Insights from Tissue Inhibitor of Metalloproteinase-2 Genotypes to Decipher the Genetic Architecture of Childhood Acute Lymphocytic Leukemia Risk

Abstract

Background/Aim

Acute lymphoblastic leukemia (ALL) is the most common pediatric hematologic malignancy, particularly affecting children aged 2~5 years. Tissue inhibitor of metalloproteinase-2 (TIMP-2), a key regulator of MMP-2 activity, has been implicated in several cancers, yet its genetic role in childhood ALL remains unexplored.

Materials and Methods

This study investigated four TIMP-2 polymorphic genotypes, rs8179090, rs4789936, rs2009196, and rs7342880, in 266 Taiwanese children with ALL and 266 matched controls using polymerase chain reaction-restriction fragment length polymorphism methodology.

Results

TIMP-2 rs8179090 exhibited a significant association with ALL risk. Individuals with the CC genotype had a markedly increased risk [odds ratio (OR)=4.01, 95% confidence interval (CI)=1.46-11.04, p=0.0076], particularly under a recessive model (OR=3.79, 95%CI=1.39-10.36, p=0.0105). The C allele frequency was also elevated in cases (20.7%) versus controls (14.5%) (p=0.0100). Stratified analysis showed stronger risk association in children aged ≤3.5 years (CC genotype: OR=5.06, p=0.0084) and in boys (CC genotype: OR=5.53, p=0.0046). Moreover, CG+CC genotypes were associated with higher clinical risk classification (OR=2.25, p=0.0031) and shorter survival (<5 years) (OR=3.68, p=0.0003), though no correlation was found with immunophenotypic subtypes. No significant associations were identified for rs4789936, rs2009196, or rs7342880.

Conclusion

TIMP-2 rs8179090, particularly the CC genotype, may serve as a novel biomarker for childhood ALL susceptibility and prognosis, especially in younger and male patients. Genotyping of this polymorphism could support early risk assessment and personalized clinical management in childhood ALL.

Introduction

Acute lymphoblastic leukemia (ALL) represents the most frequently diagnosed hematologic malignancy among pediatric populations, marked by uncontrolled proliferation of clonal lymphoid progenitor cells (1). This leukemia subtype primarily emerges in children between the ages of two and five, rendering it the leading form of blood cancer in early childhood (2,3). The disease may originate from lymphoid precursors of either T- or B-lineage, with B-lineage precursor variants accounting for more than 70~85% of pediatric instances (4,5). Although numerous studies have been conducted, the underlying mechanisms and initiating causes of childhood ALL are not yet fully delineated; current consensus suggests a multifactorial origin involving both inherited susceptibilities and environmental exposures (6-8). Presently, one of the central objectives in translational hematologic oncology is the identification of reliable and clinically applicable molecular indicators for childhood ALL (9), with a growing emphasis on research efforts within Taiwan (10-12).

Tissue inhibitor of metalloproteinase-2 (TIMP-2) is a key regulatory molecule in tumor biology, primarily through its modulation of matrix metalloproteinase (MMP) activity, particularly MMP-2 (13). While the TIMP-2/MMP axis has been extensively studied in solid tumors, its functional dynamics in hematologic malignancies remain incompletely understood. Intriguingly, over-expression of TIMP-2 has been shown to paradoxically enhance cellular invasiveness and promote proMMP-2 activation in SHI-1 cells, a human acute monocytic leukemia line known for its aggressive phenotype (14). Located on chromosome 17q25, the TIMP-2 gene harbors polymorphisms that may affect its expression or function, thereby disrupting the regulatory equilibrium between TIMP-2 and MMP-2. Such dysregulation could have profound implications for leukemogenesis and disease progression (15). A growing body of literature supports the association between TIMP-2 genetic variants and susceptibility to a broad spectrum of malignancies, including neoplasms of the head and neck, prostate, urinary bladder, stomach, gallbladder, and breast (16-21). Taken together, these observations highlight the potential of TIMP-2 polymorphisms as candidate biomarkers for risk stratification and pathophysiological insight in childhood ALL.

