Suspected renal interstitial cell tumor causing polycythemia in two dogs

Abstract

Here we report a case series of two dogs diagnosed as renal interstitial cell tumor (RICT) accompanied by elevated serum erythropoietin level and marked polycythemia. RICT is a rare tumor in dogs, originating from renal interstitial cells. While several renal tumors such as renal lymphoma, adenocarcinoma, carcinoma, sarcoma, fibrosarcoma and nephroblastoma may cause polycythemia, polycythemia caused by RICT has never been reported in dogs. The tumors in both dogs were solitary and lied within cortex or cortico-medullary junction. Histopathology revealed spindle-shaped cells suggesting mesenchymal origin, with no mitotic figures suggesting that the tumors in both dogs were benign. Following surgical removal of the affected kidney, serum erythropoietin level and polycythemia normalized in both dogs.

Renal interstitial cell tumor (RICT), which arises from renal interstitial cells is rare in dogs, and only a few cases are reported [1, 3, 17]. Diters and Wells (1986) retrospectively analyzed nine cases from the necropsy records, and the tumors were often bilateral, and all were multiple, without evidence of any mitotic figures. Cho et al. (2020) reported a malignant case in a 17-year-old spayed female Maltese dog with renal insufficiency and anemia. Rissi and Dill-Okubo (2020) reported a case with cardiac and cerebral metastasis. While RICT in human is typically benign, there seem to be both benign and malignant types in dogs from these previous reports. Histological features of RICT are similar to those of fibroma; however, demonstration of cytoplasmic lipid droplets by electron microscopy as well as presence of alcian-blue-positive matrix are requisite to diagnose RICT [13].

Polycythemia may be caused by renal neoplasm, and such cases are known in renal lymphomas [4, 8, 12, 18], renal adenocarcinoma [2], renal carcinoma [16], nephroblastoma [11], renal sarcoma [15], and renal fibrosarcoma [10] in dogs. To the author’s knowledge, polycythemia in combination with RICT have never been reported in dogs. Here we report two canine cases of marked polycythemia caused by RICT.

Case one, a 4-year-old, intact male Yorkshire Terrier weighting 4.74 kg visited a private animal hospital to receive a vaccination (day 1). Packed cell volume was found to be 65%, and polycythemia was suspected. The serum erythropoietin (EPO) concentration was 24.2 mIU/mL (reference range 1.3–13.4; FUJIFILM VET Systems Co., Ltd., Tokyo, Japan). A mass in right kidney was found by ultrasonography. For the detailed investigation and possible surgical treatment, the dog was referred to the Iwate University Veterinary Teaching Hospital on day 48. The results of blood chemistry and urinalysis were not affected. The ultrasonography revealed a single mass in the right kidney (14 × 12 mm), which was further confirmed by the computed tomography (CT) (Fig. 1A and 1B). No mass was found in the left kidney. Biopsy of the right renal mass was performed under ventrotomy. The mass consisted of interlacing bundles of spindle cells. No clear mitoses and no invasive proliferation into the surrounding tissues were noted; therefore, this tumor was suspected as benign.

Fig. 1.

Ultrasonography (A) and contrast enhanced computed tomography (CT) (B) of the mass within the right kidney in case 1. The mass is shown by arrowheads in each figure. The mass encapsulated by hypoechoic wall with the mass body showing mixed pattern was found in cortex to cortico-medullary junction (A). The mass located in cortex to cortico-medullary junction is enhanced by contrast agent (B).

Because this tumor was assumed to be benign based on the above results, and because the owner was hesitant about surgery at that time, watchful waiting approach was chosen. However, polycythemia persisted with packed cell volume (PCV) over 60% for most of the time. After considerable discussion with the owner, nephrectomy of the right kidney was performed on day 72 (Fig. 2). Histopathologic features of the mass of the surgically removed kidney were identical to those of the biopsied specimen and the tumor was unencapsulated (Fig. 3A–Band 3D). Immunohistochemistry revealed that the tumor cells reacted to vimentin (Fig. 3C and 3E), suggesting their mesenchymal origin. The tumor cells were cyclooxygenase (COX)-2 negative (Rabbit polyclonal antibody, Proteintech Inc., Rosemont, IL, USA; data not shown). The matrix of this tumor was negative for Alcian blue (data not shown). By electron microscopy, the tumor cells showed spindle-shaped morphology containing rough endoplasmic reticulum, Golgi apparatus and cytoplasmic filaments but lacked lipid droplets (Fig. 3F). Based on these results, this tumor was suspected to be RICT.

Fig. 2.

A picture of gross appearance of the mass after nephrectomy in case 1. A single, solitary, pale pink mass existed within cortex to cortico-medullary junction.

Fig. 3.

