Indonesian Genomic Landscape of Pathogenic Mutation of APC, KRAS, TP53, PIK3CA, and MLH1 in Colorectal Cancer

Background : Knowing colorectal cancer’s heterogeneity and dynamic features, recognizing its biological behaviour requires detailed identication of mutated genes involved. Colorectal cancer (CRC) requires several mutated genes to occur and those are dissimilar in each person hence essential to be discovered in specic population. Until recently, there is no known study describing genomic landscape of CRC in Indonesian population. This study aims to describe prole of pathogenic mutation of APC, TP53, PIK3CA, KRAS, and MLH1 in CRC patients treated at 3 different hospitals in Jakarta. Methods : This is a descriptive study conducted on CRC patients who underwent neoadjuvant, surgical, and adjuvant therapy at RSCM, RSKJ, and MRCCC in 2017-2018. DNA analysis was performed using next-generation sequencing and aligned against GRCh38. Pathogenic variant was identied using ACMG classication and FATHMM score. Data related to behaviour and survival were collected from medical records. Results : There were total 22 subjects in which APC, TP53, and PIKCA were mutated. KRAS mutation occurred in 64%, while MLH1 in 45%. Five types of mutation were identied, including nonsense, missense, frameshift, splice-site, and silent mutation. There are 4 groups of co-occurring mutations, which are APC, TP53, PIK3CA (triple mutation/TM) alone; TM+KRAS; TM+MLH1; and TM+KRAS+MLH1, presenting different nature and survival. Conclusion : Indonesia having various ethnicities with diverse diet and lifestyle has distinct prole of pathogenic mutation presenting mostly with locally-advanced stage with various outcome and survival rate. compared to


Introduction
Colorectal cancer has been known as one of the most well-studied malignancy. Its dynamic and heterogeneity are characterized by many interconnecting molecular etiopathogenesis exhibiting different behavior inter and intratumor. [1][2][3][4][5] Based on recent biomolecular study, genetic and epigenetic analysis can evaluate the nature of the tumor, hence able to predict heredity, progressivity, recurrency, response to therapy, and even survival rate. Those variables cannot be estimated by AJCC staging system alone. For this reason, precision medicine rooting to genomic pro le of each individual, are starting to advance.
Colorectal malignancy, which involves at least three or four genetic mutation, are feasible for next-generation sequencing methods. 2,6 Two of the three most common carcinogenic pathways are chromosomal and microsatellite instability. [7][8][9][10] Five genes which frequently involved are APC, TP53, KRAS, PIK3CA, and MLH1. Different group of age, gender, and geographic location actually have different variations of mutation and genes involved so that study on speci c population is important in advancing precision medicine. 11 Until recently, there is no publication providing genomic landscape of colorectal cancer in Indonesian population. This study aims to analyze genomic pro le of colorectal cancer in Indonesia.

Methods
This is a descriptive study in patients with colorectal malignancies who underwent surgery and/or chemoradiation and/or chemotherapy at RSCM, RSKJ, and MRCCC in 2017-2018 whose tumor tissue specimens were still properly stored in the form of formalin-xed para n-embedded (FFPE). This study has been reported in line with STROCSS criteria. 12 Sample Preparation
Quintuple mutation were identi ed in 6 patients, dominated by older age, locally-advanced stage, well-differentiated, positive lymphovascular invasion, and located in rectum or left colon.  Figure 1).
Fifty percent of subjects of cluster 1 and 2 were deceased in less than 6 months after therapy, in cluster 4, 50% of subjects were deceased before month 15. All patients of cluster 3 can survive up to 30 months after therapy and only 1 patient deceased afterwards. Cluster 2 and 4 shows highest mortality rate with highest number of deceased patients in shortest period compared to other clusters ( Figure 2).

