Etoposide, methylprednisolone, cytarabine and cisplatin successfully cytoreduces resistant myeloma patients and mobilizes them for transplant without adverse effects


Dr Shirley D'Sa, Department of Haematology, University College London Hospitals NHS Trust, Cecil Flemming House, Grafton Way, London WC1E 3DB, UK. E-mail:


Myeloma remains incurable with a median survival of 4 years, but outcome can be improved by the use of high-dose therapy. We used the etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP) regimen as second-line therapy in 42 newly diagnosed myeloma patients who had failed vincristine, adriamycin and dexamethasone (VAD)- type therapy (n = 36), responded to first-line treatment but persisted in having significant residual marrow plasmacytosis (n = 5) or failed prior stem cell harvesting (n = 1), with the dual aim of improving disease response and mobilizing peripheral blood stem cells. Fourteen of 21 (67%) patients with no change or progressive disease after VAD responded to ESHAP; seven of 12 (58%) patients with minor response converted to partial response. Marrow plasmacytosis fell from a median of 52% at diagnosis to 23·5% after primary therapy and to15% after ESHAP. ESHAP chemotherapy was well-tolerated. There were 11 admissions due to febrile neutropenia (n = 7), nausea and vomiting (n = 2), pneumonia (n = 1) and perforated bowel (n = 1). Renal function deteriorated in 13 of 42 patients after ESHAP, but none required renal support. ESHAP mobilization was performed in 32 patients of whom 87% achieved a CD34+ yield >2 × 106/kg. In all, 38 patients proceeded to high-dose therapy. The overall survival for all patients was 62% at 4 years following ESHAP. We conclude that ESHAP has acceptable toxicity and efficient stem cell mobilizing capability, effectively cytoreduced this chemoresistant group of patients, and did not appear to adversely affect transplant outcome.

Myeloma remains essentially an incurable disease, with a median survival of 4 years. Autologous stem cell transplantation (ASCT) can improve outcome in patients up to the age of 65 years of age when compared with conventional chemotherapy (Attal et al, 1996; Child et al, 2003). In chemosensitive patients, complete remission (CR) rates of 24–75% have been achieved, with an event-free survival (EFS) of 24–30 months (Anderson et al, 1993; Cunningham et al, 1994) and a second transplant may further improve EFS and overall survival (OS), especially in patients who fail to achieve a very good partial response (PR) after the first transplant (Attal et al, 2003). Results of studies which include patients with refractory/resistant disease have been less impressive, with CR rates ranging from 17–40% to EFS of 10–30 months (Harousseau et al, 1995; Bensinger et al, 1996; Rajkumar et al, 1999; Vesole et al, 1999), although direct comparisons between studies is difficult because different criteria were used to define CR. Harousseau et al (1995) reported a median OS of 54 months in patients who responded to primary treatment compared with 30 months for non-responders to primary treatment. Vesole et al (1999) reported on 72 patients with primary unresponsive disease and showed a median EFS and OS of 21 and 47 months respectively, and a CR rate of 30% [as defined by the less stringent Southwestern Oncology Group (SWOG) criteria of >75% reduction in paraprotein]. Rajkumar et al (1999) demonstrated a CR (negative by immunofixation) rate of 17% and EFS of 26 months in their subgroup of primary refractory myeloma patients. In the study by Bensinger et al (1996), which included two-thirds of the patients with resistant disease, 40% achieved a CR overall. Thus some, but not all patients with resistant myeloma can be salvaged with high-dose melphalan and ASCT, and the greatest benefit seems to be achieved if myeloablative therapy is performed early (Alexanian et al, 1994).

Other salvage regimens have been employed to treat vincristine, adriamycin and dexamethasone (VAD)-resistant myeloma, including etoposide, dexamethasone, cytarabine (Ara-C), cisplatin (EDAP) (Barlogie et al, 1989), and intensive sequential therapy with cyclophosphamide and etoposide (Dimopoulos et al, 1994) with response rates of 40% and 35% respectively. The toxicity of both approaches was moderately high with an 80% admission rate due to neutropenic sepsis after EDAP with a fatal outcome in 20%, and a 6% mortality rate related to intensive sequential therapy. More recently, thalidomide-containing regimens have been used to improve outcome in high-risk myeloma patients prior to SCT with promising results (Moehler et al, 2001; Kropff et al, 2003; Lee et al, 2003), but the role of thalidomide in the early treatment of myeloma remains to be formally established by randomized-controlled trials.

Many patients who fail to respond to first-line therapy with VAD-type regimens still remain candidates for high-dose therapy. An effective regimen that facilitates stem cell collection is needed for such patients. In an attempt to identify an alternative regimen with good cytoreductive activity, minimal toxicity and stem cell mobilizing potential, we have evaluated etoposide, methylprednisolone, Ara-C and cisplatin (ESHAP), a regimen that is used in the management of relapsed lymphoma (Velasquez et al, 1994).

The aim of this study was to determine the efficacy of ESHAP as a salvage and mobilizing regimen in patients failing or responding inadequately to VAD-type therapy but who were still suitable candidates for SCT. We report the results of using ESHAP as second-line therapy in 42 patients, with specific emphasis on antitumour efficacy, toxicity, stem cell mobilization and outcome of high-dose therapy.

Patients and methods

Patient characteristics and therapy

Forty-two patients were treated with ESHAP as salvage therapy between October 1997 and June 2003 in three Cancer Centres in the United Kingdom. Patient details including prior therapy are given in Table I. All had stages II or III disease including two patients who fulfilled the criteria for plasma cell leukaemia. Patients received primary therapy with a median of four courses of VAD-type chemotherapy (range: 3–7) and three received additional therapy, with melphalan and prednisolone (n = 2) or adriamycin, BCNU (carmustine), cyclophosphamide and melphalan (ABCM) (n = 1). Prior to treatment with ESHAP, six patients had a PR, 12 a minor response (MR), 19 had shown no change (NC), and five had progressive disease (PD) according to the European Group for Blood and Marrow Transplantation/International Bone Marrow Transplant Registry/Autologous Blood and Marrow Transplant Registry (EBMT/IBMTR/ABMTR) criteria (Blade et al, 1998).

Table I.  Patient and disease characteristics and response to induction therapy.
Patient characteristicn = 42
  1. VAD, vincristine, adriamycin, dexamethasone; VAMP, vincristine, adriamycin, methylprednisolone; ABCM, adriamycin, BCNU, cyclophosphamide, melphalan; MP, melphalan, prednisolone; Z-DEX, idarubicin, dexamethasone; ESHAP, etoposide, solumedrol (methylprednisolone), high-dose Ara-C, cisplatin; PR, partial response; MR, minor response; NC, no change; PD, progressive disease; BCNU, carmustine; Ara-C, cytarabine.

 Male/female28/14 (2:1)
Age at diagnosis (years)
Myeloma isotype (%)
 IgGλ5 (12)
 IgGκ23 (56)
 IgAλ3 (7)
 IgAκ5 (12)
 λ light chain only1 (2)
 κ light chain only3 (7)
 Non-secretory2 (4)
Stage (Durie/Salmon)
First-line therapy
 VAD or VAMP or Z-DEX39
 VAD + MP1
Response to first line Tx

Thirty-six patients (86%) had NC, MR or PD following VAD-type therapy, and hence required further cytoreductive therapy. Of the six patients who achieved a PR after primary therapy (in terms of paraprotein levels), five continued to demonstrate residual marrow plasmacytosis ranging from 20 to 79%, and one failed to mobilize adequate number of stem cells despite two cycles of cyclophosphamide/granulocyte colony-stimulating factor (G-CSF) priming.

In view of the potential renal toxicity of cisplatin in this group of vulnerable patients, each patient underwent a measurement of glomerular filtration rate (GFR) by ethylenediaminetetra acetic acid (EDTA) clearance before and after receiving ESHAP. In order to proceed with cisplatin at full dose, a minimum GFR of 80 ml/min was needed. Patients with a GFR >20 and <80 ml/min, received dose reduction of cisplatin and/or Ara-C and/or etoposide according to protocol. Patients with GFR below 20 ml/min were ineligible for ESHAP.

