Abstract
Purpose.:
To correlate in vivo spatial and spectral morphologic changes of short- to long-pulse 532 nm Nd:YAG retinal laser lesions using Fourier-domain optical coherence tomography (FD OCT), autofluorescence (AF), fluorescein angiography (FA), and multispectral imaging.
Methods.:
Ten eyes with treatment-naive preproliferative or proliferative diabetic retinopathy were studied. A titration grid of laser burns at 20, 100, and 200 milliseconds was applied to the nasal retina and laser fluence titrated to produce four grades of laser lesion visibility: subvisible (SV), barely visible (BV, light-gray), threshold (TH, gray-white), and suprathreshold (ST, white). The AF, FA, FD-OCT, and multispectral imaging were performed 1 week before laser, and 1 hour, 4 weeks, and 3 and 6 months post-laser. Multispectral imaging measured relative tissue oxygen concentration.
Results.:
Laser burn visibility and lesion size increased in a linear relationship according to fixed fluence levels. At fixed pulse durations, there was a semilogarithmic increase in lesion size over 6 months. At 20 milliseconds, all grades of laser lesion were reduced significantly in size after 6 months: SV, 51%; BV, 54%; TH, 49%; and ST, 50% (P < 0.001), with retinal pigment epithelial proliferation and photoreceptor infilling. At 20 milliseconds, there was healing of photoreceptor inner segment/outer segment junction layers compared with 100- and 200-millisecond lesions. Significant increases in mean tissue oxygenation (range, four to six units) within the laser titration area and in oxygen concentration across the laser lesions (P < 0.01) were detected at 6 months.
Conclusions.:
For patients undergoing therapeutic laser, there may be improved tissue oxygenation, higher predictability of burn morphology, and more spatial localization of healing responses of burns at 20 milliseconds compared with longer pulse durations over time.
Laser photocoagulation in diabetic retinopathy has been validated by the Early Treatment Diabetic Retinopathy Study (ETDRS) and the Diabetic Retinopathy Study.
1,2 Conventional argon lasers used in clinical practice have included spot sizes 50–500 μm with a range of pulse durations from 20 to 500 milliseconds.
2 –4
The rationale behind retinal photocoagulation was based on numerous observations from animal laser studies, and histopathological features of argon laser lesions have been correlated with time-domain optical coherence tomography (OCT).
5,6 In the last decade, laser burn thresholds have been reevaluated for argon therapy.
4,7 To achieve a visible burn, the laser energy pulse initiated at the retinal pigment epithelium (RPE) level produces a thermal temperature rise within the neuosensory retina and change in natural retinal transparency, and hence the illumination light beam will scatter to produce a visible burn.
8
The pattern scanning laser (Pascal, Optimedica, Santa Clara, CA) photocoagulator was introduced in 2006 for retinal photocoagulation.
9 It semi-automates the procedure using a short-pulse duration (20–30 milliseconds) combined with rapid raster scanning of multiple spots. There may be less outer retina and RPE damage because of less collateral thermal diffusion.
10
The extent of tissue damage that is required to generate a therapeutic outcome remains unknown. Marshall and coworkers
11 reported histopathological correlations for suprathreshold laser in
Macaca mulatta and the human retina that explain the thermal effects of laser at a cellular level. However, conventional laser parameters have been associated with burn expansion and tissue loss over time.
12
Newer laser technology has been developed to target the RPE and minimize photoreceptor injury.
13 –15 Healing responses of short-pulse photocoagulation burns have been demonstrated in animal studies and have recently been shown in vivo for the human retina at different pulse durations and power levels.
16,17 To identify a minimally traumatic laser treatment for therapeutic application, it is necessary to know whether laser lesions below the threshold visibility level reduce intraretinal tissue damage over time.
In our study, we used the pattern scanning laser (Pascal) system (532 nm) to examine laser titration burns applied to diabetic retinas. We used Fourier-domain OCT (FD-OCT) and fundus autofluorescence (AF) to noninvasively evaluate laser-tissue interactions at subclinical and low fluence levels. Multispectral imaging and fundus fluorescein angiography (FFA) tests were performed to evaluate alterations in tissue oxygenation and perfusion post-laser. The main objective of this study is to establish in vivo laser-tissue interactions for subvisible/subthreshold, barely visible, threshold, and suprathreshold photocoagulation burns by using spatial and spectral imaging techniques.
The study protocol and informed consent forms were approved by the Research Ethics Committee. Data and safety monitoring were provided by an independent panel at the University of Manchester and Manchester Royal Eye Hospital. Patients referred by the South Manchester Diabetic Screening Service with PDR or pre-proliferative diabetic retinopathy (PPDR) were prospectively recruited. Informed consent was obtained from all patients before study entry. Ten eyes of seven consecutive patients were studied between September 1, 2009, and April 6, 2010. The study adhered to the tenets of the Declaration of Helsinki, and the study was registered at the Central Manchester University Hospitals NHS Foundation Trust.
