Key findings

  • 180consecutive colorectal carcinomas from central and northern Maine, sequenced on a 50-gene hotspot panel in 2017–2018
  • 61%of tumors carried a TP53 mutation, the most frequent; APC (41%) and KRAS (34%) followed
  • HR 6.24for death with concurrent APC, KRAS and TP53 mutation, adjusted for advanced stage and age (95% CI 2.83–13.76)
  • 10.4 monthsapproximate median overall survival with the mutation triad, versus not reached in triad-negative cases

Abstract

Purpose
To test whether consecutive colorectal carcinomas treated in a rural Maine community practice recover the published driver-mutation spectrum, and whether concurrent APC, KRAS, and TP53 alteration identifies inferior overall survival independent of stage and mismatch-repair status.
Methods
We analyzed consecutive colorectal carcinoma cases diagnosed and profiled at our pathology and molecular laboratory during 2017–2018 and subsequently treated at Northern Light Cancer Institute. Molecular abnormalities identified by a 50-gene hotspot next-generation sequencing panel (AmpliSeq Cancer HotSpot Panel v2.0, Thermo Fisher) were correlated with clinical characteristics and overall survival. Additional clinicopathologic variables were abstracted from institutional records.
Results
The analytic cohort comprised 180 patients, of whom 172 had complete clinical follow-up (103 men [57%], 77 women [43%]; median age 66 years, range 28–95). Nearly all patients with documented race were White (99%); 42% were smokers; 71% had body-mass index ≥25; 48% presented with early-stage (I–II) and 51% with advanced-stage (III–IV) disease. Fifteen percent of tumors were microsatellite-high or mismatch-repair deficient. The most frequently detected mutations were TP53 (61%), APC (41%), KRAS (34%), PIK3CA (21%), BRAF (17%), and SMAD4 (10%). After a median follow-up of 17 months, 77% of the survival-evaluable cohort (n = 172) were alive. Independent of strongly predictive advanced clinical stage (p < 0.001), patients with concurrent APC, KRAS, and TP53 mutations had significantly worse overall survival (log-rank p < 0.001). This association was independent of microsatellite-high status.
Conclusion
Next-generation sequencing in a regional community oncology setting identified established prognostic and therapeutic markers and additional molecular features that correlated with outcome. The findings support prognostic stratification based on gene mutations outside current guideline-mandated testing, particularly concurrent APC, KRAS, and TP53 alterations, consistent with the multi-hit model of colorectal tumorigenesis and progression.

Introduction

Colorectal carcinoma remains the third most commonly diagnosed cancer in the United States and a leading cause of cancer death.1 Survival is still favorable when the tumor is found early and falls quickly once it has spread. Five-year relative survival is approximately 91% for localized disease, 73% for regional disease, and 15% once distant metastases are present.1 That drop is why guideline-based treatment decisions now require mutational testing. The required reportable markers, however, are only part of the genotype. Alterations outside that short list also sit in the pathways that drive progression and metastasis, and we think those additional mutations leave a survival signature. The question this paper asks is more focused than the full genomic landscape. Among the extra mutations returned on the same panel are APC, KRAS, and TP53, which together are the classic multi-hit combination in this disease. When a community laboratory sequences colorectal carcinomas as they present for care, do the familiar mutations appear at the expected frequencies, and do tumors that carry APC, KRAS, and TP53 together have a worse outcome?

The idea that colorectal carcinoma is a multi-hit disease is not new. Fearon and Vogelstein argued more than three decades ago that an adenoma becomes a carcinoma only after several genes are altered, classically APC, then KRAS, then TP53.2,3 A single mutation, even a bona fide driver, is not enough. That view has held up. Isolated cancer-associated variants are now found in phenotypically normal mucosa, and recent co-mutation analyses across more than 70,000 tumors show that Wnt- and ERK-pathway pairings are selected in colorectal carcinoma but are rare in healthy tissue.4 The event that appears to matter biologically is the combination, not the isolated hotspot.

The gene list itself is also settled. The Cancer Genome Atlas and later metastatic sequencing series place APC, TP53, KRAS, PIK3CA, and BRAF at the top of the frequency table.5,6 APC mutation has been tied to prognosis in a multigene classifier,7 and Peng and colleagues described an APC–KRAS–TP53 haplotype associated with differentiation and survival.8 In metastatic disease, Datta et al. showed that co-altered RAS or BRAF plus TP53 marks extremes of survivorship,9 and combined RAS and TP53 status stratifies survival after resection of liver metastases.10,11 Those studies were performed largely in academic or metastatic cohorts, and most emphasize two-gene rather than three-gene genotypes. Whether the full APC/KRAS/TP53 sequence still predicts outcome in consecutive, all-stage patients treated in a regional practice has not been shown.

Treatment decisions, however, still rest on one gene or one pathway at a time. Colon and rectal guidelines, multi-society biomarker statements, and European consensus documents ask for RAS and BRAF testing to select anti-EGFR and BRAF-targeted therapy, mismatch-repair or microsatellite testing to select checkpoint blockade, and PIK3CA testing in selected adjuvant settings.12–15 KRAS mutation was first linked to outcome in the RASCAL analyses and later shown to predict lack of benefit from anti-EGFR antibodies.16,17 Metastatic tumors with BRAF V600E are treated with combined BRAF and EGFR blockade, first in the BEACON regimen and now in the first-line BREAKWATER regimen.18,19 Mismatch-repair–deficient tumors respond to PD-1 blockade.20,21 Low-dose aspirin reduced recurrence among patients with PI3K-altered localized tumors in the ALASCCA trial.22 The same panels that return those results also report APC, TP53, SMAD4, and other co-alterations. Those extra findings are not used for risk stratification in community practice. We hypothesized that they should be: specifically, that concurrent APC, KRAS, and TP53 mutation identifies inferior overall survival after accounting for advanced stage and microsatellite-high or mismatch-repair–deficient status.

There is a practical reason to test that idea here rather than only in a referral series. Rural patients travel farther, see fewer oncologists, and enter fewer trials. Maine is a high-incidence rural state where those limits are part of daily care, not a footnote.23 Most of the genomic studies that shape treatment algorithms come from academic and metropolitan centers. If a community laboratory cannot recover the same mutation spectrum, or if a co-mutation survival signal is simply a product of who gets referred, those algorithms will not travel. If reflexive standing-order sequencing in a regional laboratory recovers both the published frequencies and an independent survival signal, two things follow. The biology is not confined to academic case-mix, and the local workflow is already good enough to put that information in front of the treating team.24,25 We therefore reviewed consecutive colorectal carcinomas profiled in 2017–2018 from central and northern Maine. The first aim was descriptive: do gene-level and hotspot frequencies on a 50-gene panel match published academic series? The second aim was the hypothesis test: does concurrent APC, KRAS, and TP53 mutation associate with inferior overall survival independent of stage and microsatellite or mismatch-repair status?

