Amoxicillin-clavulanate is antibiotic combination with broad spectrum activity and widely used in the world. Clavulanate was combined with amoxicillin to increase antibiotic potency by inhibit beta-lactamase enzyme (Malik et al., 2006; Dufour et al., 2005; Smith et al., 1998; Storm et al., 2004). However, there was a fact which showed that in the dosage forms, clavulanate also increase antibiotic potency of amoxicillin toward non-beta lactamase bacterial with un-explained mechanism (Smith et al., 1998; Storm et al., 2004; Quach et al., 2005). It was already founded about high variability of dosage forms quality and its pharmacokinetic profiles, beside low stability of and incompatibility of amoxicillin-clavulanate (Vree et al., 2003; Chadha et al., 2003; Vahdat and Sunderland, 2001). Chadha et al. (2003) have reported incompatibility of amoxicillin-clavulanate in the liquid dosage form and measure the interaction energy by solution calorimeter. Vahdat and Sunderland (2001) have presented amoxicillin-clavulanate instability even in the frozen condition. Vree et al. (2003) reported the high variability of the dosage form quality.
We have investigated the solid state interaction between amoxicillin-clavulanate on the reason that so much solid dosage forms of this antibiotic combination have been produced. Instead, this combination has tendency to build a co-crystal which has not been studied yet. Physically, amoxicillin and clavulanate have similar structure and similar 3-D arrangement which could co-arrange a new crystal lattice on a few ways (Cartensen, 2001; Vednere, 1990). Chemically, these compounds have p-Ka value with differences it value more than 3, the important criteria to arrange a co-crystal (Stahly, 2007). Recently, co-crystals formation becomes important issue because of its potency to develop new pharmaceutical beside useful to study the activity mechanisms which have not been solved (Stahly, 2007; Caira, 2007).
We have reported the results of preliminaries researches in several steps (Nugrahani
et al., 2007a, f). The first step have reported that interaction between
amoxicillin and clavulanate can be clearly observed under polarize microscope
and measurable with solution calorimeter in NaOH. The interaction observed extremely
at molar ratio 3:7; 5:5 and 7:3, respectively (Nugrahani et al., 2007a).
In the next research it was showed that interaction at the same molar ratio
also observed visually and measured distinctly with solution calorimeter in
HCl 0.1 N and buffered phosphate 6.8 solution as a model of non-enzymatic gastrointestinal
tract solutions (Nugrahani et al., 2007b). After that, it have concluded
that the hydrates of amoxicillin tri-hydrate could be bounded to potassium clavulanate
which predicted produced an intermolecular interaction (Nugrahani et al.,
2007c). Then, the recent research showed that the binary system of amoxicillin
tri-hydrate-potassium clavulanate after Freeze Drying (FD) showed significant
interaction points at molar ratios 3:7; 5:5 and 6:4; followed by investigation
to look the influence of the freeze dried binary system activity against the
non beta-lactamase bacterial Sarcina lutea sp. The results showed that
freeze dried system had lower potency compared to the physical mixture; but
in general had higher potency than freeze dried amoxicillin itself (Nugrahani
et al., 2007d, e).
Finally, it have been proved that solid of amoxicillin trihydrate-potassium clavulanate formed a solid solution system with heated in closed system at temperature 50°C for 30 min (Nugrahani et al., 2007f). The aim of this research was investigated the antibiotic potency of the solid solution system (simplified namely AC-H) against Sarcina lutea sp.
MATERIALS AND METHODS
This study was performed in January-June 2007. The thermal analysis study by DSC was conducted in January-March 2007 in Research and Development Laboratory of PT Tempo Scan Pacific Cikarang Indonesia; while the co-crystal preparation and microbiology evaluation was conducted in May-June 2007 in Pharmaceutical and Microbiology Laboratory of School of Pharmacy ITB Bandung Indonesia. Powder X-ray diffractometry was conducted in Meaning Faculty ITB Bandung Indonesia and Scanning Electron Microscope analysis was conducted in The Centre Research of Geology Bandung Indonesia. In segu times along January-June 2007.
Physical mixture and co-crystal preparation: Amoxicillin tri-hydrate (ex. Sandoz batch. No. 243Z from PT Tempo Scan Pacific) and potassium clavulanate potassium clavulanate (ex. Fermic, Mexico, batch Ko. CKA-2967 from PT Tempo Scan pacific) sieved with 100 mesh sieves (Rietsch, Germany). Amoxicillin tri-hydrate weighed in 10 vials with each weight 50 mg. Furthermore, one vial was filled with amoxicillin tri-hydrate alone while 9 other vials added by clavulanate to get molar ratios of amoxicillin tri-hydrate: potassium clavulanate = 1:9; 2:8; 3:7; 4:6; 5:5; 6:4; 7:3; 8:2; 9:1. In another vial, 50 mg of clavulanate weighed. All of the mixtures were homogenized manually (in about 5 min). Powder in all those 11 vials homogenized with motor stirrer for 10 min in closed condition, furthermore, heated in oven at 50°C for 30 min then identified as AC-H co-crystals (Nugrahani et al., 2007f).
DSC analysis: All of the physical mixtures and co-crystal heated were analyzed by Differential Scanning Calorimeter (DSC-6, Perkin Elmer, USA). Approximately 2-5 mg of each samples were heated in open aluminium pan from 30-350°C at a scanning rate of 10°C min-1 under a stream of nitrogen gas. Amoxicillin trihydrate alone, potassium clavulanate alone, amoxicillin trihydrate alone heated at 50°C for 30 min and potassium alone heated at 50°C for 30 min were measure by DSC as standard.
X-ray powder diffractometer analysis: The crystal structure were analyzed by X-Ray Powder Diffractometer (XRPD Philips PW1710 BASED, Netherland) under following condition: target/filter (monochromator) Cu, voltage 40 kV, current 30 mA, receiving slit 0.2 inches. The data were collected in the continue scan mode using a step size of 0.5 deg min-1. The scanned range was 5-40°.
SEM analysis: The morphology of physical mixture 1:1 and co-crystal 1:1 were observed by SEM (Jeoul, Japan) in magnification 1000, 2500 and 7500x.
Bacteria and detection of beta-lactamase produce: Sarcina lutea ATCC-9341 was identified and then grown in slant agar (Oxoid CM003, USA). The bacteria tested by beta lactamase tester kit (Hardy Diagnostics Beta-Lactamase); showed no beta lactamase enzyme was produced.
Inhibition determination: All of AC-H co-crystals and AC-PM were dissolved in buffered phosphate pH 6.8 to make at the rate of 100 μg mL-1 sample solutions in all sterile condition. Each of antibiotic solution was dropped 10 μL-1 on sterile paper disc which planted on cold growth medium. Four mL Sarcina lutea sp. ATCC-9341 were grown in a Petri dish contented 21 mL nutrient agar (Oxoid CM003, USA). After that, sample solutions were dropped to paper discs on the Petri dish, let for 1 h in ambient temperature and then entered into incubator at temperature 37°C for 24 h incubation. Then, measure the inhibition diameter.
Evaluation of technique reproducibility: to evaluate co-crystal preparation reproducibility, the DSC analysis was done 3 times. While the microbiology test was done 6 times, average and determined by Standard Deviation calculation and then evaluated by Student t-test (p = 0.95) to compare the antibiotic potency of AC-PM systems with amoxicillin trihydrate raw material and AC-H systems with amoxicillin trihydrate heated.
RESULTS AND DISCUSSION
The thermograms resulted from DSC analysis of AC-H 1:1 is showed in Fig.
1 with AC-PM counterparts. The thermograms of amoxicillin and potassium
clavulanate before and after heated were also presented as standard. The thermogram
profile of AC-H look different from AC-PM and then the diagram phase which were
arranged from all molar ratios thermogram data also shows different form were
showed in Fig. 2. The AC-Hs phase diagram showed the
solid dispersion profile in all molar ratios (Cartensen, 2001). XRPD analysis
resulted diffraction pattern like showed in Fig. 3. The morphology
of co-crystal was compare to the raw materials which were described in Fig.
