International Journal of Pediatric Otorhinolaryngology 76 (2012) 569–573

Sodium tetradecyl sulphate sclerotherapy for treating venous malformations ofthe oral and pharyngeal regions in children

Paul Stimpson *, Richard Hewitt, Alex Barnacle, Derek J. Roebuck, Ben Hartley

Great Ormond Street Hospital for Children, Great Ormond Street, London WC1N 3JH, UK

A R T I C L E I N F O

Article history:

Received 26 October 2011

Received in revised form 15 January 2012

Accepted 18 January 2012

Available online 14 February 2012

Keywords:

Venous malformations

Sclerotherapy

Vascular malformations

A B S T R A C T

Objective: Venous vascular malformations in the head and neck region present a difficult management

challenge. We describe our experience of using sodium tetradecyl sulphate injection sclerotherapy to

treat children presenting with venous malformations of the oral and pharyngeal region.

Methods: We performed a retrospective case note review of consecutive children treated at our

institution between 2004 and 2011. Patient notes were analysed for demographic details, site and size of

lesion, number and duration of treatments, treatment response and complications.

Results: Twelve patients were included (7 boys and 5 girls, mean age 7 years). Sites of lesions included

tongue, floor of mouth, pharynx, tonsillar fossae, parapharyngeal space and soft palate. All patients were

treated with 3% sodium tetradecyl sulphate (STS) foam injected trans-orally or percutaneously under

ultrasound or fluoroscopic guidance. The lesions had a mean volume of 4 ml (range 2–14 ml). An average

of 3 treatments was required (range 1–9). In 4 patients a single treatment was sufficient. For those

patients requiring multiple treatments, a mean of 4 treatments were required over an average period of

28 months. The overall response rate was 83% (10/12). Complete resolution was achieved in 4 cases (33%)

with a significant reduction in size in a further 6 cases (50%). Larger lesions generally require more

treatments than low volume lesions. Two cases recurred despite treatment. One patient suffered minor

bleeding following transcutaneous injection.

Conclusions: Injection sclerotherapy using STS foam offers an effective treatment option when managing

children presenting with venous malformations in the oral and pharyngeal. A single treatment may be

adequate for small lesions but the procedure may be safely repeated until a satisfactory result is obtained.

� 2012 Elsevier Ireland Ltd. All rights reserved.

Contents lists available at SciVerse ScienceDirect

International Journal of Pediatric Otorhinolaryngology

jo ur n al ho m ep ag e: ww w.els evier . c om / lo cat e/ i jp o r l

1. Introduction

Venous malformations (VMs) are the most common type ofvascular malformation. They are benign lesions consisting of lowflow ectatic endothelial lined venous channels. Familial andsyndromic associations have been described, but the majority ofcases are sporadic with an overall incidence of approximately 1/10 000 [1–3]. Superficial VMs classically present at birth or in earlychildhood as soft, compressible blue or purple masses that increasein size with dependency or the Valsalva manoeuvre [4,5]. Deeperlesions are often more complex and may present with functionaleffects and/or obstructive symptoms. VMs hypertrophy as thechild grows and spontaneous involution is extremely rare if theyare left untreated [1].

In the head and neck region VMs may be particularlytroublesome due to a propensity to involve deeper structuressuch as the parapharyngeal space, muscles, orbit and bone.

* Corresponding author. Tel.: +44 02074059200.

E-mail address: paul.stimpson@doctors.org.uk (P. Stimpson).

0165-5876/$ – see front matter � 2012 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.ijporl.2012.01.019

Functional effects of VMs in the head and neck may includeswallowing difficulties, obstructive sleep apnoea and airwaycompromise [4]. Episodic focal thrombosis within a VM mayresult in acute pain and swelling and resultant permanentphleboliths within lesions are common [1].

Formal resection of VMs presents an extremely difficult andpotentially dangerous surgical challenge. Injection sclerotherapy(IS) has gained in popularity as a non-surgical management option.Injections may be performed transcutaneously or via endoscopictechniques.

We describe our experience of IS utilising both endoscopic andpercutaneous approaches to treat children presenting with VMs ofthe oral and pharyngeal regions as an alternative to surgicalresection.

2. Methods

We performed a retrospective review of consecutive childrenwith VMs according to the classification system first described byMulliken and Glowaki [6]. All patients were treated at ourinstitution between 2004 and 2011. Following ethical committee

Picture 1. VM right tonsillar fossa pre treatment (patient 9).Picture 3. ‘Significant response’ 11 months post-treatment.

P. Stimpson et al. / International Journal of Pediatric Otorhinolaryngology 76 (2012) 569–573570

approval patient notes were analysed for demographic details, siteand size of lesion, number and duration of treatments, responseand complications.

Following appropriate multidisciplinary team discussion allchildren underwent IS using 3% sodium tetradecyl sulphate (STS, asfoam) under general anaesthetic. The STS was mixed with air andiohexol (240 mg/ml iodine, Omnipaque, GE Healthcare Canada Inc.,Mississauga, ON) in a 2:1:1 ratio, and agitated by repeated transferbetween two syringes, to produce a foam. The dose of STS used wasat the discretion of the operator, but did not exceed 7 ml (i.e. 14 mlof foam) for any one procedure.

Percutaneous injections were performed under ultrasound andfluoroscopic guidance by the interventional radiology team (DRand AB). For children requiring endoscopic injections, access wasachieved by the surgical team (BH) using a Boyle–Davistonsillectomy gag or via suspension laryngoscopy. Intralesionalinjection was then performed under direct vision using a 08Hopkins endoscope in combination with ultrasound wherenecessary (Pictures 1–3). Following aspiration of venous bloodto ensure correct positioning, STS foam was injected via a suitablelength and gauge needle (depending on site) with further needlesplaced within the lesion to allow venous blood displacement(acting as a low pressure exit valve) as intralesional foam volumeincreased [7]. The use of image guidance allows a more accurateassessment of foam placement and prevents spill-over into thesystemic venous circulation. Injection of sclerosant is stoppedwhen the lesion appears full. The amount of sclerosant injected atthe first treatment was recorded as a means of estimating lesion

Picture 2. Sclerotherapy needles in situ.

volume. Following injection, any surface bleeding was controlledwith pressure and topically applied adrenaline (1:10 000 solutionin normal saline) on a neuropatty. Patients were admitted to ashort stay ward for overnight airway observations due to the risk ofswelling following injection.

Treatment response was assessed based on clinical descriptionsof the effect of sclerotherapy at the end of the treatment course. Weused a modification of a previous scoring system described byAucher et al. [8]. ‘Resolution’ was used when no clinical evidence ofthe VM remained. A ‘significant response’ was recorded when asubstantial reduction in lesion size (>50% reduction or better) and/or a return to normal function was observed. ‘No response’ wasused to describe recurring lesions or those with no appreciablereduction in lesion size or functional benefit.

3. Results (Table 1)

Twelve patients were included (7 boys and 5 girls, mean age 7years). Sites of lesions included tongue, floor of mouth, pharynx,tonsil fossae, parapharyngeal space and soft palate. The sclerosantused in all cases was 3% STS foam injected trans-orally orpercutaneously under ultrasound or fluoroscopic guidance.Endoscopic injection was performed alone in 6 cases. Transcuta-neous injection only was performed in 3 cases and in the remaining3 cases both transcutaneous and endoscopic approaches wereused.

The average volume of sclerosant used was 4 ml (range 2–14 ml). An overall average of 3 treatments was required (range 1–9). In 4 patients a single treatment was sufficient. For thosepatients requiring more than one injection, a mean of 4 treatmentswere required over an average period of 28 months. Larger volumelesions appear to require more treatments than low volumelesions. The overall response rate was 83% (10/12). Resolution wasachieved in 4 cases (33%) with a significant response in a further 6cases (50%). An example of a significant response is shown inPictures 1–3 (patient 9). Two cases recurred despite treatment(17%). One patient suffered minor bleeding following transcuta-neous injection of sclerosant. There were no delayed dischargesfrom hospital and all children went home the following day.

4. Discussion

A number of management options exist for treating VMs.Treatment selection will depend upon the site and size of thelesion, patient factors such as co-morbid conditions and availabili-ty of particular techniques and materials [4]. There is no ‘one sizefits all’ solution to managing VMs and patients may require a

P. Stimpson et al. / International Journal of Pediatric Otorhinolaryngology 76 (2012) 569–573 571

number of treatment modalities [9]. An experienced multidisci-plinary team approach to managing VMs (and other vascularanomalies) in an appropriate setting is therefore essential [3,10].

Surgical excision was originally the treatment of choice andmay still offer the best curative option where possible. Over recentyears there has been a move toward less invasive therapies. This islargely owing to the fact that surgery is often difficult when there isdiffuse involvement of critical structures and may be dangerousdue to severe haemorrhage [11,12].

Non-operative treatment options include observation, com-pression and head elevation. These conservative approaches maybe suitable for very small lesions or those with no functional orcosmetic compromise but may also benefit those with largerlesions for example during sleep [3]. Modification of medicationsmay also be helpful. For example, VMs may expand duringadolescence and due to the angiogenic effects of oestrogen therapy,progesterone—only oral contraceptive agents may be preferred infemales with problematic lesions [13]. Prophylactic aspirin mayalso be indicated for recurring painful episodes of thrombophlebi-tis within a VM [12].

LASER therapy has been shown to be effective for superficialcutaneous and mucosal VMs. A number of authors have describedtheir experience of using the Nd:YAG LASER for surface lesions butalso using interstitial probes to treat submucosal disease up to adepth of approximately 8 mm [2,3]. Unfortunately LASER treat-ment has limited value for treating more complex lesions such asthose involving deeper neck structures and should be used withextreme care close to neurovascular structures. Serial LASERtreatments may be useful for debulking larger lesions, renderingthem more amenable to surgical resection.

IS has been utilised as a treatment option for small to mediumsized VMs over recent years. In common with LASER treatment, ISmay also be used as an adjunct to surgical resection for large VMs[11,12,14]. A number of sclerosant agents have been describedincluding ethanol, ethanolamine oleate, sodium tetradecyl sul-phate, sodium morrhuate, bleomycin and hypertonic salineamongst others. Ethanol is a very effective sclerosant. Althoughsome practitioners use ethanol for VMs in the head and neck,others avoid it because of concerns about potential nerve damageor thromboembolism [1,12,15–17]. Complications of ethanolsclerotherapy include cardiovascular collapse and death, althoughthese are rare. It has also been shown that toxicity is increased witheach subsequent treatment and therefore treatment should onlybe contemplated by experienced practitioners with appropriateintensive care support [1,18]. We have preferred to use STS forhead and neck VMs, although there is no good evidence to supportthis decision.

A recent large series by Zheng et al. describes the use ofpingyangmycin (a derivative of bleomycin with a differentterminal amine moiety) for IS in 179 patients (adults and children)with VMs in the oral and maxillofacial region [11]. They found thatin 154 cases ‘cure’ or ‘marked improvement’ was achieved usingsimilar clinical outcome measures as described in our study. Thisgives a comparable response rate of 85%. A poor response wasdescribed in lesions greater than 5 cm for which surgery or LASERtherapy was preferred (for debulking). A concern with pingyang-mycin (and bleomycin) is the association between cumulativedoses of the agent and pulmonary fibrosis. This has been describedin cancer patients at higher doses than those used for sclerotherapybut as a result these agents have not gained widespreadacceptance. Clearly each agent will have its own side effect profileand this must be carefully considered before contemplating anytreatment.

