METHOD AND APPARATUS FOR USING LASER-INDUCED FLUORESCENCE DURING PHOTORETRACTIVE KERATECTOMY.
TECHNICAL FIELD
The present invention relates generally to corneal surgery, and more particularly to a method of determining when a predetermined portion of the cornea has been removed during photorefractive or phototherapeutic keratectomy.
BACKGROUND ART
The 193-nm excimer laser is a device whose efficacy in treating myopia, astigmatism and anterior stromal scars has been demonstrated. See for example PJ McDonnell, H Moreira, TN Clapham, J D'Arcy, & CR Munnerlyn, Photorefractive keratectomy for astigmatism: initial clinical results. Archives of Ophthalmology 1991 ; volume 109, pages 1370-73. Over the past ten years, several aspects of excimer photorefractive technology have been examined, including refinement of refractive efficacy and reproducibility, biochemistry of wound healing, microstructural changes, and possible side effects or complications of the therapy. One concern that was raised early in the development of photorefractive keratectomy was direct exposure of corneal tissue to argon-fluoride excimer 193nm laser irradiation. Work by Setlow and others in the early 1970's had demonstrated the mutagenic and carcinogenic potential of ultraviolet light. While histological collateral damage to adjacent tissue during excimer laser ablation was known to be minimal, concerns about radiation damage to DNA of adjacent cells remained nonetheless. These concerns were allayed by several observations. First, the penetration of 193nm ultraviolet light to the nucleus is greatly diminished by the shielding effect of the surrounding cytoplasm, and it is likely that only a minute fraction ever
penetrates to the center of mammalian cells. Second, 193nm light is known to produce single-strand breaks in DNA, which are thought to be less mutagenic than the pyrimidine dimers detected after 250nm to 280nm ultraviolet light exposure. Finally, no unscheduled DNA synthesis is detected in human skin adjacent to ablative incisions made with a 193nm (ArF) excimer laser, suggesting that DNA repair mechanisms are not activated by 193nm (ArF) ablative incisions as they are with 248nm (KrF) ablative incisions.
Several research groups have also examined the potential damage to surrounding ocular structures from lower energy ultraviolet light produced by fluorescence of the ablated tissue. Loree et al examined this laser-induced fluorescence produced by ablation of porcine corneas at 193nm (ArF), 248nm (KrF) and 308nm (XeCI) wavelengths using an intensified diode array detector system. Muller-Stolzenburg et al examined laser-induced fluorescence produced during 193nm excimer laser ablation of bovine cornea, by placing a quartz fiber into the anterior chamber of the treated eye. They measured laser-induced fluorescence at several fluences but made no distinction between laser-induced fluorescence produced during epithelial or stromal ablation. Two other research groups (Tuft et al and Ediger) have made more complete analyses of laser-induced fluorescence. Both measured laser-induced fluorescence in human cornea using photodiode systems, and while their spectra showed many qualitative differences, each group calculated a similar quantum yield and a total ultraviolet energy dispersion that suggested low cataract risk from ultraviolet laser-induced fluorescence. Only Tuft's group considered any possible differences between laser-induced fluorescence produced during epithelial and that produced during stromal ablation. Technological limitations of the monochromatic photomultiplier detector used by Tuft et al required several minutes to scan the ultraviolet region in a step-wise fashion to obtain complete laser-induced fluorescence spectra, and thus necessitated relatively long, multi-shot irradiation of the tissue as well as sub-ablative fluences of 2.5 and 8 mJ/cm2 to permit enough time for
measurement of epithelial laser-induced fluorescence. Thus in order to obtain laser-induced fluorescence spectra from epithelium, fluences were kept sub- ablative at less than 10 mJ/cm2.
At present, in preparation for the photorefractive keratectomy of the stroma, the overlying epithelium is removed either by mechanical scraping with a knife or by using conventional phototherapeutic techniques such as focusing a large aperture excimer laser beam on (or scanning a small aperture excimer laser beam in a predetermined pattern across) the region of interest for a fixed period of time in order to ablate a constant "pre-set" depth into the cornea. Problems exist with both methods. Mechanical removal, for instance, can be difficult with anxious patients and may leave residual epithelial tissue which will alter the effect of subsequent refractive ablation. Using time (or number of pulses) to determine the depth is also problematic because the epithelium is known to be of variable thickness in normal eyes and is thicker and more variable in eyes that have already undergone photorefractive keratectomy. Other factors relevant to the pre-surgical eye may affect epithelial thickness as well, such as changes in epithelial hydration due to exposure and topical medications. Imprecision in removal of corneal epithelium may lead to greater variability in refractive outcome through a number of mechanisms. Wound healing, regression and epithelial thickness may be altered, as may the predicted optical correction. The theoretical implications of depth and width in estimating the power of refractive keratectomies has been analyzed previously by Munnerlyn et al. in Photorefractive keratectomy: a technique for laser refractive surgery. Journal for Cataract Refractive Surgery 1988; volume 14, pages 46-52.
