Outcomes and Complications of Sclerotherapy for Venous Malformations

Introduction

Venous malformations (VMs) are common vascular malformations that may be located in any region of the body, with a preference for head and neck. ¹ Not all VMs need to be treated, as most of them are asymptomatic. However, quite a few patients have functional and esthetical problems brought by this congenital disorder. ²

Surgical excision is one of the traditional treatments of VMs; however, symptom recurrence always occurs, as complete excision is seldom performed. Hence, percutaneous sclerotherapy as a less-invasive treatment has gradually taken the place of surgical excision to become the mainstay of treatment for VMs with good mid-term results in the past 2 decades. Many kinds of sclerosants have been reported on treating VMs. Although each sclerosant has a different mechanism of action causing varying amounts of success and complications, they all result in a thrombus that incorporates erythrocytes, organizing, and subsequent contraction and recanalization, with the consequent regression of the lesion. ³ An ideal sclerosant should meet the criteria that it is effective in endothelial cells and has low toxicity in the surrounding tissues. In this study, we reviewed the therapeutic results and safety of sclerotherapy in patients with VMs using different sclerosants. With the help of this systematic review, we aim to determine the efficacy of sclerotherapy for VMs and to assess the relationship of different sclerosants to response and the adverse effect profile.

Methods

PubMed was used to search the medical literatures for publication on the combined topics of “venous malformation and sclerotherapy” using the following key words: “venous malformation,” “low-flow vascular malformation,” “vascular malformation,” “sclerotherapy,” “embolization,” “endovascular,” “percutaneous,” and “NOT arteriovenous.” Finally, the results were combined with the key words “treatment” or “therapy”. Search took place in January 2012.

All the studies should meet the following inclusion criteria: (1) randomized controlled studies, cohort studies, or case series; (2) at least 5 patients with VMs treated with sclerotherapy; (3) sclerotherapy was used as the independent therapeutic agent; and (4) original English studies.

The following exclusion criteria were then applied to the patients in each study to eliminate inappropriate patients and produce a homogeneous population amenable to systematic review: (1) case reports, (2) Klippel-Trenaunay syndrome or other syndromes related to VMs, (3) using more than 1 sclerosant or combined with other management, (4) other vascular malformations such as arteriovenous malformation, lymphatic venous malformation, and so on, (5) patients lost to follow-up, and (6) non-English articles.

Two reviewers (Y.Q. and H.C.) independently screened the titles and abstracts of all the articles. Areas of conflicts between the reviewers were subsequently reviewed by a third party (X.L.) and were resolved through consensus. There was 100% agreement on the final interpretation of the data.

Summary tables were constructed for all included articles. The final evaluation of the venous lesions was based on both objective parameters (the volume of the lesion ascertained before and after treatment by ultrasound, fluoroscopy, digital subtraction angiography, magnetic resonance imaging, and so on) and subjective parameters (symptoms and complaints of the patient before and after treatments). The objective outcome was graded as cured (more than 90% reduction in the volume), improved (25%-90% reduction in the volume), and little change or worse (<25% reduction in the volume) at the end of the treatment. The subjective outcome was graded as cured (disappearance of symptoms), improved (relief of symptoms), and little change or worse at the end of the treatment. Complications were classified as skin damage (skin blistering, skin ulcer, skin necrosis, cutaneous fistula, and pigmentation), nerve damage (temporary numbness or paresthesias, transient nerve palsy, and permanent nerve palsy), muscle damage (muscular fibrosis), renal damage (hemoglobinnuria and renal failure), pulmonary embolism, cardiovascular collapse, and others.

Results

Using the above search strategy, 1053 publications were identified in the literature. Of these, 954 studies were excluded after the titles and abstracts were reviewed, and another 64 studies were excluded after the full text was reviewed, including 2 studies ^4,5 in which ethylcellulose was used to gelify ethanol from liquid into gel. Thirty-five articles satisfied our selection criteria for review that were published between 1986 and 2011 (Figure 1). Demographic data were collected from each article (Table 1).

OPEN IN VIEWER

Figure 1.

Study selection flowchart.

OPEN IN VIEWER

Table 1.

Quantitative Systematic Review.

