Angelika Gratzfeld-Huesgen, application chemist, Agilent Technologies, describes how fast liquid chromatography methods are speeding up impurity profiling in QA/QC
The analysis of impurities in drug substances, beginning with the initial screening and ending with the use of validated methods in routine QC and QA, is becoming an increasingly challenging task along the pharma-ceutical value delivery chain. Rapid Resolution Liquid Chromatography (RRLC), in combination with mass spectrometry, can improve the overall workflow for analysing and identifying impurities in drug substances.
In general, impurities in drug substances are addressed from two perspectives. The first is the chemical aspects, which encompass classification and identification of impurities, and require reports to be generated, appropriate specifications set and analytical procedures described. The second covers the safety aspects for the patients who will use the final product when a new drug is brought to market. Comparative studies and genotoxicity testing are of increasing importance in this context.
From the point of view of regulatory bodies, such as the US Food and Drug Administration and the European Medicines Agency (EMEA), impurities in drug substances are classified in the following categories:
- Organic impurities (process and drug related)
- Inorganic impurities
- Residual solvents
This article focuses primarily on organic impurities. There are
multiple sources for organic impurities at various concentration levels. Most frequently they may arise during the synthesis and storage of the active drug substance. However, they may also occur during the manufacturing process or storage of the final drug product and can originate from:
- starting materials
- by-products
- intermediates
- degradation products
- reagents and ligands
- packing materials
According to the FDA Guidelines Impurities in Drug Substances,1 identification of impurities below apparent levels of 0.1% is generally not considered necessary. However, identification should be attempted for those potential impurities that are expected to be unusually potent, producing toxic or pharmacologic effects at a level lower than 0.1%. In all cases, impurities should be qualified. Although it is common practice to round analytical results that fall between 0.05% and 0.09% to 0.1%, for the purpose of this guidance such values should not be rounded to 0.1% in determining whether to identify the impurities.
To summarise this guidance, highly sensitive measurement of impurities is required to reduce the risk of potential adverse reactions. Consequently, time-of-flight (TOF) mass spectroscopy, in combination with liquid chromatography (LC), is the method of choice to achieve the high sensitivity and mass accuracy that is necessary to identify trace level impurities.
Figure 1 shows a typical workflow in an analytical method development and QA/QC laboratory during drug development and commercialisation. This depicts the analysis and determination of production-related impurities in an active pharmaceutical ingredient. Impurity analysis follows the typical method development workflow, method optimisation/transfer and subsequent routine use under GMP conditions.
All these analytical procedures can be improved by using RRLC, which can speed up the analysis and, combined with TOF, provide accurate mass information in the initial phase of the investigation. RRLC can provide up to 20 times faster analysis and 60% higher resolution than conventional High-performance liquid chromatography (HPLC). This means that RRLC can process more than 2,000 samples a day on a single system, in contrast to 100 samples per day with conventional LC.
Analytical QA/QC departments are faced with increasing numbers of samples and higher demands regarding data quality and reliability. Typically, for one sample run, 10-15 ancillary runs have to be performed to ensure correctness of qualitative and quantitative data. Consequently, analytical chemists in the pharma industry require appropriate tools to increase constantly the speed of analysis while also achieving higher sensitivity in drug analysis.
Fast LC methods, which have been thoroughly validated, can help to increase sample throughput significantly without compromising data quality. This example illustrates how the Agilent 1200 Series Rapid Resolution (RRLC) system in combination with sub-2-µm-particle columns and fast LC methods can assist in increasing sample throughput and decreasing costs per analysis by a factor of three to four.
Having developed a fast analytical LC method, and after validation of this fast method, the next step is to transfer this method to the QA/QC laboratory. Typically, detailed standard operation procedures (SOPs) are available to ensure that the risk of errors and misunderstandings is minimised. The following example discusses an SOP including a system suitability check, sequencing and pass criteria for the results obtained.
Agilent's Series 1200 RRLC system was used to perform sequencing and reporting of system suitability checks and analysis of samples and calibration mixtures. The chromatographic conditions were based on a fast LC method developed and validated in previous articles.2,3 Cycle times of approximately 5 minutes should allow a significant increase in sample throughput, and consequently a reduction in costs.
Several tests are required to qualify a sample as passed or failed. Calibration mixtures and control samples have to be used, as well as blank samples containing a solvent that is also used for sample dilution. Table 1 provides a compendium of requirements to determine precision, sensitivity, resolution and other method performance parameters for the samples.
Based on the requirements defined in this table, a sequence table was then set up in the Agilent ChemStation software (Table 2). This sequence contains suitability test samples at the beginning, calibration runs before and after the analysis of three samples and one control sample. In between these runs, pure water was injected to ensure that no carryover or ghost peaks were present that might falsify the next analysis.
As an example, the system suitability test procedure is shown. To test whether the system still fulfils the method requirements, a solution between 4 and 10µg/mL of the main compound and impurities A, B, C and D was prepared. This solution was then injected every day prior to the first analysis. The following parameters were tested and limit settings fulfilled:
- Precision of areas must be <2% rsd
- Precision of retention times must be <0.5% rsd
- Resolution must be >2 for all peaks
- Maximum peak width must be <0.08 min at half height
- k' must be 5 < k" < 25
- Signal-to-noise ratio must be >50 for all peaks
The results of the system suitability sample are shown in figure 2 and summarised in table 2. All limit criteria were fulfilled. In the same way appropriate parameters and limits for the calibration standards and for the control sample were evaluated. All tests passed the limit criteria. The analysed samples also passed the limit criteria. The amounts of impurities did not exceed the limit of 0.5% for all three samples.
Fast LC methods that have been thoroughly validated, can help to increase sample throughput significantly without compromising data quality. The sequence used in this example included 47 runs, which took approximately 3.9 hours, including 30 minutes for system suitability testing. By contrast, a sequence using a conventional method with a cycle time of about 20 minutes would take approximately 15.7 hours. Consequently, the fast LC method can greatly increase sample throughput.
In addition to time-savings, solvent savings are also significant. When using a conventional method with a 20 minute cycle time, undertaking 47 runs at a flow rate of 2.2 mL/min, 2068 mL of solvent would be required. However, using the fast method described here, a 5 minute cycle time requires only 517mL solvent for 47 runs. The cost per analysis will therefore drop by approximately a factor of 3 to 4. Table 3 provides an example cost comparison of these calculations. Although revalidation of a method can take up to two weeks and cost about US$4,000, this example demonstrates that, after the analysis of about 500 runs, updating a method to a fast analysis is cost effective.
While profiling of impurities is very important in drug discovery, development and manufacturing, it can be a bottleneck in the entire process. Because of their potential toxicity, it is crucial to identify pharmaceutical impurities and by-products, and to be certain of the results. However, schedule pressures dictate that this task must be undertaken efficiently. This is why modern LC/MS systems are designed to speed the process and make the task much easier.
In summary, a good impurity analysis must start with an excellent chromatographic separation. The small particle size and increased column length of LC columns provide greater efficiency and resolution. Combine these columns with RRLC that has very low extra column volume, and resolution can be increased by as much as 60%.
In addition to the time and cost savings that RRLC can deliver, this increased resolution also inspires greater confidence when identifying impurities via retention time, and provides cleaner spectra during MS analysis.