Canine and human insulinoma

The pancreas

The pancreas is a glandular organ in the cranial abdomen that has both endocrine and exocrine functions. Ninety-eight percent of the pancreas is composed of exocrine acinar cells that produce and secrete digestive enzymes into networks of ductal cells. The remaining 2% of the pancreas is formed by the endocrine islets of Langerhans that produce hormones, which regulate glucose metabolism (Figure 1).

Fig 1 vetscite
Figure 1. Histology of the pancreas: (A) a single acinus; (B) a pancreatic islet embedded in exocrine tissue. ε-cells are not depicted, since they comprise <1% of the pancreatic islet. Figure adapted from Bardeesy et al. (2002)[1].

Acinar cells secrete 1) amylase, 2) proteases: trypsinogen, chymotrypsinogen, proelastase, procarboxy peptidase A&B, and 3) lipases: pancreatic lipase, phospholipase A2, and cholesterol esterase. These enzymes are responsible for the digestion of carbohydrate, protein, and fat. In order to prevent autodigestion of the pancreas, the proteolytic enzymes, and phospholipase A2 are secreted in inactive forms. These proenzymes require to be activated in order to exert their enzymatic function. After delivery of the proenzymes to the duodenum, they become activated through cleavage of activation peptides from the proenzymes by trypsin and enterokinase. Besides enzymes, pancreatic juice contains bicarbonate for neutralisation of gastric acid, factors that facilitate absorption of cobalamin, zinc and colipase C, and antibacterial factors.

Unlike the exocrine pancreas, where acinar cells secrete all types of enzymes, islet cells specialise in the secretion of one hormone type. Insulin, which is secreted by β-cells, is the best-studied pancreatic hormone. Insulin secretion is tightly regulated by the blood glucose concentration. In contrast to most other cells, the entrance of glucose in β-cells via facilitative glucose transporters (GLUTs) is insulin independent. With increasing blood glucose concentrations, insulin secretion gradually increases, eventually reaching a plateau level. Vice versa, when blood glucose concentrations decrease, insulin secretion is inhibited. The function of insulin is to decrease blood glucose levels through inhibition of gluconeogenesis, glycogenolysis, fatty acid breakdown, and ketogenesis, and stimulation of glycogenesis and protein synthesis. Actions of insulin are opposed by glucagon, which is secreted by -cells. Glucagon increases hepatic glycogenolysis and gluconeogenesis. Finally, ghrelin, which is secreted by α-cells influences glucose metabolism by inhibiting β-cell response to glucose, leading to a decrease in insulin release.

Surgical anatomy
The pancreas of dogs and cats is divided in a right and a left lobe that unite at the pancreatic body (Figure 2). The right or duodenal lobe is located within the mesoduodenum. In contrast to the canine pancreas, the distal third of the feline right lobe curves cranially, giving it a hook-like appearance, which ends close to the caudal vena cava. The distal or caudal part of the right lobe lies relatively unattached to the duodenum. More cranially, towards the pancreatic duct and the corpus, the pancreas is tightly associated to the duodenum and the common bile duct, which runs adjacent to the duodenum. The corpus of the pancreas is closely related to the pylorus and proximal duodenum cranially. The portal vein crosses the corpus or proximal left lobe dorsally, close to where the cranial pancreaticoduodenal artery and gastroduodenal vein enter the pancreatic body, crossed on their right side by the common bile duct [2] [3]. The left lobe, which is positioned high in the dorsal leaf of the omentum, begins just caudal to the pylorus and extends along the greater curvature of the stomach to the dorsal extremity of the spleen. The dorsal surface is associated with, from right to left, the portal vein, caudal vena cava, left gastric vein, and splenic vein. The arterial blood supply to the pancreas is tripartite. The largest artery is the cranial pancreaticoduodenal artery, a terminal branch of the gastroduodenal artery. The cranial pancreaticoduodenal artery enters the body of the pancreas, courses through the right pancreatic lobe, and also supplies the duodenum after exiting the pancreas. The pancreatic artery, which branches from the splenic artery in 80% of dogs, supplies the left lobe of the pancreas. In 20% of dogs, the pancreatic artery originates from the cranial mesenteric artery. The third and smallest source of arterial inflow is the caudal pancreaticoduodenal artery that arises from the cranial mesenteric artery, and supplies and courses through the distal portion of the right pancreatic lobe. The cranial and caudal pancreaticoduodenal arteries anastomose within the right lobe of the pancreas. The pancreaticoduodenal vein drains the right lobe of the pancreas into the gastroduodenal vein. The body and left lobe are drained via the splenic vein [4].

Pancreatic duct anatomy differs between dogs and cats. Dogs typically have two pancreatic ducts. The large accessory pancreatic duct carries secretions from the right pancreatic lobe to the minor duodenal papilla. The smaller pancreatic duct transports secretions from the left lobe and enters the major duodenal papilla next to the common bile duct, approximately 5 cm from the pylorus. Some dogs only have an accessory pancreatic duct, and another reported variation is the presence of three duodenal openings. In 80% of cats, a single pancreatic duct is present that joins the common bile duct before entering the major duodenal papilla. In the remaining 20% of cats an accessory pancreatic duct is present, which, as in dogs, opens into the minor duodenal papilla.

Multiple lymph nodes drain the pancreas. Knowledge of the location of these lymph nodes is important in cases of malignant disease. Lymph nodes known to drain the pancreas are the pancreaticoduodenal, splenic, mesenteric, and hepatic. Other lymph nodes that most likely also drain the pancreas are the colic and gastric lymph nodes.

fig 2 Vetscite
Figure 2. Anatomy of the canine pancreas and surrounding structures. Figure adapted from Evans et al. (1999) [10].

The pancreas is directly innervated by vagal nerve fibers. The coeliac and superior mesenteric plexus innervate the blood vessels of the pancreas.

Canine insulinoma

INS is the most common pancreatic endocrine tumour in the dog. Immunohistochemically, neoplastic β-cells in dogs and humans have been shown to produce a variety of hormones in addition to insulin. These hormones include glucagon, somatostatin, gastrin, serotonin, growth hormone and pancreatic polypeptide [5-10]. Although multiple hormones have been demonstrated immunohistochemically in INS, the occurrence of mixed clinical syndromes is rare and therefore INS are named after insulin, its principally-secreted hormone, and its most common clinical sign: hyperinsulinaemia-induced hypoglycaemia.

In 90% of the cases, primary canine INS are solitary tumours and their diameter is usually smaller than 2.5 cm [11,12]. Most INS are located in the left or right pancreatic lobe, rather than in the pancreatic body [13-15]. Multiple primary tumours are present in 10-14% of cases and although very rare, diffuse tumour growth in the pancreas can be observed.

In 1969, Capen & Martin stated that 60% of the INS are carcinomas and 40% are adenomas. According to these authors, hyperinsulinism was more commonly diagnosed in dogs with β-cell carcinomas than in dogs with β-cell adenomas [16]. Yet in a more recent study, INS of 35 dogs were all classified as adenocarcinomas [13]. Although there is still some controversy in the literature concerning the benign or malignant nature of β-cell tumours in the dog, in general canine INS are considered to be malignant in more than 95% of the cases, because, even though they may lack histological criteria of malignancy, they almost always tend to metastasise [25,26]. Forty to 50% of the dogs show already macroscopically visible INS metastases at surgery, primarily in regional lymph nodes and/or the liver [13,14,17]. Other metastatic sites include duodenum, mesentery, omentum, spleen, kidney, heart and spinal cord. Pulmonary metastases seem rare. Clinical staging of INS is performed in accordance with the World Health Organization’s tumour-node-metastasis (TNM) system (Table 1)[19]. Dogs with INS are divided into one of three clinical stages based on extent of neoplasia: T1N0M0 (clinical stage I), T1N1M0 (clinical stage II) and T0N0M1, T1N0M1 or T1N1M1 (clinical stage III).

INS hypersecrete insulin and cause an increased insulin concentration in the blood. The elevated insulin levels inhibit glycogenolysis and gluconeogenesis and thereby suppress glucose secretion by the hepatocytes. Moreover, the high insulin levels stimulate glucose-uptake by muscle and adipose tissue. In normal β-cells, insulin secretion is entrance of glucose into β-cells via facilitative glucose

Table 1. TNM classification for canine INS.
T Primary tumor
T0 – No evidence of tumor
T1 – Tumor present

N Regional lymph nodes
N0 – No regional lymph nodes involved
N1 – Regional lymph nodes involved

M Distant metastasis
M0 – No evidence of distant metastasis
M1 – Distant metastasis present

transporters (GLUTs) is insulin independent. With increasing blood glucose concentrations, insulin secretion gradually increases, eventually reaching a plateau level. Vice versa, when blood glucose concentrations decrease, insulin secretion is inhibited. Neoplastic -cells are less sensitive to the negative feedback of low blood glucose concentrations. Therefore INS secrete inappropriately high amounts of insulin despite declining blood glucose concentrations, resulting in a profound hypoglycaemia [12].

The clinical signs of hyperinsulinaemia-induced hypoglycaemia are due to activation of the autonomic nervous system and the lack of an energy substrate available to the central nervous system. The latter is called neuroglycopaenia and results in neurological signs since nervous tissue can use only glucose for its energy supply, and diffusion of glucose across the blood-brain-barrier as well as cerebral oxidation are severely impaired in a situation of low blood glucose concentration. The activation of the autonomic nervous system involves both neuronally-released transmitters and catecholamines released by the adrenal medulla.

Signalment and clinical signs
INS are most frequently diagnosed in middle-sized to large breed dogs. Although a real breed predilection has not been established, German Shepherds, Irish Setters, Boxers, Golden Retrievers, Poodles, Fox Terriers, Collies and Labrador Retrievers appear most frequently affected, but INS are also reported to occur in smaller breeds such as West Highland White Terriers. There is no sex predisposition for the disease [13] [17] [18]. Based on 192 dogs from 7 reports, the mean age of dogs with INS at the time of diagnosis is 9.3 years (range, 3-15 years) [9] [10] [13] [14] [15] [20] [21]. Common clinical signs of dogs with INS due to neuroglycopaenia include seizures, collapse, generalised weakness, posterior paresis, lethargy, ataxia and exercise intolerance. Hypoglycaemia-induced stimulation of the autonomic nervous system may result in muscle tremors, nervousness and hunger.

Clinical signs of canine INS often occur intermittently. In the initial stages hypoglycaemic episodes are preceded by fasting, exercise, excitement or stress, because those situations lead to increased glucose utilisation. Paradoxically, an episode might also follow directly after a meal, because the neoplastic β-cells excessively respond to the postprandial rise of the blood glucose concentration, indicating some sensitivity to the glucose feedback loop. At first, hypoglycaemic attacks occur at widely-spaced intervals, but as the disease progresses they will succeed each other at shorter intervals and will become more severe. The severity of clinical signs depends on the glucose nadir: convulsions and loss of consciousness often occur when blood glucose concentration is <2.8 mmol/L. The rate of decrease in the blood glucose concentration and the duration of the hypoglycaemia also determine the severity of the clinical signs. For example, a blood glucose concentration that gradually declines to 2 mmol/L over an extended period is less likely to result in clinical signs of hypoglycaemia compared to a blood glucose concentration of 2 mmol/L that develops rapidly over a few hours. Affected dogs usually do not have clinical signs between hypoglycaemic attacks. Prolonged and severe episodes of hypoglycaemia may eventually induce laminar cortical necrosis in the cerebrum, leading to coma and death [16]. The mean duration of clinical signs prior to diagnosis is 3.6 months (range, 1 day – 3.5 years) [13] [14] [15].

Differential diagnoses
In the elderly dog, common differential diagnoses of hypoglycaemia include INS, hypoadrenocorticism, liver insufficiency, porto-systemic shunting, sepsis and non-pancreatic neoplasia. Non-pancreatic tumours associated with hypoglycaemia include hepatocellular carcinoma, hepatoma, leiomyosarcoma, metastatic mammary carcinoma, primary pulmonary carcinoma, adrenocortical carcinoma and leukemia. In the past, several mechanisms have been suggested to explain the hypoglycaemia due to non-pancreatic tumours such as deranged tumour metabolism with excessive utilisation of glucose. Now there is convincing evidence that incompletely processed insulin-like growth factors cause the hypoglycaemia [22] [23].

Delayed separation of blood cells and plasma during sample processing is also a common differential diagnosis of hypoglycaemia. The glucose concentration in whole blood may decrease by 0.56 mmol/L per hour because erythrocytes and leukocytes continue to utilise glucose. It is advisable to collect blood for glucose measurement in a sodium fluoride-coated tube, because sodium fluoride inhibits glucose metabolism by blood cells.

Less common differential diagnoses are neonatal and juvenile hypoglycaemia, pregnancy toxaemia and hunting dog hypoglycaemia. Sporadically, hypoglycaemia is the result of glycogen storage disease, growth hormone deficiency, glucagon deficiency, severe polycythemia, renal failure, or cardiac failure. Finally, hypoglycaemia might also be iatrogenic, induced by drugs like insulin and sulfonylurea [9] [24].

A consistent abnormality in biochemistry profiles in dogs with INS is hypoglycaemia (reference interval (RI): 4.2-5.8 mmol/L, University Veterinary Diagnostic Laboratory, Utrecht, The Netherlands). Sometimes, a blood glucose concentration within the RI is found, because the blood glucose concentration can significantly fluctuate during the course of a day. In these cases, fasting samples and repeated testing may be necessary to confirm hypoglycaemia. Blood glucose concentrations below 2.8 mmol/L are often accompanied by clinical signs, but values just below the lower limit of the reference range may not. Hence the presumptive diagnosis of canine INS is not defined by hypoglycaemia alone, but commonly based on signalment and history, combined with the fulfillment of Whipple’s triad [25]:
– Presence of clinical signs
– Hypoglycaemia
– Relief of clinical signs after glucose administration or feeding

To demonstrate hypoglycaemia it may be necessary to fast the dog. Fasting for 24 hours is in most cases sufficient to reveal hypoglycaemia, but if not, fasting is prolonged for up to 72 hours [17]. Fasting should be supervised by hourly evaluation of blood glucose concentration, since in dogs with INS, blood glucose levels decrease before symptoms occur, possibly triggering sudden and severe symptoms [9] [24].

While Whipple’s triad fits any cause of hypoglycaemia, the next step in the diagnostic work-up is to exclude differential diagnoses. The plasma insulin concentration should be determined. Two commonly used insulin assays are a chemiluminescent immunoassay and a competitive enzyme-linked immunosorbent assay [10]. In case of INS, circulating insulin concentrations are typically within the reference range (RI: 10-170 pmol/L) or higher despite hypoglycaemia [9]. The simultaneous occurrence of blood glucose 70 pmol/L (>10 mU/L) is diagnostic. Plasma insulin concentrations greater than the high end of the reference interval have been reported in 56-83% of dogs with INS [9]. Several explanations have been suggested for the occurrence of INS with plasma insulin concentrations within the RI: 1) INS may episodically secrete insulin in short bursts, which causes a wide fluctuation in plasma insulin levels; 2) INS may secrete abnormal insulin, which is rapidly broken down; 3) INS may secrete excessive amounts of proinsulin, instead of insulin; 4) circulating insulin-like growth factors may contribute to hypoglycaemia [10]. Bottom line is that the plasma insulin concentration should be suppressed in case of blood glucose <3.5 mmol/L and that a plasma insulin concentration within the RI in such a case indicates inappropriate secretion of insulin.

Diagnostic imaging techniques, like transabdominal ultrasonography (US), computed tomography (CT), single-photon emission computed tomography (SPECT) and somatostatin receptor scintigraphy (SRS) can be of great help for the identification and preoperative staging of INS. Radiographs of the abdomen do not contribute to a diagnosis of INS because of the small size of the tumours and border-effacement from the surrounding soft tissues of the cranial abdomen. Although readily available to the general practitioner, in one study US was found to have a low sensitivity in detecting canine INS [26]. Only 5 out of 14 primary INS were correctly identified by US, whereas no lymph node metastases were detected by US. Similar results were obtained using SPECT. CT proved to be the most sensitive method, correctly identifying 10 out of 14 primary tumors and 2 out of 5 lymph node metastases. However, conventional pre- and postcontrast CT is not a very specific method because it also identified many false-positive lesions. More recently dual-phase CT angiography (CTA) techniques have been developed and the use of dynamic CTA for the presurgical localisation of INS in 4 dogs has been reported. With CTA, after an intravenous injection of contrast medium, CT images are acquired during the arterial and pancreatic phases. While canine INS are histologically hypervascular, their CT images are hyperattenuating at the arterial phase, compared to the normal pancreas. In all dogs CTA findings were in agreement with the surgical and histopathological findings, suggesting CTA is an accurate method for canine INS detection [27] [28]. To date the gold standard, however, remains exploratory laparotomy. Careful inspection and palpation of the pancreas and adjacent structures, with or without the use of intravenous methylene blue infusion (3 mg/kg), reveals most INS and metastases. The use of intravenous methylene blue might cause fatal haemolytic anaemia or acute renal failure for which reason it is not routinely recommended. When intra-operative inspection and palpation are unsuccessful in detecting the tumour, intra-operative US can be used to visualise both the primary INS as well as possible liver metastases. Definitive diagnosis is obtained by histological examination of tumour samples.

INS therapy can be divided into medical management and surgical treatment. Surgery, if needed combined with post-operative medical management, is the treatment of choice for long-term management, because this treatment strategy results in longest survival times [15] [18] [29]. Besides post-operative management, other indications for medical treatment are pre-operative stabilisation, or cases in which surgery is not performed.

Medical therapy
An acute hypoglycaemic crisis typically occurs at home after exercise, excitement, or eating; immediately post-operative in dogs with inoperable neoplasia; or as a result of inadvertently aggressive intravenous dextrose administration resulting in a massive release of insulin from INS and rebound hypoglycaemia. Dogs with acute and severe seizuring should be treated immediately with oral glucose syrup and/or intravenous glucose, since the shorter the period of hypoglycaemia, the lower the risk of irreversible brain damage [12] [30]. After stabilisation of emergency hypoglycaemia, dogs with INS should be fed four to six small meals a day of a high-protein, high-fat and high-complex-carbohydrate diet. This type of diet decreases postprandial hyperglycaemia, thereby preventing a marked insulin surge. Restricting exercise to brief walks on a leash might also help to reduce clinical hypoglycaemia [16]. In some dogs, these measurements are sufficient to control insulin-induced hypoglycaemia. If clinical signs persist, despite frequent feedings and restricted exercise, additional medication should be initiated.

Diazoxide is the preferred drug for treatment of INS-induced hypoglycaemia. Diazoxide raises blood glucose concentrations mainly through direct inhibition of pancreatic insulin release, but also through stimulation of hepatic gluconeogenesis and glycogenolysis and inhibition of glucose uptake by tissues [15] [17] [18]. It is recommended to start with an oral dose of 5 mg/kg twice daily. If hypoglycaemic symptoms do not disappear, the initial dose can be gradually increased to 30 mg/kg twice daily [11]. Possible side effects of diazoxide treatment are anorexia, vomiting and ptyalismus. These side effects may be prevented by dividing the daily dose and by administering diazoxide together with food. The use of diazoxide is contraindicated in patients with liver, kidney, or heart failure [11] [31].
An alternative to diazoxide therapy is glucocorticoid therapy. Glucocorticoids, such as prednisolon, antagonise the effects of insulin at the cellular level and increase gluconeogenesis [30]. The recommended initial dose of prednisolon is 0.25 mg/kg twice daily, which can be increased up to 2.0 – 3.0 mg/kg twice daily [21]. However, since dosages greater than 1.1 mg/kg orally twice daily are considered immunosuppressive and the use of high dosages of glucocorticoids often give rise to clinical signs of iatrogenic Cushing’s syndrome [31] at Utrecht University, we prefer using diazoxide over glucocorticoids.

In addition to the commonly used drugs described above, treatment with somatostatin (analogs) and cytotoxic treatment with streptozocin have been described [20,32-34]. Octreotide is a somatostatin analog that binds to the somatostatin receptors on β-cells, thereby inhibiting the synthesis and secretion of insulin. There have been conflicting reports on the effectiveness of octreotide therapy in dogs with INS. Several reports show that both somatostatin and octreotide significantly decrease plasma insulin concentrations and reduce the clinical signs of hypoglycaemia [33] [35] [36]. However, in a study conducted by Simpson et al. (1995)[32], octreotide treatment did not lead to any improvement in clinical signs or blood glucose and insulin concentrations. These contradictory results may be explained by the variable expression of somatostatin receptors on neoplastic β-cells. Another explanation for treatment failures in dogs with INS could be the relatively short (3–4 hours) suppressive effect of octreotide on plasma insulin concentration in dogs. Furthermore, some dogs become refractory to octreotide treatment [35]. The most recent study on the use of octreotide demonstrates that a single subcutaneous dose of 50 microgram octreotide results in a consistent suppression of plasma insulin concentrations and a corresponding increase of blood glucose concentrations in 12 dogs with INS [34]. Octreotide administration at this dosage was not associated with side effects, therefore further studies on the effectiveness of slow-release octreotide preparations in long-term treatment of dogs with INS are warranted.

Streptozocin is a nitrosurea compound that selectively destructs pancreatic β-cells. Because this drug is extremely nephrotoxic, induction of diuresis is mandatory to ameliorate the renal toxic effects of streptozocin. A successful treatment protocol includes a 0.9% NaCl diuresis (18 mL/kg/h, IV) for 7 hours. Three hours after initiating the diuresis, streptozocin (500 mg/m2) is administered during a 2-hour period. An antiemetic should be administered immediately before or after streptozocin therapy. Streptozocin treatment is repeated every 3 weeks until there is evidence of tumour progression, recurrence of hypoglycaemia, or development of streptozocin-induced toxicosis. Common side effects of streptozocin therapy include vomiting, diabetes mellitus and renal failure [20].

Surgical therapy
Dry food should be withheld 12 hours before surgery. Canned food can be fed until 6 hours before surgery. If dogs are clinically hypoglycaemic, liquid, easy-digestible food preparations should be given until 1-2 hours before surgery. If clinical signs occur in this immediate pre-operative period, the animal should be administered a glucose solution intravenously (1-5 mL of 50% dextrose administered over 10 minutes), because it is mandatory that the blood glucose concentration is stabilised before surgery. If not, at least during anesthaesia, dogs should receive a continuous IV infusion of a balanced electrolyte solution containing 2.5%-5.0% dextrose. Since some anaesthetics increase blood glucose concentration, administering IV glucose solutions without monitoring blood glucose, can lead to hyperglycaemia, stimulating the INS to secrete even more insulin. Additionally, manipulation of the tumour during surgery can trigger increased insulin secretion, leading to a more profound hypoglycaemia. To prevent this vicious circle from happening, it is important to regularly monitor blood glucose concentration during surgery and to adapt the rate of the IV glucose solution according to the measured blood glucose levels. At our clinic, the glucose infusion is normally stopped as soon as the primary tumour is removed, which often suddenly lowers plasma insulin concentrations and stimulates a rise in blood glucose concentration. Most of the times normoglycaemia is restored or hyperglycaemia is induced within minutes after INS resection. Continuous glucose monitoring (every 10 minutes post resection) is mandatory.

Depending on the pancreatic localisation, INS can be removed by local enucleation or partial pancreatectomy. Partial pancreatectomy is the preferred method, because it results in longer survival times than local enucleation [13]. Therefore, local enucleation should only be considered if the INS is located in the body of the pancreas, or in the most proximal portions (i.e., close to the corpus) of the right and left lobes, in which extreme caution should be taken to prevent damage to the ductal system and the pancreaticoduodenal arteries that are located near or in the pancreatic tissue [11]. With any technique, it is important to evaluate the complete pancreas to exclude the presence of multiple nodules.

Partial pancreatectomy is commonly performed using either the suture-fracture technique or the dissection-ligation technique. The suture-fracture technique is easier to perform than the dissection-ligation technique and has no obvious disadvantages [37]. After incising the mesoduodenum or omentum on each side of the pancreas, non-absorbable sutures are passed around the pancreas, just proximal to the pancreatic mass. The ligature is tightened, allowing it to crush through the pancreatic parenchyma, thereby ligating vessels and ducts. Hereafter, the pancreatic tissue distal to the ligature is excised. Using the dissection-ligation technique, the pancreatic capsule is incised and bluntly dissected down to the pancreatic duct and vessels. The duct and vessels are ligated, using double ligatures, and thereafter they are transected between the two ligatures.

Segmental pancreatectomy that leaves the distal portion of a lobe is not indicated. For instance, it is tempting to remove a tumor located in the middle of either pancreatic leg en bloc with the surrounding pancreatic tissue but leave the most distal part of the pancreas in place. This part of the pancreas may still have an adequate blood supply but often has no ductal structure that leads to the duodenum, leading to local pancreatitis or sterile pancreatic abscesses. If pancreatic abscesses occur after partial pancreatectomy, they are treated by repeated ultrasound-guided aspiration and flushing of the abscess with sterile solutions instead of reoperation. Most abscesses will resolve in this manner.

The presence of metastatic disease is evaluated in two ways: (1) gross inspection of common target organs including lymph nodes and liver, and (2) on the basis of the blood glucose concentrations after the glucose infusion has stopped. All macroscopically enlarged lymph nodes should be excised and submitted for histologic examination. Enlarged lymph nodes do not always contain tumour cells but can also be enlarged because of inflammation associated with the tumour or other causes. In case of liver metastases, our approach is aggressive: a tumour debulking approach is warranted to decrease tumour mass and increases the effects of medical therapy after surgery. Partial hepatectomy is performed of all affected liver lobes and small multiple metastases are coagulated using a neodymium:yttrium aluminum garnet (Nd:YAG) surgical laser. The Nd:YAG laser already has many applications in human soft-tissue surgery, and it is also becoming more commonly used in small animal surgery (e.g. partial prostatectomy and laparoscopic ovariectomy) [38-40]. In contact mode use, the laser beam, which is passed through a flexible optical fiber, is able to cut through tissues which are in contact with the bare fiber tip. In non-contact mode use, the 1064 nm wavelength beam has a relatively deep tissue penetration (5-10 mm), making the Nd:YAG laser a useful tool for photocoagulation.

After surgery dogs should be closely monitored and adequately treated in case of post-operative complications. The most common complications of INS surgery are acute pancreatitis, persistent hypoglycaemia and diabetes mellitus. In case of persistent very low blood glucose (<2.8 mmol/L), medical therapy using diazoxide should be initiated. In addition, these dogs might also need a glucose infusion after surgery to stabilise their blood glucose concentration. In case of subnormal to normal post-operative blood glucose concentrations dogs receive frequent meals of dry food. Diazoxide therapy is only initiated when clinical signs of hypoglycaemia reoccur. Once a dog is eating properly and is able to maintain normal blood glucose levels, diazoxide therapy is discontinued.

Canine INS has a reserved prognosis, because metastasis, tumour regrowth and return of clinical signs are almost inevitable. The prognosis for dogs treated with surgery combined with medical treatment is significantly better than the prognosis for dogs receiving only medical treatment [29]. The median disease free period and survival time of dogs that underwent partial pancreatectomy were respectively 11 months (range, 0-37 months) and 14 months (range, 0-51 months) [14,15,17,29,41]. In contrast, the median survival time of 21 dogs treated medically only was 4 months (range, 0-18 months) [15,29]. Besides treatment modality, other prognostic factors are TNM stage, age and post-operative blood glucose concentrations. Dogs with clinical stage I INS have a median disease free interval of 14 months, which is significantly longer than the disease free interval of dogs with clinical stage II and III INS, which remain normoglycaemic for a median of only 1 month. Furthermore, dogs with clinical stage I and II INS have a significantly better median survival time (18 months) than dogs with clinical stage III INS (less than 6 months) [18]. Young dogs have a worse prognosis than older dogs. Dogs that are hyperglycaemic, or normoglycaemic directly post-operative, survive significantly longer than dogs with hypoglycaemia post-operatively [14].

Human insulinoma

With a frequency of 27.2%, INS are the most common pNETs [42]. In 85% of the cases INS are solitary tumours. Remaining INS are multiple and are frequently associated with multiple endocrine neoplasia 1 (MEN1) [43]. MEN1 is an autosomal dominant disorder caused by inactivation of the MEN1 gene [44,45]. The disease predisposes to the development of multiple tumours in endocrine tissues, of which the parathyroid glands (78-94%), endocrine pancreas (35-75%), and pituitary (20-65%) are most frequently affected [46]. Eighty-five to 90% of solitary non-MEN1-associated INS are small 2cm. However, sometimes they are <2cm, making it difficult to detect their malignant potential, despite the use of all currently available prognostic factors [48].

In case of INS hypoglycaemia can be controlled using diazoxide [49]. Beta-blockers may also reduce insulin secretion by INS cells and have an effect on insulin sensitivity [50]. Another commercially available symptomatic therapy entails somatostatin analogues octreotide and lanreotide [51,52]. However, if INS cells lack somatostatin receptor subtypes 2, 3 and 5, the somatostatin analogues might paradoxically lower blood glucose levels by suppressing glucagon [53,54].

Although symptomatic treatment might control hypoglycaemia for a while, in symptomatic refractory metastatic INS, or in the situation of progressive disease, antitumour therapy is indicated. Cytoreductive surgery, chemoembolisation, radiofrequency ablation, and cytotoxic chemotherapy are amongst the antitumour therapies that have been successfully used in patients with pNETs [55]. However, specific information on their efficacy in the group of patients with INS is generally lacking in the literature. Cytoreductive surgery seems to be the most efficient antitumour therapy. With chemotherapy, a lot of side effects including cardiac, renal and/or liver toxicities, myelosuppression, mucositis, fatigue, and diarrhoea occur. Therefore, development of novel therapies specifically targeting genes involved in INS tumourigenesis is warranted.

Tumourigenesis of MEN1-associated INS is fairly well understood. The MEN1 gene encodes a protein called menin and is thought to function as a tumour suppressor gene in several tissues. The anti-proliferative effect of menin could be attributed to repression of transcriptional activation mediated by JunD and Smad3, proteins which have been shown to inhibit cell growth [56]. Menin also associates with a histone methyltransferase (MHT) complex that stimulates promotor histone methylation and thereby activates gene transcription [57]. Target genes that are regulated by the menin-HMT complex include several HOX genes and CDK-inhibitors p27 and p18. In case of MEN1-associated INS, inactivating mutations of the MEN1 gene lead to loss of function of menin and downregulation of the aforementioned target genes. The resulting deregulation of cell growth, contributes to INS tumourigenesis. In addition to affecting gene expression, it is also suggested that menin can interfere with DNA replication, and is involved in DNA damage-dependent cell cycle arrest [56]. MEN1 itself does not represent a target for treatment yet, however, the relation between menin and other druggable pathways is gaining increasing interest [58].

Although MEN1-associated tumourigenesis of INS is fairly well understood, the molecular pathogenesis of the major group of sporadic INS is largely unknown, again, because most studies are conducted on the heterogeneous group of pNETS in which specific data on INS are lacking. Signalling pathways that have been implicated to stimulate cell proliferation in INS include Ras/ERK signalling, and PI3K/Akt signalling in combination with up-regulation of cyclin D1 and down-regulation of amongst others p16 [59]. Recent studies on the molecular basis of pNETS have focused on the role of vascular endothelial growth factor (VEGF) and the mammalian target of rapamycin (mTOR) [60].

mTOR is a down-stream target of PI3K/Akt signalling. This serine-theronin protein kinase is abnormally activated in pNETs, resulting in proliferation, reduced apoptosis and stimulation of angiogenesis [61]. In 2011, everolimus was approved by the US Food and Drug Administration for the treatment of pNETs. It has been described to increase median progression free survival of pNET patients, however there was no overall survival benefit between patients treated with everolimus and patients treated with a placebo [62]. Furthermore, everolimus has been reported to be associated with significant side effects, such as hyperglycaemia and diarrhea [62].

Sunitinib is thus far the only agent targeting the VEGF receptor pathway that has been approved by the US Food and Drug Administration for the treatment of unresectable, well-differentiated pNETs. Sunitinib is a potent inhibitor of multiple tyrosine kinases. Although, sunitinib has been demonstrated to have a respone rate of only 9%, it has been shown that the median progression free survival in patients with pNETs increased to 11.4 months versus 5.4 months in the placebo group [63]. Besides the low response rate, a major concern with VEGF inhibitors are comorbidities in patients with uncontrolled hypertension, renal insufficiency, or other vascular disorders caused by the vasogenic effects of VEGF inhibitors [60].

Although the knowledge regarding tumourigenesis of pNETs has improved compared to a decade ago, more insight is needed in molecular pathogenesis of sporadic INS. Only then, efficient, and safer targeted therapies can be developed for patients with INS. The dog might be a good spontaneous, large animal model of human sporadic malignant INS. Almost all canine INS are malignant, and the estimated incidence is ~10 per million per year. Therefore, the incidence of malignant INS in dogs is approximately 50 times higher compared to the incidence in men, which underlines the attractivity of canine INS as model for human malignant INS.

Cancer stem cells

Cancer stem cells (CSCs) are considered unique subpopulations inside the heterogeneous cell population of a tumour, which are solely responsible for tumour initiation, metastasis, and recurrence [64]. The origin of CSCs is highly debated. CSCs could either derive from 1) normal stem cells that undergo malignant transformation, or from 2) ‘differentiated’ tumour cells that dedifferentiate towards a stemness state [65] [66]. CSCs have been described to be able to resist systemic anti-cancer treatment by several mechanisms: CSCs could enter a quiescence state, up-regulate expression of xenobiotic efflux pumps, and enhance anti-apoptotic and DNA repair pathways to allow cell survival [67]. Therefore, CSCs are able to cause tumour relapse after systemic treatment, making them an attractive target for novel anti-cancer drugs.

In order to improve prognosis for patients with INS novel CSC-targeting adjuvant therapies are warranted that will prevent metastasis and recurrence. CSCs have been detected in exocrine pancreatic adenocarcinoma [68], however CSCs have not been identified and isolated from PNETs yet. Indirect evidence points towards the existence of CSCs in PNETs, because embryological stem cell pathways in PNETs have been demonstrated to be up-regulated compared to adjacent normal tissue samples [69]. The direct characterisation of INS CSCs was an important aim of Floryne Buishand’s thesis [70].


1. Bardeesy, N., DePinho, R.A. 2002. Pancreatic cancer biology and genetics. Nat Rev Cancer 2: 897-909.
2. Evans, H.E. 1993. Chapter 7: the digestive apparatus and abdomen: the pancreas. In: Miller’s Anatomy of the Dog. Evans H.E. and Evans S.A., eds. Saunders, Philadelphia: 458-460.
3. Cáceres, A.V., Zwingenberger, A.L., Hardam, E., Lucena, J.M, Schwarz, T. 2006. Helical computed tomographic angiography of the normal canine pancreas. Vet Radiol Ultrasound 47: 270-278.
4. Buishand, F.O., Kirpensteijn, J. 2015. The pancreas. In: BSAVA Manual of Canine and Feline Abdominal Surgery. Wiliams J. and Niles J., eds. Wiley-Blackwell, Hoboken, New Jersey: in press.
5. Heitz, P.U., Kasper, M., Polak, J.M., Klöppel, G. 1982. Pancreatic endocrine tumors: immunocytochemical analysis of 125 tumors. Hum Pathol 13: 263-271.
6. Hawkins, K.L., Summers B.A., Kuhajda, F.P., Msith, C.A. 1987. Immunocytochemistry of normal pancreatic islets and spontaneous islet cell tumors in dogs. Vet Pathol 24: 170-179.
7. O’Brien, T.D., Hayden, D.W., O’Leary, T.P., Caywood, D.D., Johnson, K.H. 1987. Canine pancreatic endocrine tumors: immunohistochemical analysis of hormone content and amyloid. Vet Pathol 24: 308-314.
8. Robben, J.H., Van Garderen, E., Mol, J.A., Wolfswinkel, J., Rijnberk, A. 2002. Locally produced growth hormone in canine insulinomas. Mol Cell Endocrinol 197: 187-195.
9. Fernandez, N.J., Barton, J., Spotswood, T. 2009. Hypoglycemia in a dog. Can Vet J 50: 423-426.
10. Madarame, H., Kayanuma, H., Shida, T., Tsuchiya, R 2009. Retrospective study of canine insulinomas: eight cases (2005-2008). Journal Vet Med Sci 71: 905-911.
11. Steiner, J.M., Bruyette, D.S. 1996. Canine insulinoma. Comp Cont Educ Pract Vet 18: 13-23.
12. Feldman, E.C., Nelson, R.W. 2004. Canine and Feline Endocrinology and Reproduction. St. Louis: Saunders, pp. 616-644.
13. Mehlhaff, C.J., Peterseon, ME, Patnaik AK, Carillo, J.M. 1995. Insulin-producing islet cell neoplasms: surgical considerations and general management in 35 dogs. J Am Anim Hosp Assoc 21: 607-612.
14. Tryfonidou, M.A., Kirpensteijn, J, Robben, J.H. 1998. A retrospective evaluation of 51 dogs with insulinoma. Vet Q 20(Suppl 1): S114-S115.
15. Tobin, R.L., Nelson, R.W., Lucroy, M..D, Wooldridge, J.D., Freeman, E.C. 1999. Outcome of surgical versus medical treatment of dogs with beta cell neoplasia: 39 cases (1990-1997). J Am Vet Med Assoc 215: 226-230.
16. Capen, C.C., Martin, S.L. 1969. Hyperinsulinism in dogs with neoplasia of the pancreatic islets: a clinical, pathologic, and ultrastructural study. Vet Pathol 6: 309–341.
17. Leifer, C.E, Peterson, M.E., Matus, R.E. 1986. Insulin-secreting tumor: diagnosis and medical and surgical management in 55 dogs. J Am Vet Med Assoc 188: 60-64.
18. Caywood, D.D., Klausner, J.S., O’Leary T.P., Withrow SJ, Richardson RC, et al. 1988. Pancreatic insulin-secreting neoplasms; clinical, diagnostic, and prognostic features in 73 dogs. J Am Anim Hosp Assoc 24: 577-584.
19. Owen, L.N. 1980. T.N.M. Classification of Tumours in Domestic Animals. Geneva, Switzerland, p. 29.
20. Moore, A.S., Nelson, RW., Henry C.J., Rassnick, K.M., Kristal, O., et al. 2002. Streptozocin for treatment of pancreatic islet cell tumors in dogs: 17 cases (1989–1999). J Am Vet Med Assoc 221: 811–818.
21. Bryson, E.R., Snead, E.C., McMillan C., MacDougall, L., Allen, A.L. 2007. Insulinoma in a dog with pre-existing insulin-dependent diabetes mellitus. J Am Anim Hosp Assoc 43: 65-69.
22. Boari, A., Barreca, A., Bestetti, G.E., Minuto, F., Venturoli, M. 1995. Hypoglycemia in a dog with a leiomyoma of the gastric wall producing an insulin-like growth factor II-like peptide. Eur J Endocrinol 132: 744-750.
23. Zini, E., Glaus, T.M., Minuto, F., Arvigo, M., Hauser, B., et al. 2007. Paraneoplastic hypoglycemia due to an insulin-like growth factor type-II secreting hepatocellular carcinoma in a dog. J Vet Int Med 21: 193-195.
24. Rogers, A.S., Luttgen, P.J. 1985. Hyperinsulinism. The Compendium on Continuing Education 7: 829-840.
25. Whipple, A.O., Frantz, V.K. 1935. Adenoma of islet cells with hyperinsulinism. Ann Surg 101: 1299-1335.
26. Robben, J.H., Pollak, Y.W., Kirpensteijn, J., Boroffka, S.A., van den Ingh, T.S., et al. 2005. Comparison of ultrasonography, computed tomography, and single-photon emission computed tomography for detection and localization of canine insulinoma. J Vet Int Med 19: 15-22.
27. Iseri, T., Yamada, K., Chijiwa, K., Nishimura, R., Matsunaga, S., et al. 2007. Dynamic computed tomography of the pancreas in nomal dogs and in a dog with pancreatic insulinoma. Vet Radiol Ultrasound 48: 328-331.
28. Mai, W., Caceres, A.V 2008. Dual-phase computed tomographic angiography in three dogs with pancreatic insulinoma. Vet Radiol Ultrasound 49: 141-148.
29. Polton, G.A., White, R.N., Brearley, M.J., Eastwood, J.M. 2007. Improved survival in a retrospective cohort of 28 dogs with insulinoma. J Small Anim Pract 48: 151-156.
30. Rijnberk, A., Kooistra, H.S. 2010. Clinical Endocrinology of Dogs and Cats. Schluttersche, Hannover, p. 320.
31. Elie, M.S., Zerbe, Z.C. 1995. Insulinoma in dogs, cats and ferrets Comp Cont Educ Pract Vet 17: 51.
32. Simpson, K.W., Stepien, R.L, Elwood, C.M., Boswood, A., Vaillant, C.R. 1995. Evaluation of the long-acting somatostatin analog Octreotide in the management of insulinoma in three dogs. J Small Anim Pract 36: 161-165.
33. Robben, J.H., Visser-Wisselaar, H.A., Rutteman, G.R, van Rijk P.P., van Donagen, A.J., et al. 1997. In vitro and in vivo detection of functional somatostatin receptors in canine insulinomas. J Nucl Med 38: 1036-1042.
34. Robben, J.H., Van den Brom, W.E., Mol, J.A., van Haeften, T.W., Rijnberk, A. 2006. Effect of octreotide on plasma concentrations of glucose, insulin, glucagon, growth hormone, and cortisol in healthy dogs and dogs with insulinoma. Res Vet Sci 80: 25–32.
35. Lohtrop, C.D. 1989. Medical treatment of neuroendocrine tumors of the gastroenteropancreatic system with somatostatin. In: Current Veterinary Therapy X. Kirk RW, ed. Philadelphia, WB Saunders: pp. 1020-1024.
36. Meleo, K. 1990. Management of insulinoma patient with refractory hypoglycemia. Probl Vet Med 2: 602-609.
37. Allen, S.W., Cornelius, L.M., Mahaffey, E.A. 1989. A comparison of two methods of partial pancreatectomy in the dog. Vet Surg 18: 274-278.
38. L’Eplattenier, H.F., Van Nimwegen, S.A., Van Sluijs, F.J., Kirpensteijn, J. 2006. Partial prostatectomy using Nd:YAG laser for management of canine prostate carcinoma. Vet Surg 35: 406-4011.
39. Van Nimwegen, S.A., Kirpensteijn, J. 2007. Comparison of Nd:YAG surgical laser and Remorgida biopolar electrosurgery forceps for canine laparoscopic ovariectomy. Vet Surg 36: 533-540.
40. Van Nimwegen, S.A., Kirpensteijn, J. 2007. Laparoscopic ovariectomy in cats: comparison of laser and bipolar electrocoagulation. J Feline Med Surg 9: 397-403.
41. Dunn, K., Heath, M.K., Herrtage M.E., Jackson, K.F., Walker, M.J. 1992. Diagnosis of insulinoma in the dog: A study of 11 cases. J Small Anim Pract 33: 514-520.
42. Heitz, P.U., Komminoth, P., Perren, A., Klimstra, D.S., Dayal, Y. 2004. Tumours of the endocrine pancreas. In: Pathology and Genetics. Tumours of endocrine organs. De Lellis RA, Lloyd RV, Heitz PU, Eng C, eds. IARC Press, Lyon: pp. 177-208.
43. Anlauf, M., Gerlach, P., Schott, M., Raffel, A., Krausch, et al. 2011. Pathology of neuroendocrine neoplasms. Chir Z Für Alle Geb oper Medizen 82: 567-573
44. Chandrasekharappa, S.C., Guru, S.C., Manickam, P., Olufemi, S.E., Collins, F.S., et al. 1997. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276: 404-407.
45. Lemmens, .I, Van de Ven, W.J., Kas, K., Zhang, C.X., Giraud, S., et al. 1997. Identification of the multiple endocrine neoplasia type 1 (MEN1) gene. The European Consortium on MEN1. Hum Mol Genet 6: 1177-1183.
46. Pieterman, C.R.C., Vriens, M.R., Dreijerink, K.M.A., Van der Luijt, R.B., Valk, G.D. 2011. Care for patients with multiple endocrine neoplasia type 1: the current evidence base. Fam Cancer 10: 157-171.
47. Alkatout, I, Friemel,J., Stiek, B., Anlauf, M., Eisenach, P.A., et al. 2015. Novel prognostic markers revealed by a proteomic approach separating benign from malignant insulinomas. Mod Pathol 28: 69-79.
48. De Herder, W.W., Niederle, B., Scoazec, J.Y., Pauwels, S., Kloppel, G., et al. 2006. Well-differentiated pancreatic tumor/carcinoma: insulinoma. Neuroendocrinology 84: 183-188.
49. Gill, G.V. Rauf, O., MacFarlane, L.A. 1997. Diazoxide treatment for insulinoma: a nation UK survey. Postgrad Med J 73: 640-641.
50. Luna, .B, Feinglox, M.N. 2001. Durg-induced hyperglycemia. J Am Med Assoc 286: 1945-1948.
51. Vezzosi, D., Bennet, A., Rochaix, P., Courbon, F., Selves, J., et al. 2005. Octreotide in insulinoma patients: efficacy on hypoglycemia, relationships with Octreoscan scintigraphy and immunostaining with anti-sst2A and anti-sst5 antibodies. Eur J Endocrinol 152: 757-767.
52. Oberg, K., Ferone, D., Kaltsas, G., Knigge, U.P., Taal, B., et al. 2009. ENETS concensus guidelines for the standards of care in neuroendocrine tumors: biotherapy. Neuroendocrinology 90: 209-213.
53. Fjällskog, M.L., Ludvigsen, E., Stridsberg, M., Oberg, K., Eriksson, B., et al. 2003. Expression of somatostatin receptor subtypes 1 to 5 in tumor tissue and intratumoral vessels in malignant endocrine pancreatic tumors. Med Oncol 20: 59-67.
54. Portela-Gomes, G.M., Stridsberg, M., Grimelius, L., Rorstad, O., Janson, E.T. 2007. Differential expression of the five somatostatin receptor subtypes in human benign and malignant insulinomas – predominance of receptor subtype 4. Endocr Pathol 18: 79-85.
55. De Herder, W.W., Van Schaik, E., Kwekkeboom, D., Feelders, R.A. 2011. New therapeutic options for metastatic malignant insulinomas. Clin Endocrinol 75: 277-284.
56. Karnik, S.K., Chen, H., McLean, G.W., Heit, J.J., Gu, X., et al. 2005. Menin regulates pancreatic islet growth by promoting histone methylation and expression of genes encoding p27Kip1 and p18INK4c. Proc Natl Acad Sci U S A 102: 14659-14664.
57. Balogh, K., Rácz, K., Patócs, A., Hunyady, L. 2006. Menin and its interacting proteins: elucidation of menin function. Trends Endocrinol Metab 17: 357-364.
58. Capurso, G., Archibugi, L., Delle, Fave, G. 2015. Molecular pathogenesis and targeted therapy of sporadic pancreatic neuroendocrine tumors. J Hepatobiliary Pancreat Sci 22:594-601.
59. Jonkers, Y.H.M., Ramaekers, F.C.S, Speel, E.J.M. 2007. Molecular alterations during insulinoma tumorigenesis. Biochim Biophys Acta 1775: 313-332.
60. Khagi, S., Saif, M.W. 2015. Pancreatic neuroendocrine tumors: targeting the molecular basis of disease. Curr Opin Oncol 27: 38-43.
61. Kasajima, A., Pavel, M., Darb-Esfahani, S., Noske, A., Stenzinger, A., et al. 2011. mTOR expression and activity patterns in gastroenteropancreatic neuroendocrine tumours. Endocr Relat Cancer 18: 181-192.
62. Yao, J.C., Shah, M.H., Ito, T., Bohas, C.L., Wolin, E.M., et al., 2011. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med 364: 514-523.
63. Raymond, E., Dahan, L., Raoul, J.L., Bang, Y.J., Borbath, I., et al., 2011. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. N Engl J Med 364: 501-513.
64. Baccelli, I., Trumpp, A. 2012. The evolving concept of cancer and metastasis stem cells. J Cell Biol 198: 281-293.
65. Visavader, J.E. 2011. Cells of origin in cancer. Nature 469: 314-322.
66. Allegra, A., Alconi, A., Penna, G., Innao, V., Gerace, D., et al. 2014. The cancer stem cell hypothesis: a guide to potential molecular targets. Cancer Invest 32: 470-495.
67. Bomken, S., Fiser, K., Heidenreich, O., Vormoor, J. 2010. Understanding the cancer stem cell. Br J Cancer 103: 439-445.
68. Hermann, P.C., Huber, S.L, Herrler, T., Aicher, A., Ellwart, J.W., et al., 2007. Distinct populations of cancer stem cells determine tumor growth and metastatic activity in human pancreatic cancer. Cell Stem Cell 1: 313-323.
69. Grande, E., Capdevila, J. Barriuso J., Antón-Aparicio, L., Castellano, D. 2012. Gastroenteropancreatic neuroendocrine tumor cancer stem cells: do they exist? Cancer Metastasis Rev 31: 47-53.
70. Buisman, F.O. 2015. Canine and human insulinoma. PhD Thesis, Utrecht University, The Netherlands.

Author: Floryne Buishand