Medicine OPEN
Exploring the causal relationship between matrix metalloproteinases and adrenocortical carcinoma A Mendelian randomization study
Zheng Huang, MDa,*(D, Qi Zhang, PhDb
Abstract
The precise relationship between adrenocortical carcinoma (ACC) and matrix metalloproteinases (MMP) remains unclear. In this 2-sample Mendelian randomization (MR) study, exposure data regarding serum MMP levels were obtained from genome-wide association studies that included 21,758 individuals across 13 cohorts of European ancestry. The outcome data related to ACC were sourced from the FinnGen research project. The inverse-variance weighting method was employed as the primary analytical approach and was further verified using a variety of statistical techniques, including MR-Egger, weighted median, weighted mode, Bayesian weighted Mendelian randomization, constrained maximum likelihood, contamination mixture method and debiased inverse-variance weighted method. A Steiger directionality test was applied to avoid a reverse causation association. To assess pleiotropy and heterogeneity, we conducted the MR-Egger intercept test, Cochran Q test, and leave-one-out analyses. Using the IVW and Bonferroni-corrected approach, we observed that elevated serum levels of MMP-12 were associated with a reduced risk of ACC, with an odds ratio of 0.699 (95% confidence interval: 0.544-0.895; P = . 005). Sensitivity analyses indicated the absence of significant heterogeneity and pleiotropy within our study. Additionally, the Steiger directionality test did not detect a significant reverse causation effect. The causal relationship between MMP-12 levels and ACC could have significant implications for the diagnostic and therapeutic approaches employed in the management of ACC.
Abbreviations: ACC = adrenocortical carcinoma, CI = confidence interval, GWAS = genome-wide association study, IV = instrumental variable, MMP = matrix metalloproteinase, MR = Mendelian randomization, OR = odds ratio, SNP = single nucleotide polymorphism.
Keywords: adrenocortical carcinoma, causality association, matrix metalloproteinase, Mendelian randomization
1. Introduction
Adrenocortical carcinoma (ACC) is a relatively uncommon but highly cancerous tumor arising from the adrenal cortex, with a prevalence rate of nearly 0.72 cases per million annu- ally. This cancer exhibits a female predominance and has a bimodal age distribution, with peaks in children and individ- uals aged 40 to 50 years.[1-3] ACCs are generally associated with poor prognoses, characterized by a 5-year survival rate below 35% and a tumor recurrence rate ranging from 70% to 80%, although early-stage cases tend to have slightly better outcomes.[4] Approximately 1% to 11% of ACCs are detected incidentally through radiographic imaging.[5] Alarmingly, 25% to 30% of patients with ACC already have metastasis at
the point of diagnosis, contributing to their reduced survival prospects.[6]
The matrix metalloproteinase (MMP) family comprises 28 distinct enzymes that share similar structural, regulatory, and functional characteristics.[7] Early research primarily focused on their crucial role in remodeling the extracellular matrix (ECM), as these enzymes can degrade a collection of ECM components, including collagen, proteoglycans, and other non- ECM proteins.[7] Beyond ECM remodeling, MMPs regulate the immune microenvironment via tumor-infiltrating immune cells[8] and growth factor liberation.[9] In human cancers, ele- vated expression and activity of MMPs are frequently associ- ated with advanced tumor stages, increased invasiveness and metastasis, and poorer survival outcomes.[10] For example, in
The authors have completed the STROBE-MR reporting checklist. The authors have no funding and conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are publicly available.
The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).
Supplemental Digital Content is available for this article.
a Department of Urology, Shaoxing Second Hospital, Second Affiliated Hospital of Shaoxing University, Shaoxing, China, b Department of Urology, Zhejiang Province people’s Hospital, Zhejiang, China.
* Correspondence: Zheng Huang, Department of Urology, Shaoxing Second Hospital, Second affiliated hospital of Shaoxing University, No. 123 Yanan Road, Shaoxing 312000, China (e-mail: healthy365@usx.edu.cn).
Copyright @ 2025 the Author(s). Published by Wolters Kluwer Health, Inc. This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial License 4.0 (CCBY-NC), where it is permissible to download, share, remix, transform, and buildup the work provided it is properly cited. The work cannot be used commercially without permission from the journal.
