Interrogating the genetic profile of sporadic colorectal cancer

Background and aims: Colorectal cancer (CRC) is one of the most common types of cancers. Prognosis for CRC is dependent on the stage of the disease at diagnosis and a variety of molecular features including expression of biomarkers and gene mutation. This study sought to (i) develop molecular tests for use in molecular profiling of formalin-fixed paraffin embedded tissue (FFPE) and (ii) interrogate the molecular profile and biomarker expression in multiple cohorts of sporadic CRC. Materials and methods: Firstly, we developed the Quick Multiplexed Consensus (QMC) PCR protocol – a method used in conjunction with High Resolution Melting (HRM) analysis for screening of multiple mutational hotspots (10 hotspots in this study were screened) in the CRC candidate genes using DNA from FFPE tissue. The performance of the assay for the detection of single nucleotide variants (SNV) and small insertion-deletions (indels) was compared with Sanger sequencing, pyrosequencing and real-time PCR. In addition, a PCR/HRM assay was developed for testing microsatellite instability (MSI) using a panel of six quasi-monomorphic mononucleotide markers (five of which have been previously described: BAT25, BAT26, NR21, NR22, and NR24; the sixth marker, B-CAT25, is a novel in-house developed marker) and this assay was validated against the Bethesda commercial MSI PCR kit and against immunohistochemistry (IHC). Once QMC-PCR was optimised, we sought to ascertain whether diagnostic biopsies yield sufficient information for clinical decision making. The mutation profiles of CRC candidate genes (KRAS, BRAF, PIK3CA and TP53) and MSI status were evaluated in the diagnostic biopsy specimen of 30 cases and compared with the profile in matched blocks from the resection specimen. The classification of CRC into tumours with MSI and chromosomal instability (CIN) is well described. However tumours which show neither MSI nor CIN have been described and are known as microsatellite and chromosomal stable CRCs (MACS). We sought to evaluate the frequency of CIN and MACS in a series of 89 sporadic microsatellite-stable CRCs. PCR/HRM was used to exclude MSI tumous, while CIN tumours was discriminated from MACS tumours using flow cytometry. These tumours were also compared for mutations in KRAS/BRAF/TP53/PIK3CA by QMC-PCR. Some rectal tumours may receive neoadjuvant chemoradiotherapy (CRT) and we tested whether the CIN or MACS ploidy status could predict response in a group of 62 rectal cancers treated with neoadjuvant CRT. We next evaluated the clinical utility of deficient Mismatch Repair (dMMR) protein and aberrant P53 protein expression, both individually and in combination, as prognostic and predictive of response to therapy biomarkers in CRC. A total of 884 tumours from the VICTOR trial (a large phase III trial of rofecoxib in stage II and III CRC) were tested for expression of four mismatch repair (MMR) proteins (MLH1, PMS2, MSH2, MSH6) and p53 by IHC. The expression was correlated with outcome. Results: QMC-PCR worked on DNA from FFPE tissue. Spiking experiments showed the protocol could detect a minimum of 2.5% of mutant alleles compared to 20% detectable for Sanger sequencing. Precision tests showed that there was little intra-assay and inter-assay variation. Forty-three FFPE colorectal tumours were initially sequenced for hotspots in KRAS and PIK3CA and then screened by QMC-PCR. In these tests, QMC-PCR showed a sensitivity of 100%, specificity of 71%, PPV of 76% and NPV of 100%. All 43 samples were then screened for mutations in all 10 hotspots. Of 430 tests, 43 (10%) showed aberrant melting and 36 were confirmed mutant (PPV 84%). Since our technique is more sensitive than direct sequencing, the remaining 7 tests are probably sequencing false negatives. HRM analysis of KRAS (codon12/13) and BRAF (V600E) showed that 3% and 1.5% mutant alleles respectively could be reliably detected whilst pyrosequencing reliably detected 6% mutant alleles in each case. Of 110 tests performed on 22 DNA samples, in 109 cases HRM and pyrosequencing gave identical results. The results of KRAS (codon12/13) screening from the comparative study of 468 cases with a real-time PCR test (DxS), showed 89.9% concordance with the DxS test. In comparison with real-time PCR, the analysis by HRM showed 87.2% sensitivity, 91.6% specificity, 86.7% positive predictive value, 91.9% negative predictive value. After refinement of the protocol, re-testing of 100 randomly selected cases showed 99% concordance between QMC-PCR/HRM and the real-time PCR test. The performance of the HRM based MSI assay showed 100% accuracy on both cell lines and FFPE tissue in comparison to PCR and IHC. Pertaining to the results from the paired biopsy and resection specimens work, a total of 570 paired PCR tests were performed and identical results were obtained in both biopsy and resection specimens in 569 tests (>99% concordance). Four cases (13%) showed microsatellite instability, and, in all four cases, instability was seen at identical mononucleotide markers in both biopsy and matched resection specimens. Regarding the immunostaining of MMR proteins, the staining was more intense and easier to interpret in biopsies and that it faithfully replicates the diagnosis in the resection specimen. The results from the DNA ploidy study showed, in addition to the two classical forms of genetic instability (MSI, CIN), a significant third group of colorectal cancers (CRC) without either MSI or CIN i.e. Microsatellite and Chromosome Stable (MACS) CRCs. Fifty-one of 89 CRCs (57%) were aneuploid and 38 (43%) were diploid. There was no significant association seen between mutations in TP53/KRAS/BRAF/PIK3CA with ploidy. Testing of association between mutations revealed only mutual exclusivity of KRAS/BRAF mutation (p<0.001). Of the 62 rectal cancers treated with neoadjuvant chemoradiotherapy, 22 had responded (Mandard Tumour Regression Grade 1/2) and 40 failed to respond (Grade 3-5). Twenty-five of 62 (40%) tumours were diploid but there was no association between ploidy and response to the neoadjuvant CRT therapy. Evaluation of TP53 and MMR protein expression in the VICTOR trial, showed that there was dMMR in 12% (87/735) of patients. It was associated with female gender (p=0.001), proximal location of tumour (p<0.001), poor differentiation (p<0.001) and stage II disease (p<0.001). dMMR was not associated with either disease free survival (DFS) or overall survival (OS) in the group overall or when stratified by stage. However, unexpectedly, proficient MMR (pMMR) was associated with significantly improved OS (hazard ratio 0.28 (0.11, 0.68), p=0.005) and DFS (hazard ratio 0.47 (0.22, 0.99), p=0.047) in patients who did not receive chemotherapy. Aberrant p53 expression was found in 65% (482/740) of patients. It was associated with distal location (p<0.001) and stage III disease (p<0.001) and it was negatively associated with dMMR (p