Klebsiella granulomatis (Granuloma inguinale)

Authors: Nigel O'Farrell, M.D.


Calymmataobacterium granulomatis is an intracellular organism which causes an ulcerating sexually transmitted infection, known as donovanosis or granuloma inguinale (1). However, there is some debate about the classification of the causative organism. A link between C. granulomatis and Klebsiella species was suggested on the evidence of antigenic cross reactivity some time ago (40). A more recent report has demonstrated a close similarity between C. granulomatis and Klebsiella pneumoniae and K. rhinoscleromatis with the authors suggesting that the causative organism be reclassified as K. granulomatis comb nov (4, 9).


Transmission of C. granulomatis is usually sexual. Rare transmission during delivery and occasional non-sexual transmission have been reported. The main endemic areas for donovanosis have been in India, Brazil, South Africa, Papua New Guinea and among Australian aboriginals. Over the last twenty years the prevalence of donovanosis has decreased significantly and few cases are reported nowadays even in countries where the prevalence was once high.

Donovanosis has been associated with an increased risk of HIV infection. In Durban where HIV infection had been introduced only recently, the proportion of men with donovanosis and HIV infection increased significantly as the duration of lesions increased suggesting that HIV was acquired via sexual intercourse in the presence of ulcers (34).

Clinical Manifestations

Donovanosis usually commences with ulceration in the ano-genital region. Spread to local lymph nodes is followed by ulceration of skin overlying the lymph nodes. The ulcers are characterized by slow growth, absence of pain and friability. Hypertrophic lesions which stand out from the surrounding skin are common. Primary oral lesions have been described. In women, lesions may involve the cervix and upper reproductive tract. Complications include tissue destruction, scarring, the development of genital lymphoedema, haematogenous dissemination to bones and viscera and squamous carcinoma.

back to top

Laboratory Diagnosis

There are no established protocols for routine isolation and antibiotic susceptibility testing of C. granulomatis, although it is possible to culture C. granulomatis in human peripheral blood monocyte culture and in Hep 2 cells after decontaminating biopsy specimens with suitable antibiotics (10, 23). The diagnosis of donovanosis depends primarily on demonstrating the presence of intracellular organisms (termed Donovan bodies) in large mononuclear cells as Gram negative intracytoplasmic cysts filled with deeply staining bodies that may have a safety pin appearance (33).

Polymerase chain reaction (PCR) analysis using a colorimetric detection system can now be used in routine diagnostic laboratories and a genital ulcer multiplex PCR that includes C. granulomatis has been developed (11, 29).


C. granulomatis is an intracellular infection in which the organism is seen in the form of Donovan bodies sequestered within large histiocytes. It has been suggested that the inability of C. granulomatis to grow extracellularly is linked to the lack of the sucrose transporter gene (scrA) found in other Klebsiella species. A granulomatous inflammatory response is seen which leads to local tissue destruction and cutaneous ulceration.

back to top


Only three papers report results of in vitro experiments with chemotherapeutic agents (6, 12, 39). Chen et al. (12) demonstrated that streptomycin, penicillin, chlortetracycline and chloramphenicol have additive effects against C. granulomatis in vitro when administered in combination (Table 1).

Numerous unsuccessful attempts have been made to find an animal model for donovanosis. In 1931 DeMonbreun et al. reported successful infection of the eyelids of macaques with tissue taken from human lesions of donovanosis but these lesions resolved spontaneously (15). It has not been possible so far to assess antibiotic activity against C. granulomatis in an animal model.


C. granulomatis is found mainly within large histiocytes. The logical choice of therapy is thus an antibiotic with good activity against Gram negative bacteria, with good lipid solubility and capable of achieving a high intracellular to extracellular concentration ratio. Relapse may occur if treatment is taken for too short a time. Donovanosis may in the future prove a suitable infection for treatment with liposome-encapsulated antibiotics.

