Vein Wall Fibrosis Research Articles

Abstract 196: Matrix Metalloproteinases Regulation in Thrombus Resolution Is Independent of TGF-β in a PAI-1 Ko Mouse Model of Venous Thrombosis

Introduction: Plasminogen activator inhibitor (PAI-1) is the primary regulator of urokinase (u-PA) and tissue (t-PA) plasminogen activators. Observations in mice with impaired fibrinolysis have demonstrated an alternate pathway of thrombus resolution by matrix metalloproteinase (MMP)- 9 and 2. TGF-β is a potent inducer of TIMP-1 expression and is found in both thrombus and vein wall post-thrombosis. Hypothesis: PAI-1 deficiency and resulting hyperfibrinolytic state will improve thrombus resolution and result in reciprocal downregulation of MMP-2 and MMP-9 via TGF-beta induction of TIMP-1, resulting in decreased vein wall fibrosis. Methods: Inferior vena cava (IVC) ligation was performed in wild-type (WT) mice (N=194) and PAI-1 deficent mice (N=164). IVC, thrombus and blood samples were collect at 2, 6 and 14 days post procedure. The IVC and thrombus were assessed for thrombus weight and samples collected for zymography (MMP-2 and MMP-9), ELISA (TIMP-1, TGF-β) and histology. Results: PAI-1 deficient mice had similar thrombus weight (TW) versus WT at 2 days, but smaller at days 6 (p<0.0001) and 14 (p=0.0085). In thrombus and vein wall (VW), active MMP-9 was decreased in acute thrombosis (p<0.0001), and MMP-2 was decreased during late thrombus resolution (p<0.05). TIMP-1 levels in VW were elevated across all time points (p<0.05), whereas VW TGF-beta was not different compared to WT. Neutrophils were decreased at day 2 (p = 0.0309) and monocytes decreased at day 14 (p = 0.0028). Conclusions: (1) PAI-1 deficient mice exhibit a hyperfibrinolytic phenotype with normal thrombotic response to IVC ligation, but improved thrombus resolution when compared to WT mice. (2) MMP-9 and MMP-2 were suppressed, confirming a reciprocal relationship between plasmin and MMP activation as mediators of thrombus resolution. (3) TIMP-1 production was elevated in vein wall independent of TGF-beta. (4) Decreased neutrophils at day 2 correlated with decreased MMP-9 and decreased monocytes at day 14 correlated with decreased MMP-2, suggesting a plasmin-induced inhibition of leukocyte recruitment. (5) Final histology showed no difference in vein wall fibrosis or thickness at chronic time points, indicating that improved thrombus resolution does not lessen vein wall fibrosis.

Statins, inflammation and deep vein thrombosis: a systematic review

Venous thromboembolism (VTE) includes both deep vein thrombosis (DVT) and pulmonary embolism. The 2009 JUPITER trial showed a significant decrease in DVT in non-hyperlipidemic patients, with elevated C-reactive protein (CRP) levels, treated with rosuvastatin. The effects of statins on thrombosis are unclear, prompting this literature review. A literature search was performed (1950 to February 2011) with MEDLINE, EMBASE, and PUBMED databases including the following keywords: "statins", "hydroxymethylglutaryl-CoA reductase inhibitors", "VTE", "PE", "DVT", and either "anti-coagulation" or "inflammation". Editorials, reviews, case reports, meta-analysis and duplicates were excluded. Inflammatory biomarkers of DVT, include interleukin (IL)-6, CRP, IL-8, and monocyte chemotactic protein 1 (MCP-1). Statin therapy reduces IL-6 expression of CRP and MCP-1, usually elevated in VTE. Reduction of IL-6 induced MCP-1 has been linked to vein wall fibrosis, promoting post thrombotic syndrome (PTS) and recurrent DVT in patients. Also, our review suggests that the anti-thrombotic effects are likely exhibited through the anti-inflammatory properties of statins. This work supports that statin therapy has the ability to decrease the incidence and recurrence of VTE and the potential to decrease PTS. This is mainly due to the anti-inflammatory effects of statins and may explain why normolipidemic patients, with elevated CRP, appear to have the greatest reduction in VTE. Given their low risk of bleeding, statins have the potential to serve as a safe adjunctive pharmacological therapy to current treatments in select patients with VTE, however further investigations into this concept are needed and essential.

