Inhibition of interleukin‐12 and/or interleukin‐23 for the treatment of psoriasis: What is the evidence for an effect on malignancy?

Abstract Immune cells and cytokines play an important role in the pathogenesis of psoriasis. Interleukin‐12 (IL‐12) and IL‐23 promote cellular responses mediated by T cells, which contribute to an inflammatory loop responsible for the induction and maintenance of psoriatic plaques. Antibodies that inhibit IL‐12/23 or IL‐23 are key treatment options for patients with psoriasis. IL‐12 and IL‐23 also play a key role in immune responses to infections and tumors. A growing body of information from clinical trials, cohort studies, postmarketing reports, genetic studies and animal models provides insights into the potential biological relationships between IL‐12/23 inhibition and malignancies. We summarize this information in tables and provide some context for the interpretation of these data with the goal of informing dermatologists who are using IL‐12/23 or IL‐23 inhibitors to treat patients with psoriasis.

of malignancy associated with blocking IL-12 and/or IL-23 in the treatment of psoriasis.

| S TRUC TURE AND B I OLOG I C AL EFFEC TS OF IL-1AND IL-23 IN PSORIA S IS
IL-12 and IL-23 are heterodimers, sharing a common p40 (beta chain) subunit that is combined with either a p35 alpha chain (IL- 12) or p19 alpha chain (IL-23; Figure 1). [13,18] IL-12 and IL-23 signal through heterodimeric receptors, both of which contain IL-12 receptor β1 (IL-12Rβ1), which is coupled with IL-12Rβ2 to form the IL-12 receptor and with IL-23R to form the IL-23 receptor. [17,18] Signalling, mediated through the Janus kinase-signal transducers and activators of transcription (Jak-STAT) pathway, ultimately results in IL-12 and IL-23 promoting the development of cellmediated responses driven by different subsets of T helper (T H ) cells. [17,18] IL-12 plays a key role in differentiation, maintenance and activity of immune cell subsets, including T H 1 cells (which produce interferonγ) and natural killer cells ( Figure 2). [12,16,24] IL-23 has a key role in maintenance and activity of IL-17-producing T H 17 cells and IL-22-producing T H 22 cells. [25] In turn, IL-17 induces activation and proliferation of keratinocytes, which produce inflammatory cytokines, including IL-23, leading to a self-amplificatory loop. [26,27] Consequently, antibodies targeted to IL-12 and IL-23 (p40 inhibitors) affect T H 1, T H 17 and T H 22 responses, whereas those targeted to IL-23 alone (p19 inhibitors) primarily affect T H 17 and T H 22 responses. [26,27]

| INHIB ITOR S OF IL-12/2OR IL-2FOR THE TRE ATMENT OF PSORIA S IS
The development of anti-IL-12/23 antibodies was originally initiated based on the observation that mice deficient in the IL-12p40 subunit (the subunit shared by IL-12 and IL-23) and those treated with neutralizing anti-IL-12p40 antibodies showed resistance to autoimmunity. [15] The discovery of IL-23 [18] followed by the molecular characterization of IL-23 [17] and other murine and genetic studies helped identify the key pathogenic role of IL-23 in psoriasis, as summarized by Gaffen et al. [25] It is now understood that IL-12 and IL-23 act on different components of the chronic inflammatory loop associated with the formation of psoriatic plaques. [28] An association between IL-12 and psoriasis is supported by the presence of T H 1 cells and interferonγ in psoriatic lesions. [29] However, recent data from a mouse model of psoriasis suggest that IL-12 may dampen skin inflammation in psoriasis by modulating IL-23-mediated inflammatory events, decreasing skin invasion by T H 17 cells and promoting an antiinflammatory genetic programme in keratinocytes. [30] Two antibodies targeting IL-12/23p40-ustekinumab and briakinumab-have been evaluated as treatments for psoriasis and other immune-mediated diseases. [31][32][33][34][35] Ustekinumab is the only IL-12/23p40 inhibitor for the treatment of moderate-to-severe plaque psoriasis and psoriatic arthritis approved by the Food and Drug Administration (FDA). [36] Clinical development of briakinumab was discontinued, thought to be because of safety concerns reported in the clinical trials, including cardiac events and malignancies. [31,37] Guselkumab was the first antibody specifically targeting TA B L E 1 Inhibitors of IL-12/23 or IL-23 licensed or in clinical development for the treatment of psoriasis  [86] Panaccione et al [87] Guselkumab Fully human IgG1λ monoclonal antibody Binds to IL-23p19 subunit Janssen Biotech Inc. Reich et al, [11] Gordon et al [88]  Reich et al, [11] Papp et al [89] Risankizumab Krueger et al, [90] Papp et al, [39] Singh et al [91] Mirikizumab Humanized monoclonal antibody Blocks IL-23 Eli Lilly and Company Eli Lilly and Company [92] Ig, immunoglobulin; IL, interleukin. a US and European license applications withdrawn by manufacturer in 2011. [37] IL-23p19 to be approved for the treatment of moderate-to-severe psoriasis [38] , and 3 further IL-23p19 inhibitors are currently in active development for the same indication. Efficacy and safety data have been published for tildrakizumab (phase 3 [11] ) and risankizumab (phase 2 [39] ); mirikizumab (LY3074828) is currently entering phase 2 development. [40]

| MALI G NAN CIE S REP ORTED IN CLINIC AL TRIAL S
A variety of cancer types have been reported in clinical trials of IL-12/23 and IL-23 inhibitors ( Table 2). NMSCs were the most frequently reported malignancies. These are the most common malignancies in humans (albeit not routinely reported to cancer registries), with basal cell carcinomas (BCCs) more common than squamous cell carcinomas (SCCs). [41] A meta-analysis of observational studies found that the risk of SCCs was increased in patients with psoriasis compared with the general population (standardized incidence ratio [SIR] = 5.31, 95% confidence interval [CI] = 2.63-10.71) and correlated with patient exposure to 8-methoxypsoralen-ultraviolet (UV) A therapy for treatment of psoriasis. [9] Risk of BCCs was also increased in patients with psoriasis, but to a lesser extent than SCCs (SIR = 2.00, 95% CI = 1.83-2.20). [9] In trials of ustekinumab for psoriasis, as in the general population, [41] the proportion of BCCs was higher than that of SCCs. [32,[42][43][44] However, a pooled analysis of safety data from all briakinumab phase 2 and phase 3 trials and interim data from an open-label extension trial suggested that the risk of SCC was similar to the risk of BCC in patients treated with briakinumab, which may suggest a relative or absolute increase in the risk of SCC. [45] Concern about the effect of briakinumab on NMSCs was thought to be one of the reasons for discontinuing its development. [31,37] The clinical evaluation of IL-23p19 inhibitors is ongoing, and data available thus far are limited. NMSCs have been reported in some clinical trials. [10,11,39,46] Publications about tildrakizumab reported low numbers (7 cases total) of NMSCs but did not differentiate BCCs and SCCs. [11] Two cases of BCCs were reported in the phase 2 trial of risankizumab. [39] F I G U R E  [93] (Continues) Prostate and breast cancers are the most common internal malignancies in men and women, respectively. [47] Patients with psoriasis have not been found to have a significantly increased risk of prostate cancer compared with the general population, [7,8] but the relative risk of breast cancer in patients with psoriasis is less clear. Two analyses showed no significant increased risk, [7,8] and one showed a slightly increased risk compared with the general population. [9] Several cases of prostate and breast cancers occurred in trials of ustekinumab, [32,[42][43][44] briakinumab [34] and guselkumab. [10,46] These malignancy events reported in clinical trials do not prove causation but do suggest a possible biological relationship that may trigger further investigation. [48]

| MALI G NAN CIE S REP ORTED IN P OS TMARK E TING SAFE T Y DATA
Ustekinumab, which has been approved since 2009, is currently the only IL-12/23 or IL-12 inhibitor with postmarketing safety data.
The prescribing information (PI) for ustekinumab contains a general warning that it "may increase risk of malignancy," based on the observations that (i) NMSCs were reported in 1.5% of patients and malignancies excluding NMSCs (non-NMSCs) were reported in 1.7% of patients among patients treated with ustekinumab (3.2 years' median follow-up); (ii) the most frequently observed non-NMSCs were prostate, melanoma, colorectal and breast cancers, but they were similar in type and number to those expected in the general US population when adjusted for age, gender and race; and (iii) rapid appearance of multiple cutaneous SCCs was found in postmarketing reports among patients receiving ustekinumab who had pre-existing risk factors for developing NMSC. [36] The concerns raised in the ustekinumab PI are in line with an analysis of postmarketing safety data reported to the FDA, which found that patients treated with ustekinumab were 15 times more likely to report a case of cancer than were patients treated with apremilast, a phosphodiesterase 4 (PDE-4) inhibitor. [49] Furthermore, a safety signal was detected in a study of data from the FDA Adverse Event Reporting System database, which indicated that ustekinumab may be associated with several malignancies, including B-cell lymphoma; epithelioid sarcoma; and lung, oesophageal, ovarian, renal, testis and thyroid cancers. [50] In contrast, an analysis of data from the Psoriasis Longitudinal  [52] Evidence from controlled clinical trials and registries indicates that ustekinumab is well tolerated, with rates of overall mortality and malignancy comparable with those expected in the general population. [32,33,53,54] Although the postmarketing pharmacovigilance studies mentioned previously were not designed to study causality or to quantify increased cancer risk associated with specific therapies, [49,50] they can be helpful to identify safety signals that may be relevant to the advancement of overall patient safety. [55] Postmarketing safety data are not yet available for tildrakizumab, guselkumab or risankizumab. Long-term data are needed to identify any potential associations between IL-23 inhibitors and malignancies.
Clinical trials detect adverse events during trials of relatively short duration. For example, the trials of IL-12/23 and Table 2 varied in duration from 12 to 76 weeks. Consequently, drugs are usually made available for public use before rare but potentially serious adverse reactions have been identified and their probabilities quantified. [48] Although NMSCs may develop within weeks to months [56] and are relatively easy to detect by visual inspection, other malignancies may take longer to be discovered. Thus, the incidence of

| RIS K OF MALI G NAN C Y IN H UMAN S WITH G ENE TI C TR AITS S IMUL ATING NEUTR ALIZ ATI ON OF IL-12 AND/OR IL-23
In the first report of cancer in a patient with IL-12Rβ1 deficiency (the receptor subunit required to bind the p40 subunit shared by IL-12 and IL-23), the patient developed oesophageal SCC at age 25 years and died at age 29 years, an age at which this cancer is exceedingly rare (Table 3). [57] Additional studies of IL-12 genetic deficiencies have investigated whether such deficiencies are associated with increased likelihood of infections and cancer. [58,59] Genomewide association studies have shown that polymorphisms in genes encoding the IL-12p40 subunit or the IL-12p35 subunit, which result in a decreased biological effect of IL-12, are linked to increased susceptibility to oesophageal cancer, [60] osteosarcoma, [61] bladder cancer [62] and prostate cancer, [63] as well as susceptibility to and/or severity of cervical cancer. [ Increased risk of osteosarcoma [61] and oesophageal [60] and prostate [63] cancers IL-12/23 inhibitors may increase risk of osteosarcomas and bladder, cervical, oesophageal and prostate cancers 1188 AC vs AA Increased risk of bladder cancer [62] rs2569254 GG vs AA Increased risk of cervical cancer [64] IL-12A (IL-12p35) polymorphisms: rs568408 GA/AA or GA vs GG Decreased IL-12 levels Increased risk of oesophageal cancer [60] and osteosarcoma [61] IL-12/23 inhibitors may increase risk of osteosarcomas and oesophageal cancer in specific patient populations IL-23R polymorphisms: rs6682925 TC/CC or TG/GG or T>C rs1884444n T>G

Loss of IL-23 function
Increased risk of oesophageal cancer, [66] hepatocellular carcinoma, [67] and acute myeloid leukaemia [68] No association with risk of gastric cancer [65] Decreased risk of gastric cancer [65] IL-23 inhibitors may affect cancer risk of some cancers in specific patient populations IL-23 inhibitors may decrease risk of gastric cancer IL, interleukin; R, receptor.

| 743
ERGEN aNd YUSUF associated with a significantly reduced risk of gastric cancer, [65] but with a significantly increased risk of oesophageal cancer, [66] hepatocellular carcinoma [67] and acute myeloid leukaemia. [68] However, the effect of these IL-23R variants on the function of IL-23 (eg, gain, loss or no effect) was not specifically described in the studies. Taken together, these findings might lead to the hypothesis that IL-12/23 inhibitors have the potential to increase the risk of these cancers. However, a limitation of genomewide association studies is that they are not designed to investigate the causal relationship between a specific polymorphism and an increased cancer risk. For example, although several studies had shown that the TNF-238 polymorphism increased cancer risk, a meta-analysis of 34 studies did not find a significant association between this polymorphism and increased cancer risk. [69]

| RIS K OF MALI G NAN C Y IN ANIMAL MODEL S S IMUL ATING NEUTR ALIZ ATION OF IL-12 AND/OR IL-23
The malignancy data from animal models of IL-23 deficiency are conflicting. Mice that had lost IL-23 function via deficiencies in either IL-23p19 or IL-23R or by treatment with antibodies to IL-23p19 showed resistance to skin tumor growth/development (Table 4). [20] IL-23deficient mice [70] and mice treated with anti-IL-23p19 [71] have also been shown to have an increased resistance to melanoma-induced lung metastases. Furthermore, in this model of melanoma-induced metastases, anti-IL-23 antibody used in combination with IL-2 or anti-erbB2 antibody significantly inhibited subcutaneous growth of established mammary carcinomas and suppressed established and spontaneous lung metastases. [71] Deficiencies in IL-23p19 or IL-23R also resulted in decreased tumor multiplicity and growth in a mouse model of colorectal tumors. [72] These findings suggest that IL-23p19 inhibitors might prevent the growth and/or enhance the rejection of some tumors, possibly via effects on IL-22, which has been implicated in the development of epithelial tumors. [73,74] A number of studies have found that increased levels of IL-23 are associated with unfavourable outcomes in various malignancies in humans. [75][76][77][78][79] In contrast, other studies suggest that IL-23p19 deficiency might enhance the risk of certain cancers. For example, IL-23-deficient mice demonstrated an increased risk of development of chemically induced melanoma. [80] However, in a model of UV radiation, IL-23-deficient mice demonstrated both an increased risk of developing sarcoma and TA B L E 4 Malignancies in murine models of IL-23 deficiency

Effect on malignancy vs controls Potential therapeutic implications for IL-23 inhibitors
Treatment with anti-IL-23p19 antibody [20] Loss of IL-23 function a decreased risk of developing epithelial tumors compared with wildtype mice. [19] Further studies are needed to confirm this finding.
Similarly, the studies on IL-12 also show conflicting data. Several studies of mice with IL-12-specific loss of function via deficiency in IL-12p35 [20] or IL-12Rβ2 [81] showed an increased risk of tumor development (Table 5), but, in models of UV radiation [19] or chemically induced melanomas, [80] IL-12p35-deficient mice had the same risk of induced skin tumors as did their wild-type counterparts. Mouse models with loss of function in both IL-12 and IL-23 via IL-12/23p40 deficiency or treatment with anti-IL-12/23p40 antibodies showed an increase, [20,21,80] a decrease [82] or no difference [83] in tumor development (Table 6).

| CON CLUS IONS
Patients with psoriasis and/or receiving treatment for psoriasis have an increased risk of cancer. [8,9,49,84] Inhibitors of IL-12/23 and IL-23 are effective treatment approaches for psoriasis. [36,38] Existing data provide evidence to support an association between impaired IL-12 and/or IL-23 signalling and both tumor growth and resistance to tumor growth, although the nature of these relationships is not fully understood. Long-term postmarketing safety evaluations of agents targeting IL-12/23 and IL-23 are needed to fully appreciate the associated malignancy risk. The implications for therapeutic inhibition of IL-12/23 or IL-23 remain uncertain, although monitoring of patients for NMSC and malignancy seems warranted.

ACK N OWLED G EM ENTS
Both authors have read and approved the final manuscript. This work was partially supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases 1R01AR071157-01A1 to NY.
Medical writing and editorial assistance were provided by Catherine Champagne, of Fishawack Communications, and supported by Sun Pharmaceutical Industries, Inc.

CO N FLI C T O F I NTE R E S T
The authors have declared no conflicting interests.