91.

Pasqualoto K. F. M., Teófilo R. F., Guterres M., Pereira F. S., Ferreira M. M. C., "A study of physicochemical and biopharmaceutical properties of Amoxicillin formulations developed by direct compression using full factorial design with multiple responses". Águas de Lindóia, SP, Brazil, 10-15/09/2006: 10th International Conference on Chemometrics in Analytical Chemistry (CAC-2006, CAC-X), Book of Abstracts (2006) P074. Poster 074.


10th International Conference on Chemometrics in Analytical Chemistry P074

A study of physicochemical and biopharmaceutical properties
of Amoxicillin formulations developed by direct compression
using full factorial design with multiple responses

Kerly F. M. Pasqualoto*, Reinaldo F. Teófilo, Marco Guterres, Flávia S. Pereira, Márcia
M. C. Ferreira   kerlyfmp@iqm.unicamp.br

Theoretical and Applied Chemometrics Laboratory, Department of Physical Chemistry, Institute of
Chemistry, The State University of Campinas, Campinas, SP, Brazil

Keywords: factorial design, direct compression, physicochemical properties
_____________________________________________________________________________________
 

    Direct  compression  technique  presents  an  economic  advantage   in  time,   costs  and  energy,   when
compared to  the  traditional  granulation  methods.1   The  variables  that  influence  the tablets obtained  by
direct  compression  method  deserve  to  be  studied   to  minimize  formulation  costs  and   to  improve  the
physicochemical   and   biopharmaceutical    properties   of   the   resulting   compacts.   Amoxicillin  500  mg
formulations  were  previously  developed  by  direct  compression,2   and  the  adjuvants  investigated  were
microcrystalline cellulose,   spray-dried lactose,   and  croscarmellose sodium,   which is a superdisintegrant
agent.   In this study,   the  purpose  was  to  explore the adjuvants  (independent  variables)  effects  on  the
resulting     amoxicillin    tablet     formulations     considering    multiple    responses  (physicochemical    and
biopharmaceutical properties - dependent variables),  as  well as to  indicate  the  most  suitable  formulation
composition.
    A full  23  factorial design was  built to eight different  amoxicillin formulations,  each one containing three
replicate batches, and eight responses were obtained.   Each  independent variable  was investigated at two
levels:  the  microcrystalline  cellulose  (MCC)  type  Avicel®  PH-102  (low)  or  PH-200  (high),  the absence
(low)  or  presence  (high)  of  spray-dried lactocese  (LAC),  and  the absence  (low)  or  presence  (high)  of
properties  as  average  weight,   thickness  and   diameter,   hardness,   friability,   amoxicillin  concentration
(iodometric  assay),  disintegration  time,  and  dissolution  profile.2    The   responses  more  relevant  to  the
distinct formulations  for  the  experimental  design  were  hardness,  friability,  and  the amount of amoxicillin
dissolved   during   the   first  hour.   Three  models  were   built   for  each  response,   using  multiple  linear
regression (MLR).  The coefficients were considered as independent variables, and the responses  obtained
to this new set was  the  optimum response for each original variable.   A  new model was  constructed,  and
its respective coefficients  indicate  that  the level of each variable should be fixed to obtain better responses.
A  criterious  statistical  evaluation   considering   each   response  individually,  was   also   performed.   The
comparison between  the  best  levels  from  simultaneous  and  individual  response evaluation  was  carried
out.
    Regarding  each  response  individually,   the   MCC  and  DIS  levels  were  significant  and   negative  for
hardness  (see  the side  table).   The  variables  LAC  and  DIS  presented   an  interaction  for   the  friability
response.  The presence  of  LAC and  the absence  of  DIS  increased the friability.   The  MCC type  Avicel®
PH-200  also contributed to increase of that physicochemical  property.  Tablet  formulations  presenting  high
_____________________________________________________________   friability    values     are      not
                                                                                                                             recommended. The amount of
                   Hardness                 Friability      Amount Amox. Dissolv.   Multiple Response  drug disolved  presented  two
                  ________________________________________________   interactions,  one between the
                       Coeff.      p            Coeff.        p           Coeff.      p        Variables    Levels     MCC  and   DIS   levels,   and
                  ________________________________________________  another between the LAC and
Mean          107.13   0.000          0.20    0.000           3.62    0.000        MCC             0           DIS  levels.  The  combination
MCC               -6.46  0.021           0.03    0.004          -0.20    0.032        LAC       -1.33           of MCC low level (Avicel® PH-
LAC                 -2.92  0.263          0.05    0.000          -0.31    0.002        DIS          6.03          102), LAC low level (absence),
DIS                -6.13   0.027           0.00    0.895            0.07    0.394                                           and DIS high level  (presence)
MCCxLAC     0.00   1.000           0.20    0.058            0.00    0.973                                           simultaneously  increases  the
MCCxDIS     -1.63    0.527          0.02     0.081          -0.22    0.020                                           amount   of    drug    dissolved,
LACxDIS       -3.92   0.139         0.03     0.006          -0.23    0.014                                           considering the plot of marginal
_____________________________________________________________   means analysis. The individual
response  evaluation   indicated   that   the   most  suitable  amoxicillin  table  formulation  should  present  the
microcrystalline cellulose  Avicel®  PH-102  and  the  superdisintegrant agent,   croscarmellose sodium,   in its
composition.   The  simultaenous  and  individual  response  analysis  generated similar results  (table above),
except to  the MCC level.   The multiple response analysis did not allow taking any information related  to  the
type of MCC  (level 0)  in the most suitable amoxicillin table formulation.
 

Acknowledgment. CNPq and FAPESP for financial support.
_____________________________________________________________________________________
References

1 Shangraw, R.F. (1989) Pharmaceutical Dosage Forms: Tablets, v. 1, New York: Marcel Dekker Inc.
2 Pasqualoto, K.F.M., Funck, J.A.B., da Silva, F.E.B., Kratz, C.P. Acta Pharm. Bonaerense 2005, 24, 39-47.
3 Teófilo, R.F., Ferreira, M.M.C. Quim. Nova 2006, 29, 338-350.