Nontraumatic ruptures of the quadriceps tendon are a problem in middle-aged individuals1-8. The density of vascularity of the tendon may explain the pattern of spontaneous rupture of this tendon2,6. In 1958, a study by Scuderi revealed that tears of the tendon occurred in its central portion and were localized between 1.3 and 2.0 cm from the superior pole of the patella; no disruption at the "periosteal attachment" of the tendons was observed8. This finding has been corroborated by other authors9.
Inadequate blood supply of the tendons and ligaments of the knee has been a matter of study and discussion10-12, and hypoxia and decreased blood supply have been proposed as risk factors for tendon ruptures. In 1991, Kannus and Józsa analyzed biopsy specimens from eighty-two ruptured quadriceps tendons and observed hypoxic degenerative tendinopathy to be the most frequent pathological change2. Previous reports have focused more on the quadriceps muscle13,14, the patella10,11, the vascularity of the patellar tendon11,12, and tendinitis15-17. In a study reported in 1999, Petersen et al. identified a nonvascular area, oval in shape, at the deep portion of the quadriceps tendon6.
An improved understanding of the vascular anatomy of the quadriceps tendon and the tendon's capacity for healing might improve the planning of surgical repairs and reconstructions of a ruptured tendon as well as help surgeons to avoid compromising its vascularity during surgery5.
The purpose of this study was to describe the vascular anatomy of the quadriceps tendon and to correlate the density of the vasculature with the clinical patterns of rupture.
Twenty fresh human cadavers of donors who were an average of sixty-five years old (range, forty-eight to seventy years old) at the time of death were injected with a lead oxide-gelatin mixture through the right femoral artery, with a technique originally described by Rees and Taylor18 and modified by Tang et al.19,20. Fresh specimens were obtained within forty-eight hours after the death of the donor through the Dalhousie University Human Body Donation Program. The project was approved by the Dalhousie University Ethics Committee. Exclusion criteria were evidence of severe peripheral vascular disease, extensive muscle atrophy, and extensive surgical scars.
Foley catheters of appropriate sizes were inserted into the femoral artery proximally and distally by means of a longitudinal arteriotomy. The femoral vein was cannulated with a standard metallic embalming cannula that was large enough to accommodate the passage of large blood clots. A solution of tap water and potassium acetate was then injected through the femoral artery under continuous perfusion at 140 to 170 kPa until the venous outflow was clear.
The lead oxide injectate is prepared as follows. Five grams of 300 bloom pharmaceutical-grade gelatin derived from porcine skin (catalogue number G-2500; Sigma, St. Louis, Missouri) are diluted in 100 mL of tap water (at a temperature of 40°C). Subsequently, 100 g of water-soluble red lead oxide (CAS 1314-41-6; Carterchem, Montreal, Quebec, Canada) is added. The solution is stirred at regular intervals to prevent sedimentation of the lead oxide. During the process, the cadavers are placed into a waterproof 210 × 75 × 75-cm tank that is partially filled with warm (40°C) water. Adequate perfusion of the cadaver is indicated by the presence of the orange injectate in the sclera, fingers, and toes. The average amount of lead oxide mixture injected is between 20 and 30 mL/kg, depending on the cadaver's degree of obesity. The cadavers are then cooled for twenty-four hours at 4°C to allow the gelatin to solidify.
The anatomical dissection was done according to the anatomical landmarks of the knee extensor mechanism. The vascular anatomy of the different segments was outlined, dissected in coronal as well as sagittal planes, and radiographed. The angiograms were standardized at 44 kVp, 15 mA, and a distance of 102 cm (40 in). Of the forty quadriceps tendons in the twenty cadavers, thirty-three were used in our analysis. Seven tendons were not used because of poor dissection and rupture of vessels during the gel infusion. The angiograms were scanned and were analyzed with a digital picture statistical software analyzer (Scion Image for Windows; Scion, Frederick, Maryland). Pixel maps based on the radiopacity of the vessels of the tendon were created. The mean gray pixel value (MGPV) of three different selected areas of interest was studied for each of the thirty-three tendons analyzed. Gray values range from 0 to 255, with 0 being white and 255 being black. Areas of 2400 square pixels, organized in two-dimensional arrays of 80 × 30 pixels, were computed for statistical analysis.
Statistical Methods
A one-way analysis of variance was performed, and significance was determined with use of a Tukey HSD (honestly significant difference) post-hoc analysis (p < 0.05). We analyzed the MGPV (the variable) of the three different zones in the same tendon. We used an analysis of variance in order to evaluate whether there were differences between the means of the MGPV across the three vascular zones of the tendons. Analysis of variance was performed to compare the means for each of the zones in each tendon. Our observations, based on the MGPV, are described in terms of variation of the zone values among their group means. Tukey analysis added more power to the study, obtaining a true difference with use of a 95% confidence interval among the multiple comparisons of the MGPV means among the zones.
