Author + information
- Received March 31, 2017
- Revision received June 12, 2017
- Accepted June 16, 2017
- Published online February 5, 2018.
- Brett W. Sperry, MDa,∗ (, )
- Michael N. Vranian, MDa,
- Albree Tower-Rader, MDa,
- Rory Hachamovitch, MSc, MDa,
- Mazen Hanna, MDa,
- Richard Brunken, MDa,b,
- Dermot Phelan, MD, PhDa,
- Manuel D. Cerqueira, MDa,b and
- Wael A. Jaber, MDa,b
- aDepartment of Cardiovascular Medicine, Cleveland Clinic Foundation, Cleveland, Ohio
- bDepartment of Nuclear Medicine, Cleveland Clinic Foundation, Cleveland, Ohio
- ↵∗Address for correspondence:
Dr. Brett W. Sperry, Cleveland Clinic Foundation, Department of Cardiovascular Medicine, Heart and Vascular Institute, 9500 Euclid Avenue, Desk J1-5, Cleveland, Ohio 44195.
Objectives This study sought to investigate the regional uptake of technetium 99m-pyrophosphate (TcPYP) in transthyretin cardiac amyloidosis (ATTR) and its association with mortality.
Background TcPYP nuclear scintigraphy is a diagnostic and prognostic tool in ATTR. Echocardiography has identified a pattern of regional variation in longitudinal strain (LS) with a gradient of improved strain from base to apex in ATTR.
Methods Consecutive patients with ATTR were evaluated who underwent TcPYP nuclear scintigraphy with planar and attenuation corrected single-photon emission computed tomography (SPECT). Heart-to-contralateral lung (H/CL) ratio was calculated on planar images, and left ventricular (LV) uptake was determined in each of the 17 segments using SPECT. A measure of apical-sparing of myocardial TcPYP uptake, termed the apical-sparing ratio (ASR), was calculated as basal + mid / apical counts.
Results Overall, 54 patients with ATTR (age 78 ± 9 years, 76% male, 31% hereditary ATTR) were analyzed. There was increased TcPYP uptake in basal and mid relative to apical LV segments, and an apical-sparing LS pattern on echocardiography. The right ventricle similarly showed greater uptake in basal segments. There were 26 deaths over 1.8 years median follow-up. The ASR of TcPYP uptake was associated with age-adjusted all-cause mortality (p = 0.013) with worse prognosis seen at levels <2.75. Global LS was also prognostic (p = 0.01), whereas H/CL ratio and total LV uptake indexed to blood pool were not (p = 0.772 and p = 0.850, respectively). The prognostic utility of the ASR persisted in multivariable modeling (p = 0.003), whereas global LS did not.
Conclusions There is decreased TcPYP uptake in apical as compared to mid and basal segments in the LV, mimicking apical-sparing LS seen on echocardiography. Regional distribution of LV TcPYP uptake is associated with mortality, whereas overall amount of uptake as measured by H/CL ratio and indexed LV SPECT uptake is not.
- advanced cardiac imaging
- heart failure with preserved ejection fraction
- infiltrative cardiomyopathies
- nuclear cardiac imaging
Heart failure with preserved ejection fraction is sometimes a result of transthyretin cardiac amyloidosis (ATTR). Transthyretin, a protein made in the liver involved in thyroid hormone transport, misfolds and pathologically deposits into the myocardium. This leads to an increase in wall thickness and stiffness resulting in diastolic dysfunction.
Echocardiography is the primary imaging tool used in the initial differentiation among etiologies of increased left ventricular (LV) wall thickness, and regional assessment of longitudinal strain (LS) can help distinguish pressure overload states from specific cardiomyopathies (1). Cardiac amyloidosis leads to a typical pattern of regional variation in LS with relatively preserved deformation in apical segments compared to mid and basal segments, the so-called “relative apical-sparing” pattern (2). A relative regional strain ratio can be calculated which measures the degree of apical-sparing of LS (3), a pattern that is corroborated by cardiac magnetic resonance imaging (4).
