Comparing the odds of reported depression in psoriasis patients on systemic therapy: a cross-sectional analysis of postmarketing data

Abstract

Patients with psoriasis are more likely to experience depression and suicidality compared to non-psoriatic patients, though systemic therapies have been shown to improve depressive symptoms. It is unclear whether or not biologic or oral agents are more effective at improving such depressive symptoms in psoriasis patients, however. We aimed to determine an estimate of the odds of incident depression in psoriasis patients on different systemic therapies by performing a cross-sectional analysis of postmarketing data. The reporting odds ratio (ROR) for 15 different systemic agents was calculated using reports from the Food and Drug Administration Adverse Events Reporting System (FAERS). After excluding brodalumab and apremilast due to high risk of reporting bias, we found oral agents were associated with a significantly higher ROR of depression compared to biologics (OR = 2.42, 95% confidence interval: 1.93–3.04). These results suggest biologics may be more effective at reducing incident depression than oral agents. Future controlled trials are needed to confirm these findings.

Keywords:

• Suicide
• biopharmaceutical
• psychiatric disorder
• pharmacovigilance
• antidepressant

Introduction

Psoriasis can severely impact patients’ quality of life and is known to significantly increase the risk of developing depression compared to non-psoriatic patients (1). First-line therapy for the majority of patients with major depressive disorder includes pharmacologic therapy with antidepressants and cognitive behavioral therapy (2). Treatments approved for psoriasis, however, also have the potential to improve mood in these patients. A systematic review and meta-analysis showed that biologics not only improve physical disease severity of psoriasis patients, but also improve depressive symptoms (3). The question of whether some medications are more effective in improving depression than others in psoriasis does not have a definite answer, however. There is some evidence, though, that biologic mediations have a greater antidepressive effect compared to oral therapies in other chronic cutaneous diseases.

For example, an analysis of pharmacovigilance data using the Food and Drug Administration Adverse Events Reporting System (FAERS) found that in patients with hidradenitis suppurativa (HS), TNF-alpha inhibitors were associated with a decreased risk of suicidal behavior compared to other medications used to treat HS (4). FAERS is a publicly available database designed by the Food and Drug Administration (FDA) that contains worldwide adverse event and medication error reports submitted voluntarily to the FDA by healthcare professionals, consumers, and manufacturers. Reports sent directly to manufacturers from healthcare professionals or consumer are required to be sent to the FDA (5). A limitation of the FAERS database is the inability to determine if an adverse event was truly caused by the product in question. Though this can lead to reporting bias, the FAERS database can be used to determine a rough comparison of the rates of adverse events between different medications.

Given that there are a number of systemic medications approved for psoriasis, both oral medications and biologics, it is of great interest to the clinician to determine whether certain classes of medications are more effective in improving depressive symptoms than others. Such data could help guide practitioners in selecting systemic medications for psoriasis patients who are depressed knowing they may receive an additional benefit in depressive symptom reduction from one treatment compared to another. Thus, we aimed to determine an estimate of the relative risk of associated depression among systemic medications approved for psoriasis by calculating reporting odds ratios (ROR) using the FAERS database (6).

Materials and methods

We examined adverse event data for four oral agents (acitretin, apremilast, methotrexate, and cyclosporine) and eleven biologics (adalimumab, etanercept, infliximab, certolizumab, ustekinumab, secukinumab, ixekizumab, brodalumab, guselkumab, tildrakizumab, and risankizumab) approved for moderate to severe psoriasis using the publicly available FAERS website. For each medication, we extracted the number of adverse events from cases with a listed indication of ‘psoriasis.’ We considered the total number of depression events as the sum of occurrences of the terms ‘depression,’ ‘major depression,’ and ‘suicidal depression.’ The ROR is defined as the odds of an event occurring with a medication compared to the odds of such event occurring with all other considered drugs (6). We calculated the ROR of depression for each systemic medication approved for psoriasis. A higher ROR indicates a higher likelihood for depression to be reported for that medication. Since brodalumab and apremilast both include warnings for depressive symptoms and suicidality on their drug labels (7,8), it can be inferred that depressive symptoms are more likely to be closely monitored and reported. Thus, we also conducted another analysis with these two medications excluded to reduce reporting bias.

Results

Our initial analysis confirmed that both brodalumab and apremilast had the highest ROR of depression (Table 1). After excluding both medications, we found all other oral agents were associated with a significantly higher ROR of depression compared to biologics (2.42, 95% confidence interval: 1.93–3.04) (Table 2). Among biologic agents, the TNF-alpha inhibitors, certolizumab, infliximab, and adalimumab were associated with a significantly higher ROR of depression than other biologics. Etanercept, secukinumab and ixekizumab had the lowest RORs of depression.

Table 1. Reporting odds ratio (ROR) of depression for all systemic medications investigated.

	ROR	p-Value
Apremilast	4.88* (5.28–4.51)	<0.001
Brodalumab	4.08* (6.40–2.61)	<0.001
Risankizumab	2.32 (16.91–0.32)	0.41
Acitretin	1.70* (2.53–1.14)	0.009
Cyclosporine	1.28 (2.05–0.81)	0.30
Methotrexate	1.07 (1.48–0.77)	0.70
Certolizumab pegol	1.07 (1.81–0.63)	0.82
Infliximab	0.82 (1.16–0.58)	0.26
Adalimumab	0.59* (0.67–0.52)	<0.001
Ustekinumab	0.56* (0.68–0.47)	<0.001
Guselkumab	0.50* (0.78–0.32)	0.002
Ixekizumab	0.44* (0.60–0.33)	<0.001
Secukinumab	0.41* (0.48–0.35)	<0.001
Etanercept	0.37* (0.42–0.33)	<0.001
Tildrakizumab	0.00 (n/a)	n/a

An ROR greater than one indicates a higher likelihood that depression was reported as an adverse event for that drug compared to other drugs examined, and a ROR less than one indicates depression was less likely to be reported compared to other drugs. Upper and lower limits for 95% confidence intervals are reported in parentheses. ROR for tildrakizumab was unable to be calculated because there were no depression-related adverse events. Although risankizumab had a relatively high ROR, this value was not statistically significant due to a low number of reported depression events. * represents statistically significant values (p < 0.05).

Table 2. Reporting odds ratio (ROR) of depression for all systemic medications investigated excluding brodalumab and apremilast.

	ROR	p-Value
Risankizumab	4.27 (31.14–0.59)	0.15
Acitretin	3.16* (4.72–2.12)	<0.001
Cyclosporine	2.38* (3.81–1.49)	<0.001
Methotrexate	2.00* (2.78–1.44)	<0.001
Certolizumab pegol	1.97* (3.35–1.16)	0.012
Infliximab	1.53* (2.16–1.08)	0.017
Adalimumab	1.23* (1.41–1.07)	0.003
Ustekinumab	1.08 (1.31–0.90)	0.4
Guselkumab	0.93 (1.45–0.60)	0.76
Ixekizumab	0.82 (1.11–0.61)	0.21
Secukinumab	0.79* (0.93–0.67)	0.006
Etanercept	0.76* (0.86–0.67)	0.001
Tildrakizumab	0.00 (n/a)	n/a
All biologics (excluding brodalumab)	0.41* (0.51–0.33)	<0.001
All orals (excluding apremilast)	2.42* (3.04–1.93)	<0.001

An ROR greater than one indicates a higher likelihood that depression was reported as an adverse event for that drug compared to other drugs examined, and a ROR less than one indicates depression was less likely to be reported compared to other drugs. Upper and lower limits for 95% confidence intervals are reported in parentheses. ROR for tildrakizumab was unable to be calculated because there were no depression-related adverse events. Although risankizumab had a relatively high ROR, this value was not statistically significant due to a low number of reported depression events. * represents statistically significant values (p < 0.05).

Discussion

Our analysis shows that overall, psoriasis patients on biologic therapies are significantly less likely to have depression reported as an adverse event compared to psoriasis patients on oral agents. Thus, biologics may be more effective in preventing incident depression compared to oral agents. These results parallel a 2021 Cochrane review and meta-analysis which found that biologics are more effective in improving psoriasis than oral agents (9), suggesting that patients whose skin disease is more effectively being treated are less likely to experience depression.

Additionally, there may be interclass differences between anti-TNF-alpha, anti-IL-17, anti-IL-23, and anti-IL-12/23 biologics with regard to antidepressant impacts. Anti-IL-17 inhibitors (except brodalumab) had among the lowest odds of having depression reported as an adverse event. Among biologics, three of the four TNF-alpha inhibitors were associated with the highest RORs of depression with the exception of etanercept, which interestingly had the lowest ROR. The recent Cochrane review also found that TNF-alpha inhibitors were among the least effective in treating psoriasis, with the exception of infliximab (9). A limitation with comparing the results from our analysis of the FAERS database to clinical trial data is that clinical trials typically report short-term adverse events, while postmarketing data includes reports from patients who are using a drug for any period of time.

Depression in chronic inflammatory conditions like psoriasis has been also shown to be associated with imbalances in inflammatory cytokines (10). A meta-analysis of anti-cytokine biologics used in multiple different inflammatory conditions (e.g., Crohn’s disease, psoriasis, and rheumatoid arthritis) found that biologics reduced depression symptoms independently of physical disease improvement (3). Thus, biologics likely prevent depression through improvement in skin disease and possibly through a targeted anti-cytokine effect not seen in oral agents. Other explanations which were not explored in this analysis, such as medication side effect profiles (e.g., fatigue secondary to methotrexate use), can contribute to the incidence of depression (11).

One of the major limitations of this study is reporting bias. Because brodalumab and apremilast have warnings for increased risk of suicidality and depression, respectively, providers may be more likely to report depression outcomes compared to medications without these warnings. Three suicides were reported during the phase III clinical trials for brodalumab, however, postmarketing data has not shown an increased risk for suicidality in psoriasis patients on this medication, calling this association into question (12). Brodalumab, like other biologics, was actually shown to decrease depressive symptoms compared to placebo in phase III trials (13). Apremilast was shown to significantly increase the risk of depression in clinical trials compared to placebo, similarly to another PDE4 inhibitor, roflumilast (14). Another limitation of this investigation was that it was non-controlled; head-to-head trials are needed to confirm our results, though restricted by the feasibility of comparing all 15 agents. This analysis compared odds of depression among medications, thus no conclusions can be drawn in relation to patients who were untreated. Some of the more recently approved medications (e.g., risankizumab, tildrakizumab) lacked enough data for significant associations to be determined.

Overall, our results suggests biologics may be more effective in reducing incident depression compared to oral agents in psoriasis patients.These results may be useful for practitioners to consider when selecting a systemic medication for psoriasis patients who are depressed or at risk for depression, though more high-quality research is needed.

Disclosure statement

Dr. John Koo has served as an advisor for AbbVie, Amgen, Celgene, Janssen, Eli-Lilly, Leo Pharma, EPI, Novartis, Pfizer, Sun Pharma, and Ortho Dermatologics, Regeneron/Sanofi, UCB. The remaining authors have no conflicts of interest to disclose.