Bull. Pharm. Sci., Assiut University, Vol. 30, Part 2, 2007, pp. 193-212.

ــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــReceived in 18/10/2007, Received in revised form in 26/12/2007 & Accepted in 27/12/2007

THE EFFECT OF BINDER TYPE AND MAJOR DILUENTON THE MIGRATION AND BIOAVAILABILITY OFRIBOFLAVIN SODIUM PHOSPHATE DURING TABLETMAKING

Hassan M. ELSabbagh, Ahmed T. Nouh, Osama A. Soliman andMariza F. Boughdady

Department of Pharmaceutics, Faculty of Pharmacy, MansouraUniversity, Mansoura 35516, Egypt

ها

. ة منها

.ى لعتةتح.

(ة مختلفم)

٠,٥١(%) ٥١٠١٥(٢٥٠,٥١١,٥(%) ١,٥ (%.%)٢,٥٥١٠(چ

، °٧٠°٥٠ي.

، ، مل. تفهشاشة،

.

ب. . °٧٠ةتع

% ١٠چص

. .

Hassan M. ELSabbagh, et al.

194

٢,٩(% ٠,٥()١,٦٨() ٤,٥ .(

% ١٠بينما . نسبة هشاشةقلللمتجانچم) ٢٥() ٧,٩٥()٠,٨٧١(

چ% ١٠.٠,٥ %

.صحچ% ١٠

)١,٥% ٠,٥). ساعة

). ساعة٠,٥(.

The migration of riboflavin sodium phosphate (RSP) upondrying of its wet granules was studied through the formulation ofgranules using different diluents (lactose monohydrate andanhydrous dibasic calcium phosphate), different binders ofdifferent concentrations; polyvinyl pyrrolidone (PVP k25),methylcellulose (MC), hydroxypropylmethyl cellulose (HPMC) andgelatin at different drying temperatures (50°C and 70°C). Theprepared granules were compressed into tablets and evaluated. Invitro drug release from the formulated tablets was performed. Inaddition, in vivo study was conducted on some selected tabletbatches. The results showed that, the granules prepared withdibasic calcium phosphate showed lower migration for the drugthan those prepared with lactose. Also, drug migration decreasedwith increasing the binder concentration and viscosity. The degreeof tablet mottling was inversely proportional to the binderconcentration. Tablets prepared with 10% w/w gelatin were foundto be the least mottled ones. In addition, they showed the leastfriability percentage, the highest hardness value and the highestdisintegration time. Tablets prepared with 0.5% MC showed thehighest dissolution rate, however, those prepared with 10% gelatinhad the lowest dissolution rate. Generally, increasing the binderconcentration resulted in slowing the in vitro drug release fromtablets. The in vivo study showed that, tablets prepared with 10%w/w gelatin showed the lowest excretion rate and the highest Tmax

(1.5 hours). Meanwhile, tablets prepared using 0.5% w/w MCexhibited higher excretion rate and Tmax of 0.5 hour.

195

INTRODUCTION

Compressed tablets are one of themost widely used dosage forms forthe administration of orally effectivetherapeutic agents. They must beuniform in weight and in drug contentof the individual tablet. Also, theymust be elegant in appearance andmust have the characteristic shape,color, necessary to identify theproduct. Owing to their wide-rangingphysicochemical and mechanicalproperties, pharmaceutical powdersand their blends frequently exhibitpoor flow and compaction behavior1.Thus, an intermediate step, granu-lation is usually required in solid dosemanufacture to produce a free-flowing material with good com-pression characteristics. Wet granu-lation has been, and continues to be,the most widely used agglomerationprocess2. But migration of soluble,low-dosage, highly potent drugs or ofcolored additives upon drying of thewet granules may represent a seriousproblem. This problem is themovement of the solutes from granuleto granule, upon drying, which mayresult in gross maldistribution of theactive drug or the colored additive.As a result, excessive dose variationmay occur within the same batch oftablets. Also, the migration of thecolor may give rise to a tablet with amottled appearance.

The main target of this work, is toprepare riboflavin sodium phosphategranules, using various diluents,binders, and drying temperatures.Studying the effect of these variableson the extent of riboflavin sodium

phosphate migration upon drying ofits wet granules is also of our interest.In addition, the dried granules will beprepared in tablets form, in order tostudy the effect of drug migration onthe physical properties of thesetablets. In vivo evaluation of theprepared tablets will also, beinvestigated.

EXPERIMENTAL

MaterialsRiboflavin sodium phosphate

(Certa, Belgium). Lactose mono-hydrate, gelatin, H2O2 (30%), andabsolute ethanol (Adwic, EL-NasrPharmaceutical chemicals Co., Cairo,Egypt). Calcium phosphate anhydrousand Talc (El Gomhouria Co. Egypt).Polyvinyl pyrrolidone (PVP k25),methyl cellulose, and hydroxypropylmethyl cellulose (Memphis Pharm.Co. Egypt). Benzyl alcohol (RayasanChemicals, India).

