Effects of various hypoxic preconditioning strategies on arterial oxygenation in patients undergoing one-lung ventilation during thoracoscopic surgery: a randomized clinical trial

Abstract

Background

One-lung ventilation (OLV) during thoracoscopic right lung resection can lead to challenges in maintaining arterial oxygenation and postoperative pulmonary complications. Hypoxic preconditioning (HPC) may improve outcomes in this setting. This study aimed to investigate the effects of different HPC strategies on arterial oxygenation during OLV and the incidence of postoperative pulmonary complications.

Methods

This randomized controlled trial enrolled patients undergoing elective thoracoscopic right lung resection at the First Affiliated Hospital of Dalian Medical University from March 2022 to August 2022. Participants were randomly assigned to three groups in a 1:1:1 ratio: the I-HPC group receiving intermittent HPC, the C-HPC group receiving continuous HPC, and the RV group receiving regular two-lung ventilation (control group). Arterial blood gas analysis was conducted at five time points: at baseline prior to oxygen inhalation (T0), before the opening of the pleura (T1), and at 5 min (T2), 15 min (T3), and 30 min (T4) after the initiation of OLV. The primary outcome was the arterial oxygen partial pressure (PaO₂) throughout the OLV period.

Results

Baseline arterial PaO₂ at T0 did not differ significantly among the three groups. The I-HPC group showed significantly higher PaO₂ (mmHg) at T3 (214.27 ± 74.38 vs. 167.80 ± 65.68, P = 0.015) and T4 (221.80 ± 74.67 vs. 177.03 ± 70.94, P = 0.023) compared to the control group. The C-HPC group also had significantly higher PaO₂ at T3 (217.35 ± 76.10 vs. 167.80 ± 65.68, P = 0.009) and T4 (220.81 ± 78.58 vs. 177.03 ± 70.94, P = 0.023) relative to the control group. No significant differences were found among the groups regarding intraoperative oxygen saturation, arterial blood gas values, respiratory and circulatory indicators, vasoactive drug usage, or postoperative pulmonary complications (all P > 0.05).

Conclusions

Hypoxic preconditioning of the nondependent lung prior to pleural incision may enhance arterial oxygenation during OLV, suggesting that HPC could serve as a supportive measure for patients undergoing elective thoracoscopic lung resection.

Trial registration

This study was registered on March 27th, 2022, in the Chinese Clinical Trial Registry (ChiCTR2200058026), prior to enrollment of the first participant, and no protocol changes were made after trial initiation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12890-026-04314-y.

Background

One-lung ventilation (OLV), during which a bronchial catheter is inserted to provide ventilation to only one lung, is a widely used method in thoracic surgery. When the ventilation to one lung is stopped, it causes an obligatory intrapulmonary shunt that can result in low oxygen levels. Hypoxemia (defined as a pulse oxygen saturation < 95% or an arterial oxygen pressure < 69 mmHg with pure oxygen inhalation) is a common complication during OLV, occurring in approximately 5–10% of cases [1, 2]. Hypoxemia may be caused by nondependent lung collapse, which still exhibits blood perfusion and contributes to non-physiologic intrapulmonary shunt [3].

Typically, when alveolar oxygen tension decreases, hypoxic pulmonary vasoconstriction (HPV) occurs, resulting in increased pulmonary vascular resistance and redirection of blood flow to well-ventilated lung regions, thereby alleviating the ventilation/perfusion ratio mismatch [4]. This hypoxia-induced pulmonary vasoconstriction partially alleviates intrapulmonary shunt and improves arterial oxygenation during OLV [5]. Related studies have suggested that similar treatments may also reduce oxidative stress, inflammatory substances, and other lung tissue damage, enhance the body’s antioxidant capacity, and improve postoperative lung function [6–9]. However, there is no consensus on how to implement hypoxic preconditioning (HPC) in clinical practice. Several researches have reported that intermittent hypoxia seems to enhance hypoxia-induced pulmonary vasoconstriction [10, 11], which could improve arterial oxygenation during OLV and reduce the incidence of hypoxemia. Meanwhile, some studies have suggested that the hypoxic pulmonary vasoconstriction response reaches its maximum after the first hypoxic stimulus and does not increase with subsequent repetitive hypoxic stimuli [12–14]. In clinical trials, there is currently no research indicating differences between repeated hypoxic stimuli and single continuous hypoxic stimuli.

Therefore, we conducted a prospective randomized controlled trial to evaluate the effects of different hypoxic preconditioning strategies applied to the nondependent lung on arterial oxygenation during one-lung ventilation in patients undergoing thoracoscopic right lung resection. We hypothesized that hypoxic preconditioning would improve arterial oxygenation during OLV compared with standard ventilation, and that intermittent and continuous hypoxic preconditioning strategies would have comparable effects on intraoperative oxygenation.

Methods

This study was conducted in accordance with the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines [15]. This single-center randomized controlled trial was approved by the Ethics Committee of the First Affiliated Hospital of Dalian Medical University on March 7, 2022 (PJ-KS-KY-2022-46(X)) and registered in the Chinese Clinical Trial Registry (ChiCTR2200058026) on March 27, 2022, prior to enrollment of the first participant on March 29, 2022; no protocol amendments or substantive changes were made after trial initiation. The study was conducted in accordance with the principles of the Declaration of Helsinki. All participants were informed about the study protocol and provided written informed consent to participate in the study.

Participants

Patients who underwent elective right lung resection via video-assisted thoracoscopic surgery at the First Affiliated Hospital of Dalian Medical University from March 2022 to August 2022 were enrolled in this study. Only right-sided procedures were included, as right lung collapse during one-lung ventilation is associated with a larger intrapulmonary shunt due to the greater proportion of pulmonary blood flow to the right lung [16], thereby providing a more sensitive physiological context for evaluating arterial oxygenation. The inclusion criteria were: (1) Age between 18 and 70 years old, regardless of gender; (2) Patients undergoing routine video-assisted thoracoscopic right lung resection; (3) American Society of Anesthesiologists (ASA) grade I-II; (4) Patients who demonstrate a clear understanding of the study and voluntarily participate, having signed an informed consent form themselves or through their family members. All enrolled patients underwent preoperative pulmonary function testing. The exclusion criteria: (1) History of previous lung surgery; (2) Regular use of angiotensin-converting enzyme inhibitors (ACEI) or angiotensin-II receptor blockers (ARB), calcium channel blockers, or nitrates; (3) Forced expiratory volume in one second/forced vital capacity (FEV1/FVC) < 60% on preoperative pulmonary function testing; (4) Poor general condition of the patient and/or deterioration of the disease that may affect the continuation of the study; (5) Other factors that may impede the patient’s participation in the study. Patients with chronic obstructive pulmonary disease, asthma, bronchiectasis, or a history of smoking were not excluded. The intraoperative withdrawal criteria for this study included the following: (1) excessive sputum production leading to low tidal volume; (2) significant displacement of the double-lumen bronchial tube; and (3) severe hypotension resulting in a pulse oxygen saturation (SpO₂) below 95%.

Interventions

The collection, grouping, and management of subjects were carried out simultaneously. All patients underwent preoperative relevant examinations. The intervention in this study was performed by the same person, and the surgery was performed by the same team. Patients were asked to fast from solid food for 6 h and clear liquids for 2 h before anesthesia. Upon entering the operating room, arterial catheterization was inserted into the non-dominant radial artery to monitor invasive arterial blood pressure. Arterial blood was sampled for blood gas analysis without oxygen inhalation to obtain the patient’s preoperative baseline values. Anesthesia induction was achieved by intravenous administration of midazolam 0.05 mg/kg, sufentanil 0.5 µg/kg, etomidate 0.3 mg/kg, and rocuronium 0.6 mg/kg. A suitable-sized double-lumen bronchial tube was selected for intubation, and the position of the tube was confirmed by a fiberoptic bronchoscope in the supine and left lateral positions. Anesthesia maintenance was achieved by continuous infusion of propofol at a rate of 4–8 mg/kg/h and remifentanil at a dose of 0.15–2 µg/kg/min with occasional rocuronium boluses administered. OLV was initiated immediately after opening the pleura, and two-lung ventilation was restored after muscle layer closure. The bronchial tube was removed after the surgery was completed, and the patient was transferred to the post-anesthesia care unit for further observation.

Mechanical ventilation was delivered using a volume-controlled mode with a fixed inspired oxygen fraction (FiO₂) of 1.0 throughout the entire intraoperative period, including two-lung ventilation, hypoxic preconditioning interventions, and OLV. The respiratory parameters were set as follows: oxygen concentration of 1.0, inspiratory-to-expiratory ratio of 1:2, oxygen flow rate of 2 L/min, tidal volume of 6–8 ml/kg of ideal body weight, calculated as follows: 50.0 + 0.905 × (height-152.4) for males and 45.5 + 0.905 × (height-152.4) for females. The respiratory rate was set between 12 and 20 breaths per minute, while positive end-expiratory pressure was established at 5 cm H2O, and end-tidal carbon dioxide levels were maintained between 35 and 45 mmHg. Ventilatory parameters were titrated within these predefined ranges to achieve the target end-tidal carbon dioxide and to maintain airway pressures within the prespecified limits. These respiratory parameters remained unchanged throughout the OLV period. Additionally, the peak airway pressure was maintained below 25 cm H2O during two-lung ventilation and below 30 cm H2O during OLV. If the target end-tidal carbon dioxide or airway pressure could not be achieved despite adjustments within the allowed ranges (e.g., due to excessive sputum production or bronchospasm), immediate clinical management was performed, including airway suctioning and bronchodilator administration as needed. If adequate ventilation still could not be achieved and the assigned study intervention was not completed, the patient was withdrawn intraoperatively according to the predefined withdrawal criteria and was not included in the final per-protocol analysis.

