SC I A. Control of Impurities

This section provides a guide to the pharmacopoeial approach to the control of impurities in medicinal substances and formulated preparations. Other guidance relevant to manufacturers is provided in, for example, ICH guidelines on impurities in new drug substances (Q3A) and new drug preparations (Q3B) and VICH guidelines on impurities in new veterinary drug substances (GL10) and new veterinary medicinal products (GL11).

1. This section relates primarily to totally synthetic organic medicinal substances and those substances obtained by synthetic modification of a naturally-produced precursor. It is not necessarily applicable to other organic substances (e.g., those of plant or animal origin), inorganic substances and excipients.

Certain additional information of specific relevance to impurity control in the formulated preparation monographs of the British Pharmacopoeia is also provided.

2. The control provided by chemical tests limiting the levels of particular impurities or classes of impurities is often augmented by physical tests such as absorbance, specific optical rotation, melting point and clarity and colour of solution and, for a liquid, refractive index, boiling point range and weight per mL.

3. Tests such as sulfated ash and loss on drying are non-specific but they contribute to an assurance of the general quality of the material, the use of good pharmaceutical manufacturing practice in its production, the avoidance of contamination especially by inorganic substances and the removal of volatile solvents. Typical limits are 0.1% for sulfated ash and 0.5% for loss on drying.

4. Tests for purity are intended to provide appropriate limitation of known potential or actual impurities rather than to provide against all possible impurities. The tests are not necessarily designed to detect any adventitious contaminants or adulteration. Material found to contain an impurity not detectable by means of the prescribed tests is not of pharmacopoeial quality if the nature or amount of the impurity found is incompatible with good pharmaceutical practice.

5. Some medicinal substances are mixtures of closely related compounds. Where these components have similar activity they are not usually regarded as impurities and may indeed contribute to the result obtained in the assay. Examples include Erythromycin, Gentamicin Sulfate and Sodium Lauryl Sulfate. It may however be appropriate to control the relative amounts of such components, e.g., gentamicins C1, C1a, C2, C2a and C2b in Gentamicin Sulfate in order to ensure batch to batch consistency for material from one manufacturer and uniformity between supplies of the same substance from different manufacturers.

6. Many medicinal substances already on the market have been made available as racemic mixtures with little or nothing known about the biological activities of the separate isomers. This has been reflected in the monograph in the Pharmacopoeia and a test to show that the substance is the racemic mixture has not usually been included unless it was known that at least one of the separate enantiomers was also available commercially. Nevertheless, with increasing concern by regulatory authorities for substances to be made available as single isomers, tests for enantiomeric composition will become more common. When a medicinal substance is a racemate, an indication is given by means of the graphic formula [see also Supplementary Chapter I K; Stereochemistry].

Related Substances

7. It is usual to include a test for related substances in a monograph for a medicinal substance. These may be manufacturing impurities (intermediates or by-products) or degradation products or both. When preparation of a monograph is initiated the manufacturer is asked to provide information concerning the nature of such impurities, the reason for their presence, the amounts that may be encountered in material prepared under conditions of good pharmaceutical manufacturing practice and the manner in which proportions may vary on storage, together with an indication of the toxicity of any impurities in relation to that of the substance itself. Where there is only one manufacturer of a substance, pharmacopoeial limits are set in the knowledge that the level of impurities in production batches of the substance will have been accepted by the registration authority after a full consideration of the toxicity studies and clinical trials carried out before the granting of a licence. Such studies and trials will have been carried out on material with an impurity profile that is qualitatively and quantitatively similar to that of subsequent production batches. Any subsequent changes to the manufacturing process by the original manufacturer or the introduction of material from another manufacturer utilising a different route of synthesis will be subject to the need to demonstrate essential similarity or to provide equivalent data to the relevant registration authority. In some cases a change in production or source may give rise to impurities that are not adequately controlled by the published pharmacopoeial monograph. Appropriate revision of the monograph will be carried out provided that the pharmacopoeial authority is notified of the need and that it is supplied with the relevant information [see paragraph 28 below].

