Various methods for treating varicose veins have been developed, and currently, several options are available. Surgery with high ligation (HL) and stripping of the great saphenous vein (GSV) or small saphenous vein has historically been the standard treatment for varicose veins. However, in recent years, an increasing number of studies have suggested that surgery with HL and stripping is not the best treatment option for saphenous veins. Minimally invasive methods are typically employed. In the European Society for Vascular Surgery guidelines for 2022, rather than HL with stripping or ultrasound-guided foam sclerotherapy, endovenous thermal ablation is recommended as the first-line treatment for patients with GSV incompetence [1]. Endovenous thermal ablation destroys damaged veins using a laser (endovenous laser ablation [EVLA]) or radio waves (radiofrequency ablation [RFA]) through an endovenous catheter [2]. The heat generated from lasers or radio waves can coagulate the blood in veins, occlude veins, and redirect the blood flow to functional veins [3]. Thermal ablation is recommended as the primary treatment for nonpregnant patients with valvular regurgitation and symptomatic varicose veins [4,5].

Physicians can accurately quantify the effectiveness of venous surgery using air plethysmography (APG). Physiological changes in venous function can be correlated with anatomical alterations shown by venous duplex scans [4,5].

This study aimed to compare the clinical and hemodynamic changes between surgical methods (HL with stripping, EVLA, and RFA) used for the treating GSV reflux.

A total of 3,045 cases of varicose veins over a period of 26 years (between November 1995 and May 2021) at the Samsung Medical Center were retrospectively reviewed. The inclusion criteria were as follows: (1) clinical, etiological, anatomical, and pathophysiological (CEAP) class C2 (varicose vein); (2) GSV surgery; and (3) surgery performed by a single operator. The exclusion criteria were as follows: (1) history of deep venous thrombosis (DVT) or congenital malformation, (2) previous venous surgery for sclerotherapy, (3) CEAP class C3 or C4, and (4) a follow-up period less than 6 months. A total of 976 limbs of 900 patients with primary varicose veins who underwent surgical treatment were included in the study.

Hemodynamic changes, such as venous volume (VV), venous filling index (VFI), residual volume fraction (RVF), and ejection fraction (EF), obtained using an APG (ACI300, ACI Medical Inc., San Marcos, CA, USA) before and after surgery were evaluated. Duplex scans were performed by sonographers with the American Registry for Diagnostic Medical Sonography certification. Patients were checked for valvular insufficiency of the deep system, superficial system, and perforators using manual calf compression in the standing position. Venous reflux was diagnosed when the cut-off value exceeded 500 ms. Additionally, the largest GSV size and depth from the skin were measured preoperatively in the supine position by a duplex examination. After surgery, the presence of occlusion in the GSV was confirmed. Partial occlusion was defined as any region not occluded with echo findings of occlusion in the GSV after surgery.

APG was performed preoperatively and 1 and 6 months postoperatively. The APG cuff was placed on the calf of the patient, with the leg raised at 45°. Baseline values were obtained after emptying the venous blood from the leg. The patients stood without any weight on the leg being tested. The VV, venous filling time, and VFI upon standing up from the supine position were measured. To confirm the status of the GSV, duplex scanning was performed on the day after surgery and at 1 and 6 months after surgery.

The surgical methods included (1) HL and stripping of the GSV and varicosectomy (the HL with stripping group), (2) EVLA and HL in the saphenofemoral junction (SFJ) with varicosectomy (the EVLA group), and (3) RFA and HL in the SFJ with varicosectomy (the RFA group). All surgeries were performed under spinal or general anesthesia.

EVLA was performed using a 1470-nm diode laser with a bare tip (DIOTECH, Busan, Korea). The GSV was identified and dissected through a small incision near the knees. A guidewire was passed through the GSV into the SFJ. The introducer sheath was passed over the guidewire to the SFJ. The GSV was divided and ligated by making a small incision around the SFJ where the sheath was palpable [6]. This technique has been used to prevent DVT and recanalization related to femoral vein ablation [7]. Tumescent normal saline solution was injected around the GSV, depending on the length of the leg. No anesthetic agent was added to the normal saline because all patients were already under general or spinal anesthesia. After guidewire removal, the laser fiber was pulled at a fixed speed of 2 mm/s and 6 W to deliver 30 J/cm to the GSV [6].

Phlebectomies were performed around or below the knee joint. The RFA catheter system ClosureFAST (Medtronic, Inc.) was used. A RFA catheter was advanced 2 cm below the SFJ. Concomitant high ligation of the GSV was performed around the SFJ through a tiny incision where the RFA catheter was palpable. Normal saline was injected as a tumescent solution along the above-knee GSV before ablation. After starting the energy generator, radiofrequency treatment was performed for 20 s each time. The temperature of the GSV was 120oC. This process was repeated once for the initial segment.

After EVLA or RFA, full-length compression bandages (Karl Otto Braun GmbH & Co., Wolfstein, Germany) were applied to the legs for 24 h. The day after surgery, the bandage was changed to knee-length compressive elastic stockings (20–30 mmHg). The device was then worn for up to 1 month after surgery. The patients were discharged 1 day after surgery with stockings.

All statistical analyses were conducted with IBM SPSS software (version 22.0; IBM Corp., Armonk, NY, USA). Data were presented as median and interquartile range (IQR). The Wilcoxon signed-rank test was used to evaluate changes in APG values after surgery. Differences in the results among the treatment methods were compared using the Kruskal–Wallis test. Statistical significance was set at p<0.05.

This study was approved by the Institutional Review Board of Samsung Medical Center. Informed consent was obtained from all the patients.

The patient characteristics are shown in Table 1. The EVLA, RFA, and HL stripping groups included 250, 139, and 511 patients, respectively. A total of 976 limbs were treated for varicose veins. EVLA, RFA, and HL with stripping was performed in 281 (28.8%), 161 (16.5%), and 534 (54.7%) limbs, respectively (Fig. 1).

Table 1 . Demographics and clinical features.

	EVLA	RFA	HL and stripping	p
Total patients	250	139	511
Total limbs	281	161	534
Median age (yr, IQR)	56 (47, 63)	60 (51, 66)	50 (43,57)	<0.001
Gender, female (%)	131 (51.6)	75 (54.7)	330 (64.8)	0.001
Preoperative maximal GSV diameter (mm, median, IQR)	57 (49, 66)	55 (45, 65)		0.111

EVLA: endovenous laser ablation, RFA: radiofrequency ablation, HL: high ligation, IQR: interquartile range, GSV: great saphenous vein..

Figure 1. Types of operation.

All patients showed decreases in the VV, VFI, and RVF and an increase in the EF after surgery, and these changes were significant when comparing the preoperative, 1- and 6-month postoperative values (Table 2).

Table 2 . Changes of venous hemodynamics in total patients during the follow-up period.

Parameter (median, IQR)	Preoperative vs. Postoperative 1 month				Preoperative vs. Postoperative 6 months				Postoperative 1 month vs. Postoperative 6 months
	Preoperative	Postoperative 1 month	Reduction rate (%)	p	Preoperative	Postoperative 6 months	Reduction rate (%)	p	Postoperative 1 month	Postoperative 6 months	Reduction rate (%)	p
VV (ml)	114.4 (88.3, 148.8)	86.4 (62.7, 111.7)	25.2 (8.3, 42.3)	<0.001	114.4 (88.3, 148.8)	84.8 (61.9, 108.1)	27.2 (7.5, 41.9)	<0.001	86.4 (62.7, 111.7)	84.8 (61.9, 108.1)	1.8 (–24.7, 20.9)	0.152
VFI (ml/s)	4.5 (2.8, 7)	1.4 (0.9, 2)	67.6 (49, 80)	<0.001	4.5 (2.8, 7)	1.5 (1, 2.1)	67.7 (48.3, 80.2)	<0.001	1.4 (0.9, 2)	1.5 (1, 2.1)	0 (–44.7, 27.6)	0.217
EF (%)	54.4 (43.4, 66.4)	65.9 (53.7, 78.6)	–18.4 (–49.5, 7.1)	<0.001	54.4 (43.4, 66.4)	67.4 (54.8, 81.3)	–22.9 (–54.9, 3.7)	<0.001	65.9 (53.7, 78.6)	67.4 (54.8, 81.3)	–4.1 (–30.3, 16.5)	0.001
RVF (%)	39.8 (28.4, 50.4)	28.8 (17.3, 42)	26 (–7.7, 53.5)	<0.001	39.8 (28.4, 50.4)	26 (15.9, 37.8)	33.7 (–3.2, 57.2)	<0.001	28.8 (17.3, 42)	26 (15.9, 37.8)	9.2 (–53.5, 46.3)	<0.001

IQR: interquartile range, VV: venous volume, VFI: venous filing index, EF: ejection fraction, RVF: residual volume fraction..

Hemodynamic changes were investigated according to the surgical method. The EVLA group showed an increase in the EF and a decrease in the VV, VFI, and RVF at 1 month postoperatively compared with the preoperative values (Table 3); these differences were significant. Similar findings were observed in the RFA group (p<0.05; Fig. 2).

Table 3 . Changes in hemodynamic parameters after EVLA, RFA, and HL with stripping.

Parameter (median, IQR)	EVLA							RFA							HL with stripping
	Preoperative vs. Postoperative 1 month			Postoperative 1 month vs. Postoperative 6 months			Preoperative vs. Postoperative 1 month			Postoperative 1 month vs. Postoperative 6 months			Preoperative vs. Postoperative 1 month			Postoperative 1 month vs. Postoperative 6 months
	Preoperative	Postoperative 1 month	p	Postoperative 1 month	Postoperative 6 months	p	Preoperative	Postoperative 1 month	p	Postoperative 1 month	Postoperative 6 month	p	Preoperative	Postoperative 1 month	p	Postoperative 1 month	Postoperative 6 months	p
VV (ml)	101.8 (77.2, 133.6)	78.1 (55.9, 105.7)	<0.001	78.1 (55.9, 105.7)	74.7 (53, 105)	0.375	108.9 (79.4, 143)	74.5 (56.5, 99)	<0.001	74.5 (56.5, 99)	76.3 (60, 105.5)	1.000	122.7 (98.1, 160.1)	91.4 (69.1, 116.8)	<0.001	91.4 (69.1, 116.8)	89.7 (69.2, 112.3)	0.248
VFI (ml/s)	3.8 (2.5, 5,7)	1.4 (1, 2.1)	<0.001	1.4 (1, 2.1)	1.5 (1, 2.2)	1.000	4.1 (2.5, 5.9)	1.4 (1, 2.1)	0.001	1.4 (1, 2.1)	1.6 (1.1, 2.3)	1.000	5.0 (3, 8)	1.4 (0.9, 1.9)	<0.001	1.4 (0.9, 1.9)	1.4 (1, 1.9)	0.781
EF (%)	55.8 (43.7, 68.4)	63.7 (49.1, 78.2)	<0.001	63.7 (49.1, 78.2)	65.1 (52.5, 79.1)	0.740	57.5 (43.8, 73.5)	66.8 (51.2, 81.8)	<0.001	66.8 (51.2, 81.8)	65.1 (51.9, 82.2)	1.000	52.6 (43.2, 63.9)	66.6 (55.6, 78.1)	<0.001	66.6 (55.6, 78.1)	68.6 (58.3, 82.2)	0.003
RVF (%)	41.2 (30.8, 50.4)	30.5 (17.2, 43.3)	0.001	30.5 (17.2, 43.3)	28.5 (17.3, 43.4)	1.000	35.2 (23.5, 48.6)	27.2 (18.4, 50.4)	0.232	27.2 (18.4, 50.4)	28.8 (18, 44.1)	0.680	41.2 (30.8, 50.4)	28.6 (17.1, 40.3)	<0.001	28.6 (17.1, 40.3)	23.8 (15.2, 34.4)	<0.001

EVLA: endovenous laser ablation, RFA: radiofrequency ablation, HL: high ligation, IQR: interquartile range, VV: venous volume, VFI: venous filing index, RVF: residual volume fraction, EF: ejection fraction..

