Towards personalised treatment of patients with colorectal liver metastases Hof, Joost
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Hof, J. (2019). Towards personalised treatment of patients with colorectal liver metastases. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Loss of chromosome 22 predicts a poor survival after surgery for colorectal liver metastases
J. Hof1,2, K.P. de Jong1, R.H. Sijmons2, K. Kok2
1 Department of Hepato-pancreato-biliary surgery and Liver transplantation, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.
2 Department of Genetics, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.
Manuscript in preparation
Abstract
Aim: Colorectal liver metastases (CRLM) show heterogeneous behaviour, with survival rates after surgery ranging from 14-60%. This large range can partially be explained by clinicopathological characteristics, but these are currently not accurate enough for prognostication of individual patients. While copy number variations (CNVs) are described as an enabling characteristic for the established hallmarks of cancer, they are rarely studied in CRLM. We aimed to find CNVs associated with patient survival after surgery for CRLM and explore their relation to downstream mRNA expression profiles.
Methods: CNVs of resected CRLM were analysed by comparative hybridization arrays.
Two patient groups at the extremes of survival were studied: poor survivors (death from recurrent disease <30 months after surgery) and good survivors (no recurrent disease >60 months after surgery). CNV data was related to mRNA expression profiles from earlier studies.
Results: Poor survivors had a higher percentage of their genome changed compared to good survivors (47.3% vs. 36.6%, p=0.034). In multivariable analysis corrected for clinicopathological characteristics, loss of chr22 was the only independent prognostic factor associated with poor patient survival (OR: 10.9, 95%CI: 1.8-64.7, p= 0.009). CRLM with loss of chr22 had higher mRNA expression of genes related to the immune system.
Conclusion: Loss of chromosome 22 is an independent predictor of poor survival after surgery for CRLM and could potentially serve as a novel prognostic biomarker.
Introduction
Colorectal carcinoma (CRC) is the third most frequent cancer in Europe [1], and 20-25% of CRC patients have liver metastases (CRLM) at the time of diagnosis [2–4]. Thirty percent of patients are suitable for curative surgery which can achieve five-year survival rates of 15-60% [5]. Clinicopathological characteristics can partly explain this large range in survival, but are not currently accurate enough for prognostication of individual patients.
Genomic instability is an important feature of carcinogenesis and is described as an enabling characteristic for the established hallmarks of cancer. Genomic instability can induce tumour progression, invasion and development of metastases [6]. In colorectal cancer, DNA copy number variation (CNV) often occurs at the same chromosomal regions, for example at the location of the c-Myc oncogene [7]. The concordance in CNVs between the primary CRC and its liver metastases is high, indicating that CNVs might be an early event in carcinogenesis [8,9].
Prognostication using CNVs has not been widely studied in CRLM. The few studies that describe CNVs in CRLM do not link these copy number characteristics to a downstream biological effect [8,10,11]. In this study we used comparative genomic hybridization (CGH) arrays to detect CNVs in two groups of patients: poor survivors (those who died within 30 months after surgery because of recurrences) and good survivors (alive without recurrences 60 months or more after surgery). Additionally, we have mRNA expression data available which we use to identify downstream biological effects of copy number alterations.
Methods
Patient samples
Patients were selected from a prospectively maintained database. Inclusion criteria were (1) R0 partial liver resection for CRLM, (2) no neoadjuvant chemotherapy, (3) a Fong clinical risk score [12] of 3 or lower, (4) no detectable extrahepatic disease at time of surgery, (5) no other known malignant disease, and (6) availability of fresh frozen (-80°C) resected tumour tissue. Follow-up consisted of cross-sectional imaging and measurement of CEA serum levels every 3-4 months in the 2 years after liver resection and at 6 monthly intervals afterwards up to 5 years. We included samples from two groups of patients with extreme survival rates: poor survivors, who died of recurrent disease within 30 months after partial liver resection, and good survivors, who showed no evidence of recurrent disease up to
5
60 months after liver resection. Tumour tissue samples were reviewed by an experienced hepatopathologist by microscopic inspection of HE-stained frozen tissue slides to judge the quality of the tissue and the quantity of tumour cells.
