Materials and Methods

Case selection—Medical records of dogs treated surgically for an adrenal gland mass between 1999 and 2005 at The Animal Medical Center were reviewed. Dogs undergoing adrenalectomy for all types of adrenal gland disease were included. Dogs were excluded from the study when the medical record was incomplete.

Medical records review—Data from the medical records was allocated into 3 types (preoperative, intraoperative, and postoperative data).

Preoperative Evaluation

Records were reviewed for signalment, body weight at surgery, body condition score at surgery, clinical signs and their duration, blood pressure measurement before surgery, presurgical laboratory data, and preoperative treatment for adrenal gland disease. Clinical signs evaluated included polyuria, polydipsia, alopecia, vomiting, weakness or lethargy, and panting. When available, results of preoperative CBC counts, serum biochemical analysis, urinalysis, and coagulation testing; concentrations of endogenous ACTH and steroid hormones (including cortisol before and after ACTH stimulation, androstenedione, estradiol, progesterone, 17-hydroxyprogesterone, and testosterone); the urine cortisol:urine creatinine ratio; and results of LDDS and HDDS tests were collected. Systolic blood pressure was obtained indirectly by use of a Doppler system. Hypotension and hypertension were defined as systolic blood pressure < 80 mm Hg and > 180 mm Hg, respectively. Preoperative treatment for dogs with adrenal gland disease was categorized as that for hyperadrenocorticism (ie, with mitotane^a) or suspected pheochromocytoma (ie, phenoxybenzamine). Duration of treatment and treatment prior to surgery coupled with tumor type were not compared against survival time because the duration of treatment and dosages varied considerably in this small sample of dogs.

A BUN concentration > 27 mg/dL and creatinine concentration > 1.8 mg/dL were considered abnormal. Dogs were considered azotemic when BUN and creatinine concentrations were both high. Results for an ACTH stimulation test were considered increased for baseline cortisol concentrations > 6.0 μg/dL and cortisol concentrations > 22 μg/dL after ACTH stimulation. Prothrombin time > 16 seconds and PTT > 30 seconds were considered abnormal. Thrombocytopenia was diagnosed when platelet counts were < 150,000 cells/μL and platelet clumping was not detected. A urine cortisol:urine creatinine ratio > 13.5 was considered abnormal.

Intraoperative Evaluation

Adrenal gland tumors were classified as primary or as an incidental finding. Intraoperative complications (hemorrhage, hypotension, or hypertension), additional lesions (including other gross lesions or tumor thrombi) detected by the surgeon during surgery, number of affected adrenal glands (unilateral or bilateral), side of adrenalectomy, and concurrent nephrectomy were also recorded. Gross lesions included organ enlargement, nodules, or discoloration in organs other than the adrenal glands. Intraoperative hypotension and hypertension were defined as a systolic blood pressure < 60 mm Hg and > 180 mm Hg, respectively. Intraoperative hemorrhage was defined as bleeding during the surgical procedure that required a blood transfusion or was recorded by the surgeon.

Postoperative Evaluation

Factors within the immediate postoperative period were evaluated from the time of surgery until discharge from the hospital but not to exceed a period of 14 days after surgery. Factors evaluated included administration of medical treatments and development of postoperative complications. Postoperative complications evaluated included adrenal gland insufficiency, acute renal failure, pancreatitis, peritonitis, and pulmonary thromboembolism. Adrenal gland insufficiency was defined as baseline cortisol concentrations < 2 μg/dL and cortisol concentrations < 6 μg/dL after ACTH stimulation. Acute renal failure was identified on the basis of a rapid (within 10 days) development of azotemia despite IV administration of fluids in dogs that had renal values within the reference ranges before surgery or worsening azotemia despite IV administration of fluids for those dogs that were azotemic before surgery. All dogs considered to be in acute renal failure also had clinical signs consistent with excretory failure, such as vomiting, lethargy, anorexia, or anuria.

Pancreatitis and peritonitis were confirmed by abdominal ultrasonography and cytologic examination of fluid samples obtained during abdominocentesis, respectively. Pulmonary thromboembolism was diagnosed on the basis of necropsy findings.

