PDF
Citation
Print
INTRODUCTION
Colorectal cancer (CRC) ranks as the third most prevalent cancer and the second most lethal malignancy, with an estimated 1.9 million new cases and 0.9 million deaths globally in 2020.1 In Indonesia, CRC presents an incidence rate of 8.6% and a mortality rate of 7.9%, constituting a substantial national health burden.2 Approximately 70% ofCRCs are identified as sporadic, whereas 20–30% demonstrate an inheritable component, such as Lynch Syndrome and familial adenomatous polyposis (FAP).3 Colorectal cancer demonstrates a high degree of intratumoural and intertumoural heterogeneity, with molecularly distinct subgroups demonstrating varying degrees of treatment responsiveness and prognosis.4 Understanding the precise mechanisms of disease progression and metastasis may significantly improve the outcome of colorectal cancer management.
The dynamic nature and heterogeneity of CRC are defined by various interconnecting molecular etiopathogenesis displaying distinct behaviors both inter- and intratumorally.5,6 Recent biomolecular studies indicate that genetic and epigenetic analyses can assess tumor characteristics, thereby enabling predictions regarding heredity, progression, recurrence, therapeutic response, and survival rates.7 Precision medicine tailored to an individual's cancer genetic profile starts to make strides since the clinical staging system is unable to estimate these variables independently. However, the clinical translation of oncogenomic profiling in the Indonesian CRC management faces major barriers due to limited resource and intricate system of national health insurance. Variations in CRC mutations and genes are observed across different age groups, genders, and geographic locations.8 Therefore, studying specific populations is crucial for advancing precision medicine, as individual cancer genome profiling may not be universally applied.
Alteration of expression in tumor suppressor P53 (TP53) gene plays a pivotal role in a multitude of malignancies including CRC.9 A whole exome sequencing study revealed a high prevalence of TP3 loss in early-onset CRC patients, suggesting a significant contribution in tumorigenesis.10 The Kirsten rat sarcoma (KRAS) gene accounts for over 40% of CRC mutations, while the B-Raf proto-oncogene serine/threonine kinase (BRAF) gene is responsible for 10%. These mutations induce dysregulation of the mitogen-associated protein kinase (MAPK) pathway, resulting in cell proliferation, differentiation, angiogenesis, and resistance to apoptosis.11 Out of all the members of the collagen family, the one most prevalent in extensible connective tissues is collagen type IIIA1 (COL-3A1). Upregulation of COL-3A1 gene was identified in colorectal cancer tissues as compared to normal tissues in an earlier investigation.12 The transcription level of COL-3A1 was also found to be elevated from adenoma to carcinoma, suggesting its role in carcinogenesis.13 There is mounting evidence that CRC has been associated with abnormal cytokine signaling. A number of important regulatory proteins, including members of the suppressors of cytokine signaling (SOCSs) family, are required to tightly regulate cytokine responses. Evidence suggests that SOCS2 expression levels are lower in human colorectal cancer compared to normal tissues.14 The primary objective of our study was to investigate the correlation of CRC clinical staging with the expression levels of TP53, KRAS, BRAF, COL-3A1, and SOCS-2. Additionally, we aimed to evaluate the association between the expression of these five molecules and the incidence of distant metastases, along with the intercorrelation among molecules.
SUBJECTS AND METHODS
1. Patients and Tissue Samples
We included patients who had a colonoscopy and were initially diagnosed with colorectal cancer at Dr. Soetomo General Academic Hospital Surabaya from 2021 to 2024. The clinical part of the study was carried out with the approval of the Ethics Committee of the Dr. Soetomo General Hospital (No. 0179/KEPK/IV.2021). Written in-formed consent was obtained from all patients and healthy volunteers for sample and data collection. Patients were included based on the following eligibility criteria: age at least 18 years old; and diagnosis of CRC confirmed by histopathological results. Prior to recruitment, all subjects provided their written informed consent. Subjects underwent a comprehensive evaluation which comprised MRI of the abdomen and other imaging modalities as clinical suspicion of metastases suggested. Patients with the following conditions were excluded: incomplete clinical staging and metastatic status information; a prior history of other malignancies; had a known hereditary CRC syndromes; receiving of chemotherapy and/or radiotherapy; tissue specimens that did not meet the qualifications for immunohistochemical (IHC) analysis. Demographic data such as age, sex, ethnicity, smoking status, and first-degree familial history of colorectal cancer were documented. The American Joint Committee on Cancer, 8th edition (AJCC) system was used to define the clinical stage of CRC.15 The paraffin-embedded tissue blocks, which included both tumor cells and adjacent normal epithelium, were sectioned for hematoxylin and eosin (H&E) staining and selected for further IHC analysis.
