X.Pneumocystisand pneumocystosis in Europe at the end of the 20th century

Pneumocystis carinii was described as a new microorganism in the lungs of rats living in Paris by Delanoë and Delanoë in 1912 [1]. In the early 1950s, the first cases of human pneumocystosis were described in orphans and in infants living in eastern Europe [2, 3]. Then Pneumocystis carinii pneumonia (PCP) was reported as an opportunistic disease in patients treated with immunosuppressive therapy [4] for malignancy, hemopathies or cancer. The significant increase in PCP incidence at the beginning of the 1980s has been related to the emergence of the AIDS epidemic [5]. In the last 10 years, important improvements have occurred in the field of diagnosis and therapy strategies, simultaneously, due to the interest of an increasing number of research teams. From the diagnostic point of view, more sensitive techniques such as immunofluorescence with monoclonal antibodies [6] and PCR [7–10] have been proposed with less invasive samples such as bronchoalveolar lavage, induced sputum [11] and more recently oral washes [12–14]. Since the beginning of the 1990s, primary prophylaxis has been proposed for HIV-infected patients with a CD4 count below 200/mm3[15]. Trimethoprim-sulfamethoxazole (TMT-SMX) is the most efficient molecule, but others can be proposed such as aerosolized pentamidine [16]. The evolution of antiretroviral therapies, mainly new antiretroviral associations for HIV infection, will change the wide spread of PCP, but many questions are emerging and new problems may have to be taken into account over the next years. We will discuss here figures from different sources and emphasize questions for the future. In June 1997, the HIV/AIDS surveillance in Europe reported that PCP remains the first AIDS indicative disease diagnosed since 1994 in adult and adolescent men, women and children in WHO European regions [17]. The clinical group from CISIH in France reported a PCP incidence of 20±1 per 1000 patient years in the first half of 1996 with a marked decrease (49%) in the second half of 1996 [18] (Fig. 1). Those dramatic changes in incidence can likely be explained by the modification of antiretroviral prescription habits over the 1995–1996 period. But the frequency of the inaugural PCP, which had been decreasing between 1990 and 1995, has increased since the second semester of 1996. This relative increase is due to the best management of patients known to be HIV positive but a significant proportion of patients presenting with AIDS who discover their HIV status at the time of AIDS diagnosis [19]. Incidence of first occurrence of PCP (and 95% confidence interval). Another French study conducted in the Parisian area, based on the ‘Reseau AP-HP de surveillance de la pneumocystose’ from 1990 to 1996 showed a marked increase until 1994, then an 8% decrease in 1995 and a 25% decrease between 1995 and 1996. These figures were mainly explained by the improvement in HIV therapies. The number of cases occurring in patients who are not aware of their HIV status remains stable over time while it has been decreasing in patients followed for HIV infection (Fig. 2). A study performed over the same period by our team points out the large number of patients who do not know their HIV status [20]. When PCP occurs under prophylaxis, only a few cases have been reported with TMP-SMX; most of the time aerosolized pentamidine was involved because of TMP-SMX intolerance or bad compliance. The same observation has been reported recently by Kaplan [21] and Schwarcz et al. [22] in the USA. They showed that patients at higher risk of developing PCP are those with a particular social status without access to a medical center and those patients cannot benefit either from new treatment or from prophylaxis. It appears that incitement to HIV testing should be improved and that facilitation of medical access for destitute people should be provided. Evolution of PCP revealing HIV infection. The number of PCP in HIV seronegative patients remains stable. The patients present with hematologic malignancies such as acute lymphoblastic leukemia or receiving immunosuppressive therapies [23] or suffering from Wegener granulomatosis [24]. PCP occurs just after the induction of chemotherapy; the prognosis seems worse in HIV seronegative than in AIDS patients [25]. So, even more than for AIDS patients, primary prophylaxis should be considered in these patients in order to avoid the infection. Up to now, the PCP epidemiology remains poorly understood. It has long been thought that PCP occurred from reactivation with latent Pneumocystis carinii (Pc) [26–28], but recently it has been proved that PCP can result from de novo contamination [29–31]. Airborne transmission has been clearly established but the infective forms remain unknown [32]. Pc DNA gene sequences similar to those of Pneumocystis of different species such as hominis have been detected in air samples [33–37] and in water [38] suggesting that this microorganism is present in our environment. It has not been determined if Pc can be transmitted from specific environmental sources and/or directly from person to person. Nosocomial transmission of Pc from HIV infected to non-infected immunocompromised patients has been suggested in small outbreaks [39–42] but never proven. Recently, we reported [42] the usefulness of Pc sequencing to study hypothetical person to person transmission; the typing of Pc from three HIV positive couples showed that they were genetically different. Those results ruled out transmission within each couple. The transmission was probably from environmental sources. Two recent immunologic studies [43, 44] also exclude transmission from care staff to immunosuppressed patients as well as the reverse (from patients to staff). For HIV seropositive patients, the improvement of highly active antiretroviral therapies may lead to a further decrease of PCP incidence. It will therefore become questionable to continue prophylaxis for patients whose CD4 count recovers beyond 200/mm3. If primary prophylaxis has to be continued, as suggested by a recent report [46], then the risk of TMP-SMX resistance emergence will increase [45] and new molecules should be found for prophylaxis and therapeutic purposes. The detection of resistant strains may be established by sequencing [45] and would be useful for therapy guidance. If prophylaxis may be stopped, then it would be very useful to improve our knowledge about the spread of Pc and to limit exposure of immunocompromised patients to prevent PCP [32, 36]. For HIV seronegative immunocompromised patients primary prophylaxis should be proposed to avoid lethal complications. At the end of the 20th century, even in the most optimistic situation, Pc will remains a large preoccupation for physicians and researchers essentially in preventive and therapeutic aspects. The main priority will be to facilitate access to medical treatment with appropriate messages encouraging HIV testing in high-risk people; all the more, because Pc has recently been recognized to be a cofactor increasing multiplication of HIV [46]. Improvement in epidemiological knowledge and new treatment may facilitate the control of this opportunistic disease [47]. We thank P.M. Girard for his help.