Intraoperative ventilation strategies during thoracic surgery to reduce the perioperative inflammatory response: a narrative review from in vivo to clinical evidence

Summary

General thoracic surgery encompasses a range of procedures requiring one-lung ventilation. This technique is essential for surgical exposure but is associated with significant physiological changes that contribute to both local and systemic inflammatory responses from tissue injury. This narrative review summarises perioperative ventilation strategies aimed at minimising lung injury and inflammation, focusing on their impact on inflammatory markers. A literature search was conducted in PubMed, MEDLINE, and Embase up to July 2025, excluding non-English articles. When positive pressure ventilation is used, protective ventilation strategies including low tidal volume (5–6 ml kg⁻¹), individualised PEEP, and adaptive control or pressure-controlled ventilation seem to confer advantages in terms of inflammatory outcomes and ventilator-induced lung injury. Careful titration of inspired oxygen is essential in the balancing of adequate oxygenation with the risk of oxidative damage. Partial lung collapse or continuous application of positive airway pressure in the non-ventilated lung may help attenuate tissue injury from atelectasis and re-expansion injury. In selected patients, non-intubated video-assisted thoracoscopic surgery can potentially result in the attenuation of perioperative inflammatory responses by eliminating positive pressure ventilation entirely, although clinical evidence remains limited. Although these strategies have been shown to result in consistent attenuation of perioperative inflammatory responses, their translation into clinical outcomes was found to be inconsistent across the reviewed studies. Further well-designed studies with standardised cytokine measurement methodology and validated clinical endpoints are needed to establish definitive perioperative ventilation recommendations during one-lung ventilation.

Editor's key points.

• •One-lung ventilation during thoracic surgery can promote ventilator-induced lung injury and trigger local and systemic inflammatory responses linked to worse postoperative pulmonary outcomes.
• •The review suggests that lung-protective ventilation (especially low tidal volume with appropriate or individualised PEEP) is associated with lower inflammatory marker release and better lung mechanics; excessively high fixed PEEP may be harmful.
• •Although some ventilation modes and adjuncts seem promising, the effect of these strategies on clinical outcomes remains inconsistent, and further well-designed studies are needed.

Thoracic (noncardiac) surgery is used to address a range of conditions within the chest, including diseases of the lungs and airways, mediastinal tumours, and oesophageal cancer. Surgical trends are increasingly shifting towards minimally invasive approaches, driven by technological advances with the aim of improving patient outcomes.¹ However, concerns about morbidity and mortality remain. Current data indicate that pulmonary resection procedures have an in-hospital mortality rate ranging from 0.56% to 1.02% and a major morbidity rate of 6.8%. In contrast, oesophagectomy is associated with a higher in-hospital mortality rate, between 3.11% and 3.97%, and a major morbidity rate of 29.83%.²

To ensure optimal surgical exposure, patients are typically placed in one-lung ventilation (OLV) in the lateral decubitus position. Hypoxic pulmonary vasoconstriction (HPV), a protective mechanism that redirects blood flow away from the collapsed lung towards the better-ventilated lung usually occurs, enhancing overall oxygenation.³ Despite its benefits, OLV presents several challenges. The ventilated lung is prone to barotrauma and volutrauma as a consequence of high airway pressures and large tidal volumes (VTs). The resultant hyperperfusion, in combination with mechanical stress-induced endothelial injury and increased vascular permeability, promotes pulmonary oedema. Meanwhile, the non-ventilated lung undergoes collapse and hypoperfusion and upon re-expansion becomes susceptible to reperfusion injury.^4,5 These physiological changes underscore the complexity and inherent risks of thoracic surgery.

Various inflammatory markers have been used to assess the extent of inflammation during thoracic surgery. Elevated concentrations of these markers have been linked to adverse clinical outcomes.6, 7, 8, 9 The physiological consequences of OLV collectively trigger both local and systemic inflammatory responses, which have been independently associated with adverse postoperative outcomes including increased postoperative pulmonary complications, prolonged mechanical ventilation, and extended hospital stay.6, 7, 8, 9 The primary aim of this narrative review is to evaluate and synthesise available pre-clinical and clinical evidence pertinent to perioperative ventilation strategies during OLV, with a specific focus on their impact on inflammatory markers as indicators of lung injury. The secondary aim is to provide practical evidence-based guidance for clinicians in the management of patients undergoing thoracic surgery.

A literature review was conducted using PubMed, MEDLINE, and Embase databases up to July 2025. The search strategy used the following keywords: ‘one-lung ventilation’, ‘thoracic surgery’, ‘ventilation strategy’, and ‘inflammation’. Additional relevant articles were identified through supplementary searches. Relevant reports from pre-clinical and clinical studies were included. Articles published in languages other than English were excluded from the review. An overview of ventilation strategies for OLV is presented in Fig. 1.

Fig 1.

Overview of ventilation strategies for one-lung ventilation to minimise inflammatory response and ventilator-induced lung injury during thoracic surgery. (a) Interventions on the ventilated lung, (b) interventions on the non-ventilated lung, and (c) the effect of non-intubated video-assisted thoracic surgery on inflammatory response. CC-16, Clara cell protein 16; CCL-2, C-C chemokine ligand 2; CRP, C-reactive protein; CXCL-1, C-X-C chemokine ligand 1; ICAM, intercellular adhesion molecule; IL, interleukin; NE, neutrophil elastase; NA, not available; NK, natural killer; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen; Pmean, mean airway pressure; Ppeak, peak airway pressure; Pplat, plateau airway pressure; TNF, tumour necrosis factor; VT, tidal volume. ∗Data only available in the in vivo study. ^#Significant only in the in vivo study. Created in https://Biorender.com.

Effects of tidal volume on perioperative inflammatory response during one-lung ventilation

During OLV, carbon dioxide (CO₂) retention and hypoxaemia occur as a result of hypoventilation and ventilation/perfusion (V/Q) mismatch. Traditionally, higher VTs were used for the ventilated lung to maintain adequate oxygenation and prevent CO₂ retention.¹⁰ However, ventilator-induced lung injury (VILI) has long been recognised, with high VTs being a key factor.¹¹ Emerging evidence suggests that high VTs can induce an inflammatory response through many mechanisms such as endothelial injury, increased production of reactive oxygen species (ROS), and mitochondrial damage.12, 13, 14 Consequently, a low VT ventilation strategy has been proposed in patients with OLV to prevent these risks.

In an animal model of OLV, high VT ventilation was shown to increase airway pressure and result in a greater inflammatory response in the ventilated lung, as evidenced by elevated concentrations of interleukin-6 (IL-6), tumour necrosis factor-α (TNF-α), and polymorphonuclear cell (PMN) infiltration in bronchoalveolar lavage fluid (BALF). In addition, increasing inflammation was observed in the contralateral non-ventilated lung and systemic circulation in comparison with the low VT group, suggesting that high VT ventilation might induce both local and systemic inflammatory responses.¹⁵ In clinical settings, low VT ventilation (5 ml kg⁻¹) during pulmonary resection was shown to reduce concentrations of TNF-α and the soluble intercellular adhesion molecule in BALF in the ventilated lung, but clinical outcomes were not reported.¹⁶ In patients undergoing oesophagectomy with OLV, Zhang and colleagues¹⁷ compared ventilation strategies using low VT (<8 ml kg⁻¹) with those using high VT (>8 ml kg⁻¹). In the low VT pressure-controlled ventilation (PCV) group, significantly lower postoperative BALF concentrations of IL-6, TNF-α, and C-reactive protein (CRP) were recorded than in all other groups, along with a reduced incidence of postoperative pulmonary complications. It should be noted that the VT threshold of <8 ml kg⁻¹ represents the definition used in the original study and may not reflect the lower VT targets typically used in contemporary lung-protective ventilation strategies for thoracic surgery. Collectively, these findings emphasise the potential benefit of low VT ventilation to reduce the inflammatory response and improve perioperative outcomes. However, one study showed no difference in inflammatory marker concentrations in the BALF of the ventilated lung in patients undergoing thoracotomy for pulmonary resection.¹⁸ This inconsistent finding might be because of the different duration of OLV, application of PEEP, and method of anaesthesia. Given that a longer duration of mechanical ventilation typically leads to more significant injury,¹¹ the relatively short duration of OLV in the study may have been insufficient to produce significant lung injury. In addition, the protective effect of PEEP used in the study might result in reduced inflammation observed in the high VT group. Another factor was that a volatile anaesthetic agent was used in the study, and this might have induced a lower inflammatory response than an i.v. agent.19, 20, 21 Taken together, these factors potentially result in the neutral outcome observed in the study.

In summary, compared with high VT, current evidence indicates that low VT during OLV contributes to lower airway pressure and reduces the inflammatory response associated with VILI, particularly during prolonged OLV. However, only a few studies have specifically investigated the isolated effect of VT. Most of the studies compared groups with varying VTs and PEEP levels. Reports from these in vivo and clinical investigations are comprehensively presented in Supplementary Table 1.

