Detection of cytogenetic changes and chromosomal aneuploidy with fluorescent in situ hybridization in cytological specimens of oral cancers in Fanconi anemia—Proof of concept

Abstract Objectives Fanconi anemia (FA) is a rare inherited DNA instability disorder with a remarkably elevated risk of neoplasia compared with the general population, mainly leukemia and squamous cell carcinoma (SCC). Two thirds of the SCCs arise in the oral cavity and are typically preceded by visible lesions. These lesions can be classified with brush biopsy‐based cytological methods regarding their risk of a malignant transformation. As a proof of concept, this study aims to investigate genetic changes and chromosomal aneuploidy using fluorescent in situ hybridization (FISH) on oral squamous cells derived from FA affected individuals. Material and Methods Five FA oral SCC (OSCC) tumor cell lines, one FA OSCC cervical lymph node metastasis as well as tumor‐negative and atypical smears from oral brush biopsies were analyzed with FISH probes covering 5p15.2, MYC, EGFR, TERC, 9q34.1, CCND1, 9p21 and centromeres of chromosomes 3, 6, 7, 9, 11, and 17. Results OSCC specimens showed gains of all analyzed chromosomal regions. Chromosomal aneuploidy was observed in five of the six OSCC specimens in two multicolor FISH assays with panels of four probes each. Five out of six OSCC specimens displayed a relative deletion of 9p21. Applied on atypical brush biopsy‐based smears, chromosomal aneuploidy was detected in malignant lesions but not in the smear derived from a severe parodontitis. Conclusions As proof of concept, FISH was able to detect genetic changes and chromosomal aneuploidy discriminating oral cancer from noncancerous lesions in individuals with FA. This supports its application on oral brush biopsy‐based cytology.


| INTRODUCTION
Fanconi anemia (FA) is a rare inherited bone marrow failure syndrome that in most populations affects approximately one in 200,000 live births (Dufour, 2017) and is characterized by multiple malformations, bone marrow failure, endocrine disorders and a dramatically increased risk of developing hematological malignancies and solid tumors (Dufour, 2017). FA is caused by loss-of-function mutations in any of 22 known genes (FANCA-W) leading to spontaneous and inducible genetic instability, hypersensitivity to DNA crosslinkers and oxidative stress, and impaired telomere maintenance and replication (Michl et al., 2016;Nalepa & Clapp, 2018). FA patients especially display a 500-700-fold higher risk of head and neck squamous cell carcinoma (HNSCC) compared with the general population, with two thirds of these lesions located in the oral cavity (Kutler et al., 2016).
The treatment of oral squamous cell carcinoma (OSCC) is particularly challenging in FA patients since standard chemoradiation protocols cannot be applied due to the underlying genetic defect in DNA repair pathways that induces DNA-cross-linker hypersensitivity also in all normal noncancerous cells. Thus, frequent oral inspections starting at the age of ten is indicated (Fanconi Anemia Research Fund, 2020), in order to discover OSCC at an early stage of development or ideally potentially malignant lesions that likely would progress, where surgery alone is still adequate. Most often, OSCC in FA arises in visible lesions (Velleuer et al., 2020). These are observed to be chronic and frequently syn-and metachrone, which hampers repeated controls with tissue biopsies. There is evidence for noninvasive, highly sensitive and specific evaluation of oral visible lesions with brush biopsy-based oral cytology and DNA ploidy analysis, that identifies oral lesions at risk for malignant transformation (Velleuer et al., 2020). Although, lesions that do not require invasive treatment can be identified with this method, there is still a need to further reduce unnecessary tissue biopsies due to equivocal cytological results that cannot be analyzed for DNA ploidy with DNA image cytometry because of a low number of suspicious cells (<100 cells).
Thus, fluorescent in situ hybridization (FISH) emerges as a molecular method with high accuracy to detect cytogenetic changes and chromosomal aneuploidy in cytological specimens (Savic & Bubendorf, 2016) (Campbell et al., 2018;Cancer Genome Atlas Network, 2015). Moreover, many of the FISH studies on sporadic OSCC use probes for chromosomal regions harboring the above-mentioned genes or centromere probes (Kokalj Vokac et al., 2014;Lim et al., 2014;Wangsa et al., 2016). This study aims to investigate genetic amplifications, deletions, gains, losses and especially chromosomal aneuploidy in cells from OSCC in FA using FISH.
More than half of the worldwide available FA-related OSCC cell lines and in addition one cytological FA OSCC sample are analyzed. In a proof of concept, these FA OSCCs are compared with normal and some equivocal oral brush biopsy-based specimens to test the ability of discriminating OSCC from noncancerous oral squamous cells in FA.   (Feulgen & Rossenbeck, 1924) for the analysis of DNA ploidy. The processes of cover slipping, de-staining of the Papanicolaou stain (if necessary), relocalization of suspicious cells and the measurement of the nuclear DNA content, were performed as previously described (Böcking, 1995;Remmerbach et al., 2009;Schramm et al., 2011). A manual MotiCyte DNA workstation (Motic ® ) was used, that provides shading and glare correction. Technical and diagnostic recommendations for the measurements, including the definition of a DNA stemline, stemline-aneuploidy and single cells with a DNA content of >9c (single cell aneuploidy), according to the consensus reports of the European Society for Analytical Cellular Pathology (ESACP) were considered (Bocking et al., 1995;Haroske et al., 1998Haroske et al., , 2001

