Tests for fracture risk
By Robert Schneider, MD
Bone density measurements are done in order to determine if there is low bone mass, to predict risk of future fractures, to determine which patients may need drug therapy and to monitor patients on drug therapy. Bone density is highly correlated with bone strength and with fracture risk. A 1 standard deviation decrease in bone density is associated with 1.5 to 3 times increase in the risk of future fractures depending on the site of the measurement. Evaluation of bone turnover by use of biochemical markers (collagen breakdown products -N-Telopeptide), risk factors and bone density measurements are used in combination for determining for preventive therapy.
Qualitative methods such as the evaluation of radiographs and semiquantitative methods such as Singh index have been replaced by quantitative methods. These include dual X-ray absorptiometry (DXA), single X-ray absorptiometry (SXA), quantitative computerized tomography (QCT), quantitative ultrasound sonometry (QUS) and radiographic absorptiometry (photo densitometry). Sites of measurement are divided into central, which includes the spine and proximal femur, and peripheral measurements that include arm, phalanges and metacarpals, calcaneus, tibia and other sites in the appendicular skeleton.
The Bone Mass Measurement Act of 1998 provides Medicare reimbursement of bone density testing. Payments are made for both central and peripheral measurements, but at different rates with central testing receiving higher reimbursement than peripheral testing. If a peripheral test is done, a central test will be reimbursed for base line monitoring for drug therapy. Repeat measurements can be done every two years and more often for patients on steroid therapy or further medical indications.
The choice of technique to use in measuring bone density is based on availability, cost, accuracy, precision, diagnostic sensitivity and experience with the various methods. In recent years, central DXA has been the technique that is most frequently in use in clinical practice and in clinical drug trials. The World Health Organization (WHO) standards for characterizing patients as normal (with 1 SD of peak bone mass), osteopenic (1 to 2.5 SD below peak bone mass), and osteoporotic (more than 2.5 SD below peak bone mass) are based on central DXA measurements.
Central DXA measurements are done of the lumbar spine from L1-L4 or L2-L4, and also the proximal femur with regions of interest including the femoral neck, Ward's triangle, intertrochanteric region and a total femur region which includes all the these sites and the proximal shaft of the femur. Forearm densitometry also can be done with the same equipment at the junction of the middle and distal thirds of the forearm and the ultradistal site that is proximal to the wrist joint. The junction of the middle and distal thirds of the forearm or radius contains almost entirely cortical bone and may be useful in patients who have predominantly cortical bone loss, such as those with hyperparathydroidism.
Total body bone density can also be done with DXA, but is not used routinely in clinical practice. Lateral scanning provides a region of interest within the vertebral body, but excludes the posterior elements and possible calcification in the aorta and peripheral osteophytes that would be included on AP scanning. This measures predominantly cancellous bone. The majority of the scans of the spine are done posterior-anterior (AP spine) and most of the studies use AP spine and proximal femur.
There are single X-ray absorptiometry (SXA) and dual X-ray absorptiometers (DXA) that are dedicated to scanning peripheral sites including the calcaneous and forearm. These devices are smaller and cheaper than central DXA densitometers. They are mobile and may be brought from site to site. DXA densitometers also are available for scanning the fingers.
Quantitative Computerized Tomography (QCT) was developed at the same time as DXA. Densitometry of the lumbar vertebrae can be done on an ordinary CT scanner available in most hospitals with the addition of a phantom and software program for determining bone density. It provides a measurement of almost purely cancellous bone in the region of interest within the vertebral body, excluding the posterior elements, anterior osteophytes and aortic calcification. Three-dimension QCT has become available; it allows densitometry of the proximal femur as well as the spine. Devices for quantitative ultrasound sonometry (QUS) recently have been approved by the FDA for bone density testing. QUS devices mainly measure the calcaneus, although some devices measure the tibia, patella and fingers. Although quantitative ultrasound measures mainly the density of bone, bone architecturing, including conductivity of the trabecular, also play some part in the measurements. Measurements are done of broad ban ultrasound attenuation, and speed of sound through the bone. These two measurements are sometimes combined into an index called either qualitative ultrasound index (QUI) or stiffness.
Radiographic absorptiometry is a method of photo densitometry in which radiographs of the hands are done along with an aluminum step wedge on nonscreen film. The films are sent to a commercial company that digitizes the films and does computations providing a bone density reading, based on photographic density calculated by a computer. The radiographic absorptiomertry has relatively good accuracy and good precision and is inexpensive.
Measurements at any site with any of the different types of bone densitometry equipment provide information about fracture risk. Although measurements of the proximal femur are best for predicting future fractures, the other sites of measurement also can be used to predict the risk of hip fractures. Decreased density of any measurement site correlates well with the future global fracture risk. Low bone density single site is still an indicator of increased risk of fracture even if other sites are normal.
There is only a moderate correlation of 50 or 60 percent bone density among different anatomic sites. Bone density correlation of the same anatomic sites such as different sites in the proximal femur is still only moderate at 70 to 80 percent. There also is a wide difference in the prevalent sensitivity and specificity for fractures among measures of various sites. These variations or discordance among sites is most likely due to differences in bone loss with aging among the various sites. The greater the number of sites of low bone density, the higher the risk of future fracture. Conversely, peripheral screening will miss 30 percent of the osteoporotic hip and spine patients.
Central DXA measurement of the spine and proximal femur remains the gold standard for bone densitometry with 1 percent to 2 percent and 2 percent to 3 percent precision, respectively. The WHO standards for categorizing patients as normal, osteopenic and osteoporotic were based on DXA measurements of the spine and proximal femur. The femoral neck and total hip measurements are preferred for diagnosis in the elderly as spine measurements are often elevated due to degenerative disease in this age group.
For monitoring patients on drug therapy, the densitometry of the spine is preferred because a change in density occurs more in cancellous bone of the spine than in the proximal femur. Most of the peripheral densitometry measurements are not useful in monitoring patients as there is little change with drug therapy in the cortical bone mass in the forearm and fingers. There is some indication that QUS of the calcaneus may be useful for monitoring patients, however, this is not certain at this time. Peripheral densitometry is cheaper than central measurements and can be used for mass screening. It has to be recognized, however, that some patients with normal peripheral bone density will have low central bone density so that single site measurements may fail to diagnose patients who are osteoporotic.
At this time, in most cases when the peripheral densitometry scans
show a low bone density, a central bone density scan is performed
if the patient is placed on drug therapy. There are now enough
bone densitometers in the United States that bone densitometry
testing can be done on all patients in whom it is indicated.
Robert Schneider, MD, is attending radiologist at the Hospital for Special Surgery and associate professor, Cornell Medical College.