8. Tests for related substances may be specific or general.

9. Specific tests for named impurities A specific test is included where a particular impurity arising from the manufacturing process or from degradation needs to be limited on grounds of toxicity or for another special reason. Where an impurity is known to be particularly toxic, this is taken into account in setting the limit; for example, a limit of 1 ppm is specified for hydrazine in the monograph for Povidone.

9.1 Such specific tests usually employ a chromatographic or colorimetric comparison with a sample of the named substance, for example, 4-chloroaniline in Chlorhexidine Irrigation Solution and 4-aminophenol in Benorilate Tablets.
9.2 Where a specimen of the impurity is required in the test, this will be made available as a Chemical Reference Substance unless it is known that specimens of the requisite quality can readily be obtained through the usual suppliers of chemical reagents.
9.3 Specific control may be included within a more general test controlling other impurities; an example is 4-epianhydrotetracycline in the monograph for Lymecycline.
9.4 In other cases, an absolute method is more appropriate. Such a test may be for a group of potentially toxic impurities, e.g., polycyclic aromatic hydrocarbons in Liquid Paraffin.
9.5 Sometimes an impurity may be named in the Pharmacopoeia because it is necessary to use a named substance in its control for analytical reasons, such as different response factors in the specified test method. Examples are iminodibenzyl in Imipramine Tablets, and dibenzosuberone in Amitriptyline Tablets. Typical wording is as follows:
Any spot corresponding to [x] in the chromatogram obtained with solution (1) is not more intense than the principal spot in the chromatogram obtained with solution (2). [Solution (1) contains the substance being examined and solution (2) contains a named impurity [x].]

10. General tests for unnamed impurities It is unusual for the Pharmacopoeia to require the absence of a visible spot in a thin-layer chromatogram or the absence of a peak in a liquid chromatogram. Reasons for this include the difficulty of interpreting and defining absence that is a consequence of variations in the sensitivity of a method when performed in different laboratories by different analysts. It is more usual to limit the levels of impurities. This may be done in a simple test by comparison with a spot or peak obtained with a dilute solution of the substance being examined. An example is Methyl Nicotinate:

Any secondary spot in the chromatogram obtained with solution (1) is not more intense than the spot in the chromatogram obtained with solution (2). [Solutions (1) and (2) contain the substance being examined at high and low concentrations respectively.]
10.1 In the absence of evidence that the limit for a particular impurity needs to be set on the basis of its toxicity, control is often provided by a two-level test requiring, say, not more than one related substance at a nominal concentration of up to 0.5% and any others at nominal concentrations of up to 0.1%. The actual limits may be chosen on the basis of batch data for material manufactured in accordance with good pharmaceutical manufacturing practice and will take account of a number of factors, including the dose regimen of the substance and the number of impurities commonly present. An example is Oxetacaine:
Any secondary spot in the chromatogram obtained with solution (1) is not more intense than the spot in the chromatogram obtained with solution (2) (0.5%) and not more than one such spot is more intense than the spot in the chromatogram obtained with solution (3) (0.1%).

Another example, with different limits, is Phenindione:

Any secondary spot in the chromatogram obtained with solution (1) is not more intense than the spot in the chromatogram obtained with solution (2) (2%) and not more than one such spot is more intense than the spot in the chromatogram obtained with solution (3) (0.5%).
10.2 Where it is known that several impurities are likely to be present at significant concentrations, a three-level test may be appropriate. An example is Pentazocine Injection:
By each method of visualisation, in the chromatogram obtained with solution (1) any secondary spot is not more intense than the spot in the chromatogram obtained with solution (2) (1%), not more than one such spot is more intense than the spot in the chromatogram obtained with solution (3) (0.5%) and not more than four such spots are more intense than the spot in the chromatogram obtained with solution (4) (0.25%).
10.3 General tests with an ‘open’ design such as those described above have the great advantage that they provide a means of limiting the levels of related substances that may arise from modified or alternative synthetic routes not in use at the time the test was elaborated. In this context, thin-layer chromatography has the advantage over liquid chromatography and gas chromatography that it allows detection of impurities completely retained or those not retained at all by the stationary phase.

