A Conversion Protocol for Simvastatin and Gemfibrozil

Relevant data were collected, using the VistA (Veterans Health Information Systems and Technology Architecture) and CPRS medical record documentation systems. Patient demographic data, including age, gender, treatment indication, past medical history, adverse drug reaction history, and prescription history were collected for analysis of the study population baseline characteristics. In addition, pertinent laboratory data were examined, including lipid parameters (total cholesterol [TC]; LDL-C; high-density lipoprotein cholesterol [HDL-C]; and TG) and liver function (LF) markers (aspartate aminotransferase [AST]/alanine aminotransferase [ALT]), both before and after the intervention. Patient charts were manually reviewed at follow-up for any adverse events (AEs) attributed to the cholesterol medications.

The primary endpoints included the average change from baseline in lipid parameters, including TC, LDL-C, HDL-C, and TG in each group. Secondary endpoints included descriptions of any safety concerns or AEs. The study was approved on December 15, 2011, by the Medical University of South Carolina internal review board and the VA research and development committee. Statistical analysis was completed, using the paired Student t test for continuous data, and a descriptive analysis was performed for AEs.

Results

Initial patient retrieval included 151 patients who had received prescriptions for gemfibrozil and simvastatin during 2010-2011 and who also had a TG level of ≤ 150 mg/dL obtained in the previous year. Of these patients, 32 patients were missing laboratory data, 20 patients had a lifetime history of a TG level of > 500 mg/dL or pancreatitis, 9 patients had off protocol medication changes, and 1 patient had a nonfasting lipid panel. A total of 62 patients were excluded from the final analysis (Figure 2).

Eighty-nine patients were included in the primary analysis; 8 of those patients were meeting their TG goal but had an LDL-C elevated from goal at baseline, and 81 patients were meeting both their TG and LDL-C goals at baseline. The baseline demographics were reflective of a typical VA population: 99% male, average age 67.3 years, 38% white and 10% African American (52% did not disclose a racial identification).

The primary efficacy goal of change in average lipid parameters after intervention was assessed at 6 to 24 weeks postintervention. Eighty-one patients who met both TG and LDL-C goals at baseline were evaluated (Table 1, Figure 3). Average TC, LDL-C, HDL-C, and AST were not significantly different before and after intervention. There was a statistically significant difference in the average TG levels preintervention and postintervention (107.5 mg/dL vs 159.5 mg/dL; P < .001). Average ALT was significantly higher in the postintervention group (19.6 vs 27; P < .001).

In patients with TGs ≤ 150 mg/dL who did not meet the LDL-C goal at baseline, there was also a statistically significant difference in average TG levels pre- and postintervention (107.6 mg/dL vs 156.1 mg/dL; P = .01). Average LDL-C did not significantly change pre- and postintervention (119.6 mg/dL vs 119.1 mg/dL; P = .82), nor did TC, HDL-C, or AST/ALT (Table 2, Figure 4).

There were no AEs reported in the 6 to 24 weeks following the intervention that could have been attributed to the cholesterol medications. In addition, there were no new diagnoses of pancreatitis entered in the patient’s medical records in the follow-up period. Of note, the range of TG levels after discontinuation of gemfibrozil was 38 mg/dL to 341 mg/dL. After the intervention, 17 of 89 patients (19%) had TG levels of > 200 mg/dL; however, only 1 of 89 patients had a TG level of > 300 mg/dL. No patients had a measured TG level of > 500 mg/dL after the intervention.

Discussion

There was a statistically significant increase in TG concentrations in both groups after discontinuation of gemfibrozil, regardless of change in statin dose. However, the clinical significance of this increase is debatable. The NCEP ATP III guidelines define a TG goal of < 150 mg/dL, and in this study, the average TG level after discontinuation of gemfibrozil was 159.5 mg/dL in the group meeting LDL-C goal and 156.1 mg/dL in the group not meeting LDL-C goal. Following this strict definition, patients did not maintain TG control after the study intervention. Nevertheless, 80% of patients moved from the “normal” category (TG < 150 mg/dL) preintervention to the “borderline high” category (TG 150-200 mg/dL) postintervention per ATP III definitions.³

The importance of TGs as an independent marker for cardiovascular risk has been debated for decades. Data from the Copenhagen City Heart Study indicated that plasma TG levels were significantly associated with increased risk of nonhemorrhagic ischemic events. The relative risk (RR) of ischemic stroke was increased 1.12 (95% confidence interval [CI], 1.07 – 1.16) for every 88.6 mg/dL increase in TGs in that population.¹⁴