Th1/Th2 cytokine polarization in three different anaesthesia protocols administered to cats undergoing ovariohysterectomy Authors: Tarik Safak, Oznur Yılmaz-Koc, Ebru Karakaya-Bilen, Nicolò Columbano Published: 2026 ABSTRACT This study investigated perioperative Th1/Th2 cytokine polarization and biochemical changes in cats undergoing ovariohysterectomy under three different anesthesia protocols. Twenty-four healthy mixed-breed female cats were randomly allocated into three groups (n = 8/group): propofol–isoflurane (Group I), xylazine–ketamine hydrochloride (Group II), and xylazine–isoflurane (Group III). Venous blood samples were collected preoperatively (0 h) and at 6 and 12 h postoperatively. Serum concentrations of TNF-α, IFN-γ, IL-2, IL-4, IL-5, and IL-10 were quantified using feline-specific ELISA kits, and biochemical variables were measured with an automated analyzer. In Group I, tumor necrosis factor (TNF)-α, increased postoperatively (12.64 ± 2.28 ng/mL) compared to preoperative levels (6.79 ± 0.74 ng/mL; p < 0.05). In Group II, both interferon (IFN)-γ and interleukin (IL)-2 levels decreased after the operation compared to pre-op 0 h (p < 0.05). Group III showed a decrease in IL-4 and IL-10 postoperatively (124.21 ± 23.45 pg/mL and 106.05 ± 5.25 ng/mL, respectively) compared to pre-op 0 h (197.28 ± 14.40 pg/mL and 113.14 ± 11.19 ng/mL, respectively; p < 0.05). IL-5 concentrations did not change significantly in any group. Postoperative biochemical changes were transient and remained within reference ranges, with no clinical complications observed. These findings indicate that anesthesia protocols differentially influence perioperative immune responses in cats. The xylazine–ketamine combination induced the greatest cytokine fluctuations, whereas xylazine–isoflurane best preserved Th1/Th2 balance. Anesthetic selection may play a critical role in minimizing perioperative immune modulation. The xylazine–isoflurane protocol may be a preferable option for feline ovariohysterectomy, particularly in patients where immune stability is clinically relevant. These findings indicate that anesthesia protocols are associated with distinct short-term perioperative immune profiles in cats undergoing elective ovariohysterectomy. The observed differences should be interpreted as preliminary and hypothesis-generating, and larger studies with standardized pain assessment and longer follow-up are warranted to clarify clinical implications. Introduction The cellular immune system can be compromised by surgery and anesthesia, potentially adversely affecting patient outcomes [1]. Additionally, anesthesia used during surgical procedures can modulate stress responses and influence the immune system, leading to postoperative complications such as delayed wound healing and increased susceptibility to infections [2, 3]. In veterinary medicine, general anesthesia is commonly employed for ovariohysterectomy procedures in cats. Ketamine, a widely used dissociative anesthetic [4], is often combined with Propofol and Isoflurane for induction and maintenance of anesthesia [5]. General anesthesia has the dual effect of both suppressing and activating the immune system, which is essential for patients’ survival post-surgery. Anesthetic agents impact immune functions and directly affect both non-specific and adaptive immune factors [6]. T-helper (Th) lymphocytes play a crucial role in the immune response and can be categorized into two main subgroups: Th1 and Th2 cells. Recent research has highlighted a strong correlation between immune function and the imbalance between Th1 and Th2 cells during surgical procedures, with the Th1/Th2 ratio becoming a focal point of scientific investigation. Cytokines, small proteins produced by Th cells, are integral to the immune response and inflammatory processes. Th1 cells secrete cytokines such as interferon (IFN)-γ, tumor necrosis factor (TNF)-α, interleukin (IL)-2, and IL-1β, which enhance cellular immunity, while Th2 cells produce cytokines including IL-4, IL-5, IL-10, and IL-13, stimulating humoral immunity [7–9]. Anesthetic agents influence cytokine production and the balance between Th1 and Th2 cytokines. Studies in humans have shown that both ketamine [10] and propofol [11] inhibit Th1 cytokines. However, there is a lack of research on the effects of anesthesia on cytokines in cats. Therefore, it is important to investigate how different anesthesia protocols may affect cytokine levels. To the authors’ knowledge, the effects of Xylazine-Ketamine, Propofol-Isoflurane, and Propofol-Xylazine combinations on cytokines and biochemical variables in cats have not been previously studied. It is hypothesized that these anesthesia protocols will alter cytokine profiles integral to immune function. Consequently, the present study aims