Incidence of cancer in patients with spondyloarthritis treated with anti-TNF drugs - 01/11/17

Doi : 10.1016/j.jbspin.2017.08.003

Fabiola Atzeni ^a,^⁎ , Antonio Carletto ^b, Rosario Foti ^c, Marco Sebastiani ^d, Valentina Panetta ^e, Fausto Salaffi ^f, Gianluca Bonitta ^e, Florenzo Iannone ^g, Elisa Gremese ^h, Marcello Govoni ⁱ, Antonio Marchesoni ^j, Ennio Giulio Favalli ^j, Roberto Gorla ^k, Roberta Ramonda ^l, Piercarlo Sarzi-Puttini ^a, Gianfranco Ferraccioli ^h, Giovanni Lapadula ^g
on behalf of the
GISEA group
^a Rheumatology unit, L. Sacco University Hospital, Via G.B. Grassi 74, 20127 Milano, Italy
^b Rheumatology unit, university of Verona, 37129 Verona, Italy
^c Rheumatology unit, Vittorio-Emanuele University Hospital of Catania, 95123 Catania, Italy
^d University hospital of Modena, 41121 Modena, Italy
^e L’altra Statistica Consultancy and Training, Biostatistics Office, 00101 Roma, Italy
^f Polytechnic University of Marche, C. Urbani Hospital, 60035 Jesi, Italy
^g University of Bari, 70121 Bari, Italy
^h Division of Rheumatology, Institute of Rheumatology, Catholic University of the Sacred Heart, 00168 Roma, Italy
ⁱ Department of Medical Sciences, UOC of Rheumatology, Santa Anna University Hospital, 44124 Ferrara, Italy
^j G. Pini Orthopedic Institute, 20122 Milan, Italy
^k Rheumatology and Immunology Unit, Spedali Civili, 25123 Brescia, Italy
^l University of Padua, 35122 Padova Italy

^⁎Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 01 November 2017
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Abstract

Objective

To determine the incidence of cancer in TNF inhibitor (TNFi)-treated spondyloarthritis (SpA) patients entered in the GISEA registry, and identify the factors associated with its development.

Methods

This observational study involved an open cohort of 3321 SpA patients selected from the GISEA registry, designed to collect real-world clinical data concerning patients with RA or SpA treated with biological drugs. The baseline information includes demographics and clinical parameters. The overall incidence of neoplasias was compared to this observed in the general population according to the Italian Association of Medical Oncology.

Results

Of the 3321 SpA patients (1731 males, 52.2%; mean age 47±13years; median disease duration three years, interquartile range [IQR] 0–8), 50 developed at least one of 56 malignancies during the follow-up period of up to 12years of treatment with TNFi. The overall incidence was 6.3/1000 patient-years of follow-up (95% confidence interval [CI] 4.7–8.2): 7.3/1000 patient-years (95% CI 4.1–11.8) in those treated with ADA; 6.1/1000 patient-years (95% CI 3.8–9.4) in those treated with ETN; and 5.8/1000 patient-years (95% CI 3.5–9.1) in those treated with INF while in the general population was 5.1/1000 patient-years. Univariate analysis showed that age at the time of starting TNFi (P=0.001), the presence of comorbidities (P=0.012), the number of comorbidities (P<0.001), and HAQ-DI score (P=0.002) were associated with a higher risk of malignancies. Stepwise regression models showed that only previous neoplasia was a significant predictor of a new malignancy. The type of drug was not associated with the risk of malignancy.

Conclusions

The incidence of malignancies among SpA patients treated with the three TNFi was higher than in general population; having had a previous solid cancer is predictive of a new malignancy.

Le texte complet de cet article est disponible en PDF.

Keywords : Spondyloarthritis, Neoplasias, Anti-TNF drugs, Predictive factors