Necrotizing enterocolitis following a single very low dose of octreotide in a patient with congenital hyperinsulinism: a case successfully managed with 18F-DOPA PET/CT-guided surgery

Abstract

Octreotide is considered a second-line treatment for congenital hyperinsulinism unresponsive to diazoxide. Necrotizing enterocolitis (NEC) is a serious adverse effect of octreotide, typically occurring in a dose-dependent manner. Here, we report a case of necrotizing enterocolitis following a single administration of a very low dose of octreotide. A female infant was admitted on day 3 of life with severe hypoglycemia. Laboratory findings revealed hyperinsulinemia and hypoketotic hypoglycemia, confirming a diagnosis of congenital hyperinsulinism. Despite diazoxide therapy, adequate glycemic control was not achieved. As a second-line intervention, a single subcutaneous injection of octreotide (1.6 μg/kg) was administered. Two days post-administration, the patient developed abdominal distension and significant vomiting. NEC was diagnosed, necessitating bowel decompression surgery. Subsequent 18F-DOPA positron emission tomography/computed tomography revealed a focal lesion extending from the pancreatic head to the body. The lesion was successfully resected with preservation of nearly the entire normal pancreas. This case highlights that even a very low dose of octreotide may precipitate necrotizing enterocolitis, warranting close monitoring. Lesion localization using 18F-DOPA positron emission tomography/computed tomography is critical in guiding surgical management of congenital hyperinsulinism.

Highlights

● Octreotide should be used cautiously in patients at risk for necrotizing enterocolitis (NEC) .

● Adverse octreotide effects are dose-dependent; careful monitoring is needed.

● 18F-DOPA PET/CT is essential to determine the optimal surgical approach.

Introduction

Congenital hyperinsulinism (CHI) is the most common cause of persistent hypoglycemia in neonates, characterized by recurrent episodes of non-ketotic hypoglycemia resulting from abnormal insulin secretion by pancreatic β-cells (1). CHI can be classified into transient and persistent forms. According to an epidemiological study by Lapidus et al., the incidence of the persistent form is approximately 3.5 per 100,000 live births (2). Oral diazoxide is the first-line treatment for hyperinsulinemic hypoglycemia. When hypoglycemia does not respond adequately to diazoxide, subcutaneous octreotide (OTT) is considered the next therapeutic option; however, necrotizing enterocolitis (NEC) has been reported as a rare but serious adverse effect (3). In a review of 428 hospitalized infants with CHI who received 490 courses of OTT, nine (2%) developed NEC (4).

Here, we report a case of NEC that developed following the administration of a very low dose of OTT. Lesion localization was successfully achieved using 18F-DOPA positron emission tomography/computed tomography (18F-DOPA PET/CT), enabling curative surgical intervention while preserving nearly all normal pancreatic tissue. To the best of our knowledge, no previous reports have documented NEC following such a low dose of OTT. Additionally, this is the first report describing a pancreatic lesion protruding at the head–body junction. This case highlights the importance of accurate lesion localization through imaging. We also present a brief literature review of NEC in patients with CHI.

Patient Description

The patient was a 3-d-old female neonate born at 37 wk and 6 d of gestation. There were no immediate life-threatening events at birth, and her family medical history was unremarkable. Her birth weight was 3,415 g (+1.8 SD), length 52.0 cm (+2.2 SD), and occipitofrontal circumference 32.4 cm (–0.4 SD). She exhibited adequate feeding and normal activity until the second day of life. On day 3, she presented with drowsiness following discharge. At a referring hospital, her blood glucose was found to be 11 mg/dL, prompting her transfer to our institution as a result of inadequate response to multiple glucose infusions.

Upon arrival, the patient demonstrated severe agitation, peripheral coldness, increased limb tone, tachycardia, and hypotension. Cardiac echocardiography revealed mildly reduced cardiac function, without intracardiac structural abnormalities or signs of pulmonary hypertension, such as septal flattening. Comprehensive blood analyses revealed marked hyperinsulinemic hypoglycemia, accompanied by low serum ketone levels and reduced free fatty acid levels despite severe hypoglycemia (Table 1). No signs of metabolic acidosis or elevated lactate, pyruvate, or ammonia levels were observed. Amino acid and urine organic acid analyses were normal, and acylcarnitine profiling revealed no disease-specific abnormalities. The glucose infusion rate (GIR) required to maintain normoglycemia exceeded 15 mg/kg/min (Fig. 1). In addition to glucose infusion therapy, the patient required catecholamine support for several days because of overall clinical instability. Diazoxide was initiated as first-line treatment for hyperinsulinemic hypoglycemia, and the dosage was gradually increased. The diazoxide dose was escalated every 2–3 d from 3 to 5, 7, 10, and finally 15 mg/kg/d. Titration was performed while confirming that the GIR required to maintain normoglycemia did not decrease. No signs of pulmonary hypertension were observed during the process of diazoxide dose escalation. Enteral nutrition was also introduced from day 3; however, it remained unstable owing to frequent episodes of vomiting. Brain magnetic resonance imaging (MRI) obtained on day 5 of life demonstrated extensive areas of high signal intensity on diffusion-weighted imaging, predominantly involving the bilateral occipital and parietal lobes. These findings were considered consistent with hypoglycemic encephalopathy. Despite titrating diazoxide to the maximum dose, the required GIR did not decrease, leading to a diagnosis of diazoxide-unresponsive hyperinsulinemic hypoglycemia.

Table 1. Laboratory findings at the previous medical visit and admission.

Fig. 1.

Changes in glucose infusion rate (GIR) required to maintain normoglycemia after initiation of treatment.

As a second-line treatment, a single subcutaneous injection of OTT (1.6 μg/kg) was administered on day 17. Blood glucose levels temporarily increased to 400 mg/dL, indicating the effectiveness of OTT. However, 2 d later, the patient developed abdominal distension and severe vomiting. Abdominal radiography revealed marked intestinal distension, and abdominal ultrasonography detected portal venous gas (Fig. 2), leading to a diagnosis of NEC. On day 19, the patient developed NEC, and oral diazoxide administration was discontinued. Exploratory laparotomy with stoma creation was subsequently performed. After the creation of the stoma, enteral nutrition stabilized, and GIR was gradually reduced. Glucagon was also considered as a second-line treatment option; however, its use was abandoned because of previous reports indicating a high frequency of precipitation and subsequent catheter occlusion (5). On day 91, the patient developed a central venous catheter infection, which led to renewed instability in enteral nutrition. Although the infection responded to antibiotic treatment, enteral feeding failed to stabilize. Based on these episodes, it was concluded that discontinuing glucose infusion solely through medical management was challenging, necessitating surgical intervention.

Fig. 2.

Abdominal radiograph and abdominal ultrasound image on day 19 of treatment. (A) A large amount of intraluminal gas and significant intestinal distension was observed. (B) Abdominal ultrasound shows gas within the portal vein (white arrow).

After the development of NEC, a Mendelian genetic disease sequence panel analysis was performed to determine the optimal surgical intervention method. A paternally inherited variant in KCNJ11 (NM_000525.3): c.902G>A (p.Arg301His) was identified. This variant was classified as likely pathogenic (PS1+PM1+PM2+PP1+PP2+PP3+PP4) based on the guidelines of the American College of Medical Genetics and Genomics/Association for Molecular Pathology (6). The R301H variant has previously been reported in association with diazoxide-unresponsive and focal forms of CHI (7). Based on this genotype–phenotype correlation, 18F-DOPA PET/CT was performed on day 133 to localize the lesion. This scan revealed a protruding lesion at the junction of the pancreatic head and body, with 18F-DOPA accumulation confirmed at this site (Fig. 3). Focal disease was diagnosed, and local pancreatic resection along with stoma closure was performed on day 189. The surgical approach involved extending the incision from the existing stoma site, and the pancreatic head was identified first. The lesion was located slightly toward the pancreatic tail, relative to the superior mesenteric vein. It appeared as a mushroom-like protrusion from the pancreas and exhibited a subtle difference in color compared to the surrounding pancreatic parenchyma on gross inspection. Intraoperative ultrasonography revealed that the protruding mass was hypoechoic relative to the adjacent pancreatic tissue. It extended ventrally and caudally and was located at a safe distance from the main pancreatic duct, enabling resection without ductal injury. The excised lesion measured approximately 10 mm in length, 10 mm in width, and 8 mm in depth and had a moderately firm consistency. Negative surgical margins were confirmed intraoperatively, and the procedure was completed. Following resection, the GIR required to maintain normoglycemia decreased significantly, indicating the effectiveness of the surgical intervention. Pathological examination revealed diffuse proliferation of pancreatic island-like cells (Fig. 4). Postoperatively, episodes of hypoglycemia resolved, and glucose infusion was discontinued on day 210. The patient was discharged on day 236 and remained hypoglycemia-free after surgery.

Fig. 3.

18F-DOPA PET/CT on day 133. A protruding lesion is observed at the junction of the pancreatic head and body, with 18F-DOPA accumulation in the same region (white arrow).

Fig. 4.

Histopathological images of the lesion. (A) Diffuse proliferation of pancreatic islet-like cells (Hematoxylin and Eosin stain). (B) Immunostaining for insulin shows diffuse proliferation of insulin-secreting cells.

After discharge, the patient experienced no further hypoglycemic episodes. A combination of enteral tube feeding and oral feeding resulted in satisfactory weight gain. At 15 mo of age, the patient experienced several episodes of myoclonic jerking in the upper limbs. Electroencephalography revealed hypsarrhythmia, leading to a diagnosis of West syndrome, for which adrenocorticotropic hormone (ACTH) therapy was initiated. Following treatment, the patient remained seizure-free with oral valproate monotherapy. Brain MRI performed at 16 mo of age, following ACTH therapy, revealed a global reduction in cerebral cortical volume. By 24 mo of age, oral intake was fully stabilized, and enteral tube feeding was discontinued. Due to spastic paralysis caused by severe neonatal hypoglycemia, the timing of head control could not be reliably assessed in this patient. The patient achieved rolling over at 11 mo, independent sitting at 12 mo, and was able to pull to stand at 24 mo. At 24 mo of age, no meaningful words had been observed, indicating a moderate psychomotor developmental delay.

