Added Prognostic Value of Visually Estimated Coronary Artery Calcium Among Heart Transplant Recipients

Abstract

Background:

Cardiac hybrid positron emission tomography/computed tomography (PET/CT) has become a valid screening modality for cardiac allograft vasculopathy (CAV) following HT. Visually estimated coronary artery calcium (VECAC) can be quantified from CT images obtained as part of PET/CT and has been shown to be associated with adverse cardiovascular outcomes in coronary artery disease. We investigated the prognostic value of VECAC following heart transplantation (HT).

Methods:

A retrospective analysis of 430 consecutive adult HT patients who underwent ¹³N-ammonia cardiac PET/CT from 2016–2019 with follow-up through October 15, 2022, was performed. VECAC categories included: VECAC 0, VECAC 1–9, VECAC 10–99, and VECAC 100+. The association between VECAC categories and outcomes was assessed using univariable and multivariable proportional hazards regression. The primary outcome was death/retransplantation.

Results:

The cohort was 73% male, 33% had diabetes, 67% had estimated glomerular filtration rate<60 mL/min, median age was 61 years, and median time since HT was 7.5 years. VECAC alone was insufficiently sensitive to screen for CAV. During a median follow-up of 4.2 years ninety patients experienced death or retransplantation. Compared with those with VECAC 0, patients VECAC 10–99 (HR 2.25, 95% CI 1.23–4.14, p=0.009) and VECAC 100+ (HR 3.42, 95% CI 1.96–5.99, p<0.001) experienced an increased risk of death/retransplantation. The association was similar for cardiovascular death and cardiovascular hospitalization. After adjusting for other predictors of death/retransplantation, VECAC 10–99 (VECAC 10–99: aHR 1.95, 95% CI 1.03–3.71 p=0.04) and VECAC 100+ (VECAC 100+: aHR 2.33, 95% CI 1.17–4.63, p=0.02) remained independently associated with death/retransplantation.

Conclusions:

VECAC is an independent prognostic marker of death/retransplantation following HT and merits inclusion as a part of post-HT surveillance PET/CT.

Introduction

Heart transplantation (HT) is a life-saving therapy for patients with end-stage heart failure and post-transplant survival has improved by more than 50% since the 1980s.(1) However, long term survival is often impaired by the development and progression of cardiac allograft vasculopathy (CAV). CAV is characterized by diffuse intimal proliferation, luminal narrowing, and fibrotic changes within the coronary vasculature of the transplanted heart leading to decreased myocardial blood flow, restrictive physiology, and ultimately graft loss. CAV is common among heart transplant recipients, with 30% to 45% of patients having some degree of epicardial CAV by 5 years following HT and 50% to 65% at 10 years.(2,3) Effective surveillance and identification of CAV is important for guiding clinical care and optimizing outcomes. Cardiac hybrid positron emission tomography/computed tomography (PET/CT) imaging has proven to be a useful surveillance modality for assessing CAV due to its ability to provide diagnostic and prognostic information through measures of perfusion and quantification of myocardial blood flow.(4–6)

Quantification of coronary artery calcium has been clinically used for over three decades(7) and a consistent association between coronary artery calcium and adverse cardiac outcomes has been demonstrated in large population studies.(8–11) Similarly, the presence or absence of coronary calcium can reclassify risk profiles.(9,12–16) Atherosclerotic calcification begins as microcalcifications that coalesce into larger fragments parallel with atherosclerotic plaque progression.(17) While cardiac allograft vasculopathy has a distinct pathophysiology, coronary artery calcification is present in more advanced CAV and is associated with CAV progression.(18–20)

While a precise coronary artery calcium score determined using the method of Agatston et al. (7) requires a diagnostic-quality, electrocardiographically-gated CT scan obtained with breath holding, considerable information about coronary artery calcium burden can be obtained from other types of CT scans, including CT attenuation correction scans performed as part of PET/CT myocardial perfusion imaging. Several studies have demonstrated strong correlation of measures of calcium burden from such other types of CT scans with Agatston scores from dedicated calcium scoring scans.(21–23) The quantification of visually estimated coronary artery calcium (VECAC) has been applied to and has demonstrated utility in predicting adverse cardiovascular events for native coronary artery disease (CAD).(24–26) However, the application of VECAC to CAV following HT has not been thoroughly investigated. This study sought to evaluate the prognostic significance of coronary artery calcification through the assessment of VECAC in HT recipients.

Methods

In 2016, our center changed the CAV screening protocol to biennial coronary angiography, alternating with 13N-Ammonia PET with MPI. This observational retrospective cohort study included all adult HT patients who underwent ¹³N-Ammonia PET/CT with MPI between June 2016 and December 2018. Patients with prior percutaneous coronary intervention with stent placement were excluded. Similarly, for patients with repeat PET/CT scans, only the initial scan was included. Final study follow-up date was October 15, 2022. Patients were classified into four groups based on the VECAC: 0 (none), 1–9 (minimal), 10–99 (mild), and 100 or greater (moderate or greater).

Study Outcomes:

The primary study endpoint was death or retransplantation. Major secondary endpoints included cardiovascular death or retransplantation and cardiovascular hospitalization, which involved heart failure hospitalization, acute coronary syndrome, percutaneous coronary intervention, and acute rejection. Endpoints were ascertained through review of the medical record.

Cardiac PET/CT Protocol:

PET myocardial perfusion imaging (MPI) was performed using Siemens mCT 64-slice PET/CT scanners. For rest images, 8 to 12 mCi of ¹³N-Ammonia was injected intravenously before acquisition of cardiac perfusion images. For the stress portion of the examination, performed after a delay of 50 minutes (5 half-lives) from rest imaging, patients underwent pharmacological stress with dipyridamole (0.56 mg/kg), adenosine (140 μg/kg/min), or regadenoson (0.4 mg), and received another 8 to 12 mCi of ¹³N-Ammonia. After all images were obtained, the reconstructed perfusion images were analyzed using 4DM v2018 software (INVIA, Ann Arbor, MI). Myocardial ischemia or infarction was assessed using described semiquantitative assessment of a 17-segment model by board-certified nuclear cardiologists experienced with PET/CT.(27) Summed rest, summed stress, and summed difference (stress-rest) scores were calculated. Stress and rest left ventricular ejection fraction was measured from gated images. Myocardial blood flow (MBF; in milliliters per minute per gram of myocardial tissue) was calculated at rest and stress using a one-tissue compartment model (INVIA N-13 ROI 1:1). MBF reserve (MBFR) was calculated as the ratio of stress to rest myocardial blood flow adjusted for rate pressure product using a reference value of 9,000. A non-ECG-gated, free-breathing low-dose computed tomography (120 kVp, 25 effective mAs [reflecting average tube current of 75 mA, gantry rotation time of 0.5 s, and pitch of 1.5], 23 cm scan length, 1.8 mGy CTDIvol, slice thickness 3 mm, slice spacing 2 mm [slices overlap]) was used for attenuation correction and used as well for visual estimation of coronary artery calcium. Briefly, VECAC was graded on a 6-level scale, classifying patients as having an estimated Agatston score of 0, 1 to 9, 10 to 99, 100 to 300, 400 to 999, or ≥1,000 on the aforementioned attenuation correction low-dose CT. Due to the distribution of VECAC scores in the study cohort, scores greater than 100 were grouped. This technique has been previously demonstrated to have robust accuracy when compared to coronary artery calcium scores obtained from ECG gated scans across multiple different centers,(24,25,28) and was recommended in the 2016 SCCT/STR guidelines.(29) VECAC scores were assigned for each study by the same nuclear cardiologist who read the PET perfusion imaging. Average radiation effective dose for each PET study was 3.2 mSv, including 0.7 mSv from each of the two the attenuation correction CT scans, calculated using a Monte Carlo dosimetry software, Radimetrics (Bayer HealthCare, Berlin, Germany), and 0.9 mSv from each of the two 12 mCi ¹³N-Ammonia PET.(30) This is similar to the annual background radiation in the United States, 3 mSv.(31) PET CAV was graded according to a previously validated algorithm (Supplemental Table 1).(4,32)

Coronary Angiography Protocol:

All patients underwent diagnostic coronary angiography in the usual fashion through radial or femoral arterial access. Following angiography of all epicardial vessels, degree of stenosis was determined by the performing board-certified interventional cardiologist. Angiographic CAV was graded using the ISHLT criteria.(33) No patients met the restrictive physiology criteria set by the 2010 ISHLT Guidelines. Angiograms were included if they were performed within 3 months before the PET or in the subsequent 12 months to maximize the number of patients with contemporary angiographic data (median time between studies: 328 days).

Statistical Analysis:

Demographic and clinical variables were expressed as median (IQR) for continuous variables and count (%) for categorical variables. Group comparisons were made with Chi-squared, Fisher exact, and Mann-Whitney U test when appropriate. Kaplan-Meier survival analysis and Cox proportional hazards regression were performed to compare outcomes between groups of interest. Cardiovascular mortality was assessed using a Fine and Gray subdistribution hazard model to account for the competing risk of non-CV mortality. A multivariable Cox model was generated including variables with a P-value < 0.25 in univariate analysis. The proportional hazards assumption was met. To assess the clinical utility of VECAC past other clinical and imaging predictors, the C-statistic of the multivariable model including and excluding VECAC were compared. Analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina). The study was approved by the Columbia University Irving Medical Center Institutional Review Board. This study is in compliance with the ISHLT ethics statement.

Results

The study cohort included 430 consecutive adult HT recipients who underwent ¹³N-ammonia cardiac PET/CT imaging between June 2016 and December 2018. Baseline characteristics are presented in Table 1. Coronary artery calcium was present in 111 patients (25.8%), with a nearly even distribution of VECAC 1–9, VECAC 10–99, and VECAC 100+. The cohort was predominantly male, and median age was 53 at the time of transplant. Patients with coronary artery calcium were further from transplantation, had a greater prevalence of diabetes mellitus, had a prior stroke, and more commonly had chronic kidney disease. The prevalence of immunologic events (antibody mediated rejection, acute cellular rejection, or donor-specific antibodies) was similar. Angiographic CAV was more common in the group with coronary artery calcium (71.1% vs. 40.0%, p<0.0001). Additionally, individuals with coronary artery calcium had a lower median myocardial blood flow compared with the group with VEAC 0 (MBFR 2.24 [1.77–2.61] vs. 2.43 [2.00–2.82], p=0.005). Lastly, the donors for patients with coronary artery calcium were older and more commonly had diabetes, hypertension, and smoked.

Table 1.

Baseline Characteristics

	All Patients	VECAC 0	VECAC 1+	p-value
Number of patients	430	319	111
Age at inclusion PET/CT scan, y	61 (48–69)	60 (47–68)	65 (55–70)	0.003
Age at transplant, y	53.5 (39.3–61.6)	53.7 (39.1–61.7)	54.7 (43.2–62.4)	0.31
Follow Up Time from PET/CT scan, y	4.3 (3.4–5.1)	4.3 (3.6–5.2)	4.1 (3.2–5.1)	0.14
Female	114 (26.5%)	80 (25.1%)	34 (30.6%)	0.25
BMI	26.5 (23.6–29.1)	26.7 (23.6–29.1)	25.3 (23.8–29.5)	0.51
Ejection fraction	63 (58–68)	63 (58–68)	64 (59–68)	0.45
LDL, mg/dL	83 (66–100)	84 (69–102)	81.5 (59–98)	0.15
Immunosuppression
CNI	414 (96.3%)	307 (96.2%)	107 (96.4%)	0.94
Proliferation Signal Inhibitor	116 (27.0%)	85 (26.7%)	31 (27.9%)	0.79
Antimetabolite	270 (62.8%)	204 (64.0%)	66 (59.5%)	0.40
Medication
ACE inhibitor/ARB/ARNI	143 (33.3%)	106 (33.2%)	37 (33.3%)	0.98
Calcium-channel blocker	173 (40.2%)	136 (42.6%)	37 (33.3%)	0.09
Statin	366 (85.1%)	275 (86.2%)	91 (82.0%)	0.28
Aspirin	395 (92.1%)	300 (94.3%)	95 (85.6%)	0.003
Other Lipid Lower Agent*	106 (24.7%)	81 (25.4%)	25 (22.5%)	0.55
HTN	290 (67.4%)	220 (69.0%)	70 (63.1%)	0.25
Diabetes mellitus	141 (32.8%)	96 (30.1%)	45 (40.5%)	0.04
Insulin-dependent diabetes mellitus	114 (26.5%)	76 (23.8%)	39 (34.2%)	0.03
Prior Stroke	20 (4.7%)	9 (2.8%)	11 (9.9%)	0.002
CKD				0.004
eGFR > 60 mL/min	143 (33.3%)	115 (36.1%)	28 (25.3%)
eGFR 30 – 60 mL/min	201 (46.7%)	151 (47.3%)	50 (45.1%)
eGFR < 30 mL/min	64 (14.9%)	43 (13.5%)	21 (18.9%)
ESRD	22 (5.1%)	10 (3.1%)	12 (10.8%)
Prior ACR	166 (38.6%)	123 (38.6%)	43 (38.7%)	0.87
ISHLT 2R or greater	61 (14.2%)	43 (13.5%)	18 (16.2%)	0.48
Prior AMR	40 (9.4%)	28 (8.8%)	12 (10.8%)	0.54
DSA	153 (35.8%)	115 (36.4%)	38 (34.2%)	0.68
ISHLT CAV (n=371)				<0.0001
CAV 0	195 (52.6%)	169 (60.1%)	26 (28.9%)
CAV 1	166 (44.7%)	108 (38.4%)	58 (64.4%)
CAV 2–3	10 (3.1%)	4 (1.6%)	6 (6.7%)
MBFR	2.35 (1.85–2.76)	2.43 (2.0–2.82)	2.24 (1.77–2.61)	0.005
Stress myocardial blood flow	2.08 (1.62–2.51)	2.13 (1.74–2.59)	1.98 (1.50–2.33)	0.002
PET MPI Derived Ischemia	36 (8.4%)	27 (8.5%)	9 (8.1%)	0.91
Coronary Calcium
VECAC0		313 (100%)
VECAC 1–9			44 (39.6%)
VECAC 10–99			34 (30.6%)
VECAC 100+			33 (29.7%)
Donor Age, y	30.0 (22.1–40.4)	29.0 (22.0–40.0)	34.5 (23.6–47.0)	0.01
Donor Diabetes Mellitus	17 (4.0%)	8 (2.5%)	9 (8.1%)	0.008
Donor HTN	69 (16.0%)	42 (13.2%)	27 (24.3%)	0.005
Donor Cigarette Use	94 (21.9%)	54 (16.9%)	40 (36.0%)	<0.0001
Donor Cocaine Use	76 (17.7%)	56 (17.6%)	20 (18.0%)	0.86

ACE=Angiotensin converting enzyme, ACR=Acute cellular rejection, AMR=Antibody mediated rejection, ARB=Angiotensin receptor blocker, ARNI=Angiotensin receptor blocker neprilysin inhibitor, BMI=Body mass index, CAV=Cardiac allograft vasculopathy, CKD=Chronic kidney disease, CNI=Calcineurin inhibitor, DSA=Donor specific antibodies, eGFR=Estimated glomerular filtration rate calculated by the CKD-EPI creatinine equation (2021), ISHLT=International Society for Heart and Lung Transplantation, LDL=Low density lipoprotein, MBFR=Myocardial blood flow reserve, MPI=Myocardial perfusion imaging, PET=Positron emission tomography

^*Other lipid lowering agents included ezetimibe, proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, omega-3 fatty acids, and fibrates.

Coronary artery calcification was more prevalent and present in an increasing amount as the angiographic CAV grade increased (Figure 1). This was similar, though to a lesser degree, with PET CAV grading (Figure 2). However, VECAC did not demonstrate an ability to alone serve as a screening test for CAV—among patients with CAV on angiography (CAV 1 or greater), the sensitivity of VECAC was only 36.0% and the negative predictive value was only 64.9% (prevalence 44%, Supplemental Table 2). Even among patients with more significant CAV (CAV 2 or 3), the sensitivity was only 60% and the negative predictive value was acceptable at 98.8%, though was influenced by a low prevalence of CAV 2 or 3 (2.3%).

Figure 1.

Assessment of VECAC is possible during PET/CT screening for CAV. The prevalence and amount of VECAC increases as the grade of CAV increases. However, VECAC alone is insufficiently sensitive to screen for CAV. The presence of at least mild coronary artery calcification (VECAC 10+) was associated with an increased risk of death or retransplantation.

Figure 2.

Prevalence of VECAC categories for (A) angiographic ISHLT CAV grades and (B) cardiac PET CAV grades.

During a median follow-up of 4.2 years ninety patients experienced death or retransplantation and VECAC was associated with adverse events. When compared to patients with zero coronary artery calcium (VECAC 0), those with minimal coronary artery calcium (VECAC 1–9) had a similar risk of death or retransplantation (HR 0.82, 95% CI 0.35–1.91, p=0.64). Among those with a greater burden of coronary artery calcium, outcomes were worse. Patients with VECAC of 10–99 had more than twice the risk of the death or retransplantation (HR 2.25, 95% CI 1.23–4.14, p=0.009) when compared with those without coronary artery calcium. The risk was even more pronounced among those with VECAC 100+ who experienced a more than threefold greater risk of death or retransplantation (HR 3.42, 95% CI 1.96–5.99, p<0.001, Figure 3). This remained consistent when stratified by the presence of angiographic CAV, reduced myocardial blood flow reserve (MBFR<2.0), chronic kidney disease (glomerular filtration rate< 60 mL/min), and time since transplant (Table 2). There were no significant interactions between VECAC and the variables in Table 2 (Supplemental Table 3).

Figure 3.

Survival curves demonstrate similar outcomes for individuals with no (VECAC 0) or minimal (VECAC 1–9) coronary artery calcium (Dunnett p-value 0.20). Individuals with more coronary artery calcification had worse outcomes with increasing VECAC scores.

Table 2.

Sub-group univariable analysis demonstrating a consistent association between a greater burden of coronary calcium and death/retransplantation. Reference group is VECAC=0.

Sub-Group		HR	95% CI	p-value
MBFR ≤2	VECAC 1–9	0.54	0.13–2.27	0.40
	VECAC 10–99	1.81	0.70–4.67	0.22
	VECAC 100+	3.07	1.27–7.38	0.01
MBFR >2	VECAC 1–9	0.91	0.32–2.23	0.40
	VECAC 10–99	2.28	1.01–5.30	0.046
	VECAC 100+	2.75	1.31–5.77	0.007
CAV 0	VECAC 1–9	0.86	0.03–2.82	0.81
	VECAC 10–99	2.55	1.12–5.82	0.03
	VECAC 100+	4.91	2.06–11.73	0.0003
CAV 1–3	VECAC 1–9	0.78	0.23–2.63	0.68
	VECAC 10–99	1.92	0.77–4.82	0.16
	VECAC 100+	2.83	1.31–6.11	0.008
Diabetes	VECAC 1–9	0.25	0.03–1.81	0.17
	VECAC 10–99	1.79	0.70–4.60	0.23
	VECAC 100+	2.42	0.94–5.38	0.07
No Diabetes	VECAC 1–9	1.42	0.54–3.77	0.48
	VECAC 10–99	2.33	1.00–5.48	0.05
	VECAC 100+	5.01	2.35–10.66	0.0001
GFR<60	VECAC 1–9	0.75	0.10–5.68	0.78
	VECAC 10–99	1.94	0.44–8.59	0.38
	VECAC 100+	2.77	0.63–12.22	0.18
GFR≥60	VECAC 1–9	0.79	0.31–2.01	0.62
	VECAC 10–99	2.10	1.08–4.12	0.03
	VECAC 100+	3.28	1.78–6.04	0.0001
>Median Time Since Transplant	VECAC 1–9	1.02	0.34–3.04	0.97
	VECAC 10–99	3.75	1.67–8.45	0.001
	VECAC 100+	4.36	2.08–9.16	<0.0001
<Median Time Since Transplant	VECAC 1–9	0.64	0.15–2.65	0.54
	VECAC 10–99	1.20	0.43–3.39	0.73
	VECAC 100+	2.28	0.81–6.41	0.12

CAV=Cardiac allograft vasculopathy; GFR=Glomerular filtration rate; MBFR=Myocardial blood flow reserve.

The incremental implication of coronary artery calcification was consistent across both secondary endpoints. Again, there was no significant difference between VECAC 0 and VECAC 1–9, however patients with a VECAC of 10–99 (HR 2.7, 95% CI 1.40–5.32, p=0.003) and VECAC of 100+ (HR 2.66, 95% CI 2.66–5.49 p=0.008) both had twice the risk of cardiovascular hospitalization. There was a similar graded effect seen for cardiovascular death where patients with at least mild coronary calcium had twice the risk of cardiovascular death (VECAC 10–99: sHR 2.02, 95% CI 0.83–4.94 p=0.12; VECAC 100+: sHR 2.61, 95% CI 1.12–6.08, p=0.03).

In univariable analysis, a MBFR less than 2.0, diabetes mellitus, stress ejection fraction, prior stroke, chronic kidney disease, donor specific antibodies, angiographic ISHLT CAV 2 or 3, ischemia on the PET MPI (summed difference score>2), not taking a statin, and the time since transplant were identified as risk factors for death or retransplantation (Table 3). After adjusting for these other significant predictors in a multivariable model, a greater burden of coronary artery calcium remained independently associated with the primary composite endpoint of death or retransplantation (VECAC 10–99: aHR 1.95, 95% CI 1.03–3.71 p=0.04; VECAC 100+: aHR 2.33, 95% CI 1.17–4.63, p=0.02).

Table 3:

Univariable assessment for associations with death/retransplant.

	Hazard Ratio	95% Confidence Interval	p-value
Any VECAC	1.97	1.29–3.02	0.002
VECAC 1–9 *	0.82	0.35–1.92	0.65
VECAC 10–99 *	2.26	1.23–4.15	0.01
VECAC 100+ *	3.43	1.96–6.01	<0.0001
MBFR ≤2	3.05	2.07–4.51	<0.0001
Female	0.81	0.51–1.28	0.37
Age	1.00	0.99–1.02	0.63
BMI	1.00	0.95–1.06	0.92
Diabetes Mellitus	2.20	1.49–3.24	<0.0001
Stress EF	0.97	0.96–0.99	0.0003
Stroke	2.27	1.18–4.36	0.01
Chronic kidney disease	1.74	1.09–2.77	0.02
Acute cellular rejection	1.12	0.66–1.92	0.67
Antibody mediated rejection	1.32	0.72–2.41	0.37
Donor specific antibodies	1.43	0.96–2.12	0.08
Proliferation signal inhibitor	0.86	0.55–1.34	0.50
Aspirin	0.76	0.41–1.43	0.40
Statin	0.60	0.37–0.97	0.04
ISHLT CAV2 or CAV3	3.67	2.05–6.58	<0.0001
PET derived ischemia	2.29	1.35–3.89	0.002
Time since transplant	1.00	1.00–1.01	0.02
Donor Diabetes Mellitus	1.77	0.77–4.05	0.18
Donor Hypertension	1.29	0.77–2.17	0.32
Donor Cigarette Use	0.85	0.43–1.69	0.63
Donor Cocaine Use	1.44	0.80–2.61	0.26
Donor Age	0.99	0.98–1.02	0.95