2. Immunohistochemical Analysis
Consecutive 4-μm sections were meticulously sliced from the formalin-fixed paraffin-embedded (FFPE) tissue blocks, affixed to positively charged slides, and deparaffinized. These sections were then rehydrated in phosphate-buffered saline (PBS) at pH 7.4. For antigen retrieval, the slides were immersed in a citric acid solution (pH 6.0) and heated at maximum power for 15 minutes. Subsequently, endogenous peroxidase activity was inhibited using 3% hydrogen peroxide in methanol. The tissue sections were incubated overnight at 4°C with various primary monoclonal antibodies: anti-TP53 (dilution 1:100, ACRE 298 AK; Abcam's RabMAb® technology, Cambridge, MA, USA), anti-BRAF (dilution 1:50, A 0038; ABclonal, Woburn, MA, USA), anti-KRAS (dilution 1:100, A19779; ABclonal), anti- COL3A1 (dilution 1:100, M00788; ProteinTech, San Diego, CA, USA), and anti-SOCS2 (dilution 1:100, A9190; ABclonal). Following incubation, the HiDef Detection™ HRP Amplified Antibody Polymer Detection System was employed for 10 minutes at each step. The antigen-antibody complexes were visualized using the chromogen substrate liquid diaminobenzidine, and the sections were then counterstained with hematoxylin to enhance structural contrast. The expression level of each molecule was determined by the proportion of positive cells, expressed as a percentage ranging from 0% to 100%.
3. Statistical Analysis
Statistical analyses were carried out using IBM SPSS Statistics version 24.0. The mean ± standard deviation, median (interquartile range), and count (percentage) were used for presenting descriptive data. The independent T-test or Mann–Whitney test was used for numerical data in the univariate analysis, while the Chi-square test or Fischer exact test was used for categorical data. Pearson and Spearman correlations were applied to perform bivariate analysis. The correlation coefficient coefficients were categorized as follows: 0.00–0.10, negligible correlation; 0.10–0.39, weak correlation; 0.40–0.69, moderate correlation; 0.70–0.89, strong correlation; and 0.90–1.00, very strong correlation.16
RESULTS
1. Clinical Characteristics of Subjects
Inadequate clinical staging data led to the exclusion of 21 out of 84 CRC patients with confirmed biopsy results; thereby, 63 CRC patients were included in this study. The average age of the study subjects was 56.3 years, and 46.0% of them were male. The Javanese constituted the predominant ethnicity at 74.6%, followed by the Madurese and Mongoloid groups. 15.9% of the cases have a family history of CRC in first-degree relatives. According to body mass index classifications, 39.7% of patients were identified as underweight, 34.9% as having normal weight, 14.3% as overweight, and 11.1% as obese. The primary tumors were located on the left side in 81% of subjects, whereas 19% were located on the right side. Our study population primarily consisted of advanced colorectal cancer patients, with 47.6% exhibiting regional to mesenteric lymph node metastasis and 41.3% presenting distant metastasis. The liver represented the most common site for distant metastasis at 31.7%, followed by the lung at 14.3% and bone at 3.2%. Other metastases (4.8%) included two patients with axillary lymph node metastases and one patient with ovarian metastasis. Upon histological examination, it was determined that 77.8% of the biopsy samples were classified as well-differentiated adenocarcinoma (Table 1).
2. Alteration in the Expression Levels of TP53 Across Different Clinical Stages
Results from TP53, BRAF, KRAS, COL-3A1, and SOCS-2 correlation analyses with cancer staging according to AJCC 8th edition were provided in Table 2. Our analysis indicated that the expression level of the tumor suppressor P53 gene had an inverse correlation with the progression of clinical staging (ρ –0.260, p<0.05). It is important to note that the expression of TP53 was found to be absent in the majority of subjects with distant metastases (stage IV CRC) (Fig. 1). The remaining four genes—BRAF, KRAS, COL-3A1, and SOCS-2—did not show any statistically significant correlation to clinical stage (p>0.05).
