Antibody Responses against Pneumocystis jirovecii in Health Care Workers Over Time

Serena Fong

, Kieran R. Daly, Renuka Tipirneni, Leah G. Jarlsberg, Kpandja Djawe, Judy V. Koch, Alexandra Swartzman, Brenna Roth, Peter D. Walzer, and Laurence Huang

Author affiliations: San Francisco General Hospital/University of California, San Francisco, San Francisco, California, USA (S. Fong, R. Tipirneni, L.G. Jarlsberg, A. Swartzman, B. Roth, L. Huang); Veterans Affairs Medical Center/University of Cincinnati, Cincinnati, Ohio, USA (K.R. Daly, K. Djawe, J.V. Koch, P.D. Walzer)

Abstract

In a previous cross-sectional study, we showed that clinical staff working in a hospital had significantly higher antibody levels than nonclinical staff to Pneumocystis jirovecii. We conducted a longitudinal study, described here, to determine whether occupation and self-reported exposure to a patient with P. jirovecii pneumonia were associated with antibody levels to P. jirovecii over time. Baseline and quarterly serum specimens were collected and analyzed by using an ELISA that targeted different variants of the Pneumocystis major surface glycoprotein (MsgA, MsgB, MsgC1, MsgC3, MsgC8, and MsgC9). Clinical staff had significantly higher estimated geometric mean antibody levels against MsgC1 and MsgC8 than did nonclinical staff over time. Significant differences were observed when we compared the change in antibody levels to the different MsgC variants for staff who were and were not exposed to P. jirovecii pneumonia–infected patients. MsgC variants may serve as indicators of exposure to P. jirovecii in immunocompetent persons.

Pneumocystis jirovecii pneumonia (PCP) is the leading AIDS-defining illness in the United States and is a serious complication in transplant recipients and other immunocompromised persons. Although understanding of the epidemiology and transmission of Pneumocystis spp. has increased, much remains unknown. Studies have demonstrated the ubiquity of Pneumocystis isolates in the environment and their presence in the human lung; however, little is known about the precise reservoir for the Pneumocystis species that infects humans (P. jirovecii) (1–4). Traditionally, PCP was thought to occur by reactivation of latent infection acquired during childhood, but more recent studies suggest that the disease can also occur after recent exposure and infection (5–12). Animal-to-animal airborne transmission of Pneumocystis organisms has been demonstrated after brief periods of exposure (13–16). These animal studies suggest that person-to-person airborne transmission can take place after brief periods of exposure.
The occurrence of PCP outbreaks in hospital and clinic settings supports the theory that P. jirovecii can be transmitted from a patient with PCP to an immunocompromised patient at risk for PCP (17,18). Studies have also demonstrated that immunocompetent hospital workers who care for patients with PCP can become colonized with P. jirovecii as can family members of PCP-infected patients (19,20). Animal studies have shown that immunocompetent mice become transiently colonized with Pneumocystis spp. after exposure to immunocompromised PCP-infected mice and that the colonized mice subsequently transmit and infect Pneumocystis-free immunocompromised mice (21). These findings suggest that colonized persons who do not have PCP may be another source of transmission.
In the absence of a culture method, seroepidemiologic studies have provided valuable insights into the epidemiology of P. jirovecii in humans. In our prior studies, we used an ELISA to measure IgG levels against the P. jirovecii major surface glycoprotein (Msg) (22,23). The ELISA identifies overlapping recombinant fragments that span the length of the Msg from the amino terminus (MsgA) to the middle portion (MsgB) to the carboxyl terminus (MsgC) (22). In an earlier cross-sectional study of hospital staff with clinical (with patient contact) and nonclinical (without patient contact) occupations, staff with clinical occupations had significantly higher serum antibody levels against the MsgC1 variant, but not against MsgA or MsgB, than did staff with nonclinical occupations (23). To examine antibody responses in these same health care workers over time, we conducted a longitudinal study of clinical and nonclinical hospital staff in which serial serum specimens were collected. These specimens were then analyzed for IgG against MsgA, MsgB, and MsgC1 and against 3 additional Msg constructs: MsgC3, MsgC8, and MsgC9. The study objectives were to determine whether clinical occupation and self-reported exposure to a patient with PCP were associated with higher antibody levels against the different Msg variants over time. Such findings would support the theory that patient-to-provider transmission of Pneumocystis isolates occurs in the hospital setting and address the use of antibody levels against Msg as epidemiologic markers of Pneumocystis infection.