H thy that frequently presents a diagnostic dilemma to

Transcription

1 Ultrasonographic Characteristics of the Adrenal Glands in Dogs With Pituitary-Dependent Hyperadrenocorticism: Comparison With Normal Dogs Amy M. Grooters, David S. Biller, Sharon K. Theisen, and Takayoshi Miyabayashi Ultrasonographic evaluation of the adrenal glands was performed in 10 dogs with pituitary-dependent hyperadrenocorticism (PDH) and in 10 age- and weight-matched healthy control dogs. Thickness, shape, and echogenicity were determined for each adrenal gland. Adrenal thickness in dogs with PDH (median, 10 mm-left; 8.5 mm-right) was significantly greater than thickness in control dogs (median, 6 mm-left; 6 mm-right). Other ultrasonographic characteristics associated with PDH included bilaterally symmetrical adrenomegaly and maintenance of normal adrenal shape. Adrenal echogenicity was homogeneous and less than that yperadrenocorticism is a common canine endocrinopa- H thy that frequently presents a diagnostic dilemma to the small animal practitioner. Poor specificity of screening tests, such as the ACTH stimulation and low-dose dexamethasone suppression tests, often makes it difficult to distinguish between hyperadrenocorticism and nonadrenal disorders that cause similar clinical signs. Furthermore, differentiation between pituitary and adrenal causes of hyperadrenocorticism is hindered by the fact that in 15% to 40% of dogs with pituitary-dependent hyperadrenocorticism (PDH), plasma cortisol concentrations do not suppress after administration of a high dose of dexametha~one.'.~-~ Although sonographic identification of the adrenal glands in small animals was at one time considered rare, ultrasonography has since become a valuable tool for adrenal examinati~n.~.~ It has most often been used to distinguish functional adrenocortical neoplasia from PDH through the identification of large adrenal masses.*-" Recently, increasing sonographer experience has allowed more frequent identification of subtle adrenal lesions, such as bilateral adrenocortical hyperpla- ~ia.'.~ At many referral centers, ultrasonography is now well established as a discriminatory test for the differentiation of pituitary and adrenal causes of hyperadrenocorticism. Our clinical experience, as well as that of other authors:'* suggests that ultrasonography has the potential to differentiate bilateral cortical hyperplasia from normal adrenal glands, as well as from adrenocortical neoplasia. The adrenal glands can be completely evaluated in most dogs during routine abdominal ultrasonography.13 This is From the Department of Veterinary Clinical Sciences, The Ohio State University, College of Veterinary Medicine, Columbus, OH. Accepted November 6, Supported by a grant from the Canine Research College Fund at The Ohio State University. The authors thank Barbara Sauerbrei for preparation of the photographs. Reprint requests: Amy Grooters, DVM, Department of Veterinary Clinical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA Copyright 0 I996 by the American College of Veterinary Internal Medicine /96/I $3.00/0 of the adjacent renal cortex in 8 of 10 dogs with PDH and in 10 of 10 control dogs. Heterogenous echogenicity was present in 2 of 10 dogs with PDH, and was associated with nodular cortical hyperplasia in one of those dogs. Results of this study confirm the difference in sonographic appearance between PDH-induced bilateral cortical hyperplasia and functional adrenocortical neoplasia, and show a difference in sonographically determined adrenal size between healthy dogs and dogs with PDH. J Vet Intern Med 1996;10: Copyright by the American College of Veterinary Internal Medicine. especially true in small breed dogs.i3 When using ultrasound to assess adrenal size, measurement of the dorsoventral dimension (thickness) appears to correlate better with gross measurements than do length or width.i4 This is likely because of the complex shape of the gland, which makes it difficult to obtain consistent measurements in longitudinal and mediolateral planes. Normal adrenal thickness has been reported to be 3 to 5 mm in young adult medium-sized dogs.14 However, because the gross size and structure of the adrenal cortex varies with age and body weight,i5.l6 it is likely that sonographically determined adrenal thickness also varies with respect to these parameters. The ultrasonographic characteristics of normal canine adrenal glands and of adrenal neoplasia have been well documented However, although the ultrasonographic appearance of the adrenal glands in dogs with PDH has been briefly described in isolated cases?' consistent ultrasonographic characteristics have not been established. Delineation of these characteristics is needed to assess the ability of ultrasound to differentiate bilateral