Twenty-four rabbits were anesthetized following the study protocol, with successful induction of ARDS models in all subjects. The rabbits were meticulously paired into 12 pairs based on age and weight and subsequently randomized into two groups: high driving pressure (HDP) and low driving pressure (LDP) groups. No discernible differences were observed between these groups concerning age, weight, and respiratory rate. However, a notable dissimilarity existed in tidal volume, with the LDP group exhibiting a significantly lower value compared to the HDP group (6.3 ± 0.68 ml/kg vs. 10.5 ± 0.85 ml/kg, P < 0.001).

Hemodynamic parameters

As reported in Table 1, fluid loading significantly increased the CVP (P < 0.001) and decreased stroke volume variation (SVV) (P < 0.001) in both groups. Notably, in the HDP group, there was a significant rise in CO following fluid loading (P < 0.001), a response not observed in the LDP group. Moreover, mean arterial pressure (MAP) increased significantly in the HDP group after fluid infusion (P = 0.029).

Table 1 Circulatory parameters under different driving pressure and circulating volume statusCyclic “on–off” flow of pulmonary capillaries

The comparison of SDF parameters, including PPV, MFI, TVD, and PVD, between inspiration and expiration in two groups is presented in Table 2 and illustrated in Fig. 2. The change between inspiration and expiration, denoted as ΔPPV, ΔMFI, ΔPVD, and ΔTVD, were detailed in Table 3. When evaluating SDF parameters at timepoint T1, it was observed that PPV (14.2 ± 3.2 vs. 53.6 ± 3.3, p < 0.001), MFI (0.50 ± 0.04 vs. 1.50 ± 0.06, p = 0.002), TVD (38.9 ± 1.1 vs. 42.2 ± 1.5, p < 0.001), and PVD (5.5 ± 1.3 vs. 21.1 ± 1.6, p < 0.001) significantly decreased more during inspiration than expiration in the high driving pressure group. PPV (50.0 ± 3.4 vs. 59.5 ± 1.9, p < 0.001), MFI (1.38 ± 0.37 vs. 1.75 ± 0.08, p = 0.007), TVD (41.8 ± 1.3 vs. 43.2 ± 0.9, p = 0.003), and PVD (19.3 ± 0.7 vs. 25.3 ± 0.6, p < 0.001) during inspiration was also smaller than expiration in the low driving pressure group at T1. Videos depicting the pulmonary microcirculation during both inspiration and expiration are provided in the Supplementary Materials. Videos 1–4 represent HDP at T1 during inspiration, HDP at T1 during expiration, LDP at T2 during inspiration, and LDP at T2 during expiration.

Table 2 SDF parameters under different driving pressure and circulating volume statusFig. 2

SDF parameters in two ventilation groups. SDF parameters, including PPV, MFI, TVD, and PVD, were assessed in two distinct groups. At the T1 timepoint, when both high and low driving pressure groups experienced low CVP, all indexes showed a significant increase during expiration compared to inspiration. After fluid loading at the T2 timepoint, ΔPPV, ΔMFI and ΔPVD (expiration–inspiration) narrowed both in the HDP and LDP group. PPV, proportion of perfused vessel; MFI, microvascular flow index; TVD, total vessel density; PVD, perfused vessel density. * indicates a statistically significant difference between inspiration and expiration. # indicates a statistically significant difference between T1 and T2

Table 3 Change between inspiration and expiration under different driving pressure and circulating volume statusEffect of driving pressure on cyclic "on–off" flow

At T1, during inspiration, PPV (14.2 ± 3.2 vs. 50.0 ± 3.4, p < 0.001), MFI (0.50 ± 0.04 vs. 1.38 ± 0.37, p < 0.001), and PVD (5.5 ± 1.3 vs. 19.3 ± 0.7, p < 0.001) were notably lower in the HDP group compared to the LDP group. At T2, although PPV (48.7 ± 8.9 vs. 80.6 ± 4.1, p < 0.001) and PVD (20.2 ± 2.7 vs. 32.6 ± 1.3, p < 0.001) during inspiration remained lower in the HDP group, the change between the two groups diminished. Additionally, ΔPPV (expiration–inspiration) at T1 in the HDP group was substantially greater than in the LDP group (38.03 ± 3.92 vs. 11.64 ± 5.42, P < 0.001). ΔMFI (expiration–inspiration) (0.96 ± 0.33 vs. 0.38 ± 0.27, p = 0.038), ΔTVD (expiration–inspiration) (3.29 ± 1.29 vs. 1.43 ± 1.19, p = 0.025) and ΔPVD (expiration–inspiration) 15.54 ± 1.86 vs. 6.12 ± 1.87 p < 0.001) had the same trend.

Effect of fluid loading on cyclic “on–off” flow

Following fluid loading, there was a notable increase in PPV (14.2 ± 3.2 vs. 48.7 ± 8.9, P < 0.001), MFI (0.50 ± 0.04 vs. 1.50 ± 0.18, P = 0.003), and PVD (5.5 ± 1.3 vs. 20.2 ± 2.7, P < 0.001) during inspiration in the HDP group, accompanied by a decrease in ΔPPV (38.03 ± 3.92 vs. 22.99 ± 5.53, P < 0.001), ΔMFI (0.96 ± 0.33 vs. 0.40 ± 0.20, P < 0.001), and ΔPVD (expiration–inspiration) (15.44 ± 1.86 vs. 11.07 ± 2.75, P < 0.001). ΔPPV (11.64 ± 5.42 vs. -1.06 ± 4.45, P < 0.001) and ΔPVD (expiration–inspiration) (6.12 ± 1.87 vs. 0.38 ± 2.11, P < 0.001) also decreased after fluid loading in the LDP group. However, ΔTVD did not exhibit a significant difference after fluid loading in both groups.

Markers of inflammation and vascular injury

We compared four biomarkers associated with inflammation and vascular endothelial injury, namely IL-6, TNF-α, Ang-2, and vWF, at three timepoints: T0 before ARDS induction, T1 at 60 min post-modeling, and T2 at 2 h post-modeling. The results indicated elevated biomarker levels at T1 and T2 compared to T0, affirming the efficacy of the modeling. Notably, at T2, the HDP group exhibited higher levels of IL-6 (151[141,157] vs. 113[99,137], P = 0.025), TNF-α (254[239,267] vs. 181[136,206], P = 0.006), Ang-2 (3016[2574,3409] vs. 2515[2316,2538], P = 0.049) and vWF (2242[2125,2330] vs. 1696[1632,1924], P = 0.037) than the LDP group, suggesting more severe inflammation and vascular endothelial injury in the former. In addition, in the LDP group, the levels of TNF-α, vWF, and Ang-2 were higher at T1 than at T2. The detailed biomarker results are depicted in Fig. 3.

Fig. 3

Cytokine profiling in response to ventilation strategies. The cytokine levels in two distinct groups were assessed at three critical timepoints during the experiment. Blood samples were collected at T0, before the induction of ARDS; T1, following 60 min of ventilation with low CVP; and T2, after an additional 60 min of ventilation with high CVP. The cytokines analyzed included TNF-α, IL-6, Ang-2, and vWF, with a focus on elucidating differences between the two groups. At the T1 timepoint, it was observed that the high DP group exhibited a significant elevation in TNF-α levels compared to the low DP group. However, no significant differences were noted in the other cytokines at this juncture. At the T2 timepoint, the high DP group displayed higher levels of TNF-α, IL-6,Ang-2 and vWF in comparison with the low DP group. In addition, in the LDP group, the levels of TNF-α, vWF, and Ang-2 were higher at T1 than at T2, while no significant difference was found between T1 and T2 in HDP group. TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; Ang-2, angiopoietin-2; vWF, von Willebrand Factor. * indicates a statistically significant difference between LDP group and HDP group. # indicates a statistically significant difference between T1 and T2

Pulmonary edema

Pulmonary edema, represented as the ratio of the lung wet weight to the body weight and the lung wet-to-dry weight ratio, revealed significant differences between the HDP and LDP groups. The comparison of pulmonary edema is shown in Fig. 4. The HDP group exhibited a notably higher lung wet weight/body weight ratio than the LDP group (1.48 ± 0.07 vs. 1.36 ± 0.09, P = 0.019). Similarly, the lung wet-to-dry weight ratio was higher in the HDP group (7.78 ± 0.49 vs. 5.72 ± 0.58, P = 0.007).

Fig. 4

Pulmonary edema comparison. Rabbits subjected to low DP group exhibited a significantly lower proportion of wet lung/body weight and lung wet–dry ratio than those exposed to high DP. * indicates a statistically significant difference between the LDP group and the HDP group