Granulocytopenia Associated With Thymoma in a Domestic Shorthaired Cat

Introduction

Thymoma tends to be a disease of older cats and dogs with a median age of 9.5 years.¹ Paraneoplastic disorders and primarily autoimmune diseases have been linked with thymomas in humans and domestic animals. Previously, cats have been reported with myasthenia gravis ^2–4 and exfoliative dermatitis.⁵ Dogs have been presented with myasthenia gravis as well, and the presence of megaesophagus can be a negative prognostic indicator.⁶

Agranulocytosis is a rare but reported phenomenon in humans with thymoma.^7–9 Granulocytopenias are much rarer than other autoimmune disorders, such as myasthenia gravis, pure red cell hypoplasia, or hypogammaglobulinemia; however, they can also be linked to autoimmunity triggered by the presence of a thymic tumor. Granulocytopenias can be found in conjunction with other cytopenias, as red cell aplasia is associated with a decrease in leukocytes or platelets in 30% of patients; but granulocytes can also be the sole cell line affected.

In humans, the preferred approach to treatment is to resolve the neutropenia with immune suppression and then attempt surgical removal of the tumor. This can be extremely difficult, as many humans have true agranulocytosis with a complete lack of granulocyte precursors in the bone marrow. Once the thymic mass has been removed, autoimmunity may or may not resolve, and not all patients with cytopenias respond to thymectomy. An estimated 38% of human patients with red cell aplasia respond to thymectomy.⁹

To the authors’ knowledge, this syndrome has never been reported in the cat or any other domestic animal. The purpose of this paper is to describe a case of suspected immune-mediated granulocytopenia associated with a thymoma in a cat that underwent successful treatment.

Case Report

An adult, spayed female, domestic shorthaired cat was presented to the referring veterinarian for halitosis in October of 2005. The cat had been found as a spayed adult 3 years previously, and the age at that time had been estimated to be 2 years. Previous history included a dental prophylaxis and removal of the left upper canine tooth because of a tooth root abscess. The cat lived in a multicat (five total) household with access to outdoors. Previous feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) tests had been negative.

Upon examination, the referring veterinarian had noticed tachypnea, and thoracic radiographs were taken. A large, cranial mediastinal mass was noted. Complete blood count (CBC) and serum biochemical results were within reported reference intervals. An aspirate of the cranial mediastinal mass was submitted to a commercial laboratory for cytological analysis. Mast cells and moderate numbers of white blood cells (WBCs) (a mixture of small, mature lymphocytes and occasional monocytes and eosinophils) were seen. The sample was reported as suggestive of mast cell neoplasia. The cat was treated with oral amoxicillin-clavulanic acid^a (19 mg/kg q 12 hours) and prednisone^b (1.5 mg/kg q 12 hours for 5 days, then q 24 hours for 5 days), with no improvement in breathing noted.

The cat was presented for evaluation atWashington State University (WSU) Veterinary Teaching Hospital on November 7, 2005. On physical examination, the cat was bright, alert, and responsive. Temperature was 99.6°F (37.5°C), heart rate was 180 beats per minute, and respirations were 66 breaths per minute. A mild amount of dental tarter was noted. Heart and lung sounds were normal, and pulses palpated normally. Respirations were shallow, and the cranial thorax was noncompliant to compression. The abdomen and peripheral lymph nodes palpated normally. Thoracic radiographs were repeated, and a soft-tissue mass was seen occupying the majority of the thorax, particularly on the right side where the mass extended to the level of the diaphragm [Figures 1A, 1B]. Ultrasonography was performed, and the thoracic mass was found to be of heterogeneous echogenicity and echotexture, with several small, anechoic, ill-defined areas causing distal acoustic enhancement. Mild vascularity was noted within the mass with color Doppler. The vena cava was compressed within the thoracic cavity and was dilated caudal to the diaphragm with turbulent blood flow. The mass did not extend across the diaphragm, and the abdominal cavity was normal except for the presence of a small amount of free abdominal fluid seen between the liver lobes.

Repeat aspiration of the mass yielded small lymphocytes, red blood cells, and mast cells. Based on the location of the mass and the presence of small lymphocytes and mast cells, the cytological findings were suggestive of a thymoma, despite the absence of epithelial cells. Biopsy was recommended. Clotting times were within reported reference intervals (one-stage prothrombin time 11.4 seconds, reference interval 8.7 to 12.9 seconds; activated partial thromboplastin time 13.5 seconds, reference interval 11.0 to 16.7 seconds). A needle biopsy^c was taken under ultrasound guidance and submitted for histopathology. The sample consisted of large numbers of round cells forming sheets that were separated by a fine fibrovascular stroma. Approximately 30% of the cells had cytokeratin immunore-activity; occasional cytokeratin-positive large cells with intracellular lamellae (interpreted as Hassall’s corpuscles) were detected within the mass, allowing definitive diagnosis of thymoma [Figures 2A-2F].

Because of the extreme size of the mass, coarse-fractionated radiation therapy was recommended to decrease the mass prior to surgery. Beginning November 21, 2005, four fractions of 5 Gy of radiation were delivered using 6 MV photons via a linear accelerator.^d Treatments were given once weekly. The cat was placed in right lateral recumbency, and bilateral parallel-opposed portals were used. The field was decreased in size on the last treatment, because a decrease in the size of the mass was detected [Figures 3A, 3B]. A mild, peribronchiolar interstitial pattern was also seen in the aerated portions of the lungs, which had not been previously visible.

On reevaluation January 25, 2006 (6.5 weeks after finishing radiation therapy), the cat was clinically doing well. Respirations were still 64 breaths per minute, but chest excursions were less exaggerated. Thoracic radiographs were repeated, and the mass was stable in size. In preparation for surgery, serum biochemical analysis and a CBC were repeated. Serum biochemical values were normal, but the blood count revealed marked neutropenia (segmented neutrophils 374/μL, reference interval 2.5 to 12.5 × 10³/μL) with no abnormalities in other cell lines [see Table].

Bone marrow aspiration the following day yielded highly cellular samples with both erythroid and myeloid cell lines present. Within the myeloid line, the numbers of band neutrophils were greater than the numbers of segmented neutrophils. The myeloid to erythroid ratio was approximately 1:1. Low numbers of small lymphocytes and plasma cells were seen. The sample was interpreted as suggestive of a maturation arrest at the segmented neutrophil stage or as an increased production of early granulocyte precursors in response to severe peripheral neutropenia. Immune-mediated destruction of mature neutrophils was suspected, but an overwhelming tissue demand for neutrophils in excess of bone marrow production capacity could not be ruled out. However, the lack of a left shift in the peripheral blood and the absence of toxic changes in the neutrophil series in the presence of such a severe neutropenia were less supportive of increased consumption as the cause of the severe neutropenia.

As no obvious reason for increased neutrophil demand was identified, immune suppressive therapy was instituted January 26, 2006. Prednisone was initially given orally at 15 mg daily (4 mg/kg per day), and oral amoxicillin-clavulanic acid (18.4 mg/kg) was given prophylactically to prevent infection. After 4 days, the neutrophil count appeared to increase (828/μL), but the following day it dropped to 148/μL. Oral cyclosporine in liquid formulation^e was then added at 20 mg twice daily (12 mg/kg per day), and the antibiotic was discontinued because of inappetence. After 7 days of cyclosporine treatment, the WBC count was unchanged. After an additional week, the WBC count remained low, and the formulation of cyclosporine was changed to a tablet form^f to ensure the cat was receiving the prescribed dose. The dose was also increased to 25 mg twice daily (15 mg/kg daily). On March 22, 2006, 7 weeks after the initiation of immune suppressive therapy, the total WBC count was 8900/μL (reference interval 5300 to 21,000/μL) with 7209/μL segmented neutrophils. In preparation for surgical intervention, the dose of prednisone was rapidly tapered and stopped (March 26, 2006) because of concerns regarding suppression of wound healing. Cyclosporine was continued.

On March 28, 2006, the cat was presented to WSU for surgical removal of the thymoma. The owner felt the cat’s appetite and activity were normal, although the cat had lost a small amount (0.16 kg) of weight. An ulcerated proliferative mass was noted on the right nasal planum, partially occluding the right nostril. The lesion had first appeared 7 weeks prior and had increased in size slowly over time, while the cat was on immune suppressive therapy. In addition, the gingival tissues were hyperplastic. A CBC showed a total WBC count of 3800/μL with 3116/μL segmented neutrophils. Thoracic radiographs showed no change in the mediastinal mass or other thoracic structures. The cat was taken to surgery the following day for thoracic mass removal and biopsy of the nasal planum lesion and gingiva.

The thorax was entered via a median sternotomy. The mass was found to be closely adhered to the pericardium and mediastinum, but it was successfully dissected away and removed along with the majority of the pericardium. A skin biopsy punch was used to biopsy the nasal planum lesion, and a small sample of gingival tissue was also taken. The cat remained in the hospital for 8 days following surgery, primarily because excess pleural fluid accumulation had required a thoracic drain to stay in place for 7 days, and the cat had breathing difficulties that were compounded by occlusion of the right nostril. The cat was maintained on cyclosporine at 25 mg twice daily initially, but 3 days postsurgery the dose was decreased to 25 mg once daily. Neutrophil counts postsurgery were 6216/μL on day 3 and 15,936/μL on day 8.

Histopathological examination of the central tissues of the mediastinal mass revealed that 95% of the sample was composed of necrotic cells mixed with small amounts of fibrin, with no detectable inflammatory cells. Peripherally a small rim of lymphocytes mixed with loose connective tissue, capillaries, macrophages, and epithelioid cells was seen. No definitive histogenesis could be determined; it was assumed that the previous radiation therapy had altered the mass beyond recognition in the areas examined. The nasal planum lesion was determined to be a severe, chronic, lymphoplasmacytic rhinitis and dermatitis; rare cells within the epidermis contained smudged, eosinophilic, intranuclear inclusion bodies suggestive of herpesvirus infection. Areas of hyperplasia associated with the dermis were thought to be a reaction to inflammation, although squamous cell carcinoma could not be ruled out. No microorganisms were detected via Brown-Hopps, Ziehl-Neelsen acid-fast, and Gomori methenamine silver stains. The gingival samples were diagnosed as chronic, focal, gingival epithelial hyperplasia —most likely secondary to gingivitis.

The cat was discharged April 6, 2006; the owner was instructed to have the cat’s CBCs checked weekly and to wean her off cyclosporine if counts remained normal. On April 11, 2006, the segmented neutrophil count was 9345/μL, and the dose of cyclosporine was lowered to 12.5 mg once daily. On April 18, 2006, the segmented neutrophil count was 7232/μL, and cyclosporine was discontinued. One week later, the cat maintained a normal complete neutrophil count [see Table], and the nasal lesion and gingivitis were regressing. At 8 months postsurgery, the cat is reported to be clinically normal in all respects and is off all medications; clinical laboratory submissions, however, were not repeated.

Discussion

Although thymoma tends to occur in older cats, the age of the cat reported here was estimated to be only 5 years, but it may have been older than first believed. Clinical signs reportedly include dyspnea, anorexia, coughing, lethargy, and dsyphonia.³ The cat presented here had no appreciable signs, except tachypnea. Treatment for feline thymoma generally is surgical removal of the mass; however, radiation therapy has been reported as a treatment for feline thymoma either alone or as adjunctive therapy.^4,10 Radiation was chosen as the initial therapy in this cat because of the size of the mediastinal mass and concerns that anesthetic and surgical risks were unacceptably high given the cat’s compromised respiratory system. The radiation protocol chosen for this cat’s treatment was based on the author’s previous experience with coarse-fractionated radiation of feline thymomas and a desire to anesthetize the cat as few times as possible. Presurgical radiation of the mass did achieve the goal of decreasing tumor size. In fact, the cat also showed a small amount of initial weight gain (prior to immune suppressive therapy), indicating possible improvement in clinical status despite the owner’s initial thoughts that the cat had been doing fine prior to the diagnosis.

The unexpected event in this cat was the extreme neutropenia observed at 6.5 weeks postradiation therapy. Neutropenia was the only cytopenia noted, and the bone marrow showed active production of neutrophils at the time. Eosinopenia also may have been present, as no eosinophils were seen while neutrophil counts were low, and small numbers were again seen when neutrophil numbers returned [see Table]. If the problem was somehow related to radiation- induced bone marrow toxicity, suppression of all cell lines in both peripheral blood and bone marrow would have been more likely consequences, as all cell lines would have received the radiation.

