Secondary Logo

Journal Logo

Epidemiology and Social: Original Papers

Evidence of exogenous reinfection and mixed infection with more than one strain of Mycobacterium tuberculosis among Spanish HIV-infected inmates

Chaves , Fernando *; Dronda , Fernando ; Alonso-Sanz , Mercedes ; Noriega , Antonio R.*

Author Information
  • Free



After primary infection, progressive tuberculosis may develop either as a continuation of primary infection or as an endogenous reactivation of a latent focus. In some patients, however, disease may also result from exogenous reinfection by a second strain of Mycobacterium tuberculosis. There are reports of exogenous reinfection in the literature, both in immunosuppressed and immunocompetent individuals. These observations are based on clinical and epidemiological data [1-8], some supported by mycobacteriophage typing of strains [1,2,4] and, in more recent studies, employing DNA fingerprinting methods [5-8]. While most cases were associated with drug-resistant strains of M. tuberculosis [3,5-8], some exogenous reinfections by drug-sensitive organisms have also been reported [2]. Mixed-strain infection with two different strains of M. tuberculosis have also been reported [9,10].

The introduction of DNA fingerprinting of M. tuberculosis isolates by restriction fragment length polymorphism (RFLP) analysis with the insertion sequence IS6110 [11] has been a great advance in the study of the epidemiology of tuberculosis [12]. RFLP analysis permit reliable discrimination of strains derived from different clonal populations on the basis of the number and position of bands that hybridize to IS6110. This technique has a high power of discrimination, superior to that of mycobacteriophage typing [13].

Prisons are recognized as high-risk settings for tuberculosis transmission [14-16]. The high rates of HIV infections and the close contact among prisoners, together with the frequent movement of inmates within the prison system, contribute to increase the transmission of the disease. Prisoners are clearly at high risk of exposure to multiple sources of tuberculosis infection and, as a result, may develop disease with more than one strain of M. tuberculosis, particularly HIV-infected inmates. In order to evaluate the importance of this phenomenon, we retrospectively reviewed the clinical and microbiologic records of HIV-infected inmates with tuberculosis in the Madrid prison system between 1993 and 1994. From the 226 infected patients, we selected for further microbiologic studies two groups with the following characteristics. The first group comprised patients who remained culture positive 4 months or more after initial diagnosis and the second group included patients with positive M. tuberculosis culture from more than one anatomical site at the same time. In all the isolates from these two groups we did a IS6110-based RFLP analysis to compare the identity of the strains.


Patient population and data collection

At the time of this study, the prison system of Madrid consisted of seven prisons and a general penitentiary hospital, with a total of 7000 prisoners. The tuberculosis case rate was 2.3% per year and 84% of these were HIV positive. A molecular epidemiologic study of tuberculosis in the prisons of Madrid, which included 18 months of the time period in question, showed that 62% of the prisoners with tuberculosis were recently infected [16]. All clinical specimens for diagnosis of tuberculosis were sent to a central microbiology laboratory. Included in the present study were 226 HIV-positive inmates with culture-positive tuberculosis attended between January 1, 1993 and December 31, 1994. A retrospective investigation of clinical records was conducted to identify two subpopulations within this group of patients. The first consisted of patients who remained culture positive 4 or more months after initial diagnosis and the starting of the treatment. The second subgroup included all those who at the same time had culture-positive specimens from more than one anatomical site. A standardized protocol was used to identify demographic characteristics, signs and symptoms at the time of diagnosis, laboratory and radiological data, and treatment outcome. As a general rule, patients were initially treated daily with isoniazid, rifampin (rifampicin), ethambutol and pyrazinamide at standard doses for 2 months and then with isoniazid and rifampin daily for 7 months. At the time of this study, directly observed therapy was only administered while the patients were hospitalized. After discharge from the prison hospital and return to the prison, the medications were given to each patient on a daily basis and adherence to therapy for each patient was a judgment made by the health-care provider. An epidemiologic investigation was carried out for selected patients. In patients 1A, 2A and 3B we compared the DNA patterns of the isolates obtained with a database of strains of M. tuberculosis isolated in the Madrid prisons between 1993 and 1994 [16]. The records of all patients from the database with identical IS6110 fingerprint patterns to these three patients were reviewed. For these patients, it was determined exactly where each was confined at the time of diagnosis, and their movement within the prisons was tracked throughout the period extending from as far back as 2 year before the diagnosis to the beginning of chemotherapy. Our three patients were considered linked epidemiologically to other prisoners if there was evidence in the records that they were in the same prison at a time when one or more of the individuals were likely to have been infectious.

Microbiology and RFLP analysis

Clinical specimens were processed according to standard methods and inoculated in Lowenstein-Jensen and Coletsos slants (BioMerieux, France) [17]. Drug susceptibility testing for isoniazid, rifampin, streptomycin and ethambutol was performed by the proportion method. Isolates of M. tuberculosis from the two groups of patients (above) were stored at -70°C. DNA fingerprinting with the insertion sequence IS6110 was performed as described previously [11]. The computer-assisted analysis of the IS6110 fingerprints was performed by using GelCompar (Applied Maths, Kortrijk, Belgium). Patients were considered to be infected with a single strain of M. tuberculosis when isolates were identical with respect to the number and size of hybridizing bands. In selected strains, a secondary DNA typing was performed, digesting with AluI and using a polymorphic GC-rich repetitive sequence (PGRS) contained in the recombinant plasmid pTBN12 as a probe [18]. To exclude cross-contamination in the laboratory, we reviewed the microbiologic samples of patients with positive cultures processed in the same day, comparing fingerprinting patterns.


Patients who remained culture positive

Eleven patients remained culture positive. Initial and second isolates from nine HIV-infected patients were available for DNA fingerprinting. We never found isolates handled on the same day with the same RFLP patterns. We interpreted this as evidence excluding the possibility of laboratory cross-contamination. The average interval between the initial and the second isolate was 8 months (range: 4-18 months). The CD4 cell count of these patients is shown in Table 1. In seven, the DNA fingerprint patterns were identical in both isolates in spite of some phenotipical variation in the antibiotic susceptibility patterns (Table 1). In the remaining two patients (22%), who are described below, the isolates obtained 18 and 4 months after the initial isolates had different RFLP patterns with the insertion sequence IS6110 and the recombinant plasmid pTBN12.

Table 1:
Epidemiologic and clinical characteristics of nine HIV-infected inmates who were culture positive after 4 or more months of treatment.

Patient 1A

A 28-year-old male who was known to be HIV positive since 1987 entered prison in February 1992 and was diagnosed in January 1993 as having genitourinary tuberculosis. The initial isolate was fully susceptible to first-line drugs and the patient responded to chemotherapy, becoming culture negative. He finished his 9-month treatment by October of 1993 and he was considered cured. In June 1994, M. tuberculosis was again isolated, this time from sputum and blood. This isolate was resistant to isoniazid and rifampin. The IS6110 DNA fingerprint (Fig. 1, lanes 2 and 3) and the pTBN12 pattern of both isolates was clearly different. These two fingerprints were compared with a database of 210 others strains obtained from M. tuberculosis isolated in the Madrid prisons between 1993 and 1994 [16]. The urine isolate had RFLP pattern identical to that of isolates from 10 other patients. The fingerprint of the isolates from the second episode of disease was identical to that of five other patients diagnosed during the same period. One of these was linked epidemiologically to patient 1A. These findings suggest that, while in prison, this patient was exogenously reinfected with a multidrug-resistant strain of M. tuberculosis.