Methods

Patient selection

All consecutive CRC cases tested by NGS during the period of January 2017 to December 2018 were retrieved from the NGS database at Dahl-Chase Pathology Laboratory and Northern Light Cancer Institute. This work was performed in concordance with Northern Light Institutional Review Board (IRB) protocol (IRB #20-1-A-001). CRCs arising in the colon or rectum or involving metastatic sites when the primary tumor was unavailable, were utilized for testing. A molecular pathologist confirmed the diagnosis and ensured a minimum of 20% tumor content in each tested sample. Staging was performed according to the AJCC TNM classification current at the time of diagnosis. The patients received treatment in accordance with the NCCN guidelines valid at the time of this study (early targeted therapy/pre-immunotherapy era).12

Reflexive molecular testing utilizing a standing order

To perform molecular testing with short turnaround time, avoid the need for individual test requests and preserve tissue, our institution utilizes a “reflexive” molecular standing order. Updated regularly to ensure it includes all NCCN-endorsed biomarkers and meets the current standards of care, this standing order is approved by the institutional molecular tumor board in conjunction with hospital medical staff and is initiated automatically by the pathologist, as soon as the diagnosis of colorectal adenocarcinoma is established on an appropriate specimen type.

DNA extraction and sample preparation

A pathologist confirmed the presence of sufficient tumor tissue for NGS analysis by direct microscopic visualization of hematoxylin-eosin–stained slides in each case. Ten 4–5 µm formalin-fixed, paraffin-embedded sections were marked for manual microdissection of genomic DNA. Approximately 5-15 ng of DNA was amplified, fragmented, ligated to adapters, barcoded, and clonally amplified onto beads to create DNA libraries using the Ion AmpliSeq™ Cancer HotSpot Panel v2.0 and subjected to NGS on the Ion Torrent PGM or S5 instrument (all Life Technologies/Thermo Fisher, Waltham, MA).

Mutational analysis and variant interpretation

The AmpliSeq™ CHPv2.0 panel examines 2800 mutation hotspots in 50 oncogenes and tumor suppressor genes previously described as somatic mutations in human cancers, including ABL1, AKT1, ALK, APC, ATM, BRAF, CDH1, CDKN2A, CSF1R, CTNNB1, EGFR, ERBB2, ERBB4, EZH2, FBXW7, FGFR1, FGFR2, FGFR3, FLT3, GNA11, GNAQ, GNAS, HNF1A, HRAS, IDH1, IDH2, JAK2, JAK3, KDR, KIT, KRAS, MET, MLH1, MPL, NOTCH1, NPM1, NRAS, PDGFRA, PIK3CA, PTEN, PTPN11, RB1, RET, SMAD4, SMARCB1, SMO, SRC, STK11, TP53, and VHL. Sequencing reads were aligned to the hg19 (GRCh37) reference genome. Variant analysis was performed using Ion Torrent Suite (v5.8.0; Thermo Fisher) and Golden Helix VarSeq (v2.0.2; Golden Helix Inc., Bozeman, MT). The lower limit of detection was established at a 5% variant allele frequency (VAF), as published in our prior work.24,26 Genomic alterations were reviewed by a molecular pathologist, tiered by clinical relevance, and reported with guideline-based therapeutic implications, as applicable. Mismatch-repair protein expression (MLH1, PMS2, MSH2, MSH6) was assessed by immunohistochemistry using established criteria used in routine patient care. Microsatellite instability (MSI) testing was performed using the MSI Analysis System, Version 1.2 (Promega Corporation, Madison, WI) according to the manufacturer's protocol [TM255]. PCR was performed on genomic DNA (gDNA) extracted from patient tumor and normal FFPE tissue specimens, amplifying seven loci (five mononucleotide repeats (BAT25, BAT26, NR21, NR24 and MONO 27) and two pentanucleotide repeats (PentaC and PentaD). Fluorescently labeled PCR products were subjected to capillary electrophoresis using an AB3130. Allele sizing data were compared for normal and tumor samples using GeneMapper v4.0 and the presence of MSI was determined by the appearance of new alleles in the tumor sample. The pentanucleotide repeats serve only to verify specimen identity and were not used to determine the final MSI status.

Statistical analysis

Comparisons of proportions across groups were evaluated using Fisher’s exact test or chi-square test, as appropriate to sample size. Molecular abnormalities were correlated with age, gender, smoking history, BMI, tumor location, pathologic stage, and MSI status. The primary endpoint was overall survival (OS), measured from the date of diagnosis to death from any cause; patients alive at last contact were censored. The survival-evaluable cohort was restricted to patients with known vital status and documented follow-up time (n = 172). Six patients lacked vital status; two additional patients with known vital status lacked a documented follow-up interval and were excluded from time-to-event analysis (survival set n = 172). Survival curves were estimated by the Kaplan–Meier method and compared using log-rank test. Cox proportional hazards regression was used for multivariable analysis. All tests were two-sided with significance set at p < 0.05. Analyses were performed using SPSS (IBM Corp., Armonk, NY) and R Statistical Software (v4.6.1; R Core Team 2024) with survival package (v3.8-6; Therneau 2026). Tumor location was classified from the primary site as right (cecum to transverse colon, including the appendix), left colon (descending to sigmoid colon, including rectosigmoid), and rectum (rectum only). Associations between the clinical features including location and the detected mutations were tested with two-sided Fisher exact tests; a microsatellite-stable-only analysis was performed to further exclude confounding by MSI-high disease.

Visualization of mutation distribution

Domain-annotated “lollipop” plots were generated to display the spatial distribution of non-synonymous mutations relative to functional protein domains for the most frequently altered genes. Mutation counts and amino-acid positions were extracted from the institutional NGS database. Plots were constructed in R (v4.6.1; R Core Team 2024)27 using the ggplot2 package.28,29 High-resolution (300 dpi) PNG and vector PDF files were exported for manuscript inclusion.

Results

Specimen types, clinical features and treatment modalities

The specimens comprised 98 excisions and 82 biopsies, including 157 primary tumors and 23 metastatic sites (most commonly liver). Histologically, moderately differentiated adenocarcinomas predominated (>90%). The patient cohort consisted of 103 (57%) men and 77 (43%) women, ages 28 to 95 years (median approximately 66 years), living in central and northern Maine. Among patients with documented race, nearly all were White/Caucasian (130/131; 99%); race was not documented in 49 patients (27%). Approximately 42% were current or former smokers and 71% had a BMI of 25 or greater. Overall, 48.3% (87/180) presented with early-stage disease (stages I–II) and 51.1% (92/180) with advanced-stage disease (stages III–IV). MSI-high / MMR-deficient tumors comprised 27 of the 180 cases (15.0%; MSI-H and/or IHC loss of MLH1, PMS2, MSH2, or MSH6). Treatment was stage-adapted according to institutional practice aligned with NCCN guidance at the time of the study (“pre-immunotherapy era”), including surgical removal only for early stages, accompanied by systemic or targeted therapies for advanced stages. After a median follow-up of 17 months, 77% of the survival-evaluable cohort (n=172) remained alive at last contact (39 deaths). Detailed clinicopathologic features are summarized in Table 1.

Table 1. Clinicopathological features of the rural Maine colorectal carcinoma cohort profiled by next-generation sequencing.

