The Future of Chromatographic Method Development in Pharma

Jun 30, 2018

Analytical chemists are always seeking to make method development more efficient. For nearly two decades, Chris Welch led his teams at the Merck & Co. to constantly drive innovation in this area. Following his recent retirement from the company last year, he and his colleagues wrote a paper about the current state of method development for pharmaceutical research and where it is heading. He recently spoke to us about this topic.


Are there particular method development issues that chromatographers in the pharmaceutical industry face that are different from those seen in other types of analytical laboratories?

I think the need to accurately and reproducibly measure analyte molecules is pretty much the same across a wide variety of industries. But in the discovery and early development sectors of the pharmaceutical industry, the target molecules being investigated change on a very rapid basis. Consequently, creating new analytical methods becomes a daily task, which means we have to learn how to do it really well and really fast. It's a fun area of research that a number of researchers enjoy, and that's what inspired us to write our recent review in TrAC (1).

Method development screening has evolved quite a bit in recent decades. In the pharmaceutical industry, method screening for enantiomer separations has led this progress. What's the state of the art for method screening for enantiomeric separations? Is there still room for improvement?

The state of the art is constantly evolving, and the tried and true method of automated column screening is undergoing some big changes right now. In the past few years, we’ve gone from overnight screening to same-day identification of columns and conditions. This approach greatly speeds the pace of research, but further increases are coming. Recent progress in fast enantiomer separations now means that almost all enantiomers can be resolved in under a minute, some in just a few seconds. You can read about this in the free open-access article, “Are We Approaching a Speed Limit for the Chromatographic Separation of Enantiomers?” (2). It looks like chiral method development will soon become something that can be done over the course of a coffee break.

That's pretty cool.

It used to take about a week of work and then it was overnight and then it became same day. That was quite revolutionary, but it's getting faster and faster. The fastest we've ever done it was in about half an hour, but I think the routine screens that are used by most groups today still take about two to three hours. And that's going to come down to about 15 minutes.

How common is it for pharmaceutical companies to have dedicated screening instruments?

I would say that this is still the standard model. Of course, it doesn’t make sense to devote an instrument to method development screening if you have only a few researchers in an organization, but for companies with dozens of researchers, dedicated method development screening instruments simplify problem solving, saving time and labor and soon paying for themselves. We talk in the review about the historical challenges of combining screening and general utility functions within a single instrument, which looks like an area where AI [artificial intelligence] can be of assistance.

Is there any downside to having dedicated screening instruments, other than the investment?

Well, it's the investment and you need somebody to maintain it. This has often been done in central separation groups. But a lot of people are putting method development screening robots out for public use, as walkup instrumentation. I think that's the model that's going to be with us going forward.

In your experience, is chromatographic modeling software a significant time saver?

Some of my coauthors on the TraC review article are big advocates, and we’re seeing predictive tools playing an increasingly important role in method development. An exciting trend in this area is the integration of predictive and experimental tools to rapidly solve problems.

What do you think is the future of algorithms and intelligent systems for chromatographic method development for pharmaceutical analysis? Do you think we'll soon see a “self-driving” chromatography method development station?

Yes, this is certainly a very exciting area. Researchers have been pecking around the edges of this problem for many years, and I expect the field to rapidly move forward in the next few years.

It would seem that developing a self-driving method development station would require effectively harnessing the vast amounts of method development data that chromatographers produce. Are any significant advances being made to capture and interpret these data? Is the Allotrope Foundation’s model (3) a step in that direction? 

You raise several important and intertwined points here. There are already databases of analytical information that can be used for developing predictive models. We have been working with Christian Roussell and Patrick Piras at the University of Marseilles to develop models for structure-based chiral column selection based on the more than 200,000 entries of literature-reported chiral separations in the Chirbase database created and maintained by Christian over several decades. Our first efforts (4) were somewhat effective, but we were really hampered by the lack of negative data in the scientific literature—nobody publishes assay conditions that don’t work! But Patrick has recently figured out a new approach, giving us reasonably good predictive models for most chiral stationary phases (5). As you point out, the impact of this shortage of negative data in the literature can be overcome by mining the data coming from method development screening stations, where there are plenty of failures. I am optimistic that recent progress by the Allotrope Foundation in developing a common data format for analytical instrumentation can be expected to facilitate further progress in this area.

What have you been doing since your retirement from Merck last year?

Well, I’m certainly keeping busy. I started Welch Innovation, LLC, and have had some fun with consulting projects, but since December, I’ve been leading a very exciting effort called the Indiana Consortium for the Analytical Sciences, which is a joint venture between Purdue, Notre Dame, and Indiana University. I am really honored to be working with these great universities and their outstanding faculty on collaborative projects that take analytical science to the next level. We’re currently setting up an industry–university center, called the Center for Bioanalytic Metrology, (6) that will focus on addressing key measurement science challenges of industry. Uptake from industry participants has been great so far and extends well beyond traditional pharma and biopharma companies and into the energy and food sectors, and into other sectors as well. This is really a great time for our field of analytical science, and I’m certainly looking forward to some exiting times over the next few years.

That sounds like fun.

Yes, it is. Some analytical problems are the same across industries, but sometimes people come to the consortium with problems that I had never heard about. This consortium is a great mechanism to flush out key problems that need solutions. I like problem solving, so I’m having fun.

References

  1. F.T. Mattrey, A.A. Makarov, E.L. Regalado, F. Bernardoni, M. Figus, M.B. Hicks, J. Zheng, L. Wang, W. Schafer, V. Antonucci, S.E. Hamilton, K. Zawatzky, and C.J. Welch, TrAC 95, 36–46 (2017). https://doi.org/10.1016/j.trac.2017.07.021
  2. C.J. Welch, ACS Cent. Sci. 3(8), 823–829 (2017).
  3. J.M. Vergis, D.E. Vanderwall, J.M. Roberts, and P.-J. Jones, LCGC North America 33(4), 270–281 (2015).
  4. R. Sheridan, W. Schafer, P. Piras, K. Zawatzky, E.C. Sherer, C. Roussel, and C.J. Welch, J. Chromatogr. A. 1467, 206-213 (2016). https://doi.org/10.1016/j.chroma.2016.05.066
  5. P. Piras, R. Sheridan, EC Sherer, W. Schafer, C.J. Welch, and C. Roussel, J. Sep. Sci. 41, 1365-1375 (2018). DOI: 10.1002/jssc.201701334
  6. The Center for Bioanalytic Metrology. http://sites.nd.edu/cbm/.

Chris Welch is principal with Welch Innovation, LLC, an independent research and consulting firm located in Cranbury, New Jersey.

lorem ipsum