RNA/DNA isolation
Visual inspection the HE-stained slides was performed to determine whether macrodissection of tumour sections was necessary. Both genomic DNA and total RNA were isolated from 10 μm-tissue sections (RNA/DNA purification kit, Norgen Biotek Corporation, Thorold, Ontario, Canada). DNA/RNA isolation was performed according to the manufacturer’s protocol. Quality check and RNA quantification of samples was carried out by capillary electrophoresis using the LabChip GX (Perkin Elmer, Waltham, Massachusetts, USA).
CGH array
For CGH arrays, 500 ng of genomic DNA was used as starting material for both the patients (Cy5 labelled) and reference (Cy3 labelled) using the Complete Genomic SureTag DNA Enzymatic Labelling Kit protocol (Agilent, Santa Clara, CA, USA). The reference DNA was a mixture of DNA from 20 randomly selected healthy males. The hybridization was carried out according to the manufacturer’s instructions of the OligoaCGH/ChIP-on-Chip Hybridization kit (Agilent). Slides were scanned on the Agilent DNA Microarray Scanner using a 16 bit, 2 colour protocol. Thirty samples were processed on an array with design ID
‘180K_UMCG_V2_027730’ and seven samples on an array with design ID ‘021924’. Feature extraction software was used to digitalize the fluorescent signals. The corresponding grid files were ‘180K_UMCG_V2_027730_D_F_20111219’ and ‘021924_D_F_20150623’. Data analysis was performed using Nexus 7.5 (BioDiscovery Inc., El Segundo, California, USA).
The CNV plots were normalised by setting the copy number with the highest abundance to 2Log=0. Cut-off values for chromosomal gains and losses were set per tumour, and the percentage of CNVs was calculated.
mRNA sequencing and data-analysis
These methods are extensively described in chapter 3 of this thesis. Briefly, sequence libraries were generated using the Quantseq 3’ mRNA sample preparation kit (Lexogen, Vienna, Austria). The sequencing was performed on an Illumina HiSeq2500 applying a 50bp single-read protocol. Reads in genes with multiple transcript termination sites were summed up, resulting in a read count per gene.
Microsatellite instability and mutational hotspots
These methods are extensively described in chapter 3 of this thesis. Briefly, PCR products of five polymorphic mononucleotide loci (NR21, NR24, BAT25, BAT26, MONO27) were analysed on the ABI 3730xl DNA Analyzer (Thermo Fisher). Somatic driver mutations in KRAS (codon 12 and 13) and BRAF V600E were analysed by Sanger sequencing (supplementary table 1).
Statistical analysis
Summary statistics were obtained using established methods and presented as percentages, median (interquartile range, IQR) or mean (± standard deviation, SD).
Variables associated with survival were tested by univariable and multivariable binary logistic regression analyses. Variables with a p-value <0.1 in univariable analysis were entered into the multivariable model. Odds ratio (OR) and 95% confidence interval (CI) were estimated, and a p-value <0.05 was considered significant. Statistical analyses were carried out with IBM SPSS Statistics V22 (IBM, Armonk, New York, USA).
Results
Demographics
Table 1 shows the clinicopathological characteristics of all 37 patients stratified by survival.
The mean follow-up was 16.5 months (± 5.2 months) in the poor survivors and 112.6 months (± 37.0 months) in the good survivors. A high clinical risk score (CRS) tended to be associated with the poor survivor group (p=0.057), while the individual factors that are used to calculate the CRS were not associated with survival (all p > 0.1). A larger tumour size also tended to be associated with a poor survival (p=0.088). Of note, patients with poor survival died of recurrent disease in the abdomen (n=5), liver (n=5), or at multiple sites (n=8) with the lungs most often being involved.
