Comparison of cardiovascular, cancer and herpes zoster risk of tofacitinib versus etanercept: single-centre observational study

In this real-world study, incidence rates of MACE, cancer, thrombosis and zoster were similar in RA patients using tofacitinib or etanercept as first-line b/tsDMARDs. Dear Editor, Biologic/targeted-synthetic DMARDs (b/tsDMARDs) have been the game-changers in the management of RA over the last two decades. However, safety concerns have drastically increased with their widespread clinical use; for example, the ORAL Surveillance trial of tofacitinib [1]. This study point outs an increased risk of major adverse cardiovascular event (MACE) and malignancy with tofacitinib compared with etanercept and adalimumab in patients with at least one cardiovascular risk factor [1]. Additionally, observational studies are needed to test the safety of these drugs in daily clinical settings. We used the HURBIO-RA registry that was established in 2005. Rheumatoid arthritis was classified according to the treating physician and/or EULAR/ACR 2010 criteria [2]. Rheumatoid arthritis patients prescribed either tofacitinib (5 mg twice daily) or etanercept (50 mg/week) as first-line b/tsDMARDs were screened for this study, revealing 247 patients for tofacitinib and 480 patients for etanercept. Patients with missing baseline disease-activity data (as a surrogate marker of overall RA disease burden) or without a control visit were excluded, leaving a remaining 145 patients for tofacitinib and 114 patients for etanercept. All patients were questioned via telephone by the treating physician regarding cardiovascular events, metabolic comorbidities, and malignancies. National Electronic Health Records of patients who have given permission were also checked for these variables. Similar to ORAL Surveillance, we aimed to compare the incidence of adjudicated MACE (after 60 days of drug exposure), cancers excluding nonmelanoma skin cancer, thromboembolic events and herpes zoster. MACE was defined as cardiovascular death (not due to pulmonary embolism), non-fatal myocardial infarction, or non-fatal stroke. For all outcomes, we calculated incidence rate (IR) and incidence rate ratios (IRR) per 1000 patient years. As the number of MACE events was low, we did not perform multivariable analysis because the model fit and power were considered low. Type-I error <0.05 infers statistical significance. A total of 259 patients (145 in tofacitinib and 114 in etanercept groups) were included. At baseline assessment (Table 1), patients in the tofacitinib group were older, had higher odds of the female sex, exposed to active smoke, higher odds of seropositivity, lower RA activity, higher odds of hypertension and family history of early CHD, lower odds of history of cancer compared with etanercept group. Overall tofacitinib and etanercept survival rates were similar (log-rank, P = 0.17). MACE occurred in similar IR among both groups [three patients in tofacitinib vs four patients in etanercept group, IRR: 0.8 (0.2–4.1), P = 0.84]. Of the three patients in the tofacitinib group, one had smoking history, one had history of stroke and hypertension at baseline, one had familial history of early CHD; all three patients had non-fatal MI as the subtype of MACE. Of the four patients in the etanercept group, one had smoking history and hypertension at baseline (this patient had cardiovascular death due to stroke), one had smoking history, one had hypertension at baseline and one had no known risk factor; three patients in the etanercept group had non-fatal MI as the subtype of MACE. In addition, three patients in the tofacitinib group died suddenly; however, we cannot classify these patients as ‘sudden cardiac death’, we did not count these events as MACE. However, if we count them as MACE, the IRR for MACE was 1.6 (0.5–6.7) (P = 0.44). Although IRRs for venous thrombosis, cancer and herpes zoster were statistically similar, it is worth noting that all three patients with thromboembolic events were in the tofacitinib group (Table 1), and none of these three patients had a previous episode of venous thromboembolism. Baseline disease characteristics, incidence rate, incidence rate ratios for outcomes of tofacitinib and etanercept groups Mean (s.d.) (compared with independent sample t test), remaining data was presented as n(%) (compared with χ2 test). Median (interquartile range) (compared with Mann–Whitney U test). Per 1000 patient-years. Assessed similar to ORAL Surveillance. b/tsDMARDs: biologic and targeted-synthetic DMARDs; CHD: coronary heart disease; DAS28-ESR: DAS-28 joint with ESR; DVT: deep vein thrombosis; HAQ: health assessment questionnaire; HDL: high density lipoprotein; IR: incidence rate; IRR: incidence rate ratio; MACE: major adverse cardiovascular event; NA: not applicable; PTE: pulmonary thromboembolism. Baseline disease characteristics, incidence rate, incidence rate ratios for outcomes of tofacitinib and etanercept groups Mean (s.d.) (compared with independent sample t test), remaining data was presented as n(%) (compared with χ2 test). Median (interquartile range) (compared with Mann–Whitney U test). Per 1000 patient-years. Assessed similar to ORAL Surveillance. b/tsDMARDs: biologic and targeted-synthetic DMARDs; CHD: coronary heart disease; DAS28-ESR: DAS-28 joint with ESR; DVT: deep vein thrombosis; HAQ: health assessment questionnaire; HDL: high density lipoprotein; IR: incidence rate; IRR: incidence rate ratio; MACE: major adverse cardiovascular event; NA: not applicable; PTE: pulmonary thromboembolism. In this study, we found a similar risk of MACE, cancer, thrombosis and herpes zoster for patients taking tofacitinib and etanercept in daily practice, in spite of baseline characteristics (e.g. age, hypertension and familial history of early CHD) of the tofacitinib group seeming more prone to outcomes, especially MACE. Meanwhile, all patients with venous thromboembolism were in the tofacitinib group, warranting further safety assessment and caution. Although we totally agree with the ORAL Surveillance results and comments of Singh [3], especially the one that questions whether tofacitinib fails to decrease cardiovascular risk by RA itself or increases cardiovascular risk, we think more data is needed to assess the safety profile of these drugs on average and low-cardiovascular risk populations. Our study has some limitations. We had many patients excluded from the etanercept group because of missing baseline data. To understand the possible effects of this situation, we compared demographic characteristics and disease duration of included and excluded patients and they were similar. Also, recall bias is an inherent possibility. To overcome it, we have phoned every patient or their relatives, used the hospital and national health records systems. In conclusion, incidence rates of MACE, cancer, thrombosis and herpes zoster were similar for tofacitinib and etanercept in this real-life experience. Although the difference was not statistically significant, all patients with venous thromboembolism were in the tofacitinib group, which clearly needs further assessment. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article. Disclosure statement: U.K. received honoraria from Abbvie, Amgen, Johnson and Johnson, MSD, Novartis, Pfizer, Roche, UCB. İ.E. received honoraria from Abbvie, Amgen, Johnson and Johnson, MSD, Novartis, Pfizer, Roche, UCB. S.K. received honoraria from Abbvie, Amgen, Johnson and Johnson, MSD, Novartis, Pfizer, Roche, UCB. Other authors declare no conflict of interest. Ethical approval: The protocol of this study was approved by Hacettepe University ethical committee (Number: GO 21/1251; Date: 16 November 2021). The datasets used and analysed during the current study are available from the corresponding author on reasonable request.