Background

Topical corticosteroids (TCS) are used to reduce inflammation and are one of the most commonly prescribed medicines in dermatology. They were first used successfully by Sulzberger and Witten in 1952, and their success marked a cornerstone in the history of dermatology.1 Topical corticosteroids are the mainstay of atopic dermatitis treatment and used for other skin conditions such as psoriasis, where they are often required for months or years to control the disease and ultimately restore patients' quality of life. Numerous TCS are now available in different preparations, concentrations and potencies; however, when used appropriately, TCS efficacy and safety are well established.2-6

The beneficial anti-inflammatory effects of TCS are complex, being largely mediated via the cytoplasmic steroid receptor and involving actions on circulating cellular and cytokine mediators of inflammation as well as on the peripheral vasculature.7 The use of TCS is tempered by consideration of local and less frequently encountered systemic side effects. Known local side effects include skin atrophy, contact allergy, acne, mild hypopigmentation and hypertrichosis. Rarely, absorption through the skin can cause adrenal suppression. The risk of developing side effects is related to the potency, preparation, frequency and duration of use and the age of the patient. In clinical practice, these side effects are uncommon when TCS are used within their guidance.

There are two types of skin cancers: melanoma skin cancer and non-melanoma skin cancer (NMSC) (keratinocyte). Around 97% of NMSC comprise mainly of basal cell carcinomas (BCCs) or cutaneous cell carcinomas. The incidence of NMSC is increasing worldwide,8-12 with an estimated two to three million new cases of NMSC recorded each year.13 With respect to cutaneous malignant melanoma (CMM), this is the most serious form of skin cancer and has been increasing steadily in incidence over the past 30 years.14 Mortality due to CMM is much higher than that of NMSC.15

There are several observational studies that have looked at the relative risk of developing skin cancer due to oral corticosteroid exposure.16,17 These studies have provided conflicting results as to whether corticosteroids are associated with an increased risk of skin cancer. Karagas et al.16 conducted a case-control study on over 800 non-transplant squamous cell carcinoma (SCC) and BCC patients. The authors found that oral glucocorticoids may increase the risk of NMSC, whereas Baibergenova et al.17 found no association between NMSC and oral corticosteroids in a follow-up study of a chemotherapy trial with 1051 study participants. These studies highlight the clinical equipoise that exists around the impact oral corticosteroids have on the risk of skin cancer.

There have been several epidemiological studies that have explored the risk of cancer specifically among atopic dermatitis patients. Hagstromer et al.18 conducted a hospital-based study on 15,666 patients with atopic dermatitis in Sweden between 1965 and 1999.18 The authors reported that men faced a 50% increased risk of NMSC during the first 10 years of follow-up, but this did not reach statistical significance. Wang and Diepgen19 conducted a review of atopic dermatitis studies published before 2004, and no consistent associations were observed for skin cancers. This review did not look at the effect of TCS use on the risk of skin cancer. At present, we do not know in particular what impact TCS have on the risk of skin cancer in the atopic dermatitis population.

With regard to the organ transplant population, it is well established that immunosuppression increases the risk of skin malignancy.20,21 This occurs when systemic corticosteroids are used, although most studies include patients treated with a combination of immunosuppressants including azathioprine and calcineurin inhibitors.22,23 Corticosteroids are known to have an immunosuppressive effect, and TCS may have a local immunosuppressive effect. It is not known whether TCS may increase the risk of skin cancer through this mechanism.

On the other hand, it is possible that treating skin inflammation with TCS may reduce the risk of skin cancer. Several systematic reviews and meta-analyses report the benefits of anti-inflammatory drugs in reducing the risk of cancer, including skin cancers.24,25 The management of certain types of inflammatory skin diseases includes the rationale that reducing inflammation reduces the risk of SCC development in vulval and penile lichen sclerosus as well as hypertrophic lichen planus. It is also known that chronic inflammation is a risk for the development of SCC, such as in chronic ulceration and the development Marjolin's ulcer.26,27 This mainly holds true for SCC but less is known about BCC and melanoma. Therefore, overall TCS may decrease the risk of skin cancer in patients in whom TCS are used to treat inflammatory skin disease.

A search of MEDLINE and Embase revealed that no published systematic reviews or meta-analyses have been performed to collate evidence on long-term TCS use on the risk of skin cancer. Immunosuppression induced by TCS, either local or systemic, may allow these cancers to emerge from reduced innate immunosurveillance. However, TCS may also reduce the risk of skin cancer in patients in whom TCS are used to treat inflammatory skin disease. With TCS use being one of the most commonly prescribed drugs in the clinical field of dermatology and the increasing incidence of skin cancer, there is a need to review all current evidence about the possible association.

Inclusion criteria

Types of participants

The current review will consider studies that include people of all ages, genders and ethnicities.

Participants with HIV, transplant participants or participants with genetic diseases (e.g. Gorlin-Goltz syndrome) will be included.

Exposure of interest

The current review will consider studies that evaluate long-term use of TCS. Our definition of "long-term" consists of more than once a week for a month or longer.

Outcomes

The current review will consider studies that include the following outcome measures: non-melanoma skin cancer, cutaneous SCC, BCC or melanoma skin cancer. These outcomes will be measured by a clinical diagnosis and, where available, histological confirmation. Studies looking at the outcomes in oral, vulvar or genital sites only will not be included as these are not relevant in the context of our hypothesis. Precursors such as Bowen's disease will be secondary outcomes.

Types of studies

The current review will include analytical comparative observational studies including prospective and retrospective cohort studies, case-control studies and cross-sectional studies.

Search strategy

The search strategy will aim to identify both published and unpublished studies. A three-step search strategy will be utilized in this review. An initial limited search of MEDLINE and Embase will be undertaken followed by an analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search using identified keywords and index terms has been used to develop a comprehensive search strategy. The search strategy for MEDLINE is detailed in Appendix I. Third, the reference list of all identified reports and articles will be searched for additional studies. Studies published in all languages will be included. There will be no date exclusion.

Information sources

The electronic databases to be searched include: MEDLINE, Embase and LILACS all from inception to current date.

The search for unpublished studies will include: skin cancer experts who have been identified from the included studies will be contacted. We will also search EThOS at the British library (http://ethos.bl.uk) to identify other unpublished work.

Study selection

Following the search, all identified citations will be collated and uploaded in EndNote and duplicates removed. Titles and abstracts will then be screened by two independent reviewers (SR and EBT) for assessment against the inclusion criteria for the review. Studies that appear to meet the inclusion criteria will be retrieved in full and their details imported into the Joanna Briggs Institute System for the Unified Management, Assessment and Review of Information (JBI-SUMARI). The full text of selected citations will be retrieved and assessed independently by two reviewers (SR and EBT) in detail against the inclusion criteria. Full-text studies that do not meet the inclusion criteria will be excluded, and reasons for exclusion will be provided in an appendix in the final systematic review report. The results of the search will be reported in full in the final report and presented in a PRISMA flow diagram. Any disagreements that arise between the reviewers will be resolved through discussion or with a third reviewer (FBH).

Assessment of methodological quality

Included studies will be critically appraised by two independent reviewers (SR and EBT) at the study level for methodological quality in the review using standardized critical appraisal instruments from the Joanna Briggs Institute for cohort, cross-sectional surveys and/or case-control study designs (Appendices II and III). The instrument will be amended for our needs. Any disagreements that arise between the reviewers will be resolved through discussion or with a third reviewer (FBH). The results of critical appraisal will be reported in narrative form and in a table.

Data extraction

Data will be extracted from papers included in the review using the standardized data extraction tool for cohort, cross-sectional surveys and/or case-control studies in (JBI-SUMARI) (Appendices II and III) by two independent reviewers SR and EBT. We will tailor the extraction form to our needs. The data extracted will include specific details about the exposure of interest including different exposure categories if applicable, populations, study methods and relevant outcomes measures. Any disagreements that arise between the reviewers will be resolved through discussion or with a third reviewer (FBH). Authors of papers will be contacted to request missing or additional data where required.

Data synthesis

Papers will, where possible, be pooled using random effect meta-analysis methods in RevMan 5.3 (Copenhagen, Denmark: The Nordic Cochrane Centre, Cochrane). Effect sizes will be expressed as relative risks (odds ratios or risk ratios) together with their 95% confidence intervals (CIs). If we can assume that the outcome of interest is rare (i.e. <5-10%), then the odds ratio will be very similar to the risk ratio. Effect measures adjusted for confounders will be used in preference to crude effect measures. Where effect estimates and measures of precision (e.g. standard errors, 95% CI) cannot be directly extracted from the included study, we will estimate them from data presented in the paper.

Heterogeneity will be quantified using I² and [tau]². Subgroup analyses will be conducted to explore reasons for heterogeneity in the meta-analysis models based on study quality, adjusted versus crude measures of effect, dose of TCS (low versus high), non-melanoma/melanoma and type of patient population (HIV/transplant participants/participants with syndromes versus those without those conditions). Where there is insufficient data to allow for meta-analysis, the findings will be presented in narrative form including tables and figures to aid in data presentation where appropriate.

We will conduct a sensitivity analysis excluding keratinized epithelium in special sites (e.g. vulvar and penile skin) to determine if there is a difference compared to the overall result. A funnel plot will be generated using RevMan 5.3 to assess publication bias if there are 10 or more studies included in a meta-analysis. Statistical tests for funnel plot asymmetry (Egger test, Begg test and Harbord test) will be performed where appropriate. The GRADE approach for assessing confidence in the quality of evidence will be used for this review, with the results presented in a summary of findings table created using GRADEPro.

Acknowledgements

The authors would like to thank Liz Doney and Douglas Grindlay for their help with developing the search strategies.

Appendix I: Search strategy for OVID MEDLINE

Epidemiologic studies/

Exp Case-control studies/

Exp Cohort studies/

Epidemiologic$ stud$.mp.

Case control stud$.mp.

Cohort stud$.mp.

Cohort analy$.mp.

Follow up stud$.mp.

Observational stud$.mp.

Longitudinal.mp.

Retrospective.mp.

Cross sectional stud$.mp.

Cross Sectional Studies/

Exp Observational Study/

Or/1-14

Carcinoma, Basal Cell/

Neoplasms, Basal Cell/

Basal Cell Nevus Syndrome/

Basal cell carcinoma$.mp.

Basal cell cancer$.mp.

Basal cell neoplasm$.mp.

Nodular BCC.mp.

Naevoid BCC.mp.

Gorlin syndrome.mp.

Basal cell Epithelioma$.mp.

Basalioma$.mp.

BCC.mp.

Rodent ulcer$.mp.

Or/16-28

Exp Neoplasms, Squamous cell/

Exp Carcinoma Squamous Cell/

Squamous cell carcinoma$.mp.

Squamous cell cancer$.mp.

Squamous cell neoplasm$.mp.

Bowen's disease.mp.

Planocellular carcinoma$.mp.

SCC.mp.

Or/30-37

Skin neoplasms/

NMSC.mp.

Non melanoma skin cancer$.mp.

Skin cancer$.mp.

Skin tumo$.mp

Skin neoplasm$

Exp Keratinocytes/

Keratinocytes.mp.

Or/38-46

Melanoma/

Melanoma.mp.

Or/48-49

topical corticosteroid$.mp.

steroid$.mp.

corticosteroid$.mp.

exp Glucocorticoids/

alclometasone.mp.

alclomethasone.mp.

amcinonide.mp.

beclometasone.mp.

beclomethasone.mp.

exp Beclomethasone/

betametasone.mp.

betamethasone.mp.

exp Betamethasone/

clobetasol.mp.

exp Clobetasol/

clobetasone.mp.

desonide.mp.

exp Desonide/

desoximetasone.mp.

exp Desoximetasone/

diflorasone.mp.

diflucortolone.mp.

exp Diflucortolone/

fludroxycortide.mp.

flumetasone.mp.

flumethasone.mp.

exp Flumethasone/

fluocinolone.mp.

exp Fluocinolone Acetonide/

fluocinonide.mp.

exp Fluocinonide/

fluocortolone.mp.

exp Fluocortolone/

flurandrenolide.mp.

flurandrenolone.mp.

exp Flurandrenolone/

fluticasone.mp.

halcinonide.mp.

exp Halcinonide/

halobetasol.mp.

halometasone.mp.

hydrocortisone.mp.

exp Hydrocortisone/

methylprednisolone.mp.

exp methylprednisolone/

mometasone.mp.

triamcinolone.mp.

exp Triamcinolone/

Or/ 51-98

15 AND (29 OR 38 OR 47 OR 50) AND 99

Appendix II: Appraisal instruments

Appendix III: Data extraction instrument

References