Shaping Treatment Expectation to Optimize Efficacy of Interleukin 17A Antagonist Secukinumab in Psoriasis Patients

Introduction

Drugs do not operate in isolation. When a doctor prescribes a pill, its effectiveness is shaped not only by its chemical interaction with the body, but also by the patient’s psychological and biological condition. This condition is affected by various elements including genetic makeup, emotional and motivational states like the patient’s expectations of the drug’s outcomes, and prior experiences with this treatment or treatments in general.¹ Together, these components can enhance the overall effectiveness of the treatment, a phenomenon known as placebo effect.^2,3

Placebo effects modulate symptom perception, disease courses, and the efficacy and tolerability of medical treatments.⁴ They are based on complex neurobiological phenomena involving the contribution of central as well as peripheral physiological mechanisms.^5,6 Solid experimental and clinical data demonstrate convincing placebo effects in particular in pain and affective diseases.^4,7,8 Large placebo responses in clinical trials of chronic inflammatory skin diseases such as atopic dermatitis or psoriasis suggest that the mechanisms detailed above should also be considered in the therapeutic context of dermatological diseases.^9,10

This view is supported by experimental evidence in healthy subjects, allergic patients and those with atopic dermatitis.^11–13 The studies demonstrate that both subjective symptoms such as itch and objective markers such as wheal size or basophil activation can be reduced by employing verbally manipulated expectations and learning experiences in the form of classical conditioning. Preliminary outcomes furthermore suggest that harnessing placebo effects may be a viable option to enhance treatment outcome in patients suffering from chronic inflammatory diseases.^13,14

To assess the generalizability of findings from pain and affective research, studies in dermatological patient cohorts are needed. One that may provide a good model is that of persons suffering from psoriasis, as the brain-skin axis was shown to play an important role in this disease.^15,16 Psoriasis is a chronic inflammatory skin disease, characterized by scaly, erythematous plaques that can cover large portions of the body and often cause intense itch.¹⁷ With about 2–3% of the population in western societies affected, it is a comparatively common disease.^18,19 Based on an improved understanding of the underlying pathophysiology,²⁰ recent years have seen radical improvement in the availability and effectiveness of medications.²¹ However, these modern biologic treatments are costly²² and therefore mostly not prescribed as a first-line treatment.^23,24 Harnessing placebo effects to improve treatment outcomes therefore might be a viable option to reduce the dosages needed, thereby lowering costs for the healthcare system and making highly effective treatments available to more patients.

Thus, we employed a dose-reduction regimen of an effective systemic treatment for psoriasis with the interleukin (IL)-17A antagonist secukinumab in combination with the intake of a novel tasting drink accompanied by a cover story to shape patients’ expectations. We compared this group with a dose-control group and assessed both subjective (quality of life, itch, and pain) as well as objective (dermatologist-rated skin symptoms) outcomes.

Materials and Methods

Design

The protocol of this study was published²⁵ and registered online (https://drks.de/search/de/trial/DRKS00022104). The study received approval from the institutional ethics committee of the University-Duisburg Essen (IRB protocol number 19–8636-BO) and complies with the Declaration of Helsinki. It was conducted as part of a collaborative research center on treatment expectations.²⁶ Reporting followed the CONSORT criteria for randomized trials (see CONSORT checklist in supplementary Table 1).²⁷

Figure 1 illustrates the study design. Patients were randomly assigned to three groups in equal ratios. The standard dose group (treatment as usual, TAU) received 300mg of the IL-17A inhibitor secukinumab.²⁸ The dose-control group (DC) received 75mg secukinumab (75% reduction), as did the experimental group (EXP), which in addition received a novel tasting drink and suggestions of heightened treatment expectations. Over 16 weeks, patients followed the secukinumab administration protocol, with assessments of both objective and subjective outcomes at each visit. Pain was rarely reported and therefore excluded from analyses.

Figure 1 Study Design. After an initial pre-check, patients were randomly allocated to one of three treatment groups: The treatment as usual (TAU) group received the full dose of secukinumab as in clinical routine (300mg) and served as pharmacological control. The dose reduction control group (DC) was administered only 75mg of secukinumab (75% reduction) during all treatment weeks. Patients in the enhanced expectation group (EXP) were treated with the same dose reduction regimen as the DC group but received an additional verbal instruction together with the ingestion of a novel tasting beverage to increase the salience of this verbal suggestion. At follow-up, all outcome parameters were measured again and continuation of treatment outside of the study was discussed with patients.

Participants and Recruitment

Patients were eligible for participation if they had a Psoriasis Area and Severity Index (PASI)^29,30 of at least 12 and a Body Surface Area (BSA)³¹ affected by plaque psoriasis of at least 10%. Additional inclusion criteria were derived from the summary of product characteristics for secukinumab.²⁸ Patients were primarily recruited through the outpatient department of the Department of Dermatology, Venereology and Allergology, University Hospital Essen. In addition, advertisements were sent to dermatologists and posted on social media.

Procedure

All study visits were conducted at the Department of Dermatology, Venereology and Allergology, University Hospital Essen. Data collection lasted from October 2020 until November 2023 and ended with the final patient’s last scheduled visit.

During the initial visit, a responsible physician conducted a screening examination to confirm the diagnosis of plaque psoriasis and patients’ eligibility for systemic treatment with secukinumab. Patients were thoroughly informed about the study procedures and the potential risks associated with the treatment. This session also provided an opportunity for patients to ask questions prior to providing their written informed consent. Patients were then randomized to one of three groups using computer-generated random numbers. Participants’ data were pseudonymized using the software ALIIAS (Anonymization with LimeSurvey Integration and II-Factor Authentication for Scientific Research, version 0.9).³²

At each study visit, the participants were welcomed by the study team and escorted to a private treatment room where they completed various questionnaires (details in the Measures section). A dermatologist who was blinded with regard to group allocation then assessed the patients’ skin symptoms. The study physician checked on the patients for any new concerns. With the exception of screening and follow-up visits, patients then received their prescribed medication. To maintain blinding with respect to the dosage, patients were blindfolded while receiving subcutaneous injections. Additionally, patients in the EXP group were given a novel tasting gustatory stimulus (100mL of green-colored strawberry milk with 5 drops of lavender oil; for references, see^11,33) just before the injection. This drink was described by the physician as an additional treatment enhancer purported to activate the “body’s own pharmacy”, a concept introduced to the patients during the initial administration (for the script see supplementary Table 2) and briefly reminded at subsequent visits. The unusual taste was intended to increase the salience of the stimulus. Blood samples were collected every four weeks before drug administration for clinical routine assessments and IL-17A analysis.

The follow-up visit occurred 16 weeks after the initiation of treatment, when all outcome parameters were re-assessed. At this visit, patients were un-blinded, and the possibility of continuing treatment with secukinumab under usual care conditions was discussed.

Measures

Patients’ skin symptoms were assessed using the PASI. To evaluate subjective symptoms, patients completed the Dermatology Life Quality Index (DLQI)³⁴ and rated their current experience of itch and pain on a 100-point visual analogue scale (0.0–10.0 VAS). Additionally, they reported their current symptoms of depression and anxiety on the State-Anxiety-and-Depression-Scale (STADI).³⁵

Patients’ clinically meaningful outcomes were included in secondary analyses; these comprised a 75% reduction in PASI score (PASI75),³⁶ a DLQI score of zero or one (DLQI 0/1) and a reduction on the pruritus visual analogue scale of at least four points (P-VAS ≥4-point improvement).

At the screening visit and again one week after the first administration of medication, patients rated their treatment expectations in the Generic rating scale for previous treatment experiences, treatment expectations, and treatment effects (GEEE).³⁷

To validate the effect of the experimental manipulation on the patients’ immune system, IL-17A was measured in the patients’ blood plasma. To this end, blood samples were drawn into tubes containing an anti-coagulant (S Monovette EDTA, Sarstedt, Nümbrecht, Germany) and separated using centrifugation (2000g, 10 min, 4°C). Plasma samples were stored until analyses at −80°C. Analyses were performed using a Meso Scale Discovery assay (MSD, Rockville, Maryland, USA; detection limit: 1.03–9.46 pg/mL).

Statistical Analyses

The sample size of N = 120 was determined a priori employing G*Power (Version 3.1.9.2,³⁸ for details see²⁵) and taking into account an anticipated drop-out rate of 20%. Data was curated with IBM SPSS Statistics (Statistical Package for Social Science, SPSS Inc., Chicago, version 27) and analyzed using base R (version 4),³⁹ and the R-package nlme.⁴⁰ Plots were created using ggplot2.⁴¹ Up to one missing datapoint per patient was imputed using linear regression and median insertion.

Potential differences in baseline demographics were assessed using Chi²-test for dichotomous, or analysis of variance (ANOVA) for continuous variables. For each continuous dependent variable, a mixed linear model was built using group (3 levels) and study visit (8 points in time) as independent variables. Clinically meaningful outcomes were analyzed calculating each patient’s area under the curve (AUC with respect to ground)⁴² and comparing the groups in one-way ANOVAs and post-hoc pairwise t-tests. Group differences in the GEEE score of positive treatment expectations were analyzed using the Kruskall Wallis test, with False-Discovery Rate (FDR)-corrected post-hoc Dunn tests.

Results

Sample

A total of N = 120 patients were initially included in the study as illustrated in the CONSORT flow diagram (Figure 2). Of these, 34.17% were female. The sample’s mean age was 45.7 years, ranging from 20.6 to 83.7 years. 12.5% of patients suffered from comorbid psoriatic arthritis (refer to supplementary Table 3 for additional comorbidities). On average, patients had received 13.7 years of formal education (range 3.5–35 years). While these three parameters did not significantly differ between groups, both the body mass index (BMI) and baseline PASI score were significantly higher in the EXP group (BMI: F = 3.12, p = 0.048; PASI: F = 3.42, p = 0.036). Spearman’s rank correlation test indicated a significant correlation between baseline PASI and BMI (rho = 0.20, p = 0.030). There were no significant differences between the groups regarding baseline levels of the other two outcome measures DLQI and itch, nor in anxiety and depressive symptoms (Table 1).

Table 1 Baseline Demographics

Figure 2 CONSORT Flow diagram. Abbreviations: TAU, Treatment as usual group with 300mg of secukinumab application; DC, Dose control group with 75mg secukinumab application; EXP, enhanced expectation group with 75mg secukinumab application + verbal suggestion to heighten expectations.

Expectations

Figure 3 depicts positive treatment expectations at the screening visit and after one week of treatment. Groups did not differ significantly in their expectations at baseline (χ² = 3.60, p = 0.165) and showed no significant differences regarding changes in expectations between assessments (χ² = 0.61, p = 0.736). This indicates that the verbal suggestion, which occurred between assessments, did not have a significant influence on overt expectations. Notably, the initial mean expectation rating was 8.1 out of 10 points across groups, indicating a potential ceiling effect in this measure.

Figure 3 Expectation ratings. Positive treatment expectations, assessed with the GEEE scale on a 0–10 numeric rating scale. Ratings were collected prior to (week 0) and after (week 1) the experimental intervention, involving the verbal suggestion of probable treatment success supported by a novel tasting drink for the experimental (EXP) group. Abbreviations: DC, Dose control group; TAU, Treatment-as-usual group.

Primary Outcomes

Absolute numbers for each of the three primary outcomes over the course of treatment are presented in Figure 4. The analysis of PASI scores identified outliers that significantly influenced the results. Six data points that lay more than three standard deviations from the sample means were therefore excluded from the analysis. Across groups, there was a significant improvement of all assessed symptoms over the time of treatment, expressed in a significant main effect of time (PASI: F = 588.53, p < 0.001; Itch: F = 440.78, p < 0.001; DLQI: F = 524.29, p < 0.001).

Figure 4 Absolute numbers over the course of treatment for the three primary outcomes. (A) Psoriasis Area and Severity Index (PASI) over the course of treatment; (B) Itch ratings on a visual analogue scale (VAS) ranging from 0–10; (C) Dermatology Life Quality Index (DLQI), with higher numbers indicating higher burden on quality of life; (D) Interleukin (IL)-17A measured in the blood plasma. Abbreviations: TAU, Treatment as usual group with 300mg secukinumab application; DC, Dose control group with 75mg secukinumab application; EXP, enhanced expectation group with 75mg secukinumab application + verbal suggestion to heighten expectations.

The comparison of treatment groups revealed significantly higher PASI scores in the EXP group (main effect of group: F = 11.78, p < 0.001). At the same time, the EXP group showed a steeper decline in PASI scores over time compared to the DC group, with a mean improvement of 11.01 points compared to 9.02 points (time x group interaction effect; F = 4.73, p = 0.009). Detailed group comparisons can be found in Table 2. 78% of patients in the TAU group achieved a PASI75 response at week 16, compared to 50% in the EXP and 39% in the DC group. Statistical analysis of the AUC revealed a significant advantage of the TAU group, while there was no significant difference between the two dose-reduced groups (for details see Table 3).

Table 2 Mixed Model Results of Primary Outcomes

Table 3 Clinically Meaningful Outcomes

With regard to itch, there was a significant effect of group (F = 5.33, p = 0.006), driven by higher values in the EXP group. DLQI scores did not show a significant group effect (F = 1.53, p = 0.221), but there was a significant interaction effect indicating a less steep decline in symptoms in the EXP group compared to the other groups (F = 6.38, p = 0.002). At week 16, 67% of patients in the TAU group achieved a DLQI 0/1 and 81% a P-VAS improvement of ≥4 points. In the EXP group, these responses were shown by 53% and 58% respectively and in the DC group by 42% and 77%. The comparison of AUCs revealed a significant advantage of the TAU compared to the EXP group for both outcomes and no statistically significant differences between EXP and DC groups (see Tables 2 and 3 for details).

Secondary Outcomes

The analysis of IL-17A levels in patients’ blood plasma revealed a significant increase over time (F = 9.72, p = 0.002), as shown in Figure 4. The groups did not differ significantly in IL-17A levels, and there was no interaction of time and group (refer to Table 4 for details). The assessment of mood as a secondary outcome showed a significant decrease in depressive symptoms over time (F = 57.70, p < 0.001). Moreover, there was a significant interaction effect, stemming from steeper decline in the TAU group compared to the EXP group (F = 12.56, p < 0.001, mean score reduction: TAU = 3.7, EXP = 1.8). Anxiety symptoms also decreased significantly over time (F = 177.34, p < 0.001), with a notable interaction effect of time and group (F = 8.34, p < 0.001), indicating a steeper decline in the TAU group and in the EXP group compared to the DC group.

Table 4 Mixed Model Results of Secondary Outcomes

Adverse Events

A full list of reported adverse events (AEs) can be found in supplementary Table 4. The safety profile was consistent with prior studies on secukinumab and none of the serious AEs were suspected to be treatment-related. Due to the low numbers of AEs that were suspected to be treatment-related, no statistical comparison of groups was conducted.

Explanatory Variables

To analyze potential effects of gender, BMI and age, these variables were included as covariates in the linear mixed models. After FDR-correction for multiple comparisons, there were no significant main or interaction effects with regard to gender. Higher BMI led to higher symptom burden in PASI and DLQI and increased the differences between TAU and the dose-reduced groups when it came to itch. Irrespective of group, younger patients displayed more pronounced improvement over time in all three outcomes. Details of the statistical analyses can be found in supplementary Table 5.

Discussion

The presented study aimed to enhance treatment expectations of patients with psoriasis to improve the effectiveness of low-dose systemic anti-psoriatic therapy with the IL-17A antagonist secukinumab. Patients were randomized to three groups, receiving either treatment as usual or a dose reduction of 75%. Verbal suggestion of probable treatment success supported by a gustatory stimulus was used to heighten expectations in one low dose group.

While all three groups showed symptom improvement over the treatment period of four months, the TAU group was superior in achieving clinically meaningful change (PASI, DLQI, itch). At the same time, there were no clinically meaningful differences between the two dose-reduced groups in none of the three primary outcomes. Over time, the EXP-group showed faster decline in PASI scores compared to the DC-group but less improvement in quality of life. Despite the 75% reduction in dosage there was no difference in plasma IL-17A concentrations, which increased in all groups upon treatment initiation. Depressive and anxiety symptoms significantly decreased over the treatment period in all groups, with a slightly steeper decline in depressive symptoms in the TAU group and reduced anxiety in TAU and EXP groups compared to the DC group.

Past research shows that verbal suggestion can influence treatment expectations and reduce symptoms like pain⁷ and itch⁴³ and enhance medication effectiveness.⁴⁴ As many results so far come from pain studies, the generalizability to a field with differing pathophysiological background needs to be carefully tested. In dermatology, there is evidence for the importance of verbal suggestion and associative learning, as seen in allergy research.^15,45 So far, however, only a single other study has experimentally explored placebo effects in patients suffering from psoriasis.¹⁴ Adding to this small body of research, our study found that verbal suggestion alone is probably not sufficient to ameliorate symptoms in this cohort, corroborating earlier dermatological findings.^46,47

Some studies indicate that associative learning could be a stronger driver of positive expectations,⁴⁸ and combining it with verbal suggestion often produces significant placebo effects.^13,49 Moreover, the effectiveness of placebo treatments may depend on the type of outcome: One study in patients suffering from asthma for example found no difference between placebo and no treatment in expiratory volume. At the same time, placebo and active treatment had comparable effects on patient-reported symptom improvement.⁵⁰ Similar observations in other cohorts have led to the idea that patient-reported outcomes can be influenced by mere verbal suggestion, while objectively measured outcomes like changes in hormone concentrations or immune cell counts need prior experiences, ie associative learning for placebo interventions to be successful.^12,51

The presented study, however, could not find striking differences between patient-reported and physician-assessed outcomes. A potential explanation for this inconsistency could be the fact that prior experimental studies have typically looked into acute symptoms. In a longitudinal setting such as the one employed here, subjective and objective symptoms, like quality of life and observable skin symptoms, may be more closely linked.

The assessment of patients’ treatment expectations showed that these were already on a high level at baseline (8.1 of 10 points), indicating a ceiling effect. Consequently, the experimental manipulation was not successful at further increasing these expectations. Given the setting in a modern university hospital specialized in psoriasis treatment and the option to receive a novel and highly effective treatment with a monoclonal antibody, it is possible that the verbal suggestions by the treating physician were in comparison insufficient to further increase treatment expectations.

The results also show that 25% of the usual dose of secukinumab may already lead to significant symptom improvement and an increase in IL-17A levels similar to that observed in the TAU group. This mirrors what has been seen in a phase-II dose-finding study in a small group of patients (N = 25)⁵² and provides the first real-world evidence for the effects of the 75mg dosage. Considering the immense costs associated with long-term biologic treatment,⁵³ this finding is interesting from an economical perspective: currently the yearly costs of secukinumab treatment amount to approximately 18.600€ or $20.150 per year once the induction phase has been completed⁵⁴ and the application of lower dosages could make this type of treatment available to more people globally.

Strengths and Limitations

A clear strength of this study is its ecological validity, with a setting and procedures that are close to clinical routine. This notion was further promoted by the inclusion of a diverse sample, not excluding any patients based on their demographic information. At the same time, however, this ecological validity comes with certain drawbacks, like the baseline differences in BMI and PASI score seen in our experimental groups. Their starting with a higher burden of skin symptoms may have influenced the treatment trajectories of the EXP group and presents a potential source of bias. While stratification of baseline characteristics is a valuable option that should be considered in future studies, there are certainly limits to the feasibility of this method.

Future Directions

An interesting route for the future could be the employment of other strategies to harness the placebo effect in this type of treatment. Expectations may have been too high to be raised any further by verbal manipulation, but this may be achieved by other methods, like social observation⁵⁵ or associative learning.^56,57 While associative learning has already been tested successfully in a pilot study with psoriasis patients under topical treatment,¹⁴ social learning as a method for symptom relief has – to our knowledge – not been studied in dermatological patients so far.

Conclusion

Taken together, the presented results indicate that the mere verbal suggestion of treatment success might be insufficient to shape positive treatment expectations in psoriasis patients in the context of already high expectations. However, they also illustrate the feasibility of the employed model to study placebo effects in this cohort: the reduced dosage was still sufficient to elicit symptom improvement, which is an important ethical consideration when designing such studies and other mechanisms like associative learning or social observation could be tested in a similar design. The advancement of such approaches could on the one hand improve clinical care, by achieving the same treatment success while reducing the amount of medication intake necessary. On the other hand, an exploration of placebo effects in patients suffering from an inflammatory disease like psoriasis can generate important insights into the generalizability of a field that has mostly grown in the domain of pain.