Abstract
Thyroid tumors occur in many domestic species, but are most common in the dog, in which they are classified as follicular or medullary. During 2012–2016, we received tissue specimens or whole carcasses of 4 dogs with variable enlargement of the thyroid glands. The 2 males and 2 females were of mixed (mongrel) inbreeding, 3–4.5-y-old. All tumors had lobulated architecture forming follicular structures variably containing colloid. On immunohistochemistry of the tumors from 3 of the dogs, 2 were thyroglobulin positive, and all 3 were negative for calcitonin, confirming follicular thyroid carcinoma in 2 of the dogs. Thyroid carcinomas have not been reported previously in related mongrel dogs, to our knowledge.
Tumors of the thyroid gland have been reported in domestic species, including the dog, cat, horse, ferret, rodent, and bull; other species are reported infrequently.6,8,12–14,16 These neoplasms account for 1–4% of canine tumors,6,8,14,17 collectively representing 10–15% of head and neck neoplasms in dogs.1,14,17 Thyroid neoplasms are mainly epithelial in origin and are classified as adenomas or carcinomas. Thyroid carcinomas account for up to 90% of thyroid tumors in dogs.14,17 Carcinomas are subdivided into follicular cell thyroid carcinomas (FTCs) or medullary thyroid carcinomas (MTCs), which originate from the follicular thyroid cells or the parafollicular C cells, respectively. FTCs, which can produce thyroglobulin, are the tumor type identified most commonly in dogs.2,4,5 MTCs have a reported prevalence of 36% of canine thyroid carcinomas. Approximately 50–60% of dogs with thyroid neoplasia are euthyroid, 30–40% are hypothyroid, and 10–20% are hyperthyroid. 15
Dogs 10–15-y-old are susceptible to developing thyroid tumors. 14 Purebred Boxers, Beagles, Siberian Huskies, and Golden Retrievers have a greater propensity for developing thyroid carcinoma than other breeds.13,14,16,17 A sex predisposition has not been established in dogs, but in humans, females are affected more often than males. Thyroid tumors can be familial or arise sporadically. In humans, 5–15% of thyroid tumors may be familial. In dogs, thyroid carcinomas are rare, and familial origin has not been studied extensively. 16
Confirmation of a histologic diagnosis of thyroid tumors is aided by the use of antibody markers targeting thyroglobulin and calcitonin. 3 FTCs produce thyroglobulin, which is the best marker for normal or neoplastic follicular cells given that 90–100% of thyroid carcinomas demonstrate positive staining. MTCs secrete calcitonin, which is found in 70–100% of MTCs in dogs.4,5,7,10–13 Immunoreactivity with thyroglobulin and little-to-no immunoreactivity with calcitonin (suspected to be entrapped C cells) has been considered a standard way to differentiate FTCs in veterinary medicine.1,5,7,11
During 2012–2016, 4 mixed-breed dogs were presented to a private clinician, and later to the University of West Indies School of Veterinary Medicine (SVM-UWI; Champs Fleurs, Trinidad and Tobago) pathology unit. The dogs were all the result of father–daughter pairings in a pack of 9 domestic mixed-breed dogs. Dog 1 was a 4.5-y-old male with a 15 × 9-cm mass on the ventral aspect of the neck that had progressively increased in size over a few years. The dog became anorectic and seizured, was euthanized due to its deteriorating condition, and was submitted to the SVM-UWI for autopsy. Dog 2, a 3-y-old male, and dog 3, a 3-y-old female, both developed a swelling in the ventral cervical area, measuring ~30 × 40 mm in dog 2 and 35 × 45 mm in dog 3. Dog 4 was a 4-y-old female with bilaterally swollen thyroid glands and a worsening condition. Due to concerns with likely metastases, dogs 2–4 were euthanized a few months after dog 1, and were sent for autopsy. Blood samples were taken for a CBC and serum biochemistry from dogs 2–4 before euthanasia.
The whole carcasses or thyroid tissue specimens of each animal were submitted for postmortem examination and microscopic analyses. Tissue samples were fixed in 10% neutral-buffered formalin, processed routinely, and 4-µm sections stained with H&E. Immunohistochemistry (IHC) was performed using anti-thyroglobulin and anti-calcitonin at the Pathology Diagnostic Laboratory, College of Veterinary Medicine, University of Georgia (Athens, GA, USA) on samples from dogs 1, 3, and 4; tissue blocks from dog 2 had been damaged and were not available for staining (Table 1).
Details of immunohistochemical stains performed on sections of thyroid tumors in a family of mongrel dogs.
Autostainer: Sakura Prisma; chromogen: DAB.
Dog 1
Oral and conjunctival mucous membranes were pale. The thyroid glands were markedly enlarged; the right lobe measured ~150 × 70 mm, and the left lobe 70 × 30 mm. Both lobes were cream-to-pink and had raised, 1–5-mm, cream-to-brown nodules scattered throughout the glands. On cut surface, there were numerous hemorrhagic areas. Throughout and occupying ~10% of the lungs, there were multiple, raised, 1–5-mm, tan nodules. No gross abnormalities were detected in other body systems. Histologically, thyroid architecture was effaced and replaced with a lobulated mass with moderate interlobular and fine intralobular fibrous septa. Neoplastic cells were arranged in irregularly formed tubular structures lined by 1–4 layers of columnar-to-cuboidal cells with round slightly hyperchromatic nuclei, no discernable nucleoli, and moderate amounts of eosinophilic cytoplasm. The mitotic count was 2 mitotic figures in 2.37 mm2.
Dog 2
The thyroid glands were bilaterally enlarged, with the left and right lobes measuring 43 × 25 mm and 34 × 21 mm, respectively. No lesions were detected in the lungs. The neoplasm had a lobulated architecture with neoplastic cuboidal epithelium predominantly forming follicular structures of various sizes and shapes, most often filled with deeply eosinophilic and homogeneous deposits. Small solid areas were scattered throughout the section. The neoplastic cells had moderate eosinophilic cytoplasm with deeply basophilic and central nuclei. Some follicles were lined by multiple layers of cuboidal cells. There was little anisocytosis, anisokaryosis, or mitoses. There was a significant vascular component to the section with corresponding hemorrhage. Invasion of the fibrous connective tissue capsule could not be confirmed given that only 1 small area of capsule was available for inspection. Although the relatively small section of tissue presented for histologic examination did not show distinct malignant characteristics, the mass was palpable on physical examination and bilateral. The small section may have been taken from a portion of the tumor without significant malignant characteristics. Most thyroid tumors are malignant, but sections of malignant tumors may not always display overt malignant histologic characteristics.9,14
Dog 3
The thyroid glands were enlarged, with the left and right lobes measuring 51 × 33 mm and 42 × 27 mm, respectively. No lesions were detected in the lungs. The mass had a lobulated architecture composed of variably sized and shaped follicular structures lined by 1–3 layers of neoplastic cuboidal epithelial cells and variably filled with deeply eosinophilic and homogeneous material. These cells had moderate-to-marked anisocytosis and anisokaryosis, and moderately eosinophilic cytoplasm with deeply basophilic and central nuclei. Tumor cells invaded the overlying connective tissue capsule. The tumor was well-vascularized and hemorrhagic. No mitotic figures were noted in 2.37 mm2.
Dog 4
The buccal and conjunctival mucous membranes were moderately pale. Bilaterally, the thyroid glands were markedly enlarged, with the right and left lobes measuring 67 × 48 mm and 76 × 51 mm, respectively. The glands were lobular, pink-to-tan, firm, and on cut surface there was a gritty feel, and small areas of bone and necrosis were present throughout. The lymph nodes cranial to the masses were bilaterally enlarged (22 × 38 mm and 27 × 42 mm). Moderately circumscribed, coalescing, raised, tan-to-pink, 4–45-mm nodules were scattered throughout the lungs occupying >80% of the pulmonary parenchyma (Fig. 1A). A 24-mm diameter moderately circumscribed mass of similar characteristics was present at the cranial pole of the right kidney. The lateral ventricles of the brain were expanded (15 × 14 mm on cross-section); however, no other gross changes were detectable in the brain. The thyroid glands were both effaced by variably lobulated masses with light-to-marked fibrous interlobular stroma. The neoplastic cells formed variably sized and shaped follicular structures and nests with one-to-multiple layers of cuboidal epithelial cells with moderate anisocytosis and anisokaryosis, indistinct borders, and moderate amounts of eosinophilic and regularly vesicular cytoplasm (Fig. 1B). Nuclei were centrally located and rounded with heterochromatin. Mitotic count was 4 mitotic figures in 2.37 mm2. Islands of bone were scattered throughout the stroma. There were also small areas of necrosis and hemorrhage. Pulmonary and renal metastases had similar histologic characteristics except that the density of follicular structures was greater with less stroma, and no bone was seen in the sections examined. There was moderate-to-marked hepatic congestion. In the brain, there was mild-to-moderate vacuolation with a mild microglial infiltrate in the white matter adjacent to the lateral ventricles.
Gross and microscopic findings in a mongrel dog with a thyroid tumor.
Immunohistochemically, the mass from dog 1 was negative for thyroglobulin. In contrast, the mass from dog 3 had mild-to-moderate multifocal cytoplasmic staining for thyroglobulin; the mass from dog 4 had strong multifocal cytoplasmic staining for thyroglobulin (Fig. 1C). All masses submitted for IHC were negative for calcitonin (Fig. 1D).
The gross and histologic findings in this family of inbred mongrel dogs indicate the presence of FTC, confirmed by the positive thyroglobulin and negative calcitonin IHC in 2 of the dogs. The tumors in all of the dogs had lobulated architecture, with definite follicular structures in dogs 2–4, which contained colloidal substance. In dog 1, there were follicular structures; however, colloid was not seen. In dog 4, there was positive staining for thyroglobulin but negative staining for calcitonin; the histologic architecture in H&E-stained sections was typical of an FTC (Fig. 1B). As the sensitivity range of thyroglobulin immunolabeling is 90–100% for FTCs, it is possible that up to 10% of these tumors label negatively for this marker. This may explain the negative immunolabeling observed in dog 1. Mitotic counts were low, with a highest of 4 in 2.37 mm2 in dog 1, and none in dog 3.
We retrieved no cases of thyroid tumors in inbred mongrel dogs in a search of Google, PubMed, CAB Direct, Web of Science, and Scopus, using search terms “thyroid neoplasia in mongrel dogs”, suggesting that this condition has not been reported in this type of dog. Familial thyroid carcinomas are also uncommon in humans,16,18 and they are primarily medullary (parafollicular C cell) in origin. 8 They also tend to be bilateral, similar to our 4 cases. 11 There have only been 2 reports of familial thyroid carcinomas in dogs: MTCs in mixed-breed dogs with Alaskan Malamutes having the major breed influence, 8 and FTCs in a large group of German longhaired pointers (GLPs) in the Netherlands. In the latter report, a significantly higher pedigree-based inbreeding coefficient was found in affected GLPs, and therefore inbreeding contributed to development of the FTCs in this group of dogs. 18 Similarly, the occurrence of these tumors in the 4 inbred dogs in our report suggests a genetic component, although we were unable to investigate this possibility. Given that our dogs were all from the same location, an environmental contribution cannot be ruled out, although none of the resident humans at the location has developed thyroid disease to date.
Footnotes
Acknowledgements
The authors would like to acknowledge the help of the histopathology technician, Mr. Gerald Chandoo.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
Our research was funded by the internal institutional grant CRP.3.MAR16.50, which was used to fund ancillary testing in this work and was received from UWI.
