Is there a continuing role for traditional surgery?

Introduction

Open surgical treatment has been used for a century to treat uncomplicated varicose veins. As time has passed, techniques have been modified, yet the pathophysiological basis for the surgery remains – namely ligating and disconnecting either the great saphenous (GSV) or small saphenous vein (SSV) (or perforator) at its junctional connection with the deep venous system. As such, these techniques remain the gold standard against which contemporary endovenous techniques are to be compared. Despite the recent NICE guidance that advises the use of endovenous treatment (radiofrequency ablation (RFA) and endovenous laser therapy (EVLA) followed by ultrasound-guided foam sclerotherapy (UGFS)) ahead of open surgery, there remains a definite role for surgery in the management of varicose veins. ¹ This article will highlight the current debate with regard to the role of open surgery and suggest situations where open surgery may be considered to remain the “first line” treatment. A lot of the available evidence that has driven the transformation of care towards endovenous treatments is of good quality yet published by enthusiasts and as such the available data may be criticized as having a lack of general applicability to the vascular surgical community as a whole.

Open venous surgery

The current debate around the effectiveness of surgery when compared to endovenous treatment involves the early post-operative morbidity suffered by the patient and the longer term issue of recurrence.

It is widely recognised that patients suffer less post-operative discomfort and return to daily activities quicker when undergoing endovenous treatments. Furthermore, the ability to perform these treatments without the need for a general/regional anaesthetic also has significant benefits. The meta-analysis by Murad et al., as part of a Society for Vascular Surgeon/ American Venous Forum commissioned review, studied over 8000 patients undergoing all forms of intervention (including open surgery) for primary varicose veins. ² All treatments were well tolerated without significant periprocedural adverse events. However, what was evident from this study was that each form of treatment has its own local complication profile. As expected, open surgery had wound-related problems, haematoma formation and sensory nerve injury but the heat-delivered treatments (EVLA, RFA) had their own associated complications profile which included bruising and erythema, thermal skin injury with blistering, phlebitis and sensory nerve injury. UGFS was associated with significant risk of skin pigmentation as well as headaches, migraines and visual disturbances.

What of the issue of vein recurrence? With open surgery, the occurrence of recurrent varicose veins has been suggested to be in excess of 50% by Winterborn et al. who undertook long-term (>10 years) surveillance of patients enrolled into a randomised controlled trial. ³ Such recurrence rates are still associated with high levels of long-term patient satisfaction with the original venous surgery procedure performed. As such, the development of recurrent veins itself does not equate with patient dissatisfaction. A recent study by Nelzén and Fransson with a median follow-up of 12 years confirmed the results from previous studies with a poor correlation seen between duplex detected recurrences and symptoms/cosmetic result. ⁴ Overall patient satisfaction rate was high at 85% and did not seem to deteriorate over time. No endovenous studies have reported long-term results as far out from intervention as the open studies, yet in the meta-analysis from Murad et al., the authors suggested from their results that when surgery was compared with all endoluminal ablation therapies, surgery was associated with a non-significant reduction in the risk of varicose vein recurrence (RR 0.63; 95%, CI 0.37–1.07). ²

Furthermore, much of the longer term data on recurrence following open surgery for primary varicose veins dates back well over 15 years ago. Over this time, there has been a significant change of practice with regard to the assessment of patients with varicose veins, specifically with regard to the role of duplex assessment. Prior to the routine use of duplex for the examination of superficial venous disease, a number of vascular surgeons would have solely used clinical examination with or without hand-held Doppler (HHD) assessment to determine the sites of venous reflux. Such techniques deliver a suboptimal picture of the venous reflux. Indeed, comparison of duplex with the use of HHD in the evaluation of primary varicose veins has shown sensitivity of the HHD technique of between 70 and 80% when assessing the saphenofemoral junction (SFJ) and saphenopopliteal junction (SPJ) for reflux. ⁵ In this study by Mercer et al., the open venous surgery planned using HHD imaging alone would have left-residual sites of reflux in 24%. As such, there is an argument to say that current practice with routine duplex scanning even when considering open surgery would reduce the recurrence rates seen in the past with open surgery. Furthermore, vascular surgery in the UK has now become its own subspecialty with varicose vein treatment now exclusively performed by vascular surgeons rather than generalists. Such a step will only lead to improved overall patient outcomes.

