Prevalence and coexistence of KRAS,BRAF,PIK3CA,NRAS,TP53,and APC mutations in Indian colorectal cancer patients: Next-generation sequencing

Abstract

Colorectal cancer incidences are on a rise in India. In this study, we have analyzed the mutation frequencies of six potential biomarkers, their coexistence, association with clinicopathological characteristics, and tumor location in Indian colorectal cancer patients. Next-generation sequencing was performed to identify mutations in the six potential biomarker genes using formalin-fixed paraffin-embedded tissue blocks of 112 colorectal cancer patients. The mutation frequency observed in KRAS, BRAF, PIK3CA, NRAS, TP53, and APC was 35.7%, 7.1%, 16.1%, 6.3%, 39.3%, and 29.5%, respectively. The significant associations of mutations were KRAS with age less than 60 years (p = 0.041), PIK3CA with males (p = 0.032), tumor stage I–II (p = 0.013), lack of metastasis in lymph nodes (p = 0.040), NRAS with rectum (p = 0.002), and APC with T2 stage of tumor growth (p = 0.013). No single patient harbored mutations in these six genes or any five genes simultaneously. Significance was noted in coexistence of KRAS with APC (p = 0.024) and mutual exclusion of KRAS with BRAF (p = 0.029). PIK3CA exon 9 was observed to be more frequently associated with KRAS mutations than PIK3CA exon 20 (p = 0.072). NRAS mutations were mutually exclusive with BRAF and PIK3CA mutations. As per our knowledge, this is the first next-generation sequencing–based biomarker study in Indian colorectal cancer patients. Frequent coexistence of gene mutations in pairs and triplets suggests that synergistic effect of overlapping mutations might further trigger the disease. In addition, infrequent coexistence of multiple gene mutations hints toward different signaling pathways for colorectal cancer tumorigenesis.

Keywords

Colorectal cancer KRAS BRAF PIK3CA NRAS TP53 APC mutation

Introduction

Colorectal cancer (CRC) has been identified as the third most common cancer worldwide, which accounted for 1.4 million cases in 2012.¹ In Asia, CRC incidences have been rapidly increasing in most of the countries, specifically China, Japan, Singapore, and South Korea.² India is among the lowest risk of CRC in Asia where colon cancer ranks eighth in men and ninth in women, whereas rectal cancer ranks ninth among men and does not figure in the top 10 cancers in women.³ Multiple signaling pathways are associated with transition of CRC from adenomatous polyp into adenocarcinoma, which includes Wnt-β catenin pathway, MAPK/ERK pathway, PI3K/Akt pathway, and p53 pathway. The cumulative effect of activation and inactivation of numerous oncogenes and tumor suppressor genes, respectively, deregulates these signaling pathways and leads to CRC. This study includes six frequently mutated genes in CRC: KRAS, BRAF, NRAS (MAPK/ERK pathway), PIK3CA (PI3K/Akt pathway), TP53 (p53 pathway), and APC (Wnt-β catenin pathway). Mutations in these six genes have been extensively studied worldwide due to the high frequency found in CRC patients, and this study investigates these mutations with respect to Indian population. Several CRC-targeted therapies and prognosis testing are based on mutations in the signaling genes KRAS, BRAF, NRAS, and PIK3CA. In addition, TP53 and APC mutations have been associated with most human cancers and play a crucial role in CRC tumorigenesis.

KRAS (Kirsten rat sarcoma viral oncogene homolog) gene is the most commonly altered gene in the RAS family, which is found mutated in 17%–25% of all cancers and 35%–45% of CRCs.⁴ Mutations in KRAS are highly frequent in the codons 12, 13, 59, 61, and 146.^5,6 Recently, KRAS mutation has been elucidated as a negative predictive biomarker used to detect resistance to the anti-epidermal growth factor receptor (EGFR) monoclonal antibody therapy.⁷ BRAF (v-raf murine sarcoma viral oncogene homolog B1) gene mutations are persistent in 15% of all cancers, and the V600E mutation is present in 10% of all CRCs.^8,9 A significantly higher frequency of BRAF mutations has been reported in females and in the proximal colon.^10–13 Furthermore, KRAS and BRAF mutations are observed to be nearly mutually exclusive of each other.^11,12,14,15

PIK3CA (Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha) gene is involved in PI3K/Akt pathway and is found mutated in approximately 10%–20% CRCs where nearly 80% mutations are found in two hotspot regions: exons 9 and 20.¹⁶ PIK3CA mutations are more frequent in the proximal colon.^17,18 NRAS (neuroblastoma RAS viral oncogene homolog) is the second most important RAS protein involved in CRC, after KRAS.¹⁹ NRAS is generally found mutated in 15% of all cancers²⁰ and 5% of all CRCs.¹⁹ NRAS mutations have been previously reported to have a strong association with KRAS wild type (p = 0.001) and are generally found to be mutually exclusive with KRAS mutations.^21,22

The tumor suppressor (TP53) gene regulates DNA repair mechanism and apoptosis. The TP53 mutation is observed in nearly 50% CRCs, with the most frequently observed mutations between exons 5 and 8 at codons 175, 213, 245, 248, 273, and 282.^23,24 In Asia, the p53 mutation frequency varies between 23% and 52.6%.^25–28 TP53 was observed to be frequently mutated with APC.^29,30 Adenomatous polyposis coli (APC), a tumor suppressor gene involved in Wnt-β catenin signaling pathway, is found mutated in nearly 80% sporadic CRCs.³¹ The mutation cluster region (MCR) is present in exon 15 between codons 1286 and 1513. Higher frequency of APC mutations has been reported in distal colon than proximal colon.^32,33

Materials and methods

Patients and specimens

This study was performed on Indian patients undergoing surgery for CRC or diagnosed with metastatic CRC at Sir Ganga Ram Hospital, Delhi, India, between March 2013 and May 2015. Tumor samples were obtained from 146 patients; however, 34 tumor samples were filtered out during tumor content estimation, DNA extraction, and next-generation sequencing (NGS). The remaining 112 CRC patients with confirmed histopathology or biopsy of CRC were included in this study. Clinical characteristics of all patients were recorded including age, gender, tumor location, tumor size, Tumor, Lymph-node, Metastasis (TNM) stage, distant metastasis in liver, or other parts. Follow-up end point was December 2015.

