Conservative versus surgical treatment of venous leg ulcers: 10-year follow up of a randomized,multicenter trial

Introduction

The socio-economical aspects of venous leg ulcers are substantial: the prevalence of this disease is as high as 1–1.5%.^1,2 The costs of the treatment of venous leg ulcers are 1% of the total annual health care budget in Western-European countries.^3–5 In the United States, treatment costs of more than 6 million patients are approaching US$2.5 billion and 2 million workdays are lost annually. ⁶ Health care costs of chronic venous disease are spiraling; in the United Kingdom they have doubled over the last decade to £126 billion annually. ⁷ Treatment modalities are either conservative or surgical. Conservative treatment involves local treatment and compression therapy. Surgical treatment of venous ulcers is based on correcting venous hypertension, by treating incompetent superficial, deep, and perforating veins. Subfascial endoscopic perforating vein surgery (SEPS) combined with superficial vein ligation is performed to address vein incompetence in patients with chronic venous leg ulcers. Several reports describe good healing and low recurrence rates. ⁸ Still the role of perforator vein incompetence in the pathophysiology of chronic venous insufficiency (CVI) is debatable. ⁹ Between 1998 and 2001 a prospective, randomized trial was conducted, in order to compare the additive effect of venous surgery on ambulatory compression therapy on venous ulcer patients. The three years results were published in the Journal of Vascular Surgery in 2006. ¹⁰ We now present the results of 10-year follow up in which we scored healing and recurrence rates and quality-of-life (QoL) measurements.

Patients

All patients were part of a randomized, multicenter trial comparing conservative and surgical treatment of venous leg ulcers. Between 1997 and 2001, patients with an active (open) venous leg ulcer (CEAP C6) were randomized in 12 centers in the Netherlands. Both medial and lateral venous ulcers were included. Exclusion criteria were arterial pathology (ABI < 0.8), total or partial occlusion of the deep venous system, former subfascial ligation of perforating veins, severe neurological or muscular pathology, and immobility. Patients with bilateral leg ulcers were included in the study. The trial was approved by the ethics committees of all 12 participating centers. All patients provided written informed consent, and the study adhered to the Declaration of Helsinki. Between 2011 and 2013 we contacted the original patients, in order to invite them to our hospital for a medical interview, clinical examination, duplex ultrasonography (DUS), and a QoL assessment.

Methods

All patients were analyzed by medical history, physical examination, DUS, and digital photographs. Also generic QoL was assessed. (SF-36v1). Both primary and secondary causes of leg ulcers were included. Both deep axial and segmental reflux was considered to be deep venous insufficiency (DVI) in this series. Reflux was defined as retrograde flow >0.5 s after calf compression or Valsalva. We did not discriminate primary from secondary chronic venous disease.

Patients were stratified prior to randomization for three factors: first time ulcer or recurrent ulceration; presence or absence of DVI; and treatment center. Patients were divided into two treatment groups. For allocation to a treatment group, each patient was assigned by a computer program at an independent randomization center. The randomization took place within stratification groups. Patients with bilateral ulceration in the study, both legs were randomized separately.

All patients were treated by standardized ambulatory compression therapy and half of the patients underwent SEPS (Figures 1 and 2), combined with surgery of the incompetent superficial venous system, by flush saphenopopliteal ligation and/or saphenofemoral ligation and limited stripping of the great saphenous vein from groin to just below knee level. In the majority of cases, the procedure was carried out in a day-care setting. Ambulatory compression therapy was started directly after randomization. Surgical treatment was performed within five working days after randomization. Subfascial endoscopic perforating vein surgery (SEPS) was performed with an Olympus scope (Olympus®, Hamburg, Germany) as described elsewhere. ¹⁰ Patients were mobilized the first postoperative day and treated by dual layer short-stretch ambulatory compression therapy (Comprilan, Beiersdorf Medical, the Netherlands) until the ulcer had healed. Lifelong therapeutic elastic stockings (class II or III, depending on concomitant incompetence of the deep vein system) were prescribed after healing. Digital photographs were taken at each visit. After randomization, patients were seen on an outpatient clinic basis after 1, 2, 3, 6, and 12 weeks and every 3 months. After 1 year, patients returned on an annual basis. In cases of recurrences, patients were seen immediately. DUS and QoL assessment (SF-36v1) was repeated at 3 and 12 months after randomization. OPEN IN VIEWER

OPEN IN VIEWER

Figure 2.

Same patient as in Figure 1, 4 months after the procedure.

Long-term follow up

We contacted all patients enrolled in the original trial. All responsive patients were invited to visit the outpatient clinic of one of the study centers, where an interview, physical examination, and DUS exam were repeated and QoL scores were acquired. Patients who were unable to visit a study center were visited and examined at home. The same ultrasonographer as in the original trial performed the DUS examinations. QoL was assessed with the Aberdeen Varicose Vein Questionnaire (AVVQ) and Short-Form 36 version 2 (SF-36v2).

The actual clinical situation of the patient was scored according to the CEAP classification. Compliance with compressive stockings and/or bandages was noted. Follow-up visits took place during 2011 through 2013.

