Characterisation of the cross-linking and calcification associated with glutaraldehyde-treated cardic bioprostheses.

Around 170000 patients worldwide receive cardiac valve substitutes each year. Valve replacement with mechanical or bioprosthetic devices enhances patient survival and quality of life. Bioprosthetic valves have a significant advantage over mechanical valves: they do not necessarily require long-term anticoagulant therapy however, dystrophic calcification can lead to early failure. The actual mechanism of calcification is still poorly understood despite several established possible factors associated with it. Amongst these is the glutaraldehyde pre-treatment of the valves during their manufacture. Glutaraldehyde has been used for the treatment of bioprosthetic valves for the last thirty years, as a cross-linking agent and a sterilant. Whilst it is assumed to introduce stable inter- and intra-fibrillar collagen cross-links, which contribute to the durability of these valves, the specific chemistry of the fixation process is not fully understood. Additionally, glutaraldehyde is thought to be involved somehow in the process of dystrophic calcification of these same bioprosthetic valves. The primary aim of this study was to gain a greater understanding of the chemistry involved in the treatment of collagenous valve tissue with glutaraldehyde. Amino acids, peptides and proteins were thus used to mimic the effect of the glutaraldehyde treatment and to model potential reactions involved in such treatment. Techniques such as MALDI-TOF MS, ESI MS, NMR, FTIR-ATR and Raman spectroscopy were utilised to study the products of the glutaraldehyde reaction and their relationship with the calcification process. Data obtained from the products of the reactions between glutaraldehyde and model compounds showed the presence of: aldol and aldol/Michael condensation products of glutaraldehyde, Schiff base moieties (including cross-links) and various cyclisation products incorporating pyridinium and dihydropyridine ring structures. Some of these structures are in agreement with the literature, whilst others are essentially new structures that have never been proposed. Glutaric acid, used to mimic the oxidation of glutaraldehyde that can occur in-vivo, was shown to have the ability to form complexes with cations such as calcium in-vitro. A similar result was found with aqueous dilute solutions of glutaraldehyde (similar concentrations to the ones used in valve manufacture), thus leading to the hypothesis of its strong role in the initiation of calcification in-vivo. However, an extrapolation of these results to the role of the nucleophilic groups of amino acids or peptides, that could behave as the collagen macromolecule, was difficult to assess using FTIR because of the complex infrared spectra. However some findings corroborated the hypothesis that amino acids of the collagen tissue may also play a role in the initiation of calcification. Secondly, methodology was developed to allow successful analysis of tissue calcification using environmental scanning electron microscopy (ESEM). This is thought to be an important step in the analysis of tissues in their native state. Investigation of the calcification process with samples from clinical investigations (explanted human calcified valves), in-vivo screening (rat subcutaneous implantation model) and in-vitro screening (pericardial tissue incubated in metastable calcification solution) was thus undertaken using ESEM, along with other techniques such as FTIR-ATR, Raman, XRD spectroscopy and ICP-OES. The data revealed both similarities and differences between in-vivo and in-vitro calcification, although the process is unequivocally different. Late calcific deposits were assigned to poorly crystalline hydroxyapatite with high Ca/P ratios due to the probable presence of carbonate and possibly cations such as silicon and magnesium. A picture of the onset of mineralisation was hypothesised involving precursors, containing various amounts of calcium and phosphate, along with the incorporation of magnesium and silicon. These precursors phases evolved with time of implantation to the poorly crystalline form of hydroxyapatite found in the late stage of calcification. This work has provided an insight into how glutaraldehyde reacts with valve tissue and a possible explanation as to why valves fail by non-calcific or calcific mechanisms. A new approach to the study of calcified valve tissue has also been developed using ESEM methodology