Guest Column | March 20, 2024

Process Development With "The End In Mind" For Startups

By Ray Knox, founder & principal consultant, KBDG

target guideline-GettyImages-953153186

The term “the end in mind” is encountered frequently when we consider the development of the most complex medical products, but what does it mean and when does it matter? The development of a new therapeutic or medical device is a multidisciplinary effort typically beginning with bench top science and culminating in the transfer to a reproducible and predictable manufacturing process. The path from start to finish is filled with both predictable and many unknown challenges. Developing with the end in mind is to recognize this pathway and take time to consider how and when to handle those challenges. In the world of developing unique or complex products, this becomes less clear and developing with the end in mind becomes problematic.

Why Does It Matter?

Critical failures can begin at the onset of product development. Startups, primarily composed of early-stage research scientists and engineers, are often laser focused on the science — the “Can it work?” question. This discovery stage is critical for securing funding and establishing the organization. While the work may be fascinating and intoxicating during the early-stage examination of key attributes, signals from bench testing, and early analytical success, it rarely leads to a well-documented pathway of historical significance. It often fails to capture crucial details, such as the exact materials used, the specific characterization of starting materials, the exact equipment setup, consumables used, and precise process settings, with sufficient detail. Documentation controls are not well-established mechanisms in organizations that likely grew from experimental meeting to experimental meeting. Thinking about the end in mind matters, and starting at an early stage can avoid the later pitfalls of having to redefine a process or simplify a device design.

Balancing Risk, Priority, And Product Development

At the onset of a new product development, timelines often focus on the first and last milestones: initial Phase 1 success and NDA submission for a new drug product. Many new company leaders highlight these inflection points as natural fundraising events. However, taking the time to detach from those measures and assess the likely factors that will need mitigation — such as materials, processes, equipment, or toxicology results — requires the leadership to think with the end product in mind. Even more critical is the ability to step outside one’s comfort zone to prioritize the most challenging of those challenges, enabling the team to move forward while addressing possible future concerns in parallel.

This balance of risk, priority, and product development speed can be a formidable challenge. While it is preferable to have well understood and detailed documentation for all aspects of a product, some considerations should be evaluated, as the impact may be more beneficial in certain circumstances. For example, a startup facing a short runway of funds needs to prioritize speed to a result to ensure the continuation of funding. Great science is valuable, but the ability to show that your science works is essential to establish the value of the company. Don’t let perfect get in the way of good enough to continue funding of the company.

This pressure on speed is not unusual and can be achieved by using small batches or bench prototyping with limited testing and poorly understood process variation. Without assessing scale-up concerns or design for manufacturability, these rapid approaches can lead to longer delays in the future. One pitfall here can be the wanted, but unexpected, success that leads to an immediate hiatus. What if the early-stage development is outrageously successful in a Phase 1 trial, and the regulatory pathway jumps to a fast track or break through therapy? Without well understood, validated, or well documented GMP manufacturing steps, the company faces an almost impossible task of converting poorly understood processes or product definition into an expected product with a well-defined target product profile (TPP), critical quality attributes (CQA), critical process parameters (CPP), and trained staff to use GMP work steps to produce the desired product. This task is nearly impossible without having taken a staged approach to development.

Staged Development Work Mapping Phases To QbD

Most medical device developers have a stringent approach to product development that follows stage gate approaches with three to seven different stages following design control requirements as the product/process matures through early development through commercialization. A typical five-step process could include Definition, Feasibility, Verification, Validation, and Commercialization. Similarly, the pharmaceutical phase-appropriate quality by design (QbD) approach can also create a staged approach to product and process development that builds with knowledge gained. The key to applying these approaches is knowing when to start the rigorous details needed to create good technology transfer between stakeholders.

These are not new ideas, but in the world of small startups there is often a lack of appreciation for the process steps and how to apply them to unique products. There are a couple of approaches that can be successfully applied. Many contract design or manufacturing organizations mimic the practices of large medical device and pharmaceutical companies. So, one approach can be to leverage the knowledge, experience, and business processes these teams have to offer by using a more virtual structure or transferring manufacturing development to these CDMOs. To be successful, there needs to be a clearly stated expectation of the partnership as well as the explicit thought process around what is well understood and documented versus still under development and subject to change. Complete TPPs, CQAs, CPPs, and work instructions may not be available at the outset, but starting with draft outlines that identify input materials, equipment setup, and process windows around critical process settings in a table form is key to moving along at a reasonable rate. A second approach is to commit to an internally developed phase approach. But without the typical well-developed policy-driven approaches that large or mature organizations typically implement, a startup has to almost define this system on the fly. This is less than ideal, but there are tools that can help.

Knowledge Management

One way to help organizations get a handle on early development and a graduated approach is to create a knowledge management interface that makes it easier than traditional electronic lab notebook software to create a searchable database of the product and process development evolution. Larger organizations drive most development work into some looser form of its document control system leveraging an established quality management system, building engineering or prototype workflows that become more easily translatable to later-stage GMP requirements. Most startups do not have the luxury of these systems and the infrastructure required to support these systems, having developed their science on the bench without a clear understanding of the end in mind: the production of a GMP product in a repeatable and reproducible fashion. Often, institutional knowledge is embedded in lab notebooks as data, without the details required to repeat successful results. Incoming materials may not have been fully characterized, so the reason why experimental or process results change with a new batch of raw materials is not understood.

To break from these pitfalls, it is recommended that internal knowledge be managed as early as possible. This can be achieved not just through data captured in lab notebooks, or information gleaned from presentations, but instead by creating a living history. The use of systems like Helpjuice, Confluence, Sharepoint OneNote, or others can offset the lack of electronic quality management systems with change controls and change management structures in place. They require rigor in establishing the mechanism and style of capturing and reporting results. The use of searchable database tables can help build out key information from raw materials, equipment setups, consumables, process parameters, and analytical test status. By incorporating some basic risk tools into these tables, these frameworks can be used to successfully scale up or transfer the process to a CDMO.


Developing with the end in mind for startups requires a balanced approach. Planning out a series of value-creating milestones is critical to the continuing funding, while maintaining a structured advancement of the product development is key to avoiding later commercialization delays.  Establishing a staged product/process development approach from the outset can support both goals. Creating the expectation that a clearly defined product that can be manufactured at the target scale to perform in a desired manner is ideal. Establishing a tabular format for process inputs, process equipment, process variables, and process outputs that evolve with each experiment on the path to a finished product sets the mantra. The ability to risk assess ahead of time can allow organizations to balance development speed and mitigate experimentation. This can be applied to the product, the process, and the analytical and bench test methods that are needed to support the evaluation of the product performance over the same development cycle. Use of knowledge management software can help cross-functional teams build out information over time so institutional knowledge is captured and becomes the building block for future GMP activities.

About the Author:

Ray Knox has more than 35 years of life sciences experience in process development and manufacturing for complex therapeutic products in numerous technology platforms and, most specifically, with challenging drug release platforms, including bioresorbable implantable systems and long lasting oral drug delivery. He is an independent consultant with a focus on project management, CMC, technology transfer, and scale-up with internal and external operations. He has held management and senior positions at Lyndra Therapeutics, Ximedica, and 480 Biomedical, with previous experience at Boston Scientific, Medtronic, and Becton Dickinson.