Evaluation of risk factors for outcome associated with adrenal gland tumors with or without invasion of the caudal vena cava and treated via adrenalectomy in dogs: 86 cases (1993-2009)
Jessica S. Barrera, DVM, DACVS; Fabrice Bernard, DVM; E. J. Ehrhart, DVM, PhD, DACVP; Stephen J. Withrow, DVM, DACVS, DACVIM; Eric Monnet, DVM, PhD, DACVS
Objective-To evaluate risk factors for outcome for dogs with adrenal gland tumors with or without invasion of the caudal vena cava treated via adrenalectomy.
Design-Retrospective study.
Animals-86 dogs that underwent adrenalectomy for treatment of adrenal gland tumors.
Procedures-Medical records of dogs that underwent adrenalectomy for treatment of an adrenal gland tumor from 1993 to 2009 were reviewed; data collected including signal- ment, clinical signs, diagnostic test findings, treatments prior to surgery, findings at surgery including additional procedures performed and extent of caudal vena caval invasion (local invasion [caudal to the hepatic portion of the vena cava] or extensive invasion [cranial to the hepatic portion of the vena cava]), procedures performed during surgery, histopathologic diagnosis, perioperative complications, follow-up data, and necropsy findings.
Results-Of the 86 dogs, 14 had adenomas, 45 had adrenocortical carcinomas, and 27 had pheo- chromocytomas. Fourteen dogs had invasion of the caudal vena cava; of these tumors, 7 were locally invasive and 7 were extensively invasive. Risk factors for poor short-term survival (death within 14 days following surgery) were vena caval invasion, extent of invasion, pheochromocy- toma, intraoperative transfusion, and postoperative factors including disseminated intravascular coagulation, pancreatitis, hypotension, hypoxemia, and renal failure. Multivariate analysis of risk factors for poor short-term survival revealed that extensive invasion was the most important fac- tor. Regardless of extent of invasion or tumor type, long-term survival was possible.
Conclusions and Clinical Relevance-Invasion of the caudal vena cava, particularly tumor thrombus extension beyond the hepatic hilus, was associated with a higher postoperative mortality rate, but did not affect long-term prognosis in dogs undergoing adrenalectomy because of an adrenal gland tumor. (J Am Vet Med Assoc 2013;242:1715-1721)
A drenal gland tumors are rare in dogs, comprising 1% to 2% of all canine tumors.1 The most common tu- mor types include adrenocortical adenoma, adrenocorti- cal carcinoma, and pheochromocytoma.1-5 Adrenocortical tumors may or may not be cortisol secreting, which may lead to clinical signs of hyperadrenocorticism.1,2,6 For dogs with clinical signs attributable to an adrenal gland tumor, adrenalectomy is the gold-standard treatment.1-6
Invasion of the caudal vena cava by adrenal gland tu- mors has been reported to occur in 9.5% to 46% of cases.6 In the past, vena caval invasion was often determined to be inoperable or to be associated with a poor prognosis. However, more recent reports7,8 indicate that tumor inva- sion into the caudal vena cava may not affect short-term survival rate. Kyles et al7 reported that adrenalectomy for tumors with associated vena caval thrombi does not result in substantially higher perioperative morbidity and mor- tality rates, compared with surgery in dogs without inva-
Address correspondence to Dr. Monnet (Eric.Monnet@ColoState. EDU).
ABBREVIATIONS
DIC
CI Confidence interval Disseminated intravascular coagulation
sion. Those data are in agreement with the prognosis in humans with adrenal gland tumors; the presence of vena caval thrombi does not lead to a poorer prognosis.9
The purpose of the study reported here was to evalu- ate risk factors associated with short-term and long-term outcome for adrenal gland tumors with or without inva- sion of the caudal vena cava and treated via adrenalecto- my in dogs. The hypothesis was that vena caval invasion would have no association with short- or long-term sur- vival rate of dogs surgically treated for adrenal tumors.
Materials and Methods
Case selection criteria-Medical records of dogs that were treated for an adrenal gland tumor via ad- renalectomy at Colorado State University Veterinary Teaching Hospital from 1993 to 2009 were reviewed. Criteria for inclusion were adrenalectomy and histo- logic diagnosis of an adrenal tumor.
Medical records review-Surgical report evaluation was performed, and dogs were subsequently allocated to a no vena caval invasion group and a vena caval invasion group. Involvement of the caudal vena cava was defined as invasion into the caudal vena cava. For the vena caval invasion group, subgroups were formed on the basis of the severity of caval tumor thrombosis as tumor thrombi extending caudal to (local invasion) or cranial to (exten- sive invasion) the hepatic portion of the vena cava.
