Predictive Factors and the Effect of Phenoxybenzamine on Outcome in Dogs Undergoing Adrenalectomy for Pheochromocytoma
Corresponding author: Melissa A. Herrera, Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, CA 95616; e-mail: email@example.com.
Background: Some studies in dogs undergoing adrenalectomy for pheochromocytoma suggest that anesthetic complications and perioperative mortality are common. In humans, surgical outcome has improved with the use of phenoxybenzamine (PBZ) before adrenalectomy.
Hypothesis: Dogs treated with PBZ before adrenalectomy have increased survival compared with untreated dogs.
Animals: Forty-eight dogs that underwent adrenalectomy for pheochromocytoma.
Methods: A retrospective medical record review for dogs that underwent adrenalectomy for pheochromocytoma at a veterinary medical teaching hospital over the period from January 1986 through December 2005.
Results: Twenty-three of 48 dogs were pretreated with PBZ (median dosage: 0.6 mg/kg PO q12h) for a median duration of 20 days before adrenalectomy. Duration of anesthesia and surgery, percentage of dogs with pheochromocytoma involving the right versus left adrenal gland, size of tumor, and presence of vascular invasion were similar for PBZ-treated and untreated dogs. Thirty-three (69%) of 48 dogs survived adrenalectomy in the perioperative period. PBZ-treated dogs had a significantly (P= .014) decreased mortality rate compared with untreated dogs (13 versus 48%, respectively). Additional significant prognostic factors for improved survival included younger age (P= .028), lack of intraoperative arrhythmias (P= .0075), and decreased surgical time (P= .0089).
Conclusions and Clinical Importance: Results from this retrospective study support treatment with PBZ before surgical removal of pheochromocytoma in dogs.
Pheochromocytoma is a functional, often malignant, tumor of the adrenal medullary chromaffin cells.1–5 Clinical signs, when observed, are assumed to be the result of excessive secretion of catecholamines, invasion of the tumor into adjacent blood vessels or organs, or hemorrhage from the tumor into the retroperitoneal space and include weakness, panting, tachypnea, collapse, tachyarrhythmias, and seizures.1,2,4 Identification of an adrenal mass with abdominal ultrasound often is the first indication that a dog may have a pheochromocytoma. Historical complaints, findings on physical examination, results of routine blood and urine tests, and ruling out adrenal-dependent hyperadrenocorticism, hyperaldosteronism, and metastatic disease are helpful in ruling out other causes of an adrenal mass or for the clinical presentation. Increased concentrations of catecholamines or catecholamine metabolites in urine support the diagnosis of pheochromocytoma in humans6–11 but such tests are not readily available in dogs.12 The definitive diagnosis of pheochromocytoma relies on histologic evaluation of the adrenal mass.
Adrenalectomy is the treatment of choice for pheochromocytoma in humans.13 Studies evaluating the efficacy of adrenalectomy for pheochromocytoma in dogs are limited but suggest that anesthetic complications and perioperative mortality are common.1,14,15 Phenoxybenzaminea (PBZ) is an α-adrenergic antagonist that irreversibly binds to both α-1 and α-2 adrenergic receptors and blocks the α-adrenergic response to circulating epinephrine and norepinephrine.16 In humans with pheochromocytoma, PBZ typically is administered for days to weeks before adrenalectomy to minimize the deleterious effects of excessive catecholamine secretion that occur during the perioperative period.13
The role of PBZ in minimizing perioperative complications in dogs has not been reported. Pretreatment α-blockade of dogs undergoing adrenalectomy for pheochromocytoma was not done at our hospital until the mid 1990s. The purpose of this study was to assess the efficacy of preoperative PBZ and identify predictive factors for survival in dogs undergoing adrenalectomy for pheochromocytoma. The hypothesis was that dogs treated with PBZ before adrenalectomy would have a better outcome than dogs not treated with PBZ.
