Abstract
Rationale & Objective
Matrix metalloproteinase 2 (MMP-2) plays an important role in the development of fibrosis, the final common pathway of chronic kidney disease (CKD). This study aimed to assess the relationship between repeated measures of MMP-2 and CKD progression in a large, diverse prospective cohort.
Study Design
In a prospective cohort of Chronic Renal Insufficiency Cohort (CRIC) participants (N = 3,827), MMP-2 was measured at baseline. In a case-cohort design, MMP-2 was additionally measured at year 2 in a randomly selected subcohort and cases of estimated glomerular filtration rate (eGFR) halving or kidney replacement therapy (KRT) (N = 1,439).
Setting & Participants
CRIC is a multicenter prospective cohort of adults with CKD.
Exposure
MMP-2 measured in plasma at baseline and at year 2.
Outcomes
A composite kidney endpoint (KRT/eGFR halving)
Analytical Approach
Weighted Cox proportional hazards models for case-cohort participants.
Results
Participants were followed for a median of 4.6 years from year 2 and 6.9 years from the baseline. Persistently elevated MMP-2 (≥300 ng/mL at both baseline and year 2) increased the hazard of the composite kidney endpoint (HR, 1.61; 95% CI, 1.07-2.42; P = 0.09) after adjusting for covariates. The relationship of persistently elevated MMP-2 was modified by levels of inflammation, with a 2.6 times higher rate of the composite kidney endpoint in those with high-sensitivity C-reactive protein < 2.5 g/dL at study entry. Heterogeneity of effect was found with proteinuria, with a baseline MMP-2 level of ≥300 ng/mL associated with an increased risk of the composite kidney endpoint (HR, 1.30; 95% CI, 1.09-1.54) only with proteinuria ≥ 442 mg/g.
Limitations
The observational study design limits causal interpretation.
Conclusions
Elevated MMP-2 is associated with CKD progression, particularly among those with low inflammation and those with proteinuria. Future investigations are warranted to confirm the reduction in risk of CKD progression among these subgroups of patients with CKD.
Index Words: Matrix metalloproteinase 2 (MMP-2), chronic kidney disease (CKD), CKD progression, initiation of kidney replacement therapy (KRT), eGFR slope, fibrosis
Plain Language Summary
Matrix metalloproteinase 2 (MMP-2) is a matrix-degrading protease involved in fibrosis and elevated in chronic kidney disease (CKD). Longitudinal patterns of MMP-2 have not previously been assessed as a predictor of CKD progression in a large prospective cohort. Here, we found that a higher baseline level and an increasing or persistently elevated 2-year pattern of MMP-2 were associated with CKD progression, independent of all covariates except proteinuria. The association of baseline MMP-2 with CKD progression differed by level of proteinuria, whereas levels of inflammation modified the associations of 2-year MMP-2 patterns with CKD progression.
Chronic kidney disease (CKD) is a leading cause of death in the United States and a public health burden affecting ∼15% of American adults.1 CKD is typically characterized by progressive loss of kidney function and structural kidney damage and is associated with high morbidity and mortality and an increased risk of cardiovascular disease (CVD).2 Many heterogeneous disease pathways result in CKD; however, the final common manifestation of nearly all chronic and progressive nephropathies is kidney fibrosis.2,3 This fibrosis is perpetuated by chronic injury resulting in unchecked fibrotic matrix deposition that results in loss of kidney function and progression toward the need for kidney replacement therapy (KRT).3
Matrix metalloproteinase 2 (MMP-2), or gelatinase A, is a zinc-containing, matrix-degrading protease that regulates key cellular events relating to fibrosis.4 Normally found at low levels in the kidney, MMP-2 is rapidly upregulated during the progression of fibrosis.4 MMP-2 has been identified as a biomarker that is prognostic of fibrosis on biopsy,5 and higher levels of MMP-2 have been found in those with CKD.6, 7, 8, 9 MMP-2 has also been found to be associated with longitudinal decline in kidney function as measured by estimated glomerular filtration rate (eGFR) in nondiabetic, nonproteinuric patients with coronary artery disease and CKD10 and in prospective kidney disease interventional studies of patients with type 2 diabetes.11
Proteinuria is associated with increased risk of KRT and death, and early decline in proteinuria has been found to be associated with decreased risk of KRT or death, particularly in those with a high baseline proteinuria.12 It has been shown that MMPs play a role in the development of proteinuric kidney diseases13 and MMP-2 has been shown to independently correlate with proteinuria in humans.14 In animal models of kidney disease, including aldosterone and sodium induced kidney damage,15 mesangial cell-mediated forms of glomerulonephritis,16 Alport syndrome,17 and type 1 diabetes,18 development of proteinuria or albuminuria was blocked or reversed through inhibition of MMP-2. This suggests that there may be a causal relationship between MMP-2 and proteinuria in some CKD etiologies.
The objective of this study was to investigate whether baseline or 2-year patterns of MMP-2 levels were independently associated with CKD progression among men and women enrolled in the Chronic Renal Insufficiency Cohort (CRIC) Study.
Methods
Study Design and Population
From June 2003 to August 2008, 7 clinical centers across the United States enrolled 3,939 adults with moderate to advanced CKD in the CRIC Study. Baseline analyses used the entire cohort with available plasma. To investigate the 2-year patterns of MMP-2 with CKD outcomes, a case-cohort study design was used for its combination of efficiency and ability to minimize bias in estimates through inverse probability weighting of cases and a representative subcohort.19 In this study design, a subcohort of 1,300 participants was randomly selected from the 3,939 original CRIC participants. This subcohort, combined with cases of the primary outcome, a composite kidney endpoint (eGFR halving or KRT), or secondary outcome KRT, are the populations eligible to be included in the 2-year pattern case-cohort analyses. The subcohort alone was used to estimate the eGFR slope. Because outcomes were assessed in 2-year pattern analyses beginning at year 2, those who withdrew, died, did not have a year 2 visit, or did not have plasma available at both baseline and year 2 were excluded, resulting in a maximum of 990 subcohort members that were included in analyses (Fig 1). Further study design information and baseline analysis methods are included in Item S1: Supplementary Methods.
Figure 1.
Baseline and 2-year change in MMP-2 analysis flow chart.
Primary Exposure
Plasma was available for assessment of MMP-2 for 3,827 participants at baseline. For 2-year-pattern analyses, baseline, and year 2 ethylenediaminetetraacetic acid (EDTA) plasma was available for assessment in 990 participants in the randomly selected subcohort and 449 nonsubcohort cases who initiated KRT or had eGFR halving after their year 2 visit (analysis flow chart in Fig 1).
Two-year patterns of MMP-2 were defined as Low/Low (MMP-2 < 300 ng/mL at baseline and year 2), Low/High (baseline < 300 ng/mL and year 2 ≥ 300 ng/mL), High/Low (baseline ≥ 300 ng/mL and year 2 < 300 ng/mL), and High/High (≥300 ng/mL at both time points). The cut point of 300 ng/mL was selected after examining the shape of a multivariable adjusted restricted cubic spline of baseline MMP-2 and outcome that showed an increased hazard with an increasing level of MMP-2 until reaching the median (309 ng/mL), at which point the hazard plateaued. Given this nonlinear relationship, continuous 2-year change was only assessed as a secondary outcome.20
MMP-2 was measured using an enzyme-linked immunosorbent assay (R & D Systems), run in duplicate and averaged for each sample, at the CRIC Central Laboratory at the University of Pennsylvania. The coefficients of variation for control samples with concentrations of 4.02 ng/mL and 20.9 ng/mL were 9.4% and 9.7%, respectively.
Outcomes and Censoring Events
The primary CKD progression outcome was time to initiation of KRT or eGFR halving (composite kidney endpoint). Time to KRT and change in eGFR over time (eGFR slope) were assessed as secondary outcomes. Outcomes were ascertained as described in Item S1. Participant follow-up was censored at loss to follow up, death, or the end of the follow-up period. For baseline analyses, outcomes were assessed after study entry for a median of 6.9 years of follow-up. For 2-year pattern analyses, outcomes were assessed beginning at year 2 with a median of 4.6 years of follow-up.
Covariates
Covariates in a series of nested models were age, sex, race or ethnicity (non-Hispanic White, non-Hispanic African American with low-risk APOL1 genotype [0 or 1 copies of APOL1 G1 and G2 risk alleles], non-Hispanic African American with high-risk APOL1 genotype [2 copies of APOL1 G1 and G2 risk alleles], Hispanic, and other), level of education, eGFR, body mass index, systolic blood pressure (SBP), current smoking, diabetes, history of CVD, urine neutrophil gelatinase-associated lipocalin to creatinine ratio (NGAL/cr), serum phosphate, high-sensitivity C-reactive protein (hsCRP), high-sensitivity Troponin T (hsTnT), N-terminal pro-B-type natriuretic peptide (NTproBNP), and urine protein to creatinine ratio (UPCR).
For baseline analyses, covariates were measured at baseline. For 2-year-pattern analyses, year 2 covariate measurements were used with the exception of UPCR and variables not measured at year 2 (hsCRP, phosphate, hsTNT, NTproBNP, and NGAL/cr), for which baseline measurements were used. The baseline measurements of UPCR were used to avoid potential overadjustment of MMP-2 through the inclusion of contemporaneous measurements of an intermediate variable hypothesized to be on the causal pathway to outcomes.21 Further covariate information and data collection methods are described in Item S1.
Statistical Analysis
Summary statistics were assessed using χ2, analysis of variance, or Kruskal-Wallis, where appropriate. To reduce the potential bias introduced by missing data, multiple imputations were performed to replace missing covariate values, as described in Item S1.
Cox proportional hazards models were used to model the hazards of the composite kidney endpoint or secondary outcome, initiation of KRT. For case-cohort analyses assessing associations of 2-year patterns of MMP-2, participants in the randomly selected subcohort and all cases with both baseline and year 2 MMP-2 measurements who experienced the event of interest after their second annual visit were included in the case-cohort analyses (composite kidney endpoint N = 1,373; KRT N = 1,298). A flow chart of participant inclusion in analyses is depicted in Fig 1. For these case-cohort analyses, the Barlow method of weighting was used,22 which is described further in Item S1.
A priori determined models stratified by study site were used to assess the relationship between MMP-2 and the composite kidney outcome, or KRT, as follows: unadjusted (Model 1), adjusted for age, sex, race or ethnicity, education, baseline eGFR, body mass index, SBP, current smoking, history of diabetes, and CVD (Model 2), adjusted for variables from Model 2 and urine NGAL/cr, serum phosphate, hsCRP, hsTNT, and NTproBNP (Model 3), and adjusted for variables from Model 3 and UPCR (Model 4). Covariates were included in continuous form unless otherwise indicated with right-skewed variables (UPCR, NGAL/cr, hsCRP, hsTNT, and NTproBNP) log transformed. The UPCR and eGFR were modeled as restricted cubic splines with knots placed at the 25th, 50th, and 75th percentiles. Effect modification in Cox proportional hazards models was explored by an a priori selected set of covariates, including UPCR (<132.6, 132.6-441.9, or ≥442 mg/g), age (< or ≥ 60 years), sex, and race or ethnicity (non-Hispanic White or other), SBP (< or ≥120 mm Hg), hsCRP (< or ≥2.5 g/dL), and eGFR (< or ≥30 mL/min/1.73m2). Multiple comparisons were controlled for in effect modification testing using the false discovery rate.
Linear mixed effects models with random intercepts and random slopes and an unstructured covariance structure were used to model the change in eGFR over time (eGFR slope). The same a priori determined covariates are used in Cox proportional hazard. Models 1 through 4 were used in eGFR slope models, with the difference that study site and baseline eGFR were included as additional covariates in all models. Two-year-change eGFR slope analyses included only members of the randomly selected subcohort of 1,300 with both baseline and year 2 MMP-2 measurements and 1 or more eGFR measurements from their year 2 annual visit or later (N = 927). All participants with baseline MMP-2 measurements were included in the baseline eGFR slope analyses (N = 3,827).
Statistical analyses were performed using SAS software v9.4 (Cary). An α of .05 was used to determine the significance. All hypothesis tests were 2-sided. Statistical analyses are described in further detail in Item S1.
Results
Baseline Analyses
Characteristics of the cohort participants (N = 3,827) are shown by the baseline level of MMP-2 in Table 1. The baseline characteristics and results are described further in Item S2: Supplementary Results. Crude incidence by quartile is shown in Table S1.
Table 1.
Baseline Demographic and Clinical Characteristics of Participants According to Level of Baseline MMP-2
| Characteristic |
Quartile 1 |
Quartile 2 |
Quartile 3 |
Quartile 4 |
Total |
P |
|---|---|---|---|---|---|---|
| Range (ng/mL) |
0.4-241.4 |
242.0-308.8 |
309.0-396.5 |
397.0-1379.0 |
0.4-1379.0 |
|
| N | 951 | 958 | 961 | 957 | 3,827 | |
| MMP-2, ng/mL | ||||||
| Median (IQR) | 205.0 (179.9-222.0) | 275.0 (259.0-290.2) | 348.0 (327.0-369.0) | 477.0 (433.0-538.2) | 309.0 (242.0-397.0) | |
| Demography | ||||||
| Age (y), mean ± SD | 55.2 ± 11.7 | 58.2 ± 10.6 | 58.9 ± 10.4 | 58.4 ± 11.0 | 57.7 ± 11.0 | <0.001 |
| Female sex, n (%) | 465 (48.9) | 422 (44.1) | 431 (44.9) | 407 (42.5) | 1,725 (45.1) | 0.04 |
| Racial/ethnicity, n (%) | ||||||
| NH-White | 428 (47.2) | 439 (47.7) | 386 (42.0) | 345 (37.8) | 1,598 (43.7) | <0.001 |
| NH-African American, APOL1 low-risk | 274 (30.2) | 276 (30.0) | 301 (32.8) | 295 (32.3) | 1,146 (31.3) | |
| NH-African American APOL1 high-risk | 74 (8.2) | 76 (8.3) | 61 (6.6) | 65 (7.1) | 276 (7.5) | |
| Hispanic | 88 (9.7) | 94 (10.2) | 136 (14.8) | 170 (18.6) | 488 (13.3) | |
| Other | 43 (4.7) | 36 (3.9) | 35 (3.8) | 39 (4.3) | 153 (4.2) | |
