Iatrogenic Hyperadrenocorticism, Calcinosis Cutis, and Myocardial Infarction in a Dog Treated for IMT

Introduction

Calcinosis cutis is an uncommon disorder in dogs in which inorganic calcium and phosphate ions are inappropriately deposited in the dermis, epidermis, or subcutis.^1–4 Calcinosis cutis has been broadly classified into four categories: dystrophic, metastatic, idiopathic, and iatrogenic.^1–3

Arteriosclerosis and ischemic heart disease have been identified as potential causes of heart failure and sudden death in dogs.⁵ Intramural coronary arteriosclerosis results in luminal narrowing, which may lead to myocardial infarcts and altered cardiac function.⁵ Myocardial necrosis has been associated with a wide variety of conditions, including endocrine, pulmonary, hematologic, neoplastic, renal, and cardiac diseases.⁶ In humans, hyperadrenocorticism (HAC) is thought to induce or exacerbate heart disease. Possible mechanisms of damage to the cardiovascular system in human patients include hypertension, central obesity, hypercholesterolemia, hypertriglyceridemia, increased thrombotic tendency, and impairment of glucose tolerance.^7,8

This report describes a dog with dystrophic calcinosis cutis and myocardial necrosis secondary to iatrogenic HAC following immunosuppressive therapy with prednisone for immune-mediated thrombocytopenia (IMT).

Case Report

An 8 yr old male English bulldog weighing 24.4 kg presented to the Veterinary Teaching Hospital, Western College of Veterinary Medicine, University of Saskatchewan for a recheck following a 4 wk course of cephalexin^a for tail fold pyoderma. In addition to the tail fold pyoderma, the dog had a history of recurrent otitis externa and pododermatitis. During the examination, petechiation of the oral mucous membranes and ecchymoses of the perianal region were noted, and investigations were begun for suspected thrombocytopenia (day 1). Diagnostic tests performed at this time included a complete blood count (CBC), serum biochemistry profile, urinalysis, thoracic and abdominal radiographs, and an abdominal ultrasound. Testing for tick-borne diseases and leptospirosis was not pursued as these diseases were uncommon in the dog’s geographical location, and the dog had no significant travel history. As shown in Table 1, thrombocytopenia was confirmed on the CBC (platelet count was <10×10⁹/L; reference range, 200–900 × 10⁹/L ). The only significant abnormality on the serum biochemistry (Table 2) was a moderate elevation in alkaline phosphatase (ALP). All remaining diagnostic tests were unremarkable. Bone marrow aspirate cytology showed megakaryocyte hyperplasia. This, along with peripheral thrombocytopenia, supported peripheral destruction as the underlying mechanism for the thrombocytopenia. Based on these findings, treatment of IMT was initiated. Although difficult to verify, it is possible that the cephalexin^a previously administered could have triggered the immune-mediated destruction of platelets.

TABLE 1 Summary of Relevant Hematology Findings

View Fullscreen

TABLE 2 Summary of Relevant Serum Biochemical Findings

*

The corrected serum calcium values were calculated using the following equation: corrected calcium (mg/dL)=serum calcium (mg/dL)−serum albumin (g/dL)+3.5.¹¹

^†

The serum calcium×phosphorus product was calculated using the following equation: calcium×phosphorus product=corrected serum calcium (mg/dL)×serum phosphorus (mg/dL).¹¹

View Fullscreen

Immunosuppressive therapy with prednisone^b was initiated (1.02 mg/kg per os [PO] q 12 hr) with a plan to decrease the dose in 2–3 wk depending on the dog’s response to therapy. Inadequate increase in platelets by day 7 (Table 1) prompted administration of a single dose of vincristine^c (0.02 mg/kg IV on day 8). Azathioprine^d (2 mg/kg PO q 24 hr) was prescribed for its prednisone-sparing effects and for additional immunosuppression. Side effects of the prednisone therapy, including polyuria, polydipsia, and muscle weakness, were noted shortly after initiating the corticosteroid.

CBCs performed on days 11 and 21 showed that the platelet count was within the reference range, and there were no signs of azathioprine-induced myelosuppression (Table 1). These CBCs revealed a mild leukocytosis (characterized by a mild neutrophilia and a mild monocytosis), which was consistent with a stress leukogram secondary to prednisone therapy and/or inflammation. On day 21 the prednisone dose was decreased to 0.8 mg/kg PO q 12 hr with a further plan to taper the dose of prednisone by approximately 25% q 2 wk. On day 39, the dog was examined again due to the development of an abscess in the left axillary region. A CBC revealed a mild leukocytosis characterized by a mild neutrophilia with a mild left shift, mild lymphopenia, mild monocytosis, and mild toxic change. The platelet count was within the reference range (Table 1). The neutrophilia, lymphopenia, and monocytosis were attributed to glucocorticoid administration and/or inflammation. A serum biochemistry profile (Table 2) revealed marked increases in both ALP and γ-glutamyltransferase (GGT), and mild increases in alanine aminotransferase (ALT), glutamate dehydrogenase (GLDH), and sorbitol dehydrogenase (SDH). Calcium and phosphorus concentrations were within the reference range. The elevation in hepatic enzymes was attributed to steroid hepatopathy, enzyme induction, and/or azathioprine-associated hepatotoxicity. As azathioprine-associated hepatotoxicity could not be ruled out, azathioprine was discontinued on day 39. Urinalysis showed dilute urine (urine specific gravity 1.014), mild proteinuria (1+), sperm, and scant struvite crystals. The dilute specific gravity and proteinuria were attributed to the prednisone. The abscess was lanced and flushed under general anesthesia, and a 3 wk course of clindamycin^e (12.3 mg/kg PO q 12 hr) was prescribed. Cytology of the skin surrounding the abscess revealed a cellulitis characterized by neutrophilic inflammation. Culture/sensitivity showed 4+ Staphylococcus pseudintermedius, which was sensitive to all drugs except for penicillin and ampicillin. The underlying cause of the abscess was unknown, but immunosuppression may have predisposed this dog with recurrent skin disease to a deeper infection.

A CBC was repeated on day 49. No abnormalities were evident apart from a mild leukocytosis characterized by a mild neutrophilia, mild lymphopenia, and mild monocytosis. The platelet count remained in the reference range (Table 1).

By day 90 the prednisone dose had been tapered to 0.2 mg/kg PO q 24 hr. The owner reported that the dog had been itching at various “hot spots” that started to develop and that had been getting progressively worse over the preceding 3–4 wk. The dog now appeared to be sloughing the skin overlying the lateral and dorsal aspects of his neck. On physical examination, there was a severe dermatitis widely distributed over the dorsal and lateral aspects of the neck and scapulae characterized by erythema, lichenification, firm and gritty plaques, ulceration, and a malodorous purulent discharge (Figure 1). Based on the findings, calcinosis cutis was strongly suspected, but an adverse drug reaction (including erythema multiforme and pemphigus foliaceus), deep-seated bacterial or fungal pyoderma, demodicosis, and cutaneous neoplasia were other possible diagnoses. A CBC was repeated to monitor the platelet count, and a serum biochemistry profile was repeated to monitor changes in hepatic enzymes. A thorough dermatologic work-up was performed, which included the following: skin cytology (impression smears, skin scrapings), fine-needle aspirates, cultures of the ulcerated skin lesions after removal of the overlying crust, and skin biopsies. Skin biopsies were performed prior to prescribing a 3 wk course of antibiotics because the lesions were very severe and conditions such as erythema multiforme or other forms of a cutaneous drug reaction needed to be rapidly ruled out.

View Figure

• Download as Powerpoint
• Download Figure

The CBC (Table 1) revealed a moderate leukocytosis characterized by a mature neutrophilia and a lymphopenia. The neutrophilia and lymphopenia were attributed to HAC secondary to prednisone therapy and/or inflammation. There was a mild, normochromic, normocytic, nonregenerative anemia either due to a chronic inflammation or (less likely) residual myelosuppression from the azathioprine. The platelet count was within the reference range.

Abnormalities on the serum biochemistry profile (Table 2) included marked increases in both ALP and GGT and mild increases in ALT, GLDH, and SDH. There was also a mild hypoproteinemia characterized by a mild hypoalbuminemia. Calcium and phosphorus concentrations were within the reference ranges. The elevation in hepatic enzymes was attributed to a steroid hepatopathy, but hepatotoxicity from azathioprine or a primary hepatopathy were also possible. Although all hepatic enzymes ALP, ALT, GLDH, SDH, and GGT had decreased substantially they remained elevated. The low albumin was attributed to a negative acute-phase protein response associated with inflammation, but either a primary hepatopathy or loss of protein through the ulcerated skin lesions could not be ruled out.

Impression smear cytology of the skin revealed septic suppurative inflammation. Skin scrapings revealed no evidence of ectoparasitism. Fine-needle aspirates of the skin lesions were consistent with mild inflammation. S. pseudintermedius and Proteus mirabilis were isolated and sensitive to all antibiotics in the sensitivity panel. Histopathologic changes included severe mineralization of the dermal collagen fibers, outer root sheath epithelium, and superficial and follicular keratin (Figure 2). Mineralized collagen fibers were often fragmented, and these altered collagen fibers elicited a moderate granulomatous inflammatory response. These findings were consistent with a diagnosis of calcinosis cutis. Atrophy of the epidermis, dermis, and pilosebaceous units along with the arrest of the hair follicles in telogen were also observed. These changes were all consistent with HAC.

View Figure

• Download as Powerpoint
• Download Figure

Given the severity of the skin lesions and the normal platelet count, prednisone treatment was discontinued in the hopes of reversing the iatrogenically induced HAC and decreasing the risk of infection. Based on the culture and sensitivity findings, clindamycin^e (12.3 mg/kg PO q 12 hr for 3 wk) was prescribed for the secondary pyoderma. Tramadol^f (2.9 mg/kg PO q 6–12 hr as needed) was prescribed for pain. The owners were also instructed to bathe the dog as needed with benzoyl peroxide shampoo^g for its keratolytic and antibacterial properties. The owners were advised that the calcinosis cutis could take between 2 mo and 12 mo to resolve.²

Two weeks later (day 102), the dog was re-examined. The platelet count was at the very upper end of the reference range (Table 1). The skin lesions remained severe, and the dog was increasingly pruritic. Colloidal oatmeal shampoo and cream rinse^h q 3 days was added to the treatment regimen. The owner was also asked to purchase some over-the-counter diphenhydramineⁱ (2 mg/kg PO q 8 hr) and omega-3 fatty acids^j (1 capsule PO q 24 hr) to help control the pruritus as needed. Because of the dramatic rise in the dog’s platelet count, an adrenocorticotropic hormone stimulation test was planned for the dog’s next visit in 2 wk to determine whether the dog had developed spontaneous HAC or still had evidence of iatrogenic HAC.

Ten days later (day 112), the dog presented in a comatose state. The owner thought the dog had seizured that morning and reported that the dog’s breathing had been more labored over the previous 24 hr. They also felt the skin lesions were getting worse despite treatment. The dog had been receiving oral clindamycin, tramadol, diphenhydramine, and omega-3 fatty acids and was being bathed 1–2 times/wk. The owners declined any further examination, diagnostics, or treatment, and the dog was euthanized. A postmortem examination was performed. In addition to the skin lesions previously described, gross lesions included severe pulmonary edema and congestion, as well as hepatic congestion. Significant microscopic lesions were limited to the adrenal glands, skin, lungs, and heart. Histopathology confirmed the original diagnosis of calcinosis cutis, and the presence of bilateral adrenal cortical atrophy was consistent with the diagnosis of iatrogenic HAC. Pulmonary edema was confirmed and was accompanied by numerous hemosiderin-laden macrophages (so-called “heart failure cells”), indicating chronic pulmonary congestion. Mild, multifocal, mineralization of the smooth muscle of the small airways was also observed, which was attributed to HAC. In the heart, moderate, segmental arteriosclerosis of the intramural coronary arteries was noted, which was characterized by fibromuscular intimal thickenings, fragmentation of the internal elastic lamina, and narrowing of the vessel lumen (Figure 3). Vascular changes were accompanied by areas of ischemic damage within the myocardium characterized by multifocal areas of acute myocardial necrosis, mineralization, and fibrosis. No significant gross or microscopic changes were noted in the brain to account for the possible seizure.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

Calcinosis cutis occurs in association with a variety of diseases, but most commonly in dogs with either spontaneous or iatrogenic HAC.^1,3 Calcinosis cutis in humans and dogs can be classified by either the distribution of lesions or underlying cause. Based on the distribution of the lesions, there are two forms: focal (calcinosis circumscripta) and widespread (calcinosis universalis).^1,3 Depending on the underlying mechanism, cases are subclassified as dystrophic, metastatic, idiopathic, or iatrogenic (Table 3).^1,3

TABLE 3 Classification of Calcinosis Cutis by Pathogenesis

View Fullscreen

Dystrophic calcinosis cutis, the most common form in dogs, is thought to occur as a result of alterations within the skin that promote calcification in the absence of changes in serum calcium and phosphorous concentrations.^1,3 HAC is the most common cause of dystrophic calcinosis cutis in dogs and is thought to alter the structure of proteins within collagen and elastin fibers, predisposing them to calcification. Most affected dogs have widespread areas of cutaneous calcification. Few reports of calcinosis cutis resulting from exogenous corticosteroid treatment have been published in dogs.^9,10 Interestingly, calcinosis cutis does not typically develop secondary to either spontaneous or iatrogenic HAC in other animal species or in humans. The most common causes of dystrophic calcinosis cutis in humans include connective tissue disorders, panniculitis, infections, and focal trauma.²

Metastatic calcinosis cutis occurs when the serum calcium and/or phosphorous concentration is elevated, resulting in a serum calcium×phosphorous product >70 mg/dL or >5.6 mmol/L, predisposing to calcium deposition within tissues.¹¹ Kidney disease is the most common cause of metastatic calcinosis cutis reported in the dog.^1,3 It has been hypothesized that dogs with granulomatous inflammation secondary to systemic fungal infections may also exhibit a form of metastatic calcinosis cutis due to increased production of 1,25-hydroxyvitamin D (calcitriol) by mononuclear cells.¹¹

With idiopathic calcinosis cutis, there is no demonstrable underlying tissue damage or metabolic disturbance.^1–3 Generalized idiopathic calcinosis cutis lesions have been reported to develop and spontaneously regress by 1 yr of age in healthy puppies. Generalized idiopathic calcinosis cutis may also occur after systemic illness in puppies. These phenomena are uncommon after 1 yr of age and may be associated with the hyperphosphatemia that is associated with skeletal growth.¹

Iatrogenic calcinosis cutis has been reported after percutaneous absorption of a calcium-containing landscaping product and following subcutaneous injection of calcium-containing solutions administered for the treatment of hypoparathyroidism.^12,13

Given the use of immunosuppressive doses of prednisone, the calcinosis cutis in this dog most likely represents a case of dystrophic calcification secondary to iatrogenic HAC. Other signs of HAC in this patient included elevated hepatic enzymes, pyoderma, muscle weakness, and polyuria/polydipsia. With a corrected calcium×phosphorus solubility product of 38.08 (Table 2), there was no reason to suspect metastatic calcification.¹¹ However, although the literature would traditionally classify this case of calcinosis cutis as dystrophic, this theoretical framework is undoubtedly an oversimplification. Specifically, there is growing evidence that mechanisms classically associated with metastatic calcinosis cutis may play a greater role than previously thought in dogs with HAC. In one study, parathyroid hormone (PTH) and phosphate concentrations were significantly increased in many canine patients with spontaneous HAC compared with other hospitalized control dogs, and no significant difference in either total or ionized calcium concentrations were noted.¹⁴ These abnormalities resolved with treatment of the HAC using trilostane.¹⁵ As the increases in PTH observed were of the same order of magnitude as those previously reported in dogs and cats with renal and nutritional secondary hyperparathyroidism, the term “adrenal secondary hyperparathyroidism” was proposed.¹⁴ Although the pathophysiology of “adrenal hyperparathyroidism” is unknown, it is suspected to play a role in the development of calcinosis cutis, calcium oxalate urolithiasis, and pulmonary mineralization in dogs with HAC. Increased phosphorus concentrations may contribute to an elevated calcium×phosphorus product, and together with elevated PTH levels, may predispose to soft tissue mineralization.^14,15 Although total calcium and phosphate levels were within the reference range in the case described herein, serial measurements of calcium, phosphorus, and PTH concentrations during active mineralization were not available. More research is needed to establish the relationship between HAC, PTH levels, and calcium and phosphate concentrations.

Examination of both the veterinary literature and the human literature further suggests that a multifactorial approach is required to understand the calcification of tissues in patients with calcinosis cutis. In humans, calciphylaxis is a condition where calcium is deposited IV, often with lethal consequences.^16,17 This condition is most commonly reported with end-stage kidney disease. Until recently, calciphylaxis was considered a form of metastatic calcification secondary to an elevated calcium×phosphorus product.¹⁶ However, this is being questioned because the condition occurs in patients with a normal calcium×phosphorus product.^16,17 Numerous “challenging agents” are thought to trigger calcification in these patients, including administration of corticosteroids.¹⁶

In the veterinary literature, calcinosis cutis has been reported secondary to infection with blastomycosis, leptospirosis, and paecilomycosis.^11,18,19 In each of these cases, categorization of calcinosis cutis is difficult, and the underlying cause for calcification is not fully understood. For example, in dogs treated for blastomycosis, classic metastatic changes that may lead to calcification include hypercalcemia secondary to osteolysis, hyperphosphatemia secondary to kidney disease, and increased calcitriol levels with granulomatous disease. However, structural changes in the skin from cutaneous inflammation, increased levels of endogenous steroids, and a drug reaction to amphotericin B may have contributed as well. In the patient described herein, high levels of exogenous steroids and increased calcitriol levels from granulomatous skin disease may have played a role. Similar to the phenomenon of calciphylaxis in humans, further research is needed to characterize the contribution of each factor and to understand how they interact.

Interestingly, there has been significant variability in the reported prevalence of calcinosis cutis in dogs with HAC, ranging from 1.7–8% to 40%. It is likely that the lower prevalences are seen with earlier diagnosis and treatment of HAC.¹ Some authors also claim that calcinosis cutis is more common with iatrogenic than spontaneous HAC.³ In one study, 14% of dogs with iatrogenic HAC developed calcinosis cutis.¹⁰ Although the dose and duration of glucocorticoids associated with calcinosis cutis is unknown, higher doses administered for longer durations are thought to induce changes more rapidly.¹⁰ There is variability in each individual’s susceptibility to iatrogenic HAC, and genetics may play a role. It has been suggested that English bulldogs are particularly prone to developing calcinosis cutis with iatrogenic HAC.³ If this is true, this has implications for the treatment of disorders requiring immunosuppressive therapy in this breed.

The postmortem diagnosis of congestive heart failure (CHF) in the patient described in this report was unexpected given the lack of clinical signs for cardiac disease. No murmurs or arrhythmias were auscultated at any point during the dog’s life. Cough and exercise intolerance were never reported by the owner, and labored breathing was only observed the day before he was euthanized. Thoracic radiographs performed during the initial investigation for IMT showed no changes consistent with cardiac disease or failure. However, as necropsy revealed chronic changes (in addition to acute changes), there is evidence that cardiac disease had been present for some time. It is possible that heart failure was the result of multiple myocardial infarctions secondary to arteriosclerosis of the coronary arteries. It has been reported that coronary arteriosclerosis may lead to myocardial ischemic injury and CHF. As devitalized myocardial tissue is prone to poor contractility and rhythm disturbances, these patients are at risk for heart failure, sudden death, and anesthetic-related death.⁵ Unfortunately, an electrocardiogram, echocardiogram, and blood pressure measurements were not performed in this case as there was no antemortem suspicion of cardiac disease.

The clinical signs observed prior to death in the current case likely reflect an acute exacerbation of more chronic changes in the heart. As no abnormalities suggestive of cardiac disease were detected on prior physical examinations or thoracic radiographs, cardiac disease was not detected in this case until the disease was advanced. The clinical signs of collapse and seizures were similar to those described in dogs reported to have died of myocardial necrosis.^5,6 One study reported that weakness, dyspnea, and collapse within 48 hr prior to death were the most common clinical signs in dogs with myocardial necrosis.⁶ However, another group reported that dyspnea and/or seizures were the only signs observed.⁵ In the current case, the authors postulate that the dog’s comatose state on day 112 was the result of significant ischemic brain injury secondary to myocardial failure. The seizure the dog allegedly experienced likely occurred following an acute exacerbation of hypoxia following either an arrhythmia or decreased cardiac contractility. Unfortunately the presence of ischemic brain injury was not confirmed in this case as there was no evidence of neuronal degeneration and/or necrosis on histopathology. However, the absence of histopathologic lesions does not exclude the possibility of ischemic injury, as neuronal necrosis cannot be observed at the light microscopic level until 12–24 hr following the irreversible hypoxic/ischemic event.²⁰ Therefore, the lack of a histopathologic lesion may simply be a reflection of the short time period between the onset of neurologic signs and death.

The authors suspect that the corticosteroid administration may have exacerbated the underlying arteriosclerotic heart disease in this dog. HAC in people is known to induce or exacerbate several known risk factors for coronary artery disease, including hypertension, central obesity, hypercholesterolemia, hypertriglyceridemia, prothrombotic tendency, and impairment of glucose tolerance. HAC is also known to accelerate the development of arteriosclerosis.^7,8 Human patients with spontaneous HAC have a mortality rate four times higher than that observed in age- and sex-matched control patients as a result of such risk factors leading to cardiovascular complications. This risk is present not only during the active phase of the disease but long after successful treatment of HAC.⁷ It is difficult to extrapolate between species; however, it is plausible that dogs with HAC are predisposed to cardiovascular complications similar to those described in humans. Systemic hypertension associated with HAC may have accelerated the development of arteriosclerosis in this dog, but this cannot be proven because blood pressure measurements were not taken. Corticosteroid administration could also have led to a hypercoagulable state and resulted in undetected microthrombi that caused acute myocardial infarction and subsequent myocardial necrosis.

Conclusion

This report documents a case of dystrophic calcinosis cutis in an English bulldog secondary to iatrogenic HAC. The pathogenesis of calcinosis cutis secondary to HAC is likely multifactorial in origin. The role of factors such as PTH and sensitizing agents in the pathogenesis of calcinosis cutis that occurs with HAC warrants further consideration. In addition, myocardial necrosis secondary to arteriosclerosis should be considered in dogs with HAC that develop CHF or die suddenly.