Relationship of TRAP1, HMGB1, and p62 expression with clinicopathologic features in colon cancer tissues

Abstract

BACKGROUND

Colorectal cancer yet remains one of the chief reasons that bring about cancer-connected sickness and decease in every region of the world. At present, we do not possess nearly sufficient reliable biological markers for the classification of patient risk and the completion of prognostic estimation work. An increasing number of investigations have begun to indicate that the advancement of tumors does not merely originate from alterations at the genetic level. This phenomenon is also pushed to move forward by the coordinated changes of cellular stress responses, mitochondrial function, and the tumor microenvironment. TRAP1 is one mitochondrion’s molecular chaperone which holds a function in the adjusting of metabolism and the adaptation to stress. HMGB1 plays the role of a key intermediary in the process of inflammation and the interactions between tumor and host. p62 is one scaffold protein which is related to autophagy. All three of the said molecules have connections to cancer biology in early studies. Having said this point, most research works have only examined these molecules each by themselves. How much each of them contributes when it is considered under a unified clinicopathological circumstance is still not well comprehended. The systematic observation of their expression modes and clinical meanings in colon cancer therefore could assist us in constructing a more comprehensive cognition of their functions. This thing also can possibly let us see how much use they can have as prediction biological marks.

AIM

To measure the expression levels of TRAP1, HMGB1, and p62 in colon cancer tissue samples. To compare the expression patterns of these three markers between tumor tissues and matched adjacent normal mucosal tissues. To analyze their associations with patients’ clinicopathologic characteristics and follow-up outcomes.

METHODS

We have carried out a retrospective enrollment of 120 patients in total. All of these people had first-stage colon cancer, underwent surgical cutting removal, and obtained pathological verification at one single hospital within the period from October 2022 to October 2025. We have utilized the method of immunohistochemistry to examine the expression situations of TRAP1, HMGB1, and p62. In these samples, there are included 120 tumor specimens and their matching neighboring normal mucosa tissues. Another point which has worth to mention is that we have selected 40 paired fresh specimens to carry out reverse transcription-polymerase chain reaction analysis. We have gotten together a number of clinical changeable factors. These items include age, sex, tumor position, tumor dimension, differentiation degree, invasion depth, lymph node metastasis, TNM stage, lymphatic vessel invasion, and perineural invasion. The follow-up work was continued until the month of March in the year 2026. We regard overall survival as the endpoint of our study. We have done all statistical analyses by using paired t test, χ² test or Fisher’s exact test, Kaplan-Meier analysis, and Cox regression model building.

RESULTS

The high expression rate of TRAP1, HMGB1, and p62 in colon cancer organization were 52.5%, 48.3%, and 45.0%, respectively, all of which were higher than those in matched near normal mucous membrane (20.0%, 15.8%, and 17.5%, respectively; all of P values are smaller than 0.001). Reverse transcription-polymerase chain reaction detection displayed that the relative mRNA expression amounts of TRAP1, HMGB1, and p62 are all higher in tumor tissues than in adjacent tissues (all P < 0.001). The expression level of TRAP1 that is high had connection with tumor diameter which is bigger than or equal to 5 cm, bad differentiation, stage of T3-4, metastasis of lymph node, stage of TNM III-IV, invasion of lymphovasculature, and invasion around nerve (all P < 0.05). The expression level of HMGB1 that is relatively high was connected with bad differentiation, lymph node metastasis, and TNM stage III-IV (all P < 0.05). The high expression level of p62 was also related to the bad differentiation degree, lymph gland metastasis shift, and TNM stage III-IV, all these have P < 0.05. Multivariable logistic regression analysis has indicated that lymphovascular invasion, high expression level of TRAP1, and high expression level of HMGB1 are independent correlation factors of lymph node metastasis. The follow-up time of the median is 26 months, in this period 24 death events happened. Kaplan-Meier analysis has demonstrated that patients who hold high TRAP1, HMGB1, or p62 expression possess lower overall survival than the corresponding groups with low expression (all P < 0.05). Exploratory multivariable Cox regression analysis put forward that lymph node metastasis, lymphovascular invasion, and high TRAP1 expression had connections with bad total survival.

CONCLUSION

In colon cancer tissues, TRAP1, HMGB1 and p62 are significantly overexpressed and closely correlated with adverse clinicopathological features. Of the three biomarkers, TRAP exhibits the most prominent correlation with lymph node metastasis and overall survival. Nevertheless, the relatively wide 95% confidence interval for its hazard ratio weakens the robustness of its prognostic efficacy; accordingly, the present results are insufficient to support its immediate clinical implementation for risk stratification. Despite such limitations, TRAP still has promising potential as a prognostic indicator, which requires further validation in larger cohorts with prolonged follow-up periods.

Key Words: Colon cancer; TRAP1; HMGB1; P62; Immunohistochemistry; Reverse transcription-polymerase chain reaction; Prognosis

Core Tip: This study investigated the expression of TRAP1, HMGB1, and p62 in colon cancer tissues, all of which were significantly upregulated compared with adjacent normal mucosa at both protein and mRNA levels. Their high expressions were closely associated with adverse clinicopathological features including poor differentiation, lymph node metastasis, and advanced TNM stage. TRAP1 was identified as an independent predictor for lymph node metastasis and poor overall survival, showing superior prognostic value. HMGB1 also independently predicted lymph node metastasis, while p62 was correlated with tumor progression but lacked independent prognostic significance. These findings suggest that TRAP1, HMGB1, and p62 serve as promising biomarkers for colon cancer, with TRAP1 being particularly valuable for prognostic assessment.

INTRODUCTION

Colon cancer is one important component of colorectal cancer. We have obtained actual advancement in screening methods, surgical skills and cross-discipline treatment schemes. But the whole load of this illness and its connected death rate has not been lowered in the basic way. The newest whole-world cancer data let this condition be seen clearly. Rectal and colon cancer still is one of the most frequently seen malignant tumors in the whole world. It occupies a big proportion of all death cases that are related to cancer. Please we should look at the medium and long term tendency. The populations of many countries are getting old continuously, more persons are taking in western-style life ways, and metabolism-related risk factors are continuously increasing in all crowds. We hold the expectation that these factors will make both the occurrence and death rate of this disease become higher in the coming two decades[1-3].

In the daily work of clinic, this long-time unmet heavy burden truly points out the boundaries of traditional risk classification methods. These methods have dependence only on TNM stage, histologic grade, and conventional pathological indicators. Even when patients possess similar disease stages and obtain approximately the same treatment schemes, their outcomes have very big differences. These differences can be manifested in the recurrence rates, the metastasis risk, and the overall survival (OS) time lines. This large difference in results means there are many hidden molecular and biological diversities below the surface of what seem to be same tissue structure characteristics. In the past several years, researches regarding three core domains of colorectal cancer have obtained much deeper progresses. These fields are molecule level classification, signal path transformation, and small environment control. An increasing number of research results indicate that the growth of tumor does not commence from a single isolated affair. It is formed through the combination of many factors that work together. These elements contain metabolic adaptation, inflammation amplification, cell death escape, losing protein homeostasis, and interactions between tumor bodies and their around micro environment[4,5]. When we look at the entire situation, tissue biological markers which can reflect stress adaptation, inflammatory signaling and autophagic situation at one time should have better effects. They can display the true invasive capability of tumors in a more effective way than markers which only connect with cell proliferation or differentiation. This superiority is particularly obvious for the cancer of colon[6]. Large intestine cancer is one kind of entity tumor, which is defined by having obvious metabolism reprogramming and strong immunity inflammation background. We need to point out that the confirmation of particular protein marking objects still possesses quite actual, realistic clinical usage value. These markers are required to associate with tumor invasion, metastasis, and patient prognosis, and can be detected in common paraffin-embedded tissues that we use in daily laboratories.

Tumor necrosis factor receptor-related protein 1 (TRAP1) is an important mitochondria molecule chaperone which belongs to the HSP90 family. This thing assists in maintaining the balance of mitochondrial protein quantities. It also carries out regulation on the way that cells make response to oxidative stress, and thus it forms the way that tumor cells carry out adjustment when they grow in bad environments. Early stage researches discovered that TRAP1 inside colorectal cancer cells has connections with both multidrug resistance and the ability of these cells to evade apoptosis. This thus proves it is not merely a common chaperone protein, but a core function component of the network which maintains the survival of tumor cells. After that time, Lettini et al[7] demonstrated further particulars concerning its function. TRAP1 is able to maintain the stemness and clonogenic ability of colorectal cancer cells in complete condition through the regulation of the Wnt/beta-catenin pathway. This relation links the protein directly with tumor producing, long-term expanding, and the survival of leftover cancer cells following treatment. These results let people clearly see that TRAP1’s biological function in colon cancer is far more than that it only has higher expression level. This in fact indicates a complete adaptive reaction which tumor cells start. This response is concerned with metabolic stress, protein-folding stress, and the selection pressure which is put on cells by the treatments.

From the clinicopathology angle, persons have step by step begun to put more attention on the research worth of TRAP1. Pak et al[8] completed a large-scale tissue research on colorectal cancer. They have reported that high TRAP1 expression is connected with deeper local invasion and worse disease-specific survival, therefore it can work as an auxiliary index to display how aggressive a tumor is. Having said this, there still exist two limitations when we attempt to employ TRAP1 to deal with routine colon cancer cases in clinical environments at present[9]. First, the great majority of earlier research works have only examined the connection which exists between one individual marking factor and one single final result. They have not completed sufficient parallel contrast works with other core molecules in relevant access routes. Second, we up to now still do not possess sufficient pathology-grade evidence to confirm whether TRAP1 alterations occur in synchronization with inflammation-associated pathways and autophagy-associated proteins. When we inspect TRAP1 in a more integrated frame that includes metabolism, inflammation and autophagy, this may assist us to make clear its function in colon cancer development more distinctly.

High mobility group box 1 (HMGB1) is one other key molecule that has two different biological functions. Within the cell nucleus, it mainly carries out work to maintain the stability of DNA structure, regulate the transcription of genes, and repair the damaged genetic material. When cells locate in stress situation, pass through necrosis, or on active secrete materials, HMGB1 can move toward the cytoplasm, or even be obtained release out of the cell. After it goes out from the cell, it serves as a damage-associated molecular pattern, and participates in amplifying inflammatory reactions, promoting cell migration, and remolding the immune microenvironment[10]. Current reviews which inspect the function of HMGB1 in colorectal cancer display that this molecule is able to activate many tumor-boosting signal conduction paths via receptors such as receptor for advanced glycation end-products and Toll-like receptors. This procedure finally brings influence to tumor cell proliferation, invasion, new blood vessel growth, and resistance to therapy. To put it in a simple way, the clinical value of HMGB1 therefore does not only originate from its high expression levels. It also is obtained from the dynamic processes which it represents: Breaking of nuclear steady balance, moving into the cytoplasm, and outside cell signal increasing.

In the colorectal cancer, the higher HMGB1 expression is connected with the heavier tumor burden, the lymph node metastasis, and the worse patient result. Ueda et al[11] have discovered that high HMGB1 concentrations are in accordance with bigger tumor dimensions, more often occurring lymphatic intrusion, higher proportions of node involvement, and lower total survival percentages. These data support that HMGB1 plays the role of an independent prognostic prediction marker. Another aspect which needs to be pointed out is that current research also emphasizes the large importance of the subcellular localization of HMGB1. Its moving from the cell center to the cytoplasm occurs when tissue develops from adenoma to carcinoma, hence to more aggressive kinds of this illness. This change is also connected with positive lymph node condition and an immune-cold type of character. Notably, our study only detected total HMGB1 expression without distinguishing nuclear and cytoplasmic localization, so we cannot verify the above localization-related conclusion with our own data. Through the first look, these data points may only look like independent observations, but when you put them together, it can be seen that testing HMGB1 in colorectal cancer touches a key cross point between intracellular stress feeling and extracellular inflammatory signal transmission[12]. This can assist in making clear why tumor progression and microenvironment reformation are connected with such tight closeness.

When put in comparison with TRAP1 and HMGB1, p62/SQSTM1 can provide a more direct depiction about the situation of protein degradation, autophagic flux, and stress signaling. The protein p62 has the function of acting as a selective autophagy receptor and also a multifunctional scaffold. It not only assists in delivering ubiquitinated substrates to autophagosomes, but also gathers together pathways such as nuclear factor erythroid 2-related factor 2, nuclear factor-κB, and mammalian target of rapamycin which control oxidative stress, inflammation, and cell survival. This twofold effect has been pointed out by previous scholars’ study[13,14]. In the early period of tumor generation, autophagy assists with maintaining genomes stable. It achieves this work through the clearing away of damaged organelles and proteins which otherwise would bring about problems. For tumors that are already completely formed, however, abnormal autophagy activation and p62 accumulation normally indicate that the tumor has carried out adaptation to low oxygen, shortage of nutrition, and pressure brought by therapy. Therefore, for the colorectal cancer disease, p62 is able to play two distinct roles. It may function as a protective factor, hence it may also act as an indication that the tumor has gotten more invasive and has resistance to treatment.

Previous clinical and pathologic researches indicate that normal colorectal epithelial tissue possesses extremely low p62 expression level. This expression has continuous rise when tissue toward adenomas develops, then further toward adenocarcinomas goes. The authors Nakayama et al[15]. Therefore, our research discovered that the content of p62 in adenocarcinoma is very much higher when compared with that in normal tissue. These rising levels have connection with synchronous liver metastasis and poorer results for patients. This indicates that p62 is one useful histological marker for the probability that colorectal cancer cells will proliferate and encroach into neighboring tissue. Having said this, p62 is not merely one simple ending sign for autophagy. This thing in fact reflects the stress condition that tumor cells have together. This may indicate that autophagic flux has the condition of being damaged, or that cells bear a heavier burden of proteins which they must decompose. It also can give out the signal that pathways which are connected with cell survival have already been turned on. Since LC3 and Beclin-1 were not detected in our specimens, we cannot confirm actual autophagy dysfunction only based on p62 expression. The concurrent measurement of p62, TRAP1 and HMGB1 permits us to get a more overall comprehension of the biological properties of colorectal cancer. We may carry out discussion on this from three connected domains: Mitochondrion pressure, inflammation signal transmission, and autophagy steady state maintenance.

In the research of tissue biological markers, there exists a problem that is often neglected and people seldom put forward it. The existence of statistical significance does not automatically mean that you can directly use the results in clinical environment. A part of markers indeed display differences in their expression ways, but they do not possess stable connections with invasive traits, patient survival outcomes, or reasonable inherent mechanisms. Therefore, that is the reason why they do not possess high value for practical application. To TRAP1, HMGB1 and p62, this is an altogether different situation. These three labeled molecules are connected with three core life science roads: Metabolism change arrangement, inflammation degree increase, and self-eating pressure. This linkage lets the entire chain maintain consistency, from working functional mechanisms to the detection which is conducted at the tissue level, and finally to the sorting of patients into clinical subgroups. Another point which need be mentioned: Immunohistochemistry is convenient for application in conventional pathology work, and it also does not expend overly much funds. Therefore, this thus makes it extremely having great promise for real-world clinic translation work. You may also add complementary molecular validation methods such as reverse transcription-polymerase chain reaction (RT-PCR). This additional step can hence let the research conclusions of this study become more solid and dependable on the whole.

Under this background, this research did not only aim at describing the expression situation of three proteins in colon cancer. This paper it sets out to answer three questions which are of importance to real clinical work. Firstly, do the TRAP1, the HMGB1, and the p62 possess stable, unchanging distinctions between tumor organizations and adjacent normal mucous membranes? Second, these differences have the closest connection with which kinds of bad clinicopathologic characteristics? Third, when we put all three markers into the same analysis frame, which of them is therefore more easier to show relatively independent, long-term prognosis value? The resolution of these questions can therefore promote the advancement of research concerning every individual marker. It may also lay foundation for constructing an integrated evaluation model that functions better in real clinical pathology work. Such combined evaluation model is only a future research direction instead of available clinical tool at present.

In conclusion, TRAP1, HMGB1, and p62 are classified into three core functional axes. These three are the mitochondrion chaperone-medium metabolic adaptation, inflammation correlation injury signal transmission, and autophagy/protein steady state adjustment control. These three molecules lie in different molecular groups, but every one of them possesses a tight relation to the continuous living of colon cancer cells when they encounter hypoxia, inflammation, oxidative stress, and therapeutic pressure. The research on any one of these molecules when carried out alone can only obtain very limited conclusions. Carrying out synchronous comparison of their expression in tumor and adjacent tissues in the same case group, and furthermore conducting analysis of their correlations with clinicopathologic characteristics and follow-up results, therefore can assist us to more clearly make clear which marker is more proximate to a clinically usable risk grading tool. It is for this reason that this research utilized immunohistochemistry to examine the expression quantities of TRAP1, HMGB1, and p62 in colon cancer tissue samples and paired adjacent normal mucosa. It also has combined RT-PCR for molecular verification, and has analyzed the connections of these markers with clinicopathologic parameters and total survival time. This entire study has the purpose of offering new pathologic proof for biomarker screening and prognostic assessment in colon cancer.

MATERIALS AND METHODS

Study population

We went back to collect the data of 120 patients in all. Every of these people possessed primary colon cancer and underwent the surgical operation in one unique hospital during the period from October 2022 to October 2025. The condition of these patients was pathologically confirmed after they underwent the operation. For this research we have enacted the following inclusion criteria. Firstly, patients had accepted the first-time surgical cutting removal, and their primary colon cancer was verified through postoperative pathology inspection. Second, they had not accepted any neoadjuvant treatment before the operation. Third, the complete clinical information and pathologic data of these patients are obtainable, and the corresponding paraffin-embedded tissue specimens have been appropriately kept. Fourth, for each patient, the paired adjacent normal mucosal tissue can be obtained.

We have set the exclusion criteria which are listed as below. Firstly, patients who have rectal cancer will be excluded by us. Second, the people who have repeat growth tumors or transfer diseased parts do not match the requirement of this study. Third, every person that has other concurrently happening malignant tumors is hence not eligible. The final item: Cases that have obvious missing follow-up information are also excluded. This study has obtained the approval of the hospital ethics committee firstly. We have carried out a complete analysis of all case data in the anonymized form. In all stages of the entire process, no patient information that can be identified was remained in any of the data collections.

Collection of clinicopathologic data

At first, we have collected a series of indicators which are related to patients and their diseases. These factors include patient’s age, sex, tumor position, maximal tumor diameter, histological differentiation, invasion depth, lymph node transfer, TNM stage, lymphovascular invasion, and perineural invasion. The TNM staging was done according to the 8^th edition of the American Joint Committee on Cancer staging system. Two high-grade pathologists independently examined all pathologic indexes, no third party has participation in this examination procedure.

Sources and processing of specimens

To every patient, we have collected two paraffin-embedded samples to carry out immunohistochemical analysis. One sample came from the tumor tissue itself, and hence another sample came from adjacent normal mucosal tissue which is located at least 5 cm far from the tumor edge. Pathological inspection has made confirmation that the mucosal specimen possesses no tumor infiltration. Another aspect worthy of note is we have selected 40 cases which possess both fresh tumor tissue and matched adjacent normal mucous membrane. All these sample pieces were kept complete, their RNA had enough quality to satisfy the demands of analysis. We put these samples into quick freezing in liquid nitrogen just after the operation, and hence afterward we utilized them for RT-PCR analysis.

Immunohistochemistry and evaluation criteria

We by hand cut paraffin sections into a thickness of 4 micrometer. At first, we completed all the standard procedures: The removing of paraffin, the re-hydration, the antigen recovering and the blocking. After this step, we used first-stage antibodies that aim at TRAP1, HMGB1, and p62 one by one, and let them stand to incubate through the night at 4 °C. On the following day, we carried out the addition of HRP-marked secondary antibody substances. We have utilized DAB to carry out color development, thereafter we did counterstaining work for the samples by hematoxylin. Regarding the negative control group, we utilized phosphate buffered saline in the place of the primary antibody. All primary antibodies used are commercial standardized products; limited by incomplete original experimental archives, detailed antibody source, dilution ratio and positive control data are not supplemented in this manuscript and will be supplemented in future research.

Our majority work of evaluation was on TRAP1 staining within the cytoplasm. Regarding the staining of HMGB1, we have carried out observation on both the nucleus and the cytoplasm. Regarding to the p62 staining work, we mainly have observed that it presents as granular signals inside the cytoplasm. Our side utilized a half-quantitative marking system for the purpose of assessment. We have carried out the grading of staining intensity by using a scale from 0 to 3, and the proportion of positive cells we graded on a scale from 0 to 4. We have added these two scores together for obtaining the total score. The sum score from 0 to 3 indicates the low level of expression, while the score from 4 to 7 indicates the high level of expression. Two senior pathologists all completed every assessment by themselves, and they did not have the right to obtain clinical data when they carried out this work. All differences among their evaluations were processed by means of common discussion, hence they were able to get a final agreement. It is precisely this type of straightness.

RT-PCR

We by the TRIzol method have extracted total RNA. We firstly examined the pureness and consistence of the RNA to ensure that they reached the standard, hence we synthesized complementary DNA through reverse transcribe. We have utilized GAPDH to act as the internal control. We have completed the establishment of reactions and conducted the SYBR green-based RT-quantitative PCR strictly in accordance with the instructive documents provided by the manufacturer. We calculated the relative expression levels of TRAP1, HMGB1, and p62 with the 2^-ΔΔCt method. We have carried out three tests on every sample, and thus obtained the average value for the analysis that follows. It is precisely this type of straightness.

Follow-up

We have carried out follow-up work for patients by means of outpatient coming and telephone contacts. The follow-up time period was concluded in March 2026. The endpoint we utilized was OS, which is abbreviated as OS, it indicates the time length that lies between the date of operation and either the death of the patient or the last follow-up that has been confirmed. We marked patients who were still alive at the time when follow-up finished as censored, hence we used the date of their last confirmed survival check to be the cutoff point.

Statistical analysis

We carried out the statistical analyses by utilizing the SPSS 26.0. The continuous variables are manifested by mean plus-minus standard deviation. We have carried out paired-sample comparisons through the use of paired t test, and completed independent-sample comparisons through the use of independent-sample t test. Category-type variables are showed as n (%), and differences between groups were compared through utilization of the χ² test or Fisher’s exact test. We used the Kaplan-Meier method together with the log-rank test to carry out the survival analysis. We have constructed multivariable logistic regression which takes lymph node metastasis as the dependent variable, and we have established a Cox proportional hazards model which takes OS as the endpoint. A P value which is smaller than 0.05 was regarded to have statistics meaning.

RESULTS

General clinicopathologic characteristics

In this study, one hundred twenty patients who have colon cancer are totally included. The average age value was 62.9 plus-minus 10.8 years. In our study, the number of patients whose age was smaller than 60 years was 42, which accounts for 35.0%, and the number of patients whose age was not smaller than 60 years was 78, which accounts for 65.0%. Among all these patients, sixty-nine are male (57.5%), and fifty-one are female (42.5%). Forty-seven tumor masses were situated in the right-side colon (39.2%) and 73 in the left-side colon (60.8%). In 58 patients, the tumor diameter was smaller than 5 cm, which accounts for 48.3%, and in 62 patients, it was not smaller than 5 cm, which accounts for 51.7%. In 84 cases which account for 70.0%, the tumors are in good or middle differentiation status, while 36 cases which account for 30.0% have low differentiation. The T1-2 stage disease existed in 29 cases (24.2%), and the T3-4 stage disease existed in 91 cases (75.8%). The situation of lymph node metastasis was negative in 64 cases (53.3%) and positive in 56 (46.7%). TNM stage I-II and III-IV, each of them have 60 cases that account for 50.0% respectively. Regarding lymphovascular invasion, 82 cases (68.3%) had negative result, and 38 cases (31.7%) had positive result. As for perineural invasion, 91 cases (75.8%) were negative, 29 cases (24.2%) were positive (Table 1).

Table 1 General clinicopathologic characteristics of 120 patients with colon cancer, n (%).

Variable	n (%)
Age (years), mean ± SD	62.9 ± 10.8
Age < 60 years	42 (35.0)
Age ≥ 60 years	78 (65.0)
Male	69 (57.5)
Female	51 (42.5)
Right colon	47 (39.2)
Left colon	73 (60.8)
Maximum tumor diameter < 5 cm	58 (48.3)
Maximum tumor diameter ≥ 5 cm	62 (51.7)
Well/moderately differentiated	84 (70.0)
Poorly differentiated	36 (30.0)
Depth of invasion T1-2	29 (24.2)
Depth of invasion T3-4	91 (75.8)
LNM negative	64 (53.3)
LNM positive	56 (46.7)
TNM stage I-II	60 (50.0)
TNM stage III-IV	60 (50.0)
LVI negative	82 (68.3)
LVI positive	38 (31.7)
PNI negative	91 (75.8)
PNI positive	29 (24.2)

LNM: Lymph node metastasis; TNM: Tumor-node-metastasis; LVI: Lymphovascular invasion; PNI: Perineural invasion.

Comparison of immunohistochemical expression of TRAP1, HMGB1, and p62 between cancer tissues and adjacent normal mucosa

Immunohistochemistry experiment has proven that the high-expression frequencies of TRAP1, HMGB1, and p62 in colon cancer organization are 52.5%, 48.3%, and 45.0%, one by one, all are obviously higher than the corresponding frequencies in matched side normal mucous membrane (20.0%, 15.8%, and 17.5%, one by one; all P values are smaller than 0.001; Table 2).

Table 2 Comparison of TRAP1, HMGB1, and p62 expression between colon cancer tissues and paired adjacent normal mucosa, n (%).

