To the Editor, Recently, I reported, in Modern Rheumatology [1], on the characterization of the primary lack of efficacy of infliximab therapy and the efficacy of switching to tocilizumab in patients with rheumatoid arthritis (RA). In that study, nine of the 24 patients (38%) who had started induction therapy with infliximab were found to be primary nonresponders. Based on their trough concentrations of infliximab, the primary nonresponders could be classified into two groups: rapid-clearance type (six cases) and residual type (three cases). Both types of nonresponders benefited from a switch to tocilizumab therapy; eight of the nine patients (89%) were successfully treated with tocilizumab. One patient in the rapid-clearance group (Table 1, case 1), however, failed to achieve a 20% improvement according to the American College of Rheumatology criteria (ACR20) at the end of 24 weeks of tocilizumab therapy. To date, no patients who responded well to tocilizumab therapy have relapsed. Since the publication of this earlier study, it has been pointed out by several rheumatologists that our reported success rate of tocilizumab therapy seemed to be higher than that based on their experience in daily practice. To address this concern, I present the clinical characteristics and therapeutic responses to tocilizumab of primary infliximab-nonresponders who have been newly identified since the last study. I also show the efficacy of the additional use of tacrolimus with tocilizumab therapy. Following the protocol of the previous study, I conducted induction therapy with infliximab (intravenous infusions of 3 mg/kg) in 12 patients. During the infliximab therapy, methotrexate (MTX) (4–8 mg/week) and folic acid (5 mg/week) were given concomitantly. One patient, case 5, was receiving low-dose prednisolone (5 mg/day) at the start of the infliximab therapy and continued to receive this medication concomitantly throughout the period of infliximab therapy. Four of the 12 patients (33%) were characterized by a primary lack of efficacy of infliximab through the induction therapy (Table 1, cases 2–5). In one of these patients (case 2), the trough serum concentration of infliximab at week 14 was below the therapeutic level (1 lg/ml) (rapid-clearance type). The other three patients (cases 3–5) maintained high enough serum infliximab levels to produce therapeutic effects (residual type). Case 2 received a dose escalation (addition of one vial), but he nevertheless showed undetectably low trough concentrations (less than 0.1 lg/ml). All primary nonresponders were switched from infliximab to tocilizumab therapy (intravenous infusion of 8 mg/kg every 4 weeks for 24 weeks). During this therapy, all patients except for case 3 continued to receive MTX (8 mg/week) and folic acid (5 mg/week) concomitantly. MTX was discontinued in case 3 because of hepatic toxicity. Case 5 continued to receive low-dose prednisolone. As shown in Table 2, two patients (cases 3 and 4) in the residual-type group achieved an ACR70 improvement and a good response as defined by the European League Against Rheumatism (EULAR) 24 weeks after making this switch. However, the other two patients (case 2, rapid-clearance type; case 5, residual type) were refractory to tocilizumab therapy; these patients S. Mori (&) Department of Rheumatology, Clinical Research Center for Rheumatic Disease, NHO Kumamoto Saishunsou National Hospital, 2659 Suya, Kohshi, Kumamoto 861-1196, Japan e-mail: moris@saisyunsou1.hosp.go.jp