Materials and Methods
Criteria for Selection of Cases
Medical records of all dogs with a histologic diagnosis of pheochromocytoma at the Veterinary Medical Teaching Hospital (VMTH), University of California, Davis from 1986 to 2005 were reviewed. For inclusion, dogs must have undergone adrenalectomy at the VMTH. Data collected from the medical records before adrenalectomy included signalment, year of surgery, and PBZ treatment (dose and duration). Additional data collected from the medical record, when available, included medications, presence of hyperadrenocorticism, and concurrent diseases including neoplasia. Hyperadrenocorticism was ruled out based on absence of appropriate clinical signs and results of a low-dose dexamethasone suppression test, ACTH stimulation test, urine cortisol : creatinine ratio, or a combination of these.
Day-of-surgery preanesthetic medications included oxymorphone, methadone, morphine, acepromazine, and glycopyrrolate or atropine, administered SC or IV. General anesthesia was induced with a combination of opioids and benzodiazepines (eg, fentanyl, oxymorphone, or sufentanil with diazepam or midazolam), or etomidate or propofol, IV, and was maintained with a balanced opioid-inhalant technique that was similar for all dogs.
A ventral midline laparotomy was performed in each dog. Tumor side and presence or absence of vascular invasion was recorded. In dogs without vascular invasion or with vascular invasion confined to the phrenicoabdominal vein, standard adrenalectomy was performed and the thrombus was removed by ligation of the phrenicoabdominal vein distally and at its junction with the caudal vena cava. In dogs requiring venotomy owing to tumor invasion into the caudal vena cava, complete or partial temporary occlusion of the caudal vena cava was used for thrombus removal. If the thrombus was small, temporary partial occlusion was accomplished with a Satinsky clamp and the thrombus removed with a caval venotomy followed by suture closure. If the thrombus was large and required complete occlusion of the vena cava, Rummel tourniquets were placed cranial and caudal to the adrenal gland and tightened for a brief period of time.
The majority of dogs underwent deliberate hypothermia (body temperature < 34 °C) because of the likely requirement of inflow occlusion during thrombus removal. Deliberate hypothermia may protect vital organs from reduced oxygen delivery during temporary occlusion of the caudal vena cava to remove the thrombus.17,18 With temporary complete occlusion of the caudal vena cava, a caval venotomy was performed and the thrombus removed. To minimize the time of complete occlusion, a Satinsky clamp was used to secure the venotomy site after thrombus removal and the Rummel tourniquets were released before suture closure of the venotomy. Suture closure of the caval venotomy varied depending on surgeon preference and size of patient, but most often involved a simple continuous suture pattern with 5-0 silk. After removal of the adrenal gland, the surgical site was carefully examined for hemorrhage. After surgery, each dog was treated with analgesic drugs, IV administration of fluids, and IV blood products, as needed.
Blood Pressure Monitoring
Hypertension was defined as a systolic blood pressure (SBP) ≥ 180 mmHg; hypotension was defined as a SBP ≤ 80 mmHg. For each dog, the highest blood pressure, lowest blood pressure, number of hypertensive episodes, and numbers of hypotensive episodes were recorded intraoperatively and postoperatively. For each dog, spikes in SBP during the intraoperative and postoperative period were calculated by subtracting the lowest recorded blood pressure from the highest recorded blood pressure.
Intraoperative direct arterial blood pressure was measured with a pressure transducerb applied to an arterial catheter. Blood pressures were assessed continuously and recorded every 5 minutes. Postoperative blood pressure was measured by either a direct method with a pressure transducer or via an indirect oscillometric method.c Postoperatively, blood pressure was recorded every 15 minutes to 4 hours for dogs that were monitored in the intensive care unit (ICU) after adrenalectomy. For most of these dogs, blood pressure was recorded every hour. Blood pressure was not measured in dogs returned to the wards after recovery from anesthesia.
