High Agreement of the BioPlex 2200 Syphilis Total & RPR Assay with the Confirmatory Treponemal Test on Weak Positive Screening Samples

LETTER

Syphilis is a well-known sexually transmitted infection (STI) caused by the bacterium Treponema pallidum subsp. pallidum. The diagnosis is usually made by serological testing, but this remains challenging. Two main laboratory testing algorithms are used, the traditional and the reverse. The first uses a nontreponemal test (NTT), such as rapid plasma reagin (RPR), as a screening assay, followed by a specific treponemal test (TT). Due to automation of the treponemal assays, the reverse algorithm has gained popularity in clinical laboratories (1).

The BioPlex 2200 Syphilis Total & RPR assay (Bio-Rad Laboratories, Hercules, CA) combines the TT and NTT simultaneously in one reaction vessel on a fully automated analyzer, providing new possibilities for syphilis diagnostics. The group of Tesfazghi et al. (2) evaluated this method in the traditional algorithm setting and obtained a performance comparable to that of the manual methods. As many laboratories, including ours, use the reverse algorithm in which specificity is a known problem in a population with low prevalence of syphilis, we evaluated the performance of the BioPlex assay on specimens with a low-positive TT and an RPR of ≤1/4 (i.e., active syphilis excluded according to the local guideline of notifiable diseases in Flanders, Belgium [3]) and found promising results.

For our retrospective study, the specific sample group with a low analytical TT value with our routine assay, the Architect Syphilis TP assay performed on the Architect i2000 SR Plus (Abbott, Japan Co., Tokyo, Japan), i.e., those with a signal-to-cutoff ratio (S/CO) between 1 and 15 and a manual RPR of 1/4 or less (BD Macro-Vue RPR card tests; Branchburg, NJ, USA), was selected from the routine clinical setting in the Ghent University Hospital (Ghent, Belgium). In total, 97 serum samples met these criteria in the period between 2016 and 2018, i.e., 0.6% of all samples tested for syphilis serology in that period in our hospital. The patient demographics for the study population are shown in Table 1. As stated by the International Union against Sexually Transmitted Infections (IUSTI) guidelines in 2014 (4), these results need to be confirmed with a second TT, which was done in the Institute of Tropical Medicine (ITM), Department of Clinical Sciences (Antwerp, Belgium), by the Serodia-TPPA (Fujirebio, Inc., Tokyo, Japan). This method was considered the gold standard, since the ITM is the national reference center for STI diagnostics. The study samples were all analyzed in batch with the BioPlex assay. Prior to this analysis, they were stored at −20°C.

TABLE 1.

Demographics on the study population, divided by sex

Characteristic	No. (%) (n = 97)
	Male	Female	Unknowna
Age (yr)
<18	0 (0)	0 (0)	Unknown
18–35	21 (35.0)	9 (47.4)	Unknown
36–50	28 (46.7)	9 (47.4)	Unknown
>50	11 (18.3)	1 (5.3)	Unknown
HIV status
Positive	24 (40)	2 (10.5)	0 (0)
Negative	32 (53.3)	16 (84.2)	18 (100)
Unknown	4 (6.7)	1 (5.3)	0 (0)
Total	60 (61.9)	19 (19.6)	18 (18.5)

^aUnknown sex and age are due to coded samples, for example, donor samples or samples from sex workers.

The results are shown in Tables 2 to 4. Of the 97 samples with an analytical value between 1 and 15 S/CO on Architect and RPR of ≤1/4, 74 were confirmed by the ITM as truly positive (76.3% agreement). Twenty-three samples tested negative with the Serodia-TPPA (23.7% disagreement).

TABLE 2.

Pairwise comparison of the treponemal test results of the Architect i2000 SR Plus versus Serodia-TPPA for the 97 samples with low analytical value on Architect i2000 SR Plus and a manual RPR of ≤4

Architect i2000 SR Plus result	No. of specimens (%) with indicated Serodia-TPPA resulta		Total (%)
	Negative	Positive
Negative	0	0	0 (0.0)
Positive	23	74	97 (100.0)
Total	23 (23.7)	74 (76.3)	97

^aThe percent agreement between the two methods was 76.3%; the percent disagreement was 23.7%.

TABLE 3.

Pairwise comparison of the treponemal test results of the Architect i2000 SR Plus versus BioPlex 2200 Syphilis Total & RPR assay for the 97 samples with low analytical value on Architect i2000 SR Plus and a manual RPR of ≤4

Architect i2000 SR Plus result	No. of specimens (%) with indicated BioPlex 2200 Syphilis Total & RPR assay resulta		Total (%)
	Negative	Positive
Negative	0	0	0 (0.0)
Positive	22	75	97 (100.0)
Total	22 (22.7)	75 (77.3)	97

^aThe percent agreement between the two methods was 77.3%; the percent disagreement was 22.7%.

TABLE 4.

Pairwise comparison of the treponemal test results of the BioPlex 2200 Syphilis Total & RPR assay versus the Serodia-TPPA for the 97 samples with low analytical value on Architect i2000 SR Plus and a manual RPR of ≤4

BioPlex 2200 Syphilis Total & RPR assay result	No. of specimens (%) with indicated Serodia-TPPA resulta		Total (%)
	Negative	Positive
Negative	16	6	22 (22.7)
Positive	7	68	75 (77.3)
Total	23 (23.7)	74 (76.3)	97

^aThe percent agreement between the two methods was 86.6%; the percent disagreement was 13.4%.

When the same samples were tested with the BioPlex system, 22 samples were negative (22.7% disagreement between the BioPlex and Architect assays). In the comparison of the BioPlex and Serodia-TPPA, 84/97 were concordant (68 positives and 16 negatives), i.e., an agreement of 86.6% and a disagreement of 13.4%.

These results suggest that a smaller amount of weak positive TTs would need confirmation by a second method using the BioPlex assay. This would implicate lower costs and a reduction in turnaround time to produce the final result on syphilis serology. A limitation of this study is that only the low-positive TT results on the Architect i2000 SR Plus were included. A more extensive, prospective study is needed to confirm our results.