Secondary Logo

Journal Logo


Natural history of corneal haze after corneal collagen crosslinking in keratoconus using Scheimpflug analysis

Kim, Bia Z. MBChB; Jordan, Charlotte A. PhD, BOptom; McGhee, Charles N.J. DSc, FRCOphth; Patel, Dipika V. PhD, MRCOphth*

Author Information
Journal of Cataract & Refractive Surgery: July 2016 - Volume 42 - Issue 7 - p 1053-1059
doi: 10.1016/j.jcrs.2016.04.019
  • Free


Corneal collagen crosslinking (CXL) with riboflavin and ultraviolet-A (UVA) light was successfully introduced by Wollensak et al.1 in 2003 to delay or halt the progression of keratoconus.1–3 Riboflavin and UVA interact to produce free radicals, resulting in chemical bonds within the corneal stroma, thereby stiffening the cornea.4,5

After this procedure is performed, varying degrees of transient corneal haze are observed in most eyes; however, this is difficult to monitor subjectively over time, especially with different observers.6 The Pentacam Scheimpflug system (Oculus Optikgeräte GmbH) is widely used for tomography, pachymetry, and anterior chamber depth analyses. More recently, densitometry software was introduced to this Scheimpflug system, enabling quantitative analysis of corneal haze. Specifically, this software provides information regarding the amount of backscattered light in 12 areas of the cornea. This software has been used in recent studies to assess corneal haze in normal corneas and keratoconic corneas to determine the pattern of densitometry over time. There are contradicting studies of whether densitometry remains elevated 12 months after CXL in keratoconic eyes, and all previous studies included topical corticosteroid treatment in their postoperative care protocols,6–12 which might alter the natural history of haze development and regression.

Objective measurements of corneal haze using densitometry could allow more accurate determination of correlations between preoperative patient and eye characteristics and help the surgeon to anticipate preoperatively who will develop corneal haze and to what degree. Such analyses would also be valuable in understanding the pathophysiology underlying the development of corneal haze as well as perioperative management.

The aim of this prospective study was to quantitatively analyze corneal haze over a 12-month period after CXL for keratoconus using Scheimpflug imaging and densitometry software.

Patients and methods

The patients were recruited from ophthalmology and optometry practices across New Zealand and treated and examined in the Department of Ophthalmology, Greenlane Clinical Centre, Auckland District Health Board, Auckland, New Zealand, as part of a separate prospective randomized controlled clinical study of CXL in patients with progressive keratoconus.

Patient Inclusion

All patients were informed about the treatment and provided consent. The study was approved by the Northern X regional ethics committee and adhered to the tenets of the Declaration of Helsinki. All patients had bilateral progressive keratoconus, defined as an increase in the maximum keratometry (K) value of 0.75 diopter (D) or more in the preceding 3 months; a change in refractive astigmatism of 0.75 D or more in the preceding 12 months; a change in the base curve of contact lens fit of 0.2 mm or more in the preceding 12 months; or a decrease in corneal thickness of 30 μm or more in the preceding 6 months. One eye was randomized to be treated by CXL using a random-number generator.

Patients with progressive keratoconus were examined in this study. Inclusion criteria included age 14 to 30 years, maximum K of 60.0 D or less, minimal corneal thickness of 400 μm or more, corrected distance visual acuity (CDVA) of 20/80 or better, a clear cornea with no scarring, and rigid gas-permeable contact lens intolerance. Exclusion criteria included previous ocular trauma or surgery, ocular disease (other than keratoconus) or systemic disease that might affect the cornea, and stable keratoconus.

Surgical Technique

Standard corneal CXL according to Wollensak et al.1 was performed as an outpatient procedure at the eye clinic. After topical anesthesia was applied, a 7.0 to 8.0 mm diameter of the central corneal epithelium was mechanically removed using a bevel-up crescent blade. For corneas with a postepithelial debridement corneal thickness of at least 400 μm, isotonic riboflavin 0.1% ophthalmic solution was instilled in 10 mg riboflavin-5-phosphate in 10 mL dextran-T-500 (Opto Ribolink, Opto Global Pty. Ltd). For corneas with a postepithelial debridement corneal thickness of less than 400 μm, hypotonic riboflavin 0.1% ophthalmic solution was instilled (Opto Ribolink, Opto Global Pty. Ltd). After epithelial debridement, riboflavin drops were instilled every 5 minutes for 30 minutes. Light irradiation of the cornea was then started using a UVA double-diode 370 nm light source located 10.0 to 12.0 mm in front of the corneal apex. This produced a radiant energy of 3mW/cm2 or 5.4 J/cm2 (monitored via a potentiometer/UV power meter). Irradiation was performed for 30 minutes with further instillation of riboflavin drops every 5 minutes. On completion of the procedure, the eye was flushed with a balanced salt solution and a soft bandage contact lens was inserted for the immediate 24-hour postoperative period.

Patients were treated postoperatively with preservative-free chloramphenicol 0.5% drops 4 times daily for 1 week and oral tramadol (50 mg every 4 to 6 hours) and oral diazepam (4 mg at night) as required for up to 72 hours.


Both eyes of all patients were examined preoperatively and 1, 3, 6, and 12 months postoperatively. At each visit, uncorrected distance visual acuity and CDVA were recorded. Data were collected regarding age, sex, self-reported ethnicity, and personal and family ocular history.

Slitlamp microscopy was performed to subjectively assess corneal haze. Corneal tomography and pachymetry were assessed using a rotating Pentacam Scheimpflug camera. The device’s corneal densitometry software was used to analyze backscattered light over a 12.0 mm diameter area, divided into 4 annular concentric zones (0.0 to 2.0 mm, 2.0 to 6.0 mm, 6.0 to 10.0 mm, and 10.0 to 12.0 mm diameter) and 3 layers of depth (anterior 120 μm, middle, and posterior 60 μm). Densitometry is expressed in grayscale units (GSU) of backscattered light on a scale of 0 to 100 (Figure 1).

Figure 1
Figure 1:
Output screen of Scheimpflug densitometry data (N = nasal; T = temporal).

Statistical Analysis

Statistical analysis involved a general linear mixed model with densitometry as the outcome, time as a repeated measure, and person as a random effect to account for the clustering of treated eyes and untreated eyes of the same patient. Variables included treatment group, baseline densitometry, maximum K, minimum K, astigmatism, central corneal thickness (CCT), and CDVA. The effect of baseline keratometry, astigmatism, CCT, and CDVA on the difference in densitometry over time between treated eyes and untreated eyes was examined (3-way interactions). These interactions were removed if not statistically significant; the main effects of these variables on densitometry were also examined.

To evaluate whether CDVA changed in association with densitometry, a separate general linear mixed model was used with treatment, time, their interaction, and CDVA as the explanatory variables; P values less than 0.05 were considered statistically significant.


The study comprised 36 patients (25 men, 11 women). One eye of each patient had CXL (22 right eyes, 14 left eyes). The mean age of the patients was 21.9 years ± 6.1 (SD). The ethnicity distribution was 19 (52.8%) New Zealand European, 6 (16.7%) Maori, 6 (16.7%) Pacific Island Nations, and 5 (13.9%) other. Ten patients had a family history of keratoconus, 5 had a history of eczema, and 10 had asthma. Table 1 shows the preoperative data. Postoperatively, 29 patients (80.6%) returned for the follow-up at 1 month, 29 patients (80.6%) at 3 months, 32 patients (88.9%) at 6 months, and 24 patients (66.7%) at 12 months.

Table 1
Table 1:
Preoperative baseline data in eyes treated with corneal CXL and in contralateral untreated eyes.

The treated eyes had significantly higher densitometry than control eyes at 1 month after treatment (P < .01), with the estimated difference being 5.7 GSU (95% confidence interval [CI], 3.8 to 7.5). After treatment, there was a difference in the trend of the mean densitometry between treated eyes and untreated eyes over 12 months (P < .01). The estimate of the treatment slope was −0.29 (95% CI, −0.45 to −0.12) and the control slope 0.13 (95% CI, −0.05 to 0.31), showing a significant decline in densitometry in treated eyes after 1 month. The densitometry in treated eyes approached values of the untreated eyes after 6 months, with a difference of 1.1 GSU (95% CI, −0.7 to 3.1) at 12 months. Figure 2 shows the pattern of mean densitometry over 12 months in treated eyes and untreated eyes; Table 2 shows the 95% CIs.

