SEARCH

SEARCH BY CITATION

Keywords:

  • Hemostasis;
  • thrombosis;
  • special coagulation;
  • laboratory practice;
  • preanalytical variables;
  • assay interferences

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

Assays in the special coagulation laboratory are affected by numerous factors, including pre-analytical variables, anticoagulant drugs, and abnormalities of the coagulation system other than the analyte specifically being examined. By reviewing special coagulation tests as a group and in concert with clinical information, as well as understanding assay methodologies, interferences can be more easily recognized and incorrect interpretations avoided, preventing possibly unnecessary treatment of patients. Three case studies involving protein S activity, von Willebrand factor analysis, and factor V activity with Bethesda titer will highlight potential pitfalls in the interpretation of special coagulation tests.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

Special coagulation tests are affected by numerous factors, including pre-analytic variables, anticoagulant drugs, and abnormalities of the coagulation system other than the analyte specifically being examined [1]. Failure to recognize these interfering factors may lead to misdiagnosis and, possibly, to unnecessary treatment of patients. By reviewing special coagulation tests as a group and in concert with clinical information and by understanding special coagulation assay methodologies, interferences can be more easily recognized and incorrect interpretations avoided. This review will focus on three illustrative case studies, with discussion of the pitfalls demonstrated in each case as examples of the complexities inherent in special coagulation testing.

Case One

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

A 57-year-old woman was undergoing evaluation for hypercoagulability. Laboratory results are provided in Table 1. Laboratory evaluation was remarkable for a factor V Leiden mutation. Protein S activity was decreased, while both free and total protein S antigen were normal. Protein C activity was normal, as was the prothrombin time (PT). Does this patient have two hereditary risk factors for thrombophilia?

Table 1. Laboratory results for case 1 and case 3
Test nameCase 1Case 3Reference interval
  1. PT, prothrombin time; APTT, activated partial thromboplastin time; TT, thrombin time; DRVVT, dilute Russell's viper venom time; s, seconds; N/A, not applicable.

PT (s)11.535.810.0–12.8
PT 1 : 1 normal plasma mixing study (immediate) (s)N/A14.810.2–12.8
PT 1 : 1 normal plasma mixing study (60-min incubated) (s)N/A38.610.0–20.0
PT 1 : 1 normal plasma mixing study (60-min incubation control) (s)N/A22.210.0–20.0
APTT (s)26.8110.423.9–32.6
TT (s)9.413.58.0–14.0
DRVVT (s)34.0N/A28.8–42.0
Protein S activity (%)23N/A63–140
Free protein S antigen (%)100N/A62–146
Total protein S antigen (%)124N/A60–150
Protein C activity (%)138N/A55–140
Factor V Leiden mutationDetected, heterozygousDetected, heterozygousNot detected

Protein S (PS) may be evaluated by three different assays, a clot-based PS activity, free PS antigen, and total PS antigen. There is no chromogenic activity assay available. According to ISTH-WHO guidelines, the recommended screening assay is free PS antigen [2]. The free PS antigen assay immunologically measures the fraction of PS not bound to C4b-binding protein and generally correlates with and is considered a surrogate for PS activity. Total PS antigen immunologically measures both bound and unbound PS. Protein S activity assays evaluate the activated protein C (APC) cofactor activity of PS in clot-based assays that are based on the activated partial thromboplastin time (APTT), the prothrombin time (PT), or dilute Russell's viper venom time (dRVVT) assays [3-5].

Hereditary PS deficiency is described to have three different phenotypes: types I, II, and III. Type I demonstrates a decrease in all three PS assays, type II shows decreased PS activity with normal free and total antigen, and type III has decreased free PS antigen and PS activity with normal total PS antigen [4]. It is important to note that screening with free PS antigen alone may miss type II deficiencies, but these are the least common of the heritable PS deficiencies [4, 5].

Protein S activity assays have many possible causes of interference that may lead to diagnostic confusion. The interference depends on the specific type of clot-based assay performed (i.e., APTT based, PT based, or dRVVT based). These include prolonged APTT, prolonged PT, elevated factor VIII and factor VII activities, APC resistance, lupus anticoagulant, acute-phase responses (leading to increased C4b availability and therefore increased PS binding), and anticoagulant therapy. In addition, acquired decreases in PS activity occur in several clinical scenarios, such as an acute thrombotic episode, vitamin K deficiency, hepatic disease, pregnancy, or disseminated intravascular coagulation (DIC) [3, 4, 6].

Of particular interest in this case is the decreased PS activity observed in association with APC resistance due to a factor V Leiden mutation. Spuriously decreased PS activity in patients with APC resistance is well documented [2-4, 7]. Among those patients with factor V Leiden mutations, PS activity tends to be lower in patients homozygous vs. heterozygous for factor V Leiden [7]. This interference is not seen with all manufactured assay kits to the same degree, so it is essential to understand the particularities of the PS activity assay used in one's own laboratory [3]. What is the mechanism of interference? Protein S exerts its anticoagulant effect as a cofactor for APC that enhances its inactivation of factors Va and VIIIa [8]. As PS activity assays measure the APC cofactor activity of PS through its ability to prolong clotting times, causes of APC resistance, such as factor V Leiden, will lead to underestimation of PS levels [4]. However, PS deficiency has been described to occur concurrently with APC resistance due to factor V Leiden in a series of patients examined with free and total PS antigen assays, with factor V Leiden mutations identified in seven of 18 PS-deficient kindreds [9]. Therefore, not all cases of concurrent factor V Leiden mutation and decreased PS activity indicate a spurious PS value. In patients with APC resistance, with or without factor V Leiden, a free PS antigen assay is recommended.

This case demonstrates the potential pitfall of incorrectly diagnosing PS deficiency in a patient with the factor V Leiden mutation. As in this case, free and total PS antigen assays can be useful if they give normal results, as this strongly favors a spurious decrease due to interference in the PS activity assay as opposed to a true PS deficiency, given that interferences causing decreased PS activity are exceedingly more common than type II deficiency. In some cases, family history and laboratory testing of other family members may also help resolve diagnostic uncertainty.

Case Two

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

A 71-year-old man undergoing evaluation for von Willebrand disease (VWD) had the following laboratory results: factor VIII activity 118% (60–150%), von Willebrand factor (VWF) ristocetin cofactor activity (VWF:RCo) 123% (50–150%), and elevated VWF antigen (VWF:Ag) 563% (50–150%). VWF:RCo/VWF:Ag ratio was markedly decreased at 0.2, and ratios <0.5–0.7 suggest possible type 2 VWD, including types 2A, 2B, and 2M [10]. Due to the low ratio, von Willebrand factor multimer analysis was performed and demonstrated a normal pattern and distribution of multimers. An abnormal ratio with normal multimers occurs with type 2M VWD, which is an uncommon variant [10]. However, spuriously abnormal VWF:RCo/VWFAg ratios can occur if VWF:RCo is spuriously decreased or if VWF:Ag is spuriously increased [11-14].

A common cause of spuriously decreased VWF:RCo is the presence of polymorphisms in the VWF gene, seen most frequently in the African American population, that impair binding of ristocetin to VWF, thus decreasing VWF:RCo results. These polymorphisms are not thought to affect the ability of VWF to bind platelets in vivo [11, 12]. The results of the von Willebrand panel are within normal limits to elevated and are not suggestive of a diagnosis of either VWD type 1 or type 2. Furthermore, the VWF antigen is higher than would be expected with an acute-phase response. It is important to recall that these proteins are acute-phase reactants, and levels may increase to within the reference interval with stress or certain drugs [10]. In this case, however, the VWF antigen is much greater than VWF activity, suggesting that the antigen value is spurious and prompting further investigation. What could account for the marked discrepancy between VWF activity and VWF antigen values?

The latex immunoassay (LIA) is one of the most frequently used VWF:Ag assays. One of the most common and well-documented interferences in the VWF:Ag LIA is rheumatoid factor [13, 14]. In the LIA, latex microparticles are coated with monoclonal antibodies against VWF and agglutinate in the presence of VWF, leading to increased sample turbidity, which is measured photometrically. Rheumatoid factor leads to a spurious increase in VWF:Ag by LIA [13, 14]. Rheumatoid factor (IgM anti-IgG) may bind and cross-link antibodies attached to the latex microparticles, leading to an overestimation of VWF in a given sample [15]. For this reason, rheumatoid factor was evaluated on this patient sample and was >130 IU/mL (<14 IU/mL), accounting for a spurious increase in VWF:Ag by LIA. Enzyme-linked immunosorbent assay (ELISA)–based VWF:Ag assays do not generally show rheumatoid factor interference, although this methodology is less commonly performed. Failure to recognize spurious elevation of VWF:Ag due to rheumatoid factor is a possible pitfall in the diagnosis of VWD.

Case Three

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

Factor V activity with Bethesda titer was ordered for a 21-year-old woman. Laboratory evaluation is presented in Table 1. The evaluation demonstrated a prolonged PT with incomplete correction in a 1 : 1 mixing study with normal pooled plasma. A 60-min incubated mixing study demonstrated further prolongation as compared with the incubation control. The mixing study results suggested the presence of a time-dependent inhibitor. Factor V inhibitors have been reported to be time dependent [16]. The APTT was prolonged, but thrombin time (TT) was normal. Factor V activity was markedly decreased at 1% (70–150%). The Bethesda titer was initially called positive. Review of the results by the pathologist revealed increased factor V activity with increasing dilution (nonparallelism) seen in the dilutions performed for the Bethesda assay and marked prolongation of the APTT in incubated mixing studies, suggesting the presence of potassium ethylene diamine tetraacetic acid (EDTA). Although these findings are not specific for the presence of potassium EDTA, this pattern is sufficient to prompt further investigation to avoid reporting a spuriously positive Bethesda titer. Further testing of the sample revealed that the anticoagulant was, in fact, potassium EDTA and not sodium citrate.

How can laboratories that test secondary aliquots detect a sample collected in potassium EDTA and avoid this potential pitfall? First, it is useful to understand the impact that potassium EDTA has on normal plasma. Representative results of clot-based assays performed on potassium EDTA plasma from a healthy individual are presented in Table 2 (unpublished data, Colorado Coagulation). One of the most straightforward means of potassium EDTA identification is through measurement of potassium and calcium using assays available in all chemistry laboratories. In a potassium EDTA sample, potassium is expected to be markedly increased, while calcium should be essentially undetectable [17]. Potassium EDTA samples cause confusion in interpretation of factor V and VIII assays, as these factors are factitiously decreased in potassium EDTA plasmas with all other factors showing essentially normal activities. Furthermore, mixing studies using potassium EDTA plasma fail to show correction and demonstrate significant prolongation with incubation. If Bethesda assays are performed, results generally appear spuriously positive, leading to further confusion. If both factor V and factor VIII inhibitors are detected, this may be a clue to a potassium EDTA sample, as specific inhibitors to both factors would be uncommon [18].

Table 2. Representative results from potassium EDTA plasma in a healthy individual
Test namePotassium EDTA3.2% Sodium citrate
  1. PT, prothrombin time; APTT, activated partial thromboplastin time; TT, thrombin time; s, seconds; EDTA, ethylene diamine tetraacetic acid.

  2. Results vary between patients, with some showing greater decreases in factor V and VIII activity and higher factor IX activity than the values presented here.

PT (s)19.310.8–12.8
PT 1 : 1 normal plasma mixing study (immediate) (s)16.810.2–12.8
PT 1 : 1 normal plasma mixing study (60-min incubated) (s)25.410.0–20.0
PT 1 : 1 normal plasma mixing study (60-min incubated control) (s)21.710.0–20.0
APTT (s)57.323.9–32.6
TT (s)13.28.0–14.0
Factor V activity (%)4670–150
Factor VIII activity (%)1750–150
Factor IX activity (%)13060–150

This case also proved to illustrate a pitfall in assay ordering. During investigation of these unusual laboratory results, discussion with the clinical team revealed that the patient had a family history of the factor V Leiden mutation. Her father and two siblings were all heterozygous for the mutation, and her father had experienced multiple thrombotic events in the past. The patient had not yet experienced a thrombosis, but testing for factor V Leiden was desired for risk assessment, given the family history. Inadvertently, a factor V activity and inhibitor (Bethesda) titer were ordered, instead of factor V Leiden mutation analysis. Of note, potassium EDTA whole blood is the preferred specimen type for factor V Leiden mutation studies in our laboratory and may account for the choice of anticoagulant at the time of phlebotomy. A subsequent sample submitted for factor V Leiden mutation testing demonstrated a heterozygous mutation in this patient.

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References

In conclusion, special coagulation assays are subject to many sources of variation, including pre-analytical variables and effects of other, concurrent coagulation abnormalities that may cause interference. Thorough laboratory investigation, with interpretation of all screening and special coagulation assays in conjunction with clinical information, will aid in the identification of interferences and avoidance of pitfalls that may lead to incorrect diagnoses and potentially unnecessary treatments.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Case One
  5. Case Two
  6. Case Three
  7. Summary
  8. References