Three ways to overcome attrition in NCE development

Sequencing of the human genome at the turn of the 21st century promised a new era of opportunities for pharma, raising expectations as well as R&D expenditure. But while the pressure to innovate is greater than ever before, the FDA is actually approving 25% fewer drugs than it did in the 1990s.

So what’s going on? Why are pharmaceutical companies spending more money and time (often up to 15 years) developing new chemical entities (NCEs), only to see them fail before reaching the market? And how can we do better?

Why are so many NCEs failing during development?

Several studies have sought to understand the key causes of attrition in drug discovery and development. Bringing together data from multiple pharmaceutical companies, these investigations have revealed some interesting findings.

A recent study by Waring and colleagues analyzed the main causes of attrition in drug development based on a review of 812 NCEs under development by AstraZeneca, Eli Lilly, GlaxoSmithKline and Pfizer. This study also considered NCEs in the preclinical stage, reviewing products nominated to enter preclinical toxicology studies before progressing to Phase I.

Overall rates of progression from the preclinical stage through to Phase IV were low. While many drivers of attrition were identified, non-clinical toxicity, and interestingly, portfolio rationalization were cited as the most common causes of failure in drug development. When it comes to the clinical stage, clinical safety and efficacy were the most common contributing factors to program failure, accounting for just under two-thirds of overall attrition.

How can we reduce rates of attrition in NCE development?

These findings highlight the urgent need to improve overall health of the pharmaceutical pipeline. So what strategies should drug companies adopt to help overcome this challenge?

1. Drug physicochemical properties matter

There’s been much discussion of late around the value of Lipinski’s “Rule of Five” in modern drug development. Lipinski’s rules were originally developed as guidelines to improve the bioavailability of orally-administered drugs by setting cut-offs for key physicochemical factors, such as molecular weight. However, their widespread adoption across large sections of the industry has led some to question whether they are in fact contributing to the challenge of attrition.

While there’s certainly a case for adopting a more nuanced and pragmatic approach to applying Lipinski’s rules, the Waring study shows that the principles underpinning them still shape success. In fact, although a number of approved NCEs violated either the molecular weight or calculated logP rule, only 8% of drug candidates violated both.

Overall, NCEs that gain approval tend to be smaller, less lipophilic, have rotational flexibility and have fewer carbon aromatic rings. Data shows that molecules with the following physicochemical properties are more likely to become drugs:

• Molecular weight ~ 400
• clogP = 2.8 - clogD7.4 = 1.8
• sp3 fraction = 0.42
• Aromatic ring count = 2.1

Interestingly, compounds with high clogP or clogD7.4 values are more likely to fail for clinical safety reasons, while no such association was shown for preclinical toxicology failures. Zwitterionic compounds are more likely to survive preclinical toxicology, but have higher failure rates due to poor pharmacokinetics.

2. Understanding the disease-target relationship is key

While selecting better drug candidates is one way to boost success in pharmaceutical development, it’s also important to focus efforts on the right disease targets. As the disappointing progress of drugs targeting neurodegenerative disorders such as Alzheimer’s have shown, fully understanding the disease-target relationship is key to achieving better NCE survival rates through both the development and clinical phases.

Network analysis is one technique that has proven valuable in improving target selection, and helps move beyond the traditional one-effect, one-cause, one-target paradigm. Using high quality, strictly curated data, pharmaceutical companies can map node-to-node interactions. When combined with genomics data, these networks can provide a powerful set of analytics to obtain deeper insights into disease pathways. 

3. Look beyond traditional R&D dogma

It’s clear that the established pipeline for producing “blockbuster” drugs isn’t working effectively. The traditional approach requires focusing on diseases that are typically chronic and multifactorial, while relying on multilayered investment and requiring large-scale clinical trials. Ultimately, it’s a model that’s become too cumbersome and costly to maintain.

To address attrition due to portfolio rationalization, some pharmaceutical companies are moving away from the traditional model and are forming more collaborative partnerships. Rather than maintaining shadow departments and simply outsourcing areas of expertise, these companies are reconsidering the traditional, all-in approach, by empowering partner companies to act as true centers of competence. This structure is helping to increase efficiency and more quickly establish whether drugs (and their targets) are commercially viable.

Watch the webinar: The challenge of new chemical entities attrition

With attrition a significant challenge in pharmaceutical development, it’s essential that companies are using the best strategies to ensure progression of NCEs from the bench to the bedside.

Watch our webinar with Dr. Kelvin Cooper, a consultant with KC Pharma Consulting and former Senior VP at Pfizer, to gain a more complete understanding of the challenges surrounding NCE development and the strategies that can overcome them.