Metabolism and Drug-Drug Interaction Studies of Metamizole in Humans

Metamizole is a non-opioid analgesic with additional antipyretic and spasmolytic properties. It was introduced nearly 100 years ago, but due to a severe but rare side-effect agranulocytosis, it has been withdrawn from the market in several countries. Metamizole is a prodrug, which is spontaneously hydrolyzed in the gastrointestinal tract to 4-methylaminoantipyrine (4-MAA). 4-MAA is rapidly and almost completely absorbed into circulation. The main fraction is oxidized to 4-formylaminoantipyrine (4-FAA) or demethylated to 4-aminoantipyrine (4-AA), which is then acetylated to 4-acetylaminoantipyrine. Other metabolites have been described, but these 4 main metabolites reflect about 65-70% of the administered metamizole dose. The two metabolites 4-MAA and 4-AA exhibit an analgesic effect, while the end metabolites 4-AAA and 4-FAA are considered inactive. Although metamizole has been available for nearly a century, there are still knowledge gaps such as the pharmacological mode of action. Furthermore, the enzymatic system responsible for the demethylation and the oxidation of 4-MAA has not been discovered yet. Participation of cytochrome P450 (CYP) has been suspected, but the specific isoforms have not been identified so far. Additionally, there is evidence that metamizole induces both CYP2B6 and CYP3A4 isoforms, the mechanism however has not been shown yet.
A main focus of my thesis was to study the pharmacokinetics and metabolism of metamizole in clinical trials as well as in vitro experiments. These studies required reliable quantification of the metamizole main metabolites. In the first project, we therefore developed and validated a high-performance liquid chromatography tandem mass spectrometer (HPLC-MS/MS) method for the quantification of 4-MAA, 4-AA, 4-AAA, and 4-FAA in human plasma samples. A pentafluorophenyl analytical column resulted in a suitable retention of the polar metabolites by interacting with their aromatic structure. A gradient program with pre- and post-column infusion was applied for the retention and separation of the hydrophilic analytes. Formic acid was removed from the solvents and reintroduced after the chromatography to achieve sufficient ionization. The method was linear over the calibration range (R > 0.99), with an accuracy of 91.3% - 106.0% and an imprecision of <12.7%. Thus, the presented method fulfilled the criteria described in the F.D.A. guidelines for bioanalytical method validation for industry.
The second project aimed to identify the enzymes responsible for the demethylation of 4-MAA. Incubation assays of 4-MAA in human liver microsomes combined with selective CYP inhibitors
revealed that CYP2B6, CYP2C8, CYP2C9 and CYP3A4 contribute to the biotransformation of 4-MAA to 4-AA. These results were confirmed by incubation of 4-MAA in human liver homogenate and HepaRG cells. Nonetheless, the formation of 4-AA was low and did not reflect the metabolic capacity observed of a human. Thus, other enzymatic systems were considered. Incubation assays with various peroxidases demonstrated that human myeloperoxidase was also capable to demethylate 4-MAA in presence of hydrogen peroxide. Furthermore, the determined affinity of the enzyme for the demethylation was in the range of pharmacologically relevant plasma concentrations. Further experiments in human promyelocytic leukemia cells expressing myeloperoxidase confirmed the demethylation of 4-MAA in presence of hydrogen peroxide. Since there is a high quantity of myeloperoxidase expressing cells (promyelocytes, myelocytes, metamyelocytes and granulocytes) located in the bone marrow, we suggested that the demethylation of 4-MAA may take also place in the bone marrow.
In a third project, plasma samples from a pediatric, pharmacokinetic study with metamizole were analyzed. Children were treated with 10 mg/kg metamizole (intravenous application, i.v.), blood samples drawn over various time points to evaluate the dose-exposure relationship in different age cohorts. The bioanalysis of the plasma samples was performed in our laboratory, while the clinical study was conducted at the Children’s University Hospital Basel. Pharmacokinetic parameters of the metamizole metabolites were assessed by non-compartmental and population pharmacokinetic analysis. Children younger between 2-6 years showed lower exposure of 4-MAA compared to adults when treated with the same weight adjusted dose, while children < 2 year exhibited a similar exposure. Interestingly, children younger than one year displayed a higher exposure, which was probably due to unmatured metabolic capacity. The conclusion was a reduced, weight-based i.v. dose recommendation for children < 1 year compared to older children and adults (5 mg/kg vs. 10-20 mg/kg).
The last project examined the influence of metamizole treatment on the activity of various CYP isoforms. In a clinical study, the “Basel Cocktail” phenotyping approach was applied to assess the CYP1A2, CYP2B6, CYP2C9, CYP2C19, CYP2D6 and CYP3A4 activity before and after metamizole intake. The probands received 3 grams of metamizole for 7 consecutive days. We observed an induction of CYP2B6, CYP2C19 and CYP3A4 and an inhibition of CYP1A2. CYP2C9 and CYP2D6 activities were unaffected. The mechanism of induction was investigated in different HepaRG cell lines. Knock-out of the pregnane X receptor and control cells exhibited a significant mRNA upregulation of CYP2B6, CYP2C9, CYP2C19 and CYP3A4. In contrast, knock-out of the constitutive androstane receptor resulted in the abolishment of the induction. Consequently, the constitutive androstane receptor is essential for the metamizole derived induction.
In conclusion, we demonstrated that various CYP isoforms and human myeloperoxidase are able to mediate the demethylation of 4-MAA. The main responsible enzyme however has not been identified yet. In the PK study in children, we observed an overexposure of 4-MAA in children >1 year when treated with the recommended i.v. dose. Furthermore, we were able to show the impact of metamizole on the activity of various CYP isoforms. The co-administration of CYP1A2, CYP2B6, CYP2C19 and CYP3A4 substrates and metamizole may lead to altered drug-exposure and may therefore promote adverse effects.