adrenocortical hyperplasia from adrenocortical neoplasia, as well as from normal adrenal glands. The purpose of this study was to ultrasonographically evaluate the size and appearance of the adrenal glands in dogs with PDH, to compare them with the same parameters in healthy dogs of a similar age and body weight, and to compare them with previous descriptions of the ultrasonographic characteristics associated with functioning adrenocortical neoplasia. Materials and Methods Patients Ten dogs with untreated PDH evaluated at our hospital over a 30-month period were prospectively examined. Ages ranged from 7 to 13 years (mean, 10.2 years), and body weights ranged from 10 to 37.4 kg (mean, 23.0 kg). Hyperadrenocorticism was suspected in each dog on the basis of supportive historical and physical examination findings (ie, polyuria, polydipsia, pendulous abdomen, hepatomegaly, dermatologic changes), and the results of a CBC (ie, eosinopenia, lymphopenia), serum biochemistries (ie, high alkaline phosphatase activity), and urinalysis (ie, dilute specific gravity). Hyperadrenocorticism was further supported by abnormal results of a low-dose dexamethasone suppression test or ACTH stimulation test. The diagnosis of PDH was confirmed in each dog on the basis 110 Journal of Veterinary Internal Medicine, Vol 10, No 3 (May-June), 1996: pp

2 ADRENAL ULTRASONOGRAPHY IN CANINE HYPERADRENOCORTICISM 111 Fig 1. Longitudinal sonogram of the left adrenal gland of a healthy control dog imaged in a sagittal plane. Adrenal thickness is represented by the + calipers (Ao, aorta). of one or more of the following (1) suppression of plasma cortisol concentrations to less than 1.O pg/dl 8 hours after IV administration of 0.1 mgkg dexamethasone (dogs 1, 3, 6, 7, 9); (2) endogenous plasma ACTH concentrations greater than 25 pg/ml (dogs 1, 3, 8); (3) histological evidence of adrenal cortical hyperplasia (dogs 3, 9); (4) suppression of plasma cortisol concentrations to less than 1.O pg/ dl 4 hours after IV administration of mgkg dexamethasone, followed by an escape to over 1.5 pg/dl at 8 hours (dog 4); or (5) high sensitivity to treatment with o,p-ddd (dogs 2, 5, 10). The last criterion included dogs that demonstrated both of the following: (I) induction of adrenocortical hypofunction (shown by plasma cortisol concentrations <0.45 pg/dl before and after administration of ACTH) by the administration of 50 mgkg/d of o,p-ddd (mitotane) for less than 10 days, and (2) good to excellent response (based on clinical improvement and post- ACTH plasma cortisol concentrations below 6 yg/dl) to maintenance o,p-ddd therapy at standard doses (50 to 75 mgkg/wk). Matched Controls Ten clinically normal dogs owned by employees of the hospital were used as controls. Each dog was found to be healthy on the basis of history, physical examination, and serum biochemical analysis. Control dogs were matched to PDH dogs on the basis of body weight (within 15%) and age (within 12 months). Ultrasonography Technique A real-time mechanical sector scanner (Ultramark 4; Advanced Technology Laboratories, Inc, Bothel, WA) with 7.5- and 5-MHz transducers was used. The axial and lateral resolution of the transducers was determined using an ultrasound phantom. The dogs were placed in dorsal recumbency and were not sedated. Imaging of the adrenal glands was performed by one of the authors (either D.S.B. or T.M.) using standard techniques that have previously been de- ~cribed. ~. ~ Adrenal thickness was defined as the greatest dorsoventral dimension and was assessed as a single measurement made perpendicular to the long axis of the adrenal gland (Figs 1 and 2) using electronic calipers. Measurements were obtained with the transducer in a subcostal rather than an intercostal position whenever possible, so that adrenal thickness could be measured in the axial plane of the transducer. The shape, contour, and acoustic texture of each adrenal gland were subjectively assessed and described. Statistical Analysis Differences in adrenal thickness between dogs with PDH and control dogs were assessed using the Wilcoxon signed-rank test. Values for the left and right adrenal glands were considered separately. A value of P <.05 was considered significant. Results Both left and right adrenal glands were well-visualized in all 20 dogs. Ultrasonographic adrenal measurements are presented in Table I. Median adrenal thickness was 10 mm (left) and 8.5 mm (right) in dogs with PDH, and 6 mm (left) and 6 mm (right) in control dogs; there was a significant difference between dogs with PDH and matched controls for both left (P =.012) and right (P <,005) adrenal thickness. Using the ultrasound phantom, resolution of the 5-MHz transducer was determined to be 0.8 mm (axial) and 2.1 mm (lateral); resolution of the 7.5-MHz transducer was 