A continuous, standard, lead II ECG was utilized intraoperatively in all dogs and postoperatively for dogs monitored in the ICU. Hard copies of ECG tracings were examined when present in the medical record and the occurrence of cardiac arrhythmias was also obtained from the anesthesia report and the medical record.
Data evaluated, when available, included tumor size and location, presence or absence of vascular invasion, duration of anesthesia and surgery, blood pressure recordings, complications such as systemic hypertension, cardiac arrhythmias, or hemorrhage, and medications administered. Intra- and postoperative deaths were recorded. Nonsurvivors were arbitrarily defined as dogs that died during surgery or within 10 days after adrenalectomy. Outcome was defined as the ability to predict survival.
The adrenal gland containing the tumor was measured at the time of surgical removal, after submission to the pathologist, or both. Tumor size was arbitrarily defined as small (≤ 2.5 cm in maximum width), medium (2.6–4.9 cm), and large (≥ 5.0 cm). Involvement of the right versus left adrenal gland and presence of vascular invasion were recorded. The diagnosis of pheochromocytoma was based on both histologic classification and positive staining of the tumor for chromogranin A.19
Univariate logistic regression analyses were used to identify predictors of 10-day postoperative survival in all dogs. The variables evaluated included signalment, year of surgery, tumor size and location, presence of vascular invasion, pretreatment with PBZ, anesthesia and surgical time, intraoperative and postoperative hypertension, frequency of hypertensive episodes, intraoperative and postoperative hypotension, including the frequency of hypotensive episodes, intraoperative arrhythmias and hemorrhage, and intraoperative medications administered to each dog. When evaluating the effect of the surgical date on outcome, the date of adrenalectomy was arranged chronologically into 4 groups with an equal number of dogs in each group. Not all dogs had complete information in their medical records.
A Mann-Whitney test was used to compare the continuous variables between the 2 treatment groups (PBZ treated versus untreated) of dogs. Categorical variables were compared by a χ2 test of homogeneity. A statistical software program was used.d A P-value of <.05 was considered significant. Data are presented as median and range unless indicated otherwise.
From 1986 to 2005, 207 dogs were diagnosed with pheochromocytoma, and adrenalectomy was performed in 48. Median age of the 48 dogs at the time of surgery was 11 years (range, 7–16 years), 67% were male or male castrated, and 29 different breeds were represented. The breeds represented at least twice included mixed breed dogs (n = 10), Doberman Pincher (3), Miniature Schnauzer (3), Golden Retriever (2), Rottweiler (2), Standard Poodle (2), Dachshund (2), Boxer (2), and Basset Hound (2). Median body weight was 22.5 kg (range, 9–62 kg). Adrenalectomy was performed in 12 dogs during each of 4 time periods: 1986–1993, 1994–2000, 2001–2003, and 2004–2005.
Twenty-three (48%) of 48 dogs were treated with PBZ for a median of 20 days (range, 7–120 days) before adrenalectomy. The median dosage of PBZ was 0.6 mg/kg (range, 0.1–2.5 mg/kg) administered every 12 hours. One dog was treated with PBZ during the initial time period (1986–1993), 6 dogs were treated during 1994–2000, 7 dogs during 2001–2003, and 9 dogs during 2004–2005. Concurrent medications administered to PBZ dogs at the time of surgery included levothyroxine (3 dogs), aspirin (2), atenolol (1), pancreazyme (1), diphenhydramine (1), omeprazole (1), prednisone (1), insulin (1), cephalexin (1), and amlodipine (1) and concurrent medications administered to the untreated dogs included deracoxib (1), cephalexin (1), carprofen (1), amlodipine (1), amoxicillin-clavulanic acid (1), and levothyroxine (1). Tests for hyperadrenocorticism were performed in 15 PBZ and 9 untreated dogs. One PBZ dog and 3 untreated dogs were diagnosed with pituitary-dependent hyperadrenocorticism (PDH) and 1 untreated dog was diagnosed with hyperadrenocorticism of uncertain etiology. Concurrent neoplasia was identified in 2 PBZ dogs (mast cell neoplasia, soft tissue sarcoma) and 2 untreated dogs (mast cell neoplasia, thyroid tumor). Concurrent disease included hypothyroidism (3), cystic calculi (1), myelopathy (1), and diabetes mellitus (1) in PBZ dogs and osteoarthritis (3), hypothyroidism (2), diabetes mellitus (1), and atrial-ventricular block (1) in untreated dogs. Thoracic radiographs were unremarkable in 19 PBZ and 17 untreated dogs in which they were obtained before adrenalectomy. Additional findings on abdominal ultrasound performed in 21 PBZ dogs included splenic (2 dogs) and liver masses (1) and in 20 untreated dogs a kidney mass (1).
