Conflict of interest: Nothing to report
Diffuse pattern of bone marrow involvement on magnetic resonance imaging is associated with high risk cytogenetics and poor outcome in newly diagnosed, symptomatic patients with multiple myeloma: A single center experience on 228 patients†
Article first published online: 28 MAY 2012
Copyright © 2012 Wiley Periodicals, Inc.
American Journal of Hematology
Volume 87, Issue 9, pages 861–864, September 2012
How to Cite
Moulopoulos, L. A., Dimopoulos, M. A., Kastritis, E., Christoulas, D., Gkotzamanidou, M., Roussou, M., Koureas, A., Migkou, M., Gavriatopoulou, M., Eleutherakis-Papaiakovou, E., Gika, D., Koutoulidis, V. and Terpos, E. (2012), Diffuse pattern of bone marrow involvement on magnetic resonance imaging is associated with high risk cytogenetics and poor outcome in newly diagnosed, symptomatic patients with multiple myeloma: A single center experience on 228 patients. Am. J. Hematol., 87: 861–864. doi: 10.1002/ajh.23258
- Issue published online: 23 AUG 2012
- Article first published online: 28 MAY 2012
- Accepted manuscript online: 14 MAY 2012 05:24AM EST
- Manuscript Accepted: 30 APR 2012
- Manuscript Received: 8 FEB 2012
Magnetic Resonance Imaging (MRI) and specific cytogenetic abnormalities offer important prognostic information for myeloma patients. However, limited data are available about the association between cytogenetic abnormalities and MRI patterns of marrow infiltration. To address this issue, we analyzed 228 consecutive newly diagnosed, symptomatic patients who were diagnosed and treated in a single center. On bone marrow MR images, 95 (41%) patients had diffuse, 94 (41%) had focal, 35 (15%) were normal, and 4 (1.7%) patients had variegated pattern of marrow infiltration. High risk cytogenetics were more commonly observed with diffuse MRI pattern (50% vs. 31% in focal and normal patterns). Patients with diffuse MRI pattern had poorer survival compared to others and responded better to novel agent-based therapies than to conventional chemotherapy (objective response: 88% vs. 46%, P < 0.001). There was a significant improvement of patients' survival with a diffuse MRI pattern when treated upfront with novel agents compared to conventional chemotherapy (47 vs. 24 months; P < 0.001). Diffuse MRI pattern along with ISS-3 and high risk cytogenetics could identify a very high risk group of patients with extremely poor median survival (21 months) and an only 35% probability of 3-year OS. Our study shows that symptomatic myeloma patients with a diffuse MRI pattern at diagnosis very often show high risk cytogenetic abnormalities and are benefiting from upfront novel agent-based therapies. Diffuse MRI pattern in combination with high risk cytogenetics and ISS-3 can identify a subset of myeloma patients with very poor prognosis who may need innovative treatment strategies and possibly more aggressive therapies. Am. J. Hematol., 2012. © 2012 Wiley Periodicals, Inc.
During the last decade, magnetic resonance imaging (MRI) is increasingly used for the evaluation of the bone marrow in patients with multiple myeloma (MM). MRI is more sensitive compared to conventional radiography in the detection of focal bone lesions and it provides useful information on the extent of bone marrow involvement in myeloma patients [1, 2]. MRI recognizes four patterns of marrow infiltration in myeloma: focal, diffuse, variegated, and normal . A diffuse MRI marrow pattern has been associated with poor prognosis in myeloma patients treated with either conventional chemotherapy or novel agent-based therapies [4, 5]. In the era of novel anti-myeloma drugs, it becomes necessary to recognize patients with high risk disease who may benefit from more aggressive treatment. To date, the definition of high risk myeloma is based mainly on stage, according to the International Staging System (ISS) and on cytogenetic abnormalities. The International Myeloma Working Group has recently suggested that the cytogenetically detected chromosomal 13 or 13q deletion, translocation t(4;14) and deletion 17p, and detection by fluorescence in situ hybridization (FISH) of t(4;14), t(14;16), and del17p are considered as poor risk features . Other cytogenetic aberrations, such as the add1q21 and the hypodiploidy have also correlated with poor prognosis [7–9].
MRI pattern of marrow infiltration has not been included in the criteria which define high risk myeloma and there are very limited data for the association between MRI pattern and cytogenetic abnormalities in MM. Thus, the aim of the present study was to evaluate possible correlations between cytogenetic findings and MRI patterns of marrow infiltration in patients with newly diagnosed, symptomatic myeloma who were diagnosed and followed in a single center.
Patients and Methods
Two hundred and twenty-eight consecutive patients with symptomatic MM at diagnosis were studied. These patients were diagnosed, treated, and followed in the department of Clinical Therapeutics of the University of Athens, Greece. All patients had a full skeletal survey at diagnosis using conventional radiography for the detection of osteolytic disease. MRI of the thoracic spine, the lumbar spine, and the pelvis was performed in all patients before the administration of any kind of therapy, including bisphosphonates.
Conventional cytogenetics and FISH analyses were also performed at diagnosis. Response to frontline therapy was evaluated using the IMWG criteria . The study was conducted after approval by the Ethics Committee/Institutional review board. The patients gave informed consent for including their data in this analysis, whereas they gave the consent that the Radiology department demands for the MRI performance. The cost for MRIs was covered by the insurance of the patients when there was clinical suspicion of bone involvement of the respective areas and by departmental grant in all other cases.
Evaluation of MRI pattern of marrow infiltration
T1-weighted (TR/TE: 641/10, turbo factor: 4), STIR (TR/TE/TI:2000/70/150), and contrast-enhanced T1-weighted MR images (TR/TE) were obtained in the sagittal plane for the thoracic spine and for the lumbar spine and in the axial plane for the pelvis with a 1.5 T unit (Phillips Medical Systems, Eindhoven, The Netherlands). MR images were analyzed for pattern of myelomatous involvement. The pattern of marrow involvement on MR images was characterized as: (1) normal when there was no evidence of abnormal signal intensity; (2) focal, which consisted of localized areas of abnormal marrow; the lesions are darker than yellow marrow and slightly darker than or isointense to red marrow on T1-weighted images, while on T2-weighted images focal lesions are brighter than both red and yellow marrow, and on enhanced T1-weighted images they enhance to various degrees; (3) diffuse, in which normal bone marrow signal intensity is completely absent; the intervertebral disks appear brighter than or isointense to the diseased marrow; there is a diffuse decrease in the bone marrow signal intensity of the marrow on T1-weighted images, a variable increase in the signal intensity of abnormal marrow on T2-weighted images, and after the administration of intravenous contrast, the abnormal marrow enhances and the intervertebral disks appear darker than the enhanced spine; and finally (4) variegated which consists of innumerable small foci of disease on a background of intact marrow, with small dark lesions on T1-weighted images which become bright on T2-weighted image and enhance after the administration of intravenous contrast [3–5].
Detection of chromosomal abnormalities
A combined methodological approach of G-banding karyotypic analysis and FISH was performed in all patients at the time of diagnosis. Bone marrow cells were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, antibiotics, and glutamine for 2 hr (direct preparations) and for 24–96 hr (short-term cultures). G-banding analysis was performed according to the European Cytogenetic Guidelines and Quality Assurance . The clonality criteria and the karyotypic description followed the recommendations of the International System for Human Cytogenetic Nomenclature . FISH was performed according to the Recommendations for FISH in MM on uncultured BM, either on cytoplasmic immunoglobulin-enhanced cells (cIg-FISH) or on nuclei from purified CD138+ plasma cells . Commercially available DNA probes were used for the detection of del17p, del13q, t(4;14), t(11;14) and t(14;16) (Abbott-VYSIS, IL) and add1q21 (Kreatech Diagnostics, Amsterdam, the Netherlands).
Comparisons for categorical variables among different groups were made with the χ2-test, using Fisher's exact test when appropriate. Time to event curves for PFS and OS were plotted with the method of Kaplan and Meier and comparisons among groups were performed with the log rank test. For multivariate analysis, factors with prognostic significance were introduced into a Cox proportional hazards model. SPSS v17.0 software was used for the statistical analysis.
Out of 228 patients, 50% were males. The median age of all patients was 67 years (range: 31–91 years), whereas 44% of patients were younger than 65 years. Thirty-one percent of patients had ISS-1 myeloma, whereas 36.5% had ISS-2 and 32.5% had ISS-3 disease. At diagnosis, 178 (78%) patients had lytic bone lesions in conventional radiography, 88 (39%) patients had anemia (Hb <10 g/dl), and 34 (15%) had serum creatinine of ≥2 mg/dl. After the diagnosis of symptomatic myeloma, 27% of patients received conventional chemotherapy, whereas 33% received thalidomide-based regimens, 27% bortezomib-based regimens, and 13% lenalidomide-based regimens. Treatment was given according to our departmental protocols at the time of myeloma diagnosis and the treating physicians were aware of the MRI and cytogenetic results; although these did not influence the treatment decision as our departmental protocols did not take into account these results for treatment decisions.
MRI patterns and patients' characteristics
According to MRI bone marrow involvement, 95 (41%) patients had diffuse pattern of infiltration, 94 (41%) had focal pattern, 35 (15%) had normal pattern, and only 4 (1.7%) patients had variegated pattern (Table I). Because of the small number of patients with variegated pattern in this study and because we have already observed that patients with variegated pattern have similar prognosis with patients with diffuse pattern [4, 5], we have grouped and analyzed these two patterns together. As depicted in Table I, patients with diffuse and variegated patterns had more often severe anemia (51%), extensive marrow infiltration (>40% of bone marrow infiltration by myeloma cells in 88%), higher incidence of increased serum LDH ≥ 300 IU/l (15%), and ISS-3 stage (45%) compared to patients with focal and normal MRI patterns. Furthermore, patients with focal and diffuse patterns had more often osteolytic lesions on conventional radiographs compared to patients with normal pattern (91% vs. 75% vs. 49%, respectively).
|Normal||Focal||Diffuse + variegated||P value|
|All patients (N = 228)||35 (15%)||94 (41%)||99 (43%)|
|Age > 65 years||60%||61%||50%||0.255|
|Performance status ≥2||17%||53%||51%||<0.001|
|Lytic bone lesions||49%||91%||75%||<0.001|
|Hemoglobin <10 g/dl||34%||26%||51%||0.002|
|Platelet counts <130 × 109/l||3%||7%||11%||0.223|
|eGFR <50 ml/mi/1.73 m2||31%||18%||34%||0.030|
|Bone marrow PCs ≥40%||60%||56%||88%||<0.001|
|LDH ≥300 IU/l||0||7%||15%||0.021|
MRI patterns and cytogenetic abnormalities
Of the studied population, 38% of the patients had del13q detected by FISH, 27% had add1q21, 12% had del17p, 10.5% had t(11;14), 7% had t(4;14), 3.4% had t(14;16), whereas 8% of the patients had hypodiploidy. High risk cytogenetics, defined by the presence of any of del13q detected by conventional cytogenetics, hypodiploidy, t(4;14), t(14;16), del 17p, add1q21, was present in 40% of patients. Figure 1 depicts the incidence of the above cytogenetic abnormalities in each MRI pattern. Thus, del17p was more common in patients with diffuse MRI pattern (P = 0.038), whereas there was a trend for higher incidence of add1q21 and del 13q in patients with diffuse MRI pattern (P = 0.08 and P = 0.1, respectively). There were no significant differences in the frequency of t(4;14), t(11;14) and t(14;16) among patients with different MRI patterns. However, 50% of patients with diffuse MRI pattern had high risk cytogenetics versus 31% of patients with focal pattern and 31% of patients with normal pattern (P = 0.07; Fig. 1).
MRI patterns, cytogenetic abnormalities, and survival
Objective response (complete and partial responses) to first line therapy was similar for patients with different MRI patterns: 78%, 74%, and 70% for diffuse, focal, and normal pattern, respectively (P = 0.576). The median overall survival (OS) of the whole cohort was 53 months. Patients with a normal bone marrow MRI pattern had a median OS of 102 months (95% CI: 43–161 months), whereas patients with a focal pattern had a median OS of 60 months (95% CI: 41–79 months) and patients with a diffuse pattern had a median OS of 40 months (95% CI: 20–61 months; P < 0.001; Fig. 2A). Patients with high risk cytogenetics had a poorer median OS compared to those with standard risk cytogenetics (40 months vs. 57 months; P < 0.001; Fig. 2B), although the vast majority of high risk patients (93%) received novel agent-based therapies upfront. Patients who received frontline therapy with novel agents had a better median OS than those who received conventional therapy (102 months vs. 39 months; P = 0.012). This was more evident in patients with diffuse MRI pattern (Fig. 3).
A multivariate model, which did not include cytogenetic abnormalities, showed that independent prognostic factors for poor survival were: older age (>65 years, HR: 1.9, 95% CI: 1.2–3.1, P = 0.006; or >75 years, HR: 3.2, 95% CI: 1.8–5.7, P < 0.0001), ISS-3 (HR: 2.9, 95% CI: 1.5–5.5, P = 0.001), upfront therapy with conventional chemotherapy (HR: 1.9, 95% CI: 1.2–3.0, P = 0.003), elevated LDH ≥300 U/l (HR: 2.5, 95% CI: 1.2–5.1, P = 0.012), and diffuse MRI pattern (HR: 2.6, 95% CI: 1.0–6.4, P = 0.03). However, when high risk cytogenetics were included into the multivariate model, only high risk cytogenetics (HR: 3.2, 95% CI: 1.5–7.1, P = 0.004), ISS-3 (HR: 2.3, 95% CI: 1.1–5.5, P = 0.027), and high LDH (≥300 IU/l, HR: 2.0, 95% CI: 0.7–6.0, P = 0.047) were independently associated with poor survival.
Subsequently, we examined whether we could combine prognostic information obtained from MRI and cytogenetics to identify patients at very high risk, who should probably be considered for more intensive treatments. Thus, by using high risk cytogenetics, advanced ISS and diffuse MRI pattern we were indeed able to identify a subgroup of patients with extremely poor prognosis. More specifically, patients who had high risk cytogenetics, ISS-3, and diffuse MRI pattern had a median OS of only 21 months (95% CI: 6–35 months) and a probability for 3-year OS of 35%. This group of patients represented 10% of the patients in our cohort and they had all received upfront novel agent-based therapy. On the contrary, patients with only one or two of the above adverse disease features (high risk cytogenetics or ISS-3 or diffuse MRI pattern) had a median survival of 50 months (95% CI: 41–57 months) and a 62% probability of 3-year OS, whereas patients with standard risk cytogenetics, ISS-1 or -2 and normal or focal MRI pattern had a median survival which has not yet been reached and a 92% 3-year survival probability (Fig. 4).
MRI pattern of marrow involvement and specific cytogenetic abnormalities have been associated with prognosis in newly diagnosed patients with symptomatic myeloma [4–9]. However, there are very limited data in the literature for the correlation between MRI patterns and cytogenetic aberrations in myeloma patients. To address this issue, we analyzed 228 consecutive patients who were diagnosed and treated in our center and had available both MRI and cytogenetic data. We found that patients with a diffuse MRI pattern had a higher incidence of high risk cytogenetic features, including del17p, add1q21, and del 13q compared to patients with focal or normal MRI patterns. This result is described for the first time and is in accordance with a recent observation of the Heidelberg group in 87 untreated myeloma patients where the gain of 1q21 and the deletions of 17p and 13q significantly correlated with at least one abnormal finding in bone marrow dynamic contrast-enhanced MRI, that is, aberrant “focal” microcirculation pattern, increase in median microcirculation parameter amplitude A or exchange rate constant kep, suggesting that these chromosomal abnormalities trigger the angiogenic cascade in MM . Indeed, our group has described that patients with diffuse MRI pattern have increased microvessel density in their trephine biopsies and adverse myeloma features . In this study, we confirm that a diffuse MRI pattern correlated with adverse disease characteristics, such as ISS-3 stage, elevated LDH, severe anemia, and poor prognosis. In the absence of cytogenetics in the multivariate model, diffuse MRI pattern showed a strong prognostic value for poor survival, independent of the type of frontline anti-myeloma therapy. We and others have previously described that diffuse MRI pattern correlates with shorter survival when conventional chemotherapy or novel agent-based regimens are used as first line therapy in symptomatic myeloma [4, 5, 15]. The Arkansas group showed that the presence of more than seven focal lesions on whole body MRI is also an independent prognostic factor for OS in patients who received thalidomide-based first line therapy . In this study, we showed that patients with diffuse MRI patterns have almost two-fold longer median OS with the use of novel agent-based therapies upfront compared to conventional treatment. Despite the effect of MRI pattern on survival, when cytogenetic results were entered into the multivariate model, diffuse MRI pattern lost its independent significance for poor survival. This is possibly due to the strong correlation of diffuse MRI pattern with high risk cytogenetics, as 50% of patients with diffuse pattern had such adverse cytogenetic features. The Arkansas group also showed that in the presence of adverse cytogenetic abnormalities, MRI findings (presence of more than seven focal lesions) have either borderline (if FISH is used) or no (if conventional cytogenetics is used) independent significance for OS in patients treated with novel agent-based therapies upfront . Even though the Arkansas group has used a different way to express the MRI findings (presence of focal lesions) than we did in our analysis (pattern of marrow involvement), it seems that high risk cytogenetics have a stronger prognostic value than MRI. However, using MRI pattern, we were able to define a group of myeloma patients with very high risk disease. Indeed, patients with diffuse MRI pattern, high risk cytogenetics and ISS-3 had a median survival of only 21 months, despite the use of novel agent-based therapies upfront in all of them. This is extremely important as we urgently need methods to identify high risk patients with very poor outcome who may need innovative treatment strategies and possibly more aggressive therapies.
We conclude that in symptomatic myeloma patients with diffuse MRI pattern at diagnosis there is a correlation with high risk cytogenetic abnormalities and poor survival. These patients have to be treated upfront with novel agent-based therapies. Diffuse MRI pattern in combination with high risk cytogenetics and ISS-3 can identify a subset of myeloma patients with very poor prognosis who may possibly require more aggressive management.
- 13Groupe Français de Cytogénétique Hématologique (GFCH). Recommendations for the cytogenetic management of multiple myeloma proposed by the French Group for Cytogenetic Hematology. Pathol Biol (Paris) 2004; 52: 263–264.