THE JOURNAL OF UROLOGY
Vol. 87, No. 6 June 1962 Copyright © 1962 by The Williams & Wilkins Co. Printed in U.S.A.
EFFECT OF URINE pH ON THE ANTIBACTERIAL ACTIVITY OF ANTIBIOTICS AND CHEMOTHERAPEUTIC AGENTS THOMAS W. MOU From the Department of Preventive Medicine, State University of New York, Upstate Medical Center, Syracuse, N. Y.
The response of acute urinary tract infections to antibiotics generally is good unless there is prolonged obstruction of urine outflow. On the other hand, the management of chronic urinary tract infections frequently is difficult and unsatisfactory, notwithstanding the therapeutic agents that have become available during the past twenty years or more. Antibiotics currently in use ofttimes are capable of rapidly eliminating susceptible bacteria. Usually, large amounts of the antibiotics are excreted in the urine and there is a wide margin between therapeutic and systemically toxic doses. In stubborn infections (usually obstruction is present), there may be therapeutic complications with replacement of sensitive strains by resistant ones. Thus, it would appear to be of some purpose to devise a means for potentiating the activity of the antibiotic in such situations. With this in mind, it was considered worthwhile to re-investigate the possible potentiation of antibacterial activity in urine by suitable adjustment of its pH. The data to be presented are derived from laboratory and clinical studies of this relationship. MATERIALS AND METHODS
The bacteria used in the laboratory studies were obtained by culturing patients with urinary tract infections. Stock cultures were maintained in Difeo brain heart infusion broth; before each experiment, they were subcultured for two consecutive days in sterile, pooled urine adjusted to pH 7. The urine used for this purpose was a pool collected during an 8-hour period from several healthy, young, male adults. The pH was adjusted to either 5 or 7 and then rendered bacteria-free by passage through a Seitz filter. Previous studies1 had demonstrated that such Accepted for publication August 28, 19Gl. Supported by Grant E-1870, National Institutes of Health, Bethesda, Md. and additional funds from the Eaton Laboratories, Inc., Norwich, N. Y. 1 Mou, T. W., Colburn, J. and Kass, E. H.: Unpublished observations. 978
filtration did not affect the urine for these purposes. Quantitative tube dilution antibiotic sensitivities of the bacterial strains were determined by methods previously reported by "\Vaisbren and associates 2 and Rammelkamp and JV[acon. 3 • 4 The decision as to which antibiotics a specific organism ,ms to be tested against was based on whether the agent could be excreted in the urine in the range of the organism's sensitivity. 5 The experimental procedure* consisted of diluting, in physiological saline (PSS), a 24-hour urine culture of the bacterial strain under test to concentrations of 10-2 , 10-4 , 10-s, and 10-s. Inocula of 0.1 ml. of each dilution then \rere added to 5.0 ml. of sterile urine at pH 5 and 7 containing various concentrations of antibiotic. Growth controls at each pH were incubated with each experiment. Following incubation at 37°C for 6 and 24 hours, each tube was subcultured in nutrient agar pour plates, using dilutions of 10-i, 10-3, 10- 5 , and 10-7 in PSS to determine the residual number of bacteria. The pour plates were incubated at 37°C for 24 hours and the number of colonies were counted. The clinical studies were performed on patients with chronic urinary tract infections at the Syracuse Veterans Administration and the Van Duyn ?-.Iemorial Hospitals. Prior to administration of any chemotherapeutic agents, cultures 2 Waisbren, B. A., Carr, C. and Dunnette, J.: The tube dilution method of determ.ining bacterial sensitivity to antibiotics. Am. J. Clin. Path., 21: 884-891, 1951. 3 Rammelkamp, C. H.: A method for determining the concentration of penicillin in body fluids and exudates. Proc. Soc. Exp. Biol., 51: 95-97, 1942. 4 Rammelkarnp, C. H. and Macon, T.: Resistance of S. aureus to the action of penicillin. Proc. Soc. Exp. Biol., 51: 38G-389, 1942. 5 Kass, E. H.: Chemotherapeutic and antibiotic drugs in the management of infections of the urinary tract. Arn. J. Med., 18: 764-781, 1955. * Early phases of the study were performed in the Rochester Health Bureau Laboratories, Rochester, X. Y.
