Bp

LLP

a

ARRA

KABPSTM

1

psab

ietosrtat

B

0d

Lung Cancer 74 (2011) 392– 395

Contents lists available at ScienceDirect

Lung Cancer

jou rn al h om epa ge: www.elsev ier .com/ locate / lungcan

lockage of vascular endothelial growth factor (VEGF) reduces experimentalleurodesis

isete R. Teixeira ∗, Francisco S. Vargas, Milena M.P. Acencio, Sabrina C. Ribeiro, Roberta K.B. Sales,eila Antonangelo, Evaldo Marchi

leura Laboratory, Pulmonary Division, Heart Institute (InCor), University of São Paulo Medical School, Brazil

r t i c l e i n f o

rticle history:eceived 15 December 2010eceived in revised form 18 April 2011ccepted 25 April 2011

eywords:nti-VEGF antibodyevacizumableurodesisilver nitratealcalignant pleural effusion

a b s t r a c t

Background and objective: Chemical pleurodesis controls recurrent malignant pleural effusion. The mecha-nism that determines pleural symphysis involves the action of vascular endothelial growth factor (VEGF).We assessed the influence of the anti-VEGF antibody (bevacizumab) on pleurodesis induced by talc orsilver nitrate and analyzed the temporal development of pleural angiogenesis.Methods: Sixty New Zealand rabbits received intrapleural injection (2 mL) of talc (400 mg/kg) or 0.5%silver nitrate. In each group, half of the animals received an intravenous injection of bevacizumab 30 minbefore the sclerosing agent. Five animals from each group were euthanized 7, 14, or 28 days after theprocedure. Adhesions and inflammation (scores: 0–4), thickness (�m), vascular density (vessels/field),and collagen fibers (�m2) were evaluated in the visceral pleura.Results: Antibody anti-VEGF interferes in pleurodesis induced by talc or silver nitrate. Pleural inflamma-tion was discreet with no difference between the groups, regardless the anti-VEGF treatment. Concerningthe vascular density of the visceral pleura, a smaller number of neoformed vessels was noted in the ani-

mals that received bevacizumab. In the animals receiving silver nitrate, the decrement in adhesions andvascular density was associated with reduced thick and thin collagen fibers, resulting in less pleuralthickness.Conclusion: The anti-VEGF antibody inhibits adhesions between pleural layers. Despite being an exper-imental study in animals with normal pleura, the results call attention to a likely lack of success in

locke

pleurodesis when VEGF b

. Introduction

Pleurodesis is frequently indicated in the treatment of recurrentleural diseases, especially pneumothorax and malignant effu-ions [1]. The intrapleural injection of a sclerosing agent producesn inflammatory exudate that disappears when symphysis occursetween the pleural layers [2,3].

Vascular endothelial growth factor (VEGF) is one of the mostmportant inflammatory mediators involved in producing pleuralffusion. In addition to its action in vascular permeability, it hashe capacity to activate endothelial cells, stimulating the formationf new blood vessels [4–7]. Experimental studies have demon-trated that the administration of anti-VEGF produces a significanteduction in the volume of pleural fluid in the first week after

he intrapleural injection of talc or nitrate [8] and a reduction inngiogenesis and pleural adhesions 14 days after instillation of theransformation growth factor-beta (TGF-�) [9] or doxycycline [10].

∗ Corresponding author at: Rua Copacabana 415/184, São Paulo 02461-000, SP,razil. Tel.: +55 1130695125.

E-mail address: lisetepneumo@yahoo.com.br (L.R. Teixeira).

169-5002/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.lungcan.2011.04.015

rs are used.© 2011 Elsevier Ireland Ltd. All rights reserved.

These observations allow to consider two relevant issues result-ing from the use of VEGF blockers that are apparently contradictory.On the one hand, the administration of anti-VEGF could be ben-eficial in the treatment of recurrent pleural effusion, because itreduces the production of fluid. On the other hand, its use couldhinder the efficacy of pleurodesis, especially in oncology patientswith recurrent effusions who are considered candidates for thistherapeutic approach.

Within this context, the primary objective of this study is todetermine the influence of the intravenous administration of theanti-VEGF antibody, bevacizumab, in pleurodesis induced by talc orsilver nitrate. The secondary objective is to assess the deposition ofcollagen and vascular development in the experimentally inflamedarea.

