Combined injury to the ACL and lateral meniscus alters the geometry of articular cartilage and meniscus soon after initial trauma

Abstract

Combined injury to the anterior cruciate ligament (ACL) and meniscus is associated with earlier onset and increased rates of post traumatic osteoarthritis (PTOA) compared to isolated ACL injury. However, little is known about the initial changes in joint structure associated with these different types of trauma. We hypothesized that trauma to the ACL and lateral meniscus has an immediate effect on morphometry of the articular cartilage and meniscus about the entire tibial plateau that is more pronounced than an ACL tear without meniscus injury. Subjects underwent MRI scanning soon after injury and prior to surgery. Those that suffered injury to the ACL and lateral meniscus underwent changes in the lateral compartment (increases in the posterior-inferior directed slopes of the articular cartilage surface, and the wedge angle of the posterior horn of the meniscus) and medial compartment (the cartilage-to-bone height decreased in the region located under the posterior horn of the meniscus, and the thickness of cartilage increased and decreased in the mid and posterior regions of the plateau, respectively). Subjects that suffered an isolated ACL tear did not undergo the same magnitude of change to these articular structures. A majority of the changes in morphometry occurred in the lateral compartment of the knee; however, change in the medial compartment of the knee with a normal appearing meniscus also occurred.

Introduction:

Anterior cruciate ligament (ACL) injury is immediately debilitating, produces abnormal joint biomechanics,¹ and leads to early development of post-traumatic osteoarthritis (PTOA) regardless of surgical or non-surgical treatment.^2,3 This is a considerable healthcare concern because the highest incidence rate of ACL-injury occurs in young individuals between 15 and 25 years of age,^4,5 and it is expected that a substantial portion of these individuals are at significantly increased risk of experiencing PTOA within 15 years of the initial trauma.⁶

Trauma that involves injury to the ACL combined with other soft tissue structures such as the menisci and articular cartilage is associated with increased rates of PTOA compared to ACL injury in isolation.^7,8 However, little is known about the mechanism by which ACL and concomitant trauma to other articular structures influences the onset and initial progression of PTOA.³ A critical step in addressing this knowledge gap is to determine how combined ACL and meniscus injury affects knee structure and geometry soon after injury. Our previous studies revealed that the index trauma has an immediate effect on the geometric profiles of articular structures and articular cartilage thickness.^9,10 However, this work and that of others did not study ACL disruption with concomitant meniscal injury and isolated ACL disruption without meniscal injury.^9,11–13 Characterizing the initial, gross morphological changes to the articular structures about the knee following severe ligament trauma is important for informing future studies designed to understand the mechanisms and temporal events associated with the onset of PTOA. Altogether, studies targeting the time period soon after injury may provide an opportunity to understand the biomechanical and biological mechanisms of isolated ACL injury and combined ACL and meniscus injury that will aid in successful treatments that prevent PTOA. In the current study, we hypothesized that ACL disruption in combination with injury to the lateral meniscus has an immediate effect on morphometry of the articular cartilage and meniscus about the entire tibial plateau that is more pronounced than an ACL tear without concomitant meniscus injury.

Methods:

This investigation is an extension of a prospective cohort study (Level of evidence II) of the risk factors associated with suffering a first time, non-contact ACL injury.^14–17 Our institution’s Committee on Human Research in the Medical Sciences approved this investigation and each participant, as well as a parent or legal guardian if the participant was younger than 18 years of age, provided signed informed consent prior to participation.

Female and male freshman, junior varsity and varsity athletes from 28 high schools and 8 colleges were monitored prospectively throughout their athletic seasons over a four-year time interval.¹⁴ They participated in either soccer, basketball, lacrosse, field hockey, football, rugby, volleyball, baseball, softball, track and field or wrestling. Athletes that suffered their first complete non-contact ACL injury to either knee (defined as an injury occurring without direct force and contact to the knee) and had not previously suffered significant knee trauma (to either side) were identified immediately after the injury occurred. Athletes that suffered partial ACL injury or re-injury to an ACL graft were excluded. ACL injuries were diagnosed by an orthopedic surgeon and confirmed by magnetic resonance imaging (MRI) and subsequent arthroscopic visualization. At the time of enrollment of an injured subject, control subjects of the same age with normal knees and no history of significant trauma to their lower extremity were selected from the injured subjects’ teammates and were enrolled in the study. This selection process ensured the control and injured subjects had similar exposure to sport and activity associated with the trauma and were the same sex and age.