Surgical techniques have developed over time. These include endothelial closure techniques at the time of SFJ ligation and closure of the cribiform fascia which have all been postulated to reduce neovascularisation rates.^6,7 Stripping of the great saphenous vein (GSV) is now commonplace as it is felt to reduce recurrence by preventing neovascularisation from occurring and reconnecting the junction with the residual GSV in the proximal thigh. Data from randomised trials for such a technique is, however weak. ⁸ Inversion stripping and tourniquets have not been shown to provide any major clinical benefit.^9,10 Complication rates can also be mitigated. Although wound-related complications do occur with open surgery, there is now good level-one evidence that a single pre-operative prophylactic antibiotic regime significantly reduces wound-related complications in the early post-operative period. ¹¹ Furthermore, the installation of local anaesthetic with adrenaline into the stripper track has been shown to reduce tract haematoma rates. ¹²

Comparison with endovenous techniques

An increasing number of high-quality randomised studies has been performed comparing open surgery with the multiple available endovenous techniques, specifically RFA, EVLA and UGFS. Given the available data, a number of meta-analyses have been performed but in depth analysis has been difficult to perform in part due to the variable primary end points used in the randomised controlled trials. Such outcomes have included duplex derived recurrence, clinical recurrence, symptom alleviation (assessed using questionnaires including the Aberdeen varicose vein questionnaire (AVVQ)) and quality of life (QOL – either generic or disease specific).

Proponents of endovenous techniques suggest that the benefits of such techniques include fewer complications, quicker return to work, improved QoL scores, reduced need for general anaesthesia and equivalent recurrence rates.

The most recent published analysis of the available data is the recent Cochrane review, which examined a total of 13 randomised controlled studies resulting in a combined total of 3081 patients. ¹³ Three studies compared UGFS with surgery, eight compared EVLA with surgery and five studies compared RFA with surgery. The overall conclusion was that currently, the evidence suggested that UGFS, EVLT and RFA were at least as effective as surgery in the treatment of the GSV. However, incompatibilities between trials and different time point measurements for outcomes meant that the evidence lacked robustness. No major difference was seen in outcome when comparing UGFS and RFA with open surgery and the only significant differences seen were that neovascularisation rates and technical failures were reduced in those patients undergoing EVLA. Within this meta-analysis, QoL scores and complication rates including assessment of pain were not amenable to meta-analysis; however, QoL scores increased similarly in all treatment groups and complications were generally low. Pain was also similar between the treatment groups. The results suggest therefore that surgery is still a valid form of treatment for primary GSV varicose veins.

This meta-analysis did not, however, include the CLASS (Comparison of LAser, Surgery and foam Sclerotherapy as a treatment for varicose veins) trial. ¹⁴ This was a randomised controlled trial involving 11 centres in the UK that assessed the effect of treatment in 789 patients with primary varicose veins comparing the outcomes of UGFS, EVLA and open surgery. Primary outcomes at six months were disease-specific and generic QoL, as measured on several scales. Secondary outcomes included complications and measures of clinical success. The frequency of procedural complications was similar in the foam group (6%) and the surgery group (7%) but was lower in the laser group (1%) than in the surgery group. Measures of clinical success were similar among the groups but successful ablation of the main trunks of the saphenous vein was less common in the foam group than in the surgery group. The authors concluded that all treatments had similar clinical efficacy but complications were less frequent after laser treatment and ablation rates were lower after foam treatment.

A similar study performed on 580 legs by Rasmussen et al. showed that all forms of treatments (RFA / UGFS / EVLA / open surgery) were equally efficacious. ¹⁵ The technical failure rate was highest after foam sclerotherapy but both RFA and foam were associated with a faster recovery and less post-operative pain than EVLA and stripping.

