Sclerotherapy combined with sirolimus for the treatment of complex cervicofacial lymphatic malformations in infants: avoiding the need for tracheostomy

Abstract

Background

Lymphatic malformations (LMs) are low-flow, congenital lesions commonly presenting as asymptomatic masses in the head and neck. However, large lymphangiomas can significantly affect breathing or swallowing, posing considerable treatment challenges.

Methods

A retrospective analysis of complex cervicofacial LMs in infants was conducted over the past 8 years at the Department of Radiology. Patients were included if they had complex cervicofacial LMs. The size and type of LMs were assessed using ultrasound or MRI. All patients underwent sclerotherapy combined with sirolimus treatment. Treatment outcomes were evaluated through clinical examination and imaging findings.

Results

Nineteen infants with large and extensive LMs of the head and neck were identified, including 12 males and 7 females. Thirteen patients had macrocystic lesions, five had mixed lesions, and one had microcystic lesions. Posttreatment, 18 children showed a size reduction of more than 75%, and one case demonstrated a reduction of 51–75%. Mild-to-moderate fever was observed in four cases postoperatively, and two cases experienced localized swelling. None of the 19 cases developed serious adverse reactions, such as allergies, pulmonary fibrosis, nerve injury, or skin necrosis.

Conclusions

Sclerotherapy combined with sirolimus appears to be a safe and effective treatment for complex cervicofacial LMs in infants. This approach reduced the necessity for tracheotomy in affected children.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Managing complex cervicofacial lymphatic malformations (LMs) in infants presents unique challenges because of their proximity to vital structures and poor anatomical demarcation.

• Tracheostomy should be considered in cases where the lesion encircles more than half of the airway.

• A standardized treatment protocol for complex cervicofacial LMs has not been established.

WHAT THIS STUDY ADDS

• Efficacy of sclerotherapy combined with sirolimus in treating infants with complex cervicofacial LMs.

• In this study, the risk of permanent nerve damage associated with tracheotomy and surgical intervention, as well as complications such as tracheal chondromalacia, were avoided.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• We recommend that sclerotherapy combined with sirolimus be considered the first-line treatment for complex cervicofacial LMs in infants as this minimized injury, was repeatable, and significantly reduced perioperative risks.

• Surgery should be reserved as an alternative or for long-term restorative needs.

Introduction

Lymphatic malformations (LMs) are localized, slow-flow vascular malformations commonly found in the head and neck region. These malformations result from developmental anomalies characterized by abnormal lymphatic channels and cystic spaces of varying sizes. LMs are categorized into three primary types: macrocystic, microcystic, and mixed (a combination of macrocystic and microcystic components).¹ LMs rarely resolve spontaneously.²

Various treatment modalities exist for LMs, including observation, surgery, sclerotherapy, laser therapy, and drug therapy. However, a standardized treatment protocol for complex cervicofacial LMs has not been established due to their large size, cross-regional growth, high treatment complexity, and variable therapeutic outcomes. Managing complex cervicofacial LMs presents unique challenges because of their proximity to vital structures and poor anatomical demarcation.³ In cases where the lesion encircles more than half of the airway, tracheostomy should be considered.⁴Historically, surgery was the primary, and often sole, treatment option for these lesions. However, complete surgical removal is challenging due to the proximity of LMs to critical structures and their tendency to grow along muscle and vascular spaces, leading to a high risk of complications.⁵Here we present 19 cases of complex cervicofacial LMs in infants successfully treated with sclerotherapy combined with sirolimus.

Methods

Patients

This retrospective study was conducted following approval from the Institutional Research Ethics Board of Anhui Provincial Children's Hospital. We retrospectively reviewed our database of LM pediatrics treated from 2016 to 2023. All patients were evaluated by two radiologists with specialty expertise in vascular and lymphatic anomaly and underwent a comprehensive imaging evaluation, including MRI or/and ultrasound. Only patients with complex LMs in the cervical region as confirmed by two board certified radiologists were included in this study. Infants diagnosed with combined vascular malformations, such as capillary–lymphatic malformations, were excluded. Selected patients were treated with sclerotherapy combined with sirolimus at the Department of Radiology.

The cohort included 12 males and 7 females, ranging in age from 5 days to 23 months, with a median age of 6 months. All lesions had a minimum surface diameter greater than 5 cm, and none of the patients had received prior treatment. The total volume of each lesion was calculated using the method described by Lass and Brinsden, measured using Doppler ultrasonography or MRI.⁶ Lesion volume was calculated using the ellipsoid formula: volume = 0.523 × diameter 1 × diameter 2 × diameter 3.

All patients presented with cystic masses in the head and neck as their primary clinical symptom. The masses were soft, with no redness, swelling, elevated skin temperature, or tenderness. However, when associated with bleeding, the masses became rapidly enlarged, tender, and firm. Written informed consent was obtained from the parents of all participants.

Treatment

A solution was prepared by dissolving 15 000 units of bleomycin in 5 mL saline, to which 5 mL iohexol was added, yielding a concentration of 1500 units/mL. The dosage of bleomycin was 3000 units/m² of body surface area. Under digital subtraction angiography (DSA) or ultrasound guidance, the lesion was punctured using a 2.5 mL syringe or a 22G needle. For macrocystic or mixed-type (macrocystic-dominant) LMs, lymph fluid was aspirated from the most prominent area of the mass. The fluid was typically light yellow or dark red and non-coagulated. After securing the needle, anhydrous ethanol (30–50% of the extracted fluid volume) was injected into the cystic cavity, followed by lavage. After 5 min the ethanol was aspirated, and a mixture of bleomycin, saline, and iohexol was injected. Intraoperative DSA fluoroscopy was used to observe fluid distribution within the cyst. For microcystic LMs, the mixture of bleomycin, saline, and iohexol was directly injected. After injection, the needle hole was compressed with sterile cotton or gauze for 2–3 min to prevent fluid leakage. The liquid was dispersed evenly within the capsule by gently kneading the lesion. An elastic bandage or gauze was applied to promote sac adhesion closure and prevent bleeding. Treatments were conducted at 1-month intervals.

Sirolimus therapy was initiated at a dose of 0.8 mg/m² every 12 hours. Patients were evaluated biweekly to measure sirolimus serum levels, and doses were adjusted until two consecutive readings within the target range of 10–15 ng/mL were achieved. Once stable, patients underwent monthly outpatient monitoring to assess treatment complications, changes in malformation signs and symptoms, and potential drug side effects. Hematologic parameters, liver and renal function, and coagulation profiles were also monitored monthly.

Follow-up

Postintervention clinical responses were assessed through physical examinations using established sclerotherapy response criteria.^{7 8} Clinical and imaging follow-ups were conducted 6 months after the initial treatment to evaluate changes in lesion size and treatment outcomes. Imaging follow-ups were performed using available MRI or ultrasound data. An excellent response was defined as complete tumor resolution or residual induration only, with a 76–100% reduction in volume on imaging, and no skin elevation, no capsule fluid detected, no tumors visible on imaging, and no recurrence observed during follow-up. A good response was defined as a 51–75% size reduction, while a fair response was defined as a 26–50% size reduction. A poor response was defined as a 0–25% size reduction, no change, or worsening of the lesion. Clinical success was defined as either an excellent or good response.

Results

During the study period, sclerotherapy for LMs was performed on 456 patients at our institution, with 296 cases (64.91%) involving the head or neck region. Among these, 19 children met the inclusion criteria for this analysis, and 6 of the 19 cases (31.58%) were identified prenatally via ultrasonography. Symptoms at presentation included neck mass, dyspnea, and sudden enlargement due to hemorrhage. Of the 19 patients, 13 (68.42%) had macrocystic LMs, 5 (26.32%) had mixed macrocystic lesions, and only 1 child (5.26%) had microcystic lesions. The mean age at the first sclerotherapy session was 6 months. Severe airway compromise necessitated intubation in 5 patients (26.32%), and intracapsular hemorrhage before the first treatment occurred in 4 patients (21.05%).The important clinical features were collated and are summarized in table 1.

Table 1. Patient and lesion characteristics.

Case	Type	Type	Preoperative size (cm3)	Sclerotherapy sessions (n)	Sirolimus (months)	Size at 6 months (cm3)	Complications	Follow-up (months)
1	Macrocystic	V	346.43	6	12	6.45	–	89
2	Macrocystic	III	33.08	3	10	Cure	–	98
3	Macrocystic	III	218.28	1	6	Cure	Low-grade fever	94
4	Macrocystic	V	48.96	3	3	1.95	–	93
5	Macrocystic	V	94.56	1	6	8.24	–	92
6	Mixed	III	32.34	3	9	18.25	Swelling	92
7	Macrocystic	III	32.81	1	3	Cure	–	92
8	Mixed	II	21.89	1	3	7.25	Low-grade fever	88
9	Macrocystic	III	76.01	1	3	Cure	–	78
10	Macrocystic	V	39.80	5	12	2.52	–	71
11	Mixed	III	51.43	4	9	5.23	–	62
12	Microcystic	V	426.77	4	18	140.7	Low-grade fever	66
13	Macrocystic	III	80.55	4	6	3.54	–	55
14	Macrocystic	IV	120.99	2	5	28.86	–	43
15	Macrocystic	III	35.30	4	6	Cure	–	30
16	Macrocystic	V	267.62	6	12	6.44	Swelling	26
17	Mixed	III	102.01	2	6	10.29	Low-grade fever	25
18	Macrocystic	III	88.87	2	6	Cure	–	20
19	Mixed	IV	101.55	3	9	Cure	–	21

The classification of neck LMs according to the de Serres system⁹ revealed that 10 patients (52.63%) were stage III, 6 (31.58%) were stage V, 2 were stage IV, and 1 was stage II. There were no stage I lesions. Five patients were classified as having severe disease based on the Cologne Disease Score (CDS: 0–4 points).¹⁰ A total of 56 sclerotherapy treatment sessions were performed across the 19 children, with an average of 2.95 sessions per child (range: 1–6). Among the 19 cases, 5 were cured with a single injection. Patient and lesion characteristics are summarized in table 1. The bleomycin dose ranged from 0.5 to 1 mg/kg (1 unit of bleomycin=1 mg), and the maximum ethanol dose per session was between 5 mL and 25 mL.Eighteen patients required intubation for 6–48 hours after treatment. One patient experienced failed extubation attempts after the first session due to postoperative rebleeding within the lesion. Extubation was successful 3 days after completing the second treatment, 1 month later. Ultimately, all patients were extubated successfully, with no emergency or preventive tracheotomy required.

All patients received sirolimus therapy for 3–18 months, though three patients temporarily discontinued the medication for approximately 10 days due to pneumonia. Posttreatment imaging revealed no residual lesions or only a few sclerotic remnants in seven patients (figure 1), with no recurrence during follow-up. Eleven patients exhibited >75% reduction in lesion size, achieving excellent outcomes. One patient showed a 51–75% reduction, classified as a good outcome. No fair or poor outcomes were observed. The clinical success rate (excellent or good outcomes) was 100%. Follow-up periods ranged from 20 to 98 months, with a median follow-up time of 71 months. Mild-to-moderate postoperative fever occurred in four patients, while two patients experienced localized swelling, likely due to minor bleeding when the pressure bandage was removed. There were no serious adverse events, such as allergic reactions, pulmonary fibrosis, nerve injury, or skin necrosis.

Figure 1. Case 16. (A,B) T2 fat-saturated magnetic resonance imaging (MRI) showing left complex cervicofacial macrocystic lymphatic malformation surrounding the trachea. (C) Injection under digital subtraction angiography guidance. (D) MRI showing only a few residual sclerotic lesions 6 months after treatment.

Discussion

LMs are congenital benign anomalies caused by structural abnormalities of the lymphatic system, primarily affecting infants. With advancements in prenatal screening and improvements in ultrasonography, an increasing number of LMs are now diagnosed during the neonatal and even fetal periods.¹¹ Large cystic lymphangiomas are commonly observed in neonates. Their pathological features include round, oval, or lobulated sacs with smooth, thin, transparent linings and poor vascular distribution. Clinically, these lesions appear soft, with palpable fluctuation and indistinct margins. After birth, they may enlarge progressively, leading to complications, such as tracheal compression and respiratory distress, requiring urgent intervention. Approximately 75% of LMs are localized to the head and neck region, correlating with the rich lymphatic system in this area. Notably, 90% of LMs present before the age of 2 years.¹²The de Serres classification system is widely used to stage cervicofacial LMs based on their anatomical location, ranging from stage I to stage V.⁹ Patients with bilateral disease, involving both suprahyoid and subhyoid regions, generally have poorer outcomes. Conversely, unilateral subhyoid macrocystic LMs are more amenable to treatment. Previous studies have demonstrated that patients with microcystic or mixed malformations in stage V have the lowest CDS levels.¹³ This aligns with our observations, where the median CDS for stage V patients was 4 points. However, the CDS system, being based on morbidity items, is more subjective and lacks quantitative parameters compared with the de Serres classification system.

Treatment options for LMs include observation, surgery, sclerotherapy, laser therapy, radiofrequency ablation, and pharmacotherapy.¹⁴ Notably, 15–70% of LMs present with mild symptoms,¹⁵ do not cause severe dysfunction, and may only require outpatient monitoring.¹⁶ However, there is no standardized treatment protocol for complex LMs of the head, neck, and maxillofacial regions, particularly for cases characterized by large sizes, cross-regional growth, and high treatment complexity. While simple aspiration may alleviate symptoms, it carries a high risk of recurrence, bleeding, and disease exacerbation. Surgical excision was historically the primary treatment for LMs. However, in neonates, surgical risks are heightened due to poor tolerance, the intricate anatomical relationships of the cyst, and the delicate cyst wall, which increases the likelihood of vascular or nerve injury and residual cyst wall recurrence. Additionally, postoperative complications, such as infection and scarring, are common, often requiring drainage tubes that further complicate care.^{5 17}Currently, surgery is considered adjunctive, with sclerotherapy emerging as a widely accepted, minimally invasive treatment for LMs. Commonly used sclerosing agents include bleomycin, lauromacrogol, doxycycline, ethanol, and their combinations.18,22 A previous study demonstrated that bleomycin combined with other agents yields superior outcomes compared with monotherapy.¹⁸

Anhydrous ethanol, known for its aggressive sclerosing properties, induces rapid cellular dehydration and protein denaturation, leading to epithelial cell necrosis. It is frequently employed in treating vascular malformations. Standard ethanol doses range from 0.3 to 0.5 mL/kg, with a maximum limit of 1 mL/kg. In our study, alcohol consumption exceeded this limit due to the small body weight of infants. Most of the ethanol was aspirated 5 min after injection, minimizing systemic toxicity – a precaution supported by prior studies.¹⁹ Similarly, bleomycin was diluted with contrast media and injected evenly into the cavity rinsed with ethanol to mitigate risks.

Extravasation is a critical risk associated with ethanol use. If ethanol infiltrates the arterial supply of nerves, muscles, or connective tissues it can cause degeneration and necrosis.²⁰ To minimize this risk, we employed a two-needle technique. Bleomycin, an antitumor agent, acts by inducing a nonspecific inflammatory response, promoting fibroblast proliferation, and inhibiting lymphatic endothelial cell growth, resulting in capsule fibrosis and shrinkage. Pulmonary toxicity, although rare, remains a dose-related concern.¹⁸ In cases involving large cystic cavities, pressure dressing posttreatment increases the lesion’s contact area with the sclerosants, thereby reducing the likelihood of postoperative bleeding.

Mechanistic target of rapamycin inhibitors have shown promising results when used off-label for patients with vascular anomalies. Children with LMs involving the airway benefit significantly from sirolimus therapy initiated soon after birth.^{23 24} Sirolimus is an important option for refractory cases. Given the complexity of the LMs in our cohort, sirolimus was added to the treatment protocol. The combined use of sirolimus and sclerotherapy reduced the required dose of bleomycin, minimizing its toxic side effects and decreasing the number of sclerotherapy sessions. Remarkably, 8 of 19 children required only one or two interventional sclerotherapy sessions to achieve resolution.

The most common adverse events associated with sirolimus include Pneumocystis carinii pneumonia, myelosuppression (anemia, leukopenia, and thrombocytopenia), hyperlipidemia, new-onset diabetes, metabolic syndrome, infertility, and increased susceptibility to infections.²³ However, most published studies, along with our own experience, have demonstrated a low incidence of side effects with sirolimus. During the treatment period, only one patient experienced a brief treatment interruption (around 10 days) due to pneumonia. None of the children received prophylactic treatment for pneumocystosis. Furthermore, none of the patients required dose adjustments or treatment discontinuation due to adverse effects.Similar to sirolimus, bleomycin carries risks of severe complications, particularly pulmonary fibrosis, which correlates with the cumulative drug dose. Airway obstruction was a critical preoperative concern, with 41% of the children in our series experiencing severe upper respiratory and feeding difficulties that began almost immediately after birth. Previous studies have suggested that tracheotomy should be considered if the lesion encircles more than half the airway and causes acute swelling.^{4 25} However, tracheotomy can negatively impact subsequent airway development in children. To address these challenges, we administered prophylactic corticosteroids. Fortunately, all children in our cohort were successfully extubated postoperatively, with a median extubation time of 72 hours. This approach avoided the risk of permanent nerve damage associated with tracheotomy and surgical intervention, as well as complications such as tracheal chondromalacia.

For large LMs, intralesional bleeding is a hallmark of disease progression, often leading to rapid swelling and respiratory symptoms. In our cohort, 14% of children experienced intralesional bleeding prior to treatment, all of whom had large cystic LMs. During follow-up, one child with a large cystic LM experienced recurrence after trauma, and 12 mL of hemorrhagic lymph fluid was drained during re-sclerotherapy. Large cystic LMs are generally more prone to bleeding than microcystic or mixed LMs,^{21 22} likely due to increased susceptibility to external trauma or differences in gene expression.^{22 24 26}The postoperative swelling of another mixed type (Case 6) and a macrocystic type may also be related to postoperative bleeding, and the prognosis of the mixed type is a good outcome. The primary strength of this study was its demonstration that sclerotherapy combined with sirolimus can avoid surgical intervention, reducing the risk of scarring and delivering excellent esthetic outcomes. Importantly, all children were able to discontinue continuous airway support during an average follow-up period of over 5 years.

This study had some limitations. First, the small cohort size and absence of a comparison group precluded definitive conclusions. The efficacy of sclerotherapy alone versus combined therapy should be explored in future studies. Second, histopathological evidence was unavailable due to the lack of surgical resection in this cohort.

Conclusions

In summary, for infants with complex cervicofacial LMs, who have poor tolerance for surgery, low-risk and minimally invasive treatments should be prioritized. This study highlighted the safety and efficacy of sclerotherapy combined with sirolimus in treating these challenging cases. The approach minimized injury, was repeatable, and significantly reduced perioperative risks. We recommend that sclerotherapy combined with sirolimus be considered the first-line treatment for complex cervicofacial LMs in infants and young children, with surgery reserved as an alternative or for long-term restorative needs.

Data availability statement

Data are available upon reasonable request.