5
Table 1 Clinicopathological characteristics
Poor survival
(<30 months, n=18) Good survival
(>60 months, n=19) P value Patient characteristics
Mean age at liver surgery (+SD) 63.8 ± 10.4 61.3 ± 7.2 0.398
Male sex 11 (61.1%) 9 (47.4%) 0.402
Tumour characteristics
Major liver surgery (≥ 3 segments) 15 (83.3%) 11 (57.9%) 0.091
Neoadjuvant chemotherapy 0 0 -
Size largest CRLM (in cm) + IQR 4.3 (3.9 - 11.8) 3.5 (3.0 - 5.1) 0.088
Rectal primary tumour 8 (44.4%) 5 (26.3%) 0.248
Clinical risk score
CRS ≥ 3 8 (44.4%) 3 (15.8%) 0.057
Interval CRLM <12 months 8 (44.4%) 11 (57.9%) 0.413
CEA >200 mg/μl 5 (31.5%) 2 (10.5%) 0.127
More than 1 CRLM 5 (27.8%) 3 (15.8%) 0.376
CRLM larger than 5cm 13 (44.4%) 9 (31.6%) 0.420
N+ primary tumour 13 (72.2%) 9 (47.4%) 0.124
Molecular characteristics
Microsatellite instable (MSI-high) 1 (5.6%) 1 (5.3%) 1.000
KRAS codon 12/13 mutation 7 (38.9%) 8 (42.1%) 0.842
BRAF V600E mutation 0 0 -
SD = standard deviation, IQR = interquartile range, CRS = clinical risk score, CRLM = colorectal liver metastases, CEA = carcinoembryonic antigen, N+ = lymph node positive
CNV by CGH arrays
Figure 1 shows an overview of the CNV plots of all 37 patients. The most prevalent chromosomal losses, which are present in more than 50% of the tumours, were detected for chr8p, chr17p, chr15 and chr18. Similarly, chromosomal gains were seen in more than 50% of the tumours for chr8q, chr20q, chr7 and chr13. Poor survivors more often had CNVs, with a higher percentage of the genome changed (p=0.034). Poor survivors had 47.3% (± 12.3%) of their genome changed while good survivors had 36.6% (± 16.7%) of their genome changed. A complete loss of chr10 tended to be associated with a poor survival (8/18 vs. 3/19 in good survivors; p=0.057).
Chromosome 22
Loss of chr22 was associated with poor patient survival (13/18 vs. 4/19 in good survivors;
p=0.002). Of note, the deletion only encompassed 18mb of 22qter in one patients’ tumour
Figure 1 Cumulative copy number plots of all 37 patients. X-axis displays the positions of all probes across chromosomes 1-22, X and Y. Y-axis displays the percentage of patients with tumours who had a chromosomal gain (blue) or loss (red), respectively. Patients are stratified by survival.
5
from the good survivor group. This tumour was also classified as having a loss of chr22 (supplementary figure 1). All other tumours had either a complete loss, a gain or a stable chr22. We performed a binary logistic regression analysis and, by setting a threshold of p
<0.1 in univariable testing, three variables were carried over to the multivariable analysis (table 2). These factors were ‘loss of chr22’, ‘loss of chr10’ and ‘CRS’. We did not include the size of the CRLM in the multivariable analysis because this variable is related to a low or high CRS (3.8 cm (3.0-4.3) vs. 11.0 cm (5.5-14.0), respectively; p<0.001), which means that including both might cause overfitting. Similarly, we also did not include the percentage of genome changed because it is related to loss of chr22 (47.9% (39.2-58.2) vs. no loss of chr22: 39.8% (24.6-47.9); p= 0.019) and loss of chr10 (55.2% (46.2-58.8) vs. no loss of chr10: 39.8% (32.0-45.9); p=0.003). The multivariable analysis showed that loss of chr22 was the only prognostic factor (OR 10.893 (95%CI 1.833-64.734); p=0.009). We therefore conclude that loss of chr22 was independently associated with poor survival after surgery for CRLM in our patients.
Correlation of chr22 copy number status with mRNA expression profiles
We integrated mRNA sequencing data to study the biological consequences of chr22 loss in CRLM. mRNA expression data were available for 34 out of 37 patients. We performed a differential expression analysis by DESeq2 comparing the mRNA expression data of tumours with loss of chr22 vs. tumours without loss of chr22. We found 204 genes that were differentially expressed between the two groups with a p-value <0.01. For 22 of these genes, the FDR was <0.1 (supplementary table 2). Twenty-two out of the 204 (10.8%) differentially expressed genes (DEGs) were located on chr22, and we saw lower expression of all 22 DEGs in tumours with a loss of chr22. Of note, 180 of all 8931 genes studied (2.0%) were located on chr22, significantly lower than the number of DEGs on chr22 (180/8931 vs. 22/204, respectively; p<0.001). This suggests that a chromosomal loss of chr22 might result in lower mRNA expression of genes on chr22. Pathway analysis by DAVID EASE was performed using all 204 genes with a p-value <0.01. Table 3 displays the only significantly enriched pathway, showing that immune-related gene expression is associated with loss of chr22. In general, lower expression of immune-related genes was associated with loss of chr22.
Table 2 Multivariable analysis to predict survival
Univariable Multivariable
P-value OR (95% CI) P-value
Patient characteristics
Mean age at liver surgery 0.389
Male sex 0.403
Tumour characteristics
Size largest CRLM 0.053
Rectal primary tumour 0.252
Clinical risk score
CRS ≥ 3 0.066 7.866 (0.990-62.474) 0.051
Interval CRLM <12 months 0.415
CEA >200 mg/μl 0.143
More than 1 CRLM 0.381
CRLM larger than 5cm 0.422
N+ primary tumour 0.129
Molecular characteristics
Percentage of genome changed 0.049
Loss chromosome 22 0.003 10.893 (1.833-64.734) 0.009
Loss chromosome 10 0.066 5.711 (0.742-43.940) 0.094
Microsatellite instable (MSI-high) 0.969
KRAS codon 12/13 mutation 0.842
CRS = clinical risk score, CRLM = colorectal liver metastases, CEA = carcinoembryonic antigen, N+ = lymph node positive.
Table 3 Pathway analysis by DAVID EASE
Biological pathway EASE score FWER
defence/immunity protein activity 4.40e-05 2.91e-02
Genes in biological pathway
Genes Proposed immunological function
CCL19 Cytokine in lymphocyte recirculation and homing
DEFA5, DEFA6,DEFB1 Microbicidal and cytotoxic peptides made by neutrophils
IFI16 Cytokines of HIN-200 family
IFI44 Antiviral response
IGHA1, IGHG3, IGHG4, IGJ, IGKC Immunoglobulines
Pathways are a selection of the enriched pathways based on the 204 genes with a p-value < 0.01. Proposed immunological functions were extracted from www.genecards.org, except for IFI44[13].
EASE score = upper bound of the distribution of Jackknife Fisher exact probabilities given the enriched genes compared to the reference genes. FWER = EASE score adjusted for multiple testing correction by the Bonferroni method. HIN = hematopoietic interferon-inducible nuclear antigens
5
Discussion
This study presents the novel finding that loss of chr22 is an independent prognostic factor after surgery for CRLM. Moreover, loss of chr22 was associated with a lower immune- related mRNA expression. Although loss of chr22 is not an uncommon feature in oncology, the biological consequence of this phenomenon has not been studied extensively [14,15]. While loss of chr22 is described in both primary and metastatic CRC, we are the first to show an association with patient survival and possible downstream biological effects [16–18]. Loss of heterozygosity of individual genes on chr22 has previously been described in association with tumour progression and the development of metastases, with neurofibromin 2 (NF2) being the most prominent gene. This tumour-suppressor gene translates into a scaffolding protein that regulates multiple signalling pathways, which is involved in multiple types of malignancies [19]. NF2 is located on 22q12.2, which means that the CRLM of the good survivor with a deletion that encompassed 18mb of 22qter had a normal copy number state of NF2.
Both in our study and in other studies [8,10,11], characteristic patterns of chromosomal gains and losses have been seen in CRLM (i.e. gain of chr7, 8q, chr13, and chr20 and loss of chr1, chr4, 8p, and chr18). Poor survivors had a higher percentage of the genome altered in our study. Two other studies have also reported an association between patient survival and the fraction of the total genome that underwent a copy number change. Sveen et al. reported a more extensive copy number change in poor survivors based on SNPchip data [8]. In contrast, Mehta et al. showed that a higher percentage of the genome was changed in patients with a favourable survival [10] using a low resolution BAC-arrayCGH.
The contrasting results of Mehta et al. [10] with our study and that of Sveen et al. might be caused by the different cut-off values or by the different experimental method. While we analysed the percentage of genome changed as a continuous variable, Sveen et al. used a cut-off value of 25% [8] and Mehta et al. used a cut-off value of 20% [10]. In addition, we compared poor and good survivors, while Mehta et al. did not study patients at the extremes of survival [10]. Another group also performed CGH arrays and reported an association between ploidy levels of chr4 and patient survival. In their relatively small cohort of 20 patients, patients with one copy of chr4 had a lower recurrence rate and a favourable long-term survival compared to patients with two copies of chr4 [11]. We did not observe an association between chr4 and patient survival.
In conclusion, we show that a loss of chr22 was associated with a poor survival in CRLM and that this loss was also associated with a lower expression of genes from immune- related pathways. Further research is warranted to study whether this association also implies a causal relation.
