In the present study, when PCDH10 V1 and PCDH10 V2 single stable clones were treated with camptothecin, increased drug susceptibility of HCT116 cells was observed.
Although camptothecin is no longer used in the chemotherapy for the treatment of CRC due to its low solubility and adverse drug effect, however, analogues or derivatives of camptothecin, such as Irinotecan, are used in routine treatment. These kinds of analogues act in the same mechanism as that of camptothecin.38 Based on the above results, re-expression of PCDH10 V1 or PCDH10 V2 exerted impacts on apoptosis after HCT116 treated with camptothecin. This is an interesting result because PCDH10 V2 possesses a much shorter cytoplasmic tail than PCDH10 V1. Cytoplasmic domain of PCDH10 is assumed to transduce signals to the cells. Previous results from our lab on the investigation of PCDH10 V2 RNA expression reported that the expression level of PCDH10 V2 was very low and it was often associated with strong PCDH10 V1 expression. Therefore, the results observed by over-expression of
PCDH10 V2 may not be compatible with in vivo situation. Future studies on apoptosis
will recruit chemotherapeutic drugs such as 5-Fluorouracil, that are used in present clinical regimens.39 40
Loss or gain of chromosomal regions is one of the most important mechanisms
33
underlying tumorigenesis, in which tumor cells may gain abnormal growth control.
Besides, loss or gain of chromosomal regions also leads to alteration in gene expression. It is postulated that tumor suppressor genes often exist in regions where losses occurred, and is often reflected as allelic imbalance or LOH at molecular level.41 According to the allelic deletion study of PCDH10 gene, heterozygosity rates for all the four markers, except for D4P4, were relatively low. D4P1 showed the lowest heterozygosity rate (56.9%) and its location is the farthest from PCDH10. Because of its low heterozygosity rate and low LOH frequency, we conclude that this marker may be dispensable and can be excluded when determining allelic deletion of PCDH10 gene. On the other hand, marker D4P3, which is located 388bp downstream of PCDH10, showed the highest LOH frequency (32.8%). This marker alone allows us to
determine the status of allelic deletion of PCDH10 for most cases. Statistical analysis of LOH status in each marker and in allelic deletion of PCDH10 with clinicopathological characteristics of patients showed that incidence of LOH in markers D4P2 and D4P3, and allelic deletion of PCDH10, were significantly associated with distal metastasis of CRC. This result corresponded to the results from the previous studies in our lab, in which invasion ability decreased when PCDH10 V1 was re-expressed in CRC cells. Since markers D4P2 and D4P3 were critical when determining the status of allelic deletion, so when LOH of these two markers was
34
observed, allelic deletion of PCDH10 was defined. The results showed that allelic deletion of PCDH10 were correlated to a poor OS of patients as estimated by the Kaplan-Meier method and compared by log-rank test, however, it was not found to be correlated to DFS in Duke’s B and C patients. Patients in Dukes’ D, who have the worst overall survival, may contribute bias when considering overall survival.
However, in our study of allelic deletion of PCDH10 and overall survival, the number of patients in Dukes’ D was similar to the remaining groups, which can eliminate the bias from this group of patients in survival analysis. In addition, when investigating the impacts of allelic deletion of PCDH10 on survival, change in survival rate and duration of patients with CRC should be considered because due to the advances in the treatment of CRC, five-year survival rate of patients will change over time. This change may also cause a bias in the current survival analysis. At present, approximately one-fourth of patients have metastatic CRC at diagnosis, with a five-year survival rate of less than 10%. Nevertheless, with the development of target therapy and combination of chemotherapy, patients with metastatic CRC have their median overall survival raised from six to eight months, up to 20 months over the past ten years.40 42 Though patients in our study were collected throughout ten years of time, however, even with the advances in the treatment of CRC, significant difference in overall survival can still be observed in patients with PCDH10 allelic deletion. In the
35
future, multi-variable analysis on survival will be carried out.
With the availability of PCDH10 methylation status of 74 paired CRC samples, OS analysis on combined influence of incidence of LOH and promoter hypermethylation was performed. The outcome was not statistically significant, however, this may be due to the small sample size. Moreover, patients with allelic deletion only displayed poorer OS than any other conditions, while patients without both of allelic deletion and promoter hypermethylation of PCDH10 gene showed relatively better OS. More samples should be included in future study.
PCDH10 is reported as a transmembrane protein, this prompts us to investigate whether PCDH10 is involved in the EMT pathway or not. Preliminary results using TGF-β to treat CRC cell line SW480 showed induction of cells to undergo EMT. In future study, PCDH10 V1 will be transiently transfected to SW480 and treated with TGF-β in order to study PCDH10 effect on EMT. Besides, EMT is reported to be one of the mechanism underlying tumor metastasis,43 and study from the above allelic deletion of PCDH10 also revealed association of PCDH10 with distal metastasis.
Therefore, study of involvement of PCDH10 in EMT was of great interest.
Taqman®Copy number assay was used to compare with the results of LOH study, in hope of finding a method to substitute the laborious process when carrying out the LOH study. Normal PBMCs showed a copy number of two of PCDH10 gene, and up
36
to date, no copy number variations for PCDH10 have been reported in the database of genomic variants.44 Nevertheless, we cannot observe a positive correlation between copy number assay and LOH study. Most cases with incidence of allelic deletion of PCDH10 showed no evidence of a reduced copy number, and by contrast, a few cases
showed an increased copy number. This may be a consequence of normal cell contamination using patients’ primary tumors41 because our samples were not obtained by laser capture, microdissection, or because of intratumor heterogeneity. In addition, only one specific region in exon four was investigated by this copy number assay, it may not be enough to detect the loss of other regions of PCDH10. Many other target sites have to be included in further comparison of LOH study to copy number loss.
Recently, frequent occurrence of uniparental disomy revealed by SNP arrays was reported in CRC and other cancers.45 46 Uniparental disomy is a genetic change in which one allele is lost while the remaining allele gains through duplication, resulting in LOH but no copy number change. Genomic alterations of this kind were also reported to impose significant changes on gene expression, suggesting that when gene dosage was unchanged due to uniparental disomy, the gene expression level may mainly remained unchanged. This has to be taken into consideration when performing LOH study and copy number assay.
Controversial roles of PCDH10 V2 were found in the proliferation assay and
37
apoptosis assay. Investigation of the effect of PCDH10 V2 on PCDH10 V1 will be carried out using transient overexpression in PCDH10 V1 single stable clones. If the results are conclusive, we will move on to stable clones selection using pEF6/V5-His-TOPO which contains different selection marker from PCDH10 V1. The PCDH10 V2 sequence was obtained from the original PCDH10
V2/pcDNA3.1/V5-His-TOPO through enzyme digestion and ligation to the pEF6 vector. Nevertheless, as mentioned previously, in vitro culture systems may not be able to reflect in vivo condition and PCDH10 was reported to exist cell-cell adhesion, therefore, interaction of the cellular matrix or extracellular environments may have trigger different influence of PCDH10 V2 on PCDH10 V1.
Other protocadherins were also reported to be candidate tumor suppressors in various carcinomas besides Protocadherin 10.47 Decreased expression of Protocadherin 1 (PCDH1) was shown to be a predictor for poorer overall survival in medulloblastoma patients.48 Loss of Protocadherin 8 (PCDH8), which is the human ortholog of Paraxial Protocadherin in frog, fish and mouse, through mutation and epigenetic silencing was also found in breast carcinomas.49 Protocadherin 20 (PCDH20) exhibited hypermethylation at the promoter region in non-small-cell lung cancer cell lines. Hypermethylation of PCDH20 was also shown to be associated with a poorer overall survival in patients.50 These further support that the non-classic
38
cadherins may play a role in tumor suppression aside from cell-cell adhesion.
In conclusion, apoptosis and gene deletion studies on PCDH10 support its role as a putative tumor suppressor gene in CRC. The sequence of which LOH or hypermethylation takes place, occurs at which tumor stage, or which mechanism play a more important role in CRC tumorigenesis remain to be answered.
39
Figures
Figure 1. Protein expression of newly selected PCDH10 variant 1 and PCDH10 variant 2 single stable clones.
Protein levels of PCDH10 V1 and PCDH10 V2 single stable clones were determined by Western blotting with anti-V5 antibody. β-actin was used as a protein loading control. Molecular weight of PCDH10 V1 and PCDH10 V2 is 116 kDa and 100 kDa, respectively according to NCBI. Mock represents pcDNA3.1/V5-His-TOPO. V1 and V2 stand for PCDH10 V1 and PCDH10 V2, respectively.
40
A.
B.
Figure 2. Cell proliferation of four new single stable clones of PCDH10 V1 using MTT assay (A) and the corresponding protein expression (B).
The proliferation rates of four PCDH10 V1 single stable clones were compared with a pcDNA3.1 only stable line (Mock) by MTT assay (A), and the corresponding protein expression was determined by Western blotting (B). Cell proliferation decreased significantly with PCDH10 V1 re-expression when compared to Mock (*p<0.01). V1 stands for PCDH10 V1.
41
A.
B.
Figure 3. Cell proliferation of four new single stable clones of PCDH10 V2 using MTT assay (A) and the corresponding protein expression (B).
The proliferation rates of four PCDH10 V2 single stable clones were compared with a pcDNA3.1 only stable line (Mock) by MTT assay (A), and the corresponding protein expression was determined by Western blotting (B). Cell proliferation was similar to Mock with PCDH10 V2 re-expression. V2 stands for PCDH10 V2.
42
A
B.
C.
Figure 4. Re-expression of PCDH10 V1 and PCDH10 V2 induced spontaneous apoptosis of CRC cell line, HCT116.
The apoptosis percentages of PCDH10 V1 and PCDH10 V2 single stable clones were compared with a pcDNA3.1 only stable line (Mock) by Annexin V-PI staining (A), and quantitative results of flow cytometry were shown in (B). Corresponding protein expression was detected by Western blotting (C). V1 and V2 stands for PCDH10 V1 and PCDH10 V2, respectively. Apoptosis percentages increased with the re-expression of PCDH10 when compared to Mock. Mock showed similar apoptosis percentage to the parental cell line, HCT116.
43
A.
B.
C.
Figure 5. Camptothecin-induced apoptosis of PCDH10 V1 and PCDH10 V2 single stable clones.
DMSO was used to treat PCDH10 V1 and PCDH10 V2 single stable clones as control since camptothecin was dissolved in DMSO. Percentage of apoptotic cells of DMSO-treated group was shown in (A). Percentages of apoptotic cells in PCDH10 V1 and PCDH10 V2 single stable clones were shown after subtracting that of the DMSO-treated Mock, except for HCT116, which shows the percentage after subtracting its own DMSO-treated control (B). Corresponding protein expression was detected by Western blotting (C). Mock represents pcDNA3.1/V5-His-TOPO. V1 and V2 stands for PCDH10 V1 and PCDH10 V2, respectively.
44
Figure 6. Relative expression of E-cadherin in all PCDH10 single stable clones.
Western blot analysis shows protein expression of E-cadherin in various cell clones.
Mock represents pcDNA3.1/V5-His-TOPO. V1 and V2 stands for PCDH10 V1 and PCDH10 V2, respectively. Expression level of E-cadherin was normalized to that of β-actin, and the results were compared with Mock.
45
A.
B.
Figure 7. Different concentration of TGF-β treatment on CRC cell lines, SW480 and HT29, for 24 or 48h.
5, 10 or 20ng/mL TGF-β was used to treat CRC cell lines SW480 and HT29 for 24h (A) or 48h (B) to undergo EMT. Protein expression of vimentin in HT29 cannot be detected. All protein expression level was normalized to β-actin. Values are expressed as fold change compared with the untreated group.
46
A.
B.
Figure 8. Seven microsatellite markers for the detection of allelic deletion of PCDH10 by LOH study.
Preliminary test of the seven primer sets for amplifying the repetitive sequence No.1 to No.7 surrounding the PCDH10 gene. Figure 8A shows the location of the markers, in which negative values represent the numbers of base pairs (bp) upstream from the first nucleotide of the PCDH10 gene, and positive values represent the numbers of bp downstream from the last nucleotide of the PCDH10 gene. Figure 8B shows the result of PCR visualized by gel electrophoresis. DNA was obtained from normal PBMCs. M:
100bp DNA size marker.
47
Figure 9. Optimization of PCR condition for the seven markers designed.
The markers, No.1 to No.7, represent the same repetitive sequences designed in Figure 8. 60 and 58 indicate the annealing temperature during PCR. M: 100bp DNA size marker.
48
Figure 10. Detection of polymorphisms of the seven markers in a panel of eight DNA samples from normal PBMCs.
PBMCs 1-8 indicate the DNA from eight individuals. Markers No.1, 3, 4, 5 and 6 show polymorphism. M: 100bp DNA size marker.
