Comparison of polidocanol endovenous microfoam (Varithena) to physician-compounded foam in sclerotherapy treatment of truncal varicosities: A systematic review and network meta-analysis

Abstract

Background

Commercial premixed foam sclerosants such as Varithena, a polidocanol endovenous microfoam, have been developed and marketed as better alternatives to physician-compounded foams (PCFs). The aim of this study was to evaluate the efficacy of Varithena compared with PCFs in foam sclerotherapy for the treatment of truncal varicose veins.

Methods

A systematic review was performed according to PRISMA and Meta-Analysis guidelines with a registered protocol (PROSPERO: CRD42024494805) using Medline and EMBASE databases from inception to June 17, 2025, by two reviewers. Randomized controlled trials that reported on occlusion rates after truncal varicose vein treatment with foam sclerotherapy were included. A network meta-analysis was performed.

Results

Six randomized controlled trials with a total of 1726 patients were identified. Complete occlusion was defined as elimination of saphenofemoral junction reflux or occlusion of the saphenous vein on duplex imaging. The median follow-up was 39 weeks (range, 12-52 weeks). No significant difference was observed between Varithena and PCF with respect to occlusion rates at follow-up (risk ratio [RR], 1.12; 95% confidence interval [CI], 0.93-1.35; I² =71%). Moreover, similar rates of adverse events were observed between the Varithena and PCF groups (RR, 2.89; 95% CI, 0.64-12.97; I² = 0%). Compared with foam sclerotherapy with Varithena, surgical removal and endovenous laser ablation showed a significantly higher occlusion rate at follow-up (RR, 1.31; 95% CI, 1.18-1.47; I² =71% and RR, 1.32; 95% CI, 1.18-1.49; I² =71%, respectively).

Conclusions

In patients with saphenous varices, there was no indication that Varithena was superior to or associated with fewer complications than PCFs Further trials are required to define the role of Varithena in the treatment of varicose veins, especially within the context of less expensive alternatives.

Minimally invasive endovascular techniques, such as ultrasound-guided foam sclerotherapy (UGFS) and endovenous thermal ablation, are widely used to treat patients with varicose veins and chronic venous insufficiency.¹ These techniques are recommended by international guidelines over traditional open surgical procedures, owing to the faster recovery times and lower complication rates.² Foam sclerotherapy remains the most frequently performed intervention, despite lower occlusion rates compared with endovenous thermal ablation.² Common foam sclerosants include premixed commercial agents, such as polidocanol 1% endovenous microfoam (Varithena), and point-of-care physician-compounded foams (PCFs) formed using polidocanol or sodium tetradecyl sulphate (STS). Polidocanol endovenous microfoam (PEM) has been purported to have a more consistent bubble size and broader bubble size distribution compared with PCFs, which results in better foam stability and overall performance.³

However, there remains a lack of consensus on whether commercial foam sclerosants or PCFs are more effective in treating truncal varices. This systematic review and network meta-analysis aimed to review the current literature and evaluate the efficacy of Varithena compared with PCFs used in foam sclerotherapy for the treatment of truncal varicose veins.

Methods

This study was conducted in adherence to the PRISMA guidelines⁴ (Fig 1). The study protocol was prospectively registered on PROSPERO and available to access online (ID: CRD42024494805).

Fig 1.

PRISMA diagram.

Search strategy

The Medline and EMBASE databases were systematically searched from 1943 to June 17, 2025. Unpublished literature was screened by searching for trials on ClinicalTrials.gov and the World Health Organisation International Clinical Trials Registry Platform portal. The search algorithm used appropriate terms and keywords including sclerotherapy, varicose veins, polidocanol microfoam, and PCF (Supplementary Fig, online only).

Inclusion and exclusion criteria

Included studies were randomized controlled trials (RCTs) published English and available in full text comparing Varithena and PCF directly, or a comparison of these foam sclerosants with endovenous ablation techniques or surgical removal in adult patients for the treatment of superficial truncal veins (Fig 1). Studies were excluded if the indication for foam sclerotherapy was solely for tributary vein treatment, nonvascular, or for the treatment of arteriovenous malformations. All non-RCT studies were excluded. Studies which included interventions such as cyanoacrylate closure, mechanochemical ablation, liquid sclerotherapy, or mixed methods treatment were excluded.

