The Association of Body Composition Parameters and Laboratory Markers With The Severity of Hypertriglyceridemia-Induced Pancreatitis

Background: Hypertriglyceridemia has arisen as the third leading cause of acute pancreatitis. This study aimed at exploring the association between the severity of hypertriglyceridemia-induced pancreatitis (HTGP) and computed tomography (CT)-based body composition parameters and laboratory markers. Methods: Clinical parameters were collected from 242 patients with HTGP between 2017 and 2020. Severity of HTGP was evaluated by original or modied CT severity index. Body composition parameters such as area and radiodensity of muscle, subcutaneous adipose tissue and visceral adipose tissue were calculated by CT at the third lumbar vertebra level. Parameters between mild and moderately severe to severe HTGP were compared. Uni-variate and multi-variate Logistic regression analyses were employed to assess the risk factors of the severity of HTGP. Results: Seventy patients with HTGP (28.9%) were mild. Body mass index, waist circumference and all CT-based body composition parameters differed between male and female patients. None was associated with the severity of HTGP, neither in the male nor in the female. Uni-variate and multi-variate Logistic regression analysis showed that low serum albumin (<35g/L) and high C-reactive protein (>90 mg/L) were risk factors of moderately severe to severe HTGP (P<0.001, OR=4.846, 95%CI=2.122-11.068; P<0.001, OR=4.230, 95%CI=2.050-8.727, respectively). Low serum albumin was also associated with pancreatic necrosis, longer hospital stay and higher scores of APACHE II, Ranson and Marshall in HTGP patients (all P<0.05). Conclusions: Low serum albumin and high C-reactive protein upon admission are associated with the HTGP. However, the body adipose

fatty acids exert a detergent-like role and attacks platelets, vascular endothelium, and acinar cells.
Although some studies have recognized the effect of albumin on the severity and the organ failure in AP, it has not been included as a risk factor of AP in most studies and clinical practice. Furthermore, there are few studies investigating the association of serum albumin with HTGP, despite that serum albumin might be more closely linked to the deposition of free fatty acids and the severity of HTGP.
Since the prognosis and clinical intervention of AP differs largely in mild AP and severe AP, the early prediction of the severity in AP has been a great concern and research focus for clinicians. A host of predictors, including clinical, laboratory and radiological markers and various scoring systems, have been proposed. Current guidelines recommend the persistent systemic in ammatory response syndrome or organ failure for at least 48 hours to predict the severity of AP [7]. Other recognized risk factors for development of severe AP include elevated hematocrit, blood urea nitrogen and creatinine and body mass index [8,9]. Several scoring systems, such as Acute Physiology and Chronic Health Evaluation II (APACHE II), the Bedside Index for Severity in Acute Pancreatitis (BISAP), and Ranson's Criteria have proven their correlation with the severity of AP, but their clinical utilities are still limited due to the low predictive values [10,11]. Computed tomography (CT) on admission has the similar predictive accuracy for the severity of AP as clinical scoring systems [12]. Although not recommended by the guidelines [7], abdominal CT upon admission is often conducted in AP in real-world practice, especially in emergent situation. Contrastenhanced CT is advantageous to delineate pancreatic or peripancreatic uid collections and necrosis. It is the most applied method in clinic to re-evaluate the severity of cases in which deteriorated manifestation and laboratory ndings are indicated.
Recent studies have suggested that the severity of AP is also linked to some radiological parameters for the body composition [13]. Yoko Yashima and colleagues found that larger peripancreatic volume of visceral adipose tissue (VAT) is associated with severe AP [14]. Another study from Korean reported that high visceral fat with low skeletal muscle volume was strongly correlated with AP severity [15]. A most recent systematic review, which included 11 studies, concluded that VAT is an important prognostic indicator of the severity of AP and may be incorporated into the prognostic scoring systems of AP [16].
Nevertheless, Hanna Sternby and colleagues found that muscle radiodensity, rather than VAT and subcutaneous adipose tissue (SAT), correlates with severe acute pancreatitis [17]. Notably, by stratifying AP into HTGP and non-HTGP, Ting Ji and colleagues revealed that VAT and VAT/SAT were valuable factors for predicting the severity in HTGP but not non-HTGP [18].
With a focus on HTGP, a subgroup of AP with rising incidence, this study aimed at exploring body composition parameters and clinical risk factors for the severity of HTGP, which may help predict the prognosis of HTGP.

Patients
A total of 242 hospitalized patients with HTGP were retrospectively studied from the First A liated Hospital of Wenzhou Medical University between July 2017 and August 2020. The diagnosis of AP was established according to the Revised Atlanta De nitions of AP [19], when 2 of the following 3 characteristics were met: (1) the symptoms of abdominal pain are consistent with AP, (2) the levels of amylase and/or lipase are at least 3 times above the upper limit of normal, and (3) abdominal imaging is consistent with changes in AP. HTGP is considered in AP patients when the level of serum TG is (1) over 11.3 mmol/L; (2) between 5.65 mmol/L and 11.3 mmol/L, milky serum, with no other etiology of AP; (3) not tested, but the patient, without any known etiology, has previously been diagnosed as HTGP. The inclusion criteria of patients in this study were (1) the diagnosis of HTGP was established; (2) abdominal CT scanning 72 h before the occurrence of symptom; exclusion criteria were (1) poor CT imaging of the abdomen; (2) indication of biliary, alcoholic, autoimmune, drug-induced or pancreatic tumor-related etiology of AP; (3) pregnancy.
Contrast-enhanced CT scannings of abdomen were performed in 177 (73.1%) of the studied patients when deterioration of the disease was indicated during the course. The severity of HTGP was evaluated by modi ed CT severity index (MCTSI) if contrast-enhanced CT scanning has been conducted, otherwise by the CT severity index based on unenhanced CT scanning.
Clinical parameters of each patient were retrieved from the Electronic Health Record System, including sex, gender, body mass index, length of hospital stay, and C-reactive protein level, serum albumin, TG concentration upon the admission. The levels of CRP, albumin and TG were categorized as high and low with the cutoff of 90 mg/L, 35 g/L and 22.4 mmol/L [20], respectively. APACHE II, BISAP, Ranson and Marshall scorings of the patients were calculated when possible.

Ct Scanning And Body Composition Parameters
CT scanning was performed by 64-slice spiral CT scanner (Lightspeed VCT, GE healthcare, USA) or Aquilion ONE 320 Slice CT scanner (Toshiba, Japan). The scanning covered the whole abdomen region. The slice thickness was 0.625 mm for 64-slice spiral CT scanner (pitch 0.984, single-turn spiral time 0.5 s, 100 kV, 500 mA) and 0.5 mm for the 320 Slice CT scanner (single-turn spiral time 0.5 s, 100 kV, 300 mA).
For contrast-enhanced CT, an auto-injector was used to inject 60 mL of non-ionic contrast agent (iopromide 300mgI/mL) and 30 mL of saline at a speed of 4.0 mL/s, and the scanning was triggered intelligently by monitoring of the abdominal aorta, the arterial phase was delayed for 30-35 s, and the portal phase was delayed for 60-70 s. The imaging data was transferred to the post-processing workstation (Version 4.5, GE healthcare).
Body composition parameters were measured based on the non-enhanced CT scanning. Muscle and adipose tissue at the third lumbar vertebra level (L3) with supine position were analyzed by Image J software [21]. A range of -29 to 150 Houns eld units (HU) was set to highlight muscle, a range of -190 to -30 HU was set to highlight SAT, and − 150 to 50 HU for VAT. The area and density of the region of interest (ROI) were calculated automatically. Waist circumference was measured at the navel plane. CT measurements were performed by 2 experienced radiologists blinded to the clinical information. Remeasurement took place when disagreement of the measurements occurred.

Statistical analysis
Parameters between mild and moderately severe to severe HTGP were compared. Comparisons of categorical data were performed using Chi-square test or Fisher's exact test. Continuous data were expressed as mean ± standard deviation and compared by the student's t test when normal distribution was justi ed, otherwise expressed as median (interquartile range) and compared by Mann Whitney U test.
Uni-variate and multi-variate Logistic regression analyses were employed to assess the risk factors for the severity of HTGP. To test the predictive capacity of the measured parameters, receiver operating characteristic (ROC) curves with corresponding areas under the curves (AUC) were calculated. P-value of < 0.05 was considered statistically signi cant. All the statistics were performed using SPSS 18.0 (IBM. SPSS Statistics for Windows, USA).

Body Composition Parameters were divergent between male and female patients with pancreatitis
Notably, differences existed apparently between the male and female patients regarding all the body composition parameters studied in our study, including BMI, muscle area, muscle radiodensity, SAT area, SAT radiodensity, VAT area, VAT area/total adipose tissue area and waist circumference (all P < 0.05). The mean values of each parameter in male and female patients with HTGP were shown in Table 1. For the subsequent analyses, all body composition parameters were categorized as high and low subgroups using corresponding median value in the male or the female as the cutoff. There were no differences of age, CRP value, albumin concentration and TG concentration between the male and female patients (all P > 0.05).

Parameters Associated With The Severity Of Htgp
To investigate the association of the severity of HTGP with the body composition parameters, laboratory parameters and clinical data, comparisons were made between patients with mild HTGP and those with moderately severe to severe HTGP. Age and gender distribution showed no difference between the two groups (all P > 0.05). However, the serum CRP and TG level were signi cantly higher in patients with moderately severe to severe HTGP than those with mild HTGP (all P < 0.05). Besides, serum albumin concentration was signi cantly lower in the severe group (mean 34.98 vs. 40.38 g/L, P < 0.001). Since body composition parameters were signi cantly gender dependent, strati ed comparisons in either male or female patients were performed. Consequently, the muscle radiodensity (mean 42.9 vs. 44.8 HU, P = 0.020) was lower while SAT radiodensity (mean − 95.0 vs. -97.7 HU, P = 0.009) was higher in male patients with moderately severe or sever HTGP than those with mild HTGP. Uni-variate Logistic regression analysis was applied to explore parameters associated with the severity of HTGP in either male or female subgroups of patients. In male patients with HTGP, age, CRP and albumin were associated with the CT-based severity of the disease (all P < 0.05, Table 2). In female patients, high CRP and TG level, but none of the body composition parameters, were associated with the severity of HTGP (Table 2). Multi-variate Logistic analysis con rmed that CRP and albumin remained to be associated with the severity of HTGP in male patients, while CRP was found to be the only parameter associated with the severity of HTGP in female patients (all P < 0.01, Table 3). By enrolling parameters of gender, age, muscle radiodensity, CRP, albumin and TG, the multi-variate Logistic analyses on patients of both genders showed that high CRP [P < 0.001, odds ratio (OR) = 4.230, 95% con dence interval (CI) = 2.050-8.727] and low albumin (P < 0.001, OR = 4.846, 95%CI = 2.122-11.068) were associated the severity of HTGP (Table 3).

Discussion
This study showed that low albumin and high CRP levels upon admission of patients with HTGP were associated with the severity of the disease, while none of the CT-based body composition parameters was linked to the severity of HTGP. We further found that low albumin concentration was associated with pancreatic necrosis, longer hospital stay and higher APACHE II, Ranson and Marshall scorings of patients with HTGP. To our knowledge, this is the rst study suggesting that low albumin may serve as a risk factor for moderately severe to severe HTGP. Our nding indicates that HTGP patients with low albumin should be kept under close surveillance of the disease progression, and may bene t from treatment aiming to normalize albumin concentration, such as albumin supplement.
Albumin has been suggested as a predictive factor for the severity of AP [22]. More recently, a study included 708 patients with AP and additional 477 patients from validation cohort, nding that albumin was an independent predictor for SAP and in-hospital mortality in AP patients [23]. Albumin has also been incorporated into some prediction panels which showed usefulness to predict the severity of AP [24,25]. For example, serum albumin plus extrapancreatic uid collections were suggested as the best indicator of severity of AP at the time of admission [26]. Another study suggested that combining blood urea nitrogen and albumin resulted in better prediction for SAP in pediatric patients [27]. Besides, low albumin is also helpful to predict organ failure of AP [28]. It has been shown that low serum albumin is independently associated with an increased risk of persistent organ failure and death in AP [29].
However, the association of albumin concentration with the severity of the subgroup of AP, HTGP, has not been investigated before. In HTGP, the deposition of free fatty acids hydrolyzed from TG plays a crucial role in the mechanism of pancreatitis. Theoretically, decrease of albumin, which can bind to the free fatty acids in the serum, may result in accumulation of free fatty acid, thus contributing to the in ammatory progression. Decrease of albumin may also be a consequence of severe type of HTGP with lager amount of free fatty acid to be neutralized by albumin, and with higher vasopermeability which allows more albumin to permeate into tissue space [30]. Therefore, the albumin concentration may show a closer correlation with the severity of HTGP than AP caused by other etiology. In line with these assumptions, we found that HTGP patients with albumin lower than 35 g/L had four-times more possibility to be moderately severe or severe than those with albumin over 35 g/L. Notably, our strati cation study in the female patients, the albumin showed no correlation with the severity of HTGP. This might be explained by the relative smaller number of female patients and underpower of the statistical analysis. Studies with larger sample size may be of value to ascertain the association between the albumin and the severity of HTGP.
CRP level upon admission was another laboratory parameter of predictive value for the severity of HTGP in our study. This nding was consistent with the study from Yu and colleagues, in which 159 HTGP Chinese patients were retrospectively enrolled and high CRP and BMI were postulated as risk factors for severe HTGP [31]. Indeed, CRP is one of the most widely utilized biomarkers in clinical practice for AP. CRP levels of > 150 mg/L 48 h after onset of symptoms have a high sensitivity for predicting severity of AP [32]. Stirling and colleagues reported that a rise of > 90 mg/dL from admission or an absolute value of > 190 mg/dL at 48 h predicted severe disease with the greatest accuracy [33]. However, it is generally thought that the initial CRP at time of admission is too early to be predictive for the severity of AP [34].
Our nding of the predictive role of CRP upon admission for the severity AP needs validation in more studies in the future.
The body composition parameters, especially VAT, have been reported to correlate with the severity of AP. Due to the involvement of lipoprotein metabolism in HTGP, we asked whether body composition parameters were associated with the severity of HTGP. Intriguingly, no association was found between the severity of HTGP and any of these body composition parameters, including CT-measured muscle area, muscle radiodensity, SAT area/radiodensity, VAT area, and body mass index and waist circumference. In comparison, another study from Ji et al. showed that VAT area was associated with the severity of HTGP. Several factors may contribute to the discrepancy. Firstly, our study noticed that gender has a signi cant impact on all these body composition parameters, which has been reported before [35]. Therefore, the cut-off of each body composition parameter and the subsequent statistical analysis were gender dependent in our study. Secondly, our study included HTGP with any severity, while Ji and colleagues' study included only moderately severe to severe patients from intensive care unit. Lastly, there is a difference of sample size between the two studies.
Pancreatic necrosis is a parameter for CT-based evaluation of the severity of AP, and is closely associated with the morbidity and mortality of AP [36]. The management of pancreatic necrosis has been highlighted due to its signi cance for the outcome of AP [37]. Grinsven et al. has investigated the association between body composition and mortality of AP with pancreatic necrosis, concluding that loss of skeletal muscle density within the rst month after initial admission was signi cantly associated with increased mortality [38]. It has been shown that patients with HTGP had higher incidence of pancreatic necrosis compared to AP with other etiology [39]. In HTGP, excess TG is hydrolyzed by lipase from pancreatic acinar cells to produce FFAs, which in turn cause acinar necrosis [3]. Our study suggested that high SAT radiodensity and low serum albumin were associated with pancreatic necrosis in patients with HTGP.

Conclusions
We found that low serum albumin and high C-reactive protein upon admission were signi cantly associated with the severity of HTGP. Our study indicated that serum albumin may play a crucial role in the progression of HTGP, while supplement of albumin may be of therapeutic value. Besides, we noticed that gender has a signi cant impact on all body composition parameters, none of which was associated with the severity of acute HTGP. We also found that high SAT radiodensity and low serum albumin may act as an indicator for pancreatic necrosis in patients with HTGP. More studies are needed to con rm the predictive role of serum albumin and C-reactive protein in the severity of HTGP.

Declarations
Ethics approval and consent to participate The study protocol was approved by the ethics committee of the First A liated Hospital of Wenzhou Medical University. Participant informed consent was waived given the retrospective study design.

Consent for publication
Not applicable.

Availability of data and material
All relevant data and materials are included in the manuscript. For the full detailed data, please contact the corresponding author.

Competing interests
The authors declare no con icts of interest.

Funding
This work was supported by a grant from the Public Welfare Science and Technology Plan Project of Wenzhou City (Y20170829).

Authors' contributions
Chen L performed the experiments, analyzed the data and wrote the initial draft of the manuscript; Huang YB analyzed the data and assisted in manuscript writing and study conception; Yu HJ collected the clinical data and follow-up of the patients; Pan KH, Zhang Z and Man Y critically revised the manuscript for important intellectual content; Hu DY conceived the study; all authors read and approved the nal manuscript.