Article Text

Download PDFPDF

Impact of Janus kinase inhibitors on risk of cardiovascular events in patients with rheumatoid arthritis: systematic review and meta-analysis of randomised controlled trials
  1. Wenhui Xie1,
  2. Yanrong Huang1,
  3. Shiyu Xiao2,
  4. Xiaoying Sun1,
  5. Yong Fan1,
  6. Zhuoli Zhang1
  1. 1 Department of Rheumatology and Clinical Immunology, Peking University First Hospital, Beijing, China
  2. 2 Department of Gastroenterology, Peking University Third Hospital, Beijing, China
  1. Correspondence to Professor Zhuoli Zhang, Department of Rheumatology and Clinical Immunology, Peking University First Hospital, Beijing 100006, China; zhuoli.zhang{at}


Objectives To investigate the effect of Janus kinase inhibitors (Jakinibs) on cardiovascular risk in adult patients with rheumatoid arthritis (RA) via a meta-analysis of randomised controlled trials (RCTs).

Methods PubMed, Embase and Cochrane library were thoroughly searched for RCTs reporting safety issues in patients with RA receiving Jakinibs, from inception to October 2018. The primary and secondary outcomes were all cardiovascular events (CVEs) and major adverse cardiovascular events (MACEs)/venous thromboembolism events (VTEs). OR and 95% CI were calculated using the Mantel-Haenszel fixed-effect method.

Results 26 RCTs randomising 11 799 patients were included. No significant difference was observed regarding all CVEs risk following Jakinibs usage in general (OR 1.04 (0.61 to 1.76), p = 0.89), tofacitinib (OR 0.63 (0.26 to 1.54), p = 0.31), baricitinib (OR 1.21 (0.51 to 2.83), p = 0.66), upadacitinib (OR 3.29 (0.59 to 18.44), p = 0.18), peficitinib (OR 0.43 (0.07 to 2.54), p = 0.35) or decernotinib (OR 1.12 (0.13 to 10.11), p = 0.92). Likewise, there was no significant difference for Jakinibs treatment overall regarding occurrence of MACEs (OR 0.80 (0.36 to 1.75), p = 0.57) or VTEs (OR 1.16 (0.48 to 2.81), p = 0.74). Dose-dependent impact of Jakinibs on the risks of all CVEs, MACEs and VTEs was not observed in tofacitinib (5 mg vs 10 mg), upadacitinib (15 mg vs 30 mg), whereas baricitinib at 2 mg was found to be safer than 4 mg in all CVEs incidence (OR 0.19 (0.04 to 0.88), p = 0.03).

Conclusion The existing evidence from RCTs indicated no significant change in cardiovascular risk for Jakinib-treated patients with RA in a short-term perspective, but postmarketing data are sorely needed to ascertain their cardiovascular safety, especially at the higher dose, due to increased risk of thromboembolism events for both tofacitinib and baricitinib at higher dosage.

  • rheumatoid arthritis
  • Janus kinase inhibitors
  • cardiovascular events
  • randomised controlled trials
  • meta-analysis

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known about this subject?

  • Rheumatoid arthritis (RA) has been associated with an increased risk of cardiovascular events (CVEs), including major adverse cardiovascular events (MACEs) and venous thromboembolism events (VTEs).

  • Both of two licensed Janus kinase inhibitors (Jakinibs), tofacitinib and baricitinib, are found to probably correlate with a higher risk of thromboembolism events at a higher dose.

  • The impact of Jakinib therapies on the risk of CVEs in RA population remains undetermined.

What does this study add?

  • The results of largest meta-analysis of 26 randomised controlled trials did not reveal a significant change in the risks of all CVEs, MACEs and VTEs, in patients with RA initiating Jakinib therapies, as compared to placebo.

  • During the limited controlled periods, no evidence of dose-dependent impact of tofacitinib and upadacitinib on the risk of all CVEs, MACEs and VTEs was observed, while baricitinib at 2 mg was found to be safer than 4 mg one time a day regarding all CVEs risk.

How might this impact on clinical practice or future developments?

  • Based on the best available datasets, this work reveals that Jakinib therapies are not associated with increased cardiovascular risk, but in consideration of an increased cardiovascular risk of 4 mg baricitinib in present study, the US Food and Drug Administration restricts approval to only the low-dose baricitinib, as well as the most recent warning about an increased risk of pulmonary embolism in higher dose tofacitinib, further investigations are urgently needed to confirm long-term cardiovascular safety of Jakinibs, especially at the higher dose.


Patients with rheumatoid arthritis (RA) have a substantially elevated risk of cardiovascular morbidity and mortality, which cannot be entirely explained by traditional cardiovascular risk factors.1–3 This indicates that RA-specific characteristics, especially systemic inflammation and disease activity, may be associated with increased cardiovascular risk.4 5 Reaching remission or low disease activity in order to decrease cardiovascular events (CVEs), therefore, is optimally advocated in current-released European League Against Rheumatism (EULAR) recommendations.6 Disease-modifying antirheumatic drugs (DMARDs), including methotrexate (MTX) and antitumour necrosis factor (anti-TNF), have shown cardioprotective potential in patients with RA, although the evidence is inconclusive.7–10 Irrespective of anti-inflammatory effects, drug-specific mechanisms are also involved in altering cardiovascular outcomes. For instance, glucocorticoids and rofecoxib, which possess potent anti-inflammatory properties, have cardiovascular toxicity.9 11

Janus kinase inhibitors (Jakinibs), classified as targeted synthetic DMARDs (tsDMARDs), is being widely investigated in randomised controlled trials (RCTs) and two of them have been subsequently approved for the treatment of RA in a number of countries.12–14 However, the role of these agents in the modulation of cardiovascular risk remains undetermined.15 16 Limited evidence suggested that tofacitinib-based therapy positively modifies the risk of cardiovascular disease,17 18 while a growing body of literature indicated that Jakinibs adversely affect several cardiovascular risk factors (such as serum lipid profile and platelet count) and potentially increase thrombotic risk which directly leads the US Food and Drug Administration (FDA) to restrict approval to only 2 mg baricitinib.19–22 Additionally, in upadacitinib premarketing trials, major adverse cardiovascular events (MACEs) continuously reported in patients with RA receiving upadacitinib has further generated cardiovascular concerns.23

In consideration of currently not well-established cardiovascular safety profile of these agents, we set out to explore the association between Jakinibs therapies and CVEs in adult patients with RA.


This article were carried out in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.24

Data sources

A systematic search of PubMed, Embase and the Cochrane library databases was performed without language restrictions from inception through 13 October 2018. Search terms comprised rheumatoid arthritis, Jakinibs (trade names, individual drug names, and chemical names) and randomized controlled trial. An example of search strategy is available in online supplementary appendix S1. We also searched Current Index to Nursing and Allied Health Literature (CINAHL), the major annual meetings (American College of Rheumatology, EULAR) in 2016–2018 and reference lists of all included studies for additional studies. Besides, the US National Institutes of Health Ongoing Trials Register ( and relevant FDA files were procured for additional details of clinical trials. Sponsored pharmaceutical companies or the authors were contacted if ambiguity existed.

Study selection

Two investigators (WX and YH) independently screened all titles and abstracts for potential inclusion. Discordances were resolved by a third experienced investigator (ZZ). We included double-blind RCTs that reported adverse events in adult patients with RA receiving at least one dose Jakinibs compared with placebo or two comparable doses of Jakinibs (tofacitinib (5 mg vs 10 mg, two times a day), baricitinib (2 mg vs 4 mg, one time a day), upadacitinib (15 mg vs 30 mg, one time a day)) during the randomised controlled phase. Exclusion criteria included non-randomised design, non-comparative study, healthy volunteers, paediatric patients and abstracts without full-text publication. We excluded one article as the comparator in the study was MTX, an ineligible arm and three articles due to follow-up duration <12 weeks.25–28 In addition, two studies specifically recruiting elderly (50 years and older) and seropositive patients or patients with bone erosion on radiography were also excluded due to their unrepresentativeness of RA population.29 30

Data extraction and outcome measures

Data were independently extracted by two reviewers (WX and YH) with a predesigned template: authors, publication year, sites/countries, trial design, patients’ characteristics, treatment regimens and CVEs. The overall number of CVEs during the randomised controlled phase was extracted for patients who received at least one dose of study medication or placebo. For extension RCTs in which treatment assignments were switched (for instance, patients initially treated with placebo switched to a Jakinib), the occurrence of CVEs was documented at switching point (usually 12–24 weeks). The longest randomised period (usually 24–52 weeks) was chosen to compare two dosage arms (tofacitinib (5 mg vs 10 mg two times a day), baricitinib (2 mg vs 4 mg one time a day), upadacitinib (15 mg vs 30 mg one time a day)), and total number of CVEs during eligible period for each dosage was extracted to identify dose-related cardiovascular effect. The Cochrane quality assessment tool for RCTs was used to assess risk of bias.31

The primary outcome was the relationship between Jakinibs (all or individual or different dosing) and all CVEs. The secondary outcomes switched to MACEs and venous thromboembolism events (VTEs), including pulmonary embolism (PE) and deep vein thrombosis (DVT). The outcomes of CVEs included angina pectoris, myocardial infarction, congestive heart failure, carotid artery disease, aortic aneurysm, cerebral vascular diseases (stroke and transient ischaemic attack), VTEs and cardiovascular death. MACE was defined as a composite endpoint of myocardial infarction, cerebrovascular accident (ischaemic and haemorrhagic strokes) or cardiovascular death. Two kinds of comparisons were made for both primary and secondary outcomes: (1) all Jakinibs or individual Jakinib versus placebo; (2) comparisons of different dosage (tofacitinib 5 mg vs 10 mg; baricitinib 2 mg vs 4 mg; upadacitinib 15 mg vs 30 mg). In the first comparison, all dosages of study agents were combined.

Statistical analysis

Extracted data were combined using Review Manager (RevMan) software (Cochrane collaboration). OR of CVEs, MACEs or VTE in patients receiving Jakinibs compared with placebo and the same Jakinib with different dosages was calculated by Mantel-Haenszel fixed-effect method. One of sensitivity analyses was conducted with the Peto method to examine whether analysis methods had influence on the results.32 Additional sensitivity analysis was performed after excluding four exclusively Japan-based studies or the studies without an independent cardiovascular committee. Forest plots were constructed to summarise the OR estimates and their 95% CI. Heterogeneity across studies was measured by χ2 test (p<0.05 was regarded statistically significant) and I2 statistics (significant heterogeneity, I2 >50%; insignificant heterogeneity, I2 <40%). Funnel plot analysis was used to detect the potential publication bias.


In total, 26 citations comprising 11 799 patients met predefined criteria, as summarised in figure 1.33–57 The included RCTs, covering a range of 19–281 (median 88) centres, were international (6–35 countries), except 4 from Japan.33 42 48 53 Most of the eligible trials were about tofacitinib,33–42 baricitinib43–48 and upadacitinib.49–52 Other Jakinibs, peficitinib53 54 decernotinib55 56 and filgotinib,57 were relatively less reported. Of these, 14 were phase II trials33 34 36 42 43 48–50 53–57 and 12 were phase III trials.35 37–41 44–47 51 52 Baseline characteristics of patients were generally comparable with regard to age, sex composition, disease duration and disease activity across most of arms and studies (online supplementary table S1).

Figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart of included randomised controlled trials.

Of included studies, 25 were eligible for the comparison of Jakinibs and placebo and 18 were eligible for the comparison of low and high dosages. The available duration for Jakinibs compared with placebo and different-dose comparisons ranged from 12 to 52 (median 12) weeks and 12 to 108 (median 24) weeks, respectively (online supplementary table S2). The numbers of CVEs (MACEs/VTEs) during the controlled phase of these studies were summarised in online supplementary table S2. The crude incidence rates of CVEs (MACEs/VTEs) in Jakinibs and placebo group were, respectively, 1.961 (0.593/0.547) and 1.222 (0.305/0.305) per 100 patient years. Regarding the dose comparison, the crude incidence rates of CVEs (MACEs/VTEs) in low-dose and high-dose arms were 1.367 (0.365/0.228) and 1.425 (0.445/0.267) per 100 patient years, respectively.


For comparison of Jakinibs against placebo, pooled analysis of 25 trials indicated no statistically significant difference in all CVEs risk (OR=1.04, 95% CI 0.61 to 1.76, p=0.89) (figure 2). Considered separately, statistical differences remained undetectable regarding the occurrence of all CVEs for tofacitinib (OR=0.63, 95% CI 0.26 to 1.54, p=0.31), baricitinib (OR=1.21, 95% CI 0.51 to 2.83, p=0.66), upadacitinib (OR=3.29, 95% CI 0.59 to 18.44, p=0.18), peficitinib (OR=0.43, 95% CI 0.07 to 2.54, p=0.35) and decernotinib (OR=1.12, 95% CI 0.13 to 10.11, p=0.92) (figure 2). Concerning MACEs, there was no statistically significant correlation either between all Jakinibs in general (OR=0.80, 95% CI 0.36 to 1.75, p=0.57) or individual agent compared with placebo (figure 3). The statistical heterogeneity across all intergroup and intragroups analysis was very low, allowing for combination of trial results using the fixed-effect method (figures 2 and 3). Subanalysis of VTEs showed a trend towards higher rates in Jakinibs-treated patients than placebo, but the difference did not reach statistical significance (OR=1.16, 95% CI 0.48 to 2.81, p=0.74) (figure 4). A notable variation was observed in three groups of Jakinibs (tofacitinib OR=0.17, 95% CI 0.03 to 1.05, p=0.06; baricitinib OR=2.33, 95% CI 0.62 to 8.75, p=0.21; upadacitinib OR=1.77, 95% CI 0.20 to 16.00, p=0.61). Additionally, separate meta-analyses of PE and DVT showed no increased difference in Jakinibs overall, relative to placebo (OR=0.91, 95% CI 0.30 to 2.77, p=0.87 for PE; OR=1.18, 95% CI 0.35 to 3.94, p=0.79 for DVT) (online supplementary figure S1).

Figure 2

OR of all cardiovascular events (CVEs) in patients treated with Janus kinase (JAK) inhibitors compared with placebo in randomised controlled trials using the Mantel-Haenszel (M-H) fixed-effect method. P-Y, patient–year.

Figure 3

OR of major adverse cardiovascular events (MACEs) in patients treated with Janus kinase (JAK) inhibitors compared with placebo in randomised controlled trials using the Mantel-Haenszel (M-H) fixed-effect method. P-Y, patient–year.

Figure 4

OR of venous thromboembolism events (VTEs) in patients treated with Janus kinase (JAK) inhibitors compared with placebo in randomised controlled trials using the Mantel-Haenszel (M-H) fixed-effect method. P-Y, patient–year. *In consideration of the US Food and Drug Administration warning about greater risk of thromboembolic events in the higher dose of 10 mg tofacitinib, relative to both the 5 mg regimen and antitumour necrosis factor therapy, the subanalysis of the risk of VTEs between JAK inhibitors and placebo or low-dose and high-dose regimens of tofacitinib, baricitinib, as well as upadacitinib was performed.

Considering the potential increased risk of VTEs in higher dose of Jakinibs, we conducted dose comparisons for tofacitinib, baricitinib and upadacitinib.58 59 For pairwise comparisons of tofacitinib 5 mg against 10 mg, upadacitinib 15 mg against 30 mg, no significant results were identified in the risk of all CVEs (tofacitinib OR=1.02, 95% CI 0.56 to 1.87, p=0.94; upadacitinib OR=2.61, 95% CI 0.80 to 8.50, p=0.11 for upadacitinib). Conversely, baricitinib at dose of 2 mg appeared to be safer than 4 mg based on randomised controlled datasets (OR=0.19, 95% CI 0.04 to 0.88, p=0.03) (figure 5). We found very low level of heterogeneity across all included RCTs (χ2=16.82; df=14; p=0.27; I2=17%), but a high heterogeneity was detected among the three kinds of Jakinibs (χ2=7.07; df=2; p=0.03; I2=71.7%) (figure 5). Additional analyses of MACEs and VTEs showed no dose-dependent effect in general (OR=0.86, 95% CI 0.38 to 1.92, p=0.71 for MACEs; OR=0.91, 95% CI 0.36 to 2.25, p=0.83 for VTEs) (online supplementary figures S2, S3). Concretely, a comparable impact of two dosages of tofacitinib was observed, while patients receiving 2 mg baricitinib tended to have lower occurrences of MACEs and VTEs than those receiving 4 mg one (OR=0.20, 95% CI 0.01 to 4.18, p=0.30 for MACEs; OR=0.23, 95% CI 0.02 to 2.17, p=0.20 for VTEs). In comparison with 30 mg upadacitinib, the ORs (95% CI) of MACEs and VTEs in 15 mg dosage were 0.71 (0.14–3.65) and 4.36 (0.47–40.57), respectively. Results of the separate analyses of PE and DVT risk in lower dose Jakinibs in comparison with higher dose one were shown in online supplementary figure S4 (OR=0.85, 95% CI 0.27 to 2.67, p=0.78 for PE; OR=1.01, 95% CI 0.25 to 4.16, p=0.99 for DVT).

Figure 5

OR of all cardiovascular events (CVEs) in patients treated with different dosages of Janus kinase (JAK) inhibitors in randomised controlled trials using the Mantel-Haenszel (M-H) fixed-effect method. P-Y, patient–year.

The sensitivity analyses using the Peto method or excluding four Japan-based studies showed similar results for main comparisons (online supplementary figures S5, S6). After we specifically included the studies with an independent cardiovascular committee, a numerically higher, but statistically non-significant risk of all CVEs in patients with RA receiving Jakinibs (OR=1.79, 95% CI 0.85 to 3.77, p=0.13) was observed (online supplementary figure S7).

Risk of bias assessment

We found that 13 RCTs (50%) adequately reported the generation of random sequence and 15 RCTs (65%) had adequate descriptions of concealed allocation. Blinding of participants, personnel, and outcome assessor was performed in all RCTs. Incomplete outcome data were well balanced in 23 RCTs (88%). Twelve RCTs (46%) had an independent cardiovascular safety adjudication committee. Of 23 RCTs, patient baseline characteristics in all intervention groups were well balanced (online supplementary table S3). For the Mantel-Haenszel (M-H) fixed-effect and Peto methods, funnel plot analysis showed no evidence of publication bias in all comparisons (figure 6 and online supplementary figure S8).

Figure 6

Funnel plots for the meta-analysis of occurrence of all cardiovascular events (a) Janus kinase (JAK) inhibitorsversus placebo (b) Different dosages comparison; major adverse cardiovascular events (c) JAK inhibitors versus placebo (d) Different dosages comparison; and venous thromboembolism events (e) JAK inhibitorsversus placebo (f) Different dosages comparison.


To our knowledge, this is the first meta-analysis exploring the relationship between Jakinib therapies and cardiovascular risks in RA population. According to our results, short-term use of Jakinibs does not increase the risk of CVEs, relative to placebo. Furthermore, there was no adequate evidence of dose-dependent impact on the occurrence of CVEs regarding tofacitinib 5 mg against 10 mg, upadacitinib 15 mg against 30 mg, but baricitinib at 2 mg appeared to be a safer treatment option against 4 mg one.

Tofacitinib and baricitinib have been approved for treating RA, but there are residual uncertainties regarding viral infection, malignancy, as well as cardiovascular safety. A recent meta-analysis assessed the hazard of developing malignancies or serious infections in patients with RA receiving Jakinibs treatment.58 ORs of 2.48 (95% CI 0.76 to 8.11) and 1.39 (95% CI 0.21 to 9.11) alarmed safety signals regarding malignancy and serious infection risk. Regarding cardiovascular safety, only one pooled analysis suggested that tofacitinib treatment might be associated with low incidence of CVEs,17 but the reliability was limited by uncertain comparability of participants’ characteristics and non-comparative control. In this present meta-analysis, ORs of 1.04 (95% CI 0.61 to 1.76) for CVEs, 0.80 (95% CI 0.36 to 1.75) for MACEs, 1.16 (95% CI 0.48 to 2.81) for VTEs, generally indicated a ‘neutral’ role of short-term Jakinibs treatment in cardiovascular outcomes. Separately viewed, tofacitinib, the first approved tsDMARD, seemed to be safe for both of all CVEs, MACEs and VTEs risk. Tofacitinib 5 mg is recommended as the optimum dose for the RA treatment after balancing the dose-related safety risks and efficacy benefits. As outlined, two dosing regimens of tofacitinib appeared to produce almost the same cardiovascular safety profile. However, the preliminary analysis of an ongoing safety trial enrolling elder (≥50) patients with RA with at least one cardiovascular risk factor showed the risk of PE to be fivefold higher in 10 mg tofacitinib compared with anti-TNF therapy, and approximately threefold higher than tofacitinib in previous clinical programme. Additionally, all-cause mortality in 10 mg arm was higher compared with 5 mg one or anti-TNF comparator.59 60 In this safety trial, the risk-benefit profile of 5 mg tofacitinib and anti-TNF remains appropriately balanced, which are consistent with the results of Desai et al.61 This trial will be completed by the end of 2019 and is expected to provide additional information. Baricitinib both at 2 mg and 4 mg doses have been approved for the treatment of RA in various countries worldwide, except USA with 2 mg permitted only. Based on randomised controlled datasets, baricitinib showed a trend towards increased risk of all CVEs, though statistically insignificant. Subanalyses revealed that the elevation was largely driven by VTEs rather than MACEs. Notably, there was an imbalance for the occurrence of VTEs in the baricitinib clinical programme. All VTEs occurred in patients receiving 4 mg baricitinib, in contrast to none in those receiving 2 mg. Pairwise comparisons of baricitinib 2 mg and 4 mg regimens suggested that 2 mg appeared to be a far safer therapeutic option in the consideration of cardiovascular outcomes. In FDA update documents including both controlled and uncontrolled datasets from four pivotal trials and associated long-term extension study (cut-off date of 1 April 2017), there were 5 and 34 VTEs in the baricitinib 2 mg and 4 mg, respectively, accompanied by a numerically lower incidence rate of VTEs in 2 mg baricitinib than in 4 mg (0.397 vs 0.584 per 100 patient years).62 As the duration of treatment exposure and numbers of VTEs were relatively small, future investigations are desired to quantify whether cardiovascular and thromboembolic risk is drug related and further dose related. Another promising Jakinibs, upadacitinib, appear to be associated with possible increased cardiovascular risk, despite no statistical significance. Paired comparisons showed 15 mg did not seem to be safer than 30 mg regarding all CVEs risk, but the interpretation of these findings is challenging given that the limited exposure is insufficient to draw a definite conclusion now. Concerning the remaining Jakinibs, no particular red signal in CVEs was alarmed based on the limited datasets.

On the other hand, accumulating evidence has indicated the positive correlation between well-controlled inflammation and improved cardiovascular outcomes in patients with chronic inflammation,63–65 including RA.7–10 Several agents, such as MTX and anti-TNF, have been indicative of cardioprotection in RA population. However, it is still largely open to discussion that tsDMARDs-based therapy serves as friend or foe in cardiovascular outcomes in patients with RA given unfavourable alterations in cardiovascular risk factors and the limitation in current safety datasets. In light of these findings of the present study and FDA recent safety warning, cardiovascular risk assessment such as age, hypertension, prior cardiovascular disease and medication patterns should be considered before initiation of high-dose Jakinibs treatment. Besides, an independent cardiovascular safety endpoint adjudication committee reviewing all potential cardiovascular and thromboembolic events should be routinely established in subsequent Jakinibs clinical programme. To develop a more thorough understanding of the role of these small molecules in cardiovascular outcomes, both continued surveillance of emerging data and the exploration of underlying mechanisms are sorely required.

Recent findings indicated that racial disparities occur in tsDMARDs-based strategies regarding the incidence of several side effects. Higher rate of herpes zoster infection in Japanese patients than overall populations is an example.66 However, access to ethnicity-level or patient-level data was not granted by any of the study sponsors. Four cases of CVEs (including one MACEs) reported in five Japanese population-based trials indicated cardiovascular safety in Asian population need to be taken seriously.33 42 48 53 67

Limitations may exist when the findings of our meta-analysis are interpreted. First, most of included trials had a relatively short duration of randomised controlled phase (12–104 weeks), which possibly influenced the performance of this meta-analysis to detect the relationship of Jakinibs to CVEs risk. But both Peto and Mantel-Haenszel fixed-effect methods were applied in the present meta-analysis, which strengthened the statistical power. Secondary, the background therapy (such as non-steroidal anti-inflammatory drugs and oral corticosteroid) may influence the results, but the randomisation producing similar treatment patterns and background cardiovascular risk among arms could minimise the potential bias, to some extent. Besides, patient selection process of clinical trials may indirectly influence the background cardiovascular risk, such as excluding elderly patients who are at increased risk of CVEs, although the included RCTs did not explicitly restrict the patients at elevated cardiovascular risk. This may consequently limit the generalisability of the results.

In conclusion, based on the best available evidence from RCTs, the present meta-analysis indicates that neither individual Jakinib nor all Jakinibs in general significantly influences cardiovascular outcomes in adult patients with RA. Meanwhile, baricitinib at 4 mg seems to be associated with a higher occurrence of CVEs, relative to 2 mg dose. These results should be interpreted specifically in the context of limited duration of follow-up in RCTs. In consideration of these findings in present study, FDA restricts approval to only the low-dose baricitinib, as well as the most recent warning about an increased risk of PE with higher dose tofacitinib, continuous postmarketing surveillance of emerging data is urgently needed to comprehensively clarify the association of Jakinibs and cardiovascular outcomes in RA population.


The authors thank Roy Fleischmann for his kindly answer to questions, Zhan Siyan for helpful assistance of statistical information and the pharmaceutical companies for kindly feedback on research data.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.


  • Handling editor Josef S Smolen

  • Contributors ZZ conceived the study, participated in its design and coordination, and critically revised the manuscript. WX had full access to all of the data collection, analysis, interpretation and drafted the manuscript. YH, XS, YF and SX were study investigators and contributed to the process of data collection. All authors read and approved the final manuscript.

  • Funding This work was supported by the National Natural Science Foundation of China (81771740), the subproject (2010CB529103) under the National Science Technology Pillar Program of China (973 Program) (2010CB529100), the Capital Health Research and Development of Special Fund Program (2011-4021-03), the Peking University Clinical Research Program (PUCRP201305) and Beijing Natural Science Foundation (7184251).

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data are available in a public, open access repository.