New developments in laboratory diagnosis and monitoring
Diagnosis of all cases of mild haemophilia A requires both one-stage and two-stage clotting or chromogenic assays
Steve Kitchen
The most commonly performed assay for factor VIII:C worldwide for many years has been the one-stage assay. The one-stage assay is based on the activated partial thromboplastin time (APTT) and depends upon the ability of a sample containing factor VIII to correct or shorten the delayed clotting of a plasma which has a complete lack of FVIII (FVIII-deficient plasma). It should be noted however that mild haemophilia A is not excluded by the finding of a normal FVIII:C level by one-stage assay, for the reasons discussed below.
Several groups have reported that a subgroup of mild haemophilia A patients have discrepancy between the activity of FVIII as determined using different types of assay [1–3]. More than 20% of mild haemophilia A patients are associated with assay discrepancy, where a twofold difference between results obtained with different assay systems is considered as discrepant [1]. In some cases the one-stage assay result may be five times higher than the two-stage clotting or chromogenic assay [1]. The most common type of assay discrepancy is to have results of one-stage assays higher than results of two-stage clotting or chromogenic assay. In more than three-fourths of such patients all assay results are reduced below the lower limit of the reference range so that a diagnosis may be reliably made irrespective of which method is employed for analysis. However, a small proportion of patients have results by the one-stage assay which is well within the normal range with reduced levels by a two-stage clotting or chromogenic assay [3,4]. These patients have bleeding histories compatible with the lower levels obtained in a two-stage clotting or chromogenic assay. In many cases the genetic defect has been identified, so there is no doubt that these subjects do indeed have haemophilia [4,5]. In our experience about 5–10% of mild haemophilia A patients have a normal one-stage assay result. This is a prevalence similar to that described by other groups. As FVIII activity is normal in the one-stage APTT-based assay, it is not surprising that the APTT is also normal in such patients. This means that patients with a clinical history compatible with haemophilia A should have a two-stage clotting or chromogenic assay even if the APTT and one-stage assay are normal.
We screened 68 patients with mild haemophilia A and found seven in whom the result by one-stage assays was within the normal range when the two-stage was reduced, and where the one-stage activity was also at least twice the level by two-stage clotting assay. Results are shown in Table 1, below, which includes results obtained with a chromogenic assay (Siemens/Dade-Behring, Marburg, Germany).
| Case | One-stage assay (IU dL−1) | Two-stage clotting assay (IU dL−1) | Chromogenic assay (IU dL−1) |
|---|---|---|---|
| A | 101 | 34 | 13 |
| B | 88 | 15 | 28 |
| C | 63 | 30 | 40 |
| D | 55 | 24 | 40 |
| E | 58 | 21 | 33 |
| F | 72 | 21 | 36 |
| G | 84 | 19 | 45 |
It is important to note that the initial investigation of these patients included a two-stage clotting FVIII assay. If the only assay performed had been a one-stage assay on these subjects the diagnosis would most likely have been missed. The clinical phenotype correlates closely with the lower result obtained by the two-stage assay.
It has been reported that 8 of 97 patients in South Australia with mild haemophilia were also found to have normal one-stage FVIII, with reduced activity in two-stage methods [6]. In these patients the level of FVIII by chromogenic assay varied according to which commercial assay was employed. The authors reported that one of three chromogenic assays would not have been suitable to diagnose these patients, and that for two other chromogenic assays, the activity of FVIII was lower when the incubation time in the assay was extended, increasing the confidence with which mild haemophilia A was diagnosed.
We have reported a few cases of mild haemophilia A which have reduced activity by one-stage but normal results by the two-stage assay. This has been confirmed by others [7,8]. In these cases the clinical phenotype once again correlates with the two-stage result in that there is no personal or family history of bleeding with no requirement for FVIII replacement therapy [8]. Studies in Sheffield have identified this reverse discrepancy (two-stage/chromogenic FVIII:C > one-stage activity) associated with Tyr346Cys in approximately 5% of patients with mild haemophilia A [9]. More than half of these cases were initially investigated following the detection of a prolonged APTT identified prior to surgery, without any evidence of bleeding history with approximately 20% of such cases being investigated for a possible bleeding history. Thus there is an absence of bleeding in many of these cases. On the other hand, others have identified the existence of similar discrepancies in patients with normal two-stage/chromogenic FVIII activity and reduced one-stage where the bleeding symptoms are consistent with mild haemophilia (J. Oldenberg, personal communication). It may be that the presence or absence of bleeding in these reverse discrepancy patients depends on the genetic defect present and how the FVIII function is affected.
Some reviews of previously diagnosed mild haemophilia have failed to identify any cases with totally normal one-stage FVIII assay with reduced two-stage clotting/chromogenic activity [10,11]. If the population of haemophiliacs selected for these studies had been initially diagnosed by one-stage assay there could be selection bias as any patients with normal one-stage FVIII may have been originally classified as normal and therefore excluded from the group under analysis.
