Thioridazine

The label describes thioridazine's basic pharmacology as similar to other phenothiazines, with few extrapyramidal (movement) side effects. It has two active metabolites, mesoridazine and sulforidazine. Adult doses start at 50 to 100 mg three times a day and range from 200 to 800 mg a day [1].

The concern is the heart. Thioridazine prolongs the QTc interval, a measure of how long the heart takes to recharge between beats, in a dose-related way. In nine healthy men, a single 50 mg dose lengthened the QTc by about 23 milliseconds on average, and the label warns that unscreened patients may see more. QT prolongation of this kind is linked to torsade de pointes, a dangerous heart rhythm, and sudden death. The boxed warning therefore reserves thioridazine for people who have not done well on other antipsychotics, and the label recommends at least two trials of different antipsychotics first [1].

The label asks for a baseline ECG and potassium level, says people with a QTc over 450 milliseconds should not start thioridazine, and says it should be stopped if the QTc goes over 500 milliseconds [1].

CYP2D6 is a liver enzyme with many inherited variants. People with two non-working copies are poor metabolizers. CPIC estimates that about 6.5% of people of European ancestry are poor metabolizers, compared with about 2% to 3% in most other groups and under 1% of people of East Asian or Oceanian ancestry [2]. The label puts the figure at about 7% of the population [1].

The label's main evidence is a small study in 19 healthy men. After a single 25 mg dose, the 6 slow metabolizers (identified by the debrisoquine test, a functional measure of CYP2D6) had 2.4 times the peak level and 4.5 times the total exposure of the 13 rapid metabolizers [1].

Studies in patients found the same pattern. In 76 patients on thioridazine alone, the dose-corrected level fell as the number of working CYP2D6 genes rose: a median of 15.2 nmol/L per mg with none, 7.2 with one and 4.0 with two. Smokers had lower levels than non-smokers [3]. In 97 patients, the 9 with no working CYP2D6 had higher dose-corrected levels, and age and smoking also affected levels. In that study the QTc interval itself did not differ significantly by CYP2D6 genotype [4].

The FDA table notes that the predicted effect in poor metabolizers is "based on experience with CYP2D6 inhibitors" [5]. The contraindication therefore rests on higher drug levels and the known link between thioridazine levels and QT prolongation, not on trials of poor metabolizers with heart rhythm outcomes.

The FDA label contraindicates thioridazine in people "known to have a genetic defect leading to reduced levels of activity of P450 2D6", and with drugs that inhibit CYP2D6, such as fluoxetine and paroxetine. It also contraindicates fluvoxamine, propranolol and pindolol, which raise thioridazine levels by other routes, and any other drug that prolongs the QTc interval [1].

The FDA's Table of Pharmacogenetic Associations lists thioridazine and CYP2D6 in its first section and states that it is "contraindicated in poor metabolizers" [5]. A Gene2Rx report shows that statement for a poor metabolizer result.

The label's wording, "reduced levels of activity", is broader than the FDA table's "poor metabolizers". The table and the Gene2Rx report apply the contraindication to poor metabolizers only; whether an intermediate result should change the decision is for the prescriber. CPIC has not published a guideline; its gene-drug list gives the pair a provisional level of B/C [6]. The DPWG's 2024 antipsychotic guideline does not cover thioridazine [7].

If you take thioridazine or one has been suggested, show your prescriber any poor metabolizer result. Bring a full list of your medicines too. Drug interactions cause the same problem as the genes: someone with normal CYP2D6 genes who takes a CYP2D6 inhibitor can clear thioridazine like a poor metabolizer, an effect called phenoconversion. The label names fluoxetine and paroxetine; the FDA also lists bupropion, quinidine and terbinafine as strong CYP2D6 inhibitors [8]. Fluvoxamine raised thioridazine and its active metabolites about 3-fold, and propranolol raised thioridazine levels by about 50% to 400% [1].

Smoking lowers thioridazine levels [3], so stopping smoking can raise them. Do not start, stop or change thioridazine or any other medicine because of a genetic result without talking to your prescriber.

Consumer arrays (23andMe, AncestryDNA, MyHeritage) are not clinical-grade tests, and their raw data is labeled not for medical use. They read fixed positions, and Gene2Rx uses imputation to infer some variants that were not measured directly. A normal result means none of the assessed variants was found among the positions your data covers; it cannot rule out a variant at a position the data does not cover. For a contraindication tied to heart rhythm, a normal Gene2Rx result is not enough to show that thioridazine is safe; a clinical CYP2D6 test is reasonable before starting it. If a clinical test (GeneSight, Genomind or another lab) has already reported your CYP2D6 genotype, those results can be entered for a Panel Report.

A genotype says nothing about other QT risks, such as congenital long QT syndrome, low potassium or other medicines. The ECG and potassium checks in the label are the direct measures.

My report says I am a CYP2D6 poor metabolizer. Can I take thioridazine?

The FDA label contraindicates it for people with reduced CYP2D6 activity (see What the FDA label says). Talk to your prescriber; do not stop it on your own.

My result is normal. Does that make thioridazine safe for me?

No. It rules out only the variants your data covers, and the QT risk applies to everyone at higher doses or with interacting drugs (see What this result cannot tell you).

1. Thioridazine hydrochloride tablets: prescribing information (Mylan). DailyMed. Boxed warning, contraindications, CYP2D6 data, interactions and dosing.
2. CPIC. CYP2D6 frequency table (spreadsheet). Estimated phenotype frequencies by biogeographic group.
3. Berecz R et al. Thioridazine steady-state plasma concentrations are influenced by tobacco smoking and CYP2D6, but not by the CYP2C9 genotype. Eur J Clin Pharmacol (2003). Levels by number of working CYP2D6 genes and by smoking.
4. Thanacoody RH et al. Factors affecting drug concentrations and QT interval during thioridazine therapy. Clin Pharmacol Ther (2007). CYP2D6, age, smoking, drug levels and QTc in 97 patients.
5. US Food and Drug Administration. Table of Pharmacogenetic Associations. The FDA statement for thioridazine and CYP2D6 poor metabolizers.
6. CPIC. Gene-drug pairs and CPIC levels (spreadsheet). Provisional level B/C for CYP2D6 and thioridazine.
7. Beunk L et al. Dutch Pharmacogenetics Working Group (DPWG) guideline for the gene-drug interaction between CYP2D6, CYP3A4 and CYP1A2 and antipsychotics. Eur J Hum Genet (2024). The drugs the DPWG antipsychotic guideline covers.
8. US Food and Drug Administration. For healthcare professionals: FDA's examples of drugs that interact with CYP enzymes and transporter systems. Strong CYP2D6 inhibitors.