Two threats in early life: congenital hyperinsulinemic hypoglycemia and thrombosis

Introduction

Hyperinsulinemic hypoglycemia (HH) is a heterogeneous group of disorders that constitute the most common cause of hypoglycemia in infants and neonates. It is characterized by dysregulated insulin secretion from pancreatic beta cells. ¹ Because hypoglycemia can lead to severe, permanent brain damage, it is crucial to recognize and treat HH early and effectively. Clinically, the condition may be transient or persistent; persistent forms are typically congenital.

In the management of congenital HH, patients require central venous catheterization (CVC) due to the initial need for high-dose glucose infusion. Although CVC is an essential part of the treatment and care process of hyperinsulinism, it carries risks of complications such as thrombosis, phlebitis, catheter dislodgement or occlusion, and infection. ² Thrombosis is one of the most frequent and serious complications. The incidence of thrombosis in childhood varies significantly according to age, underlying disease, and clinical setting.

There is currently no data regarding the development, frequency, and risk factors of thrombosis in patients with congenital persistent HH. It is hypothesized that the likelihood of thrombosis development is high in these cases due to the requirement for long-term intravenous therapy and frequent catheter interventions. The aim of our study is to emphasize the development of catheter-related thrombosis in persistent HH cases and to identify the predisposing factors that may lead to the development of thrombosis.

Method

This retrospective, cross-sectional, descriptive study included patients diagnosed with persistent HH and followed at our university’s Pediatric Endocrinology Clinic from January 2000 to December 2025. Inclusion criteria were defined as follows: diagnosis of persistent HH with negative blood ketones during hypoglycemia, fasting insulin > 2 mIU/mL, normal counter-regulatory hormone profiles (cortisol, growth hormone, ACTH), age < 18 years at diagnosis, and active follow-up at our institution. Patients diagnosed after age 18, those with other causes of hypoglycemia (fatty acid oxidation defects, gluconeogenesis disorders, glycogen storage diseases), or those with transient/secondary hyperinsulinism (IUGR, maternal diabetes, asphyxia, post-gastrointestinal surgery) were excluded.

Anthropometric measurements (weight, height, BMI, and respective SDS scores) and biochemical parameters at the time of hypoglycemia were recorded. Therapeutic interventions (octreotide, diazoxide, glucagon, calcium channel blockers), dosages, intravenous glucose infusion rates, and surgical history were documented. Catheter-related data, including type, duration of use, and the onset and management of thrombosis, were collected. Radiographic evaluations included abdominal ultrasonography (USG), echocardiography, and Doppler USG for catheter localization. Risk factors for thrombosis (catheter site, treatment duration, glucose infusion duration, and frequency of hypoglycemic episodes) and their association with specific mutations were analyzed.

Results

A total of 20 patients were included in the study, of whom 55% (n = 11) were male and 45% (n = 9) were female. The median age at diagnosis was 1 day (range: 1–195 days). Anthropometric characteristics at diagnosis are summarized in Table 1. Regarding birth weight categories, 50% (n = 10) of the cases were large for gestational age (LGA), while two cases were small for gestational age (SGA).

OPEN IN VIEWER

Table 1.

Anthropometric characteristics at diagnosis.

Characteristics	Mean ± SD	Minimum	Maximum
Body Weight (kg)	4.4 ± 0.9	3.0	5.8
Height (cm)	53.8 ± 4.2	47.0	60.0
Height SDS	1.49 ± 1.88	−1.13	4.19
Body Mass Index (kg/m2)	14.8 ± 1.6	12.5	17.2
BMI SDS	0.69 ± 1.09	−0.88	2.42
Gestational Age (weeks)	37 + 4	29 + 4	42

BMI, body mass index; SD, standard deviation; SDS, standard deviation score.

Laboratory evaluation during hypoglycemic episodes (blood glucose < 50 mg/dL) revealed negative blood β-hydroxybutyrate levels. Hyperinsulinism was confirmed by a mean fasting insulin levels of 16.6 mIU/mL (range: 2.6–62.7) and a mean C-peptide levels of 5.2 ng/mL (range: 0.6–8.6). Liver and kidney function tests, as well as counter-regulatory hormones (ACTH, cortisol, growth hormone), lactate, ammonia, organic acids, and acylcarnitine profiles, were within normal limits.

Genetic analysis identified ABCC8 mutations in eight cases, KCNJ11 mutations in five cases, and HADH mutations in one case. No mutations were detected in the remaining six cases. Focal involvement was present in one patient with a mutation, while diffuses involvement was observed in the others.

A total of 23 central venous catheterizations were performed in 15 cases, with a median catheter dwell time of 38 days. Thrombosis developed in nine cases, all of which were of venous origin. Catheter-related thrombosis was observed in six cases, while three cases presented with intracranial thrombosis independent of the catheter. The localization of the catheters and the origins of the thrombi are detailed in Table 2. The median age at the time of thrombosis detection was 74 days (range: 14 days–9 years); in two cases, thrombosis was identified during the chronic phase at an older age. Four cases (44.4%) received enoxaparin for thrombosis treatment, all of which achieved complete resolution. No complications were observed in any of the enoxaparin-treated cases. Among the cases that developed thrombosis, one was found to be heterozygous for Factor V Leiden and prothrombin mutations, indicating a predisposition to thrombophilia. Catheter-related infection occurred in five cases (33.3%), with cultures yielding Staphylococcus aureus, Klebsiella pneumoniae, and Pseudomonas aeruginosa. Thrombosis was present in 80% of cases with catheter-related infections.

OPEN IN VIEWER

Table 2.

Catheter and thrombosis localizations.

Catheter localization: n (%)	Thrombosis localization: n (%)
Superior vena cava 9 (39.1%)	Hepatic vein 2 (22.2%)
Umbilical vein 9 (39.1%)	Left portal vein 2 (22.2%)
Internal jugular vein 4 (17.4%)	Cranial 3 (33.3%)
Femoral vein 1 (4.4%)	Torcula
	Bi-occipital
	Cortikal (Laminar Necrosis)
	Internal jugular vein 1 (11.2%)
	Intracardiac 1 (11.2%)

No significant difference in mutation characteristics was observed when comparing cases with and without thrombosis (p > 0.05). The maximum glucose infusion rate, duration of glucose infusion, and maximum doses of diazoxide and octreotide did not differ significantly between the groups. However, the thrombosis group had a significantly higher number of days with hypoglycemia and total hypoglycemic attacks (p < 0.05). There was no statistically significant difference between patients who developed thrombosis and those who did not in terms of platelet counts and C-reactive protein levels measured within the first 24 h after a hypoglycemic episode (p = 0.139 and p = 0.681). Four cases of surgery were performed within the thrombosis group due to refractory hypoglycemia, and thrombosis was identified prior to the surgical intervention in three of these cases. The evaluation of potential risk factors for the development of thrombosis is presented in Figures 1–3.

OPEN IN VIEWER

Figure 1.

Maximum therapeutic doses.

OPEN IN VIEWER

Figure 2.

Comparison of treatment parameters between patients with and without thrombosis-1.

OPEN IN VIEWER

Figure 3.

Comparison of treatment parameters between patients with and without thrombosis-2.

In the regression analysis conducted to evaluate risk factors contributing to the development of thrombosis, the number of hypoglycemic attacks was identified as the variable closest to the threshold of statistical significance (p = 0.078; Table 3).

OPEN IN VIEWER

Table 3.

Univariate regression analysis of variables associated with thrombosis risk.

Characteristics	Score	df	p-Value
Max glucose infusion rate	0.424	1	0.515
Duration of glucose infusion (days)	0.02	1	0.969
Catheter dwell time (days)	2.51	1	0.113
Number of hypoglycemic episodes	3.106	1	0.078
Days with hypoglycemia	2.628	1	0.105
Max diazoxide dose (mg/kg/g)	0.164	1	0.686
Max octreotide dose (mcg/kg/g)	0.068	1	0.795

During the follow-up period, one case was diagnosed with neuroblastoma, two cases developed diabetes mellitus following pancreatectomy, and one case underwent liver transplantation due to cholangitis and hepatic failure.

Discussion

The primary objective in the management of congenital persistent HH is to prevent recurrent and prolonged episodes of hypoglycemia, which can be clinically challenging. To prevent brain injury, serum blood glucose levels must be maintained above 70 mg/dL, ¹ consequently, patients require high-rate intravenous glucose infusion at the onset of treatment, necessitating central venous catheterization. One of the most common and severe complications of central venous catheterization is thrombosis development. The aim of this study is to determine the risk of thrombosis and the predisposing factors associated with it in cases diagnosed with congenital HH.

In our study, catheter-related thrombosis was identified in six cases (26%), with an incidence rate of 9.2 per 1000 catheter days. In pediatric population studies, the frequency of catheter-related thrombosis ranges from 4% to 22%.^3,4 This variation may be attributed to differences in clinical catheterization techniques, patient-specific risk factors, catheter types, and insertion sites. In a case series reporting on HH patients, the incidence of thrombosis was found to be 18%, or 4.2 per 1000 catheter days. ⁵ Compared to this initial report, the thrombosis rate in our study is slightly higher. Furthermore, three cases presented with intracranial thrombosis (bi-occipital, torcula, and laminar necrosis) independent of the catheter. This finding suggests the presence of predisposing factors that increase the risk of catheter-independent thrombosis in HH patients.

A study investigating the association between HH cases and thrombosis, homozygous or compound heterozygous ABCC8 and KCNJ11 mutations were found to correlate with thrombosis. ⁵ Although homozygous ABCC8 and KCNJ11 mutations were the most frequently detected genetic variants in our study, no significant association was observed between the type of mutation and the development of thrombosis. All cases of catheter-related thrombosis were asymptomatic; notably, two cases—one with portal vein thrombosis and the other with bi-occipital thrombosis—were diagnosed during the chronic phase. Regarding coagulation-related genes, Factor V Leiden and prothrombin heterozygosity were identified in only one case with thrombosis; however, these markers were not screened in cases without thrombosis. Levels of antithrombin III, protein-C, and protein-S were within normal limits. It has been demonstrated that homozygous mutations in Factor V Leiden or prothrombin increase the risk of catheter-related thrombosis fourfold. ⁶

Statistical analysis revealed no significant association between thrombosis risk and maximum glucose concentration, infusion duration, or peak doses of octreotide and diazoxide. While the mean duration of catheterization was higher in the thrombosis group compared to the nonthrombosis group, this difference did not reach statistical significance. However, the number of days with hypoglycemia and the frequency of hypoglycemic episodes were significantly higher in the group that developed thrombosis (p < 0.05). Studies indicate an association between severe, recurrent hypoglycemia and cardiovascular morbidity and mortality.^7,8 Hypoglycemia has been shown to enhance atherothrombotic mechanisms, impair fibrinolytic balance, and lead to endothelial dysfunction.^9,10 The higher frequency of hypoglycemic episodes in cases with thrombosis may be attributed to endothelial dysfunction and activation of the thrombotic cascade. Additionally, the administration of glucose boluses and continuous high-concentration glucose infusions to treat recurrent hypoglycemia may also contribute. High-concentration glucose infusions have been reported to increase the risk of catheter-related thrombosis. ¹¹ In our study, the lack of difference in maximum glucose concentration and infusion duration between the groups suggests that thrombosis may be more closely related to recurrent hypoglycemic episodes and their acute correction with high-concentration fluids.

In addition, hematologic parameters measured within the first 24 h after the thrombotic event was reviewed. Platelet counts and C-reactive protein levels did not differ significantly between patients with and without thrombosis. However, it should be noted that hypoglycemia-related inflammatory and prothrombotic responses are typically transient and dynamic.^12,13 Therefore, single measurements obtained after the thrombotic event may not adequately reflect the underlying pathophysiological processes. Prospective studies with serial measurements following hypoglycemic episodes would be more appropriate to clarify the relationship between hypoglycemia-related inflammatory activation and thrombosis risk.

Hyperinsulinemia is a condition that predisposes to thrombosis via coagulation and fibrinolytic systems, independent of other factors. It inhibits the fibrinolytic system by increasing PAI-1 levels, regardless of hyperglycemia. ¹⁴ Furthermore, a correlation has been established between insulin and various coagulation factors, including Factor VII and fibrinogen, which leads to clotting by activating the coagulation cascade.^15,16 While this mechanism is well documented in patients with type 2 diabetes mellitus, a similar process may be present in patients with HH. This hyperinsulinemia-driven prothrombotic mechanism might be one of the predisposing factors for the three catheter-independent intracranial thrombosis observed in our study.

Although the catheter-related thrombosis in our cohort were asymptomatic, they were detected in the acute phase through ultrasonographic or echocardiographic imaging performed prior to catheter removal. Chronic thrombosis, however, were identified via Doppler ultrasonography or cranial magnetic resonance imaging during complication screening in long-term follow-up. While routine sonographic screening for catheter-related thrombosis is not recommended in neonatal and intensive care units, it is suggested for patients with a history of catheter-related infection or multiple catheterizations. ¹⁷

Conclusion

In this study, which highlights thrombosis as a prevalent comorbidity in HH, frequent hypoglycemic episodes were identified as a significant predisposing factor for thrombus development. It is anticipated that the inherent nature of the disease, along with prolonged and frequent central venous catheterization and high-dose glucose infusions, further elevates the risk of thrombosis. Therefore, performing routine sonographic screening in patients with HH—even in the absence of clinical symptoms—can facilitate early diagnosis and provide complication-free treatment options. In addition to the morbidity associated with hypoglycemia itself, the increased risk of thrombosis poses a further clinical burden in cases of congenital HH; consequently, patients with treatment-resistant hypoglycemia, in particular, should be closely monitored in this regard.

Supplemental Material