Subretinal or suprachoroidal hemorrhage may develop during choroidal surgery and may lead to irreversible vision loss and other complications.
8–10 Devices that may achieve intraoperative hemostasis and avoid these complications are divided into several categories: laser modalities, electrosurgery (surgical diathermy), electrocautery (thermal cautery), and regulating the intraocular tamponade.
11,12 Also, endodiode laser thermal treatments along with endotamponade (during a chorioretinal biopsy)
13 and aggressive diathermy combined with systemic hypotension have been described previously.
14 All such methods (except tamponade), involve the application of thermal energy to limit intraocular bleeding, which may, in turn, lead to collateral tissue injury.
Our goal in designing a model to analyze this thermal tissue injury was to simulate the thermal energy transfer from ophthalmic devices into the highly vascular CBR tissue.
15,16 Although we did not fully recreate the intraocular, fluid-filled eye, our model allowed us to determine thermal tissue spread in a controlled environment and can be modified to simulate a fluid environment.
In our study, we analyzed and compared thermal spread from four modalities, each capable of addressing retinal and choroidal hemostasis: PEAK, MpL, CL, and BpC. Not surprisingly, we found that power output was directly proportional to the degree of lateral thermal spread, a finding supported by several studies.
17–19 However, the amount of tissue injury was significantly different between devices using identical power output.
BpC and MpL had the least thermal spread at all three energy levels. We suspected the BpC would have a greater lateral thermal spread than the MpL because of its continuous power output and no duty cycle; however, this was not observed. Thermal energy delivery to the tissue may have been attenuated owing to the simulated study environment with BpC, in which the fluid-device interface is limited (moist, but not immersed in fluid), and less device-to-tissue heat transfer than may occur in a more standard fluid-enriched surgical environment (wet-field cautery). In most ophthalmologic surgical conditions, wet-field cautery is immersed in a saline, aqueous solution, especially in the posterior segment. Our rationale for minimizing the fluid interface was based on the measurement requirements of the infrared camera. Fluid irrigation would have added a cooling variable (i.e., the thermal effect measured is less than in clinical use) that was difficult to control. Also, there was a meaningful loss in thermal measurement sensitivity with increasing fluid medium.
The MpL had the next-lowest thermal impact across all three energy groups, and the measurements were repeatable. We anticipated that the MpL would have the lowest thermal propagation because of its precise area of energy application, combined with a low duty cycle. Subthreshold MpLs operate in short repetitive pulses of 100- to 300-microsecond intervals within an “envelope” lasting 0.1 to 0.5 seconds.
20 The envelope consists of a duty cycle; that is, the percentage of time when the laser is active within the envelope. Laser “on” time ranges from 5% to 15%.
20 During the laser “off” time, the tissue is able to cool, which limits thermal propagation into adjacent RPE, choriocapillaris, and surrounding tissue, especially when compared to CL applications at the same energy output.
21,22 We found significantly less thermal spread of the MpL compared to the CL in all energy groups (
Table 1). We explained this by the equivalent total power output of the MpL and CL. In the clinical setting, a similar tissue effect could be obtained by using lower power settings on the MpL.
We used a subthreshold, 810-nm wavelength diode MpL. Other studies have shown that the 810-nm wavelength has a higher absorption in pigmented cells within the choroid.
23–26 Although hemostasis was not directly tested, the 810-nm diode MpL can potentially address the choroidal vasculature, even by directing the laser energy through transparent retina, while minimizing collateral thermal injury to the delicate RPE and neurosensory retina. Further testing of the MpL at lower power settings with in vivo studies will help confirm this theoretic advantage of MpL in the clinical-surgical setting.
The PEAK had the highest thermal spread of all modalities tested. Current PEAK devices have a much broader power range than we tested, and the energy output is adjusted for use in general surgery. The PEAK is the thinnest needle-delivery device. Although the power range is suboptimal for ophthalmic purposes, output range varies greatly from the original PEAK for fine cutting (PEAK-fc), which produces incisions through the retina with up to 100 μm collateral tissue damage.
27,28 The histologic analysis underestimates thermal spread because damage is measured as a function of the highest temperature reached, along with the duration of exposure.
29 Measuring the critical threshold temperature for permanent tissue damage is more likely to reflect the long-term effects on loss of delicate RPE and choriocapillaris. Our camera recorded the PEAK device to operate between the acceptable temperature ranges of 40°C and 170°C, as has been documented in other studies.
30
The PEAK medium and high energy settings could effectively make full-thickness choroidal incisions, while also applying thermal energy for hemostasis; in comparison, at the low-energy setting, PEAK only disrupted the RPE cell layer (
Figs. 7A–
C). Visible collateral thermal damage was seen with PEAK at the high energy setting, but not the low or medium energies (
Figs. 7A–
C). Thus, we conclude that the PEAK medium output is the best energy level to provide the ideal cutting and coagulation required for incisional surgery through CBR tissue.
The PEAK is limited by the formation of cavitation bubbles, as were observed at the medium and high energy levels (
Figs. 7D,
7E). PEAK technology causes the rapid evaporation of liquid, with explosive evaporation, resulting in shock waves and secondary cavitation bubbles.
31,32 Gas bubbles from the PEAK-fc are comparable in conventional electrosurgery.
27,28 In all PEAK blades we tested, we noted a change in the shape and length of the 12.5-μm tungsten electrode tip. Plasma microstreamers originate from the tungsten electrode and facilitate tissue dissection. Tungsten carbide is one of the most durable metals and is much harder than stainless steel.
33 Because the melting point of tungsten carbide is 3422°C, it is unlikely to melt at the operational temperature of this device (40°C–170°C). Interestingly, the shape of the tungsten electrode varied, even among new blades (
Fig. 8).
We also analyzed the thermal spatial variation between laser modalities. Although we did not detect meaningful variation in temperature or thermal spread when changing tissue orientation, the size and color of posttreatment RPE distortion was not uniform. Others have noted that the ophthalmoscopic appearance of laser burns varies in size and color within the same eye, despite identical laser settings.
34,35 The variation seen in our data, similar to that of Yu et al.,
34 is most likely due to the normal variation of the choroidal and RPE pigmentation,
36 choroid thickness, or perhaps other unknown variables (e.g., choroidal blood flow that could not be tested in this model). Also, the two-dimensional thermographic profile limits our ability to determine the depth of thermal transfer.
We believe this information will be helpful clinically for techniques that require incision and hemostasis through the choroid. Minimizing collateral tissue loss of the delicate RPE and neurosensory retina is especially important when working near the macula or optic nerve, because such damage would certainly result in symptomatic loss of visual acuity or visual field, respectively. Others have shown that lasers create a visible burn that propagates from the original treatment zone, resulting in progressive RPE atrophy over time.
37 Thus, managing thermal energy spread during ocular procedures may lead to less surrounding tissue injury.
In our model to test ophthalmic surgical devices delivering thermal energy to fresh, ex vivo choroidal tissue and to measure the secondary injury using standardization of energy transfer, we tested four surgical devices. Each device was a candidate technology for effective choroidal hemostasis during incisional CBR surgery. We found that power output was directly proportional to the degree of lateral thermal spread. Also, tissue injury was significantly different between devices delivering identical power outputs. The MpL and BpC were the most effective for minimizing collateral thermal damage. However, we believed that our system underestimated the thermal spread of the BpC device, because a suboptimal fluid interface was required for the thermal camera to obtain reliable results. We understood that the decreased thermal spread was most likely due to an inadequate fluid interface. However, by adding irrigation to the tissue in the BpC group, we would add a variable that would alter the thermographics for this technology. The PEAK blade had the greatest thermal spread, yet also can create a full-thickness, tractionless incision through the choroid. Thus, BpC and PEAK offer substantial surgical advantages. The PEAK may be limited by cavitation bubble formation, especially at higher energy levels. MpL and CL had similar results, yet we believed that MpL may require lower power settings to achieve hemostasis in the clinical environment. Our data are clinically relevant for exploring approaches to incisional choroidal surgery with the goal of minimizing collateral tissue damage. We believed that our analysis has implications for surgical techniques, such as chorioretinal biopsies, tumor excisions, traumatic globe repair, macular translocation surgery, and potentially other macular-regenerative surgical approaches.