This study was conducted in compliance with the Declaration of Helsinki of 1964. Protocol approval for the conduct of this study was obtained from the Medical Corporation Heishinkai OPHAC Hospital Institutional Review Board. All volunteers provided written informed consent for the use of their blood in the study.

Participants in this study were healthy Japanese male (n = 8) and female (n = 10) volunteers aged between 20 and 40 years. Main exclusion criteria included the following: (1) Meal intake within 10 h prior to blood collection; (2) Medication use within 1 week prior to blood collection; (3) Positive for infectious diseases (human immunodeficiency virus (HIV), syphilis, hepatitis B, and hepatitis C) within 6 months prior to blood collection; (4) Blood collection or donation of more than 200 mL within 1 month prior to blood collection, or more than 400 mL within 3 months prior to blood collection; (5) Other persons who are deemed unsuitable as participants by physicians. They were recruited through the blood collection volunteer enrollment process and provided blood samples (50 mL) at Medical Corporation Heishinkai OPHAC Hospital. The volunteers were not permitted to have taken any medication in the week prior to blood collection.

Whole blood was used for measuring both COX-1 and COX-2 activity [14]. No anticoagulant was used for the COX-1 assay, but blood was collected in heparin sodium-containing tubes for the COX-2 assay.

Five blood samples from each volunteer were evaluated for each drug, and the same blood was used for both the COX-1 and COX-2 assays on the same drug.

On the day of the study, five NSAIDs (diclofenac sodium [FUJIFILM Wako Pure Chemical], celecoxib [US Pharmacopeia], ibuprofen [FUJIFILM Wako Pure Chemical], flurbiprofen [MP Biomedicals, LLC], and etodolac [US Pharmacopeia]) were dissolved in dimethyl sulfoxide (DMSO), and lipopolysaccharide (LPS; from Escherichia coli O55:B5; Sigma Aldrich) and acetylsalicylic acid (FUJIFILM Wako Pure Chemical) were dissolved in physiological saline solution as controls. The preparation concentrations were as follows: diclofenac sodium (0.823–200 µg/mL), celecoxib (1.37–1000 µg/mL), ibuprofen (5.49–4000 µg/mL), flurbiprofen (2.74–2000 µg/mL), etodolac (4.12–3000 µg/mL), LPS (0.1 mg/mL), and acetylsalicylic acid (1 mg/mL). After addition to the blood samples, the final concentrations of each compound were as follows: diclofenac sodium (4.12–1000 ng/mL), celecoxib (6.86–5000 ng/mL), ibuprofen (27.4–20,000 ng/mL), flurbiprofen (13.7–10,000 ng/mL), etodolac (20.6–15,000 ng/mL), LPS (1 µg/mL), and acetylsalicylic acid (10 µg/mL).

For the COX-1 assay, 5 µL of each drug solution or DMSO was dispensed into sterile 1.5-mL tubes, and 1 mL of anticoagulant-free blood was added. Before the blood coagulated, the tubes were thoroughly mixed by inversion. After incubation at 37 °C for 1 h at rest, the tubes were centrifuged at 1600g for 10 min. Then, the supernatant was transferred to a separate sterile 1.5-mL tube and used as the sample for the COX-1 assay. To quantify the COX-1 inhibitory activity of each drug, thromboxane B2 (TXB2), which is produced by the action of COX-1 on arachidonic acid, was quantified in the COX-1 assay sample with an enzyme-linked immunosorbent assay (ELISA) kit (Enzo Biochem Inc.) according to the manufacturer’s instructions.

For the COX-2 assay, COX-1 was first inactivated by dispensing 10 µL of acetylsalicylic acid solution into sterile 1.5-mL tubes and adding 1 mL of the blood with heparin. The tubes were thoroughly mixed by inversion and incubated statically at 37 °C for 15 min. Subsequently, the following were added to the tubes: (1) 10 µL of LPS solution and 5 µL of each drug solution, or (2) 10 µL of saline solution and 5 µL of DMSO, or (3) 10 µL of LPS solution and 5 µL of DMSO. The tubes were mixed by inversion and incubated for 24 h at 37 °C with inverted agitation. Then, the tubes were centrifuged at 1600g for 10 min, and the supernatant was transferred to a separate sterile 1.5-mL tube to be used for the COX-2 assay. To quantify the COX-2 inhibitory activity of each drug, prostaglandin E2 (PGE2), which is produced by the action of COX-2 on arachidonic acid, was quantified with an ELISA kit (Enzo Biochem Inc.) according to the manufacturer’s instructions.

The COX-1 and COX-2 inhibition rates were calculated with the following equations: COX-1 inhibition rate (%) = (1 − (A/B)) × 100, where A is the value of TXB2 measured in the samples with each drug solution and B is the value of TXB2 measured in samples to which DMSO was added; COX-2 inhibition rate (%) = (1 − (C − D)/(E − D)) × 100, where C is the value of PGE2 measured in the samples to which LPS and each drug solution were added, D is the value of PGE2 measured in samples to which physiological saline and DMSO were added, and E is the value of PGE2 measured in samples to which LPS and DMSO were added.

For each drug, the inhibition rate and whole blood drug concentration at the time of inhibition were used to generate a dose–response curve with Prism software (GraphPad Software, LLC) by using a four-parameter logistic regression model with two parameters fixed as the minimum inhibition rate (0%) and maximum inhibition rate (100%). The inhibition rate of each drug was expressed as the mean ± standard error (SE).

The results of the logistic regression analysis were used to calculate the COX-1 and COX-2 inhibition rates at the Cmax [15,16,17,18,19,20,21] of clinical doses of the marketed formulations of each drug. For diclofenac sodium, we also calculated the concentrations at which COX inhibition rates were 50% and 80% (IC50 and IC80).