Comparative Activity of Amoxycillin and Ampicillin in an Experimental Bacterial Infection in Mice

Transcription

1 ANTMCROBAL AGNTS AND CHMOTHRAPY, Sept. 1973, p Copyright i 1973 Amerian Soiety for Mirobiology Vol. 4, No. 3 Printed in U.S.A. Comparative Ativity of Amoxyillin and Ampiillin in an xperimental Baterial nfetion in Mie PAMLA A. HUNTR, G. N. ROLNSON, AND DORN A. WTTNG Beeham Researh Laboratories, Chemotherapeuti Researh Centre, Bethworth, Surrey, ngland Reeived for publiation 11 June 1973 Against experimental infetions in mie with strains of sherihia oli and with Proteus mirabilis, amoxyillin was found to be more ative than ampiillin both by the oral and the subutaneous routes of administration. With the bateria used in these experiments, ampiillin and amoxyillin showed the same minimal inhibitory onentrations and, after subutaneous administration, the levels of ampiillin and amoxyillin in the plasma and tissue homogenate were also similar. However, ounts of the number of viable bateria present in the infeted tissue showed that amoxyillin exerted a more rapid and a more marked bateriidal effet than did ampiillin, and this ould be orrelated with the differene in therapeuti effet. When given by mouth, amoxyillin was more ompletely absorbed than ampiillin and gave rise to higher levels of antibioti in the plasma. This may have aounted in part for the differene in therapeuti ativity seen when these peniillins were given by the oral route. However, when appropriate oral dosages of ampiillin and amoxyillin were used so as to ahieve similar levels of antibioti in the plasma and tissue homogenate, amoxyillin was again found to exert a more marked bateriidal effet than did ampiillin, and this was aompanied by greater therapeuti ativity. n experiments in vitro, amoxyillin also showed a more rapid bateriidal effet than did ampiillin against the bateria whih were used to produe the experimental infetions. Amoxyillin is a new semisyntheti peniillin with a spetrum and level of antibaterial ativity in vitro very similar to that of ampiillin (3). When administered by mouth, amoxyillin is better absorbed than ampiillin in man (3) and in experimental animals (1), and the serum onentrations obtained with amoxyillin after oral dosage are onsiderably higher than those of ampiillin. This differene in blood levels might aount, at least in part, for the superior therapeuti ativity whih amoxyillin shows in mie ompared with ampiillin when the drugs are given by mouth (1). When given by subutaneous injetion, however, amoxyillin has also been shown to be more ative than ampiillin against experimental baterial infetions in mie, yet the blood levels of ampiillin and amoxyillin are very similar when the drugs are given by this route (1). n the work reported here the ativity of ampiillin and amoxyillin in vivo has been investigated further by using an intramusular infetion in mie resulting in a lesion whih 285 persists over a period of several days. Numbers of viable bateria present in the infeted tissue were ounted and the levels of antibioti in the blood and tissue homogenate were assayed in order to orrelate these data with the therapeuti effet of these peniillins. MATRALS AND MTHODS Cultures. The bateria used in the animal experiments were maintained as suspensions frozen in liquid nitrogen. A 6-h shaken ulture in nutrient broth was prepared, 1% glyerol was added slowly, and the suspension was dispensed in 2-ml volumes in plasti ampoules. These were plunged diretly into liquid nitrogen and stored for periods of up to 1 year. Very little loss in viability ourred after freezing, and one frozen, no appreiable loss in viability or virulene was found during storage. When required for use, an ampoule was removed and thawed quikly by immersion in a water bath at 45 C. Dilutions were then made in phosphate buffered saline (PBS) ph 7.2, to give the required number of organisms. All ultures were periodially passaged in mie to maintain virulene. The ultures used in the experiments in vitro were maintained on nutrient agar slopes. Downloaded from on September 21, 218 by guest