To the best of our knowledge, no prior study has examined the association between TIMP-2 polymorphisms and susceptibility to childhood ALL in any population worldwide. Therefore, this study aimed to evaluate the potential contribution of four TIMP-2 single nucleotide polymorphisms (SNPs), rs8179090, rs4789936, rs2009196, and rs7342880, as illustrated in Figure 1, to the risk of childhood ALL in a Taiwanese cohort. Furthermore, we aimed to provide preliminary evidence on the combined effects of these genetic variants with demographic and clinical parameters, including age, sex, and hematological indices, to better understand their predictive value in assessing childhood ALL susceptibility.

Figure 1.

Physical map for TIMP-2 rs8179090, rs4789936, rs2009196, and rs7342880 polymorphic sites.

Materials and Methods

Study population and data collection. Childhood ALL cases and cancer-free controls were recruited for this case-control study. Diagnoses of ALL were confirmed by board-certified pediatric oncologists. All participating cases completed a structured questionnaire with the assistance of their parents or legal guardians and provided peripheral blood samples for analysis. The control group consisted of healthy individuals without a history of malignancy, frequency-matched to cases by age (±2 years) and sex. All participants were of Taiwanese descent. The study protocol was reviewed and approved by the Institutional Review Board of China Medical University Hospital (CMUH111-REC1-038). Demographic and clinical characteristics of the study population are presented in Table I. There were no statistically significant differences in age or sex distribution between cases and controls (p>0.05), indicating successful matching. However, white blood cell counts were significantly elevated in the ALL group compared to the control group (p<0.0001). Among the 266 children diagnosed with ALL, 85.3% (n=227) had B-cell lineage leukemia, while 14.7% (n=39) presented with the T-cell subtype. Risk stratification of patients revealed that 48.9% (n=130) were classified as standard risk, 25.2% (n=67) as high risk, and 25.9% (n=69) as very high risk. Regarding survival outcomes, 25.9% of patients had a survival time of less than five years, whereas 74.1% survived beyond five years (Table I).

Table I. Distribution of demographics of the 266 patients with childhood acute lymphoblastic leukemia and the 266 matched controls.

ALL: Acute lymphoblastic leukemia; SD: Standard deviation; ^aBased on Student’s t-test; ^bBased on chi-square test with Yates’ correction; Statistically significant p-values are shown in bold.

TIMP-2 genotyping methodology. Peripheral blood samples were collected from all study participants, and genomic DNA was extracted within 24 h following standardized laboratory procedures, as previously described (22,23). Genotyping for TIMP-2 rs8179090, rs4789936, rs2009196, and rs7342880 was conducted using the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique. Specific primer sequences used for amplification of each SNP are provided in Table II. To ensure accuracy and reproducibility, all genotyping assays were independently performed in a blinded manner by two separate researchers. Complete concordance (100%) was observed between their results, confirming the reliability of the genotyping process.

Table II. The primer sequences, polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP) conditions for TIMP-2 genotyping.

SNP: Single nucleotide polymorphism; F and R indicate forward and reverse primers, respectively.

Statistical analysis. Differences in age distribution between childhood ALL cases and control participants were assessed using the Student’s t-test. The distribution of TIMP-2 genotypes between the two groups was compared using Pearson’s chi-square (χ²) test. Associations between individual TIMP-2 polymorphisms and the risk of childhood ALL were evaluated by calculating odds ratios (ORs) and corresponding 95% confidence intervals (CIs), based on logistic regression models under various stratification frameworks. A two-tailed p-value of less than 0.05 was considered indicative of statistical significance across all analyses.

Results

Genotypic distribution of TIMP-2 polymorphisms in childhood ALL and control groups. The genotypic frequencies of TIMP-2 polymorphisms rs8179090, rs4789936, rs2009196, and rs7342880 among 266 childhood ALL cases and 266 cancer-free controls are summarized in Table III, Table, IV, Table V, and Table VI. Genotype distributions among control subjects were consistent with Hardy-Weinberg equilibrium (HWE), with p-values of 0.7768, 0.7252, 0.7988, and 0.2470 for rs8179090, rs4789936, rs2009196, and rs7342880, respectively (Table III, Table IV, Table V, and Table VI). A statistically significant difference in the distribution of TIMP-2 rs8179090 genotypes was observed between ALL cases and controls (p for trend=0.0124, Table III). Specifically, individuals harboring the heterozygous CG or homozygous variant CC genotypes demonstrated elevated ORs of 1.23 (95%CI=0.83-1.82, p=0.3415) and 4.01 (95%CI=1.46-11.04, p=0.0076), respectively, when compared to those with the wild-type GG genotype. The association with the CC genotype was statistically significant. In the recessive genetic model, carriers of the CC genotype had a significantly increased risk of childhood ALL compared to individuals with the GG or CG genotypes combined (OR=3.79, 95%CI=1.39-10.36, p=0.0105). In the dominant model, individuals with CG or CC genotypes showed a borderline significant association with increased ALL risk relative to GG carriers (OR=1.42, 95%CI=0.98-2.06, p=0.0744). In contrast, no significant associations were detected between childhood ALL risk and the variant genotypes of TIMP-2 rs4789936, rs2009196, or rs7342880 in any of the genetic models assessed (Table IV, Table V, and Table VI).

Table III. TIMP-2 rs8179090 genotypes among the 266 patients with childhood acute lymphoblastic leukemia and 266 healthy controls.

^aBased on chi-square test with Yates’ correction; OR: odds ratio; CI: confidence interval; p_trend: p-value for trend analysis; p_HWE: p-value for Hardy-Weinberg equilibrium analysis; statistically significant p-values are shown in bold.

Table V. TIMP-2 rs2009196 genotypes among the 266 patients with childhood acute lymphoblastic leukemia and 266 healthy controls.

^aBased on chi-square test with Yates’ correction; OR: odds ratio; CI: confidence interval; p_trend: p-value for trend analysis; p_HWE: p-value for Hardy-Weinberg equilibrium analysis.

Table VI. TIMP-2 rs7342880 genotypes among the 266 patients with childhood acute lymphoblastic leukemia and 266 healthy controls.

^aBased on chi-square test with Yates’ correction (n≥5) or Fisher’s exact test (n<5); OR: odds ratio; CI: confidence interval; p_trend: p-value for trend analysis; p_HWE: p-value for Hardy-Weinberg equilibrium analysis.

Table IV. TIMP-2 rs4789936 genotypes among the 266 patients with childhood acute lymphoblastic leukemia and 266 healthy controls.

^aBased on chi-square test with Yates’ correction; OR: odds ratio; CI: confidence interval; p_trend: p-value for trend analysis; p_HWE: p-value for Hardy-Weinberg equilibrium analysis.

Allelic frequency distribution of TIMP-2 polymorphisms in childhood ALL. To further validate the genotypic association findings, allelic frequency analyses were conducted for the four TIMP-2 SNPs, with results summarized in Table VII. Among these, the variant C allele of TIMP-2 rs8179090 was significantly more prevalent in childhood ALL cases (20.7%) compared to controls (14.5%), corresponding to an OR of 1.54 (95%CI=1.12-2.12, p=0.0100). In contrast, no statistically significant differences in allele frequencies were observed between ALL cases and controls for TIMP-2 rs4789936, rs2009196, or rs7342880 (all p>0.05), indicating that these SNPs may not be associated with childhood ALL susceptibility in the Taiwanese population (Table VII).

Table VII. Distribution of allelic frequencies for TIMP-2 rs8179090 among the 266 patients with childhood acute lymphoblastic leukemia and 266 healthy controls.

^aBased on chi-square test with Yates’ correction; OR: odds ratio; CI: confidence interval; statistically significant p-values are shown in bold.

Association between TIMP-2 rs8179090 genotypes and age of onset in childhood ALL risk. To evaluate potential age-dependent effects of the TIMP-2 rs8179090 polymorphism on childhood ALL susceptibility, a stratified analysis based on age at onset (≤3.5 years versus >3.5 years) was conducted, as shown in Table VIII. A significant trend in genotype distribution was observed among participants aged 3.5 years or younger (p for trend=0.0228), but not in the older age group (p=0.4488). Notably, the homozygous variant TIMP-2 rs8179090 CC genotype was significantly associated with an elevated risk of childhood ALL in the younger subgroup (OR=5.06, 95%CI=1.39-18.34, p=0.0084). In contrast, among individuals older than 3.5 years, the CC genotype did not show a statistically significant association with ALL risk (OR=2.67, 95%CI=0.50-14.09, p=0.2738). Carriers of the heterozygous CG genotype did not exhibit a significant increase in ALL risk in either age group (both p>0.05). These findings remained consistent after adjusting for sex, further supporting a potential interaction between early age at onset and the TIMP-2 rs8179090 CC genotype in modulating childhood ALL susceptibility (Table VIII).

Table VIII. TIMP-2 rs8179090 genotypes in childhood acute lymphoblastic leukemia risk after stratification by age.

^aBy multivariate logistic regression analysis; ^bby multivariate logistic regression analysis after adjusted of sex; p_trend: p-value for rend analysis; CI: confidence interval; aOR: adjusted odds ratio; statistically significant p-values are shown in bold.

Association between TIMP-2 rs8179090 genotypes and sex in childhood ALL risk. To explore the potential interaction between TIMP-2 rs8179090 genotypes and sex in relation to childhood ALL susceptibility, a stratified analysis by sex was performed for both cases and controls, as shown in Table IX. Among boys, individuals carrying the homozygous variant CC genotype exhibited a significantly increased risk of childhood ALL (OR=5.53, 95%CI=1.54-19.84, p=0.0046). In contrast, this association was not observed among girls (OR=2.07, 95%CI=0.37-11.63, p=0.4435). Carriers of the heterozygous CG genotype did not show a significant association with ALL risk in either boys or girls (both p>0.05). These patterns remained consistent after adjusting for age, indicating that the observed sex-specific effect of the TIMP-2 rs8179090 CC genotype on ALL risk is independent of age (Table IX).

Table IX. TIMP-2 rs8179090 genotypes in childhood acute lymphoblastic leukemia risk after stratification by sex.

^aBy multivariate logistic regression analysis; ^bby multivariate logistic regression analysis after adjusted of age; p_trend: p-value for rend analysis; CI: confidence interval; aOR: adjusted odds ratio; statistically significant p-values are shown in bold.

Association of TIMP-2 rs8179090 genotypes with childhood ALL immunophenotype, risk stratification, and survival. To assess the potential clinical relevance of TIMP-2 rs8179090 polymorphism, we examined its association with immunophenotypes, risk classification, and survival outcomes in childhood ALL patients (Table X). The combined variant genotypes (CG+CC) were significantly associated with more aggressive risk classification, exhibiting an OR of 2.25 (95%CI=1.34-3.79, p=0.0031). In addition, carriers of the CG+CC genotypes demonstrated a significantly higher likelihood of shorter survival (defined as <5 years), with an OR of 3.68 (95%CI=1.82-7.45, p=0.0003). In contrast, no statistically significant association was observed between TIMP-2 rs8179090 genotypes and immunophenotypic subtypes (B- or T-cell lineages), suggesting that this polymorphism does not influence the lineage differentiation of childhood ALL (Table X).

Table X. TIMP-2 rs8179090 genotypes in childhood acute lymphoblastic leukemia risk among different immunophenotypes, risk classifications, and survival time.

CI: Confidence interval; OR: odds ratio; ^aBased on chi-square test with Yates’ correction. Statistically significant p-values are shown in bold.

Discussion

In the current study, we aimed to investigate the role of TIMP-2 genetic variants in the susceptibility to childhood ALL within a genetically and geographically representative Taiwanese pediatric population. To our knowledge, this is the first study globally to examine the association between TIMP-2 polymorphisms and childhood ALL risk. Our analysis identified a significant association between the rs8179090 genotype, located in the promoter region of the TIMP-2 gene, and increased risk of childhood ALL (Table III and Table VII). In contrast, no significant associations were found for rs4789936, rs2009196, or rs7342880 with ALL susceptibility (Table IV, Table V, and Table VI). Importantly, the sample size of our study is relatively large and drawn from a homogeneous population, enhancing the validity of our findings. We further demonstrated, for the first time, that the TIMP-2 rs8179090 CC genotype conferred a markedly increased risk of childhood ALL in younger cases (≤3.5 years) and boys (Table VIII and Table IX). Moreover, our data revealed that the CG and CC genotypes were selectively associated with more severe risk classification and shorter survival time among affected children (Table X), suggesting that this SNP may be useful not only in risk assessment but also in prognostic stratification. Collectively, these findings support the potential utility of TIMP-2 rs8179090 genotyping as a novel biomarker for identifying subgroups of pediatric patients at elevated risk for ALL. Functionally, TIMP-2 rs8179090 is located at position -418 within the Sp1 transcription factor binding site in the core promoter of the TIMP-2 gene (24). Sp1 is known to enhance transcriptional activity, and a G to C transversion at this site may reduce Sp1 binding efficiency, potentially leading to decreased TIMP-2 expression. This altered expression may disrupt the TIMP-2/MMP-2 regulatory axis, contributing to leukemogenesis.

In summary, our study provides novel evidence that TIMP-2 rs8179090 may serve as a genetic predictor of childhood ALL risk and prognosis, particularly among specific age and sex subgroups. Future functional studies and validation in independent cohorts are warranted to elucidate the mechanistic role of this variant and confirm its clinical applicability.

The anticancer properties of TIMP-2 have been documented across multiple cancer types, with evidence suggesting both tumor-suppressive and context-dependent roles. Several studies have reported down-regulation of TIMP-2 expression in malignancies such as gliomas and prostate cancer, particularly during advanced stages characterized by increased invasiveness and metastasis (25,26). Pulukuri and his colleagues (2007) specifically demonstrated that silencing of TIMP-2 in prostate cancer cell lines was closely linked to tumor progression and metastatic potential (26). In contrast, elevated TIMP-2 expression has been observed in certain tumor types, including bladder and ovarian cancers, suggesting a potential compensatory or tumor-specific response (27,28). In breast cancer, a study by Nakopoulou et al. involving 136 tumor samples found that down-regulation of TIMP-2 expression was correlated with increased tumor volume, whereas higher TIMP-2 levels were more frequently detected in low-grade tumors and were associated with longer patient survival (29). Interestingly, their analysis also revealed a positive correlation between TIMP-2 and bcl-2 expression, implying a possible anti-apoptotic role for TIMP-2 in certain cellular contexts. These findings collectively highlight the dual role of TIMP-2 in tumor biology, acting as a suppressor of invasion and metastasis in some cancers, while potentially contributing to tumor cell survival in others. Such complexity underscores the importance of context when evaluating TIMP-2 as a biomarker or therapeutic target.