4 these illustrations were obtained by SEM photography.
The inhibition diameters of amoxicillin-clavulanate physical mixtures (AC-PMs)
were showed at Table 1, while the inhibition diameters of
AC-H systems were showed at Table 2. The curves of log molar
fraction versus inhibition of bacterial growth are showed in Fig.
5. The inhibition curve of AC-H profile looks different compare to the AC-PM
||Thermograms of amoxicillin trihydrate, potassium clavulanate,
physical mixture 1:1 (AC-PM), heated amoxicillin, heated clavulanate and
heated cocrystal 1:1 (AC-H)
||(a) Phase diagram of AC-PMs and (b) Phase diagram of AC-Hs
||X-ray powder diffraction patterns of amoxicillin trihydrate-potassium
clavulanate physical mixture (AC-PM 1:1) and co-crystal heated (AC-H 1:1)
||SEM photographs of amoxicillin trihydrate (1000x), potassium
clavulanate (2500x), co-crystal heated (1000x), co-crystal heated (7500x)
|| Inhibition diameter of AC-PM
|| Inhibition diameter of AC-H
The data listed in Table 1 show that the averages of inhibition
diameter of AC-PM in the molar ratios 1:9 until 4:6 were similar and after that
there were a little increase at molar ratio of 5:5 until 9:1. Compare to amoxicillin
trihydrate raw material, all AC-PM systems were significantly lesser with calculated
t-value higher than t-table (t-table = 2.571; p = 0.05), as have been described
in table 1 and shown in Fig. 6. AC-PM-0, the pure single clavulanate
solution do not show any inhibition. In general, the inhibition diameters of
AC-PM systems 1-4 were lesser than AC-PM 5-10.
AC-H systems inhibition diameters against Sarcina lutea sp. were shown in Table 2. This table shows that after molar ratio 3:7 the inhibitions diameters tend to increase. The statistical significant differences were showed at the molar ratios = 4:6, 6:4; 8:2 and 9:1 which have t-value higher than t-table (p = 0.05).
Thermogram of AC-PM and AC-H in molar ratio 5:5 or 1:1 were selected to explain
the interaction between amoxicillin-clavulanate. From thermograms obtained by
DSC analysis (Fig. 1), amoxicillin trihydrate and potassium
clavulanate after heated at 50°C for 30 min showed have different profile
and temperature transition. Amoxicillin alone before heated has important transition
temperature at 110 (endothermic), 182 (exothermic), 194°C (endothermic)
and after heated the temperature change to 130, 178 and 194°C. The first
endothermic curve is anhydration, the exothermic showed re-crystallization from
trihydrate to anhydrate and the last endothermic indicate melting point. Potassium
clavulanate before heated has important exothermic transition temperature at
201.3°C (the second exothermic curve can be ignored because it is Avicels
curve which used to disperse clavulanate). After heated, clavulanate has 2 peaks
at 189.5 and 201.7°C. The physical mixture 5:5 or 1:1 (AC-PM) showed that
the exothermic curve of amoxicillin and clavulanate overlaid to be 1 curve at
201°C while the co-crystal (AC-H) overlaid and arrange 2 peaks at 189 and
201°C. The arrangement from 2 different peaks become one peak could indicate
co-arrangement phenomenon in the type of solid solution (Vednere, 1990; Cartensen,
All of data obtained from DSC then used to arrange a phase diagram which yielded
different phase diagram compare to the physical mixtures (Fig.
