High performance liquid chromatography (HPLC) and biomanufacturing have a long, interwoven history. HPLC supports every aspect of biotherapeutic discovery, development, and manufacturing, to the point where thinking of biotech absent HPLC becomes difficult. Through their ongoing co-evolution HPLC will assume even greater importance as the analytic engine for Biopharma 4.0.
Advances in HPLC have affected small molecule drugs too, but not to the same degree as for biotherapeutics. “Traditional” HPLC, using larger particles and sub-6000 psi operating pressures, are a mainstay in pharmaceutical quality control, a situation unlikely to change any time soon.
“The need to convert these systems to greener, more efficient UHPLC systems is low due to the large cost associated with the revalidation of methods used for the release and in-process testing of marketed drug products,” says Nivesh K. Mittal, PhD, a product manager at Shimadzu Scientific Instruments. “UHPLC use is heavier in research and development, however, since the promise of shorter method development and run times, and higher resolution, is lucrative to time-conscious researchers.”
According to Mittal, growth rates for conventional HPLC and UHPLC are 3.3% and 5.1%, respectively. Higher growth for UHPLC is partly driven by instrument makers focusing R&D on that category. Shimadzu, for example, recently introduced an inert UHPLC system, Nexera XS inert, whose non-metal fluid path is suitable for the analysis of biomolecules at high sensitivity and resolution.
“HPLC has been gravitating towards biotherapeutics for the past fifteen to twenty years,” he says, “while the small molecule application horizon is basically saturated.”
Impact of advanced detection modes
Advanced detection modes, particularly highly sensitive mass spectrometry (MS), have fueled many of the technical developments in HPLC/UHPLC. LC-MS offers numerous advantages over conventional ultraviolet or photodiode array (PDA) detection, including improved sensitivity and specificity, higher confidence in confirming compound identity, and more rapid method development, all of which have led to greater adoption of LC-MS.
“LC-MS allows simultaneous identification and quantitative analysis of multiple compounds, even if the chromatographic separation is imperfect,” Mittal explains. “This ability dramatically improves an assay’s throughput for multi-component methods.”
The reverse is true as well, as HPLC/UHPLC system quality is critical for maximizing the contribution from the MS. “The essential performance attributes of LC systems such as injection speed, sample capacity, and minimizing carryover are key areas of LC system development. The autosampler accompanying Shimadzu’s aforementioned Nexera system utilizes a needle-in-the-flow-path or total-volume-injection design, plus multiple rinse options, for near-zero carryover.
“In the simplest terms, (U)HPLC and LC-MS must complement each other to maximize performance of the entire system and achieve a faster return on investment for the laboratory,” Mittal tells GEN.
A significant factor in the evolution of HPLC/UHPLC, and a key element of 4.0-type thinking, has been the emergence of advanced data systems, particularly artificial intelligence (AI).
“Within HPLC, as in many other areas, we see the adoption of artificial intelligence in automating procedures and enabling more efficient workflows,” Mittal says. For example, “smart” flow control increases mobile phase flow rates gradually to the set point, which helps to extend column life and keep LC systems running efficiently. “Mobile phase monitoring provides a real-time eye on mobile phase levels and notifies personnel if mobile phase volumes suffice to complete the run. This can help maximize uptime,” according to Mittal.
Another emerging data method is PDA peak deconvolution. While this is not a proprietary Shimadzu technology, the company has gone the furthest in applying it to detection of hidden peaks, impurities, or coeluting intermediates, e.g., through its intelligent peak deconvolution analysis (i-PDeA II application).
“Peak deconvolution uses all available data and applies an algorithm to ‘separate’ peaks that are not resolved on-column,” Mittal says. “This allows, for example, detection and characterization of coeluting impurities in a potency assay, unexpected coeluting reaction products, or hard-to-separate degradants.”