Surgical Management Of Varicose Veins: Meta-analysis

Abstract

Open surgery remains the gold standard by which endovascular treatment of superficial chronic venous insufficiency is measured. This meta-analysis of randomized controlled trials reviews the current evidence base, comparing open and endovascular treatment of varicose veins. Systematic review of studies reporting duplex scan follow-up after open surgical, laser (endovenous laser therapy [EVLT]), or radiofrequency (VNUS Closure device, VNUS Medical Technologies, San Jose, CA) treatment of refluxing great saphenous veins was completed. Primary outcome measures were occlusion and complication rates and time taken to resume work. No significant difference in recurrence rates at 3 months between open surgery and EVLT (RR 2.19, 95% CI 0.99–4.85, p = .05) or VNUS device (RR 7.57; 95% CI 0.42–136.02) were found. Return to work is significantly faster following VNUS (by 8.24 days; 95% CI 10.50–5.97) or EVLT (by 5.02 days; 95% CI 6.52–3.52). Endovascular treatment of varicose veins is safe and effective and offers the significant advantage of rapid recovery.

Keywords

endovascular endovenous laser therapy laser radiofrequency varicose vein VNUS

Varicose veins (VVs) significantly reduce quality of life, whereas their operative treatment demonstrably improves quality of life.^1,2 This, in conjunction with their prevalence, accounts for the fact that 90,000 VV operations are performed annually in the United Kingdom, making them the most commonly performed day-surgical procedure.³

Open surgical saphenofemoral junction (SFJ) ligation combined with stripping of the great saphenous vein (GSV) above the knee and phlebectomy of incompetent tributaries remains the highest volume “gold standard” procedure.^4–7 Endovascular techniques are now well established; they have considerable advantages in recovery rate, operative time, postoperative pain, and bruising and offer the possibility of avoiding general anesthesia.^8,9 High patient satisfaction fuels demand and is giving rise to an increasing volume of endovenous practice.^10–12

Both endovenous laser therapy (EVLT) and radiofrequency ablation (RFA) thermally ablate the target vein, obviating the need for groin dissection, with its associated morbidity. The additional cost of endovascular catheters and equipment per treated leg remains in the region of £300 for both techniques. Duplex ultrasound availability and training are also required; however, overall economic costs are offset by a rapid return to work, and the cost per limb is reduced when patients have multiple vessel reflux.^13–15

Recurrence of VVs postoperatively remains a significant problem of the open technique. Recently, a published long-term follow-up from the above-cited trial found that of 133 operated legs, 83 had developed clinically recurrent VVs by 11 years (a recurrence rate of 62%).¹⁶ Much of this recurrence has been attributed to neovascularization at the SFJ, a phenomenon rarely encountered with endovascular ablation.^9,17,18

To justify their increasing use, endovascular techniques must prove their clinical noninferiority and cost-effectiveness.⁸ Here we collate and present the published data from randomized controlled trials (RCTs) and case series comparing the efficacy of open operation with laser and RFA for the treatment of primary superficial venous reflux of the GSV.

Methods

Search Strategy

Three principal electronic databases were searched, in addition to bibliographies of well-known vascular publications. The Cochrane clinical trial database was interrogated using the search term “varicose vein*,” yielding 520 RCTs, which were reviewed manually (Figure 1). Medline (1950–April 2009) and Embase (1980–April 2009) searches were performed using the terms “exp Saphenous Vein or exp Varicose Veins” and “clinical trial.mp. or clinical trial.pt. or random.mp. or tu.xs. or (randomized controlled trial or randomized or placebo or randomised controlled trial or randomized).mp,” yielding 5,388 and 2,960 results, respectively.

Figure 1.

QUORUM diagram (occlusion data). EVLT = endovenous laser therapy; RCT = randomized controlled trial; RFA = radiofrequency ablation.

Inclusion Criteria

The title and abstract of all identified articles were screened for inclusion criteria by a single reviewer. RCTs directly comparing open and endovascular treatment of GSV VVs were selected for meta-analysis. Studies without duplex follow-up or adequate technical information were excluded, as were non–English-language publications and those not relating to GSV ablation. In addition, all treatment series of open or endovascular series reporting duplex scan occlusion data, complication rates, and time taken to return to work were selected for pooled analysis.

Statistical Analysis

A meta-analysis of RCTs comparing open and endovascular treatment of GSV VVs was completed using RevMan statistical software (The Cochrane Centre, Copenhagen, Denmark), generating an aggregate relative risk (RR) with 95% confidence intervals (CIs) for the stated primary end points. Further statistical comparisons of pooled data (occlusion and complication rates from treatment series) were performed using SPSS (SPSS Inc, Chicago, IL).

Data for common and potential serious complications were pooled, and rates were calculated. Statistical significance was evaluated using a 3 × 2 Fisher exact test for all three groups, and where justified, complication rates of EVLT and RFA were compared using a two-tailed chi-square test with Yates correction.

Results

Studies Identified

Nine RCTs were included in the meta-analysis of occlusion rates (see Figure 1). Of these, six compared EVLT and three compared a VNUS Closure device (VNUS Medical Technologies, San Jose, CA) with open surgery. In addition, duplex scan follow-up data and complication rates from 87 studies (20,278 legs), in which GSV VVs were treated by open surgery, EVLT, or RFA, were pooled and analyzed (Appendix 1, Appendix 2, and Appendix 3).

Twenty-four studies (3,000 legs) with duplex ultrasound follow-up following open surgical treatment of GSV VVs were identified. Data from small RCTs have the briefest follow-up time (up to 3 years) but are useful for direct and contemporaneous comparison with endovascular treatment modalities.^8,19 Longer cohort studies follow the course of this operation over 10- (prospective) and 30- (retrospective) year periods, finding recurrence rates of 62% and 60%, respectively, much of this owing to SFJ neovascularization (type 1c recurrence).^16,20

Twenty-nine studies (4,064 legs) reported duplex ultrasound data following GSV RFA. One RCT relates to the restore catheter, which aimed to restore competence rather than ablate the vein. Given its 40% failure rate, it was soon discontinued in favor of the Closure device. The data from the restore arm of the trial are indicated as a discarded outlier on the pooled RFA results at 1 week. The Closure catheter has since been superseded by the ClosureFAST; however, this is not substantially reflected in the literature, and only two articles relate to this latest catheter.

Fifty-five studies (13,214 legs) report duplex ultrasound data following GSV EVLT. The majority of studies used 810 and 980 nm lasers, with one using 1,064 nm (252 legs) and three small series (123 legs) using 1,320 nm lasers. A number of large 810 nm series, with 1 to 3 years of follow-up, have been published recently, notably by the Italian Endovenous-Laser Working Group, Fernandez and colleagues, and Lu and colleagues, demonstrating occlusion rates of 98.6% and 91.2% at 1 year and 97% at 3 years, respectively.^21–23 Desmyttere and colleagues and Myers and Jolley have published large series with 4-year follow-up data.^24,25

Efficacy

The meta-analysis found no significant difference in recurrence rates at 3 months between open surgery and EVLT (RR 2.19, 95% CI 0.99–4.85) or VNUS Closure (RR 7.57, 95% CI 0.42–136.02). There was no difference in recurrence rate between VNUS Closure and open surgery at 2 years (RR 0.92, 95% CI 0.42–2.05) (Figure 2 and Figure 3).

Figure 2.

Endovenous laser therapy (EVLT) versus open surgery: recurrence at 3 months.^{8,9,14,59–62}

Figure 3.

VNUS Closure device versus open surgery: recurrence at 3 months and 2 years.^13,44,63,64

Pooled Gsv Occlusion Data from Published Series

Figure 4, Figure 5, and Figure 6 depict pooled recurrence rate data for GSVs treated by open surgery, RFA, and EVLT. At 2 years, the occlusion rate for the VNUS device is comparable to that of open surgery at 87.9 (95% CI 85.8–90.0%), whereas that of EVLT (all frequencies and energies) remains 91.5% (95% CI 85.9–97%). EVLT maintains durability better in comparison to RFA. Although open surgical recurrence rates are extremely high at 10 years and beyond, there are no sufficient data to confirm the long-term durability of endovascular techniques.

