Abstract
Introduction
Topoisomerase 1 (TOP1) and 2A (TOP2A) are potential predictive biomarkers for irinotecan and anthracycline treatment, respectively, in colorectal cancer (CRC), and we have recently reported a high frequency of gene gain of the TOP1 and TOP2A genes in CRC. Furthermore, Mismatch Repair (MMR) subtypes of CRC have been associated with benefit from adjuvant chemotherapy of primary CRC. Given the involvement of the topoisomerase enzymes in DNA replication and repair, we raised the hypothesis that an association may exist between TOP gene copy numbers and MMR proficiency/deficiency in CRC.
Material and methods
Test cohort: FISH analysis with an in‐house TOP1/CEN20 probe mix and a commercially available TOP2A/CEN17 (Dako, Glostrup, Denmark) probe mix was performed on archival formalin fixed paraffin embedded (FFPE) tissue samples from 18 patients with proficient MMR (pMMR) CRC and 18 patients with deficient MMR (dMMR) CRC. TOP1 and TOP2A gene copy numbers and their ratios per nucleus were correlated with MMR status using the Mann–Whitney test. Validation cohort: FFPE samples from 154 patients with primary stage III CRC (originally included in the RANX05 study) were classified according to MMR status by immunohistochemical analysis using validated antibodies for MLH1, MLH2, MSH6 and PMS2, and information on TOP1, CEN20, TOP2A and CEN17 status was previously published for this cohort.
Results
The observed TOP1 gene copy numbers in the 36 CRC test cohort were significantly greater (p < 0.01) in the pMMR subgroup (mean: 3.84, SD: 2.03) than in the dMMR subgroup (mean: 1.50, SD: 0.12). Similarly, the TOP2A copy numbers were significantly greater (p < 0.01) in the pMMR subgroup (mean: 1.99, SD: 0.52) than in the dMMR subgroup (mean: 1.52, SD: 0.10). These findings were confirmed in the validation cohort, where in the pMMR subgroup 51% had ≥2 extra TOP1 copies per cell, while all tumors classified as dMMR had diploid TOP1 status and mean TOP2A copy numbers were 2.30 (SD: 1.36) and 1.80 (SD: 0.31) (p = 0.01) in the pMMR subgroup vs. dMMR subgroup, respectively.
Discussion and conclusion
Our results show that TOP1 and TOP2A gene copy numbers are increased in the pMMR subgroup. We propose that this preference may reflect a selective pressure to gain and/or maintain the gained extra copies of topoisomerase genes whose products are required to cope with high replication stress present in the pMMR tumors, thereby providing a survival advantage selectively in pMMR tumors. Future studies should test this concept and explore potential differences between pMMR and dMMR tumors in response to Top1 and Top2 inhibitors.
Keywords: FISH, Topoisomerase 1, Topoisomerase 2A, MMR, Chromosomal instability, Gene copy number
Highlights
TOP1 and TOP2A gene copy numbers are almost exclusively elevated in pMMR CRC.
Results are validated in two independent cohorts.
More than 50% of pMMR CRC tumors in both cohorts had minimum 4 copies of the TOP1 gene.
Abbreviations
- TOP1
gene symbol for the gene encoding topoisomerase 1
- TOP2A
gene symbol for the gene encoding topoisomerase 2 alpha
- CEN17
symbol for the chromosomal region encoding centromere 17
- CEN20
symbol for the chromosomal region encoding centromere 20
- CRC
colorectal cancer
- MMR
mismatch repair
- pMMR
proficient mismatch repair
- dMMR
deficient mismatch repair
- FISH
fluorescence in situ hybridization
- IHC
immunohistochemistry
- FFPE
formalin-fixed paraffin-embedded
- MSI
microsatellite instability
- CIN
chromosomal instability
- IDL
insertion/deletion loops
- mCRC
metastatic colorectal cancer
1. Introduction
An increasing understanding of the molecular biology of cancer and the development of new targeted drugs have created a platform for the development of predictive markers aiming at introduction of individualised therapies. The topoisomerase inhibitor irinotecan is included in treatment regimens for metastatic colorectal cancer (mCRC) but no predictive biomarkers for irinotecan efficacy have so far been substantially investigated and found useful in a clinical setting (National Comprehensive Cancer Network, 2014 (http://NCCN.org)).
The topoisomerase‐1 (Top1) protein is the cellular target of campothecins and its derivatives (i.e. irinotecan) and thus represents a putative irinotecan predictive biomarker in CRC (Pommier et al., 2010). A large randomized clinical trial including material from 1313 patients with mCRC from the FOCUS trial showed a positive association between high Top1 protein expression investigated by immunohistochemistry (IHC) and benefit from irinotecan in combination with 5‐fluorouracil as first‐line treatment (Braun et al., 2008). An association between IHC Top1 protein expression and benefit from irinotecan‐containing chemotherapy has also been reported in a non‐randomized retrospective study with a cohort of 498 patients (Kostopoulos et al., 2009) but due to the design of this study no distinction between prognosis and prediction could be obtained (Boonsong et al., 2002), (Staley et al., 1999), (Paradiso et al., 2004), (Tsavaris et al., 2009). However, analyses of the CAIRO IHC study based on the same anti‐Top1 antibody as the one used in the FOCUS study failed to confirm the results from the latter (Koopman et al., 2009), thereby leaving the notion of Top1 status as a potential biomarker for response to irinotecan uncertain.
Topoisomerase‐2A (Top2A) is the cellular target of anthracyclines and two meta‐analyses recently confirmed that breast cancer patients with increased TOP2A gene copy number had increased benefit from adjuvant anthracycline treatment (Di Leo et al., 2011), (Du et al., 2011). A number of small and non‐randomized clinical studies have suggested a 10–20% objective response rate following anthracycline treatment of mCRC (Michaelson et al., 1982), (Wils, 1984) and we have now initiated a prospective phase II clinical trial in which metastatic CRC patients with TOP2A gene amplification in their cancer cells are offered treatment with the Top2 inhibitor epirubicin (EudraCTno. 2013‐001648‐79).
We have previously shown, in a cohort of 154 stage III CRC patients that both elevated TOP1 and TOP2A gene copy numbers are frequent findings in metastatic CRC (Nygard et al., 2013); (Romer et al., 2013); (Smith et al., 2013). Furthermore, a newly published study including material from 78 mCRC patients also found frequent gains of the TOP1 gene and an increased TOP1/CEN20 ratio and a borderline significant association between the TOP1 gene copy numbers and objective response to second‐line irinotecan‐therapy (Nygard et al., 2014).
