Introduction ORIGINAL RESEARCH. Daniela Simon 1, Dorina Schoenrock 1, Ingo Nolte 1, Wolfgang Baumgärtner 2, Ronnie Barron 3, Reinhard Mischke 1

Transcription

1 Veterinary Clinical Pathology ISSN ORIGINAL RESEARCH Cytologic examination of fine-needle aspirates from mammary gland tumors in the dog: diagnostic accuracy with comparison to histopathology and association with postoperative outcome Daniela Simon 1, Dorina Schoenrock 1, Ingo Nolte 1, Wolfgang Baumgärtner 2, Ronnie Barron 3, Reinhard Mischke 1 1 Small Animal Hospital and 2 Department of Pathology, University of Veterinary Medicine, Hannover, Germany; and 3 Department of Veterinary Pathology, Faculty of Veterinary Medicine, University of Glasgow, Glasgow, Scotland Key Words Cytology, dog, prognosis, sensitivity, specificity Correspondence Daniela Simon, Small Animal Hospital, University of Veterinary Medicine Hannover, Bischofsholer Damm 15, D Hannover, Germany daniela.simon@tiho-hannover.de DOI: /j X x Background: Mammary tumors are the most common neoplasms in female dogs. Malignant tumors may carry a poor prognosis and necessitate surgery. Few data are available on the value of cytologic examination as a diagnostic or prognostic tool for mammary tumors in dogs. Objectives: The objectives of this study were to determine whether cytologic findings in fine-needle aspirate specimens of canine mammary tumors correlate with histopathologic results and whether the cytologic diagnosis is associated with postoperative outcome. Methods: In this prospective study, fine-needle aspirate samples were obtained from 50 mammary tumors in 50 dogs. Results of cytologic and histopathologic examination were compared, using the histologic diagnosis as the reference method. Kaplan Meier log rank analysis was used to evaluate univariate association of the cytologic diagnosis with duration of survival, local control, and metastasis-free interval. Results: Adequate cytologic samples were obtained in 43/50 (86%) cases. The cytologic diagnosis correlated with the histologic diagnosis for benign and malignant tumors in 40/43 (93%) and 35/43 (81%) cases, respectively. Cytologic examination had a sensitivity of 88% and a specificity of 96% for the diagnosis of malignancy. The cytologic diagnosis had significant univariate association with duration of survival (P =.016), recurrence-free interval (P =.003), and metastasis-free interval (P =.014). Conclusions: Cytologic examination of mammary tumors in the dog has satisfactory accuracy, sensitivity, and specificity for the diagnosis of malignancy and is associated with postoperative outcome. Further studies on the diagnostic accuracy of cytology as well as multivariate analysis of its preoperative prognostic value in mammary tumors in the dog are warranted. Introduction Mammary tumors are the most frequent type of neoplasm in the female dog. 1,2 As a consequence of early neutering, the incidence of canine mammary tumors has noticeably decreased in the United States 2,3 ; however, they still are commonly found in unspayed female dogs. 4 7 Malignant mammary tumors often have a poor prognosis due to high rates of local recurrence as well as metastasis, necessitating surgery and in some cases adjuvant chemotherapy Preoperative determination of tumor type and differentiation between malignant and benign tumors would therefore be of value in judging an individual patient s prognosis and aiding therapeutic decision-making. Cytology often is used as a diagnostic tool in veterinary medicine and cytologic findings correlate well with histopathologic diagnoses for numerous tumor types In women, preoperative cytologic evaluation of mammary tumors is used to determine the presence of malignancy and is considered a rapid and minimally invasive alternative to surgical Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology 521

2 biopsies In several investigations, sensitivity and specificity values of 4 95% have been obtained for cytologic diagnosis of mammary tumors in comparison to histopathology. 17,18,21,22 In the dog, few data are available on the value of cytologic examination of mammary tumors. Studies in dogs report less favorable accuracy of cytologic diagnosis for mammary tumors than for other tumor types, with lower sensitivity and specificity ,23 Reasons for this have been assumed to lie, in part, with the heterogeneous composition of mammary tumors. 