• View in gallery
    Figure 1—

    Photomicrographs of collecting duct epithelial cells in the renal medulla of nondiseased canine kidney that has (A) or has not (B) undergone bleaching with hydrogen peroxide prior to labeling for COX-2. Immunohistochemical stain.

  • View in gallery
    Figure 2—

    Photomicrograph of the iridocorneal angle with adjacent ciliary processes in a section of a nondiseased canine globe. Notice finely granular intracytoplasmic COX-2 immunostaining in the trabecular meshwork of the aqueous outflow tract (inset). Immunohistochemical stain.

  • View in gallery
    Figure 3—

    Photomicrograph of a section of a canine globe with anterior uveal malignant melanoma. Notice the granular intracytoplasmic COX-2 labeling in the neoplastic tissue (inset). Immunohistochemical stain.

  • View in gallery
    Figure 4—

    Photomicrograph of a section of a canine globe with anterior uveal melanocytoma. Notice the neoplasm invading the iridocorneal angle and ciliary body and bordering the sclera, and the granular intracytoplasmic COX-2 labeling in the iridocorneal angle that is infiltrated with neoplastic tissue (inset). Immunohistochemical stain.

  • View in gallery
    Figure 5—

    Photomicrograph of a section of a canine globe with anterior uveal melanocytoma. Notice the ciliary body, ciliary processes, and adjacent neoplasm. There is finely granular intracytoplasmic COX-2 immunoreactivity in the nonpigmented ciliary body epithelium (inset). Immunohistochemical stain.

  • View in gallery
    Figure 6—

    Photomicrograph of sections of 2 diseased globes without neoplasia from dogs with chronic keratitis. Notice the granular intracytoplasmic COX-2 immunoreactivity predominantly in the corneal epithelium and occasionally in stromal keratocytes of 1 globe (A) and in the nonpigmented ciliary body epithelium of the other globe (B). Immunohistochemical stain.

  • 1.

    Radi ZA, Render JA. The pathophysiologic role of cyclo-oxygenases in the eye. J Ocul Pharmacol Ther 2008;24:141151.

  • 2.

    Sano H, Kawahito Y, Wilder RL, et al. Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 1995;55:37853789.

  • 3.

    Tsujii M, Kawano S, Tsuji S, et al. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705716.

  • 4.

    Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci U S A 1997;94:33363340.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Berg J, Christoph T, Bodenteich A, et al. Heterogeneous distribution of COX-2 over-expression in human colon carcinoma cells. Adv Exp Med Biol 1997;433:327330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Eberhart CE, Coffey RJ, Radhika A, et al. Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:11831188.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7.

    Jaeckel EC, Raja S, Tan J, et al. Correlation of expression of cyclooxygenase-2, vascular endothelial growth factor, and peroxisome proliferator-activated receptor delta with head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2001;127:12531259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Marrogi AJ, Travis WD, Welsh JA, et al. Nitric oxide synthase, cyclooxygenase 2, and vascular endothelial growth factor in the angiogenesis of non-small cell lung carcinoma. Clin Cancer Res 2000;6:47394744.

    • Search Google Scholar
    • Export Citation
  • 9.

    Han SL, Tang HJ, Hua YW, et al. Expression of COX-2 in stomach cancers and its relation to their biological features. Dig Surg 2003;20:107114.

  • 10.

    Uefuji K, Ichikura T, Mochizuki H, et al. Expression of cyclooxygenase-2 protein in gastric adenocarcinoma. J Surg Oncol 1998;69:168172.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Shirahama T. Cyclooxygenase-2 expression is up-regulated in transitional cell carcinoma and its preneoplastic lesions in the human urinary bladder. Clin Cancer Res 2000;6:24242430.

    • Search Google Scholar
    • Export Citation
  • 12.

    Marrogi A, Pass HI, Khan M, et al. Human mesothelioma samples overexpress both cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase (NOS2): in vitro antiproliferative effects of a COX-2 inhibitor. Cancer Res 2000;60:36963700.

    • Search Google Scholar
    • Export Citation
  • 13.

    Shiota G, Okubo M, Noumi T, et al. Cyclooxygenase-2 expression in hepatocellular carcinoma. Hepatogastroenterology 1999;46:407412.

  • 14.

    Goulet AC, Einsphar JG, Alberts DS, et al. Analysis of cyclooxygenase 2 (COX-2) expression during malignant melanoma progression. Cancer Biol Ther 2003;2:713718.

    • Search Google Scholar
    • Export Citation
  • 15.

    Dore M, Lanthier I, Sirois J. Cyclooxygenase-2 expression in canine mammary tumors. Vet Pathol 2003;40:207212.

  • 16.

    Brunelle M, Sartin EA, Wolfe LG, et al. Cyclooxygenase-2 expression in normal and neoplastic canine mammary cell lines. Vet Pathol 2006;43:656666.

  • 17.

    Heller DA, Clifford CA, Goldschmidt MH, et al. Cyclooxygenase-2 expression is associated with histologic tumor type in canine mammary carcinoma. Vet Pathol 2005;42:776780.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    McEntee MF, Cates JM, Neilsen N. Cyclooxygenase-2 expression in spontaneous intestinal neoplasia of domestic dogs. Vet Pathol 2002;39:428436.

  • 19.

    Khan KN, Stanfield KM, Trajkovic D, et al. Expression of cyclooxygenase-2 in canine renal cell carcinoma. Vet Pathol 2001;38:116119.

  • 20.

    Khan KN, Knapp DW, Denicola DB, et al. Expression of cyclooxygenase-2 in transitional cell carcinoma of the urinary bladder in dogs. Am J Vet Res 2000;61:478481.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Kleiter M, Malarkey DE, Ruslander DE, et al. Expression of cyclooxygenase-2 in canine epithelial nasal tumors. Vet Radiol Ultrasound 2004;45:255260.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    L'Eplattenier HF, Lai CL, van den HR, et al. Regulation of COX-2 expression in canine prostate carcinoma: increased COX-2 expression is not related to inflammation. J Vet Intern Med 2007;21:776782.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23.

    Tremblay C, Dore M, Bochsler PN, et al. Induction of prostaglandin G/H synthase-2 in a canine model of spontaneous prostatic adenocarcinoma. J Natl Cancer Inst 1999;91:13981403.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Pestili de Almeida EM, Piche C, Sirois J, et al. Expression of cyclo-oxygenase-2 in naturally occurring squamous cell carcinomas in dogs. J Histochem Cytochem 2001;49:867875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25.

    Mullins MN, Lana SE, Dernell WS, et al. Cyclooxygenase-2 expression in canine appendicular osteosarcomas. J Vet Intern Med 2004;18:859865.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    Yu HG, Huang JA, Yang YN, et al. The effects of acetylsalicylic acid on proliferation, apoptosis, and invasion of cyclooxygenase-2 negative colon cancer cells. Eur J Clin Invest 2002;32:838846.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Song X, Lin HP, Johnson AJ, et al. Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.

    Masferrer JL, Leahy KM, Koki AT, et al. Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:13061311.

    • Search Google Scholar
    • Export Citation
  • 29.

    Khan KN, Venturini CM, Bunch RT, et al. Interspecies differences in renal localization of cyclooxygenase isoforms: implications in nonsteroidal antiinflammatory drug-related nephrotoxicity. Toxicol Pathol 1998;26:612620.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Ozdal PC, Callejo S, Caissie AL, et al. Cyclooxygenase-2 expression in human irradiated uveal melanomas. Int Ophthalmol 2008;28:16.

  • 31.

    Figueiredo A, Caissie AL, Callejo SA, et al. Cyclooxygenase-2 expression in uveal melanoma: novel classification of mixedcell-type tumours. Can J Ophthalmol 2003;38:352356.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    Cryan LM, Paraoan L, Hiscott P, et al. Expression of COX-2 and prognostic outcome in uveal melanoma. Curr Eye Res 2008;33:177184.

  • 33.

    Orchard GE. Heavily pigmented melanocytic neoplasms: comparison of two melanin-bleaching techniques and subsequent immunohistochemical staining. Br J Biomed Sci 1999;56:188193.

    • Search Google Scholar
    • Export Citation
  • 34.

    Li LX, Crotty KA, Kril JJ, et al. Method of melanin bleaching in MIB1-Ki67 immunostaining of pigmented lesions: a quantitative evaluation in malignant melanomas. Histochem J 1999;31:237240.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35.

    Mohammed SI, Khan KN, Sellers RS, et al. Expression of cyclooxygenase-1 and 2 in naturally-occurring canine cancer. Prostaglandins Leukot Essent Fatty Acids 2004;70:479483.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36.

    Sellers RS, Silverman L, Khan KN. Cyclooxygenase-2 expression in the cornea of dogs with keratitis. Vet Pathol 2004;41:116121.

  • 37.

    Marshall JL, Stanfield KM, Silverman L, et al. Enhanced expression of cyclooxygenase-2 in glaucomatous dog eyes. Vet Ophthalmol 2004;7:5962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38.

    Amico C, Yakimov M, Catania MV, et al. Differential expression of cyclooxygenase-1 and cyclooxygenase-2 in the cornea during wound healing. Tissue Cell 2004;36:112.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 39.

    Bonazzi A, Mastyugin V, Mieyal PA, et al. Regulation of cyclooxygenase-2 by hypoxia and peroxisome proliferators in the corneal epithelium. J Biol Chem 2000;275:28372844.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40.

    Maihofner C, Schlotzer-Schrehardt U, Guhring H, et al. Expression of cyclooxygenase-1 and -2 in normal and glaucomatous human eyes. Invest Ophthalmol Vis Sci 2001;42:26162624.

    • Search Google Scholar
    • Export Citation
  • 41.

    Giuliano EA, Chappell R, Fischer B, et al. A matched observational study of canine survival with primary intraocular melanocytic neoplasia. Vet Ophthalmol 1999;2:185190.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement

Expression of cyclooxygenase-2 in canine uveal melanocytic neoplasms

Danielle PagliaVeterinary Medical Teaching Hospital, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by Danielle Paglia in
Current site
Google Scholar
PubMed
Close
 DVM
,
Richard R. DubielzigDepartment of Pathobiological Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, WI 53706.

Search for other papers by Richard R. Dubielzig in
Current site
Google Scholar
PubMed
Close
 DVM
,
Helen K. Kado-FongDepartment of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by Helen K. Kado-Fong in
Current site
Google Scholar
PubMed
Close
 MS
, and
David J. MaggsDepartment of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by David J. Maggs in
Current site
Google Scholar
PubMed
Close
 BVSc

Abstract

Objective—To determine whether cyclooxygenase-2 (COX-2) is expressed in benign or malignant canine uveal melanocytic neoplasms and whether expression correlates with malignancy.

Sample Population—Tissue sections from 71 globes; 57 with benign (n = 15), malignant (34), or mixed (8) uveal melanocytic neoplasms; 10 with nonneoplastic disease; and 4 with no abnormalities.

Procedures—Bleached sections from all globes and canine kidney were incubated with mouse monoclonal antibody directed against rat COX-2 protein or mouse antibody isotype control. Location, intensity, and percentage of immunolabeled cells were scored.

Results—Expression of COX-2 was detected in all but 5 globes, all of which contained neoplasms. Expression of COX-2 was detected in regions infiltrated by neoplasia in 21 globes; however, definitive labeling of tumor cells was detected in only 2 of those. In the remaining 19 globes, COX-2 expression was detected in areas also labeled in globes without disease and globes with nonneoplastic disease, especially the aqueous outflow tract and ciliary body. However, only globes with uveal malignant melanomas had detectable COX-2 expression in the iris. Expression of COX-2 was detected in the ciliary body of more globes with uveal malignant melanoma (20/34) than in those without disease (1/4), with nonneoplastic disease (4/10), or with melanocytoma (3/15) or mixed neoplasms (3/8).

Conclusions and Clinical Relevance—Canine globes with uveal melanocytic neoplasia appeared to express COX-2 in similar sites and with similar intensity as globes without neoplasia. Differentiation of benign from malignant canine uveal melanocytic neoplasms was not possible.

Abstract

Objective—To determine whether cyclooxygenase-2 (COX-2) is expressed in benign or malignant canine uveal melanocytic neoplasms and whether expression correlates with malignancy.

Sample Population—Tissue sections from 71 globes; 57 with benign (n = 15), malignant (34), or mixed (8) uveal melanocytic neoplasms; 10 with nonneoplastic disease; and 4 with no abnormalities.

Procedures—Bleached sections from all globes and canine kidney were incubated with mouse monoclonal antibody directed against rat COX-2 protein or mouse antibody isotype control. Location, intensity, and percentage of immunolabeled cells were scored.

Results—Expression of COX-2 was detected in all but 5 globes, all of which contained neoplasms. Expression of COX-2 was detected in regions infiltrated by neoplasia in 21 globes; however, definitive labeling of tumor cells was detected in only 2 of those. In the remaining 19 globes, COX-2 expression was detected in areas also labeled in globes without disease and globes with nonneoplastic disease, especially the aqueous outflow tract and ciliary body. However, only globes with uveal malignant melanomas had detectable COX-2 expression in the iris. Expression of COX-2 was detected in the ciliary body of more globes with uveal malignant melanoma (20/34) than in those without disease (1/4), with nonneoplastic disease (4/10), or with melanocytoma (3/15) or mixed neoplasms (3/8).

Conclusions and Clinical Relevance—Canine globes with uveal melanocytic neoplasia appeared to express COX-2 in similar sites and with similar intensity as globes without neoplasia. Differentiation of benign from malignant canine uveal melanocytic neoplasms was not possible.

Cyclooxygenase exists in at least 2 isoforms in many tissues, including the eye.1 The isoforms COX-1 and COX-2 both catalyze the conversion of arachidonic acid to prostaglandins; however, these 2 enzymes have different roles in the body. Cyclooxygenase-1 is widely expressed and serves as a housekeeping enzyme. In particular, COX-1 regulates production of prostaglandins involved in gastroprotection and renal function. By contrast, COX-2 is induced by inflammatory mediators and is responsible for production of prostaglandins in the inflammatory cascade. Cyclooxygenase-2 expression can be detected in inflammatory cells, endothelial cells, fibroblasts, and neoplastic cells.2 Recently, COX-2 has also been found to play a critical role in carcinogenesis,2 tumor-induced angiogenesis,3 and tumor metastasis4 in human cancers. In fact, it is overexpressed in a diverse group of spontaneously occurring human tumors, including colorectal adenoma and adenocarcinoma,2,5,6 squamous cell carcinoma,7 non-small cell lung carcinoma,8 gastric carcinoma,9,10 transitional cell carcinoma,11 mesothelioma,12 hepatocellular carcinoma,13 and malignant melanoma.14 Cyclooxygenase-2 expression has also been investigated in canine neoplasms arising from mammary tissue,15–17 intestine,18 kidney,19 bladder,20 nose,21 prostate,22,23 skin,24 and bone,25 with results closely paralleling those in human neoplasia. These data have led researchers to investigate the use of NSAIDs for treatment of COX-expressing tumors.26–28 Although melanocytic neoplasms are the most common primary intraocular neoplasm of dogs, occasionally metastasize, and frequently require that the eye be enucleated, to the authors' knowledge, there are no published studies describing COX-2 expression in uveal melanocytic neoplasms in dogs. Therefore, the purposes of the study reported here were to determine whether there is COX-2 expression in canine uveal melanocytic neoplasms and whether the degree of COX-2 expression correlates with malignancy.