For all these patients, formalin-fixed paraffin-embedded (FFPE) tissue was used for DNA isolation. All the FFPE tissue blocks were taken from the Department of Histopathology, Sir Ganga Ram Hospital, and were checked thoroughly for tumor content. FFPE tissue blocks with more than 30% tumor content were taken for the study. Five to eight sections of 10 µm were used for DNA extraction using microtome. The DNA was isolated using Qiagen AllPrep DNA FFPE Kit (Cat No./ID: 80234, Qiagen) as per manufacturer instructions.

Library preparation and sequencing

DNA was quantified for double stranded DNA (dsDNA) contents using QuBit (Thermo Fisher, USA), and protein contamination was checked by measuring the 260:280 ratio using Nanovue (Agilent, USA). Samples with 260:280 ratio within the range 1.7–2.0 were qualified. The DNA was also further qualified by running the Illumia Infinium assay. Samples with delCT value ⩽2 were qualified for further processing. As recommended by the manufacturer, 200–300 ng of dsDNA was used for generating sequencing libraries using TruSeq Amplicon cancer panel (Cat.No.FC-130-1008, Illumina, USA). The panel includes 212 amplicons of 170–190 bp length, targeting hotspot mutation in 48 cancer-related genes. DNA libraries were denatured and diluted as per Illumina protocol. Samples were run on MiSeq machine using MiSeq Reagent Kit V2 (Cat.No.MS-102-2002, Illumina).

Data analysis

The trimmed FASTQ files were generated using MiSeq Reporter (v2.6.2) from Illumina. The reads were aligned against the whole genome build hg19 in MiSeq Reporter. Subsequent analysis was done using Strand NGS v2.1.6. Reads that failed vendor QC, reads with average quality less than 20, reads with alignment score below 90, and reads with ambiguous characters were all filtered out. The Strand^® NGS variant caller was used to detect variants at locations in the target regions covered by a minimum of 10 reads with at least two variant reads. Variants with a decibel score of at least 50 were reported. Substitution variants with a StrandBias more than 100% and InDel variants in homopolymer stretches more than 4 with supporting reads less than 10% were filtered out. The variants were further analyzed for their effect on the protein using the “SNP Effect Analysis” functionality in Strand NGS. This provides the effect with respect to each valid transcript. The variants were further annotated with COSMIC (Catalogue of Somatic Mutations in Cancer) database (v71) to provide information on whether that variant has been observed before in the context of any tumor type. Samples were randomly chosen and validated through Sanger sequencing, which were found to be 100% concordant with NGS analysis.

Statistical analysis

The data were analyzed using the SPSS 17.0 statistical software. Individual information and baseline characteristics were summarized using descriptive characteristics. Continuous variables were presented as mean ± standard deviation (SD). Categorical variables were expressed as frequency and percentage. The Pearson’s chi-square test or the chi-square test of association was used to determine the relationship between two categories, which included comparison of different clinicopathological characteristics with gene mutations and coexistence of two gene mutations. The p-value less than 0.05 was considered to be statistically significant.

Results

In this study, 112 CRC samples were examined for clinicopathological characteristics related to the patients and the tumors (Table 1). CRC was more frequently observed in males (60.7%) than females (39.3%). In addition, a higher proportion of patients were observed to be aged less than 60 years (65.2%), with the median age of 53 years ranging from 23 to 86 years. Primary tumors were significantly higher in colon (82.1%) than rectum (17.9%), where the tumors were equally distributed in proximal (right) colon and distal (left) colon, accounting for 41.1% each. Cecum was included within the proximal colon while sigmoid colon was included in the distal colon. The size of tumor at the time of surgery was reported to be more than or equal to 5 cm in 77 (68.8%) patients. A higher proportion of patients were observed to have tumor stage III (40.2%), followed by stage IV (36.6%) and stage I-II combined (23.2%). The N0, N1, and N2 stages of lymph node invasion were observed to be similar. However, a noteworthy variation was observed in the depth of invasion of tumor in the layers of colon/rectum, where T3 represented 67% of all cases, followed by 25.9% cases in T4, 7.1% cases in T2, and no case observed in T1. Finally, presence and absence of distant metastasis were similar (51.8% vs 48.2%) in the study population.

Table 1.

Clinicopathological data of 112 Indian CRC patients.

Characteristics	No. of patients (%)
Total patients	112
Gender
Male	68 (60.7)
Female	44 (39.3)
Age (years)
<60	73 (65.2)
⩾60	39 (34.8)
Median	53
Range	23–86
Tumor site
Colon	92 (82.1)
Left colon	46 (41.1)
Right colon	46 (41.1)
Rectum	20 (17.9)
Size of tumor (cm)
<5	35 (31.2)
⩾5	77 (68.8)
Tumor stage
I-II	26 (23.2)
III	45 (40.2)
IV	41 (36.6)
No. of nodes
N0 (0)	34 (30.4)
N1(1–3)	41 (36.6)
N2 (>3)	37 (33.0)
Depth of invasion
T1	0 (00.0)
T2	8 (07.1)
T3	75 (67.0)
T4	29 (25.9)
Distant metastasis
Yes	58 (51.8)
No	54 (48.2)

CRC: colorectal cancer.

The most frequently mutated gene was TP53 (48 mutations), followed by KRAS (43 mutations), APC (39 mutations), and PIK3CA (19 mutations). BRAF and NRAS show the lowest case of gene mutations, with eight mutations each. Tumor profiling of CRC patients was performed in parallel to genetic alterations (Table 2). KRAS mutation was significantly linked to age, where age less than 60 years was more prevalent (p = 0.041). PIK3CA mutations were predominantly harbored by male patients (p = 0.032) and were likely to be associated with tumor stage I-II (p = 0.013) and lack of metastasis in lymph nodes (p = 0.040). NRAS mutations were frequently observed in rectum over colon (p = 0.002). Finally, occurrence of APC mutations was associated with T2 stage of tumor growth in the walls of large intestine (p = 0.013), which refers to spread of tumor into the muscle layer of colon/rectum.

Table 2.

Correlation of KRAS, BRAF, PIK3CA, NRAS, TP53, and APC gene mutations with respect to clinicopathological data of the patient cohort.

Characteristics	No. of patients		KRAS			BRAF			PIK3CA			NRAS			TP53			APC
Characteristics	N = 112	(%)	n	(%)	p-value	n	(%)	p-value	n	(%)	p-value	n	(%)	p-value	n	(%)	p-value	n	(%)	p-value
No. of patients			40	35.7		8	7.1		18	16.1		7	6.25		44	39.3		33		29.5
Total mutations			43			8			19			8			48			39
Gender
Male	68	60.7	25	36.7	0.773	7	10.3	0.107	15	22.1	0.032*	2	2.9	0.109	25	36.8	0.497	20	29.4	0.988
Female	44	39.3	15	34.1		1	2.3		3	6.8		5	11.4		19	43.2		13	29.5
Age (years)
<60	73	65.2	31	42.5	0.041*	3	4.1	0.088	13	17.8	0.494	4	5.5	0.693	27	37	0.495	23	31.5	0.517
⩾60	39	34.8	9	23.1		5	12.8		5	12.8		3	7.6		17	43.6		10	25.6
Tumor site
Colon	92	82.1	35	38	0.27	6	6.5	0.632	16	17.4	0.521	2	2.2	0.002*	33	35.9	0.112	24	26	0.093
Rectum	20	17.9	5	25		2	10		2	10		5	25		11	55		9	45
Colon side
Distal (left)	46	41.1	16	34.8	0.519	3	6.5	1.00	6	13	0.271	2	4.3	0.495	20	43.8	0.128	11	23.9	0.635
Proximal (right)	46	41.1	19	41.3		3	6.5		10	21.7		0	0		13	28.3		13	28.3
Size of tumor (cm)
<5	35	31.2	13	37.1	0.832	3	8.57	0.692	6	17.1	0.835	1	2.8	0.431	12	34.3	0.465	9	25.7	0.557
⩾5	77	68.8	27	35.1		5	6.5		12	15.6		6	7.8		32	41.6		24	31.1
Tumor stage
I-II	26	23.2	11	42.3	0.677	3	11.5	0.312	9	34.6	0.013*	1	3.8	0.214	10	38.5	0.364	6	23.1	0.622
III	45	40.2	16	35.6		4	8.9		5	11.1		5	11.1		21	46.7		13	28.9
IV	41	36.6	13	31.7		1	2.4		4	9.8		1	2.4		13	31.7		14	34.1
No. of nodes
N0 (0)	34	30.4	14	41.2	0.328	3	8.8	0.773	10	29.4	0.040*	1	2.9	0.627	11	32.4	0.455	7	20.6	0.391
N1 (1–3)	41	36.6	11	26.8		2	4.9		4	9.8		3	7.3		19	46.3		14	34.1
N2 (>3)	37	33	15	40.5		3	8.1		4	10.8		3	8.1		14	37.8		12	32.4
Depth of invasion
T1	0	0	0	0	0.83	0	0	0.107	0	0	0.561	0	0	0.518	0	0	0.686	0	0	0.013*
T2	8	7.1	3	37.5		2	25		1	12.5		0	0		2	25		6	75
T3	75	67	28	37.3		5	6.6		14	18.7		6	8		30	40		20	26.7
T4	29	25.9	9	31		1	3.4		3	10.3		1	3.4		12	41.4		7	24.1
Metastasis
Yes	58	51.8	20	34.4	0.778	2	3.4	0.116	6	10.3	0.087	3	5.2	0.71	21	36.2	0.489	17	29.3	0.971
No	54	48.2	20	37		6	11		12	22		4	7.4		23	42.6		16	29.6

Significant values.

KRAS mutations

KRAS mutations were observed in 40/112 (35.7%) CRC patients, giving rise to a total of 43 mutations. One patient harbored three KRAS mutations (G12D, G10R, and E49K), one patient harbored two KRAS mutations (G12D and E63K), while the rest harbored one KRAS mutation each. A higher frequency of KRAS mutations was observed in male patients (25/40, 62.5%) than female patients (15/40, 37.5%) and in proximal colon (19/40, 47.5%) when compared to distal colon (16/40, 40%). The detailed distribution of KRAS mutations has been given in Table 3. Codons 12 and 13 together constituted 35/43 (81.4%) of all KRAS mutations. All KRAS mutations comprised of missense substitution. The three most frequent mutations were substitution of glycine with aspartate or valine at codon 12 and substitution of glycine with aspartate at codon 13. These three mutations, G12D, G12V, and G13D, occur due to substitution at c.35G>A, c.35G>T, and c.38G>A, and have been observed in 15/40 (37.5%), 6/40 (15%), and 7/40 (17.5%) patients who were diagnosed with KRAS mutation, respectively. Other common mutations include G12C, G12S, Q61H, and A146T, which have been identified in 4/40 (10%), 2/40 (5%), 2/40 (5%), and 2/40 (5%) patients with a KRAS mutation. Finally, A59T, E49K, E63K, G10R, and G13S were identified in one patient only.

Table 3.

Mutation spectrum of KRAS, BRAF, PIK3CA, NRAS, TP53, and APC genes.

Mutation	Nucleotide change	No. of mutations (%)
KRAS		43
Codon 12		27 (62.9)
G12D	G>A	15 (34.9)
G12V	G>T	6 (14.0)
G12C	G>T	4 (09.3)
G12S	G>A	2 (04.7)
Codon 13		8 (18.6)
G13D	G>A	7 (16.3)
G13S	G>A	1 (02.3)
Codon 61
Q61H	A>C	2 (04.7)
Codon 146
A146T	G>A	2 (04.7)
Other codons^a		4 (09.3)
BRAF		8
V600E	T>A	6 (75.0)
G596D	G>A	1 (12.5)
L588R	T>G	1 (12.5)
PIK3CA		19
Exon 9		7 (36.8)
E545A	A>C	3 (15.8)
E545K	G>A	3 (15.8)
E542K	G>A	1 (05.3)
Exon 20		5 (26.3)
H1047R	A>G	3 (15.8)
Y1021C	A>G	1 (05.3)
Y1021H	T>C	1 (05.3)
Other exons^b		7 (36.8)
NRAS		8
Codon 12		3 (37.5)
G12C	G>A	2 (25.0)
G12D	G>A	1 (12.5)
Codon 60
G60E	G>T	2 (25.0)
Other codons^c		3 (37.5)
TP53		48
Codon 175
R175H	G>A	3 (06.3)
Codon 213
R213*	C>T	3 (06.3)
Codon 245
G245S	G>A	4 (08.3)
Codon 248		4 (08.3)
R248Q	G>A	2 (04.2)
R248W	C>T	2 (04.2)
Codon 273		7 (14.6)
R273C	C>T	5 (10.4)
R273H	G>A	2 (04.2)
Codon 282
R282W	C>T	2 (04.2)
Other codons^d		25 (52.1)
APC		39
Codon 244		1 (02.6)
Codon 876		3 (07.7)
Codons 1114		1 (02.6)
Codons 1306–1389		17 (43.6)
Codons 1429–1488		16 (41.0)
Codon 1556		1 (02.6)

Other codons include 10, 49, 59, and 63.

Other exons include 1, 4, 7, and 10.

Other codons include 18, 49, and 59.

Other codons include 97, 106, 131, 133, 179, 181, 195, 208, 209, 220, 229, 237, 241, 242, 244, 267, 272, 275, 278, and 342.

BRAF mutations

BRAF mutations were comparatively uncommon in CRC patients (8/112, 7.14%). A single BRAF mutation was observed in every patient. All BRAF mutations were observed to be mutually exclusive with KRAS mutations. BRAF mutations were equally distributed in proximal and distal colons (3/8, 37.5% each) but were observed to be more frequent in male patients (7/8, 87.5%) than female patients (1/8, 12.5%). All eight mutations comprised of missense substitution. The most frequent BRAF mutation in our study was V600E, which occurs due to substitution at c.1799T>A. This mutation was observed in 6/8 (75%) patients who were diagnosed with a BRAF mutation. Other two mutations, G596D and L588R, were identified in one patient each (Table 3).

PIK3CA mutations

Mutated PIK3CA was identified in 18/112 (16.1%) CRC patients. Each patient harbored one PIK3CA mutation except for one patient with two PIK3CA mutations (R88Q, H1047R). A higher frequency of PIK3CA mutations was observed in male patients (15/18, 83%) than female patients (3/18, 17%) and in proximal colon (10/18, 56%) than distal colon (6/18, 33%). Majority PIK3CA mutations were observed in exon 9 (7/19, 36.8%), exon 20 (5/19, 26.3%), and exon 1 (4/19, 21%). The remaining three exons (exon 4, exon 7, and exon 10) harbored only one PIK3CA mutation each (1/19, 5.3%; Table 3). Every PIK3CA mutation was identified to be missense substitution type. The three most frequent mutations observed were E545A, E545K, and H1047R which occur due to substitution at c.1634A>C, c.1633G>A, and c.331A>G, respectively. These three mutations individually represent 3/19 (15.8%) each and they together constitute 9/19 (47.3%) of all PIK3CA mutations. Other mutations were observed only in one patient, which include E418K, E542K, G106V, K111E, K111N, N345S, R555K, R88Q, Y1021C, and Y1021H.

NRAS mutations

The incidence rate of NRAS mutations was 7/112 (6.25%) in CRC patients. Only one patient harbored two NRAS mutations (G60E, T58I) simultaneously. NRAS mutations were frequent in female patients (5/7, 71.4%) than male patients (2/7, 28.6%) and in rectum (5/7, 71.4%) than colon (2/7, 28.6%). The two colon mutations were situated in the distal side. Coexistence of NRAS with KRAS was observed in one case. NRAS mutations mainly belonged to codon 12 (37.5%) and codon 60 (25%) which together comprised 55.6% of all NRAS mutations. One NRAS mutation was observed each in codons 18, 49, and 59 (Table 3). No mutations were observed in codon 13. All eight NRAS mutations were categorized under missense substitution. The two most frequent mutations observed were G12C and G60E which occur due to substitution at c.179G>A and c.34G>T, respectively, and have been identified in two patients each. Other mutations (A59T, E49K, G12D, and T58I) were observed only in one patient.

TP53 mutations

The frequency of TP53 mutations was 44/112 (39.3%) in CRC patients, which harbored a total of 48 TP53 mutations. One patient harbored three TP53 mutations (C275Y, V97I, R175H), two patients harbored two TP53 mutations each (R175H, N131K; C229R, R213*), while the rest harbored one TP53 mutation each. The TP53 mutations were observed to be higher in distal colon (20/44, 45.5%) over proximal colon (13/44, 29.5%) and in male patients (25/44, 56.8%) over female patients (19/44, 43.2%). The six hotspots for TP53 mutations—codons 175, 213, 245, 248, 273, and 282—individually accounted for 3, 3, 4, 4, 7, and 2 TP53 mutations, respectively (Table 3). These codons together constituted 23/48 (47.9%) of all TP53 mutations. Missense substitution mutation (45/48, 93.8%) were most common, followed by nonsense substitution mutation (2/48, 4.2%) and frameshift deletion mutation (1/48, 2.1%). The most frequent TP53 mutations were R273C, G245S, R213*, and R175H, that individually account for five, four, three, and three cases, respectively. These four TP53 mutations occur due to substitution at c.817C>T, c.733G>A, c.637C>T, and c.524G>A, respectively, and together constitute 15/48 (31.25%) of all TP53 mutations. Furthermore, six TP53 mutations (C229R, R248Q, R248W, R273H, R282W, and S106R) were observed in two patients each. The remaining TP53 mutations were identified in only one patient, which include C242F, C242G, C275F, C275Y, D208V, G244S, H179Y, I195T, M133K, M237I, M237K, N131K, P278S, R181C, R209fs*6, R267L, R342*, S241F, V272E, V97I, and Y220C.