Technetium pyrophosphate nuclear scintigraphy (TcPYP) is used off-label in the United States for cardiac amyloidosis and is both sensitive and specific for diagnosing ATTR (5). The interpretation of this imaging modality has classically been based on a visual comparison of heart uptake to bone uptake or a heart-to-contralateral lung (H/CL) uptake ratio on simple planar imaging. To date, data acquired from single-photon emission computed tomography (SPECT) using TcPYP have not been used to quantitatively or semi-quantitatively interpret variations in regional amyloid deposition or prognosis.
We hypothesized that, given the data from echocardiographic and cardiac magnetic resonance imaging studies, there would be regional variation in TcPYP uptake and that the distribution of tracer uptake would be associated with prognosis. Therefore, we sought to examine the relative regional uptake of TcPYP in patients with ATTR and its association with mortality.
We identified consecutive patients having undergone TcPYP scintigraphy at our institution from December 2011 to September 2016 for the evaluation of cardiac amyloidosis. Patients were included with ATTR confirmed by endomyocardial biopsy or by nuclear scintigraphy with visual semi-quantitative grade 2 to 3 as previously described (5). Two patients with grade 1 uptake were included due to endomyocardial biopsy confirmation. Patients with light chain (AL) amyloidosis, prior myocardial infarction, or unanalyzable images were excluded from analysis. AL was ruled out after analysis of serum free light chains and immunofixation. In the case of a monoclonal gammopathy, ATTR was confirmed based on endomyocardial or extracardiac biopsy specimens using tissue typing with immunohistochemistry and mass spectrometry as needed.
Electronic medical records were retrospectively reviewed for demographic, clinical, and follow-up data. Patients with incomplete follow-up information were contacted by telephone to assess vital status. Twelve-lead electrocardiograms were performed using standard equipment and retrospectively reviewed for heart rate, rhythm, and voltage. Vivid 7 or Vivid 9 (GE Medical, Milwaukee, Wisconsin) or EPIQ (Philips Medical Systems, Bothell, Washington) ultrasound systems were used for image acquisition. Echocardiographic parameters were measured in standard fashion as described by the American Society of Echocardiography guidelines (6). LV mass was calculated using end-diastolic dimensions measured in the parasternal long axis. Stroke volume was calculated from the LV outflow tract using Doppler flow measurements. Peak systolic LS imaging was performed in all patients using Velocity Vector Imaging (Syngo VVI, Siemens Medical Solutions, Mountain View, California) software and LV global and segmental LS values were reported. Biopsy histology was reviewed by 2 cardiac pathologists experienced in cardiac amyloidosis.
TcPYP scintigraphy was performed using SPECT-CT with Siemens Symbia T6 cameras. Patients were scanned 3 h after receiving 20 mCi ±10% of TcPYP intravenously. Whole body planar images were acquired with the heart centered on a 256 × 1,024 matrix, and cardiac SPECT images on a 128 × 128 matrix with no zoom. SPECT used a low-energy high-resolution collimator, a 15% energy window, and CT attenuation correction. Planar images were analyzed offline using 4DM software (INVIA, Ann Arbor, Michigan) by 2 experienced readers blinded to clinical and echocardiographic data. SPECT images were analyzed by an experienced reader after manual reformatting if the automated software algorithm did not adequately detect myocardial borders. Ten studies were remeasured for intraobserver and interobserver variability and evaluated using intraclass correlation coefficients.
A semi-quantitative score was obtained based on results of planar images as previously described by Perugini (7). Briefly, a score of 0 signified absent cardiac uptake, 1 indicated mild uptake less than bone, 2 indicated moderate uptake equal to bone, and 3 indicated strong uptake with mild or absent bone uptake. The H/CL ratio was calculated using the planar images by measuring radiotracer uptake over the heart using a circular region of interest and comparing to the mirror image superimposed over the contralateral lung (8,9). SPECT images were used to obtain quantitative measurements of segmental myocardial uptake. Mean tracer uptake counts per voxel were determined for each of the 17 segments. Mean blood pool counts were calculated by averaging counts obtained from 3 regions of interest in the proximal, mid, and distal ascending aorta. LV counts normalized to blood pool were calculated by dividing mean total LV by mean blood pool counts. Subsequently, regional mean count statistics were calculated by using the 6 basal, 6 mid, and 5 apical segments along with each region’s percentage of total myocardial uptake. A measure of apical-sparing of myocardial TcPYP uptake, termed the apical-sparing ratio (ASR), was calculated by adding all basal and mid segment counts and dividing by apical counts. Thus, a higher ASR signifies more apical-sparing of TcPYP uptake. SPECT images were also used to visually assess qualitative uptake in the atria and right ventricular free wall.
Categorical variables are presented as percentages and compared with the Fisher exact test. Continuous variables are expressed as mean ± SD or median (interquartile range) and compared with the paired Student t test or Wilcoxon-Mann-Whitney test where appropriate. Linear regression was used to assess the association between regional LS and TcPYP uptake. Scatter and box plots were developed to graphically represent regional TcPYP uptake and echocardiographic LS. To assess the correlation between TcPYP regional uptake and LS values, a linear mixed effects model was used with LS as the dependent variable, TcPYP uptake as a covariate, LV region as random-effects parameter, and patient identifier as the repeated measures factor.
Survival analysis was accomplished using Cox models to test the associations between age-adjusted echocardiographic and nuclear scintigraphic variables with all-cause mortality. Cumulative mortality was graphically represented using age-adjusted survival curves for ASR and global LS using median values as thresholds. Fractional polynomial curves were fitted to show the continuous relationship between ASR and global LS and mortality. Figure statistics were assessed using age-adjusted Cox models. Subsequently, a multivariable Cox model of all imaging variables was generated using backward stepwise selection with a removal significance threshold of 0.05. Harrell’s C statistic was used to compare model discrimination.
All statistical tests were 2-sided and p values <0.05 were considered statistically significant. All models were tested for linearity, collinearity, additivity, and proportional hazards where appropriate. Statistical analysis was performed using Stata version 13 (StataCorp, College Station, Texas). The study was approved by the Institutional Review Board.
There were 92 patients who underwent TcPYP nuclear scintigraphy during the study period; 32 lacked myocardial uptake on planar and SPECT images, did not have cardiac ATTR, and were excluded from the analysis. Two patients were excluded after AL was diagnosed (based on endomyocardial biopsy). Four patients were excluded due to significant bone uptake obfuscating the ability to quantify planar and regional myocardial SPECT tracer uptake. Thus, a total of 54 patients were included in the final analysis. The characteristics of these patients are seen in Table 1. Twelve patients had endomyocardial biopsy confirmation of ATTR. Seventeen had a mutation in the transthyretin gene, with the most common variant V122I (n = 15). Three patients had monoclonal gammopathy of undetermined significance. Patients who died during the study period were more likely to be older, less obese, and have higher levels of troponin T and N-terminal pro-brain natriuretic peptide.
With respect to regional TcPYP uptake patterns, less radionuclide uptake was noted in apical segments as compared to basal and mid segments (Table 2, Figure 1A). This apical-sparing uptake mimicked the pattern of impaired echocardiographic LS in basal and mid segments (Figure 1B). One-way analysis of variance confirmed a basal to apical gradient for both TcPYP uptake and regional LS (p < 0.001 for both). A mixed model confirmed that the degree of TcPYP uptake was associated with the degree of regional strain impairment (p < 0.001). A representative example of apical-sparing of TcPYP uptake is seen in Figure 2 with the corresponding echocardiographic LS analysis. The mean ASR (mean basal + mid / apical counts) in the entire cohort was 2.75. The intraclass correlation coefficients for intraobserver and interobserver variability were 0.983 and 0.963, respectively (p < 0.001 for both).
Most patients had some degree of atrial and right ventricular (RV) uptake. All patients with RV uptake had uptake in basal segments, whereas fewer had uptake in mid and apical segments. Isolated basal RV uptake and diffuse RV uptake are seen in Figures 3A and 3B, respectively.
There was a total of 63 patient-years of follow up, with a median follow up of 1.79 years per patient. Twenty-six of the 54 patients died during the study period. No patients received heart transplantation or mechanical circulatory support. There were no differences in planar imaging variables based on subsequent outcomes; semi-quantitative uptake grade was 2 or 3 in 96% of patients and mean H/CL ratio was 1.65. Quantitative uptake based on analysis of SPECT images is also seen in Table 2. LV uptake indexed to blood pool was not different between patients who lived versus died.
Results of survival analysis are seen in Tables 3 and 4⇓⇓. A more diffuse myocardial uptake of TcPYP, represented by higher percentages of uptake in apical segments, lower percentage of uptake in basal segments, and a lower ASR, were all associated with age-adjusted mortality (p = 0.013, Harrell’s C-statistic 0.722 for ASR). Age-adjusted ejection fraction and global LS were also associated with mortality (p = 0.041 and p = 0.01, respectively). Other parameters of presumed echocardiographic and nuclear disease severity, including the H/CL ratio and total indexed LV counts, were not associated with age-adjusted mortality. Fractional polynomial curves were fitted to show the continuous association of TcPYP uptake (Figure 4A) and global LS (Figure 4B) with mortality. A higher ASR of TcPYP uptake was associated with decreased mortality, with decreased risk seen at values above 2.75. Survival curves adjusted for age are seen in Figures 5A and 5B showing the association between ASR values <2.75 and increased mortality (p = 0.002) and global LS above and below the median (p = 0.089).
The multivariable stepwise model including all imaging parameters showed that ejection fraction, average E/e′, and the ASR were significantly associated with mortality with Harrell’s C statistic 0.773 (Table 4). Global LS was not statistically significant after multivariable analysis and was systematically removed from the stepwise model. These findings persisted when ASR and global LS were treated as categorical values using cut points above and below the mean (ASR p = 0.002, global LS removed from model).
Technetium-based bone scintigraphy and strain echocardiography have revolutionized the diagnostic work-up of ATTR cardiac amyloidosis and minimized the need for endomyocardial biopsy. An apical-sparing pattern of LS on echocardiography has been shown to be both diagnostic (2) and prognostic (3) in this disease. LS becomes more impaired beginning in the basal and mid segments and progresses to involve all segments leading to a decrease in global deformation. In this study, we capitalized on these findings to describe a simple, quantitative, reproducible method for the assessment of SPECT images using technetium-based bone scintigraphy in cardiac amyloidosis. We found that the uptake of TcPYP has a predilection for basal and mid segments, mimicking the pattern initially seen in strain echocardiography. TcPYP tracer uptake was also commonly noted in the atria and RV, with an apical-sparing pattern also present in the RV. More diffuse LV deposition as quantified by increased apical counts and a lower ASR is associated with increased mortality, likely reflecting late stage disease. Overall uptake density, as assessed by the semi-quantitative score and H/CL ratio on planar and indexed LV counts on SPECT, were not prognostically important in our cohort.
Prior analysis of technetium-based bone scintigraphy has centered on planar imaging techniques. This has been promoted as a simple and highly sensitive and specific diagnostic tool in ATTR cardiac amyloidosis (10). The original Perugini score (7), the heart to whole body ratio (7), and the H/CL ratio (8) have been used to differentiate ATTR from AL and other forms of non-amyloid cardiomyopathy. However, there are significant flaws inherent to planar analyses that were highlighted several decades ago when planar nuclear infarct studies were performed after myocardial infarction. Patients afflicted with ATTR are typically older and may suffer from osteoarthritis which leads to patient rotation and differential bony uptake. A simple metric such as the H/CL ratio may overcome these limitations in patients with significant disease, but may be relatively insensitive to judge disease progression. This is evidenced by the fact that the H/CL does not significantly change in serial examinations despite clinical progression of disease (11). Using SPECT also aids in localizing uptake to the myocardium to avoid measuring counts in the blood pool, which may show enhanced activity in renal dysfunction or very low cardiac output. This could be misconstrued as myocardial uptake by altering the H/CL ratio, and thus result in a false positive diagnosis of ATTR.
Bone scintigraphy and prognosis
Regional TcPYP uptake on SPECT imaging used in our study was found to identify patients with more diffuse uptake (lower ASR) in which there was an increase in mortality. This relationship was linear, with a cut point of 2.75 (in this cohort with this imaging protocol) above which, survival was improved. Neither uptake density on planar imaging (semi-quantitative Perugini grade and H/CL ratio) or SPECT (total indexed LV counts), nor echocardiographic global LS were prognostic in the multivariable model. Specifically, global LS showed borderline significant prognostic value on univariable and age-adjusted analyses; however, it was not statistically significant when included in the multivariable model along with nuclear scintigraphic variables including the ASR.
While planar imaging has gained diagnostic acclaim, efforts to turn this modality into a prognostic marker have met varied success. Prior study from our group showed that the H/CL ratio was not associated with outcomes or echocardiographic parameters of disease severity in patients with confirmed ATTR (9). A European study showed that the H/CL ratio was not associated with mortality, but was associated with outcomes when heart failure hospitalizations and new arrhythmias were included as endpoints (12). A recent multicenter pooled analysis of planar imaging in 121 patients with ATTR has concluded that the H/CL is associated with mortality (13). However, this analysis has limitations including a combined cohort of patients with 1- and 3-h scan delays (despite a marked difference in H/CL ratios between protocols) (14). The mixed results of these prior data are difficult to interpret, but suggest that the H/CL ratio may be prognostic when a large sample size is analyzed. Similar to the transition from planar infarct studies to SPECT in coronary artery disease, it is logical that SPECT based methods for image quantification will be superior with regards to diagnosis and prognosis. Glaudemans et al. (15) assessed the total indexed LV counts using LV to blood pool ratio on SPECT bone scintigraphy, and noted that patients with ATTR cardiac amyloidosis had significantly higher counts as compared to controls or genetic carriers. However, these indexed LV counts were not prognostic in our analysis.
Disease mechanisms and clinical implications
The regional distribution of TcPYP uptake in the myocardium has not previously been described, and may inform our knowledge of the disease process. Increased radiotracer uptake in basal and mid LV and RV segments may be caused by several factors. First, there could be more amyloid fibril deposition in these regions. One study noted increased amyloid deposition in basal and mid segments in the explanted hearts of 3 patients with cardiac amyloidosis as well as a lower percentage of apical segments with late gadolinium enhancement (4), but this has not been confirmed with large studies examining histopathologic or imaging deposition. This hypothesis could be tested by using radionuclides that bind specifically to the amyloid fibril. Second, there may be more calcium deposits and fibrosis in basal and mid segments, yielding increased PYP uptake in these regions. Third, due to the properties of nuclear imaging, the uptake of PYP may be noted in basal and mid segments first as that is where myocardial walls are thickest. This observation is due to the partial volume effect and the fact that increased LV thickness is often characterized by increased radionuclide signal intensity. As the disease progresses, apical uptake becomes more visible and hence the gradient lessens. Fourth, it has been postulated that as disease progresses, there may be more skeletal muscle uptake using 99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid (DPD) with relatively lower counts seen in the bone (i.e., Perugini score 3) (16). Therefore, regional differences in uptake may be seen whereas total myocardial counts may not increase significantly. However, a Perugini score of 3 versus 2 has not been shown to be prognostically detrimental, and anecdotally we do not see significant skeletal muscle uptake with the PYP radionuclide. Head-to-head comparison of uptake properties of these 2 agents is needed to answer these questions.
This study also describes uptake of TcPYP in other areas of the heart, particularly the RV and atria. Echocardiographic findings often include thickening of the RV and interatrial septum, and the diagnostic yield of RV endomyocardial biopsy is quite high, corroborating the diffuse nature of myocardial deposition in this disease. By using attenuation-corrected SPECT, we can go beyond global uptake described on planar imaging and localize uptake to various cardiac chambers. However, precise quantification of uptake outside of the LV is not feasible with SPECT given the spatial resolution and partial volume effect.
The SPECT quantification of TcPYP uptake may have clinical treatment implications. There are several pharmaceuticals in various stages of development for ATTR. As this tends to be a slowly progressive disease particularly in early stages, a surrogate imaging endpoint would be attractive. As a higher ASR of TcPYP uptake on SPECT imaging was found to portend an improved prognosis, this may be used to follow patients undergoing treatment. However, further studies must analyze longitudinal changes in SPECT quantification of radionuclide uptake, as changes in imaging parameters may not signal clinical improvement.
This study is limited by sample size, which is not uncommon in this rare disease. All patients were studied at our single institution, introducing the possibility of referral bias. Not all patients underwent endomyocardial biopsy for definitive confirmation of ATTR. However, all patients without endomyocardial biopsy were diagnosed based on noninvasive measures including semi-quantitative score 2 to 3, focal or diffuse myocardial uptake on SPECT images, and exclusion of AL amyloidosis. The ASR values described in this study are unlikely to apply to different imaging protocols with various incubation times.
SPECT analysis of TcPYP nuclear scintigraphy shows an apical-sparing pattern of uptake, mimicking the pattern seen with echocardiographic LS. A simple, reproducible ratio of count statistics in the basal and mid divided by the apical segments (ASR) can be used to quantify this pattern. More diffuse LV amyloid deposition as quantified by a greater percentage of apical uptake and a decreased ASR is associated with increased mortality, whereas planar and SPECT metrics of overall uptake are not.
COMPETENCY IN MEDICAL KNOWLEDGE 1: Technetium pyrophosphate nuclear scintigraphy shows a regional variation in uptake in patients with ATTR with sparing of the apical segments. This mimics the “apical sparing” pattern seen using echocardiographic LS analysis.
COMPETENCY IN MEDICAL KNOWLEDGE 2: An apical sparing pattern of technetium pyrophosphate uptake is associated with an improved prognosis in patients who have ATTR. More diffuse deposition is associated with a poorer prognosis.
TRANSLATIONAL OUTLOOK: Future studies should evaluate changes in SPECT quantification of radionuclide uptake over time in patients with ATTR.
Dr. Cerqueira is a consultant and on the speakers board at Astellas Pharma USA. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- light chain amyloidosis
- apical-sparing ratio
- transthyretin cardiac amyloidosis
- heart-to-contralateral lung
- longitudinal strain
- left ventricle
- single-photon emission computed tomography
- technetium 99m-pyrophosphate
- Received March 31, 2017.
- Revision received June 12, 2017.
- Accepted June 16, 2017.
- 2018 American College of Cardiology Foundation
- Phelan D.,
- Collier P.,
- Thavendiranathan P.,
- et al.
- Senapati A.,
- Sperry B.W.,
- Grodin J.L.,
- et al.
- Ternacle J.,
- Bodez D.,
- Guellich A.,
- et al.
- Gillmore J.D.,
- Maurer M.S.,
- Falk R.H.,
- et al.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Perugini E.,
- Guidalotti P.L.,
- Salvi F.,
- et al.
- Bokhari S.,
- Castaño A.,
- Pozniakoff T.,
- et al.
- Vranian M.N.,
- Sperry B.W.,
- Hanna M.,
- et al.
- ↵Dorbala S, Bokhari S, Miller EJ, et al. 99m Technetium-Pyrophosphate Imaging for Transthyretin Cardiac Amyloidosis. ASNC Pract Points. Available at: http://asnc.membershipsoftware.org//Files/Practice%20Resources/Practice%20Points/ASNC%20Practice%20Point-99mTechnetiumPyrophosphateImaging2016.pdf. Accessed March 31, 2017.
- Castaño A.,
- DeLuca A.,
- Weinberg R.,
- et al.
- Castano A.,
- Haq M.,
- Narotsky D.L.,
- et al.
- Sperry B.W.,
- Brunken R.,
- Jaber W.A.
- Glaudemans A.W.J.M.,
- van Rheenen R.W.J.,
- van den Berg M.P.,
- et al.