Background Evidence on cost-effectiveness is lacking for the comparison of biological treatments after failure of the first TNFi treatment in patients with Rheumatoid Arthritis (RA).
Objectives The objective of this study was to compare the cost-effectiveness of three biological treatment options, abatacept, rituximab or a second TNFi, after failure of the first TNFi treatment in patients suffering from RA.
Methods The inclusion criteria for this pragmatic randomized controlled trial were: failure of the first TNFi, a DAS28 >3.2, not treated before with abatacept or rituximab and no contraindications for these medications. The utility was based on the EQ5D questionnaire and was used to calculate QALYs. All medication costs in one year were taken into account. Incremental cost-utility evaluation was preformed to analyze the cost-effectiveness over one year corrected for possible confounders. Uncertainty in the cost-utility ratio was analyzed using the non-parametrical bootstrap technique. Cost-effectiveness was presented by Cost-Effectiveness Acceptability Curves (CEAC).
Results Of 144 randomized patients, 8 did not start the treatment due to infections, withdrawn from the study or waive the medication. 136 started in one of the three treatment arms: 42 on abatacept (mean age=56.2 yrs, female=88.1%, median disease duration=7.1 yrs, rheumatoid factor (RF) positive=51.4%, mean baseline DAS28=4.7), 44 on rituximab (mean age=56.7 yrs, female=63.6%, median disease duration=7.6 yrs, RF positive=79.1% mean baseline DAS28=4.9) and 50 on a second TNFi (mean age=56.2 yrs, female=74.0%, median disease duration=5.6 yrs, RF positive=55.8% mean baseline DAS28=4.9). Gender and RF were significantly different between the groups (p=0.032 and p=0.020 respectively). The mean QALY and costs in euro's were respectively 0.58 (sd=0.17) and €15.024 (sd=€11353) for patients in the abatacept group; 0.63 (sd=0.17) and €9385 (sd=€3814) for patients in the rituximab group and 0.61 (sd=0.18) and €11041 (sd=€7923) for the TNF alpha blocker group after 12 month. The corrected difference in costs between abatacept and rituximab was significant (β=4723; CI=892-8623; p=0.040). The corrected differences in costs between TNFi and abatacept and rituximab were not significant (β=-3326; CI=-8686-1031; p=0.200 and β=1397; CI=-1873-4333; p=0.394, respectively). If the Willingness To Pay is €80000, the chance rituximab is cost-effective is 85% compared to another TNFi and 98% compared to abatacept, see Figure.
Conclusions In this multi-centered pragmatic randomized controlled study in patients which their first TNFi failed, it was shown that after one year follow-up, rituximab is more cost-effective than abatacept and another TNFi.
Disclosure of Interest None declared