Correspondence

Reinfection with Multidrug-Resistant Tuberculosis

N Engl J Med 1993; 329:811-812September 9, 1993

Article

To the Editor:

In their report in the April 22 issue, Small et al.1 showed that exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis can occur either during therapy for the original infection or after therapy has been completed. They described three patients seropositive for human immunodeficiency virus (HIV) (Patients 8, 11, and 12) with drug-susceptible tuberculosis whose original tuberculosis strains became resistant to rifampin alone during therapy. This is an unusual observation. The first phenotypic change noted in virtually all persons infected with strains of tuberculosis that become drug resistant is resistance to isoniazid.

We have noted the acquisition of drug resistance to rifampin alone in three patients with the acquired immunodeficiency syndrome. These 3 patients are the only ones known to have relapsed among 57 HIV-infected persons who have received treatment through our program. Otherwise, in caring for almost 1000 patients with tuberculosis during the past 10 years, we have not observed acquired rifampin resistance except among patients with previously demonstrated resistance to isoniazid.

There are so many unique features of HIV-associated tuberculosis that it may plausibly be viewed as a new disease. The three cases reported by Small et al.,1 our three cases, and at least two others described previously2,3 suggest that the development of resistance to rifampin, rather than to isoniazid, is often the first step in the acquisition of drug resistance in HIV-infected patients who have a poor record of compliance with therapy for tuberculosis. Whether this phenomenon is related to the upsurge of multidrug-resistant tuberculosis remains to be determined.

Charles M. Nolan, M.D.
Seattle-King County Department of Public Health, Seattle, WA 98104

3 References

1. 1

   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
   Full Text | Web of Science | Medline

2. 2

   Dylewski J, Thibert L. Failure of tuberculosis chemotherapy in a human immunodeficiency virus-infected patient. J Infect Dis 1990;162:778-779
   CrossRef | Web of Science | Medline

3. 3

   Small PM, Schecter GF, Goodman PC, Sande MA, Chaisson RE, Hopewell PC. Treatment of tuberculosis in patients with advanced human immunodeficiency virus infection. N Engl J Med 1991;324:289-294
   Full Text | Web of Science | Medline

To the Editor:

Small et al. analyzed restriction-fragment-length polymorphisms (RFLPs) of M. tuberculosis from 17 HIV-infected patients. While comparing the RFLP analysis of these strains, I noticed certain similarities between the original strains from Patients 11 and 14 (11A and 14A) and the multidrug-resistant strain from Patients 13 through 17 (13B through 17B). The four larger bands generated by digestion with the restriction enzyme PvuII (5, 4.8, 4.2, and 3.0 kb) are found in the original strain from Patient 11 and the resistant strain from Patients 13 through 17 (as shown in Figure 2 and 3 of their article). The seven smaller bands (2.5 kb and smaller) are found in the original strains from Patients 11 and 14 but not in the drug-resistant strain. It is possible that strain 11A in the study represents the recombination of strain 14A and the resistant strain. The six smaller PvuII bands appear to be unique to the resistant strain and may be important for the multidrug resistance. Interestingly, the resistant strain appears to differ from the strains analyzed previously by the same group.1

The molecular mechanism of multidrug resistance in M. tuberculosis is not clear. The appearance of a single strain infecting all five patients (Patients 13 through 17) suggests that the strain may have a unique genetic structure other than a single point mutation or the acquisition of plasmids carrying genes for drug resistance. Such genes may reside in the six smaller fragments produced by PvuII digestion (13B through 17B) found only in the resistant strain. The drug resistance in Patients 7 through 12 appears to develop from the original genome, as discussed by the authors, and may represent a point mutation or small deletion or duplication not detectable by RFLP. Cloning and sequencing of the six small PvuII bands common to the resistant strain may enable us to identify genes that are not present or that differ from those found in non-resistant strains of M. tuberculosis, and thus offer clues to new treatments for this disease.

Bryan Tzy-young Lin, M.D., Ph.D.
Beth Israel Hospital, Boston, MA 02215

1 References

1. 1

   Daley CL, Small PM, Schecter GF, et al. An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus -- an analysis using restriction-fragment-length polymorphisms. N Engl J Med 1992;326:231-235
   Full Text | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: RFLP analysis of M. tuberculosis may be used to identify unambiguously patients in whom resistance has developed as a result of a mutation in the strains responsible for the original infection. As Nolan and Lin infer, this technique allows old observations to be reexamined and new questions to be raised regarding the mechanism by which resistance to antimicrobial agents develops.

Nolan suggests that the sequence by which drug resistance emerges in individual patients may be influenced by coinfection with HIV. Among patients with tuberculosis treated at Kings County Hospital, resistance to rifampin alone clearly occurs (in both HIV-seropositive and HIV-seronegative patients), although when resistance to a single agent occurs, it is more commonly due to isoniazid. We are unaware of conclusive data regarding temporal trends or the role of coinfection with HIV in relation to these observations.

We agree with Lin that genetic analysis of M. tuberculosis clones that have been shown by RFLP to have acquired drug resistance may provide key insights into the molecular basis of drug resistance. The insertion-sequence element used to produce the patterns in our study does not appear to be directly responsible for drug resistance by virtue of either its putative protein product or the sites into which it is inserted. In collaboration with other investigators, we have begun a centralized collection of epidemiologically characterized M. tuberculosis strains to serve as a communal resource for the demonstration of the molecular mechanisms of drug resistance. For example, the examination of strains that simultaneously become resistant to multiple antibiotics could lead to the identification of a global resistance mechanism involving altered permeability of the bacterial-cell envelope by a mechanism analogous to the multidrug-resistant locus in tumor cells.

Peter M. Small, M.D.
Robert W. Shafer, M.D.
Gary K. Schoolnik, M.D.
Howard Hughes Medical Institute, Stanford, CA 94306