Successful treatment with sirolimus in patients with congenital hyperinsulinism: A case series

INTRODUCTION

Congenital hyperinsulinism (CHI) is the most frequent cause of persistent hypoglycemia in early childhood, typically presenting in the neonatal period. It is a heterogeneous genetic disorder with an incidence of 1/30,000–50,000 newborns,^[1] characterized by an excessive secretion of insulin by pancreatic beta-cells regardless of blood glucose levels. These patients are at high risk of persistent severe hypoketotic hypoglycemia, which can lead to irreversible neurological sequelae.

Mutations in at least 15 genes (ABCC8, KCNJ11, GCK, GDH, HADH, HNF4A, HNF1A, UCP2, SLC16A1, PMM2, HK1, PGM1, FOXA2, CACNA1D, and EIF2S3) are currently described with CHI.^[2] The most common causes of CHI are mutations in ABCC8, followed by KCNJ11, encoding respectively the SUR1 and Kir 6.2 subunits of adenosine triphosphatase-sensitive K+ channels (K-ATP), being mostly recessive mutations.

There are three histological forms of CHI: Focal, diffuse, or atypical. In focal CHI, the surgical approach with focal pancreatectomy is curative. However, in unresponsive diffuse forms of CHI to medical therapy, subtotal pancreatectomy may be required, which is associated with the possibility of persistent hypoglycemia and, in the long-term, exocrine pancreatic insufficiency and diabetes mellitus. For this reason, medical therapy should be the first option.

Regarding medical therapy, diazoxide is the first-line drug for treating CHI. It is used in a dose of 5–20 mg/kg/day and is administered orally. The most common side effects are hypertrichosis and fluid retention, hence it is commonly used with a thiazide diuretic. Diazoxide needs an intact K-ATP channel to work, which is why it is not effective in many cases of CHI. In fact, recessive forms associated with the ABCC8/KCNJ11 genes are the most severe forms of CHI and usually are unresponsive to diazoxide. The second-line of medical treatment is octreotide, a long-acting analog of somatostatin that inhibits insulin secretion distal to the K-ATP channel. It is used in a dose of 10–50 μg/kg/day subcutaneously or intravenously. However, it has a transient efficacy due to the tachyphylaxis phenomenon. Side effects include suppression of growth, steatorrhea, cholelithiasis, abdominal distension, necrotizing enterocolitis, raised transaminases, and hepatitis.^[2,3] Other drugs used, with low efficacy, are nifedipine and glucagon.

Standard therapies have limited efficacy in CHI; therefore, alternative medical treatment alternatives should be considered especially in non-responsive forms of CHI. In 2014,^[4] sirolimus, an immunosuppressive agent, mammalian target of rapamycin inhibitor and antiproliferative of beta cells, was used in unresponsive patients with CHI. Unlike classical therapy, response to sirolimus is independent of genetic etiology. Since then, there has been limited and controversial literature on its use in CHI.

We report here two cases of diffuse CHI successfully treated with sirolimus. The characteristics of patients are summarized in Table 1.

CASE SERIES

Case 1

A preterm girl was born at 31 weeks and 5 days of gestation secondary to a hypertensive disorder of pregnancy, with a birth weight of 1920 g. It was a spontaneous vaginal delivery. The baby was hypotonic with poor respiratory effort, which improved after nasal intermittent positive-pressure ventilation was applied.

Hypoglycemia was detected immediately after birth. Continuous intravenous glucose infusion was started due to persistent hypoglycemia. The maximum glucose infusion rate (GIR) required was 28 mg/kg/min. At 48 h of life, hydrocortisone was started with a GIR of 12 mg/kg/min and at 72 h of life diazoxide therapy was started after hyperinsulinism was confirmed with blood glucose 31 mg/dL, insulin 19.5 µIU/mL and C-peptide 2.68 ng/mL with low free fatty acid levels and negative ketonemia. Hypoglycemia persisted despite maximum permissible doses of diazoxide (20 mg/kg/day), hence octreotide was started, with progressive increase up to 30 µg/kg/day. Diazoxide was discontinued after 2 weeks of treatment. Hypoglycemia improved, and therefore, intravenous glucose was withdrawn. At 1 month of life, the patient was switched to intramuscular injections of octreotide every 4 weeks and hospital discharge was decided at 2 months of life.

Positron emission tomography-computed tomography (PET-CT) using 18F-fluorodopa showed diffuse pancreatic uptake [Figure 1]. Genetic analysis revealed a heterozygous mutation for c.902 G>A and c.329G>A of the KCNJ11 gene, both inherited from the mother and father, respectively. During follow-up, there was an acceptable control of hypoglycemia episodes initially, but in the 3^rd month of life, the patient presented with a seizure due to severe hypoglycemia. Therapy with sirolimus was started at an initial dose of 0.5 mg/m²/day. Initially, sirolimus levels were measured every 5 days until achieving optimal blood levels of 5–15 ng/mL. Once the blood sirolimus levels were stable, follow-up assessments were spaced out progressively.

Close

Figure 1:

18F-DOPA positron emission tomography-computed tomography scan showing diffuse uptake body and tail of pancreas in patient 1. Physiological uptake by kidneys is also observed.

Export to PPT

Sirolimus was withdrawn at the age of 3 years, after achieving stable euglycemia controls. At the time of this report, the patient is 7 years old, and she has had a favorable evolution with normal neurodevelopment according to her age.

DISCUSSION

Management of diazoxide and octreotide in unresponsive diffuse CHI is a medical therapeutic challenge. Patients with recessive forms of CHI due to mutations in ABCC8/KCNJ11 usually have a limited response to diazoxide. Our two patients had compound heterozygous variants in the ABCC8 and KCNJ11 genes and were therefore diazoxide-unresponsive. These cases account for 60–70% of CHI and are usually the diffuse forms. In diffuse forms of CHI it is recommended to start with conservative management to avoid surgery.^[5]

In our cases, diazoxide was started in the first week of life, but required maximum doses, resulting in secondary fluid retention and persistent hypoglycemia, forcing its withdrawal and introduction of octreotide therapy. The first case initially responded to octreotide; however, tachyphylaxis was observed. The second case did not respond to octreotide, and medical alternatives were proposed (glucagon and nifedipine), which were also ineffective. Due to the lack of response and to avoid surgery, sirolimus therapy was started.

In 2014, Senniappan et al.^[4] described four cases of severe CHI unresponsive to maximal doses of diazoxide and octreotide. After initiating sirolimus, all of them showed a good response to therapy without significant adverse effects; they only presented with transient mild raised transaminases and hypertriglyceridemia. In another study,^[6] an 8-year-old boy with severe CHI presented a drastic improvement in hypoglycemia events after sirolimus. As an adverse effect, he presented with mucositis and mild acne.

Abraham et al.^[7] described the case of a neonate with severe CHI with a homozygous mutation for the ABCC8 gene that did not respond to medical or surgical treatment, and achieved normoglycemia following sirolimus therapy. Similarly, another study^[8] managed to avoid subtotal pancreatectomy with a homozygous mutation for the KCNJ11 gene, without showing any adverse effects. Another study,^[9] which included seven neonates with recessive mutations for ABCC8 and KCNJ11 genes with refractoriness to standard therapy, showed the effectiveness of sirolimus in six of them. Yet another clinical study^[10] showed a decrease in hypoglycemia events in five of six patients, with only one of them presenting mild hypertransaminasemia.

However, since its use in diffuse CHI, several publications have cautioned against its use in infants, mainly related to the lack of safety of sirolimus. The most commonly reported adverse effects are mucositis, immunosuppression, increased risk of infections, diarrhea, renal dysfunction, pneumonitis, raised transaminases, and hyperlipidemia, many of them reversible with dose reduction. Monitoring of blood sirolimus levels is therefore of vital importance, with an optimal therapeutic level between 5 and 15 ng/mL. In our cases, monitoring was performed every 5 days initially. Follow-up assessments were performed every 15 days, then monthly, and eventually at longer intervals as clinical stability was maintained.

In our patients, to avoid the adverse effects of sirolimus, particularly immunosuppression, we monitored them without detecting white blood cell count alterations. Regarding hepatotoxicity, no alterations in liver enzymes were observed, and renal function remained intact. To the best of our knowledge, therapy success was partly due to the fact that we closely monitored blood sirolimus concentration.

CONCLUSION

We consider that a trial with sirolimus should be considered before pancreatomy, in medical treatment-resistant diffuse CHI (diazoxide and octreotide) affected by ABCC8 and KCNJ11 mutations. In contrast to diazoxide, the patients who are receiving sirolimus therapy require intensive monitoring due to its immunosuppressive effects. In infants unresponsive to first-line treatment also sirolimus should be considered. In our case series, sirolimus was effective in achieving euglycemia without any side-effects. As sirolimus is not currently an approved drug for CHI, further collaborative research is required to strengthen our findings and reinforce the evidence for its regular use in CHI.