There was no significant (P= .34) difference in age between the PBZ and untreated dogs but there was a significantly (P= .036) higher number of males in the PBZ group (19 versus 13, respectively). There was no significant difference in anesthesia time (median, 5.7 versus 5.2 hours; range, 2–8 versus 2.5–9 hours) or surgical time (median, 3.1 versus 3.3 hours; range, 2–5 versus 2–6 hours) between PBZ and untreated dogs, respectively. The location of the pheochromocytoma was available in 23 PBZ and 19 untreated dogs and involved the right adrenal gland in 14 (61%) and 12 (63%) of the PBZ and untreated dogs, respectively. Size of the adrenal tumor was available in 20 PBZ and 18 untreated dogs and was classified as small, medium, and large in 2, 10, and 8 PBZ dogs and 3, 6, and 9 untreated dogs, respectively. Tumor invasion of the vasculature was available in 38 dogs. Thirty-one dogs (82%) had vascular invasion, including 16 PBZ dogs. Vascular invasion was confined to the phrenicoabdominal vein in 3 dogs (2 PBZ dogs), extended into the caudal vena cava in 27 dogs (14 PBZ dogs), and extended into the renal vein in 1 untreated dog. There was no significant difference in size or location of the pheochromocytoma or presence and location of vascular invasion by the tumor between PBZ and untreated dogs (Table 1).
Table 1. Signalment, surgical findings, intraoperative complications, and surgical outcome in PBZ-treated and untreated dogs.
|Median age (years)||10||11||.34|
| Female||4||12|| |
| 1994–2000||6||5|| |
| 2001–2003||7||6|| |
| 2004–2005||9||3|| |
| Medium||10||6|| |
| Large||8||9|| |
| PA||2||1|| |
| Renal||0||1|| |
| Left||9||7|| |
|Median surgical time (hours)||3.1||3.5||.49|
|Median anesthesia time (hours)||6||5||.30|
|Intraoperative hemorrhage|| || ||.49|
| Yes||7||4|| |
| No||16||18|| |
|Intraoperative arrhythmias|| || ||1.0|
| Yes||7||6|| |
| No||14||14|| |
|Intraoperative hypotensionb|| || ||1.0|
| Yes||11||11|| |
| No||11||9|| |
|Intraoperative hypertensionc|| || ||.54|
| Yes||12||13|| |
| No||10||7|| |
|10-day survival|| || ||.013*|
| Yes||20||13|| |
| No||3||12|| |
Complete anesthetic records were available in 42 dogs. The most common intraoperative complications were hypertension (12 of 22 PBZ dogs, 13 of 20 untreated dogs) and hypotension (11 of 22 PBZ dogs, 11 of 20 untreated dogs; Table 1). The severity of intraoperative hypertension and hypotension and the number of hypertensive and hypotensive episodes was similar for the PBZ and untreated dogs (Table 2). Additionally, there was no significant difference in blood pressure variability between the 2 groups. Intraoperative arrhythmias were identified in 7 PBZ and 6 untreated dogs and included atrial premature contractions (4 dogs), ventricular premature contractions (6), ventricular tachycardia (2), supraventricular tachycardia (2), ventricular fibrillation (1), and not described (1). Seven of the 13 dogs (1 PBZ, 6 untreated) with arrhythmias died in the perioperative period; 3 deaths (1 PBZ, 2 untreated) were attributed to fatal arrhythmias. Intraoperative hemorrhage occurred in 7 PBZ and 4 untreated dogs. There was no significant difference in occurrence of intraoperative arrhythmias or hemorrhage between PBZ and untreated dogs (Table 1).
