Abstract
Purpose
To compare the efficacy of continuous renal replacement therapy (CRRT) using either oXiris or conventional hemopurification filters in the treatment of intra-abdominal sepsis.
Methods
We conducted a retrospective analysis of septic patients with severe intra-abdominal infections admitted to our hospital from October 2019 to August 2023. Patients who meet the criteria for intra-abdominal sepsis based on medical history, symptoms, physical examination, and laboratory/imaging findings were included. Exclusion criteria: pregnancy, terminal malignancy, prior CRRT before intensive care unit admission, pre-existing liver or renal failure. Heart rate (HR), mean arterial pressure, oxygenation index, lactic acid level (Lac), platelet count (PLT), neutrophil percentage, serum levels of procalcitonin, C-reactive protein, interleukin (IL)-6, norepinephrine dosage, acute physiology and chronic health evaluation II (APACHE II), and sequential organ failure assessment (SOFA) scores before and after 24 h and 72 h of treatment, as well as ventilator use time, hemopurification treatment time, intensive care unit and hospital lengths of stay, and 14-day and 28-day mortality were compared between patients receiving CRRT using either oXiris or conventional hemofiltration. Statistical analysis was performed using SPSS Statistics 26.0 software, including the construction of predictive models via logistic regression equations and repeated measures ANOVA.
Results
Baseline values including time to antibiotic administration, time to source control, and time to initiation of CRRT were similar between the 2 groups (all p>0.05). Patients receiving conventional CRRT exhibited significant changes in HR but of none of the other indexes at the 24 h and 72 h time points (p=0.041, p=0.026, respectively). The oXiris group showed significant improvements in HR, Lac, IL-6, and APACHE II score 24 h after treatment (p<0.05); after 72 h, all indexes were improved except PLT (all p<0.05). Intergroup comparison disclosed significant differences in HR, Lac, norepinephrine dose, APACHE II, SOFA, neutrophil percentage, and IL-6 after 24 h of treatment (p<0.05). Mean arterial pressure, serum levels of procalcitonin, C-reactive protein, SOFA score, and norepinephrine dosage were similar between the 2 groups at 24 h (p>0.05). Except for HR, oxygenation index, and PLT, post-treatment change rates of △ (%) were significantly greater in the oXiris group (p < 0.05). Duration of ventilator use, CRRT time, and intensive care unit and hospital lengths of stay were similar between the 2 groups (p>0.05). The 14-day mortality rates of the 2 groups were similar (p=0.091). After excluding patients whose CRRT was interrupted, 28-day mortality was significantly lower in the oXiris than in the conventional group (25.0% vs. 54.2%; p=0.050). The 28-day mortality rate increased by 9.6% for each additional hour required for source control and by 21.3% for each 1-point increase in APACHE II score.
Conclusions
In severe abdominal infections, the oXiris filter may have advantages over conventional CRRT, which may provide an alternative to clinical treatment. Meanwhile, early active infection source control may reduce the case mortality rate of patients with severe abdominal infections.
Keywords: Intra-abdominal infection, oXiris CRRT, Conventional CRRT, Sepsis, Mortality
1. Introduction
Intra-abdominal infection (IAI) is prevalent among surgical intensive care unit (ICU) patients and may carry a mortality rate of nearly 20%.1 Etiologies include gastrointestinal perforation, intestinal obstruction, traumatic hemorrhage, and surgical complications. Among patients with severe abdominal trauma, IAI is the leading cause of death after 48 h.2
Perforation or rupture of the abdominal cavity or viscera allows the influx of digestive enzymes, bile, and other substances into the peritoneal cavity that may promote local abscess formation, diffuse peritonitis, and microbial translocation. The large surface area and vascularity of the peritoneum facilitate the rapid onset of bacteremia and endotoxemia that can cause sepsis, metastatic infection, and multi-organ failure that may increase mortality by 30%–50%.3 In addition to promptly treating the primary disease and initiating antimicrobial therapy, the timely mitigation of the maladaptive immune dysregulation caused by endotoxin and inflammatory mediators, restoration of immune homeostasis, and removal of toxins are among the most important means of reducing morbidity and mortality.4,5 Hemopurification is no longer limited to the replacement of renal function but is also used for organ support through comprehensive pathways and modes.
Conventional continuous renal replacement therapy (CRRT) can maintain water-electrolyte balance and remove inflammatory mediators; however, its survival benefit is uncertain.6 Japanese investigators introduced the use of polymyxin fiber columns (PMX adsorption columns) for hemoperfusion during the 1990s. Although numerous in vivo and clinical studies have confirmed that PMX adsorption columns can remove endotoxin and inflammatory mediators to improve hemodynamics and tissue perfusion, attenuate cytokine storms, and reduce the mortality of sepsis, controversy remains. Confidence in the evidence level of 3 randomized controlled studies showing clinical benefit was low;7 furthermore, the risks of neuro- and nephrotoxicity and expense have limited its use.
oXiris, a new type of filter marketed in Europe, has become a new focal point in the treatment of sepsis. The oXiris filter is a new iteration based on the AN69 hydrogel structure. The base membrane material, AN69 (a polymer of acrylic and sodium methanesulfonate), adsorbs a wide range of pro- and anti-inflammatory factors. An in vitro trial demonstrated that the oXiris filter conferred a more comprehensive adsorptive clearance of sepsis mediators compared to Toraymyxin or CytoSorb, with the adsorptive clearance profiles of the former 2 and an effective clearance of both endotoxin and most inflammatory factors.8,9 Consequently, we hypothesized that oXiris would outperform conventional CRRT filters in severe IAI. We retrospectively analyzed cases of IAI with sepsis, and compared the therapeutic efficacy of CRRT using conventional and oXiris filters, aiming to provide a new treatment modality.
2. Methods
2.1. Study design
We conducted a single-center retrospective cohort study of patients with severe IAI admitted to the Department of Critical Care Medicine of the First Affiliated Hospital of Fujian Medical University from October 2019 to August 2023.10 A total of 50 patients with septic shock who underwent extracorporeal hemopurification were analyzed. Patients were categorized into 2 groups according to the CCRT filter used for their therapy. Twenty-seven patients received conventional pipeline treatment (conventional group) and the other 23 received oXiris pipeline treatment (oXiris group).
2.2. Inclusion and exclusion criteria
Inclusion criteria were (1) perforation, necrosis, gangrene, or surgical complications of the digestive tract confirmed by medical history, physical examination, and abdominal CT; (2) positive bacterial smear and culture of abdominal drainage fluid or samples taken during diagnostic laparotomy, or bloodstream infection of abdominal origin; (3) meeting the case definition of sepsis (Sepsis 3.0 issued by the European Society of Critical Care Medicine and the American Society of Critical Care Medicine), requiring medication dose escalation to maintain mean dynamic pressure ≥65 mmHg (1 mmHg=0.133 kPa) after aggressive fluid resuscitation, and an arterial blood lactate >2 mmol/L.
Exclusion criteria were (1) other intestinal diseases without severe IAI; (2) antecedent hepatic, renal, or other organ system diseases; (3) end-stage malignant tumors; (4) pregnancy; (5) CRRT before entering the ICU admission.
2.3. Interventions
Conventional CRRT group: in addition to intensive care, patients received hemopurification within 24 h of ICU admission. Continuous veno-venous hemofiltration was administered by central venous cannulation (femoral vein indwelling 11.5 F catheters) and the use of a Prisma-flex hemopurification device and matching AN69ST hemofiltration membrane. The therapeutic regimen was 30–40 mL/kg/h), and blood flow was 120–180 mL/h. Either heparin anticoagulation or heparin-free treatment was selected according to the patient's coagulation function, the presence of hemorrhage, and liver function. Doses of antibiotics and other drugs were adjusted according to glomerular filtration rate.
oXiris CRRT group: the AN69-oXiris (Baxter in the USA) filter was used. All other aspects of treatment were identical to those of the conventional CRRT group.
2.4. Observation indicators
Heart rate (HR), mean arterial pressure (MAP), oxygenation index (FiO2/PaO2), blood lactate level (Lac), platelet count (PLT), neutrophil ratio (N%), serum procalcitonin (PCT), serum C-reactive protein (CRP), and interleukin (IL)-6 were measured in the 2 groups before and at 24 h and 72 h after the initiation of CRRT. Other clinical parameters compared between the 2 groups were norepinephrine dosage, acute physiology and chronic health status score II (APACHE II) and sequential organ failure assessment (SOFA) scores, duration of ventilator use, time to antibiotic administration after the first diagnosis of sepsis, time to undergoing new drainage, incision, or debridement after the first diagnosis of sepsis, time to CRRT after the first diagnosis of sepsis, duration of hemodialysis, ICU and hospital lengths of stay, and 14-day and 28-day mortality.
2.5. Statistical methods
SPSS 26.0 software was used for statistical analysis. Continuous variables that obey the normal distribution were expressed by mean ± standard deviation (SD) and compared by using the t-test; continuous variables that were not normally distributed were expressed by the median (quartile spacing) and compared using the rank and test; the 2 groups of patient count indicators were compared using the χ2 test. A p<0.05 was considered statistically significant. The logical regression equation was used to construct the prediction model, and the Hosmer-Lemeshow test was used to judge the consistency of the model. Inspection level was α=0.05.
3. Results
A total of 50 patients met the inclusion criteria (Fig. 1). Systemic heparin anticoagulation was administered to 14 (51.9%) patients in the conventional CRRT group and 12 (52.2%) in the oXiris group.
Fig. 1.
Screening flow chart. IAI: intra-abdominal infection; ICU: intensive care unit.
Pretreatment demographic and clinical characteristics were similar between the 2 groups (p>0.05) (Table 1). To reduce the bias caused by covariates, statistical analysis was performed on the mean difference of the observation content (difference method). Gram-negative bacteria were the most prevalent etiologic agents in both the conventional and the oXiris groups, accounting for 81.5% and 78.3% of cases, respectively (p>0.05). Escherichia coli caused 44.0% of our cases. Only a few fungal infections occurred, accounting for 14.8% and 4.4%, respectively (p>0.05). Gastrointestinal perforation accounted for most primary cases (no intergroup difference, p=0.324). Bloodstream infections of intra-abdominal origin accounted for 18.5% and 17.4% of patients in the conventional and oXiris groups, respectively (p>0.05).
Table 1.
