# Multiplatform Benchtop NMR Interlaboratory Study of Model Liquid Dosage Forms of Pharmaceutical Products

## TL;DR

This study tests how consistently low-field NMR can measure pharmaceutical liquid samples across different benchtop instruments and protocols.

## Contribution

The paper provides insights into harmonizing wNMR protocols to ensure reliable pharmaceutical analysis across diverse platforms.

## Key findings

- wNMR measurements from different instruments showed high linearity in tracking sample concentrations.
- Outlier samples were successfully detected using wNMR.
- Measurement variance was more influenced by protocols than by hardware differences.

## Abstract

Low-field NMR relaxometry uses water proton (1H2O) relaxation (wNMR) as a powerful
and flexible
tool to reveal deviations in critical quality attributes of liquid
pharmaceuticals such as concentration or, in the case of proteins,
aggregation. Measurements are fast and nondestructive, and many benchtop
instruments can make measurements noninvasively, with an intact drug
vial, prefilled syringe, or pen inserted directly into the probe.
Because of the diversity of low-field instrument configurations and
operating temperatures, the varying field strengths available, and
the many possible protocols, pulse sequences, and parameters for conducting
relaxation measurements, we sought to test the reproducibility of wNMR across platforms, and to identify the details of applying wNMR that impact reproducibility. To accomplish this, we
piloted an interlaboratory study, in collaboration with benchtop NMR
instrument vendors, using model liquid dosage forms of pharmaceutical
products in sealed vials. This study demonstrates that when suitable
measurement protocols are employed, wNMR measurements
from all instruments can track the concentrations of diverse samples
with high linearity. Furthermore, wNMR measurements
were capable of detecting intentionally included outlier samples.
This study revealed various operating procedures and variables that
contributed more to measurement variance than any hardware- or instrument-specific
differences. Based on these results, we offer initial considerations
for harmonizing measurement protocols to enable reliable wNMR analysis of pharmaceuticals.