Journal of Cardiovascular Diseases & Diagnosis Statins as a Primary Prevention: Which One is Most Effective? A Systematic Review and Meta-Analysis

Background: Type 2 diabetics are at high risk of atherosclerosis and Cardiovascular (CV) events due to their associated dyslipidemic profile. Adverse levels of the lipid triad: Low-Density Lipoprotein Cholesterol (LDL-C), High-Density Lipoprotein Cholesterol (HDL-C) and Triglycerides (TGs), have been shown to be major contributing factors. Statins (HMG-CoA reductase inhibitors) are recommended as first-line lipid therapy as evidence has shown their effectiveness in primary and secondary prevention of CV events. However, there is limited evidence specifically focused on primary prevention in Type 2 Diabetes Mellitus (T2DM) and dyslipidemia. The main purpose of this review was to assess the efficacy of statins in type 2 diabetics with dyslipidemia for the primary prevention of CV events. Methods: A systematic review was carried out to determine the effectiveness of statins in T2DM and dyslipidemia. A comprehensive literature search identified Randomised Control Trials (RCTs) specific to participants with T2DM and dyslipidemia without prior CV events, which investigated the effects of statins on the lipid triad. Bias and quality of the studies were assessed and data extracted was critically analysed. Meta-analyses were carried out to determine the significance of statin effects. PRISMA guidelines were adhered to throughout. Results: 10 RCTs were included investigating the effects of six statins. Statins were shown to improve the lipid triad by lowering LDL-C and TGs and increasing HDL-C, in agreement with previous research. Atorvastatin and rosuvastatin were shown to be most potent in reducing LDL-C and TGs and simvastatin showed the greatest increases in HDL-C. Meta-analyses demonstrated significance favouring statins for LDL-C and TG reductions but less evidence of significant effect of HDL-C improvement with all statins. Statins were well tolerated and adverse effects were minimal. Conclusion: Primary prevention in this high risk population is very important. Statins are effective at reducing CV event risk, and clinicians should have knowledge of most appropriate statin treatments for patients based on individual lipid levels.

need to deliver care to this high risk group in order to reduce premature mortality [7]. T2DM is considered 'a cardiovascular disease per se' and evidence suggests prevention should be sought as though CVD is present, even without signs of manifestation [8]. Studies have shown T2DM patients without CVD are just as at risk of CV events as those without DM with CVD.
First-line pharmacological intervention to improve lipids and prevent CV complications in T2DM is thus the use of statins. LDL-C, HDL-C and TG levels are all high risk factors for cardiac events in diabetics with dyslipidemia, so aim of treatment is to improve all. High risk levels include: LDL-C ≥130 mg/dL; HDL-C ≤ 40 mg/dL; and TGs>150 mg/dL [3,9]. Current statins available are: Atorvastatin, Fluvastatin, Pravastatin, Rosuvastatin and Simvastatin [10]. Statins reduce cholesterol synthesis and up-regulate LDL receptor synthesis, which promotes cholesterol removal from the blood [5]. Statins are well tolerated with generally mild side effects [11]. The potentially fatal side-effect is myositis, leading to rhabdomyolysis and progressing to renal failure.
Primary and secondary CVD prevention trials showed statins enabled significant CV event reduction. There was limited information on primary prevention in DM until results from diabetes subgroups of the Heart Protection Study and the Collaborative Atorvastatin Diabetes Study [12] (Table 1). Both reduced CV risk significantly.
However, T2DM participants recruited did not on average have dyslipidemic profiles typical of T2DM manifestation. Participants may not have been at such high risk as dyslipidemic T2DM patients, as lipid complications may not yet have risen sufficiently to cause CV damage. It is questionable whether their findings would have been so beneficial if participants had more adverse lipids.
LDL-C, HDL-C and TGs all contribute to atherogenesis in T2DM, so guidelines for lipid levels need to be targetted. The American Diabetes Association recommends target treatment goals for T2DM with statins include: LDL-C <100 mg/dL; HDL-C >40 mg/dL (men) and >50 mg/ dL (women); and TG levels <150 mg/dL [9]. From in-depth research on primary preventative trials in T2DM, it was discovered there are few studies that focus solely on this lipid triad in T2DM and dyslipidemia.
This Review focuses on statins as a primary prevention method to determine which statin is most effective in improving LDL-C, HDL-C and TG levels. By using participants with high-risk lipid profiles this Review will determine, which statin is greatest at achieving ADA target levels and at what dosage the most beneficial statin should be administered.

Methods
A SR of randomized controlled trials (RCTs) was carried out. This SR centred specifically on T2DM. Participants were required to have  or have an aspect of a dyslipidemic profile as defined by LDL-C ≥ 130 mg/dL; HDL-C <40 mg/dL and TG levels >150 mg/dL. The focus was primary prevention of CV events so only participants without macrovascular conditions were included. All ages, sex and ethnicities were included to ensure findings were globally relevant to healthcare. Dose and duration of treatment was not specified. A comprehensive database search was carried out to identify eligible RCTs from the years 2000 to 2012.
Primary outcomes were LDL-C, HDL-C and TG levels after statin treatment and whether ADA targets were achieved. Secondary outcomes included adverse effects of statin treatments. This was important as statins may be effective at improving lipids but have side effects disadvantageous in administration.
The Cochrane Collaboration's tool for assessing risk of bias and the Critical Appraisal Skills Programme (CASP) tool were used to critically judge the sources. Using the CASP tool, each study was scored on a scale of 1-9, determining overall quality (low-high) [13].
Meta-analyses were carried out to provide precise estimates of overall treatment effects [14]. The I² test assessed heterogeneity, which showed the percentage of variability in effect estimates due to heterogeneity rather than chance.
Studies investigating simvastatin showed improvement in the whole lipid triad, but demonstrated greatest improvements in HDL-C compared to other statins. Sheu et al. [19] showed 20 mg enabled the largest HDL-C increase (  dose found to cause most significant reductions in LDL-C and TGs from the simvastatin studies (p<0.01) [15,19].

Primary outcome: ADA targets achieved?
The studies investigating atorvastatin showed variation in meeting ADA targets. However, LDL-C, TGs and HDL-C targets were met by the majority of doses studied. There were differences between simvastatin studies in achieving LDL-C targets. However, all doses investigated achieved HDL-C target and no dose reached TG target.
It was not possible to determine whether rosuvastatin 20 mg investigated by Vergès et al. enabled targets due to insufficient published data, however Betteridge and Gibson showed both rosuvastatin 10mg and 20mg were successful [20,21]. Bevilacqua et al. [22] demonstrated fluvastatin enabled LDL-C and HDL-C targets but not TGs. Gentile et al. showed no statin investigated achieved LDL-C or TG targets, but all met HDL-C target [17].

Secondary outcome: Adverse effects
Studies investigating atorvastatin presented most adverse effects, although numbers of participants affected were minimal. Simvastatin treatment was reported as well tolerated. Vergès

Discussion
This SR and meta-analysis showed support for statin administration to prevent CV events as they improved the lipid triad and aided achievement of ADA targets. However, issues were raised that contribute to knowledge of statin efficacy in T2DM.
Lam et al. found neither atorvastatin 10/20/40 mg improved HDL-C. In fact, HDL-C decreased with all doses and Lam et al. stated this was "not a negative dose-response effect of atorvastatin on HDL-C as previously reported [18].
Negative dose-response describes progressively decreasing HDL-C levels as dosage increases [12]. Wierzbicki and Mikhailidis found low dose atorvastatin increased HDL-C and ApoA-1, its associated apolipoprotein, but effects diminished with increasing dose suggesting negative dose-response [23]. However, these findings were not based on T2DM. They also found 10mg still increased HDL-C, however, another study showed 10 mg decreased HDL-C, which suggests this latter study and researchers were correct in their statement [18,23].
Similarly, Lawrence et al. found HDL-C decreased with atorvastatin 80 mg [16]. They found significant negative effect on ApoA-1 and HDL-C (p<0.05) demonstrating high-dose atorvastatin potentially limits anti-atherogenic protection HDL-C provides. However, in contrast The DALI study group showed significant increases in HDL-C for both doses of atorvastation (10/80 mg) (p<0.005) [15]. The increase in HDL-C reduced the total cholesterol (TC): HDL-C ratio, which prevents against atherosclerosis.
Heljić et al. study demonstrated 6 months of simvastatin 40mg did not improve lipids as effectively as 80 mg or 20 mg [19,24,25]. Evidence has shown 40mg can have greater effects in less time than shown by Heljić's study. In hypercholesterolaemic participants Miller, Dobs, Yuan et al. found 40 mg improved lipids more significantly after six weeks and Li and colleagues demonstrated the same in two weeks [26,27]. Therefore the results from Heljić's study were potentially an anomaly.
It could be stated the results were due to obese participants. Increase in central adiposity decreases LPL activity and results in increased production of atherogenic particles [ 3 ]. Isley et al. [25] also used obese participants, but they were consuming a low-fat, low-cholesterol diet which may explain their improvement in LPL activity and therefore the lipid triad [23]. This could provide explanation for the differing results found by Heljić's study, whose participants were not diet-controlled.
Sheu et al. study showed the highest HDL-C increase (11%) with simvastatin 20mg, despite being the lowest simvastatin dose and the shortest simvastatin study [19]. However, Sheu stated treatment dose was doubled after 4 weeks if participants' LDL-C exceeded 130 mg/dL. Without individual participant data, it was not determined how many participants this involved. The results of Sheu may have represented the effects of simvastatin 40 mg as opposed to 20 mg.
Vergès et al. study found rosuvastatin 20 mg did not improve HDL-C and gave no explanation why [20]. However, they found treatment significantly increased total catabolism of VLDL. This showed, although rosuvastatin 20 mg did not increase HDL-C, it prevented HDL-C becoming TG-enriched and from becoming cleared more easily. Therefore rosuvastatin 20 mg had indirect effects, suggesting if statins can prevent HDL-C decreasing, this is beneficial even if levels are not increased. Betteridge and Gibson [21] also found rosuvastatin was not significant at increasing HDL-C. However, they did not discuss VLDL catabolism so it cannot be certain rosuvastatin consistently shows the same benefits [26].
Bevilacqua et al. demonstrated fluvastatin XL-80mg increased HDL-C but simvastatin 20 mg did not [22]. This contrasted to the simvastatin studies that demonstrated its effects on the whole lipid triad. However, despite dissimilar HDL-C findings between fluvastatin and simvastatin, improvements in ApoA-1 were similar (+8 mg/ dL, +13 mg/dL, respectively). Therefore lack of HDL-C increase with simvastatin 20 mg was possibly a discrepancy.
The duration of studies had an impact on achieving ADA targets. The studies by Lam and Lawrence showed atorvastatin 10 mg/80 mg did not achieve LDL-C targets, however, The DALI study group found they did, which was the longest study of the three. Similarly Isley [25] and Betteridge and Gibson [21] found simvastatin 80 mg A randomized, doubleblinded, placebocontrolled, cross-over trial was carried out to determine the effects of rosuvastatin versus placebo on the metabolism of TG-rich lipoproteins.
8 participants were recruited who had T2DM and diabetic dyslipidemia characterised by TG >1.70mmol/l; and HDL <1.03mmol/l in men and <1.29mmol/l in women. They had been treated with oral glucose lowering agents for at least 6 months and had stable HbA1c over this time.
Participants undertook a 4 week placebo maintenance period and were then randomized to either rosuvastatin 20mg or placebo once daily for 6 weeks. There was then a crossover for a further 6 week treatment period.
Outcomes consisted of glucose and lipid parameters and ApoB-100 kinetic parameters.
Rosuvastatin had benficial effects on the catabolism of TG-rich lipoproteins, aiding prevention of atherosclerosis in T2DM and dyslipidemia. and atorvastatin 10 mg, respectively, were near to TG target, but had treatment been longer target may have been met. Baseline lipid levels also had an impact. Very high baseline TGs were demonstrated by The DALI study group and Bevilacqua [22], which could have affected how quickly TG levels were reduced.
ADA targets for HDL-C differ for men and women (>40 mg/dL, >50 mg/dL), respectively. The RCTs consisted of male and female participants but individual participant data was unavailable. The target value of >40 mg/dL was used to represent the mixed sex samples. HDL-C targets were met by atorvastatin 10 mg/80 mg however it must be emphasized baseline HDL-C for these groups were higher than final measurements, so atorvastatin did not enable target achievement [19,20]. These shows T2DM patients with lower HDL-C would not necessarily benefit from atorvastatin to increase HDL-C.
One participant withdrew from Lawrence study because of increases in Alanine Transaminase (ALT). ALT is a liver enzyme and elevations indicate liver dysfunction. Ballantyne et al. found hypercholesterolemic participants treated with atorvastatin 80mg were more likely to have ALT elevations [28]. Neither Lam study nor The DALI study group reported this with lower doses of atorvastatin so elevation could be related to higher dosage. This is supported by Armitage who stated effect of statins on transaminase is dependent upon dose [29]. Betteridge and Gibson [21] stated incidence of myalgia was low for both atorvastatin and rosuvastation, but consistent with other clinical trials [30]. However, if a consistent finding, this may present safety risks of taking either.
Only a small number of RCTs were available to examine for this SR. More studies would have been sufficient to confirm or disagree with the findings. Two studies used Taiwanese participants so their results cannot be generalised globally due to cultural differences in lifestyle and genetics. Three studies used overweight participants so also cannot be generalizable to all with T2DM. However, obesity is a contributing factor to T2DM development and due to increasing obesity levels findings are applicable to overweight patients. The studies varied in lengths of T2DM diagnosis. If participants had T2DM a long time, baselines were possibly less adverse than if recently diagnosed due to treatments commenced. The studies varied in durations and number of participants. This heterogeneity influenced the findings and potentially impacted upon how successful statins appeared.

Implications and recommendations for clinical practice
The results from this Review imply T2DM patients should have lipid levels measured early. It is important treatment is not standardized with simvastatin 40mg, as recommended, because patients' levels may vary. How long patients go undiagnosed, due to lack of symptoms, can  For those with high LDL-C and TGs, atorvastatin or rosuvastatin may be beneficial as a primary preventative treatment. Levels are likely to reduce quickly due to their potent effects. High doses may be most effective, but caution should be taken in case of adverse effects. In contrast, those with less high risk profiles may benefit from simvastatin. Simvastatin showed overall benefit with all doses, but 80mg may present more significant effects in this population.
Care of T2DM dyslipidemic patients should be individualised. It is important to determine whether treatment is improving patients' lipids. Monitoring lipid triads and reviewing statin treatment regularly would assist in CVD prevention.

Conclusion
T2DM patients can have very adverse lipid profiles. Baseline lipids of participants in studies systematically reviewed emphasised the importance of primary prevention. As well as simvastatin, as recommended by NICE, this SR has shown the efficacy of atorvastatin and rosuvastatin. Atorvastatin and rosuvastatin were most potent in reducing LDL-C and TGs, whereas simvastatin showed the greatest increases in HDL-C. Meta-analyses demonstrated significance favouring statins for LDL-C and TG reductions, but showed less evidence of significant effect for HDL-C improvements. The ten studies showed overall statins were well tolerated and adverse effects were minimal.
Future research may consider studies all of the same intervention duration, in order to reduce heterogeneity. Research may be warranted on whether atorvastatin and rosuvastatin improve HDL-C in T2DM and dyslipidemia, and whether negative dose-responses associated with atorvastatin, and increases in ApoA-I with rosuvastatin are consistent findings. High-dosage fluvastatin showed significant benefits in this review, but further research in this population would be needed to conclude whether it could be a CV prevention measure.