Marker	High expression in tumor	High expression in adjacent tissue	Statistic	P value
TRAP1	63 (52.5)	24 (20.0)	24.635	< 0.001
HMGB1	58 (48.3)	19 (15.8)	28.756	< 0.001
p62	54 (45.0)	21 (17.5)	20.856	< 0.001

RT-PCR results of paired specimens

In our 40 newly collected matching samples, the relative expression levels of mRNA for TRAP1, HMGB1, and p62 are all higher in tumor organizations than in neighboring normal organizations, and all differences have statistics meaning (all P < 0.001; Table 3).

Table 3 Reverse transcription-polymerase chain reaction results of TRAP1, HMGB1, and p62 in 40 paired specimens, mean ± SD.

Marker	Tumor tissue (n = 40)	Adjacent tissue (n = 40)	t value	P value
TRAP1	2.31 ± 0.74	1.02 ± 0.29	10.64	< 0.001
HMGB1	2.08 ± 0.69	0.98 ± 0.25	9.87	< 0.001
p62	1.94 ± 0.63	1.01 ± 0.27	8.95	< 0.001

Relationship between TRAP1 expression and clinicopathologic characteristics

The expression of TRAP1 has not displayed any obvious connection with age, sex, or tumor position (all P > 0.05), but it is connected with tumor dimension, differentiation degree, invasion depth, lymph node metastasis, TNM stage, lymphovascular invasion, and perineural invasion (all P < 0.05; Table 4).

Table 4 Relationship between TRAP1 expression and clinicopathologic characteristics in colon cancer, n (%).

Clinicopathologic variable	Low TRAP1	High TRAP1	χ2	P value
Age < 60 years	22 (52.4)	20 (47.6)	0.617	0.432
Age ≥ 60 years	35 (44.9)	43 (55.1)
Male	31 (44.9)	38 (55.1)	0.431	0.512
Female	26 (51.0)	25 (49.0)
Right colon	23 (48.9)	24 (51.1)	0.064	0.800
Left colon	34 (46.6)	39 (53.4)
Maximum tumor diameter < 5 cm	36 (62.1)	22 (37.9)	9.555	0.002
Maximum tumor diameter ≥ 5 cm	21 (33.9)	41 (66.1)
Well/moderately differentiated	47 (56.0)	37 (44.0)	8.022	0.005
Poorly differentiated	10 (27.8)	26 (72.2)
Depth of invasion T1-2	20 (69.0)	9 (31.0)	7.066	0.008
Depth of invasion T3-4	37 (40.7)	54 (59.3)
LNM (-)	42 (65.6)	22 (34.4)	18.067	<0.001
LNM (+)	15 (26.8)	41 (73.2)
TNM stage I-II	40 (66.7)	20 (33.3)	17.678	<0.001
TNM stage III-IV	17 (28.3)	43 (71.7)
LVI (-)	46 (56.1)	36 (43.9)	7.676	0.006
LVI (+)	11 (28.9)	27 (71.1)
PNI (-)	48 (52.7)	43 (47.3)	4.158	0.041
PNI (+)	9 (31.0)	20 (69.0)

LNM: Lymph node metastasis; TNM: Tumor-node-metastasis; LVI: Lymphovascular invasion; PNI: Perineural invasion.

Relationship between HMGB1 expression and clinicopathologic characteristics

We have carried out stratified analysis for examining HMGB1 expression modes among different patient and tumor subgroups. The high expression of HMGB1 mainly appears in poorly differentiated tumor tissues, cases which have positive lymph node metastasis, and tumors that are in TNM stages III-IV. Firstly, we have carried out inspections on the differences in distribution which are connected with other clinical factors. When we carry out grouping according to age, sex, tumor location, tumor size, lymphovascular invasion, or perineural invasion, HMGB1 expression levels have no obvious differences. We have additionally researched the connection between HMGB1 expression level and invasion depth. This association has nearly achieved the statistical significance, which Table 5 displays.

Table 5 Relationship between HMGB1 expression and clinicopathologic characteristics in colon cancer, n (%).

Clinicopathologic variable	Low HMGB1	High HMGB1	P value
Age < 60 years	24 (57.1)	18 (42.9)	0.491
Age ≥ 60 years	38 (48.7)	40 (51.3)
Male	35 (50.7)	34 (49.3)	0.956
Female	27 (52.9)	24 (47.1)
Right colon	24 (51.1)	23 (48.9)	1.000
Left colon	38 (52.1)	35 (47.9)
Maximum tumor diameter < 5 cm	33 (56.9)	25 (43.1)	0.354
Maximum tumor diameter ≥ 5 cm	29 (46.8)	33 (53.2)
Well/moderately differentiated	49 (58.3)	35 (41.7)	0.042
Poorly differentiated	13 (36.1)	23 (63.9)
Depth of invasion T1-2	20 (69.0)	9 (31.0)	0.054
Depth of invasion T3-4	42 (46.2)	49 (53.8)
LNM (-)	40 (62.5)	24 (37.5)	0.018
LNM (+)	22 (39.3)	34 (60.7)
TNM stage I-II	41 (68.3)	19 (31.7)	< 0.001
TNM stage III-IV	21 (35.0)	39 (65.0)
LVI (-)	46 (56.1)	36 (43.9)	0.219
LVI (+)	16 (42.1)	22 (57.9)
PNI (-)	48 (52.7)	43 (47.3)	0.837
PNI (+)	14 (48.3)	15 (51.7)

HMGB1 was mainly associated with differentiation, lymph node metastasis, and tumor-node-metastasis stage.

LNM: Lymph node metastasis; TNM: Tumor-node-metastasis; LVI: Lymphovascular invasion; PNI: Perineural invasion.

Relationship between p62 expression and clinicopathologic characteristics

When we make comparison with HMGB1, the correlation mode of p62 has much more strong direction. The high expression of p62 mostly appears in patients who have poorly differentiated tumors, lymph node metastasis, and advanced-stage disease. We have not discovered any obvious differences when we have investigated factors such as age, sex, tumor position, tumor volume, invasion depth, lympho-vascular invasion, or perineural invasion (Table 6).

Table 6 Relationship between p62 expression and clinicopathologic characteristics in colon cancer, n (%).

Clinicopathologic variable	Low p62	High p62	P value
Age < 60 years	24 (57.1)	18 (42.9)	0.878
Age ≥ 60 years	42 (53.8)	36 (46.2)
Male	36 (52.2)	33 (47.8)	0.590
Female	30 (58.8)	21 (41.2)
Right colon	27 (57.4)	20 (42.6)	0.807
Left colon	39 (53.4)	34 (46.6)
Maximum tumor diameter < 5 cm	35 (60.3)	23 (39.7)	0.340
Maximum tumor diameter ≥ 5 cm	31 (50.0)	31 (50.0)
Well/moderately differentiated	52 (61.9)	32 (38.1)	0.034
Poorly differentiated	14 (38.9)	22 (61.1)
Depth of invasion T1-2	18 (62.1)	11 (37.9)	0.506
Depth of invasion T3-4	48 (52.7)	43 (47.3)
LNM (-)	42 (65.6)	22 (34.4)	0.020
LNM (+)	24 (42.9)	32 (57.1)
TNM stage I-II	43 (71.7)	17 (28.3)	< 0.001
TNM stage III-IV	23 (38.3)	37 (61.7)
LVI (-)	49 (59.8)	33 (40.2)	0.180
LVI (+)	17 (44.7)	21 (55.3)
PNI (-)	52 (57.1)	39 (42.9)	0.534
PNI (+)	14 (48.3)	15 (51.7)

High p62 expression was mainly associated with differentiation, lymph node metastasis, and tumor-node-metastasis stage.

LNM: Lymph node metastasis; TNM: Tumor-node-metastasis; LVI: Lymphovascular invasion; PNI: Perineural invasion.

Multivariable logistic regression analysis of factors associated with lymph node metastasis

When we constructed the multivariable logistic regression model, we took lymph node metastasis to be the dependent variable. We moreover put five other factors into our model, they are tumor dimension, differentiation degree, invasion depth, lymphovascular invasion, and the expression levels of TRAP1, HMGB1, and p62. Through the final analysis, three independent factors which are associated with lymph node metastasis have been picked out, they are lymphovascular invasion, high TRAP1 expression, and high HMGB1 expression (Table 7). These factors are the key pushing forces that we have discovered.

Table 7 Multivariable logistic regression analysis of factors associated with lymph node metastasis in colon cancer.

Variable	B	SE	Wald χ2	OR	95%CI	P value
Maximum tumor diameter ≥ 5 cm	0.531	0.367	2.093	1.70	0.80-3.60	0.148
Poor differentiation	0.642	0.377	2.892	1.90	0.90-4.10	0.089
Depth of invasion T3-4	0.470	0.447	1.106	1.60	0.60-4.00	0.293
Lymphovascular invasion positive	0.916	0.417	4.828	2.50	1.10-5.70	0.028
High TRAP1 expression	1.386	0.421	10.828	4.00	1.80-8.90	0.001
High HMGB1 expression	0.788	0.399	3.910	2.20	1.00-4.90	0.048
High p62 expression	0.405	0.350	1.346	1.50	0.70-3.40	0.246

Follow-up results and OS analysis

All the one hundred and twenty patients have been included by us into the follow-up analysis. Follow-up work was finished up to March 2026, with a middle value of duration being 26 months (extent, about 6-40 months). In this time period, 24 patients have passed away, therefore giving a 3-year total survival rate of 76.7%. Kaplan-Meier analysis has proven that patients who have high TRAP1, HMGB1, and p62 expression possess significantly lower OS when compared with the ones who have low expression (all P < 0.05). According to the single-factor survival analysis, age, differentiation degree, lymph node metastasis, lymphovascular invasion, and the expression quantities of these three proteins were put into an exploratory multi-factor Cox regression model. The outcomes showed that lymph node metastasis, lymphovascular invasion, and high TRAP1 expression have connection with worse total survival (Table 8).

Table 8 Multivariable Cox regression analysis of overall survival in patients with colon cancer.

Variable	B	SE	HR	95%CI	P value
Age ≥ 60 years	0.412	0.391	1.51	0.70-3.27	0.292
Poor differentiation	0.603	0.428	1.83	0.79-4.24	0.158
LNM positive	0.998	0.415	2.71	1.20-6.10	0.016
LVI positive	0.806	0.405	2.24	1.01-4.94	0.047
High TRAP1 expression	1.061	0.412	2.89	1.29-6.48	0.010
High HMGB1 expression	0.562	0.398	1.75	0.80-3.83	0.157
High p62 expression	0.478	0.387	1.61	0.75-3.45	0.216

LNM: Lymph node metastasis; LVI: Lymphovascular invasion; HR: Hazard ratio; CI: Confidence interval.

DISCUSSION

This research has carried out close examination on three marker substances: TRAP1, HMGB1, and p62. This research specially carried out exploration on their expression patterns in tissues, relations with clinicopathologic characteristics of colon cancer, and the value for predicting prognosis of this disease. This research team has carried out a comparison on the levels of the three marked objects inside tumor organizations and adjacent normal mucous membranes. They through research discovered that both protein and mRNA contents of all three items are higher in tumor specimens. Having said this, the marking objects did not possess the same clinical meaning meanings. TRAP1 has displayed the most broadest connections to aggressive tumor characteristics. These characteristics include tumor volume, differentiation degree, invasion depth, lymph node shift, TNM stage, and lymphovascular invasion. When the research team conducted multivariable analyses, TRAP1 therefore still remained continuously correlated with lymph node metastasis and the whole patient survival. HMGB1 has the most correlation with three particular factors: Bad differentiation, lymph node metastasis, and later tumor stage. p62, on its own side, more acts as a second-order marker that appears together with tumor progress advancement. It possessed comparatively restricted independent forecasting value.

These result points not only toward a general increasing of activity in all three markers. They also thus display an explicit hierarchy in the manner each makes contribution to colon cancer pathological changes[16]. TRAP1 may be regarded as a core biomarker which connects invasion, metastasis, and patient prognosis together. Foregoing study has already carried out exploration on the working way of TRAP1 in the condition of low oxygen[17]. It can decelerate the generation of ribosomes, and assist colorectal cancer cells in making adaptation to the pressure inside the human body. This therefore provides a relatively reasonable biology explanation for why TRAP1 content is connected with deeper tumor invasion, higher metastasis frequencies, and worse patient results.

From molecular mechanism, TRAP1 acts as upstream mitochondrial chaperone regulating multiple tumor survival pathways, while HMGB1 mainly reflects secondary inflammatory microenvironment change and p62 represents downstream autophagy-related stress accumulation, which explains why only TRAP1 owns stable independent prognostic effect in multivariate analysis[18].

Our research discovered TRAP1 possesses stronger clinical connection, and this therefore is likely linked to its function as an upstream central node in the tumor survival network. Colon cancer cells cope with continuous harsh environments as tumors develop forward. These things include anoxia, changing nourishment contents, oxidizing pressure, and the selection force that is brought by therapy. A portion of molecular matters can carry out the regulation of mitochondrial homeostasis, metabolic reprogramming, and anti-apoptotic signaling all in one same time. Tumors which possess these molecules have the tendency to be more aggressive, hence patients normally have poorer clinical results on account of this. Numerous correlative elements only exert influence upon the end segments of individual signal transmission paths[19]. TRAP1 is not same at this place, it seems to possess both the function of buffering stress and the function of maintaining phenotype. When the expression of TRAP1 obtains enhancement, it not merely represents the increment of biological activity inside the tumor. It also indicates that tumor cells have stronger capability to live through bad micro-surroundings, maintain clonal health, and resist treatment pressure. We have made comparison between TRAP1, HMGB1 and p62, thus we found TRAP1 possesses a far stronger independent connection with clinical treatment results. This result is not a random one, and it conforms to the core function of TRAP1 in adjusting the mitochondrion and supporting the adaptive ability of tumors.

The research results regarding HMGB1 indicate that the progression of colon cancer does not only depend on cell proliferation. It also brings major alterations to inflammation signal and stress-connected position movement. In this research work, we discovered that high HMGB1 expression has connection with bad differentiation, lymph gland metastasis, and more later tumor phase. But when we have carried out multivariable models, its effect magnitude has turned out to be weaker than the one of TRAP1. This indicates HMGB1 very likely reflects the inflammatory activation background when tumors grow, instead of being the uppermost upstream determinant in every situation[3]. Another aspect which needs mention: Previous mechanism studies have expounded how HMGB1 operates. It can facilitate ERK-mediated Drp1 phosphorylation by way of receptor for advanced glycation end-products, which hence promotes mitochondrial fission, autophagy activation, and chemoresistance. Such mechanism conclusion comes from published cell experiments instead of verification from our clinical data. This entire procedure enables colorectal cancer cells to possess greater capability for regrowth when they encounter treatment-related pressure. Seen from this angle, the connection we discovered among HMGB1, tumor stage and lymph node metastasis in this research therefore likely corresponds to more favorable adaptation inside the inflammation-stress-autophagy network[8]. It is after all not merely a static marking thing for the inflammation reaction. HMGB1 is present at the same time in the nucleus and in the cytoplasmic/extracellular space, therefore it possesses dual functions. Its pathological meaning generally is dependent on the tiny environment and the position that it takes inside the cell. This, therefore, can also explain why until the present moment, research works have not yet reached completely consistent conclusions regarding its prognostic influence.

Another item that has value to be mentioned. Our research discovered that HMGB1 is connected with negative pathological characteristics, however it has not displayed the same stable independent effect as TRAP1 in the survival analysis. This does not indicate that HMGB1 has no importance. Its clinical related degree just perhaps have big change according to the special situation. To patients who suffer from rectal cancer and receive preoperative chemoradiotherapy, higher expression level of HMGB1 has been associated with resistance to treatment and worse shrinkage of tumor. This gives a hint that the value it has for predicting how good treatment effect is differs from its function in monitoring how the illness develops by itself[14]. Thus, HMGB1 may more appropriately be depicted as a functional biological marker connected with inflammation signal conduction, treatment resistance, and more intense microenvironment responses, hence it is not a universal independent marker for long-term survival. According to the results gotten from our research, this perspective lets us give a more careful definition to the role of HMGB1[5]. This method can assist in discovering more aggressive tumor bodies that possess a more active micro-environment, but whether this method can function as an independent prognosis label requires more inspections from more strictly divided researches in the future.

The research results concerning p62 also provided us with useful information. In this research work, the higher expression level of p62 was associated by us with tumor differentiation degree, lymph node metastasis occurrence, and TNM stage classification. This kind of mode means that its increasing is just in accordance with the higher malignant degree of tumor. Having said this, p62 was not obtained as a stable independent correlation factor in the multivariable analysis. We cannot deduce actual autophagy pathway changes only from p62 expression without LC3 and Beclin-1 detection, so p62 is merely a correlative marker in our research. This type of tendency - in which a marker displays correlation but restricted independent forecast value - appears quite frequently in studies about autophagy-connected markers. Schmitz et al[19] have reported that the prognostic connection of p62, LC3, Beclin-1, and other autophagy marks in colorectal cancer is greatly influenced by KRAS condition, testing method, and the end points that are utilized. This indicates that there is quite large clinical heterogeneity in the performance of these markers. Therefore, the fact that p62 did not successfully enter the final independent model in this research does not mean that it possesses no biological function value at all. It more possibly acts as a compound signal that reflects the stress burden of tumor cells, the aggregation of protein, and the changed degradation pathways, instead of being a single, stable, linear mark that is linked to one particular clinical result[7]. It is needful to point out that this type of mixed outcome is quite the normal circumstance when people conduct study on complex pathway-connected biological markers.

We can carry out observation on p62 via a more simple method: It functions as a connection which binds autophagy, inflammation, and drug resistance together. Recently conducted laboratory experiments on cell specimens have examined colorectal cancer cells which do not give reaction to 5-fluorouracil and oxaliplatin. These experiments discovered that p62 maintains these cells’ survival through enhancing signals which sustain cell living, hence adjusting p62 content can alter how sensitive these cells are toward the medicines. These in vitro findings cannot be verified by our clinical specimen data. These obtained results do not conflict with the things that we observed in our research. Our research team has discovered that p62 is connected with more late-stage cancer phases and elevated risk of metastasis, however, it does not function as a strong independent prognostic marking molecule[20]. This possesses rationality when you consider the working mechanism of p62. This thing probably has a bigger function in the later phases of cancer development, especially when cells are being put under repeated pressure by treatment and continuous cell pressure. When you carry out observation on a general patient population without carrying out their division into more specific small subgroups, the effect of p62 is concealed by other more steady markers which perform action at an earlier position along the biological pathway. Therefore, p62 is not the optimal selection when you desire a single marker that alone can forecast the results of illness for patients. Having said this, it is still a very precious beginning point. This assists us to find out how colon cancer cells make adjustment through autophagy, how inflammation keeps active inside tumors, and why treatment responses can change along with time.

This research also has one other key inspiration. We have carried out direct comparison on TRAP1, HMGB1, and p62 in the identical cohort, all within a unified clinicopathological framework. The majority of previous research has respectively examined mitochondrial chaperones, inflammatory mediators, and autophagy-related markers[9]. But actual clinical situations prove these markers all occur together, in the same tissue, the same patient, and through the entire process of disease development. The result that we have got indicates one definite rule. All the three markers have ascending expression levels and are connected with malignant characters. Having said the above, the respective contributions that they make to these phenotypes do not possess equal weight.

TRAP1 appears to have a tighter connection with invasive behaviors, metastatic ability, and long-term treatment results of patients. This rule supports the view that it is a core element which determines the degree of a tumor’s aggression. We, therefore, can carry out the observation of HMGB1 from a different angle. It possesses a stronger linkage to inflammatory reconstruction of the microenvironment and the whole disease development. This result conforms to its function which depends on context, it is shaped through the interactions that exist between the tumor and the host. Now we talk about p62. It is more reasonable to regard it as a sign which appears together with cellular stress and alterations of autophagic flux, hence it is not a main push factor which directly forms how the disease develops forward.

These dissimilarities therefore mean we are not able to merely examine statistical significance when we evaluate biological markers. We must place more emphasis on three key aspects of every molecule: The pathological degree, signal order arrangement, and what clinical application it is prepared for. The true transformation value of one biological marker only appears thus when it aligns correctly with one particular clinical problem.

Several limitations exist in the present study. First, this is a single-center retrospective investigation lacking an external validation cohort, which precludes immediate clinical application of our findings for patient risk stratification, and concentrated case recruitment inevitably introduces potential selection bias. Second, the cutoff value defining TRAP1 expression was adopted from published references; we failed to verify the stability of this threshold and the reproducibility of detection across different testing platforms based on the 52.5% positive rate observed in our cohort. Third, the wide 95% confidence interval (1.29-6.48) of TRAP1’s hazard ratio for OS reduces the reliability of its prognostic implication. Fourth, the median follow-up duration was only 26 months, and disease-free survival data were not collected in this work. Moreover, the sample size for RT-PCR verification was limited. Although RT-PCR results were consistent with immunohistochemistry trends, further validation in larger cohorts is required. Additionally, we only adopted OS as the survival endpoint without analyzing disease-free survival, tumor recurrence or postoperative adjuvant therapy-related outcomes, which may underestimate the predictive value of these biomarkers under diverse treatment conditions. No cellular or mechanistic functional experiments were performed herein; accordingly, our data can only demonstrate statistical correlations rather than definitive causal relationships between biomarker expression and disease progression[8]. Future multi-center prospective studies combined with stratified treatment information and functional assays targeting the TRAP1/HMGB1/p62 axis will help clarify their exact biological functions and predictive roles in colon cancer progression and therapeutic response.

The present research utilizes a dual-evidence method which combines immunohistochemistry and RT-PCR. This method lets the research results be much more easier to explain compared with the situation that researchers only depend on one single experiment method. Immunohistochemistry can display the position of proteins inside tissue and the intensity of their expression, all when the original structural environment of the tissue is kept. It arranges itself very in accordance with the standard pathology examinations employed in common laboratory work. On the other side, RT-PCR can give the supporting evidence that comes from the transcriptional level. This tool assists in the reduction of biases which originate from non-uniform staining thresholds, changeable antibody working effects, and subjective assessments made by different observing persons[2]. At the beginning we thought the two methods might have small differences, but our final analysis result showed that their generated trends are completely consistent. This alignment proves that the higher expression of TRAP1, HMGB1, and p62 in tumor tissues originates from the actual molecular activation, not the accidental one-time staining mistakes. This kind of cross-checking lets the outcomes become much more firmer. This also provides a more reliable basis for the utilization of these markers in future translational and clinical research work.

When we observe real-world clinical translation possibility, the results we obtained from this research also indicate a clear change direction for colon cancer pathologic evaluation in the future. This research may gradually walk away from depending on single label judgments, and hence turn to layered, compound explanation of many index signs. Let us take one common clinical situation as an example[15]. If one patient’s sample displays high TRAP1 expression, and HMGB1 and p62 quantities rise in the same time, clinical doctors and pathology workers should give additional attention to three overlapped processes inside the tumor. These three items are metabolic adaptation, inflammatory activation and autophagic remodeling, and their happening together usually gives the hint that the tumor may have a more aggressive character, or possess higher resistance to conventional treatments. An additional circumstance is that only p62 obtains elevation, and TRAP1 expression maintains at a normal, not notable level. This kind of pattern more possibly is a symbol of local pressure and broken protein steady-state condition in the interior of the tumor. It usually will not act as an alone index of bad long-term results for patient people. Many people may hold the assumption that this combined interpretation method already can be immediately applied in clinical work, but at the current stage, this situation is not true. It still first needs to get verification in bigger patient groupings. Even so, these outcomes already transmit a clear information: We cannot continue to utilize the over-simplified straight line model to comprehend colon cancer’s biological conduct, the model that supposes one molecule directly connects to one sole clinical outcome. We on the other hand should construct a more feasible framework that is based on network, has layers, and brings particular environmental elements into consideration. This is without any doubt not a tiny change for this research domain.

Looked at all together, this research supports one important conclusion. TRAP1, HMGB1, and p62 cannot act as mutually replaceable parallel biological markers. Instead, they carry disease-related path information from three different dimensions of colon cancer cells: Mitochondrion adaptation, inflammation pressure, and self-eating balance. Every single one of these molecules undertakes a special function. TRAP1 possesses the most strongest comprehensive prediction ability in all experiments of this research. HMGB1 reflects the way in which the inflammation microenvironment participates in the development of tumors. p62 gives the suggestion that autophagy and stress adaptation may play an assisting role in the forming of invasive phenotypes. It is relatively uncomplicated once you examine the data one next to another. These result findings therefore bring real practical value to the clinical practice work. They do not merely contribute three positive molecules into the current reference catalogue. They also point out clearly that the assessment of colon cancer risk should not only adhere to a purely morphology-based perspective. It is necessary that it carry out a shift toward the integrated recognition of metabolic, inflammatory, and protein-homeostasis networks.

CONCLUSION

Compared with the corresponding adjacent normal mucous membranes, TRAP1, HMGB1, and p62 all display higher expression levels in colon cancer tissue samples. The RT-PCR results thus give support to this upregulation pattern on the molecular level. These three labeling things have different degrees of connection with aggressive clinical pathologic characteristics. Among these factors, TRAP1 displays a comparatively stronger connection with lymph node metastasis and OS, but wide hazard ratio confidence interval and single-center design restrict immediate clinical prognostic application. HMGB1 appears to possess a more proximate connection with the progression of disease. p62, on its own side, is relatively more able to reflect alterations of cellular homeostasis in the course of tumor formation. After all, the present research is merely one-center retrospective analysis. Further large cohort validation is required before clinical transformation. The conclusions that we have obtained here therefore need to be further verified in bigger group samples with longer follow-up time periods.