A World of Biomanufacturing: Shortages or Global Glut?

A World of Biomanufacturing: Shortages or Global Glut? Howard L. Levine, Ph.D. BioProcess Technology Consultants, Inc. BioProcess International Conference Vienna, Austria May 19-20, 2010

Steady growth in the number of biopharmaceutical products No. Products 120 100 80 60 40 EU Non-MAbs EU MAbs US Non-MAbs US MAbs 20 0 Pre- 1995 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Number of approved products in the US has steadily increased from approximately 20 in 1995 to 114 today Graph includes products approved but later withdrawn from market and produced via both mammalian cell culture and microbial fermentation Data includes 89 products approved by EMEA Data does not include all products approved by EU member states before centralized procedures were adopted Year

Coupled with continued growth of product sales 2009 US sales for 130 biologic products exceeded $95 billion 11% of total pharmaceutical market 16% annual growth rate In 2009, 27 biopharmaceutical 18 products with worldwide sales 12 in excess of $1 Billion 6 1 fewer blockbuster product than in 2008-10 manufactured by microbial fermentation Mammalian Recombinant Products Microbial Recombinant Products 17 manufactured by mammalian cell culture 9 antibody based products, including full length antibodies, antibody fragments, and Fc fusion proteins A n n ual Sal es ($ B ) 42 36 30 24 2003 2004 2005 2006 2007 2008 2009 Mammalian MAb Products Microbial MAb Products

Leads to growing demand for biopharmaceutical manufacturing Cell Culture Recombinant Proteins Microbial Fermentation 12,976 111 Cell Culture Microbial Fermentation 7,599 Approximately 13 metric tons of product manufactured by microbial fermentation 63% of total annual protein production Insulin accounts for >90% of this production 6 kg of monoclonal antibody products Approximately 7.5 metric tons produced in mammalian cell culture 7,599 kg monoclonal antibody products 111 kg recombinant proteins Monoclonal Antibody Products 6

With demand for biopharmaceutical products continuing kg Required (kg/yr) 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 - Pipeline Clinical Commercial 2009 2010 2011 2012 2013 2014 2015 Most probable demand for all products currently on the market or in development Green arrow denotes anticipated demand if all five top drivers currently in development succeed Red arrow denotes anticipated demand if all five top drivers currently in development fail

Monoclonal antibody products represent the fastest growing segment of the pharmaceutical industry 140 120 100 Mammalian Microbial Other (Plant, Insect, etc.) No. Products 80 60 40 20 0 Market BLA/NDA Phase 3 Phase 2 Phase 1** Approximately 65% of all biopharmaceutical products in development are monoclonal antibody related products Approximately 85% of mammalian cell culture products and 25% of microbial fermentation products are monoclonal antibody related products

Estimated demand for the top six monoclonal antibody products exceeds 80% of total demand for all mammalian cell culture products Bulk Requirem ents 2009 (Kg) 1,600 1,200 800 400 1,331 1,255 1,229 1,091 945 431 1,427 0 Remicade Rituxan Avastin Enbrel Herceptin Erbitux All Other Products (68)

Demand for mammalian cell culture continues Currently 57% of all commercial biopharmaceutical products and approximately 65% of all products in development produced in mammalian cell culture Does the industry have enough capacity to meet this demand? What will manufacturing operations look like in 2015?

In the late 1990 s and early 2000 s, companies struggled to obtain needed biopharmaceutical manufacturing capacity and extreme capacity shortages were forecasted We estimate that capacity will more than a triple by 2006. Unfortunately, we estimate that demand for capacity will quadruple over this time period - JP Morgan report (2002)

Despite the announcement of investments in significant amounts of additional capacity, shortages were still projected by some Current projections point to a serious shortfall developing within three years in manufacturing capacity, especially in mammalian cell culture... - DM&D Report (2003)

However, projections showed low probability of shortages Ref: Levine HL. The capacity crunch reality or myth? Presented at IBC Production Economics and Manufacturing Strategies of Biologicals; 2003 Jun 16 17; Brussels, Belgium 4,000 Reactor Volume ('000 L) 3,500 3,000 2,500 2,000 1,500 1,000 500 0 2002 2003 2004 2005 2006 2007 2010 Year Forecast Industry-wide Capacity Most Probable Case Capacity utilization increases to almost 100% by 2010 Top 5 Potential Volume Drivers Fail Capacity utilization drops to <55% through 2007 Top 5 Potential Volume Drivers Succeed Approximately 10% capacity shortage by 2007

By 2009 capacity shortages still have not materialized... Increases in product titers and operational excellence initiatives have improved overall productivity BioProcess Technology Consultants report, Published December 2008 Sufficient capacity worldwide to meet current annual production needs with a high probability of sufficient capacity for the foreseeable future

Nevertheless, some still predicting capacity constrains Nearly one-half of survey respondents predicted that their manufacturing facilities would experience capacity constraints by 2014 - BioPlan Associates report (2010)

So where do we really stand? An optimist will tell you the glass is half full; the pessimist, half empty; and the engineer will tell you the glass is twice the size it needs to be anonymous

Abundant cell culture capacity worldwide Installed Capacity (,000L) 2,000 1,750 1,500 1,250 1,000 750 500 250 Asia Europe North America 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Data from 101 companies in 28 countries worldwide indicates an increase in capacity from approximately 2.5 million liters in 2009 to approximately 4 million liters in 2015 Capacity growth will be greater in Asia than in EU or North America Year Recent economic downturn resulted in slowing of projected capacity expansion plans Ref: E. Reynolds, PhD Dissertation, MIT (2010)

However 10 companies control 80% of worldwide capacity decreasing slightly to 73% by 2015 30% By 2015, Genzyme replaced by Celltrion in top 10 25% 20% 15% 10% 5% 0% A B C D E F G H I J K L 2009 2015 A. Roche B. Amgen C. Pfizer D. Boehringer Ingelheim E. J&J/Centocor F. Lilly G. Lonza H. Novartis I. Biogen Idec J. Genzyme K. Celltrion L. All Others

Balance of worldwide cell culture capacity between product companies and CMOs favors product companies Est. Installed Reactor Volume (KL) 4,200 3,600 Product Co. Both 3,000 CMO 2,400 1,800 1,200 600 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year Product companies currently control approximately 70% of total capacity, decreasing slightly by 2015

Overall balance of cell culture capacity supply and demand Installed Capacity (,000L) 4,200 3,600 3,000 2,400 1,800 1,200 600 Available Supply Pipeline Clinical Commercial Most Probable Demand Drivers Pass Drivers Fail 0 2009 2010 2011 2012 2013 2014 2015 Total facility utilization was 41% in 2009 increasing to a projected most probable utilization of 68% by 2015 Variable capacity utilization and uneven distribution of capacity means that several large facilities currently have available capacity Impact of pipeline volume drivers on overall facility utilization If all succeed, utilization increases to 77% If all fail, utilization decreases to 61%

Impact of today s overcapacity on outsourcing Outsourcing remains an option for many companies with greater choices now available as product companies with excess capacity offer this captive capacity for contract manufacturing You can now leverage our clinical contract manufacturing capabilities for your therapeutics... - Amgen BIO conference (2010) CMOs are more flexible today in pricing and schedule and Sponsors are demanding better service and more attention than before Increased capacity utilization in the CMO market as a result of consolidation among this group and slowdown of expansion

Trends That Will Impact Future Capacity Utilization Success or failure of a limited number of high volume products currently in late stage clinical development Acquisition of volume driver product candidates by product companies having significant capacity may free up CMO capacity Fewer blockbuster drugs with greater focus on smaller markets and niche products Mergers and acquisitions, resulting in redundant facilities in larger organizations (i.e., the rich get richer) Companies are moth balling or not starting up facilities due to excess capacity

Continued innovation has increased expression levels and yields 1978 Recombinant insulin produced Commercial products have expression levels in the range 2006 of 0.2 3.0 g/l with the highest titers seen for monoclonal antibody products Industry Standard: 1-2 g/l Industry Leaders: 3-6 g/l Improved strains New technologies to improve cell line development and expression levels coupled with improved and optimized media, supplements, and bioreactor conditions have increased titers of products in development 2015-2020 Future Leaders: >10 g/l Ref: T. Charlebois, BIOMAN 2006 Conference, (2006); M. Smith, BPI Europe Conference (2005); F. Wurm, Nature Biotechnology, 22(11) (2004)

Effect of titer on demand for manufacturing capacity over time Two fold increase in industry wide productivity will reduce overall capacity requirements approximately 25% by 2013 4,000 Total Volume Requirements (,000L) 3,500 3,000 2,500 2,000 1,500 1,000 500 2013 2012 2011 2010 Year 2009 0 2008 0.5x 1x Titer Increase 1.5x 2x 2007 Higher titers and lower doses will reduce capacity demand for future commercial products

Personalized medicine initiatives will transform medicine and influence scale and demand for manufacturing capacity Targeted Therapy Rational drugs based on profiling of underlying molecular pathology Individualized Therapy Rational drugs based on comprehensive molecular profiling of individuals Personalized Care Integrated data on individual health status Transition to disease prediction and preemption Graphic adapted from: S. Burrill, IBC Development and Production (2010) Greater focus on smaller markets and niche products will result in fewer blockbuster drugs and a decrease in capacity demands for individual products

Better diagnostics will reduce waste in heathcare spending drive the development of targeted therapies Ref: S. Burrill, IBC Development and Production (2010)

Today s Biopharmaceutical Facility Most facilities today were built for low titer (<1 g/l) processes Multiple 20,000 L bioreactors each with inoculum bioreactors up to 4,000 L Current facilities struggle to match downstream capacity with bioreactor output due to large process volumes Technologies that enable higher bioreactor titers will exaggerate the DSP bottleneck Photos courtesy of Lonza Biologics

Producing 10 Metric Tons of Monoclonal Antibody But do we really need this much? Ref: B. Kelley, ACS Conference, (2006)

How much product do we need? Demand for all existing commercial monoclonal antibody products will approximately double from the current 7.7 metric tons to approximately 15.9 metric tons by 2015 Current annual product requirements for each of the top five monoclonal antibody products ranges from 941 kg to 1,331 kg Demand for products currently in development will increase the future demand for cell culture manufacturing capacity even further The anticipated demand for any new monoclonal antibody products approved between now and 2015 is expected to be less than 5 metric tons per year

The biopharmaceutical manufacturing facility of the future will Incorporate high titer (>10 g/l) processes Use disposable technologies to reduce capital investment and operating costs Anticipated reduction of over 50% (J. Roebers, BPI [2009]) Require greater DSP space and capabilities to better handle the high titer bioreactor output Ratio of bioreactor space to DSP space will decrease Use smaller bioreactors to produce similar quantities to today s larger bioreactors Reduced capital requirements may enable smaller companies to construct their own facilities rather than outsource

Single use and disposable technology will change the look and feel of future manufacturing facilities Increased facility utilization by reducing change over time Reduced fixed piping Reducing cleaning and validation costs in multiproduct operations Increased operational flexibility by minimizing or eliminating multi use equipment Improved process portability and ability to manage and implement process changes Photo courtesy of Acceleron Pharma

Approaches to debottlenecking downstream processing Use of negative chromatography as an alternative to a dedicated capture step High product titers enables capture of impurities while product flows through column Process bioreactor harvest in multiple batches Clarify and freeze bioreactor harvest for purification in smaller batches Use of disruptive technologies such as precipitation, expanded bed, and simulated moving bed chromatography Ref: A. DePlama GEN (2010 May 1)

