Comparison of navigation accuracy in THA between the mini-anterior and -posterior approaches

Transcription

1 THE INTERNATIONAL JOURNAL OF MEDICAL ROBOTICS AND COMPUTER ASSISTED SURGERY Int J Med Robotics Comput Assist Surg 2009; 5: Published online 23 December 2008 in Wiley InterScience ( ORIGINAL ARTICLE Comparison of navigation accuracy in THA between the mini-anterior and -posterior approaches Takehito Hananouchi 1 * Masaki Takao 1 Takashi Nishii 1 Hidenobu Miki 2 Daiki Iwana 3 Hideki Yoshikawa 1 Nobuhiko Sugano 4 1 Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Osaka, Japan 2 Department of Orthopaedic Surgery, Osaka Medical Centre, Osaka, Japan 3 Centre of Arthroplasty, Kyowakai Hospital, Osaka, Japan 4 Department of Orthopaedic Medical Engineering, Osaka University Graduate School of Medicine, Osaka, Japan *Correspondence to: Takehito Hananouchi, Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Osaka, Japan. hana-osaka@umin.net Abstract Background The accuracy of a CT-based hip navigation might depend on surgical approaches, resulting in varying accuracy of implant alignment. Methods We performed primary cementless total hip arthroplasty (THA) with mini-incision surgery (MIS) to 40 well-matched patients (anterior or posterior approaches, 20 hips each), using navigation with surface registration. We investigated cup alignment using postoperative computed tomography (CT) and compared the navigation accuracy between the two approaches, i.e. the difference between intra-operative and postoperative alignments of the cup. Results There was no significant difference between the two approaches. The mean navigation accuracies in abduction and anteversion were 2.0 (SD 1.4 )and2.7 (SD 1.9 ), respectively, in the anterior approach, and 2.4 (SD 2.0 )and2.0 (SD 1.4 ), respectively, in the posterior approach. All cup alignments were within 10 of the target orientation. Conclusions This CT-based navigation for MIS THA provides navigation accuracy without significant differences between the two approaches and with favourable alignment of the cup. Copyright 2008 John Wiley & Sons, Ltd. Keywords CT-based hip navigation; total hip arthroplasty; mini-incision surgery; surface registration Introduction Accepted: 12 November 2008 In total hip arthroplasty (THA), minimally invasive or mini-incision surgery (MIS) has become increasingly used to reduce soft tissue damage and promote quick functional recovery (1). Because MIS THAs are performed through a smaller skin incision than conventional THAs, it is a challenge to avoid misalignment of the implants (2). It has been reported that malpositioning of the acetabular cup increases the risk of dislocation and neck impingement on the cup liner, resulting in mechanical problems with the implant parts, such as wear and fracture (3 5). To obtain successful long-term results following THA, it is essential to avoid neck cup impingement due to malposition of the acetabular cup. To eliminate acetabular cup malposition in MIS THAs, computer navigation has been reported to be useful (6 7). However, the accuracy of navigation depends on the techniques of registration used for the navigation and secure fixation of the dynamic reference markers (8). These could be affected by the type of MIS approach, because the available anatomical portions for the registration depend on the approach. To the Copyright 2008 John Wiley & Sons, Ltd.

2 Navigation accuracy between two surgical approaches 21 best of our knowledge, there has been no previous report to investigate clinical and navigation accuracy of CTbased navigation in the mini-anterior approach, although there have been some reports on the clinical accuracy of navigation in MIS THA through a posterior approach (9). Furthermore, navigation accuracy has not been compared between the two different surgical approaches. The purpose of this study was to investigate whether the navigation accuracy of a CT-based navigation system during cup insertion differed between a minianterior approach and a mini-posterior approach, and to investigate whether navigation in the two approaches provides favourable results of cup alignment compared with the conventional technique. Methods In the period July 2005 December 2006, 91 primary cementless THAs (86 patients) with the acetabular component (Trident, Stryker Orthopaedics, Mahwah, NJ, USA) were performed, using the computed tomography (CT)-based navigation system (CT-Base HIP Navigation System, Stryker Navigation, Dallas, TX, USA). Of the 91 THAs, 20 were performed through a mini-anterior approach according to the following criteria: preoperative extension over 0 and a diagnosis of osteoarthritis or osteonecrosis. The remaining 71 THAs were performed through a mini-posterior approach, of which 20 were matched according to age, gender, body mass index and diagnosis, in order to compare the navigation accuracy (Table 1). This study was approved by the hospital s ethics committee, and all patients gave their written informed consent. Preoperative planning Preoperative CT images of each patient were obtained using a helical CT scanner (HiSpeed Advantage, GE Medical Systems, Milwaukee, WI, USA) from the level of the superior anterior iliac spine of the pelvis to the level of the femoral condyles. In the CT scan protocol for the pelvis, the slice thickness was 1 mm and the pitch was 3 mm. The CT data were transferred to the preoperative planning module of the navigation system. A three-dimensional (3D) surface model of the pelvis was reconstructed using the segmentation procedure in the planning module. The position and orientation of the cup were planned using three multiplanar reconstruction views and a 3D volume-rendering view. The target orientation of the cup was planned to be 40 of abduction and 15 of anteversion, as determined radiographically (10). Surface registration during the surgical procedure The MIS THAs were performed by two senior hip surgeons (N.S. and T.N.), who used the two surgical approaches (mini-anterior approach, 18 cases operated by N.S. and two cases by T.N.; mini-posterior approach, 14 cases by N.S. and six cases by T.N.). All operations were performed in the lateral decubitus position with a skin incision of <10 cm. The dynamic reference marker for the registration of the pelvis was placed on the anterior superior iliac spine in both approaches. In the mini-anterior approach, we performed a tissue-sparing approach which modified the Smith Peterson approach through the interval between the tensor fascia latae and the sartorius muscles, without detaching or sectioning any of the muscles and tendons around the hip joint (11), while we performed the mini-posterior approach with dissection of the tensor fascia latae, split of the gluteus maximus and incision of the short external rotators (11). The registration of the pelvis was performed after femoral head resection and acetabular exposure. Initially, coarse paired-point registration was performed by digitizing four bony landmarks of the pelvis, which were determined during preoperative planning. Then, precise surface registration was performed by digitizing 30 points on the pelvic bone surface. Digitizing for surface registration was performed through a skin incision, using the digitizing pointer with optical sensors (Figure 1). According to one study of surface registration (12), three rotational angles and three translation vectors in a coordinate system are required in computation or calculation for optimal surface registration. On the basis of that report, we considered two priorities for the optimal surface registration. First, we got wide area as much as possible for getting a lot of translation vectors. Second, we included a 3D waved area to obtain a lot of rotational angles. In order to find the 3D waved area in each approach, we determined the anterior inferior iliac spine in the mini-anterior approach, and the greater sciatic notch in the mini-posterior approach. Table 1. Patient data Parameter Mini-anterior approach (n = 20) Mini-posterior approach (n = 20) p Value Gender Male 2, female 18 Male 2, female 18 1 Age (years) 55.1 (42 68) 57.0 (42 73) 0.51 Body mass index 22.2 (15 27) 21 (17 25) 0.47 Underlying disease OA/ON 14/6 14/6 1 χ 2 test. Mann Whitney U-test. OA, osteoarthritis; ON, osteonecrosis.

3 22 T. Hananouchi et al. Then we determined the digitizing area for the surface registration of each approach as follows. For the minianterior approach (Figure 2), the anterior and superior peri-articular area within 5 cm of the acetabular rim and the antero-inferior part of the ala of the ilium, including the anterior inferior iliac spine, were included; and for the mini-posterior approach (Figure 3), the periarticular area within 5 cm of the acetabular rim and the postero-inferior part of the ala of the ilium, including the sciatic notch, were included. After the registration, a point was marked using a surgical marking pen, and the point was registered using the digitizing pointer. The stability of the reference markers was checked during the entire surgical procedure by touching the marked point with the pointer. When the tip of the pointer was displaced 2 mm from the registered point on the navigation computer, the reference