Table 2. Intraoperative blood pressure characteristics for 22 PBZ-treated and 20 untreated dogs.
|Maximum BP (mmHg) (all dogs)|
| Range||130–334||120–320|| |
|Number of hypertensive dogs||12 of 22||13 of 20||—|
|Maximum BP (mmHg) of hypertensive dogs|
| Range||180–334||180–320|| |
|Number of hypertensive episodes|
| Range||1–13||1–11|| |
|Minimum BP (mmHg) (all dogs)|
| Range||25–110||50–120|| |
|Number of hypotensive dogs||11 of 22||11 of 19||—|
|Minimum BP (mmHg) of hypotensive dogs|
| Range||25–80||50–80|| |
|Number of hypotensive episodes|
| range||1–30||1–22|| |
|BP variability (mmHg)|
| Range||50–255||50–250|| |
Information on intraoperative medications and treatment were available in 22 PBZ and 19 untreated dogs and were similar between the 2 treatment groups. The most common intraoperative medications included phentolamine for hypertension (7 PBZ and 6 untreated dogs), dopamine for hypotension (7 PBZ and 11 untreated dogs), a colloid (Hetastarch, packed red blood cells, whole blood, or plasma) for oncotic support, hemorrhage, or both (8 PBZ and 15 untreated dogs), and magnesium sulfate for hypertension (7 PBZ and 1 untreated dog).
Deliberate hypothermia was utilized in 32 dogs (19 PBZ, 13 untreated). Twenty-four of the dogs that underwent deliberate hypothermia survived (16 PBZ, 8 untreated). Deliberate hypothermia did not have a negative effect on outcome. It could not be determined whether there was a positive effect on outcome because all dogs were cooled if inflow occlusion was required.
Postoperative blood pressure measurements were available on 37 dogs (6 had died or were euthanized in surgery, 2 were not sent to the ICU, and 3 had missing data). Among the 37 dogs, the most common blood pressure abnormality was hypertension (9 of 20 PBZ dogs, 12 of 17 untreated dogs) followed by hypotension (5 of 20 PBZ dogs, 3 of 17 untreated dogs; Table 3). The severity of postoperative hypertension and hypotension and the number of hypertensive and hypotensive episodes was similar for PBZ and untreated dogs. Additionally, there was no significant difference in blood pressure variability between the 2 groups. Postoperative hypertension (P= .39) and hypotension (P= .27) were not significantly associated with 10-day survival.
Table 3. Postoperative blood pressure characteristics for 20 PBZ-treated and 17 untreated dogs.
|Maximum BP (mmHg) (all dogs)|
| Range||120–270||137–309|| |
|Number of hypertensive dogs||9 of 20||12 of 17||—|
|Maximum BP (mmHg) of hypertensive dogs|
| Range||180–270||190–309|| |
|Number of hypertensive episodes|
| Range||1–65||1–111|| |
|Minimum BP (mmHg) (all dogs)|
| Range||65–170||70–180|| |
|Number of hypotensive dogs||5 of 20||3 of 17||—|
|Minimum BP (mmHg) of hypotensive dogs|
| Range||65–80||70–75|| |
|Number of hypotensive episodes|
| Range||1–2||1–4|| |
|BP variability (mmHg)|
| Range||10–135||7–192|| |
Three (13%) of 23 PBZ dogs were nonsurvivors. All 3 dogs died in the postoperative period and death was attributed to cardiac arrhythmias, disseminated intravascular coagulation, acute respiratory distress syndrome (ARDS), pancreatitis, or a combination of these problems. The PBZ-treated dog with PDH and the PBZ-treated dogs with mast cell tumor and soft tissue sarcoma survived. Postmortem evaluation was performed in 2 dogs and did not reveal other concurrent neoplasia.
