Sage Journals: Discover world-class research

Abstract

Purpose

To date, only a few studies have investigated the role of molecular alterations in cancer recurrence. This exploratory study aimed to evaluate the impact of molecular alterations on the time and site of recurrence in patients with stage I–IV CRC and to identify the risk factors predicting recurrence-free survival in colon cancer.

Methods

A total of 270 patients were retrospectively included. We assessed the full RAS status using Sanger and pyrosequencing. MSI status was determined by immunohistochemical analysis. Molecular alterations were correlated with recurrence timing (early or late), recurrence patterns, and recurrence-free survival. Statistical analysis was performed using the Kaplan–Meier method and the log-rank test.

Results

Of the 270 patients, 85 (31%) experienced recurrence, among whom 53% had mutant full RAS status, 48% had KRAS mutations, and 31.4% had KRAS p. G12V mutation subtype. Compared with those with late recurrence, patients with early recurrence were significantly older (P = 0.02) and more likely to have poorly differentiated tumors, a higher rate of positive lymph nodes, KRAS mutations, and especially KRAS p. G12V mutation variant. RAS mutation status, KRAS mutations, and rare mutations are more common in patients with lung cancer recurrence. Multivariate logistic regression analysis revealed that differentiation, perineural invasion, full RAS mutation status, and KRAS codon 13 mutations were independent factors for recurrence-free survival in colon cancer.

Conclusion

In this cohort, the timing and patterns of recurrence appeared to be associated with the patient’s molecular profile. KRAS codon 12 mutations were the worst predictors of recurrence-free survival at all stages in our population.

Keywords

colon cancer RAS mutations the microsatellite instability status recurrence

Introduction

Colon cancer (CC) is one of the most common types of cancer worldwide. It ranks third in terms of incidence, with 1.9 million new cases diagnosed in 2020.¹ In Morocco, CC is a major public health issue given the remarkable and continuous increase in its incidence and mortality rates. According to GLOBOCAN, 4558 new colon cancer cases and approximately 1386 deaths were recorded in 2020.¹ The risk of recurrence raises concerns about postoperative surveillance to improve patient survival through earlier detection of local and metachronous recurrence at a curable stage. In colon cancer, approximately 15%–25% of patients present with synchronous metastasis at the time of diagnosis, and approximately 25%–50% develop metachronous metastasis or distant recurrence.^2-5 Previous reports have indicated that 80% of recurrences usually appear in the 3 years following curative surgery,⁶ and 95% of them occur in the first 5 years.⁷ Early recurrence was defined as recurrence within 2 years of surgery. The most common sites of distant recurrence patterns are the liver (20%-50%), followed by the lungs (5-15%), peritoneum (15%), and bone (3% to 7%).^8-11 Local recurrence (LR) occurs in 4%–11.5% of patients.¹² Many studies have demonstrated that recurrence is a major concern in stages I–III CC because it is strongly associated with shorter disease-free survival, and poor prognosis negatively affects patient survival.^13,14 Furthermore, in stage IV CC, the majority of patients die from distant recurrence, although advances in targeted therapies.¹⁵

It is well established that a higher risk of recurrence, its typical patterns, and decreased recurrence-free survival (RFS) are profoundly influenced by the clinicopathological features of the tumor, such as primary tumor location, lymph node metastasis, and TNM stage.¹⁶ However, recent studies have demonstrated that the molecular and mutation profiles of tumors are associated with a higher risk of recurrence and may affect the site of relapse and metastatic involvement.^4,17,18

In colon cancer, common oncogenetic mutations are found in the mitogen-activated kinase (MAPK) pathway genes, including KRAS (40%-45%), BRAF (10%-15%), and the NRAS (2%-6%), and in the phosphatidylinositol 3-kinase (PI3K) pathway member, PIK3CA (20%)^19-21 (p. 3). In the field of treatment, patients with metastatic colon cancer harboring KRAS and NRAS mutations do not respond to anti-epidermal growth factor receptor (EGFR) monoclonal antibodies (cetuximab and panitumumab).²² The positive relationship between full Ras mutation status and poor overall survival is well established in the literature.¹⁹ Furthermore, recent studies have demonstrated that mutated KRAS is associated with worse recurrence-free survival.^15,23,24 It has also been reported that KRAS mutations influence not only the time of recurrence but also their patterns. According to Tie et al. KRAS mutation status was associated with increased lung recurrence in stage II–III colorectal cancer (CRC).²³ Yaeger et al. reported that lung, bone, and brain metastases occur significantly more often in patients with KRAS mutations after hepactectomy.¹⁷ In addition, BRAF mutations have been reported to be associated with peritoneal spread, which may explain the poor prognosis of tumors with a BRAF-mutated status.¹⁷

To date, few studies have thoroughly investigated the relationship between RAS mutations, RFS, recurrence timing, and recurrence patterns in colon cancer. All reports have focused on stage II–III CRC with KRAS exon 2 mutations²³ or on patients with liver metastasis (LM).²⁵ Furthermore, the effect of NRAS mutations on RFS and recurrence patterns remains unknown, and no study has evaluated the association between KRAS and NRAS mutation subtypes (exon/codon) and patient recurrence, which is now mandatory, especially since it has been shown that mutational subtypes of each gene are associated with poor prognosis factors and shorter overall survival.^19,26 Therefore, in the present study, we investigated for the first time the impact of molecular alterations on the time and site of recurrence in patients with stage I–IV CRC. We identified the risk factors for recurrence-free survival.

Materials and Methods

Sample Size Calculation

To calculate the number of cases required for our study, we used a formula that has been widely used in cross-sectional studies:

N = {(Z - s c o r e)}^{2} * p * {(1 - p) / e r r o r m a r g i n}^{2}

with:

- N = sample size

- For a 95% confidence level, Z-score = 1.96

- P = the estimated proportion of the population with the characteristic (.50)

Margin of error = .05

At risk α = 5%, the N value should be equal to at least 225 patients (N = 225).

Study Design

This study enrolled patients with I–IV CC treated at the HASSAN II University Hospital Center of Fez between 2009 and 2021. A consecutive series of 270 cases was identified. Clinicopathological data, treatment protocols, and follow-up (site and time of the first recurrence) data were collected from the pathology database at the laboratory of anatomical pathology of the hospital. Follow-up was performed at 3-month intervals for the first 2 years and at 6-month intervals thereafter.

Inclusion and Exclusion Criteria

The 270 patients were included according to the following criteria: (a) patients with invasive colon carcinoma (tumor stage T1 and above) and (b) patients who underwent surgical resection at the Department of Surgery of the University Hospital HASSAN II of Fez. The exclusion criteria were as follows: (a) patients with rectal adenocarcinoma, (b) patients aged <18 years, (c) patients with unknown tumor status or tumor site, (d) surgery-related mortality, (e) death within 90 days of surgery, and (f) patients with incomplete follow-up records.

Histopathological Study

All tumor samples were subjected to routine pathological and histopathological examinations at the Department of Anatomical Pathology. All specimens were obtained from primary sites after first-line therapy. Tumor tissues were then fixed in formalin (10%) and embedded in paraffin. Hematoxylin and eosin-stained slides of formalin-fixed paraffin-embedded (FFPE) tumors were reviewed by pathologists to identify the histological characteristics of the tumor (histological type, histological differentiation, perineural invasion, and vascular invasion) and evaluate the percentage of tumor cells in the tumoral zone. All cases were staged according to the tumor, node, and metastasis (pTNM) classification system (AJCC 8th edition).

Molecular Analysis

All biomarkers were assessed using FFPE samples.

KRAS, NRAS, and BRAF Mutation Profiling

DNA Extraction

Formalin-fixed paraffin-embedded tumors were macrodissected from sections containing greater than 30% neoplastic cells. Genomic DNA was extracted using the QIAamp DNA FFPE Tissue Kit (Invitrogen) according to the manufacturer’s instructions and quantified by Qubit analysis.

Polymerase Chain Reaction and Direct Sequencing

DNA amplification of KRAS (exons 2, 3, and 4), NRAS (exons 2, 3, and 4), and BRAF (codon 15) was assessed by PCR using primers on either side of the regions of interest. The PCR products were then purified and sequenced using Sanger sequencing technology with a BigDye Terminator V3.1 Cycle Sequencing Kit (ABI Prism). The sequenced reaction products were subsequently purified using a BigDye® Xterminator™ KIT and analyzed on an Applied Biosystems 3500Dx Genetic Analyzer (Applied Biosystems).

Pyrosequencing

Pyrosequencing technology is the second method used in our laboratory to assess the RAS status of patients with colon cancer. The regions of interest (codons 12, 13, 59, 61, 117, and 146) were amplified by PCR using specific primers. After immobilizing the PCR products on the workstation, single-stranded DNA was prepared, and the sequencing primers were annealed according to the manufacturer’s instructions. The samples were analyzed using Pyromark Q24 software in AQ analysis mode to determine the mutation status of each tumor.

Analysis of the Microsatellite Status

The microsatellite instability (MSI) status was determined by immunohistochemistry as described previously.²⁷

Recurrences

Follow-up data, including patient outcomes and recurrence information, were collected for each patient. Tumor recurrence was studied according to the following parameters: cancer recurrence site (local, liver, lung, periteon, and other sites) and timing. The patients were divided into two periods of recurrence: early recurrence (<2 years) and late recurrence (≥2 years). RFS was defined as the time interval from initial treatment until the first locoregional and/or distant recurrence. Censoring was performed when the patient died or survived without recurrence at the last follow-up. Recurrence was confirmed by histological examination or imaging.

Statistical Analysis

Statistical analysis was performed using SPSS version 21(SPSS Inc., Chicago, IL, USA). The chi-square test or Fisher’s exact test was used to define the links between each mutation and recurrence pattern, as appropriate. An independent t-test was used to identify the potential association between mutation status and time of recurrence.

RFS was estimated using the Kaplan–Meier method, and differences in RFS were assessed using the log-rank test. The Cox proportional hazard regression model was used to evaluate the association between the relevant molecular and clinicopathological variables and RFS. All P-values were two-sided, and P < .05.

Results

Patients and Recurrence Characteristics

In total, 270 patients were included in this study after surgery. The baseline characteristics of the population, stratified according to the presence of recurrence, are summarized in Table 1. The mean patient’s age was 55 ± 15 with a moderate dominance of the male sex (n = 146; 54%). Of 270 patients, 199 (74%) had primary left colon cancer and 71 (26.3%) had primary right colon cancer. Most of the tumor samples were classified as adenocarcinoma (n = 229; 85%) and grade 2 (n = 150; 56%). Distant metastasis at the time of diagnosis was in 163 (60%) patients and positive lymph nodes in 85 (35%). Most patients had advanced colon cancer (III-IV) at the time of diagnosis (n = 162, 60%). After a mean follow-up duration of 30 months, the rate of recurrence after first-line therapy was 31% (n = 85). 13 (15%) patients had local recurrence and 72 (85%) had distal recurrence. The most common distal recurrence site was liver (n = 34; 40%), followed by lungs (n = 19, 22%) and peritoneum (n = 9, 11%); indeed, 10 (12%) patients had mixed recurrence or other organs (Figure 1).

Table 1.

Patients and Recurrence Features.

Variables	Overall n (%)	No Recurrence n (%)	Recurrence	P
Total	270 (100)	185 (69)	85 (31)
Sex
Female	124 (46)	82 (44)	42 (49)	.07
Male	146 (54)	103 (56)	43 (51)
Age
Mean (SD)	55 (15)	56 (15)	53 (15)	0.1
<55	119 (44)	76 (41)	43 (51)	.07
≥55	151 (56)	109 (59)	42 (49)
Tumor site
Right colon	71 (26)	46 (25)	25 (29)	.08
Left colon	199 (74)	139 (75)	60 (71)
Histological subtypes
Adenocarcinoma	229 (85)	160 (86)	69 (81)	.05
Mucinous adenocarcinoma	29 (11)	18 (10)	11 (13)
Others	12 (4)	7 (4)	5 (6)
Grading
Well	120 (44)	88 (48)	32 (38)	.03
Poor	150 (56)	97 (52)	53 (62)
Vascular invasion
Present	46 (17)	33 (18)	13 (16)	0.1
Absent	233 (83)	152 (82)	71 (84)
Perineural invasion
Present	33 (12)	16 (9)	17 (21)	.005
Absent	232 (88)	167 (91)	65 (79)
Lymph nodes
<12 Lymph nodes examined
Yes	67 (28)	41 (25)	26 (32)	.05
No	174 (72)	120 (75)	54 (68)	0.1
Nodal positive
Mean (SD)	2	1 (4)	3 (4)	.01
Presence	85 (35)	49 (30)	36 (45)	0.3
Absence	156 (65)	112 (70)	44 (55)
Lymph node ratio (mean, SD)	10⁻¹	9.10⁻²	1.10⁻²
Disease stage
I-II	108 (40%)	77 (71)	31 (29)	.04
III-IV	162 (60%)	108 (67)	54 (33)
Chemotherapy adjuvant
No	122 (49)	94 (56)	28 (33)	<.001
Yes	129 (51)	73 (44)	56 (67)

Figure 1.