2a, b). AC-PM arranged a solid dispersion or co-crystallization
at 201°C in molar ratio of amoxicillin: clavulanate = 1:9 until 5:5, but
the heated systems have arranged co-crystals in all composition. It was expressed
that amoxicillin-clavulanate after mixing and heating could form a new arrangement
powder co-crystal in all molar ratios or composition.
||The curve log concentration versus inhibition diameter of
AC-PM and AC-H
||Inhibition diameter AC-H against Sarcina lutea sp.
in the ratios = 0:10 (clavulanate alone), 3:7 (3), 4:6, 5:5 (5), 7:3 (7),
8:2 (8), 9:1 (9), 10:0 (amoxicillin alone)
XRPD analysis was conducted to investigate the lattice structure and solid phase of amoxicillin-clavulanate. Figure 3 showed that the physical mixture and co-crystal have similar pattern with amoxicillin alone. Figure 4 which show SEM observation results, could explain this phenomena. Physically, clavulanate was more amorphous than amoxicillin trihydrate. After heated, clavulanate dispersed on amoxicillin surface but it still could diffract x-ray beam because of its low crystallinity, so the diffraction pattern still shows like amoxicillin alone. Co-crystals usually formed in some kind of non-covalent bonding, included hydrogen bonding formation and could change the physicochemical properties and the antibiotic potency (Stahly, 2007; Bettinetti et al., 2000; Hickey et al., 2007; Caira, 2007).
In this report, the changes of potency is related to phenomenon which showed that amoxicillin-clavulanate enhanced the potency against non beta lactamase bacteria with unknown mechanism (Storm et al., 2004; Smith et al., 1998; Quach et al., 2005). The co-crystal as a physical intermolecular interaction examined for its inhibition diameter against Sarcina lutea sp., a non beta lactamase bacteria, compared to the AC-PM counterparts.
The inhibition diameters of AC-PM 1-4 were lesser than AC-PM 5-10 caused at much of clavulanate which freely dissolved and covered the surface of bacterial Sarcina lutea sp., physically inhibit amoxicillin tri-hydrate activities. Figure 3 shows the lesser inhibition of molar ratio 1 and 2 than amoxicillin, increase at composition 3 (molar fraction: 3:7) and increase once again at composition 5 (molar fraction: 5:5). Different from AC-PM data, the inhibition diameter of AC-H since molar ratio 3:7 until 9:1 tended to show the higher potency compare to amoxicillin alone.
These phenomena occur in the reason as follows: after heating, hydrogen bonding between amoxicillin tri-hydrate with potassium clavulanate was formed and modifies the amoxicillin tri-hydrate polarity. At the molar ratios 1:9 and 2:8 (composition 1 and 2) the solid phase were dominated by potassium clavulanate which covered the existence of amoxicillin tri-hydrate that caused potency decrease, but began in molar ratio 3:7 the amoxicillin tri-hydrate was starting to dominate the solid solution and improve the antibiotic potency. From the preliminary researches, the interaction has been proven attain an equivalent point at 5:5 (Nugrahani et al., 2007a-d). In this research, the compositions 3:7; 5:5 and 7:3 are indicated as transition points which showed lesser potency than other co-crystal heated but still higher than amoxicillin heated alone. While the molar ratios 4:6; 6:4; 8:2 and 9:1 showed significantly higher than amoxicillin alone. Confirm to preliminary data have been reported, it conclude that interaction of amoxicillin-clavulanate were occur at 3:7-7:3 molar ratio and the interaction might change the antibiotic potency.
This research proved that amoxicillin-clavulanate heated at 50°C at 30 min influence the antibiotic activity against non-beta lactam bacteria Sarcina lutea sp. That is not a high level energy, which could be equal with the energy involved in the milling, granulating, compacting and the storing.
Physical interaction between amoxicillin tri-hydrate and potassium clavulanate improve its antibiotic activity against non-beta lactamase bacterial, Sarcina lutea sp. The result can be used as information to knowing the mechanism of amoxicillin-clavulanate synergism to non-beta lactam bacteria. This research should continue with studying the interaction between the binary system interaction structures with the membrane cell of bacterial.
Authors gratefully thank to Directorate of High Education-Education Ministry Republic Indonesia, which funded the research and to PT Tempo for assistance of material and DSC analysis.