3Functional Anatomy of the Knee 3

3 Functional Anatomy of the Knee

D. G. Eckhoff

Summary

The purpose of this chapter is to identify the function- al anatomy that impacts the reconstruction of the arthritic knee with a prosthetic implant. This work does not attempt to review all the detailed soft-tissue anato- my of the knee that is covered more expansively both in description and illustration in other resources. It focus- es instead on bone morphology of the knee. The con- clusion is that morphological features of the knee are largely asymmetrical, and these features are related in both linear and angular relationships to one another in a way that will impact the function of the prosthetic re- placement.

Introduction

The knee is defined in this chapter as composed of two parts, the soft-tissue sleeve and the underlying bony architecture. The soft-tissue sleeve extends from hip to ankle and invests the bony architecture. The bony archi- tecture, both normal and pathological, is the focus of this anatomical review of the knee.

Soft-tissue Sleeve

Protection and nutritional support of the knee are pro- vided by skin,fat,capsule,and synovium.Located in these soft tissues is a network of vessels (arteries, veins, lym- phatics) and nerves. In general terms, the vessels and nerves pass from the hip to the ankle along the posterior aspect of the limb and send branches both medial and lat- eral around the knee to meet near the anterior midline.

This anatomical feature allows surgical exposure of the knee from the anterior aspect with minimal risk to neurovascular structures.A full appreciation of the three- dimensional location and relationship of the nerves and vessels to each other as well as to other soft tissues of the knee is beyond the scope of this dissertation, and is best obtained by inspection of the Visible Human (http://www.visiblehuman.org).

Muscle-tendon units lie in the soft-tissue sleeve and are a significant component of the functional anatomy of the knee.The quadriceps (rectus femoris,vastus lateralis, vastus intermedius, vastus medialis) and articularis genu lie anterior to the femur. They arise from the pelvis (rec- tus femoris), the proximal femur (vastus lateralis, vastus intermedius, vastus medialis), and distal femur (articu- laris genu),and attach by way of a conjoined tendon to the tibia to form the extensor mechanism of the knee. In- vested in the conjoined tendon is the body’s largest sesamoid bone, the patella. Retinaculum and synovium attaching to the patella and its tendon pass around the medial and lateral aspects of the knee to the distal femur and proximal tibia. Surgical approaches to the knee dis- cussed in later chapters all violate the retinacular and synovial investments of the extensor mechanism, and to a lesser extent the muscles and tendons just described.

The muscle-tendon units lying posterior to the femur are referred to collectively as the hamstrings. The lateral hamstring (biceps femoris) and the medial hamstrings (sartorius, gracilis, semitendinosis, semimembrinosis) arise from the pelvis and attach to the fibular head and medial aspect of the tibia, respectively. These muscles function collectively in knee flexion. They also function in rotating the knee, with the lateral hamstrings rotating the tibia external relative to the femur and the medial hamstrings rotating the tibia internal relative to the femur. In the arthritic knee, discussed below and else- where in this text, these muscle-tendon units become un- balanced in their effect on the knee, producing angular and rotational contractures.

Also implicated in knee contractures are the gastroc-

nemius muscles, the popliteal muscle, and the iliotibial

band. The gastrocs originate just proximal and posterior

to the femoral condyles and insert through the Achilles

tendon on the calcaneus.The popliteal muscle arises from

the posterior lateral femur and attaches to the posterior

lateral tibia. The iliotibial (IT) band arises from the later-

al pelvis and attaches to the anterolateral tibia at Gerde’s

tubercle.The latter structure,the IT band,is implicated in

an external rotation of the tibia and secondary lateral

tracking of the patella in the pathological knee. Planned

sequential release and balancing of these soft tissues,

discussed in later chapters, are integral steps in the per- formance of total knee arthroplasty.

Ligaments joining the femur and tibia are four in number, two cruciates and two collaterals. The medial collateral ligament (MCL) can be separated into two com- ponents, superficial and deep. The deep MCL originates from the area of the medial femoral epicondyle and in- serts on the mid body of the medial meniscus and the proximal medial tibial plateau,forming a confluence with the coronary ligament attaching the meniscus to the tibia. The superficial MCL has an origin similar to that of the deep MCL but lacks any attachment to the meniscus and inserts more distally along the medial tibia.The MCL slopes from posterior proximally to anterior distally. The lateral collateral ligament originates from the area of the lateral epicondyle and inserts on the fibular head.It slopes opposite the MCL, passing from anterior proximally to posterior distally. The origins of the collaterals (MCL and LCL) lie on a line joining the femoral epicondyles, also known as the epicondylar line.

There are two cruciate ligaments. The anterior cruci- ate ligament (ACL) originates from the lateral wall of the femoral intercondylar notch and inserts on the mid tibia between the articular surfaces, passing from posterior proximally to anterior distally. Passing in the opposite direction, from anterior proximally to posterior distally, is the posterior cruciate ligament (PCL), which arises from the medial wall of femoral intercondylar notch and inserts over an area approximately 2 cm in vertical length on the posterior aspect of the tibia. The origin of the cruciates (ACL and PCL) is not on the same line as the origins of the collaterals, i.e., the epicondylar line. The cruciate origins lie on a line passing through the center of the condyles, a line equidistant from points on the poste- rior articular surface of the condyles. The location and clinical significance of this line will be discussed in more detail in relation to femoral condylar geometry below,but it is important to recognize for the purpose of balancing the soft tissues and restoring the kinematics of a knee that the origins of the cruciates and collaterals are not on the same line.

Another anatomical feature of these knee ligaments worth noting is the opposite slope of the cruciates (ACL and PCL) and collaterals (MCL and LCL) described above. The clinical significance of this observation is that in the absence of the ACL, the collaterals will uncross or unwind to become more closely parallel. This occurs be- cause the tibia rotates internally relative to the femur in the absence of restraint from the ACL and/or the PCL [1].

In the course of knee replacement, one or both cruciates are removed, permitting this relative rotation of the tibia to the femur to occur, i.e., the collaterals unwind, poten- tially altering the contact pattern of the femoral and tibial components in the prosthetic knee. This issue of contact pattern and the associated issue of wear in a pros-

thetic knee are dependent on bone morphology or bony architecture of the knee, which will now be addressed.

Bony Architecture (Bone Morphology)

The distal femur has a unique three-dimensional shape marked by asymmetry. The two rounded asymmetrical prominences that articulate with the tibia, referred to as condyles, are separated by a space referred to as the in- tercondylar notch. The condyles are joined proximally by the femoral trochlear groove, the site of articulation be- tween the patella and the femur. The trochlear groove is characterized as a trough with its lowest point, called the sulcus, set between medial and lateral anterior projec- tions. These anterior projections, or ridges, are confluent with the condyles distally while the sulcus of the trochlear groove ends in the intercondylar notch. These morpho- logical features of the distal femur are covered anterior, posterior, and distal by articular cartilage.

These morphological characteristics of the distal femur have been a source of both historical and contemporary in- terest [2–8]. More than a dozen linear dimensions and half a dozen angular dimensions of the distal femur have been repeatedly measured [4,5].These measurements will not be recounted here in detail, but several documented relation- ships of functional anatomy will be highlighted. Specifical-

Femur

⊡Fig. 3-1.The Weber brothers created cross-sectional images of the femoral condyles by cutting cadaveric specimens, coating them with ink, and pressing them to paper. They found radii C1, C11, and C111 to be equal. This technique was the first to illustrate the circular profile of the condyles

3

lationship of the tibia to the femur need to be reviewed, since these are issues of functional anatomy that are integral to the practice of contemporary total knee arthroplasty.

The circular contour of the posterior condyles was first documented by the Weber brothers [2] (

).

This perception of circular geometry of the posterior condyles was challenged by Fick [3], who proposed that the condyles were more helical in shape; i.e.,he argued for a changing radius of curvature producing an instant cen- ter of flexion and extension. Fick’s interpretation still commands a large following of engineers who find it dif- ficult to reconcile the biomechanical data regarding knee motion with circular condyles. Nevertheless, abundant data now support the earlier Weber work [6]. A recent study suggests this controversy arises because authors of biomechanical studies beginning with Fick have repeat- edly selected a flexion axis perpendicular to the sagittal plane of the knee [7].While it is perhaps intuitive that the limb stays in the sagittal plane through a range of flexion and extension, there are no anatomical or kinematic data to support this idea, or the corollary that the axis of flex- ion and extension is perpendicular to the sagittal plane.

The controversy can be resolved by allowing the knee to flex about an axis not perpendicular to the sagittal plane [7]. This axis not perpendicular to the sagittal plane per- mits motion to occur about a single axis centered in the condyles and supports the concept of circular condyles.

Based on these observations, morphological studies have been conducted using modern computer techniques that confirm the circular profile of the posterior condyles, es- tablishing a single axis for flexion and extension of the knee through an arc of 10°–120° [8, 9]. This work demon- strates with careful sizing and positioning of cylinders within the condyles that the two condyles are circular in shape. It also demonstrates that the condyles share a sin- gle axis of rotation but display differing radii of curvature, with medial greater than lateral (

).

This work documents that the center of the cylinder is dif- ferent from the line joining the epicondyles (

).

Further, the data presented in this work demonstrate that the cylindrical axis, corresponding to the center of each condyle, passes through the origins of the cruciate liga- ments. As noted above, the epicondylar line incorporates the origins of the collateral ligaments, but not the origins of the cruciate ligaments.The work cited here documents that the epicondylar line and the line joining the center of the condyles are not the same. These observations of the relative relationship between the epicondylar line and the cylindrical axis based on the circular profile of the posterior condyles represent an important functional anatomical feature of the distal femur.

⊡Fig. 3-2. The cylindrical profile of the condyles can be demonstrated using computer techniques to create three-dimensional reconstructions of the distal femur from CT images with cylinders fit into the condyles. The medial cylinder (blue) is slightly larger than the lateral cylinder (red) but they share the same cylindrical axis

⊡Fig. 3a, b.The epicondylar (upper) and cylindrical (lower) axes do not lie in a single plane and are not parallel or collinear in the coronal plane (a) or the transverse plane (b)

a b

It should be noted again that the foregoing discussion of circular condyles applies to the posterior femoral condyles, i.e., that portion of the distal femur articulating with the tibia from 10° to 120° of knee flexion. The condyles articulating with the tibia in the last 10° of ex- tension have a curvature different from that of the poste- rior condyles [4,6].Further,the anterior or trochlear por- tion of the distal femur demonstrates yet another curva- ture different from the condyles. It is not the curvature of the trochlea, however, but the location and orientation of its sulcus that plays a role in functional anatomy and mer- its further attention.

The location and orientation of the sulcus have been carefully documented both in cadavers [10] and on radi- ographs [11]. The sulcus of the trochlear groove lies later- al to the midplane of the distal femur and is oriented be- tween the anatomical and mechanical lines of the femur in the coronal plane (

).

The anatomical line of the femur passes up the femoral shaft from the center of the distal femur to the greater trochanter (Fig. 4a). The mechanical line passes from the center of the distal femur to the center of the femoral head (Fig. 4b). Relative to these femoral refer- ences there is 2° deviation of the sulcus to the anatomical line and 4° deviation of the sulcus to the mechanical line

[10]. In both normal and arthritic Caucasian knees mea- sured radiographically, the sulcus lies 5±1 mm lateral to the midline of the knee [11].In a cadaveric collection from Africa the sulcus was measured by micrometer as 2.4±2.1 mm lateral to the midline [10]. The discrepancy in degree but not direction of displacement between studies is at- tributed to racial variation, an opinion supported by ear- lier work documenting that black femora are longer and narrower than Caucasian femora [10]. This issue of pop- ulation differences in functional anatomy of the knee will be revisited below.

Like the distal femur, the proximal tibia can be char- acterized as an asymmetrical three-dimensional struc- ture. Its medial surface is concave with its periphery, covered by the medial meniscus. The lateral surface is convex with its periphery, covered by the lateral menis- cus. The menisci function in conjunction with the liga- ments in the kinematics of the normal knee by guiding the femoral condyles over the surface of the tibia in flex- ion and extension. They are routinely excised along with the ACL in the process of placing a prosthetic knee, how- ever,playing no role in the functional anatomy of the knee from the perspective of total knee arthroplasty. For this reason, the functional significance of the proximal tibia anatomy lies less in its topological features and soft-tis- sue attachments, and more in its spatial position relative to the femur.

The intuitive notion that the tibia centers below the fe- mur is depicted repeatedly in anatomical illustrations and surgical manuals.This important feature of functional tib- ia anatomy is misrepresented in these illustrations,howev- er.The center of the tibia – defined as the point equidistant from the front to back and side to side – is not centered be- low the center of the femur. Studies of both normal and arthritic knees performed with three-dimensional com- puted tomography demonstrate that the center of the tibia is offset posterior (4±6 mm) and lateral (5±4 mm) to the femur center (

) [12]. The clinical significance of this relationship is that surgeons seeking to align implants congruently are often misled into centering the tibia com- ponent on the tibia and centering the femoral component on the femur with the expectation that the two components will then align or center with each other. However, the anatomical offset of the femur and tibia centers leads to translation between the two prosthetic components. This problem is compounded by the fact that engineers are de- signing implants with increasing conformity to limit wear without the recognition that most implants are translated in application. The combination of conformity and anatomical translation likely leads to increased, not de- creased wear, a topic revisited below.

Most anatomical representations and surgical manu- als also depict the tibia and femur as rotationally aligned.

This depiction of the functional anatomy appears consis- tent with studies of the normal knee but inconsistent with

Mid- plane

Sulcus axis Anatomic axis

Mechanical axis

Sulcus

⊡Fig. 3-4a, b. aThe trochlea is offset to the lateral side of the distal fe- mur and its lowest point, the sulcus, is lateral to the midplane. bThe ori- entation of the sulcus (sulcus axis) lies between the mechanical and anatomical axes of the femur

a

b

3

studies of the pathological or arthritic knee (

).

Knees demonstrating a history of anterior knee pain and early patella-femoral arthritis were found to have an ex- ternal malrotation of the tibia to the femur (7±1°) [13].

Knees undergoing total knee arthroplasty for medial compartment osteoarthritis were also found to have an external malrotaton of the tibia to the femur (5±1°) [14].

Unlike the translation discussed above, which reflects the normal morphology of the knee, this malrotation is not present in the normal knee [13, 14] but reflects a rotation- al contracture of soft tissues (hamstrings, IT band, etc.) associated with the pathological conditions of anterior knee pain and osteoarthritis.

The anatomical significance of this observation from a functional perspective is again related to the placement of components in the process of total knee arthroplasty.

A study of the rotational alignment of components in to- tal knee arthroplasty found that the tibial component was externally malrotated 5° relative to the femoral compo- nent when the component was referenced to the transtib- ial axis, and not to the femoral component [15]. Retrieval studies of failed total knee implants document a consis- tent pattern of external malrotation and translation in the wear of the tibial polyethylene [16, 17]. These studies documenting component malposition and patterns of ab- normal wear reflect differences in kinematics between the normal and the replaced knee using conventional sur- gical techniques and currently available implants.

Another significant difference between the position of a total knee tibial component and functional anatomy occurs as a result of intentionally or unintentionally al- tering the slope of the joint line. When referenced to the mechanical line of the tibia, the articular surface slopes approximately 3° down from lateral to medial and 5°

down from front to back. Historically, methods of total knee arthroplasty recreated this functional anatomy by making an anatomical cut of the proximal tibia to posi- tion the tibial component parallel to the joint line. How- ever, contemporary techniques of total knee arthroplasty often replace this sloped surface with an implant placed perpendicular to the mechanical line, the so-called clas- sical cut of the tibia. This alteration in functional mor- phology necessitates additional compensatory cuts that remove relatively more lateral than medial femur, both distal and posterior, to create rectangular spaces for the implant and to balance the soft tissues. The rationale and methods of these cuts are discussed in later chapters and they are raised here only to illustrate the normal mor- phology and the potential to alter it, intentionally or un- intentionally, in the process of performing a total knee arthroplasty.

The last morphological feature of the knee to address in this review is the patella. As previously stated, it is the largest sesamoid bone in the body, measuring 2.0–2.5 cm ventral to dorsal. When viewed from the ventral surface it is a convex oval bone. Viewed from the dorsal or artic-

⊡Fig. 3-5a–c.Femoral-tibial rotation (b) and offset (c) are illustrated on cross- sections of the femur (a, solid plane) and tibia (a, hatched plane) superim- posed on each other. The tibia is exter- nally rotated to the femur in patholog- ical knees (b), and the center of the tibia is posterior and lateral to the center of the femur in both normal and patho- logical knees (c)

Femur b

Tibia

Tibia Femur

Femur

Tibia

= femur center = tibia center a

c

ular side,there is a cartilage cap covering the surface with a ridge separating a large lateral facet from a smaller me- dial facet. A small cartilage reflection lies along the far medial side and is referred to as the odd facet.

When viewed in relationship to the femur, the patella appears to sit lateral to the midplane (

).This ob- servation is consistent with the documented shape of the trochlea and the location of the sulcus of the femur [10]

(see Fig. 4a). This relationship of the patella to the femur is present in both normal and osteoarthritic knees [11]

and should be taken into account when positioning these components in total knee arthroplasty.

The patella may tilt relative to the femur, reflecting underlying femoral pathology. In the context of the nor- mal knee, i.e., in the absence of pathology, the patella lies parallel to the coronal plane of the femur (Fig. 6a). In the pathological knee, e.g., the osteoarthritic knee and the knee with anterior pain, the patella tilts relative to the fe- mur. Traditional illustration of the tilted patella places the coronal plane of the femur parallel to the horizon and the plane of the patella inclined relative to the femur. An alternative representation is that the patella is tethered by the extensor mechanism in the coronal plane of the body and it is the distal femur that assumes a tilted orientation relative to the patella and the body (

). This rep- resentation reflects an appreciation of the normal hip morphology and the variable degrees of distal femoral anteversion that are associated with the pathological knee [13, 18]. This appreciation of abnormal anteversion leads to the intuitive notion that surgical correction of patellar tilt in total knee arthroplasty is achieved in part by ad- dressing the rotation of the femoral component in total knee arthroplasty. Failure to appreciate the presence of abnormal femoral anteversion leads to malrotation of the femoral component with an adverse effect on patella tracking, an outcome well documented in the arthroplas- ty literature [19]. These issues of surgical correction of femoral rotation and patella tilt will be addressed else- where in this book, but it is important here to appreciate that the functional anatomy of the knee varies with

pathology, shaping the perception of the problem and dictating the surgical approach to correction.

All architectural components of the knee, i.e., femur, tibia,and patella,have now been addressed along with the investing soft-tissue sleeve. However, several caveats are in order before concluding.This review addresses normal functional anatomy, but it does not address in any detail the wide range of normal, both in size and in shape, oc- curring in the human population [20]. There is also sig- nificant morphological variation in the knees of subpop- ulations, reflecting racial differences [20]. Morphological variation also occurs in the context of disease,e.g.,the os- teoarthritic knee is different from the normal knee [18].

Recognition of this anatomical variation is necessary to appreciate the art of total knee arthroplasty and to un- derstand the surgical techniques described in subsequent chapters of this text.

References

1. Kapandji I (1987) The physiology of the joints. Churchill-Livingston, New York 2. Weber W, Weber F (1992) Mechanics of the human walking apparatus.

Sect 4: The knee. Springer-Verlag, Berlin Heidelberg New York 3. Fick R (1911) Mechanik des Kniegelenkes. In: von Bardeleben K (ed) Hand-

buch der Anatomie des Menschen, Band 2, 1, vol 3. Gustav Fischer, Jena 4. Mensch J et al (1975) Knee morphology as a guide to knee replacement.

Clin Orthop Rel Res 112:231–241

5. Yoshioka Y et al (1987) The anatomy and functional axes of the femur.

J Bone Joint Surg 69-A:873–880

6. Pinskerova V et al (2001) Tibial femora movement. 1: The shapes and rel- ative movements of the femur and tibia in the unloaded cadaver knee.

J Bone Joint Surg 82-B:1189–1203

7. Hollister A et al (1993) The axes of rotation of the knee. Clin Orthop Rel Res 290:259–268

8. Eckhoff D et al (2001) Three-dimensional morphology and kinematics of the distal part of the femur viewed in virtual reality, part I. J Bone Joint Surg 83-A [Suppl 2]:43–50

9. Eckhoff D et al (2003) Three-dimensional morphology and kinematics of the distal part of the femur viewed in virtual reality, part II. J Bone Joint Surg 85-A [Suppl 4]:97–104

10. Eckhoff D et al (1996) Sulcus morphology of the distal femur. Clin Orthop Rel Res 331:23–28

11. Eckhoff D et al (1996) Location of the femoral sulcus in the osteoarthrit- ic knee. J Arthroplasty 11:163–165

⊡Fig. 3-6a, b.The patella sits lat- eral on the distal femur, consistent with the location of the sulcus of the trochlea (a). The patella tilts rel- ative to the femur in the face of al- tered femoral anteversion (b) but maintains a constant relationship to the proximal femur and the coronal plane of the body when the foot is in the sagittal plane

a b

3

Clin Orthop Rel Res 339:152–155

14. Eckhoff D et al (1994) Version of the osteoarthritic knee. J Arthroplasty 9:73–79

15. Eckhoff D et al (1995) Malrotation associated with implant alignment technique in total knee arthroplasty. Clin Orthop Rel Res 321:28–31 16. Lewis P et al (1994) Posteromedial tibial polyethylene failure in total knee

replacements. Clin Orthop Rel Res 299:11–17

18. Eckhoff D et al (1994) Femoral anteversion and arthritis of the knee. J Pediatr Orthop 14:608–610

19. Figgie H et al (1989) The effect of alignment of the implant on fractures of the patella after condylar total knee arthroplasty. J Bone Joint Surg 71-A:1031–1039

20. Eckhoff D et al (1994) Variation in femoral anteversion. Clin Anat 7:72–79