The Influence of Inhaled CorticoSteroids Adherence on Treatment Response to Mepolizumab in Severe Eosinophilic Asthma

General demographics and clinical characteristics of the mepolizumab treated population in BRSAS will be presented and these will include age, gender, ethnicity, asthma severity measures in the form of symptoms, lung function, treatment use and requirement for maintenance oral corticosteroids, exacerbation frequency, admissions and emergency department visits.

Treatment response to mepolizumab will be assessed at the one year end of treamtent in which patients will be labelled as responder if they achieved a meaningful clinical response using NICE criteria of reducing severe exacerbation frequency by 50% or more and/or reducing the dose of maintenance OCS by similar magnitude. The overall response rate in the 250 patients will be reported as a responder proportion. The overall benefit from treatment in the form of its effect on asthma measures such as asthma control questionnaire, lung function, biomarkers of blood eosinophils and exhaled nitric oxide, admissions, emergency department visits and any reported side effects will be reported.

Fraction exhaled nitric oxide (FeNO) is an inflammatory marker of asthma. FeNO has been used as a marker of non-adherence to inhaled corticosteroids (ICS). Patients non-adherent to ICS tend to have high FeNO as ICS often potently suppress FeNO. High FeNO level despite ICS treatment would indicate continuing active asthma airway inflammation and increased risk of exacerbation. We will divide patient groups into high and low FeNO groups using different cut-off levels of 25,35,50 parts per billion (PPB). FeNO high and low groups will be compared in terms of severe asthma exacerbation frequency to determine if persistent high FeNO is associated with exacerbations. An increase in exacerbation frequency in FeNO-high patients would indicate that FeNO can be used as surrogate marker of ICS non-adherence or inadequate suppression of T2 asthma inflammation due to ICS resistance with both may drive incomplete response to mepolizumab treatment.

As per outcome 4 we will categorise patients as FeNO high and FeNO low groups using cut off levels of 25,35,50 ppb and compare the response to mepolizumab in the FeNO high/low groups. Response to empolizumab will be determined as described in outcome 3. Patients continue to express high FeNO despite mepolizumab treatment are likely to have the disadvantage of unsuppressed type 2 asthma inflammation and therefore are at risk of exacerbation and are probably less likely to have be responder to mepolizumab. This assertion however has not been previously definitely confirmed and this analysis will help in clarifying this matter further.

We will look at the prescribed dose of inhaled corticosteroids (ICS) and its relation to FeNO level. ICSs are potent suppressors of FeNO and it is expected therefore that patients receiving regular ICS treatment will express lower FeNO level than those not on ICS or are non-adherent to ICS. It is not certain if there is actual relation between ICS dose as measured in microgram of beclomethasone equivalent daily dose FeNO level as too many variables may affect this (for example, non-adherence to ICS, ICS resistant patient groups, variable delivery efficiency of the drug to the airways). Therefore in this analysis we will perform correlation and association analysis of mean FeNO level and mean ICS daily dose (beclomethasone equivalent) to explore the effect of ICS of FeNO in this group of patient who are also treated with mepolizumab at the same time.

One of the measures of treatment adherence to preventer therapy (ICS) used in clinical practice the prescription possession ratio (PPR) (=total number of ICS inhaler prescriptions picked up in the 12 months period whilst on mepolizumab / expted number to be picked up in this period). A PPR ratio of 70% or higher has been used in literature to indicate adequate adherence and thefore this cut-off level has been accepted in current clinical practice in UK (NHSE severe asthma tool kit). Complementing the analysis of relationship between mean ICS dose and mean FeNO (outcome 6), the analysis of PPR relation with FeNO is expected to be more accurate as it does take into consideration the potential strong confounder of ICS non-adherence. If confirmed that PPR is related to FeNO level then interventions to enhance adherence to ICS can be instituted to suppress FeNO and in turn improve asthma outcomes.

Response rate to mepolizumab as measured by effective reduction in either maintenance oral corticosteroids dose (OCS) and /or frequency of OCS requiring exacerbations at the one year end of mepolizumab treamtent, will be compared in patients who are adherent to ICS (adherent = those who has precription possession ratio (PPR) of 70% or higher) or non-adherent. The hypothesis here is that ICS non-adherence can influence response rate to mepolizumab as part of the asthma inflammation no suppressed enough by mepolizumab will persist due to lack of adequate ICS treatment and therefore mepolizumab will less likely work in patients not taking ICS adequately. If this is demonstrated through this analysis, again more emphasis will be put to enhance the adherence to ICS that would then improve response rate to mepolizumab.

In this analysis we will compare the mean ICS daily dose and PPR in the 12 months before and the 12 months on mepolizumab treatment. Several studies showed that initiation of biologic treatment such as mepolizumab can result in reduced adherence to ICS. Asthma improvement due to mepolizumab with reduced daily symptoms of asthma may result in patients not persevering with their daily regular ICS treament or patient may feel that a potent asthma injection such as mepolizumab is sufficient and do not need ICS (often despite medical advice). In this analysis we will therefore explore if there is evidenc of change in behaviour in terms of PPR before and on biologic treamtent period and if demonstrated then factors driving such change in adherence will require further exploration.

We will compare clinical characteristics of patients with poor ICS adherence (as measured by those with PPR ration of <70%) to the adherent groups with aim to identify an predictors of non-adherence that may prompt further research into drivers of non-adherence to eventually develop tools and intervention to address these to improve patients outcomes. Clinical charactersitics that would include age, gender, ethnicity, asthma symptoms control level, quality of life, exacerbatin frequency, use of maintenance OCS, lung function and others will be explored in this analysis.