| FISH
Four panels of 13 commercially available locus specific (LSI) and centromere (CEP) FISH probes were arranged, including the UROvysion © multicolor probes a separate panel (Table 2). All probes were obtained from Abbott (Abbott Laboratories).
The cytological specimens were investigated with the four panels according to the protocol of Onofre et al. (2008), with the following modifications: After rehydration with an ethanol sequence, the uncovered cytological brush-biopsy specimens were washed in 2× saline-sodium citrate buffer for 5 min at 80°C in a water bath. The specimens were then digested by using pepsin 0.2% (Sigma-Aldrich Co.), for 15 min in a humidified chamber at 37°C and fixed in formalin 1% for 5 min. After dehydration with an ethanol sequence, the FISH probe mix consisting of 3.5 µl Vysis IntelliFISH Hybridization Buffer (Abbott Laboratories), 0.5 µl purified water, and 1 µl probe set was applied to the smears.
The hybridized areas on the specimens were analyzed for atypical nuclei (nuclear enlargement, irregular shape, patchy DAPI staining) using an Axio Imager A1 microscope (Carl Zeiss) equipped with ×63 and ×100 oil immersion objective lenses and an AxioCam MRm video camera (Carl Zeiss). Up to 50 nuclei of interest for each specimen were analyzed by two independent observers and the number of signals for each probe was registered. In addition, the OSCC specimens were analyzed in duplicates. The nuclei of inflammatory cells (e.g., neutrophils) were used as an internal reference for hybridization quality.

| FISH evaluation protocol
FISH data were analyzed with descriptive statistics. The mean number of signals per cell and the number of cells with more than two FISH signals were calculated for all cases. A CCND1 amplification was defined as the CCND1/CEP11 ratio ≥2 (panel 3, Table 2), a homozygous (biallelic) deletion of 9p21 was defined as the lack of yellow signals in the UROvysion probe set (Panel 4, Table 2). A relative deletion of 9p21 in a given nucleus was observed if the number of the signals was lower than the number of signals of the centromeric probes for chromosomes 3, 7, and 17 (Zellweger et al., 2006). The use of differentially-colored FISH probes in a multicolor panel like the UROvysion panel (panel 4,  Abbreviations: CEP, centromeric probe; LSI, locus specific probe. and losses and the analysis of chromosomal aneuploidy per cell. A cell was defined as chromosomally aneuploid with a gain of two or more of the four probes in panels 1 and 4. A tetrasomic pattern, that occurs both in malignant tumors and euploid polyploidization, was not considered as chromosomally aneuploid to prevent misinterpretation. Tetrasomy was defined as the presence of four copies of the genes in a nucleus. A deviation of ±1 copy of one of the genes, as previously suggested (Schramm et al., 2011), was accepted to consider diagnostic errors. Thus, the following patterns of gene copies were accepted as tetrasomy, regardless of the order: 4-4-4-4, 5-4-4-4, 3-4-4-4.

| Ethics Statement
The cytological specimens were leftover samples of our study 3 | RESULTS

| DNA image cytometry in OSCC samples and normal controls
The DNA ploidy results of the OSCC and tumor negative specimens are reported in Table 3. Figure 1 shows DNA histograms of cell lines VU1604-T (P1) and CCHMC-FASCC2 (P4). All OSCC specimens show stemline aneuploidy and single cell aneuploidy in the DNA image cytometry analysis, indicating a malignant transformation. In contrast, both tumor negative samples are euploid.
One of them, obtained from a leukoplakia with inflammation at the gingiva, shows an euploid-polyploid histogram with a peridiploid and a peri-tetraploid peak, consistent with euploid polyploidization.  (Table 4) in OSCC P6 in the current study as low-level (Blessmann et al., 2013). The mean number of cells that demonstrate a relative deletion of 9p21 in the respective OSCC and tumor negative specimens is shown in Table 4. With the exception of specimen P5, a relative 9p21 deletion at least in 29.5 out of 50 cells was detected in all 5 OSCC specimens. The OSCC specimens do not show any homozygous deletion of 9p21, and the tumor negative controls have always two copies of 9p21.    Table 4. Specimen A1 was cytologically classified as suspicious for malignancy, as previously reported (Velleuer et al., 2020)

