Hypertriglyceridemic acute pancreatitis in emergency department: Typical clinical features and genetic variants

To investigate the clinical characteristics of patients with hypertriglyceridemic acute pancreatitis (HTGAP), and the molecular foundation contributing to hypertriglyceridemia in such patients.


INTRODUCTION
Acute pancreatitis (AP) is a common gastrointestinal emergency with an increasing incidence in China during the past decades. Because of its severity ranging from mild to severe or even fatal, or with systemic inflammatory conditions, AP continues to arouse the interest of clinical researchers. 1,2 Although AP in most patients is mild, moderate to severe as well as recurrent AP remains a tremendous burden on both public healthcare system and the patient's family. 3,4 Of note, 10-20% of patients with AP developed severe disease with a mortality of up to 30%. [5][6][7] The most common etiologies of AP are cholelithiasis with obstruction of common bile duct or main pancreatic duct and alcohol abuse. Currently, a rapid increase in obesity nationwide and worldwide is associated with an elevated incidence of metabolic diseases and changes in disease spectra. Hyperlipidemia, characterized by serum hypertriglyceridemia, has become the third leading cause of AP and contributed to 1-10% of patients with AP. Moreover, mild to moderate hyperlipidemia can be regarded as an underlying phenomenon or comorbidity of pancreatitis. [8][9][10] Previous cohort studies have indicated that serum triglyceride (TG) of ≥1000 mg/dL (11.3 mmol/L) is a high risk factor for patients with hypertriglyceridemic acute pancreatitis (HTGAP). It has also been considered necessary to diagnose HTGAP when the patient's fasting TG level is 5.65--11.3 mmol/L with chylous blood which occurs in about 20% of all patients with AP. 11,12 Clinical studies have demonstrated that HTGAP may contribute to increased severity and mortality, higher frequencies of comorbidities and systemic complications, longer length of hospitalization, and more frequent recurrence, than other subtypes of AP. 10,13 Our emergency department has accepted many patients with AP and provided them with an early diagnosis and effective interventions. We thus conducted a retrospective study on AP patients who were admitted to our emergency department between 2012 and 2016, focusing on HTGAP. Considering that notable dyslipidemias have a strong genetic component despite the important role that secondary dietary factors play in the clinical phenotype, we further performed gene mutation detection for 11 patients with HTGAP to determine the molecular genetic characteristics of this special subtype of AP. [14][15][16] In this study, we aimed to predict the severity and recurrence risk in HTGAP with the new-generation sequencing techonology based upon clinical information of the patients.

Patients
Patients who were admitted to the Emergency Medical Ward and Emergency Intensive Care Unit (EICU) of Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University (Shanghai, China) due to AP between January 2012 and December 2016 were retrospectively included in the study. The data of the patients was anonymous and their medical insurance numbers were encrypted. Those with insufficient medical records or who were transferred to another hospital before their completion of the treatment were excluded. Patients' data and clinical details were collected from databases such as hospital inpatient enquiry, digital medical record and the hospital information management system, and the characteristics of the patients were reviewed. The clinical data of patients having AP with well-documented medical records were collected and analyzed, including their characteristics, clinical presentations, disease etiologies and severity, underlying diseases, comorbidities, length of hospital stay, laboratory test results, frequency of recurrence and outcomes, and so on. The protocol for the study was approved by the Institutional Ethics Committee of Ren Ji Hospital, and the study was conducted in accordance with the Declaration of Helsinki (2013). Three main etiologies of AP (biliary origin, alcohol abuse and hypertriglyceridemia) were identified. When cholelithiasis was identified by diagnostic imaging modalities including ultrasonography, plain or contrast-enhanced computed tomography (CT) and/ or magnetic resonance imaging (MRI) scan, biliary AP was diagnosed. Alcohol-related AP was defined as AP related to an average alcohol consumption of 80 g daily for men and 50 g daily for women for at least 5 years or an excessive amount (≥200 g for men and ≥150 g for women) of alcohol consumed immediately before the acute disease attack. Other types of AP were classified as idiopathic AP with multiple etiologies.

Diagnostic criteria, severity classification and the etiologies of AP
Peripheral blood sample was collected from the patients after overnight fasting. Serum amylase levels, liver and renal function tests, blood sugar and hemoglobin A1c (HbA1c) concentrations were analyzed using a Biochemical Analyzer (Roche, Basel, Switerland). Serum TG level was determined at day 2 or 3 of their admission. As reported previously, serum TG with a threshold of 500 mg/dL (5.65 mmol/L) has been reported to be associated with a high incidence of AP, 11,[17][18][19] HTGAP was diagnosed in this study when: (i) TG ≥11.3 mmol/L; or (ii) TG ≥5.65 mmol/L with visible chylomicronemia. Fatty liver disease was identified by ultrasonography showing a "bright" liver with increased echogenicity and/or CT/MRI scan showing a low density in the liver than in the spleen. Diabetes was diagnosed according to a clear medical history of diabetes and/or HbA1c ≥6.5%. Hypertension was diagnosed in patients with a blood pressure of over 140/90 mmHg after measured for three times under resting state or a clear history of hypertension. The patients who had at least two or three AP attacks were defined as having recurrent or repeated recurrent AP.

Patients and samples
Among the AP patients with a fasting serum TG level ≥5.65 mmol/L, 11 (including 7 men and 4 women) agreed to provide their peripheral blood samples for gene detection. Written informed consent was obtained before the collection of the blood from each patient.
Genomic DNA from blood samples was extracted using a High Pure PCR Template Preparation Kit (Roche) according to the manufacturer's instructions. Solution-based hybridization capture was used to enrich DNA fragments for sequencing on a MiSeq Sequencing Platform with a base paired-end read module (2 × 300; Illumina, San Diego, CA, USA). The procedure of hybridization capture was fulfilled by a commercial SureSelect Library Prep Kit (Agilent, Santa Clara, CA, USA) to capture a region of 97.49 kb, containing 325 regions of 32 genes known to cause hyperlipidemia. We selected eight target genes that are known to be associated with hypertriglyceridemia or familial combined hypertriglyceridemia for analysis, including APOA5, APOC3, APOE, BLK, LPL, APOC2, GPIHBP1 and LMF1. Oligonucleotide baits were designed to cover total coding exons, untranslated regions and at least 10 intronic nucleotides and all intron-exon boundaries nearby, excluding deep intronic sequences to avoid enlarging the sequence target size and subsequently reducing the sequence coverage. DNA was first quantified by applying a Qubit 2.0 (Thermo Fisher Scientific, Waltham, MA, USA), followed by 0.3-mg genomic DNA sonication sheared with a Diangenode Bioruptor Plus (Diangenode, Liege, Belgium), hybridization was then performed between the biotinylated RNA oligonucleotide baits and sheared DNA. After removing the captured fragments from the solution via a streptavidin-coated magnetic microbeads and elution step (Dynabeads MyOne Streptavidin T1; Thermo Fisher Scientific), polymerase chain reaction (PCR) amplification was performed using the enriched fragment library and primers specific to the linked Illumina adaptors. Then the PCR-produced libraries were further processed by the MiSeq Sequencing Platform (Illumina) after quantification by a quantitative PCR. All the samples could be sequenced together depending on the 6-bp index sequences (Illumina) used to differentiate different samples.

Variant analysis
After MiSeq sequencing, the raw data of each sample were sorted automatically by index sequences and the adapter sequences were trimmed by Cutadapt 1.13 (https://cutadapt.readthedocs.io/en/stable/). 20 Solex-aQA 21 V3.1.2 was used to remove low-quality bases (<Q20). The clean results were aligned to the database of the human reference genome (hg19) by applying the Burrows-Wheeler Aligner 0.7.11. 22 After alignment, the PCR duplicates were removed using the Mark Duplicates package of Picard 1.109 (https://broad institute.github.io/picard/). Realignment around indel sites known and base quality score recalibration were performed by GATK 3.3. 23 GATK HaplotypeCaller was used to call raw variants. Indels and single nuclear polymorphisms were annotated with ANNOVAR. 24 Public databases including dbSNP138, 1000 Genome project, Exome Sequencing Project, Clinic Var and the Human Gene Mutation Database 25 were used to screen variants. The prediction of functional effect was evaluated by PolyPhen and SIFT scores. 26,27 To detect the copy number variant, the sequencing depth of each region covered by the probes was calculated according to the alignment files. An Exomedepth package 28 was also used to find potential copy number variants. During copy number variant analysis, confirmed point mutation samples in the same sequence run served as controls.

Statistical analysis
Statistical analyses were performed using SPSS 17.0 (IBM, Armonk, NY, USA). Continuous variables were expressed as mean AE standard deviation (SD), whereas categorical variables were expressed as numbers and percentages. Comparisons between the groups were performed using Student's t-test (for independent samples), Pearson's χ 2 test and Fisher's exact χ 2 test when appropriate. Difference was considered as statistically significant at P < 0.05.

Characteristics of the patients
A total of 329 patients with AP were admitted to the Emergency Department (either the ward or ICU) of Ren Ji Hospital with detailed medical records. Fasting TG level was over 1.7 mmol/L, with or without hypercholesterolemia, in 118 patients, all of whom were diagnosed as having hyperlipidemia. Among them, those with a fasting serum TG of ≥11.3 mmol/L or of ≥5.65 mmol/L together with visible chylomicronemia were enrolled in the HTGAP group (n = 40, 12.2%). Other patients were enrolled in the non-HTGAP (NHTGAP) group (n = 289, 87.8%), and were further divided into three subgroups based on their etiologies: biliary AP (n = 76, 23.1%), alcohol-induced AP (n = 32, 9.7%) and idiopathic AP (n = 181, 55.0%) (Fig. 1a) Fig. 1b).

Detection of gene mutation in HTGAP and clinical outcomes
DNA sequencing showed that two patients (nos. 3 and 5) carried the same heterozygous mutation (p.G185C) and four (nos. 1, 7, 8 and 10) had another heterozygous mutation (p.V153M) of the APOA5 gene. Two patients (nos. 2 and 6) with compound heterozygous variants had p.G185C and p.V153M in the APOA5 gene presented with marked hypertriglyceridemia, and patient no. 2 with the peak concentration died of SAP. Two patients (nos. 4 and 5) carried a same homozygous variation of p.C14F in the GPIHBP1 gene and two (nos. 1 and 7) carried the same heterozygous mutation of p.C14F in the GPIHBP1 gene. One patient (no. 9), who carried a homozygous variation of p.R176C (rs7412) in the APOE gene, presented with a high TG level of 19.12 mmol/L. Two patients (nos. 1 and 10) carried the same heterozygous mutation of p.R176C in the APOE gene also had high TG levels. Two patients (nos. 3 and 5) carried the same rare heterozygous mutation, p.P562R, in the LMF1 gene. One patient (no. 11) had another rare heterozygous mutation of p.N249T in the LMF1 gene (

DISCUSSION
In this retrospective exploratory study on AP, we conducted a cohort analysis in 329 patients who were admitted to our emergency department and focused on those with HTGAP. Additionally, we obtained genetic information on hyperlipidemia from 11 patients with HTGAP. In the present study, patients with HTGAP accounted for 12.2% (40/329) of all patients with AP. We found that younger age, diabetes mellitus, fatty liver, a high recurrence rate and a more severe prognosis were more commonly observed in HTGAP, a finding similar to that reported in recent studies. 10,12,29,30 Elevated serum TG levels are known to be associated with an increased incidence of acute or recurrent pancreatitis. 6,11 Patients whose serum TG level remained ≥5.65 mmol/L had a higher probability of AP attacks. Thus we chose a fasting serum TG level of ≥5.65 mmol/L with visible chylomicronemia or TG ≥11.3 mmol/L to obtain an accurate diagnosis of hypertriglyceridemia. Population-based studies from Christian et al. 17 and Murphy et al. 11 defined TG levels of ≥5.65 mmol/L (500 mg/dL) and 1.7 mmol/L as cut-off values, respectively. The incidence of AP in individuals with serum TG levels of above 5.65 mmol/L was significantly higher than in those with a TG level of 150-500 mg/dL. Although there are controversies about the severity of HTGAP compared with other subsets of AP, 3,11,30 our study showed that HTGAP was associated with a high incidence of MSAP (65.0%) and SAP (17.5%), more common comorbidities such as diabetes and fatty liver as well as a higher mortality (7.5%). High recurrence rates or repeat recurrence rates were also found in the HTGAP group. It should be noted that mild to moderate hypertriglyceridemia (TG 175-500 mg/dL) could be considered as a comorbidity of pancreatitis, and choosing a lower cut-off value for hypertriglyceridemia might bear major flaws. This might explain the results of a British study in 43 patients with different types of AP using a cut-off value of TG >175 mg/dL (2 mmol/L) that higher severity was not found in the HTGAP group compared with other subtypes of AP. 8,30 Among the 11 patients who underwent genetic detection, we found that eight carried heterozygous APOA5 gene mutations. Genome-wide association studies have shown that genetic variations existing in the APOA5 loci, mainly p.G185C, p.V153M and p.S19W, are definitely associated with abnormal blood TG levels in humans, each with compound monogenic or polygenetic effects. 31 Although the p.G185C variant located in the APOA5 gene was first reported in 2003, it has attracted great attention from researchers during the past years because its homozygous mutation contributes strongly to severe hypertriglyceridemia and resultant HTGAP. [32][33][34] In particular, the p.G185C polymorphism in the APOA5 gene is more commonly distributed in Asians, including Chinese populations, AAP, alcohol-induced acute pancreatitis; AIP, autoimmune pancreatitis; BAP, biliary acute pancreatitis; IAP, idiopathic acute pancreatitis; HTGAP, hypertriglyceridemic acute pancreatitis; MAP, mild acute pancreatitis; MSAP, moderately severe acute pancreatitis; SAP, severe acute pancreatitis. *P < 0.05 and **P < 0.01 compared with HTGAP. than in others. 15,35,36 It has been speculated that the functional mechanism of APOA5 can decrease the concentration of blood TG by increasing lipoprotein lipase activity. 37 We inferred that a functional loss of single nucleotide polymorphisms in the APOA5 gene would result in reduced TG lipolysis and remnant accumulation, hence causing hypertriglyceridemia. 38,39 However, all the eight patients identified in this study were heterozygous for p.G185C and p.V153M, or compound heterozygous for p.G185C and p.V153M, which may have contributed partially to the abnormal serum TG levels. [40][41][42] It is interesting that two (nos. 4 and 5) of the 11 patients carried a same homozygous p.C14F variation located in the GPIHBP1 gene and two had the same heterozygous mutation of p.C14F in the GPIHBP1 gene. GPIHBP1 has been found to be expressed on capillary endothelial cells to bind lipoprotein lipase and shuttle it to its site of action in the  capillary lumen, which is crucial in the process of releasing fatty acid from blood TG for uptake in tissues. Homozygous GPIHBP1 genetic mutations may interfere with protein folding and disable the capacity of GPIHBP1 to combine and transport lipoprotein lipase, resulting in severe hypertriglyceridemia. [43][44][45] This can help to explain the pathogenesis of hypertriglyceridemia in some patients with AP in this study.
Among the 11 patients, one (no. 9) had a homogeneous APOE mutation of p.R176C (rs7412), and two had the same heterozygous mutations, which were variants that are more often found in isolated hypertriglyceridemia in APOE-associated dyslipidemia. Recent studies 46,47 have also suggested that dyslipidemia due to a polymorphism allele of the APOE gene or APOE deficiency can be identified by the accumulation of chylomicrons in the blood. This was similar to the finding in our study that one patient carried a homogeneous APOE mutation with a high TG level of 19.12 mmol/L. This indicates that dysfunctional APOE contributes to the occurrence of HTGAP, while the relationship between APOE mutations and AP needs further investigation. One unique finding in the HTGAP group was that two patients carried the same heterozygous mutation (p.P562R in the LMF1 gene), and one carried a heterozygous mutation (p.N249T in the LMF1 gene). These have been considered rare gene variants up to now and the relationship between LMF1 and HTGAP is still unclear and needs further research.
AP-associated dyslipidemias are heterogeneous disorders with a strong genetic component, characterized by the moderate to severe elevation of serum TG in combination with or without hypercholesterolemia. It results from the functional loss of mutations in a single gene or more genetic loci that damage the process of lipolysis and intravascular chylomicron clearance, with an inherited autosomal recessive mode caused by mutations in LPL, APOC2, LMF1, APOA5, GPIHBP1, APOE. 41,42 However, limited genetic information is available among individuals with HTGAP admitted to the emergency department. The present study identified several common and rare variants across the human genomes that are associated with HTGAP, indicating that genetic variants may have an important influence on the incidence of HTGAP. Once HTGAP attack has been resolved in these patients, we should focus on the strategies for early intervention and the prevention of recurrence, including early diagnosis, dietary restrictions and close monitoring of serum lipid levels. Therefore, we recommend routine genetic test combined with treatment of HTGAP, providing genetic information to patients with hyperlipemia type I or combined hypertriglyceridemia, and those with inherited mutations in APOA5, GPIHBP1, APOE and other candidate genes related to HTGAP. Pivotal factors for better outcomes include educating patients and encouraging them to adhere strictly to a low-fat diet, and using antihyperlipidemic agents to control their TG levels in order to avoid further episodes of AP. 18,48 Two large US cohort series have demonstrated that decreasing serum TG from above 500 mg/dL (5.65 mmol/L) to below 200 mg/dL (2.3 mmol/L) reduced the incidence of AP by 0.7% annually. 18 Therefore, monitoring lipids may be an option in HTGAP prevention and early intervention. The results of our study indicated that the effects of different genetic variants might have contributed to monogenic or polygenic hyperlipidemia and HTGAP. It is important to investigate the pathogenesis of hypertriglyceridemia in patients with pancreatitis because relevant management targeting such patients may be warranted.
The present study had some limitations. First, it was based on a single emergency medical center and some relevant details might not have been documented. Second, only 11 patients consented to gene detection. Further characterizations of the genetic profile influencing lipid levels with a larger sample size combined with the biological information of their family members are required to validate these findings.
In conclusion, patients with HTGAP presented with notably more severe disease process with a higher rate of recurrence compared with those without, although the differences were not statistically significant. Genetic mutations including APOA5, GPIHBP1 and APOE variants might have contributed to the occurrence of hypertriglyceridemia in patients with AP. Genetic information may be useful to investigate the pathogenesis of HTGAP and to predict the prognosis of patients with HTGAP, enabling them to offer a relevant management plan for prevention and early intervention of this disease. New-generation sequencing technology may change the current clinical diagnostic process.

ACKNOWLEDGMENTS
We thank all the patients for their kind agreement to be involved in this study. We thank Dr. Richard ZHAO from the Division of Hospital Medicine, Swedish Medical Center (Seattle, WA, USA) for his kind help on our work.