There are two main pathways described in CRC that are characterized by distinct patterns of genomic instability, the microsatellite instability (MSI) pathway comprising approximately 15% and the chromosomal instability (CIN) pathway comprising approximately 85% of sporadic CRC (Grady and Carethers, 2008). Defects in the mismatch repair (MMR) system are considered the main cause of sporadic CRC with microsatellite instability as well as the inherited Lynch syndrome. In the Lynch syndrome the MMR defect reflects germline mutations of the genes encoding MMR proteins and in sporadic CRC defective MMR is most often due to MLH1 promoter hypermethylation (Geiersbach and Samowitz, 2011). The MMR system is a posttranscriptional proofreading system that corrects base–base mismatches and insertion/deletion loops (IDL). Cells with deficient MMR (dMMR) fail to correct these errors during DNA replication resulting in base substitutions and frameshift mutations. The repetitive base sequences in DNA, the microsatellites, which are present in both gene coding regions and gene regulatory regions, are especially prone to mutations, and the longer the microsatellite the greater frequency of mutations. This pattern results in a mutator phenotype characterized by microsatellite instability (MSI) and near‐diploid DNA content (Pena‐Diaz and Jiricny, 2012). MSI is present in adenomas, indicating that MSI is an early event in the development of carcinoma and not only a result of genetic instability in carcinogenesis. The other CRC subgroup features proficient MMR (pMMR), and is characterized by chromosomal instability where large parts of chromosomes and/or whole chromosomes are amplified, deleted or rearranged. These cancer cells are aneuploid, and the CIN pattern becomes apparent usually around adenoma‐to‐carcinoma transition, in most cases associated with partial loss of chromosome 18 (Burrell et al., 2013). Although MSI and CIN are commonly conceived as being mutually exclusive, there are exceptions and several MSI CRC cases have been shown to harbour amplifications and deletions, but to a much lesser extent than in CIN CRC (Grady and Carethers, 2008). Besides having a distinct genomic instability phenotype, CRC with dMMR also has a better prognosis (Grady and Carethers, 2008). MSI positive CRC is a negative predictive marker of treatment with 5‐fluorouracil but not of treatment with irinotecan (Guastadisegni et al., 2010), (Des Guetz et al., 2009).
One previous study compared IHC expression of Top1 and Top2A proteins with MSI (dMMR) and found no correlation (Staley et al., 1999). As described above, pMMR is characterized by CIN with large parts of chromosomes and/or whole chromosomes being amplified, deleted or rearranged, implicating potential abnormalities in processes of DNA replication, DNA repair and chromosome segregation. As the topoisomerase enzymes are involved in solving topological problems during DNA replication and/or events just prior to chromosome segregation, and they create DNA breaks as part of DNA transitions including some forms of DNA repair (Wang, 2002), we hypothesized that an association may exist between TOP gene copy numbers and MMR proficiency/deficiency in CRC. In order to test our hypothesis we investigated TOP1 and TOP2A gene copy numbers in dMMR and pMMR colorectal cancer in a test cohort and validated the results in a larger cohort of patients with mCRC.
2. Material and methods
2.1. Patients and tumor tissue
2.1.1. Test cohort
For TOP1 and TOP2A FISH analysis formalin‐fixed, paraffin‐embedded (FFPE) tissue‐blocks containing tumor tissue from patients diagnosed with colorectal cancer at Herlev, Glostrup and Gentofte Hospitals in Denmark were included consecutively from January 1st 2007, until samples from 18 patients with deficient mismatch repair (dMMR) and 18 patients with proficient mismatch repair (pMMR) were obtained. Patient demographics are shown in Table 1. MMR status was established at the time of diagnosis by examination of IHC expression of MLH1 and MSH2, and this information was available for all patients. In order to obtain information on TOP1 gene copy number in unaffected mucosa cells we included 5 extra FFPE blocks from patients with pMMR tumors in addition to the 36 FFPE blocks.
Table 1.
Patient demographics: test cohort.
| All patients, N = 36 | |
|---|---|
| Age | |
| Median (range) | 73 (40–89) |
| N (%) | |
| MMR status | |
| pMMR | 18 (50) |
| dMMR | 18 (50) |
| Sex | |
| Female | 22 (61) |
| Male | 13 (36) |
| Unknown | 1 (3) |
| Primary site | |
| Right | 21 (58) |
| Left | 14 (39) |
| Unknown | 1 (3) |
| T‐stage | |
| ≤ 3 | 31 (86) |
| 4 | 4 (11) |
| Unknown | 1 (3) |
| N‐stage | |
| 0 | 23 (64) |
| ≥ 1 | 11 (31) |
| Unknown | 2 (5) |
| M‐stage | |
| 0 | 31 (86) |
| ≥1 | 4 (11) |
| Unknown | 1 (3) |
| TNM‐stage | |
| 1–2 | 21 (58) |
| 3–4 | 13 (36) |
| Unknown | 2 (6) |
| Differentiation | |
| High | 25 (69) |
| Low | 10 (28) |
| Unknown | 1 (3) |
(Primary site: Right: cecum, ascending colon, transverse colon. Left: descending colon, sigmoideum, rectum.).
2.1.2. Tissue‐micro‐array
Tissue micro arrays (TMAs) were constructed from each FFPE tissue block included for FISH analysis. Donor areas were selected based on hematoxylin and eosin (H&E) stained sections and corresponding cores were punched from the donor blocks, selected from the invasive front of the tumors and with the greatest representation of vital tumor tissue.
2.1.3. Validation cohort
A total of 154 FFPE tissue‐samples from patients with primary stage III CRC enrolled in the RANX05 randomized clinical study (Nielsen et al., 1998), were included to determine MMR status by IHC. This material was further described in a study by Rømer et al. 2013 (Romer et al., 2013) and information on TOP1 and TOP2A copy numbers was previously published for all 154 patients (Romer et al., 2013); (Nygard et al., 2013).
2.2. FISH
Test cohort: Bacterial artificial chromosome (BAC) clone CTD‐3193L13 (Invitrogen Ltd, Paisley, UK) and plasmid clone containing centromere chromosome 20 sequences pZ20 (CEN20, provided by Dr. M Rocchi, Resources for Molecular Cytogenetics, University of Bari, Italy) encompassing genes and loci of interest were cultured and DNA extracted as previously described (Poulsen, 2007a). The BAC clone CTD‐3193L13 was confirmed to contain all the exons of the TOP1 gene, by end‐sequencing the BAC clone as previously described (Poulsen and Johnsen, 2007b) followed by endonuclease‐digested using EcoRV as recommended by the enzyme manufacturer (Gibco, BRL) and exons were sequenced as described by Moisan et al. (Moisan et al., 2006) to verify that all 21 exons were present in BAC clone CTD‐3193L13. The clones were labeled with either Chromatide TexasRed–12‐dUTP (Life Technologies, Eugene, DR, USA), or (Fluorescein–12‐dUTP Roche diagnostics GmbH, Mannheim, Germany) by nick translation according to the manual supplied by the manufacturer (Roche diagnostics GmbH, Mannheim, Germany). A mixture of 100 ng TexasRed labeled BAC CTD‐3193L13 (TOP1) probe 60 ng Fluorescein labeled plasmid pZ20 (CEN20), and a 10‐fold excess of Cot‐1‐DNA (Roche diagnostics GmbH, Mannheim, Germany) was dissolved in the hybridization mixture (50% formamide, 300 mM NaCl, 30 mM sodium citrate (2× SSC), 10% dextran sulfate, and 50 mM sodium phosphate at pH 7). The FISH probes for TOP2A/CEN17 was commercial available (DAKO, Glostrup, Denmark). The FISH method was performed as described by the manufacturer of the Histology FISH kit (DAKO, Glostrup, Denmark). A fluorescence microscope (Olympus BX61) with a dual TexasRed/FITC filter was used for visualization of the signals. Signal counting was performed blinded to all patient data.
