Assessment and Treatment of Endothelial Dysfunction - 1999
Assessment and Treatment of Endothelial Dysfunction - 1999
Assessment and Treatment of Endothelial Dysfunction - 1999
1999;34;631-638 This information is current as of November 9, 2011 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://content.onlinejacc.org/cgi/content/full/34/3/631
Journal of the American College of Cardiology 1999 by the American College of Cardiology Published by Elsevier Science Inc.
The endothelium plays a key role in vascular homeostasis through the release of a variety of autocrine and paracrine substances (1). In addition to vasodilation, a healthy endothelium is antiatherogenic because of effects that include inhibition of platelet aggregation and adhesion, smooth muscle cell proliferation and leukocyte adhesion. Dysfunction of endothelial cells is a systemic process and the initiating event in atherosclerosis, and is important in the ischemic manifestations of the disease process as well. Endothelial cell dysfunction occurs in the presence of atherosclerosis or its risk factors, particularly hypercholesterolemia (2). Over the past ve years, new methodology has allowed more widespread assessment of endotheliumdependent vasodilation in patients in a variety of research settings. This review will focus on the assessment of endothelial function in humans and the therapeutic options that are now available for treating abnormalities in vascular function.
mediated by endothelin (7), angiotensin II (8) and prostacyclin (9). In the coronary circulation, endotheliumdependent stimulation causes vasodilation of epicardial vessels (10). Acetylcholine-induced vasodilation can be blocked by NO synthase (NOS) inhibition with Nmonomethyl-L-arginine (L-NMMA) (6). Resistance vessel function is determined in vivo by measuring blood ow. Administration of endothelium-dependent agonists result in an increase in blood ow in the human peripheral and coronary circulations (11). Whereas this effect has also been shown to be NO dependent (6), other mediators probably are involved in resistance vessel dilation and the control of coronary blood ow, particularly to metabolic stimuli (12,13). Leukocyte adhesion. Antiinammatory properties of the healthy endothelium are important in the prevention of atherosclerosis and ischemic coronary syndromes (14). Multiple families of adhesion molecules provide trafcking signals for the interaction between leukocytes and the endothelium. A detailed description of the events of monocyte adhesion is reviewed elsewhere (15). Leukocyte adhesion is not only important during the early stages of atherogenesis, but may contribute signicantly to plaque instability and rupture (16). Atherosclerosis and its risk factors result in upregulation of cell adhesion molecules by the nuclear factor-regulating gene expression through redox-sensitive mechanisms (17). This pathway can be mediated by a number of cytokines including interleukins, tumor necrosis factoralpha, monocyte chemotactic protein and interferon. Nitric oxide has been shown to inhibit leukocyte adhesion (18). Soluble forms of adhesion molecules can be measured in blood and serve as markers of endothelial activation (19). It was recently shown in the Physicians Health Study that sICAM-1 (soluble intercellular adhesion molecule-1) was predictive of future myocardial infarction (20). Platelet function. Nitric oxide is antithrombotic via potent antiaggregating and antiadhesive properties (21). Nitric oxide inhibits expression of P-selectin on platelets and suppresses the calcium-dependent conformational change in the glycoprotein (GP) IIb-IIIa receptor required for brinogen binding (22). Platelet activation occurs in conditions associated with impairment of endothelium-dependent va-
632
JACC Vol. 34, No. 3, 1999 September 1999:6318 Table 2. Function of the Healthy Endothelium
Abbreviations and Acronyms ACE angiotensin-converting enzyme EDHF endothelium-derived hyperpolarizing factor FMD ow-mediated vasodilation GP glycoprotein LDL low-density lipoprotein L-NMMA NG-monomethyl L-arginine NO nitric oxide NOS nitric oxide synthase
Substances NO, EDHF, PGI2, ET-1, Ang II, TXA2 PGI2, TXA2, vWF, brinogen, thrombomodulin, TF t-PA, PAI-1 P and E selectin, VCAMs and ICAMs, NF- Integrins RAGE VEGF, PDGF, TGF-
sodilation (23). Platelet function can be assessed by standard aggregation techniques. More recently, ow cytometry has been used to quantify the expression of P-selectin and GP IIb/IIIa receptors, markers of platelet activation (24). Platelet activation is particularly important in unstable coronary syndromes and coronary intervention where blockade of the GP IIb/IIIa receptor is becoming a mainstay of therapy (25). Because P-selectin is present on both platelets and the endothelium, it plays an important role in the interaction with platelets, leukocytes and the endothelium (26). It was recently demonstrated that platelets from patients with unstable coronary syndromes release less NO (27). The release products of activated platelets mediate vasoconstriction at sites of endothelial dysfunction (28). Other functions of the healthy endothelium are shown in Table 2.