Figure 4.

Open surgery: combined long (great) saphenous vein (LSV) occlusion rates.

Figure 5.

Radiofrequency ablation: combined long (great) saphenous vein (LSV) occlusion rates.

Figure 6.

Endovenous laser therapy: combined long (great) saphenous vein (LSV) occlusion rates.

Subgroup Analysis: Linear Endovenous Energy Density

One-year recurrence data from published RCTs and series using an EVLT protocol delivering more than 60 J/cm were compared to those delivering a lower energy. In the former group (linear endovenous energy density [LEED] > 60 J/cm), 46 GSVs recanalized of 2,405 legs at risk (98.1% occlusion rate). In the latter, 74 GSVs recanalized by 1 year of 1,542 legs at risk (95.2% occlusion rate). This gives a small but significantly increased risk of recanalization with low energy dosage (RR 1.03, 95% CI 1.017–1.043, p < .0001).

Return to Work

When compared to open surgical treatment, return to work is significantly faster following the VNUS Closure device (−8.24 days, 95% CI −10.50 to −5.97) and EVLT (−5.02 days, 95% CI −6.52 to −3.52). Of the four RCTs reporting time taken to return to work, only Rasmussen and colleagues reported no significant advantage from endovascular treatment (Figure 7).¹⁴

Figure 7.

Return to work following endovenous laser therapy (EVLT) and VNUS Closure device versus open surgery.^13,14,63

Safety: Complication Data

Pooling the published data suggests a significantly lower incidence of thromboembolism among patients treated endovascularly, but patients treated with RFA experience significantly greater incidence of deep venous thrombosis (DVT) (1.3 vs 0.2%, p = .0001), paresthesia (10.5 vs 4%, p = .0001), and skin burns (1.1 vs 0.5%, p = .0043) than those treated with EVLT.

The incidence of wound infection is significantly (p < .0001) higher with open surgery (0.1%; 6 of 1,186) compared to endovascular ablation (0.04% overall; 4 of 11,272), whereas skin burns are an issue only with endovascular treatment (0% of the open group reported skin burns compared to 1.1% of the RFA group and 0.5% of the EVLT group). Rates of wound infection, thrombophlebitis, and hematoma do not differ significantly between patients treated with EVLT and RFA.

Discussion

Endovascular techniques are demonstrably safe. Occlusion rates following endovascular GSV ablation are high, with early results being comparable (at 3 months EVLT and VNUS studies demonstrate 96% occlusion and open studies 93%) and midterm results being superior to open surgery (EVLT attaining 93% but open surgery achieving just 77% GSV occlusion at 3 years). The heterogeneity of the data remains a problem in establishing definitive recurrence and complication rates for all treatment modalities.

Studies vary in their follow-up protocols, and most report significant loss to follow-up, such that the pooled population studied at each of the time periods diminishes sharply. The studied RFA population consisted of 2,540 legs at 1 week, diminishing to 1,116, 673, 316, 119, and 117 patients at 1 to 5 years, respectively. Likewise, the pooled EVLT population was 5,568 legs at 1 week, decreasing to 3,947, 690, and 1,554 at 1 to 3 years, and just 45 legs from two studies were scanned at 4 years.

The great majority of GSV recanalization following endovenous treatment occurs within a year postprocedurally.^26,27 This probably represents delayed technical failure, namely, inadequate energy transferred to the venous endothelium to cause the usually observed progressive venous atrophy.^27,28 Appropriate energy dosing and scrupulous operative technique, ensuring adequate contact between the catheter tip and the vein wall, have improved results, and occlusion rates approaching 100% are reported in recent studies.^24,29–31

