A 4.5-year-old Quarter Horse gelding was referred for evaluation of a swollen right hind leg of 3 weeks duration, which had occurred after exercise. Lameness was not observed and cold hydrotherapy had resulted in little improvement. Two weeks before presentation, the gelding developed a swelling over the left zygomatic arch for which a tentative diagnosis of abscessation was made. However, the lesion was unresponsive to antibiotics, became pruritic, ulcerated, and exuded serosanguinous fluid. Two days before presentation, dexamethasone powdera was prescribed to treat the leg swelling. On the day of presentation, the gelding also developed a swelling overlying the xiphoid process and a swelling on the right lateral abdomen.

At presentation, the gelding weighed 546 kg and had normal vital signs. The right hind fetlock was swollen circumferentially from the distal third of the metatarsus to the pastern. The swelling was warm, firm, and nonpitting. The gelding was grade 1/5 lame in the affected limb.1 The mass (5 cm diameter, 2 cm depth) over the zygomatic arch was firm, ulcerated, and immoveable. The xiphoid process swelling (1 cm diameter, 1 cm depth) was soft, nonfluctuant, and nonmobile. The soft tissue swelling (15 cm diameter, 1 cm depth) on the right lateral abdomen resembled cutaneous edema.

Biochemical and hematologic values were within reference intervals except for mild eosinophilia. Radiographs of the right hind fetlock revealed a discrete osteophyte on the dorso-medial aspect of the 1st phalanx. Skull radiographs showed a discrete soft tissue swelling surrounding the left zygomatic arch.

Ultrasonography of the right metatarsal area revealed thickening of the subcutaneous tissues (1.1 cm depth) and distention (3.5 cm diameter, 1.5 cm depth) of the digital flexor tendon sheath with anechoic fluid at the level of the fetlock and pastern.

Histopathologic evaluation of a 5-mm-diameter punch biopsy from the mass on the head revealed a poorly defined tumor involving the subcutis and deep dermis. It was composed of multiple, densely cellular islands and nests of pale, neoplastic round cells embedded in large amounts of dense fibrous stroma. Mild-to-moderate infiltrates of eosinophils surrounded these neoplastic aggregates. Tumor cells had pale eosinophilic cytoplasm, uniform oval nuclei with finely stippled chromatin, and inapparent to single small nucleoli. Toluidine-blue staining revealed many fine metachromatic cytoplasmic granules. No mitotic figures were observed. This mass was diagnosed as a cutaneous mast cell tumor. Histopathologic evaluation of two 5-mm-diameter punch biopsies of the mass overlying the xiphoid process revealed marked subcutaneous fibrosis containing mild infiltrates of eosinophils with fewer lymphocytes and plasma cells. There was no evidence of neoplasia. The subcutis contained rare discrete aggregates of cell debris, amorphous hypereosinophilic material, and occasional hyalinized fragmented collagen fibers surrounded by a rim of epithelioid macrophages admixed with eosinophils consistent with foci of collagen degeneration. With the assumption that these findings were representative of the entire xiphoid lesion, a diagnosis of eosinophilic granuloma was made. No parasites were observed in either biopsy sample.

In addition to the histologic diagnoses of cutaneous mast cell tumor and cutaneous eosinophilic granuloma, a clinical diagnosis of trauma to the right hind fetlock was made. Treatments included phenylbutazonea (2 g PO single dose; followed by 1 g PO q12h for 1 day, followed by 0.5 g PO q12h for 2 days) for the swollen right hind leg, and triamcinolone acetonideb (5 mg, single dose intralesionally into the mast cell tumor and eosinophilic granuloma), and dexamethasonec (20 mg PO q24h for 7 days) for the cutaneous mast cell tumor and eosinophilic granuloma. Ivermectind (200 μg/kg PO single dose) was administered before obtaining biopsy results, because of an unknown deworming history and the possibility that the eosinophilia and the masses were related to parasitic infection. The gelding was discharged with instructions to taper the dexamethasone dosage to 10 mg, PO, every 24 hours for 7 days, followed by 10 mg, PO, every 48 hours for 7 days.

Three weeks after discharge, the gelding was re-examined because of acute exacerbation of the swelling, severe lameness with self-trauma to the right hind leg, and inappetence of 2 days duration. The owner reported that the leg swelling would decrease at a daily dose of 20 mg of dexamethasone but increased and was accompanied by the horse aggressively biting at the leg when attempting to lower the dose. The horse weighed 520 kg and had normal vital signs. A grade 4/5 lameness was present on the right hind leg which was markedly swollen from below the tarsus to the coronary band and had areas of alopecia and dried serous exudate. The leg was warm and elicited a pain response upon manipulation. Hoof testers failed to elicit a positive response. The mast cell tumor on the head was reduced in size (1.5 cm diameter, 0.5 cm depth). The eosinophilic granuloma over the xiphoid process and the edema had resolved. Biochemical and hematological evaluation revealed a normocytic normochromic anemia.

Radiographs of the right hind leg revealed discrete bone remodeling along the plantar aspect of the 2nd phalanx, corresponding to degenerative changes. There was mild distension of the metatarsophalangeal, proximal, and distal interphalangeal joints, and extensive soft tissue swelling.

Ultrasonography of the right hind limb revealed thickening of the subcutaneous tissues (0.8 cm depth). The digital flexor tendon sheath was distended (0.7 cm diameter, 2.7 cm depth) with homogenous echogenic material. The metatarsophalangeal and proximal interphalangeal joints were mildly distended with homogenously anechoic fluid and contained thickened synovial membranes (0.5 cm depth) consistent with synovitis.

Fluids from the digital flexor tendon sheath and right metatarsophalangeal joint contained 19.7 × 109 cells/L and 25.9 × 109 cells/L (reference <1.0 × 109 cells/L) and protein concentrations of 51 and 59 g/L (reference <30 g/L), respectively. Based on a 200-cell count, the tendon sheath fluid consisted of 74% eosinophils, 16% neutrophils, 4% macrophages, and 6% small lymphocytes. The metatarsophalangeal joint fluid had a similar cell distribution. A diagnosis of eosinophilic effusion in both sites was made. There was no microbial growth on aerobic or anaerobic cultures from either sample. Evaluation of a 5-mm-diameter skin punch biopsy from the pastern area of the right hind limb revealed a markedly hyperplastic epidermis which was extensively eroded and focally ulcerated. The skin surface was covered with a thick layer of proteinaceous material admixed with cell debris, degenerate neutrophils, and myriads of small coccobacilli. A diagnosis of ulcerative and suppurative dermatitis was made.

Treatments for the right hind limb included phenylbutazonee (1 g IV single dose), morphinef (50 mg in 20 mL sterile saline,g epidural, single dose), and butorphanolh (10 mg IM single dose). The owner declined tenoscopic or arthroscopic evaluation. The tendon sheath was aseptically flushed with sterile saline,g and injected with 5 mg triamcinolone acetonidec and 51 mg hyaluronate sodium.i The fetlock joint was considered inaccessible because of edema. After topical administration of nitrofurazone ointment,j a pressure bandage was applied. Additional treatments included hydroxyzine hydrochloridek (1 mg/kg PO q12h for 14 days, followed by 2 mg/kg PO q12h, duration undetermined), sodium penicillinl (22,000 IU/kg IV q6h for 8 days), dexamethasonea (10 mg PO q24h for 3 days, followed by 15 mg PO q24h for 2 days because of increased pruritus), and omeprazolem paste (2 mg/kg PO q24h for 19 days). Fenbendazolen (10 mg/kg PO q24h for 5 days) was administered on suspicion that parasites may be contributing to the weight loss. Topical hydrocortisoneo cream was applied to the mast cell tumor.

By day 6, the leg was more painful and pruritic, necessitating more aggressive analgesia. A dosage of (50 mg) dexamethasonep IV once daily for 3 days was required to prevent self-trauma. Gradual tapering of the daily IV dexamethasone dose occurred over the next 7 days. Pain and intense pruritus returned when the daily dose of dexamethasone was decreased to 25 mg necessitating increasing the dose to 30 mg until discharge.

Twenty-one days after the 2nd admission, the gelding was discharged with dexamethasone (30 mg PO q24h for 7 days) which was to be reduced by 5 mg every 7 days to the lowest effective dose. Hydroxyzine treatment was prescribed (1 mg/kg PO q12h) for an unspecified period.