CharacteristicFull NGS cohort (n = 180)
Demographic features
Age, years – median (range)66 (28–95)
Age, years – mean ± SD66.2 ± 13.5
Male103 (57%)
Female77 (43%)
White / Caucasian130 (72%)
African American1 (1%)
Race not documented49 (27%)
Specimen characteristics
Excision98 (54%)
Biopsy82 (46%)
Primary tumor sample157 (87%)
Metastatic site23 (13%)
Tumor characteristics
Right-sided (cecum–transverse)72 (40%)
Left colon (descending–sigmoid, including rectosigmoid)74 (41%)
Rectum31 (17%)
Unspecified colonic location3 (2%)
Well/moderately differentiated (low-grade)162 (90%)
Poorly differentiated (high-grade)18 (10%)
Clinical stage at diagnosis
Stage I38 (21.1%)
Stage II49 (27.2%)
Stage III48 (26.7%)
Stage IV44 (24.4%)
Stage unknown / not documented1 (0.6%)
Mismatch repair / MSI status
MSS / MSI-low153 (85.0%)
MSI-high / MMR-deficient27 (15.0%)
Follow-up and vital status
Median follow-up, months17
Alive at last follow-up135 (75%)
Deceased39 (22%)
Vital status unknown6 (3%)
Survival-evaluable setn = 172
Alive (survival set)133 (77%)
Deceased (survival set)39 (23%)

Survival analyses used 172 patients with known vital status and follow-up (133 alive, 39 deaths); six lacked vital status and two lacked a follow-up interval. Microsatellite-high or mismatch-repair–deficient includes molecular microsatellite-high status or immunohistochemical loss of MLH1, PMS2, MSH2, or MSH6; all other tumors are microsatellite-stable or microsatellite-low. Stage follows simplified AJCC grouping. Percentages may not total 100 because of rounding.

Mutation frequencies

Mutations were detected most frequently in TP53 (61%), APC (41%, 74/180), KRAS (34%), PIK3CA (21%), BRAF (17%), and SMAD4 (10%) genes (Figure 1.) Spatial distribution of mutations within their respective protein domains is shown in Figure 2 for NCCN-guideline genes (KRAS, BRAF, NRAS, and PIK3CA) and in Figure 3 for the most commonly detected extended-panel genes (TP53, SMAD4 and APC).

Figure 1. Mutation frequencies in 180 rural Maine colorectal carcinomas sequenced with a 50-gene hotspot panel. Bars are ranked by count, with number and percent annotated above each bar. Green bars are guideline therapeutic markers (KRAS, NRAS, BRAF, PIK3CA); blue bars are other panel genes above 3%. Totals exceed 180 because many tumors carried more than one mutation.
Figure 2. Domain-annotated lollipop plots of guideline-panel mutations: (A) KRAS, (B) BRAF, (C) NRAS, and (D) PIK3CA. Colored rectangles mark functional domains (UniProt coordinates).29
Figure 3. Domain-annotated lollipop plots of extended-panel mutations: (A) APC, (B) SMAD4, and (C) TP53. Recurrent hotspots (n ≥ 3) are filled red circles; rarer variants are open blue circles. Ten TP53 splice-acceptor events lack an amino-acid coordinate and are noted on the panel.

Table 2. Hotspot mutations by gene in the rural Maine colorectal carcinoma cohort, with guideline genes listed first.

Gene (category)Gene-level frequency in cohortPredominant hotspot variants (n)Reported CRC frequency (literature)
KRAS
(NCCN)
34% (61/180)G12D (18), G12V (15), G13D (17); G12C (2); Q61 (3); other (6)≈35–45% overall; among KRAS-mut: G12D ≈29–35%, G12V ≈21–24%, G13D ≈13–18%, G12C ≈6–7% (COSMIC/large NGS series)
BRAF
(NCCN)
17% (31/180)V600E (≈28); non-V600 (≈3, e.g., D594N)≈8–15% overall; V600E predominates (~80–90% of BRAF-mut CRC)
NRAS
(NCCN)
4% (7/180)Q61K (≈4); other Q61/G13 (few)≈3–5% overall; Q61 and codon 12/13 hotspots
PIK3CA
(NCCN)
21% (37/180)E545K (13), H1047R (9), E542K (3); other helical/kinase variants≈15–25% overall; E545K, E542K, H1047R are canonical hotspots (OncoKB oncogenic)
TP53
(extended)
61% (111/180)R175H (9); R248Q/W (≈8); R273H/C (≈9); R282W (5); G245S (5); splice-acceptor (10); R306* (4)≈55–65% overall; DNA-binding hotspots R175, R248, R273, R282 predominate (~30–40% of TP53-mut CRC)
APC
(extended)
41% (74/180)R1468* (≈12); R894* (≈6–7); Q1385* (≈3–4); E1327fs and other MCR truncations≈60–80% in conventional (MSS) CRC in whole-gene series; lower on hotspot-limited panels; MCR truncations classic
SMAD4
(extended)
10% (18/180)R361H (3); other MH2-domain missense/truncating (1 each)≈10–15% in advanced CRC; R361 is a recurrent MH2 hotspot

Hotspots are listed by protein change when available. Gene-level percentages use the analytic cohort (n = 180). Literature ranges are approximate frequencies from unselected multi-institutional series. MCR denotes the APC mutation cluster region.

Correlation of mutations with clinical features and survival

After a median follow-up of 17 months (range, 0.7–36.0 months), overall survival (OS) for the survival-evaluable cohort (n=172 with ascertainable vital status and follow-up time) was approximately 77% alive at last contact (39 deaths). Advanced clinical stage (III–IV) was strongly associated with shorter OS (log-rank p < 0.001; univariate Cox HR 13.5). MSI-high / MMR-deficient status was not significantly associated with OS on univariate analysis (p=0.13). Concurrent APC, KRAS, and TP53 co-mutation (“mutation triad”) was present in 16 of 180 patients (8.9%) in the full NGS cohort; 14 of these patients met survival-evaluable criteria (known vital status and documented follow-up; 9 all-cause deaths) and form the basis of the Kaplan–Meier and Cox analyses. All 9 of the triad-positive deaths occurred in patients with stage III–IV disease (8 stage IV, 1 stage III); none of the 5 survival-evaluable triad-positive patients with stage I–II disease had died at last contact. Across the full survival set, 36 of 39 deaths were in stage III–IV disease and 3 were in stage I–II disease. Kaplan–Meier analysis demonstrated that the patient subgroup with this mutation triad had significantly worse OS (log-rank p < 0.001; univariate Cox HR 5.61; median OS approximately 10.4 months versus not reached in mutation triad-negative cases). In multivariate Cox regression, the mutation triad remained independent of advanced stage and age (triad HR 6.24, 95% CI 2.83–13.76, p < 0.001; advanced stage HR 14.4, 95% CI 4.43–47.1, p < 0.001; age per year HR 1.02, 95% CI 0.999–1.05, p = 0.057; Supplemental Table S5a). Adding MSI-high/MMR-deficient status to the model did not alter the mutation triad effect (triad mutation HR 6.08, 95% CI 2.77–13.34, p < 0.001; advanced stage HR 13.4, 95% CI 4.09–44.0, p < 0.001; MSI-H HR 0.76, 95% CI 0.23–2.50, p=0.65; Supplemental Table S5b). PIK3CA mutation (37/180 cases) was not associated with all-cause OS (9 deaths among 37 PIK3CA-mutant patients versus 30 among 135 PIK3CA-wild-type patients; univariate Cox HR 1.00, 95% CI 0.48–2.12, p=0.99). KRAS co-occurred in 21 of the 37 PIK3CA-mutant tumors (56.8%) versus 40 of 143 PIK3CA-wild-type tumors (28.0%; OR 3.38, p=0.002). Individual APC, KRAS, and TP53 mutations were not significantly associated with death on univariate analysis; in a multivariate model with advanced stage, only TP53 retained a modest independent association (HR 2.21, p = 0.035), underscoring that the co-mutational triad, not single gene mutations, drove the outcome. Nested Cox models were then used to ask whether the third hit added risk beyond a double mutation. Adding the triad indicator to a model that already contained the APC+KRAS pair and stage improved fit (likelihood-ratio test p = 0.009); analogous tests versus KRAS+TP53 and APC+TP53 pairs were also significant (p < 0.001 and p = 0.0009). These incremental associations rest on 9 deaths among 14 triad-positive patients and should be interpreted as a large but imprecise effect. A larger multi-institutional cohort will be required to determine whether this third-hit increment is a stable, transportable risk signal or an effect inflated by a small number of early deaths. Clinical stage did not significantly correlate with APC, KRAS, TP53, BRAF, PIK3CA, SMAD4, or the triad mutation genotype (all Fisher exact p > 0.05); MSI-high / MMR-deficient tumors showed a non-significant trend toward early-stage disease (p = 0.059). KRAS mutation was more frequent in right-sided than left-sided colon tumors (43.1% vs 25.7%; OR 2.19, p = 0.036), whereas APC, TP53, and concurrent APC/KRAS/TP53 co-mutation did not differ significantly among tumor locations (Supplemental Tables S4a–S4b). The KRAS right-versus-left association persisted in the MSS/pMMR subset (26/54 [48.1%] vs 18/67 [26.9%]; OR 2.53, 95% CI 1.18–5.40, p = 0.022; n = 121).