49
Marker Location Repeat unit Product size (bp)
D4P1 -30477 AC(11)T AC(9) 149
D4P2 -4426 AC(14) 189
D4P3 388 AC(16) 304
D4P4 7711 AAT(13) 324
Figure 11. Four markers designed for the study of allelic deletion of PCDH10.
Negative values in location represent the numbers of base pairs (bp) upstream from the first nucleotide of the PCDH10 gene, and positive values represent the numbers of bp downstream from the last nucleotide of the PCDH10 gene. Repeat numbers in repeat unit and product sizes are according to the sequence from Ensembl (ENSG00000138650).
50
Figure 12. Results of LOH study of one paired CRC sample analyzed by GeneScan.
X-axis represents the size of the PCR product, and y-axis represents the intensity of the fluorescence. Marker D4P4 shows a result of homozygous alleles while the other markers show a result of heterozygous alleles. Markers D4P2 and D4P3 exhibit LOH.
51
A. B.
Figure 13. Kaplan-Meier survival analysis of CRC patients according to allelic deletion of PCDH10 gene.
A. Overall survival of 174 patients with CRC (p=0.0082). B. Disease-free survival of 118 patients with Dukes’ B or C only (p=0.6135).
52
A. B.
Figure 14. Kaplan-Meier survival analysis of CRC patients according to allelic deletion of PCDH10 gene.
Since promoter hypermethylation status was available for 74 patients in CRC, therefore, investigation of incidence of LOH of PCDH10 and promoter hypermethylation with OS and DFS was carried out. Independent analysis for incidence of LOH and promoter hypermethylation with OS and DFS was first performed. Here shows the result of OS for 65 patients according to allelic deletion of PCDH10 gene. (A) There are total 65 patients for OS analysis because of two patients died of postoperative complications, three without follow-up after surgery and four without informative results in LOH study. (B). For DFS analysis, only 47 patients in Dukes’ B and C were recruited. No correlation was found between incidence of LOH and OS or DFS (p=0.1273 and p=0.4680, respectively).
53
A. B.
Figure 15. Kaplan-Meier survival analysis of CRC patients according to methylation status of PCDH10 promoter region.
A. Overall survival analysis of 69 patients with CRC (p=0.3212). B. Disease-free survival analysis of 51 patients with Duke’s B and C only (p=0.1354).
54
A B.
Figure 16. Kaplan-Meier survival analysis of 65 CRC patients according to LOH and promoter hypermethylation (Me) of PCDH10 gene
A. There was no statistically significant correlation between the incidence of LOH and promoter hypermethylation of PCDH10 with OS in the four groups (p=0.4655). B.
Patients in A. are divided into two groups: neither with the presentation of LOH nor promoter hypermethylation, and with presentation of LOH and/or promoter hypermethylation. There was no statistically significant correlation between the incidence of LOH and/or promoter hypermethylation of PCDH10 with OS in the two groups (p=0.7704).
55
A.
B.
C.
Figure 17. Copy number assay of PCDH10 gene.
A. Normal PBMCs are obtained from ten individuals (1-10), and the others are paired CRC samples. B. PCDH10 copy numbers were determined with a panel of CRC cell lines. C. PCDH10 copy numbers were determined with 18 colorectal tumors with apparent LOH.
56
Tables
Table 1. Patient clinicopathological characteristics (n=211)
Characteristic No. of patients (%)
Age (yr)
Proximal colon 49 (23.2)
Distal colon 162 (76.8)
a Proximal colon includes the cecum, ascending colon and transverse colon. Distal colon includes the descending colon, sigmoid colon, rectosigmoid junction and the rectum.
b Early stage refers to Dukes’ A and B. Late stage refers to Dukes’ C and D.
c Distal metastasis: Dukes’ D; no distal metastasis: Dukes’ A, B and C.
d 209 patients in total because of two patients without pathological reports.
57
e 129 patients with Dukes’ B and C, who obtained curable surgery before December 31st, 2008, are assessed for disease recurrence.
58
Table 2. Primers for LOH study Marker Locationa Repeat unitb Primer sequencec Fluorescence dye labeledd Product size (bp)b D4P1 -30477 AC(11)T AC(9)5’ GGG ATT GTT TCT AAG GCC AA 3’ 5’GTT TCT TCA TAC CCA TAC ACA AAC ATC ACC 3’
6-FAM 149 D4P2 -4426 AC(14) 5’ CTA CCA AAA GAC ACA ATA CGT GA A G 3’ 5’ GTT TCT TTT AAA TTG GGC ACT GGT TTG 3’
VIC 189 D4P3 +388 AC(16)5’ CAA TGA CAG TTT GAG CTG CAC 3’ 5’ GTT TCT TGA CTT CAA AAG GCT GGA ACA TT 3’
6-FAM 304 D4P4 +7711 AAT(13) 5’ TCC CAG ATA CTT GAA AGG CTG AA 3’ 5’ GTT TCT TCC GAA ATG CCT CTC AAT TC 3’
VIC 324 a Negative values in location represent the numbers of base pairs (bp) upstream from the first nucleotide of PCDH10 gene, and positive values represent the numbers of bp downstream from the last nucleotide of PCDH10 gene. b Repeat numbers in the parentheses of repeat unit and product sizes are according to sequence from Ensembl (ENSG00000138650). c Seven nucleotides GTT TCT T is added to the 5’ end of each reverse primer in order to reduce the minus-A phenomenon in PCR. d Fluorescence dye is labeled at the 5’ end of each forward primer.
59
Table 3. LOH frequencies of the four microsatellite markers in 211 patients with CRC
a Total no. of samples with heterozygous alleles in normal mucosa.
b Percentage of the incidence of heterozygosity among 211 samples.
c Percentage of the incidence of LOH among informative cases.
d 209 cases in total due to failure in the PCR amplification of D4P4 in 2 samples.
Marker Informative no.a Heterozygosity (%)b LOH no. LOH frequency (%)c
D4P1 120 56.9 19 15.8
D4P2 130 61.6 26 20.0
D4P3 137 64.9 45 32.8
D4P4d 151 72.2 38 25.2
60
Table 4. Relationship between patient clinicopathological characteristics and the incidence of LOH for each marker (n=201)a No. of patients (%) D4P1 D4P2 D4P3 D4P4 CharacteristicLOHRetainedp LOHRetainedp LOHRetainedp LOHRetainedp Numberb 19 (16.2) 98 (83.8) 26 (20.8) 99 (79.2) 43 (32.6) 89 (67.4) 36 (25.0) 108 (75.0) Age (yr) 0.6623 1.0000 0.7420 0.9112 Median 7373737674717673 Range 45-9337-9846-9737-98 46-9337-9846-9737-98 Gender0.7708 0.9634 0.9484 0.6304 Male 10 (17.2) 48 (82.8) 13 (21.0) 49 (79.0) 22 (32.8) 45 (67.2) 19 (26.8) 52 (73.2) Female9 (15.3) 50 (84.7) 13 (20.6) 50 (79.4) 21 (32.3) 44 (67.7) 17 (23.3) 56 (76.7) Tumor location 0.5833 0.7405 0.9198 0.3217 Proximal colon 6 (19.4) 25 (80.6) 6 (18.8) 26 (81.2) 10 (33.3) 20 (66.7) 7 (18.9) 30 (81.1) Distal colon 13 (15.1) 73 (84.9) 20 (21.5) 73 (78.5) 33 (32.4) 69 (67.6) 29 (27.1) 78 (72.9) Dukes’ stage0.5574 0.2690 0.1992 0.3417 A 1 ( 8.3) 11 (91.7) 2 (14.3) 12 (85.7) 2 (18.2) 9 (81.8) 1 (8.3) 11 (91.7) B 7 (15.9) 37 (84.1) 8 (18.2) 36 (81.8) 18 (32.1) 38 (67.9) 13 (24.1) 41 (75.9) C 4 (12.1) 29 (87.9) 6 (15.8) 32 (84.2) 11 (26.8) 30 (73.2) 12 (24.0) 38 (76.0) D 7 (25.0) 21 (75.0) 10 (34.5) 19 (65.5) 12 (50.0) 12 (50.0) 10 (35.7) 18 (64.3)
61
Table 4. Relationship between patient clinicopathological characteristics and the incidence of LOH for each marker (n=201)a (continued) No. of patients (%) D4P1 D4P2 D4P3 D4P4 CharacteristicLOHRetainedp LOHRetainedp LOHRetainedp LOHRetainedp Tumor stagec Early stage 8 (14.3) 48 (85.7) 0.5830 10 (17.2) 48 (82.8) 0.3617 20 (29.9) 47 (70.1) 0.4976 14 (21.2) 52 (78.8) 0.3342 Late stage11 (18.0) 50 (82.0) 16 (23.9) 51 (76.1) 23 (35.4) 42 (64.6) 22 (28.2) 56 (71.8) Distal metastasisd 0.1495 0.0383 0.0440 0. 1446 Yes 7 (25.0) 21 (75.0) 10 (34.5) 19 (65.5) 12 (50.0) 12 (50.0) 10 (35.7) 18 (64.3) No 12 (13.5) 77 (86.5) 16 (16.7) 80 (83.3) 31 (28.7) 77 (71.3) 26 (22.4) 90 (77.6) Pathological differentiatione 0.2540 0.5493 0.8563 0.1739 Moderate 11 (13.8) 69 (86.2) 17 (19.5) 70 (80.5) 31 (31.6) 67 (68.4) 30 (28.0) 77 (72.0) Poor 8 (22.2) 28 (77.8) 9 (24.3) 28 (75.7) 11 (33.3) 22 (66.7) 6 (16.7) 30 (83.3) Disease recurrencef 0.7152 1.0000 0.3081 0.1925 Yes 2 (10.5) 17 (89.5) 3 (15.0) 17 (85.0) 5 (20.8) 19 (79.2) 9 (32.1) 19 (67.9) No 9 (17.3) 43 (82.7) 10 (18.2) 45 (81.8) 21 (32.8) 43 (67.2) 13 (19.7) 53 (80.3) a Only samples obtained before December 31st , 2008 were recruited. b Samples with homozygous alleles were excluded. c Early stage refers to Dukes’ A and B. Late stage refers to Dukes’ C and D. d Distal metastasis refers to Dukes’ D. No distal metastasis refers to Dukes’ A, B and C. e Pathological report of 1 case is not available. f Patients with Dukes’ B and C, who obtained curable surgery before December 31st , 2008, are assessed for disease recurrence. Student’s t test for continuous variables; Fisher’s exact test or chi-square test for categorical variables.
62
Table 5. Relationship between patient clinicopathological characteristics and allelic deletion of PCDH10 gene (n=191)a
No. of patients (%)
Characteristic Allelic deletion (+) Allelic deletion (-) p value
Number 57 (29.8) 134 (70.2)
Distal metastasisc 0.0250
Yes 17 (44.7) 21 (55.3)
a Only 201 patients who had their surgery before December 31st, 2008 were recruited for the study. Among them, 10 patients got non-informative results in all four markers studied, and were excluded from this analysis.
b Early stage refers to Dukes’ A and B. Late stage refers to Dukes’ C and D.
c Distal metastasis refers to Dukes’ D. No distal metastasis refers to Dukes’ A, B and C.
d Pathological reports of 2 cases are not available.
e 123 patients with Dukes’ B and C, who obtained curable surgery before December
63
31st, 2008, are assessed for disease recurrence.
64
Table 6. Correlation of overall survival and disease-free survival with allelic deletion of PCDH10 in CRC patients Overall survival Disease-free survivala PCDH10Case no. OS (%) Mean/Median (months)p valueb Case no. DFS (%) Mean/Median (months)p valueb n 174 118 Allelic deletion 0.0082 0.6135 Yes 54 30 (55.6) 35/35 34 26 (76.5) 48/35 No 120 87 (72.5) 49/33 84 58 (69.0) 63/33 a Only patients with Dukes’ B and C are assessed. b By Log-Rank test.
65
Table 7. Correlation of overall survival and disease-free survival with allelic deletion and/or promoter hypermethylation in patients with CRC. Overall survival Disease-free survivala Variable Case no. OS (%) Mean (months) p valueb Case no. DFS (%) Mean (months) p valueb Allelic deletion (n=65) 0.1273 0.4680 Yes 24 13 (54.2) 34 14 10 (71.4) 48 No 41 29 (70.7) 45 33 21 (63.6) 50 Promoter hypermethylation (n=69) 0.3212 0.1354 Yes 39 28 (71.8) 50 32 25 (78.1) 52 No 30 18 (60.0) 52 19 10 (52.6) 45 Allelic deletion /promoter hypermethylation (n=65) 0.4655 No/No 15 10 (66.7) 68 Yes/No 14 7 (50.0) 35 No/Yes 26 19 (73.1) 47 Yes/Yes 10 6 (60.0) 54 Allelic deletion /promoter hypermethylation (n=65) No/No 15 10 (66.7) 68 0.7704 Except No/No 50 32 (64.0) 45