13,24,25 However, more recent reports describe improved accuracy of cytologic examination for the diagnosis of mammary tumors in comparison to histologic results Data comparing the cytologic diagnosis and postoperative outcome have not been reported in canine mammary tumors to date. It was the aim of this study to evaluate the accuracy of cytologic examination of canine mammary tumors in comparison to histopathologic examination. Additionally, we investigated the association of cytologic findings with postoperative patient outcome and, therefore, prognosis. Materials and Methods Patients and pretreatment evaluation Dogs presented to the Small Animal Hospital of the University of Veterinary Medicine (Hannover) between April 1998 and March 2000 and fulfilling the following criteria were eligible for this prospective investigation: (1) presence of 1 or more mammary mass (es); (2) fine-needle aspirate specimens obtained directly before surgery; (3) surgical excision; and (4) subsequent histopathologic examination. In addition, signed owner consent was required. The study protocol was reviewed and approved by the governmental animal care committee of the state of Niedersachsen, Germany. Dogs underwent complete clinical staging, as described previously, 11 consisting of a history, physical examination, documentation of tumor size and characteristics, CBC, serum biochemistry profile, thoracic radiographs (right and left lateral, ventrodorsal), abdominal radiographs (lateral), abdominal ultrasound, electrocardiogram (ECG), and if indicated, an echocardiogram. Mammary tumors and regional lymph nodes were measured directly with calipers or by ultrasonographic imaging. Cytologic and histologic examinations Fine needle aspiration was performed using a 22-G needle on each mammary mass, all by one clinician Figure 1. Fine-needle aspirate specimen of a mammary gland tumor from a dog. The monomorphic glandular epithelial cell clusters have a low degree of pleomorphism and lack clear criteria of malignancy, indicative of an adenoma. May Grünwald Giemsa, scale bar = 10 mm. (D.S.). At least 4 fine-needle aspirate specimens from different areas were obtained from each mass and at least 1 slide was prepared in a standard fashion for each aspirate. The slides were air-dried and stained with May Grünwald Giemsa (Merck, Darmstadt, Germany). All samples were examined and classified in collaboration by 2 of the authors (R.M., R.B.) who were blinded to the histopathologic results and to clinical information on the patients such as clinical stage and type of surgery performed. Cytologic diagnosis was based on an overall judgment after evaluating all of the available specimens. Tumors were classified as malignant, benign, or benign with low numbers of atypical cells. A diagnosis of adenoma was made when most epithelial cells and clusters appeared monomorphic and 5% of epithelial cells had 4 3 nuclear criteria of malignancy (Figure 1). A diagnosis of carcinoma was made when most epithelial cells and clusters appeared markedly pleomorphic and 4 20% of epithelial cells had 4 3 nuclear criteria of malignancy such as anisokaryosis, macrokaryosis, nuclear molding, and macronucleoli (Figure 2). A diagnosis of benign with low numbers of atypical cells was made when the epithelial cells and clusters predominately had features of an adenoma, but low numbers of cells (6 20%) with 4 3 nuclear criteria of malignancy were observed (Figure 3). 30,31 Other tumor types (eg, sarcoma) were diagnosed as indicated by cell type. Mesenchymal components in an epithelial neoplasm potentially leading to a diagnosis of benign or malignant mixed tumors as well as differentiation of simple and complex carcinoma were not considered in this study. A minimum average of 20 well-stained intact cells per low-power field ( 10 objective) in at least Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology

3 Figure 2. Fine-needle aspirate specimen of a mammary gland tumor from a dog. The predominance of pleomorphic glandular epithelial cell clusters with nuclear criteria of malignancy (anisokaryosis, macrokaryosis, nuclear molding, macronucleoli) is indicative of a carcinoma. May Grünwald Giemsa, scale bar = 20 mm. specimens from each mammary mass was considered adequate for cytologic evaluation. Following surgical excision, mammary masses were fixed in 10% formalin, processed routinely, and stained with H&E. Histologic evaluation was performed by a board-certified pathologist (W.B.) according to the World Health Organization classification scheme for canine mammary tumors. 32 Follow-up examination Follow-up examination in dogs with malignant mammary tumors, based on histologic diagnoses, was performed at 6 weeks and 3, 6, 9, 12, 18, 24, and 48 months following surgery. 11 Follow-up examination included physical examination, CBC, serum biochemistry profile, thoracic radiographs, and if indicated, abdominal ultrasound, ECG, and echocardiogram. In dogs with benign mammary tumors, follow-up information on patient health as well as data on endpoints was obtained via examination at the same time points following surgery or via telephone interview with the referring veterinarians. Endpoints evaluated included the time to local recurrence, distant metastasis, and overall survival. Recurrence-free interval was defined as the time from surgery to tumor recurrence in the surgical field. Metastasis-free interval was defined as the time from surgery to the detection of distant metastasis as documented on thoracic radiographs, abdominal ultrasound, or both. Overall survival was defined as the time from surgery to death due to any cause. Statistical analysis Cytologic findings were compared with the histopathologic diagnosis, with the latter as the reference method. The value of cytologic examination as a diagnostic test in detecting malignancy was analyzed by determining the sensitivity, specificity, and positive and negative predictive values. Tumors diagnosed as benign with low numbers of atypical cells in cytologic specimens were included in the benign category. Univariate Kaplan Meier log rank analysis was used to determine whether cytologic diagnosis was associated with recurrence-free interval, metastasis-free interval, and overall survival. Dogs were censored if their tumor did not recur (recurrence-free interval analysis), if no distant metastasis was detected (metastasis-free interval analysis) at the time of death or at the end of the study period, or if they were alive (survival analysis) at the end of the study. A P-value o.05 was considered significant. Statistical analyses were performed using SPSS software (SPSS 15.0, SPSS Inc., Chicago, IL, USA). Results Patients Figure 3. Fine-needle aspirate specimen of a mammary gland tumor from a dog. This sample was classified as an adenoma with low numbers of atypical cells. At the upper right is a cluster of cells with nuclear criteria of malignancy. However, the majority of cells (4 80%) in this specimen were similar to the monomorphic glandular epithelial cell cluster seen in the lower left. May Grünwald Giemsa, scale bar = 100 mm. One hundred and thirty-four dogs with 1 or more mammary gland tumors, surgical excision, and subsequent histopathologic examination were entered into the study. Of these, cytologic samples were obtained from 50 tumors in 50 dogs. The median age was 10 years (range, 5 15 years) and median weight was 20 kg (range, 2 65 kg). Forty-five dogs were intact females Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology 523

4 and 5 were spayed females. Breeds represented by 4 3 individuals were mixed-breed (n = 18), Cocker Spaniel (n = 8), and Dachshund (n = 6). The remaining 18 dogs belonged to the following breeds: German Shepherd, Afghan, Leonberger, Poodle, Beagle, Yorkshire terrier, Bobtail, Boxer, Hovawart, Münsterländer, Bullterrier, and Schnauzer. Median tumor size (longest dimension) was 35 mm (range, mm). Clinical stage ranged from 1 to 4 (stage 1, n = 18; stage 2, n = 12; stage 3, n = 17; stage 4, n = 3). Thirty-three dogs underwent total and 17 underwent partial mastectomy. Comparison of cytologic and histopathologic diagnoses Adequate cytologic samples were obtained from 43/50 tumors. Seven of 50 (14%) samples were inadequate for diagnosis or inconclusive due to insufficient number of intact cells, excessive contamination with blood, thick smears, or smearing artifacts that impaired cell morphology. Of the remaining 43 tumor samples, 24 (56%) were classified as benign, 16 (37%) as malignant, and 3 (7%) as benign with low numbers of atypical cells by cytologic examination (Table 1). In histopathologic specimens, 26/50 (60%) tumors were diagnosed as benign and 17 (40%) as malignant. The cytologic diagnosis corresponded with the histopathologic findings in the differentiation of benign and malignant tumors in 40 (93%) samples. Correct cyto-histologic correlation for different tumor types occurred in 35/43 (81%) samples. When including the inadequate specimens, cyto-histologic correlations were correct in 40/50 (80%) samples for the diagnosis Table 1. Comparison of cytologic and histopathologic results for canine mammary tumors (n = 50). Cytologic Diagnosis n Histopathologic Diagnosis Adenoma 24 Adenoma (21) Carcinoma (1) Benign mixed mammary tumor (2) Carcinoma (13) Carcinoma 15 Adenoma (1) Squamous cell carcinoma (1) Adenoma with low numbers of atypical cells 3 Carcinoma (1) Adenoma (1) Benign complex mammary tumor (1) Fibrosarcoma 1 Malignant spindle cell tumor (1) Inadequate specimens 7 Carcinoma (1) Adenoma (2) Benign complex mammary tumor (3) Lipoma (1) Table 2. Comparison of cytologic and histopathologic results for the differentiation of benign and malignant canine mammary tumors. Histopathologic