The clinical outcome and toxicity of high-dose chemotherapy and autologous stem cell transplantation in patients with myeloma or amyloid and severe renal impairment: a British society of blood and marrow transplantation study


Dr J.M. Bird, Consultant Haematologist, Avon Haematology Unit, Horfield Road, Bristol BS2 8ED, UK. E-mail:


The outcome of high-dose chemotherapy (HDT) was evaluated retrospectively in 27 patients with myeloma and four patients with AL amyloidosis with severe renal impairment. Twenty-three patients were receiving dialysis and the rest had a creatinine clearance of <20 ml/min. The median melphalan dose was 140 mg/m2 (range: 60–200 mg/m2), but 10 patients (37%) received 200 mg/m2. Myeloid and platelet engraftment were similar to that seen in patients without renal failure. Five of 27 patients died of transplant-related toxicity before the day 100. Twenty of 27 patients had a response (70%). The median time to disease progression was 32 months (range: 6–54 months) and the median time to best response was 6·5 months. Four of 17 evaluable patients (24%) became dialysis-independent at a median of 5 months post-HDT/stem cell transplantation. At a median follow-up of 70 months, 7/23 patients with myeloma were alive but three of these seven patients had progressive disease. Two of the four patients with amyloidosis have survived. HDT is feasible in these patients and results in 5-year survival in about one-third of patients.

Clinically significant renal impairment is a common feature in multiple myeloma. In a study of 869 patients, 55% of patients had renal impairment at presentation (Kyle, 1975). A much higher percentage of patients are likely to experience renal impairment at some point during the course of their disease (Alexanian et al, 1990; Winearls, 1995) and 2–3% of patients present with renal failure that is severe enough to require dialysis (Alexanian et al, 1990; Clark et al, 1999). Many factors contribute to renal failure in myeloma patients. The kidneys are damaged directly by the light chain component of the monoclonal immunoglobulin causing proximal tubular damage and myeloma nephropathy, making patient with light chain only myeloma particularly susceptible. Hypercalcaemia, dehydration, infection, hyperuricaemia and the use of nephrotoxic drugs also contribute (Smith et al, 2006). Amyloid, light chain deposition disease and plasma cell infiltration are less frequent explanations for renal impairment.

In early trials of chemotherapy in myeloma, renal failure was found to be a predictor of poor prognosis. It is now considered that renal failure reflects a high tumour burden rather than an inherently poor prognosis and that the lower median survival seen in patients presenting with renal failure is caused by the administration of lower doses of chemotherapy and a higher treatment-related mortality (Blade et al, 1998).

High-dose chemotherapy (HDT) using melphalan 200 mg/m2 with autologous stem cell transplantation (SCT) improves the outcome of patients with multiple myeloma and has a low transplant-related mortality (TRM) of <5% (Attal et al, 1996; Barlogie et al, 1999; San Miguel et al, 2000; Child et al, 2003). TRM may be decreasing.

The largest series of patients with renal impairment treated with HDT and SCT was reported by Badros et al (2001). This initial report suggested more severe toxicity, some of it occurring in unusual organs. However, the TRM was 6% and not related to melphalan dose.

In systemic AL amyloidosis, the median survival is 1–2 years. Proteinuria, as the dominant feature, is associated with a better prognosis. Renal involvement is present in nearly one-third of newly diagnosed patients with systemic AL amyloidosis. Extracellular deposition in the kidneys of monoclonal light chain fragments forming insoluble fibrils leads to glomerular damage and proteinuria. The nephrotic syndrome is the most common renal manifestation. Mild renal dysfunction is a frequent finding but renal failure that requires dialysis is rare.

The use of high-dose melphalan therapy and peripheral blood SCT (PBSCT) in patients with AL amyloidosis was first reported in 1996, and a series of 25 patients was reported shortly afterwards by the same group (Comenzo et al, 1998). Encouraging results have since been reported in several series of patients in various centres (Sanchorawala et al, 2001; Gertz et al, 2002). HDT and PBSCT can result in reversal of the clinical manifestations of AL amyloidosis in up to approximately 60% of patients who survive the procedure. However, the efficacy of PBSCT in AL amyloidosis has not been investigated in any controlled comparative study, and procedure-related mortality has been consistently and substantially higher among patients with amyloid than those with multiple myeloma. The 100-day mortality in two experienced single-centre American studies was around 14% (Sanchorawala et al, 2001; Gertz et al, 2002) and was 39% in two multicentre European studies (Moreau et al, 1998; Gillmore et al, 1999). This reflects compromised function of multiple organ systems by amyloid, and, therefore, refinement of patient selection and improvement of peri-transplantation clinical management are priorities.

The TRM of PBSCT is associated with the number of organ systems involved with amyloid, based on the clinical evaluation of the kidneys, heart, liver and gastrointestinal (GI) system, and peripheral/autonomic nervous system. Of 43 patients transplanted in two single-centre studies (Comenzo et al, 1998; Gertz et al, 2002), patients with ≤2 organ systems involved had significantly superior 100-day survival (81%, 25 of 31) compared with those who had >2 organ systems involved (33%, four of 12; P < 0·01, Fisher's exact test). Similar conclusions have been reported in multicentre studies. The causes of death included cardiac arrhythmias, intractable hypotension, multiorgan failure and GI bleeding. Patients with poor renal function and those who are already dialysis-dependent fare very badly (Comenzo & Gertz, 2002).

There are many anecdotal single centre reports of the use of high-dose melphalan and PBSCT in myeloma and amyloidosis patients with severe renal impairment (both on and off dialysis). However, there are no larger scale national data that would help clarify the issues of morbidity and mortality to inform clinical practice. The objectives in this study were firstly to establish the numbers of patients in the UK treated in this way and to look at their outcomes. Secondly, we investigated whether there was a relationship between the dose of melphalan and toxicity.

This paper reports the results of a questionnaire-based study carried out in the UK, together with long-term follow-up data on survival and renal outcome

Patients and methods

Two questionnaires were distributed to the 41 transplant centres in the UK that perform autologous SCT. Data requested included patient characteristics, disease parameters at diagnosis, details of renal pathology, treatment prior to HDT/PBSCT, disease status at time of HDT/PBSCT, pretransplant organ function, source of stem cells and details of mobilisation, conditioning, timing of dialysis in relation to conditioning and after stem cell reinfusion (where applicable), cytokine support and engraftment details, resource usage and detailed non-haematological toxicity information, response details, long-term toxicity and quality of life. There were two questionnaires, one for patients already on dialysis and one for those with a glomerular filtration rate <20 ml/min.

Response criteria

Response to standard and high-dose chemotherapy (HDT) was assessed according to the criteria described by Blade et al (1998).


Comparisons of frequencies between groups were performed using the chi-squared test. Median values were compared using the Mann–Whitney U-test. Survival cures for overall survival, progression-free survival, and TRM were estimated by the Kaplan–Meier method and compared with the log rank test. Multivariate analysis was performed using Cox proportional regression. Variables were included in the multivariate analysis when they altered the hazard ratios for the relevant factor studied by 10%. A P-value of <0·05 was considered significant.


Completed questionnaires were returned by 14 centres regarding 28 HDT and autologous stem cell rescue procedures in 27 patients who were either receiving dialysis or who had a creatinine clearance of <20 ml/min. The majority were obtained from the patients already on dialysis (23/28 transplants) and only four of the 28 HDT procedures were performed for patients with AL amyloidosis. Patient characteristics are summarised in Table I. The median age of all patients at the time of HDT and SCT was 53·5 years (range: 39–65 years). Patients were diagnosed with myeloma or AL amyloidosis between February 1994 and October 2002 and underwent HDT and SCT between April 1995 and May 2003. The majority of patients had been on dialysis since soon after diagnosis but in seven patients dialysis was commenced during the course of treatment prior to HDT at a median time from diagnosis of 2 months (range: 1–44 months).

Table I.   Patient characteristics.
  1. HDT, high-dose chemotherapy; SCT, stem cell transplantation; CAPD, chronic ambulatory peritoneal dialysis; GFR, glomerular filtration rate.

 Median age (range)53·5 years (39–65)
Renal status prior to HDT and SCT
 On dialysis23
 GFR<20 ml/min, not on dialysis5
Type of dialysis
 Median time from diagnosis to HDT and SCT8·5 months (range: 2–58)

The median time from diagnosis to SCT in all patients was 8·5 months (range: 2–58 months) and was similar in the groups receiving (7 months) or not receiving (10 months) dialysis at the time of HDT. Eighteen patients were receiving haemodialysis and five were on chronic ambulatory peritoneal dialysis (CAPD).

Disease characteristics

The light chain involved in the myeloma or amyloid was kappa in 13 patients (one of these underwent two HDT procedures), lambda in 12 patients, one patient was non-secretory and the light chain was unknown in one patient. All myeloma patients had abnormal renal function at diagnosis – the majority were Durie-Salmon stage IIIB (two unknown); two patients with amyloidosis had renal impairment at diagnosis (one normal, one unknown). The median diagnostic creatinine concentration in all patients was 639 mmol/l.

Autologous stem cells

All but two of 27 patients received peripheral blood stem cells alone (92%). The median CD34+ cell dose was 2·35 × 106/kg (range: 0·83–14·32). Sixteen of 27 patients stem cells were collected with granulocyte colony-stimulating factor (G-CSF) alone but nine patients received cyclophosphamide (dose range: 1·25–2·8 g/m2) and G-CSF.

High-dose chemotherapy

Twenty-six of 27 patients received high-dose melphalan at a median dose of 140 mg/m2 (range: 60–200 mg/m2) (see Table I). Ten patients (37%) received 200 mg/m2. In addition, three patients received other chemotherapy in addition to melphalan and one patient received busulphan and cyclophosphamide (Table II). The median timing of the chemotherapy was 48 h prior to stem cell infusion.

Table II.   Conditioning therapy.
Melphalan dose (n = 24)
  1. BCNU, carmustine.

 60 mg/m21
 70 mg/m22
 100 mg/m21
 140 mg/m210
 200 mg/m210
Other chemotherapy (n = 4)
 Busulphan/cyclophosphamide 1


Twenty-six patients were evaluable for engraftment and the median time to engraftment was 12·5 d (range: 8–18 d). The majority of patients received G-CSF at a median of day +4 poststem cell infusion. Platelet engraftment occurred in 100% of evaluable patients at a median of 16·5 d (range: 7–45 d).

Post-transplant course and toxicity

Nineteen patients (70%) received i.v. antibiotics and there was one infective death. The median inpatient stay was 25 d (range: 9–74 d). World Health Organization toxicity occurring as a result of the procedure was moderate with only three instances of grades 3–4 toxicity occurring in 27 patients (two cardiac and one oral).

Seven out of 27 patients (26%) died before the day 100. Five patients died of transplant-related problems but three of those died of organ failure of a previously abnormal organ. Two of the 10 patients who received 200 mg/m2 melphalan died of transplant related complications. The overall TRM was 18·5%.

Disease responses and effect on renal function

Twenty of 27 patients had a response (70%); 18 of these were a complete (CR) or partial response (PR). The median time to disease progression was 32 months (range: 6–54 months). The median time to best response was 6·5 months. Four of 17 evaluable patients (24%) became dialysis-independent at a median of 5 months post-HDT/SCT. Thirteen patients remained on dialysis, but one patient required it less frequently. Of the five patients not on dialysis, two developed renal failure requiring dialysis and one died of renal failure before the day 100.Thirteen patients remained on dialysis; of these, the frequency of dialysis was decreased in one patient.

Survival and performance status

At a median follow-up of 70 months, 7/23 patients with myeloma were alive. Three of the seven patients had progressive disease. The median Karnofsky score of the survivors was 80% (range: 60–90%). Estimated median survival in the myeloma patients was 44 months (1 year 4 months to ∞, 95% CI) (Fig 1). The numbers of patients were insufficient to perform a multivariate analysis of the factors affecting the survival. Of the seven survivors, three receive dialysis and the four not having dialysis have creatinine measurements of 197, 262, 441 and 470 mmol/l.

Figure 1.

 Overall survival of myeloma patients (% surviving) versus time (years) with 95% CI.

Two of the four amyloidosis patients were alive at 51 and 89 months from transplant.


Several large national randomised studies provided strong evidence for the role of HDT and autologous SCT in the management of patients with myeloma, making it a standard of care (Attal et al, 1996; Child et al, 2003). This evidence is for 200 mg/m2 of melphalan; although it is likely that lower doses are also beneficial, this is unproven. However, this is essentially a palliative procedure; most patients relapse and there is only a 10-year survival of 20–30% (Sirohi et al, 2005). The role of this intervention in older patients and infirm patients (such as those receiving the renal replacement therapy) is less certain. Before conducting this national study, we hypothesised that, if HDT and autologous SCT could be safely delivered to myeloma patients with renal impairment, it might improve outcome and that a proportion of patients would experience improvement in their renal function.

Information concerning the feasibility, toxicity and outcome of autografting in this patient group is scant. Case reports and small single centre series have shown that stem cell collection can be carried out successfully in patients with renal failure, including those on dialysis, and time to engraftment is similar to those patients with normal renal function (Ballester et al, 1997; Rebibou et al, 1997; Reiter et al, 1999; Tosi et al, 2000). Some of these early reports suggested an increase in morbidity and mortality in patients receiving HDT on dialysis. Of particular concern were case reports of significant toxicity, especially neurotoxicity associated with melphalan use in renal impairment (Schuh et al, 1999).

There is one study, in which 27 of 38 patients received 200 mg/m2 and 11 patients 140 mg/m2 (Badros et al, 2001). Toxicities were greater in the 200 mg/m2 group, but TRM was remarkably low (5%). Renal failure did not affect the stem cell dose obtained or engraftment, and survival was not related to the dose of melphalan received. Only two of the 38 patients were able to cease dialysis.

Several other studies have since reported successful use of HDT with autologous stem cell support in patients with renal failure, even in patients on dialysis (summarised in Table III). However, non-haematological toxicity and treatment-related mortality are higher in these patients than in those with normal renal function. San Miguel et al (2000) reported a TRM of 29% in 14 patients with renal failure (four on dialysis) who received melphalan 200 mg/m2. More recently, a TRM of 16% was reported in 25 patients (eight of whom were on dialysis) who received melphalan given at a range of doses and treatment schedules. Twenty-two patients received melphalan 200 mg/m2 either as a single dose or divided over two consecutive days, with three deaths in each group (the remaining three received melphalan 140 mg/m2). A further series of 38 patients receiving melphalan 200 mg/m2 reported TRM of only 5% but none of these patients was on dialysis (Sirohi et al, 2001).

Table III.   Previously published data.
StudyTRMMelphalan dose (mg/m2)Patients taken off dialysis
  1. Overall survival (solid line) of myeloma patients (% surviving) versus time (years) with 95% CI (broken lines).

Badros et al (2001)5% of 38 patients140 or 200 (median 200)2/38
Knudsen et al (2005)17% of 29 patients (creatinine clearance <60 ml/min)140 or 200Not stated
San Miguel et al (2000)29% of 14 patients200Not stated
Present study18·5% of 27 patientsMedian 140 (200 in 10/27 patients)4/17 evaluable patients

However, single centre studies may reflect local referral patterns and do not always represent what can be achieved nationally. National multicentre data are perhaps of more use to the transplant physician and patient when deciding about HDT. In this study, nearly half the patients survived 3·5 years from transplant and two-thirds had a useful response (CR and PR). Five patients had improved renal function and >20% were weaned from dialysis.

Another major aim of this study was to evaluate the toxicity. The initial Badros study suggested severe mucositis and also significant pulmonary, cardiac and neurological toxicity (Badros et al, 2001). This led this group to reduce the dose of melphalan to 140 mg/m2. In our series, TRM occurred in 18·5%, but we were unable to demonstrate an effect of melphalan dose. The dose of melphalan associated with improved survival in the majority of published studies was 200 mg/m2. However, lower doses may be beneficial, particularly in patients with renal failure, as the pharmacokinetics of melphalan is altered. A few studies have addressed the issue of pharmacokinetics of melphalan in renal failure (Kergueris et al, 1994; Samuels & Bitran, 1995; Tricot et al, 1996). The main findings of these reports was that, although many parameters (especially the half-life and area under the curve) were correlated with creatinine clearance, renal insufficiency did not lead to a large decrease in melphalan clearance compared with inter-individual variations.

Ideally, a phase I study of doses of melphalan would be conducted to establish a dose of melphalan that was safe and effective. The timing of giving the melphalan and when to perform the dialysis was also a matter of controversy. In this study, melphalan was mainly given 48 h before the procedure and dialysis. Pharmacokinetic studies would help resolve this issue. The reasons each patient received their particular dose of melphalan are not ascertainable in a retrospective study but may relate to the uncertainty regarding the optimal doses and overall performance status.

The small number of patients with amyloidosis in this study makes conclusions about this group limited. However, HDT is also feasible in these patients and it may be worth exploring in a larger group of the selected patients.

This study may have potential selection biases in that the 14 (of 41) centres that participated in the study may have had more favourable results than all patients who received this therapy nationally. However, the study was thoroughly publicised through the UK myeloma forum and the Clinical Trials Committee and many centres contributed small numbers of patients with variable outcomes. Only a prospective study would eliminate this bias. Obtaining the historical data from notes was problematic and some details were not available.

Future studies should use the data from this and other studies to evaluate prospectively the role of HDT and autologous SCT in the management of myeloma patients with severe renal impairment. This study found that 140 mg/m2 was well tolerated and that higher doses could be explored in a phase I toxicity study. These studies should be part of national treatment protocols, such as myeloma IX, and larger studies may identify subgroups of patients who benefit most. Careful liaison with renal physicians regarding the individual antibiotic protocols is of paramount importance. The role of newer agents, such as bortezomib and lenalidomide, require evaluation in this patient group. However, for patients ineligible for entry onto studies or those declining involvement, we are able to report that HDT and autologous SCT is a feasible procedure that may produce a clinically significant benefit.


We wish to acknowledge the data managers from the patient centres that extracted data for this study.