continuous manufacturing for biologics...dp applications mainly for solid oral dosage forms....

Continuous Manufacturing for BiologicsCMC STRATEGY FORUM NORTH AMERICA 2019

MONDAY, JANUARY 28 2019

IntroductionICH Q13 Working Group started last year to work on guidance for Continuous Manufacturing

◦ Concept paper available

Session 1INDUSTRY PERSPECTIVE

ANALYTICAL TECHNOLOGIES

Industry Experience with Small MoleculesThomas García, Pfizer Inc.


What is continuous manufacturing (CM)?◦ Material continuously fed through process

◦ Material is transformed

◦ Output is continuously removed

◦ At least two unit ops in a row

◦ It is NOT a batch process

Applications for small molecules◦ API to DP

◦ Hybrid approach: batch + continuous unit ops

Example processes from Vertex, Eli Lilly, Janssen, Pfizer◦ Janssen - Conversion from batch to CM


API (DS) applications

DP applications mainly for solid oral dosage forms. Equipment is readily available for these applications: dispensing, blending, granulation, drying/tableting, film coating

Example Pfizer Equipment (PCMM) portable-continuous-modular-?◦ Capable of granulation, mixing and direct compression. Incorporates P.A.T. for

potency, content uniformity, granule sizing, moisture content using NIR and FBRM.

Industry Experience with Small MoleculeThomas García, Pfizer Inc.

Regulatory Landscape for small molecules◦ 5 products approved

◦ Meet early and frequently with the FDA

◦ ICH Q13 is pivotal to harmonization of regulatory expectations for API/DS/DP small/large molecules◦ Transitioning from legacy processes to continuous processes

◦ Challenges with guideline◦ Cannot be too detailed or high-level due to diverse processes and molecules covered under

the guideline


◦Talking points for implementation◦ Refine definition of batch size – may be based on number of units or

throughput, or consumption of input material; incorporate a validated range

◦ Stability – define number and size of batch required for implementation and post-approval changes (PACs)

◦ Maintaining dossiers for NPIs and product lifecycle

◦ Can it be part of Pharmaceutical Quality Systems instead?

◦ Are PACs necessary? How often should sponsors notify the regulatory agency?

Industry Experiences with Small MoleculeThomas García, Pfizer Inc.

Talking points for implementation (continued)◦ Validation – depart from 3X tradition and apply 3-stage approach using CPV;

must demonstrate control at beginning, end and over time

◦ Product Availability –◦ Challenges with supplies to markets that will not accept this process.

◦ May need to develop batch and continuous processes in parallel

◦ Tech Transfer – How to demonstrate “like for like” for non-identical equipment?

◦ Contingency plans when components are down

◦ Use of soft sensors and data management

◦ Leverage existing guidance: ICH Q8-Q11


Similarities between biologic and small molecule processes◦ Use of reactors

◦ Use of chromatography steps for purification

◦ Use of filtration processes

◦ Use of hybrid processes

Major differences◦ Prior knowledge is richer for small molecules

◦ Process Monitoring and analytical tools

Continuous Biomanufacturing: Development, Hybrid Implementation and OpportunitiesErik Fouts, BioMarin Pharmaceutical Inc.


Why implement CM?◦ Trend in manufacturing

◦ Improves processes, reduces costs

◦ Limited in biologics – hybrid systems in place: continuous cell culture with batch purification

◦ Most prevalent for biologics at pilot scale

The BioMarin Experience◦ 3 hybrid CM processes since 2003 using perfusion bioreactors where media is

continuously supplied to the bioreactor and incorporated clarification and harvest


Points to Consider for Perfusion Cultures◦ Maintains stability of proteins and host

◦ Increases productivity

◦ Smaller foot-print and costs

What have they learned?◦ Not a steady-state process requires constant monitoring and flexible

adjustment processes

◦ Sterility is critical

◦ Significant challenges with fluid handling capacity


Scenarios for extending CM processes◦ Continuous chrom steps using multiple single-use columns

◦ Semi-Continuous processes from perfusion step to secondary sterile containment of harvest material for input into chrom steps

◦ On-line/in-line analytical capabilities

Challenges◦ P.A.T is behind technology advancements compared to production

technologies

◦ Molecular complexity increases complexity of design space

◦ Automation produces significant increase on data outputs and require complex systems for monitoring (e.g. multivariate, predicting monitoring)

Control of Continuous Bioprocesses: Integration of P.A.T. with Time Distribution Models to Ensure Consistent Product QualityMark Brower, Merck & Co., Inc.


