Note Is Pyrazinamide Bactericidal against Mycobacterium tuberculosis?1-3

LEONID B. HEIFETS and PAMELA J. LINDHOLM-LEVY

Pyrazinamide (PZA) is now considered the third most important drug in the modern chemotherapy oftuberculosis (1). Three periods in the history of the use of PZA have been distinguished (2): (1) the initial studies (1952 to 1959) colored with reservations about its use because of the rapid emergence of drug resistanceand the drug's potential for hepatotoxicity; (2) use as a second-line drug (1958 to 1970) in the treatment of chronic casesresistant to isoniazid and streptomycin; (3) employment in clinical trials of short-course treatment of newcases (1971 to 1980). Finally, there is the present period of renewed interest in PZA followingthe reports after 1980(3-6). Analysis of the short-course studies indicated that PZA played a unique role, in a combination with rifampin (RMP) and isoniazid (lNH), in accelerating the sterilizing effect; this allowed reduction in the duration of chemotherapy from 9 to 6 months. One of the complicated issues surrounding the enigma of PZA is the so-called sterilizing activity in vivo, often interpreted as its bactericidal activity. The early reports indicated high sterilizing activity of PZA in the treatment of infected mice (7-9), as well as in clinical observations when PZA was used in combination with INH (10, 11), or INH + streptomycin (8M) (12, 13) or INH + 8M + RMP (14). Interpreting the early studies in which the organs of mice became culture negative after three months of treatment with PZA + INH (8), Grossett stressed later (15) that "such results had never been achieved before with any of the then existing drug regimens." Further studies of this sterilizing activity ofPZA in mice, reported ten years later (16, 17), indicated that after a 9O-day drugfree follow-up interval M tuberculosis was found in one-third of the animals treated with INH + PZA. The authors emphasized that the "vanishing" phenomenon did not mean that the bacilli were totally eliminated from the tissues. Even in the remaining two-thirds of the previously treated mice, M tuberculosis persisted for many months in a non-cultivable state. Nevertheless, the effect of PZA on M tuberculosis was named in the subsequent publications (18-20) as bactericidal. This judgment was based on the fact that the pulse exposure of M. tuberculosis (H. 7Rv) to PZA inhibited the regrowth of the bacteria after they werewashed and placed into a drug-free 250

SUMMARY Bactericidal activity of pyrazinamide (PZA) was tested at pH 5.6 In 7H12 broth against drug-susceptible M. fuberculosls strains. The highest tested concentrations of PZA, 500 and 1,000 /lg/ml, killed no more than 76% of the bacterial population. These concentrations are more than 32 times greater than the minimal Inhibitory concentration (MIC) and the achievable In 1';1'0 concentrations. Despite high clinical efficacy of PZA and Its so-called sterilizing activity In mouse experiments, this drug Is much less bactericidal tn l'itl0 than any other known antituberculosis drug. AM REV RESPIR DIS 1990; 141:250-252

broth (18, 19). The authors of these reports employed the term "bactericidal drug" for any antimicrobial agent that produced a delay in regrowth of several days, and the term "bacteriostatic drug" for an agent that did not affect the subsequent regrowth of bacteria after a pulse exposure. These reports did not demonstrate any killing curves with PZA, as shown for other drugs considered to be bactericidal (21). Furthermore, PZA produced an antagonism with INH when tested in a liquid medium (20), contradictory to the above described results obtained in mouse experiments and clinical observations. A recent report (22) about the lack of activity of PZA against intracellularly growing M tuberculosis added more controversy to the knowledge about PZA. It is not clear whether the inhibition of regrowth of M. tuberculosis after a pulse exposure to PZA, defined in the above cited publicatons as bactericidal, is associated with actual killing, at least of a portion of the bacterial population, or whether it is purely bacteriostatic with a significant temporary damaging effect on the bacterial growth and metabolism. But regardless of the definitions, we considered it important to quantitate the bactericidal potency of PZA in vitro in terms of the minimal bactericidal concentration (MBC) and to make an attempt to determine the minimal inhibitory concentration (MIC)/ MBC ratio, as suggested for measurement of the bactericidal activity of other antimycobacterial drugs (23). Pyrazinamide (PZA; Sigma Chemical Co., St. Louis, MO) was dissolved in sterile, distilled, bottled water for irrigation (Abbott Laboratories, North Chicago, IL) and filter sterilized (Nalge Co., Rochester, NY). Serialdoubling dilutions weremade in sterile, distilled water and aliquots of each dilution were kept at -70°C until needed, for a period not longer than 6 months. Our quality control tests