0.6 mm (axial) and 1.3 mm (lateral). In the healthy dogs, the left adrenal gland had a peanut shape when imaged in a sagittal plane (Fig 1). The right adrenal gland had an oval shape when imaged in a sagittal plane, and a V shape when imaged in a mediolateral plane. The contour of both adrenal glands was smooth, and the adrenocortical parenchyma was homogeneous and less echogenic than the adjacent renal cortex. In dogs with PDH, the adrenal margins appeared more rounded, and the glands subjectively appeared thicker, giving them a plump appearance compared with the glands of the normal dogs (Figs 2-4). The degree of adrenomegaly in each dog with PDH was subjectively similar in the left and right adrenal glands. Despite the increased thickness, the basic shape and contour of the adrenal glands in dogs with PDH was the same as that described in the healthy dogs. In 8 of 10 dogs with PDH, the adrenocortical parenchyma was homogeneous, and was hypoechoic in comparison with the adjacent renal cortex. However, in 2 dogs with PDH (dogs Fig 2. Longitudinal sonogram of the left adrenal gland of a dog with pituitary-dependent hyperadrenocorticism imaged in a sagittal plane. Adrenal thickness is represented by the x calipers; The long axis of the adrenal gland is represented by the + calipers.

3 112 GROOTERS ET AL Table 1. Ultrasonographic Measurements of Adrenal Thickness in 10 Dogs with Pituitary-Dependent Hyperadrenocorticism and in 10 Healthy Age- and Weight-Matched Control Dogs Dogs With PDH Control Dogs Left Adrenal Right Adrenal Left Adrenal Right Adrenal Age Weight Thickness Thickness Weight Thickness Thickness Dog (yr) (kg) (mm) (rnrn) (kg) (mm) (rnrn) Median and 6), the parenchyma was heterogenous, and contained variably sized focal areas of increased echogenicity (Fig 5). In one of these dogs (dog 3), histological examination of the adrenal glands indicated bilateral nodular cortical hyperplasia. Focal areas of suspected mineralization within the adrenal parenchyma were identified sonographically in dogs 3 and 6, and were confirmed histologically in dog 3. Discussion Results of this study indicate that sonographically determined adrenal thickness is significantly greater in dogs with pituitary-dependent hyperadrenocorticism than in healthy dogs of a similar age and weight. This is consistent with previous descriptions of the gross pathological findings associated with PDH." As well as being the most accurate sonographic estimation of gross adrenal size,i4 thickness also appears to be the dimension that is most severely altered by Fig 3. Longitudinal sonogram of the right adrenal gland of a dog with pituitary-dependent hyperadrenocorticism (dog 5) imaged in a mediolateral plane. Note that the normal "V" shape of the gland is maintained despite the increased thickness (CVG, caudal vena cava; Ad, adrenal gland). adrenocortical hyperplasia. In human patients, thickness is the most consistent tomographic measurement of adrenal size, and is the dimension most often changed by adrenal pathology.*' Because thickness normally is the smallest adre- nal dimen~ion,'~ diffuse cortical hyperplasia would be expected to cause a larger percent increase in adrenal thickness than in length or width. In addition, the greater variability in normal adrenal width and length (as demonstrated by gross measurement^)'^ makes it more difficult to document pathologic adrenomegaly based on these measurements. For these reasons, we believe that adrenal thickness, and not length or width, should be used to assess adrenal size in dogs suspected of having PDH. Adrenal thickness in the healthy control dogs of this study (median, 6 mm; range, 4-7 mm) was greater" than previously reported values for normal young adult mediumsized dogs (median, 4 mm; range, 2 to 5 mm).i4 This likely reflects age- and weight-associated variation in normal adrenal thickness. Based on our preliminary results in a limited number of dogs, it appears that in middle aged to older dogs, normal adrenal glands may be up to 6 mm thick in small breeds, and up to 7 mm thick in large breeds. Evaluation of adrenal thickness in a large number of healthy dogs of various ages and sizes will be necessary to more accurately establish a range of normal values for sonographically determined adrenal thickness. In addition to increased thickness, the ultrasonographic appearance of the adrenal glands in dogs with PDH was characterized by symmetrical enlargement and maintenance of normal adrenal shape and contour, giving them a "plump" appearance (Figs 2-4). This is consistent with previous gross descriptions of adrenocortical hyperplasia?' The normal canine adrenal gland is flattened dorsoventrally.20~22 In dogs with cortical hyperplasia secondary to * Data compared using the Mann-Whitney rank sum test, P <.05.

