Diagnostic value of cytologic evaluation of tumors and tumor-like conditions of the oral cavity in dogs and cats : a prospective case study on 114 cases

Background. A wide variety of benign and malignant tumors, as well as non-neoplastic conditions have been reported to originate in the oral cavity and represent a frequent occurrence in canine and feline species. Clinically, these conditions may be asymptomatic or present with pain, bleeding or mass lesions. Traditionally, a correct diagnostic approach to differentiate inflammation, neoplasia, and the correct tumor type comprises imaging techniques, surgical and/or cytological biopsies. Cytological examination has been considered a complementary technique that may represent an alternative diagnostic method of oral lesions in these species. Objectives. The purpose of the study was to prospectively compare results of cytologic examination of fine-needle biopsy (FNAB) and impression smears from diverse lesions of the oral cavity in dogs and cats with histology set as the gold standard. Methods. Oral cavity lesions from 114 animals (85 dogs and 29 cats) of different genders and breeds were included. For each case, anamnesis, signalment, clinical signs, and in some instances additional diagnostic procedures (radiology, and CT-scan) were collected. Cases were included when biopsy specimens and cytology either performed by FNAB (non-aspiration or aspiration) or by impression were available from the same lesions to analyze and compare the diagnostic potential of both methods. Accuracy, sensitivity, specificity, PPV, NPV, as well as agreement (k) between cytological and histological results were calculated. Results. Two dogs were excluded from the statistical evaluation because of the presence of two concomitant neoplastic processes. In 16 dogs, cytologic specimens were considered unsatisfactory (acellularity, hemodilution, or necrosis) and were excluded from the statistical evaluation (retrieval rate: 85.7%). Neoplasia was diagnosed in 60 out of 67 canine cases and 21 out of 29 feline cases. Results showed an overall 97% (93/96) accuracy between the diagnosis of inflammation versus neoplasia obtained by cytology and histology. Discrepancy was due to 3 false negative cases (two nixed inflammation vs. fibromatous epulis, and one reactive fibroplasia vs. osteoma). No false positive cases were obtained. For dogs sensitivity, specificity, PPV, NPV for the diagnosis of neoplasia were respectively 0.97, 1, 1, 0.78. For cats sensitivity, specificity, PPV, NPV were respectively 0.95, 1, 1, 0.89. Grouping both species, and considering as partial agreements both the three false negative cases and partial correlations – agreement referring to neoplasia but lack of agreement referring to cell type - a k value of 0.82 (CI: 0.75-0.88) graded as very good, was obtained. Separating canine from feline species, a k values of 0.80 (CI: 0.73-0.87) and 0.88 (CI: 0.83-0.94) both graded as very good, were respectively obtained. Conclusions. Cytological examination of canine and feline oral cavity lesions attested as a valid and inexpensive procedure, with an overall unexpected high sensitivity and specificity, associated with excellent accuracy when compared with the definitive histological diagnosis.