We report the outcome at a minimum of 10 years follow-up for 80 polished tapered stems performed in 53 patients less than 35-years-old with a high risk profile for aseptic loosening. Forty-six prosthesis were inserted for inflammatory hip arthritis and 34 for avascular necrosis. The mean age at surgery was 28 years in the inflammatory arthritis (17–35) and 27 years in the avascular necrosis (15–35) patients. At a mean follow-up of 14.5 years in the inflammatory arthritis group and 14 years in the avascular necrosis group respectively, survivorship of the 80 stems with revision of the femoral component for any reason as an endpoint was 100 % (95 % CI). Re-operation was because of failure of four metal-backed cups, 3 all polyethylene cups and one cementless cup. None of the stems were radiographically loose. All but two femoral components subsided within the cement mantle to a mean of 1.2 mm (0 tot 2.5) at final follow-up. Periarticular osteolysis was noted in 4 femurs in zone 7. This finding was associated with polyethylene wear and was only seen in those hips that needed revision for a metal backed cup loosening. Our findings show that the polished tapered stem has excellent medium-term results when implanted in young patients with high risk factors for aseptic loosening.

In general TKA can be divided into two distinct groups: cruciate retaining and cruciate substituting. The cam and post of the latter system is in fact a mechanical substitution of the intricate posterior cruciate ligament. In our previous work we and many other investigators have focused on the movement of the femoral component relative to the tibial tray. Little information is available about the relative movement between the cam part of the femoral component and the post of the tibial insert. In this study we determine the distance and the changes in distance between the cam of the femoral component and the tibial post during extension, flexion at 90° and full flexion. The secondary purpose is to analyse possible differences between FBPS and MBPS TKA.

Introduction

Autologous chondrocyte implantation presents a viable alternative to microfracture in the repair of damaged articular cartilage of the knee; however, outcomes for patellar lesions have been less encouraging. ChondroCelect (CC) is an innovative, advanced cell therapy product consisting of autologous cartilage cells expanded ex vivo through a highly controlled and consistent manufacturing process.

A comparative kinematic study was carried out on six cadaver limbs, comparing tibiofemoral kinematics in five different conditions: unloaded, under a constant 130 N ankle load with a variable quadriceps load, with and without a constant 50 N medial and lateral hamstrings load. Kinematics were described as translation of the projected centers of the medial (MFT) and lateral femoral condyles (LFT) in the horizontal plane of the tibia, and tibial axial rotation (TR) as a function of flexion angle. In passive conditions, the tibia rotated internally with increasing flexion, to an average of −16° (range −12/−20°, SD 3.0°). Between 0 – 40° flexion, the medial condyle translated forwards 4 mm (range 0.8/5.5 mm, SD 2.5 mm), followed by a gradual posterior translation, totaling −9 mm (range −5.8/−18.5 mm, SD 4.9 mm) between 40° – 140° flexion. The lateral femoral condyle translated posteriorly with increasing flexion completing −25 mm (range −22.6 – −28.2 mm, SD 2.5 mm). Dynamic, loaded measurements were carried out in a knee rig. Under a fixed ankle load of 130 N and variable quadriceps loading, tibial rotation was inverted, mean TR 4.7° (range −3.3°/11.8° SD 5.4°), MFT −0.5 mm (range = −4.3/2.4 mm, SD = 2.4 mm), LFT 3.3 mm (range = −3.6/10.6 mm, SD = 5.1 mm). As compared to the passive condition, all these excursions were significantly different: p=0.015, p=0.013, and p=0.011 for TR, MFT and LFT respectively. Adding medial and lateral hamstrings force of 50N each, reduced TR, MFT and LFT significantly as compared to the passive condition. In general, loading the knee with hamstrings and quadriceps reduces rotation and translation as compared to the passive condition. Lateral hamstring action is more influential on knee kinematics than medial hamstrings action.

Purpose: Precise biomechanical reconstruction of the hip joint by a hip arthroplasty is essential for the success of this procedure. With the increasing use of surface replacement arthroplasty (SRA), there is a need for better understanding of the key factors that influence the anatomical and the biomechanical parameters of the resurfaced hip joint. The goal of this study was to examine the influence of SRA on the vertical and horizontal offset of the hip.

Method: Twenty-one hips from 12 embalmed cadavers were resurfaced with a Birmingham Hip resurfacing. The thickness of the acetabular bone was measured pre- and post-reaming in 6 acetabular zones. Radiographs were taken before and after the procedure with a scaling marker. For statistical analysis, the paired Student’s T-test with a confidence interval of 95% and a significant p-value of p< 0.05 was used.

Results: The mean acetabular bone loss was 3.8 mm, 5.9 mm, 9.3 mm, 10.6 mm, 8.5 mm and 3.6 mm in zones 1 to 6. The “polar length loss” is the cumulative displacement of the femoral and the acetabular articulating surface in zones 2 to 5. This displacement indicates a shortening of the neck plus a medio-cranial displacement of the acetabular articulating surface and was 4.3 mm, 7.5 mm, 9.4 mm and 7.7 mm (zone 2–5). The radiographic center of rotation (COR) was significantly medialised (mean 6.2 mm) and displaced in the cranial direction (mean 6.9 mm) (p< 0.00001). The mean total (femoral plus acetabular) horizontal and vertical offset change was 6.4 mm and 9.5 mm respectively (p< 0.00001). There was a significantly higher vertical offset change in the acetabulum than in the femur (p=0.0006). This resulted in a significantly larger change in vertical than in horizontal offset (p=0,04).

Conclusion: The displacement of the acetabular COR was responsible for 60% of the total vertical and 99% of the total horizontal offset change. The femoral side did not compensate for this displacement. SRA did not restore the biomechanics of the native hip.

Purpose: Plate and cable alone constructs to treat periprosthetic fractures around a well-fixed femoral stem in total hip replacements, have been reported with high failure rates. The aim of this study was to evaluate the results of our surgical treatment algorithm to reliably use lateral plate and cable constructs in these fractures.

Method: One hundred and six periprosthetic fractures in 102 patients were treated between 1996 and 2006. Forty-five fractures were pre-operatively assessed as Vancouver type B1 fractures. The joint was always dislocated and stability of the implant was meticulously evaluated. This led to the identification of nine (20%) unstable stems leaving 36 fractures to be real B1-type fractures. The fracture was considered to be suitable for lateral plate and cable alone fixation if the medial cortex was not comminuted and an anatomical reduction of the medial cortex could be achieved. Twenty-nine B1- and 5 C-type fractures had been treated with a single lateral plate and cable construct. The mean length of follow-up was 43.2 months. The paired Student’s T-test with a confidence interval of 95% and a significant p-value of p< 0.05 was used to compare the pre- and postoperative UCLA hip scores.

Results: Four (12%) patients died within one month from surgery leaving 30 patients for follow-up. Twenty-nine fractures united at a mean of 6.4 months. One B1-type construct failed due to inappropriate proximal fixation. Two fractures united uneventfully with a mean of 8° of varus alignment of the proximal fragment. One patient with a C-type fracture sustained a fracture distal to the tip of the plate. There were three plate infections (8.8%). There was no significant difference between the pre- and post-operative UCLA hip scores (25 versus 23 resp.).

Conclusion: These fractures represent a difficult problem with a high complication rate of 30%. The presented treatment algorithm contributed significantly to the 97% union rate with plate and cable alone constructs that was comparable to the union rates achieved with combined plate and strut graft fixation.

Introduction: High-flexion (HF) TKA designs were introduced in order to achieve greater flexion than with conventional TKA designs. Although early clinical results are promising, recent literature raises concerns about fixation and risk for early loosening of the femoral component during high demanding activities. This study’s aim was to measure the loosening force of the femoral component of several PS-TKA designs in a deep flexion configuration.

Methods: The loosening force of the femoral component of ten contemporary PS-TKAs, including five HF and five conventional designs from the major orthopaedic companies were evaluated. To simulate a deep flexion configuration, each TKA was implanted in a femoral bone model and placed in a loading frame in 135° of flexion, with the tibia vertically. Loosening of the femoral component was induced by raising the tibial insert with constant displacement rate, maintaining the same flexion angle. The resisting force was recorded continuously. A stereo-photogrammetric system registered the relative motion between the femoral component and the bone model. The loosening force was determined when a gap of 2 mm was observed. The influence of pegs on the loosening force was also investigated.

Results: Generally, conventional femoral designs required higher forces before loosening occurred compared to HF designs (p< 0.001). In the group of the HF designs there was a statistically significant difference between the designs (p=0.015) due to the shape of the internal box cut. For some designs, the presence of pegs induced a statistically significant change in loosening force.

Discussion and Conclusion: Several design characteristics of the femoral component can alter its resistance to loosening. In this in vitro study, it was shown that the shape of the internal box cut and the presence of pegs, as well as the geometry of the pegs, are important factors for the loosening force. In the group of the HF components there was a statistically significant difference between the designs with an open and a closed box.

Aim: Component positioning may be adversely affected by minimally invasive approach in total hip replacement due to restricted visualization. Problems with proper alignement are suggested to concern anteversion more than inclination and occur particulary in the lateral position.

Method: 53 patients were enrolled prospectively randomised to each group. First group (standard group, n= 30pts) underwent conventional total hip replacement in supine position and transgluteal approach and second group (MIS group, n= 23pts) underwent THR using minimally invasive anterior approach in lateral decubitus position Every group was operated on by two experienced senior surgeons. Desired cup position was 40°–45°inclination and 15–20° anteversion for the MIS group and 45°inclination and 15 ° anteversion for standard group. Postoperatively all patients had pelvic CT scan. Inclination and anteversion were determined by an independent observer using a 3-D model and planning software, the operative definition was used according to Murray.

Results: Mean inclination/anteversion in the MIS group was 39°(26°–50°)/25°(10°–47°), and 44°(29°–57°)/22°(1°–53°) within the standard group. Standard deviation for inclination was 7° for both groups, and 10° (MIS group) vs 14° (standard group) for anteversion.

The difference in the mean values regarding inclination was greater than would be expected by chance; there was a statistically significant difference (P = 0,010).

Discussion: In general cup positioning in both groups was less steep and more anteverted as presumed. The standard deviation for inclination was the same in both groups, but the standard deviation for anteversion was less in MIS group, that means less outliers regarding anteversion. Cup positioning in minimally invasive total hip replacement is safe compared to traditional approach.

Navigation technique was discussed to equalize the drawback of MIS. However, tools like imageless navigation may further improve the cup position even in traditional approach.

Purpose: A three-year evaluation of long-term clinical efficacy of Characterized Chondrocyte Implantation (CCI) compared to microfracture (MF), in the repair of symptomatic cartilage defects of the femoral condyles at 36 months post-surgery.

Materials and Methods: In a prospective, randomized, controlled, multicenter trial, CCI was compared to MF in patients aged 18–50 years with a single symptomatic ICRS grade III–IV lesion of the knee. Clinical outcome was measured 36 months after surgery by means of the KOOS, VAS for pain and ARS, with a non-inferiority margin preset at 9 % points for KOOS and VAS. Furthermore, response to treatment and progression of knee symptoms were assessed. Treatment failure was monitored throughout the study.

Results: Improvement from baseline was higher in the CCI group (N = 41) compared to the MF group (N = 49) for all clinical outcome parameters. Mean improvement from baseline for Overall KOOS was 22.14 vs. 14.48, respectively, with VAS and ARS scores revealing a similar trend. Responder analysis showed 83% of the patients treated with CCI improving vs. 61% after MF. Additionally, we observed a shift in the proportion of knee symptoms over time (52% vs. 35% of asymptomatic knees at 36 months compared to 2% vs. 8% at baseline in the CCI and MF group respectively). At 36 months, failure rates were low in both groups (n=2 in CCI vs. n=7 in MF).

Conclusions: Previous data have described a superior structural repair after CCI compared to MF at 1 year post-surgery. Continued clinical improvement as well as a favorable responder analysis was demonstrated for CCI compared to MF at 36 months.

Introduction: The Birmingham Hip Resurfacing (BHR) is the most commonly used hip resurfacing for the treatment of hip osteoarthritis. The goal of this study was to evaluate how the surgeon could influence the biomechanical features of the navigated and non-navigated resurfaced proximal femur. METHODS 20 Cadaver hips were resurfaced with a BHR using femoral navigation. The native anteversion and neck shaft angle as indicated by the navigation system were used as a reference. The non-navigated femoral component jig was first placed in the “ideal” position aiming for 10° of valgus and neutral anteversion. The jig was then displaced 5mm in 4 directions. The anteversion and stem shaft angle (SSA) angle were measured for each position using the navigation system. A scaled XR was taken pre- and post-operatively. For statistical analysis, the paired Student’s T-test with a confidence interval of 95% and a significant p-value of p< 0.05 was used.

Results: The centre of rotation (COR) of the navigated resurfaced femur was 3,5 mm significantly (p=0,0006) more distal in the femoral neck than the native COR. This resulted in a 2.1 mm vertical caudal drop (vertical offset) and an average 2.7 mm lateral displacement of the COR (horizontal offset). The same measurements were done with 5° increments of the SSA from 120° to 140°. The vertical offset loss increased non-significantly (1.7 to 2.6 mm). The horizontal offset loss decreased non-significantly (3 to 2.2 mm). The native vertical and horizontal offset could be restored if 5 mm less bone was taken off the femur. The offset loss was significantly increased if 5 mm more bone than the normal reaming had been taken off (p< 0.0001). The “ideal” jig position on the lateral femoral cortex led to an average 137° SSA. Five millimetres of jig displacement on the lateral cortex in either direction did not lead to significant changes in the SSA or anteversion angles relative to the “ideal” position (all p> 0,13). Five millimetres of posterior displacement resulted in an average 139° SSA and 5,8° of anteversion in 95% of hips.

Conclusion: Surgical interventions can significantly change the biomechanics of the hip. Increasing the SSA with a fixed femoral head entry point, as often is done with navigation, does not significantly change the femoral offset. If the surgeon decides to take less bone off the femur, then the offset could be restored and even increased to 1 mm more than the native femur. If due to pathologic changes the bone loss would be increased to 5mm more than the “normal” bone loss, a significant offset loss of > 5 mm could be expected which might lead to detrimental biomechanical effects. The positioning of the jig is subject to surgical errors. The effect of a 5 mm error in either direction does not lead to significant changes in anteversion or SSA. Posterior displacement led to the most reproducible component positioning.

Purpose: As a one-step surgical procedure, microfracture is frequently considered to be technically easier and associated with less postoperative morbidity than autologous chondrocyte implantation (ACI), which involves both arthrotomy and arthroscopy and therefore safety was assessed in patients with symptomatic cartilage lesions of the knee treated with characterized chondrocyte implantation (CCI) or microfracture.

Methods: CCI (n=57) was compared to microfracture (n=61) in patients with grade III–IV symptomatic cartilage defects of the femoral condyles in a Phase III, prospective, multi-center, randomized, controlled trial. Safety assessments included adverse events (AEs), physical examination, vital signs, hematology and clinical chemistry.

Results: At 18 months post-surgery, similar proportions of patients experienced AEs in the CCI (88%) and microfracture (82%) groups; 67% and 59%, respectively, experienced AEs considered treatment related. The AE profile was generally similar between groups, with no significant difference for hypertrophy, although significantly more CCI-treated patients had joint swelling (19% versus 4.9%; p=0.022) and treatment-related joint crepitation (12% versus 1.6%; p=0.028). Although the proportion of patients with severe AEs was similar for CCI (12%) and microfracture (13%), slightly more microfracture-treated patients experienced serious (life-threatening or requiring hospitalization) AEs (13% versus 8.8%). No patients discontinued due to AEs or died during the study.

Conclusion: Contrary to general opinion, the two-step CCI procedure, involving arthroscopy followed by arthrotomy, has a similar safety profile to that of microfracture, a one-step arthroscopic procedure, for treating cartilage lesions of the knee.

Purpose: This study compared the efficacy and safety of Characterized Chondrocyte Implantation (CCI) to microfracture in the repair of symptomatic cartilage defects of the femoral condyle.

Methods: CCI (n=51) was compared to microfracture (n=61) in patients with grade III–IV symptomatic cartilage defects of the femoral condyles in a prospective, multicenter, randomized, controlled trial. Structural repair was assessed at 1 year by histopathologists blinded to the treatment using

computerized histomorphometry and

Results: CCI resulted in better structural repair than microfracture at 1 year post-treatment, as assessed by histomorphometry (p=0.003) and overall histology (p=0.012). Structural repair parameters relating to chondrocyte phenotype and tissue structure were also superior with CCI. Noninferiority of CCI was demonstrated for clinical outcome at 12–18 months, and both treatments were generally well tolerated.

Conclusion: At 1 year post-treatment, CCI resulted in superior tissue repair compared to microfracture. Short-term clinical outcome after 12–18 months was similar for both treatments, as was the safety profile. The superior structural repair achieved with CCI may lead to improved long-term clinical benefits.

As total knee arthroplasties (TKA) have become the gold standard procedure for severe gonarthrosis, greater interest in postperative tibiofemoral instability has developed.

Emphasizing the correlation between evaluation of symptoms and findings, offers an opportunity to elucidate the specifics of the instability.

Mandatory is the joint gap measurement during surgery to assess the effect of specific cuts or releases of the anatomic portion of the joint gap.

By performing navigation-assisted total knee arthroplasties, we are capable of measuring the joint gap in a highly reliable way.

During the ligament balancing in navigation-assisted TKA, we performed a data collection of the joint gap in 0–30 and 90 of flexion in 100 patients.

The measurements were repeated after 10 and 20 minutes in extension.

The result offers us an opportunity to assess the interesting effect of ligament-stress relaxation in TKA and to gain more insights in the further release-necessity and choice of insert during the TKA procedure.

Introduction The purpose of this study was to detect the effect of tibial slope on maximal flexion after TKA.

Methods Twenty-one cadaver implantations of a standard PCL-retaining TKA were performed with increasing tibial slope of zero degrees, four degrees and seven degrees. For every specimen all variables except slope were kept constant, including tibio-femoral contact locations in deep flexion, which were determined upon in vivo contact patterns that were obtained during maximal squatting activities in patients that had undergone TKR with the same design. Maximal flexion was determined by direct impingement of the tibial component on the posterior femoral bone fluoroscopy.

Results Maximal flexion correlated positively with increasing slope (p< 0.001, R2 = 0.8). When aimed slope was considered, flexion increased on average 2.2° for every degree of downslope. When obtained slope was considered, flexion increased on average 1.7° for every degree of downslope.

Conclusions In PCL-retaining TKA, maximal obtainable flexion icnreases on average two degrees per degree extra tibial slope.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction Although most surgeons agree that the functional results obtained with modern total knee arthroplasty are acceptable, it is clear that even with the most recent designs it is still impossible to duplicate the behaviour and functional performance of a normal knee.

Methods I present a review of the literature and personal experience.

Results Recent kinematic studies have shown that modern TKA designs consistently provoke aberrant kinematics compared to the normal knee, mainly due to the absence of the ACL and the inability to maintain a functional PCL. With regard to roll-back, PS cam-post designs appear to perform better than PCL retaining knees, but only in deeper degrees of flexion, usually only beyond 90°. Whether it is striclty necessary to try to obtain normal kinematics with our TKA designs, is still an open debate.

Conclusions It is clear however that the aberrant kinematics we have noted with the current designs, are the direct cause of the flexion limit we see in many of our patients. Furthermore they probably also are the basis for many of the discomorts associated with modern TKA, such as difficulties in stair descent, chair rise, pivoting activities, thrust instabilities etc. With regard to these issues, I believe there are two potential directions to improve our current TKA designs; (1) by introducing the concept of guided-motion (intrinsic mechanism), or (2) by maintaining or restoring the (extrinsic) determinants of kinematics, i.e. the cruciate ligaments, the joint configuration and the extra-articular structures.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction The purpose of this study was to detect the limiting mechanism of maximal active flexion that can ultimately be obtained by patients after PCL-retaining TKA.

Methods The study consisted of two parts. In the first part, 30 patients with well performing PCL-retaining TKAs were examined using videofluroscopy. In deep flexion, observations were directed towards potentiallly determinant factors of maximal obtainable flexion. Based upon these observations, a newly defined paramater, called the “posterior condylar offset”, was found to be important. The exact influence of this parameter was investigated in part two, in which 150 consecutive patients with PCL-retaining TKA were reviewed.

Results Aberrant kinematics were observed in the majority of cases. In 27 patients (93%) slide-forward of the femur was noted with flexion, with anterior translation of the medial and/or lateral femorotibial contact position. In deep squat, direct impingement of the posterior aspect of the tibial insert against the shaft of the femur was noted in 21 cases (72.4%), blocking further flexion. In part two of the sutdy, it was demonstrated that in knees with decreased post-operative “posterior condylar offset”, such impingement occurred faster and lead to decreased maximal obtainale flexion (p< 0.001).

Conclusions Maximal obtainable flexion is in the majority of cases determined by posterior tibial insertion impingement against the femoral bone and this occurs as a consequence of aberrant kinematics with anteiror sliding of the femur during flexion. Restoration of “posterior condylar offset” is important, since it allows greater degrees of flexion before impingement occurs.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction: Clinical follow-up studies of total knee arthroplasty (TKA) reveal good results. However, the range of flexion of the knee after TKA remains limited in most cases. The most important factors impacting this range are the length of the quadriceps, the capsular tightness, the tibial slope, the design of the implant, the use of physiotherapy and the surgical technique. Aims: Our aim was to measure maximal knee flexion as a function of the posterior slope of the tibial component. We tried to quantify the relationship between tibial slope and maximal knee flexion. Methods: Seven fresh-frozen cavader-knees were used, all of them had moderate osteo-arthritis. Mean age of the patients was 67. A posterior-cruciate-ligament retaining TKA was used. In each knee, the tibial component was implanted consecutively with a slope of 0°, 4° and 7°. For each slope, the maximal flexion was measured using fluoroscopy in a test rig. Results: The mean maximal flexion achieved was 104° for an implant with a 0° slope; 112° for 4° slope and 120° for 7° slope. We regressed the maximal flexion on the tibial slope, using a Pearson regression analysis and obtained a R2=0.8356; indicating a strong correlation. Conclusions: In PCL retaining TKA, there is an increase of 2° in flexion for every increase of 1° posterior tibial slope.

Performing a total knee arthroplasty in a patient with a flexion contracture or recurvatum deformity requires from the surgeon an adequate knowledge of the principles of flexion – extension space balancing.

In the standard TKA procedure, adequate balancing between the flexion and extension space is usually easily achieved, leading to an equal and symmetrical space both in flexion and extension, which results in a stable knee and maximal range of motion after implantation of the prosthetic components. The situation is different in the knee with a flexion contracture or recurvatum, where the extension space is relatively smaller (flexion contracture) or greater (recurvatum) than the flexion space. In both of these situations, the flexion and extension space should be balanced by the surgeon in order to avoid an important deficit in range of motion or an instability problem. Several surgical techniques are available for this.

In the knee with a flexion contracture, the extension space is relatively too small. Adequate removal of posterior osteophytes will increase the extension space, and this should be the first step in the flexion – extension space procedure (1). Next, the collateral structures should be balanced, with release of the tight structures that are effective in extension only (2).

These are predominantly the iliotibial band in the valgus knee, and the posterior oblique ligament in the varus knee. If these 2 steps are not sufficient, proximalisation of the femoral component by 2 to 3mm may be required (step 3), or a formal release of the posterior capsule from the posterior femoral condyles (4). When an anterior reference system is used, the surgeon can also decide to use a slightly larger femoral component with a slightly increased tibial resection to equalise the gaps (5).

In the knee with a recurvatum deformity, the extension space is relatively too large. In this situation, distalisation of the femoral component by removing 2 mm less distal femoral bone, will decrease only the extension space without altering the flexion space (1). In case of anterior referencing, the use of a slightly undersized femoral component will further equalise the gaps (2). Just using a thicker tibial insert to fill up the extension space, while increasing the flexion space by resecting the PCL or increasing the tibial slope, may be another option in the modest recurvatum knee (3).

Many surgeons consider revision total knee arthroplasty (TKA) a difficult procedure, calling for flexibility and improvisation.

However, revision TKA can be broken into a number of consecutive steps that need to be performed. Setting up a reproducible and stepwise approach is mandatory for the surgeon who performs this procedure more or less regularly.

At our institution, we have followed a five-step protocol in performing 166 revision TKA procedures. Its relatively strict guidelines leave little room for intraoperative improvisation. Our protocol covers exposition, implant extraction, implant selection, bone preparation and dealing with bony defects.

There has been acceptable ‘on the table’ reconstruction in all cases.

Although most surgeons agree that the functional results obtained with modern total knee arthroplasty (TKA) are acceptable, it is clear that even with the most recent designs it is still impossible to duplicate the behaviour and functional performance of a normal knee.

Recent kinematic studies have shown that modern TKA designs consistently provoke aberrant kinematics, mainly owing to the absence of the anterior cruciate ligament and the inability to maintain a functional posterior cruciate ligament (PCL). With regard to roll-back, PS cam-post designs appear to perform better than PCL retaining knees, but only in deeper degrees of flexion, usually only beyond 90°.

Whether it is strictly necessary to try to obtain normal kinematics remains an open debate. Clearly, aberrant kinematics are the direct cause of the flexion limitation we see in many of our patients. Further, they probably contribute to many of the discomforts associated with modern TKA, such as difficulties descending stairs, rising from chairs, pivoting and thrusting. Improvements in current TKA designs should aim at introducing the concept of guided-motion (intrinsic mechanism) and at maintaining or restoring (extrinsic) determinants of kinematics, i.e. the cruciate ligaments, the joint configuration and the extra-articular structures.