Why System Suitability Tests Are Not a Substitute for Analytical Instrument Qualification (Part 3): Performance Qualification (PQ)

In two recent Questions of Quality columns (1,2) the topic of why system suitability tests (SSTs) were not a substitute for analytical instrument qualification (AIQ) was discussed. In these two columns, the focus of the discussion was on the initial qualification of the chromatograph and that SSTs could not be a substitute for AIQ because the former is method specific and the latter is instrument specific. The columns looked at the first three phases of the 4Qs model outlined in USP <1058> (3): design qualification (DQ), installation qualification (IQ) and operational qualification (OQ).

However, these two columns did not consider system suitability tests run during routine analysis because the focus was solely on the initial qualification of a chromatograph. The last stage of the 4Qs model — performance qualification (PQ) that is performed after the operational release of the instrument — was not discussed.

After publication of the second article, I was contacted by Hermann Wätzig who mentioned that the articles had not considered how SST results could be used effectively in PQ. He, along with his colleagues, suggested that a third article focusing on the role of SSTs during PQ should be written.

What is performance qualification?

First we will consider what is meant by PQ using the definitions and explanations contained in USP <1058> (3) and then discuss how SSTs can be used to support the PQ phase of AIQ effectively. In this discussion we will use SSTs that are run during an HPLC analysis as an example, but the principles contained here can be modified and applied to GC.

Role of PQ in AIQ

PQ is an important part of analytical instrument qualification (AIQ) and according to USP <1058> (3) is defined as the documented collection of activities necessary to demonstrate that an instrument consistently performs according to the specifications defined by the user, and is appropriate for the intended use. The aim of PQ is to demonstrate that a chromatograph remains qualified and is performing as specified by the users in the chromatography laboratory.

Table 1: Performance qualification (PQ) parameters derived directly from existing SST execution.

Table 1 contained in the USP <1058> (3) defines PQ in more detail:

• Performed periodically at specified intervals for each instrument: The period can either be at specified time intervals (the traditional PQ approach) or continuously when an SST is run before and during an analysis. However, the frequency and the acceptance criteria will need to be defined by each laboratory based on their business and regulatory needs.

• Preventive maintenance and repairs: A regular service programme for the chromatographic instrument with associated records needs to be established. If an instrument fails to meet PQ test specifications, then it requires maintenance or repair. After this has occurred, then, as a minimum, the relevant PQ tests should be repeated to demonstrate that instrument performance is within specification. If a repair is expensive then a complete requalification of a module or the whole instrument may be required. However, we will not discuss this area any further in this column.

• Establish practices to address operation, calibration, maintenance and change control: This would involve instrument log books with regular peer review, how to requalify the instrument after minor or major repairs and a simple change control process. Again, this topic will not be discussed further here.

• Performance checks: We will focus on this area of PQ in the rest of the article and look at ways to generate this information from activities that are already undertaken during an analysis, as well as further information that could be obtained with modification of current analytical practices.

Performance Checks

According to USP <1058> (3) the laboratory should establish a test or series of tests to verify the acceptable performance of the instrument for its intended use. Quoting from the general chapter: PQ tests are usually based on the instrument's typical on-site applications and may consist of analysing known components or standards. The tests should be based on good science and reflect the general intended use of the instrument. Some system suitability tests or quality control checks that are performed concurrently with the test samples can be used to demonstrate that the instrument is performing suitably (3).

Some of the key points about performance checks are:

• The need to set specifications to compare the instrument performance against to determine if the instrument performance is acceptable or not.

• PQ tests may be modular (focussing on a specific module of the chromatograph. for example, the pump or detector) or holistic (looking at the whole system performance).

• It may be useful to repeat the same PQ tests each time the instrument is used so that a history of the instrument's performance can be compiled. This would allow trending of data and assessing if the instrument performance is maintained or is changing and would be a proactive approach to identifying potential problems before they adversely affected an analysis.

• Some PQ checks can be performed concurrently with the test samples to demonstrate that the chromatograph is working within a laboratory's defined limits for a specific analytical method. The development of this point is the focus of this column.

Point-of-use Instrument Checks

Although this column is discussing system suitability tests for chromatographs, other point-of-use checks are used to determine if an analytical instrument is functioning correctly before analysing samples. This is a common procedure in the majority of analytical laboratories because it is good analytical science. Consider the following examples:

• pH Meter: Before using a pH meter an analytical scientist would check that the electrode and the whole system were functioning correctly by checking with two or more calibrated solutions of known pH value that bracket the range of measurement. These solutions can be prepared in the laboratory or purchased with a calibration certificate from a supplier.

• Analytical Balance: Again, the balance will be checked before use with either masses that have been calibrated to traceable national or international standards or laboratory calibrated masses that are traceable to externally calibrated masses. The masses used would also bracket the range of measurement on the balance. This would be in addition to the balance's internal calibration mechanisms that can also be used to ensure adequate performance.

