Welcome to StopRM.org

On these pages you find a lot of material that helps you avoid  Reactive Metabolite problems in the drug design process

The purpose of the StopRM.org site is to increase the awareness / avoidance of reactive metabolites (RMs) among drug researchers involved in drug design. To this end, links to literature reviews and other educational material have been collated.

Contents on this page

  1. More on Background
  2. Can we design drug candidates with no or less RM issues?
  3. Major classes of reactive metabolites

It is a prime goal for the drug industry to provide medicines that are both therapeutically efficient and safe to patients. One safety aspect of great concern involves metabolic activation of drugs to form reactive intermediates, mostly electrophiles that are capable of covalently modifying biological macromolecules potentially leading to severe adverse drug reactions (ADRs). Such reactive metabolites (RMs) have caused, and continue to cause, much suffering for patients taking certain medicines, and numerous fatal cases have occurred during the years. Several RM-caused ADRs have led to drug withdrawal from the market causing economic losses for the drug industry and for the society as a whole.

Below are three classical examples of reactive metabolites (RMs) from drugs known to cause ADRs by  bioactivation.

The bioactivation of paracetamolnystart2

conversion of the anaesthetic halothane to a very reactive acyl chloride

1) The bioactivation of paracetamol (acetaminophene) by oxidation gives rise to a reactive quinoneimine (NAPQI), 2) shows the antipsychotic agent clozapine forming a reactive diazatropylium ion, and 3) shows the conversion of the anaesthetic halothane to a very reactive acyl chloride.

More on Background

Many drugs are converted by drug metabolizing enzymes to intermediates which can react with cellular components. These reactions leading to covalent binding to macromolecules can result in organ toxicity (liver being the organ most commonly affected), various immune mediated hypersensitivity reactions, and mutagenesis leading to tumor formation. Mutagenesis and carcinogenesis arise as a consequence of DNA damage, while other adverse events are linked to chemical modification of proteins and possibly also, in rare cases, to lipids. Some RMs are likely acting by their propensity to cause oxidative stress.

The formation of RMs causing organ injury appears to be particularly relevant to drug-induced liver injury (DILI) since drug metabolism usually occurs primarily in the liver, and any resulting electrophiles should attack hepatic proteins first because of their proximity. Consistent with such a scenario is the fact that liver injury represents one of the major drug treatment–related toxicities, accounting for approximately 50% of clinical cases of acute liver failures (for example, Lee and Senior, 2007).

Another organ often affected by RMs is the skin. Many RM-linked ADRs list such manifestations as cutaneous reactions.

Cells in the blood are also exposed to drug-induced pathologies and in many cases they are clearly caused by metabolites. Immunocytopenias (or blood dyscrasias) are feared ADRs from many classes of drugs, for example many older NSAIDs and CNS drugs, like clozapine.

The relationship between reactive metabolite formation and drug toxicity is complex. Some toxicities are highly reproducible and dose dependent, and can readily be reproduced in various preclinical species and in man and are referred to as Type A toxicity. A good example is liver toxicity caused by paracetamol (acetaminophen) which is initiated by formation in hepatocytes of the quinoneimine reactive metabolite shown above. Even though Type A toxicity is usually picked up during in vivo preclinical safety testing of candidate drugs, it can still be an important cause of inefficient drug discovery. For some drugs however, toxicity in man is an unpredictable and relatively infrequent event.  These drugs are said to cause idiosyncratic drug reactions (IDRs), which are also referred to as Type B toxicity. The toxicity is due to individual human differences in response to stimuli and often involves the immune system. They are not overtly dose dependent in a wider population and cannot be reproduced in preclinical species. However, while they may not appear to be dose dependent in a general human population that does not mean there is not an underlying dose dependency.

Can we design drug candidates with no or less RM issues?

Any effort aiming to lower the incidence of new drug candidates showing up with issues related to RM formation and, most importantly, to terminate a clinical candidate drug (CD) before  clinical development begins might consider, at least, the following aspects:

  1. To what detail are the current mechanisms of severe ADRs, thought to be caused by RMs, understood at the chemical level? This is about learning for avoidance!
  2. In many cases it is possible to demonstrate the presence of covalent binding to tissues (proteins). But the link between this observation and an immune-mediated idiosyncratic ADR is hard to verify.
  3. To what extent is exposure (dose) to an identified RM linked to severity or incidence of the ADRs? To establish a Structure (Re)activity Relationship within a series of (closely related) compounds generating RMs and severity of the ADR is a demanding task.
  4. Which tests/assays can reliably be used to filter away potentially hazardous compounds?
  5. In vitro versus in vivo observations, especially of covalent binding? A further complication is the question of which animal species translates best to human (pattern of metabolites and immune response)?

These basic questions and issues point to an unusually tough predicament for the drug designer who tries to avoid making compounds with RM liabilities – in addition to all the other optimization parameters. It is clear, however, that there are hints from past experience about which chemical groups might present hazards when acted upon by metabolizing enzymes.

Completely avoiding introduction of any RM liability might be a utopian goal. Another way of arriving at the same goal is to make very potent compounds. There is an absence of RM-related ADRs of licensed drugs which are given in a daily dose of less than approximately 20 mg.

Major classes of reactive metabolites

  • Aromatic hydrocarbons such as benzene or naphthalene form ‘arene oxides’ catalyzed by the cytochrome-dependent P450 enzymes. The labile epoxides react with nucleophiles in macromolecules or rearrange spontaneously to phenols:

Aromatic hydrocarbons

  • It is exceedingly common that aromatic compounds are set up for facile oxidation or elimination reactions to form quinones or similar compounds (quinoneimines, quinone methides and more) having other atoms than oxygen. Many of the so formed species can react with bio-nucleophiles:

Many of the so formed species can react with bio-nucleophiles

The direct precursors of the quinone-like compounds are often metabolites of a drug.

  • Benzylic alcohols, and their heterocyclic and allylic equivalents can easily form reactive carbenium ions, even more so when electron-donating substituents are in place. The common sulfotransferase enzymes can worsen the situation by converting the alcohol into a sulfate which provides a good leaving group, as the example below shows.

Benzylic alcohols

  • Aromatic amines are notorious as carcinogens and RM-forming precursors. The mechanism involves an enzymatic hydroxylation to a hydroxylamine. This, in turn, can be enzymatically acetylated to an acetate ester or sulfonated to a sulfate ester. Both of these, given the right prerequisites, can generate a nitrenium ion which can react with macromolecules (DNA or proteins).

Aromatic amines are notorious as carcinogens and RM-forming precursors

Nitro compounds can be enzymatically reduced to hydroxylamines (and nitroso compounds) and therefore also represent potential precursors of nitrenium ions.

There are many other types of reactive intermediates, some formed by intricate reaction pathways and less well explored mechanisms, but the fact remains that the listed basic types of intermediates might be responsible for over 80 % of known problems.

This was just a brief introduction to the field. Please, send your suggestions on material to be added to the page in the ”send” link!