George Eng, Marc K. Halushka, Daniel P. Judge, Marc J. Semigran, James R. Stone
Background: Accurate classification of cardiac amyloidosis, between transthyretin and light chain kappa or lambda, is paramount for optimal patient management. However, direct subtyping of amyloid deposits has significant drawbacks such as long turnaround time, high cost, and in some cases, the need to send the material to an outside reference laboratory. Potential differences in serum free light chain kappa/lambda ratios may provide a method to distinguish between different forms of cardiac amyloidosis. However, the standard upper and lower limits of normal for the assay are not optimized for cardiac amyloidosis classification. Therefore we investigated if optimization of the kappa/lambda serum free light ratios could allow for a novel method to accurately and rapidly subclassify cardiac amyloidosis.
Design: We investigated 78 cases of tissue proven cardiac amyloidosis (endomyocardial biopsy: n=65, ventricular septal myectomy: n=1, ventricular apical core: n=1, explanted heart: n=9, and autopsy heart: n=2) at two separate medical centers. All patients had serum free light chain analysis obtained prior to plasma cell neoplasm treatment in conjunction with routine classification of cardiac amyloidosis (mass spectrometry n=51, or immunofluorescence n=27). We then stratified the cases based on the ratio of kappa to lambda serum free light chains and correlated that value to their classification of cardiac amyloidosis.
Results: The serum free light chain kappa/lambda ratios were non-overlapping for the three types of amyloid identified: AL-lambda (0.01-0.41, n=29), ATTR (0.63-2.7, n=38), and AL-kappa (6.7-970, n=11). Using optimized cut-off values for the kappa to lambda serum free light chain ratio, we were able to accurately distinguish between the three common forms of ventricular cardiac amyloidosis within this cohort of patients. A kappa to lambda ratio value between 0.5 and 5.0 had 100% sensitivity and 100% specificity for distinguishing transthyretin from amyloid light-chain amyloidosis (n=78, 95% confidence interval: 89%-100% for both sensitivity and specificity). The capacity of this test to distinguish between forms of amyloidosis was equivalent for cases subtyped by either mass spectrometry or immunofluorescence.
Conclusions: Optimized ranges for serum light chain kappa/lambda ratio can provide extremely robust classification of cardiac amyloidosis. The assay is widely available, relatively inexpensive and can deliver accurate, rapid results. Cases of cardiac amyloidosis in which the kappa/lambda free light chain ratio falls close to these new cut-off values would likely benefit most from direct amyloid subtyping.