Category: Formulation and Quality
Purpose: Surface-eroding polymers are ideal for LAIs due to their ability to provide long-term sustained release based on degradation rather than diffusion. When the polymer is non-swellable and impermeable to water, this allows for zero-order release rates that are dependent on the rate of polymer surface degradation. In this work, in vitro dissolution testing of caffeine-loaded poly(glycerol sebacate) urethane (PGSU) implantable rods is investigated using a USP IV flow-through cell apparatus. PGSU is a surface-eroding polymer that is suited for drug delivery applications that require sustained high loadings over a multi-month time period. As current polymer carriers are non-degradable or month-long bulk-eroders, a dissolution method and subsequent analytical technique is needed for multi-month surface-eroding LAIs like PGSU.
Methods: Dissolution of caffeine-loaded PGSU rods was completed using a Sotax flow-through cell apparatus USP IV closed-loop system with a peristaltic pump set to 8mL/min. Dissolution was evaluated using 0.1M phosphate buffered saline (PBS) at pH = 7.40 and 37.0°C. Weekly sampling employed a “sample and replace” strategy to maintain sink conditions. Samples were analyzed by HPLC. HPLC was completed using a Zorbax Eclipse XDB-C18 4.6 x 100mm, 3.5µm column and UV detection at 275nm. The mobile phase part A was 0.2% formic acid in water and part B was 0.2% formic acid in methanol. The flow rate was 0.7ml/min and used a gradient of 70% B for zero to two minutes and 100% B from two minutes to 8 minutes. In order to prevent continued hydrolysis of degradation products half of the weekly sample was analyzed as soon as possible after sampling and half of the sample was frozen as backup.
Results: In caffeine-loaded PGSU rods (Figure 1) linear cumulative release is observed over 49 days in most samples. At day 56 several samples begin to plateau, and the concentration begins to decrease. This suggests that there is an issue of caffeine stability in the dissolution apparatus. This appears to be caused by long term exposure to PGSU degradation products in the dissolution media. When a control solution of caffeine in PBS is run alongside the dissolution samples the concentration of caffeine remains constant. However, when a known solution of caffeine and PGSU degradation products is monitored alongside the dissolution samples for 13 weeks, no caffeine is detected.
The observed lack of sample stability indicates that samples must be analyzed or frozen quickly after sampling. This effect can be seen in shaded region of Figure 2, where the release curves from days 63 to 91 show large variations, both increasing and decreasing, between weeks. In this experiment HPLC analysis of the non-frozen samples was not able to be completed immediately. This might also explain the drop observed at day 28 in all samples except the one denoted in green. When the frozen samples were analyzed the curves are as expected with reasonable variation from week-to-week, shown in Figure 1.
The analysis of surface-eroding LAI dissolution samples presents a unique set of challenges not seen in other dosage forms. In non-degradables a simple UV-Vis spectrometer can be set up in-line to monitor API release, or analysis by simple HPLC methods. The addition of degradation products from a degradable polymer prevents the use of an in-line UV-Vis due to the changes in refractive index of the solution as the polymer degrades. One solution to this problem is to separate the API from the degradation products by HPLC. One of the main challenges in method development for separating API and PGSU degradation products is the constantly changing nature of the degradation species and concentration. Because one mechanism of PGSU degradation is hydrolysis degradation products may continue to hydrolyze in solution, therefore changing the mixture which needs to be separated. A continually changing sample matrix requires a method that is robust enough to handle these changes.
Conclusion: Using an USP IV flow-through cell apparatus, linear release was observed over 49 days for a range of caffeine loadings in PGSU rods. The seeming decrease in cumulative release after 56 days appears to indicate that caffeine is not stable in the presence of PGSU degradation products. Caffeine instability in PGSU degradation products is supported by the loss of caffeine signal in samples stored on the bench prior to HPLC analysis. An HPLC method was developed that was able to separate and quantify caffeine in the presence of PGSU degradation products in dissolution samples. Large variations in cumulative release of samples, which were not analyzed soon after sampling or thawing, suggests that dissolution samples must be analyzed rapidly after sampling or thawing from its frozen state.