http://surgerybook.net/

TF, are then recruited into the area of developing thrombosis, amplifying the process. Consistently, Pselectin gene deficiency, results in less MP formation, and less thrombosis.41 In an experimental study,

the generation of procoagulant activity was shown to be dependent on P-selectin:PSGL-1 interactions

related to MP formation.42 The procoagulant nature of these MP was demonstrated by their ability to

normalize bleeding in factor VIII–deficient mice.

Figure 6-5. The endothelium is a primary interface allowing both anticoagulant functions in the resting state, with prostacyclin,

NO, plasminogen activators, and thrombomodulin. Procoagulant proteins are expressed on activated endothelium including

selectins, procoagulant proteins, such as the von Willebrand factor, TF, as well as PAI-1.

E-selectin, upregulated after P-selectin, is an important regulator of thrombus formation and fibrin

content in a mouse VT model.43,44 Endotoxin-induced TF-mediated coagulation is enhanced in humans

carrying the S128R E-selectin allele,45 and patients homozygous for the S128R E-selectin allele have an

increased risk for VTE recurrence, highlighting the importance of E-selectin in DVT.46 E-selectin has

been shown to be efficient at raising the affinity and avidity of 2 (CD18) integrins which support

neutrophil trafficking to sites of acute inflammation and recruit platelets and red blood cells.47

PSGL-1 has greatest affinity for P-selectin, and lesser affinity for E-selectin and L-selectin. The role of

P-selectin in VT has been suggested by the study of a mouse with high fourfold higher circulating levels

of P-selectin than wild type.48 These mice are hypercoagulable based on clotting tests, and a receptor

antagonist against the P-selectin receptor (rPSGL-Ig) reverses the hypercoagulability. Consistently, wildtype mice administered soluble P-selectin (sP-sel) become hypercoagulable. In models of VT, P-selectin

inhibition given prophylactically decreases thrombosis in a dose-dependent fashion, and can treat

established VT as effectively as heparin without anticoagulation.49,50

147

http://surgerybook.net/

Figure 6-6. The interaction between stasis injury and procoagulant syndromes are represented by Virchow triad. Endothelial and

vascular injuries cause leukocytes and platelets to express P- and E-selectin and the PSGL-1 receptor. Microparticles are released

which express TF. This stimulates the coagulation pathway, fibrin production, and thrombus amplification.

sP-sel is released from activated platelets and endothelial cells and levels rise significantly during

pathologic conditions.51–53 sP-sel has been studied as a biomarker for DVT,51,54–56 combining Sp-sel with

Wells score (clinical pretest probability of DVT) establishes the diagnosis of DVT. A sP-sel (≥90 ng/mL)

combined with Wells score (≥2), showed a better PPV (91%) than D-dimer (≥500 ng/mL) combined

with Wells score (≥2), which was 69%.57

Other factors of importance to venous thrombogenesis include vWF, a factor that stabilizes

procoagulant factor VIII to promote the initiation and formation of stable thrombus at the site of

vascular inflammation and TF. Using a ferric chloride model of vein wall injury, it has been shown that

occlusive thrombus formation is dependent on vWF.58 Regarding TF, mouse models using genetargeting and bone marrow transplantation technologies found that TF in the vessel wall is more

imporant than TF from leukocytes for thrombus formation when there is no flow in the experimental

model.59 TF has been linked to circulating procoagulant MPs, particularly in cancer patients, cancer is

associated with high levels of DVT, and TF activity is increased in cells treated with chemotherapeutic

agents.60 In addition, cancer patients with VTE were found to have elevated levels of microparticle TF

compared to cancer patients without VTE.61,62

Leukocytes and Thrombosis

Inflammatory cells are important to the process of thrombus recanalization and organization (Fig. 6-

7).40,63 Thrombus resolution resembles wound healing, and involves both profibrotic growth factors,

collagen deposition, matrix metalloproteinase (MMP) expression, and activation.64,65 As the thrombus

resolves, a number of proinflammatory factors are released into the local environment, including

interleukin-1 beta (IL-1β) and tumor necrosis factor-alpha (TNF-α).66 The cellular sources of these

different mediators are probably leukocytes and smooth muscle–like cells within the resolving thrombus

and adjacent vein wall. The fact that leukocytes invade the thrombus in a specific sequence suggests

their importance in the normal thrombus resolution.66

The first cell type in the thrombus is the PMN. Although PMNs may cause vein wall injury, they are

essential for early thrombus resolution by promoting both fibrinolysis as well as collagenolysis.67,68 In a

rat model of stasis VT, neutropenia was associated with larger thrombi at 2 and 7 days, and was

correlated with increased thrombus fibrosis and significantly lower thrombus levels of both uPA and

MMP-9.69 Conversely, in a nonstasis VT model, PMNs may promote thrombosis.14 It is likely the timing

of PMN influx and the amount of perithrombus blood flow that determines prothrombotic or

prothrombolytic activities. Through processes that also involve the initial activation of leukocytes and

platelets, PMNs form neutrophil endothelial traps (NETs), which are extracellular fragments of DNA

containing histones and antimicrobial proteins.14,70 NETs provide a scaffold for thrombus formation and

are prothrombotic.71

In order to investigate if plasma DNA is a surrogate for NETs in patients with DVT, and to determine

correlations with other biomarkers of DVT studied in our laboratory, patients presenting to our

diagnostic vascular laboratory were investigated.72 Patients were divided into positive for DVT by

ultrasound, leg pain but negative for DVT by ultrasound, and control volunteers. Circulating DNA was

significantly elevated in DVT patients, compared with controls. Of interest was a significant positive

correlation of circulating DNA with C-reactive protein, D-dimer, vWF, and the Wells score (all p

<0.01).

The monocyte/macrophage is likely the most important cell for later DVT resolution.73 Monocyte

influx into the thrombus peaks at day 8 after thrombogenesis, and correlates with elevated MCP-1

levels, which has been associated with DVT resolution.74 Targeted deletion of CC receptor-2 (CCR-2 KO)

in the mouse model of stasis thrombosis was associated with late impairment of thrombus resolution,

probably via impaired MMP-2 and MMP-9 activities.75 Indeed, precedent exists in humans that the

monocytic activation state may predict long-term DVT resolution.76

Thrombi have been found to contain increasing amounts of both tPA and uPA activities as they

resolve, and this activity is expressed by invading monocytes.77,78 In a recent study, mice genetically

deleted for uPA had impaired thrombus resolution with reduced monocyte infiltration at the margins of

the thrombus, with few neovascular channels present.79 Mice gene deleted for tPA, however, were not

similarly affected, suggesting that uPA, not tPA, is responsible for this activity.

148

http://surgerybook.net/

Figure 6-7. The proposed resolution mechanism involves both early thrombolysis with a large distending clot and then, over time,

a fibrotic thrombus that resembles scar tissue as produced. Note proinflammatory factors, as well as neutrophils (releasing NETs),

platelets, and MMP are present early with subsequent vein wall injury related to collagenolysis and elastinolysis. Later, vascular

smooth muscle cell proliferation appears to occur, with thrombus resolution with increased profibrotic growth factor such as IL-13

and TGF-β. This promotes collagen accumulation. Both PAI-1 and CCR7 are protective against fibrosis.

The local venous environment is by definition hypoxic, hypoxia-inducible factor-1 alpha (HIF-1α). A

major angiogenic growth factor is HIF-1α. Experimental data in a stasis model of VT suggest that

thrombosis stimulates increased vein wall HIF-1α, and that by exogenous stimulation of HIF-1α

expression, thrombus recanalization was increased and associated with accelerated VT resolution.80

Thus, thrombus resolution is in part dependent on neovascularization. The effect of mechanical stretch,

such as by a VT, may also stimulate HIF-1α as well as MMP-2 and MMP-9, both leading to reduced vein

contractility.81

Recent data linking inflammation to fibrosis demonstrate that inhibition of the inflammatory response

can decrease vein wall fibrosis. In a rat model of stasis DVT, treated with either low–molecular-weight

heparin (LMWH) or an oral inhibitor to P-selectin 2 days after establishment of thrombosis, inhibition

of P-selectin significantly decreased vein wall injury (independent of thrombus size).82 The mechanism

accounting for this protective effect is not yet known, but probably does not involve leukocyte

blockade, because no differences in influx of monocytes into the vein wall were observed.

Loss of venous endothelium likely also contributes to the vein wall fibrosis, as well as the

predisposition to recurrent thrombosis. An experimental model of DVT showed lower expression of

homeostatic endothelial genes such as NO and TM than in controls, which correlated with loss of vWF

positive cell luminal staining.83 Other investigators have found that prolonged venous stasis is

associated with decreased plasminogen activators, probably related to loss of endothelium.84

Associated with the early biomechanical injury from DVT is an elevation of profibrotic mediators,

including transforming growth factor-beta (TGF-b), interleukin-13 (IL-13), IL-6, and monocyte

chemotactic protein-1 (MCP-1).82,85,86 These mediators are present within the vein wall and thrombus

and may drive the fibrotic response. Although exogenous MCP-1 may hasten DVT resolution, it

promotes organ fibrosis in vivo. The profibrotic growth factor TGF-b is also present in the thrombus and

is activated with normal thrombolysis.87 TGF-b may be a key mechanism promoting vein wall fibrosis.

Late fibrosis has been observed in our mouse model of DVT, which demonstrated a significant

increase in vein wall collagen after stasis thrombogenesis.88–90 Based on experimental studies,

elastinolysis and collagenolysis seem to occur early, as measured by an increase in vein wall stiffness,

persisting through 14 days, and are accompanied by elevated MMP-2 and MMP-9 activities.85,89

Correlating with this increase in fibrosis is altered procollagen I and III gene expression, as well as an

increase in MMP-2 and MMP-9 gene expression and activity. Genetic deletion of MMP-2 or MMP-2/9 is

associated with decreased midterm vein wall fibrosis, possibly by modulating vein wall elastin/collagen

metabolism as well as monocyte influx.64,65 It has been demonstrated that decreasing inflammation with

a selectin inhibitor or with neutralizing IL-6 can also decrease vein wall fibrosis.43,91,92 Moreover, the

thrombus size itself does not drive the vein wall injury response; rather the mechanism of thrombosis

seems more important.85,93 Lastly, PAI-1 overexpression is associated with decreased vein wall fibrosis,

in part by decreased MMP-2/9 activity.90

ARTERIAL VERSUS VENOUS THROMBOSIS

149

http://surgerybook.net/

Thrombosis in the arterial system occurs somewhat differently than in the venous system (Fig. 6-8). The

elements required for the initiation of DVT were described by Virchow as stasis, endothelial injury, and

hypercoagulability of the blood. In the arterial circulation, endothelial vascular injury (whether acute or

chronic) is key to thrombosis. This is most clearly demonstrated by the typical atherosclerotic plaque. In

advanced lesions, the lipid core of the plaque is rich in inflammatory cells, cholesterol crystals, and TF

(generated by activated macrophages within the plaque).94 Plaque ulceration exposes highly

thrombogenic lipid to the bloodstream, activating coagulation, platelet aggregation, and leading to the

deposition of clot.95 Platelet deposition occurs at the apex of stenosis, the point of maximal shear force.

The blood flow velocity itself regulates how quickly both thrombi form and dissolve.96

A platelet-rich thrombus is observed in arterial thrombus while in contrast, venous blood stasis and

changes in its composition (leading to hypercoagulability) incite the formation of thrombus from local

procoagulant events, including small endothelial disruptions at venous confluences, saccules, and valve

pockets.11 Hypercoagulable states have classically been highly associated with VTE, but do play a role

in cardiovascular disease as well.97,98 Moreover, HMG-CoA reductase inhibitors (statin) not only

decrease atherothrombotic events, but also incident VTE,99 suggesting some common pathogenic

pathways.

