On the development of analytical methods for percutaneous transluminal coronary angioplasty catheters and drug eluting stents

Abstract

This thesis reports on analytical methods required for the development and production of interventional medical devices. A novel light obscuration method is applied to the measurement of particulate matter during an in vitro clinical procedure in the sub-visible range of 2 to 125 µm. The methodology involves a break from the standard approach of sampling a portion of the rinse solution from a medical device using a flow through syringe sampler, and introduces a closed-loop apparatus. The in-line method developed enables a more rigorous means of measuring particles coming from the devices. It has been shown that everolimus, an anti-proliferative drug used by drug eluting stents, forms 3 base-induced degradation products when prepared in acetonitrile. The mechanism involved in the degradation pathway has been identified by design of experiment. Near infra-red transmission spectroscopy was used to determine the in-process concentration of components in different drug eluting stent coating formulations. The results show that near infra-red spectrometry (NIR) is an excellent alternative to traditional wet chemistry analytical techniques for rapid and convenient analysis of these types of samples. Three different sources of device related, intrinsic particles from medical devices have been discovered using the novel closed-loop method. Light obscuration was successful in providing a fast and accurate particle count and particle size. Further analysis of the particles was carried with the complementary methods of light microscopy, scanning electron microscopy and Raman micro-spectroscopy. Firstly, drug eluting stent coatings were examined. It was discovered that a phosphorylcholine-based stent coating resulted in significantly more particles than coatings based on polymers such as polyvinylidene fluoride-co-hexafluoropropene, and poly-lactic-co-glycolic acid. Particle formation from poly-lactic-co-glycolic acid coatings, when exposed to conditions that promote hydrolysis of the polymer, has also been shown. Secondly the impact of accelerated ageing on Pebax catheters results in the formation of Nylon-12 derived particles. This phenomenon has significant ramifications for medical device manufacturers that wish to speed up device approval through the use of accelerated ageing. Finally, catheter coatings have been shown to contribute significantly to particle counts. Extensive particulate matter evaluation has been carried out on two types of hydrophilic coatings, a cross-linked neopentyl glycol diacrylate with embedded polyethylene oxide, and a cross-linked neopentyl glycol diacrylate with embedded polyvinyl pyrolidinone. These three examples of intrinsic particle sources have all led to changes in the design of medical devices to improve the particle counts. The structured experimental approach taken for the investigation into everolimus degradation has shown the power of well-constructed experimental designs. The use of statistical methods proved very successful to help with the identification of the main factors, and with establishing a key interaction involved. This interaction is between alkaline impurities from untreated soda-lime glassware and propionitrile, an impurity related to the manufacture of acetonitrile. Some everolimus degradation prevention measures have been evaluated and shown to be effective, which is hoped will benefit laboratories using this drug and its analogues. Mass spectrometry was used to identify the modified site on the everolimus molecule and structures have been proposed. Near infra-red transmission spectroscopy was used in combination with partial least squares regression to measure the components in everolimus- and paclitaxel-eluting stent coating formulations. Everolimus coating solutions contained 0.41% w/w everolimus, 2.00% w/w polyvinylidene fluoride-co-hexafluoropropene, 68.3% w/w acetone, and 29.3% w/w cyclohexanone; and the paclitaxel coating solutions contained 1.93% w/w of paclitaxel and poly-lactic acid, and 96.14% w/w n-butyl acetate. The assay of the paclitaxel coating solution demonstrated that the NIR method of measurement for the drug component at 1.93% w/w matches the accuracy of the established high performance liquid chromatography assay (root mean square error of cross-validation = 0.0318% w/w and root mean square error of prediction = 0.0313 % w/w). Near infra-red technology is not generally used as a trace level technique to provide accurate quantitative analysis < 0.5% w/w, and indeed the NIR method developed for everolimus coating solution demonstrated that the measurement of everolimus at 0.41% w/w does not reach the same level of accuracy as conventional high performance liquid chromatography assays (root mean square errors of cross-validation and prediction = 0.0135 % w/w). Measurement of the polymer, solvent and water components of the coating solutions equalled the performance of the alternative wet chemistry methods.