EBM Tools for Practice: Hypertriglyceridemia and the Risk of Cardiovascular Disease and Pancreatitis in a Veteran Population

The Veterans Health Administration (VA) is the nation’s largest integrated health care delivery system serving more than 11 million patients. There recently has been renewed emphasis on health care disparities in the US and within the VA including focus on unique Veteran demographics. Pertinent to this issue of LipidSpin is that in 2017, one quarter of the total Veteran population was residing in rural or highly rural communities including 31% of those relying on the VA for health care. The number of Veterans living in rural communities increased by 7% between 2006 and 2014, and this number is expected to continue rising. Like many rural residents, veterans are older with more than half being over age 65 and having higher rates of chronic diseases compared to their urban counterparts; 13% are minorities.(1) The VA’s clinical data warehouse is considered one of the best in the US because of the massive amounts of standardized data including pharmacy information. It therefore constitutes a unique resource with which to examine outcomes in a rural population.(2) In the past, lipid guidelines for primary and secondary prevention of cardiovascular disease (CVD) recommended triglyceride-lowering therapy (i.e., fibrates) rather than statin therapy in patients with hTG > 500 mg/dL, as an initial step to prevent hTG associate acute pancreatitis (AP).(3,4) The American Diabetes Association still suggests fibrates if the TG level is >1000 mg/dL to prevent AP.(5) Current cholesterol guidelines (6) are less specific in this regard, with a low density lipoprotein (LDL-C)-centric focus emphasizing the use of moderate-to-high-dose statins for CVD risk prevention. Thus, this may represent a lipid-lowering therapy moderately contraindicated for patients on fibrates because of adverse drug-drug interactions. A clinician therefore may be faced with choosing between drug therapy that minimizes CVD risk and one that minimizes risk of AP due to hTG.

An observational study of patients hospitalized for AP in the VA from October 2004 through September 2009 2009, Fiscal Years (FY) 2005-2009, considered the question of prescribing fibrates before statins, that is, prioritizing treatment of hypertriglyceridemia to prevent pancreatitis above LDL-c lowering to prevent coronary heart disease.(7) The authors found that, in defiance of conventional wisdom, fibrates were still prescribed less often than statins. High rates of cardiovascular events in the year following hospitalization for acute pancreatitis supported this clinical focus on reducing cardiovascular risk through pharmacological management of hypercholesterolemia. 23% of the cohort of 20,608 veterans experienced coronary heart disease, myocardial infarction, or stroke in one-year follow-up, and 13% died. We revisited the VA data for an update on outcomes of patients with acute pancreatitis in the next 10 years, FY2010-FY2019. The source data are copied from the electronic medical record system nightly for use by operations managers and researchers. Exclusion criteria included >89 years of age and missing data on clinical or demographic measures. The year prior to admission for acute pancreatitis was considered the baseline for assessment of clinical status; follow-up was measured in the 365 days following admission. The cohort of 23,422 patients admitted for acute pancreatitis in FY2010-FY2019 averaged 60 years of age (SD: 12.1; 18-89); 94% were male; 69% were White, 27% were Black, and 2% were Native American, Asian, Hawaiian Native or other Pacific Islander; independent of race, 6% were of Hispanic ethnicity. Within one year of admission for acute pancreatitis, 25% had coronary heart disease, 6% myocardial infarction, 6% stroke, and 0.2% gallbladder disease; in addition, 8% died.

Overall, 43% of patients were not prescribed antilipemic agents. About half (53%) were prescribed statins, 9% fibrates, and 3% other types of lipid-lowering drugs (LLD), primarily bile acid sequestrants. Among the patients with prescription fills, the pattern of prescribed antilipemic agents by fiscal year is shown in Figure 1. The figure shows the dominance of statins over fibrates; prescription of other types of drugs varied from 3% to 7% per year. The percentages are >100 because patients were prescribed more than one type of LLD.

In modeling each outcome with logistic regression, the unadjusted regression of triglyceride level on the outcome was contrasted with the results adjusted for demographics, Charlson Comorbidity Index at baseline,(8) hypertension, and mental and behavioral health disorders (alcohol use disorder; depression) as well as receipt of statins, fibrates, and other LLD. For myocardial infarction (MI) within one year after AP admission, elevated triglycerides in the 200-1999 mg/dL range appeared to be a risk factor in the unadjusted model. However, after adjusting for the effects of gender, race/ethnicity, comorbidity, HDL-C level, and lipid-lowering drugs, triglycerides were not significant. Other factors explained the risk of MI. Interestingly, Black race was protective in this model while statins and other LLD were risk factors (fibrates n.s.).  In the model of one-year mortality, all three levels of triglyceride elevation (200 < 500, 500 < 2000, 2000+) were protective in the unadjusted model, while none were significant in the adjusted model. Rather, the Charlson, developed to predict one-year post-discharge mortality, conferred risk while mental and behavioral health issues had protective effects, perhaps signaling patients receiving heightened attention.

This review of VA patients with acute pancreatitis illustrated the utility of applying principles of EBM and more specifically, the importance of adjusting for risk attributable to factors other than the primary one, such as LLD. It illustrates the value of prioritizing CVD risk management over acute pancreatitis risk management. Conventional wisdom should now dictate statins before fibrates. 

Disclosure statement: Dr. Copeland has no financial disclosures to report. Dr. Rajab has no financial disclosures to report.
References are listed in the 2021-2022 Winter LipidSpin .pdf on www.lipid.org