Knee arthroplasty was in its infancy in India in the 90s, when I returned from overseas having done hundreds of fairly straightforward cases, to start practice in India. I was accosted with patients with profound knee deformities, and I was desperately in need of guidance! That’s when a colleague suggested I read Ken’s book. However, in the early “90s,” in the sprawling metropolis of Mumbai, there was just one copy of The Technique of Knee Arthroplasty by Ken Krakow. It took a month of searching to track down this elusive copy, in the library of one of the medical colleges, but that book transformed my entire concepts about knee arthroplasty and Ken immediately became a superstar in my eyes. I met my idol Ken at the Current Concepts in Joint Replacement meeting in Orlando in December 1995 and since then, first as a participant and then as a faculty for 15 years, I would quiz him at every opportunity about the problems I faced. That’s when I realized what a perfectionist he was, how meticulous and precise was his thinking. I also saw another side, an ardent photographer, he was the only one permitted to amble around in the aisles and during the faculty dinner to take hundreds of photos (only a few lucky ones got to see them!). He was also the most prepared for any eventuality wherever he went—he always carried spare batteries, a laser pointer, a torch, a whistle, and some tools with him! A model of sartorial elegance, he would inspect carefully my watch and attire and share his views on how to match shoes and accessories. About 5 years ago, I tried convincing him that he needed to publish an updated version of his book which was sold out and nearly impossible to get. He said he had made several notings on pieces of paper so I suggested to him that we could capture these on camera and organize them, and that I would be only too happy to help in this endeavor. Annually, I would beseech him to update his book. Finally, a few years ago, he relented and said that Bill Mihalko would spearhead the project. It is such a delight that I have been able to contribute to this book as my tribute to one of the finest thinking surgeons that it has been my privilege to meet!
Both alignment and balancing are critical for the optimal functioning of a total knee arthroplasty (TKA) and are rendered more challenging in the face of deformity. In the first edition this chapter was prefaced by the statement that deformity correction involved “a combination of philosophy and hypothesis” (p. 249). Since its publication, many of these concepts have been substantiated by cadaveric, computational, finite element analyses, and clinical studies, whereas others have been supplemented and a few have been replaced by newer developments. It is pertinent to allude to these advances at the outset while reaffirming certain established tenets; both are equally and extremely germane to and set the stage for the vital aspect of soft tissue balancing.
Kinematics of the knee refers to the relative motion between the femur and the tibia. The knee can move in different directions, and these are known as degrees of freedom. There are three translations and three rotations that are possible. The three translations are in the anterior-posterior, medial-lateral, and proximal-distal directions. The three rotations are flexion-extension, varus-valgus, and internal rotation-external rotation. Kinematics of the prosthetic knee are governed by the implant factors (principally the design), patient factors (such as knee extensor-flexor muscle strength), and surgical factors (such as soft tissue balancing).
Unlike the intrinsically more stable ball-and-socket configuration of the hip joint, stability is conferred upon the knee by static stabilizers, such as ligaments, menisci, and capsule, and less by shape of the articulating surfaces. In addition, there are dynamic stabilizers that are the muscles and tendons around the knee such as the quadriceps, gastrocnemius, hamstrings, and the iliotibial tract (IT).
Soft tissues have an initial tension and stiffness. Tension refers to two forces pulling in directly opposite directions. Tension in a ligament controls the boundaries of laxity. Laxity refers to translation or rotation occurring in a particular direction when a force or torque is applied. Stiffness is the ability of soft tissues to resist elongation when a load is applied and shows a typical curve with an initial toe-in region, linear region, and the failure region. Ligament stiffness is defined as the force (in newtons, Nm) needed to stretch the ligament by 1 mm and is the slope of the linear region. Strain refers to the resulting deformation that occurs.
The superficial medial collateral ligament (MCL) complex has been reported to have a stiffness value of 80 N/mm, 42 N/mm for the deep MCL complex in cadaveric knees, and 58 N/mm for the lateral collateral ligament (LCL). In osteoarthritic (OA) knees undergoing TKA stiffness of the soft tissue envelope was greater in cruciate-retaining (CR) knees (80.9 and 80.8 N/mm, in flexion and extension respectively) than posterior-stabilized (PS) knees (64.3 and 68.5 N/mm, respectively), with no differences because of age or sex and no differences between extension and flexion within CR and PS groups. Stiffness in full extension was 28 N/mm for the medial compartment (not just the MCL) and 20 N/mm for the lateral compartment.
The stabilizing structures around the knee are not symmetrical in terms of anatomy and function medially and laterally nor anteriorly and posteriorly. The MCL is isometric, whereas the LCL is anisometric. Also, the articulating surfaces of the medial side of the knee differ from the lateral side in terms of shape and hence congruity and stability. The medial side of the knee is more stable, and the lateral side is more mobile; the lateral femoral condyle translates further posteriorly with flexion.
The key restraints in different directions are as follows. Valgus rotation is principally resisted by the superficial medial collateral ligament (sMCL) and varus rotation by the LCL. One-fifth of medial and lateral restraint at 5 degrees of knee flexion is conferred upon by the cruciates. Anterior translation is resisted by the anterior cruciate ligament (ACL) and MCL; posterior translation is resisted by the posterior cruciate ligament (PCL), posterior oblique ligament (POL), and posterior lateral structures (popliteus and popliteofibular ligament [PFL]). External rotation is resisted by the MCL and posterior lateral structures; internal rotation is resisted by the LCL, POL, and posteromedial capsule. Hyperextension is prevented by the oblique popliteal ligament (OPL), which is the largest ligament in the posterior capsule. The IT controls varus and internal rotation of the knee. The anterolateral ligament has been added as an additional and distinct ligament. It originates from near the popliteal tendon attachment on the lateral femoral epicondyle and courses to its tibial attachment midway between Gerdy tubercle and the fibular head. It resists anterior tibial translation but possesses much weaker tensile properties.
Flexion and extension also bring into action different restraints. For example, in flexion the posterior structures are relaxed, and therefore the main stabilizers are the medial and lateral structures. In extension the posterior structures and the medial and lateral structures provide restraint. The anterior fibers of the MCL confer more stability in flexion because of their anatomy.
Normal and osteoarthritic (OA) knee kinematics show continuous and smooth femoral external rotation with knee flexion. Maximum femoral external rotation and condylar posterior translation occur in normal knees, followed by OA knees, and the least in knees with an arthroplasty. Posterior-stabilized TKA (PS-TKA) consistently display an abnormal anterior femoral translation during the first 60 degrees of flexion; screw-home motion has not been observed in OA knees or in knees after TKA during passive motion.
Instability and Balance
Alterations or damage, due to congenital, developmental, traumatic, pathological, or surgical reasons, to the bony articulating surfaces, soft tissues, and muscles can result in changes in stability of the knee. These must be carefully assessed preoperatively as they have a direct bearing on the restoration of stability when a TKA is performed. The vastus muscles and gastrocnemius are the primary contributors to the knee joint forces during gait, followed by the rectus femoris, gluteus maximus, soleus, and hamstrings.
Instability is a symptom experienced by the patient. It can be simply described as a feeling of abnormal knee movement during normal activities. Instability may or may not be accompanied by pain. For example, a patient who has poliomyelitis ( Fig. 9.1 ) may experience instability but no pain; a patient who has OA may have both instability and pain.
Balancing refers to the art and science of restoring stability to and improving kinematics of the patient’s arthritic knee during TKA so that there is ligament stability throughout the functional range of motion (ROM). Balance means that the capsular ligamentous elements are pulled out to length but are not excessively stretched and certainly are not buckling or lax. It is one of the most complex and important issues in the whole operation. Soft tissue symmetry will need to be reestablished while the tibia and femur are aligned normally with one another. Proper balance must be established in every case, whether deformity coexists or not; severe deformity would call for more specialized techniques. Every attempt must be made during this process to not introduce instability, which may be a reason for early failure; instability may also lead to abnormal loading that may result in long-term failure of the implant. Imbalance can influence the ROM; ROM can be reduced, causing stiffness, or increased, leading to instability, subluxation, or dislocation.
Because of the presence of varying degrees of bone, cartilage damage, and ligament deficiency or excision (ACL), the implant will need a certain amount of stability or constraint. The underlying principle of TKA is to use the least constrained device and rely on the patient’s soft tissue envelope so that there are reduced stresses on its fixation in bone; this will reduce the risk of revision. When using an implant with the least constraint necessary, it becomes imperative to rely on a combination of the patient’s soft tissue tension in the capsule and ligaments in conjunction with the articular surface geometry of the implant, matching the implant constraint to the soft tissue envelope.
The most constrained implant would be a hinge; when using this design, soft tissue balancing becomes superfluous as the entire stability is afforded by the constrained design. However, a more constrained implant would lead to more stresses on its fixation and would therefore entail additional fixation in bone with resultant bone destruction, which makes a subsequent revision more challenging. If the surgeon opts for a hinge, everything that follows is irrelevant.
Adding to the complexity of soft tissue balancing are the lack of consensus on the precise definition of and quantification of stability (and therefore what constitutes a “balanced joint”), asymmetry of native soft tissue structures, presence of varying degrees and types of deformity, wide spectrum of pathology encountered, the differential effect and efficacy of various surgical steps and devices, variations in implant positioning and design, and changes in laxity that may occur over time. All these render balancing an unresolved and difficult-to-understand subject.
Furthermore, during surgery, the active muscle envelope is relaxed, and therefore the knee is in a nonloaded paralyzed condition; judgment of laxity therefore is based on passive ligament tension. Even under these conditions, assessment of balance is based on subjective “feel,” and this can be affected by the size and weight of the thigh and surgeon’s ability to stabilize the knee while assessing stability. Various newer technologies to assist in this endeavor are still under evaluation. Surgical technique is further influenced by surgical experience, training, and volume. Objective data of correlation of stability with outcome measures and patient satisfaction is still emerging (and some of it is conflicting); whereas, studies focusing on revision TKA clearly identify instability as one of the main indications for revision.
Thus not balancing the joint is clearly fraught with the real possibility of failure. Yet what constitutes a balanced joint and how to achieve it is akin to John Saxe’s “The Blind Men and the Elephant” poem of six blind men trying to describe an elephant. Everyone thinks they know what it is. I shall nevertheless attempt to examine some of these factors in greater detail, scrutinize some evidence-based conclusions, and present a systematic and rational basis for balancing the prosthetic joint.
Laxity in Normal Knees
Before I begin our quest of balancing an arthritic knee, let us first look at the normal knee. There are various studies in normal subjects and nonarthritic cadavers using radiographs, MR scans, navigation, and robotics with different stresses applied to determine what can be defined as “envelopes of passive knee joint motion” (p. 705). All these studies show that in the normal knee, medial and lateral laxity are not equal, and this difference varies with flexion. In general, the lateral gap is greater than the medial gap, and the difference increases with flexion. Even in cadaveric studies, it has been shown that there is an asymmetry of medial and lateral gaps right through from extension up to flexion; this gap is greater in flexion than in extension. There are no studies that suggest that these gaps are equal medially and laterally and that the flexion and extension gaps are equal. This is true not only of varus-valgus but also of anteroposterior and rotational laxity. Whether gaps in a prosthetic knee should mimic the normal joint or seek to establish equal gaps is a subject of debate.
With stress radiographs using a tensioner with 15 Nm torque, varus laxity of 2.8 degrees, and valgus laxity of 2.3 degrees in full extension in healthy subjects was reported. (Applying elementary trigonometry, 1 degree roughly equals 1 mm over the length of a normal femur or tibia.) When knees were flexed to 90 degrees and imaged in neutral and under a varus-valgus stress in an open MRI system, the lateral joint gap opened by 6.7 mm with a varus stress, whereas the medial joint gap opened by only by 2.1 mm, indicating that the lateral joint gap is significantly lax. One study has shown that the mean varus-valgus limits are smallest at 0 degrees flexion and increase nearly linearly with flexion; the varus limit increases more rapidly than the valgus limit. The ranges of the varus and valgus limits are both largest at 90 degrees flexion (2.6 and 1.1 degrees, respectively). A robotic study in cadavers has shown that the native medial and lateral gaps are tightest in extension at 1.3 and 2.2 mm, respectively. These gaps increase by 3 to 4 mm with 0 to 20 degrees of flexion and then plateau. The lateral native gap is 1.3 mm larger than the medial gap throughout the range. Generally, larger differences in laxity of more than 2 to 3 degrees between medial and lateral gaps over the entire range are not seen.
After ACL Release
After the ACL is resected, the most cadaveric studies using tensioners show an increase in extension gaps, with the medial side increasing by 2.1 mm and the lateral gap increasing by 2.8 mm with 100 N force and increasing 2.5 and 3.1 mm, respectively, with 200 N force. Flexion gaps increase minimally after ACL resection, medially by 0.4 mm and laterally by 0.9 mm with 100 N force and increasing 0.4 mm and 0.9 mm, respectively, with 200 N force.
After PCL Release
In contrast, PCL release leads to larger increases in the flexion than the extension gap. Older cadaveric studies have shown that the tibia can be distracted from the femur by 5.3 mm at rest and 6.4 mm under tension. In varus knees, after PCL resection and using a tensioning device, the flexion gap increased medially by 4.8 mm and laterally by 4.5 mm. The extension gap increased by <1 mm. The lateral joint gap has been shown to be 5 to 6 mm more than medial from 60 to 120 degrees of flexion. Flexion gaps increase in cadavers from 30 to 120 degrees flexion by 1 to 3 mm, but increasing the force from 50 to 100 N increases the mean gap by only 0.5 mm. Another study using a tensioning device with 10, 20, and 30 inch-pounds (in-lb) of distraction force found no change in extension gap after PCL resection; flexion gap distance enlarged by <2 mm. PCL resection increased the mean flexion gap more than the extension gap: The gaps were 2.4 mm medially and 3.3 mm laterally in flexion versus 1.3 mm medially and 1.2 mm laterally in extension. What is interesting is that the differences in joint center gap for 20, 40, and 60 lb and varus laxity at extension and flexion were not significantly different among different joint distraction forces. With the 100 N force, the medial and lateral extension gaps increased by 0.2 mm after PCL resection and by 0.2 and 0.3 mm, respectively, with 200 N; flexion gaps increased by 2.4 mm medially, 3.4 mm laterally with 100 N, and 3.4, 2.6, and 3.7 mm, respectively, with 200 N. Thus similar proportions in gap enlargement were seen with the 200 N force.24 In general, the extension gap increases by 1 mm and the flexion gap by 3 to 4 mm after PCL release.
It may be more technically challenging to achieve perfect balance when retaining the PCL. PCL should be fully released before ligament tensioning for femoral component rotation as the medial and lateral gaps increased by 0.5 to 1 mm if complete release was performed subsequent to the gaps having been measured with only partial PCL release in a cadaveric study. Similarly, in a navigational study using a 150 N tensioner force, PCL release showed no effect on the extension gap but increased the flexion gap by >3 mm in 36% of patients and >5 mm in 12% of patients, with <2 mm change in 44% of patients. Thus it is recommended to release the PCL before the femoral resections are performed, as this step determines the ratio between extension and flexion gaps.
Differential Effect of Deformity on Gaps
In varus knees the lateral extension gap increases with severity of deformity; it is larger with severe varus (>20 degrees) than in mild (<10 degrees) and moderate (10 to 20 degrees) varus, whereas there are no differences in the medial joint gaps among the groups. A 2020 navigational study found a much greater initial difference (4.7 mm) between lateral and medial gaps in extension in more deformed knees requiring releases than in those with no release required (1 mm).
On the other hand, at 90 degrees flexion, the medial-lateral laxity difference of ≤2.5 degrees was present in 91.6% of 72 knees, implying that there is no evidence of contracture in the coronal plane tissue in end-stage arthritic knees at the time of TKA. This was also seen in another study where the difference between lateral and medial gaps in flexion in more deformed knees requiring releases was 1.8 and was 0.5 mm in those with no release required.
Long-term Changes in Laxity
Under spinal anesthesia, medial laxity in extension increases by 0.6 degrees and lateral laxity by 1 degree when stress is applied.
Mediolateral joint laxity analyzed immediately intraoperatively and 30 minutes later showed that stress relaxation occurred in all cases; mediolateral laxity increased by an average of 1 mm on the medial and lateral sides. Arthrometer stress tests with 150 N force applied with the knee in 0 to 20 degrees flexion after surgery, with the patient still under anesthesia, showed no differences in laxity measurements made under anesthesia and 6 months postoperatively with either the PS or CR design.
Some studies suggest that no changes occur postoperatively in the coronal laxity that is achieved intraoperatively even up to a mean of 77 months. Both extension and flexion laxity remain unchanged at 5 years in CR knees and extension gaps in PS knees as assessed by stress radiographs. Another study showed that medial laxity remains constant postoperatively from immediately after surgery to 12 months. However, the mean lateral laxity that was 8.6 degrees immediately after surgery decreased to 5.1 degrees at 3 months. Residual medial tightness of 1 to 2 mm improved spontaneously when laxity was measured under anesthesia at the time of the staged surgery of the contralateral knee joint. There was no change in lateral laxity.
These studies suggest that about 1 mm of relaxation of gaps occurs soon after surgery and then remains unchanged. Medial tightness of 1 to 2 mm may improve, and lateral laxity may reduce by up to 3 mm over time. However, this may be influenced by the overall alignment as shown in a computer simulation study where condylar liftoff occurred in neutral coronal alignment regardless of excessive LCL laxity. Condylar liftoff occurred easily in >3 degrees varus alignment even with slight laxity in the LCL. The study also noted higher peak contact forces in the medial compartment on heel strike and lateral condylar liftoff during the single-leg stance that appeared to be influenced more by the degree of varus alignment than by the amount of LCL laxity.
Effect of Stability on Patient-Reported Outcome Measures
Residual imbalance of 2.8 and 1.3 degrees varus in extension and flexion, respectively, was not associated with significant differences in early clinical results of postoperative ROM and the subscales of (symptoms, patient satisfaction, patient expectation, functional activities) of Knee Society Score (KSS) among the groups categorized according to the varus-valgus gap angle and the laxity. However, excessive intraoperative medial joint laxity of ≥4 mm at 90 degrees flexion decreased patient satisfaction at 1 year.
Western Ontario and McMaster Universities Arthritis Index (WOMAC) scores of patients with <3 mm gaps showed worse scores for two functional items demanding knee flexion (bending to floor and getting on/off toilet), and 3 to 4 mm laxity at 90 degrees might be necessary to carry out daily life activities. Anteroposterior (AP) laxity measured with an arthrometer at 60 degrees knee flexion significantly correlated with Knee Injury and Osteoarthritis Outcome Score (KOOS) pain score. AP laxity of ≥7 was significantly associated with a subjective feeling of instability in TKA patients.
Similarly, it has been suggested that a controlled flexion gap increase of 2.5 mm may have a positive effect on postoperative flexion and patient satisfaction after TKA. In contrast, WOMAC scores were better in TKAs with a medial-lateral balanced (<2 mm) gap as were outcomes in terms of physical functioning, bodily pain, social functioning, Oxford and Knee Society scores at 6 months, and improved social functioning scores at 2 years with flexion-extension gap differences of ≤2 mm. Knees in which the difference between varus and valgus laxity was <2 degrees had greater ROM by 10 degrees whereas those with >2 mm difference had lesser ROM by 9 degrees. Smaller medial gaps at 60 and 90 degrees of flexion have been shown to play an important role in achieving medial pivot motion with tibial internal rotation; moreover, tibial internal rotation provides a better flexion angle after PS-TKA. Patients with medial-lateral instability <5 at 30 degrees at 2 years postoperatively had superior KSS functional knee scores and 36-item Short Form Health Survey (SF-36) scores than those with greater instability.
Thus there appears to be some degree of variation in the reported optimum laxity required for superior outcomes. It would appear that the medial side is key in varus knees, especially if activities involve deep flexion. It needs to be balanced to approximately within 2 mm from 0 to 90 degrees; the lateral gap especially in flexion can be more lax but not >4 mm.
This is borne out by a cadaveric study, using calibrated extensometers sutured to the LCL and sMCL, that found that the strains in both ligaments in the replaced knee are different from those in the native knee. The MCL becomes tighter in the native knee and the LCL relaxes with flexion. Both ligaments were found to be stretched in extension; in flexion the MCL was found to be relaxed, but the LCL was tight, suggesting that measured resection techniques may overstuff the joint. In another study in which transducers were attached to collateral ligaments in the native knee the MCL slackened 2 mm, whereas the LCL slackened 7 mm with flexion. Post-TKA the MCL slackened a further 3 mm and the LCL a further 4 mm during flexion. A 5 degree external rotation of the femoral component slackened the MCL by a further 2 mm and tightened the LCL by 2 mm; the opposite effect was found during 5 degrees internal rotation. Others have reported that a wide range of coronal plane laxity values are associated with highly satisfied TKA subjects and that a controlled flexion gap increase of 2.5 mm may have a positive effect on postoperative flexion and patient satisfaction after TKA. Another study determining the tibiofemoral forces and collateral ligament strain for variations in flexion and extension gaps noted that small variations in gaps had minimal effects on the soft tissue tension between 15 and 100 degrees of flexion. However, increasing the flexion gap by as little as 2 mm may reduce tibiofemoral forces beyond 90 degrees of flexion and a looser flexion gap decreases soft tissue tension beyond 120 degrees of flexion.
It is not inconceivable that it may be the postoperative kinematics of the prosthetic knee that may be responsible for patient satisfaction on patient-reported outcome measures (PROMS). In fact, a 2020 study reported that during closed kinetic chain movements, patients with poor PROM scores after TKA experience more anterior translation on the medial side followed by a medial midflexion instability and less posterior translation on the lateral side in deep flexion than patients with good PROM scores.
Given this background information regarding gaps, some of it conflicting and contradictory, the next sections of this chapter will consider relevant aspects of the patient’s history, clinical examination, and imaging that may have a bearing on balancing gaps during TKA. The surgeon often has a proclivity to bypass these steps and head directly to surgery. The surgeon is in danger of remaining oblivious about key issues that may call for a change in the surgeon’s technique to address special nuances that exist between one knee and the other, often even in the same patient ( Fig. 9.2 ).
Steps to Achieving Correct Alignment and Balance
Although some of these vital steps have been chronicled in earlier chapters, the ones described here are particularly relevant to balancing the knee and are perhaps worthy of repetition.
Prior soft tissue trauma, which may have led to a PCL or MCL injury that may have a bearing on implant type
Prior fracture with intra- or extraarticular malunion
Sudden worsening of pain above or below the knee signifying a possible stress fracture (wherein, for example, a long stem tibial component may be indicated; cutting the tibia orthogonal to its long axis would be mandated regardless of one’s views on kinematic or anatomical cuts. This would therefore impact flexion and extension gaps.) (See Fig. 9.3 .)
Prior surgery (such as high tibial ostomy [HTO], ligament reconstruction) that may be responsible for altered anatomy, kinematics, adhesions, and changes in patellar height (e.g., patella baja)
Preoperative clinical assessment
Flexion deformity (FD) is often masked by the leg lying in external rotation. The knee must be rotated internally and examined with the patella pointing upwards so that the FD manifests ( Fig. 9.4 ).
Hyperextension may not be evident as the leg is lying flat on the examining couch, and it will be missed unless the leg is passively lifted with a hand supporting the ankle ( Fig. 9.5 ). It will need a different technique of balancing than a knee with an FD.
Varus-valgus stress examination with the knee flexed 10 to 20 degrees gives a very good indication if the deformity is correctable ( Fig. 9.6 ). In full extension the posterior capsule is taut, making assessment of stability and correctability difficult; it is relaxed in slight flexion so mediolateral laxity can be better assessed.
Patellar tracking especially in valgus knees should be assessed; maltracking will need to be addressed with appropriate releases.
Tibial torsion is also masked in patients with a flexion and varus deformity. To detect intorsion, one needs to rotate the leg such that the patella points upward; if the foot and malleoli are internally rotated, it suggests tibial intorsion ( Fig. 9.7 ). Incorrect tibial tray rotational placement may affect knee kinematics.
Hip ROM has a bearing on ligament balancing; an arthritic or stiff hip or one with altered anteversion may affect orientation of the distal femur and influence femoral component orientation.
Neurological examination: weakness of quadriceps or gastrocnemius muscle (as after a stroke) may result in hyperextension, requiring a more constrained implant. Patients with severe combined valgus and flexion deformity are at risk of developing a common peroneal nerve palsy postoperatively, and the neurological status of the lower limb needs to be documented initially.
Flatfeet cause lateral deviation of the weight-bearing axis of the limb. It is more common in patients with valgus deformity ( Fig. 9.8 ), though it is frequently seen in patients with varus deformity. In the latter the surgeon may use this knowledge to slightly undercorrect the deformity, which would thus require a lesser release. Also the use of the second metatarsal for judging tibial jig placement may induce an error in rotational placement of the tibia tray because forefoot abduction is present in a severe flatfoot deformity.
Gait: In addition to observing for flatfeet, it is important for the surgeon to determine whether there is a major adductor thrust ( Fig. 9.9 ) during the stance phase in varus knees. This would signify increased laxity of the lateral structures often with severe medial tibial bone loss and will need to be factored into the balancing technique.
Weight-bearing long hip-to-ankle radiographs along with routine weight-bearing AP, lateral, and skyline views are valuable. A systematic assessment of radiographs should be undertaken. These need to be viewed carefully for the following elements:
Valgus correction angle ( Fig. 9.10 ) should be noted if conventional surgery is being performed. This can often vary from 2 to 11 degrees.
Osteophyte presence, location, and size should be noted. The larger the osteophyte, the more likely that a lesser release may be required ( Fig. 9.11 ).
Reactive bone remodeling should be sought on the radiograph. Its presence denotes significant deformity and the possibility that reduction osteotomy may be deployed to balance gaps ( Fig. 9.12 ).
Bone loss or a stress fracture that may need special attention (i.e., augment, graft, sleeve, cone, stem). If a longer stem is required, this may dictate whether a classical 90 degree tibial cut should be performed over a 3 degree or more varus tibial cut to prevent stem tip abutment against the lateral tibial cortex.
Extraarticular deformity in the femur or tibia should be identified as these knees may have to be addressed differently to align and balance them; this is discussed later. Lateral views of the length of the femur and tibia are especially mandated to determine presence and severity of extraarticular sagittal plane deformity.
Intracapsular deformity is discussed later.
Amount of tibial subluxation and extent of lateral opening often imply severe stretching of the lateral structures ( Fig. 9.13 ). The balancing technique will need to take this into account. By the same token, one should look for medial stretching in a valgus knee.
Patella tracking has been alluded to above.
Prior implant such as a staple, anchor, interference screw suggests a prior ligament reconstructive surgery and must alert the surgeon to possible ligament damage.
May be ordered if
there are bowing, torsional, and intracapsular deformities in the femur or tibia.
there is posttraumatic OA with a concern about ligament avulsion from bone, bone loss, and fracture union.
prior HTO was performed with change in tibial slope and rotational malunion.
MR scan is useful in situations where the integrity of the MCL/PCL is in doubt.
Coronal and sagittal limb alignment: the surgeon has to select and aim for classic, anatomical, kinematic, or constitutional alignment or a variant of these.
Components: the surgeon should define the acceptable limits of component placement in all three planes and resection thicknesses; they should decide on the type of implant (i.e., CR, PS, constrained, hinge) that is likely to be required and the type of instruments (i.e., conventional, patient-specific instrumentation [PSI], navigation, robot) to be used for this purpose.
Gaps: The surgeon should also be clear regarding (1) technique of gap balancing (measured resection, gap balancing, or a variation); (2) aims in terms of acceptable flexion and extension gap differences—equal and symmetrical medially and laterally or unequal and asymmetrical and, if so, by how many mm or degrees; and (3) method of assessment of gaps and their symmetry.
Interrelationship Between Bone Cuts, Joint Line, and Stability
Each of the surgeon’s decisions will have ramifications in the actual process of achieving balance. Aiming for hip-knee-ankle (HKA) axis of 177 to 180 degrees in a varus knee tight medially in flexion and extension by resecting the proximal tibia in 2 to 3 degrees varus while orienting the distal femoral cut orthogonal to its mechanical axis may result in lesser or perhaps even no release being required to achieve balanced extension and flexion gaps ( Fig. 9.14A ). A rectangular extension gap can also be achieved by cutting the distal femur in varus and externally rotating the femoral component further ( Fig. 9.14B ) while keeping the tibial cut orthogonal. After a varus distal femoral cut and correctly rotating the femoral component, the extension gap would be balanced but the lax flexion gap would be lax laterally ( Fig. 9.14C ). The dashed line represents an orthogonal bone cut. By measured resection of the posterior femoral condyles ( Fig. 9.14D , dashed lines ), there will be relative internal rotation of the femoral component and a 3- to 4-mm lateral laxity at 90 degrees if no medial release is performed. To obtain a rectangular and balanced flexion gap with the component oriented parallel to the transepicondylar axis (TEA), medial release may be required to establish equivalence of medial and lateral gaps or the femoral component must be placed in excessive external rotation ( Fig. 9.14E ). It becomes immediately apparent that it would be expedient for the surgeon to enunciate their goals at the outset, rather than struggle with the various options intraoperatively. Use of PSI can help the surgeon, as will the use of navigational and robotic software.
These goals are intricately connected with the establishment of the obliquity of the articular plane of the prosthetic joint commonly referred to as joint line . Any cut that removes relatively more from the medial femur throws the knee into varus; a deeper medial cut on the tibia would do the same (see Fig. 9.14A ). An excessively deep cut in the medial tibia might be balanced with a deep lateral cut on the femur to keep the overall tibiofemoral alignment correct. However, this combination would lead to undesirable obliquity of the joint line. In the coronal plane in normal controls the joint line is typically in 3 degrees varus, but it can vary widely, with almost half being >3 degrees and even up to 10 degrees. Likewise, in flexion if the axis of the posterior femoral condyles does not have a proper “neutral” rotational orientation, then as the condyles articulate with the tibial plateau in flexion, the femur goes into abnormal rotation or the tibia is positioned in abnormal varus or valgus attitude. Although properly oriented cuts from the perspective of joint line may imply normal alignment, they do not automatically recreate or maintain proper ligament balance.
Excess distal femoral resection ( Fig. 9.15 ) draws the MCL, LCL, and PCL closer to the distal joint line. As a result, it leads to relative laxity in extension, with the same or tighter stability in flexion. Distal migration of the femoral joint line is shown in Fig. 9.16 . It leads specifically to relative tightness in extension and possibly failure to achieve full extension. Anterior displacement leads to laxity in flexion ( Fig. 9.17 ); posterior displacement leads to relatively excess tightness in flexion ( Fig. 9.18 ).
Effect of Altering the Joint Line on the PCL
A femoral joint line that is more proximal places the PCL so that it is either lax in full extension or tightens abnormally as the knee moves into flexion ( Fig. 9.19 ). With the anterior-to-posterior vector of the PCL, the effect of its tightening should be understood. The posterior cruciate protects against abnormal posterior translation of the tibia. Its tightening pulls the tibia forward. Excessive tightening of this ligament can, in the extreme, cause the tibia to be pulled forward so that it dislocates.
Because of its somewhat anterior origin in the intercondylar notch, pathologic tightening of the PCL in extension is rarely a big problem. This is because ligaments originating more anteriorly from the femur will tend to tighten with flexion, whereas those originating more posteriorly tend to relax with flexion and tighten more with extension.
A femoral component that is too anterior would lead to relative laxity of the PCL in flexion, whereas a component that is too posterior would lead to excessive tightness in flexion.
On the tibial side, things are simpler. Displacement of the prosthesis joint line deeper into the tibia leads to increased laxity in all aspects of motion ( Fig. 9.20 ). Similarly, recession of ligament attachment from the tibia leads to laxity in all aspects of the motion cycle, whereas distal advancement leads to tightening throughout the ROM.
Understanding the Pathoanatomy of Deformity and Imbalance: a Systematic Approach to Correction
It would be instructive to look separately at the two principal ingredients of deformity, the bony elements and the soft tissues, and their interaction as this will spell out the recipe to deal with deformity with the dual objective of restoring balance and reestablishing alignment. As the deformity is often in three planes, it would be convenient to address it sequentially in each plane.
The entire task can be systematized by adopting a universal classification (A. Mullaji, submitted for publication). The pathoanatomy of any knee with its specific components of deformity and instability can be precisely summarized as follows using the ELS classification system. This takes into account:
the location of coronal deformity represented by the letter E followed by 0 for articular deformity, 1 for intracapsular deformity, and 2 for extraarticular deformity; thus E 0, 1, 2;
soft tissue balance in the coronal plane represented by the letter L followed by 0 when flexion and extension gaps are nearly equal medially and laterally, 1 when the extension gap is asymmetrical but the flexion gap is balanced, and 2 when both extension and flexion gaps are asymmetrical and unbalanced; thus L 0, 1, 2; and
soft tissue balance in the sagittal plane represented by the letter S followed by 0 when flexion and extension gaps are equal, 1 when the extension gap is less than the flexion gap, and 2 when the extension gap is greater than the flexion gap, thus S 0, 1, 2.
The Coronal Plane
The two variations in limb alignment in the coronal plane, varus and valgus deformities, can be considered as mirror images of one another. The words concave and convex are used to refer to the medial or lateral aspect of the deformed joint with varus or valgus changes. In the case of a varus deformity concave refers to the medial aspect and convex refers to the lateral aspect. The following discussion will examine more closely the overall bony contribution to deformity.
It is imperative to locate where the bony deformity resides, as this will have a bearing on the soft tissue sleeve and therefore on the principles of restoring both alignment and balance. Full-length radiographs are invaluable in determining the location of deformity. More recently, phenotypes of varus and valgus arthritic knees have been described; these may be valuable in planning and performing surgery.
Type E 0: Articular deformity
The entire arthritic process, bone loss and deformity, is confined to the articular surfaces of femur and tibia with no extraarticular deformity beyond ( Fig. 9.21 ). The TEA (i.e., the approximate attachment of the ligament origins) is normal as is its orientation with the femoral shaft and mechanical axes.
Type E 1: Intracapsular deformity.
This is perhaps the most challenging and least readily discernible deformity to the casual observer. The presence of such a deformity has vastly different implications based on whether it is located in the femur or in the tibia.
On the femoral side, the deformity ( Fig. 9.22 ) is likely to be due to congenital, developmental, or metabolic causes but rarely to traumatic causes ( Fig. 9.23 ). In addition to the intraarticular bone loss and deformity, there is distortion of the femur between the articular joint line and the collateral-capsular attachment (as denoted by the TEA) with a sinister and subtle alteration of the angle between the distal joint line and the TEA and the mechanical and anatomical axes of the femur. Generalized bowing of the femur may also be present (see E 2 in the next section) and should be an indicator of this variation. It should be understood that a prosthetic component that comes to have proper alignment with respect to the femoral axes will not simultaneously have a proper alignment with respect to the epicondylar axis. In a typical varus case ( Fig. 9.24 ) the medial epicondyle would be further away from the prosthetic surface and the lateral one would be much closer to the prosthetic surface. The lateral gap would be marked greater, and the gap would be substantially trapezoidal.
In a valgus knee hypoplasia of the distal femoral condyle is a common observation ( Fig. 9.25 ). It will contribute to exacerbation of the deformity, which may be further accentuated by lateral femoral bowing. Consequently, the lateral epicondyle may be more proximal (with a lesser lateral angle between the TEA and the femoral axes) and hence further away from the prosthetic surface; the medial epicondyle would be much closer to the prosthetic surface. The resultant medial gap would be correspondingly far greater than the lateral one.
On the tibial side, the corresponding region is between the articular joint line and the distal attachment of the sMCL. A classic example would be an overcorrected proximal tibial osteotomy ( Fig. 9.26 ). Malunited intraarticular fractures are another cause of intracapsular deformity ( Fig. 9.27 ). Here, the obliquity of the joint line will be markedly altered, and the soft tissue envelop would be altered either surgically or reactively over time to the osteotomized position. An orthogonal cut across the tibia will remove a great excess of medial bone; thus a disproportionately larger medial gap would ensue ( Fig. 9.28 ). Any mild degree of varus bowing of the femur would accentuate this disparity.
Type E 2: Extraarticular deformity.
Besides the intraarticular angulation, there is an extraarticular deformity away from the knee ( Fig. 9.29A–E ). Its effect on the putative resection levels depends on the severity of angulation (such as a malunion or prior osteotomy) and its distance from the knee joint. If the angular deformity were immediately adjacent to the hip or the ankle joint, its effect on the overall knee alignment would be minimal.
Krackow has suggested, as a rule of thumb, that such shaft deformities would contribute to knee joint deformity in nearly direct proportion to the distance from the relevant hip or ankle joint in the respective case of the femur or tibia. A 10-degree shaft angulation at the midpoint of the femur, 50% along its course, would lead to approximately a 5-degree deformity at the knee, and the deformity would be with respect to the femoral joint line. A 10-degree deformity 80% of the way along the course of the femur, which is closer to the knee, would lead to approximately an 8-degree deformity at the knee; a deformity that is at the knee joint would present 100% of the 10 degrees, or a 10-degree deformity, at the knee joint and would represent type E1 (above).
More commonly, especially in Asians, there is a lateral or medial bowing of the entire femur, the effect of which is computed by drawing the distal femoral resection line perpendicular to the femoral mechanical axis. When the tibial resection line is marked on the radiograph, the angle between the two resection lines (lines A and B in Fig. 9.29A ) would be indicative of the magnitude of the challenge in trying to achieve a balanced gap. Likewise, metaphyseal varus bowing is often seen in Asians ( Fig. 9.30 ). The likely effect of this can be appreciated by drawing the proximal tibial resection line perpendicular to the distal tibial axis. In similar manner, when the femoral resection line is marked on the radiograph, the angle between the two resection lines would be indicative of the likely difficulty in trying to achieve a balanced gap.
As the disease progresses in each of the previously discussed types, osteophytes develop on the concave side of the deformity on the femur and tibia (also in the intercondylar notch and posteriorly, which will be alluded to later). Over time, they enlarge and do not contribute to increasing the coronal deformity, but they have important implications on the soft tissues. Their excision alone can result in deformity correction and gap balancing in many cases.
Soft tissue status
After the evaluation of the bony aspect of the lower extremity from hip to ankle, this section hones in on the soft tissues within the knee. Typically, arthritis begins in one compartment, more commonly the medial one. In the initial stages with cartilage and bone loss in one compartment, articular deformity arises because one bony surface approximates the opposite surface in that compartment, which leads to an angular deformity with concavity on side of the affected compartment. If bone loss occurred only on the tibia, the deformity would be evident in both extension and flexion. Loss distally on the femoral side would appear principally as deformity in extension; if present on the posterior condyle, it would manifest in flexion.
In the initial stage the soft issues are normal; the deformity would manifest on weight-bearing. It would appear to show collateral instability, but it would be passively correctable to neutral alignment with a stress in the opposite direction and would correct until the soft tissue structures on the concave side would be out to length. (This is best done in 10 to 20 degrees flexion to relax the posterior joint capsule.)
As the unicompartmental bone loss advances, the deformity progresses and the convex structures stretch out, further increasing the angular deformity. As this knee is examined, a sense of obvious ligament instability is present on varus-valgus testing. Assuming there is no contracture of ligaments at the concave side, the knee should passively correct to neutral alignment and not overcorrect. It will not be immediately apparent that the convex side structures are lax unless a stress is applied in the same direction as the deformity and a lack of resistance is felt until the slack is taken up.
As the wear increases, the deformity progresses on weight-bearing. The convex structures may be normal or elongated. However, the concave side structures are no longer normal, but they start to undergo contracture such that attempts to passively correct the deformity with stress in the opposite direction will have partial or no success. If the contracture of the concave structures were total, there would be no sense of correction. If it were not, then there would be a partial correction.
Thus there are four possible combinations of soft tissue status:
Concave and convex structures are normal ( Fig. 9.31 ).
Concave structures are normal, and convex structures are lax ( Fig. 9.32 ).
Concave structures are contracted, and convex structures are normal ( Fig. 9.33 ).
Concave structures are contracted, and convex structures are lax ( Fig. 9.34 ).
In practical terms three possibilities have a direct bearing on the structures that need to be released to balance the gaps.
Type L 0 is when flexion and extension gaps are nearly equal medially and laterally ( Fig. 9.35A ).
Type L 1 is when the extension gap is asymmetrical (reduced on the concave side) but the flexion gap is balanced ( Fig. 9.35B ).
Type L 2 is when both extension and flexion gaps are asymmetrical (reduced on the concave side) and unbalanced ( Fig. 9.35C ).
The surgeon can specify the outer limits of the range of stability that they find acceptable. For example, they may consider a difference in extension gaps of more than 2 mm as asymmetrical, and that in flexion gap exceeding 4 mm.
Which Soft Tissue Structures Undergo Contracture?
The answer to this pivotal question is key to the surgical exercise of balancing the prosthetic joint. Intuitively, one may conclude, as has been done in the past, that the MCL undergoes contracture in a varus deformity as it lies on the concave side, and the LCL does so in case of a valgus deformity. A variety of release sequences have been described as a fallout of this premise, often with deleterious effects on stability and ligament strength, culminating in the need for more constrained implants. It has become increasingly evident that the MCL and LCL do not undergo contracture. The MR scan ( Fig. 9.36 ) depicts in no uncertain terms that the MCL, in this example of a 30-degree varus deformity, is tented substantially over the medial osteophyte rather than being contracted. I have examined over 1000 MR scans in patients undergoing arthroplasty and have yet to see a single case of contracture of the MCL. If the surgeon releases the MCL, they will compromise stability by enlarging the medial gap, thereby possibly generating an unnecessary need for a constrained implant to balance the knee. A similar fate will befall the surgeon attempting to release the LCL in a valgus knee; this will have deleterious effects particularly on the flexion gap.
The structures that undergo contracture in a varus knee are the POL, posteromedial capsule, and possibly the posterior-most fibers of the sMCL that are in close proximity to the former. In a valgus knee the analogous counterparts are the posterolateral capsule and PFL. Release of the semimembranosus and popliteus tendons have been described by some authors, but it is more likely that it is the effect of release of the adjacent structures mentioned previously rather than these structures.
Osteophytes cause tenting of the soft tissues on the concave side and diminish or eliminate passive correctability. They would also diminish the pseudoligamentous laxity, such that stresses in the opposite direction would demonstrate only limited or no correctability. However, their excision plays a pivotal role in restoration of both alignment and balance. Closely related to osteophytes is the reactive bone formation on the posteromedial tibia often indistinguishable from the osteophytes that produces a posteromedial flare of bone that furthers tents the soft tissues. Reduction osteotomy and removal of this bony protrusion after the osteophytes have been excised can provide additional assistance in equalizing gaps.
This analysis of bony deformity has been in the coronal plane with consideration of the medial and lateral soft tissues. Next, the deformity in the sagittal plane will be discussed, with a focus on the posterior capsule.
The Sagittal Plane
Flexion and Hyperextension
The two deformities possible in the sagittal plane are FD and hyperextension. Analogous to the concave and convex sides in the coronal plane, the posterior aspect is considered as being the concave side and the anterior aspect as the convex side. Likewise, bony and soft tissue components must be assessed.
In similar fashion to coronal plane deformity the locus of deformity in the sagittal plane should be sought, articular, extraarticular, or intracapsular, though these are much rarer. Articular causes of FD would be posterior tibial bone loss exceeding anterior bone deficiency; in hyperextension, the reverse is discernible more markedly on the distal femoral condyles. Extraarticular elements contributing to these deformities (hyperextension is rare, but FD is more common in the femur; in the tibia, FD is more likely) would be excessive anterior bowing of the femur or a prior fracture or osteotomy. More subtle changes in the sagittal joint line, such as from excessive or reduced posterior tibial slope ( Fig. 9.37 ) and “molding” or hypoplasia of the femoral condyles, would exemplify intracapsular causes. Trauma and proximal tibial osteotomy are the common causes for alterations in the proximal tibia. The surgeon would be well advised to look carefully at lateral and full-length lateral radiographs ( Fig. 9.38 ) so as not to miss these findings.
Soft Tissue Status
Flexion contracture is not an alignment aberration, but it is a ROM abnormality. Although the exact sequence of development of FD from soft tissue causes remains unclear, congenital causes are most uncommon; hence FD is obviously a consequence of arthritis. Pain relief is often experienced by patients when the knee is flexed over a pillow; this may be an etiological agent (cases of functional FD get completely corrected under anesthesia). FD usually begins to manifest after the ACL is disrupted and as wear progresses further posteriorly on the tibia. Ensuing contracture of the posterior capsule, in conjunction with spasm or contracture of hamstrings and gastrocnemius muscle, aggravates the situation further. Release and management of FD must not only straighten but must also lead to increased motion toward the extension side of the ROM arc. Long-standing FD can engender weakness in the extensor mechanism. Despite full passive correction of FD, there may be relapse unless the quadriceps power also recovers.
No correlation has been seen between soft tissue tension in 90-degree flexion and postoperative flexion angle, however soft tissue tension in extension affects postoperative extension angle and stability in extension and 30 degrees flexion. In fact, high tension of the soft tissue in extension intraoperatively may result in FD.
The role of the PCL as a major causative factor in a flexion contracture situation can be questioned. A lax PCL, by itself, would have no way of causing a flexion contracture, and a tight PCL pulls the tibia into a relatively anterior position. Whether this would occur with the knee in flexion, mild flexion, or extension would not appear to contribute to a flexion contracture.
Hyperextension is much less commonly seen, but it also can be more easily missed. Often it manifests only under anesthesia and if the surgeon lifts the leg up while it is unsupported except at the heel. It is imperative to have excluded a possible neurological etiology. Typically, this deformity is caused by excessive stretching of the posterior capsule and oblique popliteal ligament and possibly by attenuation of the ACL and PCL. The posterior elongated structures are relaxed in flexion, so flexion remains unaffected. The collateral ligaments are also neither elongated nor contracted; hence the flexion gap would be balanced unless other contracted structures are affected by concomitant varus or valgus deformity. As the knee extends, tension in the normal collaterals and the camming effect of the femoral component prevents hyperextension.
Based on the relative extension and flexion gaps, the combinations of soft tissue balance can be reduced to three variants:
Type S 0 when flexion and extension gaps are equal ( Fig. 9.39A )