Distal femoral valgus cut angles unreliable in total knee arthroplasty

Abstract

We hypothesized that valgus distal femoral cut angle made using a conventional cutting guide would be reproducible in a Sawbone model, regardless of training level. 3°, 5°, or 7° valgus cuts were made on lower extremity Sawbone specimens and were measured with radiographic imaging. 66 patient radiographs were also analyzed to compare pre and post-operative femoral cut angles, and VR12 measurements from each patient were collected. All femoral cuts deviated significantly from target cuts. Also, pre-TKA valgus angles showed no correlation with the angles post-TKA, and final cut angle did not correlate with functional outcomes at 1 year post-surgery.

1. Introduction

Over one million total knee arthroplasties are expected to be done annually in the United States, with surgical volume growing quickly due to the aging population, increasing body mass index, and higher level of activity at an older age. The primary goals of total knee arthroplasty (TKA) are the relief of pain for the patient and restoration of mobility and function. Functional outcomes are at least in part dependent on soft tissue balance and optimized kinematics, while also taking into account preoperative anatomy and degree of deformity based on radiological images.¹ Surgeons must navigate a variety of balancing techniques, implant designs, and surgical approaches with no clear consensus on the ideal technique or technology. Malalignment of the knee places patients at an increased risk of implant failure and revision surgery.²

In 1983, Brargen et al. found that using a slight valgus cut angle allows for the restoration of limb alignment of the knee and produces the lowest failure rate when compared to neutral or varus alignment.³ More recently, Kharwandkar et al. suggested that using a cut of 5°–6° valgus is safe for uncomplicated primary TKAs.⁴ While these measurements may provide good results for patients of average height and build, some studies have shown that cut angles should be adjusted for patients with coxa valga, coxa vara, or other hip deformities.⁵ Personalized instrumentation and robotic surgery attempt to establish normal patient anatomy with custom cut angles. As such, modern TKA techniques usually suggest adjusting the femoral cut angle to suit individual patients, with most surgeons ranging between 3 and 7° of valgus.

Personalized instrumentation, patient specific total knee designs, or robotic techniques have made the argument for customization of cutting angles and re-establishment of “normal” anatomy.⁶ Multiple studies have found that patient specific guides improve limb alignment and reduce the amount of outliers when compared to manual “old school” alignment techniques.^6,7 While such studies support the use of patient specific instrumentation, other research shows no difference in long term outcomes when using manual intramedullary guides versus patient specific cutting instrumentation.^8,9 Additionally, while robotic surgery techniques have been shown to significantly improve prosthesis fixation and more accurate limb reconstruction in TKA in some smaller studies, there are very limited high-quality randomized studies to determine the true outcomes of this method.¹⁰ Overall, the ideal techniques and alignment goals for TKA remain controversial.

We hypothesize that the majority of techniques lack the precision to reproduce desired cut angles leading to high surgical variability. The accuracy of a standard intramedullary distal femoral cutting jig was evaluated in a sawbone model, as well as in surgery. Ultimately, choosing 3°, 5°, and 7° distal femoral valgus cut likely has limited impact on post-operative alignment.

2. Methods

This is a level III, retrospective cohort study, along with a Sawbone model analysis. To test the hypothesis that attempting to cut specific distal femoral angle has no impact on the actual final alignment, both a saw bone model and a patient clinical radiographic follow up was undertaken. Forty-five femur sawbone models were obtained and mounted in a vice for cutting. The models included the soft tissues and thigh in order to mimic the clinical situation by limiting the ability to reference the axis of the femur from the diaphysis. Three different surgeons of different training levels were selected in order to assess any difference in outcome based on expertise and experience. Each surgeon was classified at a different level of experience from junior resident (Post-Graduate Year 2), senior resident (Post-Graduate Year 4), to a fellowship trained adult reconstruction surgeon (attending). The lower extremity sawbone models were separated into 3 groups of 5 samples based on target cut angle, with 15 models allocated for each surgeon.

Each surgeon identified the starting point according to standard guidelines, in the middle of the trochlea approximately 10 mm anterior to the origin of the posterior cruciate ligament. An intramedullary rod was then inserted up to the isthmus followed by a standard variable angle cutting jig (Fig. 1, Courtesy of J&J DePuy Synthes Orthopedics). Each surgeon then performed a series of five cuts at each 3°, 5°, and 7° valgus angles. After the cuts were made, these final valgus cuts were measured with standard goniometer. In addition, all specimens were taken to radiography with a standard anterior to posterior x-ray taken. The true cut angle was measured based on a line along the axis of the femur and another line parallel to the cut surface. Each cut angle was measured twice and averaged for statistical comparisons to the predicted cut angle and based on level of surgical training. Inter-observer and intra-observer correlation and variability was further assessed.

Fig. 1.

Standard distal femoral cutting jig allowing one degree adjustments including 3, 5, and 7° of valgus for either left or right knee as well as cutting depth adjustment. (Courtesy of J&J DePuy Synthes Orthopedics).

In an IRB approved study of retrospective review of implant characteristics in total joint arthroplasty, 132 total radiological images from 66 consecutive patients were analyzed preoperatively and postoperatively. All patients had a standard total knee arthroplasty with the distal femoral cut set at 5° of valgus using an intramedullary referencing jig. Similar to the sawbone component of the study, anterior to posterior radiographs of the total knee were evaluated for distal femoral valgus angle by measuring the angle formed based on a line along the axis of the femur and another line parallel to the cut surface (Fig. 2). The preoperative and postoperative valgus angle measurements were compared using a simple linear regression model. The 12-month physical component score of VR12 was then reviewed to evaluate for any correlation between cut angle and functional outcomes. Based on expected difference between samples, the study was powered for 80%. Standard descriptive statistics were obtained using STATA and the Analysis ToolPak of Microsoft Excel.

Fig. 2.

Distal femoral cut angle on x-ray. An anteroposterior (AP) x-ray was used at the first post-operative visit. The distal femoral cut angle was measured by using the Cobb tool formed between the anatomic axis of the femur and the line parallel to the joint. The femoral axis was determined by two points, proximal circle centering on the femoral diaphysis and distal circle centered on the trochlea. The joint line was determined by 2 points along the most distal aspect of medial and lateral femoral condyle. As such, the distal femoral cut angle in this image is (90–85.38) = 4.62° of valgus.

3. Results

Significant variability was observed across all cases. Starting point is likely the source of much of the variability as well as the estimation of the femoral axis. The intramedullary guide is rigid but significantly smaller than the canal space, thus allowing several degrees of error upon insertion.

All femoral cuts deviated significantly (p < 0.05) from the target cut angles of 3°, 5°, and 7° of valgus (Table 1). Interestingly, the true cut angles tended to underestimate the target cut adjustment from the distal femoral jig. The 7° cut had most accuracy, with the 3° and 5° cuts showing higher variation. At a target femoral cut of 3°, all surgeons averaged a femoral valgus angle of 2.06 ± 0.98°. At a target femoral cut of 5°, the surgeons averaged a femoral valves angle of 3.60 ± 1.01°. At a target femoral cut of 7°, all surgeons averaged a femoral valves angle of 4.82 ± 0.65°. Nevertheless, no significant differences were observed between the 3° and 5° as well as no significant differences between 5° and 7° cuts (p > 0.05).

Table 1.

All femoral cuts deviated significantly (p < 0.05) from the target cut angles of 3°, 5°, and 7° of valgus.

	Cut Guide Angle	Mean	Difference from Expected	Variability (Range)
Junior	3	1.29	1.71	4.9
Senior	3	3.03	−0.03	5.6
Attending	3	1.86	1.14	3.9
Junior	5	2.8	2.2	6.3
Senior	5	4.36	0.64	5
Attending	5	3.75	1.25	4.7
Junior	7	4.34	2.66	4.6
Senior	7	5.11	1.89	4
Attending	7	4.96	2.04	4.3

At all target cuts, the senior resident's mean cut angle was closest to the target angle showing highest accuracy. The attending surgeon showed second highest accuracy, with worst accuracy seen in the junior resident (Fig. 3). Due to the associated high variability, no statistically significant differences were seen between any of the surgeons. Mean cut angle measurements, differences from expected values, and variances (in the form of a range between lowest and highest measured cut angle for each trial) based on level of experience can be seen in Table 1. In the 3° and 5° subgroups, the attending surgeon had the least variability between cuts, and the senior resident had the least variability in the 7° subgroup. In addition to being farthest from the guide measurement with least accuracy, the junior resident had the most variability in the 5° and 7° sets, as well as the second largest range in the 3° trials.

Fig. 3.

Average distal femoral cut measurements for each surgeon based on level of experience.

Unfortunately, the overall summary of measured angles showed no correlation with the target femoral cuts. Fig. 4 suggests that measured angles scattered uniformly and unpredictably for all surgeons despite level of experience and cut angle guide used.

Fig. 4.

Measured angles scattered uniformly and unpredictably for all surgeons despite level of experience and cut angle guide used.

In a follow up clinical analysis of 66 total knee patients cut at a target 5° distal femoral valgus, the post-operative angles clustered around 3° and 5° with an overall random distribution. Nevertheless, all the cuts were valgus with a wide range from 0 to 8°. In assessing whether the preoperative deformity may drive the postoperative distal femoral angle, no statistically significant correlation was observed. The average valgus angle before surgery was 7.05° ±5.19°, while after surgery the distal femoral cut angle was 3.15° ±4.32°. Fig. 5 shows a weak trend between pre- and post-TKA valgus angle measurements (R = 0.133). To evaluate whether that variation influenced post-operative outcomes, the 12-month physical component score of VR12 was then reviewed with no significant correlation to the post-operative femoral distal cut (p > 0.05).

Fig. 5.

Weak trend between pre- and post-TKA valgus angle measurements (R = 0.133).

4. Discussion

Most orthopaedic surgeons attempt to take into account all varying factors that can affect the mechanical alignment during TKA. Height, sex, race, quality of bone, and the preoperative deformity of the knee are all considered when making cuts and balancing the knee in order to produce the best functional results for patients.^1,3, 4, 5 In order to improve on visual measurements, progressively advanced instrumentation has been introduced with several generations of improving guide and surgical solutions. Lately, the argument has been made for patient specific instrumentation and robotic navigation in order to improve the accuracy and predictability of post-operative anatomy. The accuracy of such systems may be superior to the traditional intramedullary cutting guides, but the overall clinical data has been controversial and remains the subject of debate.6, 7, 8, 9, 10 Even though the current study did not attempt to evaluate clinical measures, the data supported that there was no significant difference between various distal femoral target angles of 3°, 5°, and 7° when using traditional intramedullary cutting guides or jigs. Current guides allow for too much variation stemming from a mismatch of intramedullary guides and canal diameter, large tolerances of the blade cutting slot, differences in starting point, as well as quality of bone and patient related factors (soft tissue envelope, femoral bow, femur length, etc.).

The current data suggest that the achieved distal femoral cut angles are highly variable across different attempted target angles. Classically surgeons will cut a knee in significant preoperative valgus at 7°, while the more varus knee is cut at 3° of overall valgus. All cuts achieved a valgus angle, but the final degree of valgus was relatively unpredictable. When analyzing 3° and 7° cuts independently, they were significantly different but not clinically relevant. Furthermore, even though training and experience level may improve accuracy and reproducibility, significant variability was observed even in specialized arthroplasty surgeons. The final cut angles were scattered uniformly for all surgeons despite level of experience and cut angle guide used.

Clinical situations are often different to sawbones models due to variable anatomy, bone density, and soft tissue tension. A review of total knees always cut at 5° of valgus showed high variability of postoperative distal femoral cut angle. Furthermore, the postoperative angle did not correlate with preoperative distal femoral angle, thus theoretically independent of associated deformity. Nevertheless that may not be completely true considering that deformity usually derives from both the tibia and femur and is often perceived in at least two planes or more. Surgeons often focus on overall knee balance rather than distal femoral cut, and thus account and coordinate the tibial cut together with the femoral cut as well as medial and lateral collateral tension. Whether using a gap balancing, measured resection, or kinematic technique, the overall femoral cut is likely less relevant to the final balance and stability of the knee.

While this study did provide insight into the accuracy of cutting guides across varying angles and experience levels, there were also significant limitations. The sawbone analysis was performed on an isolated model of the lower extremity consisting only of a bone model and foam coverage. While the model was an accurate representation of the bone anatomy of an average patient, it lacks the deformity, approach variability, or true bone density of an arthritis knee. A trend towards minimally invasive approaches leads to decreased visualization and likely less precise assessment of the true femoral axis. Another significant limitation to the current study involves using a single system (DePuy Synthes PFC Sigma) intramedullary distal femoral cutting guide. While this is a widely used measurement and cutting tool comparable to other systems, variations in tolerances, intramedullary rod design or flexibility, and play in the assembly may produce different results. Finally, even though statistically powered, the clinical analysis involves a single surgeon in his early career and 1-year follow up. Variations from standard mechanical axis, even though balanced in motion, may lead to increased shear stresses and early implant failure. Extended follow up may increase the trends to allow for statistical significance and additional data on survival.

Considering the associated variability and lack of impact on clinical outcomes, using a fixed 5° cutting guide would simplify the system, remove additional play in the guides, and provide an appropriate target with consistent outcomes. Current results are well supported by previous studies looking at knee alignment and cutting techniques. Overall, various studies on radiographic, MRI, and CT images showed significant variability between valgus correction angle (VCA), individual hip-knee shaft (HKS) angle, mechanical femoral angle (FMA), and epicondylar axis, (EA) and posterior condylar axis (PCA) measurements, before and after surgery, advocating that patient specific cutting procedures.11, 12, 13, 14 However, other studies have shown that fixed angle distal femoral cut angles work well in the vast majority of patients, suggesting that fixed angle systems work well for uncomplicated patients and when personalized equipment is not available.¹⁵

5. Conclusion

Overall, we found that level of experience in joint replacement surgery does improve the accuracy and reliability of the distal femoral cut when using a standard intramedullary cutting guide. However, even the most experienced surgeon still had significant variations in cut angles. The data from this study was scattered uniformly and unpredictable for all surgeons despite level of experience and cut angle guide used. In addition, the valgus angle pre-TKA showed no correlation with the valgus angle post-TKA, and the final distal femoral cut angle did not correlate with patient functional outcomes at 1 year post-surgery.