IX. Pneumocystis: the 'carrrier state': epidemiology and transmission of human pneumocystosis
Year: 1998
Type: review
Abstract: Pneumocystis carinii pneumonia (PCP) is a common opportunistic infection in immunocompromised patients, especially in patients with acquired immune deficiency syndrome (AIDS), and in those receiving immunosuppressive and/or antineoplastic chemotherapy. An important, unresolved question in the epidemiology of PCP concerns the modes of infection and transmission. PCP was first described in premature or debilitated infants [1], and in immunosuppressed children [2]. Seroepidemiological surveys have shown that the great majority of healthy children develop antibodies against Pneumocystis carinii (Pc) during the first decade of life and suggest that asymptomatic infection occurs. According to Wakefield et al., the acquisition of immunoglobulin response to Pc starts at 6 months of age and peaks in prevalence (80%) at about 8 years old [3]. Even so, it had been demonstrated that human immunodeficiency virus (HIV)-infected patients without PCP can mount a specific IgG-mediated antibody response to Pc [4]. It is generally thought that the PCP that develops in AIDS patients with severe destruction of CD4+ cells or in other patients with immunosuppressive agents represents the reactivation of latent endogenous Pc in the lung alveoli. The hypothesis of reactivation of latent Pc sp. f. hominis in the lung had long been presumed but demonstration of latent Pc in the lungs and bronchoalveolar lavage fluids of immunocompetent patients or autopsy lung tissue from immunocompetent human subjects was rarely possible with monoclonal antibody and even when recent sensitive techniques such as DNA amplification by polymerase chain reaction (PCR) were used [5]. These results suggest that asymptomatic carriage of Pc in the lungs of immunocompetent individuals is rare. Therefore, it has been admitted that the occurrence of PCP in immunocompromised patients might be from a new infection resulting from exposure to exogenous sources of Pc through contact with environmental sources or infected persons rather than reactivation of latent organisms. The currently accepted view that Pc remains latent in the lungs after a primary infection is supported by the high prevalence of Pc antibodies detectable at an early age and persisting through life in normal persons and by the occurrence of PCP with the advent of immunosuppression [6, 7]. The period of persistence of Pc in the lungs after PCP was studied using animal model (rats, mice) which takes advantage of the fact that PCP develops in almost of 100% of rats after a period of 6–8 weeks of dexamethasone-induced immunosuppression. After a primary PCP induced with dexamethasone, experimental rats were moved to a high-efficiency particulate air-filtered isolator to prevent further exposure to environmental Pc and allowed to recover. Vargas et al. [8] assessed reactivation of PCP by histologic examination, counts of cysts per gram of lung, and DNA amplification using nested PCR. A sequential and progressive decrease in Pc was detected. Thus Pc appeared to be cleared from the lungs of ≥75% of animals within 1 year after an episode of PCP, implying that persistence of latent organisms is limited. Moreover, Chen et al. [9] reported that Pc or specific Pc DNA was not detected in the lungs of reconstituted severe combined immunodeficiency (SCID) mice killed 21 days post reconstitution or thereafter. Furthermore, depletion of CD4+ cells in reconstituted SCID mice that had resolved PCP failed to reactivate the infection. These results suggest also that latent Pc was not present in either the lungs or the extrapulmonary sites of these mice. Roux et al. [10] showed that patients with negative Pc PCR in bronchoalveolar lavage (BAL) fluid had had a first positive Pc PCR in BAL about 3 weeks earlier. This also illustrates that Pc can be cleared. All these data support and confirm the hypothesis that PCP that develops in immunocompromised patients may be a new infection resulting from exposure to an exogenous source of Pc and not from reactivation of latent infection. Genetic differences among Pc organisms isolated from the same host (rats) were first observed by differences in migration of chromosomes resolved by electrophoretic karyotyping [11]. This study supposed that the possibility of mixed infection was due to the presence of multiple strains of Pc in rats after observation, in the same host, of comigration of chromosomes present in two different Pc patterns. Electrophoretic karyotypes have been obtained for Pc sp. f. hominis[11], but attempts to apply this technique to Pc sp. f. hominis have been hampered by the small numbers of organisms that can usually be obtained from patients. Genetic divergence among isolates of Pc sp. f. hominis has been most often demonstrated by