Infectious keratitis is a sight-threatening infection of the cornea. Several pathological organisms can cause infectious keratitis, and its major risk factors include contact lens wear, trauma, and eyelid and ocular surface disorders (Al-Mujaini et al., 2009; Ting et al., 2022). Prompt treatment with topical broad-spectrum antimicrobial eye drops is the standard of care (Lin et al., 2018), and additional interventions such as corneal debridement may be used in some cases. Once the pathogen has been identified, more targeted therapy can be used according to the causative pathogen’s antimicrobial drug sensitivity profile (Barac et al., 2022). However, even with prompt treatment, outcomes are not always favorable due to various factors such as antibiotic resistance, low bacterial isolation rates, limited corneal drug penetration, plus several other predisposing risk factors (Knyazer et al., 2020).

Corneal cross-linking (CXL) is a minimally invasive procedure used to strengthen the cornea and today it is widely utilized for the treatment of ectatic corneal diseases. It is based on a photochemical reaction caused by the combination of a photoactivated chromophore (riboflavin, most commonly) and a photoactivation wavelength of light (in the case of riboflavin, is ultraviolet-A light at around 365–370 nm) (Price and Price, 2016; Kling et al., 2020).

In addition to its mechanical benefits in strengthening the cornea, photoactivation of chromophore has been shown to act like a disinfectant by various mechanisms (Hafezi and Bradley Randleman, 2014). As such, CXL was proposed as a novel treatment for infectious keratitis, termed “photoactivated chromophore for infectious keratitis—corneal cross-linking”, or PACK-CXL. It was first introduced as an adjuvant treatment to standard-of-care antimicrobial therapy (Idrus et al., 2019; Ting et al., 2019; Gulias-Cañizo et al., 2020; Knyazer et al., 2020; Achiron et al., 2021a; Saini et al., 2022), then later as a stand-alone treatment (Hafezi et al., 2022a; Kling et al., 2020; Knyazer et al., 2020; Hafezi et al., 2021).

Historically, most PACK-CXL treatments were given using the Dresden protocol settings developed for keratoconus, i.e., irradiation settings of 3 mW/cm² for 30 min to deliver a UV fluence of 5.4 J/cm² (Koppen et al., 2009; Müller et al., 2012; Price et al., 2012; Bamdad et al., 2015; Prajna et al., 2020). Accelerated protocols, employing a power density of 30 mW/cm² for a duration of 3 min (equivalent to a total energy dose of 5.4 J/cm²), have also been implemented (Richoz et al., 2014; Achiron et al., 2021b). However, because corneal ulcers are opaque, the depth of microbial invasion can be considerably deeper than the approximate 330 µm depth of cross-linking effect achieved by the standard 5.4 J/cm² applied in the Dresden protocol cross-linking (Hafezi et al., 2020). Hence, we believe that the pathogen-killing efficacy of PACK-CXL can be improved by increasing the total irradiation fluence. In a recent in vitro study, Kling et al. (Kling et al., 2020) tested the bacteria-killing effect of irradiation different intensities (ranging from 3 to 18 mW/cm²) and different fluences (ranging from 5.4 to 27 J/cm²) on six different bacterial strains. Their study showed that for Staphylococcus aureus, the ratio of bacterial killing was 50% at 5.4 J/cm² and increased to 92% at 10.8 J/cm² and 100% at 16.2 J/cm² and above (Kling et al., 2020). More in vitro studies support this result (Lu et al., 2023). In vivo studies examining high fluences on infectious keratitis, however, are scarce (Nateghi Pettersson et al., 2019; Knyazer et al., 2021; Hafezi et al., 2022a; Awad et al., 2022; Hafezi et al., 2022b; Marrie et al., 2023), mostly on animal models or human case reports. Our first aim is to show the therapeutic benefit of high fluence PACK-CXL in a case series.

UV irradiation above a certain threshold can damage the corneal endothelium. Although riboflavin shields the endothelium from UV energy, it gets consumed during the process, so there is a limit to the amount of UV energy that can be safely delivered during corneal cross-linking procedures. The safety of high fluence UV irradiation protocols was demonstrated by Mazzotta et al. (2016) who studied them in patients with keratoconus, where they found that delivering high fluences (7.2, 10, and 15 J/cm²) was as safe as both conventional (Dresden protocol) and accelerated CXL procedures. Our second aim in this study is to examine the safety of high fluence CXL protocols by assessing complications such as corneal melting of corneal perforation.

Although in practice PACK-CXL is not yet as widely adopted or established as traditional corneal cross-linking, this innovative approach holds great potential in augmenting the standard treatment options for infectious keratitis, leading to improved outcomes and preserving visual acuity for affected individuals.

The data presented here is, to our knowledge, the first trial testing the therapeutic effect of PACK-CXL at the slit lamp on patients with infectious keratitis using a high fluence of 10.0 J/cm² rather than the typical fluence of 5.4 J/cm².

This prospective interventional, nonrandomized cohort study was approved by the Institutional Review Board (IRB) of the Ben-Gurion University of the Negev, Israel (0486-20-SOR), and adhered to the tenets of the Declaration of Helsinki. All patients gave written informed consent prior to inclusion into this study.

This study reports on the outcomes of 20 eyes of 20 subjects who underwent high fluence PACK-CXL treatment for infectious keratitis at the slit lamp.

We found the procedure to be effective, with re-epithelialization occurs at an average of 8.2 ± 2.8 days, the longest period being 14 days. Based on the observed epithelial closure within 14 days, it is consistent with a non-persistent corneal ulcer (HypothesisVaidyanathan et al., 2019). As a result, we can confidently conclude that PACK-CXL as an adjuvant therapy has been successful. Our results also show that it is a safe procedure, with no complications observed during the study, despite large ulcers and a severe degree of infection (Figure 2).

PACK-CXL is thought to act through three mechanisms (Price and Price, 2016): First, the chromophore’s planar ring structure intercalates between the DNA bases, thus limiting microbial reproduction; Second, reactive oxygen species are released, directly destroying pathogen cell walls; Finally, activated chromophores form covalent connections with the corneal stroma. This alters the collagen fibers' tertiary structure, reducing the number of collagenase binding sites and making them more resistant to enzymatic destruction. Importantly, these effects are unaffected by the level of pathogen antimicrobial drug resistance, which may be of particular value in, for example, cases of bacterial keratitis caused by highly antibiotic resistant strains.

In 2000, PACK-CXL treatment for melting corneal ulcers was used for the first time in a clinical setting. A single UV-radiation with two ultraviolet diodes and a wavelength of 370 nm with an energy of 2.5 mW/cm² was administered to four patients suffering from melting ulcers of the cornea of various origins. The melting of the cornea ceased in three of the four individuals after treatment. A surgical operation could be postponed, at least temporarily (Irradiation of cornea, 2000).

Several protocols have been tried since then to optimize the effect, including the Dresden protocol developed for corneal ectasia. We have speculated that more energy is required to remove the infection effectively, particularly in deeper ulcers (Hafezi et al., 2020; Kling et al., 2020).

Our results are in line with previous studies. Awad et al. used rabbit models to test the efficiency of different PACK-CXL protocols on fungal keratitis, vs. Voriconazole as control. They found all protocols including the standard Voriconazole treatment to be effective with no statistically significant difference between them (Awad et al., 2022). Hafezi et al. compared antimicrobial therapy and PACK-CXL using of fluences of either 5.4 J/cm² or 7.2 J/cm² in a phase 3 trial that involved patients with infectious keratitis (Hafezi et al., 2022a). They found no significant difference between the groups in epithelial healing time. Of note, they used PACK-CXL as a stand-alone treatment, while in our study it was used as adjuvant to the standard of care. Pettersson et al. reported a case of a 24 years old woman with a severe Acanthamoeba keratitis, who was dramatically deteriorating despite an intensive drug regimen with adjuvant Dresden protocol PACK-CXL(24). The woman eventually got better only after high fluence PACK-CXL of 7.2 J/cm², after which no further anti-amoebic treatment was needed. Another interesting case report of fungal keratitis was authored by Hafezi et al. (Hafezi et al., 2022b). They reported a 79 years old man who had a severe ulcer which did not improve with an aggressive intra-venous and topical anti-fungal treatment. He was then received one treatment of PACK-CXL with total fluence of 7.2 J/cm² after which he got only slightly better. A week later it was decided to perform another high fluence PACK-CXL treatment, ultimately reaching a level of 14.4 J/cm², fractionated across 2 doses. Substantial improvement was observed starting at day 1 after the second treatment. Knyazer et al. reported even greater levels of fluence given to their patient suffering from Pseudomonas aeruginosa ulcer (Knyazer et al., 2021). After further deterioration was noted in spite of standard topical treatment, the patient had been through two treatments of high fluence PACK-CXL, ultimately ending up receiving 21.6 j/cm². 2 days after the last treatment, marked improvement was noted with final epithelial closure on day 6 after the procedure. The data presented in our study is, to our knowledge, the first pilot testing the therapeutic effect of high fluence PACK-CXL as an adjuvant therapy on a cohort of patients with infectious keratitis from various pathogens.

Another significant aspect of this study is that the PACK-CXL treatment was administered at the slit lamp. Cross-linking is typically a procedure performed with the patient in the supine position in the operating rooms (Hafezi et al., 2021), because irradiation time in the Dresden protocol is 30 min. However, In the setting of shorted irradiation time, such as in our protocol, the procedure can be performed at the slit lamp, allowing a more convenient, efficient, and accessible option for patients, while also maintaining the treatment’s effectiveness and safety.

The limitations of this study include the lack of a control group, short follow-up, and the small number of included patients. These limitations limit our ability to compare high fluence PACK-CXL to the standard of care and to other PACK-CXL protocols. However, we intended to show effectiveness and safety, superiority of the treatment was not one of the aims of this study. A large-scale, randomized, controlled study would be warranted to further strengthen our results and to assess the benefits of this protocol over the standard of care and over the use of lower fluence CXL protocols.

In conclusion, our study contributes valuable evidence to the growing body of knowledge that supports the advantages and efficacy of high fluence PACK-CXL. As infectious keratitis is a sight-threatening condition with often poor outcomes, we believe that the intensified treatment protocol with increased energy and exposure time in high fluence CXL has the potential to aid in the process of healing, providing a valuable therapeutic approach for this condition.