KNEE
A comparison of patient-specific and conventional instrumentation for total knee arthroplasty A MULTICENTRE RANDOMISED CONTROLLED TRIAL L. Abane, P. Anract, S. Boisgard, S. Descamps, J. P. Courpied, M. Hamadouche From Centre Hospitalier Régional Universitaire de Clermont Ferrand, Clermont Ferrand, France
L. Abane, MD, Orthopaedic Surgeon, Department of Orthopaedic and Reconstructive Surgery P. Anract, MD, Professor of Orthopaedic Surgery, Department of Orthopaedic and Reconstructive Surgery J. P. Courpied, MD, Professor of Orthopaedic Surgery, Department of Orthopaedic and Reconstructive Surgery M. Hamadouche, MD, PhD, Professor of Orthoapedic Surgery, Department of Orthopaedic and Reconstructive Surgery Hôpital Cochin, APHP, Université Paris 5, 27 Rue du Faubourg St Jacques, 75014, Paris, France. S. Boisgard, MD, PhD, Professor of Orthopaedic Surgery, Department of Orthopaedics and Traumatology S. Descamps, MD, Professor of Orthopaedic Surgery, Department of Orthopaedics and Traumatology Centre Hospitalier Régional Universitaire de Clermont Ferrand, 58 Boulevard de Montalembert, 63003, Clermont Ferrand, France. Correspondence should be sent to Professor M. Hamadouche; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B1. 34440 $2.00 Bone Joint J 2015;97-B:56–63. Received 18 May 2014; Accepted after revision 14 October 2014
56
In this study we randomised 140 patients who were due to undergo primary total knee arthroplasty (TKA) to have the procedure performed using either patient-specific cutting guides (PSCG) or conventional instrumentation (CI). The primary outcome measure was the mechanical axis, as measured at three months on a standing long-leg radiograph by the hip–knee–ankle (HKA) angle. This was undertaken by an independent observer who was blinded to the instrumentation. Secondary outcome measures were component positioning, operating time, Knee Society and Oxford knee scores, blood loss and length of hospital stay. A total of 126 patients (67 in the CI group and 59 in the PSCG group) had complete clinical and radiological data. There were 88 females and 52 males with a mean age of 69.3 years (47 to 84) and a mean BMI of 28.6 kg/m2 (20.2 to 40.8). The mean HKA angle was 178.9° (172.5 to 183.4) in the CI group and 178.2° (172.4 to 183.4) in the PSCG group (p = 0.34). Outliers were identified in 22 of 67 knees (32.8%) in the CI group and 19 of 59 knees (32.2%) in the PSCG group (p = 0.99). There was no significant difference in the clinical results (p = 0.95 and 0.59, respectively). Operating time, blood loss and length of hospital stay were not significantly reduced (p = 0.09, 0.58 and 0.50, respectively) when using PSCG. The use of PSCG in primary TKA did not reduce the proportion of outliers as measured by post-operative coronal alignment. Cite this article: Bone Joint J 2015;97-B:56–63.
Total knee arthroplasty (TKA) is a highly effective procedure which gives a good functional result1 and survival rates are greater than 90% in the major national joint registries at 15 years’ follow-up.2,3 Although still a matter of debate in the literature,4,5 there are substantial data indicating that malalignment of the components in the coronal plane is a major cause of failure and revision related to wear, loosening and instability.6,7 After more than a decade of clinical use, TKA using computer-assisted surgery (CAS) has been shown to improve both the accuracy and the precision of lower-limb alignment and positioning of the components, thereby reducing the rate of outliers.8,9 However, CAS in TKA has not been universally adopted to date owing to its cost, learning curve, additional surgical time and procedure-specific complications.10-12 As an alternative, patient-specific cutting guides (PSCG) have been developed. These are custom-made blocks for the distal femur and proximal tibia and are designed from pre-operative three-dimensional (3D) CT or MRIbased reconstructions of the knee. They allow pin placement for conventional cutting guides
or have predesigned slots to achieve a neutral post-operative mechanical axis. The theoretical advantages of PSCG over conventional instrumentation (CI) are improved accuracy and a reduction in the proportion of outliers, reduced operating time and lower costs through improved management of resources.13 However, to the best of the authors’ knowledge, the evidence to support this theory is limited, and the published data comparing PSCG and CI in primary TKA vary in terms of the actual advantages of PSCG.13-17 We therefore performed a prospective randomised multicentre trial to compare the clinical and radiographic results and consumption of resources for primary TKA using patient-specific and conventional instrumentation. We hypothesised that the use of patientspecific instrumentation would reduce the proportion of outliers with an unacceptable mechanical axis in the coronal plane.
Patients and Methods Between September 2011 and March 2012, 140 knees in 140 patients with primary or secondary osteoarthritis (OA) of the knee who THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
Enrolment
57
Assessed for eligibility (n = 183) Excluded (n = 43) • Not meeting incusion criteria (n = 27) • Decline to participate (n = 16) • Other reasons (n = 10)
Patients-specific cutting guides
Randomised (n = 140)
Conventional
Allocation Allocated to intervention (n = 70) • Received allocated intervention (n = 70) • Did not receive allocatd intervention (n = 0)
Allocated to intervention (n = 70) • Received allocated intervention (n = 70) • Did not receive allocated intervention (n = 0)
Follow-up Lost to follow-up (n = 2) Patients not available despite all efforts
Lost to follow-up (n = 0) Discontinued intervention (n = 0)
Discontinued intervention (n = 0) Analysis Analysed (n = 59) Excluded from analysis (n = 11) • Sepsis (n = 1) • Lost to follow-up (n = 2) • Inadequate or absent standing leg-length radiograph (n = 8)
Analysed (n = 67) Excluded from analysis (n = 3) • Sepsis (n = 1) • Inadequate or absent standing leg-length radiograph (n = 1)
Fig. 1 Consolidated standards of reporting trials (CONSORT) flow diagram.
needed TKA were enrolled in the study at two tertiary centres (Fig. 1). The inclusion criteria were age between 18 and 85 years, the ability to understand the information given and entitlement to social security. The exclusion criteria were active or suspected sepsis, tumour around the knee, a previous TKA, the presence of hardware that could cause artefacts on MRI, any contraindication to MRI, any extraarticular deformity that required osteotomy around the knee in conjunction with TKA and social circumstance which could impair follow-up. In all 88 women and 52 men were enrolled. Their mean age was 69.3 years (47 to 84), with a mean height of 165.9 cm (150 to 189), a mean weight of 79.4 kg (46 to 135) and a mean body mass index (BMI) of 28.6 kg/m2 (20.2 to 40.8). There were 70 knees in each group. Surgical technique. All patients in both groups had a Genesis II (Smith & Nephew, Memphis, Tennessee) posteriorstabilised fixed-bearing all-cemented TKA with patellar resurfacing. The aim was to achieve a neutral mechanical VOL. 97-B, No. 1, JANUARY 2015
axis of the lower limb, with the femoral and tibial components aligned perpendicular to the mechanical axes of the respective bones. In the sagittal plane, the aim was to place the tibial component perpendicular to the mechanical axis of the tibia, and the femoral component in 3° of flexion with respect to the anatomical axis. The operations were performed by six experienced lower limb arthroplasty surgeons (SB, PA, SD, JPC, JPL, MH). Each had performed at least ten TKAs using the PSCG system, and was familiar with the Genesis II TKA. The use of a tourniquet and any additional intra-operative releases were performed at the discretion of the individual surgeon. For the PSCG group six weeks prior to their operation they underwent MR imaging and a standing full-length anteroposterior (AP) radiograph was obtained as described in the Visionaire system (Smith & Nephew) manufacturing protocol. The images were uploaded and sent to the manufacturer for processing. From these, a preliminary surgical plan was formulated according to the preference of the
58
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
Fig. 3a
Fig. 3b
Radiographs showing a) measurement of the angle of tibial slope between the tangent and the tibial baseplate and the posterior cortex of the tibia and b) measurement of the flexion angle of the femoral component between a tangent to the anterior plane of the femoral component and the anatomical axis of the femur. Fig. 2a
Fig. 2b
Radiographs showing a) measurements of the hip-knee-ankle (HKA) angle between the mechanical axis of the femur and that of the tibia and b) measurement of the F angle between the tangent and the distal condyles of the prosthesis and the mechanical axis of the femur, and the T angle between the tangent and the prosthetic tibial baseplate and the mechanical axis of the tibia.
individual surgeon. This included the level of resection, and the alignment and size of the femoral and tibial components. This plan was reviewed and approved electronically by the operating surgeon prior to manufacture of the patient-specific cutting blocks. In the CI group, intramedullary guidance was used for both the tibial and the femoral preparation. On the femoral side the intramedullary rod was set at 5° to 7° of valgus, depending on the pre-operative hip–knee–shaft (HKS) angle between the anatomical and mechanical femoral axes as measured on the pre-operative AP standing radiograph, in order to achieve a distal femoral cut that was perpendicular to the mechanical axis of the femur. On the tibial side, the proximal tibia was cut perpendicular to the mechanical axis of the tibia and set at 0° of posterior slope. The femoral canal was occluded with a bone plug before cementing the femoral component. In the PSCG group, the cutting blocks were used for distal femoral and proximal tibial cuts, whereas the rest of the surgical procedure, including the anterior, posterior and chamfer femoral cuts, was carried out using standard instruments. All other elements, including operative preparation, wound closure and post-operative regime, were identical in the two groups. The tourniquet was released to secure haemostasis before closure of the skin. Drains were used in all patients. Prophylactic anticoagulation was performed with low molecular weight heparin (LMWH;
enoxaparin, 40 mg/day), which was continued for six weeks after surgery. Patients were mobilised on the first post-operative day using continuous passive movement exercises, and full weight-bearing was allowed as tolerated. Outcomes. The mechanical axis of the lower limb was measured after three months from digitised bipedal standing AP radiographs with the patella oriented to face directly forwards,18 using Roman radiological measurement software version 1.70 (Institute of Orthopaedics, Robert Jones and Agnes Hunt Hospital, Oswestry, United Kingdom). It was calculated from the HKA angle between the mechanical axes of the femur and tibia (Fig. 2a). An HKA angle > 180° was indicative of valgus, whereas values < 180° indicated varus. Also, the position of the components in the coronal plane was measured on both the tibial and femoral sides by the femoral (F) angle and the tibial (T) angle (Fig. 2b). A T or F angle > 90° was indicative of valgus of the respective component, whereas values < 90° indicated varus. In addition, in the sagittal plane, tibial slope (TS) and femoral component flexion (FF) angle were measured on a lateral view of the knee (Fig. 3). For the HKA, F, T, TS and FF angles, outliers were defined as a deviation > 3° from the pre-operative plan. These measurements were carried out by a single independent observer (LA) who was not involved in the surgical procedures, and who was blinded to the instrumentation used. The reproducibility of the measurements was evaluated from 20 double examinations performed in a random order. The absolute mean value ± 2.8 standard deviation (SD) of the differences was calculated. The reproducibility (99% confidence intervals (CI)) was 0.4°, 1.3°, 1.0°, 0.9° and 1.1° for the HKA, F, T TS and FF angles, respectively. The learning curve for the PSCG procedure was evaluated by correlating the post-operative HKA angle and the date of surgery. THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
59
Table I. Pre-operative data of the patients in the conventional (CI) and patient-specific cutting guides (PSCG) groups Variables*
CI group (n = 70)
PSCG group (n = 70)
p-value†
Age at surgery (yrs) Gender ratio (F/M) BMI (kg/m2) Knee Society global score Oxford Knee score Pre-operative HKA (°)
70.4 (54 to 83) 0.63 28.6 (20 to 40) 92.6 (57 to 134) 18.4 (8 to 37) 174.1 (170 to 194)
67.8 (47 to 84) 0.73 28.8 (20 to 40) 96.8 (56 to 144) 19.1 (7 to 35) 176.1 (169 to 192)
0.16† 0.43‡ 0.64† 0.29† 0.46† 0.86†
* Quantitative data expressed as mean (ranges) † Fisher's exact probability test ‡ Mann–Whitney U test HKA, hip-knee-ankle alignment
Table II. Post-operative radiological data of the patients in the conventional instruments (CI) and patient-specific cutting guides (PSCG) groups Variables*
CI group (n = 67)
PSCG group (n = 59)
p-value‡
HKA angle F angle T angle TS FF angle
178.9 (172.5 to 186.2) 90.1 (84.9 to 95.8) 89.1 (84.0 to 93.6) 0.9 (–4.2 to 4.8) 2.8 (–4.3 to 6.6)
178.2 (172.4 to 183.4) 88.6 (94.3 to 93.1) 89.8 (84.4 to 93.7) 0.5 (–5.1 to 5.2) 4.0 (–3.2 to 5.9)
0.34† 0.0005 0.03 0.95 0.65
* Quantitative data expressed as mean (ranges) † Fisher's exact probability test ‡ Mann–Whitney test HKA, hip-knee-ankle alignment; F, femoral component mechanical angle; T, tibial component mechanical angle; TS, tibial slope; FF, flexion angle of the femoral component
The sizes of the femoral and tibial components and polyethylene insert determined pre-operatively were compared with those of the implanted components to evaluate the predictive value of pre-operative planning in the PSCG group. We also assessed blood loss according to the formula of Mercuriali and Inghilleri;19 operating time from skin incision to wound closure; Knee Society (KS) knee, function and global scores;20 Oxford Knee Score (OKS; 0 to 48 worst to best);21 post-operative complications and length of hospital stay. Statistical analysis. We carried out a power analysis based on the mechanical axis of the lower limb defined by the HKA angle. We calculated the number of patients needed, assuming that 65%12 of the patients in the CI group would be aligned to within 3° of neutral (180°; SD 3) compared with 90% in the PSCG group. With a two-sided risk (α) of 2.5% and a power (1 – β) of 90%, we calculated that 64 patients would be needed in each group. Estimating that 10% of the patients would have incomplete data, we decided to include 70 patients in each arm of the study. The patients were randomised using a computer-generated block randomisation scheme based on the order of presentation, so that each patient was randomised individually regardless of the severity of their OA and deformity. To avoid surgeon bias, randomisation was stratified by the surgeon. All patients gave informed consent, and the study had approval from our ethics committee. The trial was registered at ClinicalTrials.gov (NCT 02002624). VOL. 97-B, No. 1, JANUARY 2015
Continuous variables were expressed as means and ranges. Statistical analysis was performed using nonparametric Mann–Whitney U test or Fisher’s exact probability test, as appropriate. Correlation analysis was performed using Pearson’s coefficient. Statview statistical software version 5.0 (SAS Institute, Cary, North Carolina) was used and with statistical significance defined as a pvalue < 0.05.
Results There was no significant difference in pre-operative evaluation between the two groups (Table I). Of the 140 patients included, 126 had complete clinical and radiological follow-up data (Fig. 1). Two patients in the PSCG group were treated using conventional instrumentation on the tibial side because the PSCG tibial guide fitted poorly at the time of operation. Post-operatively, there was no significant difference in the mean angles measured in the TKAs between the two groups (Table II). There was also no significant difference between the CI and PSCG groups in the number of knees that were outliers (Table III). There was also no significant correlation between the post-operative HKA angle and the date of surgery, indicating no learning curve effect (r2 = 0.009, p = 0.48). The predictive sizing value of the PSCG system was within one size in all the implanted components on both the femoral and the tibial sides (Table IV). The mean
60
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
Table III. Number of post-operative radiological outliers in the conventional instruements (CI) and patient-specific cutting guides (PSCG) groups Variables
CI group (n = 67) (n, %)
PSCG group (n = 59) (n, %)
p-value*
HKA angle F angle T angle TS FF angle
22 (32.8) 12 (17.9) 7 (10.4) 15 (22.4) 21 (31.3)
19 (32.2) 19 (25.4) 7 (11.9) 18 (30.5) 22 (37.3)
0.99 0.41 0.99 0.11 0.44
* Fisher’s Exact probability test HKA, hip-knee-ankle alignment; F, femoral component mechanical angle; T, tibial component mechanical angle; TS, tibial slope; FF, flexion angle of the femoral component
Table IV. Predictive sizing value of pre-operative planning in the patient-specific cutting guides group (implanted size compared with the pre-operatively planned size) Variables
Same size (n, %)
Within one size (n, %)
Within two sizes (n, %)
Femoral component Tibial component Insert thickness
49/70 (70) 50/70 (71.4) 53/70 (75.8)
21/70 (30) 20/70 (28.6) 12/70 (17.1)
0 0 5/70 (7.1)
Table V. Knee Society (KS) and Oxford knee scores (OKS) at three months’ follow-up in the conventional instrumentation (CI) and the patient-specific cutting guides (PSCG) groups Variables*
CI group
PSCG group
p-value†
KS score Knee Function Global OKS
83.3 (34 to 99) 78.3 (10 to 100) 161.7 (79 to 199) 38.2 (21 to 48)
82.5 (51 to 97) 82.1 (40 to 100) 164.7 (98 to 197) 37.9 (18 to 48)
0.20 0.19 0.59 0.95
* Data expressed as mean (ranges) † Mann–Whitney test
operating time was 85.7 minutes (45 to 145) in the CI group and 79.4 minutes (40 to 150) in the PSCG group (Mann–Whitney test, p = 0.09). The mean blood loss was 1358 ml (347 to 3756) in the CI group and 1285 ml (78 to 3643) in the PSCG group (Mann–Whitney test, p = 0.58). The mean length of hospital stay was 8.3 days (6 to 18) in the CI group and 8.1 days (5 to 16) in the PSCG group (Mann–Whitney test, p = 0.50). No significant difference in KS or OKS was seen between the two groups (Table V).
Discussion Lower limb malalignment in the coronal plane after TKA, defined as a deviation > 3° from neutral, has been associated with an increased risk of revision and lower functional scores.6,7,22 Patient-specific instrumentation was introduced to translate 3D planning from the intra-operative to the pre-operative setting. To date, several controlled studies13-17,23-36 have reported conflicting results from PSCG use in primary TKA (Table VI). In addition, both the 3D reconstruction of the knee computer algorithm and the block designs differ substantially from one system to another, which necessitates a specific evaluation of each PSCG system. The Visionaire system is unique, as, in addition to MRI reconstruction of the knee, it includes a fulllength view of the lower limb which provides data on joint laxity: this could make the evaluation of pre-operative
deformity and post-operative alignment more accurate. So far, the results reported for the Visionaire system have lacked the power to allow any strong conclusions to be drawn.28,31,33 We therefore conducted a prospective randomised study with a power of 90% based on a reduction of outliers with PSCG that would be similar to the results obtained with CAS in primary TKA. There were several limitations to our study. First, we did not evaluate the axial alignment of the components. Our study protocol focussed on coronal alignment and did not include post-operative CT, which would have resulted in increased radiation exposure for a relatively large number of patients, and additional cost. Second, it could be argued that radiological measurements were carried out on a standing full-length AP radiograph of the lower limb rather than on CT scans, and that we could not provide accuracy data for our measurements. However, it has recently been shown that this method has a high intra- and inter-observer correlation compared with CT and intra-operative navigation, and limits the patient’s exposure to radiation.37 Also, in our study all measurements were made by the same observer, who was blinded to the instrumentation, with a reproducibility (99% CI) < 1.3° for all measured angles. Third, 14 (10%) of the 140 patients included did not have complete follow-up data, leaving 67 patients in the CI group and 59 in the PSCG group. Our power calculation had been exacting and based THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
61
Table VI. Main published controlled studies comparing patient-specific cutting guides PSCG with conventional instrumentation (CI) in primary total knee arthroplasty Mean post-operative HKA outliers > 3° Mean operating Study
Type of study
Number of knees
Evaluation
HKA angle (°)*
(%)*
time*(mins)
Ng et al13
Retrospective
CI 55/ Signature: 105
181.1/ 180.6
22*/ 9*
NA/ NA
Nunley et al29
Retrospective matched-pairs Retrospective consecutive Retrospective first 100 cases Retrospective matched-pairs Prospective randomised Retrospective
CI: 57/ Signature: 57
Standing long-leg radiograph Scout CT scanogram
NA/ NA
18/ 26
93.4/ 89.6
Nunley et al17 Barrack et al14 Boonen et al24 Boonen et al15 Nam et al27 Roh et al33 Noble et al28 Vunderlinckx et al35 Chareancholvanich et al16 Parratte et al30 Pietsch et al32 Hamilton et al25 Woolson et al36 Victor et al34 Pfitzner et al31 Current study
Prospective randomised Prospective randomised Prospective controlled Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised
CI: 50/ Signature: 50/ Scout CT scanogram OtisMed: 50 CI: 100/ Signature: 100 Scout CT scanogram
179.9/ 180.7/ 182.8
16*/ 18*/ 44*
NA/ NA/ NA
180.5*/ 181.7*
23/ 31
NA/ NA
CI: 40/ Signature: 40
179*/ 181*
50/ 30
61/ 51
178/ 179
18/ 30
50/ 45
179.2/ 179.2
7.3/ 29.3
NA/ NA
180.3/ 179.5
10/ 12
NA/ NA
182.8*/ 181.7*
NA/ NA
128.1*/ 121.4*
Standing long-leg radiograph CI: 90/ Signature: 90 Standing long-leg radiograph CAS: 37/ Signature: 38 Standing long-leg radiograph CI: 50/ Signature: 50 Standing long-leg radiograph CI: 14/ Visionaire: 15 Standing long-leg radiograph CI: 31/ Visionaire: 31 Standing long-leg radiograph CI: 40/ PSI: 40 CT scanogram
2.7†/ 3.0†
NA/ NA
NA/ NA
179.7/ 179.7
7.5/ 2.5
68*/ 62.9*
CI: 40/ PSI: 40
Standing long-leg radiograph NA/ NA
178.3/ 179
10/ 5
NA/ NA
NA/ NA
NA/ NA
90*/ 78*
Standing long-leg radiograph CT scan
180.7/ 180.3
31/ 35
57/ 61
178.7/ 178.3
38/ 41
NA/ NA
Standing long-leg radiograph Standing long-leg radiograph Standing long-leg radiograph
178.1/ 178.6
28.1/ 24.6
NA/ NA
4.5†/ 3.0†/ 1.0†
43/ 30/ 7
76/ 63/ 58
178.9/ 178.2
32.8/ 32.2
85.7/ 79.4
CI: 40/ PSI: 40 CI: 26/ Trumatch: 26 CI: 30/ Trumatch: 33 CI: 64/ Four types PSCG: 64 CI: 30/ Trumatch: 30 Visionaire: 30 CI: 70/ Visionaire: 70
* p < 0.05 † Deviation from 0° post-operative mechanical axis Proprietry patient-specific cutting guides: Signature (Biomet, Warsaw, Indiana); OtisMed (OtisMed Corp, Alameda, California); PSI (Zimmer, Warsaw, Indiana); Trumatch (Depuy, Warsaw, Indiana) NA, not available; CAS, computer-assisted surgery
on 64 patients being included in each group so the study might have been under powered. Fourth, our study focused only on the first three post-operative months, and the data cannot be extrapolated to a longer period of follow-up. Last, our results cannot be generalised to any other commercially available system, as they differ markedly in terms of pre-operative planning and design. Using both the KS and the OKS, we found no significant difference between PSCG and CI in terms of clinical results at three months. Other studies that have compared the short-term clinical results of PSCG and CI23,31,32,35 have had similar outcomes to ours. The mean HKA angle at three months did not differ significantly between the two groups. This is consistent with most published controlled studies (Table VI). It should be emphasised that conventional systems have provided a largely acceptable mean post-operative HKA angle,12-17,24-36 and it should be expected that PSCG would reduce the proportion of outliers, thereby enhancing precision more than accuracy. However, in our study there was no significant difference in VOL. 97-B, No. 1, JANUARY 2015
the proportion of outliers for any measured angle, which refutes our hypothesis. Most of the available literature comparing PSCG and CI concurs with our results,16,24,28,30,32 whereas some showed a non-significant increase in the percentage of outliers using PSCG.14,15,17,29,33 However, apart from our study and that of Boonen et al,15 all the other studies were underpowered for this outcome. There are several explanations for this unexpected finding. In a recent review of rapid prototype surgical guides, Krishnan et al38 found multiple sources of errors in the acquisition of images, the 3D reconstruction protocol and the manufacturing of the guides, all of which can lead to some degree of inaccuracy. Also, the contact points of MRI-based guides rely on cartilage, whereas CT-based guides rely on surrounding bone and osteophytes; both have limitations when it comes to the fitting the guides. The material of the blocks is derived from nylon and has inherent elasticity that can lead to inexact placement; it can also have some mobility during fixation. The saw blade itself can lead to cutting errors, especially in the coronal plane.39 It becomes clear that if, for one
62
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
given case, all possible sources of errors are in the same direction, malalignment > 3° can occur. Some studies that used tri-planar intra-operative computer navigation have shown that these potential errors occur with both tibial and femoral cutting guides.40,41 In these studies, malalignment > 3° would have occurred in 21.7% to 45.5% of the cases if the procedure had been performed using only the PSCG guides. Similarly, some authors have described the need for frequent changes intra-operatively, and the necessity to have each case reviewed in detail by the operating surgeon prior to electronic approval of the proposed pre-operative surgical plan.42,43 We found the system was reasonably able to predict the size of implant: 70% of the implanted components were the size predicted pre-operatively, and 100% were within one size. There was more variation with the tibial insert, but PSCG systems in general do not account for joint laxity. These data, and those from Vundelinckx et al,35 will encourage the further development of fully disposable instrumentation, which is currently being undertaken by many companies. Operating time was reduced by a mean of seven minutes, which is in accordance with other reports, in which it varies between six and 18 minutes.14-16,24,28,30,31 Blood loss was not significantly different in the two groups, and nor was length of hospital stay. We conclude that PSCG does not at this point provide better clinical, radiological or resource consumption outcomes. L. Abane: Gathered the data, Analysed the data, Wrote the initial draft, Approved the final manuscript. P. Anract: Contributed to the study design, Study implementation, Approved the final manuscript. S. Boisgard: Contributed to the study design, Study implementation, Refinement of the study protocol, Ensured accuracy of data and analysis, Approved the final manuscript. S. Descamps: Contributed to the study design, Study implementation, Gathered the data, Analysed the data, Approved the final manuscript. J. P. Courpied: Contributed to the study design, Study implementation, Approved the final manuscript. M. Hamadouche: Contributed to the study design, Study implementation, Refinement of the study protocol, Analysed the data, Wrote the initial draft, Ensured accuracy of data and analysis, Approved the final manuscript. The authors thank J. P. Levai MD for contributing patients. One or more authors received benefits from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by A. Ross and first proof edited by G. Scott.