Validation cohort: FISH analyses for TOP1 (DAKO, Glostrup, Denmark) and for TOP2A (DAKO, Glostrup, Denmark) gene copy number analyses were previously described (Nygard et al., 2013 and Romer et al., 2013).
2.3. IHC
Four‐micrometer sections of FFPE tissue blocks were prepared for IHC by standard procedures. The primary monoclonal antibodies used were: MLH1 (clone: ES05, ready‐to‐use, Dako) MSH2 (clone: G219‐1129, diluted 1:400, Cellmarque) MSH6 (clone: EP49, diluted 1:50, Epitomics) and PMS2 (clone: EPR‐3947, 1:50 Dako). Slides were counterstained with Mayers hematoxylin. Normal colonic crypt epithelium adjacent to the tumor served as internal positive control for normal staining reactivity. Appendix tissue was used as external positive control and as negative control we used FLEX negative control (Dako) according to manufacturer's instruction. IHC for any of the four MMR proteins was considered positive, regardless of staining intensity, if nuclear staining was detected in >10% of tumor cells. Nuclear staining in less than 10% of tumor cells of any of the four MMR proteins classified the tumor as dMMR.
2.4. Statistical analyses
2.4.1. Test cohort
TOP1/TOP2A gene copy number and the reference probe signals CEN20/CEN17 per nucleus were calculated by dividing total number of signals counted in each tumor by number of counted nuclei. Signals were counted in as many nuclei as needed to obtain minimum 60 gene probe signals. This counting method has previously been validated statistically for HER2 FISH analysis (Olsen et al., 2004). Ratios were calculated by dividing gene copy numbers by number of reference probe signals. The Mann–Whitney test was used to compare the mean gene copy number, centromere counts and ratios to MMR status (dMMR/pMMR).
TOP1 gene copy number and CEN20 were counted in epithelial cells of unaffected colorectal mucosa and means were calculated. Because the nuclei in epithelial cells are often larger than the thickness of the tissue sections used in FISH, the nuclei are often truncated resulting in loss of some signals when counting (Wolman, 1994), (de Pender et al., 2001). The expected haploid number, n, is therefore less than 1. To estimate and apply boundaries to the counts in malignant cells we calculated theoretical ranges for normal/diploid/disomic (n = 2), triploid/trisomic (n = 3), and tetraploid/tetrasomic (n = 4) gene copy numbers. The haploid number, n, equals 0.5 times the mean in unaffected colon mucosa. The diploid/disomic range was then calculated as 2n ± 0.5*n. TOP2A and CEN17 were counted in unaffected colorectal mucosa in the validation cohort by Nygård et al. and has previously been published (Nygard et al., 2013) and theoretical ranges were calculated by the same method. This method of classifying gene and chromosome counts in FISH analysis into low, medium and high levels has previously been used in colorectal cancer and breast cancer and is considered a meaningful method in the absence of known “clinical” cut‐off values (Romer et al., 2013); (Nielsen et al., 2012).
The study was approved by the Copenhagen and Frederiksberg regional division of the Danish National Committee on biomedical Research Ethics (H‐3‐2009‐143) and conducted under the Helsinki declaration. The project was approved by the data protection agency (2009‐41‐4180).
2.4.2. Validation cohort
The 154 tumor tissue samples from the RANX05 study were classified as dMMR or pMMR according to their MMR protein expression profile as described in the Materials and Methods section. Data on cancer cell TOP1/CEN20 and TOP2A/CEN17 copy number and ratios in this patient cohort were previously published (Romer et al., 2013); (Nygard et al., 2013). The Mann–Whitney test was used to compare means of TOP1, CEN20 and TOP2A and CEN17 between the dMMR and pMMR subgroup.
3. Results
3.1. Test cohort
TOP1: Tumor tissue from one patient and normal mucosa from one patient were excluded from FISH analysis due to sparse tissue represented in the TMA or because of insufficient probe hybridization.
Tumor tissue from the 35 patients and normal mucosa from 22 patients were successfully evaluated with TOP1 and CEN20 FISH analysis. TOP1 mean (SD) in unaffected epithelia from 22 patients with pMMR tumors was 1.79 (SD = 0.13). As described above, we then calculated the theoretical normal/diploid/disomic range (n = 2) for TOP1 gene copy number as 1.34‐2.24. The theoretical triploid/trisomic range was then 2.24–3.14) and the tetraploid/tetrasomic range was 3.14–4.04. We calculated the same ranges for CEN20. The distribution of tumors according to the above definitions is listed in Tables 2 and 3.
Table 2.
Theoretical ranges for TOP1 and CEN20 copy numbers and distribution of tumors in test cohort (35 patients).
| Calculated copy number ranges: | Number of patients with copy numbers, within the range (%) counted in all tumors (n = 35): | Number of patients with copy numbers, within the range (%) counted in pMMR tumors (n = 18): | Number of patients with copy numbers, within the range (%)counted in dMMR tumors (n = 17): | |
|---|---|---|---|---|
| TOP1 : | ||||
| <1 gene copy (n) per nucleus: | <1.34 | 0 (0) | 0 (0) | 0 (0) |
| 2 gene copies (2n) per nucleus (diploid/normal): | 1.34–2.24 | 19 (54) | 2 (11) | 17 (100) |
| 3 gene copies (3n) per nucleus (triploid): | 2.24–3.14 | 5 (14) | 5 (28) | 0 (0) |
| 4 gene copies (4n) per nucleus (tetraploid): | 3.14–4.04 | 2 (6) | 2 (11) | 0 (0) |
| >4 gene copies (4n) per nucleus: | >4.04 | 9 (26) | 9 (50) | 0 (0) |
| CEN20: | ||||
| Haploid/monosomic: | <1.29 | 1 (3) | 0 (0) | 1 (6) |
| Diploid/Disomic: | 1.29–2.16 | 18 (51) | 2 (11) | 16 (94) |
| Triploid/Trisomic: | 2.16–3.03 | 8 (23) | 8 (44) | 0 (0) |
| Tetraploid/Tetrasomic: | 3.03–3.89 | 5 (14) | 5 (28) | 0 (0) |
| >Tetraploid/Tetrasomic: | >3.89 | 3 (9) | 3 (17) | 0 (0) |
Table 3.