No.	Study	Year	Country	Intervention	Study Type	Number of patients	Gender (M/F)	Type (diffusing percentage)	Location
									H&N	Limbs	Trunk	Other
1	Spence et al 6	2011	Canada	Absolute ethanol	Retrospective cohort	17	N/A	N/A	17	0	0	0
2	Orlando et al 7	2011	Brazil	Absolute ethanol	Case series	39	10/29	N/A	0	39	0	0
3	Baek et al 8	2011	Korea	Absolute ethanol	Case series	22	10/12	N/A	22	0	0	0
4	Su et al 9	2010	China	Absolute ethanol	Case series	60	26/34	N/A	60	0	0	0
5	Orlando et al 10	2010	Brazil	Absolute ethanol	Case series	81	34/47	N/A	0	69	12	0
6	Ierardi et al 11	2010	Italy	Absolute ethanol	Case series	80	35/45	N/A	0	80	0	0
7	Lee et al 12	2009	Korea	Pure or diluted ethanol	Case series	87	40/47	51.7%	87	0	0	0
8	Lee et al 13	2005	China	Absolute ethanol	Case series	5	5/0	N/A	5	0	0	0
9	Rimon et al 14	2004	Israel	Absolute ethanol	Case series	21	6/15	33.3%	2	14	5	0
10	Lee et al 15	2003	Korea	Absolute ethanol	Case series	87	N/A	100%	N/A
11	Johnson et al 16	2002	USA	Ethanol	Case series	7	4/3	N/A	7	0	0	0
12	Shireman et al 17	1997	USA	Absolute ethanol	Case series	11	5/6	45.4%	0	9	2	0
13	Berthelsen et al 18	1986	Sweden	Absolute ethanol	Case series	5	3/2	N/A	5	0	0	0
14	Hoque et al 19	2011	Bangladesh	5% EO	Case series	84 with 85 lesions	39/44	N/A	52	16	15	2
15	Kaji et al 20	2009	Japan	5% EO	Case series	73	N/A	N/A	73	0	0	0
16	Johann et al 21	2005	Brazil	1.25% or 2.5% EO	Case series	8 with 9 lesions	3/5	N/A	9	0	0	0
17	Kim et al 22	2004	Korea	5% EO	Case series	16	6/10	N/A	16	0	0	0
18	Bonan et al 23	2007	Brazil	2.5% MO	Case series	6	3/3	N/A	6	0	0	0
19	Blaise et al 24	2011	France	0.25%-3% POL	Case series	19	6/13	68.4%	8	9	1	1
20	Yamaki et al 25	2008	Japan	10% EO	RCT	40	9/31	85.0%	17	18	4	1
				1% POL		49	14/35	83.6%	25	16	7	1
21	Mimura et al 26	2003	Japan	1% POL	Case series	14	6/8	N/A	2	11	1	0
22	Cabrera et al 27	2003	Spain	0.25%-4% POL	Case series	35	13/22	45.7%	3	23	3	4
23	Jain et al 28	2002	India	1% POL	Case series	9	N/A	N/A	0	9	0	0
24	Marrocco-Trischitta et al 29	2001	Italy	1%-3% POL	Case series	9	9/0	N/A	0	0	0	9
25	Yamaki et al 30	2000	Japan	3% POL	Case series	28	2/26	N/A	16	8	2	2
26	Krokidis et al 31	2011	Italy	3% STS	Case reports	5	0/5	N/A	0	0	0	5
27	Khandpur et al 32	2010	India	3% STS	Case series	11	5/6	N/A	4	7	0	0
28	Tan et al 33	2007	Canada	3% STS	Case series	65	22/43	66.1%	0	56	9	0
29	Mahesh et al 34	2003	India	3% STS	Case series	24	N/A	N/A	N/A
30	Siniluoto et al 35	1997	Sweden	3% STS	Case series	38	14/24	N/A	38	0	0	0
31	Li et al 36	2010	China	5% SM	Case series	13	4/9	46.1	8	4	1	0
32	Li et al 37	2009	China	5% SM	Case series	23	9/14	0%	8	13	2	0
33	Spence et al 38	2010	Canada	Bleomycin	Case series	31 with 32 lesions	14/17	N/A	31	0	0	0
34	Zhi et al 39	2008	China	Pingyangmycin	Case series	82	26/56	N/A	82	0	0	0
35	Rosbe et al 40	2011	USA	3% Sotradecol	Case series	10	6/4	N/A	10	0	0	0

Abbreviations: H and N, head andneck; M/F, male/female; EO, ethanolamine oleate; MO, monoethanolamine; RCT, randomized controlled trial; POL, polidocanol; STS, sodium tetradecyl sulfate; SM, sodium morrhuate; N/A, not available.