Data extraction and quality assessment

Abstract and full-text screening were performed by two independent reviewers, with any disagreements resolved by a third reviewer. Data extraction was performed independently by two reviewers using a standardized template on Microsoft Excel. The reference lists of all included studies were screened for completion.

Quality assessment of RCTs was performed by two reviewers using the Cochrane Risk of Bias Tool for Randomized Controlled Trials⁵ demonstrated in Fig 2. Any discrepancies were mediated by a third reviewer.

Fig 2.

Risk of bias assessment of the included studies.

Outcome measures

Data were collected on study characteristics, primary outcomes (complete occlusion), and secondary outcomes including complications such as deep vein thrombosis (DVT) rate, hyperpigmentation, parasthesia, quality-of-life assessments and methods, and the rate of venous ulceration healing.

Statistical analysis

Statistical analyses were computed using R version 3.3.2 (R Core Team, GNU GPL v2 License), R Studio version 1.0.44 (RStudio Inc., GNU Affero General Public License v3) and RevMan5 (The Cochrane Collaboration). Heterogeneity was assessed using the I² statistic and visualized with a Baujat plot. Heterogeneity of >70% prompted a narrative synthesis alongside the quantitative analysis.

Outcome data were pooled with a conventional head-to-head meta-analysis using the meta package. A network meta-analysis was conducted using the netmeta package. Where overall proportions were low, a logit transformation was used to normalize the data. Where feasible, moderators of outcome collected from the baseline data were included in the statistical model.

Results

A total of 735 articles were identified from the literature search as shown by the PRISMA diagram in Fig 1. Six RCTs studies with a total of 1726 patients were identified as meeting the inclusion and exclusion criteria. The main characteristics of the included studies are presented in Table I. The sclerosant used in each study was either Varithena or PCF using polidocanol or STS. One study used Varithena, two studies used polidocanol, two studies used STS, and the remaining study used both types of PCFs.

Table I.

Study characteristics of included studies

Investigator	Study design; location	Inclusion and exclusion criteria (selected)	Sclerosant type	Comparisons	Patient demographics	Follow-up, weeks
Wright et al 20066	RCT; multinational (Europe)	Inclusion: Adults aged 18-75 years with SFJ or SPJ incompetence on Duplex scanning Exclusion: Previous DVT, BMI >32 kg/m2	Varithena	PCF CS	n = 656, CEAP C2-C4, GSV, SSV or both	12
Figueiredo et al. 20097	RCT; Brazil	Inclusion: Adults aged between 18 and 70 years Exclusion: Previous treatment of ipsilateral GSV	POL	CS	n = 60, saphenofemoral or saphenopopliteal	26
Lattimer et al. 20128	RCT; United Kingdom	Inclusion: Adults with symptomatic varicose veins, SFJ reflux on duplex Exclusion: SPJ incompetence, GSV diameter >12 mm	STS	EVLA	n = 110, CEAP C2-C6, GSV	12
Biemans et al. 20139	RCT; Netherlands	Inclusion: Adults, primary incompetent GSV above knee and incompetent SFJ with duplex Exclusion: Previous ipsilateral GSV treatment, deep venous incompetence	POL	EVLA CS	n = 223, CEAP C2-C5, GSV	52
Brittenden et al. 201410	RCT; United Kingdom	Inclusion: Adults aged >18 years, uni- or bilateral symptomatic primary varicose veins, GSV or SSV reflux >1 second Exclusion: Current DVT, acute superficial VT	STS	EVLA CS	n = 785, CEAP C2-C4, saphenous veins	6
Venermo et al. 201611	RCT; Finland	Inclusion: Unilateral uncomplicated varicose veins, GSV reflux on Duplex Exclusion: Peripheral arterial disease, BMI >40 kg/m2, previous DVT	POL STS	EVLA CS	n = 214, CEAP C2-C4, GSV	52

CEAP, Clinical-Etiology-Anatomy-Pathophysiology; BMI, body mass index; CS, surgery; DVT, deep vein thrombosis; EVLA, endovenous laser ablation; GSV, great saphenous vein; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam; POL, polidocanol; RCT, randomized controlled trial; SFJ, saphenofemoral junction; SPJ, saphenopoliteal junction; SSV, short saphenous vein; STS, sodium tetradecyl sulfate; VT, vein thrombosis.