EquipmentUSP standard sieves, Hot air

ovens (Heraeus GS model B 5042,Gering model SPA-Gelman, Instru-ment No. 16414, Germany). U.V.spectrophotometer (Jasco, V-530,Japan). Rotary viscometer (HaakeInc., Germany). MSE MinorCentrifuge (MSE scientific instru-ments, Manor Royal, CrawleyRH/0200 sussex, England). Singlepunch tablet machine, tablet hardnesstester, Roche friabilator, tabletdissolution test apparatus USP I, andUSP disintegration apparatus(Erweka-Apparatebau, G.m.b.H.,Germany). Micrometer (Mitutoyo

Hassan M. ELSabbagh, et al.

196

Corporation, Japan). thermostaticallycontrolled shaking water bath (Grantinstrument Cambridge Ltd., Barring-ton Cambridge (B2, 5002, England).Luminescence spectrofluorometer(Perkin-Elmer model LS 45,Germany).

Methodology

Effect of binder type and majordiluent on riboflavin sodiumphosphate migration during dryingof tablet granules

Preparation of different bindersolutions and determination oftheir viscosity

Four types of aqueous bindersolutions in different concentrationswere prepared. These binders are;

PVP k25 (5, 10, and 15% w/w), MC(0.5, 1, and 1.5% w/w), HPMC (0.5,1, and 1.5% w/w) and gelatin (2.5, 5,and 10% w/w).

The viscosity values of theprepared binder solutions weredetermined using rotary viscometerwhich was thermostatically controlledat 30°C ± 0.5°C.

Preparation of riboflavin sodiumphosphate wet granules

Migration of RSP was studiedthrough the preparation of RSP-containing granules using differentdiluents, binders and drying tem-peratures. Two formulations wereprepared; formula L and formula C.Their constituents are illustrated inTable (1)

Table 1: Composition of various formulations of granules containing riboflavinsodium phosphate.

Formula IngredientQuantity per 500 mg of

granules (weight of one tablet)RSP 2 mgLactose monohydrate 483 mg(L)Binder solution q.s.RSP 2 mgAnhydrous calciumphosphate

483 mg(C)

Binder solution q.s.

197

For each batch, the calculatedamount of the drug was mixedgeometrically with the major diluent(lactose monohydrate, in case offormula L), or (anhydrous calciumphosphate, in case of formula C).Sufficient quantity of the preparedbinder solution was then used for wetmassing of the mixed powders. Thewet mass was kneaded for 10 min.until homogenous colored materialwas obtained. Wet screening wasaccomplished by passing the wetgranulated mass through a 2 mmsieve (mesh number 10). A controlexperiment was performed for bothlactose and calcium phosphate byapplying the same procedure but withthe granulating fluid being wateronly.

Drying of riboflavin sodiumphosphate wet granules

For each batch, the wet granuleswere divided into two equal portionsand placed in two similar drying cells.One cell was dried at 50°C and theother was dried at 70°C in two dryingovens for 24 hours.

Drying cellRiboflavin sodium phosphate wet

granules were dried in a specialdrying cell consisting of fourconcentric rings to facilitate thedetermination of RSP concentrationsat various depths through thegranulation bed. The height of eachlayer is 0.5 cm with a total height of 2cm. The cell is open on both sides ofthe cylinder formed by the layers; the

opening was 7.9 cm in diameter(Figure 1).

(a)

(b)

Fig. 1: The drying cell employed fordrying of wet granules.

a- Assembled drying cell.b- Partially disassembled dryingcell.

When used for migration studies,the drying cell was placed on a dryingoven tray lined with a paper. Thegranulation in the bottom layer wasexposed to the paper during thedrying process. The wetted granules,after being wet sieved, was filled intothe drying cell with a spatula.

Hassan M. ELSabbagh, et al.

198

Determination of RSP concentra-tion in different layers of thegranulesSampling procedure

After drying, the top ring of thedrying cell was carefully removed.Then, the top layer of the dried bedwas removed by a flat blade, mixedthoroughly, and sampled for deter-mination of drug concentration. Threeportions, weighing 100 mg each, wereaccurately weighed from the top layerof the dried bed and used forsubsequent drug concentration deter-mination. The procedure was repeatedfor the core layers (second, and thirdlayers), and for the bottom layer also.

RSP assay methodi) For lactose granules (formula L)

Each sample was dissolved in 25ml of distilled water. The solutionswere protected from light to avoiddrug degradation. The absorbances ofthe dissolved samples were measuredat the predetermined λmax (266 nm),using distilled water as a blank. Thequantity of RSP, in each layer, wasdetermined and the percentage ofdrug in each layer was calculated.ii) For calcium phosphate granules(formula C)

Each sample was shaken with 25ml of distilled water. The mixtureswere then filtered through 0.45 µmmillipore filter, and the concentrationof RSP in the filtered samples wasmeasured at 266 nm using distilledwater as a blank. The quantity ofRSP, in each layer, was determinedand the percentage of drug in eachlayer was calculated.

The same procedure was used fordetermining the drug percentages inthe different layers for the controlexperiments.

