Imbalance between angiogenic and anti-angiogenic factors in sera from patients with large-vessel vasculitis L. Pulsatelli1, L. Boiardi2, E. Assirelli1, G. Pazzola2, F. Muratore2, O. Addimanda3,4, P. Dolzani 1, A. Versari5, M. Casali5, B. Bottazzi6, L. Magnani2, E. Pignotti3, N. Pipitone2, S. Croci7, A. Mantovani6,8,9, C. Salvarani2,10, R. Meliconi3,4

ABSTRACT Objective. Introduction Large-vessel vasculitides (LVV) are im- mune-mediated vascular diseases that typically affect the aorta and its major branches. Giant cell arteritis (GCA) and Takayasu’s arteritis (TAK) are the two major forms of LVV. GCA typically af- fects the extracranial branches of the Methods. carotid artery and is almost exclusive- ly seen in people aged over 50 years, while TAK commonly affects subjects under the age of 40 (1, 2). LVV are characterised by dysfunctional - tory lesions within the vessel wall. Vas- cular remodelling in response to aberrant immune response leads to occlusion of the lumen and ischaemic damage of me- dium-sized arteries, whereas in the aorta, aneurysm formation, rupture or dissec- Results. tion are more frequently seen (3-5). - In vasculitis immune-mediated mecha- nisms, a dual role of has - been hypothesised. Indeed, angiogen- esis might be primarily involved in - - quently, might be considered as a rescue CLINICAL AND EXPERIMENTAL RHEUMATOLOGY mechanism in ischaemic damage (6). The role of angiogenesis in LVV is also Key words: giant cell arteritis, - supported by histopathological analyses Takayasu’s arteritis, positron-emission - that have shown new capillary forma- tomography, angiogenic factors, - tion in the arterial media and intima, anti-angiogenic factors normally avascular (7, 8). In addition, a marked neoangiogenesis expanding the - capillary network of the adventitia was - recognised as a further pathogenic event. Adventitial vasa vasorum drive T cells and macrophages invasion of the vessel wall, and promote their differentiation. Recent studies have strengthened the role of interplay between tissue invasive T-cell and endothelial cells and stromal Conclusion. - cells in the vessel wall, ultimately lead- ing to formation of microvessels and

