INTRODUCTION

Atherosclerosis is a chronic inflammatory vascu-lar disease based on the immune system activ-ity. Therefore, complex interactions between ge-netic and environmental factors of the immune system and the vascular wall could act as modu-lators of the pathogenesis and development of peripheral arterial disease (PAD).

Indeed, white blood cell count (WBC) and in-flammatory markers elevation have been con-sistently associated with cardiovascular events and PAD (1-5). Moreover, there are several data pointing to a possible autoimmune origin for atherosclerosis (6-11).

CORRESPONDENCE

Cesar Varela MDAv. de Aragón 36, 3ºC, 28903, Getafe (Madrid), Spain

Tel: 034-626-98-50-93Fax: 034-91-764-15-90

E-mail: varelot@hotmail.com

CONFLICT OF INTEREST

None

ISSN 2042-4884

ABSTRACT

Background and objectives: Circulating anti-endothelial cell antibodies (AECAs) are elevated in peripheral arterial disease (PAD) patients. In this context, these autoantibodies have been associated with the endothelial dysfunction and the pro-inflammatory status that surrounds atherosclerosis. On the other hand, white blood cell (WBC) count and inflammatory markers have been consistently associated with cardiovascular events and PAD. Our aim is to investigate the relationship between circulating AECAs and WBC count in patients with PAD.

Methods: An observational translational study was conducted including 34 male patients with intermittent claudication and no previous autoimmune disease. WBC count was measured in all patients. Highly sensitive C-reactive protein levels (hsCRP) were investigated as a surrogate of inflammation. Circulating AECAs titer was detected using indirect immunofluorescence.

Results: We found higher hsCRP levels in patients with circulating AECAs (7.60 [4.25-10.20] vs. 4.90 [3.20-7.00] mg/L p=0.02). We also observed a higher monocyte count (0.72 [0.60-1.01] vs. 0.55 [0.45-0.79] X103/µL p=0.03) and eosinophil count (0.28 [0.22-0.36] vs. 0.25 [0.07-0.29] X103/µL p=0.05) in patients with these autoantibodies

Conclusions: Circulating AECAs of PAD patients could be associated with the systemic inflammatory status that surrounds the disease and with high monocyte and eosinophil blood cell count.

Relationship between white blood cell count and circulating anti-endothelial cell antibodies detected in patients with peripheral arterial disease

VASCULAR SURGERY | ORIGINAL RESEARCH

Cesar Varela MD, Joaquin de Haro MD, Leticia Esparza MD, Silvia Bleda MD, Ignacio Lopez de Maturana MD & Francisco Acin Md PhDAngiology and Vascular Surgery Department. Hospital Universitario de Getafe. Crta. Toledo KM 12.5, 28905, Getafe (Madrid), Spain. Date received: 5/5/12, Date accepted: 12/7/12Key words: Anti-endothelial cell antibodies; peripheral arterial disease; monocyte; eosinophil; inflammation DOI: 10.5083/ejcm.20424884.83

In this context, circulating anti-endothelial cell antibodies (AECAs) are a heterogeneous group of circulating antibodies targeting endothelial cells, detected in different vasculitic, inflamma-tory or autoimmune situations, whose common denominator is endothelial damage (12).

AECAs have various physiopathological effects of scarcely known mechanisms, such as induc-tion to apoptosis (13,14), cytotoxicity (15), increase in leukocyte adhesion and cytokine secretion

(16,17), activation of coagulation and thrombosis. In a recent study it has been found an associa-tion between these autoantibodies and PAD in patients with no previous autoimmune disease (18). Our aim is to investigate the relationship be-tween circulating AECAs and WBC count in pa-tients with PAD in order to provide new insights in the field of the etiological pathophysiology of atherosclerosis.

116 EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

117

METHODS

An observational translational study was conducted including male patients with intermittent claudication due to PAD after hae-modynamic confirmation of the disease by Doppler and treadmill exercise testing. None of the patients included had been previously revascularised or presented tissue lesions of the lower limbs. All pa-tients signed an informed consent form, according to the principles of the Helsinki Declaration and the study protocol was approved by our Hospital Ethics Committee.

