Raman-based real-time dissolution prediction using a deterministic permeation model.

Area of Science:

• Pharmaceutical Sciences
• Drug Delivery Systems
• Process Analytical Technology (PAT)

Background:

• Traditional dissolution prediction relies on empirical methods like PLS regression.
• Mechanistic models are typically used for off-line formulation optimization.
• Real-time prediction of drug release profiles remains a challenge.

Purpose of the Study:

• To develop a deterministic permeation model (DPM) for predicting in vitro drug release from HPMC matrix tablets.
• To utilize Raman spectra data from Process Analytical Technology (PAT) for real-time dissolution prediction.
• To compare the DPM's performance against Artificial Neural Networks (ANNs).

Main Methods:

• A 3^3 full factorial design was employed to produce HPMC matrix tablets with varying API and HPMC amounts and compression pressures.
• Raman spectra were recorded for tablets prior to dissolution testing.
• A DPM, presented as a system of differential equations, was constructed and its parameters estimated using a training dataset.

Main Results:

• The DPM successfully predicted the in vitro release profiles of HPMC matrix tablets based on Raman spectra.
• Predictions from the DPM showed similar accuracy to those from an Artificial Neural Network (ANN).
• The mechanistic DPM required fewer training samples compared to the ANN.

Conclusions:

• The developed DPM offers a novel mechanistic approach for real-time prediction of drug release profiles.
• This model enhances understanding of HPMC and compression pressure effects on release kinetics.
• The DPM shows potential for developing PAT-based surrogate dissolution methods for Real-Time Release testing (RTRt).