Background/Purpose: Tofacitinib (Tofa) was recently approved in the U.S. There are no data comparing the benefit risk profile of Tofa to tumor necrosis factor inhibitors (TNFi). This analysis included a conjoint survey of rheumatology patients’ preferences, followed by a benefit-risk assessment combining multiple benefits and risks associated with Tofa vs. TNFi.

Methods: The patient preference study investigated 10 attributes, Patient cost, ACR20, Oral vs Injection vs Infusion, Injection site/infusion reactions, Nausea, Herpes Zoster, Malignancies, and Cholesterol elevation.  Attribute “levels” were assigned from pivotal trial data, meta-analyses, FDA documents and prescribing information.  Levels encompassed a plausible range of values. Each patient was asked to complete a preference selection for 15 choice tasks, based on a randomized design using a 5-attribute partial profile. A Multi-Criteria Decision Analysis (MCDA), a tool from the EMEA risk-benefit project, was used to assess the overall benefit-risk profile, incorporating differences in outcomes and using the conjoint survey to weight each attribute’s relative importance. The base case profile identified Tofa as having a less robust ACR20 response, higher incidence of malignancy and herpes zoster risk, greater impact on cholesterol, lower levels of injection site/infusion reactions and oral vs injection/infusion. 

Results: 386 patients had evaluable data from the conjoint preference study. 71% were female and median age was between 60 and 64.  Fifty-one patients (13%) were bio-naïve.  Attributes that showed the greatest impact on preference perception between top and bottom levels included out-of-pocket costs, injection site reactions, dose interval/mode and impact on cholesterol.  There was little difference between injection site reactions levels of none and 4/100 patient years (PY), as well as between dose interval/mode of oral at 730 doses per year and injections at 12 times per year.  For dose interval/mode, the least preferred option was injections 26-52 times per year.  Based on base case profiles assigned to individual products, 42% of patients would choose golimumab, 34% an oral DMARD and 24% infliximab.  If all Tofa risk attribute differences vs TNFi were removed, 53% would choose Tofa, 28% golimumab and 19% infliximab.  Etanercept and Adalimumab preferences were sensitive to rate of injection site reactions. The MCDA estimated a greater combined score for TNFi compared to Tofa, indicating that from a patient perspective TNFi are a more favorable treatment option overall. A major contributing factor to this outcome was the relative importance that patients placed on risk attributes which favored the TNFi profile.

Conclusion: Patients appear sensitive to relatively small changes in tolerability and safety profiles of rheumatology treatments.  Benefit and risk profiles can have significant mitigating influence over patient preferences related to convenience of treatment.  This reinforces the need for full and clear disclosure to patients of the benefit and risks of treatments in RA and the importance of patient involvement in treatment decisions, since lack of preference can have an impact on adherence.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/modeling-the-benefit-risk-profiles-of-a-new-janus-kinase-inhibitor-tofacitinib-compared-tottumour-necrosis-factor-inhibitor-biologic-treatments-incorporating-conjoint-derived-patient-preference-weight/