How to cite this article: Huang Z, Zhang Q. Exploring the causal relationship between matrix metalloproteinases and adrenocortical carcinoma: A Mendelian randomization study. Medicine 2025;104:50(e46475).
Received: 21 July 2025 / Received in final form: 22 October 2025 / Accepted: 23 October 2025
http://dx.doi.org/10.1097/MD.0000000000046475
cancers such as breast, prostate, bladder, and gastric carcino- mas, higher levels of MMPs are typically associated with a worse prognosis.[11-14]
Mendelian randomization (MR) analyses were conducted to assess potential causal relationships and the strength of the asso- ciation between MMPs and ACC. This genetic epidemiological approach utilizes instrumental variable techniques to infer causal- ity from observational datasets, particularly when examining the connections between variable risk factors and clinical endpoints, such as disease occurrence.[15] Relative to conventional multivar- iate regression analyses, the MR strategy demonstrated superior robustness against 3 major limitations: measurement inaccura- cies, confounding variables, and bidirectional causation. This methodological advantage substantially strengthen the validity of the causal conclusions of our oncological study.[15] Consequently, they have gained widespread application in recent years. This study adhered to the STROBE-MR reporting Checklist.
To the best of our knowledge, relatively few studies have investi- gated the relationship between MMP levels and ACC. By clarifying the role of MMPs in ACC, we aimed to identify novel diagnostic and therapeutic targets that could ultimately improve patient out- comes and reduce the burden of this life-threatening condition.
2. Materials and methods
2.1. Study design
The causal relationships between MMP-1/2/3/7/9/10/12 and ACC were examined by MR analysis. Genetic associations for these proteolytic enzymes were initially extracted from genome- wide association study (GWAS) datasets, with stringent selec- tion of single nucleotide polymorphisms (SNPs) demonstrating a strong predictive value for MMP concentrations. The corre- sponding genetic data for ACC outcomes were subsequently retrieved from the available repositories. Our analytical frame- work incorporates eight complementary MR approaches to rig- orously evaluate the causal effects. These methods utilize SNPs that exhibit robust associations with both MMP expression profiles and ACC pathogenesis as valid instrumental variables (IVs). Figure 1 shows the flowchart of the study design.
2.2. The GWAS datasets
GWAS data for MMP-2/3/9 were obtained from a systematic pooled analysis of 1323 European-descendant samples in 2020,
which included a diverse spectrum of 18221903 SNPs.[16] MMP- 1/7/10/12 data were derived from a systematic pooled analysis of 21,758 European-descendant samples in 2020, comprising a large spectrum of 13100092 SNPs.[17] The genetic associa- tion results for adrenocortical carcinoma cases were obtained from the FinnGen Consortium Release 12 genomic database.[18] This analysis included 971 cases of adrenocortical carcinoma and 1,74,006 controls. Complete summary statistics for asso- ciation testing are publicly available in the MRC-IEU GWAS database (accessed URL: https://gwas.mrcieu.ac.uk). The details of the GWAS data for MMPs and ACC are provided in Table S1, Supplemental Digital Content, https://links.lww.com/MD/ Q939.
2.3. Instrumental variables (IVs) selection
Using R software, to integrate SNPs exhibiting genome-wide significance, as identified in the GWAS database; we imple- mented stringent quality controls to mitigate linkage disequi- librium interference, selecting SNPs based on the following thresholds: 12 < 0.001 for linkage disequilibrium, 10,000kb for physical distance, and P < 5 × 10-6 for statistical signifi- cance (relaxed threshold to capture potential MMP regula- tors); and to rule out confounding associations with renal cell carcinoma, we assessed secondary phenotypes for each SNP. The strength of the genetic instruments was assessed using F-statistics, calculated as the squared ratio of the effect size (B) to the standard error for each exposure-associated vari- ant (F = [ß/SE]2). SNPs with F < 10 were excluded to mitigate weak instrument bias[19]; and harmonization of effect alleles and beta coefficients was performed to ensure consistency across datasets.