There were 1214 patients involved in this review who received sclerotherapy as an independent cure for VMs. Postoperative patients with recurrent or residual disease of VMs were also included. The most common sclerosant was ethanol, ^6–18 as shown in Figure 2, which was used in 13 studies involving 522 patients, followed by polidocanol ^24–30 (7 studies with 163 patients), ethanolamine oleate ^25,19–22 (5 studies with 221 patients), and sodium tetradecyl sulfate ^31–33 (5 studies with 143 patients). Sodium morrhuate ^36,37 was used in 2 studies from the same clinical institution conducted at different times. Pingyangmycin, ³⁹ bleomycin, ³⁸ sotradecol, ⁴⁰ and monoethanolamine ²³ were used in 1 study each.

OPEN IN VIEWER

Figure 2.

Proportion of patients with each sclerosant.

The outcomes of each sclerosant were summarized in Table 2. Among the 4 mainstream sclerosants, the patients treated with ethanolamine oleate received the highest cure rate of up to 59.2%, followed by polidocanol (41.6%) and ethanol (27.1%), according to the objective measurement of volume by different means. Meanwhile, the total efficiency of these 4 sclerosants, defined as a more than 25% reduction in the volume, all exceeded 90% except one. Sodium tetradecyl sulfate showed the lowest cure rate (12.0%) and the highest ineffective rate (31.0%). According to the subjective expressions of symptoms and complaints by patients before and after treatments, 66.2% of the patients treated with polidocanol were symptom free at the end of the treatment, and of those treated with ethanol and sodium tetradecyl sulfate, 25% and 16.3%, respectively, were symptom free. The symptoms and complaints included pain, functional disorder, and cosmetic issue. None of the 5 studies on ethanolamine oleate listed the improvements in the clinical symptoms after the treatment.

OPEN IN VIEWER

Table 2.

Subjective and Objective Outcomes of Each Sclerosing Agent.

Interventions	Objective Outcome (Volume)				Subjective Outcome (Symptoms)
	Disappearance N (%)	Improvement N (%)	Little change or Failure N (%)	TNL	Disappearance N (%)	Improvement N (%)	Little change or Failure N (%)	TNPS
Ethanol 6–18	127 (27.1)	322 (68.8)	19 (4.0)	468	61 (25.0)	171 (70.1)	12 (4.9)	244
EO 25,19–22	132 (59.2)	74 (33.2)	17 (7.6)	223	N/A	N/A	N/A	N/A
POL 24–30	62 (41.6)	74 (49.7)	13 (8.7)	149	53 (66.2)	23 (28.7)	4 (5)	80
STS 31–35	12 (12.0)	57 (57.0)	31 (31.0)	100	17 (16.3)	61 (58.6)	26 (25.0)	104
PYMS 39	0 (0)	79 (96.3)	3 (3.6%)	82	N/A	N/A	N/A	N/A
SM 36,37	19 (52.7%)	17 (47.2)	0 (0)	36	15 (41.6)	21 (58.3)	0 (0)	36
Bleomycin 38	0 (0)	21 (58.3)	11 (30.5)	32	0 (0%)	29 (90.6)	3 (9.4)	32
Sotradecol 40	N/A	N/A	N/A	N/A	7 (70)	3 (30)	0	10
MO 23	4 (66.7)	2 (33.3)	0	6	N/A	N/A	N/A	N/A
Total	356 (32.5)	646 (58.9)	94 (8.6)	1096	153 (30.2)	308 (60.9)	45 (8.9)	506

Abbreviations: TNL, total number of lesions; TNPS, total number of patients with symptoms; EO, ethanolamine oleate; POL, polidocanol; STS, sodium tetradecyl sulfate; PYMS, pingyangmycin; SM, sodium morrhuate; MO, monoethanolamine; N/A, not available.

