3. Explain technical strategies for detection of chromosomal translocation in nondividing cells.

Transcription

1 The Oncologist Molecular Cytogenetics in Solid Tumors: Laboratorial Tool for Diagnosis, Prognosis, and Therapy MARILEILA VARELLA-GARCIA Department of Medicine, Medical Oncology Division, University of Colorado Cancer Center, University of Colorado Health Sciences Center, Denver, Colorado, USA Key Words. Cytogenetics Fluorescence in situ hybridization (FISH) Interphase FISH Biomarker Target therapy LEARNING OBJECTIVES After completing this course, the reader will be able to: 1. Explain the basic principles supporting the FISH technology and list examples of methodology variants suitable for analysis in metaphase and interphase cells. 2. Describe at least one advantageous and one limiting factor for the expansion of the applicability of FISH assays to solid tumors. 3. Explain technical strategies for detection of chromosomal translocation in nondividing cells. 4. Illustrate applications of cytogenetic markers to solid malignancies for diagnosis, prognosis, selection of therapy, and monitoring disease recurrence or response to treatment. CME ABSTRACT Access and take the CME test online and receive one hour of AMA PRA category 1 credit at CME.TheOncologist.com The remarkable progress in the understanding of leukemogenesis was soundly sustained by methodological developments in the cytogenetic field. Nonrandom chromosomal abnormalities frequently associated with specific types of hematological disease play a major role in their diagnosis and have been demonstrated as independent prognostic indicators. Molecular pathways altered by chimeric or deregulated proteins as a consequence of chromosomal abnormalities have also significantly contributed to the development of targeted therapies, and cytogenetic assays are valuable for selecting patients for treatment and monitoring outcome. In solid tumors, significantly high levels of chromosome abnormalities have been detected, but distinction between critical and irrelevant events has been a major challenge. Consequently, the application of cytogenetic technology as diagnostic, prognostic, or therapeutic tools for these malignancies remains largely underappreciated. The emergence of molecular-based techniques such as fluorescence in situ hybridization was particularly useful for solid malignancies, and the spectrum of their application is rapidly expanding to improve efficiency and sensitivity in cancer prevention, diagnosis, prognosis, and therapy selection, alone or in combination with other diagnostic methods. This overview illustrates current uses and outlines potential applications for molecular cytogenetics in clinical oncology. The Oncologist 2003;8:45-58 Correspondence: Marileila Varella-Garcia, Ph.D., University of Colorado Health Sciences Center, Campus Box B188, 4200 East 9th Avenue, Denver, Colorado 80262, USA. Telephone: ; Fax: ; marileila.garcia@uchsc.edu Received September 9, 2002; accepted for publication November 4, AlphaMed Press /2003/$5.00/0 The Oncologist 2003;8:

2 46 Molecular Cytogenetics in Clinical Oncology INTRODUCTION Early descriptions of mitotic and spindle aberrations in human tumors date from approximately a century ago; however, for many decades, the progress in cancer cytogenetics was hampered by technical challenges in assessing the nuclear chromosomal content. In the late 1950s and the 1960s, significant methodological achievements regarding culture and mitotic arresting of cells made the detection of chromosomal abnormalities possible, although without an accurate identification of the occurring changes. The typical example is the description in patients with chronic myeloid leukemia, of the Philadelphia (Ph) chromosome, which initially was depicted as a deleted chromosome 21 or 22 [1]. The flourish of chromosomal banding techniques (Q-, G-, R-banding) in the early 1970s has contributed to the establishment of karyotype-phenotype correlations for a rapidly increasing number of diseases. Characteristic recurrent translocations were rapidly identified in hematopoietic and soft tissue neoplasias and have helped in the identification of chromosomal loci containing genes involved in the genesis of these tumors [2]. Conversely, the impact of cytogenetic technology on understanding molecular mechanisms involved in the initiation and progression of solid tumors has been less valuable. Solid tumors are commonly associated with an array of orchestrated genetic changes, and the identification of changes causally related to the carcinogenic process has been frustratingly slow, mainly as a consequence of the enormous volume of secondary abnormalities reflecting the phenomenon of genomic instability. FLUORESCENCE IN SITU HYBRIDIZATION (FISH) AND INTERPHASE ANALYSIS Classical cytogenetics represented by chromosomal banding techniques was successful in correlating karyotype abnormalities with diagnosis, prognosis, and response to therapy in hematological neoplasias. However, these techniques require a high rate of cell division and good chromosomal morphology, which represent challenges for the cytogeneticists, and a long period for assaying and analyzing, which usually is a challenge for physicians (Table 1). Thus, it was not surprising that an outburst of progress in the cancer cytogenetics field followed the advent of molecular cytogenetics, with the development and optimization of FISH technology [3]. FISH technology initially focused on research issues, but soon was applied to clinical questions and has proved sufficiently sensitive and reliable to fill in the gap between classic karyotyping and highly sensitive molecular techniques. Table 1. Cytogenetic techniques most commonly used, and advantages and limitations of their applications to human solid malignancies Techniques Characteristics Applications Advantages Limitations G-banding, R-banding Protein digestion and/or special dye Identification of numerical Genome-wise screening for Low resolution, dependent on generate banding pattern specific for and structural chromosomal chromosome-level abnormalities; chromosome condensation; each chromosome anomalies Low cost for reagents and Requires mitotic cells and wellinstrumentation; spread chromosomes; Simple and robust procedures Labor-intensive analysis; Low efficacy in highly rearranged karyotypes FISH A small, labeled DNA fragment is used Identification of the presence, Applicable to interphase cells; Conclusions limited to the tested as a probe to search for homologous number of copies per cell, Fast analysis and scoring; targets; target sequences in chromosome or and localization of probe Simple and robust procedures Scarceness of specific and reliable chromatin DNA DNA reagents CGH Competitive hybridization of differentially Detection of variant DNA Applicable to fresh or preserved Dedicated instrument required for labeled total genomic tumor DNA and copy numbers at the specimens; analysis; normal reference DNA to normal human chromosome level No need for cell culture; Low resolution, dependent on metaphases used as templates Genome-wise screening for chromosome condensation of genomic imbalances normal template; Balanced rearrangements and imbalances present in low frequencies remain undetected Multicolor karyotyping Hybridization with 24 differentially Detection of rearrangements Accurate origin identification of Requires mitotic cells and well- (M-FISH, SKY) labeled, chromosome-specific probes involving one or more all segments in complex spread chromosomes; allows the painting of every human chromosomes within rearrangements; Limited accuracy in determination chromosome in a distinct color individual metaphase Clarification of marker of breakpoints; spreads chromosomes Intrachromosomal changes remain undetected; Dedicated instrument required for analysis

3 Varella-Garcia 47 Figure 1. FISH technology uses a labeled DNA segment as a probe to search homologous sequences in interphase chromatin and metaphase chromosomes. On the right, nuclei of ejaculate suspension from a patient with prostate carcinoma were hybridized with centromeric probes for chromosome 7 (labeled with SpectrumRed fluorophore) and chromosome 8 (labeled with SpectrumGreen ). Spermatozoid nuclei (small) show the expected single copy of each target because of their haploid chromosomal complement. A prostate-derived nucleus (large) shows a gain in copy number for both targets, characterizing the cell as abnormal. On the left, a metaphase spread (partially illustrated) was hybridized with locus-specific FISH probes for the HOXA (labeled with Spectrum- Green ) and HOX11L2 (labeled with SpectrumRed ) genes, which were respectively mapped at 7p15 and 5q34. The basis for the molecular cytogenetics technology is the complementary double-stranded nature of DNA. A specific DNA segment is converted into a probe through the attachment of a fluorescent tag or a reporter molecule that later in the procedure will be conjugated with a fluorescent tag. The probe is denatured, exposed to a similarly denatured target DNA, and, under proper hybridization conditions, recognizes and binds to the homologous sequences in the target DNA. After hybridization, the copy number and location of the fluorescent tags, and consequently, of the target homologous regions, are recognized under fluorescence microscopy both in chromosome spreads and in interphase nuclei (Fig. 1). FISH technology is simple and robust, and its phenomenal contribution to the cancer field largely relies on its applicability to interphase cells (Table 1). While early FISH applications aimed at the detection of numerical abnormalities in mitotic or interphase cells, as well as structural abnormalities in metaphase spreads, a milestone was achieved in 1990 when Rowley et al. [4] and Tkachuck et al. [5] developed strategies for identification of chromosomal translocations in interphase cells. Since then, variations in probe design, and consequently, in the presentation of fluorescent signals in normal cells and in cells carrying chromosomal translocations have been developed for molecularly cloned rearrangements. In the standard approach for interphase evaluation of chromosomal translocations [5], a DNA probe, comprising sequences mapped proximally to the breakpoint in one of the chromosomes involved in the reciprocal translocation, is combined with a differentially labeled DNA probe that includes sequences mapped distally to the breakpoint in the other chromosome. Positive nuclei for the translocation display one dual-color fusion signal, representing one of the derivative chromosomes generated by the translocation, and two single-color signals, one for each of the normal alleles (Fig. 2A). This standard FISH strategy has largely been used for the diagnosis of leukemia and lymphoma translocations at disease presentation or at clinical relapse, since patients at these stages usually show high frequencies of bone marrow cells with the chimeric gene. Nonetheless, for detection of residual disease, this approach lacks specificity, because cells with random spatial colocalization of normal signals with different colors, usually found in frequencies ranging from 1%-5% of scored nuclei, are seen as false positives. To minimize this problem, a dual-fusion FISH (D-FISH) approach was developed, with the probe set including DNA sequences that encompass proximally and distally the translocation breakpoints on both chromosomes involved in the translocation. As illustrated for the t(8;21)(q22;q22) in Figure 2B, sequences for each chromosome are labeled with a specific color, and the translocation generates fused signals in both derivative chromosomes [6]. Using the D-FISH approach, nuclei carrying the reciprocal translocation show two copies of the fused signals and one copy of each of the single signals representing the normal alleles. It is doubtful that such a positive pattern would be mimicked by artificial conditions; therefore, the D-FISH approach shows a higher sensitivity that favors its application in patients in clinical remission who are expected to present low frequencies of abnormal cells [7].

4 48 Molecular Cytogenetics in Clinical Oncology Figure 2. Detection of chromosomal translocations in interphase cells. A) Conventional FISH probe format was proposed by Tkachuck et al. [5] and is illustrated for the t(9;22)(q34;q11), associated with chronic myeloid leukemia. B) Dual-fusion format is illustrated for the t(8;21)(q22;q22), characteristic of the AML-FAB M2 [6]. C) Break-apart format for cancer genes with promiscuous partners is illustrated for the gene E2A [12]. These two described FISH approaches are adequate for analysis of recurrent reciprocal translocations, which comprise the majority of hematopoietic and soft tissue-specific rearrangements, but are very rare in solid tumors. Presently, the sole example of single reciprocal translocation associated with aggressive carcinomas is the t(15;19), with breakpoints in 19p13.1 and variant breakpoints in 15q11-q15 [8, 9]; the molecular events triggered by these rearrangements have yet to be characterized. Conversely, solid tumors usually present large numbers of derivative chromosomes from translocations involving two or more chromosomes [10], suggesting that some critical cancer genes are promiscuous in the selection of translocation partners. Promiscuity in translocation partnership is well known for leukemia genes such as mixed-lineage leukemia and E2A [11], and a third FISH strategy was developed to address this particular condition, namely, break-apart FISH. The break-apart probe comprises DNA sequences mapped proximally and distally to the breakpoint within a critical gene labeled with distinct fluorochromes [12]. In this case, the fused fluorescent signals represent a normal gene, whereas nuclei with disruptions within the target gene due to translocations show one fusion signal for the normal allele and two

