If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Whole Leg Radiographs (WLR) are the gold standard for diagnosing malalignment and for pre-operative osteotomy planning. Positioning can affect the reproducibility of the measured hip knee angle (HKA), resulting in insufficient diagnostics and preoperative plans. We developed an easy-to-use WLR protocol by standardizing patient positioning and focusing on reproducibility.
Objectives
This study aims on testing this reproducibility of the novel WLR protocol.
Methods
This study enrolled 30 patients for a test-retest analysis. Each patient underwent two bilateral WLRs on the same day using the investigated positioning protocol. Three observers measured the HKA, mechanical medial proximal tibial angle (mMPTA), mechanical lateral distal femoral angle (mLDFA), and joint line convergence angle (JLCA) on the two radiographs. Twice each, with one week between.
Results
The intra-observer and inter-observer reliabilities were excellent, with intraclass correlation coefficients (ICCs) between 0.990 and 0.996. The ICCs between the measured HKA (0.985), mMPTA (0.922), and mLDFA (0.903) on the two separate radiographs were excellent. The ICC between the JLCA measured on the first and second WLR was moderate with 0.632. The mean absolute error between the HKA, mMPTA, mLDFA, and JLCA measurements on the first and second WLR were respectively: 0.442°, 0.783°, 0.828°, and 0.794°.
Conclusions
The investigated novel WLR positioning protocol produced excellent and reproducible HKA measurements, with clinically acceptable degrees of error. We recommend applying this easy-to-use protocol when obtaining WLRs for osteotomy planning. Physicians still need to be aware of possible rotational and fixed flexion deformities in the lower limb present on WLRs and during physical examination.
Resultant pain and loss of function can be debilitating, with high associated socioeconomic burden, estimated to cost between 1.0% and 2.5% of the gross domestic product.
An important aetiology of knee OA is underlying varus or valgus malalignment, which can cause progressive unicompartmental knee pathology and degeneration.
Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging.
A corrective osteotomy is a joint preserving technique that aims to restore the native mechanical axis and has been observed to potentially postpone joint replacement surgery by up to 10 years.
By proceeding with corrective osteotomy prior to generalized degenerative changes, the age of patients at the time of primary total knee arthroplasty potentially substantially increases, resulting in decreased likelihood of revision surgery later in life.
Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging.
Accordingly, it is important that the radiographic technique when obtaining WLRs is consistent and accurate. Jones et al. reported a desired accuracy of 0.45° in order to achieve satisfactory intraoperative wedge accuracy for target corrections in high tibial osteotomy (HTO).
The most commonly used WLR protocol was published by Paley and Herzenberg, where the patella is used as a key landmark for antero-posterior orientation of the lower limb.
This is of clinical importance given that the reproducibility of the measured hip knee angle (HKA) on a WLR is influenced by positional differences of patients, like foot positioning, knee flexion, leg rotation, and weight bearing.
This results in sub-optimal preoperative planning and assessment of achieved surgical corrections. This can cause substantial under- and overcorrections, which could lead to inferior HTO outcomes with reported poor surgical accuracies.
Advances in modern osteotomies around the knee: report on the Association of Sports Traumatology, Arthroscopy, Orthopaedic surgery, Rehabilitation (ASTAOR) Moscow International Osteotomy Congress 2017.
Odenbring et al. performed a test-retest study including only 8 patients using a meticulous WLR protocol with a mean error of 1.3°, which improved accuracy but fell outside clinical feasibility for implementation.
Recently, our team developed a novel and easy to use WLR positioning protocol and subsequently introduced this into clinical care, with the goal of standardizing patient positioning and reproducibility. The aim of the current study was to test the reproducibility of the implemented WLR protocol using a test-retest principle. We hypothesized that WLRs would be reproducible and within the desired published osteotomy accuracy of 0.45° with this novel protocol.
Methods
Patient population
This prospective study was approved by the ethics committee of University Medical Centre (UMC) Utrecht (METC number 19–474) on 27 November 2019. Patients were recruited at the UMC Utrecht Mobility Clinic, which is a tertiary orthopaedic referral centre for knee joint preserving treatments, including cartilage repair techniques, osteotomies, and knee joint distraction.
Efficacy of one-stage cartilage repair using allogeneic mesenchymal stromal cells and autologous chondron transplantation (IMPACT) compared to nonsurgical treatment for focal articular cartilage lesions of the knee: study protocol for a crossover randomize.
Better clinical results after closed- compared to open-wedge high tibial osteotomy in patients with medial knee osteoarthritis and varus leg alignment.
Patients were eligible when they had an appointment at the outpatient clinic with scheduled WLR ordered. Exclusion criteria were: age under 18 years, inability to read, communicate, and/or speak the Dutch language, pregnancy, and patients incapable of providing informed consent. When willing to participate, patients signed an informed consent. Included patients underwent a second WLR after their appointment at the outpatient clinic, half an hour or more following their initial radiographic imaging (including WLRs). Both WLR acquisitions followed the same novel, standardized positioning protocol. This resulted in two separate episodes of patient positioning and radiograph acquisition within the same day separated by at least 30 min.
Positioning protocol
Patients were positioned with their knees in full extension. Feet were positioned with a distance of 10 cm between the heels and aligned in 10° of external rotation from the midline. This was achieved by placing the feet on a standardized positioning template (Fig. 1). X-ray technicians subsequently adjusted hip rotation, by aligning the upper body and pelvis in a straightforward (AP) position. No handlebars or supports were employed and patients were asked to place their hands alongside their body to ensure full weight-bearing. The X-ray technicians additionally instructed patients to distribute their weight equally to each leg. All WLRs were performed with the bilateral lower extremities captured on the radiographs, with a radiopaque measurement tape positioned behind the patients for subsequent image calibration. Figure 2 illustrates a patient undergoing WLR with our standardized positioning protocol.
Fig. 1Template for feet positioning during a WLR, as part of the WLR positioning protocol. Feet are pointed outwards in 10° between the midlines and placed 10 from each other from the center of the heels. This template is engraved onto a durable Trespa board (right picture). WLR, whole leg radiographs.
Each WLR was obtained using the Philips DigitalDiagnost v4.0 (Philips, Amsterdam, The Netherlands). The fixed distance between the detector plate and X-ray beam source was set to 265 cm. The X-ray beam source was fixed in height, and obtained 3 different bilateral images (proximal femur, knee joint, distal tibia) by pivoting the source towards upper, middle, and lower part of the legs. The 3 separate images were then stitched together into 1 WLR by Philips DigitalDiagnost v4.0 software. The X-ray settings were kept the same for each patient, with kV set at 81 and varying mAs.
Radiographic measurements
Radiographic measurements of the HKA were done for each leg separately by three observers in PACS IDS7 19.3 (Sectra AB, Linköping, Sweden), with the OrthoStation module. Two orthopaedic surgeons (RC&NE) and a researcher (CN) performed the measurements twice, with one week in between measurements by the same observer. The two radiographs of each patient were analysed separately and in a random order.
HKA was determined as the angle between the line from the centre of the femoral head to the centre of the femoral notch, and the line from the centre of the tibial spines to the centre of the talus (Fig. 3). The HKA was measured with 1 decimal place accuracy.
Fig. 3Leg geometry measurements on a WLR in PACS IDS7 19.3 (Sectra AB, Linköping, Sweden). Left image illustrates the HKA measurement, the right image illustrates the mMPTA, mLDFA, and the JLCA. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
OrthoStation was employed to provide a semi-automated method to determine the joint line convergence angle (JLCA), mechanical medial proximal tibial angle (mMPTA), and mechanical lateral distal femoral angle (mLDFA). The tool required input from the observers, selecting the centre of the femoral head and talus, followed by marking the joint lines of the tibia and femur (Fig. 3). The mMPTA, mLDFA, and JLCA were provided in whole numbers.
Statistical analysis
Intra-observer reliability was tested using a 2 way mixed Intraclass Correlation (ICC) for absolute agreement. The inter-observer reliabilities were tested using a 2 way random ICC for absolute agreement. Test-retest agreement was calculated using a 2 way random absolute agreement ICC and Bland-Altman analyses. The errors between the measured parameters on the 2 WLRs and different observers were reported as mean (95%-CI interval). Due to a non-normal right skewed distribution of the absolute errors, these values were reported as mean (bootstrapped resampled 95%-CI intervals). Wilcoxon signed-rank tests were performed for differences in absolute test-retest errors between sub-groups included in this study. Statistical significance was set at alpha = 0.05. All statistical calculations were performed in SPSS Statistics (IBM, version 25.0.0.2.).
