Subvisible retinal laser therapy: titration algorithm and tissue response.

PURPOSE: Laser therapy for diabetic macular edema and other retinal diseases has been used within a wide range of laser settings: from intense burns to nondamaging exposures. However, there has been no algorithm for laser dosimetry that could determine laser parameters yielding a predictable extent of tissue damage. This multimodal imaging and structural correlation study aimed to verify and calibrate a computational model-based titration algorithm for predictable laser dosimetry ranging from nondamaging to intense coagulative tissue effects. METHODS: Endpoint Management, an algorithm based on a computational model of retinal photothermal damage, was used to set laser parameters for various levels of tissue effect. The algorithm adjusts both power and pulse duration to vary the expected level of thermal damage at different percentages of a reference titration energy dose. Experimental verification was conducted in Dutch Belted rabbits using a PASCAL Streamline 577 laser system. Titration was performed by adjusting laser power to produce a barely visible lesion at 20 ms pulse duration, which is defined as the nominal (100%) energy level. Tissue effects were then determined for energy levels of 170, 120, 100, 75, 50, and 30% of the nominal energy at 1 hour and 3, 7, 30, and 60 days after treatment. In vivo imaging included fundus autofluorescence, fluorescein angiography, and spectral-domain optical coherence tomography. Morphologic changes in tissue were analyzed using light microscopy, as well as scanning and transmission electron microscopy. RESULTS: One hundred and seventy percent and 120% levels corresponded to moderate and light burns, respectively, with damage to retinal pigment epithelium, photoreceptors, and at highest settings, to the inner retina. 50% to 75% lesions were typically subvisible ophthalmoscopically but detectable with fluorescein angiography and optical coherence tomography. Histology in these lesions demonstrated some selective damage to retinal pigment epithelium and photoreceptors. The 30% to 50% lesions were invisible with in vivo multimodal imaging, and damage was limited primarily to retinal pigment epithelium, visible best with scanning electron microscopy. Over time, photoreceptors shifted into the coagulated zone, reestablishing normal retinal anatomy in lesions ≤100%, as seen in optical coherence tomography and light microscopy. Transmission electron microscopy at 2 months demonstrated restoration of synapses between shifted-in photoreceptors and bipolar cells in these lesions. Retinal pigment epithelium monolayer restored its continuity after 1 week in all lesions. No damage could be seen <30% level. CONCLUSION: A retinal laser dosimetry protocol based on the Endpoint Management algorithm provides reproducible changes in retinal morphology in animals with various levels of pigmentation. This algorithm opens doors to clinical trials of well-defined subvisible and nondestructive regimes of retinal therapy, especially important for treatment of macular disorders.

Retinal safety of near-infrared lasers in cataract surgery.

Femtosecond lasers have added unprecedented precision and reproducibility to cataract surgery. However, retinal safety limits for the near-infrared lasers employed in surgery are not well quantified. We determined retinal injury thresholds for scanning patterns while considering the effects of reduced blood perfusion from rising intraocular pressure and retinal protection from light scattering on bubbles and tissue fragments produced by laser cutting. We measured retinal damage thresholds of a stationary, 1030-nm, continuous-wave laser with 2.6-mm retinal spot size for 10- and 100-s exposures in rabbits to be 1.35 W (1.26 to 1.42) and 0.78 W (0.73 to 0.83), respectively, and 1.08 W (0.96 to 1.11) and 0.36 W (0.33 to 0.41) when retinal perfusion is blocked. These thresholds were input into a computational model of ocular heating to calculate damage threshold temperatures. By requiring the tissue temperature to remain below the damage threshold temperatures determined in stationary beam experiments, one can calculate conservative damage thresholds for cataract surgery patterns. Light scattering on microbubbles and tissue fragments decreased the transmitted power by 88% within a 12 deg angle, adding a significant margin for retinal safety. These results can be used for assessment of the maximum permissible exposure during laser cataract surgery under various assumptions of blood perfusion, treatment duration, and scanning patterns.

Therapeutic window of retinal photocoagulation with green (532-nm) and yellow (577-nm) lasers.

BACKGROUND AND OBJECTIVE: The 577-nm (yellow) laser provides an alternative to the 532-nm (green) laser in retinal photocoagulation, with potential benefits in macular treatment and through ocular opacities. To assess relative risk of thermomechanical rupture of Bruch's membrane with yellow laser in photocoagulation, the therapeutic window, the ratio of threshold powers for mild coagulation and rupture, was measured. MATERIALS AND METHODS: Retinal coagulation and rupture thresholds, visualized ophthalmoscopically, were measured with 577- and 532-nm lasers using 10- to 100-ms pulses in 34 rabbit eyes. Lesions at 1 and 7 days were assessed histologically. RESULTS: Coagulation threshold with yellow laser was 26% lower than with green laser. The therapeutic window increased linearly with log-duration for both wavelengths with a difference in parallel-slope intercept of 0.36 ± 0.20, corresponding to 8% to 15% wider therapeutic window for yellow wavelength. CONCLUSION: The therapeutic window of retinal photocoagulation in rabbits at 577 nm is slightly wider than at 532 nm, whereas histologically the lesions are similar.

Long-term safety, high-resolution imaging, and tissue temperature modeling of subvisible diode micropulse photocoagulation for retinovascular macular edema.

PURPOSE: To determine the long-term safety of high-density subvisible diode micropulse photocoagulation (810 nm), compare the clinical findings with computational modeling of tissue hyperthermia and to report results for a subset of eyes treated for diabetic macular edema (ME) documented pre- and postoperatively by spectral-domain optical coherence tomography. METHOD: All eyes treated for ME from diabetic retinopathy (diabetic ME) and branch retinal vein occlusion between April 2000 and January 2010 were reviewed for subvisible diode micropulse laser-induced retinal damage. Therapeutic outcomes were reviewed for a subgroup treated for diabetic ME with pre- and postoperative spectral-domain optical coherence tomography. Laser-induced retinal thermal effects were modeled computationally using Arrhenius formalism. RESULTS: A total of 252 eyes (212 diabetic ME, 40 branch retinal vein occlusion) of 181 patients qualified. None of the 168 eyes treated at irradiance <350 W/cm2 and 7 of 84 eyes at ≥ 590 W/cm2 had retinal damage (P = 0.0001) (follow-up 3-120 months, median, 47). Sixty-two eyes of 48 patients treated for diabetic ME with pre- and postoperative spectral-domain optical coherence tomography with median 12 months follow-up had no retinal injury by infrared, red-free, or fundus autofluorescence photos; fluorescein angiography or indocyanine green angiography; or spectral-domain optical coherence tomography. Central foveal thickness (P = 0.04) and maximum macular thickness decreased (P < 0.0001). Modeling of retinal hyperthermia demonstrates that the sublethal clinical regimen corresponds to Arrhenius integral >0.05, while damage is likely to occur if it exceeds 1. CONCLUSION: Subvisible diode micropulse can effectively treat retinovascular ME without laser-induced retinal damage, consistent with Arrhenius modeling of pulsed hyperthermia.

Improving the therapeutic window of retinal photocoagulation by spatial and temporal modulation of the laser beam.

Decreasing the pulse duration helps confine damage, shorten treatment time, and minimize pain during retinal photocoagulation. However, the safe therapeutic window (TW), the ratio of threshold powers for thermomechanical rupture of Bruch's membrane and mild coagulation, also decreases with shorter exposures. Two potential approaches toward increasing TW are investigated: (a) decreasing the central irradiance of the laser beam and (b) temporally modulating the pulse. An annular beam with adjustable central irradiance was created by coupling a 532-nm laser into a 200-µm core multimode optical fiber at a 4-7 deg angle to normal incidence. Pulse shapes were optimized using a computational model, and a waveform generator was used to drive a PASCAL photocoagulator (532 nm), producing modulated laser pulses. Acute thresholds for mild coagulation and rupture were measured in Dutch-Belted rabbit in vivo with an annular beam (154-163 µm retinal diameter) and modulated pulse (132 µm, uniform irradiance "flat-top" beam) with 2-50 ms pulse durations. Thresholds with conventional constant-power pulse and a flat-top beam were also determined. Both annular beam and modulated pulse provided a 28% increase in TW at 10-ms duration, affording the same TW as 20-ms pulses with conventional parameters.

Selective retinal therapy with microsecond exposures using a continuous line scanning laser.

PURPOSE: To evaluate the safety, selectivity, and healing of retinal lesions created using a continuous line scanning laser. METHODS: A 532-nm Nd:YAG laser (PASCAL) with retinal beam diameters of 40 µm and 66 µm was applied to 60 eyes of 30 Dutch-belted rabbits. Retinal exposure duration varied from 15 µs to 60 µs. Lesions were acutely assessed by ophthalmoscopy and fluorescein angiography. Retinal pigment epithelial (RPE) flatmounts were evaluated with live-dead fluorescent assay. Histological analysis was performed at 7 time points from 1 hour to 2 months. RESULTS: The ratios of the threshold of rupture and of ophthalmoscopic visibility to fluorescein angiography visibility (measures of safety and selectivity) increased with decreasing duration and beam diameter. Fluorescein angiography and live-dead fluorescent assay yielded similar thresholds of RPE damage. Above the ophthalmoscopic visibility threshold, histology showed focal RPE damage and photoreceptor loss at 1 day, without inner retinal effects. By 1 week, photoreceptor and RPE continuity was restored. By 1 month, photoreceptors appeared normal. CONCLUSION: : Retinal therapy with a fast scanning continuous laser achieves selective targeting of the RPE and, at higher power, of the photoreceptors without permanent scarring or inner retinal damage. Continuous scanning laser can treat large retinal areas within standard eye fixation time.

Non-damaging retinal phototherapy: dynamic range of heat shock protein expression.

PURPOSE: Subthreshold retinal phototherapy demonstrated clinical efficacy for the treatment of diabetic macular edema without visible signs of retinal damage. To assess the range of cellular responses to sublethal hyperthermia, expression of the gene encoding a 70 kDa heat shock protein (HSP70) was evaluated after laser irradiation using a transgenic reporter mouse. METHODS: One hundred millisecond, 532 nm laser exposures with 400 µm beam diameter were applied to the retina surrounding the optic nerve in 32 mice. Transcription from the HSP70 promoter was assessed relative to the control eye using a bioluminescence assay at 7 hours after laser application. The retinal pigmented epithelium (RPE) viability threshold was determined with a fluorescence assay. A computational model was developed to estimate temperature and the extent of cell damage. RESULTS: A significant increase in HSP70 transcription was found at exposures over 20 mW, half the threshold power for RPE cell death. Computational modeling estimated peak temperature T = 49°C at HSP70 expression threshold. At RPE viability threshold, T = 57°C. Similar temperatures and damage indices were calculated for clinical subvisible retinal treatment parameters. CONCLUSIONS: Beneficial effects of laser therapy have been previously shown to extend beyond those resulting from destruction of tissue. One hundred millisecond laser exposures at approximately half the threshold power of RPE damage induced transcription of HSP70, an indication of cellular response to sublethal thermal stress. A computational model of retinal hyperthermia can guide further optimization of laser parameters for nondamaging phototherapy.

Dynamics of retinal photocoagulation and rupture.

In laser retinal photocoagulation, short (<20 ms) pulses have been found to reduce thermal damage to the inner retina, decrease treatment time, and minimize pain. However, the safe therapeutic window (defined as the ratio of power for producing a rupture to that of mild coagulation) decreases with shorter exposures. To quantify the extent of retinal heating and maximize the therapeutic window, a computational model of millisecond retinal photocoagulation and rupture was developed. Optical attenuation of 532-nm laser light in ocular tissues was measured, including retinal pigment epithelial (RPE) pigmentation and cell-size variability. Threshold powers for vaporization and RPE damage were measured with pulse durations ranging from 1 to 200 ms. A finite element model of retinal heating inferred that vaporization (rupture) takes place at 180-190 degrees C. RPE damage was accurately described by the Arrhenius model with activation energy of 340 kJ/mol. Computed photocoagulation lesion width increased logarithmically with pulse duration, in agreement with histological findings. The model will allow for the optimization of beam parameters to increase the width of the therapeutic window for short exposures.

Synchronization of unidirectionally coupled Mackey-Glass analog circuits with frequency bandwidth limitations.

Synchronization of chaotic systems has been studied extensively, and especially, the possible applications to the communication systems motivated many research areas. We demonstrate the effect of the frequency bandwidth limitations in the communication channel on the synchronization of two unidirectionally coupled Mackey-Glass (MG) analog circuits, both numerically and experimentally. MG system is known to generate high dimensional chaotic signals. The chaotic signal generated from the drive MG system is modified by a low pass filter and is then transmitted to the response MG system. Our results show that the inclusion of the dominant frequency component of the original drive signals is crucial to achieve synchronization between the drive and response circuits. The maximum cross correlation and the corresponding time shift reveal that the frequency-dependent coupling introduced by the low pass filtering effect in the communication channel change the quality of synchronization.