In dogs receiving total-body radiation, granulocytes generally disappear within 4 to 14 days postradiation; monocytes disappear within 4 to 20 days; lymphocytes disappear within 1 to 20 days; platelets disappear within 8 to 20 days; and reticulocytes disappear within 3 to 14 days. Similar data have also been found in cats.^11,12 With proper marrow engraftment, granulocytes recover quickly (in 8 to 22 days). Certain bone marrow findings are expected with granulocytopenia from acute stem cell injury to the myeloid cell line following radiation therapy. Such findings are a decrease in the myeloid to erythroid ratio from granulocytic hypoplasia, with a proportionate decrease of myeloid precursors in both mitotic and postmitotic neutrophil pools. Also, the cat did receive a large field of radiation, but not total-body radiation. In half-body radiation, only transient and mild neutropenia and thrombocytopenia have been seen at 10 to 14 days postirradiation ¹³ or at 5 weeks postradiation if chemotherapy is given concurrently.¹⁴

As the bone marrow findings showed a normal myeloid to erythroid ratio and a left shift within the granulocyte lineage and no alterations in other cell lines, consumption or destruction of the mature neutrophils was suspected. The majority of the tumor was necrotic at removal secondary to radiation therapy. Therefore, temporary sequestration of neutrophils within the tumor cannot be ruled out as the initial cause of the neutropenia. This scenario seems significantly less likely as the cat was never clinically ill or febrile, and no left shift or toxic changes within the neutrophil series were ever detected.

In a recently reported retrospective study of neutropenia in dogs and cats, the largest category of diseases associated with neutropenia was nonbacterial infectious disease, such as FeLV or FIV, while immune-mediated disease was the least common.¹⁵ The neutropenia in the reported cat resolved with aggressive immune suppressive therapy, and following thymectomy the cat was able to be weaned off immune suppressive therapy. The evidence supports a diagnosis of immune-mediated neutropenia. Confirmation of this diagnosis requires demonstration of antigranulocyte antibodies. For evaluation of antineutrophil antibodies in cats, indirect leukoagglutination and fluorescence tests are available, but so far these have demonstrated very low sensitivity and are considered unreliable.¹⁶ Most recently, a study using an indirect-flow cytometric test for detecting antineutrophil antibodies in dogs found consistent results with this approach. Unfortunately, this approach has been validated only recently and only in the dog.¹⁷

Once neutropenia was identified, removal of the thymic mass was thought to be the best course of action. To reduce the risk of infection, however, it was optimal to increase the neutrophil count prior to surgery. Neutrophils are an important part of the body’s first line of defense, particularly against bacterial invasion.¹⁸ Although surgery is performed in a sterile setting, it is impossible to prevent introduction of all bacteria into surgical wounds. Neutrophils also play a minor role in tissue healing and aid with the cleanup of dead tissues.¹⁹ Although these functions are also carried out by macrophages, of which numbers were normal in this cat, it was desirable to increase neutrophil numbers prior to surgery.

Immune suppression with corticosteroids presented the additional problem of delayed wound healing. Prednisone was initially chosen to establish a quick immune suppression response. When neutropenia was resolving, prednisone therapy was quickly tapered and stopped before the surgical intervention. When further immune suppression was necessary, cyclosporine was chosen because of its potent anti-T cell activity. The dose was increased to a high level before the desired effect was achieved, which likely explains the nasal and gingival lesions that formed; cyclosporine use has been linked to mucosal hyperplasia, herpes virus infection, and formation of squamous cell carcinoma.^20–22 No evidence of tumor was found in the nasal lesion (tumors can arise secondary to cyclosporine therapy), and a local infection was considered the more likely cause.

Conclusion

This case points out the importance of examining animals for the presence of paraneoplastic disorders secondary to thymoma both before and after initiating treatment. When diseases are found that may represent autoimmune disorders, clinicians must also remember to look for thymomas as an underlying cause. Agranulocytosis or other cytopenias should be added to the list of disorders potentially occurring secondary to thymoma in domestic animals or at least in cats. Removal of the thymic mass in this cat appears to have been an important step in resolving the presumed immune-mediated granulocytopenia. Also, high-dose cyclosporine was tolerated in this cat, but epithelial complications were seen.