The Art of Reducing API Failures.


Despite the many risks for pharmaceutical companies, new drug discovery marches on.  According to PhRMA nearly 7,000 medicines are in development. Of these, 74 percent are potentially first-in-class across dozens of therapeutic areas – including cancer, diabetes, cardiovascular, and Alzheimer’s disease among others. Each one of these new drugs begins with development of the active pharmaceutical ingredient (API).

What is driving new drug discovery? There are three main reasons for pharmaceutical companies to develop new therapeutics:

  • Attempts to develop medications with fewer side effects
  • Need for new or improved treatments for formerly untreated diseases
  • Creation of new pharmaceuticals that offer an advantage over existing drugs

In the last 15 years or so the complexity of the molecules that comprise the APIs that can achieve these desired therapeutic effects has increased. This article will focus on considerations in the early stages of complex API development that can help avoid the failures in the process that could lead to delays.

Growing complexity

Last year saw strong growth for CDMOs, which is in part likely due to the increased difficulty, cost and need for specialized expertise in developing today’s APIs. Whereas a few years ago development of many APIs required just a few chemical steps, it is not uncommon today for the API development process to require many more steps than before. Getting from one step to the next adds considerable time to the development process at the same time that sponsors are under increasing competitive pressure to bring new drug candidates to market quickly.

The numerous reasons for API failures

The more complex the molecule, the more opportunity for failure. An opinion piece that appeared in Pharmaceutical Outsourcing lists no fewer than 23 ways in which API failures can occur. Written by industry veteran Max Lazar, whose 50-year career included 35 years at Hoffman-La Roche, where he was Vice President, FDA & DEA compliance in charge of APIs, includes failures in laboratory operations, in records, in raw material information, in cGMP, among others. They include such infractions as:

  • Failure to maintain complete records of your APIs
  • Failure to perform testing of APIs to ensure conformance to specifications
  • Failure to maintain complete data derived from all laboratory tests
  • Failure to ensure equipment is cleaned in a reproducible and effective manner
  • Inadequate investigations of critical deviations
  • Failure to document manufacturing operations at the time they are performed

The frighteningly long list goes on and on.  I’m sure in his long career Lazar dealt with each and every one of the infractions he listed and then some. His point is that these deviations from GMP are preventable, and I agree.

Preventing API failures begins with planning

As many of us who have been in the New Chemical Entity (NCE) business for a long time know, one misstep in the process can cascade into weeks of extra work trying to rectify the error, be it a poor choice of raw materials or failure to notate early experiments that didn’t work for whatever reason.

Preventing API failures begins early, and it is an art. Every project begins with a plan, but development of complex molecules requires more thorough planning.  The steps in synthesis are likely to be different. The raw materials are likely to be different, with special attention to controlling impurities. Some raw materials may themselves need to be synthesized and tracked.

The earlier you start in the planning process, the better off you’ll be. And the more detailed and artful the plan for approaching the project, before it begins, the more likely the project will be executed efficiently. Efficiency saves time and money.

Realizing there is complicated chemistry to be done, the plan needs to answer questions such as:

  1. What markers do we use while doing process development?
  2. What reference standards will we use as a benchmark?
  3. Do we have a protocol that will be iterative?
  4. What will we learn that’s new and will help with the next experiment?
  5. What are the lowest permissible dose limits to clean to?

Flexibility is key

We have seen many haphazard plans when technology has been transferred to us that doesn’t even come close to answering these questions. Since we make every effort to be efficient, we prefer putting together an overarching plan at the very beginning that anticipates potential problems and roadblocks to avoid.

However, that’s not to say we want the plan to be a rigid playbook from which we do not veer.  Just the opposite: process development is always about being flexible. In the course of API development, we are always reacting and adjusting the work until we come up with the desired API. With a detailed plan in place, in most instances we can move from the first process step to the last as quickly as is possible when dealing with new chemistry, which is always full of surprises, not all of which can be anticipated.

Software and technology

As the degree of complexity in developing APIs increases, as it continues to do, it’s important for sponsors to understand what CDMOs face to get them over the finish line.

Particularly with larger, more complex molecules, we have to determine the structure that gives the API its therapeutic value. We have to assure it’s scalable and pure enough to become a drug substance.

We continuously evaluate and acquire new technology that will make this work more efficient. We have detailed some of our technology upgrades in recent articles – micronization , biotage, lyophilization, and cryogenics, among others.

Our PhD scientists also employ software that helps analyze the structure of the molecule and suggest a host of experiments and provide more insight into the chemistry.

Contingency planning

Early planning includes thinking things through artfully to come up not only with a synthesis plan, but also contingency plans. For example, while producing an API that was a liquid at room temperature we encountered significant challenges in purification. We couldn’t distill it, as that would have resulted in decomposition. The molecule was also air- and light-sensitive.

Using a contingency plan, the team came up with a different way to process the compound under conditions that wouldn’t harm it and that retained the desired purity. As Louis Pasteur famously said, chance favors the prepared mind.

We have written numerous articles about saving money or preventing negative surprises during API manufacturing. If you’re interested in learning more, sign up for our newsletter or call us at (978) 462-5555.