Acknowledgements
We thank Kate McIntyre for assistance in the editorial process.
5
References
[1] Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JWW, Comber H, et al. Cancer incidence and mortality patterns in Europe: Estimates for 40 countries in 2012. Eur J Cancer 2013;49:1374–403. doi:10.1016/j.ejca.2012.12.027.
[2] Leporrier J, Maurel J, Chiche L, Bara S, Segol P, Launoy G. A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer. Br J Surg 2006;93:465–74.
doi:10.1002/bjs.5278.
[3] Jessup JM, McGinnis LS, Steele GD, Menck HR, Winchester DP. The National Cancer Data Base report on colon cancer. Cancer 1996;78:918–26. doi:10.1002/(SICI)1097-0142(19960815)78:4<918::AID- CNCR32>3.0.CO;2-W.
[4] Manfredi S, Lepage C, Hatem C, Coatmeur O, Faivre J, Bouvier A-M. Epidemiology and management of liver metastases from colorectal cancer. Ann Surg 2006;244:254–9. doi:10.1097/01.
sla.0000217629.94941.cf.
[5] Simmonds PC, Primrose JN, Colquitt JL, Garden OJ, Poston GJ, Rees M. Surgical resection of hepatic metastases from colorectal cancer: A systematic review of published studies. Br J Cancer 2006;94:982–
99. doi:10.1038/sj.bjc.6603033.
[6] Hanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell 2011;144:646–74.
doi:10.1016/j.cell.2011.02.013.
[7] Bigagli E, De Filippo C, Castagnini C, Toti S, Acquadro F, Giudici F, et al. DNA copy number alterations, gene expression changes and disease-free survival in patients with colorectal cancer: a 10 year follow- up. Cell Oncol 2016;39:545–58. doi:10.1007/s13402-016-0299-z.
[8] Sveen A, Løes IM, Alagaratnam S, Nilsen G, Høland M, Lingjærde OC, et al. Intra-patient Inter-metastatic Genetic Heterogeneity in Colorectal Cancer as a Key Determinant of Survival after Curative Liver Resection. PLoS Genet 2016;12:e1006225. doi:10.1371/journal.pgen.1006225.
[9] Mekenkamp LJM, Haan JC, Israeli D, van Essen HFB, Dijkstra JR, van Cleef P, et al. Chromosomal Copy Number Aberrations in Colorectal Metastases Resemble Their Primary Counterparts and Differences Are Typically Non-Recurrent. PLoS One 2014;9:e86833. doi:10.1371/journal.pone.0086833.
[10] Mehta KR, Nakao K, Zuraek MB, Ruan DT, Bergsland EK, Venook AP, et al. Fractional genomic alteration detected by array-based comparative genomic hybridization independently predicts survival after hepatic resection for metastatic colorectal cancer. Clin Cancer Res 2005;11:1791–7. doi:10.1158/1078- 0432.CCR-04-1418.
[11] Aust N, Schüle S, Altendorf-Hofmann AK, Chen Y, Knösel T, Dirsch O, et al. Loss of chromosome 4 correlates with better long-term survival and lower relapse rate after R0-resection of colorectal liver metastases. J Cancer Res Clin Oncol 2013;139:1861–7. doi:10.1007/s00432-013-1505-2.
[12] Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999;230:309- 318; discussion 318-321. doi:10.1097/00000658-199909000-00004.
[13] Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature 2011;472:481–5. doi:10.1038/
nature09907.
[14] Benetkiewicz M, Piotrowski A, Díaz De Ståhl T, Jankowski M, Bala D, Hoffman J, et al. Chromosome 22 array-CGH profiling of breast cancer delimited minimal common regions of genomic imbalances and revealed frequent intra-tumoral genetic heterogeneity. Int J Oncol 2006;29:935–45.
[15] Benetkiewicz M, Wang Y, Schaner M, Wang P, Mantripragada KK, Buckley PG, et al. High-resolution gene copy number and expression profiling of human chromosome 22 in ovarian carcinomas. Genes, Chromosom Cancer 2005;42:228–37. doi:10.1002/gcc.20128.
[16] González-González M, Fontanillo C, Abad MM, Gutiérrez ML, Mota I, Bengoechea O, et al. Identification of a characteristic copy number alteration profile by high-resolution single nucleotide polymorphism arrays associated with metastatic sporadic colorectal cancer. Cancer 2014;120:1948–59. doi:10.1002/
cncr.28681.
