A Case of Portfolio Metamorphosis A Case for Framing Stuart Harris, PhD Life Sciences Lead, Decision Frameworks DAAG – Indianapolis – May 2009

A Case of Portfolio MetamorphosisA Case for Framing

Stuart Harris, PhDLife Sciences Lead, Decision Frameworks

DAAG – Indianapolis – May 2009

2 | DAAG 20009 © 2009 Decision Frameworks Inc. All rights reserved. Do not copy.

Introduction

Sometimes things are not as they appear. This is a case of a decision analysis that began as a typical portfolio analysis. But over the course of framing the problem, the project went through two or three metamorphoses. …at least!


In the beginning…


Initial Primary Question

The Cardiovascular & Neurosciences Division of Cellpharm Biotech (CPB) has 8 compounds in early stage development. Given a limited budget, how do we prioritize these projects? Some are preclinical and some are first time in humans (FTIH).


Portfolio Approach

1. Create a model at the10,000 foot level for all projects.

2. Confirm with each project team and tweak for project peculiarities.

3. Collect data.4. Assess eROI (productivity for selection

based on limited budget) and eNPV (value for prioritizing)


Portfolio Approach

33%Exceptional

0 0

50% 33%Phase III Success Acceptable

0 0 0 0

50% 33%Phase II Success Fail

0 0 0 0

50% 50%Phase I Success Fail

0 0 0 0

50% 50%Preclin Success Fail

0 0 0 0

50%Go Fail

0 0 0 0

50%Fail

0 0 0

No Go

0 0

Productivity Curve

Sevoraxis hepatic encephalopathy

Travoline Suspension - GERD

Savalex for Crohn's Disease

SAV-980 for Restless Leg Syndrome

Lapotine Tablets 1100 mg - CMV

Savolexis for Stroke

Sudasalophine 1200mg Tab - Crone's

RealTab Tablets 1100 mg for UCForzipan for Ileus

$0

$1,000,000

$2,000,000

$3,000,000

$4,000,000

$5,000,000

$6,000,000

$7,000,000

$8,000,000

$0

$20,0

00

$40,0

00

$60,0

00

$80,0

00

$100

,000

$120

,000

$140

,000

$160

,000

Cumulative Cost

Cu

mu

lati

ve

NP

V

Risk vs Return (priorities)

Forzipan for Ileus


RealTab Tablets 1100 mg for UC


Sudasalophine

Sevoraxis

Lapotine Tablets

Travoline Suspension

SAV-980$0

$500,000

$1,000,000

$1,500,000

$2,000,000

$2,500,000

$3,000,000

$3,500,000

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Probability of Success

Un

ad

juste

d N

PV

1. Model

2. Data3. Select

4. Prioritize


Portfolio Approach

33%Exceptional

0 0

50% 33%Phase III Success Acceptable

0 0 0 0

50% 33%Phase II Success Fail

0 0 0 0

50% 50%Phase I Success Fail

0 0 0 0

50% 50%Preclin Success Fail

0 0 0 0

50%Go Fail

0 0 0 0

50%Fail

0 0 0

No Go

0 0

Productivity Curve

Sevoraxis hepatic encephalopathy

Travoline Suspension - GERD


SAV-980 for Restless Leg Syndrome

Lapotine Tablets 1100 mg - CMV


Sudasalophine 1200mg Tab - Crone's

RealTab Tablets 1100 mg for UCForzipan for Ileus

$0

$1,000,000

$2,000,000

$3,000,000

$4,000,000

$5,000,000

$6,000,000

$7,000,000

$8,000,000

$0

$20,0

00

$40,0

00

$60,0

00

$80,0

00

$100

,000

$120

,000

$140

,000

$160

,000

Cumulative Cost

Cu

mu

lati

ve

NP

V

Risk vs Return (priorities)

Forzipan for Ileus


RealTab Tablets 1100 mg for UC


Sudasalophine

Sevoraxis

Lapotine Tablets

Travoline Suspension

SAV-980$0

$500,000

$1,000,000

$1,500,000

$2,000,000

$2,500,000

$3,000,000

$3,500,000

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Probability of Success

Un

ad

juste

d N

PV

1. Model

2. Data3. Select

4. Prioritize


The First Transformation“Toto, I’ve got a feeling we’re not in Kansas anymore!”


Reformulated Primary Question

The Cardiovascular & Neurosciences Division of Cellpharm Biotech (CPB) has 8 compounds in early stage development. They are all pursuing the same indication − treatment of stroke. Which compound should be lead and which backups? How many backups. Some are preclinical and some are first time in humans (FTIH).