Like all randomised controlled trials, the relatively artificial environment generated by such trial design limits it general applicability to the population as a whole. Inclusion criteria for the CLASS study included primary small and great saphenous vein (GSV) with a vein diameter of >3 mm yet <15 mm. In total, only 48% of the patients screened were eligible for the study and of these, only 24% of the eligible patients agreed to take part in the study. Furthermore, even in the randomised study of EVLA vs. open surgery for GSV reflux by Carradice et al., where the inclusion criteria were slightly more lax, out of a total of 772 patients assessed for suitability, 442 did not meet the inclusion criteria. ¹⁶

These studies fail to elaborate upon reasons why patients were not eligible but potential reasons will include (a) anatomical factors and (b) patients’ preference. Why then may an open surgical approach be more appropriate in patients with primary varicose veins?

In a comprehensive study of a consecutive series of over 2000 patients attending for venous duplex, García-Gimeno et al. found that 11% had reflux within the anterior thigh vein, 12% had result from perforating veins and 8% had pure non saphenous vein reflux, equating to over 30% patients who may not be optimally treated using endothermal techniques. ¹⁷ Furthermore, patients with truncal reflux in either saphenous veins may not be suitable for endothermal treatment due to either significant tortuosity or the vein in question being too superficial and thus putting the overlying skin at risk of staining/burns. Again advocates for endovenous treatment would suggest that a number of these patients would be amenable to non-surgical treatment and specifically UGFS with high-quality results. Yet, UGFS is associated with increased incidence of phlebitis and skin staining that although often temporary may be acceptable in an older cohort of patients but is often not in younger patients.

The study by Garcia-Gimeno also showed that reflux within the anterior accessory saphenous vein (AASV) as the source of varicose veins is frequently seen. ¹⁷ This vein is often tortuous and only the most proximal aspect of this sits beneath the fascial layer, which limits the role of heat-delivered treatments. Furthermore, there is often a large burden of varicose veins that course across the thigh and down the lateral aspect of the leg. There are reports of successful treatment of the AASV with the more flexible EVLA laser fibres, yet such studies report higher failure rates in the AASV treatment group as compared with GSV treatment group when treated with EVLA. ¹⁸ Given this, treatment of AASV reflux will often only be amenable to open surgery or UGFS and given the burden of veins often seen, the complication profile of UGFS and the need for repeat procedures, open surgery may be more acceptable to patients. Furthermore, following the advent of thermal treatment, in those patients with pure SJF/AASV reflux with a normal GSV, the GSV could potentially be left in situ following the ligation of the SFJ and its tributaries in this way removing the complications of GSV stripping yet allowing endovenous treatment of the GSV if in the longer term GSV reflux was to develop.

The major benefit of endovenous treatments is that they can be delivered in an office setting using local anaesthetic/no anaesthetic techniques. This allows treatment of the truncal varicosities but limits the ability of the surgeon to treat varicose tributaries, which ideally requires a theatre-based environment to limit the infective complications that can be seen following phlebectomies. The need for further treatment for remaining varicose veins is a recognised limitation associated with endothermal techniques both from the point of view of patient acceptance and also a cost issue due to the often needed repeated out-patient appointments and duplex scans. Randomised controlled trials of concomitant vs. delayed phlebectomies suggest that between 30 and 50% of patients who did not undergo concomitant treatment required further treatment for their varicosities. ¹⁹

Furthermore, the role of ambulatory phlebectomies is limited by the amount of residual tumescence available and thus is really only suitable for those patients with low-volume calf varicosities. Therefore, in those patients with a large burden of varicosities associated with truncal reflux open surgery is likely to be an optimal treatment option that will deliver a comprehensive veins treatment in a single visit. This is specifically true in younger patients who are fit for general anaesthesia and who wish all treatment to be delivered in a single sitting.

Again the endovenous enthusiasts would suggest that such patients as described above would be able to be treated by either (a) a total or (b) near-total endovenous solution. This could be delivered by (a) UGFS combined with an endothermal solution for the truncal reflux in one sitting or (b) endothermal treatment of the truncal reflux with concomitant phlebectomies under a general anaesthetic.

Some would argue that such patients could/would be best served with UGFS for the calf varicosities but this would likely result in episodes of phlebitis occurring with associated skin staining, which is likely to be an unacceptable outcome in younger patients. Furthermore, the likely need for repetitive treatments in such patients is unattractive to younger patients at work and these patients would likely wish to have all their venous disease treated in one sitting that could be delivered using surgical techniques requiring a general anaesthesia.

Endothermal treatment of the truncal reflux with concomitant phlebectomies under general anaesthetic is likely to be one of the most optimal ways to treat patients. However, this is likely to be the least cost-effective way to treat varicose veins given the costs of both the equipment required for endovenous treatment and the theatre costs required for open surgery.

Cost- effectiveness

Given that the current evidence suggests that there is little to choose between surgery and the minimally invasive techniques in terms of efficacy or safety the relative cost of the treatments becomes a factor to consider when considering provision of service.

The REACTIV trial was published in 2006 with the primary objective of establishing the cost-effectiveness of surgery and sclerotherapy for the treatment of varicose veins. ²⁰ The results from this study suggested that treating superficial venous disease was cost-effective when compared to conservative treatment. Furthermore, the results suggested that standard surgical treatment of varicose veins (saphenofemoral ligation, stripping and multiple phlebectomies) was a cost-effective treatment for varicose veins, with an incremental cost-effectiveness ratio (ICER) well below the threshold normally considered appropriate for the funding of treatments within the NHS.

The REACTIV study was published before sufficient clinical data were available with regard to endovenous treatments. One of the major perceived benefits associated with endovenous treatments is that of reduced costs by delivering treatments in an out patient/day surgery based environment without the need for anaesthetic support. However, the heterogeneous nature of outcome measures used in the endovenous trials means that this benefit has not been fully supported by the trial evidence.

In an attempt, therefore, to assess cost-effectiveness, Gohel et al. used a Markov model constructed to compare costs and quality-adjusted life years (QALYs) for the treatment of GSV reflux. ²¹ The ICER for UGFS (vs. conservative care), EVLA (vs. UGFS) and RFA (vs. EVLA) were £1366, £5799 and £17,350 per QALY, respectively, and as such, all treatments were determined as being cost-effective based upon the NHS threshold of £20,000 per QALY. Interestingly, the ICER for traditional day case-based surgery was £19 012 when compared to RFA and as such this would also suggest that this is a cost-effective based treatment strategy for primary GSV reflux. UGFS has without doubt the lowest initial costs but the cost- effectiveness is limited by the requirement for repeat intervention.

Further attempts have been made to determine cost-effectiveness based upon the available evidence. A recent network meta-analysis with exploratory cost-effectiveness modelling was performed to assess the effectiveness of minimally invasive techniques compared with other treatments, principally surgical stripping. ²² Threshold analysis indicated that EVLA and RFA might be considered cost-effective, if their costs were similar to those for surgery but the findings of this study were subject to a number of uncertainties, specifically the risk of bias variation in reported costs.

Finally, the recent cost-effectiveness arm of the CLASS study sheds some more light on this subject. ²³ The investigators found that compared with surgery at six months, UGFS and EVLA reduced mean costs to the health service by £655 and £160, respectively. When additional overhead costs associated with theatre use were included, these cost savings increased to £902 and £392, respectively. UGFS produced 0·005 fewer QALYs, whereas EVLA produced 0·011 additional QALYs. However, the authors were only able to extrapolate the data to five years and on this basis suggested that EVLA had the highest probability of being cost-effective. We await the true 5-year data to confirm these findings.

Training in venous surgery

Lastly, there is the issue of training surgeons to undertake endovenous techniques. Scurr et al. undertook a web-based questionnaire study of 145 vascular surgery trainees within the UK. ²⁴ They received a response rate of 85%. Of the 123 trainees who responded, 78% received no venous duplex training, and experience with endovenous techniques was limited and variable. No experience of EVLA or RFA was reported by 39% and 67% of trainees, respectively. Experience and/or training with UGFS was limited to <40%. Many of those reporting no experience with endovenous ablation techniques were within the final two years of their training. A similar type of study was repeated two years later, which showed no experience of EVLA, RFA or UGFS in 33%, 49% and 46% of trainees, respectively. ²⁵ These are obviously major issues if endovenous treatments are to become first-line treatment with more time needing to be invested in the training of our vascular surgeons of the future.

Small Saphenous Vein Surgery

Unlike GSV surgery, there is a lack of high-quality data on the management of small saphenous vein (SSV) reflux. The anatomical variability of the SPJ allied with the increased complexity of operating within the popliteal fossa make surgery on the SPJ/SSV more challenging. A cadaveric study by Balasubramaniam et al. confirmed this significant anatomic variation and suggested that the risk of damaging the peroneal nerve during saphenopopliteal ligation may be higher for SPJs located above the lateral femoral epicondyle because of the proximity of the two structures. ²⁶ Such risk of nerve injury has been investigated by Atkin et al. who reported a peroneal nerve injury approaching 5% (three patients) in a consecutive series of patients undergoing SPJ ligation. ²⁷ One patients’ neurological symptoms resolved within 24 h but the other two patients eventually had symptom resolution by six months. This is often due to excessive retraction within the popliteal fossa causing a neuropraxia of the peroneal nerve. Other associated complications include sural nerve damage and wound herniation of popliteal fat through the fascia.

The difficulty in interrogating the literature on SSV surgery is that the majority of studies merge data on treating both the GSV and SSV. A study by Carradice et al. attempted to redress this in a retrospective study and confirmed that the morbidity associated with SSV incompetence is greater than suggested by its clinical severity and responds differently following treatment to that of the GSV. ²⁸ Specifically with regard to the SSV, patients scored higher (worse) following an open surgical intervention on AVVQ (P = 0.045) than GSV sufferers but lower (better) following EVLA (P = 0.042).

When undertaking an open surgical approach to the SSV/SPJ, there are a number of potential treatment options. A study undertaken on behalf of the Joint Vascular Research Group before the advent of endovenous surgery showed that there was a varied strategy to deal with SSV reflux. ²⁹ Over 50% had purely SPJ ligation with just over a quarter of patients undergoing stripping of the SSV. This approach (stripping) resulted in a non-significant reduction in the presence of clinical recurrence with no difference in neurological sequalae (numbness). Stripping of the SSV also resulted in a significant reduction in SPJ incompetence (13% vs. 32%). A more recent study by Ikponmwosa et al. looked at the role of adjunctive pre-operative duplex skin marking given the anatomical variability of the SPJ. ³⁰ In a total of 90 limbs, the authors found that the SPJ was successfully ligated in 87 (96.7%) cases. Reflux was completely abolished in 51 (56.7%) cases but persisted solely in the small saphenous vein (SSV) in 32.2%. This highlights the role of pre-operative marking for the SPJ.

There is, however, little in the way of high-quality evidence to compare open surgery to endovenous techniques. Samuel et al. randomised 106 patients with unilateral SSV reflux to either EVLA or surgery. ³¹ They found EVLA to be superior to surgery in a number of outcome measures including duplex derived outcomes, pain and return to work/function and sensory disturbance. As such, they concluded that EVLA produced the same clinical benefits as conventional surgery but was more effective in addressing the underlying pathophysiology and was associated with less periprocedural morbidity.

Perforator surgery

As technology has developed, there has become an interest in the endovenous management of perforator vein reflux. Open surgical approaches have included formal ligation or the use of subfascial endosopic perforator vein surgery (SEPS). A meta-analysis by Luebke and Brunkwall compared the role of SEPS vs. other more conventional open surgery including the Linton procedure. ³² SEPS conferred a lower rate of wound infection and a shorter hospital stay and in the longer term was associated with a reduced rate of ulcer recurrence, although the rate of initial ulcer healing was not different between the groups.

Endovenous strategies include EVLA and UGFS with the management of perforators veins being aided by the newer radial fiber slim (ELVeS-radial-slim kit™) for the 1470 nm diode laser. ³³ This technique showed very impressive immediate technical and one month outcomes with minor morbidity. However, there is a lack of longer term data available using this technique and higher quality studies including randomised controlled trials are needed to compare open vs. endovenous techniques in perforator reflux in both the thigh and the calf.

Conclusion

There is no doubt that endovenous surgery will increasingly become first-line treatment for patients with symptomatic superficial venous reflux specifically in patients with SSV reflux. However, given the device-related limitations with endothermal techniques specifically with regard to adverse anatomical features allied to the poorer results of UGFS within the treatment of GSV reflux, there is without doubt still a role for more traditional open surgical techniques that can be delivered with good short and long-term outcomes and still being delivered in a cost-effective manner.