Introduction

Systemic immunoglobulin light-chain (AL) amyloidosis is caused by abnormally produced monoclonal light chains that aggregate and deposit in various tissues, including the heart, kidneys, liver, gastrointestinal tract, peripheral nerves, skin, and other tissues [1, 2]. The outcome of AL amyloidosis varies among individuals and it depends on the dominant organ system involved. The median overall survival (OS) from the diagnosis of AL amyloidosis is reported to be approximately 3 years, but is markedly reduced to only 6–12 months in patients with advanced cardiac involvement [3]. Because of the low incidence and the diversity of clinical severity, a standard treatment for AL amyloidosis has not yet been established.

High-dose melphalan followed by autologous stem cell transplantation (HDM/ASCT) has been introduced as a treatment option for AL amyloidosis [4] and was reported to induce a higher response rate compared to conventional chemotherapy [5]. Dose manipulation of melphalan based on risk factors associated with early treatment-related mortality was reported to result in improved survival [6, 7].

We started HDM/ASCT for AL amyloidosis patients at our institution, the Japanese Red Cross Medical Center, in September 2006. The eligibility criteria have been changed in accord with the results of prior studies [6, 7] and our own experiences. In the present study, we retrospectively analyzed the cases of 40 patients with AL amyloidosis who received HDM/ASCT at our institution during the past 7 years. The patients’ responses after HDM/ASCT and the risk factors associated with overall survival are described, and new eligibility criteria are proposed.

Patients and methods

Patients

One hundred and twenty-nine patients were diagnosed as primary systemic AL amyloidosis at the Japanese Red Cross Medical Center between September 2006 and December 2013. Seventy-four of 129 patients were 65 years old or younger, and 39 of those (53 %) were considered eligible for HDM/ASCT. The diagnosis of AL amyloidosis was based on a histological examination from at least one organ. The existence of monoclonal protein in serum or urine also confirmed the diagnosis. Serum free light-chain (FLC) was measured before and after the HDM/ASCT in 15 of the 40 patients. Organ involvement was determined as defined by the consensus opinion from the 10th International Symposium on Amyloid and Amyloidosis published in 2012 [8]. Patients with age older than 65 years, those with poor performance status (ECOG 3 or 4), and those with severe organ dysfunction (e.g. NYHA III or IV, creatinine clearance <50 mL/min, or total bilirubin ≥3.0 mg/dL) were ineligible for HDM/ASCT. Until 2007, there were no exclusion criteria by plasma BNP and/or interventricular septal wall thickness (IVST) by echocardiogram. Patients with plasma BNP >1000 pg/mL and/or >IVST 15 mm were determined transplant ineligible thereafter because two patients with severe cardiac dysfunction died within 50 days after transplant in 2006 and 2007. One patients aged 70 years with a good performance status was considered eligible for HDM/ASCT. The cases of the 40 patients with AL amyloidosis treated with HDM/ASCT were retrospectively analyzed.

Treatment

The patients’ induction treatments before the HDM/ASCT varied and they included oral melphalan based (23 cases) and vincristine and dexamethasone based (four cases). Two patients received bortezomib-based and one received lenalidomide-based treatment before they were referred to our institution. Ten patients received no induction treatment before HDM/ASCT. Peripheral blood stem cells (PBSCs) were collected by apheresis procedures after mobilization with the use of granulocyte colony-stimulating factor (G-CSF), either alone or with preceding etoposide or cyclophosphamide. The minimal number of CD34-positive cells required for HDM/ASCT was 1.0 × 106 per kg of body weight. The dosage of intravenous melphalan for the conditioning for ASCT was adjusted to reduce the regimen-related toxicity according to previous reports [9, 10]. Briefly, the dosage of melphalan was reduced to 100 or 140 mg/m2 from standard 200 mg/m2 depending on the numbers of risk factors applied, including age (61–71 years), cardiac involvement, and reduced kidney function. During first 5 years, the dosage of melphalan was further reduced by the physicians’ decision.

Response evaluation

Hematologic and organ responses were evaluated according to the consensus guidelines published in 2012 [11]. Due to clinical limitations, an immunofixation (IF) assay and the measurement of FLC before HDM/ASCT were available only for patients evaluated after 2011. Immunoelectrophoresis was used instead of IF and/or FLC assay. A hematological complete response was defined as the disappearance of monoclonal protein in serum and/or urine and the normalization of FLC. Renal organ response was defined as a 50 % decrease in the patient’s 24-h urine protein (urine protein must be >0.5 g/day pretreatment). Serum creatinine and creatinine clearance must not worsen by 25 % over baseline.

Because N-terminal prohormone of brain natriuretic peptide (NT-proBNP) was measured in a limited number of patients in this series and because it is known that brain natriuretic peptide (BNP) is less affected by renal function [12], we used BNP for the evaluation of the cardiac organ response. We originally defined cardiac organ response as a 30 % decrease in BNP (indicated only for patients with BNP ≥100 pg/mL pretreatment).

Statistics

This was a retrospective cohort study. Overall survival (OS) was defined as the length of time between day 0 of the ASCT and death or the last follow-up, and was estimated using the Kaplan–Meier method. Cox proportional hazards univariate and multivariate regression models were applied to identify significant risk factors for OS. All statistical analyses were performed with Stata software ver. 13 (College Station, TX, USA).

Results

Patient characteristics

A total of 40 patients were identified who received HDM/ASCT between 1 September 2006 and 31 December 2013. Eighteen were men and 22 were women. The median patient age was 54 years (range 32–70 years). The median time from diagnosis to HDM/ASCT was 3 months (range 1–14 months). The patients’ characteristics are shown in Table 1. Renal involvement was found most commonly, in 27 patients (68 %); followed by gastrointestinal involvement in 18 patients (45 %), cardiac involvement in 17 patients (43 %), and then hepatic involvement in four patients (10 %). Two or more organs were affected in 26 patients (65 %). Six of the 40 patients (15 %) received melphalan 200 mg/m2 and 20 patients (50 %) received melphalan 140 mg/m2. The patients with severe cardiac dysfunction (35 %) received <140 mg/m2 of melphalan per their physician’s decision.

Table 1 Patient characteristics and transplant procedure
Full size table

The 36 patients who survived more than 50 days received maintenance therapy starting 2–4 months after HDM/ASCT. The maintenance therapies consisted mostly of oral melphalan plus dexamethasone (Mel/Dex) and were repeated every 2–6 months until disease progression or an adverse event occurred. Mel/Dex was continued up to 6 years and median cycles performed were 13 (range 4–25). Two patients received bortezomib-based treatment as subjects in a clinical trial. Two other patients were diagnosed with multiple myeloma (MM) in their clinical course after HDM/ASCT and received bortezomib-based treatment. Disease progression occurred in six patients. Two patients died with progressive cardiac failure and five patients progressed to renal failure requiring hemodialysis. Twenty-five of 40 patients (62.5 %) are alive without disease progression. No patient developed treatment-related myelodysplastic syndrome (t-MDS) during the observation period.

Hematologic and organ responses

The patients’ hematologic and organ responses are shown in Table 2. The hematologic response was evaluated in only 33 patients because of four patients’ early death (within 100 days after HDM/ASCT) and a lack of data for three other patients. Hematologic complete responses were achieved by 17 of those 33 patients (52 %). The cardiac organ response was evaluated in 11 patients (BNP before HDM/ASCT ≥100 pg/mL and alive for 12 months); seven patients responded (64 %). The renal organ response was evaluated in 18 patients (24-h urine protein ≥0.5 g and alive for 12 months); nine patients responded (50 %). The overall organ response was obtained in 16 of the 32 evaluable patients (50 %). An increase of serum albumin was observed in 13 of the 18 patients (72 %) whose pre-transplant albumin level was below 3 g/dL (data not shown).

Table 2 Hematologic and organ responses
Full size table

Survival

The database was closed for follow-up on 1 December 2014. Thirty of the 40 patients were alive at that time with a median follow-up of 42 months (range 12–94). The 2- and 4-year estimated overall survival rates were 85 % [95 % confidence interval (CI) 69–93 %] and 74 % (95 % CI 56–86 %), respectively (Fig. 1). The overall survival data according to representative risk factors are shown in Fig. 2.

Fig. 1
figure 1

The overall survival (OS) of 40 patients who received HDM/ASCT for systemic immunoglobulin light-chain (AL) amyloidosis, as Kaplan–Meier curves. The estimated 2- and 4- years OS rates were 85 % (95 % CI 69–93 %) and 74 % (95 % CI 56–86 %). The median follow-up period of the survivors was 42 months (range 12–94 months)

Full size image
Fig. 2
figure 2

OS according to risk factors. Kaplan–Meier curves for OS according to the existence of cardiac involvement (a), plasma brain natriuretic peptide (BNP) level (b), interventricular septal thickness (IVST) by echocardiography (c), and serum creatinine level (d). The existence of cardiac involvement, elevated plasma BNP, and elevated serum creatinine each had significant adverse effects on OS by log-rank test (p < 0.05)

Full size image

The existence of cardiac involvement, the pre-transplant plasma BNP level, and the pre-transplant serum creatinine level each had statistically significant adverse effects on overall survival by log-rank test (p < 0.05). The factors that were revealed to be associated with overall survival are shown in Table 3. Cardiac involvement and elevated plasma BNP level (≥600 pg/mL) were detected as significant unfavorable prognostic factors by Cox proportional hazards univariate regression models. The multivariate analysis failed to show any significant unfavorable factor.

Table 3 Univariate and multivariate analyses for overall survival
Full size table

Cause of death

Four patients died within 100 days after HDM/ASCT. Three patients died of heart failure; two of them had severe cardiac involvement before HDM/ASCT (BNP ≥1000 pg/mL and LVEF <50 %). One patient died of septic shock during leukocytopenia 8 days after HDM/ASCT. Thirty-six patients received oral melphalan-based maintenance therapy. One patient was diagnosed with multiple myeloma 2.5 years after the HDM/ASCT and died of the disease. Other causes of death were lung cancer and an unexpected accident.

Discussion

The results of our present analysis support the feasibility and efficacy of HDM/ASCT for systemic AL amyloidosis. Reports on HDM/ASCT for AL amyloidosis are limited and, to our knowledge, this is the largest study of HDM/ASCT for AL amyloidosis from Japan.

The eligibility of patients undergoing HDM/ASCT was discussed by Dispenzieri [6] and Wechalekar [7], and it should be carefully considered in light of patients’ performance status, cardiac function, renal function, liver function, and the severity of autonomic neuropathy. Dose modification of melphalan was also recommended by Comenzo et al. considering the patients’ age, cardiac function, renal function, and number of involved organs [9]. However, it was reported that the dose manipulation of melphalan can be associated with a lower response rate [13]. Hayashi et al. reported ten cases from a Japanese single institution [14]; they manipulated the dose of melphalan according to the patients’ pre-transplant plasma BNP level (cut off 200 pg/mL), resulting in no treatment-related mortality.

In the present series, in order to assure safety, the HDM/ASCT was started with a lower dose of melphalan, and 14 of 40 patients (35 %) received <140 mg/m2. Seven patients received <100 mg/m2 that can be too small doses considered as intermediate dose. We now consider melphalan 200 mg/m2 as standard and attenuate the dose for patients older than 65 years, those with reduced renal function, those with reduced cardiac function, and those with multiple organ involvement. In 2006 and 2007, two of the four patients died with cardiac failure; the pre-transplant plasma BNP levels were more than 1000 pg/mL in both cases. The doses of melphalan given for those two cases were 52 and 76 mg/m2, respectively, and we concluded that ASCT should not be performed for patients with severe cardiac dysfunction even with reduced dose of melphalan. The eligibility criteria have been refined, and the most recent eligibility criteria are shown in Table 4.

Table 4 Proposed eligibility criteria of ASCT for systemic AL amyloidosis
Full size table

The median age at transplant in the present patient series was 54 years old, and it is known that age at the diagnosis of AL amyloidosis is generally younger than the age at diagnosis of MM [15]. It is likely that there are patients who will not survive if early HDM/ASCT is performed, and the decision regarding whether HDM/ASCT should be done must be carefully considered. In fact, the two patients in this study whose plasma BNP levels were between 500 and 600 pg/mL have survived for more than 2 years as of this writing. Moreover, a ≥30 % decrease in plasma BNP was observed in six of the 11 patients (63 %) whose pre-transplant plasma BNP levels were ≥100 pg/mL (Table 2).

Whereas cardiac involvement critically affected the patients’ survival, renal involvement had a modest impact on survival. About one-half of the patients with Cre ≥1.0 mg/dL survived more than 2 years, and early HDM/ASCT is recommended to avoid the progression of renal insufficiency and the introduction of hemodialysis. In fact, five patients progressed to renal failure requiring hemodialysis and one of them died 35 months after HDM/ASCT. It is noteworthy that an improvement in serum albumin was observed in 13 of the 18 patients (72 %) with a pre-transplant albumin level <3 g/dL.

Because the severity of organ dysfunction varies among patients with AL amyloidosis, it is difficult to compare the superiority between HDM/ASCT and other treatments among randomized prospective clinical trials. There is only one randomized controlled trial comparing HDM/ASCT with standard-dose melphalan plus high-dose dexamethasone [16]; the study failed to show the superiority of HDM/ASCT over standard therapy, probably due to patient selection, but HDM/ASCT is still a standard of care for patients with AL amyloidosis when the patients are selected carefully [2].

Although the uses of bortezomib, thalidomide, and lenalidomide for AL amyloidosis have been reported, the roles of these agents on HDM/ASCT are controversial. Scott et al. reported that the use of bortezomib as the induction treatment before HDM/ASCT resulted in a high response rate and excellent OS without increasing the treatment-related mortality [17]. They also suggested that there might be a role for induction therapy to optimize organ function and increase the eligibility of patients. Although the usefulness of bortezomib has been reported, it may be associated with toxicities such as autonomic neuropathy and dose modification is sometimes required [18].

The uses of thalidomide and bortezomib as the consolidation therapy after HDM/ASCT were also reported [19, 20]. Landau et al. reported bortezomib and dexamethasone consolidation therapy following HDM/ASCT, and they noted that an improvement of response was observed in 86 % of the patients who received such consolidation therapy. The hematological response rate was 79 % (≥partial response) including a 55 % complete response rate, and organ response was achieved in 55 % of the patients. The feasibility and the efficacy of bortezomib either before or after HDM/ASCT should be tested in randomized controlled trials.

In conclusion, we retrospectively analyzed the cases of 40 patients with AL amyloidosis who received HDM/ASCT at a single institution in Japan. With careful patient selection and experiences, treatment-related mortality can be reduced among such patients. The optimization of induction treatment and post-transplant consolidation treatment are also warranted.