indicated that the activity of the drug remained unchanged, which is in agreement with the manufacturer's information about the stability of PZA under these conditions. Previous studies on bactericidal activity of PZA, as well as other antituberculosis drugs, against M tuberculosis, have been done with one strain only (15-18,20,21). We selected for our study the same M tuberculosis strain, H.,R v, but also included three clinical isolates of M tuberculosis, all obtained from patients before treatment. All four strains were cultivated in 7H9 broth until growth was equal to the turbidity of a No. I McFarland standard. Aliquots of the broth cultures werestored at -70°C until needed for an experiment. A seed culture was prepared by thawing a frozen vial of M tuberculosis culture and transferring 0.5 ml to a tube of 7H9 broth. When the 7H9 had incubated for four to six days and was slightly turbid, 0.1 ml of the broth was inoculated into duplicate vials of 7HI2 broth (Becton Dickinson Diagnostic Instrument Systems,Towson,MD). The 7H12 broth cultures incubated at 37°C until the daily Growth Index (01) reading on the BACTEC 460-TB instrument (Becton Dickinson) reached approximately 500, at which time the broth was used to inoculate sets of vials for MIC and MBC determination. An inoculum of such a culture (0.1 ml per vial) produces, according to our previous observa-

(Received in original form February 7, 1989 and in revised form June 1, 1989) 1 From the National Jewish Center for Immunology and Respiratory Medicine, and the Departments of Microbiology and Immunology and of Medicine, University of Colorado Health Sciences Center, Denver, Colorado. 1 Supported by the Biomedical Research Grant No. S07RR 05842-08 from the National Jewish Center for Immunology and Respiratory Medicine, Denver, Colorado. 3 Address correspondence and reprint requests to Leonid Heifets, M.D., National Jewish Center for Immunology and Respiratory Medicine, 1400 Jackson St., Denver, CO 80206.

