Review Real-World Experience with Tofacitinib for the Treatment of Rheumatoid Arthritis

Review

Real-world experience with tofacitinib for the treatment of rheumatoid arthritis R. Caporali1, D. Zavaglia2

1Division of Rheumatology, University ABSTRACT Treatment persistence and adherence of Pavia, IRCCS S. Matteo Foundation, Objective. Oral targeted synthetic dis- to tofacitinib are good overall and sim- 2 Pavia, Italy; Medical Department, ease-modifying anti-rheumatic drugs ilar to those seen for bDMARDs. Pfizer Italy, Rome, Italy (DMARDs), including the Janus kinase Roberto Caporali, MD inhibitors tofacitinib and baricitinib, are Introduction Daniela Zavaglia, MD the latest addition to the therapeutic op- Disease modification is the mainstay Please address correspondence tions for rheumatoid arthritis (RA). To- of rheumatoid arthritis (RA) treatment and reprint requests to: facitinib 5 mg, twice daily, is approved Dr Roberto Caporali, (1). Several efficacious disease-modi- Division of Rheumatology, IRCCS for treatment, with or without metho- fying anti-rheumatic drugs (DMARDs) S. Matteo Foundation, trexate, of moderate to severe active RA are currently available, including con- Piazzale Golgi 2, in adults not adequately responding to, ventional synthetic (cs) DMARDs, 27100 Pavia, Italy. or not tolerating one or more DMARDs. biologic (b) DMARDs and the recently E-mail: [email protected] In this narrative review we aimed to developed class of targeted synthetic Received on April 17, 2018; accepted provide an overview of the real-world (ts) DMARDs (1). Currently, two tsD- in revised form on June 25, 2018. evidence for tofacitinib in RA. MARDs are available, tofacitinib and Clin Exp Rheumatol 2019; 37: 485-495. Methods. The literature was reviewed baricitinib, which are both inhibitors of © Copyright Clinical and up to March 2018 for studies regard- Janus kinase (JAK). Experimental Rheumatology 2019. ing the efficacy and safety of tofacitinib Tofacitinib is a small-molecule, oral for the treatment of RA. The focus was selective inhibitor of JAK1 and JAK3 Key words: disease-modifying mainly on real-world studies with impli- and, to a lesser extent, of JAK2 (2). anti-rheumatic drug, DMARD, cations for every day clinical practice. JAKs mediate signal-transduction ac- effectiveness, JAK-inhibitor, Results. The efficacy and safety of to- tivity initiated by the surface receptors real-life, post-marketing, real-world, facitinib have been comprehensively for several cytokines, thus playing a rheumatoid arthritis, targeted synthetic assessed in a wide programme of ran- key role in lymphocyte activation, pro- DMARD, tofacitinib domised controlled trials. Extensive liferation and function (2). Tofacitinib observational research on tofacitinib (5 mg twice daily, oral administration) in RA is also ongoing worldwide and was approved in November 2012 by a substantial body of post-marketing the US Food and Drug Administration real-world data from clinical prac- (FDA) and in March 2017 by the Eu- tice is becoming available. There was ropean Medicines Agency (EMA) for a degree of consistency across the treatment, with or without methotrexate real-world studies reviewed. Tofaci- (MTX), of moderate to severe active tinib tends to be used as monotherapy RA in adult patients who have respond- more frequently than bDMARDS and ed inadequately to, or do not tolerate appears to be effective without back- one or more DMARDs (3, 4). Tofaci- ground methotrexate. The data show tinib has been included by the European a manageable safety profile, with no League Against Rheumatism (EULAR) new safety signals and a discontinua- and by the American College of Rheu- tion rate from safety issues <10%. Pa- matology (ACR) among the treatments tients initiating tofacitinib usually have recommended for second- and later longer disease duration and have been lines in RA (1, 5). exposed to longer bDMARDs than pa- The efficacy and safety of tofacitinib, tients initiating a bDMARD. alone or combined with MTX, have Competing interests: R. Caporali received Conclusion. Real-world data are a key been extensively evaluated in a series an honorarium from Pfizer in connection component of the evidence support- of phase II and phase III randomised with the development of this manuscript. ing the effectiveness of this novel drug controlled trials (RCTs) of 6–24 months D. Zavaglia is a Pfizer employee. and are of interest to all stakeholders. duration (6-18), including the ORAL

Clinical and Experimental Rheumatology 2019 485 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia programme, consisting of a series of We will then review the most relevant ble for the trials, thus showing that in- large, phase III, randomised controlled data on the use and effectiveness of to- eligible patients nevertheless had some trials (RCTs) in a variety of settings (6- facitinib in routine clinical practice. benefit from the treatment. Similar re- 13). Two global long-term extension sults were reported by the analysis of studies, including 4967 RA patients Importance of real-world data the Finnish National Register for Bio- who had completed prior phase I, II, in rheumatoid arthritis logic Treatment including patients who or III studies of tofacitinib, have con- Following the introduction of bDMARDs started a TNFi between 2004 and 2014 firmed the sustained efficacy (>6 years) for RA, several national registries have (30). A recent study reviewed the eligi- and consistent safety profile (>8 years) been implemented to evaluate the long- bility criteria of 30 RCTs that included of tofacitinib administered as either term effectiveness and safety of these bDMARDs and tofacitinib and applied monotherapy or in combination with therapies. Registries have proven par- them to two observational cohorts of RA csDMARDs (19). ticularly valuable in RA (20, 27). For patients with more than 1500 patients RCTs have the most reliable study de- example, the current status of MTX as each (32). This study, as well, found that sign to determine the efficacy and safe- the gold standard first-line of disease- only a minority of patients in both co- ty of drugs, and occupy a top position modifying interventions has emerged horts, ranging from 3.7% to 7.1%, satis- in the hierarchy of evidence support- from long-term observational studies fied the RCT inclusion criteria. Two sys- ing the approval of novel therapeutic rather than from RCTs (25). In contrast tematic reviews of studies analysing the agents (20-22). In such trials, a number with early RCTs, which failed to dem- characteristics of RA patients participat- of inclusion and exclusion criteria are onstrate the superiority of MTX over ing in RCTs and observational studies used to select a clearly defined group of other csDMARDs, subsequent obser- of bDMARDs confirmed the existence patients who are likely to benefit from vational studies revealed that drug sur- of differences between the two groups the treatment under investigation. This vival was markedly longer with MTX (31). Compared with patients enrolled can, however, compromise the external than with other csDMARDs (25). More in RCTs, those from real-world obser- validity and reduce the generalisabil- recently, an observational study recog- vational studies had worse prognostic ity of the study findings to real-world nised the increased risk of opportunistic factors, including older age, longer dis- patients (20-22). Thus, the efficacy of infections associated with inhibitors of ease duration, and >1 prior DMARDs. a therapeutic intervention does not nec- tumour necrosis factor (TNFi), which According to the authors, data from se- essarily imply its ability to work un- did not emerge from the prior RCTs (28, lected patient populations participating der real-life conditions and to provide 29). The effectiveness of tuberculosis in RCTs may result in an overestimate benefits that are important for patients, prevention and screening before treat- of the effects in patients in real-world in other words, its effectiveness. The ment with TNFi was also demonstrated settings (31). incomplete overlap of efficacy and ef- by observational studies (25, 28). Overall, these findings have contrib- fectiveness is described in the literature With the purpose of investigating the ef- uted to the general consensus among as the “efficacy-effectiveness gap” (21, ficacy-effectiveness gap in the evidence regulatory authorities, decision-makers 23). Regulatory authorities and deci- supporting current RA therapies, several and clinicians about the urgent need for sion-makers increasingly regard real- studies have compared the characteris- more real-world observational studies, world evidence as a key component tics of patients enrolled in RCTs and as RA patients participating in RCTs of the body of evidence supporting the patients included in registries, which represent only a minority of those en- effectiveness of a therapeutic interven- have found substantial differences be- countered in routine clinical practice. tion (20, 24-26). Observational studies tween the two populations (30-33). evaluating the effects of treatment in Using data from the German biologics Real-world treatment of rheumatoid routine clinical practice and outside of register Rheumatoid Arthritis Observa- arthritis with tofacitinib the highly controlled environment of tion of Biologic Therapy (RABBIT), Real-world evidence concerning tofac- RCTs are the main source of real-world it was investigated how many RA pa- itinib in RA has just begun to emerge. data (24). tients treated with a TNFi would have The most substantial body of real- Extensive observational research on to- been eligible for the RCTs that led to the world evidence comes from clinical facitinib in RA is ongoing worldwide, approval of the TNFi (33). The study practice in the US, where tofacitinib and a significant body of real-world found that only a minority of the pa- has been available since 2012. Other data from clinical practice is now be- tients (<35%) in the RABBIT register countries contributing to the body of coming available. In this narrative re- would have been eligible for the trials. real-world evidence include Japan, Tai- view, we provide an overview of the re- In these patients, ACR20 and ACR50 wan, Switzerland, Latin America and al-world evidence for tofacitinib in RA response rates were comparable with Russia. Source of real-world data are as of March 2018. We will first discuss those reported in published trials. In the administrative claim databases, clinical examples of the efficacy-effectiveness register patients considered ineligible databases, registries, including RA reg- gap in RA and the importance of real- for the trials, ACR response rates were istries, and national pharmacovigilance world evidence in bridging this gap, and lower, although absolute improvements programmes. Understanding treatment in complementing the data from RCTs. were similar to those in patients eligi- patterns and patient outcomes and

