RESULTS
Overall Results

No death or major complication was observed in this series. Partial thrombosis of the GSV occurred in 72 cases (24.9%). It spontaneously recanalized in 68 cases, at a mean follow-up of 5 months. Secondary surgical treatment or sclerotherapy of collateral saphenous veins was performed for cosmetic reasons in 27 cases. Almost all these cases involved collateral veins in the thigh with low reentry points, and hence partially visible after the procedure. All secondary procedures were performed using the same technique after marking by means of Doppler ultrasound.
Excellent improvement was obtained in 119 cases (41.2%); no visible varice was observed at any time during follow-up. Good improvement was obtained in 124 cases (43%), with residual varices and neoformation of vessels being minimal. Fair improvement was obtained in 41 cases (14.1%); the situation was better compared to the preoperative status but residual varices and/or neovessels were observed. Poor improvement was obtained in five cases (1.7%) in which the situation did not change or worsened.
Recurrences involved type A neovessels that were clinically visible in 42 cases (14.5%) and detectable by Doppler ultrasound in 72 cases (24.9%), and type B neovessles that were clinically visible in 21 cases (7.2%) and detectable by Doppler ultrasound in 30 cases (10.3%).

Table I. Comparison of patients in the D.ACHM AND ND.ACHM groups
	D.ACHM	ND.ACHM	p
Procedures (n)	198	91
Gender	145 F	65 F	NS
Mean age (years)	48.4 ± 11.7	50.8 ± 13.6	NS
Size of varices
0	2	0	< 0.001
1	66	17
2	108	48
3	22	26
Saphenous vein diameter
Patients (n)	177	77	NS
Mean diameter	6.7 ± 1.7	7.1 ± 2
Obese patients	25	13	NS
PDVI	19	11	NS
SDVI	2	2	NS
Phlebectomy	19	2	NS
NS, not significant; PDVI, primary deep vein insufficiency; SDVI, secondary deep vein insufficiency.

Table II. Comparison of outcome after D.ACHM and ND.ACHM groups
	D.ACHM (198 procedures)	ND.ACHM (91 procedures)	p
GSV thrombosis	20	52	<0.001
Improvement			<0.001
Excellent	93	26
Good	89	35
Fair	13	28
Poor	3	2
Recurrences
Type A neovessels
Clinically visible	10	32	<0.001
Detectable by ultrasound	18	54	<0.001
Type B neovessels
Clinically visible	13	8	NS
Clinically visible	13	8	NS
NS, not significant.

Results in the D.ACHM and ND.ACHM Subgroups

The D.ACHM group identified by Doppler ultrasound included 198 procedures and the ND.ACHM group included 91 procedures. These two groups were comparable with regard to age, gender, incidence of valve incompetence in the deep venous network, and incidence of obesity (Table I). Comparison of the mean preoperative diameter of the greater saphenous vein showed no significant difference between the D.ACHM and ND.ACHM groups. Nor was there any significant difference between the two groups with regard to the number of secondary phlebectomies carried out on collateral sapheous veins (R3) for cosmetic reasons. The incidence of thrombosis was 10.1% (20 cases) in the D.ACHM group versus 59% (52 cases) in the ND.ACHM group (p < 0.001). As shown in Table II, there was a highly significant correlation between saphenous vein drainage and neoformation of clinically visible or ultrasound-detectable vessels. Qualitative analysis showed no statistical difference between the D.ACHM and ND.ACHM groups with regard to formation of clinically visible or ultrasound-detectable type B neovessels.

Table III. Comparison of size of varices and diameter of GSV in patients with or without recurrence in D.ACHM and ND.ACHM groups
	D.ACHM			D.ACHM
	No recurrence	Recurrence		No recurrence	Recurrence
Procedures	175	23		33	58
Size of varices
0	2	0		0	0
1	60	6		10	7
2	95	13	NS	14	34	NS
3	18	4		9	17
GSV diameter
n	158	19		31	46
Diameter	6.6 ± 1.7	6.9 ± 19	NS	6.8 ± 1.6	7.6 ± 2.1	NS
NS, not significant

Conversely, there was a significant difference (p < 0.001) between the two groups with regard to formation of clinically visible or ultrasound-detectable type A neovessels. Within the D.ACHM and ND.ACHM groups, comparison of patients in whom recurrence was or was not observed showed no significant difference with regard to the size of varices or diameter of the GSV (Table III). In 49 cases in the ND.ACHM group, development of a perforating reentry vein along the distal GSV (P/PR2) led to saphenous vein drainage spontaneously within a period of 6 to 24 months after the procedure. In 15 of these cases, formation of type A vessels was observed before develoment of the reentry perforating vein. In contrast, formation of type A vessels was observed after development of the reentry perforating vein in only two cases. This difference was statistically significant (p < 0.005; McNemar test).

DISCUSSION

The findings of this study show that the determinant factor for outcome of ACHM is the quality of drainage of the saphenous vein system. The incidence of recurrence due to formation of type B neovessels was the same in the D.ACHM and ND.ACHM groups. Conversely, the incidence of recurrence due to formation of type A neovessels was significantly higher in the ND.ACHM group. This finding demonstrates that outcome depends on the quality of drainage of the saphenous vein system. This study also showed that secondary development of a reentering perforating vein can transform a nondraining system into a draining system, thereby greatly reducing the likelihood of formation of type A neovessels.
To obtain a draining saphenous system, ACHM strategy must take into account the hemodynamic pattern of the venous network. The original procedure proposed by Francheschi 5 for all types of venovenous shunts was, in fact, only one possible ACHM technique (ACHM-1) (Fig. 3). After identification of the perforating vein most likely to provide the best reentry point, the saphenous vein was divided and tied at the saphenofemoral junction to prevent reflux at the most proximal level. Collateral veins from the junction to the saphenous vein were left patent so as to insure optimal retrograde flow (at low pressure and low flow). At the same time, the vessel giving rise to the reentering perforating vein was interrupted about 2 cm below the mouth of the perforating vein. In this way, the reentering perforating veins was “terminalized” and the hydrostatic pressure column of the superficial venous system was relieved (Fig. 3). In cases involving nonterminal, refluxing perforating veins,9 terminalization included elimination of secondary venovenous shunts generated by the perforating veins. This technique (ACHM-1) achieved good results for patients with type I shunts (Fig. 3), which account for about 33% of the total population with varicose veins. Conversely, results were poor in patients with type 3 shunts, which account for about 60% of the total population (Fig. 4a). The drawback of ACHM-1 in patients with type III venovenous shunts was the preesence of secondary shunts gen-erated by R2-R3 reflux from the collateral that was not disconnected from the saphenous vein (Fig. 4h). This cause of failure was obvious in patients in whom this collateral was superficial (R3) and thus visible after the procedure.
This problem led Franceschi and Bailly 6 to develop a two-stage technique (ACHM-2) for type III shunts. The first stage was designed to modulate venous hemodynamics by disconnecting the origin of the collaterals giving rise to the reentering perforating vein that allowed retrograde reflux (P/R3). The goal of this step was to transform a refluxing saphenous system into a saphenous system with anterograde reflux in which the point of reentry was the junction (Fig. 4b). Following the first step, reflux should no longer be detected in the junction by Doppler ultrasound after compression and release (Fig. 4b). The second stage was carried out after formation of a functional reentry vein, thanks to the development of a P/R2 perforating vein. At this time, the system can be considered a refluxing system similar to that of a type I shunt (Fig. 4c) that can be treated by crossectomy (Fig. 4d).
However, ACHM-2 is not safe in patients presenting with type 3 short saphenous vein insufficiency, i.e., confined to the upper half of the thigh. In these cases, in fact, ACHM-2 carries the risk of extension of thrombosis from the upper saphenous vein into the common femoral vein. To prevent this complication, Franceschi and Bailly 6 proposed use of a hybrid technique (ACHM-1+2) involving simultaneous ligation of the origin of the collateral vein giving rise to the reentry perforating vein and the saphenofemoral junction (Fig. 4f). This procedure results in a saphenous vein system with practically no drainage presenting a high risk of thrombosis of the saphenous vein. Subsequently, the vessel can be recanalized by either formation of P/R2 perforating vein (Fig. 4d) or neoformation of vessels (Fig. 4g) at the level of the ligation of the collateral vein giving rise to the reentry perforating vein (ligature “bridging”). Recanalization by formation of a P/R2 perforating vein leads to a spontaneously draining saphenous system with a low propensity for formation of neovessels. Conversely, recanaliation by ligature bridging creates the need to empty the system and thus leads to formation of type A neovessles that may connect with the previous net-work or a new P/R3 perforating vein. In any case, since it results in a nondraining system, ACHM-1+2 is a conservative but nonhemodynamic technique and hence does not qualify as a true ACHM procedure. The main difference between ACHM and other conservative treatment techniques is thus clear. In over 50% of cases, crossectomy in association with interruption or phlebectomy of collaterals leads to a nondraining saphenous system with a higher incidence of neoformation of vessels. Conversely, like valvuloplasty of the saphenofemoral junction, successful crossectomy or isolated ligation of the saphenous vein at the saphenofemoral junction leads to a draining saphenous system but only treats the point of reflux in patients with primary shunts (R1- R2). Secondary shunts (R2-R3 and R2-R4) resulting from incompetent collaterals or refluxing perforating veins remain patent.2
Stripping of the saphenous vein is a good surgical procedure in terms of simplicity, rapidity, and safety. It does not require long hemodynamic evaluation or complicated mapping and the operative technique is well defined. By definition, however, stripping prevents drainge from the territory of the GSV. We thus sought to determine if the benefits of ACHM outweighed the long period of time required for hemodynamic mapping and the high investment needed for equipment.
Previous authors reporting on the results of stripping only took into account clinically visible neo-vessels. Examinations for detection of infraclinical neovessels were not performed. In our series, post-operative follow-up included both clinical examination and Doppler ultrasound, so we were able to monitor total neoformation, including both clinically visible and infraclinical neovessels.
In the past three decades, only a few authors 4,10-14 have described 3-year results of stripping of the saphenous vein. Comparison of these results with our overall results, i.e., including those from both the D.ACHM and ND.ACHM groups, indicate that ACHM compares favorably with stripping. It is likely that this comparison would probably have been even more conclusive for ACHM if only D.ACHM had been taken into account.15
These findings demonstrate that hemodynamic study, which is the basis for ACHM, is an effective method for evaluation of varicose veins. The validity of current models is confirmed by the fact that statistical methods can predict certain potential events, e.g., formation of type A neovessels. The major findings of this study can be summarized as follows:

1. The preoperative diameter of the GSV is not correlated with formation of neovessels and thus with outcome.
2. Secondary procedures carried out on R3 for cosmetic reasons do not affect the outcome of ACHM.
3. The quality of drainage from the treated GSV is the only factor correlated with outcome of ACHM.
4. Formation of type A neovessels in nondraining saphenous systems can be predicted in terms of topography and delay.
5. Formation of type A neovessels transforming a ND.ACHM into a D.ACHM and leading to a reduction in neoformation can be predicted by statistical methods.
6. Formation of type B neovessels is not correlated with the quality of drainage of the treated GSV. The incidence of type B neovessels is about 10% regardless of whether the saphenous vein system is draining or nondraining. Although this type of neoformation probably has a strong influence on the outcome of any type of surgical treatment, it has not been described in the literature. The most likely reason for this oversight is that postoperative Doppler ultrasound has not been widely used to detect infraclinical neovessels.
7. Results of ACHM at 3 years are at least as good as those reported with the same follow-up for stripping of the saphenous vein.

REFERENCES

1. Einarsson E, Eklof B, Neglen P. Scherotherapy or surgery as treatment for varicose veins: a prospective randomized study. Phlebology 1993;8:22-26.
2. Eklof B. Modern treatment of varicose veins. Br J Surg 1988; 75:297-298.
3. Haeger K. Five-year results of radical surgery for superficial varices with or without coexistent perforator insufficiency. Acta Chir Scand 1966;131:38-49.
4. Taulaniemi E. On the treatment of varices. Acta Chir Scand 1963;125:429-432.
5. Franceschi C. Théorie et Pratique de la Cure Conservatrice et Hé modynamique de l’Insuffisance Veineuse en Ambu-latroire. Pré cy-sous-thil: Editions de l’Armanc¸ on, 1988.
6. Bailly M. Résultats de la cure CHIVA. In: Cormier JM, Fichelle JM, eds. Techniques et Stratégie en Chirurgie Vasculaire. Paris: AERCV, 1992, pp 255-271.
7. Gorny P, Blanchemaison P, Chahine D, et al. Chirurgie conservatrice et ambulatoire. Phlébologie 1995;48:2:255-259.
8. Glantz SA. Primer of Biostatistics. New York: McGraw-Hill, 1992.
9. Cappelli M, Ermini S, Turchi A, Bono G. Considérations hemodynamiques sur les perforantes. Phlébologie 1994;47: 389-393.
10. Hobbs JT, Surgery and sclerotherapy in the treatment of varicose veins. Arch Surg 1974;109:793-796.
11. Jakobsen BH. The value of different forms of treatment for varicose veins. Br J Surg 1979;66:182-184.
12. Larson RH, Lofgren EP, Myers TT, Lofgren KA. Long-term results after vein surgery. Study of 1000 cases after 10 years. Mayo Clin Proc 1974;49:114-117.
13. Laurikka J, Sisto T, Salenius JP, et al. Long saphenous vein stripping in the treatment of varicose veins: self and surgeon assessed results after 10 years. Phlebology 1994;9:13-16.
14. Rivlin S. The surgical cure of primary varicose veins. Br J Surg 1975;62:913-917.
15. Capelli M, Molino Lova R, Ermini S, et al. Comparaison entre cure CHIVA et stripping dans le traitement des veines variqueuses des membres inférieurs. J Mal Vasc 1996;21:40- 46.