Impact of neuromuscular blockade depth on postoperative systemic cytokine release: a systematic review and meta-analysis

Article information

Korean J Anesthesiol. 2026;79(1):56-68

Publication date (electronic) : 2025 November 5

doi : https://doi.org/10.4097/kja.25279

Soowon Lee^1,2,3,*, Jung-Hee Ryu^1,2,*, Chang-Hoon Koo^1,2, Yu Kyung Bae¹, Ah-Young Oh^,1,2

¹Department of Anesthesiology and Pain Medicine, Seoul National University Bundang Hospital, Seongnam, Korea

²Department of Anesthesiology and Pain Medicine, Seoul National University College of Medicine, Seoul, Korea

³Center for Synaptic Brain Dysfunctions, Institute for Basic Science (IBS), Daejeon, Korea

Corresponding author: Ah-Young Oh, M.D., Ph.D. Department of Anesthesiology and Pain Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro 173beon-gil, Bundang-gu, Seongnam 13620, Korea Tel: +82-31-787-7509 Fax: +82-31-787-4063 Email: ohahyoung@hanmail.net; oay1@snubh.org

*Soowon Lee and Jung-Hee Ryu have contributed equally to this work as co-first authors.

Received 2025 April 11; Revised 2025 October 12; Accepted 2025 October 14.

Abstract

Introduction

Neuromuscular blockade (NMB) is a fundamental component of general anesthesia, widely used to facilitate intubation, optimize surgical conditions, and enhance patient outcomes. Although conflicting opinions remain, deep NMB has demonstrated potential advantages, including reduced intra-abdominal pressure, improved surgical field conditions, fewer surgical complications, and diminished postoperative pain, particularly in minimally invasive laparoscopic surgeries [1–4]. The adoption of deep NMB has become feasible owing to the introduction of sugammadex that enables rapid and complete reversal of deep NMB. However, even with sugammadex, the use of quantitative neuromuscular monitoring is essential to avoid residual NMB and ensure patient safety [4–6].

Surgical interventions inevitably result in tissue damage and stretching that activate the immuno-inflammatory responses and damage-associated molecular pattern pathways [7,8]. This activation triggers the release of pro-inflammatory cytokines, such as interleukin (IL)-1β, IL-6, and tumor necrosis factor-alpha (TNF-α), from immune cells [7,9]. These systemic cytokines mediate the inflammatory response of the body to injury and are associated with postoperative complications such as postoperative delirium (POD) and perioperative neurocognitive disorder (PND) in human and animals [10–13].

In this context, deep NMB has the potential to reduce systemic cytokine release by minimizing intraoperative tissue damage and stretching, suggesting an immunomodulatory advantage beyond its surgical benefit [14,15]. However, whether the depth of NMB differentially influences systemic cytokine release remains unclear. Therefore, in this meta-analysis, we evaluated the immunomodulatory impact of deep NMB versus that of moderate NMB on the systemic release of cytokines that have hypothetically been associated with POD and PND [10,11].

Materials and Methods

Results

Study selection

Through our systematic search of the databases, we identified 956 potentially relevant articles. After removing 590 duplicates, we screened 366 articles. Of the 366 articles screened, we excluded 362 studies because they were not focused on the study after title review (n = 173) and abstract review (n = 166), and because 23 studies were excluded after full-text review due to the absence of laboratory analysis (n = 20) or cytokine analysis (n = 3). We included an additional four RCTs via supplementary searches on relevant websites. Finally, eight RCTs were included in this meta-analysis. Our rigorous selection methodology is shown in Fig. 1.

Fig. 1.

Flow diagram of the recruitment of study participants.

Characteristics of included studies

Table 1 shows the characteristics of the RCTs included in this review [21–28]. Among all included RCTs, seven studies were on laparoscopic surgery and one study was on total hip replacement in patients with hip fracture. One study was conducted among only female patients and another among only male patients. Four studies were conducted with TIVA, three studies used inhalation anesthesia, and one study did not specify the type of anesthesia. The sampling time points and type of cytokines included in each study are described in Supplementary Table 2.

Table 1.

Detailed Characteristics of the Included RCTs

Serum cytokine levels immediately after surgery

IL-1β

The meta-analysis revealed no statistically significant difference in terms of serum IL-1β levels (pg/ml) between the patients in the deep and moderate NMB groups immediately postoperatively (SMD: −0.38, 95% CI [−1.01 to 0.26], P = 0.24) [21,23–25]. The heterogeneity among studies was high, with an I² value of 87% (τ² = 0.3674, P < 0.01) (Fig. 2A). Meta-regression analysis indicated that inhalational anesthesia was significantly associated with high IL-1β levels (QM = 11.4536, P = 0.0007), accounting for 89.36% of the observed heterogeneity (Supplementary Figs. 1A–F). No significant deviations were observed in the sensitivity analysis, and the funnel plot revealed no evidence of publication bias (Supplementary Figs. 1G and H).

Fig. 2.

Meta-analysis of serum cytokine levels immediately postoperatively. Forest plots showing the SMDs in serum cytokine levels immediately postoperatively in the deep and moderate NMB groups. (A) IL-1β (pg/ml), (B) IL-6 (pg/ml; ng/ml in the study by Helwa et al. [25]), and (C) TNF-α (pg/ml; ng/ml in the study by Helwa et al. [25]). SMD: standardized mean difference, NMB: neuromuscular blockade, IL: interleukin, TNF-α: tumor necrosis factor-alpha.

IL-6

Regarding IL-6, no statistically significant differences were observed in their serum levels (pg/ml in most studies; ng/ml in the study by Helwa et al. [25]) immediately postoperatively between the deep and moderate NMB groups (SMD: −0.54, 95% CI [−1.16 to 0.07], P = 0.08) [21,22,24,25]. However, heterogeneity among studies was high, with an I² value of 81% (τ² = 0.3271, P < 0.01) (Fig. 2B). Meta-regression analysis revealed that the number of patients included in each study was significantly associated with reduced IL-6 levels (estimate = −0.0342, 95% CI [−0.0541 to −0.0142], P = 0.0008), explaining 95.36% of the heterogeneity (Supplementary Figs. 2A–F). Sensitivity analysis revealed that when the study by Kim et al. [22] was excluded, the recalculated SMD did not include zero, suggesting a potential influence of this study on the overall effect size (Supplementary Figs. 2G and H).

TNF-α

Regarding TNF-α levels (pg/ml in most studies; ng/ml in the study by Helwa et al. [25]), a statistically significant reduction was observed in their levels in the deep NMB group compared with their levels in the moderate NMB group (SMD: −0.46, 95% CI [−0.87 to −0.06], P = 0.03) [23–25]. The heterogeneity was moderate, with an I² value of 61% (τ² = 0.0785, P = 0.08) (Fig. 2C). Meta-regression analysis demonstrated that inhalational anesthesia was significantly associated with high TNF-α levels immediately postoperatively (estimate = 0.6271, 95% CI [0.0 544–1.1 997], P = 0.032), and this accounted for 100% of the observed heterogeneity (Supplementary Figs. 3A–F). The sensitivity analysis confirmed the robustness of the findings for all the included studies, and the funnel plot did not reveal any significant publication bias (Supplementary Figs. 3G and H).

Substantial heterogeneity was observed with regard to several cytokine related outcomes, particularly regarding IL-6 and IL-1β (I² > 80%). To further explore the potential sources of heterogeneity, we conducted subgroup analyses stratified by anesthetic technique and reversal strategy (Supplementary Fig. 4). Regarding IL-1β, a larger difference between the deep and moderate NMB was observed in the inhalational anesthesia subgroup (SMD: −1.28, 95% CI [−1.76 to −0.80]) than in the TIVA subgroup (SMD: −0.07, 95% CI [−0.41 to 0.27], I² = 44%) (Supplementary Fig. 4A). Regarding IL-6, the TIVA subgroup demonstrated a statistically significant pooled SMD favoring deep over moderate NMB with low between-study heterogeneity (SMD: −0.76, 95% CI [−1.09 to −0.42], I² = 0%), whereas the inhalational subgroup retained high heterogeneity (SMD: −0.30, 95% CI [−1.72 to 1.12], I² = 93%) without statistical significance (Supplementary Fig. 4C). Regarding TNF-α, the inhalational anesthesia subgroup showed a larger difference (SMD: −0.89, 95% CI [−1.34 to −0.43]) than the TIVA subgroup (SMD: −0.26, 95% CI [−0.56 to 0.04], I² = 0%) (Supplementary Fig. 4E). Regarding subgroup analyses by reversal strategy (Supplementary Figs. 4B, D and F), comparison of (i) trials in which sugammadex was used in both deep and moderate groups with (ii) trials in which sugammadex was used for deep and neostigmine for moderate did not reveal any consistent differences for IL-1β, IL-6, or TNF-α levels, and substantial heterogeneity was observed in several subgroups.

In summary, only TNF-α levels showed a statistically significant difference immediately postoperatively, while IL-6 levels exhibited a noticeable decreasing tendency in the deep NMB group compared with the moderate NMB group, although the difference did not reach statistical significance.

Serum cytokine levels at 24 and 48 hours after surgery

Cytokines that were examined 24-h postoperatively included IL-1β (pg/ml), IL-6 (pg/ml), TNF-α (pg/ml), and C-reactive protein (CRP) (mg/dl). Plasma cytokine tests 24-h postoperatively showed no statistically significant difference in terms of the degree of NMB for IL-1β (SMD: −0.26, 95% CI [−0.60 to 0.08], P = 0.13) [21,24,26,27], IL-6 (SMD: −0.03, 95% CI [−0.21 to 0.15], P = 0.73) [19,20,22,24–26], TNF-α (SMD: −0.11, 95% CI [−0.35 to 0.14], P = 0.38) [22,25,26], and CRP (SMD: −0.07, 95% CI [−0.34 to 0.21], P = 0.64) [20,22,26]. IL-1β showed moderate heterogeneity among the included studies (I² = 60%, τ² = 0.0717, P = 0.06), while IL-6 (I² = 0%, τ² = 0, P = 0.92), TNF-α (I² = 0%, τ² = 0, P = 0.62), and CRP (I² = 0%, τ² = 0, P = 0.61) showed no heterogeneity among the studies (Figs. 3A–D). Meta-regression analysis revealed that IL-1β levels 24-h postoperatively were significantly influenced by patient age (QM = 7.19, P = 0.007), accounting for 100% of the observed heterogeneity (R² = 100%). Specifically, younger age was associated with a greater release of IL-1β in the moderate NMB group than in the deep NMB group (estimate = 0.0242, 95% CI [0.0 065–0.0 419], P < 0.01) (Supplementary Fig. 5).

Fig. 3.

Meta-analysis of serum cytokine levels 24-h and 48-h postoperatively. Forest plots illustrating the SMDs in serum cytokine levels at different postoperative time points in the deep and moderate NMB groups. (A) IL-1β levels at 24 h (pg/ml), (B) IL-6 levels at 24 h (pg/ml), (C) TNF-α levels at 24 h (pg/ml), (D) CRP levels at 24 h (mg/dl), and (E) CRP levels at 48 h (mg/dl). SMD: standardized mean difference, NMB: neuromuscular blockade, IL: interleukin, TNF-α: tumor necrosis factor-alpha, CRP: C-reactive protein.

Regarding CRP, postoperative 48-h tests were conducted in the studies included in this analysis (Fig. 3E), and they revealed no statistically significant difference between the deep and moderate NMB groups (SMD: 0.15, 95% CI [−0.12 to 0.42], P = 0.27); furthermore, no heterogeneity was observed among the studies (I² = 0%, τ² = 0, P = 0.70) [22–24].

RoB

The RoB in all included studies was evaluated using the RoB 2 tool (https://methods.cochrane.org/risk-bias-2). Four studies were classified as having a ‘low’ RoB [22,24,25,28], while the other four studies were rated as having ‘some concerns’ [21,23,26,27]. Specifically, the study by Reijnders‐Boerboom et al. [27] exhibited baseline age differences between groups that raised concerns about the comparability of the randomized groups. Additionally, Kim et al. [21], Koo et al. [23], and Li et al. [26] did not provide sufficient information on whether the allocation sequence was adequately concealed, leading to ‘some concerns’ about the randomization process (Supplementary Fig. 6).

Discussion

Our meta-analysis suggests that deep NMB may be associated with lower TNF-α levels than those observed with moderate NMB immediately postoperatively. Other cytokine levels, such as those of IL-1β, IL-6, and CRP, did not show any statistically significant difference between the groups at any postoperative time point (immediately, 24-h, and 48-h postoperatively). Inhalational anesthesia was observed to be a factor related to increased IL-1β and IL-6 levels immediately postoperatively; furthermore, at 24-h postoperatively, increased IL-1β levels were related to younger age.

Surgical tissue injury activates the body’s immune system, inducing an inflammatory response. In this process, damage-associated molecular patterns released from injured cells stimulate immune cells through pattern recognition receptors [7]. Consequently, immune cells such as macrophages and neutrophils become activated and secrete pro-inflammatory cytokines, including TNF-α, IL-1β, and IL-6 [7,29]. While the amplified inflammatory response mediated by these cytokines promotes healing at the injury site, excessive release can lead to systemic inflammatory responses, affect central nervous system (CNS) function, and potentially cause postoperative complications such as POD and PND.

Postoperative cytokines are known to reflect the degree of injury and are increasingly recognized as biomarkers for predicting POD and PND [11,30,31]. Pro-inflammatory cytokines, such as IL-1β, IL-6, and TNF-α, have been shown to affect the CNS function directly or indirectly by disrupting the blood–brain barrier (BBB), allowing neurotoxic substances and systemic inflammatory mediators to reach the CNS [32]. In Alzheimer’s disease rat models, elevated TNF-α levels were associated with altered synaptic properties, such as reduced inhibitory GABAergic neurotransmission and suppressed long-term potentiation [33,34].

While the present study identified a significant reduction in TNF-α levels immediately postoperatively in the deep NMB group compared with those in the moderate NMB group, no statistically significant differences were identified for other cytokines such as IL-1β, IL-6, or CRP. Although the pooled effect sizes for these cytokines were generally lower in the deep NMB group, these differences did not reach statistical significance and should be interpreted with caution due to the low certainty of evidence and substantial heterogeneity among the included studies.

This lack of statistical significance does not necessarily indicate the absence of a biological effect but may instead reflect methodological limitations across the included studies, particularly regarding the timing of cytokine sampling and between-study heterogeneity. The lack of significant differences observed in cytokine levels could be attributed to the timing of sample collection in the included studies, as most analyses focused on immediate and 24-h postoperative periods. Existing literature indicates that cytokine release following surgical insult demonstrates distinct peak times: TNF-α levels peak earlier, within 2–4 hours postoperatively, and IL-6 and IL-1β reach their peak levels within 6 hours postoperatively (Fig. 4). Meanwhile, CRP levels tend to peak later, around 24–48 hours [35–38]. These patterns highlight the importance of capturing samples during these critical windows to fully evaluate the inflammatory response.

Fig. 4.

Schematic timeline of systemic cytokine release postoperatively.

Among the studies included in this meta-analysis, the RCT conducted by Oh et al. [24] reported a peak in IL-6 levels at 6 h postoperatively, with a statistically significant reduction observed in the deep NMB group compared with the moderate NMB group. This finding highlights the possibility that the lack of significant differences in our pooled results may be attributed to suboptimal timing of cytokine sampling in other included studies.

Notably, IL-6, a cytokine closely linked to BBB disruption, may exhibit its strongest differences during these earlier postoperative periods [13]. Although, Oh et al. [24] reported no statistically significant increase in the incidence of POD in their moderate NMB group compared to their deep NMB group, further investigations are needed. The study’s sample size calculation was based on a 30% increase in IL-6 levels 4 hours postoperatively that may have limited its statistical power to identify smaller but clinically meaningful differences in POD incidence. Additionally, a notable trend was observed, with a higher incidence of POD in the moderate NMB group compared with that in the deep NMB group (moderate NMB: 34.15% vs. deep NMB: 17.07%, P = 0.129). This underscores the importance of future studies with larger sample sizes and standardized sampling intervals focusing on well-documented peak cytokine release times that could provide more definitive evidence on whether NMB depth differentially influences cytokine release and its potential impact on BBB integrity and postoperative neurocognitive outcomes.

The heterogeneity observed in cytokine levels can, in part, be attributed to factors such as the type of anesthesia and patient age. Meta-regression analyses revealed that inhalational anesthesia was associated with higher levels of IL-1β and TNF-α immediately postoperatively than was TIVA, suggesting that volatile anesthetics may exaggerate immune responses. Previous studies suggest that, compared with TIVA, inhalational anesthetics exaggerate the immune responses during general anesthesia. For example, isoflurane and halothane have been associated with increased systemic inflammatory responses compared with propofol, potentially owing to their effects on immune cell activity and cytokine production [39,40]. In an RCT of patients undergoing craniotomy, the patients in the TIVA group demonstrated higher levels of anti-inflammatory cytokines, such as IL-10, and a lower IL-6/IL-10 ratio than the patients in the sevoflurane group [41]. Furthermore, a mouse study demonstrated that isoflurane, without surgical stimulation, induced an increase in the levels of pro-inflammatory cytokines such as IL-1β, IL-6, and TNF-α in brain tissues, suggesting that inhalational anesthetics may independently contribute to cytokine elevation [42]. In this meta-analysis, compared with TIVA, inhalational anesthesia was associated with elevated IL-1β and TNF-α levels immediately postoperatively, suggesting that the choice of anesthetic technique can influence postoperative inflammatory profiles. These findings highlight the need for further research to optimize anesthesia strategies to minimize systemic inflammation and its potential complications.

Additionally, younger patients showed a tendency for increased cytokine release, potentially reflecting a more robust immune reaction and greater sensitivity to damage-associated molecular pattern-mediated pathways. Age-related differences in immune response are less extensively studied, particularly in the context of major surgical stressors. However, existing evidence suggests that older patients tend to exhibit a more variable and generally attenuated inflammatory response than do the younger individuals [43,44]. In this study, younger age was significantly associated with increased IL-1β levels at 24 hours postoperatively under moderate NMB, which aligns with previous findings suggesting that age-related immunosenescence may dampen the inflammatory response in older adults. This observation underscores the importance of considering patient age when evaluating cytokine-mediated outcomes in the perioperative setting. Tailoring anesthetic and surgical strategies to account for age-related differences in immune function could help optimize postoperative outcomes owing to the more robust immune response of younger patients.

A previous study by Kim et al. [22] may explain the high heterogeneity observed in the IL-6 levels immediately postoperatively; in their study, IL-6 concentrations were paradoxically higher in the deep NMB group than in the moderate NMB group. This finding contrasts with the general trend observed in other included studies, wherein deep NMB tended to attenuate cytokine responses. Notably, our sensitivity analyses demonstrated that exclusion of the study by Kim et al. [22] substantially reduced the heterogeneity (I² = 0%) and yielded a significant pooled estimate favoring deep NMB. This suggests that the dataset of the study by Kim et al. [22] may have disproportionately contributed to between-study variability. Furthermore, considering that the sample size in their study was smaller than that in most other trials, it is possible that their limited sample size amplified the study’s weight in the meta-regression model, contributing to the heterogeneity observed regarding IL-6 outcomes.

Despite our meta-regression efforts, high statistical heterogeneity pertaining to several outcomes remained, including IL-6 and IL-1β outcomes. To further explore potential sources of heterogeneity, we conducted subgroup analyses stratified by anesthetic technique and reversal strategy (Supplementary Fig. 4). These analyses showed that IL-1β and TNF-α levels exhibited larger differences between the deep and moderate NMB groups in the inhalational anesthesia subgroup compared with those in the TIVA subgroup, whereas IL-6 levels demonstrated a significant reduction in the deep NMB under TIVA group, with low heterogeneity; however, high heterogeneity was observed in the inhalational subgroup. By contrast, subgrouping by reversal strategy did not reveal consistent differences across cytokines. These findings suggest that anesthetic technique may partially contribute to the observed variability, while the role of reversal agents remains questionable. Nevertheless, substantial residual heterogeneity remained, highlighting the need for cautious interpretation of the pooled estimates.

This study had some limitations. First, the number of included studies was relatively small. Second, the timing of cytokine measurement varied among the included studies and often did not capture the peak inflammatory window (typically 2–6 hours postoperatively). This may have led to underestimation or misrepresentation of cytokine dynamics. Future trials should standardize sampling time points to better capture perioperative inflammatory peaks, ideally within 2 to 6 hours following surgical stimulus. Third, there was considerable variability pertaining to the cytokines that were measured, surgical factors, and anesthetic protocols, such as anesthetic method, NMB definition, and reversal strategy, among the included studies that likely contributed to the observed heterogeneity. Though we performed meta-regression, subgroup analyses, and sensitivity analyses to address this issue, the observed variability could not be accounted for. Hence, the pooled estimates of this meta-analysis should be interpreted with caution. Fourth, although visual inspection of funnel plots and Egger’s regression tests did not suggest any major asymmetry, the limited number of included studies per cytokine outcome substantially reduces the reliability of these assessments. Therefore, the possibility of small-study effects or selective publication bias cannot be excluded and should be considered when interpreting the pooled findings. Fifth, the analysis was limited to short-term cytokine changes, with no consideration of long-term immune or neurological outcomes. Finally, some studies lacked detailed information on randomization and allocation concealment, leading to potential biases.

To summarize, our meta-analysis suggests that deep NMB may reduce systemic cytokine levels, particularly those of TNF-α, in the immediate postoperative period, whereas no statistically significant differences were observed for other cytokine levels including those of IL-1β, IL-6, and CRP. These findings tentatively suggest that deep NMB could attenuate surgical tissue damage and modulate systemic cytokine release. Fig. 5 illustrates a conceptual pathway that links NMB to reduced tissue injury and downstream attenuation of systemic inflammation. Although CNS complications such as POD and PND were not directly assessed in the included studies, previous evidence suggests that systemic inflammation can induce neuroinflammatory responses and contribute to the disruption of BBB integrity [10–13]. Taken together, these findings are indicative of an association between cytokine modulation and CNS complications; however, this remains speculative and requires further clinical and mechanistic investigation (Fig. 5).

Fig. 5.

Hypothetical schematic mechanism of immunomodulatory effects of NMB. Solid arrows represent evidence-based relationships supported by included trials, whereas dotted arrows indicate hypothetical pathways linking cytokine modulation to CNS outcomes such as POD and PND that were not directly assessed in this meta-analysis. NMB: neuromuscular blockade, CNS: central nervous system, POD: postoperative delirium, PND: perioperative neurocognitive disorder, DAMPs: damage-associated molecular patterns.

In conclusion, these findings suggest that deep NMB may exert protective immunomodulatory effects by attenuating systemic cytokine response, particularly that associated with POD and PND. However, whether such effects extend to CNS function remains undetermined. Notably, a recent large clinical trial reported that NMB might increase the risk of POD in a dose-dependent manner, with reversal agents mitigating this risk [45]. Taken together, these contrasting perspectives highlight that the relationship between NMB and CNS outcomes remains complex and multifactorial, requiring cautious interpretation. Future research should incorporate both inflammatory biomarkers and patient-centered clinical outcomes to elucidate the mechanistic pathways and clarify the clinical implications of NMB depth in perioperative care.

Supplementary Materials

Article information Continued