Histopathology and electron microscopy of the resected mass in case 1. HE stain (A, B, D), immunohistochemistry with vimentin (C, E), and electron microscopic figure (F). The tumor was within cortex to cortico-medullary junction and its cells replaced normal structures and compressed adjacent parenchyma (A). The margin of the tumor was unencapsulated and migrated smoothly with the stromal tissue (B). The tumor consisted of interlacing bundles of spindle cells (D). No clear mitoses and no invasive proliferation into the surrounding tissues were noted (B, D). The tumor cells reacted with vimentin (C, E), suggesting that these cells were of mesenchymal origin. By electron microscopy, the tumor cells showed spindle-shaped morphology containing rough endoplasmic reticulum, Golgi apparatus and cytoplasmic filaments (F).

The serum EPO concentration was analyzed using blood sample taken at the day of surgery, and the result was even higher than before (69.3 mIU/mL, FUJIFILM VET Systems). After the nephrectomy, the serum EPO concentration drastically declined to 0.6 mIU/mL by day 75 (Fig. 4A), and polycythemia improved gradually as well (Fig. 4B). Immunohistochemistry using anti-human EPO rabbit polyclonal antibody (ab65394, Abcam, Cambridge, UK) was tried with surgically removed renal mass, however, EPO expression was not detected in the section of the tumor (data not shown). By day 839, there was no recurrence of tumor, no polycythemia nor rebounding of serum EPO concentration in this case.

Fig. 4.

Serum erythropoietin (EPO) concentration (A) and packed cell volume (PCV) (B) before and after the removal of the renal mass in case 1. Reference intervals are shown with solid gray area. The day of surgery (day 72) is shown with the white arrowhead. Both the EPO level and PCV declined to reference level after removal of the right kidney with the mass.

Case two, a 11-year-old, neutered male Toy Poodle, weighing 7.7 kg, was referred to Tokyo University of Agriculture and Technology Animal Medical Center due to five month-history of polyuria and polydipsia, accompanied by extremely PCV value at referring veterinary clinic. The dog had a history of urinary bladder calculi and otitis externa. Other than polyuria and polydipsia, the dog was otherwise healthy. On physical examination, increased redness in mucous membrane and skin and cardiac murmur of Levine 2/6 was observed. The dog was well hydrated. CBC revealed severe polycythemia (RBC 1,280 × 10⁴/µL, Hb 28 g/dL, PCV 87.4%) and mild thrombocytopenia (14.7 × 10⁴/µL). Reticulocyte count was increased even in the face of this polycythemia (145.9 × 10³/µL; reference range 10–110 × 10³/µL). There was no abnormality in thoracic radiograph and cardiac ultrasound, and blood gas analysis. Abdominal ultrasound analysis revealed no evident morphological abnormality in the kidneys or any other evidence of tumor. Blood EPO concentration was 22.0 mIU/mL (FUJIFILM VET Systems), which was high in the face of severe polycythemia without evidence of hypoxia. To search for any evidence of tumor which could not have been detected by radiography and ultrasonography, contrast-enhanced CT scan was performed, which revealed a round mass, 15 mm in diameter, enhanced in arterial phase within cortex of the left kidney (Fig. 5). Since polycythemia was severe, phlebotomy followed by fluid therapy was done multiple times until the left kidney was surgically removed.

Fig. 5.

Contrast-enhanced computed tomography scan in case 2. The round mass found within cortex of the left kidney (arrow heads). Venous phase (A), arterial phase (B), and equilibrium phase (C).

The mass within the left kidney was solid, well-circumscribed and grossly whitish compared to the adjacent cortex (Fig. 6A). Histopathology revealed that oval to spindle-shaped tumor cells proliferated forming sheets, cords, or trabeculae. The tumor cells had indistinct cell boundaries, moderate to abundant pale eosinophilic lacy cytoplasm with occasional small vacuoles, slightly anisokaryotic oval to elongated nucleus, and indistinct nucleolus. Mitotic figures were not observed in 2.37 mm² (10 hpf using the ocular lens of frame number 22). The tumor interstitium contained numerous capillaries and multifocal area of fibroplasia (Fig. 6B and 6C). Based on this characteristic, renal interstitial cell tumor was most suspected. The matrix of this tumor was weakly positive for Alcian blue (Fig. 6D). Immunohistochemistry for COX-2 and EPO (rabbit anti-human EPO polyclonal antibody, Proteintech Inc.) were both negative (data not shown).

Fig. 6.

Gross appearing of the mass within the left kidney of the case 2 (A), arrowhead showing the mass. Histopathology of the mass (B, HE stain, bar=5 mm), and its magnified image showing tumor area (C, HE stain, bar=50 µm), which is shown by the square in Fig. 6B. The area shown with rectangle (B) correspond to the magnified image (C). Alcian blue stain showing weak positive reaction within the matrix of this tumor (D), and a picture of positive control is shown at the upper left corner.

Blood EPO concentration dropped immediately on the next day after the surgery, then continued declining to below the detection limit by day six (Fig. 7A). At the same time, polycythemia gradually resolved, and PCV level was within reference range by day 14 after the surgery (Fig. 7B). Normalization of polycythemia was accompanied by resolution of polyuria and polydipsia.

Fig. 7.

Changes of erythropoietin (EPO) concentration (A) and packed cell volume (PCV) (B) before and after surgical removal of the left kidney with tumor in case 2. Reference intervals are shown with solid gray area. Pre-op represents the day of first presentation, and each number of days is the days after the surgery. The EPO level declined to subnormal level and PCV declined to within reference level after surgery.

The RICT in these two dogs was clinically and histopathologically benign, forming a single mass in right (case 1) or left (case 2) kidney. While current cases were quite different from recently reported cases of RICT in dogs which showed malignant characteristics [1, 17], characteristics of the current cases shared some similarities with a report made decades ago, except for concurrent polycythemia [3]. Based on the clinical course before and after surgical resection of the renal lesion, polycythemia observed in the current two cases was strongly suspected to be the result of inappropriate EPO production in relation to the RICT masses. However, contrary to expectations, EPO expression was not demonstrated by immunohistochemistry. The polyclonal antibody used was anti-human antibody, yet because distal tubule within the same specimen stained positive, the antibody seemed to be working. Thus, the immunohistochemistry results suggested limited possibility that tumor cells were producing EPO in current cases. Another possibility is that EPO production was enhanced by hypoxia of surrounding tissue compressed by the mass. This mechanism, yet still no more than hypothesis, has been proposed in previous reports of various renal tumors [4, 10, 15]. It is also yet to be considered whether a small mass, such as those seen in current dogs, could cause such compression. The third possibility is that substance released by the tumors stimulated EPO production. There is a study in human medicine, about leiomyoma of which the patient was accompanied by high EPO concentration. It was shown that the leiomyoma did not produce EPO itself, however, the extract of the tumor stimulated EPO production, demonstrated by injecting into mice [7]. Although similar study cannot be performed in current dogs whose tumor tissues were all treated with formalin, further study is needed to discuss whether this phenomenon would take place in RICT as well. Polycythemia have been recognized in renal lymphomas [4, 8, 12, 18], renal adenocarcinoma [2], renal carcinoma [16], nephroblastoma [11], renal sarcoma [15], and renal fibrosarcoma [9]. To the author’s knowledge, there is no report stating polycythemia in combination with RICT in dogs.

The tumor was recognizable by ultrasonography in case 1 and was not in case 2, despite similar tumor size in these two cases. In case 2, ultrasonography was repeated after CT scan targeting where the mass was observed with CT, and the mass was visualized extremely faintly in second ultrasonography (data not shown). It was concluded that the mass was not detectable during the first ultrasonography because of its nearly equal echogenicity with the surrounding tissue. Even when renal mass is undetectable with ultrasonography in polycythemia cases with other secondary reasons excluded, persistent exploration for its possibility by using different device such as contrast-enhanced CT scan is important to make confirmation.

Although morphological and histological features of the tumors in both cases were strongly suggestive of RICT, Alcian blue and COX-2 staining showed variable results. Alcian blue staining in the matrix of the tumor was negative in case 1 and was positive in case 2. There are only few documents in veterinary medicine that matrix should be Alcian blue positive for interstitial tumors and negative for fibromas [3, 5]. From the aspect of Alcian blue reaction alone, the tumors in case1 and 2 might have been fibroma and RICT, respectively. Electron microscopy of case 1 revealed insufficient microstructure preservation and suspected poor tissue fixation. Therefore, the fixation conditions may have influenced the failure to obtain a reaction with Alcian blue staining or immunohistochemistry. COX-2 staining of the tumor cells was negative in both dogs this time. Previously in humans, tumor cells of RICT were reported as COX-2 positive [14], while those of fibrosarcoma were reported as COX-2 negative [1, 9, 13]. A previous single case report in a dog diagnosed as RICT show that the tumor cells were COX-2 positive [1]. After several years, RICT in human medicine is currently termed as “renomedullary interstitial cell tumor” [6, 19], and according to most recent documentations, Alcian blue and COX-2 staining are not required for its diagnosis. Following these updates, whether expressions of Alcian blue or COX-2 by the tumor cells are necessary or not for diagnosing this tumor in dogs should be considered.

As a conclusion, RICT in dogs may cause polycythemia even when it is small. The tumor may not be evident using ultrasonography, and arduous search for any tumor using other devices such as contrast CT scan is warranted. Whether the tumor cells of RICT can produce EPO or not remains to be identified. Due to rarity of RICT in dogs, current criteria are forced to be based on the limited information. Transition of histological definition as seen in human medicine may be considered in veterinary medicine as well, and histological diagnostic criteria in detail should be updated based on thorough study.

POTENTIAL Conflicts of Interest

The authors have nothing to disclose.