ACL Research Retreat VII: An Update on Anterior Cruciate Ligament Injury Risk Factor Identification, Screening, and Prevention

March 19–21, 2015; Greensboro, NC

Important Knowns and Recent Advances

Movement Biomechanics

Noncontact ACL injuries are most commonly reported to arise from a sudden deceleration while changing direction when running or landing from a jump.¹⁶ Quantitative analyses of actual injury events demonstrate that rapid knee abduction and internal rotation during the early weight-bearing phase occur at the time of injury.^17,18 Additionally, video observations of ACL injury occurrences demonstrate a relatively extended knee,¹⁹ greater knee-abduction motion,^20–22 increased amounts of lateral trunk motion,²³ and a more posteriorly positioned center of mass.²⁴ These observational findings are largely supported by cadaveric model work and lend credence to the use of high-risk biomechanical measures as outcome variables.

To date, 2 groups have demonstrated landing biomechanics to be prospective risk factors of ACL injury. When researchers used full 3-dimensional biomechanical methods, larger peak knee-abduction moment and peak knee-abduction angle at initial contact during a drop jump were prospective risk factors in adolescent female athletes.²⁵ Use of a more clinically friendly video observation of a jump landing (Landing Error Scoring System [LESS]) in youth soccer athletes showed that participants who went on to sustain ACL tears had lower LESS scores.²⁶ However, LESS scores did not predict ACL injury in high school and college-aged athletes.²⁷

A more erect or upright posture is commonly associated with increased vertical ground reaction forces.^28,29 This is important to understand as in vivo and in vitro strain of the ACL is related to maximal load and timing of ground reaction forces.^30,31 Similarly, anterior tibial translation increases as demands on the quadriceps increase.^32,33 Thus, this upright or extended posture when contacting the ground during the early stages of deceleration tasks has been suggested to be associated with the ACL injury mechanism.^34–38 Although short-term exhaustive exercise has resulted in changes in non–sagittal-plane variables associated with ACL injury risk,^39–43 exertional protocols that aim to simulate entire game or match duration and intensity have demonstrated subtle changes only in the sagittal plane, indicative of an at-risk upright posture.^44,45

Females are known to be at greater risk of ACL injury than males,^46,47 so investigations have focused predominantly on sex differences. Several recent reviews^48,49 have indicated that females demonstrated greater knee-abduction postures during landing. Given the prospective findings of Hewett et al,²⁵ knee abduction may be a potential key biomechanical factor in defining female-specific ACL injury mechanisms.⁴⁹

Neuromuscular and Neurocognitive Insufficiency

Trunk neuromuscular control has been implicated as a risk factor for ACL injury. In a prospective study⁵⁰ of trunk displacement after sudden force release, greater trunk displacement was a prospective predictor of ACL injury in college-aged females but not males. Compared with an erect or preferred trunk posture, landing with the trunk relatively more flexed increased hip-flexion and hip-extensor moments while decreasing ground reaction forces, knee moments, and quadriceps activation amplitudes.^28,51 During various running and cutting tasks, trunk and hip movements (in all 3 planes of motion) were commonly associated, at different magnitudes, with knee-abduction angles and external knee-abduction moments.^52–55 Although trunk and hip associations with “high-risk” knee biomechanics are clearly supported by these latter studies, the correlational results do not suggest cause and effect and could instead reflect task-dependent associations. Collectively, positions and movements of segments proximal to the knee could place the knee joint in a position of high risk for ACL injury.

Neurocognitive insufficiency as a risk factor for ACL injury is supported by findings of worse baseline reaction times, processing speed, and visual-spatial awareness in athletes who went on to sustain an ACL injury compared with controls.^2,56 Evidence for a connection between central processing and neuromuscular insufficiency was provided by the finding that slower cognitive processing speed was associated with decreased trunk stability in males.⁵⁷

Unknowns and Directions for Future Research

• 1.

  External- and internal-loading profiles that cause noncontact ACL ruptures in the field, data that are central to optimal injury-prevention strategies, are unknown.⁵⁸ Video from actual injury situations (and control videos from these athletes before injury) must be accumulated^17,18,24 to allow us to better understand the mechanisms of in vivo ACL rupture. Additionally, cadaveric, mathematical, in vivo kinematic, and imaging research approaches should be combined to help us understand the postures, loads, and neuromuscular profiles that increase ACL strain or cause noncontact ACL rupture.^31,59

• 2.

  Optimal ways to assess movement in the laboratory environment are still being debated with regard to single-legged and double-legged tasks.^60,61 To better understand how movement patterns and other structures in the kinetic chain affect ACL loads, we must continue to develop, improve, and validate quality laboratory-based models (eg, computational, cadaveric) that noninvasively estimate in vivo ACL forces and strain. Care should be taken to not overgeneralize results from 1 specific task to other tasks.

• 3.

  The influence of the kinetic chain (ie, ankle, hip, lumbopelvic, and trunk biomechanics) on modeled ACL strain or forces has been studied in controlled movements commonly performed in rehabilitation.^62,63 However, the specific relationship of these kinetic chain concepts with ACL injury risk is unclear. Laboratory, cadaver, and simulation-based studies designed to evaluate cause and effect (ie, highly controlled human movement studies with 1 variable manipulated) are warranted to ascertain whether these kinetic chain factors are a cause of or compensation for poor knee biomechanics.

• 4.

  Typical biomechanical experiments rely on extracting singular measures to characterize complex human biomechanics. To better understand the richness of these data, we should consider alternative approaches to analyzing traditional biomechanical data: for example, nonlinear dynamics⁶⁴ and emerging data-analysis techniques, including statistical parametric mapping.^65–67

• 5.

  To better understand the role of the central nervous system in ACL injury, research models should include assessments of central processes (eg, automaticity, reaction time), cognitive processes⁶⁸ (eg, decision making, focus and attention, prior experience [expert versus novice]), and metacognitive processes (eg, monitoring psychomotor processes). The reader is referred to Swanik,² who briefly reviewed anecdotal, theoretical, and recent clinical research evidence attempting to unravel the brain's role in maintaining joint stability.

• 6.

  Although investigation of an initial cadaveric model showed that the ACL is susceptible to fatigue failure during repetitive simulated pivot landings and a smaller cross-sectional area of the ACL is at greater risk for such failure,⁶⁹ we do not know if it is a single episode or multiple episodes (or both) that causes gross failure of the ACL.

• 7.

  To various degrees, the amount of maximal strength is related to movement mechanics, but how this is related to injury risk is unclear.^70–75 Further work on how neuromuscular properties beyond absolute strength (eg, muscle and joint stiffness, muscle mass, rate of force production) relate to ACL function and injury risk is warranted.

• 8.

  The general consensus is that maturation influences biomechanical and neuromuscular factors affecting ACL strain.^76–87 However, evidence for how maturation affects knee biomechanics is inconsistent.⁸⁸ Examining the influence of the maturational process on knee biomechanics and specifically on ACL loads may provide unique insights into the observed difference in injury rates by sex that emerges during the early stages of physical maturation. In this regard, multiyear, longitudinal studies are needed.

• 9.

  Methodologic concerns regarding data collection, processing, and analysis continue to confound the interpretation of biomechanical variables thought to be risk factors for ACL injury. Concerns related to validity³¹ and reliability^89–91 need further study before we can trust the quality and consistency of the data. Data processing should be carefully considered for its potential effect on identifying high-risk individuals and associations with ACL injury risk.^92,93 The research community would also benefit from operational definitions of terms (eg, stiffness, stability) so that scientists and clinicians can more transparently interpret the results and conclusions.

Important Knowns and Recent Advances

Knee-Joint Laxity

Greater magnitudes of anterior knee laxity,^122,123 genu recurvatum,^{122,124–127} general joint laxity,^{84,122,124,127} and internal-rotation knee laxity¹²⁸ have been reported in the contralateral knees of ACL-injured patients compared with controls.

Knee-joint laxity varies considerably among individuals, likely because of a combination of genetic,¹²⁹ hormonal,^130–135 and anatomical^136–138 factors. On average, females have greater sagittal (anterior knee laxity, genu recurvatum)-,^{104,122,139–143} frontal (varus-valgus rotation)-, and transverse (internal-external rotation)-plane^{135,144–146} and generalized joint laxity compared with males.^104,122 Sex differences in frontal-plane and transverse-plane knee laxity persist, even in males and females with similar sagittal-plane knee laxity.^135,144,146 Females are also more likely to experience acute increases in knee laxity during exercise¹⁴⁷ and across the menstrual cycle.^130–135

Greater magnitudes of knee laxity may have both biological and biomechanical consequences. Evidence from animal studies^148–150 suggests that greater knee laxity is associated with ligament biomarkers indicative of greater collagen turnover, more immature cross-links, and lower failure loads. Biomechanically, greater magnitudes of knee laxity and general joint laxity have been associated with higher-risk landing strategies more often observed in females (eg, greater knee-extensor loading and stiffening, greater dynamic knee valgus), particularly in the planes of motion where greater knee laxity is observed.^{139,151–157} The acute changes in knee laxity observed during exercise and across the menstrual cycle are of sufficient magnitude to push an individual toward higher-risk movement strategies.^{155,156,158–160}

Lower Extremity Alignment

Although the results of 2 large prospective studies are not yet available¹⁶¹ (see also Beynnon et al¹), no clear consensus in the literature consistently links any 1 lower extremity alignment factor to ACL injury. Lower extremity alignment differs among maturational groups and also develops at different rates in males and females within those groups.¹⁶² Once fully matured, females have greater anterior pelvic tilt, hip anteversion, tibiofemoral angle, and quadriceps angle.^142,163 No sex differences have been observed in measures of tibial torsion,¹⁴² navicular drop,^142,143,163 or rearfoot angle.^142,164

Body Composition

Increasing evidence from 2 large, prospective, multivariate risk factor studies showed that an elevated body mass index predicts future ACL injury in females but not in males.¹⁰⁴ (See also Beynnon et al.¹) Higher body mass index in females is more likely to reflect a higher proportion of fat mass versus lean mass relative to body weight^165,166; less relative lean mass has been associated with greater knee-joint laxity¹³⁷ and dangerous biomechanical strategies.^51,167 This may explain why the combination of greater joint laxity and greater body mass index is a stronger predictor of ACL injury risk in females than either variable alone.¹⁰⁴ (See also Beynnon et al.¹)

Unknowns and Directions for Future Research

• 1.

  Anatomical and structural factors have often been examined independently or in small subsets of variables. Even among 3 larger, prospective multivariate risk factor studies^104,161 (see also Beynnon et al¹), the anatomical variables examined are quite disparate, making it difficult to build a clear consensus on the best prediction model of ACL injury risk for screening purposes. Further large-scale risk factor studies that account for all relevant lower extremity anatomical and structural factors are needed to reach this consensus. Prospective confirmatory analyses of currently identified risk factor models across different races and ages are also needed.

• 2.

  Most investigations of tibial plateau geometry are based on measures of the subchondral bone, and recent research¹¹⁰ suggests that accounting for the overlying cartilage and meniscal geometry when characterizing plateau slope and depth may improve prediction. However, because these measures are complex and costly, it will be important to determine the extent to which they provide additional information on ACL injury risk once more clinically accessible risk factors are taken into account. (See also Beynnon et al.¹)

• 3.

  Studies examining the combined effects of joint laxity, tibial geometry (lateral tibial slope, medial : lateral tibial slope ratio, coronal slope, medial condylar depth) and ACL morphology, as well as interactions among these variables, on tibiofemoral joint biomechanics and ACL strain and failure are encouraged.

• 4.

  Although anatomical and structural factors are often thought to be nonmodifiable, we have limited knowledge of how these structural factors change during maturation or whether physical activity (or other chronic external loads) can influence this development over time, particularly during the critical growth periods. Prospective longitudinal studies are needed to help us understand the underlying factors that result in at-risk anatomical and structural profiles during maturation while also considering relevant modifiable factors such as body composition, neuromuscular properties, and physical activity.

• 5.

  Recent evidence¹⁶⁸ suggests that individuals with specific laxity profiles may perceive functional deficits during activities of daily living and sport. We do not yet know the multiplanar laxity profiles that pose the greatest risk for ACL injury or the threshold at which the magnitude of knee laxity becomes problematic.

• 6.

  Because body mass index is not a true measure of body composition or body mass distribution, future studies using a more detailed analysis of body composition are needed to fully understand how body composition contributes to faulty lower extremity biomechanics and injury risk.

Important Knowns and Recent Advances

Since the previous ACL retreat, a growing number of common DNA sequence variants within genes involved in various biological processes have been associated with susceptibility to ACL rupture. These include variants within several genes that encode collagens^169–175 and proteoglycans¹⁷⁶ involved in the formation of the collagen fibril, the basic building block of ligaments. A subset of these variants has been specifically associated with ACL ruptures in females but not in males.^{171,173,174,177} The ACL's collagen fibril and the rest of its extracellular matrix continuously undergo remodeling after mechanical loading to maintain tissue homeostasis. In support of this, variants within genes that encode proteins involved in cell-signaling pathways, such as angiogenesis-associated signaling¹⁷⁸ and the apoptosis-signaling cascade,¹⁷⁹ as well as remodeling (specifically matrix metalloproteinases^180,181), have also been associated with ACL ruptures. Many of the anatomical, structural, and other risk factors are themselves multifactorial phenotypes determined by, to a lesser or greater extent, both genetic and environmental factors.¹⁸² Based on the current evidence, it is therefore unlikely that the identified genetic variants are independent risk factors and more likely that they modulate risk through their effects on structural differences and other biological variations.

Genetic association studies in humans have recently been strengthened by a publication reporting the association of variants within genes, including collagen and other genes proposed to have a detrimental effect on ligament structure and strength, with canine cranial cruciate ligament rupture.¹⁸³ Using a genome-wide association study design, the same investigators¹⁸⁴ have also identified 3 main chromosomal regions, which include variants within neurologic pathway genes, associated with canine cranial cruciate ligament rupture.

Unknowns and Directions for Future Research

• 1.

  Most of the genetic risk factors for ACL rupture to date have been identified in white populations, and the reported associations cannot necessarily be extrapolated to other population groups. For example, the functional Sp1-binding site polymorphism (rs1800012, G/T) within the COL1A1 gene is associated with ACL rupture in white populations.^{169,170,172,175} However, the frequency of the minor T allele of this variant is much lower in many other population groups (www.ensembl.org) and therefore unlikely to be informative within all populations.¹⁸⁵ Hence, investigators should identify appropriate informative genetic markers for other population groups.

• 2.

  Although establishment of an international consortium or registry was recommended at the last ACL retreat,¹⁷¹ most of the published case-control genetic association studies have involved relatively small sample sizes, especially with respect to the sex-specific genetic effects. The establishment of an international consortium or registry will allow us to test the hypothesis of a stronger genetic contribution in individuals who tear their contralateral ACL rather than retear the same ACL or incur no further ACL injury. The establishment of an international consortium will also make it possible to recruit large sample sizes for whole-genome screening methods, enabling us to identify all the potentially important biological pathways involved in ACL ruptures and how these might differ by population.

• 3.

  None of the associated variants alone cause ACL ruptures. Rather, the injury is the result, at least in part, of a poorly understood, complex interaction of external loading and other factors, as well as intrinsic stimuli, with the genetic background of the individual. Further research is needed to ascertain the extent to which the associated variants may result in interindividual variations in the structure and, by implication, the mechanical properties of the ACL and surrounding tissues, as well as its responses to mechanical loading and other stimuli, thereby affecting the risk of injury.

• 4.

  Although scientists are only starting to understand the contribution of genetics to ACL injury, numerous companies are marketing direct-to-consumer genetic tests for common injuries, including ACL ruptures. However, these tests are premature because the genetic data are incomplete and have not been considered together with clinical indicators and lifestyle factors, which may allow an appropriately qualified health care professional to identify an altered risk for injury.

Important Knowns and Recent Advances

Hormone receptors (eg, estrogen, testosterone, and relaxin) have been localized on the human ACL,^186–190 suggesting they are capable of regulating gene expression and collagen metabolism in a way that may influence the biology of the ACL and other soft tissue structures. This idea is supported by studies that have demonstrated associations of normal physiologic variations in sex hormone concentrations across the menstrual cycle with substantial changes in markers of collagen metabolism and production,¹⁹¹ knee-joint laxity,^130–135 muscle stiffness,¹³¹ and the muscle-stretch reflex.¹⁹² These biological changes may also have secondary neuromechanical consequences, as previously noted.^{155,156,158–160}

Although results from prior epidemiologic studies have suggested that the risk of an ACL injury appears to be greater during the preovulatory phase of the menstrual cycle compared with the postovulatory phase,^193–197 our understanding of the underlying mechanism for this increased likelihood has not advanced. Understanding these processes is difficult given the substantial variability in the timing and magnitude of hormone changes among individuals,¹³⁴ the time-dependent effect for sex hormones and other remodeling agents to influence a change in ACL tissue characteristics,^134,190 and the potential interactions among several sex hormones, secondary messengers, remodeling proteins, mechanical stresses, and genetic influences.* For example, interactions among hormones, mechanical stress, and altered ACL structure and metabolism have been observed in some animal models.^148,205,206

In recent years, the hormone relaxin has gained attention as a potential ACL injury risk factor based on a prospective study of National Collegiate Athletic Association Division I female athletes that showed elevated concentrations in those with ACL injuries compared with uninjured controls.²⁰⁷ Even though studies examining the effect of relaxin on collagen metabolism in human knee ligaments in situ are lacking, in vivo and in vitro animal studies and human cell culture studies suggest that relaxin administered at physiologic levels can have a profound effect on soft tissue remodeling.^{204,208–211} In turn, this may lead to a less organized²¹² and less dense (both in fiber diameter and density) collagen structure,^204,212 which may manifest as a weaker and more lax ACL.¹⁸⁶ This concept is supported by an animal model in which guinea pig ACLs treated with relaxin at pregnancy levels were 13% more lax and 36% to 49% weaker.²¹³

Unknowns and Directions for Future Research

• 1.

  Our understanding of the biological consequences of sex hormones on collagen metabolism and the structural integrity of the ACL in the human knee in situ remains limited. The underlying sex-specific molecular and genetic mechanisms of sex hormones on ACL structure, metabolism, and mechanical properties and how mechanical stress on the ACL alters these relationships and injury risk potential remain important areas of study.

• 2.

  Good evidence demonstrates that some (but not all) females experience substantial cyclic changes in their laxity and knee-joint biomechanics across the menstrual cycle,^155,158,160 but it is not yet possible to clinically screen for these potentially high-risk individuals. Understanding the underlying processes that regulate ligament biology and resultant ligament behavior may allow us to better screen for these individuals and prospectively examine how these factors influence injury risk potential.

• 3.

  Given the time-dependent effect of sex hormones on soft tissue structures, we need to determine how the time of injury occurrence aligns with acute changes in ACL structure and metabolism or knee-laxity changes and how the rate of increase or the time durations of amplitude peaks in hormone fluctuations across the menstrual cycle play roles in the magnitude or timing of soft tissue changes.

• 4.

  When examining hormone influences in physically active females, it is important to match the complexity of interparticipant differences in timing, magnitude, and interactive changes in sex hormone concentrations across the cycle with our study designs. Investigations should (1) verify the phases of the cycle (or desired hormone environment) with actual hormone measurements (considering relevant hormones to include estrogen, progesterone, and possibly others) rather than relying on calendar day of the cycle²¹⁴ and (2) obtain multiple hormone samples over repeated days to better characterize hormone profiles within a given female.^215,216 This may be particularly important when documenting relaxin exposure. Because relaxin rises and peaks during a relatively short window (6–10 days) after ovulation based on other hormonal events,^217–220 we do not know if the high proportion of undetectable levels within a single sample (as high as 64%–80%)^221–224 is due to timing or suggests that only a subset of females are exposed to appreciable relaxin levels due to mediating factors.^219,221,225

• 5.

  Because cyclic hormone concentrations affect soft tissues and knee-joint function, studies comparing females with males should be made during the early follicular phase in females, when hormone levels are at their nadirs (preferably 3–7 days postmenses).

• 6.

  Menstrual dysfunction in exercising females may be more prevalent than originally thought (up to 50% in exercising females).^216,226,227 Thus, the effects of menstrual disturbances on injury risk potential are poorly understood.

• 7.

  Although the authors of epidemiologic studies have reported no protective effect of oral contraceptives on ACL injury risk,^72,73 these medications are known to vary substantially in the potency and androgenicity of the progestin compound delivered, which ultimately determines the extent to which they counteract the estrogenic effects.²²⁸ We need to better understand how the different progestins influence soft tissue structures, knee function, and ACL injury risk. Furthermore, the exogenous hormones delivered in different types of contraceptives (eg, vaginal ring, transdermal patch, oral pills) are metabolized differently and may have different effects on musculoskeletal tissues.²²⁹ Relevant comparisons should then be made among users of hormonal contraceptives (of all 3 types) and eumenorrheic, amenorrheic, and oligomenorrheic females to determine if ACL injury risk or observed soft tissue changes vary.