Correlation Between UV Absorption and Riboflavin Concentration in Different Depths of the Cornea in CXL

Dear Editor,

In their article, Søndergaard et al.¹ concluded that the riboflavin during CXL treatment is only enriched in the anterior cornea. These results are in conflict with other measurements.

Riboflavin fulfills two inseparable functions in the collagen cross-linking treatment (CXL). First, it serves as a photosensitizer of the photochemical cross-linking reaction and promotes the production of reactive singlet-oxygen. Second, it acts as an UV absorber to protect the endothelium, lens, and retina from excessive UV irradiation.

Further theoretical investigations showed that riboflavin requires about 30 min to diffuse sufficiently into the stroma and to achieve a high concentration and high absorption in the posterior stroma.²

Recently, several studies were published using various methods to determine the riboflavin concentration in the cornea and especially its depth distribution. The different measurement methods may have led to different (and in some cases, contradictory) results. It is important to know that the absorption coefficient is depth dependent in the corneal stroma itself.

Søndergaard et al.¹ found by laser scanning confocal fluorescence microscopy that the riboflavin is enriched only in the anterior 200 µm. The maximum riboflavin concentration in the anterior cornea was 0.01 mg/ml = 0.001% after application of 0.1% riboflavin for 30 min. Such a low concentration would give only an additional absorption coefficient of µ = 0.6 cm⁻¹. A possible explanation for this observation may be that the exponential attenuation of the exciting light with depth, especially in highly absorbing media, was ignored. These can be checked easily by a measurement of the fluorescence in a 500-µm-thick homogenous riboflavin solution of 0.1%. For that reason these measured values of the riboflavin distribution in the cornea by the confocal fluorescence microscope are not representative.

Kampik et al.³ measured the riboflavin autofluorescence with infrared-excited two photon microscopy. They found a distribution of riboflavin throughout the corneal stroma contrary to Søndergaard. The depth-associated loss of signal caused by the scattering of both laser light and emission was used to normalize the measured riboflavin signal. Thus, after 30 min of riboflavin application, the concentration of riboflavin in a stromal depth of 400 µm was about 0.05% or 50% of the concentration in the superficial stroma.

The absorption coefficient of riboflavin for UV light depends linearly on the concentration in the range from 0–0.2%.^2,4 The absorption coefficient, µ, of riboflavin for the wavelength of 370 nm can be calculated from the concentration, c, by the equation µ = 56 × c cm⁻¹, where c is in mg/ml.⁵ From this relationship, it is easy to find the concentration from the absorption coefficient or vice versa.

We measured directly the absorption⁶ in different depths to control the absorption effect also in the posterior stroma and the correlation to the riboflavin concentration. After epithelial abrasion and 0.1% riboflavin application for 10, 20, and 30 min and without riboflavin (controls) the absorption coefficient was measured in layers of the stroma (flap thickness 180 µm), which were cut with a microkeratome from the porcine cornea (Figure 1). Our measurements of the absorption in the first flap (0–180 µm) after 30 min riboflavin application are in accordance with the measurements of Wollensak et al.⁴ They found absorption values for the stroma without riboflavin film of about 35 cm⁻¹. These measured absorption coefficients are also in line with the theoretical calculation of the concentration of 0.06% in 200 µm depth after 30 min.² The absorption coefficients are also in agreement with the concentrations measured by Kampik et al.³ After 20 to 30 min of riboflavin application, a steady state is reached and only a moderate gradient of riboflavin concentration can still be observed. However, riboflavin is enriched also in the posterior stroma, which is reflected both in our absorption measurements in the posterior layers as well as in findings with two-photon fluorescence microscopy.³ The appearance of riboflavin in the anterior chamber on slit lamp exam is further evidence for the diffusion through the posterior cornea.

FIGURE 1  Absorption coefficient and riboflavin concentration (y-axis) in different stromal depths (x-axis) before application of riboflavin solution and after application at 10, 20, and 30 min. The absorption coefficient increases with riboflavin concentration as shown by Wollensak et al.⁴

Declaration of interest: The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.