Figure 2. Comparison of hemodynamic parameters between endovenous laser ablation (EVLA), radiofrequency ablation (RFA), and high ligation (HL) with stripping.

A scatter plot analysis was performed using the EF and VFI to identify the flow of hemodynamic changes after surgery. Preoperatively, most of the dots were scattered in the right middle field. However, these dots shifted to the upper left after surgery (Fig. 3).

Figure 3. Hemodynamic changes of ejection fraction (EF) and venous filling index (VFI).

The 1- and 6-month postoperative rates of reduction in the VV, VFI, EF and RVF are shown in Table 4. When the rate of reduction was compared according to the treatment modality, the 1-month postoperative rate of reduction in VV was higher in the RFA group than in the other groups. The 1- and 6-month postoperative rates of reduction in the VFI, EF, and RVF were higher in the HL with stripping group than in the other groups (p<0.05).

Table 4 . Differences of hemodynamic changes according to treatment modality for GSV reflux during follow-up.

Parameter (median, IQR)	Preoperative vs. Postoperative 1 month				Preoperative vs. Postoperative 6 months				Postoperative 1 month vs. Postoperative 6 months
	EVLA	RFA	HL and stripping	p	EVLA	RFA	HL and stripping	p	EVLA	RFA	HL and stripping	p
Δ VV (%)	22.5 (3.1, 38.4)	30 (7.4, 47.7)	25.6 (11.6, 42.3)	0.02	25.1 (3.7, 40.8)	26 (3, 41.4)	28.9 (11.3, 42.2)	0.193	3 (–27.9, 28.3)	–1 (–36.9, 20.1)	2 (–20.3, 19.2)	0.178
Δ VFI (%)	61.2 (40.4, 75.7)	60.9 (41.8, 77.1)	73.2 (57.4, 83.7)	<0.001	63.6 (41.8, 73.7)	60.6 (36.1, 75.3)	73.6 (55.2, 83.2)	<0.001	–2.7 (–45.4, 75.7)	–0.2 (–67.2, 23.9)	0 (–38.9, 25)	0.399
Δ EF (%)	–10.6 (–46, 14.8)	–14.2 (–53.8, 11.4)	–21.9 (–50.7, –0.1)	0.002	–18.1 (–51.1, 11.4)	–9.5 (–48.2, 13.9)	–29.9 (–56.3, –4.6)	<0.001	–3.2 (–35.1, 21.4)	–0.9 (–27.7, 26.6)	–5.4 (–28.3, 13.5)	0.414
Δ RVF (%)	20 (–20.1, 54.2)	18.5 (–35, 50.5)	29.5 (–1.3, 53.5)	0.01	23.3 (–15.7, 47.9)	18.7 (–36.8, 52.9)	40.5 (9.9, 61.4)	<0.001	2.8 (–70.5, 41.6)	8.9 (–66.8, 48.4)	11.6 (–43.7, 48)	0.114

GSV: great saphenous vein, IQR: interquartile range, EVLA: endovenous laser ablation, RFA: radiofrequency ablation, HL: high ligation, VV: venous volume, VFI: venous filing index, RVF: residual volume fraction, EF: ejection fraction..

The occlusion rates of the saphenous veins in the EVLA and RFA groups were investigated using ultrasonography. The occlusion rates on the day after surgery were 97.8% and 98.1% in the EVLA and RFA groups, respectively. The GSV occlusion rates at 1 and 6 months were 85.6% and 97.5% in the EVLA group and 95.7% and 99.4% in the RFA group, respectively (Fig. 4). There was no re-intervention between 1 and 6 months after surgery.

Figure 4. Occlusion rates in endovenous laser ablation (EVLA) and radiofrequency ablation (RFA) groups. (A) Postoperative 1 day, (B) postoperative 1 month, and (C) postoperative 6 months.

The goal of treating varicose veins is to reduce the overloading of the VV in the diseased lower extremities and relieve the associated symptoms. Therefore, it is important to compare the hemodynamic parameters before and after surgery.

APG can estimate the individual value of clinical venous hemodynamics, such as calf reflux, calf muscle pump function, and outflow obstruction. This method enables exact quantitation of the volume change rate of increase of the lower limb region from the knee to the ankle.

