Abstract
Objectives
To compare the risk of new primary cancer, locoregional progression or metastasis across different disease-modifying antirheumatic drugs (DMARDs) strategies in real-world patients with rheumatoid arthritis (RA) and prior cancer.
Methods
The ARthritis and CAncer registry is a nationwide retrospective study, including patients with RA and prior cancer. Collected variables included RA characteristics, DMARDs received, cancer type and demographics. The primary outcome was the occurrence of new primary cancer, locoregional progression or metastasis during follow-up. Adjusted Cox regression models were used to estimate HRs for cancer events. Treatment strategies involving conventional synthetic (cs)DMARDs served as the reference for comparison.
Results
Among 728 patients with RA, 80 cases of cancer were identified over a cumulative exposure period of 11 022 years, with an overall cancer incidence rate of 7.3 (95% CI 5.8 to 9.0) per 1000 person-years. Adjusted HRs for all cancers (95% CI) compared with csDMARDs were 0.94 (0.73 to 1.21) for tumour necrosis factor inhibitors, 1.03 (0.61 to 1.74) for IL6-inhibitors, 0.81 (0.46 to 1.42) for CD20 inhibitors, 0.6 (0.26 to 1.4) for targeted synthetic DMARDs and 0.74 (0.41 to 1.34) for cytotoxic T-lymphocyte-associated protein 4-analogue. No differences in cancer risk were observed after adjusting for duration of exposure, line of advanced therapy or prior cancer status.
Conclusion
This real-world data did not identify differences in the risk of new primary cancer, locoregional progression or metastasis across different DMARDs strategies in patients with RA and prior cancer.
Keywords: Rheumatoid Arthritis, Biological Therapy, Arthritis
WHAT IS ALREADY KNOWN ON THIS TOPIC
The use of advanced therapies in patients with rheumatoid arthritis (RA) and a history of cancer is challenging, as both disease activity and advanced therapies may potentially increase cancer risk.
Evidence on cancer risk with advanced therapies in this population remains very limited, particularly for non-tumour necrosis factor inhibitor biologic disease-modifying antirheumatic drugs and targeted synthetic disease-modifying antirheumatic drugs, and no real-world registry has been specifically designed for this purpose.
WHAT THIS STUDY ADDS
The ARthritis and CAncer is the first multicentre observational study specifically designed to assess cancer risk in patients with RA and prior cancer.
Our real-world data did not identify differences in the risk of new primary cancer, locoregional progression or metastasis across different disease-modifying antirheumatic drugs strategies in patients with RA and prior cancer.
No differences in cancer risk were identified according to duration of exposure, line of advanced therapy or cancer status.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our real-world study contributes evidence on the use of advanced therapies in patients with RA and a history of cancer, which may help inform therapeutic decision-making.
Introduction
The use of advanced therapies, including biologic (b) and targeted synthetic (ts) disease-modifying antirheumatic drugs (DMARDs), has expanded for the treatment of rheumatoid arthritis (RA) due to their effectiveness, safety profile and the endorsement of international recommendations advocating for a treat-to-target strategy.1 On the other hand, improvements in cancer treatment over the past 50 years have led to increased disease survival,2 and new treatments have enabled a paradigm shift, redefining cancer as a chronic disease. Therefore, the number of patients with RA and a history of previous or ongoing cancer is increasing in routine care. Due to this, there are concerns in clinical practice regarding the use of advanced therapies in patients with RA and a history of cancer, as both disease activity and DMARDs could potentially increase the risk of cancer.3 This may lead to undertreatment of patients with RA and poor control of disease activity.
Patients with a history of cancer are generally excluded from clinical trials, so the safety of treatment in this specific population comes mainly from observational registries. Most studies have not found an increased risk of new cancer, recurrence or progression in patients with RA and previous cancer associated with the use of bDMARDs, particularly tumour necrosis factor inhibitors (TNFi) and CD20 inhibitors (CD20i).4–16 The use of other non-TNFi bDMARDs, such as IL6-inhibitors (IL6i) or cytotoxic T-lymphocyte-associated protein 4-analogue (CTLA4-A) (abatacept), and tsDMARDs has been poorly studied.10 16 Based on a systematic literature review,17 the European Alliance of Associations for Rheumatology (EULAR) has very recently published points to consider on the use of advanced therapies in patients with inflammatory arthritis and a history of cancer.18 Preference is given to bDMARDs targeting cytokines such as TNFi and IL6i in patients with solid organ cancer, and to CD20i in patients with a history of lymphoma. On the other hand, caution is recommended in the use of tsDMARDs and CTLA4-A. Several studies have reported an increased cancer risk associated with the use of CTLA4-A in patients without a history of prior cancer.19–23 The ORAL Surveillance trial reported a slightly increased risk of cancer with tofacitinib compared with TNFi in patients with RA enriched for certain cardiovascular risk factors.24 However, most of these recommendations are based on expert opinion due to the limited evidence regarding cancer risk in this subset of patients. To date, all available evidence on the risk of cancer associated with advanced therapies in patients with prior cancer comes from secondary analyses of large observational registries, with the major limitation being the lack of detailed information on the characteristics of prior or incident cancers. Given that patients with RA frequently receive sequential or combination DMARD therapies over time, observational studies primarily compare treatment strategies as implemented in routine clinical practice rather than the isolated effects of individual drug classes. No registry has been specifically designed to assess the risk of cancer in patients with RA and a history of cancer. Additionally, the research agenda of the EULAR points to consider includes the study of cancer risk associated with other non-TNF-advanced therapies (especially with CTLA4-A and tsDMARDs), by cancer type, as well as their safety in patients with cancer who are not in remission.
Our primary objective was to compare the risk of new primary cancer, locoregional progression or metastasis across different DMARD strategies in patients with RA with a history of cancer. As secondary objectives, we examined cancer risk by duration of exposure, treatment line and prior cancer status.
Methods
Study design
The ARthritis and CAncer (ARCA) is the first nationwide observational multicentre retrospective study primarily designed to assess cancer-related safety of DMARDs in patients with arthritis and a history of cancer. This project was promoted by the Rheumatology Society of the Community of Madrid (SORCOM), which launched an open national call for centres to participate in the registry. Phase I of the registry includes patients with RA. Phase II of the registry is currently in progress, and it includes patients with spondyloarthritis and psoriatic arthritis.
Population
For the purpose of ARCA phase I, we included all patients with RA treated with b/tsDMARDs or conventional synthetic (cs)DMARDs and a history of cancer during a 1-year recruitment period (April 2024–April 2025). Patients eligible for inclusion in the study were identified at each centre either through review of medical health records or through pre-existing databases of patients with RA. To be eligible for inclusion in the study, all of the following inclusion criteria had to be met: (a) patients aged over 18 years; (b) fulfilment of the 2010 ACR/EULAR classification criteria for RA25; (c) diagnosis of cancer, as determined by the corresponding cancer care specialist (oncologist, dermatologist, haematologist, etc), before or after the RA onset and (d) initiation, re-initiation or switch of any DMARD after the cancer diagnosis. Exclusion criteria included a diagnosis of another inflammatory arthritis or autoimmune disease (eg, spondyloarthritis, psoriatic arthritis, systemic lupus erythematosus, etc). Patients with in situ carcinoma, non-melanoma skin cancer or indolent cancers (eg, cervical intraepithelial lesions) were excluded from the registry, as these are generally considered benign lesions with a low recurrence risk. Patients were classified in the following treatment groups if they had received the treatment at any point after cancer diagnosis: csDMARDs (methotrexate, sulfasalazine, leflunomide or hydroxychloroquine), TNFi (infliximab, adalimumab, etanercept, golimumab and certolizumab pegol), IL6i (tocilizumab and sarilumab), CD20i (rituximab), tsDMARDs (tofacitinib, baricitinib, upadacitinib and filgotinib) and CTLA4-A (abatacept). Thus, each patient could potentially contribute one observation to each group, and each patient could be represented in parallel in different exposure groups.
Data collection
All variables were collected retrospectively from the electronic health records. Two online informational meetings were held prior to the start of the study in order to standardise data collection across centres. Data collection was conducted through a Research Electronic Data Capture platform, a secure web application designed to support data capture for research studies.26 The variables included sex, date of birth, date of RA diagnosis, comorbidities, each type of treatment and start/discontinuation dates and disease activity at diagnosis, previous and subsequent cancer type and treatment. Previous cancers were further classified by type as follows: melanoma, haematological cancer (lymphoma or leukaemia) or solid cancer (including lung cancer, breast cancer and other solid tumours).
Primary outcomes
The primary outcome was the occurrence of new primary cancer, locoregional progression or metastasis during follow-up, occurring after the initiation, re-initiation or switch of any csDMARD or b/tsDMARD.
Treatment exposure
Given the long latency periods of cancer, events were attributed to both current and past treatments using an ‘ever-exposed’ drug model.27 Since patients could receive several treatments over time, each recurrent cancer event could be associated with one or more treatment groups. Because treatment exposure changed over time and concomitant csDMARD use was common, comparisons should be interpreted as reflecting routine therapeutic strategies rather than mutually exclusive treatment groups.
Statistical analysis
Baseline demographic, clinical and prior cancer characteristics were summarised using absolute and relative frequencies for categorical variables. Continuous variables were expressed as means with SD or medians with lower and upper quartiles (Q1, Q3). Analyses were performed on a complete-case basis without data imputation. The study size was determined by the total number of eligible patients available in the registry during the study period rather than formal a priori power calculations.
Crude incidence rates (IRs) of cancer were calculated as the total number of events divided by the total person-years (PY) at risk and expressed per 1000 PY. The 95% CIs for the IRs were estimated using the exact Poisson method. For survival analyses, time at risk was defined from the index date (date of first exposure) until the occurrence of cancer. Patients who did not experience the event of interest were right-censored at the date of death, loss to follow-up or the end of the study observation period, whichever occurred first.
The probability of cancer over time was estimated and plotted using Kaplan-Meier survival curves. To assess the relative risk of cancer associated with different treatments, multivariable Cox proportional hazards regression models were constructed, with csDMARDs group as the reference category. To systematically select potential confounders for adjustment, a Least Absolute Shrinkage and Selection Operator (LASSO) regression approach with cross-validation was employed. To perform pure LASSO, the alpha parameter was set to 1 and the optimal penalty parameter (lambda) was determined based on the minimum cross-validation error. Prespecified clinically relevant variables, including treatment group, gender and age at disease onset, were unpenalised to guarantee their retention in the model. Covariates yielding non-zero coefficients in the optimised LASSO regression were subsequently incorporated into the final fully adjusted Cox models to estimate HRs and their corresponding 95% CIs. The proportional hazards assumption was verified for all variables using scaled Schoenfeld residuals. The results of the multivariable analyses were visually represented using forest plots. To evaluate the robustness of our findings, sensitivity analyses were conducted by subsetting the full data set according to specific clinical characteristics. First, patients with no prior exposure to advanced therapies vs >1 previous advanced therapy; second, duration of exposure (<2 years or >2 years) and third, previous cancer status (remission vs active). Within each stratum, the Cox proportional hazards models were refitted using the exact covariate structure determined by the primary LASSO regression. Subgroup-specific event counts, PY, crude IRs per 1000 PY (95% CI) and both unadjusted and fully adjusted HRs (95% CI) were reported to assess the consistency of the treatment effect. To account for the non-independence of observations from the same patient across treatment groups, all Cox models were fitted with a robust sandwich variance estimator.
All statistical tests were two-sided and a p value of <0.05 was considered statistically significant. All statistical analyses were performed using R software (R Core Team, 2025; R Foundation for Statistical Computing, Vienna, Austria).
