D. Schweitzer

[Comparison of parameters of time-resolved autofluorescence between healthy subjects and patients suffering from early AMD].

BACKGROUNDnA fluorescence lifetime mapper (FLM) was tested for quantitative estimation of early alterations in age-related macular degeneration (AMD) which are assumed to be in cellular metabolism.nnnMETHODnIn FLM time-resolved autofluorescence of the fundus is excited by picosecond (ps) laser impulses at 448 nm and detected in 2 spectral ranges (K1=490-560 nm and K2=560-700 nm) by time-correlated single photon counting. The time-dependent decrease in fluorescence intensity was approximated using 3 decay rates. The calculated lifetimes allow a comparison with endogenous fluorophores of cellular metabolism.nnnRESULTSnInitially mean lifetimes were determined for 8 healthy subjects (K1: tau1=118 ps, tau2=584 ps, tau3=2826 ps, K2: tau1=104 ps, tau2=477 ps, tau3=1623 ps). In 15 AMD patients (AREDS categories I and II) the lifetimes were longer (K1: tau1=166 ps, tau2=986 ps, tau3=3309 ps, K2: tau1=137 ps, tau2=583 ps, tau3=1924 ps). The best separation between healthy subjects and patients with early AMD was possible by parameters 1 and 2 in the short-wave channel. Fluorophore-specific alterations in the macula could be demonstrated in isolated cases with advanced AMD.nnnCONCLUSIONnMeasurements in the 30 degrees fundus field demonstrated that specific alterations were already present even in early AMD and also outside the macula. These act in the neuronal retina, because parameter tau2 is related to this layer. Increases in the lifetime of parameter tau2 in the short wave channel could at least partially be determined by an increase of protein bound NADH, the content of which increases with reduced cellular respiration.ZusammenfassungHintergrundZur quantitativen Beurteilung früher Veränderungen bei altersbedingter Makuladegeneration, die im zellulären Stoffwechsel vermutet werden, wurde ein Fluoreszenz Lifetime Mapper (FLM) erprobt.MethodeMit dem FLM wird die zeitaufgelöste Fundusfluoreszenz mit Pikosekunden-Laserpulsen bei 448xa0nm angeregt. Diese wird in 2 Spektralbereichen (K1=490–560xa0nm, K2=560–700xa0nm) durch zeitkorreliertes Einzelphotonenzählen detektiert. Der zeitabhängige Abfall der Fluoreszenzintensität wird mit 3 Abklingzeiten approximiert. Die ermittelten Abklingzeiten ermöglichen einen Vergleich mit endogenen Fluorophoren des zellulären Stoffwechsels.ErgebnisseFür 8 Gesunde wurden erstmalig die Mittelwerte der Abklingzeiten der Autofluoreszenz bestimmt (K1: τ1=118xa0ps, τ2=584xa0ps, τ3=2826xa0ps, K2: τ1=104xa0ps, τ2=477xa0ps, τ3=1623xa0ps). Bei 15xa0Patienten mit AMD (AREDS Kategorien I und II) wurden längere Abklingzeiten (K1: τ1=166xa0ps, τ2=986xa0ps, τ3=3309xa0ps, K2: τ1=137xa0ps, τ2=583xa0ps, τ3=1924xa0ps) ermittelt. Mit den Parametern 1 und 2 im kurzwelligen Kanal sind Augengesunde und Patienten bereits im Frühstadium der AMD signifikant trennbar. Fluorophorspezifische Veränderungen in der Makula konnten an Einzelbeispielen bei fortgeschrittener AMD demonstriert werden.SchlussfolgerungMessungen im 30-Grad-Fundusfeld zeigen, dass bereits im Frühstadium der AMD spezifische Änderungen auch außerhalb der Makula vorliegen. Diese wirken auch in der neuronalen Retina, der τ2 zugeordnet werden kann. Verlängerungen der Lifetime τ2 im kurzwelligen Kanal können zumindest teilweise auf eine Zunahme von an Protein gebundenem NADH zurückgeführt werden, dessen Anteil bei verringerter Zellatmung steigt.AbstractBackgroundA fluorescence lifetime mapper (FLM) was tested for quantitative estimation of early alterations in age-related macular degeneration (AMD) which are assumed to be in cellular metabolism.MethodIn FLM time-resolved autofluorescence of the fundus is excited by picosecond (ps) laser impulses at 448xa0nm and detected in 2 spectral ranges (K1=490–560xa0nm and K2=560–700xa0nm) by time-correlated single photon counting. The time-dependent decrease in fluorescence intensity was approximated using 3 decay rates. The calculated lifetimes allow a comparison with endogenous fluorophores of cellular metabolism.ResultsInitially mean lifetimes were determined for 8 healthy subjects (K1: τ1=118xa0ps, τ2=584xa0ps, τ3=2826xa0ps, K2: τ1=104xa0ps, τ2=477xa0ps, τ3=1623xa0ps). In 15 AMD patients (AREDS categories I and II) the lifetimes were longer (K1: τ1=166xa0ps, τ2=986xa0ps, τ3=3309xa0ps, K2: τ1=137xa0ps, τ2=583xa0ps, τ3=1924xa0ps). The best separation between healthy subjects and patients with early AMD was possible by parameters 1 and 2 in the short-wave channel. Fluorophore-specific alterations in the macula could be demonstrated in isolated cases with advanced AMD.ConclusionMeasurements in the 30° fundus field demonstrated that specific alterations were already present even in early AMD and also outside the macula. These act in the neuronal retina, because parameter τ2 is related to this layer. Increases in the lifetime of parameter τ2 in the short wave channel could at least partially be determined by an increase of protein bound NADH, the content of which increases with reduced cellular respiration.

Spectral and time-resolved studies on ocular structures

Measurements of endogeous fluorophores open the possibility for evaluation of metabolic state at the eye. For interpretation of 2-dimensional measurements of time-resolved auto fluorescence in 2 separate spectral ranges at the human eye, comparing measurements were performed on porcine eyes. Determining excitation and emission spectra, attention was drawn of proof of coenzymes NADH and FAD in isolated anatomical structures cornea, aqueous humor, lens, vitreous, neuronal retina, retinal pigment epithelium (RPE), choroid, and sclera. All these structures exhibit auto fluorescence, highest in lens. Excitation at 350 nm results in local fluorescence maxima at 460 nm, corresponding to NADH, in all structures. This short-wave excitation allows metabolic studies only at the anterior eye, because of the limited transmission of the ocular media. During excitation at 446 nm the existence of FAD is expressed by local fluorescence maxima at 530 nm. The composition fluorescence spectra allow no discrimination between single ocular structures. Approximating the dynamic fluorescence by a double exponential function, the shortest lifetimes were detected in RPE and neuronal retina. The histograms of mean lifetime tM cover each other on lens with cornea and also on sclera with choroid. Despite the lifetimes are close between RPE and neuronal retina, the relative contributions Q1 are wide different. The gradient of trend lines in cluster diagrams of amplitudes α2 vs. α1 allows a discrimination of ocular structures.

A new imaging technique for retinal vessel oximetry - principles and first clinical results in patients with retinal arterial occlusion and diabetic retinopathy

The oxygen saturation of blood inside retinal vessels is an essential measure for the estimation of oxygen supply to the tissue as well as its oxygen consumption. In the current approach, the blood oxygenation is measured by a dual-wavelength technique. Using a fundus camera, equipped with a special dual wavelength transmission filter and a color CCD camera, two monochromatic fundus images at 548 nm and 610 nm were recorded simultaneously. The optical densities of retinal vessels for both wavelengths and their ratio, which is known to be proportional to the oxygen saturation, were calculated. From a health control population, mean arterial and venous oxygen saturations were measured of 98±10.1% and 65±11.7% with reproducibility of 2.52% and 3.25% respectively. In 10 patients with arterial occlusion, a reduction of the arterial oxygen saturation to 78 ±17% (mean ± standard deviation, branch arterial occlusion) and 91±11% (central arterial occlusion) respectively was found in the occluded vessel. After 5 days on pentoxifilin therapy, the arterial saturation increased to an average of 93±12% or 103 ±6% respectively. In 70 eyes of 42 patients suffering from diabetic retinopathy, an increase of the venous oxygen saturation with the severity of the retinopathy was found (mild nonproliferative retinopathy: 68.4±8.2%, moderate non-proliferative retinopathy: 70.5±6.8%, severe non-proliferative retinopathy: 72.4±7.6%, proliferative retinopathy 75.7±8.3%) due to vessel shunting and diabetic changes of the permeability of vessel walls. These first clinical results demonstrate the ability of an accurate measurement of retinal vessel oxygenation with a very simple setup just requiring a special filter in the illumination path of a fundus camera and dedicated software.

Method for simultaneous detection of functionality and tomography

During clinical application of the fluorescence lifetime laser scanner ophthalmoscope, a stepped slope of ocular autofluorescence was found. Fitting of fluorescence results in wrong lifetimes if the left border of the fitting interval is set at the time channel of first appearance of fluorescence. A better fit was reached at least for the fluorescence decay if the left border of the fitting interval is set near the maximum of detected fluorescence. Analysing the appearance of the stepped slope, its origination was found by the different appearance time of fluorescence from the crystalline lens and from the fundus fluorescence. The extension of the exponential model function by a parameter tci results in an optimal fit of both the slope and of the fluorescence decay. This new parameter describes differences in the appearance time of fluorescence originating from different layers. Taking into account the refractive index between layers, the geometrical distance between them can be determined. In this way, functional information (lifetimes) and geometrical information (distances) can be determined by the same measurement. To reach geometrical resolution comparable with OCT, pulses and time resolution are required in the order of 30 fs.

Zeitaufgelöste Autofluoreszenz bei retinalen Gefäßverschlüssen

ZusammenfassungHintergrundDer zelluläre Metabolismus kann mittels zeitaufgelöster Autofluoreszenz (FLIM) beurteilt werden. Die Überlagerung verschiedener endogener Fluorophore in okulären Geweben erschwert Aussagen zu einzelnen Fluorophoren. Bei Verschlüssen retinaler Astarterien können lokale Änderungen des Stoffwechsels mit gesundem Gewebe verglichen werden.MethodeFLIM von 2xa0Patienten wurde in 2 Spektralkanälen K1xa0(490–560xa0nm), K2xa0(560–700xa0nm) gemessen, triexponentiell approximiert und mit repräsentativen Ergebnisse eines gesunden Auges verglichen.ErgebnisseIn K1 war die Lebensdauer τ1 im unterversorgten Gewebe leicht, τ2 jedoch stark gegenüber dem gesunden Gewebe verlängert. In K2 waren die Verteilungen von τ2 deckungsgleich in beiden Geweben. Am gesunden Auge waren die Verteilungen aller Lebensdauern identisch zwischen entsprechenden Fundusgebieten.SchlussfolgerungDie Verlängerung von τ1 im unterversorgten Gewebe ist auf geringeren Beitrag von proteingebundenem FAD zurückzuführen. Die Verlängerung von τ2 (ca. 500xa0ps) in gesundem Gewebe gegenüber ca. 1, 5xa0ns im unterversorgten ist wahrscheinlich durch proteingebundenes NADH verursacht, das bei der Glykolyse entsteht.AbstractBackgroundCellular metabolism can be evaluated using time-resolved autofluorescence. Because the fluorescence of ocular tissue is an accumulation of the fluorescence of several endogenous fluorophores, it is hard to determine the influence of a single fluorophore. In branch retinal artery occlusion, metabolic changes can be compared with normal tissue.MethodTime-resolved autofluorescence was measured in two patients in two spectral channels, K1 (490–560xa0nm) and K2 (560–700xa0nm), and was 3-exponentially approximated and compared with representative results of a healthy eye.ResultsIn K1, lifetime τ1 in the undersupplied tissue was weak, but τ2 was strongly elongated compared with the healthy tissue. In K2, the distribution of τ2 was identical in both tissues. In the healthy eye, there was an equal distribution of all lifetimes in corresponding fundus regions.ConclusionsThe elongation of τ1 in undersupplied tissue is probably caused by a reduced contribution of protein-bound FAD. The elongation of τ2 (about 500xa0ps) in healthy tissue, compared to about 1.5xa0ns in undersupplied tissue, is probably caused by protein-bound NADH, which is formed in glycolysis.

[Time-resolved autofluorescence in retinal vascular occlusions].

ZusammenfassungHintergrundDer zelluläre Metabolismus kann mittels zeitaufgelöster Autofluoreszenz (FLIM) beurteilt werden. Die Überlagerung verschiedener endogener Fluorophore in okulären Geweben erschwert Aussagen zu einzelnen Fluorophoren. Bei Verschlüssen retinaler Astarterien können lokale Änderungen des Stoffwechsels mit gesundem Gewebe verglichen werden.MethodeFLIM von 2xa0Patienten wurde in 2 Spektralkanälen K1xa0(490–560xa0nm), K2xa0(560–700xa0nm) gemessen, triexponentiell approximiert und mit repräsentativen Ergebnisse eines gesunden Auges verglichen.ErgebnisseIn K1 war die Lebensdauer τ1 im unterversorgten Gewebe leicht, τ2 jedoch stark gegenüber dem gesunden Gewebe verlängert. In K2 waren die Verteilungen von τ2 deckungsgleich in beiden Geweben. Am gesunden Auge waren die Verteilungen aller Lebensdauern identisch zwischen entsprechenden Fundusgebieten.SchlussfolgerungDie Verlängerung von τ1 im unterversorgten Gewebe ist auf geringeren Beitrag von proteingebundenem FAD zurückzuführen. Die Verlängerung von τ2 (ca. 500xa0ps) in gesundem Gewebe gegenüber ca. 1, 5xa0ns im unterversorgten ist wahrscheinlich durch proteingebundenes NADH verursacht, das bei der Glykolyse entsteht.AbstractBackgroundCellular metabolism can be evaluated using time-resolved autofluorescence. Because the fluorescence of ocular tissue is an accumulation of the fluorescence of several endogenous fluorophores, it is hard to determine the influence of a single fluorophore. In branch retinal artery occlusion, metabolic changes can be compared with normal tissue.MethodTime-resolved autofluorescence was measured in two patients in two spectral channels, K1 (490–560xa0nm) and K2 (560–700xa0nm), and was 3-exponentially approximated and compared with representative results of a healthy eye.ResultsIn K1, lifetime τ1 in the undersupplied tissue was weak, but τ2 was strongly elongated compared with the healthy tissue. In K2, the distribution of τ2 was identical in both tissues. In the healthy eye, there was an equal distribution of all lifetimes in corresponding fundus regions.ConclusionsThe elongation of τ1 in undersupplied tissue is probably caused by a reduced contribution of protein-bound FAD. The elongation of τ2 (about 500xa0ps) in healthy tissue, compared to about 1.5xa0ns in undersupplied tissue, is probably caused by protein-bound NADH, which is formed in glycolysis.

Detection of early metabolic alterations in the ocular fundus of diabetic patients by time-resolved autofluorescence of endogenous fluorophores

Measurements of time-resolved autofluorescence (FLIM) at the human ocular fundus of diabetic patients permit the detection of early pathologic alterations before signs of diabetic retinopathy are visible. The measurements were performed by the Jena Fluorescence Lifetime Laser Scanner Ophthalmoscope applying time-correlated single photon counting (TCSPC) in two spectral channels (K1: 490-560 nm, K2:560-700ps). The fluorescence was excited by 70 ps pulses (FWHM) at 448 nm. The decay of fluorescence intensity was triple-exponentially approximated. The frequency of amplitudes, lifetimes, and relative contributions was compared in fields of the same size and position in healthy subjects and in diabetic patients. The most sensitive parameter was the lifetime T2 in the short-wavelength channel, which corresponds to the neuronal retina. The changes in lifetime point to a loss of free NADH and an increased contribution of protein-bound NADH in the pre-stage of diabetic retinopathy.

In-vivo and in-vitro investigations of retinal fluorophores in age - related macular degeneration by fluorescence lifetime imaging

Ocular fundus autofluorescence imaging has been introduced into clinical diagnostics recently for the observation of the age pigment lipofuscin, a precursor of age-related macular degeneration (AMD). However, a deeper understanding of the generation of single compounds contributing to the lipofuscin as well as of the role of other fluorophores such as FAD, glycated proteins, and collagen needs their discrimination by fluorescence lifetime imaging (FLIM). FLIM at the ocular fundus is performed using a scanning laser ophthalmoscope equipped with a picosecond laser source (448nm or 468nm respectively, 100ps, 80 MHz repetition rate) and dual wavelength (490-560nm and 560-7600nm) time-correlated single photon counting. A three-exponential fit of the fluorescence decay revealed associations of decay times to anatomical structures. Disease-related features are identified from alterations in decay times and-amplitudes. The in-vivo investigations in patients were paralleled by experiments in an organ culture of the porcine ocular fundus. Photo-oxidative stress was induced by exposure to blue light (467nm, 0.41 mW/mm2). Subsequent analysis (fluorescence microscopy, HPLC, LC-MS) indicated the accumulation of the pyridinium bis-retinoid A2E and its oxidation products as well as oxidized phospholipids. These compounds contribute to the tissue auto-fluorescence and may play a key role in the pathogenesis of AMD. Thus, FLIM observation at the ocular fundus in vivo enhances our knowledge on the etiology of AMD and may become a diagnostic tool.

Clinical results of fluorescence lifetime imaging in ophthalmology

A laser scanner ophthalmoscope was developed for in vivo fluorescence lifetime measurements at the human retina. Measurements were performed in 30 degree fundus images. The fundus was excited by pulses of 75 ps (FWHM). The dynamic fluorescence was detected in two spectral channels K1(490-560nm), K2(560-700 nm) by time-correlated single photon counting. The decay of fluorescence was three-exponentially. Local and global alterations in lifetimes were found between healthy subjects and patients suffering from age-related macular degeneration, diabetic retinopathy, and vessel occlusion. The lifetimes T1, T2, and T3 in both channels are changed to longer values in AMD and diabetic retinopathy in comparison with healthy subjects. The lifetime T2 in K1 is most sensitive to metabolic alterations in branch arterial vessel occlusion.

Zeitaufgelöste Autofluoreszenz bei retinalen Gefäßverschlüssen@@@Time-resolved autofluorescence in retinal vascular occlusions

ZusammenfassungHintergrundDer zelluläre Metabolismus kann mittels zeitaufgelöster Autofluoreszenz (FLIM) beurteilt werden. Die Überlagerung verschiedener endogener Fluorophore in okulären Geweben erschwert Aussagen zu einzelnen Fluorophoren. Bei Verschlüssen retinaler Astarterien können lokale Änderungen des Stoffwechsels mit gesundem Gewebe verglichen werden.MethodeFLIM von 2xa0Patienten wurde in 2 Spektralkanälen K1xa0(490–560xa0nm), K2xa0(560–700xa0nm) gemessen, triexponentiell approximiert und mit repräsentativen Ergebnisse eines gesunden Auges verglichen.ErgebnisseIn K1 war die Lebensdauer τ1 im unterversorgten Gewebe leicht, τ2 jedoch stark gegenüber dem gesunden Gewebe verlängert. In K2 waren die Verteilungen von τ2 deckungsgleich in beiden Geweben. Am gesunden Auge waren die Verteilungen aller Lebensdauern identisch zwischen entsprechenden Fundusgebieten.SchlussfolgerungDie Verlängerung von τ1 im unterversorgten Gewebe ist auf geringeren Beitrag von proteingebundenem FAD zurückzuführen. Die Verlängerung von τ2 (ca. 500xa0ps) in gesundem Gewebe gegenüber ca. 1, 5xa0ns im unterversorgten ist wahrscheinlich durch proteingebundenes NADH verursacht, das bei der Glykolyse entsteht.AbstractBackgroundCellular metabolism can be evaluated using time-resolved autofluorescence. Because the fluorescence of ocular tissue is an accumulation of the fluorescence of several endogenous fluorophores, it is hard to determine the influence of a single fluorophore. In branch retinal artery occlusion, metabolic changes can be compared with normal tissue.MethodTime-resolved autofluorescence was measured in two patients in two spectral channels, K1 (490–560xa0nm) and K2 (560–700xa0nm), and was 3-exponentially approximated and compared with representative results of a healthy eye.ResultsIn K1, lifetime τ1 in the undersupplied tissue was weak, but τ2 was strongly elongated compared with the healthy tissue. In K2, the distribution of τ2 was identical in both tissues. In the healthy eye, there was an equal distribution of all lifetimes in corresponding fundus regions.ConclusionsThe elongation of τ1 in undersupplied tissue is probably caused by a reduced contribution of protein-bound FAD. The elongation of τ2 (about 500xa0ps) in healthy tissue, compared to about 1.5xa0ns in undersupplied tissue, is probably caused by protein-bound NADH, which is formed in glycolysis.