Figure 2
Figure 2:
Pattern of mean densitometry over 12 months in treated eyes and untreated eyes (GSU = grayscale units).
Table 2
Table 2:
Mean densitometry with 95% CI in eyes treated with corneal CXL and in contralateral untreated eyes over 12 months.

There was no evidence that baseline maximum K (P = .67), minimum K (P = .59), astigmatism (P = .42), or CCT (P = .32) influenced the difference in the densitometry trend between treated eyes and untreated eyes. These interactions were removed, showing that baseline CCT was inversely related to densitometry (P = .03) after adjustment for treatment and time effect (ie, eyes with thinner corneas at baseline had higher densitometry overall) (Table 3).

Table 3
Table 3:
Associations between densitometry and baseline ocular parameters in treated eyes using general linear mixed model analyses.

In untreated eyes, the highest densitometry measurements were observed in the anterior 120 μm of the cornea and the peripheral (10.0 to 12.0 mm) zone. After corneal CXL, the highest densitometry measurements were observed in the anterior 120 μm of the cornea and inner (0.0 to 2.0 mm) zone. The anterior cornea and inner cornea continued to have the highest densitometry readings for 12 months after CXL.

One month postoperatively, 11 patients had clinically apparent corneal haze on slitlamp microscopy. In contrast, 2 patients had corneal haze on slitlamp microscopy 12 months postoperatively. The mean densitometry was significantly higher in treated patients who had documented haze on slitlamp microscopy at any timepoint (mean 26.0 ± 5.8) than those who had no clinically evident haze (mean 23.2 ± 4.8) (mean difference 2.6) (P = .04). A relationship between CDVA and densitometry was not found (P = .300).


Transient corneal haze is a common clinical observation after CXL for keratoconus. The Pentacam Scheimpflug densitometry used in this study allows objective assessment of corneal haze by measuring backscattered light with a rotating camera system. This technique avoids the drawbacks associated with subjective clinical grading of haze.

The etiology of corneal haze after CXL is not clearly understood. New molecular bridges formed between collagen lamellae after CXL affect the organization of the regular stromal structure responsible for the transparency of the cornea.9 In vitro measurements of rabbit corneas showed increased collagen fiber diameter after CXL, which might reduce corneal transparency.13 In contrast, widely spaced and unevenly organized collagen fibrils after phototherapeutic keratectomy did not cause a loss in corneal transparency; therefore, they are unlikely to contribute to haze.14

Previous studies15–17 speculated that haze might develop secondary to lacunar edema in apoptotic keratocytes after CXL. In vivo confocal microscopy studies of corneas with stromal haze after corneal CXL18–20 found increased density of the extracellular fibrillar matrix and reduced density of anterior keratocytes that also became hyperreflective in the treated zone. Likewise, migratory keratocytes during wound healing cause a greater scatter of light and might contribute to transient corneal haze after CXL.21 By 6 months posttreatment, activated keratocytes repopulated the stroma, returning to baseline density by 12 months.15,19,20

In the current study, there was a significant, but transient increase in densitometry after CXL. The mean densitometry peaked 1 month after treatment and subsequently decreased, returning to baseline values after 6 months. Similar results have been reported in the literature using Pentacam Scheimpflug densitometry software.6,9,22 Greenstein et al.8 also found a significant increase in corneal haze with densitometry and slitlamp examination at 1 month, although not in all eyes. However, a significant reduction was observed between 6 months and 12 months after CXL and densitometry remained above baseline values at 12 months. The current study allowed a better comparison because patient characteristics remain the same in treated eyes and untreated eyes, whereas previous studies treated both eyes or included eyes of different patients in their control groups.8,9,22 Fluctuations in mean densitometry values were insignificant over the 12 months in contralateral keratoconic eyes that were not crosslinked, indicating the change in treated eyes is a result of the treatment itself.

