INTRODUCTION

To test whether there are differences in postoperative mechanical and component alignment, and in functional results, between conventional, navigated and patient-specific total knee arthroplasties in a low-volume centre?

Long-term implant survivorship in total knee arthroplasty (TKA) depends on the alignment of the tibial and femoral components, as well as on the mechanical alignment of the leg. Computer navigation improves component and limb alignment in TKA compared to the manual technique. However, its use is often associated with an increase in surgical time. We aimed to evaluate the use of adjustable cutting blocks (ACB) in navigated TKA. We hypothesised that the use of ACB would (1) improve tibial and femoral component positioning; (2) improve postoperative mechanical leg alignment; and (3) decrease tourniquet time, when compared to conventional cutting blocks (CCB).

This was a retrospective cohort study of 94 navigated primary TKA. Patients were classified into two groups according to whether the surgery had been performed using ACB or CCB. There were sixty-four patients in the CCB group and 30 patients in the ACB group. Charts were reviewed to obtain the following data: age, gender, body mass index (BMI), tourniquet time and operated side. Pre- and postoperative standing full-leg radiographs and lateral radiographs were reviewed.

Mean coronal femoral alignment for the CCB group was 0.8® varus (SD = 1.95®) and for the ACB group it was 1.1® varus (SD = 1.5®) (P = 0.12). Mean coronal tibial alignment for the CCB group was 0.1® valgus (SD = 1.3®) and for the ACB group it was 0.5® varus (SD = 1.01) (P = 0.15). Sagittal tibial alignment was a mean 0.5® of anterior slope (SD = 2.9®) for the CCB group and 0.7® anterior slope (SD = 2.5®) for the ACB group (P = 0.38).

Preoperatively, the CCB group had a mean mechanical alignment of 1.8® varus (SD = 9.6®), while the ACB group had a mean 1.8® varus (SD = 9.37®) (P = 0.88). After surgery, mechanical leg alignment for the CCB group improved to a mean 0.7® varus (SD = 2.7®) (P = 0.0001), while the ACB group improved to 1.8® varus (SD = 1.7®) (P<0.0001). There was significantly less variability in postoperative mechanical alignment in the ACB group (P = 0.0091).

Mean tourniquet time for the CCB group was 91 minutes (SD = 17.7 minutes). The ACB group a mean tourniquet time of 76 minutes (SD = 16.7 minutes) (P = 0.01). In the multiple linear regression model, the use of an ACB reduced tourniquet time by 16.8 minutes (P = 0.001).

Adjustable cutting blocks for TKA significantly reduced postoperative mechanical alignment variability and tourniquet time compared to conventional navigated instrumentation, while providing equal or better component alignment.

Controversies about the management of injuries to the soft tissue structures of the posteromedial corner of the knee and the contribution of such peripheral structures on rotational stability of the knee are of increasing interest and currently remain inadequately characterised. The posterior oblique ligament (POL) is a fibrous extension off the distal aspect of the semimembranosus that blends with and reinforces the posteromedial aspect of the joint capsule. The POL is reported to be a primary restraint to internal rotation and a secondary restraint to valgus translation and external rotation. Although its role as a static stabiliser to the medial knee has been previously described, the effect of the posterior oblique ligament (POL) injuries on tibiofemoral stability during Lachman and pivot shift examination in the setting of ACL injury is unknown.

The objective of this study was to quantify the magnitude of tibiofemoral translation during the Lachman and pivot shift tests after serial sectioning of the ACL and POL.

Eight knees were used for this study. Ligamentous constraints were sequentially sectioned in the following order: ACL first, followed by the POL. Navigated mechanised pivot shift and Lachman examinations were performed before and after each structure was sectioned, and tibiofemoral translation was recorded.

Lachman test: There was a mean 6.0 mm of lateral compartment translation in the intact knee (SD = 3.3 mm). After sectioning the ACL, translation increased to 13.8 mm (SD = 4.6; P<0.05). There was a nonsignificant 0.7 mm increase in translation after sectioning the POL (mean = 14.5 mm; SD = 3.9 P>0.05).

Mechanised pivot shift: Mean lateral compartment translation in the intact knee was −1.2 mm (SD = 3.2 mm). Sectioning the ACL caused an increase in anterior tibial translation (mean = 6.7 mm; SD = 3.0 mm; P<0.05). No significant change in translation was seen after sectioning the POL (mean = 7.0 mm, SD = 4.0 mm; P>0.05).

Sectioning the POL did not significantly alter tibiofemoral translation in the ACL deficient knee during the Lachman and pivot shift tests. This study brings into question whether injuries to the POL require reconstruction in conjunction with ACL reconstruction. More studies are needed to further characterise the role of the injured POL in knee stability and its clinical relevance in the ACL deficient and reconstructed knee.

Injuries to the posterior cruciate ligament (PCL) and the posterolateral corner (PLC) of the knee remain a challenging orthopaedic problem. Studies evaluating PCL and PLC reconstruction have failed to demonstrate a strong correlation between the degree of knee laxity as measured by uniplanar testing and subjective outcome or patient satisfaction. The effect that changing the magnitude of posterior tibial slope has on multiplanar, rotational stability of the PCL-deficient knee has yet to be determined. We aimed to evaluate the effect that changes in posterior tibial slope would have on static and dynamic stability of the PCL-PLC deficient knee.

Ten knees were used for this study. Navigated posterior drawer and standardised reverse mechanised pivot shift maneuvers were performed in the intact knee and after sectioning the PCL, the lateral collateral ligament (LCL), the popliteofibular ligament (PFL) and the popliteus muscle tendon (POP). Navigated high tibial osteotomy (HTO) was performed to obtain the desired change in tibial plateau slope (+5® or −5® from native slope). We then repeated the posterior drawer and the reverse mechanised pivot shift test for each of the two altered slope conditions.

Mean posterior tibial translation during the posterior drawer in the intact knee was 1.4 mm (SD = 0.48 mm). In the PCL-PLC deficient knee, posterior tibial translation increased to 18 mm (SD = 5.7 mm) (P < 0.001). Increasing the amount of posterior tibial slope by 5® reduced posterior tibial translation to 12 mm (SD = 4.7 mm) (P < 0.01). Decreasing the amount of posterior slope by 5® compared to the native knee, increased posterior tibial translation to 21 mm (SD = 6.8 mm) (P < 0.01). There was a significant negative correlation between the magnitude of tibial plateau slope and the magnitude of the reverse pivot shift (R2 = 0.71; P < 0.0001).

Mean posterior tibial translation during the reverse mechanised pivot shift test in the intact knee was 7.8 mm (SD = 2.8 mm). In the PCL-PLC deficient knee, posterior tibial translation increased to 26 mm (SD = 5.6 mm) (P < 0.001). Increasing the amount of posterior tibial slope by 5® reduced posterior tibial translation to 21 mm (SD = 6.7 mm) (P < 0.01). Decreasing the amount of posterior slope by 5® compared to the native knee, increased posterior tibial translation to 34 mm (SD = 8.2 mm) (P < 0.01). There was a significant negative correlation between the magnitude of tibial plateau slope and the magnitude of the reverse pivot shift (R2 = 0.72; P < 0.0001).

Decreasing the magnitude of posterior slope of the tibial plateau resulted in an increase in the magnitude of posterior tibial translation during the posterior drawer and the reverse mechanised pivot shift test in the PCL-PLC deficient knee. Conversely, increasing the slope of the tibial plateau reduced the amount of posterior tibial translation during the posterior drawer and the reverse mechanised pivot shift test. However, the effect of the increase in slope was not sufficient to reduce posterior tibial translation to levels similar to those of the intact knee.

Accurate and reliable registration of the ankle center is a necessary requirement in computer-assisted TKR. There is debate among surgeons over which registration procedure more accurately reflects the true center of the ankle joint. The aim of this study was to compare two different ankle registration landmarks on radiographs and determine how much they differed from the anatomic center of the talus in the frontal plane. Specifically, we asked what the average deviation in tibial mechanical axis registration would be when registering the ankle center using: A) the extreme medial and lateral points; and B) the most distal points, of the respective malleoli. A second question was whether or not BMI had any significant effect on mechanical axis registration error.

We reviewed the preoperative hip-to-ankle radiographs of 40 patients who underwent navigated TKR at our institution. The patient cohort was composed of 32 females and 7 males, with a mean age of 69 years (range, 45–84 years) and a mean BMI of 29.9 (range, 14.7–43.3). All radiographs were stored in and reviewed using PACS.

No clinically significant divergence from the anatomic center of the ankle was seen when using the Extremes Midpoint technique (mean divergence = 0.2® lateral; SD = 0.5®; 95% CI = −0.3®, −0.1®) or the Distal Midpoint technique (mean divergence = 0.2® lateral; SD = 0.6®; 95% CI = −0.39®, 0®). The mean difference between both techniques was 0.02® (SD = 0.3®; 95% CI = −0.1®, 0.1®; P = 0.68). BMI had no significant effect on the divergence from the true ankle center for either the Extremes Midpoint (R² = 0.002; P = 0.78) or the Distal Midpoint techniques (R² = 0.004; P = 0.90).(Figure 2)

The center of the ankle, as determined by using the Extremes Midpoint technique, lied 1.1 mm (SD = 2.6 mm; 95% CI = −1.9 mm to −0.3 mm) from the anatomic axis of the tibia. When determined using the Distal Midpoint technique, the center of the ankle lied 1.7 mm (SD = 2.3 mm; 95% CI = −2.5 mm to −0.98 mm) from the anatomic axis. Although statistically significant (P = 0.028), this difference was not clinically relevant (<3 mm). BMI had no significant effect on these differences (R² = 0.07; P = 0.11; R² = 0.02, P = 0.38).(Figure 3)

There is no significant difference between ankle registration using the Extremes Midpoint or the Distal Midpoint techniques and the anatomic center of the ankle. Patients' BMI does not seem to affect the registration of the ankle center with either technique.

Over the last two decades, anatomic anterior cruciate ligament (ACL) reconstructions have gained popularity, while the use of extraarticular reconstructions has decreased. However, the biomechanical rationale behind the lateral extraarticular sling has not been adequately studied. By understanding its effect on knee stability, it may be possible to identify specific situations in which lateral extraarticular tenodesis may be advantageous. The primary objective of this study was to quantify the ability of a lateral extraarticular sling to restore native kinematics to the ACL deficient knee, with and without combined intraarticular anatomic ACL reconstruction. Additionally, we aimed to characterise the isometry of four possible femoral tunnel positions for the lateral extraarticular sling.

Eight fresh frozen hip-to-toe cadavers were used in this study. Navigated Lachman and mechanised pivot shift examinations were performed on ACL itact and deficient knees. Three reconstruction strategies were evaluated: Single bundle anatomic intraarticular ACL reconstruction, Lateral extraarticular sling, Combined intraarticular ACL reconstruction and lateral extraarticular sling. After all stability tests were completed, we quantified the isometry of four possible femoral tunnel positions for the lateral extraarticular sling using the Surgetics navigation system. A single tibial tunnel position was identified and digitised over Gerdy's tubercle. Four possible graft positions were identified on the lateral femoral condyle: the top of the lateral collateral ligament (LCL); the top of the septum; the ideal tunnel position, as defined by the navigation system's own algorithm; and the actual tunnel position used during testing, described in the literature as the intersection of the linear projections of the LCL and the septum over the lateral femoral condyle. For each of the four tunnel positions, the knee was cycled from 0 to 90® of flexion and fiber length was recorded at 30® intervals, therefore quantiying the magnitude of anisometry for each tunnel position.

Stability testing: Sectioning of the ACL resulted in an increase in Lachman (15mm, p = 0.01) and mechanised pivot shift examination (6.75mm, p = 0.04) in all specimens compared with the intact knee. Anatomic intraarticular ACL reconstruction restored the Lachman (6.7mm, p = 3.76) and pivot shift (−3.5mm, p = 0.85) to the intact state. With lateral extraarticular sling alone, there was a trend towards increased anterior translation with the Lachman test (9.2mm, p = 0.50). This reconstruction restored the pivot shift to the intact state. (1.25mm, p = 0.73). Combined intraarticular and extraarticular reconstruction restored the Lachman (6.2mm, p = 2.11) and pivot shift (−3.75mm, p = 0.41) to the intact state. There was no significant difference between intraarticular alone and combined intraarticular and extraarticular reconstruction. (p = 1.88)

Isometry: The ideal tunnel position calculated by the navigation system was identified over the lateral femoral condyle, beneath the mid-portion of the LCL. The anisometry for the ideal tunnel position was significantly lower (5.9mm; SD = 1.8mm; P<0.05) than the anisometry of the actual graft position (14.9mm; SD = 4mm), the top of the LCL (13.9mm; SD = 4.3mm) and the top of the septum (12mm; SD = 2.4mm).

In the isolated acute ACL deficient knee, the addition of a lateral extraarticular sling to anatomic intraarticular ACL reconstruction provides little biomechanical advantage and is not routinely recommended. Isolated lateral extraarticular sling does control the pivot shift, and may be an option in the revision setting or in the lower demand patient with functional instability. Additionally, the location of the femoral tunnel traditionally used results in a significantly more anisometric graft than the navigation's system mathematical ideal location. However, the location of this ideal tunnel placement lies beneath mid-portion of the fibers of the LCL, which would not be clinically feasible.

Controversies about the management of injuries to the soft tissue structures of the posteromedial corner of the knee and the contribution of such peripheral structures on rotational stability of the knee are of increasing interest and currently remain inadequately characterised. The posterior oblique ligament (POL) is a fibrous extension off the distal aspect of the semimembranosus that blends with and reinforces the posteromedial aspect of the joint capsule. The POL is reported to be a primary restraint to internal rotation and a secondary restraint to valgus translation and external rotation. Although its role as a static stabiliser to the medial knee has been previously described, the effect of the posterior oblique ligament (POL) injuries on tibiofemoral stability during Lachman and pivot shift examination in the setting of ACL injury is unknown.

The objective of this study was to quantify the magnitude of tibiofemoral translation during the Lachman and pivot shift tests after serial sectioning of the ACL and POL.

Eight knees were used for this study. Ligamentous constraints were sequentially sectioned in the following order: ACL first, followed by the POL. Navigated mechanised pivot shift and Lachman examinations were performed before and after each structure was sectioned, and tibiofemoral translation was recorded.

Lachman test: There was a mean 6.0 mm of lateral compartment translation in the intact knee (SD = 3.3 mm). After sectioning the ACL, translation increased to 13.8 mm (SD = 4.6; P<0.05). There was a nonsignificant 0.7 mm increase in translation after sectioning the POL (mean = 14.5 mm; SD = 3.9 P>0.05).

Mechanised pivot shift: Mean lateral compartment translation in the intact knee was −1.2 mm (SD = 3.2 mm). Sectioning the ACL caused an increase in anterior tibial translation (mean = 6.7 mm; SD = 3.0 mm; P<0.05). No significant change in translation was seen after sectioning the POL (mean = 7.0 mm, SD = 4.0 mm; P>0.05).

Sectioning the POL did not significantly alter tibiofemoral translation in the ACL deficient knee during the Lachman and pivot shift tests. This study brings into question whether injuries to the POL require reconstruction in conjunction with ACL reconstruction. More studies are needed to further characterise the role of the injured POL in knee stability and its clinical relevance in the ACL deficient and reconstructed knee.

Robotic-guided arthroplasty procedures are becoming increasingly common. We introduced a new computer-navigated TKA system with a robotic cutting-guide into a community-based hospital and characterized the accuracy and efficiency of the technique.

We retrospectively reviewed our first 100 cases following IRB approval. Tourniquet time, intraoperative bone-cut accuracy and final limb alignment as measured by the computer were collected and divided into consecutive quartiles: Groups I, II, III, and IV; 25 cases per group. All resections were planned neutral to the mechanical axis. Postoperative component alignment and overall mechanical axis limb alignment were also measured on standing long-leg radiographs by two independent observers at minimum six weeks follow-up. Radiographic alignment was available for 62 cases.

Intraoperative Computer Data: Bone-cut accuracy was a mean 0.1° valgus, SD±0.8° for both the femur and tibia (range, femur: 2.0° valgus to 1.5° varus; range, tibia: 3.5° valgus to 1.5° varus). Final limb alignment was within 3° for 98% (97/99) of cases (range: 2.0° valgus to 3.5° varus). Radiographic Alignment: Pre-operative mechanical alignment ranged from −14.5° valgus to 21.5° varus. Radiographic femoral and tibial component alignment was within 3° of neutral in 98.4% of cases (61/62). Final limb alignment was within 3° for 87.1% (54/62) of cases (range: 4.5° varus to 4.5° valgus). Learning curve: Mean tourniquet time was 60minutes ±9.9SD (range 46–79) for Group I and 49.5minutes for Groups II, III, and IV (range 35–68), p = 0.0001. Mean tourniquet time for the first ten and second ten procedures was 65±10.6minutes and 55±8.3minutes, respectively, p = 0.034. There were no differences in accuracy among the four groups (p>0.05).

Imageless computer-navigated TKA with a robotic cutting guide allowed one surgeon to make bone resections within 3° of neutral in 98% of cases. Radiographic limb alignment was less precise, which is consistent with the known limitations inherent to this measurement technique. Surgeons can expect this procedure to take 15 additional minutes during the first ten cases and five additional minutes during the second ten cases on average, without compromising accuracy.

Unicompartmental knee replacement (UKR) has good outcomes for the treatment of compartmental osteoarthritis of the knee. Mechanical alignment overcorrection is associated with early failure of the femoral and tibial components. Preoperative mechanical alignment is the most important predictor of postoperative alignment. However, most studies do not take into consideration the magnitude of preoperative deformity when reporting on mechanical alignment outcomes after UKR.

We aimed to determine the magnitude of postoperative mechanical alignment achieved based on the magnitude of preoperative alignment; and to compare the number of cases of overcorrection into valgus to historical data.

This was a radiographic review of patients who underwent robotic medial UKR by a single surgeon between 2007 and 2011. Two examiners measured pre- and postoperative mechanical alignment for all patients on long-leg radiographs. Patients were classified into three groups of preoperative mechanical alignment: mild varus (0–5®); moderate varus (5–10®); and severe varus (>10®). Patients with valgus alignment (<0®) were excluded. Linear regression was used to estimate the magnitude of postoperative alignment for each group, adjusting for age, BMI, gender, side, implant type, and polyethylene thickness.

89 patients were included. Mean preoperative alignment was 7.3® varus (95% CI = 6.6®–8®; range, 0.1–15® varus). Mean postoperative alignment was 2.8® varus (95% CI = 1.9®–3.8®; range, 1.4® valgus–9.7® varus). There was a significant difference in postoperative mechanical alignment between the three groups (Table 1) (P<0.05). Four overcorrections (4.5%) were detected, all under 1.5® valgus. This percentage of overcorrection was significantly better than previous conventional UKR reports (mean = 12.6%; P = 0.04).

The magnitude of postoperative alignment in medial UKR depends on the severity of the preoperative deformity. Reports on radiographic outcomes of UKR should be stratified by the magnitude of preoperative alignment. The risk of overcorrection is reduced when using robotic assistance compared to using the conventional manual technique.

Introduction

Robotic-guided arthroplasty procedures are becoming increasingly common, though to our knowledge there are no published studies on robotic cutting guides in TKA. We introduced a new computer-navigated TKA system with a robotic cutting-guide into a community-based hospital and characterized the accuracy and efficiency of the technique with respect to bone cutting, component alignment and final limb alignment, and tourniquet time.