to examine the effects of these anesthesia protocols on cytokines and biochemical variables. Materials and methods Ethics approval The study was approved by the Local Ethical Committee of Kastamonu University (date: 10.01.2025, approval number: 2025/02).The owners of the cats approved the informed consent form. Animals and study groups Before finalizing the study design, a preliminary pilot study was conducted involving a limited number of cases to gather initial data, which informed the subsequent power analysis. This preliminary data enabled a more accurate estimation of the sample size required to detect statistically significant differences among groups. In the main study, variables were measured across three groups at three different time points. A G*Power analysis [12] was conducted according to the recommendations from previous studies utilizing an effect size of 0.40, an alpha level of 0.05, and a power (1 - β error probability) of 0.95. This analysis indicated that a total of 24 cats would be sufficient to reliably detect differences in the measured variables. To address uncertainty in effect size selection, we performed a sensitivity analysis across six plausible Cohen’s f scenarios and calculated the achieved power for n = 24 (8 cats per group) at α = 0.05 [13]. Twenty-four healthy female mixed-breed healthy female cats (1–4 years old; 2.8–4.2 kg) were included in this study. These cats were selected from those presented to the Kastamonu University, Faculty of Veterinary Medicine, Department of Obstetrics and Gynecology for elective ovariohysterectomy. Cats were block-randomized into three groups (n = 8 each) (Fig. 1). Prior to inclusion, cat owners were informed about the anesthesia procedures and the study, providing informed consent by signing a consent form. Cats that were clinically assessed as being in poor body condition (BCS < 3/9), pregnant, or suffering from other illnesses were excluded from the study. Preoperatively, medical histories and general examinations were conducted for each cat. As part of the preoperative preparation, a 22-gauge venous catheter was placed in the cephalic vein to allow for the infusion of an isotonic saline solution during the ovariohysterectomy procedure. Additionally, blood samples were collected in tubes containing tripotassium ethylenediaminetetraacetic acid (K3EDTA) for hematological analysis, and in tubes without anticoagulant for biochemical and cytokine analyses. Hematological parameters including hemoglobin (HGB), white blood cells (WBC), platelet count (PLT), red blood cells (RBC), and hematocrit (HCT) were measured using a device (Norma Ivet-5, Hungary). Only those cats with preoperative and postoperative hematological values within the reference range were included in the study. All cats were fasted for approximately 8 h prior to surgery. Fig. 1Criteria for inclusion of cats in the study Criteria for inclusion of cats in the study Cats scheduled for ovariohysterectomy were randomly divided into three groups according to the anesthesia protocol to be applied. In Group I (n = 8), anesthesia was induced with an intravenous (IV) injection of propofol (Propofol 1% Fresenius, Uppsala, Sweden) at a dose of 6 mg/kg. After a 10-minute interval, a well-fitted cat mask with a rubber seal was placed over the cat’s muzzle and connected to a pediatric breathing circuit. Pre-oxygenation was performed by delivering 100% oxygen at a flow rate of 1.5–2.5 L/min for 1 min. Anesthesia induction was carried out using isoflurane with an agent-specific, out-of-circuit vaporizer set to maximum output (Mindray V60 Isoflurane, Mindray Animalcare, Shenzhen, China) [14]. In Group II (n = 8), xylazine hydrochloride (2 mg/kg intramuscular, Rompun 2%, Bayer, Turkey) was administered as premedication, followed by induction with ketamine hydrochloride (Ketasol 10%, Richterpharma, Austria) at 10 mg/kg IM approximately 10 min later [15]. For Group III (n = 8), xylazine hydrochloride (2 mg/kg IM, Rompun 2%, Bayer, Turkey) was administered initially. After a 10-minute interval, a well-fitted cat mask with a rubber seal was placed over the cat’s muzzle and connected to a pediatric breathing circuit. Pre-oxygenation was performed by delivering 100% oxygen at a flow rate of 1.5–2.5 L/min for 1 min. Anesthesia induction was carried out using isoflurane with an agent-specific, out-of-circuit vaporizer set to maximum output (Mindray V60 Isoflurane, Mindray Animalcare, Shenzhen, China). Induction was deemed complete once the cat was fully relaxed and exhibited loss of mandibular tone. For all groups, following intubation, the orotracheal tube was connected to the anesthesia machine for the duration of the procedure (Mindray WATO EX-20Vet, Mindray Animalcare, Shenzhen, China). Anesthesia was maintained with isoflurane at 1.6% in oxygen (Isoflurane USP, PA, USA), with animals under spontaneous ventilation. Physiological monitoring (HR, RR, SpO₂, etCO₂, NIBP) was continuous. Manual ventilation was provided if apnea occurred or if end-tidal CO₂ (etCO₂) values exceeded 45 mmHg, as monitored by a multiparametric anesthesia monitor (uMEC12Vet, Mindray Animalcare, Shenzhen, China). The vaporizer was gradually and continuously adjusted to maintain etCO₂ levels within the desired range throughout the procedure. Surgical procedures After ensuring standard aseptic operating conditions were met, a left flank ovariohysterectomy was performed as detailed in the literature [16]. All surgeries were conducted by the single surgeon to minimize variability in technique. Following premedication and prior to surgery, each queen received an intramuscular injection of amoxicillin–clavulanic acid at a dose of 12.5 mg/kg (Synulox, Zoetis, UK) [17] and meloxicam (Bavet, Turkey) administered subcutaneously at a dose of 0.1 mg/kg for analgesic treatment [18]. Postoperative pain and recovery were clinically evaluated and recorded by the attending clinicians as part of routine care. Rescue analgesia was available if required; however, no cat required rescue analgesia. A validated feline pain scoring system was not applied; therefore, postoperative assessments were used for clinical management and are not interpreted as quantitative pain scores. A simplified behavioural pain assessment consistent with the Glasgow CMPS–Feline was used for routine monitoring; however, scores were not recorded in a format suitable for formal statistical correlation with cytokine concentrations. Blood samples and biochemical analysis Venous blood (2 mL) were collected from all cats at three different times for biochemical analysis and measurement of cytokine concentrations. The first blood sample was collected at 0 h before the operation (pre-op 0 h), when general examinations were performed and before anaesthesia was started. Second blood samples were collected after the operations were completed, approximately 6 h after the first blood samples were taken (post-op 6 h). The last blood samples were collected 12 h after the first sample was taken (post-op 12 h). All blood samples were collected from the cephalic vein into tubes that did not contain any anticoagulant. The serum was separated by centrifugation at 2,795x g for 10 min and then stored at -80 °C until analysis. After thawing, albumin (ALB), alkaline phosphatase (ALP), aspartate aminotransferase (AST), alanineaminotransferase (ALT), calcium (Ca), creatine kinase (CK), chlorine (Cl), creatine, glucose, L-lactate, magnesium (Mg), sodium (Na), phosphorus (P), and total protein (TP) concentrations were measured with an automatic biochemistry analyzer (Diasys Respons 910Vet, Holzheim, Germany) in serum samples. Cytokine analysis Cytokine concentrations (TNF-α, IFN-γ, IL-2, 4, 5, and 10) were measured by the feline species-specific commercially available Enzyme-Linked Immunosorbent Assay (ELISA) test method. TNF-α (Cat no: E0031Cat), IFN-γ (Cat no: E0013Cat), IL-2 (Cat no: E0051Cat), IL-4 (Cat no: E0121Cat), IL-5 (Cat no: E0122Cat), and IL-10 (Cat no: 162 E0030Cat, Bioassay Technology Laboratory, Shanghai, China) ELISA kits were used in the study. The manufacturer’s instructions and methods were followed during ELISA test application. After all processes, an ELISA reader read the plates at 450 nm to determine serum cytokine concentrations [19]. Statistical analysis An a priori sample size estimation was performed using G*Power. Because the final inferential analyses were non-parametric, sample size planning was expressed using an ANOVA-based approximation (F tests → ANOVA: fixed effects, omnibus, one-way; three groups; α = 0.05) to ensure transparency and reproducibility. To account for uncertainty in selecting a single effect size, a sensitivity analysis across six plausible Cohen’s f scenarios was conducted for the finalized sample size (N = 24; 8 cats per group). The normality of distribution for biochemical variables and cytokine concentrations was evaluated both visually, using probability plots and histograms, and statistically with the Shapiro-Wilk test. Based on this evaluation, it was determined that the characteristic values did not follow a normal distribution within each group. Consequently, the Kruskal-Wallis test, a non-parametric statistical technique, was employed for comparisons between groups. Following the Kruskal-Wallis test, post-hoc pairwise intragroup comparisons were conducted using the Bonferroni-corrected Mann-Whitney U test. To analyze non-parametric data for biochemical and cytokine concentrations measured in serum samples collected across three time points, the Friedman test was applied. Subsequently, the Wilcoxon test was utilized for post-hoc analysis to determine differences between the three repeated measurement points within each group. Results are presented as median ± standard error of the mean. Significance was p < 0.05. Statistical analyses were performed using SPSS version 22 (Statistical Package for the Social Sciences, IL, USA). Intra-operative variables (HR, NIBP/MAP, SpO2, etCO2, RR) were summarized for each subject during the maintenance phase of anesthesia and compared among groups using the Kruskal–Wallis test. Variables with approximately symmetric distribution are presented as mean ± standard deviation, whereas parameters potentially affected by skewness or ceiling effects (e.g., SpO2) are reported as median (interquartile range). Statistical significance was set at p < 0.05. Results The cats’ ages ranged from 12 to 48 months and their body weight ranged from 2.8 to 4.1 kg. There was no statistically significant difference in age and body weight between the groups of cats. There was also no statistical difference in the preoperative and postoperative haematological values. The duration of each operation ranged from 8.5 to 10 min. All cats recovered without complications; median surgical time was 9.2 ± 1.1 min. Pain assessment scores remained low in all animals throughout the observation period. No cat reached the predefined intervention threshold at recovery or during subsequent 30-minute evaluations; therefore, rescue methadone administration was not required in any case. Importantly, no clinically relevant differences in pain assessment outcomes or recovery quality were observed among groups. Intra-operative monitoring (HR, NIBP/MAP, SpO2, etCO2, RR) was continuous in all cats under spontaneous ventilation. Intra-operative cardiorespiratory variables did not differ significantly among groups (Table 1). Table 1Intra-operative cardiorespiratory variables during anesthesiaVariablesGroup I (n = 8)Group II (n = 8)Group III (n = 8) P HR (beats min − 1)129 ± 12104 ± 14121 ± 13NDNIBP/MAP (mmHg)88 ± 1084 ± 1186 ± 9NDSpO2 (%)99 (97–100)99 (98–100)98 (96–100)NDetCO2 (mmHg)40 (36–44)41 (37–44)39 (35–45)NDRR (breaths min − 1)16 (12–22)14 (10–18)15 (11–20)NDMajor events (Y/N)NNN—ND no statistically significant difference among groups (p > 0.05) Table 1Intra-operative cardiorespiratory variables during anesthesiaVariablesGroup I (n = 8)Group II (n = 8)Group III (n = 8) P HR (beats min − 1)129 ± 12104 ± 14121 ± 13NDNIBP/MAP (mmHg)88 ± 1084 ± 1186 ± 9NDSpO2 (%)99 (97–100)99 (98–100)98 (96–100)NDetCO2 (mmHg)40 (36–44)41 (37–44)39 (35–45)NDRR (breaths min − 1)16 (12–22)14 (10–18)15 (11–20)NDMajor events (Y/N)NNN—ND no statistically significant difference among groups (p > 0.05) Intra-operative cardiorespiratory variables during anesthesia ND no statistically significant difference among groups (p > 0.05) Table 2 provides the measurements of serum cytokine concentrations at different time points for each group, indicating the changes between the groups. Table 2Serum cytokine concentration changes between groups and three time pointsVariables/GroupsTime points p Pre-op 0 hPost-op 6 hPost-op 12 hTNF-α (ng mL− 1) Group I6.79 ± 0.74a12.64 ± 2.28b10.97 ± 4.68b* Group II6.70 ± 1.18a11.48 ± 1.40b10.61 ± 0.58b* Group III9.70 ± 1.3710.99 ± 0.489.37 ± 0.38ND PNDNDNDIFN-γ (ng mL− 1)P Group I22.06 ± 3.84A19.16 ± 2.37B22.55 ± 2.73ND Group II27.25 ± 3.63bB16.60 ± 5.39aA20.23 ± 3.04a* Group III20.08 ± 3.08A18.61 ± 2.03B21.03 ± 4.39ND P**NDIL-2 (ng mL− 1)PGroup I4.24 ± 0.70b2.86 ± 0.50abB1.75 ± 0.38a*Group II5.01 ± 0.93b2.51 ± 0.47aB1.95 ± 0.13a*Group III4.17 ± 0.83b1.98 ± 0.34aA1.00 ± 0.16a*PND*NDIL-4 (pg mL− 1)P Group I286.03 ± 38.63bB102.59 ± 16.55aA137.99 ± 19.62aA* Group II261.78 ± 24.77bB134.23 ± 20.62aB161.18 ± 14.89aB* Group III197.28 ± 14.40bA124.21 ± 23.45aB125.66 ± 20.89aA* P***IL-5 (pg mL− 1) Group I58.99 ± 2.1059.80 ± 3.5360.86 ± 3.30ND Group II61.78 ± 3.1660.73 ± 4.9059.12 ± 2.34ND Group III67.67 ± 1.9563.24 ± 2.3559.75 ± 1.16ND PNDNDNDIL-10 (ng mL− 1) Group I131.71 ± 5.30b92.27 ± 7.00a77.89 ± 6.65a* Group II119.38 ± 6.85b92.62 ± 12.04ab81.37 ± 9.89a* Group III113.14 ± 11.19b106.05 ± 5.25b66.96 ± 4.59a* PNDNDNDND No differenceA, B: difference between groups in the same column is statistically significant (p ˂ 0.05)a, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Table 2Serum cytokine concentration changes between groups and three time pointsVariables/GroupsTime points p Pre-op 0 hPost-op 6 hPost-op 12 hTNF-α (ng mL− 1) Group I6.79 ± 0.74a12.64 ± 2.28b10.97 ± 4.68b* Group II6.70 ± 1.18a11.48 ± 1.40b10.61 ± 0.58b* Group III9.70 ± 1.3710.99 ± 0.489.37 ± 0.38ND PNDNDNDIFN-γ (ng mL− 1)P Group I22.06 ± 3.84A19.16 ± 2.37B22.55 ± 2.73ND Group II27.25 ± 3.63bB16.60 ± 5.39aA20.23 ± 3.04a* Group III20.08 ± 3.08A18.61 ± 2.03B21.03 ± 4.39ND P**NDIL-2 (ng mL− 1)PGroup I4.24 ± 0.70b2.86 ± 0.50abB1.75 ± 0.38a*Group II5.01 ± 0.93b2.51 ± 0.47aB1.95 ± 0.13a*Group III4.17 ± 0.83b1.98 ± 0.34aA1.00 ± 0.16a*PND*NDIL-4 (pg mL− 1)P Group I286.03 ± 38.63bB102.59 ± 16.55aA137.99 ± 19.62aA* Group II261.78 ± 24.77bB134.23 ± 20.62aB161.18 ± 14.89aB* Group III197.28 ± 14.40bA124.21 ± 23.45aB125.66 ± 20.89aA* P***IL-5 (pg mL− 1) Group I58.99 ± 2.1059.80 ± 3.5360.86 ± 3.30ND Group II61.78 ± 3.1660.73 ± 4.9059.12 ± 2.34ND Group III67.67 ± 1.9563.24 ± 2.3559.75 ± 1.16ND PNDNDNDIL-10 (ng mL− 1) Group I131.71 ± 5.30b92.27 ± 7.00a77.89 ± 6.65a* Group II119.38 ± 6.85b92.62 ± 12.04ab81.37 ± 9.89a* Group III113.14 ± 11.19b106.05 ± 5.25b66.96 ± 4.59a* PNDNDNDND No differenceA, B: difference between groups in the same column is statistically significant (p ˂ 0.05)a, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Serum cytokine concentration changes between groups and three time points ND No difference A, B: difference between groups in the same column is statistically significant (p ˂ 0.05) a, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05) *: p ˂ 0.05 The change in cytokine concentration in cats in Group I is shown in Fig. 2. Although the TNF-α concentration, one of the Th1 cytokines, increased after the operation (12.64 ± 2.28 ng/mL) compared to before the operation (6.79 ± 0.74 ng/mL), this increase decreased 12 h after the operation (10.97 ± 4.68 ng/mL) (p ˂ 0.05). Moreover, IL-2 concentration in this group continued to decrease at postoperative time points (p ˂ 0.05). In the Group I, the concentrations of both IL-4 and IL-10, which are Th2 cytokines, decreased at postoperative time points (p ˂ 0.05). There is no statistical difference in IFN-γ and IL-5 concentrations at three different time points (p > 0.05). Fig. 2Perioperative cytokine change in Group I. Asterisks indicate statistical difference (*p < 0.05) Perioperative cytokine change in Group I. Asterisks indicate statistical difference (*p < 0.05) The Group II is show in Fig. 3. The TNF-α concentration change in this group showed the same change as in the Group I. We see that it increased after the operation (11.48 ± 1.40 ng/mL) compared to pre-op 0 h (6.70 ± 1.18 ng/mL), but decreased to 12 h after the operation (10.61 ± 0.58 ng/mL) (p ˂ 0.05). Likewise, IL-2 concentration continued to decrease after the operation (p ˂ 0.05). In this anaesthesia protocol, unlike Group I, the IFN-γ concentration paralleled the IL-2 change, decreasing after the operation (16.60 ± 5.39 ng/mL) compared to pre-op 0 h (27.25 ± 3.63 ng/mL) (p ˂ 0.05). Th2 cytokines, as in the Group I, post-op 6 h concentrations of both IL-4 (134.23 ± 20.62 pg/mL) and IL-10 (92.62 ± 12.04 ng/mL) were decreased compared to pre-op 0 h 261.78 ± 24.77 pg/mL and 119.38 ± 6.85 ng/mL, respectively (p ˂ 0.05). There is no statistical difference in IL-5 concentration at three different time points (p > 0.05). Fig. 3Perioperative cytokine change in Group II. Asterisks indicate statistical difference (*p < 0.05) Perioperative cytokine change in Group II. Asterisks indicate statistical difference (*p < 0.05) As in the other two groups, in the Group III, IL-2 concentration decreased in this group after the operation (1.98 ± 0.34 ng/mL) compared to before operation (4.17 ± 0.83 ng/mL) (p ˂ 0.05). Concentrations of Th2 cytokines IL-4 and IL-10 decreased after the operation (124.21 ± 23.45 pg/mL and 106.05 ± 5.25 ng/mL, respectively) compared to the pre-op 0 h (197.28 ± 14.40 pg/mL and 113.14 ± 11.19 ng/ mL, respectively) (Fig. 4). There is no statistical difference in TNF-α, IFN-γ, and IL-5 concentrations at three different time points (p > 0.05). Fig. 4Perioperative cytokine change in Group III. Asterisks indicate statistical difference (*p < 0.05) Perioperative cytokine change in Group III. Asterisks indicate statistical difference (*p < 0.05) In Group I, ALT, CK, Cl, and glucosa levels increased after the operation compared to the pre-op period. On the contrary, the creatine level decreased after the operation. ALP level increases in the post-op 6 h. But it decreased again at later point in time (p ˂ 0.05) (Table 3). Table 3Perioperative biochemical variables change in Group IVariablesTime points p Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)3.85 ± 0.113.21 ± 0.183.22 ± 0.15NDALP (U L− 1)59.25 ± 7.59b42.26 ± 5.08a60.31 ± 5.83b*ALT (U L− 1)23.88 ± 3.11a24.51 ± 0.01a28.99 ± 2.07b*AST (U L− 1)26.38 ± 1.43a35.56 ± 1.16b25.07 ± 3.64a*CK (U L− 1)159.86 ± 11.34a335.65 ± 27.85b486.87 ± 13.11c*Cl (mmol L− 1)109.86 ± 10.27a105.51 ± 9.37a136.07 ± 9.32b*Creatine (mg dL− 1)1.92 ± 0.10c1.02 ± 0.05b0.72 ± 0.03a*Glucose (mg dL− 1)66.75 ± 5.33a114.98 ± 7.39b230.67 ± 9.53c*L-lactate (mmol L− 1)7.06 ± 1.175.66 ± 0.976.91 ± 2.05NDND No differencea, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Table 3Perioperative biochemical variables change in Group IVariablesTime points p Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)3.85 ± 0.113.21 ± 0.183.22 ± 0.15NDALP (U L− 1)59.25 ± 7.59b42.26 ± 5.08a60.31 ± 5.83b*ALT (U L− 1)23.88 ± 3.11a24.51 ± 0.01a28.99 ± 2.07b*AST (U L− 1)26.38 ± 1.43a35.56 ± 1.16b25.07 ± 3.64a*CK (U L− 1)159.86 ± 11.34a335.65 ± 27.85b486.87 ± 13.11c*Cl (mmol L− 1)109.86 ± 10.27a105.51 ± 9.37a136.07 ± 9.32b*Creatine (mg dL− 1)1.92 ± 0.10c1.02 ± 0.05b0.72 ± 0.03a*Glucose (mg dL− 1)66.75 ± 5.33a114.98 ± 7.39b230.67 ± 9.53c*L-lactate (mmol L− 1)7.06 ± 1.175.66 ± 0.976.91 ± 2.05NDND No differencea, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Perioperative biochemical variables change in Group I ND No difference a, b, c: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05) *: p ˂ 0.05 In group II, CK, glucose, and creatine levels showed a change as in Group I. ALP level post-op 6 h (42.26 ± 5.08 U/L) time point measurements decreased compared to pre-op 0 h (59.25 ± 7.59 U/L). ALT level post-op 12 h (28.99 ± 2.07 U/L) measurement was found to be higher than both pre-op 0 h (23.88 ± 3.11 U/L) and post-op 6 h (24.51 ± 0.01 U/L) measurements (p ˂ 0.05) (Table 4). Table 4Perioperative biochemical variables change in Group IIVariablesTime points p Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)3.71 ± 0.053.98 ± 0.214.08 ± 0.18NDALP (U L− 1)71.88 ± 15.01b47.67 ± 5.83a40.63 ± 7.59a*ALT (U L− 1)23.38 ± 2.07a20.33 ± 3.11a37.87 ± 2.01b*AST (U L− 1)29.63 ± 3.16a47.33 ± 4.06b25.09 ± 1.36a*CK (U L− 1)214.38 ± 15.21a421.00 ± 36.11b500.25 ± 60.71b*Cl (mmol L− 1)118.75 ± 9.27101.67 ± 19.37111.01 ± 28.07NDCreatine (mg dL− 1)1.38 ± 0.12b0.92 ± 0.02a0.91 ± 0.02a*Glucose (mg dL− 1)72.50 ± 7.31a220.00 ± 27.39b204.64 ± 38.35b*L-lactate (mmol L− 1)5.69 ± 1.384.44 ± 1.875.48 ± 2.08NDND No differencea, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Table 4Perioperative biochemical variables change in Group IIVariablesTime points p Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)3.71 ± 0.053.98 ± 0.214.08 ± 0.18NDALP (U L− 1)71.88 ± 15.01b47.67 ± 5.83a40.63 ± 7.59a*ALT (U L− 1)23.38 ± 2.07a20.33 ± 3.11a37.87 ± 2.01b*AST (U L− 1)29.63 ± 3.16a47.33 ± 4.06b25.09 ± 1.36a*CK (U L− 1)214.38 ± 15.21a421.00 ± 36.11b500.25 ± 60.71b*Cl (mmol L− 1)118.75 ± 9.27101.67 ± 19.37111.01 ± 28.07NDCreatine (mg dL− 1)1.38 ± 0.12b0.92 ± 0.02a0.91 ± 0.02a*Glucose (mg dL− 1)72.50 ± 7.31a220.00 ± 27.39b204.64 ± 38.35b*L-lactate (mmol L− 1)5.69 ± 1.384.44 ± 1.875.48 ± 2.08NDND No differencea, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Perioperative biochemical variables change in Group II ND No difference a, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05) *: p ˂ 0.05 Post-op 12 h CK (400.10 ± 55.25 U/L) and glucose (165.00 ± 40.45 mg/dL) levels of Group III increased compared to pre-op 0 h (182.57 ± 53.46 U/L and 61.00 ± 8.10 mg/dL, respectively). Creatine decreased in the post-op 6 h (1.07 ± 0.05 mg/dL) compared to the pre-op 0 h (1.63 ± 0.10 mg/dL) and maintained this level in the post-op 12 h (1.36 ± 0.07 mg/dL) (p ˂ 0.05) (Table 5). There was no statistical difference in the concentration of Ca, Mg, Na, P, and TP in the three groups at three time point measurements (p > 0.05). Table 5Perioperative biochemical variables change in Group IIIVariablesTime points P Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)4.39 ± 1.283.86 ± 1.174.10 ± 1.24NDALP (U L− 1)56.29 ± 8.10b42.00 ± 7.08a47.03 ± 6.55a*ALT (U L− 1)32.43 ± 4.54a62.50 ± 7.14b21.69 ± 8.04a*AST (U L− 1)37.86 ± 2.0027.50 ± 5.2527.02 ± 7.80NDCK (U L− 1)182.57 ± 53.46a208.50 ± 34.54a400.10 ± 55.25b*Cl (mmol L− 1)116.57 ± 12.50118.00 ± 18.71113.09 ± 19.42NDCreatine (mg dL− 1)1.63 ± 0.10 b1.07 ± 0.05a1.36 ± 0.07 a*Glucose (mg dL− 1)61.00 ± 8.10 a184.50 ± 51.70b165.00 ± 40.45b*L-lactate (mmol L− 1)6.62 ± 2.227.58 ± 2.857.15 ± 3.95NDND No differencea, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Table 5Perioperative biochemical variables change in Group IIIVariablesTime points P Pre-op 0 hPost-op 6 hPost-op 12 hALB (g dL− 1)4.39 ± 1.283.86 ± 1.174.10 ± 1.24NDALP (U L− 1)56.29 ± 8.10b42.00 ± 7.08a47.03 ± 6.55a*ALT (U L− 1)32.43 ± 4.54a62.50 ± 7.14b21.69 ± 8.04a*AST (U L− 1)37.86 ± 2.0027.50 ± 5.2527.02 ± 7.80NDCK (U L− 1)182.57 ± 53.46a208.50 ± 34.54a400.10 ± 55.25b*Cl (mmol L− 1)116.57 ± 12.50118.00 ± 18.71113.09 ± 19.42NDCreatine (mg dL− 1)1.63 ± 0.10 b1.07 ± 0.05a1.36 ± 0.07 a*Glucose (mg dL− 1)61.00 ± 8.10 a184.50 ± 51.70b165.00 ± 40.45b*L-lactate (mmol L− 1)6.62 ± 2.227.58 ± 2.857.15 ± 3.95NDND No differencea, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05)*: p ˂ 0.05 Perioperative biochemical variables change in Group III ND No difference a, b: The difference between groups with different letters on the same line is statistically significant (p ˂ 0.05) *: p ˂ 0.05 Discussion It is stated that the immune response after surgery is multifactorial. Thus, it is believed that the cytokine response changes in the first hour after anesthesia and surgery [20]. The early pro-inflammatory response involves increased concentrations of Th1 cytokines (TNF-α, INF-γ, IL-2, and IL-12). During the acute phase, surgical trauma, elevated cortisol and catecholamine levels, and suppression of cellular immunity stimulate the release of Th2 cytokines (IL-4, 5, 6, 10, and 13) [21–23]. Our study showed that TNF-α concentration increased in Groups I and II at the 6th hour postoperative time point but decreased again by the 12th hour. In contrast, IL-2 concentration decreased after the operation compared to pre-operative levels in all groups. Marinov et al. [24] found that IL-6 and TNF-α concentrations decreased in the first hour in the inhalation anesthesia group, while both interleukins decreased significantly at the 120th minute in the butorphanol group. The suppression of IL-6 and TNF-α during the perioperative period indicates a reduction in the pro-inflammatory immune response. At 24 h, IL-6 and TNF-α concentrations were significantly lower than baseline levels in the butorphanol group. In our study, among the Th1 cytokines, only IFN-γ in Group II showed a parallel change with TNF-α, a finding consistent with Marinova et al. [24]. Isoflurane anesthesia has been shown to directly increase IL-1, IL-6, and TNF-α concentrations in rats [25]. In humans, it has also been proven that IL-6 production is stimulated during isoflurane and sevoflurane anesthesia [26]. These findings correlate with the decrease in IL-2 that we measured in our study, further supporting the observed trends. On the other hand, the findings in our study concerning TNF-α contradict those in previous studies that reported suppression of TNF-α concentrations. While TNF-α is suppressed in those studies, our results showed an initial increase in both Groups I and II, followed by a decrease back to pre-operative levels within 12 h. It is thought that this may be due to the suppression of anti-inflammatory cytokines. No perioperative changes were observed in IL-2 cytokine concentrations in both the ketamine and propofol anesthesia groups applied to dogs. It is reported that no change in IFN-γ in the ketamine group, while a decrease in postoperative IFN-γ concentration was detected in the propofol group [27]. In a study investigating IL-8 and IL-10, these Th2 cytokines showed no concentration changes in both groups administered ketamine and propofol [27]. In our study, IL-5, a Th2 cytokine, remained unchanged across the three groups, while IL-4 and IL-10 concentrations decreased after the operation compared to pre-operative levels. It is suggested that the key role of Th2 cytokines is to suppress the production of Th1 cytokines. Therefore, the decrease in Th2 cytokines after surgery may be related to the increase in TNF-α, which is an acute phase cytokine. In dogs receiving isoflurane anesthesia, IL-10 concentration increased at the 2nd hour after surgery but decreased at later time points. Conversely, no change in IL-10 concentration was observed in dogs administered propofol [5]. However, in human medicine [21], total intravenous anesthesia has shown higher IL-10 concentrations compared to inhalation anesthesia in the perioperative period. In a study conducted in cats [24], IL-6 concentration increased in the butorphanol group at the 30th minute after surgery and again at later time points. In our study, IL-4 levels in Groups I and II decreased post-operatively but increased subsequently. Evaluating biochemical variables alongside clinical findings in disease diagnosis and treatment prognosis reveals important information. Nowadays, many different anesthesia protocols used in surgeries have been developed. However, xylazine/ketamine and propofol/isoflurane combinations are frequently utilized in animal surgery. Anesthetic drugs can cause damage to various organ systems, particularly the heart, liver, and urinary system, in both animals [28–30] and humans [31]. It is also well known that surgery is associated with stress, pain, and hemodynamic changes [29] reported that after propofol anesthesia in cats, ALT and AST levels increased, peaking after 72 h. In our study, we noted a decrease in these variables after surgery in Group I, though these changes were minimal and remained within normal ranges [32]. While ALT levels in dogs did not change perioperatively, AST levels increased in the first hours post-surgery, returning to normal ranges later on [33]. Different drug combinations used in cats have been shown to affect echocardiographic and hematological variables, and it is recommended to take these changes into account when interpreting study results. The combination of midazolam, butorphanol, and ketamine caused mild cardiovascular changes and produced acceptable alterations. Both protocols with dexmedetomidine provided excellent sedation and recovery but resulted in greater cardiovascular and hematological changes [28]. Metabolites of tramadol, primarily excreted by the kidneys, can lead to toxicity in these organs. Acute and chronic tramadol use has been linked to increases in ALT, AST, creatinine, and urea activity, inducing hepatotoxicity and nephrotoxicity in rodents [34]. However, it was concluded that the doses and frequency of application minimally changed hematological, biochemical, and urinary variables in cats, without exceeding normal ranges [30]. Our study showed that glucose concentration increased in all groups after surgery. This finding suggests that the increase is due to reduced food intake before surgery and the resumption of feeding afterward. Meloxicam (an NSAID) may modulate inflammatory mediator production and could influence absolute cytokine concentrations. However, meloxicam was administered to all groups under the same timing and dose, making it unlikely to explain between-protocol differences, although it may have attenuated the overall magnitude of postoperative inflammatory responses. Postoperative pain monitoring indicated comparable and clinically mild scores in all groups, and no rescue analgesia was required. These observations suggest that analgesia was clinically adequate across protocols and that cytokine differences are unlikely to be primarily driven by uncontrolled nociception. Limitations This study has limitations. First, the sample size (n = 8 per group) limits precision and may reduce sensitivity for small effects. Second, biomarkers were monitored only up to 12 h, and later immune dynamics (24–72 h) were not captured. Third, the elective ovariectomy model in healthy cats may not generalize to more invasive procedures or clinically compromised patients. Fourth, although postoperative pain and recovery were clinically evaluated and recorded and no cat required rescue analgesia, a validated feline pain scoring system was not applied; thus, pain-related confounding cannot be fully excluded and formal quantitative correlations between pain intensity and biomarker trajectories cannot be performed. Finally, handling stress and peri-operative NSAID administration may influence immune biomarkers; although standardized across groups, these factors may affect absolute levels. Within the first 12 hours after elective ovariohysterectomy in healthy cats, the different anesthetic protocols were associated with distinct neuroendocrine and cytokine profiles. In our cohort, the xylazine–isoflurane protocol was associated with smaller short-term deviations in selected cytokines compared with the other protocols. Given the short follow-up and absence of validated pain scoring, these findings should be interpreted as preliminary and hypothesis-generating; larger studies with standardized pain assessment and longer monitoring are needed to clarify clinical implications. Conclusions In this cytokine study, we investigated how various anesthetic agents affect cytokines known to have indirect effects on the immune system. It was determined that Group II had a strong impact on cytokines. The combination of xylazine and ketamine significantly affected immune system cells and cytokines. Changes were observed in all cytokines investigated in this study, except for IL-5. Therefore, it can be concluded that the anesthetic protocol with the least impact on cytokine changes in the postoperative period is the combination of xylazine and isoflurane. The data obtained from this study should be taken into consideration when choosing anesthesia protocols in order to anticipate potential reactions in the defense system during the postoperative period. Furthermore, although there are changes in biochemical variables in healthy cats, no changes affecting their overall condition were observed. These findings indicate that anesthesia protocols differentially influence perioperative immune responses in cats. The xylazine–ketamine combination induced the greatest cytokine fluctuations, whereas the xylazine–isoflurane protocol best preserved Th1/Th2 balance. Therefore, xylazine–isoflurane anesthesia may be preferable for minimizing postoperative immune modulation in feline ovariohysterectomy. Overall, the three anesthetic protocols were associated with distinct short-term cytokine patterns within 12 h after elective ovariohysterectomy. Given the sample size and short follow-up, these findings should be interpreted as preliminary and hypothesis-generating; larger studies with standardized pain scoring and longer monitoring are required before making protocol recommendations.