Key Points
Question
Is the risk of serious infections different between biologic or targeted exposures in patients with psoriasis?
Findings
In this cohort study of 44 239 new users of biologics, the risk of serious infections was higher for new users of adalimumab or infliximab vs etanercept, whereas ustekinumab was associated with a lower risk of a serious infection. Risk of serious infections was not increased for new users of secukinumab, ixekizumab, brodalumab, guselkumab, or apremilast vs etanercept, but the risk of serious infections was increased with the use of concomitant nonsteroidal anti-inflammatory drugs or systemic corticosteroids.
Meaning
These results could help physicians choose a biologic for patients with psoriasis who are at risk of serious infections.
Abstract
Importance
Biologics and targeted therapies, such as apremilast, are efficient treatments to manage moderate to severe psoriasis. More information about the risk of serious infection is needed for the newest treatment options in a real-world setting.
Objective
To assess the risk of serious infection among biologics and apremilast used to treat psoriasis, with etanercept as the comparator.
Design, Setting, and Participants
This nationwide cohort study from France involved data from the National Health Data System covering approximately 99% of the French population. All adults with psoriasis, defined as receiving at least 2 prescriptions of a topical vitamin D derivative within a 2-year period, registered in the database between January 1, 2008, and May 31, 2019, were eligible. The study population included those who were new users of biologic agents or apremilast (ie, without any prescriptions of a biologic or apremilast during the previous year). Patients with HIV infection or a history of cancer, transplant, or serious infection were excluded. End of follow-up was January 31, 2020.
Main Outcome Measures
The primary end point was a serious infection in a time-to-event analysis using propensity score–weighted Cox proportional hazards regression models, estimating weighted hazard ratios (wHRs) and 95% CIs.
Results
A total of 44 239 new users of biologic treatment were identified (mean [SD] age, 48.4 [13.8] years; 22 866 [51.7%] men; median follow-up, 12 months [interquartile range, 7-24 months]). A total of 29 618 (66.9%) were prescribed a tumor necrosis factor inhibitor first, 6658 (15.0%) an interleukin (IL) 12/23 inhibitor, 4093 (9.3%) an IL-17 inhibitor, 526 (1.2%) an IL-23 inhibitor, and 3344 (7.6%) apremilast. The total number of serious infections was 1656, and the overall crude incidence rate was 25.0 (95% CI, 23.8-26.2) per 1000 person-years. The most frequent serious infections were gastrointestinal infections (645 patients [38.9%]). After adjusting for time-dependent covariables, risk of serious infections was higher for new users of adalimumab (wHR, 1.22; 95% CI, 1.07-1.38) or infliximab (wHR, 1.79; 95% CI 1.49-2.16) vs etanercept, whereas ustekinumab was associated with a lower risk of having a serious infection (wHR, 0.79; 95% CI, 0.67-0.94). Risk of serious infections was not increased for new users of IL-17 and the IL-23 inhibitor guselkumab or apremilast vs etanercept. Risk of serious infections was increased with concomitant nonsteroidal anti-inflammatory drugs or systemic corticosteroids.
Conclusions and Relevance
In this cohort study of individuals with moderate to severe psoriasis, risk of serious infections was increased in new users of infliximab and adalimumab vs etanercept, whereas ustekinumab users had lower risk of having a serious infection but not new users of IL-17 and IL-23 inhibitors or apremilast. Other observational studies are needed to confirm results for the most recent drugs.
This cohort study uses data from the French National Health Data System to assess the risk of serious infection among biologics and apremilast used to treat psoriasis, with etanercept as the comparator.
Introduction
Systemic nonbiologic therapies, targeted therapies, or biologics are commonly used to manage moderate to severe psoriasis, with higher efficacy for biologic than nonbiologic agents.1,2 With the fast emergence of these new therapeutic agents, evaluating long-term comparative safety in a real-world setting is needed because patients with moderate to severe psoriasis may be at increased risk of serious infection, depending on the biologic agent used.3
Clinical trials are not adequately powered to assess the risk of serious infection, defined as any infection leading to hospitalization in patients with psoriasis receiving biologic therapies,4,5 and have limited external validity because 30% to 80% of patients with psoriasis receiving biologics in national cohorts are not eligible to participate in randomized clinical trials. Moreover, ineligible patients are more likely to have serious adverse events, including serious infections, than are eligible patients.6,7,8 However, observational studies4,9,10 have had conflicting results and are difficult to compare because they use different methods, including using nonbiologic agents as a comparator. For example, Dávila-Seijo et al9 and Yiu et al4 found no increased risk of serious infection with biologics vs nonbiologics, whereas Kalb et al10 found an increased risk with adalimumab (hazard ratio [HR], 2.1; 95% CI, 1.3-3.4) and infliximab (HR, 2.5; 95% CI, 1.5-4.3) but not ustekinumab (HR, 1.0; 95% CI, 0.6-1.7) or etanercept (HR, 1.5; 95% CI, 0.9-2.4) compared with retinoids or phototherapy. More recently, a US cohort study11 of 107 707 patients with psoriasis who were new users of biologics found a lower risk of serious infection with etanercept (HR, 0.75; 95% CI, 0.61-0.93) and ustekinumab (HR, 0.7; 95% CI, 0.5-0.9) but not infliximab or adalimumab vs methotrexate. The main limitation of these observational studies is the failure to take into account the risk of immortal time bias because of prevalent users of nonbiologic agents compared with incident users of biologic agents.12 In addition, only limited data are available on the most recent biologics (interleukin [IL] 17 and IL-23 inhibitors).13 The main objective of this study was to assess and compare the risk of serious infection with several different biologics used to treat psoriasis in a large, exhaustive population of patients in France.
Methods
Data Sources
This nationwide cohort study from France is composed of health administrative data obtained from the French National Health Data System (Système National des Données de Santé [SNDS]) covering approximately 65 million individuals (98.8% of the French population) linked to the national hospital discharge database (Programme de médicalisation des systèmes d'information [PMSI]) by a unique anonymous identifier.14,15,16 The SNDS contains outpatient data (sociodemographic characteristics, including age, sex, and vital status; number of units and date of drug dispensation; date and nature of medical and paramedical interventions; attribution of long-term disease status allowing full coverage of all relevant medical costs for a renewable 5-year period; and eligibility for complementary universal health insurance, which provides free access to health care for low-income people) that can be linked with public and private hospitalization data (admission date; discharge diagnoses according to the International Statistical Classification of Diseases and Related Health Problems, 10th Revision [ICD-10] codes for the main, principal, related, or accompanying diagnosis; and medical procedures) from the PMSI. This large database has been used to conduct several pharmacoepidemiologic studies.17,18,19,20,21 This study was approved by the French data protection agency Commission Nationale de l’Informatique et des Libertés with a waiver of informed consent and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.22
Study Population and Follow-up Procedures
All adults (≥18 years of age) with psoriasis registered in the SNDS between January 1, 2008, and May 31, 2019, were eligible. End of follow-up was January 31, 2020. Adults with psoriasis were identified by having at least 2 prescriptions of topical vitamin D derivatives (Anatomical Therapeutic Chemical code D05AX, the recommended first-line treatment for psoriasis in France) within a 2-year period.19,23 Eligible patients were those with at least 1 fulfillment of a biologic prescription (biologic/biosimilar antipsoriatic agent) or apremilast (targeted synthetic antipsoriatic agent). We excluded patients with HIV infection or a history of cancer, transplant, or serious infection in the 2 years before the index date. New users of biologic/biosimilar or targeted synthetic antipsoriatic agents (defined as no prescription of a biologic or apremilast in the year before the index date) were included. The date of inclusion in the study cohort (index date) was defined as the date of the first fulfillment of biologic/biosimilar antipsoriatic agents or apremilast occurring between January 1, 2009, and July 31, 2019.
Exposure Definition
Treatment exposures were biologic/biosimilar antipsoriatic agents: etanercept, infliximab, adalimumab, certolizumab (anti–tumor necrosis factor [TNF] agents); ustekinumab (anti–IL-12/23); secukizumab, ixekizumab, and brodalumab (anti–IL-17); guselkumab (anti–IL-23); and the targeted synthetic antipsoriatic agent apremilast. Because risankizumab-rzaa (anti–IL-23) did not receive marketing authorization for psoriasis in France until January 2020, it was not assessed. Drugs were identified by Anatomical Therapeutic Chemical codes (eTable 1 in the Supplement).
Exposure to a molecule was defined as time from initiation to discontinuation. Discontinuation was identified as a period of at least 90 days without completion of a prescription for the same biologic/biosimilar or targeted synthetic antipsoriatic agents after the period covered by the previous reimbursement.18,24 The end of the 90-day period was defined as the discontinuation date. The period covered by a prescription was 28 days for all biologic/biosimilar or targeted synthetic antipsoriatic agents except infliximab and guselkumab (56 days) and ustekinumab (84 days).19 Only the first therapeutic sequence was considered in this analysis.
Combination therapy for psoriasis cure was defined as the association of a biologic/bioimilar or targeted synthetic antipsoriatic agent and a nonbiologic agent (methotrexate, cyclosporin, or acitretin). Exposure to combinations of drugs was defined as a period of less than 30 days between the prescription of a biologic/biosimilar or targeted synthetic antipsoriatic agent and a nonbiologic antipsoriatic agent and the filling of another prescription for both drugs in the following 90 days.
Outcome
The primary end point was the occurrence, after the index date, of a serious infection, defined by a hospital discharge diagnosis for an infection (ICD-10 codes are in eTable 2 in the Supplement). The first occurrence of a serious infection was considered. The following serious infection outcomes were assessed: gastrointestinal, cutaneous, eyes, ear-nose-throat, musculoskeletal, pulmonary, nervous system, and other.
Covariates
For each patient, sociodemographic characteristics, including age, sex, complementary universal health insurance, and vital status, were recorded at the index date. Date of exposure to nonbiologic antipsoriatic agents, nonsteroidal anti-inflammatory drugs (NSAIDs), and systemic corticosteroids in the 2 years preceding the index data, at the index date, and during follow-up were collected using the SNDS.
For a 2-year period before the index date, we collected data related to comorbidities (cardiovascular diseases, diabetes, inflammatory rheumatism, chronic inflammatory bowel disease [IBD], hepatic diseases, renal failure, and respiratory disease) and associated inflammatory diseases (psoriatic arthritis and IBD) by using the ICD-10 codes to extract the admission date for each event from the PMSI (eTable 1 in the Supplement).
Statistical Analysis
For descriptions of the study population, quantitative variables are reported as mean (SD) or median (interquartile range [IQR]) and categorical variables as number (percentage). The annual incidence of serious infections was estimated for the entire study population, for each biologic, and for each type of infection. The annual incidences were expressed per 1000 person-years. Patient characteristics at baseline are described in total and for each biologic/biosimilar or targeted synthetic antipsoriatic agent group.
Participants were followed up until the occurrence of the main event (serious infections) or the censorship dates: death from any cause (date of death), switch to another biologic/biosimilar or targeted synthetic antipsoriatic agent, date of the end of exposure, lost to follow-up (defined by the absence of reimbursement for 12 consecutive months), or the study point date (January 31, 2020), whichever came first.
Multivariate Cox proportional hazards regression models were used to identify factors associated with a first serious infection and to estimate the adjusted HR and 95% CI in patients who were new users of biologic/biosimilar or targeted synthetic antipsoriatic agents for psoriasis. The Cox proportional hazards regression assumption was tested formally by using Schoenfeld residuals.
A propensity score (for identifying whether a patient was likely to use a biologic/biosimilar or targeted synthetic antipsoriatic agent in particular) was estimated by using a multinomial logistic regression model that included all patient characteristics at baseline: age, sex, comorbidities (heart, liver, or respiratory kidney disease or diabetes), and a history of nonbiologic antipsoriatic agent therapy in the past 2 years. The inverse probability of treatment weighting (IPTW) method was used to reduce potential bias from treatment allocation. Weights were based on the propensity score. Because the treatment groups were of unequal size, the IPTW was stabilized by multiplying the IPTW by the marginal probability of receiving each treatment to preserve the sample size of the original data and produce an appropriate estimate of the main effect variance.25 The balance in baseline covariates was compared with standardized differences, before and after weighting. The principal analysis was performed with concomitant administration of NSAIDs or systemic corticosteroids or combination therapy with biologic/biosimilar or targeted synthetic antipsoriatic agent as time-varying variables.
To assess the robustness of our results, we performed prespecified subgroup analyses of patients without associated inflammatory disease (ie, excluding patients with psoriatic arthritis or IBD) and patients included after January 1, 2017, because 9 of the 10 studied treatments were commercialized at this time. To assess the sensitivity of the estimated weighted HRs (wHRs) with respect to several possible models, we performed the following sensitivity analyses: (1) Fine-Gray competing risks analysis, computing IPTW subhazard ratios to account for the competing risk between all-cause out-of-hospital death and hospitalization for serious infections; (2) conventional multivariate Cox proportional hazards regression model estimating adjusted HRs for the coreimbursement of NSAIDs, nonbiologic antipsoriatic agents, or prednisone, with a biologic/biosimilar antipsoriatic agents or apremilast considered as time-varying variables; (3) defining treatment discontinuation as longer than 60 days or longer than 120 days without filling a prescription for the same treatment after the period covered by the previous prescription; and (4) using a 1 × 1 propensity score–matched analysis.
All tests were bilateral. All analyses were performed with SAS Enterprise Guide, version 7.1 (SAS Institute Inc). The threshold for statistical significance was set at 2-sided P < .05.
Results
Characteristics of the Study Population
A total of 44 239 of 1 547 040 adults identified as having psoriasis were new users of biologic/biosimilar or targeted synthetic antipsoriatic agents (mean [SD] age, 48.4 [13.8] years; 22 866 [51.7%] men; median follow-up, 12 [IQR, 7-24] months). A total of 29 618 patients (66.9%) were prescribed a TNF inhibitor first: 15 925 (36.0%) received adalimumab, 9661 (21.8%) etanercept, 3002 (6.8%) infliximab, and 1030 (2.3%) certolizumab; 6658 (15.1%) received an IL-12/23 inhibitor (ustekinumab); 4093 (9.2%) received an IL-17 inhibitor: 3145 (7.1%) secukinumab, 768 (1.7%) ixekizumab, and 180 (0.4%) brodalumab; 526 (1.2%) received an IL-23 inhibitor (guselkumab); and 3344 (7.6%) received apremilast (eFigure 1 in the Supplement).