Recurrence patterns in our cohort.

As shown in Table 1, compared with patients without cancer recurrence, those with cancer recurrence were more likely to have mucinous and poorly differentiated tumors (P = 0,05, and P = .03, respectively), less than 12 total lymph nodes (P = .05), presence of positive lymph nodes (P = .01), and perineural invasion (P = .005). Additionally, patients with cancer recurrence were more likely to be diagnosed at an advanced TNM stage (P = .04) and have undergone adjuvant chemotherapy (P < .001). Moreover, in our population, the development of cancer recurrence tended to be associated with the left tumor site (P = .08), with a difference that was close to significant.

Frequencies of Molecular Alterations and Recurrence

The molecular characteristics of the study population are summarized as follows: Table 2. Of the 270 patients, full RAS mutation status was found in 107 (40%) patients. KRAS mutations were detected in 95 (35%) patients and NRAS mutations in 12 (4%). KRAS exon 2 mutations were found in 85 (79%) patients; 63 (74%) of them harbored codon 12 mutations and 22 (26%) harbored codon 13 mutations. Regarding KRAS exon 2 amino acid mutations, the G12D missense mutation was the most frequently observed variant 28 (33%), followed by the G13D 20 (24%) and the G12V 20 (24%). In addition, mutations outside KRAS exon 2 or rare mutations were identified in 22 (8%) patients. The p. Q61L (Nras exon 3) mutation had the highest frequency in the rare mutation group (n = 5, 23%), followed by the p. A146T (Kras exon 4) variant (n = 4, 18%). The p. K117N (Kras exon 4), p. G12D (Nras exon 2), and p. Q61K (Nras exon 3) mutations occurred in 13% (3/22) of the cases (Figure 2(B)).

Table 2.

Molecular Alterations and Recurrence.

Molecular Alteration	Total	No Recurrence	Recurrence	P
RAS status
Wild type	163 (60)	123 (67)	40 (47)	.002
Mutant	107 (40)	62 (33)	45 (53)
KRAS status
Wild type	175 (65)	131 (71)	44 (52)	.002
Mutant	95 (35)	54 (29)	41 (48)
NRAS status
Wild type	258 (96)	177 (96)	81 (95)	0.5
Mutant	12 (4)	8 (4)	4 (5)
Rare mutations
Rare mutant	22 (8)	12 (7)	10 (12)	.09
Wild-Type status	248 (92)	173 (93)	75 (88)
Exons
KRAS Exon 2 mutant	85 (79)	50 (81)	35 (78)	0.4
Outside KRAS exon 2	22 (21)	12 (19)	10 (22)
Mutant KRAS Exon 2
Codon 12 mutant	63 (74)	35 (70)	28 (80)	0.2
Code 13 mutant	22 (26)	15 (30)	7 (20)
KRAS exon 2 amino acid transition (n = 85)
p. G12D	28 (33)	18 (36)	10 (29)	.04
p. G12V	20 (24)	9 (18)	11 (31)
p. G13D	20 (24)	15 (30)	5 (14)
p. G12C	8 (9)	5 (10)	3 (8)
p. G12A	5 (6)	3 (6)	2 (6)
p. G12R	2 (2)	0 (0)	2 (6)
p. G13V	2 (2)	0 (0)	2 (6)
Rare mutation subtypes
p. Q61L (kras exon 3)	1 (5)	1 (8)	0 (0)	0.4
p. K117N (kras exon 4)	3 (13)	1 (8)	2 (20)
p. A146P (kras exon 4)	1 (5)	0 (0)	1 (10)
p. A146T (kras exon 4)	4 (18)	2 (17)	2 (20)
p. A146V (kras exon 4)	1 (5)	0 (0)	1 (10)
p. G12D (Nras exon 2)	3 (13)	1 (8)	2 (20)
p. G13R (Nras exon 2)	1 (5)	1 (8)	0 (0)
p. Q61K (Nras exon 3)	3 (13)	2 (17)	1 (10)
p. Q61L (Nras exon 3)	5 (23)	4 (34)	1 (10)
MSI status
MSI	17 (14)	14 (14)	3 (9)	0.2
MS	107 (86)	75 (86)	32 (91)

Figure 2.

Frequencies of RAS mutation subtypes in our population. (A) Mutation subtypes in KRAS exon2. (B) Mutation subtypes outside KRAS exon 2 (Rare mutations).

Regarding the correlation between cancer recurrence and molecular alterations, patients with cancer recurrence were more likely to have RAS mutation status (n = 45; 53%; P = .002) and to carry KRAS mutations (n = 41; 48%; P = .002). Interestingly, KRAS p. G12V mutation was significantly more common in patients with cancer recurrence than in those with other KRAS exon 2 amino acid transitions (n = 11; 31%; P = .04). In contrast, there were no significant differences in the distribution of NRAS mutation status, KRAS codon 12 vs KRAS codon 13 mutations, rare mutation subtypes, and MSI status between patients with and without cancer recurrence.

Correlation between molecular alterations and clinical pathological features

A summary of the main clinicopathological features of KRAS and NRAS mutation subtypes compared with wild-type RAS is shown in Tables 3 and 4. As shown in Table 3, female gender, left localization, classical adenocarcinoma, vascular invasion, and presence of positive lymph nodes were found to be associated with RAS-mutated tumors. Interestingly, 39% of the mutated RAS tumors had less than 12 lymph nodes removed, compared with 22% in the wild-type RAS patients.

Table 3.

Association Between KRAS and NRAS Mutation Subtypes and Clinicopathological Features.