All those patients with documented diagnosis of atopic, autoim-mune or rheumatologic disorders, transplanted or immunode-pressed, and those treated with immunosupressors or systemic and/or inhaled corticosteroids were excluded. We also excluded all those patients with an acute inflammatory status. Remaining subjects underwent a serological screening for autoimmune, rheu-matologic and neoplasic disease markers. Antibodies ANA, Anti-DNA, Anti-LKM, Anti-mitochondria and Anti-smooth muscle were analysed by means of indirect immunofluorescence and antibodies Anti-SM, Anti-Jo, Anti-LA, Anti-Scl 70, Anti basement membrane, Anti-RO and Anti-RNP were measured using the enzyme-linked Immunosorbent assay (ELISA) test. All patients seropositive to any marker were excluded from the analysis.

During two months, 248 patients were screened at our outpa-tients clinic, but only 34 met the inclusion criteria and were finally recruited for the research. The basal ankle-brachial-index higher value and the absolute claudication distance of the included sub-jects were 0.60 [0.57-0.70] and 175 [126-236] metres, respectively. Cardiovascular risk factors and medical treatment of the patients were recorded.

Basic laboratory, WBC and highly sensitive C-reactive protein measurements

Patients who met the inclusion criteria were called to attend our outpatient clinics after fasting for 12 hours. We obtained a general sample for basic laboratory determinations (glycaemia, electrolytes and renal function), lipid profile and WBC count. Using the same puncture, we also obtained an independent blood sample for plas-ma concentration of Highly sensitive C-reactive protein (hsCRP) and circulating AECAs titer measurements. Plasma concentration of hsCRP was measured using a highly sensitive, automated immu-noassay (Roche Diagnosis, Basel, Switzerland). This test provides a low detection limit of 0.2 mg/l and a variation coefficient of 4.2% in 4 mg/l and 6.3% in 1 mg/l. Each sample was performed in triplicate, and the mean of the three values was used for the analysis.

Circulating AECAs detection

The autoantibody titer was detected by indirect immunofluo-rescence using a diagnosis reagent kit from EUROIMMUN (Med-izinische Labordiagnostika AG, Luebeck, Germany) with a TITER-PLANE technique. This technique provides 100% sensitivity and specificity for AECAs detection. Cultivated umbilical vein endo-thelial cells covered the reaction areas of a BIOCHIP. Slides were incubated with patient’s diluted serum samples. Positive reactions were marked by granular staining in cytoplasm of human umbilical vein endothelial cells with fluorescein-labeled antibodies and were visualised by fluorescence microscopy. (Figure 1). According to the manufacturers, titers of <1:10 represent the reference range and the lower detection limit of the test.

LEUKOCYTES AND ANTI-ENDOTHELIAL CELL ANTIBODIES

EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

Statistical analysis

Data was processed using the SPSS 15.0 statistical software pack-age (Microsoft). Differences between groups were considered sta-tistically significant for a P<0.05 in two-tailed test. The normality of continuous variables was analysed using the Kolmogorov-Smirnov and Shapiro-Wilk tests. The association between categorical vari-ables was studied using the Chi-square test and the Fisher’s Exact test when required. The association between continuous variables was analysed using the Mann-Whitney U test and Spearman’s cor-relation test. Categorical variables were expressed as a percent-age and continuous variables as the median (interquartile range [p25-p75]). All the extreme values and outliers were identified and double-checked.

RESULTS

Circulating AECAs titer >1:10 was detected in the serum sample of 20 (59%) patients. All the autoantibodies were IgG isotype and all the serum samples with AECAs titer >1:10 reacted against beta2-glycoprotein I, proteinase 3 and collagenase antigens. The baseline characteristics of the included patients are described in Table 1.

Relationship between circulating AECAs and WBC

There were no statistical differences in lipid profile or WBC count between the analysed groups (Table 2). However, patients with AE-CAs titer >1:10 showed a higher blood monocyte count (0.72 [0.60-1.01] vs. 0.55 [0.45-0.79] X103/µL p=0.03). Eosinophils count was also higher in patients with AECAs titer >1:10 (0.28 [0.22-0.36] vs. 0.25 [0.07-0.29] X103/µL p=0.05) (Table 2 and figure 2).

Figure 1: Circulating anti-endothelial cell antibodies (AECAs) detection. Positive reaction is defined as granular fluorescence in cytoplasm of cultivated human endothelial cells from umbilical vein.

HEALTHCARE BULLETIN | VASCULAR SURGERY

118 EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

Table 1: Baseline characteristics of the included patients

AECAs titer ≥1:10 % (n=20)

AECAs titer <1:10 % (n=14) p value

Age and Cardiovascular risk factors

Age (years) 63 [60-68] 67 [57-74] 0.98

Diabetes mellitus 5 (25%) 3 (21%) 1

History of Smoking 10 (50%) 6 (43%) 0.73

Coronary heart disease 4 (20%) 4 (28%) 0.68

Hypertension 11 (55%) 7 (50%) 1

Cerebrovascular disease 2 (10%) 0 (0%) 0.50

Chronic renal failure 2 (10%) 1 (7%) 1

Chronic pulmonary disease 1 (5%) 2 (14%) 0.55

Medical treatment

Anti-platelet 15 (75%) 11 (78%) 1

Statins 6 (30%) 4 (28%) 0.61

ACE inhibitors 8 (40%) 5 (36%) 0.54

Nitrates 1 (5%) 2 (14%) 0.55

Beta-blockers 2 (10%) 2 (14%) 1

Calcium antagonist 5 (25%) 4 (28%) 0.56

Beta-agonists 2 (10%) 2 (14%) 1

Oral anticoagulants 0 (0%) 2 (14%) 0.16

Figure 2: Relationship between white blood cell count (WBC) and anti-endothelial cell antibodies (AECAs). Monocyte count (a) and eosinophil count (b) were higher in patients with circulating AECAs titer ≥1:10. *Blood cell count units: X103 /µL.

LEUKOCYTES AND ANTI-ENDOTHELIAL CELL ANTIBODIES

119EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

Table 2: Lipid profile and white blood cell count of the included patients

AECAs titer ≥1:10 (n=20)

AECAs titer <1:10 (n=14) p value

Lipid profile

Cholesterol (mg/dl) 211 [171-251] 203 [151-248] 0.43

Triglycerides (mg/dl) 148 [132-188] 139 [107-167] 0.53

High-density lipoprotein (mg/dl) 45 [38-50] 49 [42-58] 0.53

Low-density lipoprotein (mg/dl) 116 [106-166] 116 [73-173] 0.48

Very low-density lipoprotein (mg/dl) 29 [26-38] 28 [21-33] 0.53

White blood cell count

White blood cell count (X103/µL) 8.26 [5.89-12.32] 7.48 [5.83-10.15] 0.39

Neutrophils (X103/µL) 4.40 [3.63-7.21] 3.75 [2.25-5.49] 0.12

Lymphocytes (X103/µL) 2.62 [1.98-3.61] 1.67 [1.59-3.53] 0.09

Monocytes (X103/µL) 0.72 [0.60-1.01] 0.55 [0.45-0.79] 0.03

Eosinophils (X103/µL) 0.28 [0.22-0.36] 0.25 [0.07-0.29] 0.05

Basophils (X103/µL) 0.06 [0.05-0.08] 0.05 [0.04-0.06] 0.11

Relationship between circulating AECAs and hsCRP

We found higher hsCRP levels in patients with circulating AECAs titer >1:10 (7.60 [4.25-10.20] vs. 4.90 [3.20-7.00] mg/L p=0.02) (Figure 3).

Relationship between hsCRP and blood cell count

We did not find any significant statistical correlation between WBC and hsCRP in our sample.

DISCUSION

According to the findings of this study, the presence of circulating AECAs in patients with PAD could be associated with the systemic inflammatory status that surrounds the disease and with high monocyte and eosinophil blood cell count. Circulating AECAS have been related to PAD. Subjects seropositive to these autoantibod-ies presented worse endothelial dysfunction parameters (Flow-mediated arterial dilatation), higher levels of inflammation markers (hsCRP) and higher Intiman-media thickness measurements (18). This data suggest a possible role of circulating AECAs on the ath-erogenic process.

