functionality 5

International Journal of Pharmaceutics 153 (1997) 191198

Functionality testing of gellan gum, a polymeric excipientmaterial for ophthalmic dosage forms

Annouk Rozier *, Claude Mazuel, Jeffrey Grove, Bernard Plazonnet

Laboratoires Merck Sharp & Dohme-Chibret, Centre de Recherche, Route de Marsat, Riom, France

Received 27 November 1996; accepted 10 April 1997

Abstract

Low acetyl gellan gum, originally a food ingredient, has been used to devise novel ophthalmic formulations,significantly improving drug ocular bioavailability. This increase in drug bioavailability results from the unique gellingproperty of gellan gum in the presence of tear fluid cations. The aim of this study was to develop a functionality testensuring the consistency of the dosage forms gelling property and a reproducible pharmacological effect. The rupturestrength of the gel was shown to be a reliable indicator of the ocular drug bioavailability in the albino rabbit. Thetest parameters susceptible to influence the test results were identified, evaluated and optimized. The influence of theraw material characteristics and of the processing parameters on the final product gel strength were determined andoptimized, and finished product specifications established. 1997 Elsevier Science B.V.

Keywords: Functionality test; Gel strength; Ophthalmics; Gellan gum; Gelrite; Ocular bioavailability

1. Introduction

Excipients have long been considered as inertcomponents of a drug formulation. This can nolonger be considered true in view of the develop-ment of new dosage forms with excipients addedto the formulation that drastically modify thedrug release. The pharmaceutical community isbecoming acutely aware of the necessity to char-acterize and control these excipient materials: a

key issue of the Second Joint PharmacopoeiaOpen Conference on Harmonization of ExcipientsStandards held in Florida in 1994, was the func-tionality testing of such excipients.

Timoptic XE1 (Timoptol LP or TimoptolLA in Europe) is a novel ophthalmic dosageform which enhances the ocular bioavailability ofthe antiglaucoma drug timolol in the albino rabbit

1 Timoptic XE, Timoptol LP, Timoptol LA, Timoptic

and Timoptol are trademarks of Merck and Co. Inc., (White-house Station, N.J., USA).* Corresponding author.

0378-5173:97:$17.00 1997 Elsevier Science B.V. All rights reserved.

PII S 0378 -5173 (97 )00109 -9

A. Rozier et al. : International Journal of Pharmaceutics 153 (1997) 191198192

by three- to four-fold, when compared to a plaintimolol solution (Rozier et al., 1989a). In man,studies have shown that Timoptic XE given oncea day was equivalent to regular Timoptic (Timo-ptol in Europe) administered twice a day (Shedden,1994). This amelioration was obtained by theaddition to the formula of a polymer excipient,Gelrite2. Gelrite (low acetyl gellan gum), origi-nally developed as a food ingredient, has a uniqueproperty, forming viscous solutions in deionizedwater which gel in the presence of cations. Theophthalmic formulations containing gellan gum areinstilled as liquids and gel upon contact with the tearfluid cations (Mazuel and Friteyre, 1985). Oncegelled,theformulationsresistthenaturaleliminationprocess of drainage from the precorneal area; theirresidence time at the site of drug absorption isprolonged (Greaves et al., 1990), and subsequently,the amount of drug absorbed is increased. Theimportance of the gelation mechanism was demon-strated by the comparison of the intraocular druglevels obtained with Timoptic XE and a non-gellingpolymeric solution of comparable viscosity (Rozieret al., 1989b). The former are significantly higher,indicating that gelation is indeed the key parameter.It thus appeared early during the drug developmentprogram that a functionality test characterizing thisparameter was mandatory to obtain a dosage formwith a reproducible pharmacological effect.

The test had to be rapid, simple and ultimatelyusableasaroutine test,whilenotnecessitatingoverlysophisticated equipment. It had to allow the testingof both the raw material and the finished dosageform. In addition, as it was intended to be used asa product stability indicator, and because thecapacity of an ophthalmic vial is relatively small (2.5ml per vial in the case of Timoptic XE), the amountof product necessary to run the test had to be assmall as possible. Finally, the test had to be accurate,reproducible and validated.

Available methods to evaluate gel characteristicsinclude fundamental tests such as low shear oscil-latory rheology. While this kind of techniquewas used during product development, it requiressophisticated equipment that is not adapted to

routine analysis in a control laboratory. Empiricaltests include Texture Profile Analysis (TPA) and gelstrength measurement. TPA is widely used in thefood industry (Szczesniak et al., 1963) and consistsof the analysis of the force:deformation curvegenerated by compressing a gel sample using anInstron Universal Testing Machine. Gel strengthmeasurements can be obtained using a variety ofinstruments which do not necessarily measure thesame textural parameter. A gel strength test isdescribed in the USP (US Pharmacopeia XXIII,1995) for gelatin (commonly called the Bloom test),which measures gel firmness at small strain levels,but which does not measure rupture strength. TheUSP test requires too large a sample size (\100 g)for routine examination of ophthalmic samples. Italso measures the combined reaction forces of thegel itself and of the skin formed at the air-gelinterface. As such a skin would not form in theeye, the test was not thought to be representative.Another inconvenience is that the sample is notunmolded before testing and interferences from thecontainer walls can be expected. TPA provides acomprehensive understanding of gel texture, butrequires expensive equipment and is particularlyindicated when comparison of two different gellingagents is required (Sanderson et al., 1988). Since theultimate goal was to compare identical formulationsof the same gelling agent, attention was thereforedirected towards a simple gel strength test adaptedto small size samples and allowing the determinationof the gel rupture strength. The test sample prepa-ration was designed to ensure sample conservationand to avoid the interferences mentioned above. Thestrain was applied to the total sample surface areaand the force required to break the gel was deter-mined.

2. Materials and methods

2.1. Apparatus

A Stevens3 LFRA 1000 texture analyzer equip-ped with a 38 mm cylindrical plunger and a

2 Gelrite is a trademark of Kelco, unit of Monsanto Co.,St. Louis, Missouri, USA.

3 Stevens and Son Ltd., 28 Dolphin Yard, Holywell Hill,St. Albans, Hertfordshire AL1 1EX, England.

A. Rozier et al. : International Journal of Pharmaceutics 153 (1997) 191198 193

Table 1Gelling reagent formulation

Cation ConcentrationAdded as(mg:ml)

Sodium 24Sodium chloride0.225Calcium Calcium chloride,

dihydrate

spring steel strip of adjustable length. It is pro-vided with a certificate of calibration which liststhe reaction forces obtained in a standarddeflection test, for a series of different strip po-sitions. The same standard test conditions areused to measure the reaction forces with the tex-turometer to be calibrated. These should agreewithin 1% with the values of the certificate ofcalibration.

2.2. Test sample preparation

The test sample is molded from a casting so-lution prepared by boiling a known amount ofthe gellan formulation or raw material with thegelling reagent. The gelling reagent is a solutionin deionized water of the tear fluid gellingcations in physiological proportion (see Table1). After boiling, the casting solution weight isadjusted so as to obtain in the test sample afinal gellan concentration of 0.4% and 25% ofgelling reagent. The samples are cast in moldsconsisting of a Plexiglass ring sitting on a Plexi-glass plate. 2.1 ml of the casting solution isneeded for the 14.6 mm internal diameter mold.Another plate is positioned on the ring to avoidthe entrapment of air bubbles between the sam-ple and the plate. The gel is allowed to set with-out disturbance for a given time and unmoldedjust before measurement.

one-way recorder was chosen to perform themeasurements. The gel strength is evaluated bycompressing the sample up to rupture pointwith a mobile plunger moving downward at adefined rate (0.2 mm:s), for a preset distance(5.0 mm). A sensor is connected to the plungerand records the sample reaction forces to com-pression and in particular, the force necessary tobreak the gel or gel strength. Gel strength isreported in grams. It is proportional to the testsample surface area and for convenience can bereported as a mass:surface ratio in g:cm2.Thesensor is calibrated using standard masses sus-pended from the plunger. The weights of twomasses bracketing the working range (100 and1000 g) are checked against specifications (91%).

The gel strength measurement calibration iscarried out using a dummy bloom strip avail-able from the manufacturer. It consists of a

Table 2Influence of boiling time

Gel strength (g:cm29R.S.D.; number of single determinations)

14.6 mm molds 19.3 mm molds

1 min 3 min 5 min 1 min 3 min 5 min

6794 (8) 6893 (8)ND 6292 (7)6893 (6) NDND NDND 6693 (7)6793 (7) ND

ND6592 (7)6593 (5)6594 (8) ND6594 (7)6493 (8) 6492 (8)6394 (8)ND 6492 (7) ND

Avg.64Avg.65 Avg.66 Avg.66Avg.65Avg.65

R.S.D., relative standard deviation.ND, not determined.


