Background
Cellular rejection after cardiac transplantation is treatable with timely diagnosis. Because noninvasive methods for diagnosis are limited, surveillance endomyocardial biopsies are routinely performed in the first year after transplantation. The aim of this study was to test whether myocardial strain and strain rate as assessed by speckle-tracking echocardiography would be a sensitive noninvasive method for the detection of asymptomatic rejection.
Methods
Surveillance biopsies and echocardiograms obtained in the first year after transplantation were retrospectively reviewed, and patients with asymptomatic biopsy-proven cellular rejection were identified, as well as control transplantation patients without rejection or cardiac complications. Circumferential and longitudinal strain and strain rate were measured using Velocity Vector Imaging software from echocardiograms performed at three time points for patients with rejection—baseline (no rejection), rejection, and resolution (of rejection)—and three time points for control patients—baseline (within the first month after transplantation), 6 months, and 12 months after transplantation.
Results
Speckle-tracking strain and strain rate measurements were obtained from 30 patients with asymptomatic biopsy-proven rejection and 14 control transplantation patients. There were no significant differences in circumferential and longitudinal strain or strain rate between the baseline, rejection, and resolution studies. Furthermore, there were no significant differences in strain and strain rate in control transplantation patients during the first year after transplantation or compared with patients with rejection.
Conclusions
Speckle-tracking analysis was unable to detect changes on serial studies from patients with asymptomatic rejection and thus cannot replace biopsy. Other noninvasive methods for the diagnosis of cellular-mediated rejection are needed.
Cardiac transplantation is a lifesaving treatment option for select patients with end-stage heart failure. Cellular rejection remains a common but treatable complication of cardiac transplantation, particularly early after transplantation. Approximately 25% to 32% of transplantation patients experience some rejection in the early period (the first 12 months) after transplantation, but if rejection is diagnosed and treated in a timely fashion, patients can have excellent outcomes and normal cardiac function. Currently, there are no noninvasive methods for diagnosing rejection in the early period after transplantation; the gold standard is endomyocardial biopsy with histologic examination. Symptoms, physical examination findings, biomarkers such as troponin and brain natriuretic peptide, and traditional imaging assessments of cardiac function such as echocardiography and cardiac magnetic resonance imaging are not sensitive in adults and are often unchanged in early rejection, when timely treatment is indicated. As such, practice guidelines recommend a routine surveillance endomyocardial biopsy schedule in the first 12 months after transplantation to allow early diagnosis and treatment of rejection. Although biopsy can be performed safely at experienced centers, it is an expensive, invasive procedure that carries a small risk for adverse events along with anxiety and some discomfort for patients.
Because cellular rejection results in myocardial inflammation and varying amounts of myocyte damage at the cellular level, it should result in some degree of subclinical cardiac dysfunction. A number of studies have suggested that myocardial strain and strain rate are more sensitive measures than those derived from conventional echocardiography (such as ejection fraction) and thus may have a role in the detection of such early subclinical cardiac dysfunction. Several studies using tissue Doppler assessments of strain and strain rate suggest that these modalities may have a role in predicting cellular rejection. However, tissue Doppler is limited by image acquisition issues, rendering it poorly reproducible. By contrast, speckle-tracking echocardiography is not affected by the angle of acquisition and allows reliable and reproducible measurements. One animal study and two small clinical studies have suggested that strain and strain rate from speckle-tracking echocardiography may correlate with cellular rejection, but larger confirmatory studies in heart transplantation patients are needed.
Our goal was to study the utility of global longitudinal and circumferential strain and strain rate from speckle-tracking echocardiography in a large number of real-world cardiac transplantation patients within the first 12 months after transplant, the time period in which the risk for rejection is highest and scheduled surveillance endomyocardial biopsies are the standard of care. We hypothesized that changes in serial speckle-tracking echocardiographic parameters would be a sensitive noninvasive method for detection of asymptomatic early rejection.
Methods
Study Population
Medical records, surveillance endomyocardial biopsies, and echocardiographic studies were retrospectively reviewed from a single center’s cardiac transplantation program between January 2006 and December 2012. We limited our review of surveillance biopsies to those performed in the early period after cardiac transplantation, defined as within 12 months of transplantation. Patients with asymptomatic biopsy-proven cellular rejection were identified if they had no clinical signs or symptoms of rejection and had International Society of Heart and Lung Transplantation biopsy grades of 1R/2 (mild rejection) or 2R/3A (moderate rejection). In the case of multiple episodes of rejection, only the first episode of rejection was included. The few patients with biopsies showing severe pathologic grades of rejection (3R/3B or 3R/4) were not asymptomatic and were thus not included in the analysis.
By routine clinical practice at our center, most patients who have episodes of biopsy-proven rejection undergo echocardiography at the time of pathologic diagnosis to assess graft function and determine the type of treatment of the episode of rejection. Among those patients with asymptomatic biopsy-proven cellular rejection, we identified those who underwent baseline echocardiography before the episode of rejection, during their episodes of rejection, and after resolution of rejection, as confirmed by biopsy. Patients with asymptomatic biopsy-proven rejection who had echocardiograms available for all three time points were included in the analysis.
To assess for longitudinal changes in speckle-tracking parameters over the first year after transplantation, we identified a cohort of transplantation patients without rejection or other cardiac complications to serve as a control group. Echocardiograms were reviewed from this cohort of patients at baseline (within the first month after transplantation), 6 months after transplantation, and 12 months after transplantation.
Speckle-Tracking Echocardiographic Analysis of Strain and Strain Rate
Echocardiographic studies were analyzed using a two-dimensional tissue-tracking software package (Velocity Vector Imaging; Siemens Medical Systems, Mountain View, CA). For each study, the left ventricular endocardial border was manually traced in the best parasternal short-axis view at the level of the papillary muscles for calculation of circumferential strain and strain rate. To obtain valid results, adequate tracking and endocardial visualization in at least four of the six segments was required; otherwise, values from that view were discarded. Strain was defined as the greatest negative deflection before aortic valve closure on the strain curve. Systolic strain rate was defined as the greatest negative deflection on the global strain rate curve. Early diastolic strain rate was the first positive defection after aortic valve closure on the strain rate curve. The peak values of the six myocardial segments were averaged to provide a measurement of global circumferential strain, circumferential systolic strain rate, and circumferential diastolic strain rate ( Figure 1 ). The left ventricular endocardial border was manually traced in the best apical four-chamber and apical two-chamber views for calculation of longitudinal strain and strain rate. The peak values of the six segments in each view were averaged to provide a measurement of global longitudinal strain, longitudinal systolic strain rate, and longitudinal diastolic strain rate ( Figure 2 ).
Reproducibility Measurements
Interobserver variability was assessed in 30 randomly selected studies from 10 patients with speckle-tracking analysis independently performed by two investigators. Intraobserver reproducibility was assessed by blindly repeating speckle-tracking analysis from 30 randomly selected studies from 10 patients by the same investigator 12 weeks after the initial analysis.
Statistical Analysis
Data are reported as mean ± SD. Paired t tests were used to compare values; the time periods of baseline versus rejection and rejection versus resolution were examined because these are the comparisons traditionally made with endomyocardial biopsy. A two-tailed P value < .05 was considered to indicate statistical significance. A prespecified subgroup analysis was performed stratified by rejection severity (mild or moderate rejection). Bland-Altman plots were created to visualize the degree of variation, and intraclass correlation coefficients (ICCs) were calculated. Paired t tests were also used to analyze values in the control cohort of patients, and comparisons were made between the time periods of baseline versus 6 months after transplantation, baseline versus 12 months after transplantation, and 6 versus 12 months after transplantation. One-way analyses of variance were used for comparisons of rejection strain and strain rate measures compared with control measures at the three time points.
The magnitude of effect in speckle-tracking echocardiographic measures varies depending on the disease states and parameters measured, with longitudinal strain being a consistently measured and comparable parameter in most studies. In comparing different conditions within transplantation patients specifically, the largest prior speckle-tracking echocardiographic study found that global longitudinal strain was >40% lower in a cohort of transplantation nonsurvivors compared with transplantation survivors (−7.9% vs −13.7%). On the basis of the effect size from this prior study and a significance level of α < 0.05, we initially estimated that we would need a total sample size of 14 patients to have 80% power to detect a similarly sized reduction in global longitudinal strain during an episode of rejection. A post hoc power analysis was also performed on the basis of the data obtained in this study. Furthermore, using the data obtained in our study, we calculated the necessary sample size for 80% power and a significance level of α < 0.05 to detect differences in longitudinal strain of varying magnitudes for future studies using this speckle-tracking echocardiographic measure as previously described.
Results
Surveillance endomyocardial biopsies were reviewed from 98 consecutive transplantation patients. Episodes of asymptomatic biopsy-proven rejection were identified in 44 individual patients. Of these 44 patients, 33 had echocardiograms available for analysis for all three time points (baseline, rejection, and resolution). Three patients had poor endocardial visualization on one or more of their echocardiographic studies and were excluded, resulting in a technical feasibility rate of 91% for measuring speckle-tracking parameters on previously obtained echocardiographic images. Thus, 30 patients were included in the analysis cohort. Baseline patient characteristics are noted in Table 1 . Because these patients were in the early period after transplantation, immunosuppression consisted of triple therapy with corticosteroids, a calcineurin inhibitor, and an antiproliferative agent.
| Characteristic | Value |
|---|---|
| Age (y) | 44.4 ± 13.8 |
| Male gender | 23 (77%) |
| Female gender | 7 (23%) |
| Caucasian | 21 (70%) |
| African American | 5 (17%) |
| Hispanic | 3 (10%) |
| Asian American | 1 (3%) |
| Immunosuppression medications | |
| Prednisone | 30 (100%) |
| Cyclosporine | 13 (43%) |
| Tacrolimus | 17 (57%) |
| Mycophenolate mofetil | 29 (97%) |
| Azathioprine | 1 (3%) |
Eighteen control transplantation patients were identified who (1) did not have episodes of rejection in the first year after transplantation, (2) had no other cardiac complications in the first year after transplantation, and (3) had echocardiograms available for analysis for all three time points (baseline, 6 months after transplantation, and 12 months after transplantation). Of these 18 patients, four had poor endocardial visualization on one or more of their echocardiographic studies and were excluded, leaving a total of 14 patients in the control analysis cohort.
There were no significant differences in global circumferential strain, systolic strain rate, and early diastolic strain rate between the baseline, rejection, and resolution echocardiographic studies ( Figure 3 , Table 2 ). Similarly, there were no differences in global longitudinal strain, systolic strain rate, and diastolic strain rate between the baseline, rejection, and resolution studies ( Figure 3 , Table 2 ).
| Variable | Baseline | Rejection | Resolution ∗ |
|---|---|---|---|
| Global circumferential strain (%) | −22.8 ± 6.0 | −21.0 ± 6.1 | −20.0 ± 6.4 |
| Circumferential systolic strain rate (%/sec) | −1.7 ± 0.5 | −1.6 ± 0.5 | −1.5 ± 0.6 |
| Circumferential diastolic strain rate (%/sec) | 2.2 ± 0.6 | 1.9 ± 0.7 | 1.9 ± 0.8 |
| Global longitudinal strain (%) | −12.0 ± 2.5 | −12.6 ± 2.2 | −13.3 ± 3.1 |
| Longitudinal systolic strain rate (%/sec) | −1.0 ± 0.2 | −0.9 ± 0.2 | −1.0 ± 0.2 |
| Longitudinal diastolic strain rate (%/sec) | 1.1 ± 0.3 | 1.1 ± 0.3 | 1.1 ± 0.3 |
∗ P values were not significant for paired t -test comparisons of baseline versus rejection and rejection versus resolution values. More negative values indication better or greater strain and systolic strain rate, while more positive values indicate better or greater diastolic strain rate.
In the prespecified subgroup analysis, there were also no differences in any of the speckle-tracking echocardiographic parameters when limiting the analysis to those patients with moderate rejection (16 patients) versus mild rejection (14 patients) over the course of rejection from baseline to rejection to resolution, as shown in Figure 4 . Similarly, there were no differences at individual time points between speckle-tracking measurements in those with moderate versus mild rejection.
Longitudinal speckle-tracking measurements from nonrejecting control transplantation patients during the first year after transplantation are shown in Table 3 . There were no significant differences in global circumferential and longitudinal strain, systolic strain rate, and diastolic strain rate between the measurements at baseline, 6 months after transplantation, and 12 months after transplantation. In addition, there were no significant differences between any of these control time point measurements and measurements obtained from patients during the episode of rejection.