Results
Patients
During the recruitment period, 728 patients with RA and a history of cancer from 24 centres were identified, contributing a total of 1481 treatment initiations. The demographic data, RA characteristics and comorbidities are shown in table 1. The mean (SD) age was 54 (15) years at disease onset, 505 (70%) were women. The distribution of treatment initiations by type of therapy was as follows: 725 (99.6%) csDMARDs, 313 (43%) TNFi, 120 (16.5%) IL6i, 132 (18.1%) CD20i, 94 (12.9%) tsDMARDs and 97 (13.3%) CTLA4-A. The mean (SD) disease duration at the time of inclusion in the registry was 16 (11) years, and the median (Q1, Q3) follow-up since cancer diagnosis was 14 (8, 22) years (table 1).
Table 1. Demographic, RA characteristics, comorbidities and prior cancer characteristics of the patients included in the ARCA registry.
| Total treatments N=1481 |
Total patients N=728 | ||||||
|---|---|---|---|---|---|---|---|
| csDMARD (n=725) (99.6%) | TNFi (n=313) (43%) |
IL6i (n=120) (16.5%) | CD20i (n=132) (18.1%) | tsDMARDs (n=94) (12.9%) | CTLA4-A (n=97) (13.3%) | ||
| Demographics | |||||||
| Female, n (%) | 503 (70%) | 234 (75%) | 96 (81%) | 94 (71%) | 76 (81%) | 69 (71%) | 505 (70%) |
| Age at disease onset (years), mean (SD) | 54 (15) | 48 (14) | 49 (13) | 50 (14) | 49 (13) | 53 (13) | 54 (15) |
| Age at cancer diagnosis (years), mean (SD) | 63 (36) | 63 (53) | 67 (84) | 68 (80) | 59 (11) | 73 (94) | 63 (36) |
| RA characteristics | |||||||
| Disease duration at the time of inclusion (years), mean (SD) | 16 (11) | 20 (11) | 19 (11) | 19 (10) | 19 (11) | 19 (10) | 16 (11) |
| Follow-up time since cancer diagnosis (years), mean (SD)/median (Q1, Q3) | 16 (11)/14 (8, 23) | 20 (11)/19 (11, 26) | 19 (11)/16 (11, 25) | 19 (10)/18 (12, 25) | 19 (11)/16 (10, 25) | 19 (10)/17 (11, 24) | 16 (11)/14 (8, 22) |
| Rheumatoid factor positive, n (%) | 554 (77%) | 242 (78%) | 91 (76%) | 113 (86%) | 74 (79%) | 79 (81%) | 556 (77%) |
| ACPA positive, n (%) | 537 (75%) | 232 (76%) | 94 (80%) | 114 (88%) | 72 (79%) | 80 (84%) | 539 (75%) |
| DAS28-CRP at cancer diagnosis, mean (SD)/median (Q1, Q3) | 2.61 (1.32)/2.28 (1.65, 3.28) | 2.68 (1.31)/2.40 (1.65, 3.28) | 2.53 (1.26)/2.25 (1.68, 3.27) | 2.72 (1.19)/2.44 (1.97, 3.43) | 2.79 (1.44)/2.49 (1.80, 3.63) | 2.57 (1.25)/2.20 (1.77, 2.98) | 2.61 (1.33)/2.28 (1.65, 3.28) |
| DAS28-ESR at cancer diagnosis, mean (SD)/median (Q1, Q3) | 2.87 (1.39)/2.67 (1.93, 3.67) | 2.95 (1.34)/2.67 (2.00, 3.82) | 2.58 (1.33)/2.40 (1.61, 3.46) | 3 (1.32)/2.67 (1.96, 3.84) | 2.98 (1.36)/2.79 (2.06, 3.95) | 2.85 (1.28)/2.55 (1.82, 3.46) | 2.88 (1.39)/2.67 (1.95, 3.67) |
| Comorbidities | |||||||
| Obesity, n (%) | 141 (22%) | 59 (21%) | 31 (29%) | 25 (23%) | 20 (24%) | 22 (24%) | 141 (22%) |
| Smoker, n (%) | 98 (14%) | 41 (13%) | 13 (11%) | 15 (11%) | 7 (7.4%) | 12 (12%) | 98 (13%) |
| Former smoker, n (%) | 260 (36%) | 99 (32%) | 40 (33%) | 50 (38%) | 27 (29%) | 34 (35%) | 262 (36%) |
| Hypertension, n (%) | 341 (47%) | 141 (45%) | 51 (43%) | 56 (42%) | 46 (49%) | 50 (52%) | 342 (47%) |
| Hypercholesterolaemia, n (%) | 300 (41%) | 128 (41%) | 48 (40%) | 53 (40%) | 38 (40%) | 48 (49%) | 303 (42%) |
| Diabetes, n (%) | 118 (16%) | 43 (14%) | 16 (13%) | 24 (18%) | 11 (12%) | 18 (19%) | 118 (16%) |
| Coronary artery disease, n (%) | 48 (6.6%) | 24 (7.7%) | 10 (8.3%) | 14 (11%) | 2 (2.1%) | 12 (12%) | 48 (6.6%) |
| Stroke, n (%) | 34 (4.7%) | 18 (5.8%) | 10 (8.3%) | 4 (3.0%) | 5 (5.3%) | 10 (10%) | 34 (4.7%) |
| Interstitial lung disease, n (%) | 51 (7.0%) | 17 (5.4%) | 11 (9.2%) | 18 (14%) | 4 (4.3%) | 22 (23%) | 52 (7.1%) |
| Previous bDMARDs/tsDMARDs | |||||||
| No exposure, n (%) | 701 (97%) | 280 (89%) | 33 (28%) | 52 (39%) | 24 (26%) | 38 (39%) | 533 (73%) |
| >1, n (%) | 24 (3.3%) | 33 (11%) | 87 (73%) | 80 (61%) | 70 (74%) | 59 (61%) | 195 (27%) |
| Previous cancer characteristics | |||||||
| Solid, n (%) | 648 (89%) | 285 (91%) | 109 (91%) | 110 (83%) | 85 (90%) | 87 (90%) | 651 (89%) |
