In vitro and in vivo datasets of topically applied ketorolac tromethamine in aqueous humor using Raman spectroscopy

This article includes datasets acquired by Raman spectroscopy from in vivo and in vitro ocular samples collected from the dataset from Bertens and Zhang et al., “Confocal Raman spectroscopy: Evaluation of a non-invasive technique for the detection of topically applied ketorolac tromethamine in vitro and in vivo” (Bertens and Zhang, et al.). Detection of ketorolac tromethamine in pig eyes was performed in vitro and rabbit eyes in vivo. Extracted aqueous humor samples from pig and rabbit eyes were measured in vitro using a cuvette. This manuscript shows the spectral Raman data without pre-treatment or analysis from ocular tissues and provides further information towards aqueous humor research via alternative data processing methods. Furthermore, the raw data enclosed may be used for future aqueous humor investigations and pharmaceutical research.

enclosed may be used for future aqueous humor investigations and pharmaceutical research.
© 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons. org/licenses/by/4.0/).

Data
The data contains unanalyzed Raman spectra obtained from pig eyes (in vitro) (1.1, see supplementary files folder 'in vitro pig eyes' and 'in vitro cuvettes, aqueous humor from pig eyes'), rabbit eyes (in vivo) (1.2, see supplementary files folder 'in vivo rabbit eyes'), and aqueous humor samples (in vitro, see supplementary files folder 'in vitro cuvettes, aqueous humor from rabbit eyes') (1.3). Based on the differences of the samples, three types of set-ups were used on each dataset. For pig eye measurements in vitro, a long-working-distance microscope objective lens (Jena lens alone or a Gonio lens combined with an f60 lens) was utilized (see supplementary files 'in vitro pig eyes' folder 'jena lens' or 'gonio'). For the rabbit eyes measurements in vivo, a Gonio lens combine with a f60 lens was used. For cuvettes measurements, an f80 lens was used when the sample was measured in a Brand® cuvette [2]. For each experimental set-up, the fingerprint-wavenumber region (patterns specific for a drug-molecule, ranging from 350 cm À1 to 1800 cm À1 ) and the high-wavenumber region (higher energy shifted, ranging from 2500 cm À1 to 4000 cm À1 ) were included. The fingerprint spectra Specifications Table   Subject Pharmacology In vitro measurements, the enucleated eyes were immersed in ketorolac containing fluid overnight, where after they were fixed in a holder and mounted on an adjustable lens mount. In vivo measurement, the rabbits were fixed and anesthetized respectively. In cuvette measurement, the sample were load in a cuvette and fixed with a holder. Value of the Data The dataset could be used for further composition analysis of the aqueous humor and for future pharmaceutical research, to increase sensitivity of Raman systems. The dataset can be useful for researchers who are interested in the aqueous humor composition, ocular pharmaceutics, Raman spectroscopy, and software engineers. Alternative processing methods could be applied to exact other compounds in the aqueous humor or to enhance signals. This dataset offers a large cohort of animals measured on both eyes, 5 times.
dataset was used for detection of intraocular ketorolac tromethamine as described in the article of Bertens and Zhang et al. [1]. Several peaks could be identified in the fingerprint region spectrum of a ketorolac tromethamine sample (Fig. 1a). Only major peaks specific for ketorolac tromethamine were selected. Those peaks are assigned to certain chemical bonds or vibration modes. The assignment of the ketorolac related peaks is presented in Table 1 [1]. Due to the spectrometer's spectral resolution (2 cm À1 ), the peak observed at 1586 cm À1 is assigned to NH 2 deformation [3]. The peak of 1524 cm À1 is assigned to in-plane vibrations of the conjugated eC]C-. The observed peak at 1472 cm À1 is assigned to C]N stretching and the peak at 1282 cm À1 is assigned to CH 2 wagging vibrations. Because Raman spectrum of the cornea, aqueous humor, and lens show different patterns in the high-wavenumber region, spectra from this region could be used as guide for location determination in the ocular tissue ( Fig. 1b) [4e6].

In vitro, dataset
Pig eyes (enucleated) were immersed in the dark at 4 C for 24 hours in vitro in different concentrations of ketorolac solutions (0.05%e5.0%) before the measurements (see supplementary files folder 'in vitro pig eyes'). For each concentration, three eyes were measured by Raman spectroscopy. An example spectrum obtained from a pig eye is shown in Fig. 2. The location in the eye was determined using the high-wavenumber spectra (Fig. 2b).

In vivo dataset
New-Zealand white rabbits received 50 mL Acular® three times a day in their right eye. At the same time, they received a drop of buffered saline solution (BSS) in their left eye as a control (see supplementary files folder 'in vivo rabbit eyes'). The measurement parameters of the Raman system were optimized using the first four rabbits. Different integration times (10, 15, or 30 seconds) were measured to acquire the optimum Raman signal. The following measurements were performed using an integration time of 30 seconds. During these measurements, hardware influences were observed. Further optimization of the processing method can be seen in Bertens and Zhang et al. [2]. The difference of the variant integration times can be found in Fig. 3, for example, the spectrum intensity at 400 cm À1 is from 74 A .U. with 10 second integration time (Fig. 3a), 127 A .U. with 15 second integration time (Fig. 3b) and 333 A .U. with 30 second integration time (Fig. 3c). Rabbits were measured according to the schedule in Table 2.

In vitro, cuvettes dataset
Immediately after intra-ocular Raman measurements (both in vitro & in vivo), 100 mLe150 mL of aqueous humor was drawn from the pig eyes, and 50 mL was drawn from the right eye of each rabbit.
The aqueous humor samples were frozen on dry ice and stored in a À80 C freezer until use. When used, the location of focus was determined with the high wavenumber spectra, as shown in Fig. 4.
Fingerprint spectra were collected to determine ketorolac concentrations in the aqueous humor. Spectrum examples of pig and rabbit aqueous humor are show in Fig. 5a and Fig. 5b, respectively (see supplementary files folder 'in vitro cuvettes'). Further background subtraction needs to be applied for analyses.

Raman spectroscopy system
Two diode lasers were utilized as an excitation light source for Raman spectroscopy: a 26 mW 785 nm laser (Innovative Photonic Solutions SM 785 nm, Monmouth Junction, NJ, US) or a 14 mW 671 nm laser (Laser Quantum Ignis 671 and SMD 6000, Konstanz, Germany). A high-performance Raman spectrometer module (model 2500, River Diagnostics®, Rotterdam, the Netherlands) was utilized for Raman spectra recordings [8]. A 25 mm diameter pinhole was integrated within the spectrometer for the confocal Ramans spectroscopy detection. An air-cooled charge-coupled device (CCD) camera with  operating temperature À60 C was integrated within the spectrometer for signal detection. The Raman spectrometer is capable of collecting Raman scattering wavenumber ranges in 350 cm À1 -1800 cm À1 and 2500 cm À1 -4000 cm À1 with 2 cm À1 spectral resolution. A diverged laser beam out of the spectrometer is converted to a collimation beam by a lens with focus length of 80 mm (f80). Depending on the measurement, the lens setup was adapted. The system was used in single point modus and location in the sample was determined using the high wave numbers (671 nm laser).

In vitro measurement of enucleated pig eyes
Fresh domestic pig (Sus Scrofa Domesticus) eyes were obtained from a local abattoir ("Slachthuis Kerkrade Holding", Kerkrade, the Netherlands). The enucleated eyes were transported to the laboratory on ice and used within 3 hours after enucleation. Before use, the pig eyes were inspected with a stereo microscope (Olympus SZX9, Tokyo, Japan). Only eyes with clear corneas without visible corneal damage were used in the experiment. The excess tissues of the eye were removed carefully where after the eyes were washed in phosphate buffered saline (PBS) (pH of 7.4). Meanwhile, ketorolac (MSN laboratories, Telangana, India) was dissolved in PBS creating concentrations of 0.05%, 0.1%, 0.125%, 0.25%, 0.5%, 1.0%, 1.25%, 2.5%, and 5.0%. The pig eyes were submerged in 15 mL of a diluted ketorolac solution. As negative control, PBS was used, and as positive control 0.5% ketorolac ophthalmic solution (Acular®, Allergan, Dublin, Ireland) was used as submerging solution. For each concentration, three eyes were used. Before the Raman measurements, pig eyes were stored in the dark at 4 C for 24 hours. Before measurements were taken, the eyes were inserted in a home-designed holder (Fig. 6).

In vivo measurement of the rabbit eyes
Twelve New-Zealand white rabbits (weight ranged from 2.0 kg to 2.5 kg upon arrival) were obtained from Envigo (Horst, the Netherlands). The rabbits were group housed with 6 animals per cage with males and females separated. The rabbits had ad libitum access to water and food. One week was given to acclimatize before rabbits were used in the experiments. The rabbits were treated with 50 mL Acular® in the lower conjunctival fornix of their right eye. The contralateral eyes were treated with 50 mL sterile buffered saline solution (BSS, B. Braun, Melsungen AG, Germany manufacturer). Both treatments were performed three times a day. Measurements were taken on day 0, day 7, day 14, day 21, and day 28. Four rabbits were used to optimize the system parameters as shown in Table 2.  Rabbits were measured using setup as shown in Fig. 7b. During the examinations, rabbits were anesthetized intramuscularly with ketamine (Alfasan, Woerden, the Netherlands) and midazolam (Actavis, Dublin, Ireland), 50 mg/kg and 5 mg/kg, respectively. Both eyes of the rabbit were measured by the Raman system. All measurements were performed at random, 1e3 hours after receiving the eye drops. Measurement was performed with 30 second exposure times using 2 frames. All animal procedures were conducted according to the ARVO Statement for the Use of Animals in Ophthalmic and Visual Research and the Guidelines of the Central Laboratory Animal Facility of Maastricht University. All protocols were approved by the Central Committee for Animal research and were in accordance with the European Guidelines (2010/63/EU).

In vitro measurement of the aqueous humor
For cuvette detection, 50 mLe150 mL aqueous humor was obtained from an anterior chamber paracentesis from the eyes using an insulin syringe (BD Micro-Fine™, Becton Dickinson, NJ, US). 50 mL was drawn from rabbit eyes after topical sedation (1 drop 0.4% Oxybuprocaine hydrochloride solution (Bausch & Lomb Pharma, Brussels, Belgium)), 100 mLe150 mL was drawn from the pig eyes. As a negative control, 100 mL aqueous humor was drawn from seven healthy control rabbits within 10 minutes after sacrifice, no topical treatment nor anaesthetics were used.
All aqueous humor samples were frozen on dry ice immediately after sampling and stored in a À80 C freezer until measurements. Samples were measured using an f80 lens in front of the sample container (Fig. 8). The sample was measured for 3 frames in a disposable cuvette (#7592-00, Sigma-Aldrich, MO, US) with 60 seconds per frame.