2 286 HUNTR, ROLNSON, AND WTTNG Mie. Female CS1 mie from Charles River, weight 16 to 2 g, were used throughout. Antibiotis. Commerial preparations of ampiillin trihydrate (Penbritin, Beeham Researh Laboratories, 815,ug/mg anhydrous free aid) and amoxyillin (Amoxil, Benard, 85 ;sg/mg anhydrous free aid) were used. Dosages were alulated on the basis of pure anhydrous free aid peniillin. Solutions for subutaneous administration were prepared by dissolving the free aids of ampiillin or amoxyillin in 2 M sodium arbonate-sodium biarbonate buffer, ph 9.8. During solution of the peniillin the ph fell to approximately 8.5, and this was then adjusted to 7.2 with 1N HCl. Preliminary experiments showed negligible loss of ativity of ampiillin or amoxyillin after 3 min at ph 9.3. Suspensions of ampiillin or amoxyillin for oral administration were prepared by suspending the peniillin free aids, together with a little aaia, in distilled water. Solutions of ampiillin or amoxyillin for in vitro experiments were prepared by dissolving the peniillins in 1 M phosphate buffer, ph 8., and then diluting with PBS to give the desired range of onentrations. Thigh lesion test. Mie were infeted intramusularly in the right hind leg with.2 ml of a baterial suspension. The inoulation was made into the musles of the bak of the thigh as desribed by Selbie and O'Grady (2). The thigh diameters were measured by using alipers 24 h after inoulation and then daily over 4 to 6 days. A group of 1 noninfeted mie was always used as ontrol. Antibiotis were administered orally or subutaneously to infeted mie in groups of 1 and the perentage protetion from thigh enlargement was alulated as follows: % protetion = 1 x mean daily thigh enlargment of infeted ontrols - mean daily thigh enlargement of test group/mean daily thigh enlargement of infeted ontrols. All of the thigh diameter measurements in eah experiment were made by the same person, and to avoid bias the animals were presented for measurement at random. With sherihia oli strain 9, injetion of 17 or more viable bateria per thigh was required to produe a onsistent and large thigh lesion (Table 1, Fig. 1). With this infetion the thigh diameter inreased from the normal value of approximately 3.7 mm to TABL 1. Thigh lesion in mie in relation to size of inoulum using. oli strain 9a No. of No. of No. of Mean viable viable Reov- viable bateria bateria ery (%) bateria diam (mm) thigh reovered after 24 h after 24 h 1.1 X x x 1' x 1' 6.5 x x 1' x x x 1' x x x 1' x x x x x x 1' x x 1' x ~~~~~~~3.7' a Results are the mean of three animals. b Uninfeted animals..c 1 ANTMCROB. AG. CHMOTHR. nou urm 11.6 x 1 1. x x x x x x days FG. 1. Thigh lesions in mie in relation to the size of inoulum using. oli strain 9. The numbers of viable bateria injeted per thigh are shown together with the preentage mortality at 24-h intervals. Thigh diameter measurements are the mean of groups of 1 mie. over 7 mm by 24 h, and the number of viable bateria per thigh inreased to over 19 in the same period. Mortality after 24 h varied from to 3%, and approximately 5% mortality ourred over a period of several days. noulation with 4 x 16 to 8 x 1. viable bateria per thigh resulted in a smaller and variable thigh enlargement though baterial numbers inreased onsiderably over 24 h. With an infetion of 2 x 16 viable bateria or less per thigh, baterial numbers remained fairly onstant or fell slightly over 24 h, and no thigh lesion developed. With. oli strain 8 and Proteus mirabilis strain 13, 5 x 11 to 1' viable bateria were required to be injeted per thigh in order to produe a large and onsistent thigh lesion. Bateriidal ativity in vivo. Groups of 3 to 4 mie were infeted intramusularly as desribed for the thigh lesion test, and ampiillin or amoxyillin was administered orally or by subutaneous injetion. A similar group of untreated mie was used as ontrol. Animals were sarified at intervals and the infeted thighs were homogenized as desribed below. The numbers of viable bateria present in the homogenate were ounted, and in ertain experiments the levels of antibioti were also assayed in the homogenate and in blood samples taken from the same animal. Counts of viable bateria in infeted mouse thigh. Mie were killed by disloation of the nek. The outer skin, foot, and lower leg were removed from the infeted thigh whih was then weighed, hopped oarsely, and homogenized for 1 s at room temperature in 5 ml of PBS by using an Ultra-Turrax homogenizer (Janke and Kunkel KG). This homogenization proedure did not bring about any detetable loss of viability in broth suspensions of the bateria used. Tenfold dilutions of the thigh homogenate were prepared in PBS, and.2-ml drops were plaed on Downloaded from on September 21, 218 by guest

3 VOL. 4, 1973 dried nutrient agar petri dishes. A minimum of eight drops was used per dilution, and numbers of olonies were ounted after overnight inubation at 3 C. Reovery of bateria in thigh homogenate preparations, immediately after infetion, ranged from 5 to 9% of the numbers injeted (Table 1). Antibioti assays. Blood was olleted from the axillary region of mie into heparinized tubes, and the majority of these samples were assayed undiluted against ampiillin or amoxyillin standard solutions prepared in whole horse blood. Preliminary experiments showed that the inhibition zone diameters of ampiillin and amoxyillin standards in horse blood did not differ appreiably from those in mouse blood. Certain blood samples ontaining high onentrations of ampiillin or amoxyillin were diluted as required with PBS and assayed against standards prepared in orresponding dilutions of horse blood in PBS. Samples for assay were plaed in holes in nutrient agar plates (14 by 16 in, approximately 35 by 4 m) seeded with either Baillus subtilis ATCC 6633 or Sarina lutea ATCC B. subtilis was used for the majority of the samples, but those antiipated to be of low poteny were assayed using Sarina. Plasma levels were alulated from the assays of whole blood by allowing for a paked red ell volume of 4%. Tissue homogenates were diluted as required with PBS and assayed against standards similarly prepared in PBS., prior experiments having shown that the inhibition zone diameters of ampiillin and amoxyillin standard solutions in PBS did not differ signifiantly from those prepared in thigh homogenate. The lowest level of ampiillin or amoxyillin detetable in undiluted blood samples using S. lutea varied from.1 to.3 ug/ml. Beause of the dilution involved in the homogenization of the thighs, the lowest level of antibioti detetable in the tissue was.3 to.5 ug/ml. Determination of minimal inhibitory onentration and bateriidal ativity in vitro. A 1: 1, dilution of an overnight broth ulture of the test organism was prepared in nutrient broth (Oxoid, no. 2), and 9-ml volumes were dispensed in tubes. To eah tube, 1 ml of appropriate solutions of ampiillin or amoxyillin were added to give a range of twofold dilutions. Tubes were inubated at 37 C, and.5 ml was removed at intervals of time and the number of viable bateria were ounted. This was arried out by preparing 1-fold dilutions of the samples in PBS, and.2-ml drops of these dilutions were then plaed on dried nutrient agar petri dishes. A minimum of eight repliate drops was used and olonies were ounted after inubation overnight at 3 C. Turbidity of the tubes was reorded after overnight inubation, and the minimal inhibitory onentration (MC) was expressed as the minimal onentration of antibioti whih prevented visible growth. RSULTS Subutaneous dosage: infetion with. oli strain 9. Results are shown in Fig. 2 for a number of experiments in whih the ativity of AMOXYCLLN AND AMPCLLN ACTVTY N MC amoxyillin 8. * Dose (mg/kg) FG. 2. Ativity of ampiillin and amoxyillin in the thigh lesion test by using. oli strain 9. Antibiotis were administered subutaneously as a single dose at the time of infetion to groups of 1 mie. Points represent the results of individual experiments. ampiillin and amoxyillin was determined in the thigh lesion test using. oli strain 9 as the infeting organism. The antibiotis were administered subutaneously as a single dose at the time of infetion and the therapeuti effet, expressed as perentage protetion from thigh enlargement, is plotted against antibioti dosage. n in vitro tests, no differene ould be found in the MC for ampiillin and amoxyillin with this strain of. oli, the