Pneumocystis pneumonia (PCP), which is caused by Pneumocystis jirovecii (formerly P. carinii f. sp. hominis), is frequently the first serious illness encountered by HIV-infected persons. During the early years of the AIDS epidemic, PCP was the AIDS-defining illness for as many as two thirds of patients in the United States. Although a decline in incidence of PCP occurred during the era of highly active antiretroviral therapy (HAART), PCP remains the most common serious opportunistic illness in HIV-infected persons ( 1 ). Patients in the developing world without access to PCP prophylaxis or antiretroviral drugs remain at high risk, and PCP continues to develop in certain groups in industrialized countries. The drug of choice for treatment and chemoprophylaxis of PCP is trimethoprim-sulfamethoxazole (TMP-SMX). In recent years, antimicrobial drug resistance has emerged as a possible cause of failure of patients to respond to TMP-SMX. Investigators have demonstrated an association between exposure to sulfa drugs and mutations in the dihydropteroate synthase (DHPS) gene of P. jirovecii, but the relationship between these mutations and treatment (or prophylaxis) failure is unclear. Understanding whether DHPS mutations cause antimicrobial drug resistance is important in guiding clinicians who care for patients with PCP. A series of articles in this issue of Emerging Infectious Diseases highlights the continuing importance of PCP, the potential for drug resistance, and laboratory techniques that can be used to study the problem. We hope that these articles will stimulate interest in exploring the relationship between DHPS mutations and resistance of P. jirovecii to sulfa-containing drugs and in assessing DHPS mutations as possible causes of treatment failure in patients with PCP. In this introductory article, we summarize the changes in incidence of PCP since the introduction of HAART, discuss groups at risk for PCP in developing and industrialized nations, and examine possible future trends in the disease. A data collection form has been included online with this series of articles to assist in the collection of appropriate and standardized data from patients with PCP and to facilitate comparing and pooling data from different centers (Appendix). PCP before HAART The first clinical cases of PCP were reported during World War II in orphanages in Europe. These cases of "plasma cell pneumonia" were common among malnourished children and were later reported in children in Iranian orphanages. The disease was then recognized in patients who were immunocompromised because of malignancies, immunosuppressive therapy, or congenital immunodeficiencies. Solid organ transplantation increased the number of patients at risk for PCP, although rates diminished after chemoprophylaxis was introduced. Without chemoprophylaxis, rates of PCP are 5%–25% in transplant patients, 2%–6% in patients with collagen vascular disease, and 1%–25% in patients with cancer. Defects in CD4+ lymphocytes are a primary risk factor for developing PCP, but the immune response to Pneumocystis is complex. CD8+ lymphocytes seem to be important in Pneumocystis clearance, and defects in B-cells and antibody production may also predispose to PCP. The beginning of the AIDS epidemic in the early 1980s shifted the incidence of PCP from a rare disease to a more common pneumonia. Clusters of PCP cases in homosexual men and intravenous drug users were one of the first indications of the HIV epidemic ( 2 ). PCP rapidly became the leading AIDS-defining diagnosis in HIV-infected patients. In the initial stages of the epidemic, PCP rates were as high as 20 per 100 person-years for those with CD4+ cell counts 5,000 homosexual men since 1984 ( 6 ). Of these, 2,195 were either HIV-infected at time of enrollment or seroconverted to HIV during the study. Opportunistic infection rates were compared for the HAART era (1996–1998) and the era of antiretroviral monotherapy (1990–1992) ( 7 ). For persons who seroconverted during the study period, the relative hazard for development of PCP from seroconversion to initial AIDS-defining opportunistic infection was 0.06 during the HAART era compared to the time of monotherapy. For those already diagnosed with AIDS, the study found a hazard of 0.16, which demonstrated a dramatically lower risk for PCP during the HAART era. In Europe, the EuroSIDA study has followed a cohort of >8,500 HIV-infected patients. The investigators examined changes in incidence of AIDS-defining illnesses before and after HAART was introduced and found results similar to those in North America ( 8 ). PCP cases decreased over time (1994–1998). Incidence of PCP fell from 4.9 cases per 100 person-years before March 1995 to 0.3 cases per 100 person-years after March 1998 ( 9 ). Occurrence in Relation to PCP Prophylaxis PCP still occurs in industrialized nations despite the availability of HAART and anti-Pneumocystis prophylaxis. ASD investigated the history of prescriptions for PCP prophylaxis in HIV-infected adults in whom developed PCP from 1999 through 2001 (Figure 2). Almost 44% of PCP cases occurred in patients not receiving medical care, most of whom were probably not known to be HIV-infected. Forty-one percent of patients were prescribed prophylaxis but did not adhere to treatment, or PCP developed despite their taking medications appropriately. Possible explanations for PCP in the "breakthrough" group include the development of drug-resistant Pneumocystis or decreased efficacy of prophylaxis in those with low CD4+ cell counts. An additional 9.6% of patients were under medical care and should have received prophylaxis based on current recommendations, but had not been prescribed prophylaxis by their providers. Five percent of patients were under care but did not meet criteria for prophylaxis. Figure 2 Classification of Pneumocystis pneumonia cases from 1999–2001, CDC Adult and Adolescent Spectrum of HIV Disease Project, n = 1,073. Risk Factors A CD4+ cell count 200 cells/µL, the risk for PCP decreases sufficiently to safely discontinue both primary and secondary prophylaxis ( 9 , 11 ). Those in whom PCP develops while on HAART typically have low CD4+ cell levels. ASD found that the median CD4+ cell count in persons with PCP while on HAART was extremely low (29 cells/µL), although the count was somewhat higher than for those not on HAART (13 cells/µL) ( 1 ). The EuroSIDA study reported that persons on HAART in whom PCP developed had a median CD4+ cell count of 30 cells/µL, identical to those with PCP who were not receiving HAART ( 8 ). Patients without improvement in their CD4+ cell count despite use of HAART remain at risk for PCP, and PCP still rarely occurs in persons with CD4+ cell counts >200 cells/µL. Other clinical factors such as sex, race or ethnicity, and HIV transmission category have been examined as risk factors for PCP. Men and women appear to have an equivalent risk for PCP ( 12 ). One study demonstrated that African Americans have approximately one third the risk for PCP as white persons ( 10 ), but this finding has not been replicated ( 12 ). PCP risk according to HIV transmission category is also debated. One autopsy study found that PCP was less common in intravenous drug users than in other risk groups ( 13 ). Kaplan et al. found a slightly increased risk for those men who had sex with men and were intravenous drug users, but risk was equivalent in other transmission categories ( 12 ). Risk for Pneumocystis Colonization Although PCP cases have declined, polymerase chain reaction (PCR) has led to the discovery of Pneumocystis DNA in asymptomatic persons. Pneumocystis in respiratory specimens from persons who do not have signs or symptoms of clinical infection and who do not progress to infection has been defined as colonization or subclinical carriage. Often, Pneumocystis DNA is detected only by PCR, and the organism is not seen on routine histochemical staining. The clinical significance of Pneumocystis in respiratory specimens and the viability of organisms detected only by PCR are unknown. However, colonization may be important for several reasons. Pneumocystis colonization may increase the risk for progression to PCP, carriers of the organism may transmit infection to others, and latent infection may lead to inflammation that is detrimental to the lung. Most healthy persons do not have detectable Pneumocystis in respiratory specimens, but rates of colonization may be as high as 69% in HIV-infected persons ( 14 ). Recent evidence suggests that non–HIV-infected persons may also be colonized with Pneumocystis, thus increasing the potential number of persons affected ( 15 ). PCP in Children in Industrialized Countries Incidence Early in the HIV epidemic, PCP occurred in HIV-infected children at a rate of 1.3 cases per 100 child-years from infancy to adolescence and was as high as 9.5 cases per 100 child-years in the first year of life ( 16 , 17 ). In the 1990s, pediatric HIV infection decreased, primarily as a result of improved prenatal HIV testing and use of HIV treatment to prevent vertical transmission of the virus. The Pediatric Spectrum of Disease (PSD) study found significant decreases in the rates of most opportunistic infections in HIV-infected children during the HAART era (Figure 3). PCP cases declined significantly from 1992 to 1997, with an increase in the rate of decline after 1995, presumably from HAART ( 1 ). Because widespread use of HAART for children has occurred more recently than for adults, the full effect of HAART on pediatric PCP likely has not yet been realized. Figure 3 Yearly opportunistic infection rates per 1,000 HIV-infected children, CDC Pediatric Spectrum of Disease Project, 1994–2001. Bacterial, bacterial infections; CMV, cytomegalovirus; HAART, highly active antiretroviral therapy; LIP, lymphocytic interstitial pneumonia; MAC, Mycobacterium avium complex; PCP, Pneumocystis pneumonia. Incidence rates were calculated per 1,000 children at risk each year. All trends were significant at p 10 years) except for the 15 months ( 41 ). The largest autopsy series examined 180 HIV-infected children in Zambia ( 38 ). Twenty-nine percent of the children died of PCP, making PCP the third leading cause of death overall. Among children 40% among HIV-infected children hospitalized with pneumonia ( 42 , 43 ). Ruffini studied children from 2 to 24 months of age with pneumonia and found that 48.6% had PCP ( 43 ). Madhi found that in 231 episodes of pneumonia in HIV-infected children, 101 (43.7%) were due to PCP ( 39 ). PCP was most common in infants 10,000 copies/mL after 48 weeks of treatment ( 45 ). In the EuroSIDA cohort, an increasing proportion of HIV-infected patients have been exposed to all classes of antiretrovirals, with 47% of their cohort exposed to nucleoside reverse transcriptase inhibitors, protease inhibitors, and non-nucleoside reverse transcriptase inhibitors by 2001 ( 45 ). Of those patients in the cohort with multidrug-resistant HIV who received salvage regimens, a new AIDS-defining opportunistic infection developed in 11%. Growing transmission of resistant HIV is also likely. If new drugs do not become available, the number of patients with resistant virus and opportunistic infections, including PCP, will continue to climb. Not only is HIV developing resistance, but Pneumocystis may also develop resistance to standard prophylaxis and treatment regimens. Many researchers have reported mutations of Pneumocystis in response to use of sulfa- or sulfone-containing anti-Pneumocystis regimens. Whether these mutations increase the likelihood of prophylaxis or treatment failure is unclear and is reviewed in other papers in this series. Conclusion Despite the declines in death and disease from HIV in the United States and western Europe, PCP remains an important disease and is unlikely to be eradicated. In industrialized nations, PCP still occurs in those not yet diagnosed with HIV or not in medical care, those not receiving PCP prophylaxis, and those not taking or not responding to HAART. Resistance in HIV and Pneumocystis may contribute to future increases in PCP incidence. In most developing nations, AIDS patients are at high risk for PCP. In sub-Saharan Africa, the effect of disease from PCP in infants and children is high and is probably greater in adults than previously recognized. Colonization rates among both HIV-infected and non–HIV-infected populations may also be substantial. Better understanding of the epidemiology and transmission of PCP, and improved efforts in prevention and treatment, are needed. Supplementary Material Appendix PNEUMOCYSTIS CARINII: SURVEILLANCE FOR DRUG-RESISTANCE PCP CHART ABSTRACTION FORM.