A Radiographic Investigation Exploring Differences in Static Anterior Tibial Translation Expressed as a Percentage Between ACL-Deficient and -Intact Patients to Improve Interobserver Variability Between Different Centers

Abstract

Background:

Static anterior tibial translation (SATT) represents the amount of anterior translation due to axial load. It has been shown to be increased with anterior cruciate ligament (ACL) rupture, meniscal tear, and increased posterior tibial slope (PTS). It has also been shown to be correlated with ACL reconstruction failure. ACL reconstruction alone does not improve SATT. A sagittal plane slope-correcting osteotomy improves SATT, and SATT has recently been used to define the target slope correction after osteotomy. However, absolute values for SATT differ between institutions by >5 mm. Absolute measures differ based on the amount of magnification of the image, which varies based on the radiographic source to image distance, the source to object distance, rotation, and whether the medial or lateral condyle is presented to the source first. Scaled, or percentage radiographic measures, should correct for these differences.

Purpose:

To express SATT as a percentage (SATT%) of the medial plateau distance to improve accuracy and interinstitutional utilization of SATT.

Study Design:

Cross-sectional study; Level of evidence, 3.

Methods:

A consecutive series of patients without ligamentous or meniscal injuries between 2019 and 2022 was reviewed. A matched consecutive cohort of patients with nonacute ACL injuries (surgery between 6 and 12 weeks after injury) without concomitant pathology was reviewed. Preoperative SATT and PTS were measured with a previously validated technique on lateral weightbearing knee radiographs. Regression analysis was performed to investigate the relationship between SATT% and PTS.

Results:

There were 101 controls and 115 patients with an ACL injury who were included in this study. In the control cohort, the mean SATT% was 3.18% (SD, 5.92%) and mean PTS was 10.61° (SD, 3.28°). This was significantly different from our ACL cohort’s mean SATT% of 5.16% (SD, 7.41%) (P = .04) and mean PTS of 9.46° (2.85°) (P = .02). Linear regression analysis showed that for every 1° increase in PTS, there was a 0.08% increase in SATT% in the control cohort, so every 10° rise in slope was associated with an 0.8% increase in SATT%. In the ACL cohort, the effect of PTS on SATT% was larger; for every 1° of increase in PTS, there was an increase of 0.97% SATT%.

Conclusion:

The present study reports a reference SATT% value of 3.18% (SD, 5.92%) in a non–ACL injured cohort, which was lower than the ACL cohort’s mean 5.16% (SD 7.41%), despite the ACL cohort’s having a longer medial tibial plateau than the control population. The effect of slope on weightbearing anterior tibial translation was greater in the ACL population compared with the control cohort. These scaled, percentage values should improve the interinstitutional usage of SATT.

Static anterior tibial translation (SATT) is a radiographic assessment of the amount of tibial translation relative to the femur during a single-leg stance at 20° of flexion. It is therefore reflective of the axial load subjected to the anterior cruciate ligament (ACL) during the stance phase of gait as a result of translation. ⁹ It was originally described by H. Dejour and M. Bonin using ACL-injured patients, where they examined the ACL-injured leg and contralateral leg and demonstrated a side-to-side difference. However, more importantly, Dejour and Bonin demonstrated a direct correlation between SATT and posterior tibial slope (PTS), where every 10° of increased posterior slope was associated with 6 mm of increased SATT, regardless of whether the ACL was intact or not, meaning SATT is more affected by the slope than the ACL status of the knee. ⁹

The clinical importance of SATT is not reflected by a side-to-side difference between the injured leg and noninjured leg, but that increased PTS increases the SATT. It is known that increased PTS significantly increases ACL rupture rates ¹⁰ ; given the direct correlation between increased PTS and SATT, SATT may offer insight to this risk factor for failure of ACL reconstruction (ACLR). Biomechanical data have demonstrated a direct effect of increased anterior tibial translation and resultant increased ACL force associated with increasing PTS, which is most pronounced during midstance.^14,16,19,20 This supports the clinical validity of SATT’s being an in vivo measure of ACLR graft stress.

While ACLR improves dynamic laxity, clinically reflected in a reduced Lachman or anterior drawer test, ACLR does not improve SATT from preoperative values. ⁶ SATT of >6 mm has been independently associated with an odds ratio of 9.9 for ACLR failure in primary ACLR. ¹⁷ The only way to reduce the SATT is through a tibial deflexion osteotomy, where case series have demonstrated that a tibial deflection osteotomy decreases the tibial slope and subsequently decreases the SATT with no graft reruptures reported.^8,18 As such, SATT has been proposed as a way to better understand the biomechanical reason for increased graft rupture with increased tibial slope ¹⁰ and, subsequently, to help map the amount of slope correction with tibial deflection osteotomy. ⁴ More research is needed to understand this value.

Recently, normal values for normative controls (mean, 1.31 mm) and acute isolated ACL injuries (mean, 2.43 mm) have been defined. ³ Other studies have shown that increased SATT is associated with chronicity of the ACL injury, increased number of ACL ruptures, medial meniscal injury, tibial slope,^7,15 and increased risk of graft rupture with subsequent ACLR. ¹⁷ SATT within the same institution has good agreement and similar absolute values between studies with mean values differing by <0.1 mm.^3,6,7 However, the absolute values differ significantly when compared with studies from other institutions,^9,15,17 with means between institutions differing by up to 5 mm. This decreases the diagnostic utility of subsequent threshold values for SATT.

This is likely because of the fact that SATT is an absolute measure (in mm); it therefore is not scaled to the size of the patient and may differ based on the amount of magnification of the image, which varies based on the radiographic source-to-image distance, the source to object distance, rotation, and whether the medial or lateral condyle is first presented to the source. ¹²

Therefore, to decrease the variance in values of SATT between institutions due to magnification differences with imaging, we aimed to describe anterior tibial translation as a ratio or percentage of the tibial plateau distance. We hypothesize that SATT expressed as a percentage (SATT%) will still be correlated with PTS and increased in ACL-deficient knees compared with a control cohort.

Methods

Ethics

All patients provided informed consent for the use of their data for research, and the study was approved by the ethical board of GCS Ramsay Sante pour l’Enseignement et la Recherche.

Study Design

This is a retrospective radiographic cohort study that previously reported normal absolute reference values. ³ A matched cohort of isolated ACL-deficient patients without high-grade pivot shift or meniscal pathology were included between 2019 and 2022. The control population was a consecutive series of 101 patients without ligamentous or meniscal injuries from the same period that were prospectively added to our institutional registry. A matched cohort of ACL patients from the same registry was utilized.

Inclusion criteria were single-leg weightbearing lateral knee radiographs at 20° of flexion. ⁹ The exclusion criteria were age <15 years or inappropriate preoperative radiographs defined as ≥1 of the following: no superimposed posterior femoral condyles, <10 cm of the proximal tibia visualized, or the knee in a position of extension or flexion >30°. As part of our institutional protocol, all the radiographs were taken ≥1 month after the injury to mitigate the potential effect of acute injury. All surgeries were performed between 6 and 12 weeks after injury. We excluded patients with concomitant medial meniscal tear or lateral extra-articular procedures to avoid influencing the PTS effect.

Measurement

Radiographic measurements were performed on lateral knee radiographs. Two orthopaedic surgeon examiners (N.C., T.P.) independently reviewed the radiographs and calculated their measurements. Measurements were repeated twice by each reviewer. PTS and SATT measurements were performed using Horos Digital Imaging and Communications in Medicine viewer software (Version 3.3.6).

PTS was measured by calculating the angle between a line perpendicular to the tibial diaphysis and the medial tibial plateau according to Dejour and Bonnin, ⁹ because of its improved accuracy and reliability with this technique, as recommended in the ESSKA guidelines (Figure 1). ² SATT was measured as the distance between 2 lines parallel to the posterior tibial cortex, the first tangent to the posterior aspect of the medial tibial plateau and the second tangent to the posterior femoral condyles.^5,6 The medial plateau distance (MPD) was measured as the distance from the most superior aspect anteriorly to the most superior aspect posteriorly. The SATT was divided by the MPD before being multiplied by 100 to express the translation as a percentage (Figure 2).

Figure 1.

Measurement of posterior tibial slope (PTS), represented by angle a, calculated using the proximal tibial anatomic method. Line B is perpendicular to a line using midpoints of the proximal tibia, 5 cm and 10 cm (represented as the center of the circles on dotted line A), below the joint surface C. The PTS is the angle formed between lines B and C.

Figure 2.

Measurement of static anterior tibial translation (SATT) and SATT percentage. SATT is measured as the distance between 2 lines parallel to the posterior tibial cortex (dotted line A), the first tangent to the posterior aspect of the medial tibial plateau (line B), and the second tangent to the posterior femoral condyle (line C). SATT is the distance between lines B and C. The medial plateau distance (MPD) is the dotted line from the most superior point anteriorly (D) to the most superior point posteriorly (E). The SATT is then divided by the MPD and multiplied by 100 to convert the value to a percentage of translation.

Statistical Analysis

Continuous variables are expressed as the mean ± SD as appropriate, while the dichotomous variables are expressed as the number and percentage of patients. The Shapiro-Wilk normality test was used to assess the normality of distributions. A 2-tailed Student t test for independent samples was used to compare the mean values for PTS and SATT% between the control and ACL cohorts. Pearson correlation coefficient was performed to assess the correlation of sex, age, side, height, weight, and body mass index (BMI) with SATT% and slope. Univariate linear regression analysis was performed to determine the relationship between SATT% and PTS. SPSS (Version 25; IBM Corp) was used to perform these statistical analyses. Significance was set at an alpha of P < .05. The sample size and power were set from a previous investigation. ³

Results

Of the 104 controls, 101 ultimately met the inclusion criteria. Three control participants were excluded because of inadequate quality of lateral radiographs. A total of 115 ACL-deficient patients were included. Patient characteristics are presented in Table 1. There was a statistically significant difference in the mean age and weight between the cohorts (P < .001 and P = 0.05), and there was a statistical difference in sex that approached significance (P = .074). There was no statistically significant correlation between SATT and age (P = .26) or sex (P = .10), and there was no statistically significant correlation between PTS and age (P = .80) or sex (P = .11).

Table 1

Description of the Patient Groups ^a

	Control Cohort	ACL Cohort	P
Sex, male	36.6	48.7	.074
Side, right	51.5	52.2	.98
Weight, kg	65.09 (12.25)	69.20 (12.39)	.05
Height, cm	169.67 (10.44)	170.24 (9.18)	.74
BMI, kg/m2	22.49 (2.92)	23.79 (3.47)	.02
Age, y	22.91 (8.02)	32.81 (10.74)	<.001
MPD, mm	41.04 (7.07)	45.43 (8.27)	<.001
SATT, mm	1.31 (2.44)	2.27 (3.36)	.02
SATT, %	3.18 (5.92)	5.16 (7.41)	.04
PTS, deg	10.61 (3.28)	9.46 (2.85)	.02

a

Values presented as mean (SD) or percentage. Bold values indicate statistical significance (P < .05). ACL, anterior cruciate ligament; BMI, body mass index; MDP, medial plateau distance; PTS, posterior tibial slope; SATT, static anterior tibial translation.

There was excellent inter- and intraobserver reliability of SATT measurements (ICC, 0.99 and 0.97, respectively).

The mean SATT% was 3.18% (SD, 5.92), and the mean PTS was 10.61° (SD, 3.28) in the control cohort. This followed a normal distribution. There was a statistically significant correlation between SATT% and PTS (P < .01), with a Pearson correlation coefficient of 0.46, which was the same Pearson correlation coefficient for SATT (mm).

This was statistically significantly different from our ACL group’s mean SATT% of 5.16% (SD, 7.41%) (P = .04) and mean PTS of 9.46° (2.85°) (P = .02).

Linear regression analysis showed that for every 1° of increase in PTS, there was a 0.08% increase in SATT% in the control cohort (Figure 3), so every 10° rise in slope was associated with an 0.8% increase in SATT%. This effect was larger in the ACL cohort: for every 1° of increase in PTS, there was an increase of 0.97% SATT% (Figure 2), so again every 10° rise in slope was associated with a 9.7%-point increase in SATT%.

Figure 3.

Percentage SATT (SATT%) versus posterior tibial slope. Individual values for SATT% (% of static translation of tibia / y-axis) and posterior tibial slope (x-axis), with line of best fit plotted for the control cohort (blue) (y = 5.56 + 1 0.83x) compared with line of best fit plotted for the ACL cohort (red) (y = 3.84 + 1 0.97x).

There was no correlation between SATT, as an absolute value or as a percentage, and weight, height, or BMI.

The MPD was larger in the ACL-injured cohort, with the mean MPD at 45.43 mm (SD, 8.27 mm) compared with controls, at a mean of 41.04 mm (SD, 7.07 mm) (P < 0.01). To look at the effect of sex on this, a subanalysis was performed, and the results were still significant. For female patients, the MPD was larger in the ACL-injured cohort by a mean 3.34 mm (SD, 1.24) (P < .01) and in the male patients, larger by 4.80 mm (SD, 1.82) (P = .01).

When separating for sex, the MPD was correlated with height, with a Pearson correlation coefficient in female patients of 0.22 (P = .05) and in male patients of 0.28 (P = .04). No statistically significant correlation was seen between MPD and BMI or weight.

Discussion

The most important finding of this paper was that the SATT expressed as a percentage of the MPD is elevated in acute ACL-injured patients, compared with non–ACL patients, despite the ACL cohort’s having a longer medial plateau. The mean value in the non–ACL patients was 3.18% (SD, 5.92%) compared with 5.16% (SD, 7.41%) in the ACL-injured cohort.

Interestingly, the SATT did not correlate with either BMI, height, or weight of the patient. Given that SATT represents anterior tibial translation due to axial load, it would be intuitive that as the weight or BMI increased, so would the axial load and the subsequent anterior tibial translation during single-leg stance. Lachman-related laxity measures have shown an inverse correlation with BMI (ie, as BMI increases, laxity decreases, in both healthy children ¹¹ and adults ¹ ). Our findings are in agreement with previous findings by Dejour et al, ⁶ where SATT had no relationship with BMI.

While most stress radiographs for ligamentous injuries report side-to-side differences, ¹³ in the case of SATT, absolute values for the injured leg are reported instead of side-to-side differences. This is because SATT is affected primarily by the tibial slope and also by the integrity of the ACL. ⁹ SATT values reported by Cance et al ³ for ACL-injured patients were 2.43 mm, which is in agreement with Dejour et al, ⁶ who reported mean values of 2.3 to 2.4 mm and who additionally found no association with BMI. Medial meniscectomy increased this value by 2.39 mm, on average; therefore, we did not include meniscal injuries in our cohort. Both these studies included patients <6 months after ACL injury to remove chronicity. However, there is significant disagreement when these mean values are compared with mean values from other studies. Dejour and Bonnin ⁹ reported mean values of 2.9 mm in the contralateral uninjured leg and 6.4 mm in the ACL-deficient leg. Likewise, Macchiarola et al ¹⁵ reported mean values of 3.1 mm in the contralateral uninjured leg and 5 mm in the ACL-deficient leg, which increased to 7.2 mm 4 years after rupture and 9 mm in the revision ACL setting. These are significantly higher than our population. The agreement between the Cance et al ³ and Dejour et al ⁶ is likely due to these 2 studies being performed at the same institution, compared with the other 2 studies.^9,15 The size of the patient and, even more importantly, these mean values between institutions likely differ based on the amount of magnification of the image—which varies based on the radiographic source to image distance, the source to object distance, rotation, and whether the medial or lateral condyle is presented to the source first—and also likely differs between institutions. Thus, expressing values as a ratio or percentage will remove the error associated with these absolute measures. ¹²

The clinical importance of SATT is its association with PTS, and PTS is well known to be a significant risk factor for ACLR failure. ¹⁰ SATT has also been proven to be independently associated with increased odds of graft failure ¹⁷ and is not decreased by ACLR alone ⁶ but is only improved by a slope-reducing tibial deflexion osteotomy.^4,8 Moreover, SATT has been utilized to define the goal postoperative correction angle after sagittal plane osteotomy for ACL instability. ⁴ Therefore, the clinical utility of SATT as a measure is clear. Our method for expressing SATT% should improve the agreement of SATT values between institutions, allow the use of non–true size radiographs, and therefore increase the usability of threshold values such as those presented by Ni et al, ¹⁷ who found that a SATT of >6 mm increased the odds of rerupture after ACLR by 9.9 times, and the normative values found by Cance et al ³ now expressed as percentages in our study.

Limitations

There are limitations to this study. First, it is essential to note that this was a retrospective, radiographic, single-center study with limitations and biases associated with this design. A single-center study design limits the generalizability of our results; but by scaling our results, we should remove the effects of magnification between institutions, allowing other institutions to validate SATT normal results and utilize SATT to assess its effect on clinical outcomes. Additionally, our study cohorts exhibited differences in terms of age, sex, and BMI, although we conducted analyses to, concordantly with the available literature, confirm that age, sex, and BMI were not correlated with PTS and SATT. Furthermore, we reported exclusively on radiographs without direct correlation with clinical results. However, it is crucial to clarify that our study's primary aim was to perform an in vivo radiographic measurement, which was scaled to improve interinstitutional usage of this clinical laxity measurement.

Conclusion

The present study reports a reference SATT% value of 3.18% in a non–ACL injured cohort, which was lower than in the ACL-injured cohort’s mean 5.16%, despite the ACL cohort’s having a longer medial tibial plateau than the control population. The effect of slope on weightbearing anterior tibial translation was 10 times greater in the ACL population compared with the control cohort. These scaled, percentage values should improve the interinstitutional usage of SATT.