Key Points
Question
What proportion of patients with psoriasis who were taking a biologic and registered to the British Association of Dermatologists Biologics Intervention Register (BADBIR) would have been eligible for the pivotal licensing trials of biologics for psoriasis?
Findings
In this observational study of 7136 patients with psoriasis, those identified as ineligible for clinical trials had lower effectiveness and higher rates of serious adverse events while receiving biologic therapy in the first 12 months compared with patients identified as eligible.
Meaning
Clinical trial findings of biologic therapies for psoriasis are not representative of real-world patients who would have been excluded from such trials.
This observational study examines whether drug discontinuation, effectiveness, and rates of serious adverse events differ in patients with psoriasis being treated with biologics who have been categorized as eligible or ineligible for clinical trials.
Abstract
Importance
Patients with psoriasis enrolled in clinical trials of biologics may not be representative of the real-world population. There is evidence that patients ineligible for such trials have a greater risk of serious adverse events (SAEs), but the effect on drug discontinuation and effectiveness are unknown.
Objective
To determine whether (1) drug discontinuation, (2) effectiveness, and (3) rates of SAEs differ in patients with psoriasis categorized as eligible or ineligible for clinical trials.
Design, Setting, and Participants
An observational study using 157 dermatology centers in the United Kingdom and Republic of Ireland was carried out wherein we applied the eligibility criteria of clinical trials of biologic therapies for psoriasis to patients who were being followed up in the British Association of Dermatologists Biologic Interventions Register (BADBIR) and being prescribed biologics as part of standard clinical care. Patients with psoriasis registered to BADBIR who were taking etanercept (enbrel only; n = 1509), adalimumab (n = 4000), and ustekinumab (n = 1627) with at least 1 follow-up visit. Eligibility criteria were extracted from phase 3 licensing trials for etanercept, adalimumab, and ustekinumab for the treatment of moderate to severe psoriasis. Patients in BADBIR with a missing baseline Psoriasis Area and Severity Index (PASI) or baseline PASI value less than 10 (etanercept) or less than 12 (adalimumab; ustekinumab) but who would otherwise be eligible were investigated separately. Eligibility categories applied to BADBIR included: eligible, ineligible, insufficient baseline PASI only, and missing baseline PASI only.
Main Outcomes and Measures
(1) Drug discontinuation: cumulative incidence at 12 months by stop reason per eligibility category and drug; (2) effectiveness: linear regression of absolute change in PASI from baseline to 6 and 12 months; and (3) SAEs: incidence rate ratio (IRR) at 12 months between eligibility categories per drug.
Results
The mean (SD) age of the 7136 patients included in the analysis was 45 (13) years and 2924 (41%) were women and 4212 (59%) were men. Of 7136 patients, 839 (56%) etanercept, 2219 (56%) adalimumab, and 754 (46%) ustekinumab registrations were categorized as eligible. The most common reasons for ineligibility were diabetes (etanercept, 143 [9%]; ustekinumab, 201 [12%]) and nonchronic plaque psoriasis (adalimumab, 157 [4%]). Patients categorized as ineligible (etanercept, 367 [24%]; adalimumab, 282 [7%]; ustekinumab, 394 [24%]) achieved a smaller absolute change in PASI after 6 and 12 months (adalimumab, ustekinumab), and had significantly higher rates of SAEs compared with the eligible category (etanercept: IRR, 1.9; 95% CI, 1.4-2.6; adalimumab: IRR, 2.0; 95% CI, 1.5-2.6; ustekinumab: IRR, 2.8; 95% CI, 2.1-3.8). No significant differences in drug discontinuation were observed between categories.
Conclusions and Relevance
Clinical trial effectiveness and safety outcomes are not representative of real-world patients in BADBIR patients categorized as ineligible for such trials.
Introduction
Biologic therapies have revolutionized the treatment of moderate-to-severe psoriasis. However, stringent eligibility criteria applied in clinical trials of biologic therapies for psoriasis can result in selection bias and uncertainty of the external validity to real-world psoriasis populations. This is important because estimating the gap between efficacy and effectiveness is fundamental to evidence-based, cost-effective prescribing.
There are limited data quantifying the differences between clinical trial findings and clinical practice for biologic therapies in the treatment of psoriasis. One approach to investigating relevance of trial data are to examine the impact of eligibility criteria applied to a real-world population of patients with psoriasis; 29.8% patients registered in the Spanish Registry of Adverse Events Associated With Biologic Drugs in Dermatology (BIOBADADERM) would be ineligible for clinical trials using common eligibility criteria identified from a meta-analysis of randomized clinical trials investigating systemic and biologic treatments for psoriasis.1,2 An increased risk of serious adverse events (SAEs) was detected for ineligible patients in the first 12 months of treatment (incidence rate ratio [IRR], 2.6; 95% CI, 1.5-4.5), indicating that the safety of real-world patients is not adequately represented in clinical trials.1 Differences in drug discontinuation between eligible and ineligible patients are yet to be determined, and the gap between efficacy and effectiveness is also unknown.3
The British Association of Dermatologists Biologics Intervention Register (BADBIR) is a prospective pharmacovigilance register of patients with psoriasis recruited from 157 dermatology centers in the United Kingdom and Republic of Ireland.4 The BADBIR is a representative cohort of biologic prescribing in clinical practice enabling the investigation of the gap between clinical trial and real-world outcomes. Our hypothesis was that patients identified as ineligible for clinical trials would have lower effectiveness, and greater rates of drug discontinuation and SAEs compared with eligible patients. The objectives of the study were to:
identify the eligibility criteria applied in phase 3 clinical trials of biologics in psoriasis;
determine the proportion of patients registered to BADBIR who would have been eligible for those trials; and
determine whether drug discontinuation, effectiveness, and the incidence of SAEs in the first 12 months of biologic exposure differ by eligibility criteria.
Methods
Study Design and Setting
Patients registering to BADBIR receiving etanercept (enbrel only), adalimumab, or ustekinumab who completed at least 1 follow-up up to December 1, 2016, were included in the analysis. BADBIR received approval from the North West Research Ethics Committee in March 2007. All patients provided written informed consent in accordance with the Declaration of Helsinki. Participants were not compensated.
Baseline Assessments
Patients were recruited during routine appointments at dermatology centers within 6 months of initiating or switching to a biologic or conventional systemic therapy. All data were recorded onto a web-based database including Psoriasis Area and Severity Index (PASI) and clinical history (comorbidities with year of onset, previous antipsoriatic therapies, concomitant therapies). A PASI value of 10 or more is not required for patients registering on a biologic therapy in BADBIR.4
Follow-up Assessments
Assessments were undertaken by the dermatology team at 6-monthly intervals for the first 3 years, then annually. These included PASI values and changes in therapy for psoriasis (temporary cessation; discontinuation, including the stop reason; dose changes; initiation or switching). Medical records were reviewed for any adverse events since the previous visit. Serious adverse events (SAEs) were untoward medical occurrences that were considered to represent a hazard to the patient, including: death, overnight hospitalization, immediately life-threatening, intravenous antimicrobial administration, loss of function or disability, congenital malformation/birth defect, medically important event. Examples of medically important events include malignant abnormalities and pregnancies. All SAEs were subsequently coded to the Medical Dictionary for Regulatory Activities (MedDRA).
Clinical Trial Eligibility Criteria
Eligibility criteria were extracted from phase 3 licensing trials for etanercept,5,6,7 adalimumab,8,9 and ustekinumab10,11 (eTable 1 in the Supplement); further resources including http://www.clinicaltrials.gov, http://www.amgentrials.com, clinical study reports (AbbVie), and related articles12,13 were searched for additional eligibility criteria. Clinical trial study protocols were requested from AbbVie, Janssen, and Pfizer, but were not made available. Licensing trials rarely exceeded 12 months in duration and the most common primary outcomes were number of adverse events and achievement of PASI 75 at specific time points (eTable 1 in the Supplement).
The eligibility criteria used in the present study are listed in Table 1. It is possible for patients registering to BADBIR to enter a baseline PASI value of less than 10, or have a missing date for a PASI value. Patients in BADBIR would otherwise be categorized as eligible who had a baseline PASI value insufficient for entry into a clinical trial. A missing baseline PASI date, or baseline PASI recorded after the initiation of therapy (also categorized as missing) were investigated separately. The prevalence of 2 or more comorbidities in patients initiating biologic therapy in BADBIR was 54%.14 To identify an approximation for uncontrolled comorbidities (eTable 1 in the Supplement), patients with a diagnosis year matching or preceding the year of consent were designated ineligible. There was difficulty in mapping eligibility criteria from the trials to BADBIR. Criteria that were not applied to BADBIR included: washout periods (eligible participants would be washed out if commencing a clinical trial), pregnant or breastfeeding women (data not collected), and immunocompromised participants (adalimumab only).
Table 1. Eligibility Status.
| Criteriaa | No. (%) | ||
|---|---|---|---|
| Etanercept (n = 1509) |
Adalimumab (n = 4000) |
Ustekinumab (n = 1627) |
|
| Eligible | 839 (56) | 2219 (56) | 754 (46) |
| Insufficient baseline PASI only | 158 (10)b | 1166 (29)c | 370 (23)c |
| Missing baseline PASI only | 145 (10) | 333 (8) | 109 (7) |
| Ineligible | 367 (24) | 282 (7) | 394 (24) |
| Age <18 y | 6 (<1) | 19 (<1) | 5 (<1) |
| Nonchronic plaque psoriasis | 86 (6) | 157 (4) | 75 (5) |
| Comorbiditiesd | |||
| Psychiatric | 43 (3) | NA | 51 (3) |
| Myocardial infarction | 2 (<1) | NA | 10 (<1) |
| Angina | 47 (3) | NA | 62 (4) |
| Diabetes/endocrine | 143 (9) | NA | 201 (12) |
| Hypertension | 89 (6) | NA | NA |
| Renal | NA | 9 (<1) | 4 (<1) |
| Gastrointestinal | NA | 22 (<1) | 13 (1) |
| Blood/laboratory values | NA | 6 (<1) | NA |
| Infections | |||
| Recent | 7 (<1) | NA | 6 (<1) |
| HIV/HBV/HCV | 2 (<1) | 6 (<1) | 1 (<1) |
| Latent TB | NA | 39 (1) | 23 (1) |
| Cancers | 8 (<1) | 28 (1) | 2 (<1) |
| Two or more ineligibility criteria | 60 (4) | 7 (<1) | 58 (4) |
Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; NA, not applicable; TB, tuberculosis.
No exclusions based on prior exposure to therapies targeting tumor necrosis factor-α (etanercept; adalimumab) or interleukin 12/23 (ustekinumab).
Baseline PASI values of less than 10.
Baseline PASI values of less than 12.
No reports of multiple sclerosis, congestive heart failure, hepatic, or pulmonary disorders as comorbidities.
In total, 4 eligibility categories were identified: missing baseline PASI only (m-PASI), insufficient baseline PASI only (i-PASI), ineligible owing to satisfying at least 1 of the exclusion criteria in Table 1, and eligible. The criteria applied to patients initiating etanercept, adalimumab, or ustekinumab are summarized in Table 1.
Statistical Methods
Drug discontinuation was determined using the cumulative incidence of patients discontinuing therapy in the first 12 months by drug, eligibility category, and stop reason (adverse events, ineffectiveness, other, missing).
For effectiveness, baseline PASI values were identified if reported within 6 months of the drug start date (−183 to 0 days). The PASI values recorded between 4 to 8 months and 10 to 14 months after initiating therapy were used to determine response to treatment at 6 months and 12 months, respectively. Patients with a baseline, 6-month and 12-month PASI were included in the analysis. A 2-sample t test was used to determine differences between the ineligible and i-PASI categories with eligible patients for the median PASI values at all time points and absolute change at 6 and 12 months. Linear regression was used to identify predictors of absolute change in PASI at 6 and 12 months, with the eligible category as the reference group. The predictors included: ineligible and i-PASI categories, aged 70 years or older, body mass index (calculated as weight in kilograms divided by height in meters squared), 3 or more comorbidities, prevalent psoriatic arthritis, male sex, ever smoked (current or past). Selection bias between patients with complete PASI values and those with missing values at 6 and/or 12 months was investigated by determining whether the proportion of women, median age, and baseline PASI values differed (eTable 4 in the Supplement).
An incidence rate (IR) per 1000 person years was calculated by drug and eligibility category for SAEs reported in the first 12 months following treatment initiation; IRRs were calculated using Poisson regression between the eligibility categories with the eligible category as the reference group. All analyses were performed using Stata statistical software (version 13.1, StataCorp).
Results
Patient Characteristics
A total of 8533 patients were registered to the biologic cohort. Of those, 7408 (87%) completed at least 1 follow-up; 1509 (18%) patients registered to etanercept, 4000 (47%) registered to adalimumab, and 1627 (19%) registered to ustekinumab (eFigure 1 in the Supplement). The eligibility criteria extracted from phase 3 licensing trials were applied to patients registering to each biologic. Common criteria included: chronic plaque psoriasis; patients aged 18 years or older; a PASI value of 10 or higher (etanercept) or 12 or higher (adalimumab, ustekinumab); and exclusions for comorbidities, recent infections, and previous cancer (Table 2).
Table 2. Baseline Characteristics.
| Characteristic | No. (%) | ||
|---|---|---|---|
| Etanercept | Adalimumab | Ustekinumab | |
| Total, No. | 1509 | 4000 | 1627 |
| Age, median (IQR), y | 45 (36-54) | 45 (36-54) | 46 (37-55) |
| Females | 635 (42) | 1645 (41) | 644 (40) |
| Disease duration, median (IQR), y | 21 (13-30) | 21 (13-30) | 21 (12-31) |
| Missing | 12 (1) | 50 (1) | 13 (1) |
| BMI, median (IQR) | 29 (26-34) | 30 (26-34) | 31 (27-36) |
| Missing | 120 (8) | 278 (7) | 134 (8) |
| Baseline PASI, median (IQR) | 14 (11-19) | 14 (11-19) | 15 (11-20) |
| 6 mo PASI, median (IQR) | 4 (2-8) | 2 (0-5) | 2 (0-6) |
| Missing | 558 (37) | 1571 (39) | 649 (40) |
| 12 mo PASI, median (IQR) | 4 (2-7) | 2 (0-5) | 2 (0-4) |
| Missing | 533 (35) | 1482 (37) | 649 (40) |
| Age ≥70 y | 60 (4) | 109 (3) | 68 (4) |
| Disease duration >10 y | 1272 (84) | 3392 (85) | 1375 (85) |
| Concomitant therapy | 280 (19) | 645 (16) | 197 (12) |
| Registration dosea | 977 (65) | 3329 (83) | 859 (53) |
| Missing | 263 (17) | 687 (17) | 239 (15) |
| Biologic-naïve | 1386 (92) | 3372 (84) | 1037 (64) |
| Baseline PsA | 349 (23) | 963 (24) | 329 (20) |
| ≥3 Comorbidities | 107 (7) | 247 (6) | 148 (9) |
| Ever smoked | 772 (51) | 2280 (57) | 1016 (62) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); i-PASI, insufficient baseline Psoriasis Area and Severity Index; m-PASI, missing baseline PASI; PsA, psoriatic arthritis.
Doses reported are etanercept, 50 mg weekly; adalimumab, 40 mg fortnightly; and ustekinumab, 45 mg 12 weekly after loading schedules were complete.
In total, 839 (56%) etanercept, 2219 (56%) adalimumab, and 754 (46%) ustekinumab patients were categorized as eligible (Table 1). Fewer patients were identified as ineligible for adalimumab (282 [7%]; etanercept, 367 [24%]; ustekinumab, 394[ 24%]). Similar proportions of patients had missing baseline PASI values (etanercept, 145 [10%]; adalimumab, 333 [ 8%]; ustekinumab, 109 [7%]). A greater proportion of adalimumab (1166 [29%]) and ustekinumab (370 [23%]) patients had an insufficient baseline PASI value compared with etanercept patients (158 [10%]) owing to the lower PASI threshold identified for the etanercept trials.5,6,7
Patient characteristics differed between drugs. A lower proportion of ustekinumab patients were biologic naive at registration (1037 [64%]; etanercept, 1386 [92%]; adalimumab, 3372 [84%]) and a higher proportion of patients reporting having ever smoked (1016 [62%]; etanercept, 772 [51%]; adalimumab, 2280 [57%]) (Table 2). A higher proportion of etanercept patients were receiving concomitant therapy at registration (280 [19%]; adalimumab, 645 [16%]; ustekinumab, 197 [12%]) (Table 2).
Drug Discontinuation
The cumulative incidence of drug discontinuation was similar between eligibility categories per drug (Table 3). The most common reason for discontinuation in the first 12 months of treatment was ineffectiveness for each eligibility category per drug (etanercept, 11%-15%; adalimumab, 6%-8%; ustekinumab, 3%-5%) with similar proportions stopping for adverse events (etanercept, 2%-4%; adalimumab, 3%-5%; ustekinumab, 2%-3%) (Table 3).
Table 3. Cumulative Incidence of Drug Discontinuation in the First 12 Months of Therapy by Eligibility and Stop Reason.
| Stop Reason | Etanercept, Mean % (95% CI) | Adalimumab, Mean % (95% CI) | Ustekinumab, Mean % (95% CI) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Eligible | Ineligible | i-PASI | Eligible | Ineligible | i-PASI | Eligible | Ineligible | i-PASI | |
| Adverse events | 3 (2-4) | 4 (3-6) | 2 (1-5) | 3 (3-4) | 5 (3-7) | 3 (2-4) | 3 (2-4) | 2 (1-3) | 2 (1-4) |
| Ineffectiveness | 11 (9-13) | 15 (12-19) | 15 (11-20) | 8 (7-9) | 6 (4-9) | 7 (5-9) | 3 (2-4) | 5 (3-7) | 3 (2-5) |
| Other | 4 (3-5) | 2 (1-3) | 4 (2-8) | 2 (2-3) | 5 (3-7) | 3 (2-4) | 2 (1-3) | 2 (1-3) | 2 (1-4) |
| Missing | 4 (3-5) | 6 (4-8) | 8 (5-13) | 3 (2-4) | 2 (1-3) | 2 (1-3) | 1 (0-1) | 2 (1-4) | 2 (1-3) |
Abbreviation: i-PASI, insufficient baseline Psoriasis Area and Severity Index.
Effectiveness
In the adjusted analysis, significantly smaller changes in PASI values were reported at 6 and 12 months for i-PASI patients for all 3 biologics, with significantly smaller changes in PASI values also reported for ineligible adalimumab patients (Table 4). Increased BMI (adalimumab only) and 3 or more comorbidities (ustekinumab only) predicted a smaller absolute change in PASI at 6 or 12 months, whereas male sex predicted a significantly greater absolute change in PASI (etanercept at 6 months; adalimumab at 6 and 12 months) (Table 4).
Table 4. Absolute Change From Baseline PASI at 6 and 12 Months.
| Variable | Etanercept | Adalimumab | Ustekinumab | |||
|---|---|---|---|---|---|---|
| 6 Months | 12 Months | 6 Months | 12 Months | 6 Months | 12 Months | |
| Absolute change in PASI from baseline, median (IQR)a | ||||||
| Eligible | −10.4 (−13.7 to −6.9) |
−10.8 (−14.3 to −7.3) |
−14.6 (−19.5 to −11.2) |
−14.6 (−19.5 to −11.1) |
−14.5 (−18 to −11.3) |
−14.4 (−18.3 to −11) |
| Ineligible | −9.4 (−13.2 to −5.5) |
−10.5 (−15.7 to −7.2) |
−10.6 (−16.9 to −6.6) |
−9.9 (−15.5 to −8.5) |
−11.2 (−18.3 to −6.4) |
−11.1 (−17.9 to −7.7) |
| i-PASI | −2.4 (−5.5 to 0) |
−2.8 (−4.5 to −1) |
−7.4 (−9.8 to −3.6) |
−7.3 (−9.6 to −3.6) |
−7.6 (−9.6 to −2.6) |
−7.8 (−9.8 to −4.6) |
| Unadjusted linear regression of absolute change in PASI; β-coefficient (95% CI)b | ||||||
| Ineligible | 0.7 (−0.8 to 2.2) |
−0.3 (−1.8 to 1.1) |
3.3 (1.0 to 5.5) |
2.7 (0.5 to 4.9) |
2.0 (0.3 to 3.8) |
1.8 (0.1 to 3.5) |
| i-PASI | 8.8 (6.8 to 10.8) |
8.3 (6.4 to 10.2) |
9.5 (8.6 to 10.4) |
9.1 (8.3 to 10.0) |
8.6 (7.0 to 10.2) |
8.5 (6.9 to 10.1) |
| Adjusted linear regression of absolute change in PASI; β-coefficient (95% CI)b | ||||||
| Ineligible | 0.6 (−1.2 to 2.5) |
0.1 (−1.7 to 1.8) |
3.1 (0.8 to 5.5) |
2.9 (0.6 to 5.1) |
0.5 (−1.4 to 2.5) |
0.5 (−1.5 to 2.4) |
| i-PASI | 8.8 (6.8 to 10.9) |
8.3 (6.3 to 10.3) |
9.4 (8.5 to 10.3) |
9.1 (8.3 to 10.0) |
8.7 (7.0 to 10.4) |
8.5 (6.9 to 10.2) |
| Age ≥70 y | −0.1 (−3.7 to 3.5) |
0.5 (−3.0 to 3.9) |
2.3 (−0.5 to 5.2) |
0.5 (−2.3 to 3.3) |
1.0 (−2.5 to 4.5) |
0.9 (−2.5 to 4.4) |
| BMI | 0.1 (0.0 to 0.2) |
0 (−0.1 to 0.1) |
0.1 (0.0 to 0.1) |
0.1 (0.0 to 0.2) |
0.1 (−0.0 to 0.2) |
0.1 (−0.0 to 0.2) |
| ≥3 Comorbidities | 0 (−2.7 to 2.7) |
−0.3 (−2.9 to 2.3) |
0.3 (−1.5 to 2.1) |
0.1 (−1.7 to 1.8) |
3.2 (0.5 to 5.9) |
3.2 (0.6 to 5.9) |
| Prevalent PsA | 0.7 (−0.8 to 2.2) |
0.5 (−1.0 to 1.9) |
−0.5 (−1.5 to 0.5) |
0.1 (−0.9 to 1.0) |
0.6 (−1.1 to 2.3) |
0.4 (−1.3 to 2.1) |
| Male sex | −0.9 (−2.2 to 0.4) |
−1.3 (−2.5 to −0.0) |
−1.0 (−1.9 to −0.2) |
−0.9 (−1.8 to −0.1) |
−0.8 (−2.2 to 0.7) |
0.1 (−1.3 to 1.5) |
| Ever smoked | 0.2 (−1.1 to 1.5) |
−0.3 (−1.5 to 0.9) |
−0.3 (−1.1 to 0.6) |
−0.1 (−0.9 to 0.8) |
0 (−1.4 to 1.5) |
−0.2 (−1.6 to 1.3) |
Abbreviations: i-PASI, insufficient baseline Psoriasis Area and Severity Index; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); PsA, psoriatic arthritis.
Two-sample t test: eligible group vs ineligible or i-PASI group.
Eligible group as reference category.
Serious Adverse Events
The incidence of SAEs in the first 12 months of treatment was highest in the ineligible categories for each biologic (etanercept: IR, 386 per 1000 person-years [95% CI, 279-536]; adalimumab: IR, 514 [95% CI, 367-719]; ustekinumab: IR, 630 [95% CI, 490-809]) (Table 5). Ineligible patients were significantly more likely to have an SAE in the first 12 months compared with the eligible categories for each biologic (etanercept: IRR, 1.91 [95% CI, 1.40-2.60]; adalimumab: IRR, 2.00 [95% CI, 1.55-2.59]; ustekinumab: IRR, 2.81 [95% CI, 2.12-3.72]) (Table 5).
Table 5. Incidence Rates and Incidence Rate Ratios of Serious Adverse Events in the First 12 Months by Registration Therapy and Eligibility Criteria.
| Variable | (95% CI) | |||
|---|---|---|---|---|
| Eligible | Ineligible | i-PASI | m-PASI | |
| Etanercept | ||||
| SAEs, No. | 42 | 36 | 9 | 2 |
| IR per 1000 | 226 (167-305) | 386 (279-536) | 249 (130-479) | 73 (18-293) |
| IRR | 1 [Reference] | 1.9 (1.40-2.60) | 1.0 (0.58-1.64) | 1.3 (0.80-2.03) |
| Adalimumab | ||||
| SAEs, No. | 131 | 34 | 65 | 18 |
| IR per 1000 | 269 (227-319) | 51 (367-719) | 271 (213-346) | 258 (163-410) |
| IRR | 1 [Reference] | 2.0 (1.55-2.59) | 0.9 (0.74-1.11) | 1.0 (0.71-1.30) |
| Ustekinumab | ||||
| SAEs, No. | 40 | 61 | 18 | 11 |
| IR per 1000 | 282 (207-384) | 630 (490-809) | 237 (149-375) | 382 (211-689) |
| IRR | 1 [Reference] | 2.8 (2.12-3.72) | 0.9 (0.65-1.37) | 1.7 (1.04-2.74) |
Abbreviations: IR per 1000, incidence rate per 1000 person-years; IRR, incidence rate ratio; i-PASI, insufficient baseline Psoriasis Area and Severity Index; m-PASI, missing baseline PASI; SAEs, serious adverse events.
Other Analyses
Clinical trials of biologics in psoriasis most commonly report PASI 75 (the proportion of patients achieving a 75% reduction in PASI values from baseline) as the primary outcome (eTable 1 in the Supplement). Differences in PASI 75 values at 6 months, 12 months, and both 6 and 12 months by eligibility category were determined using χ2 test (eTable 2 in the Supplement). Significantly lower proportions of ineligible patients achieved PASI 75 at 6 months, 12 months, or both 6 and 12 months compared with eligible patients. The proportion of eligible patients achieving PASI 75 at 6 months (etanercept, 42%; adalimumab, 51%; ustekinumab, 56%) was lower than the proportion of patients achieving PASI 75 at 24 weeks (etanercept, 59% and 54%, respectively5,6; adalimumab, 64%8) or 28 weeks (ustekinumab, 71% and 79%, respectively10,11) in the licensing trials (eTable 1 in the Supplement).
Missing 6 and/or 12-month PASI values were common in the effectiveness analysis (796 [58%] etanercept; 2230 [61%] adalimumab; 952 [63%] ustekinumab with at least 1 missing value) (eTable 3 in the Supplement). Selection bias in baseline PASI, age, and sex was explored; baseline PASI values for ustekinumab were significantly different between complete and missing PASI values for each time point (eTable 3 in the Supplement). No other significant differences were observed for age, sex, or baseline PASI.
The number of events and incidence rates per MedDRA SOC for each biologic and eligibility status are provided in eTable 4 in the Supplement; the most common events were infections, neoplasms, and surgeries.
Discussion
Patients with psoriasis receiving biologic therapies who were identified as ineligible for clinical trials were twice as likely to experience an SAE and had significantly smaller absolute changes in PASI (adalimumab and ustekinumab) when compared with eligible patients in the first 12 months of treatment. However, eligibility status was not associated with the overall discontinuation rate because there were no differences between eligible and ineligible categories for any of the biologics investigated.
Key Findings
The eligibility criteria applied to clinical trials of biologics in patients with psoriasis creates a restricted sample in which the efficacy and safety of new therapies are investigated prior to licensing. In the present study, 367 (24%) etanercept, 282 (7%) adalimumab, and 394 (24%) ustekinumab patients were identified as ineligible for their respective biologic clinical trials. It is likely that patients categorized as m-PASI or i-PASI would be eligible for clinical trials because the drug discontinuation and SAE rates were comparable to those of patients categorized as eligible.
The cumulative incidence of discontinuing therapy in the first 12 months was similar for each stop reason and eligibility criteria per biologic (Table 3). Patients who discontinued etanercept were more likely to do so for ineffectiveness, with fewer patients discontinuing adalimumab and ustekinumab in the first 12 months. The similarities in discontinuation rates by eligibility category per drug suggest survival analyses may not be good proxy measures for either effectiveness or safety outcomes.15
Eligible patients were significantly more likely to achieve a greater improvement in PASI values at 6 and 12 months compared with i-PASI patients for all 3 biologics and ineligible adalimumab patients (Table 4). However, the proportion of eligible patients achieving PASI 75 at 6 months (etanercept, 119 [42%]; adalimumab, 461 [51%]; ustekinumab, 159 [56%]) was lower than the proportions reported in the licensing trials (etanercept 54% and 59%; adalimumab 64%; ustekinumab 71% and 79%) (eTable 1 in the Supplement). There are a number of potential reasons for the efficacy-effectiveness gap present in this study: the level of clinical care received in a clinical trial setting differs from that of standard clinical care16; adherence to therapies in clinical trials of biologics for psoriasis may exceed adherence in clinical practice, although this is yet to be robustly quantified17; and washout periods are uncommon. The use of concomitant systemic therapies in clinical practice with biologic therapies is more common (280 [19%] etanercept, 645 [16%] adalimumab, 197 [12%] ustekinumab) making it harder to generate a large change on PASI (Table 2). There were also methodological reasons contributing to the efficacy-effectiveness gap in this study: baseline PASI in BADBIR can be reported up to 6 months prior to the biologic start date, therefore the baseline PASI may not capture true disease severity and will also influence the absolute change in PASI; and patients enrolled in the licensing trials extracted all received the highest permitted doses compared with 977 (65%) etanercept and 859 (53%) ustekinumab patients in BADBIR (Table 2). The findings of the present study emphasize the difference between reports of efficacy and effectiveness18; expectations of response levels should therefore be tempered in clinical practice.
Patients categorized as ineligible were significantly more likely than eligible patients to experience an SAE in the first 12 months after initiating biologic therapy. Only 1 previous study has demonstrated the IRR of SAEs to be elevated by 2.6-fold (95% CI, 1.5-4.5) in 29.8% BIOBADADERM patients categorized as ineligible,1 which is comparable to the IRR in the present study (Table 5). However, a lower proportion of patients in BADBIR were categorized as ineligible (376 [24%] etanercept; 282 [7%] adalimumab; 394 [24%] ustekinumab) compared with BIOBADADERM owing to the application of criteria extracted from the phase 3 licensing trials for each biologic1; ineligible patients in the present study had a greater number of comorbidities, similar to the previous study (Table 1).
Limitations
Despite extracting criteria from articles of the licensing trials and additional sources where possible, only 282 (7%) patients were classified as ineligible for adalimumab resulting from fewer comorbidity exclusions extracted compared with etanercept and ustekinumab (Table 1) (eTable 1 in the Supplement). This is likely to have introduced selection bias resulting in lower effectiveness and a higher SAE rate in the eligible adalimumab patients, and therefore an underestimation of the gap between the 2 populations in our results for adalimumab. However, reporting standards of clinical trials having evolved during the past decade with the establishment of the Enhancing the Quality and Transparency of Health Research (EQUATOR) Network guidelines in 2008.19,20 Eligibility criteria were identified as underreported in articles prior to 2009 compared with the criteria extracted from clinical trial protocols of systemic psoriasis therapies obtained from the relevant pharmaceutical companies21; however, the rationale for the additional criteria is unclear. If any of the exclusion criteria are identified as unnecessary, newer biologic therapies could be tested in a more representative sample of patients with psoriasis to bridge the efficacy-effectiveness and safety gaps.
There were limitations in applying the extracted eligibility criteria to patients in BADBIR; comorbidities in BADBIR are reported with the year of onset and it was not possible to identify cases of “uncontrolled” comorbid conditions (eg, diabetes, psychiatric disorders) in the definitions reported. Exclusion of women of child-bearing potential who do not use contraceptives was also not possible to implement because only hormonal contraceptives are captured as a concomitant therapy with low levels of reporting (327 [6%] women). There was also a high proportion of missing data for 6- and/or 12-month PASI values (796 [58%] etanercept, 2230 [61%] adalimumab, 952 [63%] ustekinumab) (eTable 3 in the Supplement). Selection bias was explored between patients with missing and nonmissing PASI values for the effectiveness analysis with no significant differences in age or sex observed, but significant differences in baseline PASI values for ustekinumab only (eTable 3 in the Supplement); therefore, patients included in the effectiveness analysis are considered to be representative of those with missing PASI values at either follow-up.
Conclusions
Patients identified as ineligible in BADBIR did not have the same levels of effectiveness and were at 2 to 3 times increased risk of developing SAEs in the first 12 months following initiation of biologic treatment compared with eligible patients, similar to previous findings.1 Psoriasis patients enrolled in clinical trials of biologics are not representative of real-world patients, particularly those categorized as ineligible in the present study (higher BMI, more patients aged ≥70 years, and a greater number of comorbidities). Clinicians should be mindful of these differences when counselling patients on biologic treatment initiation based on evidence from clinical trials. All trials should publish their full protocol and justify the eligibility criteria selected; this would provide clinicians with a better understanding of the impact the trial results will likely have on their patients.
eFigure. Flowchart of included patients
eTable 1. Data extracted from licensing trial manuscripts
eTable 2. Effectiveness outcome measures by biologic, eligibility and time point
eTable 3. Selection Bias in Effectiveness Outcomes
eTable 4. Incidence rates of serious adverse events by biologic, eligibility and MedDRA System Organ Class
References
- 1.Garcia-Doval I, Carretero G, Vanaclocha F, et al. Risk of serious adverse events associated with biologic and nonbiologic psoriasis systemic therapy: patients ineligible vs eligible for randomized controlled trials. Arch Dermatol. 2012;148(4):463-470. [DOI] [PubMed] [Google Scholar]
- 2.Schmitt J, Zhang Z, Wozel G, Meurer M, Kirch W. Efficacy and tolerability of biologic and nonbiologic systemic treatments for moderate-to-severe psoriasis: meta-analysis of randomized controlled trials. Br J Dermatol. 2008;159(3):513-526. [DOI] [PubMed] [Google Scholar]
- 3.Eichler HG, Abadie E, Breckenridge A, et al. Bridging the efficacy-effectiveness gap: a regulator’s perspective on addressing variability of drug response. Nat Rev Drug Discov. 2011;10(7):495-506. [DOI] [PubMed] [Google Scholar]
- 4.Burden AD, Warren RB, Kleyn CE, et al. ; BADBIR Study Group . The British Association of Dermatologists’ Biologic Interventions Register (BADBIR): design, methodology and objectives. Br J Dermatol. 2012;166(3):545-554. [DOI] [PubMed] [Google Scholar]
- 5.Leonardi CL, Powers JL, Matheson RT, et al. ; Etanercept Psoriasis Study Group . Etanercept as monotherapy in patients with psoriasis. N Engl J Med. 2003;349(21):2014-2022. [DOI] [PubMed] [Google Scholar]
- 6.Papp KA, Tyring S, Lahfa M, et al. ; Etanercept Psoriasis Study Group . A global phase III randomized controlled trial of etanercept in psoriasis: safety, efficacy, and effect of dose reduction. Br J Dermatol. 2005;152(6):1304-1312. [DOI] [PubMed] [Google Scholar]
- 7.Tyring S, Gottlieb A, Papp K, et al. Etanercept and clinical outcomes, fatigue, and depression in psoriasis: double-blind placebo-controlled randomised phase III trial. Lancet. 2006;367(9504):29-35. [DOI] [PubMed] [Google Scholar]
- 8.Gordon KB, Langley RG, Leonardi C, et al. Clinical response to adalimumab treatment in patients with moderate to severe psoriasis: double-blind, randomized controlled trial and open-label extension study. J Am Acad Dermatol. 2006;55(4):598-606. [DOI] [PubMed] [Google Scholar]
- 9.Menter A, Tyring SK, Gordon K, et al. Adalimumab therapy for moderate to severe psoriasis: a randomized, controlled phase III trial. J Am Acad Dermatol. 2008;58(1):106-115. [DOI] [PubMed] [Google Scholar]
- 10.Leonardi CL, Kimball AB, Papp KA, et al. ; PHOENIX 1 study investigators . Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with psoriasis: 76-week results from a randomised, double-blind, placebo-controlled trial (PHOENIX 1). Lancet. 2008;371(9625):1665-1674. [DOI] [PubMed] [Google Scholar]
- 11.Papp KA, Langley RG, Lebwohl M, et al. ; PHOENIX 2 study investigators . Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with psoriasis: 52-week results from a randomised, double-blind, placebo-controlled trial (PHOENIX 2). Lancet. 2008;371(9625):1675-1684. [DOI] [PubMed] [Google Scholar]
- 12.Papp KA, Poulin Y, Bissonnette R, et al. Assessment of the long-term safety and effectiveness of etanercept for the treatment of psoriasis in an adult population. J Am Acad Dermatol. 2012;66(2):e33-e45. [DOI] [PubMed] [Google Scholar]
- 13.Saurat JH, Stingl G, Dubertret L, et al. ; CHAMPION Study Investigators . Efficacy and safety results from the randomized controlled comparative study of adalimumab vs. methotrexate vs. placebo in patients with psoriasis (CHAMPION). Br J Dermatol. 2008;158(3):558-566. [DOI] [PubMed] [Google Scholar]
- 14.Iskandar IY, Ashcroft DM, Warren RB, et al. Demographics and disease characteristics of patients with psoriasis enrolled in the British Association of Dermatologists Biologic Interventions Register. Br J Dermatol. 2015;173(2):510-518. [DOI] [PubMed] [Google Scholar]
- 15.Davila-Seijo P, Garcia-Doval I. Drug survival analysis is not a good method for assessing the safety or effectiveness of systemic therapies in psoriasis. Actas Dermosifiliogr. 2017;108(1):3-5. [DOI] [PubMed] [Google Scholar]
- 16.Kavanaugh A. Ethical and practical issues in conducting clinical trials in psoriasis and psoriatic arthritis. Ann Rheum Dis. 2005;64(suppl 2):ii46-ii48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Thorneloe RJ, Bundy C, Griffiths CE, Ashcroft DM, Cordingley L. Adherence to medication in patients with psoriasis: a systematic literature review. Br J Dermatol. 2013;168(1):20-31. [DOI] [PubMed] [Google Scholar]
- 18.Iskandar IY, Ashcroft DM, Warren RB, et al. Comparative effectiveness of biologic therapies on improvements in quality of life in patients with psoriasis. Br J Dermatol. 2017;177(5):1410-1421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Altman DG, Simera I, Hoey J, Moher D, Schulz K. EQUATOR: reporting guidelines for health research. Lancet. 2008;371(9619):1149-1150. [DOI] [PubMed] [Google Scholar]
- 20.Simera I, Altman DG, Moher D, Schulz KF, Hoey J. Guidelines for reporting health research: the EQUATOR network’s survey of guideline authors. PLoS Med. 2008;5(6):e139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kirsten N, Bulai Livideanu C, Richard MA, et al. ; French Psoriasis Research Group . Inclusion and exclusion criteria in phase III trials with systemic agents in psoriasis: the external validity of drug development. Br J Dermatol. 2016;175(3):636-638. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure. Flowchart of included patients
eTable 1. Data extracted from licensing trial manuscripts
eTable 2. Effectiveness outcome measures by biologic, eligibility and time point
eTable 3. Selection Bias in Effectiveness Outcomes
eTable 4. Incidence rates of serious adverse events by biologic, eligibility and MedDRA System Organ Class