Article Text

PDF

In-vivo pivot-shift test measured with inertial sensors correlates with the IKDC grade
  1. Giulio Maria Marcheggiani Muccioli1,
  2. Cecilia Signorelli2,
  3. Alberto Grassi1,
  4. Tommaso Roberti di Sarsina1,
  5. Federico Raggi1,
  6. Giuseppe Carbone1,
  7. Luca Macchiarola1,
  8. Vittorio Vaccari1,
  9. Stefano Zaffagnini1
    1. 1 Clinica Ortopedica II, Lab. Biomeccanica, University of Bologna, Istituto Ortopedico Rizzoli, Bologna, Italy
    2. 2 Istituto Ortopedico Rizzoli, NABI, University of Bologna, Bologna, Italy
    1. Correspondence to Dr Giulio Maria Marcheggiani Muccioli, Laboratorio di Biomeccanica ed Innovazione Tecnologica, Istituto Ortopedico Rizzoli, Bologna 40100, Italy; marcheggianimuccioli{at}me.com

    Abstract

    Objectives Kinematic Rapid Assessment (KiRA) is a wireless, non-invasive, inertial system with a single tibial sensor developed to measure the pivot-shift (PS) test. The purpose of this study was to in-vivo compare acceleration values acquired by KiRA to the objective International Knee Documentation Committee (IKDC) clinical grading of PS. The comparison was performed in non-anaesthetised patients before and after anterior cruciate ligament (ACL) reconstruction. We hypothesised the existence of a correlation between the side-to-side difference in the measured acceleration range by KiRA and the objective IKDC clinical grading of the PS.

    Methods Between 2010 and 2014, 60 non-professional football players (male/female ratio: 42/18; mean age 34±15.4 years, range 14–51 years) with ACL lesion were enrolled. They underwent over-the-top ACL reconstruction plus lateral extra-articular plasty with autologous hamstrings. All the patients were evaluated before the reconstruction and re-evaluated at 12-month follow-up. Each patient underwent a clinical examination and then was subjected to the instrumental PS examination by KiRA. The difference in the acceleration range between injured/reconstructed and contralateral limb (Δarange) was used in the analysis. Correlations between Δarange values and objective IKDC clinical grades of PS were calculated using Spearman correlation analysis.

    Results All subjective scores improved from preoperative to follow-up (P≤0.01). Objective IKDC clinical grading of the PS improved from 4B, 40C and 16D to 50A, 8B and 2C (P<0.0001). The mean Δarange measured by KiRA improved from 2.0±1.0 to 0.2±0.4 m/s2 (P<0.0001). A very strong correlation was displayed between the overall Δarange measured by KiRA and overall objective IKDC clinical grading of the PS (r=0.86, P<0.0001); correlation was strong for preoperative data (r=0.71, P<0.0001) and moderate for postoperative data (r=0.53, P<0.0001). The mean Δarange resulted 0.3±0.3 m/s2 for the IKDC A subgroup, 0.8±0.3 m/s2 for the IKDC B subgroup, 1.7±0.8 m/s2 for the IKDC C subgroup and 2.9±0.9 m/s2 for the IKDC D subgroup.

    Conclusion The side-to-side difference in the measured acceleration range by KiRA shows a correlation with objective IKDC clinical grading of PS.

    Study design Case series; level of evidence: 4.

    • knee
    • anterior cruciate ligament
    • biomechanics
    • pivot-shift
    • inertial sensors
    • accelerometers.

    Statistics from Altmetric.com

    What are the new findings

    • Kinematic Rapid Assessment (KiRA) is a new a non-invasive inertial system developed to measure the pivot shift (PS).

    • The present study showed a correlation between the side-to-side difference in the acceleration range (m/s2) measured by the KiRA system and the objective International Knee Documentation Committee clinical grading of the PS.

    Introduction

    The main objective of anterior cruciate ligament (ACL) reconstruction is correction of the joint laxity. However, even if the anteroposterior laxity correction is achieved, residual rotational laxity following ACL reconstruction has been identified as significant concern in many patients.1

    A correct instrumental assessment of knee joint laxity is a critical aspect in the management of the diagnosis treatment and rehabilitation process.

    At present, the pivot-shift (PS) test is the benchmark for ACL- injury evaluation, since it has been reported to be the most specific and highly correlated with complete or partial tear of the ACL.2 3 It is widely used for objective clinical assessment of rotatory knee laxity as part of the International Knee Documentation Committee (IKDC) objective score.4 The PS test involves combined loading of the joint, inducing movements in more than one degree of freedom during knee flexion–extension motion; this makes it difficult to perform and challenging to quantify.

    New PS measurement systems were introduced over the last 10 years, and some are non-invasive. Inertial sensor-based systems are reliable, non-invasive, low cost and easy to use.5–8

    In 2012, Lopomo et al 9 10 introduced and validated in vivo (against computer-assisted quantification of PS) a new non-invasive inertial system developed to measure the acceleration of the tibia during the PS test: the Kinematic Rapid Assessment (KiRA) triaxial accelerometer (Orthokey Italia srl, Firenze, Italy).

    Berruto et al 11 showed in clinical practice that the use of KiRA is both promising and reliable. What it was still lacking is a validation of this instrument against the objective IKDC clinical grading of the PS.

    The purpose of the present study was to compare the acceleration values acquired by KiRA to the objective IKDC clinical grading of the PS, at preoperative and 12-month follow-up after ACL reconstruction. We specifically hypothesised the existence of a correlation between the side-to-side difference in the measured acceleration range and the objective IKDC clinical grading of the PS.

    Material and methods

    Between 2010 and 2014, a total of 60 non-professional football players (male/female ratio: 42/18; mean age 34±15.4 years, range 14–51 years; mean time from injury to surgery 18 months, range 6–37 months) with ACL lesion were enrolled in this study.

    Exclusion criteria were defined as follows: (1) prior knee ligament/s reconstruction/s, (2) partial ACL lesions, (3) associated lesions of the collateral ligaments, (4) posterior cruciate ligament lesion, (5) important chondral defects (more than Outerbridge grade II and extended more than 1 cm2), (6) longitudinal and bucket handle meniscal tears involving more than 50% of the native meniscus, (7) contralateral knee surgery and (8) untreated contralateral knee lesions.

    Enrolled patients underwent over-the-top surgical reconstruction of the ACL plus lateral extra-articular plasty, performed by the two senior authors (GMMM and SZ). The surgical technique uses hamstring tendons, leaving intact their common tibial insertion, for intra-articular double-stranded reconstruction plus an extra-articular plasty performed with the remnant part of the tendons. Only one tibial tunnel was required (7 mm or 8 mm in diameter). The graft fixation was achieved by two low-profile titanium staples (8 mm) for the over-the-top and one low-profile titanium staple (6 mm) for the tibial fixation of the lateral plasty. The latter acts as a non-anatomic isometric anterolateral-ligament (ALL) reconstruction, running under the ileotibial band from the over-the-top point of graft fixation to a point just lateral and below the Gerdy’s tubercle.12

    Associated meniscal lesions were found in 33 patients (55%): 16 patients were treated with selective meniscectomy (11 medial and 5 lateral) and 17 patients were treated with meniscal suturing (9 medial, 5 lateral and 3 medial and lateral).

    Patients underwent a standardised rehabilitation protocol.12 A brace in full extension was applied for 2 weeks after the surgery only in the case of meniscal suture. Full weight bearing and unrestricted knee mobilisation were started on the first postoperative day; in the case of meniscal suture, they were started at week 4 postoperatively. All the patients were allowed to resume sport from 6 to 9 months after the surgery.

    All the patients were evaluated before the reconstruction and re-evaluated 12 months after the surgery. Each patient underwent a subjective evaluation, a clinical examination and then the instrumental PS examination.

    The subjective evaluation consisted of: global health state through the Physical Health Index and Mental Health Index of Short-Form 36 (SF-36) score13; assessment of physical activity through Tegner scale14; and subjective evaluation of knee functionality through subjective IKDC score.15

    The clinical examination included: side-to-side knee laxity evaluation through objective IKDC scoring system.4 The analysis was performed on awake patients by an expert consultant orthopaedic surgeon with more than 5 years of clinical practice, (GMMM), different from the operating surgeon, for all the enrolled subjects. The objective IKDC score considers four grades for PS grading: A (equal), B (glide), C (clunk) and D (gross).

    The instrumental PS examination was performed on awake patients using the KiRA system (Orthokey Italia srl) by the same single expert examiner (GMMM). This device consisted of a sensor embedding a triaxial accelerometer coupled with a gyroscope wirelessly connected to a common iPad (Apple, Cupertino, California, USA). The KiRA sensor was skin fixed on the tibial bone by a hypoallergenic strap between the lateral aspect of the anterior tibial tuberosity and the Gerdy’s tubercle to minimise skin artefacts during the manoeuvre (figure 1). The sensor measured the acceleration module and wirelessly transmitted the data to a customised programme (KiRA App for iPad) that analysed them. This module, purified from the gravitational component (1 G corresponds to 9.80665 m/s2), was used as the baseline signal for identifying the presence of the PS phenomenon. Once the system had automatically identified which part of the tracing is due to the PS, the output consisted of the following value: Δarange: amax – amin (acceleration variation).

    Figure 1

    Picture showing the positionament of the KiRA sensor on the leg between the lateral aspect of the anterior tibial tuberosity and the Gerdy’s tubercle during the pivot-shift manoeuvre.

    For both the clinical and instrumented evaluations, we performed the PS test using the technique (figure 2) learnt from Professor Hoshino during many years of collaboration.16 The side-to-side difference (between the injured/reconstructed and the healthy contralateral limb) in the acceleration range (Δarange) was used in the correlation analysis. Patients were divided into subgroups according to their IKDC clinical grading of the PS. For each subgroup, the corresponding mean Δarange was calculated.

    Figure 2

    Standardised pivot-shift manoeuvre used in this study.

    Statistical analysis

    The statistical analysis was performed with Analyse-it for Microsoft Excel (V.2.30) (Analyse-it Software, Leeds, UK).

    The non-parametric Wilcoxon and the paired Student’s t-test were used to prospectively assess outcomes changes, respectively, for non-parametric (Tegner activity scores) and parametric (SF-36 Physical Health Index, SF-36 Mental Health Index and subjective IKDC scores) variables.

    Statistically significant difference in the objective IKDC clinical grading of PS distribution between preoperative, and follow-up data were assessed by the Pearson’s χ2 statistics.

    Correlations between Δarange values and objective IKDC clinical grades of PS were calculated using the Spearman’s rank correlation coefficients. Correlation coefficient (r) was interpreted as weak (r<0.4), moderate (0.4<r<0.6), strong (0.6<r<0.8) and very strong (r>0.8).17

    For all the analysis, the level of significance was set at P=0.05.

    Results are expressed using mean values±SD for parametric values and median (IQR) for non-parametric values.

    Ethics

    Informed consent forms complied with European Union laws and were signed by the patients before the operation.

    Results

    All subjective scores significantly improved from preoperative to follow-up evaluation (P≤0.01) (table 1).

    Table 1

    Statistically significant improvements in subjective clinical scores from preoperative (pre-OP) to 12-month follow-up (12M FU)

    Objective IKDC clinical grading of the PS improved from 4B, 40C and 16D to 50A, 8B and 2C (P<0.0001).

    The mean Δarange measured by KIRA improved from 2.0±1.0 m/s2to 0.2±0.4 m/s2 (P<0.0001).

    A very strong correlation was displayed between the overall Δarange measured by KiRA and overall objective IKDC clinical grading of the PS (r=0.86, P<0.0001) (95% CIs 0.80 to 0.90). Correlation between the preoperative Δarange measured by KiRA and overall objective IKDC clinical grading of the PS was r=0.71 (P<0.0001; 95% CIs 0.55 to 0.81). While correlation between the postoperative Δarange measured by KiRA and overall objective IKDC clinical grading of the PS was r=0.53 (P<0.0001; 95% CIs 0.32 to 0.69).

    The correspondence between mean Δarange measured by KiRA and objective IKDC clinical grading of the PS subgroup (negative, glide, clunk and gross) was reported in table 2.

    Table 2

    Mean side-to-side difference in the tibial acceleration range (Δarange) measured by KiRA for each clinical grade of the pivot shift (PS) according to objective International Knee Documentation Committee (IKDC)

    Box plot distribution of Δarange divided according to IKDC classification for the whole data (pre and post together) is shown in figure 3. Box plot distribution of Δarange divided according to IKDC classification for preoperative data is reported in figure 4, while box plot distribution of Δa range divided according to IKDC classification for postoperative data is reported in figure 5.

    Figure 3

    Distribution of side-to-side difference in the tibial acceleration range (Δarange) measured by KiRA, divided according to objective IKDC classification, for the whole dataset. IKDC, International Knee Documentation Committee; KiRA, Kinematic Rapid Assessment.

    Figure 4

    Distribution of side-to-side difference in the tibial acceleration range (Δarange) measured measured by KiRA, divided according to objective IKDC classification, for the preoperative dataset. IKDC, International Knee Documentation Committee; KiRA, Kinematic Rapid Assessment.

    Figure 5

    Distribution of side-to-side difference in the tibial acceleration range (Δarange) measured by KiRA, divided according to objective IKDC classification, for the postoperative dataset. IKDC, International Knee Documentation Committee; KiRA, Kinematic Rapid Assessment.

    Discussion

    The most important finding of the present study was that the side-to-side difference (between the injured/reconstructed and the healthy contralateral limb) in the acceleration range measured by KiRA shows an in-vivo correlation with the IKDC clinical grading of the PS on non-anaesthetised patients.

    Several reviews have highlighted the importance, in present-day clinical practice, of quantifying the PS during the assessment of ACL injuries,18–20 also underlining the different technologies that have been developed to this end.21

    Recent studies using inertial sensors have found a correlation between the clinical grade of the PS and different kinematic parameters. Ahldén et al 22 discovered in vitro a moderate correlation (r=0.58) between the PS clinical grading and the maximal acceleration of the tibia. Kopf et al 23 found a significant in-vivo difference between the healthy and injured knees but found no correlation to the PS clinical grading. Labbé et al 24 recorded a very strong correlation (r=0.84) between the clinical PS grade and the amplitude of the drop in femoral acceleration at the time of reduction. Lopomo et al 9 pointed out a strong correlation (r=0.72) between the acceleration range measured by KiRA and the anteroposterior acceleration measured by a navigation system. Berruto et al 11 in a case series of 100 patients analysed by three physicians with different levels of expertise reported a good in-vivo interobserver reliability of KiRA (P<0.05) with good correlation to the PS clinical grades (r=0.7 for investigator 1, r=0.9 for investigator 2 and r=0.8 for investigator 3). They also highlighted that the analysed efficacy was strictly related to a learning curve and to the proper execution of the PS test. Nevertheless, in their work, they were not able to identify a single parameter that shows a very strong correlation with objective IKDC clinical grading of PS.

    In the present study, we defined the correspondence between mean Δarange measured by KiRA and each objective IKDC clinical grading of the PS subgroup (negative, glide, clunk and gross). Authors know that IKDC is a method that could be biased by the examiner, but it is well accepted that when performed by an expert examiner, it is reliable and consistent. Actually, objective IKDC is still considered the gold standard, worldwide-accepted scoring system in objective knee laxity measurement. It is included in all the most recent meta-analysis on this topic. This study reported good correlation between KiRA’s measurements and objective IKDC when executed by a single examiner. Multiple examiners might result in different results because of variability of manual testing between surgeons.25

    The KiRA system is connected by Bluetooth technology to an iPad (Apple), and this wireless and user-friendly setup allows the execution of the test by a single examiner. The sensor is applied to the proximal tibia between anterior tibial tuberosity and the Gerdy’s tubercle: this minimises skin artefacts during the manoeuvre.9 The system measures (in m/s2) the acceleration of the tibia on the femur during PS, thereby quantifying the result numerically.26 It measures Δarange of the tibia during the PS test. Δarange was chosen as reference value to evaluate the PS, and it was used in the correlation analysis, because this parameter showed the larger effect size (higher probability of discriminating between an injured and intact/reconstructed ACL) among all different parameters measured by KiRA, according to Lopomo et al 10 Moreover, the chosen position of the sensor is justified by the fact that the coupled anterior translation and the acceleration reached during the reduction of the lateral tibial compartment resulted to be the most influenced by the presence of PS phenomenon.27 28

    This is, to our knowledge, the first study to show a very strong correlation between a non-invasive wireless triaxial accelerometer with a single tibial sensor and the IKDC clinical grading of PS. The correlation was strong considering only the preoperative data and moderate considering only the postoperative data. This could be related to the abnormal distribution of data in this two very different groups (mostly C and D in the preoperative group and mostly A and B in the preoperative one). The advantage of using this non-invasive system with respect to the only objective clinical PS examination relies in the in-vivo quantification of dynamic rotatory knee laxity. The use of instrumental laxity assessment allows a more precise evaluation of the PS relative to the actual in-use clinical grading classifications (like the objective IKDC system). Moreover, such a patient-specific parameter could be useful to customise the surgery, especially in such a high demand population as football players.

    The present study had several limitations.

    The most important weakness of this study was that a single expert surgeon (GMMM) performed and graded the PS test on every knee. Kuroda et al were able to show that interphysician variability is high for the PS test, and the variation of the PS test technique results in different knee kinematics during the PSt test.29 Berruto et al 11 performed a clinical in-vivo validation of interobserver reliability of KiRA with good results (P<0.05). Moreover, we used a standardised PS manoeuvre (as reported by Hoshino et al 16 and proposed Δarange as the parameter of reference to evaluate the PS. In fact, measuring acceleration differences between injured and healthy/reconstructed knees (and not pure acceleration values) reduce the variables to a minimum and improve measurement accuracy between different subjects as showed by the very strong correlation with the clinical IKDC grading reported by the present study. Furthermore, this could improve also the measurement accuracy between different examiners. A clinical in-vivo validation of interobserver reliability of KiRA using Δarange as parameter of reference will be the topic of a future study.

    Another limitation was the following one: the inertial sensor that was used was applied over the skin. This has the potential to introduce skin-motion artefacts caused by the relative movement between the skin and underlying tissues. In a previous study we have shown that KiRA system had a good reliability and good correlation with the acceleration measured by a navigation system.9 Therefore, the KiRA is a valid method to assess dynamic joint laxity with negligible skin artefacts.

    Furthermore, authors choose 12 months as the follow-up of the present work because usually athletes usually come back to sport from 6 months to 12 months after the surgery.30 31 Twelve months is the moment when physicians need to quantify the amount of residual laxity to understand if the ACL-graft integration process is done and if the patient is ready to come back to unrestricted physical activity.

    Conclusion

    The side-to-side difference in the measured acceleration range by KiRA shows a correlation with objective IKDC clinical grading of PS.

    Acknowledgments

    The authors would like to thank the library’s staff of Istituto Ortopedico Rizzoli for their technical support.

    References

    View Abstract

    Footnotes

    • Contributors This paper has more than five authors. They all contributed significantly to the article, writing different parts of the manuscript. Moreover, they also participated to the work as the team who performed all the surgeries that was composed of GMMM and SZ; the team who supervised the postoperative rehabilitation was composed of LM, GC and VV. GMMM performed patients’ clinical evaluations, TRDS and AG collected demographic data and outcome scores. CS and FR performed the data elaboration and statistical analysis.

    • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Ethics approval Approval of the study was obtained from the Institutional Review Board.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Collaborators Bragonzoni Laura, Rinaldi Vito Gaetano, Lullini Giada, Ferretti Emil.

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.