Power calculations for reproducibility studies are not straightforward and ill-studied.
Following the guideline as proposed by Bujanga and Baharum, we need only 14 legs to find a significant difference between an ICC of 0.7 and 0.9. We however deemed this number as very small and aimed to reach a narrower confidence interval.
Following the advice of Cicchetti, we used three highly skilled readers to assess 60 legs in 30 patients, thus allowing us to include a range of alignments, male and females.
For this study, 31 patients signed an informed consent. We had to exclude one patient due to a no show for the retest radiograph. Therefore, 30 patients with 30 bilateral WLRs taken on 2 separate time-points (60 measured WLRs, 120 legs in total) were included in this study. The study included fifteen males and fifteen females, with a median age of 34.5 year (range: 18–61), and mean BMI of 25.8 (SD 3.2). Mean HKA was 179.3°, mean mMPTA was 87.1°, mean mLDFA was 87.3°, and mean JLCA was 1.3° (Fig. 4).
Fig. 4Histograms of the population mean measured HKA, mMPTA, mLDFA, and JLCA. Reported normal distributions according to Bellemans et al. are: HKA = 178.67 ± 2.34, mMPTA = 87.04 ± 2.07, mLDFA = 87.90 ± 1.74, and JLCA = 0.51 ± 1.0538. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
Test results of the intra-observer and inter-observer reliabilities are listed in Table 1. Intra-observer and inter-observer reliabilities of the HKA, mMPTA, and mLDFA were excellent (ICC ≥ 0.850). Intra-observer and inter-observer reliabilities of the JLCA were fair to good (ICC ≥ 0.463). The mean absolute differences between the 3 observers of the measured HKA, mMPTA, mLDFA, and JLCA were respectively: 0.491° (CI 0.430° – 0.552°), 0.889° (CI 0.761° – 1.013°), 0.922° (CI 0.806° – 1.056), and 0.931° (CI 0.805° – 1.069°).
Table 1Calculated Intraclass Correlation (ICC) and the 95% Confidence Interval (CI).
All results of the test-retest analyses are listed in Table 2, including all lower limbs, sub-sections based on osteotomy surgery, osteoarthritis, and (osteo-)chondral lesion. When including all lower limbs, the test-retest ICCs between the measured HKA, mMPTA, and mLDFA on the first and second radiograph were excellent. The ICC between the measured JLCA on the first and second WLR was moderate. The mean absolute test-retest errors were for the measured HKA 0.442° (CI 0.386°−0.501°), mMPTA 0.783° (CI 0.683°−0.894°), mLDFA 0.828° (CI 0.711°−0.933°), and JLCA 0.794° (CI 0.689°−0.900°). Wilcoxon signed-rank tests revealed no significant differences between all lower limbs and males and/or females, osteoarthritis, (osteo-)chondral lesions, and osteotomy surgery in terms of HKA, mLDFA, mMPTA, and JLCA.
Table 2Calculated Test-Retest errors in mean absolute degrees and 95% Confidence Interval (CI) for the HKA, mMPTA, mLDFA, and JLCA. Intraclass Correlations between the measurements performed on the test and retest radiographs, with significant (*) P-values below .05.
Fig. 5 illustrates the Bland-Altman test-retest analyses of the HKA, mMPTA, mLDFA, and JLCA, measured on the first and second WLR. With no significant systemic biases. A mean error between the HKA measurements on two separate WLRs of 0.01°, and the 95% limits of agreement between 1.15° and −1.13°
Fig. 5Bland Altman analyses of the Test-Retest results of the measured HKA, mMPTA, mLDFA and JLCA on two WLRs made at 2 different time points. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
This study demonstrated a successful introduction of a novel and easy-to-use protocol for obtaining WLRs in a real-world clinical setting. The objective of the present study was to examine the test-retest reliability of our WLR positioning protocol. We observed excellent ICCs for HKA, mMPTA, and mLDFA measured on the first and second WLRs in this test-retest study. A moderate ICC was observed for JLCA. The absolute errors of HKA, mMPTA, mLDFA, and JLCA between the first and second WLR were below the inter-observer variabilities.
To the best of our knowledge, only 1 small study with 8 participants has been published with the aim to optimize test-retest reproducibility of WLRs. Odenbring et al. used a protocol, positioning the patient on one leg and 10° of knee flexion. They made sure that the knee was orientated directly forward by superimposing the posterior aspects of the femoral condyles using a lateral fluoroscopic control to ensure optimal rotation.
This resulted in a test-retest mean absolute error of 1.3°, which is clinically inadequate, particularly given that many osteotomies are performed for deformities of 5° - 10°.
In contrast, our protocol, which is clinically readily employable, achieves a mean reliability of 0.4°, which is within the clinically desired accuracy for a HTO.
Jones et al. performed an in-silico study with the objective of describing the ideal accuracy for HTO. This was of clinical relevance given that it reported the desired measurement accuracies for lower limb geometry used as a benchmark for this study. Jones et al. found an ideal accuracy of 0.45° in order to achieve sufficient target corrections at the time of tibial osteotomy. Their findings support that our WLR positioning protocol is not only clinically readily-employed but also reliable, providing clinically acceptable WLR accuracy and reproducibility. Of note, our WLR positioning protocol was developed with the aim on user-friendliness, while achieving clinically sufficient reproducibility. The protocol is inexpensive to implement and easy to perform by X-ray technicians without direct, in-person clinician involvement.
Previous studies have demonstrated that a surgical accuracy of 0.45° can be achieved when using patient specific instrumentation (PSI) intraoperatively, however this requires a preoperative CT-scan and incurs associated radiation as well as additional costs.
In contrast, analyses of our plain film-based WLR results demonstrated that in 59% of cases the test-retest error was ≤0.45° and 95% of cases had test-retest error ≤1.15°. Of note, current standard HTO treatment is performed without 3D analyses or intraoperative use of PSI, with resulting poor accuracies far above 0.45° as reported by Van den Bempt et al.
This supports that the measurement errors of the proposed WLR protocol are substantially smaller than the subsequent technical surgical accuracy of conventional HTO, highlighting the clinical utility of the protocol.
In order to maximise user-friendliness, some positional parameters of the patients were not controlled by the investigated protocol. Weight bearing was based on patient perception of equal weight distribution on both legs. This can change between pre- and post-operative WLRs, with possible post-operative pain up to a year present after osteotomy in a unilateral fashion.
Knee joint distraction compared with high tibial osteotomy and total knee arthroplasty: two-year clinical, radiographic, and biochemical marker outcomes of two randomized controlled trials.
However, little is known about the amount of changing weight distribution following knee surgery and associated recovery. Our suggestion is to obtain post-operative assessment WLRs at the time when patients can endure full weight bearing on the operated limb during the radiograph moment.
Knee flexion was also only indirectly controlled in our protocol by asking patients to stand in full knee extension. Previous studies have demonstrated that a substantial degree of leg rotation of 10° or more is needed to alter HKA measurements in a clinically relevant fashion.
Kawahara S., Mawatari T., Matsui G., Mizu H., Akasaki Y. Malrotation of whole - leg radiograph less than 10 degrees does not influence preoperative planning in open - wedge high tibial osteotomy. 2020;1–7. doi:10.1002/jor.24845
A potential clinically relevant problem when considering that up to one-third of knee OA patients presents a knee flexion contracture of 6° or greater.
Of note, controlling leg rotation using fixed floor template is not suitable for every patient. Patients with a high degree of rotational deformity in the femur and/or tibia may not be able to participate in the WLR positioning protocol presented.
Indications for rotational errors on an AP knee radiograph are no (or too much) tibia-fibular overlap and femoral condyle asymmetry.
On the bilateral WLR with fixed feet positioning, differences between the left and right limb could also indicate rotational deformities. When patients do show rotational deformities above 10°, the surgeon should consider making a new WLR with feet in a different position using the Paley and Herzenberg protocol, compensating for the rotation deformity in the lower limb.
Additionally, we recommend considering computed tomography (CT) either biplanar radiography to assess lower limb rotation in such cases with complex rotational deformity, and which may be best addressed with multi-planar osteotomies. On CT images lower limb rotation can be compensated and analysed in a straightforward setting by aligning the tibia based on Akagi's line and the femur using the transepicondylar axis.
Our results show a moderate test-retest ICC for JLCA. At the same time, the mean absolute error in degrees (0.794°) is comparable or lower than the test-retest error of mMPTA (0.783°) and mLDFA (0.828°). The moderate ICC for JLCA can be explained by the normal distribution of the JLCA, which is between 0° and 2°.
This is narrower than the distribution of mMPTA and mLDFA suggesting that the absolute error of 0.794° is relatively higher than that of mMPTA and mLDFA given its narrowed underlying distribution. Furthermore, the measured angles of JLCA in the radiologic program used are provided in round numbers, creating an even narrower distribution for subsequent mean absolute error calculations.
This study is not without limitations. First, we used whole number measurements of mMPTA, mLDFA, and JLCA as OrthoStation provides these values in whole numbers. Therefore, it was not possible to analyse these parameters within one decimal place of accuracy. Nevertheless, given our excellent test-retest values, we believe we have obtained a clinically very reliable method for making WLRs. Second, we did not perform rotational analyses of the lower limbs. This could be valuable in order to study the effects of rotational deformities in the lower limb on the measured geometry. An implied limitation of employing single view, standardized AP radiographs is the lack of 3 dimensional information regarding the tibial and femoral rotation and knee flexion contracture.
Future studies should add a control group undergoing the Paley WLR protocol and consider measuring the test-retest mMPTA, mLDFA, and JLCA at and beyond one decimal place accuracy, understanding that further accuracy may be mathematically significant but of clinically decreasing relevance and utility. Finally, future studies should consider adding 3D analyses for the measurement of tibial and femoral rotation.
The novel WLR positioning protocol produced excellent and reproducible HKA measurements, with clinically acceptable degrees of error. We recommend applying this easy-to-use protocol when obtaining WLRs for osteotomy planning. Physicians still need to be aware of possible rotational and fixed flexion deformities in the lower limb present on WLRs and during physical examination.
Funding
This study was supported by a grant from the non-profit organization Dutch Arthritis Society (project number: 19–2–102).
Declaration of competing interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics approval
This prospective study was approved by the ethics committee of University Medical centre (UMC) Utrecht (METC number 19–474) on 27 November 2019. All included patients approved and signed informed consents.
Author contributions
HCN, NvE, and RJHC conducted this study, after developing the research question with WPG and HW. MH and HW helped with the analyses of de data. All authors contributed to the writing of the manuscript.
Acknowledgments
The authors want to thank the X-ray technicians of the radiology department of the University Medical Centre Utrecht for their help on implementing and performing the whole leg radiograph protocol.
References
Allen KD
Golightlya YM
Epidemiology of osteoarthritis: state of the evidence.
Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging.
Advances in modern osteotomies around the knee: report on the Association of Sports Traumatology, Arthroscopy, Orthopaedic surgery, Rehabilitation (ASTAOR) Moscow International Osteotomy Congress 2017.
Efficacy of one-stage cartilage repair using allogeneic mesenchymal stromal cells and autologous chondron transplantation (IMPACT) compared to nonsurgical treatment for focal articular cartilage lesions of the knee: study protocol for a crossover randomize.
Better clinical results after closed- compared to open-wedge high tibial osteotomy in patients with medial knee osteoarthritis and varus leg alignment.
Knee joint distraction compared with high tibial osteotomy and total knee arthroplasty: two-year clinical, radiographic, and biochemical marker outcomes of two randomized controlled trials.
Kawahara S., Mawatari T., Matsui G., Mizu H., Akasaki Y. Malrotation of whole - leg radiograph less than 10 degrees does not influence preoperative planning in open - wedge high tibial osteotomy. 2020;1–7. doi:10.1002/jor.24845
Template for feet positioning during a WLR, as part of the WLR positioning protocol. Feet are pointed outwards in 10° between the midlines and placed 10 from each other from the center of the heels. This template is engraved onto a durable Trespa board (right picture). WLR, whole leg radiographs.
Leg geometry measurements on a WLR in PACS IDS7 19.3 (Sectra AB, Linköping, Sweden). Left image illustrates the HKA measurement, the right image illustrates the mMPTA, mLDFA, and the JLCA. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
Bland Altman analyses of the Test-Retest results of the measured HKA, mMPTA, mLDFA and JLCA on two WLRs made at 2 different time points. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
Whole Leg Radiographs (WLR) are the gold standard for diagnosing malalignment and for pre-operative osteotomy planning. Positioning can affect the reproducibility of the measured hip knee angle (HKA), resulting in insufficient diagnostics and preoperative plans. We developed an easy-to-use WLR protocol by standardizing patient positioning and focusing on reproducibility.
Objectives
This study aims on testing this reproducibility of the novel WLR protocol.
Methods
This study enrolled 30 patients for a test-retest analysis. Each patient underwent two bilateral WLRs on the same day using the investigated positioning protocol. Three observers measured the HKA, mechanical medial proximal tibial angle (mMPTA), mechanical lateral distal femoral angle (mLDFA), and joint line convergence angle (JLCA) on the two radiographs. Twice each, with one week between.
Results
The intra-observer and inter-observer reliabilities were excellent, with intraclass correlation coefficients (ICCs) between 0.990 and 0.996. The ICCs between the measured HKA (0.985), mMPTA (0.922), and mLDFA (0.903) on the two separate radiographs were excellent. The ICC between the JLCA measured on the first and second WLR was moderate with 0.632. The mean absolute error between the HKA, mMPTA, mLDFA, and JLCA measurements on the first and second WLR were respectively: 0.442°, 0.783°, 0.828°, and 0.794°.
Conclusions
The investigated novel WLR positioning protocol produced excellent and reproducible HKA measurements, with clinically acceptable degrees of error. We recommend applying this easy-to-use protocol when obtaining WLRs for osteotomy planning. Physicians still need to be aware of possible rotational and fixed flexion deformities in the lower limb present on WLRs and during physical examination.
Planning corrective osteotomies require whole leg radiographs (WLR).5x5Issa, SN, Dunlop, D, Chang, A et al. Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging. Arthritis Care Res. 2007;
57: 398–406https://doi.org/10.1002/art.22618 Crossref | Scopus (69) | Google ScholarSee all References,9x9Lee, YS, Kim, MG, Byun, HW, Kim, SB, and Kim, JG. Reliability of the imaging software in the preoperative planning of the open-wedge high tibial osteotomy. Knee Surg Sport Traumatol Arthrosc. 2015;
: 846–851https://doi.org/10.1007/s00167-013-2700-z Crossref | PubMed | Scopus (33) | Google ScholarSee all References Accordingly, it is important that the radiographic technique when obtaining WLRs is consistent and accurate. Jones et al. reported a desired accuracy of 0.45° in order to achieve satisfactory intraoperative wedge accuracy for target corrections in high tibial osteotomy (HTO).10x10Jones, LD, Brown, CP, Jackson, W, Monk, AP, and Price, AJ. Assessing accuracy requirements in high tibial osteotomy: a theoretical, computer-based model using AP radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2017;
25: 2952–2956https://doi.org/10.1007/s00167-016-4092-3 Crossref | PubMed | Scopus (11) | Google ScholarSee all References
The most commonly used WLR protocol was published by Paley and Herzenberg, where the patella is used as a key landmark for antero-posterior orientation of the lower limb.11x11Paley, D. Radiographic Assessment of Lower Limb Deformities. Princ Deform Correct. 2002;
: 31–60https://doi.org/10.1007/978-3-642-59373-4_3 Crossref | Google ScholarSee all References This method can result in measurement biases, as patellar malalignment and/or positioning is quite common in cases of varus knee OA.12x12Harrison, MM, Cooke, TD, Fisher, SB, and Griffin, MP. Patterns of knee arthrosis and patellar subluxation. Clin Orthop Relat Res. 1994;
309: 56–63 PubMed | Google ScholarSee all References Maderbacher et al. reported a mean limb internal rotation of 8° with a range between 29.4° internal and 22.1° external rotation using this method.13x13Maderbacher, G, Matussek, J, Greimel, F et al. Lower limb malrotation is regularly present in long-leg radiographs resulting in significant measurement errors. J Knee Surg. 2021;
34: 108–114https://doi.org/10.1055/s-0039-1693668 Crossref | PubMed | Scopus (11) | Google ScholarSee all References Employing patella-based WLR protocols is even more difficult after total knee arthroplasty, with reported HKA measurement errors of up to 3.5°14x14Maderbacher, G, Baier, C, Benditz, A et al. Presence of rotational errors in long leg radiographs after total knee arthroplasty and impact on measured lower limb and component alignment. Int Orthop. 2017;
41: 1553–1560https://doi.org/10.1007/s00264-017-3408-3 Crossref | PubMed | Scopus (23) | Google ScholarSee all References
This is of clinical importance given that the reproducibility of the measured hip knee angle (HKA) on a WLR is influenced by positional differences of patients, like foot positioning, knee flexion, leg rotation, and weight bearing.15x15Sanfridsson, J, Ryd, L, Svahn, G, Fridén, T, and Jonsson, K. Radiographic measurement of femorotibial rotation in weight-bearing. Acta radiol. 2001;
42: 207–217https://doi.org/10.1080/028418501127346512 Crossref | PubMed | Scopus (13) | Google ScholarSee all References,16x16Brouwer, RW, Jakma, TSC, Brouwer, KH, and Verhaar, JAN. Pitfalls in determining knee alignment: a radographic cadaver study. J Knee Surg. 2007;
20: 210–215 Crossref | PubMed | Scopus (60) | Google ScholarSee all References,17x17Specogna, AV, Birmingham, TB, Hunt, MA et al. Radiographic measures of knee alignment in patients with varus gonarthrosis: effect of weightbearing status and associations with dynamic joint load. Am J Sports Med. 2007;
35: 65–70https://doi.org/10.1177/0363546506293024 Crossref | PubMed | Scopus (118) | Google ScholarSee all References,18x18Radtke, K, Becher, C, Noll, Y, and Ostermeier, S. Effect of limb rotation on radiographic alignment in total knee arthroplasties. Arch Orthop Trauma Surg. 2010;
130: 451–457https://doi.org/10.1007/s00402-009-0999-1 Crossref | PubMed | Scopus (91) | Google ScholarSee all References,19x19Lee, YS, Lee, BK, Lee, SH, Park, HG, Jun, DS, and Moon, DH. Effect of foot rotation on the mechanical axis and correlation between knee and whole leg radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2013;
21: 2542–2547https://doi.org/10.1007/s00167-013-2419-x Crossref | PubMed | Scopus (46) | Google ScholarSee all References,20x20Cho, YM, Ko, YJ, and Lee, W. Relationships among foot position, lower limb alignment, and knee adduction moment in patients with degenerative knee osteoarthritis. J Phys Ther Sci. 2015;
27: 1–4 Crossref | PubMed | Scopus (16) | Google ScholarSee all References This results in sub-optimal preoperative planning and assessment of achieved surgical corrections. This can cause substantial under- and overcorrections, which could lead to inferior HTO outcomes with reported poor surgical accuracies.21x21Ollivier, B, Berger, P, Depuydt, C, and Vandenneucker, H. Good long ‑ term survival and patient ‑ reported outcomes after high tibial osteotomy for medial compartment osteoarthritis. Knee Surgery, Sport Traumatol Arthrosc. 2020;
https://doi.org/10.1007/s00167-020-06262-4 Crossref | Scopus (23) | Google ScholarSee all References,22x22Van den Bempt, M, Van Genechten, W, Claes, T, and Claes, S. How accurately does high tibial osteotomy correct the mechanical axis of an arthritic varus knee? A systematic review. Knee. 2016;
23: 925–935https://doi.org/10.1016/j.knee.2016.10.001 Abstract | Full Text | Full Text PDF | PubMed | Scopus (51) | Google ScholarSee all References,23x23Gao, L, Madry, H, Chugaev, DV et al. Advances in modern osteotomies around the knee: report on the Association of Sports Traumatology, Arthroscopy, Orthopaedic surgery, Rehabilitation (ASTAOR) Moscow International Osteotomy Congress 2017. J Exp Orthop. 2019;
6https://doi.org/10.1186/s40634-019-0177-5 Crossref | Scopus (24) | Google ScholarSee all References Odenbring et al. performed a test-retest study including only 8 patients using a meticulous WLR protocol with a mean error of 1.3°, which improved accuracy but fell outside clinical feasibility for implementation.10x10Jones, LD, Brown, CP, Jackson, W, Monk, AP, and Price, AJ. Assessing accuracy requirements in high tibial osteotomy: a theoretical, computer-based model using AP radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2017;
25: 2952–2956https://doi.org/10.1007/s00167-016-4092-3 Crossref | PubMed | Scopus (11) | Google ScholarSee all References,24x24Odenbring, S, Berggren, AM, and Peil, L. Roentgenographic Assessment of the Hip-Knee-Ankle Axis in Medial Gonarthrosis. Clin Orthop Relat Res. 1993;
289: 195–196 (PMID: 8472414) Crossref | PubMed | Google ScholarSee all References
Recently, our team developed a novel and easy to use WLR positioning protocol and subsequently introduced this into clinical care, with the goal of standardizing patient positioning and reproducibility. The aim of the current study was to test the reproducibility of the implemented WLR protocol using a test-retest principle. We hypothesized that WLRs would be reproducible and within the desired published osteotomy accuracy of 0.45° with this novel protocol.
Methods
Patient population
This prospective study was approved by the ethics committee of University Medical Centre (UMC) Utrecht (METC number 19–474) on 27 November 2019. Patients were recruited at the UMC Utrecht Mobility Clinic, which is a tertiary orthopaedic referral centre for knee joint preserving treatments, including cartilage repair techniques, osteotomies, and knee joint distraction.25x25Korpershoek, JV, Vonk, LA, Kester, EC et al. Efficacy of one-stage cartilage repair using allogeneic mesenchymal stromal cells and autologous chondron transplantation (IMPACT) compared to nonsurgical treatment for focal articular cartilage lesions of the knee: study protocol for a crossover randomize. Trials. 2020;
21: 1–11https://doi.org/10.1186/s13063-020-04771-8 Crossref | PubMed | Scopus (7) | Google ScholarSee all References,26x26Jansen, MP, Mastbergen, SC, van Heerwaarden, RJ et al. Knee joint distraction in regular care for treatment of knee osteoarthritis: a comparison with clinical trial data. PLoS ONE. 2020;
15: 1–15https://doi.org/10.1371/journal.pone.0227975 Crossref | Scopus (20) | Google ScholarSee all References,27x27van Egmond, N, van Grinsven, S, van Loon, CJM, Gaasbeek, RD, and van Kampen, A. Better clinical results after closed- compared to open-wedge high tibial osteotomy in patients with medial knee osteoarthritis and varus leg alignment. Knee Surgery, Sport Traumatol Arthrosc. 2016;
24: 34–41https://doi.org/10.1007/s00167-014-3303-z Crossref | PubMed | Scopus (49) | Google ScholarSee all References Patients were eligible when they had an appointment at the outpatient clinic with scheduled WLR ordered. Exclusion criteria were: age under 18 years, inability to read, communicate, and/or speak the Dutch language, pregnancy, and patients incapable of providing informed consent. When willing to participate, patients signed an informed consent. Included patients underwent a second WLR after their appointment at the outpatient clinic, half an hour or more following their initial radiographic imaging (including WLRs). Both WLR acquisitions followed the same novel, standardized positioning protocol. This resulted in two separate episodes of patient positioning and radiograph acquisition within the same day separated by at least 30 min.
Positioning protocol
Patients were positioned with their knees in full extension. Feet were positioned with a distance of 10 cm between the heels and aligned in 10° of external rotation from the midline. This was achieved by placing the feet on a standardized positioning template (Fig. 1Fig. 1). X-ray technicians subsequently adjusted hip rotation, by aligning the upper body and pelvis in a straightforward (AP) position. No handlebars or supports were employed and patients were asked to place their hands alongside their body to ensure full weight-bearing. The X-ray technicians additionally instructed patients to distribute their weight equally to each leg. All WLRs were performed with the bilateral lower extremities captured on the radiographs, with a radiopaque measurement tape positioned behind the patients for subsequent image calibration. Figure 2Figure 2 illustrates a patient undergoing WLR with our standardized positioning protocol.
Fig. 1
Template for feet positioning during a WLR, as part of the WLR positioning protocol. Feet are pointed outwards in 10° between the midlines and placed 10 from each other from the center of the heels. This template is engraved onto a durable Trespa board (right picture). WLR, whole leg radiographs.
Each WLR was obtained using the Philips DigitalDiagnost v4.0 (Philips, Amsterdam, The Netherlands). The fixed distance between the detector plate and X-ray beam source was set to 265 cm. The X-ray beam source was fixed in height, and obtained 3 different bilateral images (proximal femur, knee joint, distal tibia) by pivoting the source towards upper, middle, and lower part of the legs. The 3 separate images were then stitched together into 1 WLR by Philips DigitalDiagnost v4.0 software. The X-ray settings were kept the same for each patient, with kV set at 81 and varying mAs.
Radiographic measurements
Radiographic measurements of the HKA were done for each leg separately by three observers in PACS IDS7 19.3 (Sectra AB, Linköping, Sweden), with the OrthoStation module. Two orthopaedic surgeons (RC&NE) and a researcher (CN) performed the measurements twice, with one week in between measurements by the same observer. The two radiographs of each patient were analysed separately and in a random order.
HKA was determined as the angle between the line from the centre of the femoral head to the centre of the femoral notch, and the line from the centre of the tibial spines to the centre of the talus (Fig. 3Fig. 3). The HKA was measured with 1 decimal place accuracy.
Fig. 3
Leg geometry measurements on a WLR in PACS IDS7 19.3 (Sectra AB, Linköping, Sweden). Left image illustrates the HKA measurement, the right image illustrates the mMPTA, mLDFA, and the JLCA. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
OrthoStation was employed to provide a semi-automated method to determine the joint line convergence angle (JLCA), mechanical medial proximal tibial angle (mMPTA), and mechanical lateral distal femoral angle (mLDFA). The tool required input from the observers, selecting the centre of the femoral head and talus, followed by marking the joint lines of the tibia and femur (Fig. 3Fig. 3). The mMPTA, mLDFA, and JLCA were provided in whole numbers.
Statistical analysis
Intra-observer reliability was tested using a 2 way mixed Intraclass Correlation (ICC) for absolute agreement. The inter-observer reliabilities were tested using a 2 way random ICC for absolute agreement. Test-retest agreement was calculated using a 2 way random absolute agreement ICC and Bland-Altman analyses. The errors between the measured parameters on the 2 WLRs and different observers were reported as mean (95%-CI interval). Due to a non-normal right skewed distribution of the absolute errors, these values were reported as mean (bootstrapped resampled 95%-CI intervals). Wilcoxon signed-rank tests were performed for differences in absolute test-retest errors between sub-groups included in this study. Statistical significance was set at alpha = 0.05. All statistical calculations were performed in SPSS Statistics (IBM, version 25.0.0.2.).
Power calculations for reproducibility studies are not straightforward and ill-studied.28x28Kottner, J, Audigé, L, Brorson, S et al. Guidelines for reporting reliability and agreement studies (GRRAS) were proposed. J Clin Epidemiol. 2011;
64: 96–106https://doi.org/10.1016/j.jclinepi.2010.03.002 Abstract | Full Text | Full Text PDF | PubMed | Scopus (1089) | Google ScholarSee all References Following the guideline as proposed by Bujanga and Baharum, we need only 14 legs to find a significant difference between an ICC of 0.7 and 0.9. We however deemed this number as very small and aimed to reach a narrower confidence interval.29x29Bujang, MA and Baharum, N. A simplified guide to determination of sample size requirements for estimating the value of intraclass correlation coefficient: a review. Arch Orofascial Sci. 2017;
12: 1–11 (Available at:)http://www.dental.usm.my/aos/docs/Vol_12/aos-article-0246.pdf. Google ScholarSee all References Following the advice of Cicchetti, we used three highly skilled readers to assess 60 legs in 30 patients, thus allowing us to include a range of alignments, male and females.30x30Cocchetti, DV. Sample size requirements for increasing the precision of reliability estimates: problems and proposed solutions. J Clin Exp Neuropsychol. 1999;
21: 567–570https://doi.org/10.1076/jcen.21.4.567.886 Crossref | PubMed | Scopus (28) | Google ScholarSee all References
Results
For this study, 31 patients signed an informed consent. We had to exclude one patient due to a no show for the retest radiograph. Therefore, 30 patients with 30 bilateral WLRs taken on 2 separate time-points (60 measured WLRs, 120 legs in total) were included in this study. The study included fifteen males and fifteen females, with a median age of 34.5 year (range: 18–61), and mean BMI of 25.8 (SD 3.2). Mean HKA was 179.3°, mean mMPTA was 87.1°, mean mLDFA was 87.3°, and mean JLCA was 1.3° (Fig. 4Fig. 4).
Fig. 4
Histograms of the population mean measured HKA, mMPTA, mLDFA, and JLCA. Reported normal distributions according to Bellemans et al. are: HKA = 178.67 ± 2.34, mMPTA = 87.04 ± 2.07, mLDFA = 87.90 ± 1.74, and JLCA = 0.51 ± 1.0538. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
Test results of the intra-observer and inter-observer reliabilities are listed in Table 1Table 1. Intra-observer and inter-observer reliabilities of the HKA, mMPTA, and mLDFA were excellent (ICC ≥ 0.850). Intra-observer and inter-observer reliabilities of the JLCA were fair to good (ICC ≥ 0.463). The mean absolute differences between the 3 observers of the measured HKA, mMPTA, mLDFA, and JLCA were respectively: 0.491° (CI 0.430° – 0.552°), 0.889° (CI 0.761° – 1.013°), 0.922° (CI 0.806° – 1.056), and 0.931° (CI 0.805° – 1.069°).
Table 1Calculated Intraclass Correlation (ICC) and the 95% Confidence Interval (CI).
Inter-observer and intra-observer reliabilities with significant
P-values when below .05.
All results of the test-retest analyses are listed in Table 2Table 2, including all lower limbs, sub-sections based on osteotomy surgery, osteoarthritis, and (osteo-)chondral lesion. When including all lower limbs, the test-retest ICCs between the measured HKA, mMPTA, and mLDFA on the first and second radiograph were excellent. The ICC between the measured JLCA on the first and second WLR was moderate. The mean absolute test-retest errors were for the measured HKA 0.442° (CI 0.386°−0.501°), mMPTA 0.783° (CI 0.683°−0.894°), mLDFA 0.828° (CI 0.711°−0.933°), and JLCA 0.794° (CI 0.689°−0.900°). Wilcoxon signed-rank tests revealed no significant differences between all lower limbs and males and/or females, osteoarthritis, (osteo-)chondral lesions, and osteotomy surgery in terms of HKA, mLDFA, mMPTA, and JLCA.
Table 2Calculated Test-Retest errors in mean absolute degrees and 95% Confidence Interval (CI) for the HKA, mMPTA, mLDFA, and JLCA. Intraclass Correlations between the measurements performed on the test and retest radiographs, with significant (*) P-values below .05.
Lower limbs are also divided based on the presence of osteoarthritis, (osteo-)chondral lesion, and osteotomy surgery.
P-values below .05.
Fig. 5Fig. 5 illustrates the Bland-Altman test-retest analyses of the HKA, mMPTA, mLDFA, and JLCA, measured on the first and second WLR. With no significant systemic biases. A mean error between the HKA measurements on two separate WLRs of 0.01°, and the 95% limits of agreement between 1.15° and −1.13°
Fig. 5
Bland Altman analyses of the Test-Retest results of the measured HKA, mMPTA, mLDFA and JLCA on two WLRs made at 2 different time points. HKA, Hip Knee Angle; JLCA, joint line convergence angle; mMPTA, mechanical medial proximal tibial angle; mLDFA, mechanical lateral distal femoral angle.
This study demonstrated a successful introduction of a novel and easy-to-use protocol for obtaining WLRs in a real-world clinical setting. The objective of the present study was to examine the test-retest reliability of our WLR positioning protocol. We observed excellent ICCs for HKA, mMPTA, and mLDFA measured on the first and second WLRs in this test-retest study. A moderate ICC was observed for JLCA. The absolute errors of HKA, mMPTA, mLDFA, and JLCA between the first and second WLR were below the inter-observer variabilities.
To the best of our knowledge, only 1 small study with 8 participants has been published with the aim to optimize test-retest reproducibility of WLRs. Odenbring et al. used a protocol, positioning the patient on one leg and 10° of knee flexion. They made sure that the knee was orientated directly forward by superimposing the posterior aspects of the femoral condyles using a lateral fluoroscopic control to ensure optimal rotation.24x24Odenbring, S, Berggren, AM, and Peil, L. Roentgenographic Assessment of the Hip-Knee-Ankle Axis in Medial Gonarthrosis. Clin Orthop Relat Res. 1993;
289: 195–196 (PMID: 8472414) Crossref | PubMed | Google ScholarSee all References This resulted in a test-retest mean absolute error of 1.3°, which is clinically inadequate, particularly given that many osteotomies are performed for deformities of 5° - 10°.10x10Jones, LD, Brown, CP, Jackson, W, Monk, AP, and Price, AJ. Assessing accuracy requirements in high tibial osteotomy: a theoretical, computer-based model using AP radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2017;
25: 2952–2956https://doi.org/10.1007/s00167-016-4092-3 Crossref | PubMed | Scopus (11) | Google ScholarSee all References,24x24Odenbring, S, Berggren, AM, and Peil, L. Roentgenographic Assessment of the Hip-Knee-Ankle Axis in Medial Gonarthrosis. Clin Orthop Relat Res. 1993;
289: 195–196 (PMID: 8472414) Crossref | PubMed | Google ScholarSee all References In contrast, our protocol, which is clinically readily employable, achieves a mean reliability of 0.4°, which is within the clinically desired accuracy for a HTO.10x10Jones, LD, Brown, CP, Jackson, W, Monk, AP, and Price, AJ. Assessing accuracy requirements in high tibial osteotomy: a theoretical, computer-based model using AP radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2017;
25: 2952–2956https://doi.org/10.1007/s00167-016-4092-3 Crossref | PubMed | Scopus (11) | Google ScholarSee all References
Jones et al. performed an in-silico study with the objective of describing the ideal accuracy for HTO. This was of clinical relevance given that it reported the desired measurement accuracies for lower limb geometry used as a benchmark for this study. Jones et al. found an ideal accuracy of 0.45° in order to achieve sufficient target corrections at the time of tibial osteotomy. Their findings support that our WLR positioning protocol is not only clinically readily-employed but also reliable, providing clinically acceptable WLR accuracy and reproducibility. Of note, our WLR positioning protocol was developed with the aim on user-friendliness, while achieving clinically sufficient reproducibility. The protocol is inexpensive to implement and easy to perform by X-ray technicians without direct, in-person clinician involvement.
Previous studies have demonstrated that a surgical accuracy of 0.45° can be achieved when using patient specific instrumentation (PSI) intraoperatively, however this requires a preoperative CT-scan and incurs associated radiation as well as additional costs.31x31Jacquet, C, Chan-Yu-Kin, J, Sharma, A, Argenson, JN, Parratte, S, and Ollivier, M. More accurate correction using “patient-specific” cutting guides in opening wedge distal femur varization osteotomies. Int Orthop. 2019;
43: 2285–2291https://doi.org/10.1007/s00264-018-4207-1 Crossref | PubMed | Scopus (26) | Google ScholarSee all References,32x32Victor, J and Premanathan, A. Virtual 3D planning and patient specific surgical guides for osteotomies around the knee: a feasibility and proof-of-concept study. Bone Joint J. 2013;
95-B: 153–158https://doi.org/10.1302/0301-620X.95B11.32950 Crossref | PubMed | Scopus (87) | Google ScholarSee all References In contrast, analyses of our plain film-based WLR results demonstrated that in 59% of cases the test-retest error was ≤0.45° and 95% of cases had test-retest error ≤1.15°. Of note, current standard HTO treatment is performed without 3D analyses or intraoperative use of PSI, with resulting poor accuracies far above 0.45° as reported by Van den Bempt et al.22x22Van den Bempt, M, Van Genechten, W, Claes, T, and Claes, S. How accurately does high tibial osteotomy correct the mechanical axis of an arthritic varus knee? A systematic review. Knee. 2016;
23: 925–935https://doi.org/10.1016/j.knee.2016.10.001 Abstract | Full Text | Full Text PDF | PubMed | Scopus (51) | Google ScholarSee all References This supports that the measurement errors of the proposed WLR protocol are substantially smaller than the subsequent technical surgical accuracy of conventional HTO, highlighting the clinical utility of the protocol.
In order to maximise user-friendliness, some positional parameters of the patients were not controlled by the investigated protocol. Weight bearing was based on patient perception of equal weight distribution on both legs. This can change between pre- and post-operative WLRs, with possible post-operative pain up to a year present after osteotomy in a unilateral fashion.33x33Jansen, MP, Besselink, NJ, van Heerwaarden, RJ et al. Knee joint distraction compared with high tibial osteotomy and total knee arthroplasty: two-year clinical, radiographic, and biochemical marker outcomes of two randomized controlled trials. Cartilage. 2019;
100https://doi.org/10.1177/1947603519828432 Crossref | Scopus (26) | Google ScholarSee all References However, little is known about the amount of changing weight distribution following knee surgery and associated recovery. Our suggestion is to obtain post-operative assessment WLRs at the time when patients can endure full weight bearing on the operated limb during the radiograph moment.
Knee flexion was also only indirectly controlled in our protocol by asking patients to stand in full knee extension. Previous studies have demonstrated that a substantial degree of leg rotation of 10° or more is needed to alter HKA measurements in a clinically relevant fashion.8x8Bayliss, LE, Culliford, D, Monk, AP et al. The effect of patient age at intervention on risk of implant revision after total replacement of the hip or knee: a population-based cohort study. Lancet. 2017;
389: 1424–1430https://doi.org/10.1016/S0140-6736(17)30059-4 Abstract | Full Text | Full Text PDF | PubMed | Scopus (371) | Google ScholarSee all References,18x18Radtke, K, Becher, C, Noll, Y, and Ostermeier, S. Effect of limb rotation on radiographic alignment in total knee arthroplasties. Arch Orthop Trauma Surg. 2010;
130: 451–457https://doi.org/10.1007/s00402-009-0999-1 Crossref | PubMed | Scopus (91) | Google ScholarSee all References,34x34Kawahara S., Mawatari T., Matsui G., Mizu H., Akasaki Y. Malrotation of whole - leg radiograph less than 10 degrees does not influence preoperative planning in open - wedge high tibial osteotomy. 2020;1–7. doi:10.1002/jor.24845 Google ScholarSee all References,35x35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022 Abstract | Full Text | Full Text PDF | Scopus (5) | Google ScholarSee all References But with added knee flexion of 10°, the error becomes at least 2°.35x35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022 Abstract | Full Text | Full Text PDF | Scopus (5) | Google ScholarSee all References,36x36Jud, L, Trache, T, Tondelli, T, Fürnstahl, P, Fucentese, SF, and Vlachopoulos, L. Rotation or flexion alters mechanical leg axis measurements comparably in patients with different coronal alignment. Knee Surgery, Sport Traumatol Arthrosc. 2019;
https://doi.org/10.1007/s00167-019-05779-7(0123456789) Crossref | Scopus (13) | Google ScholarSee all References A potential clinically relevant problem when considering that up to one-third of knee OA patients presents a knee flexion contracture of 6° or greater.37x37Ritter, MA, Lutgring, JD, Davis, KE, Berend, ME, Pierson, JL, and Meneghini, RM. The Role of Flexion Contracture on Outcomes in Primary Total Knee Arthroplasty. J Arthroplasty. 2007;
22: 1092–1096https://doi.org/10.1016/j.arth.2006.11.009 Abstract | Full Text | Full Text PDF | PubMed | Scopus (144) | Google ScholarSee all References This highlights the importance of controlling leg rotation during WLR acquisition, to minimize the effect of knee flexion on standing WLRs.35x35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022 Abstract | Full Text | Full Text PDF | Scopus (5) | Google ScholarSee all References,36x36Jud, L, Trache, T, Tondelli, T, Fürnstahl, P, Fucentese, SF, and Vlachopoulos, L. Rotation or flexion alters mechanical leg axis measurements comparably in patients with different coronal alignment. Knee Surgery, Sport Traumatol Arthrosc. 2019;
https://doi.org/10.1007/s00167-019-05779-7(0123456789) Crossref | Scopus (13) | Google ScholarSee all References Of note, controlling leg rotation using fixed floor template is not suitable for every patient. Patients with a high degree of rotational deformity in the femur and/or tibia may not be able to participate in the WLR positioning protocol presented.
Indications for rotational errors on an AP knee radiograph are no (or too much) tibia-fibular overlap and femoral condyle asymmetry.38x38Maderbacher, G, Schaumburger, J, Baier, C et al. Predicting knee rotation by the projection overlap of the proximal fibula and tibia in long-leg radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2014;
22: 2982–2988https://doi.org/10.1007/s00167-014-3327-4 Crossref | PubMed | Scopus (27) | Google ScholarSee all References On the bilateral WLR with fixed feet positioning, differences between the left and right limb could also indicate rotational deformities. When patients do show rotational deformities above 10°, the surgeon should consider making a new WLR with feet in a different position using the Paley and Herzenberg protocol, compensating for the rotation deformity in the lower limb.11x11Paley, D. Radiographic Assessment of Lower Limb Deformities. Princ Deform Correct. 2002;
: 31–60https://doi.org/10.1007/978-3-642-59373-4_3 Crossref | Google ScholarSee all References They suggested pointing the patella surface straightforward during acquisition as indicator for AP radiographs.11x11Paley, D. Radiographic Assessment of Lower Limb Deformities. Princ Deform Correct. 2002;
: 31–60https://doi.org/10.1007/978-3-642-59373-4_3 Crossref | Google ScholarSee all References Additionally, we recommend considering computed tomography (CT) either biplanar radiography to assess lower limb rotation in such cases with complex rotational deformity, and which may be best addressed with multi-planar osteotomies. On CT images lower limb rotation can be compensated and analysed in a straightforward setting by aligning the tibia based on Akagi's line and the femur using the transepicondylar axis.39x39Aglietti, P, Sensi, L, Cuomo, P, and Ciardullo, A. Rotational position of femoral and tibial components in TKA using the femoral transepicondylar axis. Clin Orthop Relat Res. 2008;
466: 2751–2755https://doi.org/10.1007/s11999-008-0452-8 Crossref | PubMed | Scopus (119) | Google ScholarSee all References,40x40Saffarini, M, Nover, L, Tandogan, R et al. The original Akagi line is the most reliable: a systematic review of landmarks for rotational alignment of the tibial component in TKA. Knee Surgery, Sport Traumatol Arthrosc. 2018;
27: 1018–1027https://doi.org/10.1007/s00167-018-5131-z Crossref | PubMed | Scopus (41) | Google ScholarSee all References
Our results show a moderate test-retest ICC for JLCA. At the same time, the mean absolute error in degrees (0.794°) is comparable or lower than the test-retest error of mMPTA (0.783°) and mLDFA (0.828°). The moderate ICC for JLCA can be explained by the normal distribution of the JLCA, which is between 0° and 2°.41x41Paley, D. Normal Lower Limb Alignment and Joint Orientation. Principles of Deformity Correction. Springer Berlin Heidelberg,
Berlin, Heidelberg; 2002: 1–18 Crossref | Google ScholarSee all References This is narrower than the distribution of mMPTA and mLDFA suggesting that the absolute error of 0.794° is relatively higher than that of mMPTA and mLDFA given its narrowed underlying distribution. Furthermore, the measured angles of JLCA in the radiologic program used are provided in round numbers, creating an even narrower distribution for subsequent mean absolute error calculations.
This study is not without limitations. First, we used whole number measurements of mMPTA, mLDFA, and JLCA as OrthoStation provides these values in whole numbers. Therefore, it was not possible to analyse these parameters within one decimal place of accuracy. Nevertheless, given our excellent test-retest values, we believe we have obtained a clinically very reliable method for making WLRs. Second, we did not perform rotational analyses of the lower limbs. This could be valuable in order to study the effects of rotational deformities in the lower limb on the measured geometry. An implied limitation of employing single view, standardized AP radiographs is the lack of 3 dimensional information regarding the tibial and femoral rotation and knee flexion contracture.35x35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022 Abstract | Full Text | Full Text PDF | Scopus (5) | Google ScholarSee all References,36x36Jud, L, Trache, T, Tondelli, T, Fürnstahl, P, Fucentese, SF, and Vlachopoulos, L. Rotation or flexion alters mechanical leg axis measurements comparably in patients with different coronal alignment. Knee Surgery, Sport Traumatol Arthrosc. 2019;
https://doi.org/10.1007/s00167-019-05779-7(0123456789) Crossref | Scopus (13) | Google ScholarSee all References Third, this study included no control group undergoing the WLR protocol as proposed by Paley.11x11Paley, D. Radiographic Assessment of Lower Limb Deformities. Princ Deform Correct. 2002;
: 31–60https://doi.org/10.1007/978-3-642-59373-4_3 Crossref | Google ScholarSee all References However, Maderbacher et al. revealed significant errors when using this protocol, especially in cases with knee arthroplasty.13x13Maderbacher, G, Matussek, J, Greimel, F et al. Lower limb malrotation is regularly present in long-leg radiographs resulting in significant measurement errors. J Knee Surg. 2021;
34: 108–114https://doi.org/10.1055/s-0039-1693668 Crossref | PubMed | Scopus (11) | Google ScholarSee all References,14x14Maderbacher, G, Baier, C, Benditz, A et al. Presence of rotational errors in long leg radiographs after total knee arthroplasty and impact on measured lower limb and component alignment. Int Orthop. 2017;
41: 1553–1560https://doi.org/10.1007/s00264-017-3408-3 Crossref | PubMed | Scopus (23) | Google ScholarSee all References Future studies should add a control group undergoing the Paley WLR protocol and consider measuring the test-retest mMPTA, mLDFA, and JLCA at and beyond one decimal place accuracy, understanding that further accuracy may be mathematically significant but of clinically decreasing relevance and utility. Finally, future studies should consider adding 3D analyses for the measurement of tibial and femoral rotation.35x35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022 Abstract | Full Text | Full Text PDF | Scopus (5) | Google ScholarSee all References,36x36Jud, L, Trache, T, Tondelli, T, Fürnstahl, P, Fucentese, SF, and Vlachopoulos, L. Rotation or flexion alters mechanical leg axis measurements comparably in patients with different coronal alignment. Knee Surgery, Sport Traumatol Arthrosc. 2019;
https://doi.org/10.1007/s00167-019-05779-7(0123456789) Crossref | Scopus (13) | Google ScholarSee all References
Conclusions
The novel WLR positioning protocol produced excellent and reproducible HKA measurements, with clinically acceptable degrees of error. We recommend applying this easy-to-use protocol when obtaining WLRs for osteotomy planning. Physicians still need to be aware of possible rotational and fixed flexion deformities in the lower limb present on WLRs and during physical examination.
Funding
This study was supported by a grant from the non-profit organization Dutch Arthritis Society (project number: 19–2–102).
Declaration of competing interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics approval
This prospective study was approved by the ethics committee of University Medical centre (UMC) Utrecht (METC number 19–474) on 27 November 2019. All included patients approved and signed informed consents.
Author contributions
HCN, NvE, and RJHC conducted this study, after developing the research question with WPG and HW. MH and HW helped with the analyses of de data. All authors contributed to the writing of the manuscript.
Acknowledgments
The authors want to thank the X-ray technicians of the radiology department of the University Medical Centre Utrecht for their help on implementing and performing the whole leg radiograph protocol.
References
1Allen, KD and Golightlya, YM. Epidemiology of osteoarthritis: state of the evidence. (Epidemiology)Curr Opin Rheumatol. 2015;
27: 276–283https://doi.org/10.1097/BOR.0000000000000161
4Palmer, JS, Jones, LD, Monk, AP et al. Varus alignment of the proximal tibia is associated with structural progression in early to moderate varus osteoarthritis of the knee. Knee Surgery, Sport Traumatol Arthrosc. 2020;
https://doi.org/10.1007/s00167-019-05840-5
5Issa, SN, Dunlop, D, Chang, A et al. Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging. Arthritis Care Res. 2007;
57: 398–406https://doi.org/10.1002/art.22618
8Bayliss, LE, Culliford, D, Monk, AP et al. The effect of patient age at intervention on risk of implant revision after total replacement of the hip or knee: a population-based cohort study. Lancet. 2017;
389: 1424–1430https://doi.org/10.1016/S0140-6736(17)30059-4
9Lee, YS, Kim, MG, Byun, HW, Kim, SB, and Kim, JG. Reliability of the imaging software in the preoperative planning of the open-wedge high tibial osteotomy. Knee Surg Sport Traumatol Arthrosc. 2015;
: 846–851https://doi.org/10.1007/s00167-013-2700-z
10Jones, LD, Brown, CP, Jackson, W, Monk, AP, and Price, AJ. Assessing accuracy requirements in high tibial osteotomy: a theoretical, computer-based model using AP radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2017;
25: 2952–2956https://doi.org/10.1007/s00167-016-4092-3
14Maderbacher, G, Baier, C, Benditz, A et al. Presence of rotational errors in long leg radiographs after total knee arthroplasty and impact on measured lower limb and component alignment. Int Orthop. 2017;
41: 1553–1560https://doi.org/10.1007/s00264-017-3408-3
17Specogna, AV, Birmingham, TB, Hunt, MA et al. Radiographic measures of knee alignment in patients with varus gonarthrosis: effect of weightbearing status and associations with dynamic joint load. Am J Sports Med. 2007;
35: 65–70https://doi.org/10.1177/0363546506293024
18Radtke, K, Becher, C, Noll, Y, and Ostermeier, S. Effect of limb rotation on radiographic alignment in total knee arthroplasties. Arch Orthop Trauma Surg. 2010;
130: 451–457https://doi.org/10.1007/s00402-009-0999-1
19Lee, YS, Lee, BK, Lee, SH, Park, HG, Jun, DS, and Moon, DH. Effect of foot rotation on the mechanical axis and correlation between knee and whole leg radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2013;
21: 2542–2547https://doi.org/10.1007/s00167-013-2419-x
20Cho, YM, Ko, YJ, and Lee, W. Relationships among foot position, lower limb alignment, and knee adduction moment in patients with degenerative knee osteoarthritis. J Phys Ther Sci. 2015;
27: 1–4
21Ollivier, B, Berger, P, Depuydt, C, and Vandenneucker, H. Good long ‑ term survival and patient ‑ reported outcomes after high tibial osteotomy for medial compartment osteoarthritis. Knee Surgery, Sport Traumatol Arthrosc. 2020;
https://doi.org/10.1007/s00167-020-06262-4
22Van den Bempt, M, Van Genechten, W, Claes, T, and Claes, S. How accurately does high tibial osteotomy correct the mechanical axis of an arthritic varus knee? A systematic review. Knee. 2016;
23: 925–935https://doi.org/10.1016/j.knee.2016.10.001
23Gao, L, Madry, H, Chugaev, DV et al. Advances in modern osteotomies around the knee: report on the Association of Sports Traumatology, Arthroscopy, Orthopaedic surgery, Rehabilitation (ASTAOR) Moscow International Osteotomy Congress 2017. J Exp Orthop. 2019;
6https://doi.org/10.1186/s40634-019-0177-5
24Odenbring, S, Berggren, AM, and Peil, L. Roentgenographic Assessment of the Hip-Knee-Ankle Axis in Medial Gonarthrosis. (PMID: 8472414)Clin Orthop Relat Res. 1993;
289: 195–196
25Korpershoek, JV, Vonk, LA, Kester, EC et al. Efficacy of one-stage cartilage repair using allogeneic mesenchymal stromal cells and autologous chondron transplantation (IMPACT) compared to nonsurgical treatment for focal articular cartilage lesions of the knee: study protocol for a crossover randomize. Trials. 2020;
21: 1–11https://doi.org/10.1186/s13063-020-04771-8
26Jansen, MP, Mastbergen, SC, van Heerwaarden, RJ et al. Knee joint distraction in regular care for treatment of knee osteoarthritis: a comparison with clinical trial data. PLoS ONE. 2020;
15: 1–15https://doi.org/10.1371/journal.pone.0227975
27van Egmond, N, van Grinsven, S, van Loon, CJM, Gaasbeek, RD, and van Kampen, A. Better clinical results after closed- compared to open-wedge high tibial osteotomy in patients with medial knee osteoarthritis and varus leg alignment. Knee Surgery, Sport Traumatol Arthrosc. 2016;
24: 34–41https://doi.org/10.1007/s00167-014-3303-z
28Kottner, J, Audigé, L, Brorson, S et al. Guidelines for reporting reliability and agreement studies (GRRAS) were proposed. J Clin Epidemiol. 2011;
64: 96–106https://doi.org/10.1016/j.jclinepi.2010.03.002
29Bujang, MA and Baharum, N. A simplified guide to determination of sample size requirements for estimating the value of intraclass correlation coefficient: a review. (Available at:)Arch Orofascial Sci. 2017;
12: 1–11
31Jacquet, C, Chan-Yu-Kin, J, Sharma, A, Argenson, JN, Parratte, S, and Ollivier, M. More accurate correction using “patient-specific” cutting guides in opening wedge distal femur varization osteotomies. Int Orthop. 2019;
43: 2285–2291https://doi.org/10.1007/s00264-018-4207-1
32Victor, J and Premanathan, A. Virtual 3D planning and patient specific surgical guides for osteotomies around the knee: a feasibility and proof-of-concept study. Bone Joint J. 2013;
95-B: 153–158https://doi.org/10.1302/0301-620X.95B11.32950
33Jansen, MP, Besselink, NJ, van Heerwaarden, RJ et al. Knee joint distraction compared with high tibial osteotomy and total knee arthroplasty: two-year clinical, radiographic, and biochemical marker outcomes of two randomized controlled trials. Cartilage. 2019;
100https://doi.org/10.1177/1947603519828432
34Kawahara S., Mawatari T., Matsui G., Mizu H., Akasaki Y. Malrotation of whole - leg radiograph less than 10 degrees does not influence preoperative planning in open - wedge high tibial osteotomy. 2020;1–7. doi:10.1002/jor.24845
35Nguyen, HC, Gielis, WP, van Egmond, N et al. The need for a standardized whole leg radiograph guideline: the effects of knee flexion, leg rotation, and X-ray beam height. J Cartil Jt Preserv. 2021;
1: 100022https://doi.org/10.1016/j.jcjp.2021.100022
36Jud, L, Trache, T, Tondelli, T, Fürnstahl, P, Fucentese, SF, and Vlachopoulos, L. Rotation or flexion alters mechanical leg axis measurements comparably in patients with different coronal alignment. (0123456789)Knee Surgery, Sport Traumatol Arthrosc. 2019;
https://doi.org/10.1007/s00167-019-05779-7
37Ritter, MA, Lutgring, JD, Davis, KE, Berend, ME, Pierson, JL, and Meneghini, RM. The Role of Flexion Contracture on Outcomes in Primary Total Knee Arthroplasty. J Arthroplasty. 2007;
22: 1092–1096https://doi.org/10.1016/j.arth.2006.11.009
38Maderbacher, G, Schaumburger, J, Baier, C et al. Predicting knee rotation by the projection overlap of the proximal fibula and tibia in long-leg radiographs. Knee Surgery, Sport Traumatol Arthrosc. 2014;
22: 2982–2988https://doi.org/10.1007/s00167-014-3327-4
39Aglietti, P, Sensi, L, Cuomo, P, and Ciardullo, A. Rotational position of femoral and tibial components in TKA using the femoral transepicondylar axis. Clin Orthop Relat Res. 2008;
466: 2751–2755https://doi.org/10.1007/s11999-008-0452-8
40Saffarini, M, Nover, L, Tandogan, R et al. The original Akagi line is the most reliable: a systematic review of landmarks for rotational alignment of the tibial component in TKA. Knee Surgery, Sport Traumatol Arthrosc. 2018;
27: 1018–1027https://doi.org/10.1007/s00167-018-5131-z
41Paley, D. Normal Lower Limb Alignment and Joint Orientation. Principles of Deformity Correction. Springer Berlin Heidelberg,
Berlin, Heidelberg; 2002: 1–18
P-values when below .05.
Article Info
00005-1&id=gr1.jpg){kind=link}
00005-1&id=gr2.jpg){kind=link}
00005-1&id=gr3.jpg){kind=link}
00005-1&id=gr4.jpg){kind=link}
00005-1&id=gr5.jpg){kind=link}
