Fax: (507) 266-4972
Ten-year survival after autologous stem cell transplantation for immunoglobulin light chain amyloidosis
Article first published online: 15 JUN 2012
Copyright © 2012 American Cancer Society
Volume 118, Issue 24, pages 6105–6109, 15 December 2012
How to Cite
Cordes, S., Dispenzieri, A., Lacy, M. Q., Hayman, S. R., Buadi, F. K., Dingli, D., Kumar, S. K., Hogan, W. J. and Gertz, M. A. (2012), Ten-year survival after autologous stem cell transplantation for immunoglobulin light chain amyloidosis. Cancer, 118: 6105–6109. doi: 10.1002/cncr.27660
- Issue published online: 3 DEC 2012
- Article first published online: 15 JUN 2012
- Manuscript Accepted: 9 APR 2012
- Manuscript Revised: 8 APR 2012
- Manuscript Received: 5 MAR 2012
- amyloidosis therapy;
- amyloidosis complications;
- immunoglobulin light chain;
- stem cell transplantation;
- therapeutic use of alkylating agents;
The current study was conducted to determine characteristics distinguishing the 10-year survivor group in patients with systemic immunoglobulin light chain (AL) amyloidosis who underwent autologous stem cell transplantation (SCT).
The study group included all 74 patients with AL amyloidosis who underwent high-dose melphalan treatment supported by autologous SCT since the beginning of the Mayo Clinic's SCT program until prior to August 2001.
A total of 32 patients (43%) patients survived for > 10 years. Statistically significant baseline differences in the 10-year survivor group included: 1) the number of organs involved; 2) septal thickness; 3) total cholesterol; and 4) urine total protein. The number of organs involved was the only predictor found on multivariable analysis. Depth of the response to therapy, as measured by the lowest posttransplantation serum free light chain level, was found to be the most significant indicator of durability of response.
Autologous SCT can offer durable benefit for patients with AL amyloidosis. The number of organs involved offers the greatest pretreatment prognostic value, whereas the lowest posttransplantation serum free light chain level offers the best posttreatment prognostic value. Cancer 2012. © 2012 American Cancer Society.
Systemic immunoglobulin light chain amyloidosis (AL) is a plasma cell dyscrasia in which monoclonal light chain protein aggregates and deposits in tissue as amyloid fibrils.1-3 Only a relatively small percentage of Ig light chains are amyloidogenic, as evidenced by the finding that AL amyloidosis occurs in only approximately 6% to 15% of patients with multiple myeloma. Amyloidogenicity is related to structural features such as the light chain isotype and the variable subgroup. The folding pathway of such amyloidogenic light chains is believed to pass through partially folded metastable conformations that aggregate either 1) at sufficiently high concentrations or 2) under specific environmental conditions.4-6
The treatment of amyloidosis has been targeted toward reducing the concentration of the amyloidogenic light chain and has paralleled that used for multiple myeloma. One of the rationales for the efficacy of such treatment is the remarkable observation that it is the growth of fibrils, in contrast to existing fibril load, that inflict cellular damage through transient defects in the cell membrane.7-10 Patients with amyloidosis differ from patients with other hematological malignancies in that they have impaired organ function with relatively normal bone marrow, which tends to be the reverse of what is found in patients with other hematological malignancies. This has implications in terms of the therapeutic window of many interventions. In what to our knowledge was the first effective chemotherapy regimen for amyloidosis, melphalan was used in conjunction with prednisone.11, 12 Since 1994, high-dose intravenous melphalan and autologous stem cell transplantation (SCT) have joined the armamentarium in the treatment of patients with amyloidosis.13-15 The role of autologous SCT in the management of patients with AL amyloidosis remains controversial in part because of the fragility of the patient population. The objective of the current study was to highlight 1) the validity of the transplantation selection criteria16 and 2) the durability and long-term results of treatment.
