Pharmacometrics (PM) and machine learning (ML) are both valuable for drug development to characterize pharmacokinetics (PK) and pharmacodynamics (PD). Pharmacokinetic/pharmacodynamic (PKPD) analysis using PM provides mechanistic insight into biological processes but is time- and labor-intensive. In contrast, ML models are much quicker trained, but offer less mechanistic insights. The opportunity of using ML predictions of drug PK as input for a PKPD model could strongly accelerate analysis efforts. Here exemplified by rifampicin, a widely used antibiotic, we explore the ability of different ML algorithms to predict drug PK. Based on simulated data, we trained linear regressions (LASSO), Gradient Boosting Machines, XGBoost and Random Forest to predict the plasma concentration-time series and rifampicin area under the concentration-versus-time curve from 0–24 h (AUC_0–24h) after repeated dosing. XGBoost performed best for prediction of the entire PK series (R²: 0.84, root mean square error (RMSE): 6.9 mg/L, mean absolute error (MAE): 4.0 mg/L) for the scenario with the largest data size. For AUC_0–24h prediction, LASSO showed the highest performance (R²: 0.97, RMSE: 29.1 h·mg/L, MAE: 18.8 h·mg/L). Increasing the number of plasma concentrations per patient (0, 2 or 6 concentrations per occasion) improved model performance. For example, for AUC_0–24h prediction using LASSO, the R² was 0.41, 0.69 and 0.97 when using predictors only (no plasma concentrations), 2 or 6 plasma concentrations per occasion as input, respectively. Run times for the ML models ranged from 1.0 s to 8 min, while the run time for the PM model was more than 3 h. Furthermore, building a PM model is more time- and labor-intensive compared with ML. ML predictions of drug PK could thus be used as input into a PKPD model, enabling time-efficient analysis.

Chronic rhinosinusitis is a common disease worldwide, and the frequently prescribed nasal sprays do not sufficiently deliver the topical medications to the target sites so that the final treatment in severe cases is surgery. Therefore, there is a huge demand to improve drug delivery devices that could target the maxillary sinuses more effectively. In the present study, different particle diameters and device pulsation flow rates, mainly used in pulsating aerosol delivery devices such as the PARI SINUS®, are considered to evaluate optimal maxillary sinus deposition efficiency (DE). Numerical simulations of the particle-laden flow using a large eddy simulation with a local dynamic k-equation sub-grid scale model are performed in a patient-specific nasal cavity. By increasing the pulsation flow rate from 4 l/min to 15 l/min, nasal DE increases from 37% to 68%. Similarly, by increasing the particle size from 1 µm to 5 µm, nasal DE increases from 34% to 43% for a pulsation flow rate of 4 l/min. Moreover, normalized velocity, vorticities, and particle deposition pattern in different regions of the main nasal cavity and maxillary sinuses are visualized and quantified. Due to the nosepiece placement in the right nostril, more particles penetrate into the right maxillary sinus than into the left maxillary sinus despite the maxillary ostium being larger in the left cavity. Lower pulsation flow rates such as 4 l/min improve the DE in the left maxillary sinus. The use of 3 µm particles enhances the DE in the right maxillary sinus as well as the overall total maxillary drug delivery.

Highly effective targeted therapies are available to treat noncommunicable chronic inflammatory skin diseases. In contrast, the exact diagnosis of noncommunicable chronic inflammatory skin diseases is complicated by its complex pathogenesis and clinical and histological overlap. Particularly, the differential diagnosis of psoriasis and eczema can be challenging in some cases, and molecular diagnostic tools need to be developed to support a gold standard diagnosis. The aim of this work was to develop a real-time PCR-based molecular classifier to distinguish psoriasis from eczema in formalin-fixed and paraffin-embedded-fixed skin samples and to evaluate the use of minimally invasive microbiopsies and tape strips for molecular diagnosis. In this study, we present a formalin-fixed and paraffin-embedded-based molecular classifier that determines the probability for psoriasis with a sensitivity/specificity of 92%/100%, respectively, and an area under the curve of 0.97, delivering comparable results to our previous published RNAprotect-based molecular classifier. The psoriasis probability, as well as levels of NOS2 expression, positively correlated with the disease hallmarks of psoriasis and negatively with eczema hallmarks. Furthermore, minimally invasive tape strips and microbiopsies were effectively used to differentiate psoriasis from eczema. In summary, the molecular classifier offers broad usage in pathology laboratories as well as outpatient settings and can support the differential diagnosis of noncommunicable chronic inflammatory skin diseases on a molecular level using formalin-fixed and paraffin-embedded tissue, microbiopsies, and tape strips.