[17] Sayagués JM, Abad M del M, Melchor HB, Gutiérrez ML, González-González M, Jensen E, et al.
Intratumoural cytogenetic heterogeneity of sporadic colorectal carcinomas suggests several pathways to liver metastasis. J Pathol 2010;221:308‒19. doi:10.1002/path.2712.
[18] Zhou C-Z, Peng Z-H, Zhang F, Qiu G-Q, He L. Loss of heterozygosity on long arm of chromosome 22 in sporadic colorectal carcinoma. World J Gastroenterol 2002;8:668–73. doi:10.3748/WJG.V8.I4.668.
[19] Petrilli AM, Fernández-Valle C. Role of Merlin/NF2 inactivation in tumor biology. Oncogene 2016;35:537–
48. doi:10.1038/onc.2015.125.
5
Supplementary table 1. Primer specifics
Forward primer Reverse primer
Microsatellite
NR21 TAAATGTATGTCTCCCCTGG ATTCCTACTCCGCATTCACA
NR24 CCATTGCTGAATTTTACCTC ATTGTGCCATTGCATTCCAA
MONO27 CACTCCAGCGTGGGAGACAG GGTGGATCAAATTTCACTTGG
BAT25 TCGCCTCCAAGAATGTAAGT TCTGCATTTTAACTATGGCTC
BAT26 TGACTACTTTTGACTTCAGCC TAACCATTCAACATTTTTAACCC
Mutation hotspot
KRAS codon 12 and 13 CGATACACGTCTGCAGTCAA GAATGGTCCTGCACCAGTAA
BRAF V600E ATAATGCTTGCTCTGATAGG TGTGAATACTGGGAACTATG
Supplementary table 2. Differential expression analysis loss chr22
Ensembl code Gene symbol Chromosome Log2 fold change P-value FDR
ENSG00000196188 CTSE 1 -3.61 5.42E-06 0.048392
ENSG00000139515 PDX1 13 1.51 1.61E-05 0.054398
ENSG00000164816 DEFA5 8 -3.61 3.24E-05 0.054398
ENSG00000104327 CALB1 8 -4.74 3.52E-05 0.054398
ENSG00000122033 MTIF3 13 0.72 3.81E-05 0.054398
ENSG00000152268 SPON1 11 -2.08 4.60E-05 0.054398
ENSG00000143222 UFC1 1 0.53 5.01E-05 0.054398
ENSG00000187742 SECISBP2 9 0.49 6.10E-05 0.054398
ENSG00000109255 NMU 4 -2.17 6.56E-05 0.054398
ENSG00000140092 FBLN5 14 -1.38 6.65E-05 0.054398
ENSG00000100425 BRD1 22 -0.62 6.70E-05 0.054398
ENSG00000131016 AKAP12 6 -1.69 7.40E-05 0.05504
ENSG00000211892 IGHG4 14 -3.23 8.87E-05 0.060908
ENSG00000130675 MNX1 7 0.98 0.000105 0.062513
ENSG00000137673 MMP7 11 -2.05 0.000105 0.062513
ENSG00000234127 TRIM26 6 0.62 0.000164 0.083769
ENSG00000170537 TMC7 16 1.16 0.000165 0.083769
ENSG00000100220 RTCB 22 -0.49 0.000169 0.083769
ENSG00000106100 NOD1 7 0.63 0.000196 0.092235
ENSG00000090920 FCGBP 19 -2.26 0.000216 0.096264
ENSG00000262074 SNORD3B-2 17 -2.31 0.000238 0.097026
ENSG00000213719 CLIC1 6 0.46 0.000239 0.097026
Genes with a FDR < 0.1 are presented in this table, ranked from low to high. Genes with a positive log2 fold change have a higher expression in the tumours with a loss of chr22; genes with a negative log2 fold change have a higher expression in the tumours without a loss of chr22. p-value is corrected by Benjamini- Hochberg method (FDR, rightmost column).
Supplementary figure 1 One patient with a partial loss of chromosome 22. A screenshot of chromosome 22 in one tumour is shown. In the tumour of this patient, the deletion only encompassed 18mb of 22qter (red bar). Of note, the small red bar on q11.21 is not judged as deletion. x-axis represents the positions of all probes across chromosomes 22. y-axis displays the normalised value for each probe, representing the copy number state.
