Relationship between metalloproteinase-9 (MMP-9) expression and clinicopathology in colorectal cancer: a cross-sectional study

Abstract

Introduction:

According to the WHO’s GLOBOCAN database, ~1,931,590 new colorectal cancer (CRC) cases and 915,607 CRC-related deaths occurred in 2020. The incidence of CRC in Indonesia is 8.6%, making it the fourth most common cancer. With CRC, matrix metalloproteinase-9 (MMP-9) has a role in tumour development and progression, such that patients with a higher MMP-9 expression had poorer survival. This study aimed to analyze the relationship between MMP-9 expression and clinicopathology in CRC patients.

Methods:

This was an analytic observational study with a cross-sectional research design. It was conducted from November 2021 to June 2022 with 52 patient tissue samples: these were subjected to MMP-9 immunohistochemistry stain, with the GeneTex (Irvine) MMP-9 monoclonal antibody. Patient data were collected with clinical variables based on medical records and histopathological examination by anatomy pathologists.

Results:

Primary tumour location, cancer staging, and histopathology grading were associated with MMP-9 (P=0.016, P=0.001, P=0.049). The more proximal to the primary tumour, the higher the stage of cancer, and the higher the histopathological grade, thus the greater the expression of MMP-9.

Conclusion:

A significant relationship existed of primary tumour location, cancer staging, and histopathology grading with MMP-9 expression in CRC patients. MMP-9 expression could be a useful indicator for the clinical assessment of tumour biologic behaviour and prognosis in CRC patients.

Introduction

Highlights

• Matrix metalloproteinase-9 (MMP-9) plays a major role in invasion and metastasis.

• Primary tumour location, staging, and grading relate to MMP-9 expression in colorectal cancer.

• Age, sex, and histopathology do not relate to MMP-9 expression.

• MMP-9 expression can assist assessment of colorectal cancer tumour behaviour and prognosis.

According to the WHO’s GLOBOCAN database, ~1,931,590 new colorectal cancer (CRC) cases and 915,607 CRC-related deaths took place in 2020. Regional estimates suggest that half the CRC deaths and 5-year-old cases were also seen in Asia. Approximately 70% of CRC cases are sporadic from environmental factors, such as habitual eating patterns, physical activity, smoking, and alcohol consumption¹. CRC in Indonesia as per GLOBOCAN 2020 saw 34,189 of 396,914 (8.6%), making it the fourth most common cancer: its mortality rate ranked eighth for colon cancer and tenth for rectal cancer^2,3.

Matrix metalloproteinase-9 (MMP-9) is a member of the most important matrix metalloproteinase type gelatinase, or gelatinase-B/92 kDa type IV collagenase. It can destroy type IV collagen in the basement membrane and the extracellular matrix, playing a major role in invasion and metastasis⁴. It was also found that surfaces binding to MMP-9 activate TGF-beta, in turn playing a role in tumour invasion and angiogenesis⁵.

MMP-9 inhibitors are now being developed as anti-metastatic cancer therapy. Their expression increases in various cancers, such as breast, prostate, nasopharyngeal, laryngeal, and in bladder transitional carcinoma. However, several studies reported conflicting results regarding the relationship of MMP-9 with clinical parameters of CRC patients⁶; this study analyzed the relationship between MMP-9 expression and clinicopathology.

Materials and methods

This is an analytic observational study with a cross-sectional research design. It was conducted in Dr. Wahidin Sudirohusodo and Hasanuddin University Hospital, Makassar, Indonesia, and strengthening the reporting of cohort, cross-sectional and case-control studies in surgery (STROCSS)⁷.

Inclusion criteria included patients over 18 years, those with a histopathological diagnosis of CRC, and those scheduled to undergo surgery, chemotherapy, or radiotherapy; histopathological samples were used for immunohistochemical examination. Exclusion criteria involved patients unwilling to participate in the study; incomplete patient examination data (demographics, location of CRC, CRC staging, histopathology characteristics, and grading); immunohistochemical examination which could not be read, and patients with autoimmune, inflammatory, or fibrotic diseases.

We included 52 patients, histopathologically diagnosed with CRC, to undergo surgery, chemotherapy, or radiotherapy. The researcher collected paraffin block preparations for CRC patients having surgery from November 2021 to June 2022, sending them to the Anatomical Pathology department of Hasanuddin University Hospital, Makassar.