Twelve (48%) of 25 untreated dogs were nonsurvivors. Seven dogs died intraoperatively and 5 in the postoperative period. Death in these 12 dogs was attributed to cardiac arrhythmias leading to cardiac arrest, ARDS, hemorrhage, acute renal failure, multiple organ dysfunction syndrome (MODS), metastatic disease, owner-elected euthanasia owing to unresectable or metastatic disease, or a combination of these problems. Necropsy was performed in 9 untreated dogs and findings included thyroid carcinoma (2 dogs), renal hemangiosarcoma (1), hepatocellular carcinoma (1), and liver hemangiosarcoma (1). Three of 6 untreated dogs with concurrent neoplasia died owing to systemic inflammatory response syndrome (SIRS)/MODS (1), owner-elected intraoperative euthanasia owing to unresectable tumor (1), and ARDS (1). Three of 4 untreated dogs with concurrent hyperadrenocorticism died owing to ARDS (1), a combination of SIRS and MODS (1), and owner-elected intraoperative euthanasia owing to unresectable tumor (1). The dog euthanized intraoperatively owing to an unresectable tumor was the dog with concurrent atrial-ventricular block for which a temporary pacemaker was placed. An adrenal cortical adenoma was identified in 1 untreated dog with hyperadrenocorticism at necropsy. The majority of nonsurvivors were humanely euthanized owing to severe complications in the intraoperative or postoperative period as opposed to having experienced spontaneous death.
Statistically significant factors associated with an improved 10-day survival in univariate models included decreased age (P= .028), lack of intraoperative arrhythmias (P= .008), reduced surgical time (P= .009) and preoperative treatment with PBZ (P= .014) (Table 4). For every 1-year age increase, the odds of 10-day survival decreased by 33%. The occurrence of intraoperative arrhythmias decreased the odds of 10-day survival by 90%. The longer the surgical time, the poorer the survival. For every 1 hour increase in surgery time the odds of 10-day survival decreased by 75%. Pretreatment with PBZ improved survival. PBZ-treated dogs were approximately 6 times more likely to survive adrenalectomy. The variables that were not significantly associated with 10-day survival included year of surgery, duration of anesthesia, presence of vascular invasion, tumor size or location, intra- and postoperative hypertension, number of hypertensive episodes, hypotension, number of hypotensive episodes, or hemorrhage (Table 4).
Table 4. Results of univariate logistic regression analyses of factors potentially associated with perioperative survival (>10 days after surgery) in 48 dogs undergoing adrenalectomy for pheochromocytoma.
|Sex (male versus female, (reference))||1.0||0.27–3.65||1.0|
|Age (1 year increase)||0.67||0.47–0.96||.028*|
|Anesthesia time (hours)||0.67||0.41–1.08||.10|
|Surgical time (hours)||0.25||0.09–0.70||.009*|
|Intraoperative maximum BP||1.0||0.99–1.01||.99|
|Intraoperative minimum BP||1.0||0.98–1.05||.41|
|Intraoperative BP variability||1.0||0.99–1.01||.77|
Adrenalectomy is the treatment of choice for pheochromocytoma in dogs. Unfortunately, the prognosis for dogs with pheochromocytoma that undergo adrenalectomy has historically been described as guarded to poor due, in part, to the high morbidity and mortality rates associated with surgery.1,15 Although prognostic indicators have been reported for dogs undergoing adrenalectomy,5 most of the dogs evaluated had adrenocortical tumors. The goal of our study was to identify predictive factors for survival in dogs undergoing adrenalectomy for pheochromocytoma. Forty-eight dogs met the criteria for inclusion in the study. The age and breed of dogs with pheochromocytoma in this study were similar to those of previous reports.1,4 The variability in breed and body size is typical for dogs with pheochromocytoma. Males were overrepresented in our study, which has not been reported previously.1,4 In our study, the significant prognostic indicators for survival included lack of intraoperative arrhythmias, decreased duration of surgery, pretreatment with PBZ, and younger age.
