Real‐world, long‐term treatment patterns of commonly used biologics in Canadian patients with moderate‐to‐severe chronic plaque psoriasis

Abstract Real‐world and long‐term data on biologic treatment changes – including switching, discontinuation, dose escalation, and interval change (both increasing and decreasing) – are required to understand treatment patterns for psoriasis (PsO) in Canada. The study objectives were to evaluate the time to first biologic treatment change and to document these changes in Canadian patients with moderate‐to‐severe chronic plaque PsO. Charts from 13 Canadian sites were queried retrospectively (2005–2019); a period covering all biologic classes commonly used for PsO in Canada. Included were patients diagnosed with, and currently using biologics for, moderate‐to‐severe chronic plaque PsO. Time to first treatment change, nature of treatment change, number of lines of treatment, proportion of patients on each drug, and drug survival were collected. Based on 1149 medical charts, adalimumab had the longest time to first treatment change (49.3 months; 95% confidence interval, 37.4–67.4). Approximately half of the patients had a treatment change, and nearly 75% of these changes were either an interval change or a biologic switch. Lack of efficacy was the most prevalent primary reason for biologic switch (67.3%), whereas 6.7% of patients switched due to adverse events. Drug survival for etanercept and infliximab was approximately twice as long for patients who had dose optimization (i.e., dose escalation or interval change) than patients without dose optimization. The survival curve of adalimumab was similar to the one of ustekinumab after 48 months of treatment, showing approximately 60% of patients remaining on treatment after 132 months, with or without dose optimization. Assessing treatment patterns of all commonly used biologics for moderate‐to‐severe chronic plaque PsO in Canada between 2005 and 2019 showed that approximately half of the patients required a treatment change (mainly interval change or biologic switch) while the other half remained on treatment.


| INTRODUC TI ON
Biologics have advanced our understanding and the treatment of psoriasis (PsO) therapy due to their efficacy and precise mechanisms of action. [1][2][3] Despite the demonstrated efficacy of biologics, ≥30% of patients show an inadequate response to these agents. 1,4,5 Treatment modifications, including dose escalations, dose reductions, switches, discontinuations, and restarts, are to be expected in the management of PsO. [6][7][8] In clinical practice, modifications of dosing regimens, [9][10][11][12][13] intermittent therapy, or interruption followed by retreatment 14,15 have been reported to impact treatment effectiveness. Biologic switches in the treatment of PsO have been evidenced to be mostly due to a lack of efficacy, to adverse events (AE) to a lesser extent, or to efforts to achieve better clinical response. 6,10,[15][16][17][18][19] Specific to Canadian real-world practices, off-label regimens are less likely to include biologics dose reductions or interval increases compared with other practices worldwide such as European practices. 11 Drug survival, defined as the duration of time from therapy initiation to discontinuation, is a proxy measure for drug effectiveness, safety, and tolerability. Predictors of biologic drug survival have been reported in specific studies as female sex, 20 psoriatic arthritis (PsA), [21][22][23] dose escalation, 24,25 and previous exposure to biologics. 21 Gradual loss of efficacy has been shown to limit biologic drug survival, 26 and several studies have reported ustekinumab as having the highest survival rate. 19,25,[27][28][29][30][31] Given the paucity of data on biologic treatment patterns for Canadian PsO patients, real-world, long-term data are needed, including data on recently approved biologics, although less extensive results are available on these biologics. The primary objective of this study was to evaluate retrospectively, in a real-world setting, the time to first treatment change -defined as switching, discontinuation, dose escalation, and interval change (both increasing and decreasing) -for commonly used biologics in Canadian patients with moderate-to-severe chronic plaque PsO. As a secondary objective, these treatment changes were documented in terms of number, types and reasons for changes, sequence of agents used as well as drug survival. These results may help identify effective therapies to clear the skin of PsO patients while minimizing treatment changes.
Retrospective data from 2005 to 2019 were obtained from 13 Canadian sites. The date range was selected to include the abovementioned biologics since their approval in Canada for treating PsO.
Sites were selected based on chart availability and relevance of geographical representation. Intrinsic to the nature of this study was the variability in the number of years since approval for these biologics, thus leading to more patients having "ongoing treatments" with the most recent biologics approved, namely no first treatment change at the time of conducting the study. The study was conducted from July 2019 to June 2020.

| Participants
Included in this study were charts of patients who were ≥18 years of age (or as per local legal age) at initiation of their first PsO biologic therapy and who were diagnosed with -and currently using biologics as primary indication for -moderate-to-severe chronic plaque PsO. In addition, the first therapy had to be initiated as per label at the starting dose, with the most recent therapy lasting ≥1 year. There were no exclusion criteria. The study was conducted in compliance with local laws and regulations and Good Pharmacoepidemiology Practices, and was approved by central and local ethic boards prior to initiation.

| Variables and data sources
To minimize chart selection bias, sites were requested to pull charts from the latest consecutive patients seen and who fit the inclusion criteria. Of charts meeting the inclusion criteria, patient demographics, PsO history, and comorbidities were retrospectively collected as well as data on PsO treatments available since 2005: treatment name, dose prescribed, administration mode, dosing start and stop dates, and reason for discontinuation or treatment change, if any.
Time to treatment change was calculated based on the start and stop/change dates. Stop dates for ongoing treatments at the time of chart review were censored at the chart review date. If patients participated in clinical trial(s) for PsO during the period covered by this study, such period(s) was/were excluded from the study.
In this study, treatment changes were defined as either switching, discontinuation, dose escalation, or interval change. Interval changes included both increasing and decreasing, which were more likely to be interval decreases than the opposite in Canadian realworld practices. Information on biologics on-label dosing and year of market entry for PsO treatment in Canada is provided in Table S1.

| Study size and statistical methods
Sample size considerations were based on time to first treatment change -the study primary end-point. Using published drug survival times for biologic treatments in PsO (23-38 months), 20,32 time to first treatment change was assumed to be 30 months. Assuming a censor rate (proportion of patients for whom treatment change was not observed) of 30%, the sample size was calculated to provide an estimate of the median time to first treatment change within a predefined confidence interval (CI) (≥23 and ≤39 months).
The primary end-point was analyzed and plotted using Kaplan-Meier (KM) estimates (reported with 95% CI). The secondary endpoints, time to treatment change as second line of treatment, and drug survival -the number of months until drug discontinuation -were analyzed using the same KM estimates as for the primary end-point. The number of biologics per patient was summarized as mean, median, standard deviation (SD), minimum, and maximum.
These analyses were performed on the full analysis set (FAS; n = 1149) population, defined as patients who met all inclusion criteria and had been treated for PsO with any of the biologics for at least 1 year. Patients whose most recent biologic treatment was <1 year, but the combined length of treatment for all biologics was ≥1 year were also included, even if not meeting protocol entry criterion. As a sensitivity analysis, the primary end-point was analyzed using the per-protocol set (PPS; n = 1059), defined as all patients of the FAS who met all inclusion criteria for the study.
Missing data were not imputed, except for the biologic treatment start/stop date. Treatments missing the start or stop month were excluded from analyses. When days of treatment were missing, the first day of the month was imputed. McDougall Scientific performed the calculations and statistical analyses using SAS version 9.4.

| Participants
A total of 1247 charts were queried from 13 Canadian sites. Of those, 1149 (92.1%) patients were included in the FAS population and 1059 (84.9%) in the PPS population, meeting all protocol inclusion criteria. The patients' mean (SD) age was 53.9 years (13.6) ( Table 1). Approximately half of the patients were male (58.0%) and most were White (75.0%). During their PsO treatment, 71.2% of patients had been exposed to at least one IL inhibitor and 53.0% to a TNFα inhibitor (Tables S2 and S3). At baseline, 25.8% of patients had PsA, and 35.2% had no reported comorbidities.

| Number and sequence of biologic treatments for PsO
Overall, the mean (SD) number of biologic treatments for PsO  46.9% taking TNFα inhibitors to 6.9% taking IL-23p19 inhibitor), and the higher the number of treatment lines (up to six drug class lines for patients who initiated treatment with TNFα inhibitors) (Table S1, Figure 3).

| Drug survival
Etanercept and infliximab showed a similar median drug survival (32-34 months) ( Figure S3), which was higher for patients who had   However, it is also recognized that not all switches lead to condition improvement, and additional data are required to make more specific recommendations. 7 On a total population basis, our findings on lack of efficacy show similarities with those reported by Kishimoto et al. 25 with regards to the TNFα inhibitors and anti-IL-17 agents, but are higher with regards to the IL-12/23 inhibitor (ustekinumab). Biologic treatment switches and discontinuations due to AE have been reported in a real-world setting to be roughly between 2% and 6%, 15,19,23,38,39 similar to our findings. Also similar to our findings is the higher proportion of patients on TNFα inhibitors who switched/discontinued treatment due to AE versus other biologic classes, which has been attributed more frequently to infliximab in some publications. 25 11,41,42 this study provides long-term data on four classes of biologics, helping to close a knowledge gap on treatment pattern practices in a real-world setting in patients with moderate-to-severe chronic plaque PsO.