Propofol-based versus volatile anesthesia and postoperative immune response in prostate cancer

Introduction

Prostate cancer remains one of the most common malignancies among men worldwide, with a persistently rising disease burden. According to the latest estimates from the International Agency for Research on Cancer (IARC) GLOBOCAN 2022 database, approximately 1,467,854 new cases and 397,430 deaths occurred globally in 2022, positioning prostate cancer as a leading cause of cancer-related mortality (1-3). Recent analyses spanning 185 countries and regions indicate significant geographic disparities in prostate cancer incidence and mortality, with rising trends observed in parts of Asia, Africa, and Latin America. This underscores enduring challenges in screening accessibility, diagnostic and therapeutic equity, and comprehensive perioperative management (4). In the United States, the American Cancer Society’s 2025 report projects 313,780 new cases and 35,770 deaths, highlighting concerns such as increasing rates of advanced-stage disease that warrant continued attention (5). Radical prostatectomy remains a cornerstone treatment for patients with localized or select locally advanced disease. However, the conventional focus on surgical resection alone often fails to account for the variability in early postoperative complication risks and individual fluctuations in immune-inflammatory status (6). Growing evidence suggests that the perioperative period, particularly the immediate postoperative phase, constitutes a critical window during which the host’s immune and inflammatory responses are substantially perturbed. These alterations are not only relevant to infection risks and recovery trajectories but may also influence subsequent treatment tolerance and the tumor biological microenvironment (7). Consequently, quantifying immune responses during this period holds significant clinical value.

From mechanistic and clinically measurable perspectives, surgical trauma and stress responses activate the sympathetic-endocrine axis, promoting the release of pro-inflammatory mediators and inducing a characteristic phenotype of adaptive immune suppression. This manifests as peripheral lymphopenia and elevated inflammatory-derived indices (8,9). Multiple studies have demonstrated associations between postoperative lymphopenia and unfavorable outcomes. For instance, following curative gastrectomy for gastric cancer, surgery-induced reductions in absolute lymphocyte counts (ALCs) correlate with recurrence-free and overall survival (10,11). Similarly, in other oncologic and surgical contexts, significant treatment-related or postoperative lymphopenia has been linked to adverse disease outcomes (12). Concurrently, the neutrophil-to-lymphocyte ratio (NLR), derived from routine complete blood counts and characterized by high reproducibility, has gained widespread use as a marker depicting the imbalance between innate immune activation and adaptive immune suppression (13). Recent reviews and clinical studies indicate consistent associations between NLR and perioperative complications as well as oncologic outcomes, establishing it as a practical proxy for perioperative immune-inflammatory responses (14). Therefore, focusing on postoperative day 1 (POD1) as a key timepoint, in conjunction with conventional metrics such as lymphocyte counts, NLR, and C-reactive protein (CRP), offers a sensitive approach to capture the magnitude and direction of early postoperative immune suppression and inflammatory amplification. This provides an operable quantitative endpoint for research on perioperative risk stratification and intervention strategies (15).

Against this backdrop, anesthetic maintenance technique, as one of the most significant and modifiable perioperative exposures, has been increasingly scrutinized for its potential impact on immune responses and tumor-related biological processes. Comprehensive reviews note the superposition of multiple immunosuppressive factors during the perioperative period and emphasize the need for their systematic evaluation and optimization (16). Regarding anesthetic agents, theoretical and experimental research suggests that propofol-based total intravenous anesthesia (TIVA) may exhibit a more immunologically favorable profile, potentially by attenuating stress responses and inflammatory amplification while relatively preserving functions associated with natural killer (NK) cells and cytotoxic T cells (17). In contrast, volatile anesthetics have demonstrated inconsistent immunomodulatory effects across studies (18). Clinically, recent systematic reviews and meta-analyses suggest that propofol-based anesthesia may be associated with improved survival or recurrence outcomes for certain cancer types (19,20). Conversely, large-scale studies and randomized trials also report potentially limited differences in long-term oncologic outcomes between strategies, underscoring the need for research designs focused on more homogeneous tumor populations and clearer mechanistic or intermediate endpoints to clarify these associations (21,22). More importantly, at the level of immune-inflammatory intermediate phenotypes, studies directly comparing different anesthetic strategies have reported effects on indices such as NLR, indicating that anesthetic agents can indeed influence intraoperative and postoperative immune-inflammatory readouts, albeit with significant heterogeneity across surgical types and patient populations (23,24). For the relatively standardized urological context of radical prostatectomy, retrospective studies have also reported suggestive differences between propofol and volatile anesthesia from an oncologic outcome perspective (25,26). Nevertheless, real-world studies employing rigorous control for clinical confounders and focusing on perioperative immune responses as a core endpoint are needed to strengthen the evidence chain.

To address this evidence gap and its potential clinical translatability, this study compares propofol-based TIVA with volatile anesthesia regarding early postoperative immune responses in patients undergoing elective radical robot-assisted prostatectomy for prostate cancer. Employing a retrospective cohort design, propensity score (PS) methods are utilized to balance baseline differences between groups where possible. The primary outcome is the absolute peripheral blood lymphocyte count on POD1, supplemented by secondary outcomes including neutrophil count, NLR, and CRP levels. Where repeated-measures data are available, the temporal dynamics of immune responses are further characterized. By focusing on perioperative immune-inflammatory intermediate phenotypes—endpoints closer to underlying mechanisms—this study aims to provide testable clinical evidence for optimizing perioperative anesthetic maintenance strategies and to establish a foundation for subsequent prospective research and higher-level evidence generation. We present this article in accordance with the STROBE reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2026-1-0096/rc).

Methods

Patient population

This retrospective cohort study utilized existing clinical data. The study population comprised patients diagnosed with prostate cancer who underwent elective radical robot-assisted prostatectomy at the Affiliated Drum Tower Hospital of Nanjing University Medical School between January 1, 2023 and December 31, 2025. Patients were identified through a retrospective search of the hospital’s electronic medical record system, anesthesia information management system, and laboratory information system. The initial search identified 215 patients who underwent radical robot-assisted prostatectomy with a confirmed pathological diagnosis of prostate cancer. Following verification of data completeness and relevance for the study, 94 patients were excluded due to comorbid immune system disorders or active infections, preoperative immunosuppressive therapy, missing key perioperative or laboratory data, or inability to definitively classify the anesthetic maintenance technique. Consequently, 121 patients were included in the final analysis.

Patients were categorized into two groups based on the primary maintenance phase drug delivery method. All anesthetic techniques represented standard clinical management strategies. All anesthetic techniques represented standard clinical management strategies implemented during routine practice. These procedures were performed by a consistent group of six surgeons and 14 anesthesiologists at our center. However, the choice of maintenance technique was at the discretion of the attending clinicians based on their professional preference and routine practice; this introduces potential provider-level confounding, as certain clinicians may have preferentially utilized one technique over the other. The study performed retrospective grouping based on the actual anesthetic maintenance technique without intervening in any perioperative anesthetic or surgical processes.

The follow-up period was defined as starting from the day of surgery (initiation of anesthesia) until the completion of predetermined monitoring for immune-related indicators, primarily covering the perioperative and early postoperative hospitalization phases. All study outcomes were derived from routine clinical monitoring and laboratory test data obtained during the hospitalization. No additional follow-up timepoints were established, and no long-term follow-up beyond discharge was conducted. The study objective was limited to assessing the association between different anesthetic maintenance techniques and postoperative immune responses and did not involve analysis of long-term oncologic outcomes.

This study was conducted entirely using clinical and laboratory data generated during prior routine diagnostic and therapeutic processes, without implementing any additional interventions or altering original clinical decisions, consistent with the fundamental characteristics of a retrospective observational study.

