Background and Objectives:: To study Haemophilus ducreyi pathogenesis, the authors developed an experimental model of infection in human volunteers. The authors analyze their cumulative experience with strain 35000 in the model and calculate the papule and pustule formation rates for estimated delivered doses (EDDs) ranging from 1 cfu to 100 cfu.
Study Design:: Sixty‐five volunteers were included in the analysis. A total of 139 sites were available for calculation of the papule formation rate, and 117 sites were available for calculation of the pustule formation rates.
Results:: The effect of EDDs and probabilities of papule formation and the pustule formation were dose‐dependent. Increasing the EDD resulted in a higher probability of papule and pustule formation.
Conclusion:: H ducreyi is highly infectious for humans. Inoculation of an EDD of 1 cfu causes a papule formation rate of 50%. Pustule formation rates are approximately 50% for 27 cfu and 90% for 100 cfu.
From the Departments of *Medicine and †Microbiology, Immunology, and Pathology and Laboratory Medicine, Indiana University, Indianapolis, Indiana
Correspondence and reprint requests: Stanley M. Spinola, MD, 545 Barnhill Drive, Emerson Hall Rm 435, Indianapolis, IN 46202.
Received for publication April 28, 1999, revised August 4, 1999, and accepted August 6, 1999.
HAEMOPHILUS ducreyi is the etiologic agent of chancroid, which facilitates transmission of HIV.1–4 Lacking specimens from naturally infected patients, we developed an experimental model of infection of the skin of human volunteers.5–11 Papules usually develop within 24 hours of inoculation, and may either resolve or progress to pustules. The histopathology of experimental papules and pustules is similar to naturally occurring ulcers, and consists of T‐cells, macrophages, and polymorphonuclear leukocytes.8,12 Lesion outcome (i.e, papule, pustule, spontaneous resolution) at multiple sites inoculated with the same dose of one isolate in an individual subject appears to be independent7; thus, site rather than subject is used as the unit of measurement for calculation of papule and pustule formation rates.
In a standardization study, inoculation of a mean estimated delivered dose (EDD) of 27.5 cfu + 9.8 cfu of H ducreyi 35000 human passaged (HP) caused papule formation at 95% of 40 sites inoculated with live bacteria (95% CI, 83.1%‐99.4%).7 The pustule formation rate of 24 sites allowed to progress to clinical outcome was 69% (95% CI, 47.1%‐86.6%).7 Our previous studies did not have sufficient numbers to determine papule or pustule formation rates outside of this narrow dose range.
In addition to dose‐response6,10 and standardization7 trials, we have recently completed a reinfection trial11 and several mutant and parent comparisons in the 35000 background9,13–15; thus, more sites inoculated with 35000 or 35000 HP are available for analysis. In this study, we report our cumulative experience with 35000 in the human model to estimate the probabilities of papule and pustule formation rates based on EDDs.
Materials and Methods
Human Volunteers and Infection Protocol
Informed consent was obtained from the participants in accordance with the human experimentation guidelines of the US Department of Health and Human Services and the Institutional Review Board of Indiana University‐Purdue University at Indianapolis. The experimental challenge protocol, preparation and inoculation of the bacteria, and clinical observations were done as described previously.5–7,9 To infect volunteers, bacterial suspensions were loaded onto a Multi‐Test applicator (Lincoln Diagnostics, Decatur, IL) and pressed into the skin of the upper arm. The device delivers approximately 1/1000 of the volume of solutions loaded on its tines into the epidermis and dermis.6,16 We did not experimentally determine the actual delivered dose; however, the EDD is likely to be 1000‐fold less than the average cfu loaded on the tines. Volunteers were observed until (1) they reached a clinical endpoint, defined as 14 days after inoculation; (2) the development of a painful pustule; or (3) the resolution of infection at all sites.
Sixty‐five participants (24 men, 41 women; mean age, 33.5 years + 9.2 years SD) were included in the analysis. Each volunteer was inoculated at 2 or 3 sites with H ducreyi 35000 (n = 18 subjects at 45 sites) or 35000 HP (n = 47 subjects at 94 sites). Only parent sites were included for subjects who participated in the mutant‐parent comparisons. Subjects who participated in the initial dose‐response trial5 and those who participated in a chemoprophylaxis trial10 were excluded from the analysis because of a short duration of infection (1‐3 days). For volunteers who participated in a reinfection trial11 and who were challenged twice, only data from their initial infection were included.
The aim of the study was to estimate the probability of papule and pustule formation based on the EDD of H ducreyi. Logistic regression modeling was used to predict the papule and pustule formation rates. The quality of the fit for each model was tested using the le Cessie‐van Houwelingen goodness of fit test.17 A nonsignificant P value (i.e., P > 0.05) indicates a model that fits well. In both analyses, we also calculated the EDD required for specified papule formation rates and the associated 95% CI. All analyses were conducted using Splus version 4.5 for Windows (Mathsoft Inc, Seattle, WA).
Papule Formation Rates
A total of 139 inoculation sites were available for calculation of the papule formation rates. The effect of EDDs and probabilities of papule formation was dose‐dependent (Figure 1). The le Cessie‐van Houwelingen goodness of fit test yielded a P > 0.8; therefore, the logistic regression model is appropriate in the analysis of papule formation rates. An EDD of 0.2 cfu (95% CI, 0‐10) gave a probability of papule formation of 0.5. An EDD of 10 cfu (95% CI, 3.9‐17.4) gave a probability of 0.7, whereas an EDD of 27.4 cfu (95% CI, 18.6‐36.2) gave a probability of papule formation of 0.9.
Pustule Formation Rates
For calculation of pustule formation rates, we included inoculation sites that achieved a definite outcome (i.e., resolution or pustule formation). Sites that did not achieve a definite outcome before treatment of subjects were excluded from analysis; thus, a total of 117 sites were available for calculation of the probabilities of pustule formation. The le Cessie‐van Houwelingen goodness of fit test yielded a P > 0.6; therefore, the logistic regression model is appropriate in the analysis of pustule data. The effect of EDDs and pustule formation was dose‐dependent (Figure 2). An EDD of 27 cfu (95% CI, 14‐40) gave a probability of pustule formation of 0.5. An EDD of 55 cfu (95% CI, 34.9‐75.3) gave a probability of pustule formation of 0.7, whereas an EDD of 99.7 cfu (95% CI, 55‐144) gave a probability of 0.9.
Papule and pustule formation rates were compared for 35000 and 35000 HP. For 35000 HP, an EDD of 0.85 cfu (95% CI, 0‐19.9) gave a probability of papule formation of 0.8, and 14.5 cfu (95% CI, 0‐30.3) gave a probability of 0.9. For 35000, an EDD of 34.4 cfu (95% CI, 18.8‐51.1) gave a probability of papule formation of 0.8 and 48.6 cfu (95% CI, 26.5‐70.7) gave a probability of 0.9. However, there were no differences in the pustule formation rates between the two strains over the dose range studied (data not shown). There were also no differences in the papule and pustule formation rates between black and white participants (data not shown).
In studies of rabbits housed at room temperature, loss of virulence was observed after in vitro passage of H ducreyi by some investigators but not others.18,19 Subsequent studies indicated that rabbits must be housed at reduced ambient temperatures to support bacterial replication20,21; thus, it is unclear whether a correlation exists between in vitro passage and the virulence of H ducreyi. We do not know the passage history of our stock of 35000; H ducreyi 35000 and 35000 HP have similar outer‐membrane and lipooligosac‐charide profiles and growth rates.7 Inoculation of 35000 HP resulted in a higher papule formation rate than 35000, but the pustule formation rates for the two strains were similar. Papule formation is more difficult to assess objectively than pustule formation in the human model. The apparent differences in papule formation rates between the two strains should be interpreted with caution, especially because the strains were not compared simultaneously.
The major conclusion of this analysis is that the effect of EDDs and probabilities of papule and pustule formations were dose‐dependent. H ducreyi clumps in vivo and in vitro18; organisms grown in broth clump less than those grown on plates.22 Although we grew all inocula in broth, the EDDs reported represent minimal estimates of the number of bacteria in a given dose. The infectious dose of H ducreyi in natural disease is not known. In this model, inoculation of as few as 30 cfu results in papule formation 90% of the time. This infectious dose of H ducreyi is consistent with the 70% transmission rate reported after single sexual exposures.23
Our previous work based on 24 sites showed that inoculation of 27 cfu resulted in a pustule formation rate of approximately 70%. However, our cumulative experience showed that inoculation of 27 cfu resulted in a pustule formation rate of 50%, and that inoculation of approximately 55 cfu is necessary to achieve a 70% pustule formation rate. To date, pustules have never resolved during the 14‐day observation period.7 Pustules frequently contain microulcerations in the epidermis and may represent a commitment to progression to the ulcerative stage of disease. For vaccine trials or mutant‐parent comparisons, change in pustule formation rate will be the measure of efficacy or virulence. Inoculation of EDDs in the range of 50‐100 cfu should be sufficient to cause pustule formation and to assess vaccine efficacy or alterations in virulence due to mutations.
1. Wasserheit JN. Epidemiological synergy: interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Transm Dis 1992; 19:61–77.
2. Jessamine PG, Ronald AR. Chancroid and the role of genital ulcer disease in the spread of human retrovirus. Med Clin North Am 1990; 74:1417–1431.
3. Telzak EE, Chiasson MA, Bevier PJ, Stoneburner RL, Castro KG, Jaffe HW. HIV-1 seroconversion in patients with and without genital ulcer disease. Ann Intern Med 1993; 119:1181–1186.
4. Kreiss JK, Coombs R, Plummer F, et al. Isolation of human immunodeficiency virus from genital ulcers in Nairobi prostitutes. J Infect Dis 1989; 160:380–384.
5. Spinola SM, Wild LM, Apicella MA, Gaspari AA, Campagnari AA. Experimental human infection with Haemophilus ducreyi.
J Infect Dis 1994; 169:1146–1150.
6. Spinola SM, Orazi A, Arno JN, et al. Haemophilus ducreyi
elicits a cutaneous infiltrate of CD4 cells during experimental human infection. J Infect Dis 1996; 173:394–402.
7. Al-Tawfiq JA, Thornton AC, Katz BP, et al. Standardization of the experimental model of Haemophilus ducreyi
infection in human subjects. J Infect Dis 1998; 178:1684–1687.
8. Palmer KL, Schnizlein-Bick CT, Orazi A, et al. The immune response to Haemophilus ducreyi
resembles a delayed-type hypersensitivity reaction throughout experimental infection of human subjects. J Infect Dis 1998; 178:1688–1697.
9. Palmer KL, Thornton AC, Fortney KA, Hood AF, Munson RS Jr, Spinola SM. Evaluation of an isogenic hemolysin-deficient mutant in the human model of Haemophilus ducreyi
infection. J Infect Dis 1998; 178:191–199.
10. Thornton AC, O'Mara EM Jr, Sorensen SJ, et al. Prevention of experimental Haemophilus ducreyi
infection: a randomized, controlled clinical trial. J Infect Dis 1998; 177:1608–1613.
11. Al-Tawfiq JA, Palmer KL, Chen C-Y, et al. Experimental infection of human volunteers with Haemophilus ducreyi
does not confer protection against subsequent challenge. J Infect Dis 1999; 179:1283–1287.
12. King R, Gough A, Nasio J, Ndinya-Achola F, Plummer F, Wilkins J. An immunohistochemical analysis of naturally occurring chancroid. J Infect Dis 1996; 174:427–430.
13. Young RS, Fortney KR, Haley JC, et al. Expression of sialyated or paragloboside-like lipooligosaccharides are not required for pustule formation by Haemophilus ducreyi
in human volunteers. Infect Immun 1999; 67:6335–6340.
14. Al-Tawfiq JA, Bauer ME, Fortney KR, et al. A pilus-deficient mutant of Haemophilus ducreyi
is virulent in the human model of experimental infection. J Infect Dis 2000 (in press).
15. Al-Tawfiq JA, Fortney KR, Katz BP, Hood AF, Elkins C, Spinola SM. An isogenic hemoglobin receptor-deficient mutant of Haemophilus ducreyi
is attenuated in the human model of experimental infection. J Infect Dis 2000 (in press).
16. Hobbs MM, SanMateo LR, Orndorff PE, Almond G, Kawula TH. Swine model of Haemophilus ducreyi
infection. Infect Immun 1995; 63:3094–3100.
17. Le Cessie S, Van Houwelingen JC. A goodness-of-fit test for binary test for binary regression models based on smoothing methods. Biometrics 1991; 47:1267–1282.
18. Morse SA. Chancroid and Haemophilus ducreyi.
Clin Microbiol Rev 1989; 2:137–157.
19. Hammond GW, Lian CJ, Wilt JC, Ronald AR. Antimicrobial susceptibility of Haemophilus ducreyi.
Antimicrob Agents Chemother 1978; 13:608–612.
20. Purcell BK, Richardson JA, Radolf JD, Hansen EJ. A temperature-dependent rabbit model for production of dermal lesions by Haemophilus ducreyi.
J Infect Dis 1991; 164:359–367.
21. Campagnari AA, Wild LM, Griffiths GE, Karalus RJ, Wirth MA, Spinola SM. Role of lipooligosaccharides in experimental dermal lesions caused by Haemophilus ducreyi.
Infect Immun 1991; 59:2601–2608.
22. Totten PA, Lara JC, Norn DV, Stamm WE. Haemophilus ducreyi
attaches to and invades human epithelial cells in vitro. Infect Immun 1994; 62:5632–5640.
23. Plummer FA, Nsanze H, Karasira P, D'Costa LJ, Dylewski J, Ronald AR. Epidemiology of chancroid and Haemophilus ducreyi
in Nairobi, Kenya. Lancet 1983; ii:1293–1295.