Regression of Myelofibrosis and Osteosclerosis following Hematopoietic Cell Transplantation Assessed by Magnetic Resonance Imaging and Histologic Grading

Article Outline

• Abstract
• Introduction
• Methods
  • Patients
  • Study Design
  • Marrow Biopsies
  • MR Technique
  • Transplantation
  • Consent
• Results
  • Pretransplantation Assessment of Myelofibrosis
  • Longitudinal Assessment of Changes in Myelofibrosis
  • Correlation of Marrow Features and Clinical Course
• Discussion
• Acknowledgments
• References
• Copyright

Abstract 

Myelofibrosis is a reactive, often inhomogeneous process in the marrow cavity, and sampling errors on biopsies obtained to diagnose and monitor the course of myelofibrosis have been a constant problem in hematopathology. We investigated the potential utility of magnetic resonance imaging (MRI) of the lumbar spine, pelvis, and femora as a diagnostic and monitoring technique for assessment of myelofibrosis. Findings on serial marrow biopsies were correlated with T1-weighted spin-echo and short inversion time inversion recovery (STIR) images in patients with chronic idiopathic myelofibrosis or myelofibrosis developing from polycythemia vera or essential thrombocythemia who underwent hematopoietic cell transplantation (HCT). Thirty-five patients were studied before HCT; 11 were followed for 3 months and 10 patients for ≥1 year after HCT with sequential marrow biopsies and MRI studies. MRI allowed direct visualization of the biopsy sites and correlation of histologic and MRI findings. MRI also provided assessment of the extent and degree of myelofibrosis in a large volume of the skeletal marrow. There was good correlation between biopsy results and MRI findings at specific biopsy sites and between successful HCT and resolution of fibrosis and osteosclerosis as determined by MRI. We conclude that in patients with myelofibrosis, MRI of the skeleton provides a comprehensive assessment of the pattern and extent of fibrosis and allows for correlation with biopsy findings. In patients undergoing HCT, MRI accurately reflects response or progression of marrow disease.

Key words: Myelofibrosis, Histology, Osteosclerosis, Magnetic resonance imaging, Hematopoietic cell transplantation

Back to Article Outline

Introduction 

Fibrosis of the bone marrow is a reactive process that may develop in patients with various malignant and nonmalignant disorders [1]. It is an essential feature of chronic idiopathic myelofibrosis with myeloid metaplasia and a frequent finding in advanced polycythemia vera and essential thrombocythemia. Chronic idiopathic myelofibrosis, polycythemia vera, and essential thrombocythemia are chronic myeloid disorders [2] and are generally easily controlled over extended periods with conservative management. However, for more advanced disease, few treatment options are available, and the only current strategy with curative potential is hematopoietic cell transplantation (HCT) [3, 4, 5, 6].

A major concern in previous studies of HCT was that myelofibrosis would interfere with engraftment of donor cells because of the severely damaged microenvironment in the marrow cavity. Some studies have supported that notion [7]. However, although the time course of donor cell engraftment tends to be slower than in patients without fibrosis, more recent trials have shown that marrow fibrosis is not a significant risk factor for graft failure [8, 9, 10]. Marrow fibrosis generally regresses after HCT [5, 6], although the kinetics have been poorly characterized.

Unfortunately, the sampling error with needle biopsies of the marrow is high because myelofibrosis is an inherently heterogeneous process. Even sampling of 5 marrow sites at autopsy, totaling up to 1700 mm² cross-sectional areas of medullary marrow, shows great variation, although such extensive sampling yields approximately 100 times the area (volume) of a sample obtained by standard needle biopsies in a living patient, which average 15 mm² at best. In contrast, the field of view of a magnetic resonance imaging (MRI) study of the pelvis covers 3-4 vertebral bodies, iliac bones, femoral heads, and proximal femoral shafts. Combined, these areas cover >200 cm² of bone compared with 17 cm² for a 5-specimen autopsy marrow examination or 0.15 cm² in a needle biopsy. The needle biopsy, in other words, samples only 1/1200 of the area visualized by a pelvic MRI. In addition, MRI obtained after the marrow biopsy documents the needle track and, thus, the site of marrow sampling; hence, the combination of marrow biopsy and MRI allows a precise assessment of the extent and pattern of fibrosis and provides an indication of whether or not the marrow biopsy was representative of the overall marrow appearance.