Diagnosis Benign of malignancy and 35/50 (70%) samples for the diagnosis of tumor type (Table 2). The sensitivity and specificity of cytologic examination for diagnosing malignancy, using histopathology as the reference method, was 88% and 96%, respectively. A false-positive result was obtained in 1 case and false-negative results were obtained in 2 cases with cytologic examination. The positive and negative predictive values of cytologic examination were 94% and 93%, respectively. Of the inadequate samples, 1 was malignant and 6 were benign by histologic examination (Table 1). When including the inadequate specimens in the calculation of diagnostic accuracy, the sensitivity and negative predictive value decreased to 83% and 91%, respectively, whereas specificity and positive predictive value were similar at 97% and 94%. Outcome analysis Cytologic Diagnosis Malignant Benign with Atypical Cells Total Benign Malignant Total Results do not include inadequate cytologic samples (n = 7). Using histopathologic examination as the reference method, cytologic examination had a sensitivity and specificity of 88% and 96% and positive and negative predictive values of 93% and 93%, respectively. Outcome analysis was performed in 39/43 dogs in which cytologic specimens were adequate; none of these dogs were lost to follow-up. Four dogs with invasive mammary carcinomas received additional chemotherapy postoperatively 11 and were therefore excluded from outcome evaluation so as not to skew analysis by the adjuvant treatment. Median overall survival in the 39 dogs was 933 days (range, days); median recurrence-free interval (range, days) and median metastasisfree intervals (range, days) were not reached. In the univariate Kaplan Meier log rank analysis cytologic diagnosis had a significant effect on overall survival (P =.016, Figure 4), recurrence-free interval (P =.003, Figure 5), and metastasis-free interval (P =.014, Figure 6). Dogs with benign mammary tumors based on cytologic examination (n = 23) had a median survival of 1327 days; median recurrence-free 524 Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology

5 Figure 4. Kaplan Meier curve of survival time for dogs with mammary tumors based on cytologic diagnosis (n = 39). The continuous line indicates dogs with benign tumors (n = 23, median survival time 1327 days; range, days), the dotted line indicates dogs with malignant tumors (n =13, median survival time 431 days; range, days), and the interrupted line indicates dogs with benign tumors with low numbers of atypical cells (n = 3, median survival time 427 days; range, days) (P =.016; log rank test). Vertical scale bars indicate dogs censored at analysis. and metastasis-free intervals were not reached, whereas dogs with malignant mammary tumors based on cytologic examination (n = 13) had a median survival duration, recurrence-free interval, and metastasis-free interval of 431, 988, and 490 days, respectively. Figure 6. Kaplan Meier curve of metastasis-free interval for dogs with mammary tumors based on cytologic diagnosis (n = 39). The continuous line indicates dogs with benign tumors (n = 23, median metastasis-free interval not reached; range, days), the dotted line indicates dogs with malignant tumors (n = 13, median metastasis-free interval 490 days; range, days), and the interrupted line indicates dogs with predominantly benign tumors with low numbers of atypical cells (n = 3, median recurrence-free interval not reached; range, days) (P =.014; log rank test). Vertical scale bars represent dogs censored at analysis. Dogs with benign tumors with low numbers of atypical cells on cytologic examination (n = 3) had a median survival duration of 427 days and a median recurrence-free interval of 294 days; median metastasis-free interval was not reached. Discussion Figure 5. Kaplan Meier curve of local control for dogs with mammary tumors based on cytologic diagnosis (n = 39). The continuous line indicates dogs with benign tumors (n = 23, median recurrence-free interval not reached; range, days), the dotted line indicates dogs with malignant tumors (n = 13, median recurrence-free interval 988 days; range, days), and the interrupted line indicates dogs with predominantly benign tumors with low numbers of atypical cells (n = 3, median recurrence-free interval 294 days); range, days (P =.003; log rank test). Vertical scale bars represent dogs censored at analysis. In humans, cytologic examination of mammary masses has high sensitivity (87 97%) and specificity (99 100%) for the detection of malignancy. 17,18,21,22,33 However, some studies had less favorable results for cytologic examination, with sensitivity and specificity values of 91% and 77%, respectively, leading to a more critical assessment of the diagnostic utility of fine-needle aspirates for the evaluation of mammary tumors. 