Ang II angiotensin II; EDHF endothelium-derived hyperpolarizing factor; ET-1 endothelin-1; ICAM intercellular adhesion molecule; NF- nuclear factor kappa beta; NO nitric oxide; PAI-1 plasminogen activator inhibitor; PDGF platelet-derived growth factor; PGI2 prostacyclin; RAGE receptor for advanced glycosylated end-products; TF tissue factor; TGF- transforming growth factor-beta; t-PA tissue plasminogen activator; TXA2 thromboxane; VCAM vascular cellular adhesion molecule; VEGF vascular endothelial growth factor; vWF von Willebrands factor.
intracoronary Doppler wires (30). Coronary blood ow increases in response to infused agonists, such as acetylcholine, and the magnitude of this increase can be used as a quantitative measure of endothelial function. The change in coronary blood ow is due to both the direct effect of the infused agonist and the resulting FMD. When compared with conduit vessel responses, the measurement of coronary blood ow is somewhat more difcult with greater variability and expense. Coronary blood ow has also been assessed by coronary sinus thermodilution techniques in laboratories equipped to do this. Positron emission tomography. Quantitative assessment of myocardial blood ow and metabolic activity can be made by positron emission tomography scanning (31,32). Both basal ow and hyperemic ow (usually to intravenous dipyridamole) can be obtained to calculate coronary ow reserve (33). Because the increase in myocardial ow is related to adenosine-induced increases and ow-mediated vasodilation, it is in part a measure of endothelial function. This technique is noninvasive and has the advantage of the potential for multiple tests per patient; however, it is very expensive and limited to a small number of laboratories (34,35). Impedance plethysmography. Hokanson et al. (36) described electrically calibrated plethysmography for direct measurement of limb blood ow. The apparatus is relatively inexpensive and versatile because direct intraarterial infusions of methacholine or acetylcholine assess endothelial function (37,38). Because forearm blood ow (ml/min/100 ml) is measured, venous occlusion plethysmography reects resistance vessel function in the forearm. Although ideally suited for one-time measurements, there is some concern about day-to-day variability, and as such, it has not been extensively used for long-term intervention studies.
633
Table 3. Conditions Associated With Impaired Endotheliumdependent Vasodilation Atherosclerosis Hypercholesterolemia Low HDL cholesterol Figure 1. High-resolution ultrasound image of brachial artery at baseline and 1 min after upper-arm occlusion cuff release and subsequent ow-mediated vasodilation. High Lp(a) Small LDL particles Susceptibility of LDL to oxidation Hypertension Hyperhomocysteinemia Aging Vasculitic conditions Transplantation atherosclerosis Syndrome X and variant angina Insulin resistance Type I and type II diabetes mellitus Hyperglycemia Acute postprandial hypertriglyceridemia Active and passive cigarette smoking Dilated cardiomyopathy Chagas disease Heart failureany cause Family history of coronary disease Postmenopausal status Post-Kawasakis disease Pregnancy-induced hypertension/ preeclampsia Pulmonary hypertension Methionine loading
Brachial ultrasound. Celermajer et al. (39) were the rst to describe a noninvasive assessment of ow-mediated vasodilation in the brachial or femoral artery. Upper-arm occlusion for 5 min results in reactive hyperemia after the cuff is released (blood ow increases ve- to sevenfold), and this increase in shear stress results in FMD (Fig. 1). The variability is acceptable (about 2%), and the measurements are reproducible in a good laboratory (40). Brachial artery FMD has been shown to correlate with measures of coronary endothelial function (41). The main advantages of this approach are the noninvasive nature and the ability to repeat multiple tests in the same patient or the study of large numbers of patients. Some laboratories are now using direct tracking ultrasound technology that may improve the precision of the technique (Drexler H, personal communication, 1998). Venous studies. Dorsal hand vein compliance can be assessed with the linear variable differential transducer technique (42). This technique uses local infusions of agonists into the hand veins with an assessment of conduit vessel compliance. In terms of invasiveness and utility, it would be similar to impedance plethysmography, except that venous conduit vessels are studied instead of forearm resistance arterial vessels.
presence of atherosclerosis (11). Part of the dysfunction in atherosclerosis is related to a decrease in NOS activity (48). In balloon-injury models of atherosclerosis, there is emerging evidence that angiotensin II blockade or ACE inhibition decreases intimal proliferation; however, this has not translated into a decrease in restenosis rates in studies of angioplasty in humans. Depending on the animal model, angiotensin II appears to be an important mediator of intimal proliferation and vasomotion (49). Increased endothelin activity also appears to play a role in endothelial dysfunction in atherosclerosis (50). Cardiovascular risk factors. Endothelial dysfunction, a systemic disturbance of function, precedes the physical presence of atherosclerosis (51,52). Oxidative stress appears to play a pivotal role in the alteration of endothelial function that characterizes all risk factors (Fig. 1). Steinberg and others have advanced the concept that oxidative modication of low-density lipoprotein (LDL) is central to the development of atherosclerosis, and recent work has suggested that oxidized LDL plays an important role in abnormal endothelial vasorelaxation (53). The detrimental effects of oxidized LDL on endothelial function are likely modulated through lysophosphatidylcholine (54), protein kinase C (55) and G proteins (56). In addition, oxygen-free radicals impair endothelial function through direct inactivation of NO (57). Oxidized LDL also increases the production of endothelin in cultured cells and intact blood vessels (58). Tetrahydrobiopterin is a cofactor for NOS, the ratelimiting enzyme responsible for the conversion of L-arginine to NO. activation of NOS at suboptimal concentrations of tetrahydrobiopterin leads to the production of hydrogen peroxide instead of NO (59). Relative deciency of tetra-
ENDOTHELIAL DYSFUNCTION
The term endothelial dysfunction is most often used to denote impairment of endothelium-dependent vasodilation, but probably encompasses those conditions leading to endothelial activation with abnormalities in endothelial interactions with leukocytes, platelets and regulatory substances (43 46) (Table 3). Atherosclerosis. It is the general consensus of vascular biologists that endothelial injury with resulting dysfunction is the initiating event in atherosclerosis (47) and plays an important role in the ischemic manifestations of coronary disease. Ludmer et al. (2) were among the rst to recognize this in human coronary arteries. In health, acetylcholine causes vasodilation as a result of endothelium-dependent release of NO; in disease, the effect of NO is decreased, and unopposed muscarinic smooth muscle cell activation leads to vasoconstriction. Atherosclerosis also impairs acetylcholineinduced increases in coronary blood ow, despite the fact that resistance vessels are rarely affected by the physical
634
Table 4. Treatment Associated With Improvement of Endothelial Dysfunction in Humans Acute LDL lowering with pheresis ACE inhibition Antioxidants (Vitamin C and E C) Estrogen L-arginine, D-arginine Tetrahydrobiopterin, methyltetrahydrofolate Deferoxamine Glutathione Calcium channel blockers Chronic LDL lowering with statins, resins ACE inhibition Antioxidants (probucol with lovastatin) Estrogen Estrogen progesterone L-arginine Exercise
nation of LDL lowering and antioxidant therapy on coronary epicardial endothelial function (76). Whereas LDL lowering resulted in an improvement in endothelial function, the combination of lovastatin and probucol resulted in near normalization in endothelial function. The acetylcholine response at follow-up, or the improvement in endothelial response over the study period, was closely related to the resistance of LDL oxidation (63). Whereas we treated patients for one year, Tamai et al. (77) recently demonstrated that forearm blood ow can be modulated within hours after LDL pheresis, demonstrating the dynamic nature of this process. Studies with other modalities to lower cholesterol have shown improved function as well, suggesting that it is the cholesterol-lowering effect and not the pleiotrophic effect of the statins that are important (31,78). Antioxidants. Animal studies by Keaney et al. (79,80) have demonstrated benecial effects of probucol and antioxidant vitamin therapy beyond protection of LDL. Recently, studies have demonstrated that vitamin C acutely improves endothelium-dependent responses in the forearm circulation in patients with a variety of risk factors (81 84). Long-term studies on vascular function in humans with vitamin C have not been reported. Although alpha-tocopherol has been shown to protect LDL against oxidation in humans (85), longer-term studies using antioxidant vitamins have not shown any effect on endothelium-dependent vasodilation (86,87). Recently, two weeks of vitamin E treatment was shown to decrease P-selectin in patients with hypercholesterolemia, suggesting attenuation of endothelial activation (19). Angiotensin-converting enzyme (ACE) inhibition. Angiotensin II has been shown to increase superoxide production via membrane-bound NADH/NADPH (88). Although it is