Materials and Methods

Sample population—Sixty-seven paraffin-embedded canine globes were selected from the Comparative Ocular Pathology Laboratory of Wisconsin collection. All globes underwent histologic review by a board-certified veterinary pathologist (RRD) and were categorized as diseased but without evidence of neoplasia or containing uveal melanocytic neoplasms (melanocytomas, malignant melanomas, or mixed neoplasms). Criteria used to diagnose each neoplasm type were based on morphology of neoplastic cells, presence of heavily pigmented cells (often characterized as benign), and presence or absence of mitotic figures. Histologic diagnoses were based on review of H&E-stained sections. Occasionally, sections bleached with potassium permanganate were also used. Four nondiseased globes obtained from the University of California Veterinary Pathology Service following postmortem removal from dogs euthanatized for nonocular disease served as additional controls in this study. These nondiseased globes lacked any histologic lesions. For all globes used in this study, data regarding clinical diagnoses were obtained from the submitting clinician and complete histologic diagnoses were obtained from the pathology report. In particular, clinical or histologic evidence of the presence or absence of glaucoma was noted.

Immunohistochemical evaluation—Tissue sections (5 μm) from all globes were mounted onto positively charged slidesa and underwent immunohistochemical detection of COX-2. Formalin-fixed, paraffin-embedded tissue sections of healthy canine (puppy) kidney served as a positive control for COX-2 expression29 and were processed in an identical manner to that used for globes. All tissue sections were deparaffinized in xylene and rehydrated in an alcohol series of decreasing concentration. Melanin was bleached, and endogenous peroxidases were depleted by incubation in a 10% solution of hydrogen peroxide in PBS solution at 22°C for approximately 3 hours. Antigen retrieval was performed in citrate bufferb heated in a microwave at 180 W for 10 minutes and allowed to cool on the benchtop for 20 minutes. Nonspecific binding was blocked by use of 10% heat-inactivated horse serum in PBS solution for 20 minutes at 22°C.

The primary antibody was a mouse monoclonal antibody directed against the rat COX-2 protein.c A purified mouse IgG1 isotyped control was used for negative controls. Both antibodies were used at a dilution of 1:30 in PBS solution. Sections were incubated with primary antibody for 90 minutes and incubated for 30 minutes with goat anti-mouse antibody linked to horseradish peroxidase.e All antibody incubations were performed in a humidified chamber at 22°C. Immunocomplexes were visualized by use of aminoethyl carbazole substratef and counterstained with Mayer hematoxylin. Slides were washed with PBS solution after bleaching, antigen retrieval, primary antibody incubation, secondary antibody incubation, aminoethyl carbazole, and counterstaining steps. Slides were preserved in mounting media and protected with a coverslip.

For each globe, an immunolabeled slide and a negative control were assessed for immunolabeling by 1 author (DP) by use of light microscopy. All regions of the globe were assessed, and any ocular tissue that expressed COX-2 was reported. Each globe was scored for positive immunolabeling of the sclera, each corneal layer, all sections of the uvea, aqueous outflow tract, all parts of the lens, and retina. For the purpose of COX-2 expression, the aqueous outflow tract was defined as the trabecular meshwork and ciliary cleft. Uveal areas containing melanocytic neoplastic tissue were closely assessed in an attempt to define whether neoplastic cells were specifically labeled. Scoring was based on labeling intensity15,19 and percentage of positively labeled neoplastic cells.16 Briefly, labeling intensity was graded semiquantitatively as negative (−), weak (+), moderate (++), or strong (+++) labeling. Percentage of positively labeled neoplastic cells was graded as 0 (negative), 1 (rare positive neoplastic cells), 2 (< 5% of cells), 3 (5% to 10% of cells), or 4 (> 10% of cells).

Results

Sample population—Histologic diagnoses for the 57 globes with uveal melanocytic neoplasia included anterior uveal melanocytoma (n = 15), anterior uveal malignant melanoma (34), and mixed melanocytic neoplasm (malignant melanoma arising within uveal melanocytoma; 8). Histologic diagnoses in diseased globes without neoplasia (n = 10) included chronic superficial keratitis (10), corneal endothelial attenuation (4), lens capsule rupture (2), preiridal fibrovascular membrane (2), peripheral anterior synechiae (2), phthisis bulbi (2), cataract (1), lens luxation (1), intraocular hemorrhage (1), asteroid hyalosis (1), and goniodysgenesis (1). Thirty-nine (68.4%) globes with neoplasia and 7 of 10 diseased globes without neoplasia had clinical or histologic evidence of glaucoma or both.

Immunohistochemical findings—Expression of COX-2 was detected in collecting duct epithelial cells in the renal medulla of all positive control canine kidney samples irrespective of whether they were bleached or not (Figure 1). Expression of COX-2 was also detected in all 4 nondiseased canine globes. Various sites in nondiseased globes were labeled, including the ciliary body musculature, the trabecular meshwork of the aqueous outflow tract (Figure 2), lens epithelium, choroidal stroma, and scleral lamina cribrosa (Table 1). The COX-2 staining was variably granular in the cytoplasm. No nondiseased globes had detectable COX-2 expression in the cornea, corneoscleral limbus, anterior sclera, iris, or retina.

Table 1—

Summary of COX-2 expression based on anatomic location in nondiseased canine globes, canine globes with melanocytic neoplasms, and canine globes with nonneoplastic disease. Values indicate number of globes.

LocationTissueNondiseased globes (n = 4)Uveal melanocytomas (n = 15)Uveal malignant melanomas (n = 34)Mixed melanocytic neoplasms (n = 8)Diseased globes without neoplasia (n = 10)
Cornea0111210
EpitheliumNA11028
StromaNA0213
Limbus02101
Sclera251853
Collagen02921
Blood vessels03410
Adjacent to neoplasmNA1410
Adjacent to AOT00211
Lamina cribrosa20001
Iris00200
Dilator muscleNANA1NANA
Anterior irisNANA1NANA
Ciliary body132034
Muscle111101
Stroma02821
NCBE01812
AOT1112028
Choroid47835
Stroma41100
Blood vessels07735
Retina01622
RPENA1522
NFLNA0100
Lens epithelium12500

AOT = Aqueous outflow tract. NA = Not applicable. NCBE = Nonpigmented ciliary body epithelium. NFL = Nerve fiber layer. RPE = Retinal pigmented epithelium.

Figure 1—
Figure 1—

Photomicrographs of collecting duct epithelial cells in the renal medulla of nondiseased canine kidney that has (A) or has not (B) undergone bleaching with hydrogen peroxide prior to labeling for COX-2. Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

Figure 2—
Figure 2—

Photomicrograph of the iridocorneal angle with adjacent ciliary processes in a section of a nondiseased canine globe. Notice finely granular intracytoplasmic COX-2 immunostaining in the trabecular meshwork of the aqueous outflow tract (inset). Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

Labeling for COX-2 was noted in 52 (91.2%) globes with uveal neoplasia. Globes in which COX-2 expression was not identified included 3 with malignant melanoma, 1 with melanocytoma, and 1 with a mixed neoplasm. A wide variety of cells and tissues in globes with neoplasia expressed COX-2 (Table 1). Of the 57 canine globes with neoplasia, 21 (36.8%) expressed COX-2 in regions infiltrated by the uveal melanocytic neoplasm. However, because these neoplasms invaded and were admixed with nondiseased tissue in many globes, it was possible to definitively identify labeling of neoplastic cells in only 2 of these 21 globes; 1 with a mixed neoplasm and 1 with malignant melanoma (Figure 3). In the remaining 19 globes, COX-2 expression was evident in those areas in which it was also seen in nondiseased globes and diseased globes without neoplasia; especially the trabecular meshwork of the aqueous outflow tract (Figure 4) and the nonpigmented epithelium of the ciliary body (Figure 5). All diseased globes without neoplasia expressed COX-2 in at least 1 ocular tissue (Figure 6). However, only globes with uveal malignant melanoma had detectable COX-2 expression in the iris. The percentage of globes in which COX-2 expression was detected in the ciliary body was notably greater in globes with uveal malignant melanoma (58.8%) than in nondiseased globes (1/4), diseased globes without neoplasia (4/10), or globes with melanocytoma (3/15) or mixed neoplasms (3/8). Among the globes with neoplasia, labeling score (percentage of cells labeled for COX-2) and labeling intensity were highly variable; however, moderate labeling intensity and scoring grades (++ and 2, respectively) were noted most commonly. Moderate labeling intensity was noted in 16 (47%) malignant melanomas, 2 melanocytomas, and 5 mixed neoplasms. A moderate scoring grade was assigned to 19 (55.9%) malignant melanomas, 3 melanocytomas, and 5 mixed neoplasms.

Figure 3—
Figure 3—

Photomicrograph of a section of a canine globe with anterior uveal malignant melanoma. Notice the granular intracytoplasmic COX-2 labeling in the neoplastic tissue (inset). Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

Figure 4—
Figure 4—

Photomicrograph of a section of a canine globe with anterior uveal melanocytoma. Notice the neoplasm invading the iridocorneal angle and ciliary body and bordering the sclera, and the granular intracytoplasmic COX-2 labeling in the iridocorneal angle that is infiltrated with neoplastic tissue (inset). Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

Figure 5—
Figure 5—

Photomicrograph of a section of a canine globe with anterior uveal melanocytoma. Notice the ciliary body, ciliary processes, and adjacent neoplasm. There is finely granular intracytoplasmic COX-2 immunoreactivity in the nonpigmented ciliary body epithelium (inset). Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

Figure 6—
Figure 6—

Photomicrograph of sections of 2 diseased globes without neoplasia from dogs with chronic keratitis. Notice the granular intracytoplasmic COX-2 immunoreactivity predominantly in the corneal epithelium and occasionally in stromal keratocytes of 1 globe (A) and in the nonpigmented ciliary body epithelium of the other globe (B). Immunohistochemical stain.

Citation: American Journal of Veterinary Research 70, 10; 10.2460/ajvr.70.10.1284

In all, 46 of 67 (68.7%) diseased globes had clinical or histologic evidence of glaucoma; 7 were diseased globes without neoplasia and 39 were globes with neoplasia. There was COX-2 expression in the aqueous outflow tract of 25 and in the ciliary body of 18 of these 46 globes. In 5 of the 18 globes with ciliary body labeling, COX-2 expression was localized to the nonpigmented ciliary body epithelium. Of the remaining 21 globes that did not have evidence of glaucoma, 16 and 12 had COX-2 expression in the aqueous outflow tract and ciliary body, respectively. Uveitis was diagnosed in 2 diseased globes without neoplasia and 8 globes with neoplasia. All 10 globes with evidence of uveitis expressed COX-2, and all but 1 expressed COX-2 in the aqueous outflow tract.

Discussion

In this study, definitive intraneoplastic expression of COX-2 was detected in only 3.5% of canine uveal melanocytic neoplasms. However, it is highly likely that this represented an underestimate of the expression of this enzyme in these neoplasms because ocular tissues that frequently were abutted or invaded by the uveal neoplasm (such as the iris, ciliary body, aqueous outflow tract, and anterior sclera) expressed COX-2 but were not included in this count unless expression was categorically identified in a neoplastic cell. The finding that COX-2 expression was infrequently (ciliary body muscle and aqueous outflow tract) or never (iris, ciliary body stroma, nonpigmented ciliary body epithelium, and anterior sclera) detected in these adjacent tissues in nondiseased globes strongly supports the theory that COX-2 expression in these surrounding tissues may actually have originated from invasive and expansile neoplastic cells. However, the proportion of globes with neoplasia in which definitive COX-2 expression was detected was lower than that reported for human uveal malignant melanomas. In 3 studies of a total of 87 human uveal malignant melanomas, the percentage of tumors in which COX-2 expression was detected was 13%,30 58%,31 and 91%.32 The different rates reported among those studies may have resulted from the methods used or interpretation of labeling. For example, none of the studies30–32 used bleached sections as in the present study. However, bleaching with hydrogen peroxide is effective when used in conjunction with immunohistochemical analysis,33,34 and this method has been used effectively in a study35 examining COX-2 expression. Additionally, in the present study, COX-2 expression was consistent between positive control samples that underwent hydrogen peroxide bleaching and those that did not. Therefore, it appears unlikely that bleaching affected COX-2 detection in the present study. It is also possible that the relatively infrequent COX-2 expression detected in canine neoplasms represents a true species difference between dogs and humans or may be because only uveal malignant melanomas were examined in the reports of human disease, whereas benign and mixed tumors, in addition to malignant melanomas, were examined in the present study.

Because COX plays a role in the release of prostaglandins during inflammation, the COX-2 expression detected in globes with uveal melanocytic tumors may have been expressed by nonneoplastic or neoplastic cells as part of the associated inflammatory response rather than as a direct result of the neoplasm. Certainly, some of the globes with neoplasia that expressed COX-2 in the present study had histologic evidence of uveitis. For this reason, a control group of inflamed globes without neoplasia was included for comparison. Although statistical comparisons among groups were not possible in this study, some subjective observations were possible. Expression of COX-2 was largely similar between the globes with neoplasia and diseased globes without neoplasia; however, 2 major differences were noted. All 10 diseased globes without neoplasia expressed COX-2 in the cornea, compared with only 7% to 32% of globes with neoplasia and none of the nondiseased globes. By contrast, COX-2 expression was never detected in lens epithelium of diseased globes without neoplasia but was detected in 12% of globes with melanocytic tumors and 1 of 4 of nondiseased globes. No expression was detected in lenses of globes with mixed tumors. Finally, diseased globes (those with neoplasia and those with nonneoplastic diseases) expressed COX-2 in a range of tissues such as the cornea, corneoscleral limbus, and retina that did not express COX-2 in any of the nondiseased globes. Because the most common histologic diagnosis in diseased globes without neoplasia of the present study was chronic superficial keratitis, and because corneal COX-2 expression has been reported, especially following corneal injury,36–39 corneal COX-2 expression in this group was expected. Taken together, these data suggest that COX-2 expression is likely upregulated by canine ocular disease in general and specifically in some globes with uveal neoplasia.

It is also possible that the coexistence of glaucoma in some of the globes with neoplasia may have altered COX-2 expression. In a study37 of 17 glaucomatous and 5 nondiseased canine globes, COX-2 expression was greater in glaucomatous globes, particularly in the cornea, sclera, trabecular meshwork, angular aqueous plexus, and ciliary body. The authors suggested that the expression of COX-2 might correlate with prostaglandin involvement in the pathogenesis of glaucoma; however, results from inflamed but nonglaucomatous globes were not reported. By contrast, less COX-2 expression has been reported in the nonpigmented ciliary body epithelium of human globes with end-stage primary open-angle glaucoma, compared with nondiseased human globes.40 In addition to species differences, it is likely that the pathogenesis of primary open-angle glaucoma, which may in part be caused by impaired COX-2 expression, varies substantially from the primary or secondary angle-closure glaucoma that occurs in canine globes. In the present study, it appeared that COX-2 expression in the aqueous outflow tract and ciliary body occurred in inflamed globes irrespective of the presence of glaucoma.

An additional purpose of the present study was to assess the value of COX-2 expression as a predictor of malignancy. When COX-2 expression of uveal malignant melanomas and mixed tumors was compared with that of benign uveal melanocytomas, 3 observations were possible. First, definitive intraneoplastic labeling was detected only in globes infiltrated by malignant melanoma or a mixed tumor. Second, most globes with uveal malignant melanoma had COX-2 expression in the ciliary body whereas this was a less common finding in globes with mixed or benign tumors. Third, only globes with uveal malignant melanoma had detectable COX-2 expression in iridal tissue. Although this was not a consistent finding in globes with uveal malignant melanoma, COX-2 expression in the iris was not observed in globes from any other category. Therefore, although detection of COX-2 expression might be used to support a diagnosis of malignancy, its absence should not be used as confirmation that a uveal melanocytic tumor is benign. By contrast, in human uveal malignant melanomas, a combination of COX-2 expression score and intensity was found to be predictive of death caused by metastasis in 1 study.32 This difference may represent differences in the way these tumors behave among species or may reflect the reported difficulty in correlating histologic and behavioral criteria of malignancy for this tumor type in dogs.41 A study35 examining COX-1 and COX-2 expression in various canine neoplasms found that COX-2 expression was apparent in 9 of 15 oral malignant melanomas. However, like us, the authors of that study35 also detected variable intensity of COX-2 expression in those neoplasms. Although limited COX-2 expression in the uveal melanocytic neoplasms in the present study was detected, therapeutic inhibition of COX-2 by use of NSAIDs may still be effective. For example, response to COX inhibitors was detected in COX-2–negative human colorectal cancer cells,26 and COX inhibitors may exert non-COX–related antitumor mechanisms.27 In particular, it is possible that COX inhibition in the vasculature and not the neoplasm itself is important for antitumor activity.28

ABBREVIATION

COX

Cyclooxygenase

a.

Catalogue No. 00206, Surgipath, Winnipeg, MB, Canada.

b.

DAKO Target Retrieval Solution, DAKO, Carpinteria, Calif.

c.

Purified mouse anti-COX-2, BD Biosciences, San Jose, Calif.

d.

Purified mouse IgG1 Isotype Control, BD Biosciences, San Jose, Calif.

e.

Envision+ anti-mouse HRP, DAKO, Carpinteria, Calif.

f.

Aminoethyl Carbazole Substrate Kit, Zymed Laboratories Inc, South San Francisco, Calif.

References

  • 1.

    Radi ZA, Render JA. The pathophysiologic role of cyclo-oxygenases in the eye. J Ocul Pharmacol Ther 2008;24:141151.

  • 2.

    Sano H, Kawahito Y, Wilder RL, et al. Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 1995;55:37853789.

  • 3.

    Tsujii M, Kawano S, Tsuji S, et al. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705716.

  • 4.

    Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci U S A 1997;94:33363340.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Berg J, Christoph T, Bodenteich A, et al. Heterogeneous distribution of COX-2 over-expression in human colon carcinoma cells. Adv Exp Med Biol 1997;433:327330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Eberhart CE, Coffey RJ, Radhika A, et al. Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:11831188.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7.

    Jaeckel EC, Raja S, Tan J, et al. Correlation of expression of cyclooxygenase-2, vascular endothelial growth factor, and peroxisome proliferator-activated receptor delta with head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2001;127:12531259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Marrogi AJ, Travis WD, Welsh JA, et al. Nitric oxide synthase, cyclooxygenase 2, and vascular endothelial growth factor in the angiogenesis of non-small cell lung carcinoma. Clin Cancer Res 2000;6:47394744.

    • Search Google Scholar
    • Export Citation
  • 9.

    Han SL, Tang HJ, Hua YW, et al. Expression of COX-2 in stomach cancers and its relation to their biological features. Dig Surg 2003;20:107114.

  • 10.

    Uefuji K, Ichikura T, Mochizuki H, et al. Expression of cyclooxygenase-2 protein in gastric adenocarcinoma. J Surg Oncol 1998;69:168172.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Shirahama T. Cyclooxygenase-2 expression is up-regulated in transitional cell carcinoma and its preneoplastic lesions in the human urinary bladder. Clin Cancer Res 2000;6:24242430.

    • Search Google Scholar
    • Export Citation
  • 12.

    Marrogi A, Pass HI, Khan M, et al. Human mesothelioma samples overexpress both cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase (NOS2): in vitro antiproliferative effects of a COX-2 inhibitor. Cancer Res 2000;60:36963700.

    • Search Google Scholar
    • Export Citation
  • 13.

    Shiota G, Okubo M, Noumi T, et al. Cyclooxygenase-2 expression in hepatocellular carcinoma. Hepatogastroenterology 1999;46:407412.

  • 14.

    Goulet AC, Einsphar JG, Alberts DS, et al. Analysis of cyclooxygenase 2 (COX-2) expression during malignant melanoma progression. Cancer Biol Ther 2003;2:713718.

    • Search Google Scholar
    • Export Citation
  • 15.

    Dore M, Lanthier I, Sirois J. Cyclooxygenase-2 expression in canine mammary tumors. Vet Pathol 2003;40:207212.

  • 16.

    Brunelle M, Sartin EA, Wolfe LG, et al. Cyclooxygenase-2 expression in normal and neoplastic canine mammary cell lines. Vet Pathol 2006;43:656666.

  • 17.

    Heller DA, Clifford CA, Goldschmidt MH, et al. Cyclooxygenase-2 expression is associated with histologic tumor type in canine mammary carcinoma. Vet Pathol 2005;42:776780.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    McEntee MF, Cates JM, Neilsen N. Cyclooxygenase-2 expression in spontaneous intestinal neoplasia of domestic dogs. Vet Pathol 2002;39:428436.

