Abstract

Objective This posthoc analysis investigated the relationship between paraoxonase-1 (PON1) genotype and activity, and risk of major adverse cardiovascular events (MACE) and malignancies in clinical studies of tofacitinib in patients with rheumatoid arthritis (RA).

Methods Data were pooled from 9 phase II/III studies and the associated long-term extension studies (all completed by October 2017). PON1 activities in plasma were measured using paraoxon (paraoxonase activity), dihydrocoumarin (lactonase activity), and phenylacetate (arylesterase activity) as substrates. PON1 Q192R genotype effect on baseline PON1 activity was assessed using linear regression for each study, with fixed-effects metaanalysis across studies. MACE and malignancy risk by time-varying enzyme activity was determined using Cox proportional hazards regression.

Results The analysis included 1969 patients with RA. Compared with the QQ genotype, the RR genotype had a significant positive association with baseline paraoxonase activity and a significant negative association with baseline lactonase and arylesterase activity (all P < 0.001). Time-varying models demonstrated a significant association of increased paraoxonase activity over time with lower risk of MACE (P < 0.001) and malignancies (excluding nonmelanoma skin cancer [NMSC]; P ≤ 0.05), even after controlling for risk factors identified in univariate analysis and RA disease activity. A similar trend was observed for lactonase and arylesterase for MACE.

Conclusion Higher paraoxonase activity over time was associated with significantly reduced risk of future MACE and malignancies (excluding NMSC), but not NMSC, in patients with RA receiving tofacitinib. Further investigation of PON1 as a novel functional lipid biomarker of MACE/malignancy risk in patients with RA is warranted. (ClinicalTrials.gov: NCT01059864, NCT00550446, NCT00687193, NCT00960440, NCT00814307, NCT00856544, NCT00853385, NCT00847613, NCT01039688, NCT00413699, NCT00661661)

Key Indexing Terms:

Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by high levels of systemic inflammation and increased morbidity and mortality.1 An increased risk of cardiovascular disease (CVD; including myocardial infarction, congestive heart failure, and stroke) and certain malignancies (lymphomas, lung cancer, and nonmelanoma skin cancer [NMSC]) is observed in patients with RA vs the general population.2-5 Chronic systemic inflammation and/or RA disease activity have been shown to be risk factors for these comorbidities.6-10

Paraoxonase-1 (PON1) is a major high-density lipoprotein cholesterol (HDL-C)–associated enzyme synthesized mainly in the liver and secreted into the blood, where it associates predominantly with HDL-C.11,12 PON1 has paraoxonase, lactonase, and arylesterase activities, promotes the antioxidant and antiinflammatory properties of HDL-C, and inhibits oxidation of low-density lipoprotein cholesterol (LDL-C).12 PON1 is considered atheroprotective and is implicated in cancer development, innate immunity, and bacterial quorum sensing.12

The PON1 gene contains several polymorphisms, including a mutation from glutamine (Q, wild type) to arginine (R, variant) at amino acid position 192 of the protein (Q192R genotype). This results in differential ability of Q vs R isoenzymes to metabolize certain substrates (eg, lipid peroxides).13 PON1 has been hypothesized to hydrolyze and regulate lactone lipid mediators from polyunsaturated fatty acids.14 The RR genotype of the PON1 Q192R polymorphism is associated with increased paraoxonase activity and lower CVD risk vs the QR and QQ genotypes in patients with RA.15 Metaanalyses have demonstrated an association between PON1 Q192R and reduced malignancy risk in the general population, although larger studies are required to define any predictive role.16,17 Overall, however, our basic understanding of PON1, such as the physiological substrates in humans and its clinical significance, is lacking.

Tofacitinib is an oral Janus kinase inhibitor used for the treatment of RA. In phase II (n = 2) and phase III (n = 1) studies, tofacitinib (5 and 10 mg twice daily [BID]) significantly increased PON1 activity levels at month 3 from baseline vs placebo.18 Here, we investigated the relationship between PON1 genotype and activity, and between PON1 activity at baseline and over time, with the risk of major adverse cardiovascular events (MACE) and malignancies in tofacitinib-treated patients with RA.

METHODS

Study design and patients. This posthoc analysis included pooled data from patients with RA who received ≥ 1 dose of tofacitinib 5 or 10 mg BID, with or without conventional synthetic disease-modifying antirheumatic drugs in 3 phase II randomized controlled trials (RCTs; ClinicalTrials.gov: NCT01059864, NCT00550446, NCT00687193),19-21 6 phase III RCTs (NCT00960440, NCT00814307, NCT00856544, NCT00853385, NCT00847613, NCT01039688),22-27 and 2 long-term extension (LTE) studies (NCT00413699, NCT00661661; data cut-off: March 31, 2015); all studies were completed by October 26, 201728,29 (Supplementary Table S1, available with the online version of this article). ORAL Surveillance (NCT02092467), an open-label, endpoint-driven, US Food and Drug Administration–required postmarketing safety study, was ongoing at the time of analysis and not included.

Studies were conducted in accordance with the Declaration of Helsinki and the International Council for Harmonisation Guidelines for Good Clinical Practice and approved by the institutional review board and/or independent ethics committee for each study center. Patients provided written informed consent.

Patient and public involvement. Patients and/or the public were not involved in the design, conduct, reporting, or dissemination plans of this research.

Posthoc analysis: patients and outcomes. Patients were required to have ≥ 1 PON1 enzyme activity measurement at individual index study baseline. To eliminate confounding effects of background genetics, genotype analysis was restricted to patients from 8 of 9 studies (excluding NCT00687193 in Japanese patients) who had been genetically determined to cluster with the most represented ethnicity, such as European/White (based on transancestry genome-wide association study [GWAS] metaanalyses with principal components analysis via EIGENSTRAT).30

PON1 Q192R genotype (QQ, QR, or RR) was determined using germline DNA extracted from peripheral blood samples at individual study baseline only.30 PON1 activity and fasting lipid levels (total cholesterol, LDL-C, HDL-C, and triglycerides) were measured in patient plasma samples. On average, 2 to 3 PON1 assays and lipid level measurements were taken per patient (at timepoints up to month 12 for PON1 and up to month 24 for lipids). Paraoxonase, lactonase, and arylesterase activity levels were measured using paraoxon, dihydrocoumarin, and phenylacetate substrates, respectively.31 The data relating to PON1 enzymatic activities were generated and validated by the laboratory of Dr. Charles-Schoeman. Disease activity was assessed using the 4-variable Disease Activity Score in 28 joints based on erythrocyte sedimentation rate (DAS28-ESR4) and C-reactive protein (DAS28-CRP4) at timepoints up to month 24. All assessment timepoints are detailed in Supplementary Data S1 (available with the online version of this article).

Safety was assessed throughout each study until completion or within 28 days after drug discontinuation (clinical trial observation period). Incident safety events included MACE (any myocardial infarction, cerebrovascular event [stroke], or cardiovascular mortality event [death caused by coronary, cerebrovascular, and cardiac events]), malignancies (excluding NMSC), and NMSC. Event adjudication is summarized in Supplementary Data S1 (available with the online version of this article).

Statistical analyses. Baseline demographic and disease characteristics were collected at individual study baseline. Average tofacitinib dose was 5 mg BID where average total daily dose (TDD) was < 15 mg/day, and 10 mg BID where average TDD was ≥ 15 mg/day.

The effect of the PON1 Q192R genotype (QQ, QR, or RR) on baseline paraoxonase, lactonase, and arylesterase activity was assessed using linear regression for each individual study, with age and sex as covariates. A fixed-effects metaanalysis was then used to assess the effects across studies, with heterogeneity assessed with Cochran Q test. In the event of significant heterogeneity, random-effects (DerSimonian and Laird) metaanalysis was conducted.

Risk of incident safety events was assessed using Cox proportional hazards regression, stratified by the studies. Time of event was based on number of days from the date of first tofacitinib exposure. For patients with multiple events, time of first event was used. For patients without an event, right-censored follow-up time was based on exit date from the study.

Association between baseline characteristics and incident safety events was determined using univariate Cox proportional hazards regression models. Univariate and multivariate regression analyses against time-varying enzyme activity used 10 models. PON1 enzyme activities were tested separately. Associations were expressed as hazard ratios (HRs) with 95% CIs. Nominal P values are reported. Model details are included in the Supplementary Data S1 (available with the online version of this article).

RESULTS

Analysis cohorts. In total, 1969 patients from 9 studies with PON1 enzyme activity assessed at individual study baseline were included (Supplementary Figure S1, available with the online version of this article). PON1 Q192R genotyping data were available from the GWAS for 1233 European/White patients from 8 studies (excluding NCT00687193; QQ, n = 645; QR, n = 485; RR, n = 103; Supplementary Figure S1).

A total of 2336 safety events from 1374 patients were identified across 9 studies (total tofacitinib exposure: 6665 patient-years [PY; tofacitinib 5 and 10 mg BID: 1719 and 4946 PY, respectively]; Supplementary Figure S1, available with the online version of this article). Of these, 798 safety events were from 471 patients with baseline PON1 enzyme activity, including 49 MACE in 39 patients, 55 malignancies (excluding NMSC) in 53 patients, and 88 NMSC in 56 patients.

Effect of PON1 Q192R genotype (RR vs QQ) on enzyme activity. The RR genotype had a significant positive association with baseline paraoxonase activity, and a significant negative association with baseline lactonase and arylesterase activity vs the QQ genotype (all P < 0.001; Figure 1).

Figure 1.

The effect of PON1 Q192R genotype (RR vs QQ) on (A) paraoxonase, (B) lactonase, and (C) arylesterase enzyme activities. Fixed-effects model. Estimates > 0 favor the RR genotype, and estimates < 0 favor the QQ genotype. PON1: paraoxonase-1.

Associations between baseline and time-varying covariates and MACE. Univariate analysis demonstrated that traditional cardiovascular risk factors at baseline, such as age (unit = 10 years; HR 2.14, 95% CI 1.54-2.96), BMI (calculated as weight in kilograms divided by height in meters squared; HR 1.05, 95% CI 1.01-1.09), smoking (ex-smoker/never smoked: HR 2.37, 95% CI 1.09-5.13; smoker/never smoked: HR 2.58, 95% CI 1.2-5.55), hypertension (HR 2.12, 95% CI 1.12-3.99), and diabetes (HR 3.65, 95% CI 1.59-8.39) were significantly associated with MACE (all P ≤ 0.05; Figure 2).

Figure 2.

Univariate analysis of associations between baseline covariates and occurrence of MACE. Univariate Cox proportional hazards regression models included baseline PON1 enzyme activities, age, sex, BMI, smoking status, RA duration, average tofacitinib dose, cardiovascular comorbidities, concomitant medications, lipid levels, and baseline disease activity as covariates. a All PON1 activities, lipids, and DAS28 scores were either log-transformed or square-root transformed and then standardized (mean 0 [SD 1]). * P ≤ 0.05. ** P < 0.01. *** P < 0.001. DAS28-CRP4: 4-variable Disease Activity Score in 28 joints based on C-reactive protein; DAS28-ESR4: 4-variable Disease Activity Score in 28 joints based on erythrocyte sedimentation rate; FDR: false discovery rate; HDL-C: high-density lipoprotein cholesterol; HR: hazard ratio; LDL-C: low-density lipoprotein cholesterol; MACE: major adverse cardiovascular events; N: no; PON1: paraoxonase-1; RA: rheumatoid arthritis; Y: yes.

Time-varying models found a significant association of increased paraoxonase activity over time with lower MACE risk in 9 out of 10 models tested, after controlling for time-varying LDL-C or HDL-C, total cholesterol, triglycerides, DAS28-ESR4, and DAS28-CRP4, and other traditional cardiovascular risk factors from univariate analysis (HR 0.49, 95% CI 0.33-0.73 to HR 0.56, 95% CI 0.41-0.75; all P < 0.001; Figure 3). The association of increased paraoxonase activity with MACE risk was not significant in a fully adjusted model controlling for PON1 genotype in European/White patients (HR 0.69, 95% CI 0.39-1.22; P = 0.21; Figure 3).

Figure 3.

Univariate and adjusted multivariable analyses of associations between baseline covariates and time-varying paraoxonase activity and lipids, and occurrences of MACE. Univariate regression was conducted with time-varying PON1 activity included in the model. Minimally adjusted multivariable regression models included time-varying PON1 activity and additional covariates of tofacitinib dose and age. Fully adjusted multivariable regression models included time-varying PON1 activity, and additional covariates of tofacitinib dose, age, BMI, diabetes, and smoking status. a Paraoxonase activities, lipids, DAS28-ESR4, and DAS28-CRP4 were time-varying, were either log-transformed or square-root transformed, and then standardized (mean 0 [SD 1]). b Paraoxonase activities were baseline, log-transformed, and then standardized (mean 0 [SD 1]). * P ≤ 0.05. ** P < 0.01. *** P < 0.001. DAS28-CRP4: 4-variable Disease Activity Score in 28 joints based on C-reactive protein; DAS28-ESR4: 4-variable Disease Activity Score in 28 joints based on erythrocyte sedimentation rate; EUR: European; HDL-C: high-density lipoprotein cholesterol; HR: hazard ratio; LDL-C: low-density lipoprotein cholesterol; MACE: major adverse cardiovascular events; N: no; PON1: paraoxonase-1; Y: yes.

Time-varying HDL-C and triglyceride levels were associated with significantly lower and increased MACE risk, respectively, in fully adjusted models controlling for baseline levels (Figure 3). Time-varying DAS28-ESR4 and DAS28-CRP4 were not significantly associated with MACE risk (Figure 3).

Similar findings were generally observed for lactonase and arylesterase activity (Supplementary Figures S2 and S3, available with the online version of this article). Increasing enzyme activity over time was significantly associated with lower MACE risk in multiple models for both lactonase (HR 0.52, 95% CI 0.27-0.99 to HR 0.56, 95% CI 0.34-0.91) and arylesterase (HR 0.68, 95% CI 0.49-0.96 to HR 0.72, 95% CI 0.53-0.99). Associations of lactonase and arylesterase activities and MACE were not significant in the fully adjusted time-varying models controlling for PON1 genotype or in multivariate models, including only baseline activity (fully adjusted and multivariate baseline models).

Associations between baseline and time-varying covariates and malignancies. In univariate analysis, age (HR 1.54, 95% CI 1.19-2.00), BMI (HR 1.04, 95% CI 1.01-1.08), RA duration (HR 1.05, 95% CI 1.02-1.08), smoking status (ex-smoker/never smoked: HR 2.83, 95% CI 1.46-5.49; smoker/never smoked: HR 3.48, 95% CI 1.8-6.73), and triglyceride levels (HR 1.42, 95% CI 1.08-1.88) were identified as risk factors for malignancies (excluding NMSC; all nominal P ≤ 0.05; Figure 4). tofacitinib 10 mg BID was significantly associated with lower malignancy (excluding NMSC) risk vs tofacitinib 5 mg BID (HR 0.42, 95% CI 0.21-0.81; P < 0.01; Figure 4). Age (HR 2.34, 95% CI 1.75-3.13), sex (HR 2.43, 95% CI 1.39-4.24), RA duration (HR 1.04, 95% CI 1.01-1.07), smoking status (ex-smoker/never smoked: HR 3.68, 95% CI 2.04-6.66), statin use (HR 2.63, 95% CI 1.39-4.98), hypertension (HR 2.67, 95% CI 1.56-4.57), and triglycerides (HR 1.35, 95% CI 1.03-1.75) were identified as NMSC risk factors (all nominal P ≤ 0.05; Supplementary Figure S4, available with the online version of this article).

Figure 4.

Univariate analysis of associations between baseline covariates and occurrence of malignancies (excluding NMSC). Univariate Cox proportional hazards regression models included baseline PON1 enzyme activities, age, sex, BMI, smoking status, RA duration, average tofacitinib dose, cardiovascular comorbidities, concomitant medications, lipid levels, and baseline disease activity as covariates. a All PON1 activities, lipids, and DAS28 scores were either log-transformed or square-root transformed, and then standardized (mean 0 [SD 1]). *P ≤ 0.05. **P < 0.01. ***P < 0.001. DAS28-CRP4: 4-variable Disease Activity Score in 28 joints based on C-reactive protein; DAS28-ESR4: 4-variable Disease Activity Score in 28 joints based on erythrocyte sedimentation rate; FDR: false discovery rate; HDL-C: high-density lipoprotein cholesterol; HR: hazard ratio; LDL-C: low-density lipoprotein cholesterol; N: no; NMSC: nonmelanoma skin cancer; PON1: paraoxonase-1; RA: rheumatoid arthritis; Y: yes.

There was a significant association between increased paraoxonase activity over time and lower malignancy (excluding NMSC) risk in 7 of 10 time-varying models, even after controlling for risk factors, including RA duration, age, and smoking status, or time-varying DAS28-ESR4 and DAS28-CRP4 (HR 0.71 [95% CI 0.55-0.91] to HR 0.75 [95% CI 0.60-0.94]; all P ≤ 0.05; Figure 5). Two other time-varying models, fully adjusted for triglycerides and HDL-C, were close to statistical significance (P = 0.055 and P = 0.06, respectively). There was no significant association in a fully adjusted model controlling for PON1 genotype (HR 0.73, 95% CI 0.47-1.12; P = 0.15; Figure 5). Baseline PON1 activity alone and genotype were not significantly associated with malignancy (excluding NMSC) risk in fully adjusted and multivariate baseline models, respectively (Figure 5). For NMSC, no significant association of increased paraoxonase activity over time, or baseline PON1 activity and genotype was observed (Supplementary Figure S5, available with the online version of this article).

Figure 5.

Univariate and adjusted multivariable analyses of associations between baseline covariates and time-varying paraoxonase activity and lipids, and occurrences of malignancies (excluding NMSC). Univariate regression was conducted with time-varying PON1 activity included in the model. Minimally adjusted multivariable regression models included time-varying PON1 activity and additional covariates of tofacitinib dose and age. Fully adjusted multivariable regression models included time-varying PON1 activity, and additional covariates of tofacitinib dose, age, RA duration, and smoking status. a Paraoxonase activities, lipids, DAS28-ESR4, and DAS28-CRP4 were time-varying, log-transformed, and then standardized (mean 0 [SD 1]). b Paraoxonase activities were baseline, log-transformed, and then standardized (mean 0 [SD 1]). * P ≤ 0.05. ** P < 0.01. *** P < 0.001. DAS28-CRP4: 4-variable Disease Activity Score in 28 joints based on C-reactive protein; DAS28-ESR4: 4-variable Disease Activity Score in 28 joints based on erythrocyte sedimentation rate; EUR: European; HDL-C: high-density lipoprotein cholesterol; HR: hazard ratio; LDL-C: low-density lipoprotein cholesterol; NMSC: nonmelanoma skin cancer; PON1: paraoxonase-1; RA: rheumatoid arthritis.

Increased lactonase and arylesterase activity over time were not significantly associated with malignancies (excluding NMSC) or NMSC in the 10 models tested (Supplementary Figures S6-9, available with the online version of this article).

DISCUSSION

This posthoc analysis investigated the relationship between PON1 activity and the PON1 Q192R polymorphism, and risk of MACE and malignancies, in almost 2000 tofacitinib-treated patients with moderate to severe RA in phase II/III RCTs and LTE studies (all studies completed by 2017). To our knowledge, this is the first analysis to study the clinical associations of PON1 (using paraoxonase, lactonase, and arylesterase activity assays) with safety outcomes in a large, longitudinal clinical study program.

Our analysis demonstrated a significant positive association between the PON1 Q192R RR genotype and baseline paraoxonase activity, and a significant negative association with baseline lactonase and arylesterase activity vs the QQ genotype. Although the association between the RR genotype and higher paraoxonase activity of PON1 is established in the general population,32 and has been reported in patients with RA,15 this is the first large study, to our knowledge, to describe a negative association of the RR genotype with PON1 arylesterase and lactonase activities. Studies in women with gestational diabetes or uncomplicated pregnancies reported a similar association of lower lactonase activity with the RR vs QQ genotypes.33

Univariate analyses demonstrated significant associations of traditional cardiovascular risk factors at baseline (age, BMI, smoking status, hypertension, and diabetes) with MACE. These risk factors (excluding smoking status) were identified in a previous posthoc analysis of data from trials of tofacitinib in patients with RA,34 and, excluding age, are consistent with those reported for other RA cohort studies.8 Baseline risk factors in univariate analysis for malignancies (excluding NMSC) and NMSC included age, RA duration, and smoking status. These findings are supported by data from ORAL Surveillance, which identified age ≥ 65 years and smoking status as risk factors for MACE and malignancies.35,36 Of note, the lower malignancy risk (excluding NMSC) associated with tofacitinib (10 vs 5 mg BID) could potentially relate to factors associated with LTE study methodology that could not be controlled for in this analysis. However, any differences in MACE and malignancies between the tofacitinib 5 and 10 mg BID groups could also be due to chance, given that these are considered rare events and the studies included in this analysis were not enriched for cardiovascular risk.

Increased paraoxonase activity over time during tofacitinib treatment demonstrated significant associations with lower risk of MACE and malignancies (excluding NMSC). This protective effect of increased paraoxonase activity was observed when controlling for baseline risk factors identified in univariate analyses and adjusting for time-varying lipid levels or disease activity. These observations are of interest given the previous reports of paradoxical LDL-C and total cholesterol associations with CVD risk in some patients with RA and indicate the need for more accurate cardiovascular risk biomarkers.9,37 Consistent with previous work,34 there were no associations between LDL-C or total cholesterol over time with MACE, but higher HDL-C levels over time remained associated with lower MACE risk. Higher triglycerides over time were also associated with increased cardiovascular risk in this analysis. Baseline levels alone for traditional lipid levels, DAS28-ESR4, DAS28-CRP4, and PON1 activity were not associated with future MACE or malignancy, likely reflecting the relatively small event numbers and, thus, low power to identify associations. Higher lactonase and arylesterase activities of PON1 over time were significantly associated with a lower MACE risk in most models but did not associate with malignancy risk. Further work is warranted to understand the physiological substrates and pathways for these 3 enzyme activities of PON1 in patients with RA, which may drive differences in their relationship with MACE vs malignancy outcomes. PON1 activities did not associate with the occurrence of NMSC over longitudinal follow-up.

Human PON1 transgene overexpression in mouse models of atherosclerosis has been associated with improved ability of HDL-C to inhibit LDL-C oxidation and reduction of atherosclerotic plaque.38 Our previous work has associated higher paraoxonase activity of PON1 with reduced cardiovascular risk in patients with RA measured by carotid atherosclerosis.15 In patients without RA, higher paraoxonase and arylesterase activities of PON1 have been associated with reduced MACE risk, with higher paraoxonase activity associated with lower levels of systemic oxidative stress.32 Oxidative stress is implicated in RA and may drive carcinogenesis and CVD39; therefore, a reduction in oxidative stress via PON1 signaling may have contributed to the reduced MACE and malignancy (excluding NMSC) risk observed here in tofacitinib-treated patients with RA.

The functional PON1 Q192R polymorphism has been associated with CVD in the general population;32 the functional PON1 Q192R polymorphism has also been associated with cardiovascular risk measured by carotid atherosclerosis in RA.15 It has also been associated with a predisposition to certain malignancies,17,40 and lower serum levels of PON1 activity have been reported in patients with cancer vs healthy participants.41 We did not identify a significant association of the PON1 Q192R polymorphism genotype with MACE or malignancy; however, this may relate to an underpowered analysis. Additionally, when the PON1 Q192R genotype was included in multivariate models with PON1 activities, the significance of PON1 enzyme activity was reduced, likely due to the strong association of this genotype with all 3 enzyme activities and known effects of multicollinearity on multivariate regression model outcomes. Further work evaluating the association of the PON1 Q192R polymorphism with safety outcomes in larger RA populations is warranted.

Increases in paraoxonase activity have been observed during treatment with tofacitinib,18 adalimumab (vs abatacept), and tocilizumab.42,43 The potential for MACE or malignancy risk reduction and improvements in RA disease activity through modulation of lipoprotein particles by RA therapies requires further evaluation. The PON1 locus has been associated with tumor necrosis factor inhibitor response in patients with RA,44 and it was demonstrated that overexpression of the human PON1 transgene in 2 RA mouse models reduced arthritic disease activity.31 Therefore, studies are warranted to assess whether PON1 activity or the PON1 Q192R polymorphism could predict responses to tofacitinib and other RA therapies.

This study is limited by its posthoc nature. Our statistical models had reduced power. Analysis of the PON1 Q192R polymorphism was restricted to the European cohort, and investigation in patients of other ethnic backgrounds is warranted. As these safety events were reported in an RCT setting, the association between paraoxonase activity levels and risk of MACE and malignancy in patients with RA receiving treatment in real-world practice remains unclear. Future studies should consider dietary effects, as antiinflammatory foods may affect PON1 activity or MACE and malignancy risk.45

In conclusion, this posthoc analysis demonstrated that higher paraoxonase activity over time was associated with a significantly reduced risk of MACE and malignancies (excluding NMSC) in tofacitinib-treated patients with moderate to severe RA. Further investigation of PON1 as a novel, functional lipid biomarker to assess MACE and malignancy risk in patients with RA is warranted.

ACKNOWLEDGMENT

The authors would like to thank the study patients and investigators. The authors would also like to thank Haiyun Fan, Pfizer, for assembling the clinical outcome data. Medical writing support, under the direction of the authors, was provided by Karen Irving, PhD, CMC Connect, a division of IPG Health Medical Communications, and Kirsteen Munn, PhD, on behalf of CMC Connect, and was funded by Pfizer in accordance with Good Publication Practice 2022 guidelines.46

Footnotes

Pfizer funded this posthoc analysis of data from tofacitinib clinical studies, as well as the original tofacitinib clinical studies included in this posthoc analysis. Publication of this article was not contingent upon approval by Pfizer.

CCS has received grants and/or research funding from AbbVie, BMS, CSL Behring, and Pfizer Inc; and has received consulting fees from AbbVie, BMS, Gilead Sciences, Pfizer Inc, and Regeneron-Sanofi. CH, SG, LS, and JA are employees and shareholders of Pfizer Inc. The remaining authors declare no conflicts of interest relevant to this article.

Accepted for publication July 4, 2023.Copyright © 2023 by the Journal of Rheumatology

This is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.

DATA AVAILABILITY

Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions, and exceptions, Pfizer may also provide access to the related individual deidentified participant data. See https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information.