Dosimetry for Rose Bengal Photodynamic Antimicrobial Therapy (RB-PDAT) for treatment of infectious keratitis

Abstract

Purpose : Rose Bengal Photodynamic Antimicrobial Therapy (RB-PDAT) is an effective treatment for infectious keratitis. RB-PDAT works by activating the photosensitizer Rose Bengal (RB) with green light to produce antimicrobial singlet oxygen (¹O₂). The purpose of this project was to demonstrate the proof-of-principle of a system that measures the concentration of ¹O₂ generated during the treatment. The system will enable control of the RB-PDAT treatment dosage.

Methods : A 520nm, 4.5 mW diode laser was filtered to remove IR emission and focused with a 20x objective (in vitro setup) or f = 15mm lens (ex vivo setup) on the sample (Fig 1). The luminescence was collected in transmission mode for in vitro experiments and in reflection mode for ex vivo experiments using an InGaAs photoreceiver. A bandpass filter centered at 1277nm and a longpass filter with a cutoff at 1150nm are placed in the detection path to select the ¹O₂ luminescence emission. We stimulated varying concentrations (0-0.1% or 1.0 mM) of RB in H₂O or in Freon 113 (0-1.6 µM, boosts ¹O₂ signal) in a 2mm quartz cuvette with the 520 nm laser. To confirm the collected signal was from ¹O₂, furfuryl alcohol (FFA) was used for ¹O₂ quenching (Fig 2). The system was tested ex vivo using 3 human donor corneas treated with 0.1% (1.0 mM) RB in H₂O for 30 minutes and stimulated with the 520 nm laser. All measurements were performed in triplicate. All results scaled to average max signal.

Results : ¹O₂ luminescence signal from RB in Freon 113 was approximately 8.6x higher than RB in H₂O at 1 µM, meaning the readout was from a boosted ¹O₂ signal. Adding 5 µl of ¹O₂ quencher, FFA, the signal for 1.61 µM (saturated) RB in Freon 113 decreased to the same signal level as pure Freon 113 (100±11.4% to 3.71±0.46%, relative units), signifying that 96.3% of the signal came from ¹O₂ (Fig 2A). For 0.01% and 0.1% RB in H₂O (0.10 and 1.0 mM), addition of 35 µl FFA lowered the signal from 17.7±2.6% and 15.8±0.65% to 6.60±1.6% and 6.78±0.37% (Fig 2B), meaning 62% and 57% of the signal was from ¹O₂. Collected signal from 1 mM RB in H₂O treated cornea was 6.62±1.0%, showing signal from cornea is possible. Units scaled to 100% ≈ 5.8 × 10¹¹¹O₂ molecules.

Conclusions : We provided a proof-of-concept for ¹O₂ luminescence dosimetry of RB-PDAT. The system can be used to create a predictive model that generates optimal RB-PDAT treatment dosage based on ¹O₂ measurement.

This is a 2020 ARVO Annual Meeting abstract.