Neoplasia in pet guinea pigs: a retrospective analysis of 2,474 autopsy examinations

Abstract

Neoplasia is a common disease in guinea pigs (Cavia porcellus); however, few studies have evaluated the prevalence of neoplasia in all organ systems. We retrospectively analyzed the tumor prevalence in pet guinea pigs and the frequency of metastasis in a multi-institutional study population of 2,474 autopsy cases. Tumors were found in 508 guinea pigs (prevalence: 20.5%), of which 95 cases had >1 tumor, resulting in a total of 627 tumors. The tumor prevalence increased from 1.4% in animals <0.5-y-old to 53.6% for guinea pigs >5-y-old. The most common tumor type was lymphoma or leukemia, affecting 174 guinea pigs (tumor prevalence: 7.0%). Lymphomas or leukemias were disseminated to various organs and/or lymph nodes in 146 (83.9%) cases and localized to 1 organ or 1 lymph node in 28 (16.1%) cases. Primary non-lymphoid tumors were most frequent in the female genital tract (62 of 1,235 cases, mostly uterus), respiratory system (116 of 2,474), skin including mammary gland (81 of 2,474), endocrine system (66 of 2,474, mostly thyroid gland), and alimentary tract (35 of 2,474). Tumors of the alimentary tract were dominated by gastrointestinal stromal tumors. Metastasis was detected in 42 of 453 non-lymphoid tumors (9.3%), with a surprisingly low frequency for pulmonary carcinoma and splenic hemangiosarcoma compared to other species. Our postmortem study demonstrates a high prevalence of disseminated lymphoma or leukemia in pet guinea pigs at the time of death or euthanasia. Additional studies are needed to further characterize these tumors.

Guinea pigs (Cavia porcellus) are one of the most common small mammal patients in veterinary practices and clinics. ²³ Neoplasia is considered one of the most common diseases in geriatric guinea pigs; ³⁰ however, few studies have evaluated tumor prevalence of all organ systems, or described demographic patient information and biological behavior.

Although the tumor diagnosis for prevalence studies should be based on histologic examination, the determined tumor prevalence is greatly influenced by the type of samples submitted for examination (i.e., autopsy or surgical biopsy). Full autopsies have the obvious advantage of examining all organs of the animal, thus each organ is equally represented, and the presence of metastasis can be evaluated. To date, only one autopsy study has evaluated a large population of pet guinea pigs (written in German), revealing a tumor prevalence of 14.9%. ³⁴ However, given the lack of exclusive focus on tumors, some information (such as frequency of metastasis) was not evaluated. Most studies describe the results of histologic examinations from surgical biopsies (some included a smaller or sometimes unknown proportion of autopsies), reporting a tumor frequency of 44.9–77.8% among all submissions.^{2,10,14,27,30,32} Studies including surgical biopsies seem to overrepresent certain organs (particularly skin) and certain diseases (particularly tumors),^{2,10,14,16,27,30,32} reflecting the surgical procedures conducted at veterinary clinics and the need for histologic diagnosis only of tissue masses detected clinically. We suggest that studies with autopsies are less biased than studies with organ biopsies for determining the prevalence of tumors and metastasis or dissemination. However, we acknowledge that findings from autopsies may somewhat underrepresent diseases that are easily diagnosed antemortem, such as skin tumors.

Using a large study cohort of autopsy cases, our objective was to determine the frequency of different tumor types in pet guinea pigs grouped by organ system, the predisposition of age and sex to developing tumors, and the potential of these tumors to undergo metastasis.

Material and methods

Case selection and data extraction

Our study methods were based on a previous study. ³ Autopsy reports from pet guinea pigs were retrospectively collected from 3 veterinary pathology institutes: 1) Freie Universität Berlin (FUB, Berlin, Germany; study period: 1995.01.01–2023.10.31), 2) University of Veterinary Medicine (UVM, Vienna, Austria; study period: 1996.01.01–2023.10.31), and 3) The Schwarzman Animal Medical Center (AMC, New York, NY, USA; study period: 1993.01.01–2023.10.31). We excluded autopsies performed on laboratory guinea pigs from our study. The data of the FUB and VMU had been described as part of a doctoral thesis or diploma thesis, respectively, with a slightly shorter study period.^2,32 Some cases from FUB (female genital tract tumors ⁵ and a cutaneous squamous cell carcinoma [SCC] ⁴ ) were published in journal articles.

Data extracted from the original autopsy reports included the age at death, sex, presence of neoplasia, organ affected by neoplasia (separated into primary organ and metastasis for non–round-cell tumors), and tumor type. Breed of the guinea pigs was inconsistently reported on the submission form and therefore not collected for our study. We excluded tissue masses solely described grossly (i.e., without a histologic examination). Diagnoses were taken from the autopsy report (i.e., re-evaluation of the histologic slides was not possible). The tumor diagnosis was based on histologic examination in all cases and, in selected cases, supported by special stains or, when requested by the submitters, immunohistochemistry (IHC). Masses of the same tumor type in the same organ system were counted as one.

Data analysis

Tumors were grouped by organ system (Table 1). For lymphoma or leukemia, a primary organ often cannot be identified and thus these tumors were considered as a separate “organ category”. Instead of a primary organ and metastatic foci, any organ affected by lymphoma or leukemia was recorded. A disseminated form of lymphoma or leukemia was diagnosed when ≥2 organs or ≥2 lymph nodes were affected. Lymphoma or leukemia was classified as localized when only one organ or one lymph node was infiltrated by the neoplastic cells. Some tumor types were combined to larger groups, particularly when IHC would have been required for an accurate classification (such as thyroid follicular vs. C-cell tumors) or when relevant inter-observer variability can be expected (such as for non-angiomatous soft-tissue sarcoma).

Table 1.

Prevalence of primary tumors of the different organ systems in 2,474 pet guinea pigs.

Organ system	No. of cases with tumors	Tumor prevalence, %*	No. of tumors	No. of cases (%) with metastasis or multi-organ dissemination
Lymphoma or leukemia	174	7.0	174	146 (83.9)
Female genital tract	62	5.0	67	7 (10.4)
Respiratory	116	4.7	116	2 (1.7)
Skin (including mammary glands)	81	3.3	87	14 (16.1)
Endocrine	66	2.7	69	3 (4.3)
Alimentary	35	1.4	35	4 (11.4)
Hematopoietic (non-lymphoid tumors)	17	0.7	17	0
Musculoskeletal	15	0.6	15	4 (26.7)
Hepatobiliary and exocrine pancreas	13	0.5	13	0
Body cavities	13	0.5	13	1 (7.7)
Urinary	10	0.4	10	1 (10.0)
Cardiovascular	5	0.2	5	0
Sensory organs	2	0.1	2	0
Male genital tract	1	0.1	1	0
Nervous	0	0	0	0
Unknown primary origin	6	0.2	6	6 (100)
All organ systems (non-lymphoid tumors)	367	14.8	453	42 (9.3)
All organ systems (all tumors)	508	20.5	627	187 (29.8)

^*Tumor prevalence is calculated for the entire study population of 2,474 guinea pigs, except for tumors of the female and male genital tract, which is based on the number of intact animals of this respective sex (1, 235 intact females; 860 intact males).

For statistical analysis, tumors were categorized as the following: 1) lymphoma or leukemia versus 2) non-lymphoid tumors, which were further subcategorized into 2A) benign histologic tumor types, and 2B) malignant histologic tumor types. Tumor names, which do not indicate malignancy (such as adrenal pheochromocytoma, pancreatic islet cell tumor, or ovarian granulosa cell tumor) were classified as benign unless otherwise specified in the autopsy report (i.e., if invasion into surrounding tissue or distant metastasis was reported).

Tumor prevalence was defined as the percentage of affected guinea pigs among all cases of the study population or the subgroup evaluated. For calculation of age-dependent tumor prevalence, age categories of 6-mo intervals were created and, if <200 animals, combined into larger age groups.

Results

Study population

At the 3 institutes, 2,474 autopsies of pet guinea pigs were documented during the evaluated time periods (1,428 cases from FUB, 909 cases from UVM, 137 cases from AMC). For this study population, age was available for 2,010 cases (81.2%; Suppl. Fig. 1) and sex was available for 2,347 cases (94.9%). Most guinea pigs were intact females or males (89.3%); spayed female animals were on average older than the other sex categories (Suppl. Table 1).