Patient characteristics are summarized in Table 1. The proportion of women was highest in the certolizumab cohort (70.4% [725 of 1030 patients]); inflammatory diseases were frequent in the infliximab cohort. Patients in the apremilast cohort had more comorbidities than those in the biologic/biosimilar antipsoriatic agent cohorts. Methotrexate was the most frequent nonbiologic antipsoriatic agent used as combination therapy with biologic/biosimilar or targeted synthetic antipsoriatic agent. Overall, 5686 patients (12.9%) had combination therapy at baseline and 7155 (16.2%) during follow-up; 2409 (5.4%) received a systemic corticosteroid at baseline and 5749 (13.0%) during follow-up, and 3583 (8.1%) received NSAIDs at baseline and 10 083 (22.8%) during follow-up. Most of the study population did not receive combination therapy, systemic corticosteroids, or NSAIDs at baseline (34 445 [77.9%]) or during follow-up (26 798 [60.6%]).
Table 1. Characteristics of the Biologic/Biosimilar–Naive or Targeted Synthetic Antipsoriatic Agent–Naive Populationa.
| Characteristic | TNF inhibitors | Ustekinumab (IL-12/23 inhibitor) | IL-17 inhibitors | Guselkumab (IL-12 inhibitor) | Apremilast (PDE4 inhibitor) | Total | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Adalimumab | Etanercept | Infliximab | Certolizumab | Secukinumab | Ixekizumab | Brodalumab | |||||
| No. (%) | 15 925 (36) | 9661 (21.8) | 3002 (6.8) | 1030 (2.3) | 6658 (15.1) | 3145 (7.1) | 768 (1.7) | 180 (0.4) | 526 (1.2) | 3344 (7.6) | 44 239 (100) |
| Follow-up, median (IQR), mo | 13 (7-26) | 12 (7-25) | 12 (10-13) | 9 (6-19) | 17 (10-31) | 13 (7-25) | 11 (6-19) | 7 (4-11) | 6 (3-9) | 7 (4-14) | 12 (7-24) |
| Sociodemographic characteristics | |||||||||||
| Age, mean (SD), y | 46.8 (13.5) | 50 (13.6) | 49.1 (14.4) | 47.6 (13.2) | 47.1 (13.3) | 49.4 (13.3) | 48 (13.7) | 48.2 (14.4) | 47.8 (12.7) | 52.7 (14.7) | 48.4 (13.8) |
| Men | 8112 (50.9) | 4711 (48.8) | 1496 (49.8) | 305 (29.6) | 3807 (57.2) | 1682 (53.5) | 453 (59.0) | 110 (61.1) | 325 (61.8) | 1865 (55.8) | 22 866 (51.7) |
| Women | 7813 (49.1) | 4950 (51.2) | 1506 (50.2) | 725 (70.4) | 2851 (42.8) | 1463 (46.5) | 315 (41.0) | 70 (38.9) | 201 (38.2) | 1479 (44.2) | 21 373 (48.3) |
| Complementary universal health coverageb | 1390 (8.7) | 755 (7.8) | 341 (11.3) | 89 (8.6) | 634 (9.5) | 270 (8.6) | 80 (10.4) | 14 (7.8) | 57 (10.8) | 232 (6.9) | 3862 (8.7) |
| Associated inflammatory diseases | |||||||||||
| Psoriatic arthritis | 1736 (10.9) | 1541 (16.0) | 632 (21.1) | 179 (17.4) | 524 (7.9) | 388 (12.3) | 45 (5.9) | 5 (2.8) | 10 (1.9) | 42 (1.3) | 5102 (11.5) |
| Inflammatory bowel disease | 2170 (13.6) | 23 (0.2) | 1122 (37.4) | 38 (3.7) | 319 (4.8) | 8 (0.3) | 1 (0.1) | 0 | 3 (0.6) | 5 (0.2) | 3689 (8.3) |
| Comorbidities | |||||||||||
| Cardiovascular events | 481 (3.0) | 437 (4.5) | 133 (4.4) | 25 (2.4) | 224 (3.4) | 111 (3.5) | 35 (4.6) | 6 (3.3) | 27 (5.1) | 173 (5.2) | 1652 (3.7) |
| Chronic obstructive pulmonary disease | 62 (0.4) | 56 (0.6) | 22 (0.7) | 6 (0.6) | 27 (0.4) | 20 (0.6) | 11 (1.4) | 2 (1.1) | 4 (0.8) | 20 (0.6) | 230 (0.5) |
| Kidney disease | 895 (5.6) | 362 (3.8) | 411 (13.7) | 67 (6.5) | 355 (5.3) | 156 (5.0) | 51 (6.6) | 13 (7.2) | 29 (5.5) | 96 (2.9) | 2435 (5.5) |
| Hepatic disease | 192 (1.2) | 212 (2.2) | 55 (1.8) | 10 (1.0) | 138 (2.1) | 40 (1.3) | 9 (1.2) | 5 (2.8) | 12 (2.3) | 55 (1.6) | 728 (1.7) |
| Diabetes | 1063 (6.7) | 824 (8.5) | 244 (8.1) | 59 (5.7) | 502 (7.5) | 278 (8.8) | 64 (8.3) | 17 (9.4) | 48 (9.1) | 339 (10.1) | 3438 (7.8) |
| Respiratory disease | 198 (1.2) | 168 (1.7) | 53 (1.8) | 21 (2.0) | 86 (1.3) | 57 (1.8) | 14 (1.8) | 4 (2.2) | 10 (1.9) | 44 (1.3) | 655 (1.5) |
| Lupus | 29 (0.2) | 29 (0.3) | 11 (0.4) | 8 (0.8) | 26 (0.4) | 6 (0.2) | 1 (0.1) | 0 | 2 (0.4) | 2 (0.1) | 114 (0.3) |
| Vasculitis | 12 (0.1) | 9 (0.1) | 3 (0.1) | 1 (0.1) | 4 (0.1) | 2 (0.1) | 0 | 0 | 0 | 0 | 31 (0.1) |
| Dermatomyositis | 4 (0.03) | 1 (0.01) | 0 | 1 (0.1) | 2 (0.03) | 0 | 0 | 0 | 0 | 0 | 8 (0) |
| Baseline concomitant treatmentsc | |||||||||||
| Conventional | 2049 (12.9) | 1790 (18.5) | 502 (16.7) | 277 (26.9) | 499 (7.5) | 317 (10.1) | 50 (6.5) | 7 (3.9) | 8 (1.5) | 187 (5.6) | 5686 (12.9) |
| Methotrexate | 1957 (12.3) | 1725 (17.9) | 488 (16.3) | 272 (26.4) | 434 (6.5) | 302 (9.6) | 37 (4.8) | 5 (2.8) | 6 (1.1) | 141 (4.2) | 5367 (12.1) |
| Cyclosporin | 35 (0.2) | 14 (0.1) | 2 (0.1) | 0 | 27 (0.4) | 4 (0.1) | 5 (0.7) | 2 (1.1) | 1 (0.2) | 0 | 90 (0.2) |
| Acitretin | 49 (0.3) | 42 (0.4) | 11 (0.4) | 4 (0.4) | 37 (0.6) | 12 (0.4) | 8 (1.0) | 0 | 1 (0.2) | 31 (0.9) | 195 (0.4) |
| Phototherapy | 9 (0.1) | 11 (0.1) | 2 (0.1) | 1 (0.1) | 5 (0.1) | 1 (0.03) | 0 | 0 | 0 | 15 (0.5) | 44 (0.1) |
| Prednisone | 1058 (6.6) | 763 (7.9) | 212 (7.1) | 115 (11.2) | 112 (1.7) | 87 (2.8) | 12 (1.6) | 2 (1.1) | 1 (0.2) | 47 (1.4) | 2409 (5.4) |
| NSAIDs | 1444 (9.1) | 1219 (12.6) | 121 (4.0) | 180 (17.5) | 264 (4.0) | 214 (6.8) | 27 (3.5) | 4 (2.2) | 8 (1.5) | 102 (3.1) | 3583 (8.1) |
| None of the above | 12 084 (75.9) | 6617 (68.5) | 2274 (75.7) | 569 (55.2) | 5873 (88.2) | 2606 (82.9) | 687 (89.5) | 167 (92.8) | 509 (96.8) | 3059 (91.5) | 34 445 (77.9) |
| Concomitant treatments during follow-upc | |||||||||||
| Conventional | 2663 (16.7) | 2180 (22.6) | 662 (22.1) | 315 (30.6) | 648 (9.7) | 386 (12.3) | 58 (7.6) | 7 (3.9) | 11 (2.1) | 225 (6.7) | 7155 (16.2) |
| Methotrexate | 2524 (15.8) | 2075 (21.5) | 645 (21.5) | 310 (30.1) | 553 (8.3) | 366 (11.6) | 44 (5.7) | 5 (2.8) | 7 (1.3) | 167 (5) | 6696 (15.1) |
| Cyclosporin | 40 (0.3) | 15 (0.2) | 2 (0.1) | 1 (0.1) | 34 (0.5) | 4 (0.1) | 5 (0.7) | 2 (1.1) | 1 (0.2) | 0 | 104 (0.2) |
| Acitretin | 96 (0.6) | 85 (0.9) | 16 (0.5) | 5 (0.5) | 64 (1.0 | 17 (0.5) | 9 (1.2) | 0 | 3 (0.6) | 39 (1.2) | 334 (0.8) |
| Phototherapy | 17 (0.1) | 15 (0.2) | 6 (0.2) | 1 (0.1) | 6 (0.1) | 1 (0.03) | 0 | 0 | 0 | 22 (0.7) | 68 (0.2) |
| Prednisone | 2481 (15.6) | 1611 (16.7) | 438 (14.6) | 239 (23.2) | 506 (7.6) | 288 (9.2) | 38 (4.9) | 3 (1.7) | 6 (1.1) | 139 (4.2) | 5749 (13) |
| NSAIDs | 3864 (24.3) | 2994 (31) | 440 (14.7) | 401 (38.9) | 1217 (18.3) | 658 (20.9) | 89 (11.6) | 17 (9.4) | 25 (4.8) | 378 (11.3) | 10 083 (22.8) |
| None of the above | 9017 (56.6) | 4805 (49.7) | 1803 (60.0) | 366 (35.5) | 4760 (71.5) | 2081 (66.2) | 607 (79.0) | 153 (85.0) | 486 (92.4) | 2720 (81.3) | 26 798 (60.6) |
| Prior treatments in the previous 2 y | |||||||||||
| Conventionald | 8423 (52.9) | 5734 (59.4) | 962 (32.0) | 542 (52.6) | 4623 (69.4) | 1765 (56.1) | 492 (64.1) | 116 (64.4) | 331 (62.9) | 1467 (43.9) | 24 455 (55.3) |
| Methotrexate | 7395 (46.4) | 5077 (52.6) | 877 (29.2) | 528 (51.3) | 3758 (56.4) | 1452 (46.2) | 396 (51.6) | 85 (47.2) | 269 (51.1) | 796 (23.8) | 20 633 (46.6) |
| Cyclosporin | 752 (4.7) | 421 (4.4) | 49 (1.6) | 16 (1.6) | 740 (11.1) | 174 (5.5) | 65 (8.5) | 19 (10.6) | 42 (8.0) | 51 (1.5) | 2329 (5.3) |
| Acitretin | 1488 (9.3) | 982 (10.2) | 103 (3.4) | 13 (1.3) | 1052 (15.8) | 350 (11.1) | 113 (14.7) | 32 (17.8) | 74 (14.1) | 780 (23.3) | 4981 (11.3) |
| Prednisonee | 4058 (25.5) | 2587 (26.8) | 740 (24.7) | 457 (44.4) | 887 (13.3) | 544 (17.3) | 103 (13.4) | 22 (12.2) | 60 (11.4) | 355 (10.6) | 9813 (22.2) |
| NSAIDs | 7232 (45.4) | 5448 (56.4) | 863 (28.7) | 720 (69.9) | 1944 (29.2) | 1297 (41.2) | 237 (30.9) | 38 (21.1) | 128 (24.3) | 947 (28.3) | 18 854 (42.6) |
| None of the above | 3024 (19.0) | 1371 (14.2) | 1211 (40.3) | 85 (8.3) | 1238 (18.6) | 708 (22.5) | 159 (20.7) | 44 (24.4) | 131 (24.9) | 1256 (37.6) | 9227 (20.9) |
Abbreviations: IL, interleukin; IQR, interquartile range; NSAIDs, nonsteroidal anti-inflammatory drugs; PDE4, phosphodiesterase 4; TNF, tumor necrosis factor.
Data are presented as number (percentage) unless indicated.
Data are missing for 74 patients.
Concomitant is defined as a period of less than 30 days between the prescription of 2 biologic/biosimilar or targeted synthetic antipsoriatic agent and that of a different conventional drug or NSAIDs or prednisone and the filling of another prescription for both drugs in the next 90 days.
At least 1 prescription.
At least 3 prescriptions.
Description of the Serious Infections
During follow-up, we identified 1656 serious infections in the study population. The overall crude incidence rate was 25.0 (95% CI, 23.8-26.2) per 1000 person-years. The most frequent serious infections were gastrointestinal infections (645 [38.9%]), then skin and pulmonary infections (324 [19.6%] and 245 [14.8%], respectively). The median time to event was 9 months [IQR, 4-20 months]. Details per therapeutic drug are available in Table 2, and serious infection details are summarized in eTable 3 in the Supplement.
Table 2. Serious Infections and Death Events According to the Biologic/Biosimilar or Targeted Synthetic Antipsoriatic Agent Used.