Theoretical ranges for TOP1 and CEN20 and distribution of tumors in validation cohort (149 patients).
| Calculated copy number ranges: | Number of patients with copy numbers, within the range (%) counted in all tumors (n = 149): | Number of patients with copy numbers, within the range (%) counted in pMMR tumors (n = 137): | Number of patients with copy numbers, within the range (%) counted in dMMR tumors (n = 12): | |
|---|---|---|---|---|
| TOP1 : | ||||
| <1 gene copy (n) per nucleus: | <1.34 | 0 (0) | 0 (0) | 0 (0) |
| 2 gene copies (2n) per nucleus (diploid/normal): | 1.34–2.24 | 37 (25) | 25 (18) | 12 (100) |
| 3 gene copies (3n) per nucleus (triploid): | 2.24–3.14 | 42 (28) | 42 (31) | 0 (0) |
| 4 gene copies (4n) per nucleus (tetraploid): | 3.14–4.04 | 40 (27) | 40 (29) | 0 (0) |
| >4 gene copies (4n) per nucleus: | >4.04 | 30 (20) | 30 (22) | 0 (0) |
| CEN20: | ||||
| Haploid/monosomic: | <1.29 | 1 (1) | 1 (1) | 0 (0) |
| Diploid/Disomic: | 1.29–2.16 | 70 (47) | 59 (43) | 11 (92) |
| Triploid/Trisomic: | 2.16–3.03 | 62 (41) | 61 (44) | 1 (8) |
| Tetraploid/Tetrasomic: | 3.03–3.89 | 16 (11) | 16 (12) | 0 (0) |
| >Tetraploid/Tetrasomic: | >3.89 | 0 (0) | 0 (0) | 0 (0) |
The mean (SD) TOP1 gene copy number in the malignant epithelia was 2.85 (SD = 1.92), CEN20 was 2.27 (SD = 1.12) and the TOP1/CEN20 ratio was 1.19 (SD = 0.24). In the pMMR subgroup mean TOP1 gene copy number and CEN20 copy number was 3.84 (SD = 2.03) and 2.89 (SD = 1.13). In the dMMR subgroup TOP1 gene copy number and CEN20 copy number was 1.50 (SD = 0.12) and 1.43 (SD = 0.15). Five (28%) tumors in the pMMR subgroup had a TOP1 ratio >1.5, suggesting numerical gain of the gene region. No tumors had a ratio <0.8 or ≥2.0. Results for TOP1 are listed in Table 4.
Table 4.
Mean TOP1 and ratio in test and validation cohort.
| Mean TOP1 gene copy number (SD): (p < 0.0001) | Mean CEN20 counts (SD): | Mean TOP1 ratio (SD): (p = 0.003) | Number of tumors with ratio >1.5 (%) | Number of tumors with ratio≥2.0 | Numbers of tumors with ratio <0.8 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Test cohort: | ||||||||||
| pMMR: | 3.84 (2.03) | p < 0.01 | 2.89 (1.13) | p < 0.01 | 1.28 (0.26) | p < 0.01 | 5 (28) | 0 | 0 | |
| dMMR: | 1.50 (0.12) | 1.43 (0.15) | 1.07 (0.14) | 0 | 0 | 0 | ||||
| All: | 2.85 (1.92) | 2.27 (1.12) | 1.19 (0.24) | 5 (28) | 0 | 0 | ||||
| Validation cohort: | ||||||||||
| pMMR: | 3.31 (1.17) | p < 0.01 | 2.32 (0.59) | p < 0.01 | 1.43 (0.43) | p < 0.01 | 43 (31) | 15 (11) | 0 | |
| dMMR: | 1.66 (0.11) | 1.62 (0.09) | 1.03 (0.03) | 0 | 0 | 0 | ||||
| All: | 3.17 (1.21) | 2.28 (0.60) | 1.39 (0.42) | 43 (31) | 15 (11) | 0 | ||||
TOP2A: TMA with tumor tissue from one patient was excluded from FISH analysis due to insufficient probe hybridization.
Tumor tissue from 35 patients was successfully evaluated with TOP2A and CEN17 FISH analysis. Using the FISH signal counts in unaffected colorectal mucosa from the validation cohort (Nygard et al., 2013), we calculated the theoretical ranges for TOP2A and CEN17 in the same manner as outlined for TOP1 and CEN20. The distribution of tumors according to above definitions is listed in Tables 5 and 6.
Table 5.
Theoretical ranges for TOP2A and CEN17 and distribution of tumors in test cohort.
| Calculated copy number ranges: | Number of patients with copy numbers, within the range (%) counted in all tumors (n = 35): | Number of patients with copy numbers, within the range (%) counted in pMMR tumors (n = 17): | Number of patients with copy numbers, within the range (%) counted in dMMR tumors (n = 18): | |
|---|---|---|---|---|
| TOP2A: | ||||
| <1 gene copy (n) per nucleus: | <1.27 | 0 (0) | 0 (0) | 0 (0) |
| 2 gene copies (2n) per nucleus (diploid/normal): | 1.27–2.12 | 30 (86) | 12 (71) | 18 (100) |
| 3 gene copies (3n) per nucleus (triploid): | 2.12–2.97 | 4 (11) | 4 (23) | 0 (0) |
| 4 gene copies (4n) per nucleus (tetraploid): | 2.97–3.82 | 1 (3) | 1 (6) | 0 (0) |
| >4 gene copies (4n) per nucleus: | ≥3.82 | 0 (0) | 0 (0) | 0 (0) |
| CEN17: | ||||
| Haploid/monosomic: | <1.31 | 4 (11) | 1 (6) | 3 (17) |
| Diploid/Disomic: | 1.31–2.18 | 30 (86) | 15 (88) | 15 (83) |
| Triploid/Trisomic: | 2.18–3.05 | 1 (3) | 1 (6) | 0 (0) |
| Tetraploid/Tetrasomic: | 3.05–3.92 | 0 (0) | 0 (0) | 0 (0) |
| >Tetraploid/Tetrasomic: | ≥3.92 | 0 (0) | 0 (0) | 0 (0) |
Table 6.
Theoretical ranges for TOP2A and CEN17 and distribution of tumors in validation cohort.
| Calculated copy number ranges: | Number of patients with copy numbers, within the range (%) counted in all tumors (n = 149): | Number of patients with copy numbers, within the range (%) counted in pMMR tumors (n = 137): | Number of patients with copy numbers, within the range (%) counted in dMMR tumors (n = 12): | |
|---|---|---|---|---|
| TOP2A: | ||||
| <1 gene copy (n) per nucleus: | <1.27 | 3 (2) | 3 (2) | 0 (0) |
| 2 gene copies (2n) per nucleus (diploid/normal): | 1.27–2.12 | 75 (50) | 64 (47) | 11 (92) |
| 3 gene copies (3n) per nucleus (triploid): | 2.12–2.97 | 56 (38) | 55 (40) | 1 (8) |
| 4 gene copies (4n) per nucleus (tetraploid): | 2.97–3.82 | 11 (7) | 11 (8) | 0 (0) |
| >4 gene copies (4n) per nucleus: | ≥3.82 | 4 (3) | 4 (3) | 0 (0) |
| CEN17: | ||||
| Haploid/monosomic: | <1.31 | 2 (1) | 2 (1) | 0 (0) |
| Diploid/Disomic: | 1.31–2.18 | 127 (85) | 115 (84) | 12 (100) |
| Triploid/Trisomic: | 2.18–3.05 | 19 (13) | 19 (14) | 0 (0) |
| Tetraploid/Tetrasomic: | 3.05–3.92 | 1 (1) | 1 (1) | 0 (0) |
| >Tetraploid/Tetrasomic: | ≥3.92 | 0 (0) | 0 (0) | 0 (0) |
Mean (SD) for TOP2A gene copy number, CEN17 and TOP2A/CEN17 ratio in the malignant epithelia from the 35 patients were 1.75 (SD = 0.43), 1.62 (SD = 0.29) and 1.08 (SD = 0.16), respectively. In the dMMR subgroup the mean TOP2A gene copy number and CEN17 copy number was 1.52 (SD = 0.10) and 1.46 (SD = 0.13). In the pMMR subgroup the mean TOP2A gene copy number and CEN17 copy number was 1.99 (SD = 0.52) and 1.79 (SD = 0.32). Two tumors (12%) in the pMMR subgroup had a TOP2A ratio >1.5, suggesting numerical gain of the gene region. No tumors had a ratio <0.8 or ≥2.0. Results for TOP2A are listed in Table 7.
