False-negative syphilis treponemal enzyme immunoassay results in an HIV-infected case-patient

Introduction

Routine syphilis screening is currently recommended for high-risk individuals, including HIV-infected persons and men who have sex with men (MSM), by both the Centers for Disease Control and Prevention (CDC) ¹ and the US Preventive Services Task Force. ² The CDC recommends syphilis screening with a non-treponemal test followed by a confirmatory, more specific, treponemal test. However, widely available automated enzyme immunoassays (EIAs) have led many laboratories to adopt a “reverse sequence” syphilis screening (RSSS) algorithm. ³ With RSSS, serum is initially tested with a treponemal EIA and followed by a quantitative non-treponemal test, if the EIA is reactive. As treponemal antibodies appear earlier, RSSS should have greater sensitivity in addition to greater specificity. We present a case report involving an HIV-infected MSM highlighting potential problems with the RSSS approach.

Case report

The case-patient was a 37-year-old HIV-infected male seen by a primary care physician in September 2015 with a desquamating rash on his palms, soles of his feet, scrotum, buttocks, and lower extremities. He was not receiving antiretroviral therapy and his CD4 cell count was <50 cells/mm³. Secondary syphilis was suspected. Syphilis serological testing was done following CDC’s recommended algorithm. His qualitative rapid plasma reagin (RPR) test was reactive, as was his quantitative Venereal Disease Research Laboratory (VDRL) test with a titer of 1:32. However, his confirmatory treponemal test (Trep-Sure EIA, Phoenix Biotech, Ontario, Canada) was non-reactive. Based on his nonreactive treponemal test, the patient was diagnosed with a “fungal” skin infection and was not treated for syphilis. Follow-up evaluation on day 7 revealed no improvement. He was retested for syphilis, and his results were unchanged. The patient was referred to the Hawaii Department of Health (HDOH) Sexually Transmitted Disease (STD) Clinic and was evaluated on day 15. Patient denied history of penile, oral, or anal ulcers. He was treated with 2.4 mU benzathine Penicillin G intramuscularly for presumptive secondary syphilis. His skin rash abated after treatment. The patient was seen at the HDOH STD Clinic as follow-up 10 weeks post-treatment and syphilis testing was repeated. His EIA remained non-reactive, while the quantitative VDRL showed a four-fold titer decrease to 1:8. The CDC was contacted, and a serum aliquot from his 10-week post-treatment visit was sent for further testing. Repeat testing by CDC revealed a quantitative RPR reactive with a titer of 1:16, non-reactive Trep-Sure EIA; however, a Treponema pallidum particle agglutination (TP-PA) test was reactive.

Discussion

Our case-patient provides clear evidence of a false-negative EIA test result. While CDC notes that “interpretation of treponemal and nontreponemal serological tests for persons with HIV infection is the same as for the HIV-uninfected patient,” they acknowledge that “rare unusual serologic responses have been observed among persons with HIV infection who have syphilis.” ¹ Most reports of unusual serologic responses in HIV-infected patients have been of biological false-positive non-treponemal test results^4–7; however, reports of false-negative non-treponemal and treponemal-specific results have also been documented.^8–10 An earlier study suggested that false-negative treponemal test results in the setting of HIV infection should be suspected when non-treponemal tests are reactive at higher titers (i.e. ≥1:8). ¹¹ Our case-patient presented in such a manner. His four-fold decreased VDRL antibody titer (1:32 to 1:8) after treatment reflects an appropriate clinical response. ¹ It is not unexpected that the RPR and VDRL titers run on identical serologic specimens differed. Sequential quantitative non-treponemal serological tests in a single patient need to be performed using the same test, as RPR titers tend to be slightly higher than VDRL titers. ¹ While treponemal tests are considered more specific and sensitive than non-treponemal tests, two recent studies have noted the Trep-Sure EIA to be less sensitive than the RPR ¹² (when screening asymptomatic high-risk STD clinic patients) and VDRL ¹³ (when screening patients suspected to have primary syphilis). It is possible that other EIAs may have reacted with our case-patient’s sera as reactivity depends on the antigens used in the assay. The TP-PA test is currently considered the most valid (sensitive and specific) confirmatory treponemal test ¹⁴ and can be used to resolve discordant test results. ⁶ The CDC has noted that “… if sera is TP-PA nonreactive, syphilis is unlikely.” ³

Our case-patient, whose low CD4 cell count reflects impaired immunity, demonstrates clear evidence of a false-negative EIA test result in the setting of secondary syphilis. False-negative EIA results can adversely impact interpretation of syphilis serological testing with both the traditional and RSSS testing algorithms. However, in the setting of a nonreactive EIA, the presence of high reactive non-treponemal titers with the traditional approach, or symptoms suggestive of syphilis, should alert a clinician to a possible false-negative result and the need for additional confirmatory testing. While treponemal tests are widely considered more sensitive and specific than non-treponemal tests, one must keep a high index of suspicion for syphilis in both HIV-infected and uninfected patients.