References 1. Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg [Am] 2004;86-A:963–974. 2. No authors listed. Swedish knee arthroplasty register. Annual report 2012. http:// www.myknee.se/pdf/117_SKAR_2012_Engl_1.0.pdf (date last accessed 14 November 2014). 3. No authors listed. The Norwegian arthroplasty register annual report 2010. http:// nrlweb.ihelse.net/eng/Report_2010.pdf (date last accessed 27 October 2014). 4. Parratte S, Pagnano MW, Trousdale RT, Berry DJ. Effect of postoperative mechanical axis alignment on the fifteen-year survival of modern, cemented total knee replacements. J Bone Joint Surg [Am] 2010;92-A:2143–2149. 5. Bellemans J, Colyn W, Vandenneucker H, Victor J. The Chitranjan Ranawat award: is neutral mechanical alignment normal for all patients? The concept of constitutional varus. Clin Orthop Relat Res 2012;470:45–53. 6. Berend ME, Ritter MA, Meding JB, et al. Tibial component failure mechanisms in total knee arthroplasty. Clin Orthop Relat Res 2004;428:26–34.
7. Ritter MA, Davis KE, Meding JB, et al. The effect of alignment and BMI on failure of total knee replacement. J Bone Joint Surg [Am] 2011;93-A:1588–1596. 8. Bathis H, Perlick L, Tingart M, et al. Alignment in total knee arthroplasty. A comparison of computer-assisted surgery with the conventional technique. J Bone Joint Surg [Br] 2004;86-B:682–687. 9. Matziolis G, Krocker D, Weiss U, Tohtz S, Perka C. A prospective, randomized study of computer-assisted and conventional total knee arthroplasty. Three-dimensional evaluation of implant alignment and rotation. J Bone Joint Surg [Am] 2007;89A:236–243. 10. Bauwens K, Matthes G, Wich M, et al. Navigated total knee replacement. A meta-analysis. J Bone Joint Surg [Am] 2007;89-A:261–269. 11. Beldame J, Boisrenoult P, Beaufils P. Pin track induced fractures around computer-assisted TKA. Orthop Traumatol Surg Res 2010;96:249–255. 12. Novak EJ, Silverstein MD, Bozic KJ. The cost-effectiveness of computer-assisted navigation in total knee arthroplasty. J Bone Joint Surg [Am] 2007;89-A:2389–2397. 13. Ng VY, De CLaire JH, Berend KR, Gulick BC, Lombardi AV Jr. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470:99–107. 14. Barrack RL, Ruh EL, Williams BM, et al. Patient specific cutting blocks are currently of no proven value. J Bone Joint Surg [Br] 2012;94(11 Suppl A):95–99. 15. Boonen B, Schotanus MG, Kerens B, et al. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomised controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21:2206–2212. 16. Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. Bone Joint J 2013;95-B:354– 359. 17. Nunley RM, Ellison BS, Zhu J, et al. Do patient-specific guides improve coronal alignment in total knee arthroplasty? Clin Orthop Relat Res 2012;470:895–902. 18. Cooke TD, Sled EA, Scudamore RA. Frontal plane knee alignment: a call for standardized measurement. J Rheumatol 2007;34:1796–1801. 19. Mercuriali F, Inghilleri G. Proposal of an algorithm to help the choice of the best transfusion strategy. Curr Med Res Opin 1996;13:465–478. 20. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989;248:13–14. 21. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg [Br] 1998;80-B:63–69. 22. Huang NF, Dowsey MM, Ee E, et al. Coronal alignment correlates with outcome after total knee arthroplasty: five-year follow-up of a randomized controlled trial. J Arthroplasty 2012;27:1737–1741. 23. Abdel MP, Parratte S, Blanc G, et al. No benefit of patient-specific instrumentation in TKA on functional and gait outcomes: A randomized clinical trial. Clin Orthop Relat Res 2014;472:2468–2476. 24. Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta Orthop 2012;83:387–393. 25. Hamilton WG, Parks NL, Saxena A. Patient-specific instrumentation does not shorten surgical time: a prospective, randomized trial. J Arthroplasty 2013;28(Suppl.1):96–100. 26. Klatt BA, Goyal N, Austin MS, Hozack WJ. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23:26–29. 27. Nam D, Maher PA, Rebolledo BJ, et al. Patient specific cutting guides versus an imageless, computer-assisted surgery system in total knee arthroplasty. Knee 2013;20:263–267. 28. Noble JW, Moore CA, Liu N. The value of patient-matched instrumentation in total knee arthroplasty. J Arthroplasty 2012;27:153–155. 29. Nunley RM, Ellison BS, Ruh EL, et al. Are patient-specific cutting blocks costeffective for total knee arthroplasty? Clin Orthop Relat Res 2012;470:889–894. 30. Parratte S, Blanc G, Boussemart T, et al. Rotation in total knee arthroplasty: no difference between patient-specific and conventional instrumentation. Knee Surg Sports Traumatol Arthrosc 2013;21:2213–2219. 31. Pfitzner T, Abdel MP, von Roth P, Perka C, Hommel H. Small improvements in mechanical axis alignment with MRI versus CT-based patient-specific instrumentsin TKA: A randomized clinical trial. Clin Orthop Relat Res 2014;472:2913–2922. 32. Pietsch M, Djahani O, Zweiger C, et al. Custom-fit minimally invasive total knee arthroplasty: effect on blood loss and early clinical outcomes. Knee Surg Sports Traumatol Arthrosc 2013;21:2234–2240. THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
63
33. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instruments comparable to Conventional TKA? A randomized controlled study of one system. Clin Orthop Relat Res 2013;471:3988–3995.
38. Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg [Br] 2012;94-B:1457–1461.
34. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific guides do not improve accuracy in total knee arthroplasty: a prospective randomized controlled trial. Clin Orthop Relat Res 2014;472:263–271. 35. Vundelinckx BJ, Bruckers L, De Mulder K, De Schepper J, Van Esbroeck G. Functional and radiographic short-term outcome evaluation of the visionaire system, a patient-matched instrumentation system for total knee arthroplasty. J Arthroplasty 2013;28:964–970.
39. Yau WP, Chiu KY. Cutting errors in total knee replacement: assessment by computer assisted surgery. Knee Surg Sports Traumatol Arthrosc 2008;16:670–673. 40. Conteduca F, Iorio R, Mazza D, et al. Evaluation of the accuracy of a patient-specific instrumentation by navigation. Knee Surg Sports Traumatol Arthrosc 2013;21:2194–2199. 41. Lustig S, Scholes CJ, Oussedik SI, et al. Unsatisfactory accuracy as determined by computer navigation of VISIONAIRE patient-specific instrumentation for total knee arthroplasty. J Arthroplasty 2013;28:469–473. 42. Pietsch M, Djahani O, Hochegger M, Plattner F, Hofmann S. Patient-specific total knee arthroplasty: the importance of planning by the surgeon. Knee Surg Sports Traumatol Arthrosc 2013;21:2220–2226. 43. Stronach BM, Pelt CE, Erickson J, Peters CL. Patient-specific Total Knee Arthroplasty Required Frequent Surgeon-directed Changes. Clin Orthop Relat Res 2013;471:169–174.
36. Woolson ST, Harris AH, Wagner DW, Giori NJ. Component alignment during total knee arthroplasty with use of standard or custom instrumentation: a randomized clinical trial using computed tomography for postoperative alignment measurement. J Bone Joint Surg [Am] 2014;96-A:366–372. 37. Babazadeh S, Dowsey MM, Bingham RJ, et al. The long leg radiograph is a reliable method of assessing alignment when compared to computer-assisted navigation and computer tomography. Knee 2013;20:242–249.
VOL. 97-B, No. 1, JANUARY 2015
A comparison of patient-specific and conventional instrumentation for total knee arthroplasty A MULTICENTRE RANDOMISED CONTROLLED TRIAL L. Abane, P. Anract, S. Boisgard, S. Descamps, J. P. Courpied, M. Hamadouche From Centre Hospitalier Régional Universitaire de Clermont Ferrand, Clermont Ferrand, France
L. Abane, MD, Orthopaedic Surgeon, Department of Orthopaedic and Reconstructive Surgery P. Anract, MD, Professor of Orthopaedic Surgery, Department of Orthopaedic and Reconstructive Surgery J. P. Courpied, MD, Professor of Orthopaedic Surgery, Department of Orthopaedic and Reconstructive Surgery M. Hamadouche, MD, PhD, Professor of Orthoapedic Surgery, Department of Orthopaedic and Reconstructive Surgery Hôpital Cochin, APHP, Université Paris 5, 27 Rue du Faubourg St Jacques, 75014, Paris, France. S. Boisgard, MD, PhD, Professor of Orthopaedic Surgery, Department of Orthopaedics and Traumatology S. Descamps, MD, Professor of Orthopaedic Surgery, Department of Orthopaedics and Traumatology Centre Hospitalier Régional Universitaire de Clermont Ferrand, 58 Boulevard de Montalembert, 63003, Clermont Ferrand, France. Correspondence should be sent to Professor M. Hamadouche; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B1. 34440 $2.00 Bone Joint J 2015;97-B:56–63. Received 18 May 2014; Accepted after revision 14 October 2014
56
In this study we randomised 140 patients who were due to undergo primary total knee arthroplasty (TKA) to have the procedure performed using either patient-specific cutting guides (PSCG) or conventional instrumentation (CI). The primary outcome measure was the mechanical axis, as measured at three months on a standing long-leg radiograph by the hip–knee–ankle (HKA) angle. This was undertaken by an independent observer who was blinded to the instrumentation. Secondary outcome measures were component positioning, operating time, Knee Society and Oxford knee scores, blood loss and length of hospital stay. A total of 126 patients (67 in the CI group and 59 in the PSCG group) had complete clinical and radiological data. There were 88 females and 52 males with a mean age of 69.3 years (47 to 84) and a mean BMI of 28.6 kg/m2 (20.2 to 40.8). The mean HKA angle was 178.9° (172.5 to 183.4) in the CI group and 178.2° (172.4 to 183.4) in the PSCG group (p = 0.34). Outliers were identified in 22 of 67 knees (32.8%) in the CI group and 19 of 59 knees (32.2%) in the PSCG group (p = 0.99). There was no significant difference in the clinical results (p = 0.95 and 0.59, respectively). Operating time, blood loss and length of hospital stay were not significantly reduced (p = 0.09, 0.58 and 0.50, respectively) when using PSCG. The use of PSCG in primary TKA did not reduce the proportion of outliers as measured by post-operative coronal alignment. Cite this article: Bone Joint J 2015;97-B:56–63.
Total knee arthroplasty (TKA) is a highly effective procedure which gives a good functional result1 and survival rates are greater than 90% in the major national joint registries at 15 years’ follow-up.2,3 Although still a matter of debate in the literature,4,5 there are substantial data indicating that malalignment of the components in the coronal plane is a major cause of failure and revision related to wear, loosening and instability.6,7 After more than a decade of clinical use, TKA using computer-assisted surgery (CAS) has been shown to improve both the accuracy and the precision of lower-limb alignment and positioning of the components, thereby reducing the rate of outliers.8,9 However, CAS in TKA has not been universally adopted to date owing to its cost, learning curve, additional surgical time and procedure-specific complications.10-12 As an alternative, patient-specific cutting guides (PSCG) have been developed. These are custom-made blocks for the distal femur and proximal tibia and are designed from pre-operative three-dimensional (3D) CT or MRIbased reconstructions of the knee. They allow pin placement for conventional cutting guides
or have predesigned slots to achieve a neutral post-operative mechanical axis. The theoretical advantages of PSCG over conventional instrumentation (CI) are improved accuracy and a reduction in the proportion of outliers, reduced operating time and lower costs through improved management of resources.13 However, to the best of the authors’ knowledge, the evidence to support this theory is limited, and the published data comparing PSCG and CI in primary TKA vary in terms of the actual advantages of PSCG.13-17 We therefore performed a prospective randomised multicentre trial to compare the clinical and radiographic results and consumption of resources for primary TKA using patient-specific and conventional instrumentation. We hypothesised that the use of patientspecific instrumentation would reduce the proportion of outliers with an unacceptable mechanical axis in the coronal plane.
Patients and Methods Between September 2011 and March 2012, 140 knees in 140 patients with primary or secondary osteoarthritis (OA) of the knee who THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
Enrolment
57
Assessed for eligibility (n = 183) Excluded (n = 43) • Not meeting incusion criteria (n = 27) • Decline to participate (n = 16) • Other reasons (n = 10)
Patients-specific cutting guides
Randomised (n = 140)
Conventional
Allocation Allocated to intervention (n = 70) • Received allocated intervention (n = 70) • Did not receive allocatd intervention (n = 0)
Allocated to intervention (n = 70) • Received allocated intervention (n = 70) • Did not receive allocated intervention (n = 0)
Follow-up Lost to follow-up (n = 2) Patients not available despite all efforts
Lost to follow-up (n = 0) Discontinued intervention (n = 0)
Discontinued intervention (n = 0) Analysis Analysed (n = 59) Excluded from analysis (n = 11) • Sepsis (n = 1) • Lost to follow-up (n = 2) • Inadequate or absent standing leg-length radiograph (n = 8)
Analysed (n = 67) Excluded from analysis (n = 3) • Sepsis (n = 1) • Inadequate or absent standing leg-length radiograph (n = 1)
Fig. 1 Consolidated standards of reporting trials (CONSORT) flow diagram.
needed TKA were enrolled in the study at two tertiary centres (Fig. 1). The inclusion criteria were age between 18 and 85 years, the ability to understand the information given and entitlement to social security. The exclusion criteria were active or suspected sepsis, tumour around the knee, a previous TKA, the presence of hardware that could cause artefacts on MRI, any contraindication to MRI, any extraarticular deformity that required osteotomy around the knee in conjunction with TKA and social circumstance which could impair follow-up. In all 88 women and 52 men were enrolled. Their mean age was 69.3 years (47 to 84), with a mean height of 165.9 cm (150 to 189), a mean weight of 79.4 kg (46 to 135) and a mean body mass index (BMI) of 28.6 kg/m2 (20.2 to 40.8). There were 70 knees in each group. Surgical technique. All patients in both groups had a Genesis II (Smith & Nephew, Memphis, Tennessee) posteriorstabilised fixed-bearing all-cemented TKA with patellar resurfacing. The aim was to achieve a neutral mechanical VOL. 97-B, No. 1, JANUARY 2015
axis of the lower limb, with the femoral and tibial components aligned perpendicular to the mechanical axes of the respective bones. In the sagittal plane, the aim was to place the tibial component perpendicular to the mechanical axis of the tibia, and the femoral component in 3° of flexion with respect to the anatomical axis. The operations were performed by six experienced lower limb arthroplasty surgeons (SB, PA, SD, JPC, JPL, MH). Each had performed at least ten TKAs using the PSCG system, and was familiar with the Genesis II TKA. The use of a tourniquet and any additional intra-operative releases were performed at the discretion of the individual surgeon. For the PSCG group six weeks prior to their operation they underwent MR imaging and a standing full-length anteroposterior (AP) radiograph was obtained as described in the Visionaire system (Smith & Nephew) manufacturing protocol. The images were uploaded and sent to the manufacturer for processing. From these, a preliminary surgical plan was formulated according to the preference of the
58
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
Fig. 3a
Fig. 3b
Radiographs showing a) measurement of the angle of tibial slope between the tangent and the tibial baseplate and the posterior cortex of the tibia and b) measurement of the flexion angle of the femoral component between a tangent to the anterior plane of the femoral component and the anatomical axis of the femur. Fig. 2a
Fig. 2b
Radiographs showing a) measurements of the hip-knee-ankle (HKA) angle between the mechanical axis of the femur and that of the tibia and b) measurement of the F angle between the tangent and the distal condyles of the prosthesis and the mechanical axis of the femur, and the T angle between the tangent and the prosthetic tibial baseplate and the mechanical axis of the tibia.
individual surgeon. This included the level of resection, and the alignment and size of the femoral and tibial components. This plan was reviewed and approved electronically by the operating surgeon prior to manufacture of the patient-specific cutting blocks. In the CI group, intramedullary guidance was used for both the tibial and the femoral preparation. On the femoral side the intramedullary rod was set at 5° to 7° of valgus, depending on the pre-operative hip–knee–shaft (HKS) angle between the anatomical and mechanical femoral axes as measured on the pre-operative AP standing radiograph, in order to achieve a distal femoral cut that was perpendicular to the mechanical axis of the femur. On the tibial side, the proximal tibia was cut perpendicular to the mechanical axis of the tibia and set at 0° of posterior slope. The femoral canal was occluded with a bone plug before cementing the femoral component. In the PSCG group, the cutting blocks were used for distal femoral and proximal tibial cuts, whereas the rest of the surgical procedure, including the anterior, posterior and chamfer femoral cuts, was carried out using standard instruments. All other elements, including operative preparation, wound closure and post-operative regime, were identical in the two groups. The tourniquet was released to secure haemostasis before closure of the skin. Drains were used in all patients. Prophylactic anticoagulation was performed with low molecular weight heparin (LMWH;
enoxaparin, 40 mg/day), which was continued for six weeks after surgery. Patients were mobilised on the first post-operative day using continuous passive movement exercises, and full weight-bearing was allowed as tolerated. Outcomes. The mechanical axis of the lower limb was measured after three months from digitised bipedal standing AP radiographs with the patella oriented to face directly forwards,18 using Roman radiological measurement software version 1.70 (Institute of Orthopaedics, Robert Jones and Agnes Hunt Hospital, Oswestry, United Kingdom). It was calculated from the HKA angle between the mechanical axes of the femur and tibia (Fig. 2a). An HKA angle > 180° was indicative of valgus, whereas values < 180° indicated varus. Also, the position of the components in the coronal plane was measured on both the tibial and femoral sides by the femoral (F) angle and the tibial (T) angle (Fig. 2b). A T or F angle > 90° was indicative of valgus of the respective component, whereas values < 90° indicated varus. In addition, in the sagittal plane, tibial slope (TS) and femoral component flexion (FF) angle were measured on a lateral view of the knee (Fig. 3). For the HKA, F, T, TS and FF angles, outliers were defined as a deviation > 3° from the pre-operative plan. These measurements were carried out by a single independent observer (LA) who was not involved in the surgical procedures, and who was blinded to the instrumentation used. The reproducibility of the measurements was evaluated from 20 double examinations performed in a random order. The absolute mean value ± 2.8 standard deviation (SD) of the differences was calculated. The reproducibility (99% confidence intervals (CI)) was 0.4°, 1.3°, 1.0°, 0.9° and 1.1° for the HKA, F, T TS and FF angles, respectively. The learning curve for the PSCG procedure was evaluated by correlating the post-operative HKA angle and the date of surgery. THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
59
Table I. Pre-operative data of the patients in the conventional (CI) and patient-specific cutting guides (PSCG) groups Variables*
CI group (n = 70)
PSCG group (n = 70)
p-value†
Age at surgery (yrs) Gender ratio (F/M) BMI (kg/m2) Knee Society global score Oxford Knee score Pre-operative HKA (°)
70.4 (54 to 83) 0.63 28.6 (20 to 40) 92.6 (57 to 134) 18.4 (8 to 37) 174.1 (170 to 194)
67.8 (47 to 84) 0.73 28.8 (20 to 40) 96.8 (56 to 144) 19.1 (7 to 35) 176.1 (169 to 192)
0.16† 0.43‡ 0.64† 0.29† 0.46† 0.86†
* Quantitative data expressed as mean (ranges) † Fisher's exact probability test ‡ Mann–Whitney U test HKA, hip-knee-ankle alignment
Table II. Post-operative radiological data of the patients in the conventional instruments (CI) and patient-specific cutting guides (PSCG) groups Variables*
CI group (n = 67)
PSCG group (n = 59)
p-value‡
HKA angle F angle T angle TS FF angle
178.9 (172.5 to 186.2) 90.1 (84.9 to 95.8) 89.1 (84.0 to 93.6) 0.9 (–4.2 to 4.8) 2.8 (–4.3 to 6.6)
178.2 (172.4 to 183.4) 88.6 (94.3 to 93.1) 89.8 (84.4 to 93.7) 0.5 (–5.1 to 5.2) 4.0 (–3.2 to 5.9)
0.34† 0.0005 0.03 0.95 0.65
* Quantitative data expressed as mean (ranges) † Fisher's exact probability test ‡ Mann–Whitney test HKA, hip-knee-ankle alignment; F, femoral component mechanical angle; T, tibial component mechanical angle; TS, tibial slope; FF, flexion angle of the femoral component
The sizes of the femoral and tibial components and polyethylene insert determined pre-operatively were compared with those of the implanted components to evaluate the predictive value of pre-operative planning in the PSCG group. We also assessed blood loss according to the formula of Mercuriali and Inghilleri;19 operating time from skin incision to wound closure; Knee Society (KS) knee, function and global scores;20 Oxford Knee Score (OKS; 0 to 48 worst to best);21 post-operative complications and length of hospital stay. Statistical analysis. We carried out a power analysis based on the mechanical axis of the lower limb defined by the HKA angle. We calculated the number of patients needed, assuming that 65%12 of the patients in the CI group would be aligned to within 3° of neutral (180°; SD 3) compared with 90% in the PSCG group. With a two-sided risk (α) of 2.5% and a power (1 – β) of 90%, we calculated that 64 patients would be needed in each group. Estimating that 10% of the patients would have incomplete data, we decided to include 70 patients in each arm of the study. The patients were randomised using a computer-generated block randomisation scheme based on the order of presentation, so that each patient was randomised individually regardless of the severity of their OA and deformity. To avoid surgeon bias, randomisation was stratified by the surgeon. All patients gave informed consent, and the study had approval from our ethics committee. The trial was registered at ClinicalTrials.gov (NCT 02002624). VOL. 97-B, No. 1, JANUARY 2015
Continuous variables were expressed as means and ranges. Statistical analysis was performed using nonparametric Mann–Whitney U test or Fisher’s exact probability test, as appropriate. Correlation analysis was performed using Pearson’s coefficient. Statview statistical software version 5.0 (SAS Institute, Cary, North Carolina) was used and with statistical significance defined as a pvalue < 0.05.
Results There was no significant difference in pre-operative evaluation between the two groups (Table I). Of the 140 patients included, 126 had complete clinical and radiological follow-up data (Fig. 1). Two patients in the PSCG group were treated using conventional instrumentation on the tibial side because the PSCG tibial guide fitted poorly at the time of operation. Post-operatively, there was no significant difference in the mean angles measured in the TKAs between the two groups (Table II). There was also no significant difference between the CI and PSCG groups in the number of knees that were outliers (Table III). There was also no significant correlation between the post-operative HKA angle and the date of surgery, indicating no learning curve effect (r2 = 0.009, p = 0.48). The predictive sizing value of the PSCG system was within one size in all the implanted components on both the femoral and the tibial sides (Table IV). The mean
60
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
Table III. Number of post-operative radiological outliers in the conventional instruements (CI) and patient-specific cutting guides (PSCG) groups Variables
CI group (n = 67) (n, %)
PSCG group (n = 59) (n, %)
p-value*
HKA angle F angle T angle TS FF angle
22 (32.8) 12 (17.9) 7 (10.4) 15 (22.4) 21 (31.3)
19 (32.2) 19 (25.4) 7 (11.9) 18 (30.5) 22 (37.3)
0.99 0.41 0.99 0.11 0.44
* Fisher’s Exact probability test HKA, hip-knee-ankle alignment; F, femoral component mechanical angle; T, tibial component mechanical angle; TS, tibial slope; FF, flexion angle of the femoral component
Table IV. Predictive sizing value of pre-operative planning in the patient-specific cutting guides group (implanted size compared with the pre-operatively planned size) Variables
Same size (n, %)
Within one size (n, %)
Within two sizes (n, %)
Femoral component Tibial component Insert thickness
49/70 (70) 50/70 (71.4) 53/70 (75.8)
21/70 (30) 20/70 (28.6) 12/70 (17.1)
0 0 5/70 (7.1)
Table V. Knee Society (KS) and Oxford knee scores (OKS) at three months’ follow-up in the conventional instrumentation (CI) and the patient-specific cutting guides (PSCG) groups Variables*
CI group
PSCG group
p-value†
KS score Knee Function Global OKS
83.3 (34 to 99) 78.3 (10 to 100) 161.7 (79 to 199) 38.2 (21 to 48)
82.5 (51 to 97) 82.1 (40 to 100) 164.7 (98 to 197) 37.9 (18 to 48)
0.20 0.19 0.59 0.95
* Data expressed as mean (ranges) † Mann–Whitney test
operating time was 85.7 minutes (45 to 145) in the CI group and 79.4 minutes (40 to 150) in the PSCG group (Mann–Whitney test, p = 0.09). The mean blood loss was 1358 ml (347 to 3756) in the CI group and 1285 ml (78 to 3643) in the PSCG group (Mann–Whitney test, p = 0.58). The mean length of hospital stay was 8.3 days (6 to 18) in the CI group and 8.1 days (5 to 16) in the PSCG group (Mann–Whitney test, p = 0.50). No significant difference in KS or OKS was seen between the two groups (Table V).
Discussion Lower limb malalignment in the coronal plane after TKA, defined as a deviation > 3° from neutral, has been associated with an increased risk of revision and lower functional scores.6,7,22 Patient-specific instrumentation was introduced to translate 3D planning from the intra-operative to the pre-operative setting. To date, several controlled studies13-17,23-36 have reported conflicting results from PSCG use in primary TKA (Table VI). In addition, both the 3D reconstruction of the knee computer algorithm and the block designs differ substantially from one system to another, which necessitates a specific evaluation of each PSCG system. The Visionaire system is unique, as, in addition to MRI reconstruction of the knee, it includes a fulllength view of the lower limb which provides data on joint laxity: this could make the evaluation of pre-operative
deformity and post-operative alignment more accurate. So far, the results reported for the Visionaire system have lacked the power to allow any strong conclusions to be drawn.28,31,33 We therefore conducted a prospective randomised study with a power of 90% based on a reduction of outliers with PSCG that would be similar to the results obtained with CAS in primary TKA. There were several limitations to our study. First, we did not evaluate the axial alignment of the components. Our study protocol focussed on coronal alignment and did not include post-operative CT, which would have resulted in increased radiation exposure for a relatively large number of patients, and additional cost. Second, it could be argued that radiological measurements were carried out on a standing full-length AP radiograph of the lower limb rather than on CT scans, and that we could not provide accuracy data for our measurements. However, it has recently been shown that this method has a high intra- and inter-observer correlation compared with CT and intra-operative navigation, and limits the patient’s exposure to radiation.37 Also, in our study all measurements were made by the same observer, who was blinded to the instrumentation, with a reproducibility (99% CI) < 1.3° for all measured angles. Third, 14 (10%) of the 140 patients included did not have complete follow-up data, leaving 67 patients in the CI group and 59 in the PSCG group. Our power calculation had been exacting and based THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
61
Table VI. Main published controlled studies comparing patient-specific cutting guides PSCG with conventional instrumentation (CI) in primary total knee arthroplasty Mean post-operative HKA outliers > 3° Mean operating Study
Type of study
Number of knees
Evaluation
HKA angle (°)*
(%)*
time*(mins)
Ng et al13
Retrospective
CI 55/ Signature: 105
181.1/ 180.6
22*/ 9*
NA/ NA
Nunley et al29
Retrospective matched-pairs Retrospective consecutive Retrospective first 100 cases Retrospective matched-pairs Prospective randomised Retrospective
CI: 57/ Signature: 57
Standing long-leg radiograph Scout CT scanogram
NA/ NA
18/ 26
93.4/ 89.6
Nunley et al17 Barrack et al14 Boonen et al24 Boonen et al15 Nam et al27 Roh et al33 Noble et al28 Vunderlinckx et al35 Chareancholvanich et al16 Parratte et al30 Pietsch et al32 Hamilton et al25 Woolson et al36 Victor et al34 Pfitzner et al31 Current study
Prospective randomised Prospective randomised Prospective controlled Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised Prospective randomised
CI: 50/ Signature: 50/ Scout CT scanogram OtisMed: 50 CI: 100/ Signature: 100 Scout CT scanogram
179.9/ 180.7/ 182.8
16*/ 18*/ 44*
NA/ NA/ NA
180.5*/ 181.7*
23/ 31
NA/ NA
CI: 40/ Signature: 40
179*/ 181*
50/ 30
61/ 51
178/ 179
18/ 30
50/ 45
179.2/ 179.2
7.3/ 29.3
NA/ NA
180.3/ 179.5
10/ 12
NA/ NA
182.8*/ 181.7*
NA/ NA
128.1*/ 121.4*
Standing long-leg radiograph CI: 90/ Signature: 90 Standing long-leg radiograph CAS: 37/ Signature: 38 Standing long-leg radiograph CI: 50/ Signature: 50 Standing long-leg radiograph CI: 14/ Visionaire: 15 Standing long-leg radiograph CI: 31/ Visionaire: 31 Standing long-leg radiograph CI: 40/ PSI: 40 CT scanogram
2.7†/ 3.0†
NA/ NA
NA/ NA
179.7/ 179.7
7.5/ 2.5
68*/ 62.9*
CI: 40/ PSI: 40
Standing long-leg radiograph NA/ NA
178.3/ 179
10/ 5
NA/ NA
NA/ NA
NA/ NA
90*/ 78*
Standing long-leg radiograph CT scan
180.7/ 180.3
31/ 35
57/ 61
178.7/ 178.3
38/ 41
NA/ NA
Standing long-leg radiograph Standing long-leg radiograph Standing long-leg radiograph
178.1/ 178.6
28.1/ 24.6
NA/ NA
4.5†/ 3.0†/ 1.0†
43/ 30/ 7
76/ 63/ 58
178.9/ 178.2
32.8/ 32.2
85.7/ 79.4
CI: 40/ PSI: 40 CI: 26/ Trumatch: 26 CI: 30/ Trumatch: 33 CI: 64/ Four types PSCG: 64 CI: 30/ Trumatch: 30 Visionaire: 30 CI: 70/ Visionaire: 70
* p < 0.05 † Deviation from 0° post-operative mechanical axis Proprietry patient-specific cutting guides: Signature (Biomet, Warsaw, Indiana); OtisMed (OtisMed Corp, Alameda, California); PSI (Zimmer, Warsaw, Indiana); Trumatch (Depuy, Warsaw, Indiana) NA, not available; CAS, computer-assisted surgery
on 64 patients being included in each group so the study might have been under powered. Fourth, our study focused only on the first three post-operative months, and the data cannot be extrapolated to a longer period of follow-up. Last, our results cannot be generalised to any other commercially available system, as they differ markedly in terms of pre-operative planning and design. Using both the KS and the OKS, we found no significant difference between PSCG and CI in terms of clinical results at three months. Other studies that have compared the short-term clinical results of PSCG and CI23,31,32,35 have had similar outcomes to ours. The mean HKA angle at three months did not differ significantly between the two groups. This is consistent with most published controlled studies (Table VI). It should be emphasised that conventional systems have provided a largely acceptable mean post-operative HKA angle,12-17,24-36 and it should be expected that PSCG would reduce the proportion of outliers, thereby enhancing precision more than accuracy. However, in our study there was no significant difference in VOL. 97-B, No. 1, JANUARY 2015
the proportion of outliers for any measured angle, which refutes our hypothesis. Most of the available literature comparing PSCG and CI concurs with our results,16,24,28,30,32 whereas some showed a non-significant increase in the percentage of outliers using PSCG.14,15,17,29,33 However, apart from our study and that of Boonen et al,15 all the other studies were underpowered for this outcome. There are several explanations for this unexpected finding. In a recent review of rapid prototype surgical guides, Krishnan et al38 found multiple sources of errors in the acquisition of images, the 3D reconstruction protocol and the manufacturing of the guides, all of which can lead to some degree of inaccuracy. Also, the contact points of MRI-based guides rely on cartilage, whereas CT-based guides rely on surrounding bone and osteophytes; both have limitations when it comes to the fitting the guides. The material of the blocks is derived from nylon and has inherent elasticity that can lead to inexact placement; it can also have some mobility during fixation. The saw blade itself can lead to cutting errors, especially in the coronal plane.39 It becomes clear that if, for one
62
L. ABANE, P. ANRACT, S. BOISGARD, S. DESCAMPS, J. P. COURPIED, M. HAMADOUCHE
given case, all possible sources of errors are in the same direction, malalignment > 3° can occur. Some studies that used tri-planar intra-operative computer navigation have shown that these potential errors occur with both tibial and femoral cutting guides.40,41 In these studies, malalignment > 3° would have occurred in 21.7% to 45.5% of the cases if the procedure had been performed using only the PSCG guides. Similarly, some authors have described the need for frequent changes intra-operatively, and the necessity to have each case reviewed in detail by the operating surgeon prior to electronic approval of the proposed pre-operative surgical plan.42,43 We found the system was reasonably able to predict the size of implant: 70% of the implanted components were the size predicted pre-operatively, and 100% were within one size. There was more variation with the tibial insert, but PSCG systems in general do not account for joint laxity. These data, and those from Vundelinckx et al,35 will encourage the further development of fully disposable instrumentation, which is currently being undertaken by many companies. Operating time was reduced by a mean of seven minutes, which is in accordance with other reports, in which it varies between six and 18 minutes.14-16,24,28,30,31 Blood loss was not significantly different in the two groups, and nor was length of hospital stay. We conclude that PSCG does not at this point provide better clinical, radiological or resource consumption outcomes. L. Abane: Gathered the data, Analysed the data, Wrote the initial draft, Approved the final manuscript. P. Anract: Contributed to the study design, Study implementation, Approved the final manuscript. S. Boisgard: Contributed to the study design, Study implementation, Refinement of the study protocol, Ensured accuracy of data and analysis, Approved the final manuscript. S. Descamps: Contributed to the study design, Study implementation, Gathered the data, Analysed the data, Approved the final manuscript. J. P. Courpied: Contributed to the study design, Study implementation, Approved the final manuscript. M. Hamadouche: Contributed to the study design, Study implementation, Refinement of the study protocol, Analysed the data, Wrote the initial draft, Ensured accuracy of data and analysis, Approved the final manuscript. The authors thank J. P. Levai MD for contributing patients. One or more authors received benefits from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by A. Ross and first proof edited by G. Scott.
References 1. Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg [Am] 2004;86-A:963–974. 2. No authors listed. Swedish knee arthroplasty register. Annual report 2012. http:// www.myknee.se/pdf/117_SKAR_2012_Engl_1.0.pdf (date last accessed 14 November 2014). 3. No authors listed. The Norwegian arthroplasty register annual report 2010. http:// nrlweb.ihelse.net/eng/Report_2010.pdf (date last accessed 27 October 2014). 4. Parratte S, Pagnano MW, Trousdale RT, Berry DJ. Effect of postoperative mechanical axis alignment on the fifteen-year survival of modern, cemented total knee replacements. J Bone Joint Surg [Am] 2010;92-A:2143–2149. 5. Bellemans J, Colyn W, Vandenneucker H, Victor J. The Chitranjan Ranawat award: is neutral mechanical alignment normal for all patients? The concept of constitutional varus. Clin Orthop Relat Res 2012;470:45–53. 6. Berend ME, Ritter MA, Meding JB, et al. Tibial component failure mechanisms in total knee arthroplasty. Clin Orthop Relat Res 2004;428:26–34.
7. Ritter MA, Davis KE, Meding JB, et al. The effect of alignment and BMI on failure of total knee replacement. J Bone Joint Surg [Am] 2011;93-A:1588–1596. 8. Bathis H, Perlick L, Tingart M, et al. Alignment in total knee arthroplasty. A comparison of computer-assisted surgery with the conventional technique. J Bone Joint Surg [Br] 2004;86-B:682–687. 9. Matziolis G, Krocker D, Weiss U, Tohtz S, Perka C. A prospective, randomized study of computer-assisted and conventional total knee arthroplasty. Three-dimensional evaluation of implant alignment and rotation. J Bone Joint Surg [Am] 2007;89A:236–243. 10. Bauwens K, Matthes G, Wich M, et al. Navigated total knee replacement. A meta-analysis. J Bone Joint Surg [Am] 2007;89-A:261–269. 11. Beldame J, Boisrenoult P, Beaufils P. Pin track induced fractures around computer-assisted TKA. Orthop Traumatol Surg Res 2010;96:249–255. 12. Novak EJ, Silverstein MD, Bozic KJ. The cost-effectiveness of computer-assisted navigation in total knee arthroplasty. J Bone Joint Surg [Am] 2007;89-A:2389–2397. 13. Ng VY, De CLaire JH, Berend KR, Gulick BC, Lombardi AV Jr. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470:99–107. 14. Barrack RL, Ruh EL, Williams BM, et al. Patient specific cutting blocks are currently of no proven value. J Bone Joint Surg [Br] 2012;94(11 Suppl A):95–99. 15. Boonen B, Schotanus MG, Kerens B, et al. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomised controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21:2206–2212. 16. Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. Bone Joint J 2013;95-B:354– 359. 17. Nunley RM, Ellison BS, Zhu J, et al. Do patient-specific guides improve coronal alignment in total knee arthroplasty? Clin Orthop Relat Res 2012;470:895–902. 18. Cooke TD, Sled EA, Scudamore RA. Frontal plane knee alignment: a call for standardized measurement. J Rheumatol 2007;34:1796–1801. 19. Mercuriali F, Inghilleri G. Proposal of an algorithm to help the choice of the best transfusion strategy. Curr Med Res Opin 1996;13:465–478. 20. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989;248:13–14. 21. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg [Br] 1998;80-B:63–69. 22. Huang NF, Dowsey MM, Ee E, et al. Coronal alignment correlates with outcome after total knee arthroplasty: five-year follow-up of a randomized controlled trial. J Arthroplasty 2012;27:1737–1741. 23. Abdel MP, Parratte S, Blanc G, et al. No benefit of patient-specific instrumentation in TKA on functional and gait outcomes: A randomized clinical trial. Clin Orthop Relat Res 2014;472:2468–2476. 24. Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta Orthop 2012;83:387–393. 25. Hamilton WG, Parks NL, Saxena A. Patient-specific instrumentation does not shorten surgical time: a prospective, randomized trial. J Arthroplasty 2013;28(Suppl.1):96–100. 26. Klatt BA, Goyal N, Austin MS, Hozack WJ. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23:26–29. 27. Nam D, Maher PA, Rebolledo BJ, et al. Patient specific cutting guides versus an imageless, computer-assisted surgery system in total knee arthroplasty. Knee 2013;20:263–267. 28. Noble JW, Moore CA, Liu N. The value of patient-matched instrumentation in total knee arthroplasty. J Arthroplasty 2012;27:153–155. 29. Nunley RM, Ellison BS, Ruh EL, et al. Are patient-specific cutting blocks costeffective for total knee arthroplasty? Clin Orthop Relat Res 2012;470:889–894. 30. Parratte S, Blanc G, Boussemart T, et al. Rotation in total knee arthroplasty: no difference between patient-specific and conventional instrumentation. Knee Surg Sports Traumatol Arthrosc 2013;21:2213–2219. 31. Pfitzner T, Abdel MP, von Roth P, Perka C, Hommel H. Small improvements in mechanical axis alignment with MRI versus CT-based patient-specific instrumentsin TKA: A randomized clinical trial. Clin Orthop Relat Res 2014;472:2913–2922. 32. Pietsch M, Djahani O, Zweiger C, et al. Custom-fit minimally invasive total knee arthroplasty: effect on blood loss and early clinical outcomes. Knee Surg Sports Traumatol Arthrosc 2013;21:2234–2240. THE BONE & JOINT JOURNAL
A COMPARISON OF PATIENT-SPECIFIC AND CONVENTIONAL INSTRUMENTATION FOR TOTAL KNEE ARTHROPLASTY
63
33. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instruments comparable to Conventional TKA? A randomized controlled study of one system. Clin Orthop Relat Res 2013;471:3988–3995.
38. Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg [Br] 2012;94-B:1457–1461.
34. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific guides do not improve accuracy in total knee arthroplasty: a prospective randomized controlled trial. Clin Orthop Relat Res 2014;472:263–271. 35. Vundelinckx BJ, Bruckers L, De Mulder K, De Schepper J, Van Esbroeck G. Functional and radiographic short-term outcome evaluation of the visionaire system, a patient-matched instrumentation system for total knee arthroplasty. J Arthroplasty 2013;28:964–970.
39. Yau WP, Chiu KY. Cutting errors in total knee replacement: assessment by computer assisted surgery. Knee Surg Sports Traumatol Arthrosc 2008;16:670–673. 40. Conteduca F, Iorio R, Mazza D, et al. Evaluation of the accuracy of a patient-specific instrumentation by navigation. Knee Surg Sports Traumatol Arthrosc 2013;21:2194–2199. 41. Lustig S, Scholes CJ, Oussedik SI, et al. Unsatisfactory accuracy as determined by computer navigation of VISIONAIRE patient-specific instrumentation for total knee arthroplasty. J Arthroplasty 2013;28:469–473. 42. Pietsch M, Djahani O, Hochegger M, Plattner F, Hofmann S. Patient-specific total knee arthroplasty: the importance of planning by the surgeon. Knee Surg Sports Traumatol Arthrosc 2013;21:2220–2226. 43. Stronach BM, Pelt CE, Erickson J, Peters CL. Patient-specific Total Knee Arthroplasty Required Frequent Surgeon-directed Changes. Clin Orthop Relat Res 2013;471:169–174.
36. Woolson ST, Harris AH, Wagner DW, Giori NJ. Component alignment during total knee arthroplasty with use of standard or custom instrumentation: a randomized clinical trial using computed tomography for postoperative alignment measurement. J Bone Joint Surg [Am] 2014;96-A:366–372. 37. Babazadeh S, Dowsey MM, Bingham RJ, et al. The long leg radiograph is a reliable method of assessing alignment when compared to computer-assisted navigation and computer tomography. Knee 2013;20:242–249.
VOL. 97-B, No. 1, JANUARY 2015
