Guidelines for the Use of Imaging in the Management of Myeloma

guideline

Guidelines for the use of imaging in the management of myeloma

Shirley D’Sa1,* Niels Abildgaard2, Jane Tighe3,* Penny Shaw4, Margaret Hall-Craggs4, 1Department of Haematology, University College Hospital, London, UK, 2Department of Haematology, Odense University Hospital, Odense, Denmark, 3Department of Haematology, Aberdeen Royal Inﬁrmary, Aberdeen and 4Department of Imaging, University College London, London, UK

for Standards in Haematology (BCSH) (Samson, 2001). Summary Reference to the use of imaging was confined to the standard In 2001, reference to the use of imaging in the British Committee use of plain X-rays in the diagnostic skeletal survey and for Standards in Haematology guidelines for the diagnosis and emergency use of computed tomography (CT) and magnetic management of myeloma was confined to the standard use of resonance (MR) imaging in the setting of cord compression. plain X-rays in the diagnostic skeletal survey and emergency use Since then, there has been a steady rise in interest in the use of of computed tomography (CT) and magnetic resonance (MR) various imaging techniques in the management of myeloma, imaging in the setting of cord compression. Since then, there has not least due to the apparently successful application of been a steady rise in interest in the use of various imaging techniques, such as positron emission tomography (PET) techniques in the management of myeloma. The purpose of scanning, and developments in MR technology. It was imaging in the management of myeloma includes the assessment therefore felt appropriate to develop stand-alone guidelines of the extent and severity of the disease at presentation, the for the use of imaging in the management of myeloma. identification and characterisation of complications, and the Myeloma is a plasma cell tumour with an annual incidence assessment of response to therapy. Plain radiography, CT, and in the UK and Scandinavian countries of approximately 50 per MR imaging are generally established examination techniques in million and a median age at presentation of about 70 years myeloma whilst positron emission tomography (PET) and (Turesson et al, 1984; Hjorth et al, 1992; Office for National 99Technetium sestamibi (MIBI) imaging are promising newer Statistics, 2001; Phekoo et al, 2004). Clinical presentation is scanning techniques under current evaluation. These stand- varied. Common presentations of myeloma include bone pain, alone imaging guidelines discuss recommendations for the use recurrent or persistent infection, anaemia, renal impairment, of each modality of imaging at diagnosis and in the follow up of symptoms of hypercalcaemia or a combination of these. Some patients with myeloma. patients are asymptomatic, abnormalities being identified on blood tests carried out for other clinical reasons. In the Keywords: guidelines, imaging, myeloma, plasmacytoma. majority of patients, the disease is characterised by plasma cell infiltration of the bone marrow, osteolytic bone lesions and the presence of a monoclonal protein in the serum or urine. In a proportion of these, however, one or more features may be 1. Introduction: Context, methodology absent, giving rise to a diagnosis of solitary plasmacytoma, asymptomatic myeloma or non-secretory myeloma. 1.1 Context In 2001, guidelines for the diagnosis and management of 1.2 Approach to imaging in the management of myeloma myeloma were prepared by the Guidelines Working Group of the UK Myeloma Forum on behalf of the British Committee The purpose of imaging in the management of myeloma includes the assessment of the extent and severity of the disease at presentation, the identification and characterisation of complications, and the assessment of response to therapy. Correspondence: Dr Shirley D’Sa Department of Haematology, Plain radiography, CT, and MR imaging are generally University College Hospital, 1st Floor Central, 250 Euston Road, established examination techniques in myeloma whilst PET London NW1 2PJ, UK. E-mail: [email protected] and 99Technetium -2-methoxy-isobutyl-isonitrile (MIBI) ima- *On behalf of the UK Myeloma Forum Guidelines Working Group, ging are promising newer scanning techniques under current On behalf of the Nordic Myeloma Study Group, and On behalf of evaluation. With the advent and development of these newer the Royal College of Radiologists techniques, it is tempting to use them in the follow up of

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 doi:10.1111/j.1365-2141.2007.06491.x Guideline challenging situations, such as patients with non-secretory Table I. Levels of Evidence. disease (especially those who have non-contributory serum Ia free light chain (FLC) measurements). Various studies of Evidence obtained from meta-ana- myeloma patients have demonstrated that more sophisticated lysis of randomised controlled imaging techniques, such as CT and MR, demonstrate the trials presence of a higher tumour burden than techniques such as 1b Evidence obtained from at least plain radiography (Laroche et al, 1996; Dimopoulos et al, one randomised controlled trial 2000; Mulligan, 2000). The important consideration is whether IIa Evidence obtained from at least one well-designed, non-random- the increased costs involved and/or greater exposure to ised study, including phase II trials radiation is justifiable in terms of improved clinical outcome. and case-control studies If clinical outcome is measured by improved survival, very few IIb Evidence obtained from at least imaging techniques can claim to achieve this goal, as it is the one other type of well-designed, treatment intervention itself that produces this effect. If, quasi-experimental study, i.e. however successful clinical outcome is measured by an studies without planned interven- improvement in quality of life measures as well, then it can tion, including observational be argued that the greater sensitivity and specificity of studies particular imaging techniques may contribute to this goal III Evidence obtained from well-de- (Hunink & Krestin, 2002). As a result of rapid advancements signed, non-experimental descrip- in diagnostic imaging techniques, comprehensive evaluation of tive studies. Evidence obtained from meta-analysis or randomised new techniques prior to their implementation is difficult controlled trials or phase II studies (Fryback & Thornbury, 1991). In practice, newer imaging which is published only in abstract techniques are often implemented on the basis of subjective form experience with a limited number of cases and subjective IV Evidence obtained from expert expectations of their usefulness (Hunink & Krestin, 2002). committee reports or opinions and/or clinical experience of respected authorities 1.3 Methods

These guidelines have been compiled by the Imaging Subgroup C. Detailed imaging technical protocols are not included; they of the Guidelines Working Group of the UK Myeloma Forum are beyond the scope of this document. and Nordic Myeloma Study Group on behalf of the BCSH, including two UK haematologists, one Danish haematologist and two UK radiologists with specialist experience of 1.4 Minimising radiation exposure oncological radiology. Their production involved the following At all times during the investigation and follow up of a patient steps: with myeloma, the potential usefulness of a proposed imaging • Review of key literature including Cochrane database, investigation should be carefully considered. A useful investi- Medline and Internet searches updated to 28 February 2006. gation is one in which the result, positive or negative, will alter • Consultation with representatives of other specialties, management. A significant number of radiological investiga- including surgeons and specialists in nuclear medicine. tions do not fulfil these aims, may increase waiting times, waste • Recommendations made based on the literature review and limited resources and add unnecessarily to patient radiation consensus of expert opinion. exposure (Royal College of Radiologists, 2003). The extent and • Completion date 31st March 2006. Table II. Grades of Recommendation. • Adherence to the BSCH procedure for guidelines development (http://www.bcshguidelines.com/process1.asp). Grade A Recommendation based on at least Evidence level Ia, Ib These guidelines are intended to set out the key areas of one randomised controlled trial of strategy in the effective use of imaging in the management of good quality and consistency myeloma. Levels of evidence and grades of recommendation addressing specific recommenda- are set out in Table I and II. However, there is a distinct lack of tion Grade B Recommendation based on well randomised controlled trials in the use of imaging in myeloma, Evidence level IIa, IIb, III conducted studies but no rand- and the relationship between diagnostic imaging information omised controlled trials on the to- and patient outcomes is difficult to demonstrate due to the pic of recommendation multiple steps and confounding factors that intervene, inclu- Grade C Evidence from expert committee ding individual patient performance status, prognostic factors Evidence level IV reports and/or clinical experiences and treatment modalities undertaken (Eisenberg et al, 1989). of respected authorities Thus all the recommendations made are of grade B and

ª 2007 British Society for Haematology 50 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline frequency of imaging investigations with inherent risks due to • Accurate clinical information should be provided to the radiation dose must be justified in terms of likely clinical Radiology Department when the imaging request is made benefit; it is important to avoid unnecessary risk due to X-ray to ensure that the right imaging technique is performed exposure (Berrington de Gonzalez & Darby, 2004). Typical at the right time, specifying whether the request pertains effective doses for common diagnostic radiology procedures to a diagnostic work up or investigation of new symptoms are shown in Table III. These values, compiled by the UK in a patient who is known to have myeloma, including National Radiological Protection Board (Hart & Wall, 2002) details of prior therapy. are based on the weighted sum of the doses to a number of body tissues, where the weighting factor for each tissue 1.5 The patient’s perspective and the multidisciplinary depends on its relative sensitivity to radiation-induced cancer team or severe hereditary defects. Although the skeletal survey is widely accepted as a standard and technically straightforward imaging technique in myeloma Recommendations patients, the impact that this and other imaging investigations • Robust systems should be in place for timely reporting may have on a patient who has skeletal fractures, deformities and secure storage of X-ray films in order to avoid or lytic lesions should not be underestimated. Adequate repeating investigations that have already been done. provision should be made for analgesia or sedation to be administered prior to the investigation being carried out. Following this initial investigation, any further imaging should ideally be discussed in the setting of a multidisciplinary Table III. Typical effective doses of common diagnostic radiological team meeting that includes an experienced radiologist and procedures. clinical oncologist. When necessary, the input of a specialist Approximate orthopaedic surgeon may be needed to advise on the need for Equivalent equivalent period surgical stabilisation of bones vulnerable to fracture. Specialist Diagnostic Typical effective numbers of of natural back- services may be available locally or in a neighbouring hospital. Procedure dose (mSv) chest X-rays ground radiation There should be clear policies and protocols for access to these services. X-ray examinations Limbs & joints (except hips) <0Æ01 <0Æ5<1Æ5d 2. Use of imaging at diagnosis Chest (single PA film) 0Æ02 1 3 d At diagnosis, it is important to establish the extent of disease to Skull permit accurate staging. This is achieved by staging investiga- 0Æ06 3 9 d tions, including a bone marrow biopsy to assess degree of Thoracic spine infiltration by plasma cells, quantification of the amount of 0Æ7 35 4 months monoclonal protein secreted in the serum (paraprotein) Lumbar spine and/or urine (Bence–Jones protein) and imaging the skeleton 1Æ0 50 5 months for evidence of osteolytic lesions. This section discusses the Hip recommended use of each imaging modality in turn; a 0Æ4 20 2 months summary of the recommendations is shown in Fig. 1. Pelvis 0Æ7 35 4 months Skeletal survey 2.1 Skeletal survey 2Æ5 125 12 months CT head The skeletal survey remains the standard method for radiolo- 2Æ0 100 10 months gical screening at diagnosis; there is a clear association between CT chest the extent of disease (in terms of the number of lytic lesions at 8 400 3Æ6 years presentation) and tumour load at diagnosis (Durie & Salmon, CT abdomen 1975). Plain radiography is universally available, allows large or pelvis 10 500 4Æ5 years areas of the skeleton to be visualised and may identify long Radionuclide studies bones at risk of impending fracture. Almost 80% of patients Bone scan with myeloma will have radiological evidence of skeletal (Tc-99m) 4 200 1Æ8 years PET whole involvement on the skeletal survey, most commonly affecting body (F-18 FDG)10 500 4Æ5 years the following sites: vertebrae in 66% of patients, ribs in 45%, PET-CT skull in 40%, shoulder in 40%, pelvis in 30% and long bones in 20 1000 9 years 25% (Collins, 1998). However, radiologically detectable lesions distal to the elbows and knees are exceptional (Healy &

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 51 Guideline

Soft tissue mass Suspected myeloma or Suspected spinal cord plasmacytoma compression

CT scan and Radiological skeletal survey Urgent MR/ CT scan consider biopsy and appropriate clinical management

Lytic lesions present?

No Yes Ambiguous findings or unexplained bone pain

Multiple Apparently Other evidence of end MRI/CT/PET solitary lesion organ damage (renal scan impairment / high calcium / BM failure

Risk of fracture? Risk of fracture?

No Yes No Yes No

Urgent orthopaedic Generalised osteopenia review & appropriate clinical management

MRI whole spine to exclude occult disease

Occult lesions identified and Normal confirmed as myelomatous

Systemic Local R/T to index lesion Local Systemic Observation therapy & systemic therapy Radiotherapy therapy and consider bisphosphonates

Fig 1. Algorithm of suggested recommendations. CT, computed tomography; MR(I) magnetice resonance (imaging); PET positron emission tomography; R/T, radiotherapy.

Armstrong, 1998). A scoring system based on clinical and (at least one lytic lesion) are at high risk of progression with a radiological findings (Table IV) has been devised to predict the median time to progression of 8 months (Wisloff et al, 1991; likelihood of fracture in long bones affected by metastatic Dimopoulos et al, 1993). Note that, in the new Interna- disease, suggesting that patients with a high score benefit from tional Classification (International Myeloma Working Group, internal fixation, whereas those with a lower score can receive 2003), patients with bone disease are classified as ‘‘sympto- radiotherapy alone (Mirels, 1989). Patients who are clinically matic’’ and requiring treatment even in the absence of clinical asymptomatic but have radiological evidence of bone disease symptoms.

ª 2007 British Society for Haematology 52 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline

Table IV. Scoring system for diagnosing impending pathological periphery of the appendicular skeleton and proceeds centrally, fractures. reaching adult proportions around the age of 20 years. Score However, there is continued replacement of red marrow throughout adult life, and this may occur in a diffuse or focal Variable 123 pattern (Ricci et al, 1990). Typical myeloma lesions have a low signal intensity on Site Upper limb Lower limb Peritrochanteric Pain Mild Moderate Functional T1-weighted images and a high signal intensity on T2-weighted Lesion Blastic Mixed Lytic and STIR (short time inversion recovery) images (Libshitz Size <1/3 diameter 1/3–2/3 >2/3 diameter et al, 1992) and generally show enhancement on gadolinium- diameter enhanced images. Four MR imaging patterns of marrow involvement in myeloma have been recognised. A normal marrow appearance is present at diagnosis in 50–75% of untreated Durie–Salmon stage I myeloma and in 20% of A major disadvantage of plain radiography is the relatively low untreated Durie–Salmon stage III disease (Moulopoulos et al, sensitivity, only demonstrating lytic disease when at least 30% of 1992; Stabler et al, 1996). Other marrow appearances of trabecular bone substance has been lost (Snapper & Khan, 1971). untreated disease include the focal pattern, diffuse pattern and In addition, this technique also provides an inadequate variegated appearance. Low tumour burden is usually associ- assessment of generalised osteopenia (Scane et al, 1994) and ated with a normal MR pattern, but a high tumour burden is has low specificity. Plain radiographs may show non-specific usually suspected when there is diffuse hypointense change on abnormalities that require further characterisation by other T1-weighted images, diffuse hyperintensity on T2-weighted techniques, such as computed tomography, and certain parts of images and enhancement with gadolinium injection. Thus the the skeleton are difficult to assess accurately, such as the sternum main methodological consideration with MR imaging is the and ribs and scapulae. The skeletal survey has been found to lack of specificity of the findings. Focal or diffuse changes may provide more accurate staging of bone involvement in myeloma exist at diagnosis, may be variations of the norm, or reflect an than limited MR imaging (Lecouvet et al, 1999). alternative pathological or physiological process such as iron loading (Isoda et al, 2001), amyloid deposition (Baur et al, 1998) or reactive marrow hyperplasia. Recommendations The prognostic value of bone marrow MR images has been • As part of the staging procedure of newly diagnosed evaluated in different studies. Patients with a single lytic lesion myeloma, the skeletal survey should include a postero- on plain radiography, who are found to have further lesions on anterior (PA) view of the chest, antero-posterior (AP) and MR imaging have a significantly shorter interval before they lateral views of the cervical spine (including an open demonstrate disease progression and the need for therapy mouth view), thoracic spine, lumbar spine, humeri and compared to those with a normal MR study (Wisloff et al, femora, AP and lateral views of the skull and AP view of 1991; Moulopoulos et al, 1995; Van de Berg et al, 1996; Weber the pelvis. In addition, any symptomatic areas should be et al, 1997). MR can detect bone marrow infiltration in 29– specifically visualised with appropriate views (Grade C 50% of patients with Durie–Salmon stage I disease and recommendation; level IV evidence). negative plain radiographs (Baur-Melnyk et al, 2005). Patients with advanced disease who have normal MR findings and receive conventional dose chemotherapy respond better to 2.2 MR imaging treatment and have a longer survival compared to those with In the event of suspected cord compression (refer to section diffuse or focal abnormalities on MR imaging (Lecouvet et al, 3.1), MR imaging is the technique of choice (Joffe et al, 1988). 1998a). The finding of more than 10 lesions on spinal MR The potential for its use in other contexts is considerable. images in patients with advanced stage myeloma was associ- Owing to its ability to visualise large volumes of bone marrow ated with a 6–10-fold higher risk of fracture than patients who without radiation exposure, magnetic resonance (MR) imaging had normal appearances or fewer than 10 lesions on MR has become a favoured imaging method for evaluating disease imaging (Lecouvet et al, 1997). MR imaging does not predict within the bone marrow. Numerous MR imaging techniques the level of fracture (Lecouvet et al, 1998b). have been developed to aid the assessment of bone marrow in The pattern of MR bone marrow involvement in myeloma haematological malignancies (Moulopoulos & Dimopoulos, also has prognostic significance, with both focal and diffuse 1997). Depending on the technical specification and set up of patterns being associated with a higher tumour burden the MR scanner, this technique can take a varying length of (Moulopoulos et al, 1992, 2005; Carlson et al, 1995; Lecouvet time to perform. The MR appearance depends on the presence et al, 1998a; Stabler et al, 1996). MR consistently demonstrates and relative proportions of trabecular bone; fat and water and more numerous spinal and pelvic marrow lesions than changes with age, particularly during the second decade. corresponding conventional radiographs in patients with Normal fatty replacement of red marrow begins in the marrow involvement (Ludwig et al, 1987; Fruehwald et al,

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 53 Guideline

1988; Tertti et al, 1995; Lecouvet et al, 1999). Given the remain a problem, and current research is ongoing to significance of detecting occult disease on the rate of progres- improve on the sensitivity and specificity of the technique for sion of the disease, the use of MR imaging in selected patients reliable and reproducible detection of malignant disease. The may contribute to the decision-making process. The other potential application of this technique to myeloma patients is setting in which MR imaging has significance is in the staging attractive, with inclusion of sites such as the sternum, skull of apparently solitary bone plasmacytoma. MR screening of the and ribs, which are usually excluded from standard MR spine and pelvis reveals lesions that are radiographically occult imaging protocols. Studies comparing MR and PET-CT in up to 80% of patients. Such patients, if treated with local imaging are underway, but MR imaging, if equivalent in radiotherapy alone, show earlier progression to systemic sensitivity and specificity, would prove to be advantageous in disease than those with a negative MR imaging survey at terms of the lack of radiation exposure. Current disadvan- diagnosis (Moulopoulos et al, 1993; Soutar et al, 2004). tages of whole body STIR MR imaging include lack of Approximately 1–2% of patients have non-secretory myel- widespread availability of specialised scanning facilities, as oma, i.e. they do not have a detectable serum or urine well as how to manage the detection of unexpected and monoclonal protein. A proportion of these patients may in fact possibly irrelevant abnormalities. In addition, the clinical have detectable free light chains in their serum, which can be experience of this imaging modality as a screening tool in serially monitored, but most patients are usually followed up myeloma patients is yet to be established. by sequential bone marrow biopsies. The use of MR imaging may be a useful tool for monitoring of non-secretory patients Recommendations in the future, once appropriate studies that examine serial MR • Urgent MR imaging is the diagnostic procedure of choice imaging in parallel with bone marrow histology have been to assess suspected cord compression in myeloma performed. As yet, there is insufficient evidence to recommend patients even in the absence of vertebral collapse (Grade the use of MR imaging in the follow up of non-secretory B recommendation; level IIB evidence). patients. • MR imaging of the whole spine should be performed in Technological advances in MR imaging is leading to an addition to the skeletal survey as part of staging in all increase in the potential use of this imaging modality in the patients with an apparently solitary plasmacytoma of staging of myeloma. In fact, MR imaging contributes to the bone irrespective of site of index lesion (Grade B Durie–Salmon Plus staging system, in which the staging suffix recommendation; level IIB evidence). B is applied in the event that additional lytic lesions on • MR imaging should be used to clarify the significance of imaging studies such as plain radiographs or PET scanning ambiguous CT findings, as these two imaging techniques and/or mild, moderate or severe diffuse spinal involvement on can give complementary information (Grade C recom- MR imaging are present (Durie et al, 2003). In this system, the mendation; level IV evidence). number of lytic lesions is quantified as is the degree of diffuse marrow involvement and their presence can lead to an Recommendation on use of MR imaging in follow up of upstaging of the classical Durie-Salmon staging system and non-secretory disease removed. may alter management (Mulligan, 2005). In another study of the redistribution of contrast agent from the interstitial to the 2.3 Computed tomography intravascular compartment in the bony pelvis and comparing this with bone marrow biopsy of the same region, dynamic Computed tomography (CT) provides an excellent tool in contrast-enhanced MR has been shown to correlate with vessel selected situations. Thus far, owing to the high levels of density and paraprotein level (Nosas-Garcia et al, 2005). These radiation exposure, CT cannot be used for screening purposes, imaging correlates, once confirmed by further study, may although low dose CT techniques are being developed as a contribute to the assessment of newly diagnosed myeloma possible alternative to plain radiography (Horger et al, 2005). patients in the future. Conventional CT has higher sensitivity than plain radiographs Recently, whole body STIR MR imaging has gained for detecting small lytic lesions and can accurately depict the popularity for the detection of occult malignant disease in presence and extent of associated extra-osseous extension of the skeleton (Eustace et al, 1997, 1999; Walker & Eustace, the disease. It can provide a high predictive value in the 2001; Hargaden et al, 2003; Kavanagh et al, 2003). With clarification of suspicious areas on plain films or symptomatic ongoing technical advancements, such as the moving table, areas that do not show abnormalities on plain films and use of multicoil elements and sophisticated image processing provides important information about vital organ (e.g. cord) technology, this technique is becoming more feasible and compression. In addition, CT can be used to assess areas of the quicker to perform (Schmidt et al, 2005). MR sequences that skeleton that are of clinical concern or that cannot be are employed include STIR, T1-weighting with or without accurately visualised by plain radiography, e.g. scapulae, CT contrast enhancement and T2-weighting. Although the use of of chest for rib and sternal lesions. CT has the practical more sequences improves the specificity of the images, this advantage of guiding needle biopsy for histological diagnosis increases acquisition time. In addition, false positive findings (Kyle, 1985) and forms the basis for radiotherapy and surgery

ª 2007 British Society for Haematology 54 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline planning (Walker et al, 2003). CT may also identify bone 2.5 DEXA scanning destruction in cases where MR is negative, and hence may provide complementary imaging information (Lecouvet et al, Dual energy X-ray absorptiometry (DEXA) scanning has been 2001). declared by the World Health Organisation to be the gold standard procedure for diagnosing osteoporosis (Kanis & Gluer, 2000), for fracture risk assessment in benign osteoporosis Recommendations (Marshall et al, 1996) and for longitudinal monitoring to assess • Urgent CT may be used to establish the presence of changes of bone mineral density (BMD) over time. The suspected cord compression in cases where MR imaging technique, which involves assessment of BMD in the lumbar is unavailable, impossible due to patient intolerance or spine, hip and distal radius, is a quick, non-invasive investigation contraindicated e.g. intraorbital metallic foreign bodies that uses a small dose (<1 lSV) of radiation (Lewis et al, 1994). or cardiac pacemakers (Grade B recommendation; level However, the estimated BMD can be influenced by spondylosis III evidence). and spinal osteophytes (Masud et al, 1993) and the presence of • CT of the spine may be considered to clarify the presence vertebral collapse, giving rise to methodological difficulties in or absence of bone destruction in cases of clinical concern using DEXA scanning in myeloma patients. Newer scanners where MR is negative (Grade B recommendation; level III are more sophisticated and can estimate vertebral BMD from a evidence). lateral view with the patient in a supine position, with improved • CT should be used to clarify the significance of ambiguous visualisation of the vertebral column. A recent study has demo- plain radiographic findings, such as equivocal lytic lesions, nstrated that reduced lumbar spine BMD at diagnosis is corre- especially in parts of the skeleton that are difficult to lated with an increased risk of early vertebral collapses visualise on plain radiographs, such as ribs, sternum and (Abildgaard et al, 2004), but this remains to be confirmed by scapulae(Grade B recommendation; level III evidence). further studies. With regard to using DEXA to distinguish • CT may identify lesions that are negative on plain between benign osteoporosis and myeloma-induced osteopor- radiography, and should be considered in patients who osis, there are no reliable data available at present. A large remain symptomatic despite having no evidence of deviation in BMD between individual vertebrae may indicate osteolysis on the skeletal survey (Grade B recommenda- local osteolysis due to myeloma, but there are no published data tion; level III evidence). to confirm this. It is important to remember that osteopenia in • CT is indicated to delineate the nature and extent of soft myeloma is not necessarily universal, but often restricted to tissue disease, and where appropriate, tissue biopsy may haematopoietically active bone. Though examination of the be guided by CT scanning (Grade B recommendation; distal radius would be easy to perform in myeloma patients, with level IIB evidence). reduced chance of confounding abnormalities, it would not allow examination of ‘‘myeloma-induced’’ BMD changes. 2.4 Bone scintigraphy

Bone scintigraphy is a radionuclide technique, using a Recommendation technetium 99-phosphorus compound, which is incorporated • Routine assessment of bone mineral density cannot be into bone areas of increased mineralisation. Owing to the recommended, owing to the methodological difficulties of excessive bone resorption/lack of osteoblastic activity that the technique and the universal use of bisphosphonates in characterises the lytic lesions of myeloma, bone scintigraphy all symptomatic myeloma patients. has lower sensitivity than conventional radiographs in the detection of osteolytic lesions in myeloma (Bataille et al, 1982; Ludwig et al, 1982; Nilsson-Ehle et al, 1982; Tamir et al, 1983). 2.6 PET and MIBI Scanning However, bone scintigraphy may be helpful in evaluating areas of the skeleton that are not well demonstrated on plain Positron emission tomography (PET) is a tomographic radiography, such as the ribs and sternum (Ludwig et al, 1982; nuclear imaging procedure that uses positrons as radiolabels Nilsson-Ehle et al, 1982; Otsuka et al, 1993). In reality though, and positron–electron annihilation reaction gamma rays to the bone scintigraph is typically normal in myeloma, or may locate the radiolabels. A low dose of a radiopharmaceutical 18 show areas of decreased uptake representing bone destruction labelled with a positron emitter, such as F is injected into and replacement of bone by myeloma cells without an the patient, who is scanned by a tomographic system; the osteoblastic reaction, and is therefore of little value in scan takes 10–40 min to perform. The most widely used 18 evaluating myeloma lesions. labelled radiopharmaceutical is Fluorine-fluoro-deoxyglucose (FDG), and numerous studies of this imaging technique applied to patients with plasma cell dyscrasias have been Recommendation published. These studies have ranged from single case • Bone scintigraphy has no place in the routine staging of reports to prospective studies of up to 66 patients, in which myeloma. PET findings were compared to other simultaneously

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 55 Guideline performed imaging techniques such as plain radiography, data available on the utility of MIBI than PET imaging in MR and CT. myeloma, but the technique has not been adopted widely into At the present time, it is fair to conclude that clinical clinical practice. The evidence from various studies of MIBI experience of PET imaging in patients with plasma cell imaging in myeloma suggests that the uptake of the Tc-99 dyscrasias is in evolution. PET imaging may detect early bone sestamibi correlates with the extent of disease (el-Shirbiny marrow involvement in patients with apparently solitary et al, 1997; Alexandrakis et al, 2001), and this can be correlated plasmacytoma (Kato et al, 2000; Schirrmeister et al, 2002), with other measurable parameters of disease activity such as can usefully assess the extent of active disease, detect lactate dehydrogenase, C-reactive protein and b2microglobulin extramedullary involvement or evaluate treatment response (Alexandrakis et al, 2001). In addition, this technique has a (Jadvar & Conti, 2002; Orchard et al, 2002; Schirrmeister et al, high sensitivity (90–100%) and specificity (88–93%) for the 2003; Bredella et al, 2005). The presence of extramedullary detection of myeloma bone lesions and good negative predic- uptake due to involvement by myeloma at diagnosis is a poor tive value in MGUS (Balleari et al, 2001; Svaldi et al, 2001; prognostic factor (Durie et al, 2002). A study by Schirrmeister Martin et al, 2005). The pattern (diffuse or focal plus diffuse et al (2002) demonstrated a 93% sensitivity for focal lesions rather than normal or focal) of Tc-99m sestamibi uptake has and a 84–92% sensitivity for diffuse lesions when verified added value in relation to known prognostic variables, such as against plain radiography, MR, CT and clinical evaluation. The C-reactive protein (Pace et al, 2001), with focal uptake being specificity ranged from 83–100% in patients with diffuse FDG more predictive of active myeloma than diffuse uptake uptake and 60% of patients with known focal osteolytic lesions (Nandurkar et al, 2006). on plain radiography were upstaged as a result of PET imaging. While MIBI imaging can detect additional lesions compared PET imaging appears to reliably predict active myeloma by to the radiological skeletal survey (Alper et al, 2003), it has been virtue of FDG uptake, and supports the diagnosis of mono- shown to underestimate the extent of spinal involvement by clonal gammopathy of undetermined significance (MGUS) by myeloma compared to MR imaging of the spine (Mirzaei et al, virtue of negative scans (Durie et al, 2002). 2003). In a comparative study of MIBI versus PET imaging The main limitation of PET scanning is limited spatial in myeloma patients, MIBI imaging detected additional resolution; subcentimetre lytic lesions seen on plain radiographs disease sites compared with the skeletal survey, and more sites may not be detectable on PET scanning, resulting in a negative of disease activity than PET imaging, and correlated better with scan (Bredella et al, 2005). The advent of fusion scanning the degree of bone marrow biopsy infiltration than PET imaging combining both PET and CT addresses the issue of limited (Mileshkin et al, 2004). More recently, in a study of 12 myeloma spatial resolution. In PET/CT fusion scanning, the patient patients at diagnosis, PET was found to compare favourably with receives an injection of FDG about 1 h before image acquisition. MIBI imaging, detecting 100% of the cases with marrow After the patient is positioned on the scanner bed, an initial involvement compared to 80% detected by MIBI, 89Æ5% of soft topogram is acquired to define the examination range for the tissue lesions (vs. 68Æ4% for MIBI) and 93Æ3% of skeletal lesions PET/CT image acquisition (usually from the ears to the hips). A (vs. 80% for MIBI) (Hung et al, 2005). spiral CT is then performed after which the scanner bed is moved back to the starting position and the PET scan commenced. Recommendations Reconstruction of the image, incorporating PET and CT data is • Based on currently available evidence, neither PET nor completed soon after PET image acquisition. The actual MIBI imaging can be recommended for routine use in the scanning time is shorter for PET/CT (approximately 30 min) management of myeloma patients. than a PET scan alone (approximately 1 h) because CT data are • Either technique may be useful in selected cases that used to perform attenuation correction. Although impressive warrant clarification of previous imaging findings, but 3-dimensional images are available on completion of the such an approach should ideally be made within the investigation, this occurs at the expense of greater cost and context of a clinical trial (Grade C recommendation; level radiation exposure (Kluetz et al, 2000). It is worth bearing in IV evidence) mind that false positive PET scans may arise from inflammatory • The evidence for the sensitivity of PET scanning is most changes due to active infection (Mahfouz et al, 2005), chemo- convincing in the setting of extramedullary disease. It is therapy within the previous 4 weeks or radiotherapy within the therefore reasonable to consider PET scanning in this previous 2–3 months (Juweid & Cheson, 2006). Studies of PET setting, to clarify the extent of extramedullary disease, in scanning and fusion PET-CT scanning in myeloma are currently cases where other imaging techniques have failed to early in development and it is too early to conclude what role this clarify the situation (Grade B recommendation; level III investigation will play in the management of the disease. evidence) 99Technetium sestamibi (MIBI) imaging is an alternative • If the decision to perform PET scanning has been taken, nuclear imaging procedure that uses Tc-99m-2-methoxy- it is advisable to avoid undertaking the procedure within isobutyl-isonitrile (Tc-99m sestamibi) as a tumour-seeking 4 weeks of chemotherapy or 3 months of radiotherapy tracer to identify areas of active disease in a variety of tumours (Grade B recommendation; level III evidence) including plasma cell dyscrasias. There are more published

ª 2007 British Society for Haematology 56 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline

2.7 Serum amyloid P component scintigraphy vertebral body collapse (Woo et al, 1986). It is common practice to perform a plain X-ray to confirm a suspected AL amyloidosis in association with monoclonal gammopathies vertebral fracture. In the event of suspected cord compression, is the most common and serious type of systemic amyloidosis MR imaging is the technique of choice (Joffe et al, 1988). It and is the cause of death in about one in 1500 people in the provides an accurate assessment of the level and extent of cord UK (Bird, 2004). Owing to the poor diagnostic yield from or nerve root compression, the size of the tumour mass and biopsies of uninvolved tissues, the considerable risk of the degree to which it has extended into the epidural space. haemorrhage from biopsies of involved tissues or organs and Once compression or impingement of the cord has been the lack of information regarding extent of disease and identified, it is important to obtain the opinions of both a response to therapy, alternative methods of confirming the clinical oncologist and spinal surgeon as to the appropriate diagnosis of suspected amyloidosis have been sought. Radio- management, particularly when extensive bony destruction is labelled serum amyloid P (SAP) component localises rapidly present. and specifically to amyloid deposits in proportion to the quantity of amyloid present. This enables the diagnosis and quantification of deposits by whole body scintigraphy 3.2 Vertebral collapse unresponsive to conservative (Hawkins, 2002). The diagnostic yield is >90% in AL measures amyloidosis, with reliable demonstration of deposits in the Recently, new interventions including percutaneous verteb- liver, spleen, kidneys, adrenal glands and bones, but insuffi- roplasty (PV) and kyphoplasty have been carried out in cient spatial resolution for gut, nerve and skin involvement. myeloma patients with vertebral body collapse. PV has been Cardiac involvement is difficult to demonstrate due to motion shown to provide rapid and effective pain relief in acute artifact and blood pool content as well as proximity to the osteoporotic vertebral fractures, and may improve subsequent spleen, which is frequently involved (Hachulla et al, 1996). rehabilitation of patients (Diamond et al, 2003; Winking SAP scintigraphy permits both the quantification of whole et al, 2004). Encouraging results have also been achieved in body amyloid load, which correlates with risks of high dose patients with myeloma (Fourney et al, 2003; Diamond et al, therapy in AL amyloidosis as well as providing information 2004; Yu et al, 2004). In the UK, National Institute for about response to therapy (Gillmore et al, 1999). The major Clinical Excellence (NICE) guidance (NICE, 2003) recom- drawback of SAP scintigraphy is that it is only available in a mends PV for the treatment of patients with vertebral few specialist centres. collapse due to osteoporosis, haemangioma and tumours, such as myeloma, whose pain is not alleviated by more Recommendations conservative analgesic measures. The following imaging • A diagnostic SAP scan should be requested if possible in investigations have been carried out prior to PV in various any patient suspected of having AL amyloidosis as a studies: complication of their plasma cell dyscrasias in addition to • Lateral and antero-posterior radiographs of the cervical, obtaining tissue biopsy evidence whenever possible thoracic and lumbar spine. (Grade B recommendation; level IIB evidence). • Cross-sectional MR imaging of affected vertebral body to • Follow up SAP scans should be performed every define the integrity and shape of the posterior vertebral 6–12 months in accordance with specialist centre policy, body cortex. to assess response to therapy or to monitor a patient with • STIR sequence of affected vertebra may confirm acute confirmed amyloidosis on a watchful waiting programme nature of vertebral body fracture by virtue of local bone (Grade B recommendation; level IIB evidence). marrow oedema (Van Gelderen et al, 1997). • Single photon emission computed tomography (SPECT) of the spine may help to determine how acute a fracture is, and 3. Use of imaging in the management of may predict for a good clinical response (Maynard et al, vertebral collapse 2000). • Calculation of percentage loss of vertebral height on MR 3.1 Acute vertebral collapse image (Diamond et al, 2004). Up to 70% of patients with myeloma experience vertebral Unfortunately, MR imaging does not enable recognition of collapse during the course of their disease due to osteopenia vertebrae that are likely to collapse, so cannot be used to and the predilection for vertebral body destruction (Carson predict for the use of preventive therapy at a particular level et al, 1955), with involvement most frequent in the lower (Lecouvet et al, 1998b). Although some authors strongly thoracic and lumbar vertebrae, although this figure may advocate the use of MR evaluation prior to the procedure become lower with the universal use of bisphosphonates in (Yu et al, 2004), there is no consensus regarding the optimum symptomatic patients. Spinal cord compression secondary to imaging protocol that should be undertaken prior to PV. vertebral body collapse occurs in up to 25% patients with