Although TIMP-2 is widely recognized as an endogenous inhibitor of MMPs, the underlying mechanisms by which its genetic variants, particularly rs8179090, may modulate individual susceptibility to childhood ALL remain incompletely understood. Emerging evidence indicates that beyond its classical MMP-inhibitory role, TIMP-2 may paradoxically promote tumor cell proliferation, enhance invasiveness and metastatic potential, and suppress apoptosis in various malignancies (30-32). In triple-negative breast cancer, for instance, Peeney et al. demonstrated that TIMP-2 can inhibit proliferation and metastasis in a mouse model, highlighting a context-specific and tumor-type-dependent function (33). Several studies have also linked elevated TIMP-2 expression with increased tumor aggressiveness in cancers such as breast (34,35), skin (36), oral (37), laryngeal (38), gastric (39), colorectal (40), liver (41), renal (42), bladder (43), prostate (44) and ovarian cancer (45). In addition, TIMP-2 rs8179090 variant was found to be associated with breast and lung cancer risk (21,46). Furthermore, the expression level of TIMP-2 is useful for discriminating between malignant ovarian tumors and ovarian tumors of low malignant potential, resulting in poor prognosis of that disease (47). However, the role of TIMP-2 in pediatric leukemia was not totally unrevealed. In a recent study, peripheral blood samples were collected before the start of treatment and then on day 33 of intensive treatment from 31 patients diagnosed with ALL. The results indicate that the levels of MMP-2, MMP-7, and TIMP-2 did not statistically significantly change before and during treatment of ALL patients (48). Levels of MMP-9, TIMP-1 and VEGF-A were higher in chronic lymphocytic leukemia than ALL and controls. On the contrary, expression of TIMP-2 and MT1-MMP mRNA was significantly higher in acute leukemia than chronic lymphocytic leukemia in dogs (49). TIMP-2 may play a dual role in pediatric leukemia by exhibiting both tumor-suppressive and tumor-promoting functions. Its specific actions are likely modulated by factors such as leukemia subtype, tumor microenvironment, and genetic polymorphisms. Further investigation into the molecular mechanisms of TIMP-2 in pediatric leukemia may offer valuable insights for prognostic assessment and the development of targeted therapeutic strategies.

Study limitations. First, we did not measure TIMP-2 expression levels in peripheral blood samples from children with ALL versus healthy controls, nor across different immunophenotypic subtypes or clinical risk groups. Second, the research was conducted at a single medical center, which may limit the generalizability of our findings to other populations. Third, our study focused solely on the TIMP-2 gene and four of its polymorphisms. While rs8179090 was found to be significantly associated with childhood ALL risk, the predictive value of a single SNP is inherently limited.

In conclusion, our findings suggest that the TIMP-2 rs8179090 polymorphism, particularly the CC genotype, may serve as a novel genetic indicator of increased susceptibility to childhood ALL. Furthermore, this variant shows potential as a predictive biomarker for identifying patients at higher clinical risk and with shorter survival outcomes. Notably, the association appears to be especially relevant among younger children (≤3.5 years) and boys within the Taiwanese population. These results underscore the potential of TIMP-2 rs8179090 genotyping to contribute to personalized risk assessment and early intervention strategies in pediatric ALL. Further validation in larger and multi-ethnic cohorts is warranted to confirm its clinical utility.

Conflicts of Interest

The Authors declare no conflicts of interest with any company or person.

Authors’ Contributions

Research design: Hsu PC, Tsai CL and Pei JS; patient and questionnaire summaries: Hsu PC, Pei JS, Chen CC and Tzeng HE; experimental work: Chang WS, Wang YC and Shih HY; statistical analysis: Hsia TC, Chang WS and Tsai CW; article writing: Chang WS and Bau DT; review and revision: Chang WS and Bau DT.

Acknowledgements

The Authors are grateful to the Tissue Bank of China Medical University Hospital and doctors/nurses for their excellent sample collection. The technical assistance from Ai-Chia Tung and Yi-Wen Hung was very helpful. This study was supported by the Ministry of Health and Welfare to Taoyuan General Hospital (PTH112043), and Taichung Veterans General Hospital (TCVGH-1144801B).

Artificial Intelligence (AI) Disclosure

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.