Continuous purification with simulated moving bed chromatography is extremely scaleable and leads to substantial reduction in manufacturing costs Fully disposable processing train can be paired with disposable bioreactors resulting in smaller process footprint Higher throughput per square foot of manufacturing space enables high titer processes to fit in existing facilities v Wash Elution Wash Photos courtesy of BioFlash

Continuous disposable downstream processing with BioSMB Process uses same fundamental phenomena as batch processes No change in media and buffer composition Same steps for binding, washing, and elution as in the corresponding batch process System volume depends on mass transfer kinetics, not binding capacity or titer Batch BioSMB Titer 3.5 gm/l Batch size 2000 L Productivity [g/l/day] 360 2630 Processing time 05:10 08:00 12:00 24:00 Protein A media [L] 88 8.0 5.2 2.6 Buffer [L] 4600 3100 3350 Number of columns 1 8 8 12 Cycles per batch 2 19 29 58 Protein A Media Costs $ 880k $ 80k $ 52k $ 26k Data courtesy of Tarpon Biosystems

BioSMB makes size exclusion chromatography fun again 1.20 1.00 UV 280 in product [AU] 0.80 0.60 0.40 0.20 0.00 Cycle 3 Cycle 4 Cycle 5 0 10 20 30 40 50 60 70 80 90 100 110 120 Time [min] - Each column switch = 10 min Continuous gel filtration chromatography of a vaccine VLP using a 12 column process set up Photos and data courtesy of Tarpon Biosystems

Future biomanufacturing facilities will need to balance multiple factors to meet performance and productivity goals Capacity Variety of product types with differing manufacturing processes Flexibility Scalability Performance/ Productivity Quality/ Regulatory Speed Antibodies, antibody fragments, fusion protein, recombinant proteins Cell culture vs microbial fermentation Low titer vs high titer Cost No one facility will fit all products or processes

Diversity of manufacturing facilities in the future Low volume, low titer products (<20 Kg/year) Little to no reason for process changes for existing products New low titer products (primarily non MAb products) will emerge Large manufacturing plants designed for multiproduct operation High volume, high titer products (>100 Kg/year) Antibodies will be drivers for platforms and very low COGS Increased use of disposable technology will reduce capital costs Dedicated plants with appropriately sized (i.e., smaller) bioreactors Adapted from W. Berthold, BPI Asia Conference (2008)

The biomanufacturing facility of the future Plant has 6 x 2,000 L bioreactors (possibly disposable reactors) 12 day fed batch CHO culture 2,000 volume, 15 g/l = 30 kg in harvest 80% purification yield = 24 kg per batch Harvest every 2 days 167 harvests (334 days) = 4 tons/year 1 Purification train serving single bioreactor Estimated facility cost <$100M Estimated COGS $70 per gram Photos courtesy of Xcellerex

Summary Sufficient capacity worldwide to meet annual production needs for the foreseeable future Uneven distribution of capacity may present difficulties to some companies trying to access capacity

Summary Fewer blockbuster drugs and greater focus on smaller, niche products with smaller commercial demands implies: Less difference in scale between pilot and commercial facilities Use of multipurpose plants with potential for continuous production

Summary Product and process innovations resulting in higher yields per batch and lower demand for bioreactor capacity implies: Investments in manufacturing facilities will continue to slow Increased use of disposables rather than steel in facilities Disposable/single use technologies possible for some commercial supply

Special thanks to Dawn Ecker BPTC database manager Data collection and analysis Companies providing data and pictures Acceleron Pharma BioFlash (Repligen) Lonza Biologics Tarpon Biosystems Xcellerex

We re moving As of June 15, 2010: BioProcess Technology Consultants, Inc. 12 Gill Street, Suite 5450 Woburn, MA 01801 1728 Email and web remain the same: hlevine@bioprocessconsultants.com www.bioprocessconsultants.com

THANK YOU!