- Email: [email protected]
Corneal crosslinking and intracorneal ring segments for keratoconus: A randomized study of concurrent versus sequential surgery
1
ARTICLE
Corneal crosslinking and intracorneal ring segments for keratoconus: A randomized study of concurrent versus sequential surgery Peter S. Hersh, MD, Reda Issa, MD, Steven A. Greenstein, MD
Purpose: To assess outcomes of corneal crosslinking (CXL) and intracorneal ring segments (ICRS) (Intacs) used adjunctively, and then compare the safety and efficacy of concurrent versus sequential surgery. Setting: Cornea and refractive surgery subspecialty practice. Design: Prospective randomized clinical trial. Methods: Patients were randomized to one of two groups: ICRS first, immediately followed by CXL during the same session (n Z 104), or ICRS followed by CXL 3 months later (n Z 94). Outcomes included changes in maximum keratometry (K) and topographic inferior–superior (I–S) difference, maximum flattening of topographic K, and changes in uncorrected (UDVA) and corrected (CDVA) distance visual acuities. These were analyzed in the entire cohort, in the two randomized groups, and in subgroups stratified to ICRS size and placement. Patients were followed for 6 months.
T
here are two unique and salient clinical characteristics of keratoconusddistortion of the corneal optical architecture and progression of the disease over time. Consequently, treatments are aimed bidirectionally to (1) decrease disease progression, and (2) improve the corneal optical contour. The first can be addressed by corneal crosslinking (CXL), a procedure that strengthens the cornea and has been shown to be safe and effective in mitigating the progression of keratoconus and ectasia.1,2 Complementing this biomechanical strengthening effect, intrastromal corneal ring segments (ICRS) (Intacs, Addition Technology, Inc.) can be implanted to achieve the second goal of improving corneal topographic symmetry.3
Results: The study comprised 198 eyes of 198 patients. Overall, maximum K decreased by an average of 2.5 D, I–S difference improved by 3.9 D, and there was an average maximum flattening of 7.5 D. The UDVA improved by 2.0 logarithm of the minimum angle of resolution lines, on average, and the CDVA improved by 1.1 lines. There was no significant difference between the sequential and concurrent groups in any of the outcomes analyzed. There were 6 clinically significant adverse events. Conclusions: CXL and ICRS can be used adjunctively with substantial improvement in corneal topography, and with no increase in safety concerns over each procedure alone. Sequential and concurrent treatment with ICRS and CXL show equivalent outcomes. Both thicker segment size and single segment placement seem to result in greater topographic improvement. J Cataract Refract Surg 2019; -:-–- Q 2019 ASCRS and ESCRS
In this prospective randomized study, we wanted to answer two questions: What are the general clinical outcomes in corneas receiving both CXL and ICRS? Does timing of the procedures matter; that is, do outcomes differ between simultaneous CXL and ICRS surgeries, and consecutive procedures in which ICRS are first performed followed by CXL 3 months later? PATIENTS AND METHODS Patients with keratoconus were enrolled as part of a prospective, randomized, controlled clinical trial (NCT01112072A) performed under a physician-sponsored Investigational New Drug. This study was approved and monitored by an investigational review board, was compliant with the United States Health Insurance Portability and Accountability Act, and it adhered to the tenets
Submitted: April 20, 2018 | Final revision submitted: December 19, 2018 | Accepted: January 16, 2019 From the Cornea and Laser Eye InstitutedHersh Vision Group (Hersh, Issa, Greenstein), CLEI Center for Keratoconus, Teaneck, and the Department of Ophthalmology (Hersh, Issa, Greenstein), RutgersdNew Jersey Medical School, Newark, New Jersey, USA. The investigational products were provided by Addition Technology, Inc., and Peschke Trade GmbH; they played no role in study design, data collection, data analysis, manuscript preparation, or the decision to submit the report for publication. Supported in part by an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness, Inc., New York, New York, USA. Corresponding author: Peter S. Hersh, MD, Cornea and Laser Eye InstitutedHersh Vision Group, CLEI Center for Keratoconus, 300 Frank W. Burr Boulevard, Suite 71, Teaneck, NJ 07666, USA. Email: [email protected]. Q 2019 ASCRS and ESCRS Published by Elsevier Inc.
0886-3350/$ - see frontmatter https://doi.org/10.1016/j.jcrs.2019.01.020
2
CXL AND ICRS
of the Declaration of Helsinki. Informed consent was obtained from each patient before any study procedures were performed. Patients with a history of keratoconus were evaluated for suitability and had the required screening procedures to determine study eligibility. Inclusion criteria were age 21 years or older, corrected distance visual acuity (CDVA) worse than 20/20, central or inferior steepening on a rotating Scheimpflug camera map (Pentacam, OCULUS Optikger€ate GmbH), and Placido-disk topography consistent with keratoconus. Exclusion criteria included corneal thickness less than 350 mm at the thinnest point measured by Scheimpflug analysis in the eye to be treated and/or less than 400 mm at the thinnest point of the anticipated ICRS placement; previous ocular condition that might predispose the eye to complications (eg, herpes simplex, herpes zoster keratitis, recurrent erosion syndrome, corneal melt, corneal dystrophy); clinically significant corneal scarring in the CXL treatment zone; history of chemical injury or delayed epithelial healing; pregnancy or lactation during the course of the study; a known sensitivity to study medications; nystagmus or any other condition that would prevent a steady gaze during the CXL treatment or other diagnostic tests; and a current condition that, in the investigator’s opinion, would interfere with or prolong epithelial healing. Patients were instructed to remove contact lenses at least 1 week before preoperative study testing and stability was confirmed by comparison of topography maps obtained at least 1 week apart. Each patient was randomized to one of two groups: ICRS first, immediately followed by CXL during the same session, or ICRS followed by CXL 3 months later. Randomization was computer generated by a technician independent from the study and, on the procedure day, a sealed envelope was opened by the study coordinator, revealing whether the eye would be in the sequential or the concurrent group. Patients were aware of their randomly assigned group. Safety monitoring throughout the study included observations at appropriate times for subjective complaints, complications, adverse events (AEs), and clinically significant findings on ophthalmic examination. Only randomized eyes were included in the study analysis. Surgical Technique Intracorneal Rings The ICRS size, symmetry, and placement angle were determined by the investigator based on analysis of individual patient topographic and refractive data. Intacs are available in 5 thicknesses (210 mm, 300 mm, 350 mm, 400 mm, and 450 mm) with more anticipated flattening with thicker segments. Typically, a symmetric pair was placed for a central cone (within a 1.0 mm zone), an asymmetric size pair for a paracentral elevation (1.0 mm to 3.0 mm), and a single segment for a more peripheral cone (outside 3.0 mm). ICRS thickness was chosen based on the degree of topographic elevation (as seen on Scheimpflug topography analysis) and distortion (as seen on Placido-disk imagery). In general, the entry incision was made 90 degrees from the axis of maximum keratometry (K) on the topography map to center the segment(s) on the cone. Immediately before surgery, the incision axis was marked at the slitlamp. After the patient received topical anesthetic, a lid speculum was placed. Using a 3.0 mm optical zone marker with gentian violet, the cornea was marked centered on the pupil. A 5.0 mm marker was then used to place another concentric mark around the pupil. The limbus was marked with a gentian violet marking pen. The specifications for intrastromal ring dissection were programmed into the femtosecond laser (Intralase, Abbott Medical Optics, Inc.). Typical settings were for an inner diameter of 6.8 mm and an outer diameter of 7.8 mm, with depth set at 75% to 80% of the thinnest area of the cornea along the anticipated channel as determined both by optical and ultrasound pachymetry, but a least 350 mm deep. After the femtosecond laser track was completed, the ICRS were place by the investigator. For patients randomized to the concurrent group, the CXL procedure was then begun. For Volume - Issue - - 2019
those in the sequential group, antibiotic and corticosteroid drops were administered at the completion of the procedure. Corneal Crosslinking CXL was performed according the methodology described by Wollensak et al.4 The central epithelium was removed either by mechanical debridement or after a 30 second application of 20% ethyl alcohol applied on a round pledget. To mitigate any potential direct effect of CXL on the ICRS, care was taken not to remove the epithelium directly overlying the ICRS. Riboflavin (0.1% in 20% dextran T500 solution, Peschke GmbH) was then administered topically every two minutes for a total of 30 minutes. Following riboflavin administration, riboflavin absorption throughout the corneal stroma and anterior chamber was confirmed by slitlamp evaluation. Ultrasonic pachymetry was performed, and if the cornea was smaller than 400 mm, hypotonic riboflavin (0.1% in sterile water; Peschke GmbH) was administered, 1 drop every 10 seconds for 2-minute sessions, after which ultrasonic pachymetry was performed to confirm that the stroma had swollen to 400 mm or more. This was repeated until adequate corneal thickness had been obtained. The cornea was aligned, and then exposed to ultraviolet-A 365 nm light for 30 minutes at an irradiance of 3.0 mW/cm2 (UV-X system, IROC Innocross). During ultraviolet exposure, riboflavin administration was continued every 2 minutes. Postoperatively, antibiotic and corticosteroid drops were administered, a soft contact lens bandage was placed, and the eye was reexamined by slitlamp evaluation. The contact lens was removed after the epithelial defect had closed, typically 4 to 5 days postoperatively. Antibiotics and corticosteroid drops were continued 4 times daily for 1 week and 2 weeks, respectively. Patient Assessment Patients had complete examinations at baseline, on the day of treatment, and at 1 day, 1 week, and 1, 3, and 6 months after treatment. For patients randomized to the sequential group, this regimen was restarted after the second intervention. Patients were instructed to leave contact lenses out for at least 1 week before examinations. Outcomes were analyzed for all eyes with data at 6 months after CXL. Outcome Measures All eyes were assessed for the prespecified primary outcome of change in maximum K on topography, and secondary outcomes of change in inferior–superior (I–S) topography value, point of maximum flattening, and changes in uncorrected distance visual acuity (UDVA) and CDVA. Corneal Topography The point of maximum K preoperatively and 6 months postoperatively, as well as the point with maximum flattening at 6 months postoperatively, were obtained using a rotating Scheimpflug camera (Pentacam HR, OCULUS Optikger€ate GmbH). The point of maximal flattening was ascertained by analysis of the postoperative–preoperative subtraction map and selecting the point of greatest difference. The I–S value was measured using the anterior sagittal curvature map. To do this, the angle of maximum K was first identified. Then, taking values at a 3.0 mm radius on the topography map at this angle, the inferior topographic K value was obtained and the superior K was taken 180 degrees away. The difference between the superior and inferior K values was noted as the I–S value (Figure 1). Visual Acuity The UDVA and CDVA were measured preoperatively and postoperatively at 6 months after the final surgical procedure. Visual acuity measurements were obtained under controlled lighting conditions using a modified Lighthouse Early Treatment Diabetic Retinopathy Study visual acuity test (2nd edition) with Sloan letters. Patients were tested 4 m from the visual acuity chart. If patients could not read any letters at 4 m, they were tested at 2 m. Visual acuity was recorded and analyzed as the logarithm of the minimum angle of resolution (logMAR) value.
3
CXL AND ICRS
Figure 1. Examples of topography indicators analyzed. Top, Left: Preoperative maximum K Z steepest point on topography axial power map (62.0 D). Top, Right: Postoperative maximum K (58.8D). Change maximum K Z 58.8 62.0 Z 3.2 D. Center, Left: Preperative I–S difference (48.5 40.4 D Z 8.1 D). Center, Right: Postoperative I–S (44.1 40.6 D Z 3.5 D). Change I– S Z 3.5 8.1 Z 4.6 D. Bottom, Left: Preoperative topography. Bottom, Middle: Postoperative topography. Bottom, Right: Difference map showing point of maximum flattening Z 17.1 D (I–S Z inferior–superior; K Z keratometry).
Clinical Safety Analysis All 198 eyes in the study comprised the safety database. Any AEs were noted at each study visit and at any unscheduled visit. In addition, the percentage of eyes that had a loss of 2 or more lines in CDVA and/or a greater than 2.0 D increase in maximum K are reported.
Statistical Analysis The primary outcome measure of this study of CXL/ICRS treatment of keratoconus was to determine whether there is a difference in the mean change from baseline to 6 months in the maximum K readings between the concurrent same day ICRS/ CXL treated group, and the randomized group that initially underwent the ICRS procedure alone followed by CXL 3 months later. A sample size of 70 eyes per treatment group was determined for a significance level of 0.05 and power of 80% to detect a 1.1 D difference between the treatment groups. Results were analyzed for the entire population and per subgroup defined by ICRS size and symmetry (single 350 mm, 400 mm, and 450 mm segments, respectively, asymmetric 450/ 210 mm segments, and symmetric 350/350 mm and 450/450 mm segments). The baseline score for all endpoints was defined as the preoperative measurement closest to the treatment date. For all analyses, only the randomized eyes were included; that is, all analyses comprised 198 eyes of 198 patients and did not include fellow eye outcomes. The primary endpoint was the difference between the sequential and concurrent ICRS/CXL randomized groups for the mean change in maximum K from baseline to month 6. The primary endpoint data were summarized using descriptive statistics, and the differences in mean changes between
the two treatment groups were evaluated using a 2-sample t test to test the following hypothesis: H0 : mcombined
mintacs Z 0 versus HA : mcombined
mintacs O0
where mcombined is the mean difference between the 6-month maximum K reading and the baseline maximum K for the concurrent group, and mintacs is the mean difference between the 6-month maximum K reading and the baseline maximum K for the sequential group. Secondary endpoints of maximum flattening, I–S difference, UDVA, and CDVA were similarly assessed. For all reported AEs, the number of distinct treatmentemergent complications and the number and percentage of participants who experienced the events were summarized by group. No formal statistical analysis was conducted on the AE data.
RESULTS Baseline Characteristics of Study Population
One-hundred ninety-eight eyes of 198 patients were included (104 eyes in the concurrent group and 94 in the sequential group). Table 1 shows the baseline characteristics of each of the subgroups. There were no differences in preoperative topography measures between the randomized groups. However, although there was no difference in preoperative visual acuity in the entire randomized populations, there was a statistically significant difference between randomized groups for both UDVA and CDVA in the single 450 mm subgroup and for CDVA in the 350/350 mm subgroup. Volume - Issue - - 2019
4
CXL AND ICRS
Table 1. Preoperative characteristics of randomized groups. Mean ± SD Segment Size/ Randomized Group All Eyes Concurrent Sequential P value Single 350 mm Concurrent Sequential P value Single 400 mm Concurrent Sequential P value Single 450 mm Concurrent Sequential P value 210/450 mm Concurrent Sequential P value 350/350 mm Concurrent Sequential P value 450/450 mm Concurrent Sequential P value
N
Female Sex (%)
Age (y)
UDVA (logMAR)
CDVA (logMAR)
Kmax (D)
I–S (D)
104 94
28 32
34 G 9 36 G 12
0.98 G 0.32 1.06 G 0.36 .10
0.41 G 0.24 0.46 G 0.32 .19
60.96 G 10.28 61.51 G 10.81 .72
13.12 G 7.28 13.88 G 8.04 .48
8 7
38 43
33 G 10 39 G 13
0.73 G 0.19 0.77 G 0.21 .66
0.28 G 0.10 0.3 G 0.13 .68
56.15 G 10.44 51.94 G 4.73 .33
8.41 G 4.37 7.6 G 4.47 .73
17 11
24 36
34 G 11 30 G 6
0.82 G 0.27 0.69 G 0.26 .23
0.25 G 0.13 0.33 G 0.18 .18
57.05 G 5.80 56.38 G 4.10 .74
9.05 G 4.52 10.73 G 4.67 .35
44 39
16 33
35 G 9 36 G 12
1.03 G 0.35 1.22 G 0.41 !.05*
0.46 G 0.26 0.65 G 0.36 !.05*
63.04 G 12.76 66.75 G 10.87 .16
16.33 G 7.72 17.11 G 9.04 .66
9 11
78 36
34 G 7 37 G 13
1.07 G 0.20 1.08 G 0.21 .87
0.44 G 0.15 0.37 G 0.22 .42
60.48 G 9.09 63.89 G 12.40 .5
12.72 G 7.33 17.08 G 9.03 .26
20 19
35 19
33 G 8 38 G 14
1.00 G 0.29 1.01 G 0.18 .84
0.38 G 0.18 0.26 G 0.16 !.05*
61.17 G 7.55 57.09 G 9.30 .14
13.22 G 6.30 10.86 G 7.71 .30
6 7
25 29
33 G 11 32 G 5
1.23 G 0.25 1.14 G 0.34 .60
0.65 G 0.41 0.43 G 0.28 .27
63.2 G 5.71 58.16 G 8.88 .26
7.6 G 5.60 10.34 G 3.34 .30
CDVA Z corrected distance visual acuity; I–S Z inferior–superior difference; Kmax Z maximum keratometry; logMAR Z logarithm of the minimum angle of resolution; UDVA Z uncorrected distance visual acuity *Statistically significant difference
Corneal Topography
Maximum Keratometry In the total study population, maximum K flattened by 2.45 D. Stratified to individual eyes, the maximum K decreased by 2.00 D or more in 112 eyes (57%), remained within 2 D in 75 eyes (38%), and increased by 2.00 D or more in 11 eyes (6%) (Figure 2). Overall, single 400 mm and 450 mm segment subgroups showed the largest decrease in maximum K (Table 2).
Figure 2. Change in maximum keratometry K in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (K Z keratometry). Volume - Issue - - 2019
In analyzing the randomized cohorts, maximum K flattened by 2.21 D (from 60.96 preoperatively to 58.75 D postoperatively) after concurrent treatment, and by 2.71 D (from 61.51 to 58.79 D) after sequential treatment (Table 2). Within groups, these postoperative improvements from preoperative values were significant. However, there was no statistically significant difference in postoperative change in maximum K between the concurrent and the sequential groups. Similarly, in analyzing the individual segment subgroups, there were no statistically significant differences between the concurrent and sequential groups. Maximum Postoperative Flattening In the total study population, the maximum postoperative flattening was 7.49 D. Regarding individual eyes, the maximum postoperative flattening was 5.0 D or more in 138 eyes (70%) and was up to 5.0 D in 57 eyes (29%); there was no flattening in 3 eyes (2%) (Figure 3). Similar to the change in maximum K, the single 450 mm segment group showed the most flattening. In analyzing the randomized cohorts, the maximum topography flattening from preoperative to postoperative was 7.34 D after concurrent treatment, and it was 7.65 D after sequential treatment (Table 3). This
5
CXL AND ICRS
Table 2. Change in topographic Kmax after corneal crosslinking/intrastromal corneal ring segments.* Mean Kmax (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
3.15 G 3.95 1.56 G 1.04 2.41 .33
3.10 G 1.52 3.88 G 2.51 3.41 .39
2.91 G 3.12 4.01 G 4.00 3.43 .17
450/210 mm (n Z 20) 2.84 G 2.19 2.68 G 2.68 2.76 .89
350/350 mm (n Z 39) C0.82 G 3.49 C0.21 G 3.46 C0.52 .59
450/450 mm (n Z 13) 1.45 G 1.39 2.80 G 2.11 2.18 .24
Overall (n Z 198) 2.21 G 3.32 2.71 G 3.72 2.45 .32
Kmax Z maximum keratometry *Intacs, Addition Technology, Inc.
difference was not statistically significant. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups. Inferior–Superior Difference In the total study population, the I–S difference decreased by 3.88 D, on average. Regarding individual eyes, the I–S difference decreased by 5.00 D or more in 51 eyes (26%), decreased up to 5.0 D in 112 eyes (57%), remained the same or increased up to 5.0 D in 32 eyes (16%), and increased by more than 5.0 D in 3 eyes (2%) (Figure 4). The single 450 mm segment group showed the greatest improvement in I–S. In analyzing the randomized cohorts, the I–S difference decreased, on average, by 29% (mean Z 3.77 D, from 13.12 preoperatively to 9.35 D postoperatively) after concurrent treatment, and by 29% (mean Z 4.01 D, from 13.88 to 9.87 D) after sequential treatment (Table 4). Within groups, these postoperative improvements were statistically significant. However, there was no statistically significant difference in change over time in I–S between the concurrent and the sequential groups. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups. Visual Acuity
Uncorrected Distance Visual Acuity In the total study population, the UDVA improved by an average of 2.0 logMAR lines. Overall, the UDVA improved by 2 lines or more in 113 eyes (57%), remained within 1 line of change in 67 eyes (34%), and worsened by 2 lines or more in 18 eyes
(9%) (Figure 5). Between the individual segment size groups, the degree of UDVA improvement was somewhat more robust in the 450 mm groups. In analyzing the randomized cohorts, the UDVA improved by 0.17 logMAR in the concurrent group and by 0.24 logMAR in the sequential group (Table 5). Within groups, these preoperative to postoperative improvements were statistically significant. However, there was no statistically significant difference in improvement in UDVA between the concurrent and the sequential groups. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups when stratified to segment size. Corrected Distance Visual Acuity In the total study population, the CDVA improved by 1.1 logMAR lines. Overall, the CDVA improved by 2 lines or more in 66 eyes (33%), remained within 1 line of change in 118 eyes (60%), and worsened by 2 lines or more in 14 eyes (7%) (Figure 6). As for UDVA between individual segment size groups, the degree of CDVA improvement was best in the 450 mm groups. Of the 14 eyes with loss of CDVA, 2 had concomitant loss of UDVA (4 lines and 2 lines, respectively), whereas 10 had improved UDVA, one had an increase in maximum K (C5.5 D), and 6 had worsening of I–S difference (range C0.4 to C6.0 D). There were no AEs in any of these eyes. In analyzing the randomized cohorts, the CDVA improved by 0.09 logMAR in the concurrent group, and by 0.14 in the sequential group (Table 6). Within groups, these postoperative improvements were found to be statistically significant. However, there was no statistically significant difference in improvement in CDVA between the concurrent and the sequential groups. In analyzing the individual segment groups, there was a statistically significant difference in the single 350 mm segment subgroup; in these eyes, the sequential group improved by 1.3 logMAR lines, whereas the concurrent group remained unchanged. However, this subgroup contained only 15 eyes. Figure 7 shows the outcomes of CXL/ICRS in the entire population and randomized subgroups. Postoperative Events
Figure 3. Maximum topographic flattening in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.).
Table 7 shows adverse events. There were two cases of infectious keratitis. The first was in a 28-year-old man who underwent concurrent CXL/ICRS. On postoperative day Volume - Issue - - 2019
6
CXL AND ICRS
Table 3. Maximum topographic flattening after corneal crosslinking/intrastromal corneal ring segments.* Mean Maximum Flattening (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
6.48 G 5,87 4.44 G 1.40 5.53 .40
7.84 G 2.98 6.39 G 2.15 7.27 .19
8.26 G 4.43 9.98 G 4.66 9.07 .09
450/210 mm (n Z 20) 7.18 G 2.08 8.08 G 2.83 7.68 .46
350/350 mm (n Z 39) 5.43 G 2.03 4.85 G 2.48 5.15 .44
450/450 mm (n Z 13) 6.97 G 1.18 6.73 G 4.18 6.84 .90
Overall (n Z 198) 7.34 G 3.85 7.65 G 4.25 7.49 .60
*Intacs, Addition Technology, Inc.
3, an infiltrate with overlying epithelial defect was noted. Cultures grew methicillin-resistant Staphylococcus aureus. Treatment with vancomycin and ciprofloxacin drops was instituted, with resolution of the infection. The patient was left with residual corneal haze. At 3 months postoperatively, the UDVA was 20/63 (20/63 preoperative) and the CDVA was 20/40 (20/40 preoperative). The second case was a 27-year-old patient who underwent sequential surgery. At 1 week after CXL, a 2.0 mm infiltrate with overlying epithelial defect was noted. Cultures grew penicillium species. Treatment with amphotericin and natamycin drops as well as oral itraconazole led to resolution of the infection 2 weeks later. The patient was left with a residual 2.0 mm corneal scar. At 3 months postoperatively, the UDVA was 10/125 (preoperative 5/200) and the CDVA was 20/32 (preoperative 20/63). There were 4 ICRS-related AEs. Most of these were related to inflammation around the segments. One patient had an overlying corneal melt. Both this patient and two others underwent removal of the Intacs segment; all had resolution of the adverse reactions. DISCUSSION Keratoconus is a disease with two notable clinical characteristicsdbiomechanical weakening of the optically functional corneal dome leads to (1) corneal distortion and decreased optical quality with consequent vision loss and (2) disease progression over time. Similarly, the two clinical interventions studied herein are directed, respectively, at mitigating these two salient features of keratoconus. Given the dual
Figure 4. Change in I–S topography power difference in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.). Negative numbers connote improvement (I–S Z inferior–superior). Volume - Issue - - 2019
nature of keratoconus characteristics and treatment goals, our study was designed, first, to determine whether the two procedures could be used adjunctively with success and second, to determine whether timingdsequential versus concurrent surgerydwas important to the ultimate clinical outcome. With riboflavin–ultraviolet-A (365 nm) crosslinking, intrastromal molecular crosslinks militate strengthening of the corneal biomechanical architecture.4 CXL has been shown to be safe and effective in decreasing progression of both keratoconus and corneal ectasia after previous refractive surgery.1,2 In addition, these studies demonstrated a modest improvement, on average, in corneal topography maximum K: 1.6 D in keratoconus and 0.7 D in ectasia. Regarding individual eye distribution, approximately 30% of eyes with keratoconus improve cone steepness by 2.0 D or more with CXL alone. In addition, we have also found modest, yet significant, improvements in corneal topography indices,5 higher-order aberrations,6 and subjective visual function7 after CXL. Notwithstanding these encouraging improvements, the essential goal of CXL is to decrease disease progression, not improve corneal optics or visual function. ICRS (Intacs) are used, specifically, to improve corneal topography and corneal optical symmetry in keratoconus.3 The procedure is not designed to mitigate progression of corneal ectatic disorders, although some studies have suggested long-term success of the procedure alone.8 The exact biomechanical mechanism of ring segments is unclear; clinically, the cornea flattens substantially in the hemimeridian of segment placement. Thus, placement of symmetric ring segments leads to a generalized flattening of the cornea and placement of a single segment typically results in flattening in the meridian of placement and steepening 180 degrees away. Given these two unique goals and mechanisms of action, in this study, we demonstrated the general outcomes in eyes receiving both CXL and Intacs, evaluated the effect of different ring segment sizes and symmetries, and analyzed timing of the proceduresdsimultaneous CXL and ICRS surgeries compared with consecutive procedures in which ICRS are first performed followed by CXL 3 months later. Previous studies have looked at the adjunctive use of CXL and ICRS. However, to our knowledge, only one study to date directly compared concurrent versus sequential ICRS and CXL: El-Raggal9 compared sequential versus
7
CXL AND ICRS
Table 4. Change in I–S topography power difference after corneal crosslinking/intrastromal corneal ring segments.* Mean I–S (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
2.81 G 4.80 5.46 G 5.41 4.05 .37
3.52 G 3.57 5.63 G 2.76 4.35 .12
5.71 G 7.05 5.31 G 4.64 5.52 .76
450/210 mm (n Z 20) 3.33 G 1.95 3.87 G 2.05 3.63 .58
350/350 mm (n Z 39) 0.99 G 2.89 0.61 G 1.71 0.80 .62
450/450 mm (n Z 13) 1.42 G 2.08 2.30 G 1.83 1.89 .47
Overall (n Z 198) 3.77 G 5.53 4.01 G 4.13 3.88 .72
I–S Z inferior–superior *Intacs, Addition Technology, Inc.
concurrent Keraring (Mediphacos Ltda.) insertion and CXL. They concluded that a same-session procedure was more effective in flattening the corneal topography; however, this study included only 16 eyes. Coskunseven et al.10 compared the effect of treatment sequencedCXL followed by ICRS versus ICRS followed by CXLdand found greater improvement in the latter group. Yildirim et al.11 followed 16 eyes for a mean of 43 months to evaluate the long-term effect of combined same-day ICRS/CXL in patients with post-laser in situ keratomileusis ectasia. In this study, there was a statistically significant improvement in visual acuity, refraction, and K values. In a retrospective analysis, Cakir et al.12 compared ICRS alone versus ICRS/CXL, and they found no statistically significant difference in outcomes between these two treatment groups. In contrast, a retrospective study by Legare et al.13 found both combined CXL/ ICRS and ICRS alone were safe and effective, but the ICRS alone group had better refractive spherical equivalent, mean and steepest K improvement, and improved higherorder aberrations after treatment. Renesto et al.14 compared CXL followed by ICRS 3 months later versus riboflavin eyedrops followed by ICRS 3 months later, and found no difference between the groups, concluding that CXL does not seem to augment the effect of ICRS before its insertion. Finally, Elsaftawy et al.15 performed a randomized prospective clinical trial comparing patients treated with Keraring implantation alone, and patients treated with Keraring insertion followed by transepithelial CXL 1 month later. They found a statistically significant improvement in both groups regarding visual acuity and K values, and
Figure 5. Change in UDVA in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (UDVA Z uncorrected distance visual acuity).
significantly more improvement in spherical refraction in patients who received the combined treatment. Analyzing outcomes in the entire treatment population (n Z 198) without regard to the randomization group demonstrated the general clinical efficacy of combined ICRS/CXL. The most substantial change was found with respect to the point of greatest flattening; overall, there was an average improvement of 7.5 D, and 70% of eyes had flattening of 5.0 D or more. Although not directly comparable because of different selection criteria between the subgroups stratified to the Intacs segment size and placement, the single 450 mm segment group showed the greatest flattening ( 9.0 D) followed by the 450/210 mm, single 400 mm, symmetric 450 mm, single 350 mm, and symmetric 350 mm groups, respectively. Thus, both single segment placement as well as thicker segments seem to militate a greater flattening effect. In addition to flattening the keratoconic cone, perhaps the primary purpose of ICRS is to improve corneal optical symmetry. By thus improving the corneal optics, corneal aberrations are diminished with the clinical goal to improve spectacle and contact lens tolerance and diminish visual function disabilities such as monocular diplopia and halos. As a topography index of corneal symmetry, we analyzed the I–S power difference. Overall, there was more than a 3.9 D (29%) improvement in I–S difference; 82% of eyes showed an improvement in I–S. In general, the single segments group led to the most improvement. This result was as expected because a single segment tends to flatten the hemimeridian of segment placement while steepening the opposite meridian, whereas opposing segments would tend to mitigate this effect. Thus, single segment Intacs placement might be more beneficial and clinically recommended in asymmetric corneas; that is, corneas with decentered rather than central cones. The aforementioned results suggest a substantial and clinically meaningful topography change after IRCS and CXL. However, studies of CXL alone have analyzed, typically, a primary outcome of change in maximum K. This is because the maximum K is an objective proxy for the severity of the keratoconic cone itself; thus, changes in maximum K have been used to determine disease progression in most clinical trials. In the study reported herein, there was, on average, a 2.5 D improvement in maximum K. Again, improvement was greatest in the single 450 mm and 400 mm subgroups (w 4.0 D); the two symmetric Volume - Issue - - 2019
8
CXL AND ICRS
Table 5. Change in UDVA after corneal crosslinking/intrastromal corneal ring segments.* Mean UDVA (logMAR) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
0.18 G 0.14 0.17 G 0.33 0.17 0.98
0.16 G 0.27 0.20 G 0.19 0.18 .67
0.18 G 0.30 0.28 G 0.39 0.24 .20
450/210 mm (n Z 20)
350/350 mm (n Z 39)
0.21 G 0.11 0.30 G 0.37 0.26 .49
0.17 G 0.22 0.19 G 0.23 0.16 .92
450/450 mm (n Z 13) 0.10 G 0.24 0.24 G 0.22 0.18 .33
Overall (n Z 198) 0.17 G 0.27 0.24 G 0.33 0.20 .11
logMAR Z logarithm of the minimum angle of resolution; UDVA Z uncorrected distance visual acuity *Intacs, Addition Technology, Inc.
segment subgroups demonstrated the least robust responses. Although not directly comparable because of study entry criteria, it is interesting to assess these results compared with results that we previously published using CXL alone.16 In that study, maximum K decreased by an average of 1.7 D (compared with 2.5 D in current study of CXL/ICRS). The maximum K decreased by 2.0 D or more in 31% (57% in current study) and increased by 2.0 D or more in 4% (6% in the current study). Thus, these findings along with the documented improvements in I–S and maximum flattening support the use of ICRS as an adjunctive procedure to CXL to substantially enhance the beneficial topography effects. For the total study population, the UDVA improved by an average of 2.0 lines. The UDVA improved by 2 lines or more in 57% the eyes and worsened by 2 lines or more in 9% the eyes. This compares with our study of CXL alone,16 in which there was an average 1.1-line improvement; 25% had improved and 9% had worsened UDVA by 2 or more lines. Thus, the UDVA results also support a dual CXL/ICRS approach to the treatment of keratoconus. With regard to the important safety as well as efficacy outcome of CDVA, we found, on average, an improvement of 1.1 lines; CDVA improved by 2 lines or more in 33% and worsened by 2 lines or more in 7%. Again, this compares with our study of CXL alone,16 in which there was an average 1.2-line improvement; 21% had improved and 1% had worsened CDVA by 2 or more lines, and also to the more extensive U.S. multicenter study of crosslinking for keratoconus,1 in which 28% and 6% gained and lost 2 or
Figure 6. Change in CDVA in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (CDVA Z corrected distance visual acuity). Volume - Issue - - 2019
more lines, respectively. Again, these studies are not directly comparable, but the CDVA outcomes of CXL/ICRS appear, generally, to be similar to CXL alone. Further studies are required to better understand why the improvement in topography seen after Intacs segment placement does not seem to dependably translate to a more substantial improvement in CDVA. For those patients in the current study with a decrease in CDVA, we could find no consistent associated findings. Given the general congruence of CDVA outcomes in this study with previous studies of CXL alone, the results might simply represent the inherent variability of outcomes in crosslinking for keratoconus. However, it might be attributed, in part, to the ring segments effect upon corneal topography; for example, 43% of the 14 eyes losing 2 or more CDVA lines had worsening of I–S difference compared with 14% of the overall population. Of these 6 eyes, 5 (83%) had placement of 2 symmetric Intacs segments. Furthermore, 7 (50%) of the total 14 eyes that lost 2 or more lines had placement of 2 segments. Although these numbers are too small to make definitive conclusions, these findings, taken with the generally more substantial topography effects of single segment placement that we report, might suggest an advantage to single segment rather than symmetric segment placement in many cases. Further research is necessary to elucidate the best solution for an individual patient. Eyes were randomized to either concurrent surgery or sequential surgery (ICRS first followed by CXL 3 months later). In our analyses of topography outcomesdmaximum K change, maximum flattening, I–S differencedthere were no significant differences between the two treatment groups, in general, nor between randomized subgroups stratified by Intacs size. Similarly, we found no difference in UDVA or CDVA outcomes between the two randomized groups nor between nor between randomized subgroups stratified by Intacs size (Figure 7). Thus, the timing of CXL/ICRS surgery does not seem to affect the ultimate efficacy and safety outcomes accomplished by the two procedures. In this study, we found 6 clinically meaningful AEsdtwo infectious keratitis and 4 segment-related complications (Table 7). One case of infection was in the concurrent group and one in the sequential group; in the latter, the infection occurred after the CXL procedure. Neither infection involved the Intacs segment and, thus, seem attributable
9
CXL AND ICRS
Table 6. Change in CDVA after corneal crosslinking/intrastromal corneal ring segments.* Mean CDVA (logMAR) ± SD Randomization Group
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
Concurrent Sequential All Eyes P value
C0.01 G 0.11 0.13 G 0.10 0.05 .03†
0.04 G 0.16 0.17 G 0.18 0.09 .06
0.12 G 0.24 0.22 G 0.25 0.16 .08
450/210 mm (n Z 20) 0.13 G 0.11 0.03 G 0.18 0.08 .16
350/350 mm (n Z 39) 0.06 G 0.23 0.07 G 0.11 0.06 .89
450/450 mm (n Z 13) 0.18 G 0.44 0.04 G 0.22 0.11 .51
Overall (n Z 198) 0.09 G 0.23 0.14 G 0.22 0.11 .12
CDVA Z corrected distance visual acuity; logMAR Z logarithm of the minimum angle of resolution *Intacs, Addition Technology, Inc. † Statistically significant difference
to the crosslinking component of the procedure. This infection rate of 0.8% compares to the U.S. multicenter study of CXL for keratoconus1 in which there was one case of infectious keratitis (0.3%) reported in 293 eyes. Regarding Intacs-related AEs, we found 4 (2.0%), with 3 requiring segment removal. Three of these ICRS events occurred after sequential surgery and 1 after concurrent surgery; in the former group, one occurred after ICRS alone and two after the CXL procedure. Procedure timing per se, therefore, does not seem to affect the likelihood of a medical complication. Does CXL increase the incidence of ICRS-related complications? In regard to the removal incidence of 1.5% reported in this study, in recently published work we reviewed 572 consecutive cases of Intacs and found that the incidence of removal for a medical complication was approximately 2.5%.17 Thus, the medical complication rate in the current study of combined CXL/ICRS does not appear to be higher than the complication rate for Intacs alone. In reviewing the complication spectrum reported in this study, it can be inferred with reasonable confidence that the two procedures, when performed on the same cornea either concurrently or sequentially, do not increase patient risk over the individual procedures alone. Limitations to this study include the small sample sizes for subgroup analysis of Intacs segment size. Therefore, although our sample sized is adequate to assess, in general, both the efficacy and safety of the ICRS/CXL procedures as well as the effect of procedure timing, we cannot extrapolate the results to the different Intacs size subgroups with statistical certitude. Further research is necessary to elucidate more fully the effect of size, symmetry, and positioning
on the outcomes of ICRS.18 Moreover, preoperative patient characteristics can influence the outcomes of CXL. In previous work analyzing characteristics influencing outcomes of CXL alone,19 we found that the only independent indicator of CDVA change after surgery was the preoperative CDVA. Eyes with poorer vision tended to have more improvement after CXL; specifically, eyes with CDVA of 20/40 or worse were 5.9 times more likely to improve 2 lines or more. Similarly, the only independent indicator of maximum K change after surgery was the preoperative maximum K. Eyes with steeper maximum K tended to have more flattening after CXL; specifically, eyes with a maximum K of 55 D or more were 5.4 times more likely to have topographic flattening of 2.0 D or more. Thus, this should be considered, in particular, when assessing subgroups such as the single 450 mm group in which there were differences in some baseline characteristics. In addition to Intacs segment thickness, there are other preoperative characteristics that might influence postoperative outcomes. Previous works by another group20–22 reported a grading system for keratoconus and ectasia, and improved ICRS outcomes in patients with worse preoperative disease; thus, such grading schemes to better define the level of preoperative disease might further enable nomogram development and prediction of postoperative outcomes. In conclusion, this study shows (1) CXL and ICRS (Intacs) can be used adjunctively, (2) ICRS substantially improves corneal topography over CXL alone, (3) ICRS size and placement patterns are important in determining the topography outcome, (4) there are no obvious increases in safety concerns of CXL/ICRS compared with each Figure 7. Average change in outcomes in total population and randomized subgroups corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (CDVA Z corrected distance visual acuity; I–S Z difference between inferior and superior topographic anterior sagittal topographic power; logMAR Z logarithm of the minimum angle of resolution; Max K Z point of maximum topographic keratometry; UDVA Z uncorrected distance visual acuity).
Volume - Issue - - 2019
10
CXL AND ICRS
Table 7. Complications after CXL/ICRS.* Complication
N (%)
Infectious keratitis
2 (1.0)
Inflammation around ICRS
3 (1.5)
Glare symptoms
1 (0.5)
Randomized Group
Surgical Intervention
1 Concurrent, 1 Sequential 2 Sequential after CXL, 1 Sequential after ICRS alone Concurrent
None 2 ICRS removal
ICRS removal
CXL Z corneal crosslinking; ICRS Z intrastromal corneal ring segments *Intacs, Addition Technology, Inc.
procedure alone, and (5) efficacy and safety outcomes are similar notwithstanding timing of the two proceduresd concurrent versus 3-month sequential surgery.
WHAT WAS KNOWN Corneal crosslinking (CXL) decreases the progression of keratoconus. CXL causes a small average improvement in corneal topography. Intracorneal ring segments (ICRS) can flatten the cone in keratoconus.
WHAT THIS PAPER ADDS CXL and ICRS (Intacs), used adjunctively, substantially improved corneal topography steepness and symmetry. Complications with CXL/ICRS did not increase over each procedure alone. Thicker segment size and single segment placement resulted in greater topographic improvement. Sequential and concurrent surgery with ICRS and CXL had equivalent outcomes.
REFERENCES 1. Hersh PS, Stulting RD, Muller D, Durrie DS, Rajpal RK, United States Crosslinking Study Group. United States multicenter clinical trial of corneal collagen crosslinking for keratoconus treatment. Ophthalmology 2017; 124:1259–1270 2. Hersh PS, Stulting RD, Muller D, Durrie DS, Rajpal RK. United States Crosslinking Study Group. U.S. multicenter clinical trial of corneal collagen crosslinking for treatment of corneal ectasia after refractive surgery. Ophthalmology 2017; 124:1475–1484 3. Colin J, Cochener B, Savary G, Malet F. Correcting keratoconus with intracorneal rings. J Cataract Refract Surg 2000; 26:1117–1122 4. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol 2003; 135:620–627 5. Greenstein SA, Fry KL, Hersh PS. Corneal topography indices after corneal collagen crosslinking for keratoconus and corneal ectasia: one year results. J Cataract Refract Surg 2011; 37:1282–1290 6. Greenstein SA, Fry KL, Hersh MJ, Hersh PS. Higher-order aberrations after corneal collagen crosslinking for keratoconus and corneal ectasia. J Cataract Refract Surg 2012; 38:292–302 7. Brooks NO, Greenstein SA, Fry KL, Hersh PS. Patient subjective visual function after corneal collagen crosslinking for keratoconus and corneal ectasia. J Cataract Refract Surg 2012; 38:615–619
Volume - Issue - - 2019
8. Torquetti L, Ferrara G, Almeida F, Cunha L, Araujo LPN, Machado AP, Lyra JM, Merayo-Lloves J, Ferra P. Intrastromal corneal ring segment implantation in patients with keratoconus: 10-year follow-up. J Cataract Refract Surg 2014; 30:22–26 9. El-Raggal TM. Sequential versus concurrent Kerarings insertion and corneal collagen cross-linking for keratoconus. Br J Ophthalmol 2011; 95:37–41 10. Coskunseven E, Jankov MR, Hafezi F, Atun S, Arslan E, Kymionis GD. Effect of treatment sequence in combined intrastromal corneal rings and corneal collagen crosslinking for keratoconus. J Cataract Refract Surg 2009; 35:2084–2091 11. Yildirim A, Uslu H, Kara N, Cakir H, Gurler B, Colak HN, Ozgurhan EB. Same-day intrastromal corneal ring segment and collagen cross-linking for ectasia after laser in situ keratomileusis: long-term results. Am J Ophthalmol 2014; 157:1070–1076 12. Cakir H, Pekel G, Perente I, Genc¸ S. Comparison of intrastromal corneal ring segment implantation only and in combination with collagen crosslinking for keratoconus. Eur J Ophthalmol 2013; 23:629–634 13. Legare ME, Iovieno A, Yeung SN, Lichtinger A, Kim P, Hollands S, Slomovic AR, Rootman DS. Intacs with or without same-day corneal collagen cross-linking to treat corneal ectasia. Can J Ophthalmol 2013; 48:173–178 14. Renesto Ada C, Melo LA Jr, Sartori Mde F, Campos M. Sequential topical riboflavin with or without ultraviolet a radiation with delayed intracorneal ring segment insertion for keratoconus. Am J Ophthalmol 2012; 153:982–993 15. Elsaftawy HS, Ahmed MH, Saif MY, Mousa R. Sequential intracorneal ring segment implantation and corneal transepithelial collagen cross-linking in keratoconus. Cornea 2015; 34:1420–1426 16. Hersh PS, Greenstein SA, Fry KL. Corneal collagen crosslinking for keratoconus and corneal ectasia: One year results. J Cataract Refract Surg 2011; 37:149–160 17. Nguyen N, Gelles JD, Greenstein SA, Hersh PS. Incidence and associations of intracorneal ring segment explantation. J Cataract Refract Surg 2019; 45:153–158 18. Chan CCK, Sharma M. Boxer Wachler BS. Effect of inferior-segment Intacs with and without C3-R on keratoconus. J Cataract Refract Surg 2007; 33:75–80 19. Greenstein SA, Hersh PS. Characteristics influencing outcomes of corneal collagen crosslinking for keratoconus and ectasia: Implications for patient selection. J Cataract Refract Surg 2013; 39:1133–1140 JL, Pin ~ero DP, Aleso n A, Teus MA, Barraquer RI, Murta J, 20. Alio rrez R, Villa C, UcedaMaldonado MJ, Castro de Luna G, Gutie Montanes A. Keratoconus-integrated characterization considering anterior corneal aberrations, internal astigmatism, and corneal biomechanics. J Cataract Refract Surg 2011; 37:552–568 JL, Vega-Estrada A, Baviera J, Beltran J, Cobo-Soriano R. 21. Brenner LF, Alio Clinical grading of post-LASIK ectasia related to visual limitation and predictive factors for vision loss. J Cataract Refract Surg 2012; 38:1817–1826 JL, Brenner LF, Javaloy J, Plaza Puche AB, 22. Vega-Estrada A, Alio Barraquer RI, Teus MA, Murta J, Henriques J, Uceda-Montanes A. Outcome analysis of intracorneal ring segments for the treatment of keratoconus based on visual, refractive, and aberrometric impairment. Am J Ophthalmol 2013; 155:575–584 OTHER CITED MATERIAL A. U.S. National Institutes of Health Clinical Trials. Corneal Collagen Crosslinking and Intacs for Keratoconus and Ectasia (CXL). NCT01112017. Available at: https://clinicaltrials.gov/ct2/show/NCT01112072. Accessed February 27, 2019
Disclosures: Dr. Hersh receives personal fees from Addition Technology, Inc. and Avedro, Inc. outside the submitted work. None of the other authors has a financial or proprietary interest in any material or methods mentioned.
First author: Peter S. Hersh, MD Hersh Vision Group, CLEI Center for Keratoconus, Teaneck, New Jersey, USA
ARTICLE
Corneal crosslinking and intracorneal ring segments for keratoconus: A randomized study of concurrent versus sequential surgery Peter S. Hersh, MD, Reda Issa, MD, Steven A. Greenstein, MD
Purpose: To assess outcomes of corneal crosslinking (CXL) and intracorneal ring segments (ICRS) (Intacs) used adjunctively, and then compare the safety and efficacy of concurrent versus sequential surgery. Setting: Cornea and refractive surgery subspecialty practice. Design: Prospective randomized clinical trial. Methods: Patients were randomized to one of two groups: ICRS first, immediately followed by CXL during the same session (n Z 104), or ICRS followed by CXL 3 months later (n Z 94). Outcomes included changes in maximum keratometry (K) and topographic inferior–superior (I–S) difference, maximum flattening of topographic K, and changes in uncorrected (UDVA) and corrected (CDVA) distance visual acuities. These were analyzed in the entire cohort, in the two randomized groups, and in subgroups stratified to ICRS size and placement. Patients were followed for 6 months.
T
here are two unique and salient clinical characteristics of keratoconusddistortion of the corneal optical architecture and progression of the disease over time. Consequently, treatments are aimed bidirectionally to (1) decrease disease progression, and (2) improve the corneal optical contour. The first can be addressed by corneal crosslinking (CXL), a procedure that strengthens the cornea and has been shown to be safe and effective in mitigating the progression of keratoconus and ectasia.1,2 Complementing this biomechanical strengthening effect, intrastromal corneal ring segments (ICRS) (Intacs, Addition Technology, Inc.) can be implanted to achieve the second goal of improving corneal topographic symmetry.3
Results: The study comprised 198 eyes of 198 patients. Overall, maximum K decreased by an average of 2.5 D, I–S difference improved by 3.9 D, and there was an average maximum flattening of 7.5 D. The UDVA improved by 2.0 logarithm of the minimum angle of resolution lines, on average, and the CDVA improved by 1.1 lines. There was no significant difference between the sequential and concurrent groups in any of the outcomes analyzed. There were 6 clinically significant adverse events. Conclusions: CXL and ICRS can be used adjunctively with substantial improvement in corneal topography, and with no increase in safety concerns over each procedure alone. Sequential and concurrent treatment with ICRS and CXL show equivalent outcomes. Both thicker segment size and single segment placement seem to result in greater topographic improvement. J Cataract Refract Surg 2019; -:-–- Q 2019 ASCRS and ESCRS
In this prospective randomized study, we wanted to answer two questions: What are the general clinical outcomes in corneas receiving both CXL and ICRS? Does timing of the procedures matter; that is, do outcomes differ between simultaneous CXL and ICRS surgeries, and consecutive procedures in which ICRS are first performed followed by CXL 3 months later? PATIENTS AND METHODS Patients with keratoconus were enrolled as part of a prospective, randomized, controlled clinical trial (NCT01112072A) performed under a physician-sponsored Investigational New Drug. This study was approved and monitored by an investigational review board, was compliant with the United States Health Insurance Portability and Accountability Act, and it adhered to the tenets
Submitted: April 20, 2018 | Final revision submitted: December 19, 2018 | Accepted: January 16, 2019 From the Cornea and Laser Eye InstitutedHersh Vision Group (Hersh, Issa, Greenstein), CLEI Center for Keratoconus, Teaneck, and the Department of Ophthalmology (Hersh, Issa, Greenstein), RutgersdNew Jersey Medical School, Newark, New Jersey, USA. The investigational products were provided by Addition Technology, Inc., and Peschke Trade GmbH; they played no role in study design, data collection, data analysis, manuscript preparation, or the decision to submit the report for publication. Supported in part by an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness, Inc., New York, New York, USA. Corresponding author: Peter S. Hersh, MD, Cornea and Laser Eye InstitutedHersh Vision Group, CLEI Center for Keratoconus, 300 Frank W. Burr Boulevard, Suite 71, Teaneck, NJ 07666, USA. Email: [email protected]. Q 2019 ASCRS and ESCRS Published by Elsevier Inc.
0886-3350/$ - see frontmatter https://doi.org/10.1016/j.jcrs.2019.01.020
2
CXL AND ICRS
of the Declaration of Helsinki. Informed consent was obtained from each patient before any study procedures were performed. Patients with a history of keratoconus were evaluated for suitability and had the required screening procedures to determine study eligibility. Inclusion criteria were age 21 years or older, corrected distance visual acuity (CDVA) worse than 20/20, central or inferior steepening on a rotating Scheimpflug camera map (Pentacam, OCULUS Optikger€ate GmbH), and Placido-disk topography consistent with keratoconus. Exclusion criteria included corneal thickness less than 350 mm at the thinnest point measured by Scheimpflug analysis in the eye to be treated and/or less than 400 mm at the thinnest point of the anticipated ICRS placement; previous ocular condition that might predispose the eye to complications (eg, herpes simplex, herpes zoster keratitis, recurrent erosion syndrome, corneal melt, corneal dystrophy); clinically significant corneal scarring in the CXL treatment zone; history of chemical injury or delayed epithelial healing; pregnancy or lactation during the course of the study; a known sensitivity to study medications; nystagmus or any other condition that would prevent a steady gaze during the CXL treatment or other diagnostic tests; and a current condition that, in the investigator’s opinion, would interfere with or prolong epithelial healing. Patients were instructed to remove contact lenses at least 1 week before preoperative study testing and stability was confirmed by comparison of topography maps obtained at least 1 week apart. Each patient was randomized to one of two groups: ICRS first, immediately followed by CXL during the same session, or ICRS followed by CXL 3 months later. Randomization was computer generated by a technician independent from the study and, on the procedure day, a sealed envelope was opened by the study coordinator, revealing whether the eye would be in the sequential or the concurrent group. Patients were aware of their randomly assigned group. Safety monitoring throughout the study included observations at appropriate times for subjective complaints, complications, adverse events (AEs), and clinically significant findings on ophthalmic examination. Only randomized eyes were included in the study analysis. Surgical Technique Intracorneal Rings The ICRS size, symmetry, and placement angle were determined by the investigator based on analysis of individual patient topographic and refractive data. Intacs are available in 5 thicknesses (210 mm, 300 mm, 350 mm, 400 mm, and 450 mm) with more anticipated flattening with thicker segments. Typically, a symmetric pair was placed for a central cone (within a 1.0 mm zone), an asymmetric size pair for a paracentral elevation (1.0 mm to 3.0 mm), and a single segment for a more peripheral cone (outside 3.0 mm). ICRS thickness was chosen based on the degree of topographic elevation (as seen on Scheimpflug topography analysis) and distortion (as seen on Placido-disk imagery). In general, the entry incision was made 90 degrees from the axis of maximum keratometry (K) on the topography map to center the segment(s) on the cone. Immediately before surgery, the incision axis was marked at the slitlamp. After the patient received topical anesthetic, a lid speculum was placed. Using a 3.0 mm optical zone marker with gentian violet, the cornea was marked centered on the pupil. A 5.0 mm marker was then used to place another concentric mark around the pupil. The limbus was marked with a gentian violet marking pen. The specifications for intrastromal ring dissection were programmed into the femtosecond laser (Intralase, Abbott Medical Optics, Inc.). Typical settings were for an inner diameter of 6.8 mm and an outer diameter of 7.8 mm, with depth set at 75% to 80% of the thinnest area of the cornea along the anticipated channel as determined both by optical and ultrasound pachymetry, but a least 350 mm deep. After the femtosecond laser track was completed, the ICRS were place by the investigator. For patients randomized to the concurrent group, the CXL procedure was then begun. For Volume - Issue - - 2019
those in the sequential group, antibiotic and corticosteroid drops were administered at the completion of the procedure. Corneal Crosslinking CXL was performed according the methodology described by Wollensak et al.4 The central epithelium was removed either by mechanical debridement or after a 30 second application of 20% ethyl alcohol applied on a round pledget. To mitigate any potential direct effect of CXL on the ICRS, care was taken not to remove the epithelium directly overlying the ICRS. Riboflavin (0.1% in 20% dextran T500 solution, Peschke GmbH) was then administered topically every two minutes for a total of 30 minutes. Following riboflavin administration, riboflavin absorption throughout the corneal stroma and anterior chamber was confirmed by slitlamp evaluation. Ultrasonic pachymetry was performed, and if the cornea was smaller than 400 mm, hypotonic riboflavin (0.1% in sterile water; Peschke GmbH) was administered, 1 drop every 10 seconds for 2-minute sessions, after which ultrasonic pachymetry was performed to confirm that the stroma had swollen to 400 mm or more. This was repeated until adequate corneal thickness had been obtained. The cornea was aligned, and then exposed to ultraviolet-A 365 nm light for 30 minutes at an irradiance of 3.0 mW/cm2 (UV-X system, IROC Innocross). During ultraviolet exposure, riboflavin administration was continued every 2 minutes. Postoperatively, antibiotic and corticosteroid drops were administered, a soft contact lens bandage was placed, and the eye was reexamined by slitlamp evaluation. The contact lens was removed after the epithelial defect had closed, typically 4 to 5 days postoperatively. Antibiotics and corticosteroid drops were continued 4 times daily for 1 week and 2 weeks, respectively. Patient Assessment Patients had complete examinations at baseline, on the day of treatment, and at 1 day, 1 week, and 1, 3, and 6 months after treatment. For patients randomized to the sequential group, this regimen was restarted after the second intervention. Patients were instructed to leave contact lenses out for at least 1 week before examinations. Outcomes were analyzed for all eyes with data at 6 months after CXL. Outcome Measures All eyes were assessed for the prespecified primary outcome of change in maximum K on topography, and secondary outcomes of change in inferior–superior (I–S) topography value, point of maximum flattening, and changes in uncorrected distance visual acuity (UDVA) and CDVA. Corneal Topography The point of maximum K preoperatively and 6 months postoperatively, as well as the point with maximum flattening at 6 months postoperatively, were obtained using a rotating Scheimpflug camera (Pentacam HR, OCULUS Optikger€ate GmbH). The point of maximal flattening was ascertained by analysis of the postoperative–preoperative subtraction map and selecting the point of greatest difference. The I–S value was measured using the anterior sagittal curvature map. To do this, the angle of maximum K was first identified. Then, taking values at a 3.0 mm radius on the topography map at this angle, the inferior topographic K value was obtained and the superior K was taken 180 degrees away. The difference between the superior and inferior K values was noted as the I–S value (Figure 1). Visual Acuity The UDVA and CDVA were measured preoperatively and postoperatively at 6 months after the final surgical procedure. Visual acuity measurements were obtained under controlled lighting conditions using a modified Lighthouse Early Treatment Diabetic Retinopathy Study visual acuity test (2nd edition) with Sloan letters. Patients were tested 4 m from the visual acuity chart. If patients could not read any letters at 4 m, they were tested at 2 m. Visual acuity was recorded and analyzed as the logarithm of the minimum angle of resolution (logMAR) value.
3
CXL AND ICRS
Figure 1. Examples of topography indicators analyzed. Top, Left: Preoperative maximum K Z steepest point on topography axial power map (62.0 D). Top, Right: Postoperative maximum K (58.8D). Change maximum K Z 58.8 62.0 Z 3.2 D. Center, Left: Preperative I–S difference (48.5 40.4 D Z 8.1 D). Center, Right: Postoperative I–S (44.1 40.6 D Z 3.5 D). Change I– S Z 3.5 8.1 Z 4.6 D. Bottom, Left: Preoperative topography. Bottom, Middle: Postoperative topography. Bottom, Right: Difference map showing point of maximum flattening Z 17.1 D (I–S Z inferior–superior; K Z keratometry).
Clinical Safety Analysis All 198 eyes in the study comprised the safety database. Any AEs were noted at each study visit and at any unscheduled visit. In addition, the percentage of eyes that had a loss of 2 or more lines in CDVA and/or a greater than 2.0 D increase in maximum K are reported.
Statistical Analysis The primary outcome measure of this study of CXL/ICRS treatment of keratoconus was to determine whether there is a difference in the mean change from baseline to 6 months in the maximum K readings between the concurrent same day ICRS/ CXL treated group, and the randomized group that initially underwent the ICRS procedure alone followed by CXL 3 months later. A sample size of 70 eyes per treatment group was determined for a significance level of 0.05 and power of 80% to detect a 1.1 D difference between the treatment groups. Results were analyzed for the entire population and per subgroup defined by ICRS size and symmetry (single 350 mm, 400 mm, and 450 mm segments, respectively, asymmetric 450/ 210 mm segments, and symmetric 350/350 mm and 450/450 mm segments). The baseline score for all endpoints was defined as the preoperative measurement closest to the treatment date. For all analyses, only the randomized eyes were included; that is, all analyses comprised 198 eyes of 198 patients and did not include fellow eye outcomes. The primary endpoint was the difference between the sequential and concurrent ICRS/CXL randomized groups for the mean change in maximum K from baseline to month 6. The primary endpoint data were summarized using descriptive statistics, and the differences in mean changes between
the two treatment groups were evaluated using a 2-sample t test to test the following hypothesis: H0 : mcombined
mintacs Z 0 versus HA : mcombined
mintacs O0
where mcombined is the mean difference between the 6-month maximum K reading and the baseline maximum K for the concurrent group, and mintacs is the mean difference between the 6-month maximum K reading and the baseline maximum K for the sequential group. Secondary endpoints of maximum flattening, I–S difference, UDVA, and CDVA were similarly assessed. For all reported AEs, the number of distinct treatmentemergent complications and the number and percentage of participants who experienced the events were summarized by group. No formal statistical analysis was conducted on the AE data.
RESULTS Baseline Characteristics of Study Population
One-hundred ninety-eight eyes of 198 patients were included (104 eyes in the concurrent group and 94 in the sequential group). Table 1 shows the baseline characteristics of each of the subgroups. There were no differences in preoperative topography measures between the randomized groups. However, although there was no difference in preoperative visual acuity in the entire randomized populations, there was a statistically significant difference between randomized groups for both UDVA and CDVA in the single 450 mm subgroup and for CDVA in the 350/350 mm subgroup. Volume - Issue - - 2019
4
CXL AND ICRS
Table 1. Preoperative characteristics of randomized groups. Mean ± SD Segment Size/ Randomized Group All Eyes Concurrent Sequential P value Single 350 mm Concurrent Sequential P value Single 400 mm Concurrent Sequential P value Single 450 mm Concurrent Sequential P value 210/450 mm Concurrent Sequential P value 350/350 mm Concurrent Sequential P value 450/450 mm Concurrent Sequential P value
N
Female Sex (%)
Age (y)
UDVA (logMAR)
CDVA (logMAR)
Kmax (D)
I–S (D)
104 94
28 32
34 G 9 36 G 12
0.98 G 0.32 1.06 G 0.36 .10
0.41 G 0.24 0.46 G 0.32 .19
60.96 G 10.28 61.51 G 10.81 .72
13.12 G 7.28 13.88 G 8.04 .48
8 7
38 43
33 G 10 39 G 13
0.73 G 0.19 0.77 G 0.21 .66
0.28 G 0.10 0.3 G 0.13 .68
56.15 G 10.44 51.94 G 4.73 .33
8.41 G 4.37 7.6 G 4.47 .73
17 11
24 36
34 G 11 30 G 6
0.82 G 0.27 0.69 G 0.26 .23
0.25 G 0.13 0.33 G 0.18 .18
57.05 G 5.80 56.38 G 4.10 .74
9.05 G 4.52 10.73 G 4.67 .35
44 39
16 33
35 G 9 36 G 12
1.03 G 0.35 1.22 G 0.41 !.05*
0.46 G 0.26 0.65 G 0.36 !.05*
63.04 G 12.76 66.75 G 10.87 .16
16.33 G 7.72 17.11 G 9.04 .66
9 11
78 36
34 G 7 37 G 13
1.07 G 0.20 1.08 G 0.21 .87
0.44 G 0.15 0.37 G 0.22 .42
60.48 G 9.09 63.89 G 12.40 .5
12.72 G 7.33 17.08 G 9.03 .26
20 19
35 19
33 G 8 38 G 14
1.00 G 0.29 1.01 G 0.18 .84
0.38 G 0.18 0.26 G 0.16 !.05*
61.17 G 7.55 57.09 G 9.30 .14
13.22 G 6.30 10.86 G 7.71 .30
6 7
25 29
33 G 11 32 G 5
1.23 G 0.25 1.14 G 0.34 .60
0.65 G 0.41 0.43 G 0.28 .27
63.2 G 5.71 58.16 G 8.88 .26
7.6 G 5.60 10.34 G 3.34 .30
CDVA Z corrected distance visual acuity; I–S Z inferior–superior difference; Kmax Z maximum keratometry; logMAR Z logarithm of the minimum angle of resolution; UDVA Z uncorrected distance visual acuity *Statistically significant difference
Corneal Topography
Maximum Keratometry In the total study population, maximum K flattened by 2.45 D. Stratified to individual eyes, the maximum K decreased by 2.00 D or more in 112 eyes (57%), remained within 2 D in 75 eyes (38%), and increased by 2.00 D or more in 11 eyes (6%) (Figure 2). Overall, single 400 mm and 450 mm segment subgroups showed the largest decrease in maximum K (Table 2).
Figure 2. Change in maximum keratometry K in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (K Z keratometry). Volume - Issue - - 2019
In analyzing the randomized cohorts, maximum K flattened by 2.21 D (from 60.96 preoperatively to 58.75 D postoperatively) after concurrent treatment, and by 2.71 D (from 61.51 to 58.79 D) after sequential treatment (Table 2). Within groups, these postoperative improvements from preoperative values were significant. However, there was no statistically significant difference in postoperative change in maximum K between the concurrent and the sequential groups. Similarly, in analyzing the individual segment subgroups, there were no statistically significant differences between the concurrent and sequential groups. Maximum Postoperative Flattening In the total study population, the maximum postoperative flattening was 7.49 D. Regarding individual eyes, the maximum postoperative flattening was 5.0 D or more in 138 eyes (70%) and was up to 5.0 D in 57 eyes (29%); there was no flattening in 3 eyes (2%) (Figure 3). Similar to the change in maximum K, the single 450 mm segment group showed the most flattening. In analyzing the randomized cohorts, the maximum topography flattening from preoperative to postoperative was 7.34 D after concurrent treatment, and it was 7.65 D after sequential treatment (Table 3). This
5
CXL AND ICRS
Table 2. Change in topographic Kmax after corneal crosslinking/intrastromal corneal ring segments.* Mean Kmax (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
3.15 G 3.95 1.56 G 1.04 2.41 .33
3.10 G 1.52 3.88 G 2.51 3.41 .39
2.91 G 3.12 4.01 G 4.00 3.43 .17
450/210 mm (n Z 20) 2.84 G 2.19 2.68 G 2.68 2.76 .89
350/350 mm (n Z 39) C0.82 G 3.49 C0.21 G 3.46 C0.52 .59
450/450 mm (n Z 13) 1.45 G 1.39 2.80 G 2.11 2.18 .24
Overall (n Z 198) 2.21 G 3.32 2.71 G 3.72 2.45 .32
Kmax Z maximum keratometry *Intacs, Addition Technology, Inc.
difference was not statistically significant. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups. Inferior–Superior Difference In the total study population, the I–S difference decreased by 3.88 D, on average. Regarding individual eyes, the I–S difference decreased by 5.00 D or more in 51 eyes (26%), decreased up to 5.0 D in 112 eyes (57%), remained the same or increased up to 5.0 D in 32 eyes (16%), and increased by more than 5.0 D in 3 eyes (2%) (Figure 4). The single 450 mm segment group showed the greatest improvement in I–S. In analyzing the randomized cohorts, the I–S difference decreased, on average, by 29% (mean Z 3.77 D, from 13.12 preoperatively to 9.35 D postoperatively) after concurrent treatment, and by 29% (mean Z 4.01 D, from 13.88 to 9.87 D) after sequential treatment (Table 4). Within groups, these postoperative improvements were statistically significant. However, there was no statistically significant difference in change over time in I–S between the concurrent and the sequential groups. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups. Visual Acuity
Uncorrected Distance Visual Acuity In the total study population, the UDVA improved by an average of 2.0 logMAR lines. Overall, the UDVA improved by 2 lines or more in 113 eyes (57%), remained within 1 line of change in 67 eyes (34%), and worsened by 2 lines or more in 18 eyes
(9%) (Figure 5). Between the individual segment size groups, the degree of UDVA improvement was somewhat more robust in the 450 mm groups. In analyzing the randomized cohorts, the UDVA improved by 0.17 logMAR in the concurrent group and by 0.24 logMAR in the sequential group (Table 5). Within groups, these preoperative to postoperative improvements were statistically significant. However, there was no statistically significant difference in improvement in UDVA between the concurrent and the sequential groups. Similarly, in analyzing the individual segment groups, there were no significant differences between the concurrent and sequential groups when stratified to segment size. Corrected Distance Visual Acuity In the total study population, the CDVA improved by 1.1 logMAR lines. Overall, the CDVA improved by 2 lines or more in 66 eyes (33%), remained within 1 line of change in 118 eyes (60%), and worsened by 2 lines or more in 14 eyes (7%) (Figure 6). As for UDVA between individual segment size groups, the degree of CDVA improvement was best in the 450 mm groups. Of the 14 eyes with loss of CDVA, 2 had concomitant loss of UDVA (4 lines and 2 lines, respectively), whereas 10 had improved UDVA, one had an increase in maximum K (C5.5 D), and 6 had worsening of I–S difference (range C0.4 to C6.0 D). There were no AEs in any of these eyes. In analyzing the randomized cohorts, the CDVA improved by 0.09 logMAR in the concurrent group, and by 0.14 in the sequential group (Table 6). Within groups, these postoperative improvements were found to be statistically significant. However, there was no statistically significant difference in improvement in CDVA between the concurrent and the sequential groups. In analyzing the individual segment groups, there was a statistically significant difference in the single 350 mm segment subgroup; in these eyes, the sequential group improved by 1.3 logMAR lines, whereas the concurrent group remained unchanged. However, this subgroup contained only 15 eyes. Figure 7 shows the outcomes of CXL/ICRS in the entire population and randomized subgroups. Postoperative Events
Figure 3. Maximum topographic flattening in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.).
Table 7 shows adverse events. There were two cases of infectious keratitis. The first was in a 28-year-old man who underwent concurrent CXL/ICRS. On postoperative day Volume - Issue - - 2019
6
CXL AND ICRS
Table 3. Maximum topographic flattening after corneal crosslinking/intrastromal corneal ring segments.* Mean Maximum Flattening (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
6.48 G 5,87 4.44 G 1.40 5.53 .40
7.84 G 2.98 6.39 G 2.15 7.27 .19
8.26 G 4.43 9.98 G 4.66 9.07 .09
450/210 mm (n Z 20) 7.18 G 2.08 8.08 G 2.83 7.68 .46
350/350 mm (n Z 39) 5.43 G 2.03 4.85 G 2.48 5.15 .44
450/450 mm (n Z 13) 6.97 G 1.18 6.73 G 4.18 6.84 .90
Overall (n Z 198) 7.34 G 3.85 7.65 G 4.25 7.49 .60
*Intacs, Addition Technology, Inc.
3, an infiltrate with overlying epithelial defect was noted. Cultures grew methicillin-resistant Staphylococcus aureus. Treatment with vancomycin and ciprofloxacin drops was instituted, with resolution of the infection. The patient was left with residual corneal haze. At 3 months postoperatively, the UDVA was 20/63 (20/63 preoperative) and the CDVA was 20/40 (20/40 preoperative). The second case was a 27-year-old patient who underwent sequential surgery. At 1 week after CXL, a 2.0 mm infiltrate with overlying epithelial defect was noted. Cultures grew penicillium species. Treatment with amphotericin and natamycin drops as well as oral itraconazole led to resolution of the infection 2 weeks later. The patient was left with a residual 2.0 mm corneal scar. At 3 months postoperatively, the UDVA was 10/125 (preoperative 5/200) and the CDVA was 20/32 (preoperative 20/63). There were 4 ICRS-related AEs. Most of these were related to inflammation around the segments. One patient had an overlying corneal melt. Both this patient and two others underwent removal of the Intacs segment; all had resolution of the adverse reactions. DISCUSSION Keratoconus is a disease with two notable clinical characteristicsdbiomechanical weakening of the optically functional corneal dome leads to (1) corneal distortion and decreased optical quality with consequent vision loss and (2) disease progression over time. Similarly, the two clinical interventions studied herein are directed, respectively, at mitigating these two salient features of keratoconus. Given the dual
Figure 4. Change in I–S topography power difference in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.). Negative numbers connote improvement (I–S Z inferior–superior). Volume - Issue - - 2019
nature of keratoconus characteristics and treatment goals, our study was designed, first, to determine whether the two procedures could be used adjunctively with success and second, to determine whether timingdsequential versus concurrent surgerydwas important to the ultimate clinical outcome. With riboflavin–ultraviolet-A (365 nm) crosslinking, intrastromal molecular crosslinks militate strengthening of the corneal biomechanical architecture.4 CXL has been shown to be safe and effective in decreasing progression of both keratoconus and corneal ectasia after previous refractive surgery.1,2 In addition, these studies demonstrated a modest improvement, on average, in corneal topography maximum K: 1.6 D in keratoconus and 0.7 D in ectasia. Regarding individual eye distribution, approximately 30% of eyes with keratoconus improve cone steepness by 2.0 D or more with CXL alone. In addition, we have also found modest, yet significant, improvements in corneal topography indices,5 higher-order aberrations,6 and subjective visual function7 after CXL. Notwithstanding these encouraging improvements, the essential goal of CXL is to decrease disease progression, not improve corneal optics or visual function. ICRS (Intacs) are used, specifically, to improve corneal topography and corneal optical symmetry in keratoconus.3 The procedure is not designed to mitigate progression of corneal ectatic disorders, although some studies have suggested long-term success of the procedure alone.8 The exact biomechanical mechanism of ring segments is unclear; clinically, the cornea flattens substantially in the hemimeridian of segment placement. Thus, placement of symmetric ring segments leads to a generalized flattening of the cornea and placement of a single segment typically results in flattening in the meridian of placement and steepening 180 degrees away. Given these two unique goals and mechanisms of action, in this study, we demonstrated the general outcomes in eyes receiving both CXL and Intacs, evaluated the effect of different ring segment sizes and symmetries, and analyzed timing of the proceduresdsimultaneous CXL and ICRS surgeries compared with consecutive procedures in which ICRS are first performed followed by CXL 3 months later. Previous studies have looked at the adjunctive use of CXL and ICRS. However, to our knowledge, only one study to date directly compared concurrent versus sequential ICRS and CXL: El-Raggal9 compared sequential versus
7
CXL AND ICRS
Table 4. Change in I–S topography power difference after corneal crosslinking/intrastromal corneal ring segments.* Mean I–S (D) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
2.81 G 4.80 5.46 G 5.41 4.05 .37
3.52 G 3.57 5.63 G 2.76 4.35 .12
5.71 G 7.05 5.31 G 4.64 5.52 .76
450/210 mm (n Z 20) 3.33 G 1.95 3.87 G 2.05 3.63 .58
350/350 mm (n Z 39) 0.99 G 2.89 0.61 G 1.71 0.80 .62
450/450 mm (n Z 13) 1.42 G 2.08 2.30 G 1.83 1.89 .47
Overall (n Z 198) 3.77 G 5.53 4.01 G 4.13 3.88 .72
I–S Z inferior–superior *Intacs, Addition Technology, Inc.
concurrent Keraring (Mediphacos Ltda.) insertion and CXL. They concluded that a same-session procedure was more effective in flattening the corneal topography; however, this study included only 16 eyes. Coskunseven et al.10 compared the effect of treatment sequencedCXL followed by ICRS versus ICRS followed by CXLdand found greater improvement in the latter group. Yildirim et al.11 followed 16 eyes for a mean of 43 months to evaluate the long-term effect of combined same-day ICRS/CXL in patients with post-laser in situ keratomileusis ectasia. In this study, there was a statistically significant improvement in visual acuity, refraction, and K values. In a retrospective analysis, Cakir et al.12 compared ICRS alone versus ICRS/CXL, and they found no statistically significant difference in outcomes between these two treatment groups. In contrast, a retrospective study by Legare et al.13 found both combined CXL/ ICRS and ICRS alone were safe and effective, but the ICRS alone group had better refractive spherical equivalent, mean and steepest K improvement, and improved higherorder aberrations after treatment. Renesto et al.14 compared CXL followed by ICRS 3 months later versus riboflavin eyedrops followed by ICRS 3 months later, and found no difference between the groups, concluding that CXL does not seem to augment the effect of ICRS before its insertion. Finally, Elsaftawy et al.15 performed a randomized prospective clinical trial comparing patients treated with Keraring implantation alone, and patients treated with Keraring insertion followed by transepithelial CXL 1 month later. They found a statistically significant improvement in both groups regarding visual acuity and K values, and
Figure 5. Change in UDVA in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (UDVA Z uncorrected distance visual acuity).
significantly more improvement in spherical refraction in patients who received the combined treatment. Analyzing outcomes in the entire treatment population (n Z 198) without regard to the randomization group demonstrated the general clinical efficacy of combined ICRS/CXL. The most substantial change was found with respect to the point of greatest flattening; overall, there was an average improvement of 7.5 D, and 70% of eyes had flattening of 5.0 D or more. Although not directly comparable because of different selection criteria between the subgroups stratified to the Intacs segment size and placement, the single 450 mm segment group showed the greatest flattening ( 9.0 D) followed by the 450/210 mm, single 400 mm, symmetric 450 mm, single 350 mm, and symmetric 350 mm groups, respectively. Thus, both single segment placement as well as thicker segments seem to militate a greater flattening effect. In addition to flattening the keratoconic cone, perhaps the primary purpose of ICRS is to improve corneal optical symmetry. By thus improving the corneal optics, corneal aberrations are diminished with the clinical goal to improve spectacle and contact lens tolerance and diminish visual function disabilities such as monocular diplopia and halos. As a topography index of corneal symmetry, we analyzed the I–S power difference. Overall, there was more than a 3.9 D (29%) improvement in I–S difference; 82% of eyes showed an improvement in I–S. In general, the single segments group led to the most improvement. This result was as expected because a single segment tends to flatten the hemimeridian of segment placement while steepening the opposite meridian, whereas opposing segments would tend to mitigate this effect. Thus, single segment Intacs placement might be more beneficial and clinically recommended in asymmetric corneas; that is, corneas with decentered rather than central cones. The aforementioned results suggest a substantial and clinically meaningful topography change after IRCS and CXL. However, studies of CXL alone have analyzed, typically, a primary outcome of change in maximum K. This is because the maximum K is an objective proxy for the severity of the keratoconic cone itself; thus, changes in maximum K have been used to determine disease progression in most clinical trials. In the study reported herein, there was, on average, a 2.5 D improvement in maximum K. Again, improvement was greatest in the single 450 mm and 400 mm subgroups (w 4.0 D); the two symmetric Volume - Issue - - 2019
8
CXL AND ICRS
Table 5. Change in UDVA after corneal crosslinking/intrastromal corneal ring segments.* Mean UDVA (logMAR) ± SD Randomization Group Concurrent Sequential All Eyes P value
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
0.18 G 0.14 0.17 G 0.33 0.17 0.98
0.16 G 0.27 0.20 G 0.19 0.18 .67
0.18 G 0.30 0.28 G 0.39 0.24 .20
450/210 mm (n Z 20)
350/350 mm (n Z 39)
0.21 G 0.11 0.30 G 0.37 0.26 .49
0.17 G 0.22 0.19 G 0.23 0.16 .92
450/450 mm (n Z 13) 0.10 G 0.24 0.24 G 0.22 0.18 .33
Overall (n Z 198) 0.17 G 0.27 0.24 G 0.33 0.20 .11
logMAR Z logarithm of the minimum angle of resolution; UDVA Z uncorrected distance visual acuity *Intacs, Addition Technology, Inc.
segment subgroups demonstrated the least robust responses. Although not directly comparable because of study entry criteria, it is interesting to assess these results compared with results that we previously published using CXL alone.16 In that study, maximum K decreased by an average of 1.7 D (compared with 2.5 D in current study of CXL/ICRS). The maximum K decreased by 2.0 D or more in 31% (57% in current study) and increased by 2.0 D or more in 4% (6% in the current study). Thus, these findings along with the documented improvements in I–S and maximum flattening support the use of ICRS as an adjunctive procedure to CXL to substantially enhance the beneficial topography effects. For the total study population, the UDVA improved by an average of 2.0 lines. The UDVA improved by 2 lines or more in 57% the eyes and worsened by 2 lines or more in 9% the eyes. This compares with our study of CXL alone,16 in which there was an average 1.1-line improvement; 25% had improved and 9% had worsened UDVA by 2 or more lines. Thus, the UDVA results also support a dual CXL/ICRS approach to the treatment of keratoconus. With regard to the important safety as well as efficacy outcome of CDVA, we found, on average, an improvement of 1.1 lines; CDVA improved by 2 lines or more in 33% and worsened by 2 lines or more in 7%. Again, this compares with our study of CXL alone,16 in which there was an average 1.2-line improvement; 21% had improved and 1% had worsened CDVA by 2 or more lines, and also to the more extensive U.S. multicenter study of crosslinking for keratoconus,1 in which 28% and 6% gained and lost 2 or
Figure 6. Change in CDVA in individual eyes after corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (CDVA Z corrected distance visual acuity). Volume - Issue - - 2019
more lines, respectively. Again, these studies are not directly comparable, but the CDVA outcomes of CXL/ICRS appear, generally, to be similar to CXL alone. Further studies are required to better understand why the improvement in topography seen after Intacs segment placement does not seem to dependably translate to a more substantial improvement in CDVA. For those patients in the current study with a decrease in CDVA, we could find no consistent associated findings. Given the general congruence of CDVA outcomes in this study with previous studies of CXL alone, the results might simply represent the inherent variability of outcomes in crosslinking for keratoconus. However, it might be attributed, in part, to the ring segments effect upon corneal topography; for example, 43% of the 14 eyes losing 2 or more CDVA lines had worsening of I–S difference compared with 14% of the overall population. Of these 6 eyes, 5 (83%) had placement of 2 symmetric Intacs segments. Furthermore, 7 (50%) of the total 14 eyes that lost 2 or more lines had placement of 2 segments. Although these numbers are too small to make definitive conclusions, these findings, taken with the generally more substantial topography effects of single segment placement that we report, might suggest an advantage to single segment rather than symmetric segment placement in many cases. Further research is necessary to elucidate the best solution for an individual patient. Eyes were randomized to either concurrent surgery or sequential surgery (ICRS first followed by CXL 3 months later). In our analyses of topography outcomesdmaximum K change, maximum flattening, I–S differencedthere were no significant differences between the two treatment groups, in general, nor between randomized subgroups stratified by Intacs size. Similarly, we found no difference in UDVA or CDVA outcomes between the two randomized groups nor between nor between randomized subgroups stratified by Intacs size (Figure 7). Thus, the timing of CXL/ICRS surgery does not seem to affect the ultimate efficacy and safety outcomes accomplished by the two procedures. In this study, we found 6 clinically meaningful AEsdtwo infectious keratitis and 4 segment-related complications (Table 7). One case of infection was in the concurrent group and one in the sequential group; in the latter, the infection occurred after the CXL procedure. Neither infection involved the Intacs segment and, thus, seem attributable
9
CXL AND ICRS
Table 6. Change in CDVA after corneal crosslinking/intrastromal corneal ring segments.* Mean CDVA (logMAR) ± SD Randomization Group
Single 350 mm (n Z 15)
Single 400 mm (n Z 28)
Single 450 mm (n Z 83)
Concurrent Sequential All Eyes P value
C0.01 G 0.11 0.13 G 0.10 0.05 .03†
0.04 G 0.16 0.17 G 0.18 0.09 .06
0.12 G 0.24 0.22 G 0.25 0.16 .08
450/210 mm (n Z 20) 0.13 G 0.11 0.03 G 0.18 0.08 .16
350/350 mm (n Z 39) 0.06 G 0.23 0.07 G 0.11 0.06 .89
450/450 mm (n Z 13) 0.18 G 0.44 0.04 G 0.22 0.11 .51
Overall (n Z 198) 0.09 G 0.23 0.14 G 0.22 0.11 .12
CDVA Z corrected distance visual acuity; logMAR Z logarithm of the minimum angle of resolution *Intacs, Addition Technology, Inc. † Statistically significant difference
to the crosslinking component of the procedure. This infection rate of 0.8% compares to the U.S. multicenter study of CXL for keratoconus1 in which there was one case of infectious keratitis (0.3%) reported in 293 eyes. Regarding Intacs-related AEs, we found 4 (2.0%), with 3 requiring segment removal. Three of these ICRS events occurred after sequential surgery and 1 after concurrent surgery; in the former group, one occurred after ICRS alone and two after the CXL procedure. Procedure timing per se, therefore, does not seem to affect the likelihood of a medical complication. Does CXL increase the incidence of ICRS-related complications? In regard to the removal incidence of 1.5% reported in this study, in recently published work we reviewed 572 consecutive cases of Intacs and found that the incidence of removal for a medical complication was approximately 2.5%.17 Thus, the medical complication rate in the current study of combined CXL/ICRS does not appear to be higher than the complication rate for Intacs alone. In reviewing the complication spectrum reported in this study, it can be inferred with reasonable confidence that the two procedures, when performed on the same cornea either concurrently or sequentially, do not increase patient risk over the individual procedures alone. Limitations to this study include the small sample sizes for subgroup analysis of Intacs segment size. Therefore, although our sample sized is adequate to assess, in general, both the efficacy and safety of the ICRS/CXL procedures as well as the effect of procedure timing, we cannot extrapolate the results to the different Intacs size subgroups with statistical certitude. Further research is necessary to elucidate more fully the effect of size, symmetry, and positioning
on the outcomes of ICRS.18 Moreover, preoperative patient characteristics can influence the outcomes of CXL. In previous work analyzing characteristics influencing outcomes of CXL alone,19 we found that the only independent indicator of CDVA change after surgery was the preoperative CDVA. Eyes with poorer vision tended to have more improvement after CXL; specifically, eyes with CDVA of 20/40 or worse were 5.9 times more likely to improve 2 lines or more. Similarly, the only independent indicator of maximum K change after surgery was the preoperative maximum K. Eyes with steeper maximum K tended to have more flattening after CXL; specifically, eyes with a maximum K of 55 D or more were 5.4 times more likely to have topographic flattening of 2.0 D or more. Thus, this should be considered, in particular, when assessing subgroups such as the single 450 mm group in which there were differences in some baseline characteristics. In addition to Intacs segment thickness, there are other preoperative characteristics that might influence postoperative outcomes. Previous works by another group20–22 reported a grading system for keratoconus and ectasia, and improved ICRS outcomes in patients with worse preoperative disease; thus, such grading schemes to better define the level of preoperative disease might further enable nomogram development and prediction of postoperative outcomes. In conclusion, this study shows (1) CXL and ICRS (Intacs) can be used adjunctively, (2) ICRS substantially improves corneal topography over CXL alone, (3) ICRS size and placement patterns are important in determining the topography outcome, (4) there are no obvious increases in safety concerns of CXL/ICRS compared with each Figure 7. Average change in outcomes in total population and randomized subgroups corneal crosslinking/intrastromal corneal ring segments (Intacs, Addition Technology, Inc.) (CDVA Z corrected distance visual acuity; I–S Z difference between inferior and superior topographic anterior sagittal topographic power; logMAR Z logarithm of the minimum angle of resolution; Max K Z point of maximum topographic keratometry; UDVA Z uncorrected distance visual acuity).
Volume - Issue - - 2019
10
CXL AND ICRS
Table 7. Complications after CXL/ICRS.* Complication
N (%)
Infectious keratitis
2 (1.0)
Inflammation around ICRS
3 (1.5)
Glare symptoms
1 (0.5)
Randomized Group
Surgical Intervention
1 Concurrent, 1 Sequential 2 Sequential after CXL, 1 Sequential after ICRS alone Concurrent
None 2 ICRS removal
ICRS removal
CXL Z corneal crosslinking; ICRS Z intrastromal corneal ring segments *Intacs, Addition Technology, Inc.
procedure alone, and (5) efficacy and safety outcomes are similar notwithstanding timing of the two proceduresd concurrent versus 3-month sequential surgery.
WHAT WAS KNOWN Corneal crosslinking (CXL) decreases the progression of keratoconus. CXL causes a small average improvement in corneal topography. Intracorneal ring segments (ICRS) can flatten the cone in keratoconus.
WHAT THIS PAPER ADDS CXL and ICRS (Intacs), used adjunctively, substantially improved corneal topography steepness and symmetry. Complications with CXL/ICRS did not increase over each procedure alone. Thicker segment size and single segment placement resulted in greater topographic improvement. Sequential and concurrent surgery with ICRS and CXL had equivalent outcomes.
REFERENCES 1. Hersh PS, Stulting RD, Muller D, Durrie DS, Rajpal RK, United States Crosslinking Study Group. United States multicenter clinical trial of corneal collagen crosslinking for keratoconus treatment. Ophthalmology 2017; 124:1259–1270 2. Hersh PS, Stulting RD, Muller D, Durrie DS, Rajpal RK. United States Crosslinking Study Group. U.S. multicenter clinical trial of corneal collagen crosslinking for treatment of corneal ectasia after refractive surgery. Ophthalmology 2017; 124:1475–1484 3. Colin J, Cochener B, Savary G, Malet F. Correcting keratoconus with intracorneal rings. J Cataract Refract Surg 2000; 26:1117–1122 4. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol 2003; 135:620–627 5. Greenstein SA, Fry KL, Hersh PS. Corneal topography indices after corneal collagen crosslinking for keratoconus and corneal ectasia: one year results. J Cataract Refract Surg 2011; 37:1282–1290 6. Greenstein SA, Fry KL, Hersh MJ, Hersh PS. Higher-order aberrations after corneal collagen crosslinking for keratoconus and corneal ectasia. J Cataract Refract Surg 2012; 38:292–302 7. Brooks NO, Greenstein SA, Fry KL, Hersh PS. Patient subjective visual function after corneal collagen crosslinking for keratoconus and corneal ectasia. J Cataract Refract Surg 2012; 38:615–619
Volume - Issue - - 2019
8. Torquetti L, Ferrara G, Almeida F, Cunha L, Araujo LPN, Machado AP, Lyra JM, Merayo-Lloves J, Ferra P. Intrastromal corneal ring segment implantation in patients with keratoconus: 10-year follow-up. J Cataract Refract Surg 2014; 30:22–26 9. El-Raggal TM. Sequential versus concurrent Kerarings insertion and corneal collagen cross-linking for keratoconus. Br J Ophthalmol 2011; 95:37–41 10. Coskunseven E, Jankov MR, Hafezi F, Atun S, Arslan E, Kymionis GD. Effect of treatment sequence in combined intrastromal corneal rings and corneal collagen crosslinking for keratoconus. J Cataract Refract Surg 2009; 35:2084–2091 11. Yildirim A, Uslu H, Kara N, Cakir H, Gurler B, Colak HN, Ozgurhan EB. Same-day intrastromal corneal ring segment and collagen cross-linking for ectasia after laser in situ keratomileusis: long-term results. Am J Ophthalmol 2014; 157:1070–1076 12. Cakir H, Pekel G, Perente I, Genc¸ S. Comparison of intrastromal corneal ring segment implantation only and in combination with collagen crosslinking for keratoconus. Eur J Ophthalmol 2013; 23:629–634 13. Legare ME, Iovieno A, Yeung SN, Lichtinger A, Kim P, Hollands S, Slomovic AR, Rootman DS. Intacs with or without same-day corneal collagen cross-linking to treat corneal ectasia. Can J Ophthalmol 2013; 48:173–178 14. Renesto Ada C, Melo LA Jr, Sartori Mde F, Campos M. Sequential topical riboflavin with or without ultraviolet a radiation with delayed intracorneal ring segment insertion for keratoconus. Am J Ophthalmol 2012; 153:982–993 15. Elsaftawy HS, Ahmed MH, Saif MY, Mousa R. Sequential intracorneal ring segment implantation and corneal transepithelial collagen cross-linking in keratoconus. Cornea 2015; 34:1420–1426 16. Hersh PS, Greenstein SA, Fry KL. Corneal collagen crosslinking for keratoconus and corneal ectasia: One year results. J Cataract Refract Surg 2011; 37:149–160 17. Nguyen N, Gelles JD, Greenstein SA, Hersh PS. Incidence and associations of intracorneal ring segment explantation. J Cataract Refract Surg 2019; 45:153–158 18. Chan CCK, Sharma M. Boxer Wachler BS. Effect of inferior-segment Intacs with and without C3-R on keratoconus. J Cataract Refract Surg 2007; 33:75–80 19. Greenstein SA, Hersh PS. Characteristics influencing outcomes of corneal collagen crosslinking for keratoconus and ectasia: Implications for patient selection. J Cataract Refract Surg 2013; 39:1133–1140 JL, Pin ~ero DP, Aleso n A, Teus MA, Barraquer RI, Murta J, 20. Alio rrez R, Villa C, UcedaMaldonado MJ, Castro de Luna G, Gutie Montanes A. Keratoconus-integrated characterization considering anterior corneal aberrations, internal astigmatism, and corneal biomechanics. J Cataract Refract Surg 2011; 37:552–568 JL, Vega-Estrada A, Baviera J, Beltran J, Cobo-Soriano R. 21. Brenner LF, Alio Clinical grading of post-LASIK ectasia related to visual limitation and predictive factors for vision loss. J Cataract Refract Surg 2012; 38:1817–1826 JL, Brenner LF, Javaloy J, Plaza Puche AB, 22. Vega-Estrada A, Alio Barraquer RI, Teus MA, Murta J, Henriques J, Uceda-Montanes A. Outcome analysis of intracorneal ring segments for the treatment of keratoconus based on visual, refractive, and aberrometric impairment. Am J Ophthalmol 2013; 155:575–584 OTHER CITED MATERIAL A. U.S. National Institutes of Health Clinical Trials. Corneal Collagen Crosslinking and Intacs for Keratoconus and Ectasia (CXL). NCT01112017. Available at: https://clinicaltrials.gov/ct2/show/NCT01112072. Accessed February 27, 2019
Disclosures: Dr. Hersh receives personal fees from Addition Technology, Inc. and Avedro, Inc. outside the submitted work. None of the other authors has a financial or proprietary interest in any material or methods mentioned.
First author: Peter S. Hersh, MD Hersh Vision Group, CLEI Center for Keratoconus, Teaneck, New Jersey, USA