Quantitative Drug Metabolite Profiling without Radiolabels Using HPLC–ICP-MS

In drug development, quantitative determination of a candidate drug and its metabolites in biofluids is an important step. The standard technique for quantitative metabolite profiling is radiolabeling followed by high performance liquid chromatography (HPLC) with radiodetection, but there are disadvantages to this approach, including cost and time, as well as safety and ethical concerns related to administering radiolabeled compounds to humans. Frank Vanhaecke and his research group at Ghent University have been developing an alternative technique, and he recently spoke to us about this work. Vanhaecke is the 2017 recipient of the Lester W. Strock award, which will be presented to him at the 2017 SciX conference. This interview is part of a series of interviews with the winners of awards presented at SciX.

You have been developing new methods for quantitative metabolite profiling in human plasma for use during drug development. First, why are new methods needed?

We have been active in this research field for a couple of years now. Several years ago, we were approached by a pharmaceutical company. The company had developed a new drug against tuberculosis (TB). The entire process of drug development is very complex and includes ADME studies, in which attention is devoted to characterizing absorption, distribution, metabolism, and excretion of the newly developed active pharmaceutical ingredient (API). For identification of the metabolites formed, high performance liquid chromatography combined with electrospray ionization mass spectrometry (HPLC–ESI-MS) is typically relied on, but for their quantitative determination, this technique usually cannot be deployed because the sensitivity is dependent on the molecular structure and typically, a standard is not available for every metabolite of the drug. As a result, a version of the drug molecule carrying a radiolabel (¹⁴C or ³H) is used for this purpose. Use of this radiolabeled version of a drug allows straightforward quantification of all metabolites because the use of HPLC–radiodetection is highly selective (specific) and provides a response independent of the molecular structure.

This specific anti-TB drug, however, showed a long residence time in the lungs, such that in vivo studies with human volunteers could not be approached in this traditional way. Therefore, the pharmaceutical company contacted us for the development of a “cold” method (without the use of a radiotracer). This request was the kickoff of this research line within my group. Because the API contained a Br-atom, we developed an approach based on HPLC with inductively coupled plasma–mass spectrometry (HPLC–ICP-MS) (1). We realized that such methods could also prove very valuable beyond the scope of this specific drug, because the synthesis of a radiolabeled version of the API can be challenging (often the synthesis needs to be started using different building blocks) as well as time-consuming and expensive. The administration of radioactive compounds to humans might also raise ethical concerns and the samples and waste products need to be adequately handled.

Why did you feel that HPLC–ICP-MS was a good technique to address this problem?

Like radiodetection, ICP-MS also provides an instrumental response independent of the molecular structure in which the nuclide monitored is present. As a result, quantification doesn’t require the availability of a standard for each and every metabolite. Moreover, ICP-MS is a very sensitive and element-specific technique.

You had to optimize the ICP-MS detection of Cl, Br, and I. What challenges did you face in that step?

Indeed, we first addressed pharmaceutical drugs containing an element that can be measured via ICP-MS. These elements include the halogens Cl, Br, and I. All of these elements are characterized by an ionization efficiency that is only modest at best. For Cl, the estimated ionization efficiency is even ≤1%. Moreover, some of these elements also suffer from spectral overlap. Present-day ICP-MS instruments, however, typically offer sufficient sensitivity and if this is not the case in a given context, we opt for preconcentration of the API and its metabolites on a precolumn. For overcoming the problem of spectral interference, we typically use chemical resolution. We are especially charmed by the capabilities of tandem ICP-MS in this context. In this setup, a collision–reaction cell is located between two quadrupole filters. The double mass selection offers a considerably enhanced control over the processes going on in the collision-reaction cell. We often rely on the use of reaction product ions formed upon selective interaction between the target ions and the reaction gas molecules. These reaction product ions can be measured interference-free at another mass-to-charge ratio that has been “freed” from other signals by the first quadrupole in the tandem setup. 

When using gradient elution in the HPLC separation (as is often necessary), accurate quantification is also a challenge. The continuously changing mobile-phase composition brings about a continuous change in the instrumental sensitivity. In our first application (Br determination), this problem was circumvented by using on-line species-unspecific isotope dilution. However, some target elements, such as P or I, are monoisotopic, while for other elements, such as S or Cl, it might not be feasible to find two nuclides, the signals of which can be monitored interference-free. Therefore, we have also investigated other approaches, such as the use of a counter-gradient or the development of a mathematical function describing how the sensitivity varies as a function of the mobile-phase composition. At least for Cl, we have seen that such a mathematical function could be derived on the basis of a set of flow injection measurements (2).

When the pharmaceutical drug does not contain an element that can be monitored using ICP-MS, you introduce such an element via derivatization. How do you select a derivatization agent?

For selecting a proper derivatization reagent, many aspects need to be taken into account. Of course, the first prerequisite is that we introduce an element into the API and all of its metabolites, rendering all of them “ICP-MS visible.” The derivatization reagent addresses a given reactive group of the API and its metabolites, and the reaction efficiency should be (close to) 100%. We also prefer a one-step derivatization method that can be carried out under sufficiently mild conditions (we prefer a reaction temperature below 80 °C and complete conversion in less than 1 h). The reagent used should be readily available and not very costly. We then search the literature for appropriate derivatization reagents and experimentally screen the promising ones to assess whether they could be fit for purpose.

What challenges do you face in optimizing the derivatization procedure?

The formation of reaction by-products is a nuisance. For example, in the derivatization of amino-group-containing compounds with tetrabromophthalic anhydride, we tried to accelerate the reaction by adding 4-dimethylaminopyridine (DMAP) as a catalyst, but this approach led to undesirable second-order derivatives (more than one derivative or reaction product for a single metabolite) due to the presence of a phenolic OH group in the target analytes. Also, solubility problems sometimes need to be addressed. For example, in the study mentioned above, the target analytes (levothyroxine and its metabolites) were not soluble in acetonitrile, which was considered the best solvent for the reaction. Finally, stability issues also need to be considered, because sometimes the analytes are thermally unstable. In the same study, degradation products were generated from levothyroxine at temperatures ≥ 40 °C.

In addition, endogenously present compounds also may become derivatized. As a result, the chromatographic separation approach should not only succeed in separating the API and its metabolites from one another, but also in separating the API from derivatized endogenous compounds or endogenous compounds naturally containing the element introduced into the API and its metabolites.

What results have you seen with your approach so far?

Well, first of all, we were very happy that the method we had developed for quantitative metabolite profiling of the Br-containing anti-TB drug (for which we were contacted by the pharmaceutical company) was successfully used in the context of an in vivo study with human volunteers (patients suffering from multidrug-resistant tuberculosis). Meanwhile, this drug is commercially available and to the best of our knowledge, it was the first time that the US Food and Drug Administration (FDA) approved a drug for which metabolite quantification was carried out using HPLC–ICP-MS. Since then, we have tackled increasingly challenging analytical problems. Each time, it gives me great pleasure to see that my coworkers (PhD students and postdoc researchers) not only succeed in developing methods suited for the task, but also come up with creative ideas to further extend the application range, improve the figures of merit, and make the methods more straightforward. Our most recent results show that for Cl-containing compounds, straightforward quantification based on a mathematical function describing how the sensitivity for Cl varies as a function of the mobile-phase composition is feasible.  We also have demonstrated that drugs (and their metabolites) containing a primary amine function can be successfully derivatized, such that ICP-MS monitoring of the Br or I atoms introduced allows for quantitative metabolite profiling via HPLC–ICP-MS.

References

1. B. Meermann, M. Bockx, A. Laenen, C. Van Looveren, F. Cuyckens, and F. Vanhaecke, Anal. Bioanal. Chem. 402, 439–448 (2012). DOI: 10.1007/s00216-011-5339-y.
2. B. Klencsár, L. Balcaen, F. Cuyckens, F. Lynen, and F. Vanhaecke, Anal. Chim. Acta974, 43–53 (2017). DOI: 10.1016/j.aca.2017.04.030.

This interview has been edited for length and clarity. To read the full interview, please visit: www.spectroscopyonline.com/quantitative-drug-metabolite-profiling-without-radiolabels-using-hplc-icp-ms