Survival Outcomes of Metastatic Colorectal Cancer Patients in Brunei Darussalam and the Impact of KRAS Mutations

Colorectal cancer (CRC) is the third most common cancer, with rising incidence due to lifestyle and diet. 40% of CRC cases are found to have KRAS mutations. In this study, we investigate the survival outcome of metastatic Colorectal cancer mCRC) patients in Brunei Darussalam restrospectively. Chisquared test was used to compare the survival outcomes of mCRC patients, and Mann-Whitney U test was used to compare the median ages of both groups. Kaplan-Meier survival curves were drawn and logrank test was used to compare the survival outcome between two groups. There was a total of 105 patients with stage IV CRC being treated during the study period. 81.6% (n=62) of mCRC patients were found to have the primary tumours on the left side of the colon. 19 of these 26 (73.1%) mutant KRAS mCRC patients died, while 23 of 50 (46.0%) wild-type KRAS mCRC patients died at the end of the study period, contributing to death rates of 45.2% and 54.8%, correspondingly. 30.3% (n=23) of the study population had a single metastatic site detected (either liver, or lung or any other organs), while 69.7% (n=53) of the 76 mCRC patients had two (double) or more metastatic sites. 69.2% (n=18) and 30.8% (n=8) of the mutant KRAS mCRC patients had mutations within codons 12 and 13, respectively. To our knowledge, this is the first study in Brunei Darussalam to analyse both the survival outcomes of metastatic CRC patients and those of mutant KRAS mCRC patients. Chi-squared analysis showed a significant difference between the survival outcomes of wild-type KRAS and mutant KRAS mCRC patients (p-value = 0.024). There was a significant difference in the survival outcome between the mutant KRAS mCRC patients with RCC and mutant KRAS mCRC with LCC patients. There was no significant difference between the survival outcomes of mutant KRAS patients with mutations in either codon 12 or 13 of the KRAS gene (Table 3). However, there is a significant difference in the median survival periods between the mutant KRAS mCRC patients with mutations in codon 12 and those with mutation in codon 13 of the KRAS gene (p-value = 0.003). In conclusion, we found that mutant KRAS mCRC patients had a significantly poorer OS, which was shown to be worse when the primary tumours were found at the left side of the colon. Mutant KRAS mCRC patients with mutations in codon 12 were found to have shorter survival median periods than those with mutations within codon 13.


INTRODUCTION
In this study, we aim to investigate the survival outcomes of mCRC patients in Brunei Darussalam with regards to the KRAS status of the patient, location of primary tumour and metastasis sites. We also attempt to correlate the various KRAS mutations to the patient survival outcomes in accordance to the patient's gender, age, site of primary tumour and metastatic location. These data may grant insights to the prognosis of patients and influence the treatment guidelines for CRC patients, especially for advanced mCRC patients treated in Brunei Darussalam.

Data collection
This is a retrospective study to investigate the survival outcome of metastatic Colorectal cancer (mCRC) patients in Brunei Darussalam. Ethical approval for the study was obtained from PAPRSB Institute of Health Sciences Medical and Research Ethics Committee (Reference: UBD/HIS/B3/8 dated 19 April 2018). The study was conducted at The Brunei Cancer Centre (TBCC), Brunei Darussalam. Data of all mCRC patients diagnosed between 1 January 2013 and 31 December 2017 were collected and followed up until 30 April 2018. The data collected were age, gender, race, date of diagnosis, site of primary tumour, metastatic sites and the patients' tumour molecular analysis of KRAS mutation status (either KRAS mutated or KRAS wild-type). KRAS mutation status was based on quantitative PCR results testing for missense mutations on codon 12 and 13 of KRAS from CAP accredited diagnostic laboratory in National University Hospital of Singapore. The diagnostic date was the date of colonoscopy or sigmoidoscopy, or other diagnostic technique done. The metastatic sites were classified into either liver, lungs or others (inclusive of bone, brain, bladder and reproductive organs). Exclusion criteria were patients with known NRAS status, uncontactable patients and absence of metastasis. Known NRAS status was excluded to obtain a non-biased population of KRAS. Uncontactable patients refer to patients that were not followed up by the end of the follow up period.

Data analysis
Overall survival (OS) refers to total duration in months between a patients' date of diagnosis to the end of the follow up period or date of termination (date of death). The status of the patients (whether the patient was alive or not at the time of study), gender, age, site of primary tumour, metastasis location, RAS status and mutated codon were documented. Primary tumours located in the ileocecal valve, cecum, ascending colon to the transverse colon was classified as the right sided colon tumour(RCC), while the primary tumours found within the splenic flexure, descending colon, sigmoid and rectum were classified as left sided colon tumours(LCC). All duration of date are recorded in months. To correlate the survival impact of the metastatic site, we have categorised the patients in accordance to the two most common metastatic organs, that is, the liver and the lungs. In addition, the number of metastatic sites were taken into account and therefore, the OS based on either single or multiple metastatic site was studied. Survival analysis of mCRC with mutated KRAS was compared with that of wild-type KRAS. A further inspection on the survival outcome of different KRAS mutations, such as G12D, G12S, G12V and G12C in codon 12 and G13D in codon 13 was done.

Statistical analysis
Chi-squared test was used to compare the survival outcomes of mCRC patients, and Mann-Whitney U test was used to compare the median ages of both groups. Kaplan-Meier survival curves were drawn and logrank test was used to compare the survival outcome between two groups (namely between gender, age groups, location of primary tumour in the colon, and KRAS status). Statistical analysis was conducted using SPSS version 17.0 and R (ver.3.6.0). A p-value of <0.05 is considered statistically significant.

Overall survival (OS)
During the study period, a total of 105 patients with stage IV CRC were treated in TBCC. These patients were initially diagnosed by either sigmoidoscopy/colonoscopic histopathologic biopsy. They were classified as stage IV patients due to presence of metastatic tumour(s). Out of the 105 cases, 29 were excluded from the study due to the following reasons: unknown KRAS status (n = 20), known NRAS status (n = 2), and inability to contact at the end of the follow-up period (n = 7). Hence, 76 metastatic CRC (mCRC) patients were included in this study, and they were followed up for a total of 1951 personmonths. Their median age was 58 years (range = 20 -79 years). (Table 1). At the end of the study period, 55.3% (n=42) of the patients died ( Table 1). The ratio of male mCRC (n=48; 63.2%) to female mCRC patients (n=28; 36.8%) was close to 2:1. At the end of the study period, only 21 (61.8%) of the males mCRC patients and 13 (38.2%) of the female mCRC patients survived. There was no significant difference between survival outcome and gender (p=0.821) ( Table 1). The median OS was 29 months (95% CI 23.7-34.3), with males and females having median survival of 27 months (95% CI 20.3-33.7) and 29 months (95% CI 23.7-34.3), respectively. There was no significant difference (p=0.764) in terms of median survival periods between the two genders ( Figure 2 left panel, Table 2) and also no significant difference in terms of the median survival based on median age distribution of 58 years old (Figure 2 Figure 2 right panel). Only 18.4% (n=14) mCRC cases with primary tumours found in the right side of the colon (abbreviated here as RCC for right-sided colorectal cancer) were recorded, out of which 6 died by the end of the study period, amounting to 14.3% of the deaths (Table 1). We were unable to calculate the median survival period for mCRC patients with RCC because more than 50% of the patients were alive at the end of the study period (  Figure 3 right panel).
The site of primary tumour was also classified either as in the colon, sigmoid colon or rectum. The primary tumours of most mCRC were found in the sigmoid colon (with n=36) constituting 47.4%, followed by rectum (n=21, 27.6 %) and colon (n=19, 25.0 %). There was no significant difference (pvalue = 0.289) with regards to the survival outcomes and sites of primary tumours (  (Table 2), as more than half of the mCRC patients with tumours in colon were alive at the end of follow-up period. Correspondingly, for mutant KRAS mCRC patients, there was no significant difference between these patients with primary tumours located in either colon, sigmoid colon or rectum (p-value=0.080, Table 3). No significant difference was found in the median survival periods between the mutant KRAS mCRC patients with primary tumours in the colon, sigmoid colon or rectum (p-value = 0.348) ( Table 4).
We further investigated the number of metastatic sites detected in these mCRC patients. 30.3% (n=23) of the study population had a single metastatic site detected (either liver, or lung or any other organs), while 69.7% (n=53) of the 76 mCRC patients had two (double) or more metastatic sites (Table 1). mCRC patients with two or more metastatic sites were classified as multiple, regardless of the site of metastasis. There was a statistical significant difference between the number of metastatic sites and survival outcome (p-value = 0.004, Table 1). There was a statistically significance value of p=0.047 between the survival outcomes of mCRC patients with a single metastatic site and mCRC patients with multiple metastatic sites ( Table 2). There was a significant difference between the numbers of metastatic sites and survival outcome (p-value = 0.047, Table 2). At the end of the study period, 7 of the 23 single metastatic site CRC patients died while 35 of the 53 multiple (two or more) metastatic sites CRC patients died, yielding deaths rates of 16.7% and 83.3%, correspondingly (p-value = 0.004, Table 1).
In order to examine the survival outcomes of mutant KRAS mCRC patients (n=26) with varying number of metastatic sites, mutant KRAS mCRC patients were categorised in accordance to the number of metastatic sites. Out of the seven mutant KRAS mCRC patients who survived till the end of the study period, 4 (57.1 %) have one metastatic site while 3 (42.9 %) have two or more metastatic sites (Table  3). Mutant KRAS mCRC patients with multiple metastatic sites were significantly more likely to die by the end of the study period (p= 0.003, Table 3), as compared to mCRC patients with wild-type KRAS.
The median survival period of mutant KRAS mCRC patients with one metastatic site was non-calculable, however, the median survival periods for mutant mCRC patients with two or more metastatic sites was 25 (95% CI: 18.0-32.0) ( Table 4). Survival analysis revealed a significant difference between these two mutant KRAS patient groups (p=0.043).
Two metastatic target organs (liver and lungs) were also examined. 78.9 % (n=60) of the mCRC patients had liver metastasis, while 55.3 % (n=42) had metastasis to the lungs. It should be noted that a mCRC patient may have metastasis to both the liver and the lungs. The survival rate of mCRC patients with liver metastasis was 46.7 % (n=28), whilst the survival rate of mCRC patients with lung metastasis was 30.9 % (n=13). Chi-squared test showed no significant difference between the number of deaths and survival of mCRC patients due to liver metastasis (p-value of 0.512), but a significant difference with p-value of 0.007 was observed between the number of deaths and survival of mCRC patients due to lung metastasis (  (Table  3).

