Abstract
Background
Skeletal scintigraphy with bone tracers is a key tool for cardiac ATTR diagnosis. However its prognostic value has not been systematically assessed.
Purpose
We evaluated the prognostic relevance of a quantitative method to assess regional 99mTc-DPD uptake by SPECT in the heart of ATTRwt patients.
Methods
All ATTRwt patients (n=229) undergoing clinical assessment and bone scintigraphy at our center (from 2012 to 2019) were enrolled. Theyreceived approximately 700 MBq of 99mTc-DPD. Planar whole body acquisition 10' after the injection followed by cardiac SPECT after 3 hours were performed. SPECT data were reconstructed into 64x64 matrices with an ordered-subset expectation maximization algorithm. For each wall region and for the apex, a circular region of interest (ROI, 20 pixels) was manually drawn and a value equating to the number of counts contained in the ROI was obtained. Partial correlation of ln-transformed ROI and biomarkers was retrieved from a multivariable regression model, while controlling for each cardiac wall region. Multivariable Cox regression was used to assess the prognostic role of lnROI while adjusting for wall region, NT-proBNP, cTnI and eGFR. Hazard ratios and 95% confidence intervals (HR, 95% CI) were computed. The Harrell's c statistic was reported for model discrimination. The interaction of biomarker and regional wall on survival was assessed; also, to account for intra-subject correlation of measures, within subject robust standard errors were computed.
Results
Median follow-up was 21 months (IQR 11, 40) and 39 (17%) patients died. Median age was 76 years (IQR, 72–80), NT-proBNP 2944 ng/L (IQR, 1815–5319), cTnI 0.095 ng/L (IQR, 0.062–0.144) and eGFR 62 mL/min (IQR, 51–77). ROI did not correlate with any of NT-proBNP, eGFR, age, cTnI or mLVWT (R<1% in all cases). All analyses were adjusted for cardiac wall. At the multivariable Cox regression (Harrell's c=0.75), there was a linear increase in the risk of death associated with lnROI (HR 2.14, P=0.014), which was independent of cardiac wall region, NTproBNP, cTnI and eGFR. Only cTnI maintained a significant prognostic value. The association of lnROI and mortality was not modified by the site of measurement test for interaction with cardiac wall p=0.818). At the predefined subgroup analysis, the risk of death was similar for all walls; we computed the optimal cut-off for 12 months survival at the apex (a region usually lately involved) to 4193 (AUC: 0.68, sensitivity 80%, specificity 68%). At the multivariable Cox regression (Harrell's c 0.76), apex ROI>4193 was an independent predictor of death (HR 3.60, 95% CI 1.45–8.93, p=0.006) and outperformed all the biomarkers tested.
Conclusions
Quantitative assessment of ROI uptake at cardiac SPECT is a powerful predictor of survival in ATTRwt patients, independent of and outperforming the other known prognostic factors. This observation warrants validation with prolonged follow-up and in independent patient series.
Funding Acknowledgement
Type of funding source: None
Inhibition of the canonical Wnt signaling pathway by a β-catenin/CBP inhibitor prevents heart failure by ameliorating cardiac hypertrophy and fibrosis