J. David Schnatz, M.D.
Director, Lipid Education Service
Saint Francis Hospital and Medical Center
114 Woodland Street, Hartford, CT 06105
Telephone: (860) 714-5555

September 1999

Baseline Lipids

In considering the effect of baseline lipids, it is important to consider the context in which this analysis of the CARE Study was published. In 1994, the 4S study showed benefit in cardiovascular events when patients with CHD and a mean LDL of 188 mg/dl were treated with Simvastatin (4). In 1995, a publication from the 4S study group presented an analysis of the effect of baseline lipids on risk reduction by Simvastatin (5). Risk reduction was the same for all quartiles of total and LDL cholesterol. Recently, the CARE Study group published an analysis of their results with Pravastatin Rx when baseline lipids were taken into account for their patients who had lower total (209 mg/dl) and LDL (139 mg/dl) cholesterol values than the 4S Study (6).

In the current analysis, baseline CHOL, LDL, HDL and TG were grouped in quartile ranges. These quartile values were used to determine the effect of baseline lipids on CHD with and without Pravastatin Rx. For the CHOL range of the CARE Study (159-239 mg/dl), the baseline value did not correlate with the risk of a CHD event. For the range of LDL values (115-174 mg/dl), there was a significant relationship. For the placebo assigned group (Plc), there was a 28% increase in relative risk of CHD death or MI for every 25 mg/dl increment of baseline LDL (p=0.015) whereas for Pravastatin assigned patients (Prav), the 6% increase in risk was not significant. This trend led to a convergence of event rates between Plc and Prav at lower LDL values.

HDL did not vary across LDL quartiles and TG decreased with increasing LDL. However, there was an 11% increase in CHD death or MI with each 10 mg/dl decrement in HDL (p=0.049). This trend was similar for both Plc and Prav groups. The benefits of Pravastatin were comparable in each HDL quartile when compared to Plc. No significant relation between baseline TG and CHD events were seen.

A post-hoc analysis was also done to compare the results and those from the 4S Study, previously mentioned. The eligibility criteria for 4S, CHOL equal to or greater than 212 mg/dl and TG < 220 mg/dl, were used to designate the CARE patients as "4S lipid-eligible" or "4S lipid-ineligible". Utilizing these criteria, 34% of the CARE Study group (1409 patients) were 4S lipid-eligible. The 4S eligible and ineligible groups were comparable in all respects, but the fact that the "eligible" group had significantly less men (80.0% vs 89.3%), a lower incidence of diabetes (10.7% vs 15.9%), took less beta blockers (37.3% vs 40.8%) and took more calcium channel blockers (42% vs 37.8%) than the 4S Study lipid-ineligible patients. Considering the Plc group, the 4S lipid-eligible patients had a greater chance of CHD death or MI than the ineligible patients (15.4% vs 12.0%). Pravastatin reduced the event rates in both groups, but the risk reduction was greater in the 4S lipid-eligible group, 32% vs 18% for the ineligible group.

In drawing conclusions from the present analysis, the CARE investigators concluded that both Pravastatin and Simvastatin benefit a broad range of patients with CHD and that "our analysis suggests that the efficacy of HMG therapies in patients with coronary artery disease with low baseline LDL cholesterol values may be more limited". The issue of efficacy with low baseline LDL values remains unsettled and a subject of current investigation.

Inflammation

In a post-hoc analysis, C-Reactive Protein (CRP) and Serum Amyloid A (SAA) were measured in the pre-randomization blood samples of 391 patients who had a subsequent MI or CHD death (cases). An equal number of age and sex matched study participants who did not develop a CHD event during 5 years of the study were the controls. The CHD cases had a significantly higher incidence of smoking (2x control, p=0.001) and of diabetes (2.3x control, p=0.001) and a higher body mass index (28.2 vs 27.3, p=0.01).

Comparing cases with controls, median CRP was higher in the cases who developed a CHD event (0.31 vs 0.28, p=0.5) as was the SAA (0.34 vs 0.28, p=0.006). When the CRP and SAA were analyzed by quintile, significant trends in relative risk were seen (p=0.044 for CRP and 0.006 for SAA). However, only the 5th quintile had a significantly increased risk for both CRP and SAA when each quintile was examined by itself. Significance of the results was maintained when controlled for smoking. The investigators state that "Further adjustment for other risk factors including diabetes had minimal impact on these risk estimates". Of the 782 study participants, 708 had concordant CRP and SAA evidence of inflammation. These were grouped into placebo (Plc) and Pravastatin Rx (Prav) with an without inflammation. The Plc group with inflammation had the highest relative risk (2.8) vs without inflammation (1.3, p=0.007). In the group with inflammation, Prav reduced the current CHD events by 54% compared to 25% in those without inflammation, despite the fact that baseline lipids were identical in both groups.

The data on inflammation extend other observations that CRP has predictive value in relation to future CHD events. The study is limited because of the post-hoc analysis as well as the differences between the case and control groups. Yet, it suggests that Pravastatin may attenuate the effect of inflammation on CHD risk. The authors state that, "It is tempting to speculate that nonlipid effects of pravastatin may be responsible for the current observations". Since there appears to be a number of risk factors other than LDL, this publication highlights the possible importance of inflammation in future CHD events and the effect that Rx with Pravastatin has on these events.

References
1. NEJ Med 335: 1001, 1996
2. Ann Int Med 129:681, 1998
3. Circulation 98:2513, 1998
4. Lancet 344:1383, 1994
5. Lancet 345:1274, 1995
6. J. Am Coll. Cardiol. 33:125, 1999
7. Circulation 98:839, 1998