|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Dialysis |




* Department of Nephrology, AP-HP, Hôpital Necker, Université Paris V, Paris, France;
Department of Nephrology, AP-HP, Hôpital Kremlin Bicêtre, Université Paris XI, Kremlin Bicêtre, France;
Department of Nephrology, AP-HP, Hôpital Pitié Salpêtrière, Université Paris VI, Paris, France;
Department of Nephrology, AP-HP, Hôpital Saint-Louis, Université Paris VII, Paris, France; || AURA Peupliers, Paris, France; ¶ AURA Centre Henri Küntziger, Paris, France; ** Department of Biostatistics, AP-HP, Hôpital Necker, Université Paris V, Paris, France
Correspondence: Dr. Fadi Fakhouri, Department of Nephrology, AP-HP, Hôpital Necker, Université Paris Descartes, Paris, France. Phone: 0033144495458; Fax: 0033144495450; E-mail: fadi.fakhouri{at}nck.aphp.fr
| Abstract |
|---|
|
|
|---|
Design, setting, participants, & measurements: We retrospectively studied patients with AA or AL amyloidosis who started dialysis in five French centers between January 1, 1995 and December 31, 2005.
Results: We identified 19 patients with AL and 20 patients with AA amyloidosis undergoing dialysis. Patients with AL amyloidosis had shorter time from diagnosis to dialysis (25.2 versus 69.3 mo, P < 0.05) and more extrarenal amyloidosis, especially cardiac (63.2 versus 5%, P < 0.0001). Mean duration of follow-up was 37.4 and 31.8 mo for patients with AL and AA amyloidosis, respectively. Fifteen patients (78.9%) with AL and three patients (15%) with AA amyloidosis died on dialysis. Median survival was shorter in patients with AL (26 mo) than AA amyloidosis [not definable (ND)] (P < 0.02). Sepsis and cardiac deaths were the main causes of mortality. Prognosis factors for death at 1 yr were AL type (P < 0.01), cardiac amyloidosis [odds ratio (OR) = 18, P < 0.01], heart failure (OR = 8, P < 0.04), and shorter time from diagnosis to dialysis (6.1 versus 56 mo, P < 0.03). Multivariate analysis indicated that AL type (P = 0.02), but not cardiac amyloidosis was independently associated with global mortality.
Conclusions: Survival of patients with amyloidosis undergoing dialysis, especially AL type, is probably better than previously reported. However, mortality is higher in AL than AA type, especially in the setting of cardiac involvement.
| Introduction |
|---|
|
|
|---|
In AL amyloidosis, the amyloidogenic protein is an immunogloblin light chain or a fragment of an Ig light chain, more frequently of the lambda isotype, synthesized by clonal plasma cells in bone marrow. Thus, AL amyloidosis is related to a lymphoproliferative disorder and can be associated with a patent hematologic malignancy, mainly multiple myeloma (MM). AA amyloidosis is due to the deposition of serum amyloid A protein in patients with longstanding inflammatory disorders such as chronic infections or rheumatologic diseases. The aim of treatment is to halt the synthesis of the amyloidogenic protein by suppressing the underlying lymphoproliferative disorder in AL amyloidosis or by controlling the inflammatory process in AA amyloidosis.
The kidney is often affected by amyloid deposits. Patients typically present with nephrotic syndrome and progressive renal failure, leading to ESRD (1). However, very few data are available regarding patients with systemic amyloidosis undergoing chronic dialysis. Poor outcome for patients with AL amyloidosis on dialysis was reported in a few published studies in the early 1990s (2–4). However, major advances made in the management of AL amyloidosis have improved the prognosis of these patients.
More data are available regarding patients with AA amyloidosis undergoing dialysis, but published series have provided conflicting results (3–8).
The aim of this study is to assess the current presentation and outcome of dialyzed patients with AL or AA amyloidosis, and to determine prognostic factors.
| Concise Methods |
|---|
|
|
|---|
Amyloidosis was histologically proven in all patients using Congo red staining, which was birefringent under polarized light. For typing amyloidosis, immunofluorescence was subsequently performed on frozen sections, with antibodies directed to kappa or lambda light chains or protein A. Diagnosis of organ involvement by amyloid deposits was made based on the criteria defined at the 10th International Symposium on Amyloid and Amyloidosis (9). None of our patients underwent endomyocardial biopsy, but echocardiography was performed in all patients. The heart was considered involved in the presence of typical echocardiographic features in a patient with a positive result of noncardiac biopsy.
Patient's charts were reviewed to collect data such as amyloidosis type, underlying disease, time from diagnosis of amyloidosis to dialysis initiation, serum troponin levels, type of monoclonal protein in urine and plasma, evolution of monoclonal protein level over time, chemotherapies administrated, echocardiographical findings, extrarenal amyloidosis, complications, duration, and reason for stopping follow-up on dialysis (lost to follow-up, renal transplantation, or death in the absence of renal function recovery).
We also recorded whether the underlying lymphoproliferative or inflammatory disorder was controlled when dialysis was initiated, and for how long during the dialysis period. In patients with MM, the stage of the disease was classified according to the Salmon and Durie classification (10), allowing us to differentiate patients with asymptomatic MM (stage I) from those with overt MM, who presented with marked anemia, hypercalcemia, bone lesions, and/or high monoclonal protein rate. To precisely determine the level of disease control, the results of repeated measurement of C reactive protein and erythrocyte sedimentation rate were analyzed in patients with AA amyloidosis. In patients with AL amyloidosis, the results of repeated urinary and serum electrophoresis were studied to evaluate the evolution of the amount of monoclonal protein over time. For all patients, these biological parameters for quantification of inflammation or monoclonal protein were measured at least once every three months, but monthly in almost all patients. Disease control was defined by the absence of clinical and biological signs of activity of the inflammatory disease in patients with AA amyloidosis, or by the absence (complete response) or decrease of at least 50% of the monoclonal protein (partial response) previously detected in plasma and/or in urine patients with AL amyloidosis, in accordance with the criteria defined at the 10th International Symposium on Amyloid and Amyloidosis (9).
Statistical Analyses Quantitative parameters are presented as median or mean ± SEM. Qualitative parameters are given as number and percentages. Continuous variables were compared using unpaired t test or Mann–Whitney U test (two groups) if the assumption tests of SD equality and Gaussian distribution were not passed. Univariate analysis of survival was made by using a log rank test for qualitative parameters or the Cox model for quantitative parameters. The independent role of factors significantly associated with survival in univariate analysis was assessed using the Cox model for multivariate analysis. Categorical variables were compared using the Fisher's exact test. P < 0.05 was considered statistically significant.
Disclosure/Conflict of Interest.
None.
| Results |
|---|
|
|
|---|
|
|
|
|
Survival ranged from 29 to 113 mo in patients with AA amyloidosis, and from 1 to 96 mo in patients with AL amyloidosis. Median of survival in patients with AL amyloidosis was 26 mo, whereas it was undefinable in those with AA amyloidosis, because only three patients (15%) died during follow-up. Log rank test demonstrated a significantly reduced survival in patients with AL compared with AA amyloidosis (P < 0.02) (Figure 1A).
|
Five patients with AL amyloidosis and three patients with AA amyloidosis manifested clinical symptoms of global heart failure, consisting of important peripheral edema and exercise dyspnea. All five AL patients had echocardiographic features of cardiac amyloidosis, whereas heart failure was related to cardiac amyloidosis, cardiac sarcoidosis, and hypertensive cardiopathy in the three AA patients, respectively. Cardiac arrhythmias occurred in six patients with AL (26.3%) and in two patients with AA amyloidosis (10%) (NS). Serum troponin levels during the period of dialysis were not available in all patients, but tended to be more often elevated in patients with AL than AA amyloidosis: three patients with AL amyloidosis of 13 (23.1%) had abnormally high troponin levels, compared with two of 14 (14.3%) with AA amyloidosis (NS). Two of the three patients with AL amyloidosis and high serum troponin levels had echocardiographic evidence of amyloidosis, and one manifested clinical symptoms of heart failure.
Hemorrhagic events, sepsis, and hospitalizations were rare and did not differ between AA and AL patients. Despite similar hemoglobin rates at dialysis onset, patients with AL required a higher dose of erythropoietin on dialysis (13,325 ± 2463 versus 6425 ± 910 IU of erythropoietin per week).
Prognosis Factors for Survival More Than 1 yr and Effect of Cardiac Amyloidosis on Survival
Patients who survived more than 1 yr on dialysis were compared with those who did not. We excluded from analysis four patients who did not die but were followed for less than 1 yr on dialysis. All patients who died during the first year of dialysis had AL amyloidosis. Factors significantly associated with death during the first year on dialysis were AL type (P < 0.01), cardiac amyloidosis (OR = 18, P < 0.01), clinical symptoms of heart failure (OR = 8, P < 0.04), and a shorter time from diagnosis of amyloidosis to dialysis initiation (6.1 ± 4.7 versus 56 ± 19.5 mo, P < 0.03) (Table 5). Clinical symptoms of heart failure were present in four of seven AL patients (57.1%) who died during the first year on dialysis, but in only one of 11 AL patients (9.1%) alive at one year (OR = 13.3, P < 0.05). AL patients who survived longer than one year on dialysis tended to have a longer time from diagnosis of amyloidosis to the start of dialysis (32.3 ± 18.7 versus 6.1 ± 4.7 mo, NS).
|
Similarly, survival on dialysis tended to be higher in AL patients with isolated monoclonal gammopathy compared with those with MM stage III (median 26 versus 13 mo, NS).
The absence of deaths among AA patients during the first year on dialysis prevented multivariate analysis of 1-yr survival. A multivariate analysis using the Cox model was performed to assess the independent role of factors significantly associated with global survival in univariate analysis. AL or AA type and cardiac amyloidosis were placed in the model, but not the time from diagnosis to dialysis, because univariate analysis using the Cox model did not demonstrate significant association between this parameter and global survival. As almost all of the patients with heart failure represented a subset of the patients with cardiac amyloidosis, this variable was also left out of the analysis given the sample size and number of deaths. Multivariate analysis indicated that AL type remained significantly associated with a worst outcome (P = 0.02), but not cardiac amyloidosis (P = 0.69) (Table 6).
|
| Discussion |
|---|
|
|
|---|
We are aware of three studies that have analyzed the outcome of patients with AL amyloidosis undergoing dialysis. In a study of 211 patients, Gertz et al. reported a very poor outcome for patients with AL amyloidosis on dialysis, with a median survival of only 8.2 mo. This study also suggested that heart failure was adversely associated with survival; however specific factors associated with mortality on dialysis were not examined (2). In contrast, in two additional studies (3,4) on heterogeneous cohorts of patients with AL or AA amyloidosis (including 61 and 48 patients, respectively), patients' median survival ranged from 25 to 52 mo, respectively, from dialysis initiation, without any significant difference in survival rates between AL and AA amyloidosis patients.
For patients with AA amyloidosis undergoing dialysis, more studies have been published, but have yielded discrepant results (3–8). In two recent reports on dialyzed patients with AA amyloidosis secondary to rheumatoid arthritis, median survival varied from 8.5 to 32 mo (5,6).
These studies have several limitations. First, studies including patients with AL amyloidosis date from the early 1990s and thus many were conducted before the era of the therapeutic advances made over the last few years, including the use of melphalan plus dexamethasone, thalidomide (11), and high-dose melphalan with autologous stem cell transplantation (12,13). These treatments undoubtedly represent important breakthroughs in the management of patients with AL amyloidosis, including those on dialysis.
Second, these studies have failed to identify factors predicting early death and long term survival of patients with AA or AL amyloidosis on dialysis. Definition of such factors are crucial, because dialysis initiation could be debated for ethical reasons in patients with systemic amyloidosis and thus potentially at a high risk of mortality on dialysis.
The study presented here provides new insights into the clinical presentation and prognostic factors of patients with systemic amyloidosis undergoing dialysis. The multicentric design of this study limited bias that could result from differences between centers in dialysis practices and treatment of onco-hematologic disorders. The results presented herein allow us to highlight several key points.
First, prognosis on dialysis is fairly good for patients with AA amyloidosis: only three patients (15%) with AA amyloidosis died on dialysis, after a mean follow-up of 31.8 mo. Although this low mortality rate did not allow us to determine the median survival, the prognosis of patients with AA amyloidosis undergoing dialysis appears far better compared with previous studies.
Second, the outcome of patients with AL amyloidosis included in our study is not as poor as described previously, with a median of survival of 26 mo; however, survival time was extremely variable, ranging from 1 to 96 mo after the initiation of dialysis. Both global and 1-yr survival were markedly reduced in AL patients compared with AA patients. After multivariate analysis, type of amyloidosis was the only variable that remained significantly associated with a worse outcome. This result suggests that AL type is the main factor affecting mortality. This could be related to several causes. In addition to heart involvement, which appears to largely explain the high mortality in AL patients, complications related to MM and chemotherapy toxicity may also account for the significant difference observed in survival between AL and AA amyloidosis. Moreover, the outcome of patients with MM undergoing dialysis regardless of amyloidosis remains controversial (14–16).
Third, cardiac amyloidotic involvement is associated with a poor outcome and probably largely accounts for the mortality observed in patients with AL amyloidosis. In our study, AL patients without cardiac amyloidosis had a more than two-fold longer median survival than AL patients with heart involvement (36 versus 16.5 mo). Multivariate analysis failed to demonstrate an independent effect of cardiac amyloidosis on survival. However, this analysis is likely to have been underpowered given the limited sample size and the rarity of cardiac amyloidosis associated with AA type. Furthermore, six of the seven patients (85.7%) with AL amyloidosis who died during the first year on dialysis had cardiac involvement. In addition, clinical symptoms of heart failure were significantly associated with death during the first year on dialysis in AL patients. Our results are in accordance with previous reports that have underlined that heart involvement is the leading cause of death of patients with systemic amyloidosis undergoing dialysis (2–4). Heart involvement is also recognized as a key prognostic factor in patients with AL amyloidosis, regardless of ESRD (17), with a median survival from diagnosis of 12.6 mo, falling to 9 mo in case of heart failure (18). However, it is noteworthy that in our study, median survival was 16.5 mo in patients with AL amyloidosis and cardiac involvement, which favorably compares with data reported in nondialyzed patients. This may be related to the fact that dialysis allows adequate control of fluid overload, preventing the occurrence of heart failure symptoms. Conversely, clinically relevant heart involvement is overall rare in patients with AA amyloidosis (7,17). The prognostic value of troponin, a powerful predictive factor of survival in patients undergoing hemodialysis (19) and in those with primary amyloidosis regardless of ESRD (20), remains to be more thoroughly evaluated in patients with AL amyloidosis on dialysis.
In addition to AL type and cardiac involvement, short time from diagnosis of amyloidosis to the start of dialysis, which probably reflects a more aggressive disease with rapid and massive systemic deposition of the amyloidogenic protein, may have an effect on patients with amyloidosis undergoing dialysis. Our study may have lacked statistical power for identifying other prognostic factors, such as the type of underlying disease and its control before and over the period of dialysis.
Our study has several limitations and biases. First, because of the rarity of patients with systemic amyloidosis on dialysis, the number of patients included was limited, restricting statistical analyses. Moreover, the duration of follow-up was variable between patients, and five patients were lost to follow-up. Likewise, we cannot exclude that a small number of patients with amyloidosis and ESRD in very poor condition were considered more suitable for palliative care and did not start dialysis. We were unable to identify such patients in our cohort. That would have resulted in a selection bias leading to an overestimation of survival of patients with systemic amyloidosis and ESRD. Finally, because the vast majority of our patients underwent hemodialysis, we were unable to determine whether the type of dialysis modified the patients' outcome.
In all, prognosis of patients with systemic amyloidosis on dialysis seems to be better than previously reported. AL type, cardiac amyloidosis, and short time from diagnosis of amyloidosis to dialysis initiation are predictive of death during the first year of dialysis. AL type appears to be the most important factor affecting survival. Our data may prove helpful for clinicians in charge of patients with amyloidosis and ESRD, especially in ethical discussions regarding the decision whether or not to initiate dialysis. Our results do not allow the elaboration of definite guidelines, and each case should be discussed individually. However, the overall good prognosis of patients with systemic amyloidosis and ESRD may plead for large indication of dialysis in these patients.
| Disclosures. |
|---|
|
|
|---|
| Acknowledgments |
|---|
| Footnotes |
|---|
Received June 13, 2007. Accepted October 31, 2007.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. A. Gertz, N. Leung, M. Q. Lacy, A. Dispenzieri, S. R. Zeldenrust, S. R. Hayman, F. K. Buadi, D. Dingli, P. R. Greipp, S. K. Kumar, et al. Clinical outcome of immunoglobulin light chain amyloidosis affecting the kidney Nephrol. Dial. Transplant., October 1, 2009; 24(10): 3132 - 3137. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Ronco and P. Aucouturier Renal involvement in AL amyloidosis: the facts, the promise and the hope Nephrol. Dial. Transplant., October 1, 2009; 24(10): 2967 - 2969. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |