In 1949, José Barraquer first proposed the use of alloplastic materials as a method to correct the refractive state of the eye.1 In 1988, Simón and Barraquer presented a method for the surgical correction of myopia by the insertion of intracorneal rings in rabbits.2 Efforts to perfect intracorneal lenses have required extensive consideration of biocompatibility, adequate permeability to nutrients, corneal and lens hydration status, shape and optical quality, as well as the level of intrastromal placement of the lens.3 Such lenses require surgical incursion into the visual axis and, although apparently tolerated by the cornea, have the potential for additional complications.4
The Intrastromal Corneal Ring (ICR®) is an investigational medical device, introducible through a small peripheral corneal incision, that is designed to modify corneal curvature to correct myopia without surgical intervention in the central cornea. The polymethylmethacrylate (PMMA) ring is implanted in the peripheral corneal stroma at approximately two-thirds depth.
Studies in human cadaver eyes have suggested the ICR flattens the central anterior corneal curvature, with increasing thickness causing a more pronounced effect.5 The ring is well tolerated and effective in altering corneal curvature in rabbit models6 as well as in blind human subjects followed for 1 year.7,8
Herein, we describe the 12-month results of ICR implantation into 10 myopic eyes. The objective of this study was to generate data on the safety of the ICR implant, the stability of the correction obtained, and the efficacy of the ICR implanted into sighted eyes.
Figure 1 : The ICR is an investigational device designed to modify corneal curvature for the correction of myopia. Composed of optically transparent polymethylmethacrylate, the ICRs used in this study were 0.30 mm thick with an outer diameter of 7.70 mm.
Figure 2: The vacuum centering guide attaches to the sclera and serves as a guide for the stromal separator, which is rotated 360° to create an intrastromal channel for the ICR in the periphery of the cornea.
SUBJECTS AND METHODS
Ten patients with preoperative refractive errors ranging from -2.63 diopters (D) to -4.25 D (mean, -3.30 D) participated in the study. One eye of each patient received an ICR. The selection criteria required that patients be over the age of 21 and in good health with two normal, functional eyes. The age range of the selected patients was 25 to 43 years old (mean, 30.8 years). The cohort was composed of seven females and three males. The eye selected to receive the ICR was determined by refractive criteria and was typically the nondominant eye. None of the patients had prior ophthalmic surgery, and all were free of diabetes, collagen vascular disease, or inherited metabolic disorders. Ophthalmic conditions contraindicating surgery included central corneal thickness of less than 0.40 mm, lid margin disease, abnormal tear status, lagophthalmos, corneas steeper than 50.00 D or flatter than 40.00 D, and intraocular pressure under 10 mm Hg or over 21 mm Hg. Other contraindications included history of herpetic eye disease in the surgical eye, unstable refractive correction, pregnancy, and systemic disease which would increase the operative risk. Additionally, contact lenses were not worn for a period of time prior to surgery. All patients underwent a comprehensive testing schedule, which included having visual acuity, manifest refraction, keratometry, slit-lamp microscopy, pachymetry, and tonometry measures taken preoperatively and at 12 intervals throughout the 2-year study period. The examinations were performed by personnel other than the surgeon. Standardized forms were used to record all data.
The protocol received approval from the Medical Ethics Committee of the Escola Paulista de Medicina, Säo Paulo Hospital, Säo Paulo, Brazil, on September 30, 1991. ICRs were implanted by the same surgeon (W.N.) in four patients on October 31 and in six patients on December 9, 1991. The ICR (Fig 1) is an optically transparent, split PMMA ring. The ICR configuration used for all 10 patients in this study was 7.70 mm in outer diameter, 6.80 mm in inner diameter, and 0.30 mm thick. This ICR configuration was designed to provide approximately -2.50 D of correction. Each end of the device has a hole to allow for manipulation by surgical instruments and to facilitate approximation of the ends.
All ICR implant procedures were performed under general anesthesia. First, the geometric center of the cornea was located using calipers. Then, using a 2.5-mm radial marker, a corneal incision site was marked in the periphery of the cornea at the 12 o'clock position, approximately 8 mm from the geometric center of the cornea. Corneal pachymetry, which was measured preoperatively at both the center of the cornea and at the intended incision site, was again measured at the incision site. A diamond knife, set to approximately two thirds of the corneal thickness, was used to create the radial incision. Next, the corneal tissue was laterally separated at the base of the incision with a Suarez Spreader (Storz, St. Louis, Mo) to prepare an entry pocket for the stromal separator.