Further subanalysis was conducted based on tumor location. In patients with left-sided CRC (n=51), a negative correlation was observed between TP53 gene expression and clinical stage (ρ –0.309, p<0.05). Contrarily, there was no such correlate in TP53 expression among the 12 patients with cancer of the right side of the colon (p>0.05). The expression levels of BRAF, KRAS, COL-3A1, and SOCS-2 consistently failed to demonstrate a correlation with the advancement of cancer staging either in right-sided or left-sided CRC patients (p>0.05). Additionally, no significant differences were found when comparing the expression of TP53, BRAF, KRAS, COL-3A1, and SOCS-2 between left- and right-sided CRC in the bivariate analysis (p>0.05).
3. Expression Levels of TP53 Differs in CRC Patients With and Without Distant Metastases
The present study additionally highlighted the comparison of expression levels of TP53, BRAF, KRAS, COL-3A1, and SOCS-2 between CRC patients with and without distant metastases. Among the five histologic markers, only the tumor suppressor P53 gene demonstrated a statistically significant difference in relation to metastatic status (Table 3). Subjects with distant metastases have a lower-level expression of TP53 compared to those without (0.00 [1.00] vs. 1.00 [23.00], p<0.05) (Fig. 2). This result indirectly indicates a downregulation phenomenon of the TP53 gene in metastatic colorectal cancer patients.
A comparative analysis of the expression levels of TP53, BRAF, KRAS, COL-3A1, and SOCS-2 across various distant metastatic organ sites was also conducted (Table 3). Expression levels of BRAF, KRAS, COL-3A1, and SOCS-2 were consistently not associated with specific organ metastases, including those in the liver, lung, and bone (p>0.05). Although there were notable differences in overall distant metastases, the expression level of the TP53 gene did not show a comparable association when specific target organs were separated (p>0.05). Further subanalysis comparing the expression of these molecules based on tumor sidedness was performed. Interestingly, we found that metastatic left-sided CRC patients exhibit a notably reduced expression level of TP53 in both lung (0.00 [0.00] vs. 1.00 [5.25], p<0.05) and overall (0.00 [1.00] vs. 1.00 [21.50], p<0.05) metastases. In contrast, no significant differences in TP53 expression were detected among patients with metastatic right-sided CRC (p>0.05).As for BRAF, KRAS, COL-3A1, and SOCS-2, there was also no difference in either right- or left-sided CRC according to metastatic status (p>0.05).
4. TP53 Expression Correlates with BRAF, KRAS, COL-3A1, and SOCS-2 Expression
Despite failing to demonstrate any significant association with clinical parameters, BRAF, KRAS, COL-3A1, and SOCS-2 were found to be directly correlated with TP53 expression (Fig. 3). The expression of BRAF and SOCS-2 was moderately correlated with the level of TP53 (ρ 0.617, p<0.01 and ρ 0.571, p<0.01, respectively). However, there was only a weak correlation between the expression level of TP53 with KRAS (ρ 0.272, p<0.05) and COL-3A1 (ρ 0.348, p<0.01). No significant correlation was found among these molecules, excluding TP53, when analyzed in relation to one another (p>0.05).
DISCUSSION
The study population predominantly consisted of subjects aged over 50 years with advanced clinical stage disease. The delay in diagnosing and managing colorectal cancer (CRC) might result from multiple factors. Low levels of education regarding colorectal cancer risk factors and the importance of screening may account for a disproportionately large percentage of patients with advanced stage disease. Patients with unspecific symptoms are less likely to visit physicians before developing a serious illness, which is partly due to the complicated system of national health insurance. Furthermore, as our population is comprised up of patients from various ethnic backgrounds, there may be dietary and lifestyle factors that contribute to the high prevalence of advanced stage CRC in Indonesia. An unexpectedly high proportion (77.8%) of well-differentiated adenocarcinoma was observed in this study, deviating notably from patterns typically seen in colorectal cancer cohorts. This finding may be attributed to factors such as variability in pathological grading between observers, distinct genetic or environmental characteristics specific to the Indonesian population, or possible selection bias inherent in the cross-sectional study design.
Tumor suppressor P53 gene is referred to as the "guardian of the genome." The protein product identifies various types of DNA injury and either pauses the cell cycle to facilitate DNA repair or, in instances of irreparable damage, triggers apoptosis.17 Our results demonstrate that TP53 expression had a negative correlation to clinical stage progression. This finding corroborates the hypothesis that the loss of TP53 occurs at a later stage in the adenoma-carcinoma transformation sequence. The rate of inactivating mutations in TP53 was relatively low in precancerous lesions, while a higher prevalence was observed in cases of invasive cancer.18 Since the expression of TP53 is relatively high in patients with early-stage cancer, our results indicate that most CRC patients in Indonesia do not originate from inflammatory bowel disease. Distinct genetic features have been identified in colitis associated with CRC. These types of cells typically carry TP53, IDH1, and MYC mutations, whereas mutant APC and KRAS are less prevalent.19-22 In colitis-associated CRC, the loss of heterozygosity for the tumor suppressor P53 gene occurs as an early event in the pathogenesis of cancer.23
In this study, the absence of TP53 expression has been identified in almost all CRC patients with distant metastases. An earlier investigation demonstrated that TP53 mutations serve as a driver of metastasis signaling in advanced colorectal cancer patients. Mutational profiles from primary and metastatic sites in patients with advanced CRC were compared in this investigation. 145 of the 171 patients analyzed demonstrated genetic alterations from primary and metastatic locations. Among 790 unique mutations, TP53 is the most prevalent, followed by APC, KRAS, PIK3CA, ATM, PTEN, NOTCH1, BRCA2, BRAF, KMT2D, LRP1B, and CDKN2A.24 The majority of our study subjects had locally advanced or distantly metastasized CRC, with the primary tumor located on the left side. We also found that the level of TP53 expression is significantly lower in patients with distant metastases compared to those without them, particularly in the left-sided CRC subpopulation. Previous study reported that the TP53 mutation is frequently observed in tumors that are growing in the rectum or distal colon, with nearly 50% of invasive colon cancers harboring this mutation.25 We additionally observed that lung metastases were associated with TP53 downregulation in our subanalysis of left-sided CRC. This finding aligns with previous investigations showing that concurrent somatic alterations of TP53 and KRAS/BRAF are significantly associated with the presence of extrahepatic metastatic sites, including the lung.26 The precise molecular mechanism responsible for the downregulation of the TP53 gene in this study is beyond the scope of our investigation. Despite frequently being inactivated by mutation or deletion, TP53 levels may be altered by post-transcriptional and post-translational events.27 In human CRC, TP53 gene copy numbers are correlated with mRNA expression, whereas TP53 protein levels are not related to these numbers.28 At the post-transcriptional level, TP53 silencing can be performed by various mechanisms, including alternative splicing and microRNA regulation. Methylation, phosphorylation, and ubiquitination are examples of post-translational modifications of the TP53 gene.29
The classification of colorectal cancer is crucial for predicting patient prognosis and guiding treatment strategies. The CRC subtyping consortium unified all previous molecular classification methods in 2015 into a single system called consensus molecular subtype (CMS).30 This study revealed a predominance of left-sided colon and rectal primary tumor sites in Indonesian CRC patients. Furthermore, the inverse correlation between TP53 expression and clinical staging progression may also indicate a late loss of tumor suppressor P53 gene activity. These two features were in accordance with the characteristics of CMS2, which is also referred to as the canonical subtype of CRC. CMS2 represents a molecular subtype of colorectal cancer characterized by mutations in the TP53 and EGFR genes. Chromosomal instability, somatic copy number changes, and low rates of hypermutation are the hallmarks of CMS2. This subtype is closely associated to the adenoma- carcinoma sequence, typically characterized by an early loss of APC, followed by a KRAS mutation and a subsequent late loss of the TP53 gene.31 In CMS2, the Wnt signaling pathway is constantly switched on, leading to the accumulation of β-catenin, which in subsequently activates key pro-oncogenic cell-cycle regulatory genes such as c-Myc and cyclin D1.32 Among all 4 molecular subtypes, CMS2 demonstrates the most favorable prognosis, having a 77% five-year survival rate and an overall survival of close to 3 years in patients with metastatic disease.33 Understanding the most prevalent molecular subtype within a specific population has significant clinical implications, particularly in predicting the cancer's response to cytotoxic drugs and molecular targeted agents. This study indicates that Indonesian CRC patients demonstrate features typical of CMS2, suggesting they may respond positively to fluoropyrimidine-based cytotoxic regimens, including FOLFOX or FOLFIRI, as well as anti-EGFR therapies such as Cetuximab and Panitumumab.34-36
Colorectal cancer primarily develops through chromosomal instability and hypermethylation pathways,37 with KRAS and BRAF V600E mutations being key drivers in each, respectively.38 However, our study found no association between KRAS or BRAF mutations and staging progression or distant metastases, possibly because most tumors were left-sided, whereas these mutations are more prevalent in right-sided tumors.39 Furthermore, COL-3A1 and SOCS-2 were not linked to metastatic status or stage advancement; despite prior evidence suggesting COL-3A1's role in liver metastasis, our findings imply its somatic changes occur late.40 This also aligns with earlier work indicating SOCS2 is not a progressive prognostic marker for CRC, though it might serve as an early diagnostic marker due to its early downregulation in adenoma.14
The study highlighted several important areas of limitation. Our reliance solely on immunohistochemistry for TP53 expression means we could not differentiate between functional wild-type protein and stable, but functionally inactive, mutant forms. This limitation suggests that while protein accumulation was observed, the precise functional status of TP53 in our cohort could not be fully ascertained, potentially impacting the interpretation of its association with clinical outcomes. While immunohistochemistry cannot differentiate between wild-type and mutant TP53, recognizing the high frequency of TP53 mutations in colorectal cancer is crucial. Although comprehensive Indonesian data is still emerging,41 studies in other Asian populations reveal TP53 mutation prevalence ranging from approximately 26% in Southeast Asian cohorts42 to 37.58% in Taiwanese patients.43 The relatively small sample size in this study represents a limitation, potentially impacting the statistical power to detect subtle associations or generalize findings to a broader population. The estimated effect size (Cliff’s delta ≈ 0.5) suggests a moderate-to-large difference, yielding an approximate statistical power of 80% at α=0.05. While this supports the robustness of the observed association, further validation in larger, independent cohorts is warranted to confirm these findings and refine their clinical relevance. Another limitation of our study is the exclusive use of univariate analysis, which precludes adjustment for potential confounding factors such as age, BMI, tumor location, and histologic grade. While our findings indicate certain associations, future studies employing multivariate models would provide a more nuanced understanding by accounting for these covariates. While certain molecular characteristics observed in our study might suggest features consistent with the CMS2 subtype, it is imperative to acknowledge that a definitive Consensus Molecular Subtype classification necessitates comprehensive transcriptomic profiling. Our current analysis, based on clinical and immunohistochemical surrogates, does not provide the required gene expression data for formal CMS categorization, thus any such inference remains speculative without further molecular validation. While CMS2 classification requires transcriptomic profiling,44,45 and known associations link microsatellite instability and BRAF status to CMS1,46,47 and KRAS mutations to CMS3,46,48 our findings might suggest a CMS2-like phenotype. This remains a hypothesis for future research, necessitating gene expression analysis to confirm CMS subtypes and characterize Wnt and MYC signaling pathways.30,46
The classic adenoma-carcinoma sequence theory posits that the presence of an adenoma is a prerequisite for the development of CRC. Carcinogenesis requires a progressive accumulation of gene alterations promoting tumor growth, which eventually results in invasive malignancy.6 Our results revealed that, although no significant association with clinical parameters was found, BRAF, KRAS, COL-3A1, and SOCS-2 demonstrated a direct correlation with TP53 expression. This finding endorses the role of TP53 as a crucial gatekeeper gene by orchestrating numerous carcinogenic responses at the molecular level throughout carcinogenesis. Consistent with our result, a prior descriptive study reported TP53 mutations in all Indonesian CRC patients.41 The downregulation of TP53 observed in this study may be attributed to the dominant prevalence of TP53 mutations, despite the fact that a straightforward comparison cannot be made, as mutations do not automatically reflect expression levels. The expression of the tumor suppressor P53 was found to be downregulated in more advanced clinical stages and in distant metastases, indicating its pivotal role in the aggressive behavior of colorectal cancer. TP53 is integral in coordinating various molecular responses promoting carcinogenesis from the initial stages, as its expression is directly correlated with BRAF, KRAS, COL-3A1, and SOCS-2. TP53 downregulation, often via mutation, is a critical therapeutic implication in colorectal cancer, contributing to chemoresistance against agents like 5-FU and oxaliplatin, and p53 overexpression can predict resistance to anti-EGFR therapies.49-53 Paradoxically, TP53 loss or mutation can also unveil novel therapeutic vulnerabilities, suggesting potential new avenues for targeted intervention.28,54,55
XML Download