In vitro and in vivo performance modelling and optimisation of different dry powder inhalers: A complementary study of neural networks, genetic algorithms and decision trees.

INTRODUCTION: Aerosol delivery from DPIs could be affected by different factors. This study aimed to evaluate and predict the effects of different factors on drug delivery from DPIs. METHODS: Modelling and optimisation for both in vitro and in vivo data of different DPIs (Diskus, Turbohaler and Aerolizer) were carried out using neural networks associated with genetic algorithms and the results are confirmed using a decision tree (DT) and random forest regressor (RFR). All variables (the type of DPI, inhalation flow, inhalation volume, number of inhalations and type of subject) were coded as numbers before using them in the modelling study. RESULTS: The analysis of the in vitro model showed that Turbohaler had the highest emitted dose compared with the Diskus and the Aerolizer. Increasing flow resulted in a gradual increase in the emitted dose. Little differences between the inhalation volumes 2 and 4 litres were shown at fast inhalation flow, and interestingly two inhalations showed somewhat higher emitted doses than one-inhalation mode with Turbohaler and Diskus at slow inhalation flow. Regarding the in vivo model, the percent of drug delivered to the lung was highly increased with Turbohaler and Diskus in healthy subjects where continuous contour lines were observed. The Turbohaler showed increased lung bioavailability with the two-inhalation modes, the Diskus showed a nearly constant level at both one and two inhalations at slow inhalation. The Turbohaler and Aerolizer showed little increasing effect moving from one to two inhalations at slow inhalation. CONCLUSIONS: Modelling of the input data showed a good differentiating and prediction power for both in vitro and in vivo models. The results of the modelling refer to the high efficacy of Diskus followed by Turbohaler for delivering aerosol. With two inhalations, the three DPIs showed an increase in the percent of drug excreted at slow inhalations.

Inhaled salbutamol from aerolizer and diskus at different inhalation flows, inhalation volume and number of inhalations in both healthy subjects and COPD patients.

The aim of the present study was to demonstrate the effect of inhalation-flow, inhalation-volume and number of inhalations on aerosol-delivery of inhaled-salbutamol from two different dry powder inhalers (DPIs) in both healthy-subjects and chronic obstructive pulmonary disease (COPD) patients. Relative pulmonary-bioavailability and systemic-bioavailability of inhaled-salbutamol, delivered by Diskus and Aerolizer, was determined in 24-COPD patients and 24-healthy subjects. The healthy-subjects and the COPD-patients participated in the study for 7 days in which they received 4 study doses of 200 µg salbutamol (one slow-inhalation, two slow-inhalations, one fast-inhalation, and two fast-inhalations) in four alternative days with 24 hr washout period after each dose. Two urine-samples were collected from each study subjects. The first was provided 30 min post inhalation (USAL0.5), as an index of relative pulmonary-bioavailability, and the second was pooled to 24 hr post inhalation (USAL24), as an index of systemic-bioavailability. Fast-inhalation resulted in significantly higher USAL0.5 and USAL24 than slow-inhalation (p˂0.05) after one-inhalation in both healthy-subjects and COPD-patients but there was no significant difference between slow and fast-inhalation after two-inhalations. One-inhalation resulted in significantly higher USAL0.5 and USAL24 in healthy-subjects compared to COPD-patient at both slow and fast-inhalation (p˂0.05) except USAL0.5 with Diskus at slow-inhalation there was no significant difference. Also, two-inhalations resulted in significantly higher USAL0.5 and USAL24 compared to one-inhalation at slow-inhalation only (p˂0.05). No significant difference was found between Aerolizer and Diskus except in USAL0.5 of one slow-inhalation in both health-subjects and COPD-patients (p = 0.048 and 0.047, respectively). Device-formula relation is present at low inhalation-flow since Diskus resulted in significantly higher USAL0.5 and USAL24 in healthy-subjects compared to COPD-patient at slow inhalation than Aerolizer. It is essential to inhale-twice and as hard and deep as possible from each dose when using DPI especially with COPD-patients having poor inspiratory efforts such as elderly patients and children.

Total emitted dose of salbutamol sulphate at different inhalation flows and inhalation volumes through different types of dry powder inhalers.

The aim of the present study was to compare the performance of two different dry powder inhalers (DPIs) at different inhalations volumes and inhalation flows. Ventolin Diskus contain blisters of 200µg salbutamol. To test the TED from Aerolizer, salbutamol in Diskus blister was emptied and placed in size 3 capsules suitable for use with Aerolizer. Total emitted dose (TED) delivered by Diskus and Aerolizer was determined using DPI sampling apparatus after one and two inhalations from the same dose. 10-60L/min inhalation flows at 2 and 4L inhalation volume were used in the determination. At inhalation flow ≤30L/min, two inhalations resulted in higher TED than one inhalation (p < 0.05) and Diskus resulted in higher TED than Aerolizer (p < 0.05). The highest TED was at inhalation flow 40L/min above which the effect of the second inhalation and formula device relation were negligible. Device formula relation is present at low inhalation flow but at flow >30L/min Diskus drug formula can be delivered by Aerolizer with no significant difference in TED produced. For the best TED patients are required to inhale as fast as possible (a minimum of 40L/min). At lower inhalation flow two inhalations results in better emitted dose than one inhalation for both DPIs. So, we recommend patients with poor inspiratory efforts to inhale twice and as hard and deep as possible from each dose as they may not receive much benefit from one inhalation even when using DPI with low resistance (Aerolizer) or medium resistance (Diskus). However, further in-vivo study are required to validate this recommendtation.