David J. Ledden

Evaluation of the CLINITEST® Human Chorionic Gonadotropin (hCG) Pregnancy Test for Susceptibility to the Hook Effect by the hCG β Core Fragment

To the Editor: In a recent report in Clinical Chemistry , Nerenz et al. described a screening method to evaluate point-of-care human chorionic gonadotropin (hCG)1 devices for susceptibility to the hook effect by the hCG β core fragment (hCGβcf) (1). Among the 11 devices they evaluated was the CLINITEST® hCG pregnancy test. Nerenz et al. did not perform the testing for the CLINITEST hCG product themselves but opted to send screening samples to a colleague to perform the test on the CLINITEK® Status+ analyzer according to the manufacturers instructions. The protocol required that the 3 screening samples be run in duplicate on a single instrument with 1 reagent lot. The results of this testing are shown in Table 1. View this table: Table 1. CLINITEST hCG performance screening method samples. The interpretation of these results by Nerenz et al. was that the test detected intact hCG at 5 × 102 pmol/L, detected hCGβcf at 5 × 105 pmol/L, and gave a false-negative and a borderline result for the sample containing 5 × 102 pmol/L intact hCG + 5 × 105 pmol/L hCGβcf, indicating that this sample is at or near the threshold of a combined hCG and hCGβcf high-dose hook …

Response to a third-party evaluation of Clinitest® hCG sensitivity to various isoforms of human chorionic gonadotropin.

The Clinitest human chorionic gonadotropin (hCG) pregnancy test from Siemens Healthcare Diagnostics (MA, USA) is a cassette-based lateral flow device read on the CLINITEK Status instrument. This instrument measures the relative intensity of a test, reference and control line at a fixed time and, if the intensities of the reference and control lines are within certain values, will provide an appropriate test result. Thus, samples with hCG concentrations greater than 25 mIU/ml will be reported as positive, and samples with concentrations of less than 25 mIU/ml will be reported as either negative (no measured test line) or borderline (a detectable test line but of less intensity than that of a 25 mIU/ml sample). In Laurence Cole’s recent publication, the author observes urine sensitivities for the Siemens Clinitest hCG assay of 50 mIU/ml (first Reference Reagent) hCG, 100 mIU/ml hyperglycosylated hCG (hCG-H), >500 mIU/ml free hCG-b and >500 mIU/ml b-core fragment (hCG-bcf ) [1]. Lot-specific information in Cole’s paper was not provided but sensitivities were derived, as the author indicates, from the lowest concentration leading to four out of four positive results when testing hCG levels of 12, 18, 24, 50 and 100 mIU/ml. While the Clinitest assay reports samples with hCG concentrations greater than 25 mIU/ml as positive, Siemens does not make claim to sensitivities for any of the various isoforms of hCG. Somewhat different results were reported with the Quidel QuickVue One-Step hCG Combo assay. Again, no lot-specific information was provided. The sensitivity was found to be 24 mIU/ml for both hCG and hCG-H, 88 mIU/ml for free hCG-b and >500 mIU/ml for the hCG-bcf. We were concerned, for many reasons, about the reported poor sensitivity of Clinitest hCG. For example, others have confirmed hCG sensitivities comparable with those of the product claim [2]. Furthermore, before the product is released for sale, every lot undergoes rigorous in-process testing for many levels of hCG, including the 25 mIU/ml level, as well as other potentially cross-reacting moieties, to ensure that various hCG concentrations are interpreted properly on the instrument. We were also concerned about Cole’s criteria for sensitivity, since we felt it was statistically challenging to distinguish, for example, three out of four positive results (when testing 24 mIU/ml) from four out of four positive results (when testing a 50 mIU/ml sample). The author did not indicate whether any of those samples that did not read positive were reported as borderline. Finally, from earlier studies, we knew that the Clinitest hCG response to various levels of hCG-H was much better than that which was reported in Cole’s paper. We recently repeated these studies by spiking male urine pools (specific gravity: 1.017) with known molar concentrations of various hCG isoform reference materials provided by the National Institute for Biological Standards and Control (NIBSC, Hertfordshire, UK). Thus, in addition to those isoforms evaluated in Cole’s paper, we also examined the fifth international standard (IS) for hCG (07/364), WHO hCG-a (99/720), WHO hCG-nicked-b (hCG-bn; 99/692) and WHO hCG-nicked (hCGn; 99/642). Known concentrations of hCG-H, spiked into the same male urine pools, were provided by Cole and assayed by us in his laboratory. At least five different concentrations of various hCG isoforms were assayed on each of five different CLINITEK Status Eric S Cowden