Work

Aggressive fever suppression in septic patients without known cardiorespiratory impairment is associated with delayed SOFA score improvement

Year: 2025

Type: article

Abstract: To the Editor: Fever is one of the oldest recognized signs of infection, yet its management in septic patients remains controversial, with inconclusive findings.[1,2] Given the diverse effects of fever in sepsis, it is crucial to identify patients who can benefit from its antimicrobial and immunomodulatory properties while tolerating the increased metabolic demands of elevated body temperature (BT). Schulman et al[3] provided indirect support for this concept, demonstrating that aggressive fever treatment in younger critically ill patients (mean age 47 years) was associated with higher mortality compared to a permissive approach. Age may be a key explanatory factor, as younger patients typically have fewer chronic cardiorespiratory impairments and greater physiological reserve, potentially enabling a more effective adaptive response to fever. A personalized approach with precise patient selection criteria may help address this knowledge gap. Therefore, our randomized controlled trial evaluated two antipyretic strategies (aggressive vs. conservative) in septic patients with a low risk of cardiorespiratory maladaptation, assessing their impact on the Sequential Organ Failure Assessment (SOFA) score.[4] This randomized controlled trial was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University Hospital of Ostrava (No. 523/2011). The trial was registered at ClinicalTrials.gov (No. NCT04227652). All adult patients admitted to the intensive care units (ICUs) of Department of Anaesthesiology and Intensive Care Medicine, University Hospital of Ostrava were screened for fever of infectious origin, defined as a BT >38.3°C, excluding those admitted after planned orthopedic surgery. Febrile patients with suspected or confirmed infection were considered eligible for inclusion. Exclusion criteria included: postcardiopulmonary resuscitation status, traumatic or nontraumatic brain injury, prior antipyretic treatment, ICU admission from another ward >72 hours, pregnancy, moribund condition, absence of informed consent, and a predefined risk of cardiorespiratory maladaptation to fever. The latter included chronic heart failure, history of myocardial infarction, coronary artery bypass grafting, valvular heart disease, cardiac arrhythmias or predisposition to arrhythmias, obstructive or restrictive lung disease, and planned lung surgery. Eligible patients were randomized into two arms using a sealed envelope method. Tinted, opaque, sealed envelopes containing the study arm label were prepared according to a computer-generated randomization sequence and distributed among the four participating ICUs before the study commenced. When a suitable patient was identified, an envelope was selected in sequence, opened, and the patient was assigned to the corresponding study arm. During screening, BT was monitored using axillary temperature sensors; post-randomization, core BT was measured via temperature-sensing Foley catheters. Both arms received pharmacological and physical antipyretic interventions. Ibuprofen (800 mg orally or via nasogastric tube) was the first-line agent; rectal administration was used if required. If contraindicated or ineffective, acetaminophen or metamizole was administered. Physical cooling methods included cooling blankets and environmental measures. In the aggressive treatment arm (AGR-arm), antipyretics were administered when BT exceeded 38.3°C, with the goal of reducing it below this threshold. In the conservative arm (CNV-arm), treatment commenced only if BT exceeded 39.5°C and was stopped once it fell below that level. The primary outcome was the SOFA score on days 3 and 7 post-randomization. Secondary outcomes included serum procalcitonin (PCT), interleukin-6 (IL-6), lactate, and central venous oxygen saturation (ScvO2), measured on days 3 and 7 post-randomization. Tertiary outcomes were duration of mechanical ventilation, ICU/hospital length of stay, and mortality. Sample size calculation was conducted using power analysis, referencing the Schulman et al’s trial.