Coronary Computed Tomographic Angiography Imaging as a Prognostic Indicator for Coronary Artery Disease: Data from a Lebanese Tertiary Center.

BACKGROUND: Coronary artery disease (CAD) is a major cause of death and disability worldwide. Coronary computed tomographic angiography (CCTA) is a noninvasive imaging technique with a high negative predictive value (NPV). Most studies were done in developed countries, where the prevalence of CAD does not reflect the actual disease burden in developing countries, such as Lebanon.
METHODS: We retrospectively evaluated the prognostic value of CCTA in predicting acute myocardial events (AMEs) in 200 Lebanese patients. We determined if specific medical and radiological characteristics are linked with AME and looked for any association between the patient's medical risk factors and the type/location of detected atheromatous plaques. Patients' records were reviewed, and the follow-up period of 5-8 years ensued. Chi-square/Fisher test and Student's t-test were used, in addition to multinomial logistic regression to adjust for the confounding variables. P <0.05 was considered statistically significant.
RESULTS: Our study showed that CCTA had a NPV that reaches 97.9% in asymptomatic patients, a positive predictive value (PPV) of 76.4% for symptomatic patients, a sensitivity of 88.9%, and a specificity of 52.5%. AMEs were significantly increased in patients with a mixed plaque type and/or a moderate-to-severe lumen reduction on CCTA.
CONCLUSIONS: CCTA is a sensitive modality for plaque detection and is found to have a remarkably high NPV for asymptomatic patients. A CCTA, along with a low pretest clinical probability of CAD, can be sufficient to rule out an AME for up to 8 years. Copyright:

INTRODUCTION

Coronary artery disease (CAD) is a major cause of death and disability worldwide, causing nearly one-third of deaths in individuals over the age of 35 years. Both CAD mortality and predisposing risk factors continue to increase,[1] with an occurrence at an earlier age in developing countries, as compared to the rest of the world.[23]

Invasive coronary angiography (ICA) is considered the “gold standard” for assessment of coronary artery lumen and degree of blood flow reduction.[4] Coronary computed tomographic angiography (CCTA) imaging is a noninvasive and cost-effective emerging tool for the evaluation of atypical chest pain. It visualizes atherosclerotic plaque burden, along with their composition and location. It also helps in diagnosing fatal causes of chest pain such as aortic dissection and pulmonary embolism.[5]

We herein studied the prognostic value of CCTA in a developing country, in terms of clinical myocardial infarction, and aimed to determine if any association between patient's risk factors and type/location of atheromatous plaques exists.

METHODS

Study design

A total of 200 patients who performed CCTA between the years 2009 and 2012 at Hotel Dieu De France Hospital in Lebanon, for various clinical indications, were retrospectively analyzed. A case–control study design was adopted and consisted of two parts:

Age, gender, body mass index (BMI), smoking status, history of personal cardiovascular risk factors or familial cardiovascular disease, initial imaging indication (asymptomatic vs. symptomatic chest pain), CCTA results, and clinical outcome in terms of acute myocardial event (AME) were extracted from the recruited patients' medical records. A follow-up period of 5–8 years ensued, depending on the imaging date and year of the last visit note with each patient's cardiologist. AME was defined as the occurrence of chest pain requiring coronarography intervention, followed by either percutaneous transluminal coronary angioplasty or, in cases of mild CAD, medical treatment alone.

Based on the abovementioned data, patients were then divided into two groups: a case group and a control group, reflecting the presence or absence of plaques on CCTA, respectively.

Risk factors and coronary computed tomographic angiography parameters

Risk factors studied were derived from the Framingham Coronary Heart Disease (CHD) Risk Score established in 1998[67] and included age, gender, smoking status, exercise, hypertension (HTN), measured systolic blood pressure (SBP) (less, equal, or higher than 140 mmHg), diabetes mellitus (DM), dyslipidemia (DL), and personal or family history of cardiovascular disease (CVD).

Regarding CCTA results, three main findings were noted. First, the plaque location is classified into left main, three-vessel disease, two-vessel disease, predominantly left anterior descending (LAD), right coronary artery (RCA), or left circumflex (LCx) for the case group versus no plaques for the controls. Both three- and two-vessel diseases included, respectively, any three or two concomitant diseased major vessels of the following: LAD, RCX, or LCx.

The second studied parameter was the plaque type, divided into predominantly calcified, noncalcified, or mixed. The calcified type was defined as plaques with an attenuation value ≥ 100 Hounsfield units on CCTA; the noncalcified type represents a vessel wall structure that is mainly fibrofatty, with a homogeneously low density detected on CCTA, and that is lower than the coronary lumen contrast. The mixed type was regarded as a plaque read by CCTA to be predominately fibrocalcified but demonstrating calcification in ≤ 50% of the plaque area. Finally, the grade of coronary artery lumen reduction as detected by CCTA was also reported and divided into four different groups: no stenosis, mild (<50%), moderate (50%–70%), and severe reduction (>70%).[8]

Patients younger than 18 years old, lost to follow-up, refused to consent, or those with prior coronary artery bypass graft were excluded from the study. The presence of a technical imaging problem (respiratory artifacts on CCTA, movement, and morbid obesity) was also an exclusion criterion.

Coronary computed tomographic angiography technique

A 64-slice, multidetector, volume computed tomography scanner with 350 ms or millisecond rotation time was used.

Statistical analysis

Data analysis was performed using IBM SPSS statistics, version 23 (SPSS Inc.). Categorical variables were compared using Chi-square or Fisher's exact test and summarized as frequencies and percentages. The means of quantitative variables were compared using Student's t-test and presented as means and standard deviation. Multinomial logistic regression was used to adjust for the different variables in relation to the different categorical outcomes in question (AME, plaque type, and location). All tests are two-sided, and P < 0.05 was considered to be statistically significant.

RESULTS

Patient characteristics

Of 200 total patients, 176 (88%) were males. The mean age was 54.2 ± 10.8 years, and 94 (47%) of them actively smoked. The mean population BMI was 28.6 ± 4.4 kg/m2. The reported comorbidities were DM in 32 patients (16%), HTN in 83 patients (41.5%), and DL in 96 patients (48%).

The majority of the patients (153 patients, 76.5%) had a normal to high-normal measured SBP (<140 mmHg). Nineteen (9.5%) had a personal history of CVD and 81 (40.5%) had a family history of CVD. One hundred and sixty-five patients (82.5%) had at least one risk factor for CAD [Table 1].

Table 1

Differences in baseline patients’ demographics and characteristics

	Control, n (%)	Case, n (%)	P
Male	66 (83.5)	110 (90.9)	0.239
Age (mean±SD)	50±10	57±10.4	<0.001
BMI (mean±SD)	29±5	28±4	0.266
Smoking status
Nonactive smoker	49 (62.8)	57 (46.7)	0.026
Active smoker	29 (37.2)	65 (53.3)
Exercise	21 (27)	39 (32)	0.448
HTN	29 (37.2)	54 (44.3)	0.321
Measured SBP (mmHg)
<140	57 (73)	96 (78.7)	0.361
≥140	21 (27)	26 (21.3)
DM	11 (14)	21 (17.2)	0.558
DL	26 (33.3)	70 (57.4)	0.001
Personal history of CVD	4 (5.1)	15 (12.3)	0.092
Family history of CVD	29 (37.2)	52 (42.6)	0.444
Presence of at least 1 risk factor (of the above 5)	57 (73)	108 (88.5)	0.005

SD: Standard deviation, BMI: Body mass index (kg/m2), SBP: Systolic blood pressure, HTN: Hypertension, DM: Diabetes mellitus, DL: Dyslipidemia, CVD: Cardiovascular disease

General coronary computed tomographic angiography findings

Seventy-nine patients (39.5%) did not have any plaques on CCTA and were considered as “control,” and 121 patients (60.5%) had a detected plaque and were classified as “case.”

Compared to controls, patients with any detected plaque on CCTA (cases) had significantly higher age (mean age of 57 years), and more than half of them were active smokers (53.3%), more commonly suffered from DL, or had at least one of the following risk factors: DM, HTN, DL, and personal and family history of CVD [Table 1].

The majority of plaque location was found to be in the LAD artery (25%), followed by either left main or three-vessel involvement (16%). As for the detected plaque type, the majority was mixed (40.7%), followed by noncalcified (12.4%) and calcified (6.2%). Among patients with detected plaques, 29.7% had a mild lumen reduction, 15.7% had a moderate reduction, and 11.9% had a severe reduction on CCTA [Table 2].

Table 2

General coronary computed tomography angiography findings

CCTA findings	n (%)
Plaque location
Left main or three-vessel disease	32 (16)
Two-vessel disease	26 (13)
LAD alone	50 (25)
RCA or left cx	13 (6.5)
Total	121 (60.5)
No plaques	79 (39.5)
Plaque type
Calcified	12 (6.2)
Noncalcified	24 (12.4)
Mixed	79 (40.7)
Total	115 (59.3)
No plaques	79 (40.7)
Grade of stenosis on CCTA
No stenosis	79 (42.7)
Mild reduction (1%-49%)	55 (29.7)
Moderate reduction (50%-70%)	29 (15.7)
Severe reduction (>70%)	22 (11.9)
Total	185

CCTA: Coronary computed tomography angiography, LAD: Left anterior descending, RCA: Right coronary artery, cx: Circumflex

Coronary computed tomographic angiography prognostic value in terms of acute myocardial event

Of the total 200 patients, 45 patients were excluded from the CCTA prognostic analysis due to loss of follow-up.

During the follow-up period of 5–8 years after performing CCTA, 55 patients (55/155; 35.5%) had an AME compared to 100 patients (64.5%) who were free of events. Patients who suffered an AME after 5-8 years of follow up were significantly older (mean difference of 6 years), and more frequently had DL or an old history personal CVD (P = 0.001, 0.051, and 0.031, respectively). No significant difference was found between both the groups in terms of gender, BMI, smoking status, regular exercise, HTN, DM, or presence of a family history of CVD [Table 3].

Table 3

Patients’ demographics, risk factors, and occurrence of acute myocardial events upon follow-up

	AME, n (%)		P
	No	Yes
Gender
Male	88 (87)	47 (87)	0.987
Female	13 (13)	7 (13)
Age (mean±SD)	52±10	58±10	0.001
BMI (mean±SD)	28.7±4.9	28.4±3.9	0.72
Smoking status
Nonactive smoker	58 (57.4)	28 (52)	0.506
Active smoker	43 (42.6)	26 (48)
Exercise	26 (25.7)	21 (39)	0.09
HTN	43 (42.6)	24 (44.4)	0.82
Measured SBP (mmHg)
<140	73 (72.3)	45 (83.3)	0.124
≥140	28 (27.7)	9 (16.7)
DM	13 (13)	11 (20.4)	0.219
DL	47 (46.5)	34 (63)	0.051
Personal history of CVD	6 (6)	9 (16.7)	0.031
Family history of CVD	42 (41.6)	25 (46.3)	0.573
Presence of at least one risk factor (of the above five)	81 (80.2)	51 (94.4)	0.017

AME: Acute myocardial event, SD: Standard deviation, BMI: Body mass index (kg/m2), HTN: Hypertension, SBP: Systolic blood pressure, DM: Diabetes mellitus, DL: Dyslipidemia, CVD: Cardiovascular disease

Regarding CCTA findings, having a mixed plaque type, left main, three- or two-vessel involvement, and a moderate-to-severe lumen reduction were all found to be significantly associated with a higher risk of AME, compared to controls (P = 0.001). Patients with AME on follow-up were predominantly symptomatic when CCTA was initially done, as compared to those who remained event free [Table 4].

Table 4

Coronary computed tomographic angiography findings and occurrence of acute myocardial events upon follow-up

	AME, n (%)		P
	No	Yes
Plaque location			0.001
Left main or three-vessel disease	9 (8.9)	18 (33.3)
Two-vessel disease	7 (6.9)	12 (22.2)
LAD alone	27 (26.7)	13 (24.1)
RCA or left Cx	4 (4)	5 (9.3)
No plaque	54 (53.5)	6 (11.1)
Plaque type			0.001
Calcified	7 (7.2)	2 (3.8)
Non-calcified	13 (13.4)	5 (9.6)
Mixed	23 (23.7)	39 (75)
No plaque	54 (55.7)	6 (11.5)
Grade of stenosis			0.001
No stenosis	54 (58.7)	6 (11.8)
Mild reduction (1-49%)	27 (29.3)	12 (23.5)
Moderate reduction (50-69%)	9 (9.8)	16 (31.4)
Severe reduction (≥70%)	2 (2.2)	17 (33.3)
Indication for CCTA (initial symptomatology)			0.0001
Asymptomatic	83 (92)	18 (27.7)
Symptomatic	7 (8)	47 (72.3)
CCTA result			0.0001
Negative	53 (52.5)	6 (11)
Positive	48 (47.5)	48 (89)

CCTA: Coronary computed tomography angiography, AME: Acute myocardial event, LAD: Left anterior descending, RCA: Right coronary artery, Cx: Circumflex

Sensitivity, specificity, negative predictive value, and positive predictive value of coronary computed tomographic angiography

Regarding CCTA sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) for AME, they were found to be 88.9%, 52.5%, 89.8%, and 50%, respectively, for the whole group of patients.

For asymptomatic patients, the CCTA sensitivity, specificity, NPV, and PPV for AME were found to be 85.7%, 57.8%, 97.9%, and 14.6%, respectively. On the contrary, the CCTA sensitivity, specificity, NPV, and PPV for AME in symptomatic patients were found to be 89.4%, 27.8%, 50%, and 76.4%, respectively [Table 5].

Table 5

Sensitivity, specificity, negative predictive value, and positive predictive value of coronary computed tomographic angiography for predicting acute myocardial events after 5-8 years

Test population	Sensitivity (%)	Specificity (%)	NPV (%)	PPV (%)
General	88.9	52.5	89.8	50
Asymptomatic	85.7	57.8	97.9	14.6
Symptomatic	89.4	27.8	50	76.4

NPV: Negative predictive value, PPV: Positive predictive value

Risk factors and location/type of plaques on coronary computed tomographic angiography

After adjusting for age, gender, BMI, smoking status, exercise, DM, HTN, DL, and personal or familial CVD, the following results regarding predominant plaque location involvement as per CCTA results are highlighted:

Old age (above 55 years) and male gender significantly increase the odds of having left main or triple-vessel involvement on CCTA, as compared to middle-aged individuals (36–55 years old) (adjusted odds ratio [AOR] =34.5, confidence interval [CI]: 9.7–125) or being a female (AOR = 15.3, CI = 2.4–96.8), respectively. No other risk factors were found to be significantly associated with a similar increased risk.

Regarding plaques detected on CCTA imaging in two different main coronary arteries, similar results of significantly increased involvement were found in older adults, males, and in addition, in active smokers as compared to middle-aged adults (AOR = 9.7, CI = 3–31.3), females (AOR: 7.7, CI = 1.2–49.8), and nonactive smokers (AOR = 6.4, CI = 2.2–18.9), respectively.

With respect to isolated LAD, a similarly increased involvement in older adults, males, and active smokers as compared to middle-aged (AOR = 4.7, CI = 1.8–12.3), younger adults aged 18–35 years (AOR = 14.3, CI = 1.1–166.7), females (AOR = 4.4, CI = 1.1–17.2), and nonsmokers (AOR = 2.4, CI = 1.1–5.3), respectively. Unpredictably, those who are overweight (defined as having a BMI ranging from 25 to 29.9 kg/m2) were found to have significantly more isolated LAD involvement as compared to those who are obese (BMI ≥ 30 kg/m2) with an AOR = 2.6 (CI = 1.1–6.5). Concerning either isolated RCA or LCx involvement, no significant results were concluded.

After adjusting for age, gender, BMI, smoking status, exercise, DM, HTN, DL, and personal or familial CVD, the following results are concluded:

Regarding mixed plaques identified on CCTA, a significantly higher involvement was found in older adults, males, and active smokers as compared to middle-aged (AOR = 9.5, CI = 3.9–23.3) and younger adults (AOR = 25, CI = 2.1–333.3), females (AOR = 7.7, CI = 2.1–28.5), and nonactive smokers (AOR = 2.7, CI = 1.3–5.8), respectively. With reference to noncalcified plaques, only having an older age was found to have a significantly higher association as compared to middle-aged individuals (AOR = 4.1, CI = 1.3–13.2). As for the calcified type, the least commonly detected plaque, a significantly higher association was found in older adults and those who actively smoke compared to middle-aged individuals (AOR = 5.1, CI = 1.1–23.3) and nonsmokers (AOR = 4.6, CI = 1.1–18.9), respectively.

In conclusion, having either left main or three-vessel disease involvement was found to be significantly associated with older age and male gender; having two concomitant vessel involvement or isolated LAD plaques with a mixed type and being older, male, or an active smoker; having mainly noncalcified plaques and being older; and finally, having predominantly calcified plaques and being older or active smoker. With respect to the remaining studied risk factors (exercise, DM, HTN, DL, and personal or familial CVD), there was no significant association with a specific location or plaque type.

DISCUSSION

The current study showed that CCTA has the best PPV (76.4%) in patients who underwent the procedure following symptomatic chest pain. However, for the general population, PPV was found to be only 50%. This PPV goes in the same average of a wide range of PPV values present in the literature (35%–97%).[91011] This relatively low PPV can be explained by the overestimation of the coronary artery narrowing detected by CCTA, secondary to blurring artifacts.

With respect to our NPV, its lower result as compared to others (97%–100%)[1112] can be explained by the difference in the sample size studied, ethnicity, and the longer follow-up period (5–8 years in our study compared to a mean of 2 years in other studies).

In addition, the sensitivity of CCTA in our study is high reaching 88.9% and is almost similar to other studies. However, we noted a discrepancy in our specificity (52.5%) compared to other studies (90%–95%),[1012] except for the one conducted by Hoffmann et al. (54%–87%).[9] This is also due to the overdetection of the coronary artery plaques by CCTA due to calcified plaques.

Zhang et al.[13] found that diabetic patients had more diffuse vessel disease, with more than one vessel involved compared to nondiabetics; however, in our study, we did not reach any significance for the association between classical CAD risk factors such as HTN, DM, or DL and a specific plaque location on CCTA. A probable explanation is our small sample size. On the contrary, it was not surprising to find that older men and active smokers are most vulnerable for left main, triple-vessel, and/or LAD plaque involvement.

Concerning the smoking status, our study showed a significant association between older age group and active smoking with the predominance of calcified plaque detected on CCTA. Add to this, it showed that having an older age was significantly linked with the detection of noncalcified plaques. Going in parallel with our results, a study conducted by North et al. similarly highlighted the role of smoking in the pathogenesis of “calcified” coronary plaque,[14] whereas Lehman et al. reported in the ROMICAT trial that smoking was independently associated with coronary atherosclerotic plaque burden progression on CCTA in patients with acute chest pain over 2 years of follow-up, and the rate of this progression was dependent on plaque composition, being higher for “noncalcified,” when compared to “calcified” plaques.[15]

Apart from smoking, male gender and older age, we did not find any significant association between other classical CAD risk factors such as HTN, DM, DL, personal or familial CVD, and a specific plaque type detected by CCTA. Although these risk factors are all included in the Framingham CHD Risk Score,[67] a score for CHD risk estimation, the latter also being increased by noncalcified plaque type, as mentioned above in the ROMICAT trial.[15] Many studies were conducted in the same context and revealed contradictory associations, both not shown in our study. Rivera et al.[16] showed that noncalcified coronary plaques were particularly associated with DM, while Zhang et al.[13] found that calcified plaques were significantly more detected in diabetic compared to nondiabetics. These contradictory results can be again explained by our small studied sample size.

CONCLUSIONS

In this study, CCTA is proven to be sensitive for coronary plaque detection regardless of symptomatology and has a high NPV for excluding CAD, reaching 97.9% in asymptomatic patients. Using this test alone, along with a low pretest clinical probability (absence of risk factors for CAD), is sufficient to rule out any AME for up to 8 years.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.