Fig. 1:
IS6110 fingerprint patterns of the clinical isolates from patients with two different M. tuberculosis strains. Lanes 1 and 6: molecular-weight markers of control strain M. tuberculosis MT14323. Lanes 2 and 3: DNA fingerprints corresponding to patient 1A (lane 2, urine isolate; lane 3, sputum isolate). Lanes 4 and 5: DNA fingerprints corresponding to patient 2A (lane 4, sputum isolate; lane 5, blood isolate). Lanes 7-9: DNA fingerprints corresponding to patient 3B (lane 7, sputum isolate; lane 8, urine isolate; lane 9, urine isolate obtained 2 days later).

Patient 2A

A 29-year-old male with HIV infection entered prison in February 1993 and 6 months later was diagnosed as having disseminated tuberculosis, with positive sputum and blood cultures. He showed good initial clinical response after 2 months of supervised therapy while in the prison hospital. After discharge, he did not receive supervised therapy. In December 1993, his clinical condition worsened and he was again diagnosed as having disseminated tuberculosis with a positive blood culture. Clinical isolates from August and December were both fully susceptible to the first-line drugs. The patient was put again under supervised therapy and had a good clinical and microbiologic response. RFLP analysis of isolates of M. tuberculosis from the two episodes demonstrated two different fingerprint patterns (Fig. 1, lanes 4 and 5). The isolates had also different pTBN12 patterns. The IS6110 RFLP pattern of the initial isolates was unique among those of the 1993-1994 database [16]. However, the second isolate had a RFLP pattern identical to another isolate from a patient diagnosed with tuberculosis in November 1993. Epidemiologic investigation could not prove direct contact between the two patients but did reveal that both had previously been in the same prison, although not at the same time. We believe this case to be an example of exogenous reinfection with a drug-sensitive strain from an inmate with unrecognized tuberculosis.

Patients with simultaneous isolation of M. tuberculosis from two or more anatomical sites

During the study period, 77 patients had multiple isolates of M. tuberculosis from a variety of anatomical sites. All of them were coinfected with HIV and 28 had two or more isolates available for DNA fingerprinting. We never found isolates handled on the same day with the same RFLP patterns. We interpreted this as evidence excluding laboratory cross-contamination. The CD4 cell count at the time of diagnosis was known in 25 patients, and in 21 (84%) it was less than 200×106 cells/l. Seven (25%) gave a history of previous treatment of tuberculosis. The sites from which the 61 isolates of M. tuberculosis that were included in this study were recovered are listed in Table 2. RFLP analysis showed that isolates from 25 patients shared an identical fingerprint pattern, leading to the conclusion that they were infected by a common strain. Two patients had isolates that differed in a single IS6110 hybridizing band. In one patient, the isolates were from sputum and blood and in the other from sputum and lymph node. Secondary fingerprinting analysis with pTBN12 showed identical patterns in the isolates of these two patients. Therefore, we did not consider these isolates to represent different strains but merely variants of a single clone. However, a third patient, who is described below, had two isolates of M. tuberculosis that showed two distinctly different RFLP patterns (Fig. 1, lanes 7, 8 and 9). These isolates showed also different pTBN12 fingerprint patterns.

Table 2:
Mycobacterium tuberculosis isolates obtained from two or more sites in 28 patients.

Patient 3B

A 35-year-old HIV-infected male entered prison in September 1991 and in March 1993 he was hospitalized with fever, productive cough and cachexia. A chest radiograph showed hilar lymph nodes and small infiltrates in the upper lobes. Three sputum samples obtained at this time were smear negative. The patient started antituberculous treatment but died 3 days later. Three weeks later, cultures of sputum, urine and blood were positive for M. tuberculosis. Susceptibility testing of the sputum and urine isolates showed it to be sensitive to first-line drugs. RFLP analysis was performed on three available isolates, one from sputum and two from urine. Different RFLP patterns were obtained from the sputum and urine cultures with IS6110 and pTBN12. These two fingerprints were compared with the 1993-1994 database [16]. The strain isolated from sputum had an indistinguishable DNA fingerprint from that of another 20 isolates obtained from 20 inmates, and the strain isolated from urine had an indistinguishable RFLP pattern from that of another 23 isolates. These two patterns were among the most frequently encountered fingerprints in the Madrid prison system between 1993 and 1994 [16]. It is hard to establish precise epidemiologic links for these patients; however, we have evidence that this patient was an inmate 3 months prior to his diagnosis in the same prison as the other three patients with an identical strain to the one isolated in the sputum and with one patient with the same strain as the one isolated in urine. Therefore, we believe that this clinical case represents an example of mixed-strain infection with two drug-sensitive strains of M. tuberculosis.


