Ultrasound-guided foam sclerotherapy for chronic venous disease with ulcer. A prospective multiple outcome cohort study *

Abstract Background Chronic Venous Disease (CVD) is the main cause of chronic leg ulcers. Varicose veins are the most frequent cause of venous leg ulcers (VLU). 50.9% of Brazilian women have varicose veins and ulcer prevalence is as high as 4%. Ultrasound-guided foam sclerotherapy (UGFS) is a low-cost treatment option for varicose veins. Objectives To analyze UGFS outcomes in patients with VLU. Methods Prospective consecutive single center cohort study. Patients with great saphenous vein (GSV) reflux and VLU were treated and followed-up for 180 days. The following were studied: quality of life (QoL), disease severity, healing, and elimination of GSV reflux. The Aberdeen questionnaire, a venous clinical severity score, and Duplex scanning (DS) results were analyzed. Results 22 patients aged 35 to 70 years were treated. There was improvement in quality of life, disease severity reduced, and ulcer diameter reduced (p < 0.001; ANOVA). 77.27% of VLU healed completely (95%CI: 59.76-94.78%). The dimensions of 20/22 VLU reduced (90.91%; 95%CI: 78.9-100%). GSV reflux was eliminated in 63.64% (95%CI: 43.54-83.74%). Men had greater QoL benefit and women had more complications. There were no severe complications. The VLU that had healed completely at the end of the study were smaller at baseline than those that did not completely heal. The GSV that were completely occluded at the end of the study were smaller at baseline than those that were not completely occluded (p < 0.05; Mann-Whitney). Conclusion The results suggest that most patients benefited from UGFS.


INTRODUCTION
Chronic Venous Disease (CVD) is the main cause of lower limb chronic ulcers. 1 Primary varicose veins with great saphenous vein (GSV) reflux is the most frequently identified condition in patients with venous ulcers (VU). 2 It is estimated that 6% of patients with varicose veins will develop VU at some point in their lives. 3 In Brazil, 50.9% of women and 37.9% of men have varicose veins and the prevalence of VU is as high as 3.6%. 4 From 2009 to 2013, the Brazilian government spent about 125 million US Dollars on treatment and social security benefits for patients because of varicose veins and its complications. 5,6 Specialized services improve rates of VU healing and reduce ulcer prevalence. 7,8 Varicose vein surgery is considered economically advantageous for reducing VU recurrence. 9 Treatment of primary CVD through compression therapy does not solve venous reflux. Compression is associated with healing of 65% of ulcers within 24 weeks, but up to 70% of the patients have VU recurrence when they end treatment. [10][11][12][13][14] Surgery improves quality of life in varicose veins patients. 15 Resection of insufficient GSVs reduces recurrence of VU. 12,16,17 Foam sclerotherapy can provoke occlusion of over 80% of GSVs treated. 18,19 Brittenden et al. reported 54% occlusion of veins in a large randomized trial. 20 Divergent results are probably a reflection of heterogeneous anatomical and clinical characteristics in the populations studied. Furthermore, there is not merely one uniform foam sclerotherapy method. Controlled trials identify lower reflux cessation rates and higher reintervention rates after sclerotherapy when compared with surgery and thermoablation. 20,21 Controlled trials also demonstrate that frequencies of adverse effects are similar after surgery and thermoablation. 21,22 Notwithstanding the inferior result in terms of reflux cessation, patients treated with sclerotherapy present less pain, better quality of life, and a faster return to daily activities than patients treated surgically. 19,21,23 Cost analyses favor sclerotherapy over other methods. 21,24 OBJECTIVES To investigate quality of life, ulcer healing and venous disease severity after foam sclerotherapy treatment and investigate factors related to main outcomes.

METHODS
Prospective and consecutive single center cohort with systematic follow-up and data collection protocol. All patients with non-healing venous ulcers and primary reflux of the GSV treated at the Hospital run by the Pontifícia Universidade Católica de Campinas (PUC-Campinas), Campinas, SP, Brazil, from June 2015 to June 2016. The study was approved by the ethics committee at the Pontifícia Universidade Católica de Campinas (PUC-Campinas). Only patients who were able to understand the risks and benefits and agreed to take part were included. We defined pathological reflux as duration greater than half a second and extension exceeding 20cm on a great saphenous vein Duplex scan (DS) performed standing up. 25 MedisonSonoace Pico and Siemens Acuson X300 PE equipment were used. A certificated researcher performed all examinations. We defined venous ulcer as an open wound of at least 1cm diameter on the skin of the leg or foot, in areas with venous hypertension. Clinical examination was confirmed by DS 26 (Figure 1).
The exclusion criteria were: • Peripheral arterial obstructive disease; • Venous thromboembolism • Foramen ovale; • Pregnancy; • Allergy to polidocanol; • Infection in the lower limbs; • Deep vein obstruction (by Duplex scan); • Age less than 18 years; • Patients who were unable to walk or were confined to bed; • Patients with unfavorable clinical conditions.
The following were evaluated: quality of life, venous disease severity, ulcer size, and venous status at baseline before treatment and at 60 days and 180 days after treatment. Versions of the Aberdeen questionnaire for venous disease (AQ) and the venous clinical severity score (VCSS) translated and validated for our language were used. [27][28][29] Ulcer healing was evaluated by the largest ulcer diameter. Venous status was characterized by DS. Clinical, anthropometric, anatomical, and social data were collected ( Figure 2).
In each session, 10 ml of foam was injected straight into the GSV with a single puncture. Puncture and foam progression were monitored by DS. Foam was produced by mixing 8ml of room air with 2ml of 3% polidocanol solution (Victalab).
Elastic bandages were applied after injections. Patients were instructed to keep the compression bandages on for 24 hours and wear 20 to 30 mmHg thigh-high compression stockings after removing the bandages. Patients were told to maintain their daily habits. Up to four treatment sessions were performed if there was persistent great saphenous vein reflux and if ulcer dimensions remained unchanged.

Data analysis
Quality of life (assessed by AQ), severity score, and ulcer diameters evaluated at pre-treatment baseline and 60 days and 180 days after treatment were compared using repeated measures ANOVA on ranks (non-parametric method). Interactions between clinical variables and outcomes were also evaluated by ANOVA on ranks. Results for QoL, VCCS, and VU healing were classified as: improvement, no improvement, or worsening. Frequencies were described at each follow-up point, 60 and 180 days after treatment. Results of the great saphenous vein treatment were classified as complete or not complete occlusion and presence or absence of residual reflux. Results classified by categories were expressed as frequencies for each follow-up point. Patients were grouped according to outcome (VU healing, occlusion and cessation of reflux in the GSV, and occurrence of adverse effects). Continuous variables were compared using the Mann-Whitney test and categorical variables were compared using Fischer's test. A 5% statistical significance level was adopted. The Statistical Analysis System for Windows 9.4 (SAS Institute Inc., Cary, NC, USA) and Minitab 16 were used for statistical analysis.

RESULTS
Twenty-two patients were treated consecutively. There were no losses. There were 42 treatments (1.9 +/-0.9 application per patient). Most patients were obese or overweight, BMI varied between 23 and 45 (30 +/-1), and only 3 patients (14%) had normal weight (BMI between 18.5 and 24.9). Ten patients (45%) did not have comorbidities whereas twelve (55%) patients did (Tables 1 and 2). Thirteen patients (59%) had no adverse effects. There were eleven adverse effects in nine patients, all of them female. Among the women, 60% had at least one adverse effect and 20% exhibited staining along the path of the vein treated. All adverse effects were mild and treated on an outpatient basis. There was a higher proportion of women in the group that presented at least one adverse effect (P = 0.017; Fisher). Other anatomical, clinical and social variables did not differ between these groups (with or without adverse effects) ( Table 3).  There was an interaction between the number of UGFS sessions and quality of life (p-value = 0.0253). The worse the QoL assessed on day 60, the more treatment sessions the patients needed (ρ = 0.5449, p-value = 0.0087, Spearman). There was an interaction between gender and QoL progress (p = 0.0309). There were no differences in men's and women's pre-treatment QoL (p-value = 0.9438; Mann-Whitney) or day 60 QoL (p = 0.1805; Mann-Whitney). On day 180, men's QoL was better than women's (p=0.0074; Mann-Whitney). Both genders exhibited quality of life improvement over time (p <0.001; ANOVA); however, women's QoL did not change from day 60 to day 180 (p = 0.0884). Among men, there were improvements at each evaluation (p <0001 baseline to day 60, p-value <0.001 baseline to day 180, p = 0.0393 from day 60 to day 180). The variables body mass index (BMI), GSV reflux pattern, VU diameter, GSV diameter, reflux in other venous territories, age, comorbidities, side affected by VU, educational level, and occupation did not have any significant influence.  VU diameters reduced over time (p <0.001; ANOVA); however, they did not vary significantly between day 60 and day 180 (p-value = 0.8903). The variables BMI, GSV reflux pattern, GSV diameter, reflux in other venous territories, age, comorbidities, side affected by VU, educational level, and occupation did not exhibit any significant influence. Ulcers that healed had smaller diameters at baseline than those that did not completely heal (p-value = 0.0336 on day 60 and p-value = 0.0115 on day 180; Mann-Whitney)

GSV occlusion and reflux
Initially, all GSVs had reflux. On day 60, reflux had been eliminated in 15 (Tables 8 and 9).

DISCUSSION
Our protocol allowed selection of individuals considered to be poor candidates for surgery and who would benefit from elimination of venous reflux. Consecutive selection according to inclusion and exclusion criteria allowed the sample to be clinically homogenous and also enabled exclusion of patients at high risk of complications. No sample size calculations were conducted prior to the study. This was because multiple outcomes would be studied that would have different frequencies and would need samples of different sizes. Considering a healing rate of 91.3%, as reported by Campos et al., our study would have needed 123 cases to attain a 95% confidence interval (CI). 22 It is important to note that small sample sizes can lead to unreliable results. Our cohort study is a longitudinal research project that aims to establish a causal link between events. It does not enable efficacy to be determined. Efficay testing would require a comparative study with control group.
We studied patients who spontaneously sought treatment and there was no active screening for these patients. It is likely such patients were more symptomatic and had worse quality of life and so presented great improvement in this regard.
A non-parametric method was used to compare VU diameter, VCSS, and AQ results. Non-parametric methods are best suited to avoid errors in small patient samples with data that are not normally distributed.
The predominance of overweight women (68%), obesity (86%), comorbidities (55%), low educational level (77% did not complete high school), and informal employment is similar to the population studied to validate the Brazilian version of the Aberdeen questionnaire. 27,28 Adverse effects Wright et al. 19 reported 11 cases of deep vein thrombosis (DVT) (incidence of 5.3%) when performing UGFS with up to 60ml of foam. European consensus guidelines recommend that foam volume should be limited to 10ml per session. 30 We did not identify any severe adverse effects (AE). The most frequent AE was painful phlebitis in 4/22 patients (18.18%; 95%CI: 20.06-34.3%). 2/22 patients (9.09%; 95%CI: 0-21.10%) had venous path hardening and hyperemia without pain. In a randomized trial with patients with GSV reflux but without VU, Thomasset et al. identified superficial phlebitis in 18% of patients, pain in 14%, and skin staining in 28%. We observed skin staining in 3/22 (13.64%; 95%CI: 0-27.98%). Women had more adverse reactions than men, which is similar to the data reported by Thomasset et al. 31 In a systematic review of 69 studies, Jia et al. reported the following frequent complications: 4.7% phlebitis, 17.8% skin staining, and 25.6% local pain. 32 These complications rates are similar to those identified in the present study.

Quality of life
We identified differences between men and women in terms of QoL improvement during the interval from 60 to 180 days after treatment. QoL is  dependent on cultural factors and factors related to patients' expectations. We believe residual varicose veins, incomplete VU healing, and occurrence of skin staining may be the causes of worse QoL in women compared to men at the end of the study. All of the patients who had VU that were not fully healed at 180 days after treatment were women. There was also a higher proportion of women in the group who had AE. Other outcomes in our study had similar results for both genders. Occurrence of residual varicose veins after treatment was not studied. The AQ aesthetic assessment would indicate whether women's poor QoL was related to cosmetic factors.

Clinical severity
Higher severity related to larger VU was to be expected, since they are scored higher on the VCSS. The influence of age is also compatible with the progressive nature of CVI. Scott et al. observed that patients with CVI were older than those with varicose veins without CVI. 33 Large cross-sectional studies, point to a higher prevalence of varicose veins in women, but do not indicate a higher prevalence of severe CVI among women. Scott et al. identified that 60% of VU patients were men while there was a predominance of women among those with varicose veins. However, multivariate analysis revealed that CVI patients were older, male and obese. The small sample in our study may be responsible for inconsistent results and since in our study we recruited patients treated at the hospital, our sample may not reveal the actual prevalence but, rather, reflect the greater demand for treatment among women. In the Edinburgh study, the severe forms of CVI did not show a predilection for gender after adjustment for age. 34

Ulcer healing
We identified complete healing of 77.27% (CI 95% 59.76-94.78%) of VU and 90.90% (CI 95%78.89-100%) of VU diameters reduced during the study. Before treatment, the VU initially had dimensions ranging from 1 to 14cm (mean of 4.11cm and median of 3.00cm). Barwel et al. identified healing of 65% of VU after treatment by saphenectomy or compression. 12 VU healing rates after UGFS range from 71 to 92%, but there are few randomized studies comparing sclerotherapy to clinical treatment and samples are also small. 22 VU open for more than 12 months, VU with an area of more than 6 cm 2 , lipodermatosclerosis, previous DVT, and bleeding were all associated with poorer cicatrization. 37 Cabrera et al. also report worse results for healing of chronic VU and when there is deep venous reflux. 36 The hypothesis that UGFS may be equivalent to surgery or thermoablation for averting recurrence of VU should be tested with long-term, controlled clinical trials with large groups. It is probable that USGFS is inferior for averting recurrence of VU, since foam sclerotherapy has higher rates of recanalization of treated veins and relapse of reflux.  20 We identified that GSVs that were completely occluded had smaller baseline diameters than GSVs that remained patent or partially patent, consistently, at both 60 days after treatment and at 180 days. Myers et al. prospectively studied 1189 sclerotherapy sessions performed in 489 patients with varicose veins, treating 454 GSVs. The occlusion rate after a single sclerotherapy session was 53.1%. The occlusion rate of non trunk veins was higher than for GSVs. Veins with a diameter greater than 6mm had worse results than those with diameters of 5mm or less. The best results were observed in patients older than 70 years when foam sclerotherapy was used and when foam volume greater than 12 ml and containing a higher concentration of sclerosants was used. 25

CONCLUSION
The benefits in terms of quality of life, ulcer healing, and reduction of the severity of the venous disease without serious complications suggest that sclerotherapy is a valid option.