Hypertriglyceridemia (HTG) is a well-established cause of acute pancreatitis (AP), and the incidence of hypertriglyceridemia-induced AP (HTG-AP) is rising globally (1). HTG-AP is more prone to progress to severe acute pancreatitis (SAP) with multiorgan dysfunction (2–4). SAP carries a mortality rate exceeding 30%, contributing to significantly higher morbidity and mortality than moderately severe AP (MSAP). The initial 24-h period following hospital admission is regarded as a critical juncture for the management of SAP, wherein timely intervention can exert a substantial influence on prognosis (5). Nevertheless, the early (< 24 h) assessment of AP severity remains a clinical challenge, and no satisfactory methodology has yet been established. Given the increased risk of severe progression in HTG-AP, early and accurate prediction of SAP risk is essential for timely ICU referral and targeted intervention (6). The implementation of such strategies is expected to reduce mortality, and complications associated with HTG-AP, thereby improving patient outcomes.

Currently, four major scoring systems are widely utilized for the early identification of SAP: the Acute Physiologic and Chronic Health Evaluation (APACHE II), Ranson score, Bedside index of AP severity (BISAP), and modified CT severity index (MCTSI). However, each system exhibits notable limitations in clinical practice. BISAP and MCTSI are more widely used in clinical settings due to their simplicity. The BISAP score, while simple and applicable within 24 h, exhibits limited sensitivity (65%) for predicting SAP (7, 8). The organ specificity of MCTSI improves its accuracy, but it needs to be graded according to local complications and requires more than 72 h to evaluate (9). Additionally, individual laboratory markers like red blood cell distribution width (RDW), serum calcium (Ca²⁺), and C-reactive protein (CRP) (10) is commonplace, yet these markers often lack sufficient sensitivity or specificity (1, 11). Therefore, advancing scoring methods is still needed to enhance predictive accuracy (12).

Pleural effusion (PE) is a prevalent complication of AP and has been identified as an early indicator of disease severity (13, 14). It has been particularly predictive of SAP within the first 48 h of hospital admission (15). A meta-analysis revealed that patients with AP and PE have a higher risk of developing complications such as pancreatic pseudocysts, accelerated pancreatic necrosis, and increased mortality compared to those without PE (16). Despite this, most existing studies on PE and AP outcomes have included heterogeneous etiologies and lack specific analyses within the HTG-AP population— a group characterized by more severe systemic inflammation and faster disease progression. Unlike previous studies that often used a 48-h window for PE detection or mixed AP subtypes, we specifically focus on the HTG-AP subtype and evaluate whether pleural effusion detected within 24 h of admission can serve as an early imaging biomarker for predicting progression to SAP and organ failure (OF) in this study. By narrowing the etiological spectrum and standardizing the imaging assessment window, we aim to determine whether early pleural effusion can facilitate more accurate and timely risk stratification in HTG-AP.

In this single-center cohort of patients with HTG-AP, PE detected within 24 h of admission was independently associated with a severe clinical course and organ failure. As an early, readily obtainable imaging sign, PE demonstrated discriminatory performance comparable to the MCTSI and superior to CRP, while remaining inferior to BISAP for predicting both HTG-SAP (ΔAUC = 0.0566) and organ failure (ΔAUC = 0.0831; both p < 0.001). These findings support the pragmatic role of PE as a simple “red-flag” marker that can inform early triage and monitoring when composite scores or complete laboratory results are not yet available. Prior multinational cohorts have linked PE to worse outcomes in acute pancreatitis overall, but most did not stratify by etiology and included only a small proportion of HTG-AP, limiting applicability to this subgroup (23–26). Etiology-specific analyses are important because HTG-AP often presents with higher inflammatory burden and earlier progression to severe disease. Our results extend existing evidence by focusing specifically on HTG-AP within a standardized 24-h imaging window and by directly comparing PE against commonly used tools (BISAP, MCTSI, CRP). We also deliberately excluded patients with chronic cardiopulmonary disease or concomitant pneumonia to reduce confounding from non-pancreatitis causes of pleural fluid, which may partially explain differences in effect sizes reported across cohorts.

Hyperlipidemia has now emerged as the second most common cause of AP globally, surpassing alcohol, and the prevalence in China of HTG-AP cases has been on a continuous rise over the past two decades attributed to socioeconomic and dietary factors (27). Studies have shown that compared with non-HTG-AP patients, HTG-AP patients have a higher proportion of males and are younger in age. HTG-AP patients also exhibit higher mortality, greater disease severity, and an increased recurrence rate (28). In our study, the robust association between early PE and HTG-SAP (OR: 6.82) may be attributed to the profound systemic inflammation characteristic of HTG-AP. The pathophysiological interplay between hypertriglyceridemia and acute pancreatitis involves the hydrolysis of excessive triglycerides by pancreatic lipase, leading to a surge in free fatty acids (FFAs). These FFAs induce direct acinar cell injury, intensify systemic inflammatory responses, markedly increases systemic vascular permeability, which facilitates fluid extravasation into body cavities, with the pleural space being a common site and contribute to microcirculatory disturbances, thereby predisposing patients to severe acute pancreatitis (SAP) with multiorgan dysfunction compared to other subtypes (29). Consequently, PE detected within the first 24 h serves as an early imaging indicator of a severe systemic inflammatory, heralding subsequent organ dysfunction.

Contrasting with Yan et al.’s (15) study, we deliberately excluded PEV as a predictive variable due to four methodological considerations. Firstly, volumetric measurement methods remain non-standardized in current clinical algorithms, with neither chest radiography (CXR) nor CT-based PEV models (e.g., ellipsoid/planimetry methods) have achieved consensus adoption under ACR/SIR guidelines (30). Secondly, specific quantitative criteria—such as fluid volume and characteristics—are not universally recognized or validated for predicting the severity of hyperlipidemic pancreatitis, making it challenging for clinicians to establish consistent and reliable assessment standards. This variability hampers the ability to develop unified clinical guidelines for evaluating the condition. Additionally, PEV assessment techniques are not widely implemented in primary care settings, and medical personnel frequently lack specialized training in its measurement. Furthermore, the absence of readily available software for accurate measurement further complicates the early recognition of critical conditions, potentially delaying timely referral to tertiary hospitals for further treatment.

The role of triglyceride levels in determining the severity of acute pancreatitis has been a subject of ongoing debate. Compared with the findings of Deng et al. (31), our findings further demonstrated that TG level was an independent high-risk factor for progression to severe HTG-AP (OR 1.07, P = 0.009). In contrast to the view held by some studies, which posit that TG acts more as a triggering factor rather than a determinant of severity, our results support a different perspective. This discrepancy may be attributed to variations in sample size and regional differences among study populations.

Supporting the association between TG levels and disease severity, a large retrospective study of 1,457 consecutive acute pancreatitis cases found that elevated TG was independently correlated with a more severe clinical course. The incidence of organ failure, persistent multiple organ failure, pancreatic necrosis, and acute collections all increased significantly with worsening severity of HTG (32). In another retrospective study involving 1,539 AP patients, Wan et al. (33) found that 461 individuals (30%) had elevated triglyceride levels, and the incidence of severe AP rose with increasing HTG severity. Among patients with severe and very severe HTG (N = 112), 29% developed acute necrotic collections and 35% developed pancreatic necrosis. These patients also exhibited higher rates of persistent organ failure, multiple organ failure, persistent systemic inflammatory response syndrome, and a greater need for ICU admission (33).

In addition, our study only investigated the relationship between unilateral or bilateral PE and the risk of HTG-SAP, and no significant correlation was found between the location of PE (unilateral or bilateral) and the development of HTG-SAP, SIRS, or OF (p > 0.05). There could be two main reasons for these results. First, since this study included only 296 patients, the small sample size may have led to different outcomes. Second, most pleural effusions associated with pancreatitis tend to occur on the left side, possibly due to anatomical proximity.

Spatial anatomical relationships further influence the laterality of PE. Pancreatic inflammation and diaphragmatic vulnerabilities predispose fluid migration toward the left thoracic cavity. Bao et al. (34) postulated that left-sided predominance may stem from the larger surface area of the left lumbar costal triangle and its proximity to the pancreas. Our findings align with this observation: bilateral PE occurred in 48.3% of cases (42/87), while left-sided PE was present at 37.9% (33/87) and right-sided PE at only 13.8% (12/87), underscoring a clear leftward bias. Some of the results in our study are consistent with those of some previous results (15, 35). For example, in the study by Yan et al. (15), the incidence of pleural effusion in AP patients was 49.9% (232/465), with 72.4% (168/232) being bilateral, 23.7% (55/232) on the left side, and 3.9% (9/232) on the right side (p < 0.001).

PE’s predictive ability is comparable to that of MCTSI (ΔAUC = 0.0035 and 0.0179 for predicting HTP-SAP and OF; p > 0.05). Since MCTSI assessment includes both pancreatic fluid collection and pancreatic necrosis, pancreatic fluid collection might not score highly on imaging at 24 h, while pancreatic necrosis typically only becomes apparent later in the disease course. This results in MCTSI having low accuracy during early evaluation, usually requiring 72 h for accurate assessment. While originally designed for mortality prediction, the BISAP score has been extensively validated in subsequent research as a reliable tool for early identification of Severe Acute Pancreatitis (SAP) and organ failure. Its predictive objective aligns directly with the primary clinical endpoint of our study—the development of SAP.

All components of the BISAP score can be obtained objectively within the first 24 h of admission, perfectly matching our study’s timeframe for evaluating early predictors. Its inclusion of dynamic criteria like SIRS already captures relevant pathophysiological processes. Although PE shows slightly lower predictive accuracy compared to the BISAP score (ΔAUC = 0.0566 and 0.0831 for predicting HTP-SAP and OF; p < 0.001), its use as a standalone indicator provides distinct advantages in clinical practice. The BISAP score includes five criteria, making it more complex to apply than pleural effusion. Additionally, some criteria in the BISAP score, such as altered mental status, rely on subjective judgment by clinicians, which may introduce assessment bias. The practical utility of PE lies in its simplicity and immediacy. In busy emergency departments or resource-limited settings where calculate a full BISAP score might be delayed due to pending laboratory results or the need for comprehensive clinical assessment, the presence of PE on a CT scan—often performed urgently to confirm the diagnosis of AP—provides an instantaneous, objective risk signal. This can prompt clinicians to prioritize these patients for intensive monitoring, aggressive fluid resuscitation, or early specialist consultation, even before a formal composite score is available.

This investigation acknowledges 6 primary limitations. First, as a single-center study, the findings are subject to regional, racial, and lifestyle influences, which may limit the generalizability of the results regarding HTG-AP. The single-center study and admission criteria may affect the representativeness of the sample. To enhance the robustness of the findings, future research should include a multicenter approach with diverse geographic and demographic populations. Second, low event rates (mortality < 8%) in this Grade III Level A hospital compromised statistical power for PE-mortality correlation analysis. Furthermore, the extended study period (2013–2024) may have introduced heterogeneity, such as imaging protocols, intensive care unit admission criteria, and treatment guidelines could have evolved over time. However, it is important to note that the core diagnostic criteria for HTG-AP, the use of early CT imaging for severity assessment, and the fundamental principles of supportive care remained largely consistent in our tertiary center throughout this period. The non-significant P-value for the comparison between unilateral and bilateral PE should be interpreted with caution due to potential limited statistical power in this subgroup analysis. Although standardized resuscitation protocols were followed, the timing and intensity of specific interventions (like fluid resuscitation or insulin/heparin therapy for HTG) during the first 24 h could have modified the clinical course and potentially affected the development of SAP. Larger prospective studies are needed to clarify the above two specific association. Lastly, in more remote or underdeveloped regions, access to diagnostic tools for detecting PE may be limited, reducing the practicality of using PE as an early predictor of disease severity in such areas.

In conclusion, early prediction and management of HTG-SAP remain critical challenges in clinical practice. This study demonstrates that PE detected within 24 h acts as a simple, readily available early warning sign, identifying HTG-AP patients at high risk for progressing to severe disease and organ failure. While inferior to BISAP for discrimination, it provides pragmatic utility for early triage when rapid scoring is not feasible.