Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?

Cytokine xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Cytokine journal homepage: www.journals.elsevier.com/cytokine

Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders? Yuehong Chen a, Marcia Friedman b, Gang Liu a, Atul Deodhar b, Cong-Qiu Chu b,c,⇑ a

Department of Rheumatology and Immunology, West China Hospital, Sichuan University, 37 Guoxue Xiang, Chengdu, Sichuan 610041, China Division of Arthritis and Rheumatic Diseases, Oregon Health & Science University, Portland, OR 97239, United States c VA Portland Health Care System, Portland, OR 97239, United States b

a r t i c l e

i n f o

Article history: Received 17 July 2016 Received in revised form 17 September 2016 Accepted 19 September 2016 Available online xxxx Keywords: Tumor necrosis factor inhibitors Immune-mediated inflammatory diseases Malignancy

a b s t r a c t Inhibition of tumor necrosis factor (TNF) activity has profoundly changed the management of several immune-mediated inflammatory diseases with great benefit for patients. The application of TNF inhibitors (TNFi), however, also brings a new concern, malignancy. We performed a systemic review to collect the studies reporting cancer incidences and risks in TNFi users regardless of indications. TNFi were most frequently used in treating patients with rheumatoid arthritis (RA) and inflammatory bowel diseases (IBD). In RA patients without prior cancer history, the incidences of malignancies ranged from the lowest rate 0 per 1000 person-years in etanercept users regarding lymphoma to the highest rate 35.62 per 1000 person-years in adalimumab users on non-melanoma skin cancer (NMSC), while in those patients with prior cancer history, the recurrent incidences of malignancies ranged from the lowest rate 5.05 per 1000 person-years regarding melanoma to the highest rate 63.20 per 1000 person-years on basal cell carcinoma (BCC) in TNFi users. In IBD patients, incidences ranged from 0 per 1000 person-years in TNFi users on lymphoma to 34.0 per 1000 person-years in infliximab users on overall cancer. However, these incidence rates of overall cancer, lymphoma and melanoma were not higher in comparison with those patients who were not treated with TNFi. Compared to general population, incidences of lymphoma were elevated in RA patients and rates of NMSC were higher in patients with psoriasis, RA and IBD. In conclusion, cancer incidences vary across different studies, indications, cancer types and studies with different individual TNFi. Treatment with TNFi is not associated with increased malignant risks of overall cancer, lymphoma or melanoma. Results of NMSC risk were inconsistent among studies. A latest prospective registry study demonstrated a small increased risk of squamous cell cancer in RA patients treated with TNFi (one additional case for every 1600 years of treatment experience). Further prospective studies are needed to verify whether TNFi users have higher NMSC risk than non-TNFi users. Ó 2016 Published by Elsevier Ltd.

1. Introduction Tumor necrosis factor (TNF), a pro-inflammatory cytokine, is mainly generated by macrophages, monocytes, and T-cells, and plays a crucial role in inflammatory response and innate immunity. Its receptors are TNF receptor type 1 (TNFR1) (also known as TNFR p55) expressed on the surface of almost all human cell types excluding erythrocytes, and TNFR2 (also called TNFR p75) expressed chiefly on immune and endothelial cells. After binding to the two different receptors, TNF induces variable biological effects [1]. However, excessive expression of TNF and its receptors ⇑ Corresponding author at: Division of Arthritis and Rheumatic Diseases, Oregon Health & Science University and VA Portland Health Care System, Portland, OR 97239, United States. E-mail address: [email protected] (C.-Q. Chu).

in the regulation process of inflammatory response and immune reaction is implicated in the pathogenesis and development of immune-mediated inflammatory diseases. TNF was first identified as an upstream cytokine in the inflammatory cytokine cascade in rheumatoid arthritis (RA) and was the first cytokine identified as therapeutic target in the treatment of a number of inflammatory diseases [2–5]. Since 1998, five branded TNF inhibitors (TNFi) have been licensed for clinical use, namely, infliximab (IFX), etanercept (ETN), adalimumab (ADA), certolizumab pegol (CZP) and golimumab (GLM). Most recently, an IFX biosimilar is also approved by US Food and Drug Administration to market in the US. Outside US, ETN biosimilars have been prescribed for many years. IFX, ADA, CZP and GLM are monoclonal antibodies. IFX, the prototype TNFi, is a chimeric; ADA is a humanized, while GLM is a full human monoclonal antibody. Certolizumab pegol is a humanized, PEGylated

http://dx.doi.org/10.1016/j.cyto.2016.09.013 1043-4666/Ó 2016 Published by Elsevier Ltd.

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

2

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

Fab fragment of monoclonal antibody. Whereas, ETN is a fusion protein of soluble TNFR2 and human IgG1 Fc fragment. While all the monoclonal antibodies specifically bind to TNF, ETN binds to TNF and lymphotoxin. The clinical efficacy of all these TNFi is comparable. Collectively, the indicated conditions for TNFi include RA, juvenile idiopathic arthritis (JIA), Crohn’s disease (CD), ulcerative colitis (UC), ankylosing spondylitis (AS), psoriasis (PsO), psoriatic arthritis (PsA) and non-infective uveitis. Although TNFi are a highly effective therapy that has drastically improved the care of many patients with immune-mediated inflammatory diseases, the risk of adverse events has continued to be a source of worry for patients and their providers. In particular, the risk of malignancy has been a major source of concern and debate. With increased passage of time and more widespread use of these medications, more robust and longer term data are available regarding the incidence and risk of malignancy in patients using these drugs. In particular, analyses from TNFi registries provide data reflecting the situation in daily clinical practice. This review is performed in order to summarize these data and our understanding to date of the risk of malignancy associated with the use of TNFi therapy. Since IFX, ETN and ADA have been in use for longer durations, most of the studies are derived from these TNFi. 2. Methods 2.1. Literature search We systematically searched the electronic databases of OVID Medline and OVID EMBASE (from January 1st, 2000 to February 28, 2016) using a combination search of MeSH terms and keywords. Our search terms were ‘‘tumor necrosis factor-alpha”, ‘‘tumor necrosis factor alpha antibody”, ‘‘adalimumab”, ‘‘certolizumab pegol”, ‘‘etanercept”, ‘‘golimumab”, ‘‘infliximab”, ‘‘malignancy”, ‘‘cancer”, ‘‘neoplasm”, ‘‘carcinoma”, ‘‘cancer epidemiology”, ‘‘prevalence rate” and ‘‘incidence rate”. We limited search results to human. 2.2. Study selection criteria We included the studies, containing original studies and studies used pooling analysis for data (i.e., meta-analysis), reported incident rates of overall malignancy, lymphoma, non-melanoma skin cancer (NMSC) and/or melanoma in TNFi users no matter what indications. TNFi should be ADA, CZP, ETN, GLM, IFX or the combination use of any these agents. The studies were excluded if they met any one of the following statements: not related to incident rates of malignancy; not correlated with TNFi therapy; could not acquire useful data; duplications (i.e., data from the same database for the same outcomes); conference abstracts, letters, narrative reviews and non-English publications. 2.3. Data extraction We extracted the following information from included studies: first author, year of publication, country, treatment indications, data sources, patient number and duration of follow-up (median years) in TNFi group and control group; malignancy type, cancer cases, follow-up duration (patient-years) and crude incident rate (per 1000 patient-years) in TNFi group. We also collected the relative risk (RR), hazard ratio (HR), incidence rate ratio (IRR), odds ratio (OR) and/or standardized rate ratio (SIR) of TNFi in contrast with control group or general population. If there were several studies from the same registry or database, we collected the longest follow-up results in a study and outcomes of the most recent report with the largest study population.

2.4. Data processing Incident rates were collected according to the report of a study. If a study only presented cancer number and follow-up duration, incident rate would be calculated via cancer number divided by follow-up patient-years and then multiplied 1000, acquiring the unit per 1000 patient-years. Relative risk and adjusted relative risk (aRR), hazard ratio and adjusted hazard ratio (aHR), incidence rate ratio (IRR) and standardized rate ratio (SIR) were acquired according to the reports of the eligible studies. RR, HR, IRR and SIR were reported for the comparisons of TNFi users to control group patients with the disease not treated with biologics or treated only with traditional disease-modifying anti-rheumatic drugs (DMARDs), and to general population. 3. Results A total of 1900 studies were retrieved from databases. After selection, 53 studies [6–58] proved eligible. Additionally, 3 studies [59–61] were included through manual searching references and 3 recently published studies [62–64] were also added. Of the 59 studies, 41 were original studies (37 were cohort studies, data in two studies were from clinical trials’ database and post marketing surveillance database, and the remaining two studies: one was the Rheumatoid Arthritis Prevention of Structural Damage (RAPID)1 trial and its open-label extension study, the other was open-label extension study of randomized controlled trial) and 18 were pooled analysis (11 were integrated studies from clinical trials, two collaborative studies from biologic registers and five were meta-analyses). Of all the included studies, 27 focused on RA, 11 focused on inflammatory bowel disease (IBD) including CD and UC, three on JIA, three on PsO, one on spondyloarthropathies (SpA), one on AS, and 13 on blended rheumatic conditions. The median follow-up duration ranged from 0.2 years to 6.8 years in TNFi group (from 0.2 years to 12.0 years in control group) in original studies. Of the 59 eligible studies, 22 studies were performed in USA, seven in Sweden, five in UK, four in China, three each in Belgium and Germany, two each in Denmark, Finland, France, Italy, the Netherlands and Canada, and one each in Australia, Japan, Spain (Supplementary Tables 1 and 2). 3.1. Cancer incidence and risk in RA patients in original studies 3.1.1. Cancer incidence and risk in RA patients without prior cancer history In RA patients, cancer incidences varied among different studies, disease indication, TNFi, and cancer type, ranging from the lowest rate 0 per 1000 person-years in ETN users regarding lymphoma and overall cancer [19] to the highest rate 35.62 per 1000 personyears in ADA users on NMSC [59] (Table 1). For RA patients using TNFi regardless of subtypes, the overall cancer incidences varied from 5.1 per 1000 person-years (44 cancer cases and 8558 person-years of follow-up) [23] to 13.8 per 1000 person-years (34 cases and 2465.3 person-years) [9]. Berghen et al. [9] and Pallavicini et al. [22] reported gender subgroup incident rates of overall higher malignancy incidence rates for males than females, with 21.79 and 25.82 respectively for males, and 11.24 and 5.23 respectively for females. With respect to lymphoma in RA patients on TNFi, the highest reported incidence rate was 3.12 per 1000 person-years [14] and the lowest report was 0.8 per 1000 person-years [16]. Askling et al. [7] and Pallavicini et al. [22] both reported higher lymphoma incidences in men (1.37 (9 cases in 6544 person-years) and 8.61 (3/348.57) respectively) than in women (0.83 (17/20436) and 0.58 (1/1702.08) respectively).

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

3

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx Table 1 Cancer incidences and risks in TNFi treated RA patients in original studies. Studies

Cancer types: malignancies (NO.)/follow-up durations (patient-years), crude incidences (per 1000 patient-years)

Cancer incidences in RA patients treated with TNFi Aaltonen [6] Overall cancer: 47/7162, 6.6 Askling (overall cancer) [8] Overall cancer: 240/25693, 9.34

Berghen [9] Buchbinder [10] Dreyer [37] Haynes [38] Kievit [16] Pallavicini [22] Strangfeld [23] Wu [25] Berghen [9] Askling (lymphoma) [7] Berghen [9] Geborek [14] Haynes [38] Kievit [16] Berghen [9] Pallavicini [22] Amari [59]

Overall cancer: 34/2465.3, 13.8 Overall cancer: 44/5752, 7.65 Overall cancer: NA Overall cancer: 9.1 Overall cancer: 28/5017, 5.6 Overall cancer: 18/2068.85, 8.7 Overall cancer: 44/8558, 5.1 Overall cancer: 89/16624.36, 5.35 Solid cancer: 26/2465.3, 10.55 Lymphoma: 26/26981, 0.96 Lymphoma: 6/2465.3, 2.433 Lymphoma: 5/1603, 3.12 Lymphoma: 1.2 Lymphoma: 4/5017, 0.8 NHL: 3/2465.3, 1.22 NHL: 4/2068.85, 1.93 NMSC: 283/11084, 25.5

Haynes [38] Kievit [16] Mercer [20] Raaschou [64] Buchbinder [10] Kievit [16] Komano [17]

NMSC: 3.1 NMSC:4/5017, 0.8 BCC: 150/43798, 3.42; SCC: 23/43798, 0.53 BCC: 236/37080, 6.36; SCC: 191/74541, 2.56 Melanoma: 10/5752, 1.74 Melanoma: 2/5017, 0.4 Neoplasms (benign, malignant, unspecified): 4/592.4, 6.75

Cancer incidences in female RA patients treated with TNFi Askling (overall cancer) [8] Overall cancer: NA Berghen [9] Overall cancer: 11.24 Dreyer [37] Overall cancer: NA Pallavicini [22] Overall cancer: 9/1702.08, 5.23 Berghen [9] Solid cancer: 8.562 Askling (lymphoma) [7] Lymphoma: 17/20436, 0.83 Berghen [9] Lymphoma: 2.68 Pallavicini [22] NHL: 1/1702.08, 0.58 Raaschou [64] SCC: NA BCC: NA Cancer incidences in male RA Askling (overall cancer) [8] Berghen [9] Dreyer [37] Pallavicini [22] Berghen [9] Askling (lymphoma) [7] Berghen [9] Pallavicini [22] Raaschou [64]

patients treated with TNFi Overall cancer: NA All cancer: 21.79 Overall cancer: NA Overall cancer: 9/348.57, 25.82 Solid cancer: 16.764 Lymphoma: 9/6544, 1.37 Lymphoma: 1.68 NHL: 3/348.57, 8.61 SCC: NA BCC: NA

Indicators and 95% confidence interval (TNFi vs control)

SIR, 1.2 (0.63–2.2) aRR (vs anti-TNFi naïve patients), 1.00 (0.87–1.17) aRR (vs starting MTX), 0.99 (0.79–1.24) aRR (vs starting non-biologic DMARDs combination therapy), 0.97 (0.69–1.36) NA aRR, 0.65 (0.37–1.17) HR, 1.02 (0.80–1.30) HR, 0.80 (0.59–1.08) NA NA aHR, 0.70 (0.44–1.12) IRR, 0.72 (0.58–0.91) NA aRR, 1.35 (0.82–2.11) NA aHR, 5.0 (0.9–27.9) HR, 0.83 (0.33–2.05) NA NA NA aHR (vs non-biologic DMARDs), 1.42 (1.24–1.63) aHR (vs MTX), 1.42 (1.23–1.65) HR, 0.83 (0.49–1.42) NA aHR, BCC: 1.20 (0.83–1.73); SCC: 1.79 (0.59–5.41) aHR, BCC: 1.14 (0.98–1.33); SCC: 1.30 (1.10–1.55) SIR, 0.54 (0.12–2.40) NA NA aRR, 1.06 (0.89–1.26) NA HR, 1.13 (0.83–1.53) NA NA aRR, 1.31 (0.75–3.30) NA NA aHR, 1.52 (1.25–1.84) 1.06 (0.89–1.27) aRR, 0.89 (0.68–1.16) NA HR, 0.83 (0.55–1.26) NA NA aRR, 1.51 (0.69–3.29) NA NA aHR, 1.26 (0.92–1.70) aHR, 1.38 (1.03–1.86)

Cancer incidences in RA patients with prior cancer history treated with TNFi Dixon [13] Recurrent overall cancer: 13/515, 25.2 Strangfeld [23] Recurrent overall cancer: 9/198, 45.5 Recurrent melanoma: 1/198, 5.05 Raaschou [64] Recurrent SCC: 10/390, 25.64 Recurrent BCC: 17/269, 63.20

SIR, 0.58 (0.23–1.43) IRR, 1.4 (0.5–5.5) NA aHR, 0.99 (0.44–2.10) aHR, 1.19 (0.67–2.15)

Cancer incidences in RA patients treated with ADA Aaltonen [6] Overall cancer: 16/2620, 6.1 Askling (overall cancer) [8] Overall cancer: 26/2160, 12.04 Burmester [11] Overall cancer: 179/18272, 9.8 Pallavicini [22] Overall cancer: 8/527.22, 15.17 Wu [25] Overall cancer: 8/2203.45, 3.63 Mercer [21] Solid cancers: 139/18027, 7.7 Burmester [11] Lymphoma: 15/18272, 0.82 Chiu [12] Lymphoma: 1/697, 1.44 Pallavicini [22] NHL: 2/527.22, 3.8 Amari [59] NMSC: 92/2583, 35.62 Burmester [11] NMSC: 43/18272, 2.35 Mercer [20] BCC: 34/12991, 2.62

SIR, 1.1 (0.51–2.2) aRR, 1.32 (0.87–1.98) NA NA IRR, 0.62 (0.30–1.31) aHR, 0.84 (0.66–1.07) NA NA NA NA NA aHR, 0.89 (0.56–1.42) (continued on next page)

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

4

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

Table 1 (continued) Studies

Cancer types: malignancies (NO.)/follow-up durations (patient-years), crude incidences (per 1000 patient-years)

Cancer incidences in RA patients treated with ETN Aaltonen [6] Overall cancer: 21/2842, 7.4 Askling (overall cancer) [8] Overall cancer: 70/9413, 7.43 Lovell [19] Overall cancer: 0/318, 0 Pallavicini [22] Overall cancer: 4/433.31, 9.23 Wu [25] Overall cancer: 71/10363.75, 6.85 Chiu [12] Lymphoma: 3/3132, 0.96 Lovell [19] ETN: lymphoma 0/318, 0. Pallavicini [22] NHL: 0/433.31, 0 Wolfe [24] NHL: 4/3474, 1.15 Amari [59] NMSC: 145/6827, 21.24 Lebwohl [18] SCC (data from clinical trials): 4/4257, 0.9; SCC (data from post marketing database): 25/250000, 0.1 Mercer [20] BCC: 57/19108, 2.98

Indicators and 95% confidence interval (TNFi vs control)

SIR, 1.3 (0.65–2.6) aRR, 0.78 (0.61–1.00) NA NA IRR, 0.91 (0.71–1.18) NA NA NA NA NA NA aHR, 1.07 (0.70–1.63)

Cancer incidences in RA patients treated with IFX Aaltonen [6] Overall cancer: 10/1700, 5.9 Askling (overall cancer) [8] Overall cancer: 144/14120, 10.20 Pallavicini [22] Overall cancer: 6/1108.12, 5.41 Amari [59] NMSC: 47/1674, 28.08 Mercer [20] BCC: 59/11700, 5.04 Wolfe [24] NHL: 6/4913, 1.22 Pallavicini [22] NHL: 2/1108.12, 1.8

SIR, 1.2 (0.44–3.1) aRR, 1.09 (0.91–1.30) NA NA aHR, 1.73 (1.14–2.62) NA NA

Cancer incidences in RA patients treated with CZP Keystone [15] CZP: malignancies excluding NMSC 23/3732, 6.2; CZP: all neoplasms (benign, malignant and unspecified) 105/3732, 28.1.