Endovenous Laser Therapy

The success of EVLT has been shown to be energy dependent. Primary nonocclusion and early reopening of the GSV are seen more frequently when the LEED is less than 70 to 80 J/cm or, factoring vein diameter into the equation, when the endovenous fluence equivalent (EFE) is less than 20 J/cm².^22,32–37 This fits with our understanding that 810 to 940 nm laser energy is transferred to the venous endothelium indirectly by means of steam bubbles whose volume, and therefore the surface area of vein treated, is directly proportional to the energy delivered.³⁸

Disselhoff and colleagues found, in a series of 93 EVLT patients followed up over a 2-year period, that all technical failures and three-quarters of treatment failures occurred in the first half of the cohort, indicating a learning curve (p < .015), and that operative time also diminishes with experience.³⁴

Placing the patient in the Trendelenburg position during the procedure empties the GSV and facilitates energy transmission to the vein wall rather than absorption by blood. When the latter occurs, early occlusion with thrombus may resolve and recanalize, leading to technical failure, as well as excessive pain and thrombophlebitis.^23,32 Failure to wear prescribed compression stockings for a minimum of 2 weeks after surgery has also been associated with an increased recurrence rate.³⁹

RCTs comparing continuous and stepwise withdrawal favor the former (Darwood and colleagues, 810 nm,⁸ and Disselhoff and colleagues, 810 nm⁴⁰), although Proebstle and colleagues failed to reproduce this finding in a very small group of 25 stepwise versus 15 continuous cases using a 940 nm laser.⁴¹ Kabnick found 810 and 980 nm devices to be essentially equivalent in outcome.⁴²

Fernandez and colleagues, analyzing their registry of almost 2,000 EVLT procedures, found body mass index > 30 kg/m² and a vein diameter > 8.5 mm to be the only factors predictive of recanalization, given equivalent energy dosing.^22,37 The latter result was reproduced independently by Desmyttere and colleagues and Gonzalez-Zeh and colleagues, who found that increasing GSV diameter was associated with failure of both foam sclerotherapy (odds ratio 1.68, 95% CI 1.24–2.27, p < .0008) and EVLT (odds ratio 1.91, 95% CI 1.02–3.59, p = .0428).^24,43 Gonzalez-Zeh and colleagues reported 90% treatment success using foam sclerotherapy for veins < 6.5 mm and laser for veins < 12 mm.⁴³

Radiofrequency Ablation

The majority of published articles describing RFA relate to the VNUS Closure device. The radiofrequency generator for this device (regulated by a feedback loop to keep a constant temperature of 85°C at the catheter tip) did not allow in vivo measurement of energy delivery. The new VNUS ClosureFAST catheter, which uses a 7 cm catheter tip to deliver radiofrequency energy, does allow accurate energy measurement and reduces operative time.

In 2008, Proebstle and colleagues reported their use of VNUS ClosureFAST in a series of 252 legs.²⁹ Treatment temperature was raised to 120°C, delivering an average LEED of 116.2 J/cm along the proximal 7 cm vein segment, 2 cm distal to the SFJ (which was treated twice), and 68.2 J/cm along the remainder of the vein. The EFE calculated from the same data set at 3 cm distal to the SFJ was as high as 82 ± 25 J/cm². Occlusion rates over 6 months using this protocol were 99.6%.

The EVOLVeS study (using the VNUS Closure device) demonstrated significant clinical improvement and significantly reduced postprocedural pain (compared to open surgery) at all follow-up points to 2 years.⁴⁴ Its recurrence rate was marginally lower than that of open surgery (14 vs 20% at 2 years), with significantly reduced neovascularization (type 1c recurrence) in the endovascular arm.

Logistic regression analysis finds catheter pullback speed (p < .0001) and body mass index (p < .0333) to be risk factors for anatomic failure.⁴⁵ The VNUS pullback speed is inversely proportional to and essentially an indirect measurement of LEED. In contrast to EVLT, however, vein diameter does not affect clinical or ultrasound-assessed outcome, as shown by Nicolini's series of 330 GSVs treated by the VNUS Closure, with a mean preoperative vein diameter of 7.7 mm (3–17.4 mm).⁴⁶

Is Adjunctive SFJ Ligation Necessary?