Ethics and informed consent

As a retrospective observational study based on pre-existing clinical data, all information utilized originated from patient medical records, anesthesia records, and laboratory reports generated during standard care. The research process involved no additional tests, interventions, or changes to the original treatment plans. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of the Affiliated Drum Tower Hospital of Nanjing University Medical School. The requirement for informed consent was waived because of the retrospective nature of the study and the use of anonymized patient data.

During data collection and analysis, researchers strictly followed relevant ethical guidelines and data protection requirements. All research data were de-identified upon extraction, containing no directly or indirectly identifiable patient information. Access to and use of the data were restricted to study team members within a controlled environment and solely for the predefined research purposes, thereby safeguarding patient privacy and data security.

Inclusion and exclusion criteria

Inclusion criteria

Patients meeting all the following criteria were included in the analysis:

• Age ≥18 years, regardless of sex.
• Pathologically confirmed diagnosis of prostate cancer via preoperative biopsy or postoperative specimen examination.
• Undergoing elective radical robot-assisted prostatectomy (including open, laparoscopic, or robot-assisted approaches) between January 1, 2023 and December 31, 2025.
• Receiving general anesthesia with complete anesthesia records, allowing clear classification of the maintenance technique as either propofol-based TIVA or volatile anesthesia based on the primary maintenance agent.
• Perioperative anesthesia and surgical management following the hospital’s standard clinical pathway, without the use of experimental or non-standard anesthetic protocols.
• Availability of complete baseline perioperative clinical data, including demographic information and perioperative anesthetic and surgical details.
• Availability of immune-related laboratory test results for at least one predefined postoperative timepoint (e.g., POD1) along with preoperative baseline values, obtained through the hospital’s routine clinical testing.

Exclusion criteria

Patients meeting any of the following criteria were excluded:

• Preoperative or perioperative active infection, sepsis, or other definitive acute inflammatory conditions likely to significantly influence immune or inflammatory markers.
• History of or recent (e.g., within 3 months) preoperative immunosuppressive therapy, including but not limited to long-term systemic corticosteroids, chemotherapy, immunotherapy, or other immunomodulatory drugs.
• Comorbid diagnosed autoimmune disorders, hematologic diseases, or severe immunodeficiency states.
• Undergoing emergency surgery, unplanned surgery, or requiring urgent conversion of the surgical plan due to severe perioperative complications.
• Significant hepatic or renal dysfunction potentially affecting inflammatory mediator metabolism or interpretation of immune markers.
• Occurrence of severe perioperative complications (e.g., major hemorrhage, shock) necessitating extraordinary rescue measures likely to substantially confound postoperative immune responses.
• Frequent switching between intravenous and inhaled anesthetic agents during the maintenance phase, or inability to definitively classify the primary anesthetic maintenance technique based on anesthesia records.
• Missing key clinical information, anesthesia records, or primary outcome data that could not be reasonably supplemented from original records or information systems.

Study procedures

Definition and grouping of anesthetic technique

Anesthetic technique was defined based on complete documentation of medication use during the induction, maintenance, and emergence phases within perioperative anesthesia records and the anesthesia information management system. All included patients received general anesthesia under standard monitoring, including continuous electrocardiography, non-invasive or invasive blood pressure monitoring, pulse oximetry, and end-tidal carbon dioxide monitoring. Induction regimens, which frequently involved propofol in both study groups, were selected by the attending anesthesiologist based on individual patient characteristics; however, these were not included in the formal exposure definition or used as grouping criteria, focusing instead on the maintenance phase.

In the propofol-based TIVA group, following induction, anesthesia was maintained primarily via continuous intravenous infusion of propofol. The mode of administration, infusion rate, and dose adjustments were individualized by the anesthesiologist based on patient vital signs and clinical response. Short-acting opioids (e.g., remifentanil or sufentanil) could be co-administered for analgesia, and neuromuscular blocking agents were used as needed to maintain optimal surgical conditions. Inhaled anesthetic agents were not routinely used during maintenance. If volatile agents were briefly used due to specific clinical circumstances but propofol remained the principal maintenance agent, the case was still classified into the TIVA group.

In the volatile anesthesia group, following induction, anesthesia was maintained primarily using inhaled anesthetic agents (e.g., sevoflurane or desflurane) delivered continuously via the anesthesia circuit to maintain adequate depth. The concentration of inhaled agents was dynamically adjusted by the anesthesiologist according to patient vital signs and surgical stimulation. This group could also receive intravenous opioids for analgesia and neuromuscular blockers as required, but inhaled agents consistently served as the primary means for maintaining anesthetic depth. If intravenous sedatives were briefly co-administered during maintenance but inhaled agents retained the primary role, the case was still classified into the volatile anesthesia group.

For cases involving concurrent use of intravenous and inhaled agents during maintenance, researchers comprehensively evaluated the anesthesia records. Group assignment was based on the drug type serving as the primary maintenance modality. This classification reflects real-world clinical practice but acknowledges the potential for crossover, such as the brief use of volatile agents in the TIVA group or intravenous sedatives during the maintenance phase in the volatile group. Cases where the maintenance technique involved frequent switching throughout the procedure or where the primary strategy could not be unambiguously determined from records were excluded from the primary analysis to minimize bias from exposure misclassification.

Perioperative management and data collection

All included patients received perioperative anesthesia and surgical management according to the standard clinical pathway. The research team did not intervene in anesthetic protocol selection or routine management. Data on perioperative adjuncts—specifically non-steroidal anti-inflammatory drugs (NSAIDs) such as ketorolac, ketamine, and dexmedetomidine—as well as intraoperative fluid administration volumes, were systematically extracted from medical records. Although these interventions were managed by attending teams according to specific surgeon- or anesthesiologist-dependent protocols, they were included in the analysis because they may independently modulate perioperative inflammatory marker levels. The study involved retrospective collation and analysis of pre-existing clinical records. To minimize information bias from variations in clinical practice, key perioperative process nodes—including surgical approach (open, laparoscopic, or robot-assisted) and elective status—were uniformly defined during data extraction. The timeline of events recorded in the anesthesia record on the day of surgery served as the primary temporal reference.

Data collection relied on routinely generated records from hospital information systems. Two researchers independently performed retrospective extraction and verification using a predefined data dictionary and standardized extraction forms. Data sources included the electronic medical record system (for demographic information, admission diagnosis, comorbidities, preoperative assessment, and perioperative progress notes), the anesthesia information management system (for anesthesia induction and maintenance medication records, anesthesia timeline summaries, intraoperative vital sign summaries, and anesthesia-related interventions), the surgical anesthesia billing and operative report system (for verifying surgical approach, start/end times, and operative notes), and the laboratory information system (for verifying test timing and completeness of laboratory result entries). The extraction workflow prioritized the sequence: “case identification → anesthetic technique determination → extraction of key perioperative variables → data consistency verification”. This approach involved confirming inclusion/exclusion status and group assignment before systematic extraction of other covariates to reduce redundant work or misentry due to uncertain grouping.

To ensure data quality, a source priority hierarchy combined with cross-verification rules addressed inconsistencies across systems. Conflicting entries for the same variable were resolved by prioritizing the source with stronger primary record attributes. For example, anesthesia medications and timeline took precedence from the anesthesia information management system; surgical approach and times from the formal operative report or corresponding anesthesia record timeline; and baseline comorbidity information from admission history and discharge diagnoses. For ambiguous cases, both researchers jointly reviewed original chart pages and anesthesia record entries. If unresolved, a third researcher (or the principal investigator) made a final determination based on predefined rules, maintaining an audit trail. Missing data were handled per predefined protocols. Cases missing key anesthetic maintenance information essential for group assignment, or core perioperative information that could not be reasonably supplemented, were excluded during the screening phase. For non-critical covariates, missing status was accurately annotated in the dataset, and the proportion of missingness was retained for reporting in the analysis phase.

All extracted data were stored in de-identified form. Researchers retained only minimal linkage information necessary for data verification, which was separated from the analysis dataset upon completion. Data access was permission-controlled, limited to study team members within a secured environment to protect patient privacy and data security. These data collection and quality control procedures were implemented using routinely generated records from prior clinical care, consistent with the data sources and operational characteristics of a retrospective observational study.

Outcome measures

Primary outcome measure

The primary outcome was the degree of change in early postoperative immune response, reflected by perioperative peripheral blood cellular immune laboratory markers. Aligned with the retrospective design, all measures were derived from routine clinical test results during hospitalization, with no study-specific assays introduced.

The ALC was used to reflect the overall status of adaptive immune function. It is among the most commonly used and stable indicators in perioperative immunosuppression research. In this study, lymphocyte counts were extracted directly from routine complete blood count results in the hospital’s laboratory information system, expressed in units of ×10⁹/L. The primary endpoint was defined as the delta (Δ) change in ALC, calculated as the POD1 value minus the preoperative baseline value, to more accurately reflect the biological magnitude of the anesthesia-related immune response relative to individual baselines. A decrease in lymphocyte count is typically viewed as indicative of surgical stress and anesthesia-related immune suppression, whereas maintenance of a higher count or a smaller decline suggests relatively milder suppression. No arbitrary thresholds were applied to categorize lymphocyte counts; they were analyzed as continuous variables to avoid information loss and bias from arbitrary categorization.

Secondary outcome measures

To supplement the assessment of postoperative immune and inflammatory responses from multiple dimensions, several perioperative routine laboratory markers were selected as secondary outcomes, capturing aspects of innate immunity, systemic inflammation, and immune-inflammatory balance.

Absolute neutrophil count (ANC)

The neutrophil count reflects the activity level of the innate immune system and acute inflammatory responses. This measure was also sourced from routine complete blood counts, expressed in ×10⁹/L. Postoperative neutrophil elevation typically correlates with surgical trauma, tissue injury, and heightened stress responses. In analyses, neutrophil counts described baseline inflammatory response characteristics under different anesthetic techniques rather than serving as a sole core indicator of immune suppression.

NLR

NLR is a composite marker of inflammation and immune balance, simultaneously reflecting the relative states of innate immune activation and adaptive immune suppression. In this study, NLR was calculated by dividing the ANC by the ALC from the same timepoint, with no additional standardization. Higher NLR values generally indicate enhanced systemic inflammation and greater immune suppression. Given variations in reference ranges across studies and populations, no fixed cut-off points were used for categorization. NLR was analyzed as a continuous variable, interpreted in conjunction with its perioperative dynamic trends.

White blood cell count (WBC)

The total WBC, a traditional inflammatory marker, reflects the overall level of postoperative inflammatory response. Sourced from routine complete blood counts (×10⁹/L), its specificity is relatively low due to influence by multiple factors. Therefore, it served primarily as a contextual descriptor of overall inflammatory response in this study, not for standalone immune function assessment.

CRP

CRP, an acute-phase reactant, is an important biomarker for systemic inflammatory response intensity. CRP values were obtained from routine biochemistry tests, with units as reported by the laboratory. Elevated CRP levels typically correlate with tissue injury, surgical trauma, and enhanced postoperative inflammation. In this study, CRP primarily assisted in evaluating differences in postoperative inflammatory response intensity between anesthetic maintenance techniques, providing supportive evidence rather than forming the primary basis for conclusions.

Assessment timepoints and data handling principles

Assessment timepoints for perioperative immune-related laboratory markers were defined based on the physiological characteristics of surgical stress response and anesthesia-related immune changes. Preoperative baseline was defined as the most recent laboratory test result within 24 hours before surgery, reflecting the patient’s baseline immune and inflammatory state prior to anesthetic and surgical intervention. This aimed to minimize the influence of other perioperative factors on baseline immune status and enhance comparability among patients.

Assessment of early postoperative immune responses primarily utilized laboratory results from POD1. This timepoint typically corresponds to the phase of most pronounced immune and inflammatory reactions induced by surgical trauma and anesthetic stress, and it exhibits high completion rates and relatively consistent testing conditions in routine clinical practice. Therefore, POD1 was pre-specified as the core timepoint for primary outcome evaluation in this study.

For a subset of patients with complete and verifiable immune-related laboratory data available on POD3 or POD7, these were additionally included to describe temporal dynamics. Given that timepoints and frequencies were not strictly uniform in this retrospective study, these later timepoints were used solely for supplementary and exploratory analyses, not for primary outcome determination or main conclusion formulation.

For patients with multiple test results within the same time window, a pre-specified selection principle was applied: the result closest to the predefined assessment timepoint was prioritized for analysis to minimize measurement error from temporal deviation. All immune-related markers were analyzed using the original test values recorded in the laboratory information system, without artificial imputation, smoothing, or re-coding into categories, thereby preserving original information and reducing bias from subjective manipulation.

Statistical analysis

All statistical analyses were performed using R software (version 4.3.2, R Foundation for Statistical Computing, Vienna, Austria) and IBM SPSS Statistics (version 26.0, IBM Corp., Armonk, NY, USA). Continuous variables were assessed for normality. Normally distributed data are presented as mean ± standard deviation; non-normally distributed data as median (interquartile range). Categorical variables are presented as frequency (percentage). Intergroup comparisons of baseline characteristics served descriptive purposes only and were not used for inferential statistical significance.

Given the retrospective cohort design and non-random assignment of anesthetic technique, PS methods were employed to balance patient groups. PSs were constructed using multivariable logistic regression models incorporating a parsimonious set of potential confounders [age, body mass index, American Society of Anesthesiologists (ASA) physical status classification, tumor stage, surgical approach, and operative duration] to maintain an adequate ratio of observations to variables and minimize the risk of model overfitting given the sample size (n=121). Other intraoperative variables, such as estimated blood loss, crystalloid volume, and dexmedetomidine use, were not included in the PS model to avoid overadjustment and collinearity, but were examined separately in descriptive analyses. To estimate the average treatment effect (ATE) across the study population, stabilized inverse probability of treatment weighting (IPTW) was utilized as the primary adjustment method, while 1:1 nearest-neighbor PS matching was performed as a sensitivity analysis to evaluate the average treatment effect on the treated (ATT). Covariate balance was assessed using standardized mean differences (SMD), with SMD <0.1 considered indicative of good balance.

Following PS adjustment, multivariable linear regression models incorporating the same set of baseline covariates were used to analyze associations. This “doubly robust” estimation approach was employed to provide a secondary layer of protection against residual confounding, ensuring unbiased effect estimates even if one of the models (propensity or outcome) was partially misspecified. For immune markers with repeated measures data, linear mixed-effects models were further employed, treating patients as random effects and including timepoint and anesthetic technique as fixed effects, while assessing interaction effects between technique and time. Variables with skewed distributions, specifically NLR and CRP, were log-transformed to meet model assumptions of normality and homoscedasticity. Model results are presented as effect estimates with 95% confidence intervals. All tests were two-sided, with a significance level set at P<0.05. To account for the increased risk of Type I error inflation arising from the analysis of multiple, biologically and mathematically correlated inflammatory markers (e.g., lymphocyte count, neutrophil count, and NLR), P values for secondary outcomes were adjusted using the Benjamini-Hochberg false discovery rate (FDR) procedure.

To evaluate the robustness of findings, the primary analyses were repeated using different PS adjustment methods. Sensitivity analyses were also conducted, excluding cases with ambiguous anesthetic maintenance classification and employing alternative outcome definitions.

Results

Patient selection and baseline characteristics

Within the study timeframe, a retrospective search of the hospital’s electronic medical records and associated information systems identified 215 patients with a pathologically confirmed diagnosis of prostate cancer who had undergone elective radical robot-assisted prostatectomy. After applying the pre-defined inclusion and exclusion criteria, 94 patients were excluded due to comorbid immune system disorders or active infections, preoperative immunosuppressive therapy, missing key perioperative or laboratory data, or inability to definitively classify the anesthetic maintenance technique. Consequently, 121 patients were included in the final analysis. A flow chart detailing patient selection is provided in Figure 1.

Figure 1 Flowchart of patient screening, grouping, and final inclusion in the analysis. EMR, electronic medical record; POD, postoperative day; TIVA, total intravenous anesthesia.

Based on the anesthetic maintenance technique documented in the anesthesia records, the included patients were divided into the propofol-based TIVA group (n=62) and the volatile anesthesia group (n=59). Baseline characteristics are presented in Table 1. Prior to PS adjustment, statistically significant differences (P<0.05) were observed between the two groups for several demographic and perioperative variables. Specifically, patients in the TIVA group were younger, had a lower proportion of ASA class III status, underwent robot-assisted surgery more frequently, and had shorter operative and anesthesia durations, lower estimated intraoperative blood loss, and reduced crystalloid infusion volumes (all P<0.05). Additionally, preoperative PSA levels were higher in the volatile anesthesia group (P<0.05).

No statistically significant differences were found for baseline variables such as BMI, smoking, or major comorbidities. Furthermore, the number of clinicians involved and the use of NSAIDs were comparable between groups (Table 1), although the TIVA group had a higher rate of dexmedetomidine use (32.3% vs. 16.9%, P=0.047), reflecting the inherent preferences of clinicians who favor intravenous strategies.

These baseline disparities likely reflect associations between the choice of anesthetic technique and patient or surgical characteristics in routine clinical practice. To mitigate their potential impact on the study outcomes, PS methods were employed to balance the two groups. Following PS matching or weighting, good balance was achieved across all pre-specified covariates included in the model (age, body mass index, ASA classification, tumor stage, surgical approach, operative duration). The SMD for all covariates were <0.1, indicating excellent comparability of baseline characteristics between groups after adjustment. Baseline characteristics before PS adjustment are presented in Table 1, while the achievement of covariate balance following adjustment is demonstrated through SMDs in Figure 2.

Figure 2 SMDs of pre-specified covariates before and after propensity score adjustment. Covariate balance improved substantially after adjustment, with all post-adjustment SMDs below 0.1. ASA, American Society of Anesthesiologists; BMI, body mass index; CCI, Charlson comorbidity index; NLR, neutrophil-to-lymphocyte ratio; PSA, prostate-specific antigen; SMD, standardized mean difference.

Primary outcome analysis

The primary outcome was the change from preoperative baseline in the absolute peripheral blood lymphocyte count on POD1 (ΔALC). After achieving balance via PS adjustment, multivariable linear regression was used to analyze the association between anesthetic maintenance technique and lymphocyte count on POD1.

In the PS-adjusted analysis, the ALC on POD1 was (1.21±0.34)×10⁹/L in the TIVA group, compared to (0.96±0.31)×10⁹/L in the volatile anesthesia group. Multivariable regression analysis revealed a significantly higher lymphocyte count in the TIVA group, with an adjusted mean difference of 0.24×10⁹/L (95% CI: 0.11–0.37, P<0.001, Figure 3).

Figure 3 Association between anesthetic maintenance technique and the primary immune outcome (ΔALC). The forest plot displays the adjusted MD in the change of absolute peripheral blood lymphocyte count from preoperative baseline to postoperative day 1 (ΔALC) between the TIVA and volatile anesthesia groups. Data are presented as adjusted MD with 95% CIs derived from the primary propensity score-adjusted multivariable regression model. A positive MD indicates a smaller reduction (better preservation) of lymphocyte counts in the TIVA group relative to the volatile group. *, P<0.05; ***, P<0.001. ALC, absolute lymphocyte count; CI, confidence interval; CRP, C-reactive protein; LMR, lymphocyte-to-monocyte ratio; MD, mean difference; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; POD, postoperative day; SII, systemic immune-inflammation index; TIVA, total intravenous anesthesia; WBC, white blood cell count.

Further analysis using a linear mixed-effects model in patients with repeated-measures data confirmed a significant change in lymphocyte count over time (main effect of time, P<0.001). A statistically significant interaction was also observed between anesthetic technique and time (interaction P=0.01, Figure 3). Compared to the volatile anesthesia group, patients receiving TIVA experienced a smaller reduction in lymphocyte count and demonstrated a trend towards faster recovery in the early postoperative period (Figure 4).

Figure 4 Association between anesthetic maintenance technique and early postoperative inflammatory markers. *, P<0.05; **, P<0.01. CI, confidence interval; CRP, C-reactive protein; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; POD, postoperative day; SII, systemic immune-inflammation index; TIVA, total intravenous anesthesia; WBC, white blood cell count.

These findings remained consistent across different analytical models, suggesting that propofol-based TIVA is associated with a relatively attenuated degree of early postoperative immune suppression. A visual summary of the primary outcome analysis is provided.

Secondary outcome analysis

Secondary analyses evaluated the association between anesthetic technique and early postoperative inflammatory markers, including peripheral neutrophil count, total WBC, NLR, and CRP levels. All analyses were conducted after PS balancing.

On POD1, the ANC in the TIVA group was (6.20±1.85)×10⁹/L, lower than the (7.10±2.05)×10⁹/L observed in the volatile anesthesia group. Multivariable regression indicated a reduction of 0.78×10⁹/L in the TIVA group compared to the volatile group (95% CI: −1.45 to −0.11; P=0.02). For total WBC, the values were (8.90±2.10)×10⁹/L (TIVA) and (9.60±2.30)×10⁹/L (volatile); the intergroup difference was not statistically significant (adjusted difference −0.55×10⁹/L, 95% CI: −1.28 to 0.18; P=0.14, Figure 4).

Regarding inflammatory indices, the median NLR on POD1 was 5.30 [interquartile range (IQR), 4.10–6.90] in the TIVA group, significantly lower than the 7.10 (IQR, 5.40–9.60) in the volatile group. Regression models estimated the NLR in the TIVA group to be 1.35 units lower than in the volatile group (95% CI: −2.20 to −0.50, P=0.002). CRP levels also differed; the median POD1 CRP was 56 mg/L (IQR, 38–82 mg/L) for TIVA and 74 mg/L (IQR, 50–108 mg/L) for volatile anesthesia. Adjusted analysis showed lower CRP levels in the TIVA group (adjusted difference per 10 mg/L increment: −1.4, 95% CI: −2.6 to −0.2, P=0.02, Figure 4).

In summary, compared to volatile anesthesia, propofol-based TIVA was associated with relatively lower levels of several early postoperative inflammatory markers. As these were pre-specified secondary outcomes, the results are interpreted for supportive and trend analysis purposes.

Temporal dynamics of immune responses

For a subset of patients, immune-related laboratory data were available not only on POD1 but also on POD3 or POD7. Seventy-eight patients had peripheral blood lymphocyte count data from at least two timepoints, qualifying for repeated-measures analysis. A linear mixed-effects model was applied to these patients to analyze the temporal dynamics of immune markers. The model included patient as a random effect, with timepoint (POD1, POD3, POD7) and anesthetic technique as fixed effects, focusing on their interaction.

The results demonstrated a significant temporal trend for ALC in the early postoperative period (main effect of time, P<0.001). Overall, lymphocyte counts reached their nadir on POD1 and showed gradual recovery by POD3 and POD7 (Figure 5).

Figure 5 Temporal dynamics of absolute peripheral blood lymphocyte counts after surgery. POD, postoperative day; TIVA, total intravenous anesthesia.

The model estimated an average daily recovery rate for lymphocyte count of 0.09×10⁹/L (95% CI: 0.05–0.13, Table 2). Further analysis revealed a statistically significant interaction between anesthetic technique and time (interaction P=0.01), indicating divergent recovery trajectories. Patients in the TIVA group exhibited a faster rate of lymphocyte count recovery compared to the volatile anesthesia group. Specifically, the model indicated an additional daily increase of 0.06×10⁹/L/day (95% CI: 0.01–0.11) in the TIVA group during the recovery phase. This difference began to emerge by POD3 and became most pronounced by POD7 (Table 2).

Exploratory analyses of the temporal dynamics for selected secondary markers (NLR, CRP) also showed significant changes over time (main effect of time: P<0.05 for both, Figure 6), with a general pattern of early postoperative elevation followed by decline. However, the interaction between anesthetic technique and time for these markers was not statistically significant (interaction P>0.05, Table 2), suggesting similar temporal patterns between groups. These results are presented for supportive interpretation.

Figure 6 Postoperative temporal dynamics of secondary immune and inflammatory markers. CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio; POD, postoperative day; TIVA, total intravenous anesthesia.

Sensitivity analyses

Multiple pre-planned sensitivity analyses were performed to assess the robustness of the primary findings against variations in confounding control strategies and grouping definitions.

First, the primary outcome analysis was repeated using different PS adjustment methods. In a matched cohort obtained via 1:1 nearest-neighbor matching (caliper 0.2 SD of the logit PS, n=104; 52 per group), the ALC on POD1 remained significantly higher in the TIVA group, with an adjusted mean difference of 0.22×10⁹/L (95% CI: 0.08–0.36, P=0.002, Table 3). In a weighted analysis using stabilized inverse probability of treatment weighting (IPTW; effective sample size ≈118), the effect direction and magnitude were consistent, yielding an adjusted mean difference of 0.25×10⁹/L (95% CI: 0.12–0.38, P<0.001, Table 3). These results demonstrate that the association between TIVA and higher early postoperative lymphocyte count persists irrespective of the matching or weighting approach.

Second, to test the potential influence of anesthetic group definition, cases with apparent intraoperative switching between agents or mixed maintenance that precluded clear classification of the primary technique were excluded (n=9 excluded, n=112 remaining). In this restricted sample, multivariable regression results were not materially altered. The lymphocyte count on POD1 remained higher in the TIVA group (adjusted mean difference: 0.23×10⁹/L, 95% CI: 0.09–0.36, P=0.001, Table 3), indicating that the primary conclusion is not driven by a small number of mixed-anesthetic cases.

Furthermore, the ALC on POD1 was utilized as an alternative outcome to verify the consistency of the findings observed in the primary ΔALC analysis. In the primary analysis, the reduction in lymphocyte count from baseline to POD1 was significantly smaller in the TIVA group compared to the volatile group (between-group difference in ΔALC: +0.21×10⁹/L; 95% CI: 0.06–0.36, P=0.006; Table 3). This directionally aligns with the sensitivity finding of significantly higher ALCs on POD1 in the TIVA group (Table 3).

Collectively, these sensitivity analyses showed consistent effect direction and similar magnitude for the primary outcome across different confounding control strategies, sample restrictions, and outcome definitions, supporting the robustness of the main findings.

Subgroup analyses

Pre-specified subgroup analyses were conducted to explore potential heterogeneity in the association between anesthetic maintenance technique and early postoperative immune response. Subgroups were defined based on clinical relevance and prior literature: surgical approach (robot-assisted vs. non-robot-assisted), ASA physical status classification (I–II vs. ≥III), and tumor stage (localized vs. locally advanced). Within each subgroup, PSs were re-estimated and balance was verified within the specific clinical subset to ensure comparability. Subsequently, multivariable regression models consistent with the primary analysis were applied, incorporating an interaction term to test for effect modification.

In patients undergoing robot-assisted surgery (n=68), the ALC on POD1 was higher in the TIVA group, with an adjusted mean difference of 0.26×10⁹/L (95% CI: 0.11–0.41, P=0.001, Table 4). In non-robot-assisted surgery patients (n=53), the corresponding difference was 0.21×10⁹/L (95% CI: 0.04–0.38, P=0.01, Table 4). No significant interaction was observed between surgical approach and anesthetic technique (interaction P=0.64), suggesting that surgical approach did not significantly modify the association.

When stratified by ASA classification, a significant association between TIVA and higher postoperative lymphocyte count was observed in ASA I–II patients (n=79; adjusted mean difference: 0.25×10⁹/L, 95% CI: 0.10–0.39, P=0.001, Table 4). In ASA ≥ III patients (n=42), the effect direction was consistent but the confidence interval was wider, and the result did not reach conventional statistical significance (adjusted mean difference: 0.19×10⁹/L, 95% CI: −0.01, 0.39, P=0.06, Table 4). The interaction test was not statistically significant (interaction P=0.28), indicating no significant heterogeneity across different perioperative risk levels.

In the analysis stratified by tumor stage, TIVA was associated with a significantly higher lymphocyte count on POD1 in patients with localized disease (n=84; adjusted mean difference: 0.27×10⁹/L, 95% CI: 0.12–0.42, P<0.001, Table 4). In patients with locally advanced disease (n=37), the point estimate was 0.18×10⁹/L (95% CI: −0.03, 0.39, P=0.09, Table 4), which was not statistically significant. However, the effect direction was consistent across stages, and no significant interaction was detected (interaction P=0.34).

Overall, the direction of the association between anesthetic maintenance technique and the primary immune outcome was consistent across all pre-defined clinical subgroups. Statistically significant effect modification was not detected by surgical approach, ASA classification, or tumor stage. These findings suggest a degree of consistency in the main results across different clinical subsets; however, estimates in smaller subgroups (e.g., locally advanced tumor group, n=37) may be susceptible to model overfitting and increased statistical uncertainty, thus these results should be interpreted as exploratory. All subgroup analyses were pre-specified but should be considered hypothesis-generating given the limited sample size within each stratum, and they are not powered to detect small-to-moderate interaction effects.

Discussion

Fluctuations in immune function following radical prostatectomy hold potential implications for postoperative recovery and oncologic prognosis (8). Surgical stress activates the sympathetic-endocrine axis and promotes the release of pro-inflammatory mediators, leading to an immune-suppressive and pro-inflammatory phenotype characterized by lymphopenia and elevated inflammatory markers. As a modifiable factor within the perioperative period, anesthetic strategy has garnered increasing attention. Theoretical and experimental work suggests that TIVA, particularly with propofol, may possess a more favorable immunological profile, whereas the immunomodulatory effects of volatile anesthetics appear less consistent. This retrospective cohort study compared propofol-based TIVA with volatile anesthesia regarding early postoperative immune responses in patients undergoing radical robot-assisted prostatectomy, aiming to elucidate their differential impacts on postoperative immune status and underlying mechanisms. While these immune and inflammatory markers reflect perioperative immune status, they are surrogate biological indicators rather than direct measures of clinical outcomes. This study did not assess postoperative complications, infectious events, length of hospital stay, or oncologic endpoints such as recurrence or survival. Therefore, the observed differences in immune markers should not be interpreted as direct evidence of improved clinical outcomes, and the magnitude of these changes should be considered within the context of surrogate endpoints.

Our findings indicate that patients receiving propofol-based anesthesia exhibited a significantly higher ALC on POD1, alongside lower neutrophil counts, NLR, and CRP levels compared to those receiving volatile anesthesia. This suggests that propofol anesthesia may attenuate the degree of early postoperative lymphopenia and the associated systemic inflammatory response. Lymphocytes are pivotal for postoperative anti-infective and anti-tumor immunity, and the extent of their postoperative decline often mirrors the degree of adaptive immune suppression (27). Our observation of a higher lymphocyte count in the propofol group implies a potential protective effect on cellular immune function. This aligns with findings from a prospective study by Cho et al. [2017] in breast cancer surgery, which reported better preservation of postoperative immune function with propofol TIVA compared to sevoflurane (28). Conversely, volatile anesthetics like sevoflurane have been shown to induce apoptosis in T and B lymphocytes, reducing peripheral lymphocyte counts (29). Our results are consistent with these reports, demonstrating more pronounced lymphopenia with volatile anesthesia. It is noteworthy that a crossover trial in healthy volunteers without surgical stress found that propofol alone could induce transient lymphopenia and result in a higher post-anesthesia NLR compared to sevoflurane (23). This highlights the complexity of the direct immunomodulatory effects of anesthetics, suggesting that while propofol might cause a brief lymphocyte decrease in the absence of surgical trauma, it appears to exert a less negative impact on lymphocytes relative to volatile agents in the context of surgical stress, resulting in a milder degree of postoperative immune suppression. This divergence may relate to differences in the activated immune axes, anesthetic depth, or stress levels induced by surgery, warranting further mechanistic investigation.

Our study also found a significantly lower NLR on POD1 in the propofol group. NLR serves as a practical marker for the imbalance between innate immune activation and adaptive immune suppression and has been linked to adverse outcomes such as postoperative infection and tumor recurrence. Our observation that TIVA was associated with a more controlled rise in NLR suggests propofol may mitigate postoperative inflammatory amplification and immune suppression. This is supported by a large-scale study in colorectal cancer by Lee et al. [2022], which reported significantly lower NLR on PODs 2 and 5 in patients receiving propofol anesthesia versus volatile anesthesia (30). Furthermore, an analysis of a randomized trial incorporating regional blockade indicated that propofol combined with epidural anesthesia led to a smaller postoperative NLR increase compared to sevoflurane-based general anesthesia (31). While these findings corroborate our results, it is important to note that the clinical significance of early NLR differences remains debated (32,33). The study by Lee et al. [2022], despite observing lower NLR with propofol, did not find significant differences in early complications or long-term oncologic survival. This suggests that attenuating postoperative inflammatory markers alone may be insufficient to improve clinical outcomes, or that larger datasets and longer follow-up are required to detect such benefits. Therefore, while our study confirms the influence of anesthetic technique on intermediate immune phenotypes like NLR, the clinical translation of these surrogate marker changes requires further validation.

The observation of lower CRP levels in the propofol group on POD1 suggests TIVA may attenuate the acute postoperative inflammatory response. CRP, an acute-phase protein induced by pro-inflammatory cytokines like IL-6, reflects the systemic inflammatory burden. The lower CRP associated with propofol likely relates to its modulatory effects on the inflammatory cascade. Randomized controlled trials in less invasive surgeries, such as microdiscectomy, have demonstrated that propofol anesthesia significantly reduces postoperative IL-6 and CRP levels compared to volatile agents (29). This may be attributed to propofol’s capacity to inhibit the release of pro-inflammatory mediators. Experimental studies show that propofol suppresses TLR4/NF-κB pathway activation in monocytes/macrophages, reducing the production of cytokines like IL-1β and IL-6 (34-36). Additionally, propofol decreases the generation of inflammatory products mediated by cyclooxygenase and lipoxygenase, such as prostaglandin E2 and leukotrienes (37,38). These anti-inflammatory properties could explain the lower CRP levels and potentially milder tissue edema observed in our study, although edema was not directly measured. Animal models have reported that propofol can reduce inflammation-induced tissue edema formation (38). Clinically, this might translate to attenuated early postoperative tissue swelling and organ dysfunction. However, it should be acknowledged that not all evidence uniformly supports propofol’s superiority in inflammation control. A meta-analysis encompassing 23 studies found no significant overall difference in postoperative inflammatory marker levels between propofol and sevoflurane anesthesia (39). Moreover, in specific contexts like pulmonary lobectomy, sevoflurane has demonstrated benefits in reducing local alveolar inflammation, showing lower IL-6 concentrations in bronchoalveolar lavage fluid during one-lung ventilation compared to propofol (40). These findings indicate that the immunomodulatory and anti-inflammatory effects of anesthetics are complex and context-dependent, influenced by surgical type and local tissue environment. While propofol may offer systemic anti-inflammatory advantages, volatile agents can exhibit anti-inflammatory or cytoprotective effects in specific settings (40,41). Therefore, while affirming our findings, it is crucial to recognize the multifaceted nature of anesthetic modulation of inflammation, necessitating further investigation across different clinical scenarios.

Differences in the molecular mechanisms of propofol and volatile anesthetics provide insights into their divergent immune effects. Beyond its sedative-hypnotic properties, propofol exhibits distinct immunomodulatory characteristics. Primarily, it directly inhibits pro-inflammatory signaling pathways in peripheral immune cells, such as blocking NF-κB activation in macrophages and neutrophils, thereby reducing cytokine and adhesion molecule expression (34-36). Secondly, its impact on lymphocytes appears comparatively modest. Some studies indicate that propofol anesthesia does not reduce overall peripheral T-cell counts nor promote leukocyte apoptosis (42,43). Conversely, propofol may potentially enhance anti-tumor immune functions. Reports suggest it can increase the infiltration and activity of NK cells and CD8⁺ T lymphocytes. For instance, in vitro and animal experiments have shown that propofol does not significantly suppress NK cell cytotoxicity and may even promote NK cell-mediated anti-tumor activity by augmenting interferon-γ production (44). Small clinical studies have also observed increased postoperative NK cell cytotoxicity with propofol anesthesia combined with non-opioid analgesia, whereas sevoflurane-opioid anesthesia was associated with reduced NK cell activity (28,45). This constellation of mechanisms may explain the higher postoperative lymphocyte counts and potentially better-preserved NK cell function in the propofol group, contributing to a less suppressed immune state.

In contrast, volatile inhaled anesthetics like sevoflurane tend to exert more inhibitory effects on immunity. Sevoflurane has been shown to induce programmed cell death in lymphocytes, particularly increasing apoptosis in T and B cells, and can shift the Th1/Th2 helper T-cell balance towards a more immunosuppressive Th2-dominant state (42). Simultaneously, volatile agents can reduce peripheral NK cell numbers and activity, impairing the innate immune system’s tumor surveillance capability. These combined effects likely contribute to the more pronounced postoperative immune dysfunction and heightened inflammatory response observed with volatile anesthesia in our study (e.g., higher NLR, lower lymphocyte counts). The opioids frequently co-administered during anesthesia may further exacerbate this immunosuppressive tendency, as opioids can promote Th2 polarization and inhibit NK cell function (31,46). It is important to note, however, that volatile anesthetics are not uniformly detrimental to immune function. In certain models, sevoflurane has been shown to induce the release of anti-inflammatory cytokines like IL-10 and enhance macrophage phagocytic function (41). Our study, focused on systemic markers, cannot delineate the net effect of these complex actions across different immune cell subsets. In summary, propofol may confer immunoprotection by suppressing stress hormone and inflammatory mediator release while preserving lymphocyte and NK cell function (29,31). Volatile anesthesia, conversely, appears more prone to triggering stress responses and immune cell apoptosis, exhibiting a more immunosuppressive profile, albeit with concurrent anti-inflammatory potential in specific contexts. These mechanistic differences provide a rationale for the clinical observations and underscore the need to balance the suppression of excessive inflammation with the preservation of anti-tumor immunity when optimizing perioperative management.

As a retrospective single-center analysis, our study has several limitations. First, although PS methods were used to balance the groups, the modest overall sample size (n=121) may limit the stability of multivariable models and increase the risk of overfitting, particularly during subgroup and interaction analyses, potentially affecting the reliability of the effect estimates. The choice of anesthetic technique and perioperative management was largely driven by individual surgeon and anesthesiologist preferences, introducing potential provider-level confounding. Moreover, despite achieving balance on measured covariates, residual confounding from unmeasured factors—such as institutional practice patterns, variations in intraoperative hemodynamic management, and subtle differences in perioperative care pathways—cannot be excluded in this retrospective design. While we tracked adjuncts and management styles, several secondary outcomes evaluated in this study—specifically ALC, neutrophil count, and the derived NLR—are inherently redundant and mathematically correlated. This collinearity between markers, coupled with multiple testing, could lead to inflated Type I error rates; therefore, the consistency across these indicators should be viewed as mutually supportive rather than entirely independent evidence. Furthermore, our exposure classification relied on the ‘primary’ maintenance technique, which does not account for the potential immunomodulatory effects of propofol induction doses administered to both groups. Since the biological threshold for propofol’s anti-inflammatory benefits remains unknown, its use during induction—along with brief maintenance crossovers—might have attenuated the observed differences between the TIVA and volatile groups. Consequently, these findings may partially reflect broader variations in individualized perioperative care rather than the pharmacological effects of the primary anesthetic agent alone. Given the modest sample size and the retrospective nature of the data, caution is required when attributing the observed immunological variations specifically to the anesthetic technique, as they may be influenced by residual confounding from unmeasured perioperative management strategies. Furthermore, the subgroup analyses presented in this study, while pre-specified, were conducted with limited statistical power due to the small sample sizes within individual strata (e.g., n=37 for locally advanced tumors). These analyses are therefore exploratory in nature, and the lack of statistically significant interaction terms does not definitively exclude effect modification. Findings from subgroup analyses should be interpreted with caution and require validation in larger, adequately powered studies. Second, while the selected markers are representative, they do not capture the full spectrum of immune function. More detailed parameters such as NK cell activity, T-cell subset functionality, and cytokine profiles were not assessed. Future studies incorporating functional immune assays are needed to validate whether the observed phenotypic differences translate to functional consequences. Third, postoperative immune responses are dynamic. Our primary outcome focused on the POD1 timepoint, which, while recognized as a sensitive window for detecting postoperative immune suppression, may not fully capture recovery patterns. Although a subset of patients (n=78) provided data for POD3 and POD7, suggesting a faster recovery trend with propofol, the retrospective nature of these additional measurements—likely obtained based on clinical indications or longer hospital stays—may introduce selection bias if repeated testing occurred preferentially in patients with different clinical recovery trajectories. Finally, this study examined short-term immunological surrogate endpoints rather than long-term oncologic outcomes such as recurrence or survival. While some literature suggests TIVA may improve prognosis in certain cancers, large randomized controlled trials have also reported no significant differences in long-term outcomes between anesthetic techniques. Therefore, we cannot extrapolate from our immunological findings to infer long-term patient benefit. This area requires validation through prospective studies, such as the ongoing VAPOR-C trial, which will provide higher-level evidence. Moreover, this study did not assess postoperative clinical endpoints such as infectious complications, length of hospital stay, or oncologic outcomes; thus, the clinical relevance of the observed immunological differences remains to be determined.

Conclusions

This study, based on real-world retrospective cohort data, systematically evaluated the association between different anesthetic maintenance techniques and early postoperative immune responses following radical robot-assisted prostatectomy for prostate cancer. After rigorously controlling for potential confounders, propofol-based TIVA was significantly associated with a higher absolute peripheral blood lymphocyte count, alongside lower neutrophil counts, NLR, and CRP levels on POD1, compared to volatile anesthesia. These findings suggest a correlation between propofol-based TIVA and a relatively attenuated degree of early postoperative immune suppression and inflammatory response. Longitudinal analysis further indicated a more favorable recovery trajectory for lymphocyte counts in patients receiving propofol anesthesia. The consistency of these results across multiple analytical strategies and sensitivity analyses strengthens their robustness.

These findings add to the understanding of the potential biological significance of perioperative anesthetic strategy in prostate cancer surgery at the level of immune-inflammatory intermediate phenotypes. They suggest that the choice of anesthesia may extend beyond a mere technical decision, potentially serving as a modifiable factor influencing postoperative physiological responses. It is important to emphasize that this study primarily demonstrates an association between anesthetic technique and early postoperative immune markers; it does not establish a causal link to long-term tumor prognosis or other clinical endpoints. Furthermore, the results should be interpreted with caution due to the potential for provider-level confounding and the limited power of the study to exclude all broader perioperative influences. Future prospective, multicenter randomized studies, incorporating more detailed immune function assessments and long-term follow-up, are warranted to validate the clinical translational value of these immunological differences and to clarify the role of perioperative anesthetic strategy within the broader context of comprehensive cancer care.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tau.amegroups.com/article/view/10.21037/tau-2026-1-0096/rc

Data Sharing Statement: Available at https://tau.amegroups.com/article/view/10.21037/tau-2026-1-0096/dss

Peer Review File: Available at https://tau.amegroups.com/article/view/10.21037/tau-2026-1-0096/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tau.amegroups.com/article/view/10.21037/tau-2026-1-0096/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of the Affiliated Drum Tower Hospital of Nanjing University Medical School. The requirement for informed consent was waived because of the retrospective nature of the study and the use of anonymized patient data.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ferlay J, Ervik M, Lam F, et al. Global Cancer Observatory: Cancer Today. Lyon: International Agency for Research on Cancer; 2024. Available online: https://gco.iarc.who.int/today
2. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
3. Ferlay J, Colombet M, Soerjomataram I, et al. Cancer statistics for the year 2020: An overview. Int J Cancer 2021;149:778-89. [Crossref] [PubMed]
4. Schafer EJ, Laversanne M, Sung H, et al. Recent Patterns and Trends in Global Prostate Cancer Incidence and Mortality: An Update. Eur Urol 2025;87:302-13. [Crossref] [PubMed]
5. ManzoorFGuptaVPinkyLA Systematic Review of Multimodal Deep Learning and Machine Learning Fusion Techniques for Prostate Cancer Classification.medRxiv [Preprint] 2025. doi: .
6. Bezu L, Akçal Öksüz D, Bell M, et al. Perioperative Immunosuppressive Factors during Cancer Surgery: An Updated Review. Cancers (Basel) 2024;16:2304. [Crossref] [PubMed]
7. Choi H, Hwang W. Anesthetic Approaches and Their Impact on Cancer Recurrence and Metastasis: A Comprehensive Review. Cancers (Basel) 2024;16:4269. [Crossref] [PubMed]
8. Guo R, Yang WW, Zhong ML, et al. The relationship between anesthesia, surgery and postoperative immune function in cancer patients: a review. Front Immunol 2024;15:1441020. [Crossref] [PubMed]
9. Silva GN, Brandão VGA, Perez MV, et al. Neuroinflammatory Approach to Surgical Trauma: Biomarkers and Mechanisms of Immune and Neuroendocrine Responses. J Pers Med 2024;14:829. [Crossref] [PubMed]
10. Aizawa M, Yabusaki H, Matsuki A, et al. Predictive significance of surgery-induced lymphopenia on the survival after curative resection for locally advanced gastric cancer: a retrospective cohort analysis. World J Surg Oncol 2023;21:7. [Crossref] [PubMed]
11. Kim YH, Choi WJ. The effectiveness of postoperative neutrophils to lymphocytes ratio in predicting long-term recurrence after stomach cancer surgery. J Korean Surg Soc 2012;83:352-9. [Crossref] [PubMed]
12. Li S, Yao W, Liu R, et al. Severe lymphopenia as a prognostic factor in rectal cancer patients receiving adjuvant chemoradiotherapy: a retrospective study. Sci Rep 2023;13:7566. [Crossref] [PubMed]
13. Kao DD, Ferrandino RM, Roof SA, et al. Neutrophil-to-lymphocyte ratio as a predictor of surgical outcomes in head and neck cancer. Head Neck 2023;45:1903-12. [Crossref] [PubMed]
14. Kawamoto M, Higashi D, Kinjo K, et al. Neutrophil-to-Lymphocyte Ratio as a Biomarker for Postoperative Complications in Crohn's Disease. In Vivo 2025;39:1580-90. [Crossref] [PubMed]
15. Ortiz-López D, Acosta-Mérida MA, Casimiro-Pérez JA, et al. First day postoperative values of the neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and C-reactive protein as complication predictors following gastric oncologic surgery. Rev Gastroenterol Mex (Engl Ed) 2022;87:142-8. [Crossref] [PubMed]
16. Ivascu R, Torsin LI, Hostiuc L, et al. The Surgical Stress Response and Anesthesia: A Narrative Review. J Clin Med 2024;13:3017. [Crossref] [PubMed]
17. Chiu WC, Wu ZF, Lee MS, et al. Propofol-based total intravenous anesthesia is associated with less postoperative recurrence than desflurane anesthesia in thyroid cancer surgery. PLoS One 2024;19:e0296169. [Crossref] [PubMed]
18. Konstantis G, Katsadouros I, Tsaousi G, et al. Immunomodulatory Effects of Anesthetic Techniques in Lung Cancer Surgery: A Systematic Review and Meta-Analysis. Medicina (Kaunas) 2025;61:1263. [Crossref] [PubMed]
19. Zhang Y, Yu P, Bian L, et al. Survival benefits of propofol-based versus inhalational anesthesia in non-metastatic breast cancer patients: a comprehensive meta-analysis. Sci Rep 2024;14:16354. [Crossref] [PubMed]
20. Enlund A, Nikberg M, Berglund A, et al. Total intravenous or inhalational volatile anaesthesia and survival after colorectal cancer surgery: a Swedish national registry study. Anaesthesia 2025;80:386-94. [Crossref] [PubMed]
21. Cao SJ, Zhang Y, Zhang YX, et al. Long-term survival in older patients given propofol or sevoflurane anaesthesia for major cancer surgery: follow-up of a multicentre randomised trial. Br J Anaesth 2023;131:266-75. [Crossref] [PubMed]
22. Bennett-Guerrero E, Romeiser JL, DeMaria S, et al. Anesthesia Type during Cancer Surgery: Results of the GA-CARES Randomized, Multicenter Trial. Anesthesiology 2026;144:51-62. [Crossref] [PubMed]
23. Hasselager RP, Madsen SS, Møller K, et al. Effect of sevoflurane versus propofol on neutrophil-to-lymphocyte ratio in healthy individuals: a sub-study of a randomised crossover trial. BJA Open 2022;2:100005. [Crossref] [PubMed]
24. Köksal BG, Bollucuoğlu K, Şahin E, et al. The effect of anesthesia methods on the neutrophil-lymphocyte ratio in patients undergoing forearm surgery: A monocentric and retrospective study. Medicine (Baltimore) 2024;103:e40290. [Crossref] [PubMed]
25. Lai HC, Lee MS, Lin KT, et al. Propofol-based total intravenous anesthesia is associated with better survival than desflurane anesthesia in robot-assisted radical prostatectomy. PLoS One 2020;15:e0230290. [Crossref] [PubMed]
26. Dubowitz J, Riedel B, Blaas C, et al. On the horns of a dilemma: choosing total intravenous anaesthesia or volatile anaesthesia for cancer surgery, an enduring controversy. Br J Anaesth 2024;132:5-9. [Crossref] [PubMed]
27. Zhang M, Chen G, Jin X, et al. Pre-Operative Immunonutrition Enhances Postoperative Outcomes and Elevates Tumor-Infiltrating Lymphocyte Counts in Colorectal Cancer Patients: A Meta-Analysis of Randomized Controlled Trials. Nutr Cancer 2024;76:499-512. [Crossref] [PubMed]
28. Cho JS, Lee MH, Kim SI, et al. The Effects of Perioperative Anesthesia and Analgesia on Immune Function in Patients Undergoing Breast Cancer Resection: A Prospective Randomized Study. Int J Med Sci 2017;14:970-6. [Crossref] [PubMed]
29. Yediyıldız MB, Durmuş İ, Ak HY, et al. Comparison of inhalation and total intravenous anesthesia on inflammatory markers in microdiscectomy: a double-blind study. BMC Anesthesiol 2025;25:238. [Crossref] [PubMed]
30. Lee S, Pyo DH, Sim WS, et al. Early and Long-Term Outcomes after Propofol-and Sevoflurane-Based Anesthesia in Colorectal Cancer Surgery: A Retrospective Study. J Clin Med 2022;11:2648. [Crossref] [PubMed]
31. Murphy O, Forget P, Ma D, et al. Tumour excisional surgery, anaesthetic-analgesic techniques, and oncologic outcomes: a narrative review. Br J Anaesth 2023;131:989-1001. [Crossref] [PubMed]
32. Liu J, Liu Y, Xiang P, et al. Neutrophil-to-lymphocyte ratio predicts critical illness patients with 2019 coronavirus disease in the early stage. J Transl Med 2020;18:206. [Crossref] [PubMed]
33. Kosidło JW, Wolszczak-Biedrzycka B, Matowicka-Karna J, et al. Clinical Significance and Diagnostic Utility of NLR, LMR, PLR and SII in the Course of COVID-19: A Literature Review. J Inflamm Res 2023;16:539-62. [Crossref] [PubMed]
34. Anderson BJ, Bagshaw O. Practicalities of Total Intravenous Anesthesia and Target-controlled Infusion in Children. Anesthesiology 2019;131:164-85. [Crossref] [PubMed]
35. Hsing CH, Lin MC, Choi PC, et al. Anesthetic propofol reduces endotoxic inflammation by inhibiting reactive oxygen species-regulated Akt/IKKβ/NF-κB signaling. PLoS One 2011;6:e17598.
36. Schläpfer M, Piegeler T, Dull RO, et al. Propofol increases morbidity and mortality in a rat model of sepsis. Crit Care 2015;19:45. [Crossref] [PubMed]
37. Inada T, Hirota K, Shingu K. Intravenous anesthetic propofol suppresses prostaglandin E2 and cysteinyl leukotriene production and reduces edema formation in arachidonic acid-induced ear inflammation. J Immunotoxicol 2015;12:261-5. [Crossref] [PubMed]
38. Jansen L, Dubois BFH, Hollmann MW. The Effect of Propofol versus Inhalation Anesthetics on Survival after Oncological Surgery. J Clin Med 2022;11:6741. [Crossref] [PubMed]
39. O'Bryan LJ, Atkins KJ, Lipszyc A, et al. Inflammatory Biomarker Levels After Propofol or Sevoflurane Anesthesia: A Meta-analysis. Anesth Analg 2022;134:69-81. [Crossref] [PubMed]
40. Yuan JL, Kang K, Li B, et al. The Effects of Sevoflurane vs. Propofol for Inflammatory Responses in Patients Undergoing Lung Resection: A Meta-Analysis of Randomized Controlled Trials. Front Surg 2021;8:692734.
41. Oliveira TB, Braga CL, Battaglini D, et al. Comparison between sevoflurane and propofol on immunomodulation in an in vitro model of sepsis. Front Med (Lausanne) 2023;10:1225179. [Crossref] [PubMed]
42. Rossaint J, Zarbock A. Anesthesia-induced immune modulation. Curr Opin Anaesthesiol 2019;32:799-805. [Crossref] [PubMed]
43. Dowling TJ, Munakomi S, Dowling TJ. Microdiscectomy. 2023 Aug 13. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2026 Jan–.
44. Inada T, Kubo K, Kambara T, et al. Propofol inhibits cyclo-oxygenase activity in human monocytic THP-1 cells. Can J Anaesth 2009;56:222-9. [Crossref] [PubMed]
45. Lim JA, Oh CS, Yoon TG, et al. The effect of propofol and sevoflurane on cancer cell, natural killer cell, and cytotoxic T lymphocyte function in patients undergoing breast cancer surgery: an in vitro analysis. BMC Cancer 2018;18:159. [Crossref] [PubMed]
46. Fang P, Zhou J, Xia Z, et al. Effects of Propofol Versus Sevoflurane on Postoperative Breast Cancer Prognosis: A Narrative Review. Front Oncol 2021;11:793093. [Crossref] [PubMed]