To the best of our knowledge, this is the first study comparing the impact of nebulizer position and gas humidification on the regional aerosol deposition using an innovative and reproducible preclinical respiratory model. This work showed with comparative conditions that these two experimental parameters can greatly change the mean deposited fractions in the RT from 18 to 55%, leading to potential wide differences in drug delivery to mechanically ventilated patients. This finding raises a major concern about practices in ICU as some international epidemiological studies highlighted disparate clinical practices (2). Moreover, this study was performed with a new generation vibrating mesh nebulizer showing a mean aerosol output higher than 90%. This point agrees with previous data (35). This is of high relevance because epidemiological studies show a majority use in ICU of less efficient jet nebulizers (2, 36). For example Ehrmann et al observed 10% of vibrating mesh nebulizer and 56% of jet nebulizers use in an international study (2). Despite economic considerations, their output is largely lower, ranging from 30 to 55% according to different studies (35, 37). Consequently, considering wide differences in aerosol output depending on nebulizer technologies and the great impact of the aerosol device position on the ventilator circuit and gas humidification, the potential cumulative differences could lead to clinical treatment failure independently of drug efficiency. Finally, the type of nebulizer has been suspected to impact ICU patients outcome as a trends to shorter length of stay for asthmatic patients (10).
Impact of nebulizer position in the ventilator circuit
The location of the nebulizer in ventilator circuit is a major determinant for the drug dose reaching the RT (38). Indeed, despite a manufacturer recommendation (39), position upstream the heated humidifier (HH), (conditions D-1 and H-1) led in this study, to a mean RT deposited fraction ranging from 18–25% while a position right before the Y-piece adapter (conditions D-2 and H-2) produced a mean RT deposited fraction ranging from 53 to 57%. Obviously, this difference is largely significant and can lead to wide differences in the treatment efficacy. However, Moraine et al failed to find a clinical difference while measuring ipratropium in urine after nebulization in mechanically ventilated patients (40). According to our results, evaluating separately the impact of gas humidification and of nebulizer position on a standardized preclinical respiratory model, Moraine et al result could be due to the interaction of both factors. Indeed, they compared whether nebulizer prior the HH or nebulizer right before the Y-piece adapter in dry conditions, leading to potential interaction of these factors. Even if the difference is not significant, the nebulizer position beyond the Y-piece adapter tend to a higher loss in expiratory limb and lower RT deposited fraction than the position 15cm before Y-piece adapter (p = 0,0508). These data are in accordance with the in vitro study from Ari et al (7) which compared the three same positions with and without gas humidification on an in vitro model. Nevertheless, due to a design with deposition on a filter instead of a RT, the expiratory loss and regional deposition could not be evaluated on their model. The expiratory loss is described as the major drawback for positions closed to expiratory limb (38) this is consequently an important interrogation. On the other hand, position 2 is characterized by the 15cm before the Y-piece, serving as a reservoir for aerosol continuously generated during expiration. This position is often recommended for this effect (4, 12, 30, 41–46). Finally, this study finds an optimal position at 15cm before the Y-piece adapter for the vibrating mesh nebulizer independently from humidification, and this result is confirmed even considering the expiratory loss. As this is different from manufacturer recommendations (39) for this specific nebulizer, the intensivists have to be aware of this point.
Impact of gas humidification
No difference was shown for humidification in terms of absolute RT deposited dose. This point differs from previous data on the humidification impact on aerosol delivery, as an important decrease, around 40%, is widely described (47). This could be due to a difference between nebulizer types, as vibrating mesh nebulizers are relatively new technologies. Moreover, Ari & al, using the same nebulizer than us, found a significant decrease with humidification for the positions before the Y-piece adapter but no difference after this latter (7). More recently, Ashraf et al, comparing conditions very similar to D-1 and H-1 in our study, i.e. the same vibrating mesh nebulizer close to the ventilator with or without humidification, found very similar results with inhalable doses around 20% equally (35). Nevertheless, a difference arose in our results for RT distribution when the position upstream HH and dry condition interacts. Indeed, these parameters (condition D-1) allowed a more peripheral deposition compared to each other condition, as previously described (25). However, this finding must be first balanced with an absolute RT deposited fraction lower than other conditions and then with the risks for patients when interrupting the humidification for nebulization. Indeed, there are two principal ways of humidification for patients under mechanical ventilation: heat and moisture exchangers needing to be removed for the duration of nebulization to allow the passing of aerosol (48) and HH. The cessation of HH just before nebulization showed an important residual humidity in limbs, being consequently ineffective to obtain “dry conditions” (29) and the removal of humidification was reported as dangerous for patients if prolonged (49, 50). The humidification question has consequently to be seen as security factor for patients mechanically ventilated. Our finding is in accordance with a clinical study on asthmatic patients showing no significant difference for clinical endpoints according to humidification or not (10). This point raises questions about evidence to interrupt humidification during nebulization with this vibrating mesh nebulizer.