value for both antibiotis being 5 ztg/ml. However, in the thigh lesion test using this strain of. oli amoxyillin was onsiderably more ative than ampiillin when given by subutaneous injetion (Fig. 2). At a dose of 5 mg/kg, amoxyillin resulted in 8% protetion, and a mean value of over 5% protetion was obtained with a dose as low as 2 mg/kg. Ampiillin, however, at a dose of 1 mg/kg was only omparable in ativity with amoxyillin at a dose of 2 mg/kg, and a dose of 2 mg/kg was required with ampiillin to give a similar degree of protetion to that ahieved with amoxyillin at 5 mg/kg. Results are shown in Fig. 3 for the plasma levels obtained in mie infeted with. oli strain 9 after single subutaneous dosage of ampiillin or amoxyillin at 25, 5, and 1 mg/kg. t will be seen that the plasma levels of the two antibiotis were very similar, partiularly at a dosage of 25 and 5 mg/kg. At 1 mg/kg the levels of amoxyillin appeared to be slightly higher than those of ampiillin, but this differene was not great. At no time was the plasma level of amoxyillin at a dose of 5 Downloaded from on September 21, 218 by guest

4 288 HUNTR, ROLNSON, AND WTTNG 1 1 CD O.1 * 1 mg/kg n ontrast, the same dosage of ampiillin o * 5 mg/kg resulted in only a slight fall in the viable ount 1 A A 25 mg/kg h after dosage, and thereafter the number of,amoxyillin viable bateria inreased at the same rate as seen -\---- in the untreated ontrol animals. The ampiillin plasma onentrations of ampiillin and amoxyillin in the animals in this experiment were \< \> 9 very similar, but at 24 h a marked differene in "z\ X\ therapeuti effet was seen. n the amoxyillin- '\Q "\\ treated animals a slight thigh enlargement was present after 24 h, orresponding to 78% prote- \\\\ tion, whereas in the ampiillin group a muh 9\\ s\, larger thigh size was present, orresponding to \\ 9 s only 46% protetion. This differene in thera- N peuti ativity was maintained, the protetion over the period of 1 to 4 days being 82 and 4%, respetively. Figure 5B shows the results obtained by using,, a dose of 1 mg/kg. At this dosage both antibiotis brought about a more marked and a minutes more sustained bateriidal effet than was obtained with a dose of 5 mg/kg. n the FG. 3. Plasma levelsofampiillin and amoxyillin amoxyillin group, the viable ount ontinued in mie infeted ir Antibiotis were dtramusularly withan olistrain 9 to fall over the first 2 h, then remained rela- subutaneously at the.. ' aidministered time of infetiont. Results are the mean of four tively onstant until 4 h, and thereafter inexperiments invol!ving 1 to 23 animals at eah time reased slowly to give a ount at 24 h only interval. slightly higher than the number present at the time of infetion. n ontrast, in the ampiillinmg/kg as high as that obtained with ampiillin treated group, the viable ount fell over the first at 1 mg/kg, whereas the results of the thigh lesion tests in Fig. 2 show amoxyillin to be onsiderably more ative than ampiillin at these respetive doses. Results are shown in Fig. 4 for the antibioti levels present in the homogenized thigh of mie infeted with. oli strain 9 after subutaneous dosage of ampiillin or amoxyillin. t will be seen that these tissue levels of ampiillin and amoxyillin were very similar indeed, partiularly when the antibiotis were given at a dose of - 1 mg/kg. At a dose of 5 mg/kg, from 45 min onwards, the levels of amoxyillin were slightly higher than those of ampiillin, but the differene was less than a fator of two. As in the ase of the plasma levels, the homogenate levels obtained with amoxyillin at 5 mg/kg were not as high as those obtained wtih ampiillin at 1 mg/kg, whereas at these dosages amoxyillin was onsiderably more ative therapeutially. The numbers of viable bateria present in the thigh of mie infeted with. oli strain 9 after subutaneous dosage with either ampiillin or amoxyillin are shown in Fig. 5. Over the first hour after a single dose of amoxyillin at 5 mg/kg, a marked fall ourred in the number of viable bateria present in the thigh (Fig. 5A). Between 1 and 2.5 h the viable ount remained relatively onstant and then inreased slowly. as C 'U C) 1 ANTMCROB. AG. CHMOTHR minutes FG. 4. Levels of ampiillin and amoxyillin in thigh homogenates of mie infeted intramusularly with. oli strain 9. Antibiotis were administered subutaneously at the time of infetion. Results are the mean of four experiments involving 1 to 23 animals at eah time interval. Downloaded from on September 21, 218 by guest