Characteristics	RAS Mutant	RAS Wild Type	P	KRAS Mutant	KRAS Wild Type	P	NRAS Mutant	NRAS Wild type	P
Age
Mean (±SD)	59 ± 16	55 ± 14	0.7	56 ± 13	55.5 ± 14	0.8	50 ± 16	56 ± 14	0.1
≤55	33 (44%)	59 (52%)	.04	42 (44%)	89 (51%)	0.1	5 (42%)	126 (49%)	0.4
>55	44 (56%)	74 (48%)		53 (56%)	86 (49%)		7 (58%)	132 (51%)
Gender
Female	60 (56%)	64 (39%)	.003	55 (58%)	69 (40%)	.003	5 (58%)	119 (46%)	0.5
Male	47 (44%)	99 (61%)		40 (42%)	106 (60%)		7 (42%)	139 (54%)
Tumor site
Right colon	22 (21%)	49 (30%)	.02	20 (21%)	51 (29%)	.09	2 (17%)	69 (25%)	0.3
Left colon	85 (79%)	84 (70%)		75 (79%)	124 (71%)		10 (83%)	189 (73%)
Histologic subtype
Adenocarcinoma	89 (83%)	140 (86%)	.05	78 (82%)	151 (86%)	.05	11 (92%)	218 (84%)	0.7
Mucinous adenocarcinoma	16 (15%)	13 (8%)		15 (16%)	14 (8%)		1 (8%)	28 (11%)
Others	2 (2%)	10 (6%)		2 (2%)	10 (6%)		0 (0%)	12 (5%)
Grading
Well	42 (39%)	78 (48%)	.03	39 (41%)	81 (46%)	0.1	3 (25%)	117 (45%)	0.1
poor	65 (61%)	85 (52%)		56 (59%)	94 (54%)		9 (75%)	141 (55%)
Vascular invasion
Presence	25 (24%)	21 (13%)	.01	19 (20%)	27 (15%)	0.2	6 (50%)	40 (16%)	.008
Absence	81 (76%)	142 (87%)		75 (78%)	148 (85%)		6 (50%)	217 (84%)
Perineural invasion
Presence	14 (14%)	19 (12%)	0.1	12 (13%)	21 (15%)	0.4	2 (17%)	31 (12%)	0.4
Absence	88 (86%)	144 (88%)		78 (87%)	154 (85%)		10 (83%)	222 (88%)
Lymph nodes
<12 Lymph nodes examined
Yes	33 (39%)	34 (22%)	.002	27 (37%)	40 (24%)	.02	6 (55%)	61 (27%)	.04
No	51 (61%)	123 (78%)		46 (63%)	128 (76%)		5 (45%)	169 (73%)
Positive lymph node:
Mean (± SD)	2 ± 3	1 ± 4	.07	1 ± 0	2 ± 4	0.2	2 ± 3	1 ± 4	0.9
Presence	45 (51%)	44 (25%)	.003	38 (50%)	59 (30%)	.002	7 (58%)	82 (35%)	.08
Absence	43 (49%)	113 (74%)	<.001	38 (50%)	134 (70%)	0.1	5 (42%)	151 (65%)	0.1
Lymph node ratio, mean ± SD	3.10⁻¹ ± 1	1 ± 0		1 ± 0	1 ± 0		1 ± 10⁻¹	2 ± 4.10⁻¹
Disease stages
I-II	49 (46%)	88 (54%)	.04	43 (45%)	94 (54%)	0.1	6 (50%)	131 (51%)	0.3
III-IV	58 (54%)	75 (46%)		52 (55%)	81 (46%)		6 (50%)	127 (49%)
Chemotherapy adjuvant
No	51 (53%)	71 (46%)	.05	45 (52%)	77 (47%)	0.2	6 (55%)	116 (48%)	0.4
Yes	46 (47%)	83 (54%)		41 (48%)	88 (53%)		5 (45%)	124 (52%)

Table 4.

Association of Rare RAS Mutations With Clinicopathological Features.

Characteristics	Rare Mutations	RAS Wild-Type Status	P
Age
Mean ((±SD)	59 ± 16	55 ± 14	0.7
≤55	11 (50%)	120 (48%)	0.5
>55	11 (50%)	128 (52%)
Gender
Female	13 (41%)	115 (46%)	0.3
Male	9 (59%)	133 (54%)
Tumor site
Right colon	3 (14%)	68 (27%)	0.1
Left colon	19 (86%)	180 (73%)
Histologic subtype
Adenocarcinoma	19 (86%)	210 (85%)	0.3
Mucinous adenocarcinoma	3 (14%)	26 (10%)
Others	0 (0%)	12 (5%)
Grading
Well	4 (18%)	116 (47%)	.007
Poor	18 (82%)	132 (53%)
Vascular invasion
Presence	8 (36%)	38 (15%)	.01
Absence	14 (64%)	209 (85%)
Perineural invasion
Presence	3 (14%)	30 (12%)	0.5
Absence	18 (86%)	214 (88%)
Lymph nodes
<12 Lymph nodes examined
Yes	10 (53%)	57 (26%)	.01
No	9 (47%)	165 (74%)
Positive lymph node:
Mean (± SD)	2 ± 3	1 ± 4	.07
Presence	10 (50%)	79 (35%)	0.1
Absence	10 (50%)	146 (65%)	<.001
Lymph node ratio, mean ± SD	3.10⁻¹ ± 1	1 ± 0
Disease stages
I-II	14 (64%)	123 (50%)	0.1
III-IV	8 (36%)	125 (50%)
Chemotherapy adjuvant
No	12 (57%)	110 (48%)	0.2
Yes	9 (43%)	120 (52%)

KRAS-mutated tumors were significantly correlated with female gender and the presence of positive lymph nodes. In total, 37% of these tumors had fewer than 12 lymph nodes removed, compared with 24% in RAS wild-type tumors. NRAS mutations were correlated with the presence of vascular invasion and <12 of removed lymph nodes.

As shown in Table 4, rare RAS-mutated tumors compared with WT tumors were more frequently poorly differentiated (82% vs 53%) and were associated with the presence of vascular invasion and a lower rate of removed lymph nodes (<12 resected lymph nodes).

Clinicopathological Features and Molecular Alterations Between Early and Late Recurrence of Colon Adenocarcinoma

Among the 85 CC patients who experienced tumor recurrence, 60 (71%) with recurrence within 2 years were considered to be in the early recurrence group, whereas 25 (29%) patients with recurrence ≥2 years were considered to be in the late recurrence group. Clinicopathological features and molecular alterations in the different groups are shown in Table 5. In our study, patients with early recurrence were significantly older (P = .02), had more poorly differentiated tumors (72% vs 40%, P = .005), and had a higher rate of positive lymph nodes (3 vs 1; P = .05) than those with late recurrence.