All patients underwent slitlamp biomicroscopy and ultra-wide-field FFA (Optos, Dunfermline, Scotland) at baseline. Two masked retina specialists graded baseline angiography images for entry into the study, and additional inclusion criteria included patients older than 18 years of age, glycosylated hemoglobin (HBA1C) level ≤ 10%, no previous laser, intraocular drug therapy, or surgery to the study eye, blood pressure < 180/110, absence of any systemic medication known to be toxic to the retina, and no history of chronic renal failure or renal transplant for diabetic nephropathy.
Safety endpoints included all adverse events reported spontaneously by study participants, elicited by investigators, and observed by investigators. Adverse events were graded as mild, moderate, or severe and were assessed as being either related or unrelated to the laser treatment. As part of the ethical and good clinical practice, we recorded all serious adverse events whether or not they were deemed to be related to the treatment.
A PRP lens (Mainster 165; Ocular Instruments Inc., Bellevue, WA) was used, with spot magnification factor 1.96 and theoretical retinal spot size 392 μm. An area of nasal retina, one disc diameter from the disc margin, was selected for grid placement with burn distribution greater than one-and-half burn widths. The clinical appearance of all laser lesions was graded by a single observer within 60 seconds of delivering the laser pulse by means of the following visual scale: no blanching (subvisible), light gray (barely visible), gray-white (threshold), and white with halo of edema (suprathreshold).
Starting at 100 mW, power was increased in 25 mW steps until a threshold burn was produced. A further 25 mW increment produced suprathreshold lesions. The threshold lesion power setting was then reduced by 25 mW (equivalent to fluence level of 3 J/cm
2) to produce a barely visible lesion, and a further 25mW reduction produced a subvisible lesion. The titration grid comprised three rows of graded intensity burns at 20, 100, and 200 milliseconds (
Fig. 1).
One Hour.
Four Weeks.
Three Months.
Six Months.
At 6 months post-laser, color photography of the laser lesions showed two zones of intraretinal damage with central pigmentation and a surrounding ring of atrophy (
Fig. 5A). Twenty millisecond lesions of subvisible, barely visible, and threshold intensity were barely visible on biomicroscopy. The 100- and 200-millisecond subvisible lesions were partially visualized on biomicroscopy. At 20 milliseconds, there was minimal central pigmentation and atrophy. However, all 100- and 200-millisecond lesions showed progressively larger areas of central pigmentation and atrophy (
Fig. 5B). Using green-free SLO imaging, the extent and accumulation of central pigment could be accurately determined (
Fig. 5B). The area of pigmentation appeared to increase from subvisible up to suprathreshold levels at all pulse durations. The amount of pigmentary changes was minimal (“hypotrophic”) for 20-millisecond lesions, and maximal (“hypertrophic”) for 200-millisecond lesions.
Over 6 months, the 100- and 200-millisecond lesions developed increasing size of the ring lacking AF that corresponded to photoreceptor/RPE atrophy (
Figs. 3C,
3D). At 20 milliseconds (subvisible, barely visible, threshold), there was a smaller ring lacking AF with minimal central hyper-AF at 6 months compared with the higher pulse duration lesions.
At 100 to 200 milliseconds, there was intraretinal contracture for all lesions and window defects of hyper-reflectivity at the edges of burns (
Fig. 6). The window defects increased in signal from barely visible to suprathreshold intensity. These window defects corresponded to tissue atrophy and lack of AF seen at the edges of laser lesions. The central parts of the lesions at 100 to 200 milliseconds showed fusion of the OS and the apical RPE (
Figs. 6A,
6B). At 6 months, there was diffuse RPE aggregation and photoreceptor atrophy at the locations of 100- to 200-millisecond laser lesions. There was a proportional increase in fluence for 100- to 200-millisecond lesions compared with 20 milliseconds, and this produced greater collateral tissue damage with intraretinal thermal disruption and rings of outer retinal atrophy around these laser lesions.
We observed the subvisible, barely visible, and threshold lesions at 20 milliseconds to remain spatially confined, with no axial or lateral spread to adjacent RPE cells, photoreceptors, or choroid at the 6-month examination (
Fig. 6C). The 20-millisecond suprathreshold lesion showed a small central area of hyper-reflectivity localized to the IS-OS and apical RPE that represented RPE aggregation and photoreceptor in-filling. At all grades of 20-millisecond lesions, the outer retinal layers could be visualized. Low fluence levels for 20-millisecond lesions resulted in minimal thermal damage, with restoration and healing of layers. The 20-millisecond subvisible lesions showed complete restoration of the outer highly reflective layers. The RPE layer showed slight thickening at the central nidus of each laser lesion (barely visible, threshold, suprathreshold), with no disruption of the photoreceptor outer segments.
Fluorescein angiography was performed at 6 months post-laser. The laser lesions showed two distinct fluorescence patterns at 6 months compared with baseline (
Fig. 5C). A central zone of hypofluorescence was present, with a surrounding zone of hyperfluorescence. At 20 milliseconds, all grades of laser lesions showed mild increased fluorescence at the center with a thin ring of hyperfluorescence. At 100 and 200 milliseconds, there was a progressive increase in the size of the central hypofluorescence and increase in the size of the surrounding hyperfluorescent ring.
The central hypofluorescence correlated with areas of increased pigmentation within outer retina that blocked the underlying fluorescence. The ring of hyperfluorescence corresponded to the atrophic rings on FD-OCT and color imaging and represented increased visualization of underlying fluorescence through atrophic window defects. The ring of hyperfluorescence increased in size from subvisible up to suprathreshold lesions for 100 to 200 milliseconds.
In this study, retinal ischemia was designated by areas of capillary dropout, vascular leakage and vessel wall staining, multiple intraretinal hemorrhages, number of microaneurysms, and presence of cotton wool spots. The fluorescein angiography at 6 months compared with baseline within the laser-treated sector, demonstrated an overall reduction in extent of vascular leakage and vessel wall staining in seven of eight eyes (two eyes showed no vascular leakage at baseline), reduced numbers of intraretinal hemorrhages and microaneurysms in 9 of 10 eyes, and improved perfusion within areas of capillary dropout in five of six eyes (four eyes showed no capillary dropout in this area at baseline). The improvements in retinal ischemia reflect the increased tissue oxygenation demonstrated within the laser titration area. There were no visible changes of retinal vessel caliber or intravascular fluorescence at 6 months compared with baseline in the nasal quadrant. The expected arteriolar constriction post-laser was not demonstrated in this study, and it would be useful to image different points in the retina and study the change in oxygenation and retinal ischemia after therapeutic retinal laser photocoagulation in a randomized clinical trial.
This study has demonstrated significant healing responses with reduction in burn size at lower levels of clinical visibility using a 20-millisecond pulse laser. A novel observation was the complete realignment of outer highly reflective layer architecture for subvisible 20-millisecond lesions. Longer pulse duration produced increased disruption of IS-OS and RPE layers, greater perilesional photoreceptor atrophy, and variable changes in lesion size over time. The spatial tissue oxygenation increased significantly after laser photocoagulation of diabetic retinas, and we report an innovative method of noninvasively measuring spatial tissue oxygenation.
All clinical grades of 20-millisecond laser burns reduced on average 49%–54% in size from the 1-hour time point. This significant reduction in size was associated with greater RPE repopulation and photoreceptor in-filling.
10,16 The 6-month FD-OCT appearances of subthreshold 20-millisecond burns appear similar to selective retinal therapy laser lesions.
23
In this study, we have shown that increasing intensity of 100- and 200-millisecond pulse lesions have a different spatial and spectral natural history compared with 20-millisecond lesions. The 1-hour FD-OCT images of 20-, 100-, and 200-millisecond pulse burns appeared similar, but the temporal laser lesion-tissue interactions were determined by fluence and pulse duration. At 6 months, increasing levels of carbonization, as defined by hyper-pigmentation and hypertrophic tissue scars, were observed for threshold and suprathreshold 100- and 200-millisecond lesions. In contrast, threshold, barely visible and subvisible 20-millisecond burns produced sustained healing responses over the long term.
Sramek and coworkers have demonstrated a more even distribution of spatial heat/temperature changes within RPE using smaller spot laser lesions.
24 Laser photocoagulation produces thermal destruction of photoreceptor inner segments that result in improved intraretinal oxygen delivery via increased diffusion of choroid-derived oxygen.
25 The rise in tissue or preretinal PO
2 after laser photocoagulation has been demonstrated in animal studies.
26 –28 Subthreshold micropulse laser can produce increased intraretinal oxygen levels and reduced tissue oxygen consumption.
29 During vitreous surgery, Stefánsson and co-workers
30 demonstrated significantly increased oxygen tensions over scatter laser-treated areas. In our study, we report for the first time in vivo the healing responses that occur with improved tissue oxygenation at both the location of laser burns and in surrounding retinal tissue using a 20-millisecond laser. The longer pulse lesions produced increased spatial oxygenation but resulted in greater collateral tissue damage over time.
At present, there is no precise and quantitative method of noninvasively measuring tissue oxygenation of laser-treated areas in vivo. Our multispectral imager demonstrated an ability to detect small but significant improvements in retinal oxygenation at the locations of laser lesions and in the surrounding tissue over time. These improvements were still apparent after correcting for measurement variability, which suggests that this technique may be of use in monitoring changes in tissue oxygenation over time and post-treatment (laser, intravitreal drugs, vitrectomy). The multispectral findings could be correlated with improvements in angiographic retinal ischemia. Although we were unable to detect any significant differences in oxygenation between individual laser lesions, this may be explained by the small sample size. Another possible explanation could be that more than one process (fluence or pulse duration) has produced the increased oxygenation within the coagulated tissue. Regarding validation with previous animal studies, the shifts in spatial oxygenation observed using our multispectral camera system produced similar oxygenation patterns reported by Zuckerman and coworkers
26 after laser photocoagulation.
The application of this technology to diabetic retinopathy may also include evaluation of tissue oxygenation levels within angiographic tissue infarction that could guide future laser management and perhaps also be used to quantify foveal oxygenation in diabetics. Although the multispectral pilot data from this study are limited, we present the data here to demonstrate the potential of this technique in diabetic patients, and this investigation may have a role in larger therapeutic clinical trials.
For clinical practice, the laser photocoagulation parameters need to be carefully controlled. Laser fluence will determine the intraretinal temperature rise and energy reflected back from the RPE layers.
8 Laser burns at 20 milliseconds have reduced fluence at all clinical grades of burn compared with 100 and 200 milliseconds. Despite lower fluence, all 20-millisecond laser lesions demonstrated effective uptake at the level of the RPE. In vivo, the shrinkage of 20-millisecond laser burns together with reparative structural changes may be explained by a potential heterogeneity in sensitivity of RPE cells to laser injury. At subvisible levels, the maximal reduction in laser lesion and restoration of cellular layers was observed, and this laser intensity was close to the minimum effective dose of laser irradiation used in our study.
A study limitation includes the low time points used for study, with analysis not performed at 1 day or 1 week after laser application. More short-term data may strengthen the observations of laser-tissue interactions. However, we previously studied the OCT and AF findings at 1 week and did not observe any significant differences in outer retina architecture between the 1-hour and 1-week time points, and this study aimed to address the laser-tissue interactions over the longer term.
15 In a recent publication by Kriechbaum et al.,
31 the OCT features of laser lesions were analyzed at 1 day. In that study, the laser lesions appear similar in appearance to the 1-hour images we have demonstrated.
A further weakness of this work could be the possibility of an SLO-based AF imaging system producing greater interpretation of the AF results in this study. The current SLO-AF systems use excitation at ∼488 nm and barrier filter at ∼500 nm. However, the new modified flash system (Topcon TRC-50DX, type IA) wavelength bandwidths used for excitation and barrier filters are closer together, with excitation closer to yellow-orange. The barrier filter now has a longer wavelength bandpass, compared with the SLO-AF, which straddles wavelengths from precursor fluorophores to improve the AF signal.
15,32
It could be argued that potential variability exists in laser uptake between each laser burn because of inconsistent aiming beam focus. However, we used FD-OCT to visualize the burns at 1-hour post-treatment, and the visibility of burns correlated with the degree of hyper-reflectivity originating from the RPE layer through retinal layers.
In any eye, there will be variations in fundus pigmentation and melanosome populations. We used the retinal quadrant nasal to the disc margin, because this is recognized to have less variation in pigmentation and is less likely to be affected by difference in ocular axial length. In all 120 laser lesions evaluated, the laser titration lesions produced similar morphologic and visible changes in retinal tissue using all our imaging modalities. Furthermore, the retinal area nasal to the optic disc did not show any significant alteration in ischemia on FFA, and this would have minimal impact on any oxygenation changes attributed directly to laser lesions. The changes in tissue oxygenation were also validated by comparison with the relative oxygen saturation within the blood vessels, and we did not detect any significant intravascular changes over time.
This study highlights the potential of multispectral imaging as a noninvasive tool to measure spatial tissue oxygenation in diabetic retina. For patients undergoing laser for ischemic and proliferative vascular retinopathy, there may be improved tissue oxygenation, higher predictability of burn morphology, and more spatial localization of healing responses of burns over time using reduced fluence 20-millisecond laser pulses compared with conventional laser pulses.
Supported by Optimedica Corp., the Manchester Academic Health Sciences Centre, and the NIHR Manchester Biomedical Research Centre. JD was funded by a College of Optometrists PhD Studentship, United Kingdom.
Disclosure:
M.M.K. Muqit, None;
J. Denniss, None;
V. Nourrit, None;
G.R. Marcellino, Optimedica Corp. (E);
D.B. Henson, None;
I. Schiessl, None;
P.E. Stanga, Optimedica Corp. (F, R)
The authors thank all staff within the Clinical Imaging Unit at MREH for performing optical coherence tomography, imaging, and fundus photography.