The search strategy yielded four studies from Europe (UK, the Netherlands, Finland), one from Brazil, and one international study. The range of Clinical-Etiology-Anatomy-Pathophysiology classifications reported in the included studies was most often C2 to C5. The truncal veins treated were primarily the great saphenous vein, with two studies limited to patients with other saphenous vein incompetence including short saphenous vein incompetence.

The risk of bias analysis of the included studies indicated that they were generally of good quality as demonstrated in Fig 2, with each study reporting randomization methods, reporting and data analysis adequately.

Synthesis of results

Of the 6 included studies, 14 pairwise comparisons were made. There were 1726 observations, 4 treatments, and 4 designs.

Primary outcome: Complete occlusion

All six of the included studies described primary end point of complete occlusion with a median follow-up of 39 weeks (range, 12-52 weeks). Complete occlusion was defined as elimination of saphenofemoral junction reflux or occlusion of the treated vein on duplex imaging. Through a network analysis, PEM was related to surgery and PCF as described in the network diagram (Fig 3). Evaluation of network consistency found substantial evidence of local and global network heterogeneity and inconsistency with a network I² of 71.1% (95% CI, 40.3%-86.0%) and tau² of 0.0087.

Fig 3.

Network nodes demonstrating comparative relationships between different modalities of treatment. CS, surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam (Varithena).

No significant difference was observed between Varithena and PCF with respect to the occlusion rate at follow-up (PCF vs Varithena: risk ratio [RR], 0.890; 95% confidence interval [CI], 0.739-1.073) (Table II). Compared with Varithena treatment, surgical ligation and endovenous laser ablation (EVLA) were associated with a higher occlusion rate at follow-up (RR, 1.172; 95% CI, 0.975-1.409 and RR, 1.182; 95% CI, 0.965-1.447, respectively), although this difference was statistically nonsignificant (Table II and Fig 4).

Table II.

Pooled relative risk of complete vein occlusion for each treatment modality when compared against Varithena using a random effects model

Treatment	RR	95% CI	z	P value
CS	1.172	0.975-1.409	1.69	.0918
EVLA	1.182	0.965-1.447	1.61	.107
PCF	0.890	0.739-1.073	−1.22	.222

CI, Confidence interval; CS, surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam; RR, risk ratio.

Fig 4.

Forest plot demonstrating comparisons between different modalities of treatments using a random effects model. CI, confidence interval; CS, surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam.

Table III shows both the pooled RRs from direct head-to-head comparisons of the various interventions and the indirect comparisons through the network. Both direct and indirect comparisons between surgery and PCF as well as EVLA and PCF showed similar RRs. This result suggests that surgery and EVLA achieving higher rates of complete vein occlusion is not only seen in individual trials, but also supported by the entire network analysis, further strengthening the evidence for using these modalities over PCF. When considering the effectiveness of Varithena, except for the direct comparison between EVLA and PEM, which was not performed, both direct and indirect comparisons suggest nonsignificant differences in vein closure rates when using Varithena. Similar efficacy was seen between PCF and PEM on both direct and indirect comparisons.

Table III.

League table of risk ratios (RRs) with corresponding 95% confidence intervals (CIs)—The top triangle represents the pooled RRs of direct head-to-head comparisons in the trials, while the bottom triangle represents the indirect head-to-head comparisons from the network meta-analysis

CS	1.01 (0.90-1.14)	1.34 (1.20-1.50)a	1.09 (0.89-1.34)
0.99 (0.89-1.11)	EVLA	1.39 (1.23-1.58)a	N.A.
1.32 (1.18-1.47)a	1.33 (1.18-1.49)a	PCF	0.96 (0.78-1.19)
1.17 (0.97-1.41)	1.18 (0.96-1.45)	0.89 (0.74-1.07)	PEM

CS, Surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam.

^aStatistically significant RRs.

Adverse events

Recurrence outcomes were only reported in one study and was not suitable for meta-analysis. Adverse events (AEs) were not reported consistently throughout all studies; therefore, four studies were selected for meta-analyses. There were 10 pairwise comparisons made with 1464 observations and 3 designs. Network nodes demonstrated once more that PEM was linked with surgery and PCF (Fig 5). Testing for inconsistency revealed that studies were largely consistent (tau² = 0; I² = 0%; 95% CI, 0.0%-79.2%).

Fig 5.

Network nodes demonstrating relationships between modalities for adverse events (AEs). CS, surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam.

Similar rates of AEs were observed between Varithena and PCF groups (RR, 0.35; 95% CI, 0.08-1.56) as well as with surgery (RR, 0.45; 95% CI, 0.09-2.20). However, when comparing EVLA with Varithena, EVLA had a significantly lower rate of AEs (RR, 0.178; 95% CI, 0.033-0.966) (Fig 6).

Fig 6.

Forest plot demonstrating comparisons between treatment modalities reporting adverse event (AE) rates. CI, confidence interval; CS, surgery; EVLA, endovenous laser ablation; PCF, physician-compounded foam; PEM, polidocanol endovenous microfoam.

Descriptors of adverse outcomes reported included DVT, hyperpigmentation, parasthesia, infection, and hematoma. There were a total of 20 reported incidences of DVT across 4 studies—with 9 superficial venous thromboembolisms (VTEs), four DVTs, 5 VTEs in the great saphenous vein, and 2 VTEs in the common femoral vein. Over the six studies, a total of seven patients experienced parasthesia, four experienced infections, and three experienced hyperpigmentation. Only one patient reported postsurgical hematoma.

Secondary outcomes

Within the included studies, other outcomes were analyzed that were unsuitable for meta-analysis. There was insufficient data on Varithena volume used to perform meta-analyses, as only one study reported mean volumes used (24.9 mL and 14.9 mL) with additional sessions.⁶ Similarly, reporting for UGFS sclerosing agent volumes used ranged from 4.7 mL to 12 mL, reported in four studies.

Quality-of-life measures were reported through CIVIQ, the Aberdeen Varicose Vein Questionnaire, and the Aberdeen Varicose Vein Severity Score, depending on the study, with a range of follow-up timeframes from 12 to 52 weeks. Studies also used Short Form-36 and Health assessment (a supplement of the EQ-5D).⁹

In Biemans et al,⁹ there was no significant differences between each treatment group for the CIVIQ, EQ-5D, and Health scores, although there was an improvement in all groups at 3 months and 1 year. This finding is similar to that from Venermo et al,¹¹ reporting no significant differences in improvement of Aberdeen Varicose Vein Severity Score scores between each treatment group. Brittenden et al¹⁰ reported QoL measures to be similar with worse disease-specific QoL scores in patients undergoing foam treatment.

Discussion

This systematic review and network meta-analysis summarizes existing evidence for Varithena compared with PCF agents used in UGFS. The main finding is that Varithena has no evidence of superior outcomes compared with PCF for the treatment of truncal varicose veins. Analysis of AEs did not reveal any statistically significant differences between Varithena and surgery or PCF. Only EVLA was associated with significantly lower AE rates when compared with Varithena. Patient-reported quality-of-life measures reported improvements post procedure that were not statistically significant between the treatment groups.