486 Clinical and Experimental Rheumatology 2019 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia characterising the safety profile of to- (61%) used it as monotherapy; the prev- (68.8% vs. 38.5%, p=0.002) and to have facitinib are the principal objectives of alence of monotherapy by line of thera- received previous bDMARDs (85.4% vs. real-world observations. py was 47%, 58% and 63% for second-, 55.4%, p=0.001). The number of previ- third- and fourth-line therapy, respec- ous bDMARDs used by non-responders Patient characteristics, treatment tively. Among patients initiating a TNFi was twice that used by responders (2.2 patterns, and outcomes (n=7976), 2531 (31%) used monother- vs. 1.1, p<0.001). Concomitant MTX Relevant real-world data are sum- apy; the prevalence of monotherapy by use was reported less frequent in non-re- marised in Table I. A study using US line of therapy was, respectively, 21%, sponders than in responders (60.4% vs. administrative claims data compared 36%, and 42%. In the matched popula- 81.5%, p=0.019). Multivariate logistic patient characteristics, treatment pat- tions, across outcome measures, TNFi regression analysis identified previous terns and costs in RA patients receiving combination therapy was more effective use of bDMARDs as a strong risk fac- tofacitinib or common bDMARDs fol- than TNFi monotherapy in second-line tor for the failure to achieve a CDAI50 lowing one previous bDMARD, from [rates of low disease activity (LDA), response [odds ratio (OR) 4.48, 95% CI 2012 to 2014 (34). Overall, 392 patients defined as Clinical Disease Activity In- 1.73–11.63, p=0.002]. The study also initiated tofacitinib, 178 adalimumab, dex (CDAI) ≤10 at 6 months, 55.6% vs. compared baseline characteristics and 118 etanercept and 191 abatacept. To- 47.1%], with differences diminishing in therapeutic outcomes between biologic- facitinib was more commonly used as third- and fourth-line. Tofacitinib com- experienced (n=77) and biologic-naïve monotherapy than bDMARDs; patients bination therapy with MTX was similar (n=36) patients. Biologic-experienced receiving tofacitinib monotherapy were to monotherapy in the third- and fourth- patients were more likely than biologic- more likely to remain on monotherapy line combined (LDA 35.2% vs. 31.1%). naïve patients to have advanced-stage during the 12-month follow-up. Per- Tofacitinib monotherapy was similar to RA (66.2% vs. 19.4%, p<0.001) and sistence in treatment and adherence TNFi combination therapy in the third- longer disease duration (12.8 years vs. were comparable with tofacitinib and and fourth-line combined (LDA 33.6% 9.6 years, p=0.004). Tofacitinib was bDMARDs. Tofacitinib was associated vs. 37.5%). prescribed in combination with MTX with lower adjusted mean RA-related Another study based on the Corrona less frequently in biologic-experienced total costs versus adalimumab, etaner- registry, which focused on safety, than in biologic-naïve patients (59.7% cept and abatacept. Tofacitinib patients showed that patients initiating tofaci- vs. 100%, p<0.001). At 6 months, bio- had a greater previous use of bD- tinib had longer disease duration than logic-experienced patients had higher MARDs than the bDMARD cohort. A those initiating bDMARDs and csD- disease activity than biologic-naïve subsequent study using the same source MARDS (respectively, 13.4, 10.1 and patients (mean CDAI scores, 11.4 vs. of real-world data, conducted between 4.8 years); patients initiating tofacitinib 4.8, p=0.001). Biologic-experienced 2014 and 2017 and focused on patients had also been exposed to a higher mean patients were significantly less likely with no pre-index use of bDMARDs and number of previous TNFi (1.8) or bD- to achieve CDAI50 (46.8% vs. 80.6%, tofacitinib, found that tofacitinib was MARDs (2.8) than patients initiating a p=0.001), CDAI70 (31.2% vs. 69.4%, more frequently prescribed as mono- bDMARD (mean number of previously p<0.001) and CDAI85 (22.1% vs. therapy than adalimumab or etanercept used biologics: 1.1 TNFi and 1.4 non- 52.8%, p=0.002) responses than biolog- and confirmed the other main findings TNFi bDMARDs) (37). ic-naïve patients; they had significantly of the previous study (35). In a prospective, observational Japanese lower remission rates (11.7% vs. 41.1%, The Corrona registry, founded in 2001, study, 113 patients who initiated treat- p=0.001), higher drop-out rates (35.1% is one of the largest US-based RA obser- ment with tofacitinib alone or in com- vs. 13.9%, p=0.025) and rates of drop- vational registries (27). A study analys- bination with MTX were followed for out due to lack or loss of efficacy (20.8% ing the data from this registry investigat- 6 months (38). RA disease activity was vs. 2.8%, p=0.011) than biologic-naïve ed the use of TNFi and tofacitinib mon- evaluated based on the CDAI at sev- patients. otherapy or combination therapy with eral time points during the observation Another 6-month observational study MTX by line of therapy in US clinical period. Within 6 months of tofacitinib in Japan enrolled 70 consecutive RA practice (36). The primary objective was treatment, 57.5% of patients achieved patients who initiated tofacitinib at a to compare the prevalence and effective- a CDAI50 response and maintained it rheumatology centre between 2013 and ness of TNFi monotherapy versus TNFi for 6 months (responders); 42.5% had 2016 (39). Clinical disease activity was combination therapy; the secondary ob- no CDAI50 response at 6 months (non- assessed using the DAS28-ESR score, jectives were to compare the prevalence responders). Sixteen patients (14.2%) simplified disease activity index (SDAI) and effectiveness of tofacitinib mono- completed the 6-month treatment with and CDAI. Overall, 68.6% of patients therapy versus tofacitinib combination tofacitinib without achieving a CDAI50 used tofacitinib in combination with therapy, as well as versus TNFi combi- response, 17 patients (15.0%) dropped MTX, and the majority had previously nation therapy. The study included pa- out for lack or loss of efficacy and 4 pa- used bDMARDs, while 31.4% were tients initiating tofacitinib or TNFi and tients (3.5%) due to adverse events. Non- biologic-naïve. At 6 months, 82.9% of followed for 6 months. Among patients responders were more likely than re- patients were still taking tofacitinib; 7 who initiated tofacitinib (n=555), 338 sponders to have advanced stage disease patients (10.0%) discontinued the drug

Clinical and Experimental Rheumatology 2019 487 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia vs. TOFA adherence in biologic-experienced NA vs. biologic-naïve pts Comparable Similar crude Trt persistence and Trt Comparable cohorts NA 15% drop-out rate for Higher drop-out rate persistence and NA persistence among NA adherence 82.9% trt persistence at 6 mths persistence and inefficacy 3.5% adherence among the 3 groups Trt persistence and Trt adherence among AEs drop-out rate for cohorts After adjustment for confounders, higher confounders, higher risk of discontinuation

57% trt persistence with TNFi

Mean 112 days, to Mean 112 any-cause stop NA

with vs. with -line in 25% of nd . bDMARD vs . bDMARD vs. biologic- pts

TOFA TOFA in TOFA TOFA Safety data in Table II. Table Safety data in 12-mth pre-index bDMARD use vs. combination therapy and Main findings stayed on monotherapy costs lower TOFA bDMARDS. Higher RA-related 12-mth pre-index Total TOFA. frequent with TNFi rates with Higher LDA TNFi N. of prior bDMARDs and II Table Safety data, see 57.5% CDAI50 response rate at response and remission rates in Prior bDMARD use strongly 77.6% in TOFA group vs. TOFA 77.6% in TNFi combination therapy Effective disease activity control Effective during short-term follow-up pts; after 1 vs. ≥2 bDMARD in 20% Disease duration at baseline longer Rapid significant decreases in associated with non-response disease activity and disability over costs and worse comorbidity score RA-related post-index costs lower combination therapy vs. monotherapy. 6 mths biologic-naïve 47.6%-59.6% in bDMARD cohort. (up to 6 mths) vs. 55% in TOFA vs. bDMARD and TOFA in Main findings Similar LDA rates for TOFA mono- TOFA rates for Similar LDA 6 mths, also with monotherapy and with TOFA. Monotherapy more TOFA. with Significantly higher CDAI50 More pts on TOFA started and TOFA More pts on Pts with ≥2 prior bDMARDs, older csDMARD initiators. after switching from TOCI after switching from age, higher BMI, or Significant and inverse association Significant and inverse association disability more likely to receive of prior bDMARD use and LDA of prior bDMARD use and LDA TOFA vs. TNFi TOFA Safety data, see Table II Table Safety data, see 63.8% vs. 53.4% pts achieved LDA or remission after 62.0 vs. 65.3 days More prior bDMARDs and higher More prior bDMARDs and higher TOFA initiated as 2 TOFA BMI associated with increased BMI associated with increased II Table Safety data, see II Table Safety data, see -line in ca. 60% vs. 38% nd

-line in 44%, after ≥1 -line in 44%, after ≥1 nd Mostly in combination with csDMARD (87%) As 2 pattern As monotherapy ≥2 bDMARDs in vs. ≥2 bDMARDs in 18% Tofacitinib treatment Tofacitinib MTX) used TOFA (21.7%) ADA than Monotherapy rate with TNFi: 31% As monotherapy or in more frequent in (53.1%) or in After 1 vs. ≥ 2 bDMARDs Combination therapy with MTX (70%) or other other or (70%) MTX with used as monotherapy NA Mostly in combination with MTX (68.6%) but significantly more as and ETN (26.5%) 61% TOFA: combination with MTX biologic-naïve pts combination with in 19.3% vs. 21.0% csDMARD (10%); as more often than TNFi more often than

Tofacitinib treatment Tofacitinib pattern Monotherapy rate with also as monotherapy monotherapy (34.2%), Combination with MTX csDMARDs (mostly monotherapy (20%) TOFA preferentially TOFA

given to pts with more

prior bDMARDs, older

age, and longer disease

86% of pts initiated after ≥1 TOFA duration with csDMARD (59%) TOFA and non-TNFi TOFA bDMARD Monotherapy (41%) 2

TOFA + TNFi vs. TOFA

Evaluate safety profile of TOFA

To compare characteristics, To Objectives trt initiation compare characteristics, To bDMARDs over 12 mths compare use and To Primary: combination therapy compare use To Secondary: of compare effectiveness To trt patterns, and costs of

Evaluate effectiveness and Evaluate effectiveness safety of TOFA safety of TOFA and to identify factors TOFA To evaluate the safety of To evaluate the effectiveness To and safety of TOFA TOFA and safety of

trt patterns, and costs in pts post-index trt initiation TNFi of effectiveness in biologic-naïve vs. TOFA and effectiveness of TOFA TOFA of and effectiveness monotherapy TOFA vs. bDMARDs over TOFA

determining the choice of TOFA Objectives

TNFi monotherapy vs. receiving TOFA vs. TOFA receiving biologic-experienced pts 12 mths pre- and post-index

TOFA or bDMARDs TOFA

To compare drug retention To

rates with TOFA, TNFi, and TOFA, rates with

non-TNFi bDMARDs

safety

To characterize pts initiating To Primary: to analyse TOFA Primary: to analyse Secondary: to determine rate Evaluate the safety profile

of TOFA of and time to LDA and remission and time to LDA

Retrospective analysis

Retrospective cohort Design and patients Retrospective cohort TOFA or Retrospective analysis Pts 6-mth prospective analysis of pts with

Retrospective analysis

in pts initiating TOFA or TOFA in pts initiating Interim analysis of a 5-yr. 6-mth prospective observational study in pts

analysis of biologic-naïve TOFA, TNFi or initiating a observational study in pts

1 prior bDMARD initiating

bDMARDs prospective observational Design and patients initiating TOFA initiating with 6-mth follow-up pts initiating a bDMARD TOFA initiating a bDMARD or TOFA TOFA a bDMARD or

study

study in pts initiating TOFA study in pts initiating

Observational cohort study Prospective observational 1.7 years mean follow-up Retrospective analysis in

pts initiating TOFA TOFA pts initiating

392, TOFA; TOFA; 392, 62, TOFA 62, n TOFA; 184, TOFA; 555, TOFA 113, 178, ADA; 178,

928, TNFi; 928, 760, TOFA; TOFA; 760,

1771, ADA; 1771, TNFi 7976, (36 biologic-naïve, 77

118, ETN; 118,

120, TOFA 120,

692, non-TNFi 4628, bDMARD; 70, TOFA 70, n

1472, ETN biologic-experienced) 191, ABA 191,

1328, csDMARD

TOFA; 376,

58, TOFA 58, TOFA 288,

US administrative claims Medical records from rheumatology centres in Argentina (2014-2016) Argentina Data source US administrative claims US Corrona Rheumatoid Rheumatology centre in database (2012-2014)

Management registry US Corrona Rheumatoid

database (2014-2017) Arthritis registry Japan (2013-2016)

research programme

(2013-2016) Arthritis registry Rheumatology centre in Japan (2013-2016) Data source

(2001-2016)

Russian REMARCA

(2012-2016)

Switzerland (2013-2017)

Swiss Clinical Quality

America St. Gallen Cohort, Rheumatology centres from 6 countries in Latin et al . 2017 et al . 2016 et al . 2017 Patient characteristics, treatment patterns and outcomes in real-world studies with tofacitinib. et al . 2016 et al . 2017 et al . 2016 et al . 2016 and et al . 2018 et al . 2017 et al . 2017 et al . 2017 et al . 2017

Table I. Table Study Harnett Smith Reed initiators Mori experienced

Finckh Kavanaugh Sansinanea

higher

Iwamoto Study

with

Karateev

ABA: abatacept; ADA: adalimumab; AE: adverse event; bDMARD: biologic DMARD; BMI: body mass index; CDAI: clinical disease activity index; csDMARD: conventional synthetic DMARD; DMARD: disease-modifying anti-rheumatic AE: adverse event; bDMARD: biologic DMARD; BMI: body mass index; CDAI: clinical disease activity csDMARD: conventional synthetic DMARD: disease-modifying anti-rheumatic ADA: adalimumab; ABA: abatacept; tofacitinib. TOFA: tocilizumab; TOCI: TNFi: tumour necrosis factor inhibitor; drug; ETN: etanercept; LDA: low disease activity; MTX: methotrexate; mth: month; N: number; RA: rheumatoid arthritis;

achievement

Kyburz

Müller discontinuation Schneeberger

488 Clinical and Experimental Rheumatology 2019 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia vs. TOFA 57% trt persistence NA adherence Mean 112 days, to Mean 112 NA with TNFi adherence Comparable 82.9% trt persistence NA at 6 mths Similar crude Trt persistence and Trt persistence and persistence and Trt After adjustment for any-cause stop persistence among adherence among confounders, higher the 3 groups cohorts risk of discontinuation Comparable persistence and adherence among cohorts

NA 15% drop-out rate for inefficacy 3.5% drop-out rate for AEs drop-out rate for Higher drop-out rate in biologic-experienced vs. biologic-naïve pts

-line in 25% of nd . bDMARD vs . bDMARD pts

TOFA TOFA in vs. TNFi TOFA TOFA Effective disease activity control Effective during short-term follow-up 63.8% vs. 53.4% pts achieved LDA or remission after 62.0 vs. 65.3 days Safety data in Table II. Table Safety data in 12-mth pre-index bDMARD use Rapid significant decreases in II Table Safety data, see disease activity and disability over (up to 6 mths) Main findings group vs. TOFA 77.6% in Main findings 6 mths, also with monotherapy and initiated as 2 TOFA Pts with ≥2 prior bDMARDs, older More prior bDMARDs and higher II Table Safety data, see pts; after 1 vs. ≥2 bDMARD in 20% after switching from TOCI after switching from BMI associated with increased 47.6%-59.6% in bDMARD cohort. age, higher BMI, or vs. 55% Significant and inverse association Significant and inverse association More pts on TOFA started and TOFA More pts on disability more likely to receive of prior bDMARD use and LDA of prior bDMARD use and LDA with stayed on monotherapy vs. with TOFA Safety data, see Table II Table Safety data, see bDMARDS. TOFA costs lower TOFA bDMARDS. Higher RA-related 12-mth pre-index costs and worse comorbidity score with TOFA. Monotherapy more TOFA. with Total TOFA. frequent with RA-related post-index costs lower RA-related post-index costs lower Higher LDA rates with TNFi rates with Higher LDA combination therapy vs. monotherapy. Similar LDA rates for TOFA mono- TOFA rates for Similar LDA vs. combination therapy and TNFi combination therapy TNFi combination therapy Disease duration at baseline longer in TOFA vs. bDMARD and TOFA in csDMARD initiators. N. of prior bDMARDs and TNFi TNFi N. of prior bDMARDs and Safety data, see Table II Table Safety data, see 57.5% CDAI50 response rate at 6 mths Significantly higher CDAI50 response and remission rates in biologic-naïve vs. biologic- Prior bDMARD use strongly Prior bDMARD use strongly associated with non-response vs. 21.0% vs. ≥ 2 bDMARDs -line in ca. 60% vs. 38% nd

-line in 44%, after ≥1 -line in 44%, after ≥1 nd Combination therapy with MTX (70%) or other other or (70%) MTX with 86% of pts initiated after ≥1 TOFA with csDMARD (59%) Mostly in combination with csDMARD (87%) As 2 As monotherapy pattern 18% Mostly in combination Monotherapy (41%) with MTX (68.6%) but 2 ≥2 bDMARDs in vs. ≥2 bDMARDs in After 1 in 19.3% csDMARD (10%); as Tofacitinib treatment Tofacitinib pattern (53.1%) or in treatment Tofacitinib also as monotherapy and non-TNFi TOFA preferentially TOFA duration bDMARD used as monotherapy monotherapy (20%) combination with given to pts with more more often than TNFi more often than csDMARDs (mostly prior bDMARDs, older

MTX) age, and longer disease

used TOFA significantly more as monotherapy (34.2%), than ADA (21.7%) ADA than and ETN (26.5%) Monotherapy rate with TOFA: 61% TOFA: Monotherapy rate with TNFi: 31% NA

As monotherapy or in combination with MTX Combination with MTX more frequent in biologic-naïve pts

TOFA + TNFi vs. TOFA

Evaluate effectiveness and Evaluate effectiveness safety of TOFA safety of safety Evaluate safety profile of TOFA

To compare characteristics, To evaluate the effectiveness To TOFA Primary: to analyse Evaluate the safety profile and safety of TOFA TOFA and safety of TOFA of trt patterns, and costs of Objectives Objectives characterize pts initiating To or bDMARDs TOFA Secondary: to determine rate TOFA and to identify factors TOFA

TOFA vs. bDMARDs over TOFA compare drug retention To and remission and time to LDA determining the choice of

12 mths pre- and post-index TNFi, and TOFA, rates with trt initiation non-TNFi bDMARDs

compare characteristics, To trt patterns, and costs in pts receiving TOFA vs. TOFA receiving bDMARDs over 12 mths post-index trt initiation Primary: To compare use and To Primary: effectiveness of TNFi of effectiveness TNFi monotherapy vs. combination therapy Secondary: To compare use To Secondary: and effectiveness of TOFA TOFA of and effectiveness monotherapy To evaluate the safety of To TOFA

To compare effectiveness of compare effectiveness To TOFA in biologic-naïve vs. TOFA biologic-experienced pts

study in pts initiating TOFA study in pts initiating Retrospective analysis

Retrospective analysis

Retrospective cohort 6-mth prospective Prospective observational Retrospective analysis in observational study in pts TOFA pts initiating analysis of pts with Design and patients Design and patients TOFA initiating Observational cohort study 1.7 years mean follow-up in pts initiating TOFA or TOFA in pts initiating

1 prior bDMARD initiating bDMARDs a bDMARD or TOFA TOFA a bDMARD or

Retrospective cohort analysis of biologic-naïve pts initiating a bDMARD or TOFA TOFA or

Retrospective analysis Pts TOFA, TNFi or initiating a with 6-mth follow-up

Interim analysis of a 5-yr. prospective observational study

6-mth prospective observational study in pts initiating TOFA initiating

120, TOFA 120, 392, TOFA; TOFA; 392, 62, TOFA 62,

178, ADA; 178, n n TOFA 70, TOFA; 376, TOFA 58, TOFA 288, 928, TNFi; 928,

118, ETN; 118, 692, non-TNFi

191, ABA 191,

TOFA; 184, 1771, ADA; 1771, 1472, ETN

555, TOFA; TOFA; 555, 7976, TNFi 7976,

760, TOFA; TOFA; 760, 4628, bDMARD; 1328, csDMARD

113, TOFA 113, (36 biologic-naïve, 77 biologic-experienced)

Switzerland (2013-2017) research programme US administrative claims America Medical records from rheumatology centres in Argentina (2014-2016) Argentina Russian REMARCA database (2012-2014) Data source Data source Rheumatology centre Swiss Clinical Quality St. Gallen Cohort, Rheumatology centres in Japan (2013-2016) from 6 countries in Latin Management registry

(2013-2016)

US administrative claims database (2014-2017)

US Corrona Rheumatoid Arthritis registry (2001-2016)

US Corrona Rheumatoid Arthritis registry (2012-2016)

Rheumatology centre in Japan (2013-2016)

et al . 2017 et al . 2016 et al . 2017 et al . 2016 et al . 2017 et al . 2016 et al . 2016 and et al . 2018 et al . 2017 et al . 2017 et al . 2017 et al . 2017

Karateev

Harnett Table I. Patient characteristics, treatment patterns and outcomes in real-world studies with tofacitinib. Table Study Study Iwamoto Kyburz Müller Schneeberger

Finckh Sansinanea

achievement

Smith

with Reed

Kavanaugh

higher initiators Mori

experienced

Clinical and Experimental Rheumatology 2019 489 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia

NA NA NA NA NA NA NA 3.2% 3.8% 3.5% 5.7% 9.6% 34.5% rate due to AEs rate due to rate due to AEs rate due to Discontinuation Discontinuation Discontinuation Discontinuation

vs. reference

increase in liver enzyme (1), facial paralysis (1) tachycardia 6 AEs: HZ (2), upper airway infection (1), transient 6 (1), pancreatitis (1) papules (1), rash (1), sinusitis (1), headache (1), folliculitis papules (1), rash sinusitis headache folliculitis Other reported AEs: diarrhoea (1), infection (1), mouth AEs: diarrhoea (1), infection mouth Other reported event. Most reported AE: abdominal discomfort (3) event. Most reported 17.9% with ≥1 AE; events mostly mild-moderate (1 severe 17.9% with ≥1 skin infections (5), HZ (4), urinary tract (4) 38 AEs reported 38 Most frequent events: upper respiratory infections (11), creatinine (1); lymphocyte count <1000/ml (4) enzymes (12); decrease in Hb levels >10% (7); abnormal enzymes (12); decrease in Hb levels >10% (7); abnormal Moderate alterations in some laboratory tests: liver Moderate alterations in some laboratory tests: liver myalgia (1) pneumonia (2), blue toe syndrome (1), thoracic pain (1), pneumonia (2), blue toe syndrome (1), thoracic pain AEs leading to discontinuation: GI events (12), flare (3), No HBV reactivation in other pts No HBV 2 of the 4 carriers with no prophylaxis had HBV reactivation 2 of the 4 carriers with no prophylaxis had HBV antiviral prophylaxis during TOFA trt TOFA antiviral prophylaxis during resolved infection). 4 of the 6 HBV carriers did not receive carriers did not receive resolved infection). 4 of the 6 HBV Stable RR of lymphoma with increasing exposure to TOFA Stable RR of lymphoma with increasing exposure to disorders; 0.12, hepatobiliary disorders Prior HBV infection in 81 pts (69.8%; 6 carriers, 75 with Prior HBV 0.60 respiratory disorders; 0.45, neoplasms; 0.43, cardiac 0.60 respiratory disorders; 0.45, neoplasms; 0.43, cardiac RR/100 py of SAEs: 2.57, infections; 0.91, GI disorders; RR/100 py of SAEs: 2.57, infections; 0.91, GI disorders; trt-related cancer risk are still imprecise Estimates of RA see Table I Table see events (0.4%) diarrhoea, 5.8%; arthralgia, 5.2% and 10.71-22.73 for other trts AE 20% reported ≥1 AE (15.7%) Infections (HZ) were the most common 25,417 AEs reported; 4352 SAEs (17.1%); 102 fatal 25,417 fatigue, 6.0%; nausea, condition aggravated, 5.9%; ABA: 2.01 (95% CI, 1.40-2.88) vs. TOFA HR of HZ for malignancies among the three groups and 3.40-7.50 TOFA for IR/1000 py of solid tumours 11.38 TOFA IR/1000 py of lymphoproliferative disease, 0 for non-TNFi bDMARDs, 1.10 liver enzymes (1.7%). SAEs 7.7%, serious infection 3.5%, Main findings 2.00-2.53; reference group (no treatment), 1.96

Main findings 34,223 py post-marketing exposure 13.2%; headache, 9.0%; pain, 6.4%; Drug ineffective, vs. 1.95-2.71/100 3.87/100 py, TOFA, Crude IR of HZ for py for bDMARDs events and Comparable IR of serious infection, CV 1456 incident cancers (812 excluding NMSC) for other trts Adjusted HR for serious infection TNFi, 1.13; 1.81; csDMARDs, 0.82; TOFA, group: AE in 33.5%; 1241 py post-marketing exposure; ≥1 most frequent: infections (12.7%), HZ (3.3%), abnormal malignancies 0.4%, died 0.6% IR/100 py of serious infection: TOFA, 3.7; other groups, TOFA, IR/100 py of serious infection:

Evaluate the safety profile of TOFA Evaluate the safety profile of

in pts treated with TOFA in pts treated with

Evaluate the risk of HBV reactivation HBV of risk the Evaluate

experienced pts TOFA of Compare effectiveness in biologic-naïve vs. biologic- Evaluate efficacy and effectiveness TOFA of safety profile of TOFA safety profile of vs. bDMARDs bDMARDs, TOFA, in pts initiating and TOFA trt with bDMARDs or Objective of serious infections among trt or csDMARDs Objective Determine the post-marketing Evaluate the risk of HZ and HSV TOFA infection associated with AEs of interest Evaluate the IR of Evaluate the association between incident cancer groups TOFA of Evaluate the effectiveness

Compare retrospectively incidence

followed-up for up to 21 mths Pts treated with TOFA (62) and TOFA Pts treated with

of 33 weeks Pts (56) treated with TOFA for a mean TOFA Pts (56) treated with

Pts (288) initiating TOFA and TOFA Pts (288) initiating followed-up for a median of 22 mths

mean 1.7 yr Consecutive pts (58) followed-up for a

(retrospective)

Taiwanese pts (116) treated with TOFA treated with pts (116) Taiwanese

(113) TOFA Pts initiating

TOFA Consecutive pts (70) initiating and followed-up for 6 mths (6967) TOCI (4628), or csDMARD (1328) TNFi; 1621, non- 4419, TOFA; 222, (n) Patients with RA (21,832) and prescribed csDMARDS Patients with RA (n) Patients with RA TOFA Pts worldwide initiating TNFi (2526), TOFA Pts initiating (5078), (12,305), RIT ABA 42,850), (760), bDMARD TOFA Pts initiating without prevalent cancer; Pts (11,582) TNFi bDMARDs; 5320 csDMARDs TNFi (61.2%), non-TNFi (24.7%), (0.8%) TOFA bDMARDs (13.3%), or TOFA Consecutive pts (2882) initiating

