Thalidomide therapy for myelofibrosis with myeloid metaplasia




Thalidomide is a putative antiangiogenesis agent with activity in several hematologic malignancies.


Forty-four patients who had myelofibrosis with myeloid metaplasia received treatment with thalidomide in a Phase II clinical trial at a dose of 200 mg daily with escalation by 200 mg weekly until the best tolerated dose (maximum, 800 mg) was reached.


Seventeen of 41 evaluable patients (41%) who received treatment for at least 15 days had a response. A complete response (without reversal of bone marrow fibrosis) was achieved in 4 patients (10%), a partial response was achieved in 4 patients (10%), and hematologic improvements in anemia, thrombopenia, and/or splenomegaly were observed in 9 patients (21%). Improvements in anemia occurred in 7 of 35 patients (20%) with hemoglobin levels <10.0 g/dL, and improvements in thrombopenia occurred in 5 of 24 patients (21%) with platelet counts <100 × 109/L. Five of 24 patients (21%) became transfusion-independent. Major or minor regression of splenomegaly was noted in 9 of 29 evaluable patients (31%), and complete regression was noted in 5 patients. Responders had a lower baseline median vascular endothelial growth factor levels (77.9 pg/mL vs. 97.7 pg/mL; P <.01) and higher median basis fibroblast growth factor levels (60.8 pg/mL vs. 37.4 pg/mL; P <.01) compared with nonresponders. Nine patients (22%) had deterioration that was attributed to thalidomide (resolved after withdrawal) with either progressive cytopenias or excessive proliferation. Two patients developed Grade 3 neutropenia with recovery and resumed therapy with dose reductions, and both later achieved a complete response. Dose-related toxicities included fatigue (50%), constipation (48%), rash or pruritis (37%), sedation (35%), peripheral edema (29%), tremors (23%), peripheral neuropathy (22%), and orthostasis (16%).


Thalidomide warrants further evaluation in patients with MMM, particularly in combination regimens, along with the investigation of newer analogs. Cancer 2006. © 2006 American Cancer Society.

Myelofibrosis with myeloid metaplasia (MMM) is characterized by left-shifted hematopoiesis with peripheral tear-drop red cells, extramedullary hematopoiesis with splenomegaly, bone marrow fibrosis, and progressive cytopenias in advanced stages. Features that predict an adverse outcome include older age, leukopenia or leukocytosis, circulating blasts, anemia, thrombocytopenia, and cytogenetic abnormalities.1, 2 Current therapy for patients with MMM is inadequate. To our knowledge, no survival benefit has been demonstrated for any particular approach to date outside of allogeneic stem cell transplantation. Thus, new treatment strategies are needed.3

Increased bone marrow angiogenesis has been well documented in MMM.4–6 Marked bone marrow stromal proliferation is attributed to increased vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), transforming growth factor β (TGF-β), platelet-derived growth factor (PDGF), and tumor necrosis factor α (TNF-α). In MMM-related angiogenesis, dysregulated proliferation of polyclonal endothelial bone marrow stromal cells results in deposition of extracellular matrix proteins, leading to characteristic pathologic findings.1, 2, 6

Increased bone marrow vascularity is an independent prognostic feature that correlates with worse survival in patients with myelofibrosis.6 Abnormally increased bone marrow angiogenesis is more pronounced in patients who have idiopathic or agnogenic myeloid metaplasia (AMM) compared with patients who have other myeloproliferative disorders (essential thrombocythemia or polycythemia vera).4, 6, 7 Blood and bone marrow VEGF levels are significantly elevated in MMM, making specific VEGF receptor antagonists and other antiangiogenic agents of particular interest for study in MMM.4, 8

TNF-α has been implicated in the inhibition of normal hematopoiesis, pathogenesis of bone marrow fibrosis, and cancer-related cachexia.9–11 TNF-α inhibitors can restore normal hematopoiesis and reduce disease-related constitutional symptoms.12, 13 Steensma et al.12 reported a beneficial effect of etanercept, a soluble TNF-α receptor, in 22 patients with MMM. Those authors noted a 20% rate of improvement in cytopenias and/or regression of splenomegaly, with 60% of patients reporting a reduction in the severity of constitutional symptoms.

Thalidomide is a putative antiangiogenic and immunomodulatory agent with activity in hematologic malignancies, including multiple myeloma (MM) and myelodysplastic syndromes (MDS).14, 15 Thalidomide inhibits neoangiogenesis by down-regulating VEGF, bFGF, and TNF-α in in vitro and in vivo animal models.16, 17 Therefore, a Phase II study of thalidomide was conducted in patients with MMM. This report comprises a summary of the experience at our institution in addition to detailed assessments of angiogenesis.


Eligibility Criteria

Patients were treated on protocols that were approved by The University of Texas M. D. Anderson Cancer Center Institutional Review Board after written informed consent was obtained according to institutional guidelines. Because of the known teratogenic potential of thalidomide, pregnant or lactating women were excluded. Participants of childbearing age were required to use appropriate methods of contraception during the study and for 4 months after the completion of therapy.

Patients with MMM (agnogenic, postpolycythemic. or postthrombocythemic) were required to meet standard diagnostic criteria.2 This included diffuse bone marrow fibrosis without BCR-ABL gene rearrangement. If splenomegaly was absent, then other features, including circulating immature myeloid cells or erythroblasts with tear-drop erythrocytes, were required to support the diagnosis. Indications for therapy included cytopenias, splenomegaly, and/or disease-related symptoms regardless of prior treatment status.

Prior investigational therapy must have been discontinued for ≥4 weeks unless progression of disease warranted earlier intervention. Other entry criteria included age older than 12 years, adequate hepatorenal functions (bilirubin and creatinine levels ≤2.5 mg/dL), and a Zubrod performance status 3 or better. Patients who had serious, active, uncontrolled infections and a history of seizures or severe neurotoxicity from prior therapy were not eligible.


Thalidomide (Celgene Corporation, Warren, NJ) was provided as 50 mg gelatin capsules and was given orally at bedtime. An initial dose of 200 mg was escalated by 200 mg weekly, as tolerated, if there was no relevant toxicity >Grade 1 (according to the National Cancer Institute Common Toxicity Criteria [version 2]), up to a maximum dose of 800 mg. Elderly patients could commence therapy at 100 mg orally with smaller increments (e.g., 50 mg or 100 mg) based on the clinical judgment of the investigator.

Dose modifications were implemented for relevant toxicity. If Grade 2 toxicity persisted for >2 weeks, then a dose reduction by 1 dose level was implemented. If patients had Grade 3 or 4 toxicity, then thalidomide was withheld until the toxicity resolved to Grade <2, and thalidomide was resumed at a lower dose if the investigator judged that the patient was benefiting from therapy. The maximum tolerated dose was maintained for as long as feasible unless intolerance or disease progression was observed.


Baseline assessments (within 2 weeks) included history and physical examination (palpating organomegaly), complete blood count with differential, sequential multiple analysis (SMA-12), and bone marrow aspirate with biopsy (including microvessel density [MVD], reticulin/trichrome stains, and cytogenetics). Imaging studies included bone marrow scans, abdominal ultrasound studies, and/or computed tomography scans of the abdomen if the organomegaly was difficult to palpate. Laboratory assessments were initially repeated weekly. Patients were evaluated prior to each dose escalation, then every 2 to 12 weeks depending on the clinical situation.

Response Criteria

All patients who had received at least 1 dose of thalidomide were evaluable for toxicity. Patients who received >14 days of thalidomide were evaluable for response. A complete response (CR) included bone marrow with ≤5% myeloblasts (normal maturation of all cell lines). Resolution of fibrosis was not required. Peripheral blood criteria for a CR included the resolution of cytopenias with hemoglobin >11 g/dL without erythropoietin or transfusion support, absolute neutrophil count (ANC) >1.5 × 109/L without granulocyte-colony stimulating factor, and platelet count (PLT) >100 × 109/L without transfusion, or normalization of white blood cells to ≤10 × 109/L and PLT to ≤450 × 109/L if those parameters were elevated prior to start of therapy. Complete resolution of hepatosplenomegaly was required.

A partial response (PR) included hematologic improvement (HI) in ≥2 of the following criteria: 1) an increase in ANC by ≥100% and ≥109/L if the baseline level was <109/L; 2) an increase in hemoglobin by ≥2 g/dL if the baseline level was <10 g/dL; 3) an increase in PLT by ≥100% and ≥50 × 109/L; 4) a reduction in splenomegaly or hepatomegaly by ≥50%; 5) a reduction in blasts to ≤5% in normocellular or hypercellular bone marrow if the baseline levels were >10%; and 6) a reduction in PLT by ≥50% if the baseline count was >450 × 109/L.

Correlative Studies

Measurement of angiogenic factors

Plasma samples were collected at baseline, 1 week, 30 days, then every 1 to 3 months during therapy when feasible (e.g., if the patient consented to participate in optional studies and remained on study for a sufficient duration). Assays were performed in duplicate, and the average values were recorded. Enzyme-linked immunosorbent assays for VEGF, bFGF, and TNF-α were performed as described previously.4 Commercially available kits from R&D Systems (Minneapolis, MN) were used, except for TNF-α, which was analyzed by using a kit from Oncogene Research Products (Boston, MA).

Measurement of bone marrow microvessel density

Bone marrow biopsies (at baseline and every 1-3 months thereafter) were assessed blindly for MVD according to previously described methods by using the standard immunoperoxidase technique for staining blood vessel endothelial cells with antifactor VIII-related antigen antibody (Dako Corporation; Santa Barbara, CA).4 The vascular area was measured using National Institutes of Health shared image-analysis software.

Statistical Considerations

A 2-stage design was implemented. If 0 of 14 evaluable patients (i.e., treated >14 days) responded, then the study would be terminated; otherwise, an additional 25 patients could be treated to determine the response rate more precisely (standard error ± 10%). Differences in pretreatment characteristics and response rates among subgroups were analyzed with the chi-square test or the Fisher exact test.18


Study Group

Between October 1999 and April 2001, 44 patients with myelofibrosis (including 31 patients with AMM) were enrolled. Their median age was 65 years (range, 27-85 yrs), and 59% of patients were male. The median time from diagnosis to treatment was 15 months (range, 1-122 mos), and 64% of patients had received previous treatment. Prior therapy included erythropoietin (n = 18 patients), hydroxyurea (n = 15 patients), interferon-α (n = 13 patients), androgens (n = 4 patients), and other chemotherapeutics (n = 3) (Table 1). Five patients underwent splenectomy.

Table 1. Pretreatment Features in 44 Patients
CategoryNo. of Patients (%)
  • WBC: white blood cells.

  • *

    Dupreiz prognostic score: 2, high risk (hemoglobin <10 g/dL or transfusion-dependent and WBC <4 × 109/L or >30 × 109/L; 1, intermediate risk (1 feature); or 0, low risk (no features).

 Primary31 (70)
 Postthrombocythemic10 (23)
 Postpolycythemia3 (7)
Prior therapy 
 None16 (31)
 Erythropoietin18 (41)
 Hydroxyurea15 (34)
 Interferon-α5 (11)
 Splenectomy5 (11)
 Androgens4 (10)
 Other9 (20)
Age in y 
 <401 (2)
 40–5913 (30)
 ≥6030 (68)
Time from diagnosis ≥12 mos24 (55)
Dupreiz score* 
 012 (27)
 121 (48)
 211 (25)
WBC <4 × 109/L or >30 × 109/L12 (27)
Circulating blasts 
 Yes27 (61)
Hemoglobin <10 g/dL29 (66)
Transfusion dependence (yes)24 (55)
Platelet count <100 × 109/L17 (39)
Splenomegaly (n = 39) (yes)32 (82)
Bone marrow blasts ≥5%6 (14)
 Diploid28 (64)
 Del20q or del13q7 (18)
 Other9 (20)
Grade of bone marrow fibrosis 
 Grade 25 (11)
 Grade 317 (39)
 Grade 420 (45)
Bone marrow osteosclerosis (yes)16 (36)

A summary of prior therapy, age, time from diagnosis, cytopenias, organomegaly, and bone marrow findings are provided in Table 1. The median white blood cell count was 9.5 × 109/L (range, 2.0-54.6 × 109/L), the median hemoglobin was 9.5 g/dL (range, 5.8-15.1 g/dL), and the median PLT was 144 × 109/L (range, 14-572 × 109/L). Fifty-five percent of patients were transfusion-dependent. Transfusion dependence correlated with hepatomegaly. Increasing age correlated with increased grade of reticulin fibrosis and higher incidence of transfusion dependence. Thirty-two of 39 patients (82%) without prior splenectomy had palpable splenomegaly.


Details related to response are provided in Tables 2 and 3. Patients were evaluable if they had received >14 days of therapy. Three patients discontinued therapy within 15 days for Grade 3 rash (Patients 12 and 17) or for rapidly progressive disease with splenic infarction (Patient 20). The median duration of thalidomide therapy was 3 months (range, from 10 days to 28 mos). Seventeen of 41 evaluable patients (41% or 39% by intent-to-treat [ITT] analysis) had an objective response to thalidomide.

Table 2. Characteristics of 17 Responders to Thalidomide
PatientAge (Years)GenderCategoryMonths from DiagnosisPrior TherapyDupriez Score*Average Dose (mg)Duration of Therapy (Days)Response
  • AMM: agnogenic myeloid metaplasia; PR: partial response; HI: hematologic improvement; PLT: platelet count; EPO: erythropoietin; CR: complete response; PT-MMM: postthrombocythemic myelofibrosis with myeloid metaplasia; SPL: spleen; IFN-α: interferon α; PPV-MMM: postpolycythemic myelofibrosis with myeloid metaplasia; HGB: hemoglobin.

  • *

    Dupreiz prognostic score: 2, high risk (hemoglobin <10 g/dL or transfusion-dependent and WBC <4 × 109/L or >30 × 109/L; 1, intermediate risk (1 feature); or 0, low risk (no features). Patients 3, 4, 6, 7, 10, 14, 18, 25, 29, 31, and 35 were transfusion-dependent: Patient 4 with del13(q14q22), Patient 7 with −7, Patient with del(20)(q11), and Patient 38 with +8 karyotypes.

  • Achieved transfusion-independence.

  • EPO failure.

645FemalePT-MMM3Hydrea, EPO1600348HI-SPL
747FemalePT-MMM95IFN-α, hydrea2800336PR
1472MalePT-MMM11Hydrea, androgens1200336HI-SPL
2972MaleAMM64Mitoxantrone plus ara-C, EPO, prednisone, splenectomy120096CR
3568MaleAMM15Hydrea, IFN-α2400101HI-PLT
Table 3. Response by Category and Angiogenesis Modulation with Thalidomide
PatientAverage Dose (mg)Days of TherapyResponse
Leukocytosis or LeukopeniaAnemiaThrombocytosis or ThrombopeniaOrganomegaly (cm below costal margin)Overall ResponseVEGF (pg/mL)bFGF (pg/mL)TNF-α (pg/mL)
  • VEGF: vascular endothelial growth factor; bFGF: basic fibroblast growth factor; TNF-α: tumor necrosis factor α; Pre: before thalidomide; Post: after thalidomide; —: not applicable; CR: complete response; PR: partial response (≥50% improvement); MR: minor response (≥30% reduction in splenomegaly); NR: no response; HI: hematologic improvement; SPL: spleen; PLT: platelets; SD: stable disease (no change from baseline); PD: progressive disease; NE: not evaluable.

  • *

    Achieved transfusion-independence.

  • Increased blasts, leukocytosis, or thrombocytosis resolved after withdrawal.

  • Worsening or new cytopenia attributed to thalidomide.

  • §

    Reduced transfusion requirements ≥50%.


Responders remained on thalidomide for a median of 8 months (range, 1-28 mos). A CR was observed in 4 patients (10%; ITT, 9%); a PR was observed in 4 patients (10%; ITT, 9%); and HI of anemia, thrombopenia, or splenomegaly was observed in 9 patients (21%; ITT, 21%). Improvements in anemia occurred in 7 of 35 patients (20%) with hemoglobin <10 g/dL and in thrombopenia for 5 of 24 patients (21%; ITT, 20%) with PLT <100 × 109/L. Five of 24 patients (21%) became transfusion-independent. Regression of splenomegaly was noted in 9 of 29 evaluable patients (31%; ITT, 29%), with complete regression noted in 5 patients.

No specific pretreatment characteristics predicted response (Table 4). Of the 4 patients who achieved a CR, 3 patients discontinued therapy after 8 to 18 months for progressive peripheral neuropathy/neuritis despite dose reductions (the fourth for anasarca). All 4 patients progressed after at least 6 months of observation, and 3 patients received supportive care. One patient failed rechallenge with thalidomide and prednisone, and was treated with the thalidomide analog lenalidomide (and achieved a PR with transfusion-independence after 6 mos of treatment).

Table 4. Response to Thalidomide Therapy by Pretreatment Features
VariableNo. of PatientsNo. of Responses (%)*
  • VEGF: vascular endothelial growth factor; bFGF: basic fibroblast growth factor; TNF-α: tumor necrosis factor α; bcm: below costal margin.

  • *

    P values for all comparisons not significant.

  • These factors were assessed at the median values based on evaluability for response and available data for VEGF (n = 38 patients), bFGF (n = 36 patients), and TNF-α (n = 31 patients).

No. of patients4117 (41)
 Primary299 (31)
 Postthrombocythemic95 (55)
 Postpolycythemic32 (66)
Prior therapy  
 Untreated145 (36)
 Treated2712 (44)
Leukocyte count  
 4.0–30.0 × 109/L2911 (38)
 <4.0 or >3.0 × 109/L126 (50)
 <10 g/dL2913 (45)
 ≥10 g/dL124 (33)
Platelet count  
 <100 × 109/L168 (50)
 ≥100 × 109/L259 (36)
Splenomegaly (n = 39)  
 <15 cm bcm137 (54)
 ≥15 cm bcm167 (43)
Bone marrow fibrosis (n = 39)  
 Grade 252 (40)
 Grade 3157 (47)
 Grade 4198 (42)
 Diploid3012 (40)
 Abnormal115 (45)
Dupriez score  
 0115 (45)
 1197 (37)
 2115 (45)
Angiogenesis factors  
  <97.7 pg/mL199 (47)
  ≥97.7 pg/mL198 (42)
  <51.7 pg/mL187 (39)
  ≥51.7 pg/mL1810 (56)
  <12.5 pg/mL167 (44)
  ≥12.5 pg/mL157 (47)

New or progressive cytopenias attributed to thalidomide (i.e., that returned to baseline after withdrawal) were observed in 9 patients (22%) (Table 3). Increased transfusion support was required in some instances. Two patients with neutropenia after initiation of therapy resumed thalidomide with dose reduction (after a brief hiatus) and achieved CRs. Proliferation was observed in 5 patients (12%) and reversed after the discontinuation of thalidomide. These events included thrombocytosis with PLT >600 × 109/L (n = 3 patients), peripheral blood basophilia (n = 1 patient), and increased peripheral/bone marrow blasts (n = 1 patient). In patients who had thrombocytosis, hydroxyurea and/or anagrelide usually were required to achieve hematologic control in addition to withdrawal of thalidomide.

Three patients transformed to acute myelogenous leukemia and underwent chemotherapy with cytarabine-containing regimens; all 3 of those patients died from infections during the induction phase. Another patient continued thalidomide off-protocol after achieving transfusion independence and regression of splenomegaly (PR). She developed malignant ovarian carcinoma (clear cell type with in utero diethylstilbestrol exposure) and died of complications from small bowel obstruction despite palliative chemotherapy.

Tolerance and Toxicity

The median average tolerated dose of thalidomide was 400 mg (range, 100-800 mg). Five patients commenced therapy at 50 mg or 100 mg because they were older. Forty-four percent of patients were escalated successfully to 800 mg, although all but 3 patients required dose reductions for intolerance. Toxicity data are detailed in Table 5. The most commonly reported side effects were constitutional in nature and included fatigue, constipation, and rash. Early discontinuation (<30 days) was related to rash, neurotoxicity (e.g., ataxia), cardiovascular symptoms (e.g., bradycardia and/or orthostatic hypotension), and/or progressive disease. There were 3 incidents of venous thromboembolism (2 pulmonary embolus, 1 extremity deep venous thrombosis). Three patients died during therapy from progressive disease (n = 1 patient), suspected pulmonary embolus (n = 1 patient), or sudden death with the cause undetermined (n = 1 patient).

Table 5. Toxicity of Thalidomide in Myelofibrosis
Adverse EventToxicity, % (n = 44)*
Grade 1–2Grade 3–4
  • *

    Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria (version 2).

  • Posthysterectomy.

  • Attributed to calcium channel blocker in 1 patient and to coronary artery disease in the other patient.

  • §

    Reversed after withdrawal, 2 with leukopenia resumed a lower dose of thalidomide with response.

 Rash or pruritis325
 Peripheral neuropathy202
 Confusion/memory loss5 
 Mood changes or depression22
 Peripheral edema209
 Pulmonary edema 5
 Pulmonary embolism 5
 Deep venous thrombosis 2
 Peripheral vascular disease2 
 Leukopenia 9
 Anemia 7
 Thrombocytosis or basophilia 7
 Pneumonia 19
 Fever of unknown origin2 
 Cellulitis 5
Splenic infarctions2 

Antiangiogeneic Effects of Thalidomide

The median MVD in pretreatment bone marrow biopsies was 7.3 (range, 1.1-60.4). There was no correlation between vascularity and response (data not shown). Baseline plasma levels were obtained for VEGF in 86% of patients enrolled, for bFGF in 82% of patients, and for TNF-α in 70% of patients (these levels were not obtained in patients who declined to participate in optional research studies or who had insufficient samples to perform all assays in the case of TNF-α). The median pretreatment levels were as follows: VEGF, 97.7 pg/mL (range, 19.5-2450 pg/mL); bFGF, 51.7 pg/mL (range, 6.3-1976.5 pg/mL); and TNF-α, 12.5 pg/mL (range, 7.6-45.4 pg/mL) (Table 3). Higher TNF-α levels correlated with older age. In patients who had serial assessments of VEGF, bFGF, and TNF-α, the “posttherapy” values that were selected for analysis were based on timing during therapy (e.g., best response or at the completion of therapy).

Figures 1 through 3 detail the median change in angiogenesis factors measured at baseline and during therapy in responders compared with nonresponders. Responders had lower median pretreatment VEGF levels (77.9 pg/mL vs. 97.7 pg/mL; P<.01) (Fig. 1) and higher bFGF levels (60.8 pg/mL vs. 37.4 pg/mL; P<.01) (Fig. 2) compared with nonresponders. There were no significant differences in pretreatment TNF-α levels, although nonresponders appeared to have increasing levels of TNF-α during therapy (Figs. 1, 3).

Figure 1.

Median vascular endothelial growth factor (VEGF) levels before and after therapy in (A) responders and (B) nonresponders.

Figure 2.

Median basic fibroblast growth factor (bFGF) levels before and after therapy in (A) responders and (B) nonresponders.

Figure 3.

Median tumor necrosis factor α (TNF-alpha) levels before and after therapy in (A) responders and (B) nonresponders.


Clinical activity of thalidomide in MMM has been reported in other clinical trials (Table 6).19–26 Similar findings were noted in our study, albeit with mixed patterns of response. Some patients clearly benefited from thalidomide with reductions in splenomegaly (31%) and improvements in anemia (20%), including achievement of transfusion-independence. The overall objective response rate was 41% in 41 evaluable patients with 4 patients (10%) who met the criteria for CR as established for this study. Conversely, others (30%) experienced clinical deterioration with worsening cytopenias or undue proliferative effects with leukocytosis, increasing peripheral/bone marrow blasts, thrombocytosis, and/or basophilia. A causal relation to thalidomide was strongly suspected, because these manifestations usually resolved after withdrawal of therapy. Two factors that may have contributed to the diversity in outcomes include 1) the inherent heterogeneity of MMM with variable clinical presentations and prognoses, suggesting unique aspects of the pathophysiology, and 2) the myriad biologic effects of thalidomide.

Table 6. Studies of Thalidomide in Myelofibrosis
ReferenceNo. of PatientsDose (Maximum), mgMedian MTD, mgNo. of Responders/Evaluable Patients*
  • MTD: maximum tolerated dose; MMM: myelofibrosis with myeloid metaplasia; HGB: hemoglobin; CR: complete response; PR: partial response; SPL: spleen; HI: hematologic improvement; PLT: platelets.

  • *

    Response definitions differed for each study.

  • Severity score: myeloproliferation index: splenomegaly/hepatomegaly (0, nonpalpable; 1, ≤10 cm below costal margin; 2, >10 cm); cm); leukocytosis (0, WBC <15 × 109/L; 1, WBC ≥15 × 109/L); thrombocytosis (0, PLT <500 × 109/L; 1, PLT ≤500 × 109/L); myelodepletion index: anemia (0, HGB >12 g/dL; 1, HGB 10-12 g/dL; 2, HGB <10 g/dL), leukopenia (0, WBC >4 × 109/L; 1, WBC ≤4 × 109/L), thrombopenia (0, PLT >150 × 109/L;1, PLT ≤150 × 109/L).

  • Compensated was defined as stable disease that did not require transfusion.

  • §

    Major response was defined as a reduction spleen size ≥25%, hemoglobin increase by >2 g/dL, or transfusion independence. Minor response was defined as <25% reduction in splenomegaly, improvement in quality of life, increase in HGB 1-2 g/dL, increase in PLT by 50% to ≥60 × 109/L, or reduction in WBC <10 × 109/L.

Canepa et al., 20011910200/day, increased by 100 every 2 wks (800)4003/10
Barosi et al., 20012021100/day, increased by 100 every mo (400)100-2008/13 (reduction in severity score ≥1 points)
Pozzato et al., 2001216100/day, increased to MTD3/6 (3/3, compensated MMM)
Piccaluga et al., 20022212100/day, increased by 100 every 2 wks (600)4007/11 (4 major, 3 minor)§
Merup et al., 20022315100-200/day (800)4000/14
Elliot et al., 20022415200/day increased by 200 mg every 28 days (1000)2003/13 (2CR, 1PR HGB, 1 PR SPL)
Strupp et al., 20022516100/day (400)30010/15 (4 CR, 6 PR)
Marchetti et al., 2004266350/day (400)40011/49 (HI HGB), 4/14 (HI PLT), 9/47 (HI SPL), 15/49 (reduction in severity score ≥1 points)
Current study44200/day, increased by 200 every wk (800)40017/41 (4 CR, 4 PR, 9 HI)

Stimulatory effects could be accounted for by the ability of thalidomide to inhibit neutrophil chemotaxis, to modify surface expression of adhesion molecules, and to induce differentiation in vitro in the human leukemia cell line K562.27–29 Reduction in apoptosis may result in unchecked hematopoiesis, which also is observed in MDS.15 Cytokine-modulatory effects include the inhibition of monocyte interleukin 12 (IL-12) production, enhanced synthesis of IL-2, and inhibition of IL-6.30, 31 Thalidomide causes differential CD8-positive T-cell stimulation with decreased CD4/CD8 ratio, shift from T-helper 1 (Th1) to Th2 T-cell responses, and inhibition of T-cell proliferation in stimulated lymphocytes.32, 33 The relevance of these modulations for MMM currently is unclear, although recent data suggest that alteration in the CD4/CD8 ratio is unlikely to have clinical significance.34