Based on these results it is recommended that all haemophilia centres have available a chromogenic or two-stage clotting assay and that this should be performed in subjects with normal APTT and one-stage FVIII activity in the presence of a personal or family history consistent with mild haemophilia A.
Standardization of platelet function testing in bleeding disorders
C.P.M. Hayward
Platelet function testing is important for the diagnosis of many inherited and acquired bleeding disorders but it has lacked standardization [12]. Furthermore, heterogeneity in the biology and laboratory manifestations of platelet function disorders poses additional challenges to standardizing the diagnostic testing [12–15]. Recent surveys, including the largest worldwide survey of clinical laboratories by the International Society on Thrombosis and Haemostasis [16], have been helpful to identify which aspects of commonly performed platelet function tests, such as light transmission aggregation (LTA), show the greatest deviation in practice [17–19]. The lack of standardization in testing has led a number of organizations to develop guidelines and recommendations, using expert opinion and/or systematic reviews of the literature [12,20–23]. Presently, many diagnostic laboratories need to update their practices in order to meet these new recommendations [22].
A number of organizations have led efforts to improve and standardize the laboratory assessment of platelet disorders [11,17–19,22,24]. Some efforts have focused on defining common practices [16–19,24] and the heterogeneity in practice stimulated the development of published guidelines from organizations such as the International Society on Haemostasis and Thrombosis and the Clinical and Laboratory Standards Institute [22]. Presently, the most common type of diagnostic assay used to investigate a known or suspected platelet function disorder is an assessment of platelet aggregation function, often by LTA [12,16,17]. Although some laboratories perform ‘screening tests’ [such as the bleeding time and Platelet Function Analyzer-100® (Siemens/Dade-behing, Marburg, Germany) closure time] [19], neither the bleeding time nor the closure time has sufficient sensitivity to rule out common platelet function disorders [23,25]. LTA has considerable diagnostic utility when performed by rigorously standardized procedures [25,26] with validated reference intervals [27] on samples from individuals referred for bleeding problems. Laboratories need to consider the potential for false positives, as LTA abnormalities with two or more agonists are much more highly predictive of a bleeding disorder than a single agonist abnormality [25]. In Canada, the Ontario Quality Management Program – Laboratory Services recently released consensus recommendations to guide performance of platelet function tests in diagnostic laboratories to promote best and standardized practices in assay performance and also to provide guidelines on LTA interpretation (Table 2) [21] which was not covered by Clinical and Laboratory Standards Institute (CLSI) guidelines [22]. In order to translate recent platelet function testing guidelines into practice, some groups are now working to develop regional, standardized operating procedures for platelet function tests, based on consensus recommendations [21].
| LTA findings | Recommended interpretation | Follow-up investigations to consider |
|---|---|---|
| Aggregation is absent or markedly reduced with arachidonic acid, normal with thromboxane analogue, and reduced with low concentrations of collagen. There is absent secondary aggregation with adrenaline | Aspirin-like defect (drug induced or inherited). The drug history should be reviewed | Repeat testing when subject is not on aspirin or other non-steroidal anti-inflammatory drugs |
| Aggregation is present only with ristocetin | Possible Glanzmann thrombasthenia (inherited or acquired) | Glycoprotein analysis to assess the platelet fibrinogen receptor αIIbβ3 |
| Aggregation is absent with high concentrations of ristocetin and the patient has thrombocytopenia with very large platelets (can be normal if the defect is acquired) | Possible Bernard Soulier Syndrome (inherited or acquired). von Willebrand factor deficiency should be excluded | Glycoprotein analysis to assess glycoprotein IbIXV, the platelet von Willebrand factor receptor |
| Aggregation is reduced with high concentrations of ristocetin and the patient does not have thrombocytopenia | The interpretation should consider the possibility of von Willebrand disease | von Willebrand factor levels |
| Aggregation is abnormally increased with low concentrations of ristocetin | Possible type 2B or platelet-type von Willebrand disease | von Willebrand factor levels. Consider genetic testing for type 2B or platelet-type von Willebrand disease |
| Aggregation is abnormal with a number of agonists but markedly impaired with ADP, with significant de-aggregation | The possibility of a platelet ADP receptor defect (P2Y12) should be considered. A drug-induced defect should be excluded as the cause | Repeat aggregation testing |
| Other abnormalities with two or more agonists | The findings suggest that a platelet function disorder is present. The finding should be confirmed on another sample, if clinically indicated | Platelet ATP release and/or electron microscopy for dense granule deficiency |
| Abnormalities seen with only one agonist (excluding collagen or ristocetin) | Aggregation responses indicate a single agonist abnormality which is non-diagnostic and could represent a false positive | Repeat aggregation testingPlatelet ATP release and/or electron microscopy for dense granule deficiency |
Recent proficiency challenges have illustrated the value of formal, standardized, external quality assessments (EQA) for the diagnostic tests used to evaluate platelet function disorders, which supplement internal quality programme initiatives [24]. As almost all platelet function tests require rapid processing of freshly collected blood samples, all platelet EQA challenges have used some type of strategy to overcome this need, ranging from performing tests on freshly collected healthy control samples spiked with or without an inhibitory addition, or by partially preparing the material required for evaluation of a platelet function disorder [24]. For example, EQA of electron microscopy assays for diagnosing platelet dense granule deficiency have used air-dried samples of platelet rich-plasma transferred onto grids, and participants have correctly identified normal and dense granule-deficient samples [28]. For LTA EQA, the North American Specialized Coagulation Association and ECAT foundation have performed post-analytical exercises, using distributed aggregation values and tracings for cases, to assess the quality of LTA interpretation. These efforts have helped to improve and standardize the diagnostic interpretation of LTA findings. One reason that EQA have focused on LTA, and the diagnosis of dense granule deficiency is because standardized LTA and standardized assays for dense granule deficiency, when performed in accordance with guidelines, have important diagnostic utility for the assessment of common bleeding problems [25]. If a higher level of platelet function test standardization can be achieved, it would probably improve the diagnostic evaluation of platelet function disorders worldwide [12]. This would be an important achievement as platelet function disorders are now recognized as one of the most common causes of abnormal bleeding worldwide [14,25,29,30].
Potential utility of thrombin generation testing in bleeding disorders
Claude Negrier, Yesim Dargaud
The diagnosis of a particular inherited bleeding disorder due to a single molecule defect is usually done by the direct measurement of this molecule using a clotting-based or a chromogenic assay. The assay in addition allows categorizing the severity of the defect, but it has a relatively poor correlation with the clinical phenotype. Conversely, a global coagulation assay such as the thrombin generation test (TGT) seems to display a relatively good correlation with the bleeding tendency but has not been used to classify the severity of the disease so far and is not used either as a screening test in case of suspicion of a bleeding diathesis. Although a statistically significant correlation was found between FVIII or factor IX clotting activity and most of the TGT parameters, the coefficient of correlation was not optimal, suggesting that additional events were indeed evaluated with the global coagulation assay. Another example of the respective clinical value of the specific and more global assays is given by factor XI deficiency. Most of the patients with FXI deficiency are mild bleeders, but it has been recognized that patients with similar FXI activity may exhibit different bleeding phenotypes. Routine laboratory assays such as measurement of FXI clotting activity is crucial for establishing the diagnosis of the deficiency, but does not help doctors to estimate the individual bleeding risk in these patients. The TGT was used to discriminate bleeders and non-bleeders in a series of 24 patients with various levels of FXI deficiency. In patients exhibiting severe bleeding tendency, independently of their FXI level, a dramatic impairment of the TGT was observed. For example, despite low plasma FXI (1 IU dL−1), a clinically non-bleeding individual exhibited normal TG results whereas another patient with severe bleeding history and mild FXI deficiency (40 IU dL−1) had a very low TG capacity. The most useful TG parameters related to the bleeding tendency in this case were thrombin peak and velocity.
With regard to treatment and prevention of bleeding in patients with inherited bleeding disorders due to a single coagulation factor defect, the main therapy principle consists of substituting the missing molecule (FVIII, FIX, FXI, FVII) in order to increase the plasma level of the clotting factor. Conversely to this substitution treatment, bypassing agents, represented by activated prothrombin complex concentrates (aPCC) and recombinant factor VIIa, are capable of triggering coagulation through different mechanisms but do not represent a substitution treatment per se. They are currently used for treatment and prevention of the bleeding complications in patients with haemophilia who developed inhibitory antibodies against FVIII. These agents trigger haemostasis at the cellular surfaces, particularly on the outer leaflet of the platelet membrane, by promoting Xase complex formation and thrombin generation, ultimately leading to fibrin deposition at the site of vascular damage. The ex vivo monitoring that would reflect achievement of haemostasis in vivo still needs further studies, though several attempts have already been initiated. In this respect, the thrombin generation assay might be used to predict the differential response to recombinant FVIIa and FEIBA® (Baxter Healthcare, Zurich, Switzerland) tested prior to in vivo administration, and might provide further insight into the optimal dose of therapy pre- and postoperatively. If one considers that there probably is a level of thrombin generated that predicts clinical efficacy, this assay could be used for monitoring of FVIII/FIX bypassing agents and for optimizing the infusion schedule by tailoring doses and frequency of injections individually for each single patient. However, correlation of the TGA parameters with in vivo clinical response needs to be further established if we believe that this assay may represent a surrogate marker for monitoring bypassing therapies in life or limb-threatening as well as in surgical situations. Finally, one should emphasize the critical importance of sampling conditions and manipulation of plasma samples as well as the use of a standardized protocol to obtain reproducible and meaningful results.
