Sonderdruck
`
`die pharmazeutische industrie
`Reprint
`ECV - Editio Cantor Veriag ~ Pharm. Ind. 57, 5, 362-369 (1995)
`
`
`1
`
`FIP Guidelines for Dissolution Testing
`of Solid Oral Products
`
`Joint Report of the Section for Official Laboratories and Medicines Control Services
`and the Section of Industrial Pharmacists of the FIP (Final Draft, 1995)
`
`Publication of the new
`Dissolution Guideline (Draft):
`FIP is asking for Comments
`
`In 1981 FIP published a Guidelines
`for Dissolution Testing of Solid Oral
`Products as a joint report of the Sec-
`tion for Official Laboratories and
`Medicines Control Services and the
`Section of Industrial Pharmacists.
`These guidelines were intended as
`suggestions primarily directed to
`compendial committees, working on
`the introduction of dissolution/re-
`lease tests for the respective pharma-
`copoeias.
`During the past decade there have
`been many developments. Biophar-
`maceutics has attracted much scien-
`tifcal as well as political interest.
`Dissolution test methodology has
`been introduced to many pharmaco-
`poeias and a number of regulations
`and guidelines on bioavailability, bio-
`equivalence and in vitro dissolution
`testing have been issued at national
`and international level.
`
`The joint working group on dissolu-
`tion of the two FIP sections therefore
`decided to establish a new dissolution
`guideline,
`taking all these develop-
`ments into consideration but adding
`proposals for further harmonization
`and for definitions and procedures
`which are not yet covered by interna-
`tional recommendations.
`
`The following guideline is the “final
`draft” version elaborated by the FIP
`working group with contributions
`from .1 M. Aiache (Clermont Fer-
`rant), H. Blume (Eschborn), H. D.
`Friedel (Leverkusen), L. T. Grady
`(Rockville),
`V. Gray (Rockville),
`B. Hubert,
`(Rockville),
`.I. Kramer
`(Eschborn),
`I. McGilveray
`(0t-
`tawa), E Langenbucher
`(Basel),
`L. Leeson (Montville), L. Lesko
`(Rockville), H. Mo'ller (Frankfurt),
`S. Qureshi
`(Ottawa), V.
`1’. Shah
`(Rockville), M. Siewert
`(Frank-
`furt), R Siiverkriip (Bonn),
`.1 0.
`Waltersson (Uppsala), E. Wirbitzki
`(Frankfurt).
`
`The FIP working group decided to
`publish this final draft version to give
`colleagues from universities, drug au-
`thorities, pharmacopoeias and the
`pharmaceutical industry the oppor-
`tunity to contribute with their com-
`ments to further improvement of the
`guideline text prior to publication of
`the final "oflicial” version. Please
`address your remarks, not later than
`September 30, 1995, to
`
`Dr. Martin Siewert
`c/o Hoechst AG
`D-65926 Frankfurt (Germany)
`
`FIP will organize a symposium on all
`biopharmaceutical aspects of in vitro
`dissolution testing of solid oral prod-
`ucts in December 1995,
`to discuss
`and issue the final guidelines for dis-
`solution testing of solid oral prod-
`ucts. Programme and registration
`forms for this conference will be pub-
`lished and distributed separately.
`
`Introductory Remarks
`
`The first Guidelines for Dissolution
`Testing of Solid Oral Products were
`published in 1981 [1] as a joint re-
`port of the Section for Official Lab-
`oratories and Medicines Control
`Services and the Section of Indus-
`trial Pharmacists of the FIP. These
`
`Guidelines were intended as sugges-
`tions primarily directed to compen-
`dial committees, working on the in-
`troduction of dissolution / release
`tests for the respective Pharmaco-
`poeias.
`During the past decade, there have
`been many developments. Biophar-
`maceutics has attracted much inter-
`
`est scientifically as well as regarding
`drug regulatory policies. Dissolu-
`
`tion test methodology has been in-
`troduced to many Pharmacopoeias
`and a number of regulations and
`guidelines on bioavailability, bio-
`equivalence and in vitro dissolution
`testing have been issued at national
`and international levels.
