Systemic AL amyloidosis is a plasma cell dyscrasia and protein conformation disorder in which Ig light chains (Ig VL) produced by clonal plasma cells form interstitial amyloid in key viscera causing organ dysfunction and death. Patients with AL usually have < 20% marrow plasma cells with low to nil proliferative indices. Paradoxically the plasma cells appear immune to the toxic amyloid-forming Ig VL possibly due to the robust cytoplasmic protein quality-control processes in plasma cells. A standard approach is to treat with high-dose melphalan and stem cell transplant (SCT) without induction therapy. Post-SCT long-term survival depends on significant reductions of the plasma cell disease and Ig VL levels as measured by the serum free light chain (FLC) assay. At 3 months post-SCT responses of the plasma cell dyscrasia are defined both by standard criteria and by normalization of abnormal FLC ratios. By standard criteria, post-SCT one-third fail to achieve a > 50% reduction in plasma cell disease (NR), one-third achieve a > 50% reduction (PR), and one-third achieve clearance of marrow plasma cells and a complete or near complete response (CR). By FLC criteria, post-SCT one-third achieve normalization (N) of the FLC ratio and two-thirds do not (A). Many factors contribute to this distribution of responses. In order to identify factors specific to AL plasma cells, gene expression profiles (GEP; Affymetrix U133 PLUS 2.0) were obtained after double amplification of RNA from FACS-sorted CD138+/DAPI- plasma cells from untreated patients with AL pre-SCT. Data were vetted based on plasma cell lineage gene expression; samples contaminated with monocytes were not used. Supervised analyses were performed based on responses at 3 months post-SCT, comparing CR (n=4) to PR (n=7) and CR to NR (n=5), and FLC N (n=5; one with CR) to A (n=7; two with CR). Differentially expressed genes between paired sets were identified using the t-test. Genes with p < 0.01 were examined using EASE, a program that identifies over-represented gene families. In the CR-PR and CR-NR comparisons, genes involved in translation, RNA ligation and protein degradation, particularly aminoacyl tRNA synthetases, were significantly over-represented with Bonferroni-adjusted EASE scores (like p-values) < 0.05. In the CR set, tryptophanyl tRNA synthetase, a protein that can also be anti-angiogenic, and IDE, an insulin-degrading enzyme that also degrades Aβ amyloid peptides, were among the most over-expressed. In the FLC N-A comparison, protein transport and detoxification gene families were also significantly over-represented with Bonferroni-adjusted EASE scores < 0.001. In the N set, CCT subunits (chaperones), UBE2B (a ubiquitination enzyme and RAD6 homolog) and glyoxalase I (detoxification) were among the most over-expressed. Our initial hypothesis is that the CR and FLC N responses to melphalan have similar but distinctive GEP related to protein folding, ER stress and protein degradation. The responses of AL plasma cells to Ig VL in the ER may influence the patterns observed, possibly modulating melphalan uptake or activity. Further studies of these differences and functional hypothesis testing are currently underway.