Table 1.

Summary of the 17 articles revised.

Authors and title	Objective of the study	Design	Material and sample	Analysis method	Conclusion
Baker et al. [4]	To measure serum levels of collagenases, stromelysins, and TIMP-1 and 2 in patients with PCa, before treatment and 6 and 12 months after starting.	Prospective cohort	Test: serum of 19 individuals with metastatic PCa and 16 with PCa without metastases.	ELISA	Increase of collagenases and TIMP-1 in patients with metastatic PCa compared to those without metastases and in the former in relation to the control group with or without rheumatoid arthritis.
			Control: 21 patients with rheumatoid arthritis and 57 healthy subjects without rheumatoid arthritis.		Reduction of TIMP-1 and collagenase levels 6 months after treatment. After 12 months, the levels of collagenases remained low; however, those of TIMP-1 returned to pretreatment values.
Bonaldi et al. [5]	To dose e-cadherin and MMP-13 at the diagnosis of PCa and three and six months after treatment, comparing with the control group.	Prospective cohort	Test: plasma (EDTA) of 29 PCa patients.	ELISA	No difference between mean MMP-13 values among test and control groups at any test period.
			Control: 10 healthy men with PSA <1.5 ng/ml.
Castellano et al. [6]	To compare levels of osteopontin (OPN), MMP-2, MMP-9, and TIMP-1.	Cross-sectional	Test: plasma (heparin) of 96 patients with PCa.	ELISA	Differences of MMP-9 and TIMP-1 (but not MMP-2) between groups; significant increase of MMP-9 and reduction of TIMP-1 in the CaP group relative to the healthy and BPH control; decreased serum levels of MMP-9 six months after radical prostatectomy.
			Control: 92 individuals with BPH and 125 healthy subjects.
Cicco et al. [7]	Correlate preoperative serum levels of 6 markers (including MMPs-2 and 9 and TIMPs-1 and 2) with tumor staging, Gleason score, and disease-free survival.	Cross-sectional	Serum of 162 PCa carriers for MMP-2 and 9 and plasma (EDTA) for TIMP-1 and 2.	ELISA	Patients with serum levels of MMP-2 < 206 ng/ml had a higher risk of PCa progression.
Gong et al. [8]	To compare TIMP-1 levels of castrated metastatic PCa patients with noncastrated (responsive to androgen ablation therapy).	Descriptive	Test: serum of 39 castrated metastatic PCa patients.	ELISA	Higher TIMP-1 serum levels in castrated PCa patients.
			Control: 24 noncastrated metastatic PCa patients.
Incorvaia et al. [9]	To compare levels of MMP-2 and 9 in individuals with PCa with bone metastases in relation to healthy individuals.	Cohort	Test: plasma (EDTA) of 35 patients with breast cancer and 44 with PCa with bone metastases.	ELISA	MMP-2 and MMP-9 significantly higher in PCa patients with bone metastases than in the control group.
			Control: 57 healthy patients.
Jung et al. [10]	To compare levels of MMP-1, MMP-3, and TIMP-1 as well as the MMP-1/TIMP-1 ratio of subjects with metastatic PCa and with nonmetastatic PCa.	Cross-sectional	Plasma (heparin) of 47 patients with prostate cancer, 29 with no metastasis (T2, 3pN0M0), and 18 with metastasis (T2, 3, 4pN1, 2M1).	ELISA	Mean MMP-1 and TIMP-1 scores were significantly higher in the metastatic PCa group than in the nonmetastatic PCa, BPH, and healthy subjects groups.
			Control: 35 healthy subjects and 29 with BPH.		10 of the 18 patients with metastatic PCa presented high levels of TIMP-1.
Morgia et al. [11]	To measure plasma levels of MMPs-2, 9, and 13 of TIMP-1, and of the enzymatic activity of MMPs-2 and 9 in patients with metastatic PCa, nonmetastatic PCa, BPH, and healthy, at diagnosis and 90 days after starting treatment.	Cohort	Plasma (heparin) of 40 patients with prostate cancer, 20 with no metastasis and 20 with metastasis.	ELISA	Plasma levels of MMP-2, 9, and 13 higher at diagnosis in the PCa group with metastasis than in the other groups, with reduction after treatment.
			Control: 20 healthy patients and 20 with BPH.		Decreased TIMP-1 in the PCa group with metastasis in relation to the healthy group but without significant difference between groups.
Oh et al. [12]	To evaluate TIMP-1 as a predictor of survival in castrated PCa patients.	Survival study	Test: plasma (EDTA) of 362 castrated PCa patients; sample was divided into two groups: one with 60 (pilot group) individuals with a follow-up time of 5.8 years and the other with 302 (primary group) participants followed by 6.6 years.	ELISA	Lower survival rates among individuals with higher levels of TIMP-1 in both groups.
Prior et al. [13]	To determine sensitivity, specificity, and predictive values for MMP-2 as a biomarker for PCa.	Diagnostic study	Test: serum of 34 PCa patients.	ELISA	Increased levels of MMP-2 among subjects with PCa compared to the group without PCa.
					Sensitivity: 24.1%; specificity: 78.6%; PPV: 31.8%; NPV: 71.4%.
			Control: 79 patients without PCa.		Cutoff of 718.36 ng/ml (mean level of MMP-2 in those without PCa).
González Rodríguez et al. [14]	To dose MMP-9 in patients who underwent prostate biopsy.	Cross-sectional	Test: serum of 32 patients with positive biopsy (PCa group).	ELISA	No difference in MMP-9 levels between groups.
			Control: 58 patients with negative biopsy.
Salminen et al. [15]	To evaluate the prognostic value of MMP-2 and MMP-9 in PCa with and without bone metastasis, comparing with ALP and PSA.	Cross-sectional and prognostic	Test: serum of 35 individuals with PCa with bone metastasis.	ELISA	No differences in MMP-2 and 9 levels between groups.
			Control: 49 individuals with PCa without bone metastasis.		MMP-2 and 9 presented low accuracy for the diagnosis of bone metastasis in PCa and were not associated with survival.
Szarvas et al. [16]	To compare serum levels of MMP-7 in PCa patients with and without metastasis and to assess its prognostic value.	Cross-sectional and prognostic	Test: serum of 93 individuals with localized PCa and 13 PCa cases with bone metastasis.	ELISA	Higher serum levels of MMP-7 in PCa patients with distant metastasis; specificity of 69% and sensitivity of 92% for detection of metastasis.
			Control: 19 healthy individuals.
Zhang et al. [17]	To search mRNA and enzymatic activity of MMP-2 and 9 in prostatic tissue and serum of PCa patients (with and without metastasis) comparing with BPH and healthy group.	Cross-sectional	Test: serum of 15 PCa patients with metastasis and 10 without metastasis.	RT-PCR and zymography	Increased expression and enzymatic activity of MMP-9 compared to the other groups.
			Control: 26 BPH patients and 20 healthy.
Gil-Ugarteburu et al. [18]	To correlate the 1562C/T polymorphism of the MMP-9 gene with its plasma levels.	Prospective cohort	Test: plasma (heparin) of 90 patients submitted to prostatic biopsy with positive results for PCa.	ELISA	No correlation between the gene polymorphism and plasma concentration of MMP-9.
			Control: 135 with negative biopsy for PCa.
Kanoh et al. [19]	To correlate the serum levels of MMP-2 and PSA with the different stages of PCa.	Cross-sectional	Test: serum of 51 PCa patients.	ELISA	MMP-2 and PSA levels associated with metastatic PCa; higher levels of MMP-2 (>950 ng/ml) and PSA (>300 ng/ml) in PCa with bone metastasis.
			Control: serum of 39 BPH.
Gohji et al. [20]	To compare MMP-2 levels between individuals with and without PCa.	Cross-sectional	Test: serum of 98 individuals with PCa without previous treatment.	ELISA	Higher levels of MMP-2 in the PCa than in the control group.
			Control: serum of 76 individuals with BPH and 70 healthy.

BPH = benign prostate hyperplasia; MMP = matrix metalloproteinase; NPV = negative predictive value; PCa = prostate cancer; PPV = positive predictive value; TIMP = tissue inhibitor of metalloproteinase.