Background: Current methods for assessing condom use are prone to uncertainty and bias. As most condoms are made from natural rubber latex (NRL), NRL proteins are an excellent candidate marker for condom use. The purpose of this study was to determine if an available assay could accurately discriminate latex and non-latex products.
Methods: We used the ASTM D6499 standard test method to quantify antigenic NRL protein concentrations in extracts from latex and non-latex products (n=166). We tested major latex condom brands and varieties (non-lubricated, lubricated, lubricated plus spermicide), non-latex condoms, vaginal lubricants, transvaginal ultrasound probe covers, and buffer (PBS). Per ASTM protocol, condoms were cut up and soaked in PBS (5mL/g) for 120 minutes at 25°C. Extracts were also prepared by soaking the outer surface of the intact condom in PBS (10mL/g) for 30 minutes at 37°C. Lubricants were prepared by dipping a swab in the product and extracting in PBS at 25°C.
Results: Extracts prepared per protocol yielded inconsistently higher NRL protein concentrations, which were not critical for distinguishing latex and non-latex products. The median concentration was 0.879µg/mL for latex products (n=106, range: 0.056 - 25.090µg/mL) and below the limit of detection (b.d., <0.008µg/mL) for non-latex products (n=60, range: b.d - 0.127µg/mL). Some non-latex products (n=13) yielded very low false-positive results. Using a cut-off of ≥0.1µg/mL afforded 95% sensitivity and 98% specificity. Adding Tween-20 to the extracts further increased specificity. High area under the ROC curve (0.996, CI: 0.991-1.00) and a high Youden Index (0.94) provide evidence of the test’s accuracy.
Conclusions: NRL proteins are detectable in extracts from latex products and almost always absent in extracts from non-latex products. A human study is necessary to establish NRL proteins as an objective marker of latex condom use. Such a marker would significantly strengthen contraceptive and HIV/STI research methods.