The influence of alignment on risk of knee osteoarthritis progression according to baseline stage of disease
Varus and valgus malalignment increase the risk of medial and lateral osteoarthritis (OA) progression, respectively. The impact of a mechanical factor such as alignment depends not only on the factor itself, but also on the state of the joint. The less-damaged joint of mild OA may be less vulnerable to malalignment effects than the more-damaged joint of moderate OA. Our goal was to explore the impact of alignment on subsequent progression of knee OA according to the baseline stage of disease.
Two hundred thirty patients with knee OA (defined by the presence of osteophytes and symptoms) recruited from the community underwent assessment of both lower limbs at baseline and at an 18-month followup. Alignment was measured on a full-limb radiograph as the angle made by the intersection of the femoral and tibial mechanical axes. Compartment-specific progression was defined as an increase between baseline and 18 months in the grade of severity of joint space narrowing on radiographs of semiflexed knees taken after fluoroscopic confirmation of position. Knees were grouped according to their baseline stage of OA as Kellgren/Lawrence (K/L) grades 0–1, 2, or 3. Progression odds ratios were estimated from logistic regression using generalized estimating equations.
There were 377 K/L grade 0–3 knees in 230 subjects (173 women and 57 men, mean age 64.0 years, mean body mass index 30.4 kg/m2) in this longitudinal study. In knees with mild OA (K/L grade 2), the odds of 18-month progression in the medial compartment were significantly increased 4-fold by varus alignment at baseline. In K/L grade 2 knees, the odds of lateral progression were increased 2-fold by valgus alignment (approaching significance). In knees with moderate OA (K/L grade 3), the risk of progression was comparably increased by varus or valgus alignment (10-fold).
While some effect of malalignment was suggested at almost all stages of knee OA examined, the impact of varus or valgus malalignment on the odds of OA progression over the ensuing 18 months was greater in knees with moderate (K/L grade 3) OA at baseline, possibly due to greater joint vulnerability with some contribution from slightly more severe malalignment.
Knee osteoarthritis (OA) is a leading cause of chronic disability (1). Few factors that contribute to the progression or advancement of OA disease at the knee have been identified. One factor that has been found to play an important role is knee alignment (i.e., hip–knee–ankle angle). Varus and valgus malalignment were shown to increase the risk of subsequent medial and lateral knee OA progression, respectively (2). The effect of malalignment potentially extends beyond its direct effect in at least two ways: malalignment is part of a vicious circle with progressive OA, and it may have effects on knee tissues other than cartilage that could further propagate OA disease.
The primary mechanism of the alignment effect is biomechanical. In the normal state, 60–80% of the total intrinsic compressive load transmitted across the knee is on the medial compartment (3). Alignment influences medial to lateral compartment load distribution. In a varus knee, the load-bearing axis is shifted so that the medial compartment experiences greater stress. In a valgus knee, stress is increased in the lateral compartment (4).
The impact of a mechanical factor depends not only on the factor itself, but also on the state of vulnerability of the knee to insult, including its OA disease stage. This state of vulnerability may change during the evolution of OA from mild to moderate stages. The less-damaged joint of mild OA may be less vulnerable to the biomechanical effects of malalignment than the more-damaged joint of moderate OA. Rarely have progression studies stratified examination of the effect of a given risk factor according to the baseline stage of OA.
The goal of this study was to examine whether the influence of alignment on subsequent progression of knee OA differs according to the baseline stage of disease. We examined the alignment/progression relationship in patients with knee OA separately in their knees without radiographic OA, with mild OA, and with moderate OA.
PATIENTS AND METHODS
The MAK (Mechanical Factors in Arthritis of the Knee) study is a natural history study of knee OA. The MAK cohort was recruited from the community through advertising in periodicals targeting elderly persons, 67 neighborhood organizations, letters to members of the registry of the Buehler Center on Aging at Northwestern University, and local medical center referrals.
All MAK participants were required to have a Kellgren/Lawrence (K/L) radiographic grade (5) of ≥2 in one or both knees and some difficulty with knee-requiring activity, defined as at least “a little” (Likert category) difficulty with 2 or more items on the Western Ontario and McMaster Universities Osteoarthritis Index physical function scale (6). Exclusion criteria were corticosteroid injection within the previous 3 months or history of avascular necrosis, rheumatoid or other inflammatory arthritis, periarticular fracture, Paget's disease, villonodular synovitis, joint infection, ochronosis, neuropathic arthropathy, acromegaly, hemochromatosis, Wilson's disease, osteochondromatosis, gout, pseudogout, or osteopetrosis.
Approval was obtained from the Office for the Protection of Research Subjects–Institutional Review Board of Northwestern University. Written informed consent was obtained from all participants.
Definition of subgroups.
All K/L grade 0–3 knees (i.e., without end-stage OA) were divided into 3 groups according to their K/L status at the baseline evaluation: grade 0–1 knees, grade 2 knees, and grade 3 knees. In the K/L system, 1 of 5 grades is assigned to each knee: 0 = normal; 1 = possible osteophytic lipping; 2 = definite osteophytes and possible narrowing of joint space; 3 = moderate multiple osteophytes, definite narrowing of joint space, some sclerosis, and possible deformity of bone contour; 4 = large osteophytes, marked narrowing of joint space, severe sclerosis, and definite deformity of bone contour.
Measurement of alignment.
To assess alignment, a single anteroposterior radiograph of the lower extremity was obtained at baseline in both limbs. A 130 × 36–cm graduated grid cassette was used to include the full limb of tall participants (7). Participants stood without footwear, with tibial tubercles facing forward. The tibial tubercle was used as a positioning landmark (8). The x-ray beam was centered on 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.
Alignment was measured as the angle formed by the intersection of the mechanical axes of the femur (the line from femoral head center to femoral intercondylar notch center) and the tibia (the line from the ankle talus center to the center of the tibial spine) (4, 8, 9). A knee was defined as varus when alignment was >0° in the varus direction, valgus when it was >0° in the valgus direction, and neutral when alignment was 0° (7, 9, 10). One experienced reader made all measurements. Reliability was high for measurements of varus (intraclass correlation coefficient [ICC] 0.99) and valgus (ICC 0.98) alignment.
At baseline and at 18 months, radiographs of both knees were obtained following the Buckland-Wright protocol (11). This protocol meets criteria for knee OA studies set by a multidisciplinary workshop (12) and the Task Force of the Osteoarthritis Research Society International (13). Knee position, criteria for beam alignment relative to knee center, radiopaque markers to account for magnification, and measurement landmarks were specified according to this protocol. All radiographs were obtained in the same unit by two trained technicians.
The standing semiflexed view of the knee in this protocol is optimal for joint space assessment because it achieves superimposition of the anterior and posterior joint margins (14, 15). The knee was flexed until the tibial plateau was horizontal, parallel to the beam, and perpendicular to the film. To control for rotation, the heel was fixed and the foot was rotated until the tibial spines were central within the femoral notch. Knee position was confirmed by fluoroscopy before films were taken. Foot maps made at baseline were used to standardize repositioning at 18 months. These protocol elements enhance the accuracy and precision of joint space assessment (14, 15).
Radiographic progression of disease.
Progression of joint space narrowing is a widely recommended primary outcome for knee OA progression studies (13, 16), and it provides a compartment-specific measure, which was required in the study. Medial or lateral progression was defined as an increase between baseline and 18 months in the grade of severity of joint space narrowing in the medial or lateral compartment, respectively. We used the 4-grade scale (0 = none, 1 = possible, 2 = definite, 3 = severe) with atlas representations from Altman et al (17).
Other approaches (i.e., osteophyte grade, K/L grade) to gauging progression had limitations in the current study. Although osteophytes can be graded per compartment, they are often more prominent in the uninvolved compartment. The K/L grade provides a global score without separate information for the medial and lateral compartments.
One experienced reader assessed radiographs using an atlas (17). Reliability for joint space grading (κ = 0.80–0.86) was very good. Reading of knee and full-limb radiographs was performed in separate sessions. The reader was blinded to knee data when assessing alignment and to alignment data when assessing knee radiographs.
Knees not at risk of progressing (i.e., those with the highest grade of joint space narrowing at baseline) were excluded. Logistic regression using generalized estimating equations to include one or both knees from an individual was applied to regress the probability of 18-month medial OA progression on varus alignment status in separate baseline K/L grade groups. These results are reported as odds ratios (ORs) with their associated 95% confidence intervals (95% CIs) for varus versus nonvarus alignment. Since age, sex, and body mass index (BMI) are recognized as potential confounders, these factors were screened using bivariate models that tested each factor as a predictor jointly with alignment for OA progression. Adjusted ORs for alignment were estimated by controlling significant factors retained from the bivariate screening. To test for varus alignment OR differences across different K/L grades, the logit of OA progression was regressed on alignment, binary variables representing the K/L grade groups, and K/L group by alignment interaction terms. Significant interaction coefficients would confirm differences due to K/L grade in the effect of alignment on progression. Similar analyses were used to estimate ORs for valgus alignment on lateral OA progression. For all analyses, an OR for which the 95% CI falls above 1 indicates that alignment is significantly associated with progression.
There were 237 MAK subjects with at least 1 knee that could progress (i.e., without the most severe grade of joint space narrowing at baseline). Seven of these 237 subjects (3%) did not return at 18 months (5 had died; 2 could not be reached) (2). The mean ± SD age of the 230 subjects constituting the study sample was 64.0 ± 10.8 years and their mean ± SD BMI was 30.4 ± 6.0 kg/m2. There were 173 women and 57 men. There were 45 K/L grade 0–1 knees (in 39 patients with a mean age of 64.0 years and a mean BMI of 28.0 kg/m2, 82% of whom were women), 241 K/L grade 2 knees (in 162 patients with a mean age of 63.4 years and a mean BMI of 29.6 kg/m2, 73% of whom were women), and 91 K/L grade 3 knees (in 75 patients with a mean age of 65.1 years and a mean BMI of 32.9 kg/m2, 70% of whom were women).
As shown in Table 1, there was a >2-fold increase in the odds of medial progression in the K/L grade 0–1 knees conferred by varus alignment, but this was not statistically significant. In the K/L grade 2 knees, the OR for medial progression associated with varus alignment rose to 4.12 and was significant. In the K/L grade 3 knees, the OR was even greater, at 10.96, and continued to be significant. The direction of change in the OR from K/L grade 0–1 to grade 2 and from grade 2 to grade 3 was consistent; the risk of medial progression related to varus alignment was greater in grade 3 knees (OR 10.96) compared with grade 0–1 knees (OR 2.50), although not significantly so (P = 0.11), which may have been due to low power related to small subgroup sizes. We recognized that a difference in the severity of malalignment between K/L grades might contribute to this difference in the odds of progression. Among varus knees, the mean ± SD severity of varus was 3.13 ± 1.25° (range 1.00–5.00) in K/L grade 0–1 knees, 3.04 ± 1.95° (range 1.00–8.00) in grade 2 knees, and 4.34 ± 2.66° (range 1.00–10.00) in grade 3 knees (P not significant for grade 0–1 versus grade 2 knees; P = 0.03 for grade 2 versus grade 3 knees).
Table 1. Risk of medial compartment progression conferred by varus alignment, stratified by baseline radiographic (K/L) grade of osteoarthritis*
|0–1||45 (39)||2.50 (0.67–9.39)|
|2||241 (162)||4.12 (1.92–8.82)|
|3||91 (75)||10.96 (3.10–38.77)|
It was not possible to examine the valgus effect in K/L grade 0–1 knees because of the small number of knees with lateral progression within this subgroup (i.e., 0 of 24 nonvalgus, K/L grade 0–1 knees and 1 of 24 valgus, K/L grade 0–1 knees progressed in the lateral compartment). As shown in Table 2, in K/L grade 2 knees, there was a >2-fold increase in the risk of lateral progression associated with valgus alignment, and this difference approached significance. In K/L grade 3 knees, there was a >10-fold increase in the odds of lateral progression, and this difference was significant. The risk of lateral progression related to valgus alignment in grade 3 knees (OR 10.44) was statistically significantly greater (P = 0.02) than the associated risk in grade 2 knees (OR 2.46). Among valgus knees, the mean ± SD severity of valgus was 3.13 ± 2.18° (range 1.00–10.00) in K/L grade 2 knees and 5.45 ± 3.07° (range 1.00–12.00) in grade 3 knees (P = 0.002 for grade 2 versus grade 3 knees).
Table 2. Risk of lateral compartment progression conferred by valgus alignment, stratified by baseline radiographic (K/L) grade of osteoarthritis*
|2||241 (162)||2.46 (0.95–6.34)|
|3||91 (75)||10.44 (2.76–39.49)|
Potential covariates considered included age, sex, and BMI. Of these factors, only BMI had a bivariate relationship to disease progression. After adjusting for BMI, the magnitude of the ORs was minimally altered.
In the present study, we found that in knees with mild OA (K/L grade 2), the odds of 18-month progression in the medial compartment were increased 4-fold by varus alignment at baseline. In K/L grade 2 knees, the odds of lateral progression were increased 2-fold by valgus alignment (approaching significance). In knees with moderate OA (K/L grade 3), the impact of malalignment in either the varus or valgus direction on subsequent 18-month progression was substantial (>10-fold increase).
Varus and valgus alignment are associated with increased force across the medial and lateral compartments, respectively. Joint pathology present at moderate stages of OA, including cartilage loss or meniscal degeneration, may render the joint less able to handle the forces imposed by a given alignment. In keeping with the concept that malalignment and OA progression are in a vicious circle, malalignment appeared somewhat more severe in K/L grade 3 knees than in K/L grade 2 knees (i.e., by 1.3° in varus knees and by 2.3° in valgus knees). The individual contributions of knee joint vulnerability and malalignment severity to the current results cannot be segregated; it is likely that both factors come into play.
In the K/L grade 2 knees, the effect of valgus alignment on lateral progression appeared weaker than the effect of varus alignment on medial progression (despite comparable absolute severity of malalignment). A biomechanical rationale exists for a smaller impact of valgus than varus in these mildly diseased knees. In the normally aligned, ambulating knee, load is disproportionately transmitted to the medial compartment (18). Varus further increases the total load passing medially during gait (9). Although valgus is associated with an increase in lateral compartment peak pressures, the medial compartment often continues to bear more load than the lateral compartment until more severe valgus is present (19, 20). Therefore, compartment load distribution in the mildly diseased valgus knee is likely to be more equitable than in the comparably diseased varus knee. A difference between the impact of varus and valgus alignment on subsequent progression was less apparent in the more vulnerable K/L grade 3 knees.
Numerous studies in the orthopedic literature demonstrate that knee alignment is associated with surgical outcome. Of particular relevance, tibial osteotomy, developed with the view that transferring load away from the stressed area of the knee would be beneficial, improves relevant outcomes in subsets of patients (21–24). That malalignment is a risk factor for the natural progression of knee OA has also been demonstrated. We previously found that the presence of varus alignment at baseline in all knees at risk of progression in subjects with knee OA increased the risk of subsequent medial progression (OR 4.09, 95% CI 2.20–7.62) (2). Valgus alignment increased the risk of lateral progression (OR 4.89, 95% CI 2.13–11.20). In a previous natural history study, a patient's recollection of having had “bow-legs or knock-knees in childhood” was associated with a 5-fold increase in the risk of OA progression (25). To our knowledge, no previous study has examined the malalignment effect according to baseline stage of disease severity.
Limitations of this study include the small number of knees in certain subsets, including the K/L grade 3 group (contributing to the wide 95% CI for the OR) and the K/L grade 0–1 group (precluding analysis of the valgus impact), which limits the power to detect statistically significant relationships. However, in spite of the size of certain subsets, with progressively worse baseline stage of disease severity, there was a clear trend toward an increase in the magnitude of the varus-associated OR and a significant increase in the valgus-associated OR.
These results have several implications. First, they add support to the concept that (among knees without end-stage OA) the influence of alignment is more pronounced as disease stage advances. This is possibly due to greater joint vulnerability, with some contribution from more severe malalignment. The alignment effect on a relatively healthy knee (the target of the epidemiologic investigation of the new development of OA) cannot be assumed to be the same as the alignment effect on a diseased knee (the target of progression studies), and needs to be separately examined. While the issue of vulnerability is directly relevant to mechanical factors, perhaps it should also be considered for risk factors in general. Second, the response to interventions for malalignment needs to be stratified according to baseline stage of disease. Interventions which reduce the stresses that malalignment places on a knee are in various phases of development and investigation (e.g., wedge insoles, “unloading” braces). The current results imply that such interventions, if they can improve compartment load distribution, may delay disease progression in varus K/L grade 2 knees and in varus or valgus K/L grade 3 knees—this warrants further study. The role of such interventions in K/L grade 0–1 knees (varus or valgus) or in valgus K/L grade 2 knees is less clear. It is important to note that the current study assessed progression during the ensuing 18 months; the effect of malalignment in knees without OA or with mild OA may take years to be fully evident. In contrast, the impact of malalignment on progression in the K/L grade 3 knees was evident within 18 months.
In summary, some effect of malalignment was suggested at almost all stages of OA examined. However, the impact of varus or valgus malalignment on the odds of OA disease progression during the ensuing 18 months was greater in knees with more advanced OA at baseline.