P. ZAMBONI, M. G. MARCELLINO, M. CAPPELLI, C. V. FEO, V. BRESADOLA, G. VASQUEZ. A. LIBONI

Objective. Evaluation of saphenous vein sparing surgical procedures alternative to high ligation and distal stab avulsion, in terms of effectiveness and suit-ability for eventual bypass surgery.
Experimental design. Prospective evaluation of 421 operations for primary varicose veins, 64 external valve-plasties of the sapheno-femoral junction (EV-SFJ), (42 performed using the hand sewing technique and 22 using the Veno-cuff device), mean follow-up 52 months, and 357 hemodynamic correction of varicose veins (French acronymis CHIVA), mean follow-up 49 months. Moreover, a subgroup of 27 patients was operated on using the CHIVA technique in two steps, mean follow-up 18 months.
Setting. Institute of General Surgery, University of Ferrara. Institutional practice, one-day surgery.
Patients. Patients were selected using clinical and duplex scanning evaluations, and classified according to CEAP criteria. Patients with varicose veins due to sapheno-femoral reflux with duplex scanning evidence of mobile valve leaflets underwent EV-SFJ. The other patients were operated on using the hemodynamic correction technique. Interventions. EV-SFJ restores valve function correcting vein wall dilatation by applying an external prosthesis. CHIVA consists of selected ligatures of the superficial veins that allow superficial blood aspiration in the deep veins through the perforators as well as the preservation of saphenous drainage.
Measures. The outcome was evaluated with independent clinical and ultrasonographic examinations; pre and postoperative AVP and LRR-RT measurements were assessed in 125 cases. Data from self-assessment of the functional and cosmetic result of the patients of the CHIVA group were also obtained using a scoring system. Moreover, scanning the preserved long saphenous vein the rate of long saphenous vein suitable as arterial conduit following sparing surgery was also evaluated.
Results. Overall long saphenous vein patency registered after EV-SFJ and CHIVA was 94%. Varicose veins recurrence rate was 12% and 11%, respectively. Postoperative AVP and LRR-RT improvement was stas-tically significant (p<0.001).
Conclusions. These two alternative procedures seem to be effective In varices treatment following the proposed indications and techniques. In addition, they appear able to preserve a more significant rate of saphenous veins suitable for eventual bypass surgery than high ligation and multiple cosmetic avulsion.

Saphenous vein sparing surgery for varicose vein disease is still controversial in the literature. Many groups note the increased risk of varicose vein recurrence if the long saphenous vein is not excised at the time of initial surgery. On the other hand, some comparative studies show no significant differences in recurrence rate between stripped and non-stripped limbs, especially if the preoperative assessment allows mapping of the insufficient veins. Alternative long saphenous vein sparing surgical procedures such as the external valve-plasty of the sapheno-femoral junction (EV-SFJ) and the hemo-dynamic correction of varices (French acronym is CHIVA), seem to be effective for varicose treatment and, maintaining a high saphenous vein patency rate more useful for future grafting as compared to high ligation and distal stab avulsion. This paper shows a prospective long-term evaluation of these treatments defining precisely the patients' population by means of CEAP classification and ultrasonographic criteria: moreover, the rate of long saphenous vein suitable as an arterial conduit following these new procedures is evaluated by means of duplex, scanning the preserved vein.

Materials and methods

Four hundred and twency-one cases of primary varicose veins disease with epical sapheno-femoral reflux have been selected for alternative saphenous vein sparing surgery at the Vascular Laboratory of our Institute.

EV-SFJ

Patients' selection and preoperative assessment.-64 patients (15 men, 49 women, 20-70 years old. mean age 41 years old, were selected for EV-SFJ using the following criteria:

Clinical.-Primary varices, symptoms of venous insufficiency, absence of prior thrombophlebites or sclerotherapy.

Duplex scanning (Ansaldo AU 530, 7.5-10 MHz probe, Italy).-Evidence of mobile valve leaflets of uniform length at the sapheno-femoral junction (SFJ). Sapheno-femoral reflux and competence of the deep venous system. Long saphenous vein (LSV) diameter at the mid thigh was measured both preoperatively and 6 months later.

