Tibial subchondral trabecular volumetric bone density in medial knee joint osteoarthritis using peripheral quantitative computed tomography technology.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 58, No. 9, September 2008, pp 2776–2785 DOI 10.1002/art.23795 © 2008, American College of Rheumatology Tibial Subchondral Trabecular Volumetric Bone Density in Medial Knee Joint Osteoarthritis Using Peripheral Quantitative Computed Tomography Technology Kim L. Bennell,1 Mark W. Creaby,1 Tim V. Wrigley,1 and David J. Hunter2 higher vBMD in the medial compartment compared with the lateral compartment, but these ratios were not influenced by disease status. Conclusion. Subregional vBMD changes were evident beneath the medial and lateral compartments of those with moderate medial knee OA. Of import, the posterior subchondral trabecular regions of the medial tibial plateau have markedly lower vBMD. Objective. Knee osteoarthritis (OA) is an organlevel failure of the joint involving pathologic changes in articular cartilage and bone. This cross-sectional study compared apparent volumetric bone mineral density (vBMD) of proximal tibial subchondral trabecular bone in people with and without knee OA, using peripheral quantitative computed tomography (pQCT). Methods. Seventy-five individuals with mild or moderate medial compartment knee OA and 41 asymptomatic controls were recruited. Peripheral QCT was used to measure vBMD of trabecular bone beneath medial and lateral tibiofemoral compartments at levels of 2% and 4% of tibial length, distal to the tibial plateau. Results. There was no significant difference in vBMD beneath the overall medial and lateral compartments between the 3 groups. However, in the affected medial compartment of those with moderate OA, lower vBMD was seen in the 2 posterior subregions compared with controls and those with mild knee OA, while higher vBMD was seen in the anteromedial subregion. Beneath the unaffected or lesser affected lateral compartment, significantly lower vBMD was seen at the 2% level in the anterior and lateral subregions of those with moderate disease. Volumetric BMD ratios showed relatively Knee osteoarthritis (OA), which most commonly affects the medial tibiofemoral compartment, is a prevalent chronic, localized joint disease (1) that involves pathologic changes in articular cartilage and subchondral bone (2). However, there is debate about the role and timing of bone and cartilage changes in disease initiation and progression. It also appears that the various subchondral bony layers beneath the articular cartilage (calcified cartilage, subchondral cortical plate, and subchondral trabecular bone) respond differently and play different roles in knee OA (2,3). Much of our understanding of bone changes in knee OA has come from studies using specimens obtained at autopsy, from animals, or from patients undergoing total knee joint arthroplasty, and thus having end-stage disease. There is limited in vivo research evaluating tibial subchondral bone changes, particularly in those with less severe disease and as the disease progresses. Our goal was to focus on the subchondral trabecular bone, since an improved understanding of the role of this tissue in knee OA may assist in developing more targeted treatment interventions. Trabecular bone has traditionally been investigated using a number of key measures that assess the quantity of bone in a fixed volume, the mineral density of that bone tissue itself, and the mass of bone mineral within the fixed volume; these are the bone volume fraction (bone volume over the total, fixed, sample Supported by an Australian Research Council Large Infrastructure Equipment grant (LEO453623), a grant from the William Buckland Foundation (ANZ Trustees), and a National Health and Medical Research Council project grant (350297). 1 Kim L. Bennell, BAppSci(Physio), PhD, Mark W. Creaby, BSc (Hons), PhD, Tim V. Wrigley, BSc (Hons), MSc: The University of Melbourne, Melbourne, Victoria, Australia; 2David J. Hunter, MBBS, PhD: New England Baptist Hospital, Boston, Massachusetts. Dr. Bennell has received speaking fees from the Hong Kong Association of Rehabilitation Medicine (less than $10,000). Address correspondence and reprint requests to Kim L. Bennell, PhD, Centre for Health, Exercise and Sports Medicine (CHESM), The University of Melbourne, 202 Berkeley Street, Melbourne, Victoria 3010, Australia. E-mail: firstname.lastname@example.org. Submitted for publication October 18, 2007; accepted in revised form May 23, 2008. 2776 VOLUMETRIC BONE DENSITY IN KNEE OA measurement volume [BV/TV]), the material density (bone mass over bone volume [BM/BV]), and apparent bone mineral density (BMD) (bone mass over fixed total volume [BM/TV]). The bone volume fraction also determines the porosity (1 ⫺ [BV/TV]), that is, the proportion of space not occupied by bone tissue. It is also clear from the above that the apparent BMD (BM/TV) is equal to the bone volume fraction (BV/TV) multiplied by the material density (BM/BV). A range of imaging technologies has been used to evaluate some or all of these parameters or related indices in subchondral tibial trabecular bone in knee OA. These technologies include high-resolution magnetic resonance imaging (MRI) (4), macroradiography (5) and micro–computed tomography (6), the latter of which is generally limited to in vitro applications (7). While unable to distinguish trabecular from cortical bone, dual energy x-ray absorptiometry (DXA) has also been used to assess the subchondral proximal tibia, where both cortical and trabecular bone are found. However, DXA does not assess any of the traditional parameters of trabecular bone. Rather, it derives the areal BMD (aBMD; gm/cm2) by dividing the bone mineral content by the projected area. As such, it is not sensitive to size differences in the third dimension (8) and may lead to erroneous conclusions (9). These imaging technologies have demonstrated a number of changes in trabecular bone associated with knee OA in vivo. MRI studies have correlated cartilage volume loss with increased bone volume fraction beneath the affected compartment, and decreased bone volume fraction beneath the unaffected compartment (4,10,11). Fractal signature analysis of macroradiographs has shown thinning of vertical trabeculae, in particular, to be characteristic of the predominantly less dense subchondral trabecular bone in OA (5,12,13). The few DXA studies, however, are contradictory, with reports of no difference (5,14), and lower (15) or higher (16,17) apparent aBMD in the OA groups compared with controls. Peripheral quantitative computed tomography (pQCT) is a broad class of x-ray–based tomographic scanners that can operate at pixel resolutions from ⬍100 m to ⬃500 m. However, the larger scanners that can accommodate a human limb tend to operate in the range from 200 m to 500 m. At this resolution, which is not high enough to identify individual trabeculae, pQCT measures the apparent volumetric density of the mineral fraction of each fixed-volume, 3-dimensional voxel (BM/ 2777 TV) (8). Thus, it cannot measure such parameters as the bone volume fraction (BV/TV). Peripheral QCT has several advantages over DXA, in particular for evaluating BMD of subchondral trabecular bone in knee OA. First, it provides 3-dimensional vBMD rather than aBMD and thus is not influenced by differences in bone size. Second, it allows for separate examination of trabecular and cortical bone. This is not merely an issue of delineation. Because cortical bone is much denser than trabecular bone (i.e., there is more of it), changes in trabecular bone are difficult to detect with DXA, in which both types of bone are present within the sampled region, i.e., cortical bone dominates. Third, because pQCT uses thin crosssectional slices, bone density profiles can be examined in subregions across a particular level. This is important for knee OA because loading varies across the tibial plateaus, and articular cartilage wear is greater in particular regions (18). Thus, it is likely that subchondral bone density will be heterogeneous across a given level beneath the plateau. However, to date, there have been no studies of in vivo proximal tibial subchondral bone in knee OA using the low ionizing technology (19) of pQCT. The specific aim of this study was to compare, using pQCT, the tibial subchondral trabecular vBMD beneath the medial and lateral joint compartments in patients with medial knee OA of varying severity and in subjects without OA. PATIENTS AND METHODS Study participants. One hundred sixteen participants were recruited, including 75 patients with medial compartment knee OA (39 women, 36 men) and 41 asymptomatic, healthy controls (24 women, 17 men). Diagnosis of knee OA was based on the American College of Rheumatology classification criteria (20). Participants were included if they were ⬎50 years of age and had knee pain on most days of the previous month (average level ⬎3 cm on a 10-cm visual analog scale). Other inclusion criteria were a predominance of pain/tenderness over the medial region of the knee and medial radiographic OA (defined as at least grade 1 medial joint space narrowing [more narrowing in the medial than in the lateral compartment] and grade 1 medial tibial or femoral osteophytes) (21). The exclusion criteria were as follows: questionable presence of or advanced radiographic knee OA (Kellgren/Lawrence [K/L] score of 1 or 4, respectively) (22), predominant patellofemoral joint symptoms based on clinical examination, knee surgery or intraarticular corticosteroid injection within the past 6 months, current or past (within 4 weeks) oral corticosteroid use, systemic arthritic conditions, a history of tibiofemoral/patellofemoral 2778 joint replacement or tibial osteotomy, or taking medication or having any condition that could affect bone density. Participants with either unilateral and bilateral disease were included, with the more painful knee studied. Controls were excluded if they reported a history of knee pain that interfered with function or caused them to seek treatment within the past year, had prior knee surgery, had any condition that affected mobility for ⬎1 week in the past 12 months, or had any condition or were taking any medications that could affect bone density. The research was approved by the University of Melbourne Human Research Ethics Committee and the Department of Human Services Radiation Safety Committee. All participants provided written informed consent. Tibial subchondral trabecular vBMD. A Stratec XCT 3000 pQCT scanner (Stratec, Pforzheim, Germany) was used to measure apparent vBMD (mg/cm3; not to be confused with bone mineral apparent density, a DXA-derived measure). Participants were seated with the extended leg placed in the scanner gantry (Figure 1a). A scout scan of the tibiofemoral joint was performed first, after which 1 medial and 1 lateral reference line were interactively placed on the scout image at half the depth of the region of highest radio-opacity near the surface of the tibia (indicating the highest cortical bone density), midway between the medial and lateral borders of the respective compartments (Figure 1b). Four