• Open Access

CA15.3, CEA and LDH in Dogs with Malignant Mammary Tumors

Authors


  • The work was done at the Department of General Pathology, Institute of Biological Sciences, Laboratory of Comparative Pathology, Federal University of Minas Gerais. Part of this work was presented as an abstract at the 28th Brazilian Congress of Pathology and Latinoamerican Society Congress of Pathology, 2011, Maceió, Brazil

Corresponding author: Geovanni Dantas Cassali, Laboratory of Comparative Pathology, Department of General Pathology, ICB/UFMG, PO Box 486 31270-901, Belo Horizonte, MG, Brazil; e-mail: cassalig@icb.ufmg.br.

Abstract

Background

Presence of tumor markers in serum might be connected to the number of secreting cells and with the stage of the neoplasm. However, there are few studies regarding these markers in veterinary clinical oncology.

Objectives

To determine the serum concentrations of cancer antigen 15.3 (CA 15.3), carcinoembryonic antigen (CEA), and lactate dehydrogenase (LDH) in female dogs with different stages of mammary cancer.

Animals

Ninety female dogs, including 30 that were healthy, 40 that had nonmetastatic cancer, 12 with regional metastasis, and 8 with distant lymph node metastasis.

Methods

Prospective case-controlled observational study. Serum samples were collected to measure CA15.3, CEA, and LDH from 60 female dogs with mammary cancer during mastectomy and 30 healthy female dogs during routine check-up. CA15.3 and CEA were determined by chemiluminescent immunoassay and LDH by ultraviolet kinetic method. Western blotting analysis was performed to confirm the specificity and possible cross-reactivity of human CA15.3 and CEA antibodies with canine serum. Group data were compared by ANOVA followed by Student-Newman-Keuls and Tukey's tests. Correlations were investigated by Pearson and Spearman tests.

Results

CEA, CA15.3, and LDH were measurable in all groups. Higher serum concentration of CA15.3 and LDH was associated with regional and distant metastases. There was a significantly higher serum CA15.3 concentration in animals with lymph node metastasis when compared with animals without metastasis. There were no significant differences in CEA among groups. Expression of CA15.3 and CEA in canine serum was confirmed by Western blotting.

Conclusions and Clinical Importance

Serum CA15.3 can be used to distinguish nonmetastatic from metastatic carcinomas.

Abbreviations
ANOVA

analysis of variance

ASCO

American Society of Clinical Oncology

CA15.3

cancer antigen 15.3

CEA

carcinoembryonic antigen

EGTM

European Group on Tumor Markers

LDH

lactate dehydrogenase

MUC1

mucin 1

UFBA

Federal University of Bahia

UFMG

Federal University of Minas Gerais

Molecules that are associated with the presence of cancer have been extensively investigated and used in clinical practice for over 30 years.[1] Tumor markers are mostly normal cellular metabolism products that have increased production because of malignant transformation.[2] These markers indicate malignancy and are characterized by substances present in the tumor, blood, or other fluids that are primarily produced by the tumor or secondarily in response to the presence of tumor.[3] In contrast to markers obtained in primary tumor tissue samples, serum tumor markers reflect a dynamic situation and different measurements can be obtained as required.

The most widely used serum markers in breast cancer are cancer antigen 15.3 (CA 15.3) and carcinoembryonic antigen (CEA). In human medicine, the ASCO (American Society of Clinical Oncology)[4] and EGTM (European Group on Tumor Markers)[5] recommend the evaluation of these tests assist in the prognostic evaluation and treatment monitoring of women with breast cancer.

CEA is a glycoprotein that is produced by normal cells of the gastrointestinal mucosa[6] and is mainly overexpressed by colon, rectum, breast, and lung adenocarcinomas. It is involved in intercellular adhesion.[7, 8] There is a positive correlation between changes in serum CEA and therapeutic responses in patients with metastatic breast cancer.[9]

CA15.3 is a high–molecular-weight mucin glycoprotein,[7] encoded by the MUC1 gene and is related to cell adhesion.[10] Among the known mucins, MUC1 is the most studied and it plays a crucial role in regulating many cellular properties, including cell proliferation, apoptosis, adhesion, and invasion. In case of tumors, cell polarization is lost and this altered cell surface expression, coupled with the disruption of the normal tissue architecture caused by the growing tumor, allows MUC1 mucin to be shed into the circulation where it can be measured by means of immunoassays.[9] This marker has shown to be specific for human mammary neoplasms and enables monitoring of treatment responses.[11]

In addition, alterations in cellular metabolism are among the most consistent hallmarks of cancer.[12] Rapid cancer cell proliferation and high metabolic demands lead to an increase in lactate dehydrogenase (LDH), an enzyme responsible for catalyzing the reversible transformation of pyruvate to lactate, having an essential position in anaerobic cellular metabolism. Increased activity of LDH is linked with intratumoral hypoxia and acidity.[13]

Spontaneous tumors of the mammary gland are the most frequent type of neoplasm in female dogs[14-17] and the prevalence of malignant lesions varies from 26 to 73%.[18] The identification of markers that could predict tumor behavior is especially important for mammary cancer, mainly attributable to the variability in clinical disease progression.

Thus far, there are no immunoassay reagents specific to canine mammary cancer, and focus on serum tumor markers in the veterinary clinic is also scarce. The use of such markers would be an additional tool for the treatment and diagnosis of canine mammary cancer. In addition, serum tests are relatively inexpensive and require a less invasive method of sample collection when compared with immunohistochemical tests, which have been used in veterinary medicine.

In this context, this study aimed to determine serum concentrations of CEA, CA15.3, and LDH in healthy female dogs and female dogs diagnosed with mammary cancer, with and without lymph node metastasis, observing possible changes in different stages of disease. Moreover, to identify the applicability of human reagent kits for canine serum samples.

Material and Methods

The study was conducted in accordance with the Animal Experimentation Ethics Committee of Federal University of Minas Gerais (CETEA/UFMG), protocol no 258/08.

Ninety female dogs were randomly selected and used without breed preference. Each dog was admitted at the Veterinary Hospital from the Veterinary School of the Federal University of Minas Gerais (UFMG) or the Veterinary Hospital from the Veterinary Medical School from the Federal University of Bahia (UFBA). After clinical evaluation, the animals were subdivided into four groups according to the following criteria of inclusion:

  • Group I (T0)—Healthy female dogs: dogs with no evidence of mammary neoplasms in routine check-up were selected;
  • Group II (T1,2,3N0M0)–—Forty female dogs with histopathologic diagnosis of malignant epithelial mammary neoplasia, but with no evidence of metastasis in regional lymph nodes (axillary and inguinal), nonregional lymph nodes, and negative to thoracic radiographic examination;
  • Group III (T1,2,3N1,2M0)–—Twelve female dogs with malignant epithelial mammary neoplasia and evidence of regional lymph node(s) metastasis (axillary and inguinal), but without evidence of metastasis in nonregional lymph nodes and in the thoracic radiographic examination;
  • Group IV (T1,2,3N1,2M1)—Eight female dogs with malignant epithelial mammary neoplasia and evidence of nonregional lymph node metastasis, confirmed by histopathologic analysis. No evidence of pulmonary metastasis was observed through examination of thoracic radiographs.

The histopathologic diagnoses were performed by veterinarians and pathologists from the Laboratory of Comparative Pathology of the Biological Science Institute of UFMG. The criteria used for lesion diagnoses were proposed by Misdorp et al.[19]

Biochemistry Analysis and ELISA Immunoassay

All samples were collected during mastectomy by vacuum collection system1 in 5-mL plastic tubes containing separating gel without anticoagulant. After collection and clot retraction, the material was centrifuged at 3000 rpm for 5 minutes. LDH serum activity was assessed on the date of collection, and additional serum was placed in eppendorfs and frozen at −20°C to perform the CEA and CA15.3 assays after the histopathologic diagnoses were known.

The serum tests were carried out according to the protocol of each manufacturer. Together with the whole battery of assays, a control serum (supplied with the kit) was measured with a known concentration range. The reagents used were the same as those used for human diagnostics.