CEAP classification.-52 patients were retrospectively and 12 prospectively classified according to the new CEAP classification criteria. Regarding the clinical class, C, all the patients were symptomatic and affected by simple varicose veins without oedema, skin changes and/or ulcerations. The etiology, E, was always primary. The anatomical distribution of cases, A, was in the LSV above the knee in 27 cases, above and below the knee in 35. Perforators were found to be incompetent at the thigh in 34 cases and at the calf in 42. Finally the pathophysiology of the disease, P, was due to reflux in all cases. The following algorithm describes the selected patients: C2s, Ep, As2-3 p 17-18. Pr.

Venous function assessment.-32 patients underwent preoperatively and 6 months postoperatively ambulatory venous pressure (AVP) and light reflection rheography-refilling time (LRR-RT) measurements.

Fig 1-Left: Private Circulation with the hemodynamic pattern of type 1 Shunt (outlet of the re-entry perforator on the LSV). Right: Procedure performed in such cases: 1) High ligation, 2) Disconnection from the LSV of the varicose tributary and, eventually, avulsion with cosmetic incisions. Legend: LSV. long saphe-nous vein: SSV short saphenous vein: DV, deep venous system: Giac, Giacomini vein: TV, tributary. PV, perforator. Arrows, blood flow direction determined by Doppler mode.

Fig. 2.-Left: Private Circulation with the hemodynamic pattern of type III Shunt (outlet of the re-entry perforator on a tributary of the LSV, type III Shunt with re-entry in Table III). Right: Procedure performed in such cases: 1) High ligation. 2) Interruption of the superficial vein just below the outlet of the insufficient perforator. Legend: see Figure 1.

Fig. 3.-Left: Private Circulation with the hemodynamic pattern ot type III Shunt (outlet of the re-entry perforator on a tributary of the LSV). Right: Procedure performed in such cases (type III Shunt without reentry in Table III): 1) High ligation. 2) Disconnection from the LSV and avulsion with cosmetic incisions of the varicose tributary.

Fig 4-Left: Private Circulalion with the hemodynamic pattern of type III Shunt "short and/or deep" (re-entry perforator on a tributary close to the origin from the LSV (short) or deep, and thus not visible). Right: Procedure performed in such cases: 1) High ligation. 2) Interruption of the superficial vein just below the outlet of the insufficient perforator, and eventual cosmetic phlebeciomy below this point. Legend: see Figure 1.

Fig. 5.-Left: Private Circulation with the hemodynamic pattern of type III Shunt (outlet of the re-entry perforator on a tributary of the LSV) and Doppier signal recorded on the saphenous trunk under squeezing manoeuver performed in standing: a forward flow of the saphenous vein during muscular contraction is followed by a reverse flow during muscular relaxation. Right: Procedure performed according to the treatment named CHIVA in two steps: 1) First step, disconnection from the LSV and eventual avulsion with cosmetic incisions of the varicose tributary Doppler trace demonstrates under squeezing manoeuver performed in standing a forward flow of the saphenous vein during muscular contraction and the disappearance of the reverse flow during muscular relaxation. Legend: see Figure 1.

Fig. 6.-Left: Private Circulation with the hemodynamic pattern of typeI Shunt (outlet of the re-entry perforator on the LSV with a secondary reflux from an incompetent PV proximal to the re-eniry PV). Right: Procedure performed in such cases: 1) High ligation. 2) Interruption of the superficial vein just below the outlet of the insufficient perforator. The ligature transforms a secondary reflux point into a re-entry point, with a Doppler detectable inward flow, and preserves at least two LSV segments. Legend see Figure 1.

Fig. 7.-Left: After development of a new re-entry perforator, Doppler signal of reflux is newly recorded on the saphenous trunk. Right: Second step of the procedure is high ligation that determines a low-pressure reverse flow toward the re-entry perforator. Legend: see Figure 1.

Operations.-Under local anaesthesia, the saplieno-femoral junction was exposed and the terminal valve encircled by an external prosthesis, using the hand-sewn technique previously reported. In 42 cases the valve-plasly was performed using a PTFEe sleeve (Gore-Tex, USA) 0.4 mm thick. 1-1.5 cm long to surround an average circumference of 1.7 cm. The graft was fixed around the vein with a 7/0 PTFEe interrupted suture. In 22 cases the valve- plasly was performed using the Veno-cuff device (Vaso Inc., USA) surrounding the valve site with a Dacron reinforced silicone cuff.
Optimal circumference for valve function restoration was considered that of the saphenous vein below the valve, previously measured in quiet standing by means of B-mode high resolution. Such a value was selected by moving the calibrator and the cuff secured firing a stainless steel surgical staples. When the terminal valve was absent or the sub terminal was also very dilated, we performed a second valve-plasty of the sub-terminal, too. This occurred in 41 cases (37 PTFEe sleeves and 4 Veno-Cuff). Patients were discharged 3-5 hours after surgery with postoperative elastic stockings.

Intraoperative valve competence tests.-Intraoperative detection of the restored valve function was made:
1. In all the treated cases intraoperative C.W. Doppier examination (Stereodop 448 S, Echomed, 4-8 MHz probe, France): The Valsalva and proximal compressive manoeuvres demonstrated disappearance of reflux. Conversely, compressive distal manoeuvres were used to detect LSV patency.
2. In 21 cases intraoperative video angioscopic assessment was used.

CHIVA

Patients' selection and preoperative assessment.-Other 357 patients with primary varicose veins syndrome with typical sapheno-femoral reflux were treated by hemodynamic correction, using the following selection criteria:

Clinical.-Primary and symptomatic chronic venous insufficiency (CVI) of all clinical classes, absence of prior chrombophlebites and/or surgical and sclerotherapic treatment.

Duplex scanning (Ansaldo AU 530. 7.5-10 MHz probe. Italy).-Preoperative duplex detected patients with typical sapheno-femoral reflux. excluding patients affected by varicose veins fed by different primary reflux points. In addition duplex made it possible to outline on the skin the points where the superficial veins should be interrupted (see Preoperative duplex mapping). The LSV diameter at mid thigh was measured boch preoperatively and 6 months later.

CEAP classification.-188 patients were retrospectively and 169 prospectively classified according to the new CEAP classification criteria. The clinical class. C, ranged from C2 to C6 (267 with simple varicose veins (C2), 60 with oedema (C3). 11 wich lipodermatosclerosis and/or other skin changes (C4). 16 with healed (C5) and 3 with active ulcer (C6): all the selected patients presented with classic symptoms of CVI. of different severity. The etiology was, obviously, primary. The anatomical distribution of cases was in the LSV above the knee in 104 cases, above and below the knee in 253 cases, perforators were found to be incompetent at the thigh in 134 cases and at the leg in 283.. Finally the pathophysiology was due to reflux in all cases. The following algorithm describes the selected patients C2-6. EP. As2-3, p 17-18, Pr.

Venous function assessment.-73 patients underwent preoperacively and 6 months postoperatively AVT and LRR-RT measurements.

Preoperative duplex mapping.-A preoperative skin map was obtained by duplex in order to identifv the points where the superficial veins had to be interrupted.
The ultrasonographic image of the so-called "saphenous eye" is a precise and constant marker clearly demonstrable in the transverse duplex access of the internal surface of both the thigh and leg. This finding is typical of few superficial veins and precisely, of the long. short and anterior saphenous and Giacomini's vein. The image is due co the duplication of the superficial fascia around the saphenous vein.
In all the selected patients duplex examination allowed us to recognize and differentiate the saphenous vein from Other superficial veins, as well as to identify the so-called Private Circulation (PC) or Shunt, a vicious circle of blood between the superficial and the deep veins in primary cases. The circle starts during muscular relaxation, when the blood from the more proximal reflux point, the SFJ, through the LSV and/or the superficial veins, flows downwards to the re-entry point represented by a Perforating Vein (PV). and then into the deep veins. The circle ends during the following muscular contraction, when the blood flows forward through the deep veins, and then again to the proximal reflux point when muscular relaxation occurs (Figs. 1. 2). Shunts in varicose veins were classified in four types according to the CHIVA theory. When the sapheno-femoral junction is incompetent, as in all the selected patients, two models of private circulation (PC) have been described: Type I Shunt and Type III Shunt. One hundred and eighty-six of the selected patients (53%) showed the hemodynamic pattern of Type I Shunt, in which the superficial branch of the PC, from the reflux point to the re-entry point. is entirely represented by the LSV. and the re-entry PV, obviously dilated, is situated on the LSV itself. In Figure I are shown the points of surgical interruption in case of Type I Shunt presentation. In contrast, 171 patients (47%) showed the hemodynamic pattern of Type III Shunt, in which the superficial branch of the PC. from the reflux point to the re-entry point, is represented either by the LSV or by a TV of the LSV. In such cases the reentry PV is situated on the TV instead of the LSV (Fig. 2).
Obviously, also the patients of this group showed a proximal reflux point, the SFJ. and a secondary reflux points at the outlet of incompetent TVs. one of which contains the outlet of the re-entry PV. In Figures 2, 3, 4, 5 are shown examples of hemodynamic presentations of Type III Shunt.

