At the 2021 ACR Convergence the primary results of ORAL-Surveillance were presented as well as an in-depth analysis of major adverse cardiovascular events (MACE), malignancy and venous thrombotic episodes (VTE). These results have eagerly been awaited by the rheumatology community since the top line results were presented in a press release earlier this year and the recent FDA advisory on what these results may mean for the label of tofacitinib (as well as other members of the Janus kinase family (JAKi) [1,2]. 

As background, it has been published that patients with rheumatoid arthritis (RA) have an increased risk of cardiovascular (CV) events and the risk correlates both with traditional CV risk factors and with markers of inflammation, such as the erythrocyte sedimentation rate [3-6]. Epidemiological studies suggest that several disease-modifying antirheumatic drugs (DMARD) treatments for RA are associated with a reduced risk of CV events, including methotrexate and tumor necrosis factor (TNFi) antagonists [7,8] with a cross-sectional association between atherosclerosis and disease activity [5,9]. It has also been demonstrated that reduction of disease activity over time in RA is associated with fewer CV events [10]. Compared to conventional synthetic disease modifying anti-rheumatic drugs (csDMARDs), mortality and cardiovascular events are significantly decreased in patients treated with TNFi as well as other biologic DMARDs (bDMARDs) without an increased risk of solid malignancies [11]. By reducing dramatically chronic inflammation in RA patients, TNFi decrease mortality, CV events without significantly increasing the risk of cancer, compared to csDMARDs [11]. Similar results have been reported with csDMARDs (methotrexate) and other bDMARDs. [12-14].

With respect to VTE, it has been demonstrated that patients with RA have a relative risk >2 for both deep vein thrombosis and pulmonary emboli [15,16] and the degree of risk is related to residual inflammation [17].

With respect to tofacitinib, neither the integrated safety data base nor a post marketing registry revealed a signal with respect to MACE, malignancy, or VTE that increased over time or differs from bDMARDs in the general RA population [18,19]. 

ORAL Surveillance was a post-marketing requirement of the FDA which assessed the relative risk of tofacitinib 5 and 10 mg BID versus TNFi for MACE and malignancies, excluding non-melanoma skin cancer, in patients with active RA, despite MTX who were at least age 50 with one or more cardiovascular risk factors. There was a prospective, non-inferiority safety trial. The reason the FDA required this study was that increases in lipids and malignancies were observed in the clinical trial program with tofacitinib treatment, especially in higher than the approved dose [20]. 

The trial was designed to assess the relative risk of MACE and malignancies in a select group of RA patients who were expected to be at increased risk of these events. Patients were aged ≥50 years and had moderate to severe RA despite MTX treatment and ≥1 additional CV risk factor (current cigarette smoker, hypertension, HDL-c <40 mg/dL, diabetes mellitus, family history of premature coronary heart disease, extra-articular disease associated with RA, history of coronary artery disease). The study was completed once at least 1,500 patients were followed for at least three years, and when at least 103 MACE and 138 malignancy events were observed.

The primary endpoint was whether the upper limit of the two-sided 95% CI for the HR for the combined tofacitinib doses (both 5 mg BID and 10 mg BID) versus TNFi was <1.8; a secondary endpoint was whether the upper limit of the two-sided 95% CI of the HR for tofacitinib 10 versus 5 mg BID was <2.0.

Incidence rates, defined as unique patients with events per 100 PY and number need to harm, defined as the number of PY of exposure needed for tofacitinib to have one more event relative to the TNFI were calculated.

More than 4,000 patients were randomized to tofacitinib 5 or 10 mg BID or TNFi, adalimumab in North America (Canada, USA, and Puerto Rico) and etanercept elsewhere in the world, all on background MTX. 

Patient baseline characteristics were similar in the three groups – median age approximately 60, primarily female, and Caucasian with mean 10 years of RA. Importantly, almost 50% of participants had a history of smoking and around one-third were age 65 or greater. The mean SDAI was 41.5. 

Non-inferiority was not met as the upper limit of the two-sided 95% CI for the HR for the combined tofacitinib doses (both 5 mg BID and 10 mg BID) versus TNFi was >1.8 both MACE (1.94) and malignancies (2.09). 

In looking at the details, the incidence rates per 100 patient-years for MACE (5 mg – 0.91; 10 mg - 1.05 and TNFi – 0.73) and malignancies (1.13 for both tofacitinib doses; 0.77 – TNFi) were low in all groups but numerically lower with TNFi. MACE and malignancies were higher for patients from North America versus elsewhere and most occurred in patients who were at least age 65 or had ever smoked. 

MACE, fatal MACE, and non-fatal MACE were all numerically higher with tofacitinib, but the confidence intervals overlapped [21]. Similar results were seen with myocardial infarction (MI) other than there were no fatal MIs with tofacitinib 5 mg BID [21]. The patients in this select group most at risk for MACE were those at least age 65, a history of smoking, male, and a high risk of MACE at baseline. Importantly, patients who were less than age 65 with no history of smoking had much lower incidences of MACE, both with tofacitinib and TNFi, with overlapping confidence intervals, suggesting that there is little difference in the risk of MACE in this population of patients although they were selected based on an increased risk of MACE. Of clinical importance, although the number of events were higher with tofacitinib, the number needed to harm for tofacitinib 5 mg BID relative to TNFi was 567 for MACE meaning that 567 more patients would need to be treated with tofacitinib for 1 more MACE than with TNFi [21].

Similar results were seen with respect to malignancies. The most frequent malignancies with tofacitinib were lung cancer and breast cancer with TNFi.  The risk for malignancies was highest in patients aged 65 or greater, male and history of smoking. Similar to MACE, importantly, patients who were less than age 65 with no history of smoking had much lower incidences of malignancies, both with tofacitinib and TNFi, with overlapping confidence intervals, suggesting that there is little difference in the risk of malignancy in this population of patients. As with MACE, the number needed to harm for malignancy was relatively high at 276 [22].  

A secondary endpoint of ORAL Surveillance was the risk of VTE. Again, there were numerically more VTE and pulmonary emboli, but not arterial thromboembolism, with tofacitinib 5 mg BID versus TNFi but no statistical differences. Tofacitinib 10 mg BID was associated with a statistically significant difference in VTE and PE versus TNFi. The number needed to harm for tofacitinib 5 mg BID versus TNFi was 763 for VTE and 870 for PE. The patients most at risk were those with a prior VTE, age 65 or greater, male sex, overweight, history of hypertension, or oral contraceptive use [23].

An analysis of efficacy showed that similar clinical benefit was seen in the three groups of patients. As seen in Figure 3, efficacy was similar in all three groups [20].

Major limitations of this analysis include that it is not known whether tofacitinib lowers the risks of MACE, malignancy, and VTE but perhaps not quite as well as TNFi, how the risks for tofacitinib compare with other DMARDs, are they specific to this high-risk population and just in RA, and if the relative risk differed between adalimumab and etanercept. What has not been seen yet is an analysis of disease activity in the patients who developed MACE, malignancy, or VTE as disease activity is a major driver of all three events in the RA population. Also, one should consider that the risk of MACE in this study with TNFi was much lower than in the ENTRACTE trial – an incidence rate of 1.27 in the on-treatment population in ENTRACTE compared to 0.73 in this study [13]. This raises the question that if this study were to be repeated, would the results be the same?

In conclusion, for MACE and malignancies, the non-inferiority criteria were not met in this RA high risk population; patients most at risk were at least age 65 or had ever smoked with little differences in the rate of events in those patients less than age 65 or who had never smoked. The NNH for tofacitinib 5 mg BID versus TNFi was 567 for MACE and 276 for malignancies. The incidence rates of VTE, including DVT and PE, were not statistically different for tofacitinib 5 mg BID versus TNFi although there were numerically more VTE with tofacitinib 5 mg BID. There are still major outstanding questions to define the group of patients who are at the highest risk of these events.

References:

1. https://www.pfizer.com/news/press-release/press-release-detail/pfizer-shares-co-primary-endpoint-results-post-marketing – January 27, 2021 downloaded November 2, 2021
2. FDA Drug Safety Communication (PDF - 255 KB) accessed 9/1/21
3. Avina-Zubieta JA, et al.  Arthritis Rheum 2008;59:1690–7.  
4.  Solomon DH, et al. Circulation 2003; 107:1303–7. 
5. Del Rincon I, et al. Arthritis Rheum 2002;48:1833–40.  
6. Myasoedova E, et al. Ann Rheum Dis 2011;70:482–7.
7. Micha R, et al. Am J Cardiol 2011;108:1362–70. 
8. Barnabe C, et al. Arthritis Care Res (Hoboken) 2011;63:522–9. 
9. Greenberg JD, et al. Ann Rheum Dis 2011;70:576–82. 
10. Solomon DH, et al. Arthritis & Rheumatology Vol. 67, No. 6, June 2015, pp 1449–1455
11. de La Forest D, et al. 2017 Joint Bone Spine 4(2)  133-140.
12. Singh, S, et al. Arthritis Care & Research Vol. 72, No. 4, April 2020, pp 561–576 
13. Giles J, et al Arthritis & Rheumatology Vol. 72, No. 1, January 2020, pp 31–40 
14. Kim, S et al. Arthritis & Rheumatology Vol. 72, No. 1, January 2020, pp 4–6
15. Choi HK, et al. Ann Rheum Dis 2013;72:1182–1187. 
16. Lee and Pope Arthritis Research & Therapy 2014, 16:435
17. Molander V et al. Ann Rheum Dis 2021 Feb;80(2):169-175
18. Cohen S, et al. 2020 RMD.Open 6(3)
19. Kremer J, et al ACR Open Rheumatology 2021, pp 1–12 DOI 10.1002/acr2.11232
20. Ytterberg S, et al. ACR 2021 Abstract 0831
21. Charles-Schoeman C et al ACR 2021 Abstract 0958
22. Curtis J, et al ACR 2021 Abstract 1940
23. Charles-Schoeman c, et al. ACR 2021 Poster 1941

Illustration by April Brust