Introduction

The anterior cruciate ligament (ACL) is important for maintaining stability of the knee joint. ACL injury may result in loss of forward and rotational stability of the knee joint, leading to secondary damage (i.e., meniscus and articular cartilage damage) [1]. Hence, evaluating and identifying risk factors for ACL injury as early as possible are of clinical importance.

The femoral intercondylar notch is a specialized anatomical structure that accommodates the anterior and posterior cruciate ligaments and the tibial intercondylar spine. It exerts a vital role in maintaining knee joint stability. The shape and width of the intercondylar notch are largely affected by sex, age, height, and ethnicity [11, 21]. Everhart et al. found that women had narrower notches and higher rates of stenosis than did men. African–American men have narrower notches than Caucasian men. Height and notch width were found to be positively associated only among men [11]. Notch width index (NWI), initially proposed by Souryal et al., describes the intercondylar notch width [41]. NWI indicates the ratio of the femoral intercondylar width to that of the internal and external condyles at the plane of the popliteal tendon. NWI excludes most anatomical factors and height and weight interference; thus, it can accurately reflect the width and stenosis of the intercondylar notch. Numerous studies have explored the relationship between tibial slope and intercondylar notch dimensions, and anterior cruciate ligament injury [53, 54]. Many studies found that tibial slope and narrow intercondylar NW are risk factors for ACL injury [10, 46, 48], while others have shown that NW does not affect ACL injury [47, 54].

Intercondylar notch is analyzed by plain radiographs, computed tomography (CT) and magnetic resonance imaging (MRI) measurements. Radiographic measurements of the intercondylar notch are sometimes inaccurate even under optimal conditions, including correction technology, projection, and magnification [4]. CT shows stable reliability between observers when measuring the femur and tibia [34]. However, because of radiation, CT use in ACL imaging is limited. MRI is more accurate than are plain radiographs, with similar accuracy to that of CT [34]. MRI has high contrast resolution for soft tissue and enables clear distinguishing of the meniscus, cruciate ligament and articular cartilage of the knee joint and can yield multiple imaging parameters [31]. MRI exerts relatively high sensitivity and specificity in diagnosing ACL injuries [8, 18].

This meta-analysis explored the relationship between the intercondylar notch dimension assessed via MRI, including NW and NWI, and the risk of anterior cruciate ligament injury. We hypothesized that MRI intercondylar notch stenosis measured via MRI was a risk factor for ACL injury.

Methods

Search strategy

This meta-analysis was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) standards. Studies reporting a correlation between ACL injury and NWI and/or NW published before 30 June 2018 were searched in the PubMed, Embase, China National Knowledge Infrastructure and Wanfang databases. The keywords “intercondylar notch” OR “notch width” OR “notch width index”and “anterior cruciate ligament” were searched. Two investigators independently and manually reviewed the studies reporting ACL and NW/NWI and their reference lists.

Inclusion and exclusion criteria

Inclusion criteria were (1) observational research, including case–control or prospective studies; (2) NWI and/or NW were exposure factors and assessed via MRI; and (3) ACL was the interest outcome. Exclusion criteria were (1) controls or ACL injury cases only; (2) lacking the mean, standard deviation (SD) or confidence interval (CI); (3) reviews, letters, abstracts or animal experiments; and (4) no MRI was conducted.

Data extraction

Two investigators independently extracted the baseline characteristics and results in the included studies. A third investigator resolved any disagreements. The following information was collected: name of first author, publishing year, nationality, ethnicity, age, case resource, control resource, study design, and the mean ± SD of the main indicators (NW and/or NWI).

Statistical analyses

Data were analyzed and managed using Stata 12.0 (StataCorp., College Station, Texas, USA). The case and control groups were assigned according to ACL injury. NW and/or NWI examined via MRI were the exposure factors, and their weighted mean difference (WMD) and corresponding 95% confidence interval (95% CI) were recorded. Heterogeneity among the enrolled literature was assessed via Chi square tests and q tests. The Mantel–Haenszel fixed-effects method or the DerSimonian and Laird random-effects method were used to assess the WMD and 95% CI. Subgroup analyses were conducted based on ethnicity, sex and control source. Begg’s and Egger’s tests were performed to detect publication bias. Sensitivity was analyzed to examine the stability of our results. P < 0.05 was considered statistically significant.

Results

Baseline characteristics

Of the 1,792 studies retrieved from the databases, 82 were selected for full-text review. Finally, 28 eligible studies were enrolled. Figure 1 shows a diagram of the literature screening. Among the 28 enrolled studies, 22 were published in English [3, 5, 6, 8, 12, 13, 16,17,18, 22, 29, 32, 33, 35, 36, 38, 39, 42, 44, 45, 51, 56], 5 were published in Chinese [19, 24, 27, 55, 57], and one was a dissertation in Chinese [26]. Patients with noncontact ACL injury were defined as the case group; these subjects exhibited noncontact ACL injuries, and patients in four studies also presented knee osteoarthritis (KOA). The population without ACL injury was defined as the control group. Controls were divided into healthy controls, patients with KOA only, self-controls (ACL injury on one side of the subject, normal on the other) and other patients (those with meniscal injuries or a diagnosis of patellofemoral pain but no ACL injury). Subjects ranged in age from 20–60 years. There were one prospective cohort study and 27 case–control studies. Eighteen studies reported NW data, and 21 reported NWI data. Table 1 lists the baseline characteristics for these studies.

Fig. 1
figure 1

Flow diagram for enrolled studies

Full size image
Table 1 The basic characteristics of enrolled studies
Full size table

NW and risk of ACL injury

Eighteen articles reported the relationship between NW and ACL, among which, three articles [33, 51, 56] distinguished the NW data between men and women, and these were divided into six studies according to sex. Finally, 21 studies were included in the meta-analysis (Fig. 2). NW was significantly narrower in ACL injury cases than in the overall controls (Table 2).

Fig. 2
figure 2

Forest plot for the intercondylar notch width and risk of anterior cruciate ligament injury. a Ethnicity subgroup analysis; b sex subgroup analysis. a, b Studies were divided into male and female group

Full size image
Table 2 NW and risk of ACL injury
Full size table

Subgroup analyses were conducted based on ethnicity and sex. Further subgroup analysis showed that NW was smaller in ACL injury cases among Asian and Caucasian populations. Three articles supplied the NW data for both women and men [33, 51, 56] and were divided into six studies. Four articles confirmed the subjects’ sex [3, 5, 29, 36]. Finally, ten studies were analyzed by sex subgroup for NW. NW was significantly narrower in ACL cases among women and men than in the control cases (Table 2).

No publication biases were identified by Begg’s test (P = 0.51 for overall and ethnicity subgroup analyses; P = 0.72 for sex subgroup analysis; Fig. 3) or Egger’s test (data not shown). Sensitivity analysis showed that the findings were robust (data not shown).

Fig. 3
figure 3

Funnel plot for the intercondylar notch width and risk of anterior cruciate ligament injury. a Ethnicity subgroup analysis; b sex subgroup analysis

Full size image

NWI and risk of ACL injury

Twenty-one articles reported a correlation between NWI and ACL injury. Among them, two articles identified NWI in men and women [33, 56], and one contained both healthy controls and KOA controls [16]. These three articles were divided into six studies. Finally, 24 studies were subjected to meta-analysis (Fig. 4a). In the overall analysis, the NWI was significantly narrowerin ACL cases than in the controls (Table 3).

Fig. 4
figure 4

Forest plot for the intercondylar notch width index and risk of anterior cruciate ligament injury. a Ethnicity subgroup analysis; b source of controls subgroup analysis; c sex subgroup analysis. a, b Studies were divided into male and female group. 1, 2 studies were divided into healthy control and KOA control

Full size image
Table 3 NWI and risk of ACL injury
Full size table

Subgroup analyses were performed on ethnicity, control source and sex (Table 3 and Fig. 4). NWI was significantly lower in ACL injury cases among Asian populations and others stratified by ethnicity. When stratified by control source, NWI was significantly lower in ACL injury cases than in self-control cases (Fig. 5). Eight studies were subjected to subgroup analysis by sex; a significant difference was found among men.

Fig. 5
figure 5

Funnel plot for the intercondylar notch width index and risk of anterior cruciate ligament injury. a Ethnicity subgroup analysis; b source of controls subgroup analysis; c sex subgroup analysis

Full size image

No publication biases were identified by Begg’s test (P = 0.083 for the overall, ethnicity subgroup, and control source subgroup analyses; P = 0.53 for the sex subgroup analysis) or Egger’s test (data not shown). After removing one study [5], sensitivity analysis identified a significant difference for NWI results in the Caucasian population, healthy controls and women.

Discussion

Internal factors that determine the risks for ACL injury include anatomical features of the pelvis and femur, intercondylar notch width, posterior slope at the medial plateau, Q angle, joint relaxation and flexibility, and hormone differences [9, 20]. The narrow intercondylar notch angle can lead to impact from forward translation forces and forced knee valgus, which can promote ACL injury [2]. The increased posterior tibial slope may promote ACL injury by increasing anterior motion of the tibia relative to the femur or intense quadricep contraction, or increasing torsional loads in cases of differences in the medial and lateral tibial slopes [14]. Two-dimensional parameters of the femoral notch are widely used because they are convenient to measure. However, they only represent the dimension at one notch location; thus, they cannot fully reflect the overall dimension and may lead to measurement bias [50]. Three-dimensional notch volume can reflect the overall dimension; however, owing to its high technical threshold and tedious measurement, its application is limited.

Two-dimensional notch parameters, including NW and NWI, are often used to evaluate femoral notch size. In most cases, two-dimensional parameters can effectively evaluate the notch size [50]. Femoral NW and NWI were often used to assess the risk of ACL injury, but the conclusions in each study were controversial [25]. To comprehensively understand how the NWI affects ACL injury, we enrolled 28 relevant studies and performed a meta-analysis.

The NW was markedly narrower in ACL injury cases than in the controls. Subgroup analyses stratified by ethnicity and sex yielded similar results. Intercondylar notch stenosis can result in ACL injury. A narrow femoral intercondylar notch can easily induce impact between the ACL and lateral wall of the femoral intercondylar bones once the knee joint is over-bent or rotated, which is much more pronounced at the position of flexion-valgus and external rotation. Frequent impact and friction easily damage the ACL fiber bundles, which cannot easily handle large external forces [12, 15]. Stijak et al. [43] and Dienst et al. [7] considered that the intercondylar notch volume is consistent with that of the ACL it contains. People with intercondylar notch stenosis present an ACL with a smaller volume, weaker strength and worse biomechanical properties and are highly susceptible to ACL rupture. Nevertheless, some studies found no significant correlation between intercondylar notch volume and that of the ACL contained within [28, 37]. In the current analysis, the NW was narrower in ACL injury cases, which may predict ACL injury. Notably, intercondylar notch width varies greatly among individuals, possibly owing to measurement methods, measurement planes or ethnicity.

NWI is an advantageous indicator that eliminates the influences of height, weight and sex and avoids individual differences and measurement errors. The critical value for NWI differed in the relevant literature because of different measurement methods, subject positions and sample sizes. In the current analysis, NWI was smaller in the ACL injury cases than in the controls. Subgroup analyses based on ethnicity revealed that the NWI was smaller in ACL injury cases among Asian and other populations. Uhorchak et al. [47] suggested that an NWI of < 0.18 would increase the risk factor of the ACL, which was 0.19 in LaPrade’s study [23]. Domzalski et al. [8] pointed out that the average NWI in ACL injury cases is 0.24, and NWI decreases with age. Souryaland Freeman [40] reported that the NWI in healthy controls was 0.231 ± 0.044 and that an NWI < 0.20 will significantly enhance the risk of bilateral ACL injury. Souryal et al. [41] recorded NWI in a 45° kneeling position via X-ray across many healthy populations. The average NWI was 0.231 ± 0.004, suggesting that an NWI ≤ 0.20 indicated severe stenosis of the intercondylar notch. Muneta et al. [30] found no significant difference in ACL volume between people with NWI > 0.20 and those with NWI < 0.20. Simon et al. [38] demonstrated that the direction of muscle travel in the ACL was closer to vertical under the pathological condition of intercondylar notch stenosis. Vertical travel direction space larger loads on the ACL and are prone to ACL injury. ACL injury is thought to be more frequent in female athletes than in male athletes, and sex influences many anatomical parameters (i.e., NWI and NW). Compared with men, women have a smaller NW and are more susceptible to ACL injury [52]. However, hormone levels, neuromuscular levels and biomechanics are all potential factors that influence ACL injury [2, 29]. Van et al. revealed no significant difference in NWI between men and women, possibly owing to different inclusion criteria and measurement methods [49]. Here, we found NWI was lower in ACL injury cases among men. Notably, when stratified by control source, subgroup analyses identified a lower NWI in ACL injury cases in the self-control group only, suggesting that the two studies including self-controls may have affected the overall results.

This analysis had some limitations. First, the included literature was significantly heterogeneous, and some resources were unclear even after subgroup and sensitivity analyses, likely owing to different measurement methods and examination planes. To get accurate measurements, the NW and corresponding NWI of different parts of the notch should implement a unified standard. These measurements should include the notch entry width (nw_-in), the exit width (nw_-ou), and the ACL attachment (nw_-aa) [56]. Second, most studies in this meta-analysis were case–control studies, and only one was a prospective cohort study. Thus, the results may have been due to ACL injury and intercondylar notch stenosis. Some other factors may have affected our findings. Third, the study by Bouras et al. [5] may have affected the stability of our findings; this study included many cases with women, and no significant correlation was found between NWI and ACL injury. Fourth, although no significant publication biases were identified, relevant studies published in languages other than English and Chinese were lacking, and some negative results may have been missing. Finally, the small sample size in this analysis likely influenced the reliability of our findings.

Conclusion

The current meta-analysis evidenced that the NW assessed via MRI was significantly smaller in ACL injury cases than in controls. NWI was lower in ACL injury cases among men. Prevention strategies for ACL injury could be applied for people with intercondylar notch stenosis.