How Often Do Varicose Veins Come Back? Recurrence Rates by Procedure, at 5, 10 and 14 Years

The question comes up before treatment and again afterwards:

"Will they come back?"

Some possibility of recurrence exists with every method. What has changed recently is that we can now answer with long-term randomised data — trials reporting at ten and fourteen years, rather than the one- and two-year figures that dominated the literature for a long time.

Those two trials appear at first to contradict each other. They do not, and understanding why is the most useful thing in this article.

---

The varicose vein treatment methods being compared

• High ligation and stripping (HL/S) — the traditional operation; the vein is tied off and physically removed
• Endovenous laser ablation (EVLA) — a fibre inside the vein closes it with heat
• Radiofrequency ablation (RFA) — the same idea using radiofrequency energy
• Ultrasound-guided foam sclerotherapy (UGFS) — a foamed sclerosant injected to close the vein
• Cyanoacrylate closure (CAC, VenaSeal) — medical adhesive; no tumescent anaesthesia required
• Mechanochemical ablation (MOCA, ClariVein) — a rotating wire damages the vein lining while a sclerosant is infused; no tumescent anaesthesia

---

Recurrence at fourteen years — the longest randomised follow-up

Published in the European Journal of Vascular and Endovascular Surgery in 2026, Rahman and colleagues at Helsinki University Hospital randomised 233 patients with symptomatic great saphenous reflux 1:1:1 to HL/S, EVLA or UGFS. 214 were included in the analysis, and the 196 treated at Helsinki University Hospital were invited to the fourteen-year follow-up.

Recurrent above-knee GSV reflux on duplex

	Recurrence
HL/S	2.6% (1 of 38)
EVLA	6.1% (3 of 49)
UGFS	30.8% (12 of 39)

The denominators in the table are those the paper reports for this outcome in each arm — 126 patients in all, fewer than the 214 included. Overall p < .001. UGFS versus HL/S: RR 11.7 (95% CI 1.6–85.6). UGFS versus EVLA: RR 5.0 (95% CI 1.5–16.6).

Note that first confidence interval. It spans 1.6 to 85.6 because the comparison rests on a single recurrence event in the stripping arm. The direction is clear; the magnitude is not precisely known.

Re-intervention on the GSV — HL/S 13.2%, EVLA 8.2%, UGFS 56.4% (overall p < .001).

Two findings that complicate the picture

Neovascularisation did not differ: 29.0% after HL/S, 22.5% after EVLA, 23.1% after UGFS (p = .76). New vessels grow back at similar rates whatever was done.

Quality of life was comparable across all three arms. Aberdeen Varicose Vein Questionnaire scores were 8.9 (HL/S), 6.1 (EVLA) and 8.8 (UGFS), overall p = .21 — including the UGFS group with its 30.8% recurrence and 56.4% re-intervention rate.

That last point deserves emphasis. Anatomical recurrence and how a patient feels fourteen years later are not the same variable. A duplex finding of reflux is not automatically a bad outcome for the person carrying it.

---

Recurrence at ten years — and an apparent contradiction

Mohamed and colleagues at Hull published ten-year outcomes in the British Journal of Surgery in 2024, comparing conventional surgery against EVLA in patients randomised for unilateral GSV reflux. Data were obtained for 206 of 280 patients (73.6%).

Clinical recurrence: EVLA 37% versus surgery 59% (p = 0.005). The number needed to treat with EVLA to avoid one clinical recurrence over ten years was five.

EVLA also scored better on quality of life — SF-36 bodily pain 84 versus 62 (p = 0.009), general health 77 versus 67 (p = 0.017), AVVQ 3.1 versus 6.3 (p = 0.029). Both groups retained significant improvement over baseline, and disease progression from baseline occurred in only 10.7%.

So: Helsinki finds stripping and laser comparable, with stripping numerically ahead. Hull finds lower clinical recurrence after laser than after surgery. Both are randomised trials with long follow-up.

Why the two trials disagree — they measure different things

The reconciliation is in the outcome definitions.

Helsinki's primary outcome was above-knee GSV reflux on duplex ultrasound — an anatomical finding, detected by scanning, whether or not the patient has visible veins.

Hull's outcome was clinical recurrence — visible, symptomatic varicose veins the patient actually presents with.

These diverge. A treated vein can show reflux on ultrasound without producing visible varicosities, and visible varicosities can arise from vessels other than the one that was treated. Stripping physically removes the trunk, so there is less of it left to reflux — which favours it on the anatomical measure. Neovascularisation after groin dissection, which stripping involves and endovenous methods avoid, plausibly feeds clinical recurrence — which disfavours it on the clinical measure.

Neither trial is wrong. They answer different questions, and the second one is closer to what a patient cares about.

---

The five-year recurrence evidence, method by method

Before the ten- and fourteen-year trials, most of what was known came from five-year data. It is still the only comparative evidence for the newer methods.

Stripping versus laser. A 2018 meta-analysis of randomised trials with five or more years of follow-up (Kheirelseid and colleagues; nine RCTs, 2,185 legs, of which 1,352 were followed to five years) found recurrence of 36.6% after EVLA and 33.3% after surgery — no significant difference (pooled RR 1.35, 95% CI 0.76–2.37, p = .3). The authors added that the numbers were too small for definitive conclusions. The 2021 Cochrane review (Whing and colleagues; 24 RCTs, 5,135 participants) agreed: recurrence was comparable within three years (OR 0.78, 95% CI 0.47–1.29; 7 studies, 1,459 participants) and at five years (OR 1.09, 95% CI 0.68–1.76; 7 studies, 1,267 participants; moderate certainty). Technical success favoured EVLA up to five years (OR 2.31, 95% CI 1.27–4.23) but not beyond (OR 0.93, 95% CI 0.57–1.50).

Radiofrequency. In the same Cochrane review, recurrence after RFA and after stripping did not differ up to three years (OR 0.93, 95% CI 0.58–1.51; 4 studies, 546 participants). A possible long-term advantage for RFA was reported (OR 0.41, 95% CI 0.22–0.75), but from a single study of 289 participants, graded low certainty. The earlier 2014 Cochrane review (Nesbitt and colleagues) found no difference between RFA and surgery in clinician-noted recurrence (OR 0.82, 95% CI 0.49–1.39), late recanalisation (OR 1.09) or neovascularisation (OR 0.31, 95% CI 0.06–1.65).

A note on neovascularisation. That 2014 review did find neovascularisation clearly lower after laser than after surgery (OR 0.05, 95% CI 0.01–0.22). The fourteen-year Helsinki data above found no difference between arms. Early and late findings do not always agree.

Cyanoacrylate (VenaSeal). A 2020 network meta-analysis (Kolluri and colleagues; 20 RCTs, 4,570 patients) estimated anatomic success to be higher with VenaSeal than with the other treatments compared, followed by RFA and EVLA. The odds of adverse events were 2.7 times higher with EVLA, 1.6 times with surgery and 1.1 times with RFA than with VenaSeal. Two things should be read alongside this: anatomic success was defined as closure within six months, and several of the authors were employed by the device's manufacturer. The Farah review, discussed below, graded its cyanoacrylate findings low certainty, and hypersensitivity-type skin reactions have been reported. Data beyond ten years do not yet exist.

Mechanochemical ablation (ClariVein). The MARADONA trial (Holewijn and colleagues, 2019) randomised 213 patients to MOCA or RFA; enrolment stopped early, at 46% of the intended number, when reimbursement was suspended. Anatomic success was lower after MOCA at one year (83.5% versus 94.2%, p = .025) and at two years (80.0% versus 88.3%, p = .066), driven mostly by partial recanalisation. Clinical success did not differ — 88.7% versus 93.2% at one year (p = .315) and 93.0% versus 90.4% at two years (p = .699). Pain over the first fourteen days was lower after MOCA (median 0.2 versus 0.5, p = .010), though the absolute difference was small, and hyperpigmentation was more frequent (7 versus 2 patients, p = .038). The Cochrane review found no clear difference in recurrence between RFA and MOCA (OR 1.00, 95% CI 0.21–4.81) and noted that long-term data are not available.

Here again, the anatomical measure and the clinical measure point in different directions.

---

What the pooled evidence on recurrence says

A 2026 meta-analysis in Scientific Reports (Sayarer and colleagues, 27 randomised trials) found anatomic success statistically indistinguishable across the thermal and surgical options:

• EVLA versus surgery: RR 0.98 (95% CI 0.90–1.06), p = .63
• EVLA versus RFA: RR 1.01 (95% CI 0.95–1.07), p = .80
• RFA versus surgery: RR 1.00 (95% CI 0.89–1.13), p = .94
• EVLA reduced nerve injury versus surgery: RR 0.47 (95% CI 0.29–0.76), p < .01
• UGFS showed inferior long-term anatomic durability

The systematic review underpinning the SVS/AVF/AVLS guidelines (Farah and colleagues, 73 studies including 45 RCTs) reached partly different conclusions and is worth reading carefully:

• HL/S showed higher anatomic closure at 30 days and 5 years than both RFA and UGFS (moderate certainty)
• No significant difference between HL/S and EVLA at 5 years
• UGFS carried increased recurrence risk versus HL/S
• EVLA had lower closure at 30 days than cyanoacrylate, and higher rates than UGFS at 1 and 5 years
• Thermal interventions scored lower on generic quality of life and carried more adverse events than cyanoacrylate (low certainty)
• Thermal interventions had lower recurrence risk than UGFS, but higher recurrence risk than cyanoacrylate

Note the disagreement: Farah found stripping outperforming RFA on closure, while Sayarer's RCT-only pooling found them equivalent. Reviews differ in what they include and how they weight it. That is normal, and it is a reason not to treat any single figure as settled.

---

Summary of reported recurrence figures

Definitions and follow-up differ between studies. Read the table for direction, not as a league table.

Method	Reported figures	Notes
Stripping (HL/S)	5-year recurrence 33.3%; 10-year clinical recurrence 59%; 14-year GSV reflux recurrence 2.6%	The figure depends heavily on the definition of recurrence
Radiofrequency (RFA)	Recurrence comparable to surgery up to 3 years; possible long-term advantage in one study	Few comparative data beyond 5 years
Laser (EVLA)	5-year recurrence 36.6%; 10-year clinical recurrence 37%; 14-year GSV reflux recurrence 6.1%	Less nerve injury than surgery (RR 0.47)
Foam (UGFS)	14-year GSV reflux recurrence 30.8%; re-intervention 56.4%	Less durable when used alone on a refluxing trunk
Cyanoacrylate (VenaSeal)	Closure within 6 months estimated higher than comparators in one network meta-analysis	No data beyond 10 years; low-certainty evidence; skin reactions reported
MOCA (ClariVein)	2-year anatomic success 80.0% versus 88.3% for RFA; clinical success similar	Trial stopped early; long-term data not available

---

Why recurrence happens at all

Neovascularisation. New vessels grow around the treated area and reflux resumes. As the fourteen-year data show, this occurs at similar rates across methods — roughly a quarter of patients.

Disease progression. Varicose veins are a manifestation of chronic venous disease. New varicosities can form in vessels that were healthy at the time of treatment. In the ten-year Hull trial, clinical disease progression from baseline was observed in 10.7% of patients.

Extent of the original treatment. If reflux in the great saphenous vein is not fully dealt with, or refluxing perforator veins are left untreated, varicosities can return by another route. This is one reason the duplex assessment before treatment matters — see why we map the veins before treatment.

Obesity, pregnancy, prolonged standing at work and family history all raise the risk. These continue to operate after treatment, which is why what happens afterwards matters as much as which method was used.

---

Key points: what recurrence rates mean in practice

In the trials and meta-analyses above, EVLA and RFA showed long-term results comparable to stripping, and — for EVLA versus surgery — one meta-analysis reported roughly half the risk of nerve injury.

UGFS showed substantially higher long-term recurrence: 30.8% anatomical recurrence and 56.4% re-intervention at fourteen years in the Helsinki trial. It has a role in smaller vessels and in patients for whom other options carry more risk, but it appears less durable when used alone on a refluxing trunk.

Cyanoacrylate showed high early closure in comparative analyses and may carry lower recurrence than thermal methods, but the ten-year-plus data do not yet exist. Its evidence is graded low certainty.

In the MARADONA trial, mechanochemical ablation closed fewer veins on ultrasound than radiofrequency, with no significant difference in clinical success and slightly lower pain scores over the first fourteen days after the procedure. Long-term comparative data are not available.

Quality of life improved and stayed improved in every group. In the fourteen-year trial it was statistically indistinguishable between arms.

That final point is the one most often lost. "Recurrence" measured on a duplex scan and "the treatment worked for me" are different things, and a single recurrence percentage is not enough to decide with unless it is clear which of the two it measured.

Ask which outcome the number refers to, over what period, and from which study. At da Re-Fit Clinic we go through the ultrasound findings with each patient and explain why a given method suits their anatomy — our source vein treatment page sets out the options in more detail.

---

References

• Rahman T, Noronen K, Vähäaho S, Halmesmäki K, Venermo M. Fourteen Year Outcomes of a Randomised Controlled Trial Comparing Endovenous Laser Ablation, High Ligation and Stripping, and Ultrasound Guided Foam Sclerotherapy for Great Saphenous Varicose Veins. Eur J Vasc Endovasc Surg. 2026 Feb 1. (PMID: 41633433)
• Mohamed AH et al. Ten-year outcomes of a randomized clinical trial of endothermal ablation versus conventional surgery for great saphenous varicose veins. Br J Surg. 2024 Aug 2;111(8):znae195. (PMID: 39162334)
• Sayarer C, Arayici ME, Gencpinar T, Bayrak S. Effectiveness of endovenous ablation techniques and surgery for great saphenous vein incompetence: a comprehensive meta-analysis of randomized controlled trials. Sci Rep. 2026;16(1):11424. (PMID: 41764322)
• Farah MH et al. A systematic review supporting the Society for Vascular Surgery, the American Venous Forum, and the American Vein and Lymphatic Society guidelines on the management of varicose veins. J Vasc Surg Venous Lymphat Disord. 2022;10(5):1155-1171. (PMID: 34450355)
• Kheirelseid EAH et al. Systematic review and meta-analysis of randomized controlled trials evaluating long-term outcomes of endovenous management of lower extremity varicose veins. J Vasc Surg Venous Lymphat Disord. 2018;6(2):256-270. (PMID: 29292115)
• Whing J, Nandhra S, Nesbitt C, Stansby G. Interventions for great saphenous vein incompetence. Cochrane Database Syst Rev. 2021;8(8):CD005624. (PMID: 34378180)
• Nesbitt C, Bedenis R, Bhattacharya V, Stansby G. Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices. Cochrane Database Syst Rev. 2014;(7):CD005624. (PMID: 25075589)
• Kolluri R et al. Network meta-analysis to compare VenaSeal with other superficial venous therapies for chronic venous insufficiency. J Vasc Surg Venous Lymphat Disord. 2020;8(3):472-481. (PMID: 32063522)
• Holewijn S et al. Two-year results of a multicenter randomized controlled trial comparing Mechanochemical endovenous Ablation to RADiOfrequeNcy Ablation in the treatment of primary great saphenous vein incompetence (MARADONA trial). J Vasc Surg Venous Lymphat Disord. 2019;7(3):364-374. (PMID: 31000063)