The gelding was treated for 2.5 months before presenting a 3rd time. Fifteen days after discharge, the owner had lowered the hydroxyzine dose to 1 mg/kg, PO, every 24 hours. The dexamethasone dose had also been decreased (by 5 mg every 10 days) as the leg swelling improved. Pruritus and self-trauma recurred and limb swelling worsened when the dose of dexamethasone was decreased to 15 mg (PO q24h). Dosage adjustments (20 mg dexamethasone PO q24h; 1 mg/kg hydroxyzine PO q12h) resulted in improvement in clinical signs. Three days before 3rd presentation, the swelling in the right hind limb increased acutely. An increase in dexamethasone to 30 mg (PO q24h) 2 days before final presentation failed to ameliorate the signs.

At 3rd presentation, vital signs were normal. Swelling of the right hind leg was more extensive and severe than noted previously. This limb was warm and sensitive to touch and manipulation. Around the fetlock and metatarsus, areas of alopecia with dried serous exudate were present. A grade 3/5 lameness was present in the right hind limb. Humane euthanasia was performed.

Necropsy revealed a thin carcass. Microscopic examination of the head mass confirmed the previous diagnosis of mast cell tumor.

The subcutis and soft tissues of the right hind leg from just above the stifle to the coronary band were markedly expanded with large amounts of yellow-tinged, clear fluid as well as fibrous tissue. Microscopic examination revealed marked expansion of the subcutis with large amounts of pale, cell-poor fluid and increased amounts of well-vascularized, loose connective tissue. A fluid sample from the metatarsophalangeal joint was normal cytologically, with no eosinophils noted. A 10 cm segment of the plantar surface of the digital flexor tendon sheath distal to the proximal sesamoid bones was markedly thickened and firm. The tendon sheath in this area was 2–4.5 cm thick, had a mild increase in clear viscous fluid, and contained several variably sized pockets of dry, gritty, granular, tan material (Fig 1). The underlying superficial and deep digital flexor tendons were unremarkable. Histologically, focal thickening of the tendon sheath was because of abundant mature fibrous connective tissue containing areas of necrosis and scattered mineral deposits interspersed with clusters of well-differentiated mast cells and small numbers of eosinophils. A diagnosis of digital flexor tendon sheath mast cell tumor with severe locally extensive edema of the distal limb was made. The synovia of the metatarsophalangeal and proximal interphalangeal joints were grossly unremarkable but were not sampled for histopathologic examination. Otherwise, the carcass was grossly unremarkable and no significant microscopic abnormalities were noted in other tissues sampled as part of a complete postmortem examination.


Figure 1.  Right hindlimb, plantar surface of the digital flexor tendon sheath, distal to the proximal sesamoid bones. Cross-sections through this area reveal multifocal, discrete pockets of tan, gritty material embedded in large amounts of dense fibrous tissue (arrows). The underlying digital flexor tendons are unaffected (star).

Download figure to PowerPoint

In summary, postmortem examination confirmed the diagnosis of a mast cell tumor on the head and identified another mast cell tumor in the digital flexor tendon sheath. The biopsy samples taken on initial presentation from the mass over the xyphoid process were small and superficial. It is therefore tempting to speculate that these samples were taken near a mast cell tumor that was not sampled. While metastatic neoplasia could not be completely discounted, this was considered unlikely. Because there was no evidence of a primary tumor, no gross evidence of regional lymph node involvement, and the tumor locations were not typical of metastatic lesions, these findings were most consistent with multicentric mast cell tumors. The tumor cells were well differentiated, exhibited minimal evidence of cellular atypia, and had a low mitotic index.

This report describes a unique equine case in which a tendon sheath mast cell tumor was associated with regional eosinophilic joint inflammation and tendon sheath effusion, severe edema of the distal limb, transient eosinophilia, and an eosinophilic granuloma.

Two cases of equine articular mastocytosis and only 1 case of a mastocytoma in a tendon sheath have been described.2–4 In both cases of articular mastocytosis, eosinophilic inflammation of joint aspirates was prominent while mast cells were either rare or not evident; biopsy of synovial membranes were required for diagnosis.2,3 One of the horses responded to joint lavage and corticosteroid treatment.3 However, similar to the present case the other horse had severe pruritus and lameness; euthanasia was the outcome.2 The eosinophilic inflammation in joint aspirates in the 2 cases with articular mastocytosis are similar to the present case. In the recently described case of a solitary equine tendon sheath mastocytoma, the horse exhibited ultrasonographic findings and clinical signs, such as lameness and tendon sheath effusion, similar to the horse in the present report.4 However, in the previously reported case pruritus, self-trauma, and edema were not described. Treatment, which consisted of surgical exploration and partial resection of the mass, and later medical treatment, was ultimately successful, with the horse returning to full function 30 months postoperatively. In contrast, euthanasia was the outcome in the horse of the present report.

Other possible causes of articular eosinophilic inflammation considered in the present case included idiopathic causes or eosinophilic synovitis secondary to unknown intraarticular injections. As synovial tissue was not evaluated histologically, primary eosinophilic synovitis or synovial mast cell tumor involvement could not be definitively ruled out. However, given the mild changes noted in joint fluid at postmortem and the presence of a mast cell tumor in the adjacent tendon sheath, this neoplasm was thought to be the cause of the earlier joint and tendon sheath eosinophilic inflammation.

Eosinophilia is rare in horses and is usually because of a type 1 hypersensitivity reaction or parasitism; only 1 report of eosinophilia in horses with mast cell tumors was found.5 In the present case, the eosinophilia may have been caused by chemoattractants released from mast cells, an underlying hypersensitivity reaction or parasite infestation. Parasite infestation was unlikely, as no parasites were demonstrated in tissue sections taken before deworming and there was no evidence of parasites in biopsy samples or at necropsy.

In horses, cutaneous mast cell tumors may resolve spontaneously and treatment of benign tumors is typically only required for cosmetic purposes or if the mass restricts normal function.6–8 Surgical resection, even if incomplete, is the treatment of choice and is usually curative.2,9–12 Other treatments include intralesional corticosteroids, cryosurgery, and radiotherapy.3,7,8,13 In this case, intralesional injection of triamcinolone and oral dexamethasone reduced the size of the mast cell tumor on the head, but the leg swelling, pruritus and self-trauma ultimately could not be controlled. Mast cell degranulation from trauma or natural tendon movement may have caused the pruritus and locally extensive edema. Release of histamine and other vasoactive is a feature of mast cell tumors in other species but has not been documented in horses.8,14 Mast cell tumors are usually nonpruritic in horses.8 In 1 of 2 reported cases of equine articular mastocytosis, pruritus was present and euthanasia was the outcome, similar to the present case.2,3


  1. Top of page
  2. Footnotes
  3. Acknowledgments
  4. References

aPhenylbutazone Tablets, Dominion Veterinary Laboratories Ltd

bTriamcinolone Acetonide, Sandoz Canada Inc, Québec, Canada

cDexamethasone powder, Dominion Veterinary Laboratories Inc, Winnipeg, MB, Canada

dEqvalan, Merial Canada Inc, Baie d'Urfé, QC, Canada

ePhenylbutazone injectable, Vétoquinol N.-A. Inc, Lavaltrie, QC, Canada

fMorphine HP25, Sandoz Canada Inc

g0.9% Sodium Chloride Irrigation USP, Baxter Corporation, Toronto, ON, Canada

hTorbugesic, Wyeth Canada, St Laurent, QC, Canada

i Legend, Bayer, Toronto, ON, Canada

j Nitro Ointment, Dominion Veterinary Laboratories Inc

kNovo-Hydroxyzin, Novopharm, Toronto, ON, Canada

l Penicillin G Sodium, Novopharm

mGastroGard, Merial Canada Inc

n Panacur Suspension 10%, Intervet, Whitby, ON, Canada

o Hyderm, TARO Pharmaceuticals Inc, Brampton, ON, Canada

p Dexamethasone 5, Vétoquinol N.-A. Inc


  1. Top of page
  2. Footnotes
  3. Acknowledgments
  4. References

The case was evaluated at the Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3.

This work was not supported by a grant or otherwise.

The authors thank Dr LeeAnn Pack for providing ultrasonographic measurement data and Leonard Doucette for the gross image.


  1. Top of page
  2. Footnotes
  3. Acknowledgments
  4. References