Navigated total knee arthroplasty: is it error-free?

Simultaneous Evaluation of Bone Cut and Implant Placement Accuracy in Robotic-Assisted Total Knee Arthroplasty

Background: This study aimed to evaluate the accuracy of bone cuts and implant placements, simultaneously, for total knee arthroplasty (TKA) performed using a system with an active robotic arm. Methods: Two experienced orthopaedic surgeons performed TKA on ten cadaveric legs. Computed tomography scans were performed to compare the bone cuts and implant placements with the preoperative planning. The differences between the planned and actual bone cuts and implant placements were assessed using positional and angular errors in the three anatomical planes. Additionally, the cut-implant deviations were calculated. Statistical analysis was performed to detect systematic errors in the bone cuts and implant placements and to quantify the correlations between these errors. Results: The root-mean-square (RMS) errors of the bone cuts (with respect to the planning) were between 0.7-1.5 mm and 0.6-1.7°. The RMS implant placement errors (with respect to the planning) varied between 0.6-1.6 mm and 0.4-1.5°, except for the femur and tibia in the sagittal plane (2.9°). Systematic errors in the bone cuts and implant placements were observed, respectively, in three and two degrees of freedom. For cut-implant deviations, the RMS values ranged between 0.3-2.0 mm and 0.6-1.9°. The bone cut and implant placement errors were significantly correlated in eight degrees-of-freedom (ρ ≥ 0.67, p < 0.05). Conclusions: With most of the errors below 2 mm or 2°, this study supported the value of active robotic TKA in achieving accurate bone cuts and implant placements. The findings also highlighted the need for both accurate bone cuts and proper implantation technique to achieve accurate implant placements.

Notching is less, if femoral component sagittal positioning is planned perpendicular to distal femur anterior cortex axis, in navigated TKA

Purpose

In navigated TKA, the risk of notching is high if femoral component sagittal positioning is planned perpendicular to the sagittal mechanical axis of femur (SMX). We intended to determine if, by opting to place the femoral component perpendicular to distal femur anterior cortex axis (DCX), notching can be reduced in navigated TKA.

Methods

We studied 171 patients who underwent simultaneous bilateral computer-assisted TKA. Femoral component sagittal positioning was planned perpendicular to SMX in one knee (Femur Anterior Bowing Registration Disabled, i.e. FBRD group) and perpendicular to DCX in the opposite knee (Femur Anterior Bowing Registration Enabled, i.e. FBRE group). Incidence and depth of notching were recorded in both groups. For FBRE knees, distal anterior cortex angle (DCA), which is the angle between SMX and DCX, was calculated by the computer.

Results

Incidence and mean depth of notching was less (p = 0.0007 and 0.009) in FBRE versus FBRD group, i.e. 7% versus 19.9% and 0.98 mm versus 1.53 mm, respectively. Notching was very high (61.8%) in FBRD limbs when the anterior bowing was severe (DCA > 3°) in the contralateral (FBRE) limbs.

Conclusion

Notching was less when femoral component sagittal positioning was planned perpendicular to DCX, in navigated TKA.

Level of evidence

Therapeutic level II.

The use of imageless navigation to quantify cutting error in total knee arthroplasty

PURPOSE

Navigated total knee arthroplasty (TKA) improves implant alignment by providing feedback on resection parameters based on femoral and tibial cutting guide positions. However, saw blade thickness, deflection, and cutting guide motion may lead to final bone cuts differing from planned resections, potentially contributing to suboptimal component alignment. We used an imageless navigation device to intraoperatively quantify the magnitude of error between planned and actual resections, hypothesizing final bone cuts will differ from planned alignment.

MATERIALS AND METHODS

A retrospective study including 60 consecutive patients undergoing primary TKA using a novel imageless navigation device was conducted. Device measurements of resection parameters were obtained via attachment of optical trackers to femoral and tibial cutting guides prior to resection. Following resection, optical trackers were placed directly on the bone cut surface and measurements were recorded. Cutting guide and bone resection measurements of both femoral and tibial varus/valgus, femoral flexion, tibial slope angles, and both femoral and tibial medial and lateral resection depths were compared using a Student's t-test.

RESULTS

Femoral cutting guide position differed from the actual cut by an average 0.6 ± 0.5° (p = 0.85) in the varus/valgus angle and 1.0 ± 1.0° (p = 0.003) in the flexion/extension angle. The difference between planned and actual cut measurements for medial and lateral femoral resection depth was 1.1 ± 1.1 mm (p = 0.32) and 1.2 ± 1.0 mm (p = 0.067), respectively. Planned cut measurements based on tibial guide position differed from the actual cut by an average of 0.9 ± 0.8° (p = 0.63) in the varus/valgus angle and 1.1 ± 1.0° (p = 0.95) in slope angle. Measurement of medial and lateral tibial resection depth differed by an average of 0.1 ± 1.8 mm (p = 0.78) and 0.2 ± 2.1 mm (p = 0.85), respectively.

CONCLUSIONS

Significant discrepancies between planned and actual femoral bone resection were demonstrated for flexion/extension angle, likely the result of cutting error. Our data highlights the importance of cut verification postresection to confirm planned resections are achieved, and suggests imageless navigation may be a source of feedback that would allow surgeons to intraoperatively adjust resections to achieve optimal implant alignment.

Valgus position of the femoral component causes abnormal kinematics in the presence of medial looseness in total knee arthroplasty: a computer simulation model of TKA for valgus knee osteoarthritis

PURPOSE

Total knee arthroplasty (TKA) for valgus knee osteoarthritis is challenging. Although overcorrection in TKA for valgus knee osteoarthritis is recommended, supportive data based on biomechanics have rarely been reported. The purpose of this study was to elucidate whether coronal rotation of the femoral compartment causes abnormal kinematics with or without medial looseness.

METHODS

Multi- and single-radius posterior-stabilised TKA implants were utilised in a computer simulation. A total of 4 mm of slack were provided in the medial collateral ligament (MCL) with varus or valgus position of the femoral component to simulate the context of valgus knee osteoarthritis. Kinematics during gait and squatting activities were evaluated in each condition.

RESULTS

During squatting, medial looseness and valgus replacement caused anterior translation of the medial femoral component in mid-flexion in the multi-radius implant. In the worst condition (7° valgus replacement with MCL looseness), there was rapid anterior translation in the multi-radius implant, and moderate anterior translation in the single-radius implant. Although medial looseness alone did not cause abnormal kinematics during gait, the worst condition exhibited an anterior translation to 4.9 mm in the multi-radius implant. This worst condition also exhibited a marked lift-off of 8.0 and 2.9 mm in the multi- and single-radius implants, respectively. Varus position caused little abnormal kinematics even with MCL looseness.

CONCLUSION

Valgus, not varus position of the femoral component caused abnormal kinematics with MCL looseness. To avoid valgus position, the safety target angle of femoral component would be slight varus rather than neutral in valgus knee OA.

Case-related factors affecting cutting errors of the proximal tibia in total knee arthroplasty assessed by computer navigation

PURPOSE

The aim of this study was to determine factors that contribute to bone cutting errors of conventional instrumentation for tibial resection in total knee arthroplasty (TKA) as assessed by an image-free navigation system. The hypothesis is that preoperative varus alignment is a significant contributory factor to tibial bone cutting errors.

METHODS

This was a prospective study of a consecutive series of 72 TKAs. The amount of the tibial first-cut errors with reference to the planned cutting plane in both coronal and sagittal planes was measured by an image-free computer navigation system. Multiple regression models were developed with the amount of tibial cutting error in the coronal and sagittal planes as dependent variables and sex, age, disease, height, body mass index, preoperative alignment, patellar height (Insall-Salvati ratio) and preoperative flexion angle as independent variables.

RESULTS

Multiple regression analysis showed that sex (male gender) (R = 0.25 p = 0.047) and preoperative varus alignment (R = 0.42, p = 0.001) were positively associated with varus tibial cutting errors in the coronal plane. In the sagittal plane, none of the independent variables was significant.

CONCLUSION

When performing TKA in varus deformity, careful confirmation of the bone cutting surface should be performed to avoid varus alignment. The results of this study suggest technical considerations that can help a surgeon achieve more accurate component placement.

LEVEL OF EVIDENCE

IV.

Does final component alignment correlate with alignment of the bone resection surfaces in cemented total knee arthroplasty?

PURPOSE

To examine, with a navigation, whether the final component alignments correlate with alignment of the bone resection surfaces in cemented total knee arthroplasty (TKA), and to evaluate the factors affecting alignment deviation.

METHODS

A total of 222 patients (276 knees) who underwent navigation-assisted TKA between September 2012 and January 2014 due to osteoarthritis were retrospectively reviewed. The deviation between the alignment of bone resection surfaces and the final alignment of femoral and tibial components was measured. Factors associated with alignment deviation of greater than 2° (outliers) were evaluated. These included age, sex, body mass index, bone mineral density (T score), preoperative and postoperative mechanical femorotibial angle, preoperative and postoperative flexion contractures, and the difference between medial and lateral gaps in knee extension or flexion.

RESULTS

Outliers consisted of 24 cases (8.6%) on the femoral coronal plane, 4 cases (1.4%) on the tibial coronal plane, and 48 cases (17.4%) on the tibial sagittal plane. In the coronal plane (femur and tibia), the outliers were associated with preoperative [p < 0.001; odds ratio (OR) 0.774; 95% confidence interval (CI) 0.672-0.891] and postoperative (p < 0.001; OR 0.240; 95% CI 0.123-0.468) flexion contractures; a difference of 3 mm or more between the medial and lateral gaps in knee extension (p < 0.041; OR 5.805; 95% CI 1.075-31.343); and a T score of less than -2.5(p < 0.024; OR 5.899; 95% CI 1.258-27.664). In the sagittal plane of the tibia, the outliers were associated with preoperative (p < 0.001; OR 0.886; 95% CI 0.829-0.946) and postoperative (p < 0.031; OR 0.803; 95% CI 0.659-0.980) flexion contractures.

CONCLUSION

There was a deviation between the alignments of the bone resection surfaces and the final alignments of components. With larger preoperative and postoperative flexion contractures in the coronal and sagittal planes, there were more outlier risks. The outliers in the coronal plane were associated with a difference of 3 mm or more between the medial and lateral gaps in knee extension and poor bone quality. Awareness of such alignment deviation and related factors can help diminish the outliers after TKA.

LEVEL OF EVIDENCE

IV.

Increased precision of coronal plane outcomes in robotic-assisted total knee arthroplasty: A systematic review and meta-analysis

BACKGROUND

Inaccuracy of component alignment in total knee arthroplasty adversely impacts outcomes. Robotic systems improve translation of pre-operative planning to intra-operative steps, theoretically resulting in greater accuracy and precision. In this study we systematically review literature data of alignment outcomes and apply meta-analysis methods to assess whether robotic-assisted knee arthroplasty provides superior outcomes when compared to conventional knee prostheses.

METHODS

A PRISMA compliant search comparing alignment outcomes in robotic vs conventional knee arthroplasty was performed. Primary outcome measures were; number of three degree outliers and mean deviation from a neutral post-operative mechanical axis.

RESULTS

In total, from five studies reporting upon 402 knees, a post-operative mechanical axis malalignment of >3° occurred in 1/181 (0.006%) of robotic knees, and 42/159 (26.4%) of conventional knees with a meta-analysis odds ratio of 0.04 (95% CI 0.01-0.14), p < 0.00001 favouring robotic-assisted instrumentation. Meta-analysis also demonstrated weighted mean differences of post-operative mechanical axis alignment to be significantly more accurate in the robotic knee group: mean difference -0.63 (95% CI: -1.18,-0.08), z = 2.25, p = 0.02. Sensitivity analysis with inclusion of only Level 1 studies showed similar findings.

CONCLUSIONS

This systematic review and meta-analysis demonstrates clear evidence of increased accuracy of alignment in robotic-assisted knee arthroplasty with specific regard to reconstituting a neutral mechanical axis and minimising number of outliers in the coronal plane. Further studies and long term data is required in order to conclude on survivorship and functional outcomes.

Articular surface mounted navigated total knee arthroplasty improves the reliability of component alignment

PURPOSE

The primary aim was to compare the early knee-specific functional outcome after articular surface mounted (ASM) navigation with non-navigated TKA. The secondary aims were to compare general physical and mental health improvement, patient satisfaction, and reliability of component alignment in the sagittal and coronal planes between ASM navigated TKA with that of non-navigated TKA.

METHODS

Prospective functional outcome and radiographic data were collect for 123 patients undergoing ASM navigation and 172 patients undergoing non-navigated TKA by a high volume single surgeon. Pre-operative and one-year Oxford knee score (OKS) and short form (SF-) 12 scores were collected. Patient satisfaction was also assessed at one year. Implant position was assessed on post-operative radiographs (alpha, beta, gamma, and sigma angles) by a blinded observer.

RESULTS

There was no significant difference for improvement in OKS, SF-12 physical or mental components, or satisfaction between the groups one year following surgery. The non-navigation group was significantly more likely to have outliers (greater than 3 degrees) in femoral varus/valgus coronal alignment [odds ratio (OR) 4.5, 95% confidence interval (CI) 1.0-20.7, p = 0.049] and for posterior tibial slope (OR 8.3, 95% CI 1.1-65.0, p = 0.03).

CONCLUSIONS

ASM navigation significantly reduces the number of outliers for the femoral and tibial components when compared to conventional non-navigation alignment. However, the short-term functional outcome is not influenced by the surgical technique used. If the surgeon wants to reduce their number of outliers, then ASM navigation should be considered but the overall functional outcome in the short term is not influenced.

LEVEL OF EVIDENCE

III Therapeutic investigation, retrospective cohort study.

A Comparison Between Chinese and Caucasian 3-Dimensional Bony Morphometry in Presimulated and Postsimulated Osteotomy for Total Knee Arthroplasty

BACKGROUND

The bone morphologies of intact knees were measured and compared between Chinese and Caucasian populations. However, to assess if distinct designs of implants are necessary for the Chinese population owing to different morphologies and sizes, the knee measurements after osteotomy performed in total knee arthroplasty were evaluated.

METHODS

Thirty-seven Caucasian and 50 Chinese patients' knees were examined using computed tomography scans. Mimics were applied to reconstruct 3-dimensional bone models. Dimensions of the 3-dimensional knee models and simulated bone resections during total knee arthroplasty were measured using Geomagic Studio and Pro/ENGINEER. The morphologic measurements of the native and resected femur and tibia included the anteroposterior (AP) depth, mediolateral (ML) width, notch width, knee physical valgus angle, tibial slope angle, and the ML-to-AP ratio of the femur, tibia, and resected femur. Statistical analysis was performed using the independent samples t test and the Pearson correlation coefficient in SPSS for Windows. Values of P < .05 were considered significant.

RESULTS

No measurements were significantly different between the Chinese and Caucasian knees. However, the Chinese female showed significant differences compared with the Chinese male on distal femoral measurements both presimulated and postsimulated osteotomy such as a smaller mean ML-to-AP ratio in presimulated (1.3 ± 0.1) and postsimulated (1.3 ± 0.1) osteotomy.

CONCLUSION

The necessity of designing a full set of total knee components specifically for the Chinese population is still undetermined. However, we suggest designing femoral components specific for the Chinese females because of different postosteotomy distal femoral ML-to-AP ratio between the Chinese males and the Chinese females.

Preserving the PCL during the tibial cut in total knee arthroplasty

PURPOSE

Previous studies have shown that the PCL insertion may be damaged during the tibial cut performed in total knee arthroplasty. We investigated the maximum thickness of a tibial cut that preserves the PCL insertion and to what extent the posterior slope of the tibial cut and that of the patient's tibial plateaus affect the outcome.

METHODS

MR images of 83 knees were analysed. The maximum thickness of a tibial cut that preserves the PCL using a posterior slope of 0°, 3°, 5° and parallel to the patient's slope of the tibial plateau, was evaluated. Correlations between the results and the degrees of the posterior slope of the patient's tibial plateaus were also investigated.

RESULTS

The maximum thickness of a tibial cut that preserves the entire PCL insertion was, on average, 5.5, 4.7, 4.2 and 3.1 mm when a posterior slope of 0°, 3°, 5° and parallel to the patients' tibial plateaus was used, respectively. When the 25th percentile was considered, the maximum thickness of a tibial cut that preserved the PCL was 4 and 3 mm with a tibial cut of 0° and 5° of posterior slope, respectively. The maximum thickness of a tibial cut that preserved the PCL was significantly greater in patients with a sagittal slope of the tibial plateaus more than 8° than in those with a sagittal slope less than 8°.

CONCLUSION

In cruciate retaining implants, the PCL insertion may be spared in the majority of patients by performing a tibial cut of 4 mm, or even less when a posterior slope of 3°-5° is used. The clinical relevance of our study is that the execution of a conservative tibial cut, followed by a second tibial resection to achieve the thickness required for the tibial component to be implanted, may be an alternative technique to spare the PCL in CR TKA.

LEVEL OF EVIDENCE

II.

A modified technique to reduce tibial keel cutting errors during an Oxford unicompartmental knee arthroplasty

PURPOSE

Bone cutting errors can cause malalignment of unicompartmental knee arthroplasties (UKA). Although the extent of tibial malalignment due to horizontal cutting errors has been well reported, there is a lack of studies evaluating malalignment as a consequence of keel cutting errors, particularly in the Oxford UKA. The purpose of this study was to examine keel cutting errors during Oxford UKA placement using a navigation system and to clarify whether two different tibial keel cutting techniques would have different error rates.

METHODS

The alignment of the tibial cut surface after a horizontal osteotomy and the surface of the tibial trial component was measured with a navigation system. Cutting error was defined as the angular difference between these measurements. The following two techniques were used: the standard "pushing" technique in 83 patients (group P) and a modified "dolphin" technique in 41 patients (group D).

RESULTS

In all 123 patients studied, the mean absolute keel cutting error was 1.7° and 1.4° in the coronal and sagittal planes, respectively. In group P, there were 22 outlier patients (27 %) in the coronal plane and 13 (16 %) in the sagittal plane. Group D had three outlier patients (8 %) in the coronal plane and none (0 %) in the sagittal plane. Significant differences were observed in the outlier ratio of these techniques in both the sagittal (P = 0.014) and coronal (P = 0.008) planes.

CONCLUSION

Our study demonstrated overall keel cutting errors of 1.7° in the coronal plane and 1.4° in the sagittal plane. The "dolphin" technique was found to significantly reduce keel cutting errors on the tibial side. This technique will be useful for accurate component positioning and therefore improve the longevity of Oxford UKAs.

LEVEL OF EVIDENCE

Retrospective comparative study, Level III.

Experimental testing of total knee replacements with UHMW-PE inserts: impact of severe wear test conditions

Aseptic implant loosening due to inflammatory reactions to wear debris is the main reason for the revision of total knee replacements (TKR). Hence, the decrease in polyethylene wear particle generation from the articulating surfaces is aimed at improving implant design and material. For preclinical testing of new TKR systems standardized wear tests are required. However, these wear tests do not reproduce the entire in vivo situation, since the pattern and amount of wear and subsequent implant failure are underestimated. Therefore, daily activity, kinematics, implant aging and position, third-body-wear and surface properties have to be considered to estimate the wear of implant components in vivo. Hence, severe test conditions are in demand for a better reproduction of the in vivo situation of TKR. In the present article an overview of different experimental wear test scenarios considering clinically relevant polyethylene wear situations using severe test conditions is presented.

The risk of sacrificing the PCL in cruciate retaining total knee arthroplasty and the relationship to the sagittal inclination of the tibial plateau

BACKGROUND

In cruciate retaining total knee arthroplasty (TKA), a partial avulsion of PCL may occur when en-bloc tibial osteotomy is performed. We evaluated the effects of a tibial cut performed with different degrees of posterior slope on PCL insertion and whether the results are affected by the sagittal inclination of the patient's tibial plateau.

METHODS

We selected 83 MRIs of knees showing mild or no degenerative changes. The effects of a simulated tibial cut performed with a posterior slope of 0°, 3°, 5° and parallel to the patient's tibial plateau inclination on PCL insertion in the proximal tibia were investigated. The results were correlated with the degree of posterior inclination of the tibial plateau.

RESULTS

Every angle we used for the tibial cut caused a PCL avulsion greater than 50%. The percentage of PCL avulsion significantly increased with increasing the posterior slope of the tibial cut. Patients with sagittal tibial plateau inclination <5° showed greater PCL avulsion than those with sagittal inclination >8°.

CONCLUSIONS

Most of the PCL insertion is likely to be sacrificed when resection of the proximal tibia is performed en-block. The risk of PCL avulsion is reduced in patients showing a marked posterior inclination of the tibial plateau, but even in this group of patients a surgical technique aimed at sparing most of the PCL insertion is necessary.