Higher TNFi Dosing Is Not Associated with More Serious Infectious Events (SIE), elevated AST/ALT or WBC<1.5 in the US CORRONA Database

Background/Purpose: There is a perception that higher TNFi doses result in increased adverse events (AEs) in selected AEs such as SIE, increased ALT/AST or WBC<1.5. We examined this perception in the USA CORRONA database.

Hypothesis: Higher doses of TNFi are associated with an increased incidence of SIE, increased ALT/AST and/or WBC<1.5.

Methods:

We examined the US CORRONA registry between Oct 1, 2001 and Feb 15, 2013. This registry is an independent, prospective, observational cohort of rheumatic disease patients (in this case rheumatoid arthritis (RA) patients (pts)) among 159 sites in the United States.

Inclusions Patients with a diagnosis of RA who started their first TNFi without prior biological use, with TNFi doses recorded and followed for ≥ 3 months.

High-dose TNFi was defined as: ≥ 5 mg/kg infliximab q8wks, > 50 mg etanercept wkly, 40 mg adalimumab wkly, >400 mg certolizumab q4wk or > 50 mg golim. q4wk for at least 3 months.

Adverse Events (AEs) were defined as: SIE (infection requiring hospitalization or IV antibiotic), AST+/-ALT > upper limit of normal or WBC<1.5.

Analysis: Descriptive statistics were done. A Cox proportional hazards model assessed the association of TNFi AEs of interest; if P≤0.2 in the univariate model, the variables were included in the multivariable model. For WBC, hazard ratio (HR) from the univariate Cox regression was used because there were too few instances for multivariate analysis.

Results: Of 4195 TNFi-starter RA patients naïve to other biologics, 2688 met inclusion criteria and constituted the test set. There were 388 high dose (14.4%), versus 2300 low dose (85.6%) pts.

The High dose group was older and generally had more active, severe disease (High versus Low P value): Age:

60 vs. 56 yrs; CDAI: 17.2 versus 14.5; erosion: 41.1 versus 18.8%; RF/CCP pos.: 80.1 versus 74.6%-all P <0.05-0.001.

Shorter F/U for the High dose group (35.8±27.9 vs. 39.7±30.0 mos; P= 0.02) was not clinically significant

(10%).

The incidence rates and dose effects for SIEs, ALT/AST and WBC<1.5 are shown in table 1.

*High dose n= 388   **Low dose n= 2300

Discussion: In this real-life cohort of RA pts, higher dose users had more active, severe disease. Since the high-dose group had lower BMI (P< 0.05), the High dose users were especially likely to be using higher milligrams/m2 doses, supporting the credibility of the results. As in all registries, however, some selection bias may have contributed to the lack of differences found in High vs. Low dose TNFi users.

Conclusion: High dose TNFi users with RA did not seem to be subject to more SIE, AST/ALT elevations or WBC<1.5 than Low dose users, within the limits of TNFi doses used within usual clinical practice in this real-life cohort of patients in the USA.