Evaluation of Three Point-of-Care Tests for Detection of Toxoplasma Immunoglobulin IgG and IgM in the United States: Proof of Concept and Challenges

Abstract Background The cost of conventional serological testing for toxoplasmosis discourages universal adoption of prenatal monthly screening programs to prevent congenital toxoplasmosis. Point-of-care (POC) technology may constitute a cost-effective approach. Methods We evaluated the diagnostic accuracy of 3 Toxoplasma POC tests against gold-standard testing performed at Palo Alto Medical Foundation Toxoplasma Serology Laboratory (PAMF-TSL). The POC tests included the following: Toxo IgG/IgM Rapid Test (Biopanda) and the OnSite Toxo IgG/IgM Combo-Rapid-test that detect IgG and IgM separately, and the Toxoplasma ICT-IgG-IgM-bk (LDBIO) that detects either or both immunoglobulin IgG/IgM in combination. Samples were selected from PAMF-TSL biobank (n = 210) and Centers for Disease Control and Prevention Toxoplasma 1998 Human Serum Panel (n = 100). Based on PAMF-TSL testing, Toxoplasma-infection status was classified in 4 categories: acute infections (n = 85), chronic infections (n = 85), false-positive Toxoplasma IgM (n = 60), and seronegative (n = 80). The POC testing was performed in duplicate following manufacturer’s instructions by investigators blinded to PAMF-TSL results. Sensitivity and specificity were calculated. Results A total of 1860 POC tests were performed. For detection of Toxoplasma IgG, sensitivity was 100% (170 of 170; 95% confidence interval [CI], 97.8%–100%) for all 3 POC kits; specificity was also comparable at 96.3% (77 of 80; 95% CI, 89.5%–98.9%), 97.5% (78 of 80; 95% CI, 91.3%–99.6%), and 98.8% (79 of 80; 95% CI, 93.2%–99.9%). However, sensitivity for detection of Toxoplasma IgM varied significantly across POC tests: Biopanda, 62.2% (51 of 82; 95% CI, 51.4%–71.9%); OnSite, 28% (23 of 82; 95% CI, 19.5%–38.6%); and LDBIO combined IgG/IgM, 100% (82 of 82; 95% CI, 95.5%–100%). Diagnostic accuracy was significantly higher for the LDBIO POC kit. The POC kits did not exhibit cross-reactivity for false-positive Toxoplasma-IgM sera. Conclusions The 3 evaluated POC kits revealed optimal sensitivity for Toxoplasma-IgG antibodies. The LDBIO-POC test exhibited 100% sensitivity for the combined detection of IgG/IgM in acute and chronic Toxoplasma infection. Biopanda and Onsite POC tests exhibited poor sensitivity for Toxoplasma-IgM detection.