PROCOAGULANT STATES

Acquired Procoagulant States

4 Most thrombotic clinical episodes have an identifiable cause, although environment risks and genetic

predispositions to thrombosis may account for many of the VTE that manifest clinically.11 Risk factors

for arterial thrombosis are primarily related to atherosclerosis, and are detailed elsewhere.

The most common risk factors for VTE are prior DVT, malignancy, immobility, intravenous catheters,

increased age, major surgery, trauma, infections such as pneumonia and urinary tract infection, and

certain chemotherapies (Table 6-1).100–102 Certain medications such as oral contraceptives and hormonal

replacement therapies also increase the risk of VTE.

Of primary importance to surgeons is how best to estimate perioperative VTE risk, and apply

appropriate prophylaxis. This can be done with a screening form, such as that devised by Caprini et

al.103 and Pannucci et al.104 Essentially, this is a focused assessment of VTE risks related to current

illnesses and history that may not be fully covered in the routine history and physical. The

recommendations are also well detailed in the routinely updated American College of Chest Physicians

VTE evidence-based guidelines.105 The higher the risk, the more intensive the prophylaxis, is the

paradigm. For example, an outpatient hernia patient may require no prophylaxis outside of early

ambulation where as a hip replacement in an obese patient would be best treated with anticoagulation

as well as sequential compression devices for the lower extremities. However, all patients should be

assessed for VTE risk.

Malignancy

Cancer represents an independent risk factor for the development of VTE, with a sixfold increased risk

in cancer patients. Importantly, cancer patients undergoing surgical procedures have twice the risk of

development of postoperative DVT.106 Highest rates of VTE are seen in pancreas, stomach, and lung

cancer. The malignancy itself activates a procoagulant phenotype by expression of TF by malignant

cells, release of TF-bearing MPs, and increased activation of neutrophil extracellular traps – the

downstream effects of which have been outlined above.107

Despite considerable evidence supporting the use of prophylactic anticoagulation in cancer patients,

many do not receive optimal pharmacoprophylaxis for DVT while inpatient.108 An extended duration of

therapy – 28 days – was found to be associated with a decreased incidence of DVT at the end of the

study period; consequently, current guidelines recommend extended thromboprophylaxis in cancer

patients after an operative intervention.108

150

http://surgerybook.net/

Figure 6-8. Depicted in A is the typical atherosclerotic thrombotic nidus, which includes rupture of a plaque, composed of smooth

muscle, foam cells, and leukocytes. Platelets are the primary intermediary in arterial thrombosis, as well as TF. In figure B,

increased coagulability, as well as vessel wall changes with procoagulant TF expression promotes thrombosis.

Inflammatory Bowel Disease

Since the first half of the last century, a link between inflammatory bowel disease (IBD) and VTE has

been recognized. While there does appear some correlation between thrombosis and the inflammatory

nature of the disease, the link between the inflammatory and coagulation cascades does not tell the

whole story. There are several possible mechanisms for this increased tendency toward thromboembolic

events under active investigation, including spontaneous platelet activation through the CD40–CD40L

pathway.109

A recent retrospective cohort study of the National Surgical Quality Improvement Program database

demonstrated that the presence of IBD is an independent risk factor for postoperative VTE with an odds

ratio of 2.110 Interestingly, there appears to be an increased tendency to form clot in unusual locations,

such as cerebral and hepatic veins. Currently, guidelines recommend only standard pharmacologic

thromboprophylaxis for patients with IBD.

Trauma

As with other acquired prothrombotic states, the systemic inflammatory milieu is associated with VTE in

trauma patients. Activation of inflammatory pathways through TNF-α with increased circulating TF and

procoagulant MP induces a prothrombotic state.111 Incidence of VTE in trauma patients is around 2%,

and this may be higher in patients with other risk factors, such as old age, obesity, and prolonged

immobility after bone fractures.112,113 LMWH with application of pneumatic compression devices has

been shown to be the preferred thromboprophylaxis in trauma patients, and this is reflected in

anticoagulation guidelines.105 However, optimal dosing regiments are not well defined.