Patients were randomly assigned to three groups at a ratio of 1:1:1: Repetitive Intermittent Hypoxia group (I-HPC group), Continuous Hypoxia group (C-HPC group), and the control group (RV group). The intervention was performed in the left lateral position after the position of the bronchial tube was confirmed by a fiberoptic bronchoscope. The I-HPC group was subjected to 2 min of lung collapse followed by 2 min of ventilation of the collapsed lung, and this cycle was repeated four times. The C-HPC group was subjected to 14 min of lung collapse followed by 2 min of ventilation of the collapsed lung, and two-lung ventilation was restored after the intervention. Before each OLV, the non-ventilated lung was suctioned at -20 kPa for 10 s and then the right side of the Y-connector of the double-lumen bronchial catheter was clamped. The RV group maintained two-lung ventilation for 16 min. OLV was initiated immediately after opening the pleura in all three groups.

During surgery, if the patient’s systolic blood pressure dropped below 90 mmHg or mean arterial pressure (MAP) below 60 mmHg, ephedrine 6 mg or phenylephrine 30–50 µg was administered intravenously. If the patient’s blood pressure increased by more than 20% above the baseline value, urapidil 15–25 mg was administered intravenously. If the heart rate (HR) dropped below 45 beats/min, atropine 0.3–0.5 mg was administered intravenously. If the SpO₂ dropped below 95%, the non-operative lung was subjected to alveolar recruitment strategy using 15–20 s of lung inflation at a pressure of 20 cmH₂O or higher until the SpO₂ increased to 98%. If the SpO₂ falls below 90%, the surgeon was required to stop operation and perform manual ventilation to both lungs until the SpO₂ reaches 98% [6]. If the patient could not reach the expected end tidal carbon dioxide and airway pressure according to the set parameters, the trail would continue within oxygen saturation above 95%. The chest X-rays and blood routine were reviewed on the first day after operation. Monitoring the patient’s temperature twice daily before discharge, and identify the occurrence of postoperative pulmonary complications by reviewing medical records.

Outcomes

The primary outcome of the study was the arterial PaO₂ measured throughout the entire duration of OLV. The secondary outcomes included the incidence of patients experiencing a SpO2 below 95% during OLV, the number of cases necessitating the restoration of two-lung ventilation due to hypoxemia, and the frequency of postoperative pulmonary complications. These complications encompassed pulmonary infection, pleural effusion, atelectasis, insertion of a new chest tube or postoperative drainage for more than 7 days, postoperative respiratory failure, and pulmonary embolism occurring during hospitalization (including postoperative day 1).

Arterial blood samples were collected from patients at five distinct time points during the surgical procedure for blood gas analysis: at baseline prior to oxygen inhalation (T0), immediately before the opening of the pleura (T1), and at intervals of 5 min (T2), 15 min (T3), and 30 min (T4) following the initiation of one-lung ventilation. Additionally, respiratory and hemodynamic parameters were recorded at each time point, including peak airway pressure (PEAK), plateau pressure (PLAT), tidal volume (VT), heart rate, and mean arterial pressure. The administration of vasoactive drugs during the surgical procedure, along with the duration of chest tube drainage and the length of hospital stay (calculated from the day of surgery to the day of discharge) for each patient group, were also systematically documented.

This study further investigated the effects of two types of HPC on early postoperative pulmonary complications, which include pulmonary infection, pleural effusion, atelectasis, prolonged postoperative air leak, postoperative respiratory failure, and pulmonary embolism [17–19]. PPCs are defined as follows: 1. Pulmonary infection: (a) Imaging findings (new unilateral or bilateral pulmonary infiltrates), (b) Two pulmonary infection symptoms present (history of antibiotic treatment / fever > 38℃/ leukocyte count (< 4000, ≥ 12,000 WBC/mm3) / purulent secretions); 2. Pleural effusion: Chest radiograph demonstrating blunting of the costophrenic angle or loss of the sharp silhouette of the hemidiaphragm on the nonoperative side. 3. Atelectasis: Chest radiograph showing opacity, mediastinal, hilar, or hemidiaphragmatic shift toward the affected area, and hyperinflation adjacent to the atelectatic region;4. Prolonged postoperative air leak: Requiring insertion of a new chest tube or postoperative drainage tube > 7 days due to air leak; 5.Postoperative respiratory failure: Requiring mechanical ventilation for ≥ 24 h, reinsertion of endotracheal tube, or noninvasive ventilation; 6.Postoperative pulmonary embolism: Thrombus documented by pulmonary angiography or ventilation/perfusion radionuclide scan.

Sample size

The primary outcome was PaO2. Originally, the sample size for this study was estimated based on data from a pilot study (n = 18, 6 per group) that was strictly comparable to the main trial conditions. Based on the pilot data, the mean ± standard deviation of the lowest PaO₂ during OLV were 167.89 ± 53.54 mmHg, 224.50 ± 77.34 mmHg, and 228.57 ± 41.09 mmHg .To ensure robust detection of clinically relevant differences in the main trial, the minimum clinically important difference (MCID) between groups was prespecified as 40 mmHg, which was set based on previous studies[39,40], yielding an f-value of 0.34. To detect a difference of 40 mmHg in PaO2 during OLV with F-test for one-way ANOVA accepting an α error of 5% and a β error of 20%, the required study size was calculated as 90 patients using Power and Sample Size Calculation version 3.1.2 (Dupont WD and Plummer WD, Vanderbilt University, Nashville, TN).Allocated in a 1:1:1 ratio, each group would require at least 30 subjects. Anticipating a dropout rate of approximately 5%, the final target sample size was set at 32 subjects per group, totaling 96 subjects.