Study endpoints

Healing and recurrence rates and QoL were assessed as study endpoints. Ulcer status after 10 years of follow up was divided into four groups:

no ulcer, no recurrence (healed directly after treatment)

Being assigned to group 1 meant the ulcer healed immediately after treatment and had not recurred in the follow-up period. This group is considered “healed” and in this group treatment is considered a technical success.

Group 2 has no active ulcer at the last follow-up visit, but has seen recurrent ulceration during the follow-up period. Group 3 has an active ulcer at the last follow-up visit, with periods of healing and recurrence during the follow-up period. Group 4 has an active ulcer at the last follow-up visit, which has never healed during the follow-up period.

Statistics

Study power was set as 90% (β = 0.10 and α = 0.01) initially, which indicated that with expected healing and recurrence rates of 70% and 35% in the conservative group and 85% and 10% in the surgical group, inclusion of 100 ulcerated legs in each arm of the study was necessary. The analysis was done by intention-to-treat (ITT). Missing data at ten years for being ulcer-free or having ulcer were imputed by last-observation-carried-forward technique using data from preceding visits made to the clinic.

Prevalence of having an ulcer-free leg and having leg ulcer is scored as percentages overall and as percentages in randomized groups. Log-likelihood chi-squares test with p-values and odds ratios (OR) with 95% confidence intervals (CI) were used to test differences between conservatively and surgically treated legs. Consequentially, logistic regression analysis was done using gender, age, DVI, and diabetes mellitus (DM) as possible predictors or confounders to again test differences between the randomized groups. Finally, we have correlated having an ulcer-free leg with the number of incompetent perforating veins.

The AVVQ index at 10 years was tested for normality of distribution by the Shapiro–Wilk test. The Student's t-test was used to test differences in AVVQ between treatment groups. A p-value of less than 0.05 was considered to be statistically significant. Data analysis was performed by SPSS-pc version 22.

Results

Original trial patient characteristics are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Original trial patient characteristics.

	Surgical (n = 94)	Conservative
	n = 94	n = 102
Gender; male: female	2:3	2:3
Age (mean in years ± SD)	64 (±15)	68 (±14)
Open ulcer (median in days)	120	120
First time ulcer	31 (33%)	36 (35%)
Deep vein incompetence (DVI)	49 (52%)	57 (56%)
Diabetes mellitus (DM)	6 (7%)	17 (17%)
Deep vein thrombosis (DVT) in medical history	29 (31%)	31 (30%)
Localization: medial	70 (74%)	62 (61%)
Ulcer size (median mm2; range)	225 (4–7800)	260 (1–27.000)

With a mean of 97 months (range 72–172 months) of follow up, 80 (41%) out of 196 legs could be inspected for a last follow up in 73 out of 170 patients (Table 2). Forty patients (23.5 %) out of the randomized 170 had deceased and 57 (33.5 %) could not be reached by our team. All 80 inspected legs were screened with DUS and the SF-36v2 plus AVVQ questionnaires were taken for 47 legs (59 %) out of 80 ones inspected (Figure 3). OPEN IN VIEWER

OPEN IN VIEWER

Table 2.

Long-term follow-up patient characteristics.

	Surgical (n = 94)	Conservative
	n = 45	n = 35
Gender; male: female	M = 19, F = 26	M = 15, F = 20
Age (mean in years ± SD)	67 (±13)	70 (±12)
	20 (44%)	11 (31%)
Diabetes mellitus (DM)	2 (4%)	2 (6%)
Deep vein thrombosis (DVT) in medical history (legs)	16 (36%)	12 (34%)
Active ulcer (C6) (legs)	12 (27%)	9 (26%)

Deep vein incompetence (DVI) (legs)

Results: Ulcer-free incidence

The main outcome parameter is having an ulcer-free leg, calculated with intention-to-treat rules by means of the last-observation-carried-forward technique. Overall, 96 (49.0%) of all 196 initially randomized legs turned out to be ulcer-free, for surgically treated ones this was 56 (58.9 %) out of 95 legs and for conservatively treated ones 40 (39.6 %) out of 101 legs (log-likelihood chi-squares: 7.38 by 1 df, p = 0.007, OR = 2.19, 95% CI: 1.24–3.88). So at 96 months after randomization chances of remaining ulcer-free for surgically treated legs significantly turned out to more than twice compared to those of conservatively treated legs. The other main outcome parameter is having an active leg ulcer (C6). Applying the same rules of imputation when missing data are concerned, overall 63 (32.1 %) of all 196 legs turned out have a current active ulcer, for surgically treated ones this was 27 (28.4%) out of 95 legs and for conservatively treated ones 36 (35.6 %) out of 101 legs (log-likelihood chi-squares: 1.17 by 1 df, p = 0.279, OR = 0.72, 95% CI: 0.39–1.31). So at 96 months after randomization chances of having ulcers for surgically treated legs seemed more than 0.7 times lower compared to those of conservatively treated legs, but this result was not statistically significant.

Results: Observed ulcer state after 10 years

Observed long-term follow-up healing rates were 95.6% (43 out of 45 legs) in the surgical group and 97.1% (34 out of 35) in the conservative group (NS). Observed recurrence rates were 48.9% for the surgical group and 94.3% for the conservative group (Figure 4). OPEN IN VIEWER

Results: Incompetent perforator veins

At 97 months the number of incompetent perforator veins was scored with DUS.

Twenty-one patients had no incompetent perforator veins, of which 6 (29%) were at risk of not being ulcer-free. Fifty-nine patients had 1 or more incompetent perforator veins, of which 52 (88%) were at risk of not being ulcer-free. The number of incompetent perforating veins appears to be a risk factor for not being ulcer-free (OR = 18.5, 95% CI: 5.42–63.70, p < 0.001, n = 80). The plotted location of the incompetent perforating veins did not show a pattern.

Results: Disease specific QoL (AVVQ)

At the last visit at 97 months after randomization, the AVVQ questionnaire was presented to the patients. Forty-two patients filled in the items concerning 45 legs. Average AVVQ at 97 months was 22.9 (sd = 10.9, range 2.63–45.0). The Shapiro–Wilk statistic had a p = 0.10, so the distribution was approximately normal. The Student's t-test between treatment groups had a p-value of 0.635, so no statistical difference could be found between the AVVQ of 22.0 (sd = 9.5, range 7.0–39.6, n = 21) for the conservative treatment group and the AVVQ of 23.6 (sd = 12.4, range 2.6–45.0, n = 24) for the surgically treated group.

Results: Relation with age and comorbidity

Using logistic regression analysis on having an ulcer-free leg using age, gender and DVI as possible predictors or confounders showed that controlled for age, treatment group (surgical or conservative) still had a statistically significant effect on having an ulcer-free leg. Net odds-ratio was 2.38, 95% CI: 1.33–4.28, backward elimination LL chi-squares = 0.003.

If—contrary to the intention-to-treat principle—having DM was used as a predictor or confounder in logistic regression analysis on having an ulcer-free leg, the original effect of treatment group appeared to be corroborated within a model using age and DM as other predictors (OR = 2.73, 95% CI: 1.41–5.31, n = 163). In violation of the protocol there were 33 (16.8%) of the DM data missing, 14% in the conservatively and 20% in the surgically treated group.

Discussion

This is the only study that reports up to 97 months of follow up after a randomized trial on superficial and perforator vein surgery in venous ulcer patients. Nelzen et al. ¹¹ published a study on SEPS combined with superficial venous surgery with a follow up of 60 months. ¹¹ They described a recurrence rate of 18% after 5 years. Both healed (C5) and active ulcers (C6) were included in the trial.

To distinguish the role of perforator vein surgery from surgery of the superficial system, the same group conducted a randomized trial of saphenous surgery with or without subfascial endoscopic perforator surgery in patients with a venous ulcer. ¹² There was no short-term (1 year) clinical benefit from adding SEPS to saphenous surgery in patients with venous leg ulcers.

Surgery of incompetent perforating veins in the management of venous leg ulceration has been described with good results, but the exact role is still unclear. ¹³ A complete performed eradication of incompetent perforators lowers the venous ulcer recurrence rate significantly, indicating the clinical importance. ⁹

In this series, we observed healing rates of 95.6% (43 out of 45 legs) in the surgical group and 97.1% (34 out of 35) in the conservative group during a median follow up of 97 months (range 72–172 months). Observed recurrence rates with the same median follow up are 48.9% for the surgical group and 94.3% for the conservative group. Both high ligation and stripping as the SEPS-procedure was performed in the surgical arm of this study.

To distinguish the additional effect of the SEPS procedure to the high ligation we scored the incompetent perforators in the study group. Twenty-one patients had no incompetent perforating veins, of which 29% was not ulcer-free. Fifty-nine patients had one or more incompetent perforating veins. Of this group 88% was not ulcer-free. This indicates the importance of incompetent perforating veins in the development and treatment of leg ulcers. Patients with venous leg ulcers should be referred to the vascular surgeon for treatment of both incompetence of superficial and perforating veins.

Another possible cause for ulcer recurrence is deep venous obstruction and/or occlusion. Deep venous pathology has been seen as a result of an increase in diagnostic and therapeutic options in the last decade, uncovering previously unknown causes such as iliac vein compression, post-thrombotic changes and anatomic variations leading to increased venous pressure and impaired venous flow. These factors can impair ulcer healing and increase recurrence rates. More research is needed to show the correlation between deep venous obstruction and venous ulceration.

Treatment of residual incompetent perforating veins by e.g. PAPS (percutaneous thermal ablation techniques) or nonthermal ablation techniques can possibly improve outcome. As the Dutch SEPS study group, we concur with the Swedish study group that fifty percent reduction in venous ulcer prevalence is achievable. ¹⁴

Conclusion

The surgically treated group has a significant higher (p = 0.007) chance (2.19×) of being “healed” (incidence of “ulcer-free”) after 97 months follow up. Treatment group does not significantly (p = 0.279) influence the actual ulcer state. Statistical correction for age, gender, and a history of DVI does not influence the correlation between treatment group and long-term disease state. QoL is not significantly different between the treatment groups. The number of incompetent perforator veins appears to be a significant predictor for long-term ulcer status.