PBZ noncompetitively blocks the α-adrenergic response to circulating epinephrine or norepinephrine. It is considered the drug of choice for preoperative management of hypertension in people with pheochromocytoma. Mortality rates have significantly decreased from 40 to 60% to 6% or less when humans are treated with PBZ for a minimum of 7 days before adrenalectomy.20–24 People with pheochromocytoma may experience chronic vasoconstriction and have decreased blood volume as a result of increased concentrations of circulating catecholamines.25 PBZ is used to reverse vasoconstriction and hypovolemia before surgery and control fluctuations of blood pressure and heart rate during anesthesia.14,26 PBZ irreversibly binds to α-receptors. Therefore, the duration of its inhibitory actions last until more receptors are produced. For this reason, PBZ has a long duration of action and these effects cannot be overridden by tumor-associated surges of catecholamine secretion.27 The dose, frequency of PBZ administration, and length of therapy required to adequately achieve desired effects has not been defined for dogs. Clinical impressions suggest that daily dosages of 1–2 mg/kg given for approximately 2 weeks before surgery are beneficial, but we were unable to demonstrate a dose and duration-to-survival correlation in this retrospective study. The dosage of 1–2 mg/kg may not be tolerated by some dogs owing to α-adrenergic antagonism-induced hypotension. PBZ was associated with improved survival in this study. The mortality rate for all dogs was high (31%) but significantly (P= .014) lower in dogs treated (13%) versus those not treated (48%) with PBZ.
The reason or reasons for the decreased mortality rate in PBZ-treated dogs was not readily apparent in this study. In theory, PBZ should lead to decreased intraoperative hypertension, decreased hypertensive episodes, and increased stability of blood pressure during anesthesia and surgery. However, there were no significant differences between PBZ-treated and untreated dogs for these perioperative variables. The difference in mortality rate may have been related to normalization of intravascular volume before surgery, improvement in the frequency, and severity of hypertensive episodes in PBZ dogs before hospitalization for adrenalectomy, or to population bias. Although age was shown to have a significant effect on survival, there was no significant difference in age between PBZ and untreated dogs, suggesting that age, by itself, cannot explain the difference in survival between the 2 treatment groups.
Two PBZ dogs were receiving medications (prednisone, atenolol) that may trigger increases in plasma catecholamine concentrations leading to vasoconstriction.28,29 The 2 PBZ dogs that received prednisone and atenolol survived. In addition to exogenous glucocorticoids and β-blockers, metoclopramide may trigger a catecholamine crisis.28 No dog, in either group, was receiving metoclopramide. Endogenous glucocorticoids have not been implicated in excessive catecholamine secretion.
One PBZ and 3 untreated dogs were diagnosed with PDH and 1 untreated dog was diagnosed with adrenal-dependent hyperadrenocorticism. Concurrent hyperadrenocorticism may have affected outcome in these dogs because of thromboembolic disease or coagulation disorders. One untreated dog with PDH developed ARDS and thromboemboli may have played a role. None of the remaining dogs with concurrent hyperadrenocorticism developed identifiable problems suggestive of thromboemboli or coagulation disorders.
Concurrent neoplasia was identified before adrenalectomy in 2 dogs from each group and in 4 additional untreated nonsurvivor dogs at necropsy. The cause of death in these dogs (SIRS/MODS, unresectable tumor, and ARDS) appeared to be unrelated to concurrent neoplasia, although identification of an unresectable pheochromocytoma and another mass may have played a role in the owner's decision to elect euthanasia intraoperatively in 1 dog. Other concurrent diseases, such as hypothyroidism and diabetes mellitus, were similar between both groups and were unlikely to have affected outcome.