APC mutations

APC mutations were persistent in 33/112 (29.5%) CRC patients, contributing toward 39 mutations. One patient harbored three APC mutations (S1389FS*26, L1488FS*19, and S1465FS*3), four patients harbored two mutations each (R1432*, R1450*; P1373S, R1450*; K1308*, R1450*; and S1355P, A1366V), while rest harbored one mutation each. Majority of the APC mutations were identified between the codons 1306 and 1556 (Table 3). APC mutations were highly frequent in male patients (20/33, 60.6%) over female patients (13/33, 39.4%) and a bit high in proximal colon (13/24, 54.2%) over distal colon (11/24, 45.8%). The APC mutations largely comprised of substitution mutations (27/39, 69.2%) followed by frameshift mutations (12/39, 30.8%). In substitution mutations, the most common mutation was nonsense (22/27, 81.5%) followed by missense (5/27, 18.5%) whereas deletion (11/12, 91.6%) and insertion (1/13, 8.3%) were observed in frameshift mutations. The most common APC mutation, R1450*, is a nonsense mutation in which substitution takes place at c.4348C>T and was observed in 7/39 (17.9%) cases. S1465FS*3, a frameshift mutation that occurs due to deletion of c.4393_4394delAG was observed in 5/39 (12.8%) cases. Further, three nonsense APC gene mutations (R876*, E1322*, and E1379*) were identified in three cases each. Mutations P1373S and Q1367* were observed in two cases each. Finally, some mutations were identified in only one patient, which include A1366V, E1306*, E1309FS*4, G244D, K1308*, L1488FS*19, Q1429*, Q1447*, R1114*, R1432*, S1346*, S1355P, S1389FS*26, and T1556fs*3.

Coexistence of two mutations

The coexistence of two gene mutations is highly fluctuating (Figure 1). The significant coexistence of KRAS and APC is reported in 18 cases (p = 0.024). Also, a significance is noted in mutual exclusion of KRAS with BRAF mutation (p = 0.029). Other frequently coexisting mutations include KRAS and TP53 (13 cases), TP53 and APC (10 cases), KRAS and PIK3CA (9 cases), PIK3CA and TP53 (6 cases), and PIK3CA and APC (6 cases). NRAS mutations were mutually exclusive with BRAF and PIK3CA mutations. A nearly significant result was found in association of KRAS and PIK3CA mutations where PIK3CA exon 9 was highly associated with KRAS mutations while PIK3CA exon 20 was not (p = 0.072).

Figure 1.

Frequency of six gene mutations with combination of double mutations.

Coexistence of three and more mutations

The correlation between different gene mutations is given in Figure 2. The six genes included in this study were not found to be mutated simultaneously in a single patient. Also, no patient harbored a mutation in any five genes. The four mutations, KRAS, PIK3CA, TP53, and APC, coexisted in two cases while KRAS, NRAS, TP53, APC and BRAF, PIK3CA, TP53, APC coexisted in one case. In addition, coinciding KRAS, TP53, and APC mutations were found in four cases. Three cases were observed for co-mutation of three genes—KRAS, PIK3CA, and APC—two cases were observed for co-mutation of each of the three genes—KRAS, PIK3CA, TP53 and PIK3CA, TP53, APC. One case was observed for co-mutation of each of the three genes—KRAS, NRAS, TP53; KRAS, NRAS, APC; BRAF, PIK3CA, APC; BRAF, TP53, APC; and NRAS, TP53, APC. No case was observed with co-mutation of BRAF, PIK3CA, and TP53.

Figure 2.

The interrelationship between occurrence of gene mutations: (a) KRAS, PIK3CA, TP53, APC; (b) BRAF, PIK3CA, TP53, APC; and (c) KRAS, NRAS, TP53, APC.

Discussion

In this study, CRC was observed to be higher in male patients (60.7%) than female patients (39.3%) which corresponds to the results reported by studies worldwide.^1,34 The cancer was observed predominantly in colon (82.1%) over rectum (17.9%) where colon cancer was equally distributed between proximal and distal colon. Previously, some studies have reported a higher rate of CRC incidence in proximal colon^35–37 whereas others report a higher rate in distal colon.^33,38 We also observed KRAS mutations to be significantly associated with younger population, with age below 60 years (p = 0.041). Previous literature reports either no significance of KRAS mutations with respect to age or higher rate in patients above 50 years.^14,39,41 Other significant values were observed in PIK3CA mutations, which were highly frequent in males (p = 0.032), tumor stage I-II (p = 0.013), and not associated with lymph nodes involvement (p = 0.040). Earlier, no significant association was observed between PIK3CA mutations and gender or tumor stage.^18,39,41 NRAS mutations were significantly linked to rectum than colon (p = 0.002); however, no such observance was made in the past.²¹ Also, APC was observed to be associated with T2 stage of tumor growth in the walls of large intestine (p = 0.013).

In this study, KRAS was found mutated in 35.7% CRC patients. This is in accordance with the average level of KRAS mutations (35%–45%) found in CRCs⁴ and the mutation rates observed in Serbia, Turkey, and Saudi Arabia (32%–34.7%); lower than the mutation rates observed in Iraq, Spain, and Slovenia (46.2%–48%); and higher than the mutation rates observed in Greece and Morocco, and previously reported rate in India (23.5%–29.3%).^14,15,38 We observed KRAS mutations to be more frequent in male patients than female patients (62.5% vs 37.5%). Earlier, a even distribution of KRAS mutations has been reported in both genders.^18,39,41 Also, KRAS mutants were frequent proximal colon (47.5%) than distal colon (40%), which is similar to previous studies that reported a high frequency in proximal colon than distal colon (60%, 40%,¹⁰ respectively, and 38%, 30%,⁸ respectively). The frequency of KRAS mutations at codons 12 and 13 were observed to be 24% and 7.1%, respectively, which is lower than the previous Indian study (47% and 19%, respectively).⁴² KRAS was chiefly observed to be mutated in codons 12 and 13, which together constituted 81.4% of all KRAS mutations. Previous studies have reported similar inference (81% and 90%, respectively).^11,43 The three most frequent mutations observed in this study, Gly12Asp, Gly12Val, and Gly13Asp (32.6%, 11.6%, and 16.3% of all KRAS mutations, respectively), are similar to the earlier reported mutation frequencies in Germany (36%, 21.8%, and 18.8% of all KRAS mutations, respectively).⁴⁴

We identified BRAF mutations in 7.14% CRC patients, which is similar to Asian reports (5.6%)³⁷ and other countries including Spain, Turkey, Morocco, Slovenia, and China (5.1%–6.2%) but lower than the mutation rates observed in Serbia (17.8%).¹⁵ Earlier, BRAF has been reported to be mutated in 9.8% CRC patients in India.¹⁴ BRAF mutations were proportional in proximal and distal colons (6.5% each) and more frequent in male patients (10.3%) over female patients (2.3%), concordant to the previous data.^6,13 However, some studies have reported higher BRAF mutations in proximal colon and in females.^10–12,14 We observed only eight cases of BRAF mutations, which might be the reason behind this disparity. The V600E mutation alone constituted 75% of all BRAF mutations. This is comparable to generally reported results (i.e. 80%).⁸ In Turkey, Spain, Slovenia, Serbia, and past Indian study, all BRAF mutations were reported to be V600E.^14,15

The mutation frequency of PIK3CA was observed to be 16.1% in this study which falls within the range reported in CRCs (10%–20%).¹⁶ This PIK3CA mutation rate is comparable to the mutation rates observed in Italy, France, Japan, and Australia (14%–17.8%), lower than the mutation rates in south Italy (28%) and higher than the mutation rates observed in Singapore, France, China, Poland, Switzerland, and past Indian study (2.2%–10.1%).^{14,17,39,41,45–47} PIK3CA mutations were higher in male patients (22.1%) than female patients (6.8%). Previous studies have either reported more PIK3CA mutations in females⁴¹ or a even distribution between males and females.^18,55 Also, PIK3CA mutations have been reported to be evenly distributed within tumor site;¹⁸ however, we observed higher mutation frequency in proximal colon (21.7%) over distal colon (13%). Together, exon 9 and exon 20 were observed to account for 63.1% of all PIK3CA mutations, which is less than earlier reports (i.e. nearly 80%).¹⁶ The PIK3CA mutation frequency at exon 9 (6.25%) and exon 20 (4.5%) is comparable to the Poland study (4.5% and 5.1%, respectively).⁴⁷

In this study, NRAS was mutated in 6.25% CRC tumors, similar to the mutation rates observed in Asia (i.e. 7.2%) and worldwide (i.e. 5%).^19,20 A higher mutation rate has been reported in Greek and Romania (9.75%)⁴⁸ and a lower rate in Japan, China, Belgium, and America (2.2%–4.2%).^18,41,45,46 Previous literature reports a even distribution of NRAS mutations with respect to gender and tumor location.^18,22 However, we observed NRAS mutations to be more frequent in female patients (11.4% vs 2.9%) and all the three NRAS mutations identified in the colon were located distally. Previously, NRAS mutations were reported to be predominant in codon 61 over codons 12 and 13.⁴⁹ In this study, codon 12 represented 33.3% of all NRAS mutations; however, no mutations were observed in codons 61 or 13.

A mutant form of TP53 was persistent in 39.3% CRC patients. This result is in accordance with the frequencies reported by Sweden, Italy, Ireland, and Netherlands (38%–41%); higher than mutation rates observed in Taiwan, Switzerland, South Korea, Australia, and previously reported rate in India (20%–36%); and lower than mutation rates observed in Austria, the United Kingdom, Scotland, Japan, the United States, China, France, Spain, Poland, and Norway (45%–84%).^{25,29,30,38,50–53} Similar to the reported literature, TP53 mutations were more frequent in male patients and in distal colon.^33,54,55 The mutation frequency in codons 175, 213, 245, 248, 273, and 282, observed in our study (6.25%, 6.25%, 8.3%, 8.3%, 14.6%, and 4.2%, respectively), can be compared to previous reports (9%, 3%, 4%, 9%, 8%, and 5%, respectively).⁵³ TP53 mutation in these six hotspots codons together accounted for 47.9% of all TP53 mutations.

The frequency of APC mutations was observed to be 29.5% in this study. This mutation rate is comparable to Korea, Netherlands, the United States, and China (30%–37.3%); lower than the mutation rates observed in Norway, France, the United Kingdom, Germany, Japan, Scotland, and past Indian study (46.7%–66%); and higher than the mutation rates observed in Iran, Tunisia, and Hungary (20.8%–23.3%).^29,38,56,57 APC mutations were observed to be evenly distributed between proximal and distal colons (28.3% vs 23.9%) as well as male and female patients (29.4% vs 29.5%). Similar results have been reported by previous literature.^29,32,33 Majority APC mutations were located between the codons 1306 and 1556. This is in accordance with study held by Attard et al.⁵⁸ that concluded that Asians have a high frequency of APC mutations between codons 1099 and 1694. There was one case with mutation at codon 1309, which is the hotspot for sporadic CRC in Chinese population.³⁶

This study provides a comprehensive evaluation of six gene mutations in Indian CRC patients, upgrading the number of genes examined, methodology used, and gene fragments previously studied in Indian population.^14,38,42 Mutual exclusion of simultaneous mutations in six or any five genes and rare combination of gene mutation in quadruplets suggests independent pathways for CRC tumorigenesis. Only two coexisting mutations were found to be significantly related. First, BRAF was significantly associated with KRAS wild type (p = 0.029). This mutual exclusion of KRAS and BRAF is supported by previous literature.^11,12,14,15 Second, KRAS and APC were found to significantly coexist in 18 cases (p = 0.024). Such relevance has not been reported earlier to the best of our knowledge. Also, NRAS and KRAS mutations were observed to be coexisting in one case. Generally, NRAS and KRAS have been observed to be mutually exclusive²¹; however, recent studies have reported rare cases of their coexistence.^18,22,40,48 In our study, both the cases with coexistence of NRAS and KRAS mutations harbored a KRAS mutation at codon 12, which corresponds to the previous reports.^48,60 In addition, three cases had coexisting NRAS and TP53 mutations, which have not been reported earlier to the best of our knowledge.

Conclusion

As per our knowledge, this is the first NGS-based biomarker study in Indian CRC patients. KRAS, BRAF, PIK3CA, NRAS, TP53, and APC are frequently altered in CRC and the mutation rates were comparable to worldwide reports. No patient harbored a mutation in six genes or five genes together, and only four patients harbored a mutation simultaneously in four genes. This suggests that CRC can be initiated through different signaling pathways. The six genes were frequently altered together in pairs and triplets suggesting that their synergistic effect might escalate CRC tumorigenesis. However, a mutual exclusion was noted between KRAS and BRAF, BRAF and NRAS, and NRAS and PIK3CA. Further studies involving a greater study population are required to validate the biomarkers for developing targeted therapies with better performance covering a wider range of therapeutic targets.

Footnotes

Acknowledgements

This work was mainly performed in the Department of Research, Sir Ganga Ram Hospital. The authors thank Strand Life Sciences for their assistance in data retrieval and preparation, and Mrs Parul Takkar for statistical analysis.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors.

Funding

This study has been funded by Merck Serono India.

Informed consent

Informed consent was obtained from all individual participants included in this study.

References

Ferlay

Soerjomataram

Ervik

. GLOBOCAN 2012 v1. 0, Cancer incidence and mortality worldwide in 2012, IARC CancerBase no. 11, International Agency for Research on Cancer, http://globocan.iarc.fr (2014, accessed 24 November 2014).

Sung

Lau

Goh

. Asia Pacific Working Group on Colorectal Cancer. Increasing incidence of colorectal cancer in Asia: implications for screening. Lancet Oncol 2005; 6(11): 871–876.

Sirohi

Shrikhande

Perakath

. Indian Council of Medical Research consensus document for the management of colorectal cancer. Indian J Med Paediatr Oncol 2014; 35(3): 192–196.

Dinu

Dobre

Panaitescu

. Prognostic significance of KRAS gene mutations in colorectal cancer—preliminary study. J Med Life 2014; 7(4): 581–587.

Bos

Fearon

Hamilton

. Prevalence of ras gene mutations in human colorectal cancers. Nature 1987; 327(6120): 293–297.

Loupakis

Ruzzo

Cremolini

. KRAS codon 61, 146 and BRAF mutations predict resistance to cetuximab plus irinotecan in KRAS codon 12 and 13 wild-type metastatic colorectal cancer. BJC 2009; 101(4): 715–721.

Amado

Wolf

Peeters

. Wild-type KRAS is required for panitumumab efficacy in patients with metastatic colorectal cancer. J Clin Oncol 2008; 26(10): 1626–1634.

Davies

Bignell

Cox

. Mutations of the BRAF gene in human cancer. Nature 2002; 417(6892): 949–954.

Barras

BRAF mutation in colorectal cancer: an update. Biomark Cancer 2015; 7(Suppl 1): 9–12.

10.

Gonsalves

Mahoney

Sargent

. Patient and tumor characteristics and BRAF and KRAS mutations in colon cancer, NCCTG/Alliance N0147. J Natl Cancer Inst 2014; 106(7): dju106.

11.

Sylvester

Huo

Khramtsov

. Molecular analysis of colorectal tumors within a diverse patient cohort at a single institution. Clin Cancer Res 2012; 18(2): 350–359.

12.

Roth

Tejpar

Delorenzi

. Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60-00 trial. J Clin Oncol 2010; 28(3): 466–474.

13.

Qiu

Zhi

. Colorectal carcinomas with KRAS codon 12 mutation are associated with more advanced tumor stages. BMC Cancer 2015; 15(1): 350.

14.

Bisht

Ahmad

Sawaimoon

. Molecular spectrum of KRAS, BRAF, and PIK3CA gene mutation: determination of frequency, distribution pattern in Indian colorectal carcinoma. Med Oncol 2014; 31(9): 1–3.

15.

Baskin

Calibasi

Amirfallah

. KRAS and BRAF mutation frequencies in a series of Turkish colorectal cancer patients. Transl Cancer Res 2014; 3(2): 160–166.

16.

Cathomas

PIK3CA in colorectal cancer. Front Oncol 2014; 4: 35.

17.

Rosty

Young

Walsh

. PIK3CA activating mutation in colorectal carcinoma: associations with molecular features and survival. PLoS ONE 2013; 8(6): e65479.

18.

Zhang

Zheng

Yang

. Molecular spectrum of KRAS, NRAS, BRAF and PIK3CA mutations in Chinese colorectal cancer patients: analysis of 1,110 cases. Sci Rep 2015; 5: 18678.

19.

Zenonos

Kyprianou

RAS signaling pathways, mutations and their role in colorectal cancer. World J Gastrointest Oncol 2013; 5(5): 97–101.

20.

Downward

Targeting RAS signalling pathways in cancer therapy. Nat Rev Cancer 2003; 3(1): 11–22.

21.

Irahara

Baba

Nosho

. NRAS mutations are rare in colorectal cancer. Diagn Mol Pathol 2010; 19(3): 157.

22.

Shen

Wang

Han

. Effectors of epidermal growth factor receptor pathway: the genetic profiling of KRAS, BRAF, PIK3CA, NRAS mutations in colorectal cancer characteristics and personalized medicine. PLoS ONE 2013; 8(12): e81628.

23.

Soussi

Béroud

Significance of TP53 mutations in human cancer: a critical analysis of mutations at CpG dinucleotides. Hum Mutat 2003; 21(3): 192–200.

24.

Soussi

Lozano

p53 mutation heterogeneity in cancer. Biochem Biophys Res Commun 2005; 331(3): 834–842.

25.

Pan

Wan

Chen

. Co-mutation of p53, K-ras genes and accumulation of p53 protein and its correlation to clinicopathological features in rectal cancer. World J Gastroenterol 2004; 10(24): 3688–3690.

26.

Rehman

Sameer

Zahoor

. Distinct pattern of mutations of conserved regions of TP53 in colorectal cancer patients in the Kashmir population: an emerging high-risk area. Ecancermedicalscience 2009; 3: 129.

27.

Jeon

Lee

Shin

. Genetic alterations of APC, K-ras, p53, MSI, and MAGE in Korean colorectal cancer patients. Int J Colorectal Dis 2008; 23(1): 29–35.

