Zenas Z.N. Yiu, Richard B. Warren

Title page:

High Standards are Paramount for the Evaluation of New Biologics and Small Molecule Therapies for Psoriasis: A Set of Criteria for Future Novel Therapeutics

Authors:

Zenas Z.N. Yiu, Richard B. Warren

Affiliation:

Dermatology Centre, Salford Royal NHS Foundation Trust, The University of Manchester, Manchester Academic Health Science Centre, NIHR Manchester Biomedical Research Centre, Manchester, UK, M13 9PT

Correspondence:

Dr Richard Warren

Address:

The Dermatopharmacology Unit, Salford Royal NHS Foundation Trust, Barnes Building, Manchester M6 8HD

Telephone:

+44 (0) 161 206 4344

Email address:

Number of references: 47; Figures: 0; Tables: 1; Word count: 3230

Article type: Review

Key words:

Psoriasis; Biologics; Novel Therapeutics

Introduction (75 word introduction):

The quality of life for a patient has been transformed in the last 15 years due to new innovations that have resulted in better treatments for severe psoriasis, a chronic inflammatory cutaneous disease. Now, novel therapies for psoriasis need to reach a high standard in order to offer patients with psoriasis a genuine alternative choice. Here, we outline a suggested critical checklist that will help industry sponsors, researchers and clinicians evaluate novel therapeutics for psoriasis.

Introduction

The quality of life for a patient with severe psoriasis, a chronic inflammatory cutaneous disease, has been transformed in the last 15 years with novel innovative therapeutics, most notably biologic therapies. Historically, patients with severe psoriasis could expect multiple disruptive admissions to hospital for treatment with coal tar and UV therapy throughout their lifetime, interspersed with long-term immunosuppressive treatments that may have a lasting detrimental effect on the kidneys and the liver; now, weekly, fortnightly or even three-monthly injections are enough to keep the discomfort and visible impact of the disease effectively controlled.

The current licensed biologic therapies for psoriasis include etanercept, infliximab and adalimumab, which are tumour necrosis factor inhibitors (TNFI); ustekinumab, an interleukin(IL)12/23 inhibitor and; secukinumab and ixekizumab, which are IL17-A inhibitors. Furthermore, apremilast, an oralphosphodiesterase 4 inhibitor, has also been licensed. Although a cure has not been forthcoming, the most recently licensed biologic therapies achieve an unprecedented level of disease control where Psoriasis Area Severity index (PASI; a common disease severity score) 100, in other words 100% of visible psoriasis removed, is not uncommon.

With the increasing quality of therapies available and an array of novel therapies in the pipeline, it is important to reflect on the progress made in the field and outline the high standard that any novel therapy for psoriasis needs to reach in order to become a valuable addition to the current armamentarium and offer patients with psoriasis a genuine alternative choice.

Targets for Efficacy

Clinical trials for psoriasis therapies have traditionally been powered to investigate the difference in the proportion of patients achieving a PASI 75 at the primary endpoint, usually around week 12-16 between the therapy and placebo. PASI 75 rates have increased from the mid-30/mid-40 percentage range, e.g. for etanercept, a fusion protein targeting tumour necrosis factor(TNF)-α(1), to the high 70 / high 80 percentage for secukinumab and ixekizumab respectively, which are both interleukin(IL)-17A inhibitors. Indeed, the PASI 100, or complete clearance, figures for secukinumab (28.6% at week 12(2)) and ixekizumab (35.3% at week 12(3)) are comparable to the PASI 75 figures for etanercept, outlining the progress in treatment efficacy that has been made in the last decade.

Novel therapeutics, ideally, need to show efficacy that is better than the gold-standard current treatments, i.e. the IL-17 inhibitors. The sample sizes for the randomised clinical trials (RCT) for the therapeutics should be calculated based upon anticipated difference for PASI 90. It would be anticipated that the yield for any gain in efficacy for novel therapeutics over the current gold-standard treatment may become ever smaller, and therefore sample sizes will need to increase for future trials.

Given the vast number of treatment options for the treatment of psoriasis, from more historical treatments such as methotrexate to the recently licensed treatments, placebo as a comparator is clinically irrelevant and inadequate. Similarly, drug companies should avoid choosing an out-dated comparator drug in the phase III clinical trials, e.g. etanercept, but instead focus on investigating for superiority to the gold-standard IL-17 inhibitors. The costs saved through not conducting a placebo-controlled trial could be directed to funding an adequately powered RCT comparing with a gold-standard treatment.

In addition to reporting the statistical tests and p-value for the measure of association, RCTs should also report measures of effect, such as the relative effect(e.g. an Odds Ratios with 95% confidence intervals) for the probability of achieving PASI 100, along with an absolute benefit measure such as the Number Needed to Treat (NNT) to help aid clinical decision making.

Occasionally, there may be a case for investigating novel therapies that do not theoretically have better efficacy than the gold-standard treatments, but have other advantages such as a better safety profile. This may be particularly pertinent for novel oral therapies, which have a more acceptable administration method than injections. For example, the license for apremilast allows for a reduction in hospital visits and monitoring blood tests as compared to the biologic therapies for psoriasis as no pre-treatment screening noractive lab monitoring is required. In those circumstances, non-inferiority RCTs to gold-standard comparators should be performed, making it transparent regarding how the non-inferiority margin was decided a priori. A separate or combined superiority RCT for specific well-defined safely outcomes should then be performed, with a good example being the recent BLISTER trial which compared doxycycline to prednisolone for treatment-related adverse events for bullous pemphigoid, an inflammatory skin condition associated with tense blisters(4).

Adequately powered A priori secondary and subgroup analyses

Subgroup analyses and/or secondary analyses are frequently reported in RCTs for psoriasis biologic therapies. For example, there were 11 different reports from subgroup analyses, some pooled from several RCTs to increase power, for secukinumab on Pubmed (search term: secukinumab AND psoriasis) (5-15). It is important that reporting of such analyses is transparent, with both specified a priori in trial registries and the power to detect differences in a particular subgroup also reported. There is a risk that these analyses may find spurious results due to multiple comparisons; conversely post-hoc subgroup analysis of a single trial may lack the power to find true effects (16). Subgroup analyses should be carefully considered and powered for a priori and a limited number performed with adjustment for multiple comparisons where necessary, while post-hoc subgroup observations should be avoided (17, 18). This should be documented in the trial protocol to increase credibility of the results from subgroup analyses (19).

There are several subgroups that all novel therapeutics for psoriasis should arguably be powered for test in. Most RCTs involve a mixed population with regards to treatment history. However, the population that stands to gain the most from novel therapeutics are patients in whom the existing therapeutics has failed. Novel therapeutics should therefore be evaluated in this group of patients; the SIGNATURE trial for secukinumab, which is yet to be published, is a good example of this approach, where patients who did not adequately respond to TNFIs or ustekinumabwere recruited. Manufacturers of future therapeutics should consider recruiting patients who have failed multiple biologics, i.e. never responded to any of the biologics or subsequent loss of response, into RCTs. It may be worthwhile to conduct mechanistic led studies in populations where patients have developed anti-drug antibodies to existing biologic therapies for psoriasis leading to a loss of response.

Another important subgroup of patients is those who suffer from both severe psoriasis and psoriatic arthritis (PsA), a destructive seronegativespondyloarthropathy affecting between 15-20% of patients with psoriasis. Efficacy has been shown for TNFIs(20), ustekinumab(21), and IL-17 inhibitors for PsA(22, 23), but as yet no trial has looked at a combined measure of psoriasis and PsA responding in the same patient. RCTs need to consider evaluating patients holistically and combination PsA and psoriasis outcomes may be useful in this group of patients.

To date, the search for clinically useful pharmacogenomics markers of response for biologic and small molecule therapeutics in psoriasis has been disappointing. Several small studies initially suggested that the presence of human leukocyte antigen (HLA)-Cw6 (C*06:02) allele would be a predictor for response for ustekinumab, but a large retrospective study based on patients from clinical trials did not show a clinically relevant difference in response between patients with and without the allele.(24)However, there continues to be active research into delineating disease endotypes in psoriasis, where clinically seemingly identical presentations of psoriasis are further stratified based upon differential pathophysiological mechanisms or pharmacogenomic signatures. Future evaluations of novel therapeutics in psoriasis should take into account new and consistent discoveries of disease endotypes in psoriasis, and through understanding of the drug mechanism identify and power the studies in the relevant endotype.

Novel metrics and Patient Reported Outcome Measures

PASI has been used as a measure of efficacy in all of the biologic therapies for psoriasis to date. However, there are several limitations to this instrument: it has not been validated; it is cumbersome to use; there is poor interrater variability; it requires estimation of body surface area affected; and it does not prioritise sites which may be more sensitive to patients, such as the face or the hands (25). PASI also does not place any emphasis on the patient’s quality of life or their own perception of severity, and often RCTs include change in the Dermatology Life Quality Index, a generic quality of life score for skin diseases that is filled by the patient, as a secondary outcome measure. There are several other instruments available, e.g. the National Psoriasis Foundation-Psoriasis Score(26); the Beer Sheva Psoriasis Severity Score(27); the Copenhagen Psoriasis Severity Index(28) and the Simplified Psoriasis Index(29). There is a need for these disease severity scores to be validated in different patient populations and cultural context, but a shift towards a better toolfor the assessment of efficacy maybe needed to better compare novel therapeutics where differences are becoming more subtle. Co-primary endpoints involving patient reported outcomes should also be considered, as these are arguably more relevant to the patient as compared to the traditional clinician-assessed disease severity outcomes.

Pragmatic Phase III trials and Observational Data for Effectiveness and Safety

Given the increasing fine margins between novel therapeutics and current therapeutics for psoriasis, it is not enough to demonstrate efficacy in an RCT. Gains in efficacy may not be the benefit of novel therapeutics anymore; with further understanding in factors underlying immunogenicity (where antibodies produced towards the drug causes a loss of response) and adherence,the goal of novel therapeutics may shift towards providing longer lasting benefit to patients and avoiding drug discontinuation. In addition, clinical trials are known to lack external validity given their strict inclusion and exclusion criteria, and it has been estimated that around 30% of patients with moderate to severe psoriasis would not have been eligible for RCTs (30). Furthermore, patients who are ineligible are estimated to have a 3-fold increase in the risk of severe adverse events as compared to those who would be eligible(30).

There is now a trend to incorporate measures of effectiveness in the pre-licensing phase of a drug in other fields. High profile examples of pragmatic clinical trials are the aforementioned BLISTER trial, or the Salford Lung Study where the real-world utility of a drug for chronic obstructive pulmonary disease was investigated (31). Although the setting is different in that psoriasis patients eligible for current biologic therapies are seen in secondary care dermatology clinics rather than in primary care, pragmatic pre-licensing trials, especially for the subgroups highlighted earlier in the article, may be an increasingly more attractive option with the advent of the use of electronic health records and other novel methods of data collection, and may be the only option that incorporates randomisation where benefits in adherence or drug persistence can be measured. They have been underutilised in the evaluation of dermatological diseases in the past(32).

Observational studies are also extremely important in the evaluation of biologic therapies in psoriasis, especially for safety outcomes. Clinical trials are not powered to investigate serious adverse events, which are often rare. A famous example in dermatology is the association of progressive multifocal leukoencephalopathy with the biologic efalizumab, which was not discovered in the clinical trials, and only emerged when cases were picked up in post-marketing pharmacovigilance studies. For research and post-marketing licensing requirements, registries of patients with psoriasis on biologic therapies have been established around the world. Most of these are designed with a comparator cohort with a similar disease severity on non-biologic systemic therapies (an example being the British Association of Dermatologists Biologic Interventions Register, a UK and Ireland based psoriasis treatment registry of over 12000 patients(33)), and are very important in quantifying the safety or risk of biologic therapies for psoriasis as well as enabling valuable insights into drug survival and real-world dosing strategies of biologics(34).

Harmonisation of Safety Outcome Terminology

The evaluation of the safety of the novel therapeutics for psoriasis is also an important consideration. RCTs, and especially pooled analysis or systematic reviews of RCTs, can be informative in picking up safety signals of more commonly associated adverse events. However, in a recent systematic review, we found that for the outcome of serious infection, trial reports varied substantially in how they labelled the event(35). Although all serious adverse events were commonly defined by the FDA, the specific serious events attracted a varied nomenclature and therefore it was unclear whether they all referred to the same type of event. We also found that it was often not transparent whether an adjudication panel was convened. The definition of adverse events, especially adverse events of interest from phase I and II trials, should be clearly defined in the trial report, and there should be clear statements of the use; or lack of; adjudication panel. There are common concerns over novel therapeutics for psoriasis, and these include the association of the drug with: serious infections; cardiovascular events; psychiatric events such as suicide; and causation or exacerbation of other inflammatory conditions such as inflammatory bowel disease. Extra a priori concerning adverse events may be driven by prior mechanistic studies. It can be argued that an adjudication panel should be convened for these important serious adverse events for all future novel therapeutics. Trials should report adverse events in accordance with the extension CONSORT statement(36).

Dosing and Pricing

Dosing regimen for most biologic therapies reflects a strict “one size fits all” approach. Adalimumab, etanercept (TNFIs), secukinumab and ixekizumab (IL-17 inhibitors) all have one licensing dosing regimen for all patients, regardless of patient background or characteristics. Often fixed dosing is preferred due to ease of preparation and reduced costs. This is at odds with most oral small molecule therapies. Infliximab and ustekinumab, on the other hand, are given by milligram per kilogram and by two different doses dichotomised by a weight of 100kg respectively. It is not a surprise, therefore, when a large real-world registry found that a higher body mass index (BMI) predicted cessation of biologic therapies In addition, drug survival, a proxy marker for effectiveness, of ustekinumab was significantly better than adalimumab(37). Obesity is known to influence biologic bioavailability for other conditions (38), and a difference in blood flow and tissue capillary density in adipose tissue at the subcutaneous injection site may also contribute to variations in real-world effectiveness of the drug.In a recent analysis of pharmacokinetic data from two RCTs for adalimumab, the group with a weight of 104kg had a 80% higher clearance of adalimumab as compared to the group with a weight <76kg(39). Subgroup analyses for efficacy stratified by BMI should be an a priori outcome, and given this real-world evidence the RCTs for novel therapeutics should be powered to investigate interactions between BMI and the drug for efficacy outcomes.

Immunogenicity of biologic therapies is a well-recognised but poorly understood phenomenon. It is not entirely known what factors drive production of anti-drug antibodies (ADAs), but they have been associated with a decrease in efficacy (40) and increased drug clearancefor adalimumab(39). In countries where there is a limited public healthcare resource such as the National Health Service in the UK, it is more common to manage the secondary loss of response of biologic therapies for psoriasis with a concomitant non-biologic systemic therapy such as methotrexate rather than increase the dose or shortening the dosing interval of the biologic therapy, as obtaining funding for this can be extremely difficult(41). However, the combination of two potent immunosuppressants may lead to more adverse effects, e.g. serious infections (42).Drug companies should consider allowing a flat price structure for different dosing regimen of the treatment to ease dose escalation.

The Importance of Transparent Reporting

Most medical journals now require trial reports to adhere to the CONSORT statement, which is a welcome step towards standardisation of trial reports. Traditionally, trial protocols were not made available with the published reports evaluating biologic therapies for psoriasis. A recent review into the external validity of phase III trials of systemic agents in psoriasis was limited by the lack of access to the original trial protocols (43). Encouragingly, recent RCTs, for example the UNCOVER trials for ixekizumab, have started to make their protocol available with the report (3). This is vitally important for a multitude of reasons, with one of them being the ability to assess which outcomes are specified a priori.