Induction of Carbon Monoxide Partially Mitigates Thrombus Resolution and Decreases Vein Wall Fibrosis in a Murine Model of Venous Thrombosis

Induction of Carbon Monoxide Partially Mitigates Thrombus Resolution and Decreases Vein Wall Fibrosis in a Murine Model of Venous Thrombosis

Interleukin-6: A Potential Target for Post-Thrombotic Syndrome

Deep vein thrombosis (DVT) and its associated sequelae, post-thrombotic syndrome (PTS), are significant health care problems in the United States. It is estimated that a maximum of 60% of patients diagnosed with DVT develop PTS, which is characterized by extensive perivenous and mural fibrosis. Interleukin-6 (IL-6) has been linked to fibrosis, and high circulating plasma levels have been found to increase the risk of developing DVT. The aim of this study was to elucidate the role of IL-6 in the progression of vein wall fibrosis by using a mouse model of DVT. C57BL/6 mice (n = 136) were treated with either anti-IL-6 monoclonal antibody or control rat-immunoglobulin G. Thrombus was induced by using an inferior vena cava ligation model. The inferior vena cava and thrombus were harvested at days 2, 6, or 14 for thrombus weight, gene expression of IL-6 and/or C-C motif chemokine ligand 2 (CCL2), inflammatory cell recruitment, and morphometric analysis of vein wall fibrosis. Mice treated with anti-IL-6 had smaller thrombus weights at day 2, decreased vein wall gene expression and protein concentration of CCL2 at day 2, and impaired vein wall influx of monocytes from days 2 to 6, as compared with controls. Intimal thickness was reduced by 44% (p < 0.05) and vein wall collagen deposition was decreased by 30% at day 14 in the anti-IL-6 group (p < 0.05). Neutralizing IL-6 throughout venous thrombogenesis decreased the production of CCL2, reduced monocyte recruitment, and decreased vein wall intimal thickness and fibrosis. These results suggest that IL-6 may serve as a therapeutic target to prevent the fibrotic complications seen in PTS.

The Generation of An Aptamer Inhibitor of Murine Von Willebrand Factor (VWF) Mediated Platelet Aggregation

AbstractAbstract 4312Venous thromboembolism (VTE) is a national health concern, with an occurrence of over 900,000 cases per year and over 300,000 deaths per year. The total number of cases of VTE and the incidence of VTE-related deaths exceeds those related to both myocardial infarction and stroke. With an aging population, the incidence of VTE has also been increasing. Current treatment of venous thromboembolism with anti-coagulation is not optimal. There is a risk of bleeding, thrombus extension, pain and swelling as well as a recurrence rate of 3–9%. A significant inflammatory response occurs with venous thromboembolism. This inflammation can influence the extent of thrombosis, vein wall fibrosis and valve damage in the thrombosed vein. In a high percentage of VTE patients a condition of venous insufficiency known as post-thrombotic syndrome (PTS) can develop. PTS is associated with stasis ulceration, dermatitis and edema. Venous thrombogenesis is influenced by platelet (PLT) and leukocyte (WBC) adhesion as well as interactions between these cells. There is growing evidence to suggest that VWF interactions with PLT GPIbα can mediate some of these early events. This is evidenced by the reduction in PLT/WBC recruitment and reduced thrombus growth seen in either VWF or GPIbα deficient mice. These data point to a role for VWF in VTE. We sought to develop an aptamer to mouse VWF that would inhibit its interactions with platelet GPIbα. The availability of this tool would support investigations into the role of VWF in mouse models of VTE.Aptamers are oligonucleotides that fold into three-dimensional structures and specifically bind to ligands with high affinity. Aptamers bound to proteins can modify and/or inhibit protein-protein interactions. Using an in vitro selection method known as Systematic Evolution of Ligands by EXponential enrichment (SELEX), we generated aptamers that bind to murine VWF (mVWF) from a modified RNA pool. Nine of these aptamers bind to mVWF with single-digit or sub- nanomolar affinity. A subset of these aptamers also binds to human VWF (hVWF). The aptamers that bind to hVWF inhibit platelet adhesion/aggregation in human whole blood. Further in vitro characterization has demonstrated that five of these aptamers specifically inhibit the interaction between mVWF and recombinant human GPIbα, but do not interfere with the binding of mVWF to collagen. These five aptamers were also active in vivo in a FeCl3-induced thrombosis model in mice. Intravenous injection of the anti-mVWF aptamers prolonged time to occlusion from a baseline of 10–15 minutes to either 25–35 minutes or >40 minutes in this model, depending on the aptamer.These results demonstrate that we have identified high affinity aptamers to mVWF that specifically disrupt mVWF binding to platelets and have an antithrombotic effect in an in vivo mouse model of thrombosis. These aptamers will allow us to investigate the role of VWF in murine models of venous thrombosis. This project was supported by Award Number R01HL095091 from the National Heart, Lung, And Blood Institute. Disclosures:Woelfel:Archemix Corporation: Employment. Wagner:Archemix Corporation: Employment. McGinness:Archemix Corporation: Employment. Schaub:Archemix Corporation: Employment.

Vein Wall Remodeling after Deep Vein Thrombosis: Differential Effects of Low Molecular Weight Heparin and Doxycycline

Venous thrombus resolution sets up an early intense inflammatory reaction, from which vein wall damage results. Tissue response to injury includes matrix metalloproteinase (MMP) activation and extracellular matrix protein turnover. This study sought to determine the effect of exogenous MMP inhibition and its potential attenuation of early vein wall injury. Rats received treatment beginning 24 hr after a stasis venous thrombosis by near occlusive ligation and until harvest at day 7. Three groups were evaluated: (1) vehicle saline controls (NaCl), (2) low molecular weight heparin (LMWH; Lovenox, 3 mg/kg daily SQ), and (3) doxycycline (DOXY, 30 mg/kg daily PO). Thrombus size (mg/mm), levels of tumor necrosis factor alpha (TNF alpha) and D-dimer by colorimetric assay, and monocytes counts by immunohistochemistry were assessed. Vein wall assessment included stiffness by tensiometry, interleukin 1beta (IL-1 beta protein levels by enzyme-linked immunosorbent assay, MMP2 and -9 by zymography, and histological analysis of intimal thickness (IT). Comparisons were by t-test to control. p < 0.05 was considered significant. Thrombus sizes were similar at days 2 and 7 for all three groups, while thrombus TNFalpha was increased in 2-day LMWH- and DOXY-treated groups (NaCl = 1.0 +/- 0.8, LWMH = 9 +/- 3, DOXY = 27 +/- 5 pg/mg protein, n = 6-8, p < 0.05) and at 7 days in the DOXY group (NaCl = 3.0 +/- 2.5, DOXY = 23 +/- 4.2 pg/mg protein, n = 5, p < 0.05). Vein wall stiffness at 7 days was less with LMWH treatment, but not with DOXY, compared to controls (NaCl = 0.33 +/- 0.05, LMWH = 0.17 +/- 0.03, DOXY = 0.43 +/- 0.09 N/mm, n = 5-7, p < 0.05). Vessel-wall IL-1 beta was reduced only in the DOXY group at 7 days (NaCl = 26 +/- 3, LMWH = 38 +/- 17, DOXY = 6 +/- 3 pg/mg protein, n = 4-6, p < 0.05), as was the IT score versus controls (NaCl = 2.2 +/- 0.6, LMWH =1.7 +/- 0.3, DOXY = 0.8 +/- 0.20, n = 4-6, p < 0.05). Zymographic MMP9 activity was significantly reduced at 2 days in the LMWH and DOXY groups (NaCl = 85 +/- 24, LMWH = 23 +/- 7( *), DOXY = 13 +/- 5 U/mg protein, n = 6-8, p < 0.05). MMP2 zymographic activity, thrombus monocyte cell counts, and D-dimer activity were not significantly different across groups. Treatment with LMWH or DOXY did not alter the size of deep vein thrombosis, mildly altered thrombus composition, and differentially affected vein wall injury, despite similar reductions in early MMP9 activity. Whether exogenous MMP inhibition affects long-term vein wall fibrosis will require further study.

Targeted Deletion of MMP-2 Inhibits Post-Thrombotic Vein Wall Fibrosis

Targeted Deletion of MMP-2 Inhibits Post-Thrombotic Vein Wall Fibrosis

Molecular characterization of post-thrombotic syndrome

The post-thrombotic syndrome represents a poorly understood and significant vascular health problem. This review focuses on our current understanding of the pathogenesis of post-thrombotic syndrome. We emphasize the cellular and molecular mechanisms that are responsible for the critical components of post-thrombotic syndrome. These include the initiation of deep venous thrombosis, the pathogenesis of elevated venous pressure, and the factors responsible for nonhealing of venous stasis ulcers.

Differential transcriptional activation of matrix metalloproteinase-2 and membrane type-1 matrix metalloproteinase by experimental deep venous thrombosis and thrombin

Resolution of deep venous thrombosis (DVT) is involved in the pathogenesis of postthrombotic syndrome. Matrix metalloproteinases (MMPs) are a family of proteolytic enzymes that are critical in angiogenesis and tissue remodeling. We hypothesized that MMP-2 and its membrane-bound activator membrane type-1 matrix metalloproteinase (MT1-MMP) expression would be expressed and activated during the resolution of DVT. DVT was generated by caval ligation in wild-type and MMP-2 transgenic reporter mice. Ligated and sham-operated (control) cavae were analyzed for MMP-2 transcription (beta-galactosidase activity in MMP-2 reporter mice) and MT1-MMP mRNA by real-time polymerase chain reaction. MMP-2 activity was determined by zymography, and immunohistochemical staining for beta-galactosidase and MT1-MMP protein was used to localize expression. Human umbilical vascular endothelial cells (HUVEC) were treated with 10 U/mL thrombin and MMP-2 and MT1-MMP mRNA levels and MMP-2 activity was determined. MMP-2 activity increased 71% (n = 5, P < .05) at day 8 in ligated vs control cavae by zymography. beta-galactosidase activity showed a 1.2-fold (n = 8, P < .05) and 1.7-fold (n = 8, P < .05) induction in MMP-2 transcription at day 3 and day 8, respectively. No significant MT1-MMP gene induction was seen at day 3 in ligated vs control cavae, but MT1-MMP mRNA was upregulated 2.5-fold (n = 8, P < .05) in ligated cavae at day 8. Immunohistochemical staining localized MMP-2 and MT1-MMP expression to the vein wall and cellular infiltrates of the thrombus. Thrombin-treated HUVEC showed differential responses of MMP-2 and MT1-MMP. Zymography of conditioned media and cell lysates illustrated a 220% (152.6 +/- 8.6 vs 69.445 +/- 5.46 pixels/unit area, n = 5, P < .05) and 150% (74.1 +/- 7.3 vs 49.2 +/- 5.7 pixels/unit area, n = 5, P < .05) increases in MMP-2 activity respectively. MMP-2 mRNA levels were downregulated 30% (0.48 +/- 0.023 vs 0.63 +/- 0.035 copies of MMP-2 mRNA/copy GAPDH, n = 5, P < .05), whereas MT1-MMP message was upregulated 250% (0.147 +/- 0.009 vs 0.059 +/- 0.005 copies of MT1-MMP mRNA/copy GAPDH, n = 5, P < .05). Resolution of DVT is associated with increased MMP-2 transcription and activity as well as MT1-MMP expression. Thrombin may mediate the increase in MT1-MMP noted in DVT. This is the first article studying MMP-2 and MT1-MMP transcription in DVT. These findings add DVT resolution to the class of inflammatory and fibrotic disorders in which transcriptional activation of the MT1-MMP/MMP-2 genes occurs and identify a potential therapeutic target to modulate this clinically relevant process. Postthrombotic syndrome remains a significant clinical problem after deep venous thrombosis (DVT), but the cellular and molecular mechanisms involved in thrombus resolution and vein wall fibrosis remain undefined. Matrix metalloproteinase (MMP) enzymes are critical to cell migration and matrix breakdown. We identify gene transcription and activity of two MMP isoforms, MMP-2 and MMP-14 (membrane type MMP 1, MT1-MMP) in the resolution phase of experimental DVT and in thrombin-treated endothelial cells. These studies define new proteases potentially important to resolution of DVT and development of postthrombotic syndrome.

Deep venous thrombosis causes vein wall fibrosis

Abstract Introduction: The most common sequelae of deep venous thrombosis (DVT) is chronic venous insufficiency (CVI), which is marked by vein wall scarring. The aim of this study was to characterize the biomechanical and biochemical effects of DVT on the vein wall. Methods: Rats underwent sham laparotomy or inferior vena cava (IVC) ligation to induce DVT, with sacrifice at 2 or 14 days. At sacrifice, venous injury was assessed by laser Doppler microflowmetry, microtensiometry, ELISAs for TNFa, TGFb, MCP-1, RANTES, and direct assays for collagen and glycosaminoglycans (GAG). Comparisons were by one-way ANOVA with P Results: IVC weight increased 43% at 2 days and 114% at 14 days compared with shams (N = 4, P Conclusions: DVT causes vein wall injury that is similar to other fibrotic processes and may be amenable to similar treatments.

P-selectin inhibition decreases post-thrombotic vein wall fibrosis in a rat model

Background. Post-deep vein thrombosis (DVT) venous insufficiency is a vexing problem despite effective anticoagulation, and is characterized by vein wall fibrosis. This study tested the hypothesis that P-selectin inhibition would decrease post-thrombotic vein wall fibrosis and associated profibrotic mediators. Methods. A rat stasis model of DVT was used to produce a 2-day-old DVT. Rats then received either intravenous saline (control), rPSGL-Ig (4 mg/kg) once, or daily subcutaneous low molecular weight heparin (LMWH) (0.5 mg/kg). Inferior vena cava wall was harvested 7 days after treatment and processed for thrombus size; leukocyte content; profibrotic mediators by enzyme-linked immunosorbent assay; collagen I and III mRNA expression by semiquantitative real-time polymerase chain reaction; and for collagen protein. Results. Thrombus mass and leukocyte counts were similar between the groups. Treatment with rPSGL-Ig and LMWH resulted in less vein wall collagen (P <.05). rPSGL-Ig treatment (and a similar trend for LMWH) was associated with decreased profibrotic mediators, including less IL-13, MCP-1, bFGH, and transforming growth factor-β (P <.05). Clollagen III gene expression, but not collagen I gene expression, was increased with LMWH treatment (P <.05). Conclusions. P-selectin inhibition with rPSGL-Ig or LMWH decreases post-DVT vein wall fibrosis, and is associated with decreased vein wall profibrotic mediators. This effect is independent of thrombus mass and vein wall leukocytes. (Surgery 2003; 134:365-71.)

P-selectin inhibition enhances thrombus resolution and decreases vein wall fibrosis in a rat model.

The purpose of this study was to compare the efficacy of P-selectin inhibition with standard anticoagulant and thrombolytic therapy in a rodent model of established deep vein thrombosis (DVT). Rats underwent temporary inferior vena cava (IVC) ligation for 2 days to create a stasis-induced thrombosis. On day 2, the animals had the IVC ligature removed and received either recombinant P-selectin glycoprotein ligand-Ig (rPSGL-Ig; 4 mg/kg) intravenously, low-molecular weight heparin (LMWH; 450 IU/kg) subcutaneously, tissue plasminogen activator (tPA; 0.5 mg/kg) intravenously, combination rPSGL-Ig plus tPA, or saline vehicle. IVC segments were harvested from rats at 4 (n = 8) and 7 (n = 3) days after treatment. All treatments were given as a single dose except for daily LMWH. Evaluation included contrast venography with computer image analysis, thrombus weight/length (mass), vein wall leukocyte counts, cytokine and tissue factor analysis with enzyme-linked immunosorbent assay, and (ED1) monocyte immunohistochemical staining. Collagen was estimated with a quantitative assay. Contrast venography revealed that rats with both rPSGL-Ig and tPA treatment had significantly smaller thrombi as compared with controls at day 7 (0.34 +/- 0.07 cm(2) and 0.34 +/- 0.05 cm(2) versus 0.68 +/- 0.13 cm(2); P <.05). LMWH and tPA groups had significantly decreased thrombus mass at harvest compared with controls on day 4 (0.06 +/- 0.009 g/cm and 0.08 +/- 0.01 g/cm versus 0.1 +/- 0.005 g/cm; P <.05), and rPSGL-Ig showed a similar trend (P =.072). Vein wall, but not thrombus, monocytes were more numerous in those rats receiving rPSGL-Ig versus controls at day 4 (30 +/- 4 cells/5 high power fields [HPFs] versus 19 +/- 2 cells/5 HPFs; P <.05) and at day 7 (32 +/- 2 cells/5 HPFs versus 20 +/- 3 cells/5 HPFs; P <.05). rPSGL-Ig treatment was associated with significantly reduced vein wall collagen at day 7 versus controls (1.3 +/- 0.6 pg/mg versus 3.7 +/- 0.5 pg/mg; P <.05) and a trend toward lower tissue factor levels. rPSGL-Ig, LMWH, and tPA showed equal DVT resolution efficacy over 7 days. However, only rPSGL-Ig was associated with a decrease in vein wall fibrosis, suggesting that purely accelerating DVT resolution may not decrease long-term vein scarring.

Insights in the Development of Primary Venous Reflux

Purpose: The relation between vein wall fibrosis and presence of primary venous reflux was investigated. Methods: Twenty-eight limbs of 27 patients undergoing coronary bypass surgery without history of superficial or deep venous thrombosis and without clinical evidence of varicose veins were subjected preoperative ultrasonic investigation of the long saphenous vein (LSV) and its tributaries in order to detect presence of reflux. Histologic examination of vein specimens, harvested perioperatively from premarked sites of the LSV (ankle, knee, and midthigh) and its tributaries, was carried out. The specimens were classified according to the fibrotic content of the wall on histology as normal, mildly, moderately, and severely affected. Results: Sixteen limbs (57%) had reflux. Saphenofemoral junction (SFJ) incompetence alone was not seen, and this always was present in association with an incompetent LSV (3/16, 19%). Reflux was equally distributed (37.5%) in the above(6/16) and belowknee (6/16) segments, whereas in three limbs it was present throughout the LSV. Fibrosis was seen in the majority of the specimens taken from the LSV (62/65, 95%) and its tributaries (21/32, 66%), irrespective of the presence or absence of reflux. Different grades of fibrosis were found on histology in different sites of the same vein. Conclusions: Venous wall fibrosis very often exists in the absence of reflux and venous insufficiency occurs at any segment of the LSV with minor involvement of the SFJ. Thus, primary venous reflux seems to be a locally developing process.

Intravascular ultrasound imaging of saphenous vein grafts in vitro: Comparison with histologic and quantitative angiographic findings

The ubiquity of coronary artery disease and the resultant widespread use of saphenous veins for coronary artery bypass surgery has stimulated considerable interest in the morphologic and pathophysiologic alterations these vessels undergo after implantation. This study was undertaken to determine the ability of intravascular ultrasound to identify and characterize abnormalities in saphenous vein grafts. Ten saphenous vein grafts excised at autopsy and nine saphenous vein segments harvested during coronary artery bypass surgery were examined with intravascular ultrasound imaging, quantitative coronary angiographic techniques and histologic analysis.Intravascular ultrasound lumen measurements were strongly correlated with quantitative coronary arteriographic measurements (r 0.91, SEE 0.5 mm). Wall thickness was significantly greater in the vein grafts after long-term implantation than in the freshly harvested veins (average thickness 1.4 ± 0.5 vs. 0.7 ± 0.2 mm, p < 0.007); this finding correlated histologically with vein wall fibrosis. There was good correlation between ultrasound imaging and histologic analysis, with the ability to distinguish among normal intima, intimal hyperplasia, vein wall fibrosis and atheromatous plaque.Thus, this preliminary study demonstrates the ability of intravascular ultrasound to provide real-time cross-sectional images of saphenous veins and morphologic characterization of their walls. This modality may have important clinical applications, including the ability to detect serial changes in vein graft intimal hyperplasia and atherosclerosis.

Pathologic findings related to the lead system and repeated defibrillations in patients with the automatic implantable cardioverter-defibrillator

The purpose of the present study was to examine at autopsy the effect of multiple defibrillations on the myocardium and the pathologic consequences of short- and long-term placement of the intravascular and interpericardial leads of the automatic implantable cardioverter-defibrillator. Twenty-five patients were examined at autopsy; 8 of them underwent lead implantation only and 17 received both leads and the automatic implantable cardioverter-defibrillator. Twelve patients (48%) died of ventricular tachycardia or ventricular fibrillation; seven (28%) died of other causes. Acute pericarditis occurred in all patients, resulting in a localized, progressive fibrosis around the apical patch lead without giving rise to pericardial restriction. Thrombus formation was associated with the superior vena cava spring electrode in four patients (17%) and the right ventricular rate-sensing electrode in one patient (4%). Asymptomatic pulmonary emboli occurred in two patients (8%). In one patient who underwent defibrillation 59 times, superior vena cava changes consisted of vein wall destruction, fibrosis and thrombus formation. Pathologic changes under the apical patch related to defibrillation were observed in seven patients; two of these had fewer than 5 defibrillations, one had 8 defibrillations and four had 21 to 74 defibrillations. These changes consisted of contraction band necrosis in four patients, vacuolar cytoplasmic clearing and loss of myocytes confined to the myocardium under the patch electrode in five patients who had multiple defibrillations. The observed pathologic changes were estimated to affect less than 2% of the total myocardial mass. Thus, the automatic implantable cardioverter-defibrillator lead system and multiple defibrillations result in localized myocardial injury confined to the tissue under the patch electrode.(ABSTRACT TRUNCATED AT 250 WORDS)