LASIK eye surgery has become a very common practice for myopic people, especially those in the military. Sometimes undertaken by people who need to keep a specific medical aptitude, this surgery could be performed in secret from the hierarchy and from the institute medical staff. However, even though the eyes have been previously described as one of the most sensitive organs to electromagnetic fields in the human body, no data exist on the potential deleterious effects of electromagnetic fields on the healing eye. The consequences of chronic long-lasting radar exposures at power density, in accordance with the occupational safety standards (9.71 GHz, 50 W/m2), were investigated on cornea healing. The metabolic and clinical statuses after experimental LASIK keratotomy were assessed on the different eye segments in a New Zealand rabbit model. The analysis methods were performed after 5 months of exposure (1 hour/day, 3 times/week). Neither clinical or histological examinations, nor experimental data, such as light scattering, 1H-NMR HRMAS metabolomics, 13C-NMR spectra of lipidic extracts, and antioxidant status, evidenced significant modifications. It was concluded that withdrawing the medical aptitude of people working in electromagnetic field environments (i.e., radar operators in the navy) after eye surgery was not justified.
LASIK surgery is routinely used to correct refraction abnormalities, such as myopia, hypermetropia, or astigmatism, without wearing glasses or contact lenses [
The eyes are known to be one of the most sensitive organs in the human body to electromagnetic field exposure. This specific susceptibility has been widely described and is mainly related to the low level of eye vascularization leading to low heat dispersion. This link was identified quickly, especially by considering links between cataracts and professional microwave exposure (radar operators) [
Besides these well-known effects, more recent works have dealt with the potential hazardous nonthermal effects of radiofrequency on the eyes. Although no mechanism has been clearly exposed, some authors have identified deleterious effects in different parts of the eye, such as the lens [
The healing eye, after situations like ocular wounds or repair following surgery, appears to be an even more sensitive organ. Scars on the cornea and refractive modification could occur, as for light and electromagnetic field dispersion, and could lead to a local specific absorption rate (SAR) increase. Lens opacities and transparent intraocular media composition would also result in local energy absorption modifications.
To our knowledge, no previous works have focused on the deleterious effects of electromagnetic fields on a healing eye after wounding or refractive surgery. By considering the other eye-aggressive agents such as UV and diabetes, corticotherapy can lead to an increase in the promotion of reactive oxygen species and to metabolic changes in the different eye targets [
All procedures were in accordance with the standards for animal care established by the Army Biomedical Research Institute (IRBA) and were approved by the IRBA ethics committee for animal experimentations (decree 87-848 19 October 1987, edited by the French government).
60 male New Zealand rabbits (Elevage Charles River, France), weighing 2.5–3 kg upon arrival, were used. Animals were housed 1 per cage at control room temperature (21°C) with 12 h/12 h light-dark cycle (dark period from 6 p.m. to 6 a.m.). Food and water were available ad libitum. The animals were randomly assigned to four different groups: surgery/exposed, surgery/sham exposed, sham surgery/exposed, and sham/sham.
Anesthesia was obtained by intramuscular injection of Zolazepam/Tiletamine 10 mg/kg (Zoletil 100, Merieux, France). In addition, topical bupivacaine hydrochloride 1% (Roche, France) was applied to each eye just before surgery. Bilateral keratotomy was realized using a one-use Moria microkeratome, set to cut a flap 8.5 mm in diameter and 100
Exposure devices were based on a pulsed X Band ESR spectrometer, ESP 380 system (Brüker, France) with a 1 kW travelling wave-tube amplifier (TWT). The TWT output was directly connected to a horn antenna (length 17 cm, gain 20 dB) located in an anechoic chamber. The system emitted a typical pulsed electromagnetic field at 9.71 GHz, with a pulse length of 9000 ns and a 1% duty cycle.
During exposure, the animal was put in polycarbonate restrainers, located in the anechoic chamber at 30 cm from the end of the antenna. According to their group, animals were then exposed or sham exposed 1 month after surgery for 1 h/day, 3 days a week for 5 months, at a power density of 50 W/m², which is the ICNIRP exposure recommendation for workers.
The electromagnetic exposure conditions were controlled and measured before each experiment using a multifrequency anisotropic field meter (FP7050, AR worldwide, USA).
