Comparison of Different Commercially Available Enzyme-Linked Immunosorbent Assays with Immunofluorescence Test for Detection of Phase II IgG and IgM Antibodies to Coxiella burnetii

Several commercially available enzyme-linked immunosorbent assays (ELISAs) for the detection of phase II IgG or IgM antibodies against Coxiella burnetii were compared. In addition, an indirect immunofluorescence test was used as a confirmation test. In all, 70 serum samples for IgG and 43 serum samples for IgM were tested. The ELISAs showed large differences in sensitivity and specificity, which led to a partially high ratio of false-negative determinations.

ABSTRACT

Several commercially available enzyme-linked immunosorbent assays (ELISAs) for the detection of phase II IgG or IgM antibodies against Coxiella burnetii were compared. In addition, an indirect immunofluorescence test was used as a confirmation test. In all, 70 serum samples for IgG and 43 serum samples for IgM were tested. The ELISAs showed large differences in sensitivity and specificity, which led to a partially high ratio of false-negative determinations. The most convincing test was PanBio from Abbott, which unfortunately can only test IgG but not IgM.

INTRODUCTION

Q fever is a worldwide zoonosis caused by the obligate intracellular Gram-negative and highly infectious bacterium Coxiella burnetii. Humans are usually infected through the inhalation of aerosols contaminated by secretions of infected animals (1). In humans, Q fever either is subclinical or results in a variety of flu-like symptoms, including fever, headache, myalgia, and other respiratory complications (2, 3). A total of 1 to 5% of acute Q fever patients develop chronic Q fever (4). This chronic form may manifest as endocarditis, granulomatous hepatitis, or vascular infections with a high morbidity and has a mortality rate of around 60% if left untreated (5, 6). The acute and chronic phases can be distinguished by serological tests. During the acute phase, antibodies against the phase II antigen are predominantly produced, whereas high titers of phase I antibodies are typical for chronicity (7). An indirect immunofluorescence test (IFAT) is the only method available to quantify the antibody titers of all immunoglobulin subclasses and to diagnose a chronic infection. In the early acute phase, PCR is also an important method for diagnosing acute Q fever (8). Thus, highly sensitive diagnostic tools are necessary for the detection of early infections. Serologic tests for the detection of specific antibodies against phase I and II antigens are usually the methods of choice.

For a primary screening, an enzyme-linked immunosorbent assay (ELISA) is often used. To confirm positive results, an IFAT is performed as well. This test is the reference method for serological Q fever diagnosis; it is used to avoid false-positive results, and it is generally accepted as the gold standard for a serological diagnosis of Q fever (7, 9–11). For both systems, different commercial kits are available; these need to be tested and compared occasionally. In the 1990s and 2000s, several laboratories tested diverse ELISAs and IFATs mainly focused on the detection of phase II IgM antibody titers (12–15).

We sought here to evaluate currently available ELISA kits in comparison to the reference method, IFAT, and to determine their sensitivities and specificities in the detection of phase II IgG and IgM.

MATERIALS AND METHODS

Serum samples.

For the comparative study, a total of up to 70 sera were included to determine phase II IgG. A total of 43 sera were used for phase II IgM detection. All sera were submitted to the German Consiliary Laboratory of Coxiella burnetii in the context of a Q fever outbreak survey and for follow-up. All patients visited a sheep farm in which the outbreak occurred with one-point source. A total of 228 persons were infected (16). All available sera were tested 3 months after the outbreak. For routine diagnosis, the ELISAs from Virion/Serion and the IFAT from Focus Diagnostics were used. For the study, all serum samples with sufficient volumes were tested by all ELISAs in parallel. Positive results were confirmed with the IFAT.

Kits and reference assay.

ELISA kits from five manufacturers were used for the comparison of the detection of phase II IgG and IgM antibodies: the Serion ELISA classic Coxiella burnetii IgG/IgM kit (Institut Virion/Serion GmbH, Würzburg, Germany); the Panbio Coxiella burnetii (Q fever) IgG ELISA kit (Alere/Abbott, Chicago, IL); the Nova Lisa, Coxiella burnetii (Q fever) phase 2 IgG/IgM kit (Mikrogen GmbH, Neuried, Germany); the Coxiella burnetii phase II IgG/IgM kit (Biomed Labordiagostik GmbH, Oberschleissheim, Germany); and the Coxiella burnetii (Q fever) phase 2 IgG/IgM ELISA kit (IBL International, GmbH, Hamburg, Germany). All assays were performed according to the manufacturer’s instructions (Tables 1 and 2) and were used for the qualitative determination of IgG and IgM antibodies against Coxiella burnetii phase II. As indicated by the manufacturer, the Rf absorption was carried out for 15 or 30 min (see Table 2).

TABLE 1.

Instruction manuals for detection of phase II IgG antibodies in serum samples^a

Factor or step	Virion/Serion	Biomed	Mikrogen	IBL	Abbott
Serum dilution	1:500	1:500	1:100	1:100	1:100
Serum incubation	37°C, 1 h	37°C, 1 h	37°C, 1 h	37°C, 1 h	37°C, 30 min
Wash step	4 × 300 μl	4 × 300 μl	3 × 300 μl	3 × 300 μl	6 × 350 μl
Incubation conjugate	37°C, 30 min	37°C, 30 min	RT, 30 min	RT, 30 min	37°C, 30 min
Wash step	4 × 300 μl	4 × 300 μl	3 × 300 μl	3 × 300 μl	6 × 350 μl
Substrate incubation	37°C, 30 min	37°C, 30 min	RT, 15 min, dark	RT, 15 min, dark	RT, 10 min
Stop	Immediately	Immediately	Immediately	Immediately	Immediately
Readout (nm)	405/650	405/650	450/620	450/620	450/620

^aRT, room temperature.

TABLE 2.

Instruction manuals for detection of phase II IgM antibodies in serum sample^a

Factor or step	Virion/Serion	Biomed	Mikrogen	IBL
Rf absorbent dilution (buffer)	1:4
Serum dilution (buffer)		1:51	1:100	1:100
Serum dilution (Rf absorbent)	1:100	1:1
Rf incubation	RT, 15 min	RT, 30 min
Serum incubation plate	37°C, 1 h	37°C, 1 h	37°C, 1 h	37°C, 1 h
Wash step	4 × 300 μl	4 × 300 μl	3 × 300 μl	3 × 300 μl
Incubation conjugate	37°C, 30 min	37°C, 30 min	RT, 30 min	RT, 30 min
Wash step	4 × 300 μl	4 × 300 μl	3 × 300 μl	3 × 300 μl
Substrate incubation	37°C, 30 min	37°C, 30 min	RT, 15 min, dark	RT, 15 min, dark
Stop	Immediately	Immediately	Immediately	Immediately
Readout (nm)	405/650	405/650	450/620	450/620

^aRT, room temperature.

A reference assay IFAT was carried out with a Q fever IFA kit (Focus Diagnostics, Cypress, CA), adhering to the manufacturer’s instructions and as reported previously (8, 10). The IFAT identified both IgG and IgM phase II antibodies against Coxiella burnetii. The serum samples were considered positive when IgG and/or IgM titers of 1:16 or higher were observed. The IFAT identified 38 of 70 serum samples positive for IgG and 29 of 43 serum samples positive for IgM.

The IFAT has a high sensitivity and specificity of 100 and 99%, respectively, compared to the complement fixation test and a relative sensitivity (phase I) of 100% (14/14) and a relative sensitivity (phase II) of 100% (16/16) compared to U.S. State Health Department’s IFA test, using data provided directly by the manufacturer (Focus Diagnostics).

Statistical analysis.

To calculate sensitivity and specificity, borderline test results in the ELISAs were considered negative. The sensitivity and the specificity and their respective 95% confidence intervals were calculated, considering IFAT phase II IgG and IFAT phase II IgM as the gold standard. P values were calculated by using a chi-square test.

RESULTS

Phase II IgG antibody detection.

All ELISA positively screened patients were subsequently confirmed with the IFAT. Thus, no false-positive results occurred and specificity in all tests was 100%. Focusing on the occurrence of false-negative results, which would pose a problem in diagnosis, a large variety among different assays was observed. In contrast to the Virion/Serion kit used for the initial screening, the other tests also detected positive serum samples, which were confirmed by the IFAT. We observed that the PanBio kit from Abbott was the only one with sensitivity of 100% with no false-negative tested sera. The ELISAs from Virion/Serion, Mikrogen, and IBL showed sensitivities of 68.42, 68.42, and 76.32%, respectively. The lowest sensitivity (55.26%) was demonstrated by the test from Biomed (Table 3).

TABLE 3.

Phase II IgG: statistical analysis of five commercial ELISA kits compared to IFAT^a

ELISA test phase II IgG	Total no. of tests	No. of IFAT-positive results	% (95% CI)		P (χ2 test)
			Sensitivity	Specificity
Virion/Serion	70	38	68.42 (51.35–82.5)	100.00 (89.11–100.00)	3.59E−09
Biomed	68	38	55.26 (38.3–71.38)	100.00 (88.43–100.00)	9.70E−07
Mikrogen	68	38	68.42 (51.35–82.5)	100.00 (88.43–100.00)	8.17E−09
IBL	68	38	76.32 (59.73–88.56)	100.00 (88.43–100.00)	2.65E−10
Abbott	70	38	100.00 (90.75–100.00)	100.00 (89.11–100.00)	5.93E−17

^aThe values for each test are reported as the determined Q fever phase II IgG level versus the IFAT result. 95% CI, 95% confidence interval.

Phase II IgM antibody detection.

In contrast to the phase II IgG detection, both false-negative and false-positive results were observed in the detection of phase II IgM antibodies. A specificity of 64.26% was observed with the Biomed, Mikrogen, and IBL tests, whereas the specificity of the Virion/Serion kit was the lowest at 50.00%. On the other hand, the assays from Virion/Serion and Biomed showed high sensitivities of 93.10 and 89.66%, respectively. However, the kits provided by Mikrogen and IBL demonstrated sensitivities of 79.31 and 75.89%, respectively (Table 4).

TABLE 4.

Phase II IgM: statistical analysis of four commercial ELISA kits compared to IFAT^a

ELISA test phase II IgM	Total no. of tests	No. of IFAT-positive results	% (95% CI)		P (χ2 test)
			Sensitivity	Specificity
Virion/Serion	43	29	93.1 (77.23–99.15)	50.00 (23.04–76.96)	1.13E−03
Biomed	43	29	89.66 (72.65–97.81)	64.26 (35.14–87.24)	2.20E−04
Mikrogen	43	29	79.31 (60.26–92.01)	64.26 (35.14–87.24)	4.94E−03
IBL	43	29	75.89 (56.46–89.7)	64.26 (35.14–87.24)	1.07E−02

^aThe values for each test are reported as the determined Q fever phase II IgM level versus the IFAT result. 95% CI, 95% confidence interval.

DISCUSSION

In the present study, different ELISA kits were tested on a homogeneous group of persons at risk of being infected with Coxiella burnetii, since these individuals had been exposed to a Q fever outbreak. All sera, which were obtained from the German Consiliary Laboratory of Coxiella burnetii, that had a sufficient volume were included in the study.

As described above, the IFAT was used as the reference method because it is generally accepted as the gold standard for the serological diagnosis of Q fever (9). Furthermore, the IFAT is the only method used to quantify antibody titer and therefore has the advantage of identifying patients at risk of developing chronic Q fever (17).

The ELISA is also an approved test for serological Q fever diagnostics. ELISAs are not only easy and quick to perform, but also, only a single dilution of serum is necessary, and ELISAs are generally less expensive than IFAT and therefore particularly well suited to screening a large number of samples. The sensitivity and specificity of many ELISAs have been screened by different working groups over the years, especially for phase II IgM, showing a variety of unreliable results (9, 18).

In the present study, both phase II IgM antibodies and phase II IgG antibodies were tested. The sensitivity for IgG phase II antibody detection varied between 55 and 100% for all ELISAs, whereas the specificity was an even 100% for all assays. It has to be noted that a phase II IgG ELISA is limited by the fact that it cannot differentiate between acute and chronic infections. Thus, the phase II IgG antibody is a good supporting parameter in Q fever diagnostics, but it is much more important in addressing epidemiological questions.

As mentioned above, ELISAs are a very useful tool in an outbreak situation where there are many serum samples to screen and to identify persons at risk. Therefore, an ELISA should have a high sensitivity rate of around 100% (especially for the phase II IgM antibody), whereas a high specificity is not mandatory. For the first 2 weeks of an outbreak, the IgM phase II is the most important parameter and would be much more useful in combination with a PCR.

The present study clearly shows that ELISAs have an important role in Q fever diagnostics, but they are not only limited in the distinction between acute, chronic, and past disease but also show large differences in sensitivity and specificity. Both our study and other studies demonstrate that the IFAT remains the most reliable test system for the serological diagnosis of Q fever (9, 12–15).

The perfect specificity in IgG phase II detection contrasts with a low IgM phase II specificity. For none of the tested ELISAs did the specificity exceed 65%, but they all showed a high sensitivity of between 76 and 96%.

To conclude, the best system to detect phase II IgG antibodies against Coxiella burnetii is the PanBio ELISA from Abbott, which is characterized by high sensitivity and specificity. The other tests had high specificity for phase II IgG detection but were unreliable with regard to other parameters. Thus, we feel the PanBio IgG ELISA from Abbott is the most accurate and effective of the tests we compared.