The following preoperative data were collected from the medical records: signalment, clinical signs, CBC, se- rum biochemical analyses, results of endocrine testing, radiography and abdominal ultrasonography findings, and use of phenoxybenzamine. Diagnosis of hyperadreno- corticism was made on the basis of adrenal gland function testing, including the ACTH stimulation test, low- and high-dose dexamethasone suppression tests, and mea- surement of endogenous serum ACTH concentrations.
Intraoperative data collected included surgical findings, including degree of invasion of the vena cava as defined, anesthetic complications, surgical compli- cations, and additional surgical procedures performed. Postoperative data collected included complications and medications. The diagnosis of DIC was made on the basis of the presence of at least 3 of the following find- ings: thrombocytopenia, increased prothrombin time, increased activated partial thromboplastin time, de- creased antithrombin III activity, presence of D-dimers, and increased fibrin degradation products, compared with reference ranges.1º A diagnosis of pancreatitis was made on the basis of clinical signs plus suggestive ultra- sonographic findings or high serum canine pancreatic lipase immunoreactivity.
Histopathologic specimens were reviewed by board-certified pathologists and were classified as ad- renocortical adenoma, adrenocortical carcinoma, or pheochromocytoma. Histopathologic criteria for ma- lignancy of adrenocortical tumors included evidence of invasion, as described.11 In addition, mitotic rate, presence or absence of vascular invasion, and degree of nuclear pleomorphism were other criteria used to dif- ferentiate benign and malignant adrenocortical tumors.
Follow-up was performed by review of the medi- cal record in combination with a telephone interview with the owner or the referring veterinarian. Short-term survival was defined as death within 14 days following surgery. Long-term survival was defined as survival > 14 days following surgery. Dogs that died from unrelated causes, were lost to follow-up, or were still alive were censored. Dogs that died or were euthanized because of their disease were not censored. Dogs that died within 14
days following surgery were not included in long-term survival rates.
Statistical analysis-Distributions of categorical data between groups were compared via x2 analysis, and continuous data were compared by use of ANOVA. Medi- an survival rate and 1-, 2-, and 3-year survival rates were analyzed via actuarial Kaplan-Meier survival analysis. Cox proportional hazard univariate analysis was used to evaluate risk factors for short-term and long-term surviv- al of dogs with different tumors as well as with and with- out invasion of the caudal vena cava. Factors that were significant at P < 0.05 following univariate analysis were entered into a proportional hazard multivariate analysis to further evaluate risk factors for short- and long-term survival. Data presented are mean + SD values. Values of P < 0.05 were considered significant. Statistical softwarea was used to perform the analyses.
Results
For the period from 1993 through 2009, 86 dogs met the inclusion criteria. Fourteen dogs had an adrenocor- tical adenoma, 45 had an adrenocortical carcinoma, and 27 had a pheochromocytoma. Breeds represented among the dogs with an adrenocortical adenoma were mixed (n = 3), Labrador Retriever (2), Lhasa Apso (2), and 1 each of American Eskimo Dog, Beagle, Belgian Malinois, Cocker Spaniel-Poodle cross, Golden Retriever, Miniature Poodle, and Siberian Husky. Breeds represented among the dogs with adrenocortical carcinoma were Labrador Retriever (n = 8), mixed (7), German Shepherd Dog (3), Basset Hound (2), Beagle (2), Bichon Frise (2), Dachshund (2), German Shorthaired Pointer (2), Golden Retriever (2), and 1 each of Australian Cattle Dog, Boxer, Cocker Spaniel-Poodle cross, Cocker Spaniel, Collie, Corgi, Dalmatian, Kees- hond, Maltese, Rhodesian Ridgeback, Saluki, Schipperke, Shih Tzu, Siberian Husky, and Standard Schnauzer. Breeds represented among dogs with pheochromocytoma were mixed (n = 6), Labrador Retriever (4), Siberian Husky (4), Rottweiler (3), and 1 each of Airdale Terrier, Austra- lian Shepherd Dog, Beagle, Flat-Coated Retriever, Golden Retriever, Miniature Dachshund, Old English Sheepdog, Puli, Rhodesian Ridgeback, and Samoyed. Eighty dogs underwent surgery for a suspected adrenal gland tumor. Six dogs underwent surgery for other reasons, including 4 dogs with hemoabdomen, 1 dog with an abdominal mass, and 1 dog with a suspected insulinoma.
Preoperative data-Age and weight were not sig- nificantly different among dogs with the different tumor types (Table 1). Systolic and diastolic arterial blood pres-
| Variable | Tumor type | P value | ||
|---|---|---|---|---|
| Pheochromocytoma | Adenoma | Carcinoma | ||
| Age (y) | 11.0 (10.1-11.8); n = 27 | 11.4 (10.1-12.5); n = 14 | 11.4 (10.8-12.1); n = 45 | 0.703 |
| Weight (kg) | 29.3 (24.5-34.1); n = 27 | 22.8 (16.1-29.5); n = 14 | 25.5 (21.8-29.2); n = 45 | 0.246 |
| Systolic arterial blood pressure (mm Hg) | 182.5 (166.1-198.9); n = 16 | 152.0 (105.7-198.3); n = 2 | 174.9 (158.5-191.2); n = 16 | 0.427 |
| Mean arterial blood pressure (mm Hg) | 143.4 (124.3-162.6); n = 7 | — | 113.6 (91.0-136.2); n = 5 | 0.049 |
| Diastolic arterial blood pressure (mm Hg) | 122.0 (100.8-143.2); n = 7 | — | 91.8 (66.7-116.9); n = 5 | 0.068 |
| Values are mean (95% CI). -= Not determined. | ||||
sures were also similar. Mean arterial blood pressure was significantly higher for dogs with pheochromocytoma than other tumor types. Fifteen dogs were treated with phenoxybenzamine for 1 to 2 weeks before surgery, in- cluding 6 dogs with pheochromocytomas and 9 dogs with adrenocortical carcinomas.12 Three dogs with inva- sion of the vena cava received phenoxybenzamine.
Surgical findings-Fourteen dogs were in the vena caval invasion group, and 72 dogs were in the no vena caval invasion group. Thirteen of the 14 invasive adre- nal gland tumors were pheochromocytomas, and 1 ad- renocortical carcinoma was invasive (P < 0.001). Seven tumors were locally invasive, and 7 tumors had exten- sive invasion into and beyond the hepatic portion of the vena cava. Five left-sided tumors and 9 right-sided tumors invaded the caudal vena cava (P = 0.833).
Intraoperative complications-Seventy-one of the 86 dogs had intraoperative complications. Seven dogs died during surgery; of these, 6 had pheochromocyto- mas and 1 had an adrenocortical carcinoma. Five of the intraoperative deaths occurred in the vena caval inva- sion group; 4 of these dogs had extensive invasion. One patient with local invasion and 2 dogs in the no vena caval invasion group died during surgery. Intraopera- tive death was significantly (P = 0.001) more frequent in dogs with invasion of the vena cava (5/14) than for dogs without invasion of the vena cava (2/72). Causes for intraoperative death included persistent hemorrhage at the surgical site (n = 2), cardiac arrest (3), perceived lack of resectability of the tumor followed by euthana- sia (1), and persistent severe hypotension unresponsive to all treatments (1). Perceived lack of resectability was determined in 1 patient because of an extensive and in- travascularly adherent tumor thrombus (length, 20 cm) that extended to the level of the right atrium.
Intraoperative complications included hypotension in 53 dogs, blood loss requiring transfusion in 22 dogs, arrhythmias in 14 dogs, hypertension in 12 dogs, and hy- poxia in 7 dogs. Twenty-five of 27 dogs with pheochro- mocytomas and 36 of 45 dogs with adrenocortical car- cinomas had intraoperative complications (P = 0.151).
Postoperative phase-Of the 79 dogs that survived surgery, 34 had postoperative complications. Fifteen
1.0
Proportion surviving
0.9
0.8
Carcinoma
0.7
0.6
0.5
0.4
Pheochromocytoma
0.3
0.2
0.1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Time (d)
dogs died in the postoperative period. Causes of post- operative death included cardiorespiratory arrest for unknown underlying causes in 3 dogs, suspected DIC in 4, respiratory failure in 4, peritonitis in 3, and eu- thanasia because of renal failure in 1. Of the dogs with complications, 12 had a pheochromocytoma, 3 had an adenoma, and 19 had a carcinoma (P = 0.132). Com- plications occurred in 10 of the dogs in the vena caval invasion group and 28 of dogs in the no vena caval in- vasion group (P = 0.063). Complications encountered included vomiting or diarrhea in 18 dogs, hypotension in 13, hypoxemia in 8, DIC in 8, pancreatitis in 5, re- nal insufficiency in 4, and peritonitis in 4. Pancreati- tis was suspected in 5 dogs in the short term, and side of surgery was not associated with the development of pancreatitis (P = 0.113), despite all 5 dogs that devel- oped suspected pancreatitis having a left-sided adrenal- ectomy. Dogs in the vena caval invasion group were not more likely to develop pancreatitis (P = 0.159)
Short-term survival-Fifty-two dogs were known to have survived the short-term period; 22 dogs died in the short-term period, and 11 dogs were lost to follow-up at a median of 14 days (range, 1 to 14 days). Given that all dogs with adenomas survived and none of these were in the vena caval invasion group, they were excluded from short- and long-term survival analysis. Fourteen- day survival rates were 79% for carcinomas and 48% for pheochromocytomas (P = 0.003; Figure 1). Ten of the 22 dogs that died in the short-term period were in the vena caval invasion group. The proportion of dogs sur- viving at 14 days was significantly (P < 0.001) different (hazard ratio, 2.29) between the no vena caval invasion group (55/71 [77%]) and the vena caval invasion group (4/14; Figure 2). Of the 10 dogs in the vena caval inva- sion group that died in the short term, 7 had extensive invasion and 3 had local invasion. Short-term mortality rates were 43% for dogs with locally invasive tumors and 100% for dogs with extensively invasive tumors (Figure 3; hazard ratio, 4.43; P < 0.001).
Univariate preoperative factors associated with short-term survival (death within 14 days following surgery) included tumor type of pheochromocytoma, invasion of the caudal vena cava, and extensive inva- sion of the caudal vena cava (Table 2). Pretreatment
1.0
0.9
Proportion surviving
0.8
No Invasion
0.7
0.6
0.5
0.4
0.3
0.2
Invasion
0.1
0.0
0
1
2
3
4
5
6
7
8
9
10
11
. 12 13 1
14
Time (d)
AVIAN
1.0
Proportion surviving
0.9
0.8
0.7
Local
0.6
0.5
0.4
Extensive
0.3
0.2
0.1
0.0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Time (d)
| Factor | Hazard ratio | 95% CI | Pvalue |
|---|---|---|---|
| Tumor type (pheochromocytoma) | 1.75 | 1.16-2.72 | 0.008 |
| Vena caval invasion | 2.29 | 1.48-3.48 | < 0.001 |
| Invasion extent | 4.43 | 2.12-9.30 | < 0.001 |
| Anesthetic complications | 1,300 | 1.62-1.15 X 10244 | 0.002 |
| Transfusion during surgery | 2.49 | 1.63-3.94 | < 0.001 |
| Postoperative complications | 1.63 | 1.05-2.65 | 0.029 |
| DIC after surgery | 3.12 | 1.88-5.21 | < 0.001 |
| Pancreatitis after surgery | 1.93 | 1.03-3.19 | 0.042 |
| Hypotension after surgery | 2.22 | 1.41-3.42 | < 0.001 |
| Hypoxemia after surgery | 1.79 | 1.02-2.89 | 0.042 |
| Renal failure after surgery | 2.25 | 1.20-3.78 | 0.015 |
with phenoxybenzamine did not make a significant difference in short-term survival for adrenalectomies overall (P = 0.441) or for pheochromocytomas alone (hazard ratio, 1.42; P = 0.322). Univariate intraopera- tive factors associated with short-term survival includ- ed transfusion and blood loss.
Univariate postoperative factors associated with short-term survival were DIC, pancreatitis, hypoxemia, renal failure, and hypotension. Nephrectomy was not identified as a significant risk factor for decreased rate of survival for > 14 days (hazard ratio, 1.65, 95% CI, 0.95 to 2.63). However, the median survival time was significantly (P = 0.009) associated with nephrectomy (Figure 4). Nephrectomy was significantly (P = 0.001) associated with subsequent development of renal fail- ure. Dogs with pheochromocytomas were significantly (P = 0.014) more likely to have DIC after surgery, com- pared with dogs with adrenocortical carcinomas. Addi- tionally, dogs that received intraoperative transfusions with blood products were significantly (P = 0.003) more likely to develop postoperative DIC. The rate of hypox- emia was not significantly different among tumor types (P = 0.678) or between the vena caval invasion and no vena caval invasion groups (P = 0.057). Twenty-seven dogs in the study, including 3 dogs with adenomas, 5 dogs with pheochromocytomas, and 19 dogs with adre- nocortical carcinomas, were treated with heparin dur- ing or after surgery. Administration of heparin was not associated with short-term survival rate (P = 0.847).
1.0
Proportion surviving
0.9
0.8
0.7
0.6
No nephrectomy
0.5
0.4
0.3
0.2
0.1
Nephrectomy
0
1
2
3
4
5
6
7
8
9
11
12
13
14
Time (d)
| Factor | Pvalue |
|---|---|
| Caval invasion and tumor type | 0.003 |
| Caval invasion | 0.012 |
| Tumor type (pheochromocytoma) | 0.409 |
| Extent of invasion and tumor type | 0.001 |
| Extent of invasion | < 0.001 |
| Tumor type (pheochromocytoma) | 0.192 |
| Caval invasion and extent of invasion | 0.001 |
| Caval invasion | 1.000 |
| Extent of invasion | 0.016 |
Multivariate analysis was performed to evaluate the effect of identified significant univariate factors on the outcome for dogs in the vena caval invasion and no vena caval invasion groups. When the vena caval inva- sion group was evaluated on the basis of adrenocorti- cal adenocarcinoma versus pheochromocytoma tumor type, invasion of the vena cava remained significant but tumor type was no longer a significant risk factor (Table 3). When the extent of vena caval invasion was evaluated on the basis of adrenocortical adenocarci- noma versus pheochromocytoma tumor type, extent of the tumor thrombus was still significant but the tumor type was no longer a significant risk factor. When the vena caval invasion group was evaluated by the extent of the tumor thrombus, extent of the tumor thrombus remained significant but invasion of the vena cava was no longer a significant risk factor.
Long-term survival-Fifty-two dogs survived > 14 days. Of these, 10 were lost to follow-up at a median of 6.1 months (range, 0.6 to 49.2 months). For those not lost to follow-up, median follow-up time was 16.9 months (range, 0.5 to 54.5 months). The median sur- vival time for dogs with adrenocortical carcinomas was 48 months (95% CI, 38.1 to 54.5 months) and was not reached for dogs with pheochromocytomas (P = 0.003; Figure 5). One-, 2-, and 3-year survival rates were 88% for carcinomas and 83%, 60%, and 60%, respectively, for pheochromocytomas.
For dogs with adrenocortical adenomas, median follow-up time was 26.3 months (range, 0.56 to 49.2
1.1
1.0
0.9
Carcinoma
Proportion surviving
..
. ..
0.8
. …
0.7
.
0.6
.
0.5
Pheochromocytoma
0.4
0.3
0.2
0.1
0.0
0
12
24
36
48
60
Time (mon)
1.0
.
0.9
No Invasion
Proportion surviving
0.8
…
.. ..
0.7
0.6
0.5
.
0.4
Invasion
0.3
0.2
0.1
0.0
0
12
24
36
48
60
Time (mon)
months). Four dogs were lost to follow-up at 0.5, 0.9, 5, and 49 months after surgery. All but I dog had long-term resolution of clinical signs, and this dog had recurrence of polyuria and polydipsia and was eutha- nized; necropsy revealed growth of an adrenocortical carcinoma in the previous tumor site.
For dogs with adrenocortical carcinomas, median follow-up time was 12.7 months (range, 0.5 to 54.5 months). Six dogs were lost to follow-up at a median of 7.1 months (range, 0.7 to 24.7 months). Nineteen dogs had long-term resolution of clinical signs, and 8 had recurrence of signs. Metastasis was suspected in 6 of these dogs on the basis of clinical signs and imaging findings but not confirmed with histopathologic find- ings in any patient. Confirmed metastasis in the long- term period was found in 3 dogs with adrenocortical carcinoma to the liver (2 dogs) and small intestines (1 dog). Metastasis to the liver or lungs was suspected in an additional 7 dogs with adrenocortical carcinoma.
For dogs with pheochromocytomas, median follow- up time was 15.8 months (range, 1.0 to 36.5 months). Twelve of these 13 dogs had resolution of their clinical signs. One dog developed weakness, anorexia, and sus- pected lung metastasis 10 months after surgery and was euthanized. Necropsy revealed vertebral metastasis with pulmonary adenocarcinoma and disseminated plasma cell tumor; it was not clear which tumor was secondary.
Confirmed metastasis in the long-term period was found in 3 dogs with a pheochromocytoma to the vertebrae, lungs, and liver. Metastasis to the lungs was suspected in an additional 2 dogs with pheochromocytoma.
Three of the 52 dogs that survived into the long- term period were in the vena caval invasion group and all had pheochromocytomas. No significant (P = 0.4544) difference was seen in long-term survival rate between dogs with and without vena caval invasion (Figure 6). Median survival time for dogs without vena caval invasion was 48 months (95% CI, 38.1 to 54.5 months) and was not reached for dogs with vena caval invasion. No factors were significant for long-term sur- vival rate as determined via univariate analysis.
Discussion
Adrenalectomy for tumors that invade the caudal vena cava was associated with a higher perioperative mortality rate, particularly if invasion was extensive, compared with tumors that did not invade the caudal vena cava in the present study. Because invasion of the caudal vena cava is much more common in dogs with pheochromocytoma, their short-term prognosis is af- fected. However, after correcting for invasion of the vena cava, the short-term prognosis for dogs with pheo- chromocytoma was not different than for dogs with ad- renocortical adenocarcinoma. Long-term survival rate in dogs that survived the perioperative period was not associated with invasion of the vena cava at the time of surgery. The population of dogs in the study was similar in signalment, body weight, history, and clinical signs to populations in previous studies.7,8,13 Age, body weight, sex, and breed were not significantly different among dogs with various tumor types and between the vena caval invasion and no vena caval invasion groups. A significant difference in the mean arterial blood pres- sure of dogs with pheochromocytomas was not de- tected, compared with other tumor types, which can be explained by catecholamine surges in patients with pheochromocytoma. In a study on adrenalectomy with and without vena caval thrombi, Kyles et al7 found a similar distribution of invasion of the vena cava as in the population of the present study. Schwartz et al13 re- ported a population with less invasion of the vena cava than in the population of the present study. In the pres- ent study, 50% of the cases of invasion of the vena cava were classified as extensive (extending to the hepatic vena cava or beyond), whereas it was 20%7 or none14 in other studies. In dogs that survived long term, metas- tasis was histologically confirmed in 11% of dogs with pheochromocytomas and in 6% of dogs with adreno- cortical carcinoma. Although this rate is lower than in previous studies,14-16 it likely reflected the low number of necropsies performed in long-term surviving dogs. Metastasis was suspected on the basis of clinical signs and diagnostic testing (thoracic radiography and ab- dominal ultrasonography) in an additional number of dogs. If these were included, metastatic rate would be 18% in dogs with pheochromocytomas and 22% in dogs with adrenocortical carcinomas.
Invasion of the caudal vena cava and extent of the tumor thrombus were risk factors for death in the short term in the population reported here. However, caval
invasion was not a risk factor for death in the long term. Dogs with caval thrombi extending cranial to the he- patic hilus were > 4 times as likely to die during the short-term period as were dogs with thrombi that did not extend past the hilus. Pheochromocytomas were significantly more likely to invade the caudal vena cava, compared with other tumor types, and all of the exten- sively invasive tumors were pheochromocytomas. Pre- vious studies7,8,12-15 have also identified the prevalence of pheochromocytomas for local vascular invasion at a rate ranging from 15% to 55%. This is in compari- son to an invasion rate for adrenocortical carcinomas of 11% to 21.5%.6,7,11,12,14 Invasion in the present study was a significant perioperative risk factor because half of these cases were extensively invasive. Surgical treat- ment for extensive invasion involves a greater degree of surgical invasiveness and cardiovascular compromise: inflow occlusion must be complete (vs partial occlu- sion options for small locally invasive tumors) and may require a greater duration of inflow occlusion and more manipulation of the pheochromocytoma, resulting in the release of catecholamines and increased risk of an- esthetic instability.
Intraoperative mortality rate was 8%, and the short-term mortality rate was 25%. This finding is simi- lar to results described in the most recent reports12-14,17 on adrenalectomy. Only 48% of dogs with pheochro- mocytomas in the present study survived the short- term period, in comparison to 81% reported by Kyles et al.7 This difference was likely attributable to the fact that 7 of the dogs with pheochromocytomas had tu- mor thrombi that extended into the hepatic vena cava (extensive invasion). After adjusting for presence of a tumor thrombus or the extent of the tumor thrombus in the vena cava, pheochromocytomas did not have a worse prognosis than did adenocarcinomas. All 7 of these dogs died in the perioperative period, and they accounted for more than one-quarter of the dogs with pheochromocytoma. The short-term mortality rate for adrenal tumors invading the caudal vena cava was 72%. Kyles et al7 reported a 30% perioperative (between sur- gery and time of discharge) mortality rate for dogs with vena caval thrombi; those data are favorable, compared with our data, but the difference is likely reflective of the extent of the thrombi. When invasion of the vena cava was adjusted for the extent of the tumor, invasion of the vena cava was no longer a risk factor for death in the short term, which is in agreement with the find- ings of Kyles et al.7 Kyles et al7 reported 8 of 10 invasive tumors as locally invasive and 2 of 10 as extensively invasive, compared with our series, in which 7 of dogs with invasive tumors had local invasion and 7 had ex- tensive invasion.
Nephrectomy was not directly identified as a risk factor in the present study. However, median survival time was negatively affected because it was associated with development of renal failure, and renal failure was identified as a risk factor for death in the short term. Nephrectomy was identified as a negative prognostic factor by Schwartz et al,13 with a hazard ratio of 142; however, nephrectomy was not identified as a direct risk factor for decreased survival rate in the present study. Nephrectomy was performed if the tumor was
invading the kidney or renal vein or wrapping around the renal vein. Although presurgical renal values may have been within reference range in these dogs, many of them were older dogs that could have had subclinical renal insufficiency or sustained or paroxysmal hyper- tension, which would mask renal insufficiency. Because the dogs with unilateral nephrectomy were more at risk for renal failure after surgery and because renal failure was identified as a significant risk factor for decreased short-term survival rate, we recommend diagnostic evaluation for adrenal gland tumor involvement of the kidney or renal vasculature and measurement of glo- merular filtration rate if nephrectomy may be needed.
Similar to the study by Schwartz et al,13 the pres- ent study identified development of postoperative pan- creatitis as a significant risk factor for decreased short- term survival rate. Manipulation of the pancreas and episodes of hypotension during surgery are factors that can induce development of pancreatitis. However, in the present study, pancreatitis was not more common in dogs with vena caval invasion, even though vena caval invasion is associated with hypotension. Also, pancre- atitis was not more common in dogs with right-sided adrenalectomy, which requires greater manipulation of the pancreas. This finding was also in agreement with those of Schwartz et al.13
Disseminated intravascular coagulation was a sig- nificant risk factor for decreased short-term survival rate and was significantly more likely to occur in dogs with pheochromocytoma and dogs receiving intraoper- ative transfusions of blood or blood products. Because DIC was not significantly more likely in dogs with vena caval thrombi, the finding was not likely confounded by the number of cases of pheochromocytoma with caval invasion. Pheochromocytoma could contribute to postoperative DIC via catecholamine surges occur- ring during surgery.18 Transfusion could contribute to increased risk of DIC either because the dog was en- tering DIC during surgery and required transfusion or because coagulation factors administered via plasma could contribute to entering the early, hypercoagu- lable state of DIC.10 Additionally, massive transfusion (replacement of entire blood volume) could contribute to development of DIC via depletion of fibrinogen and coagulation factors.19
Hypoxemia was identified as a significant risk fac- tor for death in the short term in the present study. Pos- sible causes for hypoxemia in dogs with postoperative adrenalectomy include pulmonary thromboembolism, hypoventilation, aspiration pneumonia, atelectasis, and unrelated underlying pulmonary pathological changes. Because this was a retrospective study, we could not differentiate among the different causes of hypoxemia. Pulmonary thromboembolism may be tentatively diag- nosed on the basis of hypoxemia, increased alveolar-to- arterial oxygen tension gradient, and the response to oxygen administration.20 Thoracic radiographs may or may not reveal changes such as wedge-shaped, pleural- based densities.20 Computed tomographic angiography is currently the gold-standard diagnostic test in human medicine.20 We were unable to identify a significant association between tumor type and postoperative de- velopment of hypoxemia or between invasion status
and postoperative hypoxemia. The protective effect of heparin against development of pulmonary thrombo- embolism could not be evaluated in this study owing to its retrospective nature. Heparin was administered mostly to support dogs in DIC, and different dosages and routes of administration were used. If the goal is to protect against pulmonary thromboembolism, it may be more appropriate to treat with heparin at the begin- ning of surgery with a constant rate infusion with the intent to increase activated partial thromboplastin time by 1.5 times the baseline value.10
Phenoxybenzamine did not have a protective effect in the present study. Herrera et al12 reported on the pro- tective effect of phenoxybenzamine pretreatment in dogs with pheochromocytomas, with a reduction in perioper- ative mortality rate from 48% to 13% in dogs pretreated for a mean of 20 days. Seventeen percent of the dogs in the present study received phenoxybenzamine before surgery, and only 3 cases of pheochromocytoma invad- ing the caudal vena cava were treated. Therefore, it was not possible to identify a significant effect of phenoxy- benzamine on short-term survival rate.
Although vena caval invasion, extent of invasion, and tumor type were risk factors for death in the short term, no risk factors were identified that were signifi- cantly associated with long-term survival rate. Regard- less of these factors, if a dog survived beyond the 14-day postoperative period, long-term survival was possible. Median survival time was 48 months for dogs with ad- renocortical carcinomas and was not reached for dogs with pheochromocytomas; median survival time for the 3 dogs with vena caval invasion that survived to the long-term period was not reached. These results com- pare favorably with survival times described in reports of previous studies.2,13,14,16
Limitations in the present study were primarily re- lated to its retrospective nature, particularly in regard to follow-up data and incomplete clinical information. Necropsies were rarely performed, particularly in dogs that did not die in the short term. Additionally, 4 dogs that were euthanized during surgery were included in the study because the decision to euthanize was pri- marily related to a perceived decreased quality of life related to the disease process.
It appears that the most important risk factor for decreased survival time in this population was an ex- tensive tumor thrombus in the caudal venal cava. Risk factors for survival in the short term included anesthet- ic complications, intraoperative transfusion, and post- operative DIC, pancreatitis, hypotension, renal failure, and hypoxemia. In dogs that survived the short-term period, long-term survival was possible regardless of invasion status or tumor type. Careful planning with imaging to determine extent of tumor invasion may
be warranted for owner and clinician decision making prior to surgery.
a. JMP, version 8.0.2, SAS Institute Inc, Cary, NC.
References
1. Withrow SJ, Vail DM. Withrow and MacEwen’s small animal clini- cal oncology. 4th ed. St Louis: Saunders Elsevier, 2007.
2. Anderson CR, Birchard SJ, Powers BE, et al. Surgical treatment of adrenocortical tumors: 21 cases (1990-1996). J Am Anim Hosp Assoc 2001;37:93-97.
3. Gilson SD, Withrow SJ, Orton EC. Surgical treatment of pheo- chromocytoma: technique, complications, and results in six dogs. Vet Surg 1994;23:195-200.
4. von Dehn BJ, Nelson RW, Feldman EC, et al. Pheochromocy- toma and hyperadrenocorticism in dogs: six cases (1982-1992). J Am Vet Med Assoc 1995;207:322-324.
5. Maher ER Jr. Pheochromocytoma in the dog and cat: diagnosis and management. Semin Vet Med Surg (Small Anim) 1994;9:158-166.
6. Scavelli TD, Peterson ME, Matthiesen DT. Results of surgical treatment for hyperadrenocorticism caused by adrenocortical neoplasia in the dog: 25 cases (1980-1984). J Am Vet Med Assoc 1986;189:1360-1364.
7. Kyles AE, Feldman EC, De Cock HE, et al. Surgical manage- ment of adrenal gland tumors with and without associated tu- mor thrombi in dogs: 40 cases (1994-2001). J Am Vet Med Assoc 2003;223:654-662.
8. Lang JM, Schertel E, Kennedy S, et al. Elective and emergency surgical management of adrenal gland tumors: 60 cases (1999- 2006). J Am Anim Hosp Assoc 2011;47:428-435.
9. Lee JE, Berger DH, el-Naggar AK, et al. Surgical management, DNA content, and patient survival in adrenal cortical carcino- ma. Surgery 1995;118:1090-1098.
10. Bojrab MJ, Monnet E, Clarence J. Mechanisms of disease in small animal surgery. 3rd ed. Jackson, Wyo: Teton NewMedia, 2010.
11. Labelle P, Kyles AE, Farver TB, et al. Indicators of malignancy of canine adrenocortical tumors: histopathology and proliferation index. Vet Pathol 2004;41:490-497.
12. Herrera MA, Mehl ML, Kass PH, et al. Predictive factors and the effect of phenoxybenzamine on outcome in dogs undergoing adrenalectomy for pheochromocytoma. J Vet Intern Med 2008;22:1333-1339.
13. Schwartz P, Kovak JR, Koprowski A, et al. Evaluation of prognos- tic factors in the surgical treatment of adrenal gland tumors in dogs: 41 cases (1999-2005). J Am Vet Med Assoc 2008;232:77-84.
14. Barthez PY, Marks SL, Woo J, et al. Pheochromocytoma in dogs: 61 cases (1984-1995). J Vet Intern Med 1997;11:272-278.
15. Gilson SD, Withrow SJ, Wheeler SL, et al. Pheochromocytoma in 50 dogs. J Vet Intern Med 1994;8:228-232.
16. Emms SG, Johnston DE, Eigenmann JE, et al. Adrenalectomy in the management of canine hyperadrenocorticism. J Am Anim Hosp Assoc 1987;23:557-564.
17. van Sluijs FJ, Sjollema BE, Voorhout G, et al. Results of adre- nalectomy in 36 dogs with hyperadrenocorticism caused by adreno-cortical tumour. Vet Q 1995;17:113-116.
18. Squizzato A, Gerdes VE, Ageno W, et al. The coagulation sys- tem in endocrine disorders: a narrative review. Intern Emerg Med 2007;2:76-83.
19. . Raymer JM, Flynn LM, Martin RF. Massive transfusion of blood in the surgical patient. Surg Clin North Am 2012;92:221-234.
20. Goggs R, Benigni L, Fuentes VL, et al. Pulmonary thromboem- bolism. J Vet Emerg Crit Care (San Antonio) 2009;19:30-52.