The specific absorption rate (SAR) was estimated by direct measurement of temperature increase at higher power density. It was derived according to the following equation: SAR =
Clinical ophthalmologic examination was performed the day after the surgery, one week after, one month after (before the first RF exposure), and then at three and six months (before euthanasia). This exam was performed with an ophthalmoscope and consisted of the observation of any anomalies of the cornea, the lens, and the fundus oculi. The integrity of the cornea was assessed by fluorescein examination.
At the end of the experiment, before euthanasia, blood samples were collected in lithium heparin sample tubes for oxidative stress analysis. The animals were euthanized by a lethal injection of pentobarbital (Dolethal, Vetoquinol, France). Aqueous humor was punctured and immediately frozen in liquid nitrogen for oxidative stress analysis. The corneoscleral rims were removed with Westcott scissors. One cornea was used for histological analysis and then placed in a 4% formaldehyde solution; the other was used for NMR analysis and immediately frozen in liquid nitrogen. After light scattering measurement, lenses were frozen for NMR analysis.
The optical quality of the lens was assessed by light scattering, as described by Söderberg et al. [
Light scattering principle. Position A represents the light path through a normal lens, while position B represents the light path through a lens with diffraction abnormalities.
After euthanasia, the lens was removed from the eye and immediately placed in a dark measuring chamber. The beam of the 670 nm diode laser located at 30 mm goes through the lens with a 45° incidence angle. If the lens is damaged, its ability to focus the laser beam at the various locations is altered, leading to an increase in the light diffusion.
The quantification of oxidative stress was assessed in blood serum and in aqueous humor by using an Antioxidant Assay Kit (Caymann Chemical, USA), according to the manufacturer’s instructions. This kit measures the total antioxidant capacity. The assay relies on the ability of the antioxidant in the sample to inhibit the oxidation of ABTS to ABTS+ by myoglobin. Trolox, a water-soluble tocopherol, is used as a standard; the antioxidant capacity of the sample is quantified as millimoles of Trolox equivalents. The optical density of each well was measured using a microtiter plate fluorometer Mithras LB 940 at 405 nm (Berthold Technologies, Bad Wildbad, Germany) and processed using the MikroWin software.
All high-resolution NMR spectra in liquid solution were recorded on a Bruker AM400 spectrometer (9.4T) operating at 400.13 MHz for protons.
1H-NMR spectra from aqueous and vitreous humors were recorded after lyophilization of the samples. The pellets were suspended in 500
13C-NMR experiments on lenses and corneas were recorded on the lipidic phase of methanol/chloroform/water (2/1/0.8 V/V/V) extracts obtained according to the method of Bligh and Dyer [
For the spectra of lenses, the signal-to-noise was of better quality and allowed to measure and normalize peak intensities for the different resonances used to build the index to compare groups. It should be noted that the normalization of the data used the sum of the integrals of the 180–100 ppm, 75.5–49.7 ppm, and 45–10 ppm regions to avoid dramatic errors related to the intense solvent (methanol and chloroform) resonances.
The selected peaks of interest are shown in Table
NMR peaks of interest.
Molecule | Proton nomenclature and chemical shift (ppm) | Label |
---|---|---|
Acyl chains | ||
Terminal methyl | tCH3 area 14.4–14.8 | G |
Chain methylenic | CH2n area 29.8–31.2 | H |
Methynic | CH = area 128–131 | I |
Carboxylic | CO area 174–177 | J |
|
||
Cholesterol | Peak intensities | |
C5 | 142.6 | K |
C6 | 121.7 | L |
C17 | 56.5 | M |
|
||
Sugar moieties of glucolipids | Peak intensities | |
C1 | 104.1 | N |
C6 | 62 | O |
|
||
Phospholipid head groups | Peak intensities | |
Sphingomyelin and phosphatidylcholine, PCSM | |
P, Q, R |
Phosphatidylserine, PS | |
T, U |
Phosphatidylethanolamine, PE | |
V |
Phosphatidylinositol, PI | CH2 75 | W |
Phospholipid glycerol, PG | C1–C3 64-65 area | X |
Plasmalogen, PEPL | C2′ 108; C1′ 145 | Y, Z |
NMR index comparison. The letter in formula column refers to the label of the peak of interest.
Index | Formula |
---|---|
Chain length | (G + H + I + J)/G |
Chain insaturation | I/(G + H + I + J) |
Cholesterol/glycerol | (K + L + M)/X |
Cholesterol/chains | (K + L + M)/(G + H + I + J) |
|
|
Contribution to the phospholipids of | |
Phosphatidylserine, PS | (T + U)/(P + Q + R + T + U + V + W + Y + Z) |
Phosphatidylethanolamine, PE | V/(P + Q + R + T + U + V + W + Y + Z) |
Sphingomyelin and phosphatidylcholine, PCSM | S/(P + Q + R + T + U + V + W + Y + Z) |
Plasmalogen, PEPL | (Y + Z)/(P + Q + R + T + U + V + W + Y + Z) |
|
|
Sugar to phospholipids ratio | (N + O)/(P + Q + R + T + U + V + W + Y + Z) |
Immediately after the light scattering measurement, the rabbit lenses were frozen in liquid nitrogen. The samples were then stored at −80°C to ensure conservation before NMR experiments.
For HRMAS NMR analyses, a 15 mg fragment was obtained by cutting the lenses with a surgical blade. The shape of the biopsy was a circular coin, which was obtained by using a solid matrix guiding the blade. Such samples contained almost all of the different parts of the lenses, from the inside to the external layer. Such pieces were transferred into a HRM NMR 50
1H-HRMAS NMR data were recorded at 400.13 MHZ on a Bruker Avance (Bruker Biospin, Wissembourg, France) NMR spectrometer. A modified spin-echo sequence (Carr-Purcell Meiboom Gill) was used: the echo had to be synchronized with the sample rotation speed (4 kHz). The 30 ms echo time allowed minimization of the contributions to the spectra of broad macromolecules and lipidic resonances. Π-pulses were separated by 250
Corneas from control or treated groups were fixed with 3.7% formaldehyde (Sigma Aldrich, Saint Louis, Missouri, USA) in a sodium phosphate buffer solution at a pH of 7.4. After paraffin embedding, 5
Classical tests: all results are presented as mean ± SEM. The 4 different exposure conditions were randomized: surgery/exposed, surgery/sham exposed, sham surgery/exposed, and sham/sham. Statistical comparisons were achieved using nonparametric tests.
Statistics for PCA, PLS-DA: here, the populations were gathered in “exposed” versus “nonexposed (sham)” classes following the results of nonparametric tests. The NMR data then had to be processed with the Bruker Xwin-NMR v2 software. The spectra were apodized by using a 0.3 Hz Lorentzian convolution prior to Fourier transformation. Baseline correction was also automatically realized. The final spectra were finally segmented in spectral bands of 0.2 ppm (AMIX v3.5 software, Bruker-Biospin, France). These bands were then statistically analyzed with SIMCA v12 (Umetrics, Umea, Sweden); a principal components analysis (PCA) was first performed to identify possible gathering data and unclassified aberrant criterions. PLS-DA (Projection to Latent Structure-Discriminant Analysis) was finally built to maximize intergroup variance and to minimize the intragroup variance. This processing assumed that a given result to the sham or exposed groups was realistic.
Observation of the periocular area and of the cornea did not show any inflammation, edema burns, or conjunctivitis in any case. However, if no keratitis, corneal opacity, or traces of surgery were found, limited hazes were detected in two cases on exposed rabbits. Ophthalmoscope examination showed neither lens opacity nor eye floaters within the vitreous humor, meaning that there was no major cataract formation or vitreous degeneration.
This examination was completed by light diffusion measurements of the lens, as shown in Figure
Light scattering diffusion. For each group, the value was the mean of 15 measures ± SD. Significance was determined using the Mann-Whitney
Six months after surgery, the aspect of nonexposed operated corneas (Figure
Histological aspect of the cornea in the sham/sham exposed group (a), sham surgery/exposed group (b), surgery/sham exposed group (c), and surgery/exposed group (d).
After six months of exposure, slight modifications of the epithelium were noticed. The epithelial aspect was less regular with limited proliferative areas (Figure
When exposures were performed after surgery (Figure
The spectrum in Figure
13C-NMR spectrum of cornea lipid extract in CDCl3/MeOD. Buckets are represented by black boxes and identified, as used in statistical analyses (a). Histogram representation of Kruskal-Wallis tests, using four classes: surgery/exposed (black box), surgery sham/exposed (strip box), surgery/sham exposed (light grey box), and sham surgery/sham exposed (white box) (b). PLS-DA correlation circle for type two tests: black boxes represent the sham-exposed groups; white circles represent the exposed groups (c).
1H-NMR spectra of aqueous humor (a). A two-class analysis of metabolites contributions to 1H-NMR spectra of aqueous humor, sham-exposed groups (white) versus exposed group as is shown on the bottom trace (b). Meanings of abbreviations: GLU-
Oxidative stress for the 4 different groups measured in aqueous humor and plasma. For each group, the value was the mean of 15 measures ± SD. Significance was determined using the Mann-Whitney
Lenses. PLS-DA correlation circle for type 2 tests: shams, black squares versus exposed, white squares. (a) The PCA correlation circle was built using 8 main components. (b) PLS-DA construction was built from two main components.
(a) 13C-NMR spectrum of lenses lipid extract in CDCl3/MeOD. Cholesterol resonances are labeled with stars, and other resonances are labeled as mentioned in the methods section. (b) Comparative histogram representation of Kruskal-Wallis tests, using four classes: surgery/exposed (black box), surgery sham/exposed (strip box), surgery/sham exposed (light grey box), and sham surgery/sham exposed (white box). A, chain length; B, chain unsaturation; C, chain to glycerol ratios; D, relative contribution of PS to phospholipids; E, relative contribution of PI to phospholipids; F, relative contribution of PC to phospholipids; G, relative contribution of PE to phospholipids; H, relative contribution of PEPL to phospholipids; I, relative contribution of glycidic moieties to total polar head groups; J, cholesterol/phospholipid ratio; K, cholesterol/phospholipids.
Here, also, no significant difference was found between groups taken separately by pairs or gathered in exposed versus sham groups.
(a) 1H-NMR spectra of vitreous humor were lyophilized and resuspended in D2O. The glycidic area is labeled with an arrow; other metabolites are shown with their abbreviations. (b) Comparative histogram of metabolite contributions was extracted from 1H-NMR spectra of vitreous humor, sham-exposed groups (white) versus exposed groups. Meanings of the abbreviations: GLU-
The aim of this work was to investigate whether chronic electromagnetic field exposure of the eyes after keratotomy would influence healing or at least induce metabolic or structural modifications in the eyes. Such an interrogation was supported by several professional situations, such as those for radar technicians and engineers in airports or navy personnel aboard warships. The question here was to determine whether radar exposure under chronic conditions at occupational power levels could lead to deleterious effects regarding cornea healing or metabolic perturbations in the different building blocks of the eyes. That is why the experimental model chosen in this paper used New Zealand rabbits and eye exposures at 50 W/m2, 1 hr/day, 3 days/Week over 5 months, under pulse conditions mimicking radar exposure (duty cycle of 1/1000). The different experiments performed on the different parts of the eyes, cornea observations, analyses, and histology failed to find any significant effect. This was in strong agreement with papers underlining that significantly more intense power is requested to obtain cornea lesions [
The question was more subject to controversy in the case of lenses. On the one hand, there is no doubt today that microwave radiation via thermal effects may lead to cataracts. Hence, the international guidelines for exposure limits to microwave radiations were directly related to protection against thermal effects [
Due to some limitations, extrapolation of that work to the human eye should be carefully considered. The main limitation could be addressed on the surgical operation itself. The experimental constraints do not allow the realization of photoablation in our model. However, ablation of the stromal bed could add great stress caused by the laser energy absorption. Moreover, the rabbit eye presents some differences to the human eye, such as corneal endothelium with replicative capacity, unlike in the human cornea. All of those considerations lead to the potential underestimation of possible effects of human electromagnetic exposure.
This paper dealt with the consequences of chronic electromagnetic exposures at occupational levels on the eyes after keratotomy. From the extensive study of all compartments of the eye using various approaches, including anatomy, histology, and NMR lipidic and metabolic analysis, as well as light diffusion, no dramatic deleterious effect was identified on eye healing or eye structure under these conditions. However, with regard to the suggestion that people occupationally exposed to X-band radars (10 GHz) bear no risk in the workplace when their vision requires corrective keratotomy, further investigations are required to identify or inform the bubbling process observed here. This, for instance, would be based on variations of the power of exposure and duty cycles at very low levels or conversely at levels close to thermal effects. This work, involving numerous animals and long experimental periods, is currently in progress.
The authors declare that there is no conflict of interests regarding the publication of this paper.