Vascular Arcades and Anatomical Vascular Zones of the Tendon
The blood supply to the quadriceps tendon is provided by three major vascular arcades—i.e., the medial, lateral, and peripatellar arcades. The blood supply comes from these arcades located at the periphery of the tendon and communicating with each other through a complex network of blood vessels that reach the central portion of the tendon. On the basis of the density of the vascularity that we observed, we divided the tendon into three vascular zones. Zone 1 is located between 0 and 1 cm from the superior pole of the patella, zone 2 lies between 1 and 2 cm from the superior pole of the patella, and zone 3 is located >2 cm from the superior pole of the patella. Zone 2 was found to be less vascular than zones 1 and 3.
Vascular Arcades
Medial Arcade
The medial side of the tendon is supplied by the medial arcade. This vascular arcade is created from proximal to distal by the anastomosis of muscular branches of the middle and inferior arteries of the vastus medialis divisions of the femoral artery, the descending genicular artery branch of the superficial femoral artery, and the superior medial genicular artery division of the popliteal artery. This arcade is located at the musculotendinous junction, and it sends an intratendinous vessel that reaches the rectus femoris tendon and the vastus medialis/lateralis conjoined tendon. Some vessels cross the tendon from medial to lateral, creating small anastomoses with descending branches of the axial artery of the rectus femoris and with the lateral arcade arteries. The medial arcade runs between the muscular portions of the vastus medialis and the rectus femoris and supplies the vastus medialis tendon and the medial aspects of the rectus femoris and vastus intermedius tendons (Fig. 1). Distally, the medial arcade sends branches to the peripatellar vascular ring and the patella. In some specimens, a deep as well as a superficial medial arcade was identified (Fig. 2).
Lateral Arcade
The lateral side of the tendon is supplied by the lateral arcade. This arcade is formed by the long descending branch division of the lateral circumflex femoral artery (which arises from the distal perforating artery of the vastus lateralis branch of the deep femoral artery) and by divisions of the superior lateral genicular artery (a branch of the popliteal artery). It runs between the rectus femoris tendon and the vastus lateralis tendon, sending small intratendinous branches that anastomose with those of the medial arcade below the rectus femoris tendon. This arcade sends vessels toward the superior pole of the patella, contributing to the formation of the proximal peripatellar vascular ring and the patella (Fig. 2).
Peripatellar Vascular Ring, or Peripatellar Arcade
Deep to the vastus intermedius and the capsule, a fat-pad structure was found to be rich in vascular components, arising from the deepest branches of the superior genicular arteries and the lateral and medial arcades. This arcade supplies the distal 1 cm of the tendon. Small longitudinal and transverse arterioles running below and onto the vastus intermedius insertion were noted. These branches send intratendinous vessels that reach the middle of the vastus lateralis/medialis conjoined tendon and the rectus femoris tendon. There is a very complex communication among the three different layers of the tendon, increasing the density of vascularity at the insertion of the tendons onto the superior pole of the patella. This arcade anastomoses with the peripatellar vessels through intratendinous branches. This area of rich vascular density is named the anatomical vascular zone 1 of the quadriceps tendon (Figs. 2 and 3).
The pattern of blood supply to the tendon is centripetal in nature. This supply may be regarded as a triangle, with each limb representing one of the arcades. The triangle is more vascular near the boundaries and less vascular at the center (Figs. 1 and 2).
Anatomical Vascular Zones of the Quadriceps Tendon
Each of the vascular arcades contributes to the blood flow of the tendon, and different zones can be analyzed and described on the basis of the density of the vessels that are observed (Figs. 4 and 5).
Zone 1
This zone is found between 0 and 1 cm from the superior pole of the patella. It contains the peripatellar vascular ring complex with intratendinous anastomoses. The MGPV (and standard deviation) was 139.51 ± 19.57, which identified this area as highly vascular. All of the units of the tendon receive good blood supply at their insertion onto the bone. The MGPV was found to differ significantly between zone 1 and zone 2 (p < 0.05) as well as between zone 1 and zone 3 (p < 0.05).
Zone 2
This zone is intermediate between zones 1 and 3 and is located between 1 and 2 cm from the superior pole of the patella. It was found to be hypovascular mainly in the central third, at the vastus medialis/lateralis conjoined tendon and the vastus intermedius tendon. The MGPV was 89.51 ± 19.30, which was significantly different from the MGPV in zone 3 (p = 0.02).
Zone 3
This zone is rich in vascular supply derived from the medial and lateral arcades and in close relation with the musculotendinous junction of the vastus medialis and the vastus lateralis. This zone is located >2 cm from the superior patellar pole. The MGPV was 103.16 ± 19.70.
Vascular Differences Among Subunits of the Tendon
The dissections and radiographic analysis of the various subunits of the tendon revealed that the vascular density of the rectus femoris tendon is higher than the vascular density of the middle unit (the junction of the tendinous portions of the vastus medialis/lateralis) as well as that of the deepest unit (the vastus intermedius tendon). This is due in part to the presence of the paratenon and the superficial distribution of the vessels over the rectus femoris tendon (Fig. 6).
This analysis of the pattern of blood supply of the quadriceps tendon showed that there is a characteristic distribution of blood vessels, creating areas of vascular density, and different vascular zones can be appreciated within the tendon. Vascular arcades were found close to the musculotendinous junction and the patellar insertion. These arcades offer intratendinous arteries that travel from the periphery toward the center of the tendon. Areas of the tendon adjacent to the vascular arcades showed significantly greater vascularity close to the patellar insertion and close to the musculotendinous junction. The poorest vascularity was noted in a zone located between 1 and 2 cm from the insertion of the patella.
The blood supply of the tendons and ligaments of the knee has been a matter of study and discussion2,6,11. Hypoxia and decreased blood supply have been proposed as risks for tendon rupture. As mentioned, Kannus and Józsa2 analyzed, in 1991, biopsy specimens of eighty-two ruptured quadriceps tendons and found the most frequent pathological change to be hypoxic degeneration. Histological analysis of the blood vessels of the tendon and paratenon revealed narrowing or obliteration of the lumina of the arteries and arterioles2.
Scuderi reported, in 1958, that tears in the quadriceps tendon frequently occur in the tendon's central portion, between 1.3 to 2.0 cm from the superior pole of the patella8. No disruption at the "periosteal attachment " of the tendon was identified in their group of patients (average age, 48.5 years). In 1933, McMaster studied the causes of traumatic and spontaneous ruptures of tendons and muscles21. He found that obstruction of the blood supply of a normal rabbit tendon caused rupture of the tendon, suggesting that degenerative changes and diseases might predispose tendons to rupture as the result of only slight strain. Siwek and Rao pointed out that ruptures of the quadriceps tendon were more prevalent in older patients22.
Our study showed that the quadriceps tendon is hypovascular in zone 2. This finding correlates closely with the common pattern of rupture described in the literature. The attachment of the tendon to the superior pole of the patella (zone 1) was found to be well vascularized, and ruptures at this site are rare. In another recent study on the vascular status of the quadriceps tendon, Petersen et al.6 showed an avascular zone located 2 cm proximal to the patellar insertion. Using the Spalteholtz technique, they identified an oval avascular area measuring 30 × 15 mm in which there was no immunohistochemical demonstration of laminin, indicating avascularity. This zone was found in the deep layer of the tendon, the one that is in contact with the patellofemoral groove. The study by Petersen et al. offers a detailed anatomical and histological description of the vasculature. Our technique provides not only a detailed description of the vasculature, but also a quantitative and reproducible method with which to describe the differences among the vascular zones.
Our study confirms the findings of Petersen et al.6 in that there was a hypovascular and avascular area in the deepest portions of the tendon. Also, the superficial layer of the tendon was shown to be more vascular than the deepest layer (Fig. 6). In 1967, Scapinelli described a vascular anastomotic ring in the connective tissue that surrounds the patella23. Our findings are in agreement with his work as well.
In 1994, Raatikainen et al. reported on a series of partial ruptures of the quadriceps tendon, mainly involving the deepest portion and compromising the medial side and the middle third of the tendon, in patients with an average age of twenty-eight years7. This observation correlates with our findings of a decreased number of vessels in the vastus intermedius in comparison with the rectus femoris, the most superficial layer of the tendon. This finding suggests that the articular side of the tendon may be exposed to compressive forces produced by the femoral condyles that contribute to diminished vascularity of the deepest portion of the tendon6.
We introduced the concept of vascular arcades of the quadriceps tendon in order to explain the characteristic vascular distribution of the vessels supplying the tendon. The centripetal distribution of the blood supply, reflected by the hypovascularity of the central zone (zone 2), correlates with the most frequent pattern of rupture of the tendon reported in the literature. Knowledge of the location of these hypovascular zones of the tendon may lead in the future to operative or nonoperative procedures to increase vascularity to these areas. Also, during reparative and reconstructive procedures on the tendon, the surgeon should avoid injury to the vascular arcades close to the musculotendinous units of the vastus medialis and vastus lateralis as well as the peripatellar vascular arcade because such an injury could affect the rate of tendon healing. Furthermore, clinical studies examining repairs of chronic tears of the tendon perhaps should focus on the use of the rectus femoris tendon as a rotational vascularized graft instead of the conventional V-Y techniques, which could further jeopardize these vascular arcades. This improved understanding of the anatomy of the vascular arcades may decrease the rate of complications and enhance, through directed surgery, the chances of preserving the vasculature of the tendon. 