Chapter 9 - Beta-Cell Neoplasia Insulinoma

CHAPTER 9 Beta-Cell Neoplasia: Insulinoma
Richard W. Nelson
CHAPTER CONTENTS Polycythemia, 358

Etiology, 348 Artifactual Hypoglycemia, 358
Malignant Versus Benign Potential, 348 Iatrogenic Hypoglycemia (Toxicities), 358
Pathophysiology, 349 Diagnostic Approach to Hypoglycemia: Prioritizing the Differentials, 359
Maintenance of Euglycemia in the Healthy Dog, 349 Confirming the Diagnosis of an Insulin-Secreting Beta-Cell Tumor: Serum
Hypoglycemia and the Counterregulatory Response, 350 Insulin Determination, 359
Insulin Secretion in Dogs with Beta-Cell Neoplasia, 351 Whipple’s Triad, 359
Mechanism for Insulin-Induced Hypoglycemia, 351 Determination of Baseline Insulin and Glucose Concentrations, 360
Origin of Clinical Signs, 351 Insulin-to-Glucose Ratios, 360
Clinical Features, 352 Serum Fructosamine Concentration, 361
Signalment, 352 Provocative Testing, 361
History, 353 Diagnostic Imaging, 361
Physical Examination, 353 Radiography, 361
Clinical Pathologic Abnormalities, 354 Ultrasonography, 361
Differential Diagnoses for Fasting Hypoglycemia, 355 Computed Tomography, 362
Congenital Hepatic Disease, 355 Scintigraphy, 363
Acquired Hepatic Dysfunction, 355 Treatment of Beta-Cell Neoplasia, 364
Adrenocortical Insufficiency (Addison’s Disease), 356 Surgical Versus Medical Therapy, 364
Glucagon Deficiency, 356 Perioperative Management of Dogs Undergoing Surgery, 364
Hypopituitarism, 356 Medical Therapy for an Acute Hypoglycemic Crisis, 368
Non–Beta-Cell Tumors, 357 Medical Therapy for Chronic Hypoglycemia, 369
Neonatal and Juvenile Hypoglycemia, 357 Prognosis, 372
Endotoxic or Sepsis-Induced Hypoglycemia, 358 Insulin-Secreting Beta-Cell Tumors
Kidney Failure, 358 in Cats, 372
Insulin-secreting beta-cell tumors (insulinomas) were first described beta-cell tumors has revealed a high incidence of multihormonal
in the dog by Slye and Wells in 1935. During the past seven production, including pancreatic polypeptide, somatostatin, glu-
decades, numerous publications have appeared in the veterinary cagon, serotonin, and gastrin (Hawkins et al, 1987; O’Brien et al,
literature addressing the clinical manifestations, diagnosis, treat- 1987; Minkus et al, 1997). Recently, using quantitative real-time
ment, and pathology of beta-cell tumors in dogs. Insulin-secreting polymerase chain reaction (PCR), significantly higher expression
beta-cell neoplasia is an uncommon diagnosis in dogs and a rare of genes encoding for growth hormone (GH) and insulin-like
entity in cats. Despite excellent documentation of this disease, growth factor-1 (IGF-1) have been identified in metastases, com-
increased awareness of the clinical presentations, and well-estab- pared to the primary beta-cell tumor and immunohistochemical
lished methods for establishing the diagnosis, treatment options examination of the beta-cell tumor and its metastases revealed
remain limited and the prognosis remains guarded to poor. This expression of GH, IGF-1, and GH receptor in both primary beta-
chapter summarizes current concepts regarding the diagnosis and cell tumors and metastases (Buishand et al, 2012). The authors
treatment of insulin-secreting beta-cell tumors in dogs and cats. speculated that therapeutic intervention with agents that specifi-
cally antagonize the GH/IGF-1 axis or members of their signaling
cascade may inhibit beta-cell tumor growth.
ETIOLOGY
Functional tumors arising from the beta cells of the pancreatic Malignant Versus Benign Potential
islets are malignant tumors that secrete insulin independent of
the typically suppressive effects of hypoglycemia. Insulin is the Beta-cell tumors are notorious for masking their malignant ten-
most common product demonstrated in the neoplastic cells, and dencies in the dog. Discrepancy is noted between the orderly
the clinical signs in such animals are primarily those that result arrangement of well-differentiated cells, the rarity of mitotic
from insulin-induced hypoglycemia. Beta-cell tumors, however, figures in most islet cell tumors, and the frequent metastasis of
are not completely autonomous, and they respond to provocative beta-cell tumors at the time of diagnosis (Kruth et al, 1982). Clas-
stimuli, such as an increase in blood glucose by secreting insu- sifying beta-cell tumors as adenomas or adenocarcinomas based
lin, often in excessive amounts. Immunohistochemical analysis of on their morphology often does not reflect their biologic behavior
348
CHAPTER 9 | Beta-Cell Neoplasia: Insulinoma 349
in humans and dogs (Mehlhaff, et al, 1985; Minkus et al, 1997). Plasma glucose
Differentiation of malignant from benign neoplasia is usually and/or INSULIN
caloric intake
based on identification of metastasis at surgery or necropsy or the
recurrence of hyperinsulinism and hypoglycemia days to months
after surgical removal of a “solitary” pancreatic mass. Recent his- AA mobilization Glycogenolysis Lipolysis
Gluconeogenic enzymes
topathologic evaluation of beta-cell tumors in 26 dogs revealed
that 96% were highly cellular and exhibited nuclear atypia and
83% had an infiltrative growth pattern (Buishand et al, 2010).
Glucose production Glucose utilization
Vascular invasion was common, but the mitotic index was low (liver) (fat, muscle)
in most tumors. Increased stromal fibrosis within the tumor was
the only significant morphological prognostic marker identified in
the study, a finding that illustrates the general lack of prognostic EUGLYCEMIA
significance of histopathologic criteria in beta-cell tumors (Kruth FIGURE 9-1 Hormonal and substrate changes by which euglycemia is main-
et al, 1982; Mehlhaff et al, 1985; Caywood et al, 1988). tained (and hypoglycemia is prevented) in normal subjects during a fast. The
The malignant potential of beta-cell tumors is often underesti- fall in the plasma insulin concentration is the key hormonal change resulting in
mated in the dog. In our experience, virtually all beta-cell tumors increased glucose production and decreased glucose utilization. The decline in
in dogs are malignant, and most animals have microscopic or the plasma insulin concentration is, in turn, a result of a small decrease in the
grossly visible metastatic lesions at the time of surgery. The most plasma glucose level (5 to 10 mg/dL) and/or a decrease in caloric intake. AA,
common sites of tumor spread are the regional lymphatics and Amino acid.
lymph nodes (duodenal, mesenteric, hepatic, splenic), the liver,
and the peripancreatic omentum. Pulmonary metastasis is typi-
cally not recognized until very late in the disease process. Identi- glycerol) delivered to the liver from peripheral stores. Muscle
fication of distant metastasis such as the gastrointestinal tract or and other structural tissues supply amino acids, mainly alanine;
bone marrow or gross invasion of the tumor into major blood blood cell elements supply lactate, the end product of glycolytic
vessels with tumor thrombus formation is uncommon (Pickens metabolism; and adipose tissue supplies glycerol from lipolysis of
et al, 2005; Hambrook and Kudnig, 2012). In most dogs, hypo- triglycerides (Karam, 2001). Oxidation of free fatty acids released
glycemia recurs days to weeks after surgical excision of the tumor. from adipose cells during lipolysis supplies the energy required
The high incidence of metastasis at the time afflicted dogs are ini- for gluconeogenesis and provides ketone bodies (i.e., acetoacetate,
tially examined results in part from the typically protracted time β-hydroxybutyrate), which can serve as alternative metabolic fuels
for worrisome clinical signs (e.g., collapsing episodes, seizures) for the brain during periods of prolonged fasting. Other require-
to develop and become apparent to the owner and the interval ments include a normal hepatic circulation, functioning hepato-
between the time an owner initially observes signs and when assis- cytes capable of removing substrates from the circulation, and a
tance is sought from a veterinarian. Most dogs are symptomatic complete complement of hepatic enzymes capable of converting
for 1 to 3 months before being brought to a veterinarian. these noncarbohydrate precursors into glucose.
The renal cortex also has the requisite enzymes for the produc-
tion and release of glucose into the circulation, albeit the con-
PATHOPHYSIOLOGY
tribution is only about 5% during fasting (Stumvoll et al, 1995;
Gerich et al, 2001). However, renal glucose production is regu-
Maintenance of Euglycemia in the Healthy Dog
lated and under certain circumstances (e.g., glucose counterregu-
Glucose is essentially the sole metabolic fuel for the brain under nor- lation, hepatic insufficiency) the contribution of glucose derived
mal physiologic conditions. Survival of the brain requires a continu- from renal gluconeogenesis can be as high as 40%. The kidney
ous supply of glucose from the circulation which, in turn, requires does not have glycogen stores and depends on gluconeogenesis
maintenance of blood glucose concentration within or above the as its only source of glucose production. Glutamine rather than
physiologic range. Glucose is derived from three sources: intestinal alanine is the predominant amino acid substrate for renal gluco-
absorption that occurs after digestion of dietary carbohydrates; gly- neogenesis. In addition to its contribution to glucose homeostasis
cogenolysis, which is the breakdown of glycogen; and gluconeogen- during fasting, the kidney has been shown to be an important
esis, which is the formation of glucose from precursors including contributor to increasing blood glucose (i.e., glucose counterregu-
lactate, amino acids (especially alanine and glutamine), and glycerol lation) in the event of hypoglycemia. Although glucagon does not
(Cryer, 2011). The liver is the major source of net endogenous glu- affect the kidney, the counterregulatory increase in epinephrine
cose production through glycogenolysis and gluconeogenesis. has been shown to stimulate gluconeogenesis in the renal cortex
Exogenously derived energy, in the form of ingested carbohy- (Stumvoll et al, 1995; Gerich et al, 2001).
drate, fat, and protein, provides enough fuel for 4 to 8 hours of A normally functioning endocrine system is also necessary to
cell metabolism. After this postprandial period, fuel for cellular maintain glucose homeostasis and prevent hypoglycemia. Rates of
metabolism must be derived from endogenous sources, primarily endogenous glucose influx into the circulation and glucose efflux
through production of glucose by the liver (Fig. 9-1). The liver out of the circulation into tissues other than the brain are regulated
initially provides glucose by the breakdown of stored hepatic gly- primarily by the blood glucose-lowering hormone insulin and the
cogen (glycogenolysis). Liver glycogen stores are exhausted slowly blood glucose-raising hormones, glucagon and epinephrine, such
in dogs, requiring 2 to 3 days of fasting, compared with only that systemic glucose balance is maintained, hypoglycemia and
24 hours of fasting in humans (de Bruijne et al, 1981). Hepatic hyperglycemia are prevented, and a continuous supply of glucose
glucose production is augmented by gluconeogenesis as the post- to the brain is ensured (Cryer, 2011). In humans, when the blood
prandial period increases and hepatic glycogen stores become glucose concentration exceeds approximately 110 mg/dL (6.2

depleted (Rothman et al, 1991). Gluconeogenesis is the forma- mmol/L), insulin is secreted and the blood glucose concentration
tion of glucose from precursors (e.g., alanine, glutamine, lactate, declines into the normal physiologic range (i.e., 70 to 110 mg/dL;
350 SECTION 3 THE ENDOCRINE PANCREAS
BOX 9-1 A
utonomic Nervous System Response BOX 9-2 Insulin and Counterregulatory Hormonal
to Hypoglycemia Response to Hypoglycemia
Alpha-Adrenergic Effects Insulin: Decreased Secretion

Inhibition of endogenous insulin secretion Reduced stimulation of beta cells by low glucose
Stimulation of peripheral vasoconstriction causing increase in cerebral Inhibition of insulin secretion by alpha-adrenergic nervous system and
blood flow adrenomedullary catecholamines
Primary glucose regulatory factor, first defense against hypoglycemia
Beta-Adrenergic Effects
Stimulation of hepatic and muscle glycogenolysis Glucagon: Increased Secretion
Stimulation of plasma glucagon secretion Direct stimulation of alpha cells by low glucose
Stimulation of lipolysis generating free fatty acids Stimulation of glucagon secretion by beta-adrenergic nervous system and
Impairment of glucose uptake by muscle adrenomedullary catecholamines
Increase in cardiac output causing increase in cerebral blood flow Primary glucose counterregulatory factor, second defense against
hypoglycemia
Adrenomedullary Catecholamine Effects
Augmentation of alpha- and beta-adrenergic effects Catecholamines: Increased Secretion
Direct stimulation of sympathetic nervous system by low glucose
Cholinergic Effects
Direct secretion from the adrenal medulla in response to low glucose
Stimulation of pancreatic polypeptide secretion
Involved, third defense against hypoglycemia
Increase gastric motility
Produce hunger Adrenocorticotropic Hormone and Cortisol: Increased Secretion
Direct stimulation of pituitary adrenocorticotropic hormone (ACTH) secre-
Adapted from Karam JH: Hypoglycemic disorders. In Gardner DA, Shoback D, editors:
tion in response to low glucose
Greenspan’s basic & clinical endocrinology, ed 9, New York, 2011, McGraw-Hill,
Stimulation of pituitary-adrenocortical axis by the sympathetic nervous
with permission.
system
Involved but not critical
3.9 to 6.2 mmol/L). When the blood glucose concentration decreases Growth Hormone: Increased Secretion
toward the lower limit of the normal physiologic range, insulin syn- Direct stimulation of growth hormone (GH) in response to low glucose
thesis and secretion is inhibited, which limits tissue utilization of glu- Involved but not critical
cose and allows the blood glucose concentration to increase. If the
blood glucose concentration decreases below the normal reference
range, increased secretion of glucose counterregulatory hormones,
most notably glucagon and epinephrine, increase the blood glucose events leading to endogenous glucose production is initiated by a
concentration back into the normal physiologic range. blood glucose concentration that decreases below the normal physi-
The signaling pathways in the pancreatic beta cell provide the ologic range and the hormonal and neurogenic response to hypo-
mechanism whereby insulin secretion rates respond to changes glycemia intensifies as the hypoglycemia becomes more severe.
in blood glucose concentration. Glucose enters the beta cell by Insulin is the dominant glucose-lowering hormone. Hypoglyce-
facilitated diffusion mediated by glucose transporter 2 (GLUT-2) mia suppresses insulin secretion, which facilitates the mobilization
(see Origin of Clinical Signs). Intracellular glucose is phosphory- of energy from existing energy stores (glycogenolysis, lipolysis), pro-
lated to glucose-6-phospate by the enzyme glucokinase. Evidence motes hepatic gluconeogenesis and ketogenesis, promotes renal glu-
suggests that glucokinase, by determining the rate of glycolysis, coneogenesis, and decreases glucose utilization by insulin-dependent
functions as the glucose sensor of the beta cell and is the pri- tissues (Cryer and Polonsky, 1998; Karam, 2001). Glucose-raising
mary mechanism by which the rate of insulin secretion adapts to or counterregulatory hormones include glucagon, epinephrine, GH,
changes in blood glucose (Buse et al, 2011). An increase in blood and cortisol (see Box 9-2). Insulin-induced hypoglycemia causes an
glucose levels results in an increase in the rate of glycolysis and a increase in plasma glucagon, epinephrine, and norepinephrine con-
corresponding increase in the rate of insulin secretion by the beta centrations at the onset of the glucose counterregulatory response,
cell, and vice versa. See the Insulin Synthesis, Structure, and Regu- with increases in plasma GH and cortisol occurring later. Glucagon
lation section in Chapter 7 for more information on this topic. is the key counterregulatory hormone affecting recovery from acute
hypoglycemia. In response to falling plasma glucose levels, glucagon
is secreted by the alpha cells of the pancreatic islets into the hepatic
Hypoglycemia and the Counterregulatory Response
portal circulation; it acts exclusively on the liver to activate glycoge-
The brain is wholly dependent on plasma glucose and counter- nolysis and gluconeogenesis (Cryer, 2011). Hepatic glucose produc-
acts declining plasma glucose concentrations with a carefully pro- tion increases almost immediately.
grammed neurogenic and hormonal response to mobilize storage The adrenergic-catecholamine response to hypoglycemia also
depots of glycogen and fat and raise the plasma glucose concen- plays a major role in recovery from hypoglycemia. Epinephrine has
tration (Boxes 9-1 and 9-2). Hepatic glycogen reserves and gluco- both direct and indirect effects, which stimulate hepatic glycoge-
neogenesis in the liver and kidney directly supply the brain with nolysis and hepatic and renal gluconeogenesis; provide muscle tis-
glucose, and the mobilization of fatty acids from triglyceride depots sue with an alternative source of fuel by mobilizing muscle glycogen
provides energy for the large mass of skeletal and cardiac muscle, the and stimulating lipolysis; mobilize gluconeogenic precursors (e.g.,
renal cortex, the liver, and other tissues that use fatty acids as basic lactate, alanine, and glycerol); and inhibit glucose utilization by
fuel; this spares glucose for use by tissues that remain dependent on insulin-sensitive tissues (e.g., skeletal muscle) (Cryer, 1993; Karam,
glucose, including the central nervous system (CNS), erythrocytes, 2001). The role of cortisol and GH in the acute response to hypo-
bone marrow, and renal medulla (Karam, 2001). The cascade of glycemia is minimal but cortisol and GH may play roles in the
defense against prolonged hypoglycemia (Boyle and Cryer, 1991). depends on a continuous supply of glucose from sources outside
Cortisol facilitates lipolysis, promotes protein catabolism and the the CNS. If the blood glucose concentration drops below a critical
conversion of amino acids to glucose by the liver and kidney, and level, nervous system dysfunction occurs. In mammals the cerebral
limits glucose utilization by insulin-dependent tissues. Similarly, cortex is the first area to be affected by a shortage of glucose. The
GH promotes lipolysis and antagonizes the action of insulin on metabolically slower vegetative centers in the brainstem have less
glucose utilization in muscle cells. However, the hyperglycemic demand for blood glucose and are affected after the cerebral cortex.
effects of cortisol and GH do not appear for several hours after the The entrance of glucose into the neurons of the CNS occurs
hypoglycemic episode (Cryer and Polonsky, 1998). primarily by diffusion and is not insulin dependent. Because
The adrenergic neurogenic response to hypoglycemia acts cell membranes are impermeable to hydrophilic molecules (e.g.,
directly to raise the blood glucose concentration and to stimulate glucose), all cells require carrier proteins to transport glucose
hormonal responses that augment the adrenergic mobilization of across the lipid bilayers into the cytosol. All cells except those in
energy stores (see Box 9-1). In dogs, hepatic glucose autoregula- the intestine and kidney have non–energy-dependent transporters
tion is also an important glucose counterregulatory factor (Cryer that facilitate diffusion of glucose across cell membranes. To date,
and Polonsky, 1998). That is, the rate of hepatic glucose produc- fourteen facilitative transporters have been identified in humans,
tion is an inverse function of the blood glucose concentration and they include transporters for substrates other than glucose,
independent of hormonal and neural regulatory factors. including fructose, myoinositol, and urate (Thorens and Mueck-
ler, 2010). The primary transporters involved in facilitative dif-
fusion of glucose into cells are called GLUT-1 through GLUT-4,
Insulin Secretion in Dogs with Beta-Cell Neoplasia
with the numbers designating the order of their identity (Table
In the dog or cat with an insulin-secreting tumor, neoplastic beta 9-1). GLUT-1 is present in all tissues, has a very high affinity for
cells autonomously synthesize and release insulin despite hypogly- glucose, and appears to mediate basal glucose uptake. GLUT-1
cemia. As a result, tissue utilization of glucose continues, hypogly- is an important component of the brain vascular system (blood
cemia progressively worsens, and clinical signs eventually appear. brain barrier) that ensures adequate transport of plasma glucose
The onset of clinical signs is related to both the degree of hypo- into the CNS (Fig. 9-2). GLUT-2 has very high expression in
glycemia achieved and the rate at which it occurs. For example, a pancreatic beta cells and the basolateral membranes of intestinal
blood glucose concentration that gradually drops to 35 mg/dL (2 and renal epithelial cells and hepatocytes. GLUT-3 is the major
mmol/L) over an extended period (i.e., weeks) is much less likely glucose transporter on the neuronal surface, has a very high affin-
to result in signs of hypoglycemia than is a blood glucose concen- ity for glucose, and is responsible for transferring glucose from the
tration of 35 mg/dL that develops rapidly over a few hours. cerebrospinal fluid (CSF) into the neuronal cells. GLUT-4 is the
Failure of insulin secretion to decrease during periods of hypo- major glucose transporter in adipocytes and skeletal muscle.
glycemia predisposes a dog or cat with a beta-cell tumor to develop Blood insulin concentrations do not affect neuronal glucose
clinical signs of hypoglycemia during fasting and exercise. Insulin- transport or utilization. However, if hyperinsulinemia results in an
secreting beta-cell tumors also remain responsive to many of the inadequate glucose supply for intracellular oxidative processes in
stimuli that promote insulin secretion in the healthy dog or cat, neurons, a resultant decline occurs in energy-rich phosphorylated
but the secretory response is often exaggerated, resulting in severe compounds (adenosine triphosphate [ATP]) in neurons. This in
hypoglycemia. For example, clinical signs of hypoglycemia may turn results in cellular changes typical of hypoxia, increased vas-
occur after consumption of food that is easily digestible and rap- cular permeability, vasospasm, vascular dilation, and edema. Neu-
idly absorbed or rapid intravenous (IV) administration of glucose ron death from anoxia follows. In acute hypoglycemia, histologic
to correct hypoglycemia. alterations are most marked in the cerebral cortex, basal ganglia,
hippocampus, and vasomotor centers (see Feldman and Nelson
[1987] for references). Although most of the damage from hypo-
Mechanism for Insulin-Induced Hypoglycemia
glycemia occurs in the brain, peripheral nerve degeneration and
Insulin-secreting tumors and the associated hyperinsulinemia
interfere with glucose homeostasis by decreasing the rate of glucose
release from the liver and increasing the utilization of glucose by TABLE 9-1 GLUCOSE TRANSPORTERS
insulin-sensitive tissues (e.g., muscle, adipose tissue). Insulin inter- IDENTIFIED IN HUMANS
feres with mechanisms that promote hepatic glucose output by
limiting circulating concentrations of substrates needed for gluco- NAME MAJOR SITES OF EXPRESSION AFFINITY FOR GLUCOSE*
neogenesis. This effect is accomplished by inhibiting enzymes nec- GLUT-1 Brain vasculature, red blood High
essary for mobilizing amino acids from muscle and glycerol from cells, all tissues (Km = 1 mmol/L)
adipose tissue. In addition, insulin decreases the activity of hepatic GLUT-2 Liver, pancreatic B cells, serosal Low
enzymes used in gluconeogenesis and glycogenolysis. Insulin also surfaces of gut and kidney (Km = 15-20 mmol/L)
lowers blood glucose concentrations by stimulating glucose uptake
GLUT-3 Brain neurons, also found in all High
and utilization in the liver, muscle, and adipose tissue. In essence,
tissues (Km < 1 mmol/L)
insulin increases tissue utilization of glucose already present in the
extracellular space while interfering with hepatic production of GLUT-4 Muscle, fat cells Medium
glucose. The net effect is decreasing blood glucose concentrations (Km = 2.5 to 5 mmol/L)
because of increased tissue utilization of glucose.
From Masharani U, Karam JH: Pancreatic hormones and diabetes mellitus. In Greenspan
FS, Gardner DG, editors: Basic and clinical endocrinology, ed 6, New York, 2001, Lange
Origin of Clinical Signs Medical Books/McGraw-Hill, p. 630.
GLUT, Glucose transporter.
Glucose is the primary fuel used by the CNS. Carbohydrate *Km represents the level of blood glucose at which the transporter has reached one-half
reserves in neural tissue are limited, and function of these cells of its maximum capacity to transport glucose. It is inversely proportional to the affinity.
demyelination are sometimes encountered (Braund et al, 1987). neuroglycopenic signs common to dogs include lethargy, weak-
Other major organ systems, such as the heart, kidneys, and liver, ness, ataxia, disorientation, abnormal behavior, and seizures (Table
also depend on glucose. However, an acute decrease in the blood 9-2). Clinical signs resulting from stimulation of the sympathoad-
glucose concentration results in clinical signs that involve the renal system include muscle tremors, shaking, nervousness, and
CNS before signs of any other major organ system dysfunction restlessness. In humans, the symptoms related to release of cat-
become apparent. echolamines often precede those of neuroglycopenia and act as an
Prolonged, severe hypoglycemia may result in irreversible early warning sign of an impending hypoglycemic attack (Karam,
brain damage; however, it is uncommon for a dog to die during 2001). This illustrates the rapid response of catecholamine secre-
a hypoglycemic episode. Hypoglycemia is a potent stimulus for tion to hypoglycemia and partly explains why canine patients with
the release of the counterregulatory hormones that function to insulin-secreting tumors do not always progress to generalized sei-
antagonize the effects of insulin and stimulate an increase in the zure activity during a fast.
blood glucose concentration (see Hypoglycemia and the Counter-
regulatory Response). CLINICAL FEATURES
The clinical manifestations of hypoglycemia are believed to
result from both a lack of glucose supply to the brain (neurogly- Signalment
copenia) and stimulation of the sympathoadrenal system. The
Insulin-secreting tumors typically occur in middle-aged or older
dogs. The mean age at the time of diagnosis of an insulin-secreting
CHRONIC HYPERGLYCEMIA tumor in 123 dogs in our series at University of California, Davis
Blood brain (UC Davis), was 9 years, with a median age of 10 years and an age
barrier range of 3 to 14 years (Fig. 9-3). There is no gender predilection.
Glucose A variety of breeds have been diagnosed with an insulin-secreting
GLUT-3 tumor at our hospital (Table 9-3). Labrador Retrievers and Golden
Glucose Neurons
GLUT-3 Retrievers are the breeds most commonly diagnosed with this dis-
Glucose GLUT-1 Glucose ease, which is probably a reflection of breed popularity in our region
GLUT-3
Glucose rather than a breed predisposition per se. In general, insulin-secreting
Glucose
Glucose TABLE 9-2 C
LINICAL SIGNS ASSOCIATED
WITH INSULIN-SECRETING
NORMAL GLUCOSE LEVELS TUMORS IN 117 DOGS
Blood brain
barrier CLINICAL SIGN NUMBER OF DOGS PERCENT AGE
Glucose Weakness 73 62
GLUT-3
GLUT-1 Neurons Seizures 61 52
GLUT-3
Glucose GLUT-1 Glucose Collapse 38 32
GLUT-3
GLUT-1 Tremors, shaking 25 21
Glucose Ataxia 22 19
Disoriented, abnormal 22 19
behavior
FREQUENT HYPOGLYCEMIA Lethargy 21 15
Blood brain Polyphagia 6 5
barrier
Weight gain 6 5
GLUT-1
GLUT-3
GLUT-1 Neurons
GLUT-3 20
Glucose GLUT-1 Glucose
GLUT-3
GLUT-1 16
GLUT-1
Number of dogs
GLUT-1 12
FIGURE 9-2 Glycemic regulation of glucose transporters. The center panel de-
picts the normal component of the high-affinity glucose transporter 1 (GLUT-1) 8
on the vascular cells of the central nervous system (CNS) during euglycemia.
An appropriate amount of glucose diffuses across the blood brain barrier and
4
is transported into the neurons by another high-affinity glucose transporter,
GLUT-3. The upper and lower panels, respectively, show adaptation by either
downregulation of GLUT-1 in the face of chronic hyperglycemia (upper panel) or 0
upregulation of GLUT-1 in the presence of chronic hypoglycemia. (From Karam JH: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Hypoglycemic disorders. In Greenspan FS, Gardner DG, editors: Basic and clinical FIGURE 9-3 Age of 123 dogs at the time of initial diagnosis of an insulin-
endocrinology, ed 6, New York, 2001, Lange Medical Books/McGraw-Hill, p. 701.) secreting islet cell tumor.
TABLE 9-3 B
REED DISTRIBUTION OF 115 concentrations (i.e., 20 to 30 mg/dL; 1.1 to 1.7 mmol/L) for pro-
DOGS WITH ISLET CELL TUMORS longed periods without clinical signs, and only small additional
changes in the blood glucose concentration are then required
BREED NUMBER OF DOGS PERCENT to produce symptomatic episodes. In these dogs, fasting, excite-
ment, exercise, and eating may trigger the development of clinical
Labrador Retriever 17 15 signs. The “adaptation process” to chronic severe hypoglycemia
Golden Retriever 11 10 is believed to involve “up-regulation” of the high-affinity glucose
Mixed-breeds 9 8 transporter, GLUT-1, on the vascular cells forming the blood
brain barrier (see Fig. 9-2) (Karam, 2001).
German Shepherd dog 7 6
In the healthy, exercising dog, a balance between increased glu-
Boxer 7 6 cose utilization by muscle, decreased glucose utilization by other
Terriers (Fox, Kerry Blue, West 7 6 tissues, and increased glucose production by the liver maintains
Highland White, Norwich) the circulating blood glucose concentration in the normal range,
Poodle 6 5 allowing the brain to continue to function. The exercising dog
with an insulin-secreting tumor has continuing glucose utilization
Irish Setter 6 5 not just by muscle but by all tissues, owing to the autonomous
Cocker Spaniel 6 5 and continuing secretion of insulin. In addition, hepatic release of
Collie 5 4 glucose is impaired. The potential for severe hypoglycemia is great,
Rottweiler 5 4
and this fact is supported by the number of owners who associate
symptoms in their pets with jogging, play, or long walks. A similar
Border Collie 4 3 pathophysiology is thought to explain the development of symp-
Doberman Pinscher 3 3 toms during periods of excitement. Insulin-secreting tumors are
Samoyed 2 2 responsive to increases in the blood glucose concentration, and the
insulin-secretory response can be exaggerated if the dog consumes
Staffordshire Terrier 2 2
food that is easily digestible and rapidly absorbed.
Dachshund 2 2
Other breeds (one dog each) 16 10 Physical Examination
Physical examination findings in dogs with an insulin-secreting
tumor are often surprisingly unremarkable; dogs are usually free
tumors occur more commonly in large breeds of dogs. However, the of visible or palpable abnormalities. Most abnormalities identified
size of the dog should never preclude an investigation for an insulin- on the physical examination are nonspecific (Table 9-4). Weakness
secreting tumor in a hypoglycemic dog. We have diagnosed insulin- and lethargy, the most common findings, are identified in 29% and
secreting tumors in dogs as small as a Pomeranian. 17% of our cases, respectively. Episodes of collapse and seizures may
occur during the examination but are uncommon. Afflicted dogs
History are usually free of palpable abnormalities, aside from findings com-
monly associated with aging. Weight gain is evident in some dogs
Clinical signs of an insulin-secreting tumor may have been and is a result of the anabolic effects of excess insulin in a dog with
observed for more than a year or as briefly as 1 day before veteri- a normal or increased appetite. Failure to identify abnormalities on
nary care is sought. Most dogs, however, are symptomatic for 1 the physical examination, especially in an older, large-breed dog,
to 3 months before being brought to a veterinarian. In our most is an important finding supportive of an insulin-secreting tumor.
recent 30 dogs with an insulin-secreting tumor, clinical signs had
been observed by the owners for an average of 5 weeks (range, 2 Peripheral Neuropathy
days to 4 months) before veterinary care was sought. Peripheral neuropathies have been reported in dogs with insulin-
Clinical signs of an insulin-secreting tumor typically are caused secreting tumors (Shahar et al, 1985; Braund et al, 1987; Van Ham
by hypoglycemia and an increase in circulating catecholamine et al, 1997). Clinical signs and physical examination findings range
concentrations and include weakness, seizures, collapsing epi- from paraparesis to tetraparesis, facial paresis to paralysis, sciatic
sodes, tremors, ataxia, and disorientation (see Table 9-2). One hyporeflexia to areflexia, hypotonia, and muscle atrophy of the
characteristic of hypoglycemic signs, regardless of the cause, is appendicular, masticatory and/or facial muscles. Sensory nerves
their episodic nature. Signs are generally observed intermittently may also be affected. A subclinical polyneuropathy has also been
for only a few seconds to minutes because of the compensatory reported (Braund et al, 1987). The onset of clinical signs may be
counterregulatory mechanisms that usually increase the blood acute (days) or insidious (weeks to months). Abnormalities iden-
glucose concentration after the development of hypoglycemia. If tified on electrodiagnostic testing include abnormal spontaneous
these mechanisms are inadequate, seizures may occur as the blood potentials (e.g., positive sharp waves, fibrillation potentials) and
glucose concentration continues to decrease. Seizures are usually slowed motor nerve conduction velocities (Braund et al, 1987).
self-limiting, typically lasting from 30 seconds to a few minutes. CSF analysis is usually unremarkable (Van Ham et al, 1997).
The seizure may stimulate further catecholamine secretion and Histopathologic findings in motor and sensory nerves include
activation of other counterregulatory mechanisms that increase moderate to severe axonal necrosis, nerve fiber loss, and variable
the blood glucose concentration above critical levels (see Hypo- demyelination-remyelination (Braund et al, 1987; Schrauwen
glycemia and Counterregulatory Response). et al, 1996; Van Ham et al, 1997). Muscle changes reflect neu-
The severity of clinical signs depends on the duration and sever- rogenic atrophy. The pathogenesis of the polyneuropathy is not
ity of the hypoglycemia. Dogs with chronic fasting hypoglycemia known. Proposed theories include metabolic derangements of
or with recurring episodes appear to tolerate low blood glucose the nerves induced by chronic and severe hypoglycemia or some
TABLE 9-4 PHYSICAL EXAMINATION BOX 9-3 A

Partial Listing of the Numerous Disorders
FINDINGS ASSOCIATED WITH that may Result in Weakness
INSULIN-SECRETING TUMORS
IN 78 DOGS Neuromuscular Disorders
Infectious (e.g., canine distemper, bacterial, mycotic)
PHYSICAL EXAMINATION Inflammatory
FINDING NUMBER OF DOGS PERCENTAGE Neoplasia
Weakness 23 29 Cerebrovascular hemorrhage
Spinal disorders (e.g., type II disc disease; discospondylitis)
Lethargy 13 17
Degenerative myelopathy
Collapsing 6 8 Myasthenia gravis
Tremors 5 6 Peripheral neuropathy
Obtunded 5 6 Polymyopathy/polymyositis
Peripheral neuropathy 5 6 Polyarthritis
Seizures 4 5 Immune-mediated
Infectious (e.g., rickettsial)
No abnormalities 43 55 Degenerative
identified
Cardiorespiratory
Congestive heart failure
Tachyarrhythmias/bradyarrhythmias
other tumor-induced metabolic deficiency, an immune-mediated
Heartworm disease
paraneoplastic syndrome resulting from shared antigens between
Bacterial endocarditis
tumor and nerves, or toxic factors produced by the tumor that
Pneumonia (viral, bacterial, mycotic)
deleteriously affect the nerves (Kudo and Noguchi, 1985; Das and
Pleural effusions
Hochberg, 1999; Heckmann et al, 2000). Treatment is aimed at
surgical removal of the beta-cell tumor (Jeffery et al, 1994). Pred- Hematologic Disorders
nisone therapy (initially 0.5 mg/kg every 12 hours) may improve Anemia
clinical signs (Van Ham et al, 1997). Correction of hypoglycemia, Polycythemia
by itself, may or may not improve clinical signs caused by the Hemangiosarcoma
peripheral neuropathy (Bergman et al, 1994). In our experience, Coagulopathy (e.g., warfarin-induced)
the prognosis for improvement in clinical signs is guarded to poor, Metabolic Disorders
although the occasional dog shows remarkable improvement in Hypoglycemia
ambulation following removal of the tumor and reestablishment Hypokalemia
of a normal blood glucose concentration. Hyperkalemia
Hypercalcemia
CLINICAL PATHOLOGIC ABNORMALITIES Insulinoma
Hypothyroidism
Virtually all dogs with insulin-secreting tumors remain undiag- Hypoadrenocorticism
nosed and often unsuspected of having such tumors after com- Hyperadrenocorticism
pletion of the history and physical examination. Most afflicted Pheochromocytoma
dogs have a history of episodic weakness or seizures—signs that Hepatic encephalopathy
encompass a wide variety of disorders (Box 9-3). The minimum
diagnostic evaluation for dogs with these clinical signs should
include a complete blood count (CBC), serum biochemical panel, mmol/L). The median blood glucose concentration was 38 mg/dL
and urinalysis in an effort to identify abnormalities supportive of (2.1 mmol/L). One hundred and eleven of 123 dogs (90%) had
one of the disorders outlined in Table 9-3. Therapy other than a random blood glucose concentration less than 60 mg/dL (3.4
that required in an emergency situation should be withheld until mmol/L). Dogs with insulin-secreting beta-cell tumors may occa-
a diagnosis has been made. sionally have a blood glucose concentration between 60 and 70
Results of the CBC and urinalysis in dogs with an insulin- mg/dL (3.4 and 3.9 mmol/L) on random testing. Such a finding
secreting tumor are usually normal. Results of the serum biochem- does not eliminate hypoglycemia as a cause of episodic weakness
ical profile, aside from the blood glucose concentration, are also or seizure activity. Fasting, with hourly evaluation of the blood
usually normal. Hypoalbuminemia, hypophosphatemia, hypoka- glucose concentration, should be done to induce hypoglycemia
lemia, and increased activity in alkaline phosphatase and alanine in dogs with suspected insulin-secreting beta-cell tumor. The
aminotransferase have been reported (Leifer et al, 1986), but these time required to induce hypoglycemia with fasting depends in
findings are considered nonspecific and not helpful in achiev- part on the extent of disease at the time the dog is examined and
ing a definitive diagnosis. A correlation has not been established ranges from a few hours to longer than 24 hours. Hypoglycemia
between increased liver enzyme activity and obvious metastasis of (blood glucose < 60 mg/dL) will develop in most dogs with an
beta-cell tumors to the liver. insulin-secreting beta-cell tumor within 12 hours of withholding
The only consistent abnormality identified in serum biochemis- food. A fast of 8 or fewer hours was successful in demonstrating
try profiles is hypoglycemia. The mean initial blood glucose con- hypoglycemia in 33 of 35 trials in 31 dogs with insulin-secreting
centration in 123 of our dogs at UC Davis with an insulin-secreting tumor (Kruth et al, 1982). We have had a few dogs require 12 to
tumor was 42 mg/dL, with a range of 15 to 78 mg/dL (0.84 to 4.4 24 hours of fasting before hypoglycemia became apparent and a
couple of dogs that did not develop hypoglycemia after 30 hours common problem resulting from overzealous insulin administra-
of fasting. The clinical signs in these dogs were episodic and mild, tion in diabetic dogs and cats.
and the diagnosis of beta-cell tumor was not established until 2 to
3 months after initial presentation. Congenital Hepatic Disease
Portovascular anomalies are the most common congenital cause
DIFFERENTIAL DIAGNOSES FOR FASTING
of hepatic-induced hypoglycemia. Hypoglycemia develops despite
HYPOGLYCEMIA
an appropriate reduction in circulating insulin because of insuf-
Hypoglycemia is present if the blood glucose concentration is ficient hepatic glycogen stores and inadequate hepatocellular
less than 60 mg/dL (3.4 mmol/L). It typically results from the function to support gluconeogenesis. Abnormalities suggestive of
excessive uptake of glucose by normal cells (e.g., during periods this disorder include microcytosis, hypoalbuminemia, hypocho-
of hyperinsulinism, such as that which occurs with a beta-cell lesterolemia, decreased urea nitrogen, increased total bilirubin,
tumor or xylitol ingestion) or neoplastic cells, impaired hepatic ammonium biurate crystals in the urine, abnormal preprandial
gluconeogenesis and glycogenolysis (e.g., portal shunt, hepatic and postprandial bile acid concentrations, and small liver size on
cirrhosis), a deficiency in diabetogenic hormones (e.g., hypocor- abdominal radiography or ultrasonography. Confirmatory tests
tisolism), an inadequate dietary intake of glucose and other sub- include liver biopsy, angiography, nuclear scintigraphy, and iden-
strates required for hepatic gluconeogenesis (e.g., anorexia in the tification of the shunt during abdominal ultrasound, computed
neonate or toy breeds), or a combination of these mechanisms tomography (CT) scanning or exploratory celiotomy.
(e.g., sepsis; Box 9-4; Service, 1995). Iatrogenic hypoglycemia is a Glycogen storage diseases (GSDs) are rare autosomal recessive
disorders of glycogen metabolism that result from a congenital
absolute or relative deficiency of one of the enzymes necessary to
BOX 9-4 Causes of Hypoglycemia in Dogs and Cats convert glycogen to glucose. In dogs, four breed-specific types of
GSD have been described. Type Ia in Maltese terriers (von Gierke
Beta-cell tumor (insulinoma)*
disease) is a deficiency in glucose-6-phosphatase caused by a
Extra pancreatic neoplasia (see Box 9-5)
mutated, defective glucose-6-phosphatase gene (Brix et al, 1995;
Hepatocellular carcinoma, hepatoma*
Kishnani et al, 1997; 2001). Type II in Lapland dogs (Pompe
Leiomyosarcoma, leiomyoma*
disease) is a deficiency of lysosomal acid α-glucosidase (Walvoort
Hepatobiliary disease*
et al, 1985). Type III in German Shepherd dogs and Curly-Coated
Portosystemic shunts
Retrievers (Cori disease) is a deficiency of glycogen debranching
Chronic fibrosis, cirrhosis
enzyme (amylo-1,6-glucosidase; Fig. 9-4) (Hardy, 1989; Gregory
Hepatic necrosis; toxins, infectious agents
et al, 2007). GSD in Curly-Coated Retrievers affects both liver and
Primary and metastatic neoplasia
muscle (GSD type IIIa) and is caused by a mutation of the glycogen
Sepsis*
debranching enzyme gene (AGL) (Gregory et al, 2007). Type VII in
Severe canine babesiosis
English Springer Spaniels (Tarui disease) is a deficiency in phospho-
Septic peritonitis
fructokinase (Giger et al, 1988; Smith et al, 1996). In cats, GSD
Hypoadrenocorticism*
type IV has been identified in Norwegian Forest cats, is caused by
Primary and secondary
a deficiency in a glycogen branching enzyme (alpha-1,4 glucan 6
Idiopathic hypoglycemia*
glycol transferase), and results in glycogen accumulation in skeletal
Neonatal hypoglycemia
and cardiac muscle and the nervous system (Fyfe et al, 1992).
Juvenile hypoglycemia (especially toy breeds)
Indicators for possible GSD in a juvenile dog include: clinical
Hunting dog hypoglycemia
signs suggestive of hypoglycemia; progressive hepatomegaly char-
Exocrine pancreatic neoplasia
acterized histologically by diffuse, marked hepatocellular vacu-
Pancreatitis
olation caused by glycogen accumulation; and hypoglycemia and
Glucagon deficiency (?)
increased hepatic enzyme activities identified on routine blood
Chronic kidney disease
and urine tests. Serum glucose concentrations typically fail to
Hypopituitarism
increase after injection of glucagon in dogs with type Ia and type
Severe polycythemia
III GSDs. Confirmatory tests include histologic evaluation of
Hepatic enzyme deficiencies
hepatic biopsies and specific hepatic enzyme assays.
Glycogen storage diseases (GSDs)
Severe malnutrition Acquired Hepatic Dysfunction
Prolonged storage of whole blood*
Iatrogenic
Hypoglycemia may result from progressive and severe destruction
Insulin overdose*
of the liver typically caused by primary or metastatic neoplasia or
Sulfonylurea therapy
chronic inflammation leading to fibrosis, cirrhosis, and the devel-
Ethylene glycol ingestion
opment of acquired hepatic vascular shunts. There are numerous
Xylitol ingestion
potential causes of chronic hepatic fibrosis in older dogs and cats.
Alpha lipoic acid
Hypoglycemia results from inadequate amounts of functional
Dried chicken jerky treats
hepatic tissue for adequate storage of glycogen or for sufficient
Artifact*
gluconeogenesis to sustain a normal blood glucose concentration
Portable blood glucose monitoring (PBGM) devices
during a fast. Serum insulin concentrations decline appropri-
Laboratory error
ately with worsening hypoglycemia, but this alone may be insuf-
ficient to prevent problems. Additional abnormalities suggestive
*Common cause. of hepatic insufficiency include microcytosis, hypoalbuminemia,
hypocholesterolemia, decreased urea nitrogen, increased total circulation, and it acts exclusively on the liver to activate glycoge-
bilirubin, ammonium biurate crystals in the urine, abnormal nolysis and gluconeogenesis (Cryer and Polonsky, 1998). Hepatic
preprandial and postprandial bile acid concentrations, abnormal glucose production is increased almost immediately. Abnormalities
liver size on abdominal radiography, or abnormal echotexture or in the production or secretion of glucagon prevent a normal coun-
liver size on ultrasonography. Liver biopsy is helpful in confirm- terregulatory response to decreasing blood glucose concentrations
ing severe fibrosis or cirrhosis and may even identify a cause (e.g., and predispose the animal to hypoglycemia. A classic example is
neoplasia). The etiology of hepatic fibrosis and cirrhosis goes undi- hypoglycemia unawareness in diabetic humans, which is a syn-
agnosed in most cases. drome caused by deficient glucagon and catecholamine secretion,
resulting in defective counterregulation, severe hypoglycemia, and
diabetic coma (Gerich et al, 1991; Mokan et al, 1994; Meyer et al,
Adrenocortical Insufficiency (Addison’s Disease)
1998). Isolated glucagon deficiency that causes hypoglycemia has
Hypoglycemia in dogs with hypoadrenocorticism is caused by been reported in humans but is rare (Cryer and Polonsky, 1998).
insufficient secretion of the glucocorticoids needed to stimulate Typically, glucagon deficiency occurs in conjunction with excess
hepatic mobilization and production of glucose (see Chapter insulin secretion, deficient catecholamine secretion, or increased
12). In this disorder, hypoglycemia occurs despite an appropri- tissue sensitivity to the actions of insulin; and this combination
ate reduction in the blood insulin concentration. Reduced insulin results in hypoglycemia. We have identified hypoglycemia in dogs
secretion must be accompanied by an increase in hepatic gluco- and cats with severe pancreatitis and exocrine pancreatic adenocar-
neogenesis to correct hypoglycemia. Glucose synthesis is normally cinoma, and we speculate that the destructive process associated
stimulated by gluconeogenic hormones. Without secretion of with these disorders may alter the production and/or secretion of
these hormones, as observed in hypoadrenocorticism, hypoglyce- glucagon and insulin.
mia is possible (Sherwin and Felig, 1981). Hypoadrenocorticism
is most common in young and middle-aged dogs, and there is Hypopituitarism
a gender predisposition for the female. Abnormalities on screen-
ing tests supportive of this diagnosis include a relative increase GH and cortisol are glucose counterregulatory hormones involved
in the eosinophil and lymphocyte counts, mild nonregenerative in hepatic glucose synthesis and secretion. Failure of the pituitary
anemia, mild to severe prerenal azotemia, hyperkalemia, hypona- gland to secrete ACTH, GH, or both may impact maintenance of
tremia, and hypercalcemia. Hypoglycemia may also develop with glucose homeostasis and predispose the animal to hypoglycemia,
atypical hypoadrenocorticism, which is characterized by cortisol especially in the fasting state. As in primary hypoadrenocorticism,
but not mineralocorticoid deficiency and normal serum electro- insulin secretion diminishes appropriately for the degree of hypo-
lyte concentrations. The diagnosis of hypoadrenocorticism can be glycemia, but this alone may not be sufficient to prevent clinical
confirmed by abnormal results on adrenocorticotropic hormone signs. GH deficiency is a rare cause of hypoglycemia and is usu-
(ACTH) stimulation test. ally diagnosed in young German Shepherd dogs as a congenital
defect (see Chapter 2). Pituitary failure to secrete ACTH results
in atrophy of the zona fasciculata of the adrenal cortex, impaired
Glucagon Deficiency
secretion of cortisol, and the development of secondary hypoad-
Glucagon is the key counterregulatory hormone affecting recovery renocorticism. No classic abnormalities are found on screening
from acute hypoglycemia (Cryer and Gerich, 1985). In response laboratory studies in animals with secondary hypoadrenocorti-
to falling plasma glucose concentrations, glucagon is secreted cism. These dogs may have a mild nonregenerative anemia and
by the alpha cells of the pancreatic islets into the hepatic portal fasting hypoglycemia, but serum electrolyte concentrations are
GLYCOGEN
2 9
Glucose-1-P
1 8
GLUCOSE Glucose-6-P GLUCOSE
1 Hexokinase/glucokinase Fructose-6-P
Glycogen synthase 3 7
2
Fructose-1,6-P
3 Phosphofructokinase TRIGLYCERIDES
4 Pyruvate kinase
Triose-P
5 Pyruvate carboxylase GLYCEROL
6 Phosphoenolpyruvate
carboxykinase
FATTY ACIDS
7 Fructose-1,6-bisphos- Phosphoenolpyruvate
phatase 4 6
8 Glucose-6-phosphatase Pyruvate
9 Phosphorylase
5
Acetyl CoA
Oxaloacetate FIGURE 9-4 Schematic representation of glucose
TCA metabolism. (From Cryer PE, Polonsky KS: In Wilson
Citrate
cycle KETONES JD, et al, editors: Williams textbook of endocrinology,
LACTATE α-Ketoglutarate ed 9, Philadelphia, 1998, WB Saunders, p. 940.)
Acetyl CoA, Acetyl coenzyme A; TCA, tricarboxylic
ALANINE Glutamine acid.
usually normal. An ACTH stimulation test and determination of carcinoma, hepatocellular carcinoma, and pancreatic neuroendo-
a baseline endogenous ACTH concentration are used to establish crine tumor (Boari et al, 1995; Zini et al, 2007; Finotello et al,
the diagnosis of secondary hypoadrenocorticism (see Chapter 12). 2009; Rossi et al, 2010). Blood glucose and serum insulin concen-
trations were low and serum IGF-2 concentrations were increased
at presentation to the veterinary hospital in all dogs. Blood glu-
Non–Beta-Cell Tumors
cose and serum IGF-2 concentrations returned to the reference
In humans, non–beta-cell tumors that cause hypoglycemia are range in three dogs that underwent surgical removal of the tumor
usually of mesenchymal origin (e.g., leiomyosarcoma, fibrosar- and results of immunohistochemical staining of tumor tissue were
coma). Hypoglycemia is caused less often by tumors of epithe- positive for IGF-2 in all four dogs. Interestingly, the dog with an
lial origin (e.g., hepatoma, carcinoid tumors) and hematopoietic IGF-2-secreting mammary carcinoma was an insulin-dependent
origin (e.g., lymphoma, multiple myeloma) (Cryer and Polonsky, diabetic dog prior to development of the mammary carcinoma;
1998). A variety of tumor types have also been reported to cause this dog developed problems with severe hypoglycemia as the
hypoglycemia in the dog (Box 9-5). In our hospital, hepatocellular tumor enlarged and became diabetic again after surgical removal
carcinoma, hepatoma, leiomyoma, leiomyosarcoma, and tumors of the tumor (Rossi et al, 2010).
with extensive hepatic metastasis are most commonly associated In another study involving four dogs with hypoglycemia caused
with hypoglycemia (Cohen et al, 2003). by smooth muscle tumors, the results of immunohistochemical
The pathogenesis of hypoglycemia associated with non–beta- staining for insulin were negative in the four tumors but positive
cell tumors is undoubtedly multifactorial. Proposed mechanisms for glucagon in three of the four tumors (Beaudry et al, 1995). The
include excessive glucose utilization by the tumor, impaired three smooth muscle tumors that stained positive for glucagon
hepatic glycogenolysis and gluconeogenesis as a result of tumor- originated in either the stomach or jejunum, whereas the tumor
induced hepatic destruction or inhibition of normal counter- that stained negative for glucagon originated in the spleen. Immu-
regulatory responses that prevent hypoglycemia, and secretion of nohistochemical staining for glucagon was negative in smooth
an insulin-like molecule, specifically insulin-like growth factor-2 muscle cells in normal adjacent tissue. The clinical relevance of
(IGF-2) that lowers the blood glucose concentration by enhancing this finding remains unclear, especially considering that glucagon
glucose utilization by normal cells (Cryer and Polonsky, 1998). should increase, not decrease, the blood glucose concentration.
Although the major organ responsible for circulating insulin-like Dogs with hypoglycemia caused by a non–beta-cell tumor may
growth factors is the liver, it has been demonstrated that these fac- be brought to the veterinarian with clinical signs of hypoglyce-
tors are produced ubiquitously, particularly by mesenchymal cells mia, or hypoglycemia may be a serendipitous finding on a serum
(D’Ercole et al, 1984; Barreca et al, 1992). IGF-2 is structurally biochemistry panel. In most dogs, non–beta-cell tumors that
homologous to proinsulin, can bind to insulin receptors, and has cause hypoglycemia are located in the liver or abdomen. Iden-
direct insulin-like actions that result in hypoglycemia (de Groot tification of a non–beta-cell tumor requires a thorough physical
et al, 2007). In addition, IGF-2 may suppress glucagon and GH examination of the dog or cat, thoracic and abdominal radiogra-
secretion, which may also contribute to hypoglycemia (Cryer and phy, abdominal ultrasonography, and histopathologic evaluation
Polonsky, 1998). Serum insulin concentrations are typically unde- of biopsy specimens from identifiable masses. The association
tectable or in the lower end of the reference range with non–beta- between a non–beta-cell tumor and hypoglycemia requires resolu-
cell tumors, in contrast to the high-normal to increased serum tion of hypoglycemia after surgical excision of the tumor.
insulin concentrations seen with hypoglycemia induced by a beta-
cell tumor (Beaudry et al, 1995; Bagley et al, 1996; Bellah and Neonatal and Juvenile Hypoglycemia
Ginn, 1996).
Paraneoplastic hypoglycemia affiliated with IGF-2 secretion The fetus receives a continuous source of glucose via the placenta
has been reported in a dog with gastric leiomyoma, mammary and does not depend on its own gluconeogenic capabilities to
maintain an adequate blood glucose concentration. In contrast,
the neonate depends on glycogenolysis and gluconeogenesis to
BOX 9-5 H
istologic Classification of Non–Beta-Cell maintain euglycemia during fasts, even if brief. Limited hepatic
Tumors Associated with Hypoglycemia glycogen stores, small muscle mass, lack of adipose tissue, and
decreased use of free fatty acids as an alternative energy source
Hepatocellular carcinoma, hepatoma
place the neonate at risk for developing hypoglycemia within
Leiomyosarcoma, leiomyoma
hours of fasting (Chastain, 1990). Impaired gluconeogenesis as
Hemangiosarcoma
a result of delayed induction of one or more of the rate-limiting
Adenocarcinoma
gluconeogenic enzymes is suspected in neonatal hypoglycemia of
Mammary
human infants (Cryer and Polonsky, 1998) and may play a role in
Nasal (horse)
neonatal hypoglycemia of puppies and kittens as well.
Pulmonary
Hypoglycemia often occurs in conjunction with hypothermia,
Salivary
sepsis, starvation, toxic milk syndrome, or a combination of these
Gastric
problems. The ill neonate should always be evaluated for hypogly-
Small intestine
cemia. Orally administered glucose (e.g., 0.01 mL of 5% to 10%
Splenic
solution per gram of body weight) and frequent nursing or bottle
Renal
feeding help correct and prevent hypoglycemia in the neonate.
Lymphoblastic leukemia
Hypoglycemia of toy and miniature breed dogs younger than 6
Plasmacytoma
months of age is common. Alanine deficiency has been implicated
Metastatic melanoma
in this syndrome, as it has in young children (Chew et al, 1982).
Pancreatic neuroendocrine tumor
In humans, the rate of alanine release from muscle determines the
rate of gluconeogenesis during starvation. Puppies with juvenile
hypoglycemia are usually under extreme stress. They frequently may cause hypoglycemia in humans, kidney failure-induced hypo-
have a history of recently being purchased, with an associated glycemia is a very uncommon finding in dogs and cats (Edwards
change in environment and diet. Gastrointestinal upset (vomit- et al, 1987; Cryer, 2011). Proposed mechanisms for development
ing, diarrhea, and/or anorexia) is typical and may or may not be of hypoglycemia in kidney failure include decreased renal glu-
associated with parasites. These puppies are quite fragile and are coneogenesis in conjunction with impaired glucose production
brought to veterinarians with signs that may include weakness, by the liver as a consequence of defective hepatic glycogenolysis
collapse, depression, ataxia, stupor, convulsions, hypothermia, and gluconeogenesis, limited availability of glucogenic substrates,
and/or diarrhea. IV administration of glucose usually results in inadequate glucose counterregulatory responses, decreased caloric
rapid clinical improvement. Frequent feedings prevent recur- intake, decreased renal degradation, and/or excretion of insulin
rences. This disorder virtually disappears with attainment of adult (Fischer et al, 1986; Gerich et al, 2001).
height and weight. If signs persist, a search for another disease that
may be causing the hypoglycemia should be considered. Polycythemia
Severe polycythemia (hematocrit > 65%) may cause artifactual
Endotoxic or Sepsis-Induced Hypoglycemia
hypoglycemia secondary to increased glucose utilization by the
Endotoxic or sepsis-induced hypoglycemia is a relatively uncom- markedly increased number of red blood cells in the blood sample.
mon cause of hypoglycemia in the dog and cat (Breitschwerdt Polycythemia may be primary (i.e., polycythemia vera), or it may
et al, 1981). The pathogenesis of sepsis-induced hypoglycemia is occur secondary to disorders that cause chronic systemic hypoxia
not well characterized but is believed to result from increased tis- (e.g., congenital right-to-left shunting of blood in the heart), to
sue utilization of glucose in conjunction with decreased hepatic chronic renal hypoxia (e.g., renal neoplasia), or to erythropoietin-
glucose production (Naylor and Kronfeld, 1985; Hargrove et al, producing tumors.
1988a; 1988b). Proposed mechanisms for increased glucose uti-
lization include sepsis-induced production of insulin-like sub- Artifactual Hypoglycemia
stances, interleukin-1–enhanced insulin secretion by beta cells,
cytokine-enhanced increase in glucose transport into cells, and Prolonged storage of blood before separation of serum or plasma
increased glucose utilization by bacteria and neutrophils (Com- causes the glucose concentration to decrease at a rate of approxi-
mens et al, 1987; del Rey and Besedovsky, 1987). Increased glu- mately 7 mg/dL/h (0.4 mmol/L/h). Glycolysis by red and white
cose use by macrophage-rich tissues (e.g., the liver, spleen, and blood cells becomes even more apparent in dogs and cats with
ileum) is responsible for most of the glucose utilization (Meszaros erythrocytosis, leukocytosis, or sepsis. Therefore whole blood
et al, 1988), with skeletal muscle accounting for an additional obtained for the measurement of the glucose concentration should
25% (Meszaros et al, 1987). Decreased hepatic glucose production be separated soon after collection (within 30 minutes), and the
may result from impaired hepatic oxidative metabolism, increased serum or plasma should be refrigerated or frozen until the assay is
anaerobic glycolysis of liver glucose, hypoxic injury to hepatic performed to minimize artifactual lowering of the blood glucose
cells, or sepsis-induced interference with substrate delivery to the concentration. Glucose determinations from separated and refrig-
liver (see Feldman and Nelson [1987] for references). Endotoxin erated plasma or serum are reliable for as long as 48 hours after the
may also decrease glycogenolysis through depletion of hepatic and separation and refrigeration of the specimen. Alternatively, plasma
muscle glycogen stores and may impair hepatic gluconeogenesis. can be collected in sodium fluoride tubes; sodium fluoride inhibits
In our hospital, sepsis-induced hypoglycemia is most com- glycolysis by erythrocytes, leukocytes, and platelets. Unfortunately,
monly associated with parvovirus infection, abscesses, hemor- hemolysis is common in blood collected in sodium fluoride-treated
rhagic gastroenteritis, pyothorax, pyometra, and gram-negative tubes, which can result in slight decrements in glucose values owing
septicemia. Sepsis-induced hypoglycemia should be considered if to methodological problems in laboratory determinations.
a hypoglycemic animal is suffering from severe infection or sig- Blood glucose values determined by many portable blood glu-
nificant leukocytosis (> 30,000 cells/μL). Artifactual hypoglyce- cose monitoring (PBGM) devices designed for use by human
mia caused by delays in measuring the glucose concentration in diabetics are almost always lower than actual glucose values deter-
a blood sample containing bacteria and marked leukocytosis may mined by bench-top methodologies (e.g., glucose oxidase and
contribute to the low blood glucose measurement. A diagnosis of hexokinase methods). A notable exception is the AlphaTRAK
sepsis-induced hypoglycemia is based on identification of infec- glucometer designed for use in diabetic dogs and cats. Blood glu-
tion by means of a physical examination, CBC, radiography and cose values obtained with the AlphaTRAK can be high or low
ultrasonography, appropriate bacterial cultures, and resolution of compared with actual glucose values (Cohen et al, 2009; Zini
hypoglycemia after initiation of appropriate antibiotic therapy. If et al, 2009; Fig. 9-5). Failure to consider this “error” when using
severe infection is diagnosed in a dog or cat with hypoglycemia, a PBGM device could result in an incorrect diagnosis of hypogly-
pursuit of other causes of hypoglycemia is usually not warranted cemia. Fortunately, for most PBGM devices, the more severe the
unless screening tests dictate otherwise or the hypoglycemia fails hypoglycemia, the more accurate the device becomes.
to resolve after initiation of appropriate antibiotic therapy. Laboratory error may also result in an incorrect value for any
assay. Therefore it is wise to confirm a finding of hypoglycemia
by means of evaluation of a second blood sample using bench-
Kidney Failure
top methodology before more expensive studies are performed to
The blood glucose concentration in dogs and cats with kidney identify the cause of hypoglycemia.
failure is usually within the reference range. Occasionally dogs and
cats develop hyperglycemia as a result of uremia-induced carbohy- Iatrogenic Hypoglycemia (Toxicities)
drate intolerance and insulin resistance or more commonly develop
problems with glycemic control in a dog or cat with concurrent Insulin and oral sulfonylurea drugs (e.g., glipizide, glyburide) are
diabetes mellitus. Although critical illness caused by kidney failure the only commonly available drugs that consistently lower the
100 usually caused by idiopathic hypoglycemia, starvation, congenital

portosystemic shunt, or sepsis. In young adult dogs or cats, hypo-
glycemia is usually caused by hepatobiliary disease, portosystemic
Percentage of blood samples
80
shunt, hypoadrenocorticism, or sepsis. In older dogs or cats, hepa-
tobiliary disease, beta-cell neoplasia, extrapancreatic neoplasia,
60 hypoadrenocorticism, and sepsis are the most common causes.
The blood glucose concentration tends to be greater than 45
40 mg/dL (2.5 mmol/L) and is often an incidental finding in dogs
and cats with hypoadrenocorticism or hepatobiliary disease,
20
although severe hypoglycemia causing neurologic signs may occa-
sionally occur. Additional clinical pathologic alterations are usu-
ally present (e.g., hyponatremia and hyperkalemia in animals with
hypoadrenocorticism or increased liver enzyme activities, hypo-
0
Dog/Cat Human Human Human Human cholesterolemia, hypoalbuminemia, and a low blood urea nitro-
PBGM PBGM PBGM PBGM PBGM
gen [BUN] concentration in animals with hepatobiliary disease).
Normal serum electrolyte concentrations do not rule out a corti-
#1 #2 #3 #4
sol deficiency as the cause for hypoglycemia; atypical hypoadre-
FIGURE 9-5 Frequency of low (solid) and high (hatched) blood glucose results nocorticism may be present. An ACTH stimulation test or liver
measured in the same blood sample by five portable blood glucose monitoring function test (i.e., preprandial and postprandial bile acids) may be
(PBGM) meters designed for use in human diabetics and one meter (AlphaTRAK) required to confirm the diagnosis. Severe hypoglycemia (less than
designed for use in diabetic dogs and cats, compared with reference analyzer re- 40 mg/dL; 2.2 mmol/L) may develop in neonates and juvenile
sults. One hundred fifty-eight blood samples obtained from 49 dogs were evalu- kittens and puppies (especially toy breeds) and in animals with
ated. Blood glucose concentrations measured by the reference analyzer ranged sepsis, beta-cell neoplasia, and extrapancreatic neoplasia—most
from 41 to 639 mg/dL (2.3 to 35.8 mmol/L). (From Cohen TA, et al.: Evaluation of notably hepatic adenocarcinoma and leiomyosarcoma. Sepsis is
six portable blood glucose meters for measuring blood glucose concentration in readily identified on the basis of physical examination findings
dogs, J Am Vet Med Assoc 235:279, 2009.) and abnormal CBC findings, which include a neutrophilic leu-
kocytosis (typically > 30,000/μL), a shift toward immaturity, and
signs of toxicity. Extrapancreatic neoplasia can usually be identi-
blood glucose concentration. In small animal practice, the most fied on the basis of the physical examination, abdominal or tho-
common cause of symptomatic hypoglycemia is an overdose of racic radiography, and abdominal ultrasonography findings. Dogs
insulin in a diabetic dog and cat—a diagnosis that should always with beta-cell neoplasia typically have normal physical examina-
be suspected whenever a client reports signs resembling hypogly- tion findings aside from findings suggestive of hypoglycemia (e.g.,
cemia in a diabetic pet. weakness) and no abnormalities other than hypoglycemia identi-
Clinical hypoglycemia has been reported in dogs following the fied on routine blood and urine tests. Measurement of baseline
ingestion of alpha lipoic acid (Loftin and Herold, 2009), dried serum insulin concentration when the blood glucose is less than
chicken jerky treats (Hooper and Roberts, 2011; Thompson et al, 60 mg/dL (3.4 mmol/L) is used to confirm the diagnosis of a beta-
2013), and xylitol (Murphy and Coleman, 2012). Alpha lipoic cell tumor.
acid has antioxidant properties, is available as an over-the-coun-
ter supplement, and has been investigated as possible adjunctive
CONFIRMING THE DIAGNOSIS OF AN INSULIN-
treatment for various conditions, including diabetes mellitus and
SECRETING BETA-CELL TUMOR: SERUM
diabetic neuropathy. Ingestion of dried chicken jerky treats made
INSULIN DETERMINATION
in China caused Fanconi syndrome characterized by glucosuria
and euglycemia or hypoglycemia in addition to lethargy, inappe- The diagnosis of an insulin-secreting beta-cell tumor requires an
tence, and vomiting in dogs. Xylitol is a five-carbon sugar alcohol initial confirmation of hypoglycemia followed by documentation
used as a sweetener and sugar substitute in many products includ- of inappropriate insulin secretion and identification of a pancre-
ing chewing gums, candies, baked goods, jellies, drink powders, atic mass using ultrasonography, CT, or exploratory celiotomy.
vitamins, and nutritional supplements (Murphy and Coleman, Considering the potential differential diagnoses for hypoglycemia
2012). Xylitol ingestion in dogs stimulates insulin secretion lead- (see Box 9-4), a tentative diagnosis of insulin-secreting beta-cell
ing to potentially severe and life-threatening hypoglycemia, which tumor can often be made on the basis of the history, physical
may develop within 30 minutes to 12 hours after xylitol ingestion. examination findings, and an absence of abnormalities other than
Acute hepatic failure may develop 1 to 3 days later. Treatment for hypoglycemia shown by routine blood tests.
these toxicities included IV fluids with dextrose administration in
addition to supportive care.
Whipple’s Triad
DIAGNOSTIC APPROACH TO HYPOGLYCEMIA: In 1935, the report that established insulin-secreting tumors
PRIORITIZING THE DIFFERENTIALS of the pancreas as a clinical entity included a discussion of the
three criteria to be used in confirming the diagnosis (Whipple
Hypoglycemia should always be confirmed in a second blood and Grantz, 1935). These standards, now referred to as Whipple’s
sample before initiating diagnostic studies to identify the cause. triad, are: (1) the symptoms occur after fasting or exercise; (2)
Careful evaluation of the animal’s history, physical examination at the time of symptoms, the serum glucose concentration is less
findings, and results of routine blood and urine tests (i.e., CBC, than 50 mg/dL (2.8 mmol/L); and (3) the symptoms are relieved
serum biochemistry panel, and urinalysis) usually provide clues by administration of glucose. Unfortunately, this triad can result
to the underlying cause. Hypoglycemia in the puppy or kitten is from numerous causes of hypoglycemia and as such is nonspecific.
Determination of Baseline Insulin and Glucose BOX 9-6 Interpretation of Baseline Serum Insulin
Concentrations Concentration in Dogs with Hypoglycemia
Theory Believed to be Caused by Insulin-Secreting
The diagnosis of an insulin-secreting beta-cell tumor is established Beta-Cell Neoplasia
by evaluating the blood insulin concentration at a time when
hypoglycemia is present. Hypoglycemia suppresses insulin secre- SERUM INSULIN CONCENTRATION PROBABILITY OF BETA-CELL TUMOR*
tion in normal animals, with the degree of suppression directly Above reference range High
related to the severity of the hypoglycemia. Hypoglycemia fails Upper half of reference range Possible
to have this same suppressive effect on insulin secretion if the Lower half of reference range Low
insulin is synthesized and secreted from autonomous neoplastic Below reference range Ruled out
cells, because tumor cells that produce and secrete insulin are less
responsive to hypoglycemia than normal beta cells. Invariably *Ideally, the blood glucose concentration determined by a bench-top methodology (i.e.,
the dog with an insulin-secreting tumor will have an inappropri- glucose oxidase or hexokinase method) should be less than 50 mg/dL in the same blood
ate excess of insulin relative to that needed for a particular blood sample submitted to the laboratory for insulin determination. Interpretation of serum insulin
glucose concentration. The relative excess of insulin is easiest to concentration is unreliable if the blood glucose concentration is greater than 60 mg/dL.
recognize when the blood glucose concentration is low, preferably
less than 50 mg/dL (2.8 mmol/L). If the blood glucose concentra- 150
Serum insulin concentration (µU/mL)

*
tion is low and the insulin concentration is in the upper half of 100
**
the reference range or increased, the dog has a relative or abso- ** *
lute excess of insulin that can be explained by the presence of an 80 *
**
insulin-secreting tumor that is insensitive to hypoglycemia. Con- 60
*
**
fidence in identifying an inappropriate excess of insulin depends * ***
**
on the severity of the hypoglycemia; the lower the blood glucose 40 * **
**
concentration, the more confident the clinician can be in identi- **********
20 * * *** *
fying inappropriate hyperinsulinemia, especially when the serum * * **
***********
insulin concentration falls in the reference range. 10 *** * *
* *** *
***
5 *** * * *
Protocol ****** *** * *** *
* ****
***
Most dogs with insulin-secreting neoplasia are persistently hypo- 0
glycemic. If the blood glucose concentration is less than 60 mg/dL Beta cell Hepatic Smooth muscle Sepsis
tumor tumor tumor
(3.4 mmol/L) and preferably less than 50 mg/dL (2.8 mmol/L),
serum should be submitted to a commercial veterinary endocrine FIGURE 9-6 Serum insulin concentrations in dogs with hypoglycemia caused
laboratory for determination of the glucose and insulin concen- by an insulin-secreting beta-cell tumor, hepatoma or hepatocellular carcinoma,
trations. The insulin assay must be validated for use in dogs, and leiomyosarcoma, or sepsis. Although a serum insulin concentration in the high
interpretation of insulin results should be based on the reference end of the normal range (i.e., > 10 μU/mL) is consistent with a beta-cell tumor, a
interval established by the laboratory utilized (Madarame et al, serum insulin concentration in the low end of the normal range (i.e., 5 to 10 μU/
2009; Öberg et al, 2011). If the dog’s blood glucose concentra- mL) is not diagnostic for a beta-cell tumor. Shaded regions represent the high
tion is greater than 60 mg/dL, fasting may be necessary to induce (horizontal lines) and low (hatched area) ends of the normal range.
hypoglycemia. Blood glucose concentrations should be evaluated
hourly during the fast and blood obtained for glucose and insu- insulin concentration is in the upper half of the reference range.
lin determination when the blood glucose concentration is less Insulin values near the lower end of the reference range may be
than 60 mg/dL. It is important to remember that blood glucose found in animals with other causes of hypoglycemia, as well as in
results obtained from PBGM devices are often lower than results those with insulin-secreting tumors (Fig. 9-6). Careful assessment
obtained using bench-top methodologies (Cohn et al, 2000; Wess of the history, physical examination findings, and diagnostic test
and Reusch, 2000; Cohen et al, 2009; Zini et al, 2009). Ideally a results in relation to viable differentials for hypoglycemia in that
blood sample for submission to a commercial laboratory for glu- dog and, if necessary, repeating serum glucose and insulin measure-
cose and insulin determinations should not be obtained until the ments when hypoglycemia is more severe will usually identify the
blood glucose measured on these devices is less than 50 mg/dL. cause of the hypoglycemia. In 101 of our dogs with a histologically
Once fasting has induced hypoglycemia and the blood sample has confirmed beta-cell tumor and a blood glucose concentration less
been obtained for glucose and insulin determination, the dog can than 60 mg/dL, the serum insulin concentration was above the
be fed several small meals over the next 2 to 3 hours to minimize reference range in 74 dogs (73%); in the upper half of the reference
overstimulation of the tumor and rebound hypoglycemia. range in 21 dogs (21%); and in the lower end of the reference range
in 6 dogs (6%). No dog had a serum insulin concentration less
Interpretation than the reference range. Any serum insulin concentration below
The serum insulin concentration must be evaluated from the same the reference range is consistent with insulinopenia and does not
blood sample as and in relation to the blood glucose concentration indicate the presence of an insulin-secreting tumor.
(Box 9-6). A serum insulin concentration that exceeds the upper
limit of the reference range in a dog with a corresponding blood Insulin-to-Glucose Ratios
glucose concentration of less than 60 mg/dL (3.4 mmol/L), in
combination with appropriate clinical signs and clinical pathologic Several insulin-to-glucose ratios, including the insulin-to-glucose
findings, strongly supports the diagnosis of an insulin-secreting ratio, the glucose-to-insulin ratio, and the amended insulin-to-glu-
tumor. An insulin-secreting tumor is also possible if the serum cose ratio, have been recommended to evaluate the interrelationship
between the blood glucose and insulin concentrations and to help test. By evaluating the response of blood glucose and insulin
establish the diagnosis of insulin-secreting tumor when the labo- concentrations for a period of time after administration of these
ratory results are ambiguous (e.g., hypoglycemia is marginal and agents, a differentiation between normal and neoplastic beta cells
serum insulin concentrations remain in the reference range). Of can potentially be made. We do not use any of these tests to estab-
these ratios, the amended insulin-to-glucose ratio is most com- lish the diagnosis of an insulin-secreting tumor, and they are not
monly used. The amended insulin-to-glucose ratio is determined recommended. See the first edition of Canine and Feline Endocri-
by entering the blood glucose and serum insulin concentrations in nology and Reproduction (Feldman and Nelson, 1987) for more
the following formula: information on the use of provocative testing in dogs suspected of
having beta-cell neoplasia.
serum insulin (μU/mL) × 10
blood glucose (mg/dL) - 30
DIAGNOSTIC IMAGING
The use of “− 30” in the formula is based on the theory that Diagnostic imaging is indicated to identify a pancreatic mass, to
in normal humans, serum insulin concentrations are undetect- identify metastatic disease, and to localize the site of the mass in
able when the blood glucose concentration is less than 30 mg/dL the pancreas (e.g., pancreatic body versus pancreatic limb); this
(1.7 mmol/L). Whenever the blood glucose concentration is less information provides support for the diagnosis of a beta-cell tumor
than 30 mg/dL, the number 1 is used as the divisor. Extrapolat- and a preliminary assessment of surgical resectability, likelihood
ing from the human literature, most authors have suggested that of postoperative complications (e.g., pancreatitis), and prognosis.
an amended insulin-to-glucose ratio greater than 30 is diagnostic Abdominal ultrasound is widely available and is the initial imag-
of an insulin-secreting tumor. However, this test is not specific; ing procedure used to assess the pancreas, peripancreatic tissues,
that is, some dogs with other causes of hypoglycemia may have and liver. Advanced diagnostic imaging, specifically dual-phase
abnormal amended ratios (Leifer et al, 1986). The most common CT, is currently the most sensitive imaging procedure for identi-
reason for lack of specificity is a detectable serum insulin con- fication of the primary mass and metastases and is recommended
centration, albeit usually in the lower end of the reference range, immediately prior to surgical exploration, if available.
despite hypoglycemia. This occurs most commonly with hepatic
tumors and sepsis. We do not rely on insulin-to-glucose ratios for
Radiography
interpretation of blood insulin and glucose results. Rather, we
interpret the absolute serum insulin concentration during hypo- Abdominal radiographs are not helpful in establishing the diagno-
glycemia (see Box 9-6) in conjunction with the history, physical sis of an insulin-secreting tumor, partly because of the location of
findings, and results of routine blood and urine tests. the pancreas and the small size of most insulin-secreting tumors.
Insulin-secreting tumors are typically less than 3 cm in diameter
Serum Fructosamine Concentration at the time the diagnosis is established. Displacement of viscera
or a visible mass in the right cranial quadrant of the abdominal
Serum fructosamines are glycated proteins found in blood that are cavity is extremely rare. Thoracic radiographs are of limited help
used to monitor control of glycemia in diabetic dogs and cats (see in documenting metastatic disease, primarily because beta-cell
Chapters 6 and 7). Fructosamines result from an irreversible, non- tumors rarely metastasize to the lungs until late in the course of
enzymatic, insulin-independent binding of glucose to serum pro- the disease. As such, thoracic radiographs are typically negative
teins and are a marker of the average blood glucose concentration for metastatic disease when obtained at the time the diagnosis is
during the circulating lifespan of the protein, which varies from 1 established, and surgery is contemplated. The most common sites
to 3 weeks depending on the protein. The extent of glycosylation of early metastasis are the liver, regional lymph nodes, and peri-
of serum proteins is directly related to the blood glucose concen- pancreatic omentum, which are regions where abdominal radio-
tration; the lower the average blood glucose concentration dur- graphs are also ineffective in identifying metastatic disease.
ing the preceding 2 to 3 weeks, the lower the serum fructosamine
concentration, and vice versa. Serum fructosamine concentrations
Ultrasonography
below the reference range support the existence of significant
periods of hypoglycemia and an insulin secreting tumor, assum- Abdominal ultrasonography can be used to identify a mass in
ing the history and findings on physical examination and routine the region of the pancreas and to look for evidence of potential
blood and urine test results are also consistent with the diagnosis. metastatic disease in the liver and surrounding structures (Fig.
A serum fructosamine concentration below the reference interval 9-7). Because of the small size of most beta-cell tumors and simi-
may also occur with other disorders that cause prolonged periods lar echogenicity of the tumor and the adjacent normal pancreas,
of hypoglycemia or that interfere with the assay (see Table 6-8). abdominal ultrasonographic findings are often interpreted as nor-
Documenting a low serum fructosamine concentration in a dog mal, although a pancreatic mass or metastatic lesion can be found
with suspected insulinoma and blood glucose concentrations that at surgery. A normal abdominal ultrasonographic finding does not
remain greater than 60 mg/dL despite fasting provides support rule out the diagnosis of a beta-cell tumor.
for additional diagnostics (e.g., diagnostic imaging) or exploratory Ultrasonic detection of a mass lesion in the region of the pan-
surgery (Mellanby and Herrtage, 2002). creas helps confirm the suspicion of beta-cell tumor in a dog with
appropriate clinical signs and clinical pathologic abnormalities
Provocative Testing (Fig. 9-8). Identification of mass lesions in the hepatic paren-
chyma or peripancreatic tissue suggests metastatic disease. Occa-
Several tests have been described that use agents that stimulate sionally only the metastatic sites are identified with ultrasound,
insulin secretion by normal and neoplastic beta cells; these include and the tumor in the pancreas goes undetected. Failure to identify
the glucagon tolerance test, the oral and IV glucose tolerance test, a mass lesion in the region of the pancreas or metastatic sites is
the tolbutamide tolerance test, and the epinephrine stimulation common and does not rule out the presence of a beta-cell tumor.
In one study evaluating 13 dogs, the sensitivity of ultrasonography