Intraoperative arrhythmias were significantly associated with decreased survival. Although intraoperative arrhythmias occurred with similar frequency in both groups of dogs, survival was quite different as all untreated dogs with intraoperative arrhythmias died compared with 1 of 6 PBZ dogs that developed cardiac arrhythmias. PBZ, by itself, does not prevent intraoperative arrhythmias. In humans, β-blockade has been used after adequate α-blockade to control tachycardia and tachyarrhythmias.22,25β-blockade was not instituted in our dogs but may be indicated, given the association between intraoperative arrhythmias and survival identified in this study.
A longer surgical time was significantly associated with decreased survival but there was no association between size of the tumor or presence of vascular invasion and survival. The majority of pheochromocytomas identified in this study were medium or large in size although these categories were assigned arbitrarily. Tumor thrombi were common. Caval thrombi were observed more often in the dogs reported here (71%) compared with previous studies (33–55%).4,5,30 The large tumor size and presence of vascular invasion is expected given the malignant and invasive nature of pheochromocytoma combined with the site of the study. The VMTH is a tertiary referral center and likely creates a bias toward more advanced disease (ie, larger pheochromocytomas). Larger tumors and vascular invasion may result in a more difficult surgical resection and increase the risk of perioperative complications such as hemorrhage and hemodynamic instability. However, there was no difference in survival among dogs with small, medium, or large pheochromocytomas. The presence of vascular invasion also may increase surgical time, thus influencing survival. In agreement with a previous study,5 presence of vascular invasion did not significantly affect survival.
Pheochromocytomas involved the right adrenal gland more commonly than the left in this study, which is similar to a previous retrospective study where 27 of 50 dogs had right-sided pheochromocytomas identified at necropsy or exploratory surgery.4 The overrepresentation of right-sided tumors in this study may represent a population bias because right-sided adrenal tumors have historically been regarded as more difficult to remove and may have led to greater likelihood of referral to our hospital.
Although adrenalectomies are performed with increasing frequency in our hospital, there was no significant difference between survival and the year adrenalectomy was performed. This finding was surprising as we expected an overall improvement in surgical, anesthetic, and postoperative management skills with time and experience. As with most academic teaching hospitals, our hospital has a changing cycle of residents and faculty that may have affected the impact of year of surgery on survival. Although fewer dogs received PBZ early in the study period, the increased mortality rate in untreated dogs does not appear to be related to the time of surgery because there was no significant difference between survival and the year adrenalectomy was performed.
There are several limitations in this study. One limitation is the inability to control for the various concurrent medical diseases present including neoplasia and hyperadrenocorticism. Owing to the retrospective design, blood pressure measurements were not consistently available in PBZ and untreated dogs at first presentation and could not be evaluated. Additionally, the retrospective design of the study occasionally resulted in incomplete clinical data, especially for dogs treated early in the study period. Such missing data, resulting in a smaller effective sample size, precluded the construction of multivariate statistical models that could presumably control for the effects of known confounders, leading to underestimation of P-values, and confidence intervals. Anesthetic protocols were not the same for every dog and there was variability in dose and duration of PBZ treatment before adrenalectomy. Lastly, there may have been selectivity in the dogs referred and chosen for surgical management, and dogs with large tumors or vascular invasion may have been more likely to be referred. Despite the limitations of this study, the results support treatment with PBZ before surgical removal of pheochromocytoma in dogs.
aGallipot Inc, St Paul, MN
bSorenson Transpac II, Abbott Laboratories, Abbott Park, North Chicago, IL; DTX Plus DT-XX, Becton Dickinson Critical Care Systems Inc, Sandy, UT
cCardell veterinary monitor 9301V, CAS Medical System, Brandfort, CT; Dinamap monitor 1846, Vital Signs, Totowa, NJ
dEgret, Cytel Software Corp, Cambridge, MA