for OSCCs. The patients gave their informedconsent in accordance with the regulations in the Dec-laration of Helsinki. The study was approved by theethics committee of the Faculty of Medicine of theUniversity of Coimbra. The detailed characterization ofour cohort with OSCC diagnosis is illustrated in Table 1.Controls were 15 gingival tissues from healthy subjectssubmitted to “wisdom teeth” removal. DNA from pa-tients and controls were extracted from fresh frozentissue using a High Pure PCR Template PreparationKit (Roche GmbH, Mannheim, Germany), according tothe manufacturer’s instructions and quantified by

Table 1 Patient and tumor characteristics

Patients (n=30)

Median age, years (range) 63.1 (37–84)

Sex

Male 26

Female 4

Smoking (cigarettes/day)

≥ 20 17

< 20 3

None 10

Stage

I and II 9

III and IV 21

Site

Tongue 13

Floor of the mouth 12

Buccal mucosa 4

Palate 1

Treatment

Surgery + RT 7

CT + surgery 2

Surgery + RT + CT 6

Surgery only 15

Clinical outcome

Death from the disease 10

Death from other causes 1

Alive 19

RT radiation therapy, CTchemotherapy

Tumor Biol.

UV spectrophotometric analysis using Nanodrop 1000Spectrophotometer (Thermo Scientific, USA).

MLPA probemix development and MLPA analysis

MLPA analysis was performed using a MLPA probe panelspecific for head and neck squamous cell carcinoma (HNSCC). This MLPA assay was designed and optimized at MRC-Holland (Amsterdam, The Netherlands). For selection ofprobes included in this assay, an extensive literature searchwas performed to pinpoint genes, which are suggested to be ofdiagnostic and/or prognostic importance in HNSCC. Theprobes included in this assay were selected from the MRC-Holland MLPA probe database, and if there were no existingMLPA probes available, new probes were designed. All theMLPA probes included in this assay were tested under variousconditions by changing hybridization temperature as well assalt, probe, and polymerase concentration. Moreover, the vari-ability of these MLPA probes was tested on healthy individuals(n=24), and only probes that were shown to have standarddeviation under 0.10 were included in this assay. After theoptimization and quality testing phase, this P428-B1-lot1111MLPA panel included 41 probes targeting 36 different genes,

located on chromosomes 3, 4, 5, 7, 8, 11, 13, and 18 (Fig. 1).Additionally, 11 references probes were included, which detect10 different autosomal chromosomal regions considered rela-tively quiet in HNSCC. This MLPA assay contains alsonine quality control probes to assess DNA denaturationand DNA quantity in the MLPA reaction. All MLPAreactions were performed according to the standardMLPA reaction protocol described by Schouten et al.[6] Briefly, 100 ng of DNA samples were heated at98 °C for 10 min. After the addition of the probemix,samples were heated for 1 min at 95 °C and thenincubated for 16 h at 60 °C. Ligation of the annealedoligonucleotide probes was performed for 15 min at 54 °C in abuffer containing Ligase-65 enzyme.After inactivating the ligaseenzyme by heating at 98 °C for 5 min, multiplex PCR wascarried out using FAM-labeled primers, dNTPs, and SALSApolymerase. PCR was performed for 35 cycles of 30 s at 95 °C,30 s at 60 °C, and 1 min at 72 °C. All the reactions were carriedout in a thermal cycler equipped with a heat lid (ABI 2720,Applied Biosystems, Foster City, CA, USA). PCR productswere heat-denatured and analyzed on a Gene Scan ABI PRISM3130 capillary electrophoresis system (Applied Biosystems,Foster City, CA, USA). Three controls and a negative control

Fig. 1 Chromosomal distribution of the genes studied using P428 MLPA probe panel as well as references probes (Ref)

Tumor Biol.

(without DNA) were always included in each MLPA experi-ment. The sample results are displayed as a ratio between thereference and experimental samples. Data binning of raw dataand comparative analysis was performed using Coffalyser.NETsoftware [9]. For each MLPA probe, we determined specificcutoff values for gain and loss, using the values limiting the95 % confidence interval as determined on noncancer subjects.A numerical gain was scored when the values exceeded 1.2,and a numerical loss was defined when the values were lowerthan 0.8.

HPV typing

All tumor tissue samples were analyzed for HPV infectionas described by Nobre et al. [10]. Briefly, we used PCRmethods employing established general consensus and de-generate primer sets , namely GP5+/GP6+ andMY09/MY11, which were designed to amplify a fragmentof the L1 gene of mucosatropic HPVs. For genotyping,we performed the sequencing of intra-primer DNA in theamplified fragments in order to determine the specifictypes of HPV present in the samples. In addition toDNA sequencing, we complemented our analysis by usingDNA array hybridization, using HPV CLART2 arrays(Genomica), as a way to address cases that showed infec-tion with multiple HPV genotypes.

Statistical analysis

The statistical analysis was carried out with IBM SPSS v.20software package, assuming a significance level of 0.05. De-scriptive analysis of loss and gain frequencies in genes, ineach mouth location and tumor staging, was performed. Inaddition, we created convenient charts for better understand-ing of the relationships between genes, locations, and tumorstaging. In order to verify the existence of an association, withstatistical meaning, between gains/losses in genes and loca-tion, a chi-square test was made. The same test was alsoapplied to the association between gains/losses in genes andtumor staging.

Results

Genetic profile of oral tumor tissue

This study showed the existence of genetic imbalances in26 of the 30 oral tumor specimens analyzed (Fig. 2a).Four tumor samples (13.3 %, 4 of 30 specimens) did notharbor copy number aberrations in the target regions ofthe MLPA probemix. These patient samples were furtheranalyzed with an oligonucleotide microarray 180 K(Agilent Technologies, Santa Clara, CA, USA), and we

did not detect copy number alterations in the regionstargeted by MLPA assay (data not shown). Simultaneous-ly, we did not observe any genetic imbalance for the 36genes analyzed with this MLPA probe panel in all the 15gingival tissues from healthy subjects submitted to“wisdom teeth” removal (data not shown), which corre-spond to the control group.

Regarding the genetic profile of tumor samples, overall, weverified that the number of tumors that carried gains of geneticmaterial was higher than the number of tumors that carried losses(Fig. 2b). It was also possible to observe a pattern of imbalancesin some chromosomes; therefore, the distribution of genetic

Fig. 2 Genetic imbalances (gains and losses) in the 30 OSCC patientsdetected using P428 MLPA probe panel. Losses of genetic material arerepresented by light gray, and dark gray represents gains. a Each rowrepresents one patient and each pixel in every rowcorresponds to one geneanalyzed. The gray color represents genes without alteration, and eachshade of gray shows the localization of the genes in each specificchromosome. The genes analyzed, from left to right, are ordered bychromosome from the short arm to the long arm; b the percentage ofthe imbalances in all genes analyzed. Each line represents one geneanalyzed

Tumor Biol.

alterations in terms of losses and gains was very consistent forsome chromosomes (Fig. 2a). In this sense, 3p and 8pshowed mostly losses, while 3q and 8q frequentlydisplayed gains. Chromosomes 4, 5, 13, and 18 present-ed mainly losses for the analyzed genes. On the otherhand, chromosome 7 showed more frequent gains thanlosses for the analyzed genes. Regarding chromosome11, in band q13, we verified almost exclusively gains,and in q-distal, we observed predominantly losses. Re-garding all these imbalances (Fig. 2a, b), we verifiedthat the most frequently altered chromosomes were 3, 8,and 11. Moreover, chromosome 8 showed the highestfrequency of gains (70 %) in the largest number ofpatients. The second most frequently altered chromo-some in terms of gains was chromosome 11 (66.7 %),followed by 3q (46.7 %). Regarding the losses of ge-netic material, the most commonly altered chromosomewas 3p (40 %), followed by 8p (36.7 %). No singlegene was altered in all patients. On the other hand, allgenes analyzed presented gain and/or loss of geneticmaterial in at least three patients. It is important tonotice that the following genes only presented exclu-sively loss of genetic material in these oral tumor sam-ples: RARB (3p24.2), RASSF1 (3p21.31), FHIT (3p14.2),DEPDC1B (5q12.1), WDE36 (5q22.1), BTNL3 (5q35.3),ATM (11q22.3), and SMAD4 (18q21.2). Additionally, thefollowing genes were the ones that presented exclusive-ly gain of genetic material: CCNL1 (3q25.32), PIK3CA(3q26.33), MCCC1 (3q26.33), CDK6 (7q21.12), andCTTN (11q13). The remaining 23 genes analyzed in thisstudy presented both gain and loss of genetic material(Fig. 2a, b); however, for some of those genes, it waspossible to observe that one specific event is muchmore prevalent than the other, e.g., gain or loss. In thissense, after analyzing in detail every gene in the mostcommonly altered chromosomes, we verified that fromthe 36 genes studied, it was possible to highlight someof those that frequently showed gains or losses ofgenetic material. Thus, taking only into account the genesaltered in ≥30 % of the oral tumor samples (Fig. 2a, b), thefollowing genes were the ones to present the most frequentlosses: RARB (3p24.2), FHIT (3p14.2), CSMD1 (8p23.1),GATA4 (8p23.1), and MTUS1 (8p22); and these were the onesthat presented gains: MCCC1 (3q26.33), MYC (8q24.21),WISP1 (8q24.22), PTK2 (8q24.3), CCND1 (11q13.3), FGF4(11q13.3), FADD (11q13.3), and CTTN (11q13.3).

Association between clinicopathological features and geneticimbalances

Overall, regarding the different anatomic localizations andstages of tumors (Fig. 3), it was not possible to reach statisticalsignificance in order to discriminate the patients according to

their genetic profile and clinicopathologic characteristics. Tak-ing into account only three anatomic localizations (buccalmucosa, floor of the mouth, and tongue)—and excluding thepalate due to the fact that only one tumor was located there—and performing correction for multiple comparisons, we didnot observe any gene with statistical significance to performthis separation; however, if the correction of multiple compar-isons (Bonferroni correction) was not applied, two genes fromchromosome 8 showed statistical significance. In this sense,gain in these two genes, MYC and WISP1, could suggest thatthe tumor is predominantly localized in the floor of the mouth.Regarding HPV infection, we only detected two HPV-positivepatients from the 30 patients analyzed (data not shown). Oneshowed a HPV type belonging to high-risk classes, HPV type31, and another belonging to low-risk classes, HPV type 42.The latter HPV patient did not have any other risk factors,such as tobacco or alcohol consumption.

Discussion

Genetic profile

Oral cavity tumors develop through multistep genetic path-ways, involving typically losses of tumor suppressor genesand gains of oncogenes. Presently, the probe panel P428 is theonly one exclusively developed MLPA assay for head andneck cancer. This is, therefore, the first study using this MLPAprobe panel in patients with OSCC diagnosis. In our patientcohort, the most frequently altered chromosomes were 3, 8,and 11.We observed losses at 3p and gains at 3q, a pattern thathad already been previously described for HNSCC [11]. Im-balances at 3p are associated to the early HNSCC [12]. In ourcohort, the most common losses in 3p included FHIT (40 %)and RARB (33.3 %) tumor suppressor genes. FHIT is frequent-ly deleted in epithelial cancer cell lines [13]. In 1996, Virgilioet al. [14] described that 55 % of HNSCC cell lines expressaberrant FHIT transcripts and that one or both FHITalleles aredeleted in many of these cell lines, which suggests that inac-tivation ofFHIT is important for development and progressionof HNSCC. Also, Saldivar et al. [15,16] have shown that lossof expression of Fhit protein causes DNA damage and ge-nome instability and otherwise expression of Fhit reduces thisDNA damage and contributes to protection of genome stabil-ity; thus, loss of Fhit provides a selective advantage in spo-radic cancers. Concerning RARB, in 1995, Lotan et al. [17]verified that 60 % of oral potential malignant lesions did notexpress this gene, in comparison to all the samples of normaltissue where this gene was expressed. This way, loss of RARBexpression could enhance carcinogenesis through loss of re-sponse to retinoids [18]. Hence, this receptor could eventuallybe a valuable intermediate marker in trials of retinoids for theprevention of oral carcinogenesis. This point needs to be more

Tumor Biol.

carefully explored, since retinoids have proven activity intreating early potential malignant lesions [19]. Additionally,we detected in a wide range of copy number gains at 3q,namely at MCCC1 (46.7 %). In our cohort, concerning 3q,this gene was the one that presented the highest frequency ofgains, besides aberrations in other genes, such as PIK3CA hasalready been described as extremely important for oral tumordevelopment [20].

We observed frequent losses in 8p, which is in line withliterature describing deletions at 8p in oral tumors [21]. In ourcohort, the most frequently deleted genes in 8p were CSMD1(36.7 %), GATA4 (33.3 %), and MTUS1 (30 %). CSMD1 isconsidered to be a strong candidate at 8p23 [22], and loss ofthis gene is also present in other epithelial cancers, namely inbreast and lung cancer [23]. GATA4 has been implicated incolorectal and gastric development, where it is epigeneticallysilenced and expected to contribute to tumor progression [24];however, its function in human cancers is not yet fully under-stood. A tumor suppressor role forGATA4has been described,due to the methylation and loss of expression of this geneobserved in lung, colorectal, and gastric cancers [24–26].

MTUS1 gene was initially identified as a candidate tumorsuppressor gene in pancreatic cancer; moreover, the ectopicexpression ofMTUS1gene products has been shown to inhibitcell proliferation [27]. In 2006, Zhou et al. [28] hypothesizedthat the reduction of MTUS1 expression could be associatedwith advanced oral tongue squamous cell carcinoma. In 2007,Ye et al. [29] suggested thatMTUS1 gene is a potential tumorsuppressor gene also for HNSCC and a promising candidatefor further functional analysis. Regarding the long arm ofchromosome 8, we observed more frequently gains of thegenetic material in our cohort, which is in line with theliterature [30]. The altered genes mapped in 8q in the largestnumber of patients were MYC (70 %), WISP1 (43.3 %), andPTK2 (43.3 %). Nowadays, the MYC gene is considered themost significant oncogene mapped on 8q and has been asso-ciated to the pathogenesis of several human cancers [31].Amplification and overexpression of this gene has been ob-served in 10–40% of human oral tumors [32]. In the literature,tongue carcinomas showed the highest incidence of gains(75 %) in MYC gene among oral tumors, resulting in signifi-cantly poor survival rates [33]. In our cohort, the highest

Fig. 3 Radial heat map of the genetic imbalances for tumor samples inthe three anatomic localizations (buccal mucosa, floor of the mouth, andtongue) and in stage I or II and in stage III or IVof the tumors. Each line

represents one patient. Light gray lines represent losses of geneticmaterialand dark gray lines gains

Tumor Biol.

incidence of gains in MYC gene was observed in the floor ofthe mouth. Regarding the role of WISP1 gene, it has beenobserved that this gene could enhance or inhibit tumor growth.In colon carcinoma cell lines and in colon tumors, a significantincrease of genomic copies and mRNA of WISP1 was ob-served, in comparison to normal mucosa [34]. In melanomacells, WISP1 expression was inversely correlated with prolif-eration, metastasis and growth, as well as with metastasis oflung cancer [35,36]. Further studies will be needed todetermine if WISP1 has, in fact, prognostic value inpredicting oral cancer metastasis. Amplification inPTK2 (alias, FAK) gene has already been reported inHNSCC [37]. Similarly, FAK overexpression has beenshown in tumor biopsy samples from a wide variety oftumors [37]. This overexpression has been associatedwith the invasive potential of the tumor [38].

In our patient cohort, gains at 11q13 were detectedand included CCND1 (56.7 %), FGF4 (60 %), FADD(66.7 %), and CTTN (66.7 %). Gains at band 11q13 arefrequent in human cancer, and this event is also ex-tremely common in HNSCC [39]. CCND1 and CTTNgenes were the first to be identified [39] and reported ascandidates for driving 11q13.3 amplification, due to thefact that they are both recurrently co-amplified, and thisamplification has been correlated with its overexpression[40]. Amplification of 11q13 occurs in 30–60 % ofHNSCC, which also includes FGF4 [41]. In 2007,Gibcus et al. [40] identified the FADD gene as a poten-tial driver gene in the 11q13 amplicon for head andneck cancer. Similar to this study, also in our cohort,the FADD gene presented gains in the highest number ofpatients, when compared to the other genes analyzed inthis chromosome. This gene plays a significant role incell cycle regulation, which suggests its importance inthe response to cytotoxic drugs. Thus, Gibcus et al.[40] hypothesized that HNSCC patients with 11q13.3amplification and concomitant FADD overexpressioncould benefit from the administration of Taxol-basedchemoradiotherapy over radiotherapy alone. Similarly,loss of distal 11q contributes to chromosomal instabilityand consequently to tumor progression as well as toresistance to therapy, namely reduced sensitivity toionizing radiation [42].

Correlation between genetic imbalancesand clinicopathological features

Taking into account the genetic differences among the ana-tomic localizations, we have in our cohort a relatively lownumber of patients distributed by the four anatomic localiza-tions in the oral cavity, which hinder this interpretation. In2002, Huang et al. [43] found evidence indicating that subsetsof HNSCC have different genetic patterns, allowing to

perform the distinction by site of disease within the upperaerodigestive tract. In line with that, we believe that somegains and losses could be tumor site specific (Fig. 3); however,in this first pilot study, due to the few patients enrolled in eachanatomic site, it was not possible to reach statistical signifi-cance in order to molecularly identify different subgroups ofthe oral cavity carcinoma. Besides that, in chromosome 8,gains of MYC and WISP1 genes seem to suggest higherpropensity of tumors localized in the floor of the mouth. Thisfinding could be extremely important from a clinical point ofview because it helps to subdivide the patients according tothese clinicopathological characteristics, which could signifydifferent prognoses and, ultimately, different therapies.Concerning the genetic profile between tumors in stage I orII and in stage III or IV (Fig. 3), the lack of statistical signif-icance is probably due to the small number of patients enrolledin each stage. In this sense, further studies evaluating a largercohort seem to be crucial in order to separate these tumorsaccording to genetic profile. Regarding HPV status, we onlyidentified two HPV-positive patients. Although it has beendescribed that HPV-positive and HPV-negative tumors exhib-ited distinct clinicopathological and molecular entities [44],we cannot clarify whether HPV-positive patients have indeeda relatively favorable prognosis due to the presence of specificgenetic imbalances (Fig. 2a). Nowadays, however, some clin-icopathologic variables have been validated to classify andprognosticate oral tumor patients, the predictive value ofwhich are, in general, very low [40,45]. We believe that witha higher number of patients, it might be possible to molecu-larly identify different subgroups of OSCC with differentclinical relevance.

Detection of genetic imbalances—a step forward towardsclinical personalized medicine

There is a very urgent need to establish genetic markersin order to improve diagnosis, help in the redefinitionand reclassification of histology and stage of tumors,and predict therapeutic outcome of oral tumor patients.Our results with this MLPA probe panel gave a stepforward in this direction. MLPA is a cost-efficient tech-nique with little hands-on time and compatible withroutine diagnostic. Another extremely important advan-tage in tumor analysis is the capability of this techniqueto identify imbalances even in samples with wild-typetissue contamination [7,46] as well as to provide infor-mation about intra-tumoral mosaicism.

We can conclude that our study showed the value of theMLPA technique for the analysis of oral tumor tissue in orderto detect imbalances in specific genomic regions. Moreover,our study emphasized several specific genetic targets suggest-ing its putative importance to improve diagnostic, predictpatients’ outcome, and also guide the development of novel

Tumor Biol.

molecular therapies, which must be further assessed and val-idated in a larger cohort of patients. This MLPA probe panelspecific for HNSCC detected genetic imbalances in 86.7 % ofsamples of our patient cohort, encompassing some genes thatseem to be associated to early stage of oral tumors, others thatare likely to be associated to the development of metastasis,and others possibly associated to therapeutic response, whichcould in the future represent the first step to help in theintroduction of a clinical personalized medicine. Besides thesepromising results, the need to establish more specific bio-markers for these oral squamous tumors remains evident inorder to implement a simple way of screening a wide range ofimbalances with clinical utility not only in terms of diagnosisand prognosis but also to follow up high-risk populations.

Acknowledgments The authors are grateful to Dr. Artur Ferreira, Di-rector of Maxillofacial Surgery Unit from Coimbra Hospital and Univer-sity Centre, for the contribution in the collection of the samples. Thiswork was supported in part by CIMAGO (Center of Investigation onEnvironment Genetics and Oncobiology, Faculty ofMedicine, Universityof Coimbra).

Conflicts of interest W. Rifi and S. Savola are employed by MRC-Holland, manufacturer of commercially availableMLPA probemixes. Allother authors declare no competing financial interests.

References

1. Forastiere A, Koch W, Trotti A, Sidransky D. Head and neck cancer.N Engl J Med. 2001;345(26):1890–900.

2. Mitka M. Evidence lacking for benefit from oral cancer screening.JAMA. 2013;309(18):1884. doi:10.1001/jama.2013.4913.

3. Llewellyn CD, Johnson NW, Warnakulasuriya KA. Risk factors forsquamous cell carcinoma of the oral cavity in young people—acomprehensive literature review. Oral Oncol. 2001;37(5):401–18.

4. Wittekindt C, Wagner S, Mayer CS, Klussmann JP. Basics of tumordevelopment and importance of human papilloma virus (HPV) forhead and neck cancer. GMS Curr Top Otorhinolaryngol Head NeckSurg. 2012;11:Doc09. doi:10.3205/cto000091.

5. Shah FD, BegumR, Vajaria BN, Patel KR, Patel JB, Shukla SN, et al.A review on salivary genomics and proteomics biomarkers in oralcancer. Indian J Clin Biochem. 2011;26(4):326–34. doi:10.1007/s12291-011-0149-8.

6. Schouten JP, McElgunn CJ, Waaijer R, Zwijnenburg D, Diepvens F,Pals G. Relative quantification of 40 nucleic acid sequences bymultiplex ligation-dependent probe amplification. Nucleic AcidsRes. 2002;30(12):e57.

7. Homig-Holzel C, Savola S. Multiplex ligation-dependent probe amplifi-cation (MLPA) in tumor diagnostics and prognostics. DiagnMol Pathol.2012;21(4):189–206. doi:10.1097/PDM.0b013e3182595516.

8. Wittekind C, Greene FL, Hutter RVP, Klimpfinger M. LH. S. TNMatlas. Illustrated guide to the TNM/pTNM classification of malignanttumours. 5th ed. Berlin: Springer; 2003.

9. Coffa J, Berg J. Analysis of MLPA data using novel softwareCoffalyser.NET by MRC-Holland. In: Eldin AB, editor. Modernapproaches to quality control. Rijeka, Croatia: InTech; 2011. p.125–50.

10. Nobre RJ, Cruz E, Real O, de Almeida LP, Martins TC.Characterization of common and rare human papillomaviruses in

Portuguese women by the polymerase chain reaction, restrictionfragment length polymorphism and sequencing. J Med Virol.2010;82(6):1024–32. doi:10.1002/jmv.21756.

11. Bockmuhl U, Petersen S, Schmidt S, Wolf G, Jahnke V, Dietel M,et al. Patterns of chromosomal alterations in metastasizing andnonmetastasizing primary head and neck carcinomas. Cancer Res.1997;57(23):5213–6.

12. Redon R, Muller D, Caulee K, Wanherdrick K, Abecassis J, duManoir S. A simple specific pattern of chromosomal aberrations atearly stages of head and neck squamous cell carcinomas: PIK3CAbut not p63 gene as a likely target of 3q26-qter gains. Cancer Res.2001;61(10):4122–9.

13. OhtaM, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, et al.The FHIT gene, spanning the chromosome 3p14.2 fragile site andrenal carcinoma-associated t(3;8) breakpoint, is abnormal in digestivetract cancers. Cell. 1996;84(4):587–97.

14. Virgilio L, Shuster M, Gollin SM, Veronese ML, Ohta M, HuebnerK, et al. FHIT gene alterations in head and neck squamous cellcarcinomas. Proc Natl Acad Sci U S A. 1996;93(18):9770–5.

15. Saldivar JC, Bene J, Hosseini SA,Miuma S, Horton S, Heerema NA,et al. Characterization of the role of Fhit in suppression of DNAdamage. Adv Biol Regul. 2013;53(1):77–85. doi:10.1016/j.jbior.2012.10.003.

16. Saldivar JC, Miuma S, Bene J, Hosseini SA, Shibata H, Sun J, et al.Initiation of genome instability and preneoplastic processes throughloss of Fhit expression. PLoS Genet. 2012;8(11):e1003077. doi:10.1371/journal.pgen.1003077.

17. Lotan R, Xu XC, Lippman SM, Ro JY, Lee JS, Lee JJ, et al.Suppression of retinoic acid receptor-beta in premalignant oral le-sions and its up-regulation by isotretinoin. N Engl J Med.1995;332(21):1405–10. doi:10.1056/NEJM199505253322103.

18. Zou CP, Youssef EM, Zou CC, Carey TE, Lotan R. Differentialeffects of chromosome 3p deletion on the expression of the putativetumor suppressor RAR beta and on retinoid resistance in humansquamous carcinoma cells. Oncogene. 2001;20(47):6820–7. doi:10.1038/sj.onc.1204846.

19. O’Shaughnessy JA, Kelloff GJ, Gordon GB, Dannenberg AJ, HongWK, Fabian CJ, et al. Treatment and prevention of intraepithelialneoplasia: an important target for accelerated new agent develop-ment. Clin Cancer Res. 2002;8(2):314–46.

20. Murugan AK, Hong NT, Fukui Y, Munirajan AK, Tsuchida N.Oncogenic mutations of the PIK3CA gene in head and neck squa-mous cell carcinomas. Int J Oncol. 2008;32(1):101–11.

21. Hermsen M, Guervos MA, Meijer G, Baak J, van Diest P, MarcosCA, et al. New chromosomal regions with high-level amplificationsin squamous cell carcinomas of the larynx and pharynx, identified bycomparative genomic hybridization. J Pathol. 2001;194(2):177–82.

22. Sun PC, Uppaluri R, Schmidt AP, Pashia ME, Quant EC, SunwooJB, et al. Transcript map of the 8p23 putative tumor suppressorregion. Genomics. 2001;75(1–3):17–25. doi:10.1006/geno.2001.6587.

23. Ma C, Quesnelle KM, Sparano A, Rao S, ParkMS, CohenMA, et al.Characterization CSMD1 in a large set of primary lung, head andneck, breast and skin cancer tissues. Cancer Biol Ther. 2009;8(10):907–16.

24. \Akiyama Y, Watkins N, Suzuki H, Jair KW, van Engeland M,Esteller M, et al. GATA-4 and GATA-5 transcription factorgenes and potential downstream antitumor target genes areepigenetically silenced in colorectal and gastric cancer. MolCell Biol. 2003;23(23):8429–39.

25. Hellebrekers DM, Lentjes MH, van den Bosch SM, Melotte V,Wouters KA, Daenen KL, et al. GATA4 and GATA5 are potentialtumor suppressors and biomarkers in colorectal cancer. Clin CancerRes. 2009;15(12):3990–7. doi:10.1158/1078-0432.CCR-09-0055.

26. Zheng R, Blobel GA. GATATranscription factors and cancer. Genes& Cancer. 2010;1(12):1178–88. doi:10.1177/1947601911404223.

Tumor Biol.

27. Seibold S, Rudroff C, Weber M, Galle J, Wanner C, Marx M.Identification of a new tumor suppressor gene located at chromosome8p21.3–22. FASEB J. 2003;17(9):1180–2. doi:10.1096/fj.02-0934fje.

28. Zhou X, Temam S, Oh M, Pungpravat N, Huang BL, Mao L, et al.Global expression-based classification of lymph node metastasis andextracapsular spread of oral tongue squamous cell carcinoma.Neoplasia. 2006;8(11):925–32. doi:10.1593/neo.06430.

29. Ye H, Pungpravat N, Huang BL, Muzio LL, Mariggio MA, Chen Z,et al. Genomic assessments of the frequent loss of heterozygosityregion on 8p21.3–p22 in head and neck squamous cell carcinoma.Cancer Genet Cytogenet. 2007;176(2):100–6. doi:10.1016/j.cancergencyto.2007.04.003.

30. Squire JA, Bayani J, Luk C, Unwin L, Tokunaga J, MacMillan C,et al. Molecular cytogenetic analysis of head and neck squamous cellcarcinoma: by comparative genomic hybridization, spectralkaryotyping, and expression array analysis. Head Neck.2002;24(9):874–87. doi:10.1002/hed.10122.

31. da Silva Veiga LC, Bergamo NA, dos Reis PP, Kowalski LP, RogattoSR. DNA gains at 8q23.2: a potential early marker in head and neckcarcinomas. Cancer Genet Cytogenet. 2003;146(2):110–5.

