Article Text

Download PDFPDF
Sclerotherapy treatment of orbital lymphatic malformations: a large single-centre experience
  1. Alex M Barnacle1,
  2. Maria Theodorou2,
  3. Sarah J Maling2,
  4. Yassir Abou-Rayyah2
  1. 1Department of Radiology, Great Ormond Street Hospital for Children, London, UK
  2. 2Department of Ophthalmology, Great Ormond Street Hospital for Children, London, UK
  1. Correspondence to Dr Alex M Barnacle, Department of Radiology, Great Ormond Street Hospital for Children, London WC1N 3JH, UK; Alex.Barnacle{at}


Background Percutaneous sclerotherapy is an alternative to surgery for the treatment of orbital lymphatic malformations (LMs). We present a large series of patients undergoing sclerotherapy for macrocystic LMs with detailed visual acuity (VA) outcome data.

Methods Data were collected prospectively in all patients with macrocystic orbital LMs undergoing sclerotherapy. Sclerotherapy was performed under general anaesthesia, instilling sodium tetradecyl sulfate under imaging control. Procedures were repeated at 2-week to 6-week intervals, depending on clinical response. Patients underwent ophthalmological assessment, ultrasound and/or MRI before and after treatment. Primary outcome measure: change in maximal radiological diameter of the LM. Secondary outcome measure: change in VA after treatment.

Results 29 patients underwent 71 procedures (1–8 procedures per patient) over 6.7 years. Mean age=7.31 years. 11 patients (37.9%) had undergone previous treatment, including excision biopsy, drainage and decompression. All patients presented with proptosis and/or pseudoptosis. 23 patients (79.3%) had decreased VA at presentation. Average follow-up was 21.8 months (range 3–75 months). All patients achieved a reduction in maximal lesion diameter of ≥50%, with complete radiological resolution in 51.7% (n=15). VA improved in 18/23 patients (78.2%). Average logMAR before treatment=0.43 (SD ±0.47); average after treatment=0.25 (SD ±0.32); p<0.01. There was one complication (1.4%): one patient required a lateral canthotomy for an intralesional haematoma 2 h after sclerotherapy.

Conclusions Sclerotherapy is a safe and highly effective treatment for orbital LMs with excellent VA outcomes. It should be considered as the first-line treatment for this condition.

  • Orbit
  • Treatment other

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Lymphatic malformations (LMs) or lymphangiomas of the orbit are benign vascular malformations, often diagnosed in childhood. Complications include intralesional haemorrhage, cellulitis, amblyopia and impaired visual acuity (VA). Historically, the management of orbital LMs has involved complex, aggressive surgical intervention with variable results and high recurrence rates.1 ,2

Percutaneous intralesional sclerotherapy is now well established as a first-line treatment for LMs elsewhere in the body using a variety of sclerosing agents including sodium tetradecyl sulfate (STS), ethanol, OK-432 (Picibanil), bleomycin and doxycycline.3–7 Small series of up to 13 patients have reported the use of percutaneous sclerotherapy in the management of orbital LMs with a variety of sclerosing agents and techniques.8–14 Results are promising in terms of reduction in lesion size and improvement in proptosis. Detailed VA outcomes have only been documented in one series.14 We report the largest single-centre experience of sclerotherapy in patients with orbital LMs including detailed VA outcome data.

Materials and methods

A retrospective analysis was undertaken of all patients at our institution undergoing sclerotherapy for macrocystic orbital LMs. IRB/Ethics Committee ruled that approval was not required for this study. All patients had a history of proptosis. A unilateral orbital LM was confirmed with ultrasound (US) and MRI. Only those patients with macrocystic LMs were included in this study. Patients with microcystic orbital LMs were excluded from the study as alternative primary sclerosing regimes were considered for these lesions.

The primary outcome of this study was defined as radiological change in maximal diameter of the LM, assessed with US±MRI before and after treatment. A preoperative MRI was performed in all cases; post-treatment MRIs were performed only when clinically indicated. All US were performed by a single radiologist (AMB). The imaging protocol involved US performed before treatment, between treatments, 6 weeks after final treatment and then at six monthly intervals for 2 years. Radiological response to treatment was defined as a change in maximal lesion diameter: resolved=no evidence of residual malformation in the orbit; good response=>75% reduction in maximal lesion diameter; partial response=50–75% reduction in maximal lesion diameter; poor response=<50% reduction in maximal lesion diameter; and no response=no improvement in maximal lesion diameter.

The secondary outcome was defined as change in VA following treatment. All patients underwent preoperative ophthalmological assessment and were reassessed 1 day after treatment and again at 6 weeks. Visual assessment was conducted by the orthoptic team; all patients had age-appropriate visual testing (Cardiff cards, Kay pictures or crowded logMAR). A paired Student's t test was used to compare VA before and after sclerotherapy treatment in patients in whom VA was measurable.

Sclerotherapy technique

All sclerotherapy procedures were carried out under general anaesthesia by an interventional radiologist (AMB) in a biplane angiography suite. All patients received a single dose of intravenous dexamethasone (0.25 mg/kg) on induction. Antibiotics were only given if active infection was suspected (n=2). The affected orbit was scanned using a high-resolution linear 5–13 MHz US probe. The malformation was accessed percutaneously with a 22 gauge two-part needle through an upper lid medial or lateral approach, depending on the position of the malformation. Needle placement was controlled with real-time US throughout the procedure. Where possible, cyst contents were aspirated. Contrast medium (Omnipaque 240, GE Healthcare, Oslo) was then instilled in small volumes (<1 mL) into the cyst during digital subtraction angiography (DSA) using biplane imaging. This allowed the operator to exclude communication between the cyst and the cavernous sinus, ophthalmic artery or other structures. Sclerosant was only injected if no communication was seen. STS sclerosant was then instilled in small aliquots into the largest cyst(s) under US control. The sclerosant was left to dwell in the cyst and not aspirated. Postoperative orbital bandaging was not used. Patients returned to an inpatient ward with oral morphine prescribed on an ‘as required’ basis and were closely monitored for acute proptosis caused by an intralesional bleed or severe swelling. All patients underwent ophthalmology assessment the following day prior to discharge. Routine follow-up took place at 6 weeks unless the need for a repeat sclerotherapy procedure was predicted due to a large volume of cysts, in which case sclerotherapy was repeated at 2–6 weeks.


Twenty-nine patients were referred for treatment between October 2007 and September 2014. All had macrocystic orbital LMs, with associated microcystic disease of the face in five. None of the patients had microcystic disease affecting the orbit itself, and none of the lesions had a significant venous component. Mean age at the time of treatment was 7.31 years (range 0.3–21.6 years, median 5.8 years). The commonest mode of presentation was pain and/or proptosis (figure 1). Twenty-three patients (79.3%) had decreased VA in the affected eye at presentation, ranging from several logMAR letters of visual loss to no perception of light (NPL). Eleven patients (37.9%) had undergone ophthalmological treatment prior to referral, including excision biopsy, drainage and orbital decompression.

Figure 1

A 13.6-year-old female with a macrocystic lymphatic malformation of the left orbit showing (A) proptosis and lateral deviation of the globe at presentation and (B) significant improvement in the appearances of the left orbit after five sclerotherapy procedures.

Twenty-nine patients underwent 71 procedures (1–8 procedures per patient). Repeat treatment was considered in patients with <75% reduction in lesion size on US and in conjunction with VA findings. STS was used as the sclerosant in all cases, at a strength of 1.5–3%. The average volume of sclerosant used was 1.12 mL (range 0.22–3.8 mL).

The average length of clinical follow-up was 21.8 months (range 3–75 months, median 22 months). The procedural complication rate was 1.4% (1/71 procedures). The first patient in the series (age 1.6 years) experienced an acute retrobulbar haemorrhage 2 h after the second sclerotherapy treatment for an extensive LM. 3.8 mL of STS 3% was instilled during the procedure, and a standard procedural technique was used. The haemorrhage was managed with acetazolamide and an uncomplicated lateral canthotomy. She underwent two subsequent uncomplicated sclerotherapy procedures, with complete resolution of the malformation (follow-up 75 months) and some improvement in VA (no interest in Cardiff cards pre-treatment; logMAR 0.87 post-treatment).

The primary outcome measure was radiological change in size of the lesion US±MRI after treatment. Radiological improvement was assessed in all patients on US±MRI. All patients achieved a reduction in lesion size of ≥50%. Fifteen patients (51.7%) demonstrated complete radiological resolution of their malformation, and 11 patients (37.9%) showed a good response. In many patients in whom the malformation was no longer detected, there was residual heterogeneity of the fat in the posterior orbit, taken to reflect the presence of previous inflammation/disease in the orbit. Examples of radiological appearances before and after treatment are shown in figures 24.

Figure 2

A 1.6-year-old female with a macrocystic lymphatic malformation(LM) of her right orbit. Images from a heavily T2-weighted axial MR sequence showing (A) a large LM expanding the right orbit and causing severe proptosis, and (B) radiological change after five sclerotherapy procedures. There is now minimal proptosis and the high T2 signal cystic malformation has shrunk to a small, intermediate T2 signal lesion lateral to the optic nerve complex.

Figure 3

A 5.8-year-old female with a macrocystic lymphatic malformation of the right orbit. (A) Ultrasound image of the orbit at presentation showing large thin-walled macrocysts (*) along the lateral sidewall of the orbit; (B) after one sclerotherapy procedure, the posterior orbit appears clear on ultrasound.

Figure 4

A 2.9-year-old female with a macrocystic lymphatic malformation of the left orbit. (A) T1-weighted axial MRI image showing a complex cystic lesion filling the anterior and posterior orbit and displacing the globe. The largest cyst has fluid levels within it. (B) T2-weighted axial MRI image showing a marked decrease in the size of the malformation after one sclerotherapy procedure. There is now just a small rounded residual mass at the orbital apex and the proptosis has resolved.

The secondary outcome measure was change in VA after treatment. Twenty-three patients (79.3%) had decreased VA in the affected eye at presentation. Of those 23 patients, VA improved in 18 patients (78.2%), remained stable in 3 (13%), remained unrecordable in 1 and worsened in 1 (logMAR 0.52 pre-treatment, 0.70 post-treatment). Statistical analysis of change in VA could not include two patients that had no interest in Cardiff cards at presentation or the patient who was NPL at presentation and in whom understandably this did not improve. Excluding these three patients, average VA before treatment was 0.43 (SD±0.47) and after treatment was 0.25 (SD±0.32). A paired Student's t test was used to compare VA before and after sclerotherapy treatment, p<0.01. Proptosis subjectively improved in all patients, but an accurate measure of proptosis was not possible in every patient prior to treatment due to the age range of the patients in this group.


LMs are benign vascular malformations that are present at birth and grow commensurately with the child.3 ,15 They are defined within the group of slow-flow malformations in the classification devised by Mulliken and Glowacki and recently updated by the International Society for the Study of Vascular Anomalies.16 They occur most commonly within the head and neck but are rare in the orbit.4 Histologically, LMs consist of cystic dilatation of lymphatic channels lined by vascular endothelium; they contain proteinaceous fluid.6 They are thought to be a result of erroneous embryogenesis.5 Typically there is no communication with the functioning systemic lymphatic system although lesions often fluctuate in size with systemic viral illness. LMs are classified as macrocystic or microcystic lesions, with macrocysts usually defined as cysts >10 mm in diameter.3 ,5 ,7 Subclassification of LMs into microcystic or macrocystic lesions appears to be relevant clinically and prognostically when identifying lesions that may respond to specific forms of therapy such as sclerotherapy.7 Some LMs are asymptomatic and do not require treatment but many are prone to intralesional haemorrhage and infection. In certain locations, malformations can cause significant morbidity due to mass effect, such as in the airway, mesentery or orbit.

LMs of the orbit appear to be more commonly macrocystic than microcystic, although macrocystic orbital LMs may be associated with microcystic disease elsewhere in the face, as seen in 19% of our cohort.2 ,17 A proportion of macrocysts convert to a more solid microcystic morphology after infection, haemorrhage or sclerotherapy. Orbital macrocysts occur in both intraconal and extraconal locations, often surrounding the optic nerve complex.13 Cysts vary in size and often contain altered blood, as demonstrated by varying signal characteristics on MRI (figures 2 and 4). Patients present with acute painful proptosis due to intralesional haemorrhage or with slow, insidious swelling. Once symptomatic, they cause significant ocular comorbidity, such as extraocular muscle dysfunction, infection, amblyopia and decreased VA.2 ,13 ,17 When the LM lies anteriorly within the orbit and is well circumscribed, complete surgical resection may be possible. However, more commonly the LM forms a complex, infiltrative lesion in the posterior orbit, making surgical excision difficult and often incomplete.1 ,2 Reported complication rates are high and recurrence is common.1 ,2

Percutaneous sclerotherapy is widely reported in the management of LMs throughout the body and is increasingly accepted as the first-line, non-surgical treatment for these lesions.3–7 A variety of sclerosing agents are used including STS, ethanol, OK-432 (Picibanil), bleomycin and doxycycline. Orbital sclerotherapy was first described in 1999, and a number of small series have since described this procedure.8–14

STS is an anionic surfactant buffered in benzyl alcohol 2%, which causes endothelial necrosis via direct cytotoxic action.18 The authors chose to use this agent because of familiarity with the drug in treating LMs elsewhere in the body and its relatively high potency compared with alternative agents, meaning that a smaller volume of agent is required to produce effective sclerosis; this is an advantage in the orbit where large volumes of any agent cannot be instilled. This has to be weighed against the disadvantages of STS: significant swelling in the first 1–5 postoperative days and the potential for neurological injury.19 ,20 The degree of swelling appears to correlate with the dose and strength of STS used, and both factors were altered several times in the early part of this series as the operator (AMB) gained experience. A volume of 3.8 mL of STS 3% was used in the second procedure, and this was the only procedure resulting in an acute intralesional haemorrhage postoperatively. Since procedure 19 (January 2012), standard practice has been to use a concentration of 2% STS and volumes of <2 mL. Volumes of as little as 0.4 mL of STS 1.5% have been effective in this series. Of note, STS is recognised to be more effective in macrocysts than microcysts and bleomycin should be considered as a first-line agent for solid microcystic disease.

The procedural technique used in this series is similar to that described in other papers. Lesion puncture is technically challenging, even in experienced hands. All punctures were performed under US guidance, as this allows for precise needle placement and does not involve a radiation burden to the eye. Other papers have described fluoroscopically guided punctures.9 ,10 In this series, contrast medium was instilled into the lesion during biplane DSA prior to instillation of sclerosant. This was done to exclude any communication between the cyst and the cavernous sinus, ophthalmic artery or other structures. Although there has been no evidence of contrast medium outflow in this series, the authors still advocate this technique in view of the potential for devastating complications such as cavernous sinus thrombosis, stroke or visual loss with escape of STS. Hill et al14 describe a staged percutaneous ablation and drainage technique with excellent results. This may be of particular value in patients with large collections of fluid or in lesions that are refractory to standard treatment.

All patients in this series underwent follow-up US imaging; follow-up MRI was performed rarely and only where clinically indicated. This was influenced in part by the age of the patients in this series, necessitating general anaesthesia for most MRI studies. The same operator performed all US studies. Accurate volumetric data were not available from cross-sectional imaging in every patient as a large proportion of scans were performed in referring institutions and source data that would allow accurate volume measurements was unavailable in some cases. Hence, the primary outcome measure of radiological change in size of the lesion was defined as change in maximal diameter of the lesion: resolved=no evidence of residual malformation in the orbit; good response=>75% reduction in maximal lesion diameter; partial response=50–75% reduction in maximal lesion diameter; poor response=<50% reduction in maximal lesion diameter; and no response=no improvement in maximal lesion diameter. The authors recognise this as a limitation of the study, as is the fact that both preoperative and postoperative imaging was performed by the operator with potential for bias. Allowing for this, the radiological outcomes are striking, with 51.7% of patients having no evidence of residual malformation on US±MRI after treatment.

Only one other paper has clearly documented VA data before and after sclerotherapy.14 Hill et al14 report that all patients in their series maintained or improved their preoperative VA with an average of one Snellen line improvement, with a statistically significant improvement in the subset of patients with decreased preoperative vision. VA outcomes in our series are comparable. Of the 23 patients with decreased VA, vision improved in 18 (78.2%). Of the remaining patients, one was NPL at presentation and therefore VA could not be improved, three maintained VA throughout treatment and one had a visual drop from 0.5 to 0.7. At the time of writing this paper, he was undergoing occlusion therapy. At the start of this series, the authors did not consider any significant improvement in VA to be likely in a cohort of such complex patients and the results have been both surprising and encouraging.

This large cohort study confirms that US-guided puncture, cystography and use of small aliquots of STS 2% is a reproducible and safe technique in experienced hands and that both radiological and VA outcomes are excellent. Close collaboration between a specialist interventional radiologist and an experienced ophthalmology team is mandatory. Where this is available, sclerotherapy is strongly advocated as the first-line treatment of this complex disease.


View Abstract


  • Contributors AMB designed the study, undertook and monitored data collection, performed the statistical analysis and literature review, and drafted and revised the paper. She is guarantor. MT assisted with data collection and drafted and revised the paper. SJM assisted with data collection, statistical analysis and drafted and revised the paper. YA-R assisted with data collection and drafted and revised the paper.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas