The SNP rs3834129 was genotyped by TaqMan assay (Section 2.2.9.1) (Figure 3.1) in 1193 cases and 1388 controls from the UK (Sheffield, Leeds and Dundee) and the USA (Utah) populations (these are described in Section 2.1.3). A summary some of the characteristics of the cases and controls included in the study is presented in Table 3.1.
Figure 3.1 TaqMan genotyping
Genotyping clusters of rs3834129 Taqman assay from a 384 well plate from the Sheffield population. The X and the Y axes represent fluorescent signals from VIC® and FAMTM labelled probes respectively. Blue and Red clusters represent homozygous insertion and homozygous deletion genotypes respectively. Green cluster represents heterozygous
Table 3.1 Summary of the cases and controls for rs3834129 genotyping
Sheffield Leeds Dundee Utah
Cases Controls Cases Controls Cases Controls Cases Controls n (%) Total 475 (100.0) 447 (100.0) 270 (100.0) 227 (100.0) 137(100.0) 365(100.0) 455 (100.0) 449 (100.0) Sex Male 264 (55.6) 221 (49.4) 153 (56.7) 131 (57.7) 85(62.0) 189 (51.8) 250 (55.0) 250 (55.7) Female 211 (44.4) 220 (49.2) 116 (43.0) 96 (42.3) 52(38.0) 173 (47.4) 205 (45.0) 199 (44.3) Unknown NA 6 (13.4) 1 (0.4) NA NA 3 (0.82) NA NA Family History* None 393 (82.7) 405 (90.6) 228 (84.4) 202 (89.0) 109 (79.6) 315 (86.3) 62 (13.6) 420 (93.5) 1 relative 69 (14.5) 36 (8.1) 37 (13.7) 24 (10.6) 26 (19.0) 45 (12.3) 264 (58.0) 25 (5.6) ≥ 2 relatives 13 (2.7) 6 (1.3) 5 (1.9) 1 (0.4) 6 (4.4) 5 (1.4) 129 (28.4) 4 (0.9) Age** ≤50 25 (5.3) 18 (4.0) 11 (4.1) 11 (4.9) 6 (4.4) 16 (4.4) 54 (11.8) NA 51-59 82 (17.3) 98 (21.9) 40 (14.8) 29 (12.8) 20 (14.6) 67 (18.5) 69 (15.2) NA 60-69 127 (27.6) 171 (38.3) 94 (34.8) 85 (37.4) 51 (37.2) 143 (39.5) 134 (29.5) NA ≥70 237(49.9) 160 (35.8) 124 (45.9) 102 (44.9) 60 (43.8) 136 (37.6) 185 (40.7) NA Mean (Range) 68 (30-96) 66 (26-86) 67.4 (45-89) 67.1 (45-80) 67.2 (46-80) 66.7 (45-86) 64.9 (29-91) NA CRC site*** Proximal 122 (25.7) NA 57 (24.4) NA 22 (16.0) NA 158 (32.6) NA Distal 126 (26.5) NA 81 (34.6) NA 49 (35.8) NA 147 (30.4) NA Rectal 190 (40.0) NA 92 (39.3) NA 62 (45.3) NA 115 (25.3) NA Unknown 37 (7.8) NA 4 (1.7) NA 4 (2.9) NA 64 (14.1) NA
* Family history includes 1st degree relatives only.
** Controls for Utah were samples collected from previous studies. All controls were cancer free and matched by sex and year of birth (Age will reflect the age at recruitment for previous studies, therefore it will not necessarily match for this study (Section 2.1.3.3).
*** CRC site for Utah includes 29 cases with multiple primary CRC.
3.2.1.1 Quality control
In order to assess the quality of the DNA samples and the accuracy of the TaqMan assay and genotyping results, several quality control assessments were performed. The assay genotyping call rate represents the number of successfully genotyped samples out of the total number of samples. It gives an indication of both assay and sample quality. Usually, a call rate of ~95% indicates that both the assay and the samples are of high quality. The call rate achieved for the different genotyped populations was between 92-96%, as shown in Table 3.2. In each 384 well taqman plate, 5-10% of the samples were duplicated. The duplicate concordance represents the percentage of the duplicated samples that were successfully called with the same genotype. A duplicate concordance of 98% represents high accuracy. An overall duplicate rate of 97.5% was obtained for all of the genotyped cohorts combined (range 95-100%) (Table 3.2). As mentioned in Section 2.2.21.2, testing for Hardy
Weinberg equilibrium (HWE) is an important step to investigate the quality of the genotyping assay. Table 3.2 summarises the p-values for the populations genotyped for rs3834129. The chi-squared test was performed on the control genotypes only. All the p-values were >0.05 suggesting that all the genotyped populations were consistent with HWE. In summary, the quality control analysis of the genotyping results indicates reasonable data quality.
Table 3.2 Quality control results of the genotyped cohorts Study Group Samples (n) Call rate
(%)
Overall call rate (%) Duplicate Concordance % (n)* HWE (p-value) Sheffield Cases 436 92.0 92.3 94.7 (94) 0.507 Controls 442 92.5 Leeds Cases 262 94.7 96.1 100.0 (41) 0.254 Controls 226 97.8 Dundee Cases 136 94.8 94.0 97.2 (36) 0.990 Controls 364 93.7 Utah Cases 451 92.0 93.3 100.0 (71) 0.714 Controls 444 94.6
3.2.1.2 Allelic discrimination and genotypes
Table 3.3 represents a summary of the rs3834129 genotype counts for the cases and controls from the genotyped populations. Moreover, Table 3.4 shows the observed allele frequencies in this study, together with those from previous publications and dbSNP. Comparison with the known frequencies of rs3834129 shows an agreement between our results and the results from the European and multi-ethnic American populations. However, it shows substantial difference from the Asian population.
Table 3.3 rs3834129 Genotyping results Genotyping results
n (%)
Ins/Ins Ins/Del Del/Del Total Sheffield Cases 107 (26.7) 186 (46.4) 108 (26.9) 401 Controls 119 (29.1) 197 (48.2) 93 (22.7) 409 Leeds Cases 53 (21.4) 126 (50.8) 69 (27.8) 248 Controls 58 (26.2) 102 (46.2) 61 (27.6) 221 Dundee Cases 40 (31.0) 60 (46.5) 29 (22.5) 129 Controls 80 (23.5) 171 (50.1) 90 (26.4) 341 Utah Cases 107 (25.8) 210 (50.6) 98 (23.6) 415 Controls 116 (27.6) 213 (50.7) 91 (21.7) 420
Table 3.4 rs3834129 frequencies rs3834129 allele frequencies Insertion Deletion Sheffield 0.53 0.47 Leeds 0.49 0.51 Dundee 0.49 0.51 Utah 0.53 0.47 Pittman et al. 2008 0.50 0.50 Haiman et al. 2008 0.50-0.52 0.48-0.50 European panel dbSNP(133) 0.452 0.548 Sun et al. 2007 0.76 0.24 Asian panel dbSNP(133) 0.783 0.217
Pittman et al. 2008 and Haiman et al. 2008 studies were performed in European and multi- ethnic American populations respectively. Sun et al. 2007 study was performed in a Chinese population. dbSNP allele frequency was obtained from the database release 133.
3.2.1.3 Association of rs3834129 with colon cancer
A contingency table chi-squared test (Degree of freedom (df) =2) was used to investigate whether there was a significant difference in the frequency of the genotypes between cases and controls (Section 2.2.21.1). The test was applied separately on the four different populations using observed and expected values based on the null hypothesis of no association between genotype and case status. The p-values for the 4 populations are summarised in Table 3.5.
The odds ratios (OR) for CRC relative risk of the heterozygous and the homozygous deletion (the rare allele) genotypes were calculated in comparison to the homozygous insertion (the common reference allele) (Section 2.2.21.3). OR were calculated separately for the different populations and a combined OR was also computed for the 4 populations with adjustment for study size. Finally, an additive OR model assuming an increase in CRC risk with each copy of the deletion allele (allele dose) was calculated for each study separately and with all the studies combined. The calculated OR are summarised in Table 3.5. All the calculated p-values were > 0.05 and all the estimated OR 95% confidence intervals overlapped 1.0. Therefore, these data provide no evidence of association between rs3834129 and CRC risk.
Table 3.5 Chi-squared test p-values and odds ratios Odds ratios (OR) Study p-value* Heterozygous OR
(95% CI) Homozygous Del OR (95% CI) Additive model OR (95% CI) Sheffield 0.369 1.050 (0.756-1.459) 1.292 (0.882-1.890) 1.134 (0.937-1.372) Leeds 0.429 1.352 (0.858-2.130) 1.238 (0.745-2.056) 1.105 (0.857-1.424) Dundee 0.234 0.702 (0.434-1.134) 0.644 (0.366-1.134) 0.797 (0.599-1.061) Utah 0.736 1.069 (0.773-1.479) 1.168 (0.792-1.721) 1.080 (0.890-1.310) Combined data** 1.040 (0.860-1.258) 1.110 (0.891-1.384) 1.053 (0.944-1.176) Table 3.5 summarises the chi-squared test p-values, and the odd ratios for the different populations used in rs3834129 genotyping.
* 2x3 Chi-squared p-values testing the null hypothesis that there is no significant difference in genotypes frequencies between cases and controls.
** Combined OR were adjusted for the sizes of the 4 studies based on a logistic regression analysis by Dr Cox.