This study provides evidence that HIV-infected inmates living under conditions of high environmental infectivity can be infected with more than one strain of M. tuberculosis either at the same time or in successive episodes of disease. Although this possibility has previously been considered [19,20], there are no reports that confirm it in prisoners. The few documented cases reported so far reflect diverse epidemiologic contexts. In 1975, Mankiewicz and Liivak used mycobacteriophage typing of cultures of M. tuberculosis isolated from Eskimo patients and found that 14% had mixed phage types [21]. They suggested exogenous reinfection as an explanation. In 1980, Ormerod and Skinner used drug-resistance markers in the family of a patient with drug-resistant tuberculosis and reported two patients with exogenous reinfection [3]. In 1986, Nardel and co-workers using drug resistance markers and phage typing identified seven patients with reinfection during an outbreak of tuberculosis in a shelter for the homeless [4]. Recently, molecular techniques have been used to document exogenous reinfection with multidrug- resistant M. tuberculosis strains in immunosuppressed patients with HIV infection [5,6]. Shafer and co- workers described exogenous reinfection in an alcoholic woman [7], and Turett and co-workers described exogenous reinfection in a diabetic patient [8]; both patients were reinfected with a multidrug-resistant strain. Double infection in the same patient with two different M. tuberculosis strains, one drug sensitive and one multidrug resistant, has been confirmed in an immunocompetent patient [10].

As the design of our study includes only an small subpopulation of inmates, we cannot extract conclusions regarding the real importance of this phenomenon in these settings; however, we are inclined to think that this problem is probably sufficiently frequent and important to deserve serious attention from all the health professionals responsible for these highly vulnerable populations.

In summary, this study has two conclusions. First, in HIV-infected patients tuberculous reinfection occurs, causing several consecutive episodes with different strains, not necessarily indicating acquired resistance. Second, mixed tuberculous strains may be present in HIV-infected patients. These results have implications both for clinical management of individual patients and for tuberculosis control programs as a whole. From a clinical point of view, a patient with evidence of tuberculous relapse should be evaluated for the possibility of exogenous reinfection with a drug-resistant strain, or with a drug-susceptible strain in cases of irregularities in the treatment. For tuberculosis control programs, these findings also argue against placing more than one patient with tuberculosis in one room. Immunosuppressed patients who reside in atmospheres of high infectivity, whether or not they have been treated for tuberculosis in the past, are candidates for chemoprophylaxis. Infection prevention is an important tool in breaking the chain of transmission. It is essential that correctional facilities comply fully with the recommendations for prevention and control of tuberculosis [22].


The authors thank Drs Donald Cave, Tobin Helyer, Carlos Lumbreras and Jose M. Aguado for their suggestions and comments.


1. Raleigh JW, Wichelhausen R. Exogenous reinfection with Mycobacterium tuberculosis confirmed by phage typing. Am Rev Respir Dis 1973, 108: 639-642.
2. Raleigh JW, Wichelhausen R, Rado TA, Bates JH. Evidence for infection by two distinct strains of Mycobacterium tuberculosis in pulmonary tuberculosis: report of 9 cases. Am Rev Respir Dis 1975, 112: 497-503.
3. Ormerod P, Skinner C. Reinfection tuberculosis: two cases in the family of a patient with drug-resistant disease. Thorax 1980, 35: 56-59.
4. Nardell E, McInnis B, Thomas B, Weidhaas S. Exogenous reinfection with tuberculosis in a shelter for the homeless. N Engl J Med 1986, 315:1570-1575.
5. Small PM, Shafer RW, Hopewell PC, et al. Exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis in patients with advanced HIV infection. N Engl J Med 1993, 328: 1137-1144.
6. Horn DL, Hewlett D, Haas WH, et al. Superinfection with rifampin-isoniazid-streptomycin-ethambutol (RISE)-resistant tuberculosis in three patients with AIDS: confirmation by polymerase chain reaction fingerprinting. Ann Intern Med 1994, 121:115-116.
7. Shafer RW, Singh SP, Larkin C, Small PM. Exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis in an immunocompetent patient. Tubercle Lung Dis 1995, 76: 575-577.
8. Turett GS, Fazal BA, Justman JE, et al. Exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis. Clin Infect Dis 1997, 24: 513-514.
9. Bates JH, Stead WW, Rado TA. Phage type of tubercle bacilli isolated from patients with two or more sites of organ involvement. Am Rev Respir Dis 1976, 114: 353-358.
10. Theisen A, Reichel C, Rusch-Gerdes S et al. Mixed-strain infection with a drug-sensitive and multidrug-resistant strain of Mycobacterium tuberculosis. Lancet 1995; 345:1512-1513.
11. van Embden JDA, Cave MD, Crawford JT, et al. Strain identification of Mycobacterium tuberculosis by DNA fingerprinting: recomendations for a standardized methodology. J Clin Microbiol 1993; 31: 406-409.
12. Small PM, van Embden JDA. Molecular epidemiology of tuberculosis. In Tuberculosis: Pathogenesis, Protection, and Control. Edited by Bloom BR. Washington, DC: American Society for Microbiology; 1994:569-582.
13. Snider DE, Jones WD, Good RC. The usefulness of phage typing Mycobacterium tuberculosis isolates. Am Rev Respir Dis 1984, 130:1095-1099.
14. Snider DE, Hutton MD. Tuberculosis in correctional institutions. JAMA 1989, 261: 393-397.
15. Glaser JB, Greifinger RB. Correctional health care: a public health opportunity. Ann Intern Med 1993, 118:139-145.
16. Chaves F, Dronda F, Cave MD, et al. A longitudinal study of transmission of tuberculosis in a large prison population. Am J Respir Crit Care Med 1997, 155:719-725.
17. Roberts GD, Koneman EW, Kim YK. Mycobacterium. In Manual of Clinical Microbiology. Edited by Balows A, Hausler WJ, Herman KL, Isenberg HD, Shadomy HJ. Washington, DC: American Society for Microbiology; 1991:304-339.
18. Chaves F, Yang Z, El Hajj H, et al. Usefulness of the secondary probe pTBN12 in DNA fingerprinting of Mycobacterium tuberculosis. J Clin Microbiol 1996, 34:1118-1123.
19. Stead WW. Pathogenesis of a first episode of chronic pulmonary tuberculosis in man: recrudescence of residuals of the primary infection or exogenous reinfection? Am Rev Respir Dis 1967, 95:729-745.
20. Romeyn JA. Exogenous reinfection in tuberculosis. Am Rev Respir Dis 1970, 101:923-927.
21. Mankiewicz E, Liivak M. Phage types of Mycobacterium tuberculosis in cultures isolated from Eskimo patients. Am Rev Respir Dis 1975, 111:307-312.
22. US Deparment of Health and Human Services. Prevention and treatment of tuberculosis among patients infected with human immunodeficiency virus: principles of therapy and revised recommendations. MMWR 1998, 47:1-56 (RR-20).

Mycobacterium tuberculosis; tuberculosis; tuberculosis reinfection; molecular epidemiology; DNA typing; HIV infection; prison; tuberculosis control program

© 1999 Lippincott Williams & Wilkins, Inc.