Avanir Pharmaceuticals: Uncovering Nuedexta's Potential Part I

| About: Avanir Pharmaceuticals, (AVNR)

We have prepared an in-depth two part analysis of prospects for Avanir’s (NASDAQ:AVNR) recently launched drug for pseudobulbar affect [PBA] called Nuedexta.

This first article will deliver a detailed re-analysis of the likely market for Nuedexta stratified by disease severity and consisting only of those most likely to be treated. Our next article will assess the many variables that are likely to impact Nuedexta use and provides peak US Nuedexta sales projections by using the market data developed in Part I and applying the impacting factors to these markets. There are some estimates required for this project but we hope to make them with the best available evidence incorporated and then leave the door open to refinement of the analysis.

Avanir’s current estimates of the US PBA prevalence rates are based on literature review of diagnoses associated with PBA (10-20%), physician survey data (14-18%), and a survey of 2,324 patients from the Harris Interactive Chronic Illness panel (10% moderate-severe, additional 28% mild). The latter findings were reported in poster form by W.G. Bradley et al at the 2007 AAN meeting in Boston (“Bradley survey”).

For their own estimates, Avanir focuses on the data from the Bradley survey. Their estimates of PBA prevalence from this survey are based upon the Center for Neurologic Study’s Lability Scale [CNS-LS] with scores in the 13-20 range considered to be mild and those with scores ≥21 considered to be moderate-severe PBA. The three PBA estimation methods show overall prevalence estimates running from 1.8 million to 7.3 million depending on the varying methods and definitions of PBA used. This wide range suggests the “market” for Nuedexta is imperfectly defined.

Physicians do not treat and patients do not ask for treatment based on a CNS-LS score. We believe the primary driver of a patient’s desire to seek treatment, as well as a physician’s willingness to treat, to be the “burden” of PBA on its sufferers. As such, we only considered “burdened” patients (or caretakers) to comprise the market for Nuedexta regardless of their CNS-LS. Three survey studies have been commissioned by Avanir that were designed to allow for a self-reported “burden” measure. The patient recruitment for two of the recent studies, conducted by the National Stroke Association [NSA] for stroke and the Brain Injury Association of America [BIAA] for traumatic brain injury [TBI] had, as a condition for participation, that the patient had difficulty controlling their emotions.

Unfortunately, this recruitment method introduces a bias toward patients with PBA and a likely bias toward those with more severe symptoms of PBA. The third study, conducted by the Multiple Sclerosis Association of America [MSAA], recruited from an apparently unselected group of 20,000 with roughly 5,000 participating. The relatively unbiased nature of the MSAA survey is supported by the fact that the findings match up almost exactly with the Bradley survey’s independently derived CNS-LS based estimates of the PBA prevalence/market from the Bradley survey.

If we wish to glean information as to the prevalence of PBA burden indicated by the NSA and BIAA surveys we must first correct for the bias in these data sets. We attempt this un-biasing using the expected prevalence rates of PBA from the “unbiased” Bradley survey (Table I). To this end, the NSA and BIAA survey rates of reported PBA burden were reduced in proportion to the degree that PBA appeared to be over-weighted in comparison to the Bradley survey (Table I). In addition, the rates of more severe forms of PBA were further reduced to meet ratios of mild to more severe cases.

Table I: Survey PBA Prevalence Findings

Total Mild Mod/Sev ModSev/Mild Ratio
"Unbiased" Stroke 26.0% 21.7% 4.3% 0.20
NSA 53.0% 38.7% 14.3%
Ratio 0.49 0.56 0.30
"Unbiased" TBI 48.1% 32.6% 15.5% 0.48
BIAA 80.0% 39.0% 41.0%
Ratio 0.60 0.84 0.38
"Unbiased" MS 46.5% 36.7% 9.8% 0.27
MSAA 48.0% 37.9% 10.1% 0.27
Ratio 0.97 .97 .97

Bradley survey data used as the “unbiased” comparison

The specific calculations for these bias corrections are given as an appendix at the end of this article. These adjustments suggest that prevalence findings in the NSA stroke survey needed to be reduced by factors of 0.301 and 0.561 for moderate-severe and mild PBA respectively. Likewise, prevalence findings in the BIAA TBI survey needed to be reduced by factors of 0.476 and 0.686 for moderate-severe and mild PBA respectively in order for the prevalence findings to match those of the Bradley survey. Then, we applied these same correction factors to self-reported degree of burden caused by emotional outbursts in these surveys (see appendix). Table II shows these results along with corresponding numbers of patients.

Table II: Estimates for Other Major Diagnoses

[Click to enlarge]

* Est. for ALS/PD/AD/FTD

Within Table II we show estimates of burden in ALS, Parkinson’s, Alzheimer’s, and frontotemporal dementia patients. The numbers of MS, stroke, and TBI patients reporting significant burden (“somewhat” or greater) were 2-3x more than the number thought to have moderate-severe PBA based on CNS-LS scores. Estimates of burden for ALS, Parkinson’s, Alzheimer’s, and frontotemporal dementia were based on this ratio though for conservative estimation purposes, a multiple of 1.9x was applied to ALS and Parkinson’s disease. For Alzheimer’s disease only a 1.0x multiple was used due to the greater uncertainty for that diagnosis and greater likelihood of lower burden if a patient is not aware of their episodes or does not remember them. Stratification of the degree of burden was made similar to that seen for MS, stroke, and TBI.

These findings suggest that there are 843,030 patients who consider their emotional outbursts very or extremely burdensome. Additionally, there are a total of 2,622,930 patients affected by emotional outbursts to the threshold of “somewhat burdensome or greater”. We consider this the best indicator of the potential “market” for Nuedexta. In addition, there are large groups not considered here that described their emotional outbursts as “not very burdensome”. Thus, there may be additional patients with some mild degree of PBA burden that are excluded from these totals.

These estimates of the Nuedexta market fall within currently estimated ranges and the use of burden to determine this market best reflects the impetus of both patients and physicians for treatment. In addition, a focus on the clinical impact of PBA removes lingering whisperings that somehow these patients do not need or deserve to be treated.

With a better understanding of the total potential market for Nuedexta we can now assess the many factors that will affect actual uptake. From there we can project Nuedexta sales, which is the goal of our next article in this series.

Conclusion: The potential market for PBA is large. While our estimates are below those based on CNS-LS scores, we see these estimates as more focused on those patients likely to be treated rather than as any diminution of potential for Nuedexta.

click to enlarge image


Disclosure: This article was co-authored by Seeking Alpha member OFP (OnlyFactsPlease). Both authors are long AVNR.

Continue to Part II >>