Atopic Dermatitis and Cancer Risk

As inflammatory skin diseases, psoriasis and atopic dermatitis (AD) share many similarities. For example, both are characterized by a state of persistent immune activation with histologic and immunologic overlap and have been associated with increased morbidity, other comorbid diseases, negative impact on quality of life, and a significant economic burden.^1-6 In addition, both diseases share similarities with respect to their treatment landscape, with evolution from treatments based primarily on chance observations to the development of treatments based on defined therapeutic mechanisms or targeted approaches. While these newer treatment approaches have better efficacy and/or safety profiles compared with traditional treatment options, such as immunosuppressive therapies or UV therapies, these are not without risk and many harbor warnings with regard to a potential increase in the risk of malignancy or infection.⁷  

It is important to understand that patients with psoriasis and AD may be at a higher risk for malignancy due to (1) exposure to a state of chronic, persistent, immune activation, which in it of itself can increase the risk for malignancy; (2) exposure to multiple, systemic, immunosuppressive therapies and UV therapies throughout their lifetime thus potentially increasing the risk for malignancy; and (3) an increase in known cancer-related risk factors such as smoking, increased alcohol consumption, and higher rates of obesity.^8-12 As new systemic treatment options become available for the treatment of AD, it is important for physicians to understand what the underlying risk for malignancy is as this can have an impact on treatment selection. This article explores the literature on the overall risk of malignancy in AD and the risk associated with use of common topical and immunosuppressive medications, as well as newer therapeutic options.

Prior studies have shown inconsistent results with respect to the background risk of AD. In a study by Arana and colleagues approximately a 50% increased risk of any cancer was observed in patients with AD compared to controls (standardized incidence ratio [SIR]=1.49; 95% CI, 1.39-1.61).¹³ However, a study in the Taiwanese population showed no overall increased risk among patients with AD (SIR=0.97; 95% CI, 0.87-1.09).¹⁴ With respect to cancer specific risk, Legendre and colleagues found an increased risk for any lymphoma among patients with AD compared with controls (SIR=1.43; 95% CI, 1.12-1.81).¹⁵ Hwang and colleagues found a nonstatistically significant increased risk for melanoma (SIR=2.35; 95% CI, 0.26-8.47); however, other studies have shown opposite results.^14,16 It is important to note that comparisons between these studies are limited due to inconsistencies in study design, populations studied, and the inability to disentangle the effect of disease severity from treatment effect on cancer risk thus limiting our ability to draw any meaningful conclusions. 

In the risk of cancer associated with topical or systemic treatments in patients with AD, there is still considerable debate. For example, topical calcineurin inhibitors (TCIs), such as pimecrolimus cream and tacrolimus ointment, are considered to be effective nonsteroidal treatment options for patients with AD and have been approved for use in both adults and children. However, in 2006, the FDA determined that a theoretical risk for lymphoma and skin cancer existed with use of these medications based on a limited number of studies and despite any strong evidence to support an increased cancer risk.^17,18 Furthermore, additional studies have shown no increased risk of cancer with use of TCIs. A prospective, longitudinal cohort study with more than 10 years of follow-up found no increased risk of malignancy among users of pimecrolimus in a real-world setting (SIR=1.2; 95% CI, 0.5-2.8).¹⁹ In addition, a comprehensive systematic review found no evidence that tacrolimus 0.03% or 0.1% resulted in an overall increase risk of malignancy in either children or adults.²⁰ 

As topical nonsteroidal treatment options are currently limited for patients with AD and because TCIs offer a safe and effective treatment option, careful discussion must be done with patients when prescribing these medications to eliminate or limit any fears which may preclude the use of these medications. Strategies to accomplish this, which have been previously discussed by others, include: (1) discussing the theoretical drug-related risks vs well-established risks of a chronic disease such as AD; (2) preparing patients to receive conflicting information from pharmacists or other health care professionals who may not be as well versed in prescribing TCIs; (3) emphasizing the benefits of a nonsteroidal option (ie, decrease risk of skin atrophy); and (4) discussing daily or maintenance approach as more favorable with use of TCIs vs topical corticosteroid use.^21,22 

Until recently, systemic medications used for the treatment of AD included cyclosporine (CsA), methotrexate, azathioprine (AZT), and mycophenolate mofetil, among others. Yet, while the risk of cancer associated with these medications has not been adequately explored and most of the information is based on data from the psoriasis literature, current American Academy of Dermatology guidelines for systemic treatments in AD state that “an increased risk of skin cancer and lymphoma may be observed with use of immunosuppressive drugs in AD.”⁷ Various cases of cutaneous T-cell lymphoma, lymphoproliferative disorders, and systemic lymphomas have been previously described in patients with AD treated with CsA.^23,24 

However, it is important to note that in these cases, ample variation in patient characteristics, treatment duration and dose, and exposure to prior or concomitant systemic therapies all confound the association with cancer risk and CsA use in AD. In addition, a single-center, retrospective, cohort study using CsA as monotherapy found no cases of lymphoma after 10 years of follow-up.²⁵ With respect to the risk of skin cancer, Garritsen and colleagues found an increased risk of cutaneous squamous cell carcinoma (cSCC) in users of CsA (SIR=10.7; 95% CI, 4.6-16.7).²⁶ However, a lack of association was observed for treatment duration and the prolonged time in between CsA exposure and diagnosis of skin cancer in some cases, which urges physicians to accept these findings with caution.

Methotrexate has also been a staple of treatment in AD for many years despite not having an FDA-approved indication for this treatment. Data on the risk of cancer with use of methotrexate in AD is scarce. One single study using a database from Denmark found a small increased risk for cutaneous melanoma in users of methotrexate but this included patients whose indication for treatment was any number of other diseases in addition to AD (hazard ratio=1.17; 95% CI, 1.08-1.26).²⁷ Finally, a follow-up study in this same population found no dose-response association with cancer risk among users of methotrexate.²⁸ Other agents that are used for the treatment of AD include AZT and mycophenolate mofetil. While there is a paucity of data on the risk of cancer among users of AZT in AD, data on patients with rheumatoid arthritis and inflammatory bowel disease have found an increased risk of nonmelanoma skin cancer (NMSC) and lymphoma.^29-31 In patients with AD, a small single study found no increased risk of lymphoma after 18 years of follow-up among users of AZT.³² 

Finally, there are currently no studies documenting an increased risk of lymphoma or NMSC in AD patients treated with mycophenolate mofetil monotherapy. Data on use of corticosteroids and cancer risk are also limited in this population despite these being used in the majority of patients with AD. Jensen and colleagues found no statistically significant associations with cSCC, melanoma or non-Hodgkin lymphoma but a small increased risk for basal cell carcinoma was observed (incidence rate ration=1.15; 95% CI, 1.07-1.25).³³ If this risk is attributable to drug exposure, disease severity, or other unmeasured factors remains to be seen. It should be emphasized that the use of long-term systemic steroids for AD is not currently recommended due to the increased risk of other more common adverse effects associated with chronic use. 

Our increased understanding of the inflammatory pathways in AD has led to the development of newer therapeutic options, which offer a more targeted approach to treatment. Dupilumab (Dupixent), the first biologic approved for adult patients with AD, exerts its effects by blocking the action of both IL-4 and IL-13.³⁴ In dupilumab clinical trials, no cases of malignancy have been reported thus far,³⁵ however, no animal studies have been conducted to evaluate the carcinogenic or mutagenic potential of dupilumab.³⁶ Topical phospodiesterase-4 (PDE4) inhibitors are also currently approved or in development for the treatment of AD. Crisaborole ointment 2% (Eucrisa), a PDE4 enzyme inhibitor, is approved for the treatment of AD in patients as young as 2 years of age. In clinical trials, no increased risk for malignancy was observed but the trials were short-term duration.³⁷ Janus kinase inhibitors are currently undergoing clinical trials for AD and while results are yet to be reported, currently available data on other populations, such as patients with rheumatoid arthritis, have shown a slight increased risk for lymphoma and skin cancer that remains to be further examined.³⁸

The debate on if patients with AD are at an overall increased risk of malignancy continues.  Also, it remains unclear if disease severity, treatment exposure, or both are drivers of this risk. Additionally, while no specific recommendations exist for cancer screening in the AD population, it would be reasonable to suggest that dermatologists should be familiar with at least a minimum of cancer screening guidelines as recommended by the US Prevention Services Task Force cancer screening guidelines (Table). 

Dr Chiesa Fuxench is an assistant professor of dermatology at the University of Pennsylvania Perelman School of Medicine in Philadelphia, PA. 

Disclosure:  The author reports no relevant financial relationships.