Effect of PEEP on perioperative inflammatory response during one-lung ventilation

PEEP vs no PEEP

Another strategy during the perioperative period is the application of PEEP to improve lung mechanics and prevent perioperative pulmonary complications. The protective effects of PEEP primarily prevent alveolar collapse during the expiratory phase.²² PEEP provides sustained gas exchange, reducing the shunting effect and improving both oxygenation and ventilation during the respiratory cycle. In addition, as a consequence of the prevention of the cyclic collapse and re-expansion of alveoli, PEEP enhances airway compliance. The improvement in compliance reduces the inspiratory pressure needed for ventilation, thereby minimising the risk of VILI.²³ Historically, the application of PEEP during OLV was approached with caution, owing to two principal concerns. The predominant concern was that the application of PEEP to the dependent ventilated lung would increase pulmonary vascular resistance, diverting blood flow towards the non-ventilated collapsed surgical lung and consequently worsening intrapulmonary shunt and intraoperative hypoxaemia. In addition, the risk of barotrauma as a result of elevated airway pressures in the dependent lung has been recognised as a concurrent concern. The historical practice of zero end-expiratory pressure has since been largely replaced, and contemporary surveys of both clinical practice and observational studies confirm that the application of PEEP during OLV is now widely adopted as standard practice.^24,25 There is emerging evidence to further support that the use of PEEP during OLV confers positive effects throughout the perioperative period.²⁶ A randomised controlled trial was conducted to investigate the effects of different PEEP levels in patients undergoing thoracotomy for oesophagectomy. That study demonstrated that the application of PEEP at 5 cm H₂O significantly improved oxygenation and reduced postoperative serum concentrations of inflammatory markers, including IL-6 and interleukin-10, in comparison with no PEEP application. Significantly, the clinical pulmonary infection score (CPIS), an indicator of pneumonia risk, was lower in the PEEP group. This finding shows that the application of PEEP could reduce the level of systemic inflammatory response and risk of pulmonary complication. However, when comparing PEEP 5 cm H₂O with PEEP 10 cm H₂O, the higher PEEP level was associated with an increased systemic inflammatory response, but without any significant differences in clinical outcomes between the groups. These findings suggest that moderate PEEP levels enhance oxygenation and reduce systemic inflammation during OLV compared with no PEEP or high PEEP.²⁷ Notably, the relatively high VT (8 ml kg⁻¹) used in the study might have influenced these outcomes, particularly in the PEEP 10 cm H₂O group, where high airway pressure might have contributed to increased inflammation. The combination of high VT and high PEEP resulted in an increased systemic inflammatory response. Therefore, the application of PEEP during OLV should be carefully adjusted to the VT settings to optimise outcomes and minimise the risk of VILI.

Individualised PEEP

The application of PEEP alone may not be sufficient to optimise lung mechanics, owing to the variation in airway resistance amongst patients. To address this, ‘individualised’ or ‘optimal’ PEEP has been proposed, where PEEP is titrated to maximise lung compliance for each patient. The most commonly used methods for titrating PEEP are the decremental and incremental approaches, both aimed at identifying the PEEP level that maximises lung compliance.²⁸ Two studies comparing fixed PEEP 5 cm H₂O with individualised PEEP in patients undergoing video-assisted thoracoscopic surgery (VATS) pulmonary resection found that individualised PEEP was associated with a reduced level of systemic inflammation, as indicated by lower concentrations of systemic cytokines, including IL-6, interleukin-8 (IL-8), neutrophil elastase (NE), and Clara cell protein 16. In addition, individualised PEEP significantly improved lung compliance and oxygenation, regardless of the titration method used. Although no significant differences were observed in individual postoperative pulmonary complications such as pneumonia, respiratory failure, atelectasis, and acute respiratory distress syndrome (ARDS), the composite incidence of postoperative pulmonary complications was significantly lower, and ICU length of stay was significantly reduced in the individualised PEEP group.^29,30 These findings suggest that individualised PEEP may reduce systemic inflammation more effectively than a fixed PEEP approach. This effect is likely attributable to more favourable lung mechanics during individualised PEEP, characterised by improved lung compliance and lower driving pressure secondary to reduced plateau airway pressure. Together, these physiological effects may lessen mechanical stress on the lung parenchyma and attenuate the inflammatory response. In conclusion, PEEP application during OLV enhances lung compliance, improves oxygenation, and attenuates both local and systemic inflammatory responses. However, excessively high fixed PEEP levels may increase airway pressures and exacerbate inflammation, particularly when combined with high VTs. Individualised PEEP titration, guided by lung mechanics, seems to provide superior inflammatory and mechanical benefits over fixed PEEP strategies, although clinical outcomes were mixed across the reviewed studies. The detailed findings of all included studies are presented in Table 1.

Table 1.

Effect of PEEP on perioperative inflammatory response during OLV: reports from in vivo and clinical investigations. ABG, arterial blood gas; ARDS, acute respiratory distress syndrome; ARF, acute respiratory failure; CC16, Clara cell protein 16; Cdyn, dynamic compliance; CPIS, clinical pulmonary infection score; G, group; ICU, intensive care unit; IL, interleukin; intraop, intraoperative; LOS, length of stay; NA, not available; NE, neutrophil elastase; OLV, one-lung ventilation; Pa_CO2, arterial carbon dioxide partial pressure; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen ratio; Pmean, mean airway pressure; Ppeak, peak airway pressure; Pplat, plateau airway pressure; PRVC, Pressure-Regulated Volume Control; Ref, reference; VATS, video-assisted thoracoscopic surgery; VCV, volume-controlled ventilation; ↑, significant increase; ↓, significant decrease; ↔, no significant difference.

Study model	Procedure(n)	Mechanical ventilator		OLV duration (min)	Outcome					Interpretation	Ref
		Tidal (ml kg−1)	PEEP (cm H2O)		Lung mechanics and ABG (intraop)	Cytokine			Clinical outcomes
						Lung		Serum
						Ventilated	Non-ventilated
Human with oesophageal cancer	Thoracotomy oesophagectomy(116)	8	5 vs 0	170 vs 172	↔Cdyn, ↑P/F ratio, ↔PaCO2,	NA	NA	↓IL-6, ↓IL-10	↓CPIS, ↔complications	Applying PEEP improved oxygenation and reduced inflammation without altering postoperative complications. However, application of higher PEEP was associated with increased systemic inflammation.	27
			10 vs 5	178 vs 170	↑Cdyn, ↑P/F ratio, ↔PaCO2	NA	NA	↑IL-6, ↑IL-10	↔CPIS, ↔complications
Human	VATS pulmonary resection(176)	PRVC 5–6	Decremental PEEP (8) vs fixed 5	120 vs 133	↔Ppeak, ↑Pmean, ↑Cdyn, ↔P/F ratio, ↔PaCO2	NA	NA	↓NE	↔Pneumonia, ↔atelectasis, ↔ARF, ↓ICU stay, ↔LOS	Decremental individualised PEEP improved lung mechanics and reduced systemic inflammation in both VCV and PRVC modes, without affecting postoperative pulmonary complications. A shorter ICU stay was observed only with the specific combination of decremental PEEP and PRVC mode.	29
		VCV 5–6	Decremental PEEP (10) vs fixed 5	135 vs 160	↔Ppeak, ↑Pmean, ↑Cdyn, ↑P/F ratio, ↔PaCO2	NA	NA	↓NE	↔Pneumonia, ↔atelectasis, ↔ARF, ↔ICU stay, ↔LOS
Human	VATS pulmonary resection(90)	PRVC 5–6	G1: decremental PEEP (8.7)	130	G1 and G2 vs G3: ↓Ppeak, ↓Pplat, ↑Cdyn, ↑P/F ratio, ↔PaCO2 G1 vs G2: ↔	NA	NA	G1 and G2 vs G3: ↓IL-6, ↓IL-8, ↓NE, ↓CC16 G1 vs G2: ↔	G1 and G2 vs G3: ↓pulmonary complication, ↓ICU stay, ↔pneumonia, ↔atelectasis, ↔ARF, ↔ARDS G1 vs G2: ↔	All three groups were compared. Both individualised PEEP strategies (G1 and G2) improved lung mechanics and oxygenation, reduced systemic inflammation, and decreased pulmonary complications compared with fixed PEEP (G3). No significant differences were observed between the two individualised PEEP strategies.	30
			G2: incremental PEEP (8.2)	115
			G3: fixed PEEP (5)	125

Alveolar recruitment manoeuvre

The dependent lung is vulnerable to atelectasis as a result of multiple factors, including mechanical forces, external compression from the chest wall and abdominal pressure, and absorptive atelectasis caused by a high fraction of inspired oxygen (FiO₂). These factors lead to decreased lung compliance and overdistension of the remaining open alveoli, contributing to lung injury.^4,5 The alveolar recruitment manoeuvre (ARM) is an intervention aimed at reopening collapsed alveoli and improving lung compliance. Several techniques can be used to perform ARM. One commonly used method involves the gradual application of stepwise increments in PEEP to achieve alveolar recruitment. Alternatively, sustained inflation using a predefined peak airway pressure can also be applied to reopen alveoli effectively. Both methods are commonly used during general anaesthesia.31, 32, 33 Within the scope of our review, we identified only two studies investigating the relationship between ARM as an isolated intervention and the inflammatory response during thoracic surgery. Kim and colleagues³⁴ examined the effects of performing ARM during VATS pulmonary resection. Their ARM protocol involved conducting ARM after 10 min OLV using the following sequence: peak inspiratory pressure of 30 cm H₂O for three breaths, 35 cm H₂O for three breaths, and 40 cm H₂O for six breaths. The results demonstrated a reduction in TNF-⍺ concentrations in the BALF of the dependent lung in the ARM group in comparison with the control group. However, no significant differences in postoperative complications or abnormal chest radiograph findings were observed between groups. Another study by Shi and colleagues³⁵ investigated the effects of ARM in obese patients undergoing thoracotomy for pulmonary resection. Their protocol involved performing ARM once before the initiation of OLV by applying an airway pressure of 40 cm H₂O for approximately 40 s. The findings demonstrated significantly lower postoperative serum concentrations of TNF-⍺ and IL-6 in the ARM group, although no significant differences in CPIS were observed between groups. Collectively, these results suggest that ARM performed during thoracic surgery may reduce both local and systemic inflammatory responses. However, several critical questions remain in this area of study. First, there is no standardised method for performing ARM during OLV, particularly with regard to the optimal peak airway pressure and inflation duration. Second, the timing of ARM during OLV has not been clearly defined. We hypothesise that intermittent application of ARM to the dependent lung during OLV may reduce the inflammatory response by minimising atelectasis. Third, in previous studies, intervention groups often included additional supportive measures alongside ARM, for example, AutoFlow (adaptive control [AC]) ventilation. This confounds the ability to isolate the specific effects of ARM alone. Further research is required to validate the beneficial effects of ARM on the inflammatory response during OLV. All available data regarding the inflammatory effects of ARM are presented in Supplementary Table 2.

Effect of protective lung ventilation on perioperative inflammatory response during one-lung ventilation

The preceding sections have examined the independent effects of low VT, PEEP, and ARM during OLV. This section addresses their combined application as a unified lung-protective ventilation strategy. This combined approach is in alignment with the protective ventilation strategy traditionally used in ARDS management, where the primary goal is to minimise VILI.^36,37 Lung-protective ventilation is defined by the combination of low VT (4–6 ml kg⁻¹) to prevent volutrauma and barotrauma³⁸ and adequate PEEP and ARM to prevent atelectotrauma.^36,37

In an in vivo rabbit lung model, conventional OLV with high VTs and no PEEP resulted in increased airway pressures, reduced lung compliance, and heightened systemic inflammation, as evidenced by elevated serum thromboxane B₂ concentrations in comparison with two-lung ventilation (TLV). However, when lower VTs with PEEP were implemented, lung mechanics were maintained, and the systemic inflammatory response remained comparable with that observed with traditional TLV.³⁹ These findings suggest that an appropriate ventilation strategy may help alleviate these adverse effects of VILI during OLV. Similarly, another in vivo study in a piglet lung model undergoing thoracoscopic surgery demonstrated that protective ventilation strategies were associated with reduced local and systemic inflammatory responses in comparison with conventional OLV. Inflammatory cytokine concentrations were consistently higher in the ventilated lung than in the non-ventilated lung across both groups, indicating a greater susceptibility to injury of the ventilated lung during OLV than of the non-ventilated lung.⁴⁰ Therefore, careful management of the ventilation strategy during OLV is essential to minimise tissue injury and optimise clinical outcomes. Another study in a rat lung model demonstrated that OLV with low VT and higher PEEP reduced tissue injury, as indicated by milder pathological changes and lower PMN infiltration in lung histology than with low VT with low PEEP and high VT with low PEEP. Moreover, procollagen III, a marker for lung fibrotic changes, was increased in the high VT with low PEEP group, indicating greater ventilated lung tissue injury.⁴¹ In conclusion, the evidence from in vivo models supports the lung-protective ventilation strategies of low VTs and appropriate PEEP in the OLV to reduce VILI. These approaches provide better lung mechanics, reduce inflammation, and minimise tissue injury. Clinical studies have demonstrated that protective ventilation can effectively reduce both local and systemic inflammatory responses. During oesophagectomy, protective ventilation resulted in lower concentrations of interleukin-1β (IL-1β), IL-6, and IL-8 in BALF in the ventilated lung and improved oxygenation, which contributed to a reduction in postoperative pulmonary complications.⁴² Other studies have also shown that protective ventilation lowered systemic inflammation, evidenced by reductions in systemic inflammation markers, compared with conventional ventilation. These benefits were associated with shorter ICU stays and a reduction in the duration of mechanical ventilation.^43,44 In the context of pulmonary resection, the outcomes were comparable. Liu and colleagues⁴⁵ demonstrated that, during VATS pulmonary resection, serum markers of inflammation were significantly lower in the protective ventilation group, further supporting the anti-inflammatory benefits of this strategy. Clinical outcome data were not reported in the study. Overall, these findings suggest that protective ventilation strategies can effectively reduce both local inflammatory responses in ventilated lung tissue and systemic inflammation, consequently improving lung compliance and oxygenation, and may potentially improve postoperative outcomes compared with conventional OLV. However, Kim and colleagues³⁴ reported that protective ventilation strategy alone may not be sufficient to reduce the inflammatory response. Although the concentrations of inflammatory markers in both BALF and serum tended to be lower with protective ventilation than with conventional ventilation, the difference was not statistically significant. Interestingly, the addition of ARM to the protective ventilation strategy resulted in a more significant reduction in the inflammatory response in the ventilated lung, underlining the potential benefit of combining both strategies. This suggests that although protective ventilation may help reduce inflammation, the incorporation of ARM could further enhance these effects. However, there are differences between this study and others, such as the surgical procedure, choice of anaesthetic agents, and perioperative fluid management, which may influence the results. Another study compared the effects of very low VT (3 ml kg⁻¹) with PEEP to low VT (5 ml kg⁻¹) without PEEP. No significant differences in cytokine concentrations were observed between the two groups in either local or systemic measurements. In addition, the very low VT group required more intraoperative vasopressor support,⁴⁶ indicating that this strategy may not be a practical option for protective ventilation during OLV.

The findings of this review suggest that the mechanisms by which protective ventilation strategies attenuate perioperative inflammatory responses during OLV can be understood through a sequential pathophysiological pathway. By limiting mechanical stress on the alveolar epithelium, the primary trigger of VILI, protective ventilation strategies reduce both peak and plateau airway pressures. In combination with PEEP, these lower airway pressures contribute to a reduction in driving pressure, thereby minimising cyclic alveolar stress and repetitive alveolar overdistension during mechanical ventilation. A reduction in driving pressure may also promote more homogeneous distribution of tidal ventilation and improve dynamic lung compliance. Collectively, these physiological effects may attenuate the pulmonary inflammatory response and reduce the release of local inflammatory mediators within the ventilated lung. Preservation of alveolar epithelial integrity subsequently ensures more effective gas exchange, reflected by the improved oxygenation reported across many of the included studies. Systemic inflammatory responses during OLV represent a downstream sequel of local alveolar injury, whereby attenuation of local inflammation correspondingly reduces the release of systemic inflammatory mediators, limiting their contribution to distant organ dysfunction and adverse postoperative outcomes. This biological sequence provides a coherent explanation for the clinical benefits observed across the included studies. However, neutral clinical outcomes in a subset of studies reflect the multifactorial nature of postoperative recovery after thoracic surgery, whereby the ventilation strategy represents only one component of a complex perioperative environment encompassing surgical complexity, operative duration, anaesthetic technique, fluid management, and individual patient factors. Therefore, these findings should be interpreted with caution.

In conclusion, protective ventilation incorporating low VTs (5–6 ml kg⁻¹), individualised PEEP, and ARM consistently improves lung mechanics and attenuates both local and systemic inflammatory responses. Although some studies have reported associated reductions in postoperative complications and shorter hospital stay, clinical outcomes were heterogeneous across the reviewed studies, with some demonstrating benefit and others showing no significant difference, although no study reported evidence of harm. This heterogeneity reflects both the limitations of inflammatory markers as surrogate endpoints and the multifactorial nature of postoperative recovery after thoracic surgery. Future well-designed prospective studies incorporating both inflammatory markers and clinical outcomes as co-primary endpoints, with standardised cytokine measurement methodology and adequate statistical power, are warranted to establish the full clinical significance of these findings. The detailed findings of all included studies are presented in Table 2.

Table 2.

Effect of protective lung ventilation on perioperative inflammatory response during OLV: reports from in vivo and clinical investigations. ABG, arterial blood gas; ALC, alveolar collapse; ARDS, acute respiratory distress syndrome; BAL, bronchoalveolar lavage; CAD, coronary artery disease; CC16, Clara cell protein 16; Cdyn, dynamic compliance; Com, compliance; CRP, C-reactive protein; CXR, chest X-ray; G, group; ICU, intensive care unit; IL, interleukin; intraop, intraoperative; KL6, Krebs von den Lungen 6; LVS, large volume space; MV, mechanical ventilation; NA, not available; OLV, one-lung ventilation; Pa_CO2, arterial carbon dioxide partial pressure; PCIII, type III procollagen messenger RNA expression; Pdriv, driving airway pressure; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen ratio; Pmean, mean airway pressure; PMN, polymorphonuclear cell; Ppeak, peak airway pressure; Pplat, plateau airway pressure; PPV, positive pressure ventilation; Ref, reference; Rt, right; TLV, two-lung ventilation; SP-D, surfactant protein D; sRAGE, soluble receptor for advanced glycation end-products; TC, total cell; TNF, tumour necrosis factor; TXB2, thromboxane B₂; VATS, video-assisted thoracoscopic surgery; ↑, significant increase; ↑↑, increase of greater magnitude; ↓, significant decrease; ↔, no significant difference.

Study model	Procedure(n)	Mechanical ventilator		OLV duration (min)	Outcome					Interpretation	Ref
		Tidal (ml kg−1)	PEEP (cm H2O)		Lung mechanics and ABG (intraop)	Cytokine			Clinical outcomes
						Lung		Serum
						Ventilated	Non-ventilated
In vivo study
Rabbits	Median sternotomy(18)	G1: OLV 4G2: OLV 8G3: TLV 8 (reference)	202	90	G1 vs G3: ↔Ppeak, ↔ComG2 vs G3: ↑Ppeak, ↓Com	NA	NA	G1 vs G3: ↔TXB2G2 vs G3: ↑TXB2	NA	Compared with TLV (G3), protective ventilation (G1) preserved lung mechanics and did not increase systemic inflammation, whereas conventional ventilation (G2) resulted in unfavourable lung mechanics and increased systemic inflammation.	39
Piglets	Rt lateral position thoracoscopy(20)	5 vs 8–10	5 vs 0	180	↓Ppeak, ↔Pmean, ↔Com, ↔P/F ratio, ↑PaCO2	Tissue↔TNF-⍺, ↔IL-1β, ↓IL-6, ↓IL-8	Tissue↔TNF-⍺, ↔IL-1β, ↔IL-6, ↔IL-8	↓TNF-⍺, ↔IL-1β, ↔IL-6, ↓IL-8	NA	Protective ventilation resulted in a reduction in both local and systemic inflammation in healthy piglets.	40
Wistar rats	NA(24)	G1: no PPV (reference for tissue outcomes)		–	–	NA	NA	NA	NA	Compared with no PPV (G1), all OLV strategies increased inflammatory infiltration. Low tidal volume alone (G2) did not prevent alveolar collapse; adding adequate PEEP (G3) prevented collapse and resulted in the least tissue injury, whereas higher tidal volume (G4) caused hyperinflation and the greatest degree of injury.	41
		G2: 5	2	60	(Reference for lung mechanics)	G2 vs G1:Tissue↔PCIIIHistology↑↑ALC, ↔LVS, ↑TC, ↑↑PMN
		G3: 5	5		G3 vs G2: ↔Pdriv, ↑Com	G3 vs G1:Tissue↔PCIIIHistology↔ALC, ↔LVS, ↑TC, ↑PMN
		G4: 10	2		G4 vs G2: ↔Pdriv, ↑Com	G4 vs G1:Tissue↑PCIIIHistology↔ALC, ↑LVS, ↔TC, ↑↑PMN
Clinical study
Human with oesophageal cancer	Thoracoscopic oesophagectomy(101)	5 vs 8	5 vs 0	72 vs 75	↑P/F ratio	BAL↓IL-1β, ↓IL-6, ↓IL-8	NA	NA	↓Pulmonary complication, ↔mortality	Protective ventilation enhanced oxygenation and reduced both local inflammation and pulmonary complications.	42
Human with oesophageal cancer	Thoracotomy oesophagectomy(51)	5 vs 9	5 vs 0	85 vs 89	↓Pplat, ↑P/F ratio, ↑PaCO2	NA	NA	↓IL-1⍺, ↓IL-6, ↓IL-8	↓ICU stay, ↓MV duration, ↔ARDS, ↔pneumonia, ↔mortality	Protective ventilation resulted in lower airway pressures, enhanced oxygenation, and reduced systemic inflammation.	44
Human with oesophageal cancer	Thoracotomy oesophagectomy(40)	5–6 vs 10	3–5 vs 0	142 vs 152	↓Ppeak, ↓Pplat, ↓Cdyn, ↑P/F ratio	NA	NA	↓IL-6, ↓IL-8	NA	Protective ventilation resulted in lower airway pressures and reduced systemic inflammation.	43
Human with CAD and lung cancer	VATS pulmonary resection(60)	6 vs 10	6 vs 0	210 vs 215	↓Ppeak, ↑Cdyn, ↑P/F ratio, ↔PaCO2	NA	NA	↔TNF-⍺, ↓IL-6, ↔IL-10, ↓CRP	NA	Protective ventilation improved lung mechanics, enhanced oxygenation, and reduced systemic inflammation.	45
Human with lung cancer	VATS pulmonary resection(60)	6 vs 10	5 vs 0	121 vs 131	↓Ppeak, ↓Pplat, ↔P/F ratio, ↔PaCO2	BAL↔TNF-⍺, ↔IL-1β, ↔IL-6, ↔IL-8, ↔IL-10	BAL↔TNF-⍺, ↔IL-1β, ↔IL-6, ↔IL-8, ↔IL-10	↔TNF-⍺, ↔IL-1β, ↔IL-6, ↔IL-8, ↔IL-10	↔Abnormal CXR	Protective ventilation resulted in lower airway pressures but did not reduce local or systemic inflammation.	34
Human with oesophageal cancer	Thoracoscopic oesophagectomy(29)	3 vs 5	5 vs 0	140 vs 130	↔Ppeak, ↓Pdriv, ↔P/F ratio, ↑PaCO2	BAL (fold change)↔TNF-⍺, IL-6, IL-8, sRAGE, SP-D, CC16, KL6		↔TNF-⍺, ↔IL-6, ↔IL-8, ↔sRAGE, ↔SP-D, ↔CC16, ↔KL6	↔ICU stay, ↔pulmonary complication	Protective ventilation did not reduce the local or systemic inflammatory response.	46

Effect of mode of ventilation on perioperative inflammatory response during one-lung ventilation

Historically, two common types of continuous mandatory ventilation have been widely used: PCV and volume-controlled ventilation (VCV). In VCV, a fixed VT is delivered with a constant inspiratory flow, guaranteeing minute ventilation regardless of changes in respiratory mechanics. However, when lung compliance decreases or airway resistance increases, the resultant increase in peak airway pressure increases the risk of barotrauma.⁴⁷ In contrast, PCV delivers a pre-set inspiratory pressure with a decelerating flow pattern over a controlled inspiratory time, which limits peak airway pressure and may promote more homogeneous alveolar gas distribution.⁴⁷ However, the primary limitation of PCV is the variability of the delivered VT, given that dynamic changes in airway resistance or compliance directly alter the VT achieved, potentially resulting in hypoventilation or hyperventilation. Furthermore, when compliance is reduced, a higher set inspiratory pressure is required to achieve the target VT, raising concern for pressure-related lung injury.^47,48 Despite these respective limitations, VCV has traditionally been the preferred mode during general anaesthesia, owing to its consistent and reliable delivery of VT.

Pressure-controlled ventilation

In a study by Zhang and colleagues,¹⁷ older patients undergoing oesophagectomy were randomised to receive OLV with either low VT (<8 ml kg⁻¹) with PCV or low VT with VCV. BALF concentrations of IL-6, TNF-α, and CRP were significantly lower in the PCV group, and significantly lower airway pressures and higher oxygenation index than in the VCV group were demonstrated. The number of postoperative pulmonary complications was also lower in the PCV group than in the VCV group. These findings suggest that, within the low VT setting, PCV demonstrates superior oxygenation and attenuation of inflammatory responses in comparison with VCV during OLV. However, although both groups were assigned the same VT target (<8 ml kg⁻¹), the exact delivered VTs were not reported in the results, representing an additional limitation, given that an undetected difference in actual delivered VTs between groups could confound the interpretation of mode-specific effects.

Adaptive control

AC ventilation represents a relatively novel approach in mechanical ventilation, introduced by various manufacturers under different proprietary names including AutoFlow, Adaptive Pressure Ventilation, Adaptive Support Ventilation (ASV), Volume Control Plus, and Pressure-Regulated Volume Control. Building on the limitations of both VCV and PCV described earlier, it has been proposed that AC ventilation combines the benefits of both modes while minimising their respective drawbacks. This approach relies on a breath-to-breath algorithm that continuously adjusts inspiratory flow and pressure to achieve the targeted VT.⁴⁹ The search for this review only identified one in vivo study that investigated the effect of AC ventilation during OLV. Dai and colleagues⁵⁰ investigated the effect of ASV in a porcine post-pneumonectomy model. After left pneumonectomy, the VCV group was ventilated with a VT of 20 ml kg⁻¹, whereas the ASV group was set to target 60% of the pre-pneumonectomy two-lung minute ventilation, resulting in automatically selected VTs of 6–7.7 ml kg⁻¹ at a higher respiratory rate. The concentrations of various inflammatory markers and total protein in BALF were significantly lower in the ASV group, but the concentrations of TNF-α did not differ significantly between groups. In addition, that study also investigated the expression concentration of microRNAs. The expression concentrations of miR-144-5p and miR-449b-3p were significantly suppressed in the ASV group, and serum IL-1β was significantly lower than in the VCV group. It should be noted that the substantial difference in VTs between groups represents a potential confounding variable, given that the observed attenuation of inflammatory markers in the ASV group may reflect the effect of lower VT delivery rather than the ASV mode.