32. Saranath D, Panchal RG, Nair R, Mehta AR, Sanghavi V, Sumegi J,et al. Oncogene amplification in squamous cell carcinoma of the oralcavity. Jpn J Cancer Res. 1989;80(5):430–7.

33. Vora HH, Shah NG, Patel DD, Trivedi TI, Chikhlikar PR. Prognosticsignificance of biomarkers in squamous cell carcinoma of the tongue:multivariate analysis. J Surg Oncol. 2003;82(1):34–50. doi:10.1002/jso.10183.

34. Babic AM, KireevaML, Kolesnikova TV, Lau LF. CYR61, a productof a growth factor-inducible immediate early gene, promotes angio-genesis and tumor growth. Proc Natl Acad Sci U S A. 1998;95(11):6355–60.

35. Hashimoto Y, Shindo-Okada N, Tani M, Takeuchi K, Toma H,Yokota J. Identification of genes differentially expressed in associa-tion with metastatic potential of K-1735 murine melanoma by mes-senger RNA differential display. Cancer Res. 1996;56(22):5266–71.

36. Soon LL, Yie TA, Shvarts A, Levine AJ, Su F, Tchou-Wong KM.Overexpression of WISP-1 down-regulated motility and invasion oflung cancer cells through inhibition of Rac activation. J Biol Chem.2003;278(13):11465–70. doi:10.1074/jbc.M210945200.

37. Canel M, Secades P, Rodrigo JP, Cabanillas R, Herrero A, Suarez C,et al. Overexpression of focal adhesion kinase in head and necksquamous cell carcinoma is independent of fak gene copy number.Clin Cancer Res. 2006;12(11 Pt 1):3272–9. doi:10.1158/1078-0432.CCR-05-1583.

38. Owens LV, Xu L, Craven RJ, Dent GA, Weiner TM, Kornberg L,et al. Overexpression of the focal adhesion kinase (p125FAK) ininvasive human tumors. Cancer Research. 1995;55(13):2752–5.

39. Schuuring E. The involvement of the chromosome 11q13 region inhuman malignancies: cyclin D1 and EMS1 are two new candidateoncogenes—a review. Gene. 1995;159(1):83–96.

40. Gibcus JH,Menkema L,MastikMF, HermsenMA, deBock GH, vanVelthuysen ML, et al. Amplicon mapping and expression profilingidentify the Fas-associated death domain gene as a new driver in the11q13.3 amplicon in laryngeal/pharyngeal cancer. Clin Cancer Res.2007;13(21):6257–66.

41. Lese CM, Rossie KM, Appel BN, Reddy JK, Johnson JT, Myers EN,et al. Visualization of INT2 and HST1 amplification in oral squamouscell carcinomas. Genes Chromosomes Cancer. 1995;12(4):288–95.

42. Parikh RA, White JS, Huang X, Schoppy DW, Baysal BE, BaskaranR, et al. Loss of distal 11q is associated with DNA repair deficiencyand reduced sensitivity to ionizing radiation in head and neck squa-mous cell carcinoma. Genes Chromosomes Cancer. 2007;46(8):761–75. doi:10.1002/gcc.20462.

43. Huang Q, Yu GP,McCormick SA,Mo J, Datta B,MahimkarM, et al.Genetic differences detected by comparative genomic hybridizationin head and neck squamous cell carcinomas from different tumorsites: construction of oncogenetic trees for tumor progression. Genes,Chromosomes and Cancer. 2002;34(2):224–33. doi:10.1002/gcc.10062.

44. Leemans CR, Braakhuis BJ, Brakenhoff RH. The molecular biologyof head and neck cancer. Nat Rev Cancer. 2011;11(1):9–22.

45. Nix PA, Greenman J, Cawkwell L, Stafford ND. Defining the criteriafor radioresistant laryngeal cancer. Clin Otolaryngol Allied Sci.2004;29(6):705–8. doi:10.1111/j.1365-2273.2004.00861.x.

46. Jeuken J, Cornelissen S, Boots-Sprenger S, Gijsen S, Wesseling P.Multiplex ligation-dependent probe amplification: a diagnostic toolfor simultaneous identification of different genetic markers in glialtumors. J Mol Diagn. 2006;8(4):433–43. doi:10.2353/jmoldx.2006.060012.

Tumor Biol.