Microbiologic, Pharmacokinetic, and Clinical Effects of Corneal Collagen Cross-Linking on Experimentally Induced Pseudomonas Keratitis in Rabbits

Photoactivated Chromophore for Keratitis-Corneal Cross-linking (PACK-CXL)-A Scoping Review Based on Preclinical Studies

Purpose

Photoactivated chromophore for keratitis-corneal cross-linking (PACK-CXL) stabilizes the corneal stroma and eliminates microorganisms. Numerous PACK-CXL protocols, using different energy sources and chromophores, have been applied in preclinical studies, including live animal studies, with various experimental designs and endpoints. So far, a systematic mapping of the applied protocols and consistency across studies seems lacking but is essential to guide future research.

Methods

The scoping review protocol was in line with the JBI Manual for Evidence Synthesis. Electronic databases were searched (Embase, MEDLINE, Scopus, Web of Science) to identify eligible records, followed by a two-step selection process (title and abstract screening, full text screening) for record inclusion. We extracted information on (1) different PACK-CXL protocol characteristics; (2) infectious pathogens tested; (3) study designs and experimental settings; and (4) endpoints used to determine antimicrobial and tissue stabilizing effects. The information was charted in frequency maps.

Results

The searches yielded 3654 unique records, 233 of which met the inclusion criteria. With 103 heterogeneous endpoints, the researchers investigated a wide range of PACK-CXL protocols. The tested microorganisms reflected pathogens commonly associated with infectious keratitis. Bacterial solutions and infectious keratitis rabbit models were the most widely used models to study the antimicrobial effects of PACK-CXL.

Conclusions

If preclinical PACK-CXL studies are to guide future translational research, further cross-disciplinary efforts are needed to establish, promote, and facilitate acceptance of common endpoints relevant to PACK-CXL.

Translational Relevance

Systematic mapping of PACK-CXL protocols in preclinical studies guides future translational research.

High Fluence Increases the Antibacterial Efficacy of PACK Cross-Linking

PURPOSE

Photoactivated chromophore for keratitis cross-linking (PACK-CXL) is used as an adjunct therapy to antibiotic medication in infectious keratitis. This experimental study aimed at quantifying the PACK-CXL efficacy as a function of UV fluence using several bacterial strains and irradiated volumes.

METHODS

Six distinct bacterial strains, including standardized strains and clinically isolated strains from patients with keratitis, were analyzed. Bacterial concentrations between 10 and 10 cells/mL were used (simulating small corneal ulcers). Volumes of either 11 μL (≈285 μm stromal thickness) or 40 μL (≈1000 μm stromal thickness) were irradiated within a microtiter plate at different fluences (5.4-27 J/cm) and irradiances (3, 9 and 18 mW/cm). The ratio of bacterial killing (B†) was determined to evaluate the antimicrobial efficacy of PACK-CXL.

RESULTS

B† was similar (51 ± 11%) in bacterial concentrations between 10 and 10 per ml. In 11 μL volume, Staphylococcus aureus (SA) 8325-4 ATCC 29213, Bacillus subtilis (BS) 212901, and Pseudomonas aeruginosa (PA) 2016-866624 were most sensitive to PACK-CXL at 5.4 J/cm (on average B† = 49 ± 8%), whereas Klebsiella oxytoca (KO) 2016-86624 (B† = 25%) was least sensitive. When irradiating a larger volume, B† was on average lower in 40 μL (19 ± 18%), compared with 11 μL (45 ± 17%, P < 0.001). By contrast, applying a higher UV fluence increased B† of SA ATCC 29213, from 50% at 5.4 J/cm to 92% at 10.8 J/cm, to 100% at 16.2 J/cm and above.

CONCLUSIONS

Applying higher UV fluences substantially increases the bacterial killing rates. Safety limits for clinical application require further investigation.

Depth- and direction-dependent changes in solute transport following cross-linking with riboflavin and UVA light in ex vivo porcine cornea

Diffusion is an important mechanism of transport for nutrients and drugs throughout the avascular corneal stroma. The purpose of this study was to investigate the depth- and direction-dependent changes in stromal transport properties and their relationship to changes in collagen structure following ultraviolet A (UVA)-riboflavin induced corneal collagen cross-linking (CXL). After cross-linking in ex vivo porcine eyes, fluorescence recovery after photobleaching (FRAP) was performed to measure fluorescein diffusion in the nasal-temporal (NT) and anterior-posterior (AP) directions at corneal depths of 100, 200, and 300 μm. Second harmonic generation (SHG) imaging was also performed at these three corneal depths to quantify fiber alignment. For additional confirmation, an electrical conductivity method was employed to quantify ion permeability in the AP direction in corneal buttons and immunohistochemistry (IHC) was used to image collagen structure. Cross-linked corneas were compared to a control treatment that received the riboflavin solution without UVA light (SHAM). The results of FRAP revealed that fluorescein diffusivity decreased from 23.39 ± 11.60 μm²/s in the SHAM group to 19.87 ± 10.10 μm²/s in the CXL group. This change was dependent on depth and direction: the decrease was more pronounced in the 100 μm depth (P = 0.0005) and AP direction (P = 0.001) when compared to the effect in deeper locations and in the NT direction, respectively. Conductivity experiments confirmed a decrease in solute transport in the AP direction (P < 0.0001). FRAP also detected diffusional anisotropy in the porcine cornea: the fluorescein diffusivity in the NT direction was higher than the diffusivity in the AP direction. This anisotropy was increased following CXL treatment. Both SHG and IHC revealed a qualitative decrease in collagen crimping following CXL. Analysis of SHG images revealed an increase in coherency in the anterior 200 μm of CXL treated corneas when compared to SHAM treated corneas (P < 0.01). In conclusion, CXL results in a decrease in stromal solute transport, and this decrease is concentrated in the most anterior region and AP direction. Solute transport in the porcine cornea is anisotropic, and an increase in anisotropy with CXL may be explained by a decrease in collagen crimping.

Corneal cross-linking (CXL)-A clinical study to evaluate CXL as a treatment in comparison with medical treatment for ulcerative keratitis in horses

OBJECTIVE

Compare CXL treatment with medical treatment alone in horses with stromal, ulcerative keratitis.

ANIMALS STUDIED

24 horses (24 eyes) with stromal, ulcerative keratitis were included.

PROCEDURE

12 horses were initially treated with CXL, and 12 horses were given conventional medical treatment. Topical medical treatment was added to horses in the CXL group if necessary. Parameters including cytology, microbial growth, time to fluorescein negativity, and time to inhibition of stromal melting were evaluated.

RESULTS

After the first day of treatments, a decrease in inflammatory signs and pain from the eye was observed in both groups. Stromal melting ceased within 24 hours regardless of treatment. CXL treatment alone was sufficient in 3 horses with noninfectious, superficial stromal ulcerations. Clinical signs of impaired wound healing were seen after 3-14 days in corneas with suspected or proven bacterial infection treated with CXL only, most likely because of insufficient elimination of bacteria deeper in the corneal stroma or because of re-infection from bacteria in the conjunctiva. The average decrease in stromal ulcer area per day after onset of treatment was almost identical between the groups, and no significant difference in time to fluorescein negativity was found.

CONCLUSIONS

We consider CXL a possible useful adjunct treatment of corneal stromal ulcers in horses, especially for melting ulcers and as a potential alternative to prophylactic antibiotic treatment for noninfected stromal ulcers. However, CXL should not be used alone for infected or suspected infected stromal ulcers, because topical antibiotics were required in all horses with proven infectious keratitis.

Long-term outcomes of riboflavin photodynamic antimicrobial therapy as a treatment for infectious keratitis

Purpose

To report the long-term outcomes of three patients with infectious keratitis treated with riboflavin photodynamic antimicrobial therapy (PDAT).

Observations

Case series reporting three patients with infectious keratitis unresponsive to standard medical treatment who underwent riboflavin photodynamic antimicrobial therapy (PDAT) as an adjunct therapy. One male and two female patients were treated, the median age of presentation was 58 years (range, 29-79 years). The organisms isolated and treated were Pseudomonas aeruginosa, Mycobacterium chenolae, and Curvularia spp. Different risk factors to develop corneal infection ulcers were identified, including corneal abrasion in a contact lens user, history of penetrating keratoplasty with chronic use of topical corticosteroids, and organic trauma. The median follow-up was 47 months (range 37-54 months), and there were no complications secondary to riboflavin PDAT treatment. Two cases underwent optical penetrating keratoplasty after infection was resolved and ocular surface was quiet for at least 3 years.

Conclusions and importance

Riboflavin PDAT can be used as an adjunct treatment in infectious keratitis to strengthen the corneal collagen fibers, delay keratolysis, and allow more time for antimicrobials to work and this way prevent a corneal perforation.

Comparison of corneal collagen cross-linking (PACK-CXL) and voriconazole treatments in experimental fungal keratitis

PURPOSE

To compare the antifungal efficacy of corneal collagen cross-linking with photoactivated riboflavin (PACK-CXL) and voriconazole in experimental Fusarium solani and Candida albicans keratitis models.

METHODS

Sixty-four corneas of 32 New Zealand rabbits were included and divided into two main groups. Intrastromal injection of Fusarium and Candida suspensions was performed, and it was observed that keratitis was formed on the third day. Both groups were randomly separated into the following four groups: control, PACK-CXL, voriconazole and PACK-CXL combined with voriconazole. PACK-CXL was applied using 0.25% riboflavin in an accelerated Dresden protocol (total ultraviolet A dose 5.4 J/cm²). Voriconazole was applied topically as 7x1/day with a dose of 1% (10 mg/ml). Corneal buttons were excised on the tenth day, and microbiological and pathological examinations were performed.

RESULTS

The PACK-CXL and PACK-CXL combined with voriconazole groups each had 100 colony-forming unit (CFU/ml) of reproduced micro-organisms compared with 500 CFU/ml in the voriconazole group and 1500 CFU/ml in the control group (p < 0.001) in the Fusarium keratitis model. The PACK-CXL combined with voriconazole group had 100 CFU/ml, the PACK-CXL group had 150 CFU/ml, and the voriconazole group had 200 CFU/ml of reproduced micro-organisms compared with 4000 CFU/ml in the control group (p < 0.002) in the Candida keratitis model. (p < 0.001). Fewer hyphae and non-specific stromal changes were observed in the pathological cross sections examined in subgroups that used CXL.

CONCLUSION

There was less fungus reproduction and a lower keratitis score for Fusarium solani and Candida albicans in the treatment groups compared to the control groups, especially in groups that used PACK-CXL. These results suggest that it is useful to combine PACK-CXL treatment with medical treatment in the fungal keratitis algorithm at the early stage of the disease.

Collagen Cross-linking for Microbial Keratitis

Collagen cross-linking is gaining popularity not only for arresting the progression of keratoconus but also other indications including management of corneal infections. In this review article, we analyzed the published literature to understand the level of evidence for its use in corneal ulcer. Photoactivated riboflavin and ultraviolet A light are known to possess antimicrobial properties. The treatment also induces formation of inter- and intra-fibrillar bonds, thereby making the corneal collagen resistant to the action of proteases arresting stromal melt. Both properties are well documented in in vitro experiments. The antimicrobial action is seen against bacteria, fungi, and parasites. The animal experiments have documented its efficacy against bacterial and fungal keratitis models. The literature on its application in human corneal infection is highly variable and comprises case reports, case series, and comparative nonrandomized and randomized trials. The treatment has been used as primary treatment, adjunctive treatment along with antibiotics, as the first line of treatment as well as for failed medical treatment cases. Even the cases included are of variable severity caused by a variety of microorganisms including culture-negative cases. Furthermore, the treatment protocols are also variable. While most reports show beneficial effects for bacterial corneal ulcer cases, especially those with superficial infiltrate, the effect has been mixed for fungal and parasitic keratitis. In view of these characteristics, we infer that the level of evidence for its use in corneal ulcer is at most weak. We need well-characterized, high-quality, clinical trials of sufficient power to assess its true value.