SFJ ligation has been described as an adjunctive procedure for both EVLT and VNUS Closure device.^47,48 It was considered desirable both to increase GSV occlusion rates and prevent thrombus extension into the deep femoral vein with the associated risk of pulmonary embolism.⁴⁹

The absolute risk of DVT and pulmonary embolism is small, however, and not significantly altered by performing adjunctive SFJ ligation-nor is thromboembolic risk related to proximal GSV diameter.⁵⁰

Avoiding groin dissection (with its consequent inflammatory reaction, promoting neovascularization) and leaving the SFJ intact allow anatomic drainage of proximal GSV tributaries and reduce recurrence rates.^51,52 The majority of vascular specialists have now discarded the practice of adjunctive SFJ ligation in favor of a purely endovascular technique of GSV ablation, commencing just distal to the epigastric vein to avoid proximal thrombus extension.^53,54

Extent of Ablation, Avulsions, and Anesthetic

Readily available duplex imaging has enabled selective treatment of refluxing veins while leaving functioning veins intact. Phlebectomy may not be routinely indicated. Once axial reflux is treated and venous hypertension abolished, varices diminish in size and clinical significance.⁴⁶ Morphologic studies have shown that in two-thirds of patients with GSV reflux, disease is limited to the thigh and just below the knee, but reflux extending beyond this should be treated.⁵⁵

Most study protocols ablate the GSV above the knee, with few routinely treating the GSV below the knee. When patients have reflux throughout the whole length of the GSV, extending EVLT from the groin to the midcalf (rather than limiting it to above the knee) is effective in reducing the need for concomitant ambulatory phlebectomy or sclerotherapy for residual veins to as little as 17%, without increasing the risk of saphenous nerve injury.³¹

Complications

The most serious complications associated with VV surgery, other than bleeding, are thromboembolic. The incidence of DVT after traditional open VV surgery is quoted in the literature to be as high as 5%. Principal risk factors are family history, age, and severity of venous disease (greatest with CEAP class 5 and 6, ie, ulcer patients, probably owing to their poor mobility and increased inflammatory state). Low-molecular-weight heparin thromboprophylaxis does not significantly alter the incidence.⁵⁶

Our analysis (Table 1) shows both EVLT and the VNUS Closure device to be safe, with a comparable and extremely low risk of pulmonary embolism and significantly lower risk of DVT than open surgery (0.3% for EVLT and 1.2% for the VNUS device compared to 1.8% open surgery; p = .0001).

Table 1.

Complication Rates (%)

Complication	Open (A)	RFA (B)	EVLT (C)	p Value
	13 Studies (1,186 legs)	22 Studies (2,998 legs)	42 Studies (8,274 legs)	All Groups (A vs B vs C)	RFA/EVLT (B vs C)
Skin burn	0	1.1	0.5	.0003	.0043
Paresthesia	5.3	10.5	4	< .0001	.0001
Hematoma	3.2	0.2	0.5	< .0001	.09
Thrombophlebitis	0.8	2.1	2.6	< .0001	.15
DVT	1.8	1.3	0.2	.0001	.0001
PE	0	0.06	0.01	.226	—
Wound infection	0.1	0.06	0.02	< .0001	.6

DVT = deep venous thrombosis; EVLT = endovenous laser therapy; PE = pulmonary embolism; RFA = radiofrequency ablation.

Overall comparison of three groups (A, B, and C) using a 3 × 2 Fisher exact test; analyses of RFA versus EVLT (groups B and C) using a chi-square test with Yates correction.

Pain and bruising are diminished by endovascular treatment, reflected in a rapid return to normal activity and work. Anecdotally, RFA is considered less painful than EVLT. This may be explained by the focused nature of the laser, which produces local temperatures in excess of 1,000°C, causing histologically detectable vein perforations.^28,38,57,58

VNUS Medical Technologies has sponsored the only published RCT comparing EVLT and RFA (VNUS ClosureFAST).³⁰ This study reported significantly less pain and bruising during the first 2 postprocedural weeks among the VNUS cohort, with no difference in pain at a month. Improvement in global quality of life was also significantly improved with the VNUS device at 1 and 2 weeks but was similar at 1 month. Return to work or normal activity was not reported. Eighty-seven patients were enrolled in the trial and followed for 1 month. The study was thus not designed or powered to find differences in safety and efficacy but was tailored to look at those aspects of the VNUS device that are thought to be marginally superior.