Aerosol size distribution
Interestingly, the findings in this study differ in terms of aerosol size distribution from Ashraf et al. data (35). These discrepancies, while puzzling at first, could be explained by the sampling choices made that differ from our own sampling method. Indeed, as Ashraf et al. chose to sample at the distal extremity of the endotracheal tube, we decided to sample the aerosol at the extremity of the Y-piece. However, this is not sufficient to explain these discrepancies. Another point that differs between the studies is the vacuum parameters applied to sample the aerosol. To determine the size distribution, Ashraf et al. decided to apply physiological-like sampling of the aerosol with a vacuum parameter similar to standard breathing pattern found in humans. In this study, we decided to follow our in-house bench protocol for aerosol size distribution, which is based on European Pharmacopeia monography. Therefore, these discrepancies are explained by the fact that we did not measure the same aerosol distribution. On one hand, Ashraf et al. presented granulometry data more focused on the granulometry of droplets that would efficiently reach out the distal extremities of the endotracheal tube. This method is likely to exclude the droplets which would be sedimented during the expiratory phase of the breathing-like sampling pattern, which would decrease aerosol size distribution as the biggest droplets would be excluded from the analysis. On the other hand, the data presented in this study are more focused on assessment of aerosol size distribution on bench with standardized parameters allowing the robust comparison of different nebulizers without exclusion of the biggest droplets composing the aerosol. One could argue on the advantages of one method over the other. Such assertion should be made with caution as the two sampling methods differ on the focus of the analysis. Ashraf et al. method is more patient-based, while the method presented in the study is more device-based.
Strength and limitations
Despite its interesting results, this work presents some limitations. First, as all studies using models, our ex vivo RT will never be able to mimic the complexity of a whole organism. Therefore, as preclinical data, they need to be confirmed by clinical studies. To date, results from this study seem in good agreement with the clinical study on ICU patients of Klockare et al (14) where with an ultrasound nebulizer (MMAD 4.0µm) and gas humidification, the central fraction of RT deposited fraction had a median of 62.8%. Moreover, considering that a vast majority of data for aerosol therapy provide from in vitro studies, this work adds important new information. Indeed, by the adjunction of a RT with an anatomy very close to human’s, data on regional deposition in the RT and exhaled particles are considered. Second, the model used is based on porcine RT collected from slaughterhouses and any animal was specifically sacrificed for this study. This point is equally a limitation because they presented cuts that generated leakage. This particular point was discussed on the previously published data and did not influence the homogeneity of the ventilation (25). Moreover, improvements were realized to minimize the potential leakage using cyanoacrylate glue and stitches and by a better sealing of the enclosure.
As previously stated, there is a lack of published clinical literature concerning the impact of the inhaled gas humidification and the nebulizer position on the deposition pattern of an aerosol when administered during mechanical ventilation. Among the few clinical studies, the endpoints mainly concern patients’ outcomes, such as improvement of the general state, weaning of mechanical ventilation, etc. or have difficulties to show differences, probably because of the multiplicity of influencing factors (40, 51). Moreover, most of the studies were conducted on patients under non-invasive ventilation, which is a different condition. Hence, more than 30 years after the development of ICU and aerosoltherapy, the nebulizer position in the ventilation circuit is mainly driven by empirical practices, as well as the practices concerning the humidification during nebulization. Moreover, nebulizer technologies have changed as well as ventilators and ventilation parameters. Therefore, there is a need for preclinical studies with a good reproducibility as observed here, allowing to compare humidification and position as isolated factors but equally with their interaction. Finally, these comparisons are quantitative but equally qualitative on the repartition in the RT. To our knowledge, this is the first study allowing all these comparisons.