Thyrostatic Drugs and Severe Skin Reactions

A Registry Evaluation Over 20 Years

Thyrostatic drugs have an inhibitory effect on the thyroid and influence the complex production of thyroid hormones at different points. In this analysis we will investigate the two thyrostatic active ingredients that have been observed in the context of severe skin reactions: sodium perchlorate and thiamazole/carbimazole. Propylthiouracil was not considered as it had not been taken by any patient with a severe skin reaction in the relevant exposure period.

The effect of perchlorate anions on the sodium iodide symporter is medically used to shield the thyroid gland in contrast medium examinations against radioactively marked iodide (1). Europe-wide, preparations containing sodium perchlorate are available only in Austria and Germany. Severe skin reactions are not listed among adverse drug events.

Thiamazole is a sulfur-containing thyrostatic drug. The intake of thiamazole lowers the hormone level in the thyroid and subsequently the concentration of thyroid hormones in the blood. Thiamazole is the active metabolite of carbimazole, which is also used as a medication (1). In the European Union, 40 preparations are available that contain the active ingredient thiamazole (6 manufacturers, 13 preparations in Germany) and 35 that contain carbimazole (3 manufacturers, 6 preparations in Germany).

According to the German drug prescription report for 2022, prescriptions of thyrostatic drugs continually remained low: 1,3 million defined daily doses (DDD) of sodium perchlorate, 21.6 million for thiamazole, and 12.6 million for carbimazole (2).

The severe skin reactions include Stevens-Johnson syndrome and toxic epidermal necrolysis. These are regarded as different degrees of severity of a very rare life threatening blistering skin reaction with erosive mucosal involvement and are together known as “epidermal necrolysis” (EN) (Figure). The Dokumentationszentrum schwerer Hautreaktionen (dZh, the documentation center for severe skin reactions) in Freiburg has since 1990 systematically and prospectively collected all inpatient cases of EN in Germany (3). Hospitals where patients with EN are treated actively notify suspected cases to the dZh, whose staff provide help with diagnosis and treatment. If the suspicion is confirmed, they visit patients in the treating hospital in order to lend support and collect data. All documented cases of EN are evaluated by an independent expert committee. In our analysis, we studied on the basis of these strictly validated cases the association between EN and the thyrostatic drugs mentioned earlier. In the time period between 2003 and 2022, we identified among more than 1000 EN cases 50 patients who were treated with thyrostatic drugs.

Figure

Patient with epidermal necrolysis in the context of use of sodium perchlorate:patchy, largely confluent exanthema with blister formation and skin detachment. Photograph: Maja Mockenhaupt

We used the established algorithm for assessment of drug causality in epidermal necrolysis (ALDEN; 3, 4) to test for causality in all 50 cases, which was hampered by the common occurrence of polypharmacy. We evaluated each and every medication used. The ALDEN algorithm is based on the following criteria: the time lag between the start of medication intake and the index day (the day the reaction started), presence/availability of the active ingredient in the body before the index day (which considering the half-life of the medications as well as liver and kidney function), information regarding the prior and later intake and tolerability, the possible risk potential of the active ingredient (“highly suspect,” “suspect,” “under observation,” “unclear,” or “not suspect”), and alternative causes. We used numerical scores to carry out a causality assessment for every single medication that had been taken within four weeks before the onset of the reaction. The scores ranged from “very unlikely” via “unlikely,” “possible,” “probably,” to “highly probable” (3, 4). To date, the risk potential of thyrostatic drugs has been regarded as “unclear.”

In the 50 cases with EN in whom a thyrostatic drug had been taken, allopurinol (n=9; 314.3 million DDD), the Cox-2 inhibitors etoricoxib and rofecoxib, and meloxicam (each n=1) were evaluated as (very) probable medication triggers (ALDEN point score 6, 5, 4). In one case it was not possible to differentiate between carbamazepine (28.5 million DDD) and levofloxacin. In two cases, cefuroxime and amoxicillin were recorded, respectively (Table).

Table

50 patients with epidermal necrolysis (EN) from the total population who had taken either sodium perchlorate or/and thiamazole/carbimazole

In 14 of the 50 patients treated with thyrostatic drugs, the ALDEN score categorized the drugs as “possible” (n=13) or “probable” (n=1) triggers of the severe skin reaction. This corresponds to 28% of patients with EN (n=50) who had been treated with thyrostatic drugs and 1.28% of all investigated cases of EN. These patients had not taken any additional “probable” triggers. However, some were taking co-medications, which—according to the ALDEN algorithm—were also “possible” triggers. The patients were 6 men and 8 women aged 58–87 years. The median age was 69 years. These patients were taking an average of 17 different active ingredients (range 4–29). 8 reactions developed at home, 6 during an inpatient stay. Of the 6 patients who developed EN in hospital, 5 died. Of the 8 patients who developed EN at home, 6 survived.

Even though the risk from sodium perchlorate and thiamazole/carbimazole seems lower than that from other known triggers of epidermal necrosis, these thyrostatic drugs should in future causality assessments of EN cases by means of ALDEN be categorized as “under observation” or “suspect.” Because of an earlier analysis that did not see any evidence for a necessary prescription of sodium perchlorate before administration of contrast medium (5), because of the severe adverse effects on the skin and the fact that it has been used for decades primarily in German-speaking countries, its use regarding benefits and potential harms should be carefully balanced on an individual basis.

Nadja Simone Hertenstein, Ulrich Michael Hemmeter, Maren Paulmann, Maja Mockenhaupt

Dokumentationszentrum schwerer Hautreaktionen (dZh), Klinik für Dermatologie und Venerologie, Universitätsklinikum Freiburg (Hertenstein, Hemmeter, Paulmann, Mockenhaupt)

Conflict of interest statement
The authors declare that no conflict of interest exists.

Manuscript received on 31 March 2025, revised version accepted on 19 November 2025.

Translated from the original German by Birte Twisselmann, PhD.

Cite this as
Hertenstein NS, Hemmeter UM, Paulmann M, Mockenhaupt M: Thyrostatic drugs and severe skin reactions: A registry evaluation over 20 years. Dtsch Arztebl Int 2026; 123: 80–1. DOI: 10.3238/arztebl.m2025.0216

References

1. Bundesinstitut für Arzneimittel und Medizinprodukte. 84. Sitzung (13. Juli 2021 per Videokonferenz). www.bfarm.de/DE/Arzneimittel/Pharmakovigilanz/Ausschuesse-und-Gremien/Verschreibungspflicht/Protokolle/84Sitzung/kurzprotokoll_84.html (last accessed on 08 December 2025)..

2. Ludwig WD, Mühlbauer B, Seifert R (Hrsg.): Arzneiverordnungs-Report 2022. Kap. 13 Gicht, 329–334; Kap. 24 Epilepsien, 519–32; Kap. 37 Schilddrüsenerkrankungen, 725–31. Berlin, Heidelberg, Springer 2023

3. Heuer R, Paulmann M, Annecke T, et al.: S3-Leitlinie: Diagnostik und Therapie der epidermalen Nekrolyse (Stevens-Johnson-Syndrom und Toxisch epidermale Nekrolyse) – Teil 1: Diagnostik, initiales Management und immunmodulierende Systemtherapie. J Dt Dermatol Ges 2024; 22: 1448–67

4. Diederich S, Hemmeter U, Paulmann M, Mockenhaupt M: Effects of ­dosage in new users of lamotrigine inducing epidermal necrolysis: ­Results of the German Registry of Severe Skin Reactions. Epilepsia 2023; 64: 1259–65

5. Burchardt X: Routinemäßige Schilddrüsenuntersuchungen und medikamentöse Prophylaxe vor Gabe jodhaltiger Röntgenkontrastmittel zur Risikominimierung einer jodinduzierten Hyperthyreose – Beurteilung der Vorgehensweisen unter Aspekten evidenzbasierter Medizin (EBM). Dissertation, Universität Düsseldorf; 2003. https://d-nb.info/972391495/34 (last accessed on 08 December 2025).