| DISCUSSION
Although early OSCC detection and treatment is known to improve survival for the general population (Amit et al., 2013), most OSCCs in FA are diagnosed at late stages (Kutler et al., 2016).
Brush-biopsy based cytology including the analysis of DNA ploidy of visible, oral lesions is not invasive, has a high negative predictive value and the ability to detect OSCCs and potentially malignant lesions in FA at a noninvasive or early stage (Velleuer et al., 2020). However, due to the lack of a sufficient number of atypical cells in some of those specimens, an alternative method to DNA ploidy testing is necessary to assist the microscopic cytological diagnosis. FISH requires very few atypical cells for the detection of chromosomal changes and aneusomy. The main limitation of this methodology in our proof-of-concept study is the lack of a cutoff determination for the diagnosis of aneusomy and its validation in a larger cohort. The main advantage is the transferability of the approach to oral brush biopsy-based cytology which is noninvasive and well-tolerated. All FISH probes showed high mean copy number gains ≥ 3 in the OSCC specimens, but no gains in the normal controls. This is consistent with data from "The Cancer Genome Atlas Network," that report copy number alterations in most HNSCC (Cancer Genome Atlas Network, 2015). To avoid interference with euploid polyploidization in inflammatory conditions and tissue repair (Biesterfeld et al., 1994) that is, in oral graft versus host disease following a hematological stem cell transplantation (Grein Cavalcanti et al., 2015), the mean copy numbers should be different to the value four for the FISH application in a multicolor probe set.
F I G U R E 1 DNA histograms of cell lines VU1604-T (a) and CCHMC-FASCC2 (b). DNA-content in c-units (x-axis) is plotted against number of cells (y-axis). (a) DNA aneuploidy with two abnormal DNA stemlines (modal values at 3.3c and 6.5c), and a high number of 23 9c exceeding events. (b) DNA aneuploidy with two abnormal stemlines (modal values at 2.33c and 4.77c), and one 9c exceeding event It is important to consider euploid polyploidization, that has 2 n chromosomal sets (e.g., tetrasomy), in scoring algorithms for a multicolor FISH assay applied to equivocal cytological specimens (Schramm et al., 2011) to reduce the risk of false-positive diagnosis regarding a malignant transformation. While copy number gains and relative loss of 9p21 were frequently observed in the OSCC specimens, the amplification of CCND1 or a high-level gain of MYC could be detected with FISH only in specimen P6, that was provided by a patient with a highly aggressive OSCC with multiple metastases. This finding is in line with a previous study, in which aberration in CCND1 number was described as a predictor of cervical lymph node metastasis (Myo et al., 2005).
The analysis of DNA ploidy is a valuable tool for detecting a malignant transformation of oral squamous epithelial cells in FA (Velleuer et al., 2020) and therefore we used this method to be compared with the FISH results. The analysis of DNA ploidy correlates with the chromosomal aneuploidy detected with multicolor FISH. In general, FISH is commonly used in equivocal cytology of different organs, like lungs or bile ducts (Levy et al., 2007;Schramm et al., 2011), but was not tested previously on equivocal oral cytology, especially in the context of FA. Using DNA image cytometry, we observed DNA aneuploidy in all OSCC specimens, but not in the tumor-negative controls. These results were consistent with the analysis of chromosomal aneuploidy with the two 4-probe multicolor FISH panels, which showed high mean numbers of cells with chromosomal aneuploidy in the OSCC specimens with the exception of OSCC P4. The application of multicolor-FISH on brush biopsy-based cytology was investigated with the panel 1 FISH probe set (Table 4) on three cytologically equivocal specimens. Herein the two cytological specimens obtained from a cancerogenic field and a minimal invasive OSCC showed much higher numbers of chromosomally aneuploidy cells compared with the specimen brushed from a severe parodontitis with tissue regeneration.
In conclusion, as a proof of concept, our study has shown the potential for the application of several FISH probe panels for the F I G U R E 3 Representative images of the cytological preparations (Papanicolaou stain, original magnification ×40) and fluorescent in situ hybridization (FISH) analyses (original magnification ×63) using probes from panel 1 (5p15.2 (SpectrumGreen), MYC (SpectrumGold), CEP 6 (SpectrumAqua), and EGFR (SpectrumRed). N1a: normal intermediate squamous epithelial cells in a tumor-negative brush biopsy-based specimen; N1b: Corresponding FISH shows disomic pattern of signals. Two SpectrumGreen (5p15.2) signals were counted as one split signal; A1a: Atypical superficial squamous cells with enlarged hyperchromatic nuclei, irregular nuclear contours and enhanced nuclear to cytoplasmic ratio. A1b: Corresponding FISH shows gain of 5p15.2, MYC and CEP 6 and disomy of EGFR. P6a: Malignant squamous cells from an OSCC cervical lymph node metastasis with nuclear pleomorphism and high nuclear to cytoplasmic ratio; P6b: Corresponding FISH shows gain of MYC with clusters and gain of 5p15.2, CEP 6 and EGFR