Evaluating the Evidence for Gap Balancing, Measured Resection, and Kinematic Alignment

    There is a renewed focus on understanding optimal soft tissue balancing techniques that may help improve patient satisfaction with total knee arthroplasty. In this article, the authors describe these techniques and review the literature on their use in knee replacement procedures.


    John J. Mercuri, MD, MA; Andrew M. Pepper, MD; Jordan A. Werner, MD; and Jonathan M. Vigdorchik, MD


    The authors have no disclosures relevant to this article.


    Despite the success of total knee arthroplasty (TKA), a significant proportion of patients are dissatisfied after the procedure, with surgeons unable to replicate the degree of satisfaction demonstrated with total hip arthroplasty. [1,2,3]

    For this reason, there is renewed interest in evaluating the theory and surgical techniques utilized during TKA to identify areas for potential improvement in patient outcomes. Of those, there is increasing focus on soft tissue balancing techniques and overall limb alignment, especially kinematic alignment.

    This article describes these techniques and reviews the literature on their use in TKA.

    Mechanical Limb Alignment

    Today’s surgeons are becoming increasingly focused on their patients’ native, preoperative limb alignment, joint kinematics, and ligamentous soft tissue contribution. [2] The traditional mechanical alignment technique arose from John Insall’s desire to equally distribute stresses across the articulating surfaces of his condylar prosthesis to avoid “over-loading” any one aspect. The goal was to decrease asymmetric loading and prolong the function of the implant. [2]

    A study by Hsu et al [5] determined that even load distribution occurred in a total condylar implant when using mechanical alignment techniques. In addition, early research determined that varus proximal tibial resection resulted in uneven load distribution and early clinical failure of TKA. [2]

    Mechanical alignment of the limb does not restore “normal” knee anatomy or kinematics; however, that does not appear to be necessary to obtain good results with long-term implant survival. [4,5]

    Kinematic Limb Alignment

    Kinematic alignment is a theory of limb alignment that aims to improve patient satisfaction by recreating native knee kinematics. Potential benefits of this technique include restoration of native tibiofemoral kinematics, native limb alignment with even load distribution through the implant, and native soft tissue tension postoperatively.

    When restoring the native tibiofemoral articular surface, the surgeon aligns the rotational axes of the arthroplasty components with the kinematic axes of the knee. Proponents of kinematic alignment techniques theorize that changing the tension of the supporting soft tissue structures contributes to a patient’s sensation of pain, stiffness, or instability, and it may account for the proportion of patients who are dissatisfied with their TKA. [4] They also argue that restoring all knees to a “neutral” alignment during TKA may not take into consideration patients with native constitutional varus or valgus. Placing those patients in mechanical alignment may cause greater strain deviations in the tension of the collateral ligaments.

    In flexion past 45°, mechanical alignment that utilizes the gap balancing technique results in an inadvertent tightening of the soft tissues to match the balance of the extension gap. This does not match the normal kinematic function of the knee or collateral ligament tension. The kinematic alignment technique recreates the native articular surface and the tibiofemoral native limb alignment, which theoretically leads to better maintained soft tissue tensions.

    Kinematic alignment has not been as thoroughly studied as more traditional operative techniques. Despite this, recent randomized controlled trials and multicenter studies evaluating kinematic alignment have demonstrated better pain relief, function, and motion and a more “normal-feeling” knee when compared with traditional TKA techniques. Additional research has demonstrated comparable early survival of kinematically aligned implants out to 6 years of follow-up. [6-14]

    Limitations of kinematic alignment include:

    • Limited study of the ability to correct severe fixed limb deformity
    • Steeper learning curve than mechanical alignment
    • Utilization of implants that were originally designed for mechanical alignment techniques
    • Lack of long-term survivorship data

    Long-term data are needed before the kinematic alignment technique can be formally recommended and/or broadly accepted. There is evidence to support its use in general primary TKA, and it appears that early outcomes are at least comparable to traditional mechanical alignment, if not improved. However, early adoption is limited by little knowledge of long-term survivorship, and early reports on patient outcomes should be evaluated cautiously.

    Gap Balancing Versus Measured Resection

    Soft Tissue Releases

    The gap balancing and measured resection techniques rely on soft tissue release to achieve the optimal knee balance.

    In the sagittal plane, 2 major releases address flexion and extension gap mismatch:

    • If the knee is tight in extension, the indicated release is to strip the posterior knee capsule.
    • If the knee is tight in flexion, the indicated release is to resect a portion of the posterior cruciate ligament fibers.

    With coronal plane deformity, the tissues will either remain unchanged or become contracted on the concave side. On the convex side, the soft tissues can attenuate. On the medial side of the knee, relevant soft tissue structures include: [15]

    • Pes anserine tendons
    • Superficial medial collateral ligament (MCL)
    • Posteromedial corner
    • Semimembranosus muscle insertion
    • Medial joint capsule
    • Deep MCL

    Different medial exposures to the proximal tibia are utilized for varus or valgus deformity:

    • In a valgus knee, a subperiosteal dissection of the joint capsule and deep MCL should be performed. However, the superficial MCL attachment should be preserved, and the semimembranosus attachment should be maintained.
    • In a varus knee, the exposure calls for a subperiosteal elevation of the capsule, deep MCL, part of the superficial MCL, and the semimembranous insertion.

    The classic medial soft tissue release includes a continued subperiosteal elevation of the distal aspect of the superficial MCL, the pes tendons, and even further distal periosteum and soft tissue. Alternatively, some authors advocate for a less-extensive subperiosteal release and instead utilize a pie-crusting technique to selectively lengthen the MCL with either a knife or a needle. [16] This technique is highly user dependent, with a risk of iatrogenic MCL failure.  Further study is warranted. [17]

    In a valgus knee, contracted structures include: [15]

    • Iliotibial band
    • Anterior lateral ligament
    • Lateral collateral ligament (LCL)
    • Posterolateral corner
    • Popliteus tendon

    The inability to perform a subperiosteal dissection complicates lateral releases:

    • In extension, the posterolateral capsule can be carefully released at the level of the tibia because the peroneal nerve is only about 1 cm away. The iliotibial band can also be released with a pie-crust technique.
    • In flexion, the LCL should be released with a pie-crust technique, as well as any other palpably tight structures. The popliteus tendon should be preserved due to its role as a dynamic knee stabilizer.

    Gap Balancing Surgical Technique

    With the gap balancing technique, knee balance is primarily achieved through bony resection. Ligament releases may be necessary to correct fixed deformities or to restore neutral mechanical alignment.

    The most common approach to gap balancing involves cutting the extension gap followed by the flexion gap:

    • The knee is exposed with a subperiosteal release of the deep MCL.
    • The distal femur cut is made between 3° and 7° of valgus. The proximal tibia cut is resected exactly perpendicular to its mechanical axis.
    • All osteophytes are removed from the tibial plateau and posterior femoral condyles.
    • A tensioning device is used to assess extension gap balance and alignment.
    • Soft tissues can be released sequentially until perfect balance is achieved. [18] A medial tibial reduction osteotomy may be necessary to achieve gap symmetry in large fixed varus deformities. [19]
    • The knee is flexed to 90° and each collateral ligament is equally tightened with a tensioning device.
    • Femoral rotation is then set and an appropriate-sized femoral cutting block is used to complete the femoral bone cuts.
    • When setting femoral rotation, the surgeon ensures that there is a symmetric, rectangular flexion gap that is exactly equal in dimension to the extension gap.
    • Bony landmarks such as the transepicondylar axis (TEA) and Whitesides’s line are identified and used as secondary determinants of femoral rotation to help prevent major errors in component positioning.

    Measured Resection Surgical Technique

    Either the femur or the tibia can be resected first and in their entirety with the measured resection technique because all bone resections are independent of one another. Bone resections are fixed relative to bony landmarks, implant dimensions, and patient anatomy. After all bone cuts are made, spacer blocks or trial components are utilized to evaluate soft tissue tensions and subsequently balance them as needed.

    Distal femoral rotation is determined by a combination of Whiteside’s line, the TEA, or the posterior condylar axis (PCA): [33]

    • Whiteside’s line follows the deepest point of the trochlear groove; the femoral component should be perpendicular to this line.
    • The TEA equally bisects the medial and lateral epicondylar prominences and is perpendicular to Whiteside’s; the femoral component should be parallel to this line.
    • The PCA is a line drawn across the bottoms of the posterior femoral condyles at 90° of flexion. This line is about 3° to 5° internally rotated relative to the TEA, and the subsequent femoral component should be 3° to 5° externally rotation from the PCA.

    On the tibial side, the thickness of the plateau resection should, at minimum, account for the eventual thickness of the tibial component. Depending on the alignment philosophy of the surgeon (ie, mechanical versus kinematic), this tibia cut may be made either perpendicular to the mechanical axis or in a slight degree of varus.

    After all femoral and tibial cuts are performed, additional bone resection should not be used to help balance the knee. Instead, soft tissue releases should be used to perfect the balance.

    The Case for Gap Balancing

    Katz et al [20] compared the reliability of the measured resection technique using the TEA and Whiteside’s line with the gap balancing technique in cadaveric knees. They found that the TEA axis was less reliable and resulted in more femoral component external rotation than either Whiteside’s or the gap balancing technique. The authors concluded that gap balancing may be the more reproducible technique for achieving knee balance because it is independent of obscured or poorly defined bony landmarks. At minimum, Babazadeh et al [35] found no significant difference in femoral component rotation between the measured resection technique and the measured resection technique.

    Griffin et al [21] performed an in vivo study of the gap balancing technique in 84 knees. None of the knees demonstrated a flexion versus extension gap mismatch of more than 3 mm. Similarly, 38 randomly selected knees were measured by CT scan, with 90% showing that the posterior condylar angle of the femoral component was within 3° of the TEA.

    Dennis et al [22] investigated femoral condylar lift off. They compared 20 TKAs performed using gap balancing (all posterior stabilized designs) with 40 TKAs performed using measured resection. (20 posterior stabilized design and 20 cruciate retaining design). The knees were examined fluoroscopically and an automated 3D model fitting kinematic analysis was used to determine the amount of femoral lift off at 0°, 30°, 60°, and 90° of flexion. None of the gap balanced knees had femoral lift off of more than 1 mm, compared with 60% of the CR knees and 45% of the PS knees. Maximum lift off in the gap balanced knees was 0.9 mm, compared with 3.1 mm in the CR knees and 2.5 mm in the PS knees.

    The Case Against Gap Balancing

    With the gap balancing technique, the surgeon must have a completely accurate proximal tibial cut to create the appropriate femoral rotation. Abnormal femoral rotation will overstuff one side of the flexion space. Gap balancing may also lead to over- or under-resection of the extension cuts, which would alter the joint line and make it more difficult to balance the knee through its range of motion. Finally, an overly wide extension gap will result in a similar over-resection of the posterior condyles and reduced posterior offset. This may lead to earlier impingement and a limited deep flexion arch.

    Major deficiencies in the ligamentous structures will result in erroneous bone cuts. For example, excessive resection of the lateral posterior femoral condyle and internal rotation of the femoral component will result if the medial side has a larger flexion gap than the lateral side due to a compromised MCL. Heesterbeek et al [29] reported that knees with major medial releases had less external rotation of 2° +/- 4.2°. In contrast, knees with minor lateral releases had the most external rotation (7° +/- 3.8°).

    Furthermore, correction of a varus knee with gap balancing will result in residual varus limb alignment when the knee is flexed. [36] Lee et al [37] evaluated 44 patients undergoing computer navigated TKA with the gap balancing technique. Gap balancing provided good intraoperative alignment and laxity at 0° and 90° of flexion. However, increased femoral component external rotation was found, with increased non-physiologic varus alignment at 90°. The more the femur was externally rotated, the greater the varus alignment at 90°.

    Gap balancing is also unreliable when recreating femoral bony anatomy. Matziolis et al [34] evaluated 67 knees performed with the gap balancing technique and computer navigation. Even when gap balancing is done correctly, there is rotation from the TEA that ranges from 7.4° of internal rotation to 5.9° of external rotation.

    Subluxation of a tight extensor mechanism (quadriceps tendon, patella, and patellar tendon) when at 90° of flexion can overly tighten the lateral side of the knee beyond its physiologic boundaries and lead to inappropriate femoral bone resection. Subsequent malrotation of the femoral component is associated with: [30,31]

    • Patellofemoral instability
    • Tibiofemoral instability
    • Arthrofibrosis
    • Knee pain
    • Abnormal knee biomechanics

    The Case for Measured Resection

    Babazadeh et al [35] compared 103 patients who were randomized to gap balancing or measured resection with computer navigation. Gap balancing significantly raised the joint line compared with measured resection (2.18 mm versus 0.63 mm). This was attributable to a greater distal femoral resection (11.2 mm versus 9.6 mm), with a larger tibial polyethylene insert used to achieve balance. Elevating the joint line leads to adverse patellofemoral mechanics because the patella and patellar tendon then impinge on the tibial component.

    Measured resection creates better alignment in flexion than gap balancing. Hanada et al [36] describe how gap balancing causes the knee to shift into varus when at 90° of flexion and overload the medial flexion space. Measured resection avoids this problem, likely due to the fact that gap balancing does not take into account the natural laxity of the lateral side of the knee compared with the medial side. [37] Gap balancing also over-tensions the lateral side with the tensioning device. [37]

    Placement of the femoral component parallel to the TEA is known to lessen the incidence and magnitude of femoral condylar liftoff. [28] Olcott and Scott [38] evaluated femoral component positioning and found that a rectangular flexion gap was produced 90% of the time when the femoral component was aligned with the TEA. Comparatively, a rectangular gap is achieved 83% of the time when using Whiteside’s line and 70% of the time when using the PCA.

    Measured resection also avoids compounded errors in surgical technique. A crucial component of the gap balancing technique is creation of a tibial cut that is perpendicular to the mechanical axis. Any error in the tibial cut will be compounded later in the gap balancing procedure because it will adversely affect femoral rotation and limb alignment during flexion. [22]

    The Case Against Measured Resection

    Because the flexion gap is not “balanced” against the width of the extension gap in measured resection, there is a greater risk of a gap mismatch that is difficult to balance with only soft tissue releases. There is also a significant risk of having a non-rectangular flexion gap. An asymmetric flexion gap is associated with tibial radiolucent lines, [39] as well as condylar lift-off, polyethylene surface damage, and uneven polyethylene wear patterns. [22]

    Ferhing [23] reported that 45 of 100 TKAs performed with the measured resection technique had a trapezoidal flexion gap. Similarly, Clatworthy et al [24] reported on 212 knees and compared results using measured resection and bony landmarks with computer navigation. Each of the bony landmarks had a wide divergence of femoral component rotation; reliable recreation of a rectangular flexion gap was not possible using any landmark. Only 41.5% of cases using TEA were actually within +/- 3° of the target rotation. Whiteside’s was within +/- 3° in 39.6% of cases, and the PCA was within +/- 3° in 51.9% of cases.

    The measured resection bony landmarks are often unreliable or absent. [32] Yau et al [25] used postoperative CT scans to evaluate the TEA compared with the actual femoral component in 25 TKAs in 14 patients. Outliers greater than 5° occurred in 72% of cases when the component was referenced from the PCA, 60% of cases when using Whiteside’s, 52% of cases when using TEA, and only 20% of cases when the gap balancing technique. Jerosch et al [26] demonstrated that the surgeon-identified location of the medial epicondyle varied by 22.3 mm from the true location; the lateral epicondyle varied by 13.8 mm.

    In a cadaveric study by Katz et al, [20] the identification of Whiteside’s line was less reliable in patients with significant trochlear dysplasia. Yau et al [25] conducted a study that demonstrated femoral components ranged from 15° external rotation to 17° internal rotation when using Whiteside’s line as the measured resection landmark.

    Other studies have shown that the PCA is not a reliable bony landmark for measured resection. The PCA ranges normally from -1° to 7° of external rotation. [27] The standard 3° of external rotation from this landmark is, therefore, just an average and not necessarily matching the patient’s anatomy. Furthermore, in patients with lateral femoral condyle hypoplasia or erosion (eg, valgus knees), the component will be internally rotated if measured resection is based solely on the PCA. [21]

    The measured resection technique has a higher risk of iatrogenic injury. Ligament releases are crucial to this technique because bony changes are not used to achieve the balance, only soft tissue releases. Thus, when a surgeon manipulates the soft tissues to a greater degree to achieve balance, there is an associated risk of iatrogenic injury.


    There is increasing interest in kinematic alignment, largely due to recent studies that suggest better pain relief, function, range of motion, and “normal” sensation of the operated joint compared with a traditional mechanically aligned TKA. Gap balancing and measured techniques can be used to achieve a well-balanced TKA, either in isolation or in combination.

    Regardless of the technique used, the goal of TKA is a well-aligned and well-balanced knee, achieved through whatever means the surgeon feels most comfortable using. The surgeon should also realize that the technique may make no difference in a certain percentage of patients in whom it may be impossible to achieve perfect alignment and balance.

    Author Information

    John J. Mercuri, MD, MA; Andrew M. Pepper, MD; and Jordan A. Werner, MD, are adult reconstruction fellows, Insall Scott Kelly Institute, Department of Orthopaedic Surgery, NYU Langone Health – Hospital for Joint Diseases, New York, New York. Jonathan M. Vigdorchik, MD, is an Assistant Professor of Orthopaedic Surgery, Associate Fellowship Director in Adult Reconstruction, and Co-Director of Robotics in Orthopaedic Surgery at NYU Langone Medical Center’s Hospital for Joint Diseases, New York, New York.


    1. Sierra RJ, Rachala SR. Insall & Scott Surgery of the Knee. 6th Ed. Ch. 125 Unicompartmental, Bicompartmental, or Tricompartmental Arthritis of the Knee: Algorithm for Surgical Management.
    2. Vail TP, Lang JE. Insall & Scott Surgery of the Knee. 6th Ed. Ch. 150 Surgical Techniques & Instrumentation in Total Knee Arthroplasty.
    3. Nam D, Nunley RM, Barrack RL: Patient dissatisfaction following total knee replacement: a growing concern? Bone Joint J 96-B(11 Suppl A):96–100, 2014.
    4. Howell SM, Hull ML, Mahfouz MR. Insall & Scott Surgery of the Knee. 6th Ed Ch 160 Kinematically Aligned Total Knee Arthroplasty.
    5. Hsu HP, Garg A, Walker PS, et al: Effect of knee component alignment on tibial load distribution with clinical correlation. Clin Orthop 248:135, 1989
    6. Dossett HG, Estrada NA, Swartz GJ, et al: A randomised controlled trial of kinematically and mechanically aligned total knee replacements: two-year clinical results. Bone Joint J 96-B(7):907–913, 2014
    7. Dossett HG, Swartz GJ, Estrada NA, et al: Kinematically versus mechanically aligned total knee arthroplasty. Orthopedics 35(2):e160e169, 2012.
    8. Ritter MA, Faris PM, Keating EM, et al: Postoperative alignment of total knee replacement: its effect on survival. Clin Orthop 299:153, 1994.
    9. Dossett HG, Estrada NA, Swartz GJ, et al: A randomised controlled trial of kinematically and mechanically aligned total knee replacements: two-year clinical results. Bone Joint J 96-B(7):907–913, 2014.
    10. Dossett HG, Swartz GJ, Estrada NA, et al: Kinematically versus mechanically aligned total knee arthroplasty. Orthopedics 35(2):e160e169, 2012.
    11. Howell SM, Howell SJ, Kuznik KT, et al: Does a kinematically aligned total knee arthroplasty restore function without failure regardless of alignment category? Clin Orthop 471(3):1000–1007, 2013.
    12. Howell SM, Papadopoulos S, Kuznik K, et al: Does varus alignment adversely affect implant survival and function six years after kinematically aligned total knee arthroplasty? Int Orthop 1–8, 2015.
    13. Eckhoff DG, Bach JM, Spitzer VM, et al: Three-dimensional mechanics, kinematics, and morphology of the knee viewed in virtual reality. J Bone Joint Surg Am 87(Suppl 2):71–80, 2005.
    14. Roth JD, Howell SM, Hull ML: Native knee laxities at 0 degrees, 45 degrees, and 90 degrees of flexion and their relationship to the goal of the gap-balancing alignment method of total knee arthroplasty. J Bone Joint Surg Am 97(20):1678–1684, 2015.
    15. Clarke HD, Scott WN. Knee: axial instability. Orthop Clin North Am. 2001;32(4):627–37–viii.
    16. Ha C-W, Park Y-B, Lee C-H, Awe S-I, Park Y-G. Selective Medial Release Technique Using the Pie-Crusting Method for Medial Tightness During Primary Total Knee Arthroplasty. J Arthroplasty. 2016;31(5):1005–1010. doi:10.1016/j.arth.2015.11.019.
    17. Meneghini RM, Daluga AT, Sturgis LA, Lieberman JR. Is the pie-crusting technique safe for MCL release in varus deformity correction in total knee arthroplasty? J Arthroplasty. 2013;28(8):1306–1309. doi:10.1016/j.arth.2013.04.002.
    18. Petra J. C. Heesterbeek, MSc Wilco C. H. Jacobs, MSc, and Ate B. Wymenga, MD, PhD. Effects of the Balanced Gap Technique on Femoral Component Rotation in TKA. Clin Orthop Relat Res. 2009 Apr; 467(4): 1015–1022.
    19. Arun B. Mullaji, FRCS Ed, MCh Orth, MS Orth and Gautam M. Shetty, MS Orth. Correction of Varus Deformity During TKA With Reduction Osteotomy. Clin Orthop Relat Res. 2014 Jan; 472(1): 126–132.
    20. Katz MA, Beck TD, Silber JS, Seldes RM, Lotke PA. Determining femoral rotational alignment in total knee arthroplasty: reliability of techniques. J Arthroplasty. 2001 Apr;16(3):301-5.
    21. Griffin FM, Insall JN, Scuderi GR. The posterior condylar angle in osteoarthritic knees. J Arthroplasty. 1998 Oct;13(7):812-5.
    22. Dennis DA, Komistek RD, Kim RH, Sharma A. Gap balancing versus measured resection technique for total knee arthroplasty. Clin Orthop Relat Res. 2010 Jan;468(1):102-7. doi: 10.1007/s11999-009-1112-3.
    23. Fehring TK. Rotational malalignment of the femoral component in total knee arthroplasty. Clin Orthop Relat Res. 2000 Nov;(380):72-9.
    24. Clatworthy MG, Lindberg K, Harrill W, et al. Rotational alignment of the femoral component in computer-assisted total knee arthroplasty. Reconstr Rev Joint Implant Surg Res Found. 2012 (2):37-42.
    25. Yau WP, Chiu KY, Tang WM. How precise is the determination of rotational alignment of the femoral prosthesis in total knee arthroplasty: an in vivo study. J Arthroplasty. 2007 Oct;22(7):1042-8.
    26. Jerosch J, Peuker E, Philipps B, Filler T. Interindividual reproducibility in perioperative rotational alignment of femoral components in knee prosthetic surgery using the transepicondylar axis. Knee Surg Sports Traumatol Arthrosc. 2002 May;10(3):194-7. Epub 2002 Feb 5.
    27. Poilvache PL, Insall JN, Scuderi GR, Font-Rodriguez DE. Rotational landmarks and sizing of the distal femur in total knee arthroplasty. Clin Orthop Relat Res. 1996 Oct;(331):35-46.
    28. Insall JN, Scuderi GR, Komistek RD, Math K, Dennis DA, Anderson DT. Correlation between condylar lift-off and femoral component alignment. Clin Orthop Relat Res. 2002 Oct;(403):143-52.
    29. Heesterbeek PJ, Jacobs WC, Wymenga AB. Effects of the balanced gap technique on femoral component rotation in TKA. Clin Orthop Relat Res. 2009 Apr;467(4):1015-22. doi: 10.1007/s11999-008-0539-2. Epub 2008 Oct 2.
    30. Boldt JG, Stiehl JB, Hodler J, Zanetti M, Munzinger U. Femoral component rotation and arthrofibrosis following mobile-bearing total knee arthroplasty. Int Orthop. 2006 Oct;30(5):420-5. Epub 2006 Mar 7.
    31. Romero J, Stähelin T, Binkert C, Pfirrmann C, Hodler J, Kessler O. The clinical consequences of flexion gap asymmetry in total knee arthroplasty. J Arthroplasty. 2007 Feb;22(2):235-40.
    32. Dennis DA. Measured resection: an outdated technique in total knee arthroplasty. Orthopedics. 2008 Sep;31(9):940, 943-4.
    33. Brooks P. Seven cuts to the perfect total knee. Orthopedics. 2009;32(9). doi:10.3928/01477447-20090728-27.
    34. Matziolis G, Boenicke H, Pfiel S, Wassilew G, Perka C. The gap technique does not rotate the femur parallel to the epicondylar axis. Arch Orthop Trauma Surg. 2011 Feb;131(2):163-6. doi: 10.1007/s00402-010-1113-4. Epub 2010 May 18.
    35. Babazadeh S, Dowsey MM, Stoney JD, Choong PF. Gap balancing sacrifices joint-line maintenance to improve gap symmetry: a randomized controlled trial comparing gap balancing and measured resection. J Arthroplasty. 2014 May;29(5):950-4. doi: 10.1016/j.arth.2013.09.036. Epub 2013 Oct 1.
    36. Hanada H, Whiteside LA, Steiger J, Dyer P, Naito M. Bone landmarks are more reliable than tensioned gaps in TKA component alignment. Clin Orthop Relat Res. 2007 Sep;462:137-42.
    37. Lee DS, Song EK, Seon JK, Park SJ. Effect of balanced gap total knee arthroplasty on intraoperative laxities and femoral component rotation. J Arthroplasty. 2011 Aug;26(5):699-704. doi: 10.1016/j.arth.2010.06.005. Epub 2010 Jul 20.
    38. Olcott CW, Scott RD. A comparison of 4 intraoperative methods to determine femoral component rotation during total knee arthroplasty. J Arthroplasty. 2000 Jan;15(1):22-6.
    39. Romero J, Stähelin T, Binkert C, Pfirrmann C, Hodler J, Kessler O. The clinical consequences of flexion gap asymmetry in total knee arthroplasty. J Arthroplasty. 2007 Feb;22(2):235-40.