Vol. 106 No. 6 December 2008

ENDODONTOLOGY Editor: Larz S. W. Spångberg

Immunoexpression of vascular endothelial growth factor inperiapical granulomas, radicular cysts, and residualradicular cysts

Cassiano Francisco Weege Nonaka, DDS, MSc,a Alexandre Pinto Maia, DDS, MSc,a

George João Ferreira do Nascimento, DDS, MSc,a Roseana de Almeida Freitas, DDS, PhD,b

Lélia Batista de Souza, DDS, PhD,b and Hébel Cavalcanti Galvão, DDS, PhD,b Natal, BrazilORAL PATHOLOGY DEPARTMENT, FEDERAL UNIVERSITY OF RIO GRANDE DO NORTE

Objective. Our aim was to assess and compare the immunoexpression of vascular endothelial growth factor (VEGF) inperiapical granulomas (PGs), radicular cysts (RCs), and residual radicular cysts (RRCs), relating it to the angiogenicindex and the intensity of the inflammatory infiltrate.Study design. Twenty PGs, 20 RCs, and 10 RRCs were evaluated by immunohistochemistry using anti-VEGF antibody.Angiogenic index was determined by microvessel count (MVC) using anti–von Willebrand factor antibody.Results. The PGs and RCs showed higher expression of VEGF than the RRCs. Lesions presenting few inflammatoryinfiltrate revealed the lowest immunoexpression of VEGF (P � .05). Irrespective of the intensity of the inflammatoryinfiltrate, most of the RCs and RRCs showed moderate to strong epithelial expression of VEGF. Lesions showing denseinflammatory infiltrate presented higher MVC indices (P � .05). VEGF expression and MVC did not reveal a significantcorrelation (P � .05).Conclusions. VEGF is present in periapical inflammatory lesions but at a lower level in RRCs. The expression of thisproangiogenic factor is closely related to the intensity of the inflammatory infiltrate in these lesions. (Oral Surg Oral

Med Oral Pathol Oral Radiol Endod 2008;106:896-902)

Periapical lesions, which include periapical granulomas(PGs) and radicular cysts (RCs), occur as a result of theimmunologic response to continuous antigenic stimu-lation from root canals.1,2 Persistence of the inflamma-tion is associated with resorption of adjacent bone,which is replaced by a granulation tissue to form a PG.2

As a consequence of inflammatory and immunologicresponses, the epithelial remnants of Malassez are stim-ulated to proliferate, which may result in the develop-ment of an RC.3 By definition, residual radicular cysts(RRCs) represent RCs that are inadvertently left behindafter the extraction of the involved tooth.4,5

aDoctoral student.bProfessor.Received for publication Apr 26, 2008; returned for revision Jun 28,2008; accepted for publication Jun 30, 2008.1079-2104/$ - see front matter© 2008 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2008.06.028

896

Vascular endothelial growth factor (VEGF) is a po-tent proangiogenic cytokine secreted by many celltypes which presents several pivotal functions in phys-iologic and pathologic angiogenesis.6-8 VEGF acts onthe vasculature by inducing the proliferation, differen-tiation, and migration of vascular endothelial cells.8-10

In addition, VEGF can induce microvascular perme-ability, leading to extravasation of plasma proteins,fluid accumulation, and edema.7,9,10 Therefore, VEGFhas been implicated as an important factor in granula-tion tissue development.8,11,12

Recent studies suggest a role for VEGF in enlarge-ment of cysts. It has been speculated that the presenceof VEGF in cystic lesions is capable of increasingvascular permeability, leading to accumulation of cys-tic fluid.13-17 Leonardi et al.18 proposed a similar func-tion for VEGF in periapical lesions. According to thoseauthors, the expression of VEGF in PGs and RCs has

bioactivity to increase vascular permeability, being at

OOOOEVolume 106, Number 6 Nonaka et al. 897

least partially involved in the accumulation of inflam-matory cells and cyst fluid.