Table 7.
TOP2A, CEN17 and ratio in the test and validation cohort.
| Mean TOP2A gene copy number (SD): | Mean CEN17 counts (SD): | Mean TOP2A ratio (SD): | Number of tumors with a ratio >1.5 (%) | Number of tumors with a ratio ≥2.0 (%) | Number of tumors with a ratio <0.8 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Test cohort: | ||||||||||
| pMMR: | 1.99 (0.52) | p < 0.01 | 1.79 (0.32) | p < 0.01 | 1.12 (0.21) | p = 0.4 | 2 (12) | 0 | 0 | |
| dMMR: | 1.52 (0.10) | 1.46 (0.13) | 1.04 (0.07) | 0 | 0 | 0 | ||||
| All: | 1.75 (0.43) | 1.62 (0.29) | 1.08 (0.16) | 2 (12) | 0 | 0 | ||||
| Validation cohort: | ||||||||||
| pMMR: | 2.30 (1.36) | p < 0.01 | 1.90 (0.29) | p = 0.03 | 1.20 (0.51) | p = 0.03 | 17 (12) | 3 (2.2) | 2 (1.5) | |
| dMMR: | 1.80 (0.31) | 1.75 (0.19) | 1.03 (0.11) | 0 | 0 | 0 | ||||
| All: | 2.27 (1.32) | 1.89 (0.29) | 1.19 (0.50) | 17 (12) | 3 (2.2) | 2 (1.5) | ||||
3.2. Validation cohort
Of the 154 tumors, 149 were successfully evaluated with IHC. Four tumors were excluded due to inconclusive staining, and one because of missing tumor block. Eight % (12 of 149) were dMMR and 92% (137 of 149) were pMMR. The mean TOP1 copy number for dMMR vs pMMR was 1.66 (SD = 0.11) and 3.31 (SD = 1.17) (p < 0.01), respectively. The mean TOP2A copy number for dMMR versus pMMR was 1.80 (SD = 0.31) and 2.30 (SD = 1.36) (p < 0.01), respectively. 4, 7 presents means for TOP1, CEN20, TOP1/CEN20 ratios, TOP2A, CEN17 and TOP2A/CEN17 ratios and correlations to MMR subgroup. It was seen that not only were mean TOP1 and TOP2A gene copy numbers elevated in the pMMR subgroup vs the dMMR subgroup but also mean CEN20, 2.32 (SD = 0.59) vs 1.62 (SD = 0.09) (p < 0.01) and mean CEN17, 1.90 (SD = 0.29) vs 1.75 (SD = 0.19) (p = 0.03). However, amplification of the TOP2A gene was a rare event in the pMMR subgroup. The same applied to TOP1 ratio, 1.43 (SD = 0.43) compared to 1.03 (SD = 0.03) (p < 0.01), and TOP2A ratio, 1.20 (SD = 0.51) compared to 1.03 (SD = 0.11) (p = 0.03). In the pMMR subgroup, 17 tumors (12%) had a TOP2A ratio >1.5 and 3 tumors (2.2%) a ratio ≥2.0 suggesting numerical gain of the gene region and 2 tumors (1.5%) had a ratio <0.8 suggesting deletion of the gene region. Similarly 43 tumors (31%) in the pMMR subgroup had a TOP1 ratio >1.5 and 15 tumors (11%) a ratio ≥2.0, but no tumors showed deletion of the gene region. In the dMMR subgroup no tumors had TOP1/CEN20 or TOP2A/CEN17 ratios suggesting numerical gain/deletion of the gene region (4, 7).
Graphic overviews of the distributions of the gene copy numbers and the centromere copy numbers for the test cohort and validation cohorts are shown in Figures 1 and 2.
Figure 1.
Validation and test cohort: TOP1 gene copy number and CEN20 copy number for pMMR subgroup and dMMR subgroup. Area above the solid and the dashed line represents ratios >1.5 and >2.0, respectively. The area above the horizontal dashed line represents TOP1 copy numbers >4.
Figure 2.
Validation and test cohort: TOP2A gene copy number and CEN17 copy number for pMMR subgroup and dMMR subgroup. Area above the solid and the dashed line represents ratios >1.5 and >2.0 respectively. The area above the horizontal dashed line represents TOP2A copy numbers >4.
In the test cohort, all tumors in the dMMR subgroup had a normal/diploid/disomic TOP1 and TOP2A copy number. Almost the same applied for CEN20 and CEN17 where 94% (16) and 83% (15), respectively were diploid/disomic. Similarly there were ≤2 gene copies per nucleus of TOP1 and TOP2A in 100% (12) and 92% (11) of dMMR tumors respectively in the validation cohort. One tumor in the dMMR subgroup of the validation cohort had 1 extra TOP2A gene copy. Moreover, in the validation cohort, 92% (11) of the dMMR tumors had ≤ 2 copies of the CEN20 region and 100% (12) had ≤ 2 copies of the CEN17 region.
In contrast, in the pMMR subgroup, 89% (16) of tumors from the test cohort had gain of TOP1 and 61% had ≥2 extra TOP1 gene copies per nucleus. Also, 89% (16) of these tumors had gain of CEN20 and 45% (8) had high‐level gain (≥4 CEN20 copies). Similarly, in the validation cohort, 82% (112) of the pMMR tumors had TOP1 gain and 51% (70) ≥2 extra TOP1 gene copies. CEN20 showed gain in 56% (77) in the pMMR tumors of the validation cohort (Table 5). TOP2A gain was found, in the pMMR subgroup, in 29% (5) of tumors from the test cohort and in 51% (70) from the validation cohort and there was gain of CEN17 copy number in 6% (1) of the tumors in the test cohort and 15% (20) of the tumors in the validation cohort.
4. Discussion
We have investigated TOP1 and TOP2A gene copy numbers in FFPE tissue samples from a test cohort of 36 patients, 18 with dMMR CRC and 18 with pMMR CRC. Both TOP1 and TOP2A gene copy numbers were significantly elevated in the pMMR subgroup compared to the dMMR subgroup. The difference in mean value for TOP1 in the pMMR subgroup (3.84, SD: 2.03) versus the dMMR subgroup (1.50 SD: 0.12) (p < 0.01) was more pronounced than that for TOP2A (1.99 SD: 0.52 and 1.52 SD: 0.10) (p < 0.01) but the differences were significant for both topoisomerase genes.
