• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services


  • Print
  • Share
  • E-mail

Safety of Long Term Therapy with Penicillin and Penicillin Derivatives

December 6, 2001
MO: Chuck Cooper, MD


Safety of Long Term Penicillin and Penicillin Derivatives

The safety of penicillin and penicillin derivatives when administered either intramuscularly, intravenously, or orally for extended periods of time (beyond the usual duration of use) can be extrapolated from multiple published studies. A review of the medical literature reveals studies in which such drugs have been used therapeutically for extended treatment durations. This includes studies of the treatment of recurrent acute otitis media, endocarditis, salmonella infections, prophylaxis of at risk populations (asplenic children, young children with sickle cell disease, patients with prior rheumatic fever), and the long term treatment of certain types of Lyme disease. Although most of these studies focus primarily on efficacy, they do routinely contain general statements indicating the safety and tolerability of the treatment under study. A few studies mentioned the presence of isolated cases of rash, gastrointestinal events, but none of the studies report the occurrence of a serious adverse event.

A review of the Adverse Event Reporting System (AERS) database as performed by an Office of Post-Marketing Drug Risk Assessment (OPDRA) special programmer revealed occasional cases of adverse events associated with the administration of long term penicillin therapy, however, these adverse events were consistent with the known adverse event profile of penicillin. OPDRA concluded that "no increase in or atypical adverse drug reactions were noted in the few cases involving long-term administration of oral or intramuscular penicillin G or V (all dosage forms)." This data is generated via isolated adverse event reports submitted to the FDA, and it may be difficult to determine the overall rates of adverse events in relation to treatment duration. A similar analysis is currently being performed by OPDRA for amoxicillin and Augmentin®.

There is a higher incidence of cholestatic or mixed cholestatic-hepatocellular patterns of liver injury associated with administration of Augmentin®. This has been demonstrated by several published case reports. 8, 12, 14, 17, 20, 22, 24, 27, 28 Fatalities are very rare (on the order of one in several million) and similar hepatotoxicity is not thought to occur with amoxicillin administration. A study by Garcia-Rodriguez identified specific sub-populations who might be at increased risk for hepatotoxicity.10 These groups include older patients, and patients on long term therapy. The overall risk is still very low but was noted to be significantly higher than for the comparison group receiving amoxicillin. The information from this study was submitted in a labeling supplement and resulted in the identification of these risk groups in the label.

These antibiotics have an extensive history of use, and the lack of any serious adverse events as reported in these studies support their reputation as safe drugs. Reassurance can be derived from the fact that no increase in serious adverse events has been reported in the literature despite broad usage. Clinicians should still be aware that, although quite unlikely, when using these drugs over extended period of time, the potential exists for slight increases in the rates of specific adverse events which are infrequent enough to have escaped detection. Although there is no data to suggest such an increase (except for Augmentin® and hepatotoxicity), adverse events could include increased liver function tests, serum sickness, interstitial nephritis, effects on anti-coagulant therapy, and neutropenia.


I. Review of Literature

a. Penicillin prophylaxis for rheumatic fever (n= 822)

Four large published clinical trials were identified assessing effectiveness of chronic penicillin therapy as secondary prophylaxis in pediatric patients and adolescents (1-19 years old) with a history of rheumatic fever. 4,15,16,26 These studies include a total of 511 patient treated with 1.2 million units if IM benzathine penicillin for a duration of 6 mos to 12 years. In addition, 159 patients were treated with 200,000 units orally BID, 143 were treated with 200,000 units orally BID-TID, and 9 patients were treated with an unspecified dose. In these studies, safety and tolerability of the regimen is summarized by such descriptions as the treatment was "generally well tolerated." There were no reports in any of these studies of adverse events including serious adverse events or adverse events requiring treatment.


b. Penicillin prophylaxis for patients with sickle cell disease (n= 534)
Four large published clinical trials assessing effectiveness of chronic penicillin therapy as prophylaxis of infection in children (3 - 60 mos old) with sickle cell disease were identified. 2, 3, 11, 13 There were 232 children treated with125-250 mg of oral penicillin V and 302 children treated with monthly IM doses of 600,000 units of benzathine penicillin. These studies contain general statements that "orally administered penicillin is safe" or that "the penicillin was generally well tolerated." There were no reports in any of these studies of adverse events including serious adverse events or adverse events requiring treatment.
c. Amoxicillin (n= 627)

Nine studies were identified which examined long term amoxicillin prophylaxis (4 weeks - 6 months) in children with recurrent acute otitis media. 29-37 There were 488 patients who were treated with doses ranging from 20-40 mg/kg/day of amoxicillin. In total, there were 11 adverse event reported. Four patients were reported to have experienced urticaria,31 two patients each were reported to have had vaginitis,31 diarrhea,32,37 and "rash,"37 and one patient was reported to have had an "allergic reaction."29 Three of the studies did not specifically mention safety data or adverse events. 34,35,36

Three studies were identified which examined long-term amoxicillin administration (4 weeks) in adults for the treatment of various salmonella infections. 5, 9, 18 In total there were 38 subject who took doses ranging from 250 mg of amoxicillin BID to 2,000 mg of amoxicillin TID. In one study (in which patients were treated with 2,000 mg amoxicillin TID) it was reported that 5 of 15 patients required a lowering of the dose because of "gastrointestinal side effects." 18 In one other study, an 11 year old child developed a diffuse erythematous rash three weeks after amoxicillin therapy was started and one week after completion of treatment with niridazole.9

Two studies were identified in which extended treatment with amoxicillin was studied for the treatment of Lyme disease. 7, 19 One study included 38 adult patients who were treated with 500 mg of amoxicillin TID combined with probenecid for three weeks for treatment of erythema migrans borreliosis. 7 One patient was removed from this study because of development of a macular/papular rash. There were no other adverse events or serious adverse events which were reported in this study. A second study contained 30 patients treated for disseminated Lyme disease with ceftriaxone for 14 days followed by amoxicillin 500mg/ probenecid 500mg for 100 days. 19 During this study, there were 2 episodes of C. difficile colitis, 1 episode of nausea, and 1 episode of diarrhea. There were also 22 cases in which patients had symptoms of fever, transient rash, marked worsening of symptoms, or cardicac arrythmia. These cases were attributed to possible Jarisch-Herxheimer-like reactions which resolved without altering therapy.

One study was identified in which 250 mg of amoxicillin was administered TID for an average length of 23 weeks as prophylaxis for acute exacerbation of chronic bronchitis. 6 In this study which contained 32 adult patients, 3 withdrew because of diarrhea, 3 patients reported nausea, 2 reported vomiting, 1 reported a sore mouth, and one reported dysgusia.

One case report described an adult patient treated with 42 days of oral amoxicillin plus gentamicin for the treatment of enterococcal endocarditis.21 This patient experienced an increase in creatinine which was attributed to the gentamicin. No other adverse events were reported to have occurred in this patient.

d. Augmentin (n=132)

One study was identified in which 132 patients, aged 1-10 yrs old were treated with 4 weeks of Augmentin for secretory otitis media. 25 No adverse events or withdrawals due to adverse events were reported in this study.

See Section III for information regarding Augmentin and potential hepatotoxicity.

II. AERS Database Analysis ( The following is excerpted from OPDRA review)
On October 9th, 2001, OPDRA was asked if AERS could provide information on adverse events reported in association with long-term (60 days or greater) administration of oral penicillin and penicillin G procaine. OPDRA developed an algorithm to find reports in which the duration was reported as, or could be calculated to have been, 60 days or greater. All reports found in the search were retrieved for hands-on analysis.
a. Summary of Data


A total of 61 cases involving long-term penicillin use were retrieved. Four cases were excluded as duplicates and an additional 32 were excluded as irrelevant because they did not actually involve long-term therapy or did not involve a reaction related to penicillin use. This left a total of 25 evaluable cases. Only two cases involved a penicillin G procaine component, given in combination with benzathine penicillin.

Indications: The majority of the cases (16; 5 of which involved oral therapy) were related to rheumatic fever/heart disease. The remaining indications included Lyme disease (2 cases), recurrent streptococcal infections, prophylaxis for throat and mouth infections, chronic tonsillitis, strep throat, acute annexitis, and Whipple's disease. One case had an unknown indication.

1. Cases with prolonged oral therapy (N=11)

Duration: 2 months to 26 years; median_18 months; mean_74 months

Note: 7 cases had a range from 2 to 25 months and 4 had a range from 7 to 26 years

Reactions observed:

· allergic / rash / urticaria 3

· erythema multiforme / possible SJS 2

· tooth discoloration 2

· lethargy / vertigo 1

· decreased platelet count 1 (also taking quinidine)

· cholestatic jaundice 1

· tinnitus 1 (26 year treatment duration)

2. Cases with prolonged IM therapy (N=14) (12 benzathine; 2 benzathine/procaine combo)

Monthly injections_8 cases

Duration: 3 to 60 months; median_20.5 months; mean_22 months

Every 3 weeks injections_3 cases

Duration: 5 doses; 6 doses; 9 years

Weekly injections_3 cases

Duration: 3 months; 9 months; 12 months

Reactions observed with monthly injections (all benzathine penicillin):