The importance of vascular networks in the devel-opment and maintenance of tissues have been demon-strated in many physiologic and pathologic processes,such as embryogenesis, wound healing, inflammation,and tumor progression.6-8,10,11 Angiogenesis is the pro-cess by which new blood vessels are produced bysprouting from preexisting vasculature,7,8,10 and themeasurement of its role may be important for a betterunderstanding of the pathogenesis of neoproliferativelesions.19 Although angiogenesis cannot be measureddirectly, it can be assessed by quantification of thevessels immunolabeled by specific antibodies againstendothelial cell epitopes, such as von Willebrand factor(vWF).20 In this manner, it is possible to indirectlyestablish the angiogenic activity of a tissue.20,21

To our knowledge, there are no studies analyzing theexpression of VEGF in PGs, RCs, and RRCs. There-fore, the aim of the present study was to assess andcompare the immunoexpression of VEGF in these le-sions, relating it to the angiogenic index and the inten-sity of the inflammatory infiltrate.

MATERIALS AND METHODSFifty tissue specimens, 20 PGs, 20 RCs and 10

RRCs, archived in the files of the Oral Pathology De-partment of the Federal University of Rio Grande doNorte (UFRN) were randomly selected for this study.All PGs and RCs were obtained from teeth withoutendodontic treatment. Only PGs without odontogenicepithelium were selected. Moreover, all RCs and RRCspresented unequivocal cystic cavity lined by odonto-genic epithelium. The cases were not matched for age,gender, or anatomic location. Serial sections, 3-5 �mthick, were taken from tissue blocks and processed formorphologic and immunohistochemical studies. Thestudy was approved by the Research Ethics Committeeof the UFRN, Natal, Brazil.

Morphologic analysisFor the morphologic analysis, tissue sections were

stained with hematoxylin and eosin. The intensity ofthe inflammatory infiltrate was evaluated according tothe method proposed by Tsai et al.4 Briefly, each spec-

Table I. Specificity, dilution, antigen retrieval, and incAb clone Specificity Dilution

C-1* VEGF 1:400F8/86† vWF 1:50

*Santa Cruz Biotechnology.†Dako Cytomation.

imen was graded at �200 magnification as: grade I,

inflammatory cells less than one-third per field; gradeII, inflammatory cells between one-third and two-thirdsper field; and grade III, inflammatory cells more thantwo-thirds per field. Grading of each specimen wasrecorded on the average inflammatory condition in 3consecutive microscopic fields, starting from the innerportion of the specimen and proceeding deeper intoconnective tissue. Epithelial thickness was defined asatrophic (2-10 cell layers) or hyperplastic (�10 celllayers), according to Moreira et al.22

Immunohistochemical methodsFor the immunohistochemical study, tissue sections

were deparaffinized and immersed in methanol with0.3% hydrogen peroxide to block endogenous peroxi-dase activity. The tissue sections were then washed inphosphate-buffered saline (PBS). The antigen retrieval,antibody dilution, and clone type for VEGF and vWFare shown in Table I. After treatment with normalserum, the tissue sections were incubated in a moistchamber with primary antibodies. The tissue sectionswere then washed twice in PBS and treated with strep-toavidin-biotin-peroxidase complex (SABC; Dako,Glostrup, Denmark) at room temperature to bind theprimary antibodies. Peroxidase activity was visualizedby immersing tissue sections in diaminobenzidine(D5637; Sigma Chemical, St. Louis, MO), resulting ina brown reaction product. Finally, tissue sections werecounterstained with Mayer hematoxylin and cover-slipped. Positive control samples for VEGF and vWFwere sections of normal human kidney. As negativecontrol, samples were treated as above, except that theprimary antibody was replaced by a solution of bovineserum albumin in PBS.

