Perspectives on CV safety: JAK inhibitor (Tofacitinib) and TNF inhibitors Save

Safety comparisons of JAKi (tofacitinib) vs. TNFi from Oral Surveillance showed that, in patients with active RA on MTX and aged >50 years with one or more CV risk factors, that there was numerically more MACE events, malignancies (and statistically significant malignancies with more events in combined tofacitinib groups), and VTE especially in higher dose. Data also showed more serious infections in age >65 years (0831 primary, 0958 MACE, 1940 malignancy, 1941 VTE and 1684 serious infections).

We have to realize that the FDA and Pfizer agreements on the Oral Surveillance design were not to show statistical difference (that is always a risk with a large N and an event driven trial – i.e. potentially long treatment duration in the RCT). The number needed to harm of an excess event of tofacitinib vs. TNFi added important context to the data. NNH was presented in many of the Oral Surveillance outcomes and hundreds in general needed to be treated for one year in order to have a serious adverse event in these high risk patients (such as MACE or malignancy); however, it is up to the patient and physician to determine if the small but serious excess risks are relevant.

Real world data may help frame the findings.

The STAR-RA study compared real world safety of TNFi vs. tofacitinib in several administrative databases with 102,263 patients of whom 9.5 to 13.2% were using tofacitinib and the rest TNFi (Khosrow-Khavar, F, ACR21 #1939). Overall there were no differences in CV outcomes and this was also found in the group enriched for CV events who mimicked the Oral Surveillance inclusion criteria. Drug use did not differentiate in an adjusted analysis. These data are very reassuring and may or may not help us to put the Oral Surveillance data in context.

Figure of STAR-RA in the lower panel shows the RCT matched cohort and does not show increased MACE with tofacitinib using a very large real world population.