The ESHAP was administered over 5 d via a Hickman line in an inpatient setting, followed by alternate-day follow-up of renal function and blood counts in the outpatient clinic. The ESHAP regimen consists of Etoposide 40 mg/m2/d, on days 1–4; methylprednisolone (Solumedrol) 500 mg/d, on days 1–5; High dose Ara-C 2000 mg/m2, on day 1 and cisPlatin 25 mg/m2/d by continuous intravenous (i.v.) infusion, on days 1–4 (total dose 100 mg/m2). G-CSF (Lenograstim 263 μg/d or filgrastim 300 μg/d) was administered from day 6 until stem cell harvest on day 15 (with a prerequisite white cell count of 3 × 109/l). In the event that the CD34 stem cell yield was <2 × 106/kg on day 15, a dose of G-CSF was administered after stem cell collection on day 15, followed by a further collection on day 16. During chemotherapy administration, hydration was given in the form of i.v. fluids with mannitol and potassium replacement. Low-dose diuretic use was occasionally required to control fluid balance. Other supportive measures included granisetron 1 mg, days 1–6; ranitidine 150 mg b.i.d. for 1 week, prednisolone 0·5% eye drops on days 1–8.

In some patients stem cell mobilization proceeded with cyclophosphamide (1·5 g/m2 or 4 g/m2 on day 1) and G-CSF (Lenograstim 263 μg/d or filgrastim 300 μg/d on days 2–12 until stem cell collection on day 12, with a prerequisite white cell count of 5 × 109/l) rather than ESHAP (including one patient who had two separate attempts) for logistical reasons relating to the availability of the cell separator. Stem cells were collected on a COBE spectra (COBE Laboratories Ltd, Gloucester, UK) or Baxter CS3000 (Baxter Healthcare Ltd, Berkshire, UK) cell separator when the total white blood cell count was >5 × 109/l. One patient did not undergo autologous stem cell harvest out of personal choice, and the patient with plasma cell leukaemia proceeded to an unrelated myeloablative allogeneic transplant after ESHAP therapy without a stem cell harvest.

Details regarding stem cell harvest and transplantation are shown in Table II. Conditioning therapy for the autograft consisted of melphalan 200 mg/m2 in all but one patient who received melphalan 140 mg/m2 and total body irradiation (TBI), and another who received melphalan 140 mg/m2 due to a lower CD34 stem cell yield of 0·9 × 106/kg. In all patients who underwent a reduced intensity allograft, conditioning consisted of fludarabine 30 mg/m2/d on day 7 to day 3, melphalan 140 mg/mg2 on day 2 and alemtuzumab 20 mg/d i.v. infusion on day 8 to day 4 (Peggs et al, 2003). Following transplant, patients with residual disease or mixed haematopoietic chimaerism entered a programme of dose-escalated donor lymphocyte infusions (DLI) starting at 106 CD3 cells/kg at 6 months. Conditioning therapy for T-replete myeloablative allogeneic transplantation consisted of melphalan 110 mg/m2 and TBI (12 Gy in six fractions).

Table II.  Stem cell harvest yield, transplant conditioning and outcome.
PatientMobilization regimen Autologous CD34+ cell yield (×106/kg)Type of transplant(s)Conditioning therapyNumber of days to neutrophils (>0·5 × 109/l)Clinical outcome, length of follow-up (number of days following ESHAP)
  1. ESHAP, etoposide, solumedrol (methylprednisolone), high-dose Ara-C, cisplatin; PBSCT, peripheral blood stem cell transplant; MUD, matched unrelated donor; Sib, sibling donor; TBI, total body irradiation; Flu/Mel/alemtuzumab, fludarabine, melphalan, alemtuzumab conditioned reduced intensity transplant; G-CSF, granulocyte colony-stimulating factor; GVHD, graft versus host disease; CR, complete response; PR, partial response; PD, progressive disease; NA, not applicable; NK, not known; Ara-C, cytarabine.

1ESHAP21·1Autologous PBSCTMelphalan 2009CR (652)
2ESHAP15·8Autologous PBSCTMelphalan 20010PR (638)
3ESHAP1·6Autologous PBSCTMelphalan 20015PR (481)
4ESHAP23·2Autologous PBSCTMelphalan 20010CR (577)
5ESHAP16·4Autologous PBSCTMelphalan 20011PD/further salvage therapy (654)
6ESHAP3·9Autologous PBSCTMelphalan 20010PR (500)
71. Cyclophosphamide0·3    
2. Cyclophosphamide0·2    
3. ESHAP3·2Autologous PBSCTMelphalan 20013PR (815)
8ESHAP15·8Autologous PBSCTMelphalan 20010PR (655)
9ESHAP12·1Autologous PBSCTMelphalan 20011CR (946)
10ESHAP1·4Autologous PBSCTMelphalan 20011PD/died (58)
11ESHAP11·7Autologous PBSCTMelphalan 20010PR (687)
12ESHAP1·8Autologous PBSCTMelphalan 20012CR (447)
13ESHAP29·3Autologous PBSCTMelphalan 20013PD/further salvage therapy (1302)
14ESHAP28·3Autologous PBSCTMelphalan 20011PD/died (247)
15ESHAP3·1Autologous PBSCTMelphalan 20015PR (1165)
16ESHAP12·2Autologous PBSCTMelphalan 20011CR/Thal maintenance (806)
17ESHAP3·8Autologous PBSCTMelphalan 20012PR (372)
18ESHAP12·6Autologous PBSCTMelphalan 20013PD/further salvage therapy (402)
19ESHAP5·1Autologous PBSCTMelphalan 20011PD/died (422)
20ESHAP10·9Autologous PBSCTMelphalan 20011PD/further salvage therapy (205)
21ESHAP2·3Autologous PBSCTMelphalan 20012PR (190)
22ESHAP3·6Autologous PBSCTMelphalan 2009PR (162)
23Cyclophosphamide16·7Autologous PBSCTMelphalan 2009PR/Thalidomide maintenance (1390)
24Cyclophosphamide2·1Autologous PBSCTMelphalan 20013CR (1930)
25Cyclophosphamide0·9Autologous PBSCTMelphalan 14010PD (920)
26Cyclophosphamide2·4Autologous PBSCTMelphalan 200NKPD/further salvage therapy (680)
27ESHAP8·21. Autologous PBSCTMelphalan 140/TBI13PD/further salvage therapy
2. Autologous PBSCTMelphalan 20010PR (2228)
28Cyclophosphamide8·31. Autologous PBSCTMelphalan 20013PD/further salvage therapy
2. Autologous PBSCTMelphalan 20012PR/Thalidomide maintenance (1157)
29Cyclophosphamide9·31. Autologous PBSCTMelphalan 20015PD/further salvage therapy
2. MUD low intensity allograftFlu/Mel/alemtuzumab14CR (1317)
30ESHAP16·71. Autologous PBSCTMelphalan 20013PD/further salvage therapy
2. MUD low intensity allograftFlu/Mel/alemtuzumabNKPD/died (699)
31ESHAP43·81. Autologous PBSCTMelphalan 22015PD/further salvage therapy
2. MUD low intensity allograftFlu/Mel/alemtuzumab19PD/died (773)
32Cyclophosphamide 5·91. Autologous PBSCTMelphalan 22022Stable
2. MUD low intensity allograftFlu/Mel/alemtuzumab11PD/died (889)
331. Cyclophosphamide 1·4    
2. ESHAP 0·9    
3. G-CSF alone 5SIB low intensity allograftFlu/Mel/alemtuzumab16PD/further salvage therapy (1401)
34ESHAP 2·7SIB low intensity allograftFlu/Mel/alemtuzumab10PD/died (1203)
35ESHAP14·1MUD low intensity allograftFlu/Mel/alemtuzumab12PR (1479)
36Not doneNot doneT-replete MUD allograftMelphalan110/TBI30CR (982)
37ESHAP 4·7T-replete MUD allograftMelphalan110/TBI>90 (GVHD)PD/GVHD/died (136)
38ESHAP 1·9T-replete Sib allograftMelphalan110/TBI12CR (660)
39ESHAP 0·4NoneNoneNAPD/further salvage therapy (742)
40NANot doneNoneNoneNAPR (759)
41ESHAP 4·2NoneNoneNAPD/died (181)
421. ESHAPAborted due to systemic infection    
2. Cyclophosphamide 0·9NoneNoneNAPD/further salvage therapy (422)

Response criteria

Paraprotein levels, Bence-Jones protein (BJP) levels and bone marrow plasma cell percentage were measured before and after ESHAP. Skeletal radiographs were performed every 6–12 months, or in the event of relevant symptoms. EBMT/IBMTR/ABMTR criteria were used to assess response to primary therapy, ESHAP therapy and transplant procedure (Blade et al, 1998). CR was defined as the disappearance of M-protein from serum and urine by immunofixation and <5% plasma cells in the bone marrow. PR was defined by at least a 50% reduction of the initial M-protein concentration and a reduction of Bence-Jones proteinuria by >90% or to <0·2 g/24 h. MR was defined by a 25–49% reduction of initial M-protein, and a reduction in Bence-Jones proteinuria by 50–89% but exceeding 0·2 g/24 h. In order to fulfil the criteria for CR, PR or MR, patients were required to show no signs of disease progression, such as persistent hypercalcaemia, progressive renal insufficiency, bone marrow failure or skeletal disease. In all patients, responses not satisfying the criteria for PR or MR were termed NC. PD was defined as >25% increase in the level of serum paraprotein and/or 24-h urinary light chain excretion and/or marrow plasmacytosis, an increase in the size of existing bone lesions or the development of new bone lesions or extramedullary disease. For non-secretory patients, CR was defined as <5% plasma cells, PR as >50% reduction in marrow plasmacytosis and MR as 25–49% reduction in marrow plasmacytosis, each maintained for a minimum of 6 weeks.



The ESHAP regimen was generally well-tolerated. Thirty-three patients received one course, six patients received two courses and three patients received three courses of ESHAP. No patient received prophylactic antibiotics. No patient required platelet transfusions following ESHAP. Thirty of the total of 42 patients (71%) remained asymptomatic despite documented neutropenia (neutrophils <0·5 × 109/l) in most, ranging from days 6–21 after ESHAP. Twelve patients (29%) were readmitted with complications following ESHAP. Seven patients (17%) received and responded to broad-spectrum i.v. antibiotics for neutropenic fever. None required additional support, such as antifungal drugs, inotropes or intensive care. One patient developed a perforated colon on day 3 of ESHAP (not neutropenic at the time) and underwent hemi-colectomy with antibiotic cover and fully recovered within 6 weeks of surgery. He subsequently received a further course of ESHAP without adverse event. Two patients were admitted with nausea and vomiting, requiring i.v. hydration but without evidence of sepsis. One patient, who was no longer neutropenic but had persistent and unexplained pyrexia for 3 weeks following ESHAP, developed a cough and dyspnoea and responded to treatment with clindamycin and primaquine for a presumed pneumocystis pneumonia infection. Finally, one patient developed palpitations several weeks after the neutropenic phase was over and was subsequently found to have bacterial endocarditis. This patient underwent aortic valve replacement and subsequently proceeded to ASCT. There were no treatment-related deaths.

The GFR data were available in 31 patients. Median reduction in GFR values was 21·5% (range: 7–79%), in 16 of 25 patients with GFR measurements pre- and post-ESHAP but in nine of these, GFR remained normal (>70 ml/min). Six patients had documented post-ESHAP GFR measurements only, of which two were subnormal (65 and 67 ml/min) and four were normal. Of the remaining patients for whom GFR levels could not be located, 10 had normal serum creatinine levels (<120 μmol/l) and one patient showed an increase in serum creatinine level from 120 to 250 μmol/l following ESHAP. This patient subsequently developed acute renal failure associated with an unconnected episode of pneumonia several months later and required haemodialysis. No other patient required renal replacement therapy at any stage.

Disease response

Response to treatment was assessed using urine light chain measurements in BJP-only myeloma patients (n = 4), percentage change in marrow plasmacytosis in non-secretory patients (n = 2) and serum paraprotein measurements in the remainder (n = 36). Alterations in serum or urine paraprotein levels before and after ESHAP chemotherapy for all patients with secretory disease are illustrated in Fig 1. By EBMT criteria, six patients had a PR, 12 had a MR, 19 showed NC and five had PD after first-line therapy. Of the 24 patients with NC or PD after first-line therapy, 16 (67%) responded to ESHAP (one converting to CR, eight to PR and seven to MR); disease status in seven remained unchanged and one patient showed further disease progression (Table III). Of the 12 patients with MR after VAD-type therapy, seven (58%) converted to PR, four showed NC and one had PD after ESHAP. All six patients with PR after first-line therapy remained in PR after ESHAP, but demonstrated a reduction in marrow plasmacytosis from a median of 31% (range: 10–79%) to a median of 10% (range: 5–25%). Thus, following ESHAP therapy a further disease response was seen in 52% of patients who demonstrated an upgraded clinical response. Overall, from completion of induction therapy to completion of ESHAP, serum paraprotein fell from a median of 33·5 g/l (range: 2–62) to a median of 27·5 g/l (range: 0–51), and urine light chain measurements from a median of 2 g/24 h (range: 1–3·2 g/24 h) to a median of 1·2 g/24 h (range: 0·2–1·57 g/24 h), P < 0·001 by Wilcoxon matched pairs signed rank test.

Figure 1.

Changes in serum and urine paraprotein levels pre- and post-etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP). Serum paraprotein levels (n = 36) or 24 h urinary Bence-Jones protein (BJP) excretion (n = 4) were measured following primary therapy and 4 weeks post-ESHAP. The values shown illustrate the change in levels as a result of receiving ESHAP therapy.

Table III.  Response to ESHAP.
Disease status after primary therapy (n)Disease status after ESHAP (n)
  1. *Five of six patients showed a median reduction in marrow plasmacytosis from 31% (range, 20–79) to a median of 10% (range, 5–25).

  2. ESHAP, etoposide, solumedrol (methylprednisolone), high-dose Ara-C, cisplatin; CR, complete response; PR, partial response; MR, minor response; NC, no change; PD, progressive disease; Ara-C, cytarabine.

PR (6)PR (6)*
MR (12)PR (7)
MR (3)
NC (1)
PD (1)
NC (19)PR (6)
MR (7)
NC (6)
PD (5)CR (1)
PR (2)
NC (1)
PD (1)

In addition to assessment using EBMT criteria, the effect of ESHAP on the degree of bone marrow infiltration by plasma cells was assessed where possible. In contrast to the serum and urine paraprotein data, which were available in all but the non-secretory patients at each time-point, a complete set of data regarding plasma cell infiltration at diagnosis, after VAD-type therapy and after ESHAP were available in two-thirds of patients (n = 27), showing median marrow plasmacytosis at diagnosis of 52% (range: 10–100), after first-line therapy of 21% (range: 5–100), and after ESHAP of 15% (range: 0–80), P < 0·001 for response to ESHAP by Wilcoxon signed rank test. In contrast, the paraprotein response in the same group of patients was less marked, falling from a median of 27 (range: 6–62) to 26 g/l (range: 0–51), P < 0·01. These data are illustrated in Fig 2. The durability of the response to ESHAP could not be easily assessed in this study since the majority of patients proceeded to SCT after a median of 2 months (range: 1–13) following ESHAP.

Figure 2.

Reduction in median marrow plasmacytosis and serum paraprotein in the same 27 patients. The black line indicates median values in each case, and the grey squares the individual values at diagnosis, post-vincristine, adriamycin and dexamethasone (VAD) and post-etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP). The figures indicate a relatively greater reduction in marrow plasmacytosis values compared with serum paraprotein values in response to ESHAP therapy.

Stem cell mobilization

A total of 40 patients underwent peripheral blood stem cell harvest. Thirty-three patients underwent peripheral blood stem cell mobilization with ESHAP/G-CSF and nine with cyclophosphamide/G-CSF, including two patients who underwent more than one type of priming schedule. In ESHAP-mobilized patients, median CD34+ yield was 7·3 × 106/kg (range: 0·4–43·8 × 106/kg). The minimum CD34+ cell requirement of 2 × 106/kg was achieved in 87% of ESHAP mobilized patients and included one (patient 9) with a PR following primary therapy who had failed to mobilize after two attempts with cyclophosphamide/G-CSF-priming. In the cyclophosphamide-mobilized group, the median CD34+ cell yield was 2·4 × 106/kg (range: 0·3–16·7 × 106/kg) and 67% achieved a collection in excess of the minimum requirement. Patient 42 (NC after primary therapy, MR after ESHAP) had one attempt at mobilization with ESHAP, which was aborted due to neutropenic fever commencing on day 7, and two further separate mobilization attempts with cyclophosphamide priming, yielding 0·3 and 0·6 × 106 CD34+ cells/kg. Patient 33 had 25% bone marrow plasmacytosis following ESHAP and was a poor mobiliser. He underwent three attempts at stem cell priming including cyclophosphamide/G-CSF, followed by ESHAP/G-CSF and then G-CSF alone, yielding a total of 7·3 × 106/kg CD34+ cells over a 3 month period. Patient 39 failed an attempt at ESHAP mobilization due to active sepsis related to an infected leg ulcer. All other patients underwent stem cell mobilization and harvest according to protocol. Details of stem cell yield and transplant procedure and conditioning are given in Table II.

High-dose therapy

Thirty-eight patients proceeded to high-dose therapy at a median of 2 months (range: 1–13 months) following the final ESHAP chemotherapy. Twenty-six patients had one ASCT. Two patients had two non-tandem (3 years apart in each case) ASCT due to disease progression after the first transplant. In both cases, a second aliquot of CD34+ cells was available. After a median of 9 months (range: 6–13) following ESHAP, four patients had an ASCT followed by reduced intensity allograft, of which three required intervening salvage therapy for disease progression. Three patients proceeded to reduced intensity allograft at a median of 2 months after ESHAP, and three had a conventional myeloablative allograft, at a median of 2 months following ESHAP. Median time to engraftment of neutrophils >0·5 × 109/l was 11 d (range: 9–22) following autograft, 13 d (range: 10–19) following reduced intensity allograft and 30 d (range: 12–90) following myeloablative allograft. In all but one case, the transplant procedure proceeded without unexpected complications, such as infections, renal toxicity or increased inpatient stay. The cited case (patient 37) had delayed engraftment (90 d) after her T-replete matched unrelated donor (MUD) allograft, which was complicated by acute graft versus host disease (GVHD). Four patients did not undergo a transplant procedure due to personal choice (n = 1), inadequate harvest (n = 2) and death from PD (n = 1).

Clinical outcome

Of the four patients who did not proceed to transplant, one remained refractory to ESHAP and died from PD within 6 months, one progressed within 4 months of ESHAP and two remained refractory to ESHAP. All three survivors have since responded to thalidomide-containing salvage therapy and remain in PR at a median of 20 months (range: 14–20) following initial treatment with ESHAP.

Of the remaining 38 patients that proceeded to at least one transplant procedure, six are in CR, 19 in PR, eight are receiving further salvage therapy for PD and nine have died of PD at a median of 24·3 months (range: 4·5–68·8) following initial treatment with ESHAP. The duration of follow-up for each patient, following treatment with the first course of ESHAP is shown in Table II. Of those who proceeded to one or two autografts (n = 28), the progression-free survival (PFS) was 20% (Fig 3A) and the OS was 81% (Fig 3B) at 4 years following the first autograft. Seven patients were refractory to ESHAP but for those who were ESHAP-responsive the 4-year PFS was 22% (Fig 4). The OS for all patients in this study was 62% at 4 years (Fig 5). The groups that underwent allogeneic transplantation following ESHAP were too small for subgroup analysis.

Figure 3.

Progression-free survival (A) and overall survival (B) in patients who had at least one autograft salvage with etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP) (n = 28).

Figure 4.

Progression-free survival following salvage in etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP)-responsive patients (n = 35).

Figure 5.

Overall survival following salvage in all patients who received etoposide, methylprednisolone, cytarabine and cisplatin (ESHAP) (n = 42).


In this study, ESHAP was administered as a salvage and mobilizing regimen to patients who were refractory to VAD-type therapy and to those who had a PR according to EBMT criteria, but only a minor reduction in bone marrow plasmacytosis following VAD. Our results indicate that ESHAP has acceptable toxicity, useful cytoreductive potential and efficient peripheral stem cell mobilizing capability in patients with suboptimal responses to front-line therapy. In particular, this non-cross-resistant regimen achieved a substantial degree of reduction in BM plasmacytosis in a chemoresistant patient group and was able to upgrade disease response in nearly two-thirds of these patients. In addition, it did not appear to increase the toxicity of subsequent high-dose therapy.

Following high-dose therapy with an autograft for newly diagnosed myeloma, improved survival is associated with attainment of CR (Barlogie et al, 1986; Cunningham et al, 1994; Attal et al, 1996).The issue of whether the magnitude of response to initial treatment prior to high-dose therapy also correlates with survival benefit remains uncertain. Some studies suggest that response to induction chemotherapy is not essential to obtain survival benefit from high-dose melphalan and ASCT (Singhal et al, 2002). On the contrary, relapse following or refractoriness to initial chemotherapy may result in reduced EFS and OS after ASCT (Fermand et al, 1998; Gertz et al, 2000), and chemosensitivity may accurately predict the probability of CR following high-dose therapy (Alexanian et al, 2001). Thus, although the benefit to subsequent survival of achieving a maximal degree of tumour load reduction before entry to SCT remains unproven, it cannot be dismissed on the basis of current evidence.

Consequently, a number of salvage therapies for VAD-resistant myeloma have been explored, including combination chemotherapy schedules and thalidomide-containing regimens. The aims of such regimens are usually twofold, including tumour-reduction and stem cell mobilization. Etoposide- and/or platinum-based regimens including EDAP, intensive sequential therapy (3 g/m2 cyclophophamide and 900 mg/m2 etoposide) and dexamethasone, cyclophosphamide, etoposide and cisplatin (DCEP) have been reported to have response rates of around 40% and/or good mobilization potential (Barlogie et al, 1989; Dimopoulos et al, 1994; Lazzarino et al, 2001). Our study shows that the ESHAP regimen compares favourably with other salvage regimens in its low toxicity (29% readmission and no treatment-related deaths) and good disease activity (55% response rate). An important practical advantage of ESHAP is that it enabled successful stem cell mobilization and obviated the need for additional priming in the majority of our patients. Twenty-eight of 32 patients (87%) who received ESHAP/G-CSF priming were successfully harvested after ESHAP/G-CSF priming alone, compared with those who received cyclophosphamide priming, of which 67% had a yield above the threshold of 2 × 106/kg CD34 cells.

An alternative approach to the management of VAD-resistant myeloma is to employ thalidomide-containing regimens to utilize the synergistic action between thalidomide, dexamethasone and conventional chemotherapy agents. Combinations of thalidomide and dexamethasone with agents such as cyclophosphamide and etoposide (Moehler et al, 2001); cisplatin, doxorubicin, cyclophosphamide and etoposide (Lee et al, 2003) and hyperfractionated cyclophosphamide (Kropff et al, 2003) have been used in the setting of refractory and relapsed myeloma with reasonable success. Although impressive disease responses may be seen in these patients, the effect on the outcome after high-dose therapy is not known.

Regimen-related toxicity remains a limiting, but not insurmountable, factor to this thalidomide-containing approach. In the studies cited above, the main problems included peripheral neuropathy and an increased incidence of arterial and venous thromboembolic events as well as myelosuppression and infections. These effects tend to be more prominent when thalidomide is used at higher doses, which is often necessary in the setting of relapsed or refractory disease when the disease burden is high and agents such as doxorubicin are included (Zangari et al, 2001). Neurotoxicity in particular may be a cumulative phenomenon with up to 60% patients affected to some degree after 8 months of treatment (Cellini et al, 2003) and may be irreversible if the drug is not promptly withdrawn. Venous thrombosis associated with thalidomide combinations cannot be prevented by simple strategies such as low-dose warfarin or aspirin, but requires full anticoagulation with therapeutic warfarin or low-molecular weight heparin (Zangari et al, 2002; Weber et al, 2003), which may be difficult to accurately control or result in worsening of pre-existing osteopenia respectively.

Although some patients respond rapidly to thalidomide-containing regimens, many patients require several months’ treatment to achieve maximal response, after which a mobilization regimen is still required if high-dose therapy remains a therapeutic goal. In these patients, delay in proceeding to high-dose therapy may have an adverse effect on long-term outcome.

While the long-term outcome with thalidomide- and dexamethasone-containing regimens in VAD failures is not known, a median follow-up of 23 months shows a 62% OS at 4 years in this group of patients treated with the ESHAP regimen. Based on this experience, we would suggest that the ESHAP regimen should be considered for patients who remain eligible for high-dose therapy and have normal renal function, but have evidence of disease progression within or an inadequate response to a maximum of four courses of VAD. In this setting ESHAP could be used to improve disease response and allow high-yield stem cell harvesting with a view to a transplant procedure. For those who are no longer considered candidates for high-dose therapy, a thalidomide-containing salvage regimen could be used. While the role of newer agents such as Velcade and Revimid in this category of patients awaits clarification by current studies, an easily available and effective stem cell-friendly regimen, such as ESHAP, remains an important tool in the achievement of the principal treatment aim for younger myeloma patients, namely a successful stem cell harvest and ASCT. With the role of ASCT now firmly established in the treatment of myeloma and the continued exploration of reduced intensity allografts, it is important that any chemotherapy regimen given does not increase the toxicity associated with these procedures. Our study would suggest that ESHAP is successful in this regard.

In conclusion, the results of this study indicate that ESHAP chemotherapy can be used effectively and safely to treat non-responders to first-line therapy, permitting stem cell harvest with additional therapeutic benefit. In addition, it allows patients to proceed to early transplantation without significant increase in medullary or extra-medullary toxicity and no apparent increase in early transplant-related mortality.