Discussion

This report details a case of CHI that progressed to NEC following treatment with a very low dose of OTT and was ultimately cured through surgical intervention. OTT acts on somatostatin receptors located on pancreatic β-cells, suppressing insulin secretion by inhibiting intracellular Ca²⁺ influx through calcium channels by opening ATP-sensitive potassium channels, thereby improving hyperinsulinemic hypoglycemia. However, OTT also reduces mesenteric arterial blood flow, a known risk factor for NEC. This vasoconstrictive effect is considered dose-dependent (8). To date, eight cases of NEC associated with OTT use in patients with CHI have been reported (3, 9,10,11). Many of these cases occurred in the absence of other recognized NEC risk factors, such as preterm birth, low birth weight, congenital heart disease, maternal diabetes, or small for gestational age status (12). In the present case, NEC developed after a single, very low subcutaneous dose of OTT (Table 2).

Table 2. Summary of all cases of octreotide-associated necrotizing enterocolitis in patients with congenital hyperinsulinism.

Multiple factors are believed to have contributed to the development of NEC following the administration of a very low dose of OTT in this case. The patient’s poor general condition at treatment initiation, which required the use of catecholamines, is noteworthy. One risk factor for the development of NEC is intestinal ischemia (12). At the onset of treatment, the patient was in shock with impaired cardiac function, which likely led to compromised mucosal integrity as a result of intestinal ischemia, making the patient more susceptible to the effects of OTT. Additionally, the potential side effects of diazoxides should be considered. Previous reviews have reported that diazoxide use may be a risk factor for NEC in preterm and full-term infants (13,14,15). Although the exact mechanism by which diazoxide contributes to NEC remains unclear, it has been suggested that its action on calcium-activated potassium channels in smooth muscle may lead to decreased intestinal motility and stagnation of luminal contents, possibly contributing to NEC development (15).

In the present case, at the start of treatment on day 3 of life, the patient exhibited agitation, likely as a result of hypoglycemia; however, there was no abdominal distension, and radiographs did not show any signs of ileus. Following the initiation of oral diazoxide and tube feeding, the patient began to exhibit vomiting and instability, although radiographs remained normal. Given these symptoms, it is plausible that the patient was already at a stage suggestive of NEC according to Bell’s criteria (16). In cases where complex risk factors for NEC exist—such as gastrointestinal symptoms or an unstable general condition requiring the use of vasopressors—careful consideration is needed before using OTT. Additionally, even when OTT is used, close monitoring during treatment is essential.

In this case, the lesion was located using 18F-DOPA PET/CT, enabling curative surgical treatment. 18F-DOPA PET/CT has been reported to be useful in identifying lesions in children with CHI (17). When paternally inherited pathogenic variants in KCNJ11 or ABCC8 are identified, it is recommended to evaluate their localization using 18F-DOPA PET/CT, as there may be a focal form of the disease (18). In the present case, a protruding lesion was identified at the transition between the pancreatic head and the body, enabling resection while preserving almost the entire normal pancreas. Consequently, postoperative hypoglycemia completely resolved, achieving a full cure.

Kanamori et al. reported 14 cases in Japan in which localization was assessed using 18F-DOPA PET/CT. Although localization was successful in 12 cases, none displayed a protruding lesion similar to that seen in this report (19). Although Larsen et al. described the histological characteristics of CHI and noted that lesions may protrude from the pancreatic surface (20), to our knowledge, this is the first published case to clearly demonstrate such a protruding lesion on imaging and intraoperatively. Preserving as much normal pancreatic tissue as possible lowers the risk of complications such as diabetes, highlighting the critical importance of accurate lesion localization in patients with CHI.

Although the clinical value of 18F-DOPA PET/CT in guiding surgical decision-making has been demonstrated in previous studies, access to this imaging modality remains limited in Japan because of insufficient infrastructure. This case not only reaffirms the importance of lesion localization through 18F-DOPA PET/CT in patients with paternally inherited genetic abnormalities in K_ATP channels but also highlights the urgent need to establish a nationwide system that ensures timely and equitable access to this diagnostic tool, particularly for those requiring surgical intervention.

Conclusion

Although it cannot be definitively concluded that OTT induced NEC, its administration in combination with high-dose diazoxide or under conditions of systemic instability may still precipitate NEC, even when given as a single low dose. Therefore, it is important to provide caregivers with a thorough explanation of the potential risks prior to administration, even when the dose of OTT is minimal. For patients with persistent CHI, identifying lesion localization—particularly in those with a potentially focal form—is essential for optimal management.

Conflict of interests

The authors declare no conflicts of interest.