Anatomical factors associated with the development of anterior tibial spine fractures based on MRI measurements

Anterior tibial spine fractures (ATSFs) or tibial eminence fractures are avulsion fractures of the anterior cruciate ligament (ACL) from its insertion on the tibial intercondylar eminence. ATSFs are rare, with an incidence between 3.0 and 3.5 per 100,000 individuals in the general population [1, 2]. These fractures commonly occur in children and adolescents, predominantly between the ages of 8–14 [3‐5]. However, the recent literature suggested that the prevalence of ATSFs in adults is higher than previously reported [6, 7]. The etiology of fractures is various; for pediatric patients, it may be due in part to skeletal immaturity, weak knee muscle tissue and increased ligament elasticity, while for adults, it is mostly due to traffic accidents. The most typical mechanism of injury is knee hyperextension with a valgus or rotational force [8], often resulting from a fall from a bicycle. But these fractures are increasingly common in noncontact injuries in sports, such as skiing and soccer [9].

According to the Meyers and McKeever classification system [10], ATSFs can be classified into 4 types. Type I injury is a minimally displaced fragment. Type II injury involves superior displacement of the anterior bony fragment, while type III and IV injuries involve complete separation of the fragment from the tibia. Type IV injury includes comminution of a displaced avulsion fracture. The different treatment options for ATSFs are full of hot debate, but there is a paucity of literature on the anatomical risk factors for these fractures.

Samora et al. [11] included 25 pediatric patients to evaluate risk factors for ATSF and found no significant differences in posterior tibial slope (PTS) and notch width index (NWI) compared to controls. A previous study by Messner et al. [12] explored the relationship between posterior tibial slope (PTS) and pediatric ATSFs. They found that only male patients undergoing surgical fixation of ATSF had an increased PTS compared with controls. However, their study population was pediatric patients, and their measurement of PTS was performed by plain radiographs, limiting their ability to distinguish the subtle differences between the medial and lateral compartments. In addition to PTS, we want to further explore morphological risk factors for ATSFs of the tibial plateau and femoral condyle. Various osseous morphologic risk factors associated with ACL injuries have been identified in the literature, such as shallower medial tibial depth [13], increased lateral tibial plateau slope [14, 15] and intercondylar notch stenosis [16]. Therefore, the purpose of this study was to determine which anatomic parameters are independently associated with ATSF in adults (1) and the diagnostic values of the individual anatomic parameters (2). It was hypothesized that patients with ATSF would exhibit elevated PTS compared to uninjured controls on MRI measurements.

A total of 38 patients who were diagnosed with ATSFs by surgery or MRI and met the inclusion criteria were included in the ATSF group. For the control cohort, 38 age-, sex- and BMI-matched participants with isolated meniscal tear were included to match the cases. The mean age of all participants was 33.64 ± 9.07 years (33.08 ± 8.53 years for the ATSF group vs. 34.21 ± 9.67 years for the control group), and the mean BMI of all participants was 23.36 ± 3.65 (22.63 ± 3.11 vs. 24.08 ± 4.04, respectively). There were no significant differences between the 2 groups in demographic characteristics (Table 1).

The MRI measurements are presented in Table 2. The intraobserver and interobserver reliabilities of measurements were categorized as good with the minimum ICCs of 0.864 and 0.851, respectively (Table 3). There were significant differences between the ATSF group and the control group regarding the following parameters: LPTS (P = 0.001), MPTS (P = 0.005), NWI (P = 0.005) and LFCR (P = 0.012). However, no significant differences were observed between groups for MTD, LTH, LPTS-MPTS and LPTS/MPTS (Fig. 3).

The binary logistic regression results for the LPTS, MPTS, NWI and LFCR are shown in Table 4. From the table, the LPTS (OR = 1.326, 95% CI = 1.026–1.713, P = 0.031), NWI (OR = 3.1194E−14, 95% CI = 1.2831E−23–7.60E−05, P = 0.005) and LFCR (OR = 1.295, 95% CI = 1.063–1.578, P = 0.010) were found to be independently associated with ATSF.

The ROC curves were conducted for three significant independent predictors for ATSFs to compare their diagnostic performance (Fig. 4). The most accurate predictor was the LPTS, which had a sensitivity of 63.2% and specificity of 76.3% for predicting ATSFs at an optimal cutoff of 6.9 (Table 5).

The most important finding of this study was that LPTS, MPTS, NWI and LFCR were significantly different between patients with and those without ATSFs. However, only the LPTS, NWI and LFCR were found to be independently associated with ATSFs. The LPTS had 63.2% sensitivity and 76.3% specificity at a cutoff value of 6.9, providing the best predictive performance.

There is a paucity of research comparing the injury mechanisms of ATSFs and ACL tears. Previous investigations have shown that the demographic distribution and activity associated with ATSFs differ from those of ACL tears. Leie et al. demonstrated that skiing accounted for 56% of ATSFs, followed by soccer (22%) and rugby (16%) [25], whereas ACL tears were most prevalent in girls' soccer, followed by boys' football and basketball [26, 27]. These findings suggest that the mechanisms of these two injuries are subtly distinct. A biomechanical analysis of primate specimens has indicated that the viscoelastic properties of bone and ligaments are influenced by the rate of loading applied to the specimens [28]. Avulsion fractures are more likely to occur in cases with slower loading rates, while injuries to the intrasubstance portion of the ACL are more likely to be seen in cases with faster loading rates. In children and adolescents, discrepancies in the degree of tibial eminence ossification can also give rise to differences in injury patterns between ligament rupture and avulsion fractures.

Numerous studies have reported that increased PTS is an independent risk factor for ACL injury [13, 29, 30]. Increased PTS has an adverse effect on knee kinematics and kinetics, exerting more strain on ACL and consequently leading to stretching and tearing of ACL [31]. A recent biomechanical study found that increased PTS significantly increased tension on the ACL, and even at a 2.5° increase in PTS angle, knee joint instability and larger loading on the medial meniscus were found on the ACL-deficient knee [32]. Two studies reported the relationship between PTS and ATSF in pediatric populations, but they neither found a statistically significant association of PTS with ATSF when comparing cases with controls [11, 12]. But after stratifying by sex, Messner et al. [12] found that male patients who undergo surgical fixation of ATSF tend to have increased PTS as compared with controls. Differences with our current study in imaging modality and study populations may account for the unique findings. One of the concerns with their measurements using plain radiographs was that it was not easy to discern subtle differences between the compartments due to the superimposed medial and lateral tibial plateau, limiting their ability to analyze medial and lateral PTS, respectively [33]. Articular cartilage represents the functional point of tibiofemoral articulation, and MRI can accurately assess the 3D geometry of the cartilage surface, while plain radiographs cannot [34]. Our current study found that as the LPTS increases, for every additional degree, the risk of ATSF increases by 32.6%. And according to the results of the logistic regression analysis, MPTS was not an independent risk factor for ATSF. Although many studies have evaluated the effect of medial and lateral PTS on ACL injury, there are some conflicting findings, and LPTS measured on MRI is the most consistently reported risk factor [13, 30, 35, 36]. Our findings are in line with previous studies showing that LPTS is a more significant risk factor for ACL injury than MPTS [37‐39]. Previous biomechanical literature also suggested that an increased LPTS has a greater impact on anterior tibial translation than MPTS, which creates a net internal rotation that increases the strain on the ACL [40, 41].

The association between the femoral intercondylar notch shape and dimensions and ACL injury has attracted extensive interest. In particular, NWI based on MRI measurements has widely been shown to be an important risk factor for ACL injury [42, 43]. Kocher et al. [44] evaluated skeletally immature patients with ATSF compared to matched ACL injury. They found that the ACL group had narrower notch indices than the ATSF group (0.230 vs. 0.253; P = 0.020). This may partially explain the different injury patterns of ATSF and ACL injury in the skeletally immature knee. Although we identified the decreased NWI as an independent risk factor for ATSF, the logistic regression results suggest that the OR value for NWI is too small to further indicate its clinical significance. The current study demonstrated that the LFCR—a novel anatomical risk factor for ACL injuries, is also correlated with ATSFs. This association may arise from the influence of the distal femoral morphology on knee kinematics. An enhanced depth of the posterior femoral condyle could induce changes in tibiofemoral interactions, ultimately leading to modifications in gait and loading mechanics, which may increase the susceptibility to ATSFs [45, 46]. Further investigations involving additional cohorts and laboratory biomechanical studies are warranted to elucidate the distinct mechanisms of ACL tears and ATSFs.

Although these anatomic risk factors are largely unmodifiable, a growing body of data highlights the role of critical anatomical parameters that can help clinicians take preventive measures to avoid injuries. Numerous studies have identified the morphological risk factors for ACL injury of the knee joint. People with a high risk of ACL injury can benefit from modifiable interventions, including landing biomechanics, neuromuscular training, balance training and improvements in playing surfaces and footwear [47]. A meta-analysis performed by Huang et al. [48] demonstrated significant protective effects of ACL injury prevention programs and reduced injury rates by 53%. Currently, although MRI is not routinely performed for screening, preoperative MRI for ATSF patients is necessary to identify other combined injuries, such as meniscus and ligament injuries, and is cost-effective. Identifying high-risk individuals through preoperative MRI allows us to individualize treatment and rehabilitation programs to improve postoperative outcomes and avoid reinjury, such as refracture and ACL injury. For individuals at high risk of ATSFs, such as skiers, it is postulated that training athletes to jump, land and cut in a biomechanical position and using appropriate athletic equipment may potentially diminish the incidence of ATSFs [9, 49]. Future research will also be warranted to determine if more conservative rehabilitation programs and return-to-sport protocols are necessary for people at high risk of ATSFs.

Limited studies have investigated morphological risk factors associated with ATSFs. To the best of our knowledge, this is the first study to confirm the relationship between ATSFs and anatomic parameters, including LPTS, MPTS, asymmetry of the medial and lateral slopes, MTD, LTH, NWI and LFCR. Our results showed that the increased LPTS and LFCR and decreased NWI are associated with an increased risk of ATSF. However, this study has certain limitations. First, due to the small number of pediatric patients, we included only adult patients, limiting the generalization of our findings to pediatric populations. Nevertheless, as the two groups exhibit different ATSF mechanisms, forthcoming studies should consider addressing them separately. Second, we screened in the hospital system based on surgical records. Patients with ATSF who did not undergo surgery were not included, especially those with Meyers and McKeever classification type I fractures. Third, we included patients with only meniscal injury in the control group. Although this method has been widely adopted by previous studies [24, 50, 51], the inclusion of this population may introduce some bias.

Increased LPTS and LFCR and decreased NWI are significant risk factors for the incidence of ATSF in adults. All three parameters provide good predictive performance, but LPTS is the strongest predictor. These findings may contribute to the clinician identifying risk factors for ATSF and developing related preventive strategies. Future studies with a larger population are needed to further understand the anatomical risk factors and injury mechanisms of ATSF.