Table 5.

Clinicopathological and Molecular Differences Between Early and Late Recurrence.

Variables	Early Recurrence (n = 60)	Late Recurrence (n = 25)	P
Variables	n (%)	n (%)	P
Age at diagnosis
Mean (SD)	54 (15)	52 (14)	0.1
<55	26 (43)	17 (68)	.02
≥55	34 (57)	8 (32)
Sex
Female	28 (47)	14 (56)	0.1
Male	32 (53)	11 (44)
Primary tumor site
Right colon	19 (32)	6 (24)	0.1
Left colon	41 (68)	19 (76)
Histological subtypes
Adenocarcinoma	48 (80)	21 (84)	0.1
Mucinous sumo	7 (12)	4 (16)
Others	5 (8)	0 (0)
Grading
Well	17 (28)	15 (60)	.005
Poor	43 (72)	10 (40)
Vascular invasion
Presence	8 (14)	5 (20)	0.1
Absence	51 (86)	20 (80)
Perineural invasion
Presence	12 (21)	5 (21)	0.2
Absence	46 (79)	19 (79)
Lymph nodes
<12 Lymph nodes examined
Yes	18 (31)	8 (36)	0.1
No	40 (69)	14 (64)	.05
Nodal positive
Mean (SD)	3 (4)	1 (2)	0.1
Presence	28 (48)	8 (36)	0.8
Absence	30 (52)	14 (64)
Lymph node ratio (mean, SD)	10⁻¹ (0)	10⁻¹ (0)
Stage of the disease
I-II	20 (64)	11 (36)	.08
III-IV	40 (74)	14 (26)
MSI status
MSI	3 (12)	0 (0)	0.3
MSS	23 (88)	9 (100)
RAS status
Wild type	32 (53)	8 (32)	.03
Mutant	28 (47)	17 (68)
KRAS status
Wild type	34 (57)	10 (40)	.05
Mutant	26 (43)	15 (60)
NRAS status
Wild type	58 (97)	23 (92)	0.2
Mutant	2 (3)	2 (8)
Rare mutations
Presence	6 (10)	4 (16)	0.2
Wild-type status	54 (90)	21 (84)
Exons
KRAS Exon 2 mutant	22 (79)	13 (76)	0.2
Outside KRAS exon 2	6 (21)	4 (24)
Mutant KRAS Exon 2
Codon 12 mutant	18 (82)	10 (77)	0.3
Code 13 mutant	4 (18)	3 (23)
KRAS exon 2 amino acid transition (n = 85)
p. G12D	7 (32)	3 (23)	.03
p. G12V	6 (27)	5 (38)
p. G13D	4 (18)	1 (8)
p. G12C	3 (14)	0 (0)
p. G12A	2 (9)	0 (0)
p. G12R	0 (0)	2 (15)
p. G13V	0 (0)	2 (15)
Rare mutation subtypes
p. K117N (kras exon 4)	1 (17)	1 (25)	0.2
p. A146P (kras exon 4)	1 (17)	0 (0)
p. A146T (kras exon 4)	1 (17)	1 (25)
p. A146V (kras exon 4)	1 (17)	0 (0)
p. G12D (Nras exon 2)	2 (33)	0 (0)
p. Q61K (Nras exon 3)	0 (0)	1 (25)
p. Q61L (Nras exon 3)	0 (0)	1 (25)

Regarding molecular alterations, full RAS status and KRAS gene mutations were more common in patients with late recurrence than in those with early recurrence (P = .03; P = .05, respectively). As a point of interest, the KRAS p. G12V mutation subtype was more likely to be found in the late recurrence group than in the early recurrence group (P = .03). There were no differences in the distribution of MSI status and NRAS mutations between the two groups.

Effect of KRAS and NRAS Mutation Status on the Recurrence Pattern

As shown in Table 6, RAS mutant and wild-type tumors differed significantly in their pattern of recurrence. Indeed, RAS mutation status was more likely to be common in patients with lung recurrence than in those with other recurrence sites (P = .02). Furthermore, mutational rates of KRAS were significantly higher in patients who experienced lung recurrence during the follow-up (lung = 74%; local = 54%; Peritoneum = 44%; liver = 38%; others = 30%; P = .04). Interestingly, we found that rare mutations tended to be associated with lung recurrence compared with other recurrence sites (P = .003). However, KRAS exon 2 mutant subtype (12/13), NRAS mutation status, and MSS status were not associated with recurrence in our population.

Table 6.

Recurrence Pattern According to Molecular Alterations.

Molecular Alteration	Local	Liver	Recurrence Pattern Lung	Peritoneum	Others	p
RAS status
Wild type	6 (46)	20 (59)	3 (16)	4 (44)	7 (70)	.02
Mutant	7 (54)	14 (41)	16 (84)	5 (56)	3 (30)
KRAS status
Wild type	6 (46)	21 (62)	5 (26)	5 (56)	7 (70)	.04
Mutant	7 (54)	13 (38)	14 (74)	4 (44)	3 (30)
NRAS status
Wild type	13 (100)	33 (97)	17 (89)	8 (89)	10 (100)	0.4
Mutant	0 (0)	1 (3)	2 (11)	1 (11)	0 (0)
Rare mutations
Rare mutant	0 (0)	1 (3)	7 (37)	1 (11)	1 (10)	.003
Wild-Type status	13 (100)	33 (97)	12 (63)	8 (89)	9 (90)
Mutant KRAS Exon 2
Codon 12	6 (86)	10 (77)	7 (78)	3 (75.0)	2 (100)	0.8
Codon 13	1 (14)	3 (23)	2 (22)	1 (25.0)	0 (0)
MSI status
MSI	0 (0)	2 (18)	0 (0)	0 (0)	1 (20)	0.2
MSS	5 (100)	9 (82)	8 (100)	6 (100)	4 (80)

Recurrence-Free Survival According to Clinicopathological Features and Molecular Biomarkers

The mean RFS of the entire cohort was 30 months. As illustrated in Table 7 and Figure 3, univariate analysis revealed that the clinicopathological features significantly correlated with worse RFS were poorly differentiated tumors (P = .005), advanced TNM stage (P = .01), positive lymph nodes (P = .01), and the presence of perineural invasion (P = .002). With regard to molecular alterations, RFS was significantly poorer in patients with RAS mutant status than in those with wild-type status (57% vs 95%, P < .001) and in patients harboring KRAS mutations than in those with RAS wild-type status (57 vs 94; P < .001). Interestingly, in this study, we found that RFS was significantly shorter in patient with KRAS codon 13 mutations than in those with KRAS codon 12 mutations and KRAS codon 12/13 wild type (P = .001). Rare mutation status was also correlated with worse RFS (P = .001).

Table 7.

Recurrence-Free Survival Analysis.

Variables	Univariate Analysis	P Value	Multivariate Analysis	P Value
Variables	Mean RFS months (95%CI)	P Value	Hazard Ratio (95% CI)	P Value
Age (yr)
≥55	85 (75-95)	0.3
<55	76 (64-87)
Gender
Female	74 (63-85)	0.2
Male	87 (77-97)
Tumor site
Right side	83 (69-97)	0.8
Left side	81 (71-90)
Histologic subtype
Adenocarcinoma	83 (74-91)	0.6
Mucinous adenocarcinoma	76 (53-99)
Others	56 (31-31)
Grading
Well	92 (82-102)	.005	1 (1-2)	.01
Poor	73 (61-85)
Tumor stage
I-II	91 (81-102)	.01	1 (1-2)	.06
III-IV	71 (61-82)
Lymph nodes
<12 Lymph nodes examined
Yes	69 (52-85)	0.1	1 (1-2)	0.5
No	86 (77-95)
Nodal positive
Presence	63 (51-74)	.01
Absence	89 (80-99)
Perineural invasion
Presence	47 (32-63)	.002	2 (0-1)	.003
Absence	87 (79-95)
Vascular invasion
Presence	85 (66-104)	0.6
Absence	81 (73-90)
MSI status
MSI	107 (85-129) 0.2
MSS	83. (73-92)
Full RAS status
Wild type	95 (86-104)	<.001	0 (1-4)	.05
Mutant	57 (47-67)
KRAS status
Wild type	94 (85-103)	<.001	0 (1-7)	0.2
Mutant	57 (47-67)
NRAS status
Wild type	82 (74-90)	0.7
Mutant	55 (30-80)
KRAS exon 2 codon:
Codon 12/13 WT	91 (83-100)	0.001	1 (1-0)	.01
Codon 12 mutant	59 (46-71)
Codon 13 mutant	56 (38-74)
Rare mutation status
Wild type	84 (76-92)	.05	8.10⁻³ (1-4)	.07
Mutant	45 (27-63)

Figure 3.

Kaplan–Meier recurrence-free survival curves according to molecular alterations. (A) Full RAS status, (B) KRAS gene status, (C) rare mutation status, and (D) KRAS exon 2 mutant subtypes.

Multivariate logistic regression analysis revealed that differentiation, perineural invasion, full RAS mutation status, and KRAS codon 13 mutations were independent factors for RFS. Table 7 presents the results.

Discussion

The majority of studies investigating the relationship between recurrence and patient molecular profiles have been conducted in local colon cancer patients and have focused only on KRAS gene status and liver recurrence. To the best of our knowledge, the present study is the first to report the impact of a patient’s molecular profile on cancer recurrence, time of recurrence, recurrence patterns, and recurrence-free survival in the largest cohort in Morocco and the Middle East and North Africa region (MENA region).

First, our results showed that full RAS status and the presence of KRAS mutations were significantly associated with cancer recurrence at all stages. These findings support the results of a previous study demonstrating that RAS mutations, especially KRAS mutations, correlate with poor prognostic factors and have worse oncologic outcomes with a higher risk of recurrence in both locally and advanced colon cancer stages.^4,15,19 The strength of this study is that the KRAS G12V mutation subtype was significantly more frequent in patients developing recurrence after first-line therapy . According to the KRAS amino acid substitution, different signaling pathways are activated, which leads to distinct tumor progression and invasion.

Many studies have explored novel strategies that could predict early recurrence by integrating clinicopathological features and molecular alterations, which could be useful for the surveillance of colon cancer patients.^28,29 In the present study, RAS and KRAS mutations were more common in patients with early recurrence than in those with late recurrence. Lan et al. reported that KRAS was among the top mutated genes in patients with both early and late recurrence.³⁰ Another study showed that KRAS mutations were detected in the cell-free DNA of CRC patients at all stages. Based on these results, the KRAS status may serve as a tool for the early detection of tumor recurrence.³¹ The main strength of this study is that our results demonstrated that patients with KRAS G12V amino acid substitution were more likely to develop early recurrence than those with other KRAS exon 2 mutant subtypes, which has not yet been reported. These results will be helpful in providing personalized follow-up strategies for KRAS exon 2 mutations based on amino acid substitutions.

However, our study demonstrated no difference in the distribution of NRAS mutations between patients with early and late recurrence. Lan et al. reported the same result regardless of tumor location; however, they showed that among left-sided colon cancer patients, NRAS mutations occurred more frequently in patients with early recurrence.³⁰ In the present study, MSI status was not correlated with the timing of recurrence, which may be related to the good prognostic value of MSI tumors reported in previous studies.^27,32

Regarding clinicopathological features, we reported that patients with early recurrence were significantly older than those with late recurrence. However, Lan et al. did not find any association between tumor grade and recurrence.³⁰ In addition, patients with poorly differentiated tumors were more likely to develop early recurrence, which is similar to the findings of Luo et al.¹³

Many studies have demonstrated that RAS mutations, especially KRAS mutations, not only affect the time of recurrence but also their patterns. Vauthey et al. reported that RAS mutation status is associated with an increased incidence of lung recurrence.³³ This result is supported by Kemeny et al.³⁴ Tie et al. showed that KRAS mutations in stages II and III CRC were more prevalent in lung recurrences but less prevalent in liver recurrences.²³ We also investigated the association of KRAS and NRAS mutation subtypes and MSI status with recurrence patterns in our cohort. We found that patients with full RAS mutation status and KRAS mutations were more likely to develop lung recurrence. Our study confirmed that the RAS mutation status was predictive of the involvement of this site. A novel finding of our study is that rare RAS mutations are significantly more common in patients with lung recurrence. On the basis of these results, it appears that rare RAS mutations could be used as predictive factors for lung recurrence at all stages. Further investigations using a large sample of patients with rare RAS mutations are required to validate these results. However, we did not find any association between the NRAS gene status and recurrence patterns. Our findings are supported by those reported by Tie et al.²³