The clinicopathological variables were taken from medical records and demographic and clinical data from physical and supporting examinations. These included age, sex, primary tumour location, cancer stage, histopathology and grading. We obtained ethical approval from the Ethical Committee of the Faculty of Medicine, Hasanuddin University (License No. 90/ UN4.6.4.5.31/PP36/2022), with the study registered with the Research Registry (No. 8274).

Samples underwent MMP-9 immunohistochemistry staining with the MMP-9 monoclonal antibody from GeneTex (Irvine; Catalog No. GTX60482), processed according to standard protocol. Patient data were collected, clinical variables were based on medical records, and histopathological examination results were assessed by anatomical pathologists.

Assessment of MMP-9 expression was based on an analysis of the percentage of positive tumour cells and staining intensity. We used the immunoreactive score: 0 (unstained), 1 (1–5% of all tumour cells), 2 (6–75% of all tumour cells), or 3 (76–100% of all tumour cells). Based on brown colour intensity, malignant cells showing MMP-9 overexpression were divided with a 0–3 scale: 0 (negative), 1 (weak), 2 (moderate), or 3 (strong). The percentage score of immunoreactive tumour cells was multiplied by the intensity score to obtain a total score: negative (0–2) or positive (>2).

Univariate analysis measured frequency with a percentage for each variable. Bivariate analysis used χ² tests with a significance value of P less than 0.05 to assess the comparison between categorical variables. This statistical analysis was performed with SPSS v. 27 program (Armonk: IBM Corp).

Results

Table 1 shows data distribution among 52 patients. The sample was dominated by the 50–59-year age group (34.6%), followed by the 40–49-year group (23.1%), 30–39-year age group (21.2%), 60–69-year age group (13.5%), and the 70–79-year age group (7.7%.) More patients were male (69.2%) than female (30.8%). The highest frequency of tumour location was the rectum for 30 patients, followed by the left colon (14; 26.9%), with the lowest frequency in the right colon (8; 15.4%). Grouped on embryology, the right-sided type included only 8 (15.4%) from the right colon, and the left-sided type included the left colon to the rectum in 44 (84.6%). The distribution of cancer stages was almost even: 17 (32.7%) patients had stage II cancer, 17 (32.7%) had stage III cancer, and 18 (34.6%) had stage IV cancer. Histopathological grading found that most were low-grade (61.5%), high-grade (21.2%) or moderate-grade (17.3%).

Table 1.

Participant characteristics

	Total
Variable	n	%
Age (years)
30–39	11	21.2
40–49	12	23.1
50–59	18	34.6
60–69	7	13.5
70–79	4	7.7
Sex
Mele	36	69.2
Femele	16	30.8
Location
Right colon	8	15.4
Left colon	14	26.9
Rectum	30	57.7
CRC staging
Stage II	17	32.7
Stage III	17	32.7
Stage IV	18	34.6
Histopathology characteristic
Adenocarcinoma	43	82.7
Mucinous adenocarcinoma	5	9.6
Signet ring cell carcinoma	4	7.7
Histopathology grading
High-grade	11	21.2
Moderate-grade	9	17.3
Low-grade	32	61.5

CRC, colorectal cancer.

For the MMP-9 variable, positive frequency was 44 (84.6%), and negative frequency was 8 (15.4%). Multivariate analysis assessed the strength of the relationship for each variable with expression of MMP-9. For histopathological results, Figure 1 shows MMP-9 expression.

Figure 1.

(A) Low matrix metalloproteinase-9 expression: pale brown colour with a total score of 1 (positive), and (B) moderate matrix metalloproteinase-9 expression: light brown colour with total score of 4 (positive), (C) High MMP-9 expression: dark brown color with total score of 9 (positive).

Table 2 shows a significant relationship between the primary tumour and expression of MMP-9 (P=0.016), seen with χ² analysis. The cancer stage was significantly related (P=0.001), showing that the higher the stage of cancer in grade IV, the higher the expression of MMP-9. The histopathological grading maintained a significant relationship (P=0.049).

Table 2.

