Anterior Aortic Plane Systolic Excursion: A Novel Indicator of Transplant-Free Survival in Systemic Light-Chain Amyloidosis.

BACKGROUND

Anterior aortic plane systolic excursion (AAPSE) was evaluated in the present pilot study as a novel echocardiographic indicator of transplant-free survival in patients with systemic light-chain amyloidosis.

METHODS

Eighty-nine patients with light-chain amyloidosis were included in the post-hoc analysis. A subgroup of 54 patients with biopsy-proven cardiac amyloid infiltration were compared with 41 healthy individuals to evaluate the discriminative ability of echocardiographic findings. AAPSE is defined as the systolic excursion of the anterior aortic margin. To quantify AAPSE, the M-mode cursor was placed on the aortic valve plane in parasternal long-axis view at end-diastole. Index echocardiography had been performed before chemotherapy. Median follow-up duration was 2.4 years. The primary combined end point was heart transplantation or overall death.

RESULTS

Mean AAPSE was 14 ± 2 mm in healthy individuals (mean age=57 ± 10 years; 56% men; BMI=25 ± 4 kg/m2). AAPSE < 11 mm separated patients from age-, gender-, and BMI-matched control subjects with 93% sensitivity and 97% specificity. Median transplant-free survival of patients with AAPSE < 5 mm was 0.7 versus 4.8 years (P = .0001). AAPSE was an independent indicator of transplant-free survival in multivariate Cox regression (echocardiographic model: hazard ratio=0.72 [P = .03]; biomarker model: hazard ratio=0.62 [P = .0001]). Sequential regression analysis suggested incremental power of AAPSE as a marker of transplant-free survival. An ejection fraction-based model with an overall χ2 value of 22.8 was improved by the addition of log NT-proBNP (χ2 = 32.6, P < .005), troponin-T (χ2 = 39.6, P < .01), and AAPSE (χ2 = 54.0, P < .0001).

CONCLUSIONS

AAPSE is suggested as an indicator of transplant-free survival in patients with systemic light-chain amyloidosis. AAPSE provided significant incremental value to established staging models.