CA1145158A - Method and apparatus for automatic dissolution testing of products - Google Patents
Method and apparatus for automatic dissolution testing of productsInfo
- Publication number
- CA1145158A CA1145158A CA000375654A CA375654A CA1145158A CA 1145158 A CA1145158 A CA 1145158A CA 000375654 A CA000375654 A CA 000375654A CA 375654 A CA375654 A CA 375654A CA 1145158 A CA1145158 A CA 1145158A
- Authority
- CA
- Canada
- Prior art keywords
- dissolution
- control
- signals
- flow
- output
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N13/00—Investigating surface or boundary effects, e.g. wetting power; Investigating diffusion effects; Analysing materials by determining surface, boundary, or diffusion effects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N13/00—Investigating surface or boundary effects, e.g. wetting power; Investigating diffusion effects; Analysing materials by determining surface, boundary, or diffusion effects
- G01N2013/006—Dissolution of tablets or the like
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N35/00—Automatic analysis not limited to methods or materials provided for in any single one of groups G01N1/00 - G01N33/00; Handling materials therefor
- G01N2035/00178—Special arrangements of analysers
- G01N2035/00188—Special arrangements of analysers the analyte being in the solid state
- G01N2035/00198—Dissolution analysers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/15—Medicinal preparations ; Physical properties thereof, e.g. dissolubility
Abstract
ABSTRACT OF THE DISCLOSURE
A method and apparatus for optimally performing dissolution testing of pharmaceutical dosage forms, agricul-tural products, and components of industrial products wherein the method uses dissolution profiles from a known drug dosage form, or product, as reference data for a predictive process;
and the apparatus is organized to carry out the method via both closed loop and open loop operating modes under the control of a central processor. An illustrative embodiment teaches the serial usage of the two operating modes in a single flow-through dissolution cell configuration to optimally predict the time course of in vivo bioavailability from in vitro dissolution measurements, while an alternate embodiment teaches the use of a plurality of dissolution cells and the simultaneous use of the closed and open loop operating modes to implement an Internal Standard capability. Additionally, an optimally adaptive capability is provided in the dissolution testing process via a random input modeling mode of operation.
A method and apparatus for optimally performing dissolution testing of pharmaceutical dosage forms, agricul-tural products, and components of industrial products wherein the method uses dissolution profiles from a known drug dosage form, or product, as reference data for a predictive process;
and the apparatus is organized to carry out the method via both closed loop and open loop operating modes under the control of a central processor. An illustrative embodiment teaches the serial usage of the two operating modes in a single flow-through dissolution cell configuration to optimally predict the time course of in vivo bioavailability from in vitro dissolution measurements, while an alternate embodiment teaches the use of a plurality of dissolution cells and the simultaneous use of the closed and open loop operating modes to implement an Internal Standard capability. Additionally, an optimally adaptive capability is provided in the dissolution testing process via a random input modeling mode of operation.
Description
S158 ~ ' ¦ ACKGROUND OF THE INVENTION
1. Field of th~ Invention ¦ The present invent$on relates generally to th~ fleld of automatic dlssolutlon testing of product~ whose solubility and dissolution rate properties affect product performance, and more specifioally to the optimal prediction of prsduct dissolution chara~teristics using known product data as a ¦ reference for a foedback controlled apparatu~.
I The methods disclosed and the electronically controlled 10 j apparatus descrlbed are presented in connection with in vitro ¦ ~issolution testing of pharmaceutical drus dosage forms to predict in vlvo bioava$1ability. However, both the methods and apparatus taught are equally applicable to disso-! lution testing of agricultural product~ formulated as con-trolled relea~e herbicide~, insecticide~, fertilizers, and ,~ tho 11ket and further to the di~solution testing of components ', of industrial products including solid materials who~e ¦ ~olubility propertles depend on a w$de var$ety of factor~.
1. Field of th~ Invention ¦ The present invent$on relates generally to th~ fleld of automatic dlssolutlon testing of product~ whose solubility and dissolution rate properties affect product performance, and more specifioally to the optimal prediction of prsduct dissolution chara~teristics using known product data as a ¦ reference for a foedback controlled apparatu~.
I The methods disclosed and the electronically controlled 10 j apparatus descrlbed are presented in connection with in vitro ¦ ~issolution testing of pharmaceutical drus dosage forms to predict in vlvo bioava$1ability. However, both the methods and apparatus taught are equally applicable to disso-! lution testing of agricultural product~ formulated as con-trolled relea~e herbicide~, insecticide~, fertilizers, and ,~ tho 11ket and further to the di~solution testing of components ', of industrial products including solid materials who~e ¦ ~olubility propertles depend on a w$de var$ety of factor~.
2. Des~ription of the Prior Art Dissolution testing of component~ of industr$al proau¢ts whose solubility and d$ssolut$on rate properties ¦ affect product poxformanqe can be used as a screening ~l an~ quality control tool. The solubility properties ~1 ~ solld materlal~ can depend on polymorphic crystalline 251 form, crystal habit, crystal shape, partlcle size and particle ~ size dlstributlon, and ~tate of ~olvatlon. A simple and j ' 5158 ' I
rapidly performed dissolution test can substitute for the aeterminat~on of these physical propert~es by more time con-~uming and expensive methods such as x-ray cry~tallography, differential thermal analy~is, microscopy, etc. The materials S are instead dotermined as to whether they conform to a di~solution rate ~tandard u~der specifiea conditions and in relation to a known reference sample of the same material characterized by the above physical properties and po~sessing th~ aerived aissolution rate and ~olubll *y performance.
The broad technique of determining dissolution rate propert~e~ i8 e~pecially of intere~t in the testing of drug products where the therapeutic performance of drugs i8 closely related to the drug dissolution propertie~. Seemingly minor chan~ in drvg product formulation, as well as the inadvertent varlation in matorial~ and manufacture that can occur between batches of the same productformulatiop,caninfluence the therapeutic performance of drugs. In vivo bioavailability t-~ting of drug products in human~ provides the mo~t reliable m-ans of ensur~ng bioequivalence. Howev-r, it is impractical ~0 ~o per~orm the ~xtensive and expensive human testing that would be routinely required. ~arge numbers of buman sub~ect~
would be placed ae rl~k if uch studies were conducted.
~ioavailability testing in which humans are used ~ te~t ~ub~ects can be min~mized by the development and implementation o~ in vitro di~-olution standards that reflect in vivo drug-product erformahc~. In vitro b1oequivalence require~ent-I
G ~ 5158' have been established for ~ome drugs such as digoxin. From among the various ch~ical and physical tests that can be performed on drug solid~ in vitro for correlating or predicting a drug.produc~'s in v$vo bioavailabil~ty behavior, dissolution S tosting i8 th~ most sensitive and reliable. The correlative r~lation~hips ~08t commonly reported bstween in vitro dissolu-t~on and in v~vo bioavailability are of the single-point type:
the percentage of the drug di~solved in ~ given t~me (or the t~me it takes to di~solve a given percen~age of the drug in vi.tro) and ~ome univariate characteristlc of the drug product's ln vivo re~ponse versus time profile ~such as the peak blood level, the time required to reach the pe~k or 50% of the peak, or the area under the blood-level curves) are correlated.
The sele¢tion of in vitro dissolution and in vivo bioavaila- ¦
bility parameters for such single-polnt correlations is frequently arbi.trary, and the results can be misleading.
Obviou~ly, it would be preferable to predict the entire avorage ~lood level, urinary recovery rate, pharmacological-¦ re~ponse-time, or.drug absorption rate VJ. time profile that 20¦ w~uld be elicIt~d by a arug produ¢t in a panel of human ~ub~ects rather than merely to ¢orrelate univar~ate charac-t~ri~tics of th~ dlssolution proflle with an in v~vo b~o-availability parameter. In all cases, however, the fidelity of the in vltro dlssolutlon results in oorrelat~ng and in predicting in vivo drug-product bioava~l~bility depend~
_ dl3~01u~ion-t-st prooe ~ varlabl~ uch a the ~ ~- -1 1 ~ 51 5 ~
a$ssolut~on-medium composition, the solubility Yolume of *he meaium tsink conditions that determine the extent to which the med~um become~ saturated with the drug), and the agitation rates tstirr$ng or flow rates). An improper choice of these S proces~ varia~les te.g., an excess~vely h~gh rate of agitation) can ma~k signi icant ~ioavailability differences among drug products. On thQ other ~and, the dissolution test can be overly ~ansitive in detecting difference~ that are negligible in vivo.
~n the former case, using such improper dissolution-test parameters would result in the marketing of therapeutically in~ffective drug products. In the latter case, the result ~ould be the disca~ding of drug product-~ that are ent~rely ~atisfactory in terms o in vivo performance. Serious economic losse~ could re~ult from the use of an overly sensitive in ~ltro disso1ution te-t for lot-to-lot reproducibility testing of drug produ¢ts. Therefore, whether the dissolution te~t is b~$ng used a8 a quallty control tool, a~ an ln vivo bioequi-valoncy requir~m~nt ~or multisource generic drug proaucts, or ~8 a substltute for human bioavailability test~ng during : 20 ~ th development of new drug-product formulation~, it i~
~mperative that the dlssolut~on test provide predictive result~ that ar~ biolog~cally relevant.
Developing drug-product d~s~olut~on tests that pred~ct th- t~me course of drug-product bioavailability can ~o fraught with pitfalls, some of which may be avoided through knowle~go and consid~ra~tion of the phyQlochemical propertlss o~ the oomponents of the drug product and the biological proces~es and conditions operatlve ~n the r~lea~e .
"~ 5158 of the drug from the gastrointe~tinal tract and it~ subsequent absorption. ~owever, it is not only futile, but also unnecessary to attempt to reproduce the complex of biological factors operating in vivo in the effort to d~velop a satisfactory in vitro bioavailability test, although suoh attempts have been made. The devices that re~ulted from these efforts are o~ value now only a8 museum piece~. It wnuld, however, be imprudent to lgnore such knowledge when it can be u~ed aavan-tageou~ly to circumvent a problem in the ~eqign of a dissolution test.
There are two po~sible general approaches to developing ~n vivo relevant drug product dissolution te~ts. ~oth approaches s~ek to predict thë entire time course of average blood levels that would be ob~erved for a drug product in a panel of human te~t sub~ects. In thi~ way, the di~olution test serves as a ~ubstitute for human te~ti~g.
The first approach i8 a computational method that maximizes the amount of inormation that can be ob~ined from conventlorl~l methods of in vitro di~solution testing. U~ed 20 ~08t frequently are the USP rotating-basket apparatus, the FDA paadle method, the ~tationary-b~sket/r~tat1~g filter apparatus, Sartorlu~ solub11~ty and absorption simulators ~Sartoriu~, In¢orporated, Hayward, ~alifornia), ~nd column-type flow-through as~emblie~. The last of the~e device~
offers advantages ~1th regard to the definition, flexi~ility ~l~S158 of control, standardization, and reproducibility of process variables. Thi~ apparatus has been used by the inventor of the present ~nvention to demon~trate the second approach to predicting 1~ vivo blood-lsvol curves that em~rge from the apparatus in the form of dis~olution rate versu8 time profiles.
Since the computational approach with conventional ¦ -Apparatus d~pend~ upon the relatively arbitrary selection of process variables, its usefulness is limited. However, using feedback control to continuously vary the process variables, as described below, obviates this problem. For a more complete treatment of the mathematical (and theoretical) aspects of the d~ssolution, the interested reader is dirscted to three papers co-authored by the ~nventor. These are: V.F. Smolen et al, ~Optimally Predictive In Vitro Drug Dissolution Testing for In Vivo Bioavail~bility," J. Phsrmaceutical Sci., Vol. 65, No. 12, pp. 1718 1724, December 19i6; V.F. Smolen et al, "Predicting th~ T~m~ Cour~e of In Vivo Bioavailability ~rom In Vitro D~olution Tests: Control Systems Engineering Approaches,"
Pharmaceut~aal Technology, pp. 89-102, June 1979; and V.F.
8mo1en et al, ^Predictiv~ Cnnversion of In-Vivo Drug Dissolution Dat~ lnto In Vlvo Drug R~sponse Versus T~mQ Pro~les Exemplified ~or W~rfar~n,"' J. Pharmaceutical Sci., Vol. 66, ~o. 3, pp.
297-304, M~rch 1977.
The present invention is d~rected to an improved method 4nd apparatus for carrying out the dissolution approach to opt~mally prsdicting in vivo drug bloavailability from pharmaceutlcal ~osage forms, and other applications of dissolu-tlon te3ting.
~1..1~515~3 According to one aspect of the present invention there is provided an apparatus for the automated dissolution testing of products whose solubility and dissolution rate properties affect product performance, the apparatus having a dissolution cell with an input line, an output line, a filtering chamber therein, and controllable agitation means responsive to an agitation control signal, positioned within the chamber. At least two reservoirs are connected to the input line through a like number of media supply lines for providing dissolution media from the reservoirs to the filtering chamber. A first control means is responsive to a first control signal and positioned in the input line for controlling the flow of dissolution media through the filtering chamber, and second flow control means responsive to a second control signal and positioned in at least one of the media supply lines for controlling the flow of d ssolution media to the input line. The recirculating means include third flow control means responsive to a third control signal and connected to the input and output lines for controlling the recirculatory flow therefrom. System output means is connected to the output line and contains means for measuring output parameters therein including a concentration parameter of particular constituents within the flow, and for pxoviding system output signals corresponding to the measured output parameters. Memory means provides stored control signals to the apparatus. An electronic control means is connected to the memory means and is operative in a first mode responsive to the stored control signals and to the measured output parameter signals for automatically determining a set of internal control parameters used to determine the first, second and third control signals so as to minimize the differences between the stored and measured signals, and operative in an open loop mode responsive to the determined ~4 pc/'.,,.' internal control parameters for automatically producing the first, second and third control signals in the absence of the stored signals.
According to another aspect of the present inYention there is provided a method of automatic dissolution testing of products whose solubility and dissolution rate properties affect product performance having unkown dissolution profiles by comparison against known product forms having kno~n dissolution profiles. The method includes the steps of flowing a dissolution medium of controllable content at a controlled flo~ rate through a dissolution cell containing the known product form and providing a recirculation path around the dissolution cell to provide controllable sink conditions therethrough. The concentration profile of the known product form is measured at the output of the dissolution cell and output signals are provided corresponding thereto.
Stored signals corresponding to the concentration profile of the known product form are provided and a first family of control signals are produced such that the differences between the stored signals in the output signals are minimized.
A second family of control signals are produced for controlling the medium controlled content, the controlled flow rate, and the controlled sink conditions from the first family of control signals. A dissolution medium of the controlled content at the controlled flow rate is flowed through a ; dissolution cell containing the unknown product form so that measured concentration profiles of the unknown product form are derived as being predictive of subsequent usage dissolution profiles.
According to yet another aspect of the present invention there is provided a method of simultaneously predicting in vivo dissolution profiles of a plurality of pharmaceutical drug dosage form by comparison with a pc/, ~
~ 5158 reference drug dosage form having a known in vivo dissolution profile. A reference dissolution medium formulated as a mixture of at least two constituents is passed under the control of a mixture control signals through a reference dissolution cell containing the reference drug, at a flow rate controlled by a flow control signal and at an agitation rate within the-~ dissolution cell controlled by an agitation control signal. A plurality of dissolution media, each of which is formulated as a mixture of at least two constituents lb is flowed under the control of the mixture control signal, through a corresponding plurality of dissolution cells, each of which contains a different one of the plurality of drug dosage forms. A separate recirculating path is provided around each of the dissolution cells, each path having a recirculation rate controlled responsive to a sink control signal. The concentration profile of the reference drug is measured at the output of the reference dissolution cell and provides a time series of output signals corresponding thereto. A time series of stored signals corresponding to the concentration profile of the reference drug is provided, and a set of optimized process variable control signals optimized by deriving a difference signal between the time series of output signals and the time series of stored signals is produced. The mixture control signal, the flow control signal, the sink control signal and the agitation cantrol signals are produced such that the derived difference signal is continually minimized.
B~IEF DESCRIPTION OF THE DRAWINGS
Additional objects and advantages of the invention will become apparent to those skilled in the art as the description proceeds with reference to the accompanying drawings wherein:
FIGURE 1 is an overall block diagram of the - 8a -~ pC/'/f'`-' 5~58 automatic disso~ution testing system according to the present invention;
FIGURE 2 is a simplified block diagram of the closed loop programmed control of the' automatic dissolution testing system;
FIGURE 3 iS a simplified block diagram of the open loop control system implemented in the automatic dissolution testing system;
FIGURE 4 is an overall block diagram of the dissolution testing system expanded to show operation in the Internal Standard mode; and - 8b -pc/~
~` 1~5158 FIGURe 5 i8 an overall block diagram of the closed loop ~ontrol system for u8e with an improved bioavailability predictiv~ embodiment of the present invention.
D~TAILED DESCRIPTION OF~T~E P~EFERRED EMBODIMENT
Ref~rring now to Figure 1, there is shown an overall block diagram of the automatic flow-through dissblution testing 8yBtem ~ccording to the present invention; For simplicity of expo~ition the various elements are not shown to scale, and th~ embodiment ~hown is a basic one. The description~
throughout thi~ 8pecification are expressed in tenms of the t~sting of a pharmaceutical drug dosage fonm, and the language 18 accordingly ~pecific to this usage. Of course, the embodiments !
¦dl~clo~ed are illustrative and could readily be adapted for use l wlth agricultural products, or with aontrolled releAse components 1 15 ¦ ~ indu~trial proauct~ generally. The overall ~ystem 10 i9 shown ¦
I a- comprl~ed of a ba~ic di~solution cell 12 in which i9 positioned ¦ ~ sp~cimen of the drug product 14 undergoing evaluation. The 11 12 ha~ a f~lter membrane 12A and ~ilter screen 12B, and 1- proviaed w1th a flow of various di~olution l~quid via a cell 20 lnput line 16 under the influence of a primary pump 18. Output ~ro~ the dis~olutlon ¢ell 12 i~ carr-ed by a cell output line 20 and iB routed first via the upper portion of a recirculation line 2~, and th~reafter via ~ ~ystem output line 24. Within the r~¢irculation lin- 22 i~ A recirculation pump 26 wh~ch propels the llquid ther~ln ~nto the lower portion of the recirculation I ~ 5:~58 line 22, and thereafter into a.feeder li~e 28, which serves ~8 an input to the primary pump 18. Ths two arrows 30A and 30B ~bQw the dlre¢tion of flow in the rocirculation l~ne 22 under the influence of the recir¢ulating pump 26.
A flrst r~rvoir 32 i8 used to contain a supply of a ~ir~t als~olutlon medlum 34, whioh is fed via a line 36 to the dis~olut$on cell 12. The fir~t dis~olution medium 34, herelnafter alternately called tho ac~d media, is fed to the foeder lino 28 under the influenco of a pump 38. The flow ~irection of tho acid media 34 1B ahown by the flow arrows ~OA and 40B. A second reservoir 42 18 u~ed to contain a ~upply of ~ second,.dissolut~on medium 44, hereinafter . alternately called the alkaline medium 44. The al~aline medium 44 iB fed v~a a line 46 ~nd a check valve 4B to the ~eeder line 28, ~nd subQequently through the primary pump 28 to the dis~olution ¢ell 12. The ~low direction of the alkaline ~edla 44 i~ sho~n by the flow arrows 50A and SOB. A~ w~ll b~ discu~s~d in dQta11 ~elow, the flow path 40~ of the acid modia 34 1- due to the pre~ence of the check valve 48, and 20 th~ flow p~th 50B on the alkaline medium 44 is due to ths : ~yn~mics o~ the action of the two pumps 38 and 18.
~h~ system output line 24 ~erva~ to conduct the ~low of the prooo~sed medi~containing the de~ired con-~ontration o~ th~ di~solved drug product oot of the 8y8t~m, and further ~upport~ two Xey system me~surements. A flow - ~
~ 5158 `-¦ measurement device 52, serially pos~tioned in the output ¦ line 24, p~ovides a quantitative measurement of a liguid flow rate via a group of lines 54 to a central processor 56.
(Alternatlvely, this flow rate may be obtained electronically S as the differenco in control siynals to the primary pump 18 and th~ recycle pump 26.) A spectrophotometer 58, also ~rially position~a in the output line 24, provide~ a periodic (or continuous) measurement of the drug ooncentration in the output flow, and routes this measurement.Yia a group of lines 60 to the central processor 56. Pumps 18, 26 and 38 are of the posltive displacement peristaltic type and are capable of produclng precisQly controlled flow rates in the range of 0.2 to 140 ml per minute when properly controlled. The central processor 56 provides th~s control via signals on three group~ of lines 62, 64 and 66 which modulate the excltation to the pumps as follow~. ~ontrol signals on the line 62 ar~ applied to a pump speed modulator 68, which in turn controls th- excitation of the pump 26 via the lin~ 70;
control ~ignals on the line~ 64 are applied to a pump speed modulator 72, ~hich in turn controls th~ excitation of the pump 18 via th~ lin~s 74~ tho eontrol ~ignals on the line 68 ar0 appl~ed to a pump speed modulator 76, wh~ch in turn control~ the excltat$on to the pump 38 via the lines 78.
In addltion to tho abovs threo ~ontrol s~gnals, th~ central procossor 56 further provide~ control sighals to an ayitation ~ans comprlse~ o~ a stirring paddle 80 locat~d within the dl~solution cell 12 and po~itioned ~low the filter s~reen 12B.
5158 - l The~e control s~gnals are provided on a group of lines 82 to an lnterface device 84. An output from the interface device 84 i8 applled Yia lines 86 to an agitation motor 88, ~hlch in turn activates the stirring paddle 80 via the ~ech~nical linkage shown as dashed l~ne~ 90. A group of ~ontrol ana data lin~ 92 in~erconnect the central processor 56 with a nu~ber of support~ng unit~ 6hown a a peripherals group 94. Includ~d within thi~ group 94 would be a data r~corder 94A (analog and/or digital), an output printer 94B, and input keyboard 94C, and other well known and conventional ~avice~. An ~erall measurement ~lock 96 identifies those elements ¢onsidere~ to be mea~urement apparatus, as compared to tha remaining elementæ -- 56 and 94 -- which may be considered to be the signal proce~sing an~ control portions 5 o~ th~ ~ystem..
ln. U80, the system of Figure 1 carries o~t the ~i8801utlon t~ting under the control o~ the central processor 56 a~ fol}ows. ~y WAy of a brief overv~ew, the system shown ~ opera~le in tw~ modea, the first boing a simulatlve (or 2 clo#ea loop modo) and the second being a predictiv~ tor open loop mode). First, a known drug do~age form is used to . ~alibrate~ tho ~pparatu by operating it in the closed loop mod~ using known in vivo data tfrom-the re~order 94A) on th~ partl~ul~r drug do~age form to optimize a number of 2 kQy parametera (here~nafter alternately re~erred to a~ the .
5158 - l . I
.
proci~ varl~bl-o) with1n ehe centrnl proces~or 56. An it~rat~ve spt~mizing proce3s may be used to systemmatically ~odify the process variables until the difference ~etween the measured ~in vitro) data and the known (in vivo) data are m~nimized, and lndependent o time. Secondly, predictive tests of unknown drug dosage forms are performed using open loop control o~ the apparatus ~mploying the previously `~eterm~ned values of the process varia~les. To accomplish th~o steps, cont~ol ~proportional, differential and integral) 10 1B exe~cised over one or more of th~ process variables dotermined by: ~1) the composition: t2) the recycle flow of the dissolution medium: ~3) the total flow rate of the ~lssolution mediums and ~4) the rate of agitation within the di~solution cell.
Dur~ng the course of the closed loop phase of the opQrat~on, a change in the medium p~l can ~ used to simulate the in vlvo change from the ~tomaoh to the duodenum, and the recyoling of the medium to di~olution ce1.1 12 ~llows v~rlabl~ sink condit~on~ to be achieved ~o simulate the 20 ~xisting in vi~o condition~ due to differlng barrier properties of drug ab~orbing biological membranes. Res~Qtance to biological ab~orptlon i~ simulated by mixing the fresh media w1th the solutlon leaving the cell. The recycling of solution through the di~olut$on cell in thls m~nner decreases the 2 driving force for dis~olution. Upon establi~hing de~ired . 11~5158 constant flow rate at the outlet o the dissolution cell as determln~d by the flow measuring device 52, the time varying racycl~ flow rate, and a changin~ flow from the gastric and into~tinal ~uice reser~oirs are initiated.
The spectropho~ometer 58 provides a measurement of the ¢oncsntratlon of the drug in tho liquid leaving the cell. An alternate configuration may include plac~ng the spectro-photom~ter on the recycle flow line. This measured concen-tration value i8 compared within the central processor 56 1 ~ith the known in vivo bioavailability rate, blo~d level, urinary recovery rate, or pharmacolog~cal response versus t~me profiles being simulated. Any error s~gnal produced i8 converted within the central proce~sor 56 so as to opt~mize the process variables driving the error signal 15 to a minimum v~lue.
A detailed descr~ption of the operation of this y~tem of Fiqure 1 i8 facilitated wlth additional r~erence to th~ bloc~ aiagrams of Figure~ 2 and 3. Figure 2 shows a block diagram for the closed loop control of the dissolut~on 20 to8ting 8y8tem in the simulative mode: while Figure 3 shows a block diagram for open loop control of the dissolution testing y~em in the pr~dictive mode. ~oth thetlmevariableform: R(t), ~nd th~ tran~form varla~le form:R(s) of the system parameters ~ill be u~d herein tinterchangeably a~ required) as i8 25 ~11 known in the control system art. The followinq process t . variable~ are applicables l ~
``~ 5~58 - l 1 ., .1 .
.
l Q(t) ~ volumetric flow rate (milliliteis per mlnute) ¦ C~ ~ Q~ ~ velocity through tbe cell ~centimeters per minute) l AC ~ oross-sectional area of the cell tsquare S ¢Qntimeters) R(t) - volumetric flow of recycle ~mllliliters per minut~); l.e., it can be a constant or time- !
~arying quantity M~t) - stirring rate within one dissolution chamber 10 ¦ ~l(t) - fractlon of th~ solvent which is drawn from the re~ervoir contalning, for exa~ple, s~mulated il gastric juice, water, or ~n organic solvent t34 of FIG. 1) P2tt) ~ fraction of one solvent which is drawn from the 15 l~ re~ervoir containing, for ~xample, slmulated lnte~tinal ~uice, 1.0 normal sodium hydroxide, or an organic solvent t4~ of FIG. 1).
i Tho dis-olutlon process extant withlnthesystem of F~G. 1 ~ay commonly ~e described by a simple diffuslon layer model:
~ D T (C8 - C) ~here .
- di-solution rate ~milligram~ per minute) D ~ diffuslon coefflcient for the solvent and ~olute under con~ideration (square centimeter~ per minute)-25 I aff~cted by dissolut~on media composition ll S ~ surface area for dls~olution ~square centimeters) -¦i - an lntrln~lc property of the material tdo~age ¦¦ form) be~ng tested Il . . . . .
5158 ~-¦ C~ - concentration of solute required to saturate tho solvent (m~ll$grams p~r milliliter) - affected ¦ by d~ssolution media compo~ition I C - actual solute concentration in ~olution S ¦ (mllligram~ per m~ llter) - affe~ted by recycle flow and total volume flow' ¦ T ~ e~foct~ve thickness of the film or diffusion ¦ layer ~cent~meter~? - affeoted by agitation r~te l ' and total volume flow rate.
10¦ The relation~hip between variables $n the diffusion layer ~quation and the process variables are seen as given for a ~lxed valu~ of * :
¦ ~ - a function of Q and agitation rate ¦ C ~ a function of th~ volume of the dissolution I5 chamber, V, and the ~olumotric flow rate, .
Q, 1.~., the residenoe,t~me, V/Q, and the ' ' flow of recycle, R or R(t). To avo~d changing the volume of the dio~olutlon chamb~r by changing it~ l~ngth to chang~ C, thl~ could al~o bQ
' sffect~d by ~hanging the r~ycle flow CB ~ a fun¢tion of the prope~rtles'of thè solvent.
' For example, us~ng ~imulate~ gastric and intes-tinal ~uice~ mentioned previou~ly, the process varlables to be manipulat~d here are Fl and F2.
The variables D and C8 are obviously affected by th~ solvent~ u~sa and tha relative proportiOns , 1 11451S8 ,' . I .~
.
of Qach compos~ng the dissolution medium at any j tim~. When the solvent mixture is specified, D and C8 are al~o reflective of the properties I of the solid be1n~ dissolved.
5 ¦ S ~ in addition to being a function of Q, a function o~ the in~t~al~amount of drug, mO, and the physical properties of the solid. Once these variable~ are fixed, i.e., once a drug and a dosaga form are decided upo~, the time course o~ S as the experiment proceeds is reflective i of the properties of the drug product~.
ll Referring fir~t to FIG. 2, clo~ed loop operation ~simu-.
¦l lat~ve) in its bas~G form is shown as hav~ng an input signal ¦l A(S~ -- cumulative ~n vivo availabil~ty -- applied to ~n 15 i! ~nput node wher~ it is d~ff,erenced with a fed back s~gnal C~S) --~oncentration of the stream leaving th~ d~ssolution cell -- to produce an error slgnal E~S). The error signal E(S) is applied to the lnput of a proportional-integral-dor~vative tPID) ~ontroller havlng the gain charaoteristio,~alternately transfer funct~on) GC~8). The output from the PID controller ~8 a ~lgnal M~S) ~uitable to operate the respect~ve actuators in th~ sy~tom. Th~ ~our actuators o~ FIG. 1 include three proport~onal controlled pumps and one proportional controlled ! motor. ThQ~e ac~uators a~e re~resented ~imply as having a ' partlçular transfer function Gp/m~S), whose output~ are ¦! charact~riz~d by the four proce~ variables R~S), M~S), Q(S), !jan~ 8) ~~ all ~ aes~ribed below. ~ho proce~s variable~
¦ aro applied to th~ dissolutio~ eguipment having a transfer ! ~ , I
5~58 '' .
I
¦ funotion GDE~8~ whose output i8 the de~ired parameter C(S) or ¦ Q~tS) C(S) -- as doscribed below. 9riefly, FIG. 2 depicts in ~onventional analog-like term3 the simulative funct~on of the ~solutlon testing system. The analog-like descriptive S ¦ t~rminology is u8ed for s~mpliclty ana, of cour~e, digital ¦ mbo~iments may bo u~ed to impl~mont thQ contr~l leop contemplated!.
¦ Dur~ng thi8 c108-d loop ~imulativQ, or cal~bration run) opera-¦ ticn, threQ key paramsters within th~ PID oontroller are optimized ¦ a- ~Qscribed bolo~ su~h that ~ubsoguent open loop operat~on as 10 1 hown in FIG. 3 constitutes an optimally pied~ctive operating ¦mode. As shown in FIG. 3; the open loop control system block provides optimisod value~ of M(t), ~(t), Fl~t) and Qlt) to tho a~tuators tsorvo driven pumps, and/or motors) wh~ch in turn Ilmp~ct on the di-soiution ¢ell block to produce the desired ¦Output C~t) or Q8(t)-Ctt) as above de~cri~ed.
ll The ob~ective of the system o~ FIG. 1 ~s to obtain re-¦j~ults that uniformly rofl-ct the in vivo drug aVailability w~th optlmal fldellty ovor time and v~rying d~ug release b~havior of l th~ ~o~ago ~orm-. For Any given ~ot of ~rooess varia~les, i.e., 20 ¦ ~ Fl, tF2 ~ l-Fl), the ¢losed loop opera,tion of the in 1 vltro t~ting apparatus will produce a ~unction Qs itt)Ci~t~
l for each ith ~08ag~ form 80 that the expr~ssion 4 (t)-lO(t)-Rl~t)]C~tl] or lAi~t)-~(t)] elosely approx~mate~
I ~ro. Functlon~ Ri~t), ~t), Qitt) and Fll~t) wiil be obtained 25 j¦for Qach dosage ~orm of the dr~g te~tod that was chosen to !I poss~ss differont drug release dynamics. These functions can ! be ~ead out by the ¢entral proce~sor 56 onto magnetic tape, ¦rtored on magnetic d~o, or in the central memory of the micro-!processor during tho close~ loop operat~on o~ the apparatus.
¦ lB `, ~ 5158 ~ ~
¦ At th~ ~tage, the appar~tus merely simulates the ¦ A(t) functions determined ~rom in vlvo experimentation. Analo ¦ Ri(t) function signals recorded on magnetic tape for each dosa ¦ form can be conveniently processed on the central processor S¦ 56 and their values can be averaged, over dosage forms, at ¦ each time to obtain an averagQ, R~t1 funct~on represent~ng the mean behavior of all dosage form~ ~ncluded in the closed loop operation~. A second set of open loop runs must then be l performed for each'do~age form with the R(t), M(t), Fl(t), and 10¦ Q(t) functions programmed to control the proces~ variable~.
The num~er of closed loop runs performed on different dosaqe forms of the same drug and the resulting number of Ci(tj' functions and the corre~ponding number of Ri(t), Mi(t), Qi(t) and Fli(t) proces~ variables included in the R(t), M(t), Q(t) and Fl(t) funct$ons will depend on the propertie~ of any ¦ specific drug and the drug releaso characteri~tics of the ¦ dosage forms being tested. If the ~ynamic~ of the in vivo ¦ and in vltro ~ystem approximate llnoar behavior, then only ¦ one re~renoe dosage form i~-reguired. When aopropriate, 20 ¦ an ob~ective function Fo can be formed from the M(t), Fltt), ¦ Q(t), R~t), C~t) and A(t~ functions. A min~mal value of ¦ the ob~e~tive fun¢tion i~ a¢hieved by systematically selecting ¦ different ~olvents, geometries o~ the agitator, or if one or l more proces~ variables are kept constant, different fixed values of the proce~ variables not allowed to continuously vary with time. A minimum value of the ob~ective funct~cn correspond~ to opt~mal open loop operation of t~e apparatus under ~uch conditlons. ~8 mentioned, various means can be , .
11~51S8 implemented to control the recycle flow dynamics.
The system of F~gure 1 should be operated in the simplest manner that provide~ acceptable in vitro results ~ith regard to $n vivo drug av~ilabil~ty behavior. To d~term~no the magnitude of ~onsit$vity of the fid~lity of tho te~t to di~erQnt oporating con~itions,.the test can ba initially performed ~n succes~lvo phases of increasing ¢s~plexity And equipm~nt requiraments.
. Ph~so I can ~e performed without any autamatic control, using fixed, time invariant, values of the : process variables, M, Pl, Q, and R. An optimal composition and pH of the dissolution medium may be found and thereafter maintained cotlstant.
~hase II can be performed 3~milarly to Phase I but i5 w~th the incluQion of automatic control of M(t). as a proces6 vari~ble.
Pha~o I~I can employ automatlc computer control of a t~me-vary~ng d~ssolution modia ~omposition Fl~t) in addition to Mtt).
Pha~o TV can add R~t) as an automatîcally controlled proc~s variable.
. Phaso V cnn utilize N(t), Fl~t), ~t) and Q(t) as automati~ally controlled proces~ var~able~.
Th~ order ~n which automat~c control of the proces~ variables. i8 introduced dep-nds upon the propert$es of th~ ~rug, e.g., such ns its solub$1ity and intrinslc dissolution rats in differ~nt solvent~.
. -20-l -^v ~ 5158 ., ThesR su~moaes of operation can be repeatcd for . different dosag~ form3 of the same drug to obtain the optimal condition~ ov~r all re~erenoe do~ag~ forms. The ~implest : mode of op~ration po~s~ssing an acceptable fidelity would then be choson for future studi~s with the drug.
For a ~omewhat mor~ comprQhons~ve descrlption of the : ' mathemati¢al factors involved in the abo~e, th~ interested reader is rcfQrred to the aforementioned 1976 ~rticle authorec by tho inv~ntor. A more theoretic~l treatment of thQ
~0 relationship summarized above i8 al80 contained in an addi-tional papor - V.F.,Smolen,"Theor~tical and Computations ~a~is for Drug Bioavailability Determinations Using Pharma-cological Data II Drug Input = Re~ponse Relationships,"
J. Pharmaookinetic~ and ~iopharmaceutics, Yol. 4, No. 4, PP. 355-375, 1976.
-2~- ' ', .
.
l - -I ~45158 Referring to Figure 4, thero is shown an overall block diagram of an alternate embodiment of th~ pre~ent invention directed to produclng th~ desired prQdictive dissolution profile action in an lnternal Standard operating mode. The 0mbodiment shown in part~cularly advAntageous in thQ testing of a number of drug ~osage forms s~multaneously by comp~rison to a reference drug dosage form - and may be used for slmultaneously te~ting larg~ batches of a single drug do~age form, or of simultaneou~ly evaluating a number of different drug dosage forms. The apparatus 18 basically a par~llel arrangemont of a plurality of single flow-through di~solution sy~tems as shown in Figure 1, using a single , central proces~or/peripheral forcontrol. The Internal Standard ~ystem 100 i8 shown as compr$sed o the elements o~ the embodiment o~ Figure 1, in the form of a central proce~sor 56 intorconnected with a peripherals groups 94 via a group of llne~ 92. A trunk of input/output lines lOZ from the centra proces~or 56 are routed to a refor~nc- dis~olution testing ~ubsyst~m lOR, via a ~roup of input~outout lines 102~; and to a first unknown dissolution t-sting subsystem lOA via a group of output lines 102A~ and further to an.~Nth~ un~nown dis~olution t~tlng ub~y~tem lON via a group of output lines 102N. The nu~ber of independent dis~olution subsystem~ may bc fairly largo -- a dozen, or mor~ -- being l~m~ted by purely i5 perfunctory ¢onsiderations ~uch as cost and convenlQnce in usage. With contl~ued rQf~rencQ to Flgure 4 ana occa~lonal l ~ ~ ----11~5158 I
!~ reference to Figure 1, the ~ubsystems lOR, lOA, lON (of ! Figure 4) may be identical to tho measurement ~loc~ 96 Il (of F~gure 1). The sub~y~tem lOR, ln combination with the j central procossor 56, tho peripherals group 94 and the S lntorconneatlng linos 92, 102 and 102R constitute a dissolutic te~ting sy~tem ldentical to that of Fiqure 1, operat$ng in the ¢lo~d loop mode of operat~on a0 previously descri~ed.
¦I The sub~ystems lOA-lON fun¢tion in tho open loop mode as !I previous'y d~scrlbed. The primary operating difference 0" i8 that the N ~ubsystems containing an ~nknown drug dosage for ;' and operatlng opon looy are controllod simultaneou~ly by the ,~ ldentical ¢ontrol signal~ being generated by the control proces~or 56 respon~ive to the output measurements made on , tha referen¢e dissolutlon cell, a~ compared to the reference !~
~5 drug ln vlvo dissolution proile. Thus, the in vivo dls~olution profile being outputted by the recorder 94A as i a time Beries of known data, in comblnation with a time serie~
of control ~lgnal values produced by the`oantral proce~sor 56 ~ervo~ A~ an Intornal St~ndard ln the sense that the 2`01j predlctiv profiles are produc~d in the open loop mode by ~' ~ignals whlch are slmultaneously be~ng produced by closed i~ loop mo~e of operatlon uslng a ref~renoe drug and data a~ th ba-ls. A cursory revlew o~ the operation i of the ba~lc embodiment of Figure 1, a~ compared with that of Figure 4, will reconfirm that only comparatively minor differenco- ln operation of the subsystems are involved.
For example, the ~ubsystem~ l~A-lON have no need to perform , 11~5158 the measurement of flow rate and drug concentration in their output lines. Only the subsystem lOR requires that information. In the interest of the uniformity of apparatus, and as a means of providing additional versatility to the Internal Standard system 100, any or all of the su~systems lOA-lON may include the components required to measure these output parameters and provide related signals to the central processor 56. In this latter case, the central processor 56 is merely instructed to ignore the specific output data ~roduced by those particular subsystems which are to be operated open loop.
Summerizing, the Internal Standard emhodiment`of Figure 4 includes the signal processing elements (the central processor 56 and peripherals group 941 of Figure 1, along with a plurality of the measuring blocks 96 of Figure l. Of the number of measuring blocks, one (subsystem lOR) serves as a reference subsystem and operates in a closed loop mode with the signal processing elements, while the remainder (subsystems lOA to lON) are controlled by the signal processing elements in the open loop mode. Thus, the plurality of subsystems lOA-lON each produce a predictive dissolution profile of a separate drug dosage from while all are referenced to a single reference drug dosage form.
Referring now to Figure 5, there is shown a block diagram of an improved control system for use with the dissolution testing system 10. The improved control system 200 inserts an optimally adaptiYe capability into the dissolution testing process Yia a random input modeling (RIM) mode of operation. Briefly, this mode impacts on operation in the closed loop mode wherein the average in Yivo human drug response profile A(t), for a reference drug product is reproduced by the concentration vs. time profile, C~t), output from the apparatus through feedback control of one or ~. ,, pc/~
1~ ~5158more of the process variables controlling the conditions of the dissolution testing. Random input modeling is performed to tune a PID controller for each process variable and accomplish on-line, optimally adaptive control.
These process variables may include the composition (e.g., pH) of the dissolution medium; agitation via stirring paddles;
agitation via primary flow rate; and/or sink conditions in the form of recycle flow of medium back into the dissolution cell.
The improved control system 200 may be considered as an expanded version of the closed loop control system shown in more generalized form in FIG. 2. In FIG. 5, the improved control system 200 is shown as a four channel device wherein each channel corresponds to a particular process variable to be optimized. Thus, four proportional-integral-derivative (PID) controllers 202, 204, 206 and 208 have as their common inputs an error signal E(t) derived as the difference between the input signal A(t) and the output signal C(t). Each PID controller also has an indiYidual set of adjust lines taken from the group of parameter adjust lines 210. While the impro~ed control system 20Q is clearly , shown as being a digital embodiment, the specific apparatus uæed to implement the controlling has been deemphasized -except for a few places where digital to-analog (D/A) and analog-to-digital (A/D~ converters are needed - in order to better clarify the RIM technique which is the heart of the improvement being described. The particular parameters adjusted via the lines 210 are described below. Individual outputs from the four PID controllers are routed to a corresponding number of summing junctions 212, 214, 216 and 218, respectively; each summing junction also having a pseudo-random binary signal (PRBS) applied to it from a PRBS
generator 220, via a four section low pass filter 220A.
~. pc/, ,~.:
i~5158 Individual outputs from the four summing junctions are routed to a corresponding number of D/A converters 222, 224, 226 and 228; and are further routed via a group of lines 230 to other control elements within the parameter adjust section.
A set of individual analog control signals from the four D/A's are then applied to a corresponding number of actuators 232, 234, 236 and ~38 - which correspond to the various pumps/motors described in connection with FIG. 1. The correspondence is as follows: the pH actuator 232 may correspond to the pump 38 and its associated modulator; the agitation actuator 234 may correspond to the stirring paddle 80 and its associated motor; the primary flow actuator 236 may correspond to the primary pump 18 and its associated modulator; and the recycle actuator 238 may correspond to the recycle pump 26 and its associated modulator. The four actuators function, as previously described, to control the process variablès establishing the conditions of the dissolution testing resulting in an output concentration of the drug form detected by a spectrophotometer 242 (corresponding to the element 58 of FIG. 1). The concentration vs time profile C(t) - the desired output quantit~ - is digitized in an A/D converter 244 and is applied first via a path ~46 to an input mode 248 where it is differenced with the A~t) signals;
and further via a path 250 to the inputs of four pr~cess variable tuners 252, 254, 256 and 258. The tuners are substantially identical and hence the structure and function of one only will be described. The tuners may be implemented as a discrete collection of digital circuits operating under the control of a central processor (element 56 of FIG. l);
and may also be implemented via separate, but cooperating microprocessors: and may further be embedded in the central processor 56 itself. Tuner 258, the one associated with optimizing the recycle flow rate R~t), is shown as comprised pc/ ;' ~
-~1~5158 of a cross-correlator 260 to which is applied a pair of input signals on the lines 230A and 250A. The path 250A
provides the C(t) signal, while the path 230A provides a combined signal containing the control signal plus the random signal - from the output of summing junction 218.
The output from the cross-correlator is integrated in integrator 262, whose output is in turn applied to recycle curve element 264 which produces a ~ pc/;, ~
~1~5158 process reaction curve dir~cted to optimizin~ the recycle flow rate,parameter. A control parameter determinlng element 1 266 recoives the output from tho recycle curve element 264 and perlodicaliy produce~ updated ~alue~ for three key parameter~, wh~ch are appll~d via an lnterfacing element 268, ~or,us~ in the PID~controller 208. These three key para-, meters aro the overall controller gain Kc; the integral time TI; an,d th~ delay time TD. The intere~ted reader i9 referred l to a 1953 published article wherein the parameter-tuning 10', technique of the present invention i~ described. See, Cohen, G.H. and Coon, G.A., ~Theor~tical Conside,rations of ' Retarded Control,~ Tran~. ASME, Vol. 74, 1953, pp. 827.
jj The technique has hecome very well known in the control '~ syatem arts and iB referred to herelnafter as the "Cohen-Coon lS methodn.
1~ In order to improve the'f~elity of ~he bioavailabilit~
, prediat~on when tho improved contro~ ~ystem 200 1~ operating in the open loop mode, lt is u~eful to ~irst detormihe the I proper ¢ontrol parameter ~Xc, TI and ~D) settings for aach 20 I of the PID controller~ 202, ~04, 206 and 208. This i~ best dono by r~n~om in,put'modellng during an experimental run p~rformed w1th ~ raference drug form in the elo8ea loop mode ! f operation. The recycle rate control channel i8 illustrat~ve '~ of the mothnd usod. Super~mposed on each channel control signal i8 ~ pseudo-random binary signal with an amplitude " at lea~t on- standard deviation great-r than tho noise level , I .
, -28-, . .
.
- ¦ ~ the channel and a bandwidth corr~sponding to ten time~
¦ the bandwidth of the channel. Active electronic filteriny of the output ~rom the pRss generator 220 by a low pass fllter ¦ 220A may or m~y no~ b~ necessary due the pos~ibility of auto-S j fllter$ng by the mechanlcal damping ch~racterist~cs of theactlvators u~ed. Th~ ~lgnal output of the spectrophotometer 240 will contaln the result~ of th~ control signal plu8 the i PRBS. ~his ~lgnal C~t) 1~ cros~-correlated with the combined input signal tcontrol slgnal plu8 PRBS) over a period of ~illustratively) five time con~tants to yield ~ weighting ' function at the output of the cross-correlator 260. A process reaction CUrVQ i~ produced by integratlng the we~ghtlng function in the integrator 262, and the control parameters ~ ~, TI an~ TD) ~re determined by the method of Cohen and lS Cool with~n tha element 266. Those new controller ~attings for e~ch proce~s variable are substituted into tho four PID
.. . .
, controllers an~ the proce~s 1~ r-peated for another ~ive time ~on~tant~. The procedure 18 p-rformed lndepend~ntly 1~ for Qa~h of the four PID controllor- and s~multaneou~ly for 20 il all fou~ controller~ durlng each exporimental run.
Thi- proces~ o~ controller tuning ~s initlally performo~ lnd~p~ndently (an~/or ln combinatlon with one or " more of tho other proce~s variable~) for each process varl-~ able to ok~aln initial e~timates of ~ontroller settings for each controller. In the course of an actual run, the controller -;
~ .
?l SiS8 !~ for e~ch o~ the four proces3 variables operate together in I parallel and are retuned periodically ~every S t~me constants, 1 illustrativoly) simultaneously to provide new updated values li for the controller settings. Controller ~etting values S j obtained by this adaptivQ control pro~edure during the course of an experimental run are re~ected, and previous values retalned, if the proposed value~ are outside of a range of values for each ~ett~ng which had been found to induc ~ ¦ lnstability ln the operation of onet or more of the actuators.
io ' Although the ~nvention has been described in terms of ~elected preferred em~todiments and improvements to these ', em~odiments, the lnvention should not be deemed l$mited ; -!i thereto, slnce oth~r embodiments and modifications will readlly o¢cur to one skilled in the art. It is therefore S to be undesstood that the appended olaims are intended to oover all ~uch modificaelons as fall within the true spirit ~nd ~¢ope of the lnvention.
,!
.Ij' .
! .
. I
.. .
, . . ~ .
.
~ -30-:;
,; ,
rapidly performed dissolution test can substitute for the aeterminat~on of these physical propert~es by more time con-~uming and expensive methods such as x-ray cry~tallography, differential thermal analy~is, microscopy, etc. The materials S are instead dotermined as to whether they conform to a di~solution rate ~tandard u~der specifiea conditions and in relation to a known reference sample of the same material characterized by the above physical properties and po~sessing th~ aerived aissolution rate and ~olubll *y performance.
The broad technique of determining dissolution rate propert~e~ i8 e~pecially of intere~t in the testing of drug products where the therapeutic performance of drugs i8 closely related to the drug dissolution propertie~. Seemingly minor chan~ in drvg product formulation, as well as the inadvertent varlation in matorial~ and manufacture that can occur between batches of the same productformulatiop,caninfluence the therapeutic performance of drugs. In vivo bioavailability t-~ting of drug products in human~ provides the mo~t reliable m-ans of ensur~ng bioequivalence. Howev-r, it is impractical ~0 ~o per~orm the ~xtensive and expensive human testing that would be routinely required. ~arge numbers of buman sub~ect~
would be placed ae rl~k if uch studies were conducted.
~ioavailability testing in which humans are used ~ te~t ~ub~ects can be min~mized by the development and implementation o~ in vitro di~-olution standards that reflect in vivo drug-product erformahc~. In vitro b1oequivalence require~ent-I
G ~ 5158' have been established for ~ome drugs such as digoxin. From among the various ch~ical and physical tests that can be performed on drug solid~ in vitro for correlating or predicting a drug.produc~'s in v$vo bioavailabil~ty behavior, dissolution S tosting i8 th~ most sensitive and reliable. The correlative r~lation~hips ~08t commonly reported bstween in vitro dissolu-t~on and in v~vo bioavailability are of the single-point type:
the percentage of the drug di~solved in ~ given t~me (or the t~me it takes to di~solve a given percen~age of the drug in vi.tro) and ~ome univariate characteristlc of the drug product's ln vivo re~ponse versus time profile ~such as the peak blood level, the time required to reach the pe~k or 50% of the peak, or the area under the blood-level curves) are correlated.
The sele¢tion of in vitro dissolution and in vivo bioavaila- ¦
bility parameters for such single-polnt correlations is frequently arbi.trary, and the results can be misleading.
Obviou~ly, it would be preferable to predict the entire avorage ~lood level, urinary recovery rate, pharmacological-¦ re~ponse-time, or.drug absorption rate VJ. time profile that 20¦ w~uld be elicIt~d by a arug produ¢t in a panel of human ~ub~ects rather than merely to ¢orrelate univar~ate charac-t~ri~tics of th~ dlssolution proflle with an in v~vo b~o-availability parameter. In all cases, however, the fidelity of the in vltro dlssolutlon results in oorrelat~ng and in predicting in vivo drug-product bioava~l~bility depend~
_ dl3~01u~ion-t-st prooe ~ varlabl~ uch a the ~ ~- -1 1 ~ 51 5 ~
a$ssolut~on-medium composition, the solubility Yolume of *he meaium tsink conditions that determine the extent to which the med~um become~ saturated with the drug), and the agitation rates tstirr$ng or flow rates). An improper choice of these S proces~ varia~les te.g., an excess~vely h~gh rate of agitation) can ma~k signi icant ~ioavailability differences among drug products. On thQ other ~and, the dissolution test can be overly ~ansitive in detecting difference~ that are negligible in vivo.
~n the former case, using such improper dissolution-test parameters would result in the marketing of therapeutically in~ffective drug products. In the latter case, the result ~ould be the disca~ding of drug product-~ that are ent~rely ~atisfactory in terms o in vivo performance. Serious economic losse~ could re~ult from the use of an overly sensitive in ~ltro disso1ution te-t for lot-to-lot reproducibility testing of drug produ¢ts. Therefore, whether the dissolution te~t is b~$ng used a8 a quallty control tool, a~ an ln vivo bioequi-valoncy requir~m~nt ~or multisource generic drug proaucts, or ~8 a substltute for human bioavailability test~ng during : 20 ~ th development of new drug-product formulation~, it i~
~mperative that the dlssolut~on test provide predictive result~ that ar~ biolog~cally relevant.
Developing drug-product d~s~olut~on tests that pred~ct th- t~me course of drug-product bioavailability can ~o fraught with pitfalls, some of which may be avoided through knowle~go and consid~ra~tion of the phyQlochemical propertlss o~ the oomponents of the drug product and the biological proces~es and conditions operatlve ~n the r~lea~e .
"~ 5158 of the drug from the gastrointe~tinal tract and it~ subsequent absorption. ~owever, it is not only futile, but also unnecessary to attempt to reproduce the complex of biological factors operating in vivo in the effort to d~velop a satisfactory in vitro bioavailability test, although suoh attempts have been made. The devices that re~ulted from these efforts are o~ value now only a8 museum piece~. It wnuld, however, be imprudent to lgnore such knowledge when it can be u~ed aavan-tageou~ly to circumvent a problem in the ~eqign of a dissolution test.
There are two po~sible general approaches to developing ~n vivo relevant drug product dissolution te~ts. ~oth approaches s~ek to predict thë entire time course of average blood levels that would be ob~erved for a drug product in a panel of human te~t sub~ects. In thi~ way, the di~olution test serves as a ~ubstitute for human te~ti~g.
The first approach i8 a computational method that maximizes the amount of inormation that can be ob~ined from conventlorl~l methods of in vitro di~solution testing. U~ed 20 ~08t frequently are the USP rotating-basket apparatus, the FDA paadle method, the ~tationary-b~sket/r~tat1~g filter apparatus, Sartorlu~ solub11~ty and absorption simulators ~Sartoriu~, In¢orporated, Hayward, ~alifornia), ~nd column-type flow-through as~emblie~. The last of the~e device~
offers advantages ~1th regard to the definition, flexi~ility ~l~S158 of control, standardization, and reproducibility of process variables. Thi~ apparatus has been used by the inventor of the present ~nvention to demon~trate the second approach to predicting 1~ vivo blood-lsvol curves that em~rge from the apparatus in the form of dis~olution rate versu8 time profiles.
Since the computational approach with conventional ¦ -Apparatus d~pend~ upon the relatively arbitrary selection of process variables, its usefulness is limited. However, using feedback control to continuously vary the process variables, as described below, obviates this problem. For a more complete treatment of the mathematical (and theoretical) aspects of the d~ssolution, the interested reader is dirscted to three papers co-authored by the ~nventor. These are: V.F. Smolen et al, ~Optimally Predictive In Vitro Drug Dissolution Testing for In Vivo Bioavail~bility," J. Phsrmaceutical Sci., Vol. 65, No. 12, pp. 1718 1724, December 19i6; V.F. Smolen et al, "Predicting th~ T~m~ Cour~e of In Vivo Bioavailability ~rom In Vitro D~olution Tests: Control Systems Engineering Approaches,"
Pharmaceut~aal Technology, pp. 89-102, June 1979; and V.F.
8mo1en et al, ^Predictiv~ Cnnversion of In-Vivo Drug Dissolution Dat~ lnto In Vlvo Drug R~sponse Versus T~mQ Pro~les Exemplified ~or W~rfar~n,"' J. Pharmaceutical Sci., Vol. 66, ~o. 3, pp.
297-304, M~rch 1977.
The present invention is d~rected to an improved method 4nd apparatus for carrying out the dissolution approach to opt~mally prsdicting in vivo drug bloavailability from pharmaceutlcal ~osage forms, and other applications of dissolu-tlon te3ting.
~1..1~515~3 According to one aspect of the present invention there is provided an apparatus for the automated dissolution testing of products whose solubility and dissolution rate properties affect product performance, the apparatus having a dissolution cell with an input line, an output line, a filtering chamber therein, and controllable agitation means responsive to an agitation control signal, positioned within the chamber. At least two reservoirs are connected to the input line through a like number of media supply lines for providing dissolution media from the reservoirs to the filtering chamber. A first control means is responsive to a first control signal and positioned in the input line for controlling the flow of dissolution media through the filtering chamber, and second flow control means responsive to a second control signal and positioned in at least one of the media supply lines for controlling the flow of d ssolution media to the input line. The recirculating means include third flow control means responsive to a third control signal and connected to the input and output lines for controlling the recirculatory flow therefrom. System output means is connected to the output line and contains means for measuring output parameters therein including a concentration parameter of particular constituents within the flow, and for pxoviding system output signals corresponding to the measured output parameters. Memory means provides stored control signals to the apparatus. An electronic control means is connected to the memory means and is operative in a first mode responsive to the stored control signals and to the measured output parameter signals for automatically determining a set of internal control parameters used to determine the first, second and third control signals so as to minimize the differences between the stored and measured signals, and operative in an open loop mode responsive to the determined ~4 pc/'.,,.' internal control parameters for automatically producing the first, second and third control signals in the absence of the stored signals.
According to another aspect of the present inYention there is provided a method of automatic dissolution testing of products whose solubility and dissolution rate properties affect product performance having unkown dissolution profiles by comparison against known product forms having kno~n dissolution profiles. The method includes the steps of flowing a dissolution medium of controllable content at a controlled flo~ rate through a dissolution cell containing the known product form and providing a recirculation path around the dissolution cell to provide controllable sink conditions therethrough. The concentration profile of the known product form is measured at the output of the dissolution cell and output signals are provided corresponding thereto.
Stored signals corresponding to the concentration profile of the known product form are provided and a first family of control signals are produced such that the differences between the stored signals in the output signals are minimized.
A second family of control signals are produced for controlling the medium controlled content, the controlled flow rate, and the controlled sink conditions from the first family of control signals. A dissolution medium of the controlled content at the controlled flow rate is flowed through a ; dissolution cell containing the unknown product form so that measured concentration profiles of the unknown product form are derived as being predictive of subsequent usage dissolution profiles.
According to yet another aspect of the present invention there is provided a method of simultaneously predicting in vivo dissolution profiles of a plurality of pharmaceutical drug dosage form by comparison with a pc/, ~
~ 5158 reference drug dosage form having a known in vivo dissolution profile. A reference dissolution medium formulated as a mixture of at least two constituents is passed under the control of a mixture control signals through a reference dissolution cell containing the reference drug, at a flow rate controlled by a flow control signal and at an agitation rate within the-~ dissolution cell controlled by an agitation control signal. A plurality of dissolution media, each of which is formulated as a mixture of at least two constituents lb is flowed under the control of the mixture control signal, through a corresponding plurality of dissolution cells, each of which contains a different one of the plurality of drug dosage forms. A separate recirculating path is provided around each of the dissolution cells, each path having a recirculation rate controlled responsive to a sink control signal. The concentration profile of the reference drug is measured at the output of the reference dissolution cell and provides a time series of output signals corresponding thereto. A time series of stored signals corresponding to the concentration profile of the reference drug is provided, and a set of optimized process variable control signals optimized by deriving a difference signal between the time series of output signals and the time series of stored signals is produced. The mixture control signal, the flow control signal, the sink control signal and the agitation cantrol signals are produced such that the derived difference signal is continually minimized.
B~IEF DESCRIPTION OF THE DRAWINGS
Additional objects and advantages of the invention will become apparent to those skilled in the art as the description proceeds with reference to the accompanying drawings wherein:
FIGURE 1 is an overall block diagram of the - 8a -~ pC/'/f'`-' 5~58 automatic disso~ution testing system according to the present invention;
FIGURE 2 is a simplified block diagram of the closed loop programmed control of the' automatic dissolution testing system;
FIGURE 3 iS a simplified block diagram of the open loop control system implemented in the automatic dissolution testing system;
FIGURE 4 is an overall block diagram of the dissolution testing system expanded to show operation in the Internal Standard mode; and - 8b -pc/~
~` 1~5158 FIGURe 5 i8 an overall block diagram of the closed loop ~ontrol system for u8e with an improved bioavailability predictiv~ embodiment of the present invention.
D~TAILED DESCRIPTION OF~T~E P~EFERRED EMBODIMENT
Ref~rring now to Figure 1, there is shown an overall block diagram of the automatic flow-through dissblution testing 8yBtem ~ccording to the present invention; For simplicity of expo~ition the various elements are not shown to scale, and th~ embodiment ~hown is a basic one. The description~
throughout thi~ 8pecification are expressed in tenms of the t~sting of a pharmaceutical drug dosage fonm, and the language 18 accordingly ~pecific to this usage. Of course, the embodiments !
¦dl~clo~ed are illustrative and could readily be adapted for use l wlth agricultural products, or with aontrolled releAse components 1 15 ¦ ~ indu~trial proauct~ generally. The overall ~ystem 10 i9 shown ¦
I a- comprl~ed of a ba~ic di~solution cell 12 in which i9 positioned ¦ ~ sp~cimen of the drug product 14 undergoing evaluation. The 11 12 ha~ a f~lter membrane 12A and ~ilter screen 12B, and 1- proviaed w1th a flow of various di~olution l~quid via a cell 20 lnput line 16 under the influence of a primary pump 18. Output ~ro~ the dis~olutlon ¢ell 12 i~ carr-ed by a cell output line 20 and iB routed first via the upper portion of a recirculation line 2~, and th~reafter via ~ ~ystem output line 24. Within the r~¢irculation lin- 22 i~ A recirculation pump 26 wh~ch propels the llquid ther~ln ~nto the lower portion of the recirculation I ~ 5:~58 line 22, and thereafter into a.feeder li~e 28, which serves ~8 an input to the primary pump 18. Ths two arrows 30A and 30B ~bQw the dlre¢tion of flow in the rocirculation l~ne 22 under the influence of the recir¢ulating pump 26.
A flrst r~rvoir 32 i8 used to contain a supply of a ~ir~t als~olutlon medlum 34, whioh is fed via a line 36 to the dis~olut$on cell 12. The fir~t dis~olution medium 34, herelnafter alternately called tho ac~d media, is fed to the foeder lino 28 under the influenco of a pump 38. The flow ~irection of tho acid media 34 1B ahown by the flow arrows ~OA and 40B. A second reservoir 42 18 u~ed to contain a ~upply of ~ second,.dissolut~on medium 44, hereinafter . alternately called the alkaline medium 44. The al~aline medium 44 iB fed v~a a line 46 ~nd a check valve 4B to the ~eeder line 28, ~nd subQequently through the primary pump 28 to the dis~olution ¢ell 12. The ~low direction of the alkaline ~edla 44 i~ sho~n by the flow arrows 50A and SOB. A~ w~ll b~ discu~s~d in dQta11 ~elow, the flow path 40~ of the acid modia 34 1- due to the pre~ence of the check valve 48, and 20 th~ flow p~th 50B on the alkaline medium 44 is due to ths : ~yn~mics o~ the action of the two pumps 38 and 18.
~h~ system output line 24 ~erva~ to conduct the ~low of the prooo~sed medi~containing the de~ired con-~ontration o~ th~ di~solved drug product oot of the 8y8t~m, and further ~upport~ two Xey system me~surements. A flow - ~
~ 5158 `-¦ measurement device 52, serially pos~tioned in the output ¦ line 24, p~ovides a quantitative measurement of a liguid flow rate via a group of lines 54 to a central processor 56.
(Alternatlvely, this flow rate may be obtained electronically S as the differenco in control siynals to the primary pump 18 and th~ recycle pump 26.) A spectrophotometer 58, also ~rially position~a in the output line 24, provide~ a periodic (or continuous) measurement of the drug ooncentration in the output flow, and routes this measurement.Yia a group of lines 60 to the central processor 56. Pumps 18, 26 and 38 are of the posltive displacement peristaltic type and are capable of produclng precisQly controlled flow rates in the range of 0.2 to 140 ml per minute when properly controlled. The central processor 56 provides th~s control via signals on three group~ of lines 62, 64 and 66 which modulate the excltation to the pumps as follow~. ~ontrol signals on the line 62 ar~ applied to a pump speed modulator 68, which in turn controls th- excitation of the pump 26 via the lin~ 70;
control ~ignals on the line~ 64 are applied to a pump speed modulator 72, ~hich in turn controls th~ excitation of the pump 18 via th~ lin~s 74~ tho eontrol ~ignals on the line 68 ar0 appl~ed to a pump speed modulator 76, wh~ch in turn control~ the excltat$on to the pump 38 via the lines 78.
In addltion to tho abovs threo ~ontrol s~gnals, th~ central procossor 56 further provide~ control sighals to an ayitation ~ans comprlse~ o~ a stirring paddle 80 locat~d within the dl~solution cell 12 and po~itioned ~low the filter s~reen 12B.
5158 - l The~e control s~gnals are provided on a group of lines 82 to an lnterface device 84. An output from the interface device 84 i8 applled Yia lines 86 to an agitation motor 88, ~hlch in turn activates the stirring paddle 80 via the ~ech~nical linkage shown as dashed l~ne~ 90. A group of ~ontrol ana data lin~ 92 in~erconnect the central processor 56 with a nu~ber of support~ng unit~ 6hown a a peripherals group 94. Includ~d within thi~ group 94 would be a data r~corder 94A (analog and/or digital), an output printer 94B, and input keyboard 94C, and other well known and conventional ~avice~. An ~erall measurement ~lock 96 identifies those elements ¢onsidere~ to be mea~urement apparatus, as compared to tha remaining elementæ -- 56 and 94 -- which may be considered to be the signal proce~sing an~ control portions 5 o~ th~ ~ystem..
ln. U80, the system of Figure 1 carries o~t the ~i8801utlon t~ting under the control o~ the central processor 56 a~ fol}ows. ~y WAy of a brief overv~ew, the system shown ~ opera~le in tw~ modea, the first boing a simulatlve (or 2 clo#ea loop modo) and the second being a predictiv~ tor open loop mode). First, a known drug do~age form is used to . ~alibrate~ tho ~pparatu by operating it in the closed loop mod~ using known in vivo data tfrom-the re~order 94A) on th~ partl~ul~r drug do~age form to optimize a number of 2 kQy parametera (here~nafter alternately re~erred to a~ the .
5158 - l . I
.
proci~ varl~bl-o) with1n ehe centrnl proces~or 56. An it~rat~ve spt~mizing proce3s may be used to systemmatically ~odify the process variables until the difference ~etween the measured ~in vitro) data and the known (in vivo) data are m~nimized, and lndependent o time. Secondly, predictive tests of unknown drug dosage forms are performed using open loop control o~ the apparatus ~mploying the previously `~eterm~ned values of the process varia~les. To accomplish th~o steps, cont~ol ~proportional, differential and integral) 10 1B exe~cised over one or more of th~ process variables dotermined by: ~1) the composition: t2) the recycle flow of the dissolution medium: ~3) the total flow rate of the ~lssolution mediums and ~4) the rate of agitation within the di~solution cell.
Dur~ng the course of the closed loop phase of the opQrat~on, a change in the medium p~l can ~ used to simulate the in vlvo change from the ~tomaoh to the duodenum, and the recyoling of the medium to di~olution ce1.1 12 ~llows v~rlabl~ sink condit~on~ to be achieved ~o simulate the 20 ~xisting in vi~o condition~ due to differlng barrier properties of drug ab~orbing biological membranes. Res~Qtance to biological ab~orptlon i~ simulated by mixing the fresh media w1th the solutlon leaving the cell. The recycling of solution through the di~olut$on cell in thls m~nner decreases the 2 driving force for dis~olution. Upon establi~hing de~ired . 11~5158 constant flow rate at the outlet o the dissolution cell as determln~d by the flow measuring device 52, the time varying racycl~ flow rate, and a changin~ flow from the gastric and into~tinal ~uice reser~oirs are initiated.
The spectropho~ometer 58 provides a measurement of the ¢oncsntratlon of the drug in tho liquid leaving the cell. An alternate configuration may include plac~ng the spectro-photom~ter on the recycle flow line. This measured concen-tration value i8 compared within the central processor 56 1 ~ith the known in vivo bioavailability rate, blo~d level, urinary recovery rate, or pharmacolog~cal response versus t~me profiles being simulated. Any error s~gnal produced i8 converted within the central proce~sor 56 so as to opt~mize the process variables driving the error signal 15 to a minimum v~lue.
A detailed descr~ption of the operation of this y~tem of Fiqure 1 i8 facilitated wlth additional r~erence to th~ bloc~ aiagrams of Figure~ 2 and 3. Figure 2 shows a block diagram for the closed loop control of the dissolut~on 20 to8ting 8y8tem in the simulative mode: while Figure 3 shows a block diagram for open loop control of the dissolution testing y~em in the pr~dictive mode. ~oth thetlmevariableform: R(t), ~nd th~ tran~form varla~le form:R(s) of the system parameters ~ill be u~d herein tinterchangeably a~ required) as i8 25 ~11 known in the control system art. The followinq process t . variable~ are applicables l ~
``~ 5~58 - l 1 ., .1 .
.
l Q(t) ~ volumetric flow rate (milliliteis per mlnute) ¦ C~ ~ Q~ ~ velocity through tbe cell ~centimeters per minute) l AC ~ oross-sectional area of the cell tsquare S ¢Qntimeters) R(t) - volumetric flow of recycle ~mllliliters per minut~); l.e., it can be a constant or time- !
~arying quantity M~t) - stirring rate within one dissolution chamber 10 ¦ ~l(t) - fractlon of th~ solvent which is drawn from the re~ervoir contalning, for exa~ple, s~mulated il gastric juice, water, or ~n organic solvent t34 of FIG. 1) P2tt) ~ fraction of one solvent which is drawn from the 15 l~ re~ervoir containing, for ~xample, slmulated lnte~tinal ~uice, 1.0 normal sodium hydroxide, or an organic solvent t4~ of FIG. 1).
i Tho dis-olutlon process extant withlnthesystem of F~G. 1 ~ay commonly ~e described by a simple diffuslon layer model:
~ D T (C8 - C) ~here .
- di-solution rate ~milligram~ per minute) D ~ diffuslon coefflcient for the solvent and ~olute under con~ideration (square centimeter~ per minute)-25 I aff~cted by dissolut~on media composition ll S ~ surface area for dls~olution ~square centimeters) -¦i - an lntrln~lc property of the material tdo~age ¦¦ form) be~ng tested Il . . . . .
5158 ~-¦ C~ - concentration of solute required to saturate tho solvent (m~ll$grams p~r milliliter) - affected ¦ by d~ssolution media compo~ition I C - actual solute concentration in ~olution S ¦ (mllligram~ per m~ llter) - affe~ted by recycle flow and total volume flow' ¦ T ~ e~foct~ve thickness of the film or diffusion ¦ layer ~cent~meter~? - affeoted by agitation r~te l ' and total volume flow rate.
10¦ The relation~hip between variables $n the diffusion layer ~quation and the process variables are seen as given for a ~lxed valu~ of * :
¦ ~ - a function of Q and agitation rate ¦ C ~ a function of th~ volume of the dissolution I5 chamber, V, and the ~olumotric flow rate, .
Q, 1.~., the residenoe,t~me, V/Q, and the ' ' flow of recycle, R or R(t). To avo~d changing the volume of the dio~olutlon chamb~r by changing it~ l~ngth to chang~ C, thl~ could al~o bQ
' sffect~d by ~hanging the r~ycle flow CB ~ a fun¢tion of the prope~rtles'of thè solvent.
' For example, us~ng ~imulate~ gastric and intes-tinal ~uice~ mentioned previou~ly, the process varlables to be manipulat~d here are Fl and F2.
The variables D and C8 are obviously affected by th~ solvent~ u~sa and tha relative proportiOns , 1 11451S8 ,' . I .~
.
of Qach compos~ng the dissolution medium at any j tim~. When the solvent mixture is specified, D and C8 are al~o reflective of the properties I of the solid be1n~ dissolved.
5 ¦ S ~ in addition to being a function of Q, a function o~ the in~t~al~amount of drug, mO, and the physical properties of the solid. Once these variable~ are fixed, i.e., once a drug and a dosaga form are decided upo~, the time course o~ S as the experiment proceeds is reflective i of the properties of the drug product~.
ll Referring fir~t to FIG. 2, clo~ed loop operation ~simu-.
¦l lat~ve) in its bas~G form is shown as hav~ng an input signal ¦l A(S~ -- cumulative ~n vivo availabil~ty -- applied to ~n 15 i! ~nput node wher~ it is d~ff,erenced with a fed back s~gnal C~S) --~oncentration of the stream leaving th~ d~ssolution cell -- to produce an error slgnal E~S). The error signal E(S) is applied to the lnput of a proportional-integral-dor~vative tPID) ~ontroller havlng the gain charaoteristio,~alternately transfer funct~on) GC~8). The output from the PID controller ~8 a ~lgnal M~S) ~uitable to operate the respect~ve actuators in th~ sy~tom. Th~ ~our actuators o~ FIG. 1 include three proport~onal controlled pumps and one proportional controlled ! motor. ThQ~e ac~uators a~e re~resented ~imply as having a ' partlçular transfer function Gp/m~S), whose output~ are ¦! charact~riz~d by the four proce~ variables R~S), M~S), Q(S), !jan~ 8) ~~ all ~ aes~ribed below. ~ho proce~s variable~
¦ aro applied to th~ dissolutio~ eguipment having a transfer ! ~ , I
5~58 '' .
I
¦ funotion GDE~8~ whose output i8 the de~ired parameter C(S) or ¦ Q~tS) C(S) -- as doscribed below. 9riefly, FIG. 2 depicts in ~onventional analog-like term3 the simulative funct~on of the ~solutlon testing system. The analog-like descriptive S ¦ t~rminology is u8ed for s~mpliclty ana, of cour~e, digital ¦ mbo~iments may bo u~ed to impl~mont thQ contr~l leop contemplated!.
¦ Dur~ng thi8 c108-d loop ~imulativQ, or cal~bration run) opera-¦ ticn, threQ key paramsters within th~ PID oontroller are optimized ¦ a- ~Qscribed bolo~ su~h that ~ubsoguent open loop operat~on as 10 1 hown in FIG. 3 constitutes an optimally pied~ctive operating ¦mode. As shown in FIG. 3; the open loop control system block provides optimisod value~ of M(t), ~(t), Fl~t) and Qlt) to tho a~tuators tsorvo driven pumps, and/or motors) wh~ch in turn Ilmp~ct on the di-soiution ¢ell block to produce the desired ¦Output C~t) or Q8(t)-Ctt) as above de~cri~ed.
ll The ob~ective of the system o~ FIG. 1 ~s to obtain re-¦j~ults that uniformly rofl-ct the in vivo drug aVailability w~th optlmal fldellty ovor time and v~rying d~ug release b~havior of l th~ ~o~ago ~orm-. For Any given ~ot of ~rooess varia~les, i.e., 20 ¦ ~ Fl, tF2 ~ l-Fl), the ¢losed loop opera,tion of the in 1 vltro t~ting apparatus will produce a ~unction Qs itt)Ci~t~
l for each ith ~08ag~ form 80 that the expr~ssion 4 (t)-lO(t)-Rl~t)]C~tl] or lAi~t)-~(t)] elosely approx~mate~
I ~ro. Functlon~ Ri~t), ~t), Qitt) and Fll~t) wiil be obtained 25 j¦for Qach dosage ~orm of the dr~g te~tod that was chosen to !I poss~ss differont drug release dynamics. These functions can ! be ~ead out by the ¢entral proce~sor 56 onto magnetic tape, ¦rtored on magnetic d~o, or in the central memory of the micro-!processor during tho close~ loop operat~on o~ the apparatus.
¦ lB `, ~ 5158 ~ ~
¦ At th~ ~tage, the appar~tus merely simulates the ¦ A(t) functions determined ~rom in vlvo experimentation. Analo ¦ Ri(t) function signals recorded on magnetic tape for each dosa ¦ form can be conveniently processed on the central processor S¦ 56 and their values can be averaged, over dosage forms, at ¦ each time to obtain an averagQ, R~t1 funct~on represent~ng the mean behavior of all dosage form~ ~ncluded in the closed loop operation~. A second set of open loop runs must then be l performed for each'do~age form with the R(t), M(t), Fl(t), and 10¦ Q(t) functions programmed to control the proces~ variable~.
The num~er of closed loop runs performed on different dosaqe forms of the same drug and the resulting number of Ci(tj' functions and the corre~ponding number of Ri(t), Mi(t), Qi(t) and Fli(t) proces~ variables included in the R(t), M(t), Q(t) and Fl(t) funct$ons will depend on the propertie~ of any ¦ specific drug and the drug releaso characteri~tics of the ¦ dosage forms being tested. If the ~ynamic~ of the in vivo ¦ and in vltro ~ystem approximate llnoar behavior, then only ¦ one re~renoe dosage form i~-reguired. When aopropriate, 20 ¦ an ob~ective function Fo can be formed from the M(t), Fltt), ¦ Q(t), R~t), C~t) and A(t~ functions. A min~mal value of ¦ the ob~e~tive fun¢tion i~ a¢hieved by systematically selecting ¦ different ~olvents, geometries o~ the agitator, or if one or l more proces~ variables are kept constant, different fixed values of the proce~ variables not allowed to continuously vary with time. A minimum value of the ob~ective funct~cn correspond~ to opt~mal open loop operation of t~e apparatus under ~uch conditlons. ~8 mentioned, various means can be , .
11~51S8 implemented to control the recycle flow dynamics.
The system of F~gure 1 should be operated in the simplest manner that provide~ acceptable in vitro results ~ith regard to $n vivo drug av~ilabil~ty behavior. To d~term~no the magnitude of ~onsit$vity of the fid~lity of tho te~t to di~erQnt oporating con~itions,.the test can ba initially performed ~n succes~lvo phases of increasing ¢s~plexity And equipm~nt requiraments.
. Ph~so I can ~e performed without any autamatic control, using fixed, time invariant, values of the : process variables, M, Pl, Q, and R. An optimal composition and pH of the dissolution medium may be found and thereafter maintained cotlstant.
~hase II can be performed 3~milarly to Phase I but i5 w~th the incluQion of automatic control of M(t). as a proces6 vari~ble.
Pha~o I~I can employ automatlc computer control of a t~me-vary~ng d~ssolution modia ~omposition Fl~t) in addition to Mtt).
Pha~o TV can add R~t) as an automatîcally controlled proc~s variable.
. Phaso V cnn utilize N(t), Fl~t), ~t) and Q(t) as automati~ally controlled proces~ var~able~.
Th~ order ~n which automat~c control of the proces~ variables. i8 introduced dep-nds upon the propert$es of th~ ~rug, e.g., such ns its solub$1ity and intrinslc dissolution rats in differ~nt solvent~.
. -20-l -^v ~ 5158 ., ThesR su~moaes of operation can be repeatcd for . different dosag~ form3 of the same drug to obtain the optimal condition~ ov~r all re~erenoe do~ag~ forms. The ~implest : mode of op~ration po~s~ssing an acceptable fidelity would then be choson for future studi~s with the drug.
For a ~omewhat mor~ comprQhons~ve descrlption of the : ' mathemati¢al factors involved in the abo~e, th~ interested reader is rcfQrred to the aforementioned 1976 ~rticle authorec by tho inv~ntor. A more theoretic~l treatment of thQ
~0 relationship summarized above i8 al80 contained in an addi-tional papor - V.F.,Smolen,"Theor~tical and Computations ~a~is for Drug Bioavailability Determinations Using Pharma-cological Data II Drug Input = Re~ponse Relationships,"
J. Pharmaookinetic~ and ~iopharmaceutics, Yol. 4, No. 4, PP. 355-375, 1976.
-2~- ' ', .
.
l - -I ~45158 Referring to Figure 4, thero is shown an overall block diagram of an alternate embodiment of th~ pre~ent invention directed to produclng th~ desired prQdictive dissolution profile action in an lnternal Standard operating mode. The 0mbodiment shown in part~cularly advAntageous in thQ testing of a number of drug ~osage forms s~multaneously by comp~rison to a reference drug dosage form - and may be used for slmultaneously te~ting larg~ batches of a single drug do~age form, or of simultaneou~ly evaluating a number of different drug dosage forms. The apparatus 18 basically a par~llel arrangemont of a plurality of single flow-through di~solution sy~tems as shown in Figure 1, using a single , central proces~or/peripheral forcontrol. The Internal Standard ~ystem 100 i8 shown as compr$sed o the elements o~ the embodiment o~ Figure 1, in the form of a central proce~sor 56 intorconnected with a peripherals groups 94 via a group of llne~ 92. A trunk of input/output lines lOZ from the centra proces~or 56 are routed to a refor~nc- dis~olution testing ~ubsyst~m lOR, via a ~roup of input~outout lines 102~; and to a first unknown dissolution t-sting subsystem lOA via a group of output lines 102A~ and further to an.~Nth~ un~nown dis~olution t~tlng ub~y~tem lON via a group of output lines 102N. The nu~ber of independent dis~olution subsystem~ may bc fairly largo -- a dozen, or mor~ -- being l~m~ted by purely i5 perfunctory ¢onsiderations ~uch as cost and convenlQnce in usage. With contl~ued rQf~rencQ to Flgure 4 ana occa~lonal l ~ ~ ----11~5158 I
!~ reference to Figure 1, the ~ubsystems lOR, lOA, lON (of ! Figure 4) may be identical to tho measurement ~loc~ 96 Il (of F~gure 1). The sub~y~tem lOR, ln combination with the j central procossor 56, tho peripherals group 94 and the S lntorconneatlng linos 92, 102 and 102R constitute a dissolutic te~ting sy~tem ldentical to that of Fiqure 1, operat$ng in the ¢lo~d loop mode of operat~on a0 previously descri~ed.
¦I The sub~ystems lOA-lON fun¢tion in tho open loop mode as !I previous'y d~scrlbed. The primary operating difference 0" i8 that the N ~ubsystems containing an ~nknown drug dosage for ;' and operatlng opon looy are controllod simultaneou~ly by the ,~ ldentical ¢ontrol signal~ being generated by the control proces~or 56 respon~ive to the output measurements made on , tha referen¢e dissolutlon cell, a~ compared to the reference !~
~5 drug ln vlvo dissolution proile. Thus, the in vivo dls~olution profile being outputted by the recorder 94A as i a time Beries of known data, in comblnation with a time serie~
of control ~lgnal values produced by the`oantral proce~sor 56 ~ervo~ A~ an Intornal St~ndard ln the sense that the 2`01j predlctiv profiles are produc~d in the open loop mode by ~' ~ignals whlch are slmultaneously be~ng produced by closed i~ loop mo~e of operatlon uslng a ref~renoe drug and data a~ th ba-ls. A cursory revlew o~ the operation i of the ba~lc embodiment of Figure 1, a~ compared with that of Figure 4, will reconfirm that only comparatively minor differenco- ln operation of the subsystems are involved.
For example, the ~ubsystem~ l~A-lON have no need to perform , 11~5158 the measurement of flow rate and drug concentration in their output lines. Only the subsystem lOR requires that information. In the interest of the uniformity of apparatus, and as a means of providing additional versatility to the Internal Standard system 100, any or all of the su~systems lOA-lON may include the components required to measure these output parameters and provide related signals to the central processor 56. In this latter case, the central processor 56 is merely instructed to ignore the specific output data ~roduced by those particular subsystems which are to be operated open loop.
Summerizing, the Internal Standard emhodiment`of Figure 4 includes the signal processing elements (the central processor 56 and peripherals group 941 of Figure 1, along with a plurality of the measuring blocks 96 of Figure l. Of the number of measuring blocks, one (subsystem lOR) serves as a reference subsystem and operates in a closed loop mode with the signal processing elements, while the remainder (subsystems lOA to lON) are controlled by the signal processing elements in the open loop mode. Thus, the plurality of subsystems lOA-lON each produce a predictive dissolution profile of a separate drug dosage from while all are referenced to a single reference drug dosage form.
Referring now to Figure 5, there is shown a block diagram of an improved control system for use with the dissolution testing system 10. The improved control system 200 inserts an optimally adaptiYe capability into the dissolution testing process Yia a random input modeling (RIM) mode of operation. Briefly, this mode impacts on operation in the closed loop mode wherein the average in Yivo human drug response profile A(t), for a reference drug product is reproduced by the concentration vs. time profile, C~t), output from the apparatus through feedback control of one or ~. ,, pc/~
1~ ~5158more of the process variables controlling the conditions of the dissolution testing. Random input modeling is performed to tune a PID controller for each process variable and accomplish on-line, optimally adaptive control.
These process variables may include the composition (e.g., pH) of the dissolution medium; agitation via stirring paddles;
agitation via primary flow rate; and/or sink conditions in the form of recycle flow of medium back into the dissolution cell.
The improved control system 200 may be considered as an expanded version of the closed loop control system shown in more generalized form in FIG. 2. In FIG. 5, the improved control system 200 is shown as a four channel device wherein each channel corresponds to a particular process variable to be optimized. Thus, four proportional-integral-derivative (PID) controllers 202, 204, 206 and 208 have as their common inputs an error signal E(t) derived as the difference between the input signal A(t) and the output signal C(t). Each PID controller also has an indiYidual set of adjust lines taken from the group of parameter adjust lines 210. While the impro~ed control system 20Q is clearly , shown as being a digital embodiment, the specific apparatus uæed to implement the controlling has been deemphasized -except for a few places where digital to-analog (D/A) and analog-to-digital (A/D~ converters are needed - in order to better clarify the RIM technique which is the heart of the improvement being described. The particular parameters adjusted via the lines 210 are described below. Individual outputs from the four PID controllers are routed to a corresponding number of summing junctions 212, 214, 216 and 218, respectively; each summing junction also having a pseudo-random binary signal (PRBS) applied to it from a PRBS
generator 220, via a four section low pass filter 220A.
~. pc/, ,~.:
i~5158 Individual outputs from the four summing junctions are routed to a corresponding number of D/A converters 222, 224, 226 and 228; and are further routed via a group of lines 230 to other control elements within the parameter adjust section.
A set of individual analog control signals from the four D/A's are then applied to a corresponding number of actuators 232, 234, 236 and ~38 - which correspond to the various pumps/motors described in connection with FIG. 1. The correspondence is as follows: the pH actuator 232 may correspond to the pump 38 and its associated modulator; the agitation actuator 234 may correspond to the stirring paddle 80 and its associated motor; the primary flow actuator 236 may correspond to the primary pump 18 and its associated modulator; and the recycle actuator 238 may correspond to the recycle pump 26 and its associated modulator. The four actuators function, as previously described, to control the process variablès establishing the conditions of the dissolution testing resulting in an output concentration of the drug form detected by a spectrophotometer 242 (corresponding to the element 58 of FIG. 1). The concentration vs time profile C(t) - the desired output quantit~ - is digitized in an A/D converter 244 and is applied first via a path ~46 to an input mode 248 where it is differenced with the A~t) signals;
and further via a path 250 to the inputs of four pr~cess variable tuners 252, 254, 256 and 258. The tuners are substantially identical and hence the structure and function of one only will be described. The tuners may be implemented as a discrete collection of digital circuits operating under the control of a central processor (element 56 of FIG. l);
and may also be implemented via separate, but cooperating microprocessors: and may further be embedded in the central processor 56 itself. Tuner 258, the one associated with optimizing the recycle flow rate R~t), is shown as comprised pc/ ;' ~
-~1~5158 of a cross-correlator 260 to which is applied a pair of input signals on the lines 230A and 250A. The path 250A
provides the C(t) signal, while the path 230A provides a combined signal containing the control signal plus the random signal - from the output of summing junction 218.
The output from the cross-correlator is integrated in integrator 262, whose output is in turn applied to recycle curve element 264 which produces a ~ pc/;, ~
~1~5158 process reaction curve dir~cted to optimizin~ the recycle flow rate,parameter. A control parameter determinlng element 1 266 recoives the output from tho recycle curve element 264 and perlodicaliy produce~ updated ~alue~ for three key parameter~, wh~ch are appll~d via an lnterfacing element 268, ~or,us~ in the PID~controller 208. These three key para-, meters aro the overall controller gain Kc; the integral time TI; an,d th~ delay time TD. The intere~ted reader i9 referred l to a 1953 published article wherein the parameter-tuning 10', technique of the present invention i~ described. See, Cohen, G.H. and Coon, G.A., ~Theor~tical Conside,rations of ' Retarded Control,~ Tran~. ASME, Vol. 74, 1953, pp. 827.
jj The technique has hecome very well known in the control '~ syatem arts and iB referred to herelnafter as the "Cohen-Coon lS methodn.
1~ In order to improve the'f~elity of ~he bioavailabilit~
, prediat~on when tho improved contro~ ~ystem 200 1~ operating in the open loop mode, lt is u~eful to ~irst detormihe the I proper ¢ontrol parameter ~Xc, TI and ~D) settings for aach 20 I of the PID controller~ 202, ~04, 206 and 208. This i~ best dono by r~n~om in,put'modellng during an experimental run p~rformed w1th ~ raference drug form in the elo8ea loop mode ! f operation. The recycle rate control channel i8 illustrat~ve '~ of the mothnd usod. Super~mposed on each channel control signal i8 ~ pseudo-random binary signal with an amplitude " at lea~t on- standard deviation great-r than tho noise level , I .
, -28-, . .
.
- ¦ ~ the channel and a bandwidth corr~sponding to ten time~
¦ the bandwidth of the channel. Active electronic filteriny of the output ~rom the pRss generator 220 by a low pass fllter ¦ 220A may or m~y no~ b~ necessary due the pos~ibility of auto-S j fllter$ng by the mechanlcal damping ch~racterist~cs of theactlvators u~ed. Th~ ~lgnal output of the spectrophotometer 240 will contaln the result~ of th~ control signal plu8 the i PRBS. ~his ~lgnal C~t) 1~ cros~-correlated with the combined input signal tcontrol slgnal plu8 PRBS) over a period of ~illustratively) five time con~tants to yield ~ weighting ' function at the output of the cross-correlator 260. A process reaction CUrVQ i~ produced by integratlng the we~ghtlng function in the integrator 262, and the control parameters ~ ~, TI an~ TD) ~re determined by the method of Cohen and lS Cool with~n tha element 266. Those new controller ~attings for e~ch proce~s variable are substituted into tho four PID
.. . .
, controllers an~ the proce~s 1~ r-peated for another ~ive time ~on~tant~. The procedure 18 p-rformed lndepend~ntly 1~ for Qa~h of the four PID controllor- and s~multaneou~ly for 20 il all fou~ controller~ durlng each exporimental run.
Thi- proces~ o~ controller tuning ~s initlally performo~ lnd~p~ndently (an~/or ln combinatlon with one or " more of tho other proce~s variable~) for each process varl-~ able to ok~aln initial e~timates of ~ontroller settings for each controller. In the course of an actual run, the controller -;
~ .
?l SiS8 !~ for e~ch o~ the four proces3 variables operate together in I parallel and are retuned periodically ~every S t~me constants, 1 illustrativoly) simultaneously to provide new updated values li for the controller settings. Controller ~etting values S j obtained by this adaptivQ control pro~edure during the course of an experimental run are re~ected, and previous values retalned, if the proposed value~ are outside of a range of values for each ~ett~ng which had been found to induc ~ ¦ lnstability ln the operation of onet or more of the actuators.
io ' Although the ~nvention has been described in terms of ~elected preferred em~todiments and improvements to these ', em~odiments, the lnvention should not be deemed l$mited ; -!i thereto, slnce oth~r embodiments and modifications will readlly o¢cur to one skilled in the art. It is therefore S to be undesstood that the appended olaims are intended to oover all ~uch modificaelons as fall within the true spirit ~nd ~¢ope of the lnvention.
,!
.Ij' .
! .
. I
.. .
, . . ~ .
.
~ -30-:;
,; ,
Claims (15)
1. Apparatus for the automated dissolution testing of products whose solubility and dissolution rate properties affect product performance comprising:
a) a dissolution cell having an input line, an output line, a filtering chamber therebetween, and controllable agitation means responsive to an agitation control signal, positioned within said chamber;
b) at least two reservoirs connected to said input line through a like number of media supply lines for providing dissolution media from said reservoirs to said filtering chamber c) first flow control means responsive to a first control signal and positioned in said input line for controlling the flow of dissolution media through said filtering chamber, and second flow control means responsive to a second control signal and positioned in at least one of said media supply lines for controlling the flow of dissolution media to said input line;
d) recirculating means including third flow control means responsive to a third control signal and connected to said input and output lines for controlling recirculatory flow therebetween;
e) system output means connected to said output line and containing means for measuring output parameters therein including a concentration parameter of particular constituents within said flow, and for providing system output signals corresponding to said measured output parameters;
f) memory means for providing stored control signals to said apparatus; and g) electronic control means connected to said memory means and operative in a first mode responsive to said stored control signals and to said measured output parameter signals for automatically determining a set of internal control parameters used to determine said first, second, and third control signals so as to minimize the differences between said stored and measured signals, and operative in an open loop mode responsive to said determined internal control parameters for automatically producing said first, second, and third control signals in the absence of said stored signals.
a) a dissolution cell having an input line, an output line, a filtering chamber therebetween, and controllable agitation means responsive to an agitation control signal, positioned within said chamber;
b) at least two reservoirs connected to said input line through a like number of media supply lines for providing dissolution media from said reservoirs to said filtering chamber c) first flow control means responsive to a first control signal and positioned in said input line for controlling the flow of dissolution media through said filtering chamber, and second flow control means responsive to a second control signal and positioned in at least one of said media supply lines for controlling the flow of dissolution media to said input line;
d) recirculating means including third flow control means responsive to a third control signal and connected to said input and output lines for controlling recirculatory flow therebetween;
e) system output means connected to said output line and containing means for measuring output parameters therein including a concentration parameter of particular constituents within said flow, and for providing system output signals corresponding to said measured output parameters;
f) memory means for providing stored control signals to said apparatus; and g) electronic control means connected to said memory means and operative in a first mode responsive to said stored control signals and to said measured output parameter signals for automatically determining a set of internal control parameters used to determine said first, second, and third control signals so as to minimize the differences between said stored and measured signals, and operative in an open loop mode responsive to said determined internal control parameters for automatically producing said first, second, and third control signals in the absence of said stored signals.
2. The apparatus of claim 1 wherein said dissolution cell further comprises:
a) agitation means disposed within said filtering chamber and responsive to an agitation signal for controlling the agitation rate within said chamber; and b) wherein said electronic control means further comprises means for producing said agitation signal responsive to said internal control parameters, whereby said first, second and third control signals in combination with said agitation signal serve to minimize the differences between said stored and measured signals.
a) agitation means disposed within said filtering chamber and responsive to an agitation signal for controlling the agitation rate within said chamber; and b) wherein said electronic control means further comprises means for producing said agitation signal responsive to said internal control parameters, whereby said first, second and third control signals in combination with said agitation signal serve to minimize the differences between said stored and measured signals.
3. The apparatus of claim 2 wherein said measured output parameters include a flow parameter corresponding to the flow through said system output means.
4. The apparatus of claim 2 wherein said electronic control means includes means for determining a flow parameter corresponding to the flow through said system output means by determining the difference between said first and third control signals.
5. The apparatus of claim 2 wherein said at least two reservoirs provide dissolution media of substantially different characteristics and said electronic control means includes means for controlling the ratio of said different dissolution media by adjusting the ratio of said first and second control signals.
6. The apparatus of claim 5 wherein said at least two reservoirs contain acidic and alkaline media respectively, and said ratio of said different dissolution media constitutes a pH control.
7. The apparatus of claim 6 wherein said product is a pharmaceutical drug dosage form and said stored control signals comprise in vivo dissolution profiles.
8. A method of automatic dissolution testing of products whose solubility and dissolution rate properties affect product performance having unknown dissolution profiles by comparison against known product forms having known dissolution profiles, comprising:
a) flowing a dissolution medium of controllable content at a controlled flow rate through a dissolution cell containing said known product form;
b) providing a recirculation path around said dissolution cell to provide controllable sink conditions therethrough:
c) measuring the concentration profile of said known product form at the output of said dissolution cell and providing output signals corresponding thereto;
d) providing stored signals corresponding to the concentration profile of said known product form;
e) producing a first family of control signals such that the differences between said stored signal and said output signals are minimized;
f) producing a second family of control signals for controlling said medium controlled content, said controlled flow rate, and said controlled sink conditions from said first family of control signals; and g) flowing a dissolution medium of said controlled content at said controlled flow rate through a dissolution cell containing said unknown product form whereby measured concentration profiles of said unknown product form are derived as being predictive of subsequent usage dissolution profiles.
a) flowing a dissolution medium of controllable content at a controlled flow rate through a dissolution cell containing said known product form;
b) providing a recirculation path around said dissolution cell to provide controllable sink conditions therethrough:
c) measuring the concentration profile of said known product form at the output of said dissolution cell and providing output signals corresponding thereto;
d) providing stored signals corresponding to the concentration profile of said known product form;
e) producing a first family of control signals such that the differences between said stored signal and said output signals are minimized;
f) producing a second family of control signals for controlling said medium controlled content, said controlled flow rate, and said controlled sink conditions from said first family of control signals; and g) flowing a dissolution medium of said controlled content at said controlled flow rate through a dissolution cell containing said unknown product form whereby measured concentration profiles of said unknown product form are derived as being predictive of subsequent usage dissolution profiles.
9. The method of claim 8 further compriaing the step of providing agitation at a controllable rate within said dissolution cell and wherein said second family of control signals includes signals for controlling said controllable agitation rate.
10. The method of claim 9 wherein said step of flowing a dissolution medium of controllable content comprises con-trolling the pH of said controlled medium.
11. The method of claim 9 wherein said stored signals comprise a time series representative of said concentration profile over a predetermined period of time.
12. The method of claim 9 wherein said first family of control signals comprises a set of overall control para-meters for controlling said dissolution testing process and said second family of control signals comprise a set of process variables for controlling said controlled content, flow rate, sink conditions and agitation rate.
13. The method of claim 9 wherein said product is a pharmaceutical drug dosage form and said stored signals comprise in vivo dissolution profiles and said predictive subsequent usage dissolution profiles comprise predictive in vivo dissolution profiles.
14. A method of simultaneously predicting in vivo dissolution profiles of a plurality of pharmaceutical drug dosage form by comparison with a reference drug dosage form having a known in vivo dissolution profile, comprising:
a) passing a reference dissolution medium formulated as a mixture of at least two constituents under the control of a mixture control signal through a reference dissolution cell containing said reference drug, at a flow rate controlled by a flow control signal and at an agitation rate within said dissolution cell controlled by an agitation control signal;
b) flowing a plurality of dissolution media, each of which is formulated as a mixture of at least two constituents under the control of said mixture control signal, through a corresponding plurality of dissolution cells, each of which contains a different one of said plurality of drug dosage forms;
c) providing a separate recirculation path around each of said dissolution cells, each path having a recirculation rate controlled responsive to a sink control signal;
d) measuring the concentration profile of said re-ference drug at the output of said reference dissolution cell and providing a time series of output signals corresponding thereto;
e) providing a time series of stored signals corres-ponding to the concentration profile of said reference drug:
f) producing a set of optimized process variable control signals optimized by deriving a difference signal between said time series of output signals and said time series of stored signals; and g) producing said mixture control signal, said flow control signal, said sink control signal and said agitation control signal such that said derived difference signal is continually minimized.
a) passing a reference dissolution medium formulated as a mixture of at least two constituents under the control of a mixture control signal through a reference dissolution cell containing said reference drug, at a flow rate controlled by a flow control signal and at an agitation rate within said dissolution cell controlled by an agitation control signal;
b) flowing a plurality of dissolution media, each of which is formulated as a mixture of at least two constituents under the control of said mixture control signal, through a corresponding plurality of dissolution cells, each of which contains a different one of said plurality of drug dosage forms;
c) providing a separate recirculation path around each of said dissolution cells, each path having a recirculation rate controlled responsive to a sink control signal;
d) measuring the concentration profile of said re-ference drug at the output of said reference dissolution cell and providing a time series of output signals corresponding thereto;
e) providing a time series of stored signals corres-ponding to the concentration profile of said reference drug:
f) producing a set of optimized process variable control signals optimized by deriving a difference signal between said time series of output signals and said time series of stored signals; and g) producing said mixture control signal, said flow control signal, said sink control signal and said agitation control signal such that said derived difference signal is continually minimized.
15. The method of claim 14 further comprising the step of introducing a pseudo random binary signal into a processed version of said derived difference signal to further optimize in part said optimized process variable control signals.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US141,093 | 1980-04-17 | ||
| US06/141,093 US4335438A (en) | 1980-04-17 | 1980-04-17 | Method and apparatus for automatic dissolution testing of products |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1145158A true CA1145158A (en) | 1983-04-26 |
Family
ID=22494127
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000375654A Expired CA1145158A (en) | 1980-04-17 | 1981-04-16 | Method and apparatus for automatic dissolution testing of products |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US4335438A (en) |
| EP (1) | EP0049293A4 (en) |
| JP (1) | JPS57500484A (en) |
| CA (1) | CA1145158A (en) |
| DK (1) | DK539481A (en) |
| WO (1) | WO1981003072A1 (en) |
Families Citing this family (57)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5733353A (en) * | 1980-08-06 | 1982-02-23 | Furonto Sangyo Kk | Tester for physical properties of solid agent |
| FI833181A0 (en) * | 1983-09-06 | 1983-09-06 | Labsystems Oy | FOERFARANDE FOER BESTAEMNING AV LOESNINGSHASTIGHET |
| US4678639A (en) * | 1984-03-02 | 1987-07-07 | The Perkin-Elmer Corporation | Apparatus for periodically monitoring the composition of a plurality of samples |
| DE3520034C1 (en) * | 1985-06-04 | 1986-05-07 | Erweka Apparatebau GmbH, 6056 Heusenstamm | Disintegration test device |
| EP0255026A3 (en) * | 1986-07-23 | 1988-10-19 | Takeda Chemical Industries, Ltd. | Automatic analysis method and apparatus for enzyme reaction |
| US4987083A (en) * | 1986-12-29 | 1991-01-22 | Apple Glenn D | Photometric analyzer and controller for beverages |
| EP0278374A3 (en) * | 1987-02-06 | 1989-03-01 | Pharmatest Apparatebau Gmbh | Device for testing the release of active components by pharmaceutical products |
| IT1216121B (en) * | 1988-03-17 | 1990-02-22 | Ubaldo Conte | EQUIPMENT FOR THE CONTEMPORARY DETERMINATION OF THE QUANTITY OF WATER ABSORBED BY A COMPACT SUBSTANCE AND OF THE DEVELOPED DISGREGATION FORCE. |
| US4947339A (en) * | 1988-12-01 | 1990-08-07 | Jan Czekajewski | Method and apparatus for measuring respiration, oxidation and similar interacting between a sample and a selected component of a fluid medium |
| US5076107A (en) * | 1988-12-27 | 1991-12-31 | G. D. Searle & Co. | Apparatus and method for resultant-weight measuring system |
| US5394322A (en) * | 1990-07-16 | 1995-02-28 | The Foxboro Company | Self-tuning controller that extracts process model characteristics |
| US6060024A (en) * | 1993-07-14 | 2000-05-09 | Zymark Corporation | Automatic dissolution testing system |
| JP2955365B2 (en) * | 1994-02-25 | 1999-10-04 | デイステク インコーポレイテツド | Dissolution test equipment |
| US5612187A (en) * | 1994-03-22 | 1997-03-18 | Espress Tech, Inc. | Clot lysis time determining device and method for determining the time necessary for fluid to lyse a clot, and clot supporter |
| US5807115A (en) * | 1996-01-31 | 1998-09-15 | Hu; Oliver Yoa-Pu | Dissolution apparatus simulating physiological gastrointestinal conditions |
| US6174497B1 (en) | 1997-06-04 | 2001-01-16 | Euro-Celtique, S.A. | Detection systems and methods for predicting the dissolution curve of a drug from a pharmaceutical dosage form |
| CA2257106C (en) * | 1996-06-04 | 2002-04-23 | Euro-Celtique, S.A. | Improvements in detection systems and methods for predicting the dissolution curve of a drug from a pharmaceutical dosage form |
| WO1998005957A1 (en) * | 1996-08-01 | 1998-02-12 | Distek Inc. | Apparatus for measuring heights in dissolution testing |
| US5849597A (en) * | 1996-12-18 | 1998-12-15 | Toshiba Ceramics Co., Ltd. | Treatment apparatus for high-precision analysis of impurities in silicic material |
| US6004822A (en) * | 1997-04-04 | 1999-12-21 | Alfred LaGreca | Device and method for measuring solubility and for performing titration studies of submilliliter quantities |
| US20020010550A1 (en) * | 1998-09-14 | 2002-01-24 | George M. Grass | Pharmacokinetic-based drug design tool and method |
| JP2003508748A (en) * | 1999-08-30 | 2003-03-04 | ユーロ−セルティーク，エス．エイ． | In situ method for measuring release of a substance from a dosage form |
| US7167615B1 (en) | 1999-11-05 | 2007-01-23 | Board Of Regents, The University Of Texas System | Resonant waveguide-grating filters and sensors and methods for making and using same |
| DK1330368T3 (en) * | 2000-11-04 | 2004-09-27 | Kurz Leonhard Fa | Plastic body formed as a foil, e.g. |
| US6336739B1 (en) * | 2001-03-02 | 2002-01-08 | Luke Lee | Air bath dissolution tester |
| WO2003034060A1 (en) * | 2001-10-11 | 2003-04-24 | Elan Pharma International Limited | Apparatus and method for concurrently monitoring active release and physical appearance of solid dosage form pharmaceuticals |
| US6799123B2 (en) * | 2001-11-05 | 2004-09-28 | Lyn Hughes | Dissolution test equipment and method |
| US6948389B2 (en) * | 2002-03-18 | 2005-09-27 | Distek, Inc. | Dissolution test sampling |
| AU2003243246A1 (en) * | 2002-05-17 | 2003-12-02 | Delphian Technology, Inc. | In situ methods for measuring the release of a substance from a dosage form |
| US20040115822A1 (en) * | 2002-09-12 | 2004-06-17 | Schapaugh Randal Lee | Methods of measuring the dissolution rate of an analyte in a non-aqueous liquid composition |
| US20040115837A1 (en) * | 2002-11-27 | 2004-06-17 | Schapaugh Randal Lee | Methods of measuring the dissolution rate of an analyte in a non-aqueous liquid composition |
| US7071001B2 (en) | 2003-01-10 | 2006-07-04 | Dnk Associates, Inc. | System and method for in vitro bleeding time testing |
| US7024955B2 (en) * | 2003-03-01 | 2006-04-11 | Symyx Technologies, Inc. | Methods and systems for dissolution testing |
| EP1777510A1 (en) * | 2005-10-21 | 2007-04-25 | Rohm and Haas Company | Dissolution test equipment and methods for testing |
| TW200732644A (en) | 2005-10-25 | 2007-09-01 | Rohm & Haas | Dissolution test equipment and methods for testing having improved filtration system |
| US7914741B2 (en) * | 2007-01-10 | 2011-03-29 | Agilent Technologies, Inc. | Automated microsampling dissolution test system |
| US20090064768A1 (en) * | 2007-03-13 | 2009-03-12 | Emad Alkhawam | Devices for Determining the Release Profile of Macromolecules and Methods of Using the Same |
| US20080223116A1 (en) * | 2007-03-13 | 2008-09-18 | Emad Alkhawam | Drug Release Cell and a Method for Testing the Drug Release of a Suspension in a Liquid Utilizing the Same |
| JP5612474B2 (en) * | 2007-10-17 | 2014-10-22 | トランスファーマ メディカル リミテッド | Verification of dissolution rate |
| US20090155920A1 (en) * | 2007-11-12 | 2009-06-18 | Symyx Technologies, Inc. | High throughput dissolution and precipitation apparatus and method |
| US20090165578A1 (en) * | 2007-12-06 | 2009-07-02 | Pain Therapeutics, Inc. | Dissolution test equipment and methods for testing |
| WO2009079689A1 (en) * | 2007-12-20 | 2009-07-02 | Cyantific Instruments Pty Ltd | Analytical method and apparatus |
| DE102008024840B3 (en) * | 2008-05-23 | 2009-11-12 | Merck Patent Gmbh | Device for receiving a solid in a measuring cell |
| JP5576337B2 (en) * | 2010-06-30 | 2014-08-20 | ローム アンド ハース カンパニー | Delivery device for dispensing a pharmaceutical dosage form to a dissolution test device |
| EP2529770A1 (en) | 2011-05-31 | 2012-12-05 | Nederlandse Organisatie voor toegepast -natuurwetenschappelijk onderzoek TNO | A digestion system |
| US20130074587A1 (en) * | 2011-09-28 | 2013-03-28 | Mel Merciadez | System and method for dissolution analysis of drug-eluting medical devices |
| GB201207677D0 (en) | 2012-05-02 | 2012-06-13 | Ucl Business Plc | Apparatus and method for testing medicaments |
| ITRM20130163A1 (en) * | 2013-03-18 | 2014-09-19 | Univ Roma | DEVICES AND METHOD FOR TESTING THIN FILMS AND STRIPS |
| EA035974B1 (en) * | 2014-08-06 | 2020-09-08 | Хексаль Аг | Method for determining the disintegration time of a film-shaped pharmaceutical dosage form, disintegration testing device and use of said device |
| CA2968548C (en) | 2014-11-21 | 2022-12-13 | Absorption Systems Group, Llc | System for the concomitant assessment of drug dissolution, absorption and permeation and methods of using the same |
| WO2016168445A1 (en) | 2015-04-17 | 2016-10-20 | Pion Inc. | Apparatus and method for the assessment of concentration profiling and permeability rates |
| CN105784953A (en) * | 2016-05-25 | 2016-07-20 | 延边大学 | Full-automatic dissolvability testing device of flowing cell |
| IT201600077103A1 (en) * | 2016-07-22 | 2018-01-22 | Tifqlab S R L | TREATMENT SYSTEM FOR EVALUATION OF INTERACTIONS WITH A CONTACT SURFACE |
| US11668723B2 (en) | 2019-07-09 | 2023-06-06 | Logan Instruments Corporation | Automated dissolution/permeation testing system |
| US10732080B1 (en) * | 2019-10-14 | 2020-08-04 | Sotax Corporation | Integrated dissolution processing and sample transfer system |
| CN112284822A (en) * | 2020-09-30 | 2021-01-29 | 深圳市锐拓仪器设备有限公司 | Automatic sampling device of dissolution instrument by flow cell method |
| CN114112798B (en) * | 2021-11-25 | 2024-03-12 | 上海交通大学 | Trisodium phosphate dissolution rate experimental system |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3223486A (en) * | 1962-09-12 | 1965-12-14 | Technicon Instr | Apparatus for treatment of solids for analysis |
| US3467500A (en) * | 1966-08-18 | 1969-09-16 | American Home Prod | Automatic sample digesting device |
| US3545864A (en) * | 1967-07-19 | 1970-12-08 | Desaga Gmbh C | Apparatus and method for determining gastro-enteral absorption quota |
| US3684448A (en) * | 1969-06-30 | 1972-08-15 | Boehringer Sohn Ingelheim | Apparatus and process of in vitro determination of the dissolution behavior of drugs or pharmaceutical compositions containing the same in the gastrointestinal tract |
| US3801280A (en) * | 1971-11-11 | 1974-04-02 | Upjohn Co | Solubility-dissolution test apparatus and method |
| US3791222A (en) * | 1972-04-07 | 1974-02-12 | Warner Lambert Co | Dissolution testing device |
| US3791221A (en) * | 1972-04-07 | 1974-02-12 | Warner Lambert Co | Dissolution testing device |
| US4158694A (en) * | 1978-03-30 | 1979-06-19 | The Dow Chemical Company | Automated apparatus for testing sustained release drugs |
-
1980
- 1980-04-17 US US06/141,093 patent/US4335438A/en not_active Expired - Lifetime
-
1981
- 1981-03-31 EP EP19810901227 patent/EP0049293A4/en not_active Withdrawn
- 1981-03-31 WO PCT/US1981/000423 patent/WO1981003072A1/en not_active Application Discontinuation
- 1981-03-31 JP JP56501592A patent/JPS57500484A/ja active Pending
- 1981-04-16 CA CA000375654A patent/CA1145158A/en not_active Expired
- 1981-12-07 DK DK539481A patent/DK539481A/en not_active Application Discontinuation
Also Published As
| Publication number | Publication date |
|---|---|
| EP0049293A4 (en) | 1982-09-03 |
| WO1981003072A1 (en) | 1981-10-29 |
| US4335438A (en) | 1982-06-15 |
| DK539481A (en) | 1981-12-07 |
| EP0049293A1 (en) | 1982-04-14 |
| JPS57500484A (en) | 1982-03-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1145158A (en) | Method and apparatus for automatic dissolution testing of products | |
| CA1128335A (en) | Multiple injector flow through dissolution cell, for dissolution testing apparatus | |
| US5441700A (en) | Headspace autosampler apparatus | |
| US7234365B2 (en) | Methods and systems for dissolution testing | |
| US6667009B1 (en) | Apparatus for sampling and analysis of reactions by trapping reaction components on a sorbent | |
| AU4234396A (en) | System for sensing droplet formation time delay in a flow cytometer | |
| EP3326718B1 (en) | Microplate for determining products of agglutination reactions | |
| DE102011011186A1 (en) | Testing the resolution by means of infrared temperature measurement | |
| CA2462737A1 (en) | Apparatus and method for concurrently monitoring active release and physical appearance of solid dosage form pharmaceuticals | |
| JP2003503714A (en) | Multiple test assemblies for evaluating, detecting and monitoring high pressure processes | |
| US4890484A (en) | Automated method and apparatus for determining total suspended solids in liquids | |
| EP0108404B1 (en) | Process and equipment for determining or influencing the flow properties of solid granular materials | |
| US4593563A (en) | Procedure for determining velocity of dissolution | |
| Palberg et al. | Shear modulus titration in crystalline colloidal suspensions | |
| Klier et al. | Solute and penetrant diffusion in swellable polymers: VIII. Influence of the swelling interface number on solute concentration profiles and release | |
| US5122745A (en) | Method and apparatus for determining molecular dynamics of materials | |
| Hurst | Automation in the Laboratory | |
| US20020147563A1 (en) | Method and device for accelerated stability analysis | |
| JPH05332966A (en) | Method and apparatus for inspecting exudation stability of substance | |
| CA2132965A1 (en) | Rapid hypersensitive flowthrough immunodetection system | |
| US3152251A (en) | Apparatus for automatically determining ultraviolet absorbance in a system | |
| US4820425A (en) | Liquid sensor for robotic filtration station | |
| Curtain | An automatic multiple column apparatus for analytical zone electrophoresis | |
| Matsushita et al. | A fully automated HPLC system with the function of self-diagnosis. | |
| JP4621902B2 (en) | Reaction analyzer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |