Relationship of Pelvic Positional Change with Leg Length and Offset Measurement in Experimental Total Hip Arthroplasty

Objective During total hip arthroplasty (THA), both pelvic and femur positions affect leg length (LL) and offset (OS) measurements because LL and OS calipers depend on the fixed reference points on the pelvis and femur, respectively. However, LL and OS measurement errors because of pelvic positional changes have not been described. This study aimed to clarify the effects of pelvic positional changes on LL and OS measurements in relation to the pelvic reference using a THA simulator. Methods We developed an experimental THA simulator using Sawbones models of the hemipelvis and femur that facilitated modification of the obliquity, tilt, and rotation of the pelvis. Using an LL and OS caliper, LL and OS measurement errors due to pelvic positional changes were determined with the femoral position fixed. Measurements were performed from two pelvic reference positions: the iliac tubercle (P1) and the top of the iliac crest intersecting the line of the femoral long axis (P2). Results Concerning pelvic obliquity, the total error of LL was 25.0 mm in P1 and 26.5 mm in P2, while the total error of OS was 13.0 mm in P1 and 10.9 mm in P2. For pelvic tilt, the total error of LL was 9.0 mm in P1 and 3.8 mm in P2, while the total error of OS was 0.5 mm in P1 and 1.0 mm in P2. Regarding pelvic rotation, the total error of LL was 13.8 mm in P1 and 3.2 mm in P2, while the total error of OS was 3.8 mm in P1 and 4.0 mm in P2. Conclusions Pelvic positional changes alter LL and OS measurements. The acceptable range (error <2 mm) on LL and OS measurement errors of pelvic obliquity was only 2°, regardless of the pelvic reference position. The pelvic reference position should be at the top of the iliac crest intersecting the line of the long axis of the femur because of a small LL measurement error with pelvic tilt and rotation.


Introduction
T otal hip arthroplasty (THA) is a standard treatment modality for end-stage hip arthritis. To achieve an optimal functional result after THA, it is necessary to properly adjust the leg length (LL) and offset (OS) within the hip. LL inequality after THA is a major cause of patient dissatisfaction due to unsatisfactory outcomes, such as limping, knee and back pain, early prosthesis loosening and revision surgery, and litigation. [1][2][3][4][5][6][7][8] Furthermore, improper OS adjustment also causes issues such as hip joint instability, decreased range of motion, and gait alteration. [7][8][9][10][11] To obtain appropriate LL and OS, preoperative planning is necessary to determine the extent to which LL and OS will change before and after surgery, and it is desirable to be able to confirm these changes intraoperatively. The LL and OS measurement instrument (LOMI; Smith & Nephew, Memphis, TN, USA) is a unique device that can detect changes in LL and OS before hip dislocation and after trial component or final implant insertion. 12 In clinical practice, a pin is inserted into the iliac tubercle through a stab incision for pelvic reference. Regarding the femoral reference, the lateral prominence of the greater trochanter is marked with electrical cautery. Then, the LOMI is set on a stopper of the iliac pin, and the LL and OS are measured before hip dislocation and after the trial component or final implant insertion (Fig. 1).
Given that the fixed points are distant from the center of rotation of the hip, the positions of the femur and pelvis affect LL and OS measurements. 6 To measure the LL and OS accurately, it is essential to maintain the pelvis and femur in precisely the same position between measurements. The effects of femur position on LL and OS measurements have already been investigated in different experimental settings. 13,14 The pelvic position is known to change during THA surgery, especially with pelvic tilt and obliquity, [15][16][17][18] resulting in variation in acetabular component orientation. 19 Similar to the femur, managing positional changes in the pelvis improves the accuracy of LL and OS measurements, but no studies have quantitatively examined the measurement error of LL and OS due to pelvic positional change. Moreover, the iliac tubercle, which is situated anterior to the hip rotation center, is known to be a non-ideal pelvic reference for accurate LL measurement. 14 Therefore, the purpose of this study was: (i) to clarify the effect of pelvic positional changes on measurement errors of the LL and OS; and (ii) examine the effect of two pelvic reference positions on measurement errors of the LL and OS using a THA simulator.

Experimental THA Simulator
A custom-made experimental device was developed to simulate THA surgery in the lateral decubitus position (Figs 1  and 2). In our previous study, the accuracy and repeatability of the LOMI on the THA simulator were found to be within 0.3 and 0.9 mm, respectively. 14 A 56-mm acetabular metal shell and an appropriate polyethylene liner (Reflection SP3 and XLPE, Smith & Nephew) were firmly fixed to a Sawbone model of the hemipelvis (model 1304, Sawbones, Vashon Island, WA, USA) with the usual cementless THA surgical technique. To change the position of the pelvis threedimensionally as desired, the hemi-pelvis was held using a free camera platform and a tripod (SLIK, Saitama, Japan). Using a strong bond (SU; Konishi, Osaka, Japan), a pedestal of the hemi-pelvis was attached to a square acrylic plate (900 cm 2 ; thickness, 5 mm) so that the longer edge of the pedestal was parallel to the plate. The center of rotation of the hip joint was projected onto the acrylic plate, and a 1/4-inch female screw was placed. Then, a male screw of the free camera platform was tightly twisted into the female screw. A #12 femoral broach of the Synergy Select™ hip system (Asian version of Synergy, Smith & Nephew, Memphis, TN, USA) was tightly seated into a Sawbone femur (model 1121, Sawbones, Vashon Island, WA, USA) using the usual femoral preparation. A high-offset trial neck with a 28-mm diameter and cobalt-chromium femoral head (+ 0 neck length) was attached to the femoral broach. A bar was fixed to the hole at the shoulder of the broach as an indicator of rotation. The bar was perpendicular to the long axis of the neck. The Sawbone model of the femur was held using a custom-made stand. Finally, the stand was fixed to a solid metal base board (width, 50 cm; length, 90 cm; and thickness, 2 cm) in the neutral femoral position (the long axis of the femur was horizontal and parallel to the longer edge of the base board, and the bar was horizontal). Two reference pins were placed in the ilium. The first pin was placed in the iliac tubercle (P1), and the second was placed at the top of the iliac crest, intersecting the line of the long axis of the femur when the pelvis was in the neutral position (P2). The pins were set perpendicular to the acrylic plate. The lateral prominence of the greater trochanter was chosen as the measurement point. A 3.5-mm cortical screw was placed to mark the point clearly.

Definition of Rotatory Directions
The position of the pelvis was divided into three rotatory directions according to the definition of Grammatopoulos et al., namely, obliquity, tilt, and rotation (Fig. 3). 19 Obliquity is a pelvic movement around the antero-posterior axis (coronal plane), with cranial and caudal obliquity corresponding to femoral adduction and abduction, respectively. Tilt is a pelvic movement around the transverse axis (sagittal plane), with anterior and posterior tilts corresponding to femoral flexion and extension, respectively. Rotation is a pelvic Fig. 1 The leg length and offset measurements during total hip arthroplasty. The LOMI is connected to a pin inserted into the iliac tubercle (arrow). The lateral prominence of the greater trochanter is marked with electrical cautery (arrowhead). The leg length and global offset are measured using LOMI. LOMI: leg length and offset measurement instrument movement around the longitudinal axis (axial plane), with forward and backward rotations corresponding to femoral external and internal rotations, respectively. When the acrylic plate was horizontal, obliquity and rotation were defined as 0 . When the four sides of the acrylic plate were parallel to those of the metal baseboard, the tilt was defined as 0 . The neutral position of the pelvis was defined as a 0 obliquity, tilt, and rotation.

Measurement of LL and OS Errors
To determine LL and OS measurement errors associated with pelvic position, the pelvic position was changed in one rotatory direction while the other rotatory directions were fixed. The range of obliquity was from 10 of cranial obliquity to 10 of caudal obliquity in 2 increments. The range of tilt was from 30 of anterior tilt to 30 of posterior tilt in 5 increments with reference to the scale of the free camera platform. The range of pelvic rotation was from 16 of forward rotation to 16 of backward rotation in 2 increments. For obliquity and rotation, the angle of the acrylic plate for each position was measured using a digital angle gauge (WR 300; Wixey, FL, USA). The LOMI was set to P1 or P2. LL and OS were measured three times for each pelvic position, and the mean values were calculated. The differences in the values of LL and OS between neutral and each pelvic position were defined as errors. Error of the pelvic position in one rotation is shown by line graphs. The line graphs were prepared using Microsoft Excel for Microsoft Office365 (Microsoft Corp, Redmond, WA, USA).

Analysis of Pelvic Obliquity
Pelvic obliquity had an enormous effect on the LL, irrespective of the pin position (Fig. 4). In both P1 and P2, the LL became longer in cranial obliquity and shorter in caudal obliquity; the total error was 25.0 mm for P1 and 26.5 mm for P2. Similarly, pelvic obliquity had a significant effect on the OS, irrespective of the pin position. In both P1

Analysis of Pelvic Tilt
The effect of the pelvic tilt on the LL was different for P1 and P2 (Fig. 5). For P1, the LL became shorter in anterior tilt and slightly longer in posterior tilt; the error was 9.0 mm in total. For P2, the LL became shorter in both anterior and posterior tilts; the total error was 3.8 mm. The pelvic tilt had an extremely small effect on OS, within 1.0 mm for both P1 and P2.

Analysis of Pelvic Rotation
The effect of pelvic rotation on the LL was also different between P1 and P2 (Fig. 6). For P1, the LL became longer in forward rotation, whereas it became shorter in backward rotation; the error was 13.8 mm in total. For P2, the LL became longer in both forward and backward rotations; the error was 3.2 mm in total. The effect of pelvic rotation on OS was similar for P1 and P2. OS became smaller with increases in forward and backward rotation; the error was 3.8 mm for P1 and 4.0 mm for P2.

Discussion
T his study revealed important findings. First, pelvic obliquity had a profound effect on LL and OS errors, irrespective of the pelvic reference position. Second, we showed that for pelvic tilt and rotation, the LL error was smaller for the reference on the top of the iliac crest than that for the iliac tubercle.

Impact of Positional Changes on LL and OS Measurements
Overall, the results of this study are consistent with those of previous femoral movement studies. 13,14 First, pelvic obliquity, which corresponds to hip abduction and adduction, had a profound effect on LL and OS errors, irrespective of the pelvic pin position. A pelvic obliquity of 20 resulted in an error of approximately 25 mm in the LL and 10.9-13 mm in the OS. In our previous study, LL and OS errors were approximately 24 mm and 13 mm, respectively, from 10 of adduction to 10 of abduction. 14 Under different experimental conditions, Sarin et al. found that 10 of hip abduction or adduction resulted in a 14-17 mm error for the LL and a 9.5-11.1 mm error for the OS. 13 Collectively, these results show that both pelvic obliquity and femoral abduction/ adduction greatly affect the LL and OS. Therefore, not only

Appropriate Pelvic Reference Pin Position
In the current study, the LL error for the pelvic reference in line with the long axis of the femur (P2) was smaller than that for the pelvic reference in the iliac tubercle (P1) for pelvic tilt and rotation. As for the pelvic tilt, the total LL error from 30 of anterior tilt to 30 of posterior tilt was smaller in P2 (3.8 mm) than that in P1 (9.0 mm). As for the pelvic rotation, the total LL error from 16 of forward rotation to 16 of backward rotation was much smaller in P2 (3.2 mm) than that in P1 (13.8 mm). These findings were in agreement with those of a previous report. 14 Therefore, the pelvic reference in line with the long axis of the femur when the pelvis is in the neutral position (namely, P2) should be chosen as the reference pin position in the ilium. Theoretically, no OS error should be observed because of the pelvic tilt because the tilt is movement in the sagittal plane. In this study, we observed an OS error due to a pelvic tilt of <1 mm. The result supports that the present measurements are fairly accurate. The effects of pelvic rotation on the OS were smaller than those on the LL. Moreover, the observed OS error was almost similar irrespective of the position of the pelvis pin.

Clinical Relevance of LL and OS Measurement Errors
During actual THA, modular heads are used to adjust the LL and OS. Given that the neck-shaft angles of the femoral stems are approximately 130 , a 3-mm neck-length change brings about an LL and OS change of approximately 2 mm. Therefore, a measurement error of less than 2 mm in the LL and OS is acceptable. 14 Table 1 shows the acceptable ranges of pelvic obliquity, tilt, and rotation according to the requirements. The acceptable range of pelvic obliquity for the LL and OS is extremely narrow, regardless of the pelvic reference. For pelvic tilt, the acceptable range in the LL is comparable to that in P1 (35 ) and P2 (40 ), although only 5 of anterior tilt is acceptable in P1. For pelvic rotation, the acceptable range in the LL is narrower in P1 and wider in P2, whereas the acceptable range in OS is wider, regardless of the pelvic reference. From the viewpoint of the acceptable range of LL error, P2 provides a relatively wide range of pelvic tilt and rotation. Asayama et al. 20 measured intraoperative pelvic motion in THA, using a posterolateral approach and reported that the ranges of pelvic obliquity, tilt, and rotation ranged from 6 of cranial obliquity to 9 of caudal obliquity, from 18 of anterior tilt to 10 of posterior tilt, and from 5 to 31 of forward rotation. Their results showed that LL and OS measurement errors could occur in actual THA. Furthermore, when using a portable navigation system, the accuracy of acetabular component placement has been reported to reduce with pelvic positional change in all three rotatory directions. 21 Therefore, minimizing pelvic positional change is desirable. One method for reducing pelvic motion is using the hip positioner, which fixes both the anterior superior iliac spine and sacrum as well as contributes to intraoperative pelvic stability. 16 Hence, pelvic positional change should be carefully managed.

Strengths and Limitations
To the best of our knowledge, this is the first study to quantitatively show that pelvic positional changes have an effect on LL and OS measurement errors. In previous experimental studies, the position of the pelvis was fixed, while the position of the femur was changed during LL and OS measurements. 13,14 In contrast, in the present study, the position of the femur was fixed while the position of the pelvis was changed. From the viewpoint of the relative movement between the pelvis and the femur, either experimental setting might yield similar results, although this has not been proven to date.
This study had some limitations. First, this result is probably limited to the experimental model of this study. In clinical practice, bone and implant sizes vary and the absolute value of measurement error may also vary. Second, although the pelvis could move in combination with obliquity, tilt, and rotation during an actual THA, the errors of LL and OS in the case of combined pelvic movement were not investigated. Third, not only the pelvis, but also the femur can move during the actual THA. 6,15 The combined movement of the pelvis and femur could either increase or decrease the errors in LL and OS measurements.

Conclusion
In conclusion, pelvic obliquity has an enormous effect on LL and OS measurement errors, regardless of the reference pin position. Given the small LL measurement error for pelvic tilt and rotation, the pelvic reference in line with the long axis of the femur when the pelvis is in the neutral position should be chosen as the reference pin position in the ilium.

Acknowledgments
T his research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author Contributions
A ll authors had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Disclosure Statement
T he authors declare that there is no conflict of interest.