4 ADRENAL ULTRASONOGRAPHY IN CANINE HYPERADRENOCORTICISM 113 Fig 4. Longitudinal sonogram of the left adrenal gland of a dog with pituitary-dependent hyperadrenocorticism (dog 3). Adrenal thickness (represented by the arrow) was 19 mm. Despite the severe increase in adrenal gland size, the normal "peanut" shape is maintained. PDH, the shape of the gland becomes more rounded, and with severe hyperplasia has been described as globular.'" The acoustic texture of the adrenal glands in the majority of dogs with PDH was homogeneous and less echogenic than the renal cortex, and thus was similar to that in the healthy dogs. However, in 2 dogs with PDH (dogs 3 and 6), focal areas of hyperechogenicity were present within the adrenal parenchyma (Fig S). These echogenic foci likely represent large nodules of cortical hyperplasia, which was confirmed histologically in dog 3. This is not surprising, because both nodular and diffuse patterns of bilateral cortical hyperplasia have been reported in association with PDH (although the nodular form is observed less frequently).' The results of this study support the use of ultrasonography to differentiate PDH from adrenal-dependent hyperadrenocorticism (ADH). Our findings indicate that the ultrasonographic characteristics of PDH-induced bilateral cortical hyperplasia include increased adrenal thickness, inaintenance of the normal adrenal shape and contour, bilateral symmetry. and homogeneous echogenicity in the majority of cases. In contrast, previous descriptions of the ultrasonographic appearance of adrenocortical neoplasia include variable and often heterogenous echogenicity, distortion of the normal adrenal shape and contour, adrenomegaly that is typically asymmetrical and sometimes severe, and decreased thickness of the contralateral adrenal gland."." Other findings less consistently present include adrenal mineralization, renal displacement, and compression or invasion of the caudal vena It should be noted that sonographically evident adrenal mineralization was confirmed histologically in one of the dogs with PDH in this study (dog 6, Fig 5). Thus, although adrenal mineralization has been reported to be more characteristic of adrenal neoplasia than hyperplasia,23 the presence of mineralization is not specific for neoplasia. The one situation in which ultrasonographic appearance does not appear to be helpful in differentiating PDH from ADH is in the presence of nodular cortical hyperplasia. Large hyperplastic nodules resemble adrenocortical adenomas, both grossly,'" and sonographically." In the 2 dogs of this report with suspected nodular hyperplasia, the focal areas of increased echogenicity did not alter the normal adrenal contour. However, in previous pathological descriptions,'" and in one sonographic description," nodular hyperplasia was observed to disrupt the adrenal contour. Therefore, it is not possible to differentiate PDH-induced nodular cortical hyperplasia from bilateral adrenocortical neoplasia on the basis of sonographic appearance. Although this reduces the specificity of ultrasound for PDH, this is not a major limitation to the use of ultrasound as a discriminatory test, because nodular hyperplasia occurs in less than 5% of dogs with PDH,7 and bilateral adrenocortical neoplasia is rare, accounting for less than 10% of all functioning adrenocortical tumors In general, ultrasonography appears to be an excellent discriminatory test for the evaluation of dogs with hyperadrenocorticism. When adrenocortical neoplasia is present, ultrasonography indicates the affected side, allowing adrenalectomy via a retroperitoneal approach,".2t and may identify vascular invasion or metastases of adrenocortical carcino- Ultrasound is not associated with the adverse effects that may complicate high-dose dexamethasone administration, and takes much less time to pe~form.~ The major disadvantages of ultrasonography as a discriminatory test are the experience and equipment needed to allow complete evaluation of both adrenal glands in the majority of patients with hyperadrenocorticism. The accuracy of ultrasound compared with the high-dose dexamethasone suppression test for distinguishing between PDH and ADH has yet to be determined. However, in the hands of an experienced sonographer, the only apparent limitations to its accuracy are the inability to distinguish nodular hyperplasia from bilateral adrenocortical neoplasia, and the difficulty of complete adrenal evaluation on the right side in large breed or obese dogs. I' The difference in adrenal thickness between normal dogs Fig 5. Longitudinal sonogram of the right adrenal gland of a dog with pituitary-dependent hyperadrenocorticism (dog 61 and suspected nodular cortical hyperplasia. Note the presence of a focal round area of increased echogenicity, suspected to be a hyperplastic nodule (Ndl. and 2 areas of suspected mineralization (Mn) that produce linear echogenicities within the adrenocortical parenchyma.

5 114 GROOTERS ET AL and those with PDH suggests that ultrasonography may have a role as a screening test for hyperadrenocorticism. However, like the ACTH stimulation and low-dose dexainethasone suppression tests, it is probable that sonographically determined adrenal thickness is altered by adaptation of the pituitary-adrenal axis to the stress of chronic nonadrenal illness, which will limit its specificity for hyperadrenocorticism. Although the specificity of sonographically determined adrenomegaly for PDH has been reported to be 100% by one investigator, we believe that this value was falsely elevated by the use of a control population that was not suspected of having hyperadrenocorticism. A second concern about the use of ultrasonography as a screening test for hyperadrenocorticism is its potentially inadequate sensitivity. In 3 of the dogs with PDH in our study, the thickness of one or both adrenal glands was within the range of values for the healthy control dogs, indicating that normal adrenal thickness is not sufficient to rule out PDH. Finally, it must be remembered that ultrasonography cannot distinguish functional adrenocortical neoplasia from nonfunctional adrenocortical tumors or pheochromocytomas. Therefore, as a screening test, ultrasound should be used to complement, and not to replace, functional tests of the pituitary-adrenal axis. Limitations of our study include failure to perform routine discriminatory tests in all of the study patients (which would have allowed better direct comparison of adrenal ultrasound and the HDDS), and the use of o,p-ddd sensitivity to classify 3 dogs as having pituitary-dependent disease. Feldman et alz6 reported that dogs with ADH were more resistant to o,p-ddd therapy than dogs with PDH, as shown by higher post-acth plasma cortisol concentrations during induction and maintenance therapy, higher required maintenance doses, and poorer clinical response. Although they concluded that the standard o,p-ddd treatment protocol for PDH was ineffective in the majority of dogs with ADH, there were a small number of dogs with ADH (1 out of 13, 7.7%) in which the standard protocol resulted in lowering of post- ACTH cortisol concentrations during the induction period. Thus, it is possible that our use of o,p-ddd sensitivity as a criterion for PDH resulted in misclassification of one or more dogs. However, given the fact that these dogs were highly sensitive to induction o,p-ddd therapy (which resulted in temporary adrenocortical hypofunction and discontinuation of o,p-ddd therapy for several weeks in all 3 dogs), we believe that misclassification is unlikely. In conclusion, our results indicate that the ultrasonographic characteristics of PDH-induced bilateral cortical hyperplasia include increased adrenal thickness, bilaterally symmetrical adrenomegaly, maintenance of the normal adrenal shape and contour, and homogeneous echogenicity in the majority of cases. These characteristics are in contrast to previously described sonographic characteristics of functional adrenal neoplasia, and support the routine use of ultrasonography as a discriminatory test in the evaluation of dogs with hyperadrenocorticism. When ultrasound is used in this role, the identification of 2 symmetrically enlarged adrenal glands that have a normal basic shape and contour is highly suggestive of PDH rather than adrenal-dependent hyperadrenocorticism. The identification of 2 normally shaped ad- renal glands that contain focal areas of increased echogenicity is nonspecific, and can be consistent with either bilateral nodular hyperplasia secondary to PDH, or with bilateral adrenocortical neoplasia. Sonographically determined thickness of the adrenal glands is significantly greater in dogs with PDH than in normal dogs. However, the specificity and sensitivity of ultrasonographic parameters for the diagnosis of hyperadrenocorticism may be insufficient to support its use as a screening test. Therefore, at this time, the primary role of adrenal ultrasonography in the evaluation of dogs with hyperadrenocorticism should be for the differentiation of PDH from functional adrenocortical neoplasia. References 1. Peterson ME, Smiley LE. Diagnostic testing for hyperadrenocorticism in dogs. In: Proceedings of the 15th Annual WalthadOSU Symposium (Endocrinology). Columbus, OH, 1991: Chastain CB, Franklin RT, Ganjam VK, et al. Evaluation of the hypothalamic pituitary-adrenal axis in clinically stressed dogs. J Am Anim Hosp 1986;22: Kemppainen RJ, Zenoble RD. Non-dexamethasone-suppressible, pituitary-dependent hyperadrenocorticism in a dog. J Am Vet Med Assoc 1985; 187: Peterson ME. Hyperadrenocorticism. Vet Clin North Am (Small Aniin Pract) 1984; 14: Feldman EC. Distinguishing dogs with functioning adrenocortical tumors from dogs with pituitary-dependent hyperadrenocorticism. J Am Vet Med Assoc 1983; 183: Kantrowitz BM, Nyland TG, Feldman EC. Adrenal ultrasonography in the dog: Detection of tumors and hyperplasia in hype.radrenocorticism. Vet Radiol 1986;27: Feldman EC, Nelson RW. Canine and Feline Endocrinology and Reproduction. Philadelphia, PA: WB Saunders; 1987: Saunders HM, Pugh CR, Rhodes WH. Expanding applications of abdominal ultrasonography. J Am Anim Hosp Assoc 1992;28: Schelling CG. Ultrasonography of the adrenal gland. Prob Vet Med I991 ;3: Reusch CE, Feldman EC. Canine hyperadrenocorticism due to adrenocortical neoplasia. J Vet Intern Med 1991;5: Voorhout G, Rijnberk A, Sjollema BE, et al. Nephrotomography and ultrasonography for the localization of hyperfunctioning adrenocortical tumors in dogs. Am J Vet Res 1990;51: Barthez P, Nyland TG, Feldman EC. Ultrasonographic evaluation of the adrenal glands in normal dogs and in dogs with pituitarydependent hyperadrenocorticism. Proceedings of the 12th ACVIM Forum 1994;991 (abstr). 13. Grooters AM, Biller DS, Miyahayashi T, et al. Evaluation of routine abdominal ultrasonography as a technique for imaging the canine adrenal glands. J Am Anim Hosp Assoc 1994;30: Grooters AM, Biller DS, Merryman J. Ultrasonographic parameters of normal canine adrenal glands: Comparison to necropsy findings. Vet Radiol 1995; 36: Baker DD. Studies of the suprarenal glands of dogs. I. Comparison of the weights of suprarenal glands of mature and immature male and female dogs. Am J Anat 1937;60: Hullinger RL. Adrenal cortex of the dog (Canis familiaris). I. Histomorphologic changes during growth, maturity, and aging. Zbl Vet Med C Anat Histol Embryo1 1978;7: Voorhout G. X-ray-computed tomography, nephrotomography, and ultrasonography of the adrenal glands of healthy dogs. Am J Vet Res 1990;51: Poffenbarger EM, Feeney DA, Hayden DW: Gray-scale ultra-

6 ADRENAL ULTRASONOGRAPHY IN CANINE HYPERADRENOCORTICISM 115 sonography in the diagnosis of adrenal neoplasia in dogs: Six cases ( ). J Am Vet Med Assoc 1988; 192: Glantz SA. Primer of Biostatistics, 3rd ed. New York, NY: McGraw-Hill; 1992: Kelly DF, Siege1 ET, Berg P. The adrenal gland in dogs with hyperadrenocorticism: A pathologic study. Vet Pathol ; 8: Montagne JP, Kressel HY, Korobkin M, et al. Computed tomography of the normal adrenal glands. Am J Roentgen ; 130: Evans HE, Christensen GC. Miller s Anatomy of the Dog, 2nd ed. Philadelphia, PA: WB Saunders; 1979: Penninck DG, Feldman EC, Nyland EC: Radiographic features of canine hyperadrenocorticism caused by autonomously functioning adrenocortical tumors: 23 cases ( ). J Am Vet Med Assoc 1988; 192: Ford SL, Feldman EC, Nelson RW. Hyperadrenocorticism caused by bilateral adrenocortical neoplasia in dogs: Four cases ( ). J Am Vet Med Assoc 1993;202: Johnston DE: Adrenalectomy via retropentoneal approach in dogs. J Vet Med Assoc 1977; 170: Feldman EC, Nelson RW, Feldman MS, et d. Comparison of mitotane treatment for adrenal tumor versus pituitary-dependent hyperadrenocorticism in dogs. J Vet Med Assoc 1992;200: