Sara Gale

Cardiovascular Safety of Tocilizumab Versus Tumor Necrosis Factor Inhibitors in Patients With Rheumatoid Arthritis: A Multi-Database Cohort Study

While tocilizumab (TCZ) is known to increase low‐density lipoprotein (LDL) cholesterol levels, it is unclear whether TCZ increases cardiovascular risk in patients with rheumatoid arthritis (RA). This study was undertaken to compare the cardiovascular risk associated with receiving TCZ versus tumor necrosis factor inhibitors (TNFi).

Effectiveness of Biologic and Non-biologic Antirheumatic Drugs on Anaemia Markers in 153,788 Patients with Rheumatoid Arthritis: New Evidence from Real-World Data

BACKGROUND To evaluate the impact of treatment with disease-modifying antirheumatic drugs (DMARDs), including IL-6 receptor inhibitor tocilizumab (TCZ), on anaemia markers in patients with rheumatoid arthritis. METHODS Using the Centricity Electronic Medical Records from USA, patients with rheumatoid arthritis diagnosed between January 2000 and April 2016, who initiated TCZ (n = 3732); tofacitinib (TOFA, n = 3126); other biologic DMARD (obDMARD, n = 55,964); or other non-biologic DMARD (onbDMARD, n = 91,236) were identified. Changes in haemoglobin (Hb) and haematocrit (Hct) over 2 years of treatment initiation were evaluated, adjusting and balancing for confounders. RESULTS Mean (95% CI) adjusted increase in Hb and Hct levels at 24 months in TCZ group were 0.23g/dL (0.14, 0.42) and 0.96% (0.41, 1.52) respectively. Among patients with anaemia in the TCZ group, Hb and Hct increased significantly by 0.72g/dL and 2.06%, respectively. Patients in the TCZ group were 86% (95% CI of OR: 1.43, 2.00) more likely to increase Hb ≥ 1g/dL compared to the other groups combined. No clinically significant changes in Hb were observed in the other groups. The obDMARD group demonstrated lower Hct increase than TCZ group, while no significant changes were observed in the remaining groups. Compared to those who initiated TCZ therapy after 1 year of diagnosis of rheumatoid arthritis, those who initiated earlier were 95% (OR = 1.95; 95% CI: 1.19, 3.21; p < 0.001) more likely to increase Hb within 6 months. CONCLUSIONS This real-world study suggests significant increase in Hb and Hct levels after TCZ therapy in anaemic and non-anaemic patients with rheumatoid arthritis, compared with other biologic and non-biologic DMARDs.

FRI0248 Incidence of melanoma in patients with rheumatoid arthritis treated with tocilizumab

Background There have been conflicting reports whether patients with rheumatoid arthritis (RA) receiving conventional or biologic immunosuppressive therapies are at increased risk of specific malignancies. Melanoma is an aggressive malignancy with risk factors including sex, age, fair skin and elevated cumulative UV exposure. Interleukin-6 (IL-6) has a role in pro- and antitumorigenic pathways. Whether tocilizumab (TCZ), a biologic that alters IL-6 signaling, increases the risk of melanoma in patients with RA is unclear. Objectives This age- and sex-adjusted standardized incidence ratio (SIR) analysis compared the observed reports of melanoma in patients with RA treated with TCZ in clinical trial and postmarketing settings with the expected number of cases across geographic regions. Methods SIRs for melanoma were calculated from the TCZ clinical trials all-exposure population. Postmarketing rates were estimated from the TCZ Global Safety Database population. Both databases were searched cumulatively from 11 April 2005 to 10 October 2015. For clinical trials, observed reports of melanoma in patients with RA treated with TCZ were compared with expected number of cases in the general population based on the 2012 US Surveillance, Epidemiology, and End Results using an age- and sex-adjusted SIR. Postmarketing regional SIRs were calculated based on the estimated commercial exposure in each region and the incidence of melanoma as reported by Globocan by age and sex (2012). Crude postmarketing rates were age- and sex-adjusted according to the clinical trial demographic profile to estimate the expected number of cases of melanoma in each region. Results In the clinical trial setting, 4 qualifying cases of melanoma were identified among 7093 patients with RA treated with TCZ (20,828 PY of exposure). The SIR estimate (0.71 [95% CI, 0.19–1.81]) for melanoma incidence in patients with RA treated with TCZ in clinical trials was comparable to that in the general population (Table 1). In the postmarketing setting, the number of observed reports of melanoma was comparable to the expected number of cases in Europe and Japan and fewer than expected in North America (Table 2). The exception is Australia, where SIR estimates indicated more than the expected number of cases in patients with RA receiving TCZ in Australia compared with the general population in that region (SIR 3.71 [95% CI: 2.16, 5.93]). Conclusions In clinical trials, no evidence was found to suggest there were more cases of melanoma than expected in patients with RA treated with TCZ compared with the general population. Consistent with this, no evidence was found to suggest that patients with RA treated with TCZ in Europe, North America or Japan had more cases of melanoma than expected compared with the general population in each region. In contrast, the estimated SIR of melanoma in patients with RA treated with TCZ in Australia indicated more than the expected number of cases in the general population. This finding is consistent with reports of elevated risk of melanoma in patients with RA in Australia (compared with the general population), where UV exposure is high and methotrexate is a common first-line therapy.1,2 References Buchbinder et al. Arthritis Rheum. 2008. Buchbinder et al. BMC Musculoskelet Disord. 2015. Acknowledgements Funded by Roche/Genentech. Disclosure of Interest S. Gale Employee of: Genentech, J. Wang Employee of: Roche, J. Nebesky Employee of: Roche, A. Linke Employee of: Genentech, E. Berber Employee of: Genentech

No difference in cardiovascular risk of tocilizumab versus abatacept for rheumatoid arthritis: A multi-database cohort study

OBJECTIVES While tocilizumab may increase serum lipid levels, recent studies do not suggest a link between tocilizumab use and clinical cardiovascular risk in patients with rheumatoid arthritis (RA). METHODS To compare cardiovascular safety of tocilizumab with abatacept, we conducted a cohort study using data from Medicare (2010-2013), IMS PharMetrics (2011-2014) and MarketScan (2011-6/2015). RA patients aged ≥18 years who newly started tocilizumab or abatacept entered the cohort on the day of their first use of tocilizumab or abatacept after a continuous enrollment period for ≥365 days. The primary outcome was a composite cardiovascular endpoint of hospitalization for myocardial infarction or stroke. To control for more than 60 confounders, tocilizumab starters were propensity score (PS)-matched to abatacept starters with a variable ratio of 1:3 within each database. A fixed-effects model combined database-specific hazard ratios (HR). RESULTS We included 6237 tocilizumab starters PS-matched to 14,685 abatacept starters in all three databases. Mean age was 72 years in Medicare, 51 in PharMetrics and 53 in MarketScan. The incidence rate of the composite cardiovascular events per 100 person-years ranged from 0.37 (PharMetrics) to 1.64 (Medicare) in the tocilizumab group and from 0.59 (PharMetrics) to 1.69 (Medicare) in the abatacept group. The risk of the composite cardiovascular events was similar between the two groups across all three databases, with a combined HR of 0.82 (95% CI: 0.55-1.22) in tocilizumab versus abatacept starters. CONCLUSIONS This multi-database cohort study found no difference in the risk of cardiovascular events in RA patients who newly started tocilizumab versus abatacept.

Incidence and risk of glucocorticoid-associated adverse effects in patients with rheumatoid arthritis

Using the UK Clinical Practice Research Datalink, we examined the incidence of glucocorticoid (GC)‐related serious adverse events (SAEs) in rheumatoid arthritis (RA) and non‐RA patients and quantified the risk of SAEs in patients with RA.

SAT0206 Retention of tocilizumab as monotherapy versus tnf inhibitors with conventional synthetic dmards in rheumatoid arthritis patients with inadequate response to tnf inhibitors: a study from the tocerra collaboration

Background Tocilizumab (TCZ) as monotherapy has been shown to be more efficacious than the TNF inhibitor (TNFi) adalimumab as monotherapy in patients with rheumatoid arthritis (RA). However, effectiveness data comparing TCZ as monotherapy versus TNF inhibitors in combination with csDMARDs are limited. Objectives To examine retention of TCZ administered alone (TCZ mono) versus TNFi in combination with csDMARDs (TNFi combo) in patients with RA who had an inadequate response to ≥1 TNFi (TNFi-IR). Methods Patients with RA who were TNFi-IR and treated with TCZ mono or TNFi combo with baseline (BL) data, not immediately lost to follow-up and started treatment after TCZ was available across 9 European registries in TOCERRA from 2009 to 2016 were included. The hazard for TCZ discontinuation was modeled using a country-stratified Cox proportional hazards model, adjusting for age, gender, disease duration, seropositivity, HAQ and CDAI at BL, number of previous csDMARD and biologic DMARD (bDMARD), glucocorticosteroid and calendar year of treatment initiation. Missing data on covariates were imputed using multiple imputation with chained equations. Results A total of 4748 patients were eligible, including 585 who received TCZ mono and 4163 who received TNFi combo. Patients who received TCZ mono were older with a longer disease duration, more previous bDMARDs and less glucocorticosteroids at baseline (Table 1) compared with patients who received TNFi combo. The crude median retention for TCZ mono was 1.82 years (95% CI: 1.59–2.09) and 1.54 years (95% CI: 1.43–1.64) for TNFi combo, (P=0.65). Causes of discontinuation differed between TCZ mono and TNFi combo (P<0. 001): TCZ mono stopped more frequently for ineffectiveness (25.7% vs. 13.8%) and TNFi combo stopped more frequently for safety issues (18.3% vs. 12.8%). In a country-stratified, covariate-adjusted analysis, we found that hazards of discontinuation were significantly lower among patients who received TCZ mono (HR: 0.71, P<0.001). More previous treatment with bDMARDs and a greater HAQ and CDAI at BL were significantly associated with greater risk of discontinuation. Conclusions In routine care across 9 European countries, TCZ mono retention is better than TNFi combo in patients with RA who were TNFi-IR. Acknowledgements Funding by F. Hoffmann-La Roche/Genentech. Disclosure of Interest K. Lauper: None declared, D. Nordström Grant/research support from: AbbVie, BMS, MSD, Pfizer, Roche and UCB, Consultant for: AbbVie, BMS, MSD, Roche, UCB and Pfizer, K. Pavelka Grant/research support from: AbbVie, Roche, Medis, MSD and Pfizer, Consultant for: AbbVie, Roche, Amgen, MSD, BMS, UCB and Egis, V. Hernandez: None declared, M. J. Santos: None declared, Z. Rotar: None declared, F. Iannone: None declared, C. Codreanu: None declared, G. Lukina Consultant for: BMS, Roche, MSD, AbbVie and Pfizer, S. Gale Employee of: Genentech, K. Sarsour Employee of: Genentech, A. Pethoe-Schramm Employee of: F. Hoffmann-La Roche, D. Courvoisier: None declared, C. Gabay Grant/research support from: AB2 Bio, AbbVie, Actelion, BMS, Debiopharm, MSD, Novartis, Pfizer, Regeneron, Roche, Sanofi and UCB, Consultant for: AB2 Bio, AbbVie, Actelion, BMS, Debiopharm, MSD, Novartis, Pfizer, Regeneron, Roche, Sanofi and UCB

THU0172 Risk for Serious Adverse Events Associated with Oral Corticosteroid Therapy in Patients with Rheumatoid Arthritis: A Uk Population-Based Study

Background Rheumatoid arthritis (RA) is a chronic, erosive form of inflammatory arthritis characterized by a distinctive pattern of bone and joint destruction. Along with disease-modifying antirheumatic drugs and newer biologic therapy, treatment can include oral corticosteroids (oCS), predominantly prednisolone or prednisone. CS are associated with a variety of adverse effects.1 Objectives Using data from the UKs Clinical Practice Research Datalink, we estimated the incidence rates (IRs) and incidence rate ratios (IRRs) of known oCS-related serious adverse events (SAEs) of interest in an RA cohort and in a non-RA general population (GP) comparison group matched on age, sex, and time. We also explored the role of oCS use on the risk for SAEs of interest by comparing RA patients of varying exposure levels of oCS. Methods Patients with a first-time diagnosis of RA who were ≥18 years of age were identified and matched to patients in a non-RA GP comparison group of equal size. Group IRs and IRRs were estimated for diabetes, osteoporosis, fractures, glaucoma, hypertension, thrombotic stroke or myocardial infarction (MI), gastrointestinal (GI) perforation or bleeding, and death for the RA group and the GP group. We then performed a series of nested case-control analyses examining the effect of oCS use on the risk for SAEs of interest. We used conditional logistic regression to calculate the unadjusted and multivariate adjusted odds ratios (ORs) with 95% confidence intervals (CIs) for the association between oCS use and the risk for each separate outcome of interest. We stratified the analyses according to increasing cumulative and average daily oCS use. Results The cohort consisted of 34,047 RA patients and 34,047 matched non-RA GP patients. Mean follow-up time was ∼8 years. The incidence for most SAEs was greater in the RA group than in the GP group. IRRs (95% CIs) for SAEs of interest were diabetes (1.21; 1.13–1.31), osteoporosis (2.11; 1.96–2.27), fractures (1.16; 1.08–1.24), glaucoma (1.08; 0.94–1.22), hypertension (1.13; 1.09–1.18), thrombotic stroke or MI (1.22; 1.11–1.34), GI perforation or bleeding (1.3; 1.15–1.47), and death (1.17; 1.11–1.22). In multivariate analyses, increasing risks for diabetes, osteoporosis, fractures, thrombotic stroke or MI, and death were observed with increasing cumulative oCS use (p <0.05 for trend). RA patients receiving an average daily dose >7.5 mg/day had an increased adjusted OR (95% CI) for diabetes (1.49; 1.21–1.84), osteoporosis (1.41; 1.20–1.66), fractures (1.22; 1.03–1.45), thrombotic stroke or MI (1.61; 1.27–2.04), and death (1.49; 1.27–1.74). Conclusions Patients with RA experienced an increased incidence of diabetes, osteoporosis, fractures, hypertension, thrombotic stroke or MI, GI perforation or bleeding, and death compared with age- and sex-matched controls. Increasing cumulative oCS doses and average daily oCS doses were associated with increased odds of diabetes, osteoporosis, fractures, thrombotic stroke or MI, and death. References Venables PJW. Rheumatoid arthritis treatment (beyond the basics). UpToDate, Inc. 2016. http://www.uptodate.com/contents/rheumatoid-arthritis-treatment-beyond-the-basics.Accessed January 21, 2016. Disclosure of Interest J. Wilson: None declared, K. Sarsour Shareholder of: Roche, Employee of: Roche, S. Gale Shareholder of: Roche, Employee of: Roche, A. Pethö-Schramm Employee of: F. Hoffmann-La Roche, S. Jick: None declared, C. Meier Shareholder of: Hoffmann-La Roche, Grant/research support from: Hoffmann-La Roche

FRI0298 Treatment Patterns among Adult Systemic Sclerosis Patients in A US Healthcare Claims Population

Background In the United States, systemic sclerosis (SSc) develops in adults at an annual estimated incidence rate of 20 cases per million, and the estimated prevalence is 240 cases per million.1 Published research on the epidemiology of SSc and common treatments for US SSc patient cohorts is limited. In general, SSc patients are treated based on their specific organ involvement; patients with diffuse skin involvement or severe organ involvement, or both, are treated with systemic immunosuppressive therapy. Objectives To understand the treatment patterns for and the real-world use of immunosuppressive drugs in adults with SSc using data from a large US healthcare claims database. Methods The analysis is a retrospective cohort study of patients with SSc diagnosed from 2006 to 2012 using data from Truven Health MarketScan® Research Databases, a medical insurance claims database of insured active employees and their dependents, early retirees, Medicare-eligible retirees, and Medicaid recipients. An SSc diagnosis was determined by ≥1 inpatient claim for an SSc diagnosis (ICD-9 710.1) or ≥2 outpatient claims for an SSc diagnosis. During the study time period, patients were allowed enrollment gaps of ≤30 days but had to have enrollment data 6 months before and 12 months after the SSc diagnosis. Results The prevalence of SSc in the MarketScan database from 2006 to 2013 ranged from 204 to 247 cases per million adults who were actively enrolled in health insurance plans. There were 6872 patients ≥18 years of age diagnosed with and treated for SSc during the study time period. Eighty-six percent of the treated SSc population were female, and the average age of the cohort was 54 years. The most common medication claims among patients in the first year after an SSc diagnosis were for antibiotics (54%), opioids (44%), corticosteroids (40%), and proton pump inhibitors (PPIs) (31%). Approximately 30% of SSc patients initiated ≥1 of the following disease-modifying antirheumatic drugs (DMARDs), alone or in combination, in the first year after the SSc index, with a median time to initiation of 70 days: methotrexate (21%), mycophenolate mofetil (17%), biologics (3%), cyclophosphamide (2%), or other DMARDs (62% of which included azathioprine, chloroquine, cyclosporine, hydroxychloroquine, penicillamine, quinacrine, and sulfasalazine). Among patients who filled a DMARD prescription in the first year, 15% switched to a second type of DMARD. Of the 70% of patients who did not fill a prescription for DMARD in the first year after the SSc index, 63% filled prescriptions for antibiotics, 56% opioids, 33% corticosteroids, 26% PPIs, 20% NSAIDs, 10% ACE inhibitors, 10% calcium antagonists, 5% prokinetics, 1% ERA, and <1% prostacyclin. Conclusions These data show that the prevalence of SSc among adults in a US healthcare claims population is consistent with the estimated US prevalence. Most SSc patients do not initiate DMARD treatment in the first year after diagnosis, and most SSc patients have received antibiotics, opioids, and systemic corticosteroids. References Mayes MD et al. Prevalence, incidence, survival, and disease characteristics of systemic sclerosis in a large US population. Arthritis Rheum. 2003;48:2246–2255. Disclosure of Interest S. Gale Shareholder of: Roche, Employee of: Roche, N. Mathew: None declared, C. Lin: None declared, A. Jahreis: None declared, K. Sarsour Shareholder of: Roche, Employee of: Roche

THU0110 Effectiveness of Biologic and Nonbiologic Antirheumatic Drugs on Anemia in Adult Patients (PTS) with Rheumatoid Arthritis (RA): New Evidence from Real-World Data: Table 1.

Background Anemia, a common complication of RA, has a negative impact on RA symptoms and QOL and is associated with a more severe form of RA. Roles of biologic DMARDs (BDMARDs), including the IL-6 pathway–inhibiting biologic tocilizumab (TCZ), in the pathophysiology of anemia have not been well explored at the population level. Objectives Evaluate the impact of treatment (Tx) with nonbiologic (NB) and BDMARDs on hemoglobin (Hb) and hematocrit (Hct) in a real-world setting. Methods Using the GE EMR database, a longitudinal cohort study was conducted on 153,672 adult RA pts in the US who had ≥6-mo follow-up. 4 mutually exclusive DMARDTx groups (TxGs) were identified: TCZ, tofacitinib (TOFA), other BDMARDs (OBDMARD), and other NBDMARDs (ONBDMARD). To evaluate changes in Hb and Hct at 6 and 12 mo from index date (ID) in different TxGs, unadjusted and adjusted mean (95% CI) changes were obtained, adjusting for age, sex, smoking status, RA duration, concomitant medications, CRP, and other covariates. Separate analyses were conducted for pts with Hb <12/13.5 g/dL for women/men (anemic). Results Demographics of the cohort were: 55 y mean age, 76% female, 59% white, 57% current or ex-smokers, 44% obese (mean BMI 30 kg/m2). 2% of pts had ischemic heart disease, 2% renal disease, and 7% diabetes. Mean Hb and Hct levels at ID across TxGs were 13.3 g/dL and 40%. Average CRP (1.6 mg/dL) and ESR (23 mm/h) were clinically similar across TxGs at ID. Adjusting for covariates, increases in Hb and Hct levels at 6 and 12 mo were significantly higher in the TCZ group vs the other TxGs (Table). Among anemic pts at ID, increases in Hb and Hct at 6 mo were significantly higher in the TCZ group vs the other TxGs. Significantly greater reductions in ESR and CRP levels over all follow-up periods were observed in the TCZ group vs the other TxGs.Table 1. Adjusted mean (95% CI) change in study parameters at 6 and 12 mo after initiation of Tx TxG Follow-up, mo All, n Anemia at ID, n [%] Δ Hb, g/dL Δ Hct, % All Anemia at ID All Anemia at ID TCZ 6 2,008 384 [19] 0.22 (0.18, 0.27) 0.59 (0.47, 0.71) 0.38 (0.24, 0.52) 1.18 (0.81, 1.54) 12 1,470 267 [18] 0.23 (0.17, 0.29) 0.73 (0.57, 0.90) 0.49 (0.32, 0.67) 1.45 (1.00, 1.90) TOFA 6 900 171 [19] −0.10 (−0.16, −0.03) 0.23 (0.05, 0.41) −0.34 (−0.54, −0.14) 0.57 (0.03, 1.10) 12 412 74 [18] 0.03 (−0.09, 0.15) 0.28 (−0.12, 0.69) −0.03 (−0.37, 0.31) 1.07 (−0.10, 2.25) OBDMARD 6 41,881a 7,982 [19] 0.08 (0.07, 0.09) 0.37 (0.35, 0.39) 0.18 (0.16, 0.21) 0.76 (0.70, 0.82) 12 35,632b 6,764 [19] 0.08 (0.07, 0.09) 0.46 (0.44, 0.49) 0.20 (0.18, 0.23) 1.02 (0.95, 1.10) ONBDMARD 6 59,307 11,922 [20] −0.05 (−0.05, −0.04) 0.25 (0.24, 0.27) −0.11 (−0.13, −0.09) 0.53 (0.48, 0.59) 12 47,898 9,627 [20] −0.05 (−0.06, −0.04) 0.34 (0.32, 0.36) −0.08 (−0.11, −0.06) 0.74 (0.67, 0.80) a95% TNF-α inhibitors. b96% TNF-α inhibitors. Conclusions This study suggests significant increases in Hb and Hct levels after TCZ therapy vs other nonbiologic and biologic therapies in RA pts with anemia, in line with the potential role of IL-6 in RA pts with anemia. Further study is needed to better evaluate the role of IL-6–based OBDMARDs as a Tx option for RA pts with anemia. Disclosure of Interest S. Paul Grant/research support from: Genentech, AstraZeneca, Novo Nordisk, Consultant for: GI Dynamics, Speakers bureau: AstraZeneca, GI Dynamics, Novo Nordisk, K. Klein Grant/research support from: Genentech, AstraZeneca, Novo Nordisk, Consultant for: GI Dynamics, J. Best Shareholder of: Roche, Employee of: Genentech, S. Gale Shareholder of: Roche, Employee of: Roche, A. Pethö-Schramm Employee of: F. Hoffmann-La Roche, K. Sarsour Shareholder of: Roche, Employee of: Roche