251

NOTE

tions (24-27), an initial concentration of about ent at the moment when the drug was added to the growingculture. The killingeffectof PZA was 10' CFUlml. Sets of 36 7HI2 vialswereused for each experi- determinedby the sametechniqueof samplingand ment. The vials contained 4.0 ml of 7HI2 broth plating as describedabovefor MIC determination specially manufactured(Becton Dickinson) to have and even in the same experiment. The killing efpH 6.0. The seed vials were mixed together and fect was determined in the presenceof PZA conclumpsof bacilli were brokenbyaspiratingand rap- centrations that weretwo, four, eight, 16, and 32 idly releasingthe seedbroth through a 27 g needle timeshigher than the MICs (seeabove).Compariofthe allergist's syringe (Becton Dickinson, Ruther- son of the number of CFU/ml in these vialsat the ford, NJ). Thirty-two vials were inoculated with end of a 15-day period of observation, with the 0.1 ml of this undiluted bacterial suspension.1\vo number of CFU/ml when the drug was added, more vials were inoculated with a 1:100 dilution providedinformation about the percentage of bacof the bacterial suspensionto make drug-freecon- teria killed. trols representing 1010 of the bacterial population. All vials wereincubated at 37°C and the 01 was The MIC of PZA determined in 7H12 broth read and recorded daily on the BACTEC instruat pH 5.6 was 311lg/ml for three of the tested ment. When the 01 wasapproximately50,usually achievedin 24 to 48 h, appropriate PZA solutions strains, and 62 ug/ml for one remaining strain wereadded to obtain vialscontaining 1,000, 500, of M tuberculosis. The MIC of PZA for the 250, 125,62.5, 31.25, or 15.6 ug/ml, four vials for same strains tested at pH 6.0 was 250 to 500 each concentration. Our previous observations ug/rnl, These data confirmed our previous (24-27) indicated that at this GI reading the ex- observations about the correlations between pected number of CFU/ml should be about 1()4. pH conditions and the MIC ofPZA (29). The The experiments withfourstrainsusedas teststrains lowest concentration used in these experiin this study confirmed our expectations (table I). ments, 31Ilg/ml, determined as MIC for three Half of the drug-containing and drug-free vialswere of four tested strains, produced inhibition of also inoculated with 0.5 ml of a phosphoric acid growth within 15 days of cultivation. There solution that lowered the pH to 5.6 (28). The volume of all vialswasadjusted to 5.0 ml by the addi- was no increase or detectable decrease in the tion of the 7HI2 broth. One drug-free undiluted number of viable bacteria in experiments with control vial was sampled and plated for determi- two strains, and a 27. 711fo decrease in the third nation of the numberof CFU/rnI presentwhenPZA one. The growth ofthe fourth strain, for which was added to the cultures.The 7HI2 vialswerein- the MIC was determined as 62 ug/ml, was cubated at 37°C for 2 wk and were read daily in inhibited by this concentration but there was the BACTECinstrument. At various time periods no decrease in the number of viable bacteria samplesweretaken fromalternatevials,dilutedapwithin the 15-day period of observation. propriately based in GI daily readings, and plated The bactericidal activity of PZA was evaluon 7Hll agar. After 2 wk of incubation at 37°C in a 5% CO, chamber,the plates werecounted and ated at the same pH 5.6, with the same four M. tuberculosis strains. The initial number the number of CFU/ml were determined. MIC wasdefined as the lowestconcentration of of viable bacteria was within the expected a drug inhibiting the growth of more than 99% range of CFU/ml for all strains tested: 0.6 of the bacterialpopulation in 7HI2 broth (pH 5.6). x 104 , 0.9 x 10" 1.1 x 10" and 1.7 x 104 The MIC wasdeterminedbyplating samplestaken (table 1). The differences in the initial numfrom alternate vials at various time points. These ber of CFU/ml from 0.6 x 104 to 1.7 x lQ4 samples werediluted before plating, based on the expectedrange of the number of CFU/ml to have were within a permissible tenfold range used not less than 50 and not more than 500 colonies for determining bactericidal activity of anper plate. The lowestconcentration of PZA in the timicrobial agents (30). Despite these differpresence of which there was no increase in either ences in the initial number of CFU/ml, the daily 01 reading or the number of CFU/ml during killing effect of PZA was in the same low a 15-dayperiod of observation was consideredan range. The initial number of CFU/mi for all MIC. Sincethe drug wasadded at the day of culti- four strains was sufficient to detect a 99% vation when the anticipated number of CFU/ml and even 99.9% killing effect of PZA, but shouldhavebeenin a rangeof 1()4 to 10" (seeabove), such high killing effect did not take place. the samplestaken from the vialsthat did not show For all four strains, some killing took place daily 01 increaseswerediluted 1:10 and I:100, and with a concentration twofold higher than the 0.5 ml of each dilution wasinoculated into dupliMIC (table 1). Further increase in the PZA cate agar plates. The MBC was defined as the lowestconcentra- concentration usually produced a greater killtion killing 99% of the bacterial population pres- ing effect, but even with such high concen-

TABLE 1 EVALUATION OF BACTERICIDAL ACTIVITY OF PZA AT pH 5.6 AGAINST FOUR DRUG-SUSCEPTIBLE M. TUBERCULOSIS STRAINS

MIC ofPZA Strain

{J.tglm~

Number of CFUlml When Drug Added

H37 Rv 1630 126 131

31 31 31 62

6,300 17,700 11,200 9,000

Percent of Bacterial Population Killed in Presence of Different PZA Concentrations {J.tglm~

31

62

125

250

500

1000

0 27.7 0 0

33.4 57.3 49.2 0

59.7 60.7 63.5 33.4

67.8 60.7 51.8 25.0

57.2 75.9 56.3 35.3

72.4 73.6 61.3 54.2

trations as 500 and 1,000 ug/ml the proportion of the bacterial population killed was not more than 76% for one of the strains, and was even lower for the three remaining strains (table 1). It was technically not feasible, due to the limited solubility of PZA, to test substantially higher concentrations in order to find a concentration that could kill 99% of the bacterial population. In fact, it is possible that concentrations higher than 1,000 ug/rnl would not kill 99% of the bacterial population. Therefore, we concluded that the MBC of PZA at pH 5.6 is higher than 1,000 ug/ml, and the MIC/MBC ratio should be expressed as 1:>32.

* * * The greater than 32-foJd difference between the MIC and MBC of PZA found in this study is a higher ratio than that of any other antituberculosis drug tested in vitro against M tuberculosis (24-27). These results were obtained at pH 5.6, the lowest pH that allowed cultivation of M tuberculosis. It is possible that MIC and MBC values could be lower in the more acidic environment, pH 5.0, present within macrophages. But even accepting this assumption, the highest concentration of PZA achievable in serum or within macrophages, 25 to 50 ug/ml (31, 32), most likely would not produce a 99% killing effect. It is a paradox that one of the clinically most efficient antituberculosis drugs has very low bactericidal activity. Further experiments, with mycobacteria multiplying within macrophages, are necessary to evaluate the bacteriostatic/bactericidal ratio of PZA under these conditions in comparison with the findings presented above. The low bactericidal activity of PZA found in pH 5.6 liquid medium explains the early report (16, 17) about the persistence of M tuberculosis in the tissues of mice previously "sterilized" by a combination of INH + PZA. It is possible that the high clinical efficacy of PZA, despite its low bactericidal activity, is a result of bacteriostatic effects of PZA combined with unfavorably acidic conditions within the macrophages (18, 20). Such a hypothesis may have some ground taking into account data (33) showing that the pH of small areas of the cytoplasm surrounding the phagocytized mycobacteria can drop to 4.7, probably as a result of transformation of PZA into pyrazinoic acid by mycobacterial amidase. This hypothesis, as well as many other suggestions explaining the mode of action of PZA, has to be tested and will be a subject of our further studies. Regardless of future findings explaining the mechanism of PZA activity, the fact that PZA does not have the ability to kill M tuberculosis in vitro at the concentrations achievable in vivo suggests that lack of bactericidal activity in vitro does not necessarily predict low clinical efficacy of an antimycobacterial drug. References I. O'Brien RJ, Snider DE. Thberculosis drugsold and new(editorial). Am RevRespir Dis 1985; 13l:309-11.

NOTE

252 2. Zierski M. Pharmakologie, Toxikologie und Klinische Anwendung von Pyrazinamid. Prax Klin Pneumol 1981; 35:1075-105. 3. Hong Kong Chest Service/British Medical Research Council. Controlled trial of four thriceweekly regimens and a daily regimen all given for 6 months for pulmonary tuberculosis. Lancet 1981; 1:171-4. 4. Singapore Tuberculosis Service/British Medical Research Council. Clinical trial of six-month and four-month regimens of chemotherapy in the treatment of pulmonary tuberculosis: the results up to 30 months. Tubercle 1981; 62:95-102. 5. Snider DE, RogowskiJ, Zierski M, Bek E, Long MW. Successful intermittent treatment of smearpositive pulmonary tuberculosis in six months: a cooperative study in Poland. Am Rev Respir Dis 1982; 125:265-7. 6. British Thoracic Association. A controlled trial of six months chemotherapy in pulmonary tuberculosis. Final report: results during the 36 months after the end of chemotherapy and beyond. Br J Dis Chest 1984; 78:330-6. 7. McDermott W, Ormond L, Muschenheim C, Deuschle K, McCune RM, Tompsett R. Pyrazinomide-isoniazid in tuberculosis. Am RevTuberc 1954; 69:319-33. 8. McCune RM, Tompsett R, McDermott W. The fate of M. tuberculosis in mouse tissue as determined by the microbial enumeration technique. 11. The conversion of tuberculosis infection to the latent state by the administration of pyrazinamide and a companion drug. J Exp Med 1956; 104: 763-802. 9. Grumbach F. Activite antituberculeuse experimentale du pyrazinamide (p.ZA.). Ann Inst Pasteur Lille 1958; 94:694-708. 10. Campagna M, Hauser G, Greenberg HB. The eradication of Mycobacterium tuberculosis from the sputum of patients treated with pyrazinamide and isoniazid. Am Rev Respir Dis 1%2; 86:636-9. 11. Grumbach F, Grosset J. La pyrazinamide dans Ietraitement de courte duree de la tuberculose murine. Rev Fr Malad Respir 1975; 3:5. 12. East African British Research Council. Con-

trolled trial of four short-course (6-month) regimens of chemotherapy for treatment of pulmonary tuberculosis. Third report. Lancet 1974; 2:237-40. 13. Hong Kong Tuberculosis Treatment Service/ British Medical Research Council. Controlled trial of 6- and 9-month regimens daily and intermittent streptomycin plus isoniazid plus pyrazinamide for pulmonary tubercuosis in Hong Kong. Tubercle 1975; 56:81-96. 14. East African British Medical Research Council. Controlled clinical trial of four short-course (6-month) regimens of chemotherapy for treatment of pulmonary tuberculosis. Second East African/British Medical Research Council study. lancet 1974; 2:1100-6. 15. Grosset J. The sterilizing value of rifampicin and pyrazinamide in experimental short-eourse chemotherapy. Bull Int Union Against Tuberc 1978; 53(1):5-12. 16. McCune RM, Feldman FM, Lambert HP, McDermott W. Microbial persistence. I. The capacity of tubercle bacilli to survive sterilization in mouse tissues. J Exp Med 1966; 123:445-68. 17. McCune RM, Feldman FM, McDermott W. Microbial persistence. II. Characteristics of the sterile state of tubercle bacilli. J Exp Med 1966; 123: 469-86. 18. Dickinson JM, Mitchison DA. Observations in vitro on the suitability of pyrazinamide for intermittent chemotherapy of tuberculosis. Tubercle 1970; 51:389-96. 19. Fox W, Mitchison DA. Short-course chemotherapy for pulmonary tuberculosis. Am Rev Respir Dis 1975; 1ll:325-53. 20. Dickinson JM, Aber VR, Mitchison DA. Bactericidal activity of streptomycin, isoniazid, rifampin, ethambutol, and pyrazinamide alone and in combination against Mycobacterium tuberculosis. Am Rev Respir Dis 1979; 116:627-35. 21. Dickinson JM, Mitchison DA. Bactericidal activity in vitro and in the guinea pig of isoniazid, rifampicin and ethambutol. Thbercle 1976;57:251-8. 22. Rastogi N, Potar MC, David HL. Pyrazinamide is not effective against intracellularly growing Mycobacterium tuberculosis. Antimicrob Agents

Chemother 1988; 32:287. 23. Heifets LB. Qualitative and quantitative drug susceptibility tests in mycobacteriology. Am Rev Respir Dis 1988; 137:1217-22. 24. Heifets LB, Lindholm-Levy PJ, Iseman MD. Rifabutin: minimal inhibitory and bactericidal concentrations for Mycobacterium tuberculosis. Am Rev Respir Dis 1988; 136:719-21. 25. Heifets LB, Iseman MD, Lindholm-Levy PJ. Ethambutol MICs and MBCs for Mycobacterium tuberculosis. Antimicrob Agents Chemother 1986; 30:927-32. 26. Heifets LB, Lindholm-Levy PJ. Streptomycin, amkiacin, kanamycin, capreomycin: bactericidal activity against Mycobacterium avium in comparison to M tuberculosis. Antimicrob Agents Chemother 1989; 33:1298-301. 27. Heifets LB, Lindholm-Levy PJ, Flory MA. Bactericidal activity in vitro of various rifamycins against Mycobacterium avium and M tuberculosis. Am Rev Respir Dis 1990; (In Press). 28. Heifets LB, Iseman MD. Radiometric method for testing susceptibility of mycobacteria to pyrazinamide in broth. J Clin Microbiol1985; 21:200-4. 29. Salfinger M, Heifets L. Determination of pyrazinamide MICs for Mycobacterium tuberculosis at different pHs by the radiometric method. Antimicrob Agents Chemother 1988; 32:1002-4. 30. National Committee for Clinical Laboratory Standards. Proposed guideline: methods for determining bactericidal activity of antimicrobial agents (document M26-P, ISSN 0273-3099). NCCLS, 1987, Vol. 7, No.2. 31. Subbamal J, Krishnamurthy DV,Tripathy SP, Venkataraman P. Concentrations of pyrazinamide attained in serum with different doses of the drug. Bull WHO 1968; 39:771-4. 32. Acocella G, Carlone NA, Cuffini AM, Cavallo G. The penetration of rifampin, pyrazinamide, and pyrazinoic acid into mouse macrophages. Am Rev Respir Dis 1985; 132:1268-73. 33. Pavlov EP, Tushov EG, Konyaev GA. The effect of pyrazinamide on pH of cytoplasma areas surrounding phagocytized mycobacteria. Probl 'Iuberk 1974; 1:77-9.