How useful is the monitoring of (low molecular weight) heparin therapy by anti-Xa assay? A laboratory perspective

We have conducted a series of laboratory-based surveys to assess variability in assay results utilized to monitor heparin anticoagulant therapy. These surveys involved laboratories participating in the Haematology component of the Royal College of Pathologists of Australasia Quality Assurance Program (RCPA QAP). Thirty five of 646 laboratories that were sent a preliminary questionnaire indicated that they performed anti-Xa assays and these laboratories were sent a panel of four plasma samples. These plasma samples contained respectively: (i) no added heparin, (ii) low molecular weight heparin (LMWH), enoxaparin, added to a level of approximately .5 U/mL, (iii) unfractionated heparin added to a level of approximately .5 U/mL, and (iv) LMWH added to a level of approximately 1.0 U/mL. Tests to be performed were the activated partial thromboplastin time (APTT), the thrombin time (TT), fibrinogen, and anti-Xa. As expected, returned results for APTT and TT showed some elevation in heparinized samples while fibrinogen assays were not affected. Anti-Xa assays yielded the following results (median [range]): (i) .01 [0-.11], (ii) .43 [.33-.80], (iii) .23 [.10-.49], and (iv) .90 [.60-1.30]. Thus, although median values were close to those anticipated, there was a wide variation in returned results. In a repeat exercise a few months later laboratories were also asked about their therapeutic ranges (TRs) and provided with an additional vial of LMWH-spiked (1.0 U/mL) plasma labeled as 'heparin-standard' to be used as an assay calibrant. TRs varied substantially between laboratories, from low ranges of .2-.4 to high ranges of .8-1.2. Anti-Xa assay results were similar to those of the first survey: (median [range]): (a) repeat testing: (i) .02 [0-.28], (ii) .47 [.34-.80], (iii) .25 [.14-.58], (iv) .95 [.65-1.31]; (b) repeat testing using survey provided 'heparin-standard': (i) .02 [0-.24], (ii) .55 [.4-.83], (iii) .28 [.10-.63], (iv) 1.00 [.9-1.16]. Thus using the provided 'heparin-standard' yielded lower variability in results for LMWH. In conclusion, the high variability of anti-Xa assay results coupled with the widely variable TRs suggests that therapeutic heparin monitoring is poorly standardized, and this raises some concerns over the clinical value of such monitoring.