We performed a retrospective cohort study using data collected by the Palo Alto Medical Foundation Toxoplasma Serology Laboratory (PAMF-TSL; www.pamf.org), Palo Alto, California, USA, during 1991–2010. Patient blood samples were sent from diverse laboratories from throughout the United States, and testing was conducted at the PAMF-TSL. The study was approved by the Institutional Research Board at the PAMF Research Institute.

From the PAMF-TSL database, we identified families that 1) had an index case-patient with a diagnosis of acute toxoplasmosis and 2) had >1 additional household/family member who had been tested for T. gondii infection at PAMF-TSL. Details of the process used to identify additional household/family members are described in the Technical Appendix. All identified family/household members were categorized as acutely infected (<6 months before sample collection time); recently infected (6–12 months before sample collection time); chronically infected (>12 months before sample collection time); or never infected. The criteria used for this categorization are described in the Technical Appendix. These criteria are routinely used in the daily clinical practice at PAMF-TSL to estimate the most likely time of the T. gondii infection; the accuracy of these criteria has been previously validated (7–11).

All identified families were categorized in 3 family groups (Technical Appendix). Group 1 consisted of families with an index case-patient who had acute toxoplasmosis and >1 additionally tested family/household member who had acute or recently acquired T. gondii infection. Group 2 consisted of families with an index case-patient who had acute toxoplasmosis; >1 additionally tested family/household member who had chronic T. gondii infection; and no other tested household members who had evidence of acute or recently acquired T. gondii infection. Group 3 consisted of families with an index case-patient who had acute toxoplasmosis and in which no additionally tested family/household members showed evidence of T. gondii infection.

We defined as prevalence of acute T. gondii infection in >1 family members (prevalence of group 1 families) the number of group 1 families divided by the total number of study families over the 20-year study period (primary endpoint). As secondary endpoint, we also calculated the prevalence of group 2 families. We also tested whether the IgG-Dye test titers and IgM-ELISA titers of the index case-patients were different across the 3 family groups by using the Kruskal-Wallis test. All analyses were done in Stata/SE version 12 (StataCorp LP, College Station, TX, USA).

Figure

Figure. . Flowchart for the identification of families with an index case-patient who had acute toxoplasmosis and >1 family member with acute or recent Toxoplasma gondii. infection. Data were extracted from...

Among 97,279 persons serologically tested for T. gondii in the PAMF-TSL over the 20 year study period, we identified 107 persons who had >1 person from their household with a diagnosis of acute toxoplasmosis and >1 additional household member serologically tested for T. gondii infection. Those 107 persons were grouped into 32 study families (Figure). Patient demographic and clinical characteristics are shown in Table 1; serologic test results for members of group 1 families are shown in Table 2, Appendix, and for members of groups 2 and 3 families in the Technical Appendix.

The prevalence of group 1 families in our study was 56% (18/32); group 2 families, 16% (5/32); and group 3 families, 28% (9/32) (Figure). The IgG-Dye test and the IgM-ELISA titers of the index case-patients were not significantly different across the 3 family groups (p = 0.27 for IgG and p = 0.07 for IgM) (Table 2, Appendix; Technical Appendix). For group 1 families, all additional family members with acute/recently acquired infection had serologic profiles (titers of IgG, IgM, and/or IgA/IgE and avidity) that were similar to those of the index case-patients, indicating that they were infected at about the same time (Table 2, Appendix).

Our data provide preliminary evidence that multiple cases of acute T. gondii infection may occur among family/household members. These findings are particularly critical for persons at high risk from T. gondii infection, such as women who are or may become pregnant or immunocompromised persons. Interpretation of our study findings would have been clearer had the background prevalence of acute toxoplasmosis in the United States been known. Although no such population-level empirical data exist, we have identified at PAMF-TSL 889 patients with acute T. gondii infection over the 20-year study period (estimated prevalence ≈9/1,000 patients screened at PAMF-TSL; unpub. data).

A limitation of our study is that the families tested at PAMF-TSL over this study period might represent a group in whom the prevalence of acute T. gondii infection in >1 family member has been overestimated. Only 4% of persons who had acute toxoplasmosis diagnosed at PAMF-TSL during the 20-year study period had samples sent from additional household members for T. gondii testing (32 index case-patients with acute toxoplasmosis/889 acute infections). The collection of those additional samples depended solely on the response of the referring physicians to a 1-time written request for testing of additional family members. It is possible that the response of the primary care providers to this request would have been more likely if any of those additional family/household members had symptoms suggestive of acute toxoplasmosis. In addition, the IgG-Dye test and IgM-ELISA titers of the index case-patients did not predict which families would have additional household members with acute toxoplasmosis.

Further replication of the estimated prevalence of acute T. gondii infection in consecutive US families is needed. Future studies might also compare the T. gondii serotypes among index case-patients and family members (type II vs. non–type II) (12), which could help clarify whether certain serotypes are more likely to be associated with family outbreaks. Moreover, it would be useful to screen for antibodies to sporozoite-specific antigens (13), which can provide further insight regarding the source of T. gondii infection that is more likely to be associated with acute toxoplasmosis in >1 family member (e.g., sporozoite-specific, related to contact with cat feces, vs. bradyzoite-specific, related to ingestion of undercooked meat [14]).

When a case of acute toxoplasmosis is diagnosed, screening of additional family members should be considered, especially if pregnant women or immunocompromised patients live in those households, so that appropriate preventive strategies and/or therapeutic interventions are applied. These within-family clusters of cases are not easy to predict based solely on clinical or epidemiologic information, except for situations of sharing common meal (i.e., with undercooked meat), because it is unlikely that other risk factors would be different. Thus, only routine serologic screening of household members of acutely infected persons might identify such acute T. gondii infections.