Postoperative volume size and volume score of tumors were calculated when 3 measurements (length, width, and height) of the adrenal gland were available after formalin fixation. To standardize these scores as much as possible, values were obtained from measurements obtained by personnel in the Pathology Department of the Animal Medical Center, New York, NY, prior to sectioning for histologic examination. Tumor volume size and tumor volume score were calculated by adding and multiplying the 3 measurements, respectively.

Tissue sections used for histologic examination were reviewed retrospectively by a single board-certified veterinary pathologist (SM). Slides were evaluated for confirmation of adrenal gland disease or tumor type, completeness of surgical excision, and microscopic evidence of invasion. Surgical excision was considered complete when tumor cells did not extend beyond the capsule. Evidence of metastasis was confirmed by histologic examination of other organs (eg, liver and lymph nodes) from which biopsy specimens were collected during the adrenalectomy procedure.

All preoperative, intraoperative, and postoperative factors recorded were evaluated against survival time. Survival time was defined as the interval (in number of days) between surgery and death or euthanasia. Survival was assessed in all dogs and categorized as intra-operative survival (dogs alive at the end of surgery), short-term survival (dogs survived until discharge from the hospital), and long-term survival (dogs survived until after discharge from the hospital). Cause of death was indicated as tumor related or non–tumor related.

Statistical analysis—All recorded variables were evaluated for their association with survival duration via log-rank analysis for categoric variables and by use of Cox proportional hazards analysis in a univariate model. All variables in the univariate model with values of P < 0.2 were placed into a stepwise Cox proportional hazards multivariate model. Median survival times were calculated by use of Kaplan-Meier life table analysis. For survival analysis, dogs were censored when they were still alive at the end of the study, died as a result of another disease, or were lost to follow-up monitoring. Values of P < 0.05 were considered significant.

Results

Forty-one dogs were included in the study. Mean age of dogs at the time of adrenalectomy was 10.9 years (median, 10.7 years; range, 7.0 to 14.0 years). There were 22 spayed females, 1 sexually intact female, 15 castrated males, and 3 sexually intact males. Eleven of 41 (26.8%) dogs were mixed-breed dogs, and 30 (73.2%) were purebred dogs. Breeds represented were the Maltese (n = 6), Labrador Retriever (4), Golden Retriever (2), American Staffordshire Terrier (2), and Petit Basset Griffon Ven deen (2) and 1 each of the Bichon Frise, Chihuahua, Dachshund, Doberman Pinscher, German Shepherd Dog, Keeshond, Miniature Schnauzer, Old English Sheepdog, Pomeranian, Rottweiler, Shih Tzu, Springer Spaniel, Toy Poodle, and Welsh Terrier. Mean and median body weight at the time of adrenalectomy was 19.5 kg (43.0 lb) and 19.7 kg (43.3 lb), respectively. Mean body condition score at the time of surgery was 5.4 (on a scale of 1 to 9). Age, sex, breed, body weight, and body condition score were not significantly associated with survival times.

Polyuria, polydipsia, alopecia, or hypertension was detected in 23 of 41 (56%) dogs, with a mean duration of 71 days (median, 15 days; range, 1 to 540 days) before initial examination at our facility. Detecting 1 or more of these conditions was not significantly associated with survival duration. Vomiting in 5 of 41 dogs for a mean duration of 4 days (median, 1 day; range, 1 to 10 days) and panting in 11 of 41 dogs were additional preoperative clinical signs not significantly associated with survival duration. Clinical signs consisting of weakness or lethargy (or both) were evident in 11 of 41 (26.8%) dogs at initial examination, with a mean duration of 62.1 days (median, 10 days; range, 1 to 455 days). Although weakness or lethargy (or both) were significantly (P = 0.013) associated with a shorter survival time on univariate analysis, the duration of these clinical signs did not have an effect on survival time (Table 1). Median survival time for dogs with weakness or lethargy was 138 days, compared with 908 days for dogs without these clinical signs. No dogs had hypotension at time of initial clinical examination.

Presurgical laboratory data significantly associated with a shorter survival time on univariate analysis included an increase in BUN concentration (P = 0.015), thrombocytopenia (P = 0.003), increase in PTT (P = 0.030), increase in AST activity (P = 0.039), and hypokalemia (P = 0.042).

View Full Size

Results of univariate analysis of categoric variables analyzed against survival time and a Cox proportional hazards analysis to identify relative risk for data obtained from medical records of 41 dogs that underwent adrenalectomy.

Citation: Journal of the American Veterinary Medical Association 232, 1; 10.2460/javma.232.1.77

• Download Figure
• Download figure as PowerPoint slide

The BUN concentration was obtained before adrenalectomy for 40 dogs. The BUN concentration was measured a mean of 25.9 days before surgery (median, 11 days; range, 0 to 197 days). The concentration was obtained before IV administration of fluids in all dogs, except for 1 dog that received a single bolus of fluids prior to referral to our facility. Mean BUN concentration of all dogs was 25.3 mg/dL (median, 18 mg/dL; range, 5 to 88 mg/dL). Nine of 40 (22.5%) dogs had a BUN concentration above the reference range; 4 of these 9 dogs were azotemic, and 5 of these 9 dogs died or were euthanized within 1 month after adrenalectomy (3 of them were azotemic after surgery). One dog lived 571 days after surgery, 1 dog was lost to follow-up monitoring, and 2 dogs were still alive at the end of the study. Multivariate analysis revealed that as the preoperative serum concentration of BUN increased, the risk of death also increased significantly (Cox proportional hazards, 1.075; P = 0.023; Table 2).

One of 40 dogs was thrombocytopenic. The dog was euthanized during surgery because of a nonresectable tumor. Platelet count for that dog was 111,000 cells/μL. Mean platelet count for all dogs was 341,260 cells/μL (median, 328,500 cells/μL; range, 111,000 to 600,000 cells/μL).

Two of 17 dogs had a significant increase in PTT. Mean PTT for the 17 dogs was 18.8 seconds (median, 12.2 seconds; range, 9.3 to 92.0 seconds). Both of the dogs with the significant increase in PTT died within 30 days after surgery, although neither had evidence of intraoperative hemorrhage. The first dog had a PTT of 92 seconds, developed acute renal failure, and died 3 days after adrenalectomy. The second dog (PTT of 82 seconds) was readmitted to the hospital 13 days after surgery because of lethargy, a decrease in appetite, and melena; it developed acute-on-chronic renal failure and adrenal gland insufficiency after surgery and died 30 days after adrenalectomy.

Activity of AST before surgery was available for 38 dogs. Mean AST activity was 62.1 U/L (median, 34.5 U/L; range, 3 to 709 U/L). Eleven of 38 dogs had an increase in AST activity on preoperative serum biochemical analysis. An increase in AST activity was significantly associated with a decrease in survival time on univariate analysis.

Serum concentration of potassium was recorded for 39 dogs. Mean potassium concentration for all dogs was 4.6 mEq/L (median, 4.7 mEq/L; range, 1.8 to 6.3 mEq/L). Four of 39 dogs were hypokalemic. Of these 4 dogs, 1 was euthanized during surgery because of uncontrollable bleeding, a second developed pancreatitis and acute renal failure and died 7 days after surgery, a third was euthanized 9 days after surgery because of deterioration in clinical condition and poor quality of life (this dog also had splenic hemangiosarcoma), and the fourth was euthanized 429 days after surgery because of regrowth of the adrenal gland tumor and recurrence of clinical signs. On multivariate analysis, dogs with a lower potassium concentration before adrenalectomy were almost 5 times as likely to die (Cox proportional hazards, 4.95; P = 0.005), compared with dogs that were not hypokalemic. No other abnormalities in preoperative CBC counts, serum biochemical analyses, or urinalyses were associated with survival duration.

An ACTH stimulation test was performed in 18 of 41 dogs. Cortisol concentration before ACTH stimulation was increased in 6 of the 18 dogs. The cortisol concentration after ACTH stimulation was increased in 9 of the 18 dogs. Endogenous ACTH concentration was measured in 10 of 41 dogs; 4 had concentrations below the reference range, 5 had concentrations within the reference range, and 1 had concentrations above the reference range. An LDDS test was performed in 12 dogs, and an HDDS test was performed in 4 dogs. For the LDDS test, cortisol concentrations were increased for 1 dog before, 11 dogs at 1 hour after, and 11 dogs at 8 hours after ACTH stimulation. For the HDDS test, none of the dogs had an increase in cortisol concentrations before ACTH stimulation, but cortisol concentrations were increased in 4 of 4 dogs at 4 hours after and 2 of 4 dogs at 8 hours after ACTH stimulation. The urine cortisol:urine creatinine ratio was determined for 8 of 41 dogs, and it was increased in 7 of those 8 dogs. Steroid hormone concentrations were available for 6 of 41 dogs. For 1 of the 6 dogs, only the 17-hydroxy-progesterone concentration after ACTH stimulation was reported. Steroid hormone concentrations were increased after ACTH in 2 of the 5 remaining dogs. Androstenedione concentrations after ACTH stimulation were increased in 1 of 4 dogs and not recorded in 1 dog. The 17-hydroxyprogesterone concentration after ACTH stimulation was low in 1 of the 5 and increased in 2 of the 5 dogs. Abnormal results for the ACTH stimulation, LDDS, or HDDS tests or abnormalities in concentrations of endogenous ACTH, other steroid hormones, or the urine cortisol:urine creatinine ratio were not significantly associated with survival duration.

Five of 41 dogs were treated for adrenal gland disease prior to adrenalectomy. Three of these dogs were treated with mitotane,^a and the other 2 dogs were treated with phenoxybenzamine because of a suspected pheochromocytoma. Mitotane was administered at 50 to 118 mg/kg/wk (22.7 to 53.6 mg/lb/wk), with the duration of treatment ranging from 12 to 150 days. Phenoxybenzamine was administered at 0.4 to 0.6 mg/kg (0.18 to 0.27 mg/lb), q 12 h, with the duration of presurgical treatment ranging from 2 to 12 days. Two dogs treated with phenoxybenzamine for a suspected pheochromocytoma had adenocarcinoma as the histologic tumor type. Medical treatment for adrenal gland disease prior to adrenalectomy was not significantly associated with survival duration.

View Full Size

Results of multivariate analysis by use of a Cox proportional hazards analysis to identify relative risk for data obtained from medical records of 41 dogs that underwent adrenalectomy.

Citation: Journal of the American Veterinary Medical Association 232, 1; 10.2460/javma.232.1.77

• Download Figure
• Download figure as PowerPoint slide

Four of 41 adrenal gland tumors were considered an incidental finding and not the primary reason for surgery. Only unilateral adrenalectomies were performed, 26 (63.4%) of which were on the left side. Twenty-two (53.7%) dogs had gross lesions within other organs during surgery; however, only 2 of those 22 (9.1%) dogs had histologic evidence of metastasis. Tumor thrombi within the phrenicoabdominal vessels, renal vessels, or vena cava were detected during surgery in 11 dogs. The vessels and associated thrombi were removed along with the tumor, or the thrombus was removed via venotomy. Two of 41 tumors were considered nonresectable at surgery (1 had evidence of an extensive tumor thrombus within the vena cava, and the other was in a dog with uncontrollable intraoperative hemorrhage); both dogs were euthanized during surgery. Intraoperative hypotension was detected in 14 dogs, whereas hypertension was detected in only 4. Three of these dogs had a combination of hypotension and hypertension during anesthesia. Of the dogs that had only hypotension, 6 had an adrenocortical adenocarcinoma, 4 had an adenoma, and 1 had adrenocortical hyperplasia. The 1 dog that had only hypertension had an adrenal adenoma. Two of the 3 dogs with both hypotension and hypertension had a pheochromocytoma, whereas the third dog had an adenoma. Whether the tumor was a primary tumor or incidental finding, side of adrenalectomy, evidence of gross or histologic metastasis, tumor thrombi, whether the tumor was considered resectable, and intraoperative hypotension or hypertension (or both) were not significantly associated with survival duration.

Two of 41 dogs had intraoperative hemorrhage as a complication. One dog was euthanized during surgery because of uncontrollable bleeding. The other dog received a blood transfusion during surgery but was euthanized 2 days after surgery because of development of hemoabdomen. The tumor type in both dogs with intraoperative hemorrhage was adrenocortical carcinoma. Intraoperative hemorrhage was associated with a significantly (P < 0.001) shorter survival time on univariate analysis. Median survival time for dogs that had intraoperative hemorrhage was 1 day.

A second intraoperative variable significantly (P = 0.007) associated with a shorter survival time on univariate analysis was concurrent nephrectomy during adrenalectomy. Four dogs underwent concurrent nephrectomy at the time of adrenalectomy because of the mass extending into the renal vein or parenchyma. Three of these 4 dogs died or were euthanized within 72 hours after surgery (1 dog developed anuric renal failure after surgery and was euthanized 3 days after surgery, a second dog developed hemoabdomen 2 hours after surgery and was subsequently euthanized, and a third dog developed hemoabdomen and had seizures 1 day after surgery and was euthanized 2 days after surgery). The fourth dog lived for 926 days and was subsequently lost to follow-up monitoring. Of the tumors in dogs with concurrent nephrectomy, 3 were an adrenocortical carcinoma, and 1 was a pheochromocytoma. Dogs in which a concurrent nephrectomy was performed during surgery were 4.10 times as likely to die, compared with dogs in which only adrenalectomy was performed (Cox proportional hazards, 4.098; P = 0.012), on the basis of univariate analysis. For multivariate analysis, concurrent nephrectomy significantly (P = 0.018) increased the risk for death, with dogs 142.9 times as likely to die, compared with the risk for dogs that had only an adrenalectomy (95% confidence interval, 2.4 to 8,547.0; Table 2). Median survival time of dogs that underwent concurrent nephrectomy was 2 days.

Histologically, most tumors were identified as an adrenocortical carcinoma (n = 19) or adrenocortical adenoma (14). Other masses were histologically identified as a pheochromocytoma (n = 4), adrenocortical nodular hyperplasia (2), and leiomyosarcoma (1). There were no histologic results available for the tumor in 1 dog. Median survival times for each tumor type were 230 days for an adrenocortical carcinoma, 374 days for pheochromocytoma, and 687.5 days for adrenocortical adenoma. The dog with a leiomyosarcoma survived 138 days. Median survival time for dogs with adrenocortical hyperplasia was 200.5 days. Completeness of surgical excision was reported for 26 dogs. Twenty-two of 26 (84.6%) margins were considered free of tumor cells, and 4 margins were considered incomplete. Microscopic evidence of vascular invasion was recorded for 30 dogs. Tumor type, completeness of surgical excision, and microscopic evidence of vascular invasion were not associated with survival time. Gross lesions including organ enlargement, nodules, or discoloration were evident in organs other than the adrenal glands in 22 of 40 (55.0%) dogs. Biopsy specimens were obtained during surgery from all other organs that were grossly abnormal; however, only 2 of these 22 dogs had histologic evidence of metastasis. Histologic lesions in other organs included hepatic vacuolar degeneration (n = 18), splenic nodular hyperplasia (3), omental metastasis (2), splenic hemangiosarcoma (2), endometritis (2), pancreatic nodular hyperplasia (2), and interstitial nephritis (2) and 1 each of jejunal polypoid adenoma, splenic infarction, biliary cyst adenoma, splenic myelolipoma, splenic fibrohistiocytic nodule, hepatocellular carcinoma, lymphoid hyperplasia, hepatocellular necrosis, hepatocellular glycogenosis, pyelonephritis, and renal glomerulosclerosis. Neither detection of gross lesions at the time of surgery nor histologic evidence of metastasis was significantly associated with survival time.

Tumor volume size and tumor volume score were calculated for 39 of 41 dogs. Mean and median tumor volume size were 8.45 and 7.90 cm (3.35 and 3.59 inches), respectively, whereas mean and median tumor volume score were 29.97 and 15.61 cm³ (13.62 and 7.10 inches³), respectively. Tumor volume size and tumor volume score were not significantly associated with survival time.

Thirty-three of 37 (89.1%) dogs were treated with a steroid hormone after adrenalectomy; the other 4 dogs were euthanized or died during or immediately after surgery before any steroid hormones were administered. Steroidal treatments were not standardized, and type of steroid hormone, dose, and frequency of treatment varied among patients. A postoperative ACTH stimulation test was performed in 23 of 37 (62.2%) dogs within 1 to 49 days (median, 8 days) after adrenalectomy. Results were within the reference range for 7 dogs, above the reference range for 2 dogs, and blunted for 12 dogs. Postoperative treatment with steroid hormones or evidence of adrenal gland insufficiency was not significantly associated with survival duration.

Overall, postoperative complications developed in 13 of 37 (35.1%) dogs that survived surgery. Complications were development of adrenal gland insufficiency in 9 dogs, acute renal failure in 5 dogs, pancreatitis in 3 dogs, and septic peritonitis in 1 dog. Five of these dogs developed more than 1 complication. The three dogs that developed pancreatitis also developed acute renal failure. One dog each that developed peritonitis and acute renal failure also developed adrenal gland insufficiency. Pulmonary thromboembolism was not confirmed in any dogs in the study; however, necropsy was only performed in 3 dogs. Postoperative complications significantly (P < 0.001) associated with a shorter survival time on univariate analysis included development of pancreatitis and acute renal failure. Tumor volume size and tumor volume score were calculated for 2 of the 3 dogs that developed pancreatitis. Tumor scores for both of these dogs were higher (but not significantly different) from the median scores for all dogs. Dogs that developed acute renal failure after adrenalectomy were 62.5 times as likely to die as dogs that did not develop acute renal failure (Cox proportional hazards, 62.5; P = 0.016). Of the 5 dogs that developed acute renal failure after surgery, 3 had an increase in BUN concentration in preoperative serum biochemical analysis. Median survival time was 7 days for dogs that developed pancreatitis or acute renal failure, compared with 908 days for dogs that did not develop pancreatitis or acute renal failure. The dog that developed septic peritonitis 6 days after adrenalectomy underwent another surgery, and a duodenal perforation was diagnosed and surgically corrected. That dog was still alive at the conclusion of the study (> 735 days after surgery).

View Full Size

Kaplan-Meier life table analysis for overall median survival time of 41 dogs undergoing adrenalectomy. Median survival time for all dogs was 690 days. Day of adrenalectomy was designated as day 0.

Citation: Journal of the American Veterinary Medical Association 232, 1; 10.2460/javma.232.1.77

• Download Figure
• Download figure as PowerPoint slide

Nine (22.0%) dogs died or were euthanized prior to discharge from the hospital. The intraoperative mortality rate was 4.8% (2/41 dogs). Five dogs were lost to follow-up monitoring at 2, 13, 13, 48, and 728 days. Overall Kaplan-Meier median survival time was 690 days (Figure 1). Eleven dogs were still alive at the conclusion of the study.

Discussion

Similar to findings in another study,¹ dogs in the study reported here that survived the postoperative period after adrenalectomy had long survival times. However, the short-term mortality rate is high.^1,5–7 In our study, clinical signs of weakness and lethargy were only evident in 11 of 41 (26.8%) dogs but were associated with a shorter survival time. Although these clinical signs are nonspecific, they may indicate greater severity of systemic effects for an adrenal gland tumor. Weakness and lethargy may be clinical signs associated with adrenocortical or medullary tumors. Glucocorticoids can cause a profound effect on skeletal muscle. The catabolic effects may cause substantial muscle atrophy and weakness by inhibiting myofibrillar proteins, which are mainly in type 2 muscle fibers.⁹ Episodic release of catecholamines with a pheochromocytoma can lead to cardiac arrhythmias, hypertension, weakness, and collapse.¹⁰ Local invasion into the surrounding tissues with adrenocortical or medullary tumors can cause weakness by obstruction of the caudal vena cava or via intra-abdominal hemorrhage.²

Preoperative increase in BUN concentration was significantly negatively associated with survival duration. An increase in BUN concentration may be associated with dehydration, gastrointestinal tract bleeding, heart failure, shock, or renal failure. Three of 9 dogs that had an increase in BUN concentration before adrenalectomy developed azotemia after surgery. Although it is possible that other disease processes that can cause an increase in BUN concentration may contribute to a poor outcome, attempting to use diuresis to lower BUN concentrations to the reference range prior to surgery may help improve outcome; however, additional studies are warranted to evaluate this hypothesis.