Data treatmentTo specify the extent to which the

drug migrated with a numerical value,the concept of a coefficient ofmigration was employed. It wasdeveloped by Warren and Price,(1977 a)3. Since there are four layersin the drying cell, there can be sixlayer comparison classifications, i.e.,a comparative difference, sign ignor-ed, of the average assay values oflayer 1 to layer 2, layer 1 to layer 3,layer 1 to layer 4, layer 2 to layer 3,layer 2 to layer 4, and layer 3 to layer4. These layer differences weredetermined relative to the amount ofdrug recovered and potentiallypresent in any two layers. In general,then, the equation for calculating thecomparative difference (D) betweentwo layers (j and j΄) is:

N

L

LLD

N

j j

jj

jj ∑ −

−

−=

1

'

'2

Where; Lj represents the average ofthree assay values in a given layer.Lj´ is the average assay value inanother layer.ΣN

j=1 Lj is the sum of the averageassay values for N layers.N is the number of layers in thedrying cell.

199

It is postulated that, the drugpercent of the most divergent systemusing an open drying cell after dryingcan be represented as 200%, 0%, 0%,and 200% for layers: 1 (upper layer),2 (second layer), 3 (third layer), and 4(base layer), respectively, where thedrug has migrated from the twointermediate inner layers (second andthird layers) to the two exposedevaporating surfaces (upper and baselayers). Using this equation, theclassification differences would be:D1-2 = (200-0)/200 =1, D1-3 = (200-0)/200 =1, D1-4 = (200-200)/200 = 0,D2-3 = (0-0)/200 =0, D2-4 = (0-200)/200 =1, D3-4 = (0-200)/200 =1, andtotal = 4. So, the coefficient ofmigration is 4/4 = 1.

On the other hand, the mostuniform system in a migration studyusing an open drying cell after dryingcan be represented as layer 1 = layer 2= layer 3 = layer 4 = 100%. Using thesame equation, the classificationdifferences would be zero in all cases

and the coefficient of migrationwould be 0/4 = 0.

It follows that, as the coefficientapproaches one, the extent ofmigration is greater; as the coefficientapproaches zero, the extent ofmigration is less.

Statistical analysis of dataSpearman's rank correlation

method was used to determine thecorrelation, if any, exists between theextent of RSP migration and theincrease in the binder solutionviscosity4.

Preparation of RSP tabletsFrom formula (L), for each binder;

the formulations showing the highestand the lowest coefficient ofmigration, were chosen for sub-sequent compression into tablets. Thefour layers of each granulation bedwere mixed thoroughly beforecompression. Thus, the preparedformulations are shown in Table (2).

Table 2: RSP tablet batches made using various binder concentrations.

Lactose granulation specificationsFormula

Type and concentrationof binder (w/w)

Temperature of drying(°C)

Migrationcoefficient

P1 5 % PVP 50 0.67P2 15 % PVP 70 0.198M1 0.5 % MC 50 0.423

M2 1.5 % MC 70 0.089

H1 0.5 % HPMC 50 0.525H2 1.5 % HPMC 70 0.265G1 2.5 % gelatin 50 0.467G2 10 % gelatin 70 0.004

Hassan M. ELSabbagh, et al.

200

The mixed components of eachformula were compressed into tabletsusing a single-punch tablet machine(12 mm diameter). The tablet weightwas adjusted to be 500 mg. Tabletswere visually inspected for mottlingand evaluated for; uniformity ofweight, drug content, uniformity ofthickness, friability using Rochefriabilator at 25 r.p.m, for 4 minutes,hardness using Erweka hardnesstester and disintegration time (USPprocedures were used).

In vitro release studyDrug release from the prepared

tablets was performed using tabletdissolution apparatus USP I at 50rpm. The dissolution medium was100 ml of distilled water, maintainedthermostatically at 37±0.5°C. Ali-quots of one ml were withdrawn,filtered through millipore filter (0.45µm), then, diluted with the dissolutionmedium which is replaced with thesame volume of fresh medium tomaintain it constant. The releasedamounts of RSP from each formulawere analyzed spectrophotometricallyat 266 nm using distilled water as ablank. The experiments were done intriplicates and the mean wascalculated.

In vivo evaluationTablets prepared with 0.5% w/w

MC (batch M1) and those preparedwith 10% w/w gelatin (batch G2),were chosen for the in vivo study,since they showed the lowest and thehighest in vitro dissolution rate, res-

pectively. A control batch (identicalbatch but with no binder) was usedfor comparison. Six healthy malevolunteers were selected for thestudy. The level of riboflavin in urinewas measured after administration oftwo tablets, (2 mg RSP in each), byusing a spectrofluorimetricmethod5&6.

Analysis of urine samplesThe level of riboflavin into urine

was measured by using aspectrofluorimetric method. In thiswork, the excitation and emissionwavelengths were found to be 445and 520 nm, respectively.