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 57 Guideline

Recommendations (Lecouvet et al, 2001). Contrast enhancement may be advan- • Urgent MR imaging is the diagnostic procedure of choice tageous in evaluating response to treatment (Rahmouni et al, to assess suspected cord compression in myeloma 1993). A change in MR pattern may correlate with response to patients with vertebral collapse (Grade B recommenda- therapy, including complete response shown by complete tion; level IIB evidence). resolution of the preceding marrow abnormality, and partial • Patients being considered for percutaneous vertebroplas- response demonstrated by conversion of a diffuse to a ty should undergo AP and lateral views of the cervical, variegated or focal pattern (Moulopoulos et al, 1994). Other thoracic and lumbar spine and CT or MR imaging of the features suggestive of a good response to treatment include a target area to exclude spinal cord compression (Grade C reduction in signal intensity on T2-weighted spin echo images recommendation; level IV evidence). and the absence of contrast-induced rim-enhancement that was previously present (Baur-Melnyk et al, 2005). Unfortu- nately, focal osteolytic lesions may remain hyperintense in 4. Assessment of response to therapy and both responders and non-responders to treatment due to disease relapse treatment-induced necrosis and inflammation (Rahmouni et al, 1993). These findings are not specific, however, with 4.1 Skeletal survey similar changes seen in other situations, such as growth factor- induced marrow repopulation following treatment. Following Although new or enlarging lesions undoubtedly signify stem cell transplantation, the marrow typically becomes disease progression, it is worth noting that lytic bone lesions replaced by fat, due to the effect of the high dose chemother- seldom show evidence of healing on plain radiographs and apy, and thus is seen as high signal on T1-weighted images but sequential skeletal surveys have limited value (Wahlin et al, focal residual lesions may persist (Agren et al, 1998). Residual 1982). New vertebral compression fractures on plain radio- marrow MR imaging abnormalities such as these may be graphs do not always signify disease progression and may associated with a poorer outcome (Angtuaco et al, 2004), but occur even after effective therapy, due to resolution of the this remains to be confirmed by other studies. tumour mass that was supporting the bony cortex (Collins, 2005). Furthermore, a skeletal survey is an exhausting and Paragraph on use of MR in follow up of non-secretory disease has potentially painful experience for myeloma patients and the been removed. Some patients have persistent pain despite justification for further radiation exposure should be assessed. evidence for a good response to therapy on laboratory Logistical considerations may intervene when previous radi- parameters. Such patients may benefit from further imaging ographs are misplaced, making sequential comparison to identify further deterioration in the skeleton, which can be impossible. ongoing despite adequate disease reduction. In such situations, careful interpretation is required, including correlation of the Recommendations images with the nature and timing of the treatment received • There is insufficient evidence of benefit to recommend (Lecouvet et al, 2001). routine follow up skeletal surveys in untreated asympto- MR imaging is the most sensitive and specific imaging matic patients in the absence of signs of disease progression. modality for the diagnosis of avascular necrosis of the femoral • In the event of clinical or laboratory evidence of disease head that may result from high dose steroid therapy or progression in treated or untreated patients, the skeletal radiotherapy, as demonstrated by the presence of the charac- survey should be repeated as part of the restaging process. teristic double-line sign on T2-weighted MR images (Lafforgue Any newly symptomatic areas of the skeleton should be et al, 1993). Early recognition of avascular necrosis before the specifically targeted. However, if disease progression development of a subchondral fracture is important for the occurs within 3 months of the previous skeletal survey, success of conservative management. in the absence of new skeletal symptoms, a new skeletal survey is unlikely to provide additional information Recommendations (Grade C recommendation; level IV evidence). • There is insufficient evidence to recommend routine MR imaging for the follow up of treated disease. 4.2 MR imaging Recommendation on use of MR imaging in follow up of Following treatment, there is a wide spectrum of treatment- non-secretory disease removed induced changes seen on MR for each pattern of marrow • In selected cases, where there are persisting unexplained involvement. In some cases, MR images fail to demonstrate symptoms, it is reasonable to discuss the potential useful- evidence of regression of myeloma infiltration in the marrow ness of follow up MR imaging with the radiologist (Grade C (Ishibashi et al, 1998). Focal lesions may shrink or remain recommendation; level IV evidence) unchanged in size; a fatty halo may appear following treatment

ª 2007 British Society for Haematology 58 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline

• MR imaging is the investigation of choice for suspected bone marrow examination and paraprotein measurement) was avascular necrosis of the femoral head (Grade B recom- around 90% (Balleari et al, 2001; Pace et al, 2001; Svaldi et al, mendation; level III evidence) 2001). MIBI has also been found to be predictive of disease relapse with a correlation between the pattern of extension and intensity of 99mTc-MIBI uptake and disease activity (Fallahi 4.3 Computed Tomography et al, 2005). Despite the plethora of reports of the use of MIBI in the In the follow up of treated disease, CT is more discriminating assessment of myeloma patients, the main limiting factor than plain radiography when pre- and post-treatment films are remains its limited sensitivity and specificity, the relatively low compared. The disappearance of extraosseous or extramedul- numbers of patients included and the heterogeneity of clinical lary masses and the reappearance of a continuous cortical situations described. With the increasing use of PET scanning outline and fatty marrow content may be seen in treated lytic in the management of oncological diseases, the tendency for disease in selected cases (Lecouvet et al, 2001). PET-based studies to be carried out in major myeloma treatment centres and technological advancements, such as Recommendations fusion PET-CT scanning, it is likely that PET-CT will overtake • Routine follow up CT scanning of treated disease cannot MIBI as the nuclear medicine tool of choice in myeloma. be recommended on current evidence and concern regarding radiation exposure. Recommendations • In selected cases, however, it is reasonable to use CT • Neither PET nor MIBI scanning can be recommended on scanning in the monitoring of the response of soft tissue the basis of current evidence for use in routine follow up masses to therapy (Grade B recommendation; level III of treated myeloma patients. evidence). • It would be reasonable to consider either technique for the • In selected cases, where there are persistent unexplained follow up of selected patients, such as those with predom- symptoms or there is concern about on-going fracture inant extramedullary or non-secretory disease, but this risk, or a lack of response to therapy, it is reasonable to would be best performed in the context of a clinical trial discuss the potential usefulness of performing a CT scan (Grade C recommendation; level IV evidence). in treated patients (Grade B recommendation; level III evidence). Acknowledgements 4.4 PET and MIBI scanning Dr Michael O’Doherty, Consultant in Nuclear Medicine, Guys & St Thomas’s Hospital NHS Trust, London. The results of serial PET scans in treated patients have been reported in a few patients with myeloma. Evidence of early relapse was identified in 60% of patients including some with References non-secretory disease, permitting prompt therapeutic inter- Abildgaard, N., Brixen, K., Eriksen, E.F., Kristensen, J.E., Nielsen, J.L. vention with radiotherapy (Durie et al, 2002). PET scanning & Heickendorff, L. (2004) Sequential analysis of biochemical mar- has also been shown to be useful in evaluating response to kers of bone resorption and bone densitometry in multiple myelo- therapy, particularly when other imaging techniques, such as ma. Haematologica, 89, 567–577. MR and CT, have remained abnormal, but the number of Agren, B., Rudberg, U., Isberg, B., Svensson, L. & Aspelin, P. (1998) patients evaluated was small (Bredella et al, 2005). In the MR imaging of multiple myeloma patients with bone-marrow setting of relapse, PET imaging may identify new sites of transplants. Acta Radiologica, 39, 36–42. disease, and unsuspected sites of extramedullary disease; Alexandrakis, M.G., Kyriakou, D.S., Passam, F., Koukouraki, S. & patients appear to have a particularly poor prognosis if FDG Karkavitsas, N. (2001) Value of Tc-99m sestamibi scintigraphy in uptake is present following high dose therapy and stem cell the detection of bone lesions in multiple myeloma: comparison with Tc-99m methylene diphosphonate. Annals of Hematology, 80, transplantation in either the medullary or extramedullary 349–353. compartment (Durie et al, 2002). In the era of newer therapies, Alper, E., Gurel, M., Evrensel, T., Ozkocaman, V., Akbunar, T. & including thalidomide analogues and proteasome inhibitors, Demiray, M. (2003) 99mTc-MIBI scintigraphy in untreated stage III the identification of occult residual disease may open up the multiple myeloma: comparison with X-ray skeletal survey and bone possibility of multimodal or maintenance therapy, but further scintigraphy. Nuclear Medicine Communications, 24, 537–542. diagnostic and therapeutic studies remain to be done. Angtuaco, E.J., Fassas, A.B., Walker, R., Sethi, R. & Barlogie, B. (2004) MIBI scanning has shown some utility in the follow up of Multiple myeloma: clinical review and diagnostic imaging. Radi- treated myeloma, including following the use of high dose ology, 231, 11–23. therapy, but concordance between response determined by Balleari, E., Villa, G., Garre, S., Ghirlanda, P., Agnese, G., Carletto, M., scintigraphic criteria and by conventional criteria (including Clavio, M., Ferrando, F., Gobbi, M., Mariani, G. & Ghio, R. (2001)

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 59 Guideline

Technetium-99m-sestamibi scintigraphy in multiple myeloma and management guidelines: a consensus report from the Scientific related gammopathies: a useful tool for the identification and fol- Advisors of the International Myeloma Foundation. The Hematology low-up of myeloma bone disease. Haematologica, 86, 78–84. Journal, 4, 379–398. Bataille, R., Chevalier, J., Rossi, M. & Sany, J. (1982) Bone scintigraphy Eisenberg, J.M., Schwartz, J.S., McCaslin, F.C., Kaufman, R., Glick, H. in plasma-cell myeloma. A prospective study of 70 patients. Radi- & Kroch, E. (1989) Substituting diagnostic services. New tests only ology, 145, 801–804. partly replace older ones. JAMA, 262, 1196–1200. Baur, A., Stabler, A., Lamerz, R., Bartl, R. & Reiser, M. (1998) Light el-Shirbiny, A.M., Yeung, H., Imbriaco, M., Michaeli, J., Macapinlac, chain deposition disease in multiple myeloma: MR imaging features H. & Larson, S.M. (1997) Technetium-99m-MIBI versus fluorine- correlated with histopathological findings. Skeletal Radiology, 27, 18-FDG in diffuse multiple myeloma. Journal of Nuclear Medicine, 173–176. 38, 1208–1210. Baur-Melnyk, A., Buhmann, S., Durr, H.R. & Reiser, M. (2005) Role of Eustace, S., Tello, R., DeCarvalho, V., Carey, J., Wroblicka, J.T., Mel- MRI for the diagnosis and prognosis of multiple myeloma. European hem, E.R. & Yucel, E.K. (1997) A comparison of whole-body tur- Journal of Radiology, 55, 56–63. boSTIR MR imaging and planar 99mTc-methylene diphosphonate Berrington de Gonzalez, A. & Darby, S. (2004) Risk of cancer from scintigraphy in the examination of patients with suspected skeletal diagnostic X-rays: estimates for the UK and 14 other countries. metastases. AJR. American Journal of Roentgenology, 169, 1655–1661. Lancet, 363, 345–351. Eustace, S.J., Walker, R., Blake, M. & Yucel, E.K. (1999) Whole-body Bird, J. (2004) Guidelines on the diagnosis and management of AL MR imaging. Practical issues, clinical applications, and future di- amyloidosis. British Journal of Haematology, 125, 681–700. rections. Magnetic Resonance Imaging Clinics of North America, 7, Bredella, M.A., Steinbach, L., Caputo, G., Segall, G. & Hawkins, R. (2005) 209–236. Value of FDG PET in the assessment of patients with multiple mye- Fallahi, B., Saghari, M., Fard Esfahani, A., Eftekhari, M., Iravani, M., loma. AJR. American Journal of Roentgenology, 184, 1199–1204. Beiki, D., Dabbagh Kakhki, V.R. & Sadeghi, R. (2005) The value of Carlson, K., Astrom, G., Nyman, R., Ahlstrom, H. & Simonsson, B. 99mTc-MIBI whole body scintigraphy in active and in remission (1995) MR imaging of multiple myeloma in tumour mass measure- multiple myeloma. Hellenic Journal of Nuclear Medicine, 8, 165– ment at diagnosis and during treatment. Acta Radiologica, 36, 9–14. 168. Carson, C.P., Ackerman, L.V. & Maltby, J.D. (1955) Plasma cell Fourney, D.R., Schomer, D.F., Nader, R., Chlan-Fourney, J., Suki, D., myeloma; a clinical, pathologic and roentgenologic review of 90 Ahrar, K., Rhines, L.D. & Gokaslan, Z.L. (2003) Percutaneous ver- cases. American Journal of Clinical Pathology, 25, 849–888. tebroplasty and kyphoplasty for painful vertebral body fractures in Collins, C. (1998) Multiple myeloma. In: Imaging in Oncology (ed. by cancer patients. Journal of Neurosurgery, 98, 21–30. J.E. Husband & R.H. Resnik), Vol 2. Fruehwald, F.X., Tscholakoff, D., Schwaighofer, B., Wicke, L., Collins, C.D. (2005) Problems monitoring response in multiple mye- Neuhold, A., Ludwig, H. & Hajek, P.C. (1988) Magnetic resonance loma. Cancer Imaging, 5(Suppl.), S119–S126. imaging of the lower vertebral column in patients with multiple Diamond, T.H., Champion, B. & Clark, W.A. (2003) Management of myeloma. Investigative Radiology, 23, 193–199. acute osteoporotic vertebral fractures: a nonrandomized trial com- Fryback, D.G. & Thornbury, J.R. (1991) The efficacy of diagnostic paring percutaneous vertebroplasty with conservative therapy. The imaging. Medical Decision Making, 11, 88–94. American Journal of Medicine, 114, 257–265. Gillmore, J.D., Lovat, L.B. & Hawkins, P.N. (1999) Amyloidosis and Diamond, T.H., Hartwell, T., Clarke, W. & Manoharan, A. (2004) the liver. Journal of Hepatology, 30 (Suppl 1), 17–33. Percutaneous vertebroplasty for acute vertebral body fracture and Hachulla, E., Maulin, L., Deveaux, M., Facon, T., Bletry, O., Vanhille, deformity in multiple myeloma: a short report. British Journal of P., Wechsler, B., Godeau, P., Levesque, H., Hatron, P.Y., Huglo, D., Haematology, 124, 485–487. Devulder, B. & Marchandise, X. (1996) Prospective and serial study Dimopoulos, M.A., Moulopoulos, A., Smith, T., Delasalle, K.B. & of primary amyloidosis with serum amyloid P component scinti- Alexanian, R. (1993) Risk of disease progression in asymptomatic graphy: from diagnosis to prognosis. The American Journal of multiple myeloma. The American Journal of Medicine, 94, 57–61. Medicine, 101, 77–87. Dimopoulos, M.A., Moulopoulos, L.A., Datseris, I., Weber, D., Hargaden, G., O’Connell, M., Kavanagh, E., Powell, T., Ward, R. & Delasalle, K., Gika, D. & Alexanian, R. (2000) Imaging of myeloma Eustace, S. (2003) Current concepts in whole-body imaging using bone disease–implications for staging, prognosis and follow-up. Acta turbo short tau inversion recovery MR imaging. AJR. American Oncologica, 39, 823–827. Journal of Roentgenology, 180, 247–252. Durie, B.G. & Salmon, S.E. (1975) A clinical staging system for mul- Hart, D. & Wall, BF. (2002) Radiation exposure of the UK population tiple myeloma. Correlation of measured myeloma cell mass with from medical and dental X-ray examinations. National Radiological presenting clinical features, response to treatment, and survival. Protection Board report NRPB-W4. National Radiological Protection Cancer, 36, 842–854. Board, Didcot, Oxon, UK. Durie, B.G., Waxman, A.D., D’Agnolo, A. & Williams, C.M. (2002) Hawkins, P.N. (2002) Serum amyloid P component scintigraphy for Whole-body (18)F-FDG PET identifies high-risk myeloma. Journal diagnosis and monitoring amyloidosis. Current Opinion in Ne- of Nuclear Medicine, 43, 1457–1463. phrology and Hypertension, 11, 649–655. Durie, B.G., Kyle, R.A., Belch, A., Bensinger, W., Blade, J., Boccadoro, Healy, J.C. & Armstrong, P. (1998) Radiological features of multiple M., Child, J.A., Comenzo, R., Djulbegovic, B., Fantl, D., Gahrton, G., myeloma. In: Myeloma (ed. by J. Malpas). Harousseau, J.L., Hungria, V., Joshua, D., Ludwig, H., Mehta, J., Hjorth, M., Holmberg, E., Rodjer, S. & Westin, J. (1992) Impact of Morales, A.R., Morgan, G., Nouel, A., Oken, M., Powles, R., active and passive exclusions on the results of a clinical trial in Roodman, D., San Miguel, J., Shimizu, K., Singhal, S., Sirohi, B., multiple myeloma. The Myeloma Group of Western Sweden. British Sonneveld, P., Tricot, G. & Van Ness, B. (2003) Myeloma Journal of Haematology, 80, 55–61.

ª 2007 British Society for Haematology 60 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline

Horger, M., Claussen, C.D., Bross-Bach, U., Vonthein, R., Trabold, T., Vertebral compression fractures in multiple myeloma. Part II. As- Heuschmid, M. & Pfannenberg, C. (2005) Whole-body low-dose sessment of fracture risk with MR imaging of spinal bone marrow. multidetector row-CT in the diagnosis of multiple myeloma: an Radiology, 204, 201–205. alternative to conventional radiography. European Journal of Radi- Lecouvet, F.E., Vande Berg, B.C., Michaux, L., Malghem, J., Maldague, ology, 54, 289–297. B.E., Jamart, J., Ferrant, A. & Michaux, J.L. (1998a) Stage III mul- Hung, G.U., Tsai, C.C., Tsai, S.C. & Lin, W.Y. (2005) Comparison of tiple myeloma: clinical and prognostic value of spinal bone marrow Tc-99m sestamibi and F-18 FDG-PET in the assessment of multiple MR imaging. Radiology, 209, 653–660. myeloma. Anticancer Research, 25, 4737–4741. Lecouvet, F.E., Vande Berg, B.C., Michaux, L., Jamart, J., Maldague, Hunink, M.G. & Krestin, G.P. (2002) Study design for concurrent B.E. & Malghem, J. (1998b) Development of vertebral fractures in development, assessment, and implementation of new diagnostic patients with multiple myeloma: does MRI enable recognition of imaging technology. Radiology, 222, 604–614. vertebrae that will collapse? Journal of Computer Assisted Tomo- International Myeloma Working Group (2003) Criteria for the clas- graphy, 22, 430–436. sification of monoclonal gammopathies, multiple myeloma and Lecouvet, F.E., Malghem, J., Michaux, L., Maldague, B., Ferrant, A., related disorders: a report of the International Myeloma Working Michaux, J.L. & Vande Berg, B.C. (1999) Skeletal survey in advanced Group. British Journal of Haematology, 121, 749–757. multiple myeloma: radiographic versus MR imaging survey. British Ishibashi, M., Nonoshita, M., Uchida, M., Kojima, K., Tomita, N., Journal of Haematology, 106, 35–39. Matsumoto, S., Tanaka, K. & Hayabuchi, N. (1998) Bone marrow Lecouvet, F.E., Vande Berg, B.C., Malghem, J. & Maldague, B.E. (2001) uptake of thallium-201 before and after therapy in multiple mye- Magnetic resonance and computed tomography imaging in multiple loma. Journal of Nuclear Medicine, 39, 473–475. myeloma. Seminars in Musculoskeletal Radiology, 5, 43–55. Isoda, H., Kojima, H., Shimizu, K., Kurokawa, H., Ikeda, K., Sawada, Lewis, M.K., Blake, G.M. & Fogelman, I. (1994) Patient dose in dual S., Sakaida, N. & Okamura, A. (2001) Multiple myeloma: short T2 x-ray absorptiometry. Osteoporosis International, 4, 11–15. on MR imaging. Clinical Imaging, 25, 141–143. Libshitz, H.I., Malthouse, S.R., Cunningham, D., MacVicar, A.D. & Jadvar, H. & Conti, P.S. (2002) Diagnostic utility of FDG PET in Husband, J.E. (1992) Multiple myeloma: appearance at MR imaging. multiple myeloma. Skeletal Radiology, 31, 690–694. Radiology, 182, 833–837. Joffe, J., Williams, M.P., Cherryman, G.R., Gore, M., McElwain, T.J. & Ludwig, H., Kumpan, W. & Sinzinger, H. (1982) Radiography and Selby, P. (1988) Magnetic resonance imaging in myeloma. Lancet, 1, bone scintigraphy in multiple myeloma: a comparative analysis. The 1162–1163. British Journal of Radiology, 55, 173–181. Juweid, M.E. & Cheson, B.D. (2006) Positron-emission tomography Ludwig, H., Fruhwald, F., Tscholakoff, D., Rasoul, S., Neuhold, A. & and assessment of cancer therapy. The New England journal of Fritz, E. (1987) Magnetic resonance imaging of the spine in multiple medicine, 354, 496–507. myeloma. Lancet, 2, 364–366. Kanis, J.A. & Gluer, C.C. (2000) An update on the diagnosis and Mahfouz, T., Miceli, M.H., Saghafifar, F., Stroud, S., Jones-Jackson, L., assessment of osteoporosis with densitometry. Committee of Sci- Walker, R., Grazziutti, M.L., Purnell, G., Fassas, A., Tricot, G., entific Advisors, International Osteoporosis Foundation. Osteo- Barlogie, B. & Anaissie, E. (2005) 18F-fluorodeoxyglucose positron porosis International, 11, 192–202. emission tomography contributes to the diagnosis and management Kato, T., Tsukamoto, E., Nishioka, T., Yamazaki, A., Shirato, H., of infections in patients with multiple myeloma: a study of 165 Kobayashi, S., Asaka, M., Imamura, M. & Tamaki, N. (2000) Early infectious episodes. Journal of Clinical Oncology, 23, 7857–7863. detection of bone marrow involvement in extramedullary plasma- Marshall, D., Johnell, O. & Wedel, H. (1996) Meta-analysis of how well cytoma by whole-body F-18 FDG positron emission tomography. measures of bone mineral density predict occurrence of osteoporotic Clinical Nuclear Medicine, 25, 870–873. fractures. BMJ, 312, 1254–1259. Kavanagh, E., Smith, C. & Eustace, S. (2003) Whole-body turbo STIR Martin, M.G., Romero Colas, M.S., Dourdil Sahun, M.V., Olave, P., MR imaging: controversies and avenues for development. European Alba, P.R. & Narraco, J.B. (2005) BaselinTc(99)-MIBI scanning Radiology, 13, 2196–2205. predicts survival in multiple myeloma and helps to differentiate this Kluetz, P.G., Meltzer, C.C., Villemagne, V.L., Kinahan, P.E., Chander, disease from monoclonal gammopathy of unknown significance. S., Martinelli, M.A. & Townsend, D.W. (2000) Combined PET/CT Haematologica, 90, 1141–1143. Imaging in Oncology. Impact on Patient Management. Clinical Masud, T., Langley, S., Wiltshire, P., Doyle, D.V. & Spector, T.D. Positron Imaging, 3, 223–230. (1993) Effect of spinal osteophytosis on bone mineral density Kyle, R.A. (1985) Multiple myeloma: current therapy and a glimpse of measurements in vertebral osteoporosis. BMJ, 307, 172–173. the future. Scandinavian Journal of Haematology, 35, 38–47. Maynard, A.S., Jensen, M.E., Schweickert, P.A., Marx, W.F., Short, J.G. & Lafforgue, P., Dahan, E., Chagnaud, C., Schiano, A., Kasbarian, M. & Kallmes, D.F. (2000) Value of bone scan imaging in predicting pain Acquaviva, P.C. (1993) Early-stage avascular necrosis of the femoral relief from percutaneous vertebroplasty in osteoporotic vertebral head: MR imaging for prognosis in 31 cases with at least 2 years of fractures. AJNR. American Journal of Neuroradiology, 21, 1807–1812. follow-up. Radiology, 187, 199–204. Mileshkin, L., Blum, R., Seymour, J.F., Patrikeos, A., Hicks, R.J. & Laroche, M., Assoun, J., Sixou, L. & Attal, M. (1996) Comparison of Prince, H.M. (2004) A comparison of fluorine-18 fluoro-deox- MRI and computed tomography in the various stages of plasma cell yglucose PET and technetium-99m sestamibi in assessing patients disorders: correlations with biological and histological findings. with multiple myeloma. European Journal of Haematology, 72, Myelome-Midi-Pyrenees Group. Clinical and Experimental Rheu- 32–37. matology, 14, 171–176. Mirels, H. (1989) Metastatic disease in long bones. A proposed scoring Lecouvet, F.E., Malghem, J., Michaux, L., Michaux, J.L., Lehmann, F., system for diagnosing impending pathologic fractures. Clinical Maldague, B.E., Jamart, J., Ferrant, A. & Vande Berg, B.C. (1997) Orthopaedics and Related Research, 415(suppl.), S4–13.

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 61 Guideline

Mirzaei, S., Filipits, M., Keck, A., Bergmayer, W., Knoll, P., Koehn, H., Pace, L., Catalano, L., Del Vecchio, S., Di Gennaro, F., De Renzo, A., Ludwig, H. & Pecherstorfer, M. (2003) Comparison of Technetium- Sica, G., Califano, C., Tedesco, N., Borrelli, G., Rotoli, B. & 99m-MIBI imaging with MRI for detection of spine involvement in Salvatore, M. (2001) Predictive value of technetium-99m sestamibi patients with multiple myeloma. BMC Nuclear Medicine, 3, 2. in patients with multiple myeloma and potential role in the follow- Moulopoulos, L.A. & Dimopoulos, M.A. (1997) Magnetic resonance up. European Journal of Nuclear Medicine, 28, 304–312. imaging of the bone marrow in hematologic malignancies. Blood, 90, Phekoo, K.J., Schey, S.A., Richards, M.A., Bevan, D.H., Bell, S., Gillett, 2127–2147. D. & Moller, H. (2004) A population study to define the incidence Moulopoulos, L.A., Varma, D.G., Dimopoulos, M.A., Leeds, N.E., and survival of multiple myeloma in a National Health Service Kim, E.E., Johnston, D.A., Alexanian, R. & Libshitz, H.I. (1992) Region in UK. British Journal of Haematology, 127, 299–304. Multiple myeloma: spinal MR imaging in patients with untreated Rahmouni, A., Divine, M., Mathieu, D., Golli, M., Dao, T.H., Jazaerli, newly diagnosed disease. Radiology, 185, 833–840. N., Anglade, M.C., Reyes, F. & Vasile, N. (1993) Detection of Moulopoulos, L.A., Dimopoulos, M.A., Weber, D., Fuller, L., Libshitz, multiple myeloma involving the spine: efficacy of fat-suppression H.I. & Alexanian, R. (1993) Magnetic resonance imaging in the and contrast-enhanced MR imaging. AJR. American Journal of staging of solitary plasmacytoma of bone. Journal of Clinical On- Roentgenology, 160, 1049–1052. cology, 11, 1311–1315. Ricci, C., Cova, M., Kang, Y.S., Yang, A., Rahmouni, A., Scott, W.W., Moulopoulos, L.A., Dimopoulos, M.A., Alexanian, R., Leeds, N.E. & Jr, & Zerhouni, E.A. (1990) Normal age-related patterns of cellular Libshitz, H.I. (1994) Multiple myeloma: MR patterns of response to and fatty bone marrow distribution in the axial skeleton: MR ima- treatment. Radiology, 193, 441–446. ging study. Radiology, 177, 83–88. Moulopoulos, L.A., Dimopoulos, M.A., Smith, T.L., Weber, D.M., Royal College of Radiologists (2003) Making the best use of a depart- Delasalle, K.B., Libshitz, H.I. & Alexanian, R. (1995) Prognostic ment of clinical radiology. Guidelines for Doctors. Fifth Edition. Pub- significance of magnetic resonance imaging in patients with lication No: BFCR(03)3. Royal College of Radiologists, London, UK. asymptomatic multiple myeloma. Journal of Clinical Oncology, 13, Samson, D. (2001) Diagnosis and management of multiple myeloma. 251–256. British Journal of Haematology, 115, 522–540. Moulopoulos, L.A., Gika, D., Anagnostopoulos, A., Delasalle, K., Scane, A.C., Masud, T., Johnson, F.J. & Francis, R.M. (1994) The Weber, D., Alexanian, R. & Dimopoulos, M.A. (2005) Prognostic reliability of diagnosing osteoporosis from spinal radiographs. Age significance of magnetic resonance imaging of bone marrow in Ageing, 23, 283–286. previously untreated patients with multiple myeloma. Annals of Schirrmeister, H., Bommer, M., Buck, A.K., Muller, S., Messer, P., Oncology, 16, 1824–1828. Bunjes, D., Dohner, H., Bergmann, L. & Reske, S.N. (2002) Initial Mulligan, M.E. (2000) Myeloma and lymphoma. Seminars in Muscu- results in the assessment of multiple myeloma using 18F-FDG PET. loskeletal Radiology, 4, 127–135. European Journal of Nuclear Medicine and Molecular Imaging, 29, Mulligan, M.E. (2005) Imaging techniques used in the diagnosis, sta- 361–366. ging, and follow-up of patients with myeloma. Acta Radiologica, 46, Schirrmeister, H., Buck, A.K., Bergmann, L., Reske, S.N. & Bommer, 716–724. M. (2003) Positron emission tomography (PET) for staging of so- Nandurkar, D., Kalff, V., Turlakow, A., Spencer, A., Bailey, M.J. & litary plasmacytoma. Cancer Biotherapy & Radiopharmaceuticals, 18, Kelly, M.J. (2006) Focal MIBI uptake is a better indicator of active 841–845. myeloma than diffuse uptake. European Journal of Haematology, 76, Schmidt, G.P., Schoenberg, S.O., Reiser, M.F. & Baur-Melnyk, A. 141–146. (2005) Whole-body MR imaging of bone marrow. European Journal NICE (2003) Guidance on Percutaneous Vertebroplasty. Interventional of Radiology, 55, 33–40. procedure guidance number IPG012. (September 2003) National Snapper, I. & Khan, A. (1971) Myelomatosis: fundamentals and clinical Institute for Clinical Excellence, London. features. University Park Press, Baltimore. Nilsson-Ehle, H., Holmdahl, C., Suurkula, M. & Westin, J. (1982) Bone Soutar, R., Lucraft, H., Jackson, G., Reece, A., Bird, J., Low, E. & scintigraphy in the diagnosis of skeletal involvement and metastatic Samson, D. (2004) Guidelines on the diagnosis and management of calcification in multiple myeloma. Acta Medica Scandinavica, 211, solitary plasmacytoma of bone and solitary extramedullary plas- 427–432. macytoma. British Journal of Haematology, 124, 717–726. Nosas-Garcia, S., Moehler, T., Wasser, K., Kiessling, F., Bartl, R., Zuna, I., Stabler, A., Baur, A., Bartl, R., Munker, R., Lamerz, R. & Reiser, Hillengass, J., Goldschmidt, H., Kauczor, H.U. & Delorme, S. (2005) M.F. (1996) Contrast enhancement and quantitative signal ana- Dynamic contrast-enhanced MRI for assessing the disease activity of lysis in MR imaging of multiple myeloma: assessment of focal and multiple myeloma: a comparative study with histology and clinical diffuse growth patterns in marrow correlated with biopsies and markers. Journal of Magnetic Resonance Imaging, 22, 154–162. survival rates. AJR. American Journal of Roentgenology, 167, 1029– Office for National Statistics (2001) Cancer Trends in England and 1036. Wales 1950–1999. HMSO, London. Svaldi, M., Tappa, C., Gebert, U., Bettini, D., Fabris, P., Franzelin, F., Orchard, K., Barrington, S., Buscombe, J., Hilson, A., Prentice, H.G. & Osele, L. & Mitterer, M. (2001) Technetium-99m-sestamibi scinti- Mehta, A. (2002) Fluoro-deoxyglucose positron emission tomography: an alternative approach for diagnosis and follow-up of ac- graphy imaging for the detection of occult disease in multiple tive myeloma lesions after high-dose chemotherapy and autologous myeloma. British Journal of Haematology, 117, 133–135. stem cell transplantation. Annals of Hematology, 80, 393–397. Otsuka, N., Fukunaga, M., Morita, K., Ono, S., Nagai, K., Yamada, O. Tamir, R., Glanz, I., Lubin, E., Vana, D. & Pick, A.I. (1983) Com- & Yawata, Y. (1993) Bone and 67Ga scintigraphy in the evaluation parison of the sensitivity of 99mTc-methyl diphosphonate bone scan of rib lesions in patients with multiple myeloma. Radiation Medi- with the skeletal X-ray survey in multiple myeloma. Acta Haema- cine, 11, 75–80. tologica, 69, 236–242.

ª 2007 British Society for Haematology 62 Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 Guideline

Tertti, R., Alanen, A. & Remes, K. (1995) The value of magnetic re- Walker, M.P., Yaszemski, M.J., Kim, C.W., Talac, R. & Currier, B.L. sonance imaging in screening myeloma lesions of the lumbar spine. (2003) Metastatic disease of the spine: evaluation and treatment. British Journal of Haematology, 91, 658–660. Clinical Orthopaedics, 415(suppl.), S165–175. Turesson, I., Zettervall, O., Cuzick, J., Waldenstrom, J.G. & Velez, R. Weber, D.M., Dimopoulos, M.A., Moulopoulos, L.A., Delasalle, K.B., (1984) Comparison of trends in the incidence of multiple myeloma Smith, T. & Alexanian, R. (1997) Prognostic features of asymptomatic in Malmo, Sweden, and other countries, 1950–1979.. The New multiple myeloma. British Journal of Haematology, 97, 810–814. England Journal of Medicine, 310, 421–424. Winking, M., Stahl, J.P., Oertel, M., Schnettler, R. & Boker, D.K. Van de Berg, B.C., Lecouvet, F.E., Michaux, L., Labaisse, M., (2004) Treatment of pain from osteoporotic vertebral collapse by Malghem, J., Jamart, J., Maldague, B.E., Ferrant, A. & Michaux, percutaneous PMMA vertebroplasty. Acta Neurochir (Wien), 146, J.L. (1996) Stage I multiple myeloma: value of MR imaging of the 469–476. bone marrow in the determination of prognosis. Radiology, 201, Wisloff, F., Andersen, P., Andersson, T.R., Brandt, E., Eika, C., Fjaes- 243–246. tad, K., Gronvold, T., Holm, B., Lovasen, K. & Tjonnfjord, G.E. Van Gelderen, W.F., al-Hindawi, M., Gale, R.S., Steward, A.H. & (1991) Incidence and follow-up of asymptomatic multiple myeloma. Archibald, C.G. (1997) Signiﬁcance of short tau inversion recovery The myeloma project of health region I in Norway. II. European magnetic resonance sequence in the management of skeletal injuries. Journal of Haematology, 47, 338–341. Australasian Radiology, 41, 13–15. Woo, E., Yu, Y.L., Ng, M., Huang, C.Y. & Todd, D. (1986) Spinal cord Wahlin, A., Holm, J., Osterman, G. & Norberg, B. (1982) Evaluation of compression in multiple myeloma: who gets it? Australian and New serial bone X-ray examination in multiple myeloma. Acta Medica Zealand Journal of Medicine, 16, 671–675. Scandinavica, 212, 385–387. Yu, S.W., Lee, P.C., Ma, C.H., Chuang, T.Y. & Chen, Y.J. (2004) Walker, R.E. & Eustace, S.J. (2001) Whole-body magnetic resonance Vertebroplasty for the treatment of osteoporotic compression spinal imaging: techniques, clinical indications, and future applications. fracture: comparison of remedial action at different stages of injury. Seminars in Musculoskeletal Radiology, 5, 5–20. The Journal of Trauma, 56, 629–632.

ª 2007 British Society for Haematology Journal Compilation 2007 Blackwell Publishing Ltd, British Journal of Haematology, 137, 49–63 63