Dogs with thrombocytopenia or an increase in PTT had a significantly shorter survival time. Although the 2 dogs with an increase in PTT did not have intraoperative hemorrhage, it is possible that they had postoperative abdominal bleeding, but it was not identified. Both of these patients died within 30 days after surgery, and neither had evidence of coagulopathy at the time of death. The 1 dog with thrombocytopenia was euthanized during surgery because of a nonresectable tumor attributable to an extensive caval thrombus. It is possible that this dog had thrombocytopenia as a result of an increase in peripheral consumption as a result of the caval thrombus. Another surgeon may have attempted surgical removal of the tumor in this dog. Because of the small number of patients with an abnormal PTT and thrombocytopenia in the study reported here, it is difficult to comment on the importance of these results. Coagulopathy may also cause hemorrhage in other body systems, which may not be perceived as life-threatening events. For example, melena was observed after surgery in one of the dogs with an increase in PTT. Disseminated intravascular coagulopathy should also be considered a possibility; however, necropsies were not permitted in these dogs.

In dogs, AST primarily originates from hepatic and skeletal muscle cells; however, it can also be found in renal and gastrointestinal tract cells.¹¹ Therefore, it is difficult to fully assess the clinical relevance of AST activity and a correlated negative effect on survival time. Dogs with muscle wasting may have increases in AST activity, which may indicate greater chronicity of the disease process or a negative energy balance.

In addition, lower potassium concentrations were significantly associated with a shorter survival time. Mechanisms for hypokalemia associated with adrenal gland disease include renal loss, decreased dietary intake, vomiting, metabolic alkalosis, and aldosterone secretion. Although aldosterone was not measured in any of the dogs in this study, it can be speculated that any tumor of the adrenal cortex could also cause an increase in aldosterone concentrations. Classic signs of hyperaldosteronism include lethargy, weakness, mild hypernatremia, hypokalemia, and systemic hypertension.¹² It is possible that dogs with weakness or lethargy, hypertension, or hypokalemia had increased concentrations of aldosterone; however, these data were not available because of the retrospective nature of the study.

On the basis of results of preoperative endocrine testing, survival time for dogs with functional tumors was the same as for those with nonfunctional tumors. In humans, incidental adrenal gland masses are managed differently, depending on the size and morphologic characteristics. Adenomas can be differentiated from nonadenomas by use of computed tomography and magnetic resonance imaging on the basis of intralesional fat content.¹³ It has been recommended in humans to remove adrenal gland masses > 4 cm (> 1.59 inches) in diameter because larger tumors are typically malig-nant.¹³ Depending on a person's history of cancer, patients with suspected benign adrenal gland masses may benefit from intermittent surveillance with advanced imaging techniques. The decision to biopsy or surgically remove the mass is then based on the progression from the preceding images.¹³ In dogs, mean survival time for dogs receiving medical treatment for cortisol-secreting adrenocortical neoplasms is 16.4 months.¹⁴ Additional studies are warranted to evaluate nonsurgical treatment of dogs with small, nonsecreting adrenal gland masses.

Similar to results of another study,⁶ side of adrenalectomy and the detection of thrombi during surgery were not associated with survival duration. Intraoperative hypotension or hypertension (or both) was also not associated with a difference in survival time. It is important to mention that detection of gross lesions in other organs during surgery was high; however, histologic evidence of metastasis was low. Therefore, euthanasia should not be recommended solely on the basis of gross lesions detected during surgery.

Dogs with adrenal gland masses that require concurrent nephrectomy for tumor resection have a guarded prognosis. In addition, hemorrhage during surgery significantly increased the risk for death. Hemorrhage during surgery or concurrent nephrectomy may be associated with larger, more invasive tumors. Preoperative diagnostic testing and staging varied depending on patient status, whether the adrenal gland tumor was an incidental finding, and suspicion of tumor type. Because the surgical outcome after adrenalectomy can be unpredictable, the use of improved preoperative imaging modalities, such as computed tomography or magnetic resonance imaging, may be of value in more accurately determining tumor size, invasiveness, and proximity to the kidneys and potential need for nephrectomy. This may be used to predict the need for nephrectomy or whether a tumor is nonresectable prior to surgery. Results from advanced imaging modalities may also indicate the need for excessive dissection or a close proximity to the pancreas, which would warrant concern for postoperative pancreatitis.

Postoperative pancreatitis and acute renal failure were associated with a severe decrease in survival time. Dogs with pancreatitis all died within 10 days after surgery. The incidence of postoperative pancreatitis after adrenalectomy in veterinary medicine varies from 2.5% to 11.1%.^1,6,7 It is interesting that the 3 dogs that developed pancreatitis also developed acute renal failure. Again, this may have been indicative of more invasive tumors that required excessive dissection during adrenalectomy and resulted in iatrogenic damage to the pancreas. Two of the 3 dogs that developed pancreatitis had right-sided adrenalectomies. An increase in manipulation of the pancreas during right-sided adrenalectomy may increase the incidence of pancreatitis. Dogs with hyperadrenocorticism and prolonged anesthesia or hypotension during surgery may be at increased risk for thromboembolism.¹² The increased risk of thromboembolism, in addition to manipulation of the pancreas during surgery, may further increase the risk of developing postoperative pancreatitis. Additional studies are required to determine whether rapid identification of these complications and subsequent treatment may lead to an improvement in outcome.

Pulmonary thromboembolism was not confirmed as a postoperative complication in any dogs in the study reported here. It is possible that dogs in our study had pulmonary thromboembolism but did not die; however, without confirmation during necropsy, we believed the diagnosis would be presumptive. Because of the small number of dogs on which a postmortem examination was performed, some dogs that had pulmonary thromboembolism after surgery may not have been detected.

Many causes can contribute to the development of acute renal failure after adrenalectomy. Urinary tract infections are common with hyperadrenocorticism; such infections can ascend to the kidneys. In addition, hyperadrenocorticism can lower resistance to infection as a result of glucocorticoid-induced inhibition of neutrophils and macrophages.¹² Calcium excretion is also increased with glucocorticoid excess, possibly resulting in calculi and an increase in the risk for infection.¹² Hypotension during surgery or after tumor removal may also contribute to decreased renal perfusion. Adrenal gland insufficiency after surgery affects both glucocorticoid and mineralocorticoid secretion. Glucocorticoid depletion impairs renal excretion of water, whereas the effects of hypoaldosteronism include hypotension, impaired cardiac output, and impaired renal perfusion.¹⁵ Sodium and potassium imbalances that develop with adrenal gland insufficiency result in decreased glomerular filtration rates.¹⁶ In the study reported here, dogs that developed renal failure after adrenalectomy had a significantly shorter survival time.

Similar to results reported in another study,¹ we did not detect an association between survival time and histologic diagnosis. In humans, it is routine to obtain a biopsy specimen or fine-needle aspirate via computed tomographic or ultrasonographic guidance.¹⁷ A histologic or cytologic examination may assist in the decision to move forward with adrenalectomy. However, attempts at presurgical diagnosis are not often performed in dogs because of the potential risks, such as hemorrhage, which can be fatal.¹⁸ On the basis of the results for the study reported here and for other studies, dogs that survive the immediate postoperative period after adrenalectomy have a good prognosis for long-term survival, regardless of the histologic diagnosis. Therefore, we do not recommend attempts be made (such as via fine-needle aspiration) to establish a diagnosis before surgery.

The main limitation of the study was that it was retrospective in nature. When follow-up information was not included in the records, it was obtained (when possible) by contacting the referring veterinarians or the owners. Although it appears that we had an adequate number of dogs, the same diagnostic tests were not performed in all dogs, and some abnormalities were only detected in a small number of dogs for which the clinical importance must be questioned.

Dogs surviving to discharge from the hospital after adrenalectomy had long survival times; however, there was a high short-term mortality rate. Several preoperative factors associated with a shorter survival time were weakness or lethargy, thrombocytopenia, an increase in BUN concentration, increase in PTT, increase in AST activity, and hypokalemia. Additional studies are needed to evaluate how treatment designed to control or alleviate these factors may affect or change the outcome after adrenalectomy.