• array-CGH;
  • cancer;
  • CGH;
  • fluorescence in situ hybridization;
  • genomics;
  • M-FISH;
  • molecular cytogenetics;
  • next generation sequencing;
  • spectral karyotyping


  1. Top of page
  2. Abstract
  3. Molecular cytogenetic methods
  4. Application of molecular cytogenetic methods in cancer
  5. Future perspective and conclusion
  6. Acknowledgements
  7. References

Aneuploidy or alteration in chromosome numbers is a characteristic feature in cancer that is generally a consequence of defective chromosome segregation during cell division. Molecular cytogenetic analyses have conferred substantial evidence with regards to the chromosomal architectures in cancer. Most importantly, the fluorescence in situ hybridization (FISH) technique that plays a leading role in diagnostic pathology for its single-cell analysis has provided crucial information regarding genomic variations in malignant cells. Further development of molecular cytogenetic methodologies such as chromosome specific FISH karyotyping and comparative genomic hybridization have also helped in the detection of cryptic genetic changes in cancer. But, the recent advancement of high throughput sequencing technologies have provided a more comprehensive genomic analyses resulting in novel chromosome rearrangements, somatic mutations as well as identification of fusion genes leading to new therapeutic targets. This review highlights the application of early molecular cytogenetics and the recent high throughput genomic approaches in characterizing various cancers and their invaluable support in cancer therapeutics.

A major tenet in cancer research is that the disease is caused and driven by genetic aberrations, and molecular cytogenetics plays an important role in identifying these aberrations. Molecular cytogenetics that generally refers to in situ hybridization has been a valuable addition to the conventional cytogenetics, chromosomal banding or karyotyping. It has paved the way to a more precise identification of clinically relevant genetic abnormalities which would have otherwise remain hidden if karyotype analysis was performed alone.

Fluorescence in situ hybridization (FISH) [1] is a widely established molecular cytogenetic technique that is regularly used in diagnostic pathology to locate the presence of specific DNA sequences in the chromosomes and nuclei of cancer cells. Many variations of FISH technique have since been introduced such as multiplex-FISH (M-FISH) [2], spectral karyotyping (SKY) [3], combined binary ratio labeling (COBRA) [4], color-changing karyotyping [5], cross-species color segmenting [6], high resolution multicolor-banding (mBAND) [7] and comparative genomic hybridization (CGH) [8], some of which are now used as research and clinical tools to detect cancer. However, these methods are experimentally demanding, labor intensive and resolution still limited by use of chromosomes as targets. Insight into cancer genomics is still restricted by the use of these approaches. The advent of high throughput genomic technologies such as DNA microarrays and next generation sequencing (NGS) has been a great leap forward, advancing our knowledge of somatic genomic alterations with high sensitivity and specificity and determining how each cancer occurs and behaves [9]. These technologies have helped to identify intra tumor heterogeneity between individual tumors that can contribute to treatment failure and drug resistance and further sub classify the cancer types. The data generated from these techniques can be used for accurate diagnosis and exploited to predict individual prognosis and treatment response. These techniques are faster, less invasive but more clinically useful tool for personalized detection of recurrence when compared to the early FISH based methods. This review highlights the applications of early and recent molecular cytogenetic approaches in characterizing various cancers and future prospects of these approaches.

Molecular cytogenetic methods

  1. Top of page
  2. Abstract
  3. Molecular cytogenetic methods
  4. Application of molecular cytogenetic methods in cancer
  5. Future perspective and conclusion
  6. Acknowledgements
  7. References

The early FISH based molecular cytogenetic methods detected structural variations at the chromosome level which are now detected at a whole-genome scale in a more rapid and precise manner by the advancement of high throughput genomic technologies.

Early FISH based methods

Fluorescence in situ hybridization

FISH [1] is an established molecular cytogenetic technique used in diagnostic pathology, especially the interphase FISH that can rapidly diagnose gene amplification, gene rearrangements, microdeletions and chromosomal duplication. FISH methodology includes hybridization of a DNA probe that has been labeled either directly with a fluorochrome (rhodamine or fluorscein −5-thiocyanate) or indirectly using haptens (biotin or digoxigenin).

FISH DNA probes are of three main types: (i) whole chromosome painting probes that identify the complete sequence of a given chromosome for studying the structural rearrangements, (ii) centromeric probes, designed to identify repetitive probes targeting the tandem repeat sequences of α-satellite DNA in the centromere of a given chromosome and (iii) locus specific probes designed to identify specific gene or chromosome region with copy number alterations and structural rearrangements.


M-FISH technique [2] as the name suggests, uses a series of five different dyes or fluorochromes to directly analyze the whole chromosome in one FISH experiment. Chromosomes are differentially labeled in a combinatorial approach [10] given by the equation N = 2n – 1, where N is the total number of colors achievable and n is the number of spectrally separate fluorochromes. The chromosomes are painted in the order of five chromosomes with a single fluorochrome, ten chromosomes with two fluorochromes, and nine chromosomes with three fluorochromes, respectively. Separate images are then captured for each of the five fluorochromes using narrow band pass microscope filters and a pseudo color is assigned to each unique color combination to visualize the entire chromosome pairs in 24 different colors (Fig. 1).


Figure 1. Identification of chromosomal alterations using FISH, M-FISH, SKY and COBRA-FISH and genomic imbalances using CGH, Array CGH and NGS whole-genome sequencing technology. CGH, comparative genomic hybridization; COBRA, combined binary ratio labeling; FISH, fluorescence in situ hybridization; M-FISH, multiplex-FISH; NGS, next generation sequencing; SKY, spectral karyotyping.

Download figure to PowerPoint

Spectral karyotyping

SKY [3] also uses five fluorochromes with the same approach of combinatorial labeling as in M-FISH. However, image capturing in SKY is done by creating a ‘classified’ spectral signature using a combination of epifluorescence microscopy, charge-coupled device imaging and Fourier spectroscopy unlike M-FISH. Both M-FISH and SKY techniques help to identify chromosome rearrangements, particularly translocations, insertions and marker chromosomes.

Combined binary ratio labeling

COBRA FISH technique is used to distinguish a number of targets simultaneously [4]. It uses a combination of both combinatorial labeling [10] and ratio labeling [11] to perform and quantify FISH discriminating at least 48 targets unlike M-FISH that can differentiate only 24 colors (Fig. 1).

High resolution multicolor-banding

The mBAND [7] technique helps to analyze peri and paracentric inversions in chromosomes. Overlapping micro dissected libraries are differentially labeled using five fluorochromes and a multicolour pseudo-G-banding of metaphase chromosomes with a high resolution is generated due to the changing fluorochrome intensity ratios along the chromosome.

Comparative genomic hybridization

CGH [8] is a genome-wide screening technique that uses genomic DNA to study the copy number variations in the tumor cells. Equimolar amounts of differentially labeled tumor and normal reference DNA are hybridized with normal human metaphase chromosomes and the difference in intensities of the fluorescence ratios in tumor compared with the normal DNA indicate the chromosome imbalance either deletion or amplification of tumor DNA (Fig. 1).

Other FISH based methods that are seldom used in cancer research are cross-species color segmenting [6] designed to analyze complex intrachromosomal rearrangements, color-changing karyotyping [5] similar to M-FISH and SKY method that uses combinatorial labeling with only three fluorochromes for chromosome identification and the more recently, FISH-on-a-chip that uses laminar flow to capture circulating tumor cells (CTCs) from blood samples followed by FISH analysis [12].

Recent high throughput genomic methods


In array-CGH, the metaphase chromosomes used in CGH method are replaced by high-density DNA microarray spotted with human bacterial artificial chromosome (BAC) clones [13], cDNA sequences or selected polymerase chain reaction (PCR) products obtained by amplification with either ligation-mediated PCR, degenerate primer PCR or rolling circle amplification to improve the quality of the array for a better genomic resolution. Nowadays, single nucleotide polymorphism (SNP) arrays are also used that help to genotype few hundreds to millions of SNPs to detect rare and common genomic rearrangements. These arrays require hybridization of only the test sample onto the array, unlike array CGH which relies on co-hybridization of test and reference DNA.

Next generation sequencing

The most advanced genomic technologies currently known is the second generation or NGS [14, 15]. Also known as massively parallel sequencing, these approaches employ arrays of several hundred thousand sequencing templates in parallel generating up to several hundred million short reads of DNA sequence per lane. The most commonly used platforms include 454 (Roche Applied Science, Branford, CT), Solexa (Genome Analyzer, Illumina Inc. Sandiego, CA), SOLiD (Applied Biosystems, Foster City, CA), Polonator (Dover/Harvard Systems, Salem, NH) and HeliScope (Helicos BioSciences, Cambridge, MA). NGS technologies have been used to comprehensively characterize populations of expressed genes (transcriptomes), all protein-coding genes (exomes) and even complete genomes of cancers (Fig. 1). The basic principle of NGS methods involves a process of DNA fragmentation, adapter ligation and immobilization of the fragments via the adapters to create libraries. The libraries then undergo a process of amplification, generating multiple copies of each DNA fragment which are then sequenced in parallel by fluorescence or chemiluminescence based method, yielding billions of short sequence reads. These reads are then aligned to the human reference genome and highly efficient algorithms are employed to perform the mapping of these complex genomes. Due to its cost effectiveness and enormous sequencing capacity, NGS is gradually changing the scenario of genetic diagnosis. It continues to make headway in cancer research by discovering disease-causing mutations, identifying novel drug targets and implementing therapeutic individualization way beyond the early cytogenetic methods.

Application of molecular cytogenetic methods in cancer

  1. Top of page
  2. Abstract
  3. Molecular cytogenetic methods
  4. Application of molecular cytogenetic methods in cancer
  5. Future perspective and conclusion
  6. Acknowledgements
  7. References

Cancers that have enough information regarding the molecular cytogenetic testing routinely used in the clinical management of the patients and also biomarkers identified using microarray or NGS with clinical implications have been discussed here. The rest of the cancers are out of scope of this study.

Hematologic malignancies

Hematologic malignancies are cancers affecting the immune system such as blood (leukemia), bone marrow and lymph nodes (lymphoma). Since the early 1960s when the first reciprocal translocation, t(9; 22) was identified in chronic myelogenous leukemia (CML) characterized with a fusion gene BCR-ABL by double fusion signal FISH [16], many recurrent cytogenetic abnormalities have been detected by the use of FISH based methods in almost every subtype of leukemia. For example, the translocation t(14;18)(q32;q21) has been detected by FISH that is the hallmark of follicular lymphomas and also B-cell lymphomas. Combined analysis of FISH based methods and karyotyping has been useful in the precise delineation of the karyotypes and accurate chromosome analysis of complex diagnostic cases. A complex case of acute myelogenous leukemia (AML) that showed a normal karyotype, revealed the recurrent translocation t(5;11)(q35;p15.5) with M-FISH [17] (Table 1) while another therapy-related AML case with a translocation t (8; 13) by conventional G-banding analysis, revealed an unmasked complex variant, including a four way translocation t (3; 5; 8; 21), with FISH, M-FISH and mBAND techniques showing the variant as a non-favorable prognostic factor [18] (Table 1). These findings have highlighted the importance of additional tools for accurate diagnosis and prognosis of myeloid malignancies.

Table 1. Molecular cytogenetic methods used in the detection of chromosomal abnormalities in various cancers
Cancer typeDescriptionChromosomal abnormalityGenesMolecular cytogenetic methodReferences included in the review
  1. AML, acute myelogenous leukemia; CGH, comparative genomic hybridization; CLL, chronic lymphocytic leukemia; CML, chronic myelogenous leukemia; CTC, capture circulating tumor cell; FISH, fluorescence in situ hybridization; GBM, glioblastoma multiforme; HNSCC, head and neck cancer, the squamous cell carcinoma; KRAS, v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog; M-FISH, multiplex-FISH; mBAND, high resolution multicolor-banding; NEPCa, neuroendocrine prostate cancer; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer.

Hematologic malignancies
 CLLRecurrent somatic mutationsSF3B1Whole exome sequencing[23]
 AMLt(5;11)(q35;p15.5) M-FISH[17]
 Therapy-related AMLt(3;5;8;21)AML1/ETOM-FISH & mBAND FISH[18]
 Cytogenetically normal AMLSomatic coding mutationsCDH24, EBI2, SLC15A1, KNDC1, PTPRT, PCKLC, GRINL1B, GPR123Whole-genome sequencing[20]
 AMLRecurrent somatic coding mutationsCDC42, NRAS, IDH1, IMPG2, ANKRD26, LTA4H, FREM2, C19orf62, SPRM1, PCDHA6, NPMI, CEP170, DNMT3AWhole-genome sequencing [21, 22]
  LymphomaBurkitt's lymphomat(8;14)(q24;q32)MYC-IGHFISH[82]
  Myelodysplastic syndrome Somatic mutationsSF3B1Massively parallel pyrosequencing[24]
GliomasGBMGene expressionAQP1, CHI3L1, EMP3, GPNMB, IGFBP2, LGALS3, OLIG2, PDPN, RTN1Microarray[35]
  Clear cell sarcoma t(2;22)(q34;q12)EWSR1-CREB1M-FISH[29]
  Desmoplastic round cell t(11:22)(p13;q12)EWSR1-WT1FISH[83]
  LiposarcomaWell differentiated & dedifferentiatedAmplification 12q15MDM2FISH[84]
  Head and neck cancerOral leukoplakiasTrisomy 1, 7, 10, 17 FISH[38]
  Monosomy 1, 7, 9, 10   
 Larynx, hypopharynx, oropharynx−4p, +8p, +12q, −18q CGH[39]
 Basal, mesenchymal, atypical, classical subtypesOxidative stress pathwayKEAP1/NFE2L2Copy number and expression microarray[40]
  Lineage markers OncogenesSOX2, TP63, EGFR, PIK3CA  
 HNSCC17p gene mutationsTP53p53 Genechip assay[41]
 HNSCCMutationNOTCH1, HRAS, CDKN2A, PIK3CA, TP53Exome sequencing[43]
Breast cancerSubtype associated17q amplificationHER2FISH[44]
  Somatic mutationsTP53, PIK3CA, GATA3, MAP3K1DNA copy number arrays, DNA methylation, exome sequencing, mRNA arrays, reverse phase protein arrays[85]
Lung cancerNSCLCt(2;2)(p21;p23)EML4-ALKFISH[51]
  −17p, −8pTP53, DR4, DR5SNP and CGH array[53]
  +12qCDK4, KRAS  
  MutationsKIF5B, RETWhole-genome and whole transcriptome sequencing[55]
 SCLC10q, 8p gene mutationsPTEN, FGFR1Exome sequencing[56]
  3q amplificationSOX2Exome sequencing[57]
Gastric cancerDiffuse & intestinal subtypet(1;9), −8p, polysomy of chromosome 20 M-FISH[58]
  Amplification 11p, 12p, 14q, 22q, 10q, 10p, 4p, 4q, 16q, 19pEGFR, SEC61G, CDC6, ANP32E, BYSL, FDFT1, SMAD4CGH, array CGH[61]
  3p, −18q Expression microarray 
 Primary gastric tumors, cell linesAmplification of 12p12, 10q26KRAS and FGFR2Array CGH, FISH[60]
   CTNNB1, EXOSC3, TOP2A, LBA1, LZTR1, CCL5Expression microarray[63]
Colorectal cancerKRAS wild-type colorectal tumors7p12EGFRFISH[65]
 Patients clinical trialMutationsKRAS, NRAS, EGFR, BRAF, PTEN, PIK3CA, AKT, TP53, CTNNB1Massively parallel sequencing[66]
Ovarian cancerClear cell carcinomaMutation 1pARID1AWhole transcriptome sequencing[68]
 Serous endometrialMutation 20q, 22q, 1p, Xp, 4q, 17q, 6q, 12pCDH4, EP300, ARID1A, TSPYL2, FBXW7, SPOP, MAP3K4, ABCC9Whole exome sequencing[71]
Prostate cancerParaffin-embedded tumor tissues and CTCs+8, +11, −7MYCFISH [73, 74]
 Primary tumors ERG-TMPRSS2 and ETV1-ETSFISH [75, 76]
 Prostate small cell carcinomaGene fusionsERG-TMPRSS2FISH[77]
 NEPCaAmplification 20q, 2pAURKA, MYCNRNA sequencing and oligoneucleotide arrays[79]

Other abnormalities detected by FISH based methods are also promising and have been shown to have clinical significance such as the del (11q) and del (17q) shown as a marker for poor prognosis, del (13q) or a normal karyotype for low risk disease and the presence of trisomy 12 as a marker of intermediate risk in chronic lymphocytic leukemia (CLL) [19].

With the introduction of high throughput genomic technologies such as NGS, FISH based chromosome level detection have gradually changed focus to genome-wide detection of single nucleotide and indel variants that are common in cancer. Somatic coding mutations were identified for the first time in a cytogenetically normal AML [20] (Table 1) followed by recurrent somatic coding mutations in 12 genes [21] and DNA methyl transferase gene (DNMT3A) (2p) that were associated with a poor prognosis and an intermediate risk cytogenetic profile of AML [22].

In CLL patients too recurrent somatic mutations in SF3B1 (2q) gene in almost 10% of the 384 patients sequenced was found to be associated with faster disease progression and poor overall survival [23] while this gene mutation was shown to be less deleterious in patients with myelodysplastic syndrome [24]. The heterogeneity of tyrosine kinase inhibitors responses across CML patients, with the inhibitor resistant patients showing the presence of a BIM deletion polymorphism [25] too was identified using next generation paired end sequencing technology. These findings suggested the use of genome-wide analysis as prognostic and predictive tool and validation of these results are in progress.


Sarcomas that constitute almost 1% of all solid tumors have considerable morphological overlap between different histogenetic types due to which their diagnosis is sometimes challenging. Chromosomal karyotyping is generally the classic approach for identifying genetic markers and translocations in soft tissue sarcomas, especially Ewing sarcoma and synovial sarcomas whereas FISH is used in screening the common translocation partners t(X;18)(p11.2;q11.2), SS18-SSX, for synovial sarcoma a unique translocation t(11:22)(p13;q12) involving EWSR1 and WT1 genes for desmoplastic round cell tumor and amplification of MDM2 (12q15) for well differentiated and dedifferentiated liposarcomas. In complex cases where karyotype banding analysis alone cannot accurately diagnose the disease, it is combined with either COBRA-FISH or M-FISH to identify derivative chromosomes, such as the t(X; 18) in the synovial sarcoma of the heart and larynx [26-28]. Similarly, SKY is useful in identifying the presence of t(2;22)(q34;q12) bearing the EWSR1-CREB1 fusion transcript, confirming the diagnosis of clear cell sarcoma [29]. However, targeted therapies for the treatment of sarcomas are still lacking. Recently, large scale genomic landscape analyses of sarcoma samples identified new therapeutic targets such as TP53 (17p), NF1 (17q) and PIK3CA (3q) gene mutations in seven major subtypes of sarcomas [30] and also amplification of the mouse double minute 2 homolog (MDM2, 12q), MDM2SNP309 promoter polymorphism and TP53 deletion and mutation suggesting a mechanism of TP53 pathway alteration [31] and the use of MDM2 antagonist, Nutlin for the treatment of MDM2 amplified tumors.


Gliomas are the most frequent tumors of the central nervous system. Histologically, they are classified according to cell type, non-neoplastic glial cells (ependymomas, astrocytomas, oligodendrogliomas and mixed gliomas such as oligoastrocytomas), by grade (low and high) and by location above or below the brain membrane (supratentorial and infratentorial). Initiation and progression of gliomas occur due to accumulation of multiple genetic alterations. IDH1/2 gene mutations are regarded as early event in gliomagenesis due to their presence in all the low-grade gliomas [32]. Other genetic abnormalities that accumulate during tumor progression are characteristics of different gliomas. An unbalanced translocation between chromosomes 1 and 19 with combined loss of the der (1;19)(p10;q10) detected by FISH is characteristic of oligodendrogliomas [33] and is associated with better response to adjuvant therapy. Polysomy of chromosomes 1 and 19 correlates with shorter survival in 1p/19q-codeleted tumors which is characteristic of anaplastic oligodendrogliomas whereas TP53 mutation or 17p13 loss is characteristic of astrocytic tumors [34]. The high grade astrocytoma known as glioblastoma multiforme (GBM) is the most common of all gliomas. Primary GBMs that arise de novo in older patients and have short duration of symptoms, demonstrate frequent EGFR amplification, 10q loss, and mutations in PTEN gene, but rarely have IDH mutations. On the other hand, secondary GBMs which occur in younger patients and have longer duration of symptoms, lack EGFR amplification but show mutations in TP53 and IDH genes. Therefore, evaluation of the molecular status of each of these markers is a major relevant diagnostic and prognostic tool for brain tumors and is used by neuropathologists in routine clinical practice.

Other potential markers with clinical applications have been obtained by microarray gene expression profiling of gliomas. Treatment sensitive and treatment refractory GBM tumors associated with patient outcome, revealed a consensus 38 gene survival set. A nine gene subset (Table 1) was then validated in FFPE samples and its profile was shown to be positively associated with markers of glioma stem-like cells, including CD133 and nestin showing a multigene predictor of outcome in glioblastoma that had a potential of clinical application [35]. Another study showed differential expression of a total of 4535 genes in GBM vs normal tissue and TGF-β signaling pathways mediated by SOX4 and TGFB1 as an effective therapeutic approach in GBM [36]. Recently, seven low-frequency variant SNPs at 8q24.21 have been shown to be strongly associated with risk of oligodendroglial tumors and astrocytomas with IDH1 and IDH2 mutations using NGS and SNP genotyping [37].


Carcinomas too have highly complex karyotypes like the sarcomas and enumeration of these aberrations by molecular cytogenetics has rapidly increased.

Head and neck cancer

Cancer of the head and neck region includes those in the nasal cavity, sinuses, lips, mouth, salivary glands, throat or larynx. Although there are no standard molecular tests that are routinely used in clinics, molecular markers that can be developed for risk assessment and early detection are being discovered. Mouth or oral cancer is more predominant in people with tobacco and alcohol consumption due to the malignant transformation of a patch of mucous membrane called oral leukoplakia. Frequent genetic alterations have been reported in the dysplastic oral leukoplakia compared to hyperplastic lesions or clinically healthy tissues or cell smears [38]. Others such as laryngeal tumors have also shown significant increase in −4p, +8q, +12q and −18q compared to hypo pharynegeal and pharyngeal tumors by CGH analysis indicating sub-site genetic differences in these tumors [39]. Nowadays high throughput technologies are being used to discover novel therapeutic targets so that drugs can be designed to treat these patients.

Using mRNA expression and DNA copy number microarray analysis, the most common histological type of head and neck cancer, the squamous cell carcinoma (HNSCC) that constitutes 90% of cases has been shown to have four subtypes—the basal, mesenchymal, atypical and classical with distinct patterns of chromosomal gain and loss of canonical cancer genes [40] (Table 1). Mutations in TP53 gene were identified in these patients that were significantly associated with decreased survival [41] and poor tumor response to chemotherapy (cisplatin and flurouracil) and radiotherapy [42] suggesting TP53 as a prognostic and predictive marker of HNSCC. Additionally, NGS technology showed frequent mutations in NOTCH1 (9p), a gene not previously implicated in HNSCC cancer and also mutations in HRAS (11p), CDKN2A (9p), PIK3CA and TP53 genes and their presence increased the risk of cancer progression from mild dysplasia to invasive carcinoma [43] suggesting these as important therapeutic targets.

Breast cancer

In breast cancer, the most remarkable gene amplification translated to clinical practice has been the amplification of the HER2 gene (17q) identified by FISH that is associated with poor prognosis in early-stage and metastatic breast cancer. Treatment with trastuzumab in combination with chemotherapy shows significant improvement in the overall survival of patients [44]. However, some patients are unresponsive or resistant to trastuzumab which has been found to be due to the presence of PIK3CA mutations and loss of PTEN (10q). These genes correlate with similar prognostic factors and are not mutually exclusive in breast tumors [45]. Hence, new markers are being searched for targeted therapy in these patients.

Several microarray studies have revealed the heterogenous nature of breast tumors grouping this cancer into intrinsic subtypes, namely (i) HER2 enriched or clinically HER2 positive [46], (ii) Luminal, estrogen receptor (ER) positive, consisting of Luminal A and Luminal B tumors [47] and (iii) ER negative, otherwise known as triple-negative as these tumors are also progesterone receptor (PR) negative and HER2 negative [48]. This classification has helped to diagnose and treat the patients accordingly.

With the recent explosion of NGS data, presence of recurrent gene rearrangements involving Notch family and the microtubule-associated serine-threonine (MAST) kinase family genes have been shown to play a role in breast cancer [49] and suggested as therapeutic targets. Driver mutations in at least forty cancer genes and seventy three different combinations of mutated cancer genes with several new genes (Table 1) have been demonstrated with strong correlations between number of mutations, age of the patient during diagnosis and cancer grade [50]. However, these data are yet to predict benefit from specific therapeutic agents and it is still not possible to predict prognosis or chemotherapy treatment response in specific disease subsets accurately.

Lung cancer

Histologically, lung cancer is of two main types, non-small cell (NSCLC) and small cell (SCLC) that account for 80% and 20% of lung cancers, respectively. A subset of NSCLC patients, preferentially non-smokers and Asians are known to have mutations in the EGFR gene and are treated with the anti-EGFR drug called gefitinib. However, most cases of this disease constitute smokers, hence, identification of novel transforming genes are warranted.

Recently, a gene fusion between echinoderm microtubule-associated protein-like 4 (EML4) and anaplastic lymphoma kinase (ALK) in a NSCLC patient with a smoking history made a breakthrough [51]. This fusion gene was shown to be mutually exclusive to EGFR, KRAS (12p) and BRAF mutations. This discovery led to a phase I clinical trial with ALK inhibitor (crizotinib) showing potential clinical benefit [52] which could be ultimately used in the clinic. Other studies combining genome sequencing, SNP array and array CGH analysis of tumors from NSCLC patients with smoking history, have shown a large set of novel somatic mutations, copy number loss of TP53, DR4 and DR5, gains of CDK4 and KRAS and copy neutral loss of heterozygosity of chromosome 13 including RB1 [53] and also mutations in BRAF, AKT1 (14q), ERBB2 (17q), PIK3CA and fusions involving KIF5B (10p) and RET (10q) [54, 55] suggesting these as possible stratified biomarkers and therapeutic targets.

In SCLC patients, no such cancer risk biomarkers suitable for clinical application has been validated although FISH analysis has shown alterations in chromosome 3p and amplification of MYC (8q) and EGFR gene in sputum samples of the patients for lung cancer risk or diagnosis. Exome sequencing analyses of SCLC tumors have identified new gene mutations that could lead to molecularly targeted drugs for this disease such as the PTEN gene mutations and FGFR1 gene amplifications [56] as well as SOX2 gene (3q) amplification that is present in more than a quarter of samples correlating with high expression of SOX2 and also suggesting its potential role as a driver of tumorigenesis in SCLC [57]. However, these targets are yet to be validated and assessed for their therapeutic value.

Gastric cancer

Gastric cancer (GC) is another most common cancer worldwide. It has two histological subtypes, diffuse and intestinal with distinct morphological and epidemiological features as well as fewer genetic alterations in intestinal than diffuse subtype identified by M-FISH [58]. Since molecular targeted therapy represents an emerging treatment strategy for GC patients, such therapeutic targets are being searched in molecular profiles of GC for better treatment strategy.

The HER2 gene has been validated for its amplification and overexpression in GC patients and is now being used clinically to treat these patients with trastuzumab in combination with chemotherapy [59]. GC is also known to harbor FGFR2 and KRAS gene amplifications identified by array CGH and FISH that are mutually exclusive and associated with prognosis and overall survival (Table 1) [60]. Mutually exclusive genomic alterations or single-gene amplification in a pool of other molecular targets in individual tumors has been suggested to present inter-relationships between different targets to facilitate the allocation of patients to different clinical trials.

Expression profiling by microarray in GC has shown three molecular subtypes, tumorigenic, reactive and gastric-like and the gastric-like has been shown to be associated with better overall survival compared to the other two groups, suggesting that prognostication by molecular subtypes might be clinically useful in GC, providing additional information above and beyond classical tumor staging [61]. Gene expression signatures have also helped to categorize GC as genomic intestinal (G-INT) and genomic diffuse (G-DIF) and significant association between G-INT subtype and patient prognosis have been shown by univariate and multivariate analysis (p < 0.001) [62]. Also a prognostic risk-score based on the expression levels of six genes has been described to predict relapse in GC patients after gastrectomy [63] (Table 1). These findings show the importance of microarray technology as a prognostic and diagnostic tool in GC.

Colorectal cancer

Colorectal cancer (CRC) that starts in the colon or rectum is also being studied extensively for therapeutic targets using molecular cytogenetic techniques. The EGFR gene has been shown to play an important role in CRC in several studies. Increased EGFR gene copy number in metastatic CRC patients [64], KRAS mutated wild-type EGFR FISH pattern in non-responsive cetuximab CRC patients as well as KRAS wild-type, EGFR FISH-positive (chromosome 7 disomy) in patients showing higher response to cetuximab with a longer progression-free survival [65] have demonstrated EGFR and KRAS as prognostic and predictive biomarkers. Recently, the EGFR pathway mutation analysis in a phase three randomized clinical trial for patients receiving an anti-EGFR monoclonal antibody called panitumumab, showed association with longer progression-free survival among wild-type KRAS patients (codons 12/13/61) suggesting potential clinical use of NGS for evaluating predictive biomarkers [66].

Ovarian cancer

More than 80% of ovarian tumors arise from surface epithelium that are divided into three histological subtypes: benign cystoadenoma, borderline tumors and adenocarcinoma with distinct biological behavior. The current diagnostic test for early-stage cancer is designed to measure the levels of CA125 and HE4 proteins in the blood that aid in the risk stratification of those present with pelvic mass. However, symptoms are complex and often misdiagnosed. Borderline tumors are the most difficult to diagnose among the three subtypes as it is unclear whether they are independent or part of a continuum of tumor progression that culminates in ovarian carcinoma. Using G-banding and CGH, similar imbalances like the overt ovarian carcinomas have been shown in a subset of borderline tumors [67] supporting the theory of tumor progression from borderline tumors.

In the post-genomic era, emphasis is on the identification of biomarkers to accurately predict progression of disease as well as to precisely determine drug response. By using genome-wide analysis, gene mutations in various subtypes of ovarian cancer have been identified by several groups that are associated with worse prognosis [68-72] (Table 1). However, these findings that define genomic landscapes and their association with clinical outcome warrants further investigation to allow for patient stratification and ultimately improve disease management.

Prostate cancer

Prostate cancer (PCa) is currently diagnosed by testing the elevated serum level of prostate specific antigen (PSA) but due to lack of specificity, new biomarkers of malignity are constantly being searched. Copy number alterations of MYC oncogene in prostate tumors [73] and in CTCs of progressive castration resistant metastatic PCa patients [74] have been identified as prognostic markers using FISH. The breakthrough discovery of recurrent fusion gene, TMPRSS2-ERG in PCa using microarray technology has been shown to be associated with an androgen independent, highly aggressive and lethal subtype of PCa [75, 76] and its presence in prostate small cell carcinoma has helped to distinguish this cancer from bladder carcinoma [77] demonstrating this gene as a diagnostic and prognostic marker.

Genome-wide analysis has shown a spectrum of DNA alterations in advanced PCa along with other non-ETS gene fusion events and new cancer specific gene fusions [78] (Table 1) that can act as potential therapeutic targets. The neuroendocrine PCa (NEPCa) that express neuroendocrine markers but do not express androgen receptor or PSA has been found to have significant amplification of AURKA (20q) and MYCN (2p) genes by RNA seq and oligonucleotide arrays which were completely suppressed following treatment with aurora kinase inhibitor [79] suggesting AURK as a predictive marker of NEPCa. However, these findings are still in early stage and needs to be validated in large scale clinical settings to be implemented in clinical practice.

Future perspective and conclusion

  1. Top of page
  2. Abstract
  3. Molecular cytogenetic methods
  4. Application of molecular cytogenetic methods in cancer
  5. Future perspective and conclusion
  6. Acknowledgements
  7. References

It is evident that identification of chromosomal aberrations by molecular cytogenetic techniques is important in searching novel chromosomal rearrangements and genes involved in tumourigenesis. Understanding the basis of these techniques and their application is critical in the accurate diagnosis of cancer. Traditional cytogenetic analysis and FISH based methods have no doubt played a pivotal role in the diagnosis of many tumors. FISH, has been useful in the diagnosis of gene amplification (HER2 in breast cancer), gene rearrangements (BCR-ABL in leukemias, ALK-EML4 in NSCLCs), microdeletions and chromosomal duplication. There are also an increasing number of FISH diagnostic panels and applications in other cancers such as the UroVysion test (Abbott, IL, USA) for bladder carcinoma screening that detects aneuploidy for chromosomes 3, 7, 17, and loss of 9p21 in urine. FISH on FFPE tissues are more relevant clinically since the overall tissue architecture is intact. However, the FISH based methods are labor intensive and time consuming.

Moreover, the focus that was targeted only on diagnosis previously has shifted in the past decade to prognosis with the use of microarrays and NGS technologies generating multiple markers on a genome-wide scale that can predict specific clinical end points of interest. NGS has provided new therapeutic targets increasing the power of sub classifying cancers and also allowing earlier diagnosis and more appropriate treatment for patients paving the way towards personalized medicine [60-62, 80]. In fact, based on the genome sequencing analysis of tumors, clinicians can now design patient specific DNA probe and use it in patient's blood serum to monitor the progress of the patient's treatment and relapse if any [81]. However, intra tumor heterogeneity and translation of the large oncogenomics data into an easily interpretable data accessible for cancer care is proving to be a big challenge because of which many genomic disease causal variants are still in an early phase. Nevertheless, to accelerate the rate of translating genomics data into clinical practice, collaborations between clinicians, researchers and bioinformaticians is necessary to interpret and confirm the functional relevance of variants to cancer.


  1. Top of page
  2. Abstract
  3. Molecular cytogenetic methods
  4. Application of molecular cytogenetic methods in cancer
  5. Future perspective and conclusion
  6. Acknowledgements
  7. References