Park et al. [8] reported that the overall hemodynamic values significantly improved after surgical treatment, showing decreases in the VV, VFI, and RVF, with an increase in the EF. The rates of reduction in the VV, VFI, and RVF were 25.9, 72.6, and 30.4%, respectively. The rate of increase in EF was 21.4% in the HL stripping group (p<0.001). Miyazaki et al. [9] reported that the VV, VFI, and RVF were significantly reduced in the stripping group (p<0.001); however, the EF did not improve in this group.

Nishibe et al. [10] reported the results of a retrospective study involving 91 patients (124 limbs) who underwent RFA for varicose veins. The mean preoperative VFI was 4.1±3.1 ml/s. At 1 week, the VFI was significantly reduced to 1.4±1.0 ml/s. The VFI remained decreased to 1.3±0.8 ml/s and 1.4±1.2 ml/s at 1 month and 4 months, respectively. Similar results were obtained in this study. In the group treated with RFA, the preoperative median VFI was 4.1 ml/s (IQR, 2.5–5.9 ml/s). It was significantly decreased to 1.4 ml/s (1–2.1 ml/s) at 1 month after the surgery (p=0.001), and it remained well reduced to 1.6 ml/s (1.1–2.3 ml/s) at 6 months after the surgery.

Park et al. [11] analyzed hemodynamic changes after EVLA. The changes in all hemodynamic parameters at 1 month after EVLA relative to the preoperative values were significant. This indicates that EVLA can improve calf pump function and eliminate venous reflux. In the present study, the VV, VFI, and RVF decreased, whereas the EF increased significantly after EVLA. In the group treated with EVLA, the median parameters before the surgery were as follows: VV, 101.8 ml (IQR, 77.2–133.6 ml); VFI, 3.8 ml/s (IQR, 2.5–5.7 ml/s); RVF, 41.2% (IQR, 30.8%–50.4%); and EF, 55.8% (IQR, 43.7%–68.4%). Median parameters at 1 month after the surgery were as follows: VV, 78.1 ml (IQR, 55.9–105.7 ml); VFI, 1.4 ml (IQR, 1–2.1 ml); RVF, 30.5% (IQR, 17.2%–43.3%); and EF, 63.7% (IQR, 49.1%–78.2%). The changes in hemodynamic values 1 month after surgery relative to the preoperative values were significant (p=0.001).

In a study by Zhu et al. [7] on 254 limbs of 200 patients who underwent EVLT and pinhole HL, the treatment success rate was 100%. In the present study, HL and EVLT were performed simultaneously to prevent recanalization and DVT/pulmonary embolism that might occur [12-15]. There was no recanalization of the GSV during the follow-up. No reintervention was needed for any patient. Moreover, no eHIT or DVT were detected on ultrasonography during the follow-up period. This may be one of the reasons why the patient underwent HL of the SFJ.

The combination of the EF and VFI has a good correlation with disease severity [16]. Venous hemodynamic changes were plotted in relation to the EF and VFI. Most of the limbs were placed in a hemodynamically favorable direction. We observed improvements in the calf muscle pump function and venous reflux after treatment.

In our hospital, the hospitalization period for patients with varicose veins was 2 nights and 3 days. During the hospitalization period, expenses incurred other than surgery and material costs, including anesthesia, were almost the same for all treatment modalities. The surgical costs were $350 for HL with stripping, $1,162 for EVLA, and $939 for RFA. HL with stripping was the least expensive because it is covered by insurance. EVLA required an ultrasound a machine, a laser generator, and a disposable laser fiber [17]. RFA uses an ultrasound machine, a generator, and a disposable catheter [18]. The RFA catheter costs $803.

This study has some limitations. First, the study was conducted at a single institution. Additionally, the patients in the three groups widely differed even before treatment. Therefore, corrections using matched pairs should be considered in future studies.

In conclusion, all three treatments, HL with stripping, EVLA, and RFA, resulted in significant improvements in hemodynamic parameters after surgery. Thus, an appropriate method among the three surgical methods should be selected considering the patient’s condition and surrounding environment.

The authors declare no conflicts of interest.