EFFEC'l' OF URINE pH ON ANTIBACTERIAL ACTIVITY
were performed to ascertain the degree and significance of the infection, the causativf'. organism, and its drug sensitivity spectrum. After a given period of treatment, generally four to seven three to five days without treatment were permitted before starting the next trial. This was in an opposite pH direction so that the patient could sen·c as his own controL A few sl10rt-term studies ,rnre attempted. These patients received either amn10nium chloride, 2.0 gm. every 6 hours or sodium bicarbonate, 1.0 gm. with 250 mg. 2 acetylamino-l,3,4 thiadiazole 5-sulfonarnide (diamox) on the same time schedule imn1cdiatcly after obtaining a control nrine specimen. Following the second dose of acid or alkali, urine specimens were procured at the second, third and fourth hours in order to determine the effectiveness of pH change alone. At thC' time of the third close of acid or alkali, a single dose of the antibiotic under study was administered. Urine spC:'einwns then were obtained for baeterial counts 2, :3 and 4 hours later, coinciding with maximal C'xcretion of the drug, Each patient, after t\\·o or three clays without medication, served a~ his own control and received the rnnverse agent for pH change together ,Yith the same antibiotic and dosage schedule he had rC'ccivccl previously. Every attempt was made to keep other factors in the patient's eare constant throughout the period of observation. 1\cll urine specimens were refrigerated immediately upon collection until studied in the lab-· oratory. The pH of each specimen \\'as determined with the Beckman pH meter. Bacterial enumeration ,rns rwrformed by both platC:' count and tube dilution encl point techniques. 6 RESULTS
Laboratory data on the effect of urine pH on bacterial growth inhibition by the rnrious antibiotics ,vill be limited to representative illustrations. The inhibition levPl illustrated is based, in most instances, on the findings from at least four different bacterial inocula 10~4, 10-· 5 , and 10-s dilutions of a 24-hour culture) and a minimum of three concentratiorrn of antibiotic in a mnge therapeutically attaim1ble in human beings. Three typical studies are illustrated in figure 1 and demonstrate the manner in whid1 the data 6 Mou, T. vV. and Feldman, H. A.: The enumeration and preservation of bacteria in urine. Am. J. Clin. Path., 35: 572-.575, 1961.
P. aeruginoso ond Polymyxin 8 1010
,, @ ~ ~
w ~ ~
' ' '\ CONTROL '' '
;;: S. fecal is and Chloramphenicol
10' 10' 10'
Dilution of Initial Bacterial lnoculum (Logs)
FOUR CONCENTRATIONS OF INITIAL !NOCULUM
THREE CONCENTRATIONS OF ANTIBIOTIC
Fm. l. _Growth_inhibition of three organisms by three antibiotics m unne at pH 5 and pH 7,
in figures 2 and 3 were obtained. Tlw cmTcs are based on the maximum values observed with the four dilutions of the control and the antibiotic growtb inhibition patterns. Since the several antibiotic concentrations usually yielded comparable degrees of inhibition, a single curve waf, drawn for the control values and another for the antibiotic concentrations. \Vhen polymyxin B was incorporated into the urine substrate, the inhibition of growth of Pseuclomonas aeruginosa was limited to about two logs, regardless of the inoculum size. Growth could b2 inhibited markedly by merely shifting the pH. The second pattern shows the growtb curves of Escherichia coli in the presence of streptomycin. This was ineffective at pH 5. Con 7 was versely, growth in the control tube~ at from one to two logs greater than at pH 5 but was markedly inhibited by the drug. The third pattern, illustrated with Streptococcus fae.c,alis,
THOMAS W. MOU
NO INHIBITION NO INHIBITION
DIFFERENCE IN.INHIBITORY EFFECTS AT pH 5 AND pH 7 BY LOGS
Fm. 2. Effect of urine pH on bacterial growth inhibition by various antibiotics
pH 5 6 HOURS
DIFFERENCE IN INHIBITORY EFFECTS AT pH 5 AND pH 7 BY LOGS
Fm. 3. Effect of urine pH on bacterial growth inhibition by various antibiotics
demonstrates that chloramphenicol is relatively effective at pH 7 but to no greater a degree than is an acid pH alone. The additive effect of the acid pH on the efficiency of the antibiotic results in extremely low bacterial counts. Figures 2 and 3 summarize our findings with the various organisms that were studied. Staphylococcus aureus, S. faecalis, and Ps. aeruginosa
were almost uniformly inhibited to a marked degree when the urine substrate was acid. In most instances, this appeared to be a synergistic or additive effect of antibiotic and acid, rather than the effect of acid alone. Although streptomycin, ordinarily, was more effective in alkali, if the inhibitory effect of acid was especially marked, variable effects might result. This is
EFFECT OF URINE pH ON ANTIBACTERIAL ACTIVITY TABLE
1. Bacterial growth inhibition in urine by nitrofurantoin at pH 5.0 and 7.0
0 1.0 2.0 4.0 8.0
10 10 10 10 6
--- --0 0.2 1.0 4.0 8.0
illustrated by Ps. aeruginosa when streptomycin was in the medium. Aerobacter aerogenes and E. coli strains were inhibited to greater degrees in alkaline substrates when streptomycin and polymyxin B were used. With the "broad spectrum" antibiotics (chloramphenicol, oxytetracycline, and tetracycline) growth of these two organisms was usually slightly less in acid medium. The results of studies with nitrofurantoin and an experimental analogue, nifurethazone, are summarized in tables 1 and 2. Nitrofurantoin, uniformly, was more effective against all organisms when the pH of the medium was acid. Nifurethazone (table 2) yielded opposite findings, with increased growth inhibition at an alkaline pH with A. aerogenes, E. coli, and Ps. aeruginosa, but not with Proteus vulgaris. (The acid pH may have inhibited the Proteus growth inordinately, resulting in a misleading pattern.) The growth patterns of large and small bacterial inocula were studied in acid and alkaline urine. It is of interest that small inocula (lo-s organisms per ml.) could be either partially or completely inhibited at either pH, but especially at pH 5. If the organism was able to survive at all for 6 hours, then by 24 hours it would attain the high bacterial density of the larger inocula at 6 and 24 hours. With both small and large inocula, growth at pH 5.0 was consistently two to four logs less than that at pH 7.5, in either broth or urine. The administration of acidifying and alkalinizing agents was investigated in patients. Acid and alkaline ash diets were tried in a small number but failed to contribute sufficiently to warrant
5 5 5 0 0
8 8 8 7 0
8 0 0 0
5 7 --- ---
0 0.05 0.2 1.0 2.0
4 4 4 0 0
7 7 6 3 0
0 1.6 8.0 40.0 80.0
6 2 0 0 0
5 5 6 2 0
* Of nitrofurantoin. t Bacterial count expressed to the nearest log. 2. Bacterial growth inhibition in urine by nifurethazone at pH 5.0 and 7.0 A. aerogenes
Concentration of Nifurethazone
-- - - -- -- -- -- -mg.%
3.8 7.5 15.0 30.0 60.0
8* 5 6 5 5
8 4 3 2 0
8 8 8 8 2
8 8 3 2 0
8 3 3 5 4
8 4 4 4 5
8 8 8 8 2
8 8 8 1 2
* Bacterial count expressed to the nearest log. their further use, either alone or in combination, with acidifying or alkalinizing agents. Sodium acid phosphate was used as an acidifying agent but its effect was minimal, except with large doses. Sodium bicarbonate, in doses comfortably tolerated by the patient, was found to be relatively inefficient as an alkalinizing agent. Methionine, although an excellent acidifying agent, required doses of 12-15 gm. per day and caused gastric distress in some patients. Ammonium chloride proved to be the most satisfactory material for short-term acidification. The combination of 2-acetylamino-1,3 ,4 thiadiazole-5sulfonamide (diamox) and sodium bicarbonate was found to be highly effective for alkalinizing the urine and well accepted by patients. Patient's diets, physical activity, and nursing procedures were not under complete control, but medications were given at specific times and urine collections were made at intervals calcu-
THOMAS W. MOU TABLE
3. In vivo e.ffect of pH on urinary tract infections
Streptomycin Polymyxin B Tetracycline
Coliform Coliform Pseudomonas
Sa El Sa
Pseudomonas Pseudomonas B. Alkaligenes Coliform Proteus Pseudomonas
Dr Kl Ro Ba Sh Me
Coliform Coliform Proteus Pseudomonas
Cl De Fo Dr
Growth Inhibition by -----Alkali Acid --- ---
Growth Inhibilion by
Acid Alkali ------
Coliform Coliform Coliform Coliform Proteus Proteus Pseudomonas Ps. & Coli Ps. & Coli
Sc Ba Ki El I Kl ! Sh Ki Re Le
Coliform Pseudomonas Ps. & coli
Sh Dr Sa
+ + + + + + +
+ + ±
I + * Kynex is Lederle Laboratories trade name for sulfamethoxypyridazine. t Madribon is Roche Laboratories trade name for 2,4 dimethoxy-6-sulfanilamide-1-3-diazine. t Thiosulfil is Ayerst Laboratories trade name for sulfamethizole.
Ps. & coli
lated to coincide with the maximal excretion of the substance under study. Thirty patients were studied completely; that is, the patient was treated for a sufficient period to allow for periods of both acidification and alkalinization during specific antibacterial therapy. An interval of at least 3 days, without medication, was provided between the periods of acidification and alkalinization. As with the in vitro studies, no clear-cut pattern emerged for either each organism or antibacterial agent. A more consistent pattern was observed when the data were analyzed on the basis of the antibacterial agent and representative examples are shown in table 3. Chloramphcnicol was usually most effective when used to treat patients with coliform or Pseudomonas infections and the urine was acid. Streptomycin was more effective when the urine was alkaline but polymyxin B seemed better in acid urine. The tetracycline experience was variable, appearing more effective in acid when the organism was P. vulgaris or Bacterium alkaligenes. Activity against E. coli was greater in alkaline urine, ,vhile in Pseudomonas infections, there was little to choose between acid and alkali.
The sulfonamides had greater effectiveness against Pseudomonas if the urine was acid but against the coliforms when the urine was alkaline. Nitrofurantoin was most effective when the urine was acid and the patient infected with coliforms, Proteus, or a mixed infection of coliform and Pseudomonas strains. Nifurethazone was more effective in an alkaline medium. In terms of the infecting organism, the bacterial counts fell to a greater extent (2 logs or more) with coliform infections when the urine was acid, and the chemotherapeutic agent was chloramphenicol or nitrofurantoin. If the urine was alkaline, the counts were lower when sulfamethoxypyridazine (kynex), nifurethazone, and tetracycline were used. Pseudomonas infections improved to a greater degree when the urine was acid and the drug administered was chloramphenicol, sulfamethoxypyridazine, nitrofurantoin, polymyxin B or sulfamethizole (thiosulfil). Nifurethazone, streptomycin, and possibly tetracycline were more effective in alkaline urine. Acidification of the urine when Proteus was present yielded lowest counts when sulfamethoxypyridazine and nitrofurantoin were employed. Tetracycline yielded equivocal findings,
EFFECT OF URINE pH ON ANTIBACTERIAL ACTIVITY TABLE
4. Selected references to the role of pH in the therapy of urinary tract infections Inhibitory Substance
S. albus S. aureus E.coli
Clark & Helmholz
Mandelic and nine other acids
Mandelic acid and beta-hydroxybutyric acid Lactic acid, irrigation of bladder
Acetic acid, irrigation of bladder
Urine of epileptic patients on ketogenic diets remained sterile in open containers at room temperature. E. coli By fractionating ketogenic urine, beta-hydroxybutyric acid was isolated and found to have antibacteri::11 activity. E. coli Benzoylacetic acid was most bacteriostatic in vitro, but underwent chemical alteration and was not clinically effective. Mandelic acid was effective in vitro and in vivo. E. coli Improved effectiveness was obtained by adding ammonium chloride. P. vulgaris Difficulty in lowering pH of Proteus infections limited usefulness of these drugs. Urea-splitting Bacteria that liberate lactic acid organisms were toxic to putrefactive organisms, and suggested that local bladder irrigation with lactic acid could be used clinically. P. vulgaris Attempts to overcome alkalinity of urea-splitting organisms were made using acetic acid. Various bacteria pH at which bactericidal effect occurred was often lower than attainable in vivo, but bacteriostasis occurred at physiologically attainable acid levels except for A. aerogenes and E. coli.
being perhaps somewhat more effective against Proteus strains and B. alkaligenes, if the urine was acid. With mixed Pseudomonas and coliform infections, acidification of the urine yielded best results when 2 ,4-dimethoxy-6-sulfaniamido-1,3diazine (madribon) and nitrofurantoin were used. DISCUSSION
Treatment of chronic infections of the urinary tract is most efficient if optimal conditions required for the activity of an antibacterial agent are provided. Unless obstruction is relieved, it is likely that
Bacterial growth was less m acid urine.
sensitive bacterial strains will be eliminated and replaced by resistant strains. Theoretically, the elimination of an antibiotic-sensitive organism could be a disservice to patients, since reinfection with a resistant organism may be more difficult to control. If the bacterial count can be kept low, even though the organism is not eliminated entirely, the patient's symptoms may be minimized. Kass7 noted that the lowering of the urine bacterial count by 2-3 logs frequently will reduce complaints such as dysuria and frequency. Reports of the value of pH change in the treat7 Kass, E. H.: Bacteriuria and the diagnosis of infections of the urinary tract. Arch. Int. Med.,
100: 709-714, 1957.
THOMAS W. MOu
ment of urinary tract infections have appeared for more than 60 years. A few, historically pertinent to the present discussion, are listed in table 4. One of the earliest papers by Rostoski, 8 in 1898, noted that bacterial growth was inhibited when the urine was acid. Thirty years later, Clark 9 and Helmholz 10 found that the urine of patients on ketogenic diets remained sterile, and Fuller11 shortly thereafter showed that bdahydroxybutyric acid had antibacterial activity. Rosenheim12 then studied a number of organic acids and found that mandelic acid would acidify the patient's urine and reduce bacterial numbers, while Colby13 showed that adding an urine acidifying agent (ammonium chloride) enhanced the effectiveness of mandelic acid. The problem of lowering the pH of urine infected with Proteus strains was emphasized by Helmholz 14 while Schulte15 noted limited effectiveness of bladder irrigation with lactic acid when urea-splitting organisms were present, and Marshall16 found similar results using acetic acid in treating patients infected by the same organisms. Yeaw17 pointed out that the bacteriostasis occurred at pH levels physiologically attainable in vivo but bactericidal pH values were difficult to attain. The effect of pH on the antibacterial activity of antibiotic and chemotherapeutic agents has been studied by a number of investigators whose findings are summarized in table 5. The relative efficacy of the sulfonamides in the treatment of 8 Rostoski, 0.: Ueber den Bactericiden Einfluss der Aciditat des Harns auf die Cystitiserreger. Dtsch. med. Wschr., 24: 235-236 and 249-252, 1898. 9 Clark, A. L.: E.coli bacilluria under ketogenic treatment. Proc. Staff Meet., Mayo Clin., 6: 605-
10 Helmholz, H.F.: The ketogenic diet in the treatment of pyuria of children with anomalies of the urinary tract. Proc. Staff Meet., Mayo Clin.,
6: 609-616, 1931. 11 Fuller, A. T.: The ketogenic diet, nature of the bactericidal agent. Lancet, 1: 855-856, 1933. 12 Rosenheim, M. L.: Mandelic acid in the treatment of urinary infections. Lancet, 1: 1032-1037,
13 Colby, F. H.: Progress in urology, 1936. New Eng. J. Med., 216: 1075-1078, 1937. 14 Helmholz, H. F. : A comparison of mandelic acid and sulfonilamide as urinary antiseptics. J.A.M.A., 109: 1039-1041, 1937. 15 Schulte, T. L. and Thompson, G. J.: Importance of identifying urea-splitting bacteria: report of clinical experience. J. Urol., 45: 733-743, 1941. 16 Marshall, V. F.: A reconsideration of the treatment of urinary infections. Med. Clin. N. Amer., 34: 525-533, 1950. 17 Yeaw, R. C.: Effect of pH on the growth of bacteria in urine. J. Urol., 44: 699-713, 1940.
acute urinary tract infections lessened some of the interest in urine pH. Also, urine alkalinization was recommended to prevent crystalluria, although in some instances this may have reduced the efficiency of the drugs. For example, Sickler18 found sulfanilamide to be more effective in alkaline urine, especially in coliform infections. Sung and Helmholz19 noted that the E. coli concentration was reduced to a greater extent when the urine was slightly alkaline (pH 7 .2) if sulfathiazole was used, but that the drug was more effective against S. faecalis at pH 5.5. Our clinical experience was similar with greater effectiveness of sulfonamides against coliform infections in alkaline urine. When used to treat Ps. aeruginosa, the sulfonamides appeared to be more effective in our patients in acid, but Pseudomonas strains are unusually sensitive in acid urine. Our results with S. faecalis, when antibiotics were supplied in acidified urine, would lead us to expect that the sulfonamides would act similarly and yield findings in keeping with those of Sung and HelmholzB that were noted above. Abraham and Duthie' 0 showed that penicillin was more active at pH 6.5 than at pH 7 .5, while streptomycin activity was decreased in acid. Only one organism encountered by us, S. faecalis, was susceptible to penicillin, and this was marked at pH 5. Eagle 21 had similar experience with penicillin, but found streptomycin effectiveness to be increased as the pH became more alkaline. We found streptomycin in vitro and in vivo to be more effective in an alkaline substrate except in two instances, with Pseudomonas strains and with S. faecalis. (The growth of the organisms, themselves, in these two instances ,vas probably limited in our testing system by the acidity of the medium.) 18 Sickler, J. R.: A study of the bactericidal activity of sulfanilamide in urine at various levels of hydrogen ion concentration. J. Urol., 43: 906-916,
19 Sung, C. and Helmholz, H.F.: Variation of action of sulfathiazole on E.coli and S. faecalis on changes in the pH of urine. Proc. Staff Meet., Mayo Clin., 19: 577-580, 1944. 20 Abraham, E. P. and Duthie, E. S.: Effect of pH of the medium on activity of streptomycin and penicillin and other chemotherapeutic substances. Lancet, 1:455-459, 1946. 21 Eagle, H.: The effect of the pH of the medium on the antibacterial action of penicillin, streptomycin, chloramphenicol and bacitracin. Antibiot. & Chemother., 2: 563-575, 1952.
EFFECT OF URINE pH ON ANTIBACTERIAL ACTIVITY TABLE
5. Selected references to the effect of pH on the efficacy of some antibacterial agents Year
Sung and Helmholz 19 Abraham and Duthie 20 Eagle 21
S. faecalis, E.coli
Penicillin Streptomycin Penicillin
S. aureus, E. coli
S. aureus, E. coli Various A. aerogenes, E. coli P. aeruginosa, P. vulgaris Various
Finland 22 Murray and Finland 23
Nitrofuratoin Nifurethazone Streptomycin
Finland22 presented clinical and laboratory evidence that alkalinization of the urine greatly enhanced the therapeutic effectiveness of streptomycin. Murray and Finland 23 demonstrated that an increase in the pH of the medium at any given concentration of streptomycin resulted in the inhibition of larger numbers of A. aerogenes, E. coli, Ps. aeruginosa, and P. vulgaris. Kass 24 showed that nitrofurantoin activity was increased in acid urine, while nifurethazone and streptomycin activity was better in an alkaline urine. Harris 25 raised the important question of whether the effect of streptomycin was related to the tissue or the urine pH. His results suggest that the urine pH was the more important because of 22 Finland, M., Murray, R., Harris, H. W., Kilham, L. and Meads, M.: Development of streptomycin resistance during treatment. J.A.M.A., 132: 16-21, 1946. 23 Murray, R. and Finland, M.: Effect of pH on streptomycin activity. Am. J. Clin. Path., 18: 242-252, 1948. 24 Kass, E. H.: Asymptomatic bacteriuria as a marker in the study of antibiotic-resistant infections of the urinary tract. Proc. of International Colloquium on Resistant Infections. Eaton Laboratories, Norwich, N. Y., 1958. 25 Harris, H. W., Murray R., Paine, T. F., Kilham, L. and Finland, M.: Streptomycin treatment of urinary tract infections with special reference to the use of alkali. Am. J. Med., 2: 229-250, 1947.
More effective in alkaline urine, especially in coliform infections. More effective at pH 5.5 than at 7.2. More effective at pH 7 .2 than at 5.5. More effective at pH 5 than at pH 7. Less effective at pH 5 than at pH 7. Less effective as pH increased from 5.75 to 7.7. More effective as pH increased from 5.3 to 7.3. Least effective at pH 6.6, better at higher and lower pH values. More effective in alkaline medium. More effective in alkaline medium.
Nitrofurantoin activity was increased rn acid unne. Nifurethazone and streptomycin activity was better in alkaline urine.
the better results obtained when the patients received alkali; it is unlikely that the pH of the tissues could be altered significantly. Polymyxin B appeared to be more effective in an alkaline pH when used against members of the coli-aerogenes group. Against Pseudomonas strains, this antibiotic seemed to be more effective in acid but this may be related to the marked growth inhibition of pseudomonas strains by acid. Several of the "broad-spectrum" antibiotics, chloramphenicol, tetracycline, and oxytetracycline, appeared slightly more effective when the pH of the urine was lowered. The difference was not as pronounced as with the other antibiotics and chemotherapeutic agents studied. Eagle 21 similarly found that chloramphenicol and oxytetracycline activity was less affected by pH, although chloramphenicol seemed least effective at pH 6.6. He found oxytetracycline to be more effective against E. coli when the pH was 7. 7. Our findings in urine were similar. In terms of the organisms, S. faecalis and Ps. aeruginosa appeared to be particularly sensitive to acid so that acidification should supplement antibiotic therapy. The E. coli, A. aerogenes, and S. aureus findings are varied. Here, the "broadspectrum" drugs are more effective in an acid medium (with the exception of E. coli vs. oxy-
THOMAS W. MOU TABLE
6. The pKa values of some antibacterial agents
Carbomycin 26 Chloramphenicol 27
A= 3.3* B = 7.4
Erythromycin26 Nifurethazone 28 Ni trofurantoin 28 N ovobiocin 29 Oleandomycin 26 Oxytetracycline 26
Penicillin30 Polymyxin B 26 Streptomycin20 • Sulfanilamide 31 Sulfapyridine 31 Sulfathiazole 31 Sulf adiazine 31 Tetracycline 26
Portions are converted to an inactive glucuronate in urine. Unstable in alkali at a pH of 10.82. Exists as a zwitterion in neutral solution.
C = 9.3 8.6 7.6 (pKb) 7.2 4.3, 9.1 8.6 A= 3.3* B = 7.3 C = 9.1 2.8 Approx. 9 10.43 8.43 7 .12
6.48 A= 3.3* B = 7.7 C = 9.7 6.6
Loses activity below pH 5. Primarily active in alkaline substrate. Primarily active in an acid substrate.
Exists as a zwitterion in neutral solution.
A strong acid, most effective in an acid substrate. Stable at pH 2-7. A triacidic base, most effective in an alkaline substrate. [Bacteriostatic power appears to be at a maximum when ( the pKa is close to the pH of the culture medium.. J
tricarbonyl methane system. B
Exists as a zwitterion in neutral solution.
ammonium cation system. C
tetracycline), but polymyxin Band streptomycin prefer a neutral or slightly alkaline substrate for maximum effectiveness. The degree of ionization (pKa value or ionization constant) of antibiotics and chemotherapeutic agents may be related in some instances to their effectiveness in acid or alkaline urine. The development of antibacterial agents that are effective at specific pH ranges might be useful in the management of chronic urinary tract infections. In the case of the nitrofurans, their effectiveness appears to be directly related to their pKa value, with antibacterial activity inversely proportional to the ionization of the compound. Frequently, however, the pH at which the least amount of ionization occurs is beyond the range of bacterial growth. Also, some of the antibiotics have multiple pKa values at several pH levels. This makes it difficult to relate pKa values of some of the currently available antibiotics to their antibacterial activity The pKa
phenolic diketone system.
values of 18 antibacterial agents are listed in table 6.2,-31 Information supplied by Eaton Laboratories 28 indicated that nitrofurantoin has a pKa of 7 .2 (2.0 per cent is ionized at pH 5.5, 87 per cent is ionized at pH 8.0), and that it inhibits bacterial growth more effectively when the pH of the substrate is acid. We, also, found greater inhibition of growth at pH 5.0, when the organism was 26 Antibiotics, Their Chemistry and Non-Medical Uses, ed. by H. S. Goldberg. Princeton, N. J.: D. van Nostrand Co., 1959. 27 Encyclopedia of Chemical Technology, vol. 13, ed. by R. E. Kirk and D. F. Othmer. New York: Interscience Encyclopedia, 1954. 28 Unpublished data, Eaton Research Laboratories, Norwich, N. Y. 29 Handbook of Toxicology, II, Antibiotics, ed. by W. S. Spector. Philadelphia: W. B. Saunders Co., 1957. 30 Regn a, P. P.: Chemistry of antibiotics of clinical importance. Am. J. Med., 18: 686-716, 1955. 31 The Sulphonamides and Antibiotics in Man and Animals, ed. by J. S. Lawrence and J. Francis. London: J. K. Lewis and Co., 1953.
EFFECT OF URINE pH ON ANTIBACTERIAL ACTIVITY
sensitive to nitrofurantoin. The analogue, nifurethazone, although not usable from a clinical standpoint, has important theoretical implications. Its pKb is 7.6 28 (2.0 per cent is ionized at pH 7.5 but 90 per cent at pH 5.5); thus, it is most active at an alkaline pH, where the smallest proportion is ionized. Clinically, the added effectiveness of altering the pH of the urine of patients with chronic infections is difficult to evaluate. The number of patients so treated was limited but both acidifying and alkalinizing agents were studied in each. Chloramphenicol was more effective when the patient was excreting acid urine, but the comparative reduction in bacteria was less than with nitrofurantoin. Although Eagle21 found chloramphenicol to be slightly less effective at pH 6.6 and more effective at levels both higher and lower than this, it was not possible to titrate clinical studies to this degree. Clinically, tetracycline yielded variable results, although the findings were in keeping with laboratory observations. When E. coli was the infecting organism, an alkaline urine seemed preferable but the reverse was true of B. alkaligenes, which normally prefers an alkaline medium. The theoretical and practical question of
whether antibacterial agents are most effective when conditions are favorable for the active multiplication of the organism or when this is limited by the environment remains to be determined. SUMMARY
Chloramphenicol and oxytetracycline appear to be less sensitive to pH change in vitro and in vivo. Nitrofurantoin activity is significantly enhanced in acid. The pK value of some antibacterial agents may be related to their bactericidal and bacteriostatic activities. Streptomycin is more effective in alkali except when the organism is inhibited by an acid pH. Penicillin has limited usefulness in urinary tract infections, but appears most effective when the pH is acid. Polymyxin B is most, effective against coliform organisms in alkali and against Pseudomonas organisms in acid. S. aureus, S. faecalis, and Ps. aeruginosa are inhibited most at an acid pH with all agents studied, except streptomycin. Environmental changes for the bacterial organism occasioned by pH change may be an important factor in treating urinary tract infections effectively. The author is indebted to Mrs. Ann Hawker and Mrs. Sondra Kluger for technical assistance, and to Richard Lubera for additional aid.