With regard to clinical research, Mahmoud and colleagues⁵¹ investigated 70 patients undergoing thoracotomy for pulmonary resection. Superior lung mechanics, higher PaO₂-to-FiO₂ ratios, and lower concentrations of inflammatory markers, including TNF-⍺, IL-8, albumin, and cellular components from BALF in the ventilated lung, were identified in the AC group, together with reduced serum TNF-⍺ and IL-8. Clinically, the AC group experienced fewer postoperative complications, shorter ICU stays, and reduced hospital lengths of stay. Despite the relatively short duration of OLV in comparison with other studies, AC demonstrated significant benefits in reducing inflammatory responses and improving clinical outcomes, suggesting its potential as a preferred mode of ventilation strategy. Another study examined the effects of AC ventilation in patients undergoing thoracotomy for pulmonary resection and reported similar findings. The AC group demonstrated improved lung mechanics and oxygenation, with reduced postoperative serum concentrations of TNF-⍺ and IL-6; however, clinical endpoints such as postoperative pulmonary complications and length of hospital stay were not reported.³⁵ Three other studies were identified on pulmonary resection performed via VATS, all using a protective ventilation strategy during OLV and reporting similar outcomes. AC was found to improve lung mechanics by reducing airway pressures and mitigating systemic inflammatory responses. One study reported lower serum inflammatory markers in the AC group, although no significant difference in CPIS was observed between groups,⁵² whereas the other two studies observed reduced serum NE. However, none of these studies demonstrated any significant clinical benefit associated with AC.^29,53

Beyond the established components of protective ventilation discussed previously, the mode of ventilation may represent an additional modifiable factor in further attenuating perioperative inflammatory responses during OLV. Across the reviewed studies, which predominantly used protective ventilation as the baseline strategy, AC ventilation consistently demonstrated more favourable lung mechanics than VCV at equivalent VTs, enabling lower peak airway pressures and plateau pressures to be achieved. This additional pressure-limiting effect seems to further attenuate local inflammatory mediator release in the ventilated lung with a corresponding reduction in systemic inflammatory response. Consistent with our earlier findings, clinical outcomes remained heterogeneous: some studies reported reduced postoperative complications and a shorter hospital stay, whereas others demonstrated no significant difference in pneumonia, atelectasis, or length of stay, despite the observed attenuation of the inflammatory response, although no study identified any safety concerns. These findings suggest that AC ventilation may serve as a clinically accessible adjunct to established protective ventilation strategies for further optimising the perioperative inflammatory profile during OLV, although larger well-designed studies are needed to confirm clinical benefit. Modes of ventilation during thoracic surgery are presented in Table 3.

Table 3.

Effect of mode of ventilation on perioperative inflammatory response during OLV: reports from in vivo and clinical investigations. ABG, arterial blood gas; AC, adaptive control; ARDS, acute respiratory distress syndrome; ARF, acute respiratory failure; ASV, Adaptive Support Ventilation; BAL, bronchoalveolar lavage; Cdyn, dynamic compliance; CPIS, clinical pulmonary infection score; CRP, C-reactive protein; Cstat, static compliance; ICU, intensive care unit; IL, interleukin; intraop, intraoperative; LOS, length of stay; Lt, left; MV, minute ventilation; NA, not available; NE, neutrophil elastase; OI, oxygen index; OLV, one-lung ventilation; op-time, operative time; Pa_CO2, arterial carbon dioxide partial pressure; PCV, pressure-controlled ventilation; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen ratio; Pmean, mean airway pressure; Ppeak, peak airway pressure; Pplat, plateau airway pressure; PRVC, Pressure-Regulated Volume Control; Re-ETT, reintubation; Ref, reference; TLV, two-lung ventilation; TNF, tumour necrosis factor; VATS, video-assisted thoracoscopic surgery; VCV, volume-controlled ventilation; ↑, significant increase; ↓, significant decrease; ↔, no significant difference.

Study model	Procedure(n)	Mechanical ventilator		OLV duration (min)	Outcome					Interpretation	Ref
					Lung mechanics and ABG (intraop)	Cytokine			Clinical outcomes
		Tidal (ml kg−1)	PEEP (cm H2O)			Lung		Serum
						Ventilated	Non-ventilated
In vivo study
Pigs	Lt thoracotomy pneumonectomy(15)	AC (ASV) 60% MV of TLV vs VCV 20	NA	180	↓Ppeak, ↔Cstat	BAL↓TNF-⍺, ↓IL-1β, ↓IL-6, ↓IL-8	NA	↓IL-1β, ↓IL-6	NA	ASV mode resulted in reduced local and systemic inflammation in healthy pigs.	50
Clinical study
Human	Thoracotomy pulmonary resection(70)	AC (PRVC) 6 vs VCV 6	10	83 vs 81	↓Ppeak, ↓Pplat, ↔Pmean, ↑P/F ratio, ↔PaCO2	BAL↓TNF-⍺, ↓IL-8, ↓albumin, ↓cell	NA	↓TNF-⍺, ↓IL-8	↓Pulmonary complication, ↔pneumonia, ↔ARDS, ↔atelectasis, ↔ARF, ↓ICU stay, ↓LOS, ↔mortality	PRVC mode enhanced oxygenation, reduced airway pressures, and decreased both local and systemic inflammation, leading to reduced pulmonary complications and hospital stay duration.	51
Human with lung cancer	VATS pulmonary resection(60)	AC (PRVC) 6 vs VCV 6	0	≥120	↓Ppeak, ↓Pplat, ↔Pmean, ↔P/F ratio, ↔PaCO2	NA	NA	↓TNF-⍺, ↓IL-6, ↑IL-10	NA	PRVC mode reduced airway pressures and systemic inflammation.	52
Human	VATS pulmonary resection(176)	AC (PRVC) 5–6 vs VCV 5–6	5	133 vs 160	↓Ppeak, ↔Pmean, ↔Cdyn, ↔P/F ratio, ↔PaCO2	NA	NA	↓NE	↔Pneumonia, ↔atelectasis, ↔ARF, ↔ICU stay, ↔LOS	PRVC mode reduced airway pressures and systemic inflammation. When combined with individualised PEEP, it resulted in a shorter ICU stay.	29
		AC (PRVC) 5–6 vs VCV 5–6	8 (decremental PEEP) vs 10 (decremental PEEP)	120 vs 135	↓Ppeak, ↔Pmean, ↔Cdyn, ↔P/F ratio, ↔PaCO2	NA	NA	↓NE	↔Pneumonia, ↔atelectasis, ↔ARF, ↓ICU stay, ↔LOS
Human with older age and lung cancer	VATS pulmonary resection(50)	AC (PRVC) 6 vs VCV 6	5	123 (op-time) vs 118 (op-time)	↓Ppeak↑Cdyn↔P/F ratio↔PaCO2	NA	NA	↔TNF-⍺, ↔IL-6, ↔IL-8, ↓NE	↔Re-ETT, ↔lung infection, ↔LOS	PRVC mode improved lung mechanics and reduced systemic inflammation, with no difference in clinical outcomes.	53
Human with obesity	Thoracotomy pulmonary resection(36)	AC (AutoFlow) 7 vs VCV 7	7 vs 0	125 vs 140	↓Ppeak, ↓Pplat, ↔Cdyn, ↑OI, ↓PaCO2	NA	NA	↓TNF-⍺, ↓IL-6	↔CPIS	Autoflow ventilation reduced airway pressures, enhanced oxygenation, and decreased systemic inflammation.	35
Human with oesophageal cancer	Thoracotomy oesophagectomy(180)	PCV <8 vs VCV <8	5 vs 0	72 vs 75	↓Ppeak, ↓Pplat, ↓Pmean, ↑P/F ratio, ↔PaCO2	BAL from both lungs↓TNF-⍺, ↓IL-6, ↓CRP		NA	↓Pulmonary complication	PCV with low tidal volume reduced airway pressures, enhanced oxygenation, and decreased inflammation, with fewer complications.	17

Effect of non-ventilated lung management strategies on inflammatory response during one-lung ventilation

In thoracic surgery, the lung on the operative side is typically collapsed to ensure adequate surgical exposure. Collapsed lungs are vulnerable to atelectasis, ischaemia–reperfusion, re-expansion injury, and surgical manipulation. Atelectasis activates immune and cellular responses, triggering local inflammation, whereas subsequent re-expansion after prolonged collapse adds mechanical stress and oxidative damage, further worsening pulmonary injury.^4,5 Various strategies have been explored to prevent injury to the non-ventilated lung during OLV.