Clinical significance of the specific KRAS mutation
The specific KRAS mutations detected in the mutant KRAS mCRC patients were studied. 69.2% (n=18) and 30.8% (n=8) of the mutant KRAS mCRC patients had mutations within codons 12 and 13, respectively (Tables 3 and 4). There was no significant difference between the survival outcomes of mutant KRAS patients with mutations in either codon 12 or 13 of the KRAS gene (Table 3). However, there is a significant difference in the median survival periods between the mutant KRAS mCRC patients with mutations in codon 12 and those with mutation in codon 13 of the KRAS gene (p-value = 0.003, Table 4). In our study population, the highest KRAS mutation was G12D (n=10) (38.5%), followed by G13D (n=8) (30.8%). There were 11.5 % (n=3) of G12S, followed by 7.7% each of G12V and G12C with 3.8% of G12A.

DISCUSSION
To our knowledge, this is the first study in Brunei Darussalam to analyse both the survival outcomes of metastatic CRC patients and those of mutant KRAS mCRC patients. Although a recent paper published on the survival rates and associated factors of CRC patients in Brunei Darussalam [5], investigation on survival outcomes of mutant KRAS mCRC patients and exploration on the various KRAS mutations in these patients are still in its infancy in this country. The reported median survival period was 57.0 months for CRC [5], while for mCRC patients, we estimated that the median survival period was 29.0 months. Within Asia, India was found to have the lowest survival rate of 31.2% for CRC patients while China has the highest with survival rate of 77% [27,28]. The survival rate of mCRC patients found within this study was 44.7% while the reported five year survival rate for CRC reported by Leong et al was 49.6% [5].There are more cases of male mCRC patients than female mCRC patients and a slightly shorter survival period of the male mCRC patients in our study (Table 1). Prior research has indicated that gender contributes to survival outcomes and females were deemed to have better OS [29], possibly due to the protective effects of female sex hormones against CRC [30,31].
Despite the small population of Brunei Darussalam, Chi-squared analysis showed a significant difference between the survival outcomes of wild-type KRAS and mutant KRAS mCRC patients (p-value of 0.024). This is further corroborated by Kaplan-Meier survival analysis shown in Table 2, whereby there is a statistical significant difference (p-value = 0.017). The poorer survival outcome of mutant KRAS mCRC patients as compared to those of wild-type KRAS has been established [32], as KRAS has been identified as one of the six driver genes from the TCGA database that drives metastatic CRC [33]. Large cohort studies have consistently illustrated that mutant KRAS was associated with metastasis in CRC patients, including lymphatic and distant metastases [33]. Early mutation of KRAS in the adenoma to carcinoma progression confers tumour growth advantage, enabling enhanced growth of the tumour [34]. Also, as Ras is part of Ras-Raf-MEK-ERK signalling pathway, a vital pathway that controls survival and proliferation and is linked to PI3K and AKT pathway that mediates cell death [35,36], a constitutive active Ras drives carcinogenesis. Pioneering work on the immune suppressive role of mutant KRAS (KRASG12D) revealed that mutant KRAS suppressed interferon regulatory factor (IRF2) expression, leading to increased myeloid-derived suppressor cells and poor T cell infiltration [37,38]. All these point to the aggressiveness of mutant KRAS gene in carcinogenesis.
The primary tumour of CRC has been classified as either on the right side of the colon (ileocecal valve, cecum, ascending colon to the transverse colon) or the left side of the colon (the splenic flexure, descending colon, sigmoid and rectum). A further categorisation of the primary tumour location on the left side of the colon refers to the sigmoid colon and rectum. In agreement with Buchwald et al (2018), we found that mCRC patients with primary tumours found in the rectum had better OS than those with primary tumours within the colon (Table 2) [39]. A limitation of our study is the lack of further sorting of the primary tumour, especially for the right side of the colon. There is an unusual high number of left-sided primary tumours (n=62) as compared to right-sided primary tumours (n=14) in these mCRC patients. Although most reports stated that KRAS mutation is associated with rightsided colon primary tumours in CRC [40,41], we found that most of the KRAS mutations in our study population are found within the left-sided colon primary tumours (data not shown). Nevertheless,  (2019) concluded that although KRAS mutations were in high occurrence in RCC than LCC, there was an association of KRAS mutations and poor prognosis in LCC, which was absent in RCC [42].
Increasing number of metastatic sites correlates with poorer survival outcomes in this study (Tables 1,  2 and 3). In Stage IV M1b Non-Small Cell Lung cancer, the number of metastases but not the location has been found to have an impact on the survival outcomes of the patients [43]. On the contrary, the number of metastases has no significant effect on survival in prostate cancer patients [44]. There are currently a few studies debating whether the impact of survival is dependent on the site of metastasis or the number of metastasis [45]. However, the primary tumour site of cancer may affects the survival of patients [46]. In this study, as for median survival periods, 78.9% of mCRC patients were found to have metastasis to the liver, while 55.3% of mCRC patients were found to have metastasis to the lungs. A higher proportion of mCRC patients was found to have metastasis to the liver due to the colon's anatomical situation with regards to the portal circulation [47]. In addition, 81.6% of mCRC patients in this study were found to have LCC and LCC was associated with higher incidence of liver metastasis [48]. Table 2 demonstrated that mCRC patients with metastasis to the lungs had significantly poorer survival outcome compared to mCRC patients with metastasis to the liver. Survival analysis estimated that mCRC patients with liver metastasis had an average of four months more survival period ( Table  2). Majority of the mCRC patients with lungs metastasis (n=42) had metastasis to the liver (n=60) too. Therefore, the number of metastases most likely determines the survival outcome, with metastasis to the lungs being an indicator of worse OS. On the other hand, there were more cases of metastasis to the lungs in mutant KRAS mCRC patients (69.2%), instead of 55.3% in the overall study population of mCRC patients. This is in agreement with the work of Ghidini et al (2016). Compared to wild-type KRAS mCRC patients, mutant KRAS mCRC patients were found to have higher incidence of lung and brain metastases [49]. Taken together, in this study population, the number of metastatic sites accompany worsened OS, with metastasis to the lungs being a predictor of poor OS. Mutation in KRAS gene further worsened the prognosis.
85% of KRAS mutation occurs in codons 12 and 13. Codons G12V and G12D are the most common mutated gene found in KRAS gene in codon 12 [50], accounting for 70% of all KRAS mutation [51]. Various studies have associated specific KRAS mutations with different likelihood of survival [49,51]. Bai et al (2017) determined that G12V and G12D have been associated with an increased risk of CRC associated death [50], while Jones et al (2017) established that G12C and G12V were predictors of worse OS [52]. Between G12D, G12S and G13D KRAS mutations, our study corroborates that mutant KRAS mCRC patients with G12D have the lowest survival median at 23 months (95% CI 15.4-34.6).
There is a lack of data for the median survival periods of mutant KRAS mCRC patients with G12A, G12V and G12C mutations due to the small population. In codon 13, G13D is the most common KRAS mutation [15,50] and this was the second most common specific KRAS mutation after G12D in our study with a survival median period of 29 (95% CI:24.2-33.6) months (Table 4). Similar to most studies, we obtained that mutant KRAS mCRC patient mutation in codon 13 had a comparably better OS than those with mutation in codon 12 [14,50,52].
In conclusion, we found that mutant KRAS mCRC patients had a significantly poorer OS, which was shown to be worse when the primary tumours were found at the left side of the colon. Mutant KRAS mCRC patients with mutations in codon 12 were found to have shorter survival median periods than those with mutations within codon 13.

Authors' Contributions
R.P. conceived the work. D.D, R.P., T.P.U. and S.K.L acquired data. R.P D.D., L.C. and Y.C.L analysed the data. D.D. wrote the first draft of the manuscript. D.D., R.P., L.C, S.K.L, Z.H.L., K.K and Y.C.L contributed towards the discussion and revision of the manuscript. All authors read and agreed to the final version of the manuscript.