Quantitative Bone Scintigraphy. a Study in Patients with Prostatic Carcinoma

Quantitative Bone Scintigraphy A study in patients with prostatic carcinoma

Gunilla Sundkvist

\OO -II ,

Malmö 1991 Organization Document name LUND UNIVERSITY DOCTORAL DISSERTATION Department of Clinical Physiology Date of issue Malmö Allmänna Sjukhus 1991-05-08 S-214 01 Malmö, Sweden CODEN: LUMEDW-f.MEFM-lOlOU-lOO (1991)

Authors) Sponsoring organization Gunilla Sundkvist

Title and subtitle Quantitative Bone Scintigraphy. A study in patients with prostatic carcinoma.

Abstract Quantitative bone scintigraphy was performed in patients with prostatic carcinoma before orchiectomy as well as two weeks, two and six months after operation. The count rate was recorded as serial gamma camera images over the lower thoracic and all lumbar vertebrae from 1 to 240 min and at 24 h after injection of "Tcm-MDP. In almost all abnormal vertebrae an increased count rate was observed within one hour after injection. Most of the vertebrae which were considered normal at 4 h after injection, but had an increased 24 h/4 h ratio developed into abnormal vertebrae later in the study. The patients with normal bone scintigrams showed no change in "Tcm- MDP uptake during the study. The reproducibility of quantitative bone scintigraphy was found to be ± 7% (1 SD). In response to therapy, most of the patients with abnormal bone scintigrams showed an increase in count rate two weeks after operation followed by a decrease to the pre-operative level after two months and a o further decrease after six months. This so called "flare phenomenon" was found to m ta indicate "Tc -MDP in the vascular phase as well as an active bone uptake. In some of the patients the whole-body retention of "Tcm-MDP after 24 h and the bone Is mineral density in the vertebrae were determined and found to be valuable in the a- * interpretation of skeletal metastases and the assessment of response to therapy. 81 *^w Prostatic Neoplasms, Bone Neoplasmp, Thoracic Vertebrae, Lumbar Vertebrae, Orchiectomy, Radionuclide Imaging

Gasification system and/or index tenns (if any)

Supplementary bibliographical information •fKliish

ISSN and fcry title ISBN

Recipient'i notes Number of pages JJJ Price

Security clarification

Distribution by (name and address) Department cf Clinical Physiology. Malmö Allnänna Sjukhus, S-214 01 Malmö, Sweden. I, the undersigned, being the copyright owner of the abstract of the above-mentioned dissertation, hereby grant to all reference sources permission to puMiafa and disseminate the abstract of the above-mentioned dissertation.

Signature Date 1991-03-01 Quantitative Bone Scintigraphy A study in patients with prostatic carcinoma

Akademisk avhandling som med vederl ^ligt tillstånd av Medicinska Fakulteten vid Lunds Universitet för avläggande av -doktorsexamen i medicinsk vetenskap kommer att offentligen försvaras i Kl i. ;kt Fysiologiska Avdelningens föreläsningssal, Allmänna Sjukhuset, Malmö, onsda; n den 8 maj 1991 kl. 9.00

Gunilla Sundkvist Departments of Clinical Physiology and Radiation Physics, Lund University, Malmö Allmänna Sjukhus, Malmö, Sweden

Quantitative Bone Scintigraphy A study in patients with prostatic carcinoma

Gunilla Sundkvist

Malmö 1991 ISBN91-628-0250-X © Gunilla Sundkvist Printed in Sweden Lund 1991 To Lennart Carolina, Cornelia and Christina This thesis is based on the following papers, referred to in the text by their Roman numerals:

I. Sundkvist GMG, Ahlgren L, Lilja B, Mattsson S, Abrahamsson PA, Wadström LB: Repeated quantitative bone scintigraphy in patients with prostatic carcinoma treated with orchiectomy. Eur J Nucl Med 1988; 14:203-206.

II. Sundkvist GMG, Ahlgren L, Lilja B, Mattsson S, Abrahamsson PA: Dynamic quantitative bone scintigraphy in patients with prostatic carcinoma treated by orchiectomy. Eur J Nucl Med 1990; 16:671-676.

III. Sundkvist GMG, Ahlgren L, Lilja B, Mattsson S: Additional information from quantitative 24-hour ""^Tc-MDP bone scintigraphy in patients with prostatic carcinoma. Accepted for publication in Acta Oncol.

IV. Sundkvist GMG, Ahlgren L, Lilja B, Mattsson S: Quantitative bone scintigraphy and 24-hour whole-body counting of 99mTc methylene diphosphonate in patients with prostatic carcinoma. Submitted.

V. Sundkvist GMG, Nilsson M, Ahlgren L, Lilja B, Mattsson S, Birch- Iensen M, Abrahamsson PA: Relation between 99Tcm-MDP uptake and bone mineral density in patients with prostatic carcinoma. Accepted for publication in Nucl Med Commun. Contents

Abbreviations 9

Introduction 11 Historical background 11 Bone scintigraphy 12 Whole-body counting 13 Quantitative computed tomography 14 The metastatic process 14 Prostatic carcinoma 16 Mechanism of bone metastasis in prostatic carcinoma 16 Morphological aspects of vertebral metastases in prostatic carcinoma 17 Orchiectomy 18

Aims of *he Study 19

Materials and Methods 20 Patients with prostatic carcinoma 20 Control group 21 Dynamic quantitative bone scintigraphy 21 Correction procedures 23 Time-count rate curves 24 Routine whole-body scintigraphy 24 Whole-body counting 25 Quantitative computed tomography 25 Statistical methods 26 Results 27 Quantitative bone scintigraphy in the control group 27 Quantitative bone scintigraphy in patients with prostatic carcinoma before orchiectomy 29 Quantitative bone scintigraphy in patients with prostatic carcinoma after orchiectomy 30 Quantitative bone scintigraphy in relation to whole-body counting in a control group and in patients with prostatic carcinoma before and after orchiectomy 32 Quantitative bone scintigraphy in relation to quantitative computed tomography in patients with prostatic carcinoma at the time of orchiectomy and post-operatively 33

Discussion 34

Summary and Conclusions 39

Acknowledgements 41

References 42

Papers I-V 51

8 Abbreviations

BMD bone mineral density MDP methylene diphosphonate ROI region of interest QCT quantitative computed tomography QS quantitative bone scintigraphy WBC whole-body counting WBR whole-body retention Introduction

Historical background

The discovery of artificially induced radioactivity was made by Joliot and Curie in 1934. One year later, Chiewitz and de Hevesy used the radionuclide, phosphorus-32 (32P), in the study of bone metabolism in rats. Treadwell et al. demonstrated in 1942 that strontium-89 (89Sr) accumulated chiefly in growing bone and in osteogenic tumour tissue in man. Mulry and Dudley found in 1951 that early metastases to bone could be identified through the use of gallium-72 (72Ga) before changes could be visible roentgenographically. Bauer and Wendeberg used in 1959 calcium-47 (47Ca) and strontium-85 (85Sr) in the study of localized skeletal lesions. Blau etal. found in 1962 that the uptake of fluorine-18 (18F)in normal bone is very low compared to sites of active osteogenesis. Charkes et al. employed strontium-87m (g7Srm) in 1964 in the detection of bone metastases. However, the above mentioned radionuclides showed serious limitations in the clinical use, related to radiation absorbed dose, physical or pharmacologic properties and economic factors. The introduction of the technetium-99m (99Tcm) complex of tripolyphosphate with more favourable characteristics was made in 1971 by Subramanian and McAfee. Some years later, the 99Tcm complex of methylene diphosphonate was proposed as a new skeletal imaging agent (Subramanian et al. 1973). The available evidence has suggested that the technetium-99m diphosphonates adsorb onto the surface of hydroxyapatite crystals (Francis 1969; Jones et al. 1976). Although the precise mechanism is still not completely understood, the radio- pharmaceutical is widely used.

11 In 1951 Cassen et al. invented the automatic scanner and one year later the gamma-ray pinhole camera was constructed by Anger (1952). In combination with the parallel-hole collimator (Johansson and Skanse 1953), gamma camera technology has undergone an intensive development during the last decades. The routine use of computers to acquire, process, display and store images has expanded the utility of the gamma camera, which has led to an important improvement in the imaging technique.

Bone scintigraphy

Bone scintigraphy is a useful method in the examination of the skeleton. Areas of abnormally increased uptake may represent skeletal involvement by tumours, metabolic or inflammatory diseases, Paget's disease or trauma (Thrall and Burton 1987). The routine bone scintigraphy is normally performed 3-4 hours after injection of 99Tcm-MDP and qualitatively interpreted. Visually, a generalized increased uptake in the skeleton may be misinterpreted as normal. Also, it may be difficult to identify lesions in which the radionuclide concentration is only slightly higher than in surrounding bone (Citrin etal. 1974). Quantification of the radionuclide uptake may then provide a more objective evaluation of the scintigram (Hardy eta). 1980),

For detection of skeletal metastases bone scintigraphy is a sensitive method (Osmond et al. 1975). In patients with newly diagnosed non- osseous primary tumours with bone metastases (especially carcinoma of the breast, lung and prostate) the determination of bone involvement and

12 definition of its extent is important for staging and management. Bone scintigraphy is especially useful in early detection of metastases. When the tumour cells invade the bone, remodelling results in changed skeletal metabolism with increased tracer localization (Galasko 1977). Bone scintigraphy is, however, a non specific method (Corrie et al. 1988) and it may be difficult to distinguish between uptake due to benign or malignant processes. It is therefore of special interest to further develop the method.

In the evaluation of skeletal metastases under treatment, repeated QS have shown to be of considerable value. Changes in uptake may accurately reflect the degree of the metastatic disease (Pollen et al. 1981). Initially an increased uptake can be seen, followed by a decrease after about two months. This "flare phenomenon" has been found to be an osteoblastic response as part of a healing reaction to successful therapy (Pollen and Shlaer 1979). Therefore, further knowledge about the nature of this phenomenon may be essential for the understanding of the metastatic process. It may also be important to follow the uptake during longer time after therapy to obtain new information on the response.

Whole-body counting

Quantitative determination of activity of radionuclides in the whole body (WBC) has been applied in the routine clinical work for thirty years. Cederquist and Liden introduced in 1962 a counting technique with single crystal scanning, with measurement of the patients prone and supine, in a low background room. Using WBC of 85Sr and 47Ca it was possible to distinguish between patients with and without widespread malignant skeletal diseases (Cederquist 1964).

13 Later, 24-h WBR of Tc-99m diphosphonate was used in the study of metabolic bone diseases (Fogelman et al. 1978). Also, in patients with metastatic carcinoma of the prostate, WBR of 99Tcm-MDP has been studied (Castronovo et al. 1985; Dann et al. 1987), however, without comparison to quantitative, local 99Tcm-MDP uptake.

Quantitative computed tomography

QCT and dual photon absorptiometry (DPA) are the conventional methods cf measuring bone mineral content in the spine (Genant et al. 1987; Goodwin 1987). In the study of metastatic disease, the potential advantage of QCT over DPA is its capability to separate trabecular bone from cortical bone (Reinbold et al. 1986), as metastases preferentially appear in the trabecular part of the vertebra (Jacobs 1983). In order to better understand the nature of the metastatic process and its influence on bone resorption and formation, it may be of interest to follow both the MDP uptake and the BMD value simultaneously.

The metastatic process

A metastasis is a neoplastic lesion that arises from another cancer with which it is no longer in continuity (Mirra et al. 1989). The ability to produce metastases is a property of malignant tumour cells (Poste and Fidler 1980).

•-ner - The development of a metastasis is a multi-step process depending on the interplay of intrinsic tumour cell properties, e.g., cancer cell heterogeneity and host factors, including factors in the normal tissue environment where the metastases appear (Fidler 1990; Price 1990). The pathogenesis of cancer metastasis is complex (Poste and Fidler 1980; Weiss 1985; Nicolson 1988; Fidler 1990) and a simplified sequence describing the major steps (Morgan et al. 1990) is given below.

At some point in the primary tumour growth, an angiogenesis stimulating factor is elaborated producing abnormal vascular channels, i.e., tumour vessels. This is followed by invasion of the stroma of the organ of origin and/or adjacent structures. At this point tumour cells lack normal adherence characteristics and embolize into the lymphatics and/or blood vessels. The vast majority of tumour cells which enter the circulation are distroyed by mechanical turbulence. The small number of surviving tumour cells must then attach to the endothelial barrier and basement membrane and establish themselves in the stroma of the involved organ. These tumour cells must then survive an array of local and systemic defence mechanisms including biochemical, hormonal and immunologic factors and vascularize themselves, by the action of angiogenic factors, before becoming established metastasis.

Paget proposed already in 1889 the "seed and soil" hypothesis based on the metastatic pattern seen in 735 fatal cases of breast cancer. The metastatic development was a consequence of particular tumour cells ("seed") finding a suitable environment ("soil") in order to develop and grow. Later, in 1928, Ewing was the major proponent of the anatomically oriented mechanical theory favouring the simple lodgement of tumour cell emboli in the first capillary bed (Berrettoni and Carter 1986;

15 Nicolson 1988). Both theories have been validated and they are no longer considered to be mutually exclusive (Berrettoni and Carter 1986).

Prostatic carcinoma

Adenocarcinoma of the prostate continues to have its highest incidence in the western world (Bruce and Trachtenberg 1987). In Sweden the disease accounts for about 25% of all reported cases of cancer in males (Cancer Incidence in Sweden 1987). Fatal prostatic carcinoma metastasizes to lymph nodes, lung (including pleura) and bone in about 85% each (Jacobs 1983; Brawn and Speights 1989).

Mechanism of bone metastasis in prostatic carcinoma

Bone metastasis is a pattern of selective metastasis whose mechanism is varied and complex (Berrettoni and Carter 1986). Primary prostatic tumour cells invade prostatic veins and lymphatics to gain access to the vascular system and release into the circulation. The lungs contain the first capillary bed. The cells must pass the lungs and enter the systemic circulation (Jacobs 1983). Red bone marrow is an attractive site for metastatic involvement because the vascular spaces are sinusoidal in nature and represent a relatively easy barrier for tumour cells to penetrate (Berrettoni and Carter 1986; Morgan et al. 1990). The axial skeleton is most often involved (Willis 1973; Jacobs 1983; Morgan et al. 1990) and in the spine the lumbar region is the site of the greatest number ofmetastases (Willis 1973; Schaberg and Gainor 1985, Cumming et al. 1990). Metastases in the spine usually involve the vertebral bodies, only rarely the arches or processes (Jacobs 1983; Paterson 1987).

16 Batson proposed the hypothesis that metastases to the skeleton occurred preferably to the lumbar spine via the vertebral venous system as a result of venous reflux when intrathoracic or intra-abdominal pressure is increased such as in coughing, straining and sneezing (Batson 1940). This hypothesis still has its advocates (Styles 1989; Cumming et al. 1990). However, investigations with other metastatic bone tumours have shown an almost identical distribution of bone metastases in non-prostatic tumours (Dodds et al. 1981; Morgan et al. 1990) not supporting the Batson concept.

Morphological aspects of vertebral metastases in prostatic carcinoma

Metastases to bone appear almost invariably in the red bone marrow (Willis 1973). Once established, progressive invasion of the marrow space and remodelling of the trabecular bone occurs (Jacobs 1983). In this process, the osteoclasts are the effector cells of bone resorption (Chambers 1985), while the osteoblasts are responsible for the synthesis of bone matrix (Teitelbaum and Bullough 1979). When tumour cells invade bone the morphological result is often a combination of bone destruction and formation (Milch and Changus 1956). The prostatic cancer cells may be able to directly stimulate the osteoblasts (Perkel et al. 1990) and not always require any previous resorption (Franck et al. 1982; Jacobs 1983). Skeletal metastases from prostatic carcinoma produce prevailingly osteoblastic changes with new bone formation within the cancellous bone of the marrow space and increased remodelling activity in the trabecular bone becoming broader and irregular (Jacobs 1983).

17 Orchiectomy

Half a century ago, Huggins and Hodges demonstrated the partial androgen dependence of most prostatic carcinomas (Huggins and Hodges 1941). Since that time the standard form of endocrine therapy for patients with prostatic carcinoma has been orchieciomy with surgical removal of the testes or oral administration of estrogens (Smith 1988). The goal of hormonal manipulation is to deprive the tumour cells of androgens or their by-products. Any treatment that decreases production or interferes with delivery of androgens to the cells is likely to produce a response in most patients with prostatic carcinoma (Smith 1988).

In men, who had undergone orchiectomy due to sexual delinquency, an increased bone resorption has also been observed (Stepån et al. 1989). The precise mechanism of this procedure is, however, still unclear.

18 Aims of the study

The aims of this study were to investigate:

(1) whether quantitative bone scintigraphy with 99Tcm-MDP is reproducible in the clinical routine;

(2) whether abnormal bone uptake can be predicted earlier than 4 h after an injection of 99Tcm-MDP with dynamic quantitative bone scintigraphy;

(3) whether measurements 24 h after an injection of "Tcm-MDP will provide additional information of value in distinguishing between abnormal and normal vertebrae;

(4) whether 24-hour whole-body retention of "Tcm-MDP and local bone mineral density are useful additional measurements to the local 99Tcm- MDP uptake in the interpretation of skeletal metastases;

(5) whether the measured time-count rate curves over individual vertebrae and additional 24-hour whole-body retention values can further explain the nature of the flare phenomenon;

(6) whether whole-body and local 99Tcm-MDP uptake and local bone mineral density measurements carried out as late as 6 months after orchiectomy can give new information in the assessment of response to treatment.

19 Materials and Methods

Patients with prostatic carcinoma

Seventy men (aged 60-81 years, mean age 73 years) with recently discovered, cytologically confirmed, prostatic carcinoma took part in the study (Table 1). Of the 37 patients in paper III, 31 and 34 patients also constituted the base for paper II and IV, respectively. They had no signs of metabolic or malignant diseases except prostatic carcinoma. The renal function was judged to be normal as determined by serum creatinine concentration (< 115 |a.mol/l) The patients were referred for routine whole-body scintigraphy before treatment with orchiectomy. In addition, all the patients underwent QS of the lower thoracic and all the lumbar vertebrae. No other therapy than orchiectomy was used during the course of the study.

Table 1 Paper Number of Age, years Methods and times patients mean range of registration

I 16 73 65-80 QS at 4 h after injection II 31 75 60-81 QS from injection up to 4 h III 37 74 60-81 QS at 4 h and 24 h after injection IV 34 72 60-79 QS and WBC at 1 h and 24 h after injection V 17 72 66-81 QS at 4 h after injection and QCT in close connection

20 In paper I, QS was performed before orchiectomy, two weeks and two months after operation and in papers H-FV also after six months. In paper V, the patients with normal bone scintigram were investigated at the time of orchiectomy and six months after operation, while the patients with skeletal metastases were investigated an additional time, two months post- ope ratively.

Control group

During this study, control data from men in the equivalent age interval, with no signs of metabolic or malignant disease, and no history of spine or hip trauma, was collected. As it is not ethically justified to examine healthy persons with bone scintigraphy, the control data was collected from patients referred for routine bone scintigraphy for other reasons, such as suspect fracture of the scaphoid bone. Fourteen men (aged 59-83 years, mean age 70 years) volunteered to participate in QS from injection up to 4 h and eight men (aged 60-77 years, mean age 69 years) also participated in WBC and in QS 24 h after injection.

Dynamic quantitative bone scintigraphy

With the patients supine, posterior images of the lower thoracic and all the lumbar vertebrae were recorded with a 39.0 cm (diameter) gamma camera with a high-resolution, parallel-hole collimator.

A bolus of approximately 500 MBq 99Tcm-MDP was injected intravenously. The activity in the syringe was carefully determined before

21 and after injection. By gel chromatography the amount of free pertechnetate in the preparation of 99Tcm-MDP was randomly tested to ensure that less than 5% of the activity was in the form of free pertechnetate.

In papers I and V images were stored at 240 min after injection (Table 1). In paper II serial images (2 min each) were stored from 1 min up to 11 min and at 15 min after injection, th°,n at 5-min intervals up to 30 min and thereafter at 45, 60, 90, 180, 210 and 240 minutes. In paper III images were stored at 4 h and 24 h and in paper IV at 1 h and 24 h after injection. The images were digitized in 128 x 128 matrices.

After each investigation the sensitivity of the camera was checked using a cylindrical phantom (diameter 18.0 cm, height 10.5 cm and containing a known amount of 99Tcm-MDP homogeneously distributed in water) placed on the centre of the collimator.

An ROI was defined over each vertebra from T10 to L5. As the soft tissue content of 99Tcm-MDP immediately after injection is high and the skeletal activity is low, it was not possible to place the ROIs accurately during the first 11 min. The ROIs from the image at 15 min were therefore transposed back to the initial images. This was possible as the patients had to remain in the same position. To relate the count rate in the ROI to the 99Tcm-MDP content in the vertebra, background subtraction and correction factors were applied. These factors were derived from measurements in patients with prostatic carcinoma with no signs of skeletal metastases. The count rate per pixel and unit injected activity, corrected for radioactive decay, were calculated for each ROI and measurement period.

22 Correction procedures

Subtraction of soft tissue background (paper i): To correct for soft tissue activity in front of and behind the vertebrae, a background ROI was chosen between the kidney and crista iliaca, free of projected renal and bone uptake. Measurements on a trunk-simulating phantom were performed. The count rate in the backgound ROI chosen was found to be overestimated in comparison to the real background in the vertebra ROI. According to this result, 75% of the count rate in the background ROI was subtracted from the count rate in the vertebra ROI.

Correction for varying depths (papers I and II): For each vertebra from T10 to L5, located at varying dephts, a correction was made for variations in the thickness of overlaying tissue.

Correction for background due to initial influence of highly vascularized organs (paper II): During the first few minutes after injection there is normally a high 99Tcm-MDP activity in the blood vessels and the highly vascularized organs. Overprojection of these organs particularly affects the thoracic vertebrae. From regions having a high activity content other vertebrae may also be somewhat influenced by Compton scattered radiation. Over vertebra L5 the count rate was supposed to be almost unaffected by the above-mentioned factors. Therefore, this vertebra was chosen as a standard which, after corrections for soft tissue background and depth, showed the net 99Tcm-MDP content. For T10 to L4 a set of correction factors was derived to normalize the count rate over these vertebrae to that over L5 in the time period f >m injection to 60 min. After that time no influence from this type of background was seen.

23 Time-count rate curves

For a certain time after injection the mean values (x) and standard deviations (SD) were calculated for the count rate value from the eight vertebrae in all the patients with normal bone scintigrams. The upper level for normal count rate was defined as x + 2 SD. With regard to this level all the vertebrae in the patients with skeletal metastases were divided into two groups, with normal or abnormal count rate at 240 min in the pre-operative investigation. These groups were then maintained post- operatively even if a vertebra later passed the count rate boundary between normal and abnormal. The mean values of the count rate for normal and abnormal vertebrae in patients with and without skeletal metastases were plotted against time for each investigation to get a time- count rate curve.

Routine whole-body scintigraphy

A routine whole-body scintigraphy was performed between 210 and 240 minutes after injection to detect increased tracer uptake outside the region of quantification (T10-L5). To get a standardized evaluation, all whole- body scintigrams were visually reviewed by one and the same experienced observer. If there was any doubt whether an increased uptake was due to benign or malignant disease, conventional radiographs, computed tomography or bone biopsy were obtained.

24 Whole-body counting

One hour after injection of 99Tcm-MDP, WBC was carried out. The scanning procedure was performed in a low background room with the patients in supine position. Two Nal(Tl) detectors (diameter 127 mm, height 102 mm) were located 50 cm apart, one above and one beneath the patient. The detectors were equipped with slit collimators (120 mm x 2 mm), which were made of 6 mm lead and 1.5 mm brass. To reduce the high count rate during the measurements at 1 h after injection, a 2.5 mm thick filter made of brass covered the collimator opening. The count rate was thereby reduced by a factor of 1.68. The measurements started with the centre of the detectors located approximately 30 cm cranially of jugulum. The detectors then moved caudally at a comtant speed (7.5 cm/min) giving a total scanning length of 150 cm. The net count rate in the 140 keV peak was recorded. The result was corrected for deadtime in the analyser as well as for the absorption in the filter. The initial measurement was taken as 100%. The procedure was repeated 24 hours after injection to enable the WBR to be calculated.

Quantitative computed tomography

Images through the mid-sections of each vertebra from T10 to L4 were produced on a CT unit using a slice thickness of 8 mm and an exposure at 125 kV and 350 mAs. The trabecular parts in the cross-sections of all vertebrae were outlined manually on the images as a region of interest. The BMD, which is the amount (mg) of hydroxyapatite per unit volume (cm3), was then evaluated for each individual vertebra (Nilsson et al. 1988).

25 The non-linear calibration used in this study compensates to some extent for the age-dependent fat fraction in trabecular bone. The precision of the QCT method was found to be ± 6% (1 SD) from measurements on samples placed in a body-simulating phantom. The samples contained known concentrations of hydroxyapatite in the interval 10-700 mg/cm3-

Statistical methods

Student's t-test, Wilcoxon's signed rank test and the Mann-Whitney test were used for comparison of changes within and between patient groups, respectively. The coefficient of variation was determined by variance technique (Snedecor and Cochran 1980).

26 Results

Quantitative bone scintigraphy in the control group

The time-count rate curve for the control group showed a rapid increase from injection up to 1 h. A slower increase was seen between 1 h and 3 h, with only a slight increase up to 4 h. At the next measurement at 24 h the count rate was lower than at 4 h (Fig.l). The count rate at 4 h was found to be (93 ± 2.2) 10"5 counts per pixel, s and MBq (mean ± SEM). The ratio between the count rate at 24 h and 4 h was determined to be 0.88 ± 0.01 (mean ± SEM).

27 to

Count rate/10 pixel s MBq 200

150

100 i t > t (t 1 » • 1 50

30 60 90 180 210 240 1440 Time after injection / minutes

Fig.l. Prostatic carcinoma patients with normal bone scintigram (• ): count rate over 104 vertebrae from injection up to 240 min (paper II) and over 112 vertebrae at 24 h (paper ni). Voluntary controls (o ): count rate over 112 vertebrae from injection up to 240 min and over 64 vertebrae at 24 h. Mean ± 2 SD. Quantitative bone scintigraphy in patients with prostatic carcinoma before orchiectomy

The time-count rate curves for the patients with normal bone scintigrams and for the voluntary controls are given in Fig.l. A higher individual variation was seen in the control group. No significant difference was found in the MDP uptake between the mean values for the two groups.

In the patients with a normal bone scintigram the count rate at 4 h was found to be (98 ± 1.6) 10'5 counts per pixel, s and MBq (mean ± SEM). The ratio between the count rate at 24 h and 4 h (paper III) was determined to be 0.85 ± 0.01 (mean ± SEM).

For the group of normal vertebrae in patients with abnormal scintigrams, the time-count rate curve up to 4 h was found to lie slightly above that for the group of vertebrae in patients with normal scintigrams. When measured at 24 h, the curve for the normal vertebrae in patients with abnormal scintigrams showed a smaller decrease than for the patients with normal scintigrams. The count rate at 4 h (paper II) was found to be (104 ± 1.7) 105 counts per pixel, s and MBq (mean ± SEM). The ratio between the count rate at 24 h and 4 h (paper III) was determined to be 0.97 ± 0.01 (mean ± SEM).

For the group of abnormal vertebrae, the count rate was found to be about twice as high as for the vertebrae in patients with normal bone scintigrams. A rapid increase in the time-count rate curve was seen during the first two hours after injection. Thereafter, a further increase was seen up to 24 h. The count rate at 4 h (paper II) was found to be (221 ± 8.5) 10"5 counts per pixel, s and MBq (mean ± SEM). The ratio

29 between the count rate at 24 h and 4 h (paper III) was determined to be 1.03 ±0.01 (meant SEM).

When studying the individual vertebrae, abnormal count rate (values above x + 2 SD for normal vertebrae) was seen 2 min after injection in about 25% and after 6 min in about 50% of the vertebrae, which were classified as abnormal at 4 h (paper II). Within 60 min almost all of the abnormal vertebrae showed abnormal count rates.

Quantitative bone scintigraphy in patients with prostatic carcinoma after orchiectomy

In patients with normal bone scintigrams the count rate at 4 h was found to be (99 ± 1.5), (98 ± 1.7) and (96 ± 1.4) 10"5 counts per pixel, s and MBq (mean ± SEM), when measured two weeks, two months and six months after operation, respectively. The ratios between the count rates at 24 h and 4 h (paper III) for the corresponding follow-up times were determined to be 0.83 ± 0.01, 0.86 ± 0.01 and 0.86 ± 0.01 (mean ± SEM). No significant change in count rate was seen between the pre- operative and post-operative investigations. The coefficient of variation for all count rate values measured at various times after injection gradually decreased from 22% at 2 min to 16% at 240 min (paper II). When measuring at 240 min the dominating part (14%) was due to the individual variation from patient to patient, while a smaller part (7%) was due to the reproducibility of the measurements (papers I and II).

For the group of normal vertebrae in patients with abnormal scintigrams, the time-count rate curve showed slightly higher values two weeks and

•=•&? • - two months post-operatively than before operation and after six months. The count rates at 4 h (paper II) were found to be (116 ± 4.7), (115 ± 5.0) and (108 ± 4.2) 10'5 counts per pixel, s and MBq (mean ± SEM). The ratios between the counts rate at 24 h and 4 h (paper III) for the corresponding post-operative investigations were determined to be 0.96 ± 0.01, 0.94 ± 0.01 and 0.92 ± 0.01 (mean ± SEM).

For the group of abnormal vertebrae a further increase in the time-count rate curve was seen two weeks after orchiectomy, followed by a decrease to the pre-operative level after two months. The curve continued to decrease but was still above the normal level six months post-operatively. The count rate at 4 h (paper II) was found to be (282 ± 13.3), (210 ± 9.7) and (164 ± 5.8) 10"^ counts per pixel, s and MBq (mean ± SEM), when measured two weeks, two months and six months after operation. The ratios between the count rates at 24 h and 4 h (paper HI) for the corresponding post-operative investigations were determined to be 1.07 ± 0.01, 1.02 ± 0.01 and 0.97 ± 0.01 (mean ± SEM).

When studying the individual abnormal vertebrae the flare phenomenon, with maximum count rate two weeks after operation, was found in over 80% of the vertebrae as a response to the therapy (paper II). After six months the count rate in about 30% of the vertebrae had fallen below the level for abnormal count rate (paper II).

31 Quantitative bone scintigraphy in relation to whole-body counting in a control group and in patients with prostatic carcinoma before and after orchiectomy

In a control group WBR after 24 h was determined to be (37 ± 2.1)% (mean ± SEM). After the same time WBR in patients with normal bone scintigrams (paper IV) was found to be (36 ± 1.6), (33 ± 1.5), (34 ± 1.3) and (35 ± 1.2)% (mean ± SEM), before operation, two weeks, two months and six months after operation. No significant difference in WBR was seen between the two groups of patients or between the pre-operative and post-operative investigations.

In patients with abnormal bone scintigrams, WBR after 24 h was determined to be (58 ± 2.9) before operation and (60 ± 3.3), (54 ±3.2), (51 ± 2.8)% (mean ± SEM), when mesured two weeks, two months and six months after operation. Even if the mean value after two weeks showed a tendency to increase, no significant change was found. A decrease (p < 0.050) in WBR was seen after six months compared with the values before operation. The WBR values were significantly higher (p < 0.001) than the values in the patients with normal bone scintigrams.

In the pre-operative investigation of patients with abnormal bone scintigrams, increased count rate and WBR were seen in 76% of the patients, increased count rate but normal WBR in 14% and normal count rate and normal WBR in 10%. After six months 25% of the patients with increased count rate and WBR pre-operatively showed normal values.

32 Quantitative bone scintigraphy in relation to quantitative computed tomography in patients with prostatic carcinoma at the time of orchiectomy and post-operatively

In patients with normal bone scintigrams BMD was determined to be (122 ± 4.9) and (119 ± 4.4) mg/cm3 (mean ± SEM) at the time of operation and after six months. No significant change in BMD was seen between the two investigations. For a reference material, consisting of 88 men (aged 66-81 years), BMD was determined to be (110 ± 4.3) mg/cm3 (Nilsson: Personal communication, paper V). No significant difference was found between the two groups of patients.

In patients with abnormal bone scintigrams, BMD in vertebrae with normal count rate was determined to be (126 ± 11.6), (103 ± 14.0) and (74 ± 9.6) mg/cm3 (mean ± SEM), at operation and two and six months after operation. A significant decrease (p < 0.005) in BMD was seen after six months, while the MDP uptake remained unchanged.

In patients with abnormal scintigrams, vertebrae with abnormal count rate showed significantly higher (p < 0.001) BMD values than normal vertebrae. At operation and two and six months later, BMD was determined to be (226 ± 15.3), (264 ± 16.3) and (280 ± 23.1) mg/cm3 (mean ± SEM), respectively. No significant change in mean values of BMD for the whole group was seen between the different investigations. For individual patients there were some increases and some decreases. However, regarding the MDP uptake, for patients showing stable or improved clinical status, a significant decrease (p < 0.001) was seen after six months. In one patient showing clinical signs of progression an increase in both MDP uptake and BMD was seen.

33 Discussion

During the study, normal values of 99Tcm-MDP uptake were obtained for the patients with prostatic carcinoma who had normal bone scintigrams. Later, these values were compared with the values from the control group of patients with no signs of metabolic or malignant disease and no history of spine or hip trauma. As no differences were found, the initial values were still regarded as normal. Each patient in the control group was referred for bone scintigraphy only once. Therefore, the reproducibility of QS was determined from repeated measurements on the group of patients who had normal bone scintigrams and was found to be ± 7% (1 SD) (papers I and II). This indicates a very reproducible uptake. The fraction of free pertechnetate in the 99Tcm-MDP was randomly tested and found to be < 5%. Even if various types of 99Tcm complexes may be present (Lind Handeland and Sundrehagen 1987) the uptake and distribution are highly reproducible from preparation to preparation. Because of the reproducibility of QS, this method can be used in investigations of bone-seeking tumours other than prostatic carcinoma.

For abnormal vertebrae, increased count rate was seen in about 25% of cases after two minutes and in about 50% six minutes after injection (paper II). This may be explained by neovascularization in the tumour- involved bone. To facilitate nutrition, the tumour develops numerous new capillary buds (Wirth 1979). The growth is stimulated by angiogenic factors expressed by the tumour, and there seems to be no essential morphological difference between the new capillaries responding to the malignancy and the capillary growth occurring due to physiological neovascularization (Folkman and Klagsbrun 1987).

•-*'c An increased count rate was observed in nearly all of the abnormal vertebrae within one hour. This is important for clinical practice as the count rate in a single scintigiam can be determined and, by comparison with earlier recorded time-count rate curves, abnormal count rate can be revealed within this time period.

For the various groups of vertebrae there was also a difference in the MDP content at 24 h after injection (paper III). The normal vertebrae in the patients with skeletal metastases had a net release of about 3%, while the vertebrae in patients with a normal scintigram had a net release of about 15% from 4 h up to 24 h after injection. Most of the vertebrae which were considered to have normal count rates at 4 h but had an increased 24 h/4 h ratio were found to be abnormal later in the study. This was probably due to involvement of metastases which were impossible to detect at 4 h. The abnormal vertebrae retained about 3% more MDP from 4 h to 24 h after injection. The increase in uptake, despite falling blood levels of MDP, was presumably a consequence of the increased metabolic activity in the tumour-involved bone (Citrin et al. 1975).

The value of extending bone scintigraphy measurements up to 24 h after injection with determination of the 24 h/4 h ratio has been studied earlier as a method of differentiating between benign and malignant changes, especially in single lesions visible already at 4 h (Israel et al. 1985; Guan et al. 1990). A ratio for lesion-to-normal bone uptake was determined. In the present study the advantage of the 24 h/4 h ratio was that it revealed lesions not visible at 4 h. Without quantification of the uptake it may be difficult to select normal ^one as a reference.

35 Using QS, various regions can be selected for quantification. As the spine is the most common site for skeletal metastases in prostatic carcinoma and the highest incidence is in lumbar vertebrae, the region from T10 to L5 was chosen. The WBR measurement, however, has been used as a marker indicating the total extent of skeletal metastases (Castronovo et al. 1985). Ten per cent of the patients with skeletal metastases showed pre- operatively both normal WBR and local gamma camera count rate over the spine from T10-L5 (paper IV). The lesions were small and located outside the region chosen for quantification by gamma camera. However, if the region had been extended by just two more vertebrae, the lesions would have been discovered with QS. Fourteen per cent of the patients with skeletal metastases showed small lesions in the region of quantification, not detectable with WBR. No patient showed lesions large enough to be observed by WBR but located outside the region of quantification. It may be difficult to detect small lesions with WBC. However, WBC may be valuable as a screening method, as all patients, except one, with retention values above 45% in the pre-operative investigation, had evidence of skeletal metastases. Also BMD showed pre- operatively increased values in 58% of the vertebrae with abnormal count rate (paper V). Most of the vertebrae with increased BMD showed evidence of skeletal metastases. Both WBR and BMD may thus be useful measurements in the interpretation of skeletal metastases.

The flare phenomenon with maximum count rate in the time-count rate curve two weeks after operation was seen with QS. This further increase in count rate could be observed in the abnormal vertebrae as early as 2 min after injection (paper II). This is probably due to an increase in local skeletal blood flow. Normally, capillaries are under sympathetic control, and approximately one-third of the arterioles in bone are closed (Sagar

36 et al. 1979). If the sympathetic tone is eliminated, the normally closed vessels open up, permitting a local increase in blood flow and therefore in MDP delivery (Charkes 1980). The flare phenomenon was more pronounced at 24 h than at 4 h (paper III). This indicates an active bone uptake and not only the presence of 99Tcm-MDP in the vascular phase, as the total fraction of the injected activity remaining in the blood at 24 h is very small (Subramanian et al. 1975).

In most of the patients with skeletal metastases, the flare phenomenon was seen with QS but was not apparent with WBC (paper IV). In the patients with solitary lesions, the further increase in count rate over the abnormal vertebrae two weeks post-operatively may be insufficient to be seen in the measurement of WBR. However, the patients with widespread metastatic involvement showed no change in WBR, although the local gamma camera count rate increased significantly, which may be due to redistribution of the MDP content from other parts of the body.

To assess the response to therapy, additional information can be gained if the measurements are extended to six months after operation. At this time the count rate from about 30% of the abnormal vertebrae had fallen below the level for abnormal count rate (paper II). Occasionally, patients are seen in which bone scintigraphy shows areas of improvement and areas of deterioration (Slack et al. 1980). This mixed response suggests that the metastases were composed of a heterogeneous population of cells, which can differ in response to therapy. In these patients showing a mixed response, WBR may be valuable in giving a total measurement. For the abnormal vertebrae the BMD (paper V) varied in different ways after six months. However, in the normal vertebrae of patients with skeletal metastases a decrease was seen in BMD. The general increase in bone

37 resorption at sites outside tumour areas suggests that it is mediated by factors other than those which stimulate osteoblast activity at the sites of tumour deposition (Urwin etal. 1985).

38 Summary and Conclusions

Quantitative bone scintigraphy with "Tcm-MDP showed a reproducibility of ± 7% (1 SD), when determined from four different investigations. It is thus possible to carry out repeated QS in a reproducible manner.

An increased count rate was observed in almost all abnormal vertebrae within one hour after injection of 99Tcm-MDP. For clinical practice it is important that abnormal count rates can be revealed within this time period.

Most of the vertebrae which were considered normal at 4 h, but had an increased ratio between the count rate at 24 h and 4 h, developed into abnormal vertebrae later in the study. With this additional criterion of the 24 h/4 h ratio, the vertebrae can be more accurately defined as normal or abnormal.

Measurements of the 24-hour retention of 99Tcm-MDP in the whole body and measurements of bone mineral density in the thoracic and lumbar vertebrae can provide useful information in the interpretation of skeletal metastases. However, employing quantitative bone scintigraphy on the spine, a common site for skeletal metastases, may give a more accurate evaluation.

In vertebrae with abnormal count rate the flare phenomenon was apparent as early as a few minutes after injection of 99Tcm-MDP. The flare phenomenon was found to be more pronounced at 24 h than at 4 h. This indicates an active bone uptake and not only the presence of 99Tcm-

39 MDP in the vascular phase. From the time before operation up to two months after, the WBR measurements showed a relatively constant total MDP content in the body. The flare phenomenon, with a further increase in gamma camera count rate over the abnormal vertebrae, may be due to a redistribution of MDP from other parts of the body.

The mean count rate obtained with QS over the group of abnormal vertebrae showed a decrease after six months below the pre-operative level in response to the treatment. Also, WBR showed lower values after this time period in patients with abnormal scintigrams. For abnormal vertebrae the BMD varied in different ways, while for normal vertebrae in these patients with metastatic involvement a decrease was seen. This indicates a general increase in bone resorption at sites outside the metastatic area.

40 Acknowledgements

I would like to thank all the patients without whose kind co-operation this thesis would not have been possible.

I also wish to express my sincere gratitude and appreciation to all those who have supported me through these studies, and especially to Bo Lilja, Sören Mattsson, Lars Ahlgren and my husband Lennart Sundkvist.

41 References

Anger HO. Use of a gamma-ray pinhole camera for in vivo studies. Nature 1952:170:200-1.

Batson OV. The function of the vertebral veins and their role in the spread of metastases. Ann Surg 1940:112:138-49.

Bauer GCH, Wendeberg B. External counting of Ca47 and Sr85 in studies of localised skeletal lesions in man. J Bone Joint Surg 1959;41B:558- 80.

Berrettoni BA, Carter JR. Current Concepts Review. Mechanisms of cancer metastasis to bone. J Bone Joint Surg 1986;68A:308-12.

Blau M, Nagler W, Bender MA. Fluorine-18: A new isotope for bone scanning. J Nucl Med 1962;3:332-4.

Brawn PN, Speights VO. The dedifferentiation of metastatic prostate carcinoma. Br J Cancer 1989:59:85-8.

Bruce AW, Trachtenberg J (editors). Adenocarcinoma of the prostate. Berlin, Heidelberg:Springer Verlag, 1987.

Cancer Incidence in Sweden 1987. National Board of Health and Welfare. The Cancer Registry.

42 Cassen B, Curtis L, Reed C, Libby R. Instrumentation for I131 use in medical studies. Nucleonics 1951;9:46-50.

Castronovo FP, Jr, McKusick KA, Dann J, Prout GR, Jr, Strauss HW. A simplified technique for quantifying 24-h whole body retention of 99mTc-labeled methylene diphosphonate (MDP). Int J Nucl Med Biol 1985;12:209-14.

Cederquist E. Clinical application of whole body counting of 85Sr and 47Ca in patients with and without widespread malignant skeletal disease. Acta Radiol 1964;Suppl 232.

Chambers TJ. The pathobiology of the osteoclast. J Clin Pathol 1985;38:241-52.

Charkes ND. Skeletal blood flow: Implications for bone-scan interpretation. J Nucl Med 1980;21:91-8.

Charkes ND, Sklaroff DM, Bierly J. Detection of metastatic cancer to bone by scintiscanning with strontium 87m. Am J Roentgenol Radium Ther Nucl Med 1964;91:1121-7.

Chiewitz O, Hevesy G de. Radioactive indicators in the study of phosphorus metabolism in rats. Nature 1935; 136:754-5.

Citrin DL, Bessent RG, McGinley E, Gordon D. Dynamic studies with "mTc-HEDP in normal subjects and in patients with bone tumors. J Nucl Med 1975:16:886-90.

43 Citrin DL, Bessent RG, Tuohy JB, Greig WR, Blumgart LH. Quantitative bone scanning: A method for assessing response of bone metastases to treatment. Lancet 1974; 1:1132-3.

Come D, Timmons JH, Bauman JM, Thompson IM. Efficacy of follow- up bone scans in carcinoma of the prostate. Cancer 1988;61:2453-4. dimming J, Hacking N, Fairhurst J, Ackery D, Jenkins JD. Distribution of bony metastases in prostatic carcinoma. Br J Urol 1990;66:411-4.

Dann J, Castronovo FP, Jr, McKusick KA, Griffin PP, Strauss HW, Prout GR, Jr. Total bone uptake in management of metastatic carcinoma of the prostate. J Urol 1987; 137:444-8.

Dodds PR, Caride VJ, Lytton B. The role of vertebral veins in the dissemination of prostatic carcinoma. J Urol 1981;126:753-5.

Ewing J. Neoplastic diseases. A treatise on tumors. 3rd ed. Philadelphia:WB Saunders Co., 1928.

Fidler IJ. Critical factors in the biology of human cancer metastasis: Twenty-eighth G.H.A. Clowes Memorial Award Lecture. Cancer Res 1990;50:6130-8.

Fogelman I, Bessent RG, Turner JG, Citrin DL, Boyle IT, Greig WR. The use of whole-body retention of Tc-99m diphosphonate in the diagnosis of metabolic bone disease. J Nucl Med 1978;19:270-5.

Folkman J, Klagsbrun M. Angiogenic factors. Science 1987;235:442-7.

44 Francis MD. The inhibition of calcium hydroxyapatite crystal growth by polyphosphonates and polyphosphates. Calcif Tissue Res 1969;3:151- 62.

Franck JL, Bouteiller G, Fauchier Ch, Arlet J. How and what kind of bone is made by metastatic cells of prostatic cancer. Med Hypotheses 1982;9:587-97.

Galasko CSB. The mechanism of uptake of bone-seeking isotopes by skeletal metastases. IAEA 1977;2:125-36.

Genant HK, Block JE, Steiger P, Glueer CC, Smith R. Quantitative computed tomography in assessment of osteoporosis. Sem Nucl Med 1987;17:316-33.

Goodwin PN. Methodologies for the measurement of bone density and their precision and accuracy. Sem Nucl Med 1987;17:293-304.

Guan SI, Chu LS, Hwang WS, Chen WL. The differential diagnosis of benign and malignant bony lesions in bone scanning. Using the ratio of radiouptake at different times. Clin Nucl Med 1990;l5:424-7.

Hardy JG, Kulatilake AE, Wastie ML. An index for monitoring bone metastases from carcinoma of the prostate. Br J Radiol 1980;53:869- 73. Huggins C, Hodges CV. Studies on prostatic cancer. I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1941;l:293-7.

45 Israel O, Front D, Frenkel A, Kleinhaus U. 24-Hour/4-hour ratio of technetium-99m me thy lene diphosphonate uptake in patients with bone metastases and degenerative bone changes. J Nucl Med 1985;26:237-40.

Jacobs SC. Spread of prostatic cancer to bone. Urology 1983;21:337-44.

Johansson SAE, Skanse B. A photographic method of determining the distribution of radioactive material in vivo. Acta Radiol 1953;39:317-22.

Joliot F, Curie I. Artificial production of a new kind of radio-element. Nature 1934;133:201-2.

Jones AG, Francis MD, Davis MA. Bone scanning: Radionuclidic reaction mechanisms. Sem Nucl Med 1976;6:3-18.

Lind Handeland ÅM, Sundrehagen E. Determination of the oxidation state of 99mTc in complexes with MDP and DTPA. Appl Radiat Isot 1987;38:479-84.

Milch RA, Changus GW. Response of bone to tumor invasion. Cancer 1956;9:340-51.

Mirra JM, Picci P, Gold RH. Bone tumors. Clinical, radiologic, and pathologic correlations. Philadelphia: Lea & Febiger, 1989.

Morgan JWM, Adcock KA, Donohue RE. Distribution of skeletal metastases in prostatic and iung cancer. Urology 1990;36:31-4.

46 Mulry WC, Dudley HC. Studies of radiogallium as a diagnostic agent in bone tumors. J Lab Clin Med 1951;37:239-52.

Nicolson GL. Organ specificity of tumor metastasis: role of preferential adhesion, invasion and growth of malignant cells at specific secondary si*es. Cancer Metastasis Rev 1988;7:143-88.

Nilsson M, Johneli O, Jonsson K, Redlund-Johnell I. Quantitative computed tomography in measurement of vertebral trabecular bone mass. Acta Radiol 1988;29:719-25.

Osmond JD, III, Pendergrass HP, Potsaid MS. Accuracy of "mTc- diphosphonate bone scans and roentgenograms in the detection of prostate, breast and lung carcinoma metastases. AJR 1975;125:972-7.

Paget S. Distribution of secondary growths in cancer of the breast. Lancet 1889;l:571-3.

Paterson AHG. Bone metastases in breast cancer, prostate cancer and myeloma. Bone 1987;8:Suppl 1:S17-S22.

Perkel VS, Mohan S, Herring SJ, Baylink DJ, Linkhart TA. Human prostatic cancer cells, PC3, elaborate mitogenic activity which selectively stimulates human bone cells. Cancer Res 1990;50:6902-7.

Pollen JJ, Gerber K, Ashburn WL, Schmidt JD. The value of nuclear bone imaging in advanced prostatic cancer. J Urol 1981;125:222-3.

47 Pollen JJ, Shiaer WJ. Osteoblastic response to successful treatment of metastatic cancer of the prostate. AJR 1979;132:927-31.

Pöste G, Fidler IJ. The pathogenesis of cancer metastasis. Nature 1980;283:139-46.

Price JE. The biology of metastatic breast cancer. Cancer 1990;66:1313- 20.

Reinbold WD, Genant HK, Reiser UJ, Harris ST, Ettinger B. Bone mineral content in early-postmenopausal and postmenopausal osteoporotic women: Comparison cf measurement methods. Radiology 1986; 160:469-78.

Sågar VV, Piccone JM, Charkes ND. Studies of skeletal tracer kinetics. III. Tc-99m(Sn)methylenediphosphonate uptake in the canine tibia as a function of blood flow. J Nucl Med 1979;20:1257-61.

Schaberg J, Gainor BJ. A profile of metastatic carcinoma of the spine. Spine 1985;10:19-20.

Slack NH, Karr JP, Chu TM, Murphy GP, et al. An assessment of bone scans for monitoring osseous metastases in patients being treated for prostate carcinoma. Prostate 1980; 1:259-70.

Smith JA. Hormonal therapy of prostate cancer: Current concepts and future prospects. Clin Ther 1988; 10:281-6.

48 Snedecor GW, Cochran WG. Statistical Methods. 7th ed. Ames, Iowa, USA: The Iowa State University Press, 1980.

Stepån JJ, Lachman M, Zverina J, Pacovsky V, Baylink DJ. Castrated men exhibit bone loss: Effect of calcitonin treatment on biochemical indices of bone remodeling. J Clin Endocrinol 1989;69:523-7.

Styles C. The distribution of bone metastases as shown on isotope scanning: Proposed modes of spread. Australas Radiol 1989;33:226- 8.

Subramanian G, Blair RJ, Kallfelz EA, Thomas FD, McAfee JC. "mTc- MDP (methylene diphosphonate): A superior agent for skeletal imaging. J Nucl Med 1973; 14:640.

Subramanian G, McAfee JG. A new complex of "mTc for skeletal imaging. Radiology 1971;99:192-6.

Subramanian G, McAfee JG, Blair RJ, Kallfelz FA, Thomas FD. Technetium-99m-methylene diphosphonate — A superior agent for skeletal imaging: Comparison with other technetium complexes. J Nucl Med 1975;16:744-55.

Teitelbaum SL, Bullough PG. The pathophysiology of bone and joint disease. Am J Pathol 1979;96:283-354.

Thrall JH, Burton IE. Skeletal metastases. Radiol Clin North Am 1987;25:1155-70.

49 Treadwell A de G, Low-Beer BVA, Friedell HL, Lawrence JH. Metabolic studies on neoplasm of bone with the aid of radioactive strontium. Am J Med Sci 1942;204:521-30.

Urwin GH, Percival RC, Harris S, Beneton MNC, Williams JL, Kanis JA. Generalised increase in bone resorption in carcinoma of the prostate. BrJUroll985;57:721-3.

Weiss L. Principles of metastasis. New York: Academic Press, Inc., 1985.

Willis RA. The spread of tumours in the human body. 3rd ed. London: Butterworths, 1973.

Wirth CR. Metastatic bone cancer. Curr Probl Cancer 1979;3:l-36.