The present study was intended to improve the assessment and staging of marrow fibrosis in patients undergoing HCT, to describe the kinetics of resolution of the marrow pathology, and to correlate these findings with the clinical course of patients after HCT. We also investigated whether MRI was capable of assessing the response of marrow fibrosis to allogeneic HCT. We report data on the concurrent assessment of myelofibrosis by histology and MRI in 35 patients, 21 of whom were reassessed sequentially for ≥3 months after allogeneic or autologous HCT.

Back to Article Outline

Methods 

Patients 

Thirty-five patients were enrolled. Patient and disease characteristics are summarized in Table 1. Chromosome analysis showed a normal karyotype in 16 patients, abnormalities of chromosome 20 in 4, abnormalities of chromosome 9 in 3, complex abnormalities in 3, and various structural or numerical abnormalities in 9. In 1 patient, the karyotype could not be determined.

Table 1. Patient, Disease, and Transplant Characteristics

Characteristic	Number of Patients
Number of patients	35
Age (y), range (median)	25-66(51)
Male/female	16/19
Diagnosis
CIMF	19
ET	9
PV	7
Disease duration (ms), range (median)	6-313(25)
Lille stage at HCT⁎
0	17
1	13
2	2
Cytogenetics†
Normal	16
Abnormal (see text)	18
Donor‡
HLA-identical sibling	12
HLA-identical unrelated	17
Syngeneic twin	1
Autologous	2
Conditioning regimen
tBUCY (ATG)	26(10)
FlutBU (ATG)	2(1)
BU	2
BU TBI	1
FluBU	1
GVHD prophylaxis
CSP/MTX	27
Other	2
None	3
Source of stem cells:
Peripheral blood	28
Bone marrow	4
No transplantation	3

ATG, antithymocyte globulin; BU, busulfan; CIMF, chronic idiopathic myelofibrosis; CSP, cyclosporine; ET, essential thrombocythemia; Flu, fludarabine; FlutBU, fludarabine plus targeted busulfan; GVHD, graft-versus-host disease; HCT, hematopoietic cell transplantation; MTX, methotrexate; PV, polycythemia vera; TBI, total body irradiation; tBUCY, targeted busulfan plus cyclophosphamide.

Study Design 

All patients underwent bone marrow biopsies from the posterior iliac crests before HCT and were scheduled to undergo sequential biopsies after HCT. The protocol called for biopsies followed within 2-3 days by pelvic MRI. The initial biopsy and MRI were obtained within 2-3 weeks before HCT, again at 3 months, and 1-2 years after transplantation. We were able to complete this sequence in 10 patients to ≥1 year after transplantation and in an additional 11 patients to 3 months. Among the remaining 14 patients, 3 did not undergo transplantation, and 11 declined to have repeat marrow biopsies or MRI after HCT or were unable to return to the transplantation center for re-evaluation.

Marrow Biopsies 

Marrow core biopsies were obtained in standard fashion from the posterior iliac crests with a Jamshidi biopsy needle. The patient was prone and received local anesthesia with bicarbonate buffered lidocaine. With subsequent biopsies, care was taken to avoid previous biopsy sites. After touch preparation, the biopsy was fixed in B5 or formalin and briefly decalcified before paraffin embedding. Histologic assessment of myelofibrosis was based on sections stained with hematoxylin and eosin in parallel with reticulin silver and periodic acid-Schiff stains on all biopsy sections at 5 μm using standard histologic microtomy methods. Marrow fibrosis was graded as described elsewhere [11, 12]. Trichrome stains were not routinely performed because this stain is not specific and is not used in the scale for myelofibrosis by Dekmezian et al [12]. In addition to reticulin and collagen fibrosis, extent of osteosclerosis was determined. Because currently there is no universally accepted classification scheme for osteosclerosis, we devised a semiquantitative grading for the present study (Table 2). The minimum requirement for including biopsy samples for measurement was the presence of ≥4 intact trabeculae ≥1 mm deep to the cortex of the bone. A stage micrometer was used to calibrate an eye micrometer. Estimates of the trabecular width were made at right angles to the axis of the trabecula being measured at multiple points along each measurable trabecula. More advanced cases showed knobby expanded ends on the trabeculae in addition to increased width.

Table 2. Grading of Osteosclerosis

Aspirate smears and particle sections were also prepared whenever specimens were adequate. Sections and smears were examined microscopically using a Leitz Laborlux microscope with a Leica camera attachment (Leica Microsystems, Wetzlar, Germany).

MR Technique 

The pelvic MRI technique employed T1-weighted spin-echo (TE) and short inversion time inversion recovery (STIR) techniques in the coronal and axial (transverse) planes and was designed to assess the medullary cavities of both iliac bones, the lower lumbar vertebrae, both femoral heads, and the proximal femoral shafts. Imaging was done using a Siemens Harmony 1.0-T Magnetic Resonance System (Siemens Medical Solutions, Malvern, Penn).

Table 3 schematically summarizes the appearances expected on the TE and STIR images for 6 different marrow elements, namely fibrosis, tumor, fluid, fat, normal hematopoietic cells, and osteosclerosis. Interpretation of marrow requires the understanding that the TE signal is light (high) in fatty areas and dark (low) in areas of fluid, fibrosis, sclerosis, and tumor. On STIR images, however, fibrosis and fat are dark, whereas water-containing components such as fluid or tumor are of high signal intensity and, therefore, are portrayed as bright. Normocellular marrow is approximately half fat and half water and, hence, is of intermediate brightness with both methods. Osteosclerosis and cortical bone are dark to very dark on TE and STIR images. Recent needle biopsy tracks (edema, hemorrhage) show a bright linear focus of fluid on the STIR image and a correspondingly dark appearance on TE images. Older biopsy tracks (after several weeks) generally display a dark linear image on TE images with variable intensity on STIR images (see Figures 4A and 4B).

Table 3. Patterns of Magnetic Resonance Images

TE, T1 spin echo; STIR, short inversion time inversion recovery.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 4. 

  Needle tracks. Old and new needle biopsy sites are visible on TE and STIR sequences (patient 20). A, TE series regressing from TE 4 homogeneous (H) to TE 2H at 1 year. Biopsy sites appear as vertical dark lines (black arrowheads) that remain visible in all 3 images. STIR image regresses from 2H to 1H at 1 year. The right panel (1-year follow-up) shows a new needle track, which is better recognized as new on the STIR image. B, STIR image at 1 year follow-up (compare with A, lower right). A new needle site is identified on the far right (arrow), with a crescent of bright periosteal edema and a linear white line showing the needle track filled with fluid. An old needle site is present on the left side (yellow arrowhead).

These patterns on MRI allow a broad assessment of a very large marrow volume and locations of biopsy sampling sites. Therefore, the combination of biopsy and MRI should provide a reliable estimate of the extent and severity of myelofibrosis and an indication of how representative a concurrent biopsy sample might be.

Transplantation 

Donor and transplant characteristics are summarized in Table 1.

Patients were conditioned for HCT as described previously [6]: 26 received targeted busulfan (BU) plus cyclophosphamide (with the addition of thymoglobulin in 10) [13], 2 received fludarabine plus targeted BU [14] (1 with the addition of thymoglobulin), and 1 each received BU plus total body irradiation (800 cGy) [15] or fludarabine plus total body irradiation (200 cGy) [16]. The 2 autologous transplant recipients received only BU [17]. Graft-versus-host disease prophylaxis for the 29 allogeneic transplant recipients consisted of cyclosporine plus methotrexate in 27, and 1 each received methotrexate plus sirolimus or cyclosporine plus mycophenolate mofetil, respectively [6].

Twelve patients received transplants from HLA-identical siblings, 17 from unrelated donors who were HLA-identical with the patients by high-resolution typing, 1 from a syngeneic twin, and 2 received autologous peripheral blood progenitor cells that had been harvested after granulocyte colony-stimulating factor stimulation and cryopreserved until HCT [17]. Three patients underwent the initial evaluation but, for various reasons, did not proceed to HCT.