`
`These updated Guidelines (second
`edition) are the result of careful dis-
`cussions of the joint working group
`of the two FIP, sections and are ba-
`sed on recent developments. De-
`scriptions of test methodology are
`no longer necessary, because they
`are already published elsewhere, of-
`ficially or
`semi-officially. Differ-
`ences between the regulations of dif-
`ferent countries and compendias
`were identified and proposals for
`harmonisation are made.
`
`As far as is reasonable for the pur-
`pose of these Guidelines, technical
`terms and definitions have been ad-
`opted from other harmonised re-
`commendations and mainly corre-
`spond to USP-terminology. New
`terms are “in vitro-in vivo compar-
`ison”,
`“verification”
`and
`“side
`batches”. “In vitro-in viva compar-
`ison” means any study collecting in
`vitro- and in vivo-data on the same
`
`set of test specimen to obtain in-
`formation and understanding about
`how in vitro and in vivo performance
`are related to each other. A signifi-
`cant in vivo-in vitro association can
`be a result of an in vitro-in vivo com-
`parison study, but valuable informa-
`tion could also be obtained when no
`correlation in a strict sense (e.g.
`
`Depomed Exhibit 2158
`
`

`
`USP levels) is achieved. “Verifica-
`tion” is used to define the in vivo
`data set which provides evidence
`that a chosen in vitro test method
`and the proposed specifications are
`suitable for the drug formulation in
`terms of biopharmaceutical
`per—
`formance.
`“Verification”
`is pro-
`posed as a new terminus technicus
`to avoid to extend “validation” also
`
`,,Side
`investigations.
`viva
`on in
`batches” are batches of a given drug
`formulation which represent the in-
`tended upper and lower specifica-
`tion limits. They are preferrably to
`be derived from the defined manu-
`facturing process by setting process
`parameters within the range of
`maximum variability expected from
`process validation studies. The term
`“dissolution” itself is used for all
`dosage forms, i.e. immediate—release
`(such as prompt drug releasing or
`conventional dosage forms) as well
`as controlled/modified-release prod-
`ucts (such as controlled, delayed,
`extended, modified, prolonged or
`sustained).
`
`1. Concepts of Dissolution
`Testing
`
`In vitro dissolution testing serves as
`an important tool for characterising
`the biopharmaceutical quality of a
`product at different stages in its li-
`fecycle. In early drug development
`in vitro dissolution properties are
`supportive for choosing between
`different
`alternative
`formulation
`candidates for further development
`and for evaluation of active ingredi-
`ents/drug substances. In vitro dis-
`solution data are supportive in the
`evaluation and interpretation of
`possible risks, especially in the case
`of controlled/modified—re1ease dos-
`age forms — e.g. as regards dose
`dumping, food effects on bioavail-
`ability or
`interaction with other
`drugs, which influence gastrointesti-
`nal environmental conditions. Bio-
`pharmaceutical aspects are as im-
`portant
`for stability concerns as
`they are for batch release after pro-
`duction, in vitro dissolution being of
`high relevance in quality control
`and quality assurance. Last but not
`least,
`in vitro dissolution data will
`be of great
`importance when as-
`sessing changes in production site,
`manufacturing process or formula-
`tion and assist in decision concern-
`ing the need for bioavailability stud-
`ies.
`
`None of these purposes can be ful-
`filled by an in vitro test system with-
`out sufficient reliability. Reliability
`here would be defined as the system
`being experimentally sound, yield-
`ing precise, accurate, repeatable re-
`sults and with sufficient knowledge
`
`2
`
`of the in vivo relevance of the dis-
`solution data obtained.
`
`Requirements for dissolution testing
`have been reviewed in the literature
`[2—6]. Since in vitro dissolution is a
`physical test, defined by convention
`and is of a destructive nature, prov-
`ing reliability requires special atten-
`tion. It therefore is within the scope
`of these Guidelines to define suit-
`able testing equipment and experi-
`mental design as well as to suggest
`the background for adequate phys-
`ical and analytical validation,
`to-
`gether with verification procedures
`according to the state of biopharm-
`aceutical science.
`
`The Guidelines are primarily dedic-
`ated to solid oral products. How-
`ever,
`the general concepts may be
`adapted to in vitro dissolution test-
`ing of drug substances/powders,
`semi-solid oral products, supposito-
`ries and, with destinct restrictions,
`to other non-oral products.
`
`2. Apparatus
`
`Large numbers of different dissolu-
`tion apparatuses are described in
`the literature, but only some of
`them withstand critical methodolo-
`gical examination.
`Two basic technical principles are
`applied for in vitro dissolution test-
`ing:
`the “stirred beaker method”
`and the “flow through procedure”.
`The “stirred beaker method” places
`the test specimen and a fixed vol-
`ume of fluid in a large vessel, and
`stirring provides mechanical (hydro-
`dynamic) agitation. This closed sys-
`tem design was adopted as first offi-
`cial method in USP XVIII in 1970,
`described as the rotating basket (ap-
`paratus 1, USP).
`The rotating basket and the paddle
`(apparatus 2, USP) devices are
`simple,
`robust
`and
`adequately
`standardized apparatuses which are
`used all around the world and thus
`are supported by the widest experi-
`ence of experimental use. It is be-
`cause of these advantages that the
`paddle and rotating basket appara-
`tuses are recommended in various
`guidelines as first choice for the in
`vitro dissolution testing of immedi-
`ate as well as contro1led/modified-
`release preparations.
`However, because of the “single
`container” nature of the paddle/bas-
`ket apparatus experimental diffi-
`culties may arise in terms of the
`need of a change in pH or of any
`other (partial) change in the test
`medium during an investigation.
`Furthermore, a number of sparingly
`soluble drugs and dosage forms,
`particularly aerophilic multiple unit
`forms tend to float
`initially. Pro-
`posals have been made to overcome
`some of these difficulties, e.g. to in-
`
`crease solubility by addition of an
`appropriate amount of surfactant.
`With the flow-through cell (appara-
`tus 4, USP) the specimen is placed
`in a small colurrm which is continu-
`ously flushed with a stream of fluid,
`simultaneously providing the me-
`dium and the mechanical agitation
`for dissolution of the drug sub-
`stance. It can be run as an open as
`well as a closed system. The open
`system design especially provides
`several advantages in some of the
`difficult cases mentioned above and
`was adopted first by the Deutscher
`Arzneimittelcodex (German Phar-
`maceutical Codex, DAC) in 1981.
`The flow-through apparatus is cur-
`rently monographed in USP and
`Ph.Eur. and is also proposed for Ph.
`Jap. Description of the system is
`concordant worldwide. The paddle/
`basket system is described in USP,
`the European,
`the Japanese and
`many other Pharmacopoeias. Some
`minor discrepancies are still found
`in details of the respective mono-
`graphs. Full international. harmon-
`isation is strongly recommended as
`proposed in table 1.
`As a further system (apparatus 3)
`USP describes the reciprocating cy-
`linder. With these four apparatuses,
`dissolution testing of most oral
`drug products should be possible on
`a reasonable basis. Neither too tight
`restrictions nor unnecessary prolif-
`eration of alternative dissolution
`apparatuses should be encouraged.
`If an individual drug product can-
`not be accomodated by one of the
`apparatuses, described above, alter-
`native models or appropriate modi-
`fications have
`to be developed.
`However, in such a case superiority
`of the alternative or the modifica-
`tion has to be proven in comparison
`to the well estabilshed and stan-
`dardised apparatuses. In the past,
`many papers intended to justify an
`alternative model by proving that in
`vitro dissolution results were equi-
`valent or similar to those obtained
`with e.g.
`the paddle method. Ac-
`cording to the understanding of
`these Guidelines, the latter provides
`clear
`evidence
`that
`the paddle
`method should be used!
`Modification of the apparatus as
`described in the Pharmacopoeias or
`the harmonisation proposal in table
`1 can be intended for automation
`e.g. of sampling procedure. In such
`cases, whether it is e.g. sampling via
`the hollow shaft of paddle or basket
`or permanent sampling probes in
`the beaker, which could potentially
`influence on agitation character-
`istics [7], or any other measure,
`it
`should be validated on a product-
`by-product basis that
`results are
`equivalent with and without
`the
`modification.
`
`

`
`Table 1: Dimensions of the Paddle/Basket Apparatuses (Millimeters).
`
`Vessel
`Height
`lntemal Diameter
`Paddle
`Shaft Diameter
`Blade
`Upper chord
`Lower chord
`Height
`Radius of the disk of which
`the blade is cut out
`Radius upper comers
`Thickness
`Basket
`Shaft diameter
`
`Screen
`Wire diameter
`
`Openings
`
`Height of screen
`Total height of basket
`Internal diameter of basket
`External diameter of basket
`External diameter of ring
`Vent hole diameter
`
`Height of coupling disk
`Positioning the stirring device:
`Distance between inside of the bottom
`of the vessel and the blade/basket
`
`Distance between shaft axis and
`vertical axis of the vessel
`
`Stirring characteristic
`
`160-175
`98-106
`9.4-10.1
`(before coating)
`
`0.254 (0.01 inch)
`or 0.016 inch’)
`0.381 (0.015 inch)
`or 0.034 inch3)
`
`160-175
`98-106
`9.75 2 0.35
`
`74.5 2 0.5
`42 2 1
`19.0 2 0.5
`
`41.5
`1.22)
`3-5
`
`(9.75 2 0.35)
`6.4 2 0.1
`
`No. 36 wire
`gauze
`0.425
`
`27 2 1
`36.8 2 3
`20.2 2 1
`22.2 2 1
`25.4 2 3
`2
`
`168 2 8
`102 2 4
`9.75 2 0.35
`
`74.5 2 0.5
`42.0
`19.0
`
`41.5
`1.21)
`4.0 2 1
`
`(9.75 2 0.35)
`6.4 2 0.1
`
`0.254
`
`0.381
`
`27.1 2 l
`36.8 2 3
`20.2 2 1
`22.2 2 1
`25.4 2 3
`2
`5.12 0.5
`
`25 2 2
`
`s 2
`
`Proposal
`(EFPIA)
`
`160-210
`102 2 4
`9.75 2 0.35
`
`74.5 2 0.5
`42.0 2 1.0
`19.0 2 0.5
`
`41.5 21.0
`1.2
`4.0 2 1.0
`
`9.4-10.1
`
`0.254‘)
`
`0.381“)
`
`27.0 2 1.0
`36.8 2 3.0
`20.2 2 1.0
`22.2 2 1.0
`25.4 2 3.0
`2.0 2 0.5
`5.12 0.5
`
`25 2 2
`
`s 2
`
`smoothly without
`significant wobble
`(S 0.5 mm)
`
`smoothly without
`significant wobble
`
`smoothly without
`significant wobble
`(5 0.5 mm)
`
`1) USP 23 Suppl. 2.
`2) IP Forum Vol. 3 No. 3 (July, 1994).
`3) Basket to be used is indicated in the individual monographs.
`4) Should correspond with the requirements for standards, e.g. International Standard ISO 2194 — 1972.
`
`3. Experimental Testing
`Conditions
`
`For all applications, in vitro dissolu-
`tion data should at least allow some
`
`interpretation with regard to in vivo
`biopharmaceutical performance. In
`order to increase their predictive
`value, attempts have been made to
`adjust in vitro test conditions [8—1 1]
`as close as possible to physiologic
`conditions. Nevertheless, several ex-
`amples demonstrate that such con-
`ditions can also lead to n1isinter-
`pretations and are not able to guar-
`antee in vitro results routinely rele-
`vant to the in vivo situation [12].
`In general, an aqueous medium
`should be used. It
`is not recom-
`mended to attempt to strictly mimic
`the physiologic gastrointestinal en-
`vironment (e.g. composition of gas-
`tric or intestinal fluid) but to choose
`the testing conditions as far as is
`reasonable, based on the physico-
`
`characteristics of drug
`chemical
`substance, within the range which a
`drug or dosage form could experi-
`ence
`after
`oral
`administration.
`These following ranges were estab-
`lished based on several conferences
`and recommendations [e.g. 13-15].
`For basket/paddle methods the vol-
`ume should be 500 to 1000 ml. 900
`ml had been introduced historically;
`1000 ml should be easier to handle
`in a metric system, this volume be-
`ing practicable with all equipment
`commercially available today. 1000
`ml therefore should be considered
`for new drug products or in case of
`a revision of existing test proce-
`dures. This recommendation does
`not mean that 1000 ml should be
`adopted to all existing test proce-
`dures and specifications. Although
`larger vessels, such as up to 4,000
`ml,
`could be
`advantageous
`for
`poorly soluble drugs, they are not
`described in compendia, and thus
`
`are not as well standardised and
`therefore should be regarded as mo-
`dification of a compendial method
`(see section 2.).
`The pH of the test medium should
`be set within pH 1 and 6.8 A higher
`pH needs to be justified on a case-
`by-case basis and in general should
`not exceed pH 8. For low pH in the
`acidic range 0.1N HCl should be
`used. If, in a certain case, artificial
`gastric juice without enzymes (pH
`1.2) is advantageous, this should be
`demonstrated.
`
`In the pH-range of 4.5 to 8.0 USP
`buffer solutions are recommended,
`because their buffer concentration
`(ionic strength) is not as high as
`that of e.g. buffers of Ph.Eur., which
`have not been designed for dissolu-
`tion testing.
`The use of water as dissolution me-
`dium bears the disadvantage that
`test condition details, such as pH
`and surface tension, can vary de-
`
`3
`
`

`
`pending on the source of water and
`may be changed during the dissolu-
`tion test itself, due to the influence
`of the drug products and to the
`(re-)absorption of carbon dioxide
`from air.
`
`Further additives e.g. enzymes, salts
`or surfactans, could be considered
`in specific cases. Their use should be
`justified as regards nature and con-
`centration of additive [16]. Addition
`of organic
`solvents
`should
`be
`avoided.
`
`Agitation typically should be ob-
`tained in the basket/paddle appara-
`tus by stirring at 50 to 100 rpm and
`in general should not exceed 150
`rpm. Although maximum discrim-
`inatory power should be obtained
`with lowest stirring rate,
`in many
`cases experience with 75 rpm was
`felt to represent a reliable agitation
`for paddle equipment [17]. For the
`flow through cell, flow rates should
`be set between 8 and 50 ml/min.
`
`Regarding temperature, 37 2 0.5 °C
`should generally be used for oral
`dosage forms. Slightly increased test
`temperatures (e.g. 38 :05 °C) are
`under consideration for special ap-
`plications
`e.g.
`for
`rectal dosage
`forms,
`lower temperatures (e.g. 32
`: 0.5 °C) for transdermal systems.
`Relevant parameters
`to be con-
`sidered for the definition of test
`conditions are solubility and deaer-
`ation.
`In former Guidelines
`[1],
`“sink” conditions were requested.
`“Sink“ was defined in different ways
`e.g. as 10 to 20 % [1] or approxima-
`tely 30 % [18] of solubility concen-
`tration to assure that dissolution is
`not significantly influenced by solu-
`bility characteristics. Since “sink”
`conditions per se do not guarantee
`in
`vivo-in vitro associations and
`since reliable and predictive in vitro
`profiles in certain cases can be ob-
`tained by violating “sink” condi-
`tions, solubility and drug substance
`concentrations
`during
`the
`test
`should be matter of validation stud-
`ies to demonstrate that a chosen in
`vitro test method yields biopharma-
`ceutically relevant results.
`Case-by-case validation is also re-
`quired regarding deaeration since
`some formulations will be sensitive
`whereas others are robust
`in this
`concern,
`thus making deaeration
`unnecessary. The dearation method
`has to be clearly characterised, since
`also different methods can have im-
`pact on dissolution profiles [19].
`Ph. Jap. XII is currently the only
`Pharmacopoeia that requires a spe-
`cific (very solid) sinker device for all
`capsule formulations. USP recom-
`mends a few turns of wire helix
`when
`specimen
`tend
`to
`float.
`EFPIA harmonisation
`proposal
`suggests a similar one. Sinkers can
`significantly influence the in vitro
`
`4
`
`dissolution profile of a drug [20].
`Since they are used especially with
`formulations causing problems dur-
`ing test performance, e.g. flotation,
`they will alter the dissolution pro-
`file, so that other recommendations
`[18] are not applicable.
`The use of sinkers therefore has to
`be part of case-by-case dissolution
`validation as well as of in vitro-in
`vivo comparison studies. Any strict
`requirement on use of sinkers or
`specific sinker types lacks scientific
`justification.
`
`4. Qualification and Validation
`Due to the nature of
`the test
`method, quality by design is an im-
`portant qualification aspect for in
`vitro dissolution test equipment. Be-
`sides the geometrical and dimen-
`sional accuracy and precision as de-
`scribed and commented in chapter 2
`(including table 1), any irregularities
`such as vibration or undesired
`agitation by mechanical
`imperfec-
`tion are to be avoided.
`Besides the specification of the ap-
`paratus, qualification of dissolution
`equipment has to consider critical
`parameters, e.g. temperature of test
`medium,
`rotation speed/flow rate,
`volume, sampling probes and proce-
`dures, to be monitored periodically
`during the periods of use.
`Apparatus suitability test with cal-
`ibrators is a further important as-
`pect of qualification and validation.
`The use of USP calibrator tablets
`(disintegrating as well as non-disin-
`tegrating as well as non-disintegrat-
`ing) is recommended. Since some
`individual drug products might re-
`veal similar or even higher sensitiv-
`ity against
`technical variance in
`comparison to USP calibrator tab-
`lets,
`“in-house”
`standards
`are
`judged acceptable as additional, or,
`if validated, equivalent
`for cali-
`brator tablets.
`The suitability test has to cover each
`individual apparatus and to consist
`of the full USP programme, mean-
`ing both calibrator types. Paddle
`and basket equipment, as well as 12
`mm and 22.6 mm flowthrough cell
`have to be qualified, unless only
`paddle or basket, respectively only
`small or large cell is used in one spe-
`cific piece of equipment. The system
`suitability test of USP Apparatus 3
`has to be performed with both, a
`multiparticulate and a monopartic-
`ulate standard formulation. A sys-
`tem suitability test for flow-through
`cell has just been established and
`will be soon published for USP [22].
`Apparatus suitability tests are re-
`commended to be performed not
`less than twice per year per equip-
`ment and after any occasion of
`equipment change, significant
`re-
`
`a
`pair or movement. However,
`change from paddle to "basket or
`vice versa may not require recalibra-
`tion.
`
`Additional validation aspects are
`precise product related operation in-
`structions (e.g. deaeration proce-
`dure). Dissolution results may be
`influenced by the physical behavi-
`our of the specimen such as float-
`ing, adherence to the walls, etc.
`Thus, critical inspection a.nd obser-
`vation of test performance during
`the test procedure is required. This
`approach is especially important to
`explain any “out-lying” results and
`it clearly limits the extent of auto-
`mation for a number of drug formu-
`lations.
`Validation of automated systems,
`either concerning the sampling and
`analytical part or also including me-
`dia preparation and test perform-
`ance, has to consider accuracy, pre-
`cision and avoid contamination by
`any dilutions, transfers, cleaning or
`sample or solvent preparation pro-
`cedures. There should be proof that
`there is no interference. This shall
`be evidence of no significant differ-
`ences between data obtained with
`the manual dissolution equipment
`(see 2.) and the automated system,
`including manipulations
`such as
`permanent sampling probes, addi-
`tional valves, hollow shafts, etc.
`Since sensitivity to such modifica-
`tion may be formulation related,
`validation of automated dissolution
`equipment has to be established on
`a case-by-case basis.
`Validation of the analytical proce-
`dures applied in dissolution testing,
`either automated or conventional,
`has to comply with “Validation of
`Analytical Procedures” (ICH 2) and
`“Validation of Compendial Me-
`thods” (<l225>, USP). Validation
`aspects thus are accuracy, precision
`(repeatability, reproducibility). spe-
`cificity, linearity, range. Special care
`has to be taken regarding stability
`of the drug in test medium and
`sample solutions, since the test pro-
`cedure often includes exposure to
`hydrolytic media at 37 °C over sig-
`nificant time spans.
`
`5. Formulation Characterisation
`
`During development of the drug
`formulation, as a basis for any in vi-
`tro-in vivo comparison study as well
`as for the final choice of test condi-
`tions for quality control purposes,
`the respective dosage form has to be
`thoroughly characterised in vitro
`with respect to its biopharmaceut—
`ical performance. Special attention
`has to be paid to controlled/modi-
`fied-release preparations, since suffi-
`cient information has to be gained
`about how much the dosage form
`
`

`
`itself, rather than variations in test
`conditions, “control” the rate of
`drug release.
`dissolution
`extensive
`Therefore,
`tests are necessary to understand
`the delivery system and to have a ra-
`tionale for the design of e.g. and in
`vitro-in vivo comparison study. The
`in vitro test profile will preferably
`consist of numerous individual dis-
`solution tests under many different
`test conditions, involving the pH of
`test media and agitation within the
`ranges given in section 3. Variation
`of ionic strength, surfactants, en-
`zymes or apparatus
`should be
`evaluated,
`if an influence on dis-
`solution is expected for the indi-
`vidual formulation.
`
`formulation characterisation,
`For
`dissolution tests
`should be per-
`formed under the different test con-
`ditions until actual dissolution (e.g.
`mean of six specimen) exceeds 80 %
`of labelled amount. When, even
`with test prolongation, results re-
`main significantly below 80 % and
`solubility is not the limiting para-
`meter, recovery control should be
`performed to prevent misinterpreta-
`tion of dissolution data.
`Since most in vitro Characteristics
`can be related to physiological para-
`meters (table 2)
`the information
`from formulation characterisation
`in vitro can be used later as a tool
`to demonstrate the reliability of an
`in vitro-in vivo comparison, based
`on a distinct in vitro model, as well
`as for interpretation of all those ex-
`amples where no or only a poor cor-
`relation of in vitro and in vivo data
`can be achieved. However, it is obvi-
`ous that a meaningful in vitro-in vivo
`comparison (see section 6)
`is the
`more probable, the less affected in
`vitro dissolution of a given drug for-
`mulation is by changes in the envir-
`onmental test conditions.
`
`6. In vitro-in vivo Comparison
`
`An in vitro test system for a given
`drug formulation serves as the tool
`as which it is designated only, if it
`can distinguish between “good” and
`
`“bad” batches. “Good” here means
`“of acceptable and reproducible
`biopharmaceutical performance in
`vivo”. Thus in vivo relevance of an
`in vitro test system is sought. The
`purpose of in vitro-in vivo compar-
`ison studies in this sense is the sci-
`entific verification of the in vitro test
`
`system and the respective specifica-
`tion limits for a given drug formula-
`tion.
`
`Regarding extended-release dosage
`forms the USP [18] has categorised
`correlative methods, harmonised in
`a wide international consensus, as
`correlation level A (1 :1 relation-
`ship between in vitro and in vivo dis-
`solution, calculated by numerical
`deconvolution [23, 24], according to
`Wagner-Nelson method [25] or to
`Loo-Riegelmann method [26]), cor-
`relation level B (statistical moment
`analysis [27, 28]) and correlation
`level C (single-point correlation of a
`dissolution time vs. a pharmacoki-
`netic parameter). Depending on the
`correlation level finally obtained, in
`vitro dissolution properties will be
`decisive for the necessity of how
`many batches should be included
`for a correlation study, e.g. for es-
`tablishment of in vitro dissolution
`specification limits. According to re-
`cent recommendations, one single
`batch may be sufficient for a scien-
`tifically and formally acceptable
`correlation [15, 18], only in case of
`a correlation level A and a product
`with a drug release, completely in-
`dependent from environmental con-
`ditions, which then is represented by
`only one dissolution curve. Scient-
`ific and pragmatic approaches for
`level A correlations have been pro-
`posed [29]. In case of a level A cor-
`relation,
`manufacturing
`site
`changes, minor formulation modi-
`fications,
`scale-up
`considerations
`and setting of specifications can be
`based and justified without further
`in vivo-studies.
`least two or
`In all other cases at
`three different batches have to be
`used, offering differences in their
`biopharmaceutical properties, suffi-
`cient
`for
`correlation
`purposes.
`Nevertheless, these differences have
`
`Table 2: Physiological variables contributing to biopharmaceutical performance
`especially for controlled/modified release formulations and aspects of in vitro model-
`ling.
`
`In vitro modelling possible by
`pH-profile
`Agitation-profile
`
`Physiological variable
`Intragastric pH
`
`Gastrointestinal motility, peristaltics,
`shearing forces
`Fat, lipophilic and other compounds
`
`Addition of lipids (fat and/or fatty
`acids) or other compounds (e.g. fiber
`materials)
`Addition of enzymes
`Enzymes
`Addition of surfactants
`Bile
`.
`.
`.
`.
`Gastrointestinal transit-times
`
`
`Table 3: Possible reasons for poor in
`vivo-in vitro correlations.
`
`Fundamentals
`the rate
`0 In vivo dissolution is not
`limiting step for drug absorption
`0 No in vitro test is able to model in vivo
`dissolution
`
`Study design
`0 Inappropriate in vitro test conditions
`Inappropriate in vivo test conditions
`Dosage form
`0 Drug release not controlled by the
`dosage form
`0 Drug release strongly affected by
`intestinal transport kinetics
`Drug substance
`0 Non-linear pharrnacokinetics (e.g. sa-
`turable first pass effect‘), absorption
`window, chemical degradation in the
`gastrointestinal tract
`0 Absorption of undissolved particles
`0 Large intraindividual variability
`
`to be ’effected’ by only small modi-
`fications of manufacturing variables
`within the ranges of the given pro-
`cess. In cases where differences can-
`not be achieved by these variations
`of the production process, different
`formulations of a drug substance
`are to be obtained for in vitro-in vivo
`comparison. However, any correla-
`tion received for different formula-
`tions bears the risk of being some-
`what arbitrary. A final evaluation of
`type and influence of the changes in
`the manufacturing processes
`re-
`quires thorough in vitro dissolution
`tests
`(’biopharmaceutical profile’,
`see section 5) prior to an adminis-
`tration to human volunteers in a
`clinical study.
`Concerning controlled/modified-re
`lease products there is international
`consensus that levels A to C, with
`a quality ranking A > B > C, are
`acceptable for correlation e.g.
`for
`setting specificatoins. A number of
`different reasons (see table 3) could
`be responsible for “poor” or no cor-
`relation.
`
`Even with highly sophisticated tech-
`niques it is often difficult to obtain
`meaningful in vitro-in vivo compar-
`isons,
`especially when regarding
`biopharmaceutically very sirriilar
`(bioequivalent?) products, such as
`batches of one drug formulation, re-
`presenting the upper and the lower
`specification limit.
`Recently, proposals have been made
`[30] which in vitro-in vivo compar-
`ison results scientifically and form-
`ally could suffice as verification of
`dissolution specification of con-
`trolled/modified-release
`products.
`In case _of a significant quantitative
`correlation,
`specifications can be
`dem/ed
`by . mterpolatlo-n’ when
`batches outside the specified bio-
`pharmaceutical range are tested for
`
`5
`
`

`
`in vitro-in vivo comparison. Then, at
`least three batches should be tested
`in vitro and in vivo. A qualitative, i.e.
`rank-order correlation verifies spe-
`cification ranges, when at least three
`batches are tested in vivo and in vitro
`and the dissolution data of two of
`the
`experimentally
`investigated
`batches are concluded bioequivalent
`and their dissolution characteristics
`are defined as upper and lower spe-
`cification limits.
`Where no correlation is obtained
`from an in vitro-in vivo comparison
`study, an alternative approach could
`consist of demonstrating bioequiva-
`lence of the proposed formulation
`to formulations with dissolution
`profiles at
`the upper and lower
`limits of the specification [13].
`The number of volunteers to be in-
`cluded
`in
`such
`comparative
`bioavailability studies or in an in vi-
`tro-in vivo comparison study is to be
`defined on a case-by-case basis but
`in general should not be less than
`twelve.
`The batch size of a formulation for
`in vitro-in vivo comparison studies
`need not be of full production scale.
`Parameters
`for manufacture of
`these batches, especially of formula-
`tions representing the intended spe-
`cification limits, should be defined
`from process validation studies ac-
`cording to the expected maximum
`variability of process parameters
`(“side batches”).
`Concerning immediate-release dos-
`age forms a suitable design for an in
`vitro—in vivo comparison study could
`consist of a two-way crossover be-
`tween an oral solution and a formu-
`lation representing the (lower) spe-
`cified dissolution limit.
`
`7. Dissolution Specifications
`
`The purpose of establishing dissolu-
`tion specifications
`is
`to ensure
`batch-to-batch consistency within a
`range which guarantees acceptable
`biopharmaceutical performance in
`vivo. Specification limits therefore
`have to be defined based on experi-
`ence gained during the drug devel-
`opment stage especially regarding
`clinical development andjor bio-
`equivalence studies. In most cases
`deduction of specification limits re-
`quires thorough in vitro-in vivo com-
`parison studies as described in sec-
`tion 6.
`
`For immediate—release formulations
`typically one li