As highlighted by the theoretical ranges applied to the TOP1/CEN20 and TOP2A/CEN17 counts we have shown that the majority of pMMR tumors in both cohorts had both gain of TOP1 and TOP2A copy number and, to a lesser extent, gain of the CEN20 and the CEN17 region. Gene/centromere ratios are often used as a marker of gene amplification/gain, but this parameter cannot always sufficiently describe the gene copy alterations in tumors. Larger chromosomal regions may be gained, that aside from the gene of interest, often include the centromere region as well, and gene gains can be overlooked when only using the gene/centromere ratio as an indicator of gene gain (Nielsen et al., 2012). When looking at the gene/centromere ratios in this study, it is apparent that only a smaller fraction of tumors had ratios above 1.5 or 2.0, whereas the majority of tumors had a ratio near 1. This contrasts with the large subsets of tumors with gene gain. This difference suggests that the corresponding centromere regions are involved in the extra copies of larger stretches of chromosome 20 and 17 rather than TOP genes being the subject of isolated amplification events. This finding in the pMMR subgroup, and not in the dMMR subgroup, is in agreement with the fact that pMMR tumors have been shown to have a CIN phenotype with aneuploid DNA characterized by large chromosomal aberrations including gains of whole chromosomes and chromosome arms (Lengauer et al., 1997). Studies of CRC cells from CRC tumors and cell lines with array‐comparative genomic hybridization (aCGH) have shown frequent DNA copy gains of chromosome 20q (including TOP1 and CEN20), and one study of 11 MSI and 11 CIN CRCs reported that amongst others this region on chromosome 20 is preferentially gained in CIN CRC (Boonsong et al., 2000); (Lassmann et al., 2007). The mechanism behind TOP1 gains in CRC was further analyzed in 9 CRC cell lines by investigating TOP1 gene copy number in combination with centromere‐2 (CEN‐2), the latter as a marker of ploidy level (Smith et al., 2013). The authors reported that TOP1 gain was attributed to 20q isochromosome formation and chromosome 20 aneusomy. While amplification of 20q was not observed in the 9 cell lines the same authors reported that in 10% of 154 FFPE CRC tumor samples (same cohort as the present validation cohort) the TOP1/CEN20 ratio exceeded 2 with a diploid CEN2 value indicating an isolated TOP1 gene amplification (Smith et al., 2013). These results imply that the majority of TOP1 gene aberrations are a result of CIN and are therefore predominantly found in the pMMR subgroup of CRC.
While the experimental results obtained in our present study show clear differences in terms of the preferential gains of TOP1 and/or TOP2A genes among the pMMR tumors compared to the dMMR subset, the key question is what biological mechanism(s) might account for such a preference. We believe that the mechanistic explanation may reflect the presence of a high degree of DNA replication stress selectively among the pMMR tumor subset. This view is based on the following facts. First, in the vast majority of human solid tumors including CRC, the DNA damage response is ‘spontaneously’ activated, mainly due to oncogene‐evoked DNA replication stress in tumor cells (Bartkova et al., 2005), (Halazonetis et al., 2008). Second, enhanced DNA replication stress was reported as occurring selectively in the CRC cell lines with the CIN (pMMR) phenotype, in contrast to low replication stress among the dMMR subset of CRCs (Burrell et al., 2013). Third, the topoisomerase enzymes are required for proper resolution of the DNA topological problems caused by diverse DNA intermediates and torsional stress, all events that are enhanced under conditions of replication stress (Bermejo et al., 2012). Taken together with our present results, we propose that the extra copies of TOP1/TOP2A genes provide a survival advantage for the pMMR tumors that feature high degree of replication stress. The gained copies of TOP1/TOP2A genes are unlikely to be just a ‘nonspecific byproduct’ of the overall CIN phenotype in pMMR tumors, since: i) There are also subtle amplification events that target the TOP1/TOP2A gene areas; ii) Among the large chromosomal regions gained in CIN tumors, at least the chromosome 20 q that harbours also the TOP1 gene is among the recurrent gains in CRC; and iii) Despite many chromosomal regions are commonly lost (rather than gained) in the pMMR CRCs, the chromosomal regions with the TOP1/TOP2A genes are gained, not lost. We believe that at least one reason for the observed gains of chromosomal regions harbouring the TOP1/TOP2A genes is an ‘addiction’ of the pMMR/CIN‐positive tumors to the pro‐survival, stress‐support function provided by the topoisomerases under conditions of enhanced replication stress that could otherwise grossly destabilize the genome and lead to mitotic catastrophe of such CRC cells.
CRC with dMMR is a distinct subgroup with different pathogenesis, clinico‐pathological characteristics and a better prognosis than pMMR patients (Grady and Carethers, 2008). At the molecular genetic level this subgroup has nearly diploid DNA content and we have shown in this study that there are no gene gains of the TOP1/TOP2A regions in this subgroup. We find this observation important to report because studies of predictive markers are often performed retrospectively on archival material instead of in the favourable setting of a randomised clinical trial. If we suppose that TOP1 or TOP2A gene gain are positive predictors of effect from treatment with irinotecan or anthracyclines, respectively, then the better prognosis of the dMMR subgroup could potentially obscure the beneficial effect of chemotherapy if not taking into account that this subgroup does not harbour TOP1 or TOP2A gene gains.
Last but not least, from a conceptual perspective, we speculate that targeting the topoisomerases in the pMMR subsets of CRCs that show TOP1/TOP2A gene copy gains might represent another example of targeting tumor addiction. In this scenario, an addiction to stress‐support enzymes, topoisomerases, that might help cancer cells with enhanced replication stress and chromosomal instability to maintain their genome integrity at least to a degree that allows their survival. Treatment with topoisomerase inhibitors under such conditions might tip the balance of the overall genomic instability beyond the survivable threshold (Bartek et al., 2012) thus causing cancer cell death while being less detrimental to normal cells that do not feature such a high degree of replication stress.
It should be emphasized, that the aim of the present study was not to test the predictive value of TOP FISH in relation to treatment with topoisomerase inhibitors, but future studies including relevant clinical tumor material are needed to define a clinically relevant cut‐off values in relation to drug efficacy.
5. Conclusion
We have shown in two independent patient cohorts, evaluated independently by different observers, that in CRC, TOP1 and TOP2A copy numbers are elevated in a large fraction of tumors with pMMR, whereas almost all tumors with dMMR have normal diploid gene copy number but further studies are needed to determine if this also applies in a larger data set. Since the gene/centromere ratios approximated one, the observed TOP1 and TOP2A gene aberrations most probably can be related to aberrations covering large stretches of the chromosomes as seen in CIN tumors. Using guidelines for HER2 FISH counting, a small subset of pMMR tumors could be classified as having true/high level amplification and future studies should test if these cases are the ones being responsive to Top1 or Top2a inhibitor therapy.
Additional studies are needed to validate these results, test our concept of pMMR CRC addiction to topoisomerases required to buffer the high replication stress, as well as to explore if these findings can be associated with differences in sensitivity to chemotherapeutic drugs.
Disclosure/conflicts of interest
The authors declare that they have no conflicts of interest.
Acknowledgements
The authors thank Ib Jarle Christensen and Tobias Wirenfeldt Klausen for statistical analysis.
Sønderstrup Ida Marie Heeholm, Nygård Sune Boris, Poulsen Tim Svenstrup, Linnemann Dorte, Stenvang Jan, Nielsen Hans Jørgen, Bartek Jiri, Brünner Nils, Nørgaard Peter, Riis Lene, (2015), Topoisomerase-1 and -2A gene copy numbers are elevated in mismatch repair-proficient colorectal cancers, Molecular Oncology, 9, doi: 10.1016/j.molonc.2015.02.009.
Contributor Information
Ida Marie Heeholm Sønderstrup, Email: idms@regionsjaelland.dk, Email: ida_marie83@hotmail.com.
Sune Boris Nygård, Email: whp661@alumni.ku.dk.
Tim Svenstrup Poulsen, Email: tim.svenstrup.poulsen@regionh.dk.
Dorte Linnemann, Email: dorte.linnemann@regionh.dk.
Jan Stenvang, Email: stenvang@sund.ku.dk.
Hans Jørgen Nielsen, Email: hans.joergen.nielsen@regionh.dk.
Jiri Bartek, Email: jb@cancer.dk.
Nils Brünner, Email: nbr@sund.ku.dk.
Peter Nørgaard, Email: peter.h.noergaard@regionh.dk.
Lene Riis, Email: lene.buhl.riis@regionh.dk.
References
- Bartek, J. , Mistrik, M. , Bartkova, J. , 2012. Thresholds of replication stress signaling in cancer development and treatment. Nat. Struct. Mol. Biol. 19, (1) 5–7. [DOI] [PubMed] [Google Scholar]
- Bartkova, J. , Horejsi, Z. , Koed, K. , Kramer, A. , Tort, F. , Zieger, K. , Guldberg, P. , Sehested, M. , Nesland, J.M. , Lukas, C. , Orntoft, T. , Lukas, J. , Bartek, J. , 2005. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature 434, (7035) 864–870. [DOI] [PubMed] [Google Scholar]
- Bermejo, R. , Lai, M.S. , Foiani, M. , 2012. Preventing replication stress to maintain genome stability: resolving conflicts between replication and transcription. Mol. Cell 45, (6) 710–718. [DOI] [PubMed] [Google Scholar]
- Boonsong, A. , Curran, S. , McKay, J.A. , Cassidy, J. , Murray, G.I. , McLeod, H.L. , 2002. Topoisomerase I protein expression in primary colorectal cancer and lymph node metastases. Hum. Pathol. 33, (11) 1114–1119. [DOI] [PubMed] [Google Scholar]
- Boonsong, A. , Marsh, S. , Rooney, P.H. , Stevenson, D.A. , Cassidy, J. , McLeod, H.L. , 2000. Characterization of the topoisomerase I locus in human colorectal cancer. Cancer Genet. Cytogenet. 121, (1) 56–60. [DOI] [PubMed] [Google Scholar]
- Braun, M.S. , Richman, S.D. , Quirke, P. , Daly, C. , Adlard, J.W. , Elliott, F. , Barrett, J.H. , Selby, P. , Meade, A.M. , Stephens, R.J. , Parmar, M.K. , Seymour, M.T. , 2008. Predictive biomarkers of chemotherapy efficacy in colorectal cancer: results from the UK MRC FOCUS trial. J. Clin. Oncol. 26, (16) 2690–2698. [DOI] [PubMed] [Google Scholar]
- Burrell, R.A. , McClelland, S.E. , Endesfelder, D. , Groth, P. , Weller, M.C. , Shaikh, N. , Domingo, E. , Kanu, N. , Dewhurst, S.M. , Gronroos, E. , Chew, S.K. , Rowan, A.J. , Schenk, A. , Sheffer, M. , Howell, M. , Kschischo, M. , Behrens, A. , Helleday, T. , Bartek, J. , Tomlinson, I.P. , Swanton, C. , 2013. Replication stress links structural and numerical cancer chromosomal instability. Nature 494, (7438) 492–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- de Pender, A.M. , Alers, J.C. , Vissers, K.J. , de Both, N.J. , Dinjens, W.N. , van Dekken, H. , 2001. Evaluation of oncogene amplification in intact and truncated cell nuclei of gastro-esophageal cancer cell lines by DNA in situ hybridisation. Acta Histochem. 103, (2) 127–138. [DOI] [PubMed] [Google Scholar]
- Des Guetz, G. , Uzzan, B. , Nicolas, P. , Schischmanoff, O. , Perret, G.Y. , Morere, J.F. , 2009. Microsatellite instability does not predict the efficacy of chemotherapy in metastatic colorectal cancer. A systematic review and meta-analysis. Anticancer Res. 29, (5) 1615–1620. [PubMed] [Google Scholar]
- Di Leo, A. , Desmedt, C. , Bartlett, J.M. , Piette, F. , Ejlertsen, B. , Pritchard, K.I. , Larsimont, D. , Poole, C. , Isola, J. , Earl, H. , Mouridsen, H. , O'Malley, F.P. , Cardoso, F. , Tanner, M. , Munro, A. , Twelves, C.J. , Sotiriou, C. , Shepherd, L. , Cameron, D. , Piccart, M.J. , Buyse, M. , 2011. HER2 and TOP2A as predictive markers for anthracycline-containing chemotherapy regimens as adjuvant treatment of breast cancer: a meta-analysis of individual patient data. Lancet Oncol. 12, (12) 1134–1142. [DOI] [PubMed] [Google Scholar]
- Du, Y. , Zhou, Q. , Yin, W. , Zhou, L. , Di, G. , Shen, Z. , Shao, Z. , Lu, J. , 2011. The role of topoisomerase II alpha in predicting sensitivity to anthracyclines in breast cancer patients: a meta-analysis of published literatures. Breast Cancer Res. Treat 129, (3) 839–848. [DOI] [PubMed] [Google Scholar]
- Geiersbach, K.B. , Samowitz, W.S. , 2011. Microsatellite instability and colorectal cancer. Arch. Pathol. Lab Med. 135, (10) 1269–1277. [DOI] [PubMed] [Google Scholar]
- Grady, W.M. , Carethers, J.M. , 2008. Genomic and epigenetic instability in colorectal cancer pathogenesis. Gastroenterology 135, (4) 1079–1099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guastadisegni, C. , Colafranceschi, M. , Ottini, L. , Dogliotti, E. , 2010. Microsatellite instability as a marker of prognosis and response to therapy: a meta-analysis of colorectal cancer survival data. Eur. J. Cancer 46, (15) 2788–2798. [DOI] [PubMed] [Google Scholar]
- Halazonetis, T.D. , Gorgoulis, V.G. , Bartek, J. , 2008. An oncogene-induced DNA damage model for cancer development. Science 319, (5868) 1352–1355. [DOI] [PubMed] [Google Scholar]
- Koopman, M. , Knijn, N. , Richman, S. , Seymour, M. , Quirke, P. , van Tinteren, H. , 2009. The correlation between Topoisomerase-I (Topo1) expression and outcome of treatment with capecitabine and irinotecan in advanced colorectal cancer (ACC) patients (pts) treated in the CAIRO study of the Dutch Colorectal Cancer Group (DCCG). Eur. J. Cancer 321 Suppl. 7 19114024 [Google Scholar]
- Kostopoulos, I. , Karavasilis, V. , Karina, M. , Bobos, M. , Xiros, N. , Pentheroudakis, G. , Kafiri, G. , Papakostas, P. , Vrettou, E. , Fountzilas, G. , 2009. Topoisomerase I but not thymidylate synthase is associated with improved outcome in patients with resected colorectal cancer treated with irinotecan containing adjuvant chemotherapy. BMC Cancer 9, 339 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lassmann, S. , Weis, R. , Makowiec, F. , Roth, J. , Danciu, M. , Hopt, U. , Werner, M. , 2007. Array CGH identifies distinct DNA copy number profiles of oncogenes and tumor suppressor genes in chromosomal- and microsatellite-unstable sporadic colorectal carcinomas. J. Mol. Med. (berl) 85, (3) 293–304. [DOI] [PubMed] [Google Scholar]
- Lengauer, C. , Kinzler, K.W. , Vogelstein, B. , 1997. Genetic instability in colorectal cancers. Nature 386, (6625) 623–627. [DOI] [PubMed] [Google Scholar]
- Michaelson, R. , Kemeny, N. , Young, C. , 1982. Phase II evaluation of 4'-epi-doxorubicin in patients with advanced colorectal carcinoma. Cancer Treat Rep. 66, (9) 1757–1758. [PubMed] [Google Scholar]
- Moisan, F. , Longy, M. , Robert, J. , Le Morvan, V. , 2006. Identification of gene polymorphisms of human DNA topoisomerase I in the National Cancer Institute panel of human tumour cell lines. Br. J. Cancer 95, (7) 906–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- NCCN, 2014. Guidelines Version 3 Colon Cancer National Comprehensive Cancer Network; Available at: http://www.nccn.org/professionals/physician_gls/f_guidelines.asp [Google Scholar]
- Nielsen, H.J. , McArdle, C. , Moesgaard, F. , The RANX05 Study Group1998. The effect of ranitidine on long-term survival in primary colorectal cancer. A 40 months interim analysis. GI Cancer 227–233. [Google Scholar]
- Nielsen, K.V. , Ejlertsen, B. , Moller, S. , Jensen, M.B. , Balslev, E. , Muller, S. , Knoop, A. , Mouridsen, H.T. , 2012. Lack of independent prognostic and predictive value of centromere 17 copy number changes in breast cancer patients with known HER2 and TOP2A status. Mol. Oncol. 6, (1) 88–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nygard, S.B. , Christensen, I.J. , Nielsen, S.L. , Nielsen, H.J. , Brunner, N. , Spindler, K.L. , 2014. Assessment of the topoisomerase I gene copy number as a predictive biomarker of objective response to irinotecan in metastatic colorectal cancer. Scand. J. Gastroenterol. 49, (1) 84–91. [DOI] [PubMed] [Google Scholar]
- Nygard, S.B. , Christensen, I.J. , Smith, D.H. , Nielsen, S.L. , Jensen, N.F. , Nielsen, H.J. , Vainer, B. , Brunner, N. , 2013. Underpinning the repurposing of anthracyclines towards colorectal cancer: assessment of topoisomerase II alpha gene copy number alterations in colorectal cancer. Scand. J. Gastroenterol. 48, (12) 1436–1443. [DOI] [PubMed] [Google Scholar]
- Olsen, K.E. , Knudsen, H. , Rasmussen, B.B. , Balslev, E. , Knoop, A. , Ejlertsen, B. , Nielsen, K.V. , Schonau, A. , Overgaard, J. , 2004. Amplification of HER2 and TOP2A and deletion of TOP2A genes in breast cancer investigated by new FISH probes. Acta Oncol. 43, (1) 35–42. [DOI] [PubMed] [Google Scholar]
- Paradiso, A. , Xu, J. , Mangia, A. , Chiriatti, A. , Simone, G. , Zito, A. , Montemurro, S. , Giuliani, F. , Maiello, E. , Colucci, G. , 2004. Topoisomerase-I, thymidylate synthase primary tumour expression and clinical efficacy of 5-FU/CPT-11 chemotherapy in advanced colorectal cancer patients. Int. J. Cancer 111, (2) 252–258. [DOI] [PubMed] [Google Scholar]
- Pena-Diaz, J. , Jiricny, J. , 2012. Mammalian mismatch repair: error-free or error-prone?. Trends Biochem. Sci. 37, (5) 206–214. [DOI] [PubMed] [Google Scholar]
- Pommier, Y. , Leo, E. , Zhang, H. , Marchand, C. , 2010. DNA topoisomerases and their poisoning by anticancer and antibacterial drugs. Chem. Biol. 17, (5) 421–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Poulsen, T.S. , 2007. Purification of BAC DNA. Bacterial Artificial Chromosomes Volume 1: Library Construction, Physical Mapping, and Sequencing. vol. 255, 91–100. [Google Scholar]
- Poulsen, T.S. , Johnsen, H.E. , 2007. BAC End Sequencing. Bacterial Artificial Chromosomes Volume 1: Library Construction, Physical Mapping, and Sequencing. vol. 255, 157–162. [Google Scholar]
- Romer, M.U. , Nygard, S.B. , Christensen, I.J. , Nielsen, S.L. , Nielsen, K.V. , Muller, S. , Smith, D.H. , Vainer, B. , Nielsen, H.J. , Brunner, N. , 2013. Topoisomerase 1(TOP1) gene copy number in stage III colorectal cancer patients and its relation to prognosis. Mol. Oncol. 7, (1) 101–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith, D.H. , Christensen, I.J. , Jensen, N.F. , Markussen, B. , Romer, M.U. , Nygard, S.B. , Muller, S. , Nielsen, H.J. , Brunner, N. , Nielsen, K.V. , 2013. Mechanisms of topoisomerase I (TOP1) gene copy number increase in a stage III colorectal cancer patient cohort. PLoS One 8, (4) e60613 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Staley, B.E. , Samowitz, W.S. , Bronstein, I.B. , Holden, J.A. , 1999. Expression of DNA topoisomerase I and DNA topoisomerase II-alpha in carcinoma of the colon. Mod. Pathol. 12, (4) 356–361. [PubMed] [Google Scholar]
- Tsavaris, N. , Lazaris, A. , Kosmas, C. , Gouveris, P. , Kavantzas, N. , Kopterides, P. , Papathomas, T. , Agrogiannis, G. , Zorzos, H. , Kyriakou, V. , Patsouris, E. , 2009. Topoisomerase I and IIalpha protein expression in primary colorectal cancer and recurrences following 5-fluorouracil-based adjuvant chemotherapy. Cancer Chemother. Pharmacol. 64, (2) 391–398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang, J.C. , 2002. Cellular roles of DNA topoisomerases: a molecular perspective. Nat. Rev. Mol. Cell Biol 3, (6) 430–440. [DOI] [PubMed] [Google Scholar]
- Wils, J.A. , 1984. Phase II trial of 4'-epi-doxorubicin in metastatic colorectal carcinoma. Invest New Drugs 2, (4) 397–399. [DOI] [PubMed] [Google Scholar]
- Wolman, S.R. , 1994. Fluorescence in situ hybridization: a new tool for the pathologist. Hum. Pathol. 25, (6) 586–590. [DOI] [PubMed] [Google Scholar]