2.4. MR analysis
To robustly assess potential causal relationships between the selected MMP isoforms and ACC, we employed -eight complementary MR techniques. The IVW approach, which is considered the primary MR method, operates under the assumption that genetic IVs are valid and weight estimates are based on their inverse variance. This widely adopted framework provides high statistical power when IV assumptions hold true. To corroborate these findings, we also applied weighted median, weighted mode, MR-Egger, Bayesian weighted Mendelian ran- domization, constrained maximum likelihood, contamination
X
Assumption 2
confounder
Assumption 1
IVs
Exposure MMP-1/2/3/7/9/10/12
Outcome ACC
X
Assumption 3
· Associated with exposure ( P<5x10-6)
· IVW( p< 0.05)
· LD clumping(r2=0.001,a window size of 1,000 kb)
· MR-Egger, Weighted Median, Weighted mode, cML-MA, ConMix, BWMR, and DIVW.
· Removement of IVs associated with confounders
· Bonferroni-corrected test p< 0.0071 (0.05/7)
· Steiger_test_direction
· Assessment of the strength of IVs (F- statistics>10)
· Heterogeneity Assessment ( Cochran’s Q test, p<0.05)
· Horizontal pleiotropy Assessment (MR-Egger intercept, p<0.05)
· Leave-one-out analysis
mixture method and debiased inverse-variance weighted method approaches. To account for potential pleiotropic effects, both MR-Egger and IVW integrate an intercept term into their weighted regression models. The intercept reflects directional pleiotropy, whereas the slope estimates the magnitude of the causal effect. Analyses were performed using R v4.3.0 with TwoSampleMR v0.5.6, adopting a conventional 2-tailed signifi- cance threshold (a = 0.05).
To avoid reverse causality, the Steiger directionality test was employed to identify SNPs that indicate causality in the oppo- site direction. The F-statistic was calculated for each SNP to reduce the risk of weak instrument bias. The variance explained by each SNP is represented by R2, which was calculated as R2= 2 x (1 - EAF) × 2 × EAF; F = R2/(1 - R2) x (N-2). The results of our study were visualized using a scatter plot and leave-one-out plot. Considering the multiple analyses per- formed, a Bonferroni-corrected P-value < . 0071 (.05/7) was considered to indicate statistical significance. A P-value within the range of .007 to .050 was considered suggestive evidence.
2.5. Sensitivity analysis
We primarily utilized a heterogeneity test to assess the differ- ences among the IVs. Significant heterogeneity (P <. 05) indi- cated substantial variability among the IVs. For analyses that demonstrated significant heterogeneity, random-effects IVW models were preferred to account for between-SNP variance. The fixed-effects, IVW model was preserved in the presence of homogeneous effects. The MR-Egger intercept term served as our primary pleiotropy diagnostic with values close to 0, supporting the absence of significant horizontal pleiotropy. Additionally, the MR-PRESSO test was employed to conduct a global test of heterogeneity, identify potential outliers in each SNP, and derive a corrected association result after removing these outliers. Leave-one-out validation was implemented, in which we sequentially excluded each genetic variant and re- derived the aggregated effect estimates from the retained instru- ments. This approach quantifies the influence of each variant on overall causal association. Removal of any single SNP did not significantly alter the MR results, indicating that our analysis was stable and not excessively dependent on any single SNP.
3. Results
3.1. Instrumental Variable
Seven SNPs were identified as being associated with both MMP-3 and ACC, 20 SNPs with MMP-10 and ACC, and 31 SNPs with MMP-12 and ACC (Tables S2-S4, Supplemental Digital Content, https://links.lww.com/MD/Q939). All SNPs included in the analysis demonstrated F-statistics exceeding the conven- tional threshold of 10, effectively mitigating weak IV bias and ensuring robust causal inference (Tables S2-S4, Supplemental Digital Content, https://links.lww.com/MD/Q939).
3.2. MR analysis between MMP-3/10/12 and ACC
The primary findings of this MR study were conducted using the IVW method, as depicted in Figure 2 and Table S5, Supplemental Digital Content, https://links.lww.com/MD/Q939, revealed several significant associations. The P-values of MMP-1/2/7/9 were not <. 05, so they were excluded (Table S5, Supplemental Digital Content, https://links.lww.com/MD/Q939). Prior to the application of the Bonferroni correction, three pairs of traits demonstrated statistically significant differences. Specifically, serum levels of MMP-3 were linked to an elevated risk of ACC (odds ratio [OR]: 1.361; 95% confidence interval [CI]: 1.028-1.801; P = . 031). Serum MMP-10 levels were associated with a higher risk of ACC (OR: 1.621, 95% CI: 1.075-2.446;
P = . 021). Additionally, serum MMP-12 levels were found to be associated with a decreased risk of ACC (OR: 0.698; 95% CI: 0.544-0.895; P = . 005). However, following the implementa- tion of the Bonferroni correction, the only significant associa- tion that remained was between serum MMP-12 levels and the decreased risk of ACC. The scatter plots illustrating the signifi- cant MR results prior to the Bonferroni correction are presented in Figure S1, Supplemental Digital Content, https://links.lww. com/MD/Q940, including IVW, weighted median, weighted mode, MR-Egger, Bayesian weighted Mendelian randomiza- tion, constrained maximum likelihood, contamination mixture, and debiased inverse-variance weighted method approaches. While some associations did not reach statistical significance, the direction of the results was consistent with that observed using the primary IVW method (Table S5, Supplemental Digital Content, https://links.lww.com/MD/Q939).
3.3. Sensitivity analyses
The presence of pleiotropy in this study was assessed utilizing the MR-Egger regression and MR-PRESSO global approach, and the analyses did not reveal any significant pleiotropic effects (P-value for the intercept exceeded .05 or for the PRESSO < 0.05). Additionally, the heterogeneity of the study was evaluated through the Cochran Q test, and no significant heterogeneity was identified in the analyses (Table S5, Supplemental Digital Content, https://links.lww.com/MD/Q939). The MR-Steiger directionality test consistently returned a “true” outcome across all tests, thereby indicating that reverse causal associations were not present. Sensitivity analysis employing the leave-one-out approach revealed that no individual SNP exerted a substan- tial influence on the outcome, confirming the stability of our MR estimates (Fig. S2, Supplemental Digital Content, https:// links.lww.com/MD/Q940 and Table S6, Supplemental Digital Content, https://links.lww.com/MD/Q939).
4. Discussion
This research utilized a 2-sample MR approach to explore the potential causal relationships between MMPs and ACC. Through forward and reverse MR analysis, our study revealed the potential effects of MMP-3/10/12 on ACC (P <. 05), but only MMP-12 was determined to possibly affect the risk of ACC after Bonferroni correction (P < . 0071).
Our analysis found a causal effect of MMP-12 on a lower risk of ACC. MMP-12, a metalloproteinase secreted by mac- rophages, has been identified as being aberrantly expressed in various types of tumors. It exerts significant influence on tumor progression and prognosis through multiple mechanisms.[20,21] In renal cell carcinoma and esophageal squamous cell carci- noma, patients with elevated MMP-12 levels exhibit a signifi- cantly worse prognosis compared to those with low MMP-12 levels.[22,23] In liver hepatocellular carcinoma, the upregulation of MMP-12 has been shown to promote tumor growth and pro- gression by enhancing angiogenesis, which is associated with a poorer prognosis for patients.[24] However, some studies have also suggested that MMP-12 may exert antitumor effects in colorectal cancer. Specifically, knocking out MMP-12 has been shown to lead to the accumulation of M2 macrophages, which predominantly exhibit pro-cancer effects, in the tumor micro- environment. This accumulation, in turn, promotes the growth of colorectal tumors.[25] In ovarian cancer, high levels of MMP- 12 mRNA have been associated with better overall survival.[26] Therefore, the role of MMP-12 in different types of malignan- cies may be completely opposite. Alternatively, it is conceivable that the function of MMP-12 is context-dependent. In the early stages of ACC, it may exert protective effects through tissue repair and inflammation. However, overexpression or deregu- lation of MMP-12 in more advanced or malignant stages could
| Exposure | No.of SNP | Method | OR(95% CI) | P |
|---|---|---|---|---|
| MMP-1 | 34 | IVW | 1.19 (0.87 to 1.64) | 0.283 |
| MR Egger | 0.95 (0.57 to 1.60) | 0.857 | ||
| Weighted median | 0.78 (0.49 to 1.24) | 0.287 | ||
| Weighted mode | 0.87 (0.54 to 1.38) | 0.546 | ||
| MMP-2 | 7 | IVW | 1.00 (0.72 to 1.40) - | 0.980 |
| MR Egger | 0.67 (0.27 to 1.69) | 0.438 | ||
| Weighted median | 0.96 (0.63 to 1.46) | 0.842 | ||
| Weighted mode | 0.83 (0.46 to 1.51) | 0.562 | ||
| MMP-3 | 7 | IVW | 1.36 (1.03 to 1.80) - | 0.031 |
| MR Egger | 1.99 (0.94 to 4.21) | 0.133 | ||
| Weighted median | 1.37 (0.97 to 1.94) | 0.072 | ||
| Weighted mode | 1.40 (0.99 to 2.00) | 0.108 | ||
| MMP-7 | 24 | IVW | 1.03 (0.67 to 1.59) | 0.903 |
| MR Egger | 1.12 (0.58 to 2.19) | 0.732 | ||
| Weighted median | 1.22 (0.68 to 2.20) | 0.507 | ||
| Weighted mode | 1.24 (0.69 to 2.25) | 0.481 | ||
| MMP-9 | 15 | IVW | 0.90 (0.76 to 1.07) H | 0.224 |
| MR Egger | 0.76 (0.55 to 1.05) | 0.117 | ||
| Weighted median | 0.85 (0.66 to 1.10) - | 0.225 | ||
| Weighted mode | 0.81 (0.61 to 1.07) | 0.160 | ||
| MMP-10 | 20 | IVW | 1.62 (1.07 to 2.45) | 0.021 |
| MR Egger | 1.16 (0.60 to 2.23) H | 0.668 | ||
| Weighted median | 1.53 (0.87 to 2.71) | 0.142 | ||
| Weighted mode | 1.49 (0.83 to 2.65) | 0.194 | ||
| MMP-12 | 31 | IVW | 0.70 (0.54 to 0.90) 1011 | 0.005 |
| MR Egger | 0.68 (0.48 to 0.97) 1 | 0.043 | ||
| Weighted median | 0.69 (0.51 to 0.93) | 0.015 | ||
| Weighted mode | 0.70 (0.52 to 0.95) 1 | 0.027 |
1
2
3 4
Figure 2. Forest plot of forward MR analysis between MMPs and ACC. ACC = adrenocortical carcinoma, IVW = inverse variance weighted, LD = linkage dis- equilibrium, MMP = matrix metalloproteinase, MR = Mendelian randomization, OR = odds ratio.
promote disease progression. Thus, the balance and regulation of MMP-12 activity within the adrenocortical tissue microen- vironment are probably vital in determining its overall impact on adrenocortical health and disease. Moreover, the dataset employed in the current study focused on blood MMP-12 lev- els, which may not accurately reflect the local MMP-12 levels within adrenocortical tissues. Nonetheless, additional research is required to clarify the exact role of MMP-12 in adrenocorti- cal carcinogenesis.
Our investigation revealed that MMP-3 and MMP-10 exerts a causal influence on the elevated risk of ACC. MMP-3 (stromelysin-1) and MMP-10 (stromelysin-2) are members of the stromelysin subfamily, which is subclasses of MMPs.[27] Active forms of MMP-3 and MMP-10 share 78% identity in amino acid sequence, so they have similar substrate activity, for exam- ple, ability to cleave ECM proteins such as aggrecan, fibronec- tin, and collagen types III and IV.[28] In the context of cancer, MMP-3 and MMP-10 often promotes malignancy by remodel- ing the tumor microenvironment and modulating signaling mol- ecules. Specifically, it degrades basement membranes to facilitate cancer cell invasion and activates growth factors that enhance tumor cell proliferation and metastatic progression.[29] In var- ious cancers, such as breast, oral, prostate, bladder, and lung
cancers, MMP-3 and MMP-10 are often overexpressed.[30-33] This overexpression is associated with more aggressive disease features and poorer patient outcomes. The results of this study found that MMP-3 and MMP-10 may promote the develop- ment of adrenocortical carcinoma, which may be related to the action of MMP-3 and MMP-10 on transforming growth factor, thereby stimulating cancer cell proliferation and tumor progression. Nevertheless, further studies are required to clarify the exact function of MMP-3 and MMP-10 in the development of ACC.
The present study had several strengths. First, it established the first MR analysis evidence for causal relationships between circulating MMP-3/10/12 levels and ACC risk. Rigorous quality assurance measures and analytical methodologies were system- atically employed. Simultaneously, different sensitivity examina- tions were employed to confirm the robustness of the outcomes. Finally, the use of MR methods reduced the impact of confound- ing factors to a minimum.
This study has some limitations that need to be considered. First, the study sample was limited to individuals of European descent. Moving forward, it is crucial to incorporate samples from Asia and Africa to build on our findings. Moreover, we adopted more lenient criteria (for example, considering SNPs
with P-values < 5 x 10-6). This approach might elevate the false-positive rate, but it allows for a more comprehensive and in-depth assessment of the relationship between MMPs and ACC. Last, MR analysis continues to face challenges, including biases like population stratification. Therefore, further natural experiments and prospective clinical investigations focusing on MMPs in ACC remain essential. Notably, the insights derived from our study can function as a vital element within a trian- gulation framework, thereby aiding in the establishment of a definitive causal relationship between MMPs and ACC.
5. Conclusions
In summary, our investigation uncovered a significant link between abnormal levels of MMP-12 and reduced likelihood of ACC occurrence. This finding have important clinical impli- cations. Identifying specific MMPs as pathogenic factors for ACC provides potential targets for therapeutic intervention. For example, enhancing the activity or expression of MMP-12 may reduce the risk of ACC development. Moreover, MMPs can serve as biomarkers for disease prediction, diagnosis, and prog- nosis. However, further research is needed to confirm these asso- ciations and elucidate the underlying biological mechanisms.
Acknowledgments
The authors extend their sincere thanks to all the investigators for their generous sharing of the valuable data.
Author contributions
Conceptualization: Zheng Huang.
Data curation: Zheng Huang, Qi Zhang.
Formal analysis: Zheng Huang.
Investigation: Zheng Huang.
Methodology: Zheng Huang, Qi Zhang.
Writing - original draft: Zheng Huang.
Writing - review & editing: Zheng Huang, Qi Zhang.
References
[1] Peng H, Chen Q, Ye L, Wang W.A senescence-associated gene signature for prognostic prediction and therapeutic targeting in adrenocortical carcinoma. Biomedicines. 2025;13:894-914.
[2] Shariq OA, Mckenzie TJ. Adrenocortical carcinoma: current state of the art, ongoing controvers ies, and future directions in diagnosis and treatment. Ther Adv Chronic Dis. 2021;12:1-21.
[3] Kuhlen M, Schmutz M, Kunstreich M, Redlich A, Claus R. Targeting pediatric adrenocortical carcinoma: molecular insights and emerging therapeutic strategies. Cancer Treat Rev. 2025;136:102942-53.
[4] Tella SH, Kommalapati A, Yaturu S, Kebebew E. Predictors of survival in adrenocortical carcinoma: an analysis from the national cancer database. J Clin Endocrinol Metab. 2018;103:3566-73.
[5] Fassnacht M, Arlt W, Bancos I, et al. Management of adrenal inci- dentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2016;175:G1-G34.
[6] Sarrafan-Chaharsoughi Z, Yazdian Anari P, Malayeri AA, Naraev B, Del Rivero J. Update on adrenocortical carcinoma. Urol Clin North Am. 2025;52:275-86.
[7] Berg G, Barchuk M, Miksztowicz V. Behavior of metalloproteinases in adipose tissue, liver and arterial wall: an update of extracellular matrix remodeling. Cells. 2019;8:158-71.
[8] Fang F, Wu T, Wang M, et al. Matrix metalloproteinase-9 is associ- ated with tumor microenvironment remodeling of bladder cancer. Biol Direct. 2025;20:8-20.
[9] Hu C, Zuo Y, Ren Q, et al. Matrix metalloproteinase-10 protects against acute kidney injury by au gmenting epidermal growth factor receptor signaling. Cell Death Dis. 2021;12:70-83.
[10] Cox TR. The matrix in cancer. Nat Rev Cancer. 2021;21:217-38.
[11] Fouad FA, Khali MA, Moaz I, et al. Prognostic impact of matrix metal- loproteinase 2 (MMP-2) and matrix met alloproteinase 9 (MMP-9) in Egyptian breast cancer patients. Int J Immunopathol Pharmacol. 2024;38:1-10.
[12] Zhu S, He J, Yin L, et al. Matrix metalloproteinases targeting in pros- tate cancer. Urol Oncol. 2024;42:275-87.
[13] Zhang X, Hong B, Li H, et al. Basement membrane-related MMP14 predicts poor prognosis and response t o immunotherapy in bladder cancer. BMC Cancer. 2024;24:746-59.
[14] Zhang C, Huang DL, Zhou K, et al. Human blood metabolites and gas- tric cancer: a Mendelian randomization analysis. BMC Gastroenterol. 2024;24:478-92.
[15] Dobrijevic E, van Zwieten A, Kiryluk K, Grant AJ, Wong G, Teixeira- Pinto A. Mendelian randomization for nephrologists. Kidney Int. 2023;104:1113-23.
[16] Gilly A, Park Y-C, Png G, et al. Whole-genome sequencing analysis of the cardiometabolic proteome. Nat Commun. 2020;11:6336-45.
[17] Folkersen L, Gustafsson S, Wang Q, et al. Genomic and drug target evaluation of 90 cardiovascular proteins in 30,931 individuals. Nat Metab. 2020;2:1135-48.
[18] Kurki MI, Karjalainen J, Palta P, et al; FinnGen. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613:508-18.
[19] Liu W, Wang Y, Zhou Y. Exploring the causal relationship between omega-3 and omega-6 fatty acids and kidney cancer: a Mendelian ran- domization study. Transl Cancer Res. 2024;13:1685-94.
[20] Banerjee S, Lo W-C, Majumder P, et al. Multiple roles for basement membrane proteins in cancer progression and EMT. Eur J Cell Biol. 2022;101:151220.
[21] Marcos-Jubilar M, Orbe J, Roncal C, et al. Association of SDF1 and MMP12 with atherosclerosis and inflammation: clinical and experi- mental study. Life (Basel). 2021;11:414.
[22] Beres B, Yusenko M, Peterfi L, Kovacs G, Banyai D. Matrix metallopro- teinase 12 is an independent prognostic factor predicting postoperative relapse of conventional renal cell carcinoma - a short report. Cell Oncol (Dordr). 2022;45:193-8.
[23] Han F, Zhang S, Zhang L, Hao Q. The overexpression and predic- tive significance of MMP-12 in esophageal squamous cell carcinoma. Pathol Res Pract. 2017;213:1519-22.
[24] Guo ZY, Jiang LP. Matrix metalloproteinase 12 (MMP12) as an adverse prognostic biomarker of vascular invasion in hepatic cell carci- noma. Eur Rev Med Pharmacol Sci. 2022;26:2238-49.
[25] Yang M, Zhang X, Liu Q, et al. Knocking out matrix metallopro- teinase 12 causes the accumulation of M2 macrophages in intestinal tumor microenvironment of mice. Cancer Immunol Immunother. 2020;69:1409-21.
[26] Zeng L, Qian J, Zhu F, Wu F, Zhao H, Zhu H. The prognostic val- ues of matrix metalloproteinases in ovarian cancer. J Int Med Res. 2020;48:300060519825983.
[27] Jiang J, Wu Q, Rajasekaran S, Wu R. MMP3 at the crossroads: linking molecular pathways to disease diagnosis and therapy. Pharmacol Res. 2025;216:107750.
[28] Niland S, Riscanevo AX, Eble JA. Matrix metalloproteinases shape the tumor microenvironment in cancer progression. Int J Mol Sci. 2021;23:146.
[29] Suhaimi SA, Chan SC, Rosli R. Matrix metallopeptidase 3 polymor- phisms: emerging genetic markers in human breast cancer metastasis. J Breast Cancer. 2020;23:1-9.
[30] Ewa G, Ławicki P, Grabowska H, et al. Plasma concentrations of matri- lysins (MMP-7, MMP-26) and stromelysins (MMP-3, MMP-10) as diagnostic biomarkers in high-grade serous ovarian cancer patients. Int J Mol Sci. 2025;26:5661.
[31] He L, Kang Q, Chan KI, Zhang Y, Zhong Z, Tan W. The immunomod- ulatory role of matrix metalloproteinases in colitis-associated cancer. Front Immunol. 2023;13:1093990.
[32] Kudelski J, Młynarczyk G, Gudowska-Sawczuk M, et al. Higher con- tent but not activity of stromelysin-2 (MMP-10) in comparison to stromelysin-1 (MMP-3) in human renal carcinoma. Int J Environ Res Public Health. 2022;19:12613.
[33] Kudelski J, Młynarczyk G, Gudowska-Sawczuk M, et al. Enhanced expression but decreased specific activity of matrix metallop roteinase 10 (MMP-10) in comparison with matrix metalloproteinase 3 (MMP- 3) in human urinary bladder carcinoma. J Clin Med. 2021;10:3683.