| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Musculoskeletal Imaging |
1 All authors: Department of Radiology, University of Vienna, Waehringer Guertel 18-20, Vienna 1090, Austria.
Received December 5, 2003; accepted after revision April 1, 2004.
Address correspondence to C. R. Krestan
(christian.krestan{at}univie.ac.at).
OBJECTIVE. The aim of our study was to evaluate the diagnostic agreement of quantitative sonography of the radius and proximal phalanx and dual-energy X-ray absorptiometry (DXA) of the radius, lumbar spine, and femoral neck for the detection of osteoporosis.
MATERIALS AND METHODS. In 95 women (mean age, 53 ± 13 years) and 26 men (mean age, 53 ± 13 years), DXA measurements of the lumbar spine (posterior–anterior, L1–L4) and the femoral neck, as well as quantitative sonography of the radius and proximal phalanx of the third finger were obtained. The percentage of patients below a given threshold was calculated for each imaging technique. A T score of less than –2.5 indicated presence of osteoporosis. Diagnostic agreement in identifying individuals with osteoporosis was assessed using kappa scores.
RESULTS. Between 14% and 22% of the patients were classified as
osteoporotic after DXA of the various regions of interest of the radius, 31%
after DXA of the spine, 43% after DXA of the femoral neck, 32% after
quantitative sonography of the distal radius, and 34% after quantitative
sonography of the phalanx of the third finger. Correlation coefficients
between T values for quantitative sonography and those for DXA varied between
not significant and 0.54 at the different regions. Kappa analysis showed the
diagnostic agreement among quantitative sonography and DXA to be fair to
moderate (
= 0.38–0.48). The highest agreement was between
quantitative sonography of the proximal phalanx of the third finger and DXA of
the total radius ( 0.48; p < 0.05).
CONCLUSION. Considerable diagnostic disagreement exists between quantitative sonography and DXA of the forearm, as is true for most quantitative techniques in the assessment of skeletal status. The lack of correlation makes quantitative sonography impractical for routine diagnostic use but might characterize different parameters related to bone quality.
Quantitative sonography methods have been introduced in recent years for the assessment of skeletal status in osteoporosis. The performance of quantitative sonography has been evaluated in several studies that primarily evaluated the calcaneus. Quantitative sonography is the only commercially available approach for noninvasive assessment of fracture risk that does not require ionizing radiation. The technique measures ultrasound velocity (speed of sound) and broadband sonographic attenuation, usually applied at the calcaneus, tibia, and phalanges, but has also been used at the patella and distal radius [1]. Quantitative sonography is most often used at the calcaneus and has shown valid results in clinical studies with respect to prediction of fracture risk [2, 3]. However, quantitative sonography at the calcaneus does not correlate to a clinically usable extent with DXA at the lumbar spine or hip [4] but may predict fracture independent of bone mineral density [5].
Quantitative sonography and bone mineral density can provide complementary information to improve estimates of vertebral fracture risk [6].
Additional clinical applications for quantitative sonography, specifically the assessment of the rate of change in monitoring disease progression or response to treatment, require further investigation. More investigations that assess innovative quantitative sonography techniques in well-defined research settings are important to determine and utilize the full potential of this technology for the benefit of early detection and monitoring of osteoporosis.
Significant discordance often exists among measurements taken at different skeletal sites and among measurements obtained with different techniques [7, 8]. Quantitative sonography at the radius and phalanges is a recent extension of the technical protocol for sonographic measurements. The precision of quantitative sonography was recently described [9], but its clinical usefulness, as well as its relationship to radiation-based methods, has yet to be fully established. The purpose of our study was to evaluate the diagnostic agreement of quantitative sonography of the radius and phalanges with DXA of the spine, radius, and femur in the detection of osteoporosis in Caucasian women and men.
Materials and Methods
The study group consisted of unselected clinical patients who were referred to our department for standard diagnostic workup for suspected osteoporosis. In 121 patients (95 women: mean age ± SD, 53 ± 13 years; 26 men, 53 ± 13 years), bone mineral density measurements of the lumbar spine (posterior–anterior, L1–L4), the proximal femur (neck), and the radius (proximal third, mid, ultradistal region, total radius) were performed with a QDR 4500 scanner according to the manufacturer's (Hologic) recommended standard procedures. The quantitative sonography measurements of the ultrasound velocity in the radius and proximal phalanx of the third finger of the nondominant hand were obtained with a non–imaging-guided sonography unit (Omnisense, Sunlight). Omnisense is a new, multisite quantitative sonography device that measures the acoustic velocity in the axial transmission mode along the cortex. The technique has been described by Njeh et al. [10]. The probe was aligned parallel to the bone according to the manufacturer's specifications. The midpoint of the distance between the tip of the elbow and the tip of the third phalanx was marked with a pen. The long axis of the probe was positioned parallel to the radius with the T marker of the probe placed at the skin marker. Then the probe was moved in a semiarc around the radius within 70° ulnar and 70° radial (total, 140°) rotation. For measurement at the proximal phalanx of the third finger, the length of the middle phalanx of the third finger was measured and this value was marked at the proximal phalanx. The T marker of the probe was then placed at the radial side of the proximal phalanx and moved around at an angle of 180°. The diagnosis of osteoporosis was defined, according to the World Health Organization (WHO) criteria [11], as a T score less than –2.5.
In 19 (16.5%) of 115 women and in nine (25.7%) of 35 men, quantitative sonography at the radius or at the phalanx could not be performed in a manner that permitted calculation of results. In two patients (one woman, one man), quantitative sonography results were not obtainable at either region.
Means and SDs for all examined parameters were calculated separately for men and for women. The percentage of patients having T scores under the threshold of –2.5 was calculated for both groups. Paired comparisons of all parameters and T scores were obtained using Pearson's correlation coefficients (r) and p values for testing the significance of correlations. A kappa score analysis, performed separately for women and for men to compare techniques for diagnostic ability, was achieved by classifying every subject as osteoporotic if the T score with respect to the reference group was less than –2.5.
Results
T scores and percentages of patients having T scores under the threshold of –2.5 for the different measured parameters for women and men are shown in Table 1. Classification of osteoporotic women ranged between 14% and 22% for bone mineral density at the various regions of interest for the radius, and was 43% for measurement of the bone mineral density at the femoral neck, 31% by DXA at the spine, 32% by quantitative sonography at the distal radius, and 34% by quantitative sonography at the phalanx of the third finger.
|
Linear regression analysis when comparing parameters showed moderate correlation coefficients for absolute values, which were between 0.28 (quantitative sonography of radius vs DXA for the L1–L4) and 0.51 (quantitative sonography of phalanx vs DXA of proximal third of radius), in the women (Table 2).
|
Linear regression analysis when comparing T scores showed low to moderate correlation coefficients in women for quantitative sonography and DXA and were considerably higher when comparing DXA with DXA. Coefficients ranged between 0.26 (quantitative sonography of radius vs DXA of L1–L4) and 0.98 (DXA of mid radius vs DXA of total radius) (Table 3 and Figs. 1, 2, 3).
|
|
|
|
Kappa analysis of quantitative sonography and DXA showed the diagnostic
agreement to be fair to moderate (
0.48)
(Table 4). The highest
agreement was between quantitative sonography at the proximal phalanx of the
third finger and DXA at the total radius ( = 0.48).
|
Kappa statistics for agreement of the classification of osteoporotic individuals showed only fair agreement (0.22–0.41) when comparing quantitative sonography at the radius and DXA at the radius with DXA at the axial skeleton. Agreement was moderate to fair for quantitative sonography at the phalanges compared with DXA at the various sites (0.23–0.48) (Table 4).
Correlation coefficients between the T values for quantitative sonography and those for DXA varied between not significant and 0.54 at the different regions.
A significant difference exists in the classification of osteoporosis in men and in women when measured with speed of sound of the phalanx of the third finger only (Table 1). Bone loss in osteoporosis does not always affect different body sites in a symmetric or parallel fashion. Bone loss due to hormone deficiency is usually found in women and affects the spine earlier than the peripheral skeleton. Therefore, central measurement sites might be better suited to monitor early bone loss [7].
Discussion
Several previous investigations have shown that both DXA at the neck and spine and quantitative sonography at the calcaneus can be used to differentiate between healthy premenopausal and healthy postmenopausal women and that these tests reflect age- and menopause-related bone loss [12, 13]. Quantitative sonography with the Omnisense device can be used to differentiate healthy premenopausal from postmenopausal women [14] and also to discriminate between subjects with and without fractures [9]. Weiss et al. [15] showed that quantitative sonography at the radius can be used to discriminate patients with proximal hip fracture from those without hip fracture. Guglielmi et al. [16] found that phalangeal ultrasound velocity can be used to discriminate between healthy patients and those with vertebral fractures, comparable with DXA at the spine.
Ultrasound velocity has also been applied to monitor changes and response to treatment in women with osteoporosis. Weiss et al. [17] found a significant increase of mean T scores for quantitative sonography in response to alendronate treatment with the Omnisense device in postmenopausal women with osteoporosis. Studies using the reflection ultrasound velocities method showed a significant increase in ultrasound velocity in osteoporotic patients after 9–17 months of treatment with slow-release sodium fluoride plus calcium citrate. These observations indicate that velocity measurements with the reflection ultrasound velocities method show an improvement in bone elasticity associated in part with an improvement in the rate of bone mineralization and an improvement in bone quality at the structural level, as shown by microarchitecture [18, 19].
DXA measures the bone mineral content in grams per square centimeter, which is related to the combined attenuation of X rays in cortical and cancellous bone, because DXA is a projectional 2D measurement. Sonographic waves as used in quantitative sonography, however, have a complex propagation and are affected not only by the amount of material (density of bone) but also by its elasticity and structure [20].
Ultrasound velocity is lower in cancellous bone than in cortical bone and shows a significant decline with age in postmenopausal women [21]. The sonographic unit (Omnisense) that we used in our study measures axial propagation of sonography through a thin layer beneath the surface of the cortical bone [22]. Comparative studies between bone mineral density at the lumbar spine and the femoral neck and quantitative sonography at the heel showed that the best longitudinal sensitivity after alendronate therapy could be achieved with DXA at the lumbar spine, but both quantitative sonography at the heel and femur bone mineral density showed similar abilities [23, 24].
Similar to the findings of other authors [25], our study results derived at the various sites with DXA and quantitative sonography were only moderately correlated, which precluded the prediction of one parameter from another in the individual patient and caused disagreement in diagnostic classification according to WHO criteria. DXA of the spine, which is the most widespread diagnostic technique and is generally accepted as a reliable approach for the diagnosis of osteoporosis, provided comparatively poor agreement with quantitative sonography at the distal radius or phalanx of the third finger. Correlation coefficients of absolute values of quantitative sonography at the radius compared with bone mineral density at the radius were higher than quantitative sonography compared with bone mineral density at the spine and femur, which might be attributable to the close proximity of the regions of interest (Table 2). The comparison between radiography-based methods and sonographic measurements yielded lower r values than the comparison between radiography-based measurements of the spine and hip alone. The low correlation between quantitative sonography and bone mineral density might be attributable to a nonlinear dependence on bone mineral loss, which was shown in vitro by Wu et al. [26]. T scores were correlated at practically the same levels as absolute values, which might be an indicator that the reference databases given by the systems warrant reliable results. Speed of sound in our study did not compare well with DXA for identifying individuals at risk for osteoporosis, with low kappa scores between 0.23 and 0.29 when comparing speed of sound with established techniques (DXA) at the lumbar spine and femoral neck. This discrepancy occurs because DXA and quantitative sonography assess different bone properties, including bone mineral content and biomechanical properties such as geometry, elasticity, and architecture. Other authors also found that kappa score analysis (using a –2.5 T score for osteoporosis) showed that the diagnostic agreement when comparing different techniques to classify women as osteopenic or osteoporotic was poor, with kappa scores averaging about 0.4 (exceptions were quantitative CT and DXA at lumbar spine). Often, different patients were estimated as being at risk by using different measurement sites or techniques [7].
Our study was limited by the inability of the quantitative sonography device to deliver reliable data for 31 (20%) of 152 patients, despite the fact that examinations strongly adhered to the manufacturer's specifications at both sites. In the clinical setting, limitations of the quantitative sonography device included short bone length of the proximal phalanx and deformities due to arthritis, which made it impossible for the technician to derive a valid result. In our study, as reported by other authors [27], quantitative sonography was not sufficiently related to DXA either at local sites (radius) or at axial or other peripheral sites. Because DXA is the gold standard for fracture risk prediction and the prerequisite for the initiation of antiosteoporosis therapy, the lack of correlation with quantitative sonography makes the latter impractical for routine diagnostic use but might make it practical for expression of different parameters related to bone quality that could not be assessed within the scope of the present study [3, 28, 29]. However, revised quantitative sonography T-score thresholds as suggested by Knapp et al. [30, 31], who found that the WHO definition was not suitable for use with speed-of-sound measurements, may qualify the technique to be a tool for further investigation.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