A vacuum centering guide was used to provide fixation for placement of the stromal separator (Fig 2), a blunt-tipped circular blade instrument designed to separate the layers of the stromal lamellae in the corneal periphery to create a channel for the ICR. After applying the vacuum centering guide to the perilimbal conjunctiva, the separator was introduced into the incision and slowly rotated 360° clockwise until the blunt tip reentered the incision. It was then rotated 360° counterclockwise and removed from the cornea.
Viscoat® (Alcon, Fort Worth, Tex) was used to cover the surface of the cornea and to lubricate the ICR prior to insertion. Using forceps, the ring was rotated into the intrastromal channel in a clockwise fashion. The ends of the ring were approximated and secured using a single 10-0 Prolene® (Ethicon, Sommerville, NJ) suture. The suture ends were then trimmed. The ring junction was rotated clockwise with a Sinskey Hook (Storz, St. Louis, Mo) to approximately 1 clock hour away from the radial incision. Typically, two 10-0 nylon interrupted sutures were used to reapproximate the radial entry incision. The average procedure time was 30 minutes.
Immediately following the procedure, all patients received a subconjunctival injection of 20 mg gentamicin sulphate and 3 mg betamethasone. The postoperative medication regime consisted of an antibiotic-corticosteroid combination (neomycinpolymyxin B-dexamethasone) applied four times daily. The routine postoperative medication was gradually tapered and completely discontinued by the 2-month examination for all patients, with the exception of patient no. 6 who had the tear in Descemet's membrane during surgery and continued taking medication.
The variables used to measure the effects of ICR implantation included uncorrected and spectaclecorrected visual acuity, manifest refraction, keratometry, slit-lamp microscopy observations, pachymetry, and intraocular pressure. A Student's t test was used to evaluate the differences between the pre- and postimplant values in order to determine their statistical significance (p<0.05). The 12month results discussed here pertain to the nine remaining implant patients. (The ICR was explanted from one patient after 6 months; see "Ocular Observations/Complications.") Table 1 provides a comparison of each patient's preoperative and postoperative 12-month values for uncorrected visual acuity, spectacle-corrected visual acuity, and manifest refraction. Clinically significant complications for each patient are also summarized in Table 1.
The uncorrected visual acuity (Snellen) results through month 12 are presented in Table 2 and plotted in Figure 3. Preoperative uncorrected visual acuities ranged from 20/200 to 20/400. From postoperative day 1 through month 12, the nine implant patients had uncorrected visual acuities of 20/40 or better, with the exception of patient no. 9 whose 6month Snellen reading was reported as 20/60. This was suspected to be a measurement error as the corresponding ETDRS reading for this patient was 20/25. The patient was retested and his uncorrected visual acuity was 20/20 (Snellen) and 20/25 (ETDRS). At the 12-month examination, the patient was 20/40 (Snellen). At 12 months postoperatively, 100% (9 of 9) of the patients had an uncorrected visual acuity of 20/40 or better with 33% (3 of 9) seeing 20/20 or better (Fig 4). All 10 patients had maintained a spectacle-corrected visual acuity of 20/20 or better throughout the study period, with the exception of patient no. 6 who was 20/30 at the 6-month examination. Her spectacle-corrected visual acuity returned to 20/20 a few days after the ICR was explanted and remained stable through study exit. The mean change in spherical equivalent (SE) manifest refraction achieved with the ICR at month 12 compared to preoperative readings was -2.25 D (range, -1.62 to -3.25 D; SD=0.54 D). The mean change in SE keratometric readings at 12 months compared to the preoperative readings was -2.35 D (range, +1.75 to -5.88 D; SD=2.01 D). The mean changes for keratometric and manifest refraction at each time point when compared to the preoperative measures were significant (keratometry: p<0.01; manifest refraction: p<0.001).
The most frequent positive slit-lamp microscopy findings were for conjunctival congestion, corneal cloudiness, and conjunctival swelling. These observations occurred in all implant patients during the initial 3 postoperative days. After the 1-month examination, there were no further incidences of conjunctival congestion or conjunctival swelling reported. Late findings consisted of superficial and deep wound site vascularization with associated wound site cloudiness.
Central corneal pachymetry increased transiently in three patients between the initial postoperative reading at the day 7 examination and the reading at the 2-month examination. In patient no. 6, the increase was associated with a tear in Descemet's membrane that occurred at the time of surgery, and in patients no. 4 and no. 9, the increases were bilateral, with all measurements returning to preoperative values by month 3. Comparison of mean pachymetry measurements at each point versus preoperative measures revealed no significant increases in central corneal thickness.
Results of a 0.3-mm Intrastromal Corneal Ring at 12 Months
None of the patients experienced an asymmetric increase in intraocular pressure or a pressure reading over 21 mm Hg during the 12-month study period.
Implantation of the ICR did not appear to affect astigmatism significantly. Comparison of mean corneal astigmatism (keratometric measures ) at 12 months to mean preoperative values was not significant (?=0?7). Seven cases had negative changes (or decreases) ranging from -0.25 to -1.25 D. Two cases had increases in corneal astigmatism of -1.00 D each.
OBSERVATIONS AND COMPLICATIONS
Overall, the ICR appeared to be well tolerated in the human corneal stroma. No incision healing complications were observed in any of the patients. The epithelium on the wound site was reepithelialized and completely healed by day 7 in all patients. The postoperative observations that were reported during the 12-month study period included peripheral haze, lamellar channel deposits, infiltrates, and iron Lines. These observations were not considered clinically significant.
Figure 3: Stability of uncorrected visual acuity over time after implantation of a 0.30 mm ICR (n=10*). (*The ICR was removed from patient JS-071065 after 6 months.)
Uncorrected Visual Acuity After Implantation of an Intrastromal Corneal Ring
Figure 4: Uncorrected visual acuity at 12 months after implantation of an ICR in nine eyes.
The ring was explanted from one patient (no. 6) after 6 months as a result of complications related to a posterior corneal perforation that occurred during the surgical procedure. After completion of the stromal channel dissection, a small amount of aqueous humor was noted in the channel and a limited tear in Descemet's membrane was suspected. The depth of the radial incision and the resultant dissection, which was deeper than specified in the protocol, is believed to have contributed to the approximately 1mm long tear. The tear appeared to seal up after a few minutes, so the surgeon proceeded to implant an ICR. Localized corneal edema was noted at the day 1 examination through month 3 on both sides of the ICR at approximately the 10-o'clock position in the area of the tear.
During a 4-month interim examination, the investigators noted that the patient had developed -3.50 D of astigmatism at the 150° axis. The astigmatismwas possibly induced by localized corneal edema, which was visible on slit-lamp microscopy examination and was secondary to the tear in Descemet's membrane. The patient's uncorrected visual acuity had decreased from 20/20, recorded at the 3-month examination, to 20/100. The patient's spectacle-corrected visual acuity remained at 20/20. At the 6month examination, the patient had -3.00 D of astigmatism at the 150° axis, and her uncorrected visual acuity remained at 20/100. The ICR was explanted after 6 months due to induced astigmatism and deterioration of uncorrected visual acuity. The expiant procedure for patient no. 6 was performed under peribulbar local anesthesia. A 2-mm radial incision was made over the ICR ends, the suture securing them was cut, and the ring was rotated in a clockwise fashion out of the wound without difficulty. The procedure took approximately 15 minutes to complete and was free of complications.
This patient has remained stable during the postexplant period with no complications. At the patient's 6-month postexpiant examination, there were no positive slit-lamp microscopy findings and no peripheral haze. The patient's uncorrected visual acuity was 20/100 (her preoperative level had been 20/200) with spectacle-corrected visual acuity at 20/20. The patient's manifest refraction was -3.00-1.75×135°, compared to a preoperative measurement of -2.50-0.25×180°. The patient's corneal curvature appears to be returning to its preoperative state, and there have been no adverse effects or other medically significant sequelae resulting from the ICR removal.
Figure 5: Deviation from intended correction at 12 months. Manifest refraction in nine eyes.* (The ICR was removed from patient no. 6 after the 6-month examination.)
A narrow circular band of peripheral corneal haze was noted in the intrastromal channel for all 10 patients (Fig 6). The peripheral haze is believed to be the result of the separation of the stromal lamellae by the dissecting instrument and is, therefore, not directly associated with the ICR. The peripheral haze is considered clinically inconsequential because it is located in the intrastromal channel, does not approach the central visual axis, and dissipates over time.
Seven patients developed minute lamellar channel deposits (Fig 6). These lipid-like deposits typically appeared as either crystalline and réfringent or as minute, whitish, round droplets. This material was confined to the lamellar channel with no visual consequence, appeared to be inert and nonprogressive, and did not appear to be a cellular infiltrate. These deposits were observed in three patients at the 3 -month examination, in two patients at the 4month examination, in one patient at the 6-month examination, and in one patient at the 9-month examination. No inflammatory response has been observed in conjunction with the lamellar channel deposits.
Sectoral superficial wound site neovascularization (pannus) was seen in five patients. In patient no. 2, this was associated with a loose incision suture at the 1-month examination. The suture was removed with subsequent regression of superficial vessels by the 2-month examination. For patient no. 7, superficial vessels were observed at month 1 and had resolved by month 6. For patients, no. 3, no. 5, and no. 10, pannus was noted at the 9-month examination and was either adjacent to, or superior to, a previously noted deep vessel.
Figure 6: Average change in manifest refraction over time after implantation of a 0.30 mm ICR (n=10*). (The ICR was removed from patient no. 6 after 6 months.)
Sectoral deep corneal neovascularization was noted in five patients. All cases involved a single, nonprogressive vessel. It was related to an infiltrate at the incision site in patients no. 7 and no. 3. In patients no. 1, no. 5, and no. 10, it was associated with the peripheral extent of the incision being close to the limbal vessels. None of the deep vessels were of late onset.
Patient no. 7 exhibited both superficial and deep neovascularization related to an incision site infiltrate noted at the day 7 examination. Gram and Giemsa staining, as well as culturing, were performed to determine if microorganisms could be isolated from the infiltrate. The results from the stains and cultures were negative. After treatment with topical vancomycin and oral chloramphenicol, the infiltrate and pannus slowly regressed. By 6 months, the pannus had resolved.
Patient no. 3 exhibited deep neovascularization related to an infiltrate, which developed at week 2, and the formation of a 2×2-mm corneal ulcer noted after 4 months. The suture's untrimmed ends were directed upwards in the stroma and were most likely responsible for the ulcer. The ICR ends had not been rotated away from the incision site, nor had the knot for the ring closure suture been buried in the suture hole at the time of surgery. Gram-positive bacteria (cocci) were cultured from the ulcer. The patient was treated with topical vancomycin and bacitracin, with gradual tapering of medications. Three weeks later, it was noted that the suture's untrimmed ends had eroded through the epithelial surface. The suture was removed, and healing of the defect occurred over the next few days. The patient's condition has remained stable, with only slight pannus formation at the old infiltrate site.
A superficial, inferior intraepithelial iron line (along the inner circumference of the ICR) was first noted in four patients at the 9-month examination. At the 12-month visit, no progression or clinical significance was observed related to this observation, which has been discussed in greater detail by Assil et al.9
At postoperative day 1, uncorrected visual acuity was 20/40 or better for all patients. This level of uncorrected visual acuity was maintained through the 12-month examination for the nine implant patients, with the exception of the previously mentioned patient whose 9-month reading of 20/60 was deemed to be a measurement error. All patients maintained a spectacle-corrected visual acuity of 20/20 or better during the 12-month study period, with the exception of patient no. 6 who was 20/30 at the 6-month examination. Her spectacle-corrected visual acuity returned to 20/20 a few days after the ICR was explanted and remained stable through study exit.
In this study, the average change in SE manifest refraction from the preoperative measurement was -2.25 D (range, -1.62 D to -3.25 D) at the 12month examination for the nine implant patients. All of these patients were within 1.00 D of their intended correction (Fig 5). In cadaver eyes evaluated by laser holographic interferometer (KeraMetrics Class II corneal analyzer, Kera-Metrics, Ine, Solana Beach, Calif), a 0.31-mm ICR yielded a mean flattening of 3.50 D/> Results of the ICR on corneal flattening in cadaver eyes revealed a nearly linear relationship of flattening to ring thickness.5 This cadaver eye data appears to be in reasonable agreement with this study because keratorefractive procedures routinely result in a broad range of corneal flattening in artificially deturgesced eyes.7 This topography data also compares favorably to a previous study in which the same ICR configuration was implanted into three nonfunctional eyes.8 In one eye, the ICR was removed to access the reversibility of the implant. The 12-month keratometric results for the other two eyes indicated that 2.10 D and 2.90 D of flattening were induced by the ring.
The average SE refractive change achieved at 3 months was -2.15 D (SD=0.44 D), -2.09 D (SD=0.81 D) at 6 months, and -2.25 D (SD=0.54 D) at 12 months. Despite some fluctuation in manifest refraction, five of nine patients with 12-month refractions (SE) did not vary more than 0.50 D between months 3 and 12. The average change in manifest refraction over time is shown in Figure 6. Although a 12-month follow up is not long enough for a final analysis of long-term stability, these results are encouraging. The nine implant patients in this study will continue to be followed, and the clinical study is continuing.
With the exception of patient no. 6 (explanted patient), uncorrected visual acuities in the postoperative period remained at 20/40 or better and were stable throughout the 12-month period for all patients (Fig 3). In patient no. 2, increased keratometric steepening was noted postoperatively, despite a reduction in myopia of 2.25 D. The visual results, over a significant range of preoperative refractive error and despite instances of differences in keratometric and refractive readings, may be related to altered corneal optical parameters. As shown in cadaver eyes, one unique feature of the refractive correction that occurs with the ICR is that the natural, positive asphericity of the cornea is maintained/' The positive asphericity theoretically decreases image aberration, commonly called spherical aberration, and may explain the excellent visual results of our study.10,11
Preoperatively, all patients had a spectacle-corrected visual acuity of 20/20 or better. At 12 months, all patients were still 20/20 or better, although two had lost a line and one had gained a line from their preoperative condition. The stability between the preoperative and 12-month values indicates that the ICR does not appear to affect the patients' spectacle-corrected visual acuity.
A trace amount of peripheral corneal haze along the channel was present in all patients at the 12month examination. The peripheral haze is believed to be the result of the normal stromal healing response associated with the separation of the stromal lamellae by the dissecting instrument. This haze has decreased with time in all patients and does not appear to be clinically significant.
The observed lamellar channel deposits appeared either as crystalline and refringent or as minute, whitish, round droplets, and were first noted between the 3-month and 9-month examinations in seven patients. This material was confined to the lamellar channel and has had no visual consequence for the patients. It appeared to be inert and nonprogressive, did not appear to be a cellular infiltrate, and no inflammatory response was observed. Similar crystalline deposits have been described following intracorneal lenticule implantation in monkeys.12 The authors feel that the lamellar channel deposits are the result of minute lipid deposition by keratocytes and are considered to be part of the stromal healing response.
Figure 7: An external photo taken of patient no. 10 shows peripheral haze approximately 360° in the lamellar channel. Lamellar channel deposits can be seen on the inner edge of the ring from approximately 4 to 7 o'clock.
The finding of a superficial iron line along the inner circumference of the ICR has been described elsewhere with ICR implantation9 and other refractive procedures that alter corneal topography. ms This is felt to be a result of altered tear break-up patterns and senescent basal cell mechanisms. This observation is considered to be a normal occurrence and inconsequential from a medical standpoint.
The presence of deep corneal neovascularization was found in both patients who developed an infiltrate during the early postoperative period and in three others whose corneal incisions approached the limbal area. Incision vascularization is also a common complication of radial keratotomy procedures, occurring when the incision enters the limbus or is placed too close to the limbus.16 The protocol for the ICR surgical procedure has been revised to specify that the incision is not to extend closer than 1 mm from the limbus.
Although limited in scope, this study demonstrates that the ICR can be successfully implanted and its complications to date have been manageable in sighted eyes, therefore demonstrating the feasibility of this technique of refractive corneal surgery (Figs 7, 8, and 9). Most postoperative slit-lamp microscopy findings were transient and without clinical significance. Some of the complications appeared to be related to deviations in the surgical technique. None of the complications led to a decrease in spectacle-corrected visual acuity. Further investigation is merited in order to more fully evaluate the ICR's potential as an approach to modifying anterior corneal curvature for the correction of refractive errors, without surgically invading the center of the cornea. A larger clinical study is needed to draw valid conclusions about the 0.30-mm thick ICR's efficacy and predictability. Indications that ICRs of thicker configurations may correct higher degrees of myopia warrant additional study as do initial impressions about the reversibility of the ring's refractive effect.
Figure 8: An external photo taken of ICR implant. Patient no. 9 shows the cornea to be clear at 12 months postoperatively. Refraction and keratometry results were stable.
Figure 9: At the 12-month slit-lamp microscopy examination, patient no. 9's cornea is clear and refraction and keratometry results were stable. The peripheral haze can also be observed in the slit-lamp microscopy beam.
1. Barraquer JI. Queratoplasia refractiva, estudios e informaciones. Oftalmologicas. 1949;2:10-30.
2. Simón G, Barraquer RI. Queratocricoemfitesis: nuevo procedimiento de cirugia refractiva (estudio experimental). Arch Soc Esp Oftal Invest. 1988;1:87-94.
3. McCarey BE. Alloplastic refractive keratoplasty. In: Sanders DR, ed. Refractive Corneal Surgery. Thorofare, NJ: Slack, Ine; 1986:531-548.
4. Lindstrom RL, Lane SS. Polysulfone intracorneal lenses. In: Sanders DR, ed. Refractive Corneal Surgery. Thorofare, NJ: Slack, Ine; 1986:551-563.
5. Burris TE, Baker PC, Ayer CT, Loomas BE, Mathis ML, Silvestrini TA. Flattening of central corneal curvature with Intrastromal Corneal Rings of increasing thickness: an eyebank eye study. J Cataract Refract Surg. 1993;19(suppl):182-187.
6. Fleming JF, Reynolds AE, Kilmer L, Burris TE, Abbott RL, Schanzlin DJ. The Intrastromal Corneal Ring: two cases in rabbits. J Refract Surg. 1987;3:227-232.
7. Neves R, Nosé W, Belfort R Jr. The Intrastromal Corneal Ring - implantation in blind and myopic eyes. ARVO abstracts. Invest Ophthalmol Vis Sci., 1992;33(4):998.
8. Nosé W, Neves RA, Schanzlin DJ, Belfort R Jr. The Intrastromal Corneal Ring: one-year results of first implants in humans, a preliminary nonfunctional eye study. Refract Corneal Surg. 1993;9:452-458.
9. Assil K, Quantock A, Barrett A, Schanzlin DJ. Corneal iron lines associated with the Intrastromal Corneal Ring. Am J Ophthalmol. 1993;116:350-356.
10. Fleming JF. Should refractive surgeons worry about corneal asphericity? Refract Corneal Surg. 1990;6:455-457.
11. Holladay JT, Lynn MJ, Waring GO III, Gemmili M, Keehn GC, Fielding B. The relationship of visual acuity, refractive error and pupil size after radial keratotomy. Arch Ophthalmol. 1991;109:70-76.
12. Rodrigues MM, McCarey BE, Waring GO, Hidayat AA, Kruth HS. Lipid deposits posterior to impermeable intracorneal lenses in rhesus monkeys: clinical, histochemical, and ultrastructural studies. J Refract Corneal Surg. 1990;6:32-37.
13. Steinberg EB, Wilson LA, Waring GO III, Lynn MJ, Coles WH. Stellate iron lines in the corneal epithelium after radial keratotomy. Am J Ophthalmol. 1984;98:416-421.
14. Mannis MJ. Iron deposition in the corneal graft. Another corneal iron line. Arch Ophthalmol. 1983;101:1858-1861.
15. Keonig SB, McDonald MB, Yamaguchi T, Friedlander M, Ishii Y. Corneal iron lines after refractive keratoplasty. Arch Ophthalmol.. 1983;101:1862-1865.
16. Rowsey J, Balyeat H, Rabinovitch B, Burris TE, Hays J. Complications of radial keratotomy. In: Henkind P, ed. International Congress of Ophthalmology. San Francisco, Calif: JB Lippincott; 1983:1223-1229.
Results of a 0.3-mm Intrastromal Corneal Ring at 12 Months
Uncorrected Visual Acuity After Implantation of an Intrastromal Corneal Ring