| Event | No. (%) of events | Incidence rate, per 1000 person-years (95% CI) | Time before event, median (IQR), mo |
|---|---|---|---|
| Total (N = 44 239) | |||
| Serious infections | 1656 (3.7) | 25 (23.8-26.2) | 9 (4-20) |
| Gastrointestinal | 645 (1.5) | 9.8 (9-10.5) | NA |
| Skin and subcutaneous tissues | 324 (0.7) | 4.9 (4.4-5.4) | NA |
| Pulmonary | 245 (0.6) | 3.7 (3.2-4.2) | NA |
| Urinary | 157 (0.4) | 2.4 (2-2.7) | NA |
| Death | 179 (0.4) | 2.7 (2.3-3.1) | 12 (6-24) |
| Adalimumab (n = 15 925) | |||
| Serious infections | 697 (4.4) | 27.7 (25.7-29.8)) | 10 (4-23) |
| Death | 41 (0.3) | 1.6 (1.1-2.1) | 12 (8-29) |
| Etanercept (n = 9661) | |||
| Serious infections | 367 (3.8) | 24.8 (22.3-27.4) | 9 (4-24) |
| Death | 52 (0.5) | 3.5 (2.6-4.5) | 13 (7-25) |
| Infliximab (n = 3002) | |||
| Serious infections | 171 (5.7) | 43 (36.5-49.4) | 7 (4-19) |
| Death | 21 (0.7) | 5.3 (3-7.5) | 8 (3-13) |
| Certolizumab (n = 1030) | |||
| Serious infections | 35 (3.4) | 26.7 (17.9-35.6) | 9 (4-17) |
| Death | 5 (0.5) | 3.8 (0.5-7.2) | 16 (2-16) |
| Ustekinumab (n = 6658) | |||
| Serious infections | 202 (3.0) | 16.3 (14-18.5) | 11 (5-22) |
| Death | 36 (0.5) | 2.9 (2-3.8) | 16 (5-29) |
| Secukinumab (n = 3145) | |||
| Serious infections | 95 (3.0) | 21.9 (17.5-26.3) | 9 (4-17) |
| Death | 10 (0.3) | 2.3 (0.9-3.7) | 11 (8-16) |
| Ixekizumab (n = 768) | |||
| Serious infections | 15 (2.0) | 17.7 (8.8-26.7) | 7 (6-12) |
| Death | 2 (0.3) | 2.4 (<0-5.6) | 13 (13-13) |
| Brodalumab (n = 180) | |||
| Serious infections | 2 (1.1) | 17.8 (<0-42.4) | 4 (2-5) |
| Death | 0 | NA | NA |
| Guselkumab (n = 526) | |||
| Serious infections | 8 (1.5) | 29.9 (9.2-50.7) | 2 (0-3) |
| Death | 2 (0.4) | 7.4 (<0-17.9) | 6 (5-8) |
| Apremilast (n = 3344) | |||
| Serious infections | 64 (1.9) | 21.7 (16.4-27) | 8 (3-14) |
| Death | 10 (0.3) | 3.4 (1.3-5.5) | 9 (5-13) |
Abbreviations: IQR, interquartile range; NA, not applicable.
Risk of Serious Infection Among New Users of Biologics/BioSimilars or Targeted Synthetics
A pseudocohort was obtained with use of the stabilized propensity score; distributions of patient characteristics were balanced, with standardized differences less than |0.1| between the treatment classes (eFigure 2 in the Supplement). The results of the crude and main analysis are presented in Table 3. After adjustment for time-dependent covariables, risk of a serious infection was higher for new users of biologic/biosimilar or targeted synthetic antipsoriatic agents using adalimumab (wHR, 1.22; 95% CI, 1.07-1.38) or infliximab (wHR, 1.79; 95% CI, 1.49-2.16) than etanercept. In contrast, the risk was reduced with ustekinumab (wHR, 0.79; 95% CI, 0.67-0.94). The risk was not increased with use of secukinumab, ixekizumab, brodalumab, guselkumab, and apremilast vs etanercept. Risk of serious infection was increased with concomitant use of NSAIDs (wHR, 1.47; 95% CI, 1.25-1.73) or systemic corticosteroids (wHR, 2.32; 95% CI, 1.95-2.76).
Table 3. Risk of Serious Infection Among 44 239 Patients by the Biologic/Biosimilar or Targeted Synthetic Antipsoriatic Agent Used in the Unadjusted Cox Proportional Hazards Regression and IPTW Cox Models.
| Variable | Unadjusted Cox model | IPTW Cox model | ||
|---|---|---|---|---|
| cHR (95% CI) | P value | wHR (95% CI) | P value | |
| Treatments | ||||
| Etanercept | 1 [Reference] | 1 [Reference] | ||
| Adalimumab | 1.12 (0.99-1.27) | .08 | 1.22 (1.07-1.38) | <.001 |
| Infliximab | 1.70 (1.42-2.05) | <10−4 | 1.79 (1.49-2.16) | <.001 |
| Certolizumab | 1.06 (0.75-1.50) | .75 | 1.15 (0.83-1.59) | .87 |
| Ustekinumab | 0.67 (0.56-0.79) | <10−4 | 0.79 (0.67-0.94) | .01 |
| Secukinumab | 0.87 (0.69-1.09) | .23 | 0.94 (0.75-1.18) | .61 |
| Ixekizumab | 0.69 (0.41-1.16) | .16 | 0.82 (0.50-1.35) | .44 |
| Brodalumab | 0.65 (0.16-2.60) | .54 | 0.79 (0.21-2.95) | .73 |
| Guselkumab | 1.07 (0.53-2.16) | .86 | 1.37 (0.70-2.67) | .36 |
| Apremilast | 0.83 (0.63-1.08 | .16 | 0.83 (0.63-1.10) | .19 |
| Time-dependent covariables | ||||
| NSAIDs | NA | NA | 1.47 (1.25-1.73 | <.001 |
| Concomitant conventional treatments | NA | NA | 0.99 (0.87-1.13 | .90 |
| Prednisone | NA | NA | 2.32 (1.95-2.76 | <.001 |
Abbreviations: cHR, crude hazard ratio; IPTW, inverse probability of treatment weighting; NA, not applicable; NSAIDs, nonsteroidal anti-inflammatory drugs; wHR, weighted hazard ratio.
Subgroup Analyses
After the exclusion of patients with associated inflammatory disorders, either psoriatic arthritis or IBD, the results did not differ from the main analysis except for in the adalimumab cohort, which had no difference compared with etanercept in the risk of serious infection (wHR, 1.14; 95% CI, 0.98-1.31) (Table 4). The results did not differ in the subgroup starting use of a biologic/biosimilar or targeted synthetic antipsoriatic agent after January 2017, except for the adalimumab cohort, which had no difference compared with etanercept (wHR, 1.09; 95% CI, 0.76-1.56) (Table 4).
Table 4. Risk of Serious Infection by the Biologic/Biosimilar or Targeted Synthetic Antipsoriatic Agent Used in the IPTW Cox Model in Subgroup Analyses.
| Variable | Population without IBD or psoriatic arthritis (n = 35 494 patients and 1205 events) | Population starting agent from January 2017 (n = 19 215 patients and 475 events) | ||
|---|---|---|---|---|
| wHR (95% CI) | P value | wHR (95% CI) | P value | |
| Treatments | ||||
| Etanercept | 1 [Reference] | 1 [Reference] | ||
| Adalimumab | 1.14 0.98-1.31) | .09 | 1.09 0.76-1.56) | .65 |
| Infliximab | 2.00 1.56-2.55) | <10−4 | 1.66 1.12-2.44) | .01 |
| Certolizumab | 1.11 0.77-1.62) | .58 | 1.22 0.67-2.21) | .52 |
| Ustekinumab | 0.73 0.60-0.88) | 10−3 | 0.81 0.54-1.19) | .28 |
| Secukinumab | 0.84 0.65-1.09) | .19 | 0.92 0.62-1.37) | .69 |
| Ixekizumab | 0.76 0.45-1.31) | .33 | 0.88 0.50-1.58) | .67 |
| Brodalumab | 0.84 0.23-3.14) | .80 | 0.81 0.21-3.15) | .76 |
| Guselkumab | 1.25 0.61-2.57) | .55 | 1.56 0.77-3.16) | .22 |
| Apremilast | 0.85 0.64-1.13) | .25 | 0.82 0.54-1.26) | .37 |
| Time-dependent covariables | ||||
| NSAIDs | 1.57 1.30-1.89) | <10−4 | 1.62 1.18-2.22) | <.001 |
| Concomitant conventional treatments | 1.00 0.86-1.17) | >.99 | 1.05 0.83-1.33) | .68 |
| Prednisone | 2.51 (2.05-3.08) | <10−4 | 2.06 1.43-2.96) | <.001 |
Abbreviations: IBD, inflammatory bowel disease; IPTW, inverse probability of treatment weighting; NSAIDs, nonsteroidal anti-inflammatory drugs; wHR, weighted hazard ratio.
Sensitivity Analyses
The results of the Fine-Gray competing risks analysis are available in eTable 5 in the Supplement. Pairwise results are given in eTable 4 in the Supplement. These results were consistent with the main analysis and with a classic Cox proportional hazards regression model adjusted by age, sex, comorbidities (heart, liver, kidney, or respiratory disease or diabetes), history of receiving nonbiologic antipsoriatic agents in the past 2 years, and time-dependent covariables (eTable 6 in the Supplement), as well as after modifying the gap to 60 and 120 days after the period covered (eTable 5 in the Supplement) and using the propensity score–matched analysis (eTable 7 in the Supplement).
Discussion
This large, population-based, nationwide cohort study of 44 239 new users of biologic/biosimilar or targeted synthetic antipsoriatic treatments with no history of serious infection, cancer, HIV infection, or transplant suggests a greater risk of a serious infection for new users of infliximab and adalimumab vs new users of etanercept, after controlling for confounding factors. Risk of serious infection was lower with ustekinumab. For the other biologic/biosimilar or targeted synthetic antipsoriatic agents, the IL-17 inhibitors (secukinumab, ixekizumab, and brodalumab), the IL-23 inhibitor (guselkumab), and apremilast were not associated with increased risk of serious infections compared with etanercept, a finding that persisted in subgroup and sensitivity analyses. The most frequent serious infections were gastrointestinal, skin, and subcutaneous tissue and pulmonary infections. This distribution of the serious infections was expected,10,26 but with an unexpectedly high proportion of gastrointestinal serious infections in our study (especially cholangitis and diverticulitis). These variations could be attributable to the differences between national health care systems, leading to differences in hospitalization rates for serious infection.
The findings of this study are important because recent biologic agents, including secukizumab, ixekizumab, brodalumab, and guselkumab, were compared with one another. Prior observational studies10,12,13,27 have provided conflicting results on the risk of infection related to use of biologics and mostly used conventional systemic immunosuppressive treatment as a comparator. The differences in the results could be explained by the following. First, immortal time bias cannot be avoided by including both prevalence and incident disease10,13,27 or when prevalent disease is excluded. Immortal time bias is defined as the period during which the outcome of interest cannot occur because of the study design.12 For example, immortal time bias is introduced when there is a typical sequence of treatments (eg, starting use of a biologic/biosimilar only after nonbiologic antipsoriatic agents). For example, in a cohort study that compares the risk of serious infections among new users of biologic/biosimilar antipsoriatic agents with nonusers, there is a wait time (ie, immortal time) for the biologic/biosimilar antipsoriatic agents group because patients need to survive alive or event free to initiate use of a biologic/biosimilar antipsoriatic agent but nonusers do not. If patients have an event before starting use of a biologic/biosimilar antipsoriatic agent, their person-time and the event are attributed only to the nonuser group. Regarding the previous studies,4,11,28 the authors did not take into account previous exposure, whereas they used a nonbiologic antipsoriatic agent as a comparator, which may lead to overestimating the risk of this comparator (usually methotrexate). Participants initiating use of a biologic/biosimilar antipsoriatic agent mainly received a nonbiologic antipsoriatic agent as a previous exposure. However, this previous exposure, considered low risk in terms of serious adverse events (otherwise, the patients would not have received biologics), is not taken into account in their analyses. Second, confounding by indication could also explain these discordances because the treatment choice depends on patient characteristics, comorbidities, or medical preferences.29,30,31 Third, differences in study population may explain differences in results: most previous studies4,9,10,13,27,28 included patients with psoriatic arthritis, whereas 1 study11 did not, and exclusion criteria could be too restrictive,11,13 with combination therapies not always taken into account.10,11,13,27,28
This study found that IL-17 inhibitors (secukizumab, ixekizumab, and brodalumab) and an IL-23 inhibitor (guselkumab) were not associated with increased risk of serious infections compared with etanercept, which remained in subgroup and sensitivity analyses. Taking into account the low number of participants initiating use of guselkumab and the low number of events, this result needs to be confirmed. Compared with etanercept, which is also a TNF inhibitor, infliximab and adalimumab were associated with higher risk of serious infection. This result (for infliximab) is consistent with those published for the British Association of Dermatologists Biologic and Immunomodulators Register (BADBIR) cohort.26 In the subgroup that excluded patients with psoriatic arthritis or IBD, adalimumab was no longer associated with increased risk of serious infection. Thus, a differential risk of serious infections for adalimumab depending on the underlying diseases (psoriasis or psoriasis associated with other inflammatory disorders) cannot be excluded. Compared with etanercept, ustekinumab was associated with a reduced risk of serious infection as has been already reported10,11,13 and is supported by differences in the mechanism of action compared with anti-TNF agents.32 Using ustekinumab as a comparator, a recent study33 of 123 383 patients with psoriasis or psoriatic arthritis found that anti-TNF agents and the anti–IL-17 agents secukinumab and ixekizumab were associated with a 1.5 to 3 times higher risk of serious infections. The authors did not provide results on brodalumab (another anti–IL-17) or guselkumab (anti–IL-23). The current study found that ixekizumab was not associated with increased risk of serious infections compared with ustekinumab. Thus, a differential risk of serious infections could not be excluded within the anti–IL-17 class, and other observational studies are needed.
This study also found that concomitant use of systemic corticosteroids or NSAIDs, but not nonbiologic systemic treatment for psoriasis (methotrexate, acitretin, or cyclosporin), was associated with increased risk of serious infection in new users of biologic/biosimilar or targeted synthetic antipsoriatic agents. Previous studies10,11,13,27,28 did not consider these combination medications like the current study did (time-dependent variables).
Limitations
This study has limitations. First, the definition of psoriasis was based on the prescription of at least 2 topical vitamin D derivatives within a 2-year period. The latter drugs constitute the first-line treatment for psoriasis,34 and the definition used in this study has been validated in a study of a Danish health insurance database by Egeberg et al.35 The sensitivity of the definition was estimated at 85%19 and probably underestimated the number of cases of psoriasis. Second, drug exposure was defined based on health care fulfillment of prescription data, which is not necessarily equivalent to days of use. However, in psoriasis, adherence rates for biologic/biosimilar antipsoriatic agent are generally higher than for other treatment categories.36 Identification of serious infections was based on discharge diagnosis only, but the validity of the outcomes was assessed, with positive predictive values of recorded cases and type of infections of 97%.37
This large, population-based study in a real-world setting assessed the risk of serious infection with each drug separately in new users of biologic/biosimilar or targeted synthetic antipsoriatic agents using etanercept rather than methotrexate as a comparator (to avoid the risk of immortal time bias). This study involved a large number of patients from a national database, with annual quality control of coding, which captures information during routine medical care. This framework minimizes selection bias. Results were adjusted on concomitant treatments as time-dependent variables, and a stabilized IPTW mode was used based on a propensity score created from patients’ baseline characteristics to reduce confounding bias and immortal time bias. However, such bias may not be eliminated. In addition, several sensitivity analyses performed support the integrity of these results.
Conclusions
This study found an increased risk of serious infections in new users of infliximab and adalimumab vs etanercept with moderate to severe psoriasis. Risk of serious infection was reduced with the IL-12/23 inhibitor (ustekinumab) and was not increased with the newest drugs, including apremilast, IL-17 inhibitors (secukinumab, ixekizumab, and brodalumab), and an IL-23 inhibitor (guselkumab). Risk of serious infection was increased with concomitant use of NSAIDs or systemic corticosteroids during the biologic/biosimilar treatment. Other observational studies are needed to confirm results for the most recent drugs.
eTable 1. Definitions of Exposures, Associated Inflammatory Diseases, and Comorbidities
eTable 2. All Diagnoses of Infections With Related ICD-10 Codes Included as Any Serious Infection With Subdivision by Site-Specific Groups
eTable 3. Characterization of Serious Infection Cases, n = 1,656
eTable 4. Risk of Serious Infection With One b/bs/ts Antipsoriatic Agent Relative to Another Displayed by IPTW Fine-Gray Model
eTable 5. Risk of Serious Infection According to the b/bs/ts Antipsoriatic Agent Used in the IPTW Cox and Fine-Gray Models, in Sensitivity Analyses (N = 44,239)
eTable 6. Risk of Serious Infection by the b/bs/ts Antipsoriatic Agent Used in the Classical Cox Adjusted Model for the Entire Study Population (N = 44,239)
eTable 7. Risk of Serious Infection Events According to the b/bs/ts Antipsoriatic Agent Used in the Propensity Score-Matched Cox Model, for the Entire Study Population (N = 44,239)
eFigure 1. Flowchart of the Study
eFigure 2. Standardized Differences Before and After Inverse Probability of Treatment Weighting (IPTW) Weighting (A to I)
References
- 1.Nast A, Gisondi P, Ormerod AD, et al. European S3-Guidelines on the systemic treatment of psoriasis vulgaris: update 2015: short version-EDF in cooperation with EADV and IPC. J Eur Acad Dermatol Venereol. 2015;29(12):2277-2294. doi: 10.1111/jdv.13354 [DOI] [PubMed] [Google Scholar]
- 2.Sbidian E, Chaimani A, Afach S, et al. Systemic pharmacological treatments for chronic plaque psoriasis: a network meta-analysis. Cochrane Database Syst Rev. 2020;1:CD011535. doi: 10.1002/14651858.CD011535.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Singh JA, Wells GA, Christensen R, et al. Adverse effects of biologics: a network meta-analysis and Cochrane overview. Cochrane Database Syst Rev. 2011;CD008794(2):CD008794. doi: 10.1002/14651858.CD008794.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yiu ZZN, Exton LS, Jabbar-Lopez Z, et al. Risk of serious infections in patients with psoriasis on biologic therapies: a systematic review and meta-analysis. J Invest Dermatol. 2016;136(8):1584-1591. doi: 10.1016/j.jid.2016.03.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Afach S, Chaimani A, Evrenoglou T, et al. Meta-analysis results do not reflect the real safety of biologics in psoriasis. Br J Dermatol. 2021;184(3):415-424. doi: 10.1111/bjd.19244 [DOI] [PubMed] [Google Scholar]
- 6.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: 10.1001/archdermatol.2011.2768 [DOI] [PubMed] [Google Scholar]
- 7.Mason KJ, Barker JNWN, Smith CH, et al. ; BADBIR Study Group . Comparison of drug discontinuation, effectiveness, and safety between clinical trial eligible and ineligible patients in BADBIR. JAMA Dermatol. 2018;154(5):581-588. doi: 10.1001/jamadermatol.2018.0183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Masson Regnault M, Castañeda-Sanabria J, Diep Tran MHT, et al. Users of biologics in clinical practice: would they be eligible for phase III clinical studies? cohort study in the French Psoriasis Registry PSOBIOTEQ. J Eur Acad Dermatol Venereol. 2020;34(2):293-300. doi: 10.1111/jdv.15878 [DOI] [PubMed] [Google Scholar]
- 9.Dávila-Seijo P, Dauden E, Descalzo MA, et al. ; BIOBADADERM Study Group . Infections in moderate to severe psoriasis patients treated with biological drugs compared to classic systemic drugs: findings from the BIOBADADERM Registry. J Invest Dermatol. 2017;137(2):313-321. doi: 10.1016/j.jid.2016.08.034 [DOI] [PubMed] [Google Scholar]
- 10.Kalb RE, Fiorentino DF, Lebwohl MG, et al. Risk of serious infection with biologic and systemic treatment of psoriasis: results from the Psoriasis Longitudinal Assessment and Registry (PSOLAR). JAMA Dermatol. 2015;151(9):961-969. doi: 10.1001/jamadermatol.2015.0718 [DOI] [PubMed] [Google Scholar]
- 11.Dommasch ED, Kim SC, Lee MP, Gagne JJ. Risk of serious infection in patients receiving systemic medications for the treatment of psoriasis. JAMA Dermatol. 2019;155(10):1142-1152. doi: 10.1001/jamadermatol.2019.1121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Suissa S. Immortal time bias in observational studies of drug effects. Pharmacoepidemiol Drug Saf. 2007;16(3):241-249. doi: 10.1002/pds.1357 [DOI] [PubMed] [Google Scholar]
- 13.Li X, Andersen KM, Chang H-Y, Curtis JR, Alexander GC. Comparative risk of serious infections among real-world users of biologics for psoriasis or psoriatic arthritis. Ann Rheum Dis. 2020;79(2):285-291. doi: 10.1136/annrheumdis-2019-216102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Weill A, Dalichampt M, Raguideau F, et al. Low dose oestrogen combined oral contraception and risk of pulmonary embolism, stroke, and myocardial infarction in five million French women: cohort study. BMJ. 2016;353:i2002. doi: 10.1136/bmj.i2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Tuppin P, Drouin J, Mazza M, Weill A, Ricordeau P, Allemand H. Hospitalization admission rates for low-income subjects with full health insurance coverage in France. Eur J Public Health. 2011;21(5):560-566. doi: 10.1093/eurpub/ckq108 [DOI] [PubMed] [Google Scholar]
- 16.Tuppin P, Rudant J, Constantinou P, et al. Value of a national administrative database to guide public decisions: from the système national d’information interrégimes de l’Assurance Maladie (SNIIRAM) to the système national des données de santé (SNDS) in France. Rev Epidemiol Sante Publique. 2017;65(suppl 4):S149-S167. doi: 10.1016/j.respe.2017.05.004 [DOI] [PubMed] [Google Scholar]
- 17.Lemaitre M, Kirchgesner J, Rudnichi A, et al. Association between use of thiopurines or tumor necrosis factor antagonists alone or in combination and risk of lymphoma in patients with inflammatory bowel disease. JAMA. 2017;318(17):1679-1686. doi: 10.1001/jama.2017.16071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kirchgesner J, Lemaitre M, Carrat F, Zureik M, Carbonnel F, Dray-Spira R. Risk of serious and opportunistic infections associated with treatment of inflammatory bowel diseases. Gastroenterology. 2018;155(2):337-346.e10. doi: 10.1053/j.gastro.2018.04.012 [DOI] [PubMed] [Google Scholar]
- 19.Sbidian E, Mezzarobba M, Weill A, Coste J, Rudant J. Persistence of treatment with biologics for patients with psoriasis: a real-world analysis of 16 545 biologic-naïve patients from the French National Health Insurance database (SNIIRAM). Br J Dermatol. 2019;180(1):86-93. doi: 10.1111/bjd.16809 [DOI] [PubMed] [Google Scholar]
- 20.Meyer A, Rudant J, Drouin J, Weill A, Carbonnel F, Coste J. Effectiveness and safety of reference infliximab and biosimilar in crohn disease: a French equivalence study. Ann Intern Med. 2019;170(2):99-107. doi: 10.7326/M18-1512 [DOI] [PubMed] [Google Scholar]
- 21.Sbidian E, Billionnet C, Weill A, Maura G, Mezzarobba M. Persistence of apremilast in moderate-to-severe psoriasis: a real-world analysis of 14 147 apremilast- and methotrexate-naive patients in the French National Health Insurance database. Br J Dermatol. 2020;182(3):690-697. doi: 10.1111/bjd.18047 [DOI] [PubMed] [Google Scholar]
- 22.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. doi: 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 23.Mason AR, Mason J, Cork M, Dooley G, Hancock H. Topical treatments for chronic plaque psoriasis. Cochrane Database Syst Rev. 2013;CD005028(3):CD005028. doi: 10.1002/14651858.CD005028.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Warren RB, Smith CH, Yiu ZZN, et al. Differential drug survival of biologic therapies for the treatment of psoriasis: a prospective observational cohort study from the British Association of Dermatologists Biologic Interventions Register (BADBIR). J Invest Dermatol. 2015;135(11):2632-2640. doi: 10.1038/jid.2015.208 [DOI] [PubMed] [Google Scholar]
- 25.Xu S, Ross C, Raebel MA, Shetterly S, Blanchette C, Smith D. Use of stabilized inverse propensity scores as weights to directly estimate relative risk and its confidence intervals. Value Health. 2010;13(2):273-277. doi: 10.1111/j.1524-4733.2009.00671.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yiu ZZN, Ashcroft DM, Evans I, et al. ; BADBIR Study Group . Infliximab is associated with an increased risk of serious infection in patients with psoriasis in the U.K. and Republic of Ireland: results from the British Association of Dermatologists Biologic Interventions Register (BADBIR). Br J Dermatol. 2019;180(2):329-337. doi: 10.1111/bjd.17036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dobry AS, Quesenberry CP, Ray GT, Geier JL, Asgari MM. Serious infections among a large cohort of subjects with systemically treated psoriasis. J Am Acad Dermatol. 2017;77(5):838-844. doi: 10.1016/j.jaad.2017.07.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Carneiro C, Bloom R, Ibler E, et al. Rate of serious infection in patients who are prescribed systemic biologic or nonbiologic agents for psoriasis: a large, single center, retrospective, observational cohort study. Dermatol Ther. 2017;30(5):e12529. doi: 10.1111/dth.12529 [DOI] [PubMed] [Google Scholar]
- 29.Sbidian E, Giboin C, Bachelez H, et al. Factors associated with the choice of the first biologic in psoriasis: real-life analysis from the Psobioteq cohort. J Eur Acad Dermatol Venereol. 2017;31(12):2046-2054. doi: 10.1111/jdv.14406 [DOI] [PubMed] [Google Scholar]
- 30.Davison NJ, Warren RB, Mason KJ, et al. Identification of factors that may influence the selection of first-line biological therapy for people with psoriasis: a prospective, multicentre cohort study. Br J Dermatol. 2017;177(3):828-836. doi: 10.1111/bjd.15551 [DOI] [PubMed] [Google Scholar]
- 31.Garcia-Doval I, Rustenbach SJ, Stern R, et al. ; Psonet Network . Systemic psoriasis therapy shows high between-country variation: a sign of unwarranted variation? cross-sectional analysis of baseline data from the PSONET registries. Br J Dermatol. 2013;169(3):710-714. doi: 10.1111/bjd.12344 [DOI] [PubMed] [Google Scholar]
- 32.Winthrop KL, Mariette X, Silva JT, et al. ESCMID Study Group for Infections in Compromised Hosts (ESGICH) Consensus Document on the safety of targeted and biological therapies: an infectious diseases perspective (soluble immune effector molecules [II]: agents targeting interleukins, immunoglobulins and complement factors). Clin Microbiol Infect. 2018;24(suppl 2):S21-S40. doi: 10.1016/j.cmi.2018.02.002 [DOI] [PubMed] [Google Scholar]
- 33.Jin Y, Lee H, Lee MP, et al. Risk of hospitalized serious infection after initiating ustekinumab or other biologics for psoriasis or psoriatic arthritis. Arthritis Care Res (Hoboken). 2021. doi: 10.1002/acr.24630 [DOI] [PubMed] [Google Scholar]
- 34.Griffiths CEM, Armstrong AW, Gudjonsson JE, Barker JNWN. Psoriasis. Lancet. 2021;397(10281):1301-1315. doi: 10.1016/S0140-6736(20)32549-6 [DOI] [PubMed] [Google Scholar]
- 35.Egeberg A, Mallbris L, Gislason GH, Skov L, Hansen PR. Risk of multiple sclerosis in patients with psoriasis: a Danish nationwide cohort study. J Invest Dermatol. 2016;136(1):93-98. doi: 10.1038/JID.2015.350 [DOI] [PubMed] [Google Scholar]
- 36.Aleshaki JS, Cardwell LA, Muse ME, Feldman SR. Adherence and resource use among psoriasis patients treated with biologics. Expert Rev Pharmacoecon Outcomes Res. 2018;18(6):609-617. doi: 10.1080/14737167.2018.1512408 [DOI] [PubMed] [Google Scholar]
- 37.Sahli L, Lapeyre-Mestre M, Derumeaux H, Moulis G. Positive predictive values of selected hospital discharge diagnoses to identify infections responsible for hospitalization in the French national hospital database. Pharmacoepidemiol Drug Saf. 2016;25(7):785-789. doi: 10.1002/pds.4006 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Definitions of Exposures, Associated Inflammatory Diseases, and Comorbidities
eTable 2. All Diagnoses of Infections With Related ICD-10 Codes Included as Any Serious Infection With Subdivision by Site-Specific Groups
eTable 3. Characterization of Serious Infection Cases, n = 1,656
eTable 4. Risk of Serious Infection With One b/bs/ts Antipsoriatic Agent Relative to Another Displayed by IPTW Fine-Gray Model
eTable 5. Risk of Serious Infection According to the b/bs/ts Antipsoriatic Agent Used in the IPTW Cox and Fine-Gray Models, in Sensitivity Analyses (N = 44,239)
eTable 6. Risk of Serious Infection by the b/bs/ts Antipsoriatic Agent Used in the Classical Cox Adjusted Model for the Entire Study Population (N = 44,239)
eTable 7. Risk of Serious Infection Events According to the b/bs/ts Antipsoriatic Agent Used in the Propensity Score-Matched Cox Model, for the Entire Study Population (N = 44,239)
eFigure 1. Flowchart of the Study
eFigure 2. Standardized Differences Before and After Inverse Probability of Treatment Weighting (IPTW) Weighting (A to I)