Relationship of meniscal damage, meniscal extrusion, malalignment, and joint laxity to subsequent cartilage loss in osteoarthritic knees

Authors


Abstract

Objective

Progressive knee osteoarthritis (OA) is believed to result from local factors acting in a systemic environment. Previous studies have not examined these factors concomitantly or compared quantitative and qualitative cartilage loss outcomes. The aim of this study was to test whether meniscal damage, meniscal extrusion, malalignment, and laxity each predicted tibiofemoral cartilage loss after controlling for the other factors.

Methods

Laxity and alignment were measured at baseline in individuals with knee OA. Magnetic resonance imaging included spin-echo coronal and sagittal imaging for meniscal scoring and axial and coronal spoiled gradient echo sequences with water excitation for cartilage quantification. Tibial and weight-bearing femoral condylar subchondral bone area and cartilage surface were segmented. Cartilage volume, denuded bone area, and cartilage thickness were quantified in each plate, with progression defined as cartilage loss >2 times the coefficient of variation for each plate. Qualitative outcome was assessed as worsening of the cartilage score. Logistic regression analysis with generalized estimating equations yielded odds ratios for each factor, adjusting for age, sex, body mass index, and the other factors.

Results

We studied 251 knees in 153 persons. After full adjustment, medial meniscal damage predicted medial tibial cartilage volume loss and tibial and femoral denuded bone increase, while varus malalignment predicted medial tibial cartilage volume and thickness loss and tibial and femoral denuded bone increase. Lateral meniscal damage predicted every lateral outcome. Laxity and meniscal extrusion had inconsistent effects. After full adjustment, no factor except medial laxity predicted qualitative outcome.

Conclusion

Using quantitative cartilage loss assessment, local factors that independently predicted tibial and femoral loss included medial meniscal damage and varus malalignment (medially) and lateral meniscal damage (laterally). A measurement of quantitative outcome was more sensitive at revealing these relationships than a qualitative approach.

Progressive knee osteoarthritis (OA) is believed to result from local mechanical factors acting in a systemic environment (1, 2), although there is as yet little direct evidence of this from magnetic resonance imaging (MRI)–based natural history studies. Healthy menisci, more neutral alignment, and joint stability all protect the articular cartilage from concentrations of stress (3–5). When these factors are altered or impaired, stress is not well distributed, and it increases focally, potentially leading to articular cartilage damage. Meniscal damage, meniscal extrusion, varus–valgus malalignment, and medial–lateral laxity are local factors that may be present in primary knee OA (6–15). Their effect on load distribution and attenuation are especially important in the damaged and more vulnerable OA knee.

Natural history studies of knee OA ideally should consider meniscal damage, malalignment, and laxity together to determine if any effect on cartilage loss persists after adjusting for the other local factors, and thereby address the strong possibility of confounding. In the very few progression studies that have evaluated these factors, either meniscal damage (16–18) or alignment (12) has been considered, but not both together, and no studies have included laxity.

The emphasis on the radiographic joint space in the literature on knee OA progression reflects the intent to capture cartilage loss. However, radiographic outcomes are inherently limited in their ability to detect the impact of certain local factors. Without contrast, radiography cannot distinguish articular cartilage and meniscal tissue; since the meniscus contributes to the radiographic joint space, joint space change cannot be used to study the effect of meniscal damage on OA progression. Medial laxity (medial joint line opening) and lateral laxity (lateral joint line opening) each stresses cartilage in the compartment opposite the side of opening. A separate assessment of medial laxity effect versus lateral laxity effect requires an outcome tool capable of detecting progression equally well in the medial and lateral tibiofemoral compartment of the same knee. Knee radiography can only reveal progression in the tibiofemoral compartment that is predominantly involved; reciprocal widening of the other compartment makes it impossible to gauge progression there. MRI approaches, in contrast, reveal outcome equally well in each tibiofemoral compartment and surface within the same knee.

Notably, almost all published knee OA progression studies with MRI outcome measures have used qualitative cartilage assessment to define cartilage loss. It is unclear at present whether a more quantitative assessment of cartilage loss is more sensitive than the qualitative approach previously used. To date, no study of OA risk factors has included both outcomes.

Using MRI-based quantitative measures of cartilage loss, we tested 2 hypotheses. First, the local factors of medial meniscal damage, medial meniscal extrusion, varus malalignment, and lateral laxity are each associated with a reduction in cartilage volume and thickness and an increase in denuded bone area in the medial tibial and medial femoral surfaces after adjusting for age, sex, body mass index (BMI), and the other local factors. Second, the local factors of lateral meniscal damage, lateral meniscal extrusion, valgus malalignment, and medial laxity are each associated with a reduction in cartilage volume and thickness and an increase in denuded bone area on the lateral tibial and lateral femoral surfaces after adjusting for age, sex, BMI, and the other local factors. We then tested both hypotheses by applying a qualitative MRI outcome measure, i.e., worsening of the cartilage integrity score.

PATIENTS AND METHODS

Sample.

Study participants were members of a cohort participating in a natural history study of knee OA, the MAK-2 (Mechanical Factors in Arthritis of the Knee, second cycle). MAK-2 participants were recruited from community sources, e.g., through advertising in periodicals targeting elderly persons, neighborhood organizations, letters to members of the registry of the Beuhler Center on Aging at Northwestern University, and via medical center referrals.

Inclusion criteria were the definite presence of tibiofemoral osteophytes (grade ≥2 on the Kellgren/Lawrence [K/L] radiographic scale [19]) in 1 or both knees, and a Likert category of at least “a little difficulty” for ≥2 items on the Western Ontario and McMaster Universities Osteoarthritis Index physical function scale (20). Exclusion criteria were having received a corticosteroid injection within the previous 3 months; a history of avascular necrosis, rheumatoid or other inflammatory arthritis, periarticular fracture, Paget's disease, villonodular synovitis, joint infection, ochronosis, neuropathic arthropathy, acromegaly, hemochromatosis, gout, pseudogout, osteopetrosis, or meniscectomy; or meeting exclusion criteria for MRI, such as the presence of a pacemaker, artificial heart valve, aneurysm clip or shunt, metallic stent, implanted device (e.g., pain control/nerve stimulator, defibrillator, insulin/drug pump, or ear implant), or any metallic fragment in an eye.

Approval was obtained from the Office for the Protection of Research Subjects and the Institutional Review Boards of Northwestern University and Evanston Northwestern Healthcare. Written consent was obtained from all participants.

Meniscal damage and meniscal extrusion.

Images of both knees of all participants were obtained using a commercial knee coil and 1 of 2 whole-body scanners, a 1.5T Symphony (Siemens, Erlangen, Germany) or a 3T Genesis Signa (GE Healthcare, Waukesha, WI). The protocol included coronal T1-weighted spin-echo (SE), sagittal fat-suppressed dual-echo turbo SE, and axial and coronal T1-weighted, fat-suppressed 3-dimensional fast low-angle shot sequences.

Meniscal damage and extrusion were graded in each knee using the Whole-Organ MRI Score (WORMS) (21). The anterior horn, posterior horn, and body of the medial and lateral meniscus were each graded on a scale of 0–4, where 0 = intact, 1 = minor radial or parrot-beak tear, 2 = nondisplaced tear, 3 = displaced tear or partial maceration, and 4 = complete maceration and destruction. The highest grade for each meniscus was used in the analysis. Extrusion of the body segments of the medial and lateral menisci was graded (0 if none, 1 if less than half of the meniscus, or 2 if more than half) using coronal images at the level of the medial collateral ligament and lateral collateral ligament, respectively. MRIs were read by 1 of 3 experienced readers (AG). The interobserver reliability of the readers applying this scoring system has been published previously (21); the intraclass correlation coefficient (ICC) was 0.94 for medial meniscal damage and 0.81 for lateral meniscal damage. Readers were blinded with regard to all outcome data.

Varus–valgus alignment.

To assess the hip–knee–ankle angle, a single anteroposterior radiograph of both lower extremities was obtained using a 51 × 14–inch graduated grid cassette (Reina Imaging, Crystal Lake, IL) to include the full limb of tall participants. By filtering the x-ray beam in a graduated manner, this cassette accounted for the unique soft tissue characteristics of the hip and ankle. The tibial tubercle, a site adjacent to the knee that was not distorted by OA, was used as the positioning landmark. Participants stood without footwear, with the tibial tubercles facing forward. The x-ray beam was centered at the knee at a distance of 2.4 meters. Settings of 100–300 mA/second and 80–90 kV were used, depending on limb size and tissue characteristics. All radiographs were obtained in the same unit by 2 trained technicians.

Alignment, i.e., the hip–knee–ankle angle, was measured as the angle formed by the intersection of the line connecting the centers of the femoral head and the intercondylar notch with the line connecting the centers of the surface of the ankle talus and the tips of the tibial spines. Our reading reliability, in image sets including both varus- and valgus-aligned knees, was excellent (ICC 0.98–0.99) (12). Alignment was analyzed as a continuous variable.

Medial–lateral laxity.

Medial and lateral laxity were measured using a previously described device (15), consisting of a bench and an attached arc-shaped, low-friction track (30- cm in radius measured from the center of the knee and running medially and laterally) and providing thigh and ankle immobilization and a stable knee flexion angle. The distal shank (30 cm from the knee) was immobilized in a sled that traveled within the track. A hand-held dynamometer fitted into either side of the sled was used to apply a fixed varus and valgus load. Laxity was measured in degrees as the angular deviation after fixed load in each direction. All laxity measurements were performed by the same examiner and assistant (DK and MM). Our reliability with this device testing persons with knee OA and varying body habitus was very good (within-session ICC 0.85–0.96, between-session ICC 0.84–0.90) (15). Medial and lateral laxity were separately analyzed as continuous variables.

Quantitative measurement of cartilage loss.

Articular cartilage was quantified at baseline and 2 years later. For quantitative measurements, coronal spoiled gradient echo sequences with water excitation were acquired, with a slice thickness of 1.5 mm and an in-plane resolution of 0.31 mm (field of view 16 cm, 512 × 512–pixel matrix, number of excitations 1). The repetition time, echo time, and flip angle, respectively, were 18.6 msec/9.3 msec/15° on the 1.5T scanner, and 12.2 msec/5.8 msec/9° on the 3T scanner. The total area of subchondral bone and the area of the cartilage surface were segmented for the medial tibial and lateral tibial surfaces, and in the weight-bearing portion of the medial and lateral femoral condyle using proprietary software (Chondrometrics, Ainring, Germany) (22–24). Cartilage volume, percentage of subchondral bone covered with cartilage, denuded subchondral bone area, and the average thickness of cartilage, including areas of denuded subchondral bone as 0 mm, were quantified. Baseline and 2-year scans were read together, with the reader blinded with regard to the order of acquisition.

Using the methodology applied here (image acquisition at 1.5T, double-oblique coronal acquisitions, 1.5-mm slice thickness, image analysis by experienced readers, and Chondrometrics software), the precision errors (coefficient of variation [CV] for 2 acquisitions with repositioning) for cartilage volume, cartilage thickness, and denuded bone area, respectively, were as follows: for the medial tibia 2.6%/2.1%/1.1%, for the medial weightbearing femur 3.2%/3.0%/1.4%, for the lateral tibia 2.1%/2.1%/1.2%, and for the lateral weight-bearing femur 3.7%/3.0%/1.7% (23). These reliability measures are consistent with those reported in the literature and recently summarized (25); in other studies, interscan (intraobserver) precision errors for cartilage volume have ranged from 2.1% in the medial tibia to 6.7% in the lateral tibia.

Qualitative assessment of cartilage loss.

Specifically, 3 regions (anterior, central, and posterior) of the medial and lateral femoral condyles and tibial plateaus were each scored separately for cartilage morphology following a detailed reading protocol, including visual illustrations of each grade (21). In each region, cartilage morphology was graded on a scale of 0–6, where 0 = normal thickness and signal, 1 = normal thickness but increased signal on T2-weighted images, 2 = solitary focal partial- or full-thickness defect ≤1 mm in width, 3 = multiple areas of partial-thickness loss or a grade 2 lesion >1 mm, with areas of preserved thickness, 4 = diffuse, >75%, partial-thickness loss, 5 = multiple areas of full-thickness loss, or a full-thickness lesion >1 mm, with areas of partial-thickness loss, and 6 = diffuse, >75%, full-thickness loss. The ICC for these readers was 0.98 for cartilage integrity in the medial regions and 0.99 in the lateral regions (21).

Acquisition and reading of radiographs.

All participants had bilateral, anteroposterior, weight-bearing knee radiographs performed at baseline in the semiflexed position with fluoroscopic confirmation of superimposition of the anterior and posterior tibial plateau lines and centering of the tibial spines within the femoral notch (for full protocol, see ref. 26). To describe the knees, the K/L global radiographic score was used (0 = normal, 1 = possible osteophytes, 2 = definite osteophytes without definite joint space narrowing, 3 = definite joint space narrowing, some sclerosis, and possible attrition, and 4 = large osteophytes, marked narrowing, severe sclerosis, and definite attrition). Reliability of radiographic grading for the single reader was high (κ = 0.85–0.86).

Statistical analysis.

Two-year progression was defined as cartilage loss >2 times the CV (previously determined for each measure of cartilage loss in each cartilage plate; see ref. 23). This definition of progression yielded the following cut points: in the medial tibia, cartilage volume loss 5.2%, cartilage thickness loss 4.2%, denuded bone area 2.2%; in the medial weight-bearing femur, cartilage volume loss 6.4%, cartilage thickness loss 6.0%, denuded bone area 2.8%; in the lateral tibia, cartilage volume loss 4.2%, cartilage thickness loss 4.2%, denuded bone area 2.4%; and in the lateral weight-bearing femur, cartilage volume loss 7.4%, cartilage thickness loss 6.0%, denuded bone area 3.4%. The unit of analysis was the knee. Generalized estimating equations were used to estimate logistic regression models that validly included potentially correlated observations between knees in the same individual. Separate modeling was performed for each tibiofemoral cartilage surface. The effects of baseline meniscal damage, meniscal extrusion, medial–lateral laxity, and varus–valgus malalignment on cartilage loss measurements were expressed as odds ratios (ORs) with the associated 95% confidence intervals (95% CIs).

All logistic regression models controlled for age, sex, and BMI. Full logistic models for the lateral cartilage surface outcomes additionally controlled for lateral meniscal damage, lateral meniscal extrusion, valgus malalignment severity (continuous variable), and medial laxity (continuous variable). Full logistic models for the medial cartilage surface outcomes additionally controlled for medial meniscal damage, medial meniscal extrusion, varus malalignment severity, and lateral laxity.

Two-year progression was then examined, based on qualitative cartilage integrity scores. First, the original WORMS system was used, and then the recently introduced modification (18). In this modification, the original WORMS values of 0 and 1 were collapsed to 0, the original scores of 2 and 3 were collapsed to 1, and the original scores of 4, 5, and 6 were considered to be 2, 3, and 4, respectively. For both of these qualitative approaches, progression was defined as a worsening of the cartilage integrity score in 1 or more regions of a given compartment from the baseline to the followup evaluation.

RESULTS

Of 165 persons evaluated at baseline, 12 (7%) were lost to followup because of poor health or because they had moved away or had undergone bilateral total knee replacement. We studied 251 knees in 153 persons with knee OA. The mean ± SD age of the participants was 66.4 ± 11.0 years and the mean ± SD BMI was 30.1 ± 5.9 kg/m2. The K/L score in most knees was 2 (41%) or 3 (33%) at baseline. Mean ± SD varus–valgus alignment at baseline was 0.11 ± 4.8 degrees in the varus direction. Baseline medial and lateral meniscal damage scores were 1.29 ± 1.59 and 0.77 ± 1.46, respectively, and the medial and lateral meniscal extrusion scores were 0.52 ± 0.74 and 0.25 ± 0.57, respectively. Mean ± SD medial laxity at baseline was 2.52 ± 1.28 degrees and mean lateral laxity was 4.14 ± 1.53 degrees. Baseline values and values for the 2-year change in cartilage volume, cartilage thickness, and area of denuded bone at baseline are described in Table 1. Tables 2 and 3 show the correlations between local factors at baseline. Although some factors were moderately correlated, their association did not introduce harmful multicollinearity into our multiple regression models based on standard criteria (27).

Table 1. Baseline and 2-year change in cartilage measurements*
 Medial compartmentLateral compartment
BaselineBaseline to 2-year changeBaselineBaseline to 2-year change
Mean ± SDMedian (IQR)Mean ± SDMedian (IQR)Mean ± SDMedian (IQR)Mean ± SDMedian (IQR)
  • *

    Change is described/defined as a decrease for the cartilage volume and thickness and an increase for the area of denuded bone. Data are from the right knee of each participant. IQR = interquartile range.

  • The proportion of knees with no denuded bone at baseline was 79% for the medial tibia, 82% for the medial femur, 79% for the lateral tibia, and 83% for the lateral femur.

  • The proportion of knees with no denuded bone at both baseline and the 2-year followup was 74% for the medial tibia, 78% for the medial femur, 77% for the lateral tibia, and 78% for the lateral femur.

Tibia        
 Cartilage volume, mm31,943.30 ± 551.791,879 (1,610, 2,286)53.63 ± 160.3537 (−17, 114)1,861 ± 686.831,773 (1,361, 2,281)41.71 ± 122.1430 (−21, 108)
 Cartilage thickness, mm1.64 ± 0.321.68 (1.48, 1.83)0.05 ± 0.110.04 (−0.01, 0.09)1.83 ± 0.491.88 (1.54, 2.17)0.05 ± 0.110.04 (−0.01, 0.01)
 Area of denuded bone, cm20.41 ± 1.160 (0, 0)0.23 ± 0.620 (0, 0)0.45 ± 1.320 (0, 0)0.08 ± 0.490 (0, 0)
Femur (weight-bearing portion)        
 Cartilage volume, mm31,028.13 ± 371.261,016 (787, 1,238)32.06 ± 91.0916 (−21, 64)1,184.63 ± 334.741,161 (929, 1,429)22.13 ± 93.6712 (−40, 80)
 Cartilage thickness, mm1.62 ± 0.471.66 (1.38, 1.96)0.05 ± 0.140.03 (−0.02, 0.1)1.77 ± 0.351.76 (1.56, 2)0.03 ± 0.120.02 (−0.05, 0.1)
 Area of denuded bone, cm20.33 ± 0.960 (0, 0)0.09 ± 0.310 (0, 0)0.13 ± 0.420 (0, 0)0.07 ± 0.250 (0, 0)
Table 2. Spearman's correlation coefficients for associations between local medial compartment factors at baseline*
 Medial meniscal damageMedial meniscal extrusionVarus malalignmentLateral laxity
  • *

    Coefficients were determined using data from the right knee of each participant.

Medial meniscal damage1.000.620.49−0.09
Medial meniscal extrusion0.621.000.36−0.07
Varus malalignment0.490.361.00−0.27
Lateral laxity−0.09−0.07−0.271.00
Table 3. Spearman's correlation coefficients for associations between local lateral compartment factors at baseline*
 Lateral meniscal damageLateral meniscal extrusionValgus malalignmentMedial laxity
  • *

    Coefficients were determined using data from the right knee of each participant.

Lateral meniscal damage1.000.540.290.13
Lateral meniscal extrusion0.541.000.280.002
Valgus malalignment0.290.281.000.14
Medial laxity0.130.0020.141.00

As shown in Table 4, in models adjusted for age, sex, and BMI, medial meniscal damage significantly increased the likelihood of cartilage volume loss, cartilage thickness decrease, and denuded bone increase in both the medial tibial and the medial weight-bearing femoral cartilage plates. In the fully adjusted models (i.e., adjusted for age, sex, BMI, medial meniscal extrusion, varus malalignment, and lateral laxity), a significant relationship between medial meniscal damage and the following outcomes persisted: medial tibial cartilage volume loss, medial tibial denuded bone increase, and medial weight-bearing femoral denuded bone increase. Medial meniscal extrusion was significantly associated with every outcome for both plates in the models adjusted for age, sex, and BMI, but in none of the fully adjusted models, although the relationship approached significance for medial weight-bearing femoral cartilage thickness loss and denuded bone increase.

Table 4. Relationship between local mechanical factors and quantitative cartilage loss outcomes in the medial tibiofemoral cartilage plates*
Measure of cartilage loss, baseline to 2 years (dependent variable)
Local mechanical factor, at baseline (independent variable)Medial tibiaMedial weight-bearing femur
Cartilage volume loss, OR (95% CI)Cartilage thickness loss, OR (95% CI)Denuded bone increase, OR (95% CI)Cartilage volume loss, OR (95% CI)Cartilage thickness loss, OR (95% CI)Denuded bone increase, OR (95% CI)
Adjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariates
  • *

    Odds ratios (ORs) and 95% confidence intervals (95% CIs) for the baseline to 2-year quantitative cartilage loss outcome associated with each local factor were first adjusted for age, sex, and body mass index (BMI), and were then adjusted for all covariates (age, sex, BMI, medial meniscal damage, medial meniscal extrusion, varus malalignment, and lateral laxity). Progression, defined as cartilage loss >2 times the coefficient of variation, was observed in 67 knees (27%) for tibial cartilage volume, 77 knees (31%) for tibial cartilage thickness, 49 knees (20%) for tibial area of denuded bone, 68 knees (27%) for femoral cartilage volume, 75 knees (30%) for femoral cartilage thickness, and 39 knees (16%) for femoral area of denuded bone. Data are from 251 knees in 153 participants.

  • OR per unit score.

  • OR per 1 degree.

Medial meniscal damage1.57 (1.29, 1.91)1.29 (1.02, 1.64)1.40 (1.16, 1.69)1.07 (0.84, 1.37)3.44 (2.37, 5.01)2.42 (1.56, 3.75)1.32 (1.09, 1.60)1.10 (0.87, 1.38)1.39 (1.16, 1.68)1.19 (0.94, 1.50)2.42 (1.76, 3.32)1.69 (1.25, 2.28)
Medial meniscal extrusion1.99 (1.36, 2.91)1.21 (0.79, 1.87)1.81 (1.23, 2.65)1.27 (0.78, 2.06)4.98 (2.94, 8.43)1.77 (0.82, 3.85)1.68 (1.15, 2.44)1.28 (0.84, 1.96)1.93 (1.35, 2.77)1.46 (0.97, 2.20)3.55 (2.25, 5.61)1.62 (0.98, 2.68)
Varus malalignment1.16 (1.07, 1.26)1.11 (1.01, 1.21)1.20 (1.10, 1.31)1.18 (1.07, 1.30)1.41 (1.20, 1.65)1.22 (1.06, 1.41)1.12 (1.04, 1.20)1.09 (1.00, 1.18)1.10 (1.02, 1.17)1.05 (0.97, 1.14)1.35 (1.22, 1.50)1.21 (1.10, 1.32)
Lateral laxity1.24 (1.02, 1.51)1.22 (1.00, 1.49)1.17 (0.95, 1.44)1.17 (0.94, 1.44)1.11 (0.87, 1.42)1.04 (0.75, 1.43)1.03 (0.85, 1.25)1.02 (0.83, 1.24)0.94 (0.77, 1.15)0.91 (0.74, 1.12)1.01 (0.82, 1.24)0.91 (0.72, 1.15)

Varus malalignment was significantly associated with every outcome for the medial tibial and femoral plates after adjusting for age, sex, and BMI. In the fully adjusted models, a significant relationship persisted between varus malalignment and each of the following medial outcomes: tibial cartilage volume loss, tibial cartilage thickness loss, tibial denuded bone increase, and weight-bearing femoral denuded bone increase. In contrast, lateral laxity was associated only with medial tibial cartilage volume loss.

In the lateral compartment, lateral meniscal damage was significantly associated with every outcome in both the lateral tibial and the lateral femoral surfaces in the models adjusted for age, sex, and BMI, and the relationship persisted in all of the fully adjusted models (i.e., adjusted for age, sex, BMI, lateral meniscal extrusion, valgus malalignment, and medial laxity) (Table 5). Lateral meniscal extrusion was significantly associated with every outcome for both joint surfaces after adjustment for age, sex, and BMI; in the fully adjusted models, a significant relationship persisted only for lateral tibial denuded bone increase. In the models adjusted for age, sex, and BMI, valgus malalignment was significantly associated with lateral tibial denuded bone increase, lateral weight-bearing femoral cartilage volume loss, and lateral weight-bearing femoral cartilage thickness loss, but the relationship did not persist in any of the fully adjusted models. In the fully adjusted models, medial laxity was significantly associated only with an increase in lateral weight-bearing femoral denuded bone.

Table 5. Relationship between local mechanical factors and quantitative cartilage loss outcomes in the lateral tibiofemoral cartilage plates*
Measure of cartilage loss, baseline to 2 years (dependent variable)
Local mechanical factor, at baseline (independent variable)Lateral tibiaLateral weight-bearing femur
Cartilage volume loss, OR (95% CI)Cartilage thickness loss, OR (95% CI)Denuded bone increase, OR (95% CI)Cartilage volume loss, OR (95% CI)Cartilage thickness loss, OR (95% CI)Denuded bone increase, OR (95% CI)
Adjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariates
  • *

    ORs and 95% CIs for the baseline to 2-year quantitative cartilage loss outcome associated with each local factor were first adjusted for age, sex, and BMI, and were then adjusted for all covariates (age, sex, BMI, lateral meniscal damage, lateral meniscal extrusion, valgus malalignment, and medial laxity). The proportion of knees with progression was 97 of 251 (39%) for tibial cartilage volume, 95 of 251 (38%) for tibial cartilage thickness, 49 of 251 (20%) for tibial area of denuded bone, 53 of 251 (21%) for femoral cartilage volume, 61 of 251 (24%) for femoral cartilage thickness, and 37 of 251 (15%) for femoral area of denuded bone. Data are from 251 knees in 153 participants. See Table 4 for definitions.

  • OR per unit score.

  • OR per 1 degree.

Lateral meniscal damage1.54 (1.26, 1.87)1.45 (1.14, 1.85)1.69 (1.39, 2.06)1.62 (1.28, 2.06)2.46 (1.96, 3.09)2.11 (1.64, 2.72)1.78 (1.43, 2.20)1.62 (1.27, 2.07)1.75 (1.42, 2.17)1.66 (1.30, 2.12)1.97 (1.54, 2.53)1.80 (1.37, 2.37)
Lateral meniscal extrusion2.11 (1.28, 3.47)1.41 (0.80, 2.48)2.25 (1.36, 3.73)1.33 (0.74, 2.40)4.54 (2.66, 7.74)2.19 (1.18, 4.04)2.30 (1.38, 3.85)1.22 (0.66, 2.27)1.93 (1.21, 3.06)0.95 (0.52, 1.75)2.95 (1.81, 4.81)1.66 (0.95, 2.87)
Valgus malalignment1.05 (0.99, 1.12)0.99 (0.93, 1.05)1.07 (1.00, 1.15)0.99 (0.93, 1.06)1.20 (1.08, 1.34)1.02 (0.92, 1.12)1.13 (1.01, 1.26)1.02 (0.93, 1.13)1.14 (1.03, 1.26)1.04 (0.96, 1.14)1.12 (0.97, 1.30)0.97 (0.87, 1.08)
Medial laxity1.12 (0.90, 1.38)1.07 (0.85, 1.35)1.08 (0.86, 1.34)1.01 (0.79, 1.29)1.11 (0.86, 1.43)0.99 (0.72, 1.36)1.26 (0.97, 1.65)1.18 (0.82, 1.67)1.32 (1.04, 1.69)1.25 (0.91, 1.71)1.50 (1.16, 1.95)1.48 (1.06, 2.07)

Considering the medial tibiofemoral progression outcome identified from the 2 qualitative approaches, the original and the modified WORMS systems, significant relationships were detected for medial meniscal damage, medial meniscal extrusion, and varus malalignment after adjusting for age, sex, and BMI but not after further adjustment for the other local factors (Table 6). In the lateral compartment, a significant relationship was not detected for any of the local factors other than medial laxity (Table 7).

Table 6. Relationship between local mechanical factors and qualitative cartilage loss outcomes in the medial tibiofemoral compartment*
 WORMS score of cartilage integrity, OR (95% CI)Modified WORMS score, OR (95% CI)
Adjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariates
  • *

    Odds ratios (ORs) and 95% confidence intervals (95% CIs) for the baseline to 2-year qualitative cartilage loss outcome associated with each local factor were first adjusted for age, sex, and body mass index (BMI), and were then adjusted for all covariates (age, sex, BMI, medial meniscal damage, medial meniscal extrusion, varus malalignment, and lateral laxity). The proportion of knees with progression was 42 of 251 (17%) using the Whole-Organ Magnetic Resonance Imaging Score (WORMS) and 35 of 251 (14%) using the modified WORMS. Data are from 251 knees in 153 participants.

  • OR per unit score.

  • OR per 1 degree.

Medial meniscal damage1.38 (1.12, 1.71)1.22 (0.90, 1.66)1.51 (1.21, 1.88)1.32 (0.95, 1.83)
Medial meniscal extrusion1.91 (1.21, 3.01)1.48 (0.82, 2.68)2.07 (1.27, 3.38)1.42 (0.75, 2.70)
Varus malalignment1.06 (0.98, 1.15)1.01 (0.91, 1.11)1.09 (1.01, 1.17)1.02 (0.93, 1.12)
Lateral laxity1.11 (0.89, 1.37)1.09 (0.88, 1.36)1.15 (0.93, 1.43)1.14 (0.91, 1.42)
Table 7. Relationship between local mechanical factors and qualitative cartilage loss outcomes in the lateral tibiofemoral compartment*
 WORMS score of cartilage integrity, OR (95% CI)Modified WORMS score, OR (95% CI)
Adjusted for age, sex, and BMIAdjusted for all covariatesAdjusted for age, sex, and BMIAdjusted for all covariates
  • *

    ORs and 95% CIs for the baseline to 2-year qualitative cartilage loss outcome associated with each local factor were first adjusted for age, sex, and BMI, and were then adjusted for all covariates (age, sex, BMI, lateral meniscal damage, lateral meniscal extrusion, valgus malalignment, and medial laxity). The proportion of knees with progression was 30 of 251 (12%) using the WORMS and 23 of 251 (9%) using the modified WORMS. Data are from 251 knees in 153 participants. See Table 6 for definitions.

Lateral meniscal damage†1.10 (0.87, 1.38)0.93 (0.70, 1.24)1.22 (0.96, 1.54)1.09 (0.84, 1.43)
Lateral meniscal extrusion†1.07 (0.56, 2.03)0.93 (0.47, 1.83)1.28 (0.68, 2.41)1.03 (0.52, 2.06)
Valgus malalignment‡1.10 (1.00, 1.20)1.09 (0.98, 1.22)1.09 (0.98, 1.21)1.05 (0.95, 1.17)
Medial laxity‡1.62 (1.22, 2.14)1.59 (1.21, 2.08)1.49 (1.16, 1.93)1.45 (1.12, 1.88)

DISCUSSION

After considering other local factors, medial meniscal damage and varus malalignment were independently associated with baseline to 2-year quantitatively measured cartilage loss from each medial surface, both tibial and weight-bearing femoral. In the fully adjusted models, neither medial meniscal extrusion nor lateral laxity was associated with cartilage loss in either plate (although the relationship between meniscal extrusion and femoral cartilage loss measurements approached significance). Lateral meniscal damage predicted cartilage loss in each of the lateral surfaces after adjusting for the other local factors. In these fully adjusted models, lateral meniscal extrusion was linked to one outcome for the tibial surface, valgus malalignment was not significantly associated with cartilage loss in either surface, and medial laxity was linked to one outcome for the femoral surface.

In these analyses, which used the more quantitative cartilage assessment, factors bearing the strongest relationship to cartilage loss were medial meniscal damage and varus malalignment for the medial surfaces and lateral meniscal damage for the lateral surfaces. Using the qualitative cartilage assessment, no significant relationship with outcome was detected for these local factors (except medial laxity) in the fully adjusted models.

The findings for meniscal damage are not surprising, especially given that the menisci function to reduce contact stresses by enlarging the contact surface, distributing load, and increasing stability (28–31). Meniscectomy reduces the contact area, with a corresponding 2–3-fold increase in stress (32). Total meniscectomy causes knee OA changes; the risk of OA is high, even after partial meniscectomy (33).

In a study of patients with knee injury, 22% of cartilage lesions in the presence of a meniscal tear worsened, as compared with 14.9% of lesions in the absence of a meniscal tear (16). Berthiaume et al found that knees in a severe medial meniscal tear group had greater medial cartilage volume loss than did knees without a tear (3.0% versus 2.7%) (17). In analyses adjusting for medial meniscal extrusion, extrusion, but not tear, predicted loss. A trend toward more lateral cartilage loss in knees with lateral meniscal tear approached significance. In the Boston Osteoarthritis of the Knee Study, meniscal damage and extrusion independently predicted worsening in the cartilage morphology score (18). Our study uniquely considered malalignment and laxity; meniscal damage relationships persisted, but extrusion was not linked to cartilage loss in any surface after controlling for the other local factors.

A compelling biomechanical rationale also exists for the relationship between malalignment and cartilage loss. Malalignment (i.e., when the center of the knee does not lie close to the mechanical axis of the limb, which is represented by a line from the center of the hip to the center of the ankle) alters stress distribution in the knee (4, 5, 10). Varus malalignment is a major determinant of the adduction moment at the knee during gait (11, 34), which, in turn, is strongly related to medial load (35), a correlate of the medial to lateral subchondral bone density ratio, and a predictor of OA progression (36). We previously found that varus and valgus malalignment increased the likelihood of radiographic medial and lateral tibiofemoral OA progression, respectively (12). Cicuttini et al found an average annual loss of medial femoral cartilage of 17.7 μl (95% CI 6.5–28.8) for every 1 degree of increase in baseline varus angulation, with a trend toward a similar relationship with medial tibial cartilage volume loss (37). For every 1 degree of increase in valgus angle, there was an average loss of lateral tibial cartilage volume of 8.0 μl (95% CI 0.0–16.0). After adjusting for the other local factors, we found a consistent and independent effect of varus malalignment, but not valgus malalignment, as a continuous variable on cartilage loss. The valgus impact on the lateral joint surfaces may be weaker than the varus impact, which is consistent with previous findings suggesting that compartment load distribution is more equitable in valgus than in varus knees (38–40).

We did not find consistent evidence of a relationship between medial or lateral laxity and cartilage loss. The effect of frontal plane laxity on primary knee OA progression has not been previously reported. Instability leads to abrupt motion, with larger displacements, altered fit and contact regions of opposing joint surfaces, and an increase in regional shear and compressive forces (41). Several studies support a link between instability and posttraumatic OA development (42). The stable knee is a result of several interacting systems: articular and periarticular restraints (e.g., condylar geometry, tibial tubercle, iliotibial tract, cruciate and collateral ligaments, capsule, menisci, and muscle); contact forces generated by muscle activity and gravitational forces; mechanoreceptors providing proprioceptive input for reflex and centrally driven muscle activity, with feedforward and feedback neuromuscular control mechanisms; and visual, vestibular, and somatosensory subsystems (3, 43–45). In the OA setting, dynamic instability may originate, in theory, from any combination of impairments in these systems. Our results raise the possibility that static and non–weight-bearing frontal plane laxity does not capture key aspects of joint-protective dynamic stability.

MRI enhances the ability to assess local factor impact; the impact of meniscal damage and laxity, as described above, could not be examined using knee radiography. Several studies (25, 46–49) summarize evidence of the validity and long-term reliability of cartilage quantification in OA knees. Whether qualitative or quantitative cartilage loss outcomes are superior is an area of great interest; this is the first study to include measures of both. Previous studies examining local risk factors have often showed progression as a cartilage integrity score that worsens in any region within the tibiofemoral compartment of interest, relying on the original or modified WORMS scoring system. We included these outcomes to be able to compare our results with those published reports and to be able to compare a quantitative approach with the previously applied qualitative approach. The WORMS scoring system is invaluable in characterizing an array of tissue lesions in the OA knee joint organ comprehensively and reliably. For outcome assessment, however, the quantitative approach was more sensitive in revealing independent relationships in the fully adjusted models.

From these results, we cannot conclude that this quantitative approach is superior to any qualitative approach. It is possible that other qualitative approaches may be more sensitive. The skewed distribution of quantified cartilage loss precluded handling it as a continuous variable (Table 1). Cartilage loss in a large population may be better distributed, allowing for exploration of additional ways to handle longitudinal quantitative cartilage data. In addition to its ability to reveal relationships, advantages of the quantitative approach of the current study include interpretability, i.e., as an amount of change ≥2 times the measurement error, as opposed to change as a continuous measure, which may be difficult to interpret. The current study had a 2-year followup; findings may or may not be the same over a longer period of time.

Finally, it is important to note that local factors, such as those examined in this study, may participate in vicious circles with the worsening of knee OA. In a knee with OA and any of these impairments, it is often not possible to determine which came first, local impairment or knee OA. Whenever along the OA disease timeline a local impairment develops, it may contribute to subsequent OA progression and cartilage loss, especially given the vulnerable milieu of the already damaged OA knee. Local change has mechanical consequences, which may lead to other structural changes, in a vicious circle of progressive damage. This and amplification loops further increase the impact of local factors. Ultimately, strategies that interrupt these vicious circles may be especially powerful.

The absence of disease-modifying therapy increases the need to identify factors underlying progressive cartilage loss as potential targets for intervention. These factors may modify the effect of drugs that emerge in the future; targeting them may enhance drug response. Their presence also defines higher-risk subsets, useful at the individual level and at the public health level in the development of progression-prevention programs. These results support further work to define optimal interventions for meniscal tears in the setting of knee OA and the study of emerging interventions targeting the varus-malaligned OA knee.

In conclusion, using quantitative approaches to assess cartilage loss, local factors that independently predict tibial and femoral cartilage loss included medial meniscal damage and varus malalignment for the medial compartment and lateral meniscal damage for the lateral compartment. Quantitative cartilage loss outcome measures were more sensitive in revealing these relationships than a previously applied qualitative approach.

AUTHOR CONTRIBUTIONS

Dr. Sharma had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Sharma, Song, Cahue, Dunlop.

Acquisition of data. Sharma, Guermazi, Prasad, Kapoor, Cahue, Marshall.

Analysis and interpretation of data. Sharma, Eckstein, Song, Guermazi, Hudelmaier, Dunlop.

Manuscript preparation. Sharma, Eckstein, Song, Guermazi, Dunlop.

Statistical analysis. Song, Dunlop.

Ancillary