2523 - RA, Anti-TNF Therapy, and Risk of Malignant Melanoma-a Nationwide Population-Based Study From Sweden

Tuesday, November 8, 2011: 4:30 PM
W375a (McCormick Place West)
Pauline Raaschou1, Julia F. Simard2, Martin Neovius2, Marie Holmqvist2, Jonas Eriksson2, Johan Askling2 and the ARTIS studygroup3, 1Clinical Pharmacology Unit, Dept of Medicine, Karolinska Institutet, Stockholm, Sweden, 2Clinical Epidemiology Unit, Dept of Medicine, Karolinska Institutet, Stockholm, Sweden, 3Karolinska Institutet
Presentation Number: 2523

Background/Purpose: States of immune suppression, such as post-transplant therapy, have recently emerged as a possible risk factor not only for non-melanoma skin cancer but also for malignant melanoma (Vajdic CM et al, JAMA 2006). An increased risk of melanoma associated with anti-TNF therapy has been reported (Wolfe F et al, Arthritis Rheum. 2007; Askling J. EULAR Conference Abstract 2009). We aimed to investigate the risk of malignant melanoma in patients with RA compared to the general population, and to investigate whether anti-TNF treatment influences melanoma risk in RA.

Method: A nationwide and population-based cohort of patients with RA were identified using data from the Swedish outpatient register from 2001 through 2009 (n=56,336). Patients starting anti-TNF therapy were identified through linkage to the Swedish Biologics Register ARTIS and to the national Prescribed Drug Register (n=8,453) To each RA patient, 5 general population comparators were matched for sex, year of birth and county of residence. Occurrence of first-ever invasive malignant melanoma, as well as occurrence of first-ever cancer, irrespective of type (all-site), was assessed through linkage to the national Swedish Cancer Register. Relative risks (RRs) for malignant melanoma and all-site cancer were calculated using Cox regression with age as time-scale and anti-TNF therapy treated as a time-dependent variable, adjusted for selected co-morbidities. RRs were assessed overall and by time since start of anti-TNF therapy.

Result: Based on 135 incident melanomas during 253,572 person-years of follow-up among biologics-naïve patients with RA vs. 718 melanomas during 1,460,120 person-years of follow-up in the general population comparator, the RR for malignant melanoma in biologics-naïve patients with RA was 1.1 (95% CI 0.9-1.3). Based on 32 incident malignant melanomas during 44,858 person-years of follow-up from start of anti-TNF therapy, compared to biologics-naïve RA patients, the RR was 1.8 (95% CI 1.2-2.7). Based on 418 incident all site cancers during 42,418 person-years of follow-up among the anti-TNF exposed, the RR was 1.0 (0.9-1.1) compared to biologics-naïve RA patients (Table). Sensitivity analyses also including in situ melanomas resulted in a relative risk of 1.5 (95% CI 1.1-2.1).

Conclusion: In the absence of anti-TNF therapies, RA patients are not at elevated risk of malignant melanoma. Patients selected for and treated with anti-TNF have a higher risk of malignant melanoma than biologics-naïve RA patients. 

Table. Relative risks adjusted for age, sex and co-morbidities (95% confidence intervals) and number of cases among patients treated with anti-TNF, comparing Swedish patients with RA treated with anti-TNF (n=8,453) to biologics-naïve patients with RA (n=47,883).


Anti-TNF-exposed vs. Biologics-naïve RA

Time since first anti-TNF treatment start



< 1 year

1-2 years

≥2 years

Melanoma   Relative risk (95%CI), Cases 

1.8 (1.2-2.7), 32

1.5 (0.6-4.1), 4

2.9 (1.4-5.9), 8

1.7 (1.0-2.7), 20

All-Site Cancer Relative risk (95%CI), Cases 

1.0 (0.9-1.1), 418

1.0 (0.8-1.3), 74

0.9 (0.7-1.2), 63

1.1 (1.0-1.3), 281

Keywords: Anti-TNF therapy, Cancer and rheumatoid arthritis (RA)

Disclosure: P. Raaschou, None; J. F. Simard, None; M. Neovius, None; M. Holmqvist, None; J. Eriksson, None; J. Askling, Pfizer Inc, Bristol Meyer Squibb, 5 .