Pre-treatment demographic and clinical characteristics.
| Variables | Conventional group | oXiris group | p value |
|---|---|---|---|
| Male, n (%) | 14 (51.9) | 17 (73.9) | 0.109 |
| Age (year), mean (Q1, Q3) | 61 (48, 73) | 55 (39, 67) | 0.202 |
| BMI (kg/m2), mean ± SD | 24.76 ± 4.24 | 25.51 ± 4.56 | 0.549 |
| AKI stage, mean (Q1, Q3) | 2.00 (1.00, 2.00) | 2.00 (2.00, 3.00) | 0.072 |
| Heart rate (beats/min), mean ± SD | 115.92 ± 3.55 | 123.00 ± 5.23 | 0.258 |
| MAP (mmHg), mean ± SD | 76 (63.33, 84.33) | 70.33 (61.33, 84.33) | 0.559 |
| Lactate (mmol/L), mean (Q1, Q3) | 4.85 ± 3.545 | 8.99 ± 11.86 | 0.09 |
| Oxygenation index (mmHg), mean (Q1, Q3) | 197.14 (144, 273) | 188 (139, 256) | 0.586 |
| Platelets (10^9/L), mean (Q1, Q3) | 102.50 (39.75, 147.25) | 96.00 (59.00, 136.00) | 0.948 |
| Neutrophil ratio (%), mean (Q1, Q3) | 88.40 (81.43, 92.08) | 91.50 (87.70, 93.50) | 0.060 |
| PCT (mg/mL), mean (Q1, Q3) | 24.54 (12.84, 63.45) | 23.08 (2.78, 78.32) | 0.674 |
| CRP (mg/L), mean (Q1, Q3) | 104.50 (61.91, 159.25) | 133.00 (101.00, 196.30) | 0.318 |
| IL-6 (pg/mL), mean (Q1, Q3) | 1324.50 (245.93, 4386.60) | 3751 (762.00, 6978.00) | 0.055 |
| Norepinephrine dosage (μg/kg/min), mean ± SD | 0.55 ± 0.65 | 0.89 ± 0.77 | 0.097 |
| APACHE II, mean ± SD | 23.91 ± 4.82 | 22.81 ± 5.04 | 0.114 |
| SOFA, mean ± SD | 13.44 ± 2.38 | 12.61 ± 3.89 | 0.429 |
| Route of infection, n (%) | |||
| Blood infection | 5 (18.5) | 4 (17.4) | 1.000 |
| Gastrointestinal perforation | 6 (22.2) | 8 (34.8) | 0.324 |
| Intestinal obstruction | 5 (18.5) | 4 (17.4) | 1.000 |
| Biliary system | 5 (18.5) | 1 (4.3) | 0.271 |
| Acute pancreatitis | 4 (14.8) | 6 (26.1) | 0.523 |
| Anastomotic leak | 3 (11.1) | 2 (8.7) | 1.000 |
| Abdominal abscess | 1 (3.7) | 1 (4.3) | 1.000 |
| Source of infection, n (%) | |||
| Gram-negative bacteria | 22 (81.5) | 18 (78.3) | 1.000 |
| Gram-positive bacteria | 11 (40.7) | 7 (30.4) | 0.449 |
| Fungal infection | 4 (14.8) | 1 (4.3) | 0.449 |
| Timing of antibiotics (h), mean (Q1, Q3) | 1 (1, 2) | 1 (1, 2) | 0.616 |
| Timing for source control (h), mean (Q1, Q3) | 16 (9, 29) | 16 (9, 20) | 0.619 |
| Timing for RT (h), mean (Q1, Q3) | 34 (24, 40) | 24 (10, 46) | 0.074 |
SD: standard deviation; BMI: body mass index; AKI: acute kidney injury; MAP: mean arterial pressure; PCT: serum procalcitonin CRP: C-reactive protein; IL: Interleukin; APACHE II: acute physiology and chronic health evaluation II; SOFA: sequential organ failure assessment.; RT: continuous renal replacement therapy.
The conventional CRRT group displayed significantly changed HR between baseline and 24 h and 72 h timepoints, but no statistically significant differences in MAP (p=0.343 and p=0.963, respectively), Lac (p=0.691 and p=0.506, respectively), PCT (p=0.514 and p=0.154, respectively), CRP (p=0.672 and p=0.263, respectively), norepinephrine dose (p=0.746 and p=0.971, respectively), APACHE II (p=0.201 and p=0.239, respectively), and SOFA scores (p=0.682 and p=0.782, respectively) (Fig. 2). The oXiris group exhibited improvement of all parameters except PLT (p=0.080) at the 72 h time point (p < 0.05). Intergroup comparison of the change rate of indexes at 24 h disclosed significant differences in HR, MAP, Lac, and norepinephrine dose (all p<0.01). Evaluation of inflammatory markers revealed intergroup differences in N% and IL-6 (p<0.05). The effects of treatment on APACHE II and SOFA scores (p<0.01) and other various indexes (p<0.05) were significantly greater in the oXiris group (Table 2, Fig. 2).
Fig. 2.
Results of various indicators using repeated measures analysis of variance within and between groups.
a, b, c indicated time point differences using the alphabetical labeling method (within-group comparisons), ∗ denotes significant differences between groups (between-group comparisons).
HR: heart rate; MAP: mean arterial pressure; OI: oxygenation index; Lac: lactate; PLT: platelet; N%: neutrophil ratio; PCT: procalcitonin; CRP: C-reactive protein; IL-6: interleukin 6; APACHE II: acute physiology and chronic health evaluation; SOFA: sequential organ failure assessment.
Table 2.
Comparison of the 24 h∖72 h change rate (△%) of the 2 groups.
| Variables | Comparison of the 24 h change rate (△%) of the 2 groups of treatment |
Comparison of the 72 h change rate (△%) of the 2 groups of treatment |
||||
|---|---|---|---|---|---|---|
| Conventional group | oXiris group | p value | Conventional group | oXiris group | p value | |
| HR | -38.5 (-13.0, -10.47)‡a | -17.74 (-25.61, -3.53)‡a | 0.001 | -6.99 (-20.88, 6.50)‡ | -22.76 (-33.33, 3.53)‡ | 0.012 |
| MAP (mmHg) | 3.82 ± 2.23† | 11.76 ± 20.89† | 0.121 | 4.75 ± 23.72†a | 17.26 ± 21.56†a | 0.030 |
| OI (mmHg) | 10.68 (-8.33, 23.25)‡ | 24.47 (2.62, 23.25)‡ | 0.329 | 0.73 (-13.19, 38.97)‡ | 27.80 (-4.19, 71.43)‡ | 0.116 |
| Lactate (mmol/L) | -7.14 (-23.40, 39.13)‡a | -34.48 (-57.29, -7.26)‡a | 0.006 | -18.18 (-51.45, 46.67)‡a | -54.29 (-74.16, -20.11)‡a | 0.011 |
| PLT (10^9/L) | -18.30 ± 28.64† | -8.06 ± 34.20† | 0.436 | -33.33 (-64.14, 10.71)‡ | -35.38 (-51.55, -11.94)‡ | 0.335 |
| N% | 3.40 (-37.70, 2.12)‡ | -1.20 (-4.77, 5.30)‡ | 0.055 | -3.20 (-3.87, 6.48)‡a | -3.81 (-8.67, 4.60)‡a | 0.009 |
| PCT (mg/mL) | -18.78 (-34.00, 0.00)‡a | -26.00 (-153.90, -1.87)‡a | 0.046 | -38.12 (-56.62, -12.72)‡a | -66.85 (-84.82, -37.22)‡a | 0.038 |
| CRP (mg/L) | -10.00 (-25.05, 13.76)‡ | -18.11 (-47.89, 0.00)‡ | 0.069 | -27.70 (-43.01, 2.83)‡a | -49.68 (-57.66, -24.06)‡a | 0.017 |
| IL-6 (pg/mL) | -30.1 (-68.92, -8.37)‡a | -60.45 (-80.74, -34.78)‡a | 0.001 | -53.16 (-79.02, -13.65)‡a | -83.47 (-90.89, -40.08)‡a | 0.017 |
| Norepinephrine dosage (ug/kg/min) | 0.00 (-29.03, 62.50)‡a | -37.46 (-76.40, -25.00)‡a | 0.007 | 0.00 (-77.00, 67.48)‡a | -75.23 (-93.10, -34.21)‡a | 0.001 |
| APACHE II | -7.69 (-14.81, 9.52)‡a | -18.18 (-23.53, -9.09)‡a | 0.000 | -7.41 (-19.05, 8.70)‡a | -28.57 (-43.33, -15.00)‡a | 0.000 |
| SOFA | 0.00 (-12.50, 6.67)‡a | -12.50 (-20.00, -7.69)‡a | 0.003 | -3.87 ± 22.82†a | -20.52 ± 13.51†a | 0.001 |
Data presented as mean (Q1, Q3) or mean ± SD, excepted p value.
HR: heart rate; MAP: mean arterial pressure; OI: oxygenation index; PLT: platelet; N%: neutrophil ratio; PCT: procalcitonin; CRP: C-reactive protein; IL-6: interleukin 6; APACHE II: acute physiology and chronic health evaluation II; SOFA: sequential organ failure assessment.
p < 0.05; †: t/χ2 test; ‡: Ζ test.
Ventilator use time, CRRT time, ICU and hospital lengths of stay, and 14-day mortality were similar between the 2 groups (all p>0.05) (Table 3). After excluding patients for whom CRRT was interrupted, the 28-day mortality rate was lower in the oXiris group than in the conventional group (25.0% and 54.2%, respectively; p=0.050).
Table 3.
Observational indicators in the conventional and oXiris groups.
| Item | Conventional group | oXiris group | p value |
|---|---|---|---|
| Respiratory support time (day), mean (Q1, Q3) | 7.21 (4.04, 11.00) | 8.00 (5.00, 11.00) | 0.690 |
| RT time (day), mean (Q1, Q3) | 7.00 (5.33, 13.00) | 7.00 (4.00, 14.00) | 0.733 |
| ICU hospitalization time (day), mean (Q1, Q3) | 10.00 (6.00, 16.00) | 10.00 (8.00, 17.00) | 0.396 |
| Total hospitalization time (day), mean (Q1, Q3) | 19.00 (8.00, 30.00) | 21.00 (11.00, 32.00) | 0.365 |
| 14-day mortality rate, n (%) | 12 (44.4) | 5 (20.8) | 0.091 |
| 28-day mortality rate, n (%) | 13 (54.2) | 5 (25.0) | 0.050 |
RT: continuous renal replacement therapy; ICU: intensive care unit.
Binary logistic regression was performed to jointly analyze the relationship between time to antibiotic administration (h), time to source control (h), time to initiation of CRRT (h), APACHE II scores, and 28-day mortality in the 2 groups (Table 4, Fig. 3).
Table 4.
Analyzing the relationship between the timing of antibiotics, timing for source control, timing for RT, APACHE II and 28-day mortality rate by logistic regression.
| Parameter | β | SE | Waldc2 | p | OR | 95% CI |
|---|---|---|---|---|---|---|
| Constant | -8.190 | 2.801 | 8.550 | 0.003 | 0.000 | – |
| Timing of antibiotics | 0.647 | 0.488 | 1.759 | 0.185 | 1.910 | 0.734–4.967 |
| Timing for source control | 0.092 | 0.045 | 4.116 | 0.042 | 1.096 | 1.003–1.198 |
| Timing for RT | 0.012 | 0.026 | 0.214 | 0.644 | 1.012 | 0.962–1.065 |
| APACHE II | 0.193 | 0.089 | 4.645 | 0.031 | 1.213 | 1.018–1.445 |
SE: standard error; OR: odds ratio; CI: confidence interval; RT: continuous renal replacement therapy; APACHE II: acute physiology and chronic health evaluation II.
Fig. 3.
Analyzing the relationship between the timing of antibiotics, timing for source control, timing for RT, APACHE II and 28-day mortality rate by logistic regression. RT: continuous renal replacement therapy; APACHE II: Acute Physiology and Chronic Health Evaluation II; ROC: receiver operating characteristic curve.
The Hosmer-Lemeshow test indicated a good fit (χ2=5.966, p=0.651). The receiver operating characteristic curve was plotted according to the calculation of the prediction model, and disclosed an area under the receiver operating characteristic curve of 0.827 (95% CI: 0.704–0.949). The 28-day mortality rate increased by 9.6% for every 1 h delay of source control and by 21.3% for each 1-point increase of APACHE II score.
4. Discussion
IAI includes primary and secondary peritonitis. Primary peritonitis arises without an antecedent intra-abdominal focus of infection, but occurs after bacterial translocation across either the intestinal wall, lymphatic vessels, or the female reproductive system; or by hematogenous metastasis. Primary peritonitis usually occurs in immunocompromised patients. Etiologic agents include Escherichia coli, Streptococcus pneumoniae, and hemolytic streptococci.
Secondary peritonitis follows perforation of hollow abdominal viscera or abdominal surgery. Enterobacteriaceae are still the leading pathogens, of which Escherichia coli is the most common.11 In this study, secondary IAI was the main cause of sepsis, and 80.0% of patients were infected with Gram-negative bacteria. Escherichia coli accounted for 44.0% of our cases. A study of 5510 IAI clinical isolates from the Asia-Pacific region identified Escherichia coli, Klebsiella pneumoniae, Klebsiella oxytoca, Citrobacter freundii, and Proteus mirabilis as the 5 most prevalent extended-spectrum β-lactamase producers.12 Although cultures in several cases in this study yielded only Gram-positive bacteria or fungi, the possibility of false-negative cultures for Gram-negative bacteria cannot be excluded considering the complexity of IAI. Therefore, these patients were not excluded from our analysis.
On the other hand, the role of enterogenic Gram-positive bacterial infections cannot be underestimated. Honore et al.13 proposed that endotoxins translocate following disruptions of the gastrointestinal barrier due to diverse infectious and non-infectious etiologies such as Gram-positive bacterial infections, pancreatitis, hypovolemic shock, and mesenteric ischemia, and thereby cause endotoxemia. Endotoxins may catalyze the development of sepsis. Dysregulated inflammation, high metabolic states, and multi-organ dysfunction that complicate sepsis may be triggered directly or indirectly by endotoxins.4,14 Unfortunately, our center has not yet conducted quantitative and activity assays of endotoxin levels, and consequently cannot document the effects of treatment on endotoxins. We observed changes in sepsis indirectly by comparing the levels of inflammatory indicators such as IL-6, CRP, PCT, and N%.
PCT levels are widely used in sepsis monitoring. Its synthesis is regulated by cytokines and endotoxins.15,16 The use of PCT to guide antibiotic therapy of sepsis may reduce 28-day mortality.17,18 Because PCT levels are closely related to all-cause mortality,19 they may be used to directly or indirectly monitor the severity of sepsis.
IL-6 is a key participant in the cytokine cascade. Elevated levels are closely related to mortality.20 Therefore, we used PCT and IL-6 levels to observe the severity of inflammation. Patients with severe IAI are in a state of pro-inflammatory cytokine storm. The removal of endotoxins and excess inflammatory factors and restoration of immune homeostasis may confer clinical benefits to improve clinical outcomes.21
oXiris consists of a hollow fiber of polyacrylonitrile methanesulfonate (AN69) coated with positively charged polyethylene imine and wrapped in heparin. In vitro experiments demonstrate that oXiris is the only CRRT filter that has both endotoxin adsorption and inflammatory mediator clearance abilities, and can also provide good renal support.9,22
Active control of the underlying infection is essential in the treatment of IAI. As shown in Table 1, both groups received antimicrobial agents within a relatively brief period. The time to initiation of antibiotic therapy did not differ between the 2 groups (p<0.05). A previous retrospective study showed that neglecting infection source control increased 28-day mortality from 26.7% to 42.9%.23 The prevailing consensus is that early intervention benefits clinical outcomes.24 We found that the time between the first diagnosis of sepsis and a new drainage, incision, or debridement was similar between the 2 groups (p<0.05).
CRRT was originally used for patients with acute renal failure who could not tolerate conventional hemodialysis. Currently, CRRT is indicated for sepsis therapy not only for renal support but also to abrogate immune dysregulation through multiple pathways and treatment modes to achieve multiple treatment goals and restore multi-organ function. In this study, clinical and laboratory parameters of the conventional and oXiris groups were similar before hemopurification (Table 1). Nevertheless, to reduce the bias of the covariates, we performed statistical analyses of the mean difference (difference method) of the observations, i.e., comparing the changes rate in relevant indicators between the two groups at 24 h and before starting treatment, and between 72 h and before starting treatment (△%).