Partially collapsed lung

Liu and colleagues⁵⁴ investigated the effect of controlled lung collapse in a canine model using a specially designed pulmonary sequestration tube, a novel single-lumen device that controls the degree of bronchial occlusion through adjustment of sealing elements within the bronchial cavity. This allows partial alveolar ventilation to be retained on the operative side while maintaining an adequate surgical field. Animals were divided into three groups according to the degree of collapse: complete collapse (group 1), controlled collapse with 90% collapse (group 2 [G2]), and controlled collapse with 50% collapse (group 3 [G3]). Compared with complete collapse, both controlled collapse groups demonstrated significantly lower serum concentrations of TNF-α, intercellular adhesion molecule-1, and IL-6, and less histological lung injury. No significant differences were found between G2 and G3 in any measured parameter. These findings suggest that partial ventilation of the operative lung during OLV reduces atelectasis and attenuates inflammatory responses compared with complete lung collapse, regardless of the degree of partial ventilation maintained. However, it should be noted that this technique was achieved using a purpose-built experimental device not currently available in standard clinical practice, which limits the direct applicability of these findings.

Application of CPAP

Another strategy involves applying CPAP to the non-ventilated lung. In a rat lung model study by Tojo and colleagues,⁵⁵ 5 cm H₂O of CPAP was applied to the non-ventilated lung, and outcomes were compared with leaving it open to the atmosphere. The CPAP group demonstrated improved oxygenation during OLV, along with significantly lower concentrations of IL-6, C-C chemokine ligand 2 (CCL-2), and C-X-C chemokine ligand 1 (CXCL-1) in lung tissue. Both CCL-2 and CXCL-1, produced by alveolar epithelial cells after injury, play a key role in recruiting macrophages, fibroblasts, and neutrophils, contributing to tissue damage and fibrosis.^56,57 This study demonstrated that application of CPAP could help prevent further lung injury. Interestingly, the study also found that CPAP application reduces cytokine concentrations in the ventilated lung, suggesting that its effects extend to both lungs. In a clinical study, this strategy was also evaluated during robotic-assisted thoracoscopic oesophagectomy. A CPAP of 5 cm H₂O was applied to the non-ventilated lung during OLV. BALF from the non-ventilated lung showed lower concentrations of inflammatory markers. All measured markers were significantly reduced in the CPAP group, demonstrating its beneficial effects on local inflammation.⁵⁸ However, the concentrations of serum inflammatory markers did not differ significantly between groups, leaving the impact of CPAP on systemic inflammation uncertain. In addition, there were no significant differences in clinical outcomes including pneumonia, ICU stay, and length of hospital stay between groups. There are a few concerns regarding the application of CPAP. One major issue is its potential impact on surgical exposure: more than half of the patients in the CPAP group experienced disturbances in the surgical field. Another limitation is that CPAP may only be suitable for non-pulmonary resection procedures. During pulmonary resections, an expanded lung can create tension on the lung parenchyma, increasing the risk of complications during resection.

Across the reviewed studies, both partial lung collapse and CPAP application to the non-ventilated lung consistently attenuated local inflammatory responses at the surgical site. The mechanistic basis relates to two principal pathophysiological processes occurring in the non-ventilated lung during OLV. First, complete lung collapse promotes atelectasis and re-expansion injury upon restoration of ventilation, generating local inflammatory mediator release. Second, HPV-induced hyperperfusion of the ventilated lung predisposes it to oedema and tissue damage. By maintaining partial alveolar patency, both strategies mitigate atelectasis, attenuate hyperperfusion, and limit re-expansion injury, with evidence suggesting that these anti-inflammatory benefits extend to the ventilated lung. However, interpretation of these findings is significantly constrained by the quality of available evidence, given that the evidence supporting partial lung collapse derives exclusively from animal models, whereas clinical evidence for CPAP is limited to a single oesophagectomy study, restricting any translation to broader thoracic surgical practice. CPAP application was also associated with surgical field disturbance in a significant proportion of patients and may be unsuitable during pulmonary resection procedures. The impact on systemic inflammation remains unclear, and no significant differences in clinical outcomes were demonstrated. Although these strategies show mechanistic promise, further well-designed clinical studies are needed to establish their efficacy, safety, and optimal application before definitive recommendations can be made. Reports from these in vivo and clinical investigations are presented in Table 4.

Table 4.

Effect of non-ventilated lung management strategies on inflammatory response during OLV: reports from in vivo and clinical investigations. ABG, arterial blood gas; BAL, bronchoalveolar lavage; CCL-2, C-C chemokine ligand 2; Com, compliance; CPAP, continuous positive airway pressure; CXCL-1, C-X-C chemokine ligand 1; G, group; ICAM, intercellular adhesion molecule; ICU, intensive care unit; IL, interleukin; intraop, intraoperative; LOS, length of stay; Lt, left; MV, mechanical ventilation; NA, not available; OLV, one-lung ventilation; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen ratio; Ppeak, peak airway pressure; Pplat, plateau airway pressure; Ref, reference; Rt, right; TNF, tumour necrosis factor; VT, tidal volume; ↑, significant increase; ↓, significant decrease; ↔, no significant difference.

Study model	Procedure(n)	Mechanical ventilator		OLV duration (min)	Outcome					Interpretation	Ref
					Lung mechanics and ABG (intraop)	Cytokine			Clinical outcomes
		Tidal (ml kg−1)	PEEP (cm H2O)			Lung		Serum
						Ventilated	Non-ventilated
In vivo study
Dogs	Thoracotomy(18)	G1: 100% collapsed, no ventilation (reference)		120	G2 and G3 vs G1: ↑P/F ratioG2 vs G3: ↔	NA	NA	G2 and G3 vs G1: ↓TNF-⍺, ↓ICAM, ↓IL-6G2 vs G3: ↔	NA	A partially collapsed, partially ventilated lung (G2 and G3) improved oxygenation and caused less systemic inflammation than a completely collapsed lung (G1) in healthy dogs. No differences were observed between G2 and G3.	54
		G2: 90% collapsed, 10% ventilation
		G3: 50% collapsed, 50% ventilation
Rats	Thoracotomy(32)	Lt 6	Rt 5 vs Rt open to atmosphere	240	↔Ppeak, ↑P/F ratio	Tissue (Lt)↔TNF-⍺, ↔IL-1β, ↓IL-6, ↔IL-10, ↓CCL-2, ↓CXCL-1	Tissue (Rt)↔TNF-⍺, ↔IL-1β, ↓IL-6, ↔IL-10, ↓CCL-2, ↓CXCL-1	NA	NA	Application of CPAP to a collapsed lung during OLV reduced local inflammation in both lungs of healthy rats.	55
Clinical study
Human with oesophageal cancer	Robotic thoracoscopic oesophagectomy(30)	Lt VT 6	Rt CPAP 5 vs Rt no CPAP	150 vs 170	↔Ppeak, ↔Pplat, ↔Com	NA	BAL (Rt)↓TNF-⍺, ↓IL-1⍺, ↓IL-1β, ↓IL-6, ↓IL-8, ↓IL-10	↔TNF-⍺, ↔IL-1⍺, ↔IL-1β, ↔IL-6, ↔IL-8, ↔IL-10	↔MV duration, ↔pneumothorax, ↔pneumonia, ↔ICU stay, ↔LOS	Application of CPAP to a collapsed lung during OLV reduced local inflammation but had no effect on systemic inflammation or clinical outcomes.	58

Effect of miscellaneous interventions on perioperative inflammatory response during one-lung ventilation

Fraction of oxygen

Oxygen supplementation is commonly used to maintain adequate tissue oxygenation, particularly during periods of physiological stress.⁵⁹ In thoracic surgery, V/Q mismatch and hypoxaemia frequently occur, necessitating adjustments in FiO₂ to prevent such complications.¹⁰ However, it has been shown that excessive oxygen administration promotes the generation of ROS, which can lead to oxidative stress and tissue injury, highlighting the importance of optimising oxygen delivery to balance its benefits with potential adverse outcomes.60, 61, 62 In an in vivo study using a pig model, Olivant Fisher and colleagues⁶³ compared the effects of hyperoxia (FiO₂ 1.0) and normoxia (FiO₂ ≤0.5) during OLV. After 180 min of OLV, lung tissue from the normoxia group exhibited significantly lower concentrations of pro-inflammatory cytokines. In contrast, hyperoxia was associated with elevated protein carbonyl concentrations, increased myeloperoxidase concentrations, and reduced superoxide dismutase activity, indicative of increased ROS production and oxidative tissue damage. Histological analysis further revealed an increased presence of red blood cells in the alveoli under hyperoxic conditions, suggesting vascular leakage or damage. A similar study in rabbit models yielded comparable results, showing that lower FiO₂ during OLV reduced local inflammatory responses. Histological analysis revealed significantly higher lung injury scores in the high FiO₂ group.⁶⁴ Collectively, these results suggest that high FiO₂ during OLV exacerbates local tissue inflammation and oxidative damage via ROS production during OLV. Large observational cohort studies have reported a correlation between supratherapeutic oxygen administration during surgery and an increased risk of postoperative lung injury. Subgroup analyses of thoracic procedures further demonstrated that high concentrations of oxygen supplementation were associated with an increased risk of lung injury.⁶⁰ Similarly, a retrospective cohort study by Frei and colleagues⁶¹ demonstrated that higher intraoperative oxygen supplementation was associated with an increased incidence of postoperative respiratory complications. These effects are not limited to direct damage to local tissue caused by ROS production. Elevated partial pressures of oxygen can contribute to lung atelectasis by increasing alveolar oxygen absorption. Recent meta-analyses have confirmed a correlation between high intraoperative oxygen supplementation and an increased risk of postoperative lung atelectasis. These findings highlight the importance of carefully titrating oxygen delivery to ensure adequate tissue oxygenation while minimising the risks of oxygen toxicity.⁶⁵ To minimise the risk of oxygen toxicity while ensuring adequate oxygenation, FiO₂ should be carefully adjusted to the lowest concentration that maintains sufficient oxygen delivery.

Inspiration-to-expiration ratio

During short inspiratory times, airflow predominantly reaches low-resistance alveoli, potentially leaving high-resistance, diseased alveoli under-recruited as a result of early cessation of inspiration. Inverse ratio ventilation (IRV), characterised by an inspiratory-to-expiratory (I:E) ratio greater than 1, is used to overcome this limitation in patients with lung injury. By prolonging the inspiratory phase, IRV promotes more uniform alveolar recruitment, particularly in regions with uneven ventilation.⁶⁶ Owing to its beneficial effects, the strategy of IRV was initially used in patients with ARDS to enhance oxygenation. In recent years, the application of IRV has expanded to the perioperative period. In particular, in the context of laparoscopic surgery, the evidence suggests that IRV not only improves oxygenation but also enhances functional residual capacity and lung mechanics.^67,68 Within the scope of this review, Zhang and Zhu⁶⁹ investigated the effects of IRV during OLV in patients undergoing pulmonary resection. The study compared two groups of patients who were ventilated with identical settings, with the exception of differing I:E ratios (1:2 in the control group and 2:1 in the IRV group). The IRV group demonstrated improved oxygenation and enhanced lung mechanics. The BALF from the ventilated lung indicated lower concentrations of pro-inflammatory cytokines, including IL-1β, IL-6, and IL-8. The prolonged inspiratory time in the IRV group generated an auto-PEEP of approximately 1.5 cm H₂O, resulting in a total PEEP of 6.5 cm H₂O in comparison with 5 cm H₂O in the control group. This higher PEEP in the IRV group is likely to have contributed to improved lung compliance and oxygenation, resembling the effect of individualised PEEP titration. In addition, the IRV group exhibited lower plateau pressures, which translated into reduced driving pressures. These combined effects suggest that IRV might reduce atelectotrauma and barotrauma during OLV. Unfortunately, clinical outcomes were not reported in the study. However, the clinical application of IRV must be carefully considered in thoracic surgical patients, who frequently present with underlying obstructive lung disease. The shortened expiratory time inherent to IRV may lead to progressive gas trapping, dynamic hyperinflation, and unintended auto-PEEP accumulation. These can result in haemodynamic compromise and barotrauma, risks that are significantly amplified in patients with increased airway resistance during OLV.^70,71 In conclusion, current evidence suggests that IRV can reduce both local inflammation and the systemic inflammatory response. However, there are limited studies investigating the application of IRV during OLV. Further research is needed to confirm the effects of IRV on the inflammatory response and its correlation with clinical outcomes. All available data regarding the inflammatory effects of various miscellaneous ventilation strategies are presented in Supplementary Table 2.

Effect of non-intubated video-assisted thoracic surgery on perioperative inflammatory response during one-lung ventilation

Despite the advances in lung-protective ventilation strategies described in the preceding sections, all positive pressure ventilation techniques carry inherent risks of mechanical lung injury. Therefore, non-intubated video-assisted thoracic surgery (NIVATS) has emerged as an alternative approach that avoids many of these challenges. NIVATS is described in the literature using several terms, including awake video-assisted thoracic surgery, non-intubated thoracic surgery, and spontaneous-ventilation VATS (SV-VATS).72, 73, 74 Although the terminology varies, the defining feature common to all these approaches is the avoidance of endotracheal intubation and positive pressure mechanical ventilation while maintaining spontaneous ventilation throughout the procedure. NIVATS encompasses a spectrum of anaesthetic techniques, ranging from true awake surgery under regional and local anaesthesia to general anaesthesia with spontaneous ventilation maintained via a laryngeal mask airway or face mask. By avoiding endotracheal intubation and positive pressure mechanical ventilation, NIVATS eliminates the associated risk of VILI.72, 73, 74 These advantages suggest that NIVATS may help reduce the potential increase in the perioperative inflammatory response during OLV. In an in vivo study in a rabbit lung model, the effect of spontaneous ventilation was investigated during thoracotomy. Over a 240-min intervention period, BALF from the ventilated lung in the spontaneous-ventilation group demonstrated lower TNF-α concentrations, whereas histological analysis revealed reduced inflammatory changes and decreased the level of expression of TNF-α messenger RNA in lung tissue, indicating a mitigation of the inflammatory response in comparison with OLV.⁷⁵ However, caution is warranted when interpreting these findings in a clinical context, given that anatomical differences between rabbits and humans necessitate endotracheal intubation without neuromuscular blocking agents, which may not fully replicate the conditions of NIVATS in humans. With regard to the clinical aspects, Vanni and colleagues⁷⁶ investigated the benefits of NIVATS in patients with benign lung diseases. Although there were no significant differences in perioperative complications or hospital stay between the NIVATS and control groups, the NIVATS group demonstrated less suppression of natural killer (NK) cell activity and total lymphocyte count after surgery than the control group. Similarly, another group study examined the immune and inflammatory responses in patients with pulmonary oligometastasis undergoing NIVATS. The results of this study were in alignment with previous findings, revealing the reduced suppression of lymphocyte count and NK cell activity in the NIVATS group. Postoperative serum IL-6 concentrations were also significantly lower in the NIVATS group than in the controls. Clinically, the NIVATS group also showed a lower incidence of major morbidity and shorter hospital stays.⁷⁷ One study provided detailed mechanical ventilator settings and found that, even when a protective ventilation strategy was used during OLV, NIVATS resulted in a reduced systemic inflammatory response compared with protective ventilation. In particular, serum TNF-⍺ and IL-6 concentrations were significantly lower in the NIVATS group after surgery. However, there were no significant differences between the groups in terms of postoperative opioid consumption or duration of hospital stay.⁷⁸ NK cells are known to play a significant role in immune surveillance and have been implicated in influencing malignancy outcomes.⁷⁹ Many recent studies demonstrated that a higher number of NK cells was associated with better oncologic outcomes.80, 81, 82 Although the preservation of NK cell activity observed with NIVATS is theoretically promising, direct evidence pertinent to oncologic outcomes is limited but emerging. Zheng and colleagues⁸³ conducted a propensity score-matched retrospective study of patients with non–small cell lung cancer undergoing SV-VATS vs mechanical ventilation VATS lobectomy. The results demonstrated significantly better overall survival and disease-free survival rates in the SV-VATS group at a median follow-up of 4.78 yr, with the benefit most pronounced in stage III patients. However, these findings should be interpreted with caution, given that the survival advantage may reflect the combined influence of several associated factors rather than the avoidance of positive pressure ventilation alone.

NIVATS represents a technique that is an option in thoracic surgery, offering an approach to perioperative inflammatory modulation by avoiding positive pressure ventilation, thereby removing the primary mechanical trigger for VILI. There is pre-clinical evidence to support the attenuation of local inflammatory responses in the dependent lung through this mechanism, whereas the available clinical evidence demonstrates corresponding reductions in systemic inflammatory markers and preservation of immune function. The oncologic implications of these have been discussed previously. With respect to clinical outcomes, the evidence remains limited and mixed, consistent with the multifactorial influences on postoperative recovery. The applicability of NIVATS is further limited by the requirement for specific surgical and anaesthetic expertise and careful patient selection, limiting its generalisability across broader thoracic surgical practice. Although the biological rationale for NIVATS as an inflammatory-attenuating strategy is compelling, further larger well-designed clinical studies incorporating both local and systemic inflammatory endpoints alongside validated clinical outcomes are needed to confirm its benefit. The effects of NIVATS on the perioperative inflammatory response during OLV, based on in vivo and clinical studies, are presented in Table 5.

Table 5.

Effect of non-intubated video-assisted thoracic surgery on perioperative inflammatory response during OLV: report from in vivo and clinical investigations. ABG, arterial blood gas; BAL, bronchoalveolar lavage; CD, cluster of differentiation; CRP, C-reactive protein; FiO₂, fraction of inspired oxygen; IL, interleukin; intraop, intraoperative; LOS, length of stay; mRNA, messenger RNA; NA, not available; NIVATS, non-intubated video-assisted thoracoscopic surgery; NK, natural killer cell; OLV, one-lung ventilation; op-time, operative time; Pa_CO2, arterial carbon dioxide partial pressure; P/F ratio, arterial oxygen partial pressure to fractional inspired oxygen ratio; Ppeak, peak airway pressure; PS, pathological score; Ref, reference; Rt, right; SV, spontaneous ventilation; TNF, tumour necrosis factor; VATS, video-assisted thoracoscopic surgery; VT, tidal volume; WBC, white blood cell; ↑, significant increase; ↓, significant decrease; ↔, no significant difference.

Study model	Procedure/(n)	Mechanical ventilator		OLV duration (min)	Outcome					Interpretation	Ref
					Lung mechanics and ABG (intraop)	Cytokine			Clinical outcomes
		Tidal (ml kg−1)	PEEP (cm H2O)			Lung		Serum
						Ventilated	Non-ventilated
In vivo study
Rabbits	Rt thoracotomy pulmonary resection(36)	SV-to-FiO2 ratio 0.4 vs OLV-to-VT ratio 10 and FiO2 0.4	NA vs 0	NA vs 240	↑P/F ratio, ↑PaCO2	BAL ↓TNF-⍺ Tissue ↓TNF-⍺ mRNA, ↓PS	NA	NA	NA	SV was associated with improved oxygenation and decreased local inflammatory response in healthy rabbits.	75
Clinical study
Human with benign lung disease	VATS(50)	Venturi mask vs OLV-to-Ppeak ≤30	NA vs 5	44 (op-time) vs 43 (op-time)	NA	NA	NA	↑Total lymphocyte, ↔CD3, ↔CD4, ↔CD8, ↔CD19, ↑NK	↔Complication, ↔LOS	NIVATS resulted in less suppression of lymphocyte count and function.	76
Human with pulmonary metastasis	VATS pulmonary resection(68)	Venturi mask vs OLV	NA	NA	NA	NA	NA	↓IL-6, ↔IL-10, ↔WBC, ↑total lymphocyte, ↑NK	↓Morbidity, ↓LOS, ↔mortality	NIVATS resulted in reduced inflammation and caused less suppression of lymphocyte count and function.	77
Human with lung cancer	VATS pulmonary resection(40)	Mask with bag 5–10 L min−1vs OLV 6	NA vs 5	NA vs 151	NA	NA	NA	↓TNF-⍺, ↔IL-1β, ↔IL-4, ↓IL-6, ↔IL-10, ↔CRP, ↔WBC	↔Opioid usage, ↔LOS	NIVATS resulted in reduced systemic inflammation.	78

Clinical implications

The findings of this review support a multimodal perioperative ventilation approach in which several complementary strategies may be applied to minimise inflammatory responses during OLV. Low VT ventilation, defined in this review as 5–6 ml kg⁻¹, remains the cornerstone of lung-protective ventilation during OLV, representing the most consistently supported strategy across the available evidence. Individualised PEEP titration, guided by lung mechanics rather than fixed values, offers additional inflammatory benefit beyond low VT alone, minimising the risk of overdistension. An ARM serves as a complementary strategy to optimise lung aeration, although its independent contribution remains difficult to isolate, given its frequent application as part of a combined strategy. AC ventilation may offer incremental inflammatory benefit over conventional VCV through its pressure-limiting properties. Careful titration of inspired oxygen concentration is recommended to balance adequate oxygenation against the risk of hyperoxia-induced oxidative stress. IRV may further improve oxygenation and attenuate local inflammatory responses through uniform alveolar recruitment and auto-PEEP generation. However, its clinical application during OLV requires careful patient selection given the risk of progressive gas trapping and haemodynamic compromise, and evidence is currently limited, with no clinical outcomes reported. Management of the non-ventilated lung through partial lung collapse or CPAP application may further attenuate local inflammatory responses at the surgical site, although current evidence is predominantly pre-clinical, and clinical applicability remains limited. Where feasible and in appropriately selected patients, NIVATS represents an option for further attenuation of the perioperative inflammatory response by avoiding positive pressure ventilation entirely, although clinical evidence remains limited. Although these strategies demonstrate reasonably consistent attenuation of perioperative inflammatory responses, their potential for translation into clinical outcomes was mixed across the reviewed studies. This likely reflects the complexity of the perioperative environment in thoracic surgery, where ventilation strategy represents one important but not isolated modifiable factor amongst many, including surgical complexity, operative duration, anaesthetic technique, and individual patient characteristics. Therefore, these strategies are most effective when applied collectively as part of a comprehensive perioperative care pathway, and further well-designed multicentre studies with standardised inflammatory and clinical outcome measures are needed to establish evidence-based guidelines for perioperative ventilation during OLV.

Limitations

Several limitations of the current evidence base warrant acknowledgement. A significant proportion of the reviewed studies used animal models, which, while providing valuable mechanistic insights, may not fully replicate human perioperative physiology, and therefore, conclusions drawn from pre-clinical evidence should be regarded as hypothesis generating rather than directly applicable to clinical practice. Most clinical studies were small single-centre trials with limited sample sizes, rendering them underpowered to detect meaningful differences in clinical outcomes. Considerable heterogeneity exists across included studies with respect to study design, patient population, surgical procedure, ventilation protocols, and selection of inflammatory markers, significantly limiting cross-study comparability. As is consistent with narrative review methodology, formal risk of bias assessment was not performed. Furthermore, the absence of standardised cytokine measurement methodologies across studies limits the comparability of inflammatory findings. These limitations collectively underscore the need for larger, multicentre, well-designed prospective studies with standardised outcome measures to establish definitive evidence-based recommendations for perioperative ventilation strategies during OLV.

Gaps in current knowledge and future direction

Several important knowledge gaps remain that warrant further investigation. The clinical significance of perioperative inflammatory marker reductions as surrogate endpoints has not been definitively established, and future studies should incorporate both the production of inflammatory markers and validated clinical outcomes as co-primary endpoints. Optimal targets for individual ventilation parameters during OLV, including VT, PEEP level, and ARM protocol, remain incompletely defined, and a standardised ARM protocol in particular is needed to enhance reproducibility across future studies. The independent inflammatory contribution of each protective ventilation component has rarely been examined in isolation, and future studies examining individual components independently would clarify their relative contributions. The evidence base for partial lung collapse, CPAP application to the non-ventilated lung, and IRV during OLV remains critically limited, predominantly derived from animal models or single clinical studies, and clinical validation across broader thoracic surgical procedures is needed. No studies have directly compared NIVATS with conventional lung-protective ventilation in terms of perioperative inflammatory response, representing an important priority for future randomised controlled trials. Finally, standardisation of cytokine measurement methodologies and consistent reporting of confounding variables, including anaesthetic techniques and OLV duration, are needed to enhance the reproducibility and clinical relevance of future research in this field. Extended follow-up periods to assess long-term clinical impact are vital in this area of research.

Conclusion

General thoracic surgery is a complex specialty where perioperative ventilation management plays a pivotal role in the determination of surgical outcomes. OLV induces significant physiological consequences that trigger both local and systemic inflammatory responses. The findings of this narrative review demonstrate that protective ventilation strategies, particularly in low VT ventilation and individualised PEEP titration, most consistently attenuate perioperative inflammatory responses during OLV. ARM, AC ventilation, FiO₂ titration, IRV, and non-ventilated lung management represent complementary strategies with varying levels of supporting evidence. In appropriately selected patients, NIVATS offers a unique approach to inflammatory attenuation by eliminating positive pressure ventilation entirely, although its applicability remains limited to specific procedures and patient populations. The translation of all these inflammatory benefits into clinical outcomes was inconsistent across the reviewed studies, reflecting both the limitations of inflammatory markers as surrogate endpoints and the multifactorial nature of postoperative recovery. Optimising perioperative ventilation strategies during OLV represents a clinically accessible and modifiable approach to the attenuation of lung injury and the inflammatory response, although their consistent translation into meaningful clinical benefit remains to be established.

Authors’ contributions

Conceptualisation: AS, SC, NC

Data curation: AS, SC, NC

Formal analysis: AS, SC, NC

Investigation: AS, KC, KS, SS, SC, NC

Methodology: AS, KC, KS, SS, SC, NC

Project administration: SC, NC

Supervision: SC, NC

Validation: AS, SS, KS, SC, NC

Visualisation: AS, SC, NC

Writing—original draft: AS

Writing—review and editing: AS, KC, SS, KS, SC, NC

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used the artificial intelligence language tool ChatGPT to enhance English grammar and clarity during the manuscript preparation process. The scientific content and interpretations remain solely the responsibility of the authors. After using the tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Funding

The Distinguished Research Professor grant from the National Research Council of Thailand (N42A690147) to NC; the Chiang Mai University Centre of Excellence Award to NC.

Declaration of interests

The authors declare that they have no conflicts of interest.

Handling Editor: Susan M Goobie

Appendix A. Supplementary data

The following is the Supplementary data to this article: