Large osteophyte removal from the posterior femoral condyle significantly improves extension at the time of surgery in a total knee arthroplasty

Abstract

Removing osteophytes from the posterior compartment of the femur eliminates the tenting effects on the joint capsule and consequently increases the extension gap in total knee arthroplasty. However, there is no clear association with the size of osteophytes removed and the potential degree of additional extension achieved at time of surgery.

Aims

Correlate the size of posterior osteophytes removed with the degree of extension gained intraoperatively in total knee arthroplasty and develop a radiological classification system to grade these osteophytes.

Methods

Patients who underwent a TKA had pre and post operative sagittal radiographs assessed and classified according to 4 different categories of a proposed classification system. Knee extension was then assessed by a computer navigated system before incision and after implant insertion. Confounding factors were controlled and considered on the analysis. The study was done retrospectively.

Results

147 patients were included in the study. Ninety-three (63.2%) patients had osteophytes on the posterior aspect of the femur completely removed and fifty-four patients (36.8%) did not have radiological evidence of osteophytes on the posterior aspect of the femur. There was a positive and linear correlation (Pearson correlation 0.327, p .005) between osteophyte size and degree of extension corrected at time of surgery. On Multivariate Logistic Regression Analysis, we found that small osteophytes (Grade 1) did not seem to affect the extension, while removing Grade 2 or Grade 3 osteophytes lead to a gain in extension of 2.7 and 4.5° respectively.

Conclusion

Removing large osteophytes (Grade 2 and Grade 3) from the posterior femoral compartment can be used as an adjuvant strategy to ensure that intraoperative extension is optimal. However removing small osteophytes (Grade 1) should not be expected to affect extension at the time of surgery in TKA and could increase intra-operative time and morbidity.

1. Background

Fixed flexion following total knee arthroplasty (TKA) is a major concern and even minimal flexion deformities (less than 5°) may result in pain, limp, early fatigue with walking and an unsatisfactory outcome at mid-term following total knee replacement.¹,²,³ Multiple intra-operative strategies are described to help achieve full extension in knee arthroplasty with the most common recognising differences in the extension gap as a key contributor to flexion contracture correction. These include distal femur resection, soft tissue balance and femoral component placement in flexion.⁴,⁵,⁶,⁷ Theoretically, removing the posterior condyle of the femur or osteophytes from the posterior compartment eliminates tenting effects on the joint capsule caused by these structures that consequently increases the extension gap. Sriphirom P et al. assessed the effect of osteophyte removal from the posterior compartment of the femur in flexion and extension gaps in 40 varus knees. The authors found that 52% of patients had the medial extension gap increased by 1 mm (range 1–3). However, 48% of patients did not achieve any gains in the extension gap⁸ and they did not correlate their findings with flexion contracture correction or changes in sagittal alignment. We postulate that removing large osteophytes from the posterior compartment of the femur may considerably increase the extension gap and result in improved extension at the time of surgery. However, removal of minimal or small osteophytes may not result in significant improvement in extension at the time of surgery. To the best of our knowledge, there is no validated radiological classification system to grade the size of osteophytes on the posterior aspect of the femur. The primary objective of the study was assessing if osteophyte removal from the posterior aspect of the femur was an independent variable affecting extension at the time of surgery. The secondary objective was to correlate the findings with a proposed classification system to grade the radiological thickness of osteophyte removed during TKA.

We tested the hypotheses that (1) large osteophytes removed from the posterior aspect of the femur may affect the sagittal alignment (extension) at the time of surgery whilst (2) small osteophytes removed from the posterior aspect of the femur may not.

1.1. Clinical relevance

If the hypotheses are true, (1) removing specific sized osteophytes from the posterior aspect of the femur may help in the correction of fixed flexion contractures at the time of surgery.

2. Material and methods

From November 2008 to July 2018 patients who underwent total knee replacement by a single senior surgeon using NexGen Cruciate-Retaining CR-Flex Fixed Bearing Knee (Zimmer-Biomet, Warsaw, IN, USA) were eligible for the study. Inclusion criteria included idiopathic knee osteoarthritis, flexion contracture ≥ 5°. The following patients were excluded from the analysis: (1) pre-operative coronal deformity > 10°; (2) post-traumatic or inflammatory osteoarthritis; (3) missing pre or post-operative radiographies; (4) sagittal radiographs with more than 5 mm of femoral malrotation (Fig. 1), (5) patients with pre-operative knee instability (6) patients who had the joint line changed more than 2 mm (7) patients with posterior femoral offset changed more than 2 mm, (8) patients with concomitant extensive soft tissue release (9) or evidence of posterior capsular release (10) and patients with postoperative radiological evidence of incomplete osteophyte removal. The database has been collected and kept confidential by our institution. This study was approved by the review board of our institution. Our primary outcome was the change in extension (Δ Extension) from the pre-operative navigation data compared to the post-operative extension at the time of surgery after definitive implant insertion.

Fig. 1.

Sagittal radiography with more than 5 mm of femoral malrotation.

2.1. Surgical technique

All procedures were performed by the same senior surgeon using the same technique. The patient is placed supine with the knee at 90° of knee flexion. No tourniquet is applied. Two threaded pins (3.5 mm) are bicortically drilled into the anteromedial border of the tibia, 4 cm below tibial tubercle and 2 threaded pins (3.5 mm) drilled into the anterior femur to mount trackers for a computed navigation system (OrthoSoft Knee Universal®, Zimmer Biomet - Montreal, CA). Sagittal and coronal pre-operative deformities are recorded. A medial parapatellar approach is used. The anterior cruciate ligament (ACL) is resected with the posterior cruciate ligament (PCL) retained. Visible femoral and tibial osteophytes are removed from the anterior compartment of the knee. Recommended anatomical landmarks are registered with the pointer according to navigation system requirements. Femoral component is sized and selected according to surgeon preference based on information from an anteroposterior sizer device and navigation data. A Multi-Reference 4-in-1 Instrument guide creates a rectangular and symmetrical flexion gap between the femur and tibia. Neutral mechanical coronal alignment was aimed for on the femur in the axial plane parallel to transepicondilar axis (TEA)and 3° of femoral flexion in the sagittal plane. After posterior femoral and postero-oblique chamfer cuts were made, patients who had radiological evidence of osteophytes had the posterior compartment of the femur assessed and osteophytes were removed with a curved osteotome but the posterior capsule of the knee was preserved. The proximal tibia was then cut in neutral coronal alignment using the extramedullary alignment guide with 6° of posterior slope in the sagittal plane. Once implant trials were inserted, ligament balance was checked in full extension, 30° and 90° flexion with manual varus and valgus stress tests. More than 2 mm of asymmetry was considered suboptimal and adjusted with sequential soft tissue releases.

2.2. Pre-operative and post-operative sagittal assessment

Patients had total knee replacements performed under navigation (OrthoSoft Knee Universal®, Zimmer Biomet - Montreal, CA). Before the incision, anatomical landmarks were collected to determine coronal and sagittal alignments of the limb. The surgeon then raised the limb from the heel 30 cm from the table and the coronal and sagittal deformities (including flexion and extension) were recorded. At the end of the surgery, after definitive implants were inserted, the same procedure was repeated and extension was again recorded (Fig. 2). Negative measures represent hyperextension and positive measures represent flexion contracture.

Fig. 2.

Post-operative assessment of extension with navigation. The leg is raised and held from the heel 30 cm from the table and the coronal and sagittal alignments are recorded. This procedure is also performed pre-operative before bone cuts.

2.3. Posterior femoral osteophytes assessment

As part of the pre-operative assessment, patients had weight-bearing anteroposterior and non-weight bearing lateral radiographs performed within one month of the surgery. Post-operative radiographs confirmed the complete removal of posterior osteophytes (Fig. 3) and those with incomplete osteophyte removal were excluded from the study. In order to quantify the size of the osteophytes removed, sagittal thickness was measured on lateral digital radiography with the following technique using a picture archiving and communication system (PACS) workstation: a line is drawn on the blumensaat line from anterior to the most posterior aspect of the femoral cortex (Line A) (Fig. 4). A second perpendicular line was then drawn on the most posterior aspect of femoral cortex (Line B). The third line, parallel to Line B, was drawn on the most posterior edge of the osteophyte (Line C). The distance from Line B to Line C was then recorded as the sagittal osteophyte thickness and they were classified into 4 different grades: Grade 0: absence of osteophytes, Grade 1: ≤10 mm, Grade 2: 11–15 mm, Grade 3: >15 mm (Table 1, Fig. 5).

Fig. 3.

Figure a evidences a 9.2 mm thick osteophyte on posterior femoral compartment. Figure b: post-operative sagittal radiography confirms a complete osteophyte removal.

Fig. 4.

Method of posterior femoral osteophyte measurement according to Leie et al: Figure a. yellow arrow evidence the edge of posterior wall in a lateral radiography with femoral condyle with acceptable rotation. White arrow evidences the edges of posterior femoral osteophyte. Figure b: A line is drawn on the blumensaat line up to the posterior edge of femoral cortex (Line A). Figure c: A second line (Line B) is drawn on the most posterior aspect of femoral cortex, perpendicular to Line A. Figure d: a third line (Line C) is drawn parallel to line B and on the most posterior edge of osteophyte. The distance between Line B and Line C is recorded as osteophyte sagittal thickness (14.8 mm).

Table 1.

Leie Classification for Posterior Femoral Osteophytes.

Grade	Sagittal Thickness
Grade 0	absence
Grade 1	0–10 mm
Grade 2	11–15 mm
Grade 3	>15 mm

Fig. 5.

Sagittal radiography evidences osteophytes of different sizes according to Leie Classification. Figure a: Grade 0 osteophyte – absent. Figure b: Grade 1 osteophyte: 1–10 mm. Figure c: Grade 2 osteophyte: 11–15 mm, Grade 3 osteophyte: > 15 mm.

2.4. Femoral and tibial bone thickness resection assessment

A calliper measurement device (Fig. 6) was used to confirm the level of femoral and tibial resection. The thickest point of every femoral and tibial resection was reported in millimetres in the following order: femoral distal medial (DM), femoral distal lateral (DL), femoral posteromedial (PM), femoral posterolateral (PL), medial tibial plateau (MP), lateral plateau (LP). A thickness of 1 mm was added to every cut to account for bone saw thickness.

Fig. 6.

Calliper device used to measure posteromedial femoral resection (left) and posterolateral femoral resection (Right).

2.5. Con-founder factors assessment

In order to assess the effect of osteophyte removal from the posterior aspect of the femur, recognized factors that affect the correction of sagittal deformities were controlled.⁹,¹⁰,¹¹,¹² The joint line was determined considering the femoral resection of the healthy compartment of the knee. To determine the posterior femoral condylar offset we considered the amount of bone resected on the posteromedial and posterolateral femoral condyle minus the thickness of the posterior femoral component chosen for each patient. Soft tissue balance was performed according to surgeon experience and those patients who had extensive soft tissue releases (posterolateral corner, posterior capsule release) were also excluded from the study.

2.6. Statistical analysis

We used SPSS version 23.0 (IBM Corp., Armonk, NY, USA) to conduct statistical analyses. Descriptive statistics are provided with counts and percentages for categorical variables and mean ranges for continuous variables. Pearson's chi-square test was used to discover if there was a relationship between two categorical variables. The one-way analysis of variance (ANOVA) was used to determine whether there were any statistically significant differences between the means of two or more independent (unrelated) groups. One-way analysis of covariance (ANCOVA) was used to determine whether there were any significant differences between two or more independent (unrelated) groups on a dependent variable. Intra and interclass correlation (ICC) were performed to compare the reliability of the measurements of the proposed classification between two blinded researchers. A Post hoc power analysis (G*Power 3.1.9.3 for Mac OS X 10.7 to 10.3) indicated that to detect a small effect size (f² = 0.10), the final model with four predictors had a power (1-β) = 0.98. A p value of less than 0.05 was considered significant.

3. Results

Between November 2008 and July 2018, patients who underwent total knee replacement by senior surgeon (MC) and had pre and post-operative radiographies available were eligible for the study. After considering inclusion and exclusion criteria, 147 patients were included in the study. 101 (68.7%) patients had pre-operative coronal varus alignment and the average of pre-operative flexion contracture was 8.6 ± 4.1°. All patients had optimal knee coronal and sagittal balance in flexion and extension at the time of surgery according to the operating surgeon. Table 2 summarizes demographic and intraoperative data. Ninety-three (63.2%) patients had osteophytes on the posterior aspect of the femur which were completely removed and the average thickness measured on sagittal radiographs was 8.1 ± 3.8 mm. Fifty-four patients (36.8%) did not have radiological evidence of osteophytes on the posterior aspect of the femur. Twenty-two (15%) patients were classified as having Grade 1 osteophytes, 47 (32%) patients classified as having Grade 2 osteophytes and 24 (16.3%) patients were classified as having Grade 3 osteophytes. The average sagittal osteophyte thickness for patients classified as Grade 1 was 6.3 ± 1.9 mm, 12.0 ± 1.34 mm for Grade 2 and 19.3 ± 4.0 for Grade 3. On analysis of the reliability of the method of measuring the osteophytes on sagittal radiographs, we found an intra-observer ICC of 0.994 (IC95 0.983 - 0.997) and inter-observer ICC of 0.997 (IC95 0.991 - 0.998).

Table 2.

Demographics and intraoperative data.

	Frequency (percentage)/Mean (SD or range)
Gender
Female	93 (63.3%)
Age	64.4 ± 4.2
Right side	87 (59.2%)
Pre-operative Flexion Contracture (degrees)	8.57 ± 4.1
Post-operative Extension (degrees)	0.21 ± 2.5
Sagittal Correction (degrees)	8.19 ± 4.1
Pre-operative coronal deformity (degrees) a
Varus alignment	101 (68.7%)
Mean deformity	- 4.2 (−9.5 −0.5)
Valgus deformity	46 (32.3%)
Mean deformity	2.9 (7–0.5)
Post-operative coronal deformity	.03 (3–−2.5)
Soft tissue release	77 (52.0%)
Minimal medial	60 (40.5%)
Osteophytes removed	93 (63.2%)
Osteophyte Thickness (mm)	8.11 ± 3.8
Leie Classification for Osteophyte
Grade 0	54 (36.7%)
Grade 1	22 (15.0%)
Grade 2	47 (32%)
Grade 3	24 (16.3%)

^aNegative values for coronal alignment mean varus deformities.

On analysis of strength and direction of an association that exists between sagittal osteophyte thickness and change in degree of extension at time of surgery, we found a positive and linear correlation (Pearson correlation 0.327, p .005) between these two variables (Fig. 7). On multivariate regression model we assessed how each variable independently contributed to the correction of the flexion contracture at the time of surgery in our patients. Posterior femoral osteophyte removal was a categorical and continuous variable statistically significant in affecting extension at time of surgery (Table 3). When osteophytes grades were individually assessed, we found that only Grade 2 osteophytes (R 0.10, B 2.75, p = .01) and Grade 3 osteophytes(R 0.38, B 4.54, p = .000) seemed to affect extension at the time of surgery. Removing Grade 2 osteophytes (11–15 mm) from the posterior aspect of the femur was responsible for correcting 2.75° of flexion contracture in our model, while Grade 3 osteophytes (>15 mm) removal was responsible for correcting up to 4.5° of flexion contracture at the time of surgery. Using a one-way ANOVA posthoc test to compare change in extension and osteophyte Grade, we found that patients with Grade 0 and Grade 1 osteophytes did not have any statistical difference in change in extension (7.0 ± 3.7° vs 6.7 ± 3.5°, p = .99). However, patients who had Grade 2 and Grade 3 osteophytes had significantly more correction of their flexion deformity time of surgery (8.3 ± 3.13° vs 11.9 ± 4.85°, p = .000 respectively). There was no difference in the frequency of distribution amongst categories of osteophyte size with regards to soft tissue release (p = .922), tibial level of resection (p = .765) tibial slope (p = .793), femoral flexion in the sagittal plane (p = .982) or insert thickness (p = .265).

Fig. 7.

Pearson correlation between Osteophyte sagittal thickness and Change in Extension (Pearson'r = 0.32. p = .005).

Table 3.

Multivariate logistic regression for change in sagittal alignment.

Model	Unstandardized Coefficient B	R	Significance
Change in coronal alignment (degrees)	.06	.06	.19
Femoral Implant Size	-.13	.03	.50
Tibial Implant Size	.01	.00	.90
Insert thickness	-.23	.04	.35
Soft tissue release	-.21	.04	.43
Minimal medial	-.54	.06	.25
Tibial slope (degrees)	-.30	.03	.96
Femoral flexion (degrees)	.12	.00	.96
Osteophyte Release categories (0,1,2,3)	1.31	.36	.00
Grade 1	.70	.08	.30
Grade 2	2.75	.10	.01
Grade 3	4.54	.38	.00
Osteophyte Sagittal Thickness (mm)	.41	.38	.00

4. Discussion

Our results show a positive and linear correlation between the sagittal thickness of osteophytes removed from the posterior aspect of the femur and the degree of correction of flexion contracture at the time of surgery. Furthermore, osteophyte removal from the posterior aspect of the femur was an independent factor in correcting the flexion contracture at the time of a TKA. We found that the removal of large osteophytes (Grade 2 and Grade 3) were associated with a flexion deformity correction of 2.7° and 4.5° respectively. However, the removal of small osteophytes (Grade 1) from the posterior aspect of the femur did not affect knee extension at the time of surgery.

Osteophyte removal is an important principle of soft tissue balancing in the total knee arthroplasty.¹³ Specifically, osteophytes located in the posterior aspect of the femur can cause a tenting effect on the posterior capsule of the knee (Fig. 8) and failure to remove them may result in unnecessary soft-tissue releases or lead to additional bone resection. However, posterior femoral osteophytes can be difficult to remove because they are hard to see, and surgeons may be concerned about passing a sharp instrument posteriorly, near the popliteal neurovascular structures. Despite being relatively rare, vascular complications in TKA may result in a 7% risk of mortality, 42% risk of amputation,¹⁴ increased hospital charges and length of stay.¹⁵ Unnecessary manipulation of the posterior compartment can also result in additional intraoperative bleeding from the posterior genicular arteries. Persistent drainage after TKA, for instance, has been related to a failure in haemostasis or a non-identified source of bleeding from the posterior capsule and genicular arteries¹⁶ and patients requiring an acute return to surgery to manage it had an increased rate of deep infection at 2-year follow-up (6.0%).¹⁷ Galat et al. also found that acute returns to surgery within 30 days to specifically evacuate hematomas following TKA had a 12.3% risk of needing additional major surgery within 2 years and a 10.5% risk of a deep infection.¹⁷ Since we did not find any benefits to extension at the time of surgery with removal of small Grade 1 osteophytes, we do not recommend their removal as this could result in unnecessary bleeding and potentially increase the risk of wound complications with no benefit in knee extension at time of TKA.

Fig. 8.

Magnetic Resonance Imaging evidencing the potential tenting effect caused by posterior femoral osteophytes (red arrows) on the posterior capsule (blue arrows).

Sriphirom P et al. assessed 40 primary varus osteoarthritic knees with posterior condylar osteophytes that underwent nagivated TKA. The authors showed that removing osteophytes with an average of 7.7 ± 5.3 mm sagittal thickness from the posterior aspect of the femur resulted in a small change in the extension gap of 0.64 ± 0.80 mm.⁸ In a different way, Yau WP et al. assessed the relationship between residual osteophytes left on the posterior femoral condyle and knee flexion at 12 months following TKA. The authors found that the most significant independent surgical factors to predict the amount of post-operative flexion was the presence of residual posterior femoral condyle osteophytes.¹⁸ Baldini et al. also measured the flexion and extension gaps in 50 consecutive primary posterior-stabilized knee arthroplasties to determine whether posterior cruciate ligament (PCL) release or posterior condylar osteophyte removal increased the gap width. The authors found that after osteophyte removal, the gaps in extension increased on average 1.8 mm medially and 1.8 mm laterally with respect to the base line value. Flexion increased an average of 2 mm medially and 2.2 mm laterally.¹⁹ Both studies reported only a small effect in increasing the extension gap but they did not check the applicability of their results in terms of degrees of extension acquired intraoperatively or the effect of larger osteophytes removed from the posterior femur. In the current study we may confirm these findings, as we found that small, Grade 1 osteophytes <10 mm (average 6.3 ± 1.9 mm) did not seem to affect extension at the time of surgery of gap-balancing navigated TKA. However, with larger osteophytes (Grade 2 and Grade 3), removal led to a significant improvement in extension at the time of surgery by 2.7 and 4.5° respectively.

Multiple intra-operative strategies are described to help achieve full extension in knee arthroplasty, and the most common include changes in the extension gap as key contributors to flexion contracture correction. These include distal femoral resection, soft tissue balancing and femoral component placement in flexion.⁴,⁵,⁶,⁷ Cross et al. were the first authors who showed that every two millimetres of distal femoral resected resulted in four degrees of the fixed flexion contracture corrected. They called this the “2 by 4” rule.²⁰ Liu DW et al. have also shown that an additional 2 mm of distal femoral resection resulted in an improvement of 3.37° in extension, and 4 mm resulted in up to 6.68° of improved extension intraoperatively.²¹ However, extra distal bone resection is not an attractive option for every surgeon, as it can potentially change the tibiofemoral joint line, impacting the patellofemoral joint and knee stability²² as well as reducing knee scores when the joint line is raised above a certain level.²³ Particular in our study we included only patients who had the joint line preserved in relation to the native joint line and did not have posterior condylar offset changed significantly in order to assess the effect of the posterior femoral osteophyte removal in isolation.

To our best knowledge, there is no reliable classification system to grade osteophyte size on sagittal radiography. Altman et al. assessed individual radiographs of osteoarthritic knees, hands, hips and classified the osteophytes from grade 0–3 based on degrees of change (i.e. 0-normal, 3 -severe change).²⁴ However, the authors did not include the posterior aspect of the femur nor the size or sagittal thickness of the osteophytes. With our proposed classification system, we found excellent intra and inter-class correlation (ICC 0.994 and ICC 0.997 respectively) in grading osteophytes exclusively on the posterior aspect of the femur. Our classification considers only the sagittal thickness of the osteophytes and a three-dimensional assessment with magnetic resonance image (MRI) or computed tomography scan (CT) may be more appropriate to assess the whole volume of the osteophytes. However, the purpose study was to suggest a classification system applicable for routine pre-operative assessment and radiographs are the most common imaging modality for pre operative planning in total knee replacements. We also did not intra-operatively measure osteophytes with a calliper because they are usually fragmented when the surgeon removes them and sizing fragments would make the method unreliable and not practical. As part of our inclusion criteria, only patients with no femoral malrotation (less than 5 mm) on lateral radiography (Fig. 1) and those patients whose post-operative radiographs evidenced complete removal of femoral osteophytes in comparison to pre-operative x-rays were included.

There are several other factors that contribute to correction of flexion contractures in a gap-balancing total knee replacement. Lustig et al. found that sagittal alignment of greater than 3.5° from the mechanical axis was shown to increase the relative risk of a mild flexion contracture at one-year follow-up by 2.9 times, independent of other variables.⁷ Nowakowski et al. on the other hand also showed that increasing the tibial slope significantly widened both the extension and the flexion gaps.²⁵ Kim et al. also found that releasing medial osteophytes, the deep medial collateral ligament and semimembranosus were responsible for a change in flexion contracture for up to 5.2° ± 2.8°.¹ The role of soft-tissue structures in the treatment of fixed flexion contractures is surprisingly poorly described. Asano et al. found that increasing “soft tissue tension” of the knee in extension correlated with increasing the flexion contracture. “Soft tissue tension” was defined as the amount of axially applied distraction force required to open the extension gap to the thickness of the implant.²⁶ However, in our sample those factors did not show any positive effects in intraoperative flexion correction. Once they were evenly distributed, a multivariate logistic regression analysis showed no statistical significance for those factors in our sample.

4.1. Limitations

There are a number of important limitations to our study and results should be carefully interpreted. Firstly, it is a retrospective study and perhaps a prospective intra-operative sequential analysis of extension before and after removing the osteophytes would be more appropriate. However, in our multivariate regression model, the unstandardized coefficients indicated how much the dependent variable (change in knee extension) varied until time zero of TKA with an independent variable (osteophyte sagittal thickness) when all other independent variables were held constant. Secondly, 98 patients were excluded due to low quality of radiographies or when they were not localized, and this ensured that the osteophytes thickness was accurately measured on sagittal radiographs. We only measured the sagittal thickness of osteophytes and perhaps CT scan or MRI assessment would be more accurate to measure the three dimensional volume of the osteophytes. As part of a gap-balancing TKA, the majority of our patients had minimal soft tissue release performed. These mostly involved minimal pie-crusting of the superficial medial collateral ligament and we excluded patents who had extensive releases that could interfere on the results. These results are only applicable to patients who underwent posterior cruciate retaining (CR) arthroplasty and not PCL sacrificing (PS) implants.

The strengths of the study include a large sample size to detect a large effect size and the final model with three predictors had a power (1-β) = 0.98. Secondly, surgeries were performed by single surgeon who also measured and accurately recorded every measure of femoral and tibial bone resected over the last ten years and the method to assess sagittal alignment has been the same for all patients. To our knowledge, this is also the first study that reports a clinical improvement in intraoperative flexion contracture in relation to the removal of posterior femoral osteophytes. Our system of osteophyte classification also showed excellent reproducibility, with an intra-observer ICC of 0.994 (IC95 0.983 - 0.997) and inter-observer ICC of 0.997 (IC95 0.991 - 0.998).

5. Conclusion

Based on the results of this study, the data supports that osteophyte removal from the posterior aspect of the femur was an independent variable affecting knee extension at the time of surgery in a gap-balancing TKA. Furthermore, this is dependent on the thickness of the osteophytes removed. Removing small Grade 1 osteophytes should not be expected to affect extension at the time of surgery in TKA and could increase intra-operative time and morbidity. However, large Grade 2 and Grade 3 osteophyte removal can be used, in combination with other surgical techniques to ensure that intraoperative extension is optimal.

Declaration of competing interest

The Authors declare that there are no conflicts of interest.