Malignancy and Rheumatoid Arthritis: Epidemiology, Risk Factors and Management

Diederik De Cock1 PhD

Kimme Hyrich1,2 MD PhD FRCPC

1 Arthritis Research UK Centre for Epidemiology, Centre for Musculoskeletal Research, University of Manchester, Manchester, United Kingdom2. National Institute of Health Research Manchester Biomedical Research Centre, Manchester NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester, United Kingdom

Corresponding Author:

Professor Kimme HyrichArthritis Research UK Centre for EpidemiologyDivision of Musculoskeletal and Dermatological SciencesSchool of Biological SciencesFaculty of Biology Medicine and HealthThe University of Manchester Stopford BuildingOxford Road, Manchester, M13 9PTUnited KingdomEmail: Kimme.hyrich@manchester.ac.uk

Disclosure: The Authors declare no conflicts of interest.

Funding: No specific funding was received for preparation of this manuscript.

Word Count: 7943

Abstract:

Rheumatoid arthritis (RA) is a chronic inflammatory condition which can result in pain and functional

disability. It is also associated with an increased occurrence of comorbidities, including an increased

risk of certain cancers such as lung cancer and lymphoma. The aetiopathogenesis of this increased

cancer risk is likely multifactorial and includes shared risk factors as well as chronic inflammation.

There is also a concern that the treatment for RA itself may increase this risk further, particularly

treatment with biologic disease modifying anti-rheumatic drugs (DMARDs). This paper aims to

review the evidence for the increased risk of cancer in RA as well as the latest evidence exploring the

association between DMARDs and tumorigeneses. It also discusses the evidence for management of

patients with biologic DMARDs in the setting of existing cancer.

Word Count: 115

Keywords: rheumatoid arthritis, disease modifying anti-rheumatic therapies, biologics, malignancy

Practice points:

1. Patients with rheumatoid arthritis have an increased risk of lymphoma and lung cancer which does not appear to be related to the use of biologics or methotrexate.

2. To date, no large observational studies have found a significant association between lymphoma or solid organ cancers and TNFi and these findings should be shared with patients.

3. Patients with rheumatoid arthritis are at an increased risk of skin cancer and information about skin surveillance and sun safety should be provided.

4. Data surrounding the safety of using TNFI in patients with a history of cancer are limited and decisions should be made at the individual patient level.

1. Introduction

Rheumatoid arthritis (RA) is a common musculoskeletal disease, affecting approximately 0.5-1.0% of

the adult population in Europe (1). RA is an insidious inflammatory auto-immune condition which

presents with joint swelling, pain and stiffness. It is neither preventable nor curable, and if left

uncontrolled disease progression will damage articular bone and cartilage (2).

In the past, RA was considered to be a disabling destructive disease associated with high mortality

and morbidity, including a recognised increased risk of certain cancers (3). Over the past decades,

the treatment of RA has been revolutionized, with more aggressive initial treatment strategies, such

as treat-to-target, as well as the introduction of biologics and other targeted therapies, leading to

dramatically improved clinical outcomes for many patients. Disease progression, deformities and

excess mortality, which characterized the disease as recently as two decades ago, can now be

avoided (2, 4); however, the increasing use of biologic therapies has also come with a concern about

whether the treatment of RA itself can increase the risk of certain comorbidities, including cancer.

Therefore, the aim of this review is to present what is known about the risk of cancer in RA, what

evidence there is for any additional risk associated with disease modifying anti rheumatic drugs

(DMARDs), and the evidence for management of RA with DMARDS in the setting of existing or prior

cancer.

2. Does rheumatoid arthritis increase the risk of developing cancer?

An association between RA and malignancy was first described by Isomaki et al. in 1978, who

identified that patients with RA have an increased risk of lymphoma (3). Since then, many studies

looking at this association have been published. The result of two meta-analyses, one being an

update from the other, including 23 studies published between 1990 and 2014. (5, 6). These reviews

included studies published in English with an observational study design and which compared

malignancy risk with the general population. The pooled standardised incidence ratio (SIR) of all site

malignancy risk was 1.09 (95% Confidence interval (CI) 1.06–1.13). In addition to all malignancies,

site specific malignancy risks were also reported, although not all studies in the analysis reported on

all cancer sites.

The overall pooled SIR (95 % CI) (12 studies) for lymphoma was 2.46 (2.05–2.96: Hodgkin disease (14

studies) pooled SIR (95 % CI) 3.21 (2.42–4.27); non-Hodgkin lymphoma (17 studies) pooled SIR (95 %

CI) of 2.26 (1.82–2.81). The analysis also identified an increased risk of lung cancer (23 studies;

pooled SIR 1.64 (95% CI 1.51–1.79); however, the risk of colorectal cancer (23 studies) (pooled SIR

(95 % CI) 0.78 (0.71–0.86)) and breast cancer (19 studies; pooled SIR (95 % CI) 0.86 (0.73–1.01)) were

both decreased. There was little to no association with other site cancers including cervical cancer

(15 studies) pooled SIR 0.87 (95% CI 0.72–1.05) and prostate cancer, pooled SIR (95% CI) 1.15 (0.98–

1.34). The more recent meta-analysis also suggested there may be an increased risk of malignant

melanoma (17 studies; pooled SIR (95% CI) 1.23 (1.01–1.49)) although a recent very large cross-

European study not included in the original meta-analysis has not confirmed this association (7).

These data suggest that although there is a small overall increased risk of cancer in patients with RA,

this risk is not the same across all sites, with a lowered risk compared to the general population

observed at some sites.

.

3. Why might patients with RA have an increased risk of cancer at certain sites and a

decreased risk at others?

Following on from these findings, it is important to consider why this differential risk of cancer

across sites may exist in patients with RA. One theory is that both certain cancers and RA have

shared risk factors. In part, this may explain the observed increased risk of lung cancer. The causal

relationship between lung cancer and tobacco use has been long established and smoking is

estimated to be responsible for an up to 85% cases of lung cancer (8). It is also known that smoking

is a risk factor for susceptibility to RA, increasing the risk by up to 40% (9). Hence, smoking may act

as a shared risk factor for both lung cancer and RA in certain individuals (10).

The theory of shared risk factors cannot explain all of the increased risk of cancer seen in RA. A study

by Hellgren et al. could not find any indication for a shared susceptibility between lymphoma and RA

(11). This study demonstrated that rates of lymphoma preceding the diagnosis of RA were not

increased compared with the rates that would be expected in the general population. Following the

diagnosis of RA, the risk increased substantially, particularly within the first 10 years after RA

diagnosis, with the highest rates earlier in the disease course. Although this study cannot rule out a

shared susceptibility to both diseases, either due to a shared genetic or environmental risk factor, it

suggests that the contribution of such shared aetiologies to an increased lymphoma risk after RA

diagnosis is likely to be low. Of course, one cannot rule out that there could be shared risk factors for

lymphomas which develop in older age, after the diagnosis of RA, risk factors for more lethal forms

of lymphoma, such that certain patients do not survive to develop RA, or that treatment for more

indolent forms of lymphoma may prevent the development of RA. There is also evidence for the

pathogenic effect of chronic immune stimulation/chronic inflammation in lymphomagenesis,

suggesting that RA itself could lead to the increased risk of lymphoma (12).

The observation of a reduced risk of colon cancer may relate to the common use of non-steroidal

anti-inflammatory drugs (NSAIDs) and previously aspirin in RA. NSAIDs are frequently used for

symptomatic pain and inflammation control in RA patients. It is known from use in the general

population that NSAID use is correlated with a decrease in colorectal cancer risk (13) and it is

possible a similar effect is being seen among patients with RA (14).

4. Risk of cancer with anti-rheumatic therapies

With the increasing use of immunosuppressive DMARDs and biologic therapies in the management

of RA, there has been heightened interest in whether or not the treatment for RA can increase the

risk of certain cancers, including contributing to the increased risk of lymphoma. Although the

association between cancer and RA had already been observed prior to the widespread use of

methotrexate and biologic therapies, more recent studies of malignancy risk in RA have not found

that the risk has been significantly reduced with the widespread use of DMARD therapies and better

disease control (15).

The mainstay of managing RA is DMARD therapy (4, 16). These can be categorised into conventional

synthetic (cs) DMARDs, biologic (b) DMARDS and most recently introduced, targeted synthetic (ts)

DMARDs, such as the Janus Kinase (JAK) Inhibitors. This next section summarises the evidence

surrounding the association between csDMARDs, bDMARDs (primarily tumour necrosis factor

inhibitors (TNFi)) and cancer. There have been minimal reports concerning the association between

tsDMARDs and cancer and these are also briefly summarised later in this review.

4.1 csDMARDs

Conventional synthetic DMARDs represent the most common group of DMARDs used in the

treatment of RA. The most common of these is methotrexate (MTX) although others include

sulphasalazine, hydroxychloroquine and leflunomide. Other less commonly used csDMARDs include

ciclosporin and azathioprine. csDMARDs, particularly MTX, are the gold standard first line therapies

for patients with RA and can be used as monotherapy or in combination with other csDMARDs,

bDMARDs and tsDMARDs.

Although originally used as a chemotherapeutic agent in patients with certain malignancies, MTX, at

significantly lower doses, is one of the safest and most effective therapies for RA (17). Whether or

not this low dose MTX, an immunosuppressing csDMARD, can itself increase the risk of cancer has

been evaluated. Overall MTX does not appear to increase the risk (18), although one study has

suggested a link between MTX exposure and risk of melanoma in patients with RA compared to the

general population. As this latter study was performed within an Australian population with a very

high baseline risk of melanoma, it is possible that the risk is different within this population. An

increased risk of non-Hodgkin lymphoma and lung cancer was also observed in MTX treated patients

when compared to the general population (19), but as this study did not have an untreated RA

control group, it is not possible to know if this risk was associated with MTX or the disease itself.

