Epidemiological and observational evidence suggests associations among OA and systemic and skeletal vascular pathology and hypercoagulability [32]. Patients with OA of the hip exhibit a high prevalence of vascular comorbidities. In a cohort of 1000 patients undergoing hip replacement, 550 (55 %) had a cardiovascular, peripheral vascular, or respiratory comorbidity [1]. Of 660 cases of unilateral, atraumatic OA, 65 % had hypertension, 40 % had arteriosclerotic cardiovascular disease, and 10 % had arteriosclerotic peripheral vascular disease. Seventy-eight percent had elevated BMI and this was associated with the prevalence of comorbidities (p < 0.001). Other studies as well have indicated associations among obesity, vascular disease, and OA [33–35]. One study reported that among 241 patients with OA, 75 % had clinically relevant cardiovascular symptoms [2].

An association between OA and atheromatous vascular disease has been suggested [3]. A study of 5650 individuals found an association in women between OA of the knee and carotid wall thickness with an adjusted odds ratio of 1.7 (p < 0.05) and a lesser, but significant, association between OA of the hand and carotid atheromatous plaque in women with an adjusted odds ratio of 1.4 (p < 0.001) for distal interphalangeal joint, and 1.5 (p < 0.05) for metacarpophalangeal joint, OA. No associations were demonstrated in men and other associations, e.g. hip OA, were not observed [36].

Of interest in peripheral vascular disease is intima-media wall thickness (IMT) as a measure of vascular damage, and flow patterns, expressed as volume (ml/min) or peak systolic velocity (cm/s). Our clinic studied 40 lower extremities in 20 patients with established OA of the knee with Doppler ultrasound for arterial blood flow volume and velocity (Fig. 3.1). Color Doppler sonography was performed bilaterally with a GE Logiq 9, 9 MHz transducer. Color Doppler imaging was used to measure common femoral, superficial femoral, and popliteal artery flow volume (ml/min) and flow velocity (cm/s). Common femoral, superficial femoral, and popliteal IMT was measured with B-mode duplex scanning as an assessment of arterial damage. The IMT was defined by two parallel echogenic lines that corresponded to the lumen-intima and the media-adventitia interfaces. An increased IMT is a hallmark of arterial wall disease associated with atherosclerosis.

Reported studies of arterial wall thickness have been inconsistent. One study found no changes in femoral or popliteal wall thickness with OA [37]. Another study reported an association between generalized OA and popliteal artery wall thickness. This study, of 42 patients with symptomatic generalized OA, defined as two or more different sites (hand, spine, hip, knee), and 27 individuals without symptoms or radiographic OA, examined popliteal artery wall thickness by magnetic resonance imaging (MRI). Compared to a mean arterial wall thickness of 0.96 mm. in normal individuals, OA patients exhibited a mean wall thickness of 1.09 mm [38]. Our data is consistent with this study showing a mean (± SEM) popliteal arterial wall IMT of 1.8 ± 0.1 mm in patients with OA of the knee.

Observations of arterial flow patterns of the lower extremity have been reported in 39 female patients with OA of the knee with variable but suggestive results [37]. Notably, peak systolic velocity and flow volume were higher than normal in the external iliac and superficial femoral arteries but not in the common femoral arteries in patients with OA. Flow measured by velocity or volume was not different between OA and normal individuals in vessels distal to the femoral artery. Vessel diameter was smaller in OA compared to controls without OA in the popliteal, anterior tibial, and posterior tibial arteries. Data from our clinic has demonstrated significant increases in common femoral, superficial femoral, and popliteal artery flow volumes in patients with OA of the knee (Table 3.2). Although these data are emerging and confirmatory studies are needed, these initial studies suggest an association of structural vascular disease with OA.

Hypercoagulation has been suspected in OA due to both thrombophilia and hypofibrinolysis. In patients with OA an increase in factor VIII, elevated D-dimer, euglobulin lysis times, and PAI-1 have been described in association with relative hyperlipidemia consisting of hypercholesterolemia and hypertriglyceridemia [40]. These observations are consistent with a pro-thrombotic state in OA. Elevated plasma lipids enhance platelet sensitivity to aggregating agents and this has been proposed as a significant pathway for increased PAI-1 activity in OA [40]. PAI-1 blocks the cleavage of plasminogen to plasmin and is the major regulator of fibrinolysis. Systemic elevations in PAI-1 observed in OA are associated with systemic hypofibrinolysis and contribute to hypercoagulability.