MATERIALS AND METHODS
All patients with AL amyloidosis who underwent autologous SCT at the Mayo Clinic over the 5-year interval from the beginning of our transplantation program in July 1996 until July 2001 were included in this study (74 patients). Pertinent clinical and laboratory data and information concerning autologous SCT and follow-up were extracted from the patient records. All patients had authorized the use of their medical records for research in accordance with Minnesota privacy statutes. The study was approved by the Institutional Review Board at the Mayo Foundation in accordance with the Declaration of Helsinki and the Health Insurance Portability and Accountability Act guidelines.
All patients underwent baseline evaluation, which included bone marrow biopsy, serum and urine electrophoresis and immunofixation, and serum light chain measurement. Amyloidosis was identified in biopsy specimens by positive Congo red staining. The subtyping of amyloidosis into AL amyloidosis was done by immunofluorescence staining of κ and λ light chains, immunoperoxidase staining for serum amyloid A, and genetic screening for transthyretin (TTR) mutations.
Eligibility criteria for autologous SCT were: 1) “physiological” age ≤ 70 years; 2) Eastern Cooperative Oncology Group performance score of ≤ 2; 3) creatinine clearance of ≥ 30 mL/minute (unless patient was undergoing chronic dialysis); and 4) New York Heart Association classes I or II (unless patient was undergoing chronic dialysis).16, 17 Before 2001,cardiac biomarkers were not routinely available as part of the eligibility criteria, but were performed on frozen sera when available.
Stem cell mobilization was accomplished with sargramostim (granulocyte-macrophage–colony-stimulating factor) or filgrastim (granulocyte–colony-stimulating factor). Apheresis was performed until a minimum of 2 × 106 CD34+ cells/kg were collected. All patients underwent conditioning with high-dose melphalan before the autologous SCT. Neutrophil and platelet engraftment was defined as the achievement of 500 neutrophils/μL for 3 consecutive days and 50,000 platelets/μL without platelet transfusion for 3 days.
Hematological response to therapy was assessed by criteria from the consensus opinion of the 10th International Symposium on Amyloid and Amyloidosis, held in Tours, France in 2004.18 Complete hematological response requires an absence of monoclonal protein by immunofluorescence and a normal free light chain ratio. Overall survival was calculated from the day of bone marrow transplantation.
Statistical analyses were performed using the R (version 2.13.0; R Foundation, Vienna, Austria) and S-Plus (version 6.2; Insightful Corporation, Seattle, WA) statistical software packages. The characteristics of the long-term survivor (LTS) and non–long-term survivor (NLTS) groups were compared using the Wilcoxon rank sum test. The median values of all recorded numeric and ordinal baseline and posttherapy characteristics of the population are shown in Tables 1 and 2 alongside the corresponding P values from the Wilcoxon rank sum test. All probabilities reported are 2-tailed; P values < .05 were considered to be statistically significant.
|Alkaline phosphatase, U/L||87.2||93.5||.429|
|Creatinine clearance, mL/min||72||59.5||.306|
|Serum M-protein spike, g/dL||0.3||0.1||.48|
|Size of urine M-protein spike, g/dL||0.236||0.113||.159|
|Urine total protein, g/d||6.24||3.07||.0343|
|Urine M-protein, %||0.06||0.0918||.345|
|Plasma cells, %||10.5||7||.608|
|Estimated bone marrow plasma cells, %||6.5||5||.965|
|Pretransplantation free light chain, mg/dL||7.54||16.8||.418|
|Pretransplantation uninvolved FLC, mg/dL||0.6||0.55||.743|
|Labeling index, %||0||0||.68|
|Autologous stem cell transplantation|
|Time from diagnosis to transplantation, mo||4.87||6.78||.0771|
|Time from diagnosis to M-protein recognition, mo||-0.0164||-0.0987||.781|
|CD34 collected, ×106 cells/kg||7.21||6.19||.549|
|Mononuclear cells collected, ×106 cells/kg||7.49||5.8||.105|
|No. of apheresis collections||2||3||.156|
|Ejection fraction, %||66||65||.558|
|Septal thickness, mm||10||13||.0122|
|NT pro-BNP, pg/mL||215||969||.0781|
|Liver edge RCM, cm||0||0||.615|
|No. of organs involved||1||2||.00611|
|Time to ANC of 500 cells/μL, d||13||13.5||.338|
|Time to 20,000 platelets/μL, d||13.5||14||.241|
|Time to 50,000 platelets/μL, d||20||20.5||.541|
|Lowest posttransplantation FLC, mg/dL||1.51||4.22||.0085|
|Autologous stem cell transplantation|
|Time to response, mo||6.45||3.55||.129|
Survival data were fit to the Cox proportional hazards model.19, 20 Parameter estimates were obtained in the usual fashion by maximizing the partial likelihood. Estimates of the 95% confidence interval variance were computed via the Tsiatis method.21 In assessing multivariate models, all combinations of parameters were fit and the best model was chosen on the basis of 1) optimizing the partial likelihood across models with the same number of parameters and 2) accepting a model with a larger number of parameters only if the P value for the likelihood ratio test was statistically significant for the difference in degrees of freedom between the 2 models.
Analysis of variance (ANOVA) was used to compare the means of survival times for several factors, such as the melphalan conditioning regimens. The F-statistic was computed in the usual fashion and factors for which the F distributions were computed to be >0.95 were further analyzed for all cluster pairs via the Tukey-Kramer multiple comparisons of means method.22
A total of 74 patients with AL amyloidosis who underwent autologous SCT were included in the study. Of these, 41 patients (55%) were men and 33 patients (45%) were women. Patients were retrospectively divided into 2 groups: those surviving >10 years, termed the long-term survivor (LTS) group (32 patients), and those surviving <10 years, termed the non–long-term survivor (NLTS) group (42 patients). The median ages of the patients in the LTS and NLTS groups were 52.8 years (range, 37.4-67.4 years) and 55.2 years (range, 31.0-69.9 years), respectively.
Table 1 demonstrates the baseline characteristics of the study population. Four baseline characteristics were determined to be statistically significantly different between the LTS and NLTS groups: 1) the number of organs involved; 2) interventricular septal thickness; 3) total cholesterol; and 4) urine total protein.
Table 2 demonstrates the posttherapy characteristics of the study population. Only 1 characteristic was determined to be statistically significantly different between the LTS and NLTS groups: the lowest post-SCT serum free light chain level. Although the median best hematological response was different for the LTS and NLTS groups, this is related to the lowest post-SCT serum free light chain level and thus was not independent.
To assess the importance of the 4 characteristics on survival, we fit the survival data to Cox proportional hazards models. On the basis of the criteria described earlier, the best fit to the survival data was found to be a Cox proportional hazards model with the number of organs involved as the single covariate (predictor). Figure 1 depicts the Kaplan-Meier plot.
We computed the hazards ratios in the multivariate Cox proportional hazards model for all 4 characteristics. These are shown in Table 3. In the overfit model, only the number of organs involved was found to rise to the level of statistical significance.
|Characteristic||Difference in Median||HR||95% CI|
|Urine total protein, g/d||3.17||0.897||0.657-1.23|
|Septal thickness, mm||-3||0.866||0.627-1.19|
|No. of organs involved||-1||0.532||0.344-0.824|
To assess the importance of the posttherapy characteristics on survival, we fit survival data (censored for therapy-related mortality) to a Cox proportional hazards model. We computed the hazards ratio and tabulated it in Table 4.
|Characteristic||Difference in Median||HR||95% CI|
|Lowest posttransplantation FLC, mg/dL||-2.79||0.899||0.836-0.968|
The frailty of the patients with AL amyloidosis necessitates the careful selection of patients who are eligible for high-dose melphalan treatment supported by autologous SCT (HDM/SCT) to minimize treatment-related mortality. The results of the current study demonstrate that response can be durable for patients with AL amyloidosis who undergo HDM/SCT. Of the 74 patients who underwent HDM/SCT, 32 (43%) survived > 10 years.
The kidneys were the most commonly involved organs in patients with AL amyloidosis who underwent SCT2, 23; renal involvement was present in 24 (75%) patients in the LTS group and 26 (62%) patients in the NLTS group (62%). The median serum cholesterol levels of the patients in the LTS group (329 mg/dL) were higher than those of patients in the NLTS group (213 mg/dL), as were the total urine protein levels (6.24 g/dL for the LTS group vs 3.07 g/dL for the NLTS group). The creatinine clearance of patients in the LTS and NLTS groups were 72 mL/minute and 60 mL/minute, respectively, a difference that was not found to be statistically significant.
Cardiac involvement was the second most common organ involvement in patients with AL amyloidosis,2, 24 and was present in 9 (28%) patients in the LTS group and 11 (26%) patients in the NLTS group. The extent of cardiac infiltration as measured by septal thickness was lower in the LTS group (median, 10 mm) compared with the NLTS group (median, 13 mm), although the ejection fractions were preserved between the 2 groups and were found to be very similar. Systolic function was found to be preserved in patients with AL amyloidosis, at least in the early stages of the disease. The median number of organs involved in the LTS and NLTS groups were 1 (range, 1-3 organs) and 2 (range, 1-3 organs), organs, respectively.
In the current study, we identified 4 baseline characteristics that demonstrated statistically significant differences between the LTS and NLTS groups (ordered by statistical significance): 1) number of organs involved; 2) septal thickness; 3) total cholesterol; and 4) urine total protein. The number of organ systems involved is the key prognostic characteristic; the median number of organ systems involved in the LTS group was 1, whereas that in the NLTS group was 2. Cardiac involvement, as assessed by septal thickness, possesses significant prognostic value that is however subordinate to the number of organ systems involved. The counterintuitive prognostic value of renal involvement, as assessed by urine total protein or serum cholesterol, may be rationalized on this basis: The median cholesterol levels between the LTS and NLTS groups were well beyond levels that could be explained through differences in nutritional status, a factor that has been shown to impact prognosis in patients with AL amyloidosis.25 In the LTS group, patients manifesting any organ involvement were more likely to manifest renal rather than cardiac involvement. Equivalently, the higher incidence of proteinuria noted in the LTS group appears to be more than mitigated by their superior cardiac function. It should be noted that 3 of the patients in the current study underwent kidney transplantation; 2 of these patients ended up as members of the LTS group, whereas 1 patient ended up in the NLTS groups and was the recipient of both a cardiac and a renal transplantation.
Finally, ANOVA analysis of the survival times versus the dose of melphalan conditioning therapy did not find any statistically significant differences (Prob(> F) = 0.354), an effect that has been studied previously.26
The depth of the response to therapy, as measured by the lowest posttransplantation serum free light chain level, is the most significant leading indicator of the durability of treatment benefits.27 Unfortunately, because this cannot be assessed before therapy, it has no predictive value in terms of selecting patients. It should be noted that N-terminal pro-b-type natriuretic peptide (NT pro-BNP) was available in only 34 of the 74 patients (46%) and therefore could not be systematically assessed for impact in this model, although the medians for the LTS and NLTS groups appear to be quite different.
The impact of serum and urine M-proteins on survival was examined, but no statistically significant differences were found.
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURE
The authors made no disclosures.
- 16Validation of the criteria of response to treatment in AL amyloidosis. Blood. 2010; 116: 586-587., , , et al.
- 18Definition of organ involvement and treatment response in immunoglobulin light chain amyloidosis (AL): a consensus opinion from the 10th International Symposium on Amyloid and Amyloidosis. Tours, France. April 18-22, 2004, Am J. Hematol. 2005;79:319-328., , , , , , , , , , , , .