for detecting insulinomas was 69% and the sensitivity for detect-
ing hepatic and lymph node metastasis was 44% (Lamb et al,
1995). In a more recent study evaluating 14 dogs, the sensitiv-
ity of ultrasonography for detecting insulinomas was 35% and
ultrasonography was negative in all five dogs that had lymph node
metastasis at surgery, and it was negative in two of four dogs with
hepatic metastasis (Robben et al, 2005). Random evaluation of 58
dogs with surgically and histologically confirmed beta-cell neo-
plasia seen at UC Davis that underwent ultrasound evaluation of
the abdomen prior to surgery identified a mass lesion in only 24
dogs (41%). Ultrasonography failed to identify diffuse infiltration
A of the pancreas by the tumor in two dogs and a mass that was not
grossly visible at the time of surgery in two dogs. In one of these
dogs, a 2.3 × 1.3 cm pancreatic mass was identified 1 year after the
initial exploratory surgery. Tumors in the left limb of the pancreas
were identified by ultrasound more often than those in the body
or right limb. Similarly, larger tumors were more likely to be iden-
tified with ultrasonography although tumors greater than 2 cm ×
2 cm were not identified.
The accuracy of ultrasound depends on the experience of the
operator, quality of the ultrasound machine and sonogram images,
size of the pancreatic mass, and presence of extraneous factors that
affect results (e.g., bowel gas, obesity, movement of the dog). Fail-
ure to identify a pancreatic mass in a dog suspected of having a
beta-cell tumor does not rule out the diagnosis or the presence of
metastatic disease but is a further indication for exploratory sur-
B gery in the hopes of finding a small, excisable tumor.
FIGURE 9-7 Ultrasonogram of the pancreas, showing an islet beta-cell tumor
(arrow; A) and an enlarged hepatic lymph node (arrows; B) resulting from me- Computed Tomography
tastasis of the beta-cell tumor to the liver in a 9-year-old Cocker Spaniel. (From
Nelson RW, Couto CG, editors: Small animal internal medicine, ed 5, St Louis, Conventional pre- and postcontrast CT is reported to have bet-
2014, Elsevier, p. 817.) ter sensitivity than ultrasonography for the detection of beta-cell
A B
C
FIGURE 9-8 Ultrasonograms of the pancreas showing an islet beta-cell tumor (arrows) in a 13-year-old Borzoi (A)
and a 14-year-old Miniature Poodle (B). C, Ultrasonogram of the peripancreatic tissue showing a metastatic beta-
cell tumor (arrows) in a 5-year-old Golden Retriever.
tumors in dogs. In a series of 14 dogs with insulinoma, the presence or absence of somatostatin receptors in tumor biopsy
sensitivity of conventional pre- and postcontrast CT for detec- samples. Five somatostatin receptor subtypes, designated sst1 to
tion of insulinoma and lymph node metastasis was 71% and sst5, are recognized in human tissue (Reubi et al, 2001). One
40%, respectively (Robben et al, 2005). Unfortunately, CT also variable influencing the success of somatostatin receptor scin-
identified 28 false-positive lesions in lymph nodes. Conven- tigraphy is the predominant somatostatin receptor subtype
tional pre-and postcontrast CT has been replaced by dual-phase expressed by the insulinoma, which dictates its affinity for pente-
computed tomographic angiography (CTA) for the identifica- treotide. Ligand binding studies on beta-cell tumors in humans
tion and localization of insulinomas and metastases in humans have identified different subtypes of somatostatin receptors, with
(Chatziioannou et al, 2001). Dual phase CTA techniques have variable binding capacities for somatostatin and somatostatin
been developed in dogs and preliminary studies in dogs with analogs, which may explain the variability of results (Lamberts
insulinoma have been promising (Caceres et al, 2006; Iseri et al, 1991; Bruns et al, 1994).
et al, 2007; Mai and Caceres, 2008). During dual-phase CTA, In the dog, the predominant somatostatin receptor containing
images are acquired during the arterial and venous phases after high-affinity binding sites for the somatostatin analog octreotide
IV injection of contrast medium. Most human insulinomas are and the radiopharmaceutical In-111 pentetreotide in insulin-
histopathologically hypervascular, and the CT images obtained secreting tumors has been sst2 (Robben et al, 1997; Garden et al,
during the arterial phase may clearly delineate enhancing tumor 2005). Somatostatin receptor scintigraphy using radio-labeled
lesions (Gritzmann et al, 2004). Insulinomas in dogs are also
assumed to be hypervascular. A study evaluating dual-phase
CTA of the pancreas in 10 healthy Beagle dogs identified an
arterial phase peak at 15 ± 2 seconds after contrast medium
injection followed by a venous phase peak where the pancre-
atic parenchyma was clearly delineated at 28 ± 9 seconds and
appearance of the equilibrium phase approximately 70 seconds
after injection (Fig. 9-9; Iseri et al, 2007). Caceres, et al., (2006)
identified a purely arterial pancreatic window of 5 to 6 seconds a
after contrast administration in nine healthy Beagles. Evaluation c
of a dog with insulinoma revealed a hyperattenuating mass at the p
arterial phase of the dual-phase CTA, and the size and location Sp
of the tumor observed on the CT images were consistent with
those seen at surgery (Iseri et al, 2007).
In a subsequent study involving three dogs with insulinoma, St
dual-phase CTA identified lesions not seen on ultrasonography,
including the primary insulinoma in two dogs (Mai and Caceres,
2008). Findings with dual-phase CTA were in agreement with the Liver
surgical findings in all three dogs. In two dogs, the insulinomas
had marked contrast enhancement during the arterial phase of
the study with less enhancement during the venous phase and was
L
isoattenuating to the rest of the pancreas during the delayed phase
of the study. In the third dog, a metastasized lymph node but not A
the pancreatic insulinoma had strong enhancement at the arterial
phase comparable to that seen in the primary insulinoma in the
other two dogs. Lack of arterial enhancement of an insulinoma L
has been reported in humans with up to 45% of pancreatic insu-
linomas being hypo- to isoattenuating to the rest of the pancreas St
during the arterial phase (Van Hoe et al, 1995). Occasionally
the tumor is hyperattenuating at the venous but not the arterial
phase of the study. Evaluation of the arterial and venous phase of
a contrast-enhanced CT is currently recommended immediately Sp
prior to surgery to identify the location of the primary insulinoma
and its metastatic sites. The decision to either pursue surgery with
medical treatment to follow, or cancel surgery and initiate medi- RK
cal treatment will be dependent, in part, on the findings of the
CT study.
L
B LK
Scintigraphy
FIGURE 9-9 Transverse (A) computed tomography (CT) image and a maximum
Somatostatin receptor scintigraphy using the radiopharmaceuti- intensity projection CT image in the dorsal plane (B) of a pancreatic beta-cell
cal drug indium (In)-111 pentetreotide has been used to image tumor obtained during the arterial phase of dual-phase CT angiography in a dog
pancreatic islet cell tumors in humans (Kvols et al, 1993; Lam- that presented with severe hypoglycemia and inappropriate hyperinsulinemia.
berts et al, 1993). In humans, localization of beta-cell tumors Note the marked distinction between the beta-cell tumor (arrows) and the re-
with somatostatin receptor scintigraphy has been inconsistent maining normal pancreas (arrow heads). (Images courtesy of Dr. Eric Johnson.)
and of limited value (Lamberts et al, 1991; Buetow et al, 1997). a, aorta; c, caudal vena cava; p, portal vein; St, stomach, Sp, spleen; RK and LK,
Positive and negative scan results have been correlated with the right and left kidney.
octreotide or In-111 pentetreotide was used to identify beta-cell Perioperative Management of Dogs Undergoing Surgery
neoplasia in seven dogs with inconclusive findings on abdominal
ultrasonography (Robben et al, 1997; Lester et al, 1999). Soma- The intent of surgery should be to remove as much abnormal tis-
tostatin receptor scintigraphy was effective in identifying the sue as possible, including resectable sites of metastases. The success
primary insulin-secreting tumor in five of seven dogs and larger of surgery depends in part on providing appropriate fluid therapy,
metastases in the regional lymph nodes and liver in three of three dextrose, and supportive care during the perioperative period to
dogs and two of three dogs, respectively. Small metastases in the avoid severe hypoglycemia and postoperative pancreatitis and to
liver were not detected in one dog. In a subsequent study by Gar- improve the likelihood of an uneventful recovery. Euthanasia is
den, et al., (2005) using In-111 pentetreotide, scintigraphic scans not recommended regardless of the findings at surgery. Many
in four of five dogs with insulinoma showed that abnormal foci dogs with metastatic disease can be managed medically for several
of In-111 pentetreotide activity attributed to the insulinoma but months to more than a year.
the anatomic location of the tumors was correctly predicted in Until surgery is performed, the dog should be protected from
only one of these dogs. The scan in the fifth dog was equivo- episodes of severe hypoglycemia. This can usually be accomplished
cal; a 1.5 cm insulinoma was identified in the distal left limb of through frequent feeding of small meals and administration of
the pancreas at surgery. In a series of 14 dogs with insulinoma, glucocorticoids (Box 9-7). A continuous IV infusion of a balanced
the sensitivity of single-photon emission CT using radiolabeled
octreotide was 43% and lymph node metastasis was identified
in none of five dogs (Robben et al, 2005). Negative scan results BOX 9-7 L
ong-Term Medical Therapy for Dogs with
could reflect the expression of somatostatin receptors with low a Beta-Cell Tumor
affinity for pentetreotide, low expression density of sst2 receptors,
or small size of the tumor, rather than absence of the tumor itself Standard Treatments
(Garden et al, 2005). 1. Dietary therapy
Somatostatin receptor scintigraphy offers intriguing options for a. Feed canned or dry food in three to six small meals daily.
identifying insulin-secreting tumors and determining potential b. Dietary fat, complex carbohydrates, and fiber help prolong postpran-
responsiveness of the tumor to octreotide therapy. Presumably, dial glucose absorption.
positive scintigraphic scans would also predict a positive response c. Avoid foods containing monosaccharides, disaccharides, propylene
to treatment with the somatostatin analog octreotide (see Soma- glycol, and corn syrup.
tostatin Therapy). 2. Limit exercise to walks; avoid strenuous exercise.
3. Glucocorticoid therapy
a. Prednisone, 0.5 mg/kg divided into two doses initially.
TREATMENT OF BETA-CELL NEOPLASIA
b. Gradually increase dose and frequency of administration, as needed.
c. Goal is to control clinical signs, not to reestablish euglycemia.
Surgical Versus Medical Therapy
d. Consider alternative treatments if signs of iatrogenic hypercorti-
Treatment options for a beta-cell tumor include surgical explo- solism become severe or glucocorticoids become ineffective.
ration, medical treatment for chronic hypoglycemia, or both. Additional Treatments
Surgery offers a chance to cure dogs with a resectable solitary 1. Diazoxide therapy
mass. In dogs with nonresectable tumors or with obvious meta- a. Continue standard treatment; reduce glucocorticoid dose if polyuria
static lesions, removal or “debulking” of as much abnormal tissue and polydipsia is unacceptable.
as possible frequently results in remission, or at least alleviation, b. May initiate diazoxide early when glucocorticoid dose is low, or later
of clinical signs and an improved response to medical treatment. when glucocorticoids become ineffective or polyuria and polydipsia
Survival time is longer in dogs that undergo surgical exploration becomes unacceptable.
and tumor debulking followed by medical therapy, compared with c. Diazoxide, 5 mg/kg every 12 hours initially.
dogs only treated medically (Tobin et al, 1999). Despite these d. Gradually increase dose as needed, not to exceed 60 mg/kg/day.
benefits, surgery remains a relatively aggressive mode of diagnosis e. Goal is to control clinical signs, not to reestablish euglycemia.
and treatment, in part because of the high prevalence of metastatic 2. Somatostatin therapy
disease, the older age of many dogs at the time beta-cell neopla- a. Continue standard treatment; reduce glucocorticoid dose if polyuria
sia is diagnosed, the potential for postoperative pancreatitis, and and polydipsia is unacceptable.
the unpredictable response to surgery as it relates to improvement b. Octreotide (Sandostatin), 10 to 40 μg/dog administered subcutane-
in hypoglycemia and clinical signs. As a general rule, we are less ously every 12 hours to every 8 hours.
aggressive about recommending surgery in aged dogs (i.e., older 3. Streptozotocin therapy
than 12 years of age), dogs with extensive metastatic disease iden- a. Effectiveness in improving hypoglycemia, controlling clinical signs,
tified by diagnostic imaging, and dogs with concurrent disease and prolonging survival is variable and potentially severe adverse
that significantly enhances the anesthetic risk. reactions are common (see Streptozotocin).
Medical management of chronic hypoglycemia should be initi- b. Continue standard treatment; reduce glucocorticoid dose if polyuria
ated when an exploratory celiotomy is not performed or when and polydipsia is unacceptable.
metastatic or inoperable neoplasia results in recurrence of clinical c. 0.9% saline diuresis for 3 hours, then streptozotocin, 500 mg/m2,
signs. Medical therapy revolves around nonspecific antihormonal in 0.9% saline and administered intravenously over 2 hours, then
therapy designed to increase the blood glucose concentration and 0.9% saline diuresis for 2 additional hours.
decrease the occurrence of clinical signs. Many dogs with met- d. Administer antiemetics immediately after streptozotocin adminis-
astatic disease can be managed medically for several months to tration to minimize vomiting.
more than a year. Medical therapy, however, has no potential for e. Repeat treatment every 3 weeks until hypoglycemia resolves or ad-
providing a “cure” or for preventing metastasis of malignant beta- verse reactions develop (e.g., pancreatitis, renal failure).
cell neoplasia.
electrolyte solution containing 2.5% to 5% dextrose before, dur- development of severe hypoglycemia (i.e., a blood glucose con-
ing, and immediately after surgery is important. Although this centration < 40 mg/dL). Fortunately, it is uncommon for a dog
does not restore euglycemia, these solutions provide a substrate in stable condition with a beta-cell tumor to require more than
for CNS function, thereby minimizing CNS signs in most dogs. a 5% dextrose solution given intravenously during surgery. This
Concentrations of dextrose in excess of 5% should be avoided to infusion usually maintains the blood glucose concentration above
prevent overstimulation of the pancreatic tumor and rebound, 40 mg/dL. If a 5% dextrose infusion is ineffective in preventing
sometimes fatal, hypoglycemia. The IV dextrose infusion can be severe hypoglycemia during surgery, a constant-rate infusion of
initiated the evening before surgery, at the time food and water glucagon should be considered (see Medical Therapy for an Acute
are withheld, and continued throughout the perioperative period. Hypoglycemic Crisis).
Initiation of fluid therapy before surgery also helps ensure ade- Adequate fluid therapy just prior to, during, and immediately
quate circulation to the pancreas, thereby minimizing the risk of after surgery is extremely important for minimizing the devel-
postoperative pancreatitis. The goal of the dextrose infusion is to opment of pancreatitis. Digital manipulation and dissection of
prevent clinical signs of hypoglycemia and to maintain the blood the pancreas cause inflammation. The severity of inflammation
glucose concentration at greater than 40 mg/dL (2.2 mmol/L), depends on the gentleness of the palpation, circulation to the
not to reestablish a normal blood glucose concentration. If the pancreas, and surgical procedures performed. Providing adequate
dextrose infusion is ineffective at preventing severe hypoglycemia fluid therapy prior to and during surgery ensures that every means
during the perioperative period, a constant-rate infusion of glu- of maintaining circulation through the microvasculature of the
cagon should be considered (see Medical Therapy for an Acute pancreas has been used and helps minimize the development of
Hypoglycemic Crisis). Glucagon is a potent stimulant of hepatic pancreatitis. We routinely administer fluids at a rate of 60 to 100
gluconeogenesis and is effective in maintaining normal blood glu- mL/kg/24 hr during surgery and for 24 to 72 hours after the pro-
cose concentrations in dogs with an insulin-secreting tumor when cedure, unless concurrent problems (e.g., heart failure, hypopro-
administered by constant-rate infusion. teinemia) are present that may affect the dog’s ability to handle
IV fluids.
Intraoperative Considerations During surgery, as much of the pancreas as possible should be
Attention to the patient’s blood glucose concentration and main- examined visually. A complete, gentle digital inspection of this
tenance of adequate fluid therapy during surgery are imperative organ should then be undertaken. The importance of gentle han-
for the dog with beta-cell neoplasia. In a recent study, the addition dling of the pancreas cannot be overemphasized; failure to handle
of medetomidine (5 μg/kg IM) to the preanesthetic medication the organ gently may result in severe, potentially life-threatening
protocol significantly decreased plasma insulin concentrations, pancreatitis. A thorough examination of the liver, surrounding
increased plasma glucose concentrations, and decreased the intra- lymph nodes, and omentum for metastatic sites should also be
operative glucose administration rate in 12 dogs undergoing done.
surgery for beta-cell tumor, compared with 13 dogs that did not Frequency of Tumor Identification. Most dogs with insulin-
receive medetomidine prior to surgery (Guedes and Rude, 2013). secreting tumors have masses that are easily visible to the sur-
These findings support the judicious use of medetomidine at low geon inspecting the pancreas (Fig. 9-10). In a minority of dogs,
doses as an adjunct to the anesthetic management of dogs with the tumor is not visible but can be palpated during gentle but
beta-cell neoplasia. thorough digital examination of the pancreas. Multiple pancre-
Monitoring the blood glucose concentration every 30 to 60 atic masses may also occur. Ninety-nine (88%) of 111 dogs with
minutes during surgery using a point-of-care or PBGM device insulin-secreting tumors at UC Davis had an obvious mass in the
allows objective assessment of the dog’s blood glucose status. The pancreas at the time of surgery.
goal is to maintain the blood glucose concentration greater than Tumor Location. There is no predisposition for tumor loca-
40 mg/dL (2.2 mmol/L), not to establish a normal blood glu- tion in the pancreas (Fig. 9-11). In our 99 dogs in which a mass
cose concentration per se. Moderate changes in the blood glu- was identified in the pancreas, the mass was located in the right
cose concentration can be monitored and adjustments made in (duodenal) limb of the pancreas in approximately 41%, in the left
the rate of IV dextrose administration, as needed, to prevent the (splenic) limb of the pancreas in 40%, and in the central region
A B
FIGURE 9-10 A and B, Photographs of pancreatic insulin-secreting islet beta-cell tumors (arrows).
(body) of the pancreas in 19%. In five of the remaining 12 dogs, and hypoglycemia typically recurs shortly after surgery. Prior to
a diffuse, microscopic islet cell carcinoma was recognized histo- surgery we routinely discuss with the client the possible locations
logically in an arbitrarily resected portion of the right limb of the of the tumor and the implications that location has on attain-
pancreas. Diffuse thickening of the pancreas was evident on digital ing a successful outcome. We strongly recommend against tumor
palpation of the pancreas at the time of surgery in three of these removal if the insulin-secreting tumor is located in the body of
five dogs; the pancreas was visually and digitally normal in two the pancreas because of the high probability of life-threatening
dogs. In three dogs, there was no visible mass and no metastatic postoperative complications. We inform the client that there is a
sites, the pancreas was normal on digital palpation, and histologic one in five chance of the dog having inoperable disease and that
examination of a portion of the right limb of the pancreas failed if such disease is found, we advise closing the abdomen and treat-
to identify an insulin-secreting tumor. One year later a pancreatic ing the dog medically rather than risk the development of severe
mass subsequently confirmed as an insulinoma was identified in pancreatitis by trying to remove the tumor.
one of the dogs. In two dogs, there was no visible mass and no Failure to Identify a Mass: Use of Methylene Blue. IV methy-
metastatic sites, the pancreas was normal on digital palpation, and lene blue infusion has been advocated for intraoperative identifica-
histologic examination of a portion of the right limb of the pan- tion of a beta-cell tumor in the dog (Fingeroth and Smeak, 1988;
creas failed to identify an insulin-secreting tumor but subjectively Fingeroth et al, 1988). Methylene blue is an azo dye that, when
there was marked increase in the number of islets and “islands” administered intravenously, is concentrated in the parathyroid glands
of beta cells scattered throughout the pancreas. Periodic evalua- and endocrine pancreas. Methylene blue intensely stains hyperfunc-
tions of the dogs identified persistent hypoglycemia and hyper- tional, adenomatous, or carcinomatous areas of these organs. Nor-
insulinemia but failed to identify a pancreatic mass. One dog was mal pancreatic endocrine tissue is stained a dusky slate blue, whereas
lost to follow-up at 18 months after surgery and a necropsy was hyperfunctioning tissue is stained more intensely, often a reddish
performed on the other dog approximately 5 years after surgery; blue. In one dog, methylene blue also successfully identified an
histologic changes suggestive of “beta cell hyperplasia” were still ectopic islet cell tumor and differentiated metastatic from nonmeta-
evident, and a pancreatic mass was not identified. Enlargement static nodules in surrounding tissue (Smeak et al, 1988).
of a mesenteric lymph node adjacent to the pancreas was evident Methylene blue is administered as an IV infusion by mixing
in two dogs, but a mass was not identified in the pancreas per appropriate volumes of methylene blue in 250 mL of normal iso-
se. Histologic examination of the excised lymph node confirmed tonic saline solution to obtain a total dose of 3 mg methylene
metastatic beta-cell tumor. Hyperinsulinism and hypoglycemia blue per kilogram of body weight (Fingeroth and Smeak, 1988).
persisted in both dogs after surgery. The entire solution is given over a period of 30 to 40 minutes.
Tumor location has important ramifications for the success of Maximal staining of the endocrine pancreas occurs approximately
surgery. In general, solitary tumors in the left or right limb of 30 minutes after initiation of the infusion. Complications with
the pancreas are readily excisable with minimal damage to the methylene blue infusion include Heinz body hemolytic anemia,
pancreas and a low prevalence of postoperative pancreatitis. In acute kidney failure, pseudo-cyanosis (i.e., blue-appearing oral
contrast, tumors in the body of the pancreas are often intimately mucous membranes), green-tinged urine, and possibly pancreati-
intertwined with the pancreatic ducts, blood vessels, and lymphat- tis. Hemolytic anemia is common, with the hematocrit declining
ics. Surgical removal often requires extensive manipulation and to less than 25% 2 to 3 days after surgery.
dissection of the pancreas. Severe and potentially life-threatening We do not routinely use methylene blue because of its postop-
pancreatitis that requires aggressive and often extended treat- erative complications, the routine use of dual-phase CTA prior to
ment is a common postoperative complication despite appropri- surgery, and the ability to grossly identify abnormal tissue in the
ate perioperative treatment aimed at preventing its development. vast majority of our dogs with beta-cell neoplasia. If our surgeon
In addition, complete excision of the tumor is almost impossible fails to recognize a mass and the diagnosis has been confirmed
by glucose and insulin measurements, the recommendation is to
remove the right or left limb of the pancreas in the hope of remov-
ing the portion that contains the tumor. In theory, 90% of the
pancreas could be removed without causing overt diabetes mel-
litus or exocrine pancreatic insufficiency.
Sites of Metastasis. Little correlation appears to exist between
18 tumor size or shape and its malignant potential. A complete
Stomach inspection of the abdominal contents is imperative to identify
unsuspected abnormalities as well as sites of metastasis. The most
common sites of tumor spread include the regional lymphatics and
lymph nodes (duodenal, mesenteric, hepatic, splenic), the liver,
and the peripancreatic omentum. Failure to identify metastatic
40 disease is common during surgery. A solitary pancreatic mass is
41 commonly removed in toto, with the belief that the dog has been
Pancreas “cured,” only to have clinical signs of hyperinsulinism recur weeks
to months later. In our experience, almost all beta-cell tumors
in the dog are malignant. Unfortunately, initial clinical signs are
often vague and not worrisome to the owner; weeks to months
may elapse between the onset of clinical signs and establishment
of the diagnosis, and as a result, the likelihood of metastasis at the
Duodenum time of exploratory surgery is high.
FIGURE 9-11 Diagram of tumor location in the pancreas in 99 dogs with an Recommendations if Metastasis Is Identified. Ideally, all
insulin-secreting islet beta-cell tumor. abnormal-appearing tissue should be removed, if possible, and
submitted for histologic evaluation. When abnormal tissue cannot neoplastic beta cells in the pancreas, liver, lymph nodes, or peri-
be entirely removed, debulking of the tumor mass may be benefi- pancreatic tissues that multiply and eventually reach a population
cial. Biopsy of suspected tumor tissue is the least a surgeon should density capable of secreting enough insulin to cause hypoglycemic
accomplish. The surgeon must always weigh the potential gains signs to recur. For these dogs, resolution of diabetes is followed
obtained with aggressive tumor removal and debulking against the by a variable period of euglycemia, which eventually progresses
potential complications that may develop as a result of the surgi- to hypoglycemia. Owner evaluation of the pet’s urine glucose is
cal procedure. This is especially important when dealing with the helpful in identifying when insulin therapy is no longer needed.
pancreas, because life-threatening pancreatitis can develop after Persistently negative urine glucose in conjunction with cessation
extensive manipulation and dissection of the gland. Because medi- of polyuria and polydipsia is an indication to discontinue insu-
cal treatment is a viable option after surgery, euthanasia at the time lin therapy. If hyperglycemia and glycosuria recur, insulin therapy
of surgery and heroic attempts to remove all abnormal tissue are can be reinstituted, but at a lower insulin dosage. The develop-
not recommended in a dog with metastatic disease, especially if ment of permanent insulin-requiring diabetes mellitus after sur-
the latter course increases the risk of postoperative complications. gical removal of a solitary insulin-secreting tumor is uncommon
Postoperative Complications. The most common postopera- and implies additional abnormalities involving the beta cells (e.g.,
tive complications are pancreatitis, hyperglycemia, and hypogly- beta-cell degeneration, islet hypoplasia; see Chapter 6). Perma-
cemia. The development of these complications is directly related nent diabetes mellitus has developed in only two of the dogs that
to the expertise of the surgeon in handling the pancreas and excis- underwent surgical removal of an insulin-secreting tumor at our
ing these tumors, the location of the tumor in the pancreas (i.e., hospital. Both dogs were lost to follow-up after 1 to 2 years, and
peripheral limb versus body), the presence or absence of func- at that time both dogs were still receiving insulin injections twice
tional metastases, and the adequacy of fluid therapy during the a day to control hyperglycemia.
perioperative period. Persistent Postoperative Hypoglycemia. Dogs that remain
Pancreatitis. IV administration of polyionic fluids with 2.5% hypoglycemic after surgical removal of an insulin-secreting tumor
to 5% dextrose (60 to 100 mL/kg/24 hr) and nothing by mouth have functional metastatic disease. Medical therapy should be initi-
prior to, during, and for 24 to 48 hours after surgery, followed by ated in dogs with persistent postoperative hypoglycemia. During
appropriate dietary therapy during the ensuing week, is helpful in the initial 48 to 72 postoperative hours, IV infusion of 2.5% to 5%
minimizing the development of pancreatitis. We rely on physical dextrose should be continued. The goal is to prevent clinical signs
examination findings in determining when to initiate oral water of hypoglycemia (especially seizures), not to reestablish a normal
and a bland diet. Circulating pancreatic enzyme concentrations blood glucose concentration. Additional therapy may be needed if
(e.g., canine pancreatic-specific lipase [cPL]) are usually not deter- hypoglycemic seizures occur (Box 9-8; also see Medical Therapy for
mined after surgery. Arbitrarily treating the dog for pancreatitis an Acute Hypoglycemic Crisis). Small meals should be fed every
without determining the serum pancreatic enzyme concentra- 4 to 6 hours, beginning as soon after surgery as possible. A diet
tions beforehand has produced excellent results. Despite gentle acceptable for the treatment of pancreatitis should be fed initially.
handling of the pancreas during surgery, aggressive fluid therapy Additional therapy may be needed, depending on the efficacy of
during the perioperative period, and appropriate dietary therapy the frequent feedings in maintaining remission of clinical hypo-
during the postoperative period, nine (13%) of 70 dogs undergo- glycemia (see Medical Therapy for Chronic Hypoglycemia). If a
ing surgery for beta-cell tumor at UC Davis still developed clinical dog becomes symptomatic despite the frequent feedings, medical
signs of acute pancreatitis. Three of the nine dogs died as a result therapy should be attempted before euthanasia is recommended.
of pancreatitis; the tumor was located in the body of the pancreas
and was difficult to excise in all three dogs.
Diabetes Mellitus. Occasionally dogs develop transient diabe-
BOX 9-8 M
edical Therapy for Hypoglycemic Seizures
tes mellitus after surgical removal of an insulin-secreting tumor.
Caused by an Insulin-Secreting Beta-Cell Tumor
Diabetes mellitus is believed to result from inadequate insulin
secretion by “atrophied” normal beta cells. Removal of all or a
Seizures at Home
majority of the neoplastic cells acutely deprives the animal of insu-
Step 1: Rub sugar solution on pet’s buccal mucosa.
lin. Until the normal beta cells regain their secretory capability, the
Step 2: Once pet is sternal, feed a small meal.
dog is hypoinsulinemic and may require exogenous insulin injec-
Step 3: Call the veterinarian.
tions to maintain euglycemia. It was once thought that postsurgical
hyperglycemia and glycosuria were excellent prognostic signs indi- Seizures in Hospital
cating total removal of insulin-secreting neoplastic cells. However, Step 1: Administer 1 to 5 mL (depending on dog size) of 50% dextrose
most of our dogs have required exogenous insulin only transiently (diluted) intravenously slowly over 1 to 2 minutes followed by continuous
after surgery and ultimately have required medical management IV infusion of 5% dextrose in water (i.e., D5W).
for an exacerbation of an insulin-secreting tumor several weeks to Step 2: Once animal is sternal, feed a small meal.
months after their need for insulin therapy dissipates. Step 3: Initiate long-term medical therapy (see Box 9-7).
Postsurgical insulin therapy is initiated only when hyperglyce- Intractable Seizures in Hospital
mia and glycosuria persist for 1 to 2 days after discontinuation Step 1: Administer 2.5% to 5% dextrose in water intravenously at 1.5 to 2
of all dextrose-containing IV fluids. Initial insulin therapy should times maintenance fluid rate.
be conservative—that is, 0.25 U of Lente or neutral protamine Step 2: Add 0.5 to 1 mg of dexamethasone/kg to IV fluids and administer
Hagedorn (NPH) insulin per kilogram of body weight given once over 6 hours; repeat every 12 to 24 hours, as necessary.
daily. Subsequent adjustments in dosage or frequency of admin- Step 3: If above fails, administer glucagon USP (Eli Lilly) intravenously by
istration should be based on clinical response and blood glucose constant rate infusion at an initial dosage of 5 to 10 ng/kg/min.
determinations (see Chapter 6). Step 4: If necessary, control seizure activity with diazepam or phenobarbital
The need for insulin treatment is usually transient, lasting until medical treatment becomes effective in controlling hypoglycemia.
from a few days to a few months. Most of these dogs still have
Evaluating the Long-Term Success of Surgery: Is the Dog Cured? increases blood glucose through stimulation of hepatic glycogenol-
The long-term success of surgery can be difficult to predict in ysis and gluconeogenesis. In a recent study, subcutaneous admin-
dogs with a “solitary” mass that is removed in toto and subse- istration of glucagon resulted in a rapid and significant increase
quent blood glucose concentration returns to normal. The most in serum glucose concentrations in healthy Beagles but the effect
efficient and logical initial method for evaluating these patients was short-lived (Zeugswetter et al, 2012; Fig. 9-12). Although
for recurrence of beta-cell neoplasia is periodic measurement (i.e., clinical trials are needed, at home glucagon emergency kits used
every month initially) of a fasting blood glucose concentration. to treat severe hypoglycemia in human diabetics may become a
The fasting blood glucose concentration should be consistently viable option for the short-term treatment of severe hypoglycemia
greater than 70 mg/dL (3.9 mmol/L) if beta-cell neoplasia has not in dogs or cats and provide the time needed to get the dog or cat to
recurred. Recurrence of beta-cell neoplasia should be suspected if an emergency veterinary hospital for care (Niessen, 2012).
the blood glucose concentration is less than 70 mg/dL. Confir- In the hospital, clinical signs of hypoglycemia can usually be
mation of recurrence requires measurement of the serum insulin alleviated initially with IV administration of 50% dextrose,
concentration when the blood glucose concentration is less than diluted, followed by continuous IV infusion of 5% dextrose (i.e.,
60 mg/dL (see Confirming the Diagnosis of an Insulin-Secreting dextrose 5% in water [D5W]). In dogs with beta cell neoplasia,
Beta-Cell Tumor: Serum Insulin Determination). dextrose should be administered in small amounts slowly (e.g., 1
to 5 mL increments depending on the size of the dog over a period
of 1 to 2 minutes) to effect. Rapid administration of large boluses
Medical Therapy for an Acute Hypoglycemic Crisis
of glucose to a dog with suspected or proven beta cell neoplasia
The acute onset of clinical signs caused by hypoglycemia typi- can result in severe rebound hypoglycemia caused by excessive
cally occurs at home after exercise or consumption of food that insulin secretion by the tumor in response to the rapid increase in
is easily digestible and rapidly absorbed; during the immediate the blood glucose concentration. The goal of therapy is to control
postoperative period in the dog with functioning metastases or neurologic signs (primarily seizures), not correct hypoglycemia.
inoperable neoplasia; or as a result of inadvertently aggressive Once neurologic signs have been controlled with judicious IV
IV dextrose administration at the time hypoglycemia is initially administration of dextrose, frequent feedings and glucocorticoids
identified. Therapy depends on the severity of clinical signs and can be initiated (see Box 9-7).
the location of the dog (i.e., home versus hospital) and initially If the dextrose infusion is ineffective in preventing severe hypo-
involves administration of glucose, either as food or sugar solution glycemia or breaking the cycle of hypoglycemia and hyperglyce-
by mouth or as an IV dextrose solution. mia, a constant-rate infusion of glucagon should be considered.
If an owner contacts a veterinarian by telephone and reports Glucagon is a potent stimulant of hepatic glycogenolysis and glu-
that the pet is having a hypoglycemic seizure, we do not recom- coneogenesis and is effective in maintaining normal blood glucose
mend transporting the dog to a veterinary hospital. Rather, the concentrations in dogs with beta-cell neoplasia when administered
owner should be instructed to rub a sugar solution on the pet’s by constant-rate infusion (Fischer et al, 2000; Fig. 9-13). One
buccal mucosa. Hypoglycemic dogs usually respond in 1 to 2 min- milligram of lyophilized glucagon USP (Eli Lilly) is reconstituted
utes. The owner should be instructed to never place fingers in, or with the diluent provided by the manufacturer, and the solution
pour the sugar solution down, the pet’s mouth. Once the dog or is added to 1 L of 0.9% saline, making a 1 μg/mL solution, which
cat is sternal and cognizant of its surroundings, it should be fed a can be administered by syringe pump. The initial dosage is 5 to 10
small meal and brought to the veterinarian. ng per kilogram of body weight per minute. The dosage is adjusted
At home subcutaneous administration of glucagon is used to as needed to maintain the blood glucose concentration between
treat severe hypoglycemia in human diabetics. Glucagon quickly 60 and 100 mg/dL (3.4 and 5.6 mmol/L). When discontinuing
12 216
Group 1 (Intravenous)
a,b a,b a Group 2 (Subcutaneous)

10 Group 3 (Placebo) 180
Glucose (mmol/L)
Glucose (mg/dL)
8 144
FIGURE 9-12 Mean glucose concentrations

6 108 (mmol/L and mg/dL) over 180 minutes after
the subcutaneous (SC) and intravenous (IV)
injection of synthetic glucagon or 0.9% saline
solution (placebo). (From Zeugswetter FK, et al.:
4 72 Metabolic and hormonal responses to subcuta-
neous glucagon in healthy beagles, J Vet Emer
Crit Care 22:558, 2012.) Bars indicate standard
deviations. a, Significant difference of SC versus
2 36 placebo injection; b, significant difference of
0 20 40 60 80 100 120 140 160 180 200 SC versus IV injection; P < 0.05 is considered
Sampling Points (minutes) significant.
Blood glucose (mg/dL) dogs benefit from surgical debulking after medical treatment has
150 become ineffective in controlling clinical signs of hypoglycemia.
Glucagon stopped
One of our dogs underwent surgical debulking on three separate
100
occasions; the dog survived 3 years before succumbing to meta-
Sx Glucagon
tapered
static disease involving the lungs.
Alloxan and streptozotocin are drugs with specific toxicity
50 directed at beta cells. The potential for serious adverse reactions
Glucagon infusion has limited the use of these drugs for the treatment of insulin-
0 secreting tumors in dogs. However, a viable treatment protocol
1 3 5 7 9 using streptozotocin in dogs has been described, and studies to
Days determine its value in the treatment of insulin-secreting tumors
FIGURE 9-13 Blood glucose concentrations in a 13-year-old female spayed Po- have been reported (Moore et al, 2002; Northrup et al, 2013).
meranian before and after surgical removal of an insulin-secreting islet beta-
cell tumor. Pancreatitis and severe hypoglycemia developed postoperatively. Frequent Feedings
The hypoglycemia resolved, and euglycemia was maintained after initiation of Dogs with insulin-secreting tumors have a persistent absolute or
a constant-rate intravenous (IV) infusion of glucagon. The dosage of glucagon relative excess of circulating insulin. Frequent feedings provide
was gradually tapered beginning on day 5, feeding of small amounts of food a constant source of calories as a substrate for the excess insulin
was begun on day 7, and the IV glucagon infusion was stopped on day 8. Severe secreted by the tumor and help to reduce the frequency of hypo-
hypoglycemia did not recur. Sx, Surgery. glycemic episodes. Diets high in fat, complex carbohydrates, and
fiber delay gastric emptying, slow intestinal glucose absorption,
and help minimize a rapid increase in the portal blood glucose
glucagon, the dose should be gradually decreased over 1 to 2 days concentration that could stimulate excessive pancreatic insu-
and the blood glucose concentration monitored for recurrence of lin secretion. Simple sugars are rapidly absorbed, have a potent
severe hypoglycemia. stimulatory effect on insulin secretion by neoplastic beta cells, and
Occasionally, a hypoglycemic dog with CNS signs fails to should be avoided. A combination of canned and dry food, fed
respond to glucose or glucagon administration. These signs in three to six small meals daily, is recommended. Daily caloric
could be the result of a disorder unrelated to hypoglycemia. intake should be controlled because hyperinsulinemia promotes
However, irreversible cerebral lesions may result from long- obesity. Exercise should be limited to walks on a leash.
term, severe hypoglycemia and the resultant cerebral hypoxia.
Cerebral hypoxia predisposes the nervous tissue to edema, caus- Glucocorticoid Therapy
ing increased CSF pressure and cell death. These animals have Glucocorticoid therapy should be initiated when dietary manipu-
a guarded to grave prognosis. Therapy is directed at providing lations are no longer effective in preventing clinical signs of hypo-
a continuous supply of glucose as a 5% solution given intrave- glycemia. Glucocorticoids antagonize the effects of insulin at the
nously or by stimulating hepatic glucose production with a con- cellular level, stimulate hepatic glycogenolysis, and indirectly pro-
stant-rate infusion of glucagon. Simultaneously, seizure activity is vide the necessary substrates for hepatic gluconeogenesis. Pred-
controlled with diazepam or stronger anticonvulsant medication nisone (dog) or prednisolone (cat) are the glucocorticoids most
(e.g., phenobarbital). Last, if cerebral edema is suspected, treat- often used. The initial dosage is 0.25 mg/kg by mouth every 12
ment with mannitol, furosemide, and/or dexamethasone should hours. Adjustments in the dose are based on clinical response. The
be considered (Fenner, 1995). dose of prednisone required to control clinical signs increases with
time in response to growth of the tumor and its metastatic sites.
Medical Therapy for Chronic Hypoglycemia Eventually, the adverse effects of prednisone, specifically polyuria
and polydipsia, become unacceptable to clients. This typically
See Box 9-7. occurs when the prednisone dosage approaches 1 mg/kg twice
daily, although there is dog to dog variability in development of
Background adverse effects and owner tolerance of the adverse effects. When
Medical management for chronic hypoglycemia should be initi- adverse effects become intolerable, the dose of prednisone should
ated when an exploratory celiotomy is not performed or when be reduced by 25% to 50% (not stopped) and additional therapy
metastatic or inoperable neoplasia results in recurrence of clinical considered.
signs. The goals of medical therapy are to reduce the frequency
and severity of clinical signs and to avoid an acute hypoglycemic Diazoxide Therapy
crisis, not to establish euglycemia per se. Medical therapy typi- Diazoxide (Proglycem) is a benzothiadiazide diuretic that inhibits
cally involves nonspecific antihormonal therapy. Antihormonal insulin secretion, stimulates hepatic gluconeogenesis and glyco-
therapy is palliative and should minimize hypoglycemia by pro- genolysis, and inhibits tissue use of glucose. The net effect is the
viding a continuous source of glucose from the gastrointestinal development of hyperglycemia. Diazoxide does not inhibit insu-
tract (frequent feedings), increasing hepatic glycogenolysis and lin synthesis and does not have cytotoxic (antineoplastic) effects.
gluconeogenesis (glucocorticoids), or inhibiting the synthesis, Diazoxide therapy can be initiated early in the medical treatment
secretion, or peripheral cellular actions of insulin (glucocorticoids, of a beta-cell tumor when the glucocorticoid dose is low and poly-
diazoxide, somatostatin). Antihormonal therapy consists primar- uria and polydipsia are acceptable to the client or can be initiated
ily of frequent feedings and glucocorticoids (see Box 9-7). Surgi- later when glucocorticoids are no longer effective in controlling
cal debulking of functional masses may enhance the effectiveness clinical signs of hypoglycemia or the severity of polyuria and poly-
of medical therapy. The best results are obtained when surgical dipsia has become unacceptable to the client. In the later situa-
debulking is performed shortly after the diagnosis of an insulin- tion, glucocorticoids should be continued but at a lower dose. The
secreting tumor has been established, although we have had a few initial dosage of diazoxide is 5 mg/kg by mouth every 12 hours.
The dosage may gradually be increased as needed to control signs receptors have been identified in humans (Patel, 1999). These sub-
of hypoglycemia but should not exceed 60 mg/kg/day. Thiazide types show a tissue-specific distribution and differences in affinity
diuretics may potentiate the effects of diazoxide. The two drugs for somatostatin and its analogs (Bruns et al, 1994). In humans,
can be administered together to enhance hyperglycemic effects if some insulin-secreting tumors have receptor subtypes that do not
diazoxide alone is not effective. The dosage of hydrochlorothiazide or only minimally bind octreotide, resulting in minimal to no
is 1 to 2 mg/kg by mouth every 12 hours. effect by the analog on serum insulin and glucose concentrations
The goal of diazoxide therapy is to establish a dosage at which (Lamberts et al, 1991; 1996). Autoradiography performed in dogs
hypoglycemia and its clinical signs are reduced or absent. In addi- with insulin-secreting neoplasia suggests the presence of only one
tion, the dosage should be low enough to avoid hyperglycemia somatostatin receptor (sst2 receptors) in canine insulin-secreting
(blood glucose concentrations > 180 mg/dL; 10 mmol/L) and its tumors (Robben et al, 1997). The somatostatin receptor identified
associated clinical signs. Reports of diazoxide use have appeared in canine insulin-secreting tumors contains high-affinity binding
in the veterinary literature only sporadically (Leifer et al, 1986; sites for octreotide and the radiopharmaceutical pentetreotide (see
Feldman and Nelson, 1987). Thirteen of 17 dogs with an insulin- Scintigraphy). In that study, baseline plasma insulin concentra-
secreting tumor in our series had a good clinical response, lasting tions, although varying widely, decreased significantly in all 10
6 weeks to 20 months. In another report, nine of 14 dogs had a dogs after octreotide administration. Unfortunately, octreotide is
good response to diazoxide therapy (Leifer et al, 1986). extremely expensive, must be administered by injection, has a rela-
The most common adverse reactions to diazoxide administra- tively short (< 6 hours) suppressive effect on serum insulin con-
tion are anorexia and vomiting. Administering diazoxide with a centrations in some dogs, clinical response to octreotide treatment
meal or decreasing the dosage, at least temporarily, is usually effec- is unpredictable, and some dogs that initially respond become
tive in controlling adverse gastrointestinal signs. Other potential refractory to octreotide treatment (Lothrop, 1989). Nevertheless,
complications include diarrhea, tachycardia, bone marrow sup- octreotide (10 to 40 μg SC twice or three times per day) is well
pression, aplastic anemia, thrombocytopenia, pancreatitis, diabe- tolerated and can be used for the management of both acute and
tes mellitus, cataracts, and sodium and fluid retention (Feldman chronic hypoglycemia in some dogs with insulin-secreting neo-
and Nelson, 1987). Diazoxide is metabolized in the liver, and plasia. Adverse reactions have not been reported at these dosages.
the metabolites are excreted via the kidneys and biliary system. Streptozotocin. Streptozotocin is a naturally occurring nitro-
Adverse reactions or complications may develop more rapidly or sourea that is similar in structure to glucose and is taken up by
at a lower dosage of diazoxide in a dog with concurrent hepatic the GLUT-2 transmembrane carrier protein but not by other glu-
dysfunction. cose transporters (Schnedl et al, 1994). Because pancreatic beta
cells have high concentrations of GLUT-2 transporters, strepto-
Somatostatin Therapy zotocin selectively destroys pancreatic beta cells by depressing the
Octreotide (Sandostatin) is an analog of somatostatin that inhibits pyridine nucleotides nicotinamide adenine dinucleotide (NAD)
the synthesis and secretion of insulin by normal and neoplastic and reduced nicotinamide adenine dinucleotide (NADH). Two
beta cells. IV administration of octreotide can rapidly decrease the dogs with confirmed hyperinsulinism were treated with strepto-
serum insulin concentration, causing a corresponding increase in zotocin by Meyer in the 1970s. The first dog developed neph-
the serum glucose concentration in dogs with insulin-secreting rotoxicosis and was euthanized 3 weeks after a single treatment
neoplasia (Robben et al, 1997; Fig. 9-14). The inhibitory actions with streptozotocin at a dosage of 1000 mg/kg body weight given
of octreotide on insulin secretion can be maintained for several intravenously over a 1-minute period. The second dog developed
hours with subcutaneous administration (Fig. 9-15). The respon- temporary remission of hypoglycemia that lasted approximately
siveness of insulin-secreting tumors to the suppressive effects of 50 days after two treatments with streptozotocin at a dosage of
octreotide varies and depends on the presence of membrane recep- 500 mg/m2 given intravenously over a 30-second period. The
tors on the tumor cells that bind somatostatin (Lamberts et al, treatments were given 1 week apart, and mannitol was infused for
1990; Simpson et al, 1995). To date, five subtypes of somatostatin 20 minutes before and after each streptozotocin treatment. The
200 50
160 40
Serum insulin (µU/mL)
Serum insulin (µU/mL)
120 30
80 20
40 10
0 0
0 10 20 30 0 2 4 6 8
Time (minutes) Time (hours)
FIGURE 9-14 Mean (± SD) serum insulin concentration prior to and after intra- FIGURE 9-15 Serum insulin concentration prior to and after subcutaneous (SC)
venous (IV) administration of 100 μg of octreotide in six dogs with an insulin- administration of 20 μg of octreotide to a dog with an insulin-secreting islet cell
secreting islet cell tumor. Arrow, Octreotide administration; hatched area, normal tumor. Arrow, Octreotide administration; hatched area, normal range for fasting
range for fasting serum insulin concentration. serum insulin concentration.
dog developed a nephropathy and hepatopathy after a third treat- of hypoglycemia, detection of local recurrence or metastasis, or
ment administered at day 97 and was euthanized shortly there- death because of any cause, was 196 days (range, 20 to 840 days).
after. As a result of these clinical reports, streptozotocin was not Response rate to streptozotocin could not be determined from
considered a viable treatment for insulin-secreting tumors in dogs. the results of this study because there was no control group of
In 2002, Moore and colleagues described a fluid diuresis pro- dogs with comparable metastatic or nonresectable insulinoma that
tocol that allowed streptozotocin to be administered to dogs with were not treated, and survival times in the streptozotocin-treated
insulin-secreting tumors with a minimum of adverse reactions. dogs were confounded by concurrent symptomatic therapy, use of
Fluid diuresis has been reported to ameliorate the renal toxic- other cytotoxic therapies, and owner decision to euthanize. Bell,
ity of streptozotocin in humans, presumably as a result of less et al., (2005) reported on a Springer Spaniel treated with gluco-
contact time between the drug and the renal tubular epithelium corticoids and one treatment of streptozotocin for metastatic insu-
(Tobin et al, 1987; Kintzel, 2001). In Moore’s study, diuresis linoma that subsequently developed diabetes mellitus. The dog
with 0.9% sodium chloride at a rate of 18.3 mL/kg/hr adminis- was euthanized 118 days after streptozotocin treatment because of
tered through a peripherally located over-the-needle catheter was cervical pain caused by metastasis of the tumor.
performed for 7 hours. Streptozotocin (Zanosar) was adminis- In our experience, the effectiveness of streptozotocin in improv-
tered over a 2-hour period beginning 3 hours after initiation of ing hypoglycemia, controlling clinical signs, and prolonging
the saline diuresis. The dose of streptozotocin (500 mg/m2) was survival time has been unpredictable and adverse events to strep-
diluted in an appropriate volume of 0.9% saline to maintain tozotocin (severe vomiting, acute pancreatitis, potentially severe
the same rate of fluid administration for 2 hours. Saline diuresis kidney injury) are common and can be life-threatening. A thor-
was continued at the same fluid rate for an additional 2 hours ough discussion of potential complications with streptozotocin
after completion of the streptozotocin administration. Butorph- treatment should always be undertaken with the owner prior to
anol (0.4 mg/kg IV) was given immediately after streptozotocin initiating treatment.
administration as an antiemetic. Streptozotocin treatments were Phenytoin. Phenytoin is an anticonvulsant that inhibits the
repeated at 3-week intervals until there was evidence of tumor release of insulin by beta cells and may also directly impair the
progression (i.e., increase in tumor size by greater than 50%), effects of insulin on peripheral tissues (Haemers and Rottiers,
recurrence of hypoglycemia, or streptozotocin-induced toxicity 1981). Unfortunately, phenytoin is not usually successful in
that required supportive treatment. controlling clinical signs of hypoglycemia. Only 30% of human
Fifty-eight treatments were administered to 17 dogs with an patients with hyperinsulinism showed any beneficial effects after
insulin-secreting tumor at variable times after surgery (Moore phenytoin administration (Haemers and Rottiers, 1981). Con-
et al, 2002). Sixteen of 17 dogs had metastatic disease. One dog current diazoxide administration is not recommended, because it
developed azotemia, several dogs developed increases in serum ala- results in a decrease in blood concentrations of phenytoin. Phe-
nine aminotransferase activity that appeared to resolve with ces- nytoin has not been critically evaluated in dogs with beta-cell
sation of treatment, and vomiting occurred in 18 (31%) of 58 neoplasia.
streptozotocin treatments and was occasionally severe. Two dogs Propranolol. Propranolol is a nonselective beta-adrenergic
developed diabetes mellitus after receiving five treatments; two blocking drug that has no intrinsic sympathomimetic activity. Its
of three dogs had rapid resolution of paraneoplastic peripheral potential usefulness in patients with beta-cell neoplasia probably
neuropathy; and two dogs had a measurable reduction in tumor involves its ability to block insulin secretion by beta cells. Insu-
size. Although the median survival time was longer in dogs treated lin secretion is stimulated by the beta-adrenergic nervous system.
with streptozotocin than in 15 control dogs with a similar stage of However, propranolol may also induce hypoglycemia by impair-
disease (163 versus 90 days, respectively), this difference was not ing hepatic gluconeogenesis and glycogenolysis, normally induced
statistically significant. The range for survival time was also similar by endogenous catecholamines. Propranolol has not been criti-
between the two groups of dogs (streptozotocin-treated dogs, 16 cally evaluated in dogs with beta cell neoplasia.
to 309 days; control dogs, 0 to 426 days).
Because myelosuppression was not observed in the Moore study,
Northrup, et al., (2013) investigated increasing dose intensity by 100
decreasing the interval between streptozotocin dosing from 3 to
2 week intervals. Nineteen dogs with residual, local, metastatic,
80
or recurrent insulinoma were treated with the streptozotocin and
Percentage alive
saline diuresis protocol described by Moore, et al., (2002) but

administered biweekly rather than once every 3 weeks. Treatment 60
was initiated after surgery for insulinoma or at the time of recur-
rence. The planned treatment protocol was five treatments per dog; 40
however, 13 dogs received fewer than five treatments (median, 3;
range, 1 to 4) primarily because of development of adverse events.
20
Adverse events occurred in all dogs and included nausea, anorexia,
acute emesis or regurgitation, diarrhea, increased liver enzyme
activity, renal tubular injury, and hypoglycemic collapse or sei- 0
zure during treatment. Mild to moderate gastrointestinal toxicity 0 200 400 600 800 1000 1200 1400
was the most common adverse event. Myelosuppression was not Survival time (days)
identified. Eight dogs developed diabetes mellitus and six of these FIGURE 9-16 Kaplan-Meier curve depicting overall survival of 19 dogs with in-
dogs subsequently died or were euthanized. Median survival time sulinoma treated with streptozotocin at 2-week intervals. The median survival
for the 19 dogs was 308 days (range, 20 to 1404 days; Fig. 9-16). time was 308 days (range, 20 to 1404 days). (From Northrup NC, et al.: Prospec-
Median progression-free survival time, defined as the number tive evaluation of biweekly streptozotocin in 19 dogs with insulinoma, J Vet Intern
of days from the first streptozotocin treatment until recurrence Med 27:483, 2013.)
PROGNOSIS INSULIN-SECRETING BETA-CELL TUMORS

IN CATS
Owing to the extremely high likelihood of malignancy in any dog
with an insulin-secreting tumor (greater than 95%), the long- Insulin-secreting beta-cell tumors are rare in cats with only a few
term prognosis is guarded to poor at best because beta-cell tumors single cat case reports in the literature (McMillan, 1985; O’Brien
almost always metastasize. The disease-free interval and survival et al, 1990; Hawks et al, 1992; Kraje, 2003; Greene and Bright,
time are difficult to predict. Survival time depends partly on the 2008) and one additional report in which the beta-cell tumor was
owner’s willingness to treat the disease. Predictors of disease free an incidental finding at necropsy in a cat with ductal pancreatic ade-
interval and survival time in dogs with an insulin-secreting beta- nocarcinoma (Carpenter et al, 1987). We have seen only two cats
cell tumor include tumor size, TNM (tumor, node, and metasta- with beta-cell neoplasia during the past three decades, in contrast to
sis) stage, stromal fibrosis within the tumor, and the Ki67 index more than 150 dogs with the disease during the same time interval.
(Caywood et al, 1988; Buishand et al, 2010). Ki67 is a prolifera- The clinical characteristics of insulin-secreting beta-cell tumors in
tion marker expressed during the active phases of the cell cycle cats appear to be similar to those in dogs. To date, the disorder has
and absent in resting cells (Scholzen and Gerdes, 2000). The Ki67 affected aged cats, 12 to 17 years old. Interestingly, three cats have
index has been correlated with clinical outcome in humans with been Siamese. Immunohistochemical analysis of the beta-cell tumor
insulinoma (La Rosa et al, 2009). in a few of these cats has revealed multihormonal productivity, with
Tobin and colleagues (1999) reported median survival time insulin and chromogranin A being the most common peptides dem-
after diagnosis of 74 days (range, 8 to 508 days) and 381 days onstrated in the neoplastic cells (Kraje, 2003; Jackson et al, 2009).
(range, 20 to 1758) in dogs treated medically versus dogs that ini- Clinical signs result from the effects of hyperinsulinism and included
tially underwent surgery followed by medical treatment, respec- seizures, weakness, ataxia, lethargy, disorientation, and muscle twitch-
tively. Polton and colleagues (2007) reported median survival time ing. Hypoglycemia (blood glucose concentration < 60 mg/dL; 3.4
after diagnosis of 196 days (95% confidence interval [CI] 0 to 549 mmol/L) was documented in each cat, and an inappropriately
days) and 785 days (95% CI: 190 to 1380 days) in dogs treated increased blood insulin concentration was documented in four cats in
medically versus dogs that initially underwent surgery followed which insulin was measured. Abdominal ultrasound failed to identify
by medical treatment, respectively. The shorter survival time for a 1.5 to 2.0 cm pancreatic mass in one cat and a 0.4 cm pancreatic
dogs treated medically in the Tobin study was partly due to a more mass in another cat in which ultrasound was performed prior to sur-
severe stage of the disease at the time of diagnosis and the own- gery. A pancreatic mass was identified in five cats that underwent sur-
ers’ feelings of hopelessness, which translated into early acceptance gery, and hypoglycemia and clinical signs resolved in four of five cats
of euthanasia when clinical signs recurred. Polton, et al., (2007) after surgical excision of the mass. Four cats died or were euthanized
reported a median disease free interval of 496 days (95% CI: 302 4 weeks, 5 weeks, 18 months, and 2 years after surgery, and three of
to 690 days) in 19 dogs that underwent partial pancreatectomy. these four cats were hypoglycemic at the time of death. Necropsy in
The median disease free interval in 31 dogs that underwent partial one of these cats revealed metastasis to the liver and pancreatic lymph
pancreatectomy at Utrecht University was 244 days (range, 0 to nodes. One cat was alive with no clinical signs suggestive of recur-
1116 days) and survival time was 258 days (range, 1 to 1146) rence of the insulinoma 32 months after surgical removal of a solitary
(Tryfonidou et al, 1998). 0.4 cm pancreatic carcinoma (Greene and Bright, 2008).
The extent to which surgery can alter the prognosis depends Until more experience is gained with beta-cell neoplasia in cats,
on the clinical stage of the disease, most notably the extent of it seems prudent to approach this disorder in a manner similar to
metastatic lesions. In one multi-university study, dogs with that used for dogs. Beta-cell neoplasia should be included in the list
tumors confined to the pancreas (stage I) were normoglycemic for of differential diagnoses for persistent hypoglycemia in the older cat
a median of 14 months after surgery, whereas dogs with metasta- (see Box 9-4). The index of suspicion for beta-cell neoplasia should
sis to regional lymph nodes (stage II) or distant metastasis (stage be heightened after a thorough review of the history, physical exam-
III) were normoglycemic for a median of approximately 1 month ination, and results of clinicopathologic tests and diagnostic imag-
(Caywood et al, 1988). Dogs with stage I or stage II disease had a ing. However, as with the dog, failure to identify a pancreatic mass
median survival time of approximately 18 months, whereas those with abdominal ultrasound does not rule out a beta cell tumor. The
with stage III disease had a median survival time of less than 6 diagnosis should be confirmed by documentation of an inappro-
months. Approximately 50% of dogs with metastases to the liver priate serum insulin concentration despite the presence of hypo-
were dead by 6 months, and all were dead by 18 months from the glycemia (see Confirming the Diagnosis of an Insulin-Secreting
time of diagnosis. Beta-Cell Tumor: Serum Insulin Determination). Many commer-
In our experience, approximately 10% to 15% of dogs under- cially available radioimmunoassays for insulin work well in the dog
going surgery for an insulin-secreting tumor die or are euthanized but do not work in the cat (Lutz and Rand, 1993). It is imperative
at the time of or within 1 month of surgery because of severe that the radioimmunoassay used to measure feline insulin is vali-
metastatic disease, uncontrollable postoperative hypoglycemia, or dated for cats and that species specific reference ranges are available.
complications related to pancreatitis. An additional 20% to 25% Surgical exploration should be the initial treatment of choice
of dogs die or are euthanized within 6 months of surgery because for beta-cell neoplasia in the cat. However, the age of the cat,
of severe metastatic disease and recurrence of clinical hypoglyce- identification of metastatic disease using ultrasonography, or the
mia. The remaining 60% to 70% live beyond 6 months postop- existence of concurrent disease that increases the anesthetic risk
eratively, many beyond 1 year after surgery, before uncontrollable may warrant a more conservative medical approach in some cats.
hypoglycemia develops, resulting in death or necessitating eutha- Frequent feedings and prednisolone therapy have been effective
nasia. Additional surgery to debulk metastatic lesions may improve in controlling clinical signs of hyperinsulinism and should be the
the animal’s responsiveness to medical therapy and prolong the mainstay of medical therapy for chronic hypoglycemia. The use
survival time in some dogs that become nonresponsive to medi- of diazoxide has not been reported in the cat and is not recom-
cal treatment after the initial surgery. Some dogs with metastatic mended until its safety and dosing schedule have been investi-
disease do remarkably well (i.e., survive longer than 2 years) after gated in cats. Similarly, the efficacy of octreotide for the treatment
aggressive surgical debulking of the tumor and its metastases. of feline beta-cell neoplasia remains to be reported.
REFERENCES
Bagley RS, et al.: Hypoglycemia associ- Chastain CB: Endocrine and metabolic Fenner WR: Diseases of the brain. In Ettinger
ated with intraabdominal leiomyoma and systems. In Hoskins JD, editor: Veterinary SJ, Feldman EC, editors: Textbook of veteri
leiomyosarcoma in six dogs, J Am Vet Med pediatrics, Philadelphia, 1990, WB nary internal medicine, ed 4, Philadelphia,
Assoc 208:69, 1996. Saunders, p 249. 1995, WB Saunders, p 578.
Barreca A, et al.: In vitro paracrine regulation Chatziioannou A, et al.: Imaging and localization Fingeroth JM, Smeak DD: Intravenous methylene
of human keratinocyte growth by fibroblast- of pancreatic insulinomas, Clin Imaging blue infusion for intraoperative identification
derived insulin-like growth factors, J Cell 25:275, 2001. of pancreatic islet cell tumors in dogs. II.
Physiol 151:262, 1992. Chew DJ, et al.: Hyperglycemia and hypoglyce- Clinical trials and results in four dogs, J Am
Beaudry D, et al.: Hypoglycemia in four dogs mia. In Klenner WR, editor: Quick reference Anim Hosp Assoc 24:175, 1988.
with smooth muscle tumors, J Vet Intern to veterinary medicine, Philadelphia, Fingeroth JM, et al.: Intravenous methylene
Med 9:415, 1995. 1982, JB Lippincott, p 432. blue infusion for intraoperative identifica-
Bell R, et al.: Treatment of insulinoma in a Cohen M, et al.: Gastrointestinal leiomyosarcoma tion of parathyroid gland and pancreatic
springer spaniel with streptozotocin, J Sm in 14 dogs, J Vet Intern Med 17:107, islet cell tumors in dogs. I. Experimental
Anim Pract 46:247, 2005. 2003. determination of dose-related staining
Bellah JR, Ginn PE: Gastric leiomyosarcoma Cohen TA, et al.: Evaluation of six portable efficacy and toxicity, J Am Anim Hosp
associated with hypoglycemia in a dog, blood glucose meters for measuring blood Assoc 24:165, 1988.
J Am Anim Hosp Assoc 32:283, 1996. glucose concentration in dogs, J Am Vet Finotello R, et al.: Pancreatic islet cell tumor
Bergman PJ, et al.: Canine clinical peripheral Med Assoc 235:276, 2009. secreting insulin-like growth factor type-II
neuropathy associated with pancreatic islet Cohn LA, et al.: Assessment of five portable in a dog, J Vet Intern Med 23:1289,
cell carcinoma, Prog Vet Neurol 5:57, 1994. blood glucose meters, a point-of-care ana- 2009.
Boari A, et al.: Hypoglycemia in a dog with a lyzer, and color test strips for measuring Fischer JR, et al.: Glucagon constant-rate infu-
leiomyoma of the gastric wall producing an blood glucose concentration in dogs, J Am sion: a novel strategy for the management of
insulin-like growth factor II–like peptide, Vet Med Assoc 216:198, 2000. hyperinsulinemic-hypoglycemic crisis in the
Eur J Endocrinol 132:744, 1995. Commens PJ, et al.: Interleukin-1 is a potent dog, J Am Anim Hosp Assoc 36:27, 2000.
Boyle PJ, Cryer PE: Growth hormone, cortisol, modulator of insulin secretion from Fischer KF, et al.: Hypoglycemia in hospitalized
or both are involved in defense against but isolated rat islets of Langerhans, Diabetes patients: causes and outcomes, N Engl J
are not critical to recovery from prolonged 36:963, 1987. Med 315:1245, 1986.
hypoglycemia in humans, Am J Physiol Cryer PE: Catecholamines, pheochromocy- Fyfe JC, et al.: Glycogen storage disease type IV:
260:E395, 1991. toma and diabetes, Diabetes Rev 1:309, inherited deficiency of branching enzyme
Braund KG, et al.: Insulinoma and subclinical 1993. activity in cats, Pediatr Res 32:719, 1992.
peripheral neuropathy in two dogs, J Vet Cryer PE: Hypoglycemia. In Melmed S, Polonsky Garden OA, et al.: Somatostatin receptor imaging
Intern Med 1:86, 1987. K, Larsen PR, Kronenberg HM, editors: in vivo by planar scintigraphy facilitates
Breitschwerdt EB, et al.: Hypoglycemia in four Williams textbook of endocrinology, ed 12, the diagnosis of canine insulinomas, J Vet
dogs with sepsis, J Am Vet Med Assoc Philadelphia, 2011, Elsevier, p 1552. Intern Med 19:168, 2005.
178:1072, 1981. Cryer PE, Gerich JE: Glucose counterregula- Gerich J, et al.: Hypoglycemia unawareness,
Brix A, et al.: Glycogen storage disease type tion, hypoglycemia, and intensive insulin Endocr Rev 12:356, 1991.
Ia in two littermate Maltese puppies, Vet therapy in diabetes mellitus, N Engl J Med Gerich JE, et al.: Renal gluconeogenesis: its
Pathol 32:460, 1995. 313:232, 1985. importance in human glucose homeostasis,
Bruns C, et al.: Molecular pharmacology of Cryer PE, Polonsky KS, et al.: Glucose homeo Diabetes Care 24:382, 2001.
somatostatin receptor subtypes, Ann N Y stasis and hypoglycemia. In Wilson JD, edi- Giger U, et al.: Metabolic myopathy in canine
Acad Sci 733:138, 1994. tor: Williams textbook of endocrinology, ed 9, muscle-type phosphofructokinase deficiency,
Buetow PC, et al.: Islet cell tumors of the pan- Philadelphia, 1998, WB Saunders, p 939. Muscle Nerve 11:1260, 1988.
creas: clinical, radiological, and pathologic Das H, Hochberg FH: Metastatic neoplasms Greene SN, Bright RM: Insulinoma in a cat, J
correlation in diagnosis and localization, and paraneoplastic syndromes. In Goetz Sm Anim Pract 49:38, 2008.
Radiographics 17:453, 1997. CG, Pappert EJ, editors: Textbook of clini Gregory BL, et al.: Glycogen storage disease
Buishand FO, et al.: Evaluation of clinico- cal neurology, Philadelphia, 1999, WB type IIIa in curly-coated retrievers, J Vet
pathological criteria and the Ki67 index as Saunders, p 957. Intern Med 21:40, 2007.
prognostic indicators of canine insulinoma, D’Ercole AJ, et al.: Tissue concentrations Gritzmann N, et al.: CT in the differentiation of
Vet J 185:62, 2010. of somatomedin C: further evidence for pancreatic neoplasms-progress report, Dig
Buishand FO, et al.: Expression of insulin-like multiple sites of synthesis and paracrine or Dis 22:6, 2004.
growth factor-1 by canine insulinomas and autocrine mechanisms of action, Proc Natl Guedes AG, Rude EP: Effects of pre-operative
their metastases, Vet J 191:334, 2012. Acad Sci USA 81:935, 1984. administration of medetomidine on plasma
Buse JB, et al.: Type 2 diabetes mellitus. In de Bruijne JJ, et al.: Fat mobilization and insulin and glucose concentrations in
Melmed S, Polonsky K, Larsen PR, plasma hormone levels in fasted dogs, healthy dogs and dogs with insulinoma, Vet
Kronenberg HM, editors: Williams textbook Metabolism 30:190, 1981. Anaesth Analg 40:472, 2013.
of endocrinology, ed 12, Philadelphia, de Groot JW, et al.: Non-islet cell tumor- Haemers S, Rottiers R: Medical treatment of
2011, Elsevier, p 1371. induced hypoglycaemia: a review of the insulinoma, Acta Clin Belg 36:199, 1981.
Caceres AV, et al.: Helical computed tomographic literature including two new cases, Endocr Hambrook LY, Kudnig ST: Tumor thrombus
angiography of the normal canine pancreas, Relat Cancer 14:979, 2007. formation in two dogs with insulinoma, J
Vet Radiol Ultrasound 47:270, 2006. del Rey A, Besedovsky H: Interleukin 1 Am Vet Med Assoc 241:1065, 2012.
Carpenter JL, et al.: Tumors and tumorlike affects glucose homeostasis, Am J Physiol Hardy RM: Diseases of the liver and their treat-
lesions. In Holsworth J, editor: Diseases of 253:R794, 1987. ment. In Ettinger SJ, editor: Textbook of
the cat: medicine and surgery, Philadelphia, Edwards DF, et al.: Hypoglycemia and chronic veterinary internal medicine, ed 3, Phila-
1987, WB Saunders, p 406. renal failure in a cat, J Am Vet Med Assoc delphia, 1989, WB Saunders, p 1479.
Caywood DD, et al.: Pancreatic insulin-secreting 190:435, 1987. Hargrove DM, et al.: Adrenergic blockade does
neoplasms: clinical, diagnostic, and Feldman EC, Nelson RW: Canine and feline en not abolish elevated glucose turnover
prognostic features in 73 dogs, J Am Anim docrinology and reproduction, Philadelphia, during bacterial infection, Am J Physiol
Hosp Assoc 24:577, 1988. 1987, WB Saunders. 254:E16, 1988a.
Hargrove DM, et al.: Adrenergic blockade Lamberts S, et al.: Parallel in vivo and in vitro Minkus G, et al.: Canine neuroendocrine tu-
prevents endotoxin-induced increases detection of functional somatostatin mors of the pancreas: a study using image
in glucose metabolism, Am J Physiol receptors in human endocrine pancreatic analysis techniques for the discrimina-
255:E629, 1988b. tumors: consequences with regard to tion of metastatic versus nonmetastatic
Hawkins KL, et al.: Immunocytochemistry of diagnosis, localization and therapy, J Clin tumors, Vet Pathol 34:138, 1997.
normal pancreatic islets and spontane- Endocrinol Metab 71:566, 1990. Mokan M, et al.: Hypoglycemia unaware-
ous islet cell tumors in dogs, Vet Pathol Lamberts SJW, et al.: The role of somatostatin ness in IDDM, Diabetes Care 17:1397,
24:170, 1987. and its analogs in the diagnosis and treat- 1994.
Hawks D, et al.: Insulin-secreting pancreatic ment of tumors, Endocrinol Rev 12:450, Moore AS, et al.: Streptozotocin for treatment
(islet cell) carcinoma in a cat, J Vet Intern 1991. of pancreatic islet cell tumors in dogs: 17
Med 6:193, 1992. Lamberts SJW, et al.: Octreotide and related cases (1989-1999), J Am Vet Med Assoc
Heckmann JG, et al.: Hypoglycemic senso- somatostatin analogs in the diagnosis and 221:811, 2002.
rimotor polyneuropathy associated with treatment of pituitary disease and soma- Murphy LA, Coleman AE: Xylitol toxicosis in
insulinoma, Muscle Nerve 23:1891, tostatin receptor scintigraphy, Neuro dogs, Vet Clin N Am Small Anim 42:307,
2000. endocrinology 14:27, 1993. 2012.
Hooper AN, Roberts BK: Fanconi syndrome in Lamberts SWJ, et al.: Somatostatin analogs: Naylor JM, Kronfeld DS: In vivo studies of hy-
four non-basenji dogs exposed to chicken future directions, Metabolism 45:104, poglycemia and lactic acidosis in endotoxic
jerky treats, J Amer Anim Hosp Assoc 1996. shock, Am J Physiol 248:E309, 1985.
47:178, 2011. Leifer CE, et al.: Insulin-secreting tumor: Niessen SJM: Glucagon: are we missing a
Iseri T, et al.: Dynamic computed tomography diagnosis and medical and surgical (life-saving) trick? J Vet Emerg Crit Care
of the pancreas in normal dogs and in management in 55 dogs, J Am Vet Med 22:523, 2012.
a dog with pancreatic insulinoma, Vet Assoc 188:60, 1986. Northrup NC, et al.: Prospective evaluation of
Radiol Ultrasound 48:328, 2007. Lester NV, et al.: Scintigraphic diagnosis of biweekly streptozotocin in 19 dogs with in-
Jackson TC, et al.: Cellular and molecular insulinoma in a dog, Vet Radiol Ultra sulinoma, J Vet Intern Med 27:483, 2013.
characterization of a feline insulinoma, sound 40:174, 1999. Öberg J, et al.: Validation of a species-
J Vet Intern Med 23:383, 2009. Loftin EG, Herold LV: Therapy and outcome optimized enzyme-linked immunosorbent
Jeffery ND, et al.: Letter to the editor, Prog of suspected alpha lipoic acid toxicity in assay for determination of serum concen-
Vet Neurol 5:135, 1994. two dogs, J Vet Emerg Crit Care 19:501, trations of insulin in dogs, Vet Clin Path
Karam JH: Hypoglycemic disorders. In 2009. 40:66, 2011.
Greenspan FS, Gardner DG, editors: Lothrop CD: Medical treatment of neuroen- O’Brien TD, et al.: Canine pancreatic endocrine
Basic and clinical endocrinology, ed 6, docrine tumors of the gastroenteropan- tumors: Immunohistochemical analysis of
New York, 2001, Lange Medical Books/ creatic system with somatostatin. In Kirk hormone content and amyloid, Vet Pathol
McGraw-Hill, p 699. RW, editor: Current veterinary therapy 24:308, 1987.
Kintzel PE: Anticancer drug–induced kidney X, Philadelphia, 1989, WB Saunders, p O’Brien TD, et al.: Pancreatic endocrine
disorders, Drug Safety 24:19, 2001. 1020. tumor in a cat: clinical, pathological, and
Kishnani PS, et al.: Isolation and nucleotide Lutz TA, Rand JS: Comparison of five com- immunohistochemical evaluation, J Am
sequence of canine glucose-6-phospha- mercial radioimmunoassay kits for the Anim Hosp Assoc 26:453, 1990.
tase mRNA: identification of mutation in measurement of feline insulin, Res Vet Sci Patel YC: Somatostatin and its receptor family,
puppies with glycogen storage disease 55:64, 1993. Front Neuroendocrinol 20:157, 1999.
type Ia, Biochem Mol Med 61:168, Madarame H, et al.: Retrospective study of Pickens EH, et al.: Unique radiographic appear-
1997. canine insulinomas: eight cases (2005- ance of bone marrow metastasis of an
Kishnani PS, et al.: Canine model and genomic 2008), J Vet Med Sci 71:905, 2009. insulin-secreting beta-cell carcinoma in a
structural organization of glycogen storage Mai W, Caceres AV: Dual-phase computed dog, J Vet Intern Med 19:350, 2005.
disease type Ia (GSD Ia), Vet Pathol tomographic angiography in three dogs Polton GA, et al.: Improved survival in a retro-
38:83, 2001. with pancreatic insulinoma, Vet Radiol spective cohort of 28 dogs with insuli-
Kraje AC: Hypoglycemia and irreversible Ultrasound 49:141, 2008. noma, J Sm Anim Pract 48:151, 2007.
neurologic complications in a cat with in- McMillan F: Functional pancreatic islet cell Reubi J, et al.: Somatostatin receptor sst1-sst5
sulinoma, J Am Vet Med Assoc 223:812, tumor in a cat, J Am Anim Hosp Assoc expression in normal and neoplastic human
2003. 21:741, 1985. tissues using receptor autoradiography
Kruth SA, et al.: Insulin-secreting islet cell Mehlhaff CJ, et al.: Insulin-producing islet with subtype-selective ligands, Eur J Nucl
tumors: establishing a diagnosis and the cell neoplasms: surgical considerations Med 28:836, 2001.
clinical course of 25 dogs, J Am Vet Med and general management in 35 dogs, Robben JH, et al.: In vitro and in vivo detec-
Assoc 181:54, 1982. J Am Anim Hosp Assoc 21:607, 1985. tion of functional somatostatin receptors
Kudo M, Noguchi T: Immunoreactive myelin Mellanby RJ, Herrtage ME: Insulinoma in in canine insulinomas, J Nucl Med
basic protein in tumor cells associated a normoglycaemic dog with low serum 38:1036, 1997.
with carcinomatous neuropathy, Am J Clin fructosamine, J Sm Anim Pract 43:506, Robben JH, et al.: Comparison of ultrasonog-
Pathol 84:741, 1985. 2002. raphy, computed tomography, and single-
Kvols LK, et al.: Evaluation of a radio-labeled Meszaros K, et al.: Increased uptake and photon emission computed tomography for
somatostatin analog (I-123 octreotide) in phosphorylation of 2-deoxyglucose by the detection and localization of canine
the detection and localization of carcinoid skeletal muscles in endotoxin-treated insulinoma, J Vet Intern Med 19:15,
and islet cell tumors, Radiology 197:129, rats, Am J Physiol 253:E33, 1987. 2005.
1993. Meszaros K, et al.: In vivo glucose utilization Rossi G, et al.: Paraneoplastic hypoglycemia
La Rosa S, et al.: Improved histologic and by individual tissues during nonlethal in a diabetic dog with an insulin growth
clinicopathologic criteria for prognostic hypermetabolic sepsis, FASEB J 2:3083, factor-2-producing mammary carcinoma,
evaluation of pancreatic endocrine tumors, 1988. Vet Clin Path 39:480, 2010.
Human Path 40:30, 2009. Meyer C, et al.: Effects of autonomic neuropa- Rothman DL, et al.: Quantitation of hepatic
Lamb CR, et al.: Ultrasonography of pancreatic thy on counterregulation and awareness of glycogenolysis and gluconeogenesis in
neoplasia in the dog: a retrospective review hypoglycemia in type 1 diabetic patients, fasting humans with 13C NMR, Science
of 16 cases, Vet Rec 137:65, 1995. Diabetes Care 21:1960, 1998. 254:573, 1991.
Schnedl WJ, et al.: STZ transport and cytotoxi Smith BF, et al.: Molecular basis of canine Van Hoe L, et al.: Helical CT for the preopera-
city; specific enhancement of GLUT2- muscle type phosphofructokinase defi- tive localization of islet cell tumors of the
expressing cells, Diabetes 43:1326, 1994. ciency, J Biol Chem 271:20070, 1996. pancreas: value of arterial and paren-
Scholzen T, Gerdes J: The Ki-67 protein: from Stumvoll M, et al.: Uptake and release of glucose chymal phase images, Am J Roentgenol
the known and the unknown, J Cell Physiol by the human kidney: postabsorptive rates 165:1437, 1995.
182:311, 2000. and responses to epinephrine, J Clin Invest Walvoort HC, et al.: Comparative pathology
Schrauwen E, et al.: Peripheral polyneuro 96:2528, 1995. of the canine model of glycogen storage
pathy associated with insulinoma in the Thompson MF, et al.: Acquired proximal renal disease type II (Pompe’s disease), J Inherit
dog: clinical, pathological, and electro tubulopathy in dogs exposed to a common Metabl Dis 8:38, 1985.
diagnostic features, Prog Vet Neurol dried chicken treat: retrospective study Wess G, Reusch C: Evaluation of five portable
7:16, 1996. of 108 cases (2007-2009), Aust Vet J blood glucose meters for use in dogs, J Am
Service FJ: Hypoglycemic disorders, N Engl J 91:368, 2013. Vet Med Assoc 216:203, 2000.
Med 332:1144, 1995. Thorens B, Mueckler M: Glucose transporters Whipple AO, Grantz VK: Adenoma of islet cells
Shahar R, et al.: Peripheral neuropathy in a in the 21st century, Am J Physiol Endocrinol with hyperinsulinism: a review, Ann Surg
dog with functional islet B-cell tumor and Metab 298:E141, 2010. 101:1299, 1935.
widespread metastasis, J Am Vet Med Assoc Tobin MV, et al.: Forced diuresis to reduce Zeugswetter FK, et al.: Metabolic and hormonal
187:175, 1985. nephrotoxicity of streptozocin in the treat- responses to subcutaneous glucagon in
Sherwin RS, Felig P: Hypoglycemia. In Felig P, ment of advanced metastatic insulinoma, healthy beagles, J Vet Emerg Crit Care
et al.: Endocrinology and metabolism, New Br Med J 294:1128, 1987. 22:558, 2012.
York, 1981, McGraw-Hill, p 869. Tobin RL, et al.: Outcome of surgical versus Zini E, et al.: Paraneoplastic hypoglycemia
Simpson KW, et al.: Evaluation of the long- medical treatment of dogs with beta-cell due to an insulin-like growth factor type-II
acting somatostatin analogue octreotide neoplasia: 39 cases (1990-1997), J Am secreting hepatocellular carcinoma, J Vet
in the management of insulinoma in three Vet Med Assoc 215:226, 1999. Intern Med 21:193, 2007.
dogs, J Small Anim Pract 36:161, 1995. Tryfonidou MA, et al.: A retrospective evaluation Zini E, et al.: Evaluation of a new portable
Slye M, Wells HG: Tumor of islet tissue with of 51 dogs with insulinoma, Vet Quart glucose meter designed for use in cats,
hyperinsulinism in a dog, Arch Pathol 20:114, 1998. Schweiz Arch Tierheilkd 151:448, 2009.
19:537, 1935. Van Ham L, et al.: Treatment of a dog with an
Smeak DD, et al.: Intravenous methylene insulinoma-related peripheral polyneuro
blue as a specific stain for primary and pathy with corticosteroids, Vet Rec 141:98,
metastatic insulinoma in a dog, J Am Anim 1997.
Hosp Assoc 24:478, 1988.

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Copyright:

Available Formats

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Copyright:

Available Formats

CHAPTER 9 Beta-Cell Neoplasia: Insulinoma

CHAPTER CONTENTS Polycythemia, 358

Alpha-Adrenergic Effects Insulin: Decreased Secretion

TABLE 9-4 PHYSICAL EXAMINATION BOX 9-3 A

100 usually caused by idiopathic hypoglycemia, starvation, congenital

Serum insulin concentration (µU/mL)

In one study evaluating 13 dogs, the sensitivity of ultrasonography

a,b a,b a Group 2 (Subcutaneous)

FIGURE 9-12 Mean glucose concentrations

Serum insulin (µU/mL)

saline diuresis protocol described by Moore, et al., (2002) but

PROGNOSIS INSULIN-SECRETING BETA-CELL TUMORS

You might also like

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Copyright:

Available Formats

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 9 - Beta-Cell Neoplasia Insulinoma

Uploaded by

Copyright:

Available Formats

CHAPTER 9 Beta-Cell Neoplasia: Insulinoma

CHAPTER CONTENTS Polycythemia, 358

Alpha-Adrenergic Effects Insulin: Decreased Secretion

TABLE 9-4 PHYSICAL EXAMINATION BOX 9-3 A

100 usually caused by idiopathic hypoglycemia, starvation, congenital

Serum insulin concentration (µU/mL)

In one study evaluating 13 dogs, the sensitivity of ultrasonography

a,b a,b a Group 2 (Subcutaneous)

FIGURE 9-12 Mean glucose concentrations

Serum insulin (µU/mL)

saline diuresis protocol described by Moore, et al., (2002) but

PROGNOSIS INSULIN-SECRETING BETA-CELL TUMORS

You might also like

TABLE 9-4 PHYSICAL EXAMINATION BOX 9-3 A