Expression Profile of KRAS and p16 in Periampullary Cancer

Abstract

Activating point mutations in codons 12, 13, and 61 of the KRAS gene and loss of p16 expression, a tumor suppressor gene, are common genetic alterations in periampullary cancer (PAC). The present study explores expression profile of KRAS and p16 genes in PAC and its prognostic relevance. A total of 50 patients with PAC who underwent potentially curative pancreaticoduodenectomy were included in the study. Formalin-fixed, paraffin-embedded tissue samples were analyzed for point mutations in codons 12 and 13 of KRAS and codon 9 of p16 using polymerase chain reaction. KRAS mutation in codon 12/13 was found in 32 (64%) and loss of p16 expression in 36 (72%) cases. KRAS mutation was significantly associated with higher grade, higher pathological tumor (pT) stage, lymphovascular invasion (LVI), perineural invasion (PNI), and pathological lymph nodes (pN) involvement on univariate analysis. On multivariate analysis, significant association of KRAS remained with higher grade (p = 0.031), pT stage (p = 0.09), and LVI (p = 0.028). On univariate analysis, loss of p16 expression was significantly associated with higher grade, pN involvement, LVI, PNI, and pT stage whereas on multivariate analysis, statistical significant association of p16 was found with higher grade of tumor only (p = 0.04). Patients with KRAS mutation had significantly (p = 0.018) worse disease-free survival (DFS) whereas no significant association was found in overall survival (OS). Loss of p16 expression had no association with either DFS or OS. The presence of p16 and KRAS alterations in patients with PAC suggests aggressive tumor biology. KRAS mutations confer a significantly poor DFS in PAC.

Introduction

Periampullary carcinomas (PAC) are a heterogeneous group of tumors that arise around 2 cm of the confluence of the common bile duct (CBD) with the main pancreatic duct (MPD) and have different anatomical origins: pancreatic head (60%), ampulla of Vater (20%), distal common bile duct (10%), and duodenum (10%) [1]. PAC account for only about 0.5–2% of all gastrointestinal malignancies and 20% of all tumors of the extrahepatic biliary tree [2–4]. The prognosis of PAC is deemed better compared to pancreatic ductal adenocarcinomas (PDAC) with comparative 5-year survival rates of 33–50% vs. 10–20%, respectively [5, 6]. Two histomolecular subtypes of ampullary carcinoma have been described; the intestinal type exhibiting similarity to colon/duodenal carcinomas and pancreatobiliary (PB) type having an aggressive behavior resembling PDAC and cholangiocarcinoma [7].

While the molecular alterations in the etiopathogenesis of PDAC have been extensively studied and reported, there exists paucity of data regarding genetic aberrations in PAC. Activating mutations of KRAS gene and inactivating mutations of genes p16 (also known as CDKN2A or INK4A), p53, and SMAD4 are the commonly found in invasive PDAC. Since both PDAC and PAC occur in proximity, they also share a few genetic alterations that include genes like KRAS, BRAF, PIK3CA, and EGFR [8].

We have herein tried to assess the incidence of KRAS and p16 mutations in 50 patients with PAC and its correlation, if any, with adverse pathological features, disease-free (DFS) and overall survival (OS).

Patients and Methods

This study was conducted in the Hepatopancreatobiliary and Gastrointestinal Division, Department of Surgical Oncology, of a tertiary care hospital in India between January 2015 and December 2016. The Institute’s Ethical Committee approved the study.

Patients

After obtaining an informed consent, 50 diagnosed consecutive patients of PAC who underwent pancreaticoduodenectomy were included in this study. A detailed history, examination, and histopathological (HPE) findings were recorded in each case. No patient received any neoadjuvant therapy. The patients were staged using a pancreatic protocol triple-phase contrast-enhanced computed tomography (CECT) scan of the thorax, abdomen, and pelvis. A preoperative upper gastrointestinal (UGI) endoscopy ± endoscopic ultrasound (EUS) and biopsy was performed in every patient. A standard pylorus resecting pancreaticoduodenectomy was done in all cases with Blumgart’s duct-to-mucosa pancreaticojejunostomy (DMPJ). Patients were administered gemcitabine-based adjuvant chemotherapy. No patients received any radiotherapy. Patients were followed up from the day of discharge from hospital every 3 months for the first 2 years and every 6 months thereafter till date. A CECT scan of thorax, abdomen, and pelvis was repeated every 6 months and tumor markers CEA and CA 19-9 every 3 months.

Methods

Formalin-fixed, paraffin-embedded tissue samples were taken from each of the 50 resected specimens after pancreaticoduodenectomy. The methodology used is described below.

DNA Isolation and Elution from Paraffin-Embedded Tissue Samples

DNA was isolated using standard proteinase K phenol chloroform method and reagents provided by QIAamp QIAGEN. The details of the methodology used are included in the supplementary file.

Primers for KRAS and p16 Genes

1. KRAS gene

Previously used KRAS specific primers were tested using Primer3 program (http://simgene.com/Primer3) for codons 12 and 13. For confirmation of primers, input parameters used were product size 100–300 bp, primer length 18–30 bp, primer annealing temperature 56–65 °C, and GC content 30–70% (Table 1).

• 2.p16 gene

Table 1.

Primers of specific genes, their sequences, and amplification product

Targeted gene	Sequences	Product size (bp)
KRAS*	5′-TGGTGGAGTATTTGATAGTGTATTAACCT-3′	282
	5′-ATGAAAATGGTCAGAGAAACCTTTATC-3′
p16	5′ TTATTAGAGGGTGGGGCGGATCGC3′	234
	3′ CCACCTAAATCGACCTCCGACCG5′
PCR program: 94 °C, 60 s; 65 °C, 60 s; 72 °C 60 s (30 cycles)

Legend: bp base pairs, PCR polymerase chain reaction

*KRAS specific fragment primers exon2, includes codons 12 and 13

Previously used primer of p16 at location 9p21.3 and 8 exon were procured and tested using Primer3 program (http://simgene.com/Primer3). We had selected accessionNC_000009.12 and NC_000009.11, 21967751, 21994490, 21967752, 21995043, complement of chromosome 9 for primer design. Genomic position primers were placed near the transcriptional start site. Details of polymerase chain reaction (PCR) primers are given in Table 1.

PCR Amplification

PCR amplification was performed as per standard protocol of Mullis et al. [9] with minor modifications [10]. Reaction was carried out in a 25-μl volume reaction mixture containing 10× PCR buffer having 10 mM Tris HCl pH 8.4, 50 mMKCl, 1.5 mM MgCl₂, 200 μM of each deoxynucleotide triphosphate (MBI, Fermentas), 5 pmol of each oligonucleotide primer (Operon, Cologne, Germany), and 100 ng of DNA with 0.5 U of Taq polymerase (Bangalore Genie, India), for amplification of p16 and KRAS genes. All the amplifications were carried out in a thermal cycler (Biometra, Goettingen, Germany) at initial denaturation of 95 °C for 4 min followed by 30 cycles 94 °C for 30 s annealing at 65 °C and 62 °C for 30 s respectively p16 and KRAS and extension at 72 °C for 30 s which was extended for 4 min in the final cycle.

Gel Electrophoresis of PCR Product (Amplicon) and Visualization

To check for the specific DNA amplification, an agarose gel electrophoresis was performed. For carrying out gel electrophoresis, 3% agarose was dissolved in 1× TAE buffer in a microwave oven. After cooling it to just below the 45 °C, 0.5 mg/ml ethidium bromide (EtBr) was added to the final concentration. Gel slabs were prepared by pouring the dissolved agarose in the required plate. A comb having at least 0.5 to 1.0 mm space above the base of the gel was inserted and the gel was allowed to solidify for 25 min. Then, the comb was removed and the tank was filled with 1× TAE buffer. Next, 10 μl sample and 100 bp marker with gel 2 μl loading buffer (0.025% bromophenol blue and xylene cyanol, 10 mM EDTA pH 8.0, 50% glycerol) was loaded in the sample well and electrophoresis was carried out at approximately 80 V. Amplified DNA bands were visualized and imaged by Alpha imager EC Alpha Innotech.

Statistical Analysis

A statistical analysis was performed using SPSS (version 25.0, Chicago, IL, USA) software. For comparing two groups, mean Student t-test was done. The chi-square test was used to analyze the statistical significance between p16 and KRAS mutation with other clinicopathological variables. Tests were considered significant if p value <0.05. Survival curves were estimated by the Kaplan-Meier method and differences between the curves were measured using the log rank test.

Results

Ampullary cancers were the commonest of the PAC (40; 80%), of which 36% (n = 22) had intestinal and 34% (n = 18) PB morphology, followed by cancers of pancreatic origin 10% with duodenal and biliary cancers constituting 6% and 4%, respectively. Out of 50 resected pancreatoduodenectomy specimens, all had clear and uninvolved margins. One patient underwent concomitant portal vein resection to obtain negative margin. Details of all clinicopathological parameters are shown in Table 2.

Table 2.

Clinicopathological parameters

Location	Number (Percent)
Ampulla	40 (80%)
Pancreatic head	5 (10%)
Duodenum	3 (6%)
CBD	2 (4%)
Histology
Adenocarcinoma	45 (90%)
Adenosquamous	2 (4%)
Signet ring cell	3 (6%)
Histological subtype: ampullary
Intestinal	22 (44%)
Pancreatobiliary (PB)	18 (36%)
Histological differentiation
Well differentiated	25 (50%)
Moderately differentiated	22 (44%)
Poorly differentiated	3 (6%)
Margin status: negative	50 (100%)
LVI
Present	26 (52%)
Absent	24 (48%)
PNI
Present	21 (42%)
Absent	29 (58%)
pT stage
T1	5 (10%)
T2	16 (32%)
T3	19 (38%)
T4	10 (20%)
pN stage
N0	22 (44%)
N1/2	28 (56%)

Legend: LVI lymphovascular invasion, PNI perineural invasion, pT pathological tumor stage

KRAS Gene

KRAS mutation was present in 32 (64%) cases (Table 3). The PCR amplification pattern of KRAS is shown in Fig. 1. No statistically significant association was found between KRAS mutation and age, gender, location of tumor, or histological variants. KRAS mutation was found to be significantly associated with higher grade of tumor, higher pathological tumor (pT) stage, lymphovascular invasion (LVI), perineural invasion (PNI), and pathological lymph node (pN) involvement on univariate analysis. However, on the multivariate analysis, KRAS mutation had significant association only with higher grade of tumor (p = 0.031), pT stage (p = 0.009), and LVI (p = 0.028) (Table 4).

Table 3.

KRAS mutation

KRAS gene	Number	Percent (%)
Present	32	64.0
Absent	18	36.0
Total	50	100.0

Fig. 1.

PCR amplification pattern of loaded sample in KRAS of 282 bp. mMr lane is of marker and amplification seen in lanes 3, 4, 5, 6, 7, and 1 is negative control. Legend: PCR polymerase chain reaction, bp base pairs, mMr molecular marker

Table 4.

Relationship between KRAS mutation and clinicopathological characters

Clinicopathological parameters	KRAS present	KRAS absent	p value (univariate analysis)	p value (multivariate analysis)
Gender
Male	25	11	p = 0.203	p = 0.613
Female	7	7
Location
Ampulla	29	11	p = 0.083	p = 0.14
Pancreatic	1	4
Duodenum	1	2
CBD	1	1
Histology
Adenocarcinoma	29	16	p = 0.519	p = 0.075
Adenosquamous	2	0
Signet ring	1	2
Grade
2 + 3	20	5	p = 0.018*	p = 0.031*
1	12	13
LVI
Present	22	4	p = 0.002*	p = 0.028*
Absent	10	14
PNI
Present	17	4	p = 0.034*	p = 0.198
Absent	15	14
pT stage
T3, T4	22	7	p = 0.04*	p = 0.009*
T1, T2	10	11
pN stage
N1/2	18	4	p = 0.02*	p = 0.93
N0	14	14

Legend: CBD common bile duct, pT pathological tumor stage, pN pathological node stage, LVI lymphovascular invasion, PNI perineural invasion

*p < 0.05

p16 Gene

Loss of p16 expression was found in 36 (72%) cases (Table 5). The PCR amplification pattern of p16 is shown in Fig. 2. On univariate analysis, loss of p16 expression was significantly associated with higher grade of the tumor, pN involvement, LVI, PNI, and higher pT stage of tumor whereas no significant association was found between age, gender, histological variants, and location of tumor. However, on multivariate analysis, only higher grade of the tumor (p = 0.04*) remained statistical significant (Table 6).

Table 5.

p16 mutation/loss of expression

p16 gene	Number	Percent (%)
Loss of p16 expression	36	72.0
Expressed	14	28.0
Total	50	100.0

Fig. 2.

PCR amplification pattern of loaded sample in p16 primer of 234 bp. mMr lane is of marker and amplification seen in lanes 2, 3, 5, 6, and 1 is negative control. Legend: PCR polymerase chain reaction, bp base pairs, mMr molecular marker

Table 6.

Relationship between p16 expression and clinicopathological characters

Clinicopathological parameters	p16 loss of expression	p16 expressed	p value (univariate analysis)	p value (multivariate analysis)
Gender
Male	28	8	p = 0.521	p = 0.453
Female	8	6
Location
Ampulla	32	8	p = 0.477	p = 0.275
Pancreatic	1	4
Duodenum	2	1
CBD	1	1
Histology
Adenocarcinoma	33	12	p = 0.881	p = 0.466
Adenosquamous	2	0
Signet ring	1	2
Grade
2 + 3	23	2	p = 0.002*	p = 0.04*
1	13	12
LVI
Present	23	3	p = 0.007*	p = 0.269
Absent	13	11
PNI
Present	19	2	p = 0.013*	p = 0.451
Absent	17	12
pT stage
T3, T4	24	5	p = 0.046*	p = 0.678
T1, T2	12	9
pN stage
N1/2	20	2	p = 0.008*	p = 0.418
N0	16	12

Legend: CBD common bile duct, pT pathological tumor stage, pN pathological node stage, LVI lymphovascular invasion, PNI perineural invasion

*p < 0.05 significant

We did a subset analysis of KRAS and p16 expression profiles in between two groups made by clubbing the PB type of ampullary cancer with pancreatic and biliary PAC together as group A and comparing it with the intestinal type of ampullary cancer and duodenal PAC as group B (Table 7). No significant association was found between KRAS and p16 expression among the two groups, A and B, although KRAS was more frequently found mutated in group A (p = 0.08).

Table 7.

KRAS and p16 expression profile in different subtypes of PAC

	PB (18) + biliary (2) + pancreatic (5) Group A (25)	Intestinal (22) + duo (3) Group B (25)	Total	p value
KRAS
Present	19	13	32	p = 0.08
Absent	6	12	18
p16
Loss of expression	20	16	36	p = 0.21
Expressed	5	9	14
Total			50

Legend: PB pancreatobiliary type of ampullary carcinoma, Intestinal intestinal type of ampullary carcinoma, duo duodenal carcinoma

Survival Analysis

Clinical follow-up of all post-operative patients revealed that 45 out of 50 patients survived with an average follow-up of 28 months (range from 9 to 60 months). In all, there were 5 deaths and all died of metastatic PAC. Overall, 13 patients developed metastases during follow-up.

Significantly more patients with KRAS mutation had disease recurrence (p = 0.018). Median DFS (Fig. 3a) with and without KRAS mutation was 34 (SE = 0.936) months and 44 (SE = 7.999) months, respectively (p = 0.018), whereas median OS (Fig. 3b) was 36 (SE = 0.737) months and 44 (SE = 8.960) months, respectively (p = 0.472).

Fig. 3.

a KRAS mutation versus disease-free survival (DFS). Legend: Group 1 mutated KRAS, group 2 wild KRAS, *p = 0.018. b KRAS mutation versus overall survival (OS). Legend: Group 1 mutated KRAS, group 2 wild KRAS, p value = 0.472

No significant difference was observed between DFS and OS among patients with loss of p16 expression versus normal expression. Median DFS (Fig. 4a) was found to be 36 (SE = 0.925) months and 60 (SE = 0.000) months, respectively (p = 0.257), and the median OS (Fig. 4b) was 36 (SE = 0.936) months and 44 (SE = 5.690) months, respectively (p = 0.752).

Fig. 4.

a p16 expression versus disease-free survival (DFS). Legend: Group 1 loss of p16 expression, group 2 p16 expressed, p = 0.257. b p16 expression versus overall survival (OS). Legend: Group 1 loss of p16 expression, group 2 p16 expressed, p = 0.752

Discussion

Molecular alterations in PAC are less extensively studied due to albeit rarity of this disease although significant literature does exist on PDAC. Although these cancers share some similarities in their genetic profile, important differences exist between subtypes of PAC, including incidence of driver mutations, natural history, biology, and response to chemotherapy [11].

Compared to biliary and pancreatic cancers, ampullary cancers have significantly better biologic behavior and outcome, with 5-year survivals ranging between 37 and 68% [12]. Ampullary cancers are further characterized by variable histologic and morphologic features. While some studies demonstrate that intestinal type ampullary carcinomas were associated with a better prognosis compared to the PB type [7, 13, 14], others reported that stage-by-stage, they tended to have the same prognosis [15, 16].

KRAS

KRAS mutations are generally single amino acid substitution in codon 12 or less frequently codon 13. The KRAS mutations found in PDAC are mainly G-T transversions at the first or second base and a G-A transition at the second base of codon 12. This mutation spectrum is similar to the spectrum found in lung carcinomas, but differs from that in colon tumors. Activating mutations at codons 12, 13, and 61 have been detected with variable frequencies in PDAC showing highest incidence of 90% and are associated with poor prognosis. However, unlike in PDAC, PAC have a far lower incidence of KRAS mutation. Mutations in KRAS are an early event in carcinogenesis [17] and are found in 30 to 40% of patients with ampullary adenocarcinoma [18–20]. In a meta-analysis by Kim et al., the incidence of KRAS mutation varied from 30 to 67% among the 5 studies included in the analysis. Out of 388 patients, 175 (45%) had KRAS mutation. Of 175 patients with mutant KRAS, 134 (76.5%) had mutation at codon 12 [21]. Since KRAS mutations in various studies have been reported most commonly to occur at codon 12 and less frequently codon 13, we confined mutation analysis of KRAS oncogene to codons 12 and 13. KRAS mutation was found in 64% (32) out of 50 PAC in our study and this was higher than reported in most other studies.

In our study, a statistically significant association was found between KRAS mutation and higher T stage. Valsangkar et al. [12] too reported higher stage ampullary carcinoma (56% in T4) to be associated with deleterious KRAS mutation (G12D). In another study by Chen et al., KRAS mutation was found to be in statistically significant association with higher T stage (p = 0.033) [22]. However, Howe et al., in their study of 92 patients, found no significant association between T stage and KRAS mutation [23].

We did not find any significant difference in the incidence of KRAS mutation between the intestinal vs. PB subtype of PAC. Similar findings have been reported by Mikhitarian et al. [24] 52% vs. 42%, Matsubayashi et al. [20] 78% vs. 64% respectively and also by Howe et al. [23] and Kwon et al. [25].

On univariate analysis, KRAS mutation in our study was found to be significantly associated with higher grade and pT stage, LVI, PNI, and pN involvement. It was not associated with age, gender, histological type or subtype, and location of tumor.

However, on multivariate analysis, KRAS mutation was found to have significant association only with higher grade of tumor (p = 0.031), pT stage (p = 0.09), and LVI (p = 0.028). KRAS mutation was significantly associated with worse DFS (p = 0.018) although it had no effect on OS. Two of the 5 studies included in the meta-analysis by Kim et al. [21] also reported significantly worse relapse-free survival in KRAS mutant ampullary carcinomas (HR = 2.74, 95% CI: 1.52–4.92, p = 0.0008).

p16

The p16 tumor suppressor gene located on 9p21 encodes a 16-kDa protein that acts as a cyclin-dependent kinase (CDK) 4/6 inhibitor [26]. The mutation of p16 gene in PAC is caused by homozygous deletion, point mutation, or methylation of the CpG island located in the p16 gene promoter region [27–29]. The incidence of p16 gene mutation and protein expression in PDAC has been reported between 30 and 87% [29–31]. Genetic abnormalities inactivating the p16 gene thus confer a growth advantage to the cell contributing to tumorigenesis.

Loss of p16 expression was found in 36 (72%) of our patients. On univariate analysis, p16 expression loss was significantly associated with higher pT stage, pN involvement, higher grade of the tumor, presence of LVI, and PNI whereas no significant association was found between age, gender, histological variety, and location of tumor in this study. On multivariate analysis, statistical significant association was found only between p16 and higher grade of the tumor. Loss of p16 gene expression has been reported by various authors to be associated with higher pT stage [32], lymph nodal status [32, 33], histological grade [34], and PNI [35]. We, however, could not find any study in literature that reveals an association between LVI and loss of p16 expression in PAC to the best of our knowledge.

Hechtman et al. [36] reported p16 to be more frequently deleted in intestinal type ampullary carcinoma vs. in PB type ampullary carcinoma (30% vs. 6%), although these trends did not reach statistical significance (p > 0.05). Our study did not reveal any significant association between histological subtypes and loss p16 expression.

We found no association between loss of p16 expression and DFS and OS. Chang et al. [37] similarly found no significant association between loss of p16 expression and survival in their series of 68 patients with PAC. There are publications showing that lack of expression of p16 was related to better survival [34, 38].

KRAS mutation and loss of p16 expression play an important role in the pathogenesis of PAC. Their expression profiles may trigger aggressive behavior and hence affect disease prognosis thereby the need for adjuvant therapy. The expression of both these genes may be further assessed in larger samples and open doors to their clinical use as therapeutic targets and prognostic biomarkers. The present study has some limitations because it is a single-center study, with a relatively small sample size and short duration of follow-up.

Conclusions

KRAS mutation and loss of p16 expression plays an important role in the pathogenesis of PAC. This is one of the very few studies published to date reporting the incidence of these two important mutations in PAC and its correlation with histological parameters and survival. The presence of p16 and KRAS alterations in patients with PAC suggest aggressive tumor biology. Multicenter studies exploring the genetic aberrations in PAC are required that will increase our understanding of molecular pathogenesis of PAC and its effect on prognosis of the disease. It will further pave the way for research in to new therapeutic and diagnostic approaches for this deadly disease.

Declarations

Competing Interests

The authors declare no competing interests.