Dear Editor,

We read with interest the recent article by Pavelka and colleagues [1] and the accompanying editorial by van Vollenhoven [2] regarding dose escalation of infliximab in the treatment of patients with rheumatoid arthritis (RA). Pavelka and colleagues evaluated infliximab dose escalation in patients who had initially responded (based on the criteria of an improvement of 1.2 in the 28-joint count Disease Activity Score [DAS28]) but who failed to achieve DAS28 remission (DAS28 score of < 2.6) after 12 months. Patients were randomly assigned to either ontinue receiving 3 mg/kg or increase their dose to 5 mg/kg. The results showed that the mean change from baseline to 1 year in the DAS28 score was not significantly different between the treatment groups.

We agree with the editorial that this study is an important addition to the medical literature, as it is the first controlled, randomized, double-blinded study evaluating infliximab dose escalation conducted after our double-blinded, prospective, dose-escalation study, the START trial.[3] However, we cannot fully agree with the conclusions in this article and the accompanying editorial, which suggest that the study provides definitive proof that infliximab dose escalation is not useful under any circumstances. Although some of the study limitations were discussed by Pavelka and colleagues as well as by van Vollenhoven (e.g., additional dose adjustment above 5 mg/kg was not permitted), the following important limitations were not discussed.

First, to be eligible to participate in the study, patients must have received infliximab 3 mg/kg for at least 12 months; whereas, current EULAR treatment guidelines suggest that patients should be evaluated every 3 months and have their treatment adjusted whenever clinical response is inadequate,[4] which generally reflects actual clinical practice. Thus, the study entry criteria may have excluded patients who would have been more likely to benefit from dose escalation, such as those who switched to another therapy or received a dose increase before 12 months. This left fewer patients in the study population who might have benefitted from a dose increase. In addition, the baseline mean swollen joint count was 4.5, which may be too low to show a difference between two dosages that differ by only 2 mg/kg. No information on baseline median swollen joint count and tender joint count was provided, which is more meaningful in a population that is not normally distributed.

Second, the primary study endpoint was the mean change in DAS28 score, which may not be the most sensitive efficacy measure for evaluating dose escalation given that the majority of RA patients do not need infliximab dose escalation. In the START trial, only 30% of patients met the criteria for dose escalation, and approximately half of these patients needed an infliximab dose higher than 5 mg/kg to achieve a response.[3]
Evaluating “the proportion of patients achieving a specific level of response” such as the EULAR response, low disease activity, or DAS28 remission would have been a more sensitive endpoint in this situation. A recently published, randomized, double-blind study (RISING Study) showed that a significantly greater proportion of nonresponders who received dose escalation achieved EULAR responses than those who did not receive dose escalation.[5] Although the number of nonresponders who received dose escalation in the RISING study was smaller than that of the study reported by Pavelka and colleagues, the study design and evaluation period in the RISING study were more consistent with the EULAR treatment recommendations and clinical practice.

Third, the article by Pavelka and colleagues did not report the C-reactive protein (CRP) concentration at baseline; however, it is evident from Figure 3 of their article that the CRP concentration at study period week 0 was substantially lower in the 5-mg/kg group than in the 3-mg/kg group. We have shown in our START trial that all 7 nonresponders to infliximab dose escalation had low CRP concentrations at baseline.[3]
Thus it is possible that the 5-mg/kg group had more patients who were less likely to respond to infliximab dose escalation than the 3-mg/kg group.
Furthermore, 18 patients (25%) in the 3-mg/kg group were noncompliant with the study treatment compared with 4 patients (6%) in the 5-mg/kg group. The authors did not elaborate on the nature of this noncompliance. If 25% of patients received higher dose of infliximab (de facto dose escalation) or other treatment adjustments to improve clinical response, then the decrease in CRP level and DAS28 score in the 3-mg/kg group could be attributed to this noncompliance, depending on how data from these noncompliant patients were handled in the analysis. It would be interesting to know how the outcome would have been affected if an efficacy measure that does not include CRP (e.g., CDAI) had been used to evaluate an appropriately selected study population. The actual infliximab dose received in the 3-mg/kg group ranged from 2.7 to 5.5 mg/kg and from 3.3 to 6.3 mg/kg in the 5-mg/kg group. The difference in response rates between the two groups could easily be eliminated if some patients in the 3-mg/kg group who would have responded to dose escalation actually received infliximab 5.5 mg/kg instead and if some patients in the 5-mg/kg group who would have responded to dose escalation, only received 3.3 mg/kg of infliximab and consequently did not respond.

Fourth, radiographic progression was not evaluated by Pavelka and colleagues. Results from both the ATTRACT [6] and the RISING [5] studies have indicated that higher doses of infliximab inhibit radiographic progression more effectively than lower doses.

We applaud the efforts of Professor van Vollenhoven to educate rheumatologists on the phenomenon of regression to the mean, which is inevitable given the waxing and waning nature of RA disease activity.
Nonetheless, if infliximab dose escalations were unnecessary for all patients, there would be no reason to expect a relationship between the need for dose escalation and infliximab trough serum concentrations. The results of the START study showed that some patients who required dose escalation had inadequate serum concentrations of infliximab (lower than that of patients who did not require dose escalation), and these patients responded after their serum infliximab concentrations were increased by dose escalation.[3] This is further illustrated in Figure 1A-C herein, which shows that the mean trough serum infliximab concentrations for patients who received dose escalations in START were below those of patients who did not require dose escalation at any time (Figure 1A, 1B, and 1C). Trough serum concentrations for patients who required multiple dose escalations reached those of patients who did not require dose escalation at the final dose increase after which response was achieved (except nonresponders; Figure 1D). A relationship between serum infliximab concentrations and response has also been demonstrated in the ATTRACT study [6] and most recently in the RISING study.[5] Despite the limitations of these studies, their findings indicate that some patients receiving 3 mg/kg every 8 weeks have an insufficient infliximab serum concentration and may be more likely to respond to a higher dose or a shorter dosing interval.

Some patients, however, do not respond to an increased dose. In the START trial, 7 patients met the criteria for double-blinded dose escalation but did not respond even after 4 dose escalations and despite having adequate serum infliximab concentrations (Figure 1D). These patients had normal CRP concentrations at baseline, suggesting that they did not have active inflammation. Durez and colleagues [7] also showed that patients requiring dose escalation had higher levels of CRP and higher joint counts at baseline. We do not advocate the measurement of serum infliximab concentrations to determine the need for dose escalation due to high individual variability.[6] However, as indicated by the studies mentioned above,[3, 6] a careful clinical evaluation may allow for the identification of patients who would benefit from dose escalation.

Finally, Pavelka and colleagues reported an increased rate of nonserious adverse events in the 5-mg group compared with the 3-mg group.
The clinical significance of these events is unclear. The rates of serious adverse events and serious infections did not differ between treatment groups. This is consistent with the results of the START trial in which patients with dose escalation did not have increased rates of serious adverse events or serious infections.[3] Since dose escalation in selected patients only served to increase the serum trough concentrations to values comparable to those in patients who do not require dose escalation, we postulated that the risks of serious infection and serious adverse events would not be increased in patients who received dose escalation.

The results from the study reported by Pavelka and colleagues are important but are not the final verdict on the issue of infliximab dose escalation. Rather than abandon dose flexibility altogether, efforts must be made to distinguish patients who would benefit from dose escalation from those who would not benefit.
Rheumatologists must be cognizant of the regression to the mean phenomenon, but clearly some patients require dose escalation, and physicians should be given the flexibility to optimize treatment for these patients. Academic clinicians, practicing physicians, and the pharmaceutical industry must collaborate to further elucidate the optimal use of infliximab and other treatments for RA.

Figure

Figure 1: Mean (± standard deviation) trough serum infliximab concentrations for patients in the START study who were eligible to receive dose escalations. Serum levels before and after the last dose escalation, when patients achieved a clinical response, are shown. Patients who required dose escalation and responded to dose escalation had lower mean serum concentrations than those who did not require dose escalation.

References

1. Pavelka K, Jarosova K, Suchy D, et al. Increasing the infliximab dose in rheumatoid arthritis patients: a randomised, double blind study failed to confirm its efficacy. Ann Rheum Dis 2009;68:1285-9.

2. van Vollenhoven RF. How to dose infliximab in rheumatoid arthritis: new data on a serious issue. Ann Rheum Dis 2009;68:1237-9.

3. Rahman MU, Strusberg I, Geusens P, et al. Double-blinded infliximab dose escalation in patients with rheumatoid arthritis. Ann Rheum Dis 2007;66:1233-8.

4. Combe B, Landewé R, Lukas C, et al. EULAR recommendations for the management of early arthritis: report of a task force of the European Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis 2007;66:34-45.

5. Takeuchi T, Miyasaka N, Inoue K, et al. Impact of trough serum level on radiographic and clinical response to infliximab plus methotrexate in patients with rheumatoid arthritis: results from the RISING study. Mod Rheumatol 2009 Jul 22 [Epub ahead of print].

6. St Clair EW, Wagner CL, Fasanmade AA, et al. The relationship of serum infliximab concentrations to clinical improvement in rheumatoid arthritis: results from ATTRACT, a multicenter, randomized, double-blind, placebo-controlled trial. Arthritis Rheum 2002;46:1451-9.

7. Durez P, Van den Bosch F, Corluy L, Veys et al. A dose adjustment in patients with rheumatoid arthritis not optimally responding to a standard dose of infliximab of 3 mg/kg every 8 weeks can be effective: a Belgian prospective study. Rheumatology (Oxford) 2005;44:465-8.