Background

Release of new ESC Guidelines

In August 2019 the European Society of Cardiology (ESC) released its most recent dyslipidaemia management [1] (subsequently hereafter referred to as ‘ESC 2019 guidelines’).

The outcomes of several major clinical trials form the basis for these revisions. In particular, the FOURIER [2] and ODYSSEY OUTCOMES [3] trials. These trials demonstrated that the low-density lipoprotein cholesterol (LDL-C) reductions previously seen with the addition of a proprotein convertase subtilisin/kexin type 9 serine protease inhibitor (PCSK9i) to those high-risk patients on maximally tolerated statin therapy translate to a lower incidence of major cardiovascular events. Additionally, there appears to be no threshold for clinical benefit in lowering LDL-C levels in these high-risk patients. As such, these new guidelines have introduced lower LDL-C targets for those in the high and very high-risk categories and have recommended the use of a PCSK9 inhibitor in those at ‘very high risk’ who fail to meet their LDL-C target while prescribed a high intensity statin and ezetimibe.

The 2019 ESC guidelines now specify that high and very high-risk patients should also aim to reduce their LDL-C levels below an absolute target, as well as a reduction in their LDL-C levels > 50% from baseline. Additionally, it is now recommended that those at ‘very high risk’ who fail to achieve their LDL-C target using both a high intensity statin and ezetimibe, a combination with a PCSK9i is recommended [1].

Cardiac rehabilitation

Cardiac rehabilitation is a supervised programme designed to improve cardiovascular health. It includes education on risk factor management, heart healthy living, exercise counselling and training. It is most often offered as part of a secondary prevention strategy for acute coronary syndrome (ACS) patients, but increasingly is being offered for non-ACS indications such as heart failure or valvopathies. It takes place over a four-phase process, with phase one and two occurring peri-admission. During this index admission to an acute hospital, a metabolic risk panel, including a lipid profile is measured. Phase three consists of a 6–12 week structured exercise and education programme, and it is during this time that lipid management is optimised.

Dyslipidaemia management and LLT titration

Conventionally, a lipid profile is obtained during an index admission, with lipid lowering therapy (LLT) prescribed or altered upon discharge from hospital. Patient lipid profiles are again measured at the initiation of phase 3 to ensure adequate clinical response, and LLT may be titrated by an advanced nurse prescriber (ANP).

Baseline LDL-C levels

The term ‘baseline’ refers to the LDL-C level in a person not taking any LDL-C-lowering medication. In people who are taking LDL-C-lowering medication(s), the projected baseline (untreated) LDL-C levels should be estimated, based on the average LDL-C-lowering efficacy of the given medication or combination of medications [1] (Fig. 1).

Fig. 1
figure 1

Adapted from ‘ESC guidelines for the management of dyslipidaemia’, 2019

Full size image

Aims and objectives

To determine the proportion of our patients who completed cardiac rehabilitation in 2019 that reached their LDL-C targets, before and after the introduction of 2019 ESC guidelines.

To examine the choice of lipid lowering therapy (LLT) prescribed in our cohort, including changes to prescribing practice following the introduction of the 2019 ESC guidelines.

Methods

Retrospective chart review

A list of patients who completed phase 3 of cardiac rehab in 2019 was created using our centre’s Patient Administration System (PAS). Those patients who did not complete phase 3 before Dec 31, 2019 were excluded. Four patients with missing LDL-C data were excluded. This returned 163 patients.

A retrospective chart review for these patients was undertaken. A database was created containing baseline patient characteristics including gender, age, BMI, cardiac risk factors, reason for cardiac rehabilitation (CR) referral and laboratory results. These results included prior lipid profile results and their contemporary lipid lowering therapy (LLT), final lipid profile at the end of phase 3 (taken as their ‘post rehab’ lipids) and corresponding LLT. For those patients prescribed a statin, the ‘intensity’ of the statin was classified as ‘mild’, ‘moderate’ or ‘high’, based on the expected LDL-C lowering effect the agent at a specific dose (Table S1).

For those patients who did not have a previous diagnosis of atherosclerotic cardiovascular disease (ASCVD) (as defined in the 2016 and 2019 ESC guidelines) and were managed as part of a primary prevention strategy, further fields were generated in order to risk stratify and determine their LDL target.

These fields included blood pressure and prior diagnosis of diabetes (with time since diagnosis and presence of end-organ damage, if applicable). Baseline serum creatinine was determined, and the Modification of Diet in Renal Disease Study (MDRD) equation was used to determine the estimated glomerular filtration rate. Those with chronic kidney disease were stratified as severe (< 30 mL), or moderate (30–60 mL/min). The presence of a markedly elevated single risk factor (e.g. BP > 180 systolic or LDL > 8 mmol/L) was recorded. A SCORE risk was then calculated. These patients were then risk stratified using their highest scoring criteria.

For our secondary prevention cohort, the presence of a second vascular event within 2 years was considered; however, none of our patients fulfilled this criterion.

LDL-C targets

Either 2016 or 2019 ESC guideline LDL-C targets were applied to patients depending on if they started phase 3 of cardiac rehab prior to or post September 9, 2019, respectively (Fig. 2). This date represents a 10-day period to account for delays in implementation. Confirmation was obtained that the implementation of these guidelines was immediate.

Fig. 2
figure 2

LDL-C Level Targets for either cardiovascular risk category, stratified by guidelines used

Full size image

For those patients treated using the 2016 guidance, achievement of either an LDL level below that recommended for their risk category or a reduction in LDL level over 50% was recorded as successful adherence to guidelines.

For those treated using the 2019 guidance, only those patients who achieved both their risk-specific LDL-C level and an over 50% reduction in their LDL-C level were recorded as adhering to the guidelines. For those patients in whom only a post-CR LDL-C level was available, achieving an LDL-C level below their recommended target was recorded as successfully adhering to guidance.

Baseline lipids

Baseline LDL-C values were determined where possible. Where available, this was taken as a lipid profile obtained while the patient was not prescribed LLT.

As most patients were already receiving LLT upon presentation, baseline LDL-C was calculated by recording the results of a lipid profile taken within the year prior to starting CR, as well as the LLT they were prescribed at this time. The projected LDL-C lowering effects of the LLT (Fig. 1) were used to calculate a baseline LDL-C.

It was not possible to calculate the baseline LDL-C values for 45 patients, due to missing data regarding their LLT or having only one lipid profile on file.

LLT prescribing practices

The agent(s) and dose(es) that formed each patient’s LLT were recorded at both the start and completion of phase 3 of CR. Increase in agent dosage or the introduction of a new agent was classified as an increase in LLT. A dose decrease or withdrawal of an agent was recorded as a decrease in LLT. Patients prescribed the same LLT at both the start and end of phase 3 were recorded as no change in LLT. Those switched to an equivalent agent(s) of similar expected LDL-C lowering efficacy (Fig. 1) were recorded as being changed to an equivalent LLT. Those who were recorded as not being prescribed LLT at the start of CR and then prescribed LLT were recorded as having been initiated on LLT. Alterations in LLT which could not be inferred for patients who did not have their initial or completion LLT recorded were recorded as an unknown change.

Predicted required LLT

For those who did not achieve their absolute LDL-C target at the end of CR and for whom contemporary LLT data was available (n = 55), the percentage reduction in their latest measured LDL-C required to reach their absolute LDL-C target was calculated. Using the estimated LDL-C lowering effects of different LLT regimes (Fig. 1), the approximate LDL-C lowering effect of their current LLT was noted, and the LLT regime that would be likely to achieve the additional percentage reduction in LDL-C was determined.

As an example, a ‘very high’ risk patient treated under the 2019 guidelines would have an LDL-C target of 1.4 mmol/L. If their LDL-C is measured as 2.0 mmol/L at the end of phase 3, then a 30% reduction in their current LDL-C is required to meet this target. Their prescribed LLT is atorvastatin 20 mg monotherapy (a moderate intensity statin at this dose), which can be expected to reduce LDL-C 30% from baseline. As can be seen from Fig. 1, up-titration of their LLT to a high intensity statin therapy in combination with ezetimibe would likely be required to achieve an additional 30% reduction in their LDL-C. Therefore, their predicted required LLT is recorded as a high intensity statin in combination with ezetimibe.

Statistical analysis

All statistical computations were performed in SPSS statistics (version 25; IBM, New York, USA). The alpha value was set at 0.05. Categorical data was described using counts and percentages where appropriate. When comparing the pre-2019 to post-2019 guideline cohorts, respective baseline categorical variables were compared using a Pearson’s X2 test for independent samples to control confounding factors and ensure both groups were well matched. When comparing column proportions, the chi-squared test was used.

Numerical data was described using mean and standard deviation where appropriate. Normality was confirmed both graphically and using the Shapiro-Wilks normality test. Due to the significant difference in population sizes, homogeneity of variances was confirmed using Levene’s test before comparing continuous variables. Scale variables were described using the mean, 95% confidence intervals for the mean and an unpaired t test was performed when comparing groups.

Results

Baseline characteristics

Table 1 displays the baseline characteristics of all the patients who finished phase 3 of CR in 2019, stratified by the guidelines under which they were treated. The mean (SD) age of those who completed CR in 2019 was 62 (9) years, 76% (123/163) were male, with the majority (87%; 142/163) being treated as part of a secondary prevention strategy. Of those who did not have a diagnosis of ASCVD, 5 (3%) were in the ‘high-risk’ category, 16 (10%) were of ‘moderate’ risk, and there was 1 patient deemed to be of ‘low’ risk.

Table 1 Baseline characteristics of those patients who completed CR in 2019
Full size table

There were no significant differences with regard to age (p = 0.234), gender (p = 0.560), proportion of patients being treated for primary or secondary prevention (p = 0.520) or risk category (p = 0.491). The proportion of patients with established ASCVD was comparable across both 2016 guideline and 2019 guideline cohorts (p = 0.520).

Significantly, more patients in the 2019 guideline cohort were on no LLT at the beginning of phase 3 (p < 0.005) and a higher proportion of our 2016 guideline cohort began CR with a high intensity statin monotherapy (p < 0.005).

Choice of lipid lowering therapy

Table 2 displays the LLT prescribed for each patient upon completion of CR in 2019, stratified by guidelines used. Overall, the most common LLT on completion of CR was a high intensity statin monotherapy (91/163, 56%), followed by a moderate intensity statin monotherapy (41/163, 25%). Two patients (1%) were treated with a high intensity statin and ezetimibe. Fourteen patients were on no lipid lowering therapy, 4 of whom had a documented refusal of LLT (2%). There was no significant difference, on the completion of CR, in the proportion of patients prescribed a high intensity statin therapy, either alone or in combination with ezetimibe, between the 2016 (64/112; 57%) and 2019 (29/51; 57%) cohort (p = 0.553).

Table 2 Lipid lowering therapy at completion of phase 3 of CR in 2019
Full size table

In total 5/163 (3%) of our cohort were prescribed ezetimibe therapy, either as a monotherapy or in combination with other LLT.

Titration of lipid lowering therapy

Overall, there was a significant difference in the prescribing practices between the 2016 and 2019 cohorts, which are summarised in Table 3. Patients treated under the 2019 guidelines were more likely to be initiated on LLT (31% vs. 13%; p < 0.005) or have their LLT up-titrated (16% vs. 4%; p = 0.014) and were less likely to be maintained on their current LLT (27% vs. 73%; p < 0.005). There was no difference in the proportion of patients who had their LLT down-titrated (8% vs. 4%; p = 0.430) or changed to another LLT with an equivalent expected LDL-C reducing effect (8% vs. 4%; p = 0.098). There was a significantly higher proportion of patients treated under the 2019 guidelines for whom the titration of LLT could not be determined due to missing or incomplete data (16% vs. 3%; p = 0.023).

Table 3 Comparison of LLT prescribing practices between those treated under the 2016 and 2019 guidelines
Full size table

Estimation of predicted required LLT

The estimated required LLT for those who completed CR in 2019 are summarised in Table 4.

Table 4 Predicted required LLT for those who completed CR in 2019 (n = 55)
Full size table

Eighty-two percent (55/67) of patients who failed to meet their contemporary LDL-C level in 2019 had sufficient LDL-C and LLT data to estimate a predicted LLT. Seven percent (4/55) and 27% (15/55) are likely to require either a moderate or high intensity statin monotherapy, respectively, in order to meet their contemporary, absolute LDL-C target. Twenty-five (14/55) would likely require a high intensity statin and ezetimibe combination to meet their absolute LDL-C target. It is estimated 40% (22/55) of these patients would require a PCSK9i to reach their absolute LDL-C target, either in combination with a high intensity statin only (16%; 9/55) or in combination with a high intensity statin and ezetimibe (24%; 13/55).

Cholesterol

Table 5 displays the mean LDL-C results for those patients attending CR, both prior to starting (where available) and upon completion of CR. It also displays the number of patients who achieved their absolute LDL-C targets for their relevant risk category and applicable guidelines. The mean percentage reduction in LDL-C, where both a pre- and post-CR LDL-C was available, as well as the number of patients achieving an over 50% reduction in their baseline LDL-C is also displayed.

Table 5 Summary of LDL-C values for those patients who completed CR in 2019, stratified by treatment guidelines and risk category
Full size table

For those patients who completed Cr in 2019, 96/163 (59%) patients met their absolute LDL-C targets. The mean (95% CI) LDL-C reduction was 57% (52–62), with 61 patients of the 118 for whom pre-CR LDL-C was available (52%) achieving an over 50% reduction in their LDL-C levels. One hundred four (64%) patients were found to be treated in compliance with their contemporary guidelines. The distribution of either measured or calculated baseline LDL-C was not significantly different between both groups (p = 0.006).

With regard to those patients treated using the 2016 guidelines, 70/112 (63%) patients met their absolute LDL target. The mean (95% CI) reduction in LDL-C was 61% (56–66), and 50 out of the 81 patients for whom LDL data was available, 62% achieved LDL-C reductions > 50% of baseline. Eighty-two (73%) of those patients treated under the 2016 guidelines achieved their LDL-C targets.

Those patients treated under the 2019 guidelines had fewer patients meeting their absolute (26/51, (51%)) or relative reduction (11/37, 30%) LDL-C targets when compared to the 2016 cohort. The mean reduction in LDL-C from baseline was 48% (95% CI 37–59%).

Missing data

During data collection, there were 4 patients who had no LDL-C data available, either pre- or post-CR.

Choice of LLT was not recorded for 8 patients. Baseline LDL-C was either unavailable or could not be estimated for 45 (28%) patients.

Discussion

Baseline characteristics

The majority of patients who attended cardiac rehabilitation in 2019 had established ASCVD. The majority of those patients without ASCVD were referred following cardiac valve surgery, a diagnosis of heart failure and insertion of implantable cardiac defibrillator (ICD).

Choice of LLT

While the most common choice of LLT, only slightly more than half of our cohorts were prescribed a high intensity statin monotherapy (56%), and only a very small minority were prescribed ezetimibe. As the majority of our cohorts (88%) were considered to be at ‘very high risk’, this seems to suggest that many patients are being undertreated. This could perhaps be due to intolerance of these medications, therapeutic inertia, or a lack of awareness of available therapeutic strategies. As there were no patients who did not meet their LDL-C target while prescribed ezetimibe in combination with a high intensity statin therapy, no patients met the criteria for a combination therapy with a PCSK9i, as recommended by the ESC.

Prescribing practices

Allowing for the proportion of missing data, those treated under the 2019 ESC guidelines appeared to have been subject to a more aggressive LLT strategy. This may be as a result to changes in practice in response to the new lower targets set out in the 2019 and may account for the lower LDL-C levels seen in those treated under the 2019 ESC guidelines.

LDL-C target achievement

For many of those patients treated under the 2016 guidelines who had not met their absolute LDL-C targets, determining their baseline LDL-C levels found that they achieved-over 50% reduction in LDL-C from baseline, and as such could be said to be treated in accordance with best practice. This resulted in overall greater number of patients meeting their LDL-C targets.

However, given that the 2019 ESC dyslipidaemia guidelines specify that both an absolute and relative target needed to be met for those at high and very high risks, many who met their absolute LDL-C target did not meet their relative reduction from LDL-C baseline. This meant that for this cohort, consideration of their baseline LDL-C levels meant fewer patients were meeting their LDL-C targets.

Comparison with international standards

Numerous large international cross-sectional and prospective studies, both European and worldwide, have demonstrated poor risk factor control in those with established ASCVD [4,5,6,7,8]. The proportion of patients who attended CR in 2019 in our centre and reached their absolute LDL-C target compares favourably when compared to large European studies. A survey of 8261 patients with ASCVD from the European Action on Secondary and Primary Prevention Intervention to Reduce Events V (EUROASPIRE V) [9] registry demonstrated that only 29% (2396/8261) of these patients met their LDL-C target. The majority of these patients (> 80%) were under specialist care (Cardiologist, Cardiac ANP, Diabetologist/endocrinologist).

The ‘EU-Wide Cross-Sectional Observational Study of Lipid-Modifying Therapy Use in Secondary and Primary Care’ (DAVINCI) [8] study examined LDL-C compliance in both secondary and primary prevention cohorts, and further stratified these patients as being managed in either primary or secondary care. Of those at ‘very high risk’, 39% (795/2029) met their absolute LDL-C targets, with fewer patients treated in primary care meeting their LDL-C targets when compared to those under specialist supervision (35% (188/526) vs. 41% (620/1513).

Table 6 compares the results of the current study with both the EUROASPIRE and DAVINCI cohorts.

Table 6 Comparison of LDL-C target achievement between those who completed CR in 2019 with the EUROASPIRE and DAVINCI studies
Full size table

Estimated required lipid-lowering therapies

It was a notable finding that 13% (22/163) of our cohort would likely require a PCSK9i combination therapy to achieve their absolute LDL-C targets. While none of these patients met the ESC criteria for this combination as they had not been trailed on a high intensity statin and ezetimibe, this finding is likely significant given the cost of these therapies [10] and the strict reimbursement criteria outlined by the Health Services Executive (HSE) in Ireland [11]. The current HSE reimbursement criteria for those with ASCVD require a LDL-C > 4 mmol/L while maintained on a high intensity statin and ezetimibe combination [11]. None of the 22 patients identified in our cohort met the current reimbursement criteria due to either an LDL-C that was already < 4 mmol/L on current therapies (91%; 20/22) or an LDL-C level that would likely fall below 4 mmol/L with up-titration of their LLT to a high intensity statin and ezetimibe combination (9%; 2/22).

As a corollary, the prescription of a high intensity statin to all patients in a ‘very high’ risk group is estimated to have led to at least an additional 12% (19/163) of patients, who completed CR with LDL-C levels above their recommended level, achieving their absolute LDL-C targets. The addition of ezetimibe to high intensity statin therapy in select patients who were unlikely to meet their targets with a high intensity statin alone is estimated to have led to a further 14 patients meeting their absolute LDL-C targets (14/163; 9%). Combined, these interventions are likely to have led to a further 20% (33/162) of patients to meet their LDL-C targets.

As such, while it is likely that a sizable minority of patients are likely to require a PCSK9i combination to meet the targets outlined by the ESC, increased prescription of high intensity statins and ezetimibe combination therapy may represent a more cost-effective area of focus to ensure a greater number of patients meet their LDL-C targets.

Limitations

The single centre, retrospective design of this study poses several limitations. As there was no systematic protocol in place for the assessment and recording of LLT compliance or side effects, it is not clear to what degree this may have influenced DL-C target achievement or choice of LLT. Additionally, this may have led to misclassification bias for those who were non-compliant. The sample sizes of the 2016 and 2019 cohorts are unequal, due to the suspension of CR services in Jan. 2020 as a result of the COVID-19 pandemic, which may reduce the statistical power of the study. Information regarding baseline LDL-C, and LLT was missing for a proportion of patients, which may have confounded our results. While a rapid implementation of new ESC guidelines within the CR unit was confirmed, the present study cannot account for improvements in practice that may arise due to increased familiarity practicing within these guidelines. Estimates for the ‘predicted required LLT’ are based on the average LDL-C lowering effects of each LLT as outlined by the ESC and do not account for the between-person variability of treatment effect.

Conclusion

With the introduction of lower absolute LDL-C targets and the additional requirement for a reduction in LDL-C levels > 50% from baseline, fewer patients attending CR are meeting the LDL-C targets set out in the 2019 ESC dyslipidaemia guidelines. LDL-target achievement following CR in our centre compares favourably with international standards of usual primary or secondary care, and patients likely benefit from specialist input. Both high intensity statin therapy and ezetimibe are under prescribed. Of the minority of patients likely to require PCSK9i combination therapy to meet their LDL-C targets, none currently qualify for reimbursement in Ireland.