NA NA

NO. = number; NA = not acquired; RA = rheumatoid arthritis; TNFi = tumor necrosis factor inhibitors; IFX = infliximab; ETN = etanercept; ADA = adalimumab; CZP = certolizumab pegol; GLM = golimumab; MTX = methotrexate; DMARDs = disease-modifying anti-rheumatic drugs; IRR = incidence rate ratio; SIR = standardized rate ratio; aRR = adjusted relative risk; HR = hazard ratio; aHR = adjusted hazard ratio; NHL = non-Hodgkin lymphoma; NMSC = non-melanoma skin cancer; SCC = squamous cell carcinoma; BCC = basal cell carcinoma.

Regarding melanoma incidences in RA patients on TNFi, Buchbinder et al. [10] and Kievit et al. [16] reported incidence rates of 1.74 (10/5752) and 0.4 (2/5017) respectively. Five studies assessed the incident rates of NMSC in TNFi users. Three studies reported rates of NMSC were 0.8 (4/5017) [16], 3.1 [38] and 25.5 (283/11084) [59] respectively. Mercer et al. [20] and Raaschou et al. [64] reported basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) of NMSC separately, showing higher incidence rates on BCC (3.42 and 6.36) than on SCC (0.53 and 2.56). Compared to control group patients with the same disease but not treated with biologics, the majority of studies did not find significant differences in risk of overall cancer, lymphoma, and melanoma. In a nationwide population based prospective cohort study from Sweden, a 50% increased risk of invasive melanoma was observed [65], but this was not confirmed by a latest collaborative study including 11 European biologic registries [63]. As for NMSC risk in TNFi users compared to non-TNFi users, inconsistent results were reported among different studies [20,38,59,64]. A prospective observational study based on data from British Society for Rheumatology Biologics Register (BSRBR) showed an increased risk of non-melanoma skin cancers in treated RA patients, but the risk was not exacerbated by TNFi [20]. A collaborative safety study was performed in the US by combining data from four large cohorts. In this Safety Assessment of Biological Therapeutics (SABER) retrospective study, patient population was extended to include RA, IBD, PsO, PsA and AS. No increased incidence of cancers was observed in patients who were exposed to TNFi [38]. In contrast, data from the national administrative databases of the Department of Veterans’ Affairs (VA), it was observed that there was a higher risk of developing NMSC in RA patients who were treated with TNFi (HR 1.42; 95% CI 1.24, 1.63) [59]. It was further noted that risk factors for NMSC included older age, male gender, non-steroidal anti-inflammatory drugs and glucocorticoid use and a history of prior malignancies. The most recently

published cohort study based on nationwide prospectively recorded data from Sweden reported a small to moderately increased risk of basal cell cancer in biologics-naïve RA patients with no further increase by TNFi use. For squamous cell cancer, the risk was nearly doubled in biologics-naïve RA patients compared with general population, with a further 30% increase in risk among patients treated with TNFi [64] (Table 1). Among RA patients on ADA, the highest reported cancer incidence was 15.17 (8/527.22) for overall cancer [22] and the lowest was 0.82 (15/18272) per 1000 person-years for lymphoma [11], showing no incidence difference (SIR = 1.1, 95%CI 0.51– 2.2) when compared to control RA group [6]. For RA patients on ETN, the incidences of cancer ranged from 0 (0/318) regarding overall cancer and lymphoma [19] to 21.24 (145/6827) on NMSC [59], with no higher incident rate and cancer risk than the control group. With respect to IFX use in RA patients, overall malignancy rates were 5.9 (10/1700) [6], 10.20 (144/14120) [8], 5.41 (6/1108) [22] respectively. For non-Hodgkin lymphoma (NHL), rates were reported with 1.8 (2/1108.12) [22] and 1.22 (6/4913) [24]. Mercer et al. [20] reported a higher risk of BCC in RA patients treated with IFX (5.04 (59/11700) per 1000 person-years) compared to control groups (aHR = 1.73, 95%CI 1.14–2.62). Only one study reported malignancy rates for CZP users. Keystone et al. [15] found that after excluding NMSC, the rate of malignancy was 6.2 per 1000 person-years with a total of 23 cases and 3732 person-years of follow-up. The rate for all neoplasms including benign, malignant and unspecified was 28.1 (105/3732) per 1000 person-years (Table 1). This study did not include a control group for comparison, however the authors noted that rates of malignancy did not increase with increased CZP exposure, and that rates of malignancy were comparable with previously published incidence rates in the general population. There are currently no data for GLM use and malignancy risk.

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

3.1.2. Incidence and risk of recurrent malignancy in RA patients Three studies reported recurrent cancer incident rate in RA patients with prior malignancy history. Dixon et al. [13] included 177 TNFi users with a prior history of malignancy; 147 had prior solid malignancy, 13 had prior lymphoproliferative malignancy, and 17 had melanoma malignancy; patients with prior NMSC were excluded from the study. In the total 515 person-year follow up time (3.1 person years per person), 13 malignancies were observed yielding a recurrent malignant incidence of 25.2 per 1000 personyears in TNFi users, which was not higher than the control group of RA patients with prior malignancy not treated with biologics (SIR = 0.58, 95%CI 0.23–1.43). A subset of melanoma recurrence, however, found that 3 out of 17 recurrences in the TNFi treated group compared to 0 out of 10 in the non-biologic treated group. Strangfeld et al. [23] studied the second malignancy incidence using data from the German biologics register (RABBIT). A total of 58 TNFi users with a prior malignancy were included. Nine recurrences were observed in a total of follow-up time 198 person-years with the average time span 9.5 years from the first tumor to second cancer, resulting in a recurrent cancer incidence of 45.5 per 1000 person-years—which was not statistically different from patients with prior malignancies treated with DMARDs alone (IRR = 1.4, 95% CI 0.5–5.5). Raaschou et al. [64] reported the second SCC in 10 and BCC in 17 RA patients were detected who had a prior SCC or BCC history before the start of follow-up, yielding second cancer incidences 25.64 (10/390) for SCC and 63.20 (17/63.20) for BCC, without higher recurrent risk of SCC and BCC in TNFi users than in the control group (Table 1). 3.2. Malignancy incidence and risk in IBD patients in original studies A total of 8 studies [26–32,38] assessed cancer incidence rates in IBD patients including CD and UC patients, with highest reported rate of 34.0 per 1000 person-years of overall cancer among IFX treated CD patients [32] and lowest reported rate of 0 per 1000 person-years of lymphoma among TNFi treated IBD patients [38] and among IFX treated UC patients [29]. Overall cancer incidence was higher in males 4.9 (38/7776) than in females 4.0 (43/10665) [30]. The highest incidence of lymphoma was 3.84 per 1000 person-years in patient with IBD treated with both thiopurine and IFX [29]. No increased incidences and cancer risks of overall cancer, lymphoma, NMSC, or melanoma were reported in IBD patients receiving TNFi treatment compared to the control group treated receiving DMARDs (Table 2). 3.3. Malignancy incidence in JIA patients in original studies Three studies [33,34,36] reported cancer incidences in JIA patients. In evaluating overall cancers in TNFi users, one study reported one case during 386.73 person-years [34] and another study reported one case during 2922 person-years [33], yielding respective cancer incidence 2.6 per 1000 person-years and 0.34 per 1000 person-years. No cases on malignant neoplasms or lymphoma were detected in JIA patients who received ETN, IFX or ADA therapy [34,36] (Table 2).

5

users on NMSC, and lowest rate was 0.9 (8/8922) in ETN users on lymphoma [35]. However, cancer risks in TNFi users were not higher than in the control group. Regarding incident rates in patients with PsA who received TNFi treatment, two studies were found [38,39]. Incidences of lymphoma in TNFi users were 0 per 1000 person-years and 0.73 per 1000 person-years, of any solid tumor was 7.7 per 1000 personyears, and of NMSC was 2.9 [38,39]. There were no differences of solid cancer (HR = 0.74, 95%CI 0.20–2.76) and NMSC (HR = 0.74, 95%CI 0.06–8.72) between TNFi group and control group in PsA patients [38] (Table 2). 3.5. Malignant incidence and risk in patients with other indications in original studies For TNFi treated AS patients, the reported incidence rates of lymphoma, solid tumor and NMSC were 0.26, 4.4 and 1.5 per 1000 person-years respectively [38,39]. Risk of solid cancer in TNFi users was slightly lower than control group (HR = 0.03, 95%CI 0.002–0.45) (Table 2). For SpA patients, Westhovens et al. reported overall cancer incidence of 5.0 per 1000 person-years in TNFi users, with higher rate in women than in men (7.7 vs 3.7) [42] (Table 2). For cancer incidences in other immune-mediated or rheumatic diseases regardless of indications, five studies were identified [37,40,41,43,44]. In TNFi users, the highest reported incidences of overall cancer, melanoma, NMSC, lymphoma and NHL were 11.5, 0.38, 2.7, 0.66, and 0.54 per 1000 person-years [37,43]. Risks of melanoma, NMSC, and NHL in TNFi users were not higher than non-biological users (aHR = 1.54, 95%CI 0.37–6.34; aHR = 1.10, 95%CI 0.69–1.76; aHR = 0.63, 95%CI 0.18–2.20) [37]. In ETN users, the highest cancer incidences of overall cancer and lymphoma were 5.1 and 0.15 [43,44]. In studies of IFX users, incidences of overall cancer was reported as 4.0 in one study and 3.1 in another; incidence of lymphoma in IFX users was 0.69 [40,43,44]. In two studies of ADA users, incidences of overall cancer and lymphoma were 3.0 and 0.65 respectively [40,43]. In one study of GLM users, incidence of overall cancer was 0 per 1000 person-years [44] (Table 2). 3.6. Cancer incidence of TNFi users versus general population in pooled analyses In RA patients receiving TNFi treatment, higher incidences of lymphoma and NMSC were reported as compared to general population, however, no differences were found on overall cancer (Table 3). In AS, PsA and multiple indications patients, incidences of overall cancer, lymphoma and NMSC in TNFi users were no higher than those in the general population (Table 3). In PsO and IBD patients, incidences of NMSC in TNFi users were higher than the general population, however, no differences were found regarding incidences of overall cancer and lymphoma compared to the general population (Table 3). 4. Discussion

3.4. Cancer incidence and risk in PsO and PsA patients in original studies Two studies were found evaluating cancer incidence rates in TNFi treated PsO patients [35,38]; neither study found increased rates of malignancy compared to control groups (Table 2). Haynes et al. reported incidence rates of 7.8/1000 person-years for solid tumors, 0/1000 person-years for lymphomas, and 5.2/1000 person-years for NMSC [38]. Kimball et al. evaluated cancer risk by TNFi subtype and found highest rate was 26.0 (18/691) in IFX

TNFi are frequently used in patients with RA and IBD as well as other chronic immune-mediated inflammatory diseases (e.g., AS, PsA, PsO, JIA) [66]. Incidences of malignancies varied among studies, indications, cancer types, cancer history, and the use of different individual TNF inhibitors. When compared to general population, however, incidences of lymphoma and NMSC were higher in patients with RA. TNFi users did not have increased risks for overall cancer, lymphoma or melanoma. These suggest that malignancy is associated with disease activity rather than with

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

6

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

Table 2 Cancer incidences and risks in TNFi treated patients in original studies. Studies

Indications

Cancer incidences in IBD patients treated with TNFi Nyboe Andersen [30] IBD Biancone [26] CD Haynes [38] IBD Nyboe Andersen [30] IBD Haynes [38] IBD Herrinton [28] IBD Haynes [38] IBD Nyboe Andersen [30] IBD Cancer incidences in IBD patients treated with IFX Fidder [27] IBD Lichtenstein [31] CD Liu [32] CD Lichtenstein [31] CD Liu [32] CD Fidder [27] IBD Lichtenstein [31] CD Liu [32] CD Khan [29] UC Fidder [27] Lichtenstein [31]

IBD CD

Fidder [27] Lichtenstein [31] Fidder [27]

IBD CD IBD

Cancer incidences in JIA patients treated with TNFi Beukelman [33] JIA Kok [34] JIA Tarkiainen [36]

JIA

Cancer incidences in PsO patients treated with TNFi Haynes [38] PsO

Kimball [35]

PsO

Cancer incidences in PsA patients treated with TNFi Hellgren [39] PsA Haynes [38] PsA

Cancer incidences in SpA patients treated with TNFi Westhovens [42] SpA

Cancer incidences in AS patients treated with TNFi Haynes [38] AS

Hellgren [39]

AS

Cancer incidences in patients with rheumatic diseases received TNFi therapy Dreyer [37] Arthritis patients (RA, AS, PsA and other arthritides) Hernandez [40] Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Dreyer [37] Arthritis patients (RA, AS, PsA and other arthritides) Hernandez [40] Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases)

Cancer types: malignancies (NO.)/follow-up durations (patient-years), crude incidences (per 1000 patient-years)

Indicators and 95% confidence interval (TNFi vs control)

Overall cancer: 81/18440, 4.4 Overall cancer: TNFi only 2.7 Overall cancer: TNFi with IMM 3.4 Solid cancer: 8.2 Malignant melanoma: 9/18440, 0.5 Lymphoma: 0 Lymphoma: 1/316, 3.16 NMSC: 2.5 Overall cancer: female 43/10665, 4.0 Overall cancer: male 38/7776, 4.9

SIR, 1.07 (0.85–1.36) NA NA HR, 1.42 (0.47–4.26) SIR, 1.31 (0.63–2.74) NA NA HR, 0.08 (0.01–0.82) SIR, 0.96 (0.69–1.33) SIR, 1.20 (0.85–1.69)

Overall cancer: 23/3775, 6.1 OR, 0.97 (0.56–1.65) Overall cancer: 113/17712, 6.4 RR, 0.88 (0.66–1.19) Overall cancer: 34.0 NA Solid tumor: 75/17712, 4.2 RR, 0.92 (0.66–1.28) Solid tumor: 29.1 NA Lymphoma: 2/3775, 0.5 NA Lymphoma: 8/17712, 0.5 RR, 0.98 (0.34–2.82) Lymphoma: 3.3 NA Lymphoma: IFX 0/482, 0 NA Lymphoma: thiopurine and IFX 1/260, 3.84 IRR, 6.2 (0.8–44.2) NMSC: 8/3775, 2.1; BCC: 6/3775, 1.6; SCC: 2/3775, 0.5 NA NMSC: 29/17712, 1.6; BCC: 16/17712, 0.9; RR, NMSC: 0.89 (0.45–1.74) SCC: 9/17712, 0.5 Melanoma: 2/3775, 0.5 NA Melanoma 7/17712, 0.4 NA NHL: 1/3775, 0.26 NA TNFi, overall cancer: 1/2922, 0.34 TNFi, overall cancer: 1/386.73, 2.6 TNFi, lymphoma: 0/386.73, 0 ETN, malignant neoplasms: 0/710, 0 IFX, malignant neoplasms: 0/591, 0 ADA, malignant neoplasms: 0/188, 0

NA NA NA NA NA NA

TNFi, solid cancer: 7.8 TNFi, lymphoma: 0 TNFi, NMSC: 5.2 ETN, overall cancer except NMSC: 93/8643, 10.8 ETN, lymphoma: 8/8922, 0.9 ETN, NMSC: 146/8646, 16.9 ADA, overall cancer except NMSC: 14/1566, 8.9 ADA, lymphoma: 4/1589, 2.5 ADA, NMSC: 38/1535, 24.8 IFX, overall cancer except NMSC: 9/694, 13.0 IFX, lymphoma: 1/724, 1.4 IFX, NMSC: 18/691, 26.0

HR, 0.58 (0.10–3.31) NA NA IRR, 0.7 (0.5–1.1) IRR, 0.4 (0.1–m1.4) IRR, 1.1 (0.8–1.6) IRR, 0.5 (0.2–1.0) IRR, 0.8 (0.2–2.6) IRR, 1.2 (0.7–1.9) IRR, 1.3 (0.5–3.2) IRR, 3.6 (0.5–29.0) IRR, 1.1 (0.4–2.5)

TNFi, TNFi, TNFi, TNFi,

NA NA HR, 0.74 (0.20–2.76) HR, 0.74 (0.06–8.72)

lymphoma: 8/10912, 0.73 lymphoma: 0 solid cancer: 7.7 NMSC: 2.9

TNFi, overall cancer: total 6/1199.83, 5.0 TNFi, overall cancer: female 7.7 TNFi, overall cancer: male 3.7

NA NA NA

TNFi, TNFi, TNFi, TNFi,

NA HR, 0.03 (0.002–0.45) NA NA

lymphoma: 0 solid cancer: 4.4 NMSC: 1.5 lymphoma: 2/7790, 0.26

TNFi, overall cancer: 180/15592, 11.5 TNFi, overall cancer: 54/17330, 3.1

NA NA

TNFi, melanoma: 6/15592, 0.38 TNFi, melanoma: 3/17330, 0.17

aHR, 1.54 (0.37–6.34) NA

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

7

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx Table 2 (continued) Studies

Indications

Cancer types: malignancies (NO.)/follow-up durations (patient-years), crude incidences (per 1000 patient-years)

Indicators and 95% confidence interval (TNFi vs control)

Dreyer [37] Hernandez [40]

Arthritis patients (RA, AS, PsA and other arthritides) Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Arthritis patients (RA, AS, PsA and other arthritides) Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Any indication (i.e., RA, SpA, AS, PsA, IBD, PsO) Arthritis patients (RA, AS, PsA and other arthritides) Any indication (i.e., RA, SpA, AS, PsA, IBD, PsO) Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Rheumatic diseases (RA, SpA, PsA and others) Any indication (i.e., RA, SpA, AS, PsA, IBD, PsO) Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Rheumatic diseases (RA, SpA, PsA and others) Any indication (i.e., RA, SpA, AS, PsA, IBD, PsO) Rheumatic diseases (RA, AS, PsA, chronic arthritis and chronic immune-mediated diseases) Chronic inflammatory conditions (age 6 30 years old: JIA, RA, AS, PsA, PsO and other) Rheumatic diseases (RA, SpA, PsA and others) Chronic inflammatory conditions (age 6 30 years old: JIA, RA, AS, PsA, PsO and other) Any indication (i.e., RA, SpA, AS, PsA, IBD, PsO) Chronic inflammatory conditions (age 6 30 years old: JIA, RA, AS, PsA, PsO and other) Rheumatic diseases (RA, SpA, PsA and others)

TNFi, NMSC: 42/15592, 2.7 TNFi, NMSC: 34/17330, 1.96

aHR, 1.10 (0.69–1.76) NA

TNFi, lymphoma: 7/15592, 0.45 TNFi, lymphoma: 4/17330, 0.23

NA NA

TNFi, lymphoma: 38/57711, 0.66 TNFi, NHL: 5/15592, 0.32 TNFi, NHL: 31/57711, 0.54 IFX, overall cancer: 4.0

NA aHR, 0.63 (0.18–2.20) NA NA

IFX, overall cancer: 4/1308, 3.1 IFX, lymphoma: 0.69 ADA, overall cancer: 3.0

NA NA NA

ADA, overall cancer: 0/473, 0 ADA, lymphoma: 0.65 ETN, overall cancer: 3.0

NA NA NA

ETN, overall cancer: 103/231404, 0.45

NA

ETN, overall cancer: 6/1176, 5.1 ETN, lymphoma: 23/231404, 0.10

NA NA

ETN, lymphoma: 0.15 ETN, melanoma: 7/231404, 0.03 ETN, NHL: 7/231404, 0.03 GLM, overall cancer: 0/163, 0

NA NA NA NA

Dreyer [37] Hernandez [40] Mariette [43] Dreyer [37] Mariette [43] Hernandez [40] Mok [44] Mariette [43] Hernandez [40] Mok [44] Mariette [43] Hernandez [40] Hooper [41] Mok [44] Hooper [41] Mariette [43] Hooper [41] Mok [44]

NO. = number; NA = not acquired; IBD = inflammatory bowel disease; CD = Crohn’s disease; UC = ulcerative colitis; JIA = juvenile idiopathic arthritis; AS = ankylosing spondylitis; PsO = psoriasis; PsA = psoriatic arthritis; SpA = spondyloarthropathies; TNFi = tumor necrosis factor inhibitors; IFX = infliximab; ETN = etanercept; ADA = adalimumab; GLM = golimumab; IMM = immunosuppressant; IRR = incidence rate ratio; SIR = standardized rate ratio; RR = relative risk; OR = odds ratio; HR = hazard ratio; aHR = adjusted hazard ratio; NHL = non-Hodgkin lymphoma; NMSC = non-melanoma skin cancer; SCC = squamous cell carcinoma; BCC = basal cell carcinoma.

treatment. Chronic inflammation is an established risk factor for cancer, and is estimated to be associated with 25% of cancers [67,68]. It is further reported that the elevated lymphoma risk in RA patients is related to high inflammatory activity itself rather than the treatment of inflammation [69]. RA is also reported to be a risk factor for the development of skin cancer [70]. Consequently, it is not surprising that RA and other immunemedicated chronic inflammatory disorders are associated with increased incidences of lymphoma and NMSC. However, a small risk in squamous cell cancer associated with TNFi was observed in a national cohort of veterans with RA in the US [59] and a prospective cohort study in Sweden [64]. A strong association between a prior history of malignancy was also found in veterans with RA [59]. Although the risk is small, it is advised that practicing physicians be vigilant in screening patients for risk factors for NMSC prior to considering a TNFi for an RA patient. There is biologic plausibility for increased risk of cancer due to TNFi treatment. TNF is a major cytokine that regulates the inflammatory response and may influence tumor occurrence. Tumor necrosis activity was first observed in the serum of mice treated with bacillus Calmette-Guerin (BCG) and bacterial endotoxin which induced hemorrhagic necrosis of subcutaneously transplanted sarcomas [71]. Interestingly, methylcholanthrene (Meth A)-induced sarcoma cells do not express TNF receptors and therefore are not responsive to TNF in vitro. The in vivo killing of Meth A sarcoma by TNF is actually mediated by neutrophil-mediated destruction and thrombosis of tumor blood vessels (reviewed in [72]). It is now clear that TNF has complex and different functions in cancer biology. TNF signaling can activate pathways resulting in cell survival and proliferation, transcription of proinflammatory cytokines and cell death [73]. Under different circumstances, TNF can either promote death or survival of tumor cells. Therefore, investigations have been directed to promote TNF or to target

TNF for therapy of cancer [72]. Nonetheless, from this clinical epidemiology review in patients with inflammatory diseases who have been treated with TNFi, there is no significant increase in risk of cancer compared to those treated with conventional DMARDs. In our review, TNFi themselves do not appear to be associated with higher rates of overall malignancies, lymphoma and melanoma compared with patients with the same underlying disease not taking biologics. Most of the control patients not on biologics, however, are taking traditional DMARDs, which may also be associated with increased cancer risk. Patients exposed to thiopurine had higher lymphoma incidence rates compared to unexposed UC patients (IRR = 3.6, 95%CI 2.2–6.0) [29], and higher NMSC risk in IBD patients (aHR = 2.28, 95%CI 1.50–3.45) [74]. It has also been reported that methotrexate might increase the risk of malignant melanoma in RA patients and increase the risk of NMSC in psoriasis patients [70]. The use of methotrexate for one year or more has also been reported to increase the risk of developing a second NMSC in IBD patients with a prior NMSC surgery history (HR = 1.24, 95%CI 1.04–1.48) [75]. Cyclosporine has also been found to distinctly elevate NMSC risk, though is uncommonly used in the inflammatory conditions discussed in this review [70]. Our review also found an increased cancer risk in patients exposed to immune-suppressants compared with non-exposed patients. In CD patients there were higher incidences of overall cancer in a TNFi plus immunomodulator group (3.4 per 1000 person-years) than in TNFi monotherapy group (2.7) [26]. UC patients on IFX plus thiopurine had a higher lymphoma incidence rate than IFX monotherapy patients (3.84 versus 0 per 1000 person-years) [29]. IBD patients on ADA plus immunomodulator therapy were finally found to have higher incidence of overall cancer than an ADA monotherapy group (3.04 versus 2.78 per 1000 patient-years) [50].

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

8

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

Table 3 Standardized incidence ratios in TNFi users versus general population in pooled analysis. Studies

TNFi

Cancer types

SIRs in RA patients treated with TNFi Burmester [56] ADA Overall cancer excluding NMSC Lymphoma Schiff [47] ADA Lymphoma Burmester [56] ADA NMSC Bykerk [46] CZP Overall cancer excluding NMSC (vs GLOBOCAN database) Overall cancer excluding NMSC (vs SEER database) Lymphoma (vs GLOBOCAN database) Lymphoma (vs SEER database) Gottlieb [57] ETN Lymphoma Kay [58] GLM Overall cancer Lymphoma Mercer [63] TNFi Melanoma SIRs in AS patients treated with TNFi Burmester [56] ADA Overall cancer excluding NMSC lymphoma NMSC van der Heijde ETN Overall cancer excluding NMSC [55] BCC/SCC (vs Arizona) SCC (vs Minnesota) Kay [58] GLM Overall cancer Lymphoma Hellgren [62] TNFi Overall cancer Melanoma

TNFi

IFX

Overall cancer Lymphoma Melanoma

SCC Overall cancer excluding NMSC lymphoma

SIRs in IBD (regardless of CD and UC) patients treated with TNFi Burmester [56] ADA Overall cancer excluding NMSC Colombel [48] ADA Overall cancer Osterman [50] ADA Overall cancer excluding NMSC Overall cancer excluding NMSC Overall cancer excluding NMSC Burmester [56] ADA Lymphoma Colombel [48] ADA Lymphoma Burmester [56] ADA NMSC Colombel [48] ADA NMSC Osterman [50]

ADA

Lichtenstein [49]

IFX

Siegel [51]

TNFi

NMSC NMSC NMSC Overall cancer excluding Overall cancer excluding Overall cancer excluding Overall cancer excluding Overall cancer excluding Overall cancer excluding Lymphoma (IBD) Lymphoma (IBD) Lymphoma (CD) Lymphoma (CD) Lymphoma (UC) Lymphoma (UC) NHL

NMSC NMSC NMSC NMSC NMSC NMSC

0.93 2.74 3.19 1.39 1.27

Age and sex

(0.82–1.06) (1.83–3.93) (1.78–5.26) (1.19–1.60) (0.99–1.61)

1.06 (0.82–1.33) 2.72 (0.88–6.34) 1.81 (0.59–4.23) 3.45 (1.83–5.89) GLM 50 mg: 1.64 (0.92–2.71); GLM 100 mg: 1.15 (0.68–1.81) GLM 50 mg: 0.00 (0.00–7.81); GLM 100 mg: 8.89 (3.30–19.55) 1.2 (0.99–1.6)

0.68 (0.22–1.59) 5.88 (0.66–21.2) 1.25 (0.46–2.72) GLM 50 mg: 1.30 (0.27–3.81); GLM 100 mg: 0.64 (0.08–2.32) GLM 50 mg: 0.00 (0.00–25.76); GLM 100 mg: 0.00 (0.00– 19.68) 0.9 (0.7–1.1) 1.1 (0.4–2.60) 1.5 (0.7–3.3)

SIRs in PsO patients treated with TNFi Burmester [56] ADA Overall cancer excluding NMSC Lymphoma NMSC Pariser [54] ETN Overall cancer excluding NMSC SCC/BCC

Menter [53]

Standardized variables

0.51 (0.16–1.19) 1.93 (0.03–10.7) 0.80 (0.29–1.74) 1.47 (0.54–3.21) BCC: 0.12 (0.00–0.65); SCC: 0.56 (0.01–3.11) 1.37 (0.03–7.65) GLM 50 mg: 0.07 (0.02–3.89); GLM 100 mg: 0.00 (0.00–2.19) GLM 50 mg: 11.97 (0.30–66.71); GLM 100 mg: 0.00 (0.00–38.96) 0.9 (0.7–1.2) 1.4 (0.6–3.6)

SIRs in PsA patients treated with TNFi Burmester [56] ADA Overall cancer excluding NMSC Lymphoma NMSC Kay [58] GLM Overall cancer Lymphoma Hellgren [62]

SIRs and 95% confidence interval (TNFi users vs general population)

(IBD) (IBD) (CD) (CD) (UC) (UC)

Age and sex Age and sex Age and sex

Age and sex NA Age and sex Age and sex

NA

NA Age and sex

Age and sex

NA

Age and sex

0.96 (0.65–1.36) 0.63 (0.01–3.49) 1.76 (1.26–2.39) 1.15 (0.78–1.64) SCC (vs Arizona): 1.78 (1.11–2.69); BCC (vs Arizona): 0.55 (0.37–0.80) SCC (vs Minnesota): 4.28 (2.68–6.47) 0.39 (0.05–1.42) 0.00 (0.00–12.76)

Age and sex

1.52 (0.95–2.31) 1.56 (0.94–2.43) ADA monotherapy: 0.63 (0.17–1.62) ADA + any immunomodulator: 3.04 (1.66–5.10) ADA + thiopurine 2.78 (1.33–5.11) 2.58 (0.29–9.31) 2.94 (0.33–10.6) 2.29 (1.44–3.47) NMSC: 2.09 (1.22–3.35); SCC: 5.26 (1.92–11.5); BCC: 1.43 (0.69–2.63) ADA monotherapy: 1.2 (0.39–2.80) ADA + any immunomodulator: 4.59 (2.51–7.70) ADA + thiopurine: 5.05 (2.61–8.82) IFX without immunomodulator: 0.78 (0.02–4.33) IFX with immunomodulator: 0.99 (0.02–5.50) IFX without immunomodulator: 0 (0.00–6.92) IFX with immunomodulator: 0 (0.00–8.24) IFX without immunomodulator: 1.22 (0.03–6.82) IFX with immunomodulator: 1.54(0.04–8.57) IFX without immunomodulator: 0 (0.00–54.61) IFX with immunomodulator: 0 (0.00–62.71) IFX without immunomodulator: 0 (0.00–157.21) IFX with immunomodulator: 0 (0.00–163.60) IFX without immunomodulator: 0 (0.00–83.69) IFX with immunomodulator: 0 (0.00–101.69) 3.23 (1.5–6.9)

Age and sex Age Age and sex

NA

Age, sex and race

Age and sex Age Age and sex Age Age and sex

Age, sex, and race

Age and sex

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

9

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx Table 3 (continued) Studies

TNFi

Cancer types

SIRs in patients with multiple indications received TNFi therapy Gottlieb [57] ETN Overall cancer (RA, JIA, PsA, AS, PsA) Kay [58] GLM Overall cancer (RA, PsA, AS) Lymphoma (RA, PsA, AS) Hellgren [62] TNFi Overall cancer (AS, SpA, PsA) Lymphoma (AS, SpA, PsA) Melanoma (AS, SpA, PsA)

SIRs and 95% confidence interval (TNFi users vs general population)

Standardized variables

1.00 (0.83–1.19) GLM 50 mg: 1.48 (0.89–2.31); GLM 100 mg: 0.99 (0.61–1.53) GLM 50 mg: 1.71 (0.04–9.55); GLM 100 mg: 6.69 (2.45–14.56) 0.9 (0.7–1.0) 0.8 (0.4–1.7) 1.3 (0.7–2.3)

Age and sex NA Age and sex

NA = not acquired; RA = rheumatoid arthritis; IBD = inflammatory bowel disease; CD = Crohn’s disease; UC = ulcerative colitis; JIA = juvenile idiopathic arthritis; AS = ankylosing spondylitis; PsO = psoriasis; PsA = psoriatic arthritis; SpA = spondyloarthropathies; TNFi = tumor necrosis factor inhibitors; ADA = adalimumab; CZP = certolizumab pegol; ETN = etanercept; IFX = infliximab; GLM = golimumab; SIR = standardized rate ratio; NHL = non-Hodgkin lymphoma; NMSC = non-melanoma skin cancer; SCC = squamous cell carcinoma; BCC = basal cell carcinoma.

Of all the eligible studies, five studies [7,9,22,30,42] reported gender subgroup incidences of overall cancer, solid cancer, lymphoma and NHL in RA, IBD and SpA patients. Higher cancer incidences were typically found in male patients except for overall cancer incidence in SpA patients, where female patients had a higher cancer incidence (7.7 per 1000 person-years in females versus 3.7 per 1000 person-years in males). The explanation for higher malignant incidence in females in this study was mainly due to more frequently breast cancer cases in women, which comprised roughly 54% of all malignancies [42]. To summarize, cancer incidences in TNFi users, most frequently reported in RA and IBD patients, are varied among different studies, indications, cancer types and individual TNF inhibitors. Cancer incidence rates of lymphoma are higher than the general population in RA patients, and incidence rates of NMSC are higher than the general population in PsO, RA and IBD patients, but not in AS and PsA patients. The exact cause of the increased malignancy rates in these patients compared to the general population is not precisely known, but is may be due to the underlying disease itself or confounding immunomodulatory medications. Incidences and risks of overall cancer, lymphoma and melanoma in TNFi users do not appear to be increased compared to patients with a similar disease not treated with biologics, whereas, higher NMSC risk in TNFi in RA patients remains to be confirmed by more prospective studies. Acknowledgements CQC is supported by Rheumatology Research Foundation Innovative and Pilot grants. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cyto.2016.09.013. References [1] H. Zelova, J. Hosek, TNF-alpha signalling and inflammation: interactions between old acquaintances, Inflammat. Res.: Official J. Eur. Histamine Res. Soc. [et al.] 62 (2013) 641–651. [2] C. Blandizzi, P. Gionchetti, A. Armuzzi, R. Caporali, S. Chimenti, R. Cimaz, et al., The role of tumour necrosis factor in the pathogenesis of immune-mediated diseases, Int. J. Immunopathol. Pharmacol. 27 (2014) 1–10. [3] Y. Lai, C. Dong, Therapeutic antibodies that target inflammatory cytokines in autoimmune diseases, Int. Immunol. 28 (2016) 181–188. [4] P.C. Taylor, Pharmacology of TNF blockade in rheumatoid arthritis and other chronic inflammatory diseases, Curr. Opin. Pharmacol. 10 (2010) 308–315. [5] M.A. Willrich, D.L. Murray, M.R. Snyder, Tumor necrosis factor inhibitors: clinical utility in autoimmune diseases, Transl. Res.: J. Lab. Clin. Med. 165 (2015) 270–282.

[6] K.J. Aaltonen, J.T. Joensuu, L. Virkki, T. Sokka, P. Aronen, H. Relas, et al., Rates of serious infections and malignancies among patients with rheumatoid arthritis receiving either tumor necrosis factor inhibitor or rituximab therapy, J. Rheumatol. 42 (2015) 372–378. [7] J. Askling, E. Baecklund, F. Granath, P. Geborek, M. Fored, C. Backlin, et al., Antitumour necrosis factor therapy in rheumatoid arthritis and risk of malignant lymphomas: relative risks and time trends in the Swedish Biologics Register, Ann. Rheum. Dis. 68 (2009) 648–653. [8] J. Askling, R.F. van Vollenhoven, F. Granath, P. Raaschou, C.M. Fored, E. Baecklund, et al., Cancer risk in patients with rheumatoid arthritis treated with anti-tumor necrosis factor alpha therapies: does the risk change with the time since start of treatment?, Arthritis Rheum 60 (2009) 3180–3189. [9] N. Berghen, L.A. Teuwen, R. Westhovens, P. Verschueren, Malignancies and anti-TNF therapy in rheumatoid arthritis: a single-center observational cohort study, Clin. Rheumatol. 34 (2015) 1687–1695. [10] R. Buchbinder, S. Van Doornum, M. Staples, M. Lassere, L. March, Malignancy risk in Australian rheumatoid arthritis patients treated with anti-tumour necrosis factor therapy: analysis of the Australian Rheumatology Association Database (ARAD) prospective cohort study, BMC Musculoskeletal Disord. 16 (2015) 309. [11] G.R. Burmester, M. Matucci-Cerinic, X. Mariette, F. Navarro-Blasco, S. Kary, K. Unnebrink, et al., Safety and effectiveness of adalimumab in patients with rheumatoid arthritis over 5 years of therapy in a phase 3b and subsequent postmarketing observational study, Arthrit. Res. Therapy 16 (2014) R24. [12] Y.M. Chiu, H.C. Lang, H.Y. Lin, M.T. Yang, C.H. Fang, Y.W. Yang, et al., Risk of tuberculosis, serious infection and lymphoma with disease-modifying biologic drugs in rheumatoid arthritis patients in Taiwan, Int. J. Rheum. Dis. 17 (Suppl 3) (2014) 9–19. [13] W.G. Dixon, K.D. Watson, M. Lunt, L.K. Mercer, K.L. Hyrich, D.P. Symmons, Influence of anti-tumor necrosis factor therapy on cancer incidence in patients with rheumatoid arthritis who have had a prior malignancy: results from the British Society for Rheumatology Biologics Register, Arthrit. Care Res. 62 (2010) 755–763. [14] P. Geborek, A. Bladstrom, C. Turesson, A. Gulfe, I.F. Petersson, T. Saxne, et al., Tumour necrosis factor blockers do not increase overall tumour risk in patients with rheumatoid arthritis, but may be associated with an increased risk of lymphomas, Ann. Rheum. Dis. 64 (2005) 699–703. [15] E. Keystone, R. Landewe, R. van Vollenhoven, B. Combe, V. Strand, P. Mease, et al., Long-term safety and efficacy of certolizumab pegol in combination with methotrexate in the treatment of rheumatoid arthritis: 5-year results from the RAPID 1 trial and open-label extension, Ann. Rheum. Dis. 73 (2014) 2094– 2100. [16] W. Kievit, J. Fransen, E.M. Adang, A.A. den Broeder, H.J. Bernelot Moens, H. Visser, et al., Long-term effectiveness and safety of TNF-blocking agents in daily clinical practice: results from the Dutch Rheumatoid Arthritis Monitoring register, Rheumatology (Oxford) 50 (2011) 196–203. [17] Y. Komano, M. Tanaka, T. Nanki, R. Koike, R. Sakai, H. Kameda, et al., Incidence and risk factors for serious infection in patients with rheumatoid arthritis treated with tumor necrosis factor inhibitors: a report from the Registry of Japanese Rheumatoid Arthritis Patients for Longterm Safety, J. Rheumatol. 38 (2011) 1258–1264. [18] M. Lebwohl, R. Blum, E. Berkowitz, D. Kim, R. Zitnik, C. Osteen, et al., No evidence for increased risk of cutaneous squamous cell carcinoma in patients with rheumatoid arthritis receiving etanercept for up to 5 years, Arch. Dermatol. 141 (2005) 861–864. [19] D.J. Lovell, A. Reiff, N.T. Ilowite, C.A. Wallace, Y. Chon, S.L. Lin, et al., Safety and efficacy of up to eight years of continuous etanercept therapy in patients with juvenile rheumatoid arthritis, Arthritis Rheum. 58 (2008) 1496–1504. [20] L.K. Mercer, A.C. Green, J.B. Galloway, R. Davies, M. Lunt, W.G. Dixon, et al., The influence of anti-TNF therapy upon incidence of keratinocyte skin cancer in patients with rheumatoid arthritis: longitudinal results from the British Society for Rheumatology Biologics Register, Ann. Rheum. Dis. 71 (2012) 869–874.

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

10

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx

[21] L.K. Mercer, M. Lunt, A.L. Low, W.G. Dixon, K.D. Watson, D.P. Symmons, et al., Risk of solid cancer in patients exposed to anti-tumour necrosis factor therapy: results from the British Society for Rheumatology Biologics Register for Rheumatoid Arthritis, Ann. Rheum. Dis. 74 (2015) 1087–1093. [22] F.B. Pallavicini, R. Caporali, P. Sarzi-Puttini, F. Atzeni, C. Bazzani, R. Gorla, et al., Tumour necrosis factor antagonist therapy and cancer development: analysis of the LORHEN registry, Autoimmun. Rev. 9 (2010) 175–180. [23] A. Strangfeld, F. Hierse, R. Rau, G.R. Burmester, B. Krummel-Lorenz, W. Demary, et al., Risk of incident or recurrent malignancies among patients with rheumatoid arthritis exposed to biologic therapy in the German biologics register RABBIT, Arthrit. Res. Therapy 12 (2010) R5. [24] F. Wolfe, K. Michaud, Lymphoma in rheumatoid arthritis: the effect of methotrexate and anti-tumor necrosis factor therapy in 18,572 patients, Arthritis Rheum. 50 (2004) 1740–1751. [25] C.Y. Wu, D.Y. Chen, J.L. Shen, H.J. Ho, C.C. Chen, K.N. Kuo, et al., The risk of cancer in patients with rheumatoid arthritis taking tumor necrosis factor antagonists: a nationwide cohort study, Arthrit. Res. Therapy 16 (2014) 449. [26] L. Biancone, S. Zuzzi, M. Ranieri, C. Petruzziello, E. Calabrese, S. Onali, et al., Fistulizing pattern in Crohn’s disease and pancolitis in ulcerative colitis are independent risk factors for cancer: a single-center cohort study, J. Crohn’s Colitis 6 (2012) 578–587. [27] H. Fidder, F. Schnitzler, M. Ferrante, M. Noman, K. Katsanos, S. Segaert, et al., Long-term safety of infliximab for the treatment of inflammatory bowel disease: a single-centre cohort study, Gut 58 (2009) 501–508. [28] L.J. Herrinton, L. Liu, X. Weng, J.D. Lewis, S. Hutfless, J.E. Allison, Role of thiopurine and anti-TNF therapy in lymphoma in inflammatory bowel disease, Am. J. Gastroenterol. 106 (2011) 2146–2153. [29] N. Khan, A.M. Abbas, G.R. Lichtenstein, E.V. Loftus Jr., L.A. Bazzano, Risk of lymphoma in patients with ulcerative colitis treated with thiopurines: a nationwide retrospective cohort study, Gastroenterology 145 (1007–15) (2013) e3. [30] N. Nyboe Andersen, B. Pasternak, S. Basit, M. Andersson, H. Svanstrom, S. Caspersen, et al., Association between tumor necrosis factor-alpha antagonists and risk of cancer in patients with inflammatory bowel disease, JAMA 311 (2014) 2406–2413. [31] G.R. Lichtenstein, B.G. Feagan, R.D. Cohen, B.A. Salzberg, R.H. Diamond, W. Langholff, et al., Drug therapies and the risk of malignancy in Crohn’s disease: results from the TREAT Registry, Am. J. Gastroenterol. 109 (2014) 212–223. [32] J. Liu, G. Sylwestrzak, A.P. Ruggieri, A. DeVries, Intravenous versus subcutaneous anti-TNF-alpha agents for Crohn’s disease: a comparison of effectiveness and safety, J. Managed Care Specialty Pharm. 21 (2015) 559–566. [33] T. Beukelman, K. Haynes, J.R. Curtis, F. Xie, L. Chen, C.J. Bemrich-Stolz, et al., Rates of malignancy associated with juvenile idiopathic arthritis and its treatment, Arthritis Rheum. 64 (2012) 1263–1271. [34] V.C. Kok, J.T. Horng, J.L. Huang, K.W. Yeh, J.J. Gau, C.W. Chang, et al., Population-based cohort study on the risk of malignancy in East Asian children with juvenile idiopathic arthritis, BMC Cancer 14 (2014) 634. [35] A.B. Kimball, J. Schenfeld, N.A. Accortt, M.S. Anthony, K.J. Rothman, D. Pariser, Cohort study of malignancies and hospitalized infectious events in treated and untreated patients with psoriasis and a general population in the United States, Br. J. Dermatol. 173 (2015) 1183–1190. [36] M. Tarkiainen, P. Tynjala, P. Vahasalo, P. Lahdenne, Occurrence of adverse events in patients with JIA receiving biologic agents: long-term follow-up in a real-life setting, Rheumatology (Oxford) 54 (2015) 1170–1176. [37] L. Dreyer, L. Mellemkjaer, A.R. Andersen, P. Bennett, U.E. Poulsen, T. Juulsgaard Ellingsen, et al., Incidences of overall and site specific cancers in TNFalpha inhibitor treated patients with rheumatoid arthritis and other arthritides – a follow-up study from the DANBIO Registry, Ann. Rheum. Dis. 72 (2013) 79–82. [38] K. Haynes, T. Beukelman, J.R. Curtis, C. Newcomb, L.J. Herrinton, D.J. Graham, et al., Tumor necrosis factor alpha inhibitor therapy and cancer risk in chronic immune-mediated diseases, Arthritis Rheum. 65 (2013) 48–58. [39] K. Hellgren, K.E. Smedby, C. Backlin, C. Sundstrom, N. Feltelius, J.K. Eriksson, et al., Ankylosing spondylitis, psoriatic arthritis, and risk of malignant lymphoma: a cohort study based on nationwide prospectively recorded data from Sweden, Arthrit. Rheumatol. 66 (2014) 1282–1290. [40] M.V. Hernandez, R. Sanmarti, J.D. Canete, M.A. Descalzo, M. Alsina, L. Carmona, et al., Cutaneous adverse events during treatment of chronic inflammatory rheumatic conditions with tumor necrosis factor antagonists: study using the Spanish registry of adverse events of biological therapies in rheumatic diseases, Arthrit. Care Res. 65 (2013) 2024–2031. [41] M. Hooper, D. Wenkert, B. Bitman, V.C. Dias, Y. Bartley, Malignancies in children and young adults on etanercept: summary of cases from clinical trials and post marketing reports, Pediat. Rheumatol. Online J. 11 (2013) 35. [42] I. Westhovens, R.J. Lories, R. Westhovens, P. Verschueren, K. de Vlam, Anti-TNF therapy and malignancy in spondyloarthritis in the Leuven spondyloarthritis biologics cohort (BIOSPAR), Clin. Exp. Rheumatol. 32 (2014) 71–76. [43] X. Mariette, F. Tubach, H. Bagheri, M. Bardet, J.M. Berthelot, P. Gaudin, et al., Lymphoma in patients treated with anti-TNF: results of the 3-year prospective French RATIO registry, Ann. Rheum. Dis. 69 (2010) 400–408. [44] C.C. Mok, K.Y. Chan, K.L. Lee, L.S. Tam, K.W. Lee, Factors associated with withdrawal of the anti-TNFalpha biologics in the treatment of rheumatic diseases: data from the Hong Kong Biologics Registry, Int. J. Rheum. Dis. 17 (Suppl 3) (2014) 1–8. [45] T. Bongartz, F.C. Warren, D. Mines, E.L. Matteson, K.R. Abrams, A.J. Sutton, Etanercept therapy in rheumatoid arthritis and the risk of malignancies: a

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

systematic review and individual patient data meta-analysis of randomised controlled trials, Ann. Rheum. Dis. 68 (2009) 1177–1183. V.P. Bykerk, J. Cush, K. Winthrop, L. Calabrese, O. Lortholary, M. de Longueville, et al., Update on the safety profile of certolizumab pegol in rheumatoid arthritis: an integrated analysis from clinical trials, Ann. Rheum. Dis. 74 (2015) 96–103. M.H. Schiff, G.R. Burmester, J.D. Kent, A.L. Pangan, H. Kupper, S.B. Fitzpatrick, et al., Safety analyses of adalimumab (HUMIRA) in global clinical trials and US postmarketing surveillance of patients with rheumatoid arthritis, Ann. Rheum. Dis. 65 (2006) 889–894. J.F. Colombel, W.J. Sandborn, R. Panaccione, A.M. Robinson, W. Lau, J. Li, et al., Adalimumab safety in global clinical trials of patients with Crohn’s disease, Inflamm. Bowel Dis. 15 (2009) 1308–1319. G.R. Lichtenstein, P. Rutgeerts, W.J. Sandborn, B.E. Sands, R.H. Diamond, M. Blank, et al., A pooled analysis of infections, malignancy, and mortality in infliximab- and immunomodulator-treated adult patients with inflammatory bowel disease, Am. J. Gastroenterol. 107 (2012) 1051–1063. M.T. Osterman, W.J. Sandborn, J.F. Colombel, A.M. Robinson, W. Lau, B. Huang, et al., Increased risk of malignancy with adalimumab combination therapy, compared with monotherapy, for Crohn’s disease, Gastroenterology 146 (2014) 941–949. C.A. Siegel, S.M. Marden, S.M. Persing, R.J. Larson, B.E. Sands, Risk of lymphoma associated with combination anti-tumor necrosis factor and immunomodulator therapy for the treatment of Crohn’s disease: a metaanalysis, Clin. Gastroenterol. Hepatol.: Official Clin. Pract. J. Am. Gastroenterol. Assoc. 7 (2009) 874–881. E.D. Dommasch, K. Abuabara, D.B. Shin, J. Nguyen, A.B. Troxel, J.M. Gelfand, The risk of infection and malignancy with tumor necrosis factor antagonists in adults with psoriatic disease: a systematic review and meta-analysis of randomized controlled trials, J. Am. Acad. Dermatol. 64 (2011) 1035–1050. A. Menter, K. Reich, A.B. Gottlieb, M. Bala, S. Li, M.C. Hsu, et al., Adverse drug events in infliximab-treated patients compared with the general and psoriasis populations, J. Drugs Dermatol.: JDD 7 (2008) 1137–1146. D.M. Pariser, C.L. Leonardi, K. Gordon, A.B. Gottlieb, S. Tyring, K.A. Papp, et al., Integrated safety analysis: short- and long-term safety profiles of etanercept in patients with psoriasis, J. Am. Acad. Dermatol. 67 (2012) 245–256. D. van der Heijde, D. Zack, J. Wajdula, S. Sridharan, A.S. Koenig, Rates of serious infections, opportunistic infections, inflammatory bowel disease, and malignancies in subjects receiving etanercept vs. controls from clinical trials in ankylosing spondylitis: a pooled analysis, Scand. J. Rheumatol. 43 (2014) 49–53. G.R. Burmester, R. Panaccione, K.B. Gordon, M.J. McIlraith, A.P. Lacerda, Adalimumab: long-term safety in 23 458 patients from global clinical trials in rheumatoid arthritis, juvenile idiopathic arthritis, ankylosing spondylitis, psoriatic arthritis, psoriasis and Crohn’s disease, Ann. Rheum. Dis. 72 (2013) 517–524. A.B. Gottlieb, K. Gordon, E.H. Giannini, P. Mease, J. Li, Y. Chon, et al., Clinical trial safety and mortality analyses in patients receiving etanercept across approved indications, J. Drugs Dermatol.: JDD 10 (2011) 289–300. J. Kay, R. Fleischmann, E. Keystone, E.C. Hsia, B. Hsu, M. Mack, et al., Golimumab 3-year safety update: an analysis of pooled data from the longterm extensions of randomised, double-blind, placebo-controlled trials conducted in patients with rheumatoid arthritis, psoriatic arthritis or ankylosing spondylitis, Ann. Rheum. Dis. 74 (2015) 538–546. W. Amari, A.L. Zeringue, J.R. McDonald, L. Caplan, S.A. Eisen, P. Ranganathan, Risk of non-melanoma skin cancer in a national cohort of veterans with rheumatoid arthritis, Rheumatology (Oxford) 50 (2011) 1431–1439. J. Askling, K. Fahrbach, B. Nordstrom, S. Ross, C.H. Schmid, D. Symmons, Cancer risk with tumor necrosis factor alpha (TNF) inhibitors: meta-analysis of randomized controlled trials of adalimumab, etanercept, and infliximab using patient level data, Pharmacoepidemiol. Drug Saf. 20 (2011) 119–130. J.P. Leombruno, T.R. Einarson, E.C. Keystone, The safety of anti-tumour necrosis factor treatments in rheumatoid arthritis: meta and exposureadjusted pooled analyses of serious adverse events, Ann. Rheum. Dis. 68 (2009) 1136–1145. K. Hellgren, L. Dreyer, E.V. Arkema, B. Glintborg, L.T. Jacobsson, L.E. Kristensen, et al., Cancer risk in patients with spondyloarthritis treated with TNF inhibitors: a collaborative study from the ARTIS and DANBIO registers, Ann. Rheum. Dis. (2016). Published Online First 4 May 2016. L.K. Mercer, J. Askling, P. Raaschou, W.G. Dixon, L. Dreyer, M.L. Hetland, et al., Risk of invasive melanoma in patients with rheumatoid arthritis treated with biologics: results from a collaborative project of 11 European biologic registers, Ann. Rheum. Dis. (2016). Published Online First 15 June 2016. P. Raaschou, J.F. Simard, C. Asker Hagelberg, J. Askling, Rheumatoid arthritis, anti-tumour necrosis factor treatment, and risk of squamous cell and basal cell skin cancer: cohort study based on nationwide prospectively recorded data from Sweden, BMJ 352 (2016) i262. P. Raaschou, J.F. Simard, M. Holmqvist, J. Askling, Rheumatoid arthritis, antitumour necrosis factor therapy, and risk of malignant melanoma: nationwide population based prospective cohort study from Sweden, BMJ 346 (2013) f1939. Y. Chen, J. Sun, Y. Yang, Y. Huang, G. Liu, Malignancy risk of anti-tumor necrosis factor alpha blockers: an overview of systematic reviews and metaanalyses, Clin. Rheumatol. 35 (2016) 1–18. F.R. Balkwill, A. Mantovani, Cancer-related inflammation: common themes and therapeutic opportunities, Semin. Cancer Biol. 22 (2012) 33–40.

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013

Y. Chen et al. / Cytokine xxx (2016) xxx–xxx [68] A. Korniluk, O. Koper, H. Kemona, V. Dymicka-Piekarska, From inflammation to cancer, Ir. J. Med. Sci. (2016). Published Online First 7 May 2016. [69] E. Baecklund, A. Iliadou, J. Askling, A. Ekbom, C. Backlin, F. Granath, et al., Association of chronic inflammation, not its treatment, with increased lymphoma risk in rheumatoid arthritis, Arthritis Rheum. 54 (2006) 692–701. [70] M.S. Krathen, A.B. Gottlieb, P.J. Mease, Pharmacologic immunomodulation and cutaneous malignancy in rheumatoid arthritis, psoriasis, and psoriatic arthritis, J. Rheumatol. 37 (2010) 2205–2215. [71] E.A. Carswell, L.J. Old, R.L. Kassel, S. Green, N. Fiore, B. Williamson, An endotoxin-induced serum factor that causes necrosis of tumors, Proc. Natl. Acad. Sci. USA 72 (1975) 3666–3670.

11

[72] J.P. Waters, J.S. Pober, J.R. Bradley, Tumour necrosis factor and cancer, J. Pathol. 230 (2013) 241–248. [73] J.P. Waters, J.S. Pober, J.R. Bradley, Tumour necrosis factor in infectious disease, J. Pathol. 230 (2013) 132–147. [74] J. Ariyaratnam, V. Subramanian, Association between thiopurine use and nonmelanoma skin cancers in patients with inflammatory bowel disease: a meta-analysis, Am. J. Gastroenterol. 109 (2014) 163–169. [75] F.I. Scott, R. Mamtani, C.M. Brensinger, K. Haynes, Z.C. Chiesa-Fuxench, J. Zhang, et al., Risk of nonmelanoma skin cancer associated with the use of immunosuppressant and biologic agents in patients with a history of autoimmune disease and nonmelanoma skin cancer, JAMA Dermatol. 152 (2016) 164–172.

Please cite this article in press as: Y. Chen et al., Do tumor necrosis factor inhibitors increase cancer risk in patients with chronic immune-mediated inflammatory disorders?, Cytokine (2016), http://dx.doi.org/10.1016/j.cyto.2016.09.013