Four ml of urine were introducedinto a centrifugation tube and 0.4 mlacetic acid buffer of pH 3.5 wasadded. The oxidation was started bythe addition of 1 ml of KMnO4

solution (4%, w/v). After 1 min, theexcess of KMnO4 was reduced byaddition of 0.1 ml of H2O2 solution(30%). The residue was thenextracted by the addition of 4 g ofammonium sulphate and 3 ml ofbenzyl alcohol saturated with water.The tubes were shaken for 3 min andthen centrifuged at 3000 rpm for 5min. One ml of the supernatant wastaken and diluted with 7 ml of asolution containing 45% (v/v) ofethanol in a solution of acetic acidand sodium acetate with aconcentration of 0.1 N of each6. Thesamples were then measured spectro-fluorimetrically at excitation andemission wavelengths of 445 nm and520 nm, respectively.

201

Pharmacokinetic analysisThe prepared batches were

compared in terms of urinaryrecovery of riboflavin during the first24 hr after administration, maximumurinary excretion rate (R max), and thetime (Tmax) required to reach Rmax. Allparameters were determined from theindividual urinary excretion rate–timecurves, a plot of urinary excretion rateagainst the mid-point of urinecollection interval.

RESULTS AND DISCUSSION

The viscosity values for theprepared binder solutions areillustrated in Table (3). Following wetgranulation and drying of the twoformulations (L and C) containingRSP, it was found that, the granuleswere no longer uniform as to RSPcontent. Migration of the drugsubstance had occurred during drying.

Table 3: The viscosity values ofbinder solutions in differentconcentrations.

Bindertype

Binder concentration(% w/w)

Viscosity(mPa.s)

5 6.4410 11.5PVP15 17.440.5 15.61 24.9MC

1.5 69.80.5 16.11 35.7HPMC

1.5 72.52.5 10.55 13.9Gelatin

10 66.55

RSP, being water soluble, followedthe solvent as it moved upward anddownward during the drying period.Higher concentrations of the drugwere found in the outer layers, with areduced amount toward the center ascompared with the initial blendedmaterial.

Table (4) shows that, for a specificbinder, the values of migrationcoefficient had a descending order, onincreasing the binder solutionconcentration or viscosity. Forformula (L), dried at 50°C, the valuesof migration coefficient were 0.67,0.418 and 0.291 for granules preparedwith 5, 10 and 15% w/w PVP,respectively. For MC, the values ofmigration coefficient were 0.423,0.265 and 0.188 for granules preparedwith 0.5, 1 and 1.5% w/w MC,respectively. For HPMC, the valuesof migration coefficient were 0.525,0.375 and 0.275 for granules preparedwith 0.5, 1 and 1.5% w/w HPMC,respectively. While for gelatin, thevalues of migration coefficient werefound to be 0.467, 0.098 and 0.008for granules prepared with 2.5, 5 and10% w/w gelatin, respectively.

So, there is an obvious decrease inthe values of the coefficient ofmigration with increasing the binderconcentration and hence viscosity.These results are in agreement withthat reported by Warren and Price,(1977 b)4, who proved that, drugmigration decreased with increasedbinder solution viscosity, whereincreasing the concentration andtherefore, the viscosity of PVPsolution has been shown to slow the

Hassan M. ELSabbagh, et al.

202

migration of propoxyphene hydro-chloride in fixed bed of wet granules.This also was proved by Kiekens etal., (1999)7, who conclud-ed that, theintragranular migration decreased asthe granulating liquid viscosity ofpovidone increased. The viscosity ofgranulating fluids im-pedes themovement of moisture by increasingthe fluid friction8.

Spearman's rank correlationmethod was used to determine if acorrelation exists between the extentof RSP migration and increasing thebinder solution viscosity. The valuesof r were; - 0.7198, - 0.6264, - 0.6758and - 0.7363 for formula L (dried at50ºC), Formula L (dried at 70ºC),formula C (dried at 50ºC) andformula C (dried at 70ºC), respec-tively, which are considered signifi-cantly different than zero. (When rvalue approaches 1, whatever thesign, there is a good correlation

between the two studied variables.While, in case of r value equals zero,there is no correlation.). A significantnegative correlation indicates that,there was an inverse relationshipbetween the extent of RSP migrationand increasing the binder solutionviscosity.

Table (4) reveals that, themigration coefficient and hence themigration of the drug at 70°C is lessthan its analogue obtained at 50°C.

Newitt & Papadopoulos, (1959)9

and Pietsch & Rumpf, (1966)10

postulated that, solute migration ondrying is reduced as the dryingtemperature is raised. They suggestedthat, at high temperatures, the flow ofliquid through the granule cannotmaintain the higher rate of drying atperiphery and evaporation takes placefrom progressively further inside thegranule. Travers, (1975)11 stated that,

Table 4: Migration coefficient of RSP using various binders and diluents fromgranules dried at different temperatures.

Lactose (L) Calcium phosphate (C) Binder type andconc. (%w/w) 50°C 70°C 50°C 70°C

PVP5

1015

0.670.4180.291

0.5970.3370.198

0.4750.2870.207

0.4250.2820.143

MC0.51

1.5

0.4230.2650.188

0.2860.2270.089

0.1770.0880.046

0.1190.0890.024

HPMC0.51

1.5

0.5250.3750.275

0.4390.3360.265

0.2530.1910.132

0.2030.1720.025

Gelatin2.55

10

0.4670.0980.008

0.3350.0760.004

0.1330.0510.008

0.110.043

0.0Water 0.765 0.57 0.602 0.434

203

solute migration is likely to begreatest when the evaporation rate isfairly slow at low rates of heattransfer. While, at higher rates, fluidfriction will act to retard the moisturemovement so that, the granule surfacebecomes dry at a stage when lessoverall movement has taken place.

To investigate the effect of thetype of major diluent, it is observedfrom Table (4) that, formula C has alower coefficient of migration thanformula L, i.e. calcium phosphateshows lower migration for RSP thanlactose. These results are inagreement with those obtained byChaudry and King (1972)12, whoassumed that, calcium phosphate hadmore drug migration inhibitorycharacteristics than both lactose andcalcium sulphate. Also, Stewart et al.,(1979)13 reported the adsorption ofriboflavin onto water insolublediluents, including Emcompress(dibasic calcium phosphate). Theadsorption of riboflavin onto thesurface of Emcompress molecules,may account for the retardation of itsmigration to the surface of thegranulation bed during drying, andhence, lower values of migrationcoefficient.

Tablet evaluationVisual inspection of the preparedtablets

Figures (2-5) show differentbatches of the prepared tablets,exhibiting different degrees ofmottling. From Figure (2), it can beobserved that, tablets prepared fromthe granules made with 5% PVP

(batch P1), exhibit high degree ofmottling, while tablets prepared fromthe granules made with 15% PVP(batch P2), are somewhat moreuniform in color. Figure (3) showsthat, tablets prepared from thegranules made with 0.5% MC (batchM1), exhibit high degree of mottling.On the other hand, those preparedfrom the granules made with 1.5%MC (batch M2), are more uniform incolor. Figure (4) reveals mottling forbatches H1 and H2, (tablets preparedfrom the granules made with 0.5 and1.5% HPMC, respectively). Also,Figure (5) shows mottling in tabletsprepared from the granules made with2.5% gelatin (batch G1), while thoseprepared from the granules made with10% gelatin (batch G2) are found tobe the most homogenously coloredones. Thus, the migration is very clearwith the less viscous binder solutions(5% PVP, 0.5% MC, 0.5% HPMC,and 2.5% gelatin), and so contributesfor the mottled appearance of theprepared tablets, as observed inbatches: P1, M1, H1, and G1. For thehigher concentrations of the samebinders, it could be observed that,more uniformly colored tablets wereobtained. This means that, a moreuniform distribution for the drug wasobtained upon increasing the binderconcentration. This is due to theincrease in the viscosity of the bindersolutions which in turn, retarded theinter-granular migration of thecolored drug into the surface of thegranulation bed and thus reducedmottling.

Hassan M. ELSabbagh, et al.

204

(a)

(b)Fig. 2: RSP tablets prepared with PVP

K25 bindera-From granules made with concen-

tration 5 % w/w (batch P1).b-From granules made with concen-

tration 15 % w/w (batch P2).

(a)

(b)

Fig. 3: RSP tablets prepared with MCbinder

a-From granules made withconcentration 0.5 % w/w (batchM1).

b-From granules made withconcentration 1.5 % w/w (batchM2).

(a)

(b)Fig. 4: RSP tablets prepared with

HPMC bindera-From granules made with concen-

tration 0.5 % w/w (batch H1).b- From granules made with concen-

tration 1.5 % w/w (batch H2).

(a)

(b)Fig. 5: RSP tablets prepared with

gelatin bindera-From granules made with

concentration 2.5 % w/w (batchG1).

b-From granules made withconcentration 10 % w/w (batchG2).

205

Uniformity of weight and thicknessTable (5) reveals that, all the

batches of the prepared RSP tabletsshowing a good uniformity of weightand thickness.

FriabilityFrom the same Table, it could be

observed that, batches P2, and G2comply with the USP XXVIIrequirements for the friability (% lossdoes not exceed 1%). Also, it couldbe observed that, for PVP batches, thepercentages of friability were 1.2%and 0.923% for batches of tablets P1,P2 respectively. Thus, increasing thebinder concentration from 5% w/w(batch P1) to 15% w/w (batch P2),led to a decrease in friabilitypercentage. These results are inagreement with that obtained by El-Sabbagh et al., (1981a)14, who foundthat, the increase in the concentrationof a specific binder produces morestrong granules which on compre-ssion, produces tablets of greatest

hardness, lowest friability percen-tages, and highest disintegration time.

However in case of MC, bothbatches M1 and M2, did not complywith the USP XXVII requirements forthe percent friability since the valueswere > 1% (2.9% and 1.8% forbatches M1 and M2, respectively).

Table (5) shows that, the samepattern was obtained for HPMC, witha value of 2.1% and 1.1% for batchesH1 and H2, respectively.

In the case of gelatin bindersolution, Table (5) shows that, 2.5%w/w concentration produced tabletsof 1.93 percent friability (batch G1).On increasing gelatin concentration to10% w/w, the percent friability wasdecreased to 0.871% (batch G2).

HardnessFor all binders, the hardness

values were increased with theincrease in binder concentration, asillustrated in Table (5).

Table 5: Physical properties of riboflavin sodium phosphate tablets, preparedwith different concentrations of binders.

FormulaWeight

(mg)Mean ± SD

Thickness(mm)

Mean ± SD

%Friability

Hardness(Kg)

Mean ± SD

Disintegrationtime (min.)

Dissolutionrate (T90)

(min.)

P1 496 ± 2.00 2.998 ± 0.021 1.2 5.56 ± 0.625 2.6 10

P2 501.6 ± 1.89 3.036 ± 0.009 0.923 7.93 ±1.1 7.7 15.5M1 498 ± 2.05 3.034 ± 0.015 2.9 4.5 ± 0.866 1.68 9

M2 498.2 ± 2.2 3.02 ± 0.012 1.8 5.44 ± 0.987 3.9 17

H1 500.2 ± 2.44 3.046 ± 0.011 2.1 4.8 ± 0.629 2.34 8.5H2 499.9 ± 2.02 3.044 ± 0.011 1.1 6.3 ± 1.02 10.55 37

G1 499 ± 2.28 3.04 ± 0.007 1.93 5.75 ± 0.762 4.07 40

G2 499.6 ± 2.36 3.036 ± 0.019 0.871 7.95 ± 1.2 25 170

Hassan M. ELSabbagh, et al.

206

Batch M1, (made with 0.5% w/wMC), produced tablets with thelowest hardness values while batchG2, (made with 10% w/w gelatin),produced the highest values within allbatches. The high hardness valuesobtained with gelatin correspondswith its known property of producinghard tablets, as stated by Healy et al.,(1974)15. The hardness results wereparallel and confirm those offriability, i.e. as the hardness increas-ed, the percent of friability decreasedand vice versa. Also, these resultswere in agreement with thoseobtained by El- Sabbagh et al.,(1981b)16.

DisintegrationFrom Table (5), it could be

observed that, all batches complywith the USP XXVII requirements fordisintegration time in distilled water.Increasing the binder concentration isfollowed by an increase in thedisintegration time.

Esezobo and Pilpel, (1976 &1977)17&18 confirmed these results.Also, Davies and Gloor, (1972)19

observed an increase in disintegrationtime of lactose tablets prepared with

higher concentrations of bindingagents and they attributed this to theincreased binding capacity of thesegranulating agents at higherconcentrations.

Uniformity of drug contentTable (6) reveals that, the range of

assay values was the widest in case ofbatch M1 (which exhibited aconsiderable drug migration). On theother hand, batch G2 (whichexhibited the least drug migration),showed a narrow range of assayvalues. The high range of assayvalues may indicate that, the drug isless homogeneously distributed in thisbatch, and vice versa.

Warren and Price (1977b)4,proved that, the formula granulatedwith binder solution of viscosity 3 cps(formula exhibiting considerable drugmigration), had an average assayvalues of 88.1-102.6% of thetheoretical amount of drug/tablet.While, the formula granulated with600 cps binder solution (formulashowing little drug migration), had anaverage assay values of 96.1-102.4%of the theoretical amount ofdrug/tablet.

Table 6: Riboflavin sodium phosphate content in different tablet batches.

Batch codeAverage amount recovered per

tablet (mg)Range of assay values of

RSP / tablet (%)P1 1.756 ± 0.126 77.70 – 96.35P2 1.826 ± 0.075 84.75 – 97.75M1 1.754 ± 0.212 74.00 – 108.5M2 1.878 ±0.069 88.10 – 99.00H1 1.722 ± 0.129 76.00 – 94.80H2 1.824 ± 0.142 75.50 – 98.00G1 1.843 ± 0.109 80.50 – 99.35G2 1.904 ± 0.055 91.00 – 98.20

207

In vitro drug releaseResults of drug release from

different formulations are illustratedin Figures (6-9). Figure (6) shows theeffect of PVP concentration on thedrug release from different batchesprepared with various concentrationsof PVP. It could be observed that,batch P1 has a higher dissolution ratethan batch P2. This was indicated bythe t90 value, being 10 and 15.5 minfor batches P1 and P2, respectively.Zubair et al., (1988)20, explained whyincreasing the binder concentration,increases markedly the disintegrationand dissolution times of the tablets.This is because, binders are forcedinto the interparticular spaces,thereby, increasing the area of contactbetween particles, leading to theformation of additional solid bonds.

Fig. 6: Effect of PVP concentration onRSP release from its tablets.

It was also suggested that, at lowercompression forces, dissolution rate isrelated to the structure and crushingcharacteristics of the granules, and,therefore, to the concentration of PVPbinder21.

Chalmers and Elworthy, (1976)22

noted that, increasing concentrationof PVP, resulted in a decrease in therate of tablet dissolution. Theyattributed this to the heavier coatingof the powder particles with PVP athigh concentrations which may act byslowing the rate at which invadingwater reaches the surface of thepowder particles, and slowingdiffusion away from the surface of thedrug, due to higher viscosity of PVPsolutions compared with that of purewater.

Figures (7-9) illustrate the effectof MC, HPMC and gelatin,respectively on the dissolution rate ofRSP from tablets. It could be noticedthat, the batches prepared withbinders of low viscosity (batches M1,H1 and G1) exhibited more rapiddissolution rate (t90 = 9, 8.5 and 40min, respectively), than thoseprepared with binder solutions of highviscosity (batches M2, H2 and G2which had t90 of 17, 37 and 170 min,respectively).

Fig. 7: Effect of MC concentration onRSP release from its tablets.

0

20

40

60

80

100

120

0 10 20 30 40 50

Time (min.)

% D

rug

rele

ased

Formula P1

Formula P2

0

20

40

60

80

100

120

0 10 20 30 40 50Time (min.)

% D

rug

rele

ased

Formula M1Formula M2

Hassan M. ELSabbagh, et al.

208

Fig. 8: Effect of HPMC concentrationon RSP release from its tablets.

Fig. 9: Effect of gelatin concentrationon RSP release from its tablets.

Pharmacokinetic analysisFigure (10) shows that, tablets

prepared with 10% gelatin as a binder(batch G2) exhibited the least amountexcreted, while tablets prepared with0.5% w/w MC (batch M1) and the

control batch, prepared with water,showed the highest amount excretedin urine. It could be observed that,there is no difference between M1and the control batches.

0

0.5

1

1.5

2

2.5

3

3.5

0 2 4 6 8 10 12 14 16 18 20 22 24 26

Time (hr)

Mea

n cu

mul

ativ

e am

ount

of

ribo

flav

in e

xcre

ted

in u

rine

(m

g)

Batch G2

Batch M1The control batch

Fig. 10: Average cumulative amount ofriboflavin excreted in the urineafter oral administration of 4mg RSP from different tabletbatches.

The average excretion rates ofriboflavin from different tabletbatches are shown in Table (7). Thedata were analyzed by the analysis ofvariance test (ANOVA) and theresults were illustrated in Table (8).From these Tables, it could beobserved that, there is a significantdifference between the averageexcretion rates of the three testedbatches at time of 0.5 hr.

0

20

40

60

80

100

120

0 20 40 60 80

Time (min.)

% D

rug

rele

ased

Formula H1Formula H2

0

20

40

60

80

100

120

0 50 100 150 200 250

Time (min.)

% D

rug

rele

ased

Formula G1Formula G2

209

Table 7: The average excretion rate * (mg/hr) of riboflavin after oraladministration of 4 mg RSP from different tablet batches.

Excretion rate (mg/hr)(Mean ± SD)Time (hr)

Batch (G2) Batch (M1)The control

batch0.5 0.546 ± 0.135 1.445 ± 0.299 1.259 ± 0.3251.5 0.629 ± 0.124 0.569 ± 0.031 0.563 ± 0.1432.5 0.338 ± 0.067 0.284 ± 0.118 0.303 ± 0.1413.5 0.198 ± 0.044 0.165 ± 0.095 0.251 ± 0.1355 0.071 ± 0.027 0.070 ± 0.055 0.091 ± 0.0367 0.058 ± 0.014 0.049 ± 0.029 0.055 ± 0.020

11 0.032 ± 0.010 0.024 ± 0.026 0.028 ± 0.01319 0.018 ± 0.009 0.008 ± 0.013 0.011 ± 0.006

• Average of six volunteers

Table 8: Analysis of variance test forthe rate of excretion ofriboflavin after oraladministration of 4 mg RSPfrom different tabletbatches.

Time (hr) F. value0.5 19.04 *1.5 0.6393 •2.5 0.0353 •3.5 1.143 •5 0.509 •7 0.221 •

11 0.264 •19 1.642 •

* Extremely significant difference • Non significant difference among the

tested batches by analysis of varianceat 5 % level.

In Figure (11), the urinaryexcretion rate was plotted against the

mid-point of a urine collectioninterval. From this curve, it could benoticed that, the average peaks (R max)were 0.629, 1.445, and 1.259 mg/hrfor batches G2, M1, and the controlbatch, respectively.

Fig. 11: Average urinary excretion rateof riboflavin after oral admini-stration of 4 mg RSP fromdifferent tablet batches.

Time (hr)

Rate

of e

xcre

tion

of ri

bofla

vin

(mg/

hr)

0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.00.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

Batch G2

Batch M1

Control batch

Hassan M. ELSabbagh, et al.

210

The values of Tmax were found tobe 1.5 hr for batch G2, and 0.5 hr forboth, batch M1 and the control batch.Analysis of data between each twopairs of batches, by student's t- test(Table 9), revealed a significantdifference between batches G2 andM1, as well as between batches G2and the control batch for the value ofRmax. On the other hand, nosignificant difference was foundbetween batches M1 and the controlbatch for the value of Rmax. Theseresults emphasize that, formulationchanges in a drug product may affectthe rate of drug bioavailability. Thenature and quantity of the binderemployed has a profound effect onthe bioavailability of the drug fromthe tablet dosage form. The value ofTmax for batch G2 was more than itsanalogues for both, batch M1 and thecontrol batch. This may be attributedto the difference in the dissolutionrate of the tested batches, where, thevalues of T90% were; 170 and 9 minfor G2 and M1 batches, respectively.

Table 9: Statistical t-test for themaximum excretion rate(Rmax) of riboflavin for G2,M1 and control batches.

The two compared batches Value of t

Batch G2 and the control batch 4.442 *

Batch M1 and the control batch 1.030Batch G2 and batch M1 6.188 **

* Highly significant** Extremely significantT value critical at 10 degree offreedom and 95% confidence level is2.23.

On studying the bioavailability ofriboflavin, Khalil et al., (1991)23,found that, tablets belonging to thebrand that contained gelatin in thesubcoat exhibited poorer dissolutionprofiles than the other brandcontaining no gelatin.

ConclusionFrom the previous work, it could

be concluded that;•Migration of riboflavin sodium

phosphate as an example of colored,low dose, water soluble drug withthe solvent during drying of wetgranules results in, an unevendistribution of the drug substancewithin the granulation bed.

•Increasing the binder solutionviscosity, decreases the inter-granular migration of the testeddrug.

•Riboflavin sodium phosphate mostlyshowed higher migration at a dryingtemperature of 50°C than 70°C.

•The type of major diluent employedhas a profound effect on the drugmigration, where dibasic calciumphosphate showed lower migrationthan lactose for riboflavin sodiumphosphate.

•Mottling is extensively observed forbatches prepared with low viscositybinder solutions, while, it diminisheson using high viscous bindersolutions.

•Using 10% w/w gelatin, as a binderin the wet granulation method, and70ºC as the drying temperature ofthe wet granules, minimized themigration of the studied drug, a casewhich led to homogenous

211

distribution of the drug in theprepared tablets which alsoprovoked almost no mottling.

•Tablets prepared with 10% w/wgelatin showed the most uniformdrug distribution, but had slowerdissolution rate, and higherdisintegration time, compared to theother batches.

•The in vivo study revealed a lowerexcretion rate, higher Tmax and goodcorrelation with the in vitro data forriboflavin sodium phosphate tabletsprepared with 10% gelatin (batchG2). However, tablets prepared with0.5% w/w MC (Batch M1) exhibitedhigher excretion rate than batch G2.

REFERENCES

1- I. Krycer, D. G. Pope and J. A.Hersey, Powder Technol., 34, 39(1983).

2- L. L. Augsburger and M. K.Vuppala, "Theory ofGranulation" in: "Handbook ofPharmaceutical GranulationTechnology", D. M. Parikh,Marcel Dekker, New York,1997, pp. 7-13.

3- J. W. Warren and J. C. Price, J.Pharm. Sci., 66, 1406 (1977a).

4- J. W. Warren and J. C. Price,ibid., 66, 1409 (1977b).

5- H. B. Burch, O. A. Bessey andO. H. Lowry, J. Biol. Chem.,175, 457 (1948).

6- J. Hamdani, J. Goole, A. J. Moësand K. Amighi, Int. J. Pharm.,323, 86 (2006).

7- F. Kiekens, R. Zelko and J. P.Remon, Pharm. Dev. Technol.,4, 415 (1999).

8- M. Aulton, "Drying", chapter 26,in: "Pharmaceutics: The Scienceof Dosage Form Design", 2nd

Ed., M. E. Aulton, ChurchillLivingstone, Edinburgh, 2002, p.395.

9- D. M. Newitt and A. L.Papadopoulos, Proc. Fert. Soc.,55, 3 (1959).

10- W. B. Pietsch and H. Rumpf,Colloq. Int. C.N.R.S., 160, 213(1966).

11- D. N. Travers, J. Pharm.Pharmacol., 27, 516 (1975).

12- I. A. Chaudry and R. E. King, J.Pharm. Sci., 61, 1121 (1972).

13- A. G. Stewart, D. J. W. Grantand J. M. Newton, J. Pharm.Pharmacol., 31, 1 (1979).

14- H. M. El-Sabbagh, A. H.Ghanem and H. M. Abdel-Alim,Pharmazie, 36, 548 (1981a).

15- J. N. C. Healy, M. H. Rubinsteinand V. Walters, J. Pharm.Pharmacol., 26 supp., 41P(1974).

16- H. M. El-Sabbagh, A. H.Ghanem and M. H. El-Shaboury,Pharmazie, 36, 488 (1981b).

17- S. Esezobo and N. Pilpel, J.Pharm. Pharmacol., 28, 8 (1976).

18- S. Esezobo and N. Pilpel, J.Pharm. Sci., 66, 852 (1977).

19- W. L. Davies and W. T. Jr., ibid.,61, 618 (1972).

20- S. Zubair, S. Esezobo and N.Pilpel, J. Pharm. Pharmacol., 40,278 (1988).

Hassan M. ELSabbagh, et al.

212

21- H. L. Smith, C. A. Baker and J.H. Wood, ibid., 23, 536 (1971).

22- A. A. Chalmers and P. H.Elworthy, ibid., 28, 288 (1976).

23- S. A. Khalil, N. S. Barakat andN. A. Boraie, STP. Pharma. Sci.,1, 189 (1991).