Abnormal NT-ProBNP in AL Amyloidosis Patients without Cardiac Involvement at Diagnosis - A Predictor of Subsequent Cardiac Involvement.

Cardiac involvement in AL amyloidosis is associated with a poor prognosis and greatly increased treatment related morbidity and mortality, and regression of cardiac amyloid deposits is extraordinarily slow following chemotherapy that suppresses the underlying aberrant light chain production. Diagnosis of cardiac amyloidosis is normally made by echocardiography, by which time significant diastolic dysfunction has usually developed. Atrial natriuretic peptides (ANP, BNP and its N-terminal fragment NT-ProBNP) are useful in early diagnosis of myocardial dysfunction. Serum NT-ProBNP concentration has been reported to be a promising marker of cardiac dysfunction in AL amyloidosis, and patients with normal NT-ProBNP values at diagnosis have superior outcomes. We report here the outcome of patients attending the UK National Amyloidosis Centre (NAC) who had elevated NT-ProBNP at diagnosis of AL amyloidosis but who did not have accompanying evidence of cardiac involvement using conventional consensus criteria. To exclude the confounding effect of renal failure which is associated with substantial elevation of NT-ProBNP, we studied patients with serum creatinine <150 μmol/L and creatinine clearance of >50ml/min at diagnosis in whom there was less than 10% change in renal function after treatment. AL type amyloidosis was confirmed in all patients histologically with corroborating genetic studies to robustly exclude hereditary amyloidosis as indicated. Organ involvement and responses/progression were defined according to recent international consensus criteria (Gertz et al 2005). 102 patients who had no evidence of cardiac involvement by these conventional parameters and who otherwise conformed with our study criteria were identified. Median creatinine was 87 μmol/L (44–128), albumin 33g/L (10–65), bilirubin 7 μmol/L (1–65) and alkaline phosphatase 89 units/L (36–2649). The median interventricular septal and left ventricular posterior wall thickness was 9 mm (7–11 mm). 62 (61%) patients had NT-ProBNP ≤ 35pMol/L at diagnosis while 40 (39%) had NT-ProBNP of >35 pMol/L. There was no significant difference in the baseline characteristics of either group. 5 patients in each group did not respond to the initial chemotherapy (p=0.46). With median follow-up of 60 months, 19/40 (47%) of patients with NT-ProBNP >35pMol/L at diagnosis developed evidence of cardiac involvement compared to only 6/62 (10%) of whose baseline NT-ProBNP was ≤ 35 pMol/L (p<0.001). The Kaplan-Meier estimated median overall survival has not been reached for either group but the estimated 7 year survival was significantly better in the group with NT-ProBNP of ≤35pMol/L compared to those with greater values (92% vs. 82%, p=0.03). In conclusion, these preliminary findings suggest that patients who have elevated NT-ProBNP concentration but no conventional evidence of cardiac involvement at diagnosis of AL amyloid appear to be at greater risk of developing cardiac amyloidosis during follow-up, and have a poorer prognosis. It reasonable to speculate that such patients have early cardiac involvement at diagnosis that cannot be identified by conventional non-invasive methods, and that their risk of subsequently developing clinically significant cardiac amyloidosis may be reduced by striving to achieve complete remission of their underling clonal plasma cell disease.