Following injury and prior to surgery, all injured participants underwent 3T MRI scans using the same scanner (Phillips Acheiva 3T MRI system) on both their injured and uninjured knees using a previously described protocol that has been shown to be reliable.¹⁸ Subjects were positioned supine within an eight channel SENSE knee coil by the same technician with the knee in extension, and three-dimensional proton density weighted scans (slice thickness 0.7 mm; resolution 0.4 mm by 0.4 mm) and T1-weighted fast field echo (FFE) scans (slice thickness 1.2mm; in plane resolution 0.3 mm by 0.3 mm) were acquired. All control subjects were imaged using the same protocol.

Forty-three subjects met our entry criteria and were treated at the University of Vermont Medical Center. Each injured subject had arthroscopic findings regarding the status of their articular cartilage and menisci of their ACL injured knee documented at the time of surgery by the same orthopaedic surgeon. Arthroscopy revealed 3 subjects suffered injuries to the ACL and both medial and lateral menisci, 4 subjects experienced injuries to the ACL and medial meniscus, 17 subjects suffered injuries to the ACL and lateral meniscus, and 19 subjects had an ACL injury without arthroscopic evidence of injury to their articular cartilage and menisci.(Table 1; supplemental text) This distribution of injuries led us to establish two groups; subjects that suffered injuries to the ACL and lateral meniscus (The ACL-LM group; n=17) and subjects that only suffered injury to the ACL (The ACL group; n=19). The sample size was identical for the matched control subjects (n=17 and 19, correspondingly)

Characterization of the position of the tibia relative to the femur during MRI acquisition was accomplished by establishing three-dimensional, bone-based coordinate systems within the femur and tibia (Fig. 1).^9,18–20 The following three-dimensional (3D) measurements of the geometry of the tibial articular cartilage surface, underlying tibial subchondral bone, and meniscus were made relative to the 3D bone based coordinate systems using the approaches that have been described (Table 1).^15,16 In the medial and lateral compartments of the tibia, the profiles of the articular cartilage surfaces were characterized as the slopes of the middle and posterior regions at that sagittal MRI slice that was located at the point of maximum depth of concavity in the articular surface (M_edialMCS, M_edialPCS, L_ateralMCS, L_ateralPCS: Fig. 2), the wedge angles formed between the superior surface of the posterior horn of the meniscus relative to the underlying articular cartilage (M_edialMCA, L_ateralMCA: Fig. 2) and relative to subchondral bone (M_edialMBA, L_ateralMBA: Fig. 2) were characterized. In addition, we measured the slopes of the underlying subchondral bone in both compartments of the tibial plateau (M_edialBS and L_ateralBS: Fig. 2), heights of the posterior aspect of the meniscus relative to the surface of the articular cartilage (M_edialMCH, L_ateralMCH: Fig. 2), and the posterior surface of the articular cartilage relative to subchondral bone (M_edialCBH, L_ateralCBH: Fig. 2). Tibial articular cartilage thicknesses maps were derived from the perpendicular distance between the three-dimensional profiles of articular cartilage and subchondral bone surfaces at 1 mm by 1 mm grid points evenly spaced over the tibia bone surface as previously described,⁹ and this was done for the lateral and medial compartments. Geometry of the tibial articular cartilage surfaces, underlying tibial subchondral bone, menisci, and cartilage thickness maps were obtained with a high level of measurement reliability.^9,15,16

Figure 1.

Bone-embedded anatomical coordinate systems for the tibia and femur^9,18–20. Anatomical coordinate systems were used to determine the position of the tibia relative to the femur.

Table 1:

Description of angle, slope and height measurements made on articular cartilage and meniscus^15,16. Measurements were made in the lateral (Lat) and medial (Med) compartments.

	Units	Description
Lat MBA	degrees	Lateral Meniscus Bone Angle
Lat MCA	degrees	Lateral Meniscus Cartilage Angle
Lat PCS	degrees	Lateral Posterior Cartilage Slope
Lat MCS	degrees	Lateral Middle Cartilage Slope
Lat BS	degrees	Lateral Bone Slope
Lat MCH	mm	Lateral Meniscus Cartilage Height
Lat CBH	mm	Lateral Cartilage Bone Height
Med MBA	degrees	Medial Meniscus Bone Angle
Med MCA	degrees	Medial Meniscus Cartilage Angle
Med PCS	degrees	Medial Posterior Cartilage Slope
Med MCS	degrees	Medial Middle Cartilage Slope
Med BS	degrees	Medial Bone Slope
Med MCH	mm	Medial Meniscus Cartilage Height
Med CBH	mm	Medial Cartilage Bone Height

Figure 2:

Top panel presents articular cartilage and bone slopes. Middle and posterior cartilage slopes (MCS and PCS) and subchondral bone slope (BS) measured relative to the posterior-anterior directed axis of the tibia. The middle panel presents the posterior meniscus angles. Posterior meniscus wedge angle measured relative to posterior surface of the articular cartilage (MCA) and relative to subchondral bone (MBA). The bottom panel presents the meniscus and cartilage heights. Height of the posterior aspect of the meniscus measured relative to the cartilage surface (MCH), and height of the posterior articular cartilage surface measured relative to subchondral bone (CBH). The same measurements were made in the medial and lateral compartments.

To test our hypothesis we initially made within subject (injured-to-normal side) comparisons of the before mentioned geometric features of the articular cartilage, meniscus, subchondral bone and position of the tibia relative to the femur using a Student’s paired T-test after confirming normality of the data. Comparisons were done separately for the ACL group and then the ACL-LM group. This was followed by comparisons between the two groups (ACL vs. ACL-LM groups) using a non-paired T-test after confirming normality.

A two-step process was used to analyze the cartilage thickness maps. The first step was to process data from the matched (based on age, sport and sex), uninjured control subjects.⁹ At each point on the 1 mm by 1 mm grid we calculated the side-to-side difference in cartilage thickness with corresponding 95% confidence intervals and used this data to establish the normal limits of cartilage thickness difference values outside of which we have the capacity to detect an injured-to-normal side difference for the ACL injured subjects. Second, each compartment was divided into three regions (anterior, central, and posterior) as described previously.⁹ A Student’s paired T-test was applied to each compartment and region of the ACL inured group (Injured-to-contralateral normal side) while accounting for multiple comparisons with the Benjamini-Hochberg method²¹ using the approach we have described.⁹ Injured-to-normal thickness differences that were greater than the 95% CI of normal and statistically different were considered thicker, while differences that were less than the 95% CI of normal and statistically different were considered thinner.

Results:

Subjects in the ACL and ACL-LM groups had similar sex, age, height, weight and BMI (Table 2). Likewise, the time interval between the index trauma and acquisition of the MRI data was similar between the groups: The average time interval between the index injury and MRI acquisition was 13.5 days (2 to 38 days) and 20.5 days (1 to 65 days) for those in the ACL-LM and ACL groups, correspondingly. Trauma occurred to the knee of dominant limb in 11 (65%) and 9 (47%) subjects in the ACL-LM and ACL groups, respectively.

Table 2:

Demographic and anthropometric data for subjects in the ACL + LM and ACL groups.

	ACL-LM	ACL	p-value
Sex	9 F, 8 M	13 F, 6 M
Age	17.8 (14 – 23) years	16.9 (14 – 21) years	0.26
Height	174.5 (157.5 – 193.0) cm	171.0 (188.0 – 160.0) cm	0.28
Weight	72.8 (55.8 – 98.9) kg	68.7 (54.4 – 124.7) kg	0.40
BMI	23.8 (19.3 – 26.6) kg/m2	23.3 (20.0 – 35.8) kg/m2	0.68

Following acquisition of the MRI scans subjects underwent arthroscopically assisted ACL reconstruction and treatment of concomitant injuries, and these data are presented to provide insight into the severity of concomitant trauma. For the ACL-LM group, 3 (17.5%) subjects underwent partial lateral meniscectomy, 10 (59 %) had tears that underwent repair, 3 (17.5%) had combined partial lateral meniscectomy and repair, and 1 (6.0%) tear was a partial thickness undersurface tear that was not treated (see supplement text for details on the location, type and treatment of the meniscus tears). Two subjects in the ACL-LM group had minor softening of their tibial articular cartilage (grade I changes) that were observed in the medial compartment in one subject and the lateral compartment in the other. There were no tibial articular cartilage lesions in the ACL group. There were no femoral or patella articular cartilage lesions in both groups.

Effect of combined ACL and lateral meniscus injury on morphometry of articular cartilage, meniscus, and subchondral bone

Analysis of subjects in the ACL-LM group revealed statistically significant differences in geometric profile of the articular cartilage and meniscus when comparing the injured leg to the contralateral uninjured leg; however, no differences in the measures of tibial plateau subchondral bone geometry were detected.(Table 3) In the lateral compartment of the ACL-LM group, the injured knee demonstrated increased posterior-inferior directed cartilage slopes. Compared to the normal contralateral side, L_ateralMCS and L_ateralPCS of the injured knee increased an average of 1.7 degrees (p = 0.03) and 3.6 degrees (p = 0.04), correspondingly. Further, there was an increase of the meniscus-cartilage wedge angle that approached statistical significance (L_ateralMCA of the injured knee increased an average of 3.3 degrees compared to the contralateral side; p = 0.06). In the medial compartment of the ACL-LM group, the injured knee had a 0.3 mm decrease of the cartilage-bone height (M_edialCBH) compared to the contralateral normal knee.(p = 0.01)

Table 3:

ACL-LM injury group. Data are presented for within subject comparisons with the convention used to calculate the difference based on the injured side minus the uninjured side.

	Units	Injured Knee (Mean)	Normal Knee (Mean)	Within subject injured-uninjured knee difference
				(Mean)	(Std Dev)	p value
Lat MBA	degrees	27.4	27.0	0.4	3.5	0.70
Lat MCA	degrees	43.5	40.2	3.3	6.7	0.06
Lat PCS	degrees	15.3	11.7	3.6	6.9	0.04
Lat MCS	degrees	1.2	−0.5	1.7	3.1	0.03
Lat BS	degrees	−0.8	−1.5	0.7	3.4	0.42
Lat MCH	mm	6.8	6.7	0.1	1.4	0.88
Lat CBH	mm	2.7	2.7	0.0	0.7	0.92
Med MBA	degrees	24.5	24.5	0.0	2.0	0.95
Med MCA	degrees	24.0	22.7	1.3	3.5	0.17
Med PCS	degrees	−3.8	−5.2	1.4	4.4	0.21
Med MCS	degrees	−0.9	−0.7	−0.2	2.0	0.64
Med BS	degrees	−3.3	−3.5	0.2	1.8	0.64
Med MCH	mm	5.6	5.6	0.0	1.1	0.88
Med CBH	mm	1.6	1.9	−0.3	0.4	0.01

The medial tibial compartment had a 30 mm² area of articular cartilage that underwent an increase in thickness (mean increase was 0.4 mm; p value range 0.01 to 0.05) relative to the contralateral normal knee, while a 1 mm² area decreased in thickness (mean decrease was 0.3 mm; p = 0.04) relative to the normal side (Figure 3). No statistically significant differences in cartilage thickness were detected in the lateral compartment of the tibia.

Figure 3:

Articular cartilage thickness differences. Cartilage thickness differences (injured – uninjured knees) that were greater than the 95% confidence interval based on side-to-side comparisons made in matched controls with normal knees (gray region) were observed for the tibial cartilage (top row) of the ACL-LM and ACL injury groups. Red and yellow regions represent an increase in cartilage thickness in the injured knee relative to the normal side while blue regions represent a decrease in cartilage thickness (p < 0.05). T-tests accounting for multiple comparisons identified locations with a statistically significant difference (p < 0.05) between injured and contralateral uninjured knees (bottom row).

At the time of acquisition of the MRI data, a side-to-side difference was observed in anterior-posterior (AP) position of the tibia relative to the femur in the ACL-LM group. In comparison to the contralateral uninjured knee, the tibia of the injured knee was located 2.5 mm more anterior relative to the femur (p = 0.001). No significant differences in the medial-lateral, flexion-extension, and internal-external rotation positions of the tibia relative to the femur were detected.

Effect of ACL injury on morphometry of articular cartilage, meniscus and subchondral bone

In the lateral compartment there was a decrease of the meniscus-cartilage height that approached statistical significance (L_ateralMCH of the injured knee decreased 0.5 mm; p = 0.06; Table 4).

Table 4:

Isolated ACL injury group. Data are presented for within subject comparisons with the convention used to calculate the difference based on the injured side minus the uninjured side.

	Units	Injured Knee (Mean)	Normal Knee (Mean)	Within subject injured-uninjured knee difference
				(Mean)	(StdDev)	p value
Lat MBA	degrees	26.4	28.2	−1.8	5.4	0.15
Lat MCA	degrees	41.2	40.8	0.4	7.7	0.82
Lat PCS	degrees	14.9	13.3	1.6	5.8	0.25
Lat MCS	degrees	1.2	1.2	0.0	3.1	0.99
Lat BS	degrees	0.0	0.7	−0.7	3.4	0.38
Lat MCH	mm	6.1	6.6	−0.5	1.0	0.06
Lat CBH	mm	2.7	2.8	−0.1	0.6	0.55
Med MBA	degrees	24.9	24.9	0.0	4.0	0.97
Med MCA	degrees	23.7	22.7	1.0	4.1	0.30
Med PCS	degrees	−5.4	−6.5	1.1	3.2	0.12
Med MCS	degrees	−1.4	−1.5	0.1	2.0	0.84
Med BS	degrees	−4.2	−4.3	0.1	2.0	0.75
Med MCH	mm	5.4	5.6	−0.2	0.8	0.35
Med CBH	mm	1.6	1.6	0.0	0.6	0.72

A 3 mm² area of cartilage on the medial side of the tibial plateau increased in thickness (mean 0.3 mm; p value range 0.04 to 0.05) (Figure 3). No regions underwent cartilage thinning, and no significant differences in cartilage thickness were observed in the lateral compartment of the tibial plateau.

Significant differences in the anterior-posterior position of the tibia relative to the femur at the time of acquisition of the MRI were found between ACL-injured knees in comparison to the contralateral uninjured knee. As with the ACL-LM injury group, tibias were more anteriorly displaced (mean = 3.3 mm) in ACL injured knees in comparison to the contralateral uninjured knees (p = 0.01). No differences in the medial-lateral, flexion-extension, and internal-external rotation positions of the tibia relative to the femur were detected.

Comparison of the injured knee between ACL-LM and isolated ACL groups

There was an order of magnitude increase of cartilage area that underwent thickening for the ACL-LM group (30 mm²) compared to the isolated ACL injury group (3 mm²) (Fig. 3). No differences in the measurements of articular cartilage, meniscus and subchondral bone geometry between the ACL-LM injury and the isolated ACL injury groups were detected (Table 5). Similarly, no differences were observed for position of the tibia relative to the femur between the ACL and ACL-LM groups.

Table 5:

Comparison between ACL-LM injury group versus isolated ACL injury groups. Convention used to calculate the mean difference used the combined injury group minus the isolated injury group.

	Units	Between group comparisons (ACL-LM injury - isolated ACL injury group differences)
		(Mean)	p value
Lat MBA	degrees	2.2	0.16
Lat MCA	degrees	2.9	0.25
Lat PCS	degrees	2.0	0.34
Lat MCS	degrees	1.8	0.10
Lat BS	degrees	1.4	0.23
Lat MCH	mm	0.5	0.20
Lat CBH	mm	0.1	0.76
Med MBA	degrees	0.1	0.95
Med MCA	degrees	0.2	0.86
Med PCS	degrees	0.2	0.86
Med MCS	degrees	−0.3	0.63
Med BS	degrees	0.1	0.93
Med MCH	mm	0.1	0.65
Med CBH	mm	−0.3	0.16

Discussion:

Findings from this study lead us to accept the hypothesis that combined ACL and LM injury alters meniscus and articular cartilage geometries of both medial and lateral compartments compared to the contralateral uninjured knee, and isolated injury to the ACL produces substantially less change to these structures. Notably, these changes were detected within an average of 3 weeks of the incident injury. The ACL-LM group underwent changes in the lateral compartment (increases in posterior-inferior directed slopes of articular cartilage surface, and wedge angle of the posterior horn of the meniscus) and medial compartment (cartilage-bone height decreased posteriorly and the articular cartilage underwent both thickening and thinning). Interestingly, we also observed change in the medial compartment of the ACL-LM knees that did not have a concomitant medial meniscus injury that was visible at the time of arthroscopy. This finding indicates that knee injuries involving combined trauma to the ACL and lateral meniscus affect both compartments of the knee, even if the meniscus and articular cartilage appears normal in the medial compartment upon arthroscopic visualization. Our study also demonstrated that the magnitude of change of the posterior-inferior directed articular cartilage slopes (Table 3; Lat PCS and Lat MCS) for the ACL-LM group were more than 2 times greater than the corresponding changes in subjects that suffered an isolated ACL injury (Table 4; Lat PCS and Lat MCS). It is unclear if this 2-fold increase was produced by transmission of an increased magnitude of compressive and shear stresses across the tibiofemoral joint at the time of the index trauma, the initial adaptive response of the knee to the complex changes in knee biomechanics associated with the trauma, or some combination of both. This increase in posterior-inferior slope is substantially larger than the side-to-side differences that were observed in the isolated ACL injury group (Table 4), and the side-to-side differences in control subjects with normal knees described previously.^9,10 Likewise, the change in articular cartilage thickness were considerably greater than the side-to-side differences observed in the isolated ACL injury group, and the side-to-side differences in control subjects with normal knees.^9,10,17

The ideal study design to assess the effect of knee trauma on articular cartilage and meniscus would characterize the knee prior to and then immediately after trauma (i.e. a within knee comparison); however, this approach is not feasible in the present design as it would require evaluation of thousands of subjects to generate the number and types of injuries that were included in the current study. Consequently, we chose a study design that was based on injured side-to-normal side comparisons (i.e. a within subject comparison). This approach was supported by our observation of side-to-side symmetry of articular cartilage and meniscus geometry in a similar group of subjects with normal knees and no history of trauma,¹⁰ and that all ACL-injured subjects had a contralateral knee that was normal upon MRI examination. Our study identified subjects at the time they suffered serious knee trauma and acquired MRI data within an average of three weeks of the incident trauma. It was not possible to acquire MRI data immediately after ACL trauma as the diagnosis of injury, enrollment, and acquisition of the MRI data requires time. This is important because the change in thickness and geometry of cartilage and meniscus associated with the index trauma is thought to occur after the initiation of compositional changes of cartilage matrix components.²² Consequently, the short time interval between the index injury and acquisition of MRI data captured the initial structural responses of the knee to trauma, and the underlying mechanism for these changes may have been produced by a change in cartilage matrix components that was initialed at the time of injury. This is supported by the work of Li et al. who demonstrated subjects with ACL and concomitant meniscus injury have a greater increase in T1rho and T2 times in adjacent articular cartilage from the time of injury through one year follow-up in comparison to those with ACL injury and normal menisci.²³ These findings emphasize the need for studying both the initial and temporal responses of the articular cartilage and meniscus to trauma in an effort to determine if the changes we observed either resolve or continue to progress over time as subjects undergo surgery, rehabilitation, and return to preinjury activities.

Prior work has shown that increased posterior-inferior directed slope of the tibial plateau is associated with increased risk of suffering an ACL^18,19 and ACL graft injury,²⁴ and the increased lateral compartment posterior-inferior directed slopes found in the current study (3.6 degrees for the ACL-LM group and 1.6 for the ACL group) further exacerbates the chance of a second injury to the ACL graft.^24–26 In addition, there is evidence that geometry of the articular cartilage and tibial plateau subchondral bone are associated with worsening cartilage degenerative changes in the lateral tibial plateau.²⁷ Considered in combination, these findings introduce the hypothesis that once ACL injury has occurred the same geometric characteristics associated with increased risk of suffering the index injury --- increased posterior-inferior directed slope of the articular surfaces of the tibia^10,17–19 --- are also associated with increased risk of developing PTOA. If future work supports this hypothesis, our finding of a two-fold increase in slope for the combined ACL-LM group in comparison to the isolated ACL group may provide a biomechanical explanation for why the risk of PTOA is greater for those that suffer combined injuries to the ACL and meniscus in comparison to those that suffer isolated ACL trauma.

Many different approaches have been used to analyze cartilage thickness maps^9,11,28–33 and our approach was based on a two-step process. First data from the matched control subjects with normal knees was used to determine the side-to-side thickness differences and corresponding 95% confidence limits (i.e. normal thickness maps) at each 1mm by 1mm grid location. This established the limits (both thickening and thinning) outside of which subjects in ACL-LM and ACL groups were considered to have injured-to-uninjured side differences in cartilage thickness that were greater than the normal side-to-side differences of articular cartilage thickness, and consequently were meaningful and could be detected. These limits were applied to the injured subjects and used to identify the regions of the thickness difference maps where the injured-to-normal side differences were greater and less than the 95% confidence limits of normal (Fig 3 top left and right panels). The second step involved statistical analysis of the data that were outside the limits of normal with corrections for multiple comparisons using the Benjamini-Hochberg method (Fig 3 bottom left and right panels).²¹ This revealed that both ACL-LM and ACL groups underwent cartilage thickening, and this has been observed during the first five years after injury in humans,^9,11 and in an animal study of combined trauma to the ACL and meniscus with 3 months of healing.³⁴ In the later study mechanical tests revealed a 19% increase in tibial articular cartilage thickness 3 months after injury and this supports our finding of a 14% increase in thickness (calculated as a mean increase in thickness of 0.4 mm relative to a mean 2.8 mm thick cartilage obtained from normal knees at the same location). Our measurements were taken soon after injury (mean 17 days) and are concerning as there is strong evidence that articular cartilage and its matrix components change soon after ACL injury, reconstruction and return to activity.^23,35–40 In addition, following severe knee trauma that involves the ACL, increased T1rho and T2 relaxation times are thought to reflect loss of proteoglycan and change in water content that precede changes to cartilage thickness associated with the progression of PTOA. This may be the underlying mechanism associated with cartilage thickening that was observed in the current study.²³

Our investigation acquired MRI data with subjects non-weightbering and found the injured knee in both ACL and ACL-LM groups had anterior positioning of the tibia relative to the femur compared to the contralateral uninjured side at the time of data acquisition. Amongst the control subjects, there were no side-to-side differences in the position of the tibia relative to the femur. These findings demonstrate that the ACL disruption produced the anterior positioning of the tibia relative to the femur even when the subject was non-weightbearing at the time of MRI data acquisition, and substantiates prior reports in the literature.^1,41,42 Similarly, our prior work that was based on a larger sample size used the same MRI-based measurement approach and revealed both anterior and medial positioning of the tibia relative to the femur for ACL injured knees in comparison to the contralateral normal side with no side-to-side differences in controls with normal knees.⁹ Likewise, prior work using a weight-bearing radiographic technique demonstrated similar magnitudes of anterior¹ and medial⁴³ positioning of the tibia relative to the femur following ACL injury with no side-to-side differences in control subjects.⁴³ Abnormal anterior and medial positioning of the tibia relative to the femur represent concerns as they may be related to degenerative changes about the knee as indicated by Siriwanarangsun and colleagues.⁴⁴

Strengths of the current study include evaluation of subjects soon after injury and data analyses based on three-dimensional bone-based coordinate systems as this allowed us to establish the complete 6 degree-of-freedom position of the tibia relative to the femur during MRI acquisition and make reliable measures of the articular structures. The cases and controls had similar body composition (Table 2). In addition, we compared two groups that suffered their first noncontact ACL injury and arthroscopic examination of the knee revealed that both groups had articular cartilage that appeared normal. Consequently, the two groups only differed by the condition of the lateral meniscus (i.e. either normal or torn). For all ACL injured subjects, the contralateral limb had no history of injury, was normal on MRI examination, and this allowed within subject comparisons to be made with an approach that eliminated between subject variability. Further, the study included control subjects with normal MRIs that were matched with the injured subjects (based on sex, age, and participation on the same sports team at the time of ACL injury) and this provided control of the level, type and magnitude of activity that subjects participated in prior to suffering knee trauma. A consistent definition of non-contact injury was applied to all subjects, and subjects were contacted soon after injury to minimize recall bias, confirm that the injury was indeed produced by a non-contact mechanism, and determine the time interval between the index injury and data acquisition. This minimized confounding factors that may have been introduced into the study by including subjects with differences in joint biomechanics between non-contact and contact injury mechanisms.

There are potential limitations associated with our study. The time interval between injury and MRI acquisition was not different between the groups (average of 13.5 and 20.5 days for subjects in the ACL-LM and ACL injury groups, respectively); however, the time interval was not the same across subjects. Our prior work on the larger cohort from which this sample was selected revealed that differences in the time interval between the index ACL injury and acquisition of MRI data were not associated with differences in articular cartilage thickness.⁹ This suggests that the comparisons made between the two groups were not affected by the time interval between the index ACL injury and acquisition of the MRI data.

In our study, gross changes to the macro-level geometry of the subchondral tibial bone were not identified less than 3 weeks after injury. Bowes et al. observed flattening and bone protuberances at the periphery of the femoral condyles at 3 month follow-up; however, in their cohort of patients suffering ACL injury.¹³ This suggests that early geometric changes soon after ACL injury occur within the soft tissue structures, while bony changes evolve a few months later. On the other hand, Bowes et al. also noted bony changes in the medial femur, which was not measured in our study. Their findings indicate that characterizing femoral geometry may be an important area for future work. Altogether, direct comparison of our study to Bowes et al. is limited because their cohort included patients undergoing surgical and nonsurgical treatment of their ACL injury, and did not compare those with combined ACL and meniscus injury to those with ACL injury and a normal meniscus.

We examined the early effects of trauma on the geometry of the articular structures of the knee and it is not possible for us to differentiate between transient changes created by the index trauma and long-term changes that progress to the onset of PTOA. It is possible that the changes we have observed following combined ACL and lateral meniscal injury resolve and do not lead to PTOA, given that a previous study found that partial lateral meniscectomy was associated with better patient reported outcomes.⁴⁵ Of the variables measured in the current study, it appears as if the earliest macro-level geometric changes of the articular structures of individuals that suffer their first severe knee trauma involves the meniscus and articular cartilage, and not gross changes to the macro-level geometric characteristics of the underlying subchondral tibial bone. Future work is necessary to characterize the micro-level temporal response of the articular structures and subchondral bone to knee trauma and this work should consider all of the articular cartilage and meniscus injuries that occur in combination with the index ACL injury.^46,47

Clinical significance:

Knee injuries that involve combined trauma to the ACL and meniscus directly affect both compartments of the knee, even if the meniscus and articular cartilage appears normal upon arthroscopic examination.