transverse CT scans were then obtained at 2% and 4% of tibial length (defined as the distance from the medial joint line to the tip of the medial malleolus) distal to the reference lines. These levels were designed to sample subchondral trabecular bone deep to the subchondral cortical plate, consistent with our objective. The 2% level in our participants corresponded to ⬃6.4–8.6 mm from the reference line, while the 4% level corresponded to ⬃12.7–17.1 mm from the reference line. The XCT 3000 has a fixed slice thickness of 2.3 mm. The voxel size was set at 500 ⫻ 500 m, and the scan speed was 20 mm/second. Total scan time was ⬃15 minutes. The analysis was conducted using custom-written MatLab software (MathWorks, Natick, MA) because currently available commercial software is not suited to subregional analysis of subchondral trabecular bone. With our custom software (23), soft tissue was removed and the outer tibial cortex was identified. Ellipses were then automatically placed to define the regions of interest (ROIs) in the slices taken from beneath the medial and lateral compartments. The width of the ellipses was 0.35 of the maximum medial–lateral tibial outer cortical dimension, and the height was 0.40 of the maximum anteroposterior dimension at the medial–lateral centroid of each compartment. These dimensions ensured that cortical regions and osteophytes were not included. Six subregions of interest within each ellipse were then created (anteromedial, anterolateral, middle medial, middle lateral, posteromedial, and posterolateral) at the 2% and 4% levels (Figure 1c). The variable of interest was the mean trabecular apparent vBMD (mg/cm3), in the whole medial and lateral ellipses and in each of the subregions. Test–retest reliability was established in 16 subjects with and without knee OA, who were measured twice with repositioning. Intraclass correlation coefficients (ICC 2.1) BENNELL ET AL Figure 1. a, Patient positioning for measurement of tibial subchondral trabecular volumetric bone mineral density. b, Scout view, showing medial and lateral reference lines and measurement positions at 2% and 4% tibial length distal to each reference line. c, Transverse peripheral quantitative computed tomography scan of the subchondral tibia, showing the medial and lateral ellipses and demarcation of lateral and medial subregions of interest. ranged from 0.86 to 0.99 for the 2% level ROIs and from 0.83 to 0.99 for the 4% ROIs. Disease severity and knee alignment. In patients with knee OA, standardized semiflexed posteroanterior knee radiographs were obtained with the participant standing barefoot. Radiographic severity of tibiofemoral OA was assessed using the K/L scale, where 0 ⫽ normal, 1 ⫽ possible osteophytes, 2 ⫽ minimal osteophytes and possible joint space narrowing, 3 ⫽ moderate osteophytes, some narrowing, and possible sclerosis, and 4 ⫽ large osteophytes, definite joint space narrowing, and VOLUMETRIC BONE DENSITY IN KNEE OA 2779 Table 1. Characteristics of the control and OA groups* Age, years Female, no. (%) Height, meters Body mass, kg BMI, kg/m2 Knee alignment, degrees§ Controls (n ⫽ 41) Mild OA (n ⫽ 40) Moderate OA (n ⫽ 35) 60.5 ⫾ 8.5 24 (58.5) 1.67 ⫾ 0.1 74.6 ⫾ 14.0 26.5 ⫾ 3.5 – 62.8 ⫾ 7.6 23 (57.5) 1.67 ⫾ 0.1 80.6 ⫾ 15.1 28.8 ⫾ 4.4‡ 179.6 ⫾ 1.3 67.4 ⫾ 7.9† 19 (54.3) 1.70 ⫾ 0.1 91.0 ⫾ 12.6† 31.9 ⫾ 4.6† 177.7 ⫾ 1.9¶ * Except where indicated otherwise, values are the mean ⫾ SD. OA ⫽ osteoarthritis; BMI ⫽ body mass index. † P ⱕ 0.05 versus controls and versus mild OA. ‡ P ⱕ 0.05 versus controls. § Lower numbers indicate greater varus. ¶ P ⱕ 0.05 versus mild OA. severe sclerosis (21,22). The radiographs were assessed by an experienced musculoskeletal researcher and physiotherapist (KLB) whose intrarater reliability was 0.87 (linearly incremental weighted kappa statistic). Anatomic knee alignment was measured using the posteroanterior radiographs. Alignment measured in this manner is strongly correlated with the mechanical axis obtained from a long leg radiograph (ICC 0.86; data from 40 patients in our laboratory) (24) and avoids the additional cost and radiation associated with a long leg radiograph. A prediction equation was used for conversion (24). The intrarater and interrater reliability of this technique was excellent (ICC ⫽ 0.98 and 0.97, respectively). Statistical analysis. Results are expressed as the adjusted mean and SE. Data were analyzed using SPSS, version 13.0 (SPSS, Chicago, IL). Checks were made for normality and violation of statistical assumptions. Patients with OA were divided into 2 groups based on the K/L score; those with a K/L score of 2 were categorized as having mild disease, while those with a K/L score of 3 were categorized as having moderate disease. Differences between the mild OA, moderate OA, and control groups in each ROI at 2% and 4% levels were determined using generalized linear models (GLMs). Group was included as a fixed factor, while age, body mass, and sex were included as covariates. The adjusted models were followed by pairwise contrasts to locate the source of any significant differences. Differences between groups in the ratio of vBMD beneath the medial and lateral compartments were also calculated, using the same GLM approach as described above. P values less than 0.05 were considered significant. RESULTS Characteristics of the OA and control groups are shown in Table 1. The groups were similar in sex representation and were comparable in height. However, the moderate OA group was significantly older, heavier, and had a higher body mass index (BMI) compared with the control and mild OA groups. The mild OA group had a significantly higher BMI than the control group. The moderate OA group had greater varus malalignment than the mild OA group. Table 2. Trabecular volumetric bone mineral density (mg/cm3) overall and in each subregion of interest in the affected medial compartment Controls (n ⫽ 41) 2% level Whole ellipse Posteromedial Posterolateral Middle medial Middle lateral Anteromedial Anterolateral 4% level Whole ellipse Posteromedial Posterolateral Middle medial Middle lateral Anteromedial Anterolateral Mild OA (n ⫽ 40) Moderate OA (n ⫽ 35) Adjusted mean SE Adjusted mean SE Adjusted mean SE 205.419 221.120 150.362 254.990 151.671 244.256 213.647 4.990 7.512 5.490 8.134 4.125 7.684 6.702 205.655 217.544 149.790 258.902 151.372 249.730 208.267 4.656 7.009 5.123 7.589 3.849 7.170 6.254 198.931 190.428* 131.647* 257.093 143.295 271.351† 199.758 5.601 8.430 6.162 9.129 4.629 8.624 7.522 193.459 240.491 166.103 251.043 122.112 224.164 164.454 5.213 8.743 5.995 7.584 4.950 6.919 5.095 189.261 225.439 155.215 240.351 123.949 227.755 169.880 4.864 8.158 5.593 7.077 4.619 6.456 4.754 182.779 200.588† 149.897 229.082 114.243 241.262 170.901 5.851 9.812 6.728 8.512 5.556 7.765 5.718 * P ⱕ 0.05 versus controls and versus mild osteoarthritis (OA) (generalized linear model with age, sex, and body mass as covariates). † P ⱕ 0.05 versus controls (generalized linear model with age, sex, and body mass as covariates). 2780 BENNELL ET AL Figure 2. Percentage difference in subchondral trabecular volumetric bone mineral density beneath the medial (a) and lateral (b) compartments at 2% distal to the tibial plateau reference line, in the osteoarthritis (OA) patient groups compared with the control group. The color bar provides a visual representation of increasingly greater differences in bone density in the mild and moderate OA groups compared with controls. ⴱ indicates P ⱕ 0.05 versus controls; † indicates P ⱕ 0.05 versus mild OA. Beneath the medial compartment, the trabecular vBMD of the whole ellipse was similar in the OA and control groups at both the 2% and 4% levels. However, subregional differences in vBMD existed between groups (Table 2). At the 2% level, the moderate OA group had a significantly lower vBMD than both the controls and the mild OA group in the posteromedial and posterolateral subregions. These differences ranged from 12.1% lower to 13.9% lower than in the control and mild OA groups, respectively. However, the moderate OA group had a significantly higher vBMD than the controls in the anteromedial subregion, with a similar trend also observed when compared with the mild OA group (P ⫽ 0.06). The percentage differences in vBMD in the OA groups compared with controls are shown in Figure 2. At the 4% level, the deficit in posteromedial vBMD was seen in the moderate OA group compared with controls, with bone density being 16.6% lower (P ⫽ 0.006). No other significant subregional differences in bone density were noted between groups at the 4% level. Beneath the unaffected or less affected lateral compartment (Table 3), trabecular vBMD of the whole ellipse at the 2% level was 10% lower in the moderate OA group compared with controls, which almost reached significance (P ⫽ 0.057). Subregional vBMD differences existed between the groups at the 2% level. Volumetric BMD was significantly less in the moderate OA group compared with controls in the posteromedial (14% less), posterolateral (13.1% less), and middle lateral (11.4% less) subregions. It was also significantly lower when compared with the mild OA group in the posterolateral subregion (9.4% less). Overall and subregional vBMD at the 4% level was similar between groups. The medial to lateral vBMD ratios from the whole ellipse at both the 2% and 4% levels were ⬎1, indicating a relatively higher vBMD beneath the medial VOLUMETRIC BONE DENSITY IN KNEE OA 2781 Table 3. Trabecular volumetric bone mineral density (mg/cm3) overall and in each subregion of interest in the unaffected or less affected lateral compartment Controls (n ⫽ 36) 2% level Whole ellipse Posteromedial Posterolateral Middle medial Middle lateral Anteromedial Anterolateral 4% level Whole ellipse Posteromedial Posterolateral Middle medial Middle lateral Anteromedial Anterolateral Mild OA (n ⫽ 37) Moderate OA (n ⫽ 29) Adjusted mean SE Adjusted mean SE Adjusted mean SE 187.032 219.545 241.276 158.126 199.318 143.585 167.436 5.515 8.014 7.227 5.435 6.019 5.777 6.394 179.449 206.942 231.357 155.442 188.579 137.245 164.290 4.959 7.207 6.499 4.887 5.413 5.195 5.749 168.924 188.885* 209.563† 149.916 176.568* 138.718 155.291 6.524 9.482 8.550 6.430 7.121 6.835 7.564 160.034 222.370 212.501 108.718 171.841 123.927 137.964 5.239 7.123 7.358 5.972 6.470 5.256 7.026 155.218 207.536 206.771 108.094 164.417 118.566 143.106 4.711 6.405 6.616 5.371 5.818 4.726 6.318 151.840 200.833 191.246 112.947 161.266 119.805 139.069 6.198 8.427 8.705 7.066 7.655 6.218 8.312 * P ⱕ 0.05 versus controls (generalized linear model with age, sex, and body mass as covariates). † P ⱕ 0.05 versus controls and versus mild osteoarthritis (OA) (generalized linear model with age, sex, and body mass as covariates). compartment compared with the lateral compartment in all 3 groups (Figure 3). The ratios ranged from 1.1 to 1.2, indicating 10–20% greater vBMD beneath the medial compartment. Consistent with the findings that vBMD in the whole medial and lateral ellipse was similar between groups, there was no difference in ratios between groups (P ⬎ 0.05). Figure 3. Ratio of overall medial to lateral trabecular volumetric bone mineral density in the osteoarthritis (OA) and control groups, shown as the adjusted mean and SE at 2% and 4% distal to the tibial plateau reference line. DISCUSSION This study is the first to assess apparent vBMD of tibial subchondral trabecular bone in patients with medial knee OA using pQCT, and the first to assess in vivo regional variations in vBMD. In general, it showed no differences in apparent vBMD beneath the whole medial and lateral compartment, or in their ratio, when comparing medial knee OA and control groups. However, an examination of subregional vBMD beneath each compartment revealed significant group differences. Beneath the affected medial compartment of those with moderate OA disease, subregions of both lower and higher vBMD were found when compared with controls and those with mild knee OA. Beneath the less affected or unaffected lateral compartment, subregions of significantly lower vBMD were seen in patients with moderate OA disease. A greater number of vBMD differences between groups were noted at the 2% rather than the 4% tibial level, which makes intuitive sense, since its closer proximity to the diseased joint surface makes it more likely to be implicated in the OA process. Apparent trabecular vBMD beneath the medial compartment as a whole did not differ between groups. This finding is consistent with some DXA studies (14), but not others, in which both increased and decreased 2782 aBMD have been reported in those with knee OA (15–17). However, it is difficult to make direct comparisons given the differences between DXA and pQCT. In particular, the differences may relate to the inability of DXA to distinguish between cortical and trabecular bone and to its measurement of aBMD. Since the latter measure may be associated with bone size–related artifact (25), size differences between individuals and changes in bone size with OA disease may be important (26) and may subsequently influence DXA results. Given these issues, we believe that our results using pQCT provide a more accurate reflection of proximal tibial trabecular BMD in patients with knee OA. Our results indicate that assessment of apparent vBMD in the compartment as a whole can mask subregional variations that exist. In those with moderate knee OA, we found significantly lower vBMD in the posteromedial and posterolateral subregions of the affected medial compartment compared with controls and those with mild OA. In fact, two-thirds of the subregions exhibited a similar pattern. However, in the anteromedial subregion, we found significantly higher vBMD in the group with moderate knee OA. These variations could reflect regional differences in mechanical load and cartilage degeneration. Overall, the medial knee joint compartment experiences greater load than the lateral compartment (27). This load differential is exacerbated by more severe disease (28) and joint malalignment (17,29). However, within the medial compartment, the actual distribution of loading and cartilage wear patterns across the tibial plateau are complex, heterogeneous, and not consistently established. Loading and cartilage wear within the medial compartment appear to vary depending on factors such as the integrity of the anterior cruciate ligament (18,30,31), meniscal status (32,33), and joint malalignment (18,31). For example, one study found the greatest cartilage wear in the anterior regions in medial knee OA (34), whereas others have shown greater involvement in the posterior regions (18,31). Subregional vBMD differences between groups might reflect these loading variations given that bone responds to the loads that are placed upon it. We must be somewhat cautious in these interpretations, since it is also conceivable that some differences may actually originate in the subchondral trabecular bone prior to development of OA. One difficulty in understanding the trabecular bone changes in OA is reconciling results derived from different measures. An increase in trabecular bone volume fraction (BV/TV) is a characteristic finding in BENNELL ET AL OA. This has been shown in the late stages of OA (35), as well as early in its course (36). However, the material density of this tissue typically indicates relative undermineralization. Thus, at the tissue level, OA is actually associated with reduced material bone density (2). Remembering that the apparent BMD (BM/TV) is equal to the bone volume fraction (BV/TV) multiplied by the material density (BM/BV), changes in either of the latter 2 quantities will change the vBMD. The exact effect depends on the relative magnitude and direction of the changes in bone volume and material density, although only to the extent that these change bone mass (given that total volume is fixed in vBMD). Studies that have shown an increase in subchondral trabecular bone (e.g., BV/TV) have also generally investigated the region very close to the subchondral cortical plate (36–38). A reduction in bone may be more characteristic of trabecular bone at a greater distance from the cortical plate (5,38). In this study, the localized lower vBMD beneath the affected medial compartment was marked (up to 17% lower), particularly in the posterior subregions. The interpretation of this result is that it is due to reduced mineral mass. This is most likely related to a decrease in bone tissue volume (e.g., thinning of trabeculae); however, if remodeling is occurring, reduced mineralization of the new bone could also partly explain a loss of overall mineral mass. Remodeling is likely to be driven by changes in loading, which would result in remodeling due to either direct bone effects or the production of microdamage (39). If the posterior subregions are directly exposed to lower load, then the subchondral trabecular bone would respond by remodeling with greater bone resorption, thus reducing its bone mass and hence its apparent vBMD. In contrast, although seemingly counterintuitive, there are possible mechanisms by which increased joint loading could also lead to localized lower subchondral trabecular vBMD. One is related to the thickening of the interposed subchondral cortical plate that occurs in OA, and has been shown in animal studies to occur prior to cartilage deterioration (6,40). This new, thickened bone may be poorly mineralized (41) due to the increased bone remodeling. This may, in turn, reduce the stiffness of the cortical plate (42–44). However, the timing of these cortical plate changes is complex and not particularly well understood (3,38). However, it has been suggested that the thickened but less stiff cortical bone would absorb more of the load through the knee joint, thus potentially shielding the underlying subchondral trabecular bone (2,3,5,13). In response to this reduced VOLUMETRIC BONE DENSITY IN KNEE OA loading, the trabecular bone may undergo increased remodeling with subsequent loss of mass and therefore apparent vBMD. This effect has been likened to “stress shielding” of bone around a metal implant. Another possibility is that high loading could lead to microdamage accumulation (39) and a transient reduction in apparent vBMD. However, given that our results were seen in those with more advanced disease, this is a less likely explanation. Regardless of the mechanism, the timing of these reductions in bone density is also not well understood (3,6,13,38). While vBMD differences with tibiofemoral OA have not previously been reported in vivo, our finding of lower vBMD has recently been shown by pQCT in patellae of human cadavers (45). In 14 patellae, those with advanced or moderate cartilage degeneration had reductions in vBMD of ⬃25% and ⬃10%, respectively, compared with those patellae with no or minimal degeneration. Also consistent with our results, loss of trabeculae (5,13) and reductions in aBMD (15) have been found in human in vivo studies using other technologies. The area of higher vBMD beneath the anteromedial subregion in the group with moderate OA could reflect an area subjected to increased loading. Since this increased density was isolated to 1 region confined to the 2% level, it is possible that some regions of subchondral trabecular bone closer to the cortical plate may experience increased density similar to that of the cortical plate (5,37). Regional variations in the thickness and depth of the cortical plate (46) might have contributed to the finding being limited to this subregion. Also, as the disease progresses, changes that may initially occur more proximally, closer to the cortical plate, may extend more distally (37). This may occur at different rates in different regions. The findings of altered subregional vBMD were only significant in those with moderate knee OA. While similar trends were seen in the group with mild OA, the differences were of a lesser magnitude. This suggests that vBMD changes as detected by pQCT become more apparent as the disease progresses. This is consistent with findings using other techniques, such as macroradiography (13). Further investigation is required to elucidate the pattern of vBMD change in humans through longitudinal studies. Beneath the less affected or unaffected lateral compartment, the adjusted mean overall vBMD was 10% lower in the group with moderate OA, with the result just failing to reach significance (P ⫽ 0.057). This group also demonstrated substantially lower vBMD at 2783 the 2% level in half the subregions. It has been surmised that lower bone density beneath the lateral compartment may simply represent disuse osteopenia (4,47). This could be due to the fact that more of the load is shifted to the medial compartment due to increasing varus malalignment and joint changes as the disease progresses (17,28,29). It is also possible that reduced physical activity levels as a person’s disease severity worsens (48) could lower the overall amount of loading taken through the knee. The medial:lateral vBMD ratios of all 3 groups at both the 2% and the 4% levels were ⬎1, reflecting relatively higher vBMD beneath the medial tibial compartment compared with the lateral compartment. This is consistent with some DXA studies of aBMD ratios (17,49–51). However, it is important to note that medial: lateral ratios ⬎1 are unable to distinguish between increased density on the medial side and decreased density on the lateral side. Since both phenomena seem to be apparent in OA, precise interpretation of this measure is problematic. Despite the findings of regional variations in the extent of the relative increased or decreased trabecular vBMD among OA participants and controls, our pQCT-based medial:lateral vBMD ratios, which are not subject to the possible artifacts of DXA, did not show any differences between controls and those with mild or moderate OA. Thus, while several DXA studies have found that those with more severe disease and greater joint space narrowing have higher aBMD ratios (17,50), we do not believe that interpretation of such findings is clearcut in relation to changes beneath either the medial or the lateral compartment. We would therefore advise that medial:lateral density ratios not be used as primary outcome measures. Our results indicate that the relative proportion of trabecular vBMD beneath the compartments as a whole at the 2 levels we measured is not altered in those with mild to moderate OA. In summary, subregional vBMD changes in the proximal tibia are apparent in those with moderate medial knee OA compared with those with mild OA and in healthy controls. Of import, the posterior subchondral trabecular regions beneath the affected medial tibial plateau, as well as a number of regions beneath the unaffected lateral compartment, appear to have markedly lower vBMD. In general, our findings using pQCT to measure vBMD are consistent with results of some in vitro studies and some in vivo studies, for example, those using macroradiography. It is apparent that further studies are needed. In particular, more focused attention on subregional vBMD as it relates to regional alterations 2784 BENNELL ET AL in cartilage and mechanical loading, as well as measurement of longitudinal changes in vBMD as the disease progresses, would provide important data. 12. ACKNOWLEDGMENTS We wish to thank Associate Professor Mark R. Forwood for his valuable feedback on an earlier revision of the manuscript, Sue Kantor and Ben Metcalf for assisting with pQCT measurements, and Kelly-Ann Bowles and Georgina Morrow for assisting with participant recruitment. AUTHOR CONTRIBUTIONS Dr. Bennell 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. Bennell, Wrigley, Hunter. Acquisition of data. Creaby. Analysis and interpretation of data. Bennell, Creaby, Wrigley, Hunter. Manuscript preparation. Bennell, Creaby, Wrigley, Hunter. Statistical analysis. Creaby. Acquisition of funding. Bennell, Creaby, Wrigley. 13. 14. 15. 16. 17. 18. REFERENCES 1. Felson DT, Zhang Y, Hannan MT, Naimark A, Weissman BN, Aliabadi P, et al. The incidence and natural history of knee osteoarthritis in the elderly: the Framingham Osteoarthritis Study. Arthritis Rheum 1995;38:1500–5. 2. Burr DB. The importance of subchondral bone in the progression of osteoarthritis [review]. J Rheumatol Suppl 2004;70:77–80. 3. Burr DB. Anatomy and physiology of the mineralized tissues: role in the pathogenesis of osteoarthrosis [review]. Osteoarthritis Cartilage 2004;12 Suppl A:S20–30. 4. Lindsey CT, Narasimhan A, Adolfo JM, Jin H, Steinbach LS, Link T, et al. Magnetic resonance evaluation of the interrelationship between articular cartilage and trabecular bone of the osteoarthritic knee. Osteoarthritis Cartilage 2004;12:86–96. 5. Messent EA, Buckland-Wright JC, Blake GM. Fractal analysis of trabecular bone in knee osteoarthritis (OA) is a more sensitive marker of disease status than bone mineral density (BMD). Calcif Tissue Int 2005;76:419–25. 6. Muraoka T, Hagino H, Okano T, Enokida M, Teshima R. Role of subchondral bone in osteoarthritis development: a comparative study of two strains of guinea pigs with and without spontaneously occurring osteoarthritis. Arthritis Rheum 2007;56:3366–74. 7. Boutroy S, Bouxsein ML, Munoz F, Delmas PD. In vivo assessment of trabecular bone microarchitecture by high-resolution peripheral quantitative computed tomography. J Clin Endocrinol Metab 2005;90:6508–15. 8. Ferretti JL, Cointry GR, Capozza RF. Noninvasive analysis of bone mass, structure, and strength. In: An YH, editor. Orthopaedic issues in osteoporosis. Boca Raton (FL): CRC Press; 2002. p. 145–67. 9. Seeman E. From density to structure: growing up and growing old on the surfaces of bone [review]. J Bone Miner Res 1997;12:509–21. 10. Beuf O, Ghosh S, Newitt DC, Link TM, Steinbach L, Ries M, et al. Magnetic resonance imaging of normal and osteoarthritic trabecular bone structure in the human knee. Arthritis Rheum 2002;46: 385–93. 11. Blumenkrantz G, Lindsey CT, Dunn TC, Jin H, Ries MD, Link 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. TM, et al. A pilot, two-year longitudinal study of the interrelationship between trabecular bone and articular cartilage in the osteoarthritic knee. Osteoarthritis Cartilage 2004;12:997–1005. Messent EA, Ward RJ, Tonkin CJ, Buckland-Wright C. Tibial cancellous bone changes in patients with knee osteoarthritis: a short-term longitudinal study using Fractal Signature Analysis. Osteoarthritis Cartilage 2005;13:463–70. Messent EA, Ward RJ, Tonkin CJ, Buckland-Wright C. Cancellous bone differences between knees with early, definite and advanced joint space loss: a comparative quantitative macroradiographic study. Osteoarthritis Cartilage 2005;13:39–47. Clarke S, Wakeley C, Duddy J, Sharif M, Watt I, Ellingham K, et al. Dual-energy X-ray absorptiometry applied to the assessment of tibial subchondral bone mineral density in osteoarthritis of the knee. Skeletal Radiol 2004;33:588–95. Karvonen RL, Miller PR, Nelson DA, Granda JL, FernandezMadrid F. Periarticular osteoporosis in osteoarthritis of the knee. J Rheumatol 1998;25:2187–94. Bruyere O, Dardenne C, Lejeune E, Zegels B, Pahaut A, Richy F, et al. Subchondral tibial bone mineral density predicts future joint space narrowing at the medial femoro-tibial compartment in patients with knee osteoarthritis. Bone 2003;32:541–5. Wada M, Maezawa Y, Baba H, Shimada S, Sasaki S, Nose Y. Relationships among bone mineral densities, static alignment and dynamic load in patients with medial compartment knee osteoarthritis. Rheumatology (Oxford) 2001;40:499–505. Moschella D, Blasi A, Leardini A, Ensini A, Catani F. Wear patterns on tibial plateau from varus osteoarthritic knees. Clin Biomech (Bristol, Avon) 2006;21:152–8. Braun MJ, Meta MD, Schneider P, Reiners C. Clinical evaluation of a high-resolution new peripheral quantitative computerized tomography (pQCT) scanner for the bone densitometry at the lower limbs. Phys Med Biol 1998;43:2279–94. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–49. Altman RD, Hochberg M, Murphy WA Jr, Wolfe F, Lequesne M. Atlas of individual radiographic features in osteoarthritis. Osteoarthritis Cartilage 1995;3 Suppl A:3–70. Kellgren JH, Lawrence JS. Radiological assessment of osteoarthritis. Ann Rheum Dis 1957;16:494–502. Wrigley TV, Creaby MW, Bennell KL. Peripheral quantitative computed tomography (pQCT) of tibial subchondral bone in knee osteoarthritis [abstract]. J Biomech 2007;40 Suppl 2:S591. Hinman RS, May RL, Crossley KM. Is there an alternative to the full-leg radiograph for determining knee joint alignment in osteoarthritis? Arthritis Rheum 2006;55:306–13. Carter DR, Bouxsein ML, Marcus R. New approaches for interpreting projected bone densitometry data. J Bone Miner Res 1992;7:137–45. Wluka AE, Wang Y, Davis SR, Cicuttini FM. Tibial plateau size is related to grade of joint space narrowing and osteophytes in healthy women and in women with osteoarthritis. Ann Rheum Dis 2005;64:1033–7. Schipplein OD, Andriacchi TP. Interaction between active and passive knee stabilizers during level walking. J Orthop Res 1991; 9:113–9. Sharma L, Hurwitz DE, Thonar EJ, Sum JA, Lenz ME, Dunlop DD, et al. Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis. Arthritis Rheum 1998;41:1233–40. Hurwitz DE, Ryals AB, Case JP, Block JA, Andriacchi TP. The knee adduction moment during gait in subjects with knee osteoarthritis is more closely correlated with static alignment than VOLUMETRIC BONE DENSITY IN KNEE OA 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. radiographic disease severity, toe out angle and pain. J Orthop Res 2002;20:101–7. Andriacchi TP, Briant PL, Bevill SL, Koo S. Rotational changes at the knee after ACL injury cause cartilage thinning. Clin Orthop Relat Res 2006;442:39–44. Harman MK, Markovich GD, Banks SA, Hodge WA. Wear patterns on tibial plateaus from varus and valgus osteoarthritic knees. Clin Orthop Relat Res 1998;352:149–58. Hunter DJ, Zhang YQ, Niu JB, Tu X, Amin S, Clancy M, et al. The association of meniscal pathologic changes with cartilage loss in symptomatic knee osteoarthritis. Arthritis Rheum 2006;54: 795–801. Kurosawa H, Fukubayashi T, Nakajima H. Load-bearing mode of the knee joint: physical behavior of the knee joint with or without menisci. Clin Orthop Relat Res 1980;149:283–90. Weidow J, Pak J, Karrholm J. Different patterns of cartilage wear in medial and lateral gonarthrosis. Acta Orthop Scand 2002;73: 326–9. Fazzalari NL, Parkinson IH. Fractal properties of subchondral cancellous bone in severe osteoarthritis of the hip. J Bone Miner Res 1997;12:632–40. Ding M, Odgaard A, Hvid I. Changes in the three-dimensional microstructure of human tibial cancellous bone in early osteoarthritis. J Bone Joint Surg Br 2003;85:906–12. Bobinac D, Spanjol J, Zoricic S, Maric I. Changes in articular cartilage and subchondral bone histomorphometry in osteoarthritic knee joints in humans. Bone 2003;32:284–90. Burr DB, Schaffler MB. The involvement of subchondral mineralized tissues in osteoarthrosis: quantitative microscopic evidence [review]. Microsc Res Tech 1997;37:343–57. Burr DB, Radin EL. Microfractures and microcracks in subchondral bone: are they relevant to osteoarthrosis? [review]. Rheum Dis Clin North Am 2003;29:675–85. Carlson CS, Loeser RF, Purser CB, Gardin JF, Jerome CP. Osteoarthritis in cynomolgus macaques. III. Effects of age, gender, and subchondral bone thickness on the severity of disease. J Bone Miner Res 1996;11:1209–17. 2785 41. Grynpas MD, Alpert B, Katz I, Lieberman I, Pritzker KP. Subchondral bone in osteoarthritis. Calcif Tissue Int 1991;49:20–6. 42. Lereim P, Goldie IF. Relationship between morphologic features and hardness of the subchondral bone of the medial tibial condyle in the normal state and in osteoarthritis and rheumatoid arthritis. Arch Orthop Unfallchir 1975;81:1–11. 43. Li B, Aspden RM. Mechanical and material properties of the subchondral bone plate from the femoral head of patients with osteoarthritis or osteoporosis. Ann Rheum Dis 1997;56: 247–54. 44. Lereim P, Goldie I, Dahlberg E. Hardness of the subchondral bone of the tibial condyles in the normal state and in osteoarthritis and rheumatoid arthritis. Acta Orthop Scand 1974;45:614–27. 45. Lammentausta E, Kiviranta P, Toyras J, Hyttinen MM, Kiviranta I, Nieminen MT, et al. Quantitative MRI of parallel changes of articular cartilage and underlying trabecular bone in degeneration. Osteoarthritis Cartilage 2007;15:1149–57. 46. Milz S, Putz R. Quantitative morphology of the subchondral plate of the tibial plateau. J Anat 1994;185:103–10. 47. Zysset PK, Sonny M, Hayes WC. Morphology-mechanical property relations in trabecular bone of the osteoarthritic proximal tibia. J Arthroplasty 1994;9:203–16. 48. De Groot IB, Bussmann JB, Stam HJ, Verhaar JA. Actual everyday physical activity in patients with end-stage hip or knee osteoarthritis compared with healthy controls. Osteoarthritis Cartilage 2008;16:436–42. 49. Hurwitz DE, Sumner DR, Andriacchi TP, Sugar DA. Dynamic knee loads during gait predict proximal tibial bone distribution. J Biomech 1998;31:423–30. 50. Lo GH, Zhang Y, McLennan C, Niu J, Kiel DP, McLean RR, et al. The ratio of medial to lateral tibial plateau bone mineral density and compartment-specific tibiofemoral osteoarthritis. Osteoarthritis Cartilage 2006;14:984–90. 51. Thorp LE, Wimmer MA, Block JA, Moisio KC, Shott S, Goker B, et al. Bone mineral density in the proximal tibia varies as a function of static alignment and knee adduction angular momentum in individuals with medial knee osteoarthritis. Bone 2006;39: 1116–22.