unclear whether angiotensin II blockers result in improved endothelial function, these data are available for ACE inhibitors (8,89). The ACE inhibitors could potentially improve endothelium-dependent vasodilator responses through decreased levels of angiotensin II, increased levels of bradykinin and NO. In humans, acute administration of ACE inhibitors augmented endothelium-dependent vasodilation in both the coronary and peripheral circulation (90,91). Mancini et al. (92) recently reported that the tissue-specic ACE inhibitor quinapril attenuated coronary endothelial dysfunction in patients with coronary artery disease. To assess the potential differences in tissue specicity among ACE inhibitors (93) and the importance of the bradykinin effect, we recently compared quinapril, enalapril, losartan and amlodipine in a cross-over design in 80 patients with coronary artery disease. Over an 8-week treatment period, improvement in brachial artery FMD was only seen in the quinapril group, suggesting potential differences among vasoactive medications (94). Two recent studies using enalapril or lisinopril reported negative results as well
hydrobiopterin has been implied as contributing to endothelial dysfunction in some conditions (60). Hypercholesterolemia induces a number of changes on vascular homeostasis, including a decrease in NO bioactivity, an increase in superoxide production, an increase in endothelin immunoreactivity (61), an increase in adhesion molecules (19) and attenuation of endothelium-dependent vasodilation (37). The combination of diabetes and hypertension appears to have an additive effect on monocyte adhesion (62). It seems that cholesterol-induced endothelial dysfunction is related to the degree of LDL oxidation and not LDL concentration itself (63,64). Acute elevations of free fatty acids and triglycerides can attenuate vasodilator responses over the course of several hours (65,66). Acute hyperglycemia (6 h) can also impair vasomotor responses (67), adding evidence to the belief that it is the milieu of the artery that determines vascular function, and this can be modulated over minutes to hours. Attenuation of endothelium-dependent responses is an important manifestation in many other cardiac and noncardiac conditions. Detailed discussion is beyond the scope of the current review (38,68 72) (see Table 3).
635
Figure 2. Oxidative stress leads to attenuation of endothelium-dependent vasodilation by: 1) decreased production of NO through oxidized LDL-mediated mechanisms, and by 2) increased destruction of NO by superoxide. See text for details. Gi pertussis sensitive G protein; O2 superoxide anion; ONOO peroxynitrite anion; PKC protein kinase C; cNOS constitutive form of nitric oxide synthase; ET endothelin; AT II angiotensin II; ox-LDL oxidized LDL; lyso PC lysophosphatidylcholine; sGC soluble guanylate cyclase; cGMP cyclic guanosine monophosphate; NFKB nuclear factor kappa beta.
(95,96). The effect of lowering blood pressure by other means on endothelial function is less clear, but it appears that ACE inhibitors have unique vascular protective properties. Hormone replacement therapy. Hormone replacement with estrogen has been shown to improve endotheliumdependent vasorelaxation acutely in a number of animal models, including primate coronary arteries (97). It is likely that benet is both NO dependent and independent (98). Sack et al. (99) demonstrated that estrogen attenuates the susceptibility of LDL to oxidation in women. Studies in postmenopausal women demonstrated that acetylcholineinduced coronary vasoconstriction (100,101), but not metabolic vasodilation (102), can be attenuated in as little as 10 min by either intracoronary or intravenous administration of estrogen. Guetta et al. have demonstrated that the acute benecial effects of estrogen on blood ow are NO related because this can be attenuated with L-NMMA (103). Improved endothelial vascular responses have been demonstrated in the peripheral circulation after nine weeks of estradiol therapy (104), and these effects were not attenuated by the coadministration of progesterone (105). Medroxyprogesterone, which is commonly used in combination with estrogen, probably has a detrimental effect on endothelial responses (106). Long-term studies evaluating the effect of hormone replacement on coronary endothelium-dependent vasodilation are underway. Other interventions. Augmentation of NO production by L-arginine supplementation has been shown to improve
vascular relaxation acutely in certain conditions (107). A recent study demonstrated improved brachial artery FMD in hypercholesterolemic subjects after four weeks of oral L-arginine supplementation (108). It is interesting to note that Quyyumi (109) more recently demonstrated augmentation of Ach-induced changes in forearm blood ow acutely with both L- and D-arginine, suggesting other mechanisms of action of arginine. A variety of other interventions that have been shown to modulate vasomotor responses in humans are shown in Table 4. Most of these were administered acutely, and further studies are required to assess their long-term use.
PERSPECTIVE
The clinical manifestations of coronary disease depend on a multitude of interrelated pathophysiologic processes, of which endothelial dysfunction is only one. A wealth of indirect evidence argues that the endothelium plays a vital role in atherosclerosis and its manifestations. Endothelial dysfunction is generally believed to be the inciting event in atherosclerosis (47), and is probably important in ischemic manifestations as well. We demonstrated that endothelial dysfunction was associated with the development of atherosclerosis as assessed by intravascular ultrasound in patients postcardiac transplantation (110). Those participants with normal vasodilator responses to acetylcholine developed atherosclerosis at a rate one-third that of those with endothelial dysfunction. Studies of cholesterol lowering, antioxidant therapy and ACE inhibition also suggest that atten-
636
uation of endothelial dysfunction provides the link between basic scientic observations and the decrease in cardiac events observed in the megatrials. However, it is not clear at this stage whether this is causal and if improvement in endothelium-dependent vasodilation is clinically relevant. However, we are currently at a stage where the importance of measuring endothelial function in individual patients is unknown. The noninvasive technique is well suited for group studies in laboratories that are expert in its application and measurement (Fig. 2). Even in these laboratories, the day-to-day variability is at least 20% to 25% (just as it is with other measures of endothelial function). In laboratories that have not done a lot of testing and do not use careful off-line measurements, the variability will be at least twice that. Thus, before we openly embrace endothelial function testing, signicantly more research has to be done. Standard methodology needs to be established. Large population studies over the next decade are required to determine if a single measure of vasoreactivity in an individual patient predicts the development of atherosclerosis or its complications, and such studies are now in the planning stages. In addition, the relationship between vasomotor responses and other measures of endothelial activation (platelet and leukocyte markers) needs to be established. Finally, intervention studies would need to be done to determine if intervening is prudent and effective in those patients at increased risk because of vascular dysfunction. Whether endothelial dysfunction in a primary prevention population can be used as a marker for participants at higher risk of developing atherosclerosis is yet to be proved, and this remains a very important goal of the clinical vascular biologist over the next two decades.
CONCLUSION
Endothelial function can now be readily measured in humans and is a very useful research tool to assess the effect of risk factors and their treatment on vascular function. A growing list of therapeutic modalities have been shown to modulate endothelial dysfunction, which has important implications for the treatment of participants at risk of developing atherosclerotic complications. For measurement of endothelial function to become a clinically useful tool, much work needs to be done. However, it is probable that endothelial function testing will assume a prominent role in the evaluation and treatment of patients at risk of developing coronary atherosclerosis and its sequelae.
Reprint requests and correspondence: Dr. Todd J. Anderson, 8th Floor, Foothills Hospital, 1403-29th Street NW, Calgary, Alberta, Canada T2N-2T9. E-mail: todd.anderson@crha-health. ab.ca.
REFERENCES
1. Moncada S, Higgs A. The L-arginine-NO pathway. N Engl J Med 1993;329:200212.
637
51.
55.
65.
70. 71.
72.
Mechanical pressure and stretch release endothelin-1 from human atherosclerotic coronary arteries in vivo. Circulation 1997;95:357 62. Reddy KG, Nair RN, Sheehan HM, Hodgson JM. Evidence that selective endothelial dysfunction may occur in the absence of angiographic or ultrasound atherosclerosis in patients with risk factors for atherosclerosis. J Am Coll Cardiol 1994;23:833 43. Anderson TJ, Gerhard MD, Meredith IT, et al. Systemic nature of endothelial dysfunction in atherosclerosis. [Review]. Am J Cardiol 1995;75 Suppl:71 4B. Steinberg D. Antioxidants and atherosclerosis: a current assessment. Circulation 1991;84:1420 5. Kugiyama K, Ohgushi M, Sugiyama S, et al. Lysophosphatidylcholine inhibits surface receptor-mediated intracellular signals in endothelial cells by a pathway involving protein kinase C activation. Circ Res 1992;71:1422 8. Ohgushi M, Kugiyama K, Fukunaga K, et al. Protein kinase C inhibitors prevent impairment of endothelium-dependent relaxation by oxidatively modied LDL. Arterioscler Thromb 1993;13:1525 32. Liao JK, Clark SL. Regulation of G-protein alpha i2 subunit expression by oxidized low-density lipoprotein. J Clin Invest 1995; 95:1457 63. Rubanyi GM, Vanhoutte PM. Superoxide anions and hyperoxia inactivate endothelium-derived relaxing factor. Am J Physiol 1986; 250:H8227. Boulanger CM, Tanner FC, Bea ML, Hahn AW, Werner A, Luscher TF. Oxidized low density lipoproteins induce mRNA expression and release of endothelin from human and porcine endothelium. Circ Res 1992;70:11917. Heinzel B, John M, Klatt P, Bohme E, Mayer B. Ca2/calmodulindependent formation of hydrogen peroxide by brain nitric oxide synthase. Biochem J 1992;281:62730. Stroes E, Kastelein J, Erkelens W, et al. Tetrahydrobiopterin restores endothelial function in hypercholesterolemia. J Clin Invest 1997;99: 41 6. Lerman A, Webster MWI, Chesebro JH, et al. Circulating and tissue endothelin immunoreactivity in hypercholesterolemic pigs. Circulation 1993;88:2923 8. Tsao PS, Niebauer J, Buitrago R, et al. Interaction of diabetes and hypertension on determinants of endothelial adhesiveness. Arterioscler Thromb Vasc Biol 1998;18:94753. Anderson TJ, Meredith IT, Charbonneau F, et al. Endotheliumdependent coronary vasomotion relates to the susceptibility of LDL to oxidation in humans. Circulation 1996;93:164750. Heitzer T, Yla-Herttuala S, Luoma J, et al. Cigarette smoking potentiates endothelial dysfunction of forearm resistance vessels in patients with hypercholesterolemia: role of oxidized LDL. Circulation 1996;93:1346 53. Plotnick GD, Corretti MC, Vogel RA. Effect of antioxidant vitamins on the transient impairment of endothelium-dependent brachial artery vasoactivity following a single high-fat meal. JAMA 1997;278: 1682 6. Steinberg HO, Tarshoby M, Monestel R, et al. Elevated circulating free fatty acid levels impair endothelium-dependent vasodilation. J Clin Invest 1997;100:1230 9. Williams SB, Goldne AB, Timimi FK, et al. Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 1998;97:1695701. Johnstone M, Creager SJ, Scales K, Cusco JA, Lee B, Creager MA. Impaired endothelium-dependent vasodilation in patients with insulin-dependent diabetes mellitus. Circulation 1993;88:2510 6. Celermajer DS, Sorensen KE, Georgakopoulos D, et al. Cigarette smoking is associated with dose-related and potentially reversible impairment of endothelium-dependent dilation in healthy young adults. Circulation. 1993;88(Pt 1):2149 55. Celermajer DS, Adams MR, Clarkson P, et al. Passive smoking and impaired endothelium-dependent arterial dilation in healthy young adults. N Engl J Med 1996;334:150 4. Celermajer DS, Sorensen K, Ryalls M, et al. Impaired endothelial function occurs in the systemic arteries of children with homozygous homocystinuria but not in their heterozygous parents. J Am Coll Cardiol 1993;22:854 8. Anderson TJ, Meredith IT, Uehata A, et al. Functional signicance
638
73. 74.
92.
93. 94.
75. 76.
95. 96.
97.
80.
81.
87. 88.
89.
90. 91.
Assessment and treatment of endothelial dysfunction in humans Todd J. Anderson J. Am. Coll. Cardiol. 1999;34;631-638 This information is current as of November 9, 2011
Updated Information & Services References including high-resolution figures, can be found at: http://content.onlinejacc.org/cgi/content/full/34/3/631 This article cites 108 articles, 60 of which you can access for free at: http://content.onlinejacc.org/cgi/content/full/34/3/631#BIBL This article has been cited by 47 HighWire-hosted articles: http://content.onlinejacc.org/cgi/content/full/34/3/631#otherarticl es Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://content.onlinejacc.org/misc/permissions.dtl Information about ordering reprints can be found online: http://content.onlinejacc.org/misc/reprints.dtl
Citations
Reprints