Analysis of clinicopathological relationships with MMP-9 expression

	MMP-9
	Positive		Negative		Total
Variable	n	%	n	%	n	%	p
Age (years)
30–39	8	18.2	3	37.5	11	21.2	0.661*
40–49	11	25.0	1	12.5	12	23.1
50–59	15	34.1	3	37.5	18	34.6
60–69	6	13.6	1	12.5	7	13.5
70–79	4	9.1	0	0.0	4	7.7
Sex
Mele	31	70.5	5	62.5	36	69.2	0.654*
Femele	13	29.5	3	37.5	16	30.8
Primary tumour location
Right colon	5	11.4	3	37.5	8	15.4	0.016 *
Left colon	10	22.7	4	50.0	14	26.9
Rectum	29	65.9	1	12.5	30	57.7
Cancer staging
Stage II	10	22.7	7	87.5	17	32.7	0.001 *
Stage III	16	36.4	1	12.5	17	32.7
Stage IV	18	40.9	0	0.0	18	34.6
Histopathology characteristic
Adenocarcinoma	36	81.8	7	87.5	43	82.7	0.543*
Mucinous adenocarcinoma	5	11.4	0	0.0	5	9.6
Signet ring cell carcinoma	3	6.8	1	12.5	4	7.7
Histopathology grading
High-grade	11	25.0	0	0.0	11	21.2	0.049 *
Moderate-grade	9	20.5	0	0.0	9	17.3
Low-grade	24	54.5	8	100.0	32	61.5

Bold values indicate significance (P<0.05).

MMp-9, matrix metalloproteinase-9.

^*χ² test.

Discussion

This study showed a relationship of MMP-9 expression to clinicopathology in CRC patients. Results show that the 50–59-year age group was the most common (34.6%), but saw no relationship with MMP-9 expression. Otero-Esteves et al.⁸ found the most common age group was younger than or equal to 49 years, followed by older than or equal to 60 years, with the least common of 50–59 years. This study showed a significant relationship between age group and MMP-9 expression (P<0.001), but the 50–59-year age group showed the lowest MMP-9 expression among three groups. Yang et al.⁹ found the group older than or equal to 58 years was more common than the one less than 58 years (36 vs. 32 patients). The group older than or equal to 58 years was more positive for MMP-9 than the one less than 58 years (26 vs. 21 positive cases of MMP-9), but no association existed between age and MMP-9 expression (P=0.557).

Many studies examined the relationship between age and progression of cancer cells: age affected development of tumour cells due to genetic mutations, genomic instability, and epigenetic alterations, prone to occur with increased age. Evidence from epidemiological studies and radiation-induced cancer models suggests that radiation exposure in midlife may increase risk more than exposure at other ages^10,11. Given this, several factors that affect men relate to CRC: sedentary lifestyle and obesity, alcohol consumption, and smoking. Smoking tobacco increases the chance of CRC by up to 10.8% with its high content of carcinogens such as nicotine, whose metabolites can reach the intestine and produce polyps^12,13. The sample of Yang et al.⁹ was dominated by men (44 vs. 24 women), but subsequent analysis found no relationship with MMP-9 expression (P=0.383): this study had 58.9% more women than men⁸.

The majority of primary tumour sites were in the rectum, with a significant link between primary tumour and the MMP-9 biomarker. Huang et al.¹⁴ found a majority of tumours originating from the colon (61 patients; 54.46%) with only 51 patients (45.54%) with tumours originating from the rectum. This study compared eight biomarkers, but only that of the MMP-7 biomarker showed a significant association between primary tumour site and biomarker levels (P=0.035).

The dominant microbiome is different between left and right locations of the primary tumour; the right was dominated by Prevotella, Pyramidobacter, Selenomonas, and Peptostreptococcus, whereas the left is dominated by Fusobacterium, Escherichia, Shigella, Leptotrichia, and Helicobacter pylori. This illustrates the difference in gene expression in the colonic mucosa. In some clusters, this was related to expression of pro-inflammatory genes that can develop into CRC. On the right, mutations in the KRAS, BRAF, and TGFβR2 genes are often found. On the left, there are more mutations of TP53, APC, and KRAS with chromosomal aberrations such as 8p and 17p arm deletions (in the TP53 gene mutation) plus aneuploidy. Compared to the right, left-side tumours have chromosomal instability in ~75% of cases¹⁵. Yet, a meta-analysis showed that right-sided CRC with mutations had a worse prognosis^16,17.

Moreover, the higher the cancer stage, the higher the expression of MMP-9. Huang et al.¹⁴ showed the highest prevalence of the stage III group of 37 patients (33.04%), followed by stage IV (36; 32.14%), stage II (34; 30.36%), and stage I (5; 4.46%). Subsequent analysis of MMP-9 expression and other biomarkers showed no significant relationship between stage and MMP-9 expression, as levels were higher in the stage I–II group (312.98 ng/ml) vs. the stage III–IV group (272.78 ng/ml). In accord, Sato et al. found more patients in the late stages: 33 in stage III–IV compared to 22 in stage I–II, but there was no significant relationship. This study assessed the clinicopathological association with LGR5 expression as a biomarker for stem cell cancer, which in high levels may lead to a poor prognosis, yet this is still controversial¹⁸.

Our findings demonstrate the capacity to quantify MMP-9 overexpression in colorectal cancer. This may enable invasion of malignant cells or the development of new blood vessels. MMP-9 expression is related to aggressive carcinomas. The stroma surrounding the tumour was shown to overexpress MMP-9¹⁹, which may be a response to other mediators that trigger MMP expression. An important mechanism by which proteolysis and subsequent invasion may occur is by recruiting normal stromal cells surrounding tumour cells to generate MMPs. Swallow et al.²⁰ examined the capacity of colorectal cancer cell lines with and without potential for metastasis, prompting monocytes to produce MMP-9 in this in-vitro research. MMP-9 activity was evaluated while aforementioned cell lines were cocultured with monocytes, but kept apart by a membrane. The cells of colorectal cancer did not generate either MMP. This research shows the ability of colorectal cancer cell lines with metastatic potential to activate stromal monocyte MMP-9. In-vitro data support that cancer cells cause stromal cells to degrade the extracellular matrix and basement membrane via a paracrine-type effect for specific MMPs. Others claim that compared to nonmetastatic basal cell carcinomas, MMP-9 was more expressed in stromata surrounding squamous cell carcinomas of the epidermis²¹. Although our data were not analyzed in cells from which the MMP mRNA was removed or localized, these studies confirm our findings that MMP-9 expression is elevated in colorectal carcinomas.

One extra mechanism seen in MMPs was found to enhance the local infiltration that encourage angiogenesis. Local basement cell membrane deterioration must occur for endothelial and tumour cells to infiltrate normal tissue. In this process, MMPs play a function supported by two recent studies. In rodents lacking MMP-9, Itoh et al.²² examined the ability of cancer cells to invade. They demonstrated how invasive tumour properties implant in MMP-9-deficient mice, despite a reduction, indicating that MMP-9 is crucial for invasion with normal stromal cells to aid invasion by upregulation and activation of MMP-9. An investigation into how endostatin affects tumour invasion lends credence to this view²³; it significantly decreased MMP-9 activation and expression, reducing invasive properties of both tumour and normal endothelial cells. MMP-9 and other MMPs may collectively play a role in angiogenesis, although its inhibitors may have the means to prevent this.

Fakhri and Agus²⁴ found positive MMP-9 expression in 28 of 35 samples, in the group with a T1 and T2 invasion depth of 35.7% (10 samples) and the group with a T3 and T4 invasion depth of 64.3% (18 samples). This showed that the deeper the invasion of CRC cells, the greater the percentage of MMP-9 expression, although the relationship was not statistically significant (P>0.05). Bendardaf et al.²⁵ found a relationship between MMP-9 expression and metastasis in CRC with a total of 359 research samples, 70% with positive MMP-9 expression and the presence of metastases, with the remaining 30% of samples without metastases. Shinta et al.²⁶, with 50 samples showed positive MMP-9 expression. The authors also showed a significant correlation between tumour invasion depth and MMP-9 expression (P=0.002).

MMP-9 degrades the extracellular matrix, other matrix metalloproteinases, and plasminogen, and releases and activates growth factors. This has a role in metastasis, angiogenesis, and rapid tumour growth, and the progression of CRC. The inhibition of MMP took place by tissue inhibitors of metalloproteinases (TIMP), divided in four groups: TIMP-1, -2, -3, and -4. The imbalance between MMP activation and inhibition has a major role in cancer pathophysiology²⁷.

Our study showed that the most common histopathological type was adenocarcinoma (43 people; 82.7%). The most common histopathological grading was low (32 people; 61.5%); 11 patients had high-grade tumours (21.2%), and a significant correlation with MMP-9 expression (P = 0.049). Otero-Estévez et al.⁸ found that the histological type is divided into tubular and tubulovillous/villous. The sample was dominated by the tubular type, with higher MMP-9 expression in the tubulovillous/villous group (mean 780±420 ng/ml) vs. the tubular group. This finding is sufficient to show a difference in the mean of MMP-9, highest in the advanced neoplasia group, with an average of 712 ng/ml, vs. non-advanced adenomas (668 ng/ml) and non-neoplasias (642 ng/ml). Advanced neoplasias were divided into advanced adenomas and cancer. The advanced adenomas had the highest MMP-9 average, at 722 ng/ml, with a significant relationship (P=0.029). The histopathological grading was assessed as well—differentiated/low-grade, moderately differentiated/moderate-grade, and poorly or undifferentiated/high-grade. In the study of Yang et al.⁹, the sample was dominated by low-grade tumours (32 patients, 23 positive for MMP-9), followed by moderate-grade (19 patients, 14 positive for MMP-9) and high-grade tumours (17 patients, 10 positive for MMP-9), but no significant relationship. In contrast, Huang et al.¹⁴ showed that it was dominated by moderate-grade tumours (98 patients; 87.5%), followed by high-grade (12 patients; 10.71%) and low-grade tumours (2 patients, 1.79%). Analysis of MMP-9 expression saw that the combined low–moderate category had a higher mean MMP-9 expression of 286.81 ng/ml than the high-grade group at 267.71 ng/ml, without a correlation test. Undifferentiated high-grade should have high MMP-9 expression, but we found the opposite. This could be due to the abnormal distribution of the data.