28.

Oki

Zhao

Yoshida

. The difference in p53 mutations between cancers of the upper and lower gastrointestinal tract. Digestion 2009; 79(Suppl. 1): 33–39.

29.

Smith

Carey

Beattie

. Mutations in APC, Kirsten-ras, and p53—alternative genetic pathways to colorectal cancer. Proc Natl Acad Sci U S A 2002; 99(14): 9433–9438.

30.

Conlin

Smith

Carey

. The prognostic significance of K-ras, p53, and APC mutations in colorectal carcinoma. Gut 2005; 54(9): 1283–1286.

31.

Fearnhead

Britton

Bodmer

WF.

The ABC of APC. Hum Mol Gen 2001; 10(7): 721–733.

32.

Lüchtenborg

Weijenberg

Roemen

. APC mutations in sporadic colorectal carcinomas from The Netherlands Cohort Study. Carcinogenesis 2004; 25(7): 1219–1226.

33.

Sugai

Habano

Jiao

. Analysis of molecular alterations in left-and right-sided colorectal carcinomas reveals distinct pathways of carcinogenesis: proposal for new molecular profile of colorectal carcinomas. J Mol Diagn 2006; 8(2): 193–201.

34.

Majek

Gondos

Jansen

. Sex differences in colorectal cancer survival: population-based analysis of 164,996 colorectal cancer patients in Germany. PLoS ONE 2013; 8(7): e68077.

35.

Weiss

Pfau

O’Connor

. Mortality by stage for right-versus left-sided colon cancer: analysis of surveillance, epidemiology, and end results—Medicare data. J Clin Oncol 2011; JCO-2011.

36.

Yuan

Sun

Zhang

. APC and K-ras gene mutation in aberrant crypt foci of human colon. World J Gastroenterol 2001; 7(3): 352–356.

37.

Yoon

Shi

Alberts

. Racial differences in BRAF/KRAS mutation rates and survival in stage III colon cancer patients. J Natl Cancer Inst 2015; 107(10): djv186.

38.

Malhotra

Anwar

Nanda

. Alterations in K-ras, APC and p53-multiple genetic pathway in colorectal cancer among Indians. Tumor Biology 2013; 34(3): 1901–1911.

39.

Palomba

Colombino

Contu

. Prevalence of KRAS, BRAF, and PIK3CA somatic mutations in patients with colorectal carcinoma may vary in the same population: clues from Sardinia. J Transl Med 2012; 10(1): 178.

40.

Phipps

Buchanan

Makar

. KRAS-mutation status in relation to colorectal cancer survival: the joint impact of correlated tumour markers. Br J Cancer 2013; 108(8): 1757–1764.

41.

Barault

Veyrie

Jooste

. Mutations in the RAS-MAPK, PI(3)K (phosphatidylinositol-3-OH kinase) signaling network correlate with poor survival in a population-based series of colon cancers. Int J Cancer 2008; 122(10): 2255–2259.

42.

Sinha

Hussain

Mehrotra

. KRAS gene mutation and RASSF1A, FHIT and MGMT gene promoter hypermethylation: indicators of tumor staging and metastasis in adenocarcinomatous sporadic colorectal cancer in Indian population. PLoS ONE 2013; 8(4): e60142.

43.

McLellan

Owen

Stepniewska

. High frequency of K-ras mutations in sporadic colorectal adenomas. Gut 1993; 34(3): 392–396.

44.

Neumann

Zeindl-Eberhart

Kirchner

. Frequency and type of KRAS mutations in routine diagnostic analysis of metastatic colorectal cancer. Pathol Res Pract 2009; 205(12): 858–862.

45.

Colombino

Sperlongano

Izzo

. BRAF and PIK3CA genes are somatically mutated in hepatocellular carcinoma among patients from South Italy. Cell Death Dis 2012; 3(1): e259.

46.

Phua

Yeo

. Prevalence of KRAS, BRAF, PI3K and EGFR mutations among Asian patients with metastatic colorectal cancer. Oncol Lett 2015; 10(4): 2519–2526.

47.

Stec

Semeniuk-Wojtaś

Charkiewicz

. Mutation of the PIK3CA gene as a prognostic factor in patients with colorectal cancer. Oncol Lett 2015; 10(3): 1423–1429.

48.

Negru

Papadopoulou

Apessos

. KRAS, NRAS and BRAF mutations in Greek and Romanian patients with colorectal cancer: a cohort study. BMJ Open 2014; 4(5): e004652.

49.

Peeters

Kafatos

Taylor

. Prevalence of RAS mutations and individual variation patterns among patients with metastatic colorectal cancer: a pooled analysis of randomised controlled trials. Eur J Cancer 2015; 51(13): 1704–1713.

50.

Calistri

Rengucci

Seymour

. Mutation analysis of p53, K-ras, and BRAF genes in colorectal cancer progression. J Cell Physio 2005; 204(2): 484–488.

51.

Thomas

Association of p53 mutations with short survival in colorectal cancer. Gastroenterology 1994; 106: 42–48.

52.

Slebos

Baas

Clement

. Clinical and pathological associations with p53 tumour-suppressor gene mutations and expression of p21WAF1/Cip1 in colorectal carcinoma. Br J Cancer 1996; 74(2): 165–171.

53.

Russo

Bazan

Iacopetta

. The TP53 colorectal cancer international collaborative study on the prognostic and predictive significance of p53 mutation: influence of tumor site, type of mutation, and adjuvant treatment. J Clin Oncol 2005; 23(30): 7518–7528.

54.

Golmohammadi

Namazi

Nikbakht

. Characterization and prognostic value of mutations in exons 5 and 6 of the p53 gene in patients with colorectal cancers in central Iran. Gut Liver 2013; 7(3): 295–302.

55.

Samowitz

Curtin

. Prognostic significance of p53 mutations in colon cancer at the population level. Int J Cancer 2002; 99(4): 597–602.

56.

Masjedizadeh

A unique profile of adenomatous polyposis coli gene mutations in Iranian patients suffering sporadic colorectal cancer. Cell 2014; 16(1): 17.

57.

Ling