TNF-α enhances neoangiogenesis and contributes to the systemic symptoms of advanced malignancy.11 Thalidomide inhibits TNF-α production by stimulated monocytes, macrophages, and neutrophils.35 Proposed mechanisms include 1) accelerated degradation of TNF-α mRNA, 2) binding to α1-acid glycoproteins with intrinsic anti-TNF-α activity, and 3) decreased binding activity of nuclear factor κB (which controls activation of the TNF-α gene).36, 37 Etanercept is a soluble TNF-α receptor that has potent systemic anti-TNF-α activity.12, 13 In clinical trials of etanercept in MMM, clinical benefit was observed, suggesting that inhibition of TNF-α may account for some of the efficacy of thalidomide. Although there was no difference in outcomes by pretreatment TNF-α levels, TNF-α levels appeared to increase during therapy in nonresponders (Table 4; Fig. 3).

The inhibition of bone marrow angiogenesis has been correlated with response to thalidomide in patients with MMM by a few investigators.22, 34 No direct correlation between reduction in bone marrow vascularity and outcome was detected in our study, similar to the findings from initial studies of thalidomide in patients with previously treated MM.20 Decreases in levels of VEGF and bFGF occurred during therapy but did not correlate directly with response (Figs. 1 and 2). Responders had lower pretreatment VEGF levels and higher bFGF levels compared with nonresponders; however, other occult disease features may have accounted for these differences. The clinical relevance of these findings is unclear and will require corroboration in future clinical trials. Factors that may account for differences in the degree of modulation of angiogenesis among thalidomide studies include heterogeneity of patient populations, potential differential effect of dosing (high vs. low), differences in timing of bone marrow and blood collection during therapy, and differences in assay methodology.

In the current study, clinical benefit of thalidomide was apparent; however, no specific parameters were predictive of outcome according to our prognostic factor analysis. A pooled analysis of earlier studies of thalidomide in patients with MMM determined that patients with more advanced disease (e.g., major disease severity score and significant splenomegaly) had a higher overall response rate of 62%, although no CRs were observed.38 In our study, a small proportion (21%) of patients achieved a CR (albeit with persistent bone marrow fibrosis). The median age and response rates (achievement of transfusion-independence, reductions in splenomegaly, and improvements in thrombopenia) were similar in our study and in the pooled analysis.

In the current study, there was no appreciable relationship between thalidomide dose and response. Because of the generally poor tolerance of thalidomide at higher doses, particularly related to sedation, constipation, peripheral neuropathy, and autonomic dysfunction, other therapeutic strategies have been explored. Mesa et al.39 reported improved tolerance with low-dose thalidomide and oral prednisone in 21 patients; approximately 70% of those patients achieved an objective response (particularly improvements in anemia) with the combination. The improvement in efficacy may have been related to the longer exposure to therapy in addition to the potential synergism of thalidomide with corticosteroids (like with thalidomide and dexamethasone in MM).40 The activity of erythropoietin and thalidomide with or without corticosteroids in patients with previously treated MMM suggests that further study with such combinations may be warranted.41, 42

Newer thalidomide analogs (e.g., lenalidomide) have significant potency with respect to TNF-α inhibition and have a more favorable side-effect profile. Activity of lenalidomide in MM and M.D.S has been reported.43, 44 Early results from a Phase II clinical trial of lenalidomide in 15 patients with previously treated MMM showed a response rate of 27% (improvements in anemia, splenomegaly, and/or constitutional symptoms).45 Therapy was well tolerated with expected myelosuppression, although proliferative effects, particularly thrombocytosis were noted (as with thalidomide). Future clinical trials of thalidomide should incorporate other targeted therapeutics to exploit potentially synergistic mechanisms of action (e.g. tyrosine kinase inhibitors, such as imatinb mesylate,46 or farnesyltransferase inhibitors, such as R11577747). The biologic effects of thalidomide and its analogs should be investigated further in a systematic fashion. Findings may account for the differing relevance of the modulation of angiogenesis in predicting outcome among clinical trials and may have prognostic implications, allowing tailored therapeutic approaches.