Endothelial dysfunction acts as a primary pathogenic event for atherosclerosis, as it occurs before structural changes are evident on angiogram or ultrasound scan and it is not correlated with the disease’s severity (19). The loss of endothelial regulation has been attributed to a reduction in nitric oxide (NO) bioactivity, and to an increased oxygen-free radical formation in the context of the pro-inflammatory status found in the disease (20,21).

Figure 3: Relationship between C-reactive protein (hsCRP) and anti-endothelial cell antibodies (AECAs). Levels of hsCRP were higher in subjects with circulating AECAs titer ≥1:10.

HEALTHCARE BULLETIN | VASCULAR SURGERY

120 EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

In this context, circulating AECAs may cause endothelial damage through several mechanisms. Polyclonal IgG fractions of these au-toantibodies can mediate endothelial cell lysis when incubated in vitro with the complement or in the presence of mononuclear cells to induce an antibody-dependent cellular cytotoxicity (15). Circu-lating AECAs can also induce endothelial apoptosis by themselves or together with natural killer cells. Signaling for apoptosis could be initiated through the interaction of these autoantibodies with the Fas receptor or with the integrin α3B1 and α6B1-NAG-2 protein complex (12). Furthermore, some published data have suggested that circulating AECAs might cause endothelial cell apoptosis through the NO pathway (22,23).

On the other hand, Inflammation has a key role in every aspect of the atherosclerotic process. Recent investigations reported that the presence of high-risk carotid plaques is associated with increased levels of inflammatory markers, such as hsCRP and leu-kocyte count (24-26). Another study found that the prevalence of unstable hypoechoic plaques was higher in PAD than in coronary artery disease patients, probably because, consistent with previous studies, the PAD group has more pronounced inflammatory profile (27). Neutrophil count was associated with a more severe clinical presentation of atherosclerosis independently of sex, age and clas-sic risk factors (27).

Relationship between circulating AECAs, monocyte blood cell count and hsCRP levels

In the present research we have observed an association between AECAs titer ≥1:10 and high monocyte blood cell count in patients with PAD. These findings are of great interest, considering that monocytes contribute actively to atherosclerotic lesion develop-ment. The infiltration of circulating monocytes into the artery wall is one of the earliest events in the development of the atheroscle-rotic lesion. These white blood cells are believed to play a critical role in the atherosclerosis initiation, plaque instability and artery re-modeling stages (28-30). Blood monocytes have been also indepen-dently associated with subclinical PAD after adjustment for other inflammatory markers (4).

In this context, circulating AECAs could activate endothelial cells through a NF-κB dependent mechanism, leading to the expression of leukocyte adhesion molecules (intercellular adhesion molecule 1 [ICAM-1], vascular cell adhesion molecule 1 [VCAM-1], E-selectin) and increasing the release of pro-inflammatory cytokines (interleu-kin 1 [IL-1], interleukin 6 [IL-6]) in vitro (16,17). These observations are consistent with the finding of raised hsCRP levels in patients with PAD and circulating AECAs. The relationship found between these autoantibodies and hsCRP levels suggest that autoimmunity could play an important role in the genesis of the chronic inflam-mation observed in patients with PAD. On the other hand, the hypothetical endothelial inflammatory chemotaxis enhancement related to circulating AECAs could justify higher monocyte count levels when the antibody is present.

Relationship between circulating AECAs and eosinophil blood cell count

We also found higher eosinophils count in patients with circulat-ing AECAs. Recent observations suggest that eosinophils may play a role in coronary atherosclerosis (31). Eosinophils count has been associated with an increased risk for future cardiovascular events (32,33). When activated, eosinophils secrete several proteins, among which eosinophil cationic protein (ECP) stands up (34). This mole-cule interacts with other immune cells and plasma proteins such as coagulation factors and the complement system (35). ECP also up-regulates endothelial ICAM-1 expression (36) allowing monocytes adhesion on endothelium. Therefore, eosinophils might cooperate with, or amplify, circulating AECAs immune mechanisms in mediat-ing endothelial damage.

Our findings do not establish a causal relationship. However, the observed association of circulating AECAs with monocytes and eo-sinophils count in patients with PAD supports the hypothesis of an alleged autoimmune origin of atherosclerosis, given the clear rela-tionship of these blood cells with autoimmune diseases and circu-lating AECAs pathogenic mechanisms. Although the data obtained from this study does not have a direct clinical application, its poten-tial benefits lie in the improvement of the etiopathogenic insights of the atherogenic process.

CONCLUSION

The presence of circulating AECAs in patients with PAD could be as-sociated with the systemic inflammatory status that surrounds the disease and with high monocyte and eosinophil blood cell count. However, further prospective studies are required to establish a causal relationship.

REFERENCES

Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: a meta-analyses of prospective studies. J Am Med Assoc 1998;279:1477-82.

Ridker PM, Stampfer MJ, Rifai N. Novel risk factors for systemic atherosclerosis: a comparison of C-reactive protein, fibrinogen, homocysteine, lipoprotein (a), and standard cholesterol screening as predictors of peripheral arterial disease. J Am Med Assoc 2001;285:2481-85.

Selvi E, Erlinger TP. Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 1999-2000. Circulation 2004;110:738-43.

Nasir K, Guallar E, Navas-Acien A, Criqui MH, Lima JA. Relationship of monocyte count and peripheral arterial disease: results from the National Health and Nutrition Examination Survey 1999-2002. Arterioscler Thromb Biol 2005;25:1966-71.

De Haro J, Acin F, Lopez-Quintana A, et al. Direct association between C-reactive protein serum levels and endothelial dysfunction in patients with claudication. Eur J Vasc Endovasc Surg 2008;35:480-6.

1

2

3

4

5

LEUKOCYTES AND ANTI-ENDOTHELIAL CELL ANTIBODIES

121EUROPEAN JOURNAL OF CARDIOVASCULAR MEDICINE VOL II ISSUE II

De Haro J, Martínez E, Florez A, Varela C, Bleda S, Acín F. Nitric oxide: link between endothelial dysfunction and inflammation in patients with peripheral arterial disease of the lower limbs. Interact Cardiovasc Thorac Surg 2009:9:107-12.

Lin YS, Lin CF, Lei HY, et al. Antibody-mediated endothelial cell damage via nitric oxide. Curr Pharm Des 2004;10:213-21.

Lin CF, Lei HY, Shiau AL. Endothelial cell apoptosis induced by antibodies against dengue virus non-structural protein 1 via production of nitric oxide. J immunol 2002;169:657-64.

Lombardo A, Biasucci LM, Lanza GA, et al. Inflammation as a possible link between coronary and carotid plaque instability. Circulation 2004;109:3158-63.

Rossi A, Franceschini L, Fusaro M, et al. Carotid atherosclerosis plaque instability in patients with acute myocardial infarction. Int J Cardiol 2006;111:263-6.

Noto AT, Bogeberg Mathiesen E, Amiral J, Vissac AM, Hansen JB. Endothelial dysfunction and systemic inflammation in persons with echolucent carotid plaques. Thromb Haemost 2006;96:53-9.

Brevetti G, Sirico G, Lanero S, De Maio JI, Laurenzano E, Giugliano G. The prevalence of hypoechoic carotid plaques is greater in peripheral than in coronary artery disease and is related to neutrophil count. J Vasc Surg 2008;47:523-9.

Schmitz G, Herr AS, Rothe G. T-lymphocytes and monocytes in atherogenesis. Herz 1998;23:168-77.

Boring L, Gosling J, Cleary M, Charo IF. Decreased lesion formation in CCR2-/- mice reveals a role for chemokines in the initiation of atherosclerosis. Nature 1998;394:894-7.

Libby P. Molecular bases of the acute coronary syndromes. Circulation 1995;91:2844-50.

Niccoli G, Ferrante G, Cosentino N, et al. Eosinophil cationic protein: A new biomarker of coronary atherosclerosis. Atherosclerosis 2010;211:606-11.

Prentice RL, Szatrowski TP, Fujikura T, Kato H, Mason MW, Hamilton HH. Leukocyte counts and coronary heart disease in a Japanese cohort. Am J Epidemiol 1982;116:496-509.