5 VOL. 4, 1973 AMOXYCLLN AND AMPCLLN ACTVTY N MC 289 A 1 r B C 1 2 C Q 1. A. a C o 1 r O.1 L io 9r h., Q 19 r 1~17._ o~ 16 lo L S hr lolor 191- o 18.a o = m 24hr v 2 C -> 'a 2 ' 1 Downloaded from -a -U 17 o 16 Os 14L, / { 24 hr O C ;12 FG. 5. Thigh enlargement, plasma levels, and numbers of viable bateria present in the thigh of mie infeted with. oli strain 9 after subutaneous dosage with ampiillin or amoxyillin. Antibiotis were administered at the time of infetion at the dosage indiated. Viable ounts were made using groups of three mie; the lines show the geometri mean and the points indiate the values for the individual animals. Thigh enlargements are the mean values at 24 h for the group of animals as a whole. Plasma onentration:, ampiillin;, amoxyillin. Viable bateria per thigh: A, ampiillin; A, amoxyillin. A, Dosage at 5 mg/kg; B, dosage at 1 mg/kg; C, dosage at 2 mg/kg. hour, though not so markedly as in the amoxyillin group, and then remained onstant up to 2 h, after whih time the baterial numbers inreased to give a ount at 24 h approahing that in the untreated ontrol animals. The differene in bateriidal effet of ampiillin and amoxyillin was again aompanied by a marked differene in the thigh enlargement. n the amoxyillin group, after 24 h only a very slight thigh lesion ourred, orresponding to 89% protetion, whereas in the ampiillin group a onsiderable enlargement ourred, orresponding to only 48% protetion. Over the period of 1 to 4 days the perentage protetion was 93 and 4% for amoxyillin and ampiillin, respetively. a i C.C -S.- on September 21, 218 by guest

6 29 HUNTR, ROLNSON, AND WTTNG nrease in antibioti dosage to 2 mg/kg (Fig. 5C) resulted in almost omplete prevention of thigh enlargement in the amoxyillintreated group and only a slight enlargement in the ampiillin group. The perentage protetion at 24 h was 92 and 74% for amoxyillin and ampiillin, respetively, and over the period of 1 to 4 days the values were 93 and 82%. A marked bateriidal effet also ourred in both groups, but it will be seen that the rate and the extent of this bateriidal ation was more marked with amoxyillin than with ampiillin. This differene in bateriidal effet was apparent not only over the first few hours after dosage but also 24 h later, when the mean viable ount in the amoxyillin-treated group was still below that present at the time of infetion, whereas in the ampiillin group the mean viable ount at 24 h had risen to nearly 19 bateria per thigh. n the experiments shown in Fig. 2 and 5, the antibiotis were administered as a single dose at the time of infetion. Results of thigh lesion tests involving repeated subutaneous dosage of ampiillin or amoxyillin are shown in Table 2.. oli strain 9 was used as the infeting organism and the antibiotis were administered at the time of infetion and again at 2 and 4 h after infetion. t will be seen that amoxyillin again showed greater therapeuti ativity than did ampiillin. nfetion with. oli strain 8. The MC of ampiillin and amoxyillin for. oli strain 8 was found to be 2.5 sg/ml, and repeated serial dilution tests failed to indiate any differene in sensitivity to these two antibiotis. The results of a thigh lesion experiment arried out using this strain of. oli are shown in Fig. 6. t will be seen that a single subutaneous dose of TABL 2. Ativity of ampiillin and amoxyillin by repeated subutaneous dosage in the thigh lesion test using. oli strain 9 Dosage (mg/kg) xpt Ampiillin Amoxyillin Mean aantibiotis were administered to groups of 1 mie at the time of infetion and at 2 and 4 h later at the dosage indiated. Results indiate perentage protetion. 