1 "This is the peer reviewed version of the following article: Ribeiro JL, Moraes RM, Carvalho BFC, Nascimento AO, Milhan NVM, Anbinder AL. Oral pyogenic granuloma: an 18-year retrospective clinicopathological and immunohistochemical study. J Cutan Pathol. 2021;Jul;48(7):863-869, which has been published in final form at https://doi.org/10.1111/cup.13970. This article may be used for non- commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions." Oral Pyogenic Granuloma: an 18-Year Retrospective Clinicopathological and Immunohistochemical Study Running title: Oral Pyogenic Granuloma: retrospective study Authors: Jaqueline Lemes Ribeiro, Renata Mendonça Moraes, Bruna Fernandes do Carmo Carvalho, Anderson de Oliveira Nascimento, Noala Vicensoto Moreira Milhan, Ana Lia Anbinder. Affiliation: Department of Bioscience and Oral Diagnosis, Institute of Science and Technology, São Paulo State University (Unesp), São José dos Campos, São Paulo, Brazil. Correspondence: Ana Lia Anbinder Avenida Engenheiro Francisco José Longo, nº 777 – Jardim São Dimas São José dos Campos, São Paulo, Brazil Zip Code: 12245-000 ana.anbinder@unesp.br Conflict of Interest: None. Acknowledgment: This project was supported by Coordination of Improvement of Higher Education Personnel CAPES. Funding information: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – CAPES. mailto:ana.anbinder@unesp.br 2 Oral pyogenic granuloma: an 18-year retrospective clinicopathological and immunohistochemical study Abstract Background: Pyogenic granuloma (PG) is a lesion characterized by the proliferation of blood vessels, commonly affecting the skin and the mouth. We aimed to compare clinical, microscopic, and immunohistochemical features of the two types of oral PG: lobular capillary hemangioma (LCH) and non-LCH (NLCH). Methods: Epidemiological and clinical data from 2000 to 2018 were collected from the archives of our institution, and histopathological sections of PG were reviewed. Immunohistochemical analyses (CD34, D2-40, SMA, mast cell, and Ki-67) were performed in 34 cases. Results: Sixty-two LCH and 107 non-LCH samples were included. The mean (±SD) age of the patients was 38.59 ± 16.96 years; 55.62% were female; 39.64% of cases occurred in the gingiva, 44% of the nodules were pedunculated, and 13.02% of patients reported a history of trauma. NLCH was more prevalent among older patients than LCH. The most prevalent site of LCH was the lips, while NLCH occurred more in the gingiva (P < 0.05). Epithelial atrophy, microvessels, SMA-positive areas, and Ki-67-positive nuclei were more prevalent in LCH (P < 0.05). Conclusions: PG accounted for 2.25% of lesions archived in the pathology service and most cases were NLCH. LCH and NLCH exhibited clinicopathological differences in terms of age, site, epithelial atrophy, vascularization, and proliferation rate. 3 Introduction Pyogenic granuloma (PG) is a lesion characterized by blood vessels proliferation,1-4 which is quite common in the skin and oral cavity.5 Low-intensity trauma, poor oral hygiene, and hormonal factors are the most cited etiological factors for oral lesions.2,6-8 It is estimated that 3% of PGs occur during pregnancy; in these cases, the lesion is called “granuloma gravidarum”.9 The International Society for the Study of Vascular Anomalies (ISSVA) classifies PG as a benign vascular tumor, which can also be termed “lobular capillary hemangioma” (LCH).10 In the literature, PG has been classified as two histopathological variants, LCH and non-LCH (NLCH): while LCH is characterized by the proliferation of blood vessels arranged in lobular aggregates separated by bundles of fibrous connective tissue, NLCH exhibits highly vascularized proliferation that resembles granulation tissue.11-13 Although studies have evaluated the clinical, histopathological, and immunohistochemical characteristics of PG,1,4,11-16 the two types have rarely been compared in the literature,11-13 and there is still controversy whether the two histopathological variants represent, in fact, different entities.11 Some authors believe that the more organized pattern with little or no inflammation in nonulcerated lesions observed in the LCH, combined with the lack of evident etiological factors in most cases, support spontaneous development,8 while the NLCH probably represents a reactive lesion, in which the identification of the etiological factor is important to avoid recurrence or further growth of the lesion.11 Thus, we aimed to reevaluate cases of PG from our files in, to the best of our knowledge, the largest retrospective study to compare the two subtypes of PG in the English literature to date, correlating histopathological and immunohistochemical findings with epidemiological data to better understand and clinically describe these two subtypes of PG. Materials and methods The present study was approved by the Institutional Ethics Committee (protocol Nº. CAAE: 43557715.3.0000.0077). Biopsy material diagnosed as PG between January 2000 and October 2018 was selected from the authors’ Oral Pathology Laboratory archives, using the search terms “pyogenic 4 granuloma” and “lobular capillary hemangioma.” Histopathological sections stained with hematoxylin and eosin in all cases were reviewed by two examiners. The lesions were classified as NLCH or LCH according to the predominance of nonorganized granulation tissue or lobular aggregates of capillary-sized vessels separated by fibrous bundles, respectively. Cases without available slides or blocks, or with insufficient tissue remnants for new histological processing, were excluded. Clinical and epidemiological data, including sex, age, pregnancy, affected site, clinical appearance (sessile or pedunculated), and clinical diagnostic hypothesis (CDH), were collected from requisition forms. Other common aspects evaluated included the following: presence of ulceration; predominance of congested vessels or atrophic epithelium over the lesion surface; foci of eosinophilic inflammation; intravascular papillary endothelial hyperplasia; presence of superficial bacterial colonies; or plasma pools in the epithelium. Immunohistochemistry A sample size calculation was performed and 17 cases were needed to detect a significant difference at the desired confidence levels. Sections (3-µm thick) were obtained from the selected blocks on adhesive coated slides. Briefly, the sections were deparaffinized in xylol and hydrated in a descending series of ethanol solutions. Antigenic recovery was performed using the appropriate solution for each primary antibody (Table 1). The endogenous peroxidase was blocked with methanol and 20% hydrogen peroxide solution (1:1). The sections were incubated with the primary antibody (Table 1), followed by the secondary antibody (Envision – Dako, Carpinteria, California) for 30 min and, subsequently, with diaminobenzidine chromogen (DakoCytomation). Finally, they were counterstained using Mayer’s hematoxylin. Slides stained with D2-40 antibodies were evaluated descriptively, and histometric analysis was performed for the others. Histometric analysis Six images from each scanned slide (Pannoramic Desk, 3DHistech, Budapest, Hungary) stained with CD34, SMA, or mast cell antibodies were captured using the Pannoramic Viewer program 5 (3DHistech)—three located close to the ulcer, and three distant from the ulcer—at 20× zoom. In nonulcerated cases, only three images were captured. The most vascularized areas were selected for microvascular evaluation on slides stained with CD34 and SMA antibodies. The number of microvessels was counted using ImageJ 1.31p software (National Institutes of Health, Bethesda, Maryland). Clusters or single cells with positive CD34 staining, even without lumen, were considered to be countable microvessels. The mean value of the six images (or three in the case of nonulcerated lesions) was used for statistical analyses. Mast cells positively stained for the mast cell antibody were counted in the same manner as microvessels. For SMA analysis, the area of positive staining was determined with the aid of an imaging processing software (Opticam, Nashville, Tennessee). For Ki-67 analysis, five 200× fields (“hot spots”) from a scanned slide were obtained, positive stained nuclei were counted using ImageJ 1.31p, and the mean value was used for statistical analyses. Statistical analyses Epidemiological and clinical data were compared between groups (LCH and NLCH) using the chi-squared test (Sphinx, Canoas, RS, Brazil). Data, including age, number of blood vessels stained by CD34, area stained by SMA, number of mast cells, and Ki-67 positive nuclei, were compared between groups using the Student's t test. To determine the correlation between the number of mast cells and the number of blood vessels stained by CD34, the Pearson’s correlation coefficient was used (GraphPad Prism 6, La Jolla, California). For all tests, differences with P < 0.05 were considered to be statistically significant. Results The preliminary search of the archives retrieved 197 cases of PG, corresponding to 2.25% of all lesions diagnosed in the Oral Pathology Laboratory during the period. After initial analysis, 28 cases were excluded due insufficient tissue remnants for new histological processing or discordance with the previous diagnosis. Thus, 169 cases of PG were ultimately included in the present study, 107 of which were NLCH (Figure 1A) and 62 were LCH (Figure 1B). 6 Epidemiological and clinical data From the clinical data obtained (Table 2), the mean (± SD) age at involvement was 38.59 ± 16.96 years, with a slight predominance of females (1.25:1). In 8 cases, age was not reported. The most affected site was the gingiva (39.64%), followed by the lips (22.49%). The CDH provided by the clinicians were varied. The most cited hypothesis was PG (63.31% of cases), followed by inflammatory fibrous hyperplasia/irritation fibroma (23.08%), peripheral giant cell granuloma (10.65%), mucocele (8.88%), and peripheral ossifying fibroma (2.96%). CDH different from those were cited in 20.71% of cases and, in 14 (8.28%) cases, a CDH was not provided. Twenty-two (13.02%) cases recorded a history of mechanical trauma, 12 of which were LCH and 10 NLCH. Among all cases, only 17 (10.06%) occurred during pregnancy, and, when only women were considered, pregnancy was associated with 18.08% of cases. There was a statistically significant difference between the LCH and NLCH groups in relation to average age. NLCH was more prevalent among older patients than LCH (41.27 ± 16.96 vs 33.55 ± 17.02 years; P = 0.006 [Student’s t test]) (Table 2), and its most prevalent site was the gingiva, while LCH was more common in the lips (P = 0.0001 [chi-squared test]). There was no statistically significant difference between the groups in terms of sex, CDH, nodule characteristics (sessile or pedunculated), and pregnancy (Table 2). Histopathological features In most lesions, an acute inflammatory infiltrate of varying intensity was observed near ulcers, with chronic or mixed inflammation in the depth of the lesion (Table 3). Foci of eosinophilic inflammation were found in both LCH and NLCH (17.16%) and, in 50.89% of lesions, there was a predominance of congested blood vessels. In a few cases (4.14%), foci of endothelial papillary proliferation were found, as well as plasma pooling (12.43%). Ulceration was present in 85.8% of cases, with no statistically significant difference between the groups. However, epithelial atrophy occurred more frequently in LCH (P = 0.0108 [chi-squared test]). Immunohistochemical features 7 LCH exhibited a significantly greater number of CD34-positive vessels and more positive SMA stained areas than NLCH; however, there was no statistical difference in the number of positively stained mast cells (Table 4, Figure 1). Although SMA stained mainly pericytes and smooth muscle around vessels, a population of myofibroblasts was also evidenced in the lesions. In addition, there was no statistically significant correlation between the number of mast cells and the number of CD34-positive vessels when considering all cases together (r = 0.1606; P = 0.36425), or the PG subtypes separately (NLCH: r = 0.252; P = 0.327; LCH: r = 0.16; P = 0.5396). A significantly greater number of nuclei positively stained for Ki-67 was found in LCH compared with NLCH (Table 4). Lymphatic vessels were positive for D2-40. Lesions exhibited more D2-40-positive lymphatic vessels near the ulcer and the epithelial lining than at the depth. Discussion PG is a common oral lesion that exhibits two distinct histopathological subtypes: LCH and NLCH. Based on reports in the literature, we expected some clinical, histopathological, and immunohistochemical differences between LCH and NLCH,11-13 which were confirmed in the present study. PG was more prevalent in females and in individuals with an average age in the fourth decade of life, and appeared clinically as a nodule in the gingiva, similar to previous studies.1,2,4,7,12 The preference for females is believed to be due to the vascular effects of female hormones.6 In vitro studies have demonstrated that female sex hormones are potential regulators of the production of several growth factors, such as vascular endothelial growth factor (VEGF), in various cell types.17 In our survey, the most cited CDH was PG, followed by inflammatory fibrous hyperplasia/irritation fibroma and peripheral giant cell granuloma. Irritation fibroma, PG, and peripheral giant cell granuloma were also among the most common CDH in previous studies.1,4,16 Many factors can influence CDH, including experience of the clinician and clinical aspect(s) of the lesion. In most cases, only benign lesions were included among the differential diagnosis. When comparing the clinical features of LCH and NLCH, there was no statistically significant difference in relation to the type of nodule after evaluation of data from 100 PG cases from 8 which this information was available (59.52% of our cases). However, as previous reported,12 LCH affected younger patients and was more common in the lip, while NLCH was more common in the gingiva. Since it was previously suggested that PG subtypes may differ in etiology11 with NLCH being correlated with traumatic etiological factors, such as overhanging edges of restorations or biting, while LCH does not present any obvious correlation with traumatic events, we investigated these parameters in our cases. Differently from what it was previously found,11 we could not detect any associations between trauma and PG subtypes. This might have been because this information was available in only 13.02% of the cases (12 LCH and 10 NLCH). Similarly, in the study by Koo et al.,3 only 14% of cases reported a history of trauma. Regarding microscopic aspects, we found that surface atrophic epithelium and a higher number of CD34 and SMA (µm2) stained vessels, with clearly higher vascularization, were observed in LCH compared with NLCH, as previously reported.11 Although SMA staining the area was mostly occupied by pericytes and smooth muscle cells around the vessels, a population of myofibroblasts was also included in the region selected with the aid of an imaging processing software. Inflammatory cells as mast cells and eosinophils are involved in angiogenesis and tissue repair processes. Mast cells release mediators that increase vascular permeability and vasodilation, facilitating the migration of inflammatory cells.18 The outcome of mutual mast cell-fibroblast interactions promotes granulation tissue formation, a hallmark of PG,18 which encouraged us to evaluate the mast cells numbers in both histopathological subtypes. Several authors19 have found a positive correlation between the number of micro vessels and mast cells in a group of so-called reactive lesions, including LCH, fibrous hyperplasia, inflammatory fibrous hyperplasia, and peripheral giant cell lesion, while others found a negative correlation.20 In our study, we found a positive, yet not significant correlation between microvessels and mast cells, similarly to previous results,18 and no difference in the number of mast cells between groups. Eosinophil major basic protein has pro-angiogenic effects and can participate in tissue inflammation and remodeling.21 Moreover, besides VEGF, eosinophils were shown to produce other proangiogenic factors such as IL-8 and nerve growth factor.22 The contribution of 9 eosinophils to the pathogenesis of oral reactive lesions was previously studied, but eosinophils were not found to be involved in the fibrotic process and the variation of microscopic features of those lesions.23 In our study, foci of eosinophils were found in 17.16% of the cases, but they do not differ between the two subtypes of PG. Thus, our results indicate that both variants have similar inflammatory cells, and that angiogenesis might be linked to another factor. To investigate the proliferative rate of the lesions, we analyzed Ki-67 marker. It confirmed a higher proliferative activity of LCH compared with NLCH. The same difference was reported by Rezvani et al.,12 who suggested that LCH is more similar to a neoplasm than to a reactive lesion. Although Ki-67 is a marker of cell proliferation, we believe that the presence of positive cells in these lesions is insufficient to confirm a neoplastic nature, only reflecting a higher proliferation. Several studies24-26 have described cytogenetic alterations that may indicate the neoplastic nature of PG; however, they are not totally conclusive regarding genetic mutations.27 In cutaneous lesions, BRAF mutation was identified in endothelial cells as a major driver mutation in the pathogenesis of PG,25 explaining the occurrence of multiple PGs in patients treated with BRAF inhibitors. However, recently, it was reported the activation of the MAPK/ERK pathway in oral LCH by demonstrating pERK1/2 immunopositivity in the endothelial cells, independently of BRAF mutations.28 This pathway is activated by the oncogene RAS, which promotes an angiogenic phenotype in endothelial cells and has been shown to play a key role in tumor microenvironment and angiogenesis.29,30 Other studies investigating alterations of MAPK/ERK pathway in NLCH variants could help defining the possible different origins between the two PG subtypes. Another feature we evaluated was the presence of foci of intravascular papillary endothelial hyperplasia, which is an uncommon finding that can be associated with nonneoplastic vascular lesions.31 In seven cases, we found this association, and one was recently published32; however, there was no statistical difference between LCH and NLCH. Similarly, no difference was found between the two PG subtypes regarding the presence of plasma pooling in the epithelium, which is generally associated with traumatized lesions. In this retrospective study, we demonstrated that PG variants clinically differ with LCH occurring mostly in younger patients with a preference for the lips, while the NLCH is more common on the gingiva. Besides, on histopathology, LCH showed atrophic epithelium, higher 10 vascularization and proliferative activity when compared with NLCH. Even though this was the largest study in the literature, the differences found by us may be insufficient to conclude whether they are distinct entities. Nevertheless, new studies evaluating the nature of these lesions, specially focusing on molecular pathways, are important for academic purposes. 11 References 1. Krishnapillai R, Punnoose K, Angadi P, Koneru A. Oral pyogenic granuloma — a review of 215 cases in a South Indian Teaching Hospital, Karnataka, over a period of 20 years. Oral Maxillofac Surg. Sep 2012;16(3):305–309. http://dx.doi.org/10.1007/s10006-012-0315-z. 2. Sharma S, Chandra S, Gupta S, Srivasta S. Heterogeneous conceptualization of etiopathogenesis: Oral pyogenic granuloma. Natl J Maxillofac Surg. Jan-Jun 2019;10(1):3–7. http://dx.doi.org/10.4103/njms.NJMS_55_15. 3. Koo MG, Lee SH, Han S. Pyogenic Granuloma: a retrospective analysis of cases treated over a 10-year. Arch Craniofac Surg. Mar 2017;18(1):16–20. http://dx.doi.org/10.7181/acfs.2017.18.1.16. 4. Gordón-Núñez MA, Carvalho MV, Benevenuto TG, Lopes MFF, Silva LMM, Galvão HC. Oral pyogenic granuloma: A retrospective analysis of 293 cases in a Brazilian population. J Oral Maxillofac Surg. Sep 2010;68(9):2185–2188. http://dx.doi.org/10.1016/j.joms.2009.07.070. 5. Yao T, Nagai E, Utsunomiya T, Tsuneyoshi M. An intestinal counterpart of pyogenic granuloma of the skin. A newly proposed entity. Am J Surg Pathol. Sep 1995;19(9):1054– 1060. http://dx.doi.org/10.1097/00000478-199509000-00009. 6. Jafarzadeh H, Sanatkhani M, Mohtasham N. Oral pyogenic granuloma: a review. J Oral Sci. Dec 2006;48(4):167–175. http://dx.doi.org/10.2334/josnusd.48.167. 7. Kamal R, Dahiya P, Puri A. Oral pyogenic granuloma: various concepts of etiopathogenesis. J Oral Maxillofac Pathol. Jan 2012;16(1):79–82. http://dx.doi.org/10.4103/0973-029X.92978. 8. Fechner R, Cooper P, Mills S. Pyogenic granuloma of the larynx and trachea. An etiological and pathologic misnomer for granulation tissue. Arch Otolaryngol Head Neck Surg. Jan 1981;107(1):30–32. http://dx.doi.org/10.1001/archotol.1981.00790370032006. 9. Bett J, Batistella E, Melo G, et al. Prevalence of oral mucosal disorders during pregnancy: A systematic review and meta-analysis. J Oral Pathol Med. Apr 2019;48(4):270–277. http://dx.doi.org/10.1111/jop.12831. 10. International Society for the Study of Vascular Anomalies. Classification of vascular anomalies 2018. Acessed July 9, 2020. http://www.issva.org/UserFiles/file/ISSVA-Classification- 2018.pdf. 11. Epivatianos A, Antoniades D, Zaraboukas T, et al. Pyogenic granuloma of the oral cavity: comparative study of its clinicopathological and immunohistochemical features. Pathol Int. Jul 2005;55(7):391–397. http://dx.doi.org/10.1111/j.1440-1827.2005.01843.x. 12. Rezvani G, Azarpira N, Bita G, Zeynab R. Proliferative activity in oral pyogenic granuloma: a comparative immunohistochemical study. Indian J Pathol Microbiol. Jul-Sep 2010;53(3):403– 407. http://dx.doi.org/10.4103/0377-4929.68242. http://dx.doi.org/10.1007/s10006-012-0315-z http://dx.doi.org/10.4103/njms.NJMS_55_15 http://dx.doi.org/10.7181/acfs.2017.18.1.16 http://dx.doi.org/10.1016/j.joms.2009.07.070 http://dx.doi.org/10.1097/00000478-199509000-00009 http://dx.doi.org/10.2334/josnusd.48.167 http://dx.doi.org/10.4103/0973-029X.92978 http://dx.doi.org/10.1001/archotol.1981.00790370032006 http://dx.doi.org/10.1111/jop.12831 http://dx.doi.org/10.1111/j.1440-1827.2005.01843.x http://dx.doi.org/10.4103/0377-4929.68242 12 13. Isaza-Guzmán D, Teller-Carrero C, Laberry-Bermúdez M, Gonzáles-Pérez L, Tobón-Arroyave S. Assessment of clinicopathological characteristics and immunoexpression of COX-2 and IL-10 in oral pyogenic granuloma. Arch Oral Biol. May 2012;57(5):503–512. http://dx.doi.org/10.1016/j.archoralbio.2011.11.004. 14. Vassilopoulos SI, Tosios KI, Panis VG, Vrotsos JA. Endothelial cells of oral pyogenic granulomas express eNOS and CD105 ⁄ endoglin: an immunohistochemical study. J Oral Pathol Med. Apr 2011;40(4):345–351. http://dx.doi.org/10.1111/j.1600-0714.2010.00969.x. 15. Park S, Lee J, Tak M, Lee H, Choi H. A research of pyogenic granuloma genesis factor with immunohistochemical analysis. J Craniofac Surg. Nov 2017;28(8):2068–2072. http://dx.doi.org/10.1097/SCS.0000000000004148. 16. Toida M, Hasegawa T, Watanabe F, et al. Lobular capillary hemangioma of the oral mucosa: clinicopathological study of 43 cases with a special reference to immunohistochemical characterization of the vascular elements. Pathol Int. Jan 2003;53(1):1–7. http://dx.doi.org/10.1046/j.1440-1827.2003.01434.x. 17. Kanda N, Watanabe S. Regulatory roles of sex hormones in cutaneous biology and immunology. J Dermatol Sci. Apr 2005;38(1):1–7. http://dx.doi.org/10.1016/j.jdermsci.2004.10.011. 18. Komi D, Khomtchouk K, Maria P. A review of the contribution of mast cells in wound healing: involved molecular and cellular mechanisms. Clin Rev Allergy Immunol. Jun 2020;58(3):298- 312. http://dx.doi.org/10.1007/s12016-019-08729-w. 19. Ferreira S, Xavier F, Freitas M, Nunes F, Gurgel C, Cangussu M. Involvement of mast cells and microvessels density in reactive lesions of oral cavity: A comparative immunohistochemical study. Pathol Res Pr. Sep 2016;212(9):761–766. http://dx.doi.org/10.1016/j.prp.2016.05.007. 20. Kouhsoltani M, Khiavi MM, Tahamtan S. Evaluation of mast cell counts and microvessel density in reactive lesions of the oral cavity. J Dent Res Dent Clin Dent Prospects. 2016;10(4):241–246. http://dx.doi.org/10.15171/joddd.2016.038. 21. Puxeddu I, Berkman N, Nissim Ben Efraim AH, et al. The role of eosinophil major basic protein in angiogenesis. Allergy Eur J Allergy Clin Immunol. Mar 2009;64(3):368–374. http://dx.doi.org/10.1111/j.1398-9995.2008.01822.x. 22. Nissim Ben Efraim AH, Levi-Schaffer F. Roles of eosinophils in the modulation of angiogenesis. Chem Immunol Allergy. 2014;99:138–154. http://dx.doi.org/10.1159/000353251. 23. Aghbali A, Akbarzadeh A, Kouhsoltani M. The role of macrophages and eosinophils in reactive lesions of the oral cavity. J Oral Maxillofac Pathol. Jan-Apr 2018;22(1):147. http://dx.doi.org/10.4103/jomfp.JOMFP_15_16. 24. Lim Y, Douglas S, Ko C, Antaya R, McNiff J, Zhou J. Somatic activating RAS mutations cause vascular tumors including pyogenic granuloma. J Invest Dermatol. Jun 2015;135(6):1698– 1700. http://dx.doi.org/10.1038/jid.2015.55. http://dx.doi.org/10.1016/j.archoralbio.2011.11.004 http://dx.doi.org/10.1111/j.1600-0714.2010.00969.x http://dx.doi.org/10.1097/SCS.0000000000004148 http://dx.doi.org/10.1046/j.1440-1827.2003.01434.x http://dx.doi.org/10.1016/j.jdermsci.2004.10.011 http://dx.doi.org/10.1007/s12016-019-08729-w http://dx.doi.org/10.1016/j.prp.2016.05.007 http://dx.doi.org/10.15171/joddd.2016.038 http://dx.doi.org/10.1111/j.1398-9995.2008.01822.x http://dx.doi.org/10.1159/000353251 http://dx.doi.org/10.4103/jomfp.JOMFP_15_16 http://dx.doi.org/10.1038/jid.2015.55 13 25. Groesser L, Peterhof E, Evert M, Landthaler M, Berneburg M, Hafner C. BRAF and RAS Mutations in Sporadic and Secondary Pyogenic Granuloma. J Invest Dermatol. Feb 2016;136(2):481–486. http://dx.doi.org/10.1038/JID.2015.376. 26. Truss L, Dobin SM, Donner LR. Deletion (21)(q21.2q22.12) as a sole clonal cytogenetic abnormality in a lobular capillary hemangioma of the nasal cavity. Cancer Genet Cytogenet. Oct 2006;170(1):69–70. http://dx.doi.org/10.1016/j.cancergencyto.2006.04.016. 27. Li G, Adams E, Eshleman J, Eberhart C. No BRAF V600E mutation identified in 28 periocular pyogenic granuloma. Ophthalmic Plast Reconstr Surg. Nov-Dec 2018;34(6):525–527. http://dx.doi.org/10.1097/IOP.0000000000001075. 28. Pereira TSF, de Amorim LSD, Pereira NB, et al. Oral pyogenic granulomas show MAPK/ERK signaling pathway activation, which occurs independently of BRAF, KRAS, HRAS, NRAS, GNA11, and GNA14 mutations. J Oral Pathol Med. Nov 2019;48(10):906–910. http://dx.doi.org/10.1111/jop.12922. 29. Meadows KN, Bryant P, Vincent PA, Pumiglia KM. Activated Ras induces a proangiogenic phenotype in primary endothelial cells. Oncogene. Jan 2004;23(1):193-200. http://dx.doi.org/10.1038/sj.onc.1206921. 30. Sharma A, Trivedi NR, Zimmerman MA, Tuveson DA, Smith CD, Robertson GP. Mutant V599EB-Raf regulates growth and vascular development of malignant melanoma tumors. Cancer Res. Mar 2005;65(6):2412-2421. http://dx.doi.org/10.1158/0008-5472.CAN-04-2423. 31. Inoue H, Miyazaki Y, Kikuchi K, et al. Intravascular papillary endothelial hyperplasia of the oral cavity. J Oral Sci. Dec 2011;53(4):475–480. http://dx.doi.org/10.2334/josnusd.53.475. 32. Milhan N, Torquato L, Costa V, de Marco A, Carvalho Y, Anbinder A. A mixed form of intravascular papillary endothelial hyperplasia in an uncommon location: case and literature review. Dermatol Online J. Feb 2018;24(2):13030. Accessed December 20, 2020. https://escholarship.org/uc/item/2dk039r1. http://dx.doi.org/10.1038/JID.2015.376 http://dx.doi.org/10.1016/j.cancergencyto.2006.04.016 http://dx.doi.org/10.1097/IOP.0000000000001075 http://dx.doi.org/10.1111/jop.12922 http://dx.doi.org/10.1038/sj.onc.1206921 http://dx.doi.org/10.1158/0008-5472.CAN-04-2423 http://dx.doi.org/10.2334/josnusd.53.475 https://escholarship.org/uc/item/2dk039r1 14 Table 1 – Antibodies and immunostaining conditions Antibody Manufacturer Concentration Clone Incubation (hours) Antigen retrieval SMA Dako 1:200 1A4 1 Citrate pH 6.0 D2-40 Dako 1:50 D2-40 1 Citrate pH 6.0 CD34 Dako 1:50 QBend 10 1 Citrate pH 6.0 Mast cell Imgenex 1:700 AA1 1 Citrate pH 6.0 Ki-67 SpringBioscience 1:50 SP6 1 EDTA pH 9.0 Table 2 – Clinical data and comparison between LCH and NLCH Data All cases (n=169) LCH (n=62) NLCH (n=107) P-value (LCH X NLCH) Age range (mean ± SD) 4 – 80 years (38.59 ± 16.96) 4-70 (33.55±17.02) 5 – 80 (41.27±16.96) 0.0056* (Student t test) Sex, n (%) Female Male Not specified 94 (55.62%) 74 (43.79%) 1 (0.59%) 30 (48.39%) 31 (50%) 1 (1.61%) 64 (59.81%) 43 (40.19%) 0.1819 (chi-squared test) Site, n (%) Gingiva Lip Tongue Buccal mucosa Alveolar ridge Palate 67 (39.64%) 38 (22.49%) 25 (14.79%) 16 (9.47%) 12 (7.1%) 11 (6.51%) 7 (11.29%) 29 (46.77%) 13 (20.97%) 11 (17.74%) 2 (3.23%) 0 60 (56.07%) 9 (8.41%) 12 (11.21%) 5 (4.67%) 10 (9.35%) 11 (10.28%) 0.0001* (chi-squared test) Clinical diagnostic hypothesis, n (% ) PG mentioned PG not mentioned 107 (63.31%) 62 (36.69%) 35 (56.45%) 27 (43.55%) 72 (67.29%) 35 (32.71%) 0.3615 (chi-squared test) Nodule, n (%) Sessile Pedunculated Not specified 24 (14.2%) 76 (44.97%) 69 (40.83%) 7 (11.29%) 26 (41.94%) 29 (46.77%) 17 (15.89%) 50 (46.73%) 40 (37.38%) 0.64 (chi-squared test) Pregnancy, n (% of all cases) Yes Male/Not specified 17 (10.06%) 152 (89.94%) 7 (11.29%) 55 (88.71%) 10 (9.35%) 97 (90.65%) 0.6854 (Chi-squared test) Abbreviation: PG, pyogenic granuloma. *P < 0.05 15 Table 3 – Histopathological features – comparison between LCH and NLCH Microscopic features All cases (%) LCH (%) NLCH (%) P-value Ulceration 145 (85.8%) 51 (82.26%) 94 (87.85%) 0.3155 Congested vessels 86 (50.89%) 33 (53.23%) 53 (49.53%) 0.6435 Atrophic epithelium 66 (39.05%) 32 (51.61%) 34 (31.78%) 0.0108* Eosinophil foci 29 (17.16%) 12 (19.35%) 17 (15.89%) 0.5645 Foci of intravascular papillary endothelial hyperplasia 7 (4.14%) 2 (3.23%) 5 (4.67%) 0.6491 Superficial bacterial colonies 85 (50.30%) 35 (56.45%) 50 (46.73%) 0.2707 Plasma pooling 21 (12.43%) 6 (9.68%) 15 (14.02%) 0.4096 *P < 0.05 (chi-squared test). 16 Table 4 – Number of blood vessels (CD34), smooth muscle actin stained area (SMA), mast cell, and Ki-67 positive cells – comparison between LCH and NLCH All cases LCH NLCH P-value CD34 positive blood vessels (mean ± SD) 83.28 ± 45.04 113.99 ± 40 50.59 ± 21.69 < 0.0001* SMA stained area (mean ± SD) 4741.08 ± 3420.79 µm2 6547.05 ± 3656.08 µm2 2822.24 ± 1755.54 µm2 0.0009* Mast cells (mean ± SD) 27.41 ± 17.68 28.02 ± 17.21 26.79 ± 18.65 0.843 Ki-67 positive nuclei (mean ± SD) 56.97 ±39.9 74.46 ± 43.31 38.31 ± 26.06 0.0092* *P < 0.05 (Student’s t test). 17 Figure 1. Histopathological and immunohistochemical characteristics of pyogenic granuloma. (A) NLCH — vascular proliferation similar to granulation tissue, associated with chronic inflammatory infiltrate, with fibroblasts and newly formed vessels of varying sizes (HE staining). (B) LCH — vascular proliferation in lobular aggregates, separated by bundles of fibrous connective tissue (HE staining). (C) NLCH — positive staining of vascular endothelium — immunohistochemical reaction (CD34). (D) LCH — positive staining of vascular endothelium — immunohistochemical reaction (CD34). A greater number of vessels is observed in comparison to C. (E) NLCH — positive staining for SMA immunohistochemical reaction. (F) LCH — positive staining for SMA immunohistochemical reaction. A larger positive marked area is observed in comparison to E. The slides were digitized using a whole slide scanner (Pannoramic Desk, 3DHistech), with x20 objective. Images A and B were captured with x1.4 zoom (scale bar = 1000 μm) and images C to F were captured with x20 zoom (scale bar = 50 μm).