Comparison of techniques for quantification of fluorescence from tissue

This paper summarizes ten approaches to quantifying fluorescence in tissues, and contrasts their strengths and weaknesses, relative to what their common applications are, and should be. The major issues involved in this analysis are to compare the accuracy of the method and the ability to quantify the active (i.e. non-aggregated) fraction of fluorophore in the tissue. In addition, issues of the depth of penetration and the availability of the method come into play when clinical applications are required. In general, tissue extraction and liquification methods are the 'gold standard' in this field, yet these are plagued by large variance in the values, raising questions about their ability to report on the true active fraction of drug in the tissue. Confocal and fiber optic microsampling methods allow direct quantification of the active fluorescence in vivo and are able to quantify the heterogeneity in the tissue. Yet both of these methods sample the most superficial layers of a tissue, unless invasive injection of the probe is done. Macroscopic sampling of the tissue is therefore the preferred choice for clinical use, yet there is truly no optimum method which can sample the drug concentration to arbitrary accuracy. Empirical bulk tissue sampling methods are the most commonly used, yet without model-based interpretation of the values it is generally not possible to be quantitative. Even relative uptake values can be distorted by the shape of the tissue, and so raster scanning or model-based assessment of the fluorescent yield is preferable, if available. Extending this concept further, tomographic methods can be implemented to quantify fluorescence, and can even be coupled into existing clinical imaging systems, but development and optimization of these methods will be required in the coming years. These are outlined, and case examples illustrated in this paper.