These results are supported by a previous meta-analysis comparing Varithena with endovenous thermal ablation by Kabnick et al,¹² finding no statistically significant differences for vein closure. A review article by Athavale et al,¹³ exploring ablation methods in larger saphenous vein disease, revealed no significant differences between vein closure and radiofrequency ablation. Varithena studies (VANISH I and II) describe a primary end point of the Varicose Vein Symptom Questionnaire post procedure instead of complete closure, with a range of closure rates of 59% to 83% at the 8-week time point.^14,15

Treatment using PEM is reported to be minimally invasive and safe as a nonsurgical procedure. A consensus statement from the American vein and lymphatic society addresses PEM as an effective treatment option for a wide range of venous anatomies.¹⁶ Recommendations from the manufacturers include limitation of PEM to 15 mL per session in a single treatment; however, patients may require more treatments with extensive disease.^14,17,18 A review of Varithena after its inception summarizes physical properties of Varithena alongside trial data surrounding clinical efficacy, safety, and AEs in placebo-controlled trials. Its physical properties include formation of microfoams with a fixed proportion of O₂:CO₂ and uniform bubble sizes. These physical properties may reflect the noninferiority of Varithena to other treatment modalities; however, more research is required to explore the mechanisms behind these improvements.¹⁹

UGFS has shown great advancements in recent years, with research supporting its use despite more frequent recurrences, citing its cost effectiveness and simple procedures. Comparisons between UGFS, open surgery, and EVLA are difficult to compare owing to significant variance between studies and reporting measures. A systematic review by Hamann et al²⁰ reports that UGFS had lower anatomical success rates; however, Darvall et al²¹ report similar patient reported outcome measures with durable results up to 6 years.¹¹

Limitations

This systematic review and network meta-analysis is limited owing to insufficient data and quality for several outcomes, which results in difficulty in making comparisons between treatment modalities. These limitations also arise from significant variance in the papers. There are limited numbers of papers for Varithena, resulting in difficulties drawing conclusions. Significant heterogeneity between included studies was observed—this is hypothesized to arise from intrinsic differences in study design, follow-up durations, and procedural techniques. A key limitation is the heterogeneity in technique between each operator with PEM, PCF, and PCM—lack of consistent reporting of air/CO₂ volumes and concentrations, mixing techniques and amounts used within each included study increases the risk of bias owing to inconsistent comparisons.

The interchangeable use of key terms such as truncal veins and varicosities needs to be distinguished in future research to clarify key findings. The reporting of occlusion rates may also not accurately reflect recurrence rates for patients. Reporting of AEs is also poor, with small numbers of patients in each group, making it difficult to draw comparisons. There are wide 95% CIs for our results. Our study also reports poor quality network inconsistency measures owing to low power and the limited numbers of studies. Therefore, there may be limited practical application of this evidence owing to poor generalizability and power, suggesting the need for further large-scale, head-to-head trials to investigate the validity of Varithena in the context of UGFS.

We suggest future research on the topic to address these limitations, including studies that distinguish anatomical occlusion and absence of saphenofemoral junction reflux, and between DVTs and endothermal heat-induced thrombi. Future studies should address standardized reporting of volumes of foam, adverse outcomes, and longer follow-up periods to address recanalization rates and AEs. The direction of future research should aim to clarify and detail differences in techniques to decrease heterogeneity and allow true comparisons between techniques.

Conclusions

This systematic review has demonstrated that, in patients with truncal varices, Varithena did not show any superiority to PCF. Higher rates of occlusion with surgical removal or EVLA of truncal varices were observed, which is concordant with the current literature and international guideline recommendations. Nevertheless, this systematic review has limitations, including heterogeneity and limited sample sizes. Further large-scale trials are required to define a role of Varithena in the treatment of varicose veins, especially within the context of less expensive alternatives.

Author contributions

Conception and design: SO, AHD

Analysis and interpretation: RC, CS, MS, BT, MT

Data collection: RC, CS, MS

Writing the article: RC, CS, MS, MT

Critical revision of the article: CS, MT, SO, AHD

Final approval of the article: RC, CS, MS, MT, BT, SO, AHD

Statistical analysis: CS, MS, MT, BT

Obtained funding: Not applicable

Overall responsibility: AHD

RC and CS contributed equally to this article and share co-first authorship.

Funding

R.C. and C.L.S. are supported by Health Education England Specialized Foundation Post and acknowledges infrastructure support for this research from the National Institute for Health Research Imperial Biomedical Research Centre.

Disclosures

None.