At 3 months, before being discharged from the transplantation center, patients underwent a complete “departure workup,” including marrow biopsy and MRI. Patients who at that time were receiving treatment for graft-versus-host disease with glucocorticoids were started on prophylactic bisphosphonates (usually Fosamax 35 mg once a week). Subsequently, patients were cared for by their hometown physicians and returned to the transplantation center for a re-evaluation approximately 1 year after transplantation.

Consent 

All patients gave informed consent for enrollment into this study according to the procedures required by the institutional review board of the Fred Hutchinson Cancer Research Center (Seattle, Wash).

Back to Article Outline

Results 

We first carried out a cross-sectional review of biopsies before HCT and MRI findings in all 35 patients to describe the spectrum and patterns of marrow fibrosis (and osteosclerosis) that were observed. We next reviewed sequential studies in the same patients to define longitudinal (chronologic) changes in biopsy and MRI, and we correlated those findings with the patients’ clinical course. Histologic and MRI analysis occurred independently (GES and BAP, respectively); MRI reading was blinded as to biopsy findings and patient condition.

Pretransplantation Assessment of Myelofibrosis 

The initial biopsies, available in all 35 patients, showed fibrosis grades ranging from 2 to 4 as described previously [11, 12]. Osteosclerosis ranged from absent (grade 0) to severe (grade 6). MRI findings before HCT demontrated several patterns as illustrated in Figure 1, Figure 2, Figure 3, Figure 4, ranging from diffuse fibrosis to patchy patterns. There was no correlation of MRI findings and Lille score [18].

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Patient with complete resolution of myelofibrosis 1 year after transplantation. A, Bone marrow reticulin stain before transplantation. Diffuse reticulin positivity (300×). B, Reticulin stain at 1 year after transplantation (grade 0-1). C, T1 spin-echo (TE) sequence shows progressive diminution of the dark (low signal) medullary pattern (TE 4 homogeneous [H]) and transition to a pale gray (intermediate) pattern typical of more normal marrow at 3 months (TE 2H), and return to near normal, with scattered foci of low signal at 1 year after transplantation (TE 1.5H). This coronal section shows femoral and acetabular regions of the ilia and portions of the lower lumbar vertebrae on a selected image of 20 acquired through the pelvic marrow. D, TE transverse images with the same timing as in C; similar changes from dark to light are seen in the posterior iliac crests. E, STIR images show progression from moderately light (STIR 3, mildly hypercellular) to dark (STIR 0.5, normocellular) with resolution of fibrosis; same time sequence as in C and D.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  Patient with refractory disease. A, Bone marrow reticulin stain before transplantation, grade 3-4 fibrosis (300×). B, Reticulin stain 1 year after transplantation; no improvement (fibers are still dense 300×). C, TE coronal MRI sequence documents fibrosis before transplantation (TE 4 homogeneous [H]) with little or no change at 3 months (TE 4H) and 1 year (TE 4 patchy [P]). There is only slight improvement in the femoral head appearance at 1 year. D, TE MRI of posterior iliac crests shows lack of change in improvement of skeletal appearance over 1 year. E, STIR sequence shows only modest change from before transplantation (STIR 3H) to 1 year (STIR 1P). This change was interpreted as primarily due to decreased fluid content without much change in fibrosis. Note fresh needle site on the right side of the left image, with extraosseous and intraosseous edema seen as a bright white linear signal (arrowhead) and a linear vertical line. This image was representative of the marrow as a whole.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 3. 

  Heterogeneous fibrosis, TE coronal view. Heterogeneous pattern with a focal high signal region in the medial right femur (arrow) indicats an area of uninvolved hypocellular normal fatty marrow.

Longitudinal Assessment of Changes in Myelofibrosis 

Data on 21 patients in whom sequential studies were obtained are summarized in Table 4, Table 5. MRI allowed a distinction of uniform from heterogeneous patterns of myelofibrosis in 3-4 lumbar vertebrae, both femoral heads, shafts, and iliac crests. Recent and old biopsy sites were identified, and agreement (or disagreement) between the biopsy sites and the general marrow appearance was determined.

Table 4. Course and Outcome in 11 Patients with Myelofibrosis followed for 3 Months after Hematopoietic Cell Transplantation

H, homogeneous; MRI, magnetic resonance image; OS, osteosclerosis grade; P, patchy; STIR, short tau inversion recovery; TE, T1-weighted spin echo; NED, no evidence of disease; GVHD, graft-versus-host disease.

Table 5. Course and Outcome in 10 Patients with Myelofibrosis followed for at least 1 Year after Hematopoietic Cell Transplantation

MF, myelofibrosis grade; OS, osteosclerosis grade (OS grading was not available on all samples); TE, T1 spin echo (grades 1-4); STIR, short inversion time inversion recovery (grades 1-4); H, homogenous; P, patchy; MRI, magnetic resonance imaging; NA, not available; NED, no evidence of disease.

Among the 11 patients with follow-up studies for 3 months (Table 4), 1 showed complete clearing of myelofibrosis by biopsy and MRI over this interval. The remaining patients showed minimal evidence of changes in the marrow, although splenomegaly was usually markedly decreased.

Among 10 patients followed for ≥1 year (Table 5), complete clearing of fibrosis was observed in 6. In 3 patients, the degree of fibrosis decreased from grades 3-4 to grades 1-2, and in 1 patient from grade 4 to grade 3. However, in this patient, fibrosis subsequently progressed again, associated with disease recurrence as also documented by reappearance of patient cells (see below).

When there was discordance between biopsy results and overall MRI findings, the needle site was usually shown to be in marrow different in appearance from that visualized elsewhere, ie, located in an area of marrow not considered representative. A representative biopsy was defined as a site that matched the overall appearance of the anatomic areas visualized on MRI (posterior iliac crests, the acetabular region of the ileum, the femoral heads and upper shafts, and the lumbar vertebrae). MRI findings were graded from 1 to 4 (for fibrosis intensity) and designated as homogeneous or patchy on TE and STIR images in each case at each time point studied (before HCT, 3 months, and ≥1 year).

In patients with homogeneous patterns, the biopsy site was deemed representative by definition. Discrepancies were possible, however, in cases with patchy distributions of fibrosis. Here, the identification of the needle track was useful in assessing possible sampling “error.”

The resolution of fibrosis appeared fastest in the femora, intermediate in iliac crests, and slowest in the vertebrae.

Detailed image sequences of representative patients, 1 whose disease resolved and 1 whose disease was refractory, are illustrated in Figure 1, Figure 2. The MRI studies of patient 13 showed regression from a TE image with diffuse dark marrow (TE 3 homogeneous) to a light (high signal) homogeneous pattern (TE 0) characteristic of clearing fibrosis over a period of 15 months (Figures 1A and 1B). The STIR images showed a hypercellular-appearing (STIR bright) pattern correlating with marrow histology by biopsy, which showed high cellularity along with interstitial fibrosis before transplantation. Over 1 year, the STIR pattern reverted to a normal cellularity appearance. The myelofibrosis grade (3-4) and osteosclerosis grade (4) decreased to 0 over that interval. Patient 19 (Figure 2) showed refractory disease. The scores before HCT were 4 for fibrosis and 4 for osteosclerosis, with a suggestion of histologic improvement at 3 months, but at 14 months diffuse fibrosis was present with little or no change by MRI on the TE or the STIR images. Clinically the patient showed disease progression.

Several patients showed a heterogeneous MRI pattern before HCT, and the extremes of regression versus refractory disease were more difficult to assess. Patient 12, for example, showed a large cellular patch of fatty hematopoiesis in the femoral neck region (Figure 3), which persisted throughout the year’s follow-up. Other marrow regions remained patchy by MRI, although there was a heterogeneous pattern of improvement.

In patients with patchy MRI patterns, correlation with the biopsy site data appeared particularly important. For example, Figures 4A and 4B show new and old needle sites, with the new sites displaying edema fluid as white lines in the posterior iliac crest and a white fluid crescent (edema) adjacent to the periosteum. This finding was easily interpreted as an indication that the biopsy site was not representative of the overall marrow status when it was compared with the marrow in the iliac crests and other bones within the pelvic MRI study.

A subset of 10 patients (including 2 who had MRIs only until 3 months; Table 4) had serial biopsies evaluable for osteosclerosis up to 1 year. Six of the 10 patients showed complete resolution of osteosclerosis, and 4 showed regression by ≥2 grades. When total histologic scores (myelofibrosis plus osteosclerosis) were tallied for this subset of 10 patients, the median improvement in total scores was from 7.0 before transplantation to 0.5 after transplantation. Histologic features of 1 of these patients are shown in Figure 5, Figure 6.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 5. 

  Osteosclerosis. A, Before transplantation, an extremely wide trabecular bone with multiple osteoid lines and knobby projections characterizes this patient’s marrow biopsy (osteosclerosis grade 5). B, One year after transplantation, the average trabecular width was within normal limits (osteosclerosis grade 1). Hematoxylin and eosin stain, 250×).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 6. 

  Documentation of fibrosis by vimentin staining. A, Vimentin immunohistology (immunoperoxidase stain) shows the linear staining of fibroblasts and round cytoplasmic staining of other mesenchymal cells before transplantation. B, At 1 year after transplantation, the vimentin staining shows nearly total disappearance of the linear markings of fibroblasts.

Together the data suggest that ≥3 months and, more frequently, 1 year of follow-up was required to confidently diagnose histologic and MRI resolution of marrow fibrosis after HCT.

Correlation of Marrow Features and Clinical Course 

Details of transplantation results have been presented elsewhere [6, 17]. Clinical data in comparison with biopsy and MRI findings are summarized in Table 4, Table 5. In the group of 11 patients who had the last MRI at 3 months (Table 4) but continued to be followed, 7 showed partial clearing. There were 2 additional patients who showed histologic clearing at 1 year (but did not have MRI data available at 1 year) and no evidence of disease. Two patients, 1 of whom had received an autologous transplant (patient 4), showed no regression of fibrosis at 3 months by MRI or histologically and subsequently succumbed to their disease. One patient showed significant clearing of marrow fibrosis already by 3 months (patient 3) and was shown to be a complete donor chimera.

All 10 patients who had longitudinal biopsy and MRI studies to 1 year achieved engraftment (Table 5). Six patients in this group who showed complete resolution of fibrosis at 1 year also were complete donor cell chimeras (100% of CD3⁺ and CD33⁺ cells in the marrow and blood were donor derived). One patient, who had received stem cells from an unrelated donor (patient 19), never showed any clearing of marrow fibrosis and had disease progression. Three patients in this group showed only partial clearing by 1 year but subsequently all achieved normal hematopoietic parameters with complete donor chimerism; no follow-up MRI studies were available.

Back to Article Outline

Discussion 

In patients with hematologic malignancies, marrow fibrosis is a secondary phenomenon due to abnormal cytokine secretion by clonogenic hematopoietic precursor cells. Fibrosis may lead to severe clinical manifestations due to peripheral blood cytopenias, splenomegaly, and hypermetabolism. Manifestations are progressive and generally refractory to conventional therapy. We and others have shown that HCT can cure the underlying hematopoietic malignancy and lead to regression of fibrosis [5, 6, 10, 17]. However, accurate assessment of the degree of marrow fibrosis (and response to therapy) has been notoriously difficult, requiring serial marrow biopsies that are subject to significant sampling error. It has been suggested that the use of MRI might be useful in disease assessment [19, 20].

In the present study we showed that MRI can be used to assess the extent of marrow fibrosis and to monitor marrow responses to treatment with HCT or disease recurrence and progression. In successfully treated patients, regression of fibrosis and osteosclerosis with bone remodeling was apparent as soon as 3 months after HCT. The tempo of regression of fibrosis and osteosclerosis differed between patients, but a correlation with other parameters or predictive factors was not possible within this limited cohort of patients. Response assessment was more difficult in patients who presented with a patchy pattern of fibrosis. It is conceivable that the kinetics of resolution were related to the clearance of residual clonal cells of the underlying malignancy, and prolonged mixed chimerism, ie, the persistence of patient cells in addition to engrafted donor cells after transplantation, as has been observed in patients with myelofibrosis [6]. In support of such a concept is the fact that, in patients in whom the hematologic malignancy recurred, as determined by the presence of clonal chromosomal markers or progressive increase in the proportion of patient cells, MRI studies showed regrowth or progression of marrow fibrosis. The number of patients with progressive/recurrent disease in the present cohort was too small to determine whether MRI findings revealed evidence of disease recurrence before it was evident clinically or on the basis of changes in peripheral blood cell counts.

An unexpected observation was that the resolution of myelofibrosis over time proceeded at different rates in different skeletal sites. The most rapid changes were seen in the femora and adjacent supra-acetabular bone. Although the explanation is not clear, it is plausible that constant weight-bearing stress in these areas may hasten bone remodeling and turnover. The different rates of remodeling further underscore the heterogeneity of fibrosis in the skeleton, and caution against basing a global assessment of the marrow status on a single marrow biopsy. The biopsy needle may, by chance, sample a site that is not indicative of the current level of disease activity and may not detect changes associated with disease recurrence after HCT.

The present study suggests that MRI is useful for the classification of myelofibrosis. However, a larger cohort of patients will need to be studied to develop a more definitive MRI staging system for myelofibrosis (and osteosclerosis) before the use of MRI can be recommended as a standard monitoring tool in patients with myelofibrosis. Nevertheless, the present study shows that MRI more than complements the findings from needle biopsies of marrow and adds accuracy and a more global perspective to the evaluation of the status and evolution of myelofibrosis. This approach coupled with in situ molecular studies on biopsy specimens from MRI-defined sites should enable us to shed new light on processes involved in the progression and resolution of marrow fibrosis and osteosclerosis.

Back to Article Outline

Acknowledgments 

This study was supported by grants HL 36444, CA 15704, and CA 18029 from the National Institutes of Health, Bethesda, Md. We thank Ngan Nguyen and Joanne Greene for maintaining a meticulous database, Clara Bryan for processing biopsy specimens, the staff at First Hill Diagnostics Imaging for securing all MRI studies, and Helen Crawford and Bonnie Larson for manuscript preparation.

Back to Article Outline

References 

1. Hoffman R. Agnogenic myeloid metaplasia. In:  Hoffman R,  Benz EJ,  Shattil SJ,  Furie B,  Cohen HJ,  Silberstein LE,  McGlave P editor. Hematology: Basic Principles and Practice. New York: Churchill Livingstone; 2000;p. 1172–1188
   • View In Article
2. Tefferi A. Chronic myeloid disorders: Classification and treatment overview (review). Semin Hematol. 2001;38(suppl 2):1–4
   • View In Article
   • Full-Text PDF (117 KB)
   • CrossRef
3. Platzbecker U, Gooley T, Anasetti C, et al. Curative therapy of advanced essential thrombocythemia or polycythemia vera by hemopoietic stem cell transplantation. Leuk Lymphoma. 2002;43:1409–1414
   • View In Article
   • MEDLINE
   • CrossRef
4. Kroger N, Zabelina T, Schieder H, et al. Pilot study of reduced-intensity conditioning followed by allogeneic stem cell transplantation from related and unrelated donors in patients with myelofibrosis. Br J Haematol. 2005;128:690–697
   • View In Article
   • MEDLINE
   • CrossRef
5. Rondelli D, Barosi G, Bacigalupo A, et al. Allogeneic hematopoietic stem-cell transplantation with reduced-intensity conditioning in intermediate- or high-risk patients with myelofibrosis with myeloid metaplasia. Blood. 2005;105:4115–4119
   • View In Article
   • MEDLINE
   • CrossRef
6. Deeg HJ, Gooley TA, Flowers MED, et al. Allogeneic hematopoietic stem cell transplantation for myelofibrosis. Blood. 2003;102:3912–3918
   • View In Article
   • MEDLINE
   • CrossRef
7. Rajantie J, Sale GE, Deeg HJ, et al. Adverse effect of severe marrow fibrosis on hematological recovery after chemoradiotherapy and allogeneic bone marrow transplantation. Blood. 1986;67:1693–1697
   • View In Article
   • MEDLINE
8. Doney KC, Buckner CD, Thomas ED, et al. Allogeneic bone marrow transplantation for chronic granulocytic leukemia. Exp Hematol. 1981;9:966–971
   • View In Article
   • MEDLINE
9. Soll E, Massumoto C, Clift RA, et al. Relevance of marrow fibrosis in bone marrow transplantation: a retrospective analysis of engraftment. Blood. 1995;86:4667–4673
   • View In Article
   • MEDLINE
10. Guardiola P, Anderson JE, Bandini G, et al. Allogeneic stem cell transplantation for agnogenic myeloid metaplasia: a European Group for Blood and Marrow Transplantation, Société Française de Greffe de Moelle, Gruppo Italiano per il Trapianto del Midollo Osseo, and Fred Hutchinson Cancer Research Center collaborative study. Blood. 1999;93:2831–2838
    • View In Article
    • MEDLINE
11. Bauermeister DE. Quantitation of bone marrow reticulin—a normal range. Am J Clin Pathol. 1971;56:24–31
    • View In Article
    • MEDLINE
12. Dekmezian R, Kantarjian HM, Keating MJ, Talpaz M, McCredie KB, Freireich EJ. The relevance of reticulin stain-measured fibrosis at diagnosis in chronic myelogenous leukemia. Cancer. 1987;59:1739–1743
    • View In Article
    • MEDLINE
    • CrossRef
13. Deeg HJ, Storer BE, Boeckh M, et al. Reduced incidence of acute and chronic graft-versus-host disease with the addition of thymoglobulin to a targeted busulfan/cyclophosphamide regimen. Biol Blood Marrow Transplant. 2006;12:573–584
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (201 KB)
    • CrossRef
14. Bornhäuser M, Storer B, Slattery JT, et al. Conditioning with fludarabine and targeted busulfan for transplantation of allogeneic hematopoietic stem cells. Blood. 2003;102:820–826
    • View In Article
    • MEDLINE
    • CrossRef
15. Jurado M, Deeg HJ, Storer B, et al. Hematopoietic stem cell transplantation for advanced myelodysplastic syndrome after conditioning with busulfan and fractionated total body irradiation is associated with low relapse rate but considerable nonrelapse mortality. Biol Blood Marrow Transplant. 2002;8:161–169
    • View In Article
    • Abstract
    • Full-Text PDF (152 KB)
    • CrossRef
16. Stuart MJ, Cao TM, Sandmaier BM, et al. Efficacy of non-myeloablative allogeneic transplant for patients with myelodysplastic syndrome (MDS) and myeloproliferative disorders (MPD) (except chronic myelogenous leukemia). Blood. 2003;102(Pt 1):185a;Abstract 644
    • View In Article
17. Anderson JE, Tefferi A, Craig F, et al. Myeloablation and autologous peripheral blood stem cell rescue results in hematologic and clinical responses in patients with myeloid metaplasia with myelofibrosis. Blood. 2001;98:586–593
    • View In Article
    • MEDLINE
    • CrossRef
18. Dupriez B, Morel P, Demory JL, et al. Prognostic factors in agnogenic myeloid metaplasia: a report on 195 cases with a new scoring system. Blood. 1996;88:1013–1018
    • View In Article
    • MEDLINE
19. Olipitz W, Beham-Schmid C, Aigner R, Raith J, Linkesch W, Sill H. Acute myelofibrosis: multifocal bone marrow infiltration detected by scintigraphy and magnetic resonance imaging. Ann Hematol. 2000;79:275–278
    • View In Article
    • MEDLINE
    • CrossRef
20. Gualdi GF, Di Biasi C, Serafini G, Casciani E. The role of MRI in the valuation of disorders involving the bone marrow. [in Italian] Clin Ter. 1997;148:407–417
    • View In Article
    • MEDLINE