Ultimately, this study showed a relationship between location of the primary tumour, cancer stage, and histopathological grading. The tumour’s location was more proximal; the higher the biomarker MMP-9, the higher the stage of cancer. Stage IV had higher or positive MMP-9, although with higher grading or which was poorly/undifferentiated, the MMP-9 looked negative. This result is still controversial, and more testing is needed with non-normal data distribution.

Most CRC patients come to health facilities at an advanced stage with metastases to various organs, so a noninvasive screening tool is needed for early diagnosis to assist colonoscopy assessment, considered an invasive procedure. MMP-9 is expected to be used for screening with its involvement in angiogenesis, invasion, and tumour metastasis in CRC, but it has a sensitivity of ~82% and a specificity of 46%¹⁴. In CRC, MMP-9 plays a role in tumour development and progression. The prognostic role of MMP-9 expression in CRC patients found that those with higher MMP-9 expression had poorer survival. MMP-9 plays an important role in colon cancer invasion and metastasis, and may be a useful indicator for clinical assessment of tumour biologic behaviour and prognosis of CRC. Further studies to compare other biomarkers are needed to determine optimal screening for both diagnosis and treatment success in CRC.

Conclusions

A significant relationship exists for primary tumour location, cancer staging, and histopathology grading with MMP-9 expression in CRC in Makassar. This study showed no correlation of other clinicopathological variables: age, sex, and histopathology characteristics with MMP-9 expression in CRC patients: this may be a useful indicator for clinical assessment of tumour biologic behaviour and prognosis of CRC patients. This can also be used for further research: any investigations must be done with a larger number of samples and continued research time.

Ethical approval

All procedure for human experiment has been approved by Health Research Ethical Committee of the Faculty of Medicine, Universitas Hasanuddin, Number: 90/UN4.6.4.5.31/PP36 /2022.

Consent

Written informed consent was obtained from all the subjects for participation and publication of the data. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

Sources of funding

None.

Author contribution

Mudatsir: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, validation, visualization, writing—original draft, writing—review and editing. I.L.: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, writing—original draft, writing—review and editing. J.A.U.: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, validation, visualization, writing—original draft, writing—eview and editing. J.H.: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, software, validation, visualization, writing—original draft, writing—review and editing. Warsinggih: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, software, validation, visualization, writing—original draft, writing—review and editing. R.E.L.: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, visualization, writing—original draft, writing—review and editing. Mappincara: conceptualization, data curation, investigation, resources, supervision, visualization, writing—original draft, writing—review and editing. S.S.: conceptualization, data curation, formal analysis, investigation, resources, software, supervision, visualization, writing—original draft, writing—review and editing. M.I.K.: conceptualization, data curation, formal analysis, investigation, methodology, resources, software, supervision, validation, visualization, writing—original draft, writing—review and editing. E.S.: conceptualization, data curation, formal analysis, investigation, methodology, resources, software, supervision, validation, visualization, writing—original draft, writing—review and editing. A.A.: conceptualization, data curation, formal analysis, investigation, resources, software, validation, visualization, writing—review and editing. M.F.: data curation, formal analysis, investigation, methodology, project administration, resources, software, validation, visualization, writing—review and editing.

Conflicts of interest disclosure

The authors declare that they have no conflicts of interest.

Research registration unique identifying number (UIN)

This research has been registered with the Research Registry number: 8274. https://researchregistry.knack.com/researchregistry#home/registrationdetails/6317c16b38f8ac0024891392/

Guarantor

Ibrahim Labeda.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author [Ibrahim Labeda]. The data are not publicly available due to [restrictions e.g. their containing information that could compromise the privacy of research participants].