| Education level, n (%) | ||||||
| Less than high school | 135 (14.2) | 193 (20.2) | 214 (22.3) | 260 (27.2) | 802 (21.0) | <0.001 |
| High school graduate | 179 (18.8) | 180 (18.8) | 186 (19.4) | 177 (18.5) | 722 (18.9) | |
| Some college | 292 (30.7) | 287 (30.0) | 251 (26.1) | 276 (28.8) | 1,106 (28.9) | |
| College graduate or higher | 344 (36.2) | 298 (31.1) | 310 (32.3) | 244 (25.5) | 1,196 (31.3) | |
| Anthropometry | ||||||
| BMI, kg/m2 | ||||||
| Mean ± SD | 32.1 ± 7.3 | 32.0 ± 7.4 | 32.1 ± 8.0 | 32.3 ± 8.5 | 32.1 ± 7.8 | 0.88 |
| Blood pressure | ||||||
| Systolic, mmHg | ||||||
| Mean ± SD | 122.2 ± 18.6 | 125.8 ± 20.9 | 130.6 ± 22.2 | 135.6 ± 24.4 | 128.6 ± 22.2 | <0.001 |
| Medical history, n (%) | ||||||
| Diabetes | 303 (31.9) | 393 (41.0) | 510 (53.1) | 655 (68.4) | 1,861 (48.6) | <0.001 |
| Current smoker | 157 (16.5) | 120 (12.5) | 114 (11.9) | 102 (10.7) | 493 (12.9) | 0.001 |
| History of CVD | 222 (23.3) | 312 (32.6) | 339 (35.3) | 404 (42.2) | 1,277 (33.4) | <0.001 |
| Kidney function measure | ||||||
| eGFR, mL/min/1.73m2 | ||||||
| Mean ± SD | 50.1 ± 18.5 | 45.8 ± 16.0 | 44.0 ± 16.0 | 39.5 ± 14.6 | 44.8 ± 16.8 | <0.001 |
| Kidney injury markers | ||||||
| UPCR, mg/g | ||||||
| Median (IQR) | 93.7 (46.9-324.4) | 124.6 (52.2-483.5) | 165.3 (61.9-907.0) | 393.4 (84.9-2167.6) | 152.9 (57.5-777.0) | <0.001 |
| NGAL/Cr, ng/mg | ||||||
| Median (IQR) | 19.2 (8.7-47.9) | 20.1 (8.7-49.1) | 23.7 (10.1-60.7) | 33.1 (12.2-96.5) | 23.4 (9.8-61.0) | <0.001 |
| Metabolic marker | ||||||
| Phosphate, mg/dL | ||||||
| Mean ± SD | 3.6 ± 0.6 | 3.7 ± 0.7 | 3.7 ± 0.6 | 3.9 ± 0.7 | 3.7 ± 0.7 | <0.001 |
| Inflammatory marker | ||||||
| hsCRP, mg/L | ||||||
| Median (IQR) | 3.3 (1.3-7.6) | 2.7 (1.1-7.0) | 2.4 (1.0-5.5) | 2.0 (0.8-5.3) | 2.6 (1.1-6.4) | <0.001 |
| Cardiovascular risk markers | ||||||
| hsTNT, pg/mL | ||||||
| Median (IQR) | 7.7 (3.4-14.8) | 10.9 (5.6-19.7) | 13.8 (6.8-26.1) | 20.0 (10.4-38.1) | 12.3 (5.8-23.9) | <0.001 |
| NTproBNP, pg/mL | ||||||
| Median (IQR) | 81.3 (38.2-196.0) | 129.8 (58.5-309.2) | 171.2 (75.4-418.7) | 309.6 (124.1-820.2) | 152.5 (63.3-413.3) | <0.001 |
Note: SI conversion: To convert serum phosphate to mmol/L, multiply by 0.323.
Abbreviations: BMI, body mass index; CI, confidence interval; Cr, creatinine; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; hsCRP, high-sensitivity C-reactive protein; hsTNT, high-sensitivity Troponin T; IQR, interquartile range; MMP-2, matrix metalloproteinase-2; NGAL, neutrophil gelatinase-associated lipocalin; NH, non-Hispanic; NTproBNP, N-terminal pro-B-type natriuretic peptide; SD, standard deviation; UPCR, urine protein to creatinine ratio; yr, year.
In adjusted analyses, baseline MMP-2 was found to be independent of all covariates except proteinuria (Table 2). Proteinuria was found to be a significant effect modifier, with elevated MMP-2 levels (≥300 ng/mL) related to increased hazard of the composite kidney endpoint (HR, 1.30; 95% CI, 1.09-1.54) only in those with UPCR ≥ 442 mg/g (Fig 2). A similar relationship was found with KRT (Table S2 and Fig S1). In models including baseline MMP-2 as a restricted cubic spline without adjustment for UPCR or only including those with UPCR ≥ 442 mg/g, a plateauing of hazard was found, in which the hazard of time-to-event CKD outcomes increased until median levels of MMP-2 (309 ng/mL), at which point the hazard did not increase further (Fig S2). This plateauing of hazards informed the decision to investigate 2-year patterns of MMP-2 as above or below 300 ng/mL.
Table 2.
Hazard Ratios of the Composite Kidney Endpoint by Quartile of Baseline MMP-2
| Model 1 |
Model 2 |
Model 3 |
Model 4 |
|||||
|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | |
| Composite kidney endpoint (KRT/eGFR halving) | ||||||||
| Quartile 1 | (Reference) | <0.001 | (Reference) | <.001 | (Reference) | 0.007 | (Reference) | 0.41 |
| Quartile 2 | 1.26 (1.04-1.51) | 1.23 (1.01-1.48) | 1.19 (0.98-1.45) | 1.11 (0.92-1.35) | ||||
| Quartile 3 | 1.69 (1.41-2.02) | 1.39 (1.15-1.67) | 1.31 (1.08-1.58) | 1.18 (0.97-1.43) | ||||
| Quartile 4 | 2.67 (2.25-3.16) | 1.66 (1.39-1.99) | 1.38 (1.14-1.67) | 1.13 (0.93-1.37) | ||||
Model 1: Unadjusted. All Cox proportional hazard models were stratified by study site.
Model 2: Adjusted for age, sex, race/ethnicity, level of education, eGFR, body mass index, systolic blood pressure, current smoking, diabetes, and history of cardiovascular disease.
Model 3: Adjusted for variables from Model 2 and urine neutrophil gelatinase-associated lipocalin to creatinine ratio, serum phosphate, high-sensitivity C-reactive protein, high-sensitivity Troponin T, and N-terminal pro-B-type natriuretic peptide.
Model 4: Adjusted for variables from Model 3 and urine protein to creatinine ratio.
Results are pooled from 20 imputed datasets and shown by quartile of MMP-2 (ranges, ng/mL: Quartile 1 [0.4-241.4]; Quartile 2 [242.0-308.8]; Quartile 3 [309.0-396.5]; Quartile 4 [397.0-1379.0]. Analyses included 3,827 participants with 1,192 composite kidney endpoint events.
Abbreviations: CI, confidence interval; eGFR, estimated glomerular filtration rate; HR, hazard ratio; KRT, kidney replacement therapy; MMP-2, matrix metalloproteinase-2; SE, standard error.
Figure 2.
Effect modification of elevated baseline MMP-2 on the composite kidney endpoint Heterogeneity of the effect of elevated matrix metalloproteinase-2 (MMP-2; ≥300 ng/mL) on the composite kidney endpoint of initiation of kidney replacement therapy (KRT) and estimated glomerular filtration rate (eGFR) halving across subgroups was tested through inclusion of interaction terms. Models were stratified by study site and adjusted for Model 4 covariates: age, sex, race or ethnicity, level of education, eGFR, body mass index, systolic blood pressure (SBP), current smoking, diabetes, history of cardiovascular disease, urine neutrophil gelatinase-associated lipocalin to creatinine ratio, serum phosphate, high-sensitivity C-reactive protein (hsCRP), high-sensitivity Troponin T, and N-terminal pro-B-type natriuretic peptide, and urine protein to creatinine ratio (UPCR). The UPCR, age, race or ethnicity, SBP, hsCRP, eGFR, and baseline MMP-2 were categorized as shown for effect modification testing and included as main effects and interactions with all model variables. A false discovery rate correction (FDR q value) was used to correct for multiple comparisons using the P values from the interaction term between elevated MMP-2 and the indicated subgroup variables. Results are pooled estimates of 20 imputed datasets.
Two-year Change Analyses
Two-year change analyses were performed, including a categorical predictor indicating whether participants had above or below 300 ng/mL MMP-2 at baseline or at year 2. The baseline characteristics of the 1,439 participants included in the 2-year primary or secondary analyses are shown in Table 3 and described further in Item S2. The crude rates of outcomes for the subcohort are shown in Table S3.
Table 3.
Baseline Demographic and Clinical Characteristics of Participants According to Level of Baseline and Year 2 MMP-2
| MMP-2 (ng/mL) Low <300, High (≥300) |
Baseline: Low |
Baseline: Low |
Baseline: High |
Baseline: High |
Total |
P |
|---|---|---|---|---|---|---|
| Year 2: Low |
Year 2: High |
Year 2: Low |
Year 2: High |
|||
| N | 322 | 307 | 187 | 623 | 1,439 | |
| MMP-2 at baseline, ng/mL | ||||||
| Median (IQR) | 230.5 (193.0-266.0) | 250.0 (215.0-275.8) | 371.0 (333.0-436.0) | 409.0 (349.1-493.0) | 322.0 (252.0-416.0) | |
| MMP-2 at year 2, ng/mL | ||||||
| Median (IQR) | 244.5 (218.0-274.0) | 385.0 (338.0-443.0) | 250.0 (228.0-275.0) | 427.0 (361.0-514.0) | 347.0 (270.0-442.0) | |
| MMP-2 Δ, ng/mL | ||||||
| Mean ± SD | 13.8 ± 58.9 | 159.9 ± 91.2 | −146.1 ± 90.7 | 13.6 ± 127.1 | 24.1 ± 135.5 | |
| Demography | ||||||
| Age (y), mean ± SD | 55.5 ± 11.3 | 57.2 ± 11.0 | 56.8 ± 11.8 | 58.4 ± 10.3 | 57.3 ± 10.9 | 0.001 |
| Female sex, n (%) | 158 (49.1) | 122 (39.7) | 082 (43.9) | 260 (41.7) | 622 (43.2) | 0.09 |
| Racial/ethnicity, n (%) | ||||||
| NH-White | 145 (46.6) | 133 (44.9) | 081 (44.0) | 202 (33.9) | 0561 (40.5) | <0.001 |
| NH-African American, APOL1 low-risk | 087 (28.0) | 096 (32.4) | 054 (29.3) | 210 (35.3) | 0447 (32.3) | |
| NH-African American, APOL1 high-risk | 030 (9.6) | 025 (8.4) | 018 (9.8) | 044 (7.4) | 0117 (8.4) | |
| Hispanic | 027 (8.7) | 034 (11.5) | 023 (12.5) | 109 (18.3) | 0193 (13.9) | |
| Other | 022 (7.1) | 008 (2.7) | 008 (4.3) | 030 (5.0) | 0068 (4.9) | |
| Education level, n (%) | ||||||
| Less than high school | 040 (12.5) | 059 (19.2) | 036 (19.3) | 162 (26.0) | 0297 (20.7) | <0.001 |
| High school graduate | 058 (18.1) | 058 (18.9) | 033 (17.6) | 112 (18.0) | 0261 (18.2) | |
| Some college | 096 (29.9) | 089 (29.0) | 053 (28.3) | 171 (27.4) | 0409 (28.4) | |
| College graduate higher | 127 (39.6) | 101 (32.9) | 065 (34.8) | 178 (28.6) | 0471 (32.8) | |
| Anthropometry | ||||||
| BMI, kg/m2 | ||||||
| Mean ± SD | 32.1 ± 7.2 | 32.1 ± 7.6 | 31.5 ± 7.5 | 32.5 ± 8.2 | 32.2 ± 7.8 | 0.45 |
| Blood pressure | ||||||
| Systolic, mm Hg | ||||||
| Mean ± SD | 122.2 ± 18.6 | 127.4 ± 19.4 | 126.0 ± 21.6 | 136.3 ± 23.9 | 129.9 ± 22.3 | <0.001 |
| Medical history, n (%) | ||||||
| Diabetes | 112 (34.8) | 140 (45.6) | 085 (45.5) | 422 (67.7) | 0759 (52.7) | <0.001 |
| Current smoker | 046 (14.3) | 034 (11.1) | 025 (13.4) | 066 (10.6) | 0171 (11.9) | 0.34 |
| History of CVD | 086 (26.7) | 108 (35.2) | 055 (29.4) | 242 (38.8) | 0491 (34.1) | 0.001 |
| Kidney function measure | ||||||
| eGFR, mL/min/1.73m2 | ||||||
| Mean ± SD | 48.2 ± 17.8 | 43.0 ± 15.4 | 46.1 ± 17.6 | 39.3 ± 13.6 | 43.0 ± 15.9 | <0.001 |
| Kidney injury markers | ||||||
| UPCR, mg/g | ||||||
| Median (IQR) | 118.5 (53.0-445.5) | 243.1 (57.5-978.6) | 253.7 (64.5-905.2) | 480.0 (105.2-2135.7) | 267.9 (68.1-1186.3) | <0.001 |
| NGAL/Cr, ng/mg | ||||||
| Median (IQR) | 22.4 (8.3-54.2) | 19.5 (9.0-56.5) | 20.7 (10.2-53.6) | 32.7 (12.0-88.8) | 25.4 (10.2-67.0) | <0.001 |
| Metabolic Marker | ||||||
| Phosphate, mg/dL | ||||||
| Mean ± SD | 3.6 ± 0.6 | 3.7 ± 0.7 | 3.7 ± 0.6 | 3.8 ± 0.7 | 3.7 ± 0.7 | <0.001 |
| Inflammatory marker | ||||||
| hsCRP, mg/L | ||||||
| Median (IQR) | 3.0 (1.3-7.2) | 2.7 (1.1-6.2) | 2.0 (0.9-5.0) | 2.0 (0.8-4.9) | 2.4 (1.0-5.6) | <0.001 |
| Cardiovascular risk markers | ||||||
| hsTNT, pg/mL | ||||||
| Median (IQR) | 8.0 (4.1-15.1) | 12.5 (5.9-22.6) | 10.2 (5.1-17.9) | 19.5 (9.9-37.5) | 12.9 (6.3-26.4) | <0.001 |
| NTproBNP, pg/mL | ||||||
| Median (IQR) | 85.6 (43.4-206.3) | 139.1 (64.4-334.7) | 157.9 (62.2-376.7) | 264.2 (120.4-597.4) | 173.5 (71.7-430.8) | <0.001 |
Note: SI conversion: To convert serum phosphate to mmol/L, multiply by 0.323.
Abbreviations: BMI, body mass index; CI, confidence interval; Cr, creatinine; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; hsCRP, high-sensitivity C-reactive protein; hsTNT, high-sensitivity Troponin T; IQR, interquartile range; MMP-2, matrix metalloproteinase-2; MMP-2 Δ, Change in MMP-2 from baseline to year 2; NGAL, neutrophil gelatinase-associated lipocalin; NH, non-Hispanic; NTproBNP, N-terminal pro-B-type natriuretic peptide; SD, standard deviation; UPCR, urine protein to creatinine ratio; yr, year.
Table 4.