Tumor prevalence

Tumors were found in 508 of the 2,474 cases, thus the entire study population had a tumor prevalence of 20.5%. The tumor prevalence per laboratory was 291 of 1,428 (20.4%) for FUB, 172 of 909 (18.9%) for UVM, and 45 of 137 (32.8%) for AMC. Of the 508 guinea pigs with tumors, 95 cases had multiple tumors (2 tumors in 79 cases; 3 tumors in 11 cases; 4 tumors in 3 cases; 5 and 6 tumors in 1 case each) for a total of 627 tumors. We grouped tumors into lymphoma or leukemia (n = 174, see separate section below) and non-lymphoid tumors (n = 453 tumors in 367 guinea pigs). Lymphoma or leukemia (Fig. 1A) was the most common tumor in both male and female pet guinea pigs; non-lymphoid tumors arose most commonly in the uterus (Fig. 1B), lungs (Fig. 1C), skin including the mammary gland (Fig. 1D), thyroid gland (Fig. 1E), and gastrointestinal (GI) tract (Fig. 1F; Table 1). The 453 non-lymphoid tumors were classified as malignant tumor types in 223 (49.2%) instances, affecting 204 guinea pigs. Metastasis of malignant non-lymphoid tumors was reported in 42 of 223 (18.8%) tumors affecting 41 guinea pigs; thus, only 1.7% of the entire study population had metastatic non-lymphoid tumors.

Figure 1.

Common tumors in guinea pigs. A. Disseminated lymphoma in the spleen. B. Endometrial adenoma. C. Pulmonary adenoma in the left cranial lung lobe (arrow). D. Subcutaneous lipoma. E. Thyroid follicular adenoma (arrow). F. Gastrointestinal stromal tumor in the jejunum.

Guinea pigs with tumors had a mean age of 55.8 mo at death compared to 30.5 mo for guinea pigs without tumors (Suppl. Table 2). Tumor prevalence increased from 1.4% in animals <0.5-y-old to 53.6% for guinea pigs >5-y-old (Fig. 2; Suppl. Table 3). Tumors were found more commonly in female intact guinea pigs, representing 57.5% of the affected animals (compared to 52.6% of the entire study population), which can be explained by the higher presence of genital tract neoplasms in females compared to males (Table 1; Suppl. Table 4). Spayed female animals represented 1.8% of all tumor cases (compared to 1.1% of the entire study population), which is probably associated with the higher age distribution of this subgroup.

Figure 2.

Tumor prevalence in 2,010 pet guinea pigs comparing different age groups, with 213–330 cases within each age group. Note that latter age groups encompass larger periods to account for unequal age distribution of the study population. Note that some guinea pigs had concurrent non-lymphoid tumors and lymphoma or leukemia.

Lymphomas or leukemias

Lymphomas or leukemias were detected in 174 guinea pigs (tumor prevalence: 7.0%), thus being the most common tumor type in guinea pigs, particularly in younger animals (Fig. 2). The mean age of affected animals (50.8 mo; Suppl. Table 2) was lower than for non-lymphoid tumors (59.9 mo). There was no apparent predisposition of either sex to developing lymphoma compared to the distribution of the study population (Suppl. Table 4).

Lymphomas or leukemias were localized to 1 organ or 1 lymph node in 28 (16.1%) cases and disseminated to multiple organs and/or multiple lymph nodes in 146 (83.9%) cases. The organs most frequently affected (Fig. 3) were the lymph nodes (Fig. 4A), liver (Fig. 4B), and spleen (Fig. 1A; Suppl. Table 5). Of the 132 guinea pigs with lymphoma in the lymph nodes, specific locations were reported in 109 animals as the following: peripheral lymph nodes in 97 cases (89%, particularly common in the cervical lymph nodes; Suppl. Figs. 2, 3), abdominal lymph nodes in 35 cases (32%), and thoracic lymph nodes in 17 cases (16%).

Figure 3.

Frequency of organ involvement by disseminated and localized lymphomas or leukemias in pet guinea pigs. Organs involved <7 times were excluded from this graph and listed in Suppl. Table 4.

Figure 4.

Histologic images of common tumors in guinea pigs. H&E. A. Lymphoma in a lymph node. B. Lymphoma in the liver with fatty degeneration. C. Pulmonary adenoma. D. Higher magnification of the pulmonary adenoma. E. Papillary thyroid adenoma. F. Higher magnification of the thyroid adenoma.

Tumors of the female genital tract

Tumors of the female genital tract were found in 62 guinea pigs (67 tumors; tumor prevalence for female intact animals: 5.0%). Uterine tumors were reported in 52 guinea pigs (54 tumors) including 23 leiomyomas (42.6%), 10 adenomas (18.5%; Fig. 1B), 10 adenocarcinomas (18.5%), 9 leiomyosarcomas (16.7%), 1 malignant mixed Müllerian tumor (1.9%), and 1 hemangioma (1.9%; Suppl. Tables 2–4). Metastasis was reported for 5 of the 10 endometrial adenocarcinomas (50%) and 2 of the 9 uterine leiomyosarcomas (22.2%).

In the ovaries of 10 guinea pigs were 6 rete-ovarii adenomas, 3 ovarian surface adenomas, 1 granulosa cell tumor, and 1 leiomyoma. A uterine tube adenoma and a vaginal leiomyoma were diagnosed in one guinea pig each.

Tumors of the respiratory system

Pulmonary tumors were found in 116 guinea pigs (tumor prevalence: 4.7%): 59 adenocarcinomas and 57 adenomas (Figs. 1C, 4C, 4D; Suppl. Tables 2–4). Only 2 of the 59 adenocarcinomas (3.4%) developed metastases. Tumors of the upper respiratory tract were not documented.

Tumors of the skin and mammary gland

Tumors of the skin (including the subcutis, pinna, external ear, and mammary gland) were reported in 81 cases (tumor prevalence: 3.3%); 6 cases had multiple skin tumors. The tumor types or groups diagnosed included 28 lipomas (32.2%; Fig. 1D), mammary adenocarcinomas (19.5%), 15 tumors with hair follicle differentiation (17.2%), 12 soft-tissue sarcomas (13.8%), 5 mammary adenomas (5.7%), 2 liposarcomas (2.3%), 2 SCCs (2.3%), 2 sebaceous adenomas (2.3%), and 1 case each of an undifferentiated carcinoma, apocrine adenocarcinoma, hemangiosarcoma, and osteosarcoma. Metastasis was found in 7 of 17 mammary adenocarcinomas, 3 of 12 soft-tissue sarcomas, 1 of 2 SCCs, and in the undifferentiated carcinoma, apocrine adenocarcinoma, and cutaneous osteosarcoma. Mammary tumors were common in intact male guinea pigs with a frequency of 31.8% of all mammary tumors (Suppl. Table 4).

Tumors of endocrine system

Tumors of endocrine glands were found in 66 guinea pigs (tumor prevalence: 2.7%); 3 cases had 2 endocrine tumors. Most commonly, the thyroid glands were affected in 51 guinea pigs (77.3%; tumor prevalence: 2.1%), and the diagnosed tumor types included thyroid adenoma in 40 cases (Figs. 1D, 4E, 4F) and thyroid carcinoma in 13 cases. One thyroid carcinoma had metastasized to distant organs.

The 16 tumors of the other endocrine organs included 8 pancreatic islet cell tumors, 3 adrenocortical carcinomas, 1 adrenocortical adenoma, 2 pheochromocytomas, 1 parathyroid adenoma, and 1 pituitary adenoma. One islet cell tumor and one adrenocortical carcinoma developed distant metastases.

Tumors of the alimentary system

Tumors of the alimentary system were reported for 35 guinea pigs (tumor prevalence: 1.4%), most commonly affecting the GI tract, and rarely the salivary glands or gingiva. Twenty-seven tumors were non-angiomatous, non-lymphomatous mesenchymal tumors (77.1%) of which 17 were subclassified by IHC (using antibodies for KIT/CD117 and smooth muscle actin) into 13 GI stromal tumors (GISTs; Fig. 1F), 2 leiomyomas, and 2 leiomyosarcomas, according to the original autopsy report. Those mesenchymal tumors occurred at various sites within the GI tract (11 in the stomach; 16 in the small or large intestine). Additional tumors of the intestine included an adenoma and an adenocarcinoma. Three of the 27 mesenchymal tumors (11.1%) and the intestinal adenocarcinoma developed distant metastases.

There were 2 salivary gland adenocarcinomas and 2 salivary gland adenomas. In the oral cavity, one ameloblastoma and one fibrosarcoma were diagnosed. None of these cases had metastases.

Non-lymphoid tumors of the hematopoietic system

The spleen had primary tumors in 17 guinea pigs (tumor prevalence: 0.7%), including 10 hemangiosarcomas, 4 myelolipomas, 2 soft tissue sarcomas, and 1 hemangioma. None of the tumors developed metastases. Non-lymphoid tumors were not reported in the other hematopoietic organs (thymus, lymph node, bone marrow).

Tumors of the musculoskeletal system

Tumors of the bone were diagnosed in 14 cases (prevalence: 0.6%) and in the skeletal muscle of 1 case. In the bone, 10 cases were diagnosed as osteosarcoma, 2 cases as fibrosarcoma, 1 case as chondrosarcoma, and 1 case as ossifying fibroma. The bone affected included the mandible (3), ribs (3), femur (2), humerus (2), maxilla (1), sacrum (1), and tympanic bulla (1); location was unknown in 1 case. The skeletal muscle tumor was interpreted as a fibrosarcoma. Metastasis was reported in 3 of 10 skeletal osteosarcomas and the skeletal muscle fibrosarcoma.

Tumors of the hepatobiliary system and exocrine pancreas

Tumors of the hepatobiliary system were found in 7 guinea pigs (prevalence: 0.3%). Most common were cholangiocellular tumors (3 adenomas, 1 adenocarcinoma), followed by hepatocellular tumors (1 adenoma, 1 adenocarcinoma), and 1 fibrosarcoma. These tumors did not develop metastases.

The exocrine pancreas of 5 guinea pigs had tumors (prevalence: 0.2%), including 2 adenomas and 3 adenocarcinomas, none of which metastasized.

Tumors of the body cavities

Within the abdominal cavity (i.e., not associated with a parenchymal organ), 13 guinea pigs had tumors (prevalence: 0.5%) including 5 lipomas, 3 liposarcomas, 2 soft-tissue sarcomas, and 1 case each of leiomyoma, hemangiosarcoma, and undifferentiated metastatic tumor.

Tumors of the urinary system

The urinary system of 10 guinea pigs had tumors (prevalence: 0.4%), most of which occurred in the urinary bladder. The tumor types included 5 cases of urothelial cell carcinoma of the urinary bladder, and 1 case each of a SCC of the urinary bladder, leiomyoma of the urinary bladder, soft-tissue sarcoma of the urinary bladder, sarcoma of the kidney, and undifferentiated tumor of the kidney. Metastasis was present in the urinary bladder SCC and renal sarcoma.

Tumors of the cardiovascular system

Five cases were diagnosed with primary non-angiomatous spindle-cell tumors of the heart (tumor prevalence: 0.2%); no IHC was done to clearly characterize the tumor type. Metastasis was not reported for the cardiac tumors. Tumors of large vessels were not reported.

Tumors of the sensory organs

In the orbit (eye and ocular adnexa), a lacrimal gland adenoma and an undifferentiated tumor was reported in one guinea pig each. Tumors of the other sensory organs were not reported.

Tumors of the male genital tract

One guinea pig was diagnosed with a prostatic adenocarcinoma without metastasis. There were no tumors of the remaining portions of the male genital tract.

Tumors of unknown primary origin

Tumors with unknown/uncertain primary origin included 3 metastatic sarcomas, 2 metastatic adenocarcinomas, and 1 abdominal metastatic teratoma.

Discussion

We evaluated the tumor prevalence in all organ systems of pet guinea pigs autopsied at 3 institutions. Tumor prevalence has traditionally been considered low based upon studies of laboratory guinea pigs, ¹³ which are typically younger, with different environments than pet guinea pigs. However, we have found a tumor in every fifth pet guinea pigs autopsied. Tumor prevalence was higher for pet guinea pigs than that of pet rabbits when comparing data from the same institution (tumor prevalence of 20.4% for pet guinea pigs and 14.5% for pet rabbits at FUB). ³ The one comparable study for pet guinea pigs did not focus exclusively on tumors and did not include information such as the frequency of metastasis; ³⁴ the tumor prevalence was 14.9% compared to 20.5% in our study (average of all 3 institutions). The authors of the 2004 study ³⁴ did not clearly state the age distribution of the study population; they mention that animals >4-y-old had a tumor prevalence of 39%, which is also somewhat lower than the 48.1% prevalence in our study for the same age group. The most commonly affected organ systems (disseminated lymphoma or leukemia, lungs, thyroid gland, and skin including subcutis and mammary gland) were similar between these 2 studies, with the exception that we found tumors of the female genital tract to be more common. Organ systems with particularly low tumor prevalence in guinea pigs included the nervous system, male genital tract, and sensory organs (such as eyes), which is supported by the low number of these tumor types described in the literature.^12,19,30,34

It is important to note that most of the previous studies evaluated tumor frequency from predominantly surgical biopsies, for which skin tumors are by far the most common. Accordingly, reports of tumors from visceral organs that are more difficult to surgically access (e.g., lung tumors) are extremely rare.^{7,10,14,16,27,30} This discrepancy can be explained by several factors: 1) tumors of the skin are more easily detected on clinical examination, and 2) skin masses can be more routinely removed surgically compared to some internal organ masses. While studies involving surgical biopsies deal with the tumor types of interest for clinicians, autopsies more closely reflect the actual tumor prevalence of all organ systems (including the internal organs) at the time of death or euthanasia. These tumors may or may not be of importance to clinicians, as we included all tumors regardless of their suspected clinical relevance (main finding vs. incidental finding). It is possible that some tumor types, for which an antemortem diagnosis is easily obtained (particularly skin tumors), are underrepresented in autopsy cases.

We did not find any previous study of a systematic exploration of the frequency of metastasis in guinea pigs in a literature search in Google Scholar and PubMed using search terms such as “guinea pig” AND “tumor” OR “neoplasia” OR “metastasis”. Considering that all guinea pigs in our study died or were euthanized due to their primary disease and that full autopsy evaluation has a high diagnostic sensitivity for detecting metastasis, the proportion of metastasis among all non-lymphoid tumors seems low (42 of 453; 9.3%). In a comparable study on autopsied pet rabbits, metastasis occurred in 74 of 229 non-lymphoid tumors (32.3%). ³ Of note, metastasis was infrequent or absent in several tumor types, which are known to be aggressive in other animal species, including pulmonary carcinoma (compared to dogs ²⁵ and cats ³³ ) and splenic hemangiosarcoma (compared to dogs ³¹ ). The presumably low risk for metastasis of splenic hemangiosarcoma is supported by the literature (suspected metastasis in 1 of 6 cases)^29,37; supporting literature is lacking for pulmonary tumors. Compared to dogs and cats with mostly malignant pulmonary tumors, pulmonary adenoma was more commonly diagnosed in guinea pigs. Histologic criteria for pulmonary adenoma have not been reported or validated for guinea pigs (or any other animal species); however, general characteristics include a well-demarcated, expansile nodule with minimal features of malignancy. Although data regarding the long-term behavior of these tumors are lacking in many species, our series confirms that metastasis did not occur in 57 pulmonary adenomas in guinea pigs. Interestingly, of 59 pulmonary adenocarcinomas (with histologic signs of malignancy in the primary tumor), only 2 (3.4%) developed metastases. Thus, development of a consensus for histologic criteria of pulmonary adenoma and adenocarcinoma of guinea pigs correlated with biological behavior would be useful to pathologists and clinicians.

Further descriptive studies of potentially aggressive tumors in guinea pigs are needed to evaluate prognostic markers and effectiveness of different treatment procedures. Studies with more detailed evaluation of the histologic and immunohistochemical tumor characteristics associated with clinical follow-up and potentially other clinical data are needed for further information about the clinical impact of these neoplasms. Lymphoma or leukemia in guinea pigs is of particular interest for future studies due to the high prevalence and the aggressive biological behavior with dissemination to several organs in 83.9% of our cases.

The main limitation of our study was that the tumor diagnoses were obtained from the original autopsy reports because reevaluation of all tumors was not possible. It should be noted that the diagnostic criteria for classification of tumor types (such as pulmonary adenoma vs. adenocarcinoma) were likely inconsistent throughout the long study period and between different pathologists. Due to the inability to subclassify some tumor types without IHC and potential inter-observer variability between pathologists involved in autopsy service at the authors’ institutes, we combined some of the tumor types into larger groups. Previous case reports or case series of neoplasms in guinea pigs have provided details regarding use of IHC to assist with tumor classification, such as for B-cell versus T-cell lymphoma, ⁸ GISTs versus other GI mesenchymal tumors, ³⁸ cardiac tumors, ⁴² and thyroid follicular versus C-cell tumors. ¹⁰ Another limitation was that the autopsy submission form and/or postmortem report was not always complete, thus demographic information for some patients was not available. This led to the decision to exclude breed as a risk factor for tumor development. Interpretation of tumor prevalence should be done in conjunction with the age distribution of the study population, which is why we calculated the prevalence for different age groups. While all cases were examined by a “full autopsy” following the protocols of each institution at the respective time of the study period, the CNS was not examined in some cases. However, this is not expected to have a significant impact on the tumor prevalence of the CNS given that autopsy procedures are adapted based on clinical findings (e.g., nervous signs). Inclusion of clinical parameters possibly associated with the tumors (i.e., endocrinopathies of endocrine tumors) was beyond the scope of our study.

In the literature, there are several large case series with detailed information about tumors in specific organ systems, including the skin (including subcutis and mammary gland),^7,18,28,36 female genital tract,^5,22,41 and thyroid tumors. ¹⁰ In the following paragraphs we will discuss some selected tumor types of particular importance in guinea pigs.

Lymphoma or leukemia was the most common tumor in our study and, in most cases, was the main autopsy finding based upon dissemination into several organs in 83.9% of our cases. In fact, this tumor was the most common neoplasm of any organ system. Considering the high prevalence of lymphoma or leukemia in our study, the paucity of current literature about this tumor type in guinea pigs is surprising. Although some studies included several cases of lymphoma, complete organ distribution and subtypes according to the WHO system ³⁹ were not described.^27,34 In a series of 31 cases of disseminated lymphoma, 30 were found in multiple lymph nodes; however, involvement of other organs was not described. ³⁴ In our study, we identified lymphoma or leukemia to be most common in lymph nodes, liver, spleen, lungs, and heart. It should be noted that most of our cases were late-stage lymphomas or leukemias (which resulted in death or euthanasia). In these cases, identification of the primary organ (i.e., where the tumors had initially arisen) was not possible, nor were we able to differentiate disseminated versus multicentric disease. Subclassification of lymphomas is available in a study using the REAL-System ³⁴ and in several case reports.^8,20,24,35 Based on these studies, B-cell lymphoma might be somewhat more common. It was beyond the scope of our study to conduct IHC, thus we are unable to explore lymphoma or leukemia subtypes. Further studies are needed in guinea pigs with more detailed descriptions of lymphoma (i.e., cell size, mitotic activity, immunophenotype, and subtypes using the WHO system ³⁹ ), potential viral etiology, and their clinical course.

Thyroid tumors are one of the most common tumors and by far the most common endocrine tumor in guinea pigs.^7,10,30,34 All 19 thyroid tumors in a previous study had a follicular origin, as determined by IHC. ¹⁰ For our study, IHC was performed only for a small subset of cases and was not available for reevaluation (thus was not systematically analyzed in the results section), revealing that few thyroid tumors in guinea pigs were interpreted to be of C-cell origin. Metastasis of thyroid tumors occurs rarely and was detected only in 1 of 66 (1.5%) of our cases and 1 of 19 published cases. ¹⁰ In addition to thyroid tumors, previous studies with large study populations have reported very few endocrine tumors of other types.^27,30,34 There are a few case reports on pancreatic islet cell tumors,^1,15,40 which were also occasionally seen in our study. Similar to other animal species, thyroid tumors have been associated with a clinical syndrome of hyperthyroidism and pancreatic islet cell tumors with hypoglycemia. ²¹

The most common tumor of the alimentary tract were GI non-angiomatous, non-lymphomatous mesenchymal tumors, which requires IHC to differentiate GISTs from other mesenchymal tumors (mostly smooth-muscle tumors). ¹¹ A subset of our cases was evaluated immunohistochemically using antibodies against KIT and smooth muscle actin. These results indicated that 13 of 17 tumors were GISTs. This finding is consistent with previous case reports and a small case series that described 6 of 7 GISTs and 1 of 7 smooth-muscle tumors in the GI tract of guinea pigs.^9,17,26,38 Another specific marker for GIST is DOG1 (discovered on GIST protein 1), which has been used in guinea pigs ³⁸ and is described to have a higher sensitivity than KIT in dogs. ⁶ The location of GISTs in the stomach seems to be more common in guinea pigs, compared to dogs.^{9,11,17,26,38} Metastasis occurred occasionally in 11% of our cases and was reported in 1 of 5 published cases.^17,38