LAUSR

Compared with conventional systemic treatment modalities such as methotrexate, the superior efficacy and absence of cumulative toxicity have led to a widespread uptake and use of biologics for treatment of psoriasis and a number of other chronic inflammatory diseases. As with any therapy intended for long-term use, efficacy and safety must be carefully balanced when choosing the most appropriate drug for the individual patient. The rapidly evolving landscape of biological treatment options for inflammatory diseases may provide not only opportunities for hitherto unimagined levels of efficacy but could at the same time pose new challenges as certain safety signals may only be found after long-term use in large populations. Efalizumab, a recombinant humanized monoclonal antibody, was withdrawn from the market in 2009 after reports of progressive multifocal leucoencephalopathy development following long-term treatment with this drug. Although just four cases of progressive multifocal leucoencephalopathy among patients with psoriasis treated with efalizumab were reported, the seriousness of this adverse event was in stark contrast to the risk willingness of most dermatologists and patients with psoriasis. On the other hand, a recent phase III trial reported that treatment-emergent conjunctivitis occurred in 28% of patients treated with dupilumab, an interleukin-4 receptor alpha antagonist used for treatment of atopic dermatitis. Broadly speaking, adverse events may be acceptable if they are common but mild, or severe but rare. In some cases, they may even be acceptable when they are both severe and common, if there are no other viable treatment alternatives. Recently, data from the phase III development programme for the interleukin-23 inhibitor guselkumab showed an increased incidence of male breast cancer compared with patients with psoriasis in general. Importantly, the absolute number of events was very low, and it is likely that this was a random finding, yet whether a true risk exists for such rare events cannot be properly determined in a randomized controlled trial. On the other hand, biologic registries are specifically designed to monitor safety signals, and thus provide a unique opportunity to examine rare events over extended periods of time using nonrandomized observational data. In this issue of the BJD, Garcia-Doval and colleagues undertook the ambitious task of combining registry data from four different countries through the Psonet collaboration. Using combined data from Israel, Italy, Spain, the U.K. and Republic of Ireland, the authors performed parallel case–control studies of patients with psoriasis with first-time cancer and examined the association with previous exposure to biologics. After meta-analysing these data, the authors found no significant association between cancer and biologics use. Regrettably, the authors combined not only all biological agents (i.e. adalimumab, infliximab, etanercept and ustekinumab) into a fixed exposure, but also pooled multiple different types of cancers into a composite cancer estimate. It has been convincingly demonstrated in previous studies that targeted inhibition of selected cytokine signalling pathways is associated with a dose–response protective effect for example on the incidence and mortality from lung cancer. On the other hand, tumour necrosis factor (TNF) is a pleiotropic cytokine with both proand antitumoral effects, and use of biologics such as TNF-inhibitors may therefore potentially increase the risk of certain types of cancers, while reducing the risk of other cancers. Theoretically, one biological agent may increase the risk of some cancer types, while another biologic could protect against different cancers. By using composite exposures and outcomes, such increases and decreases in risk could cancel each other out, leading to a false null finding. Moreover, because of differences in data quality and completeness, the four included registries represented somewhat heterogeneous data sources. In meta-analyses, study groups are carefully evaluated to assess whether or not those groups can be pooled, and study heterogeneity is scrutinized before making any firm conclusive statement. This is also the case for the present study, and the authors are commended for their contribution to the literature, in which clinicians may gain comfort from knowing that no major malignancy risk signals have been shown so far with use of established biological therapies for psoriasis. Notwithstanding this, it is important to remember that absence of evidence is not evidence of absence, and patients may be funnelled into other nonbiological treatment options if they are deemed to be at high risk of cancer development, thus leading to a healthy user effect. However, for now at least, biologics as a group appear to be an oncologically safe treatment option for patients with psoriasis without a history of previous malignancy, and the absolute cancer risk associated with use of biologics, whether causal or not, is likely very small. Over the past two decades, biologics have revolutionized our approach to treating moderate-to-severe psoriasis, and the clear benefits of treating psoriasis with these drugs likely outweighs most concerns for risk of developing firsttime cancer.