Hazard Ratios of the Composite Kidney Endpoint by MMP-2 Baseline and Year 2 Levels
| MMP-2 (ng/mL) Low <300, High ≥300 |
Model 1 |
Model 2 |
Model 3 |
Model 4 |
|||||
|---|---|---|---|---|---|---|---|---|---|
| Baseline | Year 2 | HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P |
| Composite kidney endpoint (KRT/eGFR halving) | |||||||||
| Low | Low | (Reference) | <0.001 | (Reference) | 0.004 | (Reference) | 0.01 | (Reference) | 0.09 |
| High | Low | 1.34 (0.93-1.93) | 1.45 (0.90-2.35) | 1.37 (0.84-2.25) | 1.10 (0.65-1.86) | ||||
| Low | High | 2.11 (1.55-2.90) | 1.71 (1.13-2.60) | 1.65 (1.09-2.50) | 1.34 (0.87-2.06) | ||||
| High | High | 3.12 (2.38-4.10) | 1.94 (1.34-2.82) | 1.86 (1.26-2.75) | 1.61 (1.07-2.42) | ||||
Model 1: Unadjusted. All models were stratified by study site.
Model 2: Adjusted for age, sex, race/ethnicity, level of education, eGFR, body mass index, systolic blood pressure, current smoking, diabetes, and history of cardiovascular disease.
Model 3: Adjusted for variables from Model 2 and urine neutrophil gelatinase-associated lipocalin to creatinine ratio, serum phosphate, high-sensitivity C-reactive protein, high-sensitivity Troponin T, and N-terminal pro-B-type natriuretic peptide.
Model 4: Adjusted for variables from Model 3 and urine protein to creatinine ratio.
Results are pooled from 30 imputed datasets and included 1,373 subcohort members and incident cases after year 2 with 663 events.
Abbreviations: CI, confidence interval; eGFR, estimated glomerular filtration rate; HR, hazard ratio; KRT, kidney replacement therapy; MMP-2, matrix metalloproteinase-2; SE, standard error.
Compared with those in the Low/Low MMP-2 group and after adjustment for Model 3 covariates, the Low/High MMP-2 two-year pattern is associated with over one and a half-fold rates of the composite kidney endpoint (HR, 1.65; 95% CI, 1.09-2.50) and the High/High 2-year pattern is associated with nearly 2-fold rates of the composite kidney endpoint (HR, 1.86; 95% CI, 1.26-2.75; Table 4). With adjustment for baseline UPCR in Model 4, the hazard ratio for High/High 2-year pattern was still significantly different from the Low/Low pattern (HR, 1.61; 95% CI, 1.07-2.42) though the overall type 3 P value testing for differences between all levels of the variable was marginally nonsignificant at P = 0.09.
Adjusted for Model 4 covariates, a significant interaction was detected by level of inflammation, with those with hsCRP < 2.5 mg/L having over two and a half times the rates of the composite kidney endpoint (HR, 2.60; 95% CI, 1.40-4.82) with the High/High 2-year pattern compared with those with the Low/Low pattern, whereas those with hsCRP ≥ 2.5 mg/L did not have a significantly elevated hazard (HR, 1.18; 95% CI, 0.63-2.21; Fig 3).
Figure 3.
Effect modification of baseline and year 2 levels of MMP-2 on the composite kidney endpoint. Heterogeneity of the effect of high Matrix metalloproteinase-2 levels (MMP-2; ≥300 ng/mL) at baseline or year 2 on a composite kidney endpoint of initiation of kidney replacement therapy (KRT) and estimated glomerular filtration rate (eGFR) halving across subgroups was tested through inclusion of interaction terms. Hazard ratios (HRs) are shown for participants with high MMP-2 levels at both baseline and year 2 (High/High) in comparison to participants with low levels of MMP-2 at both baseline and year 2 (Low/Low). A false discovery rate correction (FDR q value) was used to correct for multiple comparisons using the overall P values from the interaction term between the MMP-2 two-year pattern (4-level variable) and indicated subgroup variables. All models were stratified by study site and adjusted for Model 4 covariates: age, sex, race or ethnicity, level of education, eGFR, body mass index, systolic blood pressure (SBP), current smoking, diabetes, history of cardiovascular disease, urine neutrophil gelatinase-associated lipocalin to creatinine ratio, serum phosphate, high-sensitivity C-reactive protein (hsCRP), high-sensitivity Troponin T, N-terminal pro-B-type natriuretic peptide, and urine protein to creatinine ratio (UPCR). The UPCR, age, race or ethnicity, SBP, hsCRP, and eGFR were categorized as shown for effect modification testing and included as main effects and interactions with all model variables. Results are pooled estimates of 30 imputed datasets. Participants included in above models were members of the subcohort and cases that experienced the event of interest after year 2.
Two-year patterns in MMP-2 were also tested for associations with secondary outcomes, initiation of KRT, and change in the eGFR over time (eGFR slope), described further in Item S2. Associations with these outcomes were similar to those with the composite kidney outcome, with the highest associations found in those with persistently elevated MMP-2 and significant effect modification with low inflammation at the initiation of KRT (Tables S3 and S4 and Fig S3).
Change in the MMP-2 from baseline to year 2 was also explored as a continuous predictor and is characterized in Table S5. A significant increase in hazard of the composite kidney endpoint was found with 1 standard deviation (136 ng/mL) increase in MMP-2 from baseline to year 2 with adjustment for baseline MMP-2 and all covariates except UPCR, which attenuated the effect enough to become marginally nonsignificant (Table S6; HR, 1.17; 95% CI, 0.98-1.40). KRT followed a similar pattern to the composite kidney endpoint, but the eGFR slope was not associated with a continuous 2-year change in MMP-2 after adjusting for baseline eGFR (Table S6).
Discussion
In this large, well-characterized, and diverse prospective cohort of adults with CKD, we report associations between baseline and persistent (2-year) elevation of MMP-2 ≥ 300 ng/mL with an increase in the hazard of the composite kidney endpoint and KRT, independent of key demographic, kidney function, medical history, anthropometric, and inflammation covariates, with the exception of proteinuria. Proteinuria significantly modified the effect of elevated MMP-2 (≥300 ng/mL), with an increased risk of the composite kidney endpoint and KRT found only in participants with UPCR ≥ 442 mg/g. Effects of persistent elevation were significantly modified by level of inflammation, with a 2.6-fold higher rate of both CKD progression outcomes in those with a high-sensitivity C-reactive protein of <2.5 g/dL at study entry. We report associations between a single measurement of MMP-2 with CKD progression, independent of all tested covariates except proteinuria.
Potential reasons for the increased hazard of outcomes with a 2-year pattern of elevated MMP-2 among those with a C-reactive protein of <2.5 g/dL at study entry are manyfold and would require a complex analysis approach that is beyond the scope of this observational study. First, the relationship between MMP-2 and inflammation is complex. MMP-2 has previously been found to positively correlate with CRP in those with angina23,24 and coronary artery disease.25 However, CRP has not been found to have a substantive correlation with MMP-2 in multiple studies in the setting of CKD,26,27 including this one (Spearman ρ = −0.1). A potential cause of the association or lack thereof between MMP-2 and CRP among different populations could be that MMP-2 synthesis can be regulated by different factors such as proinflammatory cytokines or oxidative stress, which has previously been shown in multiple disease settings, including CKD.28, 29, 30, 31 Medication usage could potentially affect inflammation or MMP-2 expression. A recent publication by Kopanko et al,32 found differential expression of MMP-2 and inflammatory biomarkers and the relationship between those markers in subgroups with different medication usage; β-blocker usage was associated with lower MMP-2 and inflammation markers interleukin 6 (IL-6) and tumor necrosis factor-α, whereas calcium channel antagonists were associated with higher levels of these factors. In the group not using β-blockers, variations in MMP-2 levels were associated with the oxidative stress marker Cu/Zn superoxide dismutase (Cu/Zn SOD) and CKD etiology, neither of which are available in the CRIC cohort. In addition, though inflammation can be elevated at various stages of CKD progression and vary depending on the underlying cause, it can be particularly pronounced in the early stages of kidney fibrosis when injury triggers an inflammatory response in the injured cells and inflammatory cells in the kidney, which can in turn promote the activity and expression of MMP-2 and the development of renal interstitial fibrosis.33 It is possible that for some participants, the co-occurrence of elevated CRP and MMP-2 was a result of early stages of fibrosis and thus less likely to be associated with the hazard of outcomes during the follow-up period of this study. Last, MMP-2 is involved in multiple mechanistic pathways of multiple diseases, with some resulting in kidney damage and others in vascular damage.13 There is the potential that for some participants, co-elevation of CRP and MMP-2 was caused by a comorbid condition rather than kidney fibrosis. These complex relationships between inflammation and MMP-2 warrant further mechanistic studies to fully elucidate the relationship between inflammation and MMP-2 within specific CKD etiologies and subpopulations.
MMP-2 has also been shown to have an increasing trend with increasing CKD stages;34 however, we did not find effect modification by eGFR, suggesting the interaction between MMP-2 and proteinuria cannot fully be explained by CKD stage. MMP-2 has previously been found to not be associated with eGFR decline in a general nondiabetic population.35 Of the outcomes assessed in our study, the association of MMP-2 with eGFR slope, ie, rate of eGFR decline, was the weakest.
MMPs, including MMP-2, effect CKD and CVD through multiple pathophysiological mechanisms13 and are involved in healthy vascular remodeling and acute kidney injury repair, and inappropriate vascular remodeling and kidney fibrosis.36,37 Dimas et al38 suggested an approach of considering MMPs as beneficial regarding angiogenesis before the development of proteinuria and harmful after the early stages of atherosclerosis, nephritis, and glomerular basement membrane structural defects. Although potentially useful, this approach does not explain 2 previous prospective cohort studies finding MMP-2 to be an independent predictor of CKD progression in specific nonproteinuric or combined proteinuric and nonproteinuric subpopulations, in which cohorts were limited to those without diabetes with coronary artery disease10 or those with diabetes,11 respectively. Proteinuria is considered a marker of systemic vascular damage13 and also occurs consistently with diabetic kidney disease.2 A possible explanation of these seemingly inconsistent findings with our study and previous reports may be that these high-risk or “pro-proteinuric” populations have already traversed a threshold of disease progression after which increased MMP-2 expression is more likely to be due to upregulation of MMP-2 during the progression of fibrosis than to angiogenesis.
Reductions in proteinuria, especially in those with high baseline proteinuria, have previously been associated with a decreased risk of KRT or death.12 In multiple animal models of kidney disease, development of proteinuria or albuminuria was blocked or reversed through inhibition of MMP-2,15, 16, 17, 18 which has previously suggested MMP-2 could be a potential target for intervention in specific proteinuric CKDs. Here, we show participants that reduced their MMP-2 levels from above 300 ng/mL at baseline to below 300 ng/mL at year 2 had a similar risk of kidney endpoints as those with persistently low levels after 7 years of follow-up, which lends further support for MMP-2 as a potential promising target for intervention in established CKD. Specific and broad spectrum synthetic MMP inhibitors and inhibitory antibodies have been developed and used in animal models.13,39 Renin-angiotensin-aldosterone system inhibitors, which have been shown to reduce the risk of CKD progression and to reduce proteinuria,40 have been shown to inhibit MMP gelatinases.41
Strengths of this study include the large size and diversity of the CRIC cohort, with follow-up for as long as 9.8 years from the baseline. Inclusion of covariates informed by previous research reduces the likelihood of unmeasured confounders. Effect modification testing elucidated the relationship between single measurements of MMP-2 and proteinuria and 2-year patterns of MMP-2 and hsCRP. Multiple imputations of missing data (Table S7) reduced the potential bias in estimates and preserved sample size and statistical power. This study is the first to report associations between MMP-2 and CKD progression within a large prospective cohort containing a diverse CKD population or to report the relationship between 2-year patterns in MMP-2 and CKD outcomes, which were confirmed in sensitivity competing risk analyses (Table S8). This is also the first to report a higher risk associated with persistent elevation of MMP-2 in those with lower levels of inflammation. To our knowledge, the observation that a single measurement of MMP-2 is predictive of CKD progression in those with proteinuric CKD has not been reported elsewhere.
There are several limitations of this study. A major limitation is that we are unable to differentiate whether increased hazard with elevated MMP-2 is a direct effect of circulating MMP-2 in the kidney or whether the leakage of kidney-derived MMP-2 is largely responsible for the increase in MMP-2 levels that is associated with CKD progression. The CRIC Study population is not entirely representative of all patients with CKD in the United States, and it is possible that important covariates have not been considered.
Single elevated measurements of MMP-2 are associated with CKD progression in those with proteinuria. Persistently elevated MMP-2 over 2 years above 300 ng/mL is associated with CKD progression, with a 2.6-fold higher rate of the composite kidney outcome in those with C-reactive protein <2.5 g/dL at study entry. Future investigations should confirm if elevation of the fibrosis marker MMP-2 relates to an increased risk of CKD progression among these subgroups of patients with CKD. Further mechanistic studies identifying MMP-2’s role in specific proteinuric CKD pathways are warranted.
Article Information
The CRIC Study Investigators include Lawrence J. Appel, MD, Debbie Cohen, MD, Laura Dember, MD, Alan S. Go, MD, James P. Lash, MD, Robert G. Nelson, MD, PhD, Mahboob Rahman, MD, Panduranga S. Rao, MD, Vallabh O. Shah, PhD, and Mark L. Unruh, MD
Authors’ Full Names and Academic Degrees
Robin L. Baudier, MSPH, Paula F. Orlandi, MD, MSCE Wei Yang, PhD, Hsiang-Yu Chen, MS, Nisha Bansal, MD, J. Walker Blackston, MSPH, Jing Chen, MD, Rajat Deo, MD, MSCE, Mirela Dobre, MD, Hua He, PhD, Jiang He, MD, PhD, Ana C. Ricardo, MD, Tariq Shafi, MD, MHS, Anand Srivastava, MD, Dawei Xie, PhD, Katalin Susztak, MD, PhD, Harold I. Feldman, MD, MSCE, Amanda H. Anderson, PhD, and the CRIC Study Investigators
Authors’ Contributions
AHA designed the study; CRIC Study personnel carried out data and specimen collection; RLB and AHA analyzed the data; RLB made the figures; RLB and AHA drafted and revised the article; all authors either contributed to the conception or design of the work or the acquisition, analysis, or interpretation of data as well as gave input on revisions and approved the final version of the article.
Support
The CRIC Study is funded under cooperative agreements between the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK; U01DK060990, U01DK060984, U01DK061022, U01DK061021, U01DK061028, U01DK060980, U01DK060963, U01DK060902, and U24DK060990). Funding for this ancillary study of MMP-2 was provided by funding from the NIDDK (R01DK104730). In addition, this work was supported in part by: the Perelman School of Medicine at the University of Pennsylvania Clinical and Translational Science Award NIH/NCATS UL1TR000003, Johns Hopkins University UL1TR000424, University of Maryland GCRC M01RR16500, Clinical and Translational Science Collaborative of Cleveland, UL1TR000439 from the National Center for Advancing Translational Sciences (NCATS) component of the National Institutes of Health and NIH roadmap for Medical Research, Michigan Institute for Clinical and Health Research (MICHR) UL1TR000433, University of Illinois at Chicago CTSA UL1RR029879, Tulane COBRE for Clinical and Translational Research in Cardiometabolic Diseases P20GM109036, Kaiser Permanente NIH/NCRR UCSF-CTSI UL1RR024131, Department of Internal Medicine, University of New Mexico School of Medicine Albuquerque, NM R01DK119199. The views expressed in this article are those of the authors and do not reflect those of the National Institutes of Health (NIH). The NIH was not involved in the design, data collection, data analysis, interpretation of the data, writing of the report, or the decision to submit for publication. Dr Feldman is currently an employee of the Patient-Centered Outcomes Research Institute (PCORI) but the views presented in this article are solely the responsibility of the authors and do not necessarily represent the views of PCORI, its Board of Governors, or its Methodology Committee.
Financial Disclosure
The authors declare they have no relevant financial interests.
Acknowledgements
The authors would like to thank the CRIC Study participants, personnel, and investigators.
Peer Review
Received December 13, 2023, as a submission to the expedited consideration track with 5 external peer reviews. Direct editorial input from the Statistical Editor and the Editor-in-Chief. Accepted in revised form May 6, 2024.
Footnotes
Complete author and article information provided before references.
Figure S1: Effect modification of elevated baseline MMP-2 on initiation of KRT.
Figure S2: Adjusted hazard ratio for the composite kidney endpoint and KRT by level of baseline MMP-2 for participants overall or with protein to creatinine ratio above or below 50 mg/mmol.
Figure S3: Effect modification of baseline and year 2 levels of MMP-2 on kidney replacement therapy (KRT)
Item S1: Supplementary Methods.
Item S2: Supplementary Results.
Table S1: Crude Event Rates of the Composite Kidney Endpoint and KRT Endpoints by Quartile of Baseline MMP-2
Table S2: Hazard Ratios of KRT and Slopes of Estimated Glomerular Filtration Rate (mL/min/1.73m2/y) by Quartile of Baseline MMP-2.
Table S3: Post Year 2 Crude Event Rates of the Composite Kidney Endpoint and Initiation of KRT by MMP-2 Baseline and Year 2 levels.
Table S4: Hazard Ratios of Initiation of KRT and Slopes of Estimated Glomerular Filtration Rate (mL/min/1.73m2/y) by MMP-2 Baseline and Year 2 levels.
Table S5: Baseline Demographic and Clinical Characteristics by Change in MMP-2 from Baseline to Follow up.
Table S6: Hazard Ratios of the Composite Kidney Endpoint and KRT with Decrease in MMP-2 from Baseline to Year 2.
Table S7: Missingness by Variable and Timepoint.
Table S8: Hazard Ratios of the Composite Kidney Endpoint by Quartile of Baseline MMP-2 by Cause-Specific (primary analysis) and Competing Risks Fine and Gray analysis.
Contributor Information
Amanda H. Anderson, Email: ahanders@uab.edu.
CRIC Study Investigators:
Lawrence J. Appel, Debbie Cohen, Laura Dember, Alan S. Go, James P. Lash, Robert G. Nelson, Mahboob Rahman, Panduranga S. Rao, Vallabh O. Shah, and Mark L. Unruh
Supplementary Materials
Figure S1-S3; Item S1-S2; Table S1-S8.
References
- 1.US Department of Health and Human Services. Centers for Disease Control and Prevention . 2021. Chronic Kidney Disease in the United States. [Google Scholar]
- 2.Webster A.C., Nagler E.V., Morton R.L., Masson P. Chronic kidney disease. Lancet. 2017;389(10075):1238–1252. doi: 10.1016/S0140-6736(16)32064-5. [DOI] [PubMed] [Google Scholar]
- 3.Humphreys B.D. Mechanisms of renal fibrosis. Annu Rev Physiol. 2018;80:309–326. doi: 10.1146/annurev-physiol-022516-034227. [DOI] [PubMed] [Google Scholar]
- 4.Narula S., Tandon C., Tandon S. Role of matrix metalloproteinases in degenerative kidney disorders. Curr Med Chem. 2018;25(15):1805–1816. doi: 10.2174/0929867325666171205143441. [DOI] [PubMed] [Google Scholar]
- 5.Mansour S.G., Puthumana J., Coca S.G., Gentry M., Parikh C.R. Biomarkers for the detection of renal fibrosis and prediction of renal outcomes: a systematic review. BMC Nephrol. 2017;18(1):72. doi: 10.1186/s12882-017-0490-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lu L.C., Yang C.W., Hsieh W.Y., Chuang W.H., Lin Y.C., Lin C.S. Decreases in plasma MMP-2/TIMP-2 and MMP-9/TIMP-1 ratios in uremic patients during hemodialysis. Clin Exp Nephrol. 2016;20(6):934–942. doi: 10.1007/s10157-015-1221-0. [DOI] [PubMed] [Google Scholar]
- 7.Kobusiak-Prokopowicz M., Krzysztofik J., Kaaz K., Jolda-Mydlowska B., Mysiak A. MMP-2 and TIMP-2 in patients with heart failure and chronic kidney disease. Open Med (Wars) 2018;13:237–246. doi: 10.1515/med-2018-0037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bauvois B., Mothu N., Nguyen J., Nguyen-Khoa T., Noel L.H., Jungers P. Specific changes in plasma concentrations of matrix metalloproteinase-2 and -9, TIMP-1 and TGF-beta1 in patients with distinct types of primary glomerulonephritis. Nephrol Dial Transplant. 2007;22(4):1115–1122. doi: 10.1093/ndt/gfl743. [DOI] [PubMed] [Google Scholar]
- 9.Sanders S.F., Huitema M.G., Hanemaaijer R., van Goor H., Kallenberg C.G., Stegeman C.A. Urinary matrix metalloproteinases reflect renal damage in anti-neutrophil cytoplasm autoantibody-associated vasculitis. Am J Physiol Renal Physiol. 2007;293(6):F1927–F1934. doi: 10.1152/ajprenal.00310.2007. [DOI] [PubMed] [Google Scholar]
- 10.Hsu T.-W., Kuo K.-L., Hung S.-C., Huang P.-H., Chen J.-W., Tarng D.-C. Progression of kidney disease in non-diabetic patients with coronary artery disease: predictive role of circulating matrix metalloproteinase-2, -3, and -9. PLOS ONE. 2013;8(7) doi: 10.1371/journal.pone.0070132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mayer G., Heerspink H.J.L., Aschauer C., et al. Systems biology-derived biomarkers to predict progression of renal function decline in type 2 diabetes. Diabetes Care. 2017;40(3):391–397. doi: 10.2337/dc16-2202. [DOI] [PubMed] [Google Scholar]
- 12.Inker L.A., Levey A.S., Pandya K., et al. Early change in proteinuria as a surrogate end point for kidney disease progression: an individual patient meta-analysis. Am J Kidney Dis. 2014;64(1):74–85. doi: 10.1053/j.ajkd.2014.02.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Provenzano M., Andreucci M., Garofalo C., et al. The association of matrix metalloproteinases with chronic kidney disease and peripheral vascular disease: A light at the end of the tunnel? Biomolecules. 2020;10(1) doi: 10.3390/biom10010154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Nagano M., Fukami K., Yamagishi S., et al. Circulating matrix metalloproteinase-2 is an independent correlate of proteinuria in patients with chronic kidney disease. Am J Nephrol. 2009;29(2):109–115. doi: 10.1159/000151439. [DOI] [PubMed] [Google Scholar]
- 15.Butler M.J., Ramnath R., Kadoya H., et al. Aldosterone induces albuminuria via matrix metalloproteinase-dependent damage of the endothelial glycocalyx. Kidney Int. 2019;95(1):94–107. doi: 10.1016/j.kint.2018.08.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Steinmann-Niggli K., Ziswiler R., Kung M., Marti H.P. Inhibition of matrix metalloproteinases attenuates anti-Thy1.1 nephritis. J Am Soc Nephrol. 1998;9(3):397–407. doi: 10.1681/ASN.V93397. [DOI] [PubMed] [Google Scholar]
