The Outcome of Corneal Collagen Cross-Linking in Patients with Advanced Progressive Keratoconus: A 2-Year Follow-up Study.

PURPOSE: The aim of this study was to evaluate the safety and efficacy of collagen cross-linking (CXL) in advanced progressive keratoconus with a maximum keratometry (Kmax) value of more than 58 diopters (D).
METHODS: This prospective interventional case series involved patients with advanced progressive keratoconus with a Kmax of more than 58 D. The best-corrected visual acuity (BCVA), uncorrected visual acuity (UCVA), Kmax, mean keratometry (Kmean) value, corneal astigmatism, and thinnest corneal thickness before surgery and 24 months after CXL were determined for 30 eyes of 27 patients. A Pentacam was used to measure the paraclinical parameters.
RESULTS: The mean age of the patients was 24.47 ± 3.33 years. The mean logarithm of the minimum angle of resolution (logMAR) of the uncorrected visual acuity UCVA decreased from 0.73 ± 0.36 D at baseline to 0.48 ± 0.30 D (P = 0.01), while the mean thinnest point thickness of the cornea decreased from 438.65 ± 40.11 μm to 431.43 ± 61.92 μm (P = 0.005) after 24 months. The decreases in the mean logMAR of the BCVA, Kmax and Kmean values, and corneal astigmatism were not statistically significant (P > 0.05) at the 24-month follow-up. Progression was halted in 29 eyes (96.6%); only 1 eye (3.3%) showed an increase in the Kmax value of more than 2.0 D, which was indicative of treatment failure. In contrast, most other eyes showed a decrease in the Kmax value although it was not statistically significant. There were no major complications in any of the patients during the study period.
CONCLUSION: Standard CXL treatment was safe and stabilized both the visual acuity and tomographic parameters at the 2-year follow-up in eyes with advanced progressive keratoconus.

Introduction

Keratoconus is a noninflammatory and progressive disease that causes thinning of the cornea and is characterized by a cone-like shape protrusion of the corneal surface.[1] Although it affects all ethnicities, studies have found a higher incidence and prevalence in Asian populations compared with European populations.[23] In addition, a young age at onset is associated with a high severity of the disease.[4] In Iran, the prevalence of keratoconus is 2.5%–3.3%, which is higher than in other countries.[56]

The main histopathology of keratoconus is the progressive degeneration of the corneal basal epithelial cells and the basement membrane,[7] followed by the degeneration of Bowman's layer and the fragmentation of tiny fibers of protein, called collagen, which help hold the cornea in place and maintain its shape.[8] Further structural changes within the cornea cause it to thin and form a conical shape.

The classical treatment of keratoconus is based on correcting the visual impairment. However, recent studies have suggested a method that can biologically correct the disease based on its histopathology. Collagen cross-linking (CXL) is a procedure that induces the cross-linking of corneal collagen using a combination of riboflavin and ultraviolet (UV) light.[9] CXL is the only treatment that reverses the pathophysiology of keratoconus, and thus, it is practiced around the world.[101112] In addition to improving the biomechanical stability of the corneal tissue, CXL halts the progression of ectasia, flattens the cornea, and improves the keratometric indices, although it seems less effective in correcting the refractive errors.[13] However, several complications of CXL have been reported, and many cases require repeating the CXL procedure or providing additional treatments.[14] CXL is also not feasible for everyone;[15] for example, patients with advanced grades of keratoconus might not benefit from this procedure.[16] Few studies have addressed the outcome of CXL in patients with advanced progressive keratoconus. Thus, this study aimed to evaluate the long-term safety and efficacy of CXL in patients with advanced keratoconus with a maximum keratometry (Kmax) value >58 diopters (D).

Methods

Study design

This prospective study recruited all patients with advanced progressive keratoconus who were referred to the Cornea Clinic of Amir-Almomenin Hospital, Rasht, Iran from December 2014 to September 2015, underwent CXL, and were followed up for at least 2 years. The protocol of the study was approved by the Ethics Committee of Guilan University of Medical Science, Rasht, Guilan, Iran Before being recruited, the design and objectives of the study were explained to the patients and written informed consent was obtained. Each step of the study was performed in accordance with the principles of the Declaration of Helsinki.

Patients who met the inclusion criteria were more than 12 years of age and had a definitive diagnosis of keratoconus for which CXL was indicated. Patients were diagnosed with keratoconus if they showed at least one typical sign of the disease during a slit lamp examination (SLE) such as a Fleischer ring, Vogt striae, corneal stromal scar, stromal thinning, or conical bulging in the cornea apex, or if they showed topographic characteristics of keratoconus. Pregnant and lactating women and patients with a positive history of systemic disease, previous corneal and intraocular surgery, a history of chemical damage or delay in corneal epithelium repair, or a neurological or retinal disease causing reduced vision were excluded from the study.

After taking a complete history of the patients and recording their demographic characteristics, the researcher performed the following eye examinations and recorded the data: SLE, best-corrected visual acuity (BCVA), uncorrected visual acuity (UCVA), ophthalmoscopy, and imaging with corneal tomography using a rotating Scheimpflug imaging system (Pentacam HR; Oculus, Wetzlar, Germany). In this study, the severity of keratoconus was scored according to the results by Ishii et al.[17] Patients with corneal scarring due to severe keratoconus were excluded from the study because the Scheimpflug imaging could give false results in these cases. The progression of keratoconus was defined as a deterioration of BCVA (loss of 1 or more lines on the Snellen chart) and an increase in Kmax of 1.0 D or more at the 1-year follow-up. The treatment failure of CXL was defined as an increase in the Kmax reading of more than 1.00 D over the preoperative value. After the data had been collected and recorded, patients underwent CXL surgery. The operation was performed by one cornea surgeon (R.S.M) who used the same protocol for all patients.

Surgical technique

Patients underwent CXL using the standard Dresden protocol. However, hypotonic riboflavin was used in cases in which the corneal thickness was <400 μm after epithelial removal. The patient was given a topical anesthetic, and the central 8.0 mm of the corneal epithelium was debrided using a blunt spatula; the anesthetic was 1.0% tetracaine (Anestocaine; Sina Daru, Tehran, Iran) eye drops. Iso-osmolar riboflavin solution with 0.1% riboflavin in 20% dextran (Ricrolin; Sooft Italia SpA, Montegiorgio, Italy) was used for most eyes. However, a hypo-osmolar riboflavin solution was administered to eyes with a denuded corneal thickness <400 μm for the procedure and applied topically every 3 min for 30 min. The eyes were exposed to UVA irradiation (CSO-Vega X-linker; Scandicci, Florence, Italy) at a wavelength of 370 nm with a surface irradiance of 3.0 mW/cm2 for 30 min (surface dose 5.4 J/cm2). Throughout the irradiation phase, the riboflavin solution was applied every 3 min to ensure that the stromal surface was kept moist. A soft bandage contact lens was applied until re-epithelialization was complete. After surgery, the patients applied topical 0.5% chloramphenicol (Chlobiotic; Sina Darou, Tehran, Iran) 4 times a day for 7 days and 0.1% betamethasone (Betasonate; Sina Darou, Tehran, Iran) 3 times a day for 20 days. Preservative-free artificial tears (Sina Tears, Sina Darou, Tehran, Iran) were administered every 6 h for 45 days to lubricate the ocular surface.

Patients were followed up 1 week, 1, 3, 6, 12, and 24 months after surgery. During each visit, they underwent full ophthalmic examinations. Images were obtained using the Pentacam HR, 12 and 24 months after CXL to complete the data collection.

Statistical analysis

The distribution of the qualitative variables was given as a frequency (percentage) while the quantitative variables were reported as the mean ± standard deviation. The Wilcoxon test was used to assess the data of the variables that showed a normal distribution; the paired sample t-test was used to compare the ophthalmic measurements before and after surgery. Data were analyzed using IBM SPSS Statistics for Windows (version 21.0; IBM Corp. Armonk, NY, USA). Statistical significance was considered at P < 0.05.

Results

A total of 30 eyes from 27 patients were studied; the mean age of the patients was 24.47 ± 3.33 years (range of 17–30 years). Seventeen eyes (56.7%) belonged to male patients and 13 eyes (43.3%) belonged to female patients. The CXL was performed bilaterally in 3 patients and unilaterally in 24 patients. The left eye was studied in 60% of the cases (n = 18), and the right eye was studied in 40% of the cases (n = 12).

The mean logarithm of the minimum angle of resolution (logMAR) of the UCVA decreased from 0.73 ± 0.36 D at baseline to 0.47 ± 0.31 D and 0.48 ± 0.30 D, 12 and 24 months after CXL, respectively (P < 0.05) [Figures 1 and 2]. The mean logMAR of the BCVA decreased from 0.59 ± 0.34 before CXL to 0.44 ± 0.33 after 12 months and 0.45 ± 0.32 after 24 months although the difference was not statistically significant at 24 months (P = 0.07). The mean Kmax also decreased from 62.19 ± 4.56 D before CXL to 60.91 ± 4.36 D after 12 months and 60.95 ± 4.42 D after 24 months while the mean keratometry (Kmean) decreased from 51.37 ± 4.27 D before CXL to 50.62 ± 3.06 D after 12 months and 50.62 ± 3.06 D after 24 months; these changes were not statistically significant (P > 0.05) [Table 1]. Progression was halted in 29 eyes (96.7%) while progression or treatment failure was noted in 1 eye (3.3%) over the 2-year follow-up period. The thinnest point thickness (TPT) decreased from 438.65 ± 40.11 μm to 431.43 ± 61.92 μm after 24 months (P = 0.005) [Figure 3]. The TPT was also found to be significantly decreased after 24 months; however, this change was unlikely to be clinically significant. The mean corneal astigmatism before CXL was 4.51 ± 1.70 D, which changed to 4.69 ± 1.67 D after 24 months, although the change was not statistically significant (P = 0.18). There were no major complications in the patients during the study period.

Figure 1

Mean logarithm of the minimum angle of resolution of the uncorrected visual acuity before collagen cross-linking

Figure 2

Mean logarithm of the minimum angle of resolution of the uncorrected visual acuity 24 months after collagen cross-linking

Table 1

Mean and standard deviation of clinical and tomographic ophthalmologic measurements compared before and 12 and 24 months after surgery

Variable	Before surgery, (mean±SD)	12 months after surgery, (mean±SD)	P*	24 months after surgery, (mean±SD)	P*
UCVA (logMAR)	0.73±0.36	0.47±0.31	0.008	0.48±0.30	0.010
BCVA (logMAR)	0.59±0.34	0.44±0.33	0.052	0.45±0.32	0.073
Kmax (D)	62.19±4.56	60.91±4.36	0.091	60.95±4.42	0.106
Kmean (D)	51.37±4.27	50.62±3.06	0.251	50.62±3.06	0.554
Thinness point thickness (μm)	438.65±40.11	430.46±41.05	0.039	431.43±61.92	0.005
Corneal astigmatism (D)	4.51±1.70	4.70±1.64	0.243	4.69±1.67	0.179

*Significance level is 0.05, the results of paired samples t-test. D: Diopter, logMAR: The mean logarithm of the minimum angle of resolution, Kmean: Mean keratometry, Kmax: Maximum keratometry, BCVA: Best-corrected visual acuity, UCVA: Uncorrected visual acuity, SD: Standard deviation

Figure 3

Preoperative thinnest point thickness and 24 months after collagen cross-linking

Discussion

Some studies have suggested that there is an increased failure rate of CXL in corneas with a Kmax of more than 58 D. However, few studies have investigated CXL in patients with advanced keratoconus. In this study, we examined the results of CXL on 27 patients with advanced progressive keratoconus in corneas with a Kmax≥58 D. The 30 eyes studied showed a significant improvement in UCVA and a decrease in TPT although the improvement in BCVA and the decreases in Kmax, Kmean, and corneal astigmatism were not statistically significant at the 2-year follow-up. In the present study, the research team ophthalmologist confirmed the diagnosis of advanced progressive keratoconus, and patients with Grades I–IV according to the Amsler–Krumeich classification underwent CXL. The results confirmed the feasibility and safety of CXL on different stages of advanced keratoconus.

The present study considered brief changes in cone steepness (Kmax and Kmean) after CXL in advanced keratoconic eyes, which was not statistically significant. In 29 eyes (96.6%), no progression was found, which indicated that the treatment had stabilized the cone. Continued Kmax progression 2 years after CXL, which indicated treatment failure, was observed in 1 eye (3.3%). After CXL, visual improvement occurred and the mean UCVA improved significantly. The improvement in BCVA was not statistically significant at 24 months. One explanation for this is that advanced cases do better with contact lens fittings and we only evaluated the spectacle-corrected visual acuity.

Keratoconus is a progressive disease that affects visual acuity, which in some cases cannot be easily corrected using traditional methods.[18] The introduction of CXL as a unique method to halt the pathophysiology of keratoconus was considered to be a revolution in the treatment of the disease. However, nearly two decades after the introduction of CXL, there are still some areas that need to be further investigated.[19] While several studies have investigated the efficacy and safety of CXL, some studies have suggested that CXL is associated with complications such as fungal and microbial infections[202122] and damage to the cornea.[2324] Although these studies mainly involved case reports, other studies have presented strong evidence that suggests that the safety of CXL depends on the techniques used during the procedure such as the homogeneous irradiance of the light source design at a wavelength of 370 nm, the removal of the epithelium to facilitate the diffusion of the riboflavin, which is applied 30 min before UV exposure, and a minimum corneal stroma thickness of 400 μm.[252627] The patients in this study had no complications, which confirmed the safety of CXL in cases of advanced keratoconus and justified the adherence to standardized and appropriate techniques, as well as reflected the surgical skills and experience of the team.

Appropriate patient selection is considered to be an important issue that can affect the complication and failure rate of CXL. As suggested, patients >35 years of age are considered to be at risk for a high rate of complications.[14] All patients included in the present study were >12 years of age and <30 years of age.

A few studies have reported treatment failure after CXL.[2829] In one study, failure was defined as the continued progression of the pathophysiology and was reported in 7.6% of patients with Kmax>58 D despite 2.8% in cases with lower Kmax in 1-year follow-up.[14] Another study indicated that patients with a preoperative Kmax<55 D had better outcomes. Furthermore, it has been suggested that CXL fails more often in cases of advanced keratoconus, as determined by the Kmax values, and thus, a safety margin of 58 D was suggested.[15] Compared with other studies,[1030] the present study on patients with Kmax>58 D had a low failure rate because there was only 1 case (3.3%) of treatment failure in which there was >2 D increase in the Kmax value at the 2-year follow-up; in 29 eyes, keratoconus progression was stopped. Other studies have found a significant increase in the Kmax and Kmean at the 1-year follow-up.[3132] In the present study, the Kmax decreased by 1 D in 15 cases (almost half of the study population) after CXL and by 1.5 D in one-third of the cases; these results are similar to the results of a previous study.[16] Because the preoperative Kmax value may play a role in the failure rate, the mean preoperative Kmax in the present study was determined to be 62.19 ± 4.56 D, which was lower than the previous study.[16] Another study suggested that CXL corrects the refractive errors, which allows for clinical improvement, rather than topographic measurements.[16] Therefore, considering the improved visual acuity in the present study, CXL can be an effective method for the treatment of advanced keratoconus.

The results of the present study also indicated that there was a statistically significant decrease in the logMAR of the UCVA (0.24 ± 0.46 D) at the 24-month follow-up; in other words, there was a significant improvement in visual acuity. This result agreed with the findings of previous studies.[113133] In addition, there was a decrease in the logMAR of the BCVA (0.13 ± 0.39) after 24 months, which was not statistically significant, although 13 patients improved more than 2 lines in the Snellen chart. The unchanged astigmatism in the present study showed the efficacy of the treatment method in stabilizing the cone and improving visual acuity, which were similar to other studies that reported stability in the corneal keratometry after CXL.[3435] The statistically significant decrease in the TPT was not clinically significant despite the fact that changes in the corneal thickness could be related to stromal compaction from the CXL treatment, rather than the progression of the disease.[36]

In the current study, there were no major complications, such as infections or significant loss of vision, in the patients. Patients with advanced keratoconus who underwent CXL were successfully followed for 2 years, and the results showed the relatively long-term effect of CXL on the disease. However, there were several limitations to the study, including the limited sample size, the lack of a control group, and the nonrandomized inclusion of patients into the study.

Conclusion

The present study showed that CXL prevented the further progression of advanced progressive keratoconus in 29 of 30 eyes at the 2-year follow-up. Therefore, CXL is an effective and safe method to treat patients with advanced progressive keratoconus and improves the tomographic and clinical indices of the disease, and the risk of complications and failure of treatment may equal to less progressed cases. Our study emphasized that it is not necessary to exclude cases of advanced progressive keratoconus from CXL. Further studies with a larger sample size and for a longer follow-up period could enhance the results of the present study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.