5 Varella-Garcia 49 single-color signals, one for each of the derivative chromosomes, regardless of which chromosome is the partner in the translocation (Fig. 2C). FISH VARIANTS: COMPARATIVE GENOMIC HYBRIDIZATION AND MULTICOLOR KARYOTYPING Two major variants of FISH technology, comparative genomic hybridization (CGH) and multicolor karyotyping, have been extensively applied to solid tumors; their advantages and shortcomings are summarized in Table 1. CGH was developed by Kallioniemi et al. [13] and allows the visualization of gains and losses in the copy numbers of DNA sequences from specific chromosomal regions without involving cell culture, as illustrated in Figure 3A. The genomic DNA from the tumor specimen is isolated, fragmented, and labeled with a given fluorophore, for instance, the green fluorophore fluorescein isothiocyanate (FITC). A sample of normal genomic DNA, taken as a reference, is similarly isolated and fragmented but is differentially labeled, for instance, with the red fluorophore Texas Red. Subsequently, equal amounts of tumor and reference DNA are coprecipitated with an excess of repeat-rich DNA (Cot-1 DNA) for the blocking of the repetitive sequences, and this DNA mixture is allowed to hybridize to normal metaphase spreads. Sequences of DNA from each specific chromosomal region recognize their homologous regions in the normal template metaphases and hybridize to them according to the proportion in which they are represented in the DNA mixture. Comparative analysis of the proportion of each distinctly labeled tumor and reference DNA that was incorporated into the template metaphase provides a ratio of tumor to normal signal against a baseline of the metaphase chromosomes. Therefore, the normal number of chromosome regions with copies in both the tumor and reference DNA of a chromosomal region will display a balanced profile of Texas Red and FITC fluorophores, whereas regions deleted in the tumor will show a higher profile of Texas Red, and regions with gains in copy number in the tumors will show a higher profile of FITC. Figure 3B illustrates the typical pattern for presentation of results generated by CGH analysis. In the prostate carcinoma illustrated in Figure 3B, for instance, gene amplifications were found for the oncogene MYC (at 8q24) and the vitamin D receptor gene (at 12q12-q14), which mediates growth inhibitory effects of vitamin D in prostate cancer. The CGH technique has been extensively used for solid tumor analysis since its inception, and a nonrandom pattern of genomic imbalances has emerged from these studies [14]. To this point, CGH has been the most helpful cytogenetics tool for identification of oncogenes and tumor suppressor genes involved in the initiation and progression of solid tumors, guiding the more detailed genetic analysis to specific chromosomal regions. Two major multicolor karyotyping techniques are spectral karyotyping (SKY) and multiplex-fish (M-FISH). These techniques are similarly based on combinatorial probe-labeling schemes with five spectrally discrete fluorochromes that uniquely paint all 24 human chromosomes (Fig. 4) but differ in their analytical procedures. SKY relies on a single digital image acquired with a charge-coupled device (CCD) in conjunction with a customized multiband optical filter, and on Fourier spectroscopy [15]. An interferometer is used to access the spectrum of fluorescence wavelengths for each pixel of the image. A dedicated computer program identifies the components of the spectrum, applying a classification algorithm, and generates a composite image in which each chromosome is pseudocolored based on its fluorochrome signature. Since each chromosome is represented by one specific color, rearrangements involving different chromosomes are easily recognized (Fig. 5). On the other hand, the M-FISH technique combines monochrome CCD images acquired separately for each of the five spectrally distinguishable fluorochromes (Fig. 6) and has them subsequently analyzed by a dedicated software program that also builds a composite image by assigning a particular pseudocolor to each chromosome [16]. Thus, both multicolor karyotyping techniques are specially tailored to identify and characterize the complex chromosomal rearrangements found in solid tumor cells. Nevertheless, these methodological approaches are unable to detect intrachromosome rearrangements, such as inversions and deletions, since these specific abnormalities maintain the correct color for abnormal chromosomes. Consequently, multicolor karyotyping cannot replace standard banding karyotyping; rather, both techniques should be used in conjunction for a comprehensive evaluation of solid tumors. Application of multicolor karyotyping technology to solid tumors faces most of the constraints of classic cytogenetics regarding the need for dividing cells and long chromosomes with good morphology in well-spread metaphases. These features are very difficult to achieve simultaneously in primary tumors, and both the SKY and M-FISH techniques have been more commonly applied to cell lines established from primary tumors. The immortal lines offer more suitable material for karyotype analyses, and identification of chromosomal regions/genes involved in rearrangements provides interesting clues in the mechanisms of tumorigenesis [17]. CURRENT PROGRESS ON SOLID TUMORS: SLOWLY BUT SURELY The interphase FISH technology opened a new window of research opportunities in solid tumors, largely due to the availability of formalin-fixed, paraffin-embedded tumor blocks that could be used both for prospective and retrospective studies. The ability to keep the tissue architecture

6 50 Molecular Cytogenetics in Clinical Oncology Figure 3. Searching for genomic imbalances using CGH. A) Schematic representation of the procedure. B) Multiple imbalanced genomic regions detected in the prostate adenocarcinoma cell line ALVA-31. Each human chromosome is represented by its ideogram on the left and a graphic on the right. In the graphic, the black line represents the balanced hybridization, the red line on the left represents the threshold for loss, and the green line on the right represents the threshold for gain. The blue line shows the CGH profile, i.e., the ratio of tumor/reference DNA along the chromosome. Chromosomal regions that are deleted or gained above the threshold are highlighted by, respectively, red and green bars at the sides of the chromosome ideograms. Gains in copy numbers were found for specific regions in chromosomes 1, 3, 8, 10, 12, and 19 (green bars at right). Losses were found for regions in chromosomes 8, 10, 13, 17, 18, and 19 (red bars at left). Since DNA from a female donor was used as reference DNA, loss for chromosome X and gain for chromosome Y were not considered.

7 Varella-Garcia 51 SKY (Spectral karyotyping) Schorck et al. [15] M-FISH (Multiplex FISH) Speicher et al. [16] Principle: Combinatorial labeling scheme with 5 fluorochromes to distinctly paint the 24 human chromosomes Counterstaining with DAPI Abs/Em Abs/Em SA SG SGo SR FR M-FISH 433/ / / / /675 DAPI FITC Rh TxR Cy5 Cy5.5 SKY 367/ / / / / / nm Figure 4. Probe-labeling scheme used in each technique and for diamidinophenylindolc (DAPI), used as counterstain. Fluorophores largely used for M-FISH are SpectrumAqua (SA), SpectrumGreen (SG), SpectrumGold (SGo), SpectrumRed (SR), and Spectrum- FarRed (FR). SKY uses the fluorophores fluorescein isothiocyanate (FITC), Rhodamine (Rh), Texas Red (TxR), Cy5, and Cy5.5 and, consequently, to reveal cell-to-cell heterogeneity and detect small clones of genetically distinct cells is very appealing and may be essential in achieving breakthroughs in current cancer therapy, as stressed by Kobayashi et al. [18]. Those authors established four sublines from a primary pulmonary adenocarcinoma characterized by a poor prognosis. These sublines differed in morphological, biochemical, and genetic findings, and one of them exhibited a 12-fold amplification of the c- MYC oncogene and an increased sensitivity to cytosine arabinoside. However, a serious caveat for the broader application of FISH technology as a diagnostic tool for solid tumors is the scarceness of specific abnormalities associated with them. Recent progress in the DNAlabeling strategies using new fluorophores has partially circumvented this lack of specific targets, allowing the design of multicolor FISH probes that simultaneously target at least four different chromosomal regions in a single cell. The multicolor strategy is particularly useful for analysis of solid malignancies characterized by polyploidy or polysomy for multiple chromosomes, as addressed later in this review. In addition, among the latest publications, there is an emerging variability in applications of FISH as an adjunct diagnostic tool, which attests to the encouraging progress recently achieved. Examples are described below and summarized in Table 2. Figure 5. Spectral karyotype of a cell representative of the small cell lung carcinoma cell line UMC19. The top panel shows the classified image, and the bottom panel shows the inverted DAPI image. Chromosome derivatives from complex rearrangements involving two to five chromosomes are shown.

8 52 Molecular Cytogenetics in Clinical Oncology Figure 6. Images captured for the M-FISH karyotyping analysis: DAPI (counterstain) alone and combined with SpectrumAqua, SpectrumGreen, SpectrumGold, SpectrumRed, and SpectrumFarRed. Distinct chromosomes fluoresce in distinct colors due to the combinatorial probe-labeling scheme. Diagnosis and Follow-Up An interesting illustration is the differential diagnosis between teratoid/rhabdoid (AT/RT) and medulloblastomas/central primitive neuroectodermal (MB/PNET) tumors through the deletion of chromosome 22q11.2 sequences. AT/RT is a rare intracranial neoplasm typically unresponsive to therapy and rapidly fatal, and approximately 90% of patients present with a 22q deletion [19]. Bruch et al. [20] have shown the diagnostic use of a Table 2. Strategies used for application of FISH technology to clinical oncology 22q11.2 FISH probe in specimens with equivocal primitive central nervous system neoplasms. Difficult to classify tumors were assigned to AT/RT (when 22q deletion was present) or MB/PNET groups (when 22q deletion was not present), and after diagnosis reclassification, the study concluded that misdiagnosed cases of AT/RT had probably accounted for the worse prognosis associated with MB/PNET in younger infants. An attempt has been made to use FISH to monitor surgical margins. The combination of multiple DNA probes successfully identified chromosome imbalances associated with malignancy in head and neck squamous cell carcinomas that were also present in clinically normal adjacent cells, thereby detecting subclinical tumorigenesis [21]. In some patients, aneusomy in surgical margins was found in specimens for which the pathological evaluation was negative; these patients relapsed and died within the 3-year follow-up [22]. Although a question remained regarding the practicability of FISH as an intraoperative tool because, currently, results cannot be provided within an intraoperative time frame, the FISH results from these head and neck carcinomas strongly reinforced the need for adjunct therapy for patients with molecularly abnormal margins. Additional examples are emerging regarding the applicability of FISH assays to accurately monitor tumor response to therapy and to offer the clinician extra arguments for changing ineffective therapy at an earlier stage in the course of treatment. The cytologic examination of cerebrospinal fluid (CSF) is the primary method for evaluation of response to therapy for metastatic spread to the leptomeninges; a diffuse and often multifocal infiltration of the leptomeninges is estimated to occur in up to 8% of patients Application Strategy Reference Differential diagnosis of teratoid/rhabdoid (AT/RT) and Detection of 22q11.2 deletion as a Bruch et al. [20] medulloblastomas/central primitive ectodermal (MB/PNET) tumors marker for AT/RT Monitoring of surgical margins in head and neck carcinomas Detection of chromosomal aneusomy Barrera et al. [21, 22] in adjacent margins Monitoring response to therapy in metastatic cancer Detection of chromosomal aneusomy Van Oostenbrugge et al. [24] in cerebrospinal fluid Selection of breast cancer patients for target therapy with Detection of HER-2 gene amplification Pauletti et al. [33] trastuzumab in tumor cells Kakar et al. [34] Bartlett et al. [35] Tubbs et al. [36] Early detection of bladder cancer recurrence Detection of aneusomic cells in Halling et al. [45] voided urine Identification of patients at risk for carcinoma relapse Detection of aneusomic epithelial cells Engel et al. [49] in peripheral blood Assessment of thyroid cancer risk in therapeutic exposure to Detection of TP53 loss as a marker for Ramirez et al. [51] radioactive iodine lack of cellular response to DNA damage

9 Varella-Garcia 53 with systemic cancer [23]. However, this cytodiagnosis loses much of its sensitivity through the course of protracted therapy because CSF undergoes a decrease in cellularity, changes in cell morphology, and presentation of reactive ependymal cells. Van Oostenbrugge et al. [24] tested leptomeningeal metastases from three cases of non-hodgkin s lymphoma, three cases of breast cancer, and one malignancy with unknown primary site using a single FISH probe (chromosome 1 centromere), and found a better correlation between FISH and treatment response than routine cytodiagnosis. This study was performed retrospectively; however, the authors conceded that had the FISH results been taken into account, the treatment would have changed for four of the seven patients with normal or suspicious cytology in the ventricular CSF but aneusomic cells detected by FISH. Interestingly, in that study, the chromosome 1 probe was selected as a nonspecific indicator of chromosome aneusomy rather than a specific marker for the patients diseases. Markers for Prognosis and Targets to Therapy Examples of biomarkers that reliably predict responses to chemotherapy for solid tumors are rare but nonetheless rapidly increasing, and FISH technology has expanded the availability of molecular targets to be evaluated. The dualcolor FISH assay for evaluation of HER-2 (erbb2) gene amplification in breast cancer is probably the utmost investigated FISH test that has proven to be effective as a prognostic marker and predictor for response to therapy in solid malignancies. Overexpression of HER-2 has been found in 20%-30% of breast cancers and has been associated with a poor overall survival [25, 26] and response to therapy [27, 28]. Trastuzumab, a humanized monoclonal antibody that recognizes the HER-2 protein, produced objective responses, alone or in combination with chemotherapy, in breast cancer patients previously untreated and in the secondline setting [29-32]. Although it might be predicted that protein expression would be more effective for assessing response to trastuzumab therapy, because the antibody binds to the cell surface protein, results from a FISH assay using HER-2 and chromosome 17 centromere sequences showed that the presence of gene amplification (Fig. 7) correlated better with survival, was more accurate and reliable for selecting patients eligible for treatment with trastuzumab, and was superior for predicting response to therapy [33-36]. The success of targeting HER-2 for therapy in breast cancer has yet to be validated in other epithelial tumors. In lung and prostate carcinomas, it is clear that only a minority of patients overexpress HER-2 due to gene amplification [37, 38]. Intriguingly, the expression patterns of HER-2 are different in breast and lung carcinomas (Fig. 8), and most lung tumors display gains in copy number of HER-2 gene per cell due to chromosomal aneusomy rather than gene amplification. Initial reports show that prostate adenocarcinoma patients without HER-2 amplification have not responded to trastuzumab [39], while experiments in lung cancer cell lines indicate that trastuzumab was effective in reducing cell growth when multiple copies of the gene were present, even as a consequence of aneusomy [40]. Therefore, studies to determine the effectiveness of trastuzumab in relation to the molecular profiling of patients regarding HER-2 status are urged. Another member of the HER family of receptors, the epidermal growth factor receptor (EGFR), has been postulated as a potential marker for prognosis and a target for therapy in solid malignancies. Recent studies have shown promising results with tyrosine kinase inhibitors of EGFR (ZD1839 and OSI-774) in patients with esophageal [41] and lung carcinomas. In non-small-cell lung carcinomas, ZD1839 (Iressa ; AstraZeneca; London, UK) produced objective responses in about 15%-20% of patients with advanced disease who had failed one or two chemotherapy regimens (at least one platinum based) and in 10% of patients who had failed two or more prior chemotherapy regimens containing platinum and docetaxel, with a favorable adverse event profile [42, 43]. These findings raised considerable excitement regarding the potential role of Figure 7. Dual-color FISH assay in breast adenocarcinomas with sequences for the HER-2 gene labeled in red (SpectrumOrange ) and the chromosome 17 centromere labeled in SpectrumGreen. A) Ideogram of chromosome 17 with the location of each probe. B) Metaphase spread of cell line SKBr3 showing multiple chromosomes with multiple copies of the HER-2 gene. C) Interphase nuclei of a primary tumor with HER-2 gene amplification identified by clusters of fluorescent red spots.

10 54 Molecular Cytogenetics in Clinical Oncology Figure 8. Evaluation of HER-2 protein staining by immunohistochemistry and gene status by FISH in lung tumors [38]. Normal epithelium showing apical staining (A), well-differentiated tumor showing basolateral staining (B), moderately differentiated tumor scored as 2 + (C), and poorly differentiated tumor scored as 3 + (D). FISH assays showed three major patterns: balanced disomy for HER-2 gene and chromosome 17 (E), balanced gain for HER-2 gene and chromosome 17 (F), and HER-2 gene amplification with a gene/ chromosome ratio >2 (G). such therapies for lung cancer patients with less advanced disease and their potential to improve survival. Nevertheless, there is little information on how patients who might benefit from therapy with EGFR inhibitors should be selected. An interesting finding from the University of Colorado (Hirsch et al. submitted for publication) was the positive correlation between the level EGFR protein expression and copy number of the gene per cell, suggesting that an additive effect of gene copies is an important mechanism for EGFR protein expression in lung cancer cells. Other key regulator genes for cell growth, proliferation, and differentiation processes, such as the MYC (8q24), MYCN (2p24.1), and CCND1 (cyclin D1, 11q13), have been less evaluated clinically in solid tumors. However, when these genes are erroneously expressed due to genetic rearrangements, such as translocation or gene amplification, neoplastic transformation is activated, and their molecular pathways are potential therapy targets. FISH probes are commercially available for these genes, and in the future, may be utilized similarly to the HER-2 probe. Diagnosis of Recurrent Disease Detection of residual disease in patients with solid malignancies is an essential goal in clinical oncology, and FISH technology has already been introduced as a tool in this area. One of the promising applications of interphase FISH is the detection of tumor cells in body fluids through noninvasive procedures, as illustrated by the detection of tumor cells in urine of patients with urothelial carcinomas. Urinary cytology has been the reference test for the evaluation of symptomatic patients, for the detection of lesions in high-risk patients, and for the follow-up of patients with a prior history of transitional cell carcinoma (TCC). Although routine cytology has a relatively high level of specificity for the detection of high-grade in situ and invasive lesions, the sensitivity for patients with low-grade TCC is low, because these tumors may exfoliate cells that are cytologically indistinguishable from normal. Urine cytology is further limited by poor specificity for clinically significant lesions in cases that are classified as atypical on the basis of cells with equivocal cytologic features. The cytologic diagnosis of atypia may result from degenerative changes related to prior therapy or reactive changes that are of limited clinical significance. Therefore, the clinical outcomes for individual cases cannot be accurately determined by cytologic examination alone. As a result, many patients undergo unnecessary procedures to rule out malignancy based on limited specificity for clinically significant lesions. The U.S. Food and Drug Administration (FDA) recently approved the UroVysion Bladder Recurrence Cancer Test (Vysis; Downers Grove, IL) for the detection of TCC in urine cells. The UroVysion test was designed for interphase cell quantification of chromosomes 3, 7, 17, and 9p21 region (p16/cdkn2a gene), which were recognized as frequently involved in numerical anomalies in bladder cancer [44]. In a series of 265 patients, Halling et al. [45] demonstrated a significantly greater overall sensitivity of FISH over cytology for the detection of ptis, pt1- pt4, and grade 3 urothelial carcinoma. Validation of the UroVysion FISH assay as a screening tool for selection of patients with a history of bladder cancer who should be submitted to cystoscopy has been performed in numerous institutions, including the University of Colorado. Although false-negative FISH results can be expected in low-grade TCC due to frequent diploid chromosomal content, in our

11 Varella-Garcia 55 Figure 9. A highly aneusomic cell found in voided urine using the UroVysion bladder cancer. The top panel segmented images for each of the three chromosome enumeration probes (CEP) including centromeric sequences for chromosomes 3 (CEP 3 labeled in SpectrumRed ), 7 (CEP 7 labeled in SpectrumGreen ), and 17 (CEP 17 in labeled SpectrumAqua ) and the locusspecific indicator (LSI) probe for p16 (9p21 labeled in SpectrumGold). The bottom panel shows nuclei stained with DAPI and the multicolor image. The large nucleus at the top left in each image of the top panel shows, respectively, 6, 11, 10, and 6 copies of the DNA targets. small series of 19 patients, urine specimens from two patients with G1 tumors and one patient with a G2 tumor were classified as negative by cytology but abnormal by FISH. Figure 9 illustrates an abnormal cell found in the G2 TCC patient confirmed by cystoscopy in our series, showing multiple copies of each of the four DNA targets included in the UroVysion probe set. Interestingly, the UroVysion test also was proven to be effective in detecting tumor recurrence before evidence of urothelial carcinoma was found on biopsy, likely due to the ability of voided urine to present the entire urothelium versus the regional evaluation of the biopsy [45]. A follow-up biopsy revealed recurrent urothelial carcinoma in 7 out of 11 patients identified by Halling at al. [45] as abnormal by FISH but with a negative initial biopsy. Therefore, patients with positive FISH but negative biopsy should be carefully followed, as they are at risk of harboring occult disease. Other recently developed multicolor FISH probes were demonstrated to be potentially useful laboratory tools in the effort to reduce mortality from solid tumors, due to their greater ability to diagnosis early changes, and therefore, prevent disease progression. Examples include the analysis of cells collected through minimally invasive approaches, such as epithelial cells obtained in fine needle aspirates from ductal carcinoma of the breast [46] and in bronchial biopsies from smokers at risk for lung carcinoma [47]. The multicolor probe used in the bronchial biopsies study, LAVysion (Vysis), addresses four chromosomal regions or genes commonly involved in gains in lung carcinomas: 5p15, 6 centromere, 7p12 (EGFR), and 8q24 (MYC), as illustrated in Figure 10. Normal bronchial cells are shown with two copies of each of these DNA targets, and a nucleus from a squamous cell carcinoma is shown with multiple copies of each target. The development of innovative technologies for unbiased recovery and enrichment of tumor cells from Figure 10. Multitarget probe set (LAVysion ; Vysis) for detection of aneusomy in lung tumors. A) Four chromosomal targets are addressed: 5p15 (labeled in SpectrumGreen ), chromosome 6 centromere (in SpectrumAqua ), EGFR sequences at 7p12 (in SpectrumRed ), and MYC sequences at 8q24 (in SpectrumGold ). B) Normal bronchial epithelial nuclei showing two copies of each probe. C) Highly aneusomic nucleus from a squamous cell lung carcinoma showing additional copies for all four DNA targets.

12 56 Molecular Cytogenetics in Clinical Oncology peripheral blood or bone marrow aspirates using magneticactivated cell-sorting systems [48], coupled with accurate tests to characterize these cells, is expected to better monitor the response to cytotoxic or hormonal therapy and to identify patients at risk for relapse. FISH has been proposed as an essential test to be combined with immunophenotyping for evaluation of potential tumor cells selected from blood or bone marrow. Engel et al. [49], in dilution assays with the breast cancer cell line MCF7 and normal peripheral lymphocytes, demonstrated the higher sensitivity of FISH tests using centromeric probes compared with flow cytometry sorting using anticytokeratin and anti-cd45 antibodies. In addition, these authors successfully identified circulating tumor cells in the peripheral blood of 30 out of 43 patients with breast or ovarian cancer using one locusspecific (17q11-12) and three centromeric probes (chromosomes 7, 12, and 17) in combination with anticytokeratin and anti-cd45 antibodies. Risk Assessment There are persisting concerns regarding the carcinogenic risk to therapeutically irradiated patients, because it is well known that ionizing radiation is a major aneugenic and clastogenic agent [50]. The application of emerging molecular cytogenetic methods to risk assessment may help to clarify the uncertainties of low-risk exposure to cancer treatments, such as radioactive iodine for thyroid disease. Ramirez et al. [51] tested a new FISH probe set, that included the centromere of chromosome 17 and the p53 locus (17p13.1), in buccal cells from patients with thyroid cancer before and after radioactive iodine treatment. The rationale for the selection of these probes was that the failure to express a functional nuclear phosphoprotein p53 plays a critical role in the cellular response to DNA damage, leads to aneuploidy in vitro and in vivo, and increases the incidence of spontaneous tumors [52, 53]. Inactivation of p53 is frequently associated with the loss or mutation of its encoding gene, TP53. In that study, the radiationinduced chromosome breakage resulted in a significant increase in 17p gains and losses in exposed individuals. Therefore, the 17cen/p53 FISH assay may be a potential biomarker for cancer risk in people therapeutically exposed to radiation. CONCLUSIONS The identification of specific chromosomal abnormalities in leukemias, lymphomas, and soft tissue tumors has provided critical insight into the molecular changes that underlie carcinogenesis. In solid malignancies, the progress was less substantial until the development of molecular cytogenetic techniques such as interphase FISH, CGH, and multicolor karyotyping. SKY and M-FISH, the two major multicolor karyotyping techniques, contribute to the identification of complex chromosomal changes associated with solid tumors, and CGH has highlighted critical regions harboring oncogenes and tumor suppressor genes. Interphase FISH technology has been demonstrated to be specially suitable to bridge basic research to clinical practice, and its applicability has increased substantially in a short time. These new methodological strategies are rapidly expanding the application of cytogenetic assays to critical areas in solid malignancies at both the basic science and clinically applied levels. Availability of reliable and specific markers for individual diseases or stages is a critical shortcoming in the current application of FISH to solid tumors, and the development of a substantial number of DNA probes is warranted for a more powerful use of the technology. In this regard, the massive volume of molecular information that has been constantly released in both the fields of genome databases and tumor profiling is expected to minimize the current difficulties. Additionally, automated devices are currently under development to process and analyze interphase FISH assays, which are expected to reduce the burden of the labor-intensive aspect of the molecular cytogenetics procedures. As a consequence, molecular cytogenetic technology is shortly expected to extend its contribution to more accurate tumor profiling; improved early detection, diagnosis, and prognosis; selection of target therapeutic approaches, and monitoring clinical outcome in patients with solid malignancies. ACKNOWLEDGMENTS Supported in part by the following National Cancer Institute grants: Cancer Center Core Grant P30-CA46934, Specialized Program of Research Excellence P01-CA58187, and Early Detection Research Network U01-CA Critical comments provided on this review by Dr. Fred Hirsch were deeply appreciated. REFERENCES 1 Nowell P, Hungerford D. A minute chromosome in human granulocytic leukemia. Science 1960;132: Rowley JD. The role of chromosome translocations in leukemogenesis. Semin Hematol 1999;36(suppl 7): Pinkel D, Straume T, Gray JW. Cytogenetic analysis using quantitative, high-sensitivity, fluorescence hybridization. Proc Natl Acad Sci USA 1986;83: Rowley JD, Diaz MO, Espinosa R et al. Mapping chromosome band 11q23 in human acute leukemia with biotinylated

13 Varella-Garcia 57 probes: identification of 11q23 translocation breakpoints with a yeast artificial chromosome. Proc Natl Acad Sci USA 1990;87: Tkachuck DC, Westbrook C, Andreef M et al. Detection of bcr-abl fusion in chronic myelogenous leukemia by in situ hybridization. Science 1990;250: Paskulin G, Philips G, Morgan R et al. A pre-clinical evaluation of probes to detect t(8;21) AML minimal residual disease by fluorescence in situ hybridization. Genes Chromosomes Cancer 1998;21: Varella-Garcia M, Murata-Collins JL, Ai H et al. Interphase cytogenetic analysis of minimal residual disease in t(8;21) AML and enrichment using CD34 selection. Leukemia 2001;15: Dang TP, Gazdar AF, Virmani AK et al. Chromosome 19 translocation, overexpression of notch3 and human lung cancer. J Natl Cancer Inst 2000;92: French C, Miyoshi I, Aster JC et al. BRD4 bromodomain gene rearrangement in aggressive carcinoma with translocation t(15;19). Am J Pathol 2001;159: Saunders WS, Shuster M, Huang X et al. Chromosomal instability and cytoskeletal defects in oral cancer cells. Proc Natl Acad Sci USA 2000;97: Hayashi Y. The molecular genetics of recurrent chromosome abnormalities in acute myeloid leukemia. Semin Hematol 2000;37: Boomer T, Varella-Garcia M, McGavran L et al. Detection of E2A translocations in leukemias via fluorescence in situ hybridization. Leukemia 2001;15: Kallioniemi A, Kallioniemi OP, Sudar D et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992;258: Struski S, Doco-Fenzy M, Cornillet-Lefebvre P. Compilation of published comparative genomic hybridization studies. Cancer Genet Cytogenet 2002;135: Schrock E, du Manoir S, Veldman T et al. Multicolor spectral karyotyping of human chromosomes. Science 1996;273: Speicher MR, Ballard SG, Ward DC. Karyotyping human chromosomes by combinatorial multi-fluor FISH. Nat Genet 1996;12: Schrock E, Padilla-Nash H. Spectral karyotyping and multicolor fluorescence in situ hybridization reveal new tumor-specific chromosomal aberrations. Semin Hematol 2000;37: Kobayashi S, Noda M, Isogami K et al. c-myc amplification and enhancement of sensitivity to cytosine arabinoside: an in vitro and in vivo study on four sublines established from a pulmonary adenocarcinoma. Surg Today 2002;32: Biegel JA, Allen CS, Kawasaki K et al. Narrowing the critical region for a rhabdoid tumor locus in 22q11. Genes Chromosomes Cancer 1996;16: Bruch LA, Hill DA, Cai DX et al. A role for fluorescence in situ hybridization detection of chromosome 22q dosage in distinguishing atypical teratoid/rhabdoid tumors from medulloblastomas/central primitive neuroectodermal tumors. Hum Pathol 2001;32: Barrera JE, Hong Ai, Pan X et al. Malignancy detection by molecular cytogenetics in clinically normal mucosa adjacent to head and neck tumors. Arch Otolaryngol Head Neck Surg 1998;124: Barrera JE, Varella-Garcia M. Chromosomal aneuploidy as a predictor for poor outcome in patients with head and neck cancer. Arch Otolaryngol Head Neck Surg 2001;127: Posner JB. Neurologic Complication of Cancer. Oxford University Press USA, New York, NY, Van Oostenbrugge RJ, Hopman AH, Arends JW et al. Treatment of leptomeningeal metastases evaluated by interphase cytogenetics. J Clin Oncol 2000;18: Slamon DJ, Clark GM, Wong SG et al. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235: Slamon DJ, Godolphin W, Jones LA et al. Studies of the HER-2/neu proto-oncogene in human and ovarian cancer. Science 1989;244: Muss HB, Thor AD, Berry DA et al. C-erb-B2 expression and response to adjuvant therapy in women with node-positive early breast cancer. N Engl J Med 1994;330: Paik S, Bryant J, Park C et al. c-erb-b2 and response to doxorubicin in patients with axillary lymph node-positive, hormone receptor-negative breast cancer. J Natl Cancer Inst 1998;90: Baselga J, Trupathy D, Mendelsohn J et al. Efficacy and safety of weekly intravenous recombinant humanized anti-p185her2 monoclonal antibody in patients with HER2/neu overexpressing metastatic breast cancer. J Clin Oncol 1996;14: Baselga J, Norton L, Albanell J et al. Recombinant humanized anti-her antibody (Herceptin) enhances the antitumor activity of paclitaxel and doxorubicin against HER2/neu overexpressing human breast cancer xenografts. Cancer Res 1998;58: Cobleigh MA, Vogle CL, Tripathy D et al. Efficacy and safety of Herceptin (humanized anti-her2 antibody) as a single agent in 222 women with HER2 overexpression who relapsed following chemotherapy for metastatic breast cancer. Proc Am Soc Clin Oncol 1998;17:97a. 32 Slamon D, Leyland-Jones B, Shak S et al. Addition of Herceptin (humanized anti-her2 antibody) to first line chemotherapy for HER2 overexpressing metastatic breast cancer (HER2 + /MBC) markedly increases anticancer activity: a randomized, multinational controlled phase III trial. Proc Am Soc Clin Oncol 1998;17:98a. 33 Pauletti G, Dandekar S, Rong HM et al. Assessment of methods for tissue-based detection of the HER2/neu alteration in human breast cancer: a direct comparison of fluorescence in situ hybridization and immunohistochemistry. J Clin Oncol 2000;18: Kakar S, Puangsuvan N, Stevens JM et al. HER-2/neu assessment in breast cancer by immunohistochemistry and fluores-

14 58 Molecular Cytogenetics in Clinical Oncology cence in situ hybridization: comparison of results and correlation with survival. Mol Diagn 2000;5: Bartlett JM, Going JJ, Mallon EA et al. Evaluating HER2 amplification and overexpression in breast cancer. J Pathol 2001;195: Tubbs RR, Pettay JD, Roche PC et al. Discrepancies in clinical laboratory testing of eligibility for trastuzumab therapy: apparent immunohistochemical false-positives do not get the message. J Clin Oncol 2001;19: Bubendorf L, Kononen J, Koivisto P et al. Survey of gene amplifications during prostate cancer progression by highthroughput fluorescence in situ hybridization on tissue microarrays. Cancer Res 1999;59: Hirsch FR, Varella-Garcia M, Franklin WA et al. Evaluation of HER-2/neu gene amplification and protein expression in nonsmall cell lung carcinomas. Br J Cancer 2002;86: Morris MJ, Reuter VE, Kelly WK et al. HER-2 profiling and targeting in prostate carcinoma. Cancer 2002;94: Bunn Jr PA, Helfrich B, Soriano A et al. Expression of Her- 2/neu in human lung cancer cell lines by immunohistochemistry and fluorescence in situ hybridization and relationship to in vitro cytotoxicity by trastuzumab and chemotherapeutic agents. Clin Cancer Res 2001;7: Cohen EEW, Rosen F, Dekker A et al. Phase II study of ZD1839 (Iressa) in recurrent or metastatic squamous cell carcinoma of the head and neck (SCCHN). Proc Am Soc Clin Oncol 2002;21:225a. 42 Fukuoka M, Yano S, Giaccone G et al. Final results from a phase II trial of ZD1839 (Iressa) for patients with advanced non-small cell lung carcinoma (IDEAL 1). Proc Am Soc Clin Oncol 2002;21:298a. 43 Kris R, Natale B, Herbst RS et al. A phase II trial of ZD1839 (Iressa) in advanced non-small cell lung carcinoma (NSCLC) who had failed platinum- and docetaxel-based regimens (IDEAL 2). Proc Am Soc Clin Oncol 2002;21:292a. 44 Sokolova IA, Halling KC, Jenkins RB et al. The development of a multitarget, multicolor fluorescence in situ hybridization assay for the detection of urothelial carcinoma in urine. J Mol Diag 2000;2: Halling KC, King W, Sokolova IA et al. A comparison of cytology and fluorescence in situ hybridization for the detection of urothelial carcinoma. J Urol 2000;164: Heselmeyer-Haddad K, Chaudhri N, Stoltzfus P et al. Detection of chromosomal aneuploidies and gene copy number changes in fine needle aspirates is a specific, sensitive, and objective genetic test for the diagnosis of breast cancer. Cancer Res 2002;62: Romeo M, Sokolova IA, Morrison LE et al. Chromosomal abnormalities in non-small cell lung carcinomas and in bronchial epithelium of high risk smokers detected by multitarget interphase fluorescence in situ hybridization. J Mol Diagn (in press). 48 Franklin W, Shpall EJ, Archer P et al. Immunocytochemical detection of breast cancer cells in marrow and peripheral blood of patients undergoing high dose chemotherapy with autologous stem cell support. Breast Cancer Res Treat 1996;41: Engel H, Kleespies C, Friedrich J et al.. Detection of circulating tumour cells in patients with breast or ovarian cancer by molecular cytogenetics. Br J Cancer 1999;81: Natarajan AT, Vyas RC, Wiegant K et al. A cytogenetic follow up study of the victims of a radiation accident in Goiania (Brazil). Mutat Res 1991;247: Ramirez MJ, Puerto S, Galofre P et al. Multicolour FISH detection of radioactive iodine-induced 17cen-p53 chromosomal breakage in buccal cells from therapeutically exposed patients. Carcinogenesis 2000;8: Chen LC, Neubauer A, Kurisu W et al. Loss of heterozygosity on the short arm of chromosome 17 is associated with high proliferative capacity and DNA aneuploidy in primary human breast cancer. Proc Natl Acad Sci USA 1991;88: Livingstone LR, White A, Sprouse J et al. Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell 1992;70: