These two authors equally contributed to this work.
A non-randomised dose-escalating phase II study of thalidomide for the treatment of patients with low-risk myelodysplastic syndromes: the Thal-SMD-2000 trial of the Groupe Français des Myélodysplasies
Version of Record online: 8 NOV 2005
British Journal of Haematology
Volume 131, Issue 5, pages 609–618, December 2005
How to Cite
Bouscary, D., Legros, L., Tulliez, M., Dubois, S., Mahe, B., Beyne-Rauzy, O., Quarre, M. C., Vassilief, D., Varet, B., Aouba, A., Gardembas, M., Giraudier, S., Guerci, A., Rousselot, P., Gaillard, F., Moreau, A., Rousselet, M. C., Ifrah, N., Fenaux, P., Dreyfus, F. and For The Groupe Français des Myélodysplasies (GFM) (2005), A non-randomised dose-escalating phase II study of thalidomide for the treatment of patients with low-risk myelodysplastic syndromes: the Thal-SMD-2000 trial of the Groupe Français des Myélodysplasies. British Journal of Haematology, 131: 609–618. doi: 10.1111/j.1365-2141.2005.05817.x
- Issue online: 8 NOV 2005
- Version of Record online: 8 NOV 2005
- Received 21 July 2005; accepted for publication 22 August 2005
- myelodysplastic syndromes;
- vascular endothelial growth factor;
Patients (n = 47) with low-risk myelodysplastic syndrome were treated with thalidomide [200 mg/d, increased by 200 mg/d/4 weeks up to week 16]. Responses were evaluated according to the International Working Group criteria at week 16 for 39 patients who received at least 8 weeks of treatment. Twenty-three (59%) patients showed haematological improvement (HI): four major erythroid response (HI-EM), 15 minor erythroid response, six major neutrophil response, two major platelet response. Side effects caused 22/39 to stop thalidomide before week 16. Nine of 23 responders continued thalidomide after week 16 [19% of trial patients] with sustained response in eight of nine. Six reached week 56, including the four HI-EM patients [13% of trial patients]. Nineteen of 36 red blood cell transfusion-dependent patients (53%) showed erythroid response, but only four became transfusion-independent. Among the 23 responders, the median duration of response was 260 d (range 30–650). Responses were sustained in all patients except one, and were observed between week 4 and week 8 in 85% of patients, at doses ranging from 200 to 400 mg. Only two patients responded at 600 mg/d and none at 800 mg/d. No clinical characteristics of responding versus non-responding patients were identified. The erythroid response rate was identical in all cytogenetic subgroups, including 5q31.1 deletions. Pretreatment vascular endothelial growth factor levels were lower in responders compared with non-responders (P = 0·004). Microvessel density (MVD) increased and apoptosis decreased in four of six and in all six responders studied respectively whereas MVD and apoptosis were unchanged in three non-responders.
The myelodysplastic syndromes (MDS) are one of the most common haematological malignancies affecting the adults. They represent a group of haematological disorders characterised by peripheral cytopenias and dysplastic marrow (Heaney & Golde, 1999). Low-risk MDS, including refractory anaemia (RA), RA with ring sideroblasts (RARS) and RA with <10% excess blasts (RAEB <10%) are characterised by ineffective haematopoiesis that may be explained in part by excessive apoptosis of haematopoietic progenitors and of more mature cells (Raza et al, 1995; Bouscary et al, 1997; Westwood & Mufti, 2003). Compensatory marrow proliferation may occur with later escape of clonal disease and potential for transformation to acute leukaemia (Rosenfeld & List, 2000). In particular, accelerated and abnormal apoptosis regulation in haematopoietic progenitors and their progeny has been studied regarding erythropoiesis (Claessens et al, 2002). The death receptor Fas is overexpressed on MDS CD34+ cells and on MDS erythroid cells; apoptosis in these cells coincides with overproduction of Fas-ligand (Bouscary et al, 1997). Pro-apoptotic cytokines, such as tumour necrosis factor alpha (TNF-α), interleukin-1 beta (IL-1-β) and transforming growth factor beta (TGF-β) may also promote the apoptotic death of long-term initiating cells as well as of primitive and committed progenitors (Maciejewski et al, 1995; Raza et al, 1996). Increased bone marrow microvessel density (MVD) is observed in trephine biopsies and correlates with the percentage of blasts (Pruneri et al, 1999). Vascular endothelial growth factor (VEGF) is overexpressed and secreted by myelomonocytic precursors in MDS and neutralisation of VEGF inhibits leukaemia colony formation (Bellamy et al, 2001). VEGF may also promote the expansion of myeloid progenitors while inhibiting the formation of erythroid progenitors from MDS bone marrow mononuclear cells (Bellamy et al, 2001). TNF-α also seems to have a role in angiogenesis by upregulating the expression of endothelial integrin, which is crucial for this process. Finally, immunological abnormalities with autoreactive T cells have been suggested as possible initiating events in the development of some cases of MDS (Smith & Smith, 1991; Barrett et al, 2000).
Thalidomide, a derivative of glutamic acid, was introduced in Europe in 1954 as a sedative/hypnotic agent, but was removed from the market because of its teratogenic effects. Thalidomide has been recently rediscovered, and even if its use is limited by very strict guidelines, it is now an option for a diverse range of clinical applications (Bartlett et al, 2004). Thalidomide suppresses the synthesis of TNF-α by activated monocytes, partly inhibits angiogenesis by inhibiting beta-Fibroblast Growth Factor (β-FGF) and has been shown to co-stimulate T cells and exert complex immunomodulating activities (Sampaio et al, 1991; Moreira et al, 1993; D'Amato et al, 1994; Haslett et al, 1998). Because of these activities, thalidomide has been used for the treatment of MDS. In this phase II study conducted by the Groupe Français des Myélodysplasies (GFM), an escalating dose of thalidomide was given to a cohort of 47 low-risk MDS patients. We also investigated bone marrow apoptosis and angiogenesis and serum levels of TNFα, VEGF, vascular cell adhesion molecule (VCAM-1) and E-selectin in a subset of those patients to search for possible mechanisms of action of thalidomide in MDS.
Patients and methods
Eligibility criteria included: (i) MDS according to the French–American–British (FAB) criteria, RA, RARS and RAEB <10%; (ii) low, intermediate (INT)-1 or INT-2 according to the International Prognostic Scoring System (IPSS); (iii) either of the following: transfusion-dependent anaemia with a transfusion requirement of at least two packed red blood cell (PRBC) units per month in the 2 months before inclusion; neutropenia <1 × 109/l and related bacterial infection; thrombocytopenia <50 × 109/l; (iv) other requirements: Eastern Co-operative Oncology Group (ECOG) performance status of 0–2, renal function (creatinine <300 μm/l), no peripheral neuropathy and normal electromyography (EMG), no previous venous thrombosis, a negative pregnancy test, and effective contraception in premenopausal women. Exclusion criteria were RAEB >10%, intensive chemotherapy in the previous 3 months, treatment with erythropoietin (Epo) and/or granulocyte colony-stimulating factor (G-CSF) in the previous 2 months, ECOG index ≥3, or concomittant treatment with drugs known to interfere with thalidomide, including cyclosporin, ciprofloxacin, dexamethasone or pentoxifillin. All patients, after signing an informed consent form approved by the Institutional Review Board of Hospital Cochin, Paris, participated in the study, Thal-SMD-2000, entitled ‘ Phase II trial of escalating doses of thalidomide to improve the cytopenias of patients with low-risk myelodysplasia’.
Patients were treated with thalidomide (Laphal, Allauch, France) given at bedtime for 4 weeks at a starting dose of 200 mg/d. Thalidomide was increased by 200 mg/d every 4 weeks, up to week 16 (maximum dose 800 mg/d). The 4-week increase was performed if no grade 2, 3 or 4 side effects [World Health Organisation (WHO) classification] had occurred. Patients who showed no response at week 16 were taken off study. Doses adaptations were performed as follows: Grade 1: no change; Grade 2: decrease to the previous dosage when possible, then reintroduction at the same dose, returning to the previous dose if the side effect reappeared, if a haematological response had been observed. Definite stop if no haematological response had been observed. Grades 3–4: stop 1 week and reintroduce at the previous dose if possible, then reintroduce at the same dose. If the side effect reappeared, continue at the lower dose if haematological response or definite stop if not. Responding patients at week 16 were allowed to receive maintenance therapy until week 56, as long as the haematological response was maintained, and no intolerable side effects were observed.
Evaluation During Study
Complete history and physical examination as well as complete blood count, bone marrow aspirate and biopsy with cytogenetics were assessed at baseline, and IPSS score calculated. Serum ferritin, vitamin B12 and folate levels were determined before the treatment. Complete blood counts were measured every 4 weeks. Bone marrow aspirate, biopsy, and cytogenetics were evaluated at week 16. Plasma samples were collected and frozen before treatment and at weeks 4, 8, 12 and 16. All patients had EMG and nerve conduction studies at inclusion, at week 16 and every 6 months in patients treated after week 16.
All patients who completed at least 8 weeks of treatment were considered as evaluable for response. Responses were assessed according to International Working Group (IWG) criteria (Cheson et al, 2000), and were evaluated only for patients who met inclusion criteria for the given lineage.
Analysis of Microvessel Density and Apoptosis on bone marrow biopsies
Bone marrow biopsy (BMB) was performed before the treatment in 31 patients, in six responders and three non-responders at week 16, and in two patients at week 56. BMB samples were routinely fixed in phosphate-buffered formalin (10%), decalcified, paraffin-embedded, and 4 μ sectioned for haematoxylin eosin, Periodic acid-Schiff (PAS), reticulin according to Gordon Sweet and Perls staining. Immunohistochemistry used the following antibodies for endothelial cells: anti-CD31 (DAKO 1/100, Trappes, France), anti-CD34 (DAKO 1/100), anti-VIII factor (DAKO 1/300, Paris, France), anti-ULEX (Biosys 1/50) and StreptAPAAP method (DAKO) for detection. MVD was estimated twice, blindly and semi-quantitatively graded for extent of CD34 and CD31 staining at ×10, ×25, ×40 power objective lens: 0 negative, + faint positivity, ++ moderate, +++ intensive, ++++ very intensive staining. Large vessels and open sinuses throughout the samples were excluded. For the nine patients (six responders and three non-responders) analysed at week 16 and the two at week 56, six digitised micrographs were made at ×250 and ×400 for each BMB in the ‘hot spot’ grade, and comparatively analysed for each patient: first (week 0) versus second (week 16) BMB, first versus second versus third (week 56) BMB.
Apoptotic cells were detected using the deoxy-transferase uridin triphosphate ‘nick’ end labelling (TUNEL) method on BMB. The DNA strand breaks that are characteristic of apoptosis were identified by labelling the free 3′OH nucleotidic end with fluorescein deoxy uridin triphosphate. The incorporated fluorescein was detected by antifluorescein antibody Fab fragments from sheep, conjugated by horseradish peroxidase (POD) and then substrate reaction with diaminobenzidine (DAB) as described by the manufacturer (In Situ Hybridisation Cell Death Detection Kit POD; Roche). A semi-quantitative grading was used: + few apoptotic cells, ++ moderate, +++ numerous.
Plasma VEGF, TNF-α, VCAM-1 and E-Selectin Enzyme-Linked Immunosorbent assay (ELISA)
Serum levels of VEGF, TNF-α, VCAM-1 and E-selectin were measured in triplicate using a sandwich enzyme-linked immunosorbent assay (ELISA) (Quantikine, R&D System, Minneapolis, MN, USA) according to the manufacturer's instructions. Serum controls were obtained from seven healthy volunteers that were sex- and age-matched. For serum cytokine levels, data were presented as medians and interquartile ranges [low quartile-high quartile (LQ-HQ)].
Distribution of numerical values are described as means and standard deviation or as median and percentiles as appropriate. Comparisons of numerical values over time or between groups were made by Wilcoxon signed rank or paired tests. P-values of <5% were considered statistically significant.
For statistical analysis of serum cytokine levels, the values were compared between groups using the non-parametric test of Kruskal–Wallis.
Forty-seven patients from 10 French Centers of the GFM [http://www.gfmgroup.org] were entered onto the study between January and August 2001. Patient baseline characteristics are reported in Table I. Males predominated and the median age was 67 years. The FAB diagnosis was RA (44·5%), RAEB <10% (32%) and RARS (23·5%). IPSS was INT-1 (55%), Low (17%), INT-2 (15%), but no patient had a high IPSS score. The IPSS score could not be calculated in 13% of patients because of failure of cytogenetic analysis. Median time from diagnosis was 455 d (145–824). Inclusion criteria were transfusion-dependent anaemia in 76% of patients, thrombocytopenia <50 × 109/l in 17%, and neutropenia with related bacterial infection in 15%.
|Median age, years||67|
|RAEB <10%||15 (32%)|
|Median time from diagnosis||455 d (145–824)|
|Undetermined (absence of cytogenetics)||6 (13%)|
|Thrombocytopenia <50 × 109/l|
|Neutropenia <1 × 109/l and related bacterial infection|
|Karyotype –n (%)|
Among the 47 patients treated in this study, one was lost to follow up after 2 weeks of treatment and three deaths not related to thalidomide occurred: one at week 9, from intracerebral haemorrhage in a patient with severe thrombocytopenia at inclusion, and two from septic shock because of pre-existing neutropenia not worsened by thalidomide and previous bacterial episodes, at weeks 7 and 9. Twenty-two patients stopped thalidomide before week 16 because of side effects [eight of 22 had haematological improvement (HI), including eight minor erythroid responses (HI-Em) and three major neutrophil responses (HI-NM)]. Only seventeen of the 47 patients (36%) completed 16 weeks of treatment, and 15 out of these 17 (88%) had IWG HI [four major erythroid responses (HI-EM), seven HI-Em, two major platelet responses (HI-PM,) three HI-NM]. No patient achieved the criteria for complete haematological response at week 16. Eight of the 17 patients who reached week 16 stopped treatment at that time because of side effects, including six responders. Only nine of 17 could continue thalidomide after week 16. Among these nine patients, three had to stop at weeks 32, 48 and 53 because of side effects with persistent response, whereas the remaining six [corresponding to 12·5% of trial patients], including the four with HI-EM, reached week 56 of treatment with sustained response in five of six. These results are summarised in Fig 1.
Among the 23 responders, the median duration of response was 260 d (range 30–650). Responses were sustained in all patients except one. However, all patients have actually stopped thalidomide owing to intolerable side effects.
Table II shows the characteristics of the four patients, including three RA and one RARS, who reached week 56 with a sustained major erythroid response. At that time, all four patients were responsive to treatment, requiring no more PRBC transfusions. Two of these patients (Patients 25 and 45) had 5q-deletion. HI-EM was achieved in Patient 14 at 400 mg/d of thalidomide as early as week 4, at 200 mg/d and at week 8 in Patients 25 and 50, and at 100 mg/d at week 8 in Patient 45. The major response was maintained with the same dose in three (Patients 14, 25, 45), but needed an increase from 200–400 mg/d in Patient 50.
|Patient||Age (years)||Sex||FAB||RBC Trans/month||Cytogenetic findings||HI-EM observed at||Trial dose at week 56||Adverse event Type: grade|
|P14||64||M||RA||2||46XY||400 mg||8||400 mg||Sedation: 1 Constipation: 2|
|P25||66||M||RA||4||46XY,del5q(50%)||400 mg||8||200 mg||Constipation: 1 Neuropathy: 2|
|P45||51||F||RA||4||46XY (40%) T(X;11)(p13;q23) del5q(q15;q13) (50%)||100 mg||12||100 mg||Sedation: 1 Constipation: 1|
|P50||61||M||RARS||5||46XY||200 mg||8||400 mg||Constipation: 2 Peripheral oedema|
Overall, erythroid responses represented 70% of all observed responses (19/27), and 19 of the 36 patients (53%) enroled in the trial for red blood cell transfusions requirement had an erythroid response. However, only four of these became transfusion-independent (11%). The major characteristics of the 19 responding and 17 non-responding patients for the erythroid lineage are shown in Table III. There was no statistical difference between the two groups for the following parameters: FAB, IPSS score, cytogenetic, age and sex, time from diagnosis, haemoglobin concentration, number of PRBC transfusions per month before inclusion, neutrophils and platelets counts, BM blasts (data not shown). In particular, the cytogenetic pattern did not correlate significantly with the haematological response: 50% of patients with a deletion of 5q31.1 had a response, but 57% and 50% of patients with normal karyotype or other cytogenetic abnormality respectively also obtained erythroid response (Table III).
|Variable||Number of patients||Erythroid response No. (%)||P-value|
|RAEB <10%||11||5 (45)|
An objective of the trial was to determine a dose–effect relationship of thalidomide. Thirteen of the 35 (37%) patients who could continue after week 4 increased thalidomide from 200–400 mg/d at week 4, 13 of 27 (48%) from 400–600 mg/d at week 8 and 5 of 15 (33%) from 600–800 mg/d at week 12. However, haematological responses were essentially seen between weeks 4 and 8 in 19 of the 23 responding patients and the dose range at response was between 200 and 400 mg. Only two of the 13 patients who received thalidomide at 600 mg attained a response at that dose, and one at 100 mg/d. None of the five patients who received 800 mg/d of thalidomide achieved a response at that dose (among these, two patients obtained a HI-Em at week 8 under 400 mg/d, which was not increased at 800 mg/d and persisted until week 56 at 400 and 200 mg/d, respectively).
Fourteen of the 17 patients who received at least 16 weeks of treatment had a BM examination at that time. No change in myelodysplastic features was observed in these patients. No cytogenetic improvement was noted at week 16 in eight informative patients, including the two with HI-EM (Patients 25 and 45) (Table II).
Side effects of thalidomide are shown in Table IV. Irrespective of their haematological response, 37 of 47 (78%) patients had to stop thalidomide because of side effects: 26 before week 16 (55%), eight at week 16, and three after week 16. Among these, 38% stopped at 200 mg, 35% at 400 mg, 18% at 600 mg and 9% at 800 mg. The most common side effects were sedation (72%), constipation (40%), fatigue (25%), dizziness (25%) and muscle cramps (12%). Symptoms of peripheral neuropathy were observed in 8% of patients. Transient worsening of neutropenia was detected in 7% of cases but resolved spontaneously in all cases. No worsening of thrombocytopenia related to thalidomide was observed. Headache, nausea, cutaneous reaction, dehydration, dyspnoea or palpitations (the last signs generally linked to significant anaemia), mostly grade 1–2 and reversible, were observed in 1·5–5% of patients. No venous thrombosis occurred in this trial. Sixty-five per cent of the patients experienced more than two side effects. The four patients with HI-EM treated for 56 weeks had a good tolerance to thalidomide with only grade 1 or 2 adverse events (Table II).
|Adverse event||Number of patients||All patients any grade (%)|
|Grade 1 or 2||Grade 3 or 4|
Angiogenic factors and apoptosis assays
Serum levels of angiogenic factors were measured at baseline in 30 patients and compared with the level of 7 sex- and age-matched normal volunteers. VEGF levels were significantly lower in MDS patients compared with healthy volunteers [141·1 pg/ml (LQ-HQ: 105·5–242·5) vs. 364·6 pg/ml (297·1–1558-1) respectively] (P = 0·017). TNF-α levels were increased in MDS patients [3 pg/ml LQ-HQ: 2·3–4·3) vs. 0·9 pg/ml (0·8–1·2)] (P = 0·006). Serum levels of VCAM-1 and E-Selectin were not different from those observed in controls. Cytokine assays were also performed in 26 patients at 4 weeks including 11 responders and 15 non-responders and repeated at week 8 in 10 responders and 10 non-responders. Pretreatment VEGF levels were significantly lower in responders compared with non-responders [105·5 pg/ml (LQ-HQ: 48·9–129·5) vs. 242·5 pg/ml (LQ-HQ: 161·1–402·5)] (P = 0·004) and remained unchanged in responders during treatment, whereas they reached a peak level at week 4 in non-responders. TNF-α levels were not different between responders and non-responders at baseline, and remained statistically unchanged in responders and non-responders under treatment. VCAM-1 and E-selectin levels were not different at baseline between responders and non-responders and remained unchanged during the treatment, irrespective of response.
Angiogenesis was also quantified on the BMB of 32 patients at baseline, in six responders (Patients 14, 25, 45 with HI-EM and Patients 7, 12, 34 with HI-Em) and three non-responders at week 16, and repeated in Patients 14 and 45 at week 56. Given the heterogeneity of FAB subtypes and marrow cellularity, it was not possible to perform inter-patient comparisons of MVD. Only serially examined BM samples provided significant information. MVD significantly increased between weeks 0 and 16 of thalidomide treatment in four of the six responders, including the three patients with HI-EM, remained unchanged in one patient (Patient 12) and decreased in one patient (Patient 7) (Table V). No change in MVD was observed in any of the three non-responders. MVD further increased at week 56 in Patient 45 with persistent HI-EM and remained stable, compared with week 16, in Patient 25 (Table V). Figure 2 illustrates the MVD increase in Patient 45 under thalidomide, using immunohistochemical detection of CD34 antigen expressed on endothelial cells. Similar results were obtained using the anti-CD31 antibody (not shown).
Apoptosis was quantified on the BMB of the same six responders and three non-responders at week 16. Apoptosis decreased in all six responders at week 16 but remained unchanged in the three non-responders (Table V). Figure 3 illustrates the decrease of apoptosis observed in Patient 45 from baseline to week 16.
Only few studies have investigated the role of thalidomide in the treatment of MDS (Raza et al, 2001; Musto et al, 2002; Strupp et al, 2002; Musto, 2004). We conducted a non-randomised phase II dose-escalating trial to study the efficacy and safety of thalidomide in a homogeneous population of 47 low-risk MDS patients. In these patients, supportive care generally continues to be recommended. Some patients benefit from administration of Epo alone or combined with G-CSF, but these are mainly patients with low endogenous Epo serum levels and limited red blood cells transfusion requirement (Hellstrom-Lindberg et al, 1998; Casadevall et al, 2004). Other agents, including retinoids, arsenic trioxide or organic thiols, such as amifostine, have limited efficacy (Casadevall et al, 2004). Our results confirm the high activity of thalidomide on erythropoiesis in low-risk MDS (Raza et al, 2001; Musto et al, 2002; Strupp et al, 2002; Musto, 2004). Overall, haematological response was achieved in 59% of the patients who received at least 8 weeks of thalidomide, the majority of whom (70%) had erythroid responses. On an intention-to-treat analysis, 27% of trial patients reached week 16 of treatment, and 84% of them had improved erythropoiesis; only 12·5% could be treated for 56 weeks with a sustained response in 83% of them. The inclusion criteria were PRBC transfusion-dependent in 76% of the patients (36/47), and erythroid response was observed in 53% of them. However, only 11% of patients with transfusion-dependent anaemia became transfusion-independent in our trial. These results are in agreement with those observed by Raza et al (2001): 15 of 51 evaluable patients (29%) who completed 12 weeks of treatment achieved red blood cell transfusion independence or a >50% decrease in PRBC transfusion need. Similarly, the German study (Strupp et al, 2002) reported that 56% of patients treated with thalidomide, at a median daily dose of 400 mg, obtained HI by intention-to-treat analysis. Another group recently reported their experience with thalidomide in 40 essentially low-risk transfusion-dependent MDS patients (Musto et al, 2002; Musto, 2004). Using an initial dose of 100 mg, which was progressively increased to a maximum of 300 mg, 20% of trial patients became completely transfusion-free (Musto et al, 2002; Musto, 2004). The comparative results of these studies are presented in Table VI.
|Study||Patients (n)||Dose (mg/d)||Responses (ITT) N/(%)||HI-EM/HI-Em (ITT)(%)||Drop-out (%)|
|Raza et al (2001)||83||100–400||31 (19)||13/5||38|
|Strupp et al (2002)||34||100–500||65 (56)||12/17·6||15|
|Musto et al (2002)||40||100–300||32 (20)||20/?||37|
|Present study||47||200–800||23 (49)||8·5/31||55% at W16 78% at W56s|
However, we could not identify a specific population of low-risk MDS patients more likely to respond to thalidomide. In particular, the erythroid response rate was the same in patients with 5q31.1 deletion (50%) or with normal or other cytogenetic abnormalities. In contrast, lenalidomide, the novel analogue of thalidomide, induced the highest response rates in patients with clonal interstitial deletion involving chromosome 5q31.1 (List et al, 2005).
In agreement with previously reported results (Raza et al, 2001), our trial confirms the low efficacy of thalidomide on granulocytic and megakaryocytic lineages, even if our response criteria were stringent, as only patients with neutropenia <1 × 109/l and related bacterial infection or thrombocytopenia <50 × 109/l at inclusion were assessed for granulocytic or megakaryocytic responses, respectively.
The optimal dose of thalidomide and the time required to obtain an erythroid response in MDS patients are not well defined. The majority of responses in this study were observed at 200 or 400 mg/d and it generally took 4–8 weeks until HI occurred. Only two patients had a response at 600 mg/d. Among the five RA patients who received 800 mg/d of thalidomide, all were enroled in the trial for transfusion-dependant anaemia. However, they either had no response at 800 mg/d or did not improve the response achieved with a lower dose. Our results strongly suggest a lack of dose–response effect of thalidomide in MDS. Doses of thalidomide of <200 mg/d may be effective, as suggested by the HI-EM obtained in patient 45 at 100 mg/d and maintained up to week 56 at the same dose. Most patients in the study published by Raza et al (2001) received between 150 and 200 mg/d. The short-time interval of 4 weeks chosen for our study to increase the dose of thalidomide probably lowered the percentage of responses that could have been observed at 200 mg/d if given for a longer period, because this leads to early trial discontinuation in many patients, owing to side effects.
Indeed, all published studies have confirmed that thalidomide is associated with side effects in MDS patients (Raza et al, 2001; Strupp et al, 2002). The side effects observed in this study were very similar to those previously reported and generally correlated with the dose of thalidomide. Night-time administration and high fibre diet were recommended to the patients. The combination of thalidomide and low-dose prednisone could decrease the incidence of side effects, as opposed to thalidomide given alone, as recently reported by Mesa et al (2004) for myelofibrosis with myeloid metaplasia. Such an association should be tested in MDS. Of note, no venous thrombosis was detected in this population, and such an adverse event of thalidomide seems to be very rare in MDS, when given alone. However, as described in multiple myeloma, the combination of thalidomide with prednisone and/or Epo may increase the risk of thrombosis (Steurer et al, 2003).
The second aim of our study was to monitor in MDS patients the impact of thalidomide on several factors implicated in angiogenesis, including TNF-α, VEGF, VCAM-1 and E-selectin. As previously described, we found, when compared with controls, significantly lower pretreatment VEGF levels and significantly higher TNF-α levels in MDS patients (Zorat et al, 2001). Sudhoff et al reported soluble VCAM-1 and E-selectin expression in MDS but had no normal controls for their work (Sudhoff et al, 2002). In our study, we found no significant difference between controls and MDS patients for serum VCAM-1 and E-selectin levels. Because of its known properties, we could have expected a decrease of TNF-α levels. However, TNF-α levels remained unchanged during the treatment. This seemed to exclude down regulation of TNF-α as one of the mechanisms of action of thalidomide in MDS. These results are in agreement with those of Zorat et al (2001) who found no significant decrease in TNF-α levels in responders. Concerning VEGF, we found significantly lower pretreatment levels in responders compared with non-responders, suggesting that VEGF could be a predictive factor of thalidomide response in MDS. Moreover, the VEGF expression pattern seemed to differ between responders and non-responders. Indeed, non-responders had a peak of VEGF at week 4 whereas VEGF levels remained unchanged in responders. In the whole patient population treated by thalidomide, a trend for decrease of VEGF and VCAM-1 levels during treatment was observed, possibly related with the antiangiogenic properties of thalidomide. However, we observed an increase in MVD in four of six responders at week 16 under thalidomide. As VEGF levels remained unchanged in these patients, it can be suspected that other angiogenetic growth factors, such as β-FGF or hepatocyte growth factor, may be increased by thalidomide. Similarly, increased MVD has been reported by Musto (2004) in thalidomide responders, with increased marrow plasma values of VEGF and HGF. Their results and ours do not suggest an antiangiogenesis effect as a mechanism of action of thalidomide. Decreased apoptosis was observed in all six responders that could be analysed by TUNEL at week 16. Similarly, decreased apoptosis of bone marrow cells has been reported in MDS patients treated with thalidomide (Invernizzi et al, 2005) and also in erythroid or CD34+ cells in patients who responded to Epo ± G-CSF treatment (Rajapaksa et al, 1996; Schmidt-Mende et al, 2001). The mechanisms that lead to increased apoptosis in MDS have been studied mainly in the erythroid lineage. Apoptosis is caspase-dependent, implicating the mitochondria and a deregulated Fas/Fas ligand pathway (Hellstrom-Lindberg et al, 2001; Claessens et al, 2002). It has been suggested that endogenous Epo levels may increase in MDS patients responding to thalidomide (Musto, 2004). This increase could contribute to the reduction of erythroid progenitors apoptosis. However, a direct effect of thalidomide on erythropoíesis cannot be excluded.
In conclusion, thalidomide appears to be useful mainly in low-risk MDS patients who need red blood cell transfusions and is less effective on other cytopenias. Low doses of thalidomide that are tolerable in elderly individuals may be effective to stimulate erythropoíesis. We are currently testing the effect of thalidomide given at an initial dose of 50 mg/d for 12 weeks. However, a combination of low-dose thalidomide with prednisone and Epo ± G-CSF may improve the response rates, as opposed to single-agent thalidomide. The mechanisms of action of thalidomide remain unclear, but an anti-TNF-α effect seems to be excluded. Increased angiogenesis and decreased apoptosis may be associated to response, but the exact mechanism still needs to be defined. Moreover, and in contrast to Lenalidomide (List et al, 2005), we confirmed that thalidomide very rarely induced cytogenetic remissions (Strupp et al, 2003). Finally, the precise definition of the population of responding patients is necessary. Pretreatment low VEGF level may be useful as a predictor of response, but this needs to be confirmed in larger studies of MDS patients, along with a comparator group.
We thank greatly Pharmion Laboratory, Paris, France for the help provided to conduct this trial, and particularly Robert Zerbib. We also thank Anne Marie Hagnere for technical assistance. This work was supported by the Direction Régionale à la Recherche Clinique (DRRC): Project CRC00105.
- 2000) Myelodysplastic syndrome and aplastic anemia: distinct entities or diseases linked by a common pathophysiology? Seminars in Hematology, 37, 15–29. , & (
- 2004) The evolution of thalidomide and its IMiD derivatives as anticancer agents. Nature Review Cancer, 4, 314–322. , & (
- 2001) Vascular endothelial cell growth factor is an autocrine promoter of abnormal localized immature myeloid precursors and leukemia progenitor formation in myelodysplastic syndromes. Blood, 97, 1427–1434. , , , , , , & (
- 1997) Fas/Apo-1 (CD95) expression and apoptosis in patients with myelodysplastic syndromes. Leukemia, 11, 839–845. , , , , , , , & (
- 2004) Health, economic, and quality-of-life effects of erythropoietin and granulocyte colony-stimulating factor for the treatment of myelodysplastic syndromes: a randomized, controlled trial. Blood, 104, 321–327. , , , , , , , , , , , , , , , , , , & (
- 2000) Report of an international working group to standardize response criteria for myelodysplastic syndromes. Blood, 96, 3671–3674. , , , , , , , , , , , , , & (
- 2002) In vitro proliferation and differentiation of erythroid progenitors from patients with myelodysplastic syndromes: evidence for Fas-dependent apoptosis. Blood, 99, 1594–1601. , , , , , , , , & (
- 1994) Thalidomide is an inhibitor of angiogenesis. Proceedings of the National Academy of Sciences of the United States of America, 91, 4082–4085. , , & (
- 1998) Thalidomide costimulates primary human T lymphocytes, preferentially inducing proliferation, cytokine production, and cytotoxic responses in the CD8+ subset. Journal of Experimental Medicine, 187, 1885–1892. , , & (
- 1999) Myelodysplasia. New England Journal of Medicine, 340, 1649–1660. & (
- 1998) Treatment of anemia in myelodysplastic syndromes with granulocyte colony-stimulating factor plus erythropoietin: results from a randomized phase II study and long-term follow-up of 71 patients. Blood, 92, 68–75. , , , , , , , , , , , , , , , , , & (
- 2001) Apoptosis in refractory anaemia with ringed sideroblasts is initiated at the stem cell level and associated with increased activation of caspases. British Journal of Haematology, 112, 714–726. , , , , & (
- 2005) Thalidomide treatment reduces apoptosis levels in bone marrow cells from patients with myelodysplastic syndromes. Leukemia Research, 29, 641–647. , , , , , , & (
- 2005) Efficacy of lenalidomide in myelodysplastic syndromes. New England Journal of Medicine, 352, 549–557. , , , , , , , , & (
- 1995) Fas antigen expression on CD34+ human marrow cells is induced by interferon gamma and tumor necrosis factor alpha and potentiates cytokine-mediated hematopoietic suppression in vitro. Blood, 85, 3183–3190. , , & (
- 2004) Durable responses to thalidomide-based drug therapy for myelofibrosis with myeloid metaplasia. Mayo Clinic Proceedings, 79, 883–889. , , & (
- 1993) Thalidomide exerts its inhibitory action on tumor necrosis factor alpha by enhancing mRNA degradation. Journal of Experimental Medicine, 177, 1675–1680. , , , , & (
- 2004) Thalidomide therapy for myelodysplastic syndromes: current status and future perspectives. Leukemia Research, 28, 325–332. (
- 2002) Thalidomide abolishes transfusion-dependence in selected patients with myelodysplastic syndromes. Haematologica, 87, 884–886. , , , , & (
- 1999) Angiogenesis in myelodysplastic syndromes. British Journal of Cancer, 81, 1398–1401. , , , , , , , & (
- 1996) Altered oncoprotein expression and apoptosis in myelodysplastic syndrome marrow cells. Blood, 88, 4275–4287. , , & (
- 1995) Apoptosis in bone marrow biopsy samples involving stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes. Blood, 86, 268–276. , , , , , , , , , , , , , , , & (
- 1996) Novel insights into the biology of myelodysplastic syndromes: excessive apoptosis and the role of cytokines. International Journal of Hematology, 63, 265–278. , , , , , , , , , , , , , , , , , , & (
- 2001) Thalidomide produces transfusion independence in long-standing refractory anemias of patients with myelodysplastic syndromes. Blood, 98, 958–965. , , , , , , , , , , , , & (
- 2000) A hypothesis for the pathogenesis of myelodysplastic syndromes: implications for new therapies. Leukemia, 14, 2–8. & (
- 1991) Thalidomide selectively inhibits tumor necrosis factor alpha production by stimulated human monocytes. Journal of Experimental Medicine, 173, 699–703. , , , & (
- 2001) Granulocyte colony-stimulating factor inhibits Fas-triggered apoptosis in bone marrow cells isolated from patients with refractory anemia with ringed sideroblasts. Leukemia, 15, 742–751. , , , , , , & (
- 1991) The occurrence subtype and significance of haemopoietic inhibitory T cells (HIT cells) in myelodysplasia: an in vitro study. Leukemia Research, 15, 597–601. & (
- 2003) Thromboembolic events in patients with myelodysplastic syndrome receiving thalidomide in combination with darbepoietin-alpha. British Journal of Haematology, 121, 101–103. , , & (
- 2002) Thalidomide for the treatment of patients with myelodysplastic syndromes. Leukemia, 16, 1–6. , , , , & (
- 2003) Cytogenetic response to thalidomide treatment in three patients with myelodysplastic syndrome. Leukemia, 17, 1200–1202. , , , & (
- 2002) Levels of circulating endothelial adhesion molecules in patients with myelodysplastic syndromes. International Journal of Oncology, 20, 167–172. , & (
- 2003) Apoptosis in the myelodysplastic syndromes. Current Hematology Reports, 2, 186–192. & (
- 2001) The clinical and biological effects of thalidomide in patients with myelodysplastic syndromes. British Journal of Haematology, 115, 881–894. , , , , , , , , , & (