Drug interactions in targeted therapies: CYP3A4 and QT interval prolongation

Targeted therapies changed the profile of the drug interaction problem in oncology. Many are taken orally, continuously and at home, with narrow therapeutic margins and plasma exposure sensitive to external factors. In that setting, a drug added for another reason, an antifungal or even a herbal product can significantly alter the concentration of the oncology treatment.

This article describes general mechanisms and reference sources. It contains no treatment recommendations, dose adjustments or monitoring schedules: those decisions belong to the care team, usually in collaboration with hospital pharmacy and, where appropriate, cardiology.

CYP3A4 is an enzyme of the cytochrome P450 system, present mainly in the liver and intestine, responsible for metabolising a very high proportion of commonly used drugs. Many tyrosine kinase inhibitors and other oral targeted therapies are predominantly metabolised through this pathway, which makes them sensitive to any substance that alters its activity.

Inhibition

A CYP3A4 inhibitor reduces the enzyme's metabolic capacity and can therefore increase exposure to a drug that uses it as its main route of elimination. The expected result is a higher plasma concentration and a greater risk of exposure-dependent toxicity. Groups classically described as potent inhibitors include several azole antifungals, some macrolides and certain boosted antiretrovirals; grapefruit juice is the most frequently cited dietary example.

Induction

An inducer produces the opposite effect: it increases expression of the enzyme and accelerates the drug's metabolism, with the possible consequence of insufficient exposure and loss of efficacy. Relevant inducers described include certain antiepileptics, some antituberculosis drugs and St John's wort, an over-the-counter product patients often do not regard as medication and therefore do not mention.

Three practical nuances are worth remembering. First, the magnitude of the effect depends on the specific drug and cannot be extrapolated between members of the same family. Second, induction has slow kinetics: it takes days or weeks to establish and to disappear, so stopping an inducer does not normalise the situation immediately. Third, CYP3A4 is not the only relevant pathway: other isoenzymes and transporters such as P-glycoprotein take part in interactions with clinical implications.

To this must be added the reverse direction of the relationship: some targeted therapies are themselves inhibitors or inducers of metabolic enzymes and can alter exposure to the patient's chronic medication, including anticoagulants or other narrow-margin treatments.

Not all interactions are metabolic. Several oral targeted therapies require an acidic environment to dissolve properly, so drugs that reduce gastric acidity — proton pump inhibitors, H2 antagonists, antacids — can reduce their absorption and systemic exposure. This interaction is particularly relevant because acid-suppressing drugs are very widely used and are often continued for long periods without review.

Likewise, the conditions of administration in relation to food can modify bioavailability in a clinically significant way. Each drug's summary of product characteristics specifies those conditions, and adhering to them is part of the treatment, not an incidental detail.

The QT interval on the electrocardiogram reflects the duration of ventricular depolarisation and repolarisation. Excessive prolongation is associated with an increased risk of serious ventricular arrhythmias, including torsade de pointes. The mechanism most frequently involved in drug-induced prolongation is interference with the potassium currents responsible for cardiac repolarisation.

Some targeted therapies are among the drugs with this effect described in their summary of product characteristics, and the risk does not depend on the oncology drug alone. It usually arises from the concurrence of several factors:

• Concurrent use of other drugs with a known effect on the QT interval, such as certain antiarrhythmics, antipsychotics, antidepressants, antiemetics or antibiotics.
• Electrolyte disturbances, particularly hypokalaemia and hypomagnesaemia, which are far from uncommon in oncology patients.
• Pharmacokinetic interactions that increase exposure to the QT-prolonging drug, so that the metabolic and cardiac effects compound one another.
• Patient factors: structural heart disease, bradycardia, history of long QT syndrome, advanced age.

For that reason, reviewing QT risk is not a one-off check on a single drug but an assessment of the whole medication list and of the patient's clinical context, to be carried out by the care team with whatever monitoring it considers appropriate.

The reference source for a specific drug is its authorised summary of product characteristics, which sets out in structured form the information on interactions, warnings, precautions for use and conditions of administration. It is the document that underpins any statement about the drug's behaviour, and its version matters, because it is updated.

• CIMA, the medicines information centre of the Spanish Agency for Medicines and Medical Devices (AEMPS), provides summaries of product characteristics and package leaflets for medicines authorised in Spain.
• The European Medicines Agency (EMA) publishes product information for medicines authorised through the centralised procedure, together with assessment reports.
• The FDA provides prescribing information for products authorised in the United States, useful as a comparative reference.
• BIFIMED, from the Spanish health administration, provides information on the reimbursement status of medicines.
• Clinical practice guidelines and indexed literature, with their PMID identifier, provide the evidence context; they should always be distinguished from authorised product information.

When information comes from different sources, it is important to state which source each statement was taken from. A warning set out in a summary of product characteristics and an observation published in a study do not have the same status, and presenting them interchangeably diminishes both.

Several operational consequences follow from the above. The first is that the list of concomitant medication must be complete and up to date, and must explicitly include over-the-counter products, supplements and herbal products, which patients do not always regard as medication.

The second is that reviewing interactions is not a single act. Every change in the patient's medication, including one introduced by another specialist or started during an admission, reopens the question. Collaboration with hospital pharmacy is the usual mechanism for making that review happen systematically.

The third is that the review must be documented: what was checked, against which source and in which version. That record is what makes it possible to reconstruct the reasoning later and to distinguish between an undetected risk and a risk assessed and consciously accepted.

Automating this documentary review — cross-referencing concomitant medication against structured information from summaries of product characteristics and regulatory sources, and linking every warning to its origin — is one of the functions of Sphera OncoCore. Interpretation and decision remain with the clinical team.