Society: AGA

Background: The incidence of hypertriglyceridemia-associated acute pancreatitis (HTG-AP) has been increasing recently. Plasmapheresis can effectively remove triglyceride(TG) from plasma, but its clinical value in HTG-AP is unclear due to the lack of solid evidence. This study aimed to assess the impact of plasmapheresis on the evolution of organ failure in HTG-AP patients.
Methods: This study is a prospective, multicenter cohort study. HTG-AP patients with clinical worrisome features and abdominal pain onset within 72 hours were included. Patients who underwent plasmapheresis were compared with those receiving conventional medical treatment. The primary outcome was organ failure-free days (OFFD) to 14 days of enrollment. Secondary outcomes included other measures for the evolution of organ failure, need for intensive care unit (ICU) admission, length of ICU and hospital stay, 60-day mortality, and the incidence of infected pancreatic necrosis. Propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) analysis were used to control potential confounders.
Results: During the study period, 267 HTG-AP patients were enrolled from 28 sites, among whom 211 underwent conventional medical treatment and 56 underwent plasmapheresis. The PSM created 47 pairs of patients with balanced baseline characteristics. In the matched cohort, no difference was detected concerning OFFD between patients undergoing plasmapheresis or not (median (IQR), 12(8,14) vs. 13 (8,14); P = 0.94) (Table 1). Kaplan-Meier curves and Cox proportional hazards model also showed similar probability of organ failure resolution in both groups (HR:1.05, 95%CI:0.69-1.60, P=0.82) (Figure 1). Moreover, patients in the plasmapheresis group had higher rates of ICU admission (44(93.6%) vs. 24(51.1%), P<0.01). No difference was found in other secondary outcomes. IPTW results conform to the results from the PSM analysis.
Conclusion: In conclusion, early plasmapheresis may not be effective in improving organ failure and may be associated with more ICU admissions.

Table 1. Outcomes of plasmapheresis and conventional groups after propensity score matching

Table 1. Outcomes of plasmapheresis and conventional groups after propensity score matching

Figure 1. Kaplan–Meier plots for the OFFD in matched cohort

Figure 1. Kaplan–Meier plots for the OFFD in matched cohort

Background: While transient bacteremia is common during dental and endoscopic procedures [PMID: 18541739, 6354885], infections in sterile diseases like AP can be fatal. Fat necrosis occurs in human AP [PMID: 3950033, 3094241] and contains abundant NEFA from fat lipolysis [PMID: 25500204]. Here we study how NEFAs released in clinical AP may impair bacterial clearance, causing this sterile to infected transition.
Methods: All studies were approved by the IRB, IACUC. Mayo Clinic AZ emergency room patients with ≥ 2/3 AP criteria were included. Those with active malignancy, immuno-suppression, chemotherapy, age >90 years were excluded. Patients who had microorganisms isolated, developed sepsis, or required antibiotics for infection were grouped as infected. Antibiotic prophylaxis was not considered as infection. Admission serums were analyzed for NEFA, albumin-unbound NEFA, bacterial DNA amounts, microbiome, and immune cell injury (annexin V staining). Mice after IL12,18 pancreatitis [PMID: 18515422] +/- the triglyceride of linoleic acid [LA; PMID: 33514548] were analyzed similarly. Multiple groups were compared by ANOVA. A P value <0.05 was considered significant.
Results: There were 53 normal controls, 178 non-infected, 21 infected AP patients. Age, sex, BMI in all three groups were similar. Serum lipase, etiology of AP were similar in both AP groups. Infected AP had higher (p<0.001) serum unsaturated NEFA (737±526µM) vs. non-infected AP (475±293µM), controls (170±99µM). Unbound-NEFA were similarly higher in infected AP (5.6±3.0µM, 2.8±1.9µM, 0.9±0.4µM, P<0.001). Specifically unbound linoleic acid (LA; 3.4±3.1µM vs. 1.4±0.8µM in non-infected, 0.8±0.7µM control), and unbound-oleic acid (OA) were higher in infected AP. Bacterial 16S DNA copies (670±1201, 170±340, 2.9±6.3) were increased in infected AP with altered beta-diversity (Jaccard-Emperor, PERMANOVA <0.0001), and enrichment in Pseudomonadales on level 4 analysis. Annexin V positive myeloid (CD14) and CD3 cells increased only in infected AP to 15±20%, 11±11% respectively vs <3% in controls. In vitro 2µM Unbound-LA, and -OA depolarized mitochondria of macrophages unlike other -medium and -long chain NEFA. The patterns of cell injury, NEFA elevation, microbiome changes, enrichment of Pseudomonadales were replicated only in mice with AP and lipolysis of LA’s triglyceride.
Conclusions: Release of LA, OA during fat necrosis in AP can increase their circulating unbound levels, which injure immune cells. This injury may explain sterile to infected transition during AP and is consistent with bacteremia being common but transient during dental, endoscopic procedures in normal individuals with healthy immune cells. In a separate submission [Control ID:3866559] we mechanistically detail this amphipathic immune cell injury, and the consequent impairment of phagocytosis and bacterial clearance.

Bryce McFayden

Appearances