Other ndings
Early recurrence (<5 years) occurred in 2 patients of cluster 4, of which 1 patient underwent neoadjuvant chemoradiation and adjuvant chemotherapy (MFOLFOX6) and another was given XELOX after surgery. Both patients have disease-free interval of 15 months.
One patient was given anti-EGFR therapy (cetuximab) + MFOLFOX6. Patient's PCR result for KRAS was wild-type. There is no data of therapeutic response due to patient's death during midcycle (127 days after surgery). This patient was included in cluster 2 (with KRAS mutation) and also had EGFR mutation (rs121913467).
One patient was given anti-VEGF therapy (bevacizumab) + MFOLFOX6 after diagnosed with local recurrence after 1-year of oral capecitabine. This patient was included in cluster 3 with noted BRAF mutation as well (rs121913353). Patients was known to have complete response to bevacizumab.
Two of 22 patients had family history of malignancy (Table 1). In patient with family history of colon cancer, there was 1 germline mutation identi ed in STK11, meanwhile in patient with family history of breast cancer, 2 germline TP53 mutations were identi ed. The theory of colorectal neoplasia, namely adenoma-carcinoma sequence (ACS) states that the formation of colorectal carcinoma must be preceded by the presence of an adenoma. 1,2 Changes in the normal intestinal mucosal epithelium to adenoma are triggered by mutations in the tumor suppressor gene, that is APC. APC can be detected in the aberrant crypt foci (ACF), which is a precursor lesion that occurred early in the beginning of the formation of adenomatous polyps and can only appear in dysplastic lesions. 15 All subjects (100%) in this study had nonsynonymous mutations in APC. Only 2 patients had adenomas on colonoscopy. One of those had tubulous adenomas with mild dysplasia on colonoscopy and rst-degree relative with CRC. Nonsense mutated APC was found at codons 879, 1095, 1123, which completely stopped glutamine production (Q). Meanwhile, in another patient with villous adenomas and well-differentiated adenocarcinoma, nonsense mutations were found at codons 876, 879, 1096, 1291, 1294 and 1517 that stopped the production of the amino acids glutamine (Q) and arginine (R). Mutations in APC are known to have high-penetrance that can reach 100% for FAP and CRC. [16][17][18][19] In contrast to the Japanese population, whose APC mutations scattered at codons 142 -1513, subjects in this study had APC mutations occur at codons 876 -1517 with mutation cluster region (MCR) in exons 14 -17. 20,21 After the normal mucosal epithelium turned into an early adenoma, KRAS mutation occurred subsequently triggering the change of early into intermediate adenoma. However, in contrast to APC, KRAS can act on nondysplastic ACF precursor lesions. 15 In this study, mutations in the KRAS gene occurred in 14 of 22 samples (63.6%) at 9 codons and were most commonly found in older age group, locally-advanced stage, well-differentiated/low grade, with positive lymphovascular invasion, and located at the rectum. There were differences of codon location in missense mutation between Jakarta (Indonesia) and United States population, i.e. codons 13,14,34,58,59,146 VS 12,13,61,146.22 In addition, nonsense mutations were also found at codon 22 which only occurred in 1 patient. This patient diagnosed with stage 2A (pT3N0M0) undergoing elective curative resection and was given 8 cycles of capecitabine adjuvant chemotherapy with complete response. Mutation located in codon 12 has more aggressive behavior than codon 13, because patients were commonly presented in advanced stage. 22 Nevertheless, number of cases with metastases involving KRAS mutation in this study was found in 3 of 5 samples without involvement of codon 12.

Discussion
KRAS mutation can occur concomitantly with APC mutation leading to increased accumulation of b-cathenin in the cytoplasm, by destroying its binding to E-cadherin, which has actually been increased due to loss of mutated APC degradation function.
This causes the Wnt signal to become more active so that motility and cell invasion are more aggressive. 15,18,21,[23][24][25][26] In CRC, the combination of APC and KRAS mutations (co-occurring mutations) can occur up to 80%, whereas in this study only occurred in 63.6% of subjects. 27 In this study, patients with APC, TP53 and KRAS mutations were predominantly 50 years old, with locally-advanced stage and positive lymphovascular invasion. Two shortest median life expectancy were found in patients with KRAS mutation (Figure 1), in addition 50% of patients died within 6 months after therapy ( Figure 2).
Before turning into carcinoma, intermediate adenomas differentiate into late adenomas triggered by mutations in the SMAD4, CDC4, and DCC genes. 2,7 In this study, we found SMAD4 nonsense and missense mutations in 18 of 22 patients (82%).
In ACS theory, late adenomas who developed into carcinomas have mutations in TP53, TGFBR2, BAX, and IGF2R. Mutated TP53 was found in all subjects in this study in the form of nonsense, missense, frameshift, splice-site, and silent mutation. Five most frequently occurred codon locations in this study were 237, 238, 127, G245S, and R248Q. Those are different compared to world database in The Cancer Genome Atlas Program (TCGA) portal which stated that the ve codon positions with the highest frequency were 175, 282, 248, R273H, and R273C. 28 In contrast to the UK population, in 64% (14 out of 22) subjects, TP53 and KRAS mutations co-occurred. 18,21 In Indian population, these two combinations were only found in 13 of 112 cases, whereas the study by Timar can occur in up tõ 40%. 27,29 TP53 and KRAS activate different carcinogenesis pathways so that they rarely coexist. 30 Similar to APC and TP53, PIK3CA mutations were found in all subjects (100%) with 9 SNVs. PIK3CA has no role in the aggressive behavior of CRC. 31,32 Even so, when it occurs concurrently with KRAS mutations, it will show evident aggressive behavior, especially when occurs in exons 9 and 20. While in this study mutations occurred in exons 2, 3, and 4, aggressive behavior presenting as locally-advanced stage and positive lymphovascular invasion can be found.
Mutations in MLH1 can also occur in non-hereditary/sporadic CRC. The existence of microsatellite instability due to mutations in genes that play roles in the MMR system such as MLH1 actually provides a good prognosis with a higher survival rate. 33 In this study, the group of cases with MLH1 mutations alone, had the highest median life expectancy and had a 30-month survival rate of up to 100%.
Referring to the colorectal neoplasia principle mentioned above, all subjects in this study indeed involved activation of oncogenes (PIK3CA and KRAS) and inactivation of tumor suppressor genes (APC, TP53 and MLH1) and also involved a range of 8-19 mutated genes per person. In this study, mutated APC and KRAS, which supposed to be occurred on early sequence of ACS, supports what Fearon stated about unimportance of mutational sequence in determining tumor biologic behavior. 1,2 We are strongly aware of our study's limitation. Limited number of samples is the lack of this study. Further research is truly required to accomplish complete mapping of Indonesian pro le, especially in investigating our unique ndings in each of genes described.
Nevertheless, this is the rst study fully describes the pro le of pathogenic mutations of CRC in Indonesian population with its unique characteristics, compiled of various ethnicities with diverse diet and lifestyle which may have roles in contributing natures of Indonesian version of CRC presenting in locally-advanced stage with large tumor size and moderate-severe malnutrition status. This study is also the rst in the world to examine the co-occurring mutations of APC, TP53, PIK3CA, KRAS, and MLH1.

Declarations
Authors declare no con ict of interest. Provenance and peer review. Not commissioned, externally peer-reviewed.

Availability of data and materials
The datasets generated and/or analysed during the current study are not publicly available due to presence of individuals' personal details but are available from the corresponding author on reasonable request.

Authors' Contribution
VMGM analysed and interpreted data related to colorectal cancer, gathered information in medical records, and synchronized patients' data with VCF data les, and was a major contributor in writing the manuscript.
LE carried out the molecular genetics study and participated in sequence alignment.
TJML participated in its design and coordination and helped to draft the manuscript.
All authors read and approved the nal manuscript.

Consent for publication
Consent was not required in this study because no details on individuals reported within the manuscript. This was already reviewed by the Ethics Committee mentioned above. Figure 1 Median life expectancy in days