Debra Goode, Medfusion, sent this message: I do not see methods or data for performing functional sensitivity tests for coagulation assays, particularly clotting factor assays. Is this not done conventionally and if not, why?
George confirmed that the inquiry refers to analytical sensitivity of clot-based coagulation factor assays.
Debra, thank you, and yes, for each clot-based factor assay we establish analytical sensitivity. Using the factor VIII assay as an example, and paraphrasing the Mayo Medical Laboratories protocol summary, the assay is performed using the partial thromboplastin time (PTT, APTT) method and factor deficient substrate, which is normal plasma depleted of factor VIII by immunoadsorption. Patient plasma is combined with the factor VIII deficient substrate to ensure normal activity levels of all but factor VIII. PTT reagent is added, and after a specified incubation, calcium is added to trigger coagulation. The time to clot formation is recorded and compared to a calibrator curve to derive factor activity. The patient specimen is tested at multiple dilutions and results from these dilutions are compared mathematically to validate the result and rule out the presence of an inhibitor.
The calibration curve is prepared from various dilutions of a calibrator plasma of known factor VIII activity, and the greatest dilution, yielding the lowest factor VIII activity, is typically 1%, establishing thereby the limit of detection (LOD) at 1%. If a patient has a factor VIII activity level below 1% we report the result as less than 1%, as we don’t extrapolate beyond the assay’s ability to distinguish smaller activity levels. This is clinically valid, as the typical factor VIII reference interval is 50–150%. LOD, in conjunction with imprecision data, is typically used by instrument manufacturers to determine the limit of quantification (LOQ). This would not necessarily be communicated with the end users but they will see it indirectly as a limit in the in factor assay report which is usully set above the LOQ. For instance, a FVIII report of less than 1% means activity level was below LOQ of the assay system.
We also establish analytical sensitivity, which is the magnitude of change in response (such as clotting time in seconds), for 1 unit difference in measurement (factor level). Ali Sadeghi-Khomami, PhD, lead scientist for Precision BioLogic Inc. adds: Yes, analytical sensitivy is done but is usually called “factor sensitivity.” The main reason that this kind of data cannot easily be communicated is due to its dependence on the entire system including the analyzer, protocol, diluent, reagents and the range of the factor level of interest. in many cases we employ both a low-range and a normal or high-range calibration curve.
In other words, functional sensitivity in factor assays is the same as the “slope of calibration curve,” provided the protocol uses linear regression. There are instruments and protocols that use polynomial orders for factor assay; that is why communicating data around factor sensitivity (slope of calibration curve) is not always doable. Typically, people who run factor sensitivity tests work on an identical system with exception of one variable to avoid this problem. Alternatively if data comparison is required between various platforms, they should test identical samples containing various factor levels and report clotting time differences as an indicator for factor sensitivity of their assay system. They need to take care of all manageable variables in their study before jumping to any meaningful conclusion.
Dave McGlasson describes a second way in which we establish PTT reagent sensitivity. Most PTT reagent manufacturers design their reagent to yield a result exceeding the reference interval limit when the patient’s factor level is at 30–40%, reflecting clinical needs. When the manufacturer delivers a new lot of reagent, the laboratory scientist confirms PTT reagent sensitivity by testing against a series of plasma dilutions. We do this typically for factors VIII, IX, and XI, the three factors often implication in coagulopathies, and record this value for as long as the reagent lot is being employed, typically, one year.