Clearance of Pneumocystis carinii cysts in acute P carinii pneumonia: assessment by serial sputum induction

STUDY OBJECTIVES

To determine the feasibility of repeat sputum induction in acute Pneumocystis carinii pneumonia (PCP) and to define the rate of clearance of P carinii cysts from the respiratory tract of HIV-seropositive patients with acute PCP.

DESIGN

Prospective cohort evaluation.

SETTING

University medical center.

PARTICIPANTS

Twenty-four HIV-seropositive subjects with acute PCP.

MEASUREMENTS

Sputum induction for P carinii 2, 3, 4, and 6 weeks after initial diagnosis, and follow-up for 1 year.

RESULTS

Eighty-eight percent of subjects had residual cysts at 2 weeks, 76% at 3 weeks, 29% at 4 weeks, and 24% at 6 weeks postdiagnosis. A prior AIDS-defining illness (p = 0.033) or prior PCP (p = 0.004) predicted relapse within 6 months, but persistent cysts at 3 weeks did not; 8 of 16 sputum-positive subjects and 1 of 5 sputum-negative subjects experienced a relapse within 6 months (p = 0.34). Secondary prophylaxis with trimethoprim-sulfamethoxazole was associated with a reduced risk of relapse.

CONCLUSIONS

Serial sputum induction coupled with direct fluorescent antibody staining is a feasible, noninvasive method of respiratory tract surveillance for the eradication of P carinii during and after acute PCP. Three-quarters of HIV-seropositive patients with acute PCP have persistent cysts in their lungs at the end of antimicrobial treatment, despite clinical recuperation, but only one quarter have residual cysts 6 weeks postdiagnosis. A prior AIDS-defining illness and prior PCP are positively associated, and subsequent trimethoprim-sulfamethoxazole prophylaxis is negatively associated, with relapse within 6 months, while persistent organisms at 3 weeks do not appear to be a significant predictor of relapse risk.

Diversity of host species and strains of Pneumocystis carinii is based on rRNA sequences

We have amplified by PCR Pneumocystis carinii cytoplasmic small-subunit rRNA (variously referred to as 16S-like or 18S-like rRNA) genes from DNA extracted from bronchoalveolar lavage and induced sputum specimens from patients positive for P. carinii and from infected ferret lung tissue. The amplification products were cloned into pUC18, and individual clones were sequenced. Comparison of the determined sequences with each other and with published rat and partial human P.carinii small-subunit rRNA gene sequences reveals that, although all P. carinii small-subunit rRNAs are closely related (approximately 96% identity), small-subunit rRNA genes isolated from different host species (human, rat, and ferret) exhibit distinctive patterns of sequence variation. Two types of sequences were isolated from the infected ferret lung tissue, one as a predominant species and the other as a minor species. There was 96% identity between the two types. In situ hybridization of the infected ferret lung tissue with oligonucleotide probes specific for each type revealed that there were two distinct strains of P. carinii present in the ferret lung tissue. Unlike the ferret P. carinii isolates, the small-subunit rRNA gene sequences from different human P. carinii isolates have greater than 99% identity and are distinct from all rat and ferret sequences so far inspected or reported in the literature. Southern blot hybridization analysis of PCR amplification products from several additional bronchoalveolar lavage or induced sputum specimens from P. carinii-infected patients, using a 32P-labeled oligonucleotide probe specific for human P. carinii, also suggests that all of the human P. carinii isolates are identical. These findings indicate that human P. carinii isolates may represent a distinct species of P. carinii distinguishable from rat and ferret P. carinii on the basis of characterization of small-subunit rRNA gene sequences.

Polyclonal infections due to Mycobacterium avium complex in patients with AIDS detected by pulsed-field gel electrophoresis of sequential clinical isolates

Invasive infection with organisms of the Mycobacterium avium complex (MAC) is common among patients with advanced human immunodeficiency virus infection. In previous studies, we analyzed multiple individual colonies of MAC isolated from specimens obtained at the same time and observed that 14 to 20% of patients are simultaneously infected with more than one strain. In this study, we examined sequential isolates from 12 patients with AIDS who had two or more MAC isolates available from clinical specimens collected more than 1 week apart; the intervals between the first and last specimens ranged from 8 to 192 (median, 46) days. For each isolate, restriction digests of genomic DNA were analyzed by pulsed-field gel electrophoresis; DNA was prepared by using a protocol, described here in detail, which had been optimized for conditions of bacterial growth and lysis. The pulsed-field gel electrophoresis analysis identified four patients (33%) infected with two different MAC strains. Both M. avium and M. intracellulare were cultured from blood specimens from two patients. In each of the four patients, the second strain was identified from a culture taken within 14 days of the initial study isolate, and in three of these patients, the first strain was detected again in a subsequent culture. These observations suggest that the presence of two different strains among isolates from sequential cultures may reflect ongoing polyclonal infection. We conclude that polyclonal infection with MAC is common among patients with AIDS. The identification of such infections may be critical in the development of effective treatments.