Cyanoacrylate glue used to treat great saphenous reflux: Measures of outcome

Study population

These were nonrandomized consecutive patients with varicose veins (mainly bilateral) who presented to our tertiary vascular center. Inclusion criteria included fully consenting adult patients aged more than 18 years, who had primary varicose veins with pre-operative duplex ultrasound proven incompetence of saphenofemoral junction (SFJ) and/or reflux in the great saphenous vein (GSV).

All the pre-procedural duplex ultrasound scans were performed by certified vascular sonographers using the Sequoia 512 (Siemens, CA, USA) or Philips iU22 (Philips Healthcare Solutions, WA, USA) devices in our vascular laboratory. Saphenous vein reflux was defined by the presence of retrograde flow detected in the duplex ultrasound scan over the SFJ, with the patient examined in the standing position (more than or equal to half a second). ⁸ The diameter of GSV was measured, while patient supine, at three levels (proximal thigh near SFJ, mid thigh, and distal thigh above knee). Mean diameter of GSV was defined as the average of the above three measurements. All our patients gave written consent.

This study had local Institutional Review Board approval (reference UW 15-212). There was no deliberate selection bias by the authors to only perform the operation with favorable lesions, and all patients followed a fixed protocol of follow up.

Study procedure

All the endovenous procedures were performed in our Minimally Invasive Surgical Center under local anesthesia with the use of standard sterile technique. Since this was an IRB-approved project with a very new treatment modality in our region, we preferred to perform these procedures in the presence of an anesthetist.

The endovenous technique has been described previously. ⁹ In summary, after injection of local anesthesia in puncture site, percutaneous ultrasound-guided (uSmart 3200T, Terason, Burlington, USA) puncture of the GSV at the knee level was done using micro-puncture set (Angiodynamics, Queensbury, New York, USA). The 0.035 propriety guidewire was passed to the SFJ, and then exchanged to the propriety 5Fr long sheath. The cyanoacrylate adhesive was prepared and attached to the delivery catheter. Patient was then put in head-down position. The tip of the 5Fr introducer sheath/cyanoacylate catheter was advanced to the SFJ and positioned 4 cm distal to the SFJ, as measured on the ultrasound screen. There was no need for peri-venous tumescence anesthesia. With proximal compression by the ultrasound probe, the cyanoacylate was injected intravenously as per protocol, with two injections of approximately 0.09 mm were given 1 cm apart at this location, followed by a 3-min period of local compression, and then repeated 3 cm spaced injections and 30 s ultrasound probe compression sequences until the entire length of the target vein was treated. Treatment length of GSV was defined as the length of GSV that received cyanoacrylate adhesive treatment. GSV occlusion and the lack of deep vein thrombosis (with compressibility) were confirmed by duplex ultrasound intraoperatively. Small stab avulsions of varicosities under local anesthesia were performed in the same operative setting.

Patients were discharged same-day post procedure from the vascular ward. Post-procedure compression stockings were used in view of avulsions and the small wounds dressed by community nurses.

Peri-procedure assessment and follow up

All the patients underwent serial clinical and duplex ultrasound examination at 1 week, 1 month, 6 months, and 1 year. The primary endpoint was procedural success rate with obliteration of the GSV, lack of deep vein thrombosis, and lack of clinical recurrence of varicose veins. The complete closure of the saphenous vein was defined as duplex ultrasound noncompressibility/absence of flow signal along the entire treated saphenous vein segment with no discrete segments of patency exceeding 5 cm as read by a qualified specialist vascular ultrasonographers. ¹⁰

We also looked at postoperative pain score, ecchymosis at site of puncture and avulsion, side-effects and serious adverse events possibly related to the treatment at discharge time and follow up, time to return to work, improvement in disease specific, and general quality of life scores. The perioperative pain score will be recorded with the patient being asked to rate pain experienced immediate postoperatively on a 0 to 10 numeric rating scale. The ecchymosis at site of puncture and avulsion will be recorded at 1 week and rated as minimal, mild, moderate, or severe. The venous clinical severity score (VCSS) is a well-recognized system in recording venous severity.^11,12 To avoid confusion with the CEAP scale and to generate a dynamic score, the categories of the VCSS are presented on an elemental basis to add emphasis to the most severe sequelae of venous disease that are likely to show the greatest change in response to therapy. ¹³ The VCSS was measured for every leg sequentially preoperatively, at 1 month, 3 months, 6 months, 1 year, and 2 years postoperatively. The Aberdeen varicose vein questionnaire (AVVQ) was developed and validated by Garratt et al. ¹⁴ for measuring quality of life in patients with lower limb varicose veins. The AVVQ was asked for every patient preoperatively, at 1 month, 3 months, 6 months, 1 year, and 2 years postoperatively. The Short Form Health Survey 36 Item (SF-36) is a widely-used and well-validated generic quality-of-life instrument.^15,16 It assesses two domains: physical health (assessed as the patient’s level of functioning) and mental health (assessed as an indication of well-being). When complete, the survey generates a score ranging from 0 to 100, with higher scores indicating better general health perception. The SF-36 was measured for every patient preoperatively, at 1 month, 3 months, 6 months, 1 year, and 2 years postoperatively. All data were entered prospectively into our departmental computerized database.

Statistical analysis

Incomplete closure or recanalization of the GSVs would be detected by duplex ultrasound. Closure rates over time were analyzed using Kaplan–Meier analysis. Risk factors of mean diameter of GSV, treatment length of GSV, and preoperative clinical score of CEAP were tested for prediction of recanalization using Cox proportional hazard regression analysis. All analyses were performed using SPSS (version 22.0). A p-value of <0.05 was regarded as statistically significant.

Patients and disease characteristics

A total of 57 legs in 29 consecutive patients with duplex ultrasound-proven incompetent SFJ/GSV reflux were included in this study between September 2014 and October 2015. Twenty-eight had bilateral lower limb varicose veins. We preferentially chose bilateral cases because of cost. All were day cases, and all the patients returned for all the scheduled follow-up visits. There were 20 (69%) women. Thirty-three legs (58%) had CEAP clinical class 3 venous disease, twenty legs clinical 4a, two legs clinical 4b, one leg clinical 5, and one leg clinical 6 (Table 1). The predominant symptoms were leg pain, cramps, and aching. The VCSS and AVVQ scores were consistent with mild to moderate venous reflux disease at baseline (Figure 3).

OPEN IN VIEWER

Table 1.

Baseline characteristics.

	Number	Percentage
Patients’ characteristics (n = 29)
Age	Median 63 (range 39–80)
Male: Female	9:20
Co-morbidities
DM	3/29	10.3
HT	11/29	37.9
Cardiac	7/29	24.1
Renal	1/29	3.4
COPD	0/29	0
Leg varicose vein characteristics (n = 57)
CEAP clinical classification
C3	33/57	57.9
C4a	20/57	35.1
C4b	2/57	3.5
C5	1/57	1.8
C6	1/57	1.8
Mean diameter of GSV	Median 7.1 mm (range 3.9–11.4 mm)
Treatment length of GSV	Median 27.0 cm (range 17–33 cm)

The average diameter of the great saphenous vein in the treatment zone as determined by pre-operative duplex ultrasound with the patient supine was 7.1 mm (range 3.9–11.4 mm). The median treatment length of GSV was 27 cm (range 17–33 cm) (Table 1).

Operating procedures

There was no deviation of the protocol: all the GSVs were successfully punctured percutaneously, cannulated according to protocol, and all the cyanoacrylate injections were at precisely 4 cm from the SFJ. All the patients had immediate on-table completion duplex ultrasound to document successful obliteration of the GSV in the treatment segment, and lack of deep vein thrombosis as shown by calf squeeze and compressibility in the common femoral vein. The mean procedural time from aseptic prepping and draping to completion of all avulsions and application of support stockings was 64 min (range 28–99 min).

Postoperative results

The mean pain ratings were low at 0.86 (range 0–3) at immediate postoperative period. All the patients walked home independently. The severity of the ecchymosis at the puncture and avulsion sites as assessed at postoperative day 7 was low: no bruises in 40 legs, mild in 16 legs, and moderate in 1 leg. All the bruises were from avulsion phlebectomy sites and not from the GSV puncture sites.

Post operation, the median time of return to work or normal life was 1 day, ranging from 1 to 16 days. Most of the delay was due to dressings of wounds for the avulsion sites. All of the patients resumed normal daily activities on discharge. Operative complications included two minimal extension of thrombus to deep vein. Both required treatment of subcutaneous low molecular weight heparin injection for 1 week. Reassessment duplex ultrasound afterwards showed completion resolution of the deep vein thrombus. Other operative complications were minor with one minor wound infection over an avulsion surgical site which resolved with wound dressings. One patient had mild paraesthesia over dorsum of the foot which improved upon follow up. With a mean follow-up period 9 months (range 1–13 months), there was no clinical varicose vein recurrence.