Figure 4. Kaplan–Meier overall survival by concurrent APC, KRAS, and TP53 mutation. The triad-positive group (red) had inferior survival compared with all other genotypes (blue) (log-rank p < 0.001). Tick marks are censored observations. Time is months from diagnosis.

Discussion

This study demonstrates that extended NGS profiling is feasible in a routine community/rural laboratory setting. The mutation spectrum detected in our patients is similar to those reported from larger academic and multi-institutional series6,30. Two of the five most frequently detected alterations (TP53 and APC) are not currently mandated by NCCN guidelines for therapeutic decision-making, yet they carry biologic and potential prognostic weight. The central finding of this series is that concurrent APC, KRAS, and TP53 mutation (“mutation triad”) identifies a molecularly defined high-risk subset with significantly inferior overall survival, independent of advanced clinical stage, MSI-H status, or tumor location (Supplemental Tables S4a–S4b, S5a–S5b). This observation supports the emerging view that co-alteration of multiple canonical driver genes is clinically meaningful and drives patient prognosis.7

Gene-level mutation frequencies ranked TP53 (61%), APC (41%, 74/180), KRAS (34%), PIK3CA (21%), BRAF (17%), and SMAD4 (10%) as the most common alterations (Figure 1; Table 2). Among NCCN-guideline genes, KRAS mutations clustered at codons 12 and 13 (G12D, G12V, and G13D predominating); BRAF was dominated by V600E; NRAS alterations centered on codon 61; and PIK3CA mutations concentrated in the helical (E545K, E542K) and kinase (H1047R) domains. Among extended-panel genes, APC truncations clustered in the mutation cluster region (including recurrent R1468*), SMAD4 involved the MH2 domain (including R361), and TP53 mapped predominantly to DNA-binding-domain hotspots (R175, R248, R273, R282, G245), with additional splice-acceptor events. Domain-annotated lollipop plots displaying the spatial distribution of the above mutations are shown in Figure 2. These allele hierarchies match population-scale resources: COSMIC, cBioPortal/TCGA colorectal datasets, and OncoKB report comparable gene-level frequencies and the same dominant hotspots.5,31–34 Concordance with multi-institutional series indicates that the rural Maine cohort is molecularly similar rather than geographically unique.

The clinical significance of these hotspot subtypes is increasingly actionable. BRAF V600E confers a poor prognosis and relative chemoresistance; these tumors are treated with combined BRAF and EGFR blockade using encorafenib plus cetuximab with chemotherapy in the metastatic setting. BRAF V600E meets NCCN- and FDA-endorsed criteria for encorafenib plus cetuximab (with or without chemotherapy) in metastatic disease, supported by the phase 3 BREAKWATER trial demonstrating improved progression-free and overall survival with first-line encorafenib–cetuximab–mFOLFOX6 versus standard care.19 Activating KRAS mutations predict resistance to anti-EGFR monoclonal antibodies in metastatic colorectal cancer.16,17 Importantly, KRAS subtype matters both prognostically and therapeutically. Several series have reported differences in overall survival among specific KRAS alleles (including G12C versus G12D or other codon-12 variants), supporting allele-level stratification rather than treating all KRAS mutations as equivalent.35 Patients harboring KRAS G12C mutations may benefit from combined KRAS G12C and EGFR inhibition with adagrasib plus cetuximab or sotorasib plus panitumumab after prior fluoropyrimidine-, oxaliplatin-, and irinotecan-based chemotherapy.36,37 These agents are allele-specific covalent inhibitors that require the unique cysteine residue of G12C and therefore have no activity against G12D-mutant tumors. The FDA granted accelerated approval to adagrasib plus cetuximab in June 2024 and approved sotorasib plus panitumumab in January 2025 for these previously treated populations. Of note, the confirmatory phase 3 KRYSTAL-10 trial subsequently found that adagrasib plus cetuximab did not significantly improve progression-free or overall survival compared with second-line chemotherapy, although its response rate was higher; therefore, claims of superior survival for this combination are not supported at present. Although rare in our series (n=2), KRAS G12C is now an approved therapeutic target in the indicated setting. KRAS G12D, the most frequent KRAS allele in this series, is not yet paired with an approved inhibitor in colorectal cancer but is an active inclusion criterion for investigational G12D-selective and pan-RAS agents. With only two KRAS G12C cases in the present cohort, allele-specific survival comparisons were not feasible. Larger multi-institutional series will be required to determine whether individual KRAS codon-12/13 alleles (G12C, G12D, G12V, G13D) carry independent prognostic weight beyond their contribution to the APC/KRAS/TP53 triad. Unlike breast and prostate cancer, colorectal cancer currently lacks approved PIK3CA-selective inhibitors and dedicated PI3K-inhibitor registration trials. PIK3CA mutation (particularly exon 20 kinase-domain variants such as H1047R) is a negative predictive marker associated with reduced benefit from anti-EGFR monoclonal antibodies in RAS/BRAF wild-type metastatic disease. The 21% PIK3CA frequency and KRAS co-occurrence observed here are consistent with plasma ctDNA profiling of advanced CRC (PIK3CAm 19.2%, with increased co-alteration of APC and KRAS); that series is an external landscape check, not a survival comparator, because it is an advanced-disease plasma cohort rather than consecutive institutional FFPE.38 In the adjuvant setting, the phase 3 ALASCCA trial demonstrated that aspirin 160 mg daily for 3 years significantly reduced recurrence among patients with resected stage II–III colorectal cancer harboring PIK3CA hotspot mutations in exons 9 or 20 (or other PI3K-pathway alterations).22 Immune checkpoint blockade is now the standard of care for MSI-high / MMR-deficient colorectal cancer, supported by the phase 3 KEYNOTE-177 trial demonstrating superior progression-free survival with first-line pembrolizumab versus chemotherapy and by subsequent NCCN and FDA endorsements of PD-1 ± CTLA-4 inhibitors in this molecular subset.21 Although these therapies were not yet standard of care during the 2017–2018 study period, our reflexive MSI/MMR testing already identifies patients who currently qualify for immunotherapy. Thus, reflexive molecular testing in a rural laboratory already identifies patients for both approved precision regimens and emerging trial criteria.

Substantial genetic heterogeneity exists in colon cancer, and concurrent APC, KRAS, and TP53 mutation is not universal.39 That rarity is expected if co-mutation is a selected event rather than a chance collision of common drivers.4 In our cohort this mutation triad occurred in only a minority of cases yet conferred a disproportionate prognostic impact, remaining independent of advanced stage in multivariate analysis (Supplemental Table S5a). Among KRAS-mutant tumors, G12D and G13D frequently co-mutated with APC (approximately 60–65%), underscoring recurrent pairing of Wnt- and MAPK-pathway alterations. Complementary profiling of stage II–III colon cancers has shown that APC mutation is associated with poorer survival in stage III disease, an effect enhanced when MAPK-pathway mutations coexist.40 Exploratory stratification by KRAS allele showed that concurrent G12D or G13D with APC and TP53 defined a high-event subset (n = 8; 6 deaths) with inferior overall survival (log-rank p < 0.0001; Supplemental Figure S1); median OS was approximately 10.7 months versus not reached among remaining patients. These 8 cases are a subset of the 14 triad-positive tumors and 6 of the 9 triad deaths; Supplemental Figure S1 is therefore an allele-level slice of Figure 4, not an independent confirmation of the triad result. Residue-level APC or TP53 combinations within the triad were each present in only one or two patients and should be regarded as hypothesis-generating. Across major driver backgrounds (Supplemental Table S1), APC was significantly enriched among KRAS-mutant tumors (OR 3.07; p = 0.0007) and depleted among BRAF-mutant tumors (OR 0.22; p = 0.002). Allele-level analysis (Supplemental Table S2) confirmed high APC co-mutation rates for G12D (61.1%) and G13D (64.7%); pooled G12D|G13D tumors harbored APC more often than KRAS wild-type tumors (OR 3.61; p = 0.0015). Prior studies have linked co-alteration of components of the APC–KRAS–TP53 axis to adverse outcomes, although most emphasize pairwise rather than full-triad genotypes. Peng and colleagues described a prognosis-related APC/KRAS/TP53 haplotype associated with higher stage and lower tumor-specific survival in TCGA colorectal cases.8 In metastatic CRC, Datta et al. showed that co-altered RAS/BRAF and TP53 defined an extremes-of-survivorship cluster,9 and after resection of colorectal liver metastases, combined RAS and TP53 status stratified post-hepatectomy survival beyond either gene alone.10,11 Our finding that the full triad independently predicts inferior overall survival in a consecutive rural NGS cohort extends this literature from pairwise combinations and selected metastatic series to a population-based practice setting.

Biologically, APC functions as a classic tumor-suppressor “gatekeeper” whose product negatively regulates the Wnt/β-catenin pathway. Biallelic APC inactivation is an early event in the majority of conventional colorectal adenocarcinomas and initiates adenoma formation. Beyond initiation, residual APC mutation retains prognostic information, particularly when accompanied by concurrent MAPK-pathway alterations.40 The enrichment of truncating APC mutations within the mutation cluster region in our cohort is consistent with selection for complete loss of β-catenin regulation and supports the interpretation that APC disruption remains relevant after invasive carcinoma has developed.

KRAS encodes a small GTPase that transduces receptor tyrosine kinase signals through the MAPK pathway, promoting proliferation and survival. Experimental and clinical data indicate that KRAS activation cooperates with other driver lesions to accelerate progression: in genetically engineered models, KRAS activation combined with mutant p53 and TGF-β pathway suppression promotes metastasis,41 and clinical series confirm that MAPK-pathway mutations amplify the adverse prognostic effect of APC mutation.40 In our cohort, KRAS mutations occurred at frequencies comparable to multi-institutional series and, when present together with APC and TP53, defined the highest-risk molecular subgroup.

TP53 encodes the p53 transcription factor, a central guardian of genomic integrity coordinating DNA repair, cell-cycle arrest, and apoptosis. TP53 mutation is found in approximately 60% of colorectal cancers; most are missense substitutions at DNA-binding-domain hotspots that confer both loss of wild-type function and oncogenic gain-of-function properties.41 Mouse models have shown that specific missense mutants drive submucosal invasion and, when combined with Kras activation, promote metastatic dissemination.41 Reviews of targetable abnormalities in colon cancer similarly emphasize that concurrent alteration of KRAS, PI3K-pathway genes, and TP53 contributes to progression and metastasis.42 In our series TP53 was the most frequently mutated gene (61%), with classic hotspots (R175, R248, R273) predominating, and its co-occurrence with APC and KRAS marked the subset with the poorest survival.

In the classic Fearon–Vogelstein multi-step model of colorectal carcinogenesis, APC inactivation initiates adenoma formation, subsequent KRAS activation drives intermediate adenoma growth, and TP53 loss coincides with the transition to invasive carcinoma.2,3 Consistent with this model, concurrent APC/KRAS/TP53 alteration was observed in a minority of our cohort and identified the subgroup with the poorest survival. Our findings, together with contemporary genomic and functional studies,39–43 indicate that this linear sequence retains clinical relevance: tumors that complete the full sequence of APC, KRAS, and TP53 alterations exhibit a distinctly adverse prognosis. Rather than viewing these events as merely chronological steps in tumorigenesis, the present data support their use as a composite molecular risk biomarker ascertainable from routine NGS panels already deployed in community practice.

The practical contribution of this series is operational: reflexive standing-order NGS in a rural laboratory recovers a driver spectrum that matches published academic series, with a particularly adverse APC/KRAS/TP53 triad detectable in that setting. The frequencies of KRAS, BRAF, and MSI-high / MMR-deficient status closely paralleled published rates. The APC mutation rate in this series (41%, 74/180) should not be read as a geographic finding. Paired tumor–normal whole-exome analysis in the TCGA colorectal project detected somatic APC mutations in 81% of non-hypermutated tumors,5 and sequencing of the entire APC coding region in sporadic CRC has shown truncating somatic mutations in approximately 70% of cancers (about 75% when loss of heterozygosity is included).44 Those figures describe somatic tumor genotypes, not germline APC syndromes. Because AmpliSeq CHPv2 interrogates APC hotspots rather than the full coding sequence, the 41% rate observed here is the expected result of an incomplete somatic APC profile and is consistent with under-ascertainment relative to whole-gene assays. This supports the generalizability of molecularly guided treatment algorithms derived from broader populations to patients treated in regional settings.

Our experience also highlights the practical benefits of reflexive molecular testing via an institutional standing order. Automatic initiation of NGS (and other methods such as fluorescence in situ hybridization or immunohistochemistry, where relevant) at the time of diagnosis minimizes tissue wasting, shortens turnaround time, and ensures that results are available when systemic therapy decisions are made - an advantage particularly relevant in a regional practice that serves a large geographic catchment area with long patient commute times.

Limitations of our study include its retrospective nature, the small number of events in the mutation triad-positive arm (9 deaths among 14 patients), evolving composition of the NGS panel content over the study period, and incomplete stage or follow-up information for a small minority of cases. Because APC detection on this hotspot panel is incomplete relative to somatic whole-gene APC sequencing,5,44 some biologically triad-positive tumors may be classified as triad-negative; the reported hazard ratio is therefore a lower-bound estimate of the true mutation triad effect. Formal multivariable modeling with full adjustment for all covariates was constrained by sample size (39 events). Age was added as the one additional covariate a modest event count can support; treatment is collinear with stage, 23 samples were metastatic biopsies, and their tumor site was not applicable to the colonic location analyses. A larger cohort will be needed to refine the magnitude of the third-hit increment and to support penalized or resampling-based survival estimates. Median follow-up was 17 months (range, 0.7–36.0) in a cohort that is approximately half stage I–II, which is short for colorectal overall survival; the endpoint is all-cause OS, not cancer-specific or disease-free survival. Cause of death could not be abstracted in a structured way for the 39 deaths, so a cancer-specific sensitivity analysis was not performed and thus an all-cause OS in this window should not be interpreted as direct evidence that the triad drives metastasis. The strengths include consecutive ascertainment, uniform molecular testing methodology, and integration of molecular and clinical outcomes in a single regional oncology practice. Mapping of hotspot genotypes to precision-therapy eligibility is constrained by factors beyond the capacity of this series. The assessment of ultimate benefit from genotype-base treatments of eligible patients with access to approved agents or trials would require prospective follow-up in future series.

Identification of high-risk molecular profiles such as the APC/KRAS/TP53 triad may ultimately aid treatment stratification and selection of patients for intensified surveillance or novel pathway-targeted therapies, complementary to the genotype-directed eligibility outlined above. Expansion of the cohort with more recently treated patients is planned to further validate the prognostic model and to quantify real-world uptake of approved and investigational genotype-directed options.

Conclusions

Routine DNA mutation analysis by NGS is readily performed in a community/rural setting and can be integrated into standard diagnostic workflows via reflexive standing orders. The gene mutation spectrum in this rural Maine cohort is comparable to published multi-institutional series, and the APC/KRAS/TP53 triad represents an adverse all-cause OS marker in community practice. Two of the top five most frequent mutations (TP53 and APC) are not part of current NCCN-mandated therapeutic testing. Concurrent TP53/APC/KRAS co-mutations are associated with adverse overall survival, independent of tumor stage and MSI status. Nested models indicated that the third hit confers incremental risk beyond pairwise APC–KRAS or RAS–TP53 genotypes, consistent with completion of the canonical APC–KRAS–TP53 sequence rather than simple addition of three independently adverse genes. Whether that incremental third-hit risk is a stable feature of CRC biology, rather than a large effect observed in a small event set, will require investigation in a larger multi-institutional cohort.

Declarations

Conflicts of interest statement
The authors have no conflicts of interest to declare.
Funding statement
No external funding was received for this study.
Acknowledgments
The authors thank the pathology, molecular, and oncology teams at Dahl-Chase Diagnostic Services and Northern Light Cancer Institute for their contributions to specimen processing, sequencing, data abstraction, and clinical care of the patients included in this study.

Supplementary Material

Figure S1. Overall survival for the exploratory KRAS G12D or G13D plus APC plus TP53 subset (n = 8; 6 deaths) versus all others (n = 164; 33 deaths; log-rank p < 0.0001). These eight cases are 8 of 14 triad-positive tumors and 6 of 9 triad deaths in Figure 4 and are not an independent confirmation of the triad. Median survival was about 10.7 months in the allele-defined subset and was not reached in the remaining cohort.

Table S1. APC co-mutation enrichment among major driver-mutant subgroups

Driver subgroupnAPC co-mutated, n (%)ORp-value
KRAS+6136 (59.0%)3.070.0007
PIK3CA+3719 (51.4%)1.690.19
TP53+11149 (44.1%)1.390.35
FBXW7+115 (45.5%)1.210.76
SMAD4+186 (33.3%)0.690.62
BRAF+315 (16.1%)0.220.002
NRAS+74 (57.1%)1.960.45
Overall cohort18074 (41.1%)——

APC co-mutation rates among major driver subgroups in 180 tumors. Odds ratios compare APC frequency in tumors mutant versus wild-type for the listed gene. APC was enriched with KRAS (odds ratio 3.07; p = 0.0007) and depleted with BRAF (odds ratio 0.22; p = 0.002). No APC–CTNNB1 co-occurrence was observed.

Table S2. APC co-mutation rates by KRAS allele

KRAS statusnAPC co-mutated, n (%)
G13D1711 (64.7%)
G12D1811 (61.1%)
G12V157 (46.7%)
Q6133 (100%)
Other KRAS64 (66.7%)
G12C20 (0%)
Pooled G12D|G13D3522 (62.9%)
KRAS wild-type11938 (31.9%)

APC co-mutation by KRAS allele. Pooled G12D or G13D versus other KRAS alleles was not significant (odds ratio 1.45; p = 0.60). Pooled G12D or G13D versus KRAS wild-type was enriched (odds ratio 3.61; p = 0.0015). G12D and G13D dominate absolute triad counts without allele-specific enrichment relative to other KRAS mutants.

Table S3. Rare and uncommon hotspot-panel variants in the rural Maine CRC cohort (n ≤ 2 occurrences)

GeneProtein changenOncoKB oncogenicity (summary)Preclinical / literature note
KRASA146T1Oncogenic (gain-of-function)Activating in cell models (increased pathway activation); OncoKB Oncogenic
KRASG12A1Oncogenic (gain-of-function)Canonical codon-12 allele; well-established oncogenic
KRASG12S1Oncogenic (gain-of-function)Canonical codon-12 allele; well-established oncogenic
KRASQ61H / Q61L1 eachOncogenic (gain-of-function)Codon-61 alleles; OncoKB Oncogenic
KRASK117N1Oncogenic (gain-of-function)OncoKB Oncogenic
BRAFD594N1Oncogenic (class III / kinase-impaired)Class III BRAF; RAS-dependent CRAF signaling in preclinical models (Yao et al.)
BRAFD634N / V5111 eachNot established / limited annotationInsufficient functional data in this series
NRASQ61H1Oncogenic (gain-of-function)Codon-61 NRAS; established oncogenic class
NRASG13R / G15R1 eachLimited / not fully annotatedFunctional validation needed
PIK3CAH1047L / H1047Y1 eachLikely oncogenic (kinase domain)Related to H1047R hotspot; OncoKB supports oncogenicity of H1047 family
PIK3CAQ546E / Q546K / Q546L1 eachLikely oncogenic (helical domain)Helical-domain alleles near E545; curated as likely oncogenic in OncoKB when annotated
PIK3CAG451R, V346A, Y1021H, R916H1 eachNot established in this seriesRare non-hotspot; functional validation needed
SMAD4R361G1Likely oncogenic (MH2)R361 is a recurrent MH2 hotspot; R361H has LOF preclinical data
SMAD4Other MH2/truncating (e.g., R445*, E526*, D351del)1 eachLikely LOF when truncating; missense variableTruncating alleles presumed LOF; uncommon missense need validation
TP53Y220C, P278H, C176R2 eachLikely oncogenic / DNA-bindingRecurrent DNA-binding alleles; broadly supported as pathogenic
TP53Selected singleton missense/fs (e.g., C242F, G266V, M246del)1 eachVariable / not fully annotatedFunctional validation needed for private alleles
APCUncommon MCR truncations (n=1–2)1–2Likely oncogenic (truncating)Truncating MCR alleles presumed LOF; residue-level specificity limited

Variants detected in ≤2 tumors and outside the predominant recurrent alleles for each gene are listed. OncoKB oncogenicity is summarized from public annotations (accessed August 2026). Pathogenicity is suggested only when supported by published preclinical functional data; otherwise, pathogenicity is listed as not established in this series.Rare alleles observed in two or fewer tumors. Several uncommon KRAS and NRAS alleles and BRAF D594N are annotated as oncogenic in OncoKB; D594N is a class III mutant.45 Private non-hotspot missense changes lack sufficient functional data and are not treated as pathogenic from frequency alone. Pathogenicity should be validated before clinical use.

Table S4a. Driver mutation prevalence: right-sided vs left-sided colon

MutationRight mut/n (%)Left colon mut/n (%)OR95% CIp
KRAS31/72 (43.1)19/74 (25.7)2.191.09–4.410.036
APC34/72 (47.2)28/74 (37.8)1.470.76–2.840.32
TP5339/72 (54.2)48/74 (64.9)0.640.33–1.250.24
Triple (APC+KRAS+TP53)7/72 (9.7)6/74 (8.1)1.220.39–3.830.78

Right colon (cecum to transverse, n = 72) versus left colon (descending to sigmoid, including rectosigmoid, n = 74) in 180 tumors; three unspecified sites were excluded. Two-sided Fisher tests. KRAS was right-enriched (43.1% versus 25.7%; odds ratio 2.19; p = 0.036) and the association persisted in microsatellite-stable tumors (p = 0.022). APC, TP53, and the triad did not differ by side.

Table S4b. Driver mutation prevalence: right-sided vs left-sided colon plus rectum

MutationRight mut/n (%)Left+Rectum mut/n (%)OR95% CIp
KRAS31/72 (43.1)29/105 (27.6)1.981.05–3.730.037
APC34/72 (47.2)39/105 (37.1)1.510.82–2.780.21
TP5339/72 (54.2)69/105 (65.7)0.620.33–1.140.16
Triple (APC+KRAS+TP53)7/72 (9.7)8/105 (7.6)1.310.45–3.780.78

Right colon (n = 72) versus left colon plus rectum (n = 105). KRAS remained right-enriched (odds ratio 1.98; p = 0.037). APC, TP53, and the triad did not differ. Site definitions match Supplemental Table S4a. Three tumors with unspecified site were excluded.

Table S5a. Multivariable Cox proportional hazards model — concurrent APC/KRAS/TP53 triple mutation, advanced stage, and age

CovariateHR95% CIp
Triple APC/KRAS/TP53 co-mutation6.242.83–13.76< 0.001
Advanced stage (III–IV)14.44.43–47.1< 0.001
Age (per year)1.020.999–1.050.057

Cox model in 172 patients with follow-up; 14 were triad-positive. Covariates were the APC/KRAS/TP53 triad, advanced stage (III–IV versus I–II), and age in years. The triad hazard ratio was 6.24 after age was added and 6.14 without age. Age itself was of borderline significance (p = 0.057).

Table S5b. Multivariable Cox proportional hazards model — concurrent APC/KRAS/TP53 triple mutation, advanced stage, and microsatellite-high or mismatch-repair–deficient status.

CovariateHR95% CIp
Triple APC/KRAS/TP53 co-mutation6.082.77–13.34< 0.001
Advanced stage (III–IV)13.44.09–44.0< 0.001
MSI-high / MMR-deficient0.760.23–2.500.65

Cox model in 172 patients with follow-up. Covariates were the APC/KRAS/TP53 triad, advanced stage (III–IV), and microsatellite-high or mismatch-repair–deficient status. The triad hazard ratio remained 6.08 after mismatch-repair status was added. Mismatch-repair–deficient status itself was not associated with overall survival (hazard ratio 0.76; p = 0.65).

Survival-evaluable cohort (n = 172) with known vital status and documented follow-up time. Overall survival measured from date of diagnosis; patients alive at last contact were censored. MSI-high / MMR-deficient includes molecular MSI-H and/or immunohistochemical loss of MLH1, PMS2, MSH2, or MSH6. Legend. HR = hazard ratio from Cox proportional hazards regression. Adding MSI-high / MMR-deficient status to the model in Supplemental Table S6A does not materially change the triple-mutation effect (HR 6.18 → 6.11) and MSI-H itself is not significantly associated with overall survival in this multivariable model. These results support the interpretation that the adverse prognostic effect of concurrent APC/KRAS/TP53 co-mutation is independent of both advanced stage and MSI status in this rural Maine cohort. Estimates are from a Cox model on the locked survival-evaluable set (n = 172; 39 deaths). Stage remains independently associated with inferior overall survival in the three-covariate model; the wide confidence interval for stage reflects the small number of deaths among early-stage patients.

References

  1. Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73(1):17-48.
  2. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell. 1990;61(5):759-767.
  3. Smith G, Carey FA, Beattie J, et al. 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.
  4. Iyer A, Petrovic M, Sesia D, et al. Evolving patterns of co-mutations from tumor initiation to metastatic progression. Nat Genet. 2026;58:2053-2062. doi:10.1038/s41588-026-02661-4
  5. Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 2012;487(7407):330-337.
  6. Yaeger R, Chatila WK, Lipsyc MD, et al. Clinical sequencing defines the genomic landscape of metastatic colorectal cancer. Cancer Cell. 2018;33(1):125-136.e3.
  7. Schell MJ, Yang M, Teer JK, et al. A multigene mutation classification of 468 colorectal cancers reveals a prognostic role for APC. Nat Commun. 2016;7:11743.
  8. Peng X, Zhang T, Jia X, et al. Impact of a haplotype (composed of the APC, KRAS, and TP53 genes) on colorectal adenocarcinoma differentiation and patient prognosis. Cancer Genet. 2022;268-269:115-123.
  9. Datta J, Smith JJ, Chatila WK, et al. Co-altered Ras/B-raf and TP53 is associated with extremes of survivorship and distinct patterns of metastasis in patients with metastatic colorectal cancer. Clin Cancer Res. 2020;26(5):1077-1085. doi:10.1158/1078-0432.CCR-19-2390
  10. Kawaguchi Y, Kopetz S, Newhook TE, et al. Mutation status of RAS, TP53, and SMAD4 is associated with survival outcome after resection of colorectal liver metastases. Ann Surg Oncol. 2019;26(7):1948-1956. doi:10.1245/s10434-019-07302-1
  11. Chun YS, Passot G, Yamashita S, et al. Deleterious effect of RAS and evolutionary high-risk TP53 double mutation in colorectal liver metastases. Ann Surg. 2019;269(5):917-923. doi:10.1097/SLA.0000000000002450
  12. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Colon Cancer. Version 2.2026. Accessed August 8, 2026. https://www.nccn.org/professionals/physician_gls/pdf/colon.pdf
  13. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Rectal Cancer. Version current as of August 8, 2026. Accessed August 8, 2026. https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf
  14. Sepulveda AR, Hamilton SR, Allegra CJ, et al. Molecular biomarkers for the evaluation of colorectal cancer: guideline from the American Society for Clinical Pathology, College of American Pathologists, Association for Molecular Pathology, and the American Society of Clinical Oncology. J Clin Oncol. 2017;35(13):1453-1486.
  15. Van Cutsem E, Cervantes A, Adam R, et al. ESMO consensus guidelines for the management of patients with metastatic colorectal cancer. Ann Oncol. 2016;27(8):1386-1422.
  16. Andreyev HJ, Norman AR, Cunningham D, et al. Kirsten ras mutations in patients with colorectal cancer: the RASCAL II study. Br J Cancer. 2001;85(5):692-696.
  17. Karapetis CS, Khambata-Ford S, Jonker DJ, et al. K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med. 2008;359(17):1757-1765. doi:10.1056/NEJMoa0804385
  18. Kopetz S, Grothey A, Yaeger R, et al. Encorafenib, binimetinib, and cetuximab in BRAF V600E-mutated colorectal cancer. N Engl J Med. 2019;381(17):1632-1643.
  19. Elez E, Yoshino T, Shen L, et al. Encorafenib, cetuximab, and mFOLFOX6 in BRAF-mutated colorectal cancer. N Engl J Med. 2025;392(24):2425-2437. doi:10.1056/NEJMoa2501912
  20. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med. 2015;372(26):2509-2520.
  21. André T, Shiu KK, Kim TW, et al. Pembrolizumab in microsatellite-instability–high advanced colorectal cancer. N Engl J Med. 2020;383(23):2207-2218. doi:10.1056/NEJMoa2017699
  22. Martling A, Hed Myrberg I, Nilbert M, et al. Low-dose aspirin for PI3K-altered localized colorectal cancer. N Engl J Med. 2025;393(11):1051-1064. doi:10.1056/NEJMoa2504650
  23. Levit LA, Byatt L, Lyss AP, et al. Closing the rural cancer care gap: three institutional approaches. JCO Oncol Pract. 2020;16(7):422-430. doi:10.1200/OP.20.00174
  24. Skacel LE, Babcock MJ, Harb A, et al. Lung adenocarcinoma: next generation sequencing and outcome in regional community oncology setting. J Surg Res. 2021;4:804-819.
  25. Pennell NA, Arcila ME, Gandara DR, West H. Biomarker testing for patients with advanced non-small cell lung cancer: real-world issues and tough choices. Am Soc Clin Oncol Educ Book. 2019;39:531-542.
  26. McCall CM, Mosier S, Thiess M, et al. False positives in multiplex PCR-based next-generation sequencing have unique signatures. J Mol Diagn. 2014;16(5):541-549.
  27. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2025. Accessed August 8, 2026. https://www.R-project.org/
  28. Wickham H. ggplot2: Elegant Graphics for Data Analysis. 2nd ed. Springer-Verlag; 2016.
  29. UniProt Consortium. UniProt: the Universal Protein Knowledgebase in 2025. Nucleic Acids Res. 2025;53(D1):D609-D617. doi:10.1093/nar/gkae1010
  30. Guinney J, Dienstmann R, Wang X, et al. The consensus molecular subtypes of colorectal cancer. Nat Med. 2015;21(11):1350-1356.
  31. Forbes SA, Beare D, Boutselakis H, et al. COSMIC: somatic cancer genetics at high-resolution. Nucleic Acids Res. 2017;45(D1):D777-D783. doi:10.1093/nar/gkw1121
  32. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5):401-404. doi:10.1158/2159-8290.CD-12-0095
  33. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1. doi:10.1126/scisignal.2004088
  34. Chakravarty D, Gao J, Phillips SM, et al. OncoKB: a precision oncology knowledge base. JCO Precis Oncol. 2017;1(1):1-16. doi:10.1200/PO.17.00011
  35. Fakih M, Tu H, Hsu H, et al. Real-world study of characteristics and treatment outcomes among patients with KRAS p.G12C-mutated or other KRAS mutated metastatic colorectal cancer. Oncologist. 2022;27(8):663-674. doi:10.1093/oncolo/oyac077
  36. Yaeger R, Weiss J, Pelster MS, et al. Adagrasib with or without cetuximab in colorectal cancer with mutated KRAS G12C. N Engl J Med. 2023;388(1):44-54. doi:10.1056/NEJMoa2212419
  37. Fakih MG, Salvatore L, Esaki T, et al. Sotorasib plus panitumumab in refractory colorectal cancer with mutated KRAS G12C. N Engl J Med. 2023;389(23):2125-2139. doi:10.1056/NEJMoa2308795
  38. Milsap A, Tsai J, Jafry B, Kim A, Kazmi SMA, Verma N. Comprehensive analysis of PIK3CA mutations in advanced colorectal cancer using circulating tumor DNA profiling. NPJ Precis Oncol. 2025;9(1):220. doi:10.1038/s41698-025-00975-w
  39. Samowitz WS, Slattery ML, Sweeney C, Herrick J, Wolff RK, Albertsen H. APC mutations and other genetic and epigenetic changes in colon cancer. Mol Cancer Res. 2007;5(2):165-170. doi:10.1158/1541-7786.MCR-06-0398
  40. van den Broek E, Krijgsman O, Sie D, et al. Genomic profiling of stage II and III colon cancers reveals APC mutations to be associated with survival in stage III colon cancer patients. Oncotarget. 2016;7(45):73876-73887. doi:10.18632/oncotarget.12510
  41. Nakayama M, Oshima M. Mutant p53 in colon cancer. J Mol Cell Biol. 2019;11(4):267-276. doi:10.1093/jmcb/mjy075
  42. Baviskar T, Momin M, Liu J, Guo B, Bhatt L. Target genetic abnormalities for the treatment of colon cancer and its progression to metastasis. Curr Drug Targets. 2021;22(7):722-733. doi:10.2174/1389450121666201119141015
  43. Su Z, El Hage M, Linnebacher M. Mutation patterns in colorectal cancer and their relationship with prognosis. Heliyon. 2024;10(17):e36550. doi:10.1016/j.heliyon.2024.e36550
  44. Christie M, Jorissen RN, Mouradov D, et al. Different APC genotypes in proximal and distal sporadic colorectal cancers suggest distinct WNT/β-catenin signaling thresholds for tumourigenesis. Oncogene. 2013;32(38):4675-4682. doi:10.1038/onc.2012.486
  45. Yao Z, Yaeger R, Rodrik-Outmezguine VS, et al. Tumours with class 3 BRAF mutants are sensitive to the inhibition of activated RAS. Nature. 2017;548(7666):234-238. doi:10.1038/nature23291

Article information

Authors

How to cite

Babcock, M., Skacel, M., et al., 2026. Extended Mutational Profiles and Clinical Outcome of Colorectal Carcinoma Patients from Rural Maine. Medical Research Archives, [online] 14(10). https://doi.org/10.18103/mra.2026.0600

Record

Journal
Medical Research Archives
Section
Research Articles
Published
31 October 2026
ISSN
2375-1924
Downloads
127

Copyright and licence

Published by the European Society of Medicine under a Creative Commons Attribution 4.0 International licence (CC BY 4.0), which permits unrestricted use, distribution and reproduction in any medium, provided the original authors and source are credited.