Lee CD, Folsom AR, Nieto FJ, Chambless LE, Shahar E, Wolfe DA. White blood cell count and incidence of coronary heart disease and ischemic stroke and mortality from cardiovascular disease in Africa-American and White men and women: atherosclerosis risk in communities study. Am J Epidemio 2001;154:758-64.

Venge P, Byström J, Carlson M, et al. Eosinophil cationic protein (ECP): molecular and biological properties and the use of ECP as a marker of eosinophil activation in disease. Clin Exp Allergy 1999;29:1172-86.

Venge P, Byström J. Eosinophil cationic protein (ECP). Int Biochem Cell Biol 1998;30:433-7.

Chihara J, Yamamoto T, Kurachi D, Kakazu T, Higashimoto I, Nakajima S. Possible release of eosinophil granule proteins in response to signalling from intercellular adhesion molecule-1 and its ligands. Int Arch Allergy Immunol 1995;108:52-4.

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

REFERENCES

Marrack P, Kappler J, Kotzin BL. Autoimmune disease: why and where it occurs. Nat Med 2001;7:899-905.

Hansson GK, Libby P. The immune response in atherosclerosis: a double edged sword. Nat Rev Immunol 2006;5:508-18.

Zhou X, Nicoletti A, Elhage R, Hansson GK. Transfer of CD4+ T cells aggravates atherosclerosis in immunodeficient apolipoprotein E knockout mice. Circulation 2000;102:2919-22.

Xu Q, Kleindienst R, Waitz W, Dietrich H, Wick G. Increased expression of heat shock proteine 65 coincides with population of infiltrating T lymphocytes in atherosclerosis lesions of rabbits specifically responding to heat shock protein 65. J Clin Invest 1993;91:2693-702.

Phershinka H, Wellezohn B, Parson W, et al. Identification of athero-sclerosis-associated conformational heat shock protein 60 epitopes by phage display and structural alignment. Atherosclerosis 2007;194:79-87.

Shoenfeld Y, Gerli R, Doria A, et al. Accelerated atherosclerosis in autoimmune rheumatic diseases. Circulation 2005;112:3337-47.

Meroni P, Ronda N, Raschi E, Borghi MO. Humoral autoimmunity against endothelium: theory or reality? Trends Immunol 2005;26:275-81.

Bordron A, Dueymes M, Levy Y, et al. The binding of some human antiendothelial cell antibodies induces endothelial cell apoptosis. J Clin Invest 1998;101:2029-35.

Worda M, Sgonc R, Dietrich H, et al. In vivo analysis of the apoptosis-inducing effect of anti-endothelial cell antibodies in systemic sclerosis by the chorionallantoic membrane assay. Arthritis Rheum 2003;48:2605-14.

Bordron J, Révélen R, Dueymes M, Youinou P. Pathogenicity of antiendothelial cell autoantibodies. In: Schoenfeld Y, Harats D, Wiik G, editors. Atherosclerosis and Autoimmunity. Elsevier 2001, 2001:203-10.

Del Papa N, Guidali L, Sironi M, et al. Anti-endothelial cell IgG anti-bodies from patients with Wegener’s granulomatosis bind to human endothelial cells in vitro and induce adhesion molecule expression and cytokine secretion. Arthritis Rheum 1996;39:758-66.

Yazici ZA, Raschi E, Patel A, et al. Human monoclonal anti-endothelial cell IgG-derived from systemic lupus erythematosus patient binds and activates human endothelium in vitro. Int Immunol 2001;13:349-57.

Varela C, de Haro J, Bleda S, Esparza L, Lopez de Maturana I, Acin F. Anti-endothelial cell antibodies are associated with peripheral arterial disease and markers of endothelial dysfunction and inflammation. Interact Cardiovasc Thorac Surg 2011;13:463-7.

Medina F, De Haro J, Martínez E, Florez A, March JR, Acín F. Relationship between non-invasively measured endothelial function and peripheral arterial disease. Angiology 2009;60:725-31.

Ignarro LJ, Cirino G, Casini A, Napoli C. Nitric oxide as a signalling molecule in the vascular system: an overview. J Cardiovasc Pahrmacol 1999;34:879-86.

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20