- 17.Zeisberg M., Khurana M., Rao V.H., et al. Stage-specific action of matrix metalloproteinases influences progressive hereditary kidney disease. PLOS Med. 2006;3(4) doi: 10.1371/journal.pmed.0030100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ramnath R.D., Butler M.J., Newman G., et al. Blocking matrix metalloproteinase-mediated syndecan-4 shedding restores the endothelial glycocalyx and glomerular filtration barrier function in early diabetic kidney disease. Kidney Int. 2020;97(5):951–965. doi: 10.1016/j.kint.2019.09.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.O’Brien K.M., Lawrence K.G., Keil A.P. The case for case-cohort: an applied epidemiologist’s guide to reframing case-cohort studies to improve usability and flexibility. Epidemiology. 2022;33(3):354–361. doi: 10.1097/EDE.0000000000001469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Glymour M.M., Weuve J., Berkman L.F., Kawachi I., Robins J.M. When is baseline adjustment useful in analyses of change? An example with education and cognitive change. Am J Epidemiol. 2005;162(3):267–278. doi: 10.1093/aje/kwi187. [DOI] [PubMed] [Google Scholar]
- 21.Schisterman E.F., Cole S.R., Platt R.W. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology. 2009;20(4):488–495. doi: 10.1097/EDE.0b013e3181a819a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Barlow W.E., Ichikawa L., Rosner D., Izumi S. Analysis of case-cohort designs. J Clin Epidemiol. 1999;52(12):1165–1172. doi: 10.1016/s0895-4356(99)00102-x. [DOI] [PubMed] [Google Scholar]
- 23.Sun H., Zhang J., Zheng Y., Shang S. Expressions and clinical significance of factors related to acute coronary syndrome. J Biol Regul Homeost Agents. 2018;32(2):299–305. [PubMed] [Google Scholar]
- 24.Zeng B., Prasan A., Fung K.C., et al. Elevated circulating levels of matrix metalloproteinase-9 and -2 in patients with symptomatic coronary artery disease. Intern Med J. 2005;35(6):331–335. doi: 10.1111/j.1445-5994.2005.00822.x. [DOI] [PubMed] [Google Scholar]
- 25.Muzzio M.L., Miksztowicz V., Brites F., et al. Metalloproteases 2 and 9, Lp-PLA(2) and lipoprotein profile in coronary patients. Arch Med Res. 2009;40(1):48–53. doi: 10.1016/j.arcmed.2008.10.011. [DOI] [PubMed] [Google Scholar]
- 26.Rysz J., Banach M., Stolarek R.A., et al. Serum metalloproteinases MMP-2, MMP-9 and metalloproteinase tissue inhibitors TIMP-1 and TIMP-2 in patients on hemodialysis. Int Urol Nephrol. 2011;43(2):491–498. doi: 10.1007/s11255-009-9683-1. [DOI] [PubMed] [Google Scholar]
- 27.Preston G.A., Barrett C.V., Alcorta D.A., et al. Serum matrix metalloproteinases MMP-2 and MMP-3 levels in dialysis patients vary independently of CRP and IL-6 levels. Nephron. 2002;92(4):817–823. doi: 10.1159/000065464. [DOI] [PubMed] [Google Scholar]
- 28.Pawlak K., Mysliwiec M., Pawlak D. Peripheral blood level alterations of MMP-2 and MMP-9 in patients with chronic kidney disease on conservative treatment and on hemodialysis. Clin Biochem. 2011;44(10-11):838–843. doi: 10.1016/j.clinbiochem.2011.03.143. [DOI] [PubMed] [Google Scholar]
- 29.Pawlak K., Pawlak D., Mysliwiec M. Extrinsic coagulation pathway activation and metalloproteinase-2/TIMPs system are related to oxidative stress and atherosclerosis in hemodialysis patients. Thromb Haemost. 2004;92(3):646–653. doi: 10.1160/TH04-02-0128. [DOI] [PubMed] [Google Scholar]
- 30.Pawlak K., Pawlak D., Mysliwiec M. Urokinase-type plasminogen activator and metalloproteinase-2 are independently related to the carotid atherosclerosis in haemodialysis patients. Thromb Res. 2008;121(4):543–548. doi: 10.1016/j.thromres.2007.06.008. [DOI] [PubMed] [Google Scholar]
- 31.Pawlak K., Tankiewicz J., Mysliwiec M., Pawlak D. Systemic levels of MMP2/TIMP2 and cardiovascular risk in CAPD patients. Nephron Clin Pract. 2010;115(4):c251–c258. doi: 10.1159/000313483. [DOI] [PubMed] [Google Scholar]
- 32.Kopanko M., Zabludowska M., Pawlak D., et al. The possible effect of beta-blocker use on the circulating MMP-2/TIMP-2 system in patients with chronic kidney disease on conservative treatment. J Clin Med. 2024;13(7) doi: 10.3390/jcm13071847. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cheng Z., Limbu M.H., Wang Z., et al. MMP-2 and 9 in Chronic kidney disease. Int J Mol Sci. 2017;18(4) doi: 10.3390/ijms18040776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Peiskerova M., Kalousova M., Kratochvilova M., et al. Fibroblast growth factor 23 and matrix-metalloproteinases in patients with chronic kidney disease: are they associated with cardiovascular disease? Kidney Blood Press Res. 2009;32(4):276–283. doi: 10.1159/000243050. [DOI] [PubMed] [Google Scholar]
- 35.Enoksen I.T., Svistounov D., Norvik J.V., et al. Serum matrix metalloproteinase 7 and accelerated glomerular filtration rate decline in a general non-diabetic population. Nephrol Dial Transplant. 2022;37(9):1657–1667. doi: 10.1093/ndt/gfab251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Galis Z.S., Khatri J.J. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res. 2002;90(3):251–262. [PubMed] [Google Scholar]
- 37.Parrish A.R. Matrix metalloproteinases in kidney disease: role in pathogenesis and potential as a therapeutic target. Prog Mol Biol Transl Sci. 2017;148:31–65. doi: 10.1016/bs.pmbts.2017.03.001. [DOI] [PubMed] [Google Scholar]
- 38.Dimas G., Iliadis F., Grekas D. Matrix metalloproteinases, atherosclerosis, proteinuria and kidney disease: linkage-based approaches. Hippokratia. 2013;17(4):292–297. [PMC free article] [PubMed] [Google Scholar]
- 39.Wozniak J., Floege J., Ostendorf T., Ludwig A. Key metalloproteinase-mediated pathways in the kidney. Nat Rev Nephrol. 2021;17(8):513–527. doi: 10.1038/s41581-021-00415-5. [DOI] [PubMed] [Google Scholar]
- 40.Ames M.K., Atkins C.E., Pitt B. The renin-angiotensin-aldosterone system and its suppression. J Vet Intern Med. 2019;33(2):363–382. doi: 10.1111/jvim.15454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sorbi D., Fadly M., Hicks R., Alexander S., Arbeit L. Captopril inhibits the 72 kDa and 92 kDa matrix metalloproteinases. Kidney Int. 1993;44(6):1266–1272. doi: 10.1038/ki.1993.378. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1-S3; Item S1-S2; Table S1-S8.