Dexmedetomidine Dose-Dependent Modulation of Lung Function and AQP1 Expression During One-Lung Ventilation in Adults Undergoing Thoracoscopic Surgery

Introduction

With the advancement of thoracoscopic techniques for complex thoracic surgeries, one-lung ventilation (OLV) has become increasingly prevalent. OLV optimizes surgical exposure and effectively isolates the lungs, protecting healthy tissue from contamination [1]. However, prolonged OLV may contribute to acute lung injury (ALI), characterized by inflammatory cell infiltration, increased pulmonary capillary permeability, and impaired gas exchange, often manifesting as pulmonary edema and hypoxemia [2]. ALI is a significant cause of morbidity and mortality following thoracic surgery, adversely impacting patient prognosis [3,4].

The inflammatory cascade during OLV is a recognized driver of gas exchange disturbances in ALI [5]. Within this response, tumor necrosis factor-α (TNF-α) increases pulmonary capillary permeability by activating various inflammatory mediators. This leads to pulmonary edema, impairing gas exchange and reducing lung compliance [6]. Alveolar neutrophil levels correlate with both pulmonary capillary permeability and the severity of hypoxemia [5]. Interleukin-8 (IL-8), a key chemokine, activates neutrophils and promotes their accumulation within the lungs, thereby exacerbating inflammation and lung injury, ultimately resulting in pulmonary dysfunction [7]. While these processes are well-studied, their interplay with alveolar fluid regulation requires further elucidation.

Pulmonary edema remains a critical pathological hallmark of ALI [5]. Under physiological conditions, efficient alveolar fluid clearance (AFC) is essential for maintaining gas exchange. Aquaporin-1 (AQP1), a highly selective water channel, facilitates water transport across alveolar and capillary membranes [8]. Decreased AQP1 expression impedes fluid transport in alveolar capillaries and epithelial cells, hindering the clearance of interstitial and alveolar fluid, which is closely associated with the pathogenesis and progression of pulmonary edema [9–11]. Experimental evidence suggests that inflammation can downregulate AQP1 expression, potentially impairing edema resolution [12]. However, this relationship appears to be context-dependent and is incompletely defined in human clinical settings [13]. Notably, inflammatory mediators like TNF-α exhibit complex interactions with aquaporins [14]. Recent in vitro evidence using human bronchial epithelial cells (BEAS-2B) demonstrates that TNF-α induces a dose-dependent downregulation of AQP1 expression at both mRNA and protein levels [15], aligning with established mechanisms linking inflammation to AFC impairment [16]. These findings suggest that inflammation-induced AQP dysregulation can contribute to fluid imbalance in lung injury [13,14], but its impact on OLV-related lung injury remains to be verified.

Dexmedetomidine (DEX), a highly selective α-2 adrenergic receptor agonist, provides sedation, analgesia, antisympathetic effects, and organ protection. Previous studies indicate that DEX ameliorates intrapulmonary shunting and hypoxemia by inhibiting inflammatory responses and reducing oxidative stress induced by endotoxemia or ischemia-reperfusion, thereby exerting pulmonary protective effects [17–19]. However, data on the impact of DEX on lung function during OLV and its underlying mechanisms remain limited [20]. Although rodent models suggest DEX can attenuate lung injury and pulmonary edema by upregulating AQP1 [21,22], a direct mechanistic link between DEX and AQP1 regulation in humans remains unestablished.

Based on these insights, we hypothesized that DEX can mitigate OLV-induced ALI through concurrent modulation of inflammatory cytokines and attenuation of AQP1 suppression. Therefore, this trial aimed to evaluate the dose-dependent effects of DEX on lung function during OLV and explore the relationship between AQP1 expression and inflammation in human lung tissue.

Material and Methods

STUDY DESIGN AND ETHICS APPROVAL:

This prospective, double-blind, single-center, randomized controlled trial was conducted at Nanxishan Hospital, Guangxi Zhuang Autonomous Region, China. The study protocol was approved by the Nanxishan Hospital Ethics Committee (Approval No. 2016NXSYYEC-007) and complied with the Declaration of Helsinki and CONSORT guidelines. Written informed consent was obtained from all participants prior to enrollment.

PATIENTS:

Patients scheduled for elective thoracoscopic radical resection of lung cancer between January 2018 and March 2019, requiring intraoperative OLV exceeding 2 hours, were eligible. Inclusion criteria were: American Society of Anesthesiologists (ASA) physical status I or II, age 40–70 years, and body mass index (BMI) 18–30 kg/m². Exclusion criteria were: preoperative tumor radiotherapy or chemotherapy; immune or endocrine disorders; hepatic or renal dysfunction; preoperative corticosteroid use; recent respiratory infection or chronic lung disease; sick sinus syndrome or atrioventricular block; or preoperative anemia (hemoglobin <90 g/L).

RANDOMIZATION AND BLINDING:

Participants were randomly assigned to 1 of 3 groups using a computer-generated sequence. Allocation concealment was ensured via sealed, opaque envelopes opened on the morning of surgery. The 3 groups were: Control (Group C, saline infusion), DEX 0.3 μg/kg/h (Group D1), and DEX 0.5 μg/kg/h (Group D2) (allocation ratio 1: 1: 1). An independent anesthesia nurse prepared study infusions according to allocation. Patients, anesthesiologists, surgeons, nursing staff, and outcome assessors remained blinded throughout the study.

ANESTHETIC MANAGEMENT AND INTERVENTION:

Standard monitoring upon operating room arrival included non-invasive blood pressure (NBP), pulse oximetry (SpO₂), electrocardiogram (ECG), end-tidal carbon dioxide (EtCO₂), and bispectral index (BIS). Invasive arterial pressure monitoring was established via radial artery catheterization. Anesthesia was induced with intravenous midazolam (0.05 mg/kg), etomidate (0.3 mg/kg), sufentanil (0.5 μg/kg), and cisatracurium (0.3 mg/kg). A double-lumen bronchial tube was placed under fiberoptic bronchoscopy guidance. Mechanical ventilation (Mindray A5 anesthesia machine, China) was initiated in volume-controlled mode with the following settings: tidal volume (VT) 8 mL/kg, inspiratory: expiratory ratio (I: E) 1: 2, respiratory rate (RR) 12–14 breaths/min, and fraction of inspired oxygen (FiO₂) 100%. The bronchial tube position was re-confirmed bronchoscopically after lateral positioning. During OLV, VT was reduced to 6 mL/kg and the positive end-expiratory pressure (PEEP) was 3 cmH₂O; RR was adjusted to maintain EtCO₂ at 35–45 mmHg. Central venous access was established via the right internal jugular vein. Following induction, Group D1 and D2 received a DEX loading dose (0.5 μg/kg over 10 min), followed by continuous infusions of 0.3 μg/kg/h or 0.5 μg/kg/h, respectively, until 30 minutes before the end of the operation. Group C received an equivalent volume of 0.9% saline. Anesthesia maintenance consisted of intravenous remifentanil (0.1–0.3 μg/kg/min), propofol (4–10 mg/kg/h), and intermittent cisatracurium (0.08 mg/kg). The intraoperative BIS value was maintained within the range of 40–60. Assisted ventilation techniques were applied to patients with SpO₂ less than 90% after OLV, and these patients were withdrawn from the final analysis. Under the premise of satisfactory depth of anesthesia, intraoperative vasoactive drugs were used to maintain hemodynamic stability. If hypotension (mean arterial pressure [MAP] <65 mmHg or a decline of more than 20% of the baseline) or bradycardia (heart rate [HR] <50 beats/min) occurred, ephedrine 5–10 mg or atropine 0.3–0.5 mg, respectively, was administered.

OUTCOME MEASURES:

Preoperative and intraoperative patient characteristics were recorded. Intraoperative hypotension and bradycardia events were documented. Anesthesia records underwent a weekly audit, and deviations exceeding 10% led to exclusion.

Serum TNF-α and IL-8 concentrations (radioimmunoassay kits, Furui Bio, Beijing, China) were measured at 5 time points: pre-DEX administration (Before DEX), 60 min (OLV_{60 min}), 90 min (OLV_{90 min}), 120 min (OLV_{120 min}) after OLV initiation, and 30 min after reinstitution of two-lung ventilation (ReTLV_{30 min}).

AQP1 expression, serving as an indicator of pulmonary interstitial water permeability, was assessed by immunohistochemistry (IHC) performed on lung tissue samples. Two histologically normal specimens (1×1×1 cm3 each) were collected per patient: 1) Pre-OLV: ≥5 cm from tumor margin in planned resection area; 2) Post-resection: ≥5 cm from margin in isolated lungs. This sampling distance (>2×NCCN-recommended surgical margin [≥2 cm] [23]) ensured non-neoplastic tissue from routinely discarded pathological material, posing no added risk. Tissues were fixed in 10% neutral buffered formalin (24 h), paraffin-embedded, and sectioned (3 μm). IHC used rabbit anti-human AQP1 primary antibody (Abcam #ab168387, 1: 300) with brown diaminobenzidine (DAB) chromogen; membrane staining was considered positive. Two blinded, board-certified pathologists independently scored slides. Inter-rater reliability was assessed using Cohen’s kappa (κ). Discrepancies were resolved by consensus. Five random high-power fields (HPF, 400x) per section were evaluated. The percentage of positive cells (0: 0–25%; 1: 26–50%; 2: 51–75%; 3: 76–100%) and staining intensity (0: none; 1: light yellow; 2: brown; 3: dark brown) were scored. The total IHC score (range 0–9) was calculated as (percentage score x intensity score); specimens were classified as low (≤4) or high (>4) AQP1 expression [24]. The requirements of ethics approval for waste tissue use and individual patient consent were waived under relevant regulations.

Radial arterial blood samples were drawn for blood gas analysis at the specified time points. Dynamic lung compliance (Cdyn) was recorded from the anesthesia machine. The respiratory index (RI) and oxygenation index (OI) were calculated as:

P(A-aDO₂) – alveolar-arterial blood oxygen partial pressure difference, PaCO₂ – partial arterial pressure of CO₂, Pb – barometric pressure (760 mmHg), pH₂O – partial pressure of water vapor (47 mmHg), RQ – respiratory quotient (0.8).

STATISTICAL ANALYSIS:

Sample size estimation was based on IHC scores of post-resection lung tissue AQP1 expression. Preliminary data yielded mean±SD scores: Group C (2.4±0.6), Group D1 (2.8±0.8), Group D2 (3.4±0.9). Using PASS 2021 (β=0.2, α=0.05, two-tailed), 17 patients per group were required. Accounting for a 15% dropout rate, 20 patients were enrolled per group.

Continuous data are presented as mean±standard deviation (SD) following normality confirmation (Shapiro-Wilk test). Inter-group comparisons utilized one-way ANOVA with Bonferroni post hoc tests. Within-group changes over time were analyzed using repeated-measures ANOVA; Mauchly’s test assessed sphericity (Greenhouse-Geisser correction applied if violated; P<0.05). Bonferroni adjustment was applied for multiple comparisons. Cohen’s κ coefficient (95% CI via 10 000 bootstraps) quantified IHC scoring agreement. Categorical data are presented as frequencies (%) and compared using the chi-square or Fisher’s exact tests, as appropriate. Pearson’s correlation analysis assessed relationships between continuous variables. Missing data were handled using multiple imputations. All analyses were performed using SPSS 26.0 (IBM Corp., USA), with P<0.05 considered statistically significant.

Results

CHARACTERISTICS OF PARTICIPANTS:

The CONSORT flow diagram details participant screening, allocation, and analysis (Figure 1). Sixty patients (20 per group) completed the study. No significant inter-group differences existed in baseline demographics (age, sex, BMI), ASA physical status, surgical site, operation duration, OLV duration, blood loss, or intraoperative fluid intake (all P>0.05). The incidence of intraoperative bradycardia or hypotension was also comparable across groups (P>0.05) (Table 1).

COMPARISON OF SERUM TNF-α AND IL-8 CONCENTRATIONS:

Serum TNF-α and IL-8 levels increased significantly from baseline in all groups starting at OLV60 min (P<0.05). Compared to Group C (control), Group D2 (DEX 0.5 μg/kg/h) exhibited significantly lower TNF-α and IL-8 concentrations starting at OLV60 min (all P<0.05). Group D1 (DEX 0.3 μg/kg/h) showed significant reductions versus Group C only at OLV60 min and OLV90 min (P<0.05). Furthermore, Group D2 demonstrated significantly lower TNF-α concentrations than Group D1 at OLV120 min and ReTLV30 min (P<0.05), and significantly lower IL-8 concentrations than Group D1 starting at OLV90 min (P<0.05) (Tables 2, 3).

COMPARISON OF AQP1 EXPRESSION AND IMMUNOHISTOCHEMISTRY SCORING:

Inter-rater agreement for AQP1 immunohistochemistry (IHC) scoring was excellent (Cohen’s κ=0.82, 95% CI 0.75–0.89, P<0.001). AQP1 immunoreactivity localized predominantly to alveolar capillary endothelial cell membranes, visualized as light yellow to dark brown granular staining (Figure 2). Compared to pre-OLV levels, post-resection AQP1 IHC scores were significantly lower in all groups (Group C, D1, and D2; P<0.05), accompanied by visibly lighter staining intensity. However, the magnitude of reduction differed significantly: Group D2 maintained significantly higher post-resection AQP1 IHC scores and staining extent than Group C and Group D1 (P<0.05) (Figure 2, Table 4).

COMPARISON OF INTRAOPERATIVE RI, OI, AND CDYN:

During the transition from OLV to ReTLV, the RI increased, and the OI and Cdyn significantly decreased in all 3 groups (P<0.05). Compared with Group C, the RI values of Groups D1 and D2 were significantly lower starting at OLV60 min, the OI was higher in Group D1 at ReTLV30 min and in Group D2 from OLV60 min to ReTLV30 min, and Cdyn was higher at OLV120 min in Group D1 and OLV90, 120 min in Group D2 (P<0.05). Compared to Group D1, Group D2 had significantly lower RI values from OLV90 min to ReTLV30 min and significantly higher values of the OI (starting at OLV60 min) and Cdyn (at OLV90 min and OLV120 min) (P<0.05) (Tables 5–7).

CORRELATIONS BETWEEN AQP1 AND TNF-α/IL-8:

Pearson correlation analysis revealed significant inverse relationships between post-resection lung tissue AQP1 IHC scores and inflammatory cytokine levels measured at OLV90 min (Figure 3). AQP1 scores demonstrated moderately negative correlations with both cytokine concentrations: TNF-α (r=−0.672, P<0.001) and IL-8 (r=−0.744, P<0.001).

Discussion

This randomized controlled trial evaluated the dose-dependent effects of DEX on lung function and AQP1 expression during OLV. Dosing regimens (0.5 μg/kg loading dose followed by 0.3 or 0.5 μg/kg/h maintenance) were selected based on established efficacy and safety profiles [18,25,26]. Crucially, these doses did not increase the incidence of bradycardia or hypotension, confirming their clinical safety. Our findings demonstrate that DEX significantly attenuates OLV-induced physiological disturbances in a dose-dependent manner, evidenced by improved intraoperative lung mechanics (increased OI and Cdyn, decreased RI), reduced inflammatory markers (TNF-α, IL-8), and partial preservation of AQP1 expression. These results support our hypothesis that DEX confers pulmonary protection by concurrently modulating inflammatory cytokine and alveolar fluid regulation pathways.

OLV induces lung collapse, surgical compression, and subsequent ischemia-reperfusion upon re-expansion, triggering the release of inflammatory mediators like TNF-α and IL-8 from alveolar epithelium [2]. Preclinical evidence indicates DEX mitigates ventilator-induced lung injury [27] and modulates inflammation in OLV models [28]. Clinically, perioperative DEX attenuates the OLV-induced inflammatory response and improves oxygenation, particularly reducing TNF-α and IL-8 levels at later OLV time points [29]. Our data corroborate these findings, showing DEX dose-dependently suppressed TNF-α and IL-8 production during OLV, highlighting its anti-inflammatory and anti-chemotactic properties as key protective mechanisms.

AQP1, predominantly localized to alveolar capillary endothelia, mediates critical water transport functions under osmotic gradients, influencing pulmonary interstitial fluid dynamics and vascular permeability [9,30,31]. Its downregulation is associated with edema formation in ventilator-induced lung injury (VILI) [9,32,33], and Lin et al specifically demonstrated reduced AQP1 expression correlating with OLV duration and injury severity [34]. Consistent with prior reports [31,32,34], we observed decreased AQP1 expression in non-ventilated lungs after OLV. Notably, DEX treatment, particularly at the higher dose (0.5 μg/kg/h), resulted in modest but significant preservation of AQP1 expression compared to controls, representing partial attenuation of OLV-induced suppression rather than pathological upregulation. While this AQP1 preservation may secondarily support fluid homeostasis, its contribution appears subsidiary to DEX’s primary anti-inflammatory action. The role of aquaporins in lung injury is complex and context-dependent [30,35]. Although AQP1 facilitates alveolar fluid clearance physiologically [9,30], its function during inflammation remains incompletely defined. Concerns exist that overexpression might exacerbate edema in inflammatory settings [36]; our observed preservation levels likely mitigate this risk. Furthermore, DEX-induced hemodynamic stabilization [19,20], potentially improving pulmonary perfusion and reducing shunt fraction (Qs/Qt), may also contribute indirectly to edema resolution, though this mechanism was not directly assessed here.

Supporting the link between inflammation and AQP1 dysregulation, meta-analyses confirm DEX suppresses TNF-α and IL-8 during OLV [29], and studies demonstrate inflammatory mediators like TNF-α downregulate AQP expression [37–39]. In our study, serum TNF-α and IL-8 levels at OLV90 min showed significant moderate negative correlations with post-resection lung tissue AQP1 expression (r=−0.672 and r=−0.744, respectively; both P<0.001). While these associations align with proposed inflammation-mediated AQP1 suppression [21,37–39], caution is warranted. These represent temporal correlations at a single intraoperative phase (OLV90 min), not proven causation. Although DEX groups exhibited higher relative AQP1 scores, absolute expression remained modest (≤4.2 on a 0–9 scale). The observed correlations could reflect shared responsiveness to DEX intervention rather than a direct cytokine-mediated regulation of AQP1 expression. Therefore, we postulate that DEX’s anti-inflammatory effects can partially mitigate inflammation-driven AQP1 downregulation. However, the modest absolute changes (Δ≤1.4 points) and absence of formal mediation analysis preclude definitive claims about AQP1 as a primary effector mechanism in humans. Although rodent studies suggest that AQP1 upregulation mediates the protective effects of DEX [21], our human data support only an association between preserved AQP1 expression and clinical improvement, warranting future mechanistic validation.

OLV-induced pulmonary dysfunction encompasses impaired ventilation (reduced compliance) and gas exchange (hypoxemia). Indices like OI and RI effectively reflect lung oxygenation and diffusion capacity [40]. In our study, prolonged OLV progressively decreased OI and Cdyn while increasing RI across all groups, indicating worsening gas exchange and lung mechanics, consistent with prior reports linking OLV duration to injury severity [34]. DEX administration significantly improved these parameters in a dose-dependent manner, suggesting enhanced lung diffusion and reduced ventilation-perfusion mismatch. This protective effect likely stems from attenuation of inflammatory cascades and mitigation of diffuse interstitial edema. Importantly, improvements in RI and OI persisted until 30 minutes after reinstitution of two-lung ventilation (ReTLV30 min), indicating sustained protection encompassing the critical re-expansion phase. However, 2 key limitations constrain clinical extrapolation: (1) The absence of standardized postoperative pulmonary assessments (eg, arterial blood gases, spirometry at 24–72 h) precludes evaluation of DEX’s impact on clinically relevant outcomes like atelectasis, pneumonia, or ARDS; and (2) Whether intraoperative benefits translate to sustained postoperative protection requires dedicated investigation.

Beyond the lack of postoperative assessment, several limitations warrant acknowledgment: (1) As a single-center trial, generalizability may be limited to comparable surgical populations and settings; (2) The semi-quantitative nature of the AQP1 immunohistochemistry scoring system, though validated and demonstrating excellent inter-rater reliability, has inherent precision limits compared to quantitative methods (eg, Western blot, qPCR); and (3) The specific molecular pathways underlying DEX’s potential modulation of AQP1 expression in human lung tissue remain unexplored and represent an important avenue for future research.

Conclusions

In summary, dexmedetomidine improves intraoperative lung function during OLV in a dose-dependent manner, primarily through potent anti-inflammatory effects that reduce TNF-α and IL-8 levels. Partial preservation of AQP1 expression may secondarily contribute to fluid balance, although its role appears subsidiary. The significant inverse correlations observed between inflammatory markers and AQP1 expression warrant further mechanistic investigation to elucidate potential causal links. Higher-dose DEX (0.5 μg/kg/h) demonstrated superior efficacy in attenuating OLV-induced lung injury.