Drug of Choice

Table 2 summarizes data from the most important therapeutic trials. For a comprehensive bibliography of drug trials for donovanosis up to 1991 readers are referred to a review paper (44).

Drugs of the tetracycline group have been used extensively in the treatment of donovanosis for many years, generally with excellent results. Many published guidelines for the treatment of donovanosis (e.g., World Health Organization, Centers for Disease Control) recommend use of tetracyclines as first choice therapy. Individual well-documented cases of tetracycline resistance have been reported (35) and the drug appeared ineffective in infections contracted in Vietnam (47). It may be assumed that the different forms of tetracycline have equivalent efficacy. Doxycycline is now generally preferred over alternatives for ease of administration (twice daily).

Excellent results with co-trimoxazole have been reported from India (26) and Africa (28). Two treatment failures with co-trimoxazole have been reported from South America (38).

Chloramphenicol is widely used for treatment of donovanosis in Papua New Guinea (30), generally with good results. Lengthy experience with the use of chloramphenicol for a variety of infections in Papua New Guinea has shown that haematological toxicity is rare in this population. While chloramphenicol may be considered a treatment of choice in Melanesians, concerns about potential toxicity would limit its use elsewhere. Recent work from South America suggests that thiamphenicol, a congener of chloramphenicol, which has the convenience of once daily administration and reputedly does not carry the risk of haematological toxicity, is of comparable efficacy (5).

Ceftriaxone can give good results in chronic relapsing cases which have failed to respond a variety of other antibiotics (32). Weekly or daily azithromycin has useful activity which may make it a valuable drug for use in poorly compliant patients (7). Good results with azithromycin have been also been reported in patients who have failed on other therapies (31). In India, an initial study using norfloxacin showed good results (43).

Erythromycin and lincomycin both give good results in donovanosis although only a few trials have been conducted and experience is limited. Streptomycin was for many years the main drug used in the treatment of donovanosis. It continued in use in India up to 1971 (27) and more recently it has been evaluated in combination with tetracycline (28). Most clinicians now prefer to select antibiotics with less toxicity and greater ease of administration. Gentamicin also shows useful activity but has not been used much (30).

Ampicillin gave good results when used in American soldiers in Vietnam (8) but results in other trials have been poor. A treatment failure with ampicillin was reported by Johnson in 1991 (22). On the strength of available evidence ampicillin cannot be recommended as first line treatment.

Combination Therapy

Only one trial has been reported comparing monotherapy with combination therapy in the treatment of donovanosis (28). In this study streptomycin with tetracycline was compared to co-trimoxazole and no significant difference in outcome was noted in the two groups. Good results in the treatment of pregnant women using combined erythromycin and lincomycin have been reported from Australia (3). Small numbers of patients have been treated with combinations of streptomycin plus penicillin or chloramphenicol and with chloramphenicol plus tetracycline.

The efficacy of antibiotics in the treatment of donovanosis has largely been assessed in small-scale open trials. No randomized comparisons of therapy have been reported and the trial of streptomycin and tetracycline compared with co-trimoxazole described above (28) is one of a very small number of non-randomized comparisons published. Published treatment guidelines tend to favor antibiotics of the tetracycline group over alternatives but they do not offer explicit reasons for adopting this choice. On the basis of the published data set out in Table 2, and known adverse reaction profiles, the following antibiotics should be considered first line therapies for donovanosis: azithromycin, erythromycin, fluorinated quinolones, doxycycline. The dosage used in reported trials is set out in Table 2. In general most antibiotics are given at conventional dosage. Most clinicians continue therapy until lesions heal and some extend the treatment period further in the hope of reducing relapse. One or two weeks of therapy is often sufficient for small early lesions but therapy may need to be continued for 2-3 months for female patients with extensive pelvic infection.

Special Situations

Disseminated Infection

Hematogenous disseminated spread is a life-threatening complication of donovanosis. In a case report and review of 18 earlier published reports, 7 of the 19 cases died, including the subject of the case report (37).Among those cured, five were cured with tetracycline, two with streptomycin and one with chloramphenicol.Two authors have reported successful treatment of disseminated disease with combined streptomycin and tetracycline (13, 14). Paterson suggests the effectiveness of azithromycin in other forms of the disease make it the most promising choice for patients with disseminated infection, although its use in this rare situation has not yet been reported (37). He advocates initial daily therapy for a week followed by weekly dosing for 4-6 weeks.


Donovanosis has a tendency to extend rapidly during pregnancy and to show a diminished responsiveness to antibiotic therapy. Many of the reported cases of haematogenous dissemination have been linked to tearing during delivery of an infected cervical lesion. Cordero has described such a patient who responded well to combination treatment with streptomycin and minocycline (14). Many first line antibiotics are contraindicated in pregnancy. Erythromycin is considered safe for use in pregnancy and satisfactory results among pregnant women have been reported with erythromycin alone (20) or in combination with lincomycin (3).

Donovanosis in Patients with HIV

Jardim has described two patients with HIV infection who failed to respond to extended treatment with combinations of co-trimoxazole, tetracycline and thiamphenicol (21). This would suggest that some patients may require vigorous treatment with high dose, parenteral, combination regimens. In a recent study from South Africa 18 pregnant women with HIV and donovanosis did not differ significantly in terms of outcome when compared with women without HIV (20).

Alternative Therapy

Second line drugs that may be considered for patients failing to respond to, or intolerant of the first line drugs listed above include ampicillin, chloramphenicol, thiamphenicol, lincomycin, streptomycin, co-trimoxazole and gentamicin. Prior to the introduction of antibiotics the following treatments were used with success for treatment of donovanosis: intravenous potassium antimony tartrate and other trivalent antimonials, surgical excision of lesions, diathermic fulguration, local treatment with podophyllin, ultra-violet radiation and radiotherapy (44).

back to top


Adjunctive surgical measures are required for patients with complications such as abscess formation, fistulae, strictures and elephantiasis (36). Surgery carries the risk of disseminating active infections if carried out without antibiotic cover.

Patients with extensive malodorous ulcers benefit from the addition of penicillin to treatment regimens and by bathing ulcers in solutions of dilute potassium permanganate.


Patients can be monitored clinically by observing the healing and re-epithelialization of ulcers. Repeat smears may be made from lesions to monitor the disappearance of Donovan bodies, though this is rarely undertaken outside the context of clinical trials. Follow-up should ideally be extended up to 18 months as late relapse may occur following healing. In long-standing cases a biopsy should be done to exclude malignant change.


As donovanosis is sexually transmitted, management should always address the issues of partner management, health education and screening for other sexually transmitted infections, especially syphilis which often accompanies donovanosis. Exposed sexual partners should be offered examination and treated if lesions are found. Epidemiological treatment can be offered to asymptomatic exposed individuals who are concerned about becoming infected.


Debate still continues about the best nomenclature for the causative organism. Aragao and Vianna originally cultured a pleomorphic bacterium from ulcer lesions and identified it as C. granulomatis (2) but a number of studies suggested a link to Klebsiella species. After the advent of PCR, DNA sequencing of the 16SrRNA and phoE genes demonstrated that C. granulomatis had a greater than 99% similarity with Klebsiella pneumoniae and K. rhinoscleromatis and a proposal was put forward to reclassify the causative organism as K. granulomatis comb nov (4, 9). However, Kharsany et al performed a phylogenetic analysis of C. granulomatis based on 16S rRNA gene sequences and found that strains had similarities of only 95% and 94% respectively to the genera Klebsiella and Enterobacter (24).

back to top


1. Anderson K. The cultivation from granuloma inguinale of a microorganism having the characteristics of Donovan bodies in the yolk sac of chick embryos. Science 1943;97:560-1.

2. Aragao HD, Vianna G. Pesquizas sobre o granuloma venereo. Memn Inst Oswaldo Cruz 1913;5:211-238.

3. Ashdown LR, Kilvert GT. Granuloma inguinale in Northern Queensland. Med J Aust 1979;1:146-148 [PubMed]

4. Bastian I, Bowden FJ. Amplification of Klebsiella-like sequences from biopsy samples from patients with donovanosis. Clin Infect Dis 1996 23:1328-30. [PubMed]

5. Belda W, Velho P, Arnone M, Romitti R. Donovanosis treated with thiamphenicol. Braz J Infect Dis 2007;11: 388-9. [PubMed]

6. Beveridge WI. The action of antimony and some other bacteriostatic substances on Donovania granulomatis isolated in the chick embryo. J Immunol 1946;53:215-223. [PubMed]

7. Bowden FJ, Mein J, Plunkett C, Bastian I. Pilot study of azithromycin in the treatment of genital donovanosis. Genitourin Med 1996;72:17-19. [PubMed]

8. Breschi LC, Goldman G, Shapiro SR. Granuloma inguinale in Vietnam: successful therapy with ampicillin and lincomycin. Journal of the American Venereal Diseases Association 1975;1:118-120. [PubMed]

9. Carter JS, Bowden FJ, Bastian I, Myers GM, Sriprakash KS, Kemp DJ. Phylogenetic evidence for reclassification of Calymmatobacterium granulomatis as Klebsiella granulomatis comb. nov. Int J Systematic Bacteriol 1999;49:1695-1700. [PubMed]

10. Carter J, Hutton S, Sriprakash KS, Kemp DJ, Lum G, Savage J, Bowden FJ. Culture of the causative organism of donovanosis (Calymmatobacterium granulomatis) in HEp-2 cells. J Clin Microbiol 1997;35:2915-7. [PubMed]

11. Carter J, Kemp DJ. A colorimetric detection system for Calymmatobacterium granulomatis. Se Trans Inf 2000;76:134-6. [PubMed]

12. Chen CH, Dienst RB, Greenblatt RB. Antibiotics versus Donovania granulomatis. Am J Syph Gonorrhea Vener Dis 1951;35:383-392. [PubMed]

13. Cliff S, Wilson A, Wansborough-Jones M, Ash S. Disseminated granuloma inguinale secondary to cervical infection. J Infect 1998;36:129-30. [PubMed]

14. Cordero FA. Granuloma venereo. Su manifestacion clinica en genitales y otras partes del organismo. Med Cutan Ibero Lat Amer 1975;3:125-32. [PubMed]

15. DeMonbreun WA, Goodpasture WE. Infection of monkeys with Donovan organisms by injections of tissue from human lesions of granuloma inguinale. Am J Trop Med 1931;11:311-323.

16. Greenblatt RB, Barfield WE, Dienst RB, West RM. Terramycin in the treatment of granuloma inguinale. J Vener Dis Info 1951;32:113-15. [PubMed]

17. Greenblatt RB, Kupperman HS, Dienst RB. Streptomycin in the therapy of granuloma inguinale. Proc Soc Exp Biol Med 1947;56:1-6. [PubMed]

18. Greenblatt RB, Wammock VS, Dienst RB, West RM. Chloromycetin in the therapy of granuloma inguinale. Am J Obstet Gynecol 1950;59:1129-1133. [PubMed]

19. Harb FW, Simpson WG, Wood CE. Intramuscular injections of chloromycetin in the treatment of granuloma inguinale. J Vener Dis Info 1951;32:177-181. [PubMed]

20. Hoosen AA, Mphatsoe M, Kharsany AB, Moodley J, Bassa A, Bramdev A. Granuloma inguinale in association with pregnancy and HIV infection. Int J Gynaecol Obstet 1996;53:133-8. [PubMed]

21. Jardim ML, Barros ER, Silveira M. Donovanose em pacientes portadores de AIDS. Relato de dois casos. Anais Brasileiros de Dermatologia e Sifilografia 1990;65:175-7.

22. Johnson SH, Cherubin C. Case report. Infectious Diseases Newsletter 1991;10:14-15.

23. Kharsany AB, Hoosen AA, Kiepiela P, Naicker T, Sturm AW. Culture of Calymmatobacterium granulomatis. Clin Infect Dis. 1996;22:391. [PubMed]

24. Kharsany AB, Hoosen AA, Kiepiela P, Kirby R, Sturm AW. Phylogenetic analysis of Calymmatobacterium granulomatis based on 16S rRNA gene sequences. J Med Microbiol. 1999;48:841-7. [PubMed]

25. Kupperman HS, Greenblatt RB, Dienst RB. Streptomycin in the therapy of granuloma inguinale. JAMA 1948:136:84-89.

26. Lal S, Garg BR. Further evidence of the efficacy of co-trimoxazole in granuloma venereum. British Journal of Venereal Diseases 1980;56:412-413. [PubMed]

27. Lal S. Continued efficacy of streptomycin in the treatment of granuloma inguinale. British Journal of Venereal Diseases 1971;47:454-455. [PubMed]

28. Latif AS, Mason PR, Paraiwa E. The treatment of donovanosis (granuloma inguinale). Sex Transm Dis 1988;15:27-29. [PubMed]

29. Mackay IM, Harnett G, Jeoffreys N, Bastian I, Sriprakash KS, Siebert D, Sloots TP. Detection and discrimination of herpes simplex viruses, Haemophilus ducreyi, Treponema pallidum, and Calymmatobacterium (Klebsiella) granulomatis from genital ulcers. Clin Infect Dis 2006;42:1431-8. [PubMed]

30. Maddocks I, Anders EM, Dennis E. Donovanosis in Papua New Guinea. British Journal of Venereal Diseases 1976;52:190-196. [PubMed]

31. Mein J, Bastian I, Guthridge S, Farmer B, Bowden F. Donovanosis: sequelae of severe disease and successful azithromycin treatment. Int J STD AIDS 1996;7:448-51. [PubMed]

32. Merianos A, Gilles M, Chuah J. Ceftriaxone in the treatment of chronic donovanosis in central Australia. Genitourin Med 1994;70:84-9. [PubMed]

33. O’Farrell N, Hoosen AA, Coetzee K, van den Ende J. A rapid staining technique for the diagnosis of granuloma ingunale (donovanosis). Genitourin Med 1990;66:200-201. [PubMed]

34.O’Farrell N, Windsor I, Becker P. Risk factors for HIV-1 in heterosexual attenders at a sexually transmitted diseases clinic in Durban. S Afr Med J 1991;80:17-20. [PubMed]

35. Pariser RJ. Tetracycline-resistant granuloma inguinale. Arch Dermatol 1977;113:988. [PubMed]

36. Parkash S, Radhakrishna K. Problematic ulcerative lesions in sexually transmitted diseases: surgical management. Sex Transm Dis 1986;13:127-133. [PubMed]

37. Paterson DL. Disseminated donovanosis (granuloma inguinale) causing spinal cord compression: case report and review of donovanosis involving bone. Clin Infect Dis 1998;26:379-83. [PubMed]

38. Pradinaud R, Grosshans E, Basset A, Bertin C. 24 cases of donovanosis in French Guiana Bull Soc Pathol Exot Filiales 1981;74:30-36. [PubMed]

39. Rake G, Dunham W. Action of disinfectant, chemotherapeutic, and antibiotic agents on the organism of granuloma inguinale. American Journal of Syphilis 1947;31:610-617.

40. Rake G. The antigenic relationships of Donovania granulomatis(Anderson) and the significance of this organism in granuloma inguinale. Am J Syph Gonorrhea Vener Dis 1948;32:150-158.

41. Rajam RV, Rangiah PN. Further observations on streptomycin therapy in venereal granuloma. Indian J Venereal Diseases and Dermatology 1952;18:1-8.

42. Rajam RV, Rangiah PN, Sowmini CN, Krishnamoorthy N. Tetracyclene (Achromycin) in venereal diseases. The Antiseptic 1956;53:9-19.

43. Ramanan C, Sarma PS, Ghorpade A, Das M. Treatment of donovanosis with norfloxacin. Int J Dermatol 1990;29:298-299. [PubMed]

44. Richens J. The diagnosis and treatment of donovanosis (granuloma inguinale). Genitourin Med 1991;32:441-452. [PubMed]

45. Robinson HM Jr, Cohen MM. Treatment of granuloma inguinale with erythromycin. J Invest Dermatol;1953;20:407-409. [PubMed]

46. Rosen T, Tschen JA, Ramsdell W, Moore J, Markham B. Granuloma inguinale. J Am Acad Dermatol 1984;11:433-437. [PubMed]

47. Shapiro SR, Breschi LC. Venereal disease in Vietnam: clinical experience at a major military hospital. Milit Med 1974;139:374-9. [PubMed]

48. Velasco JE, Miller AE, Zaias N. Minocycline in the treatment of venereal disease. JAMA 1972;220:1323-1325. [PubMed]

back to top


Table 1. Bactericidal Concentrations of Antibiotics Against C. granulomatis Reported by Chen (1951) (12).

Antibiotic Bactericidal concentration per ml of medium
Streptomycin 0.2 mg
Penicillin G 1 unit
Chlortetracycline 5 mg
Chloramphenicol 10 mg

From Chen CH, Dienst RB, Greenblatt RB. Antibiotics versus Donovania granulomatis. Am J Syphilis 1951:35:383-392.

Table 2. Key Trials of Antibiotics for Treatment of Granuloma Inguinale

Drug Dose Route Duration M:F Total patients Smear positive Relapse Days to smear negative Days to heal Ref
Ampicillin 500 mg qid O 2-4wk 31 67% 8
Chloramphenicol 500mg-1g 6h O 5-20d 0/23 23 2 2-4 days 18
2-4g every 3-4 days IM 8-12d 20/23 43 43 5 1-7 days 7-30d 19
500mg qid O 10-55d 50 1 av. 7 days av.17d 30
Thiamphenicol 2.5g stat, then 1Gram/day O 2 wk 10/0 10 10 2 15 days 15-30d 5
Co-trimoxazole 160/800mg, bid O Mean 12.5d 84/32 116 116 2 5 days 7-22d 26
160/800mg, bid O 20-49d 19/1 20 16 2-5wk 46
160/800mg, bid O 14d 12/6 18 18 7-21d 28
Erythromycin 1-300mg 6h O 16-37d 5/4 9 9 1 9-30d 45
Lincomycin 500mg qid O 14d 5 8
Streptomycin 0.3-1g 4h IM 6-46d 23 23 3 5-9 days 17
0.3-4g/d IM 5-62d 32/16 48 48 3 2-11 days av. 3wk 25
1g bid IM 3-70d 146/81 227 219 1 av. 6 days 41
1g bid IM av.12.5d 83/39 122 27
Gentamicin 1mg/kg tds IM mean 22d 9 mean 12 days mean 22d 30
Minocycline 200mg stat, then 100mg bid O 15-45d 5/1 6 6 15-90d 48
Oxytetracycline 2-3g/d O 2-29d 32 32 2 4-5 days 16
Tetracycline hydrochloride 1g/d O 20d 14/6 20 20 5-7 days 25-45d 42
Norfloxacin 400mg bid O 2-11d 10/0 10 10 8 by 3 days mean 7.3d 43
Ceftriaxone 1g od IM/IV 10d 4/8 12 8 5 32
Azithromycin 1g per week or 500mg daily O 1-4 wk 4/7 11 9 0 7
Antibiotic Combinations
Streptomycin with 1g/d IM 14d 7/6 13 13 7-21d 28
tetracycline 500mg 6h O 14d
Erythromycin with 500mg qid O 10-14d 0/4 4 4 "rapid" 3
Lincomycin 500mg qid O 10-14d


Klebsiella granulomatis