^*Reference is VECAC 0

Coronary artery calcium assists in cardiovascular risk-stratification in atherosclerosis, and it demonstrated a similar ability for CAV as well. Figure 4 demonstrates the differences in 4-year survival free from retransplant across various degrees of CAV (defined by angiography [Figure 4A] and cardiac PET/CT [Figure 4B]) and the severity of coronary artery calcification. For instance, patients with CAV 0 or PET CAV 0 but a VECAC of >10 had a 4-year survival that was similar to patients with CAV 3 or PET CAV 3 and a VECAC of 0. Alternatively, the added valued of VECAC was demonstrated by the C-statistic of a multivariable model including all variables listed in Table 4, but making VECAC binary (VECAC≥10) being 9% greater when VECAC was included than when it was omitted (C-statistic 0.75 [95% CI 0.67–0.83] vs. C-statistic 0.69 [95% CI 0.64–0.75], p=0.03). Likewise, the model with VECAC was superior as assess by Akaike’s Information Criterion (964.4 vs. 979.1) and Schwartz’s Bayesian Criterion (989.4 vs. 999.1). Among patients with PET CAV, coronary calcium was a non-significant risk enhancer for patients with PET CAV 0 (HR 1.42, 95% CI 0.78–2.59, p=0.25), however it helped identify individuals with PET CAV 1 that carried a higher risk (HR 2.25, 95% CI 1.14–4.42, p=0.02). A similar trend was seen with angiographic CAV.

Figure 4.

4-year survival by strata of VECAC score and CAV as assessed by (A) angiography and (B) cardiac PET scan.

Table 4.

Multivariable assessment demonstrating the association between coronary calcium and death/retransplant persists.

	Hazard Ratio	95% Confidence Interval	p-value
MBFR ≤2	1.93	1.23–3.04	0.005
Time since transplant	1.00	1.00–1.00	0.52
Diabetes Mellitus	1.87	1.20–2.91	0.006
ISHLT CAV2 or CAV3	3.44	1.31–9.06	0.01
Stress EF	1.00	0.97–1.02	0.88
Statin	0.58	0.33–1.04	0.07
PET derived ischemia	2.60	1.42–4.75	0.002
Chronic kidney disease	1.41	0.80–2.46	0.23
Stroke	1.74	0.79–3.83	0.17
VECAC 1–9 *	0.66	0.28–1.58	0.35
VECAC 10–99 *	1.95	1.03–3.71	0.04
VECAC 100+ *	2.33	1.17–4.63	0.02

^*Reference is VECAC 0

Discussion

Coronary artery calcification is an imaging finding that may be used to assist in cardiovascular risk stratification. In this cohort study of adult heart transplant recipients, we sought to assess if VECAC obtained from a cardiac PET/CT scan had a role in CAV risk assessment. We found that: 1) The presence of coronary artery calcium was more common with more advanced CAV but was not reliable alone as a screening test for CAV in this cohort. 2) At least mild coronary artery calcification (VECAC 10+) was able to identify patients at an increased risk of death or retransplantation, cardiovascular death, and cardiovascular hospitalization. 3) At least mild coronary artery calcification enhances the risk of death/retransplant over the next four years beyond standard CAV grading.

To date, several studies have demonstrated the ability to identify prognostic CAV. The 2010 ISHLT angiographic grading scale has been demonstrated to effectively identify individuals with an increased risk of future events.(3,34) Numerous intravascular imaging studies have demonstrated that the additional information obtained through intravascular ultrasound and optical coherence tomography is associated with future adverse events.(18,35–40) And more recently, physiologic parameters of coronary and myocardial blood flow, both invasive and noninvasive, have been shown to provide prognostic information.(4–6,41) In this cohort, we found that coronary artery calcification (VECAC≥10) also is a CAV assessment tool that provides prognostic information.

In the practice of medicine, diagnostic tests provide information that helps clinical decision making and guide care. Within the field of cardiology, there has been an effort to glean more information from diagnostic tests (e.g. electrocardiogram, chest x-ray) that are already performed and potentially obviate the need for additional testing.(42,43) While non-invasive screening for CAV is not uniform around the world(44), the visual estimation of coronary artery calcification does not require dedicated cardiac imaging and is a possible way to leverage imaging being performed for other purposes for CAV assessment. With this in mind, we sought to explore what role the visual estimation of coronary artery calcium has in the diagnosis and management of heart transplant recipients. A prior study found that in a cohort of 133 heart transplant recipients coronary artery calcium score had a sensitivity of 88% and negative predictive value of 97% for CAV 2–3 (CAV 2–3 prevalence 12%).(45) Unfortunately, in the current cohort with nearly 300 more patients, the presence of coronary artery calcium did not display a sensitivity or negative predictive value that would make it a sensible screening test for CAV. This was not unexpected, knowing that prior studies demonstrated coronary artery calcification occurs later in the evolution CAV than in typical atherosclerosis.(18–20) Nevertheless, coronary artery calcification did demonstrate utility.

Coronary artery calcium’s role as a risk modifier is well established in atherosclerosis.(14–16,46) In this cohort, we found that it appears to function in a similar way for CAV. In this cohort, an individual without angiographic CAV (CAV 0) but a VECAC≥10 had a four-year survival that was akin to individuals with moderate to severe CAV (CAV 2–3) and no coronary artery calcification. The same was seen when non-invasive PET CAV grading was utilized. Figure 3 demonstrates that for CAV assessed by PET/CT and by angiography, the presence of at least mild coronary artery calcification enhanced the risk at each grade. The reason for this warrants further examination. An increased burden of comorbid conditions and older age appeared possible based on the baseline differences, however coronary artery calcification remained an independent predictor of death/retransplantation after adjusting for comorbidities. Additionally, the risk for overall mortality was similar for cardiovascular specific mortality. It also may be that coronary artery calcification may simply be a visual marker of more advanced, diffuse CAV in both the epicardial coronary arteries and the microcirculation. More patients with coronary calcification had CAV and CAV 2–3 remains the strongest predictor of future adverse events. Further, patients with coronary artery calcification had lower MBFR and evidence of microvascular dysfunction (normal perfusion, but reduced MBFR). Finally, coronary artery calcium in this population may represent individuals who have developed atherosclerotic lesions in the transplanted heart. In a study utilizing optical coherence tomography, features typical of atherosclerosis became increasingly prevalent further from transplant.(47)

The 2010 study that demonstrated the high agreement between VECAC assessed by free breathing, low-dose, non-ECG-gated CT attenuation correction scans and coronary artery calcium as measured by Agatston score exhibited a high weighted kappa of 0.89.(24) However, in that study the greatest source of error occurred in the VECAC 1–9 group, where the actual Agatston score was 0 in 36% and was ≥10 44% of the time. The attenuation correction CT scans in our study were non-ECG-gated and free-breathing, which may have limited the accuracy of identifying VECAC 1–9 and potentially limited the strength of the finding in that group. In this way, it is possible that an Agatston score of 1–9 obtained from an end-inspiratory breath hold, ECG-gated CT scan may have prognostic information for heart transplant recipients.

There are limitations to our study that must be acknowledged. The first is that this is a single center, non-randomized cohort study of consecutive patients who underwent a PET/CT scan. It is possible that the findings may differ among different populations. Secondly, these findings may differ among attenuation correction CT scanning protocols – we used a low tube current (averaging 75 mA), however other centers use the minimal possible tube current which results in adequate image quality for purposes of attenuation correction and lower radiation dose, but noisier images that render assessment of coronary calcification and other findings more challenging. Thus, our results may not be generalizable to minimum-dose CT attenuation correction scans, which would require further validation. Thirdly, while the study included 430 heart transplant recipients and is the largest to evaluate the role of coronary artery calcification in a transplant population, the sample size remains modest when compared with studies of coronary artery calcification in non-transplant populations. Since this was an observational study, patients with clinically more severe CAV were often continued on a yearly invasive protocol, introducing the possibility of selection bias to a healthier subset of transplant recipients and the cohort was not enriched with a significant group with moderate-to-severe CAV. Lastly, since angiography and PET/CT scans were not contemporaneous, the inclusion of angiograms that were performed up to one year after the PET/CT scan may have inflated the influence of angiographic measures given the progressive nature of CAV.

Conclusion

The results of this study suggest that coronary artery calcium is associated with an increased risk of death or retransplantation. The presence or absence of at least mild coronary artery calcium adjusts future risk of death or retransplantation beyond standard CAV grading.

Abbreviations

CAV

Cardiac allograft vasculopathy

HT

Heart transplantation

ISHLT

International Society for Heart and Lung Transplantation

MPI

Myocardial perfusion imaging

PET/CT

Cardiac hybrid positron emission tomography/computed tomography

sHR

Subdistribution hazard ratio

VECAC

Visually estimated coronary artery calcium