Effect of Lipid Management on Coronary Heart Disease Risk in Patients with Diabetes

Coronary Heart Disease Risk Among Patients with Diabetes and Dyslipidemia

To better understand CHD risk for patients with versus without DM in the context of lipid management, it is informative to analyze baseline characteristics of DM patients participating in landmark lipid-lowering clinical trials such as trials focused exclusively on patients with DM, or those that report data on a sufficient number of patients making up DM subgroups. For the present summary, data were considered from trials with a high proportion of patients with DM at baseline (> 15%) and/or trials that enrolled at least 100 patients with prevalent DM. Data from 47 lipid trials comprising 198,930 patients, of whom 65,558 had prevalent DM, are summarized in Tables 15-1 (alphabetically by trial name acronym) and 15-2 (ordered by descending numbers of participants with DM), with abbreviations used to describe respective trial outcomes defined in Table 15-3 . Except for the Acute Coronary Syndrome Israeli Survey (ACSIS), all studies were randomized controlled trials, most of which evaluated monotherapy with a statin or a fibrate versus placebo. A few randomized controlled trials studied a combined lipid-lowering intervention (statin plus fibrate; statin plus ezetimibe); Steno-2 was a randomized comparison of a multifactorial intervention versus usual care, with statins and/or fibrates used as part of a comprehensive global cardiovascular disease (CVD) risk intervention strategy in patients with DM and albuminuria.

The baseline characteristics of patients participating in clinical trials and substudies that included only patients with DM (n = 46,326 patients) are described in Table 15-4 , and ranked by baseline low-density lipoprotein (LDL) cholesterol (LDL-C). Mean age at entry was 60.3 years and, similar to lipid intervention trials not focusing specifically on DM, men accounted for more than two thirds of the patients. Most DM patients in these studies are assumed to have had type 2 diabetes mellitus (T2DM) given the relative prevalence of T2DM versus type 1 diabetes mellitus (T1DM) in the age groups studied (see Chapter 1 ), with a few trials specifically allowing inclusion of patients with T1DM, and most trials not differentiating DM type for eligibility.

Average baseline lipid levels in the trials surveyed did not meet contemporary targets for lipid management in DM patients (see later), with mean LDL-C 129 mg/dL (3.3 mmol/L); non–high-density lipoprotein (HDL) cholesterol (non–HDL-C) 166 mg/dL (4.3 mmol/L); HDL-C 42 mg/dL (1.1 mmol/L); and triglycerides (TGs; triacylglycerols) 180 mg/dL (2.0 mmol/L). In most diabetes trials and DM subgroup analyses, LDL-C was calculated, with routine direct measurement in only three studies: 4D; HPS—MRC/BHF (DM subgroup); and PROACTIVE. ^{, , ,} Mean baseline apolipoprotein B100 (apo B) concentration was 115 mg/dL, a value also beyond generally accepted targets, as inferred from studies in which baseline apo B level was available in the reports. *

* References , , , , , , , , .

In addition to elevated and/or unsatisfactory LDL-C levels at study entry, the frequent findings of low HDL-C together with elevated fasting TGs are consistent with the assumption of a high prevalence of atherogenic dyslipidemia in these DM patients at enrollment (see Table 15-4 and Chapter 10 ). Similarly, data from studies in which baseline apo B was measured concomitantly with LDL-C reveal a high prevalence of increased small, dense LDL particles—a pattern commonly observed in T2DM (see Chapter 10 ).

Given the high cardiovascular (CV) risk associated with DM (see Chapters 7 , 9 , and 10 ), observed event rates in reported lipid trials are influenced by the proportion of patients with DM enrolled. The rates of primary CV outcome events among patients with DM across lipid intervention trials in primary prevention, secondary prevention, and post–acute coronary syndrome (ACS) patient populations are summarized in Table 15-5 , arranged by decreasing hazard, with significantly higher rates among those with versus without DM across the spectrum of clinical indication.

To estimate CHD risk or to better characterize it for specific patients groups, there are other complementary determinations in addition to total cholesterol and LDL-C levels measurements. Non–HDL-C, apo B, and LDL particle (LDL-P) concentration are closely associated with obesity, DM, insulin resistance or hyperinsulinemia, and other markers of dysmetabolism in conditions of increased cardiometabolic risk, such as T2DM. Their determination, in addition to LDL-C measurement, may provide information before and after introduction of lipid-lowering therapies, to assess CV risk at baseline and residual vascular risk after intervention. Compared with on-treatment LDL-C, these may help to clarify some aspects of risk and response to therapy, but their combined measurement is not routinely recommended in DM.

In addition to non–HDL-C, apo B, and LDL-P determination, screening for atherogenic dyslipidemia, before any lipid-lowering intervention, is an easy and inexpensive means to determine residual vascular risk associated with low HDL-C, high TGs, and their determinants. As a result of lack of agreement on cutoffs for HDL-C and TGs, this is rarely performed in routine practice. An alternative approach to defining atherogenic dyslipidemia as the combined occurrence of high TG levels plus low HDL-C uses the ratio of TG to HDL-C. Because both TG and HDL-C levels are continuous risk variables with mutually additive associations with residual vascular risk, computing a ratio of fasting TGs to HDL-C provides a summary metric for the severity of atherogenic dyslipidemia, calibrated as a continuous rather than a dichotomous variable. Given the extreme right skewness of TG values, assessing atherogenic dyslipidemia using log(TGs)/HDL-C may be more clinically informative than an untransformed ratio. ^,

Therapeutic Lifestyle Changes

The cardiometabolic abnormalities underlying CHD risk in T2DM and their potentially reversible components under the influence of various therapeutic lifestyle changes make the population with these abnormalities particularly suitable to respond positively to such interventions, provided they are applied early (i.e., in primary prevention) and maintained long term (see Chapters 5 and 12 ). Long-term compliance with therapeutic lifestyle changes must be optimized and regularly assessed and reinforced, because it is often difficult for patients with DM to comply long term with dietary and lifestyle advice. This is all the more relevant because the hallmark of atherogenic dyslipidemia (low HDL-C and high TGs) and all three other defining components of the metabolic syndrome are responsive to therapeutic lifestyle changes (see Chapters 4 , 5 , and 12 ).

Lifestyle approaches and dietary strategies to lower LDL-C and TGs and raise HDL-C as well as the effects of dietary carbohydrate restriction on atherogenic dyslipidemia, fatty acid partitioning, and metabolic syndrome have been previously reviewed. ^, The American Diabetes Association (ADA) recommends that DM patients on low-carbohydrate diets undergo lipid profile monitoring and that the ratios of dietary intake of carbohydrates, proteins, and fat be individually adjusted to the metabolic requirements and preferences of patients. With regard to dietary intake of saturated and trans -saturated fatty acids, both modulators of LDL-C levels, there are no specific data available for patients with DM, and recommended dietary goals for DM patients are currently those, by default, of individuals with established CHD. Thus, saturated fat intake should not exceed 7% of total calories, and intake of trans -saturated fatty acids should be reduced.

Unfortunately, for most T2DM patients in real-life conditions it is very hard to follow the current recommendations regarding dietary and physical activity for DM for the long term and/or with the required intensity. This underscores the importance of systems-based lifestyle interventions (see Chapter 12 ). Even for those who are able to implement lifestyle changes, many will not achieve sufficiently low LDL-C, non–HDL-C, and apo B and/or improve atherogenic dyslipidemia components through therapeutic lifestyle changes alone and will require lifelong therapy with one or more lipid-lowering drugs. For example, in the Action for Health in Diabetes (Look AHEAD) trial, 5145 overweight or obese adults with T2DM were randomized to intensive lifestyle intervention focusing on weight loss and increased leisure-time physical activity versus standard care diabetes support and education. Despite greater and sustained weight loss, greater reductions in glycated hemoglobin, and improvements in fitness in the intensive lifestyle arm, there was no difference achieved between the groups in LDL-C. The overall trial failed to demonstrate mortality improvement, as discussed in Chapter 12 .

Drugs Targeting Low-Density Lipoprotein Cholesterol

Given the predominant role of LDL-C as the major lipid-related modifiable risk factor for CHD in nondiabetic and diabetic patients, most of the pivotal studies have investigated the benefit on CHD outcomes of pharmacologic agents whose main effect is to reduce LDL-C levels ( Fig. 15-1 ). Most of the clinical evidence of a beneficial effect on CHD of a decrease in total cholesterol and/or LDL-C was derived from landmark clinical trials with statins in nondiabetic and diabetic cohorts. In contrast, it is currently not established whether other nonstatin drugs specifically targeting LDL-C (ezetimibe; bile acid binders; proprotein convertase subtilisin/kexin type 9 [PCSK9] inhibitors) ( Fig. 15-2 ), or with an LDL-C– lowering component among other lipid- and lipoprotein-modulating effects (niacin, fibrates) have a beneficial influence on cardiovascular events in nondiabetic or diabetic patients.

3-Hydroxy-3-Methylglutaryl-Coenzyme a (HMG-CoA) Reductase Inhibitors: the Statins

Statins inhibit the 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase, upregulating hepatic expression of LDL receptors and increasing uptake of circulating LDL ( Fig. 15-3 ), and thereby decreasing circulating cholesterol (total cholesterol, LDL-C, and non–HDL-C) as a result of lowered LDL-P numbers. As a direct consequence, statins also reduce levels of apo B, the major atherogenic apolipoprotein, of which a single structural molecule is present on each LDL-P, as well as on each of their TG-rich lipoprotein precursors (very low-density lipoproteins [VLDLs]; intermediate-density lipoproteins [IDL]; and VLDL remnants).

In general, patients with and without diabetes respond similarly to statins with regard to LDL-C reduction, with response similarly dependent on drug choice, dose choice, and individual response. Proportional reductions in LDL-C levels on statin treatment range from 20% to 40% for less potent statins (fluvastatin, lovastatin, and pravastatin), 30% to 45% for simvastatin, and 40% to more than 50% for the most potent statins (atorvastatin, rosuvastatin, and pitavastatin). In the vast majority of diabetic and nondiabetic patients, long-term statin use is safe and effective.

Patients with Diabetes in Key Statin Trials

Numerous studies have demonstrated the effectiveness of statins to reduce primary CHD outcomes in primary and secondary prevention settings and post-ACS events; the risk reduction after LDL-C lowering parallels the magnitude of the achieved LDL-C decrease in populations with and without diabetes. The studies having established the beneficial effect of statins on CHD risk in patients with DM selected for the present review comprised a total of 20,103 DM patients, followed for a mean duration of 4.0 years. Average (1 standard deviation) on-treatment LDL-C decreased to a mean 81 (standard deviation [SD] 18) mg/dL (2.1 [SD 0.5] mmol/L). This corresponds to absolute and relative reductions of 48 mg/dL (1.2 mmol/L) and 36%, respectively. With respect to non–LDL lipids, mean on-statin HDL-C was 46 (4) mg/dL (1.2 [0.1] mmol/L); non–HDL-C was 114 (19) mg/dL (3.0 [0.5] mmol/L); and TG was 144 (12) mg/dL (1.6 [0.1] mmol/L) ( Table 15-6 ).

Table 15-6

Statins Effects on Lipids and apo B and Primary Outcome (Ranked by Low-Density Lipoprotein Cholesterol [LDL-C] Reduction) in Diabetes Trials and Substudies

Acronyms	Active Arm (n)	Control arm (n)	Follow-up (yr)	Therapy	Dosage (mg)	Control	LDL-C	HDL-C	Non– HDL-C	TG	apo B	Delta LDL-C (mg)	Delta LDL-C (%)	Events Active Arm (n)	Annual Rate (%)	Events Control (n)	Annual Rate (%)	HR	95% CI	P	ARR (%)	RRR (%)	5-Year NNT
GREACE (DSS)	161	152	3	atorva	23.7	Usual care	97	43	123	142		− 92	− 49	20	4.1	46	10.1	0.41		< .0001	17.8	59	3
4S (DSS)	105	97	5.4	simva	20-40	Placebo	119	46	143	134		− 67	− 36	15	2.6	24	4.6	0.58		.087	10.5	42.3
4D	619	636	4	atorva	20	Placebo	72					− 53	− 42	226	9.1	243	9.6	0.96	0.77-1.1	.37	1.7	4.4
SPARCL (DSS)	395	399	4.9	atorva	80	Placebo	81	48	106	137		− 50	− 38
ASCOT-LLA (DDS)	1258	1274	3.3	atorva	10	Placebo	81	47	108	136		− 47	− 37	116	2.8	151	3.6	0.78	0.61-0.98	.036	2.6	22.2	25
CARDS	1428	1410	3.9	atorva	10	Placebo	70	55	99	140	90	− 47	− 40	83	1.5	127	2.3	0.65	0.48-0.83	.001	3.2	35.5	24
PROVE IT–TIMI 22 (DSS)	499	479	2	atorva	80	prava 40	57					− 44	− 44	142	14.2	152	15.9	0.88		0.28	3.3	10.3
AURORA (DSS)	388	343	2.8	rosuva	10	Placebo	54	45	83	146		− 43	− 39	85	7.8	104	10.8	0.72	0.51-0.90	0.008	8.4	27.7	7
CARE (DSS)	282	304	5	prava	40	Placebo	96	40	130	143		− 40	− 27	81	5.7	112	7.4	0.78		< .0001	8.1	22	12
HPS— MRC/BHF (DSS)	2978	2985	4.8	simva	40	Placebo	89	41	136	177	84	− 35	− 28	601	4.2	748	5.2	0.81	0.19-0.30	< .0001	4.9	19.5	20
ASPEN	1211	1199	4	atorva	10	Placebo	79	48	108	141		− 34	− 30	166	3.4	180	3.8	0.91	0.73-1.12	.341	1.3	8.7
TNT (DSS)	753	748	4.9	atorva	80	atorva 10	77	45	106	145		− 19	− 20	103	2.8	135	3.7	0.76	0.58-0.97	.026	4.4	24.2	22
Total (n = 20,103)	10,077	10,026												1638		2022
Mean			4.0				81	46	114	144		− 48	− 36		5.3		7.0	0.76			6.0	25	16

All lipids measurements correspond to on-treatment values (in mg/dL).

See Table 15-1 for acronyms definition and Table 15-2 for primary outcomes description.

ARR = Absolute risk reduction; atorva = atorvastatin; CI = confidence interval; NNT = number needed to treat; prava = pravastatin; rosuva = rosuvastatin; RRR = relative risk reduction; simva = simvastatin.

In the active arms of statin trials (n = 10,077), a total of 1638 primary outcome events (5.3%/year) were observed, versus 2022 outcomes (7.0 %/year) in the comparator arms (n = 10,026), with a weighted and adjusted hazard ratio (HR) of 0.76 (95% confidence interval [CI] 0.65 to 0.84) favoring statin treatment. As a class, for each 1 mg/dL (0.03 mmol/L) reduction in LDL-C achieved on statin, the HR of incident CHD was reduced on average by 0.5%. This translated into a 19.5% primary outcome reduction for every 1 mmol/L (40 mg/dL) decrease of LDL-C. The mean absolute risk reduction for composite major adverse CV events in statin trials was 6.0%; the mean relative risk reduction (RRR) was 25%. The average number needed to treat for 5 years to prevent one major adverse CV event was 16 patients. Qualitatively similar effects are evident when analyzing the component endpoints of all-cause mortality (HR 0.88; 95% CI 0.65-1.01) and fatal CHD events (HR 0.59; 95% CI 0.48-0.97), although pooled analysis of death alone failed to achieve statistical significance (see Table 15-6 ).

Meta-analyses of Statin Efficacy Among Diabetes Mellitus Patients

In the Cholesterol Treatment Trialists’ Collaboration (CTT) prospective meta-analysis of data from 90,056 participants in 14 statin randomized controlled trials (among whom 21% had DM), statin therapy safely reduced 5-year incidence of major adverse coronary events (MACEs) and coronary revascularization by approximately 20% per 40 mg/dL (1.0 mmol/L) LDL-C reduction, and largely independent of baseline LDL-C. Other meta-analyses have confirmed these observations that statins effectively reduce CHD outcomes with or without DM in both primary and secondary prevention populations, including meta-analyses focused only on statin efficacy among DM participants.

Patients with DM could possibly benefit even more from the cardioprotective effects of statins than those without DM, and the authors of the CTT meta-analysis of 18,686 diabetic patients from 14 randomized controlled trials went as far as to recommend considering statin therapy for all patients with DM and elevated risk for incident CV events, based on the 21% proportional reduction in MACE for every 40 mg/dL (1.0 mmol/L) LDL-C reduction. However, this may overestimate statin efficacy to some degree because of the exclusion from the analyses of data from the ASPEN trial, which was a DM-specific trial of atorvastatin that failed to demonstrate statistically significant differences in CV outcomes.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here