In vitro validation and characterization of pulsed inhaled nitric oxide administration during early inspiration

Abstract

Purpose: Admixture of nitric oxide (NO) to the gas inspired with mechanical ventilation can be achieved through continuous, timed, or pulsed injection of NO into the inspiratory limb. The dose and timing of NO injection govern the inspired and intrapulmonary effect site concentrations achieved with different administration modes. Here we test the effectiveness and target reliability of a new mode injecting pulsed NO boluses exclusively during early inspiration.

Methods: An in vitro lung model was operated under various ventilator settings. Admixture of NO through injection into the inspiratory limb was timed either (i) selectively during early inspiration ("pulsed delivery"), or as customary, (ii) during inspiratory time or (iii) the entire respiratory cycle. Set NO target concentrations of 5-40 parts per million (ppm) were tested for agreement with the yield NO concentrations measured at various sites in the inspiratory limb, to assess the effectiveness of these NO administration modes.

Results: Pulsed delivery produced inspiratory NO concentrations comparable with those of customary modes of NO administration. At low (450 ml) and ultra-low (230 ml) tidal volumes, pulsed delivery yielded better agreement of the set target (up to 40 ppm) and inspiratory NO concentrations as compared to customary modes. Pulsed delivery with NO injection close to the artificial lung yielded higher intrapulmonary NO concentrations than with NO injection close to the ventilator. The maximum inspiratory NO concentration observed in the trachea (68 ± 30 ppm) occurred with pulsed delivery at a set target of 40 ppm.

Conclusion: Pulsed early inspiratory phase NO injection is as effective as continuous or non-selective admixture of NO to inspired gas and may confer improved target reliability, especially at low, lung protective tidal volumes.

Keywords: ARDS; Artificial lung; Inhalation; Mechanical ventilation; Nitric oxide; PiNO.

Fig. 1

Schematic setup of the artificial lung model. Mechanical breaths from the ventilator inflate lung 1 (valve A), thereby raising a lifting bar to cause aspiration of ambient air into lung 2. Expiration into the breathing system occurs from lung 2 (valve B), while lung 1 releases to ambient air. Sites of NO injection and sampling and the distance from the ventilator or Y-piece are indicated along the breathing circuit. Injection: 20 cm distal from the ventilator, 10 cm proximal of Y-piece. Sampling: 40 and 120 cm distal from ventilator, 4 cm proximal of Y-piece, Y-piece, mid-tracheal, and artificial lung 1. Figure adapted and modified from [13]

Fig. 2

a Inspiratory NO concentrations achieved with “pulsed delivery”. NO was injected as a bolus during early inspiration into the breathing circuit at 20 cm after the mechanical ventilator, sampled at 120 cm thereafter, and quantified with ozone-based chemiluminescence. Ultra-low (230 ml), low (450 ml) or traditional (750 ml) tidal volumes were applied via both pressure and volume-controlled ventilation with I:E ratios of 1:1 and 1:1.9 (4 conditions). Mean NO concentrations were determined for 120 s (n = 30 respiratory cycles) per each individual ventilation condition (i.e. n = 120 per bar representing 4 ventilation conditions). Means ± SD. *p < 0.05 vs. target. b Target reliability of NO administration modes. NO was administered via “pulsed”, “flow proportional” or “continuous delivery” through injection into the breathing circuit at 20 cm after the mechanical ventilator, sampled at 120 cm thereafter, and quantified with ozone-based chemiluminescence. For each NO target concentration, mechanical ventilation was performed with ultra-low (230 ml), low (450 ml) or traditional (750 ml) tidal volumes applied via both pressure and volume-controlled ventilation with I:E ratios of 1:1 and 1:1.9 (12 conditions). NO concentrations were determined for 120 s (n = 30 respiratory cycles) per each individual ventilation condition (i.e. n = 360 per data point representing 12 ventilation conditions; n = 354 for “flow proportional delivery” at 5 ppm). The level of agreement is depicted by the mean inspiratory NO concentration expressed as a percentage of the target value ± SD. *p < 0.05 between respective modes

Fig. 3

Sensitivity of NO quantification methods. The graph depicts the theoretical temporal resolution of ozone-based chemiluminescence (OBC, blue) versus an electrochemical sensor (ECS, black) in retracing fluctuations of the true NO concentration (red)