11. Total impurity limits In gas chromatographic and liquid chromatographic tests, it is increasingly common to limit the total areas of peaks due to related substances. In monographs currently in force the limit for the sum is commonly in the range 1 to 2% but these values are applicable at the end of the shelf-life. This procedure is rarely adopted in thin-layer chromatographic tests, because of the semi-quantitative nature of estimating individual spots and resulting imprecision in expression of results for the totals. This apparent drawback to the use of thin-layer chromatography is largely overcome by means of two- and three-level tests as described in paragraphs 10.1 and 10.2, above.

12. In response to requests from users of the Pharmacopoeia, statements of the approximate real or nominal levels of impurities controlled by tests for impurities have been widely introduced, where appropriate, for information [see paragraphs 14 and 15 below]. Conformity with the requirements will still be determined on the basis of compliance or otherwise with the stated test.

Test design and expression of limits

13. For identified impurities, several aspects are taken into account in designing the test. These include the nature of the impurity, its toxicity and the levels likely to be found in routine production. Analytical considerations such as the correction factor (defined in paragraph 16) for the impurity and practical issues such as availability of the impurity as a reference material or reagent also influence the test design.

14. If a major and/or toxic impurity in a material is known to have a significantly different response (more than ±25%) from that of the main peak in the substance being examined in the conditions of the test, the preferred manner of limiting this impurity is to use a reference substance of the impurity. If this is not possible, a reference solution of the substance being examined containing a known amount of the impurity may be used. Using either of these approaches, the concentration limit indicated in the monograph for information in parentheses expresses the approximate limit as a real percentage of the impurity in question. This is also referred to as % w/w. When neither of these approaches is possible, a dilution of the solution of the substance being examined may be used as a reference solution. This approach is also commonly used in tests where an impurity that is known (but not named within the test) has a response within ±25% of that of the main peak in the substance being examined.

15. Unless explicitly stated otherwise, an indication of the approximate concentration limit provided in any test where a dilution of the substance being examined is used as the reference solution should be interpreted as an expression in terms of a nominal percentage of the substance being examined (in accordance with the General Notices) rather than as a real percentage of the impurity.

16. No reference is made in a test to a correction factor for an identified impurity unless unavoidable. When used, the term is defined as follows:

The correction factor (1/k) is a relative term, being the reciprocal of the peak response of equal concentrations of one substance relative to that of another in the conditions described in the test.

In the context of a related substances test where a correction factor is quoted for an impurity, unless otherwise stated, this is the expected correction for that impurity in relation to a response of unity for the substance being examined. The way in which a correction factor is to be used in any subsequent calculation is stated in the monograph.

17. Correction factors of less than 0.2 or more than 5 are not generally used. If the difference between the response of an impurity and that of the substance being examined is outside these limits, a different method of determination, such as a different detection wavelength (λ) or a different method of visualisation, is used.

18. For a correction factor quoted in a pharmacopoeial test, the following points are observed:

(a) the peak to which the factor applies is identified unambiguously, recognising the difficulties associated with such identification in chromatograms showing peaks of similar retention times [the use of a sample containing an unquantified amount of the relevant impurity can sometimes be used to assist such identification],
(b) the correction factor quoted is a confirmed value and preferably is based on relative peak areas of equal concentrations of the impurity and the sample under the conditions of the test [alternatively the reciprocal of an absolute figure based on the A(1%, 1 cm) for the detection wavelength used (or similar for other methods of detection) of the impurity and the sample may be used],
(c) the correction factor is stated in a way that does not imply an unrealistic or misleading degree of precision [declaration to several significant figures is not meaningful when placed in context of the nature of the test and the amount of the impurity likely to be present (typically, less than 0.5%) (e.g., 1.4, 2.5, 3.0, 0.5) or 1 decimal place (e.g., 1.4, 7.5) are generally appropriate],
(d) older monographs of the BP may contain the term response factor. The term will be revised to correction factor as these monographs are updated.

