Identification Of Reversible Organ Damage and Of Early Markers Of Response To Treatment In Renal AL Amyloidosis: A Study On 461 Patients

In the present study we evaluated 461 consecutive previously untreated patients with AL amyloidosis and renal involvement (defined as albuminuria >0.5 g/24h as per current criteria) diagnosed between 2004 and 2012. Median age was 64 years, and 264 patients (57%) were males. The heart was involved in 311 patients (67%) and 160 (35%) were Mayo stage III. Thirty-one (7%) patients had liver and 21 (4%) peripheral nervous system involvement. At diagnosis the median estimated glomerular filtration rate (eGFR) by the CKD-EPI formula was 62 mL/min per 1.73 m² and median proteinuria was 5 g/24h. The CKD stage was 1 in 94 patients (20%), 2 in 148 (32%), 3 in 139 (30%), 4 in 67 (15%), and 5 in 13 (3%). Seventy-one patients (15%) required dialysis after a median time of 10 months (range 1-94 months). Patients who died off-dialysis were considered censored for the purpose of the analysis. The median survival from diagnosis was 47 months, and the median survival from dialysis initiation was 39 months. At univariate analysis the only baseline variables predicting renal survival were eGFR and proteinuria. We adopted values close to the medians as clinical cutoffs. At multivariate analysis eGFR <60 mL/min (HR 10.0, 95%CI 2.5-40.6, P=0.001) and urinary protein loss >5 g/24h (HR 3.5, 95%CI 2.0-6.3, P<0.001) were independent predictors of renal survival. These cutoffs allowed discriminating 3 groups with increasing risk of progression to dialysis (Figure 1). Importantly, no patient with proteinuria £5 g/24h and eGFR ³60 mL/min required dialysis during follow-up. Both eGFR and proteinuria did not affect overall survival. The analysis of early response/progression criteria was performed on the 314 patients who had a complete dataset at 3 months. A ROC analysis showed that the cutoffs best predicting renal survival were a 25% eGFR decrease, a 20% proteinuria decrease, and a dFLC concentration <40 mg/L, which, remarkably, corresponds to the current definition of very good partial response (VGPR). A multivariate analysis showed that these cutoffs were independent predictors of renal survival and were also independent from the baseline stage based on proteinuria and eGFR (Table 1). Partial hematologic response (dFLC decrease by >50%) and the current definition of renal response (proteinuria decrease of >50% in the absence of progressing renal failure) were less powerful predictors not reaching statistical significance.