Pts previously treated with MTX

Databases ® (2014-2016) Medical records from rheumatology centres inArgentina

(2015-2016) Rheumatology centre in Colombia

Rheumatology centres from America 6 countries in Latin

(2013-2017) St. Gallen Cohort, Switzerland

(2015-2017)

Taiwan Medical centre in

Rheumatology centre in Japan (2013-1016) (2012-ongoing; 3-yr. interim data) (2012-ongoing; 3-yr. Rheumatology centre in Japan (2013-2016) surveillance safety database registry Rheumatic Diseases (1998-2015) (6-mth analysis) Data source (2010-2014) Data source post-marketing Manufacturer’s US health plans (2010-2014) Arthritis US Corrona Rheumatoid interim analysis) (2012-2016; 3-yr. US National Data Bank for US MarketScan Japanese post-marketing surveillance data et al . 2017 Summary of safety data from real-world studies with tofacitinib. et al . 2016 et al . 2017 et al . 2017 et al . 2017 et al . 2016 et al . 2017 et al . 2017 et al . 2017 et al . 2018 et al . 2016 et al . 2017 et al . 2017 ABA: abatacept; AE: adverse event; bDMARD: biologic DMARD; CI: confidence interval; csDMARD: conventional synthetic DMARD; CV: cardiovascular; DMARD: disease-modifying anti-rheumatic drug; GI: gastrointestinal; HBV: hepa - HBV: gastrointestinal; GI: drug; anti-rheumatic disease-modifying DMARD: cardiovascular; CV: DMARD; synthetic conventional csDMARD: interval; confidence CI: DMARD; biologic bDMARD: event; adverse AE: abatacept; ABA: rituximab; RR: reporting herpes simplex virus; IR: incidence rate; MTX: methotrexate; mth: month; NMSC: non-myeloma skin cancer; py: patient-year; RA: rheumatoid arthritis; RIT: titis B virus; HR: hazard ratio; HZ: herpes zoster; HSV: tofacitinib. TOFA: tocilizumab; TOCI: TNFi: tumour necrosis factor inhibitor; rate;

Sansinanea

Malpica

Schneeberger Schneeberger

Müller

Chen

Mori

Iwamoto Study

Table II. Table Study Cohen Curtis Kavanaugh England Machado Mimori

490 Clinical and Experimental Rheumatology 2019 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia

- NA NA NA NA NA NA NA NA 9.6% 3.5% 3.8% 3.2% 5.7% 34.5% rate due to AEs rate due to rate due to AEs rate due to Discontinuation Discontinuation Discontinuation

Estimates of RA trt-related cancer risk are still imprecise Estimates of RA and 10.71-22.73 for other trts IR/1000 py of lymphoproliferative disease, 0 for TOFA TOFA IR/1000 py of lymphoproliferative disease, 0 for for other trts IR/1000 py of solid tumours 11.38 for TOFA and 3.40-7.50 TOFA for IR/1000 py of solid tumours 11.38 1456 incident cancers (812 excluding NMSC) malignancies among the three groups Comparable IR of serious infection, CV events and Comparable IR of serious infection, CV HR of HZ for TOFA vs. ABA: 2.01 (95% CI, 1.40-2.88) vs. TOFA HR of HZ for py for bDMARDs py for bDMARDs Crude IR of HZ for TOFA, 3.87/100 py, vs. 1.95-2.71/100 3.87/100 py, TOFA, Crude IR of HZ for Stable RR of lymphoma with increasing exposure to TOFA Stable RR of lymphoma with increasing exposure to disorders; 0.12, hepatobiliary disorders 0.60 respiratory disorders; 0.45, neoplasms; 0.43, cardiac 0.60 respiratory disorders; 0.45, neoplasms; 0.43, cardiac RR/100 py of SAEs: 2.57, infections; 0.91, GI disorders; RR/100 py of SAEs: 2.57, infections; 0.91, GI disorders; malignancies 0.4%, died 0.6% diarrhoea, 5.8%; arthralgia, 5.2% liver enzymes (1.7%). SAEs 7.7%, serious infection 3.5%, liver enzymes (1.7%). SAEs 7.7%, serious infection 3.5%, fatigue, 6.0%; nausea, 6.0%; condition aggravated, 5.9%; fatigue, 6.0%; nausea, condition aggravated, 5.9%; 1241 py post-marketing exposure; ≥1 AE in 33.5%; 1241 py post-marketing exposure; ≥1 most frequent: infections (12.7%), HZ (3.3%), abnormal Drug ineffective, 13.2%; headache, 9.0%; pain, 6.4%; 13.2%; headache, 9.0%; pain, 6.4%; Drug ineffective, non-TNFi bDMARDs, 1.10 events (0.4%) creatinine (1); lymphocyte count <1000/ml (4) group: TOFA, 1.81; csDMARDs, 0.82; TNFi, 1.13; TNFi, 1.13; 1.81; csDMARDs, 0.82; TOFA, group: 25,417 AEs reported; 4352 SAEs (17.1%); 102 fatal 25,417 reactivation in other pts No HBV enzymes (12); decrease in Hb levels >10% (7); abnormal (1), pancreatitis (1) resolved infection). 4 of the 6 HBV carriers did not receive carriers did not receive resolved infection). 4 of the 6 HBV Adjusted HR for serious infection vs. reference

pneumonia (2), blue toe syndrome (1), thoracic pain (1), pneumonia (2), blue toe syndrome (1), thoracic pain 34,223 py post-marketing exposure reactivation 2 of the 4 carriers with no prophylaxis had HBV Moderate alterations in some laboratory tests: liver papules (1), rash sinusitis headache folliculitis see Table I Table see 2.00-2.53; reference group (no treatment), 1.96 increase in liver enzyme (1), facial paralysis (1) tachycardia AEs leading to discontinuation: GI events (12), flare (3), AEs reported 38 Most frequent events: upper respiratory infections (11), AE: abdominal discomfort (3) event. Most reported AEs: HZ (2), upper airway infection (1), transient 6 Main findings Main findings AE 20% reported ≥1 AE (15.7%) Infections (HZ) were the most common infection in 81 pts (69.8%; 6 carriers, 75 with Prior HBV trt TOFA antiviral prophylaxis during myalgia (1) skin infections (5), HZ (4), urinary tract (4) AEs: diarrhoea (1), infection mouth Other reported IR/100 py of serious infection: TOFA, 3.7; other groups, TOFA, IR/100 py of serious infection: 17.9% with ≥1 AE; events mostly mild-moderate (1 severe 17.9% with ≥1

incident cancer trt with bDMARDs or TOFA and TOFA trt with bDMARDs or Evaluate the association between or csDMARDs in pts initiating TOFA, bDMARDs, TOFA, in pts initiating Evaluate the IR of AEs of interest Evaluate the IR of vs. bDMARDs infection associated with TOFA infection associated with Evaluate the risk of HZ and HSV

Evaluate the effectiveness of TOFA of Evaluate the effectiveness

safety profile of TOFA safety profile of groups

Determine the post-marketing TOFA of Compare effectiveness TOFA in pts treated with in biologic-naïve vs. biologic- of serious infections among trt

Objective Objective Evaluate efficacy and effectiveness experienced pts reactivation HBV of risk the Evaluate TOFA Evaluate the safety profile of TOFA Evaluate the safety profile of TOFA Evaluate the safety profile of of TOFA of Compare retrospectively incidence Evaluate the safety profile of TOFA Evaluate the safety profile of

TNFi bDMARDs; 5320 csDMARDs 222, TOFA; 4419, TNFi; 1621, non- 4419, TOFA; 222, Pts (11,582) without prevalent cancer; Pts (11,582) (4628), or csDMARD (1328) Pts initiating TOFA (760), bDMARD TOFA Pts initiating TOCI (6967) TOCI 42,850), ABA (12,305), RIT (5078), (12,305), RIT ABA 42,850), Pts initiating TOFA (2526), TNFi (2526), TOFA Pts initiating

Consecutive pts (2882) initiating TOFA Consecutive pts (2882) initiating

bDMARDs (13.3%), or TOFA (0.8%) TOFA bDMARDs (13.3%), or

(24.7%), TNFi (61.2%), non-TNFi (24.7%), mean 1.7 yr TOFA Pts worldwide initiating (113) TOFA Pts initiating (retrospective)

(21,832) and prescribed csDMARDS followed-up for up to 21 mths followed-up for a median of 22 mths of 33 weeks Patients with RA (n) Patients with RA (n) Patients with RA TOFA Consecutive pts (70) initiating TOFA treated with pts (116) Taiwanese Consecutive pts (58) followed-up for a and TOFA Pts (288) initiating (62) and TOFA Pts treated with and followed-up for 6 mths Pts previously treated with MTX Pts (56) treated with TOFA for a mean TOFA Pts (56) treated with

Databases ® US MarketScan

Rheumatic Diseases (1998-2015) US National Data Bank for (2012-2016; 3-yr. interim analysis) (2012-2016; 3-yr. registry US Corrona Rheumatoid Arthritis US Corrona Rheumatoid

US health plans (2010-2014)

(6-mth analysis)

Japanese post-marketing surveillance data

(2012-ongoing; 3-yr. interim data) (2012-ongoing; 3-yr. surveillance safety database

(2014-2016) (2013-2017) post-marketing Manufacturer’s Rheumatology centre in Japan (2015-2017) (2013-1016) (2010-2014) 6 countries in Latin America 6 countries in Latin (2015-2016) rheumatology centres inArgentina Data source Data source Rheumatology centre in Japan Taiwan Medical centre in St. Gallen Cohort, Switzerland Rheumatology centres from Medical records from (2013-2016) Rheumatology centre in Colombia et al . 2017 et al . 2016 et al . 2017 et al . 2017 et al . 2017 et al . 2016 et al . 2017 et al . 2017 et al . 2017 et al . 2018 et al . 2016 et al . 2017 et al . 2017 Machado

England

Kavanaugh

Curtis

Mimori

Cohen Mori

Table II. Summary of safety data from real-world studies with tofacitinib. Table Study Study Iwamoto Chen Müller Schneeberger Malpica Sansinanea hepa HBV: gastrointestinal; GI: drug; anti-rheumatic disease-modifying DMARD: cardiovascular; CV: DMARD; synthetic conventional csDMARD: interval; confidence CI: DMARD; biologic bDMARD: event; adverse AE: abatacept; ABA: rituximab; RR: reporting herpes simplex virus; IR: incidence rate; MTX: methotrexate; mth: month; NMSC: non-myeloma skin cancer; py: patient-year; RA: rheumatoid arthritis; RIT: titis B virus; HR: hazard ratio; HZ: herpes zoster; HSV: tofacitinib. TOFA: tocilizumab; TOCI: TNFi: tumour necrosis factor inhibitor; rate;

Clinical and Experimental Rheumatology 2019 491 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia for lack of efficacy and 4 (5.7%) due to post-marketing surveillance of the drug. tofacitinib compared with abatacept, adverse events. Treatment with tofaci- Interim analysis of the data entered in which was taken as reference given its tinib was associated with rapid and sig- the safety database from November common use as a second- or subsequent- nificant improvements in all disease ac- 2012 to November 2015 was recently line therapy in RA (HR 2.01, 95% CI tivity scores from baseline to 6 months. published (46). Most data (99.1%) were 1.40–2.88). Tofacitinib was also associ- Subgroup analysis by use or non-use from post-marketing reports sponta- ated with the greatest incidence of the of concomitant MTX showed that to- neously submitted to the drug manu- combined outcome (7.61/100 patient- facitinib was associated with significant facturer by healthcare professionals years, HR 1.40, 95% CI 1.09–1.81). improvements in disease activity scores or patients. Most originated in the US Rates and adjusted HRs for all other regardless of whether used in combina- (73.9%). Overall, the worldwide ex- bDMARDs were comparable to each tion with MTX or as monotherapy. The posure to tofacitinib during the 3-year other and abatacept. Of note, older age, study also analysed the effectiveness of reporting period was estimated to be female sex, prednisone >7.5 mg/day, tofacitinib in patients who failed treat- 34,223 patient-years. In total, 25,417 prior infection and greater number of ment with the interleukin (IL)-6 inhibi- adverse events, 4352 (17.1%) severe hospitalisations were associated with tor tocilizumab. These patients also ex- adverse events and 102 (0.4%) fatal increased HZ risk, whereas vaccination perienced significant improvements in cases were reported. Adverse events was associated with a lower risk. disease activity scores from baseline to with a frequency ≥5% were: ineffective A 5-year study based on the US Corro- 6 months, even though less substantial drug (13.2%), headache (9.0%), pain na registry and evaluating the standard- than those reported by the overall popu- (6.4%), fatigue (6.0%), nausea (6.0%), ised incidence rates of adverse events lation; 25% achieved disease remission condition aggravated (5.9%), diarrhoea of interest in three populations – tofaci- at 6 months. Of several baseline vari- (5.8%) and arthralgia (5.2%). The esti- tinib initiators, bDMARD initiators, ables tested in a multivariate regression mated incidence rates of severe adverse csDMARD initiators – is ongoing real- model, only the number of previously events per 100 patient-years were: 2.57 world (37). According to an interim used bDMARDs was significantly and for infections, 0.91 for gastrointestinal analysis including 760 tofacitinib, 4628 independently associated with achieve- disorders, 0.60 for respiratory disorders, bDMARD, and 1328 csDMARD initia- ment of LDA at 6 months. 0.45 for neoplasms, 0.43 for cardiac tors, the reported incidence rates of se- The effectiveness of tofacitinib has also disorders and 0.12 for hepatobiliary rious infections, cardiovascular events emerged from other small observational disorders. Non-melanoma skin cancer and malignancies in the three groups studies, which have confirmed that in was the most frequently reported malig- were similar despite differences in clinical practice tofacitinib is used both nancy during the 3-year post-marketing baseline characteristics (37). In tofaci- in combination with MTX and as mono- surveillance programme of tofacitinib, tinib, bDMARD and csDMARD initia- therapy, and is usually reserved for sec- with 16 serious cases reported. There tors the respective incidence rates per ond- and later lines of treatment, as cur- were 15 cases of lymphoma, with no 100 patient-years were 3.69, 3.29 and rently recommended (Table I) (40-45). increase in reporting frequency with in- 2.39 for severe infections, 1.43, 0.94 Notably, analysis of the Swiss Clinical creased exposure to tofacitinib. No new and 0.47 for HZ, 1.81, 1.77 and 1.81 for Quality Management registry including safety issues were reported compared malignancies, and 1.69, 2.57 and 2.59 almost 2000 patients initiating treat- with the safety profile of tofacitinib for cardiovascular disease. Of note, ment with tofacitinib, TNFi, or non- seen in the RA clinical development the incidence rate of HZ in tofacitinib TNFi bDMARDs between 2013 and programme. initiators (1.43/100 patient-years) was 2016 found similar crude drug retention Based on US health plan data, a study substantially lower than that reported rates for the three cohorts (40-45). After evaluated the risks of herpes zoster by other real-world studies (47) and in adjustment for potential confounders, a (HZ) and herpes simplex virus (HSV) the RCTs with tofacitinib (7). higher risk of discontinuation was asso- infection among RA patients with no A longitudinal observational study that ciated with TNFi versus tofacitinib [haz- history of HZ or HSV infection, initi- included RA patients without prevalent ard ratio (HR) 1.27, 95% CI 1.02–1.57, ating tofacitinib or bDMARDs (47). cancer participating in the US National p=0.03], but not with non-TNFi versus A total of 2526 patients initiating to- Data Bank for Rheumatic Diseases ex- tofacitinib (HR 1.03, 95% CI 0.83–1.28, facitinib were compared with patients amined the association of bDMARDs p=0.76). A higher number of prior bD- initiating TNFi (n=42,850), abatacept and tsDMARDs with incident cancer MARDs and greater BMI values were (n=12,305), rituximab (n=5078) and in RA patients (48). Among 11,582 also significantly associated with an in- tocilizumab (n=6967). The estimated patients, 6262 biologic/tofacitinib and creased risk of discontinuation. crude incidence of HZ associated with 5320 csDMARD initiators were iden- tofacitinib was 3.87/100 patient-years tified; overall, 1456 incident cancers Safety profile and was in agreement with that seen in occurred during the observation period Relevant safety information from real- the tofacitinib clinical trial programme (1998–2015). Increased rates of solid world data is summarised in Table II. In (3.3/100 patient-years) (47). After mul- tumours were found with tofacitinib; 2012, the manufacturer of tofacitinib in- tivariate adjustment, HZ risk was found non-TNFi biologics were associated stituted a safety database for worldwide to be significantly more elevated with with increased rates of lymphoprolif-

492 Clinical and Experimental Rheumatology 2019 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia erative malignancies. After adjustment one-third of the patients experienced world data also show that persistence in for confounding variables, no increased at least one adverse event. Infections treatment with tofacitinib and adherence risk of cancer with bDMARDs or tofac- were reported by 367 patients (12.7%). to treatment are good overall and similar itinib was found relative to MTX mono- The most frequent adverse events were to those seen for bDMARDs. therapy. Overall, the results were incon- HZ (3.4%) and abnormality in liver en- Two real-world studies have addressed clusive and highlighted the difficulty of zymes (1.7%). Severe adverse events the impact of prior bDMARD use on determining treatment-related cancer occurred in 221 patients (7.7%). The outcomes and shown that biologic-naïve risk. Regarding non-melanoma skin most frequent severe adverse events patients initiating tofacitinib are more cancer, which was the most frequently were pneumonia (0.7%), interstitial likely to achieve LDA or disease remis- reported malignancy identified dur- lung disease (0.5%) and worsening sion than biologic-experienced patients ing post-marketing surveillance (46), of condition (0.4%). Infections were (38, 39). Furthermore, a significant and an analysis of data from 18 studies in severe in 101 patients (3.5%). Malig- inverse association of the number of the tofacitinib RA clinical programme nancies were reported in 21 patients bDMARDs used with the achievement identified 83 cases among 6,092 tofac- (0.7%). There were 16 deaths (0.6%) of LDA has been identified (38, 39). itinib-treated patients (15,103 patient- over the 6-month observation period. As a consequence, it may be advisable years) (49). There was no clear dose A safety profile in line with the findings to introduce tofacitinib before failure relationship, and the incidence rate was from RCTs has also been described in a of multiple bDMARDs. However, it is consistent with those observed in pa- number of other real-life observational unclear at present which patients may tients with RA treated with TNFi and studies as summarised in Table II (42- benefit most from early introduction of other bDMARDS. 44, 52, 53). Of note, a recent retrospec- tofacitinib, and further investigations A retrospective analysis of real-world tive analysis in a real-life cohort of 116 are needed to clarify this point. data from the US MarketScan® data- Taiwanese patients RA treated with As for the safety profile of tofacitinib in bases compared the incidence of seri- tofacitinib evaluated the risk of hepa- real-world patients, with the exception ous infections associated with various titis B virus (HBV) reactivation (52). of one study in which about one-third treatments including tofacitinib in RA Overall, 81 (69.8%) patients had a past of patients interrupted treatment with patients previously treated with MTX HBV infection; 6 were defined as HBV tofacitinib due to adverse events (42), (50). The analysis included 21,832 carriers, while in 75 HBV infection was discontinuation rates for safety issues patients: 0.8% treated with tofaci- considered resolved. Among the 6 car- appear to be <10% (38, 39, 43, 44, 51). tinib, 24.7% with csDMARDs, 61.2% riers, only 2 received antiviral prophy- Overall, no unexpected safety events with TNFi and 13.3% with other bD- laxis during treatment with tofacitinib. have been reported in the real-world MARDs. The incidence of serious in- Of the 4 carriers without antiviral setting and incidence rates of adverse fections per 100 patient-years with to- prophylaxis, 2 experienced HBV reac- events of interest (HZ, serious infec- facitinib was 3.7; in the other treatment tivation during treatment with tofacitin- tions, opportunistic infections, malig- groups, it ranged from 2.00 to 2.53 and ib. No HBV reactivation was observed nancies, cardiovascular disease) appear was 1.96 for the reference group of in patients with resolved infection. to be similar to those reported in the patients with no prescribed treatment. RCTs (7, 19, 47). The adjusted HRs for infections requir- Discussion and conclusions In conclusion, this review provides ing hospitalisation were 1.81 (95% CI The present narrative review provides, a first glimpse into the real-world use 1.08–3.01) for tofacitinib, 0.82 (95% for the first time to the best of our and effectiveness of tofacitinib. Though CI 0.62–1.07) for csDMARDs, 1.13 knowledge, a summary of the real- essential for establishing the effective- (95% CI 0.92–1.38) for TNFi and 1.10 world evidence concerning tofacitinib ness of a treatment, data from post- (95% CI 0.87–1.38) for non-TNFi bD- in the management of patients with RA. marketing surveillance programmes are MARDs versus the reference group. Despite the heterogeneity of the avail- not without limitations: most reports The initial 6-month interim data from able sources of real-world evidence, a are spontaneous, and the magnitude of the ongoing post-marketing surveil- number of findings are consistent across non-reported cases is unknown (46). In lance programme of tofacitinib in Ja- the studies reviewed. With regards to pa- addition, relevant clinical information pan have not yet revealed any new or tient characteristics, treatment patterns is often missing or incomplete so that unexpected safety issues compared and outcomes, in the real-world setting confounding effects that may contribute with the adverse event profile reported patients initiating tofacitinib usually to reported adverse events cannot be in the registration RCTs (51). Overall, have longer disease duration and have identified (46). Clearly, more extensive 2882 patients were enrolled, with an been exposed to longer bDMARDs than and long-term data are needed to fully estimated exposure to tofacitinib of patients initiating a bDMARD. Tofaci- characterise the safety profile of tofaci- 1241.4 patient-years. Of these patients, tinib appears to be more frequently used tinib, the optimal timing of treatment 686 (23.8%) discontinued treatment. as monotherapy than bDMARDs and to initiation and position of Discontinuations were due to adverse be effective without background MTX, JAK inhibition among the various tar- events in 276 cases (9.6%) and to lack in agreement with the results from the geted strategies available for treatment of efficacy in 260 cases (9.0%). About ORAL RCT programme (8). The real- of RA.

Clinical and Experimental Rheumatology 2019 493 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia

Acknowledgments diographic study. Arthritis Rheum 2013; 65: Special Task Force on real-world evidence in Medical writing support was provided 559-70. health care decision making. Pharmacoepi- 13. van VOLLENHOVEN RF, FLEISCHMANN R, demiol Drug Saf 2017; 26: 1033-9. by Lorenza Lanini on behalf of Health COHEN S et al.: Tofacitinib or adalimumab 25. FINCKH A, COURVOISIER D: Lessons learned Publishing & Services Srl and was versus placebo in rheumatoid arthritis. from rheumatoid arthritis registries. Joint funded by Pfizer. N Engl J Med 2012; 367: 508-19. Bone Spine 2018; 85: 271-4. 14. FLEISCHMANN R, CUTOLO M, GENOVESE 26. KIM SY: Efficacy versus effectiveness. MC et al.: Phase IIb dose-ranging study of the Korean J Fam Med 2013; 34: 227. References oral JAK inhibitor tofacitinib (CP-690,550) 27. KREMER JM: The Corrona US registry of 1. SMOLEN JS, LANDEWE R, BIJLSMA J et al.: or adalimumab monotherapy versus placebo rheumatic and autoimmune diseases. Clin Exp EULAR recommendations for the manage- in patients with active rheumatoid arthritis Rheumatol 2016; 34 (Suppl. 101): S96-S99. ment of rheumatoid arthritis with synthetic with an inadequate response to disease-mod- 28. GOMEZ-REINO JJ, CARMONA L, VALVERDE and biological disease-modifying antirheu- ifying antirheumatic drugs. Arthritis Rheum VR, MOLA EM, MONTERO MD, Biobadaser matic drugs: 2016 update. Ann Rheum Dis 2012; 64: 617-29. Group: Treatment of rheumatoid arthritis 2017; 76: 960-77. 15. WALLENSTEIN GV, KANIK KS, WILKINSON with tumor necrosis factor inhibitors may 2. KAWALEC P, SLADOWSKA K, MALINOWS- B al.: Effects of the oral Janus kinase inhibi- predispose to significant increase in tubercu- KA-LIPIEN I, BRZOSTEK T, KOZKA M: Euro- tor tofacitinib on patient-reported outcomes losis risk: a multicenter active-surveillance pean perspective on the management of rheu- in patients with active rheumatoid arthritis: report. Arthritis Rheum 2003; 48: 2122-7. matoid arthritis: clinical utility of tofacitinib. results of two Phase 2 randomised controlled 29. KEANE J, GERSHON S, WISE RP et al.: Ther Clin Risk Manag 2018; 14: 15-29. trials. Clin Exp Rheumatol 2016; 34: 430-42. Tuberculosis associated with infliximab, 3. EMA (European Medicines Agency) 2017. 16. KREMER JM, BLOOM BJ, BREEDVELD FC a tumor necrosis factor alpha-neutralizing XELJANZ (tofacitinib citrate) 5 mg film- et al.: The safety and efficacy of a JAK in- agent. N Engl J Med 2001; 345: 1098-104. coated tablets: Summary of Product Char- hibitor in patients with active rheumatoid 30. AALTONEN KJ, YLIKYLA S, TUULIKKI JOEN- acteristics. Available from: http://www.ema. arthritis: Results of a double-blind, placebo- SUU J et al.: Efficacy and effectiveness of europa.eu controlled phase IIa trial of three dosage lev- tumour necrosis factor inhibitors in the treat- 4. US FDA (Food and Drug Administration) els of CP-690,550 versus placebo. Arthritis ment of rheumatoid arthritis in randomized 2017. XELJANZ(R) (tofacitinib) tablets / Rheum 2009; 60: 1895-905. controlled trials and routine clinical practice. XELJANZ(R) XR (tofacitinib) extended re- 17. KREMER JM, COHEN S, WILKINSON BE et Rheumatology (Oxford) 2017; 56: 725-35. lease tablets: Prescribing information. Avail- al.: A phase IIb dose-ranging study of the 31. KILCHER G, HUMMEL N, DIDDEN EM, EGG- able from: https://www.accessdata.fda.gov/ oral JAK inhibitor tofacitinib (CP-690,550) ER M, REICHENBACH S, GETREAL WORK 5. SINGH JA, SAAG KG, BRIDGES SL, Jr. et al.: versus placebo in combination with back- PACKAGE: Rheumatoid arthritis patients 2015 American College of Rheumatology ground methotrexate in patients with ac- treated in trial and real world settings: com- Guideline for the Treatment of Rheumatoid tive rheumatoid arthritis and an inadequate parison of randomized trials with registries. Arthritis. Arthritis Rheumatol 2016; 68: 1-26. response to methotrexate alone. Arthritis Rheumatology (Oxford) 2018; 57: 354-69. 6. BURMESTER GR, BLANCO R, CHARLES- Rheum 2012; 64: 970-81. 32. VASHISHT P, SAYLES H, CANNELLA AC, SCHOEMAN C et al.: Tofacitinib (CP- 18. TANAKA Y, SUZUKI M, NAKAMURA H, MIKULS TR, MICHAUD K: Generalizability 690,550) in combination with methotrexate TOYOIZUMI S, ZWILLICH SH, Tofacitinib of patients with rheumatoid arthritis in bio- in patients with active rheumatoid arthritis Study I: Phase II study of tofacitinib (CP- logic agent clinical trials. Arthritis Care Res with an inadequate response to tumour ne- 690,550) combined with methotrexate in (Hoboken) 2016; 68: 1478-88. crosis factor inhibitors: a randomised phase 3 patients with rheumatoid arthritis and an in- 33. ZINK A, STRANGFELD A, SCHNEIDER M et trial. Lancet 2013; 381: 451-60. adequate response to methotrexate. Arthritis al.: Effectiveness of tumor necrosis factor 7. COHEN SB, TANAKA Y, MARIETTE X et al.: Care Res (Hoboken) 2011; 63: 1150-8. inhibitors in rheumatoid arthritis in an obser- Long-term safety of tofacitinib for the treat- 19. FLEISCHMANN R, WOLLENHAUPT J, TA- vational cohort study: comparison of patients ment of rheumatoid arthritis up to 8.5 years: KIYA L et al.: Safety and maintenance of according to their eligibility for major rand- integrated analysis of data from the global response for tofacitinib monotherapy and omized clinical trials. Arthritis Rheum 2006; clinical trials. Ann Rheum Dis 2017; 76: combination therapy in rheumatoid arthritis: 54: 3399-407. 1253-62. an analysis of pooled data from open-label 34. HARNETT J, GERBER R, GRUBEN D, KOENIG 8. FLEISCHMANN R, KREMER J, CUSH J et al.: long-term extension studies. RMD Open AS, CHEN C: Evaluation of real-world ex- Placebo-controlled trial of tofacitinib mono- 2017; 3: e000491. perience with tofacitinib compared with therapy in rheumatoid arthritis. N Engl J Med 20. BARNISH MS, TURNER S: The value of prag- adalimumab, etanercept, and abatacept in RA 2012; 367: 495-507. matic and observational studies in health patients with 1 previous biologic DMARD: 9. FLEISCHMANN R, MYSLER E, HALL S et al.: care and public health. Pragmat Obs Res Data from a U.S. Administrative Claims Da- Efficacy and safety of tofacitinib monothera- 2017; 8: 49-55. tabase. J Manag Care Spec Pharm 2016; 22: py, tofacitinib with methotrexate, and adali- 21. EGGER M, MOONS KG, FLETCHER C, GETRE- 1457-71. mumab with methotrexate in patients with AL WORKPACKAGE: GetReal: from efficacy 35. SMITH T, HARNETT J, GRUBEN D, CHEN C, rheumatoid arthritis (ORAL Strategy): a phase in clinical trials to relative effectiveness in AGARWAL E, WOOLCOTT J: Real-world ex- 3b/4, double-blind, head-to-head, randomised the real world. Res Synth Methods 2016; 7: perience with tofacitinib versus adalimumab controlled trial. Lancet 2017; 390: 457-68. 278-81. and etanercept in biologic-naive patients 10. KREMER J, LI ZG, HALL S et al.: Tofacitinib in 22. FAVALLI EG, BUGATTI S, BIGGIOGGERO with RA previously treated with methotrex- combination with nonbiologic disease-modi- M, CAPORALI R: Treatment comparison in ate: Data from a US Administrative Health- fying antirheumatic drugs in patients with ac- rheumatoid arthritis: head-to-head trials and care Insurance Claims Database. Arthritis tive rheumatoid arthritis: a randomized trial. innovative study designs. Biomed Res Int Rheumatol 2017; 69 (S10): Abstract 2831. Ann Intern Med 2013; 159: 253-61. 2014; 2014: 831603. 36. REED GW, GERBER RA, SHAN Y: TNFI and 11. LEE EB, FLEISCHMANN R, HALL S et al.: 23. EICHLER HG, ABADIE E, BRECKENRIDGE tofacitinib monotherapy and comparative ef- Tofacitinib versus methotrexate in rheuma- A et al.: Bridging the efficacy-effectiveness fectiveness in clinical practice: results form toid arthritis. N Engl J Med 2014; 370: 2377- gap: a regulator’s perspective on addressing Corrona registry [Oral presentation]. Ann 86. variability of drug response. Nat Rev Drug Rheum Dis 2017; 76 (Suppl. 2): 60. 12. van der HEIJDE D, TANAKA Y, FLEIS- Discov 2011; 10: 495-506. 37. KAVANAUGH AF, GEIER J, BINGHAM CI et CHMANN R et al.: Tofacitinib (CP-690,550) 24. BERGER ML, SOX H, WILLKE RJ et al.: Good al.: Real world results from a post-approval in patients with rheumatoid arthritis receiv- practices for real-world data studies of treat- safety surveillance of tofacitinib (Xeljanz): ing methotrexate: twelve-month data from a ment and/or comparative effectiveness: Rec- Over 3 year results from an ongoing US- twenty-four-month phase III randomized ra- ommendations from the joint ISPOR-ISPE based rheumatoid arthritis registry. Arthritis

494 Clinical and Experimental Rheumatology 2019 Real-world experience with tofacitinib in RA / R. Caporali & D. Zavaglia

Rheumatol 2016; 68 (S10): Abstract 2595. Tofacitinib in rheumatoid arthritis: real life Abstract 3066. 38. MORI S, YOSHITAMA T, UEKI Y: Tofacitinib experience. Ann Rheum Dis 2017; 76 (Suppl. 49. CURTIS JR, LEE EB, MARTIN G et al.: Analy- therapy for rheumatoid arthritis: A direct 2): 1202. sis of non-melanoma skin cancer across the comparison study between biologic-naive 44. SCHNEEBERGER EE, SALAS A, MEDINA LF et tofacitinib rheumatoid arthritis clinical pro- and experienced patients. Intern Med 2018; al.: Real-world use of tofacitinib in rheuma- gramme. Clin Exp Rheumatol 2017; 35: 614- 57: 663-70. toid arthritis: data from Latin America [Ab- 22. 39. IWAMOTO N, TSUJI S, TAKATANI A et al.: stract]. Ann Rheum Dis 2017; 76 (Suppl. 29): 50. MACHADO M, MOURA CS, FERREIRA GUER- Efficacy and safety at 24 weeks of daily clini- 196. RA S et al.: Serious infections associated with cal use of tofacitinib in patients with rheuma- 45. FINCKH A, HERZOG L, SCHERER A: Drug tofacitinib in rheumatoid arthritis patients toid arthritis. PLoS One 2017; 12: e0177057. retention of tofacitinib versus biologic anti- previously treated with methotrexate. Arthri- 40. KARATEEV D, LUCHIKHINA E, ABDUL- rheumatic agents in rheumatoid arthritis: ob- tis Rheumatol 2017; 69 (S10): Abstract 2785. GANIEVA D et al.: Tofacitinib in rheumatoid servational data from the Swiss SCQM regis- 51. MIMORI T, HARIGAI M, ATSUMI T et al.: arthritis: results of post-approval investiga- try [Poster]. Ann Rheum Dis 2017; 76 (Suppl. Post-marketing surveillance of tofacitinib in tor initiated trial. Ann Rheum Dis 2016; 76 2): 267. Japanese patients with rheumatoid arthritis: (Suppl. 2): 1036. 46. COHEN S, CURTIS JR, DEMASI R et al.: World- an interim report of safety data. Ann Rheum 41. KYBURZ D, RIEK M, HERZOG L et al.: Use wide, 3-year, post-marketing surveillance ex- Dis 2017; 76 (Suppl. 2): 1200. of tofacitinib in rheumatoid arthritis: Data perience with tofacitinib in rheumatoid arthri- 52. CHEN YM, HUANG WN, WU YD et al.: from the Swiss Clinical Quality Management tis. Rheumatol Ther 2018; 5: 283-91. Reactivation of hepatitis B virus infection in RA Registry. Arthritis Rheumatol 2016; 68 47. CURTIS JR, XIE F, YUN H, BERNATSKY S, patients with rheumatoid arthritis receiving (S10): Abstract 1637. WINTHROP KL: Real-world comparative tofacitinib: a real-world study. Ann Rheum 42. MÜLLER R, MATTOW F, POPP F, von KEMPIS risks of herpes virus infections in tofacitinib Dis 2018; 77: 780-2. J: Effectiveness, tolerability, and safety of to- and biologic-treated patients with rheumatoid 53. MALPICA A, SANTOS-MORENO P, MENDEZ facitinib in rheumatoid arthritis: A retrospec- arthritis. Ann Rheum Dis 2016; 75: 1843-7. K, BUITRAGO-GARCIA D: Safety profile of tive analysis of real-world data from the St. 48. ENGLAND BR, PEDRO S, MIKULS TR, tofacitinib in patients with rheumatoid ar- Gallen Cohort. [Poster]. Arthritis Rheumatol MICHAUD K: Risk of incident cancer with bi- thritis – A real world experience with good 2017; 69 (S10): Abs 536. ologic and tofacitinib therapy in rheumatoid results. Ann Rheum Dis 2017; 76 (Suppl. 2): 43. SANSINANEA P, COSTI AC, VULCANO A et al.: arthritis. Arthritis Rheumatol 2016; 68 (S10): 1502.

Clinical and Experimental Rheumatology 2019 495