[3] It was assumed that aggressive fever treatment would result in a 14% worsening in outcome. To detect this effect with 80% power and a 5% significance level, each trial arm required 57 patients. The power analysis was conducted using PASS 13 (NCSS, LLC, Kaysville, Utah, USA). More detailed methodology is provided in the Supplementary Material, https://links.lww.com/CM9/C685. Between September 2013 and February 2019, 5998 ICU patients were screened, and 609 (10.2%) met the inclusion criteria. Following exclusions, 154 patients were randomized: 79 to the AGR-arm and 75 to the CNV-arm [Supplementary Figure 1, https://links.lww.com/CM9/C685]. Fourteen patients were lost due to protocol violations, resulting in 69 patients in the AGR-arm, and 71 in the CNV-arm included in the final analysis. Baseline characteristics were comparable between the arms [Supplementary Table 1, https://links.lww.com/CM9/C685]. The mean Acute Physiology and Chronic Health Evaluation II (APACHE II) scores confirmed similar illness severity (21.3 ± 9.2 vs. 20.1 ± 8.9; P = 0.453). All patients fulfilled the criteria for sepsis, severe sepsis, or septic shock as defined by Levy et al[5]. Over half of the patients in both arms presented with septic shock, and more than 80% required mechanical ventilation. Details of septic episodes are provided in Supplementary Table 2, https://links.lww.com/CM9/C685. Pulmonary infection was the most common source of sepsis. The proportion of patients with microbiologically confirmed infection and the adequacy of initial antibiotic therapy were comparable between the arms. The total duration of fever was significantly shorter in the AGR-arm (median 12.0 h [interquartile range [IQR] 37.0 h]) compared to the CNV-arm (median 34.0 h [IQR 86.5 h]; P = 0.002). Mean BT during fever was also lower in the AGR-arm (38.6°C ± 0.2°C vs. 38.8°C ± 0.3°C; P = 0.001). A comparison of daily fever duration between the arms over 7 days post-randomization is shown in Supplementary Figure 2, https://links.lww.com/CM9/C685. Ibuprofen use alone was more frequent in the AGR-arm (81.2% [56/69] vs. 57.7% [41/71]; P <0.01), while other antipyretic and physical cooling use was comparable. The primary, secondary, and tertiary endpoints are summarized in Table 1. The evolution of the SOFA score, IL-6, PCT, and lactate levels over the study period is presented in Supplementary Figure 3, https://links.lww.com/CM9/C685. The primary and secondary endpoints at randomization were comparable between the arms. On day 3, the SOFA score remained similar between the arms; however, by day 7, it was significantly higher in the AGR-arm (median 8.5 [IQR 7.5]) compared to the CNV-arm (median 6.0 [IQR 5.0]; P = 0.028). When comparing day 7 to randomization, the SOFA score remained unchanged in the AGR-arm but decreased significantly in the CNV-arm (P <0.001). Table 1 - Study endpoints of different fever suppression strategies in septic patients without known cardiorespiratory impairment Endpoints Aggressive arm (n = 69) Conservative arm (n = 71) P values Primary endpoint SOFA score on RND 9.0 (6.0) 10.0 (8.0) 0.351 SOFA score on day 3 8.0 (9.5) 9.0 (6.0) 0.271 SOFA score on day 7 8.5 (7.5) 6.0 (5.0)* 0.028 Secondary endpoints PCT level on RND 1.21 (4.78) 0.89 (5.6) 0.852 PCT level on day 3 0.82 (2.04) 0.81 (2.03) 0.799 PCT level on day 7 0.61 (0.79)† 0.32 (0.67)* 0.245 IL-6 level on RND 176 (412.5) 183 (509.6) 0.926 IL-6 level on day 3 51.4 (83.2)* 64.6 (95.9)* 0.570 IL-6 level on day 7 41.7 (53.4)* 27.6 (68.9)* 0.918 Lactate level on RND 1.4 (1.0) 1.45 (1.2) 0.477 Lactate level on day 3 1.1 (0.9)† 1.1 (0.6)‡ 0.744 Lactate level on day 7 1.5 (0.9) 1.1 (0.6)‡ 0.036 ScvO2 level on RND 73 (17.3) 74 (13.1) 0.444 ScvO2 level on day 3 73.5 (13.5) 72.8 (10.2) 0.432 ScvO2 level on day 7 71.3 (8.3) 72 (9.6) 0.794 Tertiary endpoints Mechanical ventilation (h) 207.8 (290.5) 237.3 (328.1) 0.545 Length of ICU stay (days) 12 (16) 14 (17) 0.435 Length of hospital stay (days) 26 (31) 23 (17) 0.525 ICU mortality 15 (21.7) 15 (21.1) 0.999 In-hospital mortality 17 (24.6) 17 (23.9) 0.999 Data are shown as n (%) or median (interquartile range). The primary endpoint was the total SOFA score on days 3 and 7 following randomization. Secondary endpoints included serum levels of PCT, IL-6, lactate, and ScvO2, measured on days 3 and 7 post-randomization. *P <0.001 compared to RND; †P <0.01 compared to RND; ‡P <0.05 compared to RND. IL-6: Interleukin-6; PCT: Procalcitonin; RND: Randomization; ScvO2: Central venous oxygen saturation; SOFA: Sequential Organ Failure Assessment. No significant differences were observed between the arms in IL-6, PCT, or ScvO2 levels. However, the serum lactate level was significantly higher on day 7 in the AGR-arm (1.5 mmol/L [IQR 0.9 mmol/L]) compared to the CNV-arm (1.1 mmol/L [IQR 0.6 mmol/L]; P = 0.036). There were no significant differences in the duration of mechanical ventilation, ICU or hospital length of stay, or mortality (ICU mortality: 21.7% [15/69] vs. 21.1% [15/71]; in-hospital mortality: 24.6% [17/69] vs. 23.9% [17/71]). To account for incomplete SOFA data due to early discharge or death, a subgroup analysis was conducted in patients who remained in ICU ≥7 days. Findings aligned with the overall trial and are detailed in the Supplementary Material, https://links.lww.com/CM9/C685. A total of 2890 microbiological cultures were obtained, yielding 1050 pathogens. The lungs were the most common source (45%), consistent with pulmonary infection being the leading cause of sepsis in the trial. While the number of cultures per patient was similar, the AGR-arm showed a significantly higher number of pathogens per examination (median 0.41 [IQR 0.27] vs. 0.32 [IQR 0.27]; P = 0.034). This randomized controlled trial effectively fulfilled its aim to evaluate the impact of two antipyretic strategies in septic patients at low risk of cardiorespiratory maladaptation to fever. Comparable baseline characteristics, including APACHE II scores and primary and secondary endpoints at randomization, confirmed similar illness severity and underscored the homogeneity of the study population. The significantly shorter duration of fever and lower body temperature in the AGR-arm indicate good protocol adherence and confirm that fever was more actively present and sustained in the CNV-arm. Importantly, the absence of elevated lactate and reduced ScvO2 in the CNV-arm suggests that fever-related cardiorespiratory deterioration did not occur. On the other hand, aggressive fever suppression was associated with delayed resolution of organ dysfunction, as reflected by the higher SOFA score on day 7 [Table 1 and Supplementary Figure 3A, https://links.lww.com/CM9/C685], and a greater number of pathogens cultured in microbiological examinations. Among the secondary endpoints, IL-6 and PCT were chosen to assess the immunomodulatory and antimicrobial effects of fever, respectively. IL-6, a marker of systemic inflammation, showed nearly identical trends in both arms [Supplementary Figure 3B, https://links.lww.com/CM9/C685], suggesting that the observed difference in the SOFA score was not attributable to an exaggerated inflammatory response in the AGR-arm. The antimicrobial effect of fever was evaluated using PCT, a surrogate marker for infection resolution. Although differences in PCT levels between the arms did not reach statistical significance, there was a consistent trend toward higher PCT values in the AGR-arm [Supplementary Figure 3C, https://links.lww.com/CM9/C685]. This, together with the increased number of cultured pathogens, raises the possibility that aggressive fever suppression may impair bacterial clearance and increase susceptibility to superinfections, thereby prolonging the recovery of organ function. In conclusion, our findings suggest that tolerating fever up to 39.5°C in septic patients without significant cardiorespiratory comorbidities is a safe alternative to routine early antipyretic treatment. This strategy was not associated with adverse hemodynamic outcomes, prolonged ICU/hospital stay, or increased mortality. Furthermore, conservative fever strategy may promote faster resolution of organ dysfunction, possibly by reducing the risk of secondary infections. While the observed difference of more than 2 points in the SOFA score on day 7 may be clinically relevant, confirmation in larger, multicenter trials focusing on patient-centered outcomes is needed before definitive practice recommendations can be made. Acknowledgement The authors sincerely thank the physicians and nursing staff of the Department of Anesthesiology and Intensive Care in Ostrava for their invaluable assistance and cooperation in conducting this trial. Funding The trial was supported by a grant from the Ministry of Health of the Czech Republic-conceptual development of research organization (No. 2/RVO/FNOs/2013). Conflicts of interest None.

Source: Chinese Medical Journal

DOI: 10.1097/cm9.0000000000003892