Acute Pancreatitis in Patients With Severe Hypertriglyceridemia in a Multi-Ethnic Minority Population

Ambika Amblee, MD; Divyanshu Mohananey, MD; Micheal Morkos, MD; Sanjib Basu, PhD; Ayokunle T. Abegunde, MD; Malini Ganesh, MD; Neil Bhalerao, MD; Amrutha Mary George, MD; Milli Jain, MD; Leon Fogelfeld MD


Endocr Pract. 2018;24(5):429-436. 

In This Article


The prevalence of HTG-AP in our cohort was 9.2%, which is slightly lower than the weighted mean prevalence of 14.0% with a wide range of prevalence estimates (8 to 31%) reported by a recent systematic review of observational studies.[13] However, the size of our cohort was larger than the entire pooled patient population from that review. Plus, this is the first HTG-AP report in a multi-ethnic minority population.

This study clearly shows that the AP risk increases with higher levels of TGs, especially >2,000 mg/dL. This study validates the Endocrine Society's suggested cut-off of TGs >2,000 mg/dL as a risk factor to develop HTG-AP and shows a 4.3-fold increase in the prevalence of AP at this level compared to the group with TGs 1,000 to 1,999 mg/dL (22.0% vs, 5.1%). This trend was further noted when the TG values were divided into four incremental groups, with the risk of AP increasing 11-fold in the highest group.

Alcohol is a well-known independent risk factor for AP.[15,18,19] Our study shows that patients with a history of alcohol intake have 4 times the odds of developing AP. These results are concordant with previous literature on HTG-AP.[20] It is well known that alcohol leads to markedly higher TG levels through inhibition of lipolysis, especially in patients with pre-exiting lipoprotein lipase deficiency.[9] Several additional mechanisms have been suggested for the causative role of alcohol intake in AP, including sphincter of Oddi dysfunction, plugging of the pancreatic ductules resulting in acinar injury, stimulation of pancreatic enzyme secretion predisposing to endotoxin-induced damage, direct toxic effect mediated by metabolites such as fatty acid ethyl esters, and, finally, through localized hypoxia produced by increased oxygen requirements during ethanol metabolism.[21–25]

Although smoking was not significantly associated with development of HTG-AP, we observed that approximately 50% of patients with a history of alcohol intake were also smokers, and amongst those with a history of smoking, 82.5% had a history of alcohol intake. Therefore, it is evident that these two modifiable risk factors often co-exist. In previous studies, smoking has been shown to increase progression of chronic alcoholic pancreatitis and also potentiate pancreatic ischemia and damage in acute alcoholic pancreatitis.[26,27]

Gallstone disease is also a known independent risk factor for AP.[15,18,19] Our analysis revealed that gallstone disease was an independent risk factor for HTG-AP development. We also found that in patients with AP, prevalence of gallstone disease was significantly higher in women compared to men, which has been well described in the past.[15,18,19]

In addition to the degree of HTG, excessive alcohol intake and presence of gallstones, other factors were associated with AP. Patients who developed AP were younger, and younger age showed to be an independent risk factor associated with HTG-AP. It is not clear why older patients with similar risk factors were less susceptible to developing AP. A recent retrospective comparison of HTG-AP, biliary AP, and alcoholic AP highlighted this fact by reporting that patients with HTG-AP are significantly younger as compared to patients with biliary AP or alcoholic AP.[10] We also observed that men clearly outnumbered women in both the overall cohort and the AP subgroup. This finding is similar to a study by Sandhu et al,[12] who reported that 70% of the patients in their large cohort of HTG patients were men. Interestingly, despite the lower overall prevalence of women with HTG in our study, there was an equal prevalence of AP in male and female patients (9.1% vs 9.9%).

We found that a large number of patients in our cohort had a previous diagnosis of DM (70.9%), with an average HbA1c of 9.6% (81 mmol/mol), indicating poor glycemic control, especially in men (10.1% [87 mmol/mol] in men vs. 8.1% [65 mmol/mol] in women). Uncontrolled DM has been previously reported in studies on HTG-AP at varying prevalence, ranging from 43 to 72%.[4,28] Our data reveal that there is no difference in either prevalence of DM or in mean HbA1c values between patients with and without HTG-AP. These findings are concordant with two studies, one retrospective and one prospective.[12,20]

Our study is retrospective and, therefore, has limitations inherent to this type of study. However, we subjected our data to rigorous checks, including periodically choosing random patients and by sorting each continuous variable in ascending order to evaluate accuracy and look for outliers. In addition, the large size of our cohort makes the statistical analysis robust and the results credible. Since our study group is unique in racial composition, it cannot be fully generalized to the general population. Also, when stratifying the patients by TG levels, the number of patients in the highest two quartiles were small. Therefore, caution needs to be observed in interpreting data from these two subgroups. Our data also lacked details of medication usage and body mass index values. These could possibly contribute to unmeasured confounding variables in our regression model.

In this study, the prevalence of AP increased within each TG group with the concomitant presence of excessive alcohol intake and gallstones. In absence of these risk factors, the prevalence of AP was very low (2 to 3%) in the lower TG level groups. These risk predications are further corroborated by the ROC analyses and the predictive model developed. In addition, the predictive model shows the importance of young age as additional important risk factor (Figure 3). Using this model, younger patients had a much higher risk, as much as 4-fold, to develop pancreatitis than their older counterpart with similar risk factors. However, this model will need further validation. The findings of the study might help risk-stratify patients with severe HTG in the outpatient setting to assess their risk for developing AP. It may also have therapeutic implications and help the clinician to decide on the urgency and intensity of management of severe HTG based on the patient's risk.