Abstract
The vast majority of mammary tumors in cats are malignant with high metastatic and mortality rates. This study evaluated clinical, cytological and histopathological aspects of 20 cases of cat mammary masses deriving from the city of Rio de Janeiro. Data on history, macroscopic description, evidence of regional lymph node and distant metastases, clinical examination and treatment were collected. The cats were treated with unilateral mastectomy, unilateral mastectomy plus chemotherapy or chemotherapy alone. Tissues were submitted for cytological and histological evaluation. Histopathological analysis described one case of mammary hyperplasia and 19 cases of adenocarcinoma. Inadequate surgical margins were obtained in two cats while six showed evidence of microscopic invasion in other mammary glands. Multiple mammary involvement and greater tumor size (diameter over 7.05 cm) were correlated with increased chance of local metastases. Ulceration of tumors and metastases were common. These facts support the highly malignant nature of this neoplasia and emphasize the risks associated with delay in diagnosis and treatment of feline patients. In conclusion, mammary palpation should be performed routinely in the clinical practice.
Mammary tumors are known to be the third most frequently occurring tumor in the cat, following hematopoietic neoplasms and skin tumors (Dorn et al 1968, Carpenter et al 1987). The vast majority of mammary tumors reported in cats are malignant, reported previously as 85–90% of all mammary gland masses (Hayes et al 1981, Jubb et al 1993) and is an important cause of mortality in middle-aged and elderly female cats. Mammary carcinoma has a high metastatic potential and mortality rate (Hayes et al 1981, Carpenter et al 1987).
The majority of affected cats are intact females; however, the disease is frequently seen in oophorectomized females (Hayes et al 1981, Weijer and Hart 1983). One study suggests that intact female cats have a sevenfold greater risk of developing mammary cancer than oophorectomized cats (Dorn et al 1968). Cats that are spayed before 1 year of age are less likely to develop mammary carcinoma (Overley et al 2005).
Many investigators have documented an abnormal mammary gland growth as a side effect of progestational contraceptive therapy, including mammary cancer and fibroepithelial hyperplasia (Hayden et al 1981, 1989).
Mammary tumors have been reported to occur in cats from 9 months to 23 years of age, with a mean age of occurrence of 10–12 years (Weijer and Hart 1983, Tomlinson et al 1984, Carpenter et al 1987, Ito et al 1996, Viste et al 2002).
The tumors are usually firm and nodular, single or multiple, and some studies suggest that the caudal glands are more frequently involved (Weijer and Hart 1983, Carpenter et al 1987). At least 25% of the affected cats had ulcerated masses in one study (Carpenter et al 1987) and some studies report the presence of metastasis in mammary cancer cats to vary from 22.7% (Hayes et al 1981) to 70.6% (Weijer et al 1972).
A definitive diagnosis is obtained by tissue biopsy and histopathology (Ogilvie 1992). More than 80% of the feline mammary tumors are histologically classified as adenocarcinomas (Hayes et al 1981). Unilateral or bilateral mastectomies are recommended because they reduce the chance of tumor recurrence when compared with mastectomy or lumpectomy (Weijer et al 1972, MacEwen et al 1984).
The most common benign mammary growth in cats is fibroepithelial hyperplasia, which involves one or more glands and is frequently seen 1–2 weeks following estrus or hormone therapy and in pregnant females (Hayden et al 1981, Jubb et al 1993). This is a non-neoplastic lesion responsive to progesterone stimulation (Hayden et al 1981, 1989).
The objective of this paper was to evaluate 20 cases of mammary masses in cats from Rio de Janeiro, using clinical findings, cytology and histopathology.
Materials and methods
Twenty cats with mammary masses were evaluated at veterinary hospitals in Rio de Janeiro, Brazil, between February 2002 and December 2004. The animals were followed during the clinical examination, diagnosis and treatment.
Data concerning age, breed, history of hormonal therapy, mammary hyperplasia, and pregnancy, reproductive state, age at the time of ovariohysterectomy and time between detection by the owners and diagnosis, were obtained through a questionnaire given to the owners. Information about the site of mammary lesions, maximum diameter of the largest mass and gross appearance of the lesion were obtained during physical examination.
Regional lymph node metastasis was identified through palpation and confirmed with histological examination. Distant metastasis was verified using three-view thoracic radiographs and ultrasound of the abdomen; these were later confirmed using necropsy or surgical biopsy histological evaluation.
Fine-needle aspiration was performed on all masses and two slides were prepared from each mammary lesion. These slides were stained with the Papanicolau method. The specimens were collected with incisional biopsy in one cat, excisional biopsy or necropsy in 19 cats and then were fixed with 10% formalin, trimmed, processed in an automatic tissue processor, sectioned at 6 μm thickness and stained with hematoxylin and eosin (H&E). Histopathological evaluation of the complete unilateral mammary chain was performed in 10 animals.
Clinical staging was according to the WHO (World Health Organization) system for cats (Ito et al 1996).
The choice of treatment was chosen based on client choice and financial resource, gross appearance of the lesion, presence of metastasis at the time of diagnosis and cytological evaluation. Surgery, surgery plus chemotherapy, chemotherapy only or no treatment, together with the prognosis of the cats were discussed with the owners. The type of surgical treatment performed was unilateral mastectomy; the chemotherapy protocol used was as published by Jeglum et al 1985, which consisted of five doses of doxorubicin (25 mg/m2 IV every 3 weeks) combined with cyclophosphamide (100 mg/m2 PO daily for 4 days). Two animals were euthanased at their owner's request, following discussion (Beaver et al 2000).
For statistical comparison, the following statistical methods were used:
Person correlation test was used to assess
time between detection by the owners and diagnosis of lesion and maximal tumor diameter;
maximal tumor diameter and time of survival.
Wilcoxon–Mann–Whitney test was used to assess
presence of local metastasis and time between detection by the owners and diagnosis of lesion;
presence of local metastasis and maximal tumor diameter;
number of tumors and time of survival;
presence of ulceration and time of survival;
cellular pleomorphism and time of survival;
presence of metastasis and time of survival;
clinical staging and time of survival;
Fisher's Exact test was used to assess
number of tumors and presence of local metastasis.
A value of P<0.05 was considered significant.
Results
General data
Of the 20 cats, 11 were domestic shorthair (DSH), seven were Siamese and two were Persian (Table 1). The age of the cats ranged from 2 to 17 years, with an average of 10 years. Four cats had been treated with progestational drugs for contraceptive therapy annually during their life; three owners did not know whether their cats had received hormone therapy. Two cats had a history of primary mammary hyperplasia before the diagnosis of the current mammary masses; one of these developed a mammary carcinoma after 2 years and the other had a recurrence of the hyperplasia in 2 months. Also, five animals had a history of pregnancy with an average age at the time of parturition of 2.6 years. Fourteen cats were spayed and only one had ovariohysterectomy before 1 year of age. The other 13 cats were spayed recently, with an average age at the time of surgery of 5 years (Table 1). The time between detection by the owner and diagnosis was a median of 10.5 weeks (range 1–48 weeks).
Results of general data of the studied animals
DSH=domestic shorthair.
Clinical data
Of the 20 animals studied 14 (70.0%) had one tumor and six had more than one tumor, in one or both mammary chains (Table 2). No mammary gland was spared, but tumors occurred less frequently in the cranial glands. Ten mammary masses (37.04%) were localized in the cranial glands (M1 and M2) and 17 masses (62.96%) were present in the caudal glands (M3 and M4). The maximum diameter of the largest mass had a median of 5.37 cm, with a range of diameter of 0.3–30.0 cm (Fig 1). The 30.0 cm gland was a recurrence of a large mammary hyperplasia. Seven cats had mammary ulceration (Table 3).
Site distribution of mammary lumps
Macroscopy of feline mammary adenocarcinoma. (A) Cat 2. Mammary lump in left inguinal breast measuring 2.8cm. Note the swollen nipple. (B) Cat 18. Mammary lump in right abdominal breast, with 14.0cm of maximum diameter and fixed to underlying muscle. (C) Cat 1. Mammary lump in left abdominal breast. It is ulcerated and fixed to abdominal muscle, with 14.0cm of maximum diameter.
Characteristics of gross appearance of the lesions and frequency of local metastasis and distant metastasis of the cats studied
M1=axillary; M2=thoracic; M3=abdominal; M4=inguinal; ALN=axillary lymph node; ILN=inguinal lymph node; R=right; L=left.
Eight cats had local metastasis in the regional lymph nodes and also had tumor extension into the abdominal muscle. One of these also had metastasis in the cutaneous tissue. Distant metastases were verified with pulmonary metastasis in four cats and one cat also had metastasis to the liver.
Cytological and histopathological data
In five cats cytologic evaluation of the mammary mass was declined by the owner. In the other 15 cats, cytological evaluation suggestive of carcinoma was consistent with the histopathological diagnosis. The cells were arranged in clusters, which were composed of malignant epithelial cells with well-defined, round to oval nuclei. Also, a marked variation in cell and nuclear size, well-defined nucleoli and cell pleomorphism were seen (Fig 2). Two cases had a discrete pleomorphism, nine had moderate while nine had high pleomorphism on cytologic evaluation.
Cytological analysis of feline mammary carcinoma. Note the mitosis figure (→) (Papanicolau − 400× original).
Using the WHO classification scheme described by Misdorp et al (1999), the histopathologic examination revealed one case of mammary hyperplasia and 19 cases of adenocarcinoma. Six malignant tumors were classified as tubulopapillary carcinomas, and there were also four tubular solid carcinomas, four papillary carcinomas, three tubular carcinomas, one papillary cystic carcinoma and one solid carcinoma (Fig 3).
Examples of microscopy of histopathological classification of feline mammary adenocarcinoma. (A) Cat 4, papillary carcinoma (H&E − 100× original). (B) Cat 16, tubular carcinoma (H&E − 40× original). (C) Cat 17, solid carcinoma (H&E − 100× original). (D) Cat 1, tubulopapillary carcinoma (H&E − 40× original).
The cat with mammary hyperplasia was 2 years old, had a history of mammary hyperplasia and annual progestational contraceptive therapy (cat 5). This cat was pregnant at the time of diagnosis with a 30.0 cm mass in the left inguinal mammary gland. Microscopic analysis consisted of a proliferation of mammary ducts with numerous layers of normal epithelial cells and connective tissue (Fig 4). There were rare mitotic figures and some ducts contained acidophilic material within them.
Feline mammary hyperplasia of cat 5. (A) Macroscopy of the left inguinal breast lump with a maximum diameter of 30.0cm. (B) Microscopy showing proliferation of mammary ducts and abundant connective tissue (H&E − 40× original). (C) Microscopy of the numerous layers of normal cubic and disorganized epithelial cells (H&E − 100× original).
Histopathological evaluation of unilateral excised complete mammary chain was performed in 10 animals. Inadequate surgical margins were revealed in two cases, five had evidence of tumor close to the surgical margin, and six had microscopic invasion in other mammary glands (cats 10, 12, 13, 16, 19 and 20).
Clinical staging
Three cats were classified in stage I, five stage II, seven stage III and four stage IV (Table 4). The cat with mammary hyperplasia was not classified as it was a benign condition.
Characteristics of type of treatment instituted, survival, time of recording and clinical stagings of the animals included in the study
OM=observation month; DM=death month; CM=month of conclusion of the study; T=tumor (a=not fixed; b=fixed to underlying tissue; c=fixed to abdominal muscle); N=lymph node (a=not fixed; b=fixed; (−)=negative in histology; (+)=positive in histology); M=metastasis.
Treatment and survival
Two cats were euthanased and one animal did not receive any treatment due to the owner's decision. Nine cats underwent unilateral mastectomy only. Seven received chemotherapy after surgery, which consisted of five doses of doxorubicin (25 mg/m2 IV every 3 weeks) combined with cyclophosphamide (100 mg/m2 PO daily for 4 days). Only one cat was treated with chemotherapy alone because of the presence of widespread metastasis at the time of diagnosis.
Survival was recorded for 18 cats, excluding the two cats that were euthanased at the time of presentation. Eleven cats were alive in the end of the study and seven had died because of the tumor, with a survival median of 5.29 months (Table 4).
Statistics
Only three comparisons were significant in the statistical analysis. One was the presence of local metastasis (regional lymph node and cutaneous metastasis) and maximal tumor diameter. Using Wilcoxon–Mann–Whitney test, with 5% significance, animals with larger tumors had more local metastasis (P=0.001) (Fig 5).
Local metastasis presence versus maximal tumor diameter.
Other parameters analyzed were number of tumors and presence of local metastasis. Using the Fisher's exact test, with 5% of significance, animals with more tumors had also more local metastasis (P=0.047) (Fig 6).
Frequency of number of tumors versus number of animals versus local metastasis presence.
The last parameters that were significant were the presence of ulceration and survival time. We considered time of survival to be zero for the two cats that were euthanased at presentation, and December 2004 for those that were alive at the end of the study. Using Wilcoxon–Mann–Whitney test, with 5% of significance, animals that had ulcerated tumors had a shorter time of survival (P=0.017) (Fig 7).
Presence of ulcerated tumors versus time of survival (months).
Discussion
Nineteen (95%) cats had malignant mammary gland neoplasia, suggesting a high prevalence of these tumors in this species in Rio de Janeiro, Brazil. Hence, the presence of a lump in a feline mammary gland should be interpreted with caution and prompt diagnosis is appropriate so that suitable and rapid treatment can be provided to avoid treatment failure.
The age at diagnosis ranged from 2 to 17 years, with a median of 10 years. This interval of risk age is in agreement with other studies (Weijer et al 1972, Hayes et al 1981, Weijer and Hart 1983, MacEwen et al 1984, Tomlinson et al 1984, Carpenter et al 1987, Ito et al 1996, Viste et al 2002). DSH and Siamese cats were predominant as in other studies (Hayes et al 1981, MacEwen et al 1984, Tomlinson et al 1984). This may be a reflection of breed distribution in Brazil (Souza et al 2002).
The diagnosis of mammary hyperplasia in a young pregnant female cat that had been treated with progestational contraceptive therapy was the only benign case of the study, and was consistent with the history and clinical presentation described in the literature (Hayden et al 1981, 1989).
Four animals had been treated with progestational contraceptive therapy before the diagnosis, including the one with mammary hyperplasia. This fact supports the strong predisposition of development of benign and malignant mammary lesions when this type of medication is instituted (Hayden et al 1981, 1989, Carpenter et al 1987).
One study suggests that cats spayed before 1 year of age are less likely to develop mammary carcinoma (Overley et al 2005). This fact was not supported by the current study but, among the 14 cats that were ovariohysterectomized, only one was spayed before 1 year of age.
Abdominal and inguinal mammary glands were more likely to be affected. This type of disposition is common in dogs and agrees with some reports (Weijer and Hart 1983, Carpenter et al 1987). The majority (14) of the cats had involvement of only one gland. This finding is in conflict with some reports that describe multiple mammary lesions predominating in cats with mammary carcinoma (Carpenter et al 1987, Ogilvie 1992).
Seven (35%) cats had ulceration of the mammary gland mass reflecting that these cats were presented for veterinary care after a long period of time and with advanced disease. Alternatively, ulceration may have occurred as a result of rapid tumor growth. Also, 12 (63.16%) cats with malignant neoplasia had metastasis. Eight (42%) had local metastasis and four (21%) had distant metastasis. Some studies report the presence of metastasis in mammary cancer cats to vary from 22.7% (Hayes et al 1981) to 70.6% (Weijer et al 1972). Regional lymph node metastasis occurs in 6% (Hayden and Nielsen 1971) to 82.8% (Weijer et al 1972) of cats in some studies. Distant metastasis was present in the lung in 9.8% (Hayes et al 1981) to 83.6% (Weijer et al 1972). There was evidence of liver metastasis in 4.0% (Hayden and Nielsen 1971) to 23.6% (Weijer et al 1972) of the cats. So metastasis occurs frequently in cats with mammary carcinoma and represents an important cause of death in these animals (Hayes et al 1981, Weijer and Hart 1983, Carpenter et al 1987, Ito et al 1996). Consequently, we advise that early diagnosis is important in the management of feline mammary gland tumors.
Animals with multiple tumors (P=0.047) and cats with large tumors (P=0.001) also had higher rate of local metastasis. Cats with tumor ulceration had a shorter survival time (P=0.017). This suggests that tumors were not only large but also aggressive.
The cytological characteristics of the mammary changes concurred with the histopathological results. This demonstrates that cytology had a great importance in the evaluation of the cats to make a differential diagnosis and in the determination of the prognosis.
It is imperative that the entire excised mammary chain be submitted for histopathological evaluation. In the 10 animals that had complete histopathological evaluation of the mammary chain, it was determined that the tumor extended to the surgical margin in two animals. Five cats had evidence of tumor close to the surgical margin, and six had microscopic tumor invasion in other mammary glands. These results raise some concerns, for example, when a simple mastectomy is performed and the adjacent mammary glands are not evaluated for evidence of tumor involvement. Also, this form of histopathological analyses can determine which cats possess greater risk of recurrence, and which cats should have chemotherapy postoperatively.
The cats were evaluated between February 2002 and December 2004. This short period did not allow the evaluation of the parameters for prognosis, such as type of treatment instituted, clinical staging and survival time. But when we compared staging to survival, all the cats from group I were alive in December 2004 and all cats that were treated from group IV were dead by the end of the study. For these reasons, it is suggested that additional research be done to evaluate the efficacy of treatment and survival time of feline mammary cancer, over a longer time period.
The most important factor for improving the outcome for feline mammary cancer is early diagnosis. Therefore, palpation of the mammary gland chain must be instituted as part of a routine physical examination in veterinary medicine. A correct diagnosis must be established quickly and treatment must be instituted rapidly when alteration is noted in the mammary glands.