Operations.-All operations were performed under local anaesthesia. The SFJ was exposed in the usual way, maintaining the tributaries. The SFJ was clipped or disconnected (flush ligation). The preserved tributaries allow the superficial and pelvic venous systems to drain into the LSV, where the blood flow will be reversed toward the re-entry perforators. In Type I Shunt, we ligated all the incompetent TVs from their origin on the LSV. If the TV is particurarly dilated cosmetic avulsion of its proximal tract can be performed, while the PV on the LSV main trunk allows blood re-entry into the deep veins without sapheno-femoral reflux overload. Moreover, in Type I Shunt a secondary reflux from an incompetent PV proximal to the re-entry PV was detected in 31 cases (9% of the operated cases). (Fig. 6). In such a case we interrupted the LSV just below the origin of the proximal PV. The ligature transforms a secondary reflux point into a re-entry point, with a Doppler detectable inward flow and preserves at least two LSV segments (Fig. 6). In contrast, in Type III Shunt, the original technique proposed by Franceschi in 1988, the so-called "CHIVA 1", consists of disconnecting either the SFJ (the proximal reflux point) and all the incompetent TVs from the LSV (the secondary reflux points), except chat containing the re-entry PV (Fig. 2). Such a lactic allows us to maintain the LSV patency and the drainage function, and was attempted in 72 cases (20% of ehe operated cases); in addition the TV was ligated just below the outlet of the re-entry PV (Fig. 2).
Thirty-five patients were operated on in the same way but presented with the segment of the TV, which contains the re-entry PV, close to the origin from the LSV (short) or deep, and thus not visible (Fig. 4). Finally, other 97 cases with the hemodynamic pattern of Type III Shunt and large varicose veins (27%) for cosmetic reasons, were operated on disconnecting and excising from the LSV all the incompetent varicose tributaries (Fig. 3). Multiple cosmetic phlebectomy was also performed (conservative but not hemodynamic treatment). Patients were discharged 3-5 hours, after surgery with elastic stockings.

Clinical assessment of the results.-The assessment was performed by an independent assessor who had not been involved in previous surgical decision making and operative procedure (MGM) according to the following criteria previously proposed in the literature:

Objective assessment:
- class A: no visible and palpable varicose veins,
- class B: a few visible and palpable varicose veins with diameter <5 mm;
- class C: remaining or newly formed varicose veins with diameter >5 mm;
- class D: insufficient main trunks and perforator. In addition, functional and cosmetic results were self-assessed by the patients, at the time of the last examination in Hospital, using a simple analogue scale well explained by MGM to the patients themselves:

Subjective assessment:
- class A: no inconvenience;
- class B: slight functional or cosmetic imperfection, but satisfaction with the result;
- class C: appreciable functional or cosmetic failure; improvement but dissatisfaction with the result;
- class D: unaltered or increased inconvenience.

The four classes, both subjectively and objectively assessed, were divided in accord with the preoperative hemodynamic pattern (Shunt I or III) and cesced for significance by the x2 test.

Selection criteria adopted in a sub-group of patients.-After 4 years of mean follow-up we selected other 27 patients for the so called CHIVA in 2 steps treatment, proposed to avoid treatment failures chat both the first part of the present study and the literature had shown to coincide with the hemodynamic patterns of Type III Shunt. In this way a sub-group of patients affected by superficial incompetence of all stages, not previously treated, was selected according to the following ultrasonographic criteria:
Varices supplied by SFJ reflux with re-entry perforator located in a TV (Type III Shunt). Competent deep venous system.
We selected 27 patients, 22 women and 5 men, mean age 44 years old (range 24 to 54 years). They also underwent preoperative LSV diameter assessment at mid thigh, repeated 6 months post-operatively.

CEAP classification.-Patients were prospectively classified in accord with CEAP. The clinical class, C, ranged to C2 to C5 (12 C2, 10 C3. 4 C4, 1 C5) and all the patients were symptomatic. The etiology, E, was primary. The anatomical distribution of cases, A, was in the LSV above the knee in 9 cases, above and below the knee in 18 cases; perforators were always incompetent at the leg. Finally the pachophysiology, P, was due to reflux in all cases. The following algorithm describes the selected patients C2-5, Ep, As2-3 p 18. Pr.

Preoperative duplex mapping.-In all the selected patients duplex examination allowed the identification of the superficial branch of the PC represented by the LSV from its junction to the origin of the TV on which the outlet of the re-entry PV is located (Fig. 5).

Operations.-The treatment is performed in two steps. The first step is represented by the disconnection of the origin of the TV containing the "reentry" PV from the main trunk of the LSV, thus transforming the refluent LSV into an LSV with a forward flow during muscular contraction, but no Doppler-detectable reverse flow during muscular relaxation (Fig. 5). However, Doppler-detectable reverse flow could be demonstrated under Valsalva manoeuvre. The second step, represented by the section-ligature of the SFJ, is performed when the LSV again shows a reverse flow due to the development of a new re-entry PV situated on the LSV itself (transformation into a Shunt type 1) or on a new insufficient TV.

Table I.-Pre and postoperative venous funcion assessment
Parameters	Preoperative	Postoperative	Student's "t" and Wilcoxon tests
EV-SFJ
AVP	31.78±4.74	20.96±3.24	p<0.001
LRR/RT	12.50±5.18	23.68±4.20	p<0.001
CHIVA 1
AVP	50.13±6.56	28.82±7.14	p<0.001
LRR/RT	10.12±2.61	19.80±4.91	p<0.001

Table II.- Results of EV-SFJ (mean follow-up 52 months)
EV-SFJ	N total	%
Varicose vein recurrences	7/62	11
Long saphenous vein patency	58/62	94
early postop. thrombosis	2/64	3.1
late occlusion	2/62	3
Sapheno-femoral reflux recurrence	6/62	10
New regurgitations points	8/62	13
Graft infections	0/62	0

Table III. - Results in CHIVA 1 group/mean follow-up 72 months.
Parameters	Type I Shunt		Type III Shunt with re-entry		Type III Shunt without re-entry		Type III Shunt "short or deep"		Total %
	N/total	%	N/total	%	N/total	%	N/total	%	Total %
Varicose veins disapprarance	172/186	92	16/27	59	81/92	87	32/35	91	89
Varicose veins reduction with exercise	9/186	5	8/27	30	4/92	4	2/35	6	-
Recurrent reflux site: Sapheno-femoral	2/186	1	1/27	4	3/92	3	1/35	3	2
Perforators	6/186	3	1/27	4	6/92	7	1/35	3	4
Sapheno-tributar	6/186	3	27/27	100	2/92	2	3/35	9	11
Saphenous vein patency	184/186	99	27/27	100	74/92	80	35/35	100	94
Saphenous vein thrombosis	2/186	1	0/27	0	18/92	20	0/35	0	6
Symptoms improvement	183/186	98	27/27	100	84/92	91	35/35	100	97
Subject Eval
Class A	162/186	87	8/27	30	74/92	80	30/35	86	81
Class B	21/186	11	16/27	59	14/92	15	3/35	9	16
Class C	21/186	1	2/27	7	3/92	3	2/35	6	3
Class D	1/186	1	1/27	4	2/92	2	0/35	0	1
Object Eval
Class A	172/186	92	16/27	59	81/92	87	32/35	91	89
Class B	4/186	2	3/27	11	5/92	1	4/35	11	5
Class C	6/186	3	7/27	26	3/92	1	1/35	3	4
Class D	2/186	1	1/27	4	3/92	1	0/35	0	2

Table IV.-Results in two steps CHIVA group: mean follow-up 18 months
Type III Shunt	N cases	Rate (%)
Varicose vein disappeareance	27/27	100
S-F reflux disappeareance:
Postop	27/27	100
3 months	9/27	33
6 months	19/27	70
>12 months	4/27	15
Site of the new re-entry:
Saphenous vein	10/27	37
Tributaries	13/27	48
Neither reflux nor re-entry	4/27	15
Saphenous vein patency	27/27	100
Saphenous vein thrombosis	0/27	0
Symptoms improvement	26/27	96

Results

Of the 421 selected patients who entered the study 2 of the EV-SFJ and 17 of the CHIVA group did not finish it. We have divided the results evaluation into three sections.

EV-SFJ

For these 62 patients follow-up lasted 52 months, ranging from 12 to 84. The outcome evaluation consisted of clinical and duplex scanning examinations for all the patients every 3 months for the First three years and then every year. 12 patients underwent descending and 4 ascending venography. Clinical results are summarized in Table 1. In the early postoperative period we had 2 saphenous thrombophlebites due to a technical error, requiring emergency short stripping. Both cases had been operated on with the hand sewn technique above described and were lost for further follow-up. All patients were discharged at the day of surgery. Total varices recurrence race was 12% (7/60). Ultrasonographic follow-up showed the long saphenous vein completely preserved in 58 cases (94%). Mean preoperative diameter at middle thigh was 5.6 mm versus 4 mm recorded after surgery.

Venous function assessment.-AVP and LRR/RT measurements were performed in 32 cases preoperatively and 6 months postoperatively. AVP and LRR/RT modifications after surgery evaluated both by Student's "t" and Wilcoxon tests, demonstrated a highly significant variation (p<0.001)). (Table II).

CHIVA 1

In this patients' group the mean follow-up lasted 49 months, ranging from 72 to 12. Clinical and duplex evaluations were made every 6 months for the first 3 years and then every year. Operations were well tolerated under local anesthesia. Postoperative analgesic administration was not necessary. Patients resumed working activity within 3-7 days after surgery.

Ultrasonographic findings and clinical correlation.-Results are summarized in Table III, dividing the cases into four sub-groups, Type I Shunt (186 cases. Fig. 1), Type III Shunt operated on maintaining the reverse flow from the LSV to a long and superficial TV containing the re-entry PV(27 cases, Type III Shunt with re-entry in Table III, Fig. 2); Shunt III operated on without preserving the LSV drainage function (92 cases, Type III Shunt without re-encry in Table III, Fig. 3) and finally, Type III Shunt with a short or deep TV containing the reentry PV (35 cases, Type III Shunt short or deep in Table III, Fig. 4). Overall saphenous vein patency recorded was 94%, with a mean diameter measured at mid thigh of 4.6 mm as compared to 6.2 mm recorded preoperatively. When patency was demonstrated the saphenous flow was reversed and with low velocity. Two patent segments of LSV were demonstrated in 29 cases of the 31 LSV interrupted (94%), whereas in 2 cases we ligated the main trunk above the re-entry perforator causing a symptomatic LSV thrombosis for technical error. Total recurrences/residual varicose veins registered were 11%, 8% for Shunt I and 16 % for Shunt III, respectively. Symptoms improved in 97% of cases, no-ulcer recurrences were observed in the outcome of the 19 patients in pre-operative clinical class 5 or 6.

Subjective and objective assessment of the results.- These results are also summarized in Table III. Better results obtained with this technique in patients with Type I Shunt as compared to those objectively and subjectively assessed in Type III Shunt are statistically significant (x2=22.144, p<0.0001). However, overall evaluation of the technique demonstrated 84% of patients in class A, 11% in class B, 4% in class C and 1% in D (x2 p<0.0001),

Hemodynamic results.-AVP and LRR/RT postoperative evaluations were done 6 months after surgery. The difference between pre and postoperative measurements was statistically significant using both Student's "t" and Wilcoxon's tests (p<0.001), (Table 1).

CHIVA in two steps/or Type III Shunt

The last group of 27 patients with Type III Shunt was operated on using the two steps CHIVA strategy (Figs. 5, 7). Although the SFJ is not treated in the first operation LSV reflux and symptomatology disappeared immediately after. When a new re-entry point-was developed LSV reflux was newly detected and symptoms worsened. This occurred in 33% of cases after 3 months, in 70% after 6 but in 15°/o of cases after 12/18 months follow-up neither reflux nor re-entry was detected. The new re-entry perforator was detected in the LSV main trunk in 37% of cases (Type III Shunt transformed in Type 1 Shunt) and in 48% newly in a TV, but deep, short or still not varicose and visible (Shunt type III transformed in Type III Shunt short and/or deep). The follow-up lasted on average 18 months, ranging from 14 to 24. Clinical and ultrasonographic evaluations were made every 3 months.
Poscoperative course was quite similar to that described in the first group. Results are summarized in Table IV.

Discussion

Long saphenous vein sparing surgical procedures have two main end-points: to perform an ambulatory and effective varicose veins treatment and to save the long saphenous vein for an eventual future grafting. This paper also introduces the concept of differentiating the surgical treatment on the basis of the ultrasonographic features of patients with varicose veins. We will discuss these three concepts in regard to the two operations evaluated.

EV-SFJ

Edwards and other authors showed that in early varicose stages valve incompetence is due to a parietal dilatation with normal valve leaflets. Hallberg and other authors showed the effectiveness of the external valve repair both in the deep and in the superficial venous system when this early pathologic condition has been identified. Current duplex scanning images can demonstrate the presence of mobile valve flaps at the sapheno-femoral junction. This examination is crucial in order to plan external surgical repair.
On the other hand, the present series suggests a low rate of SFI with high resolution B-Mode evidence of mobile and of uniform length valve leaflets, thus suitable for the proposed operation (64/421-15%). These series, with a long follow-up, demonstrate the effectiveness of this treatment when the proposed indications are observed. Sapheno-femoral recurrences are fewer than those described in literature after high ligation, although this is a verv carefully selected group, all with mild severity of the disease (clinical class 2 according to CEAP criterial) and thus not comparable with those previously reported.
In addition, another advantage with respect to high ligation is the reported higher rate of patent long saphenous veins suitable for vascular bypasses.
Schanzer comparing in a randomized study high ligalion-avulsion versus external valve-plasty-avulsion found no differences in terms of recurrences but a significant difference in LSV patency race. The present study shows 94% of long saphenous veins preserved with a mean diameter of 4 mm at the thigh, highly suitable as an arterial conduit. Furthermore the Veno-cuff device allows an easy and rapid procedure and its calibrator permits competence control before firing. We would also underline the absence of prostheses infection in the outcome. Reliability of CW Doppler for intraoperative assessment of valve competence was confirmed by intraoperacive angioscopy. Finally, it could be questionable to assess the SFJ competence in the outcome by the combination of duplex with the squeezing manoeuver. We showed above and in Figure 5 that when a varicose network with the hemodynamic pattern of Type III Shunt occurs, the simple disconnection-avulsion of the LSV tributary containing the re-entry PV is able to abolish the reflux signal in the saphenous vein without any additional treatment of the SFJ. Reflux in such a case can also be detected by means of duplex under Valsalva, and we recommend adding such a manoeuvre in the outcome evaluation of this group of patients.

CHIVA

We consider the application of CHIVA in patients with the outlet of the re-entry perforator on the LSV (Shunt I) to he successful in our long term follow-up (Table III): 92% of disappearance of varicose veins, 99% of saphenous vein patency and when the functional and cosmetic results were subjectively and objectively assessed, we registered excellent and good results (Class A and B of the scale) in 98% and 94% of the cases, respectively. This finding is not surprising and previous clinical reports of CHIVA technique showed satisfactory results for Type I Shunt but disappointing cosmetic results for persistent or recurrent varicose veins in the Type III Shunt, the Achille's heel of the CHIVA theory.
Such a situation is well apparent also in the present study, as comparing the results registered in Type I Shunt sub-group with that named, in Table III, as Type III Shunt with re-entry. In the latter the application of the original CHIVA technique maintains a reflux between the LSV and the TV containing the re-entry PV (Fig. 2). The varicose tributary, even if decreased in size, can often be visible after the operation, especially if dilated, long and superficial. Although this technique allowed appreciable functional results, the negative impact in subjective and objective assessment of cosmetic results in this sub-group of patients is well apparent in Table III. Conversely, in 35 patients with the hemodynamic pattern of Type III Shunt, and a deep or very short segment of the tributary between the LSV and the re-entry perforator, functional and cosmetic results were not significantly different from that obtained in the Shunt I (Table III). It is intuitive that the depth or the shortness of the insufficient tributary in these cases assures an acceptable cosmetic result (Fig. 4), Finally in other 92 patients with the hemodynamic pattern of Type III Shunt we treated the patients with flush ligation plus disconnection and avulsion of the varicose tributaries, in order to avoid the above mentioned cosmetic problems (Fig. 3). Such a treatment can be assimilated to the classic conservative but not hemodynamic treatment described in the literature, planned on the basis of careful pre-operative duplex mapping. The graded responses for symptomatic and cosmetic outcome were classified as successful for patients in class A and B, and, according to this classification the subjective and objective outcome was judged to be successful in 89% and 88% of cases, respectively.
Fligelstone using the same procedure after four years follow-up reported a successful subjective and objective outcome in 84% and 87% of cases, respectively and the group of Hammarsten in Sweden a successful subjective outcome in 94% of cases following the same surgical procedure. In the latter such a result was not significantly different from that assessed using stripping in the contralateral limb. As far as LSV patency is concerned, only when the re-entry PV is preserved we are capable of detecting by means of duplex a flow in LSV (Figs. 1. 2. 4). The data analysis of Table III well demonstrates this finding: -In Type I Shunt, 2 cases of LSV occlusion were due to the incorrect ligation of the saphenous trunk above the PV. In 31 cases we ligated the LSV below the PV and we preserved at least two long and patent segments of LSV, suitable for eventual grafting. In 153/186 cases we preserved a patent LSV for its whole length. -When Type III Shunt, was operated on maintaining the re-entry PV the LSV was found to be patent in 100% of cases (sub-groups Type III Shuni with re-entry and Type III Shunt shon or deep in Table III).
-Non-draining and occluded LSV were found in 20% of the cases operated without planning a reentry for the reverse saphenous flow (sub-group Type III Shunt without re-entry in Table III). Moreover, previous reports demonstrate CHIVA effectiveness, when the draining function of the saphenous vein is maintained. Non-draining saphenous vein presented in the outcome a higher rate of recurrences in addition to its obvious unsuitability as arterial conduit. In the present study, the Type I Shunt group presented 99% LSV patency and 92% varicose veins disappearance, whereas in Type III Shunt we registered a race of 88% and 84%, respectively (p<0.001).
Refluent LSVs present during muscular contraction a forward flow due to the action of the muscular pump, and during muscular relaxation a reverse flow directed towards re-entry PVs (Fig. 5a). SFJ closure causes the disappearance of the forward flow but does not interfere with the reverse flow (Fig. 7). On the other side, the ligation of the re-entry PV, or of the origin of the TV which contains it, causes the disappearance of the reverse flow but does not incerfere with the forward flow, which remains preserved (Fig. 6). In both cases the LSV is able to empty. In contrast, when reflux and re-entry points are suppressed at the same time, no Doppler-detectable flow is present, and thus the LSV becomes a "non-draining" vessel.
For this reason, we were surprised when Campanello reported a 100% saphenous vein patency in the outcome of high ligation and distal avulsion with meticulous ligalion of all incompetent perforators, localized by means of preoperative venography. This result is highly questionable since the saphenous trunk does not have any more a possibility- of emptying. Following high ligation and multiple avulsions. Rutherford reported a 21% rate of saphenous occlusion and Fligelstone only a 64% of patent and suitable LSV for bypass surgery. Finally, based on such hemodynamic considerations and on the results of CHIVA treatment of Type III Shunt a new technique, named CHIVA in two steps, was proposed. The first step is represented by the disconnection of the origin of the TV containing the re-entry PV, thus transforming the refluent LSV into a LSV with a forward How during muscular contraction, but no Doppler-detectable reverse flow during muscular relaxation (Fig. 5). The second step. represented by the section-ligature of the SFJ, can be performed when the LSV again shows a reflux due to the development of a new re-entry represented by a PV situated on the LSV or again, on a TV (Fig. 7). Based on the above mentioned results, when reflux re-appears with the hemodyna'mic patterns of type I Shunt or of Type III Shunt with short or deep re-entry (Figs. 2, 3) the second step was represented by simple high ligation (Fig. 7). This tactic was performed in 85% of cases with successful functional and cosmetic results (Table IV), even if with a very short and thus still inconclusive follow-up. We would like to underline in 15% of cases the absence of a detectable reverse flow after 12 - 18 months follow-up from the first procedure. Moreover, until now we never observed the development of an incompetent TV long and superficial. analogue to that avulsed with the First procedure. However, a larger number of treated cases and a long term follow-up is required in order to evaluate the possible role of CHIVA in two steps in varicose vein treatment sparing the saphenous vein.

Conclusions

Long saphenous vein sparing surgical procedure alternatives to high ligation and distal stab avulsion seem to be advantageous for future grafting because a higher rate of long saphenous vein paiencv was found. All the illustrated techniques suppress reflux while maintaining a saphenous flow. Following CHIVA I and the second step of CHIVA 2 procedures the LSV drains blood with a reverse flow toward the re-entry PV, whereas after EV-SFJ and the first step of CHIVA 2 with the physiological flow through the junction. Further studies are warranted in order to verify the association between draining and patent LSV and successful outcome of varicose vein treatment on one hand, and with the suitability of the vein for arterial reconstruction