molecular biology techniques such as sequence analysis of the mitochondrial large subunit (mt LSU) rRNA gene and the internal transcribed spacers (ITS) regions of Pc [12–14]. Samples from patients with PCP contained several mt LSU rRNA gene or ITS gene sequences in the same sample, showing that these patients were coinfected with genetically distinct strains of Pc sp. f. hominis[12, 13, 15, 16]. Despite effective prophylaxis, recurrent episodes of PCP are relatively common in HIV-infected patients. Recurrence of PCP in immunocompromised hosts may be the result of the inability of the hosts to clear all Pc organisms that the hosts harbored. Sequencing gene fragments derived by PCR showed that genetically distinct isolates were associated with each episode of PCP [15–17]. Therefore, patients who had two episodes of PCP separated by at least 4 months have different Pc sp. f. hominis isolates in the first and second disease episodes. These cumulative data support the hypothesis that recurrence of Pc infection occurs in immunocompromised hosts and may result from de novo acquisition of the organism from an exogenous source. Pc infection is usually a disease of the lungs, but cases of extrapulmonary and/or disseminated pneumocystosis have been reported, exceptionally before the AIDS epidemic (16 cases), much more frequently since its beginning (92 cases) [18, 19]. Nevertheless, extrapulmonary pneumocystosis (EPP) remains a rare disease: the incidence in HIV-infected patients has been reported to range from 0.06 to 2.5%[19]. But this incidence is certainly underestimated since not all diagnoses are made either before or after death. The tissues of the reticuloendothelial system (lymph nodes, spleen, liver, bone narrow) are the most common extrapulmonary foci, but almost all tissues have been reported as sites of infection. The frequency of lymph node involvement was 75% in non-HIV-infected patients, and 35% in HIV-infected people [20]. Other clinical features distinguished between these two populations [21]: (i) 81% of non-HIV-infected patients had concomitant PCP versus 47% of AIDS patients, (ii) clinical signs were present in only 56% of non-HIV-infected patients, so the diagnosis was recognized by autopsy in almost all of the cases (81%), whereas EPP was clinically evident and diagnosed by biopsy in 70% of HIV-infected patients, (iii) some extrapulmonary sites have never been described in non-HIV-infected patients i.e. the eye, ear, or central nervous system (CNS) [20]. In the two populations, two or more non-contiguous sites (that is defined as disseminated infection) were involved in 50% of cases. The prognosis was better in patients with infection restricted to a single site. The pathogenesis of EPP is not well understood. The results of recent experimental studies lend support to the hypothesis that the spread of Pc from the lungs to various organs, which was observed even in immunocompetent animals, is not an exceptional event and so might be the natural outcome of infection in the host [22–25]. The route of dissemination might be the lymphatic route because of the frequent involvement of structures belonging to the reticuloendothelial system. The existence of a hematogenous route is supported by infection of distant sites such as kidney or CNS, and the vascular and perivascular invasion by Pc in numerous cases. Pc DNA was detected by PCR in blood specimens from animals in experimental studies [22, 26, 29] or from immunosuppressed patients, infected or not with HIV, with or without PCP [26–28, 30–33]. Nevertheless, Pc DNA was not detected in blood samples from immunocompetent patients. At the present time, it is not possible to determine if the circulating form of Pc is viable and infectious. Because of its lack of systemic activity, aerosolized pentamidine has been reported to be a risk factor for dissemination of Pc [19, 21, 34]. But cases of EPP have been described in patients receiving dapsone-pyrimethamine prophylaxis [34, 35]. The most important pathogenic factor might be the virulence of the Pc strain. In experimental studies [22–25], the proliferation of Pc was lower in extrapulmonary sites than in the lungs and the histopathologic features of human EPP (i.e. eosinophilic exudate and necrosis) were not observed. More aggressive strains might be responsible for a massive proliferation and transition from ‘physiological’ to ‘pathological’ involvement of an extrapulmonary site. Genetic diversity among Pc sp. f. hominis strains has been already demonstrated by molecular typing methods. It should be possible to determine whether (i) the organism found in the extrapulmonary site and that found in lung when concomitant PCP is present belong to the same strain, (ii) Pc strains with special adaptability and capacity for development are involved in the occurrence of EPP. Since 1990, PCR has provided the most sensitive technique for Pneumocystis detection [36]. This technique has made it possible to detect low levels of parasite in pulmonary samples from immunocompromised patients without PCP. These cases have been reported as Pc sp. f. hominis carriage, essentially among HIV-positive patients [36–40], whereas there are few data about Pc sp. f. hominis carriage in HIV-negative patients. Moreover, different frequencies of parasite carriage in HIV-negative patients have been reported: 10% among patients with chronic obstructive lung disease [41], 40% among immunosuppressed patients [42], 8.6% among severely immunodepressed patients [43], 33% among patients who were more or less immunodepressed [44]. Moreover, it is quite difficult to compare these data because of differences in: (i) patient immunodepression status, (ii) pulmonary samples, and (iii) DNA extraction and amplification procedures. However, it is noticeable that parasite carriage was much more reported in severely immunodepressed patients, whereas few parasite carrier patients were not. For example, Calderon et al. [41] noted that two of their five Pc sp. f. hominis carrier patients received any corticosteroid therapy. In the same way, Ribes et al. [44] described that only 14 patients of their 23 carriers (60.9%) had recently taken or were currently receiving steroid therapy. Thus, a severe immunodepression would not be always necessary to have parasite carriage, as PCP has been sometimes observed without any severe immunodepression [44–46]. Nevez et al. [47] compared the Pc sp. f. hominis carriage in HIV-negative patients versus HIV-positive patients. Two hundred and three patients, admitted to two different hospitals, were investigated. The difference between the frequencies of parasite carriage in the HIV-positive population and the HIV-negative population was not significant because of the too low number of HIV-positive patients enrolled. However, the number of carrier patients was higher than the number of patients with PCP. Among these carrier patients, HIV-negative patients were the most numerous. With regard to this last observation, it appears particularly interesting to investigate more precisely parasite carriage among HIV-negative populations. Furthermore, it is important to establish if carrier persons are potentially infectious for other severe immunodepressed patients. It is well documented that immunosuppressed animals may become infected when housed with infected animals and airborne transmission has been suggested in an immunosuppressed animal model of PCP [48, 49]. It is therefore assumed that human infection could be similarly acquired. Although person-to-person transmission has not been documented, clusters or outbreaks of infection suggest that this does occur [50]. The reports of cases of PCP in premature and/or malnourished children in institutions or other crowded conditions in the years following World War II point to person-to-person spread [1, 51–53]. The occurrence of small outbreaks in immunosuppressed patients, in organ transplant units, hospitalized simultaneously with HIV-seropositive subjects suggests that nosocomial infection or person-to-person transmission can occur [54]. A recent study, using typing of Pc sp. f. hominis strains, in three HIV-infected couples with PCP has shown that each member of the three couples was infected by a genetically distinct strain of Pc sp. f. hominis[55]. These results thus rule out transmission within each couple and therefore, according to recent reports on Pc sp. f. hominis nucleic acids detected by PCR in ambient air of an animal room, a patient room, and Oxford county [56, 57], the most probable mode of transmission in these cases is environmental. Transmission to hospital personnel was first supposed by Schluger with the detection of asymptomatic PCP by amplification of the DHFR gene of Pc in blood samples from the staff [28]. Previous serological studies indicated that HIV-positive patients with PCP had a higher rate of antibody response to the major surface antigen (MSG) of Pc than HIV-negative controls [4]. Exposure of staff to Pc may occur especially in hospital departments treating patients with PCP. A recent report studying the possible transmission of Pc from patients to staff has been described by analysis of the serological response in hospital staff with and without exposure to patients with PCP [58, 59]. Among 38 people handling patients with documented PCP at two infectious disease departments, 26% had antibodies to MSG and in a group of 40 staff without exposure to patients with PCP, 37% had antibodies to MSG. These results are unable to demonstrate that transmission to staff increases with contact of patients with PCP.
Institutions: Hôpital Saint-Antoine, Institut Pasteur, Institut Pasteur de Lille
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2058648331