Higher doses of MTX, compared to lower doses, were also associated with an increased risk of

lymphoproliferative diseases in Japanese patients with RA (20). However, the concept of a direct role

of MTX potentiating the development of cancer is limited by confounding factors such as disease

activity and co-medication.

Other frequently used csDMARDs including sulphasalazine and hydroxychloroquine lack evidence for

an association with an increased risk of malignancies, although neither are felt to be significant

immunosuppressive DMARDs.

Some less commonly used csDMARDs, including ciclosporin and azathioprine, have been reported to

be associated with an increased risk of malignancy. Ciclosporin, although more commonly used now

in the management of transplantation rather than RA, is an immunosuppressing csDMARD. The

majority of studies regarding the development of malignancies following treatment with ciclosporin

stem from transplantation research, where a twofold increase in overall malignancy risk is observed,

with skin and genitourinary malignancies the most frequent (21). A meta-analysis of ciclosporin trials

in RA showed a 3 fold increase in the risk of malignancy (22) while a retrospective case control in RA

showed no effect compared to other patients with RA (23). While both studies are limited to give an

answer to the association between RA and ciclosporin, the evidence from other conditions seem to

demonstrate a carcinogenic effect of ciclosporin use, especially for non-melanoma skin cancer(24);

however, this relationship does not seem to be straightforward. Duration and dose of ciclosporin

therapy might be the driver for its malignancy risk, as shown for skin cancer in patients receiving

treatment with ciclosporin for skin conditions (25). Azathioprine, although also rarely used now in

the management of RA, has also been associated with an increased risk for malignancy, particularly

lymphoma (26, 27).

Leflunomide is another immunosuppressive csDMARD, used mostly for RA and psoriatic arthritis.

Although a signal for a possible increase in the rates of pancreatic cancer was observed in a single

German study, this finding was not replicated in 2 further analyses based in Sweden and the UK,

suggesting this may have been a spurious association (28).

4.2 bDMARDs.

Since the early 2000s, bDMARDs have become a second common class of DMARD used in the

treatment of RA. These drugs are large, complex molecules or mixtures of molecules derived from

living organisms and work by targeting specific points in the inflammatory process. Because of the

high cost associated with bDMARD treatment, these drugs are often reserved for people who have

not completely responded to csDMARDs and for those who cannot tolerate csDMARDs in doses

large enough to control inflammation. Biologics may be used alone as monotherapy, but research

has shown that many seem to be more effective when used in combination with MTX (29). As will

be discussed in the following sections, there has been a concern that bDMARDs, and in particular

tumour necrosis factor alpha inhibitors (TNFi), may increase the risk of malignancy. It is therefore a

requirement from regulators (such as the European Medicines Agency (EMA) or the United States

Food and Drug Administration (FDA)) that all pharmaceutical companies who bring bDMARDs to the

market have a risk management plan that includes continued monitoring of these therapies for

malignancy. For many bDMARDs, this includes monitoring of these drugs within national bDMARD

registries.

4.2.1 Tumour necrosis factor inhibitors

The original bDMARDs used in RA blocked the action of tumour necrosis factor alpha (TNF). This

cytokine was shown to play a key role in the inflammation in RA. In vivo and in vitro, TNF has also

shown the ability to lyse tumours. Hence, when introduced there was a concern that inhibition of

TNF may also have the ability to induce or increase the risk of malignancies. There are now 5

different TNFi available: etanercept, adalimumab, infliximab, certolizumab pegol, and golimumab.

These biologics differ subtly from each other. Etanercept is a recombinant TNF receptor antagonist.

Infliximab, adalimumab, and golimumab are monoclonal antibodies directed against TNF.

Certolizumab pegol is also a monoclonal antibody but is pegylated, in that it does not contain an Fc

portion and hence does not induce complement activation, antibody-dependent cellular cytotoxicity,

or apoptosis. The full impact of these structural differences on the efficacy and safety of these 5

agents is unknown, although generally their effectiveness appears to be similar. The 5 TNFi came to

market at different times: infliximab and etanercept in early 2000, adalimumab in 2004, and

certolizumab pegol and golimumab in 2009. As cancer is a rare event requiring large populations

studied for long periods of time, the majority of evidence around TNFi and cancer risk is based on

exposure to etanercept, infliximab and adalimumab.

4.2.2 TNF inhibitors and the overall risk malignancy risk in patients with RA

Data on the occurrence of malignancies in patients receiving TNFi have come primarily from two

sources: meta-analysis of clinical trial data and results of observational cohort studies. Two pooled

analyses of early clinical trial data on TNFi have suggested an increased overall cancer risk in patients

on TNFi (30-33). A 2 to 3-fold increase in risk was observed with a potential dose dependant relation

between treatment and malignancy. However, as these studies were meta-analyses of short-term

trials, these analyses have limitations due to their design. Firstly, the trials included all have small

numbers of patients with short periods of follow-up, in general being limited to 3 months to 1 year

of follow-up. Secondly, as in every controlled trial with a placebo group, this control group tend to

have a higher frequency of treatment discontinuation and thus drop out, resulting in a shorter

follow-up of the control group. For a relatively rare outcome such as malignancy, the difference in

follow-up may have influenced the cancer risk, especially as malignancy may have a longer latency

period compared to outcomes such as serious infection. A further meta-analysis which included a

wider range of trials did not confirm this original finding (33). Therefore, validation from

observational data, particularly from registries with large patient numbers with sufficient follow-up,

is needed. Overall, observational data have not confirmed an association between TNFI and an

increased risk of cancer (15, 34-38).

Many biologic register and observational drug cohorts exist across the world, forming a formidable

source for data concerning RA patients in daily practice (39, 40). As discussed, data from many of

these registries feed into the risk management plans for ongoing monitoring of bDMARDs in post-

marketed routine use. Most data about the risk of cancer has come from 4 national biologic

registries. The practice-based national Swedish Biologics Register (ARTIS) have included and followed

up patients with RA 16 years of older starting anti-TNF therapy since 1998 (41). This register has the

ability to link to other registries including general population, biologic naive and cancer registries

which makes it a very strong source of evidence for malignancy risk in patients with RA. The second

is the Danish DAN-BIO registry (42). Since 2000, registration to DANBIO is mandatory for all patients

with RA regardless of treatment. The British Society for Rheumatology Biologics Register for RA

(BSRBR-RA) has recruited patients starting biologics, alongside a parallel cohort of patients receiving

csDMARDs, since 2001. Patient recruitment is encouraged but not mandatory. (43, 44). Like the

Swedish register, both DAN-BIO and BSRBR-RA have the possibility to link to national cancer

registries as a source of cancer data. The German RABBIT register, also initiated in 2001, recruits

patients with RA starting a bDMARD or csDMARDs (45, 46). Unlike the other 3 registers, the German

register relies on direct report of cancer from the treating rheumatologist. These long-standing large

biologic registries form a formidable source of data within which to assess cancer risk in patients

with RA treated with different DMARDs.

In the Swedish Biologics Registry (35), no overall increased risk of malignancy risk was observed

during 25,693 person-years of follow-up in 6366 patients newly started on TNFi, compared with a

biologic-naïve RA cohort of 61,160 patients. In a second analysis, they also did not identify a risk

when compared to a specific population of patients with RA starting MTX or other csDMARD

combinations. Studies from other databases including the Danish, German and British biologic

registries and a large North American cohort have also found no increase in the overall risk of cancer

in TNFi treated patients compared to patients with RA not receiving these therapies (15, 34, 36, 37).

Of note, statistically significant differences in cancer risks between adalimumab, etanercept, and

infliximab were observed during the first year following the start of treatment in the Swedish study,

but these differences subsequently disappeared and have not been confirmed in other studies. This

could point out to subtle differences between TNFi or a possible bias by indication, especially as

these therapies were not all introduced at the same time.

Hence, most evidence from observational studies indicates no increase in overall malignancy risk.

The assumption here is that the effect of TNF inhibitor therapies is the same across all anatomical

sites and malignancy subtypes. A recent comprehensive systematic review and meta-analysis of 21

full manuscripts and 8 abstracts in registries and prospective observational studies have also

confirmed this finding. (47). In this analysis, patients treated with TNFi did not have an additional

increased risk for malignancies overall with a pooled risk estimate of 0.95 (95% CI 0.85 to 1.05).

However, as seen with overall cancer risk in RA in general, it is possible that the risk for cancer at

individual cancer sites may be different.

4.2.3 TNFi inhibitors and the risk of lymphoma

Early reports had suggested a potential association between lymphoma occurrence in patients with

RA receiving TNFi (48, 49). A review of the post marketing surveillance system run by the US FDA,

MedWatch, identified 26 cases of lymphoproliferative disorders following treatment with TNFi.

Unfortunately this case series could not confirm any causality in the role of TNFi due to the lack of

overall denominator of exposed patients as well as the lack of a comparison cohort (48).

Fortunately, the majority of recent evidence in multiple large robust population- and registry-based

studies from various countries, including the United States, United Kingdom, Sweden, Denmark, and

France have not been able to confirm any increased risk in the occurrence of lymphoma in patients

receiving TNFi compared to other csDMARDs, especially after adjusting for any differences in disease

severity (30, 36, 50-52) (Table 1).

A population-based study from Sweden (30) included prevalent RA cases, incident RA cases, and

patients treated with TNFi. Data from these patients were linked to the Swedish Cancer Registry. The

study confirmed the previous finding of a two-fold increased risk of lymphoma among patients with

RA when compared to the general population, including a 2.9 fold increased risk in TNFi treated

patients compared to the general population. However, after adjustment for age, sex, and disease

duration, the risk of lymphoma among patients in the TNFi cohort was not higher than those

patients with RA not receiving TNFi therapy with a relative risk of 1.1 (95% 0.60 -2.1).