The Endovascular Varicose Vein VNUS versus EVLT RCT (or EVVERT study, ISRCTN 63135694) is a single-center study under way at St George's Vascular Institute (London, UK). It is independent of corporate sponsorship and seeks to enroll 200 patients with primary GSV VVs and randomize them to receive EVLT (810 nm, Varilase Bright-tip, Vascular Solutions Inc., Minneapolis, MN) or RFA (VNUS ClosureFAST). The study is powered to detect differences in occlusion rate under carefully standardized conditions and as such should provide valuable information regarding the relative efficacy of these two techniques.

The incidence of phlebitis increases with venous diameter and may occur more often with the 810 nm than with the 980 nm laser.⁴² However, Lu and colleagues reported a total absence of thrombophlebitis in their series of over 1,100 patients treated by 810 nm laser.²³ They stressed the importance of using intraoperative limb elevation and stocking compression to limit this complication.

Return to Work

The meta-analysis of RCTs confirmed significantly faster recovery rates of endovascular ablation compared to conventional VV surgery by 8.24 days following the VNUS device (95% CI 10.50–5.97) and 5.02 days after EVLT (95% CI 6.52–3.52).¹⁴ Shorter recovery rates are decisive in significantly decreasing overall economic cost; however, patients' subjective interpretations of pain and of readiness to return to work are multifactorial and remain difficult to evaluate objectively.¹³

Conclusion

Endovascular treatment of VVs is safe, and short- to medium-term efficacy is at least equivalent to that of open surgery. Pooled data demonstrate that the 2-year occlusion rate for the VNUS device is comparable to that of open surgery at 87.9% (95% CI 85.8–90.0), whereas it is 91.5% (95% CI 85.9–97) for EVLT (pooling all frequencies and energies). The latest catheters and protocols offer substantial improvements over these results.

RFA has marginally higher recanalization rates than EVLT, but the heterogeneity of reported data and the fact that few studies relate to the new VNUS ClosureFAST RFA catheter render this finding open to question.

The heterogeneous nature of published data shows the need for well-designed RCTs in this field of surgery to accurately evaluate differences in recanalization and complication rates between the latest EVLT and VNUS catheters, under rigorously standardized conditions. Accurate information regarding postoperative pain and bruising and the time taken to return to work by patients undergoing these two increasingly used endovascular treatment modalities would be of benefit to the vascular specialist and inform patient choice.

Acknowledgment

R. Brar was awarded a research fellowship by the Royal College of Surgeons of England.

Financial disclosure of reviewers: None reported.

Footnotes

Appendix 1.

Open Surgery

Lead Author	Year	Study Design	Center	Legs (GSV)	Longest Duplex Follow-Up (mo)
Allegra⁶⁵	2007	Series	Single	862	60
Belcaro⁶⁶	2000	RCT vs foam	Multi	42	120
Bountouroglou*⁶⁷	2006	RCT vs foam	Single	30	3
Darwood⁸	2008	RCT vs EVLT	Single	32	3
De Maeseneer⁶⁸	2005	Series	Single	100	12
De Medeiros*⁵⁹	2005	RCT vs EVLT	Single	20	1
Disselhoff*⁹	2008	RCT vs EVLT	Single	60	0
Dwerryhouse⁶	1999	RCT (strip/ligate)	Single	52	60
Fischer²⁰	2001	Series	Single	125	408
Frings*⁶⁹	2004	Series	Single	500	24
Hinchliffe*¹⁹	2006	RCT vs RFA	Single	16	12
Kalteis*⁶¹	2008	RCT vs EVLT	Single	48	3
Kianifard*¹⁸	2006	NRCT vs RFA	Single	55	12
Lurie*⁶³	2003	RCT vs RFA	Multi	36	3
Lurie⁴⁴	2005	RCT vs RFA	Multi	36	24
Mekako⁶⁰	2006	NRCT vs EVLT	Single	62	3
Perala*⁶⁴	2005	RCT vs RFA	Single	13	36
Rasmussen*¹⁴	2007	RCT vs EVLT	Single	68	6
Rautio*¹³	2002	RCT vs RFA	Multi	13	0
Sarin⁷	1994	RCT (strip/ligate)	Single	43	24
Smith⁷⁰	2002	RCT (duplex/no)	Single	189	12
Van Rij*⁵⁶	2004	Series	Single	377	0
Vuylsteke*⁷¹	2006	RCT vs EVLT	Single	124	6
Winterborn¹⁶	2004	RCT (strip/ligate)	Single	133	120

EVLT = endovenous laser therapy; GSV = great saphenous vein; no=no duplex; NRCT = non randomized controlled trial; RCT = randomized controlled trial; RFA = radiofrequency ablation.

*Complication data.

Appendix 2.

Radiofrequency Ablation

Lead Author	Year	Study Design	Center	Legs (GSV)	Duplex Follow-Up (mo)	Device
Almedia*³⁰	2009	RCT	Multi	46	1	ClosureFAST
Almedia⁷²	2006	Series	Single	128	-	Closure
Boros*⁵²	2008	Series	Single	77	0.25	Closure
Boros*⁵²	2009	Series	Single	142	0.25	Closure
Chandler*⁴⁷	2000	Series	Multi	301	3	Closure
Fassiadis*⁷³	2002	Series	Single	18	12	Closure
Fassiadis⁵¹	2002	Series	Single	127	12	Closure
Goldman*^19,74	2000	Series	Single	12	0	Closure
Hinchliffe*¹⁹	2006	RCT	Single	16	12	Closure
Hingorani*⁵³	2004	Series	Single	73	0.25	Closure
Kianifard¹⁸	2006	NRCT	Single	55	12	Closure
Labropolous*⁷⁵	2006	Prospective series	Single	46	0.25	Closure
Lurie*⁶³	2003	RCT	Multi	44	3	Closure
Lurie⁴⁴	2005	RCT	Multi	44	24	Closure
Manfrini*⁷⁶	2000	RCT	Multi	152	6	Closure
Manfrini	2000	RCT	Multi	68	0.25	Restore
Marston⁷⁷	2006	NRCT	Single	58	1	Closure
Merchant*⁷⁸	2002	Series	Multi	319	24	Closure
Merchant*^45,79	2005	Series	Multi	1,154	60	Closure
Nicolini*⁴⁶	2005	Series	Multi	330	36	Closure
Ogawa*⁸⁰	2005	Series	Single	25	1	Closure
Perala*⁶⁴	2005	RCT	Single	15	36	Closure
Pichot¹⁷	2003	Series	Multi	63	24	Closure
Proebstle*²⁹	2008	Series	Multi	252	6	ClosureFAST
Puggioni*⁸¹	2005	Series	Single	130	0.25	Closure
Rautio*¹³	2002	RCT	Multi	15	0	Closure
Ravi⁸²	2006	Series	Single	159	0.25	Closure
Salles-Cunha⁸³	2004	Series	Single	106	6	Closure
Sybrandy*⁸⁴	2002	Series	Single	26	6	Closure
Weiss*⁸⁵	2002	Series	Single	140	24	Closure

GSV = great saphenous vein; NRCT = non randomized controlled trial; RCT = randomized controlled trial.

*Complication data.

Appendix 3.