In both of these cases, the point-of-use check determines if the instrument is within acceptable limits from a holistic perspective. In exactly the same vein, a system suitability test confirms that a chromatograph is functioning and is fit to analyse samples for a specific analytical procedure.

System Suitability Tests as Performance Checks

All of the SST parameters are defined in the appropriate chapters of the Pharmacopoeias, for example the United States Pharmacopoeia (USP) <621> (4) or European Pharmacopoeia (EP) 2.2.46 (5). During the development and validation of an analytical procedure, the SST parameters most applicable to demonstrating that the chromatographic system works will be selected by the method developer and then validated as part of the overall procedure. As an SST is run before every chromatographic analysis in a regulated environment, the following question arises: How can the information contained in the SST injection samples be analysed to produce meaningful performance data of a chromatograph?

Traditional versus Continuous Performance Checks

Traditionally, a system check would be performed every three to six months depending how often the chromatograph was used. However if a problem was identified as a result of the performance check, up to the last six months of chromatographic results could be considered suspect. This is not an appealing situation for the manager of the analytical laboratory to be in. Therefore, an alternative approach should be considered to provide an effective and efficient means of assuring the performance of a chromatograph.

An alternative is continuous PQ, which can be regarded as a major improvement because this modern approach allows a laboratory to reduce time and effort in gathering performance statistics without losing data quality. Even better, it allows for the continuous monitoring of the instrument performance (6,7). This approach uses a holistic look at the most relevant performance parameters of a liquid chromatograph such as:

• Precision and linearity of the injection volume

• Injection carry-over

• Flow-rate precision

• Thermostatting precision of the column oven

• Linearity of the detector response

• Signal-to-noise ratio (S/N).

There are five parameters listed in Table 1, which can be determined from data obtained after running an existing SST (6,7). By using this approach there is the benefit of saving extra effort as these parameters can be checked on a daily basis compared with the traditional PQ approach where system performance is just checked every 3–6 months. Furthermore, it avoids raising the question that the results generated since the last traditional PQ may be questioned because issues with a chromatograph can be identified and resolved much sooner that using the traditional PQ. The tolerance values in the right hand columns of both Tables 1 and 2 were chosen according to reference 8 and this table also complies with the EDQM specifications (9) for HPLC performance.

Table 2: Additional performance qualification (PQ) parameters that can be derived by modifying an SST sequence.

Many of the parameters in Table 1, for example injection volume precision, mobile phase proportioning (if used), flow-rate precision and precision of column oven temperature, can be derived from the consistency of the retention time of a single peak over the time frame of an analysis.

Table 2 shows the additional seven parameters that can be monitored holistically with little additional effort by adapting and extending the SST design. The thermostatting precision and accuracy of the autosampler tray, flow-rate accuracy, mobile phase proportioning, wavelength accuracy and absolute values for noise and drift have to be measured separately as outlined in Table 2 (6,7).

The total list of PQ parameters shown in Tables 1 and 2 consists of up to 12 items that are necessary to monitor the PQ of an HPLC instrument thoroughly. This list combines both PQ parameters and some of the modular parameters of an operational qualification (OQ). Most of these parameters can be derived holistically by applying either normal (Table 1) or modified (Table 2) SSTs. Additionally, these parameters, which require modular testing, can be investigated without changing parts such as flow cells, columns or the mobile phase reservoir. Thus extra time for the sample and mobile phase preparation and long equilibration procedures is no longer necessary.

Note, however, that not all instrument PQ parameters can be evaluated by this proposed approach. For example, detector lamp energy cannot be measured by the continuous PQ but this parameter would be a less frequent performance check in a PQ and is easily accomplished manually.

General Procedure for Continuous PQ

If the principle of continuous PQ is of interest to you, how should it be accomplished? We suggest the following approach which is outlined in Figure 1 and in the text below (7).

Figure 1: A suggested approach for continuous performance qualification (PQ).

The basic continuous PQ test can derive five parameters from the system suitability test injections and the enhanced PQ can determine up to 12 parameters in a single run.

1. Examine the SST of your method. If there are multiple methods running alternately on the system, take the simplest one(s). However, you must make sure that the respective standard substances are well defined.

2. Look for a well-separated peak. You should find at least one as this is typically claimed for an SST (6,7).

3. Run the injection sequence consisting of the following samples, which for the basic performance determination should be structured as follows:

(a) One blank injection for the determination of baseline noise or drift.

(b) A minimum of five standard injections for the determination of all area and retention time parameters.

(c) One blank injection for the determination of baseline noise or drift.

(d) One injection of standard impurities if the method is used for impurity determination.

(e) Unknown samples for analysis (these may have standards between them depending on the calibration method used in the analysis).

(f) One standard at the end of the injection sequence for the determination of peak area drift.

4. Calculation of the various parameters can be performed by a chromatography data system (CDS) or a validated Excel spreadsheet. However for trending of the SST data, most CDS systems are not set up for this as standard and, unfortunately, a validated Excel spreadsheet will typically be used.

Perhaps suppliers of chromatography data systems (CDSs)should develop functionality to ensure that these PQ parameters can be derived and trended for each analysis across all chromatographs, individual chromatographs or individual chromatographers to identify potential issues and take preventative action before a problem occurs.

5. The parameters that are not related to peak area or retention can be controlled with a little extra effort using the routine analysis data. Depending on the applications, these parameters should be checked at appropriate intervals, for example, as defined in the classic PQ (1).

The determination of the remaining parameters takes a little bit of extra time (approximately one hour). Depending on the applications, these parameters should be checked at appropriate intervals, for example as defined in the traditional PQ. However, instrumental downtimes and extra manpower are not required.

Conclusions

Continuous PQ is a major improvement compared with the traditional PQ approach and offers several benefits. First, it avoids additional working time as the basic information shown in Table 1 can be collected each time an SST is run. Second, there is the continuous survey of critical instrument parameters which enhances analytical certainty and hence the overall data quality because it provides not only a snapshot of system performance, but an ongoing performance history. We suggest using control charts for documenting chromatograph performance and archiving. This is also the best way to detect a performance drift and to take appropriate counter measures in time before a malfunction occurs. This compares with the time-based assessment of performance using the traditional approach.

Lukas Kaminski is a PhD student at the Institute of Pharmaceutical Chemistry, Technical University Braunschweig, Braunschweig, Germany.

Joachim Ermer is the head of quality control at Sanofi-Aventis Deutschland, Frankfurt am Main, Germany.

Claus Feuäner is the vice president, quality control at Vetter Pharma-Fertigung, Ravnsburg, Germany.

Armin Groh is group leader, quality control at Nycomed, Konstanz, Germany.

Heidemarie Höwer-Fritzen is head of analytical development at W. Spitzner Arzneimittelfabrik, Karlsruhe, Germany.

Peter Link is head of laboratory at LAZ, Tübingen, Germany.

Bernd Renger is principal of Bernd Renger Consulting, Randolfzell, Germany.

Martin Tegtmeier is head pharma production at Schaper & Brümmer, Salzgitter, Germany.

Hermann Wätzig studied pharmacy at the Freie Universität Berlin. He recieved his PhD in 1989 about an HPLC topic, then became lecturer at the Institut für Pharmazie in Würzburg. In 1999 he was appointed to a professorship in pharmaceutical chemistry at the Technische Universität Braunschweig. Since 2001 he has been the chair of the division of pharmaceutical analysis/quality control at the German Pharmaceutical Society. To contact Professor Dr Wätzig e-mail h.waetzig@tu-bs.de

"Questions of Quality" editor Bob McDowall is principal at McDowall Consulting, Bromley, Kent, UK. He is also a member of LCGC Europe's Editorial Advisory Board. Direct correspondence about this column should be addressed to "Questions of Quality", LCGC Europe, 4A Bridgegate Pavillion, Chester Business Park, Wrexham Road, Chester CH4 9QH, UK or e-mail Alasdair Matheson, the editor, at amatheson@advanstar.com

References

1. R.D. McDowall, LCGC Europe, 23(7), 369–374 (2010).

2. R.D. McDowall, LCGC Europe, 23(12), 585–589 (2010).

3. United States Phamacopoeia, General Chapter <1058> Analytical Instrument Qualification.

4. United States Pharmacopoeia, General Chapter <621> Chromatography.

5. European Pharmacopoeia 6.4., Chapter 2.2.46 (2009) Chromatographic Separation Techniques.

6. L. Kaminski et al., J. Pharm. Biomed. Anal., 51(3), 557–564 (2010).

7. L. Kaminski et al., Consensus Paper Efficient and economical HPLC Performance Qualification of the Working Group Drug Quality Control / Pharmaceutical Analytics of the German Pharmaceutical Society (DPhG), in collaboration with the Arbeitsgemeinschaft für Pharmazeutische Verfahrenstechnik (APV; engl.: International Association for Pharmaceutical Technologyhttp://www.pharmchem.tu-bs.de/forschung/waetzig/dokumente/Consensus paper EQ approved by working group AM-K on 6 Nov 2009 gs.pdf.

8. P. Bedson and D. Rudd, Accred. Qual. Assur., 4(1–2), 50–62 (1999).

9. http://www.edqm.eu/medias/fichiers/UPDATED_Annex_1_Qualification_of_HPLC_Equipment.pdf