This finding was confirmed in a longitudinal study in the United States (50), a national Danish

analysis (36) and most recently in the British registry (52).

Unfortunately as lymphoma is a relatively rare event in RA, estimated to occur at a rate of

130/100,000 person years (53), it is possible that small risks in lymphoma have not been ruled out

due to the fact that even large national registers will not be powered enough to detect small

increases in the risk of rare events.

It is increasingly recognised that non-Hodgkin’s lymphoma is not a single disease and although the

overall risk for NHL does not appear to be increased, there has been a concern that TNFi may

increase the risk of hepatosplenic T-cell lymphoma, a very rare form of non-Hodgkin lymphoma

following case reports, primarily among young men receiving TNFi for Crohn’s disease. Many of

these patients were also receiving azathioprine (54). To explore this further, a recent collaboration

between 12 European national biologic registries including more than 120 000 patients with RA (>

45000 ever treated with TNFi) did not identify any difference in the distribution of patterns of

individual NHL subtypes between TNFi treated and untreated patients and specifically, among this

large cohort with RA, did not identify any cases of hepatosplenic T-cell lymphoma among 533 cases

of lymphoma.

4.3.4 TNF Inhibitors and the risk of solid organ cancers

A meta-analysis of 18 clinical trials including adalimumab, infliximab, and etanercept treatment, did

not show any increased risk of solid cancers in TNFi treated patients compared to placebo (33).

However, the short follow-up time per patient, which was on average 10 months, was too short to

draw any conclusions regarding the risk over a longer term.

A number of large observational studies have now explored this in more detail (Table 2). A Swedish

study set within the Swedish RA registries concluded that the risk of solid malignancies in the TNFi

cohort was not greater than that of the general population in Sweden (55). The British biologics

register also confirmed these findings in patients with RA ever treated with a TNFi therapy,

compared with those never treated with such agents (15). This UK analysis also included a study of

the most common solid organ cancers, including lung, breast, colorectal, and gastroesophageal

cancers and again could not identify any difference in risk between TNFi treated patients and those

receiving csDMARDs only. An American study gathering data from 4 large administrative databases

found in 19750 TNFi treated patients a HR of 0.80 (95% CI 0.59 – 1.08) compared to 9805 patients on

csDMARD treatment only. Follow-up was short, less than 6 months on average, raising questions

about the validity of findings (56). A retrospective study from Taiwan observed a decreased hazard

ratio (0.46 (95% CI 0.27 – 0.79)) in patients on TNFi, all on etanercept, compared to matched biologic

naïve patients (38).

A recent Swedish study which focussed on cervical cancer in women with RA (57) suggested that

women receiving TNFi may be at an increased risk of cervical cancer (cervical intraepithelial

neoplasia 2 or higher-grade changes and invasive cervical cancer) as compared to patients with RA

not receiving this therapy. However, as immunosuppression may drive this risk and the exposure to

DMARDs was different between the 2 groups, they undertook a sensitivity analysis restricting to

patients with a more recent diagnosis of RA and to those who had received 2 or more csDMARDs. In

this analysis, which aimed to more closely match TNFi patients with controls, the observed increased

risk was no longer evident.

There has been a note regarding a possible increased risk of cancer among patients with ANCA-

associated vasculitis who have received cyclophosphamide in combination with etanercept (58).

However, this is rarely used in the treatment of RA. Although the overall number of malignancies

observed in a clinical trial of etanercept versus placebo was small, the high number of solid

malignancies observed over a period of only two years suggested that the combination of TNF

inhibition and cyclophosphamide heightens the risk of cancer beyond the risk observed with

cyclophosphamide alone. Hence, the manufacturer's labelling for etanercept now states that use of

etanercept in patients receiving concurrent cyclophosphamide is not recommended due to an

increased risk of developing cancer (59).

4.3.5 TNF inhibition and the risk of skin cancer

Clinical trial evidence seems to indicate an increased risk of non-melanoma skin cancer among

patients treated with TNFi compared with those not receiving these agents (47, 50, 60-62). The

evidence for an increased risk of non-melanoma skin cancer with use of TNF inhibitors is

demonstrated by two meta-analyses (47, 61). One used data from more than 15 000 patients across

74 clinical trials and showed an increased relative risk of 2.02 (95%CI 1.11–3.95).The other included

data from observational studies including around 28 000 participants and showed a pooled estimate

of risk of 1.45 (95% CI 1.15 to 1.76). An observational study published at the same time using data

from the US Department of Veterans Affairs showed similar results in more than 20 000 American

patients with RA (60). However, this has not been a consistent finding (Table 3) and two large

observational studies with generally longer periods of follow-up have not observed an increased risk,

especially after adjusting for confounding variables using propensity scores (36, 63). The Danish

biologics registry found no increased risk of non-melanoma skin cancer in 3347 TNFi treated patients

and the British biologics register also found no increased risk among 11 881 TNFi treated patients

compared to 3629 csDMARD treated controls; however, both studies observed an increased risk of

non-melanoma skin cancer in patients with RA, regardless of whether treated with TNFi or

csDMARD, when compared with the general population. This risk increase ranged from a 76 to 83

percent increase in patients with RA on csDMARD treatment and a 72 to 92 percent increase among

patients treated with TNFi. These findings suggest that patients with RA have an overall increased

risk of non-melanoma skin cancer, either due to immunosuppressive therapy, the underlying

disease, or other unmeasured factors, such as differential sun exposure. These studies could not

differentiate between these. Based on these findings it is suggested that all patients with RA,

regardless of treatment, should be screened for skin cancer. Patients should also be educated

regarding skin cancer prevention, including use of broad-spectrum sunscreens when in the sun, sun

avoidance, and skin self-examination.

These studies were largely limited to patients who were not known to have a history of skin cancer.

However, the risk of new skin cancers may be different among patients receiving TNFi therapy who

already have a history of this cancer (Table 4). Among 106 csDMARD and 177 TNFi treated patients

with a history of non-melanoma skin cancer, the BSRBR-RA did not identify any difference in

occurrence of second or further skin cancers between patients receiving TNFI or csDMARDs;

however, in both groups, the crude incidence rate of a new skin cancer was almost 10-fold higher

compared with those without a previous skin cancer history (63).

The risk of melanoma may be different from that of non-melanoma skin cancer (Table 3). TNF

infusions have been used to successfully treat melanoma and therefore there is a concern that

blocking TNF may increase the risk of this cancer (64). Studying melanoma is challenging, as this

form of skin cancer is rare. A large Swedish population-based cohort study that included patients

with RA who were either treated with a TNFi or were biologic naïve and a third control cohort from

the general population, found that the risk of a first invasive melanoma was modestly increased in

patients receiving TNFi therapy compared with patients with RA who were not treated with biologic

agents with a hazard ratio of 1.5 (95%CI 1.0; 2.2) (62). A subsequent meta-analysis of observational

studies (65) found a meta-SIR of 1.7 (95% CI 1.2 to 2.3) in TNFi treated patients when compared to

the general population, but the meta-SIR for patients treated with TNFi compared with those

without TNFi was not increased (1.4 (95% CI 0.8 to 2.3)). An imprecise estimate was found likely due

to required adjustment for heterogeneity and the authors could therefore not rule out that patients

with RA treated with TNFi are at a raised risk of melanoma in comparison with patients TNFi naïve.

Therefore, given the rarity of this cancer, a pooled analysis of 11 European biologic registries was

undertaken (7).This analysis, which included 93 melanoma cases in 220 259 patient years of TNFi

exposure among patients with RA did not confirm the increased risk of melanoma observed in the

Swedish cohort (meta-SIR 1.2 (95% CI 0.9 to 1.6) when compared to the general population. There

was a large variation in risk noted across the different countries, although this was felt to be

explained largely by the low rates of melanoma within each registry.

4.3.6 TNF Inhibitors and the Risk of Malignancy in Other Autoimmune Diseases

The subtle interplay between auto-immune conditions, inflammation, and malignancy onset has

been investigated in many conditions (66), yet there are few studies of cancer risk in autoimmune

conditions commonly treated with TNFi other than RA.

A Danish nationwide registry study of 4553 patients receiving TNFi for inflammatory bowel disease

compared to patients with IBD not receiving these therapies did not find any significant association

between TNFi exposure and cancer (67).

A 2018 publication looked at the risk of all-site malignancy and non-melanoma skin cancers among

patients with PsA registered in the BSRBR who had received TNFi (68). Compared to the general

population, this UK study did not find an overall increased risk of malignancy in patients with PsA

receiving TNFi, but it did find a significant increased risk of skin cancer (SIR 2.12; 95% CI: 1.19, 3.50).

It is not possible given this study design to know whether this increased risk in skin cancer was due

to TNFi or an already increased risk among patients with psoriatic arthritis and psoriasis, as no

untreated PsA population was included. It has been shown that patients with psoriasis are at an

increased risk of non-melanoma skin cancer, like due to UVA exposure (68). Reassuringly, a

Swedish/Danish nationwide study set within the ARTIS and DANBIO registries found that among

8703 patients with spondyloarthritis (SpA) including PsA and ankylosing spondylitis, who had

received TNFi there was no increase in risk of lymphoma when compared to 28,164 patients

biologic-naïve patients with SpA (69).