The LDH tests were evaluated by means of commercial kit2 and continuous ultraviolet kinetic method. The equipment Labmax 2403 was used for reaction readings. The results were expressed in U/L and compared to reference values established for the canine species (45 to 233 U/L).[20] LDH was not able to be measured in 3 samples.

To test for CEA and CA15.3, commercial solid phase, noncompetitive ELISA immunoassay kits were used (both from CanAg4). The Labmax 2403 was used for measurements. The CEA results were expressed in ng/mL, and CA15.3 was expressed in U/mL. For both tests, no reference values had been established for veterinary medicine.

SDS-PAGE and Western Blot

Thirty micrograms (30 μg) of total serum proteins was separated by electrophoresis on a denaturing 4–8% polyacrylamide-SDS gel and electro-transferred to PVDF membranes. Membranes were blocked during 1 hour with PBS containing 5% (w/v) nonfat dry milk and 0.1% Tween-20, washed 3 times with PBS containing 0.1% Tween-20 and then incubated overnight at 4°C with specific primary antibodies (CEA-HRP and CA15.3- HRP) by a dilution of 1:100 in phosphate-buffered saline containing 5% (w/v) BSA and 0.1% Tween-20. After washing, membranes were incubated with appropriate horseradish peroxidase-conjugated secondary antibody (1 : 2000). Immunoreactive bands were visualized by ECL detection system, as described by the manufacturer.5

Statistical Analysis

Statistical analysis and graphs were performed with GraphPad Prisms.5 D'Agostino & Pearson omnibus normality test was used to determine if data presented normal distribution. When possible, the analysis of variance (ANOVA) was followed by Student-Newman-Keuls (SNK) and Tukey's tests (CEA determinations were analyzed after logarithmic transformation before testing). In the same way, possible correlations were investigated by Pearson or Spearman tests. Associations and concordance were considered significant when the test significance probability had P < .05.

Results

Among the diagnosed tumors, the most frequent type was carcinoma in a mixed tumor (30/60; 50.0%), followed by carcinoma in situ and tubular carcinoma (both 6/60; 10.0%) (Table 1). The age of the animals ranged from 6 to 15 years old (mean 10.3 ± 2.1). Breeds were not used as selection criteria, but a greater frequency of the Poodle breed (28/60; 46.6%) was observed, followed by mixed-breed (13/60; 21.6%). All female dogs from group IV (T1,2,3N1,2M1) showed nonregional lymph node metastases, with absence of lung or liver metastasis, confirmed by radiograph and ultrasound techniques.

Table 1. Frequency of mammary tumor histological types in the female dogs in study (n = 60)
DiagnosisFrequency%
Carcinosarcoma12
Carcinoma in a mixed tumor3050
Carcinoma in situ610
Tubular carcinoma610
Tubulo-papillary carcinoma58
Papillary carcinoma23
Micropapillary carcinoma23
Solid carcinoma35
Solid tubular carcinoma47
Secretory carcinoma12
Total60100

The serum concentration of CA15.3 was significantly less in group I (T0) compared with groups II (T1,2,3N0M0) (P = .0029), III (T1,2,3N1,2M0) (P < .001), and IV (T1,2,3N1,2M1) (P < .001). Statistically significant differences were not observed between groups III (T1,2,3N1,2M0) and IV (T1,2,3N1,2M1). A positive (r = 0.535) and significant correlation (P < .001) between serum marker concentration and disease staging was observed (Fig 1).

Figure 1.

Positive correlation (r = 0.535, P < .001) between serum concentration of CA15.3 and staging of groups. 1. Healthy female dogs (n = 30, 1.19 ± 0.51). 2. Female dogs with mammary cancer without evidence of metastasis (n = 40, 1.61 ± 0.61). 3. Female dogs with mammary cancer and evidence of regional lymph node(s) metastasis (n = 12, 2.39 ± 1.02). 4. Female dogs with nonregional lymph node metastasis (n = 8, 2.46 ± 1.00).

The CEA test did not show a normal distribution; thus, it was log-transformed and further analyzed statistically, showing no significant difference between groups (Tukey's test: P > .05; Table 2).

Table 2. CEA and CA15.3 serum levels in female dogs from groups I, II, III, and IV (mean, SD)
Group (n)CEA (ng/mL)CA15.3 (ng/mL)
  1. The means followed by different letters in the same column differ statistically with P-value < .05. Group I: female dogs without mammary cancer; group II: female dogs with mammary cancer without metastasis; group III: female dogs with regional metastasis; group IV: female dogs with nonregional lymph node(s) metastasis.

Group I (30)0.19 ± 0.20 (a)1.19 ± 0.51 (a)
Group II (40)0.12 ± 0.12 (a)1.61 ± 0.61 (b)
Group III (12)0.29 ± 0.36 (a)2.39 ± 1.02 (c)
Group IV (8)0.07 ± 0.04 (a)2.46 ± 1.00 (c)

Increase in serum CEA and CA15.3 levels presented a statistically insignificant and weak positive correlation (r = 0.162; P = .125).

Higher LDH serum levels presented a statistically significant positive correlation (r = 0.409; < .001) for canine mammary cancer with advanced staging (Fig 2). Statistically significant differences were observed among groups I compared with groups II, III, and IV (Table 3).

Figure 2.

Positive correlation (r = 0.409, P < .001) between serum concentration of LDH and staging of groups. 1. Healthy female dogs (n = 30, 201.7 ± 84.7). 2. Female dogs with mammary cancer without evidence of metastasis (n = 40, 282.1 ± 163.9). 3. Female dogs with mammary cancer and evidence of regional lymph node(s) metastasis (n = 9, 384.4 ± 268.6). 4. Female dogs with nonregional lymph node metastasis (n = 8, 414.0 ± 178.8).

Table 3. LDH serum activity in female dogs from groups I, II, III, and IV (mean, SD)
Group (n)LDH (U/L)
  1. The means followed by identical letters in the same column do not differ statistically with P-value < .05. Group I: female dogs without mammary cancer; group II: female dogs with mammary cancer without metastasis; group III: female dogs with regional metastasis; group IV: female dogs with nonregional lymph node(s) metastasis.

Group I (30)201.7 ± 84.7 (a)
Group II (40)282.1 ± 163.9 (b)
Group III (9)384.4 ± 268.6 (b)
Group IV (8)414.0 ± 178.8 (b)

Western blotting analysis with anti MUC-1 (CA15.3) and CEA antibodies revealed the presence of reactive protein bands in human and canine serum. The band with approximate molecular weight of 250 KDa corresponds to MUC-1, whereas the approximate molecular weight of the CEA protein band appears to be 180 KDa (Fig 3). Positive and negative controls were used to confirm these results.

Figure 3.

Western blot analysis of serum proteins from a woman and bitch presenting mammary gland malignant neoplasms. (A) Overexpression of MUC-1 protein in human and canine serum (Anti- CA15.3 monoclonal mouse antibody, CanAg, Fujirebio Diagnostics. Diluition 1 : 100). (B) Overexpression of CEA protein in human and canine serum (Anti-CEA monoclonal mouse antibody, Autobio Diagnostics. Diluition 1:100).

Discussion

In this study, we demonstrated that it is possible to detect CEA and CA15.3 in serum of dogs by use of kits designed to detect the human form of these antigens. Both serum CA15.3 and LDH concentrations had positive correlations with disease staging. These findings support the potential use of CA15.3 as a prognostic factor for mammary cancer in dogs. The latter antigen was found to be specific for the identification of mammary neoplasia and useful for monitoring of the disease.[11] Furthermore, we observed that CEA could not be used as tumor marker because the serum concentration was not significantly different between healthy animals and animals with cancer.

The biology of canine mammary tumors is of interest because these tumors have been proposed as a comparative model for the study of human female breast cancer. These comparative studies are feasible because of the epidemiological and biological similarities of mammary tumors in both species.[15]

In addition, there is a great interest in veterinary medicine to achieve better monitoring of tumor development and response to treatment, aiming to increase survival rates and disease-free intervals and prioritizing the quality of life for the animal.

The assessment of some serological tests can assist in the evaluation of neoplasia development, enable the detection of metastasis and recurrences,[21] and be useful for evaluating disease progression after treatment.[4] However, at the first time, it is necessary to determine if it is possible to detect biomarkers in serum of canine species. In this study, CEA and CA15.3 have been proven to be measurable in all samples assayed with a kit to identify human antigen.

The serum tumor marker CA15.3 has been broadly studied in human medicine since 1980s; however, in veterinary medicine, there are few references about tumor markers in the literature. In this study, the results demonstrated a significant difference in the CA15.3 concentrations of group I compared with the others groups, suggesting that the increased blood concentration of the marker is proportional to the clinical stage of the disease. In human medicine, serum CA15.3 varies according to the cancer staging of the patient and increases 5 to 30% from the normal values in stage I, 15 to 50% in stage II, 60 to 70% in stage III, and 65 to 90% in stage IV.[6] Moreover, it is well known that, in humans, very high serum levels are associated with poor survival.

A previous study that aimed to establish a method for CA15.3 testing in veterinary medicine, obtained a mean serum level of 2.77 ± 0.28 for the marker in dogs without neoplastic diseases.[11] This result differs from the data obtained in our study, where in healthy female dogs, the mean serum level of CA15.3 was 1.37 ± 0.46. This difference might have occurred mainly because of different ELISA methodologies used and the use of different reagents.

The role of the CEA tumor marker in animals with neoplasms was studied and a significant increase in its serum level was observed.[22] However, in this study, the groups compared were heterogeneous in relation to sex, species (cat and dog), histopathologic diagnosis, and clinical staging. In our study evaluating only female dog mammary neoplasias, there were no statistically significant differences between the groups.

Levels of serum CEA could be influenced by the rate and amount of its production by carcinoma cells, the existence of tumor necrosis, and perivascular infiltration. Other factors that might influence the determination of serum CEA levels are the hepatic metabolism and the renal elimination rate.[23]

Another important result was the biggest relationship between the increase in CA15.3 serum levels and the disease staging compared with the increased CEA serum levels. These data suggest the possibility of using CA15.3 as a tumor marker, which is in agreement with another study[24] that compared four tumor serum markers, between CEA and CA15.3, for the detection of breast cancer in women. They concluded that the CA15.3 tumor marker had the best performance regarding diagnosis because of its higher sensitivity compared with CEA.

The LDH levels showed a positive and significant relationship with tumor staging and disease evolution. These data agree with previous studies[25, 26] that state that the bloodstream concentration of this enzyme is increased with disease evolution, indicating a poorer prognosis and affirmed that neoplastic cells observed experimentally have an increased glycolysis rate and proposed an increase in LDH activity. This is because of up-regulation of LDH in cancer cells ensures an efficient anaerobic/glycolytic metabolism for tumors and reduced dependence on oxygen.[27] According to this important observation, the enzyme demonstrated statistically significant differences among the groups of healthy dogs and the groups of dogs with mammary neoplasia.

The main finding in Western blotting analysis demonstrated good interaction between human and canine MUC-1 and CEA antibodies. This finding agrees with a previous study that demonstrated a very high similarity in the DF3 epitope of the MUC-1 protein sequence in these 2 species.[28]

In conclusion, this study observed an association between the presence of advanced mammary cancer in female dogs and the increase in serum CA15.3 levels. We investigated the potential prognostic value of these biomarkers; however, additional researches are needed to confirm these possibilities. Measurement of this tumor marker could be of value to the veterinary clinic for monitoring disease progression and the response to chemotherapy.

Acknowledgments

This study was supported, in part, by Fundação de Amparo a Pesquisa de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

Conflict of Interest: Authors disclose no conflict of interest.

Footnotes

  1. 1

    Vacutainer, BD Company, Franklin Lakes, NJ

  2. 2

    Gold Analisa Diagnóstica Ltda, Belo Horizonte, MG, Brazil

  3. 3

    Labtest Diagnóstica SA, Lagoa Santa, MG, Brazil

  4. 4

    Fujirebio Diagnostic Inc, PA

  5. 5

    GE Healthcare, Piscataway, NJ

  6. 6

    GraphPad Prisms v. 5.0; GraphPad, San Diego, CA

Ancillary