Successful cessation of tumor necrosis factor inhibitor treatment in rheumatoid arthritis patients and potential predictors for early flare: An observational study in routine clinical care

Abstract

Objectives: To investigate the prevalence and the consequence of tumor necrosis factor inhibitor (TNFi) cessation after clinical improvement in rheumatoid arthritis (RA) patients in clinical practice and predictors of flare after TNFi cessation.

Methods: We retrospectively assessed the prevalence of TNFi cessation after achieving sustained improvement, disease flare and joint damage progression after TNFi cessation in consecutive RA patients who started TNFi due to insufficient response to methotrexate were studied. Predictors for flare after TNFi cessation were investigated using Cox regression analysis.

Results: In 135 patients who started TNFi with methotrexate, 95 stopped TNFi after sustained improvement and continued methotrexate thereafter. Over 1 year, 33 patients had a flare and 26 restarted TNFi therapy. In 78 patients whose radiographs adequate for evaluation were available, 73 did not exhibit joint damage progression. Female gender, smoking, the interval from starting methotrexate to starting TNFi of more than 9 months and glucocorticoid use at starting TNFi were independently associated with shorter time to flare.

Conclusion: Sixty-five percent of patients were successfully discontinued TNFi over 1 year. Radiographic joint damage progression was rare. Early intervention with TNF inhibitor may contribute to successful TNF inhibitor cessation in patients with insufficient response to methotrexate.

Key message

• Successful TNF inhibitor cessation is achievable in two-third of RA patients after achieving sustained remission.

• Female gender and smoking may predispose to flare after TNF inhibitor cessation.

• Early intervention with TNF inhibitor may contribute to successful TNF inhibitor cessation.

Keywords:

• Biological therapy
• clinical remission
• recurrence
• withdrawing treatment

Acknowledgments

The authors would like to thank Masayoshi Iida and Koji Shimoda, medical information system engineers at Nagoya City University Hospital, who supported to develop clinical template for arthritis patients and extracted patient data from electric medical record in Nagoya City University Hospital, Yumiko Matsuura and Naomi Niimi, medical technologists in Takeuchi Orthopedics & Internal Medicine, who collected patients' data, Drs Shinji Maeda, Shinya Tamechika, Shin-ichiro Ohmura, Tohru Yamabe, and Koji Uehara, Division of Rheumatology, Department of Internal Medicine. Nagoya City University Hospital, who cared for the patients, and Prof. Akio Niimi, Department of Respiratory Medicine, Allergy and Clinical Immunology, Nagoya City University Graduate School of Medical Sciences, who provided general support as a department chair.

Conflict of interest

T. Naniwa has received research grants and lecture fees (<$5000) from Eisai Co., Ltd., Mitsubishi Tanabe Pharma Corporation, Pfizer Co. Ltd, and lecture fees (<$5000) from AbbVie GK, Astellas Pharma Corporation, Janssen Pharmaceutical K.K., and UCB Japan Co. Ltd. The other authors declare no conflicts of interest.