Table 6-1 Acquired Hypercoagulable States

151

http://surgerybook.net/

Lupus Anticoagulant/Antiphospholipid Syndrome (Antiphospholipid Antibody)

Antiphospholipid antibody syndrome is a particularly virulent hypercoagulable state that results in both

arterial thrombosis and VT and consists of an elevated antiphospholipid antibody titer in association

with thrombosis, recurrent fetal loss, thrombocytopenia, and livedo reticularis. Strokes, myocardial

infarction, visceral infarction, and extremity gangrene may also occur. Although the lupus anticoagulant

has been noted often in patients with systemic lupus erythematosus (SLE), it does occur in patients

without SLE. It may also be induced in patients by medications, cancer, and certain infections.114

A number of possible thrombotic mechanisms have been suggested, including inhibition of PGI2

synthesis or release from endothelial cells,115 inhibition of APC by thrombin/TM,116 elevated PAI-1

levels,117 platelet activation,118 endothelial cell activation,119 and interference with the endothelial cell–

associated annexin V anticoagulant activity.120 Increased TF expression on monocytes and low free

protein S plasma levels have also been found with the antiphospholipid syndrome and a history of

thrombosis.121,122

At least one-third of patients with lupus anticoagulants have a history of one or more thrombotic

events, 70% or more being VTE.123 Graft thrombosis has been observed in 27% to 50% of patients

positive for antiphospholipid antibody.124,125

For the diagnosis of antiphospholipid syndrome, both clinical and laboratory criteria are necessary.

For the clinical criteria, ≥1 must be present:

1. ≥1 clinical episode of, objectively confirmed, arterial, venous, or small vessel thrombosis.

2. Pregnancy morbidity: ≥1 unexplained fetal death @ ≥10 weeks EGA; ≥1 premature birth (≤34th

week of gestation) due to eclampsia, severe preeclampsia, or placental insufficiency; ≥3 unexplained

consecutive spontaneous abortions @ <10 weeks EGA.126

For the laboratory criteria, ≥1 must be present:

1. LA (+) ≥2 occasions, at least 12 weeks apart, according to ISTH guidelines (prolonged PL-based

clotting assay, lack of correction with 1:1 mix, and correction with excess PL).

2. ACLA and/or anti-β2 glycoprotein-I antibody (medium or high immunoglobulin G (IgG) and/or IgM

isotype titer ≥2 occasions, at least 12 weeks apart using standardized ELISA assays).

The prolongation in the aPTT is strictly a laboratory phenomenon. The dilute Russell viper venom

time confirms the presence of a lupus anticoagulant.

There is imperfect agreement between diagnostic tests for this abnormality. Approximately 80% of

patients with a prolonged aPTT will have a positive antiphospholipid antibody, but only 10% to 50% of

patients with a positive antiphospholipid antibody will have a prolonged aPTT.127

Heparin followed by warfarin has been recommended for the treatment of the antiphospholipid

syndrome.114,123 For recurrent fetal loss, heparin or LMWH use through pregnancy is recommended. In

patients with lupus anticoagulants, heparin is monitored by antifactor Xa levels.

Heparin-Induced Thrombocytopenia and Thrombosis Syndrome

5 Heparin-induced thrombocytopenia (HIT) occurs in 0.6% to 30% of patients in whom heparin is given,

although severe thrombocytopenia associated with thrombosis (HITTS) is seen much less

frequently.128,129 Approximately one half of HIT patients have thrombosis that is noted clinically or by

duplex imaging. In an analysis of 11 prospective studies, the incidence was reported to be 3%, with

152

http://surgerybook.net/

thrombosis in 0.9%.130 Although earlier morbidity and mortality rates of 61% and 23% had been

reported,116 with early diagnosis and appropriate treatment, these rates have declined to 6% and 0%,

respectively.131 HIT is caused by a heparin-dependent IgG antibody that, when bound to platelet factor

4 (PF4), induces platelet aggregation in part by inducing MP formation.132,133 The antibody may not be

heparin specific, as the degree of sulfonation of the heparin-like compound has been suggested to be

critical for this aggregation.134

Both porcine and bovine UFH as well as LMWH have been associated with HIT.135 The syndrome

usually begins 3 to 14 days after heparin is begun. Both arterial and venous thromboses have been

reported, and even small exposures to heparin (heparin coating on catheters) can cause the

syndrome.128,136 VTE associated with HIT is <1% for LMWH, while it is approximately 12% to 13% for

UFH with a higher risk for mortality.137

The diagnosis should be suspected when a patient experiences a 50% or greater decline in platelet

count, when there is a fall in platelet count below 100,000/mL during heparin therapy, or in any patient

who experiences thrombosis during heparin administration.138 The syndrome may be difficult to

diagnose as many hospitalized patients have multiple reasons for low platelet counts, and vigilance is

important. A platelet count should be checked about every 2 days when on heparin therapy.

The laboratory diagnosis of HIT/HITTS is made by a number of assays. The serotonin release assay

was the “gold standard” in the past. An ELISA test detecting the antiheparin antibody in the patient’s

plasma directed against the heparin–PF4 complex is now commonly used.128 This assay is less specific

but more sensitive and is easier to perform and interpret than the serotonin assay. However,

inappropriate testing is common when HIT is suspected. The 4Ts score is a validated measure of pretest

probability with a high negative predictive value that, used appropriately, can limit expensive overtesting and unnecessarily subjecting patients to cessation of heparin and alternative

anticoagulation.139,140

When the diagnosis is made (clinically), cessation of heparin is mandatory. This includes removing

heparin from intravenous catheters and flushes.136 Warfarin should not be administered until an

adequate alternative anticoagulant has been started to prevent thrombotic complications. A number of

anticoagulants are now available to substitute for patients with this diagnosis. The direct thrombin

inhibitor, argatroban, is FDA approved for this indication and shows no cross-reactivity to heparin

antibodies.138,141 When using argatroban, it is important to keep in mind that the INR is artificially

elevated with this agent. Fondaparinux has also been used for this indication.142

Hypercoagulability Testing: Clinical signs of possible thrombophilia include thrombosis at a young

age (<40), unprovoked thrombosis, recurrent thrombosis, thrombosis at unusual locations such as the

mesentery, CNS venous sinus, and portal vein, a family history of thrombosis, particularly unprovoked,

severe, and in those <50 years of age.143 Testing for hereditary defects in patients with thrombosis

with no family history has pros and cons. On the pro side, testing may: improve understanding of

pathogenesis of thrombosis, identify and counsel affected family members, and obviate expensive

diagnostic testing (e.g., CT scans) looking for a malignancy. On the con side, testing may identify

patients with defects whose management would change, with the potential for overaggressive

management, insurance implications for the patient arise, and the testing is rather costly.

In order to determine the efficacy of hypercoagulable testing, a multicenter international

observational registry on clinical characteristics, treatment patterns, and outcome in consecutive

patients with symptomatic, objectively confirmed acute VTE, was done. In this study, 22,847 patients

were enrolled and 4,503 tested for thrombophilia. Of the total patient population, 8.4% had with factor

V Leiden, 6.8% had PTG20210A, and 3% had activated protein C resistance (APC-R). The authors

concluded that for the low incidence and in accordance with the recommendations made by other

authors and international bodies, “the undertaking of thrombophilia testing on patients with a first

episode of VTE is not advisable.”144

A hypercoagulable screen should include routine coagulation tests such as the aPTT and platelet

count, AT activity and antigen assay, protein C antigen and activity levels, protein S antigen level, and

mixing studies to identify a lupus anticoagulant (if indicated); APC-R assay and factor V Leiden gene

analysis; prothrombin G20210A genetic analysis; homocysteine level; an antiphospholipid antibody

screen that includes anticardiolipin antibody; fibrinogen level; FVIII, FIX, and FXI levels and a

functional plasminogen assay.

Inherited Procoagulant States

Defects with High Risk for Thrombosis

153