Randomization and blinding

Enrolled patients were assigned identification numbers based on the date of surgery. An assistant, who was not involved in the study, assigned participants to groups in a 1:1:1 ratio utilizing SPSS software. The order of assignment was concealed within serially numbered, sealed, and opaque envelopes. Another assistant, also independent of the study, was tasked with recruiting patients. Upon enrollment completion, the investigator opened the envelope and executed the experimental intervention in accordance with the group assignment. There were no modifications to the protocol after the initiation of the trial. The trial was conducted in a blinded manner for the patients and concluded once all participants were recruited. Patients who did not meet the inclusion criteria were excluded prior to randomization. After randomization, five patients experienced unexpected intraoperative events that led to interruption of data collection and did not receive the assigned study intervention. Consequently, no outcome data were available for these patients. An intention-to-treat analysis was therefore not feasible, and a per-protocol analysis was performed, including 91 patients in the final analysis.

Statistical analysis

For normally distributed continuous variables, data are presented as mean ± standard deviation (SD). One-way ANOVA was used for between-group comparisons, and the LSD test was used for post hoc pairwise comparisons. Additionally, to reduce the risk of inflated type I error, we added Holm Bonferroni for supplementary analysis. Repeated measures ANOVA was used for within-group comparisons. For non-normally distributed continuous variables, data were presented as median (interquartile range), and nonparametric tests were used for between-group comparisons. Categorical variables were presented as frequency or percentage, and the chi-squared test or Fisher’s exact test was used for between-group comparisons. All statistical analyses were performed using SPSS 26.0 software (IBM Corp., Armonk, NY, USA), and a two-sided p-value of < 0.05 was considered statistically significant.

Results

Initially, 96 patients were enrolled in the study. All eligible participants were randomly assigned to one of three groups, each consisting of 32 patients. During the surgical procedures, one patient was excluded due to excessive sputum leading to low tidal volume (in I-HPC group), two patients experienced significant displacement of the double-lumen bronchial tube (one in I-HPC group and the other one in RV group), and two patients presented with severe hypotension, resulting in SpO2 levels falling below 95% (one in C-HPC group and the other one in RV group), these five patients did not receive the study treatment and had no follow-up data. Ultimately, 91 patients who underwent right lung resection via thoracoscopy were included in the data analysis, comprising 30 patients in the I-HPC group, 31 patients in the C-HPC group, and 30 patients in the RV group (Fig. 1).

Fig. 1.

Flowchart of patient enrollment

Among the total patient population, there were 37 males and 54 females. The mean ages of the three groups were 56.63 ± 9.07, 56.06 ± 9.44, and 59.30 ± 7.35 years, respectively. No significant differences were observed among the three groups regarding sex, age, weight, height, Body Mass Index (BMI), lung function, blood gas levels at T0, type of surgery, duration of OLV, surgical procedure, anesthesia, postoperative chest tube drainage, and length of hospital stay (all P > 0.05). The general characteristics are presented in Table 1.

Table 1.

Baseline Characteristics of Patients

	RV (n = 30)	C-HPC (n = 31)	I-HPC (n = 30)	P
Gender (male/female)	11/19	13/18	13/17	0.857
Age (years)	56.63 ± 9.07	56.06 ± 9.44	59.30 ± 7.35	0.306
Height (cm)	166.13 ± 7.90	165.65 ± 7.06	163.83 ± 6.89	0.440
Weight (kg)	67.37 ± 10.80	66.42 ± 10.15	63.27 ± 9.06	0.258
Predicted body weight (kg)	59.43 ± 8.67	59.45 ± 8.28	57.83 ± 8.03	0.688
BMI (kg.m-2)	24.35 ± 3.09	24.15 ± 2.87	23.55 ± 2.99	0.561
ASA score(I/II)	13/17	15/16	12/18	0.802
Smoker	3	1	2	0.522
Preoperative pulmonary function
FEV1 (liter)	2.49 ± 0.55	2.54 ± 0.64	2.33 ± 0.52	0.352
% of predicted FEV1 (%)	89.43 ± 13.90	93.23 ± 15.05	92.40 ± 15.54	0.581
FVC (liter)	3.03 ± 0.66	3.08 ± 0.77	2.86 ± 0.60	0.439
% of predicted FVC (%)	89.73 ± 13.12	93.32 ± 12.85	91.70 ± 14.64	0.587
FEV1/FVC (%)	82.77 ± 9.36	82.52 ± 5.67	81.73 ± 7.25	0.862
Preoperative arterial blood gas
pH	7.42 ± 0.03	7.40 (0.07)	7.42 ± 0.03	0.295
PaO2(mmHg)	87.63 ± 9.97	84.90 ± 8.32	86.43 ± 6.37	0.444
PaCO2(mmHg)	38.50 ± 3.61	39.52 ± 3.96	38.80 ± 3.70	0.734
HCO3-(mmol/L)	25.52 ± 1.33	25.98 ± 1.85	25.96 ± 1.83	0.481
Hemoglobin (g/L)	121.17 ± 10.76	123.13 ± 14.72	122.73 ± 14.78	0.838
Surgical related data
OLV time (min)	90.00 (48)	74.00 (37)	103.00 (79)	0.052
Operation time (min)	103.50 (48)	90.00 (42)	114.50 (82)	0.051
Anesthesia time (min)	143.50 (38)	119.00 (45)	151.00 (85)	0.053
Postoperative chest tube drainage time(d)	3.00 (2)	3.00 (2)	3.50 (2)	0.399
Postoperative hospital stay (d)	4.50 (3)	5.00 (3)	5.00 (2)	0.613
Surgery type (wedge resection/segmentectomy/lobectomy)	11/3/16	15/3/13	10/3/17	0.785

Abbreviations: RV regular ventilation, C-HPC continuous hypoxic preconditioning, I-HPC intermittent hypoxic preconditioning, BMI body mass index, ASA American Society of Anesthesiologist, FEV1 forced expiratory volume in the first second, FVC forced vital capacity, PaO₂ arterial oxygen partial pressure, PaCO2 arterial carbon dioxide oxygen partial pressure, OLV one-lung ventilation

Regarding arterial oxygen partial pressure, intragroup comparisons indicated that each group experienced the lowest PaO₂ during the measurement period at T3. Intergroup comparisons showed that, compared with the RV group, PaO₂ was higher in the I-HPC group at T3 by 46.47 mmHg (214.27 ± 74.38 vs. 167.80 ± 65.68; 95% CI 9.40–83.54; P = 0.015) and at T4 by 44.77 mmHg (221.80 ± 74.67 vs. 177.03 ± 70.94; 95% CI 6.37–83.17; P = 0.023). Similarly, PaO₂ was higher in the C-HPC group than in the RV group at T3 by 49.56 mmHg (217.35 ± 76.10 vs. 167.80 ± 65.68; 95% CI 12.79–86.32; P = 0.009) and at T4 by 43.77 mmHg (220.81 ± 78.58 vs. 177.03 ± 70.94; 95% CI 5.68–81.86; P = 0.025). No significant differences in PaO₂ were observed between the I-HPC and C-HPC groups at T3 (mean difference − 3.09 mmHg; 95% CI − 39.86 to 33.68; P = 0.868) or T4 (mean difference 0.99 mmHg; 95% CI − 37.10 to 39.08; P = 0.959) (Table 2, Figure S1). Table S1 presents multiplicity-adjusted results using the Holm–Bonferroni method, which showed consistent effect estimates, although statistical significance at T4 was attenuated after correction.

Table 2.

Intraoperative PaO₂ and SpO₂ among three groups

Group	RV (n = 30)	C-HPC (n = 31)	I-HPC (n = 30)	P
PaO2(mmHg)
T1 (before OLV)	408.03 ± 72.37	398.68 ± 50.87	413.87 ± 47.76	0.588
T2 (OLV 5 min)	233.83 ± 96.72	242.48 ± 77.19	238.00 ± 70.79	0.919
T3 (OLV 15 min)	167.80 ± 65.68	217.35 ± 76.10	214.27 ± 74.38	0.014a, b
T4 (OLV 30 min)	177.03 ± 70.94	220.81 ± 78.58	221.80 ± 74.67*	0.034a, b
Number of hypoxemic patients with SpO2 <95%	3 (10%)	1 (3.2%)	1 (3.3%)	0.526

Abbreviations: RV regular ventilation, C-HPC continuous hypoxic preconditioning, I-HPC intermittent hypoxic preconditioning, OLV one-lung ventilation, PaO₂ arterial oxygen partial pressure, SpO2 pulse oxygen saturation

^aP < 0.05 for I-HPC group compared to the RV group

^bP < 0.05 for I-HPC group compared to the RV group

Regarding the decrease in oxygen saturation, three patients in the RV group(incidence rate: 10.00%, 95% CI: 2.12–26.53%) and one patient each in the C-HPC(incidence rate: 3.23%, 95% CI: 0.08–16.70%) and I-HPC groups(incidence rate: 3.33%, 95% CI: 0.08–17.22%) exhibited SpO2 levels below 95%, with no statistically significant difference observed among the groups (P = 0.526). None of the patients in any group recorded an SpO2 level below 90%. The analysis of other arterial blood gas values, along with respiratory indicators (VT, PEAK, PLAT) and circulatory indicators (HR, MAP) at each measurement time point, revealed no significant differences among the three patient groups (All P > 0.05) (Table 3). Additionally, the utilization of vasoactive drugs, including Ephedrine (P = 0.413), Phenylephrine (P = 0.541), and Urapidil (P = 0.471), during surgery did not demonstrate significant differences (Table 4).

Table 3.

Results of other arterial blood gas values, respiratory and circulatory indicators

	Time	RV	C-HPC	I-HPC	P
pH
	T1	7.40 ± 0.05	7.39(0.08)	7.40 ± 0.05	0.930
	T2	7.40 ± 0.05	7.41 ± 0.05	7.40 ± 0.05	0.617
	T3	7.42 ± 0.04	7.43 ± 0.06	7.41 ± 0.06	0.406
	T4	7.42 ± 0.05	7.43 ± 0.06	7.41 ± 0.06	0.538
PaCO2(mmHg)
	T1	41.00(8)	41.00(9)	43.50(8)	0.485
	T2	40.00(6)	39.00(7)	42.50(8)	0.42
	T3	38.23 ± 4.42	38.13 ± 5.85	40.30 ± 6.29	0.239
	T4	37.63 ± 3.86	37.52 ± 5.86	39.93 ± 5.78	0.166
HCO3-(mmol/L)
	T1	25.70 ± 1.96	25.89 ± 1.83	26.50 ± 1.45	0.193
	T2	25.40 ± 1.86	25.71 ± 1.77	26.02 ± 1.41	0.370
	T3	24.82 ± 1.69	25.21 ± 1.91	25.38 ± 1.70	0.458
	T4	24.54 ± 1.65	24.84 ± 1.82	25.50 ± 1.49	0.075
Hemoglobin (g/L)
	T1	115.13 ± 14.15	118.74 ± 11.11	118.20 ± 13.73	0.513
	T2	114.00 ± 14.12	115.74 ± 12.43	115.07 ± 15.33	0.887
	T3	113.23 ± 14.87	113.97 ± 13.61	113.70 ± 14.61	0.98
	T4	113.03 ± 13.52	112.00 ± 12.95	113.00 ± 15.03	0.946
VT (mL)
	T1	445.63 ± 53.73	438.58 ± 51.79	417.47 ± 38.25	0.069
	T2	438.73 ± 57.45	441.35 ± 48.23	414.67 ± 42.67	0.077
	T3	441.33 ± 52.12	436.81 ± 49.19	415.30 ± 44.83	0.093
	T4	440.43 ± 52.19	430.23 ± 48.96	411.50 ± 41.32	0.064
PEAK (cmH2O)
	T1	17(5)	18(4)	17(3)	0.440
	T2	22.77 ± 3.79	22.32 ± 3.63	22.00 ± 4.27	0.747
	T3	23.43 ± 4.27	22.87 ± 3.71	22.47 ± 3.66	0.628
	T4	23.57 ± 4.59	22.77 ± 3.48	22.47 ± 4.07	0.559
PLAT (cmH2O)
	T1	15(5)	15(3)	14(3)	0.097
	T2	19.53 ± 3.90	18.94 ± 3.28	17.97 ± 4.13	0.275
	T3	19.80 ± 4.34	19.26 ± 3.40	18.73 ± 83.74	0.563
	T4	20.07 ± 4.43	19.29 ± 3.33	18.37 ± 3.99	0.251
HR (bpm)
	T1	59(15)	61(21)	70(20)	0.073
	T2	65(12)	70.32 ± 13.31	71.87 ± 11.19	0.149
	T3	70.37 ± 11.76	71.68 ± 15.73	74.87 ± 10.75	0.388
	T4	64.93 ± 10.38	67.45 ± 12.76	72.27 ± 13.04	0.064
MAP (mmHg)
	T1	90.13 ± 12.38	86.90 ± 16.12	88.03 ± 13.86	0.670
	T2	90.37 ± 11.71	94.81 ± 12.92	92.70 ± 13.14	0.392
	T3	84.93 ± 12.13	80.13 ± 10.82	84.10 ± 12.69	0.245
	T4	83.73 ± 9.02	78.94 ± 13.68	79.83 ± 10.30	0.214

Abbreviations: RV regular ventilation, C-HPC continuous hypoxic preconditioning, I-HPC intermittent hypoxic preconditioning, PaCO2 arterial carbon dioxide oxygen partial pressure, VT tidal volume, PEAK peak airway pressure, PLAT plateau pressure, HR heart rate, MAP mean arterial pressure

Table 4.

Comparison of intraoperative vasoactive drugs used in patients (number of cases)

	RV (n = 30)	C-HPC (n = 31)	I-HPC (n = 30)	P
Ephedrine	11	15	16	0.413
Phenylephrine	0	2	2	0.541
Urapidil	5	5	2	0.471

Abbreviations: RV regular ventilation, C-HPC continuous hypoxic preconditioning, I-HPC intermittent hypoxic preconditioning

Postoperative pulmonary complications demonstrated no significant differences among the groups (Table 5), which included pneumonia (P = 0.298), pleural effusion (P = 0.437), atelectasis (P = 0.319), and prolonged air leak (P = 0.522). Furthermore, no serious postoperative complications, such as pulmonary embolism or respiratory failure, were reported in any of the patients.

Table 5.

