SEARCH

SEARCH BY CITATION

Keywords:

  • multiple myeloma;
  • bone lesion;
  • magnetic resonance imaging;
  • bone densitometry;
  • prognostic factor

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. References

We assessed the role of spinal magnetic resonance imaging (MRI) and bone densitometry as prognostic factors in patients with asymptomatic stage I multiple myeloma (MM) and negative skeletal survey. 55 consecutive patients underwent spinal MRI and 41 of them underwent bone densitometry by dual-energy X-ray absorptiometry (DEXA).

Spinal MRI studies showed evidence of bone marrow involvement in 17/55 patients (31%). A diffuse pattern was present in three patients and a focal pattern in 14 patients, nine of them with only one nodular lesion. During a median follow-up of 25 months, 10 patients had disease progression, 8/17 patients with abnormal MRI and 2/38 patients with normal MRI. Median time to disease progression was not reached in both groups but was significantly different for patients with normal and those with abnormal patterns on MRI (P < 0.0001). Lumbar BMD was only slightly decreased compared with normal people (median lumbar Z score − 0.43) and was not of prognostic value. Using a multivariate analysis the only two independent significant prognostic parameters were abnormal MRI (P < 0.001, HR 30.4, 95% CI 4.3–213) and bone marrow plasmacytosis > 20% (P = 0.004, HR 16.4, 95% CI 2.6–104).

Thus, spinal MRI but not bone densitometry, appeared to be justified in patients with stage I asymptomatic MM and negative skeletal survey.

In recent years increasing numbers of patients are diagnosed with asymptomatic multiple myeloma (MM). In these patients with indolent myeloma the time to progressive disease requiring treatment is variable, with a range of a few months to several years (Dimopoulos et al, 1993; Facon et al, 1995; Kyle & Greipp, 1980; Weber et al, 1997; Wisloff et al, 1991).

Various criteria have been proposed to distinguish patients who are likely to remain stable without therapy for a long period from those who are at high risk for a serious complication and might benefit from early chemotherapy. They include the presence of one lytic bone lesion on skeletal survey, a high level of serum myeloma protein or urinary Bence-Jones protein, the degree of bone marrow plasmacytosis, anaemia and the IgA isotype of the monoclonal component (Dimopoulos et al, 1993; Facon et al, 1995; Kyle & Greipp, 1980; Weber et al, 1997; Wisloff et al, 1991). Since the presence of a lytic bone lesion on skeletal survey is constantly associated with a median time to progression of about 8–10 months, such patients should be treated soon after diagnosis.

In patients without any lytic bone lesions on skeletal survey, two sensitive techniques may disclose bone abnormalities related to MM: spinal magnetic resonance imaging (MRI) and bone densitometry. One group assessed the value of spinal MRI in 47 patients with asymptomatic MM and concluded that MRI appeared to be justified only for patients at intermediate risk defined by one of the three following features: serum peak level > 30 g/l, IgA myeloma, Bence-Jones protein excretion > 50 mg/d (Moulopoulos et al, 1995; Weber et al, 1997). The prognostic value of bone densitometry in patients with asymptomatic MM has never been studied, although its value as a marker of treatment response has been established in patients with advanced MM (Mariette et al, 1992, 1995).

In this study concerning 55 patients we assessed the role of spinal MRI and bone densitometry as prognostic factors in patients with asymptomatic stage I MM without any lytic bone lesions on skeletal survey.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. References

Patients

Between August 1993 and October 1997 we studied, in one centre, 55 consecutive, newly diagnosed patients with asymptomatic MM. Eligibility criteria included Durie and Salmon criteria of low tumour mass (or stage I) MM with the exclusion of patients with one lytic bone lesion on skeletal survey since such patients should be treated soon after diagnosis. In addition, patients with a serum and/or urine monoclonal immunoglobulin (MIg) whose bone marrow contained < 10% plasma cells were also included, providing that myeloma cells could be unequivocally identified on bone marrow smears. In all of these, bone marrow specimens were independently reviewed by two experienced cytologists and stringent criteria were used to define myeloma cells, especially the presence of nuclear abnormalities: presence of nucleoles and of a thin chromatin (Greipp & Kyle, 1983).

After diagnosis, patients were re-evaluated at least every 4 months. Disease progression was defined by the appearence of a symptomatic lytic bone lesion, by a sustained increase in the MIg by > 25% or by occurrence of hypercalcaemia or anaemia (haemoglobin < 10 g/dl).

MRI

All patients underwent MRI of the lumbar spine, including the last three dorsal vertebras and the sacral spine. MRI of the spine was performed using a 0.5 T MRMax system (General Electric Inc.) with a spinal coil. Sagittal images included three sequences: (1) gradient echo T2-weighted images (TR 440–540 ms) with a TE of 20 ms. At this TE, water and fat protons are in opposite phase at 0.5 T, resulting in a significant decrease of the fat signal. Because the lesions in myeloma have a bright signal on T2-weighted images, fat suppression (i.e. decrease in the high signal of fat) improves tumour delineation. In addition, gradient echo sequences are very sensitive to susceptibility artefacts. Therefore trabecular (uninvolved) bone appears darker than tumourous bone. The combination of opposed-phase imaging and gradient-echo imaging allows for a better lesion conspicuity; (2) spin-echo T1-weighted images (TR 520–640 ms, TE 34 ms); and (3) spin-echo T1-weighted images (TR 520/640 ms, TE 34 ms) after injection of gadolinium-DOTA with fat and water suppression (chemical shift imaging, Dixon technique). The records were independently reviewed by two radiologists with a consensual agreement. MRI were classified as (1) no vertebral lesion; (2) one or multiple vertebral focal lesions; or (3) a diffuse vertebral bone marrow replacement.

Bone densitometry

Lumbar spine (antero-posterior view L1–L4) and whole body bone mineral density (BMD) were measured by dual-energy X-ray absorptiometry (DEXA) using a standard osteodensitometer QDR-1000/W (Hologic Inc.). A Z score, i.e. the number of standard deviations related to normal age- and sex-matched subjects, was calculated for each BMD measurement. The long-term precision on a hydroxyapatite spine phantom was < 0.5%. As assessed by in vivo control studies in our department, reproducibility was 0.7% for lumbar spine and 1.5% for total body. Fractured vertebrae were discarded for lumbar BMD analysis.

Statistical analysis

Clinical features and laboratory parameters of patients with normal and abnormal MR studies were compared using Mann-Whitney and χ2 tests. Curves for event-free survival were plotted according to the Kaplan-Meier method and were compared by the log-rank test using the BMDP software. The influence of concomitant variables on event-free survival (EFS) was examined using the Cox proportional-hazards regression model with time to progression as the outcome variable. The following variables were included in the univariate analysis: age, serum peak level, bone marrow plasmacytosis, haemoglobin level, β2-microglobulin level, Bence-Jones protein level, isotype of the monoclonal component, lumbar Z score and result of MR imaging. The Cox proportional-hazards regression model was used first to study the influence of each factor on time to progression, and then to identify independent prognostic factors. For the latter, a multivariate analysis was applied using a stepwise procedure to those of the above variables having P leqslant R: less-than-or-eq, slant 0.20 in the univariate analysis. In the multivariate procedure, P < 0.05 was considered as the level of significance. Hazards ratios (HR) of disease progression were presented along with 95% confidence intervals.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. References

Patients

There were 26 men and 29 women. The median age was 62 years (range 37–83). The median level of serum myeloma protein was 21.5 g/l (range 3–47). An IgG type was present in 76%, an IgA type in 22% and a Bence-Jones protein alone in 2% of patients. Bence-Jones protein excretion exceeded 50 mg/d in 13% of patients. The median percentage of bone marrow plasma cells was 9% (range 1–30), i.e. 27/55 patients had < 10% bone marrow plasma cells. The median β2-microglobulin level was 1.5 mg/l (range 0.9–6.4) (normal 0.9–1.9). The median haemoglobin level was 13.2 g/dl (range 10–15.9); 10/55 patients had a haemoglobin value between 10 and 12 g/dl.

MRI

MRI of the lumbar spine and of the last three dorsal vertebras was performed in all patients. Evidence of bone marrow involvement was detected in 17/55 patients (31%). A diffuse bone marrow involvement was present in three patients, a single nodular lesion in six patients, and multiple nodular lesions in eight patients (Figs 1–3). The diffuse bone marrow involvement appeared as a slightly higher signal intensity on T2-weighted images and as an homogenous low signal on T1-weighted images which enhanced after contrast administration. The focal lesions had a bright signal on T2-weighted images, were iso- or hypo-intense relative to the uninvolved bone marow on T1-weighted images and enhanced after contrast administration.

image

Figure 1. . Diffuse involvement of the bone marrow of the lumbar spine. The sagittal MR T2*-weighted image demonstrates a slight homogenous hypersignal of the bone marrow (A). The unenhanced T1-weighted image (B) shows a diffuse low signal of the bone marrow which was enhanced after gadolinium-DOTA injection (C).

Download figure to PowerPoint

image

Figure 2. . Isolated nodular vertebral lesion. The sagittal MRI show a focal nodular hypersignal on T2*-weighted image of the L2 vertebral body (A). This lesion disclosed a hyposignal on T1-weighted image (B), which was enhanced after gadolinium-DOTA injection on T1-weighted image with fat suppression (Dixon sequence) (C).

Download figure to PowerPoint

image

Figure 3. . Multiple nodular vertebral lesions. The T2*-weighted sagittal MRI of the dorso-lumbar spine shows multiple areas of high signal intensity (A) corresponding to areas of low signal intensity on the T1-weighted image (B). Those lesions were enhanced after gadolinium-DOTA injection on T1-weighted image with fat suppression (Dixon sequence) (C).

Download figure to PowerPoint

Bone densitometry

Lumbar examination was performed in 41/55 patients. Whole-body examination was performed in 32 patients. The median lumbar BMD was slightly decreased compared to normal levels: 0.865 g/cm2 (range 0.611–1.538) (median lumbar Z score: − 0.43; range − 3.5–4.4). The median whole body BMD was 1.058 g/cm2 (range 0.858–1.247).

Factors associated with disease progression

1Table I lists initial clinical features of the patients according to MRI results. Patients with an abnormal MRI were slightly older and more often anaemic than the others (P = 0.05). In addition, they had a significantly higher β2-microglobulin level (1.8 v 1.4 mg/l) (P = 0.01). Other characteristics of the two groups were similar. Median follow-up after MRI was similar in the two groups: 26 months (range 5–56) and 25 months (range 5–56), respectively. During follow-up, 10 patients had disease progression, 8/17 patients with abnormal MR imaging and 2/38 patients with normal MRI. Among these 10 progressive patients, four (all with abnormal MRI) presented with vertebral fractures, associated with epiduritis in one patient, and six (four of them with abnormal MRI) developed a sustained increase in the MIg. Median time to disease progression was not reached in both groups but was significantly different for patients with normal versus those with abnormal patterns on MRI (P < 0.0001) (Fig 4). Beside MRI, the other prognostic features were a β2-microglobulin level geqslant R: gt-or-equal, slanted 2 mg/l (P = 0.02), a bone marrow plasmacytosis geqslant R: gt-or-equal, slanted 20% (P = 0.04) and to a lesser degree, a urinary Bence-Jones protein geqslant R: gt-or-equal, slanted 50 mg/d (P=0.07) and a serum peak level geqslant R: gt-or-equal, slanted 30 g/l (P = 0.14) (Table II). Using a multivariate analysis including these five variables, the only two independent significant prognostic parameters were abnormal MRI (P < 0.001, HR 30.4, 95% CI 4.3–213) and bone marrow plasmacytosis > 20% (P = 0.004, HR 16.4, 95% CI 2.6–104). The two patients with both bone marrow plasmacytosis > 20% and an abnormal pattern on MRI progressed within only 2 months. On the other hand, only one of the 33 patients without either of these two features had disease progression with a median follow-up of 25 months.

Table 1. Table I. Clinical characteristics of patients with normal or abnormal bone marrow MR pattern.Thumbnail image of
image

Figure 4. . Variable time to progression of asymptomatic stage I MM patients according to MRI status.

Download figure to PowerPoint

Table 2. Table II. Correlation of features at diagnosis with time to progression.*Thumbnail image of
  • a

    * As the number of patients who progressed was low, the median time to progression was not reached for most of the features. These two patients had β2-microglobulin levels of 1.7 and 1.5 mg/l and bone marrow plasma cells of 18% and 20%.

  • DISCUSSION

    1. Top of page
    2. Abstract
    3. PATIENTS AND METHODS
    4. RESULTS
    5. DISCUSSION
    6. References

    Most patients with MM are diagnosed with one or more symptoms which require immediate treatment. In recent years the increased frequency of multichemical blood screening has led to the incidental diagnosis of many patients in early phase of the disease and who are free of symptoms. These patients with asymptomatic MM do not require chemotherapy until disease progression, because there is no evidence that prompt treatment provides any survival benefit (Hjorth et al, 1993). However, a subgroup of these asymptomatic patients is at high risk for a serious complication and therefore might benefit from early chemotherapy. Several studies have defined criteria that distinguish patients with early disease progression from those with disease that may remain stable for many years (Dimopoulos et al, 1993; Facon et al, 1995; Kyle & Greipp, 1980; Weber et al, 1997; Wisloff et al, 1991). In most of them the presence of a lytic bone lesion was associated with early progression, usually within a year. In this report we show that, in patients without any lytic bone lesion on skeletal survey, abnormal spinal MRI was an independent prognostic feature of disease progression, whereas decreased lumbar bone mineral density assessed by dual-energy X-ray absorptiometry did not carry any adverse prognosis.

    The presence of an abnormal MRI pattern was not correlated with the degree of osteoporosis, evaluated by a very sensitive method: lumbar and total body bone densitometry. We confirmed that, in contrast to patients with stage III MM, patients with asymptomatic stage I MM have a roughly normal bone density (median lumbar Z score − 0.43) whatever their MRI pattern (Abildgaard et al, 1996). Moreover, bone density had no prognostic value in these patients. Thus, bone densitometry, which might be useful in some stage III MM patients, showing decreased lumbar BMD which may improve with effective chemotherapy (Mariette et al, 1992, 1995), appears to have, in most cases, no indication in patients with asymptomatic stage I MM.

    Spinal MRI is a sensitive technique to detect focal or diffuse bone marrow involvement in patients with MM (Bellaiche et al, 1997; Fruehwald et al, 1988; Kusumoto et al, 1997; Ludwig et al, 1987; Moulopoulos et al, 1994; Tertti et al, 1995). This procedure has been shown to be useful in cases where epiduritis was suspected and to define the staging and treatment of patients with solitary plasmacytoma of bone (Moulopoulos et al, 1993). In patients with stage III symptomatic myeloma, spinal MR imaging is abnormal in > 80% of the cases but does not provide information which could modify the treatment (Bellaiche et al, 1997; Kusumoto et al, 1997; Tertti et al, 1995). Changes in MRI seem to be fairly well correlated with response to chemotherapy (Moulopoulos et al, 1994) and therefore sequential MRI might be useful in some patients without any biological markers to assess the response to treatment, although this remains to be definitely established. In patients with stage I asymptomatic MM and one lytic bone lesion on skeletal survey, spinal MR imaging is abnormal in most cases: during the period of the present study it was performed in eight patients with a single lytic bone lesion and was abnormal in seven cases (data not shown). In these patients, spinal MR imaging patterns did not yield to changes in treatment which was required soon after diagnosis. In patients with stage I asymptomatic MM and negative skeletal survey, we detected an abnormal spinal MRI pattern in 31% (17/55) of the patients, which is in accordance with two other studies of the literature which included enough patients with asymptomatic stage I MM patients: 17/47 (36%) in the study by Weber et al (1997) and 7/24 (29%) in the study by Van de Berg et al (1996). The pattern of bone marrow involvement was focal in 14 patients (a single nodular lesion in nine patients and multiple nodular lesions in five patients) and diffuse in three patients. In our study, MRI of the lumbar spine and of the last three dorsal vertebras was performed in all patients but 19 patients also underwent thoracic examination. However, the latter could not detect lesions in patients whose lower spine examination (including T10, T11, T12, lumbar spine and sacrum) was normal. Contrary to Weber et al (1997), we did not detect any patients with a variegated pattern which, in any case, appears rare.

    An abnormal MRI pattern, when it is present, might be an additive tool to differentiate asymptomatic MM from MGUS. Indeed, in patients with MGUS, MRI is always normal (Bellaiche et al, 1997). The distinction between asymptomatic MM and MGUS is difficult and is based on bone marrow examination (Greipp & Kyle, 1983). Some authors take into account only the percentage of bone marrow plasma cells to diagnose MM. However, marrow infiltration is frequently uneven in MM and the failure to demonstrate plasma cells in excess of 10%, even on several marrow specimens, does not rule out the possibility of MM. Other cytologists take into account not only the percentage of bone marrow plasma cells but also the presence of plasma cell nuclear abnormalities to diagnose MM. Like Weber et al (1997), which included 12/47 patients with < 10% bone marrow plasma cells, we considered this latter definition of MM in our study. Interestingly, the incidence of patients with abnormal spinal MR imaging was the same in patients with < 10% bone marrow plasma cells (33%) and in those with > 10% (28%). Likewise, the risk of progression to symptomatic MM was not different in patients with more or less than 10% bone marrow plasma cells: 5/28 and 5/27 (17%), respectively. These data confirm that we dealt with patients with asymptomatic stage I MM and not with MGUS.

    The most interesting finding of the current study is the multivariate analysis of all prognostic features which points to the value of two independent parameters: an abnormal MRI pattern and marrow plasmacytosis. Marrow plasmacytosis had been already mentioned in a multivariate regression analysis as an independent prognostic feature in stage I asymptomatic MM (Facon et al, 1995). In another report evaluating spinal MRI in patients with asymptomatic MM, marrow plasmacytosis appeared as a prognostic factor in a univariate regression analysis (Weber et al, 1997; Moulopoulos et al, 1995). In our study, as in the study by Weber et al (1997), spinal MRI appeared to be of important prognostic value. Because the number of patients with different abnormal MRI patterns was small, we could not determine if one of the three observed patterns may also have prognostic value. The two patients with both bone marrow plasmacytosis > 20% and an abnormal pattern on MRI progressed within only 2 months. On the other hand, only one of the 33 patients with normal spinal MRI and marrow plasmacytosis < 20% had disease progression within a median follow-up of 25 months. Therefore spinal MR imaging appeared to be justified in stage I asymptomatic MM patients with negative skeletal survey, which must remain under close supervision when MRI is shown to be abnormal. Median time to progression in patients with abnormal spinal MRI seems to be in the order of 2 years. Whether or not the presence of abnormal MRI findings must lead to initiation of treatment, either with chemotherapy or bisphosphanates, needs further evaluation.

    References

    1. Top of page
    2. Abstract
    3. PATIENTS AND METHODS
    4. RESULTS
    5. DISCUSSION
    6. References
    • 1
      Abildgaard, N., Brixen, K., Kristensen, J.E., Vejlgaard, T., Charles, P. & Nielsen, J.L. (1996) Assessment of bone involvement in patients with multiple myeloma using bone densitometry. European Journal of Haematology, 57, 370376.
    • 2
      Bellaiche, L., Laredo, J.D., Liote, F., Koeger, A.C., Hamze, B., Ziza, J.M., Pertuiset, E., Bardin, T. & Tubiana, J.M. (1997) Magnetic resonance appearance of monoclonal gammopathies of unknown significance and multiple myeloma. Spine, 22, 25512557.
    • 3
      Dimopoulos, M.A., Moulopoulos, A., Smith, T., Delasalle, K.B. & Alexanian, R. (1993) Risk of disease progression in asymptomatic multiple myeloma. American Journal of Medicine, 94, 5761.DOI: 10.1016/0002-9343(93)90120-E
    • 4
      Facon, T., Menard, J.F., Michaux, J.L., Euller-Ziegler, L., Bernard, J.F., Grosbois, B., Daragon, A., Azais, I., Courouble, Y., Kaplan, G., Laporte, J.P., De Gramont, A., Duclos, B., Leonard, A., Mineur, P., Delannoy, A., Jouet, J.P., Bauters, F. & Monconduit, M. (1995) Prognostic factors in low tumour mass asymptomatic multiple myeloma: a report on 91 patients. European Journal of Haematology, 48, 7175.
    • 5
      Fruehwald, F.X., Tscholakoff, D., Schwaighofer, B., Wicke, L., Neuhold, A., Ludwig, H. & Hajek, P.C. (1988) Magnetic resonance imaging of the lower vertebral column in patients with multiple myeloma. Investigative Radiology, 23, 193199.
    • 6
      Greipp, P.R. & Kyle, R.A. (1983) Clinical, morphological and cell kinetic differences among multiple myeloma, monoclonal gammopathy of undetermined significance and smoldering multiple myeloma. Blood, 62, 166171.
    • 7
      Hjorth, M., Hellquist, L., Holmberg, E., Magnussen, B., Rodjer, S. & Westin, J. (1993) Initial versus deferred melphalan–prednisone therapy for asymptomatic multiple myeloma stage I: a randomized study. European Journal of Haematology, 50, 95102.
    • 8
      Kusumoto, S., Jinnai, I., Itoh, K., Kawai, N., Sakata, T., Matsuda, A., Tominaga, K., Murohashi, I., Bessho, M., Harashima, K. & Heshiki, A. (1997) Magnetic resonance imaging patterns in patients with multiple myeloma. British Journal of Haematology, 99, 649655.DOI: 10.1046/j.1365-2141.1997.4213236.x
    • 9
      Kyle, R.A. & Greipp, P.R. (1980) Smoldering multiple myeloma. New England Journal of Medicine, 302, 13471349.
    • 10
      Ludwig, H., Fruehwald, F.X.J., Tscholakoff, D., Rasoul, S., Neuhold, A. & Fritz, E. (1987) Magnetic resonance imaging of the spine in multiple myeloma. Lancet, ii, 364366.DOI: 10.1016/S0140-6736(87)92383-X
    • 11
      Mariette, X., Bergot, C., Ravaud, P., Roux, C., Laval-Jeantet, M., Brouet, J.C. & Fermand, J.P. (1995) Evolution of bone densitometry in patients with myeloma treated with conventional or intensive therapy. Cancer, 76, 15591563.
    • 12
      Mariette, X., Khalifa, P., Ravaud, P., Frija, J., Laval-Jeantet, M., Chastang, C., Brouet, J.C. & Fermand, J.P. (1992) Bone densitometry in patients with multiple myeloma. American Journal of Medicine, 93, 595598.DOI: 10.1016/0002-9343(92)90190-M
    • 13
      Moulopoulos, A., Dimopoulos, M.A., Alexanian, R., Leeds, N. & Libshitz, H.I. (1994) Multiple myeloma: MT patterns of response to treatment. Radiology, 193, 441446.
    • 14
      Moulopoulos, A., Dimopoulos, M.A., Smith, T., Weber, D.M., Delasalle, K.B., Libshitz, H.I. & Alexanian, R. (1995) Prognostic significance of magnetic resonance imaging in patients with asymptomatic multiple myeloma. Journal of Clinical Oncology, 13, 251256.
    • 15
      Moulopoulos, A., Dimopoulos, M.A., Weber, D., Fuller, L., Libshitz, H.I. & Alexanian, R. (1993) Magnetic resonance imaging in the staging of solitary plasmacytoma of bone. Journal of Clinical Oncology, 11, 13111315.
    • 16
      Tertti, R., Alanen, A. & Remes, K. (1995) The value of magnetic resonance imaging in screening myeloma lesions of the lumbar spine. British Journal of Haematology, 91, 658660.
    • 17
      Van de Berg, B.C., Lecouvet, F.E., Michaux, L., Labaisse, M., Malghem, J., Jamart, J., Maldague, B.E., Ferrant, A. & Michaux, J.L. (1996) Stage I multiple myeloma: value of MR imaging of the bone marrow in the determination of prognosis. Radiology, 201, 243246.
    • 18
      Weber, D.M., Dimopoulos, M.A., Moulopoulos, A., Delasalle, K.B., Smith, T. & Alexanian, R. (1997) Prognostic features of asymptomatic multiple myeloma. British Journal of Haematology, 97, 810814.DOI: 10.1046/j.1365-2141.1997.1122939.x
    • 19
      Wisloff, F., Andersen, P., Andersen, T.R., Brandt, E., Eika, C., Fjaestad, K., Gronvold, T., Holm, B., Lovasen, K. & Tjonnfjord, G.E. (1991) Incidence and follow-up of asymptomatic multiple myeloma. European Journal of Haematology, 47, 338341.