To examine the relationship of obesity to activation of the PAI-1 gene, genetically obese diabetic mice lacking the PAI-1 gene were used to test the hypothesis that elevated PAI-1 contributes to the hyperglycemia, hyperinsulinemia, and insulin resistance associated with the obese and diabetic phenotype [41]. Lack of PAI-1 reduced adiposity and improved hyperglycemia and hyperinsulinemia associated with obesity. Lack of PAI-1 also reduced TNFα expression. TNFα is a major inflammatory mediator in OA. Since TNFα can be produced by adipose tissue, these observations provide substantial linkages among obesity, hypofibrinolysis, and inflammation, and may be relevant to the association of MS and OA.

Inflammation and coagulation are intimately linked and activation of both coagulation and fibrinolytic pathways in arthritic joints are linked to articular inflammation [42]. Fibrinolysis is activated during joint inflammation as shown by a correlation of CRP, D-dimer levels, and plasminogen activators [42]. The inflammatory cytokine, IL-1, a key cytokine in the pathophysiology of OA, induces thrombin formation and increases PAI-1 [7]. Coagulation of whole blood is a stimulus for the production of IL-1 and this may be a mechanism by which thrombosis promotes inflammation [43]. Fibrin and fibrinogen degradation products have been shown to induce the expression of cytokines important in the pathogenesis of OA including IL-1 and TNFα [7, 44]. Inflammatory and coagulation proteins have deleterious effects on articular cartilage. Thrombin induces aggrecan release from both normal and OA cartilage [45]. Components of the fibrinolytic pathway, plasminogen, plasminogen activators, and plasmin have been found in increased concentration in OA cartilage and are capable of activating MMPs and contributing to cartilage matrix breakdown [46]. Fibrin formation during inflammation is also catabolic to cartilage matrix.

Aside from direct effects of coagulation and inflammatory proteins on joint tissues, it has been speculated that changes in the physicochemical environment of subchondral bone may have direct regulatory effects on OA osteoblasts and that bone may emerge as a therapeutic target [25, 32]. Alterations in the circulation of OA subchondral bone consist of venous stasis and intraosseous hypertension and have been termed a “venous outlet syndrome” [47]. Decreased perfusion results in local hypoxia and reduced interstitial fluid flow. Osteoblasts are sensitive to their physicochemical environment and, in response to changes in pressure, fluid flow, and oxygen concentration, alter their expression of cytokines relevant to OA [48–50]. Notably, OA osteoblasts can degrade cartilage matrix and alter chondrocyte phenotype [51–53]. OA chondrocytes co-cultured with OA osteoblasts express less SOX9, type-2 collagen, and PTHrP, downregulate aggrecan, and upregulate MMP expression [54, 55]. In these ways, the effects of changes in arterial and venous circulation, coagulation proteins, inflammatory cytokines, and adipokines may be mediated by OA osteoblasts and play pathophysiologic roles in OA providing mechanistic linkages among obesity, hypercoagulability, and inflammatory components of the MS and OA.

We have studied hypercoagulation due to both thrombophilia and fibrinolysis in 40 patients with histologically proven OA of the hip compared to an age, gender, and BMI-matched cohort of 25 patients without OA. Proteins C and S, resistance to activated protein C, factor V_Leiden, lipoprotein A, antiphospholipid antibodies, lupus anticoagulant and LA (lupus anticoagulant) ratio were measured to assess thrombophilia; PAI-1 and tissue plasminogen activator were used to assess fibrinolysis. The presence of circulating antiphospholipid antibodies, anticardiolipins and lupus anticoagulant is associated with an increased incidence of venous and arterial thrombosis. These antibodies react against β-2 glycoproteins in cell membranes and activate clotting factors on platelet membranes. Mean values for LA ratio were 1.12 ± 0.02 in OA patients compared to 0.5 ± 0.01 in normal subjects (p = 0.001) (Fig. 3.2); the prevalence of abnormal lupus anticoagulant and LA ratio was 8/40 (20 %) in OA patients compared to 0/25 (0 %) in normal subjects (p = 0.005). The prevalence of anticardiolipins in normal subjects was 1/25 (4 %) while OA patients had a prevalence of 12/40 (30 %) (p = 0.002). Measures of risk and precision are displayed in Table 3.3. Impaired fibrinolysis was suggested primarily by elevated PAI-1. Mean levels of PAI-1 were 55.7 ± 6.8 in individuals with OA compared to 30.2 ± 3.6 in normal subjects (p = 0.003) (Fig. 3.2). The presence of elevated PAI-1 and LA ratio was highly specific, and had high positive predictive value, but was not sensitive, for OA. The prevalence of PAI-1 abnormalities was 21/40 (53 %) in individuals with OA compared to 7/25 (28 %) in normal subjects (p = 0.005). Together, these data indicate a greater prevalence of hypofibrinolytic and pro-thrombotic proteins in OA patients compared to normal subjects.