Toxoplasma gondii, the causative agent of toxoplasmosis, is an obligate intracellular protozoan of worldwide distribution that infects one third of the world's human population [1]. When T gondii infection occurs in immunocompetent patients, it typically follows an asymptomatic and benign clinical course, whereas in immunocompromised hosts, toxoplasmosis can lead to life-threatening manifestations if diagnosed or treated late [2]. Of note, severe presentations of the acute infection in immunocompetent patients have been reported in association with atypical genotypes of the parasite [3,4]. Primary T gondii infection that occurs during pregnancy or within 3 months before conception can result in congenital toxoplasmosis (CT) if not promptly diagnosed and treated [5]. Congenital toxoplasmosis can have devastating consequences for the fetus including stillbirth, neurological sequelae, and severe ocular disease [5]. The risk of CT and the severity of its clinical manifestations are influenced mainly by the gestational age at the time of maternal infection, T gondii strain, and antenatal anti-Toxoplasma treatment [6]. Maternal T gondii infections are usually asymptomatic; approximately half of T gondii-infected pregnant women who gave birth to infants with CT did not report any risk factors or any symptoms suggestive of infection [7]. Thus, timely diagnosis of T gondii infection during pregnancy relies largely on serological testing throughout gestation in the context of universal prenatal screening programs [8].
Maternal screening programs for CT have been widely adopted in France, Austria, Germany, and other European countries [9][10][11]. These programs consist of serial serological screening (eg, monthly in France, bimonthly in Austria, and every 2-3 months in Germany) of Toxoplasma-seronegative pregnant women coupled with prompt initiation of anti-toxoplasma treatment upon maternal seroconversion [9][10][11][12][13]. Several observational studies over the past decade concluded that antenatal anti-Toxoplasma therapy reduces rates of mother-to-child transmission [9][10][11]14] and mitigates the severity of clinical manifestations among congenitally infected infants [9,11,[15][16][17]. Early initiation of treatment, within 3-4 weeks from maternal infection, is also critical [14,16]. However, despite the fact that CT is a preventable and treatable condition, most pregnant women in the United States and worldwide are not routinely screened and treated accordingly for Toxoplasma infections [18]. In the United States, CT is usually diagnosed upon appearance of clinical and/or ultrasound findings in the fetus or at birth [5]. Consequently, reported cases of CT in the United States have higher rates of severe clinical manifestations at birth (ie, chorioretinitis, intracranial calcifications, and hydrocephalus) compared with their European counterparts [19][20][21].
Because approximately 5 to 8 serological tests would be performed in each Toxoplasma-seronegative woman, a major factor impeding uptake of monthly maternal screening programs worldwide is high cost of conventional serological testing, which is generally performed at commercial laboratories using automated platforms [22,23]. Indeed, a decision-analytic economic model in the United States showed that universal monthly maternal screening can be cost saving in the United States, even with an incidence of acute maternal T gondii infection as low as 0.2 per 1000 pregnant women, if screening test costs ~$12.00 per test, a figure far below conventional laboratory-based testing cost [24]. Moreover, in Europe, and France in particular, where Toxoplasma-seroprevalence in the general population has declined (from 80% in 1960 to 36.7% in 2010) [25], numbers of seronegative pregnant women necessitating screening has increased, thereby endangering further financial viability of screening programs [25,26]. Hence, alternative diagnostic tools are needed for toxoplasmosis diagnostics.
Point-of-care (POC) rapid tests have improved access to efficient diagnostics across infectious diseases and global health, especially in low-income country settings [27,28]. Point-ofcare technology for antibody detection usually consists of immunochromatographic (ICT) membrane-based assays that can be performed without sophisticated laboratory infrastructure. Previous studies have shown optimal diagnostic performance for the Toxoplasma ICT IgG-IgM (LDBIO Diagnostics, Lyon, France) POC kit when compared with reference-serological methods in France [29,30] and the United States [31]. However, direct comparisons of performance of various POC kits against reference-standard methods used in the United States, where T gondii strains are more diverse, have not been previously performed.
In this study, we sought to evaluate diagnostic accuracy for 3 Toxoplasma-IgG-IgM POC kits using a large number of sera from US patients previously tested by reference methods available at Palo Alto Medical Foundation, Toxoplasma Serology Laboratory ([PAMF-TSL] http://www.pamf.org/serology/), the reference laboratory for study and diagnosis of toxoplasmosis in the United States.

Serological Samples
A total of 310 patient serum samples tested at PAMF-TSL were included: 210 were obtained from the PAMF-TSL biobank, and 100 from the Centers for Disease Control and Prevention (CDC) Toxoplasma 1998 Human Serum Panel (CDC-HSP) ( Table 1). The PAMF-TSL samples were selected from archived specimens submitted between from February 27, 2013 to June 23, 2017. The CDC-HSP samples were selected from a biobank that contains known positive and negative sera available to researchers and laboratories in the United States. These samples are obtained by the CDC to help evaluate accuracy

Serological Testing at Palo Alto Medical Foundation Toxoplasma Serology Laboratory
Toxoplasma-IgG testing was performed using Sabin-Feldman IgG DT, the gold-standard method for Toxoplasma-IgG detection in the United States. The DT is a very sensitive and specific test in which live tachyzoites are lysed in the presence of complement and T gondii-specific IgG derived from patients' samples. Toxoplasma-IgM testing was performed using a laboratory-developed, double-sandwich IgM-ELISA that uses sonicated T gondii antigen prepared from live tachyzoites (www. pamf.org/serology/clinicianguide.html). When indicated (eg, sera with positive Toxoplasma-IgM test results), additional tests such as Toxoplasma-IgG avidity, differential agglutination (AC/HS), Toxoplasma-IgA, and Toxoplasma-IgE, were performed by PAMF-TSL [23]. According to PAMF-TSL results, sera from PAMF-TSL biobank and CDC were classified in 4 groups, described above under "Serological Samples": (1) acute Toxoplasma infection; (2) chronic Toxoplasma infection; (3) seronegative samples uninfected with T gondii; and (4) FP Toxoplasma-IgM.

Diagnostic Accuracy of Point-of-Care Testing for Toxoplasma-Immunoglobulin M
Sensitivity calculations for detection of Toxoplasma-IgM was performed for 82 (instead of 85) Toxoplasma-IgM-positive serum samples because 3 of 35 positive Toxoplasma-IgM samples from the CDC-HSP serum panel were dilutions of 3 true Toxoplasma-IgM-positives specimens, and their corresponding PAMF-TSL IgM-ELISA results were <2.0 (0.5, 0.8, and 1.6 units). Therefore, these 3 samples were excluded from sensitivity calculation in results reported to us by the CDC (Tables 2  and 3

Cross-Reactivity of Point-of-Care Kits With Nonspecific Immunoglobulin M
All 3 POC kits showed negative results for IgM (0 of 60) when tested against FP Toxoplasma-IgM samples (nonspecific IgM) ( Table 5). In 3 of 60 (5%) samples, the Biopanda POC kit tested falsely positive for IgG. Of note, IgM-ELISA titers from patients with nonspecific IgM were significantly lower compared with

Interobserver and Intertest Agreement
There were instances when Biopanda and OnSite POC cassettes displayed poorly visualized testing bands (ie, very faint colored bands-ambiguous bands) (Supplementary Material, Figure S1). Ambiguous bands were seen in 44 of 620 (   Toxoplasma-IgM results. Hence, our results support use of the LDBIO-POC platform as a simple, low-cost, and rapid diagnostic tool capable of efficiently detecting Toxoplasma-IgG and/ or Toxoplasma-IgM antibodies. Our study also supports the LDBIO-POC test as an initial serological screening test in programs for universal Toxoplasma prenatal testing. The underperforming diagnostic accuracy for Toxoplasma-IgM detection noted for the 2 POC tests that were designed to detect IgG and IgM separately deserves further discussion. There are some technical differences between these 3 POC tests. The Biopanda and OnSite POC tests use recombinant T gondii antigens for binding to specific anti-Toxoplasma antibodies (IgG and IgM, respectively), whereas the LDBIO POC test uses as antigen whole-cell lysates of tachyzoites from the T gondii RH Sabin Type I strain. Additional mechanistic differences on lateral flow assays design may also account for the higher diagnostic performance observed with the LDBIO POC kit (see Supplementary Material). Biopanda and OnSite offer in their POC cassettes detection of Toxoplasma-IgM antibodies in a distinct testing band (M band, Figure 1). False-negative results for IgM may lead to delay confirmatory testing, missing the opportunity for timely initiation of anti-Toxoplasma therapy in seroconverting pregnant women. In addition, the Biopanda and the OnSite POC tests exhibited the presence of ambiguous bands, defined as poorly visualized testing bands with very faint color, which were seen in 7.1% (44 of 620) and 4.5% (28 of 620) of the tests performed by Biopanda and OnSite, respectively. The above, added to an excessive number of FP results (11.5% for Biopanda IgM), represents a significant drawback that, in clinical practice, may lead to confusion and unnecessary confirmatory testing at expenses of increasing testing cost and patient's anxiety. The failure of the Biopanda (62.2% sensitivity) and OnSite POC tests (28% sensitivity) to optimally detect Toxoplasma IgM represents a serious handicap for the implementation of such POC tests in universal screening programs during pregnancy.
The LDBIO Toxoplasma POC kit showed robust performance in our study, with 100% sensitivity for the combined detection of IgG and IgM detection and low rate of FP IgG/IgM POC test results (1.3%, 1 of 80 in IgG/IgM seronegative sera). Of note, the LDBIO reports Toxoplasma-IgG and Toxoplasma-IgM antibodies combined in a single testing band (T band); therefore, independent calculation of IgM sensitivity and specificity for this POC test, without taking into account the presence of IgG, was not possible (Tables 2 to 3). The diagnostic accuracy for LDBIO has been documented in prior studies using various reference standard methods for Toxoplasma-IgG and IgM detection [29][30][31]. To date, at least ~2000 sera and whole blood samples from France, United States, and Morocco have been tested (including the 310 sera from this study). Chapey et al [29] evaluated the LDBIO POC test in 400 sera from France and demonstrated diagnostic sensitivity and specificity of 97% and 96%, respectively, for the combined detection of Toxoplasma IgG-IgM when compared with the fully automated chemiluminescence Architect test (Abbott North, Chicago, IL  [34]. In light of accumulated evidence for LDBIO POC test's accuracy in comparison with reference-standard methods in the United States and France, the LDBIO POC kit can be deemed as an optimal low-cost technology for population-based serological screening and/or universal monthly prenatal screening. This POC test has been shown to perform well independently of implicated T gondii strain and across a wide range of Toxoplasma-IgG antibody titers [31]. If cost of LDBIO POC kit can be maintained at <$5.00, this assay may become an attractive option for first-line testing across maternal screening programs. Our study has some limitations. Despite large sample size and inclusion of sera from acute and chronic Toxoplasma infection and FP IgM results, we did not include sera from patients with early T gondii infection (early seroconversion) to assess for IgM diagnostic accuracy in the absence of Toxoplasma IgG. These sera are scarce given lack of routine prenatal screening for Toxoplasma infection in the United States, impeding access to patients with documented seroconversion at early stages of infection. Therefore, in our study, IgM titers from patients Our study was a proof-of-concept study and the first to directly compare 3 commercial POC tests for Toxoplasma diagnostics. As such, our results have notable implications for obstetricians, pediatricians, neonatologists, infectious diseases specialists, and others including policy makers in public health. As a neglected disease, CT represents a significant burden to healthcare systems around the world, with 190 000 infants per annum affected worldwide carrying 1.2 million disability-adjusted life years [35][36][37]. Integration of POC platforms into Toxoplasma diagnostics holds enormous promise given its critical elements such as rapid turnaround time, simplicity, low cost, and completion of testing during same clinical encounter [38]. However, careful evaluation of POC test accuracy and diagnostic performance is a pivotal step before any consideration for clinical use. The World Health Organization has proposed the "ASSURED" criteria for the ideal POC assay: Affordable, Sensitive, Specific, User-friendly, Rapid/robust, Equipment-free and Deliverable to users [28]. In this frame, our study represents the initial stage to move forward POC diagnostics in toxoplasmosis. Given their low complexity and minimal risk for incorrect results, POC kits with robust analytical performance such as the LDBIO POC test, which has also been shown to have excellent diagnostic accuracy when tested with whole blood collected directly from a fingerstick (via a capillary tube) [34], can receive Clinical Laboratory Improvement Amendments waivers, enabling them to be used in POC settings and thereby contribute to mitigate costs associated with serological testing in maternal screening programs [24]. Despite enthusiasm, several questions regarding the application of POC diagnostics in clinical grounds remain to be addressed, and future studies are needed to prospectively evaluate effectiveness of POC test-based prenatal screening strategies, including cost effectiveness in individual country settings, impact on clinical outcomes, and their integration to clinical workflows in perinatal care [39]. Work is in progress for the evaluation of POC test-based prenatal screening integration in the clinical workflow in prenatal care (Thrasher Research Fund award no. 13798).
Positive POC test results indicating acquisition of an acute Toxoplasma infection during gestation (eg, seroconversion from IgG − /IgM − serostatus to an IgM + or IgG + or combined IgG + /IgM + status between 2 consecutive screenings) would always require confirmatory testing with conventional laboratory-based methods to confirm the diagnosis and guide clinical management. Moreover, positive POC test results at the first prenatal care visit (either IgG + /IgM − or positive for the combined IgG/IgM detection) also require confirmatory testing, to confirm that these reflect chronic infection, acquired before gestation. Pregnant women found to be chronically infected early in gestation would not need additional monthly screening for the rest of their pregnancy because they are not at risk of transmitting the infection to their fetus, unless severely immunocompromised and not on prophylactic therapy. Despite the need for confirmatory testing for positive results, POC test-based prenatal screening programs are still cost saving because they significantly limit the number of women who will need such confirmatory testing (even in settings with seropositivity rates among pregnant women as high as 50%, a POC test-based screening program can cut the need for laboratory-based screening by half).

CONCLUSIONS
In conclusion, we demonstrated that all 3 POC tests (Biopanda, OnSite, and LDBIO POC tests) showed optimal diagnostic sensitivity and specificity for Toxoplasma-IgG detection across different Toxoplasma-infection states. However, the LDBIO POC test showed 100% sensitivity for the combined detection of IgG/IgM in acute and chronic Toxoplasma infection, whereas Biopanda and OnSite POC tests showed poor diagnostic sensitivity for Toxoplasma IgM. An accurate, low-cost, and reliable POC diagnostic test for Toxoplasma infection can benefit patients and mitigate the cost of serological testing in universal prenatal screening programs for Toxoplasma infections.

Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Potential conflicts of interest. R. M. and J. G. M. performed a literature review for Sanofi-Pasteur during the time this manuscript was prepared; no compensation was received for this review. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.