Before comparing the efficacy of conventional and oXiris CRRT, we first assessed the therapeutic effect within each group. The conventional CRRT group displayed a lower HR but no other improvements in organ function (Table 2, Fig. 2), and no statistically significant differences in other relevant indexes before the start of the treatment (p>0.05). This finding differed from those of several recent studies.25 This discrepancy might be explained by: (1) Timing of the initiation of CRRT. Although international guidelines indicate when to initiate CRRT in patients with acute renal failure,26 there are currently no uniform standards for patients with sepsis. All patients in this study initiated CRRT treatment within 24 h of ICU admission. KIDGO (Kidney Disease: Improving Global Outcomes) staging was similar between the 2 groups (p > 0.05); however, renal function values were unknown before ICU admission, at which time renal injury may have been irreversible and therefore unresponsive to timely hemopurification treatment.6,27 (2) Observation time points: whether 24 h or 72 h time points are chosen for observation, false-negative results may result from premature observation. Severe organ dysfunction may not have improved by these time points, and similarly, these time points may also be associated with negative results after organ function has been restored to a stable state; the lack of significant improvement of indicators at 24 h in the oXiris group (Table 2, Fig. 2) also may have been due to the above reasons, as oXiris treatment improved organ function improvement at 48 h in one study.28 In this study, due to the blood collection procedure, systematic data loss, and the patient's family economic status, only 24 h and 72 h clinical data could be reviewed. Consequently, the time points were too few to illustrate the entire treatment course. Rigorous prospective studies are needed to further explore this issue. (3) Selection of control group: most studies have confirmed that CRRT improves patient outcomes.6,29 This study did not include a comparative analysis of patients without hemopurification. In addition, the frequency of line replacement, antibiotic use, the nutritional status at different stages, the number of drains, and drainage volume are also variables that should not be ignored.
Subsequently, our study compared the amount of change rate (△%) in the observed indexes in the 2 groups after 24 h and 72 h of treatment (Table 2, Fig. 2). Improvements in PCT, CRP, neutrophil ratio, and IL-6 levels in the oXiris group were significantly higher than those in the conventional CRRT group (p<0.05), suggesting the efficient removal of inflammatory mediators by the oXiris filter and subsequent prompt removal of endotoxins in these patients. Pro-inflammatory cytokine storms and immune dysregulation were mitigated, as confirmed by Bateman and other investigators.30 Congruent with recent studies,31,32 we found that compared to conventional filters, the oXiris filter significantly improved not only hemodynamics (HR, MAP) and Lac but also permitted a significant reduction in the dose of vasoactive medication (p<0.05) and also improved organ function (decreased APACHE II, SOFA scores, p<0.01). Although ICU and hospital lengths of stay did not differ between the 2 groups (p>0.05), which is consistent with the study by Navas et al.33 our study demonstrated that oXiris improved the clinical outcome in terms of 28-day mortality after removal of autodischarged patients (p<0.01).34
Logistic regression analysis of the relationships between time to antibiotic administration, time to source control, time to CRRT, APACHE II, and 28-day mortality rate in the 2 groups found that time to source control affected the 28-day mortality rate. However, our intervention time was relatively early, which of course is related to the availability of multiple medical specialty teams, management processes, patients’ families and social support, and many other factors at our hospital.
Our retrospective study design may have limited the scope of the investigation. More comprehensive prospective study designs are needed.35,36 Timing of CRRT did not differ between the 2 groups (p>0.05), and regression analysis showed that earlier CRRT initiation did not improve 28-day mortality. Similarly, the STARRT-AKI (Timing of Initiation of Renal-Replacement Therapy in Acute Kidney Injury) study that included a total of 3019 patients randomized to either an early group (initiation of CRRT within 12 h of meeting the KDIGO criteria for AKI stage 2) or a standard strategy group (CRRT was discouraged unless traditional indications were present or AKI persisted for >72 h), showed no intergroup difference in the 90-day mortality rate.37 These findings suggest that diagnostic evaluations should be patient-specific, with careful observation of the patient's dynamic clinical evolution following the basic principles of timely initiation of therapy in the presence of life-threatening disturbances of fluid, electrolyte, and acid-base balance. Our analysis did not disclose the benefit of early antibiotic intervention, but due to the small sample size, among other reasons, we still believe that empiric anti-infective therapy should be initiated within 1 h whenever possible in patients with IAI.
oXiris filters are pre-grafted with heparin, which can induce thrombocytopenia and increase the risk of hemorrhage. In this study, the prevalence of systemic heparinization was similar between the oXiris and conventional CRRT groups (52.17% vs. 51.85%, respectively; p > 0.05), and PLT was similar in both groups before and after treatment. Because oXiris is hemocompatible and does not affect PLT, the withholding of systemic heparinization does not affect filter life;38 consequently, heparin-free treatment with oXiris can be offered to patients with hemorrhagic diatheses and is, therefore, a new option for patients with contraindications to anticoagulation.39 Notably, the in vitro line portion of the complete filter does not exert anticoagulant activity and has the potential to activate the coagulation cascade, deplete coagulation factors, and reduce PLT, as reported in previous studies.39
Limitations of this study must be acknowledged: (1) Because of the retrospective study design, the timing of initiation of hemopurification could not be controlled. A comparison with patients without hemopurification was not conducted; consequently, the results cannot fully reflect the advantages of conventional CRRT in sepsis caused by IAI, and thus fail to fully demonstrate the advantages of oXiris filtration; (2) We could not assay endotoxin activity and the endotoxin removal effect of either filter; (3) The sample size of this study was small, inclusion criteria were strict, and results were limited; consequently, further studies with larger sample sizes and further subgroup analysis will be necessary; (4) Choices of filter replacement frequency, treatment mode, anticoagulation regimen, treatment dose, and time off the devices still require confirmation by more large-scale prospective randomized controlled studies.
In summary, oXiris filter may have advantages over conventional CRRT in severe abdominal infections, which may provide an alternative to clinical treatment. Meanwhile, early active infection source control may reduce the case mortality rate of patients with severe abdominal infections.
CRediT authorship contribution statement
Ye Zhou: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Ming-Jun Liu: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Xiao Lin: Writing – review & editing. Jin-Hua Jiang: Writing – review & editing. Hui-Chang Zhuo: Writing – review & editing, Funding acquisition, Conceptualization.
Ethical statement
The authors’ affiliations stated that retrospective studies are not applicable to ethical approval.
Funding
Startup Fund for Scientific Research of Fujian Medical University. 2022QH1102.
Declaration of competing interest
The authors report no conflicts of interest in this work.
Footnotes
Peer review under responsibility of Chinese Medical Association.
References
- 1.Vincent J.L., Rello J., Marshall J., et al. International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009;302:2323–2329. doi: 10.1001/jama.2009.1754. [DOI] [PubMed] [Google Scholar]
- 2.Marshall J.C. Intra-abdominal infections. Microb Infect. 2004;6:1015–1025. doi: 10.1016/j.suc.2014.08.005. [DOI] [PubMed] [Google Scholar]
- 3.Marshall J.C. Sepsis definitions: a work in progress. Crit Care Clin. 2018;34:1–14. doi: 10.1016/j.ccc.2017.08.004. [DOI] [PubMed] [Google Scholar]
- 4.Marshall J.C. Endotoxin in the pathogenesis of sepsis. Contrib Nephrol. 2010;167:1–13. doi: 10.1159/000315914. [DOI] [PubMed] [Google Scholar]
- 5.Kaizu K., Inada Y., Kawamura A., et al. Current status of blood purification in critical care in Japan. Contrib Nephrol. 2010;166:4–10. doi: 10.1159/000314844. [DOI] [PubMed] [Google Scholar]
- 6.Gaudry S., Hajage D., Schortgen F., et al. Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med. 2016;375:122–133. doi: 10.1056/NEJMoa1603017. [DOI] [PubMed] [Google Scholar]
- 7.Yamashita C., Moriyama K., Hasegawa D., et al. Evidence and perspectives on the use of polymyxin B-immobilized fiber column hemoperfusion among critically ill patients. Contrib Nephrol. 2018;196:215–222. doi: 10.1159/000485725. [DOI] [PubMed] [Google Scholar]
- 8.Viallet N., Brunot V., Kuster N., et al. Daily urinary creatinine predicts the weaning of renal replacement therapy in ICU acute kidney injury patients. Ann Intensive Care. 2016;6:71. doi: 10.1186/s13613-016-0176-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Malard B., Lambert C., Kellum J.A. In vitro comparison of the adsorption of inflammatory mediators by blood purification devices. Intensive Care Med Exp. 2018;6:12. doi: 10.1186/s40635-018-0177-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wu X.W., Ren J.A. Chinese guidelines for the diagnosis and treatment of abdominal infections (2019 edition) Chin J Pract Surg. 2020;40:1–16. doi: 10.19538/j.cjps.issn1005-2208.2020.01.01. [DOI] [Google Scholar]
- 11.Hawser S., Hoban D.J., Badal R.E., et al. Epidemiology and antimicrobial susceptibility of Gram-negative aerobic bacteria causing intra-abdominal infections during 2010-2011. J Chemother. 2015;27:67–73. doi: 10.1179/1973947814Y.0000000164. [DOI] [PubMed] [Google Scholar]
- 12.Sheng W.H., Badal R.E., Hsueh P.R. Distribution of extended-spectrum β-lactamases, AmpC β-lactamases, and carbapenemases among Enterobacteriaceae isolates causing intra-abdominal infections in the Asia-Pacific region: results of the study for Monitoring Antimicrobial Resistance Trends (SMART) Antimicrob Agents Chemother. 2013;57:2981–2988. doi: 10.1128/AAC.00971-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Honore P.M., De Bels D., Preseau T., et al. Efficacy of polymyxin B hemoperfusion in and beyond septic shock: is an "endotoxin severity score" needed? Crit Care. 2018;22:205. doi: 10.1186/s13054-018-2093-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Buttenschoen K., Radermacher P., Bracht H. Endotoxin elimination in sepsis: physiology and therapeutic application. Langenbecks Arch Surg. 2010;395:597–605. doi: 10.1007/s00423-010-0658-6. [DOI] [PubMed] [Google Scholar]
- 15.Assicot M., Gendrel D., Carsin H., et al. High serum procalcitonin concentrations in patients with sepsis and infection. Lancet. 1993;341:515–518. doi: 10.1016/0140-6736(93)90277-n. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kyriazopoulou E., Liaskou-Antoniou L., Adamis G., et al. Procalcitonin to reduce long-term infection-associated adverse events in sepsis. A randomized trial. Am J Respir Crit Care Med. 2021;203:202–210. doi: 10.1164/rccm.202004-1201OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Vouloumanou E.K., Plessa E., Karageorgopoulos D.E., et al. Serum procalcitonin as a diagnostic marker for neonatal sepsis: a systematic review and meta-analysis. Intensive Care Med. 2011;37:747–762. doi: 10.1007/s00134-011-2174-8. [DOI] [PubMed] [Google Scholar]
- 18.Simon L., Saint-Louis P., Amre D.K., et al. Procalcitonin and C-reactive protein as markers of bacterial infection in critically ill children at onset of systemic inflammatory response syndrome. Pediatr Crit Care Med. 2008;9:407–413. doi: 10.1097/PCC.0b013e31817285a6. [DOI] [PubMed] [Google Scholar]
- 19.Liu D., Su L., Han G., et al. Prognostic value of procalcitonin in adult patients with sepsis: a systematic review and meta-analysis. PLoS One. 2015;10 doi: 10.1371/journal.pone.0129450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Abasıyanık M.F., Wolfe K., Van Phan H., et al. Ultrasensitive digital quantification of cytokines and bacteria predicts septic shock outcomes. Nat Commun. 2020;11:2607. doi: 10.1038/s41467-020-16124-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ronco C., Tetta C., Mariano F., et al. Interpreting the mechanisms of continuous renal replacement therapy in sepsis: the peak concentration hypothesis. Artif Organs. 2003;27:792–801. doi: 10.1046/j.1525-1594.2003.07289.x. [DOI] [PubMed] [Google Scholar]
- 22.Atan R., Peck L., Prowle J., et al. A double-blind randomized controlled trial of high cutoff versus standard hemofiltration in critically ill patients with acute kidney injury. Crit Care Med. 2018;46:e988–e994. doi: 10.1097/CCM.0000000000003350. [DOI] [PubMed] [Google Scholar]
- 23.Cecconi M., Evans L., Levy M., et al. Sepsis and septic shock. Lancet. 2018;392:75–87. doi: 10.1016/S0140-6736(18)30696-2. [DOI] [PubMed] [Google Scholar]
- 24.Ceresoli M., Lo Bianco G., Gianotti L., et al. Inflammation management in acute diverticulitis: current perspectives. J Inflamm Res. 2018;11:239–246. doi: 10.2147/JIR.S142990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Chen G., Zhou Y., Ma J., et al. Is there a role for blood purification therapies targeting cytokine storm syndrome in critically severe COVID-19 patients? Ren Fail. 2020;42:483–488. doi: 10.1080/0886022X.2020.1764369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120:c179–c184. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
- 27.Zarbock A., Kellum J.A., Schmidt C., et al. Effect of early vs delayed initiation of renal replacement therapy on mortality in critically ill patients with acute kidney injury: the ELAIN randomized clinical trial. JAMA. 2016;315:2190–2199. doi: 10.1001/jama.2016.5828. [DOI] [PubMed] [Google Scholar]
- 28.Shum H.P., Chan K.C., Kwan M.C., et al. Application of endotoxin and cytokine adsorption haemofilter in septic acute kidney injury due to Gram-negative bacterial infection. Hong Kong Med J. 2013;19:491–497. doi: 10.12809/hkmj133910. [DOI] [PubMed] [Google Scholar]
- 29.Ronco C. Evolution of technology for continuous renal replacement therapy: forty years of improvement. Contrib Nephrol. 2018;194:1–14. doi: 10.1159/000485596. [DOI] [PubMed] [Google Scholar]
- 30.Bateman R.M., Sharpe M.D., Jagger J.E., et al. 36th international symposium on intensive care and emergency medicine : Brussels, Belgium. 15-18 march 2016. Crit Care. 2016;20:94. doi: 10.1186/s13054-016-1208-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Turani F., Barchetta R., Falco M., et al. Continuous renal replacement therapy with the adsorbing filter oXiris in septic patients: a case series. Blood Purif. 2019;47:1–5. doi: 10.1159/000499589. [DOI] [PubMed] [Google Scholar]
- 32.Schwindenhammer V., Girardot T., Chaulier K., et al. oXiris® use in septic shock: experience of two French centres. Blood Purif. 2019;47:S1–S7. doi: 10.1159/000499510. [DOI] [PubMed] [Google Scholar]
- 33.Navas A., Ferrer R., Martínez M.L., et al. Impact of hemoperfusion with polymyxin B added to hemofiltration in patients with endotoxic shock: a case-control study. Ann Intensive Care. 2018;8:121. doi: 10.1186/s13613-018-0465-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ma J., Xia P., Zhou Y., et al. Potential effect of blood purification therapy in reducing cytokine storm as a late complication of critically ill COVID-19. Clin Immunol. 2020;214 doi: 10.1016/j.clim.2020.108408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Hecker A., Reichert M., Reuß C.J., et al. Intra-abdominal sepsis: new definitions and current clinical standards. Langenbecks Arch Surg. 2019;404:257–271. doi: 10.1007/s00423-019-01752-7. [DOI] [PubMed] [Google Scholar]
- 36.Rhodes A., Evans L.E., Alhazzani W., et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock. Intensive Care Med. 2016;43:304–377. doi: 10.1007/s00134-017-4683-6. 2017. [DOI] [PubMed] [Google Scholar]
- 37.Joannes-Boyau O., Honoré P.M., Perez P., et al. High-volume versus standard-volume haemofiltration for septic shock patients with acute kidney injury (IVOIRE study): a multicentre randomized controlled trial. Intensive Care Med. 2013;39:1535–1546. doi: 10.1007/s00134-013-2967-z. [DOI] [PubMed] [Google Scholar]
- 38.Wang G., He Y., Guo Q., et al. Continuous renal replacement therapy with the adsorptive oXiris filter may be associated with the lower 28-day mortality in sepsis: a systematic review and meta-analysis. Crit Care. 2023;27:275. doi: 10.1186/s13054-023-04555-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Zhang L., Yan Tang G.K., Liu S., et al. Hemofilter with adsorptive capacities: case report series. Blood Purif. 2019;47:S1–S6. doi: 10.1159/000499357. [DOI] [PubMed] [Google Scholar]