markers were considered to have been loosened. Cup fixation with the navigation system The position of the surgical tools and the target orientation of the cup to the pelvis were presented on the monitor after the registration step was completed. Checking the alignment and position of the cup on the monitor, the cup was placed using 1.8 mm press-fit at the rim without screws. While looking at the orientation of the cup on the monitor, the two surgeons were encouraged to match the intra-operative orientation of the cup with the preoperative target orientation of the cup as much as possible. The final abduction and anteversion angles were recorded. This cup orientation was defined as intraoperative cup orientation. We confirmed whether the reference marker had deviated after the final cup fixation. After the acetabular procedure, the remaining steps of the THA, including femoral rasping and stem fixation, were performed. Postoperative measurements of the cup All patients had a CT scan 3 weeks after the operation. The postoperative orientation of the cup was measured after adjusting for differences in pelvic orientation between the preoperative and postoperative CT images. Semiautomatic volume registration by 3D image-processing software (Virtual Place-M; Medical Imaging Laboratory, Tokyo, Japan) was performed to adjust for the differences in pelvic orientation. The reproducibility of the volume registration procedure was previously reported to be 0.7 and 0.8 mm (13). Intra- and inter-observer variability of the measurements of the postoperative cup orientation were evaluated using Pearson correlation coefficients. Ten cases were randomly selected for assessment of the variability. The intra- and inter-observer variabilities were 0.95 and 0.83, respectively, as evaluated by two surgeons. This cup orientation was defined as postoperative cup orientation. Navigation accuracy, which was defined as the absolute difference between the intra-operative cup orientation and the postoperative cup orientation, was compared between the two approaches in terms of its mean difference and variance. Since the press-fit technique might be affected by the surgical approach for MIS THA, which could affect the final cup alignment, cup alignment deviation during implantation, which was defined as the absolute difference between the target orientation and the intra-operative cup orientation, was compared between the two approaches in terms of its mean difference and variance. Figure 1. Digitizing for surface registrationis performed through the surgical incision, using a pointer of the CT-based navigation. In this figure the scheme of the pelvis is overlaid to show the relationship between the pointer and the surface of the pelvis during digitizing Figure 2. Representative distribution of the digitizing points through the mini-anterior approach

4 Navigation accuracy between two surgical approaches 23 According to some previous reports that indicated the outlier of the cup alignment (14,15), surgeons with conventional manual technique could not achieve the criterion, which was determined as being within 10 from the target orientation. In the context, we determined this criterion as the outlier in our study and investigated whether all our cases achieved this criterion. The following parameters relating to navigated surgery were also compared: time for paired point registration; time for surface registration; cup size; operative time; and intra-operative blood loss. The occurrence of dislocations within 1 year after surgery was also determined. Statistical analysis Ameandifferenceof3 in navigation accuracy of the cup insertion was identified as significant. This is because 2 3 could be reasonable statistically and clinically when investigating angle difference between two different interventions in THA, according to one previous report (15). Therefore, a sample size to detect a difference of 3 in clinical accuracy was calculated. Twenty hips in each approach were sufficient to determine whether there was a significant difference (power = 0.8; p < 0.05). We used the χ 2 test for categorical data and the Mann Whitney U- test and analysis of variance for continuous data. We also investigated whether differences between the surgeons in this study affected the comparison between the two approaches as a confounding factor, using multiple analysis of variance. p < 0.05 was considered statistically significant. Results There was no significant difference between the two approaches in terms of the mean differences and the variance of the navigation accuracy (mean difference, abduction p = 0.61 and anteversion p = 0.29; variance; abduction p = 0.73 and anteversion p = 0.18). The mean accuracy of the navigation system was 2.0 (SD 1.4,range )forabductionand2.7 (SD 1.9, range 0 6.0) for anteversion in the mini-anterior approach. The mean navigation accuracy was 2.4 (SD 2.0, range ) for abduction and 2.0 (SD 1.4, range ) for anteversion in the mini-posterior approach. There was no case in which the reference marker deviated > 2mm during THA. With regard to the cup alignment deviation during implantation, there was no significant difference between the two approaches (mean difference, abduction p = 0.61 and anteversion p = 0.29; and variance, abduction p = 0.22 and anteversion p = 0.34). The mean cup alignment deviation during implantation was 1.1 (SD 0.8,range ) of abduction and 1.4 (SD 2.3, range ) of anteversion in the mini-anterior approach. The mean cup alignment deviation during implantation was 1.8 (SD 2.0,range0 7.0 ) of abduction and 1.5 (SD 1.7,range ) of anteversion in the mini-posterior approach. The postoperative orientations of the cup were 39.3 (SD 2.7, range ) of abduction and 15.0 (SD 2.8, range ) in the mini-anterior approach (Figure 4). The postoperative orientations of the cup