· allergic / anaphylaxis 2

· serum sickness 1

· convulsions / hypertonia / cardiac arrest 1

· bradycardia / cardiac arrest / death 1

· dizziness / death 1

· nausea / headache / fatigue 1

· injection site reaction 1

Reactions observed with every three weeks injections (all benzathine penicillin):

· anaphylaxis/death

· syncope/cardiac arrest/death 3 to 4 minutes following injection (9 year therapy duration)

· dizziness/cyanosis/death within 10 minutes of injection

Reactions observed with weekly injections:

· allergic reaction (benzathine/procaine combo x 9 months)

· chest pain / SOB / palpitations (benzathine/procaine combo x 12 months)

· injection site reaction (benzathine penicillin)
b. Conclusion:
No increase in or atypical ADRs were noted in the few cases involving long-term administration of oral or intramuscular penicillin G or V (all dosage forms).

Medical Officer Comment: It should be noted that these data are generated from individual adverse event reports. Therefore, the total number of patients receiving long-term treatment is unknown and this precludes assessment of the relative incidence of such events in long-term vs. short term therapy.

III. Issues regarding Long Term Augmentin Use and Hepatotoxicity.

Although amoxicillin is generally not regarded as a hepatotoxic drug 23, there have been several cases of hepatoxicity associated with the use of Augmentin®. 8, 12, 14, 17, 20, 22, 24, 27, 28 This toxicity is thought to be related to the presence of clavulanic acid in Augmentin®. Fatalities have been reported 12 but are thought to be very rare (less than one in 4 million prescriptions per the product label). The vast majority of cases of hepatotoxicity associated with Augmentin resolve upon discontinuation of the drug.

In a study published by Garcia Rodriguez in 1996, amoxicillin was compared to amoxicillin plus clavulanate to assess the risk of acute liver injury.10 This was a retrospective cohort study in the United Kingdom and the data were derived from a cohort of 93,433 users of the combination of amoxicillin and clavulanate and 360,333 users of amoxicillin alone. The subjects were aged between 10 and 79 years of age and were followed up from 1991 through 1992. Data was compiled from identifying subjects with suspected liver injury reported in the General Practitioner's Research Database. The charts of identified subjects were then requested for review from the attending general practitioner. The results of this study revealed that there was a 20.2 fold increase in the relative risk of acute liver injury for Augmentin (when compared to amoxicillin amongst patients not exposed to other potential hepatotoxic drugs. The study also revealed that older subjects (65-79 yo) and (independently) those on increased duration of treatment (defined as having 2-4 prescriptions) also resulted in increased relative risk of development of acute liver injury. When the two independent risk factors of increased age and increased duration of use were combined, the largest increase in incidence was found with a rate of 9.3 per 10,000 for such patients taking Augmentin®.

IV. Conclusion

Although there are no specific studies that directly assess the safety of these antibiotics when given over an extended period of time, there is a significant amount of information that supports the safety of such therapy. Despite very wide usage for many years, no reports were found in the literature which described specific adverse events related to long term use of penicillin or amoxicillin. In addition, no unusual safety issues were identified in several published reports that studied the extended use penicillin and amoxicillin for a variety of infections and prophylactic indications. Potential hepatotoxicity associated with the use of Augmentin particularly in older individuals and when used for an increased duration was identified as a potential concern, however, such hepatotoxicity only very rarely resulted in death.



1. Adverse Drug Reactions Advisory Committee. Augmentin jaundice. Australia Adverse Drug Reactions Bull. November 1990:2

2. Angelin DL, et al. Effect of penicillin on nasopharyngeal colonization with s. pneumoniae in children with sickle cell disease. Pediatirics 1984;104:18-22

3. Buchanan GR, et al. Pneumococcal septicemia despite penicillin prophylaxis in children with sickle cell disease. Am J Dis Child 1986;140:428-32

4. Bassili A, et al. Profile of secondary prophylaxis among children with rheumatic heart disease in Egypt. East Mediterr Health J 2000;6(2/3):437-446

5. Bassily SB, et al. Chronic salmonella bacteriuria with intermittent bactermia treated with low doses of amoxicillin or ampicillin. Antimicrob Agents Chemother 1981 Nov;20(5):630-3

6. Cooper J, et al. Prophylactic treatment of chronic bronchitis comparing co-trimoxazole and amoxicillin. Br J Clin Pract 1975 NOV;29(11):307-310

7. Dattwyler RJ, et al. Amoxicillin plus probenecid vs. doxycycline for treatment of erythema migrans borreliosis. Lancet 1990;336(8728):1404-1406

8. Dowsett, JF, et al. Amoxcillin.clavulanic acid induced intrahepatic cholestasis. Dig Dis Sci 1989;34:1290-1293

9. Farid Z, et al. Treatment of chronic enteric fever with amoxicillin. J Infect Dis 1975 Dec;132(6):698-701

10. Garcia Rodriguez LA, et al. Risk of acute liver injury associated with the combination of amoxicillin and clavulanic acid. Arch Intern Med 1996;156:1327-1332

11. Gatson MH, et al. Prophylaxis with oral penicillin in children with sickle cell disease. NEJM 1986;314(25):1593-9

12. Hebbard GS, et al. Augmentin-induced juandice with a fatal outcome. Med J Aust 1992;156:285-286

13. John, AB, et al. Prevention of pneumococcal infection in children with homozygous sickle cell disease. Brit Med J 1984 May;288(6430):1567-70.

14. Larrey D, et al. Hepatitis associated with amoxicillin-clavulanic acid combination report of 15 cases. Gut 1992;33:368-371

15. Lue HC, et al. Long-term outcome of patients with rheumatic fever receiving benzathine penicillin G prophylaxis every three weeks vs. every four weeks. J Pediatr 1994 Nov;125(5 pt 1):812-816

16. Massell BF, et al. Prophylaxis of streptococcal infections and rheumatic fever. JAMA 1979 Apr;241(15):1589-94.

17. Medicine Control Agency. Cholestatic juandice with co-amoxclav. Curr Probl Pharmacovig 1993;19:2

18. Nolan CM, et al. Treatment of typhoid carriers with amoxicillin. Correlates of successful therapy. JAMA 1978 Jun 2;239(22):2352-4

19. Oski J, et al. Comparison of oral cefixime and intravenous ceftriaxone followed by oral amoxicillin in disseminated lyme borreliosis. Eur J Clin Microbiol Infect Dis 1998; 17:715-719.

20. Schneider JE, et al. Cholestatic hepatitis after therapy with amoxicillin/clavulanate. NY State J Med 1989;6:355-356

21. Seligman S, et al. Treatment of enterococcal endocaridits with oral amoxicillin and IM gentamicin. J Infect Dis 1974;129(suppl):S213-215

22. Silvain C, et al. Granulomatous hepatitis sue to combination of amoxicillin and clavulanate. Dig Dis Sci 1992;37:150-152

23. Stricker BH, et al. Drug induced hepatic injury. Amsterdam, the Netherlands: Elsevier Science Publishers; 1992

24. Stricker BH, et al. Cholestatic hepatitis due to amoxicillin and clavulanic acid. Dig Dis Sci 1989;34:1576-1580

25. Thomsen J, et al. Long-term antibiotic treatment of children with secretory otitis media. Acta Otolaryngol 1998;449:49-50.

26. Wood HF, et al. The comparative effectiveness of three prophylactic regimens in preventing rheumatic recurrences. Ann Int Med 1964;60(suppl):31-46

27. Van den Broek JW, et al. Cholestatic hepatitis due to amoxicillin and clavulanic acid. Ned Tijdschr Geneeskd. 1988;132:1495-1497

28. Verhamme M, et al. Cholestatic hepatitis due to an amoxicillin/clavulanic acid preparation. J Hepatol 1989;9:260-264

29. Berman S, et al. Effectiveness of continuous vs. intermittent amoxicillin to prevent episodes of otitis media. Pediatr Infect Dis J 1992;11(2):63-67.

30. Brook I, et al. Prophylaxis with amoxicillin or sulfisoxazole for otitis media: effect on the recovery of penicillin-resistant bacteria from children. Clin Infect Dis 1996;22:143-145

31. Casselbrant ML, et al. Efficacy of antimicrobial prophylaxis and of tympanostomy tube insertion for prevention of recurrent otitis media: results of a randomized clinical trial. Pediatr Infect Dis J 1992;11(4):278-286.

32. De Diego JI, et al. Comparison of amoxicillin and azithromycin in the prevention of recurrent otitis media. Int J Ped Otorhinolaryng 2001;58:47-51.

33. Marchisio P. Comparative study of once-weekly azithromycin and once-daily amoxicillin treatments in prevention of recurrent acute otitis media in children. Antimicrob Agents Chemo Dec 1996;40(12):2732-2736.

34. Mandel EM. Efficacy of antimicrobial prophylaxis for recurrent middle ear effusion. Pediatr Infect Dis J 1996;15(12):1074-82

35. Principi N, et al. Prophylaxis of recurrent acute otitis media and middle ear effusions. Am J Dis Childr Dec 1989;143:1414-1418.

36. Roark R, et al. Continuous twice daily or once daily amoxicillin prophylaxis compared with placebo for children with recurrent acute otitis media. Pediatr Infect Dis J 1997;16(4):376-81.

37. Teele DW, et al. Antimicrobial prophylaxis for infants at risk for recurrent acute otitis media. Vaccine 2001;19:S140-143.