Limits...
Contribution of neutrophil elastase to the lysis of obliterative thrombi in the context of their platelet and fibrin content.

Rábai G, Szilágyi N, Sótonyi P, Kovalszky I, Szabó L, Machovich R, Kolev K - Thromb. Res. (2010)

Bottom Line: Leukocytes invade newly formed thrombi through interactions with platelets and fibrin and later contribute to the removal of fibrin deposits mainly through the action of neutrophil elastase.The digitalized fluorescent microscopic images were decomposed according to the color channel of each thrombus constituent.Association between NE-FDP and leukocyte content of thrombi is evidenced by a significant Pearson correlation coefficient of 0.71 (p=0.00002).

View Article: PubMed Central - PubMed

Affiliation: Department of Medical Biochemistry, Semmelweis University, Budapest, Hungary.

Show MeSH

Related in: MedlinePlus

Clustering of patients according to the elastase-digested fibrin, leukocyte, platelet antigen and fibrin content of their thrombi. The thrombus constituents were quantified as illustrated in Fig. 1. Based on the NE-FDP, leukocyte and platelet antigen content two main clusters of patients (ELP1 and ELP2) were revealed by hierarchical, agglomerative clustering technique using Ward's method and Euclidean distances (A) and two main clusters (ELF1 and ELF2) were identified on the basis of NE-FDP, leukocyte and fibrin content (E). In panel B univariate comparison of the NE-FDP (red), leukocyte (blue) and platelet (green) content in clusters ELP1 and ELP2 is shown, whereas panel F presents the same analysis for clusters ELF1 and ELF2, where the green color stands for fibrin (mean and 95 % confidence intervals are indicated). Multiple correlation analysis was performed separately for the patients belonging to the clusters ELP1 shown with red symbols and ELP2 shown with green symbols (C, D).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2927910&req=5

fig4: Clustering of patients according to the elastase-digested fibrin, leukocyte, platelet antigen and fibrin content of their thrombi. The thrombus constituents were quantified as illustrated in Fig. 1. Based on the NE-FDP, leukocyte and platelet antigen content two main clusters of patients (ELP1 and ELP2) were revealed by hierarchical, agglomerative clustering technique using Ward's method and Euclidean distances (A) and two main clusters (ELF1 and ELF2) were identified on the basis of NE-FDP, leukocyte and fibrin content (E). In panel B univariate comparison of the NE-FDP (red), leukocyte (blue) and platelet (green) content in clusters ELP1 and ELP2 is shown, whereas panel F presents the same analysis for clusters ELF1 and ELF2, where the green color stands for fibrin (mean and 95 % confidence intervals are indicated). Multiple correlation analysis was performed separately for the patients belonging to the clusters ELP1 shown with red symbols and ELP2 shown with green symbols (C, D).

Mentions: Cluster analysis is an exploratory data analytic method, which does not require any a priori hypotheses about the phenomena under investigation, but rather lets the data speak. Thus, this analytic approach has definite advantages under the clinical conditions of our study, when little or no objective data are available about the age of thrombi. When the amount of elastase-digested fibrin, leukocyte content and platelet GPIIb/IIIa antigen were used for classification of thrombi, two main clusters (ELP1 and ELP2) emerged and these were verified by the discriminant function analysis. The size of cluster ELP1 is 18 and the cluster is more homogeneous than cluster ELP2, the size of which is 10 (Fig. 4A). We attempted to get a better insight into the clustering principles underlying the existence of these classes of thrombi by calculating multivariate means and variances of the clusters. Hotelling's multivariate statistics identified the clusters as significantly different in their thrombus composition (T2 = 96.0 F(3,24)=29.5, p < 0.00001). Thrombi isolated from patients of cluster ELP2 had significantly higher elastase-digested fibrin (p = 0.0005) content than thrombi of patients belonging to cluster ELP1 (Fig. 4B). The difference in platelet-related antigen content was nearly significant (p = 0.06). Since F-ratio test indicated inhomogeneity of variances in both cases, mean values were probed by Welch's method of separate variance estimate [17]. The mean values for leukocyte content (estimated from nuclear staining) were also different (p = 0.005), while variances did not differ significantly (Table 1). Moreover, we found that leukocyte content of thrombi predicted elastase-digested fibrin significantly in cluster ELP2 (multiple correlation R2 = 0.58), but not in cluster ELP1 (R2 = 0.25) (Fig. 4C-D). This association explains the greater variability of cluster ELP2, and that consequently it is less homogeneous than cluster ELP1. These results can be interpreted as a dynamical effect; inhomogeneity (greater variability) of cluster ELP2 might result from the patients being in a later phase of cell-related thrombolysis (due to variable invasion of leukocytes into the thrombus and their activation), while digestion has just begun in cluster ELP1.


Contribution of neutrophil elastase to the lysis of obliterative thrombi in the context of their platelet and fibrin content.

Rábai G, Szilágyi N, Sótonyi P, Kovalszky I, Szabó L, Machovich R, Kolev K - Thromb. Res. (2010)

Clustering of patients according to the elastase-digested fibrin, leukocyte, platelet antigen and fibrin content of their thrombi. The thrombus constituents were quantified as illustrated in Fig. 1. Based on the NE-FDP, leukocyte and platelet antigen content two main clusters of patients (ELP1 and ELP2) were revealed by hierarchical, agglomerative clustering technique using Ward's method and Euclidean distances (A) and two main clusters (ELF1 and ELF2) were identified on the basis of NE-FDP, leukocyte and fibrin content (E). In panel B univariate comparison of the NE-FDP (red), leukocyte (blue) and platelet (green) content in clusters ELP1 and ELP2 is shown, whereas panel F presents the same analysis for clusters ELF1 and ELF2, where the green color stands for fibrin (mean and 95 % confidence intervals are indicated). Multiple correlation analysis was performed separately for the patients belonging to the clusters ELP1 shown with red symbols and ELP2 shown with green symbols (C, D).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2927910&req=5

fig4: Clustering of patients according to the elastase-digested fibrin, leukocyte, platelet antigen and fibrin content of their thrombi. The thrombus constituents were quantified as illustrated in Fig. 1. Based on the NE-FDP, leukocyte and platelet antigen content two main clusters of patients (ELP1 and ELP2) were revealed by hierarchical, agglomerative clustering technique using Ward's method and Euclidean distances (A) and two main clusters (ELF1 and ELF2) were identified on the basis of NE-FDP, leukocyte and fibrin content (E). In panel B univariate comparison of the NE-FDP (red), leukocyte (blue) and platelet (green) content in clusters ELP1 and ELP2 is shown, whereas panel F presents the same analysis for clusters ELF1 and ELF2, where the green color stands for fibrin (mean and 95 % confidence intervals are indicated). Multiple correlation analysis was performed separately for the patients belonging to the clusters ELP1 shown with red symbols and ELP2 shown with green symbols (C, D).
Mentions: Cluster analysis is an exploratory data analytic method, which does not require any a priori hypotheses about the phenomena under investigation, but rather lets the data speak. Thus, this analytic approach has definite advantages under the clinical conditions of our study, when little or no objective data are available about the age of thrombi. When the amount of elastase-digested fibrin, leukocyte content and platelet GPIIb/IIIa antigen were used for classification of thrombi, two main clusters (ELP1 and ELP2) emerged and these were verified by the discriminant function analysis. The size of cluster ELP1 is 18 and the cluster is more homogeneous than cluster ELP2, the size of which is 10 (Fig. 4A). We attempted to get a better insight into the clustering principles underlying the existence of these classes of thrombi by calculating multivariate means and variances of the clusters. Hotelling's multivariate statistics identified the clusters as significantly different in their thrombus composition (T2 = 96.0 F(3,24)=29.5, p < 0.00001). Thrombi isolated from patients of cluster ELP2 had significantly higher elastase-digested fibrin (p = 0.0005) content than thrombi of patients belonging to cluster ELP1 (Fig. 4B). The difference in platelet-related antigen content was nearly significant (p = 0.06). Since F-ratio test indicated inhomogeneity of variances in both cases, mean values were probed by Welch's method of separate variance estimate [17]. The mean values for leukocyte content (estimated from nuclear staining) were also different (p = 0.005), while variances did not differ significantly (Table 1). Moreover, we found that leukocyte content of thrombi predicted elastase-digested fibrin significantly in cluster ELP2 (multiple correlation R2 = 0.58), but not in cluster ELP1 (R2 = 0.25) (Fig. 4C-D). This association explains the greater variability of cluster ELP2, and that consequently it is less homogeneous than cluster ELP1. These results can be interpreted as a dynamical effect; inhomogeneity (greater variability) of cluster ELP2 might result from the patients being in a later phase of cell-related thrombolysis (due to variable invasion of leukocytes into the thrombus and their activation), while digestion has just begun in cluster ELP1.

Bottom Line: Leukocytes invade newly formed thrombi through interactions with platelets and fibrin and later contribute to the removal of fibrin deposits mainly through the action of neutrophil elastase.The digitalized fluorescent microscopic images were decomposed according to the color channel of each thrombus constituent.Association between NE-FDP and leukocyte content of thrombi is evidenced by a significant Pearson correlation coefficient of 0.71 (p=0.00002).

View Article: PubMed Central - PubMed

Affiliation: Department of Medical Biochemistry, Semmelweis University, Budapest, Hungary.

Show MeSH
Related in: MedlinePlus