This study is the first to report RFS according to full RAS status, KRAS mutation subtypes (codon/exon), and rare mutations. The majority of studies have evaluated the prognostic impact of only KRAS gene status and have been conducted in patients with II–III CC.^15,35 Our cohort analysis revealed that RAS mutation status was a predictive factor for worse RFS. Similar results were obtained in a large cohort of 2720 patients by Sinicrope et al.³⁶ In addition, KRAS mutant status was associated with a shorter RFS (P < .001) in our population. In line with our study, Vauthey et al. found that patients with KRAS mutant CRC experienced worse RFS (P = .005).³³ This result has been confirmed by many studies.^20,37,38 Furthermore, it has been suggested that KRAS mutations reflect a more invasive tumor process that tends to recur early and after metastatic resection, which explains their impact on RFS.³⁹

We found that rare mutations appeared to have worse RFS compared with the wild-type status. However, this finding has not yet been reported in the literature. In this study, we also documented that KRAS codon 12 mutant status was associated with the worst RFS compared with the KRAS codon 13 mutant and wild-type statuses. Based on our analysis, it appears that the prognostic value of KRAS codon 12 mutations is related to the association between KRAS G12 V substitutions and early recurrence. Consequently, patients with this mutation require specific surveillance modalities. Our findings are consistent with those of previous reports.⁴⁰ Yoon et al. evaluated a cohort of 2478 patients receiving adjuvant FOLFOX alone or in combination with cetuximab and reported that KRAS mutations in codons 12 or 13 were associated with shorter DFS.⁴¹ The mutational status of KRAS codons 12 and 13 appears to be a predictive biomarker for the benefit of adjuvant chemotherapy.

This study has both strengths and limitations. First, this was a single-institution study, which may limit the generalizability of our results to the entire population of our country. Second, detailed data on chemotherapy were not available in this study. Patients who received oral chemotherapy, intravenous chemotherapy, or incomplete chemotherapy were classified as having received adjuvant chemotherapy. Third, other molecular biomarkers strongly involved in CC carcinogenesis (PI3K and P53…), were not tested in the present study; hence, we did not analyze the effect of these variables on the prognosis of our patients. Above all, our study is the first and largest in Morocco and the MENA region.

Conclusion

Our study demonstrated that patients with early recurrence had specific clinicopathological features and harbored more KRAS G12 V substitutions. Patients with KRAS mutations are more likely to develop lung cancer recurrence. Cox regression analysis demonstrated that KRAS codon 12 mutations were the worst predictors of RFS at all stages in our cohort. In contrast, there was no correlation between MSI status and recurrence risk in cases of colon cancer.

Footnotes

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.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Statement

ORCID iDs

Fatima El Agy

Sara Boukansa

References

Sung

Ferlay

Siegel

, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021;71:209-249.

Mantke

Schmidt

Wolff

Kube

Lippert

. Incidence of synchronous liver metastases in patients with colorectal cancer in relationship to clinico-pathologic characteristics. Results of a German prospective multicentre observational study. Eur J Surg Oncol 2012;38:259-265.

van der Pool

AEM

Damhuis

Ijzermans

JNM

, et al. Trends in incidence, treatment and survival of patients with stage IV colorectal cancer: a population-based series. Colorectal Dis 2012;14:56-61.

Bonnot

Passot

. RAS mutation: site of disease and recurrence pattern in colorectal cancer. Chin Clin Oncol 2019;8:55.

Nishizawa

Haraguchi

Kim

, et al. Randomized phase II study of SOX+B-mab versus SOX+C-mab in patients with previously untreated recurrent advanced colorectal cancer with wild-type KRAS (MCSGO-1107 study). BMC Cancer 2021;21:947.

Kunst

Alarid-Escudero

Aas

Coupé

VMH

Schrag

Kuntz

. Estimating population-based recurrence rates of colorectal cancer over time in the United States. Cancer Epidemiol Biomarkers Prev 2020;29:2710-2718.

Caldarella

Crocetti

Messerini

Paci

. Trends in colorectal incidence by anatomic subsite from 1985 to 2005: a population-based study. Int J Colorectal Dis 2013;28:637-641.

Van

Cervantes

Nordlinger

, et al. ESMO guidelines working group. Metastatic colorectal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol Off J Eur Soc Med Oncol 2014;25:1-9.

Kim

Cho

Lee

Kim

. Pulmonary metastasectomy for colorectal cancer: how many nodules, how many times? World J Gastroenterol 2014;20:6133-6145.

10.

Franko

Shi

Meyers

, et al. Prognosis of patients with peritoneal metastatic colorectal cancer given systemic therapy: an analysis of individual patient data from prospective randomised trials from the Analysis and Research in Cancers of the Digestive System (ARCAD) database. Lancet Oncol 2016;17:1709-1719.

11.

Christensen

Jensen

Larsen

Nielsen

. Systematic review: incidence, risk factors, survival and treatment of bone metastases from colorectal cancer. J Bone Oncol 2018;13:97-105.

12.

Brown

KGM

Koh

. Surgical management of recurrent colon cancer. J Gastrointest Oncol 2020;11:513-525.

13.

Luo

Yang

Shan

, et al. Clinicopathological features of stage I–III colorectal cancer recurrence over 5 Years after radical surgery without receiving neoadjuvant therapy: evidence from a large sample study. Front Surg 2021;8:666400.

14.

Böckelman

Engelmann

Kaprio

Hansen

Glimelius

. Risk of recurrence in patients with colon cancer stage II and III: a systematic review and meta-analysis of recent literature. Acta Oncol 2015;54:5-16.

15.

Kim

Heo

Lee

, et al. The impact of KRAS mutations on prognosis in surgically resected colorectal cancer patients with liver and lung metastases: a retrospective analysis. BMC Cancer 2016;16:120.

16.

Zhang

Mei

Pei

, et al. A modified tumor-node-metastasis classification for primary operable colorectal cancer. JNCI Cancer Spectr 2021;5;pkaa093.

17.

Yaeger

Cowell

Chou

, et al. RAS mutations affect pattern of metastatic spread and increase propensity for brain metastasis in colorectal cancer. Cancer 2015;121:1195-1203.

18.

Saadat

Boerner

Goldman

, et al. Association of RAS mutation location and oncologic outcomes after resection of colorectal liver metastases. Ann Surg Oncol 2021;28:817-825.

19.

El Agy

El Bardai

El Otmani

, et al. Mutation status and prognostic value of KRAS and NRAS mutations in Moroccan colon cancer patients: a first report. PLoS One 2021;16: 0248522.

20.

Brudvik

Kopetz

Conrad

Aloia

Vauthey

. Meta-analysis of KRAS mutations and survival after resection of colorectal liver metastases. Br J Surg 2015;102:1175-1183.

21.

Bisht

Ahmad

Sawaimoon

Bhatia

Das

. Molecular spectrum of KRAS, BRAF, and PIK3CA gene mutation: determination of frequency, distribution pattern in Indian colorectal carcinoma. Med Oncol 2014;31:124.

22.

Petrelli

Coinu

Cabiddu

Ghilardi

Barni

. KRAS as prognostic biomarker in metastatic colorectal cancer patients treated with bevacizumab: a pooled analysis of 12 published trials. Med Oncol 2013;30:650.

23.

Tie

Lipton

Desai

, et al. KRAS mutation is associated with lung metastasis in patients with curatively resected colorectal cancer. Clin Cancer Res 2011;17:1122-1130.

24.

Diamantis

Ioannis

Fabio

, et al. Clinical significance and prognostic relevance of KRAS, BRAF, PI3K and TP53 genetic mutation analysis for resectable and unresectable colorectal liver metastases: a systematic review of the current evidence. Surg Oncol 2018;27:280-288.

25.

Brudvik

Vauthey

. Surgery: KRAS mutations and hepatic recurrence after treatment of colorectal liver metastases. Nat Rev Gastroenterol Hepatol 2017;14:638-639.

26.

Rimbert

Tachon

Junca

Villalva

Karayan-Tapon

Tougeron

. Association between clinicopathological characteristics and RAS mutation in colorectal cancer. Mod Pathol 2018;31:517-526.

27.

El Agy

Otmani

Mazti

, et al. Implication of microsatellite instability pathway in outcome of colon cancer in Moroccan population. Dis Marks 2019;2019:3210710.

28.

Dai

, et al. A robust gene signature for the prediction of early relapse in stage I-III colon cancer. Mol Oncol 2018;12:463-475.

29.

Dai

Xiang

, et al. Prognostic and predictive value of an autophagy-related signature for early relapse in stages I-III colon cancer. Carcinogenesis 2019;40:861-870.

30.

Lan

Chang

Lin

, et al. Clinicopathological and molecular features of patients with early and late recurrence after curative surgery for colorectal cancer. Cancers 2021;13:1883.

31.

Wang

Hsieh

Chang

, et al. Molecular detection of APC, K- ras, and p53 mutations in the serum of colorectal cancer patients as circulating biomarkers. World J Surg 2004;28:721-726.

32.

Guinney

Dienstmann

Wang

, et al. The consensus molecular subtypes of colorectal cancer. Nat Med 2015;21:1350-1356.

33.

Vauthey

Zimmitti

Kopetz

, et al. RAS mutation status predicts survival and patterns of recurrence in patients undergoing hepatectomy for colorectal liver metastases. Ann Surg 2013;258:619-626.

34.

Kemeny

Chou

Capanu

, et al. KRAS mutation influences recurrence patterns in patients undergoing hepatic resection of colorectal metastases. Cancer 2014;120:3965-3971.

35.

Hutchins

Southward

Handley

, et al. Value of mismatch repair, KRAS, and BRAF mutations in predicting recurrence and benefits from chemotherapy in colorectal cancer. J Clin Oncol 2011;29:1261-1270.

36.

Sinicrope

Shi

Smyrk

, et al. Molecular markers identify subtypes of stage III colon cancer associated with patient outcomes. Gastroenterology 2015;148:88-99.

37.

Schneider

Eden

Pache

, et al. Mutations of RAS/RAF proto-oncogenes impair survival after cytoreductive surgery and HIPEC for peritoneal metastasis of colorectal origin. Ann Surg 2018;268:845-853.

38.

Passiglia

Bronte

Bazan

Galvano

Vincenzi

Russo

. Can KRAS and BRAF mutations limit the benefit of liver resection in metastatic colorectal cancer patients? A systematic review and meta-analysis. Crit Rev Oncol Hematol 2016;99:150-157.

39.

Serova

Astorgues-Xerri

Bieche

, et al. Epithelial-to-mesenchymal transition and oncogenic Ras expression in resistance to the protein kinase Cbeta inhibitor enzastaurin in colon cancer cells. Mol Cancer Ther 2010;9:1308-1317.

40.

Pereira

Rego

Morris

, et al. Association between KRAS mutation and lung metastasis in advanced colorectal cancer.Br J Cancer 2015;112:424-428.

41.

Yoon

Tougeron

Shi

, et al. KRAS codon 12 and 13 mutations in relation to disease-free survival in BRAF-wild-type stage III colon cancers from an adjuvant chemotherapy trial (N0147 alliance). Clin Cancer Res 2014;20:3033-3043.

RAS Mutations Predict Recurrence-Free Survival and Recurrence Patterns in Colon Cancer: A Unicenter Study in Morocco

Abstract

Purpose

Methods

Results

Conclusion

Keywords

Introduction

Materials and Methods

Sample Size Calculation

Study Design

Inclusion and Exclusion Criteria

Histopathological Study

Molecular Analysis

KRAS, NRAS, and BRAF Mutation Profiling

DNA Extraction

Polymerase Chain Reaction and Direct Sequencing

Pyrosequencing

Analysis of the Microsatellite Status

Recurrences

Statistical Analysis

Results

Patients and Recurrence Characteristics

Frequencies of Molecular Alterations and Recurrence

Correlation between molecular alterations and clinical pathological features

Clinicopathological Features and Molecular Alterations Between Early and Late Recurrence of Colon Adenocarcinoma

Effect of KRAS and NRAS Mutation Status on the Recurrence Pattern

Recurrence-Free Survival According to Clinicopathological Features and Molecular Biomarkers

Discussion

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

Ethical Statement

ORCID iDs

References