Background on Stroke Treatment

Current treatment of stroke is similar to heart attack treatment. A small cluster of fatty material breaks from a blood vessel and lodges in a small blood vessel surrounding the heart or the brain. Treatment is to dissolve the clot.

The danger is that the smaller clot can re-deposit itself and the problem remains, but somewhere else. …and possibly creates a worse problem.

A new class of treatment would dissolve the clot from the inside out and avoid the problem of re-depositing.


Backup Approach

1. The backup receives its value from the lead failing.

2. Sometimes the backup is better than the lead. A decision needs to be made of whether to replace the lead. The trade-off is quality vs time. This occurs ONLY IF there is competition.

3. Another strategy is to delay the backup in order for it to become a follow-on, a next generation therapeutic.


Backup Approach0.5

Lead Success

Lead ONLY 0 0

0 0 0.5Failure

0 0

0.5Lead Success

0 0Lead AND Backup

0.50 0 Backup Success

0 0.5Failure 0 0

0 0 0.5Failure

0 0

0.5Backup Success

Backup ONLY 0 0

0 0 0.5Failure

0 0

1. Arrange Chronologically

2. assess lead & First Backup

3. Select Lead Only, backup only, or lead & backup

4. Add next backup & return to step 2

5. When done, rank by value metric


Backup Approach

We now have a prioritized list of leads and backups. But some backups have lost value, not due to quality, but due to time to market.

Consider a follow-on strategy. This introduces cannibalization and a timed launch to ensure competitive advantage.

It is easier to switch an existing customer to an improved treatment than to a new brand.


Backup Approach

We now have a prioritized list of leads and backups. But some backups have lost value, not due to quality, but due to time to market.

Consider a follow-on strategy. This introduces cannibalization and a timed launch to ensure competitive advantage.

It is easier to switch an existing customer to an improved treatment than to a new brand.


The Final Metamorphosis


Primary Question

The Cardiovascular & Neurosciences Division of Cellpharm Biotech (CPB) has 8 compounds in early stage development. What is the best CVN Strategy to maximize eNPV and ensure future growth (cash flow)?


Disease Area Strategy Approach

Framing1. Continue collecting and sorting the issues.

2. Develop Strategic Decisions (strategy table)

3. Create Decision-Risk Timeline

4. Construct Project Influence Diagram

5. Construct Portfolio Influence Diagram to be fed by Project Influence Diagram

Analysis1. Calculate eNPV for each strategy

2. Show cash flow from now until patent expiry


Framing – Questions & Issues

Issue List© Copyright 2002-2008 Decision Frameworks Inc. All Rights Reserved.

State the decision problem

List the key questions the evaluation needs to answerWhat's the hightest valued compound for each of the 3 classes?How do our compounds compare to the competition in each class?How likely will each successive generation cannabolize the preceeding generation?

List and categorize the issues

Decision Type Issue Type IssueFocus Decision Should we develop the best of each of the 3 types of compounds or just select one, possibly two types?Focus Decision Should we advance more than one of any type?Given Decision A stroke treatment will be a priority and at least one solution and, as many as three, will be pursued.Given Decision A reformulation should be launched ASAPTactic Decision Which if any of the clot dissovers should be given in combination with the neural protectants.

Fact Current stroke unmet needFact The current CVN portfolio has 5 in-house and 3 available for in-licensing. Fact The current treatment will be generic in 2013. Patents range from 2018 to 2026.Fact The risk and time to market increase in order of improved formulation, 2nd generation dissolution, neuroprotectants.Fact Royalty paymentsFact There are 3 types of compounds – improved formulation, improved clot dissolution, dual action neural protectant.Fact 2 improved formulations will be to market quickly through bridging studies.Fact 3 are "clot dissolvers" and are 2 years behind the improved formulations.Fact 2 are neural protectants.Uncertainty FormulationsUncertainty Safety and Efficacy of 2nd Generation TreatmentsUncertainty Competition in each of the 3 classes of compoundsUncertainty Value of 1st and 2nd classes when current treatment goes off patentUncertainty Will there be cannibalization and is it worth it to protect market shareUncertainty How do we compare to the competition?Value/Objective Value measures will be eNPV, eROI, and Cash Flow


Issue List© Copyright 2002-2008 Decision Frameworks Inc. All Rights Reserved.

State the decision problem

List the key questions the evaluation needs to answerWhat's the hightest valued compound for each of the 3 classes?How do our compounds compare to the competition in each class?How likely will each successive generation cannabolize the preceeding generation?

List and categorize the issues

Decision Type Issue Type IssueFocus Decision Should we develop the best of each of the 3 types of compounds or just select one, possibly two types?Focus Decision Should we advance more than one of any type?Given Decision A stroke treatment will be a priority and at least one solution and, as many as three, will be pursued.Given Decision A reformulation should be launched ASAPTactic Decision Which if any of the clot dissovers should be given in combination with the neural protectants.

Fact Current stroke unmet needFact The current CVN portfolio has 5 in-house and 3 available for in-licensing. Fact The current treatment will be generic in 2013. Patents range from 2018 to 2026.Fact The risk and time to market increase in order of improved formulation, 2nd generation dissolution, neuroprotectants.Fact Royalty paymentsFact There are 3 types of compounds – improved formulation, improved clot dissolution, dual action neural protectant.Fact 2 improved formulations will be to market quickly through bridging studies.Fact 3 are "clot dissolvers" and are 2 years behind the improved formulations.Fact 2 are neural protectants.Uncertainty FormulationsUncertainty Safety and Efficacy of 2nd Generation TreatmentsUncertainty Competition in each of the 3 classes of compoundsUncertainty Value of 1st and 2nd classes when current treatment goes off patentUncertainty Will there be cannibalization and is it worth it to protect market shareUncertainty How do we compare to the competition?Value/Objective Value measures will be eNPV, eROI, and Cash Flow



Decision Problem Focus

Decision Hierarchy TRUE© Copyright 2002-2008 Decision Frameworks Inc. All Rights Reserved.

TRUE

Values/ ObjectivesValue measures will be eNPV, eROI, and Cash Flow

Made Decisions/ GivensA stroke treatment will be a priority and at least one solution and, as many as three, will be pursued.A reformulation should be launched ASAP

Focus DecisionsShould we develop the best of each of the 3 types of compounds or just select one, possibly two types?Should we advance more than one of any type?

Tactical DecisionsWhich if any of the clot dissovers should be given in combination with the neural protectants.

Show Decision Triangle


TRUE







Decision Problem Focus


FALSE







FALSE







Issues

Decisions Should we develop the best of each of the 3 types of compounds –

improved formulation, improved clot dissolution, dual action neural protectant – or just select one, possibly two types?

Should we advance more than one of any type?

Facts Current stroke unmet need The current CVN portfolio has 5 in-house and 3 available for in-licensing. The current treatment will be generic in 2013. Patents range from 2018

to 2026. The risk and time to market increase in order of improved formulation,

2nd generation dissolution, neuroprotectants. Royalty payments

Uncertainties Formulations Safety and Efficacy of 2nd Generation Treatments Competition in each of the 3 classes of compounds Value of 1st and 2nd classes when current treatment goes off patent Will there be cannibalization and is it worth it to protect market share


Influence Diagram - Project

Competitor Entries

Formulation, Dose, & Regimen

Compound SelectionGo/No-Go

Market Size& Share,

less cannibalization

PricingStrategy

DevelopmntCosts

$$$eNPV

Net Sales Revenues

Ph3 efficacy & safety results Regulatory

success

COGS

Product Profile

LaunchTiming

Accelerated Review

Clinical Trial Timing

Timing

Costs


First Entry Launch Timing

Compound SelectionGo/No-Go

Market Size& Share

cannibalization

First Entry Development

Costs

$$$eNPV / eROI

First EntryNet SalesRevenues

First Entry Product Profile

Total Development

Costs

Second Entry Launch Timing

Second EntryNet SalesRevenues

Second Entry Development

Costs

Second Entry Product Profile

Influence Diagram – Multiple Launches


Cash Flow

2010 2015 2020 2025 2030

0.0

0.5

1.0

1.5

TIME

Reformulation

Clot DissolutionNeuroprotectant


Conclusion / Observations / Lessons

Not always NPV Regarding the actual portfolio problem, one of the three strategies dominated. Although eNPV was similar for two, a major uncertainty had the potential to be mitigated.

Don’t Presume This was neither a portfolio problem nor a back-up issue. Without proper framing I may have tried the wrong approach. I believe that I was doing what I try to get the project teams avoid – i.e., making assumptions about the problem and its solution.

Representativeness Bias We are all subject to heuristics and it is well to remember the value of not presuming. The representative bias (aka guilt by association) had me approaching the problem incorrectly. Too often we immediate create trees or influence diagrams because we KNOW what the problem is all about.

Documents

A Case of Portfolio Metamorphosis A Case for Framing Stuart Harris, PhD Life Sciences Lead, Decision Frameworks DAAG – Indianapolis – May 2009