Relative apical sparing of longitudinal strain usingtwo-dimensional speckle-tracking echocardiographyis both sensitive and specific for the diagnosis ofcardiac amyloidosis

Dermot Phelan, Patrick Collier, Paaladinesh Thavendiranathan, Zoran B Popovi
c,
Mazen Hanna, Juan Carlos Plana, Thomas H Marwick, James D Thomas
ABSTRACT
Background The diagnosis of cardiac amyloidosis (CA) is
challenging owing to vague symptomatology and nonspecific echocardiographic findings.
Objective To describe regional patterns in longitudinal
strain (LS) using two-dimensional speckle-tracking
echocardiography in CA and to test the hypothesis that
regional differences would help differentiate CA from
other causes of increased left ventricular (LV) wall
thickness.
Methods and results 55 consecutive patients with CA
were compared with 30 control patients with LV
hypertrophy (n¼15 with hypertrophic cardiomyopathy,
n¼15 with aortic stenosis). A relative apical LS of 1.0,
defined using the equation (average apical LS/(average
basal LS + mid-LS)), was sensitive (93%) and specific
(82%) in differentiating CA from controls (area under the
curve 0.94). In a logistic regression multivariate analysis,
relative apical LS was the only parameter predictive of
CA (p¼0.004).
Conclusions CA is characterised by regional variations
in LS from base to apex. A relative ‘apical sparing’
pattern of LS is an easily recognisable, accurate and
reproducible method of differentiating CA from other
causes of LV hypertrophy.
BACKGROUND
Cardiac amyloidosis (CA) is an infiltrative cardiomyopathy characterised by deposition of b-pleated
fibrils in the myocardium and usually portends
a poor prognosis.1 The diagnosis is challenging and
relies on a high degree of clinical suspicion,
a combination of imaging techniques and often an
endomyocardial biopsy.2 Symptoms early in the
disease course are vague and non-specific, which
often leads to a delay in diagnosis. Echocardiography remains the preferred imaging study for the
initial assessment of cardiac involvement; however,
findings suggestive of CA, such as increased wall
thickness, impaired systolic and diastolic function
and reduced myocardial tissue velocities, are regularly attributed to more prevalent pathologies.
More contemporary echocardiographic techniques play an adjunctive role in the diagnosis of
CA.3 Specifically, strain and strain rate imaging
parameters obtained by tissue Doppler techniques
have been shown to be significantly reduced in
primary amyloidosis with cardiac involvement in
comparison with patients without cardiac
involvement.4 Furthermore, two-dimensional (2-D)
speckle tracking of global radial, circumferential
and longitudinal strain (LS) analysis can help to
differentiate CA from hypertensive heart disease
and hypertrophic cardiomyopathy (HCM).
However, the specificity of this technique is
diminished because any myopathic process can
result in a reduction in global strain parameters.
Patients with left ventricular hypertrophy (LVH) or
HCM and impaired left ventricle (LV) systolic
function may have reduced global strain in the
same range as patients with CA, making it difficult
to differentiate these entities solely on the basis of
this measurement.
In our echocardiography laboratory, a pattern of
regional differences in LS of the LV has been noted
among patients with CA. The aims of this study
were (1) to describe the pattern of regional LS in
patients with CA using 2-D speckle-tracking echocardiography and (2) to assess whether this pattern
is specific to CA and helps to differentiate patients
with CA from patients with other causes of
increased LV wall thickness.
PATIENTS AND METHODS
Patient population
This was a caseecontrol study involving consecutive patients referred to the Cleveland Clinic
Foundation (CCF) echocardiography laboratory
with a presumptive diagnosis of CA between
October 2010 and March 2012. Study patients were
identified from the echocardiography database
using the search term ‘amyloid’ and its variants and
from a clinical database maintained at our local
amyloid outpatient clinic. Medical records of all
these patients were reviewed to confirm the diagnosis of CA. A confirmatory diagnosis consisted of
either (1) an endomyocardial biopsy consistent
with CA,5 or (2) a positive non-cardiac biopsy for
amyloidosis and cardiac magnetic resonance
imaging (CMR) with classic features,6 or (3)
a positive non-cardiac biopsy with characteristic
non-strain-based echocardiographic parameters
together with a consistent clinical history and ECG
findings.5 As the clinical challenge relates to the
correct identification of amyloidosis among the