4.2.6 Non-TNF inhibitors and the risk of malignancy in RA

Following the success of TNFi in RA, a number of other classes of bDMARD therapies have now been

developed and are in increasingly wide spread use. These include rituximab, an antibody that binds

to the cell surface protein CD20. CD20 is widely expressed on B cells, from early pre-B cells to later in

B-cell differentiation. The mode of action of rituximab in controlling disease activity in RA is not

completely understood, but treatment does result in the elimination of B cells. Abatacept is a CTLA-4

inhibitor, which works by blocking the co-stimulatory signal between antigen presenting cells and T

cells, thus blocking activation of T-cells. Tocilizumab and sarilumab are Interleukin 6 (IL-6) inhibitors,

Anakinra is an IL-1 inhibitor. As these therapies are newer, there is less evidence surrounding any

association between exposure and the development of malignancy. This next section summarises

this evidence.

Rituximab was the first non-TNFi available for management of RA and most evidence regarding

malignancy risk is available for this non-TNFi bDMARD. Rituximab was first introduced as a successful

treatment of B-cell lymphoma and therefore, there has been less concern that rituximab may

increase the risk of lymphoma in RA (70). A pooled analysis of safety data from patients with RA

treated with rituximab in randomized control trials with follow-up out to 11 years, providing 14 816

patient years of observation, did not reveal any increase in the risk on malignancy overall compared

to a placebo population(71). This report concluded that the observed rates of cancer overall and per

type of cancer were within the expected ranges for the RA population. A Finnish register study,

albeit in a low number of patients with RA treated with rituximab with limited follow-up, saw the

highest crude incidence rates for rituximab (72). However, after adjusting for confounding, the

observed differences between TNFi, RTX and csDMARDs disappeared. A small French study

confirmed these findings of no increased risk of cancer with rituximab compared to a previously

studied csDMARD treated cohort (73). A recent large Swedish population further strengthened these

findings in an analysis which included multiple comparator cohorts consisting of TNFi or biologic

naïve treated patients, and the general population (74). One challenge in studying the risk of cancer

in patients receiving rituximab, however, is the almost universal past exposure to TNFi, and

therefore with time it may be difficult to disentangle any observed risk associated with rituximab

with any carried risk from previous treatments.

There is considerably less experience with the risk of malignancies associated with long-term

treatment with other non-TNFi bDMARDs, including abatacept, sarilumab and tocilizumab. Although

immunosuppression with these drugs could theoretically facilitate tumour development, so far there

have been no signals for increased malignancies in patients treated with these agents (74-82). The

majority of these studies were done as meta-analyses of clinical trials of these agents, comparing

exposure to placebo, but most lacked a high number of patients or follow-up.

A Swedish population study did observe an increased risk of squamous cell skin cancer in patients on

abatacept with a hazard ratio of 2.12 (95%CI 1.14 – 3.95) compared to the TNFi treated patients (74).

It is to be noted that, as the authors admit, this study suffered from the risk of multiple testing and

this association would be non-significant after adjusting with a multiple test correction. Additionally,

patients on abatacept received more prescription drugs and had a much higher proportion of prior

TNFi therapy. Although the analysis corrected for these differences in drug history, it is still possible

drug interactions or overlapping drug exposure times are a possible explanation for this increased

observed risk, and not abatacept on its own; however, the trend of a possible association between

abatacept and squamous cell skin cancer had already been observed in earlier research on

abatacept. The relative risk factor of a first non-melanoma skin cancers was found to be increased

for RA patients on abatacept compared to methotrexate with a relative risk of 15.3(95% CI 2.05–114)

(37) Confidence intervals were wide due to low cases of non-melanoma skin cancer, showing the

possibility of an imprecise estimate.

4.2.7 Risk of malignancy with targeted synthetic DMARDs

The armamentarium for RA treatments is still being increased. Recently, a new type of anti-

rheumatic therapy has been introduced in daily clinical practice: JAK inhibitors, which are classified

as tsDMARDs. Current available tsDMARDs include tofacitinib, a JAK1/JAK3 inhibitor and baricitinib,

a JAK1/JAK2 inhibitor. Like biologics, they target a specific point in the immune pathway but unlike

biologics, these therapies are small chemical molecules and both are currently available in oral

formulations. As these drugs were only recently introduced, data on how or if these novel agents

influence the risk on malignancies for patients with RA are limited.

The more studied tsDMARD is tofacitinib, with most data emerging from the USA, where it received

marketing authorisation in 2012, preceding market authorisation in Europe by many years. Most

evidence stems from the clinical trial program (83). A meta-analysis of malignancy risk with

tofacitinib included over 4 000 patients totalling 12 664 patient years with a median exposure of

2.35 years per patient. No increase in the risk of malignancies overall or any specific type of

malignancy in RA patients had been observed nor has any dose related risk been observed compared

to csDMARD use. A study of the post marketing surveillance database of Pfizer Inc. , which markets

tofacitinib, concerning 3 year data on tofacinitib confirmed these findings (84).

Evidence for the other approved tsDMARD, baricitinib, is even more limited. Current meta-analyses

do not contain enough cancer cases nor follow-up time to draw any conclusions (85). As most

evidence is limited to meta-analyses of trials or risk observed within individuals trials, the need for

data from observational data is clear in order establish if an increased risk on malignancies exists in

patients with RA treated with tsDMARDs.

4.3 The risk of cancer following bDMARD treatment in patients with a prior malignancy risk

The majority of evidence discussed thus far is limited to the occurrence of a first cancer in patients

with no known cancer history. However, the use of bDMARDs in patients with a history of cancer

may confer a different risk on the risk of a second new cancer or recurrence of the prior cancer.

Patients with prior cancer history have been excluded from clinical trials and therefore, have not

been included in any meta-analysis of these same trials. However, observational registries allow the

study of outcomes among these patients who do go on to receive bDMARD therapy in routine care.

Having a prior malignancy increases the risk of current malignancy in an individual (86). Agents such

as TNFi could theoretically influence this risk as TNF has many physiologic roles, including host

defence and tumour surveillance. Hence, TNFi inhibition could trigger malignancy recurrence or

development of new cancers in patients with a prior malignancy history.

Currently, only a limited number of studies have looked at the further occurrence of malignancy in

patients with a history of malignancy prior to starting bDMARDs (Table 4). The first, a small nested

case-control study set within the German national biologics register did not find any significant

difference between recurrent cancers in patients with prior cancer treated with either TNFi (n=67) or

csDMARDs (n=55) (34).

A small study set within the British national biologics register investigated the risk of overall cancer

occurrence in patients with a history of prior cancer treated with either csDMARDs (n = 159), TNFi (n 

= 243) or RTX (n = 23) (87). The study found a decreased risk of new cancer diagnosis in patients

receiving TNFi or RTX compared to csDMARDs. The authors hypothesised that patients with a higher

risk of cancer were steered away from bDMARD treatment, although they did not have access to

cancer severity or detailed history other than site in any patient. There was a higher rate of

recurrence of the same cancer and a shorter time between prior cancer and registration in the study

in patients receiving csDMARDs compared to TNFi.

As stated, neither the German nor the British study had access to information on cancer severity,

stage or treatment, which may influence the risk of recurrence independent of any other

treatments. To overcome this to some degree, a recent publication from Sweden matched study

patients on TNFi with a comparator of unexposed patients using the metastasis classification system

for their prior cancer (88). This study found no difference in the risk of a recurrent cancer overall or

in any of the types of solid organ malignancy included in the study between TNFi and biologic naïve

patients.

The three studies presented above looked at recurrence of cancer across a wide group of cancer

sites and it may be that recurrent cancer risk is higher at some sites compared to other. A

retrospective cohort study using data from the US Department of Veterans Affairs administrative

databases studied head and neck cancer recurrence (89). This study included 180 patients with RA

and a previous head and neck cancer and showed that recurrence rates did not differ significantly

between patients subsequently treated with TNFi and those treated with csDMARDs.

Two studies have looked into the risk of recurrent breast cancer (90, 91). The first, from Sweden,

found no difference in the rate of recurrent breast cancer among a matched cohort of 120 women

starting TNF inhibitor therapy and 120 biologic-naïve patients with RA and previous breast cancer.

The average time from previous breast cancer to start of the study was 9.4 years. Nine women

developed breast cancer recurrence in each group after a median of 4.9 years of follow-up. A second

study using United States Medicare data (91) also found no increase in risk of recurrent breast

cancer among 291 women with either RA or inflammatory bowel disease receiving TNFi treatment

over a median follow-up of 2.7 years, compared with 1164 matched nonusers. In this study, the

majority of women had had their initial breast cancer surgery one year prior to the start of follow-

up.

Studying the risk of recurrent or new cancers among patients receiving TNFi who have a prior history

of cancer is challenging. Although a number of studies have attempted to look at this issue, even

those studies set within very large databases have only been able to identify at most a few hundred

patients with this history. In part this likely relates to anxiety about use in this situation and also

national guidelines, which often state that a period of time should have passed before bDMARDs are

considered. This suggested time has not been evidence based but more likely relates to a period

after which cancers are more likely to have been considered cured. This has led, however, to the

mean time between previous cancer and start of bDMARD to be greater than 5 years in most

instances and therefore the safety of using bDMARDs in patients with more recently diagnosed

cancers remains largely unstudied. Although some studies have attempted to match on cancer stage

and severity, these can only be matched on what has been recorded in medical records and there

are likely to be other unmeasured factors involved in treatment decisions that may differ between

treated groups of patients. Moreover, because of the lack of any data, treating rheumatologists

could be reluctant to treat such patients with TNFi, resulting in a selection bias, with only patients

perceived as having a low(er) risk for cancer going on to receive TNFi treatment.

5. Confounders for estimating the risk of malignancies in patients with RA

Understanding and studying the association between bDMARDs and the risk of cancer induction is

not without its challenges. This next section outlines some of these key considerations.

5.1 Channelling bias

Channelling bias is a form of drug allocation bias where groups of patients with different prognostic

profiles are prescribed drugs with similar therapeutic indications. Comparison of outcomes between

these drugs then leads to observations of differences between treatments, not due to the different

drugs but due to differences in the groups of patients receiving these drugs. This bias should be

considered when studying the relationship between cancer and choice of DMARDs. An example here

is the study of lymphoma. Patients with RA have an increased risk of lymphoma compared to the

general population. Therefore, the observation of an increased risk of lymphoma in comparison to

the general population cannot differentiate between the risk associated with disease and any

additional risk conferred by the treatment. It is also recognised that chronic inflammation may also

increase this risk of lymphoma further (12). Patients who start a TNFi or other bDMARD usually do so

as they have high disease activity uncontrolled by csDMARDs. Therefore these same patients may

also be those at higher risk of lymphoma. Therefore, when analysing the risk of this and other

cancers it is important to consider any differences between treated and untreated patients and

allow for these in statistical analysis by taking these confounders into account (92).

Frisell et al. recently illustrated this channelling bias in the choice of bDMARDs. Older and less

healthy patients with RA seem to be selected for non-TNFi bDMARD therapy, in particular as first

bDMARD. This selection bias can substantially confound the results of non-randomised comparative

studies including non-TNFi agents (93).

5.2 Treatment exposure and cancer latency

Another challenge is the need for long-follow-up in cancer research. Trials usually have short follow-

up, so have limited evidence of the long-term risk. Moreover, in observational studies, patients are

exposed to multiple therapies throughout their treatment, hampering the attribution of cancer

development to one treatment type. Only one Swedish study in patients with RA on TNFi therapy

looked at the risk of cancer overall over time in different time intervals and found no trends toward

increasing or decreasing risk over time on TNFi or biologic naïve therapy, nor between these two

groups (94).

5.3 Cancer and its subtypes are a relatively rare event

As discussed above, the risk of cancer in patients with RA is not the same for all patients with RA.

Therefore, most studies have looked either in cancer risk overall or in one specific subtype. However,

researching some subtypes of cancer in a national register can be limited, as some types of

malignancies are exceedingly rare. Although lymphoma is one of the 10 most common cancer types

in the general population, the age-standardised incidence rate (IR) in Europe for lymphoma is

approximately 25/100 000 (53, 95). Although many registries have a few thousand patients with RA

on a bDMARD therapy, this may not be enough to detect very small increases in risk. Therefore,

consortiums of different registries working together to estimate these risks of a cancer subtype will

be necessary (52, 96).

6. Clinical Practice Points and Research Agenda

The evidence to date does not support a strong association between TNFi and cancer occurrence in

patients receiving these therapies who have no prior history of cancer and patients should be made

aware of this information when bDMARDs are being prescribed. Due to the rarity of certain types of

individual cancers, however, small increases in risk cannot be ruled out but are likely, in terms of

absolute risk, to be very small. However, most studies have only followed patients for on average 5

years and therefore, research should continue in this area to identify late risks of malignancy, such

as through national registers.

Whether there is an increased risk of certain types of cancers such as non-melanoma skin cancer,

remains less clear, with some studies finding an increase and others not. Two large studies did

identify an increased risk across all patients with RA receiving either csDMARDs or bDMARDs and

therefore, patients thus should be advised on preventative skin care and skin surveillance and should

be prompted to report any new persistent skin lesions.

Although the data on the risk of recurrent cancers in patients receiving bDMARDs who have had

prior cancer is reassuring, all studies to date have been very small and have included a select group

of patients, many with a history of cancer many years prior to starting TNFi. Therefore whether

these same findings will be true if a group of patients with very recent cancer are studied is

unknown. The data available does not suggest that prior cancer is an absolute contraindication to

bDMARDs. Until further information becomes available, the decision to treat a patient with a prior

cancer should be a joint decision between the patient, their rheumatologist and their oncologist,

taking into account the nature of the previous cancer, the treatment received but also the activity of

the patient’s arthritis and their quality of life, considering what alternative therapies there may be. It

is currently unknown whether switching class of bDMARD is a better decision that continuing the

same bDMARD and this question should form part of a future research agenda. It is also currently

unknown whether bDMARDs increase the risk of cancer in patients with pre-malignant conditions,

such as Barret’s oesophagus, or in those with a strong family history of cancer, such as women with

a family history of breast and/or ovarian cancer. These patients are already at an increased risk of

cancer and this background risk must be considered in any analysis of outcome in patients

subsequently treated with bDMARDs. What remains unknown is how physicians should manage

anti-rheumatic therapies in patients who develop a cancer while receiving bDMARDs. The current

British guidelines recommend that bDMARD therapy should be stopped but do not offer specific

advice on when it is safe to restart (97). The American guidelines give specific recommendations

which DMARD to use for the following types of cancer: previously treated or untreated skin cancer,

previously treated lymphoproliferative disorders and previously treated solid organ malignancies

(16).This lack of clear guidelines is not surprising given there is very little evidence to inform them.

Only one study, in abstract form only, has reported on clinical decision making with respect to

bDMARDs following diagnosis of cancer in patients already receiving TNFi (98). Among 404 cancers

among over 12 000 bDMARD treated patients, over two thirds of patients who survived at least 6

months following their cancer diagnosis had their TNFi stopped at point of cancer diagnosis. Over

the next 4.5 years, over half remained off biologic therapy and for those who did restart, a majority

switched class of bDMARD.

7. Conclusion

To conclude, it has long been recognised that RA in associated with an increased risk of malignancy

overall although this is largely reflected in large increases in risk of lymphoma and lung cancer and

lowered risks of colon and breast cancer. The role of csDMARDs in this increased risk is not clear, as

many estimates of risk predate the widespread use of MTX. Some csDMARDs, such as azathioprine

and ciclosporin are now rarely used to manage RA.

The advent of bDMARDS has raised this concern again but reassuringly the overall risks associated

with bDMARDs, primarily TNFi, appear to be low after taking into account confounding, but as

discussed, not enough research has been done to rule out this risk completely.

A majority of research on the risk associated with bDMARDs has been limited to etanercept,

adalimumab and infliximab. The risks associated with certolizumab pegol and golimumab may be

similar but this is unknown. The risks associated with non-TNFi bMDARDS are largely unknown,

although to date, no concerning safety signals have arisen in meta-analysis of clinical trial data,

albeit within their limitations. A clear need exists for ongoing research into the safety, with respect

to malignancies, across all bDMARD therapies as well as the new tsDMARDs (99).

References

1. Alamanos Y, Voulgari PV, Drosos AA. Incidence and prevalence of rheumatoid arthritis, based on the 1987 American College of Rheumatology criteria: a systematic review. Seminars in arthritis and rheumatism. 2006;36(3):182-8.2. Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. Lancet. 2016;388(10055):2023-38.3. Isomaki HA, Hakulinen T, Joutsenlahti U. Excess risk of lymphomas, leukemia and myeloma in patients with rheumatoid arthritis. Journal of chronic diseases. 1978;31(11):691-6.4. Smolen JS, Landewe R, Bijlsma J, Burmester G, Chatzidionysiou K, Dougados M, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update. Annals of the rheumatic diseases. 2017;76(6):960-77.5. Smitten AL, Simon TA, Hochberg MC, Suissa S. A meta-analysis of the incidence of malignancy in adult patients with rheumatoid arthritis. Arthritis research & therapy. 2008;10(2):R45.6. Simon TA, Thompson A, Gandhi KK, Hochberg MC, Suissa S. Incidence of malignancy in adult patients with rheumatoid arthritis: a meta-analysis. Arthritis research & therapy. 2015;17:212.7. Mercer LK, Askling J, Raaschou P, Dixon WG, Dreyer L, Hetland ML, et al. Risk of invasive melanoma in patients with rheumatoid arthritis treated with biologics: results from a collaborative project of 11 European biologic registers. Annals of the rheumatic diseases. 2017;76(2):386-91.8. Warren GW, Cummings KM. Tobacco and lung cancer: risks, trends, and outcomes in patients with cancer. American Society of Clinical Oncology educational book American Society of Clinical Oncology Annual Meeting. 2013:359-64.9. Sugiyama D, Nishimura K, Tamaki K, Tsuji G, Nakazawa T, Morinobu A, et al. Impact of smoking as a risk factor for developing rheumatoid arthritis: a meta-analysis of observational studies. Annals of the rheumatic diseases. 2010;69(1):70-81.10. Peacock A, Leung J, Larney S, Colledge S, Hickman M, Rehm J, et al. Global statistics on alcohol, tobacco and illicit drug use: 2017 status report. Addiction. 2018.11. Hellgren K, Smedby KE, Feltelius N, Baecklund E, Askling J. Do rheumatoid arthritis and lymphoma share risk factors?: a comparison of lymphoma and cancer risks before and after diagnosis of rheumatoid arthritis. Arthritis and rheumatism. 2010;62(5):1252-8.12. Baecklund E, Iliadou A, Askling J, Ekbom A, Backlin C, Granath F, et al. Association of chronic inflammation, not its treatment, with increased lymphoma risk in rheumatoid arthritis. Arthritis and rheumatism. 2006;54(3):692-701.13. Din FV, Theodoratou E, Farrington SM, Tenesa A, Barnetson RA, Cetnarskyj R, et al. Effect of aspirin and NSAIDs on risk and survival from colorectal cancer. Gut. 2010;59(12):1670-9.14. Kauppi M, Pukkala E, Isomaki H. Low incidence of colorectal cancer in patients with rheumatoid arthritis. Clinical and experimental rheumatology. 1996;14(5):551-3.15. Mercer LK, Lunt M, Low AL, Dixon WG, Watson KD, Symmons DP, 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. Annals of the rheumatic diseases. 2015;74(6):1087-93.16. Singh JA, Saag KG, Bridges SL, Jr., Akl EA, Bannuru RR, Sullivan MC, et al. 2015 American College of Rheumatology Guideline for the Treatment of Rheumatoid Arthritis. Arthritis care & research. 2016;68(1):1-25.17. Lopez-Olivo MA, Siddhanamatha HR, Shea B, Tugwell P, Wells GA, Suarez-Almazor ME. Methotrexate for treating rheumatoid arthritis. The Cochrane database of systematic reviews. 2014(6):CD000957.18. Raheel S, Crowson CS, Wright K, Matteson EL. Risk of Malignant Neoplasm in Patients with Incident Rheumatoid Arthritis 1980-2007 in relation to a Comparator Cohort: A Population-Based Study. International journal of rheumatology. 2016;2016:4609486.

19. Buchbinder R, Barber M, Heuzenroeder L, Wluka AE, Giles G, Hall S, et al. Incidence of melanoma and other malignancies among rheumatoid arthritis patients treated with methotrexate. Arthritis and rheumatism. 2008;59(6):794-9.20. Kameda T, Dobashi H, Miyatake N, Inoo M, Onishi I, Kurata N, et al. Association of higher methotrexate dose with lymphoproliferative disease onset in rheumatoid arthritis patients. Arthritis care & research. 2014;66(9):1302-9.21. Cockburn IT, Krupp P. The risk of neoplasms in patients treated with cyclosporine A. Journal of autoimmunity. 1989;2(5):723-31.22. Arellano F, Krupp P. Malignancies in rheumatoid arthritis patients treated with cyclosporin A. British journal of rheumatology. 1993;32 Suppl 1:72-5.23. van den Borne BE, Landewe RB, Houkes I, Schild F, van der Heyden PC, Hazes JM, et al. No increased risk of malignancies and mortality in cyclosporin A-treated patients with rheumatoid arthritis. Arthritis and rheumatism. 1998;41(11):1930-7.24. Durnian JM, Stewart RM, Tatham R, Batterbury M, Kaye SB. Cyclosporin-A associated malignancy. Clinical ophthalmology. 2007;1(4):421-30.25. Muellenhoff MW, Koo JY. Cyclosporine and skin cancer: an international dermatologic perspective over 25 years of experience. A comprehensive review and pursuit to define safe use of cyclosporine in dermatology. The Journal of dermatological treatment. 2012;23(4):290-304.26. Williams CA, Bloch DA, Sibley J, Haga M, Wolfe F, Raynauld JP, et al. Lymphoma and luekemia in rheumatoid arthritis: are they associated with azathioprine, cyclophosphamide, or methotrexate? Journal of clinical rheumatology : practical reports on rheumatic & musculoskeletal diseases. 1996;2(2):64-72.27. Bernatsky S, Clarke AE, Suissa S. Hematologic malignant neoplasms after drug exposure in rheumatoid arthritis. Archives of internal medicine. 2008;168(4):378-81.28. Strangfeld A, Hyrich K, Askling J, Arkema E, Davies R, Listing J, et al. Detection and evaluation of a drug safety signal concerning pancreatic cancer: lessons from a joint approach of three European biologics registers. Rheumatology. 2011;50(1):146-51.29. Gabay C, Riek M, Scherer A, Finckh A, physicians Sc. Effectiveness of biologic DMARDs in monotherapy versus in combination with synthetic DMARDs in rheumatoid arthritis: data from the Swiss Clinical Quality Management Registry. Rheumatology. 2015;54(9):1664-72.30. Askling J, Fored CM, Baecklund E, Brandt L, Backlin C, Ekbom A, et al. Haematopoietic malignancies in rheumatoid arthritis: lymphoma risk and characteristics after exposure to tumour necrosis factor antagonists. Annals of the rheumatic diseases. 2005;64(10):1414-20.31. Bongartz T, Sutton AJ, Sweeting MJ, Buchan I, Matteson EL, Montori V. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infections and malignancies: systematic review and meta-analysis of rare harmful effects in randomized controlled trials. Jama. 2006;295(19):2275-85.32. Bongartz T, Warren FC, Mines D, Matteson EL, Abrams KR, Sutton AJ. Etanercept therapy in rheumatoid arthritis and the risk of malignancies: a systematic review and individual patient data meta-analysis of randomised controlled trials. Annals of the rheumatic diseases. 2009;68(7):1177-83.33. Leombruno JP, Einarson TR, Keystone EC. The safety of anti-tumour necrosis factor treatments in rheumatoid arthritis: meta and exposure-adjusted pooled analyses of serious adverse events. Annals of the rheumatic diseases. 2009;68(7):1136-45.34. Strangfeld A, Hierse F, Rau R, Burmester GR, Krummel-Lorenz B, Demary W, et al. Risk of incident or recurrent malignancies among patients with rheumatoid arthritis exposed to biologic therapy in the German biologics register RABBIT. Arthritis research & therapy. 2010;12(1):R5.35. Raaschou P, Simard JF, Neovius M, Askling J, Anti-Rheumatic Therapy in Sweden Study G. Does cancer that occurs during or after anti-tumor necrosis factor therapy have a worse prognosis? A national assessment of overall and site-specific cancer survival in rheumatoid arthritis patients treated with biologic agents. Arthritis and rheumatism. 2011;63(7):1812-22.

36. Dreyer L, Mellemkjaer L, Andersen AR, Bennett P, Poulsen UE, Juulsgaard Ellingsen T, 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. Annals of the rheumatic diseases. 2013;72(1):79-82.37. Solomon DH, Kremer JM, Fisher M, Curtis JR, Furer V, Harrold LR, et al. Comparative cancer risk associated with methotrexate, other non-biologic and biologic disease-modifying anti-rheumatic drugs. Seminars in arthritis and rheumatism. 2014;43(4):489-97.38. Lan JL, Tseng CH, Chen JH, Cheng CF, Liang WM, Tsay GJ. Reduced risk of all-cancer and solid cancer in Taiwanese patients with rheumatoid arthritis treated with etanercept, a TNF-alpha inhibitor. Medicine. 2017;96(7):e6055.39. Kearsley-Fleet L, Zavada J, Hetland ML, Nordstrom DC, Aaltonen KJ, Listing J, et al. The EULAR Study Group for Registers and Observational Drug Studies: comparability of the patient case mix in the European biologic disease modifying anti-rheumatic drug registers. Rheumatology. 2015;54(6):1074-9.40. Nikiphorou E, Buch MH, Hyrich KL. Biologics registers in RA: methodological aspects, current role and future applications. Nature reviews Rheumatology. 2017;13(8):503-10.41. Askling J, Fored CM, Geborek P, Jacobsson LT, van Vollenhoven R, Feltelius N, et al. Swedish registers to examine drug safety and clinical issues in RA. Annals of the rheumatic diseases. 2006;65(6):707-12.42. Ibfelt EH, Jensen DV, Hetland ML. The Danish nationwide clinical register for patients with rheumatoid arthritis: DANBIO. Clinical epidemiology. 2016;8:737-42.43. Watson K, Hyrich K. The role of the British Society for Rheumatology Biologics Register (BSRBR) and the NICE guidelines for anti-TNF therapy. Musculoskeletal care. 2003;1(1):58-64.44. Hyrich KL, Watson KD, Isenberg DA, Symmons DP, Register BSRB. The British Society for Rheumatology Biologics Register: 6 years on. Rheumatology. 2008;47(10):1441-3.45. Zink A, Listing J, Kary S, Ramlau P, Stoyanova-Scholz M, Babinsky K, et al. Treatment continuation in patients receiving biological agents or conventional DMARD therapy. Annals of the rheumatic diseases. 2005;64(9):1274-9.46. Strangfeld A, Eveslage M, Schneider M, Bergerhausen HJ, Klopsch T, Zink A, et al. Treatment benefit or survival of the fittest: what drives the time-dependent decrease in serious infection rates under TNF inhibition and what does this imply for the individual patient? Annals of the rheumatic diseases. 2011;70(11):1914-20.47. Mariette X, Matucci-Cerinic M, Pavelka K, Taylor P, van Vollenhoven R, Heatley R, et al. Malignancies associated with tumour necrosis factor inhibitors in registries and prospective observational studies: a systematic review and meta-analysis. Annals of the rheumatic diseases. 2011;70(11):1895-904.48. Brown SL, Greene MH, Gershon SK, Edwards ET, Braun MM. Tumor necrosis factor antagonist therapy and lymphoma development: twenty-six cases reported to the Food and Drug Administration. Arthritis and rheumatism. 2002;46(12):3151-8.49. Geborek P, Bladstrom A, Turesson C, Gulfe A, Petersson IF, Saxne T, 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. Annals of the rheumatic diseases. 2005;64(5):699-703.50. Wolfe F, Michaud K. The effect of methotrexate and anti-tumor necrosis factor therapy on the risk of lymphoma in rheumatoid arthritis in 19,562 patients during 89,710 person-years of observation. Arthritis and rheumatism. 2007;56(5):1433-9.51. Mariette X, Tubach F, Bagheri H, Bardet M, Berthelot JM, Gaudin P, et al. Lymphoma in patients treated with anti-TNF: results of the 3-year prospective French RATIO registry. Annals of the rheumatic diseases. 2010;69(2):400-8.52. Mercer LK, Galloway JB, Lunt M, Davies R, Low AL, Dixon WG, et al. Risk of lymphoma in patients exposed to antitumour necrosis factor therapy: results from the British Society for

Rheumatology Biologics Register for Rheumatoid Arthritis. Annals of the rheumatic diseases. 2017;76(3):497-503.53. Thomas E, Brewster DH, Black RJ, Macfarlane GJ. Risk of malignancy among patients with rheumatic conditions. International journal of cancer. 2000;88(3):497-502.54. Deepak P, Sifuentes H, Sherid M, Stobaugh D, Sadozai Y, Ehrenpreis ED. T-cell non-Hodgkin's lymphomas reported to the FDA AERS with tumor necrosis factor-alpha (TNF-alpha) inhibitors: results of the REFURBISH study. The American journal of gastroenterology. 2013;108(1):99-105.55. Askling J, Fored CM, Brandt L, Baecklund E, Bertilsson L, Feltelius N, et al. Risks of solid cancers in patients with rheumatoid arthritis and after treatment with tumour necrosis factor antagonists. Annals of the rheumatic diseases. 2005;64(10):1421-6.56. Haynes K, Beukelman T, Curtis JR, Newcomb C, Herrinton LJ, Graham DJ, et al. Tumor necrosis factor alpha inhibitor therapy and cancer risk in chronic immune-mediated diseases. Arthritis and rheumatism. 2013;65(1):48-58.57. Wadstrom H, Frisell T, Sparen P, Askling J, group As. Do RA or TNF inhibitors increase the risk of cervical neoplasia or of recurrence of previous neoplasia? A nationwide study from Sweden. Annals of the rheumatic diseases. 2016;75(7):1272-8.58. Wegener's Granulomatosis Etanercept Trial Research G. Etanercept plus standard therapy for Wegener's granulomatosis. The New England journal of medicine. 2005;352(4):351-61.59. SPC Enbrel. http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000262/WC500027361.pdf. Last accessed 31/10/2018.60. Amari W, Zeringue AL, McDonald JR, Caplan L, Eisen SA, Ranganathan P. Risk of non-melanoma skin cancer in a national cohort of veterans with rheumatoid arthritis. Rheumatology. 2011;50(8):1431-9.61. Askling J, Fahrbach K, Nordstrom B, Ross S, Schmid CH, Symmons D. Cancer risk with tumor necrosis factor alpha (TNF) inhibitors: meta-analysis of randomized controlled trials of adalimumab, etanercept, and infliximab using patient level data. Pharmacoepidemiology and drug safety. 2011;20(2):119-30.62. Raaschou P, Simard JF, Holmqvist M, Askling J, Group AS. Rheumatoid arthritis, anti-tumour necrosis factor therapy, and risk of malignant melanoma: nationwide population based prospective cohort study from Sweden. Bmj. 2013;346:f1939.63. Mercer LK, Green AC, Galloway JB, Davies R, Lunt M, Dixon WG, 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. Annals of the rheumatic diseases. 2012;71(6):869-74.64. Lienard D, Ewalenko P, Delmotte JJ, Renard N, Lejeune FJ. High-dose recombinant tumor necrosis factor alpha in combination with interferon gamma and melphalan in isolation perfusion of the limbs for melanoma and sarcoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 1992;10(1):52-60.65. Olsen CM, Green AC. Risk of invasive melanoma in patients with rheumatoid arthritis treated with biologics: an updated meta-analysis. Annals of the rheumatic diseases. 2018;77(8):e49.66. Franks AL, Slansky JE. Multiple associations between a broad spectrum of autoimmune diseases, chronic inflammatory diseases and cancer. Anticancer Res. 2012;32(4):1119-36.67. Nyboe Andersen N, Pasternak B, Basit S, Andersson M, Svanstrom H, Caspersen S, et al. Association between tumor necrosis factor-alpha antagonists and risk of cancer in patients with inflammatory bowel disease. JAMA. 2014;311(23):2406-13.68. Chiesa Fuxench ZC, Shin DB, Ogdie Beatty A, Gelfand JM. The Risk of Cancer in Patients With Psoriasis: A Population-Based Cohort Study in the Health Improvement Network. JAMA Dermatol. 2016;152(3):282-90.69. Hellgren K, Dreyer L, Arkema EV, Glintborg B, Jacobsson LT, Kristensen LE, et al. Cancer risk in patients with spondyloarthritis treated with TNF inhibitors: a collaborative study from the ARTIS and DANBIO registers. Ann Rheum Dis. 2017;76(1):105-11.

70. Maloney DG, Press OW. Newer treatments for non-Hodgkin's lymphoma: monoclonal antibodies. Oncology. 1998;12(10 Suppl 8):63-76.71. van Vollenhoven RF, Fleischmann RM, Furst DE, Lacey S, Lehane PB. Longterm Safety of Rituximab: Final Report of the Rheumatoid Arthritis Global Clinical Trial Program over 11 Years. The Journal of rheumatology. 2015;42(10):1761-6.72. Aaltonen KJ, Joensuu JT, Virkki L, Sokka T, Aronen P, Relas H, et al. Rates of serious infections and malignancies among patients with rheumatoid arthritis receiving either tumor necrosis factor inhibitor or rituximab therapy. The Journal of rheumatology. 2015;42(3):372-8.73. Slimani S, Lukas C, Combe B, Morel J. Rituximab in rheumatoid arthritis and the risk of malignancies: report from a French cohort. Joint, bone, spine : revue du rhumatisme. 2011;78(5):484-7.74. Wadstrom H, Frisell T, Askling J, Anti-Rheumatic Therapy in Sweden Study G. Malignant Neoplasms in Patients With Rheumatoid Arthritis Treated With Tumor Necrosis Factor Inhibitors, Tocilizumab, Abatacept, or Rituximab in Clinical Practice: A Nationwide Cohort Study From Sweden. JAMA internal medicine. 2017;177(11):1605-12.75. Simon TA, Smitten AL, Franklin J, Askling J, Lacaille D, Wolfe F, et al. Malignancies in the rheumatoid arthritis abatacept clinical development programme: an epidemiological assessment. Annals of the rheumatic diseases. 2009;68(12):1819-26.76. Nishimoto N, Ito K, Takagi N. Safety and efficacy profiles of tocilizumab monotherapy in Japanese patients with rheumatoid arthritis: meta-analysis of six initial trials and five long-term extensions. Modern rheumatology / the Japan Rheumatism Association. 2010;20(3):222-32.77. Campbell L, Chen C, Bhagat SS, Parker RA, Ostor AJ. Risk of adverse events including serious infections in rheumatoid arthritis patients treated with tocilizumab: a systematic literature review and meta-analysis of randomized controlled trials. Rheumatology. 2011;50(3):552-62.78. Schiff MH, Kremer JM, Jahreis A, Vernon E, Isaacs JD, van Vollenhoven RF. Integrated safety in tocilizumab clinical trials. Arthritis research & therapy. 2011;13(5):R141.79. Alten R, Kaine J, Keystone E, Nash P, Delaet I, Genovese MC. Long-term safety of subcutaneous abatacept in rheumatoid arthritis: integrated analysis of clinical trial data representing more than four years of treatment. Arthritis & rheumatology. 2014;66(8):1987-97.80. Ramiro S, Gaujoux-Viala C, Nam JL, Smolen JS, Buch M, Gossec L, et al. Safety of synthetic and biological DMARDs: a systematic literature review informing the 2013 update of the EULAR recommendations for management of rheumatoid arthritis. Annals of the rheumatic diseases. 2014;73(3):529-35.81. Westhovens R, Kremer JM, Emery P, Russell AS, Alten R, Barre E, et al. Long-term safety and efficacy of abatacept in patients with rheumatoid arthritis and an inadequate response to methotrexate: a 7-year extended study. Clinical and experimental rheumatology. 2014;32(4):553-62.82. Amano K, Matsubara T, Tanaka T, Inoue H, Iwahashi M, Kanamono T, et al. Long-term safety and efficacy of treatment with subcutaneous abatacept in Japanese patients with rheumatoid arthritis who are methotrexate inadequate responders. Modern rheumatology / the Japan Rheumatism Association. 2015;25(5):665-71.83. Curtis JR, Lee EB, Kaplan IV, Kwok K, Geier J, Benda B, et al. Tofacitinib, an oral Janus kinase inhibitor: analysis of malignancies across the rheumatoid arthritis clinical development programme. Annals of the rheumatic diseases. 2016;75(5):831-41.84. Cohen S, Curtis JR, DeMasi R, Chen Y, Fan H, Soonasra A, et al. Worldwide, 3-Year, Post-Marketing Surveillance Experience with Tofacitinib in Rheumatoid Arthritis. Rheumatology and therapy. 2018;5(1):283-91.85. Kunwar S, Collins CE, Constantinescu F. Baricitinib, a Janus kinase inhibitor, in the treatment of rheumatoid arthritis: a systematic literature review and meta-analysis of randomized controlled trials. Clinical rheumatology. 2018.86. Vineis P, Wild CP. Global cancer patterns: causes and prevention. Lancet. 2014;383(9916):549-57.

87. Silva-Fernandez L, Lunt M, Kearsley-Fleet L, Watson KD, Dixon WG, Symmons DP, et al. The incidence of cancer in patients with rheumatoid arthritis and a prior malignancy who receive TNF inhibitors or rituximab: results from the British Society for Rheumatology Biologics Register-Rheumatoid Arthritis. Rheumatology. 2016;55(11):2033-9.88. Raaschou P, Soderling J, Turesson C, Askling J, Group AS. Tumor Necrosis Factor Inhibitors and Cancer Recurrence in Swedish Patients With Rheumatoid Arthritis: A Nationwide Population-Based Cohort Study. Annals of internal medicine. 2018;169(5):291-9.89. Phillips C, Zeringue AL, McDonald JR, Eisen SA, Ranganathan P. Tumor Necrosis Factor Inhibition and Head and Neck Cancer Recurrence and Death in Rheumatoid Arthritis. PloS one. 2015;10(11):e0143286.90. Raaschou P, Frisell T, Askling J, Group AS. TNF inhibitor therapy and risk of breast cancer recurrence in patients with rheumatoid arthritis: a nationwide cohort study. Annals of the rheumatic diseases. 2015;74(12):2137-43.91. Mamtani R, Clark AS, Scott FI, Brensinger CM, Boursi B, Chen L, et al. Association Between Breast Cancer Recurrence and Immunosuppression in Rheumatoid Arthritis and Inflammatory Bowel Disease: A Cohort Study. Arthritis & rheumatology. 2016;68(10):2403-11.92. Lobo FS, Wagner S, Gross CR, Schommer JC. Addressing the issue of channeling bias in observational studies with propensity scores analysis. Research in social & administrative pharmacy : RSAP. 2006;2(1):143-51.93. Frisell T, Baecklund E, Bengtsson K, Di Giuseppe D, Forsblad-d'Elia H, Askling J, et al. Patient characteristics influence the choice of biological drug in RA, and will make non-TNFi biologics appear more harmful than TNFi biologics. Annals of the rheumatic diseases. 2018;77(5):650-7.94. Askling J, van Vollenhoven RF, Granath F, Raaschou P, Fored CM, Baecklund E, 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 and rheumatism. 2009;60(11):3180-9.95. Sant M, Allemani C, Tereanu C, De Angelis R, Capocaccia R, Visser O, et al. Incidence of hematologic malignancies in Europe by morphologic subtype: results of the HAEMACARE project. Blood. 2010;116(19):3724-34.96. Mercer LK, Regierer AC, Mariette X, Dixon WG, Baecklund E, Hellgren K, et al. Spectrum of lymphomas across different drug treatment groups in rheumatoid arthritis: a European registries collaborative project. Annals of the rheumatic diseases. 2017;76(12):2025-30.97. Ding T, Ledingham J, Luqmani R, Westlake S, Hyrich K, Lunt M, et al. BSR and BHPR rheumatoid arthritis guidelines on safety of anti-TNF therapies. Rheumatology. 2010;49(11):2217-9.98. Druce K, De Cock D, Watson K, Symmons DPM, Hyrich KL. Treatment Decisions Following Diagnosis of Cancer during TNFi Inhibitor Treatment in Patients with Rheumatoid Arthritis: Results from the BSRBR-RA [abstract]. Arthritis & rheumatology. 2016;68(Suppl 10).99. Ramiro S, Sepriano A, Chatzidionysiou K, Nam JL, Smolen JS, van der Heijde D, et al. Safety of synthetic and biological DMARDs: a systematic literature review informing the 2016 update of the EULAR recommendations for management of rheumatoid arthritis. Annals of the rheumatic diseases. 2017;76(6):1101-36.100. Scott FI, Mamtani R, Brensinger CM, Haynes K, Chiesa-Fuxench ZC, Zhang J, 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 dermatology. 2016;152(2):164-72.

Table 1: Risk of lymphoproliferative malignancies in patients with RA on TNFi or csDMARD in selected observational studies

Author Country Registry Year

Number of patients Relative Risk (95% CI)TNFi vs csDMARD as

referenceCancer outcome

TNFi csDMARD

Geborek et al. (49) Sweden SSATG 2005 757 800 4.90 (0.90-26.20) Lymphoma

Askling et al. (30) Sweden ARTIS 2005 4160 53067 1.10 (0.60 -2.10) Malignant lymphoma

Wolfe et al. (50) USA NDBRD 2007 10834 8757 1.00 (0.60–1.80) Lymphoma

Dreyer et al. (36) Denmark DANBIO 2013 5345 4351 0.63 (0.18 - 2.20) Non-Hodgkin lymphoma

Haynes et al. (56) USA SABER 2013 19750 9805 0.83 (0.33–2.05) Lymphoma

Wadstrom et al. (57) Swedish ARTIS 2016 10756 46358 0.85 (0.65-1.10)First invasive hematologic

malignant neoplasm

Mercer et al. (52) UK BSRBR-RA 2017 11931 3367 1.00 (0.56 - 1.80) Lymphoma

TNFi –Tumour Necrosis Factor inhibitor; csDMARD - classical synthetic disease modifying anti rheumatic drugs ; CI – confidence interval; USA – United States of America; UK- United Kingdom; SSATG - South Swedish Arthritis Treatment Group register ; ARTIS - Anti rheumatic Therapies in Sweden; NDBRD - National Data Bank for Rheumatic Diseases; DANBIO - The Danish Registry for Biologic Therapies in Rheumatology ; SABER - Safety Assessment of Biological Therapeutics study; BSRBR-RA – British Society Rheumatology Biologic Register Rheumatoid Arthritis

Table 2: Risk of Solid organ cancer excluding non-melanoma skin cancer in patients with RA on TNFi or csDMARD in selected observational studies

Author Country Registry YearNumber of patients Relative Risk (95% CI)

TNFi vs csDMARD as reference

TNFi csDMARD

Haynes et al. (56) USA SABER 2013 19750 9805 0.80 (0.59 – 1.08)

Mercer et al. (15) UK BSRBR-RA 2015 11767 3249 0.83 (0.64- 1.07).

Lan et al. (38) Taiwan Registry of Catastrophic Illness Database 2017 1111 16812 0.46 (0.27-0.79)

Wadstrom et al. (57) Sweden ARTIS 2016 10325 46358 0.94 (0.85-1.03)

TNFi–Tumour Necrosis Factor inhibitor; csDMARD - classical synthetic disease modifying anti rheumatic drugs; CI – confidence interval; USA – United States of America; UK- United Kingdom; SABER - Safety Assessment of Biological Therapeutics study; BSRBR-RA- British Society Rheumatology Biologic Register Rheumatoid Arthritis ; ARTIS - Anti rheumatic Therapies in Sweden

Table 3: Risk of skin cancer in patients with RA on TNFi or csDMARD in selected observational studies

Author Country Registry YearNumber of patients Relative Risk (95% CI)

TNFi vs csDMARD as referenceTNFi csDMARD

Non-melanoma skin cancer

Amari et al. (60) USADepartment of Veterans’

Affairs (VA) national administrative databases

2011 4088 18396 1.42 (1.24-1.63)

Mercer et al. (63) UK BSRBR-RA 2012 11 881 3629 0.95 (0.53 -1.71).

Dreyer et al. (36) Denmark DANBIO 2013 5345 4351 1.76 (1.26 - 2.46)

Haynes et al. (56) USA SABER 2013 19750 9805 0.83 (0.49–1.42)

Wadstrom et al. (74)* Sweden ARTIS 2017 10760 46358 1.24 (0.95 – 1.62)

Melanoma skin cancer

Rauschou et al. (62) Sweden ARTIS 2013 10878 42198 1.50 (1.00 – 2.20)

Dreyer et al. (36) Denmark DANBIO 2013 5345 4351 1.00 (0.32 - 3.11)

Wadstrom et al. (74) Sweden ARTIS 2017 10744 46358 0.85 (0.60 – 1.18)

Mercer et al. (7) Europe Multiple European registers 2017 68 411 48 304 1.14 (0.80- 1.60)

*Limited to squamous cell skin cancer only.

TNFi–Tumour Necrosis Factor inhibitor; csDMARD – classical synthetic disease modifying anti rheumatic drugs; CI – confidence interval; USA – United States of America; UK- United Kingdom; BSRBR-RA - British Society Rheumatology Biologic Register Rheumatoid Arthritis; DANBIO - The Danish Registry for Biologic Therapies in Rheumatology ; SABER - Safety Assessment of Biological Therapeutics study ; ARTIS - Anti rheumatic Therapies in Sweden

Table 4: Risk of cancer recurrence in patients with RA on TNFi or csDMARD in selected observational studies

Author Country Registry Year

Number of patients Relative Risk (95% CI)

TNFi vs csDMARD as reference

Prior Cancer site Cancer outcomeTNFi csDMARD

Strangfeld et al. (34) Germany RABBIT 2010 67 55 1.40 (0.50-5.50) Any prior cancer except NMSC Recurrence of cancer

Silva-Fernandez et al. (87) UK BSRBR-RA 2016 243 159 0.55 (0.35-0.86) Any prior cancer except NMSC

Any new cancer diagosis (recurrence

or new primary) except NMSC

Rauschou et al. (88) Sweden ARTIS 2018 467 2164 1.06 (0.73- 1.54) Prior solid organ cancer (selected list) Recurrence of cancer

Scott et al. (100) USA Medicare claim database 2016 109 335 1.49 (1.03–2.16) NMSC NMSC

Mercer et al. (63) UK BSRBR-RA 2012 177 106 1.18 (0.60-2.32) NMSC NMSC

Rauschou et al. (90) Sweden ARTIS 2015 120 120 1.10 (0.40- 2.80). Breast cancer Breast cancer

Mamtani et al. (91) USA Medicare claim database 2016 17 48 1.13 (0.65–1.97) Breast cancer Breast cancer

Philips et al. (89) USA

Department of Veterans’ Affairs administrative

database

2015 14 88 0.75 ( 0.31–1.85) Head and Neck Cancer

Head and Neck Cancer

TNFi–Tumour Necrosis Factor inhibitor; csDMARD - classical synthetic disease modifying anti rheumatic drugs ; CI – confidence interval; USA – United States of America; UK- United Kingdom; RABBIT - Rheumatoid Arthritis: Observation of Biologic Therapy; BSRBR-RA - British Society Rheumatology Biologic Register Rheumatoid Arthritis ; ARTIS – Anti rheumatic Therapies in Sweden; NMSC- Non-melanoma skin cancer