Minor and major complications were summarized in Table 3. Skin damage including skin blistering, skin ulcer, skin necrosis, cutaneous fistula, and pigmentation was the most common minor complications with the overall incidence of 10.0%. Among them, the patients treated with ethanol had the highest (15.5%) incidence of skin damage. Other common minor complications after ethanol treatment involved nerve damage (4.8%), muscle fibrosis (1.5%), and transient hemoglobinuria (0.95%). Most cases of nerve damage were reversible. Only 3 cases ^6,15 showed permanent nerve palsy. Two of them were facial nerve palsy and the other was peroneal nerve palsy. One patient ⁶ had transient facial muscle function but was left with “crocodile tears.” Major complications included 1 case of pulmonary embolism and 1 case of cardiovascular collapse. ¹⁴ Pulmonary embolism happened 2 days after sclerotherapy in a patient with a hypercoagulable state. That patient recovered eventually after treatment with no sequelae. Cardiovascular collapse was manifested in the form of acute pulmonary hypertension. After resuscitation and 24-hour stay in the intensive care unit, this patient was discharged home 3 days later.

OPEN IN VIEWER

Table 3.

Minor and Major Complications of Each Sclerosant.

	TNP	Skin Damage N (%)	Nerve damage			Muscle Damage N (%)	Renal Damage N (%)	Pulmonary Embolus N (%)	Cardio-vascular Collapse N (%)	Others
			Paresthesia N (%)	Transient Nerve Palsy N (%)	Permanent Nerve Palsy N (%)
Ethanol 6–18	522	81 (15.5)	10 (1.9)	12 (2.3)	3 (0.57)	8 (1.5)	5 (0.95)	1 (0.19)	1 (0.19)	Crocodile tears(n = 1), 6 reversible ophthalmic venous thrombosis(n = 1)
EO 25,19–22	221	15 (6.7)	–	–	–	5 (2.3)	35 (15.8)	–	–	–
POL 24–30	163	15 (9.2)	2 (1.2)	–	–	–	8 (4.9)	–	1 (0.61)	–
STS 31–35	143	8 (5.6)	7 (4.9)	1 (0.7)	–	1 (0.7)	–	–	–	Allergy(n = 2), 33,35 gangrene (n = 1), 34 blindness (n = 1), 35 infection (n = 1) 35
PYMS 39	82	–	–	–	–	–	–	–	–	Inflammation(n = 3), 39 fever(n = 1) 39
SM 36,37	36	2 (5.5)	–	–	–	–	–	–	–	Mild fever(n = 1) 36
Bleomycin 38	31	1 (3.2)	–	–	–	–	–	–	–	infection(n = 1), 38 nausea(n = 2) 38
Sotradecol 40	10	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A
MO 23	6	–	–	–	–	–	–	–	–	–
Total	1214	122	19	13	3	8	48	3	1
		10.0%	1.5%	0.10%	0.25%	0.66%	3.9%	0.25%	0.08%

Abbreviation: TNP, total number of patients.

Renal damage (15.8%) was the most common complication in the patients treated with ethanolamine oleate, involving 34 cases of hemoglobinuria ^20,25 and 1 case of renal failure. ²⁵ The patient developed acute renal failure with the use of ethanolamine oleate and recovered within 1 week by infusion of K⁺-free fluid and diuretics. Skin damage occurred in 6.7% of the patients. Five patients ^20,22 with lesions on the head and neck experienced trismus. Two ²² of the trismus disappeared after 1 week of conservative management, and the reversibility ²⁰ of the rest was not mentioned in the literature.

Among the patients treated with polidocanol, 15 (9.2%) cases showed skin damage, 8 cases ^25,29,30 showed transient hemoglobinuria, and 2 (1.2%) cases ²⁶ showed numbness of limbs. One patient ²⁶ experienced decrease in blood pressure and bradyarrhythmia during the treatment and was recovered with fluid replacement and intramuscular injection of atropine sulfate.

For those treated with sodium tetradecyl sulfate, skin damage (5.6%) and nerve damage (5.6%) were the most common complications. All the nerve damage caused by sodium tetradecyl sulfate were reversible, including 7 cases ^34,35 of temporary numbness or paresthesias and 1 case ³⁵ of transient slight facial nerve paresis. Allergic reaction was also seen in 2 cases. ^33,35 Other rare complications involved 1 case ³⁴ of gangrene of toes, 1 case ³⁵ of unilateral blindness, 1 case ³⁵ of trismus, and 1 case ³⁵ of infection.

Discussion

Venous malformations (VMs) are common congenital vascular malformations that are believed to be the result of faulty embryonic development. ⁴¹ They can be seen at birth as well as in childhood or adulthood. Most VMs show progressive enlargement and never regress.

Surgery, once the mainstream treatment for VMs, has been gradually replaced by sclerotherapy for its limits of the functional and esthetic sequelae and the high recurrence rate. ⁷ However, it cannot be completely eliminated, as surgery is still the only method to remove the residual fibrosis and phleboliths. ¹⁹

Percutaneous sclerotherapy has been widely used for its unique advantage of being an easy, simple, fast, ambulatory, and low-cost procedure with good tolerability, without the need for hospital admission. ³ A variety of sclerosing agents were described in the literature. Although in our review the patients treated with ethanolamine oleate received the highest cure rate, it is still not possible to make the conclusion that ethanolamine oleate is the most effective sclerosant. Many objective parameters should be taken into account, such as the volume of the lesions and the total number of procedures, for which we failed to extract data in a unified way from the literature, in that they were partially missing or presented in different ways. Symptoms are the other important evaluation criteria of outcome, for they are the main reason on the patients’ attendances. After the sclerotherapy, most of the patients’ symptoms were relieved or disappeared no matter what sclerosant they used.

Among the available sclerosing agents, ethanol seems to be the most powerful according to its side effects and complications. It worked through inducing thrombosis by denaturing blood proteins, denuding the vascular wall of endothelial cells, precipitating their protoplasm, and segmentally fracturing the vascular wall to the level of the internal elastic lamina. ⁴² Pain and swelling were so common that they happened to all of the patients treated with absolute ethanol. When treating the lesions on the specific area, intubation is required in order to avoid potential breathing difficulties ¹⁶ caused by serious swelling. Skin damage, nerve damage, and muscle damage were the main complications. Diluted ethanol was tried to be used in the lesions adjacent to the skin to prevent skin necrosis ¹² in 1 study, which seemed to meet with success. This is promising but more studies are needed. Although most of the nerve and muscle damage were reversible in our review, they still should be avoided as much as possible. Tender sclerosants such as polidocanol could take place of ethanol when treating the lesions near the nerves or muscles. Furthermore, controlling the amount of ethanol on each procedure is very important to the safety of patients. That case of cardiovascular collapse happened probably due to the large amount of ethanol injected (1 mL/kg). ¹⁴

Sodium tetradecyl sulfate seemed to have a potential to trigger allergic reaction. Nerve damage could not be underestimated as well. Apart from 7 cases of temporary numbness or paresthesias and 1 case of transient slight facial nerve paresis, 1 case of monocular blindness was also a result of optic nerve injury probably caused by inflammation and thrombosis of the territory of the malformed ophthalmic vein. ³⁵ Renal damage was the main complication of ethanolamine oleate, recommending that patients with poor renal function must avoid using this sclerosant agent. Polidocanol seemed to be very tender and tolerated, but there was still 1 case of cardiac complication in our review. ²⁶ It was hard to differentiate this reaction from the cause of polidocanol itself or the vasovagal reflux associated with anxiety and pain during puncture. However, the amount of polidocanol should be kept low to prevent this reaction from happening again.

Other sclerosants such as pingyangmycin, sodium morrhuate, bleomycin, sotradecol, and monoethanolamine showed both good efficacy and low complication rates. However, studies on these sclerosant agents were limited and less convincing. More studies on these sclerosants are required.

Several limitations of this review are evident. Most of the studies included were retrospective case series. Only 1 randomized clinical trial was available. There is very limited evidence from randomized clinical trials to support the use of any kind of sclerosants. Besides, data were not always sufficient. Also, this review shows substantial heterogeneity among patients in factors that predict outcomes such as lesion size, lesion location, total number of procedures, and so on. These factors are especially relevant when comparing different sclerosants. In addition, the lack of non-English studies and single database could potentially impact the conclusions.

In conclusion, sclerotherapy is effective for venous malformations. Based on the existing review, it is still hard to figure out which sclerosant is the most effective on VMs. There is a need to lift evidence on sclerotherapy to a higher level by performing randomized controlled trials on the different sclerosants. Furthermore, choosing proper sclerosant according to the patient’s health condition and the lesion’s location may be a potential method to decrease complications.