2. Material and methods

After approval of the project by the Institutional Ethics Commit-tee, 60 New Zealand rabbits, weighing between 1.5 and 3.0 kg, werestudied.

g Canc

2

aiPBwtatpfimwobb

np1s

2

mpctpaFn3ot

2

smm

tibsemgcfa(mSeidtgpp

L.R. Teixeira et al. / Lun

.1. Animal experiment

According to previously established methodology [11,12], thenimals were sedated and given analgesia with intramuscularnjections of 35 mg/kg of ketamine hydrochloride (Cristalia, Sãoaulo, Brazil) and 5 mg/kg of hydrochloride (Bayer, São Paulo,razil). After trichotomy of the right hemithorax and antisepsisith iodopovodine, a 0.5-cm incision was made in the skin (along

he hemiclavicular line), following the projection of the sixth costalrch. Next, a catheter (8 Fr) was tunneled through subcutaneousissue impeding possible attempts of removal. In sequence, thearietal pleura was exposed, and after the insertion, the drain wasxed. Using groups of 30 animals each, 400 mg/kg of talc (USP Phar-acy; São Paulo, Brazil) or 0.5% silver nitrate (Merck, Germany)as injected (2 mL) into the right pleural cavity. Half of the animals

f each group received, according to previous studies, 5 mg/kg ofevacizumab (Avastin®, Roche, Switzerland) intravenously 30 minefore the sclerosing agents [8,11].

After 7, 14, or 28 days, 5 animals from each group were eutha-ized by intravenous injection (marginal ear vein) of 40 mg/kg ofentobarbital [12,13]. The thorax was removed en bloc, and by using0% formalin, the lungs were expanded (60 mL) and maintainedubmersed for at least 48 h [12,13].

.2. Macroscopic evaluation

The macroscopic analysis was done according to previousethodology. Samples of visceral pleura were obtained from all

ulmonary lobes. When adhesions or symphysis were presentareful dissection was done in order to adequately collect represen-ative fragments [12,13]. A consensus grading of the macroscopicleurodesis was reached by two independent investigators (LRTnd MMPA) who were blinded concerning the treatment group.or assessment of the pleural adhesions, a 0–4 score was used (0:ormal pleural space; 1: up to 3 discrete adhesions; 2: more than

sparse adhesions; 3: generalized adhesions allowing separationf the lung from the thoracic wall, and 4: complete obliteration ofhe pleural space) [12,13].

.3. Microscopic evaluation

Fragments of visceral pleura were removed, processed andtained by hematoxylin–eosin according to previously usedethodology. The microscopic evaluation of the inflammation wasade using the 0–4 score (0: normal and 4: severe changes) [12,13].For the evaluation of collagen fibers, slides containing thin sec-

ions of visceral pleura were stained with 2% Sirius red dissolvedn saturated aqueous picric acid solution. The enhancement ofirefringence measured by the Picrosirius polarization method ispecific for collagen structures that consist of aggregates of ori-nted molecules. Thus, thin fibers (collagen III) corresponding toore immature collagen appear green and are weakly birefrin-

ent, whereas thick fibers (collagen I) corresponding to matureollagen appear orange–red and are strongly birefringent. Five dif-erent fields were randomly selected for the evaluation of collagent 400× magnification. The Leica Q500IV image analysis systemHeerbrugg, Switzerland), which permits geometric and densito-

etric quantitative measurements, was used for quantification.pecific thresholds for thin and thick fibers were established forach slide after an ideal contrast had been obtained allowing thedentification of the fibers. The area occupied by these fibers wasetermined by digital densitometry, with the device calibrated to

ransform pixel values into micrometers. The amount of total colla-en was expressed in terms of area (�m2), corrected for 103 �m2 ofleural area. This standardization allowed us to establish the pro-ortion between thick and thin fibers for each agent (talc or silver

er 74 (2011) 392– 395 393

nitrate). Total collagen corresponds to the sum of the amount ofthick and thin fibers (collagens I + III). Pleural thickness was deter-mined based on the linear measurement of the distance betweenthe internal and external margins of the pleura and is expressed inmicrometers [14].

Assessment of vascular density was carried out with slidesstained by immunohistochemistry for the factor VIII antigen (vonWillebrand factor), utilizing the polyclonal anti-factor VIII anti-body (Dako, AO82, Dako, CA, USA) [15] diluted at 1/7000. TheUniversal HRP Polymer kit (Biocare Medical, CA, USA) was usedto provide visible staining. Slides were counterstained with Mayerhematoxylin, dehydrated, and cover slipped. For the evaluation ofvascular density, 5 randomly selected fields were analyzed in eachcase. In this way, the degree of angiogenesis was expressed as themean number of vessels positively stained in the fields analyzed. Anobserver (LA) blinded to the treatment of the rabbit performed theanalysis.

2.4. Statistical analysis

The results are presented as means and standard error. For com-parison between the groups, the nonpaired t test or Mann–Whitneyrank sum test was used. The value of p < 0.05 was considered sig-nificant. The statistical analysis was performed using SigmaStat 3.1software (Systat, CA, USA).

3. Results

The intravenous administration of anti-VEGF resulted in asmaller degree of pleural adhesions in both groups and at all times(Fig. 1A). In animals injected with talc, pleural adhesion scoreswere significantly lower after the injection of bevacizumab: 7thday (1.6 ± 0.2 vs. 0.5 ± 0.2; p = 0.01); 14th day (2.2 ± 0.1 vs. 1.0 ± 0.2;p = 0.002); and 28th day (2.6 ± 0.2 vs. 1.4 ± 0.1; p = 0.008). A simi-lar behavior was observed in the group injected with silver nitrate:7th day (2.1 ± 0.1 vs. 0.9 ± 0.2; p = 0.001); 14th day (3.1 ± 0.6 vs.1.9 ± 0.1; p < 0.01); and 28th day (3.8 ± 0.2 vs. 1.5 ± 0.3; p = 0.029).Fig. 1B demonstrates the influence of anti-VEGF on the devel-opment of pleural thickness (�m). In the animals that receivedtalc, no significant changes were observed. On the other hand, inthose that received silver nitrate, there was a significant reduc-tion at all times: 7th day (116.5 ± 39.5 vs. 35.9 ± 5.2; p = 0.043);14th day (120.5 ± 10.5 vs. 55.6 ± 22.3; p = 0.047); and 28th day(195.0 ± 27.0 vs. 95.3 ± 22.5; p = 0.007). The collagen compositionof the pleura, assessed by the area (�m2) occupied by thick fibers(collagen I) or thin fibers (collagen III) is represented in Fig. 1C–F.In the animals that received talc, no influence of bevacizumab wasobserved. However, in those that received silver nitrate, we noteda reduction in total collagen around the 14th day (3139.0 ± 925.0vs. 1182.0 ± 424.0; p = 0.01) and the 28th day (5106.0 ± 1593.0 vs.1275.0 ± 255.0; p = 0.049), by the decrease in both thick and thinfibers.

Pleural inflammation was discreet (Fig. 2A), and no differ-ence was noted between the groups, regardless of the treatmentwith anti-VEGF. Concerning the vascular density of the visceralpleura (Fig. 2B), we noted a smaller number of neoformed ves-sels in the animals that received bevacizumab, regardless of thesclerosing agent utilized. In the animals that received talc, thedensity (number of vessels/field) was greater in those that hadnot received anti-VEGF: 7th day (10.2 ± 2.3 vs. 2.9 ± 0.9; p = 0.039);14th day (12.9 ± 0.8 vs. 3.2 ± 0.3; p < 0.001); and 28th day (4.1 ± 0.5

vs. 0.7 ± 0.05; p = 0.002). A similar behavior was seen in the ani-mals that received silver nitrate: 7th day (7.8 ± 1.6 vs. 2.8 ± 0.1;p = 0.035); 14th day (11.8 ± 4.6 vs. 3.5 ± 1.3; p = 0.03); and 28th day(6.6 ± 2.1 vs. 2.5 ± 0.7; p = 0.049).

394 L.R. Teixeira et al. / Lung Cancer 74 (2011) 392– 395

F on of

m 0.05 (

4

gsTdcit

F(

ig. 1. It shows the influence of the anti-VEGF antibody after intrapleural injectiicroscopic thickness (B), total collagen (C and F), thick (D) and thin (E) fibers. *p <

. Discussion

This study demonstrates that blockage of vascular endothelialrowth factor (VEGF) interferes with the architecture of pleurode-is induced by the intrapleural instillation of talc or silver nitrate.he decrement of macroscopic pleural adhesions and vascularensity observed after administration of the sclerosing agents is

oncomitant (in the animals that received silver nitrate) with anntense reduction in collagen fibers, causing a reduction in pleuralhickness.

ig. 2. It shows the changes in the visceral pleura after intrapleural injection of the scleroA) and vascular density (B). *p < 0.05 (with × without bevacizumab).

the sclerosing agents. Pleural cavity: macroscopic adhesions (A). Visceral pleura:with × without bevacizumab).

The pathophysiological mechanism associated with the symph-ysis of the pleural layers involves multiple factors, although thecommon denominator is the injury provoked in the mesothelialcells. It is believed that this injury is responsible for the inflamma-tory process that develops and culminates with the formation of theexudative pleural effusion, preceding the deposition of fibrin andcollagen, which determine the pleural thickness, and consequently,

symphysis of the cavity. In this way, the exudation of pleural fluidreflects the degree of inflammation and signals the intensity ofpleurodesis [2,3].

sing agents and the influence of the anti-VEGF antibody. Microscopic inflammation

g Canc

mvtvpc

frttaaTriwqottrattp

pitrtahpciot

atVrteigtadoca

ort

[

[

[

[

[

[

[

[[

L.R. Teixeira et al. / Lun

The role of the vascular endothelial growth factor in the for-ation of pleural effusion is widely known, whether by increasing

ascular permeability or by formation of new blood vessels. Withinhis context, the use of VEGF blockers produces a reduction in theolume of fluid [8,9]. However, this finding was not confirmed inatients with malignant pleural effusion due to nonsmall cell lungancer [16].

With the objective of better understanding the multiplicity ofactors involved in this process, Guo et al. [9,10] assessed theole of angiogenesis in pleurodesis induced by TGF-�, highlightinghe role of VEGF in modulating pleural symphysis. Corroboratinghese facts, they demonstrated that the intravenous injection ofnti-VEGF antibodies reduces the number of pleural adhesions innimals that undergo intrapleural instillation of doxycycline orGF-� [9,10]. The authors credited this lower intensity of pleu-odesis to the blockage of VEGF, which by causing a reductionn vascular permeability and in the formation of new vessels,

ould decrease the volume of liquid formed with the conse-uent action on the turnover of fibrin deposition, and therefore,f fibrosis. Previously we confirmed these findings demonstratinghat bevacizumab interferes in the acute phase of inflamma-ion induced by talc or silver nitrate. In summary, the anti-VEGFeduces the production of pleural fluid, pleural vascular perme-bility and IL-8 pleural fluid levels [8]. Both studies highlighthe importance of VEGF in pleural fluid production and suggesthat anti-VEGF agents should be avoided in patients requiringleurodesis.

These facts evidence the importance of the binomial vascularermeability/angiogenesis in pleural exudation resulting from the

nstillation of sclerosing agents. One should also consider, however,hat despite the importance of VEGF in vascular permeability, theeason why it is also known as the vascular permeability factor,he development of new vessels depends on the participation ofngiogenic stimuli coming from other factors, such as epidermal,epatic, placentary, and fibroblastic growth factors, in addition tolasminogen activators [17]. Nevertheless, in the present study weoncentrated on the role of VEGF in the architecture of pleurodesisnduced by talc or silver nitrate, considering that the inactivationf this potent inflammatory agent should influence the collapse ofhe pleural space.

Within this context, the extrapolation of the results obtained innimal experimentation for clinical practice allows one to infer thathe induction of pleurodesis in patients who use drugs that blockEGF [11,18,19] may result in incomplete symphysis of the pleu-al cavity. When the sclerosing agent used is talc, which is knowno produce less pleural exudation, the reduction in angiogenesisxhibits little influence on the deposition of collagen, and whats predominantly noted is a reduction in pleural adhesions, sug-esting incomplete collapse of the pleural cavity. However, despitehe observation of similar results, when silver nitrate is used, therchitectural changes of the pleura are more exuberant. The visibleecrement in number of neoformed vessels, the smaller depositionf collagen fibers with a resulting reduction in pleural thickness,ulminates with a slighter pleurodesis and a smaller number ofdhesions between the pleural layers.

Despite the limitations of this study, among them the method-logy (induction of pleurodesis in animals with normal pleura), ouresults call attention to a possible lack of therapeutic success whenhe objective is the symphysis between the visceral and parietal

[

er 74 (2011) 392– 395 395

pleural layers in patients with malignant pleural effusion who areusing VEGF blockers.

Financial support

Research Support Foundation of the State of São Paulo (FAPESP)and National Research Council (CNPq), Brazil.

