Hydrochlorothiazide: risk of non-melanoma skin cancer, particularly in long-term use

Advise patients taking hydrochlorothiazide-containing products of the cumulative, dose-dependent risk of non-melanoma skin cancer, particularly in long-term use, and the need to regularly check for (and report) any suspicious skin lesions or moles. Counsel patients to limit exposure to sunlight and UV rays and to use adequate sun protection.

Advice for healthcare professionals:

  • pharmacoepidemiological studies have shown a dose-dependent increased risk of non-melanoma skin cancer (basal cell carcinoma [BCC] and squamous cell carcinoma [SCC], including SCC lip cancer) with exposure to increasing cumulative doses of hydrochlorothiazide (see table of data below)
  • inform patients taking hydrochlorothiazide-containing products of the risk of non-melanoma skin cancer, particularly in long-term use, and advise them to regularly check for and report any new or changed skin lesions or moles
  • reconsider the use of hydrochlorothiazide in patients who have had previous skin cancer
  • examine all suspicious moles or skin lesions (potentially including histological examination of biopsies)
  • advise patients to limit their exposure to sunlight and UV rays and use adequate protection when exposed to sunlight and UV rays to minimise the risk of skin cancer
  • report suspected adverse reactions associated with medicines to the Yellow Card Scheme

Study data showing increase risk of skin cancer

Two recent pharmaco-epidemiological studies 1 2 in Danish nationwide data sources (including the Danish Cancer Registry and National Prescription Registry) have shown a cumulative, dose-dependent, association between hydrochlorothiazide and non-melanoma skin cancer. The known photosensitising actions of hydrochlorothiazide could act as possible mechanism for this risk.

Table: Study findings

Type of cancer Number of cases Number of population controls Adjusted odds ratios with ever-use of hydrochlorothiazide (95% CI) Adjusted odds ratios with high3 cumulative use of hydrochlorothiazide (95% CI) Adjusted odds ratios with highest4 cumulative use of hydrochlorothiazide (95% CI)
BCC 71, 533 1,430,833 1.08 (1.05–1.10) 1.29 (1.23–1.35) 1.54 (1.38–1.71)
SCC 8,629 172,462 1.75 (1.66–1.85) 3.98 (3.68–4.31) 7.38 (6.32–8.60)
Lip cancer 633 63,067 2.1 (1.7–2.6) 3.9 (3.0–4.9) 7.7 (5.7–10.5)
BCC = basal cell carcinoma. SCC = squamous cell carcinoma. CI = confidence interval.      

The study authors’ analyses did not find a similar association for risk of BCC or SCC1 and SCC lip cancer2 with overall or cumulative use of other diuretics and other hypertensives, including bendroflumethiazide, calcium channel blockers, angiotensin-converting enzyme inhibitors, angiotensin II receptor antagonists, and furosemide.

Pedersen and colleagues reported that, assuming causality, 9 in 100 SCC cases and fewer than 1 in 100 BCC cases that were diagnosed during the study period may have been attributed to hydrochlorothiazide use.1 Pottegård and colleagues reported that 11 in 100 of SCC lip cancer cases occurring in the study period may have been attributed to hydrochlorothiazide use.2

About non-melanoma skin cancer

Non-melanoma skin cancer is a rare event. Incidence rates highly depend on skin phenotypes and other factors, which leads to different baseline risks and varying incidence rates in different countries. Estimated incidence rates in Europe range from 1–34 cases per 100,000 people per year for SCC and 30–150 per 100,000 people per year for BCC.

In the UK, rates of SCC and BCC vary by region. One systematic review estimated average incidence rates in England of 23 cases of SCC per 100,000 person-years and 76 cases of BCC per 100,000 person-years.5 Average rates in Scotland were 27 cases of SCC per 100,000 person-years and 90 cases of BCC per 100,000 person-years, with similar incidence also reported for Northern Ireland (31 cases and 87 cases per 100,000 person-years, respectively).5

Based on the results of the two Danish epidemiological studies, a best estimate of the increased risk is 7.7-fold for SCC and 1.5-fold for BCC based on a length of usage of hydrochlorothiazide 12.5mg daily for 44 years or 25 mg daily for 22 years. For hypertension, products containing 25 mg of hydrochlorothiazide are indicated only if patients are not adequately controlled on lower-dose products.

The Summary of Product Characteristics and Patient Information Leaflets for all the concerned products have been updated to inform of the risk of non-melanoma skin cancer. A letter about the risk and advice has also been sent to prescribers and dispensers of hydrochlorothiazide-containing medicinal products.

Background

Hydrochlorothiazide-containing medicinal products are used to treat hypertension, as well as oedema associated with cardiac or hepatic disease and chronic heart insufficiency (heart failure). In the UK, hydrochlorothiazide is only available in fixed-dose combination with other medicines. We estimate that approximately 28,000 patients in the UK take medicines containing hydrochlorothiazide.6

Report suspected adverse drug reactions on a Yellow Card

Please continue to report suspected adverse drug reactions associated with hydrochlorothiazide to MHRA through the Yellow Card Scheme.

Article citation: Drug Safety Update volume 12, issue 4: November 2018: 1.

  1. Pedersen SA, et al. Hydrochlorothiazide use and risk of non-melanoma skin cancer: A nationwide case-control study from Denmark. J Am Acad Dermatol 2018; 78: 673–81.  2 3

  2. Pottegård A, et al. Hydrochlorothiazide use is strongly associated with risk of lip cancer. J Intern Med 2017; 282: 322–31.  2 3

  3. High cumulative hydrochlorothiazide use for BCC and SCC1 is ≥50,000 mg (corresponding to 12.5 mg hydrochlorothiazide taken daily for about 11 years), high cumulative hydrochlorothiazide use for SCC lip cancer2 is ≥25,000 mg. 

  4. Highest cumulative hydrochlorothiazide use for BCC and SCC is ≥200,000 mg, highest cumulative hydrochlorothiazide use for lip cancer is ≥100,000 mg. 

  5. Lomas A, et al. A systematic review of worldwide incidence of nonmelanoma skin cancer. Br J Dermatol 2012; 166: 1069–80.  2

  6. This data has been extrapolated to the UK population from the Clinical Practice Research Datalink (CPRD), which is a representative sample of approximately 750 primary care practices across the UK. The data are based on prescriptions issued to patients and therefore it is not possible to confirm whether the medicine was dispensed and subsequently consumed by the patient. It also does not include medicines prescribed in hospital settings. The CPRD data are obtained under licence from the UK MHRA, however the interpretation and conclusions from this data are that of the authors alone. 

Published 14 November 2018