Several observers have noted and reported a visible bluish light produced during excimer laser keratectomy that appears to lessen in intensity after ablation of the epithelium, and have speculated that if the visible fluorescence
is visually monitored in a darkened room, the surgeon may be able to more accurately control the ablation of the epithelial layer.
DISCLOSURE OF INVENTION The present invention provides apparatus and method for measuring laser-induced fluorescence in "real-time" during actual excimer laser ablation, to more accurately control the photorefractive ablation of the corneal stroma. Laser-induced fluorescence of at least one wavelength of the ultraviolet region is monitored in real-time during ablation of the different layers of corneal tissue. In accordance with one aspect of the invention, a marked change in the monitored fluorescence indicates the moment when substantially all the epithelium has been ablated and Bowman's layer and the adjoining stroma is about to be penetrated, whereupon the excimer laser is switched to a photorefractive mode in which the stroma is selectively ablated (for example, by changing the diaphragm opening or the scanning pattern during the course of the stromal ablation) to produce a reshaped stroma having a predetermined surface profile corresponding to the desired refractive correction.
In accordance with another aspect of the invention, the monitored fluorescence is integrated over time to provide a quantitative measurement of the ablated tissue.
BRIEF DESCRIPTION OF DRAWINGS
Fig 1 is a schematic of an experimental apparatus for ablating different layers of the cornea with an excimer laser, and for monitoring the laser-induced fluorescence spectra during the excimer laser ablation. Fig 2 is a spectrum of the Laser-induced fluorescence obtained during ablation of corneal epithelium at a fluence of 150 mJ/cm2. Fig 3 is a spectrum similar to that of Fig 2, but obtained during ablation of corneal stroma at a fluence of 150 mJ/cm2.
Fig 4 is a three dimensional representation of multiple consecutive laser-induced fluorescence spectra obtained during an ablation 100 //m into corneal stroma, beginning with the initial ablation of a superficial region of the epithelium on the extreme right of the graph and ending with the deepest extent of the stromal ablation on the far left.
Fig 5 is a photomicrograph of a human corneal button subjected to two excimer laser ablations, one (on the right) being carried well into the corneal stroma, and the other (on the left) being halted in response to a marked decrease in the monitored ultraviolet laser-induced fluorescence.
Fig 6 is a higher magnification view of ablation shown at left in Fig 5.
Fig 7 shows a practical embodiment of the apparatus aspects of the invention in which the radiation measurement is used to directly control the excimer laser during a photorefractive procedure. Fig 8 is a flow chart showing an exemplary implementation of the method aspects of the invention.
PREFERRED MODE FOR PRACTICING THE INVENTION
In a series of laboratory experiments using the apparatus of Fig 1, laser ablations were accomplished using a commercially available, 193nm argon-fluoride laser L (Twenty-Twenty excimer laser, VISX Inc, Sunnyvale, California). For all trials, fluence was between 150 and 165 mJ/cm2 and repetition rate was 6 Hz. Freshly enucleated human globes were obtained and used within six hours of procurement, from which nine millimeter corneal buttons C were obtained from the globes, taking care during trephination not to disturb the epithelium. These buttons were then placed on a UV permissive plastic tissue stage T and ablated.
A 0.5 mm fiberoptic detector D was placed directly below the tissue stage T. To measure laser-induced fluorescence LIF, we used an ISA, Model HR 320 spectrographic grating G connected to a cooled, intensified charge-coupled
device CCD (578 x 384 Thomson CCD chip) from Princeton Instruments (Princeton, NJ) having a built-in intensifier I. Spectra were displayed and recorded with Princeton Instruments CSMA software using a desktop IBM-compatible computer PC.
The ablated corneas CB were fixed in formalin and embedded in paraffin. Histological sections were cut to a thickness of 4 to 5 μvn and stained with hematoxylin and eosin.
The spectra shown in Figs 2, 3 and 4 are the result of averaging fluorescence over a 2-second period of 12 laser pulses, during which the corneal tissue was being ablated with a 193nm excimer laser beam having a nominal fluence (peak instantaneous energy) of 150mJ/cm2. Shorter integration periods, including full laser-induced fluorescence spectra from single 20-nanosecond excimer laser pulses, yielded qualitatively identical laser-induced fluorescence spectra, with some increase in signal-to-noise ratio. As noted in Fig 2, two main peaks in laser-induced fluorescence are noted initially during 193nm excimer laser ablation of epithelium from the corneal buttons CB. The higher frequency peak at about 346nm is noticeably more intense than the secondary peak at about 405nm, which stretches into the visible region (hence the bluish visible light previously noted by others during ablation). Identical spectra were obtained using frozen eye bank globes.
Deeper ablation into the same corneal button CB yielded the stromal laser-induced fluorescence spectra shown in Fig 3, which was also integrated over two seconds.
Laser-induced fluorescence spectra from consecutive 2-second integration periods during the entire ablation are shown in Fig 4, beginning with the initial ablation of a superficial region of the epithelium on the extreme right of the
graph (Integration period 1) and ending with the deepest extent of the stromal ablation on the far left (integration period 30).
Note that as shown in Fig 4, the peaks P1, P2 do change in frequency, with the greatest change occurring in the vicinity of the 346nm peak P1 and corresponding to a threefold decrease in the measured intensity. This pronounced shift in intensity commenced approximately 20 seconds into the ablation (ie, integration period 10), and occurred abruptly during a transition period spanning approximately 2 integration periods (ie, integration periods 10- 12). Since the intensified charge-coupled device displayed a full spectrum from 290nm to 560nm which was recalculated at the end of each integration period (ie, once every two seconds), it was possible to monitor the displayed spectra visually and manually halt the ablation shortly after the beginning of this transitional phase (ie, at integration period 12 of Fig 4), which we believe corresponds to the relatively thin basement membrane and anterior aspect of Bowman's layer which separates the epithelium from the stroma.
The cornea shown in Fig 5 was subjected to two different excimer laser ablations. The first ablation A1 , shown on the right side of the photomicrograph, was continued 100 μm into stroma S, allowing demonstration of the change in fluorescence as epithelium E, Bowman's layer B and anterior stroma S are transversed. On the opposite side of the cornea, a separate ablation A2 was intentionally halted when the spectrum of ultraviolet laser-induced fluorescence was noted to change. Fig 6 offers a higher magnification view of this laser-induced ultraviolet fluorescence-guided ablation, from which it will be noted that substantially all the epithelium E has been removed from the area of interest, but that Bowman's layer B remains substantially intact.
Since our intensified charge-coupled device detector system shown in Fig 1 simultaneously measured laser-induced fluorescence at hundreds of individual wavelengths six times a second for the entire near-ultraviolet region (300 to 500nm), we were able to monitor in real-time sequential changes in corneal fluorescence during simulation of a therapeutic excimer laser keratectomy at 150 mJ/cm2. Further, real-time measurement of laser-induced fluorescence enabled us to monitor the progression of ablation through the cornea under conditions typical of excimer laser keratectomy (at least 6 pulses per second of a 193nm excimer laser with a fluence of at least 50mJ/cm2 per pulse) and to halt (or otherwise modify) the ablation process when the epithelial-stromal interface had been reached.
In fresh human tissue, two peaks were noted at approximately 405nm and 346nm. In a practical embodiment, it should be sufficient to monitor only one of these peaks (preferably the stronger 345nm peak), rather than the entire ultraviolet spectrum, since the observed change in fluorescence is particularly noticeable at each peak. By observing the approximate threefold reduction in ultraviolet laser-induced fluorescence upon completion of epithelial ablation, the photorefractive surgeon is then able to gauge when to terminate the constant density profile ("phototherapeutic mode") used during the temporary removal of the epithelium and when to commence the variable density profile ("photorefractive mode") used for the permanent reshaping of the stroma. Those skilled in the art will recognize that the resultant more precise removal of the epithelial layer in the region of interest, regardless of individual variations in epithelial thickness (or other possible variations in physical characteristics such as density and hydration), permits a more controlled and reliable change in refraction during the photorefractive phase of the surgery. Moreover, experiments on animals indicate that removal of the epithelium with the excimer laser, as opposed to mechanical scraping, results in a substantial reduction in the time for the epithelial layer to be restored, presumably because the surface is smoother after laser removal, and because the area being removed is
confined to the region of interest (typically a 6mm circle defining the outer boundary of the photorefractive radiation). In particular, we observed in three clinically acceptable animal model subjects (lagomorphs) healing of comeal epithelium in two days after laser removal, compared to three days after mechanical removal in three other subjects.
The system shown schematically in Fig 7 is a practical embodiment of the experimental setup discussed in detail above, but adapted for photorefractive surgery on human subjects. In particular, the excimer laser L is mounted above a patient support T', and the measured fluorescence F' emanates from the outer surface of the cornea C, and is collected by a optical collimator O and passes through a bandpass filter B before its intensity is measured by a suitable photoreceptive sensor P. In order to improve the signal to noise ratio, the sensor P is preferably responsive to a range of ultraviolet frequencies, for example from 300nm to 500nm; however bandpass filter B preferably excludes any visible background radiation (such as light used by the surgeon to align the apparatus with the patient's cornea or to otherwise monitor the patient during the procedure), as well as any background radiation in the far ultraviolet such as direct or reflected 193nm radiation from the laser L. The optical collimator O merely increases the amount of the emergent fluorescence that impinges of the photo receptor P. thereby further improving the system's sensitivity. Measuring background radiation between pulses, using multiple sensors, combining multiple measurements, and other known physical, optical and computational expedients for increasing the signal to noise ratio and thereby improving the speed and accuracy of the monitoring process will doubtless be apparent to those skilled in the art.
In another practical embodiment of the method aspects of the invention, as shown in Fig 8, rather than visually monitoring the intensity of the radiation (either the entire spectrum, or an isolated region containing one or more peaks) after processing of the sensor data in a computer and displaying the computed
spectrum in real-time on the computer monitor PC, the computer may be programmed to cause the excimer laser L to automatically commence the photorefractive phase upon the detection of a predetermined quantitative decrease in the measured ultraviolet fluorescence. More particularly, as new data are received (block 100) and any necessary initialization has been performed (blocks 102, 104), the new data are stored (block 106). Assuming that the device has not already been switched to the photorefractive mode (block 108), the current measurement is compared with the initial measurement (block 110) and if the ratio is greater than a predetermined experimentally determined threshold (for example 3) (block 112), the machine is switched to a photorefractive processing mode (blocks 114 through 124). In the photorefractive processing mode, the size and shape of the effective aperture of the excimer laser beam can be controlled in accordance with a stored time- dependent profile, as is done conventionally following mechanical removal of the epithelium. However, rather than merely determining the moment of time when the decrease in intensity occurs and thereupon commencing a programmed photorefractive mode in which the duration, intensity and location of the radiation is controlled in accordance with a predetermined profile corresponding to the desired refractive correction, the apparatus may continues to monitor the measured intensity of the fluorescence during the photorefractive ablation of the stroma and integrate (blocks 116, 120) a quantitative measurement of the intensity to provide a quantitative measurement of the amount of actual tissue removed to thereby determine (block 122) if the procedure should be terminated (block 124) even before the predicted time has elapsed (blocks 114, 126). The inventors believe that such a quantitative measurement will be able to compensate for changes in corneal hydration between one subject and the next, and thus provide a more accurate prediction of the ultimate refractive correction that will be obtained.
The foregoing discussion assumes a wide beam radiation pattern which remains at full aperture (approximately 6 mm) during the photoablation of the
epithelium, and a variable radiation pattern obtained with a programmed time-varying aperture during the photorefractive re-shaping of the stroma. However, the principles of the invention are equally applicable to a scanning beam system in which the control system of the excimer laser determines not only the size and shape of the beam, but also a specific path traversed by the beam across the region of interest, how many times each point in the region of interest is radiated, and for how long. In that case, by making separate fluorescence measurements in real-time at different locations within the region of interest and by using the results of those real-time measurements to adaptively control the scanning path and/or the timing of the individual pulses ablation at each location along the path, it should be possible to compensate for thickness variations in the same epithelium (such as can result from prior injury or disease) and to more accurately control the amount of tissue actually ablated from different locations of the stroma, thereby better compensating for variations in the hydration of the stroma and other physical properties of the cornea that may occur from one patient to the next, or as the result of medication, scrubbing and irrigation of the exposed corneal surfaces in preparation for or during the keratectomy procedure.
Those skilled in the art will also recognize that the principles of the present invention may also be applied to photorefractive keratectomy using a relatively narrow scanning beam with a fixed aperture rather than a relatively wide fixed beam with a variable aperture, in which case it should be possible to obtain an accurate removal of the epithelial layer from the entire region of interest, even if the thickness varies from one location to the next. The invention should also find application to other photorefractive procedures using other types of lasers operating at other wave lengths, for example a solid state laser operating at about 213nm. The quantification aspects of the invention may also find application in other photo refractive procedures in which stroma is ablated without ablation of the epithelium.