| Breast cancer, n (%) | 196 (27%) | 94 (30%) | 41 (34%) | 36 (27%) | 32 (34%) | 28 (29%) | 196 (27%) |
| Lung cancer, n (%) | 86 (12%) | 40 (13%) | 19 (16%) | 12 (9.1%) | 13 (14%) | 16 (16%) | 87 (12%) |
| Colorectal cancer, n (%) | 77 (11%) | 21 (6.7%) | 8 (6.7%) | 11 (8.3%) | 9 (9.6%) | 8 (8.2%) | 78 (11%) |
| Other, n (%) | 289 (40%) | 130 (42%) | 41 (34%) | 51 (39%) | 31 (33%) | 35 (36%) | 290 (40%) |
| Haematological, n (%) | 48 (6.6%) | 18 (5.8%) | 9 (7.5%) | 18 (14%) | 6 (6.4%) | 8 (8.2%) | 48 (6.6%) |
| Lymphoma, n (%) | 36 (5.0%) | 15 (4.8%) | 6 (5.0%) | 15 (11%) | 4 (4.3%) | 5 (5.2%) | 36 (5.0%) |
| Leukaemia, n (%) | 12 (1.7%) | 3 (1.0%) | 3 (2.5%) | 3 (2.3%) | 2 (2.1%) | 3 (3.1%) | 12 (1.7%) |
| Melanoma, n (%) | 27 (3.7%) | 9 (2.9%) | 2 (1.7%) | 3 (2.3%) | 3 (3.2%) | 2 (2.1%) | 27 (3.7%) |
ACPA, anticitrullinated protein antibody; ARCA, ARthritis and CAncer; bDMARDs, biologic disease-modifying antirheumatic drugs; CD20i, CD20 inhibitors; CRP, C reactive protein; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; CTLA4-A, cytotoxic T-lymphocyte-associated protein 4 analogue; DAS28, Disease Activity Score in 28 joints; ESR, erythrocyte sedimentation rate; IL6i, IL6-inhibitors; RA, rheumatoid arthritis; TNFi, tumour necrosis factor inhibitors; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs.
Previous cancer characteristics
All patients included in the ARCA registry had a diagnosis of cancer: 651 (89%) had solid tumours, 48 (6.6%) had haematological malignancies and 27 (3.7%) had melanoma. The mean (SD) age at cancer diagnosis was 63 (36) years. The type and frequency of prior cancer according to the therapy received are shown in table 1. Breast cancer was the most common prior cancer (27%), followed by lung cancer (12%) and colorectal cancer (11%). The frequency of lymphoma was numerically higher among patients who received CD20i compared with the other therapies. The frequency of other cancer types was similar across all exposure groups (table 1).
Rates and relative risk of all recurrent cancers
A total of 80 cases of cancer events were detected during the exposure period (sum=11 022 years), with an overall cancer IR of 7.3 (95% CI 5.8 to 9.0) per 1000 PY. 47 (59%) patients had a new primary cancer, 25 (31%) patients experienced progression of previous cancer and eight (10%) patients had metastases of the prior cancer. Figure 1 presents a Kaplan-Meier survival curve depicting the probability of all cancers associated with b/tsDMARD use. Table 2 shows the number of events, crude IRs, the non-adjusted and fully adjusted HRs for all cancer events and specific cancer subtypes. Crude IRs per 1000 PY for all cancers were: 7.0 for csDMARDs, 5.4 for TNFi, 5.0 for IL6i, 5.7 for CD20i, 3.5 for tsDMARDs and 5.1 for CTLA4-A. Based on 77 cancer cases in the csDMARD group, 33 in TNFi, 11 in IL6i, 14 in CD20i, six in tsDMARDs and nine in CTLA-4, the fully adjusted HR (95% CI) versus csDMARDs was 0.94 (0.73 to 1.21) for TNFi, 1.03 (0.61 to 1.74) for IL6i, 0.81 (0.46 to 1.42) for CD20i, 0.60 (0.26 to 1.40) for tsDMARDs and 0.74 (0.41 to 1.34) for CTLA4-A (table 2). Older age at disease onset (HR 1.06; 95% CI 1.04 to 1.09) was associated with increased cancer risk. Figure 2 presents the results of the adjusted Cox proportional hazards regression analysis.
Figure 1. Kaplan-Meier survival curve depicting the probability of cancer associated with b/tsDMARD use. bDMARDs, biologic disease-modifying antirheumatic drugs; CD20i, CD20 inhibitors; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; CTLA4-A, cytotoxic T-lymphocyte-associated protein 4 analogue; IL6i, IL-6 inhibitors; TNFi, tumour necrosis factor inhibitors; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs.
Table 2. Number of events, PY, IR and HRs for all cancer events, solid cancers, melanoma and haematological cancers according to type of therapy.
| Events | PY | Crude IR per 1000 PY (95% CI) | HR (95% CI) | Fully adjusted HR* (95% CI) | |
|---|---|---|---|---|---|
| All cancers | |||||
| csDMARD | 77 | 11 009.2 | 7.0 (5.5, 8.7) | 1.0 (reference) | 1.0 (reference) |
| TNFi | 33 | 6073.5 | 5.4 (3.7, 7.6) | 0.73 (0.57, 0.94) | 0.94 (0.73, 1.21) |
| IL6i | 11 | 2180.6 | 5.0 (2.5, 9.0) | 0.73 (0.43, 1.24) | 1.03 (0.61, 1.74) |
| CD20i | 14 | 2474.4 | 5.7 (3.1, 9.5) | 0.79 (0.50, 1.26) | 0.81 (0.46, 1.42) |
| tsDMARDs | 6 | 1720.9 | 3.5 (1.3, 7.6) | 0.51 (0.24, 1.08) | 0.60 (0.26, 1.40) |
| CTLA4-A | 9 | 1749.4 | 5.1 (2.4, 9.8) | 0.72 (0.40, 1.30) | 0.74 (0.41, 1.34) |
| Solid cancer | |||||
| csDMARDs | 63 | 11 009.2 | 5.7 (4.4, 7.3) | 1.0 (reference) | 1.0 (reference) |
| TNFi | 26 | 6073.5 | 4.3 (2.8, 6.3) | 0.71 (0.53, 0.94) | 1.00 (0.76, 1.32) |
| IL6i | 9 | 2180.6 | 4.1 (1.9, 7.8) | 0.72 (0.40, 1.30) | 1.07 (0.60, 1.90) |
| CD20i | 10 | 2474.4 | 4.0 (1.9, 7.4) | 0.68 (0.39, 1.21) | 0.70 (0.35, 1.40) |
| tsDMARDs | 4 | 1720.9 | 2.3 (0.6, 6.0) | 0.41 (0.16, 1.07) | 0.66 (0.25, 1.73) |
| CTLA4-A | 7 | 1749.4 | 4.0 (1.6, 8.2) | 0.68 (0.35, 1.33) | 0.68 (0.35, 1.34) |
| Haematological cancer | |||||
| csDMARDs | 8 | 11 009.2 | 0.7 (0.3, 1.4) | 1.0 (reference) | — |
| TNFi | 4 | 6073.5 | 0.7 (0.2, 1.7) | 0.79 (0.41, 1.52) | — |
| IL6i | 2 | 2180.6 | 0.9 (0.1, 3.3) | 1.31 (0.39, 4.43) | — |
| CD20i | 3 | 2474.4 | 1.2 (0.3, 3.5) | 1.74 (0.75, 4.05) | — |
| tsDMARDs | 2 | 1720.9 | 1.2 (0.1, 4.2) | 1.59 (0.51, 5.01) | — |
| CTLA4-A | 1 | 1749.4 | 0.6 (0.0, 3.2) | 0.77 (0.13, 4.53) | — |
| Melanoma | |||||
| csDMARDs | 3 | 11 009.2 | 0.3 (0.1, 0.8) | 1.0 (reference) | — |
| TNFi | 2 | 6073.5 | 0.3 (0.0, 1.2) | 1.26 (0.56, 2.83) | — |
| IL6i | 0 | 2180.6 | 0.0 (0.0, 1.7) | — | — |
| CD20i | 1 | 2474.4 | 0.4 (0.0, 2.3) | 1.45 (0.28, 7.42) | — |
| tsDMARDs | 0 | 1720.9 | 0.0 (0.0, 2.1) | — | — |
| CTLA4-A | 1 | 1749.4 | 0.6 (0.0, 3.2) | 2.01 (0.40, 10.15) | — |
Fully adjusted HR using Cox regression model using LASSO. Adjusted for age at disease onset, gender, smoking, obesity.
CD20i, CD20 inhibitors; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; CTLA4-A, cytotoxic T-lymphocyte-associated protein 4 analogue; IL6i, IL-6 inhibitors; IR, incidence rate; LASSO, Least Absolute Shrinkage and Selection Operator; PY, person years; TNFi, tumour necrosis factor inhibitors; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs.
Figure 2. Adjusted Cox proportional hazards regression analysis on the risk of (A) overall cancer and (B) solid cancer. CD20i, CD20 inhibitors; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; CTLA4-A, cytotoxic T-lymphocyte-associated protein 4 analogue; IL6i, IL-6 inhibitors; TNFi, tumour necrosis factor inhibitors; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs.
Subanalysis of the risk of new primary cancer
Online supplemental table 1 shows the number of events, crude IRs, the non-adjusted and fully adjusted HRs for new primary cancer events.
Subanalysis in patients with prior solid cancer
The analyses in patients with previous solid cancer did not identify statistically significant differences in overall cancer risk across different DMARD strategies (online supplemental table 2 and online supplemental figure 1).
Subanalysis by line of b/tsDMARD, duration of exposure and remission of prior cancer
When stratifying patients by line of b/tsDMARD (none or more than one previous advanced therapy), the fully adjusted HR for all cancers did not differ significantly compared with the main analysis (online supplemental figure 2). Considering the duration of exposure (<2 years or >2 years), the fully adjusted HR for all cancers was similar to the main analysis (online supplemental figure 3). Finally, after performing a sensitivity analysis according to cancer status (remission or active), the results were similar to the main analysis (online supplemental figure 4).
Discussion
In this large nationwide real-world study, we did not identify differences in the risk of new primary cancer, locoregional progression or metastasis across different DMARD strategies in patients with RA and prior cancer. In addition, no differences in cancer risk were identified in sensitivity analyses stratified by duration of exposure, line of therapy or cancer status. However, our findings should be interpreted as comparisons between treatment strategies used in routine clinical practice rather than causal estimates of the independent effects of individual DMARD classes.
To date, no increased risk of cancer has been demonstrated in association with the use of advanced therapies in patients with RA and a history of prior malignancy.17 18 28 29 The majority of studies report safety data in patients with RA treated with TNFi and CD20i,4–16 whereas the safety of other non-TNFi bDMARDs or tsDMARDs has been poorly investigated.10 16 In a recent EULAR systematic literature review and meta-analysis including 4428 patients with inflammatory arthritis and a history of cancer, no increased risk of malignancy was identified with advanced therapies compared with csDMARDs (HR 0.92, 95% CI 0.74 to 1.15).17 However, all studies included in the EULAR review reported the risk for TNFi and CD20i, and no study included tsDMARDs. Our study also did not identify differences in cancer risk, but additionally provides safety data on other non-TNFi bDMARDs and tsDMARDs, for which evidence is very limited.10 16 Furthermore, none of the studies included in the EULAR review involved patients whose cancer was not in remission, a situation where there is no safety evidence for the use of advanced therapies. In our subanalysis stratified by cancer status (remission or active), we did not detect differences in cancer risk. These results may help inform therapeutic decisions until new safety data with larger sample sizes become available.
The safety of TNFi use in patients with prior malignancy is generally well established. Several large observational studies have similarly not identified an increased risk of cancer recurrence among TNFi-treated patients with prior malignancy.5 6 16
After TNFi, CD20i are the second most studied therapy in patients with prior malignancy. Their use has even been favoured in international recommendations, considering their effectiveness in the treatment of haematological neoplasms.30 31 Previous observational studies have also not identified an increased risk of cancer recurrence with CD20i in patients with prior malignancy.13 16 32 Our findings are consistent with these data and we did not identify differences in overall or solid cancer risk associated with CD20i use.
Safety data on other non-TNFi bDMARDs and tsDMARDs are very limited. Westerman et al, in 720 patients with RA with prior malignancy, did not identify an increased risk of cancer associated with the use of any bDMARD (HR 0.92, 95% CI 0.38 to 1.73) compared with csDMARDs; however, they did not provide therapy-specific data for non-TNFi treatments, and their study included only patients with cancer in remission. In the Spanish BIOBADASER registry, including 352 patients with inflammatory arthritis with prior malignancy, we did not identify an increased risk of cancer in patients exposed to tsDMARDs (HR 0.5, 95% CI 0.2 to 1.7), CD20i (HR 0.4, 95% CI 0.1 to 1), IL6i (HR 1.1, 95% CI 0.5 to 2.4), CTLA4-A (HR 1.5, 95% CI 0.7 to 3.1) or IL17i (HR 0.7, 95% CI 0.2 to 2.4) compared with TNFi therapy.10 In the ARCA registry, we did not identify differences in cancer risk associated with the use of other non-TNFi bDMARDs, such as IL6i or CTLA4-A, in either the overall analysis or the sensitivity analyses. However, CTLA4-A has been reported to be associated with an increased risk of cancer versus other advanced therapies in patients without a history of malignancy.19 22 33 EULAR recommends against the use of CTLA4-A when other alternatives are available,18 but this recommendation is based on studies in patients without a history of malignancy.
Finally, EULAR recommends caution regarding the use of tsDMARDs in patients with a history of malignancy, and suggests using them only in the absence of therapeutic alternatives.18 This recommendation is based on the results from the ORAL Surveillance trial, a randomised controlled study comparing the occurrence of major cardiovascular events and cancers between tofacitinib and TNFi in patients with RA without a history of malignancy but enriched for certain key risk factors, which showed that tofacitinib was not non-inferior to TNFi with respect to cancer incidence.24 Beyond clinical trials, real-world data in patients without prior malignancy are limited, and most studies have not identified an increased risk of cancer associated with tsDMARDs use.34–37 However, Huss et al, in a study including 10 447 patients with RA, detected an increased risk of non-melanoma skin cancer particularly after 2 or more years since treatment initiation.34 Real-world safety data on tsDMARDs in patients with prior malignancy remain very limited.10 The ARCA registry contributes data from 94 tsDMARDs treatment initiations in patients with prior malignancy, and we did not detect differences in cancer risk, although the limited number of events warrants cautious interpretation.
The main strength of the ARCA study is that it is the first registry specifically designed to evaluate the risk of cancer associated with advanced therapies in patients with RA with prior malignancy. Its multicentre real-world design enhances the generalisability of the results. Among the limitations are the retrospective design, with the potential for information bias or missing data due to death. Furthermore, given the observational nature of the study, we cannot exclude indication bias or unmeasured factors influencing the decision to initiate advanced therapy in patients with prior malignancy, such as better prognostic factors for cancer recurrence. Furthermore, we cannot exclude the possibility that certain covariates, such as obesity and smoking status, changed during follow-up, as these variables were only assessed at the time of RA diagnosis. Additionally, the limited sample size and the small number of certain events, such as haematological cancers or melanomas, did not allow for multivariable analyses in certain subgroups. Another limitation is the lack of a defined minimum exposure treatment duration necessary to attribute a particular cancer to that treatment; however, the heterogeneous carcinogenic processes underlying different cancer types complicate such an analysis. Our study includes a smaller number of more recently commercialised b/tsDMARDs compared with TNFi, although this reflects routine clinical practice. Finally, as the majority of patients received csDMARDs at some point during follow-up, the independent contribution of csDMARD exposure to cancer risk cannot be fully isolated. Accordingly, the HRs reported should be interpreted as comparisons between treatment strategies as used in routine clinical practice, rather than as causal estimates of the effect of individual drug classes.
In summary, the ARCA study did not identify differences in the risk of new primary cancer, locoregional progression or metastasis across different DMARD strategies in patients with RA with prior malignancy. Considering the long latency period of rare adverse events such as cancer, very long-term pharmacovigilance studies are needed to confirm these findings.
Supplementary material
Acknowledgements
The authors would like to thank all the patients who participated in the study and the Rheumatology Society of Madrid for promoting the ARthritis and CAncer study and providing the Research Electronic Data Capture tool.
Footnotes
Funding: This study was partly supported by grants PI20/00847, PI23/01226 and RD24/0007/0029 from the Instituto de Salud Carlos III and co-financed by the European Regional Development Fund. The funds were used exclusively for statistical analysis expenses.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study was approved by the Research Ethics Committee of Hospital General Universitario Gregorio Marañón, Madrid, Spain, which acted as the central committee (PI 177/24 ARCA), and was conducted in accordance with Good Pharmacoepidemiology Practice standards and the principles of the Declaration of Helsinki. Written informed consent was deemed not mandatory for participation in the registry.
Collaborators: ARCA project collaborative group: Alejandro Escudero, Jerusalem Calvo Gutierrez, Lourdes Ladehesa Pineda, Desirée Ruiz Vilchez, Ignacio Gómez García, Mari Carmen Ábalos, Julia Uceda, Marta Ruíz-Valdepeñas Almansa, Cristina Lajas, María López, Luis Coronel, Alejandra López Robles, Teresa González and Ana Isabel Rebollo-Giménez.
Presented at: The results of this study were presented at the 52nd National Congress of the Spanish Society of Rheumatology.38
Contributor Information
on behalf of the ARCA project collaborative group:
Alejandro Escudero, Jerusalem Calvo Gutierrez, Lourdes Ladehesa Pineda, Desirée Ruiz Vilchez, Ignacio Gómez García, Mari Carmen Ábalos, Julia Uceda, Marta Ruíz-Valdepeñas Almansa, Cristina Lajas, María López, Luis Coronel, Alejandra López Robles, Teresa González, and Ana Isabel Rebollo-Giménez
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
References
- 1.Smolen JS, Landewé RBM, Bergstra SA, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2022 update. Ann Rheum Dis. 2023;82:3–18. doi: 10.1136/ard-2022-223356. [DOI] [PubMed] [Google Scholar]
- 2.Hemminki K, Försti A, Hemminki A. Survival in colon and rectal cancers in Finland and Sweden through 50 years. BMJ Open Gastroenterol. 2021;8:e000644. doi: 10.1136/bmjgast-2021-000644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Szekanecz Z, Gomez I, Soós B, et al. Eight pillars of oncorheumatology: crossroads between malignancies and musculoskeletal diseases. Autoimmun Rev. 2020;19:102658. doi: 10.1016/j.autrev.2020.102658. [DOI] [PubMed] [Google Scholar]
- 4.Raaschou P, Simard JF, Asker Hagelberg C, et al. 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. 2016;352:i262. doi: 10.1136/bmj.i262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Strangfeld A, Hierse F, Rau R, et al. Risk of incident or recurrent malignancies among patients with rheumatoid arthritis exposed to biologic therapy in the German biologics register RABBIT. Arthritis Res Ther. 2010;12:R5. doi: 10.1186/ar2904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Raaschou P, Söderling J, Turesson C, et al. Tumour necrosis factor inhibitors and cancer recurrence in Swedish patients with rheumatoid arthritis: a nationwide population-based cohort study. Ann Intern Med. 2018;169:291–9. doi: 10.7326/M17-2812. [DOI] [PubMed] [Google Scholar]
- 7.Raaschou P, Frisell T, Askling J, et al. TNF inhibitor therapy and risk of breast cancer recurrence in patients with rheumatoid arthritis: a nationwide cohort study. Ann Rheum Dis. 2015;74:2137–43. doi: 10.1136/annrheumdis-2014-205745. [DOI] [PubMed] [Google Scholar]
- 8.Dreyer L, Cordtz RL, Hansen IMJ, et al. Risk of second malignant neoplasm and mortality in patients with rheumatoid arthritis treated with biological DMARDs: a Danish population-based cohort study. Ann Rheum Dis. 2018;77:510–4. doi: 10.1136/annrheumdis-2017-212086. [DOI] [PubMed] [Google Scholar]
- 9.Waljee AK, Higgins PDR, Jensen CB, et al. Anti-tumour necrosis factor-α therapy and recurrent or new primary cancers in patients with inflammatory bowel disease, rheumatoid arthritis, or psoriasis and previous cancer in Denmark: a nationwide, population-based cohort study. Lancet Gastroenterol Hepatol. 2020;5:276–84. doi: 10.1016/S2468-1253(19)30362-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Molina-Collada J, Alonso F, Otero L, et al. Cancer risk with biologic and targeted synthetic DMARDs in patients with rheumatic diseases and previous malignancies: results from the BIOBADASER register. Semin Arthritis Rheum. 2024;64:152341. doi: 10.1016/j.semarthrit.2023.152341. [DOI] [PubMed] [Google Scholar]
- 11.Mamtani R, Clark AS, Scott FI, et al. Association between breast cancer recurrence and immunosuppression in rheumatoid arthritis and inflammatory bowel disease: a cohort study. Arthritis Rheumatol . 2016;68:2403–11. doi: 10.1002/art.39738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dixon WG, Watson KD, Lunt M, et al. 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. Arthritis Care & Research . 2010;62:755–63. doi: 10.1002/acr.20129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Silva-Fernández L, Lunt M, Kearsley-Fleet L, 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 (Oxford) 2016;55:2033–9. doi: 10.1093/rheumatology/kew314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Scott FI, Mamtani R, Brensinger CM, 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. 2016;152:164–72. doi: 10.1001/jamadermatol.2015.3029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mercer LK, Green AC, Galloway JB, 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. 2012;71:869–74. doi: 10.1136/annrheumdis-2011-200622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Westermann R, Cordtz R, Duch K, et al. Cancer recurrence risk with bDMARD treatment in patients with rheumatoid arthritis and a history of cancer: a nationwide Danish register-based cohort study. RMD Open. 2025;11:e005247. doi: 10.1136/rmdopen-2024-005247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sebbag E, Molina-Collada J, Ndoye R, et al. Systematic literature review and meta-analysis informing the EULAR points to consider on the initiation of targeted therapies in patients with inflammatory arthritis and a history of cancer. Ann Rheum Dis. 2024;S0003-4967:00002–0. doi: 10.1136/ard-2024-225981. [DOI] [PubMed] [Google Scholar]
- 17.Sebbag E, Lauper K, Molina-Collada J, et al. 2024 EULAR points to consider on the initiation of targeted therapies in patients with inflammatory arthritis and a history of cancer. Ann Rheum Dis. 2024 doi: 10.1136/ard-2024-225982. [DOI] [PubMed] [Google Scholar]
- 19.Simon TA, Boers M, Hochberg M, et al. Comparative risk of malignancies and infections in patients with rheumatoid arthritis initiating abatacept versus other biologics: a multi-database real-world study. Arthritis Res Ther. 2019;21:228. doi: 10.1186/s13075-019-1992-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.de Germay S, Bagheri H, Despas F, et al. Abatacept in rheumatoid arthritis and the risk of cancer: a world observational post-marketing study. Rheumatology (Oxford) 2020;59:2360–7. doi: 10.1093/rheumatology/kez604. [DOI] [PubMed] [Google Scholar]
- 21.Wadström H, Frisell T, Askling J, et al. Malignant neoplasms in patients with rheumatoid arthritis treated with tumour necrosis factor inhibitors, tocilizumab, abatacept, or rituximab in clinical practice: a nationwide cohort study from Sweden. JAMA Intern Med. 2017;177:1605–12. doi: 10.1001/jamainternmed.2017.4332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Montastruc F, Renoux C, Dell’Aniello S, et al. Abatacept initiation in rheumatoid arthritis and the risk of cancer: a population-based comparative cohort study. Rheumatology (Oxford) 2019;58:683–91. doi: 10.1093/rheumatology/key352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Huss V, Bower H, Wadström H, et al. Short- and longer-term cancer risks with biologic and targeted synthetic disease-modifying antirheumatic drugs as used against rheumatoid arthritis in clinical practice. Rheumatology (Oxford) 2022;61:1810–8. doi: 10.1093/rheumatology/keab570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ytterberg SR, Bhatt DL, Mikuls TR, et al. Cardiovascular and cancer risk with tofacitinib in rheumatoid arthritis. N Engl J Med. 2022;386:316–26. doi: 10.1056/NEJMoa2109927. [DOI] [PubMed] [Google Scholar]
- 25.Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62:2569–81. doi: 10.1002/art.27584. [DOI] [PubMed] [Google Scholar]
- 26.Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–81. doi: 10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dixon WG, Carmona L, Finckh A, et al. EULAR points to consider when establishing, analysing and reporting safety data of biologics registers in rheumatology. Ann Rheum Dis. 2010;69:1596–602. doi: 10.1136/ard.2009.125526. [DOI] [PubMed] [Google Scholar]
- 28.Shasti-Nazem B, Chandra S, Strouse J, et al. Treatment of persons with rheumatoid arthritis with a history of cancer. Curr Rheumatol Rep. 2025;27:34. doi: 10.1007/s11926-025-01200-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Xie W, Xiao S, Huang Y, et al. A meta-analysis of biologic therapies on risk of new or recurrent cancer in patients with rheumatoid arthritis and a prior malignancy. Rheumatology (Oxford) 2020;59:930–9. doi: 10.1093/rheumatology/kez475. [DOI] [PubMed] [Google Scholar]
- 30.Chatzidionysiou K, Delcoigne B, Frisell T, et al. How do we use biologics in rheumatoid arthritis patients with a history of malignancy? An assessment of treatment patterns using Scandinavian registers. RMD Open. 2020;6:e001363. doi: 10.1136/rmdopen-2020-001363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Singh JA, Saag KG, Bridges SL, Jr, et al. 2015 American College of Rheumatology guideline for the treatment of rheumatoid arthritis. Arthritis Care Res (Hoboken) 2016;68:1–25. doi: 10.1002/acr.22783. [DOI] [PubMed] [Google Scholar]
- 32.Risk Of Cancer Recurrence Or New Tumors In RA Patients With Prior Malignancies Treated With Various Biologic Agents ACR meeting abstracts. [15-Feb-2022]. https://acrabstracts.org/abstract/risk-of-cancer-recurrence-or-new-tumors-in-ra-patients-with-prior-malignancies-treated-with-various-biologic-agents/ Available. Accessed.
- 33.Ozen G, Pedro S, Schumacher R, et al. Safety of abatacept compared with other biologic and conventional synthetic disease-modifying antirheumatic drugs in patients with rheumatoid arthritis: data from an observational study. Arthritis Res Ther. 2019;21:141. doi: 10.1186/s13075-019-1921-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Huss V, Bower H, Hellgren K, et al. Cancer risks with JAKi and biological disease-modifying antirheumatic drugs in patients with rheumatoid arthritis or psoriatic arthritis: a national real-world cohort study. Ann Rheum Dis. 2023;82:911–9. doi: 10.1136/ard-2022-223636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Kremer JM, Bingham CO, 3rd, Cappelli LC, et al. Postapproval comparative safety study of tofacitinib and biological disease-modifying antirheumatic drugs: 5-year results from a United States-based rheumatoid arthritis registry. ACR Open Rheumatol . 2021;3:173–84. doi: 10.1002/acr2.11232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Khosrow-Khavar F, Desai RJ, Lee H, et al. Tofacitinib and risk of malignancy: results from the safety of tofacitinib in routine care patients with rheumatoid arthritis (STAR-RA) study. Arthritis Rheumatol . 2022;74:1648–59. doi: 10.1002/art.42250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Molina-Collada J, Otero-Varela L, Vela P, et al. Cancer risk in patients with rheumatoid arthritis receiving biologic and targeted synthetic disease-modifying antirheumatic drugs: results from the BIOBADASER III registry. Rheumatology (Oxford) 2026;65:keaf556. doi: 10.1093/rheumatology/keaf556. [DOI] [PubMed] [Google Scholar]
- 38.Molina-collada J. P017. RISK OF CANCER ASSOCIATED WITH BDMARDS AND TSDMARDS IN PATIENTS WITH RHEUMATOID ARTHRITIS AND A HISTORY OF CANCER: INSIGHTS FROM THE ARCA REGISTRY. Reumatol Clin. 2026;22 doi: 10.1136/rmdopen-2026-007113. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data relevant to the study are included in the article or uploaded as supplementary information.