STS is a detergent based sclerosant that acts on endothelial lipidmolecules causing surface damage and collagen exposure. Thiscreates an inflammatory response with resultant fibrosis, scarring

and shrinkage of the VM with minimal thrombus formation [2,13].It has been used extensively since the 1940s for the treatment ofother low flow vascular lesions such as varicose veins andoesophageal varices. STS has a well documented safety profileand is generally preferred over ethanol as an initial option forsclerotherapy [19]. The injection of STS as foam is thought to bemore beneficial by preventing washout and layering whilstproviding optimal sclerosant contact with endothelial surfaceswithin the VM over a period of up to several weeks [2–4].

We have demonstrated that IS using STS foam is an effectivetechnique for the management of VMs in the oropharynx inchildren. An overall response rate of 83% is encouraging and iscomparable to reported outcomes from other studies in adultpatients. A return to normal function and satisfactory cosmeticoutcome is the ultimate goal of treatment and this was achieved inthe majority of cases.

In our study a total of 37 injections were performed with anaverage of 3 treatments required per patient. In four patients asingle treatment was adequate with complete resolution in 3 ofthese cases. In a further case a significant response was achievedfollowing a single treatment and additional sclerotherapy was notrequired. In two patients treatment response was minimal and thelesion quickly recurred. One patient did not want to try furthertreatment and in the other patient the VM was difficult to controland continued to grow despite multiple attempts.

The age range of children in our series ranged from 4 to 12 years(mean 7 years) at initial presentation. We would advocate earlytreatment in children presenting with VMs as the natural historywould suggest that lesions become more difficult to treat as thechild grows [3].

We evaluated response based on clinical criteria and accuratedocumentation of outcome in the case notes. Clearly this may biasour results as treatment effect is subjective and may be over-estimated in a retrospective study such as this. However, theprincipal indication for treatment is cosmesis and loss of functionso the use of a subjective clinical outcome tool seems reasonable.Previous authors have used family questionnaires to assessoutcomes from a range of interventions but again this may besubject to recall bias in a retrospective study [9]. An objective,validated means of measuring treatment response for these lesionswould be helpful for research purposes but is currently lacking.Magnetic resonance imaging may be useful but may subjectyounger patients to general anaesthesia or sedation. In addition, itis not clear that response in terms of change in imagingappearances is a clinically useful parameter.

Sclerotherapy is performed under image guidance and infiltra-tion is immediately stopped if there is evidence of overspill fromthe VM. We measured the quantity of sclerosant injectable at thefirst treatment as a proxy measure of lesion volume. Using thismethod, average lesion volume was 4 ml (range 2–14 ml).Although potentially less accurate than formal three-dimensionalvolumetric analysis of the entire lesion, the volume of sclerosantinjected gives a useful estimate of lesion size. Traditional methodssuch as palpation and direct visual estimation of size may bedifficult due to the depth and complexity of the lesions. From ourresults it appears that smaller lesions required fewer injectionsthan larger VMs (Table 1 and Fig. 1). This would be in keeping withanecdotal reports but a larger study is required to confirm thisobservation. Inevitably there will be exceptions to this general ruleand in our series one patient required 9 treatments for a lesion ofonly 2.5 ml volume in the tongue (patient 1). A significant responsewas eventually achieved. This example highlights the suitability ofusing STS foam for repeated treatments. Although this may resultin a prolonged treatment strategy (60 months for patient 1), allinjections were well tolerated with no complications in this case.The surgical alternative for this particular lesion would be very

Table 1Results.

Patient Age Sex Site of lesion Volume

(ml)

Number of

treatments

Length of treatment

(months)

Response Complications

1 12 M Tongue 2.5 9 60 Significant Nil

2 10 F Floor of mouth/parapharyngeal space 3 2 Single dose to 2 sites Resolution Nil

3 2 M Tongue 2 1 Single dose Reduced, then recurred Nil

4 13 M Naso/Parapharyngeal 5 7 48 Partial, difficult to control.

