International working group for myelofibrosis research and treatment response assessment and long-term follow-up of 50 myelofibrosis patients treated with thalidomide-prednisone based regimens


  • Conflict of interest: Nothing to report.

We have previously reported the benefits of combination therapy with thalidomide and prednisone for the improvement of cytopenias and splenomegaly for patients with myelofibrosis (MF); both primary and those arising from an antecedent polycythemia vera (PV) and essential thrombocythemia (ET) [1]. We report here the overall efficacy of three different thalidomide-prednisone based regimens (alone or in combination with either oral cyclophosphamide for three months or continuous weekly etanercept) when assessed by the IWG-MRT response criteria [2] which were developed after completion of these studies. Additionally we report on the durability of response and long-term follow-up. We observed an overall response rate of 28% by IWG-MRT criteria (mostly clinical improvement—22% anemia, 8% splenomegaly) with a median duration of response of 8.5 months. Toxicities primarily consisted of neuropathy (Grade 3 or higher neuropathy in 6%) and cytopenias (Grade 3 or higher in 20%). After a median follow-up of three years, fourteen patients (28%) had expired from MF at the time of this analysis with median survival of 36 months.

MF is a chronic, clonal myeloproliferative neoplasm characterized by progressive anemia, leukoerythroblastic peripheral blood smear picture, splenomegaly and the potential transformation to acute leukemia [3]. Clinically, patients experience a heterogeneous course with problems of significant constitutional symptoms [4], progressive anemia, and splenomegaly. The treatment of patients with MF remains unsatisfactory. Allogeneic stem cell transplantation is a potentially curative option [5], but its wide applicability is limited by significant toxicity, lack of suitable donors, and the advanced age of many MF patients. Currently available medical therapies have been only of modest efficacy and used mainly for palliation of MF-associated cytopenias (erythropoietin [6], androgens like danazol [7, 8]) or splenomegaly (hydroxyurea [9], interferon [10]). Median survival of patients with MF has not improved with medical therapy.

Thalidomide, originally developed as a sedative, hypnotic agent in the 1950s exhibits potent anti-angiogenic properties by inhibiting basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) [11, 12]. Initial tests of thalidomide in MF [13, 14] demonstrated improvement in cytopenias without much impact on splenomegaly, and none on the histologic manifestations of the disease. Tolerance at doses in excess of 100 mg/day was poor (100–800 mg) with significant neurotoxicity which was dose limiting. We performed a subsequent Phase 2 study demonstrating that low dose thalidomide (50 mg) with a tapering dose of prednisone had a much better tolerance with improvement in cytopenias with modest efficacy for splenomegaly [1]. Among objective responses based upon the criteria of that trial 62% of patients experienced improvement in anemia, 40% achieved transfusion independence and 19% had a reduction in their spleen size.

Subsequently we tried to augment the efficacy seen with thalidomide-prednisone by the addition of low dose oral cyclophosphamide (as there was some evidence of synergy when it was used with thalidomide and steroids in multiple myeloma [15]). Additionally, a second combination trial of thalidomide-prednisone with the TNF-alpha inhibitor etanercept (demonstrated single agent symptomatic benefit [16]) was undertaken. In this report we assess the combined long term follow-up on our three thalidomide-prednisone based clinical trials. Finally, we analyzed the outcomes from Thal-pred based MF therapy in light of the IWG-MRT prognostic and response criteria both of which were developed subsequent to these trials [2, 17].

A total of 50 patients with MF were enrolled between the three trials (See Table I for demographic information) with demographic characteristics typical for the disorder. Median age at enrollment was 68 years (range, 46–85 years) and 14 were females (28%). Thirty-four (68%) had PMF and the remainder had evolved from pre-existing PV and ET. Forty-five (90%) had Intermediate-2 or higher risk disease by IWG-MRT prognostic criteria. Thirty-one patients were transfusion-dependent prior to the study (62%). The median spleen size was 17.2 cm (range, 0–34 cm). Overall, 78% patients reached the three month juncture on the trials (at which point prednisone was stopped), with 40% reaching six months. Ninety-five percent patients were able to complete the first three months of treatment with thalidomide-prednisone alone whereas it dropped to 73% and 57% for thalidomide-prednisone-etanercept and thalidomide-prednisone-cyclophosphamide, respectively.

Table I. Demographic Characteristics for 50 Patients with Myelofibrosis Treated on Thalidomide-Prednisone Based Regimens
DemographicsTPC (N = 14)TPE (N = 15)TP (N = 21)Totals (N = 50)
  • TP, thalidomide prednisone; TPC, thalidomide prednisone cyclophosphamide; TPE, thalidomide, prednisone, etanercept; M: male.

  • a

    Requiring more than 2 units of red cells in a month prior to enrollment.

  • b

    The cm below the left costal margin in mid-clavicular line.

Median age (years)72.5 (range 46–85)67 (54–78)65 (43–78)68 (43–85)
SexM - 7 (50%)M −13 (86%)M −16 (76%)M-36 (72%)
Type of MFPrimary −11 (79%)9 (60%)14 (67%)34 (68%)
Post-PV-3 (21%)2 (13%)2 (9%)7 (14%)
Post-ET-04 (27%)5 (24%)9 (18%)
IWG-MRT prognostic group [17]Low-001 (5%)1 (2%)
Intermediate1–1 (7%)1 (7%)2 (9%)4 (8%)
Intermediate2–1 (7%)7 (47%)8 (38%)16 (32%)
High risk-12 (86%)7 (47%)10 (48%)29 (58%)
Performance Status (ECOG)
Transfusion dependent anemiaa8/14 (57%)7/15 (47%)16/21 (76%)31 (62%)
Median platelets (×10 9/L)76 (11–229)127 (9–512)154 (23–448)117 (11–512)
Median WBC (×10 9/L)11(2.8–70.1)8.15 (2.6–116)9.6 (1.6–23.5)11 (1.6–116)
Abnormal karyotype at enrollment5/14 (35%)8/15 (53%)12/21 (57%)25 (50%)
Median spleen sizeb17.2 (0–29)19.7 (0–23)16.8 (0–34)17.2 (0–34)

In general across the trials there was good tolerance except for the thalidomide-prednisone-cyclophosphamide regimen (Table II). Toxicity during active therapy included myelosuppression with three cases of Grade 3 anemia, three cases of Grade 3 neutropenia, and four cases of Grade 3 thrombocytopenia (mostly in the cyclophosphamide containing regimen). Initial neuropathy was uncommon with only three cases of Grade 3 neuropathy observed. There was no treatment-attributable mortality. When responses were assessed BY IWG-MRT criteria [2], none of the patients were able to achieve complete response (CR) and there was only one partial remission (PR). Fourteen (28%) obtained clinical improvement (CI) by the IWG-MRT response assessment criteria and the majority (48%) of responses were stable disease (SD). Among the achievers of CI, 11 had response for anemia (22%) and 4 (8%) had response for splenomegaly (one patient in Thalidomide-Prednisone regimen had combined response to anemia and splenomegaly).

Table II. Efficacy and Toxicity Observed in Three Thalidomide-Prednisone Based Regimens for Myelofibrosis
ToxicityTPC (N = 14)PET (N = 15)TP (N = 16)Totals (N = 50)
Toxicity observed (≥ Grade 3)
Fluid retention/edema1012
Efficacy and outcomes
Median duration of follow up (Months)22.5 (3–57)58 (6–75)42 (6–106)36 (3–106)
Completion of 3 cycles8/14 (57%)11/15 (73%)20/21 (95%)39 (78%)
Reason for stopping therapyNo response (N = 7)(N = 8)(N = 8)23 (46%)
Toxicity (N = 4)(N =1)(N = 2)7 (14%)
Response by IWG criteria [2]
Complete response0000
Partial response0011
Clinical improvement
 • Anemia14611
 • Splenomegaly0044
Stable disease96924
Progressive disease35210
Median time to response (weeks)4488
Median duration of response (months)6 (only one response)9.5 (6–16)8 (3–42)7 (3–42)
Median time to next therapy (months)2 (1–4)3 (1–45)7 (2–50)3 (1–50)

After a median follow up of 36 months across this cohort, we observed an overall median duration of response of 8.5 months (range, 3–42 months). Median time to institution of next therapy was three months (range, 1–50 months). There were patients with periods of prolonged stabilization after cessation of therapy. At the time of this analysis, 14 patients (28%) had expired from MF. Median survival across the entire cohort was 36 months (range, 3–106).

Although comparison of three independent trials is not an ideal way to compare and contrast regimen effectiveness, this report suggests greater toxicity, inferior response rate, duration of response, and time to next therapy with the cyclophosphamide containing regimen. There were more females in the regimen containing cyclophosphamide than the other two. More patients had transfusion dependent anemia in the thalidomide-prednisone alone regimen whereas more patients were thrombocytopenic in the cyclophosphamide containing regimen. The median spleen size was comparable across the three studies but more patients had less chromosomal abnormalities in the cyclophosphamide containing. Eastern cooperative group (ECOG) performance status was comparable among these three studies but significantly higher percentage of patients turned out to be high risk by the new IWG-MRT risk assessment criteria in the cyclophosphamide and etanercept containing studies. Despite reported synergy of cyclophosphamide in multiple myeloma and activity of etanercept in alleviating constitutional symptoms in MF, addition of these agents to thalidomide and prednisone does not seem to augment the efficacy. This could partly be secondary to inability of patients to complete the planned total duration of treatment due to toxicity and intolerance. It could also be secondary to more patients belonging to a higher risk prognostic category (as suggested by IWG-MRT criteria) in the other two regimens. Nevertheless, thalidomide-prednisone based regimens seem to be active in a subset of patients with MF, with modest response rates by IWG-MRT criteria but limited by neuropathy.

The therapeutic landscape in MF is changing fast. The discovery of JAK2-V617F mutation, and other molecular insights into the pathophysiology of myeloproliferative neoplasms, has opened an era of potential new therapies for MF. Over a dozen JAK2 inhibitors are in development [18]. Epigenetic changes are being targeted with agents like decitabine [19, 20]. Reduced intensity allogeneic transplantation is being explored which may broaden its applicability to more patients with curative potential [21, 22]. Newer Jak-2 inhibitors seem to have their main effect on splenomegaly with potential worsening of cytopenias [23, 24]. Pomalidomide, a distinct IMiD® immunomodulatory drug, is being tested in MF [25, 26]. Published response rates of pomalidomide appear equal or exceed that of thalidomide-prednisone based regimens without treatment limiting neuropathy.

Methods and Study Design

The patients were enrolled in all the three studies after approval from the Mayo Clinic Institutional Review Board. We retrospectively analyzed the data taking into account IWG-MRT response and prognostic criteria. Patients with primary MF and those with MF secondary to Polycythemia Vera (PV) and Essential Thrombocythemia (ET) were eligible. All patients had a pretreatment physical examination, baseline laboratory assessment, assessment of circulatory hematological parameters and bone marrow examination with cytogenetics and fluorescent in situ hybridization (FISH) to exclude occult chronic myelogenous leukemia. Thalidomide was administered per the Celgene System for Thalidomide Education and Prescribing Safety (STEPS) program. In the thalidomide-prednisone study; patients received thalidomide 50-mg orally daily. Prednisone was given in a tapering fashion; 0.5 mg/kg/d for first month, 0.25 mg/kg/d for second month, and 0.125 mg/kg/d for the third month. Patients continued to receive additional three months of thalidomide at the 50 mg daily dose if they showed any response for cytopenias and organomegaly. In the thalidomide-prednisone-cyclophosphamide study; patients received oral cyclophosphamide 25 mg daily for three months along with thalidomide 50 mg daily and prednisone taper as before. Thalidomide was continued as maintenance treatment if there was a response. In the thalidomide-prednisone-etanercept study; 25 mg dose of etanercept was administered subcutaneously twice-a-week along with thalidomide-prednisone as in the aforementioned studies. Patients remained on etanercept and thalidomide if they had any response per protocol. The National Cancer Institute Common Toxicity Criteria (NCI CTC versions 3.0) was used to evaluate toxicity. The descriptive statistics were calculated using means, standard deviations, medians, and ranges for continuous variables; frequencies and relative frequencies for categorical variables. The survival was calculated using Kaplan-Meier survival statistics. 1

Figure 1.

Survival estimate in 50 patients with myelofibrosis treated on thalidomide-prednisone based regimens

Prakash Thapaliya*, Ayalew Tefferi†, Animesh Pardanani†, David P. Steensma‡, John Camoriano*, Wenting Wu§, Susan Geyer¶, Ruben A. Mesa*, * Division of Hematology/Oncology, Mayo Clinic Scottsdale, Arizona, † Division of Hematology, Mayo Clinic Rochester, Minnesota, ‡ Dana-Farber Cancer Institute, Wellesley, Massachusetts, § Department of Biostatistics, Mayo Clinic Rochester, ¶ Department of Biostatistics, Ohio State University.