Treatment Patterns Among Patients with Psoriatic Arthritis Treated with a Biologic in the United States: Descriptive Analyses from an Administrative Claims Database

BACKGROUND: In patients with psoriatic arthritis (PsA), limited data exist regarding patterns of biologic therapy use. OBJECTIVE: To examine treatment patterns and therapy modifications in U.S. patients with PsA receiving a tumor necrosis factor inhibitor (TNFi) or an anti-interleukin (IL)-12/23 inhibitor. METHODS: Adults with PsA who newly initiated a biologic therapy (index biologic) between January 1, 2013, and January 31, 2015, were included from the Optum Research Database. Biologic therapies comprised those that were approved by the FDA for the treatment of PsA at the time of the study initiation (adalimumab, certolizumab pegol, etanercept, golimumab, infliximab, or ustekinumab). Outcomes included adherence, persistence, and discontinuation of the index biologic; initiation of adjunctive medications (nonbiologics, including those commonly used for pain and/or inflammation); and dose escalation of the index biologic during the 12-month follow-up period. RESULTS: Of the 1,235 patients included, 52.5% were female, and mean (SD) age was 50.3 (12.1) years. The mean (SD) duration of persistence with a newly initiated index biologic (etanercept [48.1%], adalimumab [24.0%], infliximab [10.4%], golimumab [8.3%], ustekinumab [7.2%], or certolizumab pegol [2.0%]) was 246 (128) days; 44.5% of patients persisted with the index biologic for ≥ 12 months. During the 12-month follow-up period, 22.9% of patients switched to a different biologic, 26.8% discontinued without switching or restarting, and 5.8% discontinued and restarted the index biologic. Of the 1,010 patients who persisted with the index biologic for > 90 days, 45.6% received ≥ 1 adjunctive medication during the period from 90 days after the index date to the end of persistence or 12 months. The most commonly initiated adjunctive medications were corticosteroids (22.0%), opioids (17.1%), and nonsteroidal anti-inflammatory drugs (12.9%). Overall, 9.6% of patients had a dose escalation of the index biologic in the immediate 12-month post-index period. CONCLUSIONS: This real-world study of treatment patterns for PsA, which used a large U.S. claims database, demonstrated that the majority of patients with PsA discontinued their index biologic (TNFi or anti-IL-12/23 inhibitor) before 12 months. Nearly half of patients initiated an adjunctive medication, many of which were pain and conventional anti-inflammatory medications.

• Treatment persistence is important for achieving optimal outcomes in patients with psoriatic arthritis (PsA) receiving biologic therapy. • Updated estimates are needed of treatment adherence and persistence that reflect new therapies approved for PsA, as well as studies of dose escalation of biologic therapy and adjunctive modifications.

What is already known about this subject
• The majority of patients with PsA (56%) discontinued their index biologic (either a tumor necrosis factor inhibitor [TNFi] or an anti-interleukin [IL]-12/23 inhibitor) before 12 months; the mean duration of persistence was approximately 8 months. • Nearly half (45%) of the patients initiated an adjunctive medication, many of which were pain and anti-inflammatory medications. • Outcomes from this study of real-world treatment patterns for PsA from a single U.S. claims database suggest that in this population of patients with PsA, additional options for concomitant therapies or alternatives to TNFis and anti-IL-12/23 inhibitors may be important.

Treatment Patterns Among Patients with Psoriatic Arthritis Treated with a Biologic in the United States: Descriptive Analyses from an Administrative Claims Database
using the International Classification of Diseases, Ninth/Tenth Revision, Clinical Modification (ICD-9/10-CM) codes (ICD-9-CM code 696.0; ICD-10-CM codes L40.50, L40.51, L40.52, L40.53, L40.54, and L40.59). Patients had to have ≥ 2 non-rule-out diagnoses of PsA ≥ 30 days apart during the 12 months before the index date (baseline period) through 12 months after the index date, with the first diagnosis occurring during the baseline period or on the index date. The index date was defined as the date of first pharmacy fill or medical infusion of the index biologic. Only patients with no evidence of biologic use in the prior 12 months were included. Patients who had claims for ≥ 2 different biologics on the same day during the 12-month baseline period through the 15-month follow-up period were excluded. Patients had to be continuously enrolled in a commercial or Medicare Advantage health plan for 12 months before (baseline period) and 15 months after (follow-up period) the index date. The 12-month baseline period was used to assess patient characteristics and demographics. In the 15-month followup period, only the first 12 months post-index were used to assess outcomes (including biologic persistence, discontinuation, switching, dose escalation, and initiation of adjunctive medications), and the last 3 months were allowed for time to observe gaps in therapy following fills received at the end of the 12-month follow-up period (Appendix A, available in online article). To reduce confounding, the addition of adjunctive medications and dose escalations were identified only in patients who persisted with their index biologic for >90 days.
The efficacy of biologics is generally not assessed in the first 3 months because of the time needed to see a clinical response in some patients; therefore, the addition of adjunctive medications or dose escalations in the first 3 months may not be related to lack of efficacy. The add-on period was defined as the period from 90 days after the index date to either the end of persistence or 12 months post-index (whichever came first).

Study Outcomes
Patient characteristics recorded at the index date included age, sex, geographic region, and insurance type. A baseline comorbidity score (Quan-Charlson Comorbidity Index) based on the presence of diagnosis codes on medical claims in the baseline period was calculated, 39,40 and comorbid conditions were recorded. Baseline comorbid conditions were defined using the Quan-Charlson Comorbidity Index.
The biologic treatment was identified on the index date, and the biologics approved by the FDA for the treatment of PsA at the start of the study included adalimumab, certolizumab pegol, etanercept, golimumab, infliximab, and ustekinumab. Adherence and persistence were measured in the follow-up period using claims for the index biologic, inclusive of fills or infusions on the index date. Adherence was measured as the proportion of days covered (PDC) in 12 months, calculated as the number of days covered by prescription claims for the prognostic indicators, synthetic disease-modifying antirheumatic drugs (DMARDs) and biologic DMARDs are recommended. 4,12,13 Biologics approved for the treatment of PsA include the tumor necrosis factor inhibitors (TNFis) adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab; 14-21 the human anti-interleukin (IL)-12/23 monoclonal antibody ustekinumab; 22 the selective T-cell costimulation modulator abatacept, 23,24 approved by the U.S. Food and Drug Administration (FDA) for the treatment of PsA in 2017; and the anti-IL-17A monoclonal antibodies secukinumab and ixekizumab, [25][26][27][28] which were approved by the FDA for the treatment of PsA in 2016 and 2017, respectively.
Treatment persistence is important for achieving optimal outcomes in patients with PsA who are receiving biologic therapy. 29 Although standard dosing of biologic therapies is effective for many patients, dose escalations, adjunctive medications, and switching between biologic therapies are often necessary for inadequately controlled pain and inflammation. 4,[30][31][32][33] However, real-world studies examining biologic therapy use in U.S. patients with PsA are limited, especially those involving treatment modifications. Recently published studies have examined biologic therapy treatment patterns, 32,34-38 but updated estimates of treatment adherence and persistence that reflect new therapies approved for PsA, and studies of dose escalation of biologic therapy and adjunctive medications, are still needed. The objective of this study was to describe treatment patterns and therapy modifications in U.S. patients who newly initiated an approved FDA biologic therapy for PsA at the time of the study (either a TNFi or an anti-IL-12/23 inhibitor).

■■ Methods Data Source
This retrospective, observational study used U.S. administrative claims data obtained from the Optum Research Database (ORD) between January 1, 2012, and April 30, 2016. The ORD includes commercial and Medicare Advantage health plan members. In 2015, data were available for approximately 13.5 million individuals with both medical and pharmacy benefit coverage. Claims for pharmacy services are submitted electronically by the pharmacy when prescriptions are filled. Pharmacy claims data include National Drug Code numbers, dosage form, strength, fill date, days of supply, and cost information. Because this study used only de-identified patient records and did not involve the collection, use, or transmittal of individually identifiable data, it is not subject to the U.S. Department of Health and Human Services regulation for the protection of human subjects.

Study Design and Patient Selection
This study included patients aged ≥ 18 years who initiated therapy with an FDA-approved biologic for PsA between January 1, 2013, and January 31, 2015. Patients with PsA were identified Treatment Patterns Among Patients with Psoriatic Arthritis Treated with a Biologic in the United States: Descriptive Analyses from an Administrative Claims Database index biologic divided by 365 days. Persistence was measured as the number of days from the index date to the date of discontinuation of the index biologic or switching to a different biologic. Discontinuation was defined as a gap in therapy of > 90 days based on clinical expert opinion; the date of discontinuation was defined by the run-out-of-days supply of the last prescription filled before the gap in therapy.
Postpersistence treatment patterns were evaluated for all patients who were not persistent for 12 months and were classified as (a) discontinuing the index therapy without switching or restarting; (b) discontinuing and restarting the index therapy; and (c) switching to a different biologic. Among patients who switched to a different biologic, the biologic that a patient was switched to was identified, and the number of days from the index date to the switch date (defined as the earliest claim that indicates a switch after the index date) was calculated.
For patients who were persistent with the index biologic for > 90 days, initiation of adjunctive medications and dose escalation of the index biologic were identified from 90 days after the index date until the end of persistence or 12 months post-index (add-on period). Patients had to have no claims for the adjunctive medication during the first 90 days after the index date for it to be considered an adjunctive medication. The adjunctive medications included conventional synthetic DMARDs (csDMARDs), NSAIDs, opioids, corticosteroids, antidepressants, anxiolytics, targeted synthetic DMARDs, sleep aids, and topical   analgesics. Dose escalation was defined as a patient receiving a dose > 110% of the reference dose from the product label for ≥ 90 continuous or intermittent days. The time-to-dose escalation was measured as the number of days from the index date until the 90th day with a dose > 110% of the reference dose.

Statistical Analyses
All data were analyzed descriptively. Patient-level analyses included baseline demographics, number of patients initiating each biologic, and number of patients who discontinued or switched treatments. Subgroup analyses by index biologic reported the mean (standard deviation [SD]) duration of time patients persisted on the index biologic, time to switch to a different biologic, number of patients with dose escalation of the index biologic, and number of patients who initiated adjunctive medications.
Overall, 44.5% of patients persisted with the index biologic for ≥ 12 months (Table 2); infliximab had the highest percentage of patients persistent at 12 months (61.2%) and certolizumab pegol had the lowest (32.0%). Among the total patients, 55.5% were not persistent with the index biologic at 12 months: 26.8% discontinued without switching or restarting, 5.8% discontinued and restarted the index biologic, and 22.9% switched to a different biologic ( Table 2). Among patients who switched to a different biologic, the mean (95% confidence interval) time to switch was 177.4 (166.6-188.2) days (approximately 6 months; Table 2). The most common biologic therapy that patients were switched to was adalimumab (48.4%), followed by etanercept (14.5%; Figure 2).

Treatment Patterns Among Patients with Psoriatic Arthritis Treated with a Biologic in the United States: Descriptive Analyses from an Administrative Claims Database
In the present study, the mean baseline Quan-Charlson Comorbidity Index score was 0.73 as opposed to 0.52 in the study by Zhang et al. (2014), which showed a 12-month persistence of 54%. 34 Furthermore, the higher comorbidity burden observed in patients with certolizumab pegol as their index biologic may have contributed to their lower persistence relative to patients receiving other biologics. Lower 12-month persistence with biologic treatment in the present study may also be related to changes in costs or insurance coverage of biologics, health care provider prescribing patterns, and patient preference, or to lack of efficacy, adverse effects, and insufficient pain control. 30,32,38 The proportion of patients who switched biologics in the present study (22.9%) was similar to other studies. 34,35,38 Although this study did not capture the reason patients switched biologics, a previous study of patients with PsA showed that adverse effects and lack of efficacy were the most common reasons patients switched their TNFi. 43 Additionally, it is possible that some patients may discontinue biologic therapy because of minimal symptoms of PsA, although it is unlikely that the high proportion of patients switching biologic therapy or initiating adjunctive anti-inflammatory or pain medications had adequate symptom control.
(subcutaneously administered), 5.3% had a dose escalation. The mean (SD) number of days from index date to dose escalation was 192 (75) days.

■■ Discussion
In this descriptive, administrative claims-based study, the majority of patients with PsA (55.5%) discontinued their index biologic before 12 months, with a mean persistence of approximately 8 months. Among patients who persisted for > 90 days, 45.6% received ≥ 1 adjunctive medication during the period from 90 days after the index date to the end of persistence with the index biologic or 12 months, and 9.6% had a dose escalation of the index biologic in the immediate 12-month post-index period.
The percentage of patients persistent at 12 months in this study (44.5%) was lower than in other studies; previous U.S. studies showed 12-month persistence with biologic therapy ranging from 51% to 66% in patients with PsA. 32,34,35,41 One reason for the difference in persistence among the studies may be related to varying severity of comorbidities in the patients studied. Comorbidities may affect the safety of and response to biologic therapy and lead to discontinuation or switching of the biologic. 42 adjunctive DMARDs and not other additional medications. 34,38 Notably, the majority of claims for adjunctive medications in this study were for treatment of pain and inflammation (corticosteroids, 22% of patients; opioids, 17%; NSAIDS, 13%), which suggests that some patients have inadequate disease control while receiving biologic therapy. Similarly, a previous study that compared use of concomitant medications during the 12 months before and after initiating a biologic found that 46.1% of patients were receiving pain medication at 12 months after starting biologic therapy. 35 Currently, no recommendations exist for the use of adjunctive medications for patients with PsA who have an inadequate response to biologics alone or with the addition of csDMARDs. It is also important to note that long-term administration of systemic corticosteroids is neither recommended nor supported by evidence for patients with PsA, but short-term use may be indicated for rescue therapy. 46 In addition to pain and anti-inflammatory medications, 9.7% of patients received either an antidepressant or anxiolytic. More research is needed to better understand why some patients with PsA require additional pain and mental health medications.
Although the reasons for dose escalation are not entirely understood, they may involve the availability of insurance approval for dose escalations and caution in administering the initial doses. Notably, a larger proportion of patients receiving an intravenously administered biologic had dose escalation than patients receiving a subcutaneously administered biologic. Infliximab also had the highest 12-month persistence in the present study.
High rates of discontinuation and switching of biologic therapies, along with high rates of dose escalation, suggest a high frequency of suboptimal biologic experience in patients with PsA. The reasons for more frequent dose escalations are unclear, but possible reasons include dosing instructions, caution in administering the initial doses, increased flexibility of IV infusion, and insurance approval policies. Furthermore, PsA is a heterogeneous disease, and treatment responses to biologics can vary substantially between patients.
More research to guide patients and health care providers in making decisions about choosing first-and subsequent-line biologics is needed. In addition, more research is needed to evaluate differences between lines of TNFi therapies. A recent systematic literature review of 18 real-world studies, most of which were outside the United States, reported limited evidence comparing different lines of therapy; however, first-line TNFi therapy was found to yield better results than secondline TNFi therapy, and significant differences were observed between second-and third-line TNFis. 44 Moreover, recent studies have shown that biologic-naive patients were more persistent with their first biologic than were patients with a history of biologic use. 38,45 Additional treatment options may be beneficial to address the heterogeneity of the disease and patients' responses to treatment.
To our knowledge, this study is one of a limited number that have examined adjunctive medications for patients with PsA initiating a biologic therapy. In the findings presented here, a large percentage of patients modified their treatment. Use of adjunctive DMARDs was similar between this study and previous studies; however, the previous studies only evaluated   The relationship between dose escalation and persistence is unclear. Dose escalation requires time for insurance approval and may provide more effective dosing; therefore, these factors may have contributed to longer persistence. In addition, patients receiving infliximab may have had higher rates of persistence because infliximab is administered by a health care provider. Persistence may also be influenced by the administration of infliximab every 8 weeks as opposed to every 1-4 weeks with subcutaneously administered biologics.
With the emergence of additional non-TNFi treatment options for PsA, future studies will be needed to examine the impact of these newer therapies on treatment patterns and therapy modifications in patients with PsA.
The strengths of the present study include the use of data from a large database covering all regions of the United States, which provides valuable real-world information. In addition, these findings expand on previous studies by examining the use of adjunctive medications in patients receiving a biologic for PsA.

Limitations
This study was limited by its retrospective design and includes observational analyses of administrative claims databases with descriptive findings only. The claims data also did not capture the reasons for switching, discontinuation, dose escalation, or adding adjunctive therapy; thus, it is unknown how various factors influence these treatment decisions/patterns. Because the study was limited to 12 months, longer-term treatment patterns are unknown.
Furthermore, this study was limited to patients with commercial health coverage or Medicare Advantage, and 49% of patients were from 1 U.S. region (South); it is unclear if these findings can be extended to patients in all regions of the United States and to those with Medicaid or no insurance. Finally, this study could not consider the potential effects of rebates, discounts, or other price concessions on choice of biologic therapy, and utilization rates may differ based on the formulary of each payer/employer group.

■■ Conclusions
The majority of patients with PsA discontinued their index biologic (TNFi or anti-IL-12/23 inhibitor) before 12 months, and mean persistence was approximately 8 months. Nearly half of patients initiated an adjunctive medication, many of which were pain and anti-inflammatory medications. Insufficient control of symptoms may lead patients to discontinue biologic therapy, which can contribute to disease progression. Further research is necessary to routinely monitor patient outcomes; to understand the reasons patients with PsA discontinue, switch, or modify biologic therapy; and to assess the impact of insufficient control of pain and inflammation.

DISCLOSURES
This study was sponsored by Novartis Pharmaceuticals. Optum was commissioned by Novartis to conduct this study, but employment was not contingent on results of the study. Walsh is a paid consultant for Novartis. Adejoro was an employee of Optum at the time of the study and writing of the manuscript.
Chastek is an employee of Optum. Palmer and Hur are employees of Novartis.
Results of this study were presented as an abstract and poster at the Academy of Managed Care Pharmacy Nexus 2017; October 16-19, 2017; Dallas, TX; and the EULAR 2017 Annual European Congress of Rheumatology; June 14-17, 2017; Madrid, Spain.