Statin use in patients with coronary atheroma identified on computed tomography coronary angiography: current practice in South Auckland, New Zealand (ANZACS-QI 82)

Computed tomography coronary angiography (CTCA) has a high accuracy for the diagnosis of coronary atherosclerosis and is strongly endorsed by international guidelines in the assessment of chest pain syndromes.^1,2 CTCA can non-invasively identify the extent of obstructive coronary artery disease (CAD) with comparable effectiveness to invasive coronary angiography.³ Moreover, CTCA also detects coronary atherosclerosis in the absence of obstructive CAD. The finding of any anatomic evidence of atherosclerosis is an independent predictor of adverse cardiovascular events even after adjusting for known cardiovascular risk factors.⁴ Its presence is an indication for more intensive prevention, including use of statin medications to reduce low-density lipoproteins (LDL) cholesterol.⁵ Atherosclerosis associated with obstructive CAD is a well-established indication for statin therapy. Observational studies also suggest that statin therapy in patients with CTCA evidence of atherosclerosis but without obstructive CAD is associated with improved outcomes.^6,7

With the more routine use of CTCA in the investigation of chest pain there is an opportunity to better target lipid-lowering therapy according to whether or not a patient has coronary atherosclerosis. We have recently implemented a CTCA registry within the Aotearoa New Zealand All Cardiology Services Quality Improvement (ANZACS-QI) programme. The registry is linked with national administrative datasets to track statin dispensing. We therefore have an opportunity to assess whether CTCA finding of coronary atherosclerosis, both with or without obstructive CAD, is associated with changes in the subsequent use of statin therapy.

Methods

Study cohort and data collection

The study cohort was identified from the ANZACS-QI cardiac computed tomography (CT) registry at Middlemore Hospital, Auckland, New Zealand. The ANZACS-QI cardiac CT reporting tool generates a structured clinical report while simultaneously collecting an audit and a research-enabled dataset. The reporting tool is incorporated in the national ANZACS-QI web-based platform.⁸ Patient demographics, cardiovascular risk factors and CT findings are captured in this registry, and the de-identified study cohort was linked to national datasets to track the use of statins in patients with chest pain who underwent CTCA, from initial dispensing 6–12 months prior to CTCA to subsequent maintenance up to 9–12 months post-CTCA.

For this study, at least a year of post-CTCA follow-up time was required. Therefore, the study cohort included consecutive patients who underwent CTCA at Middlemore Hospital for investigation of chest pain syndrome from 21 September 2020 to 31 December 2021 in both acute inpatient and outpatient settings. The inpatient cohort comprised patients presenting to hospital with acute chest pain in whom myocardial infarction (MI) was ruled out, and the outpatient cohort comprised those with stable chest pain who had been referred to the cardiology clinic for evaluation. Only the first CTCA for each patient in the time period was used. Medication dispensing data were obtained by individual linkage to routine national datasets, as previously described.^9,10 Data on LDL cholesterol, coronary angiography, percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) were obtained through linkage to national datasets. An encrypted version of the National Health Index (NHI) number, a unique identifier assigned to everyone who uses health and disability support services (>98% of the population), was used to link de-identified ANZACS-QI cardiac CT patient records to the national datasets. The encryption and linkage methodology has previously been described.^8–10

Variables used for this study included age, gender, ethnicity, diabetes, smoking status, family history of cardiovascular disease (CVD), hypertension and LDL cholesterol within 3 months before CTCA. Patients were grouped into those with or without coronary artery atheroma using the segment involvement score (SIS). The SIS score is determined on CTCA by designation of a score of 1 for each one of the coronary artery segments with a detectable atherosclerotic coronary plaque, irrespective of the plaque size or individual plaque burden in that segment.^11,12 In this study, presence of coronary atheroma was defined as a SIS score of 1 or more, and categorised as 0, 1, 2–3 and >4 segments.⁷ Patients with coronary atheroma were further divided into those with obstructive CAD (defined as >50% reduction in the luminal diameter of a major coronary artery) and without obstructive CAD (<50% diameter stenosis).¹³

Predictors of statin dispensing within 3 months post-CTCA was assessed. Sociodemographic variables were derived from the linked national dataset. For patients with more than one ethnic group recorded, ethnicity was prioritised in accordance with health sector protocols, in the following order: Indigenous Māori, Pacific, Indian, and NZ European/Other.¹⁴ Socio-economic deprivation was assessed by the New Zealand Index of Deprivation 2018 (NZDep2018) score, a census-based small-area 10-point index of deprivation based on the person’s domicile.¹⁵

Linkage of the ANZACS-QI CT registry and national datasets has been approved by the National Multi Region Ethics Committee (MEC07/19/EXP) and Auckland Health Research Ethics Committee (AHREC AH26614).

Statistical analysis

Categorical variables were summarised as frequency and percentage and continuous variables as mean and standard deviation (SD) or median and interquartile range (IQR). Comparisons between those with atheroma and those without were made using Pearson’s Chi-squared test for categorical data and the Mann–Whitney U test for continuous data as it is not normally distributed. All tests of statistical significance were two-tailed and a p-value <0.05 was considered statistically significant.

For regression analysis, multivariate (modified) Poisson regression models with robust error estimates were used to estimate the relative risk (RR) with a 95% confidence interval (CI) and associated p-value for variables associated with statins dispensed within 3 months post-CT. Variables included in the model were whether patients had atheroma, age, gender, ethnicity, NZDep2018, current smoker, hypertension and diabetes. A two-sided p-value of less than 0.05 was considered to indicate statistical significance.

Data were analysed using SAS version 9.4 (SAS Institute, Cary, North Carolina, United States of America).

Results

Study cohort

Between 21 September 2020 and 31 December 2021, there were 1,294 patients who underwent CTCA and had at least a year of available follow-up. Five-hundred and forty-eight patients had CTCA as an inpatient and 746 patients underwent CTCA in the outpatient setting. Table 1 summarises the clinical characteristics of the cohort.

Acute chest pain cohort

The mean age of the patients was 56.1+11.5 years and 287 (52.4%) were men (Table 1). Of the 548 consecutive patients, 204 (37.2%) had no atheroma, 208 (38.0%) had non-obstructive CAD and 136 (24.8%) had obstructive CAD. There were more men in the sub-groups with CAD and they were older compared with those with no atheroma. Hypertension and diabetes were more frequent in the CAD sub-groups. Of the 136 patients with obstructive CAD, 87 (64%) patients underwent invasive coronary angiography within 2 weeks after CTCA. Thirty-eight (27.9%) patients subsequently had inpatient revascularisation (PCI or CABG) within 2 weeks post-CTCA.

Table 2 shows the initial statin dispensing and maintenance over the year post-CTCA. Of the 548 patients, 52.9% of those with obstructive CAD, 45.7% with non-obstructive disease and 23.5% with no atheroma were on a statin pre-CTCA (p<0.001, Figure 1). Within 3 months post-CTCA, 92.6% of those with obstructive disease and 67.8% with non-obstructive disease were on a statin compared to 33.8% for those with no atheroma (p<0.001). More than 80% of those with obstructive CAD were continued on a statin up to 9–12 months post-CTCA compared with only 59.6% of those with non-obstructive CAD, and 28.9% with no atheroma. Statin use pre- and post-CTCA in those without obstructive CAD also did not differ significantly according to SIS score (Appendix Table 1).

Multivariate regression (Table 3): After adjustment for demographic and cardiovascular risk factors, patients with obstructive CAD were twice as likely to be prescribed statins within 3 months post-CTCA (RR 2.19, 95% CI 1.76–2.71) and the finding of non-obstructive CAD was associated with 55% higher statin dispensing within 3 months post-CTCA (RR 1.55, 95% CI 1.25–1.94). Other variables independently associated with higher statin dispensing were hypertension, male gender, older age and Indian ethnicity.

Stable chest pain cohort

The clinical characteristics of patients who underwent CTCA as outpatient were shown in Table 1. The mean age was 58.6+10.5 years, and 47.5% were men. More than half of the patients (n=454, 60.9%) had atheroma on CTCA and obstructive disease was identified in 140 (30.8%) patients. Those with and without obstructive CAD were younger (p<0.001) and had more cardiovascular risk factors compared with those with no atheroma. There were more men, Māori and Pacific people in the CAD group compared with the no atheroma group. Fifty-eight (41.4%) out of 140 patients with obstructive disease underwent invasive angiography and 36 (25.7%) subsequently had revascularisation (PCI or CABG) within 6 months post-CTCA.

Statin dispensing before and after CTCA is described in Table 2 and Figure 2. Statin dispensing increased in the three groups within 6 months pre-CTCA. Statin dispensing decreased in the no atheroma group post-CTCA, whereas there was a significant increase in statin dispensing in the obstructive disease group 3 months and up to 9–12 months post-CTCA (p<0.001). There was no significant change in statin dispensing in the non-obstructive disease group post-CTCA. Within the group without obstructive CAD there was higher use of statins post-CTCA in those with a greater extent of atheroma (SIS 1: 49.5%; SIS 2–3: 70.3%; SIS >4: 66.7 [p=0.002]) (Appendix Table 1).

Multivariate regression (Table 3): The finding of atheroma with and without obstructive disease on CTCA after adjustment for demographic and cardiovascular risk factors was associated with higher statin prescribing within 3 months post-CTCA (without obstructive disease: RR 1.80, 95% CI 1.49– 2.16; with obstructive disease: RR 2.14, 95% CI 1.75–2.62). Other variables independently associated with higher statin prescribing were hypertension, diabetes and older age.

Discussion

In this cohort of symptomatic patients referred for CTCA, over two-thirds have evidence of coronary atherosclerosis. The CTCA finding of atherosclerosis with obstructive CAD was associated with appropriate initiation and maintenance of statin dispensing. In those without coronary atherosclerosis, statin dispensing fell towards baseline levels after increasing prior to CTCA. However, around a quarter of these patients continued a statin despite the documented absence of atheroma. Just over a third of the cohort had atheroma documented but without obstructive CAD. The use of statins in this sub-group was intermediate between those with obstructive CAD and those with no atheroma. After adjustment for available covariates, the presence of atheroma was strongly associated with statin dispensing in patients with obstructive and non-obstructive CAD.

Statin use in patients with no atheroma

Subsequent atherosclerotic CVD event rates are very low in asymptomatic and symptomatic patients with calcium score of zero, and in primary prevention a calcium score of zero is used to reclassify patients with intermediate CVD risk estimated using traditional risk factors as lower risk.^16–18 Because coronary atheroma is the precursor pathological lesion to coronary calcium, the absence of atheroma would be expected to confer a very low risk. In a Danish registry of patients studied with CTCA for suspected CAD, those with no CAD had a low rate of all-cause mortality or MI (4.13/1,000 patient years).¹⁹ It is likely that many of these patients will have non-cardiac chest pain. However, this group will also include some patients who have ischaemia without obstructive disease. In patients with known non-coronary atherosclerotic CVD, familial hypercholesterolaemia and very high CVD risk estimated from traditional risk factors, continuation of statins may still be considered appropriate.

Statin use in patients with non-obstructive CAD

In our cohort, more than 60% of patients with non-obstructive CAD were dispensed a statin post-CTCA. This finding was similar to the recently published meta-analysis²⁰ where CTCA resulted in higher statin prescriptions, and the rate of statin dispensing is higher than in prior published series where rates were around 50%.^19,21,22 Neither current primary prevention nor secondary prevention guidelines contain recommendations for medical therapy in patients with non-obstructive CAD beyond traditional cardiovascular risk factors. Several recent observational studies have reported a benefit of statins in patients with non-obstructive CAD.^7,19 A large Danish registry of patients with suspected CAD found a graded relationship between the burden of CAD measured by coronary calcium score (CCS) and CVD outcomes. After adjustment for covariates, the use of statins was associated with improved outcomes even in those with either no CAD or mild non-obstructive CAD with CCS 0–99.¹⁹ However, in that study only a third of patients with mild CAD received a statin and unmeasured confounders may have influenced the results.

Multiple observational studies have shown that the direct demonstration of atherosclerosis on coronary imaging studies is a strong predictor of subsequent atherosclerotic CVD events. The burden of atherosclerotic disease can be reported using CCS or the SIS score. The presence of any atheroma or coronary calcium is associated with increased CVD risk but there is a graded increase in risk with increasing disease burden.^7,23 Statins are clearly indicated in patients with extensive CAD documented by high CCS or SIS scores in whom CVD risk is in the range seen for those with known clinical atherosclerotic CVD. In the Multi-Ethnic Study of Atherosclerosis (MESA),^17,23 a CCS score >300 confers an absolute 10-year hard CVD event rate of over 15%. Even intermediate CCS scores of 100–300 were shown to confer a >7.5% 10-year risk in all the demographic sub-groups studied—the risk threshold endorsed for considering statin initiation in the American College of Cardiology/American Heart Association (ACC/AHA) cholesterol and prevention guidelines.⁵ This threshold is similar to the >5% all CVD 5-year risk threshold endorsed in the New Zealand guidelines.²⁴

It is, however, less clear what to recommend for those with a lesser extent of disease, in whom the absolute increase in risk may be small.⁴ While guidelines have endorsed the use of CCS to re-stratify risk in those with intermediate risk estimated using the traditional risk factors, there is no guidance about what to do with the finding of atheroma on coronary imaging. Multiple studies have shown that the addition of CCS to traditional risk factors improves risk prediction,²⁵ and it is likely that when atheroma volume is small, the combination of imaging information and a 5- or 10-year CVD risk estimated using traditional risk factors will best inform clinical decision making. All patients, regardless of the anatomic extent of coronary atheroma, should have traditional CVD risk factors measured as these are important modifiers of disease and targets for treatment.

Statin use in patients with obstructive CAD

Use of statins is a Class I indication for secondary prevention in patients with known CVD. The major clinical trials supporting this recommendation were done prior to CTCA availability and largely included patients with symptomatic disease. Although we studied cohorts with suspected CAD, not all those with CTCA evidence of obstructive CAD will have had symptoms clearly attributable to CAD, and our cohort probably includes some patients with asymptomatic obstructive CAD. The 2018 ACC/AHA guideline⁵ recommends starting statins even in asymptomatic obstructive CAD as an indication for secondary prevention. Use of CTCA has therefore increased the number of patients identified as appropriate for secondary prevention. In our cohorts around half the patients with CTCA-proven disease were already on a statin prior to presentation, consistent with a cohort that includes patients with known high CVD risk due to either risk factors or known CVD. The finding of obstructive CAD increased the proportion maintained on a statin to over 80%. The maintenance rates for statins in this cohort is similar to the previous published report using the 2015–2017 New Zealand acute coronary syndrome cohort.²⁶ In that study, which included all acute coronary syndrome patients, 89.8% were dispensed a statin within 3 months, and 79.8% had a medication possession ratio ≥0.8 during the first year, indicating that a high proportion of patients had adequately maintained dispensing over the year post-discharge.²⁶ The one in five patients with obstructive CAD not on a statin at 1 year is likely due to known intolerances or barriers to getting medication dispensed early after discharge. They are a group who may benefit from strategies to improve medication adherence, and are appropriate as a target for quality improvement.

Temporal changes in statin dispensing

Statin dispensing increased in those who presented with acute chest pain despite no atheroma on CTCA. It is routine practice to prescribe a statin alongside other medications, such as aspirin, for patients with acute chest pain in the hospital, so it is not surprising to find that statin dispensing increased early post-CTCA and subsequently the rate dropped in the first year post-discharge. The rate of statin dispensing also dropped in the non-obstructive disease group post-CTCA. These patients were routinely hospitalised under the care of general medical specialists at our institution. In contrast, outpatients with stable chest pain were seen by the cardiology team, and those with non-obstructive disease were maintained on a statin at 1 year. A prior registry-based study²⁷ found that patients seen by cardiologists were more likely than others to receive guideline-directed statin therapy in both primary and secondary prevention contexts.

Study limitations

This study provides information on statins dispensed, which may or may not be consumed. However, this method is considered to have high specificity (identify those not taking medication) and is more likely than clinical trials to reflect a “real world” situation.²⁸ We have also not taken into account the dosing and type of statin used, but we did include anyone who is on a statin regardless of type or dosage. We use the SIS score as a measure of the extent of coronary atherosclerosis instead of the CCS as our CTCA studies did not have a concomitant calcium score performed. SIS is a semi-quantitative measure of extent of atherosclerotic disease, irrespective of severity of plaque. Quantifying extent of CAD as such is consistent with the move from qualitative to standardised quantitative CTCA reporting in the Society of Cardiovascular Computed Tomography’s Coronary Artery Disease Reporting and Data System (CAD-RADS) guidelines.¹¹ Unlike the CCS, it also accounts for non-calcified plaque, which may be the substrate for major adverse cardiac events.²⁹

Conclusion

In this study, the finding of atheroma with obstructive disease on CTCA is associated with an appropriate increase in statin use and absence of atheroma leads to appropriate down-titration/discontinuation of statin use. Statin use in patients with non-obstructive CAD is still suboptimal despite evidence supporting the use of statins is associated with a clinically important reduction in the risk of MI and death in patients with non-obstructive CAD. Therefore, initiatives to improve longer-term maintenance of statins in this group of patients are required to improve patient outcomes.

View Table 1–3, Figure 1–2.

View Appendix.