Continuous renal replacement therapy with adsorbing filter oXiris in the treatment of sepsis associated acute kidney injury: a single-center retrospective observational study Authors: Feng Zheng, Yi-lan Wang, Wei-yi Zhou, Jing Zhang, Min Lu, Ni-fang Pan, Jian He, Qian Zhang, Lan Cao, Jiang-song Wu, Yan Gu, Li-hua Qiu, Hong-wei Ye Published: 2024 ABSTRACT AbstractBackground and objectiveCritical bedside ultrasound is widely used in clinical practice, and it can monitor renal perfusion. The reduction of renal perfusion and inflammatory injury are two contributing factors to sepsis-associated acute kidney injury (SA-AKI).The aim of this study was to examine whether the oXiris filter was useful in the continuous renal replacement therapy(CRRT) treatment of SA-AKI patients.Design, setting, participants, and measurementsWe performed a retrospective single-center observational study and enrolled two hundred and forty-three SA-AKI patients from January 2022 to December 2023, who were divided into the oXiris group (n = 88) and the control group (n = 155). The primary endpoints were the 28-day recovery of renal function and 28-day all-cause mortality. The secondary endpoints included renal Doppler markers (RRI, RVSI, and PDU), SOFA, vasoactive-inotropic score (VIS), inflammatory markers (PCT, CRP, IL-10 and TNFα), lactate level, and length of stay in ICU and hospital.ResultsFor the primary endpoint, the rates of complete recovery, partial recovery, and dialysis dependence were observed to be 60.3%, 13.6%, and 26.1% in the oXiris group, respectively, compared to 63.9%, 15.5%, and 20.6% in the control group. The 28-day all-cause mortality was not different in the two groups (22.7% vs. 27.1%). For the secondary endpoint, the oXiris group exhibited greater reductions in VIS scores compared to the control group within the first 24 h (p = 0.001) and 48 h (p < 0.001) of CRRT. Following 48-h of CRRT, lactate levels in the oXiris group were significantly lower than those in the control group (p = 0.014). Prior to CRRT, levels of IL-6 were higher in the oXiris group (p = 0.036), but these differences were not significant after CRRT (p > 0.05). The levels of RRI at T1 (p = 0.002) and T2 (p = 0.001) were lower in the oXiris group than in the control group. Even after adjusting for AKI stage, multivariable Cox regression analysis showed that SOFA and inflammatory factors (TNFα, IL-10, and IL-6), oXiris were significantly associated with a lower 28-day mortality among SA-AKI patients when compared to M150 [HR = 0.466, 95%CI 0.233–0.934, p = 0.031].ConclusionOur findings suggest that the use of the oXiris filter in CRRT is associated with reduced inflammatory injury and improvement in renal perfusion. However, it is not associated with improved 28-day recovery of renal function and 28-day all-cause mortality.Keywords: Variation of renal ultrasound markers, Continuous renal replacement therapy, Adsorbing filter oXiris, Sepsis associated acute kidney injury Introduction Sepsis is characterized by life-threatening organ dysfunction resulting from the dysregulation of the host response due to infection. In individuals presenting with infection or suspected infection, the diagnosis of sepsis is established when there is an increase of more than 2 points in the Sequential (Sepsis-related) Organ Failure Assessment (SOFA) score from the baseline [1]. Sepsis-associated acute kidney injury (SA-AKI) is a term used to describe the occurrence of acute kidney injury as a result of sepsis. The pathophysiological mechanisms underlying the development of SA-AKI involve the reduction in renal perfusion and inflammation-associated injury [2, 3]. Consequently, monitoring renal perfusion is imperative for the detection and treatment of SA-AKI. The advent and widespread adoption of critical care ultrasound technology have rendered it a necessary clinical skill for emergency and critical care physicians. Three renal perfusion ultrasound markers, renal resistive index (RRI), power Doppler ultrasound (PDU) score and renal venous stasis index (RVSI), have been reported as potential ultrasound markers for detecting AKI [4]. A study of 371 critical cases, including 123 patients with persistent AKI showed that RVSI had a high sensitivity and low specificity for the differentiating persistent AKI [5]. PDU functions by analyzing the reflected signals of ultrasound waves as they interact with the dynamics of blood flow. Specifically, the amplitude of the signal transmitted by red blood cells per unit area is utilized for imaging purposes [6]. Moreover, PDUs meet the criteria for convenient clinical application, non-invasiveness, and real-time monitoring [7]. The PDU score exhibits a negative correlation with the KDIGO stage of AKI, as evidenced by the fact that only 7.69% of patients with a PDU score of 3 are classified as AKI stage 3. Conversely, there is a notable increase in the proportion of AKI stage 3 patients with a PDU score below 2 [8]. Inflammatory injury is another key point in the mechanism of SA-AKI. Endotoxin plays a significant role in initiating the inflammatory cascade within the body. Upon entering the bloodstream in substantial quantities, endotoxin can stimulate the activation of macrophages, neutrophils, the complement system, and the coagulation system, thereby triggering a cascade of inflammatory responses and resulting in sepsis [9]. Consequently, multiple tissues and organs sustain secondary damage, with the kidney being particularly susceptible, leading to the development of SA-AKI [10]. Hence, it can be hypothesized that effective elimination of the endotoxins and inflammatory mediators, regulation of the sepsis cascade, and enhancement of the microcirculation might be beneficial in the management of SA-AKI. Among the various blood purification therapies, continuous renal replacement therapy (CRRT) is one of the most widely employed way due to its ability to rapidly and comprehensively improve the internal environment, making it the preferred treatment for SA-AKI [11]. It has been suggested that the membrane material of the filter, dialyzer, and hemadsorption device may impact the prognosis of sepsis patients undergoing blood purification treatment [2, 12, 13]. Optimal therapeutic outcomes can be attained by carefully selecting the appropriate membrane material based on specific treatment objectives. The oXiris is a surface modified membrane of the AN69 (a classical membrane of acrylic and sodium mesylate polymer). This base material has the ability to adsorb a diverse range of pro-inflammatory and anti-inflammatory factors [14]. Additionally, it can effectively adsorb negatively charged endotoxins through ionic bonds during the filtration process, thereby achieving the dual objective of endotoxin removal and prevention of upstream inflammatory reactions [15]. Therefore, the initial utilization of oXiris has the potential to effectively eliminate inflammatory factors and endotoxins in individuals with sepsis, thereby mitigating kidney damage and ultimately enhancing the overall clinical outlook for patients suffering from SA-AKI. Nevertheless, the improvement of renal perfusion after oXiris-CRRT treatment is still unclear. Thus, in this study, we conducted a retrospective observational study at a single center, enrolling a total of 243 cases, to examine the impact of oXiris-CRRT treatment in 28-day recovery of renal function, 28-day all-cause mortality and renal perfusion evaluated by renal ultrasound in patients with SA-AKI. Materials and methods Study design and participants After approval by the Clinical Research Ethics Committee of Suzhou University affiliated Changshu Hospital (X202323), we performed a retrospective single-center observational study. Due to the retrospective retrieval of the patients’ data, the informed consent was waived. Participants Our study enrolled two hundred and forty-three SA-AKI patients from January 2022 to December 2023 at Suzhou University affiliated Changshu Hospital (Jiangsu, China). Patients who met the inclusion criteria were eligible to: (1) age > 18, (2) Met the diagnostic criteria outlined in the sepsis 3.0 guidelines [16]; (3) matched the diagnostic criteria for AKI from KDIGO 2012 [17]; (4) CRRT for at least 24 h. Patients with the following criteria were excluded: (1) other factors contributing to AKI, including prior use of nephrotoxic medications, contrast-induced nephropathy, and urinary blockages; (2) pre-existing CKD; (3) pregnancy; (4) treatment with immunosuppressive drugs; (5) active neoplasia; (6) insufficiency clinical data; (7)coexisting disease with a high probability of death (< 6 months). All the patients were divided into the oXiris group (n = 88) and the control group (n = 155). The oXiris filter and M150 filter were used at least 72 h in CRRT treatment in oXiris group and control group, respectively. Study interventions All the enrolled patients were under CRRT treatment after SA-AKI diagnosis with either oXiris (Gambro lndustries, France) or M150 (Gambro lndustries, France). The Gambro Prismaflex was used during the CRRT process. CRRT model was continuous veno-venous hemofiltration with a blood flow rate of 150–180 ml/min, a pre-filter reinfusion flow of 20 ml/kg/h and a post-filter reinfusion flow of 15 ml/kg/h. The filter was replaced every 24 h. The oXiris filter was replaced on a daily basis. After 72 h of treatment with oXiris in the oXiris group, the M150 filter was replaced. Unfractionated heparin was used as the anticoagulant during CRRT. The femoral vein was chosen for vascular access, and a double-lumen 11.5 F hemodialysis catheter (DIALL, China) was inserted. Data collection Clinical data, encompassing baseline demographic characteristics (age and sex), comorbidities, infection sites, blood routine, and liver-kidney function, were documented at three distinct time points: upon the diagnosis of SA-AKI (T0), 24 h later (T1), and 48 h later (T2) following the initiation of CRRT treatment. The APACHE II score and SOFA score were documented at T0, T1, and T2. The severity of acute kidney injury and the level of kidney function were evaluated using the KDIGO 2012 criteria [17], which categorizes severity into three stages based on the elevation of creatinine levels from baseline, with stage 1 denoting the least severe and stage 3 indicating the most severe. Assessment of renal perfusion improvement All the ultrasound tests were performed by doctors experienced in critical care ultrasound to reduce personnel error. The medical ultrasound instrument used in this study was the M9 (Mindry, China). The methods of measurement were followed by previous reports [18–20]. Briefly, using a 2–6 MHz convex array probe, the long axis section of the kidney was obtained from the posterolateral side of the abdomen. Color Doppler was used to locate an interlobar artery for pulse Doppler examination, and peak systolic flow speed (VS) and minimum diastolic flow speed (DS) were measured. The RRI was calculated using the formula: RRI = (VS-VD)/VS. RI was determined in three different segmental arteries (upper, middle, and lower) of both kidneys and expressed as the mean of these values. The estimation of RVSI was conducted by assessing the peak systolic velocity minus the end-diastolic velocity, divided by the peak systolic velocity, within the renal vein. However, if the spectral waveform exhibited a monophasic or continuous pattern, the RVSI measurement was not performed, as it necessitates the presence of multiple phases for accurate calculation. The renal blood perfusion was observed with energy Doppler, and the optimal image was saved for off-line semiquantitative PDU analysis according to previous report [20]. The scoring criteria are as follows: a score of 0 indicates the absence of identifiable vessels; a score of 1 denotes the presence of a few visible vessels near the hilum; a score of 2 signifies the visibility of hilar and interlobar vessels throughout most of the renal parenchyma; and a score of 3 represents the identification of renal vessels extending to the arcuate arteries across the entire field of view. The right kidney was primarily evaluated for PDU due to its typically lower position and easier accessibility, while the left kidney was examined as a contingency in case the visualization of the right kidney was insufficient. Primary and secondary endpoints setting The primary endpoints were the 28-day recovery of renal function and 28-day all-cause mortality compared between the Oxiris group and the control group. The categorization of renal function recovery included complete recovery, partial recovery, and dependence on dialysis. Complete recovery was characterized by the restoration of Scr and eGFR levels to baseline levels and the presence of a normal urine test. Partial recovery was indicated by the absence of Scr and eGFR levels returning to basal level and/or the persistence of proteinuria and hematuria, without the requirement for hemodialysis. Dependence on dialysis was defined as the necessity for hemodialysis. The term 28-day refers to 28 days following the initiation of CRRT treatment. The secondary endpoints were renal Doppler markers (RRI, RVSI, and PDU), SOFA score, vasoactive-inotropic score (VIS), inflammatory markers (PCT, CRP, IL-10, and TNFα), lactate level, and length of stay in the ICU and hospital. Statistical analysis IBM SPSS Statistics (version 21.0, IBM, Armonk, NY, USA) and Stata (Stata SE, version 18) were used for data analysis. Continuous data were presented as means and standard deviations (SDs), or as medians and interquartile ranges (IQRs, 25th and 75th percentiles). Differences between two groups were tested using the Kruskal–Wallis test with the Mann Whitney U test, χ2 test, or Fisher’s exact test. Survival analysis was performed using the Kaplan–Meier curve, and the rates were compared using the log-rank test. Univariate and multivariate Cox regression analyses were performed to estimate the relative effect of outcome variables by calculating unadjusted odds ratios (ORs). The standardized mean difference (SMD) and its 95% confidence interval were calculated to determine the effect size. The interpretation of the SMD is as follows: no effect (0 < SMD < 0.2), small effect (0.2 ≤ SMD < 0.5), moderate effect (0.5 ≤ SMD < 0.8), and large effect (SMD ≥ 0.8) [21]. The significance level for the p value was set at 0.05. Results Enrollment and baseline demographic and clinical characteristics First, as shown in Fig. 1, our study enrolled 243 SA-AKI patients, including 88 oXiris cases and 155 M150 cases. As shown in Table 1, the results of baseline demographic and clinical characteristics before CRRT treatment indicated that the oXiris group had a higher number of hypoproteinemia cases (39.8% vs. 18.7%, p < 0.001) and electrolyte disorder cases (48.9% vs. 32.3%, p = 0.013) than the control group. The MAP levels in the oXiris group were higher than those in the control group (71.41 ± 6.99 mmHgvs. 67.65 ± 6.16 mmHg, p < 0.001). The SpO2/FiO2 levels in the oXiris group were lower than those in the control group (257.69 ± 64.66vs. 285.68 ± 79.25, p = 0.003). Flow chart of the patient enrollment. SA-AKI = sepsis associated acute kidney injury, CRRT = continuous renal replacement therapy, CKD = chronic kidney disease Table 1Baseline demographic and clinical characteristics before CRRT treatmentVariablesAll patientsoXiris groupM150 groupp valueSMD (95%CI)n = 243n = 88n = 155Age (years)70(56,77)72(63,78)70(54,76)0.1080.218(-0.044, 0.481)Female gender, n (%)85(35.0)29(33.0)56(36.1)0.676-BMI (Kg/m2)22.12 ± 2.0221.82 ± 2.0222.29 ± 2.000.083-0.234 (-0.497,0.028)AKI stage before CRRT treatment0.194- Stage 186(35.4)33(37.5)53(34.2)- Stage 289(36.6)26(29.5)63(40.6)- Stage 368(28.0)29(33.0)39(25.2)-History of Smoking, n (%)40(16.5)18(20.5)22(14.2)0.213-History of alcohol consumption, n (%)58(23.9)26(29.5)32(20.6)0.121-Hypertension, n (%)84(34.6)31(35.2)53(34.2)0.889-DM, n (%)59(24.3)16(18.2)43(27.7)0.119-Hyperlipidemia, n (%)38(15.6)12(13.6)26(16.8)0.584-CHD, n (%)27(11.1)9(10.2)18(11.6)0.834-COPD28(11.5)12(13.6)16(10.3)0.531-Respiratory failure146(60.1)58(65.9)88(56.8)0.175-Heart failure53(21.8)16(18.2)37(23.9)0.335-Coagulation disorders51(21)24(27.3)27(17.4)0.074-Hypoproteinemia64(26.3)35(39.8)29(18.7) < 0.001-Electrolyte disorder93(38.3)43(48.9)50(32.3)0.013-Thrombocytopenia97(39.9)40(45.5)57(36.8)0.220-MAP (mmHg)69.01 ± 6.7171.41 ± 6.9967.65 ± 6.16 < 0.0010.581 (0.314, 0.848)HR (bpm)112.52 ± 13.91114.78 ± 15.54111.23 ± 12.780.0560.257(-0.006, 0.519)SpO2/FiO2275.55 ± 75.37257.69 ± 64.66285.68 ± 79.250.003-0.377(-0.64, -0.113)CVP (mmHg)8.27 ± 3.548.44 ± 3.868.19 ± 3.350.5890.071(-0.191, 0.332)APACHE II score27.55 ± 8.3127.35 ± 8.1527.66 ± 8.420.345-0.037(-0.299, 0.224)SOFA score9.82 ± 2.5310.02 ± 2.489.70 ± 2.550.7830.127(-0.135, 0.389)Mechanical ventilation, n (%)106(43.6)40(45.5)66(42.6)0.688VIS score*11.22 ± 4.9012.95 ± 5.0010.23 ± 4.56 < 0.0010.576(0.309, 0.842)Blood transfusions, n (%)75(30.9)28(31.8)47(30.3)0.885-Data were presented as mean ± SD or median (lower quartile, upper quartile)BMI Body mass index, DM Diabetes mellitus, CHD Coronary heart disease, MAP Mean artery pressure, CVP Central venous pressure, HR Heart rate, bpm beat per minute, COPD Chronic obstructive pulmonary disease, VIS score Vasoactive-inotropic score*First 24 h of ICU admission Baseline demographic and clinical characteristics before CRRT treatment Data were presented as mean ± SD or median (lower quartile, upper quartile) BMI Body mass index, DM Diabetes mellitus, CHD Coronary heart disease, MAP Mean artery pressure, CVP Central venous pressure, HR Heart rate, bpm beat per minute, COPD Chronic obstructive pulmonary disease, VIS score Vasoactive-inotropic score *First 24 h of ICU admission Laboratory characteristics and conditions of infection Then, we analyzed the laboratory characteristics, and conditions of infection before CRRT treatment. As shown in Table 2, levels of BNP (p = 0.003), IL-6 (p = 0.036) and fibrinogen (p < 0.001) in the oXiris group were higher than those in the control group. There were no significant differences in the distribution proportion of main infection site and culture-proven infections between these two groups (p > 0.05). Table 2Laboratory characteristics and conditions of infection before CRRT treatmentVariablesAll patientsoXiris groupM150 groupp valueSMD(95%CI)n = 243n = 88n = 155White Blood Cells, *10919.4(12.5,24.8)19.8(15.2,24.2)18.9(12.5,26.5)0.6950.04(-0.222,0.302)CRP, mg/L145.3(67.8,204.5)167.6(91.3,218.8)143.2(67.5,199.9)0.0980.231(-0.031,0.494)Platelet, *109201(108,265)195.5(88.5,251.5)207(143,265)0.178-0.215(-0.477,0.047)Hematocrit, %31.59 ± 8.5132.42 ± 8.1131.12 ± 8.710.2550.152(-0.11,0.414)PaO2, mmHg90.48 ± 28.5489.66 ± 23.9990.95 ± 30.890.718-0.045(-0.307,0.217)PaCO2, mmHg38.96 ± 7.5737.79 ± 9.0839.62 ± 6.490.071-0.242(-0.505,0.02)AST, U/L57(45,76)57.5(45,87.5)56(45,76)0.5830.581(0.314,0.848)ALT, U/L45(32,66)44.5(27,76)45(33,66)0.6220.443(0.178,0.707)Total bilirubin, umol/L12.8(8.7,17.6)11.4(7.6,17.6)14.3(9.8,17.6)0.058-0.028(-0.289,0.234)BNP, ng/L655(453,6265)1300.5(509,12,290.5)598(443,3567)0.0030.465(0.2,0.73)Cardiac troponin I, ug/L0.07(0.03,0.13)0.10(0.03,0.17)0.05(0.03,0.12)0.1170.355(0.091,0.618)Procalcitonin, ng/mL17.2(9.5,35.4)19.55(8.63,54.8)15.4(9.6,34.3)0.2920.217(-0.045,0.48)TNFα, pg/mL37.65(12.33,62.60)32.17(10.18,62.60)43.22(13.24,61.09)0.521-0.112(-0.374,0.149)IL-10, pg/mL752.18(280.88,899.03)801.29(302.12,928.54)673.09(280.88,879.76)0.2200.167(-0.095,0.429)IL-6, pg/mL2834.33(875.42,3055.08)2884.00(1134.40,4196.75)2543.20(798.45,3050.59)0.0360.282(0.019,0.545)pH7.34 ± 0.117.34 ± 0.147.33 ± 0.090.3930.128(-0.134, 0.39)Lactate, mmol/L5.5(3.2,8.5)5.55(4.15,8.35)5.50(3.00,9.40)0.668-0.004(-0.266, 0.257)Serum creatinine, umol/L244(165,387)245.5 (161,365)244(165,388)0.265-0.211(-0.473, 0.052)BUN, mmol/L12.7(9.9,21.2)12.9(9.3,20.9)12.7(10.9,21.2)0.331-0.102(-0.364, 0.16)PT, s11.4(9.7,17.7)12.6(9.9,16.4)11(9.4,21.9)0.0570.01(-0.252,0.271)APTT, s46.44 ± 15.4746.92 ± 19.3646.16 ± 12.810.7140.049(-0.213,0.311)D dimer, mg/L3.55(1.6,8.8)3.94(2.12,8.76)3.44(1.65,8.76)0.4850.154(-0.108, 0.416)Fibrinogen, ug/mL2.55(2.13,3.45)3.45(2.55,4.56)2.54(1.65,8.76) < 0.0011.233(0.95, 1.517)Main infection site, n (%)0.970- Lung94(38.7)33(37.5)61(39.4)- Abdominal78(32.1)30(34.1)48(31)- Urinary45(18.5)15(17)30(19.4)- skin and soft tissue14(5.8)5(5.7)9(5.8)- Other12(4.9)5(5.7)7(4.5)-Culture-proven infection, n (%)0.646- Gram positive88(36.2)36(40.9)52(33.5)- Gram negative111(45.7)38(43.2)73(47.1)- Gram positive and negative15(6.2)4(4.5)11(7.1)- Fungal infection29(11.9)10(11.4)19(12.3)-Data were presented as median (lower quartile, upper quartile)AST aspartate aminotransferase, ALT alanine aminotransferase, BNP brain natriuretic peptide, BUN blood urea nitrogen, PT prothrombin time, APTT activated partial thromboplastin time Laboratory characteristics and conditions of infection before CRRT treatment Data were presented as median (lower quartile, upper quartile) AST aspartate aminotransferase, ALT alanine aminotransferase, BNP brain natriuretic peptide, BUN blood urea nitrogen, PT prothrombin time, APTT activated partial thromboplastin time Primary endpoint As shown in Table 3, the rates of complete recovery, partial recovery, dialysis-dependent were 60.3%, 13.6%, 26.1% in the control group and 63.9%, 15.5%, 20.6% in the oXiris group, respectively. There was no significant difference in 28-day recovery of renal function (p = 0.608). As shown in Fig. 2, the 28-day all-cause mortality was lower in oXiris group than that in control group (22.7% vs. 27.1%) but the difference was not significant between the two group (p = 0.58). Table 328-day recovery of renal functionVariablesAll patientsoXiris groupControl groupp valuen = 243n = 88n = 155Complete recovery, n (%)152(62.6)53(60.3)99(63.9)0.608Partial recovery, n (%)36(14.8)12(13.6)24(15.5)Dialysis-dependent, n (%)55(22.6)23(26.1)32(20.6) 28-day recovery of renal function The 28-day all-cause mortality. In the Kaplan–Meier curve analysis, the number of death at 28 days after CRRT treatment were 20 (22.7%) and 42 (27.1%) in oXiris group and control group, respectively. The difference of mortality was not significant between the two groups (p = 0.58) Secondary endpoints In terms of the critical score, reduction of SOFA score in the oXiris group was higher than that in the control group in second 24 h of CRRT treatment [-1.00 (-3.00, 0) vs. -0.25 (-2.00, 1.00), p = 0.028] (Fig. 3A). Reductions of VIS score in oXiris group were higher than those in control group in both first 24 h (p = 0.001) and 48 h (p < 0.001) of CRRT treatment (Fig. 3B). The levels of lactate in oXiris group were lower than control group after 48 h CRRT treatment [2.2 (1.2, 3.85) vs. 2.6 (2.0, 4.2), p = 0.014] (Fig. 3C). There were no significant differences in the length of ICU and hospital stay between oXiris group and control group (p > 0.05) (Fig. 3D). Secondary endpoints. A Reduction of SOFA score within the 48-h CRRT treatment period; B Reduction of VIS score within the 48-h CRRT treatment period; C levels of lactate; D length of stay in ICU and hospital; E levels of procalcitonin; F levels of TNFα; G levels of IL-6; H levels of IL-10. All the panels show the median of each index. T0 = the day SA-AKI was diagnosed (before CRRT treatment); T1 = 24 h after CRRT treatment; T2 = 48 h after CRRT treatment In terms of the inflammatory markers, after 24 h CRRT treatment, the levels of TNFα and IL-10 in the oXiris group were lower than those in the control group (Fig. 3F and H). The levels of IL-6 in the oXiris group were higher than those in the control group before CRRT treatment [2884.0 (1134.4, 4196.8)vs.2543.2 (798.5, 3050.6), p = 0.036] and these differences disappeared after CRRT treatment (p > 0.05) (Fig. 3G). There were no significant differences in the levels of procalcitonin between oXiris group and control group (p > 0.05) (Fig. 3E). In the fields of renal ultrasound markers, the levels of RRI, PDU and RVSI in the oXiris group presented no significant differences compared to those in the control group before CRRT treatment (p > 0.05). After initiation of CRRT treatment, we found that the levels of RRI at T1 (0.688 ± 0.045 vs. 0.709 ± 0.049, p = 0.002) and T2 (0.666 ± 0.044 vs. 0.687 ± 0.047, p = 0.001) in oXiris group were lower than those in control group. The levels of PDU at T1 (1.81 ± 0.69 vs. 1.59 ± 0.66, p = 0.018) and T2 (2.48 ± 0.68 vs. 2.21 ± 0.61, p = 0.004) in oXiris group were higher than those in control group (Fig. 4). Renal ultrasound markers. A RRI; B PDU; C RVSI. All the panels show the mean value and standard deviation. RRI = renal resistive index; PDU = power Doppler ultrasound; RVSI = renal venous stasis index; T0 = the day SA-AKI was diagnosed (before CRRT treatment); T1 = 24 h after CRRT treatment; T2 = 48 h after CRRT treatment Cox regression analysis Then, in order to analyze whether oXiris presented more a protective effect than M150 in the prognosis of SA-AKI patients, we performed the univariate and multivariable Cox regression analyses. As shown in Table 4, the results of univariate logistic regression indicated that oXiris (vs. M150) did not show a significant association with 28-day mortality [HR = 0.605, 95%CI 0.350–1.047, p = 0.072]. After adjusting for the AKI stage and SOFA score, oXiris showed a negative association with the 28-day mortality compared to M150 [HR = 0.468, 95%CI 0.237–0.925, p = 0.029]. Furthermore, after adjusting for the AKI stage, SOFA score and inflammatory factors (TNFα, IL-10 and IL-6), the oXiris remained associated with a lower 28-day mortality among SA-AKI patients[HR = 0.466, 95% CI 0.233–0.934, p = 0.031]. Table 4Cox regression analysis of variables associated with 28-day mortality of SA-AKI patientsModelHR (95%CI)p-value Model 10.605(0.350, 1.047)0.072 oXiris (vs. M150)Model 2 oXiris (vs. M150)0.468(0.237, 0.925)0.029AKI stage before CRRT0.040 Stage 1 (vs. Stage 3)0.510 (0.276, 0.941)0.031 Stage 2 (vs. Stage 3)0.407 (0.176, 0.941)0.036SOFA score1.091(1.010, 1.180)0.028Model 3 oXiris (vs. M150)0.466 (0.233, 0.934)0.031AKI stage before CRRT0.037 Stage 1 (vs. Stage 3)0.488(0.243, 0.979)0.043 Stage 2 (vs. Stage 3)0.397(0.167, 0.945)0.037SOFA score at T01.095(1.003, 1.195)0.042TNFα at T00.999(0.998, 1.011)0.925IL-6 at T01.000 (1.000, 1.000)0.859IL-10 at T01.000 (0.999, 1.001)0.696Adjust variables in Model 2 are the AKI stage and critical score, in Model 3 are the inflammatory markers. T0 = the day SA-AKI diagnosed (before CRRT treatment) Cox regression analysis of variables associated with 28-day mortality of SA-AKI patients Adjust variables in Model 2 are the AKI stage and critical score, in Model 3 are the inflammatory markers. T0 = the day SA-AKI diagnosed (before CRRT treatment) Discussion Sepsis is characterized by the presence of organ dysfunction, with research suggesting that approximately 40% to 50% of individuals with sepsis may experience the development of SA-AKI. SA-AKI has been associated with a significant 6- to eightfold rise in mortality rates and an increased risk of transitioning to Chronic Kidney Disease [22–24]. Currently, SA-AKI is believed to be associated with inflammatory damage and decreased renal perfusion [2, 3]. This study conducted a retrospective observational analysis at a single medical center, enrolling a total of 243 cases to assess the impact of oXiris on the 28-day recovery of renal function and mortality, as well as its potential to improve renal perfusion and reduce inflammatory injury in patients with SA-AKI. The findings of our study suggest that the use of the oXiris filter in CRRT can effectively mitigate inflammatory injury, enhance renal perfusion, and was associated with a lower 28-day mortality in SA-AKI patients. Our study set two primary endpoints. The findings indicated that there were no statistically significant differences between the two groups in terms of 28-day recovery of renal function and 28-day all-cause mortality. This was a retrospective study. The selection of membrane type was based on our clinicians’ experience. Normally, the patients with unstable hemodynamics and severe toxemia were more likely to be treated with oXiris-CRRT. Thus, the patients in the oXiris group were more severely ill than those in the control group. It is noteworthy that the oXiris group exhibited a significantly higher incidence of hypoproteinemia and electrolyte disorders compared to the control group. These findings suggest that the patients in the oXiris group has higher severity of illness than the patients in the control group. Additionally, the oXiris group displayed markedly elevated levels of VIS and BNP in comparison to the control group, indicating a more severe circulatory status in the oXiris group. Furthermore, the elevated levels of IL-6 and fibrinogen in the oXiris group compared to the control group indicate a greater degree of inflammatory injury and hypercoagulability. These findings suggest that the oXiris group exhibited more severe symptoms prior to receiving CRRT treatment. Therefore, we hypothesize that the lack of a significant difference in the primary endpoint between the oXiris and control groups may have been due to a higher severity of illness at the beginning of treatment in the oXiris group compared to the control group. A meta-analysis enrolling 695 cases indicated that treatment with the oXiris filter during CRRT in sepsis patients might be associated with lower 28-, 7-, and 14-day mortalities, as well as reduced lactate levels, SOFA scores, NE doses, and a shorter length of ICU stay [25]. Another study indicated that while there was no significant difference in 90-day mortality, the use of oXiris may potentially lower short-term (< 14-day) mortality compared to the ST150 groups in patients with septic shock and acute kidney injury [15]. Our Cox regression analysis showed that the use of oXiris (vs. M150) was significantly associated with a lower 28-day mortality among SA-AKI patients after adjusting for the AKI stage, SOFA score, and inflammatory factors (TNFα, IL-10 and IL-6). Additional research in randomized controlled trials or high-quality prospective studies is necessary to confirm these results. In terms of secondary endpoints, our study revealed that the reduction in SOFA score within the oXiris group was significantly greater than that observed in the control group during the second 24 h of CRRT treatment. Additionally, reductions in the VIS score within the oXiris group were found to be higher than those in the control group during both the initial 24 h and the subsequent 48 h of CRRT treatment. Furthermore, levels of lactate within the oXiris group were significantly lower than those within the control group following 48 h of CRRT treatment. These findings suggest that oXiris filter may be associated with improvements in severity and organ perfusion in SA-AKI patients. A previous study also found that oXiris significantly decreased the VIS score at 24 and 48 h following the commencement of CRRT [15]. Another study demonstrated that the SOFA scores of patients following CRRT with oXiris therapy exhibited a statistically significant decrease compared to baseline scores on days 1, 3, and 7 post-treatment [26]. We tested three inflammatory markers and procalcitonin levels before and after CRRT treatment as one of the secondary endpoints. Our study revealed that the concentrations of TNFα and IL-10 in the oXiris group were significantly reduced compared to the control group following 24 h of CRRT. Additionally, the levels of IL-6 in the oXiris group were initially elevated relative to the control group prior to CRRT treatment, but this disparity was no longer observed post-treatment. These findings demonstrate the ability of oXiris to clear inflammatory markers, indicating its potential anti-inflammatory properties. However, this notion is not supported by the literature, as there are limited RCTs to substantiate this claim. Another retrospective investigation demonstrated that the utilization of CRRT with the oXiris filter could potentially enhance hemodynamic stability through the suppression of inflammatory mediators, which may be associated to a more expedited resuscitation process and a reduction in overall fluid accumulation during the resuscitative stages [26]. In this retrospective study, the authors concluded that only associations could be drawn, and no causation was reported. Kang's research findings indicate that the application of oXiris in non-AKI COVID patients should be approached with caution to prevent exacerbating societal and individual burdens [27]. Nevertheless, the data from their study demonstrate that oXiris may effectively reduce inflammatory markers, including IL-2, IL-6, and IL-10. Similar studies have shown the effectiveness of oXiris in eliminating the inflammatory factors [28–30]. Renal ultrasound was one of the secondary endpoints and used to evaluate renal perfusion. Following the commencement of CRRT treatment, it was observed that the levels of RRI at time points T1 and T2 were lower in the oXiris group compared to the control group. Conversely, the levels of PDU at time points T1 and T2 were higher in the oXiris group than in the control group. The findings suggest that oXiris filter CRRT treatment may be more effective in enhancing renal perfusion. Though this could be due to improved MAP or a reduced need for vasopressors, as observed by the decline in VIS and SOFA. The improved MAP and decreased need for vasopressors are more common after oXiris treatment, according to our experiments and other reports [15, 31, 32]. In our opinion,these associations are potentially due to the clearance of inflammation markers by oXiris treatment which could lead to the improved MAP or a reduced need for vasopressors. There are few studies focusing on the change in renal perfusion using ultrasound for evaluation in SA-AKI following oXiris treatment. A case report observed that 48 h of oXiris treatment resulted in enhanced clinical perfusion parameters and decreased inflammatory markers [33]. Our research assessed the impact of oXiris therapy on enhancing renal perfusion through the utilization of ultrasound imaging. These findings serve as a theoretical adjunct to the use of oXiris in the management of SA-AKI cases. There are some limitations in this study. First, due to the retrospective nature of this single-center study, there is a potential for bias. Future research should include a prospective study with a larger population to confirm the validity of our findings. Second, we conducted a 28-day follow-up period, which may have resulted in missing cases of AKI that progressed to CKD. A longer duration of follow-up would be more appropriate for capturing these cases. Third, the baseline levels exhibited heterogeneity; the disease severity between the two groups of patients did not align completely, and the illness status was more severe in the oXiris group. In conclusion, our study included 243 SA-AKI patients, with 88 patients treated with the oXiris filter and 155 patients treated with M150. For the primary endpoints, there was no significant difference in 28-day recovery of renal function and 28-day all-cause mortality between the oXiris group and the control group. However, for the secondary endpoints, our findings suggest that the use of the oXiris filter in CRRT may effectively reduce inflammatory injury, improve renal perfusion, and potentially enhance the prognosis of SA-AKI patients. Acknowledgements The authors would like to thank all of the participants for their time and effort. Disclosure statement No potential conflict of interest was reported by the author(s). Abbreviations Sepsis associated acute kidney injury Sequential organ failure assessment Serum creatinine Renal resistive index Power Doppler ultrasound Renal venous stasis index Continuous renal replacement therapy Peak systolic flow speed Minimum diastolic flow speed Vasoactive-inotropic score Standard deviations Odds ratios Authors’ contributions FZ, YLW, WYZ and JZ conceived the research. LHQ, YLW, WYZ, JZ, ML, NFP, JH, QZ, LC, JSW and YG collected and analyzed the clinical data. FZ designed the study. FZ drafted the manuscript. HWY and LHQ reviewed the final manuscript. All authors read and approved the final manuscript. Funding This study was supported by the special subject of Suzhou key clinical diseases diagnosis and treatment technology (LCZX2021123), and the youth subject of Changshu Health Commission science and technology plan (CSWSQ202101). Data availability The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The Clinical Research Ethics Committee of Suzhou University affiliated Changshu hospital (2024X01). All methods were carried out in accordance with relevant guidelines and regulations. Due to the retrospective retrieval of the patients data, the informed consent was waived. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Feng Zheng, Yi-lan Wang, Wei-yi Zhou and Jing Zhang contributed equally to this work and share first authorship. Contributor Information Li-hua Qiu, Email: zhouyuling2013@163.com. Hong-wei Ye, Email: yehongwei@foxmail.com. References Associated Data Data Availability Statement The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.