Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

15.1 Radionuclide Imaging of Bone

It was the development of the cyclotron by Ernest Lawrence at Berkeley during the 1930s that first made radionuclides available in the quantity and variety to allow investigations of their medical use. Lawrence’s brother John, who was a physician, directed the first clinical studies, which included an investigation by Charles Pecher of the treatment of metastatic bone disease with 89Sr [1]. An important finding made at Berkeley was the discovery of 99mTc in 1938 by Segrè and Seaborg in a sample of irradiated molybdenum [2]. But further developments had to await the end of the war.

Since calcium has no radioisotopes suitable for imaging, the next obvious element to exploit was strontium. In 1959 Bauer and Wendeberg [3] described the first clinical study with 85Sr (Eγ = 514 keV; T1/2 = 65 days), but its use was limited by the long half-life and consequent high radiation dose. In 1964 Charkes, Sklaroff and Bierly [4] described skeletal scintigraphy with 87mSr (Eγ = 388 keV; T1/2 = 2.8 h), a short half-life isomeric radionuclide conveniently available from a generator. 87mSr was the bone agent used in Sheffield where one of us (GB) started work in 1972. Its disadvantage was the high renal tubular reabsorption of strontium, which combined with the short half-life led to indistinct images with a high soft tissue background (Fig. 15.1).

Fig. 15.1
figure 1

Colour ribbon printout of an 87mSr bone scan performed on a rectilinear scanner (Reproduced with permission from Kemp et al. [20])

Full size image

A much superior agent, 18F-fluoride, was described by Blau, Nagler and Bender in 1962 [5], but its use required an onsite cyclotron which was available in very few centres. By the early 1970s the situation was ripe for the introduction of 99mTc labelled bone radiopharmaceuticals.

Interest in 99mTc for Nuclear Medicine imaging began in the early 1960s following the development of the 99mTc generator by Tucker and Greene at the Brookhaven laboratory [6]. The first technetium labelled bone agent, 99mTc-polyphosphate, was described by Subramanian and McAfee [7] in 1971 (Fig. 15.2), and so radical was the improvement in image quality that by 1973 87mSr had fallen out of use.

Fig. 15.2
figure 2

Black and white film scintiscan from a rectilinear scanner showing a 99mTc-polyphosphate bone scan in a patient with no abnormalities (Reproduced with permission from Redman and Turley [21])

Full size image

In 1975 Subramanian described a superior agent, 99mTc-methylene diphosphonate [8], and with the introduction of the wide-field-of-view gamma camera and the replacement of spot views with whole-body scans the Nuclear Medicine bone scan took its modern form.

Recent years have seen a further change in the choice of optimum tracer as the wider availability of PET scanners has brought renewed interest in 18F-fluoride. 18F is superior to 99mTc-MDP for skeletal imaging because of its higher plasma clearance to bone and absence of protein binding, both factors that lead to improved bone to soft tissue uptake. In addition PET is a superior imaging technique with higher spatial resolution and sensitivity compared with the gamma camera. The 99mTc-MDP bone scan is an old and trusted friend that continues to perform with some distinction, but it is apparent that we can do significantly better with 18F. As PET scanners become more widely available there appears to be a compelling case for the introduction of 18F-fluoride as the preferred agent [9].

15.2 Quantitation of Bone Tracer Kinetics

There is a large early literature describing non-imaging studies of bone tracer kinetics using nuclides such as 45Ca (T1/2 = 163 days), 47Ca (T1/2 = 4.5 days) and 85Sr (T1/2 = 65 days). Imaging studies using short half-life tracers such as 99mTc-MDP and 18F-fluoride provide more restricted information, but have the advantage of allowing regional as well as whole skeleton measurements. Perhaps the most widely known bone quantitation method using an imaging tracer is the 24-h 99mTc-MDP whole-body retention test described by Fogelman in the late 1970s while working at the Glasgow Royal Infirmary [10]. In this test the patient is injected with a tracer amount (~1 MBq) of 99mTc-MDP and has a head-to-foot measurement in a shadow shield whole-body counter. The measurement is repeated 24 h later and, on the assumption the tracer is either cleared to bone or excreted through the kidneys, the counts are corrected for background and radioactive decay and the retention of 99mTc-MDP calculated. With the lack today of whole-body counter facilities, Brenner described a gamma camera measurement of whole-skeleton uptake based on whole-body scans acquired at 3 min and 6 h after injection [11]. By drawing a large region of interest over the adductor muscles the percentage of tracer still in soft tissue at 6 h is inferred and subtracted from the whole body retention to estimate the amount in bone.

A new method of using gamma camera scans to perform whole-skeleton and regional measurements of 99mTc-MDP bone plasma clearance (analogous to the measurement of renal function using GFR) was developed by Amelia Moore and colleagues at Guy’s Hospital [12]. Fast (~10 min) anterior and posterior whole-body scans are performed at 10 min, 1, 2, 3 and 4 h after injection, and blood samples taken at 5, 20, 60, 120, 180 and 240 min. The latter are centrifuged by ultrafiltration to determine the plasma concentration of free (non-protein bound) 99mTc-MDP and the Patlak method used to determine whole-skeleton and regional (skull, spine, pelvis, arms, legs) bone plasma clearance. The method was used by Moore to investigate the effect of teriparatide on the bone scan in postmenopausal women treated for osteoporosis [13].

Just as PET imaging with 18F-fluoride is superior to the 99mTc-MDP gamma camera scan, the same is true for bone quantitation. Hawkins et al. [14] described a method of measuring 18F bone plasma clearance in the lumbar spine (units: mL min−1 mL−1) by applying compartmental modelling to a 60-min dynamic PET scan with arterial sampling of the plasma curve. The method was simplified by Michelle Frost and colleagues at Guy’s Hospital [15] so that from a series of venous blood samples and bed positions regional plasma clearance could be measured across the entire skeleton with a single injection of tracer. With the ability of 18F PET scans to make measurements at the hip, the most important fracture site, we believe this is the best method for studying the effect of osteoporosis treatments on regional bone metabolism.

15.3 Bone Densitometry

The first bone densitometer based on photon absorptiometry was described by Cameron and Sorenson in 1963 [16] and used the 27 keV radiation from a 125I source to measure bone mineral content in the radius, a method known as single photon absorptiometry (SPA). Because the beam contained photons with just a single energy it was necessary to place the patient’s forearm in a water bath to simulate a constant thickness of soft tissue across the wrist.

During the 1970s Medical Physics groups at the University of Wisconsin and other centres pioneered the development of dual photon absorptiometry (DPA) using a rectilinear scanning device with a 153Gd source with emissions at 44 and 100 keV [17]. Unlike SPA, these devices were able to scan the spine and hip. By the early 1980s several companies were manufacturing DPA scanners and in 1982 Frans Verlaan, founder of Vertec Scientific, organised a conference in London for those interested in the new technology that was attended by many from the Nuclear Medicine community. In some UK centres medical physicists such as Victor Poll in Southampton built their own systems from old rectilinear scanners. Because of the use of a radioactive source, the DPA devices were largely confined to Nuclear Medicine Departments with experience in the safe handling of radionuclides.

The limitation of DPA was that it is impossible to contain sufficient radioactivity in a small enough volume to achieve both the necessary count rate and adequate spatial resolution. In 1988 in the recently opened Guy’s Osteoporosis Unit patient appointment times for the Novo BMC-LAB 22a DPA system were 1 h long, as it took 30 min to scan and analyse the spine, and a similar time for the hip (Fig. 15.3a, b).

Fig. 15.3
figure 3figure 3

(a) A Novo BMC-LAB 22a DPA scanner as used in the mid-1980s. (b) A lumbar spine DPA scan being analysed on a Novo BMC-LAB 22a system. Note the facility for the line-by-line adjustment of the soft tissue reference baseline on the left hand side of the image

Full size image

In the same year bone densitometry was revolutionised when Jay Stein, founder of Hologic, replaced the 153Gd source with an X-ray tube and produced the first dual-energy X-ray absorptiometry (DXA) system. This provided an improvement in scan quality for bone densitometry just as radical as 99mTc-MDP had for the radionuclide bone scan a decade earlier. Overnight the market for DPA systems was dead!

With a rapidly growing number of centres offering a DXA service it was necessary to devise clearer ways of reporting scans. In 1994 a group of physicians led by John Kanis recommended that DXA scans should be interpreted using T-scores defined as the BMD measurement expressed in standard deviation units relative to a population of healthy young adults, with osteoporosis defined as a T-score of less than −2.5 at the spine, hip or forearm [18]. This simple definition caught the imagination of clinicians and ever since the T-score has been a cornerstone of DXA scan reporting. But it ignores the fact that fracture risk increases progressively with diminishing BMD and does not take into account factors such as age and previous fractures in evaluating risk. In 2008 a collaboration under John Kanis placed the interpretation of DXA scans on a more secure footing by launching the FRAX website [19] where a clinician can enter details of a patient’s hip T-score and clinical risk factors and obtain an estimate of their 10-year risk of fracture.