Characteristics of postoperative pulmonary complications

	RV(n = 30)	C-HPC (n = 31)	I-HPC(n = 30)	P
Pneumonia	4	1	2	0.298
Pleural effusion	9	5	7	0.437
Atelectasis	2	0	1	0.319
Prolonged air leak	2	1	3	0.522
Total PPCs	14	6	12	0.066

Abbreviations: RV regular ventilation, C-HPC continuous hypoxic preconditioning, I-HPC intermittent hypoxic preconditioning, PPC postoperative pulmonary complication

Discussion

The results of this study indicated that, in comparison to the control group, both the I-HPC group and the C-HPC group exhibited elevated PaO₂ values at both 15 and 30 min following OLV. No statistically significant differences were observed in the incidence of intraoperative hypoxemia or postoperative pulmonary complications among the groups. Preconditioning the nondependent lung with hypoxia prior to pleural incision has the potential to enhance arterial oxygenation during OLV. These findings may provide important insights into the practical application of HPC in clinical settings. These findings may offer valuable insights on the implementation of HPC in clinical practice.

Arterial oxygenation during OLV is influenced by multiple perioperative factors, including airway management, anesthetic technique, and hemodynamic stability. Tube malposition remains a major cause of intraoperative hypoxemia, although its incidence has decreased with routine fiberoptic confirmation [20–22]. The effects of intravenous versus inhalational anesthesia on oxygenation during OLV remain inconsistent [23, 24]. Therefore, total intravenous anesthesia was used to minimize confounding. Hemodynamic instability and vasoactive drug use may also affect hypoxic pulmonary vasoconstriction [25, 26]. However, vasoactive drug use did not differ among groups in the present study, suggesting minimal influence on the observed oxygenation differences. Finally, to ensure that the baseline characteristics of the patients in each group were as consistent as possible, we compared the gender, age, weight, height, BMI, lung function, preoperative arterial blood gas values, surgical type, duration of OLV, duration of surgery and anesthesia, postoperative chest tube drainage, and length of hospital stay among the groups, and found no statistically significant differences. Studies have shown that 20–30 min after OLV, the PaO₂ tends to reach its lowest point and then gradually increases. In our experiment, PaO₂ decreased significantly at T2, T3, and T4 in all groups compared with T1, and PaO₂ increased slightly at T4 in all three groups compared with T3, but the difference was not statistically significant. All groups of patients reached the lowest PaO₂ at T3 during the measurement period. In the comparison of PaO₂ between groups at T3 and T4, both the C-HPC group and I-HPC group had higher PaO₂ than the RV group, and the difference was statistically significant, indicating that HPC of the nondependent lung can improve arterial oxygenation during OLV. Although the Holm-Bonferroni correction indicated that the experimental groups exhibited significantly higher PaO2 at T3, no statistically significant difference was observed at T4. However, as this study is exploratory in nature, we aim not only to test the primary hypothesis but also to identify potential, non-prespecified patterns of intergroup differences within the data. Therefore, the LSD test was ultimately selected to demonstrate the results. A previous study on 136 patients undergoing OLV showed that hypoxemia was less frequent in patients that received five intermittent hypoxic stimuli [6]. They also found that PaO₂ in the intermittent hypoxia group was higher 30 and 60 min after starting OLV, which is consistent with our results. However, there was no significant difference between the C-HPC group and I-HPC group at T3 and T4, indicating that the form of hypoxia did not have a significant impact on improving arterial oxygenation during OLV. A retrospective study showed that the lowest oxygen pressure during OLV may be most closely related to the largest oxygen pressure during two lung ventilation [27]. Therefore, in this study, the oxygen pressure of each patient was measured before thoracotomy, and the results showed that there was no difference in oxygen pressure values among the three groups of patients before OLV, ensuring the consistency of oxygen pressure before OLV in the three groups of patients.

Hypoxic preconditioning has been shown to enhance hypoxic pulmonary vasoconstriction at the level of the lung, lung lobes, and pulmonary vasculature [28–30]. These mechanisms have been consistently demonstrated in experimental and animal models of hypoxic exposure, supporting a causal link between hypoxic preconditioning, pulmonary vascular reactivity, and improved oxygenation during one-lung ventilation. Hypoxic pulmonary vasoconstriction develops rapidly after hypoxic exposure and persists transiently after reoxygenation [31]. Although repeated hypoxic stimuli can augment vasoconstrictive responses, this effect does not increase indefinitely, with several experimental and clinical studies suggesting that maximal responses are reached after initial exposures [12–14, 28, 32, 33]. Consistent with these findings, both intermittent and continuous hypoxic preconditioning improved arterial oxygenation during OLV in the present study, without a significant difference between strategies, suggesting that hypoxic preconditioning itself, rather than the specific stimulation pattern, is the primary contributor to improved oxygenation. However, because hypoxic pulmonary vasoconstriction and intrapulmonary shunt fraction were not directly measured in this study, this mechanistic interpretation should be regarded as hypothesis-generating; future studies incorporating direct measurements of pulmonary artery pressure, shunt fraction, or cardiac output during OLV would be required to formally test this mechanism.

Unfortunately, although the number of patients with hypoxemia was lower in the two experimental groups compared to the standard bilateral ventilation group, there was no statistical significance in the differences. We speculate that this may be related to the following reasons: First, it may be related to the ventilation parameter settings. In this experiment, we always maintained pure oxygen ventilation in the non-operative lung during OLV, which to some extent reduced the probability of hypoxemia during the surgery. Second, it may be related to the insufficient sample size. In this study, the oxygen partial pressure during OLV was used as the main indicator, rather than oxygen saturation. Moving forward, to further investigate the differences in hypoxemia incidence among the groups, we should expand the sample size and continue to refine the respiratory parameter settings to quantify the role of hypoxia pre-treatment in thoracoscopic surgery in a more scientific and effective way.

This study further investigated the effects of two types of HPC on early postoperative pulmonary complications, which include pulmonary infection, pleural effusion, atelectasis, insertion of a new chest tube or indwelling chest tube for more than 7 days, postoperative respiratory failure, and pulmonary embolism [17–19]. The results showed that there was no statistically significant difference in the incidence of postoperative pulmonary complications between the experimental group and the control group. Similarly, a previous study on patients undergoing thoracoscopic lobectomy also found no significant differences in the incidence of pulmonary complications or overall postoperative morbidity between patients receiving HPC and control group [7]. This may be because the benefits of HPC may be more evident in improving oxygenation and lung function, rather than reducing postoperative complications. Related studies have also suggested that HPC may reduce the level of inflammatory cytokines to alleviate lung injury and provide lung protection [34, 35]. Unfortunately, our study did not measure the levels of related inflammatory cytokines, preventing us from evaluating the incidence of postoperative pulmonary inflammatory injury in each group of patients.

Previous studies have shown that lung recruitment strategies seem to improve arterial oxygenation during OLV and to some extent reduce the incidence of hypoxemia [36]. Lung recruitment usually refers to the expansion of the dependent lung for 15–20 s using a pressure of 20 cm H₂O or higher, reducing atelectasis and dead space by briefly increasing airway pressure and expanding the lung [37]. It can be used in both thoracic and non-thoracic surgery. However, there are risks associated with using lung recruitment strategies to improve oxygenation in thoracic surgery. Sustained intrathoracic pressure can lead to hemodynamic instability, especially in patients with low blood volume. Similarly, elevated airway pressure poses a risk of lung injury, particularly in patients with existing lung disease. Conversely, HPC in the nondependent lung is unlikely to cause such complications. Our study was conducted preoperatively so it would not interfere with the surgical procedure in any way.

The limitations of this study are mainly reflected in the following aspects: Firstly, this study was conducted at a single center, representing a homogeneous population, and the sample size was relatively small. As a result, potential biases in the conclusions cannot be entirely excluded. Secondly, previous studies have shown that patients with lower cardiopulmonary reserve are more prone to hypoxemia during surgery. However, the patients included in this study had relatively healthy cardiopulmonary reserves, so our results cannot be extrapolated to patients with poor cardiopulmonary function. In addition, all patients were ventilated with an FiO₂ of 1.0 during one-lung ventilation, and the cohort consisted exclusively of ASA I-II patients, representing a population at relatively low risk of intraoperative hypoxemia. Therefore, the generalizability of these findings to higher-risk patients remains uncertain. Lastly, most previous studies have suggested that the beneficial effects of hypoxia preconditioning are often achieved by enhancing HPV. To accurately evaluate HPV and calculate the shunt fraction, we often need to measure pulmonary artery pressure, mixed venous oxygen content, and cardiac output using a pulmonary artery catheter [25]. However, due to the invasiveness of catheterization and the relatively healthy status of the patients included in our study, we did not quantify HPV. Future studies that build on this foundation may further elucidate the effectiveness of hypoxia preconditioning in one-lung ventilation.

Conclusion

In conclusion, both repeated intermittent hypoxia stimulation and single persistent hypoxia stimulation of the nondependent lung prior to pleural incision may improve arterial oxygenation during subsequent OLV. HPC could potentially be considered as a supportive measure for patients undergoing OLV in elective thoracoscopic lung resection.

Abbreviations

One

lung ventilation (OLV)

HPC

Hypoxic preconditioning

ARB

Angiotensin-II receptor blockers

FEV₁/FVC

Forced expiratory volume in one second/forced vital capacity

ACEI

Angiotensin-converting enzyme inhibitors

SpO₂

Pulse oxygen saturation

MAP

Mean arterial pressure

HR

Heart rate

PEAK

Peak airway pressure

PLAT

Plateau pressure

VT

Tidal volume

SD

Standard deviation

Authors’ contributions

Yue Wu and Mei Yang carried out the studies, participated in collecting data, and drafted the manuscript. Liangfei Tian and Jingyin Yuan performed the statistical analysis and participated in its design. Qingping Wen: performed research guidance, final draft, and financial support. All authors read and approved the final manuscript.

Funding

This study received no external funding. The funder had no role in the study design, data collection, data analysis, interpretation of data, or manuscript preparation.

Data availability

The raw data supporting the findings of this study are available from the corresponding author upon reasonable request. Requests for access to the dataset should be directed to Yue Wu (Email: wuyue0903@126.com).

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the First Affiliated Hospital of Dalian Medical University PJ-KS-KY-2022-46(X). The study was conducted in accordance with the principles of the Declaration of Helsinki. All participants were informed about the study protocol and provided written informed consent to participate in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.