were 36.6 (SD 3.5, range ) of abduction and 16.0 (range ,SD4.1 )ofanteversionin the mini-posterior approach (Figure 4). The postoperative orientations of the cup in all cases with both approaches were within 10 of the target orientation of the cup (abduction 40 and anteversion 15 ) (Figure 4). None of the patients developed postoperative dislocation within 1 year after surgery. Except for the time for surface registration, there were no significant differences between the two approaches in the other parameters measured (time for the paired point registration, p = 0.88; cup size, p = 0.76; operative time, p = 0.11; and intra-operative blood loss, p = 0.77) (Table 2). There was a statistically significant difference (p < ) in the time for surface registration, but on average the difference was only about 1 min. The differences between the surgeons did not affect these results (p = 0.88). Discussion To investigate whether different surgical approaches affect the navigation accuracy of CT-based hip navigation using surface registration, the navigation accuracies of the mini-anterior and mini-posterior approaches were compared. In the present study, there was no significant Figure 3. Representative distribution of the digitizing points through the mini-posterior approach

5 24 T. Hananouchi et al. Figure 4. Scattergram of the cup alignment in the two approaches. All cups are within 10 of the target orientation (greyframe, 10 from abduction 40 and anteversion 15 ) difference in the navigation accuracy between the minianterior and mini-posterior approaches. Intra-operative confirmation was obtained that the reference markers did not loosen in any of the cases. Thus, the navigation accuracy in the present study was mainly affected by the registration accuracy. Previous reports dealing with CT-based navigation in only one surgical approach have shown that the accuracy of surface registration varies according to the digitizing area (12,16). To the best of our knowledge, no previous studies have compared the registration accuracy between different surgical approaches. One study of surface registration for CT-based navigation surgery showed that a 3D waved area should be included in the digitizing area, because the 3D waved area determines three rotation angles (12). We previously reported that the sciatic notch was one 3D waved area for surface registration during THA through the posterior approach. Navigation accuracy was greater using the sciatic notch than without its use (16). In the present study, the anterior inferior iliac spine was included for surface registration during THA through the anterior approach. The anterior inferior iliac spine appears to be another 3D waved area. This would be one reason why surface registration with the mini-anterior approach is as accurate as that with the mini-posterior approach. In order to investigate the navigation accuracy precisely, postoperative CT images were used, and volume registration was performed to adjust for preoperative and postoperative pelvic orientations. There has been only one previous clinical study about the navigation accuracy of CT-based navigation for THA using postoperative CT images and adjustment of the pelvic coordination system (15). In that report, the mean navigation accuracy was 3.0 (SD 2.6 ) of abduction and 3.3 (SD 2.3 ) of anteversion. The accuracy of navigation determined in the present study was comparable to that of the previous report. It is not always possible to acquire the target orientation of the cup, even under navigation guidance. A previous report showed the undesirable effect of the press-fit procedure on the final alignment of the of the cup (9). In the initial 10 cases of that study, the mean absolute cup alignment deviations during implantation were 3.7 of abduction (range 0 9,SD2.8 )and5.1 of anteversion (range 0 10,SD2.3 ). In the present study, cup fixation was performed while monitoring cup orientation. The mean cup alignment deviation during implantation in both approaches was within 2 in both abduction and anteversion. This indicates that the difference in the surgical approaches did not significantly affect the pressfit procedure. However, we consider that the effect of the press-fit itself on the final alignment of the cup could not be ignored in terms of the maximum angle of the alignment deviation due to the press-fit procedure. Therefore, we think that the current press-fit techniques need to be developed to allow surgeons to be as accurate as they would like to be. Since the accuracy of navigation was acceptable and the cup alignment deviation during press-fit fixation was small in both MIS approaches, all cases achieved the criterion, which was within 10 of the target cup orientation. We consider our results to be more favourable than those in previous reports of THAs using the conventional technique because they could not achieve the criterion. As favourable cup alignment reduces the risk of neck/liner impingement, which may lead to dislocation and mechanical problems of the load-bearing parts, both MIS THA approaches using the current CT-based navigation system are expected to achieve successful longterm outcomes. The time for surface registration was the only measurement that was significantly different between the two approaches. However, the actual difference in the Table 2. Intraoperative and postoperative measurements in the two approaches Parameter Mini-anterior approach (n = 20) Mini-posterior approach (n = 20) p Value Time for registration (min) Paired-point registration 0.6 (range ; SD 0.3) 0.6 (range , SD 0.2) 0.88 Surface registration 3.4 (range ; SD 0.9) 2.1 (range , SD 0.4) < Cup size (mm) 50.7 (range 48 56) 50.4 (range 46 58) 0.76 Operative time (min) (range ) (range ) 0.11 Intraoperative blood loss (ml) (range ) (range ) 0.77 Mann Whitney U-test.

6 Navigation accuracy between two surgical approaches 25 time was only 1 min. Furthermore, the total operation time was not significantly different between the two approaches. Therefore, this difference does not appear to be of clinical importance. The difference between the two approaches may be associated with the learning curve for digitizing through the mini-anterior approach, because we performed fewer CT-based navigation THAs through the anterior approach than through the posterior approach, with over 300 cases since The present study had some limitations. First, it was not randomized; however, the patients demographic factors were unlikely to have affected the results, since the two approaches were comparable in terms of age, gender, body mass index and underlying disease. Second, the number of patients enrolled in the present study was small and the result, which showed that there was no difference in the navigation accuracy between the two approaches, even though each approach had 20 hips in order to detect adifferenceof3 in navigation accuracy according to the power analysis, may have been the product of a type 2 error. However, we believe that this CT-based navigation system is effective for MIS THA through both approaches, because the cup alignments with the navigation in both approaches were favourable in all cases. In conclusion, this CT-based navigation for MIS THA provides navigation accuracy without significant differences between the two approaches; the mean difference between the intra-operative and postoperative cup alignments is , with a variance of Even considering the effect of the press-fit procedure, this CTbased navigation also provides the favourable result that all cases are within 10 of the target alignment of the cup. References 1. Murphy SB, Ecker TM, Tannast M. THA performed using conventional and navigated tissue-preserving techniques. Clin Orthop Relat Res 2006; 453: Woolson ST, Mow CS, Syquia JF, et al. Comparison of primary total hip replacements performed with a standard incision or a mini-incision. JBoneJtSurgAm2004; 86A(7): Lewinnek GE, Lewis JL, Tarr R, et al. Dislocations after total hip-replacement arthroplasties. JBoneJtSurgAm1978; 60: Nishii T, Sugano N, Miki H, et al. Influence of component positions on dislocation computed tomographic evaluations in a consecutive series of total hip arthroplasty. JArthroplasty2004; 19: Sugano N, Nishii T, Miki H, et al. Mid-term results of cementless total hip replacement using a ceramic-on-ceramic bearing with and without computer navigation. JBoneJtSurgBr2007; 89(4): DiGioia AM, Plakseychuk AY, Levison TJ, et al. Mini-incision technique for total hip arthroplasty with navigation. J Arthroplasty 2003; 18(2): Malik A, Dorr LD. The science of minimally invasive total hip arthroplasty. Clin Orthop Relat Res 2007; 463: Mayr E, de la Barrera JL, Eller G, et al. The effect of fixation and location on the stability of the markers in navigated total hip arthroplasty: a cadaver study. JBoneJtSurgBr2006; 88(2): DiGioia AM, Jaramaz B, Blackwell M, et al. The Otto Aufranc Award. Image guided navigation system to measure intraoperatively acetabular implant alignment. Clin Orthop Relat Res 1998; 355: Murray DW. The definition and measurement of acetabular orientation. JBoneJtSurgBr1993; 75(2): Kennon RE, Keggi JM, Wetmore RS, et al. Total hip arthroplasty through a minimally invasive anterior surgical approach. JBone Jt Surg Am 2003; 85: (suppl 4): Herring JL, Dawant BM, Maurer CR Jr, et al. Surface-based registration of CT images to physical space for image-guided surgery of the spine: a sensitivity study. IEEE Trans Med Imaging 1998; 17(5): Watanabe Y, Masumoto J, Sasama T, et al. Preprocessing method for rigid registration between pre- and postoperative CT images in total hip replacement. Med Imag Tech 2003; 21(5): [in Japanese]. 14. Bosker BH, Verheyen CC, Horstmann WG, et al. Poor accuracy of freehand cup positioning during total hip arthroplasty. Arch Orthop Trauma Surg 2007; 127(5): Kalteis T, Handel M, Bathis H, et al. Imageless navigation for insertion of the acetabular component in total hip arthroplasty: is it as accurate as CT-based navigation? JBoneJtSurgBr2006; 88(2): Sugano N, Sasama T, Sato Y, et al. Accuracy evaluation of surface-based registration methods in a computer navigation system for hip surgery performed through a posterolateral approach. Comput Aided Surg 2001; 6(4):