9 ontrol 8 7 ampiillin 2 mg/kg 5) a) 6 / n o/& ampiillin 5 mg/kg s 5 / /O 4 ANTMCROB. AG. CHMOTHR. amoxyillin 2 mg/kg ampiillin 1 mg/kg days FG. 6. Therapeuti effet of ampiillin and amoxyillin in the thigh lesion test using. oli strain 8. Antibiotis were administered subutaneously at the time of infetion. Thigh diameter is shown together with perentage mortality at 24-h intervals. amoxyillin at 2 mg/kg ompletely prevented the formation of any thigh enlargement, whereas with ampiillin, a dose of 1 mg/kg was required to bring about a omparable effet. The numbers of viable ells of. oli strain 8 in the thigh of mie after subutaneous dosage with ampiillin or amoxyillin at 5 mg/kg are shown in Fig. 7. As in the experiments with. oli strain 9, amoxyillin exerted a more rapid and a more marked bateriidal effet than did ampiillin, and this was aompanied by a orresponding differene in therapeuti effet in terms of thigh enlargement. nfetion with P. mirabilis strain 13. The ativity of ampiillin and amoxyillin in the thigh lesion test using a strain of P. mirabilis is shown in Table 3. With this strain of P. mirabilis, a heavy infetion was required to produe a onsistent thigh lesion, and single dosage with ampiillin or amoxyillin at the time of infetion gave rise to variable results. More reproduible results were obtained with repeated dosage, and in the experiment shown in Table 3 the antibiotis were administered subutaneously at the time of infetion and again at 2 and 4 h after infetion. As in the experiments with. oli, amoxyillin was more ative than ampiillin in terms of prevention of thigh enlargement. Mortality was also lower in the amoxyillintreated animals than in the ampiillin group. n in vitro tests with this strain of P. mirabilis, the Downloaded from on September 21, 218 by guest

7 mg/kg. With ampiillin, on the other hand, a dose of 5 or 1 mg/kg was without effet on the thigh enlargement and a dose of 2 mg/kg showed only a slight therapeuti effet, ompa- -~~~~~~~~~ ontrol~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ VOL. 4, 1973 AMOXYCLLN AND AMPCLLN ACTVTY N MC 291 Dosage (mg/kg) io9. h-il looro C t ~~~~~~~~19 * an _ F. 6~~~~~ a *U;s~~~~~~~iiii * $~~~~~~~~-ii.. e 4 C mpiillin *7 X Z.Q 2 n o 'a t1 o 15 amo.yillin / 2. h r, hr FG. 8. Thigh enlargement and number of viable FG. 7. Thigh enlargement and number of viable bateria present in the thighs of mie infeted with bateria present in the thighs of mie infeted with. P. mirabilis strain 13, after subutaneous dosage with oli strain 8 after subutaneous dosage with ampiillin or amoxyillin. Antibiotis were administered at a tered at a dose of 1 mg/kg at the time of infetion and ampiillin or amoxyillin. Antibiotis were adminis- dose of 5 mg/kg at the time of infetion. Viable also at 2 and 4 h after infetion. Viable ounts were ounts were made using groups of three mie; the lines made by using groups of three mie; the lines show the show the geometri mean and the points indiate the geometri mean and the points indiate the values for values for the individual animals. Thigh enlargements the individual animals. Thigh enlargements are the are the mean values at 24 h for the group of animals as mean values at 24 h for the group of animals as a a whole. whole. TABL 3. Thigh lesion test in mie using P. mirabilis strain 13a Dosage (mg/kg) Desription Ampiillin Amoxyillin Protetion (%) No. of mie i 7 dead after 48 h a Ampiillin or amoxyillin was administered subutaneously to groups of 1 mie at time of infetion and at 2 and 4 h after infetion at the dosage indiated. MC of both ampiillin and amoxyillin was 1 pg/ml. Viable ounts of P. mirabilis in the infeted thigh of mie after subutaneous dosage with ampiillin or amoxyillin are shown in Fig. 8. The antibiotis were administered at a dose of 1 mg/kg at the time of infetion and at 2 and 4 h subsequently. t will be seen that the numbers of viable bateria were lower in the amoxyillin-treated group, and this was partiularly marked at 24 h. At this time the ount in the amoxyillin-treated animals had remained low, but in the ampiillin group onsiderable baterial multipliation had ourred and this was aompanied by a marked differene in the thigh enlargement. Oral dosage: infetion with. oli strain 9. Results in Fig. 9 show the ativity of ampiillin and amoxyillin in the thigh lesion test C 1 amoxyillin 8 ( 6 C\ ) L 4 2 ampiillin Dose (mg/kg) FG. 9. Ativity of ampiillin and amoxyillin in the thigh lesion test using. oli strain 9. Antibiotis were administered orally at the time of infetion to groups of 1 mie. Points represent the results of individual experiments. when the antibiotis were administered by mouth.. oli strain 9 was used as the infeting organism and the antibiotis were administered at the time of infetion. t will be seen that amoxyillin was onsiderably more ative than ampiillin. At a dose of 5 mg/kg, amoxyillin resulted in 8% protetion from thigh enlargement and showed a slight but measurable therapeuti effet even at a dose as low as 2 Downloaded from on September 21, 218 by guest

8 292 HUNTR, ROLNSON, AND WTTNG rable with that obtained with amoxyillin at a dose of 2 mg/kg. When given by mouth, amoxyillin was better absorbed than ampiillin, and the plasma levels obtained in mie after single oral dosage are shown in Fig. 1. t will be seen that ampiillin and amoxyillin showed a similar plasma halflife, but the levels obtained with amoxyillin were onsiderably higher than those obtained with ampiillin. However, although amoxyillin was ertainly better absorbed than ampiillin, the differene in the plasma levels obtained was not as great as the differene in therapeuti effet. For example, the plasma levels obtained with amoxyillin at a dose of 5 mg/kg were similar to those obtained with ampiillin at 2 mg/kg, but at these respetive dosages amoxyillin was onsiderably more effetive therapeutially (Fig. 9). Results are shown in Fig. 11 and 12 in whih the effet of oral dosage of ampiillin at 25 mg/kg was ompared with oral dosage of amoxyillin at 5 mg/kg. These dosages were hosen in an attempt to ahieve similar levels of antibioti in the blood. t will be seen from Fig. 12, 5 lo 5 *" t LO -5 FG. 1. Plasma levels of ampiillin and amoxyillin after oral dosage in mie. Antibiotis were administered at the dosages indiated (mg/kg). Results are #& , 1A #- AA -J,a --#n ;;- tne moean of U tw KU aimats at ean time intervat.,lol 19 lo _ ~ ~ ~ ~ ~ 2 however, that the levels of amoxyillin both in the plasma and in homogenized tissue were slightly lower than those of ampiillin. Nevertheless, it will be seen from Fig. 11 that a more rapid bateriidal effet ourred in the amoxyillin-treated animals and a orresponding differene was also seen in the therapeuti effet. n the amoxyillin group almost omplete protetion from thigh enlargement was ahieved, whereas in the ampiillin-treated animals after 24 h a onsiderable thigh lesion ourred, orresponding to only 2% protetion. Bateriidal ativity of ampiillin and amoxyillin in vitro. The bateriidal effet of ampiillin and amoxyillin in vitro against. oli strain 8 is shown in Fig. 13. t will be seen that amoxyillin resulted in a more rapid fall in the numbers of viable ells than did ampiillin, the differene in the ount at 1 h with 5 and 1 gg/ml being approximately 1-fold. Similar results were obtained with. oli strain 9. With P. mirabilis strain 13, amoxyillin was also found to exert a more rapid bateriidal effet, though with this organism the differene between the two antibiotis was not as marked as it was with the strains of. oli. DSCUSSON The experiments reported here using an intramusular infetion in mie show amoxyillin to be more ative than ampiillin both by oral and by subutaneous administration. n these experiments, the infeting organisms used showed the same sensitivity to ampiillin and amoxyillin in tests arried out to determine the 3. ANTMCROB. AG. CHMOTHR. Downloaded from on September 21, 218 by guest O ts 7 f f 24 h r FG. 11. Thigh enlargement and number of viable bateria present in the thighs of mie infeted with. oli strain 9 after oral dosage of ampiillin at 25 mg/kg or amoxyillin at 5 mg/kg. Antibiotis were administered at the time of infetion. Mean viable ounts and range of values are shown for groups of 1 mie. Thigh enlargements are the mean values at 24 h for groups of 2 mie. 5 hours FG. 12. Ampiillin and amoxyillin levels in plasma and homogenized thighs of the same mie used in the experiment shown in Fig. 11. Ampiillin was administered orally at a dose of 25 mg/kg and amoxyillin at 5 mg/kg. Values shown are the mean of 1 animals.

9 VOL. 4, 1973 u a) g/m 1 2 FG. 13. Bateriidal a amoxyillin in vitro usin ounts were arried out ontaining antibioti at thi bols:, ontrol; A, ampi AMOXYCLLN AND AMPCLLN ACTVTY N MC Lg /m 1,g/mi tissue, it would appear that the differene in therapeuti ativity between these two peniillins is related to the bateriidal ation of the drugs in vivo, that of amoxyillin being more rapid and more marked than that of ampiillin against the organisms used in these experiments. When administered by the oral route, amoxyillin is absorbed more ompletely than ampiillin, resulting in substantially higher levels of antibioti in the blood, and this may aount in part for the greater therapeuti effet obtained with amoxyillin when the antibiotis are given by mouth. However, when appropriate pp... oral dosages of ampiillin and amoxyillin were used whih resulted in similar levels of antibihours oti in the blood and tissue, amoxyillin was again found to be therapeutially more effetive itivity of ampiillin and than ampiillin. These results after oral dosage gt 37Cli strain 8. Viablh are in agreement with those obtained with e onentration shown. Sym- subutaneous administration and show that illin; A, amoxyillin. despite similar levels of antibioti in the body, amoxyillin exerted a more rapid bateriidal MC. The differene: in ativity observed in effet than did ampiillin, and it is suggested vivo, therefore, annot be explained on the basis that this differene in bateriidal effet is of a differene in senssitivity as measured by responsible for the superior therapeuti ativity using onventional seriial dilution tests. Simi- obtained. larly, the superior theralpeuti ativity of amox- A differene in the rate of bateriidal ation yillin ompared with aimpiillin in these exper- between ampiillin and amoxyillin ould also iments annot be aotanted for on the basis of be seen in experiments arried out in vitro. higher levels of antibi ti in the animal body. Whether this differene in vitro aounts ennistration, the peak lev- tirely for the superior therapeuti ativity of By subutaneous admi: els of ampiillin and famoxyillin both in the amoxyillin in vivo is not ertain. blood and in the tissu e homogenate were very The experiments reported here all involve similar indeed, and altkiough the levels of amox- infetion with. oli or P. mirabilis. Work is in yillin tended to fall r^ather more slowly than progress using other bateria as infeting orga- differene was slight nisms. Preliminary experiments with a strain of those of ampiillin, thlis and would seem unlil kely to aount for the Staphyloous aureus in the thigh lesion test differene in therapeutii ativity. For example, indiate amoxyillin to be slightly more ative although the plasma arid tissue levels obtained than ampiillin when administered subutanedose of 5 mgokg were ously, though the differene is not as marked as with amoxyillin at a slightly higher than tldose obtained with the it is with. oli or P. mirabilis. same dose of ampiillin1, the levels were learly ACKNOWLDGMNT lower than those obtairned with ampiillin at a The authors thank Linda Mizen for dose of 1 mg/kg, but arrying out ertain of in the thigh lesion tests the blood level determinations in mie. the therapeuti effet off ampiillin at 1 mg/kg was onsiderably less t;han that of amoxyillin LTRATUR CTD at a dose of 5 mg/kg. Furthermore, in experi- 1. Ared, P., P. A. Hunter, L. Mizen, and G. N. Rolinson. ments in whih diret omparisons were made a-amino-p-hydroxybenzylpeniillin (BRL 2333), of subutaneous dosage> of amoxyillin at 25 or a new broad-spetrum semisyntheti peniillin: in vivo evaluation. Antimirob. 5 mg/kg with ampiilllin Ag. Chemother. 197, p undosage at 5 or mg/kg, greater therapeiuti ativity and a more 2. Selbie, F. R., and F. O'Grady A measurable marked bateriidal effet was seen with amox- tuberulosis lesion in the thigh of the mouse. Brit. J. yillin despite the lowe r dosage. xp. Pathol. From the results of t] *r dosage. 35: Sutherland, R.,. A. P. Croydon, and G. N. Rolinson. he ounts of the number Amoxyillin: a new semi-syntheti peniillin. Brit. of viable organisms p)resent in the infeted Med. J. 3: Downloaded from on September 21, 218 by guest