Dynamic Anatomy of the Popliteal Artery: Hinge Point and Accessory Flexions

EDITOR’S NOTE: This article by Diaz et al. contains extremely valuable information. The importance of assessing the “dynamic anatomy” of the popliteal artery (and other vessels) was only recently appreciated as a result of developments with endovascular therapy and the increasing use of fracture-prone intraluminal metallic stents. The findings described by the Argentinian group should prove useful to interventionalists who are planning to perform a stenting procedure in a given patient. But even more so, they will likely have an impact on current R&D efforts and concepts surrounding stent technology for treatment of infra-inguinal disease — a very significant area in interventional medicine indeed! — Frank J. Criado, MD, Director, Center for Vascular Intervention, Chief, Division of Vascular Surgery, Union Memorial Hospital/MedStar Health, Baltimore, Maryland.* *“Editor's Note” reprinted with permission from Diaz, Villegas, Tamashiro, et al. Flexions of the Popliteal Artery: Dynamic Angiography. J Invas Cardiol 2004;16(12):712–715.

Abstract

The advent of endovascular technology has brought about a need to better understand the dynamic anatomy of the popliteal artery (PA). It has been theorized that vessel compression and movement promote development of hinge points (HPs), which ultimately lead to stent fractures in the PA. Using dynamic angiography, we have identified a HP of the PA. We established its relationship with bone structures using a geometric model. We were then able to create a system of endovascular-dynamic classification. In addition, we identified other flexions called accessory flexions. In this study, we present an update of our previous results and described data not reported previously.

Introduction

Stenting is an alternative in treating several diseases affecting the popliteal artery (PA).^1-7 Significant advances in stent technology have produced enhanced flexibility of these devices. However, fractures of self-expandable stents in the PA have been observed.^8-10 Vessel compression and movement promote development of hinge points (HPs), which are thought to play a key role in the genesis of stent fractures. In a recent study, we introduced a dynamic diagnostic method that has allowed us to obtain accurate morphologic information about the PA during knee flexion.¹¹ Using this new technique, which we call “dynamic angiography” (DA), we were able to identify HPs of the PA as the main curve observed during knee flexion and establish their relationship with bone structures. We believe that maximizing our understanding of the “dynamic anatomy” of the PA will allow for a more comprehensive approach to patients requiring stenting of the PA as well as the technology surrounding endovascular therapy. In the present study, we review and update our previous series¹¹ and present additional data regarding the dynamic anatomy of the PA not described previously.

Patients and Methods

Patients Adult patients who had been referred to our department by their primary physician for angiography of the lower extremities to diagnose arterial disease were included. Only symptomatic legs were studied. DA was not performed in patients with renal failure (serum creatinine ? 1.40 mg/dL), history of allergy or intolerance to the contrast material injected during conventional angiography (CA), and/or a segment of the PA that was not visible during CA. Methods Seventy-four DA procedures performed in the Department of Hemodynamics of the Hospital Nacional Alejandro Posadas in Buenos Aires, Argentina, between February 2000 and February 2004, were retrospectively reviewed. The study was approved by the Institutional Ethics Committee.

Angiography Procedure

DA was performed immediately after CA, using the same arterial access. An infusion of nonionic, low-osmolality contrast in 2 boluses of 10 cc each in 1 second was administered through a pump. Images were captured with 35 mm film at 30 photograms/second. The DA had a static phase and a dynamic phase. During the static phase, the patient laid on the side to be studied, with the knee in flexion at 100° (flexion of 0° is defined as the axis of the leg that continues to the axis of the thigh).¹¹ Focus was on the knee joint, using lateral and medial projection. Once opacification of the PA had occurred, the leg was passively extended to total extension (dynamic phase). Evaluation of the artery took place during the entire recording process using the same projection and focal point. Definitions HP of the PA: First curve in the PA in an acute angle toward the femur that appeared during knee flexion. Accessory flexion (AF) of the PA: Any curve in the PA (other than the HP) identified during knee flexion.

Interpretation of DA

Images were evaluated with a Tagarno projector. Dynamic morphology of the PA was analyzed by running the film in antegrade and retrograde directions and evaluating each of the photograms. Geometric model A geometric model was developed to establish a relationship of proximity between the medial supracondylar tubercle MSCT of the femur and the HP. With the film stopped at the moment of maximum flexion, we drew a circle with its center on the MSCT. The radius of this circle was 3 times the diameter of the PA. We then evaluated whether or not the HP was inside this circle.¹¹

Statistical Analyses

Univariate and multiple logistic regression analysis were used to analyze the presence of AFs and how they are related to risk factors for arteriosclerosis (ie, age, hypertension, smoking, dyslipidemia, and diabetes mellitus). Analyses were performed using SAS software, version 8.2.

Results

We evaluated 74 PAs in 68 patients. All patients were able to bend their knees to 100°. Presence of an HP was identified in 73 out of 74 PAs (98.6%). The HP was the main and first curve that appeared during knee flexion moving toward the femur. In addition, the pre-hinge point (pre-HP) and the post-hinge point (post-HP) segments were defined. We were unable to identify the presence of an HP in only 1 PA, as it was visualized as an extensive curve. HPs were never observed at the level of the joint line of the knee joint. The joint line observed in the CA corresponded with the post-HP segment of the PA in the DA in all cases. We established a relationship of proximity between the MSCT and the HP using the previously published circle geometric model.¹¹ We assessed the relationship in 58 DAs (78%) in which we had identified the MSCT. We found an HP within an area equivalent to 3 times the diameter of the PA in 57 DAs (98.2 %) and within an area equivalent to 2 times the diameter of the PA in 48 DAs (82.7 %). In addition, we found that the upper edge of the patella and the MSCT were aligned on the horizontal plane. Interestingly, we identified AFs when the knee was bent in 52 of 74 PAs (70.3 %). We identified a total of 87 AFs in 52 PAs. Fifty AFs occurred in the pre-HP segment of the PA; 46 of them (92%) were moving away from the bone surface (posterior convexity). Thirty-seven AFs were identified in the post-HP segment of the PA moving away (57%) and close (43%) to the joint surface. AFs were always observed in the presence of an HP. In those cases in which an artery had more than 1 flexion of the PA, the DA allowed us to differentiate the HP from the AFs because the HP was the first to appear in a semiflexion position. There were no complications associated with the DAs. The association between presence of AFs and high blood pressure was significant by both univariate and multivariate analyses (P = 0.0018 and 0.0051). None of our patients were obese.

Discussion

Because of its anatomic location, the PA is an artery exposed to movement and external forces, including compression, torsion, elongation, and flexion.^2,12 The dynamics of this vessel may influence the results of stenting procedures and present a challenge to stent technology for treatment of infrainguinal disease. Self-expandable stents are advantageous in mechanically exposed areas of the human body.^8,12 However, fractures of self-expandable stents in the PA have been reported.^8-10 In addition, severe stent fractures have been associated with stent restenosis or reocclusion.⁸ These fractures have been attributed to repeated external compression, increased length of the stented vessel, and stent overlaps that ultimately result in abnormal HPs,¹⁰ particularly when the stent is placed at flexion points of the vessel.^8,9 In the current study, we identified presence of an HP in the PA (defined as the first curve moving toward the femur). We also established a relationship of proximity between the HP and the MSCT using a geometric model. These bone parameters allowed us to correlate our findings with those obtained from CAs for each patient.¹¹ The DA allowed us to observe the sequential development of multiple flexions (HP and AFs) that occur as the knee bends. This observation is similar to those of Avisse et al.,¹³ Vernon et al.,¹⁴ and Zocholl et al.,¹⁵ who demonstrated that with increasing degrees of knee flexion, flexions of the PA (HP and AFs) appear to be more pronounced. Using DA, we introduce a dynamic anatomic concept (from an endovascular perspective) and describe presence of an HP and division of the PA into pre-HP and post-HP segments, as opposed to describing it in 3 segments (superior or supraarticular, middle or articular, and lower or infraarticular), as described by classic (static) anatomy, with the PA in the extension position. Other AFs appeared during knee bend in 70% of the PAs studied. Eighty-seven AFs were observed. Fifty-seven percent appeared in the pre-HP segment and 43% appeared in the post-HP segment. Avisse et al.¹³ and Vernon et al.14 describe the superior or supraarticular PA that extends from the adductor ring to the superior border of the femoral condyle as an “adaptation zone.” In our series, we observed that presence of an AF in the pre-HP moving away from the femur along with the HP moving toward the femur, give the PA an “S” shape that represents adaptation of this vessel to the joint flexion in this region. However, presence of AFs in the post-HP moving away and toward the joint are in contrast to the concept that the medial and distal anatomic segments of the PA play a minimal role in the origin of flexions¹³ Taken together, our findings suggest that, at least in this patient population, the HP, the pre-HP, and the post-HP segment are free and allow for adaptation of the PA to flexion of the knee. We conclude that the PA makes efficient use of the space available, following 2 basic dynamic-anatomic models: 1) A simple scheme represented by an HP; 2) A complex scheme, which occurs when more than one flexion is present during knee flexion (HP plus AFs). In conclusion, we have summarized the morphologic changes of the PA during knee bend in 2 types of flexions (the HP and AFs) to simplify the commonly used description of the PA as a “tortuous vessel.” This information may be useful in the technology surrounding endovascular therapy, which ultimately will benefit patients who require stenting of the PA. Further studies are needed to assess the clinical implications of our findings.

Acknowledgements The authors would like to thank Alejandra Moranden-Brown and Simone Matlock for their enthusiasm and their belief that our efforts were worthwhile and useful; Lydia Dong for her excellent work in statistical analyses; and the nursing, secretarial, and technical staff of the catheterization laboratory of the Hospital Nacional Alejandro Posadas for their excellent work. Address for correspondence: Jose Antonio Diaz, MD, Hospital Nacional Alejandro Posadas, Pte. Illia S/n y Marconi, EL Palomar (CP 1706), Pcia. de Buenos Aires, Argentina. E-mail: jadhemosurg@hotmail.com