Immunostaining assessment and statisticalanalysis

After the immunohistochemical treatment, the tissuesections were examined by light microscopy. Immuno-histochemical expression of VEGF was evaluated bothin the connective tissue of PGs, RCs, and RRCs and inthe epithelial lining of RCs and RRCs. In the connec-tive tissue, a quantitative assessment of the immunopo-sitive cells, irrespective of the color intensity, wasperformed according to the method proposed by Freitas

n of the antibody (Ab) clonesAntigen retrieval Incubation

rypsin pH 7.9, oven 37°C, 60 min Overnight (18 h)itrate pH 6.0, steamer, 30 min 60 min

ubatio

TC

et al.20 Tissue sections were examined by light micros-

OOOOE898 Nonaka et al. December 2008

copy using �100 magnification to identify 5 fields withthe largest number of immunostained cells. Using�400 magnification, the counting of the immunoposi-tive cells was performed in each one of these fields.

The immunoexpression of VEGF in the epitheliallining of RCs and RRCs was semiquantitatively eval-uated, using �100 magnification. Performing an adap-tation of the method proposed by Leonardi et al.,18 theepithelial immunoexpression of VEGF was classifiedaccording to the following parameters: no staining in�10% of cells; weak, staining in 11%-25% of cells;moderate, staining in 26%-75% of cells; strong, stain-ing in �76% of cells.

Angiogenic index in PGs, RCs, and RRCs was de-termined based on the number of vessels immunoreac-tive to anti-vWF antibody. Adopting the methodologyproposed by Freitas et al.,20 a microvessel count(MVC) was performed. Tissue sections were examinedby light microscope at �40 magnification, and 5 areasshowing the highest vascularization were identifiedsubjectively. In these areas, vessels were counted at�200 magnification.

The results obtained were submitted to statisticalanalysis. Computations were made using the StatisticalPackage for the Social Sciences (SPSS 13.0). To ana-lyze the immunohistochemical expression of VEGF inthe epithelial lining of RCs and RRCs, Fisher exact testwas performed. To compare the number of cells immu-noreactive for VEGF in the connective tissue of PGs,RCs, and RRCs, the nonparametric Kruskal-Wallis testwas performed. The analysis of the number of vesselsin PGs, RCs, and RRCs was evaluated by the Kruskal-Wallis and Mann-Whitney nonparametric tests. Finally,Spearman correlation test was performed to verify pos-sible correlations between the number of immunoposi-tive cells for VEGF and the number of MVC. For alltests, significance level was set at .05 (P � .05).

RESULTSMorphologic analysis

Analysis of the inflammatory infiltrate in PGs re-vealed 19 cases (95%) with inflammatory infiltrategrade III and only 1 specimen (5%) with inflammatoryinfiltrate grade II. In RCs, 14 cases (70%) showedinflammatory infiltrate grade III, 5 cases (25%) inflam-matory infiltrate grade II, and only 1 case (5%) inflam-matory infiltrate grade I. In the group of RRCs, 4 cases(40%) showed inflammatory infiltrate grade I, 3 cases(30%) inflammatory infiltrate grade II, and 3 cases(30%) inflammatory infiltrate grade III.

Morphologic analysis of the epithelial thickness inRCs revealed the presence of a hyperplastic epitheliumin 14 cases (70%) and an atrophic epithelium in only 6

cases (30%). In RRCs, 6 cases (60%) presented an

atrophic epithelium and 4 cases (40%) showed a hy-perplastic epithelium.

Considering all cystic lesions with hyperplastic epi-thelium (n � 18), 13 cases (72.2%) presented inflam-matory infiltrate grade III, 3 cases (16.7%) inflamma-tory infiltrate grade II, and 2 cases (11.1%)inflammatory infiltrate grade I. Considering cystic le-sions with atrophic epithelium (n � 12), 5 cases(41.7%) presented inflammatory infiltrate grade II, 4cases (33.3%) inflammatory infiltrate grade III, and 3cases (25%) inflammatory infiltrate grade I.

Immunohistochemical analysisThe mean number of cells immunostained with anti-

VEGF antibody was 564.90 (range 254-916) and565.05 (range 214-792) in PGs and RCs, respectively.In RRCs, the mean number of immunopositive cellswas 443.90 (range 222-660). Grouping all lesions ac-cording to the intensity of the inflammatory infiltrate,the mean number of immunopositive cells for VEGFwas 390.40 (range 222-548) in lesions with inflamma-tory infiltrate grade I (n � 5) (Fig. 1). In lesions withinflammatory infiltrate grade II (n � 9), the meannumber of immunoreactive cells was 428.22 (range214-792). In lesions with inflammatory infiltrate gradeIII (n � 36), the mean number of immunopositive cellswas 589.78 (range 314-916) (Fig. 2).

The analysis of the immunoreactivity of VEGF in theepithelium of RCs revealed weak expression in 2 cases(10%), moderate expression in 7 cases (35%), andstrong expression in 11 cases (55%). Only moderateand strong immunoreactivities for VEGF were ob-served in the epithelium of RRCs. Three cases (30%)presented moderate expression and 7 cases (70%)

Fig. 1. Immunoexpression of VEGF in specimen of residualradicular cyst presenting inflammatory infiltrate grade I(SABC method, original magnification �400).

strong expression.

OOOOEVolume 106, Number 6 Nonaka et al. 899

Grouping RCs and RRCs according to epitheliumthickness, in lesions with hyperplastic epithelium (n �18) 8 cases (44.4%) presented moderate epithelial ex-pression of VEGF and 10 cases (55.6%) strong immu-noreactivity (Fig. 3). In lesions with atrophic epithe-lium (n � 12), 2 cases (16.7%) showed weak epithelialexpression of VEGF, 2 cases (16.7%) moderate expres-sion, and 8 cases (66.6%) strong expression (Fig. 4).

The mean number of blood vessels determined byMVC in PGs and RCs was 210.45 (range 124-279) and250.85 (range 159-350), respectively (Fig. 5). In RRCs,the mean number of blood vessels was 217.00 (range114-366). Grouping all lesions according to the inten-sity of the inflammatory infiltrate, the mean number ofblood vessels was 192.20 (range 152-252) in lesions

Fig. 2. Immunoexpression of VEGF in specimen of periapi-cal granuloma presenting inflammatory infiltrate grade III(SABC method, original magnification �400).

Fig. 3. Strong epithelial expression of VEGF in specimen ofresidual radicular cyst with hyperplastic epithelial lining(SABC method, original magnification �400).

with inflammatory infiltrate grade I (n � 5). In lesions

with inflammatory infiltrate grade II (n � 9), the meannumber of blood vessels was 202.89 (range 114-350).Lesions with inflammatory infiltrate grade III (n � 36)showed a mean number of blood vessels of 239.14(range 124-409).

Statistical analysisAnalysis of the number of immunopositive cells for

VEGF in the connective tissue of PGs, RCs, and RRCs,performed by nonparametric Kruskal-Wallis test,showed no statistically significant difference (P � .05)(Table II). Nevertheless, comparison of the number ofimmunopositive cells for VEGF according to the inten-sity of the inflammatory infiltrate revealed a statisticallysignificant difference between groups (P � .05) (Table

Fig. 4. Strong epithelial expression of VEGF in specimen ofradicular cyst with atrophic epithelial lining (SABC method,original magnification �400).

Fig. 5. Vessels immunoreactive to anti-vWF antibody inspecimen of residual radicular cyst (SABC method, originalmagnification �100).

III).

OOOOE900 Nonaka et al. December 2008

Fisher exact test was performed to compare the ep-ithelial expression of VEGF in RCs and RRCs. Nodifferences were observed between groups (P � .05).In addition, comparison of the immunoexpression ofVEGF according to the epithelial thickness revealed nostatistically significant differences between lesions withatrophic epithelium and lesions with hyperplastic epi-thelium (P � .05).

Regarding the number of blood vessels of PGs, RCs,and RRCs, nonparametric Kruskal-Wallis test revealedno difference between groups (P � .05). In addition,comparison of the number of blood vessels according tothe intensity of the inflammatory infiltrate did not re-veal significant differences (P � .05). Nevertheless,when lesions showing inflammatory infiltrate grades Iand II were grouped and compared with those present-ing inflammatory infiltrate grade III, nonparametricMann-Whitney test revealed a significant difference (P� .05) (Table IV).

Spearman correlation test disclosed no significantcorrelation between the number of immunopositivecells for VEGF and the angiogenic index (P � 0.118).

DISCUSSIONDespite intense investigation,2,3,22-24 the molecular

mechanisms involved in the formation of granulationtissue and enlargement of jaw cysts have not been fullyclarified. Recent studies suggest an important role forVEGF in the pathogenesis of PGs and RCs.18,19 Adimeric glycoprotein with a selective mitogenic effecton vascular endothelial cells, VEGF is an extremelypotent and specific angiogenic factor.6-10 Moreover,VEGF is capable of inducing microvascular permeabil-ity with a potency some 50,000 times that of hista-mine, leading to extravasation of plasma proteins anda predictable sequence of proangiogenic stromalchanges.6,18 Therefore, VEGF has been implicated asan important factor in granulation tissue develop-ment7,8,11,12 and cyst enlargement.13-17

In the present study, we evaluated the expression of

Table II. Parameters used for the calculation of theKruskal-Wallis (KW) test for the evaluation of thenumber of immunopositive cells for VEGF in PGs,RCs, and RRCs

Lesion n Median Q25-Q75

Mean ofthe ranks KW

Pvalue

PG 20 586.00 406.25–690.25 27.55 3,315 .191RC 20 590.50 505.75–688.75 27.20RRC 10 460.00 256.00–618.50 18.00

PG, Periapical granuloma; RC, radicular cyst; RRC, residual radicularcyst; Q, quartile.

VEGF in PGs, RCs, and RRCs. The results revealed

that VEGF is expressed in the connective tissue of all ofthese lesions and in the epithelial lining of RCs andRRCs, corroborating earlier studies.18,19 According toLeonardi et al.,18 the expression of VEGF in periapicallesions has bioactivity to increase vascular permeabil-ity, being partially involved in the accumulation ofinflammatory cells and cyst fluid. Therefore, in accor-dance with the biologic functions described for VEGF,such as inducing microvascular permeability and beinga potent mitogenic factor for endothelial cells,6-8 thepresent findings suggest the participation of this proan-giogenic factor in the development and enlargement ofPGs, RCs, and RRCs.

Despite the absence of statistical significance (P �.05), PGs and RCs presented a higher mean number ofimmunopositive cells for VEGF compared with RRCs.PGs and RCs present a constant source of stimulationby bacterial toxins released from the infected root canalwhich are not present in RRCs.3 In addition, Muglali etal.3 verified higher levels of interleukin-1�, monocytechemotactic protein 1, and regulated upon activationnormal T-cell expressed and secreted (RANTES) incystic fluids of RCs compared with cystic fluids ofRRCs. Interleukin-1� is a cytokine that has beeninvolved in up-regulation of the expression ofVEGF.7,25,26 Monocyte chemotactic protein 1 andRANTES are important chemokines that regulate mac-rophage infiltration in tissues.3,27 Besides their func-tions as antigen-presenting cells and phagocytes duringwound repair, macrophages present a proangiogenicpotential,28 suggesting a functional contribution ofthese cells in wound angiogenesis.27,29 Therefore, thelow mean number of immunopositive cells for VEGF inRRCs observed in the present study might be related toa decreased level of cytokines in these lesions, as aconsequence of the reduction of the inflammatory stim-uli.

Leonardi et al.18 verified marked differences in theexpression of VEGF between PGs and RCs. In theirstudy, many inflammatory cells in PGs were immu-nopositive for VEGF, but there was a progressive de-crease in the number of such immunolabeled cells toalmost none in RCs. Therefore, Leonardi et al.18 sug-gested that inflammatory cells and fibroblasts would beresponsible for synthesis of VEGF in the early phasesof periapical lesion development and epithelial cells inthe later ones.

In the present study, we could not verify such pro-gressive decrease in the immunoexpression of VEGF.The differences between our results and those reportedby Leonardi et al.18 might be due to the small sample ofRCs used in their research. Moreover, Leonardi et al.18

did not report if the intensity of the inflammatory in-

filtrate was evaluated in their study. In the present

.75–259

OOOOEVolume 106, Number 6 Nonaka et al. 901

research, lesions with inflammatory infiltrate grade Irevealed a lower number of immunopositive cells forVEGF than lesions showing inflammatory infiltrategrade III (P � .05). Similarly, Graziani et al.19 verifieda positive correlation between the intensity of the in-flammatory infiltrate and the expression of VEGF inperiapical lesions.

Most of the RCs and RRCs evaluated in the presentstudy presented a moderate to strong epithelial expressionof VEGF, irrespective of the intensity of the inflammatoryinfiltrate or the thickness of the epithelial lining. Theepithelial expression of VEGF in RCs and RRCs mightconstitute an additional mechanism for the enlargement ofthese lesions, maintaining the stimulus for angiogenesisand vascular permeability, a role also proposed by Leo-nardi et al.18 Results from studies of the expression ofVEGF in brain tumor cysts13,15,17 and cyst fluid of thyroidnodules14,16 also corroborate this suggestion. Accordingto Vaquero et al.,15 VEGF might enhance accumulation ofcyst fluid, together with an increase in oxygen supplyboosting the development of the cyst.

The epithelium status of the RCs (hyperplastic or atro-phic) has been suggested as a reliable histologic parameterof biologic activity and/or inactivity of cystic growth.22,30

On quiescent lesions (atrophic epithelium), despite thepresence of antigens and enzymes able to induce immu-nologic responses, epithelium proliferation, and bone re-sorption, cyst enlargement does not occur.22 In suchphases, immunosuppressor effectors22 or apoptoticevents31 may operate to regulate cyst enlargement. De-spite those reports, the present results revealed that mostof the RCs and RRCs with atrophic epithelial liningpresent strong epithelial expression of VEGF. These find-ings might suggest the existence of a potential for expan-sion in cystic lesions with atrophic epithelium.

The present results showed that lesions with a dense

Table III. Parameters used for the calculation of the Kimmunopositive cells for vascular endothelial growth fInflammatory infiltrate n Median

Grade I 5 373.00 242Grade II 9 298.00 246Grade III 36 611.00 542

Q, Quartile.

Table IV. Parameters used for the calculation of the Maimmunostained by anti–von Willebrand factor antibodyInflammatory infiltrate n Median

Grade I/II 14 191.00 155Grade III 36 232.50 198

Q, Quartile.

inflammatory infiltrate (grade III) present both a high

number of immunopostive cells for VEGF (P � .05)and a high MVC (P � .05). These findings are inaccordance with other studies and highlight the impor-tance of inflammatory cells, particularly neutrophilsand macrophages,27 on VEGF expression.19,32 In thisway, the low level of inflammation, probably due toreduced antigenic stimulation,3 could be suggested as apossible explanation for the low expression of VEGF,and consequently low MVC, in RRCs.

Graziani et al.19 observed a positive correlation be-tween the microvessel density and the expression ofVEGF. However, in our research, we could not find astatistically significant correlation between MVC andthe number of immunopositive cells for VEGF. Thedifferences between our results and those reported byGraziani et al.19 could be related to differences in themethods selected for evaluation of the immunoreactiv-ity. In addition, differing from our study, Graziani etal.19 only evaluated RCs.

Despite the absence of statistical significance, RCspresented a higher number of blood vessels comparedwith PGs and RRCs. To our knowledge, there are noprevious studies in the literature that compared theangiogenic index between PGs, RCs, and RRCs. Thelow MVC observed in RRCs could be related to thelower number of cells immunopositive for VEGF inthese lesions compared with RCs. However, PGs andRCs showed a similar number of cells immunopositivefor VEGF. It could be hypothesized that the differencein the number of blood vessels might be related to apossible conversion of granulation tissue to a cyst. Inthis way, it can be speculated that the limits of neovas-cularization are reached and, as a result, the center ofthe granuloma necroses. Additional studies are neces-sary to clarify these observations.

In conclusion, our results revealed that VEGF is

l-Wallis (KW) test for the evaluation of the number ofaccording to the grade of the inflammatory infiltrate

Mean of the ranks KW P value

.50 11.70 9.736 .008

.00 17.56

.25 29.40

hitney U test for the evaluation of the number of vesselsrding to the grade of the inflammatory infiltrate.

Mean of the ranks U P value

.50 18.57 155.00 .036

.00 28.19

ruskaactor

Q25-Q75

.00–547

.00–633

.75–690

nn-Wacco

Q25-Q75

.00–231

present in periapical inflammatory lesions, but at a

OOOOE902 Nonaka et al. December 2008

lower level in RRCs. The expression of this pro-angio-genic factor is closely related to the intensity of theinflammatory infiltrate in these lesions. Further studiesneed to be performed to clarify the role of VEGF andother angiogenic factors in the pathogenesis of PGs,RCs, and RRCs.

REFERENCES1. Garcia CC, Sempere FV, Diago MP, Bowen EM. The post-

endodontic periapical lesions: histologic and etiopathogenic as-pects. Med Oral Patol Oral Cir Bucal 2007;12:585-90.

2. Walker KF, Lappin DF, Takahashi K, Hope J, Macdonald DG,Kiname DF. Cytokine expression in periapical granulation tissueas assessed by immunohistochemistry. Eur J Oral Sci 2000;108:195-201.

3. Muglali M, Komerik N, Bulut E, Yarim GF, Celebi N, Sumer M.Cytokine and chemokine levels in radicular and residual cysts.J Oral Pathol Med 2008;37:185-9.

4. Tsai CH, Weng SF, Yang LC, Huang FM, Chen YJ, Chang YC.Immunohistochemical localization of tissue-type plasminogenactivator and type I plasminogen activator inhibitor in radicularcysts. J Oral Pathol Med 2004;33:156-61.

5. Moldauer I, Velez I, Kuttler S. Upregulation of basic fibroblastgrowth factor in human periapical lesions. J Endod2006;32:408-11.

6. Dvorak HF, Detmar M, Claffey KP, Nagy JA, van de Water L,Senger DR. Vascular permeability factor/vascular endothelialgrowth factor: an important mediator of angiogenesis in malig-nancy and inflammation. Int Arch Allergy Immunol 1995;107:233-5.

7. Hoeben A, Landuyt B, Highley MS, Wildiers H, Van OosteromAT, De Bruijn EA. Vascular endothelial growth factor andangiogenesis. Pharmacol Rev 2004;56:549-80.

8. Byrne AM, Bouchier-Hayes DJ, Harmey JH. Angiogenic andcell survival functions of vascular endothelial growth factor(VEGF). J Cell Mol Med 2005;9:777-94.

9. Ferrara N, Davis-Smyth T. The biology of vascular endothelialgrowth factor. Endocrine reviews 1997;18:4-25.

10. Takahashi H, Shibuya M. The vascular endothelial growth factor(VEGF)/VEGF receptor system and its role under physiologicaland pathological conditions. Clin Sci 2005;109:227-41.

11. Pokharel RP, Maeda K, Yamamoto T, Noguchi K, Iwai Y,Nakamura H et al. Expression of vascular endothelial growthfactor in exuberant tracheal granulation tissue in children.J Pathol 1999;188:82-6.

12. Nissen NN, Polverini PJ, Koch AE, Volin MV, Gamelli RL, DiPietro LA. Vascular endothelial growth factor mediates angio-genic activity during the proliferative phase of wound healing.Am J Pathol 1998;152:1445-52.

13. Strugar JG, Criscuolo GR, Rothbart D, Harrington WN. Vascularendothelial growth/ permeability factor expression in humanglioma specimens: correlation with vasogenic brain edema andtumor-associated cysts. J Neurosurg 1995;83:682-9.

14. Sato K, Miyakawa M, Onoda N, Demura H, Yamashita T, MiuraM, et al. Increased concentration of vascular endothelial growthfactor/vascular permeability factor in cyst fluid of enlarging andrecurrent thyroid nodules. J Clin Endocrinol Metabol 1997;82:1968-73.

15. Vaquero J, Zurita M, Oya S, Coca S, Salas C. Vascular perme-ability factor expression in cerebellar hemangioblastomas: cor-relation with tumor-associated cysts. J Neurooncol 1999;41:3-7.

16. Hofmann A, Gessl A, Girschele F, Novotny C, Kienast O,Staudenherz A, et al. Vascular endothelial growth factor in

thyroid cyst fluids. Wien Klin Wochenschr 2007;119:248-53.

17. Stockhammer G, Obwegeser A, Kostron H, Schumacher P,Muigg A, Felber S, et al. Vascular endothelial growth factor(VEGF) is elevated in brain tumor cysts and correlates withtumor progression. Acta Neuropathol 2000;100:101-5.

18. Leonardi R, Caltabiano M, Pagano M, Pezzuto V, Loreto C,Palestro G. Detection of vascular endothelial growth factor/vascular permeability factor in periapical lesions. J Endod2003;29:180-3.

19. Graziani F, Vano M, Viacava P, Itro A, Tartaro G, Gabriele M.Microvessel density and vascular endothelial growth factor(VEGF) expression in human radicular cysts. Am J Dent2006;19:11-4.

20. Freitas TMC, Miguel MCC, Silveira EJD, Freitas RA, GalvãoHC. Assessment of angiogenic markers in oral hemangiomas andpyogenic granulomas. Exp Mol Pathol 2005;79:79-85.

21. El-Gazzar RF, Macluskey M, Ogden GR. The effect of theantibody used and method of quantification on oral mucosalvascularity. Int J Oral Maxillofac Surg 2005;34:895-9.

22. Moreira PR, Santos DFM, Martins RD, Gomez RS. CD57� inradicular cyst. Int Endod J 2000;33:99-102.

23. Rodini CO, Lara VS. Study of the expression of CD68� macro-phages and CD8� T cells in human granulomas and periapicalcysts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod2001;92:221-7.

24. Danin J, Linder LF, Lundqvist G, Andersson L. Tumor necrosisfactor-alpha and transforming growth factor-beta 1 in chronicperiapical lesions. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2000;90:514-7.

25. Elaraj DM, Weinreich DM, Varghese S, Puhlmann M, HewittSM, Carroll NM, et al. The role of interleukin 1 in growth andmetastasis of human cancer xenografts. Clin Cancer Res2006;12:1088-96.

26. Ma J, Sawai H, Matsuo Y, Ochi N, Yasuda A, Takahashi H, etal. Interleukin-1� enhances angiogenesis and is associated withliver metastatic potential in human gastric cancer cell lines.J Surg Res 2008;148:197-204.

27. Eming SA, Brachvogel B, Odorisio T, Koch M. Regulation ofangiogenesis: wound healing as a model. Prog Histochem Cyto-chem 2007;42:115-70.

28. Moldovan L, Moldovan NI. Role of monocytes and macrophagesin angiogenesis. In: Clauss M, Breier G, editors. Mechanismsof angiogenesis. Massachusetts: Birkhäuser Verlag; 2005. p.127-46.

29. Crowther M, Brown NJ, Bishop ET, Lewis CE. Microenviro-mental influence on macrophage regulation of angiogenesis inwounds and malignant tumors. J Leukoc Biol 2001;70:478-90.

30. Cury VC, Sette PS, Silva JV, Araújo VC, Gomez RS. Immuno-histochemical study of apical periodontal cysts. J Endod1998;24:36-7.

31. Loyola AM, Cardoso SV, Lisa GS, Oliveira LJ, Mesquita RA,Carmo MAV, et al. Apoptosis in epithelial cells of apical radic-ular cysts. Int Endod J 2005;38:465-9.

32. Artese L, Rubini C, Ferrero G, Fioroni M, Santinelli A, PiattelliA. Vascular endothelial growth factor (VEGF) expression inhealthy and inflamed human dental pulps. J Endod 2002;28:20-3.

Reprint requests:

Profa. Dra. Hébel Cavalcanti GalvãoDepartamento de OdontologiaAv. Senador Salgado Filho, 1787Lagoa Nova-Natal-RNBrasilCEP 59056-000.

hebel.galvao@yahoo.com.br