Design of a flexible process within a laboratory space at pilot scale

Perfusion cell culture◦ Viable cell-density reaches a steady state along with CQA indicators

Designing the physical space◦ Single use components and warehousing

◦ Maintaining sterile boundary

◦ Integrated P.A.T

◦ Provides flexibility to improve or expand the process/facility


Merck’s PROLab – Pilot Plant Design◦ Sincle-used closed processing

◦ DeltaV Control

◦ Integrated PAT

◦ Speed compensation for flexible batch size

◦ Flexible for removing, adding process components

◦ Non-monitored process steps

◦ May be operated by reduced number of personnel

Proof of concept◦ Controlled perfusion feed rate allows to control and maintain product quality attributes


P.A.T. Toolbox for Biologics◦ Eliminates need for sampling, chain of custody, QC testing and reporting in

electronic systems

◦ Integrates raw materials spectroscopy tools - Raman

◦ Raman for nutrients, metabolites and VCD

◦ Online LC – aggregates, impurities, concentration

◦ ELISA and sensors – potency, biomass

◦ Multi-attribute Methods – identity, glycans, process impurities, product impurities

◦ Product Attribute Controls - MVDA

◦ Rapid Micro methods – sterility, bioburden, air monitoring


Residence Time Distribution (RTD) Models◦ An old technique for a modern application

◦ Used to model each unit operation or a single unit operation in tandem

◦ Allows predictive models for the entire process

◦ Requires increased testing frequencies

◦ Allows adjustments in the process to exclude output material with deviating quality attribute from further processing

Technical challenges◦ Need more advances with analytical tools

◦ PAT

◦ MAMs

◦ Adventitious Virus testing

Session 1: Panel Discussion- Q&AHighlight of small-molecules vs biologics similarities/differences for CM

◦ Similarities described in presentation; stability assessments, validation

◦ Differences◦ Raw materials are powders vs. slurries/liquid require different approaches for ensuring fluid process, residual removal

Does it get easier after the first filing? ◦ Yes. Prior knowledge was leveraged, the process was easier for filing.

Given limited experience with CM, what are the expectations for content for ICH Q13? The expectation is that Q13 covers small molecules and biologics alike.

◦ Industry partnerships have been working on a platform approach for development and submissions

◦ It may be too soon for biologics, however implementation may take 3+ years. In that period, biologics have time to come up with harmonized strategies

◦ Q13 concept paper bases expectations on small-molecule model. An addendum is expected for biologics for further guidance.

◦ The guidance is meant to be forward-looking

Session 1: Panel Discussion- Q&AHow do you tailor analytical method development and validation for CM for inclusion in a new marketing application? For post-marketing approvals?

◦ For biologics, the intent should be to submit application using CM platform. Not an option to submit both batch process and CM for manufacturing

◦ For small molecules, it is more feasible to incorporate both processes. It prevents issues with markets that will not approve a CM process

◦ Perfusion is usually an approach to get faster to the clinic. The intention is not to go back to batch processing.

What was the process to decide whether a process would be developed as CM or batch?◦ Risk assessments involved considerations of suitability of the formulation for CM

Hybrid processes seem easier to introduce as scope of change is narrower, and allows use of existing facilities. What is the argument for introducing P.A.T for a highly characterized process with already minimized analytics?

◦ It is necessary to demonstrate cycle consistency and process understanding to determine the extent of P.A.T.

Session 1: Panel Discussion- Q&AWhat is the opinion for comparability exercise needed to switch from batch to CM?

◦ It is important to meet often with the regulators.

◦ The expectations for comparability do not change. You have to define design space and critical quality attributes.

◦ It is necessary to demonstrate comparability of commercial material to clinical material

How could you deal with cell-age effect on CM processes (eg. shift in quality attributes)? How do you ensure consistency throughout the process?

◦ The concept of infinite run times for biologics is not viable. However, a finite process may be characterized to the point where cell-age shifts quality attributes

◦ Another point to consider are the limitations of cell-doubling. Defining a batch would be based on steady quality attributes.

◦ It may be necessary to re-think control strategies. What and where should we quantitate the attributes?

Session 1: Panel Discussion- Q&ARegarding future of CM, where do you see minimization of product testing and more reliance of in-process testing? Do you anticipate Q1 will address those questions?

◦ Use of modeling is critical for real-time release. Tests may be streamlined to determine which testing is necessary.◦ An argument can be made on the submission to use minimized release testing and rely on in-process testing

Has industry analyzed cost-benefits from semi-CM to fully-CM state? Are steps such as VF worth the investment needed to support CM?

◦ Typical approach is to apply “baby steps” and progressive scale-up. As equipment evolves (smaller size, increased capability, reduced cost), the process becomes simpler. “Dive head first”.

◦ Typically the cost of raw materials remains static. Thus, cost-reduction lies on smaller facility needs, reduced personnel. However, the personnel must be highly skilled.

What are the arguments to push for CM of mAbs? Current process highest cost is ProA resin, all else is fairly inexpensive

◦ Reduction of bioreactor size◦ Reduction of chromatography resins◦ Need to consider demand and use of the mAb therapeutic

◦ Current process for high demand

◦ CM for low demand products

Session 1: Panel Discussion- Q&AFor approved small molecule CM processes, was there a reduction of in-process sampling? What about biologics?

◦ P.A.T. is used for monitoring only. Release specification remained similar to batched process. However, P.A.T. was used to scrap deviating material at blending step.

◦ It is possible to take P.A.T. results to automate action limits. However, validation of that approach is challenging.

◦ Intent is to incorporate elements of validation in Q13 to potentially realize real-time release

Are you looking at rapid methods for AdVirus and micro detection? What should be the sampling points?

◦ These are the tests that are particularly challenging to incorporate in a real-time release strategy. They take days to provide output.

◦ Sampling pre-post filtration is currently being examined

Session 1: Panel Discussion- Q&AHow does process/equipment transfer influence the selection of technology for CM?◦ For small-molecules, Pfizer example, all modules are mobile and adaptable. They can be

installed in any suitable space. Equipment may be duplicated at a different site.

Are there drivers to push for CM by regulatory agencies?◦ This is one of the reasons the working group wants to include biologics in Q13.

◦ CM could also be a solution for shortage in drug supplies

◦ The FDA is very supportive of CM, thus partnerships from industry and FDA are highly supported to overcome technical challenge sin implementation

◦ FDA has seen positive outcomes for the feasibility of CM

Session 1: Panel Discussion- Q&AWho are the current participants in Q13 comments? Any opportunities for additional representatives to provide feedback? What is the current timeline? ◦ PMDA, ANVISA, Singapore, Taiwan, Health Canada, EMA, Kazakhstan, Industry

◦ Feedback is allowed under any of the participating pharma companies/members. There is an established procedure in place by ICH steps.

◦ Timeline is 3-5 years; step 4 scheduled for NOV2021

Are the working groups for Q12 on Established Conditions communicating with Q13 working groups to avoid conflicting messages/approaches?

Session 1: Panel Discussion- Q&ACM seems more suitable for smaller-scale processes? Do you foresee CM becoming more prevalent for larger scale processed (e.g. insulin)?◦ For small-molecules , size adjustments can be incorporated for CM. Thus far, no known

process for blockbuster drugs are identified to go to CM

◦ CM is ideal for scale-up, down and out

How are CM processes going to affect the development and characterization packages for submission?◦ About 6-7 molecules have been transitioned to ProLab. The tools for characterization have

not changed. The characterization space may be planned into the CM runs. For upstream manufacturing, multiplex experiments may be possible in the near future with new analytical development tools.

◦ Demonstrating viral clearance will be challenging. It may be necessary to put additional criteria for screening raw materials in order to validate a CM process.

Session 2AUTOMATION APPROACHES AND GAPS

EQUIPMENT AND FACIL IT IES

BUSINESS CASE STUDIES

BioPhorum Technology Road Map: CM for Downstream Processes for BiopharmaceuticalsCarl Carson, Exyte


Charter on Continuous Downstream Processing◦ In scope – definitions, business case, gap analysis, regulatory considerations◦ Out of scope – technical solutions

Working on a white paper, out by Q1 2019, focusing on◦ Technology Gaps◦ Bioburden Control◦ Multi-column chrom, continuous flow-through column separation◦ Continuous or semi-continuous flow to VI◦ Continuous VF, UF/DF, formulation◦ Continuous buffer preparations◦ On-line monitoring and instrument probes◦ Start-up◦ Automation◦ Leveraging benefits for other processes (e.g. cell therapy)


Model created for mAb molecules

Gaps were identified regarding topics such as◦ Terminology, definitions

◦ Limited knowledge of full CM or hybrid processes

◦ Adaptability of bioburden control requirements

◦ Defining novel viral inactivation validation

◦ Limited availability of at-scale equipment

◦ Maintaining sterile boundary at scale

◦ Lack of “plug and play” options

◦ Understanding business case

◦ Real-time data analysis

◦ Disposition and batch strategy


Why aren’t we there yet?◦ Aversion to risk, initial cost, technology gaps

Why should we have CM?◦ Improve PQ, reduce footprint, reduce operator error risk by automation

Integrated CM Bio Platform-Addressing Challenges in Automation and GapsMarina Hincapie, Sanofi

Integrated CM Bio Platform (ICB) - Addressing Challenges in Automation and GapsMarina Hincapie, Sanofi

mAb platform hybrid approach. Continuous process from perfusion through mAb capture steps. The rest of the process is batched

R&D process is developed at the same scale as manufacturing, reducing tech transfer risk

Approaches used◦ Chemically defined-media increases cell density and productivity

◦ A state of control is maintained for 30 days

Challenges with developing end-to-end process at-scale◦ Synchronizing flow rate for purification with perfusion rate

◦ Slow establishment of PAT

◦ Maintaining sterility

◦ High cost of implementation

Integrated CM Bio Platform- Addressing Challenges in Automation and GapsMarina Hincapie, Sanofi

Analytical support increases proportionally with molecular targets when testing off-line◦ PAT implementation is high cost for implementation, but has a quick ROI by reducing

analytical support

Comparison and contrast of PAT use in R&D vs Manufacturing◦ Overall it is more complex and invested in R&D for the technology, but requires GMP

validation and maintenance when used for manufacturing

It is key to design PAT in R&D space that is “fit for purpose” to gather process understanding. Then, proceed with defining parameters for monitoring product quality at scale◦ Using non-destructive methods (eg. Raman and FTIR spectroscopy)

◦ Raman is more suited for USP and FTIR for DSP

◦ Establish data management platforms and infrastructure

Integrated CM Bio Platform- Addressing Challenges in Automation and GapsMarina Hincapié, Sanofi

FTIR provides excellent prediction models for attributes such as titer, charge variants, aggregation compared to off-line methods

Issues with PAT◦ Great amount of data collected

◦ Data mining and identifying critical attributes requires complex models

◦ Need integration for bioburden detection

◦ Need autoclavable probes

◦ Automation challenges connecting acquisition between different hardware and software

Advances in Next-Gen CM Platform for BiologicsGanesh Vedantham, Amgen Manufacturing Limited


CM at Amgen and its role in strategic goals◦ Focus on more products, targeted to different diseases

◦ Incorporate flexible development and discovery: biosimilars, strategic acquisitions

◦ Expand global presence

Current bioreactor facilities operate at high capacity and low cost

For next-gen facilities, an assessment of true benefits was performed◦ Perfusion-based at 2000L single-use bioreactors match the productivity of a 15KL reactor

◦ Some in-line analytics for control of product quality attributes

◦ Options for purification◦ Pooling and batch purification

◦ Semi-continuous purification

◦ Continuous purification


Next-Gen Facility benefits◦ Reduced footprint in size

◦ Manufacturing at less cost and less time

◦ High productivity in small reactors

◦ Single-use bioreactors – 95% surface contact is single use

◦ Connected purification processes; no pool vessels needed

◦ Cell culture and VF is a closed process.

◦ Closed processes allows most operations in ISO 9

◦ In-line and at-line testing

◦ Mixing vessels are single-use

◦ Harvest technologies for high-density ell culture◦ Flocculation at end of production with cationic propertie swith minimal filtration steps for clarification (4 hr process)


Downstream Process: Connected Process◦ Large pool vessels not required

◦ Column equilibration occurs in parallel with previous purification step

◦ Column cleaning occurs right after purification process step

◦ UF/DF steps was used to control the upstream purification processes

Process and product control using real-time analytics based in light scattering technology

Microbial control is achieved by using the closed process and systems◦ Pool holds are not necessary, thus reducing potential for microbial proliferation

◦ Single-use vessels and connectors eliminate liquid hold ups on pipes and tanks

Reduced environmental impact◦ Less water utilization, reduced CO2 production, reduced utilities in mostly ISO 9 environment


Next Gen Facility◦ Modular facility incorporating QC Labs, utilities, warehouse, mechanical support, manufacturing,

administrative offices and amenities

◦ Reduces CAPEX and footprint

◦ Flexible scaleability and capacity

◦ Agile tech transfers

◦ Reduction in facility time to deploy

iSKID: Next-Gen CM Platform for BiologicsNuno Fontes, Boehringer IngelheimPharmaceuticals in collaboration with Pfizer

iSKID: Next-Gen CM Platform for BiologicsNuno Fontes, Boehringer Ingelheim Pharmaceuticals in collaboration with Pfizer

Integrated skid for cell culture and purification◦ Non-steady state cell culture with short duration (14 days)◦ Hybrid downstream platform: ProA, VI, flow-through AEX plus Nano-filtration, UF & Formulation

◦ Two tandem ProA columns alternating in use

◦ One batch = 1 bioreactor + 1 VF = 1 DS◦ Intended for early stage clinical development; no development required fro Ph1 platform

Key features◦ Short-term◦ Cincentrated medua and buffer feeds◦ No holds, reducing bioburden opportunities◦ Flexible bioreactor sizes 100L – 2kL◦ Flexible flow rates◦ High cell line productivity◦ Low complexity for tandem ProA◦ 3- printed incubation chamber for VI


Pooling after AEX to reduce complexity◦ Stable intermediate

◦ Inexpensive downstream unit ops

◦ Easy to add additional polishing step at the batch stage

Some Challenges◦ TFF fouling

◦ Difficult osmo control in bioreactor

◦ Too big of batch for single-use facilities


Business Drivers◦ Productivity – up to 30-60 kg from the same cell line used for fed

batch; 5-10X more material than fed batch

◦ Reliability – enables single-use; no CIP/SIP required; more automation/less human errors

◦ Flexibility – enables faster response for clinical demand; delays investment for pivotal trails; less re-supply campaigns◦ CONS – less manufacturing experience by PPQ/BLA


Implementation scenarios◦ End to end proof of concept

◦ Approved product process improvement

Impact to CMO business◦ May compete with current installations

◦ May encourage customers to making more and buying less manufacturing slots

◦ Cost of customizing hardware/automation platform vs off-the-the shelf platforms

Session 2 Panel Discussion Q&A

Session 2 Panel Discussion Q&AIs the industry ready for CM? RoW and RoEU?

◦ Not quite. We are still working with coming up with solutions to develop solid processes◦ Not ready yet fro full CM, but close for hybrid processes◦ From a regulatory perspective, consider Q8-Q10 are still works in progress. Q13 may be able to come online faster◦ Technology is evolving rapidly and we hope the technical regulatory challenges are solve quickly. The pause is making a

worthy business case◦ We have basic elements of CM in place for 2 decades, so we should not think we are not ready yet.◦ P.A.T. and sensors are not quite there yet.

Two approaches were discussed: CM at early development and CM for late stage/commercial. Is Amgen ready to invest in CM for early development?

◦ Business decision to not use pipeline molecules, but a process established in 2 commercial facilities.◦ A second approach was to develop a platform for FIH processes

How did BI determine not to modify the platform when it is used in early stage, when the practice is to modify as development progresses?

◦ You always have to alter your process if the CQA profile is not adequate.◦ Rapid iteration of manufacturing is important to reduce the costs in clinical development.

Session 2 Panel Discussion Q&AIs there a discussion at BI to implement CM for commercial manufacturing?◦ Yes, but cannot share publicly. BI is more likely to favor CM for early stage

material.

How was comparability assessed for Amgen’s CM process vs commercial process? Were elements of Q5 used?◦ Elements of Q5 were assessed and a single CQA from commercial parsed out

from consideration (glycation = non-CQA for CM process). Glycosylation was controlled given the flexibility of producing media. pK studies were included in comparability package. CM facility and process were licensed successfully.◦ Active consultation with several agencies was employed for the approach

Session 2 Panel Discussion Q&ANow that Amgen’s CM processed is approved, what will be the comparability approach for transferring the processes to other facilities?

◦ No clinical comparability will be assessed for this one process/product. Only analytical comparability will be assessed. This approach will be discussed with multiple regulatory agencies.

Can you give examples or success stories of partnerships of the Gates Foundation with industry on MoW establishments?

◦ None have gone into markets. Currently working on clinical stage for a polio vaccine. Investing in process improvements of legacy products to reduce costs. Investing in education to regulators in African countries to increase their skills for improving processes and quality systems.

Based on Amgen’s experience in Singapore, would you consider moving the sterile boundaries for the process?

◦ It is best to continue with the segregation and no future plans to eliminate the boundary.

Session 2 Panel Discussion Q&AAmgen CM process:

For DSP, do you use closed equipment or is it placed in an isolator? Have you received challenges from inspections/lessons learned?

◦ No isolators are used in this facility.

◦ Initially, there were leaks that needed rectification. Dealing with vendors was key to resolve this issue.

◦ For RoEU the approach is to inspect on site with various support team members (assessors) to truly understand the process before evaluating.

How will the space constraint work when implementing multiproduct in a this facility?

Does eliminating hold times result in less flexibility for manufacturing process?◦ Amgen has chosen not to hold, because chrom steps are efficient.

Session 2 Panel Discussion Q&AFor regulators:

What kind of information are you looking for in a submission?◦ Establish a clear difference between regular and integrated process

◦ Demonstrate Chrom capacity and control of quality attributes

◦ Demonstrate with validation input, outputs and capacity

◦ Show how to investigate process failures (eg. filter integrity failures)

◦ Real-time data to demonstrate modelling is accurate

◦ Existing guidance must be followed

Do you foresee a convergence of models for CM? ◦ All 4 approaches presented today had a different flavor to them, because IP will rarely be shared by

companies

◦ We need to come to a harmonized nomenclature for CM. Harmonize the term “integrated”, for example.

Session 2 Panel Discussion Q&AHow do you segregate the non-conforming material when the analytical results are far behind the CM process step?◦ It will depend on the definition of a “batch”. Pool of material, pooling rate,

etc. The data has to be representative of the entire batch.

◦ It is a reality for non-CM processes to discard material if contamination can not be traced back to a particular step or time. The same should apply to CM.

◦ This decision must be taken in conjunction with the regulatory agency,

How many failures in the CM process can be tolerated before scrapping the material?◦ It’s not about the number, it is about investigation, root cause determination

and addressing/controlling the issue.

Session 2 Panel Discussion Q&ACan you incorporate re-processing/re-work in a CM process?

▪ Some process steps lend themselves to re-processing

▪ CM downfall is when they fail, you will very likely lose the material

▪ It may boil down to making a business decision on how long will you run a CM process

Example: Finite 14 day process vs 30 days

How reliable are the probes and sensors used in these CM processes? What is the Plan B if a failure occurs?▪ This should be included as part of development

How do you validate and define ranges for input parameters?▪ Based on the target length of the process, you would have to demonstrate at which point components

such as resins will foul

▪ Once you develop a process in steady-state, product quality parameters may be tightened.

Session 2 Panel Discussion Q&AAre we focusing on the right attributes picked up by P.A.T.? Should we be focusing on controllable quality attributes?

◦ P.A.T. is not appropriate at early development of process. It makes more sense to use P.A.T. after there is a good working understanding of the process.

Seems like P.A.T. methods are developed by PD, not by AD. Is this a true observation?◦ At Sanofi, BI and Amgen, P.A.T. method development belongs in the analytical space. Parallel assessments

between off-line methods and P.A.T. are performed by AD

◦ It should be a combined effort. PD provides process information, AD develops the method.

How close are we to establish parametric release approach?◦ Not close…

Has CBER established a mechanism to enhance pre-dialogue with the sponsors for emerging new technologies?

◦ CBER encourages sponsors to approach for advice. No mechanism exists similar to CDER’s.

Session 2 Panel Discussion Q&AHow to obtain sufficient representative material to set specifications and demonstrate process control, if the CM process lends itself to manufacturing less batches/less manufacturing experience?

◦ May take a similar approach as small molecules, generating worst-case material?

◦ All would be needed is data from every batch used in the clinic demonstrating safety

◦ Need to have data at the commercial scale

◦ Evaluated in a case-by-case basis for those processes with little experience. May incorporate post marketing agreements to revise specifications.

continuous manufacturing for biologics...dp applications mainly for solid oral dosage forms....

Documents