33,34 In comparison to the relatively large amount of data available in women, little data are available on the accuracy of cytologic examination for the diagnosis of canine mammary tumors. Two studies comparing cytology and histopathology of canine mammary tumors reported sensitivities of 21% and 65% and specificities of 56% and 94%. 13,23 Other studies also found low accuracy of cytologic examination for the diagnosis of canine mammary tumors. 12,14 In part, these results are due to the heterogeneity of canine mammary tumors, in which cell Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology 525

6 morphology is variable in different areas of the tumor. This may lead to false-negative or, in the case of highly proliferative benign lesions, false-positive diagnoses if the cytologic specimen is not adequately representative of the neoplasm. 13,24,25 Low diagnostic accuracy also has been attributed to the presence of necrosis or inflammation in mammary tumors, because these processes can result in cellular atypia. 13 In contrast to previous studies, the cytologic findings in the present study showed good accuracy, with a 93% agreement in differentiating benign and malignant tumors, when compared with histopathologic diagnoses, and 81% agreement in determining specific tumor type. In addition, sensitivity and specificity for the diagnosis of malignancy were relatively high (86% and 96%, respectively). Slide preparation and staining techniques in this study were comparable to those used in previous human and veterinary studies. Possibly, the evaluation of at least 4 aspirate specimens per tumor increased the likelihood of obtaining a representative sample for assessing multiple different areas of a heterogeneous tumor, thus improving the diagnostic accuracy. Additionally, the consensus cytologic diagnosis of 2 observers may have added to the accuracy of cytologic examination, as was shown previously for the collaborative histologic examination of human breast cancer cases. 35 The accuracy of cytologic examination found in more recent reports support the findings in the present study for canine mammary tumors. Cassali et al. 29 found a cyto-histological correlation of 93% and sensitivity and specificity of 88.6% and 100%, respectively, comparable to the results in the present study. Furthermore, Simeonov and Simeonova found good correlation of cytologic and histologic diagnoses of canine mammary tumors using computer-assisted morphometric techniques to differentiate benign and malignant tumors. Reasons for the discrepancies in cytologic accuracy between the more recent studies and prior investigations remain unclear. The use of computer-assisted techniques may have contributed to this development. Additionally, in 1 study, 29 a different staining technique was used, specimens evaluated as suspicious or inconclusive were not included into the calculations, and tumor classification was based on human tumor types rather than canine tumor types. A subset (14%) of cytologic samples in this study was of insufficient quality for diagnostic purposes. In most cases this was due to hypocellular samples. In human mammary tumor aspirates, up to 27% of samples were inadequate, making tissue biopsies for histopathologic diagnosis necessary. 17,34 In a recent study of canine mammary tumors, 25% of samples were of insufficient quality for cytologic diagnosis. 29 Factors contributing to the lower percentage of inadequate specimens in our study may have been the higher number of aspirations made of each tumor and more individual experience and training in tumor sampling, slide preparation, and slide examination. This aspect of interindividual variability raises the question of the general applicability of our study to other clinical settings; future multi-institutional studies will be needed to clarify this. Although all diagnostic procedures, including histopathologic examination, may result in errors, both false-negative and false-positive results may have unfavorable consequences in individual patients when the findings form the basis for therapeutic decisions such as whether or not to perform surgery or to the type of surgery performed. 33,34 A drawback of the cytologic evaluation in the present study was that specific tumor subtypes (eg, simple vs complex carcinoma) were not differentiated. This was not done because high number of subgroups with low sample sizes makes statistical evaluation more difficult. A limitation of this study was the low number of patients; this study was a part of a larger clinical study in which patients were accrued during a certain time period, and logistical or client-associated reasons affected sample size. Higher patient numbers would have improved the power of the study and resulted in more definite information. Additionally, the group of samples classified as adenomas with low numbers of atypical cells was very small and insufficient for analysis, so conclusions cannot be drawn. A comparison of cytologic and histopathologic results in a larger number of canine mammary tumor cases, possibly in multiinstitutional studies, is warranted to verify the results of the present investigation. An important way of evaluating diagnostic accuracy is to consider patient outcome. To date, the association between cytologic diagnosis and postoperative patient outcome in canine mammary tumors has not been investigated. In 1 recent study, cytologic diagnosis of feline mammary gland tumors via computed-assisted morphometry was not found to be associated with postoperative patient survival in 35 cats. 36 In a study of 213 cases of human mammary carcinoma, Ciatto et al 37 found a significant correlation between cytologic tumor grade and 5-year survival rate. In the present study, the cytologic diagnosis was significantly associated with postoperative survival, duration of local tumor control, and metastasis-free interval. Preoperative cytologic evaluation of mammary tumors, therefore, may provide prognostic as well as diagnostic information and help with clinical decision-making; for example, deciding whether or not to perform 526 Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology

7 surgery. However, because only univariate analysis was performed, the cytologic diagnosis does not represent an independent prognostic variable but rather may be influenced by other confounding factors. Ciatto et al, 37 for example, found that the significant prognostic association of cytologic grading in univariate analysis was not reproducible in the subsequent multivariate analysis adjusting for additional significant variables. Histologic parameters of malignancy and prognosis, such as tumor type, grade, degree of invasion, and stage, cannot be determined cytologically. 8,9,11 Because the cytologic findings correlated well with histologic findings in this study, its influence on outcome may be skewed. However, the preoperative availability of cytologic examination raises its potential as a diagnostic tool for assessing prognosis. Other preoperative parameters such as clinical stage, tumor size, and ulceration also been an association with prognosis 11,38,39 and may be more important than cytologic finding on treatment outcome. Evaluation of unique or specific cytopathologic features that may have prognostic value in canine mammary tumors, and multivariate analyses that include cytologic findings as a variable, should be evaluated further. In summary, cytologic examination has satisfactory sensitivity and specificity for the differentiation of malignant and benign canine mammary tumors. Furthermore, the cytologic diagnosis is significantly associated with postoperative patient outcome in univariate survival analysis. Further studies on the diagnostic accuracy of cytology as well as multivariate analysis of its preoperative prognostic value in mammary tumors in the dog are warranted. References 1. Dorn CR, Taylor DO, Frye FL, et al. Survey of animal neoplasms in Alameda and Contra Costa Counties, California. I. Methodology and description of cases. J Natl Cancer Inst. 1968;40: Rutteman GR, Withrow SJ, MacEwen EG. Tumors of the mammary gland. In: Withrow SJ, MacEwen EG, eds: Small Animal Clinical Oncology. 3rd ed. Philadelphia, PA: WB Saunders; 2001: Sorenmo K. Canine mammary gland tumors. Vet Clin North Am Small Anim Pract. 2003;33: Eichelberg H, Seine R. Life expectancy and cause of death in dogs. I. The situation in mixed breeds and various dog breeds. Berl Munch Tierarztl Wochenschr. 1996;109: Simon D, Goronczy P, Stephan I, et al. Canine mammary gland tumors: investigation on incidence and course of the disease. Der Praktische Tierarzt. 1996;77: Moe L. Population-based incidence of mammary tumours in some dog breeds. J Reprod Fertil. 2001;57(Suppl): Dobson JM, Samuel S, Milstein H, et al. Canine neoplasia in the UK: estimates of incidence rates from a population of insured dogs. J Small Anim Pract. 2002;43: Gilbertson SR, Kurzman ID, Zachrau RE, et al. Canine mammary epithelial neoplasms: biologic implications of morphologic characteristics assessed in 232 dogs. Vet Pathol. 1983;20: Kurzman ID, Gilbertson SR. Prognostic factors in canine mammary tumors. Semin Vet Med Surg (Small Anim). 1986;1: Karayannopoulou M, Kaldrymidou E, Constantinidis TC, et al. Adjuvant post-operative chemotherapy in bitches with mammary cancer. J Vet Med A Physiol Pathol Clin Med. 2001;48: Simon D, Schoenrock D, Baumgärtner W, Nolte I. Postoperative adjuvant treatment of invasive malignant mammary gland tumors in dogs with doxorubicin and docetaxel. J Vet Intern Med. 2006;20: Griffiths GL, Lumsden JH, Valli VE. Fine needle aspiration cytology and histologic correlation in canine tumors. Vet Clin Pathol. 1984;13: Allen SW, Prasse KW, Mahaffey EA. Cytologic differentiation of benign from malignant canine mammary tumors. Vet Pathol. 1986;23: Stockhaus C, Teske E. Clinical experiences with cytology in the dog. Schweiz Arch Tierheilkd. 2001;143: Stockhaus C, Schoon HA, Grevel V, Oechtering G, Teske E. The diagnostic value of cytology in the diagnosis of soft tissue sarcoma in the dog and cat. Tierärztl Prax. 2003;31: Reinhardt S, Stockhaus C, Teske E, Rudolph R, Brunnberg L. Assessment of cytological criteria for diagnosing osteosarcoma in dogs. J Small Anim Pract. 2005;46: Cheung PS, Yan KW, Alagaratnam TT. The complementary role of fine needle aspiration cytology and tru-cut needle biopsy in the managment of breast masses. Aust NZ J Surg. 1987;57: Ciatto S, Rosselli del Turco M, Ambrogetti D, et al. Solid nonpalpable breast lesions. Success and failure of guided fine-needle aspiration cytology in a consecutive series of 2444 cases. Acta Radiol. 1997;38: Costa MJ, Tadros T, Hilton G, Birdsong G. Breast fine needle aspiration cytology. Utility as a screening tool Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology 527

8 for clinically palpable lesions. Acta Cytol. 1993;37: Orell SR, Miliauskas J. Fine needle biopsy cytology of breast lesions: a review of interpretative difficulties. Adv Anat Pathol. 2005;12: Tschubel K, Helpap B. Comparative studies of simultaneous intraoperative frozen sections and imprint cytology of the mamma. Z Krebsforsch Klin Onkol Cancer Res Clin Oncol. 1975;84: Saravanja S, Kubik-Huch RA, Komminoth P, Jöstingmeier S, Hohl M, Otto RC. Ultrasound-guided fine-needle aspiration of the breast. Schweiz Rundsch Med Prax. 2005;94: Hellmén E, Lindgren A. The accuracy of cytology in diagnosis and DNA analysis of canine mammary tumours. J Comp Pathol. 1989;101: Misdorp W, Hart AA. Canine mammary cancer. I. Prognosis. J Small Anim Pract. 1979;20: Rutteman GR. Mammary tumors in the dog. In: Kessler M, ed: Small Animal Oncology. Berlin, Germany: Parey; 2000: Simeonov R, Simeonova G. Fractal dimension of canine mammary gland epithelial tumors on cytologic smears. Vet Clin Pathol. 2006;35: Simeonov R, Simeonova G. Computerized morphometry of mean nuclear diameter and nuclear roundness in canine mammary gland tumors on cytologic smears. Vet Clin Pathol. 2006;35: Simeonov R, Simeonova G. Computerized cytomorphometric analysis of nuclear area, nuclear perimeter and mean nuclear diameter in spontaneous canine mammary gland tumours. Vet Res Commun. 2007;31: Cassali GD, Gobbi H, Malm C, Schmitt FC. Evaluation of accuracy of fine needle aspiration cytology for diagnosis of canine mammary tumours: comparative features with human tumours. Cytopathology. 2007;18: Alleman AR, Bain PJ. Diagnosing neoplasia: the cytological criteria for malignancy. Vet Med. 2000;95: Meinkoth HJ, Cowell RL. Recognition of basic cell types and criteria of malignancy. Vet Clin North Am Small Anim Pract. 2002;32: Misdorp W, Hellmen E, Else RW. Mammary Tumors and Dysplasias in Dogs and Cats. Histologic Classification of Tumors of Domestic Animals. Second Series, 7 (2). Washington, DC: Armed Forces Institute of Pathology; Park IA, Ham EK. Fine needle aspiration cytology of palpable breast lesions. Histologic subtype in false negative cases. Acta Cytol. 1997;41: Layfield LJ, Glasgow BJ, Cramer H. Fine-needle aspiration in the management of breast masses. Pathol Annu. 1989;24: Perkins C, Balma D, Garcia R, Susan G, Members of the Consensus Group. Komen for the cure. Why current breast pathology practices must be evaluated. Breast J. 2007;13: Simeonov R, Simeonova G. Nuclear cytomorphometry in feline mammary gland epithelial tumours. Vet J. 2009;179: Ciatto S, Bonardi R, Herd-Smith A, Cariaggi P, Confortini M, Bulgaresi P. Prognostic value of breast cancer cytologic grading: a retrospective study of 213 cases. Diagn Cytopathol. 1993;9: Hellmen E, Bergstrom R, Holmberg L, et al. Prognostic factors in canine mammary tumors: a multivariate study of 202 consecutive cases. Vet Pathol. 1993;30: Chang SC, Chang CC, Chang TJ, Wong ML. Prognostic factors associated with survival two years after surgery in dogs with malignant mammary tumors: 79 cases ( ). J Am Vet Med Assoc. 2005;227: Vet Clin Pathol 38/4 (2009) c2009 American Society for Veterinary Clinical Pathology