19. Identification of peaks is generally not based on absolute retention times since these may be too ‘system dependent’; however, advice such as ‘the principal peak has a retention time of about x minutes’ may be given. In some cases, for example, where a simple chromatogram is expected to show a limited number of impurities, an expected relative retention may be given to designate impurities. In other cases where potential impurities have similar retention, a sample containing the components of interest and a sample chromatogram may be provided.

20. Unidentified impurities may be limited by reference to a dilution of the solution of the substance being examined used as a reference solution together with an open design of statement limiting ‘any’ or ‘any other’ secondary peak or spot. Such a reference solution may be used in addition to those containing named impurities (any other secondary peak/spot) or, in some simple tests, control of unspecified and specified (but unnamed) impurities may be exerted by means of a comparison between the sample solution and a dilution of this solution (any secondary peak/spot).

21. An indication of the approximate limit concentration for an unidentified impurity would only be given in terms of a nominal percentage of the substance being examined (see paragraph 15) since no assumption can be made about the response of an unidentified impurity. While such nominal limits are not quantitatively fully transparent and caution is needed in attempting to use them to set a total impurity limit, their use is less misleading than quoting an arbitrary/assumed correction factor of 1 for an unidentified impurity.

Formulated Preparations

22. Many monographs for formulated preparations in the British Pharmacopoeia and the British Pharmacopoeia (Veterinary) also include tests for impurities. In general, wherever possible a test for impurities based on that in the monograph for the active ingredient is included with any necessary modification.

23. Wider limits and/or additional controls may be required for impurities arising on manufacture or storage of the dosage form.

24. Tests for impurities in monographs for formulated preparation are used to control not only degradation products but also by-products of the synthetic route used for manufacture of the active ingredient. It has been argued, for example in the ICH guideline Q3B, that by-products of synthesis have been controlled already during examination of the substance before formulation and that further testing for these impurities is unnecessary. Clearly it would be repetitious and wasteful of resources for tests, often complex in nature, to be repeated routinely simply to demonstrate acceptably low levels of impurities that could arise only during synthesis (as opposed to degradation) of the active ingredient. However, this information is available only to those who know the detailed attributes of the active raw material that has been used. For an analyst who has access only to the dosage form, the profile of synthesis-related impurities offers one means of establishing whether or not the dosage form has been prepared from an active ingredient of pharmacopoeial quality. It is for this reason that such tests are included in British Pharmacopoeia and British Pharmacopoeia (Veterinary) monographs for formulated preparations [see paragraph 3 of General considerations in the introduction to Supplementary Chapter I].

Current and future developments

25. Transparency A statement giving the identities of impurities that are known to be limited by the specifications is being added to appropriate monographs for medicinal substances and formulated preparations. For example the monograph for Diamorphine Hydrochloride contains the following information at the end of the monograph:

IMPURITIES The impurity limited by the this monograph is: 6-O-acetylmorphone

This increase in the transparency of pharmacopoeial specifications is of assistance to licensing authorities and others when considering whether the standards in the monograph are appropriate to a new source of supply. It is emphasised that other, unnamed impurities may also be limited. The Commission is actively seeking information that will allow the statements to be extended in future.

26. System suitability In chromatographic tests increasing use is being made of system suitability tests to enable the analyst to confirm that performance of the chosen column or plate is satisfactory under the chosen conditions. In liquid chromatographic and gas chromatographic tests, peak separation between impurities and the substance being examined is generally considered to offer the best indication of performance of the system.

27. Residual solvents A general test for residual solvents has been included in the European Pharmacopoeia (Appendix VIII L) together with guidelines on its application (Supplementary Chapter IV D). At present, a test for solvent residues is included in a specific monograph only where variation in levels of known solvents requires control, e.g., methanol in Gentamicin Sulfate.

28. The British Pharmacopoeia Commission welcomes suggestions for improving the monographs. In particular, where it is found that significant impurities are not controlled by the monograph, the Commission would be glad to receive details of validated methods that can be considered for adoption.