Conflict of interest statement

We have no conflicts of interest.

Acknowledgements

We thank the biologists Carlos S.R. Silva and Vanessa Alvarengaand pharmacist Gabriela G. Carnevale for help animal care and withpreparation and storage of tissue samples.

References

[1] Light RW, Vargas FS. Pleural sclerosis for the treatment of pneumothorax andpleural effusion. Lung 1997;175:213–23.

[2] Marchi E, Vargas FS, Teixeira LR, Acencio MMP, Antonangelo L, Light RW.Intrapleural low-dose silver nitrate elicits more pleural inflammation and lesssystemic inflammation than low-dose talc. Chest 2005;128:1798–804.

[3] Marchi E, Vargas FS, Acencio MM, Antonangelo L, Teixeira LR, Genofre EH, et al.Talc and silver nitrate induce systemic inflammatory effects in the acute phaseof experimental pleurodesis in rabbits. Chest 2004;125:2268–77.

[4] Breier G, Albrecht U, Sterrer S, Risau W. Expression of vascular endothelialgrowth factor during angiogenesis and endothelial cell differentiation. Devel-opment 1992;114:521–32.

[5] Ferrara N. Molecular and biological properties of vascular endothelial growthfactor. J Mol Med 1999;77:527–43.

[6] Cheng DS, Rodriguez RM, Perkett EA, Rogers J, Bienvenu G, Lappalainen U, et al.Vascular endothelial growth factor in pleural fluid. Chest 1999;116:760–5.

[7] Grove CS, Lee YCG. Vascular endothelial growth factor: the key mediator inpleural effusion formation. Curr Opin Pulm Med 2002;8:294–301.

[8] Ribeiro SCC, Vargas FS, Antonangelo L, Marchi E, Genofre EH, Acencio MM,et al. Monoclonal anti-vascular endothelial growth factor reduces fluid vol-ume in experimental model of inflammatory pleural effusion. Respiration2009;14:1188–93.

[9] Guo YB, Kalomenidis I, Hawthorne M, Parman KS, Lane KB, Light RW. Pleu-rodesis is inhibited by vascular endothelial growth factor antibody. Chest2005;128:1790–7.

10] Guo YB, Xie CM, Light RW. Effect of anti-vascular endothelial growth factoron pleurodesis induced by transforming growth factor-beta or doxycicline inrabbits. Zhonghua Jie He He Hu Xi Za Zhi 2006;29:39–43.

11] Reck M, Crino L. Advances in anti-VEGF and anti-EGFR therapy for advancednon-small cell lung cancer. Lung Cancer 2009;63:1–9.

12] Teixeira LR, Wu W, Cheng DS, Light RW. The effect of corticosteroids on pleu-rodesis induced by doxycycline in rabbits. Chest 2002;121:216–9.

13] Teixeira LR, Vargas FS, Acencio MMP, Bumlai RUM, Antonangelo L, Marchi E.Experimental pleurodesis induced by antibiotics (macrolides or quinolones).Clinics 2006;61:559–64.

14] Antonangelo L, Vargas FS, Teixeira LR, Acencio MM, Vaz MA, Filho MT, et al.Pleurodesis induced by talc or silver nitrate: evaluation of collagen and elasticfibers in pleural remodeling. Lung 2006;184:105–11.

15] Detmar M, Brown LF, Schön MP, Elicker BM, Velasco P, Richard L, et al. Increasedmicrovascular density and enhanced leukocyte rolling and adhesion in the skinof VEGF transgenic mice. J Invest Dermatol 1998;111:1–6.

16] Lee YCG, Devore R, Novotny W, Light RW. The effects of recombinant humanmonoclonal antibodies against vascular endothelial growth factor on malig-nant pleural effusions in patients with non-small cell lung carcinoma [abstract].Chest 2001;120(Suppl.):193S.

17] Folkman J, Klagsbrun M. Angiogenic factors. Science 1987;235:442–7.18] Ferrara N, Hillan KJ, Gerber HP, Novotny W. Discovery and development of

Bevacizumab, an anti-VEGF antibody for treating cancer. Nat Rev Drug Discov2004;3:391–400.

19] Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J, Heim W,Berlin J, et al. Bevacizumab plus irinotecan, fluorouracil and leucovorin formetastatic colorectal cancer. N Engl J Med 2004;350:2335–42.