  • 19.

    Khan KN, Stanfield KM, Trajkovic D, et al. Expression of cyclooxygenase-2 in canine renal cell carcinoma. Vet Pathol 2001;38:116119.

  • 20.

    Khan KN, Knapp DW, Denicola DB, et al. Expression of cyclooxygenase-2 in transitional cell carcinoma of the urinary bladder in dogs. Am J Vet Res 2000;61:478481.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Kleiter M, Malarkey DE, Ruslander DE, et al. Expression of cyclooxygenase-2 in canine epithelial nasal tumors. Vet Radiol Ultrasound 2004;45:255260.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    L'Eplattenier HF, Lai CL, van den HR, et al. Regulation of COX-2 expression in canine prostate carcinoma: increased COX-2 expression is not related to inflammation. J Vet Intern Med 2007;21:776782.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23.

    Tremblay C, Dore M, Bochsler PN, et al. Induction of prostaglandin G/H synthase-2 in a canine model of spontaneous prostatic adenocarcinoma. J Natl Cancer Inst 1999;91:13981403.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Pestili de Almeida EM, Piche C, Sirois J, et al. Expression of cyclo-oxygenase-2 in naturally occurring squamous cell carcinomas in dogs. J Histochem Cytochem 2001;49:867875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25.

    Mullins MN, Lana SE, Dernell WS, et al. Cyclooxygenase-2 expression in canine appendicular osteosarcomas. J Vet Intern Med 2004;18:859865.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    Yu HG, Huang JA, Yang YN, et al. The effects of acetylsalicylic acid on proliferation, apoptosis, and invasion of cyclooxygenase-2 negative colon cancer cells. Eur J Clin Invest 2002;32:838846.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Song X, Lin HP, Johnson AJ, et al. Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.

    Masferrer JL, Leahy KM, Koki AT, et al. Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:13061311.

    • Search Google Scholar
    • Export Citation
  • 29.

    Khan KN, Venturini CM, Bunch RT, et al. Interspecies differences in renal localization of cyclooxygenase isoforms: implications in nonsteroidal antiinflammatory drug-related nephrotoxicity. Toxicol Pathol 1998;26:612620.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Ozdal PC, Callejo S, Caissie AL, et al. Cyclooxygenase-2 expression in human irradiated uveal melanomas. Int Ophthalmol 2008;28:16.

  • 31.

    Figueiredo A, Caissie AL, Callejo SA, et al. Cyclooxygenase-2 expression in uveal melanoma: novel classification of mixedcell-type tumours. Can J Ophthalmol 2003;38:352356.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    Cryan LM, Paraoan L, Hiscott P, et al. Expression of COX-2 and prognostic outcome in uveal melanoma. Curr Eye Res 2008;33:177184.

  • 33.

    Orchard GE. Heavily pigmented melanocytic neoplasms: comparison of two melanin-bleaching techniques and subsequent immunohistochemical staining. Br J Biomed Sci 1999;56:188193.

    • Search Google Scholar
    • Export Citation
  • 34.

    Li LX, Crotty KA, Kril JJ, et al. Method of melanin bleaching in MIB1-Ki67 immunostaining of pigmented lesions: a quantitative evaluation in malignant melanomas. Histochem J 1999;31:237240.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35.

    Mohammed SI, Khan KN, Sellers RS, et al. Expression of cyclooxygenase-1 and 2 in naturally-occurring canine cancer. Prostaglandins Leukot Essent Fatty Acids 2004;70:479483.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36.

    Sellers RS, Silverman L, Khan KN. Cyclooxygenase-2 expression in the cornea of dogs with keratitis. Vet Pathol 2004;41:116121.

  • 37.

    Marshall JL, Stanfield KM, Silverman L, et al. Enhanced expression of cyclooxygenase-2 in glaucomatous dog eyes. Vet Ophthalmol 2004;7:5962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38.

    Amico C, Yakimov M, Catania MV, et al. Differential expression of cyclooxygenase-1 and cyclooxygenase-2 in the cornea during wound healing. Tissue Cell 2004;36:112.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 39.

    Bonazzi A, Mastyugin V, Mieyal PA, et al. Regulation of cyclooxygenase-2 by hypoxia and peroxisome proliferators in the corneal epithelium. J Biol Chem 2000;275:28372844.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40.

    Maihofner C, Schlotzer-Schrehardt U, Guhring H, et al. Expression of cyclooxygenase-1 and -2 in normal and glaucomatous human eyes. Invest Ophthalmol Vis Sci 2001;42:26162624.

    • Search Google Scholar
    • Export Citation
  • 41.

    Giuliano EA, Chappell R, Fischer B, et al. A matched observational study of canine survival with primary intraocular melanocytic neoplasia. Vet Ophthalmol 1999;2:185190.

    • Crossref
    • Search Google Scholar
    • Export Citation

Contributor Notes

Dr. Paglia's present address is Animal Eye Center, 5175 Pacific St, Ste A, Rocklin, CA 95677.

Supported by an unrestricted gift from Novartis Animal Health US Incorporated.

The authors thank Diane Naydan, Dr. Christopher Reilly, and John Doval for technical assistance.

Address correspondence to Dr. Maggs (djmaggs@ucdavis.edu).