Clinical and Experimental Rheumatology 2020 S-23 Angiogenesis-related factors in LVV / L. Pulsatelli et al. expansion of the intimal layer (9, 10). Materials and methods indicative of vasculitis. Disease was Angiogenesis is a complex and highly Between January 2003 and September least two new or worsened items (32). circulating factors as well as locally 2010, 47 patients with LVV (33 with For the purpose of this study, due to produced mediators (11, 12). TAK and 14 with GCA) who had been the heterogeneity of the angiographic Physiological formation of new blood referred to the Rheumatology Unit modalities, worsening of angiographic vessels derives from the net balance at the Arcispedale S. Maria Nuova in - between angiogenic inducers and in- Reggio Emilia, Italy were enrolled in ening of luminal changes in the affect- hibitors (positive and negative regula- the cross-sectional study. Patients were ed vessel or the appearance of new vas- tors of angiogenesis) (11, 12). diagnosed according to the ACR classi- cular lesions. MRA, CTA and/or colour A wide array of factors is involved in this Doppler ultrasound were used to assess process. Vascular endothelial growth (30), as appropriate. luminal changes. As a tertiary reference centre for vascu- Clinical and 18F-FDG PET/CT data (FGF), platelet-derived growth factor litis, we admit patients with suspected, from an additional 20 patients (5 with (PDGF), , soluble adhe- early, or established LVV for re-evalua- TAK and 15 with GCA) were also in- sion molecules (such as soluble vascular tion of treatment and/or disease activity. cluded in the cluster analysis (no serum cellular adhesion molecule sVCAM-1) Among our patients, 36 (11 GCA) samples were available from the pa- are recognised as positive regulators were on treatment: 19 received steroid tients in the latter group). of angiogenesis, whereas the major in- only (prednisone: 16.07±12.35 mg/ Written informed consent was obtained hibitors include , angiostatin, day, mean ± standard deviation), 13 from all patients according to the Dec- received immunosoppressive agents laration of Helsinky and the study was (methotrexate, azathioprine, mycophe- approved by the local ethics committee. of angiogenesis, and -2 nolate, cyclophosphamide) with or (11-15). without steroid, and 4 received biolog- VEGF, FGF and PDGF are expressed in vasculitic lesions (7, 8, 16, 17), as well 1 tocilizumab). Positron emission tomography scans as the up-regulation of circulating levels Patients with LVV are offered a stand- were performed using a hybrid PET/ of some proangiogenic factors was ob- CT scanner (Discovery, GE) with 3.30 served in patients with LVV (6, 18-20). then yearly in order to obtain a compre- min-emission scan/bed and CT- attenu- On the other hand, pentraxin (PTX) hensive evaluation of the disease activ- ation correction. All subjects fasted for 3, the prototypical long PTX that has ity and extension by means of duplex been recognised as an antiangiogenic sonography, computerised tomography (37 Mbq of 18F-FDG/13 kilograms of factor by binding to FGF-2 and inhib- (CT), or magnetic resonance (MR), an- patient weight). Mean time from in- iting FGF-2-dependent endothelial cell giography (CTA, MRA), and 18F-FDG jection to acquisition was 60 minutes. proliferation and neovascularisation PET/CT. This protocol includes the de- Whole-body 18F-FDG PET/CT scan- (21-23), is to be up-regulated at both termination of erythrocyte sedimenta- ning was performed from mid-femora vascular and circulating levels in LVV tion rate (ESR) and C-reactive to external auditory meatus. Vascular (19, 24, 25) and also in small-vessel (CRP) at the time of 18F-FDG PET/CT uptake was visually graded using a vasculitis (26-28) evaluation. four-point scale (33, 34) ranging from In order to gain further insights into the A semi-quantitative clinical evaluation 0 to 3, where 0 = no uptake, 1 = low- molecular mechanisms driving angio- grade uptake (lower than liver uptake), genesis dysregulation in LVV, we in- the unit (considered as the baseline in 2 = intermediate-grade uptake (similar vestigated the serum levels of a panel this study) using the ITAS (31) and the to liver uptake) and 3 = high-grade up- of angiogenic inducers (VEGF, FGF-2, NIH/Kerr (32) indices. The ITAS lists take (higher than liver uptake). soluble vascular cellular adhesion mol- constitutional and organ manifestations Four aortic segments (ascending tho- ecule sVCAM-1, 1, -2,) with particular emphasis on cardiovas- racic aorta, aortic arch, descending tho- and inhibitors (angiostatin, endosta- racic aorta and abdominal aorta) and tin, PTX3) in patients with GCA and markers (ESR and CRP) and the physi- the carotid, subclavian, axillary, iliac TAK. Furthermore, we analysed the cian’s global opinion. A score is gen- and femoral arteries were evaluated bi- soluble factor levels according to dis- erated on the basis of present (=1) or laterally and 18F-FDG vascular uptake ease activity assessed by means of 18F- absent (=0) clinical manifestations, and Fluorodeoxyglucose-positron emission disease was considered active if one for vasculitis and scores of 0 and 1 tomography/computerised tomography or more organ system scores positive (18F-FDG PET/CT) and clinical indi- (31). The NIH/Kerr index assesses four Assessment of the 18F-FDG PET/CT ces including the National Institutes of items: constitutional manifestations, data was carried out by two nuclear medi- Health (NIH)/Kerr criteria and the In- raised ESR, manifestations of vascular cine specialists who were blinded to clini- dian Takayasu’s Activity Score (ITAS). ischaemia and angiographic features

S-24 Clinical and Experimental Rheumatology 2020 Angiogenesis-related factors in LVV / L. Pulsatelli et al.

Table I. Characteristics of normal controls (NC), giant cell arteritis (GCA) patients and Takayasu’s arteritis (TAK) patients at baseline. NC GCA NC TAK Two groups of healthy subjects, matched for age to the GCA and TAK No. of subjects 21 14 25 33 patients, respectively, were evaluated Age, mean (SD) 64 (10) 61 (8) 40 (9.0) 42 (14) Sex, no. female (%) 12 (57)a 8 (57) 17 (68)a 27 (81) as normal controls (NC). FDG-PET/CT scan, no. positive (%) NA 6 (43) NA 20 (61) Venous blood was drawn from all the Active disease according to Kerr score NA 7 (50) NA 16 (48) patients and all the NC between 8.00 no. positive (%) a.m. and 10.00 a.m. and collected in Active disease according to ITAS score, NA 7 (50) NA 15 (45) no. positive (%) tubes without anticoagulant. Serum ESR, median (range), mm/hr NA 12.5 (2.0-120.0) NA 28.0 (2.0-116.0) was separated by centrifugation for 10 CRP, median (range), mg/dl NA 0.64 (0.08-9.59) NA 0.87 (0.08-13.39) minutes 1000 x g. Samples were ali- No. of treated patients (%)b NA 11 (79) NA 25 (74) quoted and stored at -80°C until analy- 18F-FDG PET/CT: 18F-Fluorodeoxyglucose-positron emission tomography/computerised tomography; sis. Serum levels of VEGF, sVCAM-1, ITAS: Indian Takayasu’s activity score; ESR: erythrocyte sedimentation rate; CRP: C-reactive protein; FGF-2, endostatin, angiopoietin-1, an- NA: not applicable. giopoietin-2, angiostatin were evaluat- a ed using commercial sandwich enzyme (using Fisher’s exact test) (NC GCA: =1.000: NC . TAK: =0.353). bCorticosteroid with or without immunosuppressive drug (methotrexate/azathioprine/cyclophospha- immunoassay following the manufac- mide) /biological drugs. turer’s instructions. We used ELISA kits from R&D Systems, Minneapo- lis, MN, USA for detection of VEGF, sVCAM-1, FGF, Angiopoietin-1 An- giopoietin-2 and endostatin; serum angiostatin levels were measured us- ing ELISA kits from Cusabio College Park, MD, USA. Serum levels of PTX3 were evaluated by means of a sandwich ELISA as pre- viously described (35). TAK TAK ESR and CRP levels were measured at the same time points of sample collec- tion for circulating cytokines.

Continuous data were expressed as median and 25th–75th percentiles or as means and standard deviation depend- ing on the distribution of data, unless otherwise indicated. The Kolmogorov Smirnov test was performed to test TAK TAK normality of continuous variables. The Fig. 1. Serum concentrations of angiogenic fac- Levene test was performed to assess tors in normal controls (NC) and in patients with the homoscedasticity. Takayasu’s arteritis (TAK) and giant cell arteritis (GCA). Comparison of angiogenic/anti-angi- Boxes show 25th and 75th percentiles. Lines ogenic factor levels between the two within boxes show medians. Vertical lines below independent groups was performed and above boxes show 10th and 90th percentiles. by non-parametric analysis using the Solid circles indicate outliers. Statistical analysis was performed using Mann-Whitney test. Mann-Whitney test, since distribution VEGF: vascular endothelial growth factor; FGF- of normality and homoscedasticity. TAK vascular cellular adhesion molecule-1. The Spearman rank correlation was used to assess correlations between As already reported in a previous study A analysis among clusters was continuous variables. The Fisher Chi (36), two-step Cluster analysis was performed to identify which variables square test was performed to investi- performed to identify any subgroups of characterised each cluster (36). gate the relationships between dichoto- LVV patients with peculiar 18F-FDG In the present study, comparison of an- mous variables. giogenesis-related factor levels among

Clinical and Experimental Rheumatology 2020 S-25 Angiogenesis-related factors in LVV / L. Pulsatelli et al. clusters was performed by applying the to those with inactive disease (Suppl. Kruskal-Wallis test followed by the Dunn’s post-hoc correction. VEGF/Angiostatin, VEGF/endostatin Ordered differences among classes were and VEGF/PTX3 ratios were found in analysed using the Jonckheere Terpstra test. patients with active disease compared Statistical analysis was carried out us- to those with inactive disease (data not ing SPSS version19.0 (IBM Corp., Ar- shown). monk, NY, USA) and GraphPad Prism for Windows v. 5.0 (CA, USA). TAK Results As already reported in a previous study (36), we analysed 18F-FDG PET/CT vascular uptake in four aortic segments The baseline characteristics of NC, GCA (ascending thoracic aorta, aortic arch, and TAK patients are reported in Table I. descending thoracic aorta and abdomi- Among the angiogenic factors, only nal aorta) and the carotid, subclavian, axillary, iliac and femoral arteries, bi- higher in TAK patients compared to laterally, by cluster analysis to identify NC (=0.0029) (Fig. 1). any subgroups of LVV patients with TAK No difference was found for angiogen- peculiar 18F-FDG PET/CT vascular ic factor serum levels between GCA patients and NC. were obtained by 18F-FDG PET/CT Anti-angiogenic factor (Angiostatin, assessment performed in 67 patients Endostatin, PTX3) serum levels were with LVV (38 TAK and 29 GCA) and, TAK patients compared to NC (Fig. 2). and were characterised as previously - detailed (36) (Fig. 3). tween GCA and TAK patients. - angiogenic factor levels of LVV pa- cant difference between the treated and TAK tients included in each cluster (avail- untreated patients in our series (Data in Fig. 2. Serum concentrations of anti-angiogenic able for 47 patients: 33 with TAK and factors in normal controls (NC) and in patients 14 with GCA) according to diagnosis No correlation was found between with Takayasu’s arteritis (TAK) and giant cell and then we compared the angiogene- arteritis (GCA). VEGF and anti-angiogenic factor (An- Boxes show 25th and 75th percentiles. Lines sis-related factors levels in each cluster giostatin, Endostatin, PTX3) serum within boxes show medians. Vertical lines below to NC. levels. Furthermore, no correlation and above boxes show 10th and 90th percentiles. Given the small sample size of the Solid circles indicate outliers was observed between angiogenic/ant- Statistical analysis was performed using Mann- GCA patient group in Cluster 1 and angiogenic factors and either ESR or Whitney test. PTX3: pentraxin 3. Cluster 2 (36), this analysis was limited CRP in both LVVs (data not shown). to TAK patients. Among the angiogenic factors, as only active scan, compared to those with an When considering these patients, inactive scan, except for PTX3 in TAK VEGF, angiostatin and PTX3 levels patients (Suppl. Fig. S2). Similar results NC, we also evaluated the VEGF/An- were obtained when considering VEGF/ only in Cluster 3 compared to the NC giostatin, VEGF/endostatin and VEGF/ Angiostatin, VEGF/endostatin and group (Figs. 4-5), whereas endostatin PTX3 ratios in order to better investi- VEGF/PTX3 ratios (data not shown). levels were found to be higher in Clus- ters 1 and 3 (Fig. 5). VEGF and the anti-angiogenic factors. was found between soluble factor lev- - Only the VEGF/endostatin ratio was sig- els and 18F-FDG PET/CT total score ences in VEGF and angiostatin serum - in either TAK or GCA patients. levels among clusters were also ob- pared to NC (=0.0325) (Suppl. Fig. S1). When disease activity was evaluated served, with the highest concentration according to NIH/Kerr and ITAS in- corresponding to Cluster 3 (Jonckheere dices, only VEGF serum levels ap- Terpstra test: =0.001 and <0.0005 respectively) (Figs. 4-5). When stratifying LVV patients accord- in GCA patients with active disease, As already reported, in the three clusters, ing to 18F-FDG PET/CT assessment, whereas in terms of all the other solu- the percentage of patients with positive ble factors, no difference was found in levels was observed in patients with an patients with active disease compared 25% and 89%, respectively (36).

S-26 Clinical and Experimental Rheumatology 2020 Angiogenesis-related factors in LVV / L. Pulsatelli et al.

Fig. 3.., 2017; 35 (Suppl. 103): S102-10, with permission of

Discussion Recently, plasma levels of the N-ter- In contrast with our results, high circu- In this study, we performed a paired minal fragment of Chromogranin-A lating levels of VEGF have been pre- evaluation of circulating levels of a set (CgA), which have a recognised in- viously reported in GCA patients (19, of positive (VEGF, FGF-2, Angiopoi- hibitory effect on angiogenesis, have 43, 44). etin-1, -2, sVCAM-1) and negative been found markedly increased in TAK In these papers, VEGF up-regulation (endostatin, angiostatin, PTX3) regu- in GCA may be due to the inclusion lators of angiogenesis in patients with of the anti-angiogenic inducers in this of various subgroups of GCA patients GCA and TAK, in order to gain further LVV (37). (. biopsy-proven GCA clini- insight into the angiogenesis-driving Furthermore, a genome-wide associa- cally diagnosed GCA patients, and mechanisms in LVV. GCA patients with very recent optic Our overall results demonstrated a cir- within gene associated with the nerve ischaemia patients with- risk of developing GCA. Interestingly, out optic ischaemia), which typically by a prevailing expression of anti-angi- this gene encodes a secreted blood zy- showed higher levels of VEGF com- ogenic soluble factors in LVV. mogen that can be converted to pared to the other GCA patients. Indeed, among the angiogenic factors, and angiostatin and, considering the It should also be brought to mind that - opposing role of these factors in the an- the variability of VEGF circulating ly higher in TA patients compared to NC. giogenic process, risk alleles could levels is also determined by a func- On the other hand, anti-angiogenic fac- contribute to the imbalance of the angio- tional genetic variant that affects gene tor (angiostatin, endostatin, PTX3) serum genic mechanism in GCA (38). expression and circulating protein lev- It is noteworthy that in other vasculitides els (45) and by platelet release in serum GCA and TAK patients compared to NC. angiogenic factors are elevated. Indeed, samples (43, 46, 47). The cluster analysis limited to TAK in Kawasaki disease (KD), a systemic patients and the VEGF/endostatin ra- vasculitis that predominantly affects in- increase in VEGF circulating levels compared to NC. These results may be GCA patients compared to NC) further VEGF levels and lower endostatin lev- considered in line with those obtained els have been observed (39). by Dogan . (20), that showed high- angiogenic pattern. In small vessel ANCA-associated vas- The up-regulation of the anti-angiogenic culitis, increasing levels of VEGF, VEGF levels in TA patients compared mediators could, on the one hand, coun- sVCAM-1, Angiopoietin-2 and PTX3 to NC. In the same study, in contrast have been reported (27, 40-42). with our results, the authors reported a angiogenesis, but, on the other, it might Thus, considering the above-mentioned contribute to ischaemic damage. patients with active disease compared Up to now, the modulation of anti-angio- hypothesis of a distinctive proangio- to those with inactive disease. These genic factors has barely been investigated.

Clinical and Experimental Rheumatology 2020 S-27 Angiogenesis-related factors in LVV / L. Pulsatelli et al.

TAK

TAK TAK

TAK

TAK TAK

Fig. 4. Angiogenic factor levels in three TAK Boxes show 25th and 75th percentiles. Hori- zontal lines within boxes show medians. Verti- cal lines below and above boxes show 10th and 90th percentiles. Solid circles indicate outliers. Comparison of angiogenic factor levels among clusters was performed using the Kruskal- Wallis test followed by the Dunn’s correction. TAK - ters (Jonckheere Terpstra test): =0.001. Anti-angiogenic factor levels in three TAK: Takayasu’s arteritis; VEGF: vascular Fig. 5. Boxes show 25th and 75th percentiles. Horizon- growth factor-2; sVCAM-1: soluble vascular TAK tal lines within boxes show medians. Vertical cellular adhesion molecule-1. lines below and above boxes show 10th and 90th percentiles. Solid circles indicate outliers. considering the acknowledged limi- myalgia rheumatica (PMR), a disease Comparison of anti-angiogenic factor lev- tation of currently utilised criteria to closely related to GCA, we found sig- els among clusters was performed using the monitor disease activity in LVV (48). - Kruskal-Wallis test followed by the Dunn’s correction. - pared to NC, whereas no difference in - PTX3 levels was found between PMR (Jonckheere Terpstra test): <0.0005. els of PTX3 in LVV compared to NC. and NC, either at disease onset or dur- TAK: Takayasu’s Arteritis; PTX3: pentraxin 3. Currently, the role of PTX3 as a bio- ing follow-up. marker of disease activity in LVV is still Since, in this study, opposing results study on GCA and PMR patients should debated. Indeed, even if several lines of have been shown in GCA (no elevation verify this hypothesis. - of VEGF together with an up-regulation Two main limitations of this study tentially useful molecule for assessing should be addressed. disease activity in LVV (49, 50), other suggest a role for combined evaluation Firstly, the relatively small sample size of VEGF and PTX3 as a potential bio- - - fected the statistical power of the analy- PTX3 serum levels and disease activity cult GCA in patients with polymyalgic sis of patient subgroups obtained follow- in LVV (24, 25). This discrepancy may features. Prompt recognition of GCA - - has a relevant clinical impact leading to tive disease adopted in the studies. prompt glucocorticoid treatment, which group was only partially investigated In two previous studies on VEGF (51) prevents the cranial ischaemic compli- and the analysis performed on pro-an- and on PTX3 (52) serum levels in poly- cations of GCA. A direct comparison giogenic/anti-angiogenic factors levels

S-28 Clinical and Experimental Rheumatology 2020 Angiogenesis-related factors in LVV / L. Pulsatelli et al.

fore, our main results ( normal lev- AIFA, FOREUM, UpToDate. A. Man- analysis were limited to TAK patients. els of angiogenic factors and increased tovani has received research grants from Secondly, since our patients were en- anti-angiogenic factor levels) may be a Novartis, Roche, Ventana, Pierre Fabre, rolled in a tertiary Rheumatology Cen- consequence of the treatment effects on Verily, AbbVie, Compugen, Macrophage tre with special interest in vasculitis, factor serum levels. Therapeutics, AstraZeneca, Biovelocita, the great majority of LVV patients BG Fund, Third Rock and Verseau, is an evaluated in the study were already on may modify serum levels of angiogen- iventor of patents related to PTX3 and corticosteroid treatment, either with or esis-related factors differently and the other innate immunity molecules, and without immunosuppressive drugs/bio- receives royalties for reagents related to logic agents (53). - innate immunity. The other authors have The effect of treatment on angiogenic/ lating levels of angiogenic/anti-angio- declared no competing interests. anti-angiogenic factors has been previ- genic molecules should be investigated ously described. Several and References studies on cancer and placenta an- Finally, the main point of interest raised 1. MILLER DV, MALESZEWSKI JJ: The pathol- ogy of large-vessel vasculitides. giogenesis reported a down-regulation by this study lies in the novel and origi- 2011; 29 (Suppl. 64): S92-8. of VEGF expression induced by gluco- nal results concerning the imbalance of 2. ESPIGOL-FRIGOLE G, PRIETO-GONZALEZ S, corticoids (54, 55). the circulating molecular pattern to- ALBA MA: Advances in the diagnosis of large vessel vasculitis. In a recent study, VEGF and angiopoi- 2015; 41: 125-40, ix. etin-2 serum levels have been showed pattern of angiogenic/anti-angiogenic 3. WEYAND CM, GORONZY JJ: Immune mecha- to be lowered in glucocorticoid-treated response should be evaluated in treat- nisms in medium and large-vessel vasculitis. GCA patients compared to baseline lev- ment-naive patients in the early phase 2013; 9: 731-40. 4. WATANABE R, ZHANG H, BERRY G, els (56). of the disease. GORONZY JJ, WEYAND CM: Immune check- point dysfunction in large and medium vessel were reported concerning serum angi- Acknowledgements vasculitis. - The authors thank Alexandra Teff for 2017; 312: H1052-H9. 5. MARUOTTI N, CANTATORE FP, NICO B, deed, cyclophosphamide and steroids English language assistance and - VACCA A, RIBATTI D: Angiogenesis in vascu- have been shown to increase angiopoi- for litides. 2008; 26: 476-83. etin-1 (57), whereas other observations granting permission to publish Figure 3. 6. CID MC, HERNANDEZ-RODRIGUEZ J, ESTE- have revealed that several medications, BAN MJ: Tissue and serum angiogenic activity is associated with low prevalence of such as steroids, cyclosporine, colchi- ischemic complications in patients with giant- cine, did not have any effects on serum 1Laboratory of Immunorheumatology cell arteritis. 2002; 106: 1664-71. angiopoietin-1 and angiopoietin-2 (58). and Tissue Regeneration, IRCCS Istitu- 7. KAISER M, YOUNGE B, BJORNSSON J, GORONZY JJ, WEYAND CM: Formation of Among the anti-angiogenic factors, to Ortopedico Rizzoli, Bologna, Italy; new vasa vasorum in vasculitis. Production of 2 increased endostatin levels were ob- Division of Rheumatology, Azienda angiogenic cytokines by multinucleated giant served in RA patients treated with a Unità Sanitaria Locale-IRCCS di Reg- cells. 1999; 155: 765-74. combination of disease-modifying anti- gio Emilia, Italy; 3Medicine and Rheu- 8. WEYAND CM, GORONZY JJ: Arterial wall injury in giant cell arteritis. rheumatic drugs (59). Similarly, PTX- matology Unit, IRCCS Istituto Ortope- 1999; 42: 844-53. 4 3 circulating levels were up-regulated dico Rizzoli, Bologna, Italy; Dept. of 9. ZHANG H, WATANABE R, BERRY GJ : by glucocorticoids in mouse and hu- Biomedical and Neuromotor Sciences, man treated with dexamethasone (60). University of Bologna, Italy; 5Nuclear medium and large vessel vasculitis. 2017; 114: E970-E9. Opposing results were reported by our Medicine Unit, Azienda Unità Sani- 10. WATANABE R, MAEDA T, ZHANG H: MMP previous study performed on PMR pa- taria Locale-IRCCS di Reggio Emilia, (matrix metalloprotease)-9-producing mono- Italy; 6- cytes enable T cells to invade the vessel wall and decrease of PTX3 serum levels com- munology, Humanitas Clinical and cause vasculitis. 2018; 123: 700-15. 11. TAS SW, MARACLE CX, BALOGH E, SZE- pared to baseline was observed during Research Center IRCCS, Milan, Italy; KANECZ Z: Targeting of proangiogenic sig- corticosteroid treatment (52). Similarly, 7Clinical Immunology, Allergy and Ad- tocilizumab (TCZ) has been shown to vanced Biotechnologies Unit, Azienda 2016; 12: 111-22. 12. WIETECHA MS, CERNY WL, DIPIETRO LA: lower circulating levels of PTX3 in Unità Sanitaria Locale-IRCCS di Reg- Mechanisms of vessel regression: toward an 8 GCA patients (61). gio Emilia, Italy; Humanitas Univer- understanding of the resolution of angiogenesis. - sity, Milan, Italy; 9The William Harvey 2013; 367: 3-32. gard to angiogenic/anti-angiogenic fac- Research Institute, Queen Mary Uni- 13. KOCH AE, HALLORAN MM, HASKELL CJ, 10 SHAH MR, POLVERINI PJ: Angiogenesis me- tor serum levels between treated and versity of London, UK; University of diated by soluble forms of E-selectin and vas- untreated LVV patients. However, this Modena and Reggio Emilia, Italy. cular cell adhesion molecule-1. 1995; 376: 517-9. account that the small sample size of Competing interests 14. KAZEROUNIAN S, LAWLER J: Integration of pro- and anti-angiogenic signals by endothelial the untreated group should affect the N. Pipitone has received fees from cells. 2018; 12: 171-9. statistical power of the analysis. There- AIM Group, leC, Alfa-Wassermann, 15. BORNSTEIN P: function as

Clinical and Experimental Rheumatology 2020 S-29 Angiogenesis-related factors in LVV / L. Pulsatelli et al.

regulators of angiogenesis. 30. HUNDER GG, BLOCH DA, MICHEL BA: 45. CARMONA FD, GONZALEZ-GAY MA, MAR- 2009; 3: 189-200. The American College of Rheumatology TIN J: Genetic component of giant cell arteri- 16. KAISER M, WEYAND CM, BJORNSSON J, tis. (Oxford) 2014; 53: 6-18. GORONZY JJ: Platelet-derived growth factor, arteritis. 1990; 33: 1122-8. 46. WEBB NJ, BOTTOMLEY MJ, WATSON CJ, intimal hyperplasia, and ischemic complica- 31. MISRA R, DANDA D, JAYASEELAN L, SIVA- BRENCHLEY PE: Vascular endothelial growth tions in giant cell arteritis. KUMAR R, LAWRENCE A, BACON P: ITAS factor (VEGF) is released from platelets dur- 1998; 41: 623-33. and DEI.TAK - scores for clinical disease ac- ing blood clotting: implications for measure- 17. O’NEILL L, ROONEY P, MOLLOY D: tivity and damage extent in Takayasu aorto- ment of circulating VEGF levels in clinical - arteritis (TA). (Oxford) 2008; disease. (Lond) 1998; 94: 395-404. esis in giant cell arteritis by acute-phase se- 47: 101. 47. FERRERO S: Serum and plasma vascular en- rum Amyloid A. 2015; 32. KERR GS, HALLAHAN CW, GIORDANO J dothelial growth factor levels in testicular 67: 2447-56. : Takayasu arteritis. 1994; cancer patients. 2004; 15: 989-90. 18. CID MC, GRANT DS, HOFFMAN GS, AUER- 120: 919-29. 48. DIRESKENELI H: Clinical assessment in BACH R, FAUCI AS, KLEINMAN HK- 33. WALTER MA, MELZER RA, SCHINDLER C, Takayasu’s arteritis: major challenges and cation of as an angiogenic factor MULLER-BRAND J, TYNDALL A, NITZSCHE controversies. 2017; 35 in sera from patients with systemic vasculi- EU: The value of [18F]FDG-PET in the diag- (Suppl. 103): S189-93. tis. 1993; 91: 977-85. nosis of large-vessel vasculitis and the assess- 49. DAGNA L, SALVO F, TIRABOSCHI M: 19. BALDINI M, MAUGERI N, RAMIREZ GA ment of activity and extent of disease. Pentraxin-3 as a marker of disease activity : Selective up-regulation of the soluble 2005; 32: 674-81. in Takayasu arteritis. 2011; pattern-recognition receptor pentraxin 3 and 34. MELLER J, STRUTZ F, SIEFKER U : 155: 425-33. of vascular endothelial growth factor in giant Early diagnosis and follow-up of aortitis with 50. ISHIHARA T, HARAGUCHI G, KAMIISHI T, cell arteritis: relevance for recent optic nerve [(18)F]FDG PET and MRI. TEZUKA D, INAGAKI H, ISOBE M: Sensitive ischemia. 2012; 64: 854-65. 2003; 30: 730-6. assessment of activity of Takayasu’s arteritis 20. DOGAN S, PISKIN O, SOLMAZ D: Markers 35. DEYAB G, HOKSTAD I, WHIST JE: Anti- by pentraxin3, a new biomarker. of endothelial damage and repair in Takayasu rheumatic treatment is not associated with re- 2011; 57: 1712-3. arteritis: are they associated with disease activ- duction of pentraxin 3 in rheumatoid arthritis, 51. MELICONI R, PULSATELLI L, DOLZANI P ity? 2014; 34: 1129-38. psoriatic arthritis and ankylosing spondylitis. : Vascular endothelial growth factor pro- 21. RUSNATI M, CAMOZZI M, MORONI E: 2017; 12: e0169830. duction in polymyalgia rheumatica. 36. PULSATELLI L, BOIARDI L, ASSIRELLI E 2000; 43: 2472-80. factor-2 by the long pentraxin PTX3 inhibits : Interleukin-6 and soluble interleukin-6 52. PULSATELLI L, PERI G, MACCHIONI P: angiogenesis. 2004; 104: 92-9. receptor are elevated in large-vessel vasculi- Serum levels of long pentraxin PTX3 in pa- 22. ALESSI P, LEALI D, CAMOZZI M, CANTEL- tis: a cross-sectional and longitudinal study. tients with polymyalgia rheumatica. MO A, ALBINI A, PRESTA M: Anti-FGF2 ap- 2017; 35 (Suppl. 103): 2010; 28: 756-8. proaches as a strategy to compensate resist- S102-10. 53. ELEFANTE E, BOND M, MONTI S: One ance to anti-VEGF therapy: long-pentraxin 3 37. TOMBETTI E, COLOMBO B, DI CHIO MC year in review 2018: systemic vasculitis. as a novel antiangiogenic FGF2-antagonist. : Chromogranin-A production and frag- 2018; 36 (Suppl. 111): S12-32. 2009; 20: 225-34. mentation in patients with Takayasu arteritis. 54. MARTENS B, DREBERT Z: Glucocorticoid-me- 23. LEALI D, INFORZATO A, RONCA R: 2016; 18: 187. diated effects on angiogenesis in solid tumors. Long pentraxin 3/tumor necrosis factor-stim- 38. CARMONA FD, VAGLIO A, MACKIE SL: 2019; 188: 147-55. ulated gene-6 interaction: a biological rheo- 55. OZMEN A, UNEK G, KIPMEN-KORGUN D, risk alleles in plasminogen and P4HA2 as- CETINKAYA B, AVCIL Z, KORGUN ET: Glu- angiogenesis. sociated with giant cell arteritis. cocorticoid exposure altered angiogenic fac- 2012; 32: 696-703. 2017; 100: 64-74. tor expression via Akt/mTOR pathway in rat 24. TOMBETTI E, DI CHIO MC, SARTORELLI S 39. TAKESHITA S, KAWAMURA Y, TAKABAYASHI placenta. 2015; 198: 34-40. - H, YOSHIDA N, NONOYAMA S: Imbalance in 56. VAN SLEEN Y, SANDOVICI M, ABDULAHAD lar enhancement and progression in Takayasu the production between vascular endothelial WH : Markers of angiogenesis and arteritis. 2014; 16: 479. growth factor and endostatin in Kawasaki macrophage products for predicting disease 25. ALIBAZ-ONER F, AKSU K, YENTUR SP, KESER disease. 2005; 139: 575-9. G, SARUHAN-DIRESKENELI G, DIRESKENELI 40. LI CG, REYNOLDS I, PONTING JM, HOLT PJ, in giant cell arteritis. (Oxford) H: Plasma pentraxin-3 levels in patients with HILLARBY MC, KUMAR S: Serum levels of 2019 [Epub ahead of print]. Takayasu’s arteritis during routine follow-up. vascular endothelial growth factor (VEGF) 57. BASSYOUNI IH, SHARAF M, WALI IE, MAN- 2016; 34 (Suppl. 97): are markedly elevated in patients with We- SOUR HM- S73-6. gener’s granulomatosis. etin-1 in Behcet’s disease patients with vas- 26. VAN ROSSUM AP, PAS HH, FAZZINI F : 1998; 37: 1303-6. cular involvement. 2016; 31: Abundance of the long pentraxin PTX3 at 41. SCHNEEWEIS C, RAFALOWICZ M, FEIST E 918-24. sites of leukocytoclastic lesions in patients : Increased levels of BLyS and sVCAM-1 58. CHOE JY, PARK SH, KIM SK: Serum angiopoie- with small-vessel vasculitis. in anti-neutrophil cytoplasmatic tin-1 level is increased in patients with Behcet’s 2006; 54: 986-91. (ANCA)-associated vasculitides (AAV). disease. 2010; 77: 340-4. 27. FAZZINI F, PERI G, DONI A : PTX3 in 2010; 28 (Suppl. 57): S62-6. 59. NAGASHIMA M, ASANO G, YOSHINO S: Im- small-vessel vasculitides: an independent 42. MONACH PA, KUMPERS P, LUKASZ A: balance in production between vascular en- indicator of disease activity produced at sites Circulating angiopoietin-2 as a biomarker dothelial growth factor and endostatin in pa- 2001; 44: in ANCA-associated vasculitis. tients with rheumatoid arthritis. 2841-50. 2012; 7: e30197. 2000; 27: 2339-42. 28. SAHIN S, ADROVIC A, BARUT K: Pen- 43. SMETS P, DEVAUCHELLE-PENSEC V, ROU- 60. DONI A, MANTOVANI G, PORTA C: Cell- traxin-3 levels are associated with vasculitis ZAIRE PO, PEREIRA B, ANDRE M, SOUBRIER and disease activity in childhood-onset sys- M: Vascular endothelial growth factor levels hormones in hematopoietic and nonhematopoi- temic lupus erythematosus. 2017; 26: and rheumatic diseases of the elderly. - etic cells. 2008; 283: 29983-92. 1089-94. 2016; 18: 283. 61. GLOOR AD, YERLY D, ADLER S: Immu- 29. AREND WP, MICHEL BA, BLOCH DA : 44. GOODFELLOW N, MORLET J, SINGH S: no-monitoring reveals an extended subclini- The American College of Rheumatology 1990 Is vascular endothelial growth factor a useful cal disease activity in tocilizumab-treated - biomarker in giant cell arteritis? giant cell arteritis. (Oxford) ritis. 1990; 33: 1129-34. 2017; 3: e000353. 2018; 57: 1795-801.

S-30 Clinical and Experimental Rheumatology 2020