Endovenous Laser Therapy

Lead Author	Year	Study Design	Center	Legs (GSV)	Duplex Follow-Up (mo)	Wavelength (nm)
Agus (IEWG)*²¹	2006	Series	Multi	1,076	36	810–980
Almedia^† ⁷²	2006	Series	Single	819	12	810–980
Almedia*^† ³⁰	2009	RCT (vs RFA)	Multi	41	1	980
Bush*⁸⁶	2005	Series	Single	620	6	940
Chang*⁴⁸	2002	Series	Multi	252	-	1,064
Corcos*⁸⁷	2008	Series	Single	44	24	808
Darwood^† ⁸	2008	RCT	Single	71	3	810
De Medeiros*⁵⁹	2005	RCT	Single	20	1	810
Desmyttere*²⁴	2007	Prospective series	Multi	511	48	980
Disselhoff*⁹	2008	RCT	Single	60	24	810
Disselhoff*⁴⁰	2008	RCT	Single	86	24	810
Disselhoff*³⁴	2005	Series	Single	93	24	810
Fernandez*^† ²²	2008	Series	Single	1,985	24	810
Gonzalez-Zeh*^† ⁴³	2008	RCT	Single	45	12	980
Hayes^† ⁸⁸	2008	RCT	Multi	19	1	1,320
Huang*⁸⁹	2005	Series	Single	230	12	810
Jung*^† ⁹⁰	2008	Prospective series	Single	135	3	810
Janne d'Othee	2008	RCT	Single	157	3	810
Kalteis*⁶¹	2008	RCT	Single	47	3	810
Kabnick*⁴²	2006	RCT	Single	30	12	810
Kabnick*	2006	RCT	Single	30	12	980
Kavuturu*^† ⁹¹	2006	Series	Single	66	3	940
Labropolous*⁷⁵	2006	Prospective series	Single	56	0.25	-
Lu*^† ²³	2008	Series	Single	1,186	12	810
Marston⁷⁷	2006	NRCT	Single	31	1	810
Mekako*^† ⁹²	2006	Series	Single	70	3	810
Mekako^† ⁶⁰	2006	RCT	Single	70	3	810
Min*²⁷	2003	Series	Multi	499	24	810
Min*⁹³	2001	Series	Single	90	12	810
Myers*⁹⁴	2006	Series	Multi	404	36	810
Myers²⁵	2009	Series	Single	509	48	810
Navarro*⁹⁵	2001	Series	Multi	40	6	810
Oh*⁹⁶	2003	Series	Single	15	3	980
Perkowski*³⁹	2004	Series	Single	203	12	940
Pleister⁹⁷	2008	Series	Single	50	-	980
Proebstle*⁴¹	2003	Series	Single	41	12	940
Proebstle*²⁸	2003	Series	Single	109	12	940
Proebstle³²	2004	Series	Single	106	3	940
Proebstle*³⁸	2002	Series	Single	31	1	940
Proebstle*^† ³⁵	2006	NRCT	Single	249	3	940
Proebstle*^† ⁹⁸	2005	NRCT	Single	33	3	1,320
Puggioni*⁸¹	2005	Series	Single	77	0.25	810
Rasmussen*^† ¹⁴	2007	RCT	Single	69	6	980
Ravi⁸²	2006	Series	Single	990	0.25	810
Sadick*^† ²⁶	2007	Series	Single	94	48	810
Theivakumar^† ³¹	2008	RCT	Single	68	3	810
Theivakumar*³⁶	2007	Series	Single	644	3	810
Theivakumar^† ⁹⁹	2008	NRCT	Single	84	12	810
Timperman*^† ³³	2004	Series	Multi	111	6	810 and 940
Timperman*^† ³⁷	2005	Prospective series	Multi	100	6	810 and 940
Van den Bremer¹⁰⁰	2008	Series	Single	403	1	980
Viarengo^† ¹⁰¹	2007	RCT	Single	27	12	980
Vuylsteke*⁷¹	2008	Prospective series	Single	129	6	980
Vuylsteke*⁶²	2006	RCT	Single	118	6	980
Yang¹⁰²	2006	Series	Single	71	12	1,320

GSV = great saphenous vein; IEWG = Italian Endovenous-Laser Working Group; NRCT = non randomized controlled trial; RCT = randomized controlled trial; RFA = radiofrequency ablation.

*Complication data.

^†High-energy protocol.

References

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.