The patients included in our study did not receive topical corticosteroid treatment as part of their postoperative management. The trend in densitometry in the current study highlights the natural history of haze development and resolution over time without the influence of topical corticosteroids. However, the pattern is very similar to that observed in other studies that used topical corticosteroids after reepithelialization. Although short-term topical steroids might assist with suppressing inflammation in the immediate weeks after CXL, results in the current study suggest that corticosteroid treatment is unlikely to affect the course of postoperative haze. Longer-term application is unnecessary or ineffective in reducing the haze because persistent haze after CXL is resistant to topical corticosteroids.12 Future studies evaluating the effect of topical corticosteroids on corneal haze after CXL might be beneficial.

A transient reduction in corneal thickness because of compaction of the stroma after CXL might increase the refractive index and cause a temporary myopic shift.9 Greenstein et al.8 found that the development of haze did not correlate with a change in visual acuity; however, the absolute degree of haze was correlated with poorer visual acuity, likely caused by advanced disease with greater haze. The current and previous studies found that visual acuity did not necessarily correlate with transient corneal haze.6

On the other hand, permanent haze that persists beyond 12 months should be considered distinct from transient haze. Permanent haze corresponds to hyperreflective fibrotic tissue devoid of cells on in vivo confocal microscopy and could contribute to poor visual outcomes.11,12 Lim et al.23 found a significant increase in astigmatism in 2 of 30 cases that was caused by late-onset deep stromal scarring.

Risk factors for permanent haze after CXL include uncontrolled stromal dehydration during epithelium-off CXL, minimum corneal thickness less than 400 μm, age over 35 years, preoperative activated keratocytes in the anterior stroma, forwarded defocus of the UVA source, lack of riboflavin 0.1% administration during irradiation, excessive riboflavin–dextran 20% solution (causing stromal dehydration), noncompliance with postoperative therapy, postoperative infections, therapeutic contact lens intolerance, and hypoxia causing stromal edema and intense inflammatory response.20,22,24 There is also a suspected increase in permanent haze with the presence of Langerhans cells after removal of the therapeutic contact lens. Mazzotta et al.12,20 found that intensive prolonged topical steroid therapy is indicated in these cases, in contrast to results in previous studies of steroid resistance. Furthermore, eyes with preoperative corneal scarring and Vogt striae with a reticular pattern of dark hyporeflective microstriae on in vivo confocal microscopy, indicative of advanced disease, have an increased risk for corneal haze following CXL.11,12,20

Advanced keratoconus is associated with greater densitometry, and keratoconic corneas are reported to have higher densitometry than normal eyes, even before CXL.10,25 In a large study of normal corneas,7 corneal densitometry also increased with age. These studies led to postulations that haze might be caused by the disruption of the normally well-organized uniform collagen fibers and keratocytes in advanced keratoconus, resulting in more backscatter of light.10,11,16,26 Eyes that developed haze after CXL also had higher K values and thinner corneas preoperatively than those that had clear corneas after CXL.8,11,27,28 In the current study, a lower preoperative CCT correlated with higher densitometry. This suggests that as densitometry increases further after treatment, corneas with a thinner central area are more likely to develop haze.

In a white cohort with no corneal pathology, densitometry was highest in the anterior and peripheral cornea.7 The current study confirms that densitometry is highest in the anterior and peripheral area of keratoconic corneas. After corneal CXL, the anterior 120 μm of the cornea in the inner 0.0 to 2.0 mm concentric zone had the highest densitometry, corresponding to the target area receiving maximum treatment, as found by Pircher et al.29 Even at 12 months, although the mean densitometry had returned to baseline, the anterior inner area had the highest densitometry of all areas. Therefore, treatment with CXL appears to have a persistent effect in terms of altering the distribution of corneal densitometry.

In conclusion, this study strengthens current evidence of the transient development and resolution of corneal haze over a 12-month period after corneal CXL, allowing more informed preoperative counseling for patients, especially in terms of expected outcomes. Specifically, it highlights the resolution of corneal haze without application of corticosteroids in a timeline similar to that in earlier studies that used topical corticosteroids in the treatment regimen. A lower baseline CCT was associated with higher densitometry; thus, haze development might be anticipated in patients with advanced keratoconus. The analysis of densitometry in 3 layers and 4 annular zones of the cornea also provides information about the specific area affected by treatment. Overall, the reported natural history of corneal haze after CXL might enable clinicians to form more appropriate future management plans for the affected eye as well as for the contralateral eye when using CXL.

What Was Known

  • Varying degrees of transient corneal haze is observed in most eyes after corneal CXL.
  • There are contradicting reports of whether densitometry remains elevated 12 months after CXL in keratoconic eyes.

What This Paper Adds

  • The mean corneal densitometry peaked in the anterior 120 μm and inner 0.0 to 2.0 mm zone 1 month after CXL then subsequently decreased to baseline values after 6 months.
  • The preoperative CCT was inversely related to densitometry.
  • Densitometry remained stable in untreated contralateral eyes; thus, the change in densitometry can be attributed to the CXL treatment.


1. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A–induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 135, 2003, p. 620-627, Available at: Accessed May 24, 2016.
2. Raiskup-Wolf F, Hoyer A, Spoerl E, Pillunat LE. Collagen crosslinking with riboflavin and ultraviolet-A light in keratoconus: long-term results. J Cataract Refract Surg. 2008;34:796-801.
3. Vinciguerra P, Albè E, Trazza S, Rosetta P, Vinciguerra R, Seiler T, Epstein D. Refractive, topographic, tomographic, and aberrometric analysis of keratoconic eyes undergoing corneal cross-linking. Ophthalmology. 2009;116:369-378.
4. Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine corneas after riboflavin–ultraviolet-A-induced cross-linking. J Cataract Refract Surg. 2003;29:1780-1785.
5. Greenstein SA, Fry KL, Hersh PS. In vivo biomechanical changes after corneal collagen crosslinking for keratoconus and corneal ectasia: 1-year analysis of a randomized, controlled, clinical trial. Cornea. 31, 2012, p. 21-25, Available at: Accessed May 22, 2016.
6. Chang CY, Hersh PS. Corneal collagen cross-linking: a review of 1 year outcomes. Eye Contact Lens. 40, 2014, p. 345-352, Available at: Accessed May 24, 2016.
7. Ní Dhubhghaill S, Rozema JJ, Jongenelen S, Ruiz Hidalgo I, Zakaria N, Tassignon M-J. Normative values for corneal densitometry analysis by Scheimpflug optical assessment. Invest Ophthalmol Vis Sci. 55, 2014, p. 162-168, Available at: Accessed May 24, 2016.
8. Greenstein SA, Fry KL, Bhatt J, Hersh PS. Natural history of corneal haze after collagen crosslinking for keratoconus and corneal ectasia: Scheimpflug and biomicroscopic analysis. J Cataract Refract Surg. 36, 2010, p. 2105-2114, Available at: Accessed May 24, 2016.
9. Gutiérrez R, Lopez I, Villa-Collar C, González-Méijome JM. Corneal transparency after cross-linking for keratoconus: 1-year follow-up. J Refract Surg. 28, 2012, p. 781-785, Available at: Accessed May 24, 2016.
10. Lopes B, Ramos I, Ambrósio R Jr. Corneal densitometry in keratoconus. Cornea. 2014;33:1282-1286.
11. Raiskup F, Hoyer A, Spoerl E. Permanent corneal haze after riboflavin-UVA-induced cross-linking in keratoconus. J Refract Surg. 2009;25:S824-S828.
12. Mazzotta C, Balestrazzi A, Baiocchi S, Traversi C, Caporossi A., 2007. Stromal haze after combined riboflavin-UVA corneal collagen cross-linking in keratoconus: in vivo confocal microscopic evaluation [letter], Clin Experiment Ophthalmol, 35, 580-582.
13. Wollensak G, Wilsch M, Spoerl E, Seiler T. Collagen fiber diameter in the rabbit cornea after collagen crosslinking by riboflavin/UVA. Cornea. 2004;23:503-507.
14. Connon CJ, Marshall J, Patmore AL, Brahma A, Meek KM. Persistent haze and disorganization of anterior stromal collagen appear unrelated following phototherapeutic keratectomy. J Refract Surg. 2003;19:323-332.
15. Mazzotta C, Balestrazzi A, Traversi C, Baiocchi S, Caporossi T, Tommasi C, Caporossi A. Treatment of progressive keratoconus by riboflavin-UVA–induced cross-linking of corneal collagen; ultrastructural analysis by Heidelberg Retinal Tomograph II in vivo confocal microscopy in humans. Cornea. 2007;26:390-397.
16. Wollensak G, Herbst H. Significance of the lacunar hydration pattern after corneal cross linking. Cornea. 2010;29:899-903.
17. Mazzotta C, Caporossi T, Denaro R, Bovone C, Sparano C, Paradiso A, Baiocchi S, Caporossi A. Morphological and functional correlations in riboflavin UV A corneal collagen cross-linking for keratoconus. Acta Ophthalmol. 90, 2012, p. 259-265, Available at: Accessed May 24, 2016.
18. Mazzotta C, Traversi C, Baiocchi S, Caporossi O, Bovone C, Sparano MC, Balestrazzi A, Caporossi A. Corneal healing after riboflavin ultraviolet-A collagen cross-linking determined by confocal laser scanning microscopy in vivo: early and late modifications. Am J Ophthalmol. 146, 2008, p. 527-533, Available at: Accessed May 24, 2016.
19. Jordan C, Patel D, Abeysekera N, McGhee CNJ. In vivo confocal microscopy analyses of corneal microstructural changes in a prospective study of collagen cross-linking in keratoconus. Ophthalmology. 2014;121:469-474.
20. Mazzotta C, Hafezi F, Kymionis G, Caragiuli S, Jacob S, Traversi C, Barabino S, Randleman JB. In vivo confocal microscopy after corneal collagen crosslinking. Ocul Surf. 2015;13:298-314.
21. Jester JV, Moller-Pedersen T, Huang J, Sax CM, Kays WT, Cavangh HD, Petroll WM, Piatigorsky J. The cellular basis of corneal transparency: evidence for ‘corneal crystallins’. J Cell Sci. 112, 1999, p. 613-622, Available at: Accessed May 24, 2016.
22. Koller T, Mrochen M, Seiler T. Complication and failure rates after corneal crosslinking. J Cataract Refract Surg. 2009;35:1358-1362.
23. Lim LS, Beuerman R, Lim L, Tan DTH. Late-onset deep stromal scarring after riboflavin-UV-A corneal collagen crosslinking for mild keratoconus. Arch Ophthalmol. 129, 2011, p. 360-362, Available at: Accessed May 24, 2016.
24. Mazzotta C, Caragiuli S. Intraoperative corneal thickness measurement by optical coherence tomography in keratoconic patients undergoing corneal collagen cross-linking. Am J Ophthalmol. 2014;157:1156-1162.
25. Anayol MA, Sekeroglu MA, Ceran BB, Dogan M, Gunaydin S, Yilmazbas P. Quantitative assessment of corneal clarity in keratoconus: a case control study of corneal densitometry. Eur J Ophthalmol. 2016;26:18-23.
26. Rozema JJ, Koppen C, Bral N, Tassignon M-J. Changes in forward and backward light scatter in keratoconus resulting from corneal cross-linking. Asia Pac J Ophthalmol. 2013;2:15-19.
27. Raiskup F, Spoerl E. Corneal crosslinking with riboflavin and ultraviolet A. Part II. Clinical indications and results. Ocul Surf. 2013;11:93-108.
28. Wittig-Silva C, Chan E, Islam FMA, Wu T, Whiting M, Snibson GR. A randomized, controlled trial of corneal collagen cross-linking in progressive keratoconus; three-year results. Ophthalmology. 121, 2014, p. 812-821, Available at: Accessed May 24, 2016.
29. Pircher N, Pachala M, Prager F, Pieh S, Schmidinger G. Changes in straylight and densitometry values after corneal collagen crosslinking. J Cataract Refract Surg. 2015;41:1038-1043.
© 2016 by Lippincott Williams & Wilkins, Inc.