This Annex contains detailed advice on a number of miscellaneous points that are likely to arise during the evaluation process. It is not a complete checklist of points to consider. Advice already included in the main text of the manual is not repeated here. The glossary, abbreviations, references and annexes are as those appearing in the main text.

When consulting this Annex, DRAs may fin it useful to refer to the decision tree for marketing authorizations using WHO-type product certificates appearing in Part IV of the manual, under “Evaluation of data on quality”.

1. WHO-type certificates of pharmaceutical product (WHO-type certificates)

2. Quality control tests and limits

3. Identity of active ingredient

4. Impurities in the API

5. Assay of APIs

6. Physicochemical properties of APIs

7. Quality of excipients

8. Assays of active ingredients in finished products

9. Testing for impurities in the finished product

10. Tests on the finished product

11. Stability

12. Assay methodology

13. Container labelling

14. Trade names

15. Product information (PI)

16. Interchangeability

17. Papers to be presented to the Expert advisory body for a decision on a multisource pharmaceutical product

18. Variations

19. Periodic reviews of marketing authorizations

1. WHO-type Certificates of Pharmaceutical Product (WHO-type certificates)

It should be noted that “Requests for provision of certificates offering more limited attestations - for instance, that the manufacturer complies with GMP or that the product is authorized for “free sale” within the country of export - are discouraged” (Annex 2, section 3.2).

What the certificate means

If the exporting country has not authorized the product to be placed on its own market the WHO-type certificate amounts to certification of the manufacturing standard at the site in question. Neither an officially approved product information nor an evaluation report will be available.

If the exporting country has authorized the product to be placed on its own market, the WHO-type certificate, in addition to certifying the manufacturing standard at the site in question, implies that the country issuing the certificate accepts that the product is of adequate quality, safety and efficacy to remain on its own market. Therefore an officially approved product information document, such as the European summary of product characteristics, should normally accompany the certificate (Annex 2, Appendix 1, 2A.5). If this document is not available, then the product is probably “grandfathered” (see glossary) and has not been fully evaluated. Some countries also issue a summary basis of approval that summarizes the technical basis on which the product has been authorized.. When this is available, it should also be attached to the certificate (Annex 2, Appendix 1, 2A.4).

Where a WHO-type certificate is issued and the product has marketing authorization in the issuing country, it is not necessary to seek separate certification of the sites of manufacture of the API.

If the product is not identical to that in the issuing country, the model application form (Annex 6) allows applicants to list any differences (under “Certification by a responsible person in the applicant company”) and to justify the differences (under “justification for any differences to the product in the country or countries issuing the submitted WHO-type certificates”). DRAs must then decide whether the differences are minor and have been adequately justified, and consequently whether the WHO-type certificate is relevant.

More than one country of manufacture

Manufacture of a single product may occur in more than one country, for example in the following circumstances.

(a) Different phases of manufacture may be conducted in different countries. For example, a batch of tablets may be prepared in bulk in country A, and packaged and subjected to quality control testing in country B.

(b) The product may be fully manufactured in both countries A and B, and the applicant wishes to obtain approval to use both sites of manufacture.

The WHO-type certificate should normally be prepared by the country that directly exports the product to the importing country, i.e. country B in example (a) above. If the product has marketing authorization in this exporting country and the product to be imported is identical to that authorized, the sites of manufacture in third countries will have been approved by the exporting country. In that case, no further review of GMP in third countries is necessary. Indeed, some exporting countries might voluntarily attach a GMP certificate for sites in other countries (Annex 2, Appendix 1, explanatory note 16).

If the exporting country has not authorized the product to be placed on its own market, the importing country may seek WHO-type certificates from more than one country, i.e. countries A and B in the above examples.

In example (b), both countries can separately be asked for a WHO-type certificate.

Currency and validity

The WHO Certification Scheme recommends that the exporting country prepare a separate certificate to be made available for each importing country, and that certificates should reflect any variations made to marketing authorizations (Annex 2, section 3.5). Consequently, certificates should always be current (recent) and each DRA may wish to define what it considers to be “current”. If there is any doubt as to the status or validity of a certificate, the DRA should request an identical copy, clearly marked as duplicate, directly from the certifying authority (Annex 2, section 4.9).

If there is doubt that the authority issuing a WHO-type certificate satisfies the prerequisites for the scheme, there is a provision (see Annex 2, section 2.5) which reads:

“Each Member State assumes the responsibility to determine, through a process of self-evaluation, whether it satisfies these prerequisites. The Scheme contains no provision for external inspection or assessment under any circumstances, either of a competent national authority or of a manufacturing facility. However, should a Member State so wish, it can approach WHO, or a well recognized drug regulatory authority, occasionally to delegate consultants to act as advisers in the course of both national inspections and inspector training activities.”

DRAs may wish to prepare a list of Member States from whom WHO-type certificates are considered acceptable, perhaps in conjunction with other authorities with which it collaborates. Any such list should be kept up to date.

The source of the certificate

All requests for WHO-type certificates should be channelled through the agent (applicant) in the importing country (Annex 2, section 3.6).

2. Quality control tests and limits

Tests and limits for the API and the finished product should, as a minimum, comply with relevant pharmacopoeial requirements, including general requirements such as those in the International Pharmacopoeia (IP) for parenteral preparations, capsules, and so forth. Additional tests and limits may be necessary, for example for the dissolution rate of solid dosage forms of water-insoluble active ingredients when the pharmacopoeia does not include a test. Examples might be amoxicillin capsules and primidone tablets.

Pharmacopoeias sometimes do not set limits because they are not necessary for all products. An example is the particle size of a water-insoluble API. It is not appropriate for the pharmacopoeia to set a limit because suitable limits may vary with the use to which the starting material is put. If the API is present in the product as solid particles and does not go into solution during manufacture, a limit on particle size would normally be appropriate. But the same material may be used to manufacture a solution, for which a particle size would be irrelevant. For example haloperidol has a solubility in water of less than 0.001% (BP 1998). It would be appropriate to include limits on the particle size range of API used to manufacture haloperidol tablets because the drug is present as a solid and does not go into solution during manufacture. The particle size range may well effect bioavailability. However the same API is in solution in Strong Haloperidol Solution BP, and particle size limits would not be needed for API used in its manufacture.

When a product is given marketing authorization, and the specifications include compliance with a pharmacopoeial monograph, it is usual to allow automatic updates as new editions of the same monograph are published in the same pharmacopoeia. Any tests additional to those of the pharmacopoeia must continue to be applied, such as particle size and limit tests for impurities relevant to a particular synthetic route.

3. Identity of active ingredient

There should be a test (or tests) in quality control of the API to ensure that it is the intended substance. Some active ingredients are available either as a single enantiomer, as a racemate (e.g. dexamfetamine and amfetamine), or as multiple isomers (e.g. labetalol), in which case it is necessary to ensure that the active ingredient is the same from batch to batch (in the case of labetalol, the ratio of the two pairs of racemates should be controlled). The salt form and/or state of solvation may also need to be controlled, e.g. amoxicillin sodium versus amoxicillin trihydrate. Control of the ratio of (E)- and (Z)- isomers (cis- and trans-) may be necessary for older active ingredients such as clomifene.

Test procedures for identity should be validated (14).

4. Impurities in the API

For many, but not all, existing multisource pharmaceutical products, data on safety of the API are available from toxicological and clinical studies. At the very least, the fact that the comparator has been marketed for some time is assumed to provide de facto evidence of lack of major toxicity. This is an assumption and may not be valid in all circumstances, such as when the product is marketed only in countries where there is no effective adverse reaction monitoring system.

Accepting this limitation, to be able to consider that information on the safety and efficacy of established products is applicable to new multisource pharmaceutical products the impurity content of the new brand must be no greater than that of the comparator. Quantitative limits on individual impurities in the API must be no greater than those for the comparator, and ideally there should be no new impurities in the new brand. If these conditions do not hold, then strictly speaking new safety studies should be considered.

To make the necessary comparison of impurity content, ideally the limits on impurities in the comparator product should be known, together with information as to what impurities are actually found. The impurities actually found may not be detected or controlled by the methods and limits in pharmacopoeias, so that additional controls are sometimes necessary. Unfortunately, information on impurities actually found in the comparator is often not available to the DRA because the comparator began to be marketed before premarket assessment was introduced. To seek information on impurities in the comparator in order to be able to assess the new brand may not be considered equitable in all jurisdictions, but some may consider it justifiable on grounds of public safety. A complete assessment of the appropriateness of impurity limits also requires information on the route of synthesis, and the opinion of an experienced organic chemist as to what by-products are likely following synthesis by that particular route. It is then possible to assess whether the possible by-products are likely to be detected by the proposed methodology for impurity testing. There should be a similar assessment of likely routes of degradation and whether the likely degradation products will be detected by the proposed impurity testing.

By far the most desirable situation is for the API to be controlled by tests and limits that have been shown to be suitable for the particular synthetic route and site of synthesis. This is achieved in countries that have well resourced agencies. However, it is recognized that this outcome may not at present be achievable by all DRAs, for two reasons:

(1) The DRA may decide that its resources do not permit assessments in this detail.

(2) The reality at the present time is that batches of pharmaceutical starting materials pass through multiple hands, so that their origin and route of synthesis becomes obscure. Purchase from intermediaries is not recommended (13), but it is recognized that it occurs in some circumstances, for example when price is a factor in determining accessibility. It is emphasized that this is not a desirable situation.

In both these situations, smaller DRAs may decide to rely upon compliance of an API with the impurity requirements of a pharmacopoeia. In that case, the following is recommended:

X The latest edition of the pharmacopoeia should be specified so as to take advantage of the most up-to-date information available on important properties of starting materials, especially in relation to contaminants, and on testing methodology.

X Each batch of API should be fully tested for compliance with specifications by the finished product manufacturer.

X A confirmatory certificate of analysis from the supplier should be available for each batch, and the certificate should be satisfactory as defined in WHO’s GMP guidelines (13 p.57).

The WHO draft guideline for the certification of APIs, which seeks GMP certification of all sites of manufacture of APIs, will, once enforced worldwide, be a major step in achieving better control.

5. Assay of APIs

Assay procedures for APIs should if possible be specific for the API in the presence of impurities. However, this may not be essential if the remainder of the monograph fully controls the content of all types of impurity, including as necessary degradation products, synthetic by-products, heavy metals and non-combustible contamination.

6. Physicochemical properties of APIs

It may be necessary to control the physical properties of APIs, for example (as necessary) particle size and polymorphic form.

7. Quality of excipients

Information on the route of synthesis is less readily obtainable for excipients. Consequently, for reasons similar to those outlined for APIs, it is recommended that excipients comply with the latest edition of a pharmacopoeia, and that each batch obtained via an intermediary rather than directly from the producer be fully tested by the finished product manufacturer before use.

Only colours listed in the European Union’s “List of permitted food colours”, FDA’s “Inactive ingredient guide” or the “Japanese Pharmaceutical Excipients” list (19-22) should be permitted.

8. Assays of active ingredients in finished products

Assay methodology in pharmacopoeias must be validated for each formulation. An example of what can go wrong is a formulation which, when processed by the pharmacopoeial method, produces a opalescent solution at a point at which ultra violet absorption is to be measured.

9. Testing for impurities in the finished product

If the DRA is convinced that a company is adequately ensuring the absence of significant synthetic by-products in the API, the company need not test for those by-products in the finished product. It is necessary, however, to test for impurities that may be breakdown products of the active ingredient.

The DRA may wish to test for synthetic by-products in the finished product as a part of its regulatory role in monitoring quality.

Test procedures for detecting and/or quantitating impurities should be validated (14).

10. Tests on the finished product

In addition to tests for identity and assay of API and for absence of impurities, it is often appropriate to conduct tests on the physicochemical properties of finished products.

Examples include particulate contamination of solutions for injection, lack of phase separation of emulsified injectables, and the particle size of suspensions.

11. Stability

Products must maintain their quality for the full length of the shelf-life when stored in the final marketing pack under any storage conditions permitted by the label. The permitted storage conditions must be attainable in local circumstances, for example in warehouses and distribution systems.

Products must also be stable when processed according to labelled instructions, such as during reconstitution or dilution.

12. Assay methodology

All assay methodology should have been validated. A number of suitable guidelines are available concerning validation of analytical procedures, for example those of ICH (14).

13. Container labelling

In line with GMP guidelines (13), container labels should comply with national legislation and should include at least all of the following information:

X The name of the product;

X A list of the active ingredients (using INNs if applicable; 23), showing the amount of each present, and a statement of the net contents, e.g. number of dosage units, weight or volume;

X The batch number assigned by the manufacturer;

X The expiry date in an uncoded form;

X Any special storage conditions or handling precautions that may be necessary;

X Directions for use, and any warnings or precautions that may be necessary;

X The name and address of the manufacturer or the company or the person responsible for placing the product on the market.

Some concession is appropriate for small containers of less than or equal to 10 ml capacity that are marketed in an outer pack such as a carton. If the outer pack bears all the required information, the immediate container need only contain items (b), (c), (d), (g) - or a logo that unambiguously identifies the company - and the name of the dosage form or the route of administration (see section on “Container labelling” in Annex 6).

14. Trade names

DRAs may or may not have the legal authority to approve proposed trade names, depending on national legislation. However, even if this function is not written explicitly into legislation, DRAs may nevertheless have the authority to disallow a trade name on the grounds that it constitutes a safety hazard or is misleading. Examples include names that imply unapproved claims, such as an unapproved:

X indication, e.g. “Bacta.....”, or “Hypno.....”;
X patient population, e.g. “Pregna.....”, or “Infa.....”; or
X dosage regimen, e.g. “Uni.....”, or “Mono.....”.

If a product’s principal indication(s) is/are changed or there is a change in the API, it may be necessary to consider whether the trade name is still appropriate. For example if a product is indicated for angina and has the trade name “Cardioproduct”, and later the principal indication is changed to claudication, the trade name would no longer be appropriate.

Names that are based on generic names, especially INNs, are also to be discouraged. The Forty-sixth World Health Assembly in May 1993 requested Member States “to develop policy guidelines on the use and protection of international nonproprietary names, and to discourage the use of names derived from INNs (23), and particularly names including established INN stems as trade marks” (resolution WHA46.19). The request was based on a statement by the WHO Expert Committee on the Use of Essential Drugs (5) that use of a trade mark based on an INN, and particularly one based on an INN stem, “can frustrate the rational selection of further INNs for related substances, and it will ultimately compromise the safety of patients by promoting confusion in drug nomenclature”.

15. Product information

Organization and content

Product information should normally be based on WHO’s “Sample Product Information Sheet” (Annex 8, Appendix). Where a monograph exists in either WHO Model Prescribing Information or the model formulary, product information should usually include all the information in that monograph. Information on adverse reactions should conform to CIOMS guidelines (24). Product information should not allude to unapproved (off-label, undocumented) indications, nor should it be speculative. Favourable comparisons with other products should be avoided and only permitted if there is a sound clinical basis, documented in the literature, as such claims may later be used in promotion (see below).

A second, shorter version of product information may be appropriate for use in primary health care, and for this WHO’s series entitled Model Prescribing Information (25) would be a suitable format. The product information itself may then be considered as a reference text which provides more in-depth information.

All promotion should be based on product information

It should be a condition of marketing authorization that all promotion of the product be consistent with the agreed product information and with WHO’s guidelines on ethical criteria for Medicinal drug promotion (Annex 8, Appendix). Monitoring by a DRA of all promotion would be resource- intensive and is beyond the capacity of even well resourced authorities. Therefore the usual means of control is a system of random audits with suitable sanctions for breaches.

Note that the World Health Assembly has recently adopted a resolution on “Cross-border advertising, promotion and sale of medical products using the Internet” (26).

Product information should be consistent with the container labelling and with the product

The product information should be consistent with both the nature of the product and with container labelling. Examples of inconsistencies might include:

X The container label states that the product contains a particular steroid whereas the product information states that it contains the acetate ester of the steroid;

X The container label states that the product should be diluted before infusion, whereas the product information refers only to administration intravenously;

X The product information states that the product is not to be used in children, whereas the brand name on the container implies use in children;

X The product information refers to a dose of 25 mg when the product is a 50 mg capsule;

X The product information recommends a course of 5 ml twice daily for three days when the product contains only 25 ml of liquid.

Comparison with existing product information

It is not usually acceptable to have two marketing authorization holders of the same multisource product on the same market when their product information are inconsistent. Normally the evaluator should compare the product information for the new product with that of the comparator brand, carefully consider any differences, and determine whether amendments are necessary. In general, one or both of the product information will have to be amended so that they are not inconsistent and match as close as possible, although the wording need not be identical. Sources of information for this exercise should ideally be independent, reliable and complete, which may involve a literature search and review. In practice, reference is usually made to standard textbooks of medicine and pharmacology, national compendium of approved product information, and any existing national formulary. If a product information is available for the same product in another country, particularly one with a well resourced DRA, a copy should be sought for comparison. When a new API has recently come out of patent (for example in the last five years), the innovator’s product information is normally a suitable starting point for a multisource product information.

Differences in product information may have to be tolerated in certain situations, particularly where local legislation allows new uses to be patented (new indications in the case of pharmaceuticals) or where market exclusivity arrangements apply.

It should be noted that not all existing product information meet current standards. Product information for older products in particular need careful scrutiny and in some cases may need to be substantially rewritten.

The advice of the expert advisory body is useful in finalizing product information, particularly to take advantage of expertise relating to local circumstances and endemic diseases.

Patient information or package inserts

Any patient information or package inserts should be consistent with the agreed product information and should comply with guidelines in WHO’s Ethical criteria for medicinal drug promotion (see Annex 8).

16. Interchangeability

WHO’s guideline on “Multisource (generic) pharmaceutical products” (Annex 3, Part One, section 2) mentions a number of features that are important to interchangeability:

X compliance with appropriate quality standards and at least compliance with relevant pharmacopoeial standards;

X stability;

X possible differences in sensitizing potential due to the use of different excipients;

X therapeutic equivalence in terms of, as appropriate, bioequivalence, pharmacodynamic studies, clinical studies or in vitro dissolution rate;

X product information and labelling.

Each of these features will be briefly examined in turn.

Quality

Various options for control of quality have been outlined earlier in the manual. There is no objection to a new multisource pharmaceutical product complying with a higher standard of quality than an existing product, but any difference in specifications should be considered carefully in terms of what effect it might have on interchangeability. For example, a faster in vitro dissolution rate may entail a difference in plasma concentration/time profile, which in turn could have therapeutic implications depending on the active ingredient and its indications. On the other hand, a narrower limit for the same impurities, or for assay of the active ingredient, implies a product of higher quality and it is difficult to envisage an adverse clinical consequence.

Stability

Stability should be examined as a part of the premarket evaluation of quality. Data should be available as described in WHO guidelines (see Annex 11). Different shelf-lives are acceptable for different brands of pharmaceutical equivalents as long as each is clearly labelled with the correct expiry date appropriate to that product and batch.

Different excipients

Different brands usually contain different excipients. Idiosyncratic reactions to excipients are difficult to eliminate completely, but can be minimized by avoiding excipients known to have a high incidence of such reactions. Information may be available in the published literature (19-22, 27 and 31). Safety data should be sought for new excipients. Well resourced agencies may have already completed a scientific report on data on new excipients.

Therapeutic equivalence

Pharmaceutical equivalents on the same market should be interchangeable as defined in the WHO guidelines (see Annex 3). Interchangeability must therefore be considered during evaluation of applications for new products that are pharmaceutical equivalents of well established drugs on the same market. If there is a product on the same market that has been fully evaluated in terms of quality, safety and efficacy (often the innovator), it will be the most relevant comparator. However, some manufacturers market products with different characteristics in different countries (12). If the study of equivalence has compared the new product with a brand available on another market, the applicant should provide evidence which convincingly demonstrates the relevance of the study.

For some products, it is not essential for pharmaceutical equivalents to have exactly the same rate and extent of bioavailability. The definition of “bioequivalence” in Annex 3 allows for this:

Two pharmaceutical products are bioequivalent if they are pharmaceutically equivalent and their bioavailabilities (rate and extent of availability), after administration in the same molar dose, are similar to such a degree that their effects can be expected to be essentially the same.” [emphasis added]

When studies show that the bioavailability of a new product is not the same as that of the existing pharmaceutical equivalent, it will usually be appropriate to seek the advice of the expert advisory body as to whether the products can be designated bioequivalent.

Bioequivalence is difficult to establish for some products, such as slow-release formulations, and high precision is required for others, such as anticoagulants and anticonvulsants, which have a steep dose-response curve and/or a low therapeutic index. Some DRAs simply declare this type of product “not interchangeable”. But for countries with scarce resources and limited distribution facilities, it may be unrealistic to expect that the same brand will always be available for a particular patient. Advice from primary health care workers and pharmacists responsible for distribution of pharmaceuticals (e.g. those who are members of the expert advisory body) is important on this point. It is not an option simply to stipulate that only those products distributed to remote areas must be interchangeable because the appropriate studies would then be even less affordable in the more limited market. It may be appropriate to seek an opinion from the expert advisory body on the risks and benefits of permitting multiple brands to be made available in this situation before the DRA makes a decision.

In evaluating bioequivalence, evaluators should bear in mind the possibility that the existing (comparator) product may not have adequate bioavailability. If the area under the plasma concentration-time curve of the new brand is greater than that of the comparator, that may be either because the new product uses a different technology (such as absorption enhancers) or because the comparator is of poor quality. To achieve interchangeability, the DRA’s options in the latter case are either to cancel the marketing authorization of the comparator or to require that generics mimic the comparator’s plasma concentration-time curve. The decision should normally be in favour of the comparator unless it is clearly a poor product. (See also “Non-interchangeability” below)

It should be noted that a faster rate of absorption with the same area under the curve does not necessarily mean a better product. For example, fast absorption of carbamazepine and nifedipine has been shown to result in side-effects in at least some patients (29, 30). There is no substitute for careful and informed consideration of each case, after having fully reviewed the literature and consulted appropriate experts.

In the event that a DRA decides to cancel the marketing authorization of the existing product, it is essential that the change-over be carefully planned. Health professionals should be advised in advance so that they can take appropriate steps to monitor patients during the change-over period. The sponsor of the newly authorized product should issue suitable letters to prescribers and pharmacists. The text of all such letters should be agreed in advance with the DRA.

If, in assessing an application for a new product, the DRA intends to rely on a report or decision made by another authority, the question of interchangeability with a suitable locally marketed comparator must be considered. Therefore, even if the authority providing a report or WHO-type certificate considers that interchangeability has been adequately addressed, it should be reconsidered by the local DRA (i.e. for each market) because the comparator may be different. If the applicant can demonstrate to the DRA’s satisfaction that the comparator is identical in the two countries, that may be sufficient assurance of bioequivalence. If the products are similar but not identical, the applicant should provide argument and/or data to support interchangeability with the local comparator.

Product information and container labelling

As noted above, the product information of new products should be consistent with that of any existing brand.

It should be clear from the labelling of the new and existing brands that these are generic versions of the same multisource pharmaceutical product. This is achieved by use of generic names (INNs) in Product information and container labelling.

Non-interchangeability

If a study shows that two multisource products are not interchangeable, the DRA has three options:

1. Not to allow marketing of both products simultaneously;
2. To consider the second product to be a different dosage form; or
3. To allow simultaneous marketing of non-interchangeable products.

In the first case, the DRA would normally reject the second application. But, if the data show that the product already-marketed is of poor quality, there may occasionally be a case for approving the new product and withdrawing the authorization for the existing product. The change-over would have to be handled carefully, ensuring that the two products are not available simultaneously and warning prescribers and pharmacists in advance that patients must be monitored carefully during the transition (normally by means of letters to prescribers and pharmacists).

In the second case, the difference between the two products is so significant that they can be considered different dosage forms. This is an exceptional situation and should be avoided if possible because there will always be a chance of confusion between the products. If, for example, the new product has the same extent of absorption (area under the curve) in a bioequivalence study but is much more rapidly absorbed, it may be appropriate to consider it a “rapid absorption” or “prompt release” product. Clinical evidence would be needed to show that the product is safe and efficacious, for example that the higher maximum plasma concentration does not cause adverse effects, and that the (probably) lower trough plasma concentration does not lead to reduced efficacy. It would be essential to ensure that prescribers and pharmacists are aware that the products are different, and the new product should be given a different generic name, such as “..... rapid release tablets”. The holder of the new market authorization would be responsible for ensuring that the market launch makes the difference very clear.

The third case is undesirable, particularly for countries with scarce resources and limited distribution facilities (see above under “Therapeutic equivalence”).

17. Papers to be presented to the expert advisory body for a decision on a multisource pharmaceutical product

In considering an application for a new multisource pharmaceutical product, the expert advisory body must have sufficient information before it to be able to make a decision that is independent of both the applicant and the DRA. Papers prepared by the DRA should be objective rather than subjective, and reasons should be given for all recommendations. At least the following information should appear in papers presented to the expert advisory body.

X The company’s letter of application.

X The evaluator’s summary of the data set and his/her recommendations, with reasons. Include or attach at least:

X Copies of the specifications of the API and the finished product, with notes as to any aspects the evaluator finds unacceptable;

X A summary of stability data, the applicant’s recommended shelf-life and the evaluator’s recommended shelf-life;

X A summary and evaluation of the evidence as to interchangeability; the evaluator should state what calculations, if any, he or she has duplicated during the evaluation;

X A copy of the proposed product information with the evaluator’s comments and recommendations.

X Details of sites of manufacture and certification of GMP. The evaluator should state whether he or she believes these to be satisfactory.

X The results of any quality control laboratory testing of the product and any inspection of local manufacturing activities, such as packaging.

18. Variations

As a matter of principle, applications to vary an existing marketing authorization should not be discouraged because they are often intended to improve quality (such as stability, batch-to-batch consistency, and analytical methodology) or product information (for example, updates to information on adverse reactions). Prior advice of such changes is to be encouraged; variations should therefore be processed as quickly as possible. A balance must be maintained between not placing the company at a disadvantage for making the application and nevertheless ensuring that the change has been adequately validated.

Variations that are submitted for approval should not automatically trigger a request for a full data set or a full evaluation. Data requirements, and the evaluation, should be limited to those needed to validate the variation. Changes that require a new marketing authorization number may not necessarily need to be supported by a full data set. For example, a change of or additional trade name and/or market authorization holder may require a new marketing authorization number, but would not require data provided that the new authorization holder certifies that (a) the product is identical in every respect to the produce that is currently authorized and (b) the holder of the old marketing authorization agrees to the change.

19. Periodic reviews of marketing authorizations

The intention of a periodic review is to consolidate the information held by the DRA.

Any pharmaceutical (manufacturing, quality control, shelf-life, container labelling, etc.) variations/changes made to the product since the first marketing authorization or the last periodic review should have been either changes for which prior approval has been obtained or those not requiring prior approval. Any safety updates should have been made when the relevant safety information became sufficiently well established as defined by CIOMS (24, p.39). GMP practices should have been maintained and, as necessary, improved at all sites of manufacture.

The following information should be submitted with each periodic review.

1. Updated WHO-type certificates for each foreign site of manufacture of the finished product (see Annex 2, section 3.5). When the proposed WHO guidelines for certification of APIs have been finalized, it may also be necessary to supply WHO-type API certificates for sites of manufacture of APIs.

2. Updated certificates of GMP for domestic manufacturing sites.

3. A chronological list of all:

(a) Approved pharmaceutical variations of any type, and all amendments made to product information since the first marketing authorization or the last periodic review. The dates of approval and file reference for each change should be provided; and

(b) A set of current specifications for the API(s) and finished product, including test methods in sufficient detail for them to be replicated by another laboratory.

The periodic review should not be used as an opportunity to seek authorization for new variations. These should be subject to separate application. Nor should it be a means of regularizing unauthorized changes to pharmaceutical information or to product information.