Duplex ultrasound follow up

The mean follow-up period of this cohort of patients was 9 months (range 1–13 months). Fifty-seven (100%), fifty-three (75%), and seventeen (30%) legs had passed the 3, 6, and 12 months follow-up points, respectively. None of our patients defaulted follow-up visits. On day 7, all the GSVs in the intended treatment segments were successfully obliterated. At post-op 1 week, 1 month, 6 months, and 12 months, the closure rates were 100%, 95.3%, 90.3%, and 78.5%, respectively (Figure 1). There were no total recanalization. Partial recanalization (defined as any further recanalization compared to previous scans) occurred in seven legs, with one, two, two, and two recanalizations of the proximal GSVs noticed in 1 week, 1 month, 6 months, and 1 year respectively during the follow up (Figure 2). None of the recannalized GSVs resulted in clinical recurrence or deterioration of the CEAP score. OPEN IN VIEWER

OPEN IN VIEWER

Figure 2.

Example of a 6-month follow-up duplex showing recanalization of the previously obliterated proximal great saphenous vein.

OPEN IN VIEWER

Figure 3.

Graphical representation of the venous severity scores (a), Aberdeen varicose vein questionnaires (b), and SF 36 questionnaires (c).

Prediction of recanalization

Cox regression analysis showed mean GSV diameter ≥8 mm, which was a significant independent predictor of recanalization (hazard ratio 6.92, 95%CI 1.34–35.67, p = 0.021) (Table 2).

OPEN IN VIEWER

Table 2.

Predictors for recanalization.

	Hazard ratio	95% Confidence interval	p-Value
Mean GSV diameter ≥8 mm	6.92	1.34–35.67	0.021
Treatment length <30 cm	2.59	0.34–19.81	0.359
Pretreatment CEAP score C4-6	2.29	0.51–10.23	0.280

Closure rates of those GSV <8 mm remained high at 100%, 93.1%, 93.1%, and 93.1% at post-op 1 week, 1 month, 6 months, and 12 months, respectively; whilst that for GSV ≥8 mm, closure rates were 90.9%, 90.9%, 68.2%, and 34.1% at post-op 1 week, 1 month, 6 months, and 12 months, respectively (log-rank test, p = 0.006) (Figure 1(b)).

On the other hand, treatment length of the GSV (hazard ratio 2.59, 95%CI 0.34–19.81, p = 0.359) and preoperative clinical score of the CEAP (hazard ratio 2.29, 95%CI 0.51–10.23, p = 0.280) were not predictive of recanalization (Table 2).

Questionnaire follow up

The results of the VCSS, AVVQ, and the SF-36 surveys are tabulated in Figure 3. Our VCSS scores decreased from a mean ± standard deviation of 6.91 ± 3.45 at baseline to 2.43 ± 1.49, 2.23 ± 1.60, 1.76 ± 1.71, and 1.69 ± 0.95 at 1 month, 3 months, 6 months, and 1 year of follow up, respectively. Those with partial recanalizations presented higher baseline and subsequent scores (Figure 3(a)).

The AVVQ scores decreased from a mean ± standard deviation of 23.66 ± 11.10 at baseline to 6.10 ± 5.55, 5.77 ± 5.88, 4.74 ± 5.46, and 4.13 ± 2.80 at 1 month, 3 months, 6 months, and 1 year of follow up, respectively. The patients with partial recanalizations presented higher baseline and subsequent scores (Figure 3(b)).

Physical scores of the SF-36 questionnaire remained similar from a mean ± standard deviation of 44.24 ± 7.50 at baseline to 43.85 ± 8.22, 43.85 ± 8.22, 43.52 ± 11.08, and 52.90 ± 4.30 at 1 month, 3 months, 6 months, and 1 year of follow up, respectively. Similarly, mental scores of the SF-36 questionnaire remained similar from a mean ± standard deviation of 54.26 ± 9.56 at baseline to 52.50 ± 10.16, 52.50 ± 10.16, 51.26 ± 12.88, and 62.55 ± 1.15 at 1 month, 3 months, 6 months, and 1 year of follow up, respectively (Figure 3(c)).