Recurred

Minor bleeding from

needle puncture

5 9 M Parapharyngeal 5 1 Single dose Resolution Nil

6 7 M Parapharyngeal 2 2 8 Significant Nil

7 8 F Left submandibular/jugular area 5 2 5 Significant response but

refused further Rx

Nil

8 4 F Submental 3 1 Single dose Resolution Nil

9 4 M Tonsil fossa 2 2 11 Significant Nil

10 6 F Soft palate 2 1 Single dose Significant Nil

11 6 M Floor of mouth, tongue, lips, soft palate 14 6 20 Significant Nil

12 6 F Tonsil/tongue base 3.5 3 20 Resolution Nil

Volume of lesion versus number of treatments

0

2

4

6

8

10

1614121086420

Volume (mls)

Nu

mb

er

of

treatm

en

ts

Fig. 1. Relationship between lesion size and number of treatments required

(including non-responders).

P. Stimpson et al. / International Journal of Pediatric Otorhinolaryngology 76 (2012) 569–573572

difficult without compromising function and risking uncontrolla-ble bleeding. It is not clear why some lesions respond to singletreatments where others require multiple injections over manymonths or even years. VMs are diagnosed on clinical grounds andthere may be subtle pathological and structural differencesbetween lesions that may render them more or less sensitive toa particular IS agent. Without tissue analysis this inference ispurely speculative but may be an avenue for further research in thefuture.

Unfortunately we do not have long term follow up data for ourpatient group. VMs are rare lesions and patients are sometimesreferred from long distances for treatment. Routine follow up istherefore difficult and local specialists will assess and managepatients following their course of treatment at our institution.However, of the 10 children for whom treatment was successful,none have been re-referred for further therapy. A prospectivestudy with comprehensive outcome information is required and isthe subject of ongoing study in our department.

Risks of using STS are similar to those of other sclerosantsand include pain, swelling and local ulceration. Rarer complica-tions include extravasation, anaphylaxis and cardiovascularcollapse [3,4]. In our series one patient suffered bleeding duringtranscutaneous injection. This was due to puncture of anoverlying vessel (superficial temporal artery) and not fromthe VM itself. Haemostasis was achieved easily with pressurebut clearly bleeding complications on mucosal surfaces may bemore difficult to control. No other significant complicationswere reported. One patient (patient 11) unfortunately died dueto massive blood loss from residual VM during a separate

procedure after the study period. This highlights the fact thatthese lesions, although benign, can be very dangerous iftraumatised and lead to massive blood loss. It is for this reasonthat surgery is hazardous and alternative strategies must besought and properly evaluated.

5. Conclusions

Injection sclerotherapy using STS foam offers a safe andeffective treatment option when managing children presentingwith head and neck venous malformations. It is cheap, accessibleand avoids the risks of complex surgical intervention. A singletreatment may be adequate but the procedure may be safelyrepeated until a satisfactory result is obtained. Further study isrequired to evaluate long-term outcomes and predictive factors forfavourable response.

References

[1] D.J. Choi, A.I. Alomari, G. Chaudry, D.B. Orbach, Neurointerventional managementof low-flow vascular malformations of the head and neck, Neuroimaging Clin. N.Am. 19 (2009) 199–218.

[2] K. Scherer, M. Waner, Nd:YAG lasers (1,064 nm) in the treatment of venousmalformations of the face and neck: challenges and benefits, Lasers Med. Sci.22 (2007) 119–126.

[3] L.M. Buckmiller, G.T. Richter, J.Y. Suen, Diagnosis and management of haeman-giomas and vascular malformations of the head and neck, Oral Dis. 16 (2010) 405–418.

[4] K.B. Puttgen, M. Pearl, A. Tekes, S.E. Mitchell, Update on paediatric extracranialvascular anomalies of the head and neck, Childs Nerv. Syst. 26 (2010) 1417–1433.

[5] R. Glade, K. Vinson, G. Richter, J.Y. Suen, L.M. Buckmiller, Endoscopic managementof airway venous malformations with Nd:YAG Laser, Ann. Otol. Rhinol. Laryngol.119 (5) (2010) 289–293.

[6] J.B. Mulliken, J. Glowaki, Classification of paediatric vascular lesions, Plast.Reconstr. Surg. 70 (1982) 120–121.

[7] S. Puig, H. Aref, F. Brunelle, Double-needle sclerotherapy of lymphangiomas andvenous angiomas in children: a simple technique to prevent complications, Am. J.Roentgenol. 180 (May (5)) (2003) 1399–1401.

[8] B.M. Aucher, C.J. Chang, V.M. Vander Kam, Management of haemangioma ofinfancy (review of 245 patients), Plast. Reconstr. Surg. 99 (5) (1997) 1301–1308.

[9] R.S. Glade, G.T. Richter, C.A. James, J.Y. Suen, L.M. Buckmiller, Diagnosis andmanagement of paediatric cervicofacial venous malformations: retrospectivereview from a vascular anomalies centre, Laryngoscope 120 (2) (2010) 229–235.

[10] W.L. Chen, J.T. Ye, L.F. Xu, Z.Q. Huang, D.M. Zhang, A multidisciplinary approach totreating maxillofacial arteriovenous malformations in children, Oral Surg. OralMed. Oral Pathol. Oral Radiol. Endod. 108 (2009) 41–47.

[11] J.W. Zheng, X.J. Yang, Y.A. Wang, Y. He, W.M. Ye, Z.Y. Zhang, Intralesional injectionof pingyangmycin for vascular malformations in oral and maxillofacial regions:an evaluation of 297 consecutive patients, Oral Oncol. 45 (2009) 872–876.

[12] A.K. Greene, A.I. Alomari, Management of venous malformations, Clin. Plast. Surg.38 (2011) 83–93.

[13] G.M. Legiehn, M.K.S. Heran, Venous malformations: classification, development,diagnosis and interventional radiological management, Radiol. Clin. North Am. 46(2008) 545–597.

P. Stimpson et al. / International Journal of Pediatric Otorhinolaryngology 76 (2012) 569–573 573

[14] A.C.B.R. Johann, M.C.F. Aguiar, A.V. do Carmo, R.S. Gomex, W.H. Castro, R.A.Mesquita, Sclerotherapy of benign oral vascular lesions with ethanolamineoleate: an open clinical trial with 30 lesions, Oral Surg. Oral Med. Oral Pathol.Oral Radiol. Endod. 100 (2005) 579–584.

[15] B. Minkow, D. Laufer, D. Gutman, Treatment of oral hemangiomas with localsclerosing agents, Int. J. Oral Surg. 8 (1979) 18–21.

[16] C.H. Lee, S.G. Chen, Direct percutaneous ethanol instillation for treatment ofvenous malformation in the face and neck, Br. J. Plast. Surg. 58 (2005) 1073–1078.

[17] D.C. Pappas Jr., M.S. Persky, A. Berenstein, Evaluation and treatment of head andneck venous vascular malformations, Ear. Nose. Throat J. 77 (1998) 914–916,918–922.

[18] B.B. Lee, Y.S. Do, H.S. Byun, I.W. Choo, D.I. Kim, S.H. Huh, Advanced management ofvenous malformation with ethanol sclerotherapy: mid term results, J. Vasc. Surg.37 (2003) 533–538.

[19] L. Reiner, The activity of anionic surface active compounds in producing vascularobliteration, Proc. Soc. Exp. Biol. Med. 62 (1946) 49–54.