Steep tibial slope correlates with inferior patient‑reported knee function independent of tunnel widening

Abstract

Purpose

The purpose of this study was to evaluate any association between posterior tibial slope (PTS), graft failure, clinical outcome and bone tunnel widening following anterior cruciate ligament reconstruction (ACLR) using a hamstring autograft.

Methods

A retrospective, case–control study was conducted comparing 350 revision ACL reconstructions (ACL‐RRs) with 350 primary ACL reconstructions (ACL‐Rs). A hamstring autograft was used for reconstruction in all cases. PTS and femoral/tibial tunnel diameters were measured on plain anteroposterior (AP) and on lateral radiographs after surgery, prior to ACL revision surgery and at follow‐up (FU). Clinical outcome was assessed at FU using the Lysholm score. Associations between PTS and Lysholm score were tested with linear regression (adjusted for age, sex and body mass index [BMI]). Tunnel enlargement was analyzed using Welch's test and correlated with PTS. The study demonstrated 97% power for the primary Lysholm model.

Results

ACL graft failure was significantly associated with a higher PTS (7.9 ± 1.8° in ACL‐R vs. 10.0 ± 2.1° in ACL‐RR, p < 0.001). At FU (35.5 ± 9.2 months), no significant widening of bone tunnels related to tibial inclination occurred, whereas subjective outcome was significantly affected (p < 0.001) by a steep PTS. In knees exceeding a PTS threshold of 10°, each degree increase in PTS was associated with a 1.1‐point decrease in Lysholm score. The difference in clinical outcome of knees with a PTS < 10° and  those ≥10° measured 5–6 points and exceeded the established minimal clinically important difference (≈5.5 pts). The proportion of patients reaching a patient acceptable symptom state (≥80 pts) lowered from ≈85% to ≈60%. Adjusted for sex and age, in subjects with a PTS ≥ 10°, the Lysholm score was approximately 4.3 points lower than in those with a PTS < 10°. Regardless of PTS, BMI had a negative effect on clinical outcome (p < 0.001).

Conclusion

An elevated PTS was not only associated with ACL graft failure but also contributed to poorer outcomes following ACLR. However, no association between PTS and bone tunnel expansion could be detected, suggesting other mechanisms for graft failure and reduced knee function.

Level of Evidence

Level III, retrospective comparative study.

Abbreviations

ACL

anterior cruciate ligament

ACL‐R

primary anterior cruciate ligament reconstruction

ACL‐RR

revision anterior cruciate ligament reconstruction

ANOVA

analysis of variance

AP

anteroposterior

BMI

body mass index

FU

follow‐up

ICC

intraclass correlation coefficient

ICRS

International Cartilage Repair Society

IQR

interquartile range

MCID

minimal clinically important difference

MRI

magnetic resonance imaging

PASS

patient acceptable symptom state

PTS

posterior tibial slope

INTRODUCTION

The posterior tibial slope (PTS) is recognized as a modifiable patient‐related risk factor for anterior cruciate ligament (ACL) graft failure [5, 6, 7, 35, 36, 37]. The PTS is defined as the angle formed between a tangent to the tibial plateau and the longitudinal axis of the tibia [5].

Biomechanically, an increased PTS is associated with increased anterior tibial translation, especially when knee valgus moments are present, posing higher forces on the ACL, finally leading to linear elongation of the ACL [36]. Additional external loads acting on the ACL further emphasize the correlation between PTS and ACL strain forces, indicating that a steep PTS may be particularly harmful for athletes generating large out‐of‐plane knee moments during landing tasks [1, 4, 15].

Slope‐correcting procedures have been developed to mitigate PTS‐related risks for ACL graft failure, yielding satisfactory functional outcomes [2, 9, 22, 26, 28, 29, 30, 31, 32].

However, the exact mechanism for ACL graft failure in knees with a steep PTS has not been fully resolved. Graft failure might be the result of graft lengthening, bone tunnel widening or both. We assumed that with increased anterior tibial translation, bone tunnel widening might occur due to graft impingement against the tunnel wall when the tibia translates anteriorly. Moreover, different inclination angles might affect the extent of graft failure being reflected in functional outcome. Therefore, this study aims to evaluate the influence of PTS on functional outcome and bone tunnel widening following ACLR.

We hypothesized that steeper PTS (≥10°) would be associated with a higher rate of ACL graft failure and lower Lysholm scores, and greater tibial tunnel widening, independent of age, sex and body mass index (BMI).

MATERIALS AND METHODS

A retrospective analysis was conducted on subjects who had undergone primary or revision ACL reconstruction (ACL‐R or ACL‐RR). All subjects were treated at our institution between 2010 and 2022. Subjects were divided into two groups: ACL‐R and ACL‐RR. Participants were assessed for age, sex, BMI and PTS. Bone tunnel diameters following ACLR were recorded after ACLR, prior to revision surgery in knees with a failed ACL graft and at follow‐up (FU). Patient satisfaction using the Lysholm score (range: 0–100) was assessed at FU of at least 24 months.

Inclusion criteria

In order to create a uniform study group and avoid graft‐related risks for retearing, only subjects who received a hamstring autograft in ACL‐R or ACL‐RR were included. Subjects younger than 17 or older than 45 years of age at the time of surgery were excluded as were subjects with missing documentation or unclear medical history. Subjects with a retear of the ACL less than 18 months after the index surgery were also excluded to rule out biological graft failure. Subjects were excluded if radiographs demonstrated signs of a previous lower extremity fracture or osteotomy, if their body habitus precluded appropriate visualization of critical radiographic landmarks, or if the lateral knee radiographs did not include at least 15 cm of the tibia distal to the joint line. Subjects demonstrating radiographic evidence for malplacement of the femoral, tibial or both drill tunnels following ACLR were also excluded. Knees demonstrating irreparable meniscal tears or subtotal meniscectomy or chondral lesions greater than International Cartilage Repair Society Grade 2 at the time of surgery were excluded, as were subjects with concomitant medial or lateral collateral ligament injury or posterior cruciate ligament injury.

A Strengthening the Reporting of Observational Studies in Epidemiology‐style screening diagram is presented in Figure 1. All subjects followed the same postoperative pathway, with visits scheduled at 3, 6, 12 and 24 months, and, when clinically indicated, annually thereafter. The questionnaire was administered at the latest routine visit that met the ≥24‐month requirement (mean 35.5 ± 9.2 months).

Figure 1.

Eligibility cascade for study cohorts. Flow diagram depicting the stepwise screening and selection of knees from the institutional ACL registry. ACL, anterior cruciate ligament; ACL‐R, primary anterior cruciate ligament reconstruction; ACL‐RR, revision anterior cruciate ligament reconstruction; FU, follow‐up; ICRS, International Cartilage Repair Society; mo, months.

Graft integrity was verified at every visit by Lachman and pivot‐shift testing. Magnetic resonance imaging (MRI) was triggered by equivocal findings.

Radiographic measurement

Lateral knee radiographs were used to determine the proximal anatomic PTS. The PTS was formed by the perpendicular to the proximal anatomic tibial shaft axis and the tangent to the medial tibial plateau.

Only radiographs that passed a predefined two‐tier rotation‐quality screen were used for any measurement; the full exclusion algorithm (posterior‐condyle overlap ≤ 3 mm, centred‐patella criteria, etc.) is detailed in Supplement 1. Using this filter removed 412 of 5034 knees (8.2%) before cohort matching.

In anteroposterior (AP) radiographs, femoral and tibial tunnel diameters were measured at a depth of 10 mm below the tunnel exit point (Figure 2). In lateral radiographs, the femoral tunnel diameter was measured at its exit point parallel to Blumensaat's line, whereas the tibial tunnel diameter was determined 10 mm below the tunnel exit point (Figure 1). Differences in tunnel diameters at the time points (directly after ACLR, prior to revision in cases with graft failure and at FU) defined bone tunnel expansion (i.e., tunnel widening) and were reported in millimetres. Measurements were conducted using standardized digital imaging software.

Figure 2.

Radiographic analysis of tibial slope and bone tunnel diameters following ACL reconstruction. (A) Anteroposterior (AP) radiograph showing measurement of femoral and tibial tunnel diameters at 10 mm depth from the respective tunnel exit points. The green dashed line highlights the tibial and femoral tunnel, as well as two perpendicular lines that represent the 10 mm depth from the tunnel exit points; the yellow line represents the measured diameter. (B) Lateral radiograph used to assess the proximal anatomic posterior tibial slope (PTS), defined as the angle between a line perpendicular to the proximal tibial shaft axis and the tangent to the medial tibial plateau. Additionally, femoral and tibial tunnel diameters are measured at standardized positions: at the femoral tunnel exit parallel to Blumensaat's line and 10 mm distal to the tibial tunnel exit. Green line (tibial tunnel), orange line (PTS measurement), blue line (double‐circle method to measure the PTS), yellow line (tunnel diameter measurements and 10 mm distance from the tunnel exit point). ACL, anterior cruciate ligament.

Femoral and tibial tunnel positioning was assessed using the Bernard–Hertel and Montreuil anatomic zones [8, 21]. Knees with gross malplacement (2.5% of the initial pool) were excluded.

In order to minimize investigator‐related bias of the radiographic measurements described, inter‐ and intraobserver reliability was assessed. Two observers (A.D. and D.Z.), trained fellows in sports medicine, independently evaluated the PTS, femoral and tibial bone tunnel diameters in a subset of 50 randomly selected patients. Measurements were repeated by each investigator 2 weeks apart.

Statistical analysis

All statistical analyses were performed using R (version 4.2.3, R Foundation for Statistical Computing). A p value < 0.05 was considered statistically significant. Continuous variables were summarized with means, standard deviations, ranges and interquartile ranges (IQRs); categorical variables were summarized with frequencies and percentages.

Linear regression was applied to test for correlation. Groups were frequency‐matched on age and sex to minimize gross imbalance; however, within‐stratum variability was not eliminated by frequency matching. The ACL‐R and ACL‐RR cohorts were frequency‐matched on age and sex, meaning their overall age‐band distributions and male‐to‐female ratios were made similar, instead of performing a 1:1 individual match. Nevertheless, these parameters were retained as covariates in subsequent regression models to absorb residual within‐stratum variability, lower the standard error of the PTS coefficient by approximately 10%, and align with methodological guidance indicating that matched variables should remain in the outcome model to avoid bias. Only age and sex were used in the matching algorithm; BMI and other clinical variables were intentionally left unmatched to allow for natural variability to be modelled as covariates in multivariable analyses.

To test for differences in tunnel widening, Welch's two‐sample t test was used. Since the two cohorts were created with frequency (group‐level) matching rather than 1:1 pairwise matching, individual observations remained statistically independent, making an independent‐samples test appropriate. Welch's version was selected over Student's t test because unequal variances were indicated by Levene's test in two of the four tunnel‐expansion metrics. Normality was verified with Shapiro–Wilk tests (all p > 0.05), and a non‐parametric sensitivity check using the Mann–Whitney U test reproduced the same conclusions (all p values within ±0.01).

The primary binary explanatory variable (PTS < 10° vs ≥10°) was defined a priori since large clinical studies reported sharply elevated graft‐failure risk when the slope exceeded approximately 10°; for example, 10° was identified as the inflection point for a ≥5‐fold increase in ACL injury or retear [11, 13].

A two‐way analysis of variance (ANOVA) (factors: PTS category <10° vs. ≥10°, surgery type ACL‐R vs. ACL‐RR; covariates: age, sex and BMI) was pre‐specified to test for a PTS × surgery‐type interaction on Lysholm score.

Observed power for the primary Lysholm model was calculated at 0.9; full power calculations and sensitivity analyses that repeated the model at the 8°, 10°, 12° and 13° PTS cut‐offs are presented in Supplement 1.

RESULTS

From 5034 ACLRs in the institutional registry, 4334 knees were excluded through a nine‐stage screening cascade (Table S2, Figure 1). The primary reasons for exclusion were non‐hamstring graft type (n = 1241; 24.6%), concomitant pathology (n = 986; 19.6%), inadequate radiographic quality—including mal‐rotation (n = 412; 8.2%)—and tunnel malplacement (n = 127; 2.5%). This resulted in 851 knees meeting all baseline criteria; 700 of these (350 ACL‐R and 350 ACL‐RR) were available for clinical and radiographic FU at 35.5 ± 9.2 months (range: 24–174), yielding a FU‐rate of 82.2%.

Age and sex distributions were comparable across cohorts, while BMI was ~1 kg/m² higher in ACL‐RR (Table 1).

Table 1.

Baseline characteristics of the study population.

Variable	ACL‑R (n  =  350)	ACL‑RR (n  =  350)	p
Age, y (mean ± SD)	27.5 ± 5.9	28.1 ± 5.3	n.s.
Sex, F/M	133/217	160/190	n.s.
BMI, kg/m2 (mean ± SD)	24.0 ± 3.7	25.1 ± 4.0	<0.001
Follow‑up, mo (mean ± SD)	35.3 ± 9.0	35.7 ± 9.4	n.s.
Posterior tibial slope, ° (mean ± SD)	7.94 ± 1.79	10.0 ± 2.08	<0.001

Note: Values are presented as mean ± standard deviation (SD) unless otherwise indicated. p values derive from Welch's two‑sample t tests for continuous variables and chi‐square tests for categorical variables.

Abbreviations: ACL‐R, primary anterior cruciate ligament reconstruction; ACL‐RR, revision ACL reconstruction; BMI, body‐mass index; F/M, number of females/males; mo, months; n.s., not significant (p > 0.05), y, years.

PTS and bone tunnel measurement proved excellent interrater reliability (intraclass correlation coefficient [ICC] ≥ 0.9; Table 2). Objective tunnel‐placement data for the 350 revision cases are presented in Supplement 1. More than 99% of the bone tunnels were within the anatomic zones.

Table 2.

Inter‐ and intra‐observer reliability for radiographic measurements.

Metric	Inter‑observer ICC (95% CI)	Intra‑observer ICC (95% CI)
Posterior tibial slope	0.93 (0.88–0.96)	0.95 (0.91–0.97)
Femoral tunnel Ø – AP	0.91 (0.85–0.95)	0.94 (0.89–0.97)
Femoral tunnel Ø – Lat	0.92 (0.87–0.96)	0.94 (0.90–0.97)
Tibial tunnel Ø – AP	0.90 (0.83–0.95)	0.93 (0.88–0.96)
Tibial tunnel Ø – Lat	0.92 (0.86–0.96)	0.94 (0.90–0.97)

Note: Intraclass correlation coefficients (ICCs) with 95% confidence intervals (CIs) are shown.

Abbreviations: AP, anteroposterior; Lat, lateral; Ø, diameter.

The PTS was approximately 2° steeper in the ACL‐RR cohort (Table 1, Figure 3), confirming the expected group difference.

Figure 3.

Boxplots comparing posterior tibial slope (PTS) in degrees for primary ACL reconstruction (ACL‐R) versus revision ACL reconstruction (ACL‐RR). Each box depicts the median (horizontal line within the box) and interquartile range (box boundaries). Whiskers extend to 1.5 times the interquartile range from each quartile, and individual points beyond the whiskers denote outliers. Overall, the ACL‐RR cohort exhibits a higher median PTS distribution than the ACL‐R cohort, indicating a steeper tibial slope among patients undergoing revision surgery. ACL, anterior cruciate ligament.

No differences with respect to bone tunnel enlargement were found between the groups. Even when analyzing irrespective of group belonging, no significant differences in bone tunnel diameters could be detected when comparing knees with a PTS < 10° to those with a PTS ≥ 10° (Table 3).

Table 3.

Tunnel enlargement and Lysholm score outcomes stratified by surgery type and posterior tibial slope (PTS) category.

Outcome	ACL‑R	ACL‑RR	p	PTS < 10°	PTS ≥ 10°	p
Tunnel enlargement, mm (mean ± SD)
Femoral Ø Δ AP	1.62 ± 0.40	1.61 ± 0.44	n.s.	1.59 ± 0.42	1.64 ± 0.44	n.s.
Femoral Ø Δ Lat	1.74 ± 0.42	1.76 ± 0.40	n.s.	1.70 ± 0.45	1.75 ± 0.37	n.s.
Tibial Ø Δ AP	2.72 ± 0.34	2.72 ± 0.35	n.s.	2.71 ± 0.34	2.71 ± 0.38	n.s.
Tibial Ø Δ Lat	3.19 ± 0.33	3.21 ± 0.35	n.s.	3.21 ± 0.30	3.16 ± 0.42	n.s.
Lysholm score
Follow‐up at Lysholm assessment, months (mean ± SD)	35.3 ± 9.0	35.7 ± 9.4	n.s.	‐	‐	‐
Mean ± SD	89.2 ± 5.3	78.8 ± 4.9	<0.001	86.8 ± 6.7	81.4 ± 6.2	<0.001
PASS ≥ 80%	93.7	50.3	<0.001	84.6	58.9	<0.001

Note: Values are mean ± SD unless otherwise noted.

Abbreviations: ACL, anterior cruciate ligament; ACL‐R, primary ACL reconstruction; ACL‐RR, revision ACL reconstruction; AP, anteroposterior radiograph; Lat, lateral radiograph; mo, months; ns, not significant; PASS, patient acceptable symptom state (Lysholm ≥ 80); SD, standard deviation; Δ, change from baseline to final follow‐up; Ø, diameter.

The ACL‐R group demonstrated an average Lysholm score of 89.2 ± 5.2 (74.3–99.0, IQR: 7.5, 35.3 ±  9.0 months FU), compared to 78.8 ± 4.9 (67.0–95.2, IQR: 6.4, 35.7 ±  9.4 months FU) in the ACL‐RR cohort (Table 3, Figure 4).

Figure 4.

Boxplots comparing Lysholm scores between the primary ACL reconstruction (ACL‐R) group and the revision ACL reconstruction (ACL‐RR) group. The Lysholm score was collected at a mean 35.5‐month follow‐up (FU) in the ACL‐R group, and at a mean 35.5‐month FU in the ACL‐RR group. Mean FU did not differ between cohorts (two‑sample t test, t = 0.5, p = 0.5; Mann–Whitney U test, p = 0.5). The box denotes the interquartile range, with the horizontal line indicating the median. Whiskers extend to 1.5 times the interquartile range, and individual points beyond that range appear as outliers. The ACL‐R group demonstrates a higher median Lysholm score compared to ACL‐RR, indicating better self‐reported knee function among primary reconstruction patients. ACL, anterior cruciate ligament.

When regrouping subjects (350 primary + 350 revisions) depending on tibial inclination, knees with a PTS < 10° achieved an average Lysholm score of 86.8 ± 6.7 (67.4–99.0) compared to 81.4 ± 6.1 (67.0–99.0) in knees with a PTS ≥ 10° (p < 0.001). Thus, subjects with a PTS ≥ 10° achieved an average Lysholm score that was more than 5 points lower than those with a PTS < 10°. However, in the multivariable regression model adjusting for age, sex and BMI, the adjusted difference between PTS < 10° and PTS ≥ 10° was noted as 4.3 points (p < 0.001) (Figure 5). When analyzing the PTS continuously, in knees with a PTS ≥ 10°, each degree increase in PTS was associated with a 1.1 decrease in Lysholm score (β = −1.1, p < 0.001) (Figure 4). Patient acceptable symptom state (PASS) was also reduced by a slope ≥10° from 85% to 59%.

Figure 5.

Partial effect plot from a generalized additive model (GAM) controlling for age, sex, and BMI, illustrating how the outcome changes across the predictor of interest (horizontal axis); the blue ribbon denotes the 95% confidence interval, and the dashed red lines highlight key threshold values. BMI, body mass index.

Applying the commonly used cut‐off value of 12° to the population yielded similar patterns of association. Knees with a PTS > 12° reached a Lysholm score of 79.4 ± 6.1 (n = 123) versus 86.0 ± 6.7 (n = 577) in knees with a PTS ≤ 12°; the 6.6‐point gap was statistically significant (p < 0.001).

Using the two‐way ANOVA, no interaction between PTS category (<10° and ≥10°) and group belonging (ACL‐R or ACL‐RR) on Lysholm score could be detected (interaction p = 0.58), indicating that the effect of PTS on Lysholm score was similar in both ACL‐R and ACL‐RR. When ‘revision status’ was added as a covariate to the continuous‐PTS regression, the slope coefficient was found to change by <2% (β = −1.09 pts/deg), confirming that PTS was not acting as a proxy for revision surgery.

BMI had a modest independent association (−0.2 pts per kg/m², p < 0.001). In both models, whether the PTS was treated continuously (coefficient = −0.23, p < 0.001) or dichotomously (coefficient = −0.27, p < 0.001), each 1‐unit increase in BMI predicted a modest but statistically significant decrease (0.23–0.27 points) in Lysholm score.

Post hoc power was determined to be 0.9 for the observed PTS effect, and residual diagnostics revealed no model violations. Low‐grade meniscal or cartilage lesions (13% of knees) were found to be unrelated to Lysholm score or tunnel enlargement and did not influence any PTS estimates. Inclusion of these covariates did not materially change the effect estimates for PTS on graft failure, Lysholm score or tunnel dimensions (maximum change in β = 3.2%). Model fit (adjusted R ²) improved by <0.01, indicating negligible explanatory gain.

DISCUSSION

The main finding was that a steep PTS was associated with ACL hamstring graft failure without affecting bone tunnel diameters. Clinical outcome was negatively influenced by PTS, with significantly lower Lysholm scores in knees with a PTS exceeding 10°.

Significant differences in PTS values were found between ACL‐R and ACL‐RR. Thus, a steep PTS was associated with hamstring graft failure. Clinical outcome was less favourable in ACL‐RR compared to ACL‐R, with significant differences in Lysholm scores. When subjects were analyzed independent of group belonging, there was a negative association between PTS and clinical outcome. In knees with a PTS exceeding 10°, every increase in PTS was linked to a 1‐point decrease in Lysholm score. The association between clinical outcome and PTS was not influenced by concomitant intra‐articular pathology, as adjustments for low‐grade meniscal or cartilage lesions left the point estimates for PTS essentially unchanged. Other studies attributed poorer results to higher rates of meniscal injuries and chondral damage in the revision group [16, 19, 34, 36]. In contrast to these findings, this study was designed to minimize the impact of meniscal or cartilage injuries on subjective outcomes by creating uniform study groups according to the inclusion criteria applied.

The PTS is recognized as an important risk factor contributing to ACL graft failure, particularly in ACL revisions. However, there is still debate on the critical value for the PTS. Song et al. and Dejour et al. proposed thresholds of 13° and 13.2°, respectively, with favourable outcomes after slope‐reducing high tibial osteotomy [9, 29]. Ni et al. suggested a cut‐off of 17° [25], whereas de Sousa Filho et al. found that a PTS exceeding 8° tripled the risk of ACL injury [10]. Fares et al. reported that a 10° threshold carried a fivefold risk of ACL tears relative to healthy controls [14]. In a 15‐year FU, Webb et al. demonstrated that a PTS exceeding 12° bore a fivefold risk for sustaining another ACL injury [33]. More recently, Dracic et al. proved that knees with a PTS exceeding 10° were associated with an 11‐fold risk for ACL graft failure [13].

In the current study, knees with a PTS of less than 10° were found to have an approximately 5‐point higher average Lysholm score compared to those with a PTS exceeding 10°. Further evaluation revealed that in knees with a PTS exceeding 10°, a 1° increase in PTS was associated with a 1.1° decrease in Lysholm score. Only a few studies have been conducted investigating the influence of PTS on clinical outcomes. Hinz et al. could not detect any relationship between the International Knee Documentation Committee, Lysholm or Tegner activity score with the PTS [18]. However, these studies used different thresholds (5.6° and 12°) for the PTS.

Although the absolute score reduction associated with a steep PTS numerically appears modest, it is clinically consequential. The 5‑to‑6‑point decrement straddles the accepted minimal clinically important difference (MCID) for the Lysholm score after ACLR (≈5.5 pts) and translates into a 26‑percentage‑point absolute fall in patients who attain a symptom state they regard as satisfactory (PASS ≥ 80) [12]. Put differently, elevating the PTS from 10° to 14° predicts an ~8‑point loss—well beyond every published MCID—and cuts the chances of feeling ‘acceptable’ from roughly five in six to three in five. These data confirm that a PTS ≥ 10° not only carries statistical but also clinical consequences for everyday knee function.

With higher tibial inclination, an increase in anterior tibial translation occurs, and a larger force is exerted on the ACL, potentially leading to graft insufficiency due to widening of the ACL bone tunnels [1, 15]. Only a few studies have investigated the impact of the PTS on bone tunnels after ACLR. Nakazato et al., Luo et al. and Sabzevari et al. found an association between an increased PTS and tibial bone tunnel expansion following ACLR [20, 24, 28]. However, tibial tunnel expansion had no effect on knee function [24]. In the current study, lateral radiographs of knees with a PTS exceeding 10° revealed an expansion of the tibial bone tunnel. However, the effect just missed statistical significance. Thus, a change in bone tunnel morphology did not seem to mediate graft failure or reduced knee function.

A high BMI is acknowledged as a risk factor for ACL graft failure [23]. However, conflicting results exist regarding the influence of BMI on subjective outcomes [3, 11, 17, 26]. Recent studies provided evidence that overweight patients with a BMI exceeding 25 kg/m² have a poorer outcome following ACLR [17, 26]. This study supports the finding that overweight patients with a BMI exceeding 25 kg/m² experience poorer outcomes following ACLR. The effect of BMI on subjective outcomes appeared to be present, but the main impact on graft failure and subjective outcomes was attributed to the PTS.

The current findings underscore that a slope of 10° or more independently increases the likelihood for ACL graft failure and reduces Lysholm scores by amounts exceeding the MCID. Slope‐reducing osteotomy should thus be considered in high‐slope knees, particularly when revision surgery is planned. Even outside elite sports settings, a steep tibial slope imposes a measurable biomechanical burden that translates into poorer subjective knee function, making routine assessment of slope advisable.

Strengths and limitations

This study possesses several strengths that enhance the credibility of its findings. First, a large and well‐defined cohort provides over 80% statistical power for every primary endpoint. Second, strict inclusion and exclusion criteria eliminated high‐grade meniscal or chondral pathology, allowing for the examination of PTS and BMI in relative isolation. Third, highly reproducible measurement techniques and frequency matching on age and sex, followed by multivariable adjustment, produced balanced study arms and minimized residual confounding.

Several methodological and population features explain why a clinically relevant slope effect was detected in this work when it was not in earlier reports. The cohort primarily consisted of recreational athletes, with a mean BMI at the lower border of ‘overweight’, and biomechanical models suggest that in such heavier, less‐conditioned knees, a slope of 12° or more can increase graft strain by 15%–20%, leading to subtle but symptom‐generating micro‐instability.

The retrospective nature of the study may introduce selection bias related to patient inclusion and exclusion criteria.

Bone tunnel diameters were determined in plain radiographs using standardized digital imaging software. Without the use of a measuring sphere for calibration, there might be some inaccuracies in the measurement of bone tunnel diameters. However, computed tomography and 3D MRI provide higher spatial resolution and have been shown to detect subtle tunnel enlargement that standard radiographs can miss or underestimate. Accordingly, our measurements may under‑represent the absolute magnitude of tunnel widening. However, the excellent inter‑ and intra‑observer reliability (ICC ≥ 0.9) suggests that any modality‑related error is non‑differential between groups and thus would bias slope‑related comparisons towards the null.

Detailed neuromuscular control patterns—that is, the timing and coordination of quadriceps, hamstrings and hip‑stabilizer activation that provide dynamic joint stability—were not consistently recorded. Consequently, unmeasured variation in these patterns could have influenced both subjective and objective measures of knee performance (e.g., Lysholm score, range of motion and strength) as well as graft integrity and clinical success. Finally, despite the region‐wide electronic medical record and active patient queries, the possibility that a small number of participants underwent ACL‐RR at unaffiliated centres cannot be entirely excluded.

CONCLUSION

A steep PTS exceeding 10° is considered a risk factor for ACL graft failure and appears to impact subjective outcomes. In knees with a PTS exceeding 10°, every degree increase in PTS was associated with a 1.1‐point decrease in Lysholm score. In contrast, the PTS did not affect bone tunnels following ACLR.

AUTHOR CONTRIBUTIONS

Anel Dracic has co‐designed the study, collected data, performed statistical analysis, interpreted data and co‐edited the manuscript. Domagoj Zeravica contributed to data collection. Semin Becirbegovic and Marcus Jäger critically revised the manuscript. Sascha Beck has co‐designed the study, interpreted results and co‐edited the manuscript. All authors read and approved the final manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The study was approved by the Ethics Committee of the University of Duisburg‐Essen (study number 17‐7523‐BO). All subjects gave their informed consent to participate in the study.

Supporting information

Supplement_1.

DATA AVAILABILITY STATEMENT

The data sets generated and analyzed in the current study are available from the corresponding author on reasonable request.