Table 3Influence of setting time

Sample Gel strength (g:cm29R.S.D.; number of single determinations)

14.6 mm molds 19.3 mm molds

6 h2 h4 h 6 h2 h

6691 (6) 6892 (6)Gellan solution (autoclaved) 6591 (6) 6593 (8)ND160 (2) ND15992 (3)Gellan solution (non-autoclaved) ND16995 (3)

Raw material184 (2) NDLot 1 182915 (3) 179 (2) ND

NDNDLot 2 9598 (3) 96 (2) 102 (2)186923 (3) 174917 (3)Lot 3 ND183914 (3) ND

R.S.D., relative standard deviation.ND, not determined.

2.3. Method 6alidation

The influence of boiling time (1 to 5 min),setting time (2 to 6 h), casting temperature range(8488 or 6269C) and weighing variations wereevaluated and the operator-to-operator reproduci-bility checked.

2.4. Ocular bioa6ailability e6aluation

Ocular bioavailability was evaluated in albinorabbits. Bilateral instillations, 50 ml of 0.25% Tim-optic XE solutions with different gel strengths,were made into the conjunctival sac of the albinorabbits. Cornea, aqueous humor and irisciliarybody were subsequently sampled at 10 minutes,0.5, 1, 2 and 4 h post-treatment for quantitationof their timolol content by HPLC and UV detec-tion at 294 nm.

2.5. Gellan gum molecular weight e6aluation

The differences in the raw material molecularweights were evaluated using intrinsic viscositydata extrapolated from the viscosities of diluteaqueous gellan gum solutions at various concen-trations and measured using a Contraves LowShear 30 viscometer. To ensure complete polymerdissolution and to perform the measurementsabove the conformational transition in the ran-dom coil conformation (Robinson et al., 1991),measurements were carried out on 45C solutionscontaining 25 mM Na.

3. Results and discussion

Each of the test parameters susceptible to influ-ence the final sample gel strength was evaluatedand optimized, when deemed necessary, to insuremethod reproducibility. The correlation betweengel strength results and ocular bioavailability wasestablished and limits set. Finally, the influence ofvarious factors such as raw material batch andprocessing parameters on the final product gelstrength, was evaluated.

3.1. Method 6alidation

3.1.1. Influence of boiling timeBecause of potential polymer degradation dur-

ing boiling of the casting solution, the influence of

Table 4Influence of casting temperature (determination with 14.6 mmmolds)

Gel strength (g:cm29R.S.D.; number of singledeterminations)

Casting temperature:Casting temperature:8488C 6269C

6691 (5) 6392 (7)6091 (8)6894 (6)


Table 5Influence of gellan concentration, reagent concentration and precision of final q.s. on gel strength

Gel strength (g:cm29R.S.D.; number of single determinations)

2%1% 5%Gelling reagent added: Nominal5% 2% 1%6493 (13) 6592 (20)5794 (12) 6193 (21) 6593 (15) 6593 (9)6293 (32)

Nominal 1.25% 2.5% 5%Water added for q.s.: 5% 2.5% 1.25%5492 (12)6293 (12)6192 (13)6293 (27)7192 (11) 7193 (11) 6392 (15)

10%5%Gellan formulation added: Nominal10% 5%7295 (9) 7394 (3) 8695 (6)6294 (6) 6796 (6)

R.S.D., relative standard deviation.

boiling time on gel strength was studied. Gelsamples prepared with 14.6 and 19.3 mm diametermolds were allowed to stand for 4 h before gelstrength determination. Results shown in Table 2indicate no difference between 1 and 5 min ofboiling time. For practical reasons, a time windowof 3 to 5 min was chosen. The data show nodifference in gel strength between 14.6 and 19.3mm molds.

3.1.2. Influence of setting timeAfter the samples have been molded, some time

is required for the gel to set. In order to determinethe influence of setting time on gel strength re-sults, various gellan samples (raw materials, non-autoclaved and autoclaved solutions) were tested.Results given in Table 3 indicate that a settingtime between 2 and 6 h does not affect gelstrength results.

3.1.3. Influence of casting temperatureThe gel strength of the sample will be influ-

enced if the casting temperature approaches thegel setting temperature (around 40C). The cast-ing temperature may therefore be critical. It is afunction of the time that elapses between the endof boiling and the actual casting of the solution,which includes the final q.s. adjustment by weight.Its influence was evaluated by allowing the castingsolution to cool down to different temperaturesbefore molding the test sample. Results shown inTable 4 reveal a small effect on gel strengthvalues. However, instead of setting molding tem-perature limits which proved to be impractical,

the time elapsed between the end of boiling andthe casting of the solution was set to 5 min, whichresulted in a solution temperature within therange 8488C.

3.1.4. Influence of weighing 6ariationsSample preparation involves several weighing

operations: test formulation, reagent and waterfor q.s. adjustment, which are not all done underanalytical conditions (the q.s., for example, needsto be done with the hot solution). Therefore, theeffect of variation in the amounts of gellingreagent, gellan formulation and water was evalu-ated in several exploratory experiments. Results in

Table 6Reproducibility study

Sample 95% Confi-Average gel strength (g:cm29R.S.D., N12) dence interval

(g:cm2)

10393.2Raw material 101105lot 1

183195Raw material 18995.0lot 2

0.5% Timoptic 15216215795.1XE lot 1

5194.6 50520.5% TimopticXE lot 2

0.25% Timoptic 88979297.7XE lot 1

6397.3 60660.25% TimopticXE lot 2

Averages are obtained from the results obtained by threeanalysts on two different days, each day in duplicate.R.S.D., relative standard deviation.


Fig. 1. Corneal concentration:time profile for different gelstrengths.

borderline statistical differences among treatments(P0.05) (see Table 7).

Gel samples are obtained under conditions(heating the ophthalmic solution with the gellingcations to boiling point, then cooling the mixtureto room temperature) that are different from thephysiological ones. Gelation occurs in the eye (at33C; Adlers Physiology of the Eye, 1981) bydiffusion of the gelling cations, mostly sodiumand calcium, into the instilled formulation. Never-theless, these data confirm that ocularbioavailability is influenced by gel strength asmeasured by the method described above. Whengel strength is within set limits, a consistent phar-macological effect can be expected. This has beenconfirmed in man by clinical observations madewith Timoptic XE using gel strengths within thisrange.

3.3. Raw material characterization

Gel strength results obtained from different gel-lan raw material batches show a significant batch-to-batch variation. Since gellan gum is producedby a fermentation process, it contains residualcations in variable amounts. Its intrinsic viscosity,which is a function of the polymer molecularweight was also found to vary. The amounts ofmonovalent and divalent cations in the test gelsample were a hundred-fold and a ten-fold, re-spectively, that of the raw material content, there-fore no correlation was found between gelstrength and raw material residual cations (datanot reported). On the other hand, gel strength wasfound to be proportional to gellan gum intrinsicviscosity (Fig. 3). Although intrinsic viscosity iscommonly used to assess relative polymer molecu-lar weight, the gel strength method was foundmore practical for routine testing. Specificationswere set for the raw material gel strength, as thiswas found to influence the final product gelstrength.

3.4. Influence of the manufacturing process onfinished product gel strength

Since the route of administration for TimopticXE is ocular, a major concern was to assure

Table 5 show that all these parameters impact onthe final gel strength and especially those influenc-ing the gum concentration (formulation weightand final q.s.). For this reason, it was decided toset a limit of 90.05 g for all these variables,which corresponds to 91.7% for the gellingreagent, 90.4% for the water added for q.s. and90.6% for the gellan formulation.

3.1.5. ReproducibilityIn order to evaluate the reproducibility of the

gel strength test, measurements were made on rawmaterials and formulations with gel strength val-ues bracketing the range normally encountered.Samples were assayed by three analysts on twodifferent days and each day in duplicate (i.e. 12measurements per sample) (see Table 6).

3.2. In 6i6o performances:gel strength correlation

In the albino rabbit, the pharmacokineticparameters Cmax, Tmax and T0.5 of timolol in theeye were similar for the three different gel strengthformulations (See, for example, the concentration:time profiles for cornea in Fig. 1.) In addition,increasing gel strength produced a gradual con-comitant increase in ocular bioavailability, as as-sessed by the measurement of the area under theconcentration versus time curve, AUC(04 h), atthe three ocular sites (Fig. 2). An analysis ofvariance (StudentNewmanKeuls) showed only


Fig. 2. AUC(0-4h) by tissue and gel strength.

Table 7Influence of gel strength on ocular bioavailability as assessed by AUC(04 h) measurements at three ocular sites

Cornea (mg h:g) Aqueous humor (mg h:ml) Irisciliary body (mg h:g)Gel strength (g:cm2)

5.39 8.52b48 87.9103.7 6.8677 11.30

11.907.4598 113.5a

0.62 0.91Standard error 7.1

a Statistically (PB0.05) greater than 48 g:cm2 gel strength.b Statistically (PB0.05) lower than 77 or 98 g:cm2 gel strength.

sterility of the finished product. Steam sterilizationwas proven the only viable option. Autoclavinggellan solutions leads to a significant reduction inthe finished product gel strength (Fig. 3). Thisreduction was found to be proportional to theautoclaving time. During manufacturing processdevelopment and scale-up, the gel loss duringsterilization was used as an indicator to determineoptimal sterilization conditions.

4. Conclusion

The development of a functionality test, based ongelation properties, has been described above for

gellan gum (Gelrite). This test satisfies the require-ments of simplicity and rapidity for small samplevolumes encountered in ophthalmics. The influenceof experimental parameters has been investigatedand conditions defined to obtain a reproducible androbust test for both the raw material and thefinished product. The functionality test has provento be an invaluable tool at each stage of thedevelopment of the dosage form and has enabledthe establishment of finished product specificationsensuring a consistent pharmacological effect. Inaddition, the control of the raw material based ongel strength measurements was shown to be possi-ble. Finally, the test allowed the establishment ofa manufacturing process that retained the funda-


Fig. 3. Intrinsic viscosity vs. gel strength.

mental gelling properties of the excipient for ex-ploitation in ophthalmic drug delivery.

The present test for a polymeric excipient isperhaps too specific to find a place in Pharmaco-poeias. Nevertheless, this work highlights howmuch this specificity should be the basic require-ment for a functionality test and how fundamen-tal a well-focused test can be during development.Taking the reasoning one step further, it can alsobe reflected that if a functionality test is generaland is included in a monograph, it may not berelevant to all applications. This highlights thedifficulties encountered by the Pharmaceutical In-dustry to agree on a definition of functionalitytests. It is especially true for polymer excipientsthat are incorporated into numerous different for-mulations, where they do not necessarily fulfill thesame function.

Acknowledgements

The authors are indebted to E.R. Morris, SilsoeCollege, Cranfield, UK for the measurement ofintrinsic viscosity, and to R. Clark, Kelco, SanDiego, USA for sharing his expertise in TPA andGel Strength Determination.

References

Adlers Physiology of the Eye, Moses R.A. (Ed.), C.V. Mosby,St. Louis, Missouri, 1981, p. 74.

Greaves, J.L., Wilson, C.G., Rozier, A., Grove, J., Plazonnet,B., 1990. Scintigraphic assessment of an ophthalmic gellingvehicle in man and rabbit. Curr. Eye Res. 9 (5), 415420.

Mazuel, C., Friteyre, M.C. Composition pharmaceutique de typea` transition de phase liquide-gel. French Patent Fr 85 14689,3 October 1985.

Robinson, G., Manning, C.E., Morris, E.R., Properties of thebacterial polysaccharides Gellan, Welan and Rhamsan. In:Dickinson E. (Ed.), Food Polymers, Gels and Colloids. TheRoyal Society of Chemistry, Cambridge, UK, 1991, p. 2233.

Rozier, A., Grove, J., Mazuel, C., Plazonnet, B., Gelrite

solutions: novel ophthalmic vehicles that enhance ocular drugpenetration. Proc. Int. Symp. Control. Release Bioact. Mater.(1989a) 109110.

Rozier, A., Mazuel, C., Grove, J., Plazonnet, B., 1989b. Gelrite:a novel, ion activated, in situ-gelling polymer for ophthalmicvehicles. Effect on bioavailability of timolol. Int. J. Pharm.57, 163168.

Sanderson, G.R., Bell, V.L., Clark, R.C., Ortega, D., The textureof gellan gum gels. In: Phillips G.O. et al. (Ed), Gums andStabilizers for the Food Industry 4. IRL Press Ltd, Oxford,UK, 1988, pp. 219229.

Shedden, A.H., 1994. Timolol maleate in gel forming solution:a novel formulation of timolol maleate. Chibret Int. J.Ophthalmol. 10, 3236.

Szczesniak, A.S., Brandt, M.A., Friedman, H.H., 1963. Devel-opment of standard rating scales for mechanical parametersof texture and correlation between the objective and sensorymethods of textural evaluation. J. Food Sci. 28, 397.

US Pharmacopeia XXIII, 1995. Gel Strength of Gelatin, p. 1922

functionality 5

Documents