| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Pathology, University of Virginia, Charlottesville, VA;
2 Ohio State University, Columbus, OH;
aaddress correspondence to this author at: Department of Pathology, University of Virginia, MR5 Building, Rm. 3330, 415 Lane Rd., PO Box 800904, Charlottesville, VA 22908-0904; fax 434-924-1545, e-mail mahadevan{at}virginia.edu
Although the role of factor V as a coagulation factor is more familiar, it has an equally important alternative role as a cofactor for protein C. Activated protein C (APC) is important in a naturally occurring anticoagulant pathway in which it cleaves factor V, thereby controlling the concentrations of factor V. The factor V Leiden mutation (1), which has a frequency of 
1% in Caucasian populations and accounts for most cases of (APC) resistance, makes factor V resistant to cleavage by APC. Heterozygosity for the factor V Leiden mutation confers an increased lifelong relative risk for venous thrombosis, whereas homozygosity for the factor V Leiden mutation confers an even greater increased lifelong risk. Because of its high prevalence and association with thrombophilic disorders, a variety of assays have been developed to detect the G
A mutation at nucleotide 1691, codon 506, of the factor V gene, including assays based on use of the LightCyclerTM (2)(3). In this report we present a case of an anomalous result obtained with the Roche LightCycler assay for factor V Leiden and discuss its implications.
Recently, a 52-year-old female was diagnosed with deep venous thrombosis at our institution. As part of her assessment, she underwent a routine work-up for hypercoagulation. Her partial thromboplastin time and prothrombin time were normal before she was treated with anticoagulants. Measurements of protein C, protein S, and anti-thrombin III were deferred until she was finished with coumadin. Meanwhile, results from molecular diagnostics testing included a negative result for the GA mutation at nucleotide 20210 in the prothrombin gene. However, an assay for the factor V Leiden mutation performed on the LightCycler showed an abnormal melting peak, distinct from the usual factor V Leiden mutation (Fig. 1A
).
|
We used the Roche analyte-specific reagent for factor V Leiden (cat. no. 3028526). This assay is performed by real-time PCR followed by melting curve analysis with fluorescence resonance energy transfer (FRET) probes targeted at the factor V Leiden mutation sequence. According to the Roche package insert, the acceptable intervals for the melting temperatures for the PCR products are 62.5–67.5 °C for the wild-type factor V allele and 54.5–59.5 °C for the factor V Leiden allele. As part of our quality-control (QC) procedures, for the past 2 years we have been running DNA samples from patients confirmed to be heterozygous for the factor V Leiden mutation in all of our factor V assays, tracking the melting temperatures of the wild-type and factor V Leiden PCR products. The mean (SD) melting temperatures (Tms) for our control materials were 64.59 (0.50) °C and 56.52 (0.50) °C for the wild-type and factor V Leiden alleles, respectively. Thus, in our hands, these results show that the assay has a much narrower range of variability (CV <1.0%) than that suggested by Roche. The mean (SD) difference in melting temperature (
Tm; i.e., the Tm of the wild-type allele minus the Tm of the mutant allele), an adjunct QC measure, was 8.07 (0.17) °C with a CV of 2.1%. For 27 random patients identified as heterozygous for the factor V Leiden mutation, the mean (SD) Tms were 64.50 (0.53) °C and 56.36 (0.54) °C for the 2 alleles, with a mean (SD) Tm of 8.14 (0.15) °C, in keeping with the results from our QC material.
The patient in this report was heterozygous for the wild-type allele and a second allele (Fig. 1A
). The wild-type allele had a melting temperature of 65.7 °C. This patient’s second allele showed melting temperatures of 58.7 and 59.7 °C for 2 separate DNA extracts, significantly different from that for our positive QC sample. However, if we used the package insert temperature range guidelines of 54.5–59.5 °C, the former value (58.7 °C) was within the range for the factor V allele and the latter value (59.7 °C) was just slightly out of that range. In addition, the Tms (6–7 °C) were also significantly less than those for our QC materials or our usual results from heterozygous patients with the factor V Leiden mutation. Thus, because of our quantitative QC monitoring of this assay and the fact that the patient’s results were significantly outside the established variability for our heterozygous control material, we considered that the results from the Roche LightCycler assay were consistent with an anomalous melting peak and did not correspond to a factor V Leiden mutation. This prompted us to consider another mutation lying within the region detected by the FRET probes.
At least 19 different mutations in the factor V gene have been described (4)(5). Most of these mutations have been described in factor V-deficient patients (4)(5). Of significance to this patient, a mutation causing a "false positive factor V Leiden" result has been reported previously and was obtained by a strategy combining PCR and restriction fragment length polymorphism analysis with Mnl1 and Hph1 restriction enzyme digestion (6). Sequencing of the PCR product from that patient revealed the mutation to be a CT transition at position 1690 of the factor V gene (6). Given this information, we used the software associated with the LightCycler to show that this mutation would give a Tm that corresponded to that seen in our patient’s specimen (i.e., 59 °C). Direct sequencing of the PCR product from our patient’s specimen revealed that she was indeed heterozygous for the CT mutation at position 1690 (Fig. 1B
). This mutation causes the arginine at position 506 to change to a stop codon, producing a null allele and probably causing a 50% decrease in the measured factor V activity in a heterozygote. Unfortunately, quantitative assays for factor V activity or protein were not available.
Obviously, the premature stop codon that we identified would have effects on the coagulation pathway vastly different from those of a factor V Leiden mutation, as well as significantly different clinical and therapeutic implications. The CT change at position 1690 would produce a factor V deficiency, whereas the factor V Leiden mutation (GA at position 1691) causes a thrombophilic predisposition. Because of the proximity of the 2 melting peaks seen in the Roche LightCycler factor V assay, the results have the potential to be misinterpreted as positive for factor V Leiden mutation if the CT 1690 mutation is present instead of the GA 1691 mutation. Accordingly, based on our results and DNA sequence confirmation, the patient’s thrombotic condition was considered not related to a factor V Leiden mutation.
This report highlights the importance of recognizing the difference between these 2 peaks so that a patient with a potentially hemorrhagic condition is not mistakenly diagnosed with a thrombophilic condition. Although recognition of the CT mutation at position 1690 by other methods has been described previously, to our knowledge this is the first report of this mutation detected with the LightCycler methodology. Given the increasing popularity of such assays, providers need to be cognizant of this potentially false-positive result. Furthermore, it underscores the value of quantitative QC measures for allele-specific assays that use melting-temperature analyses. We emphasize the importance of establishing and maintaining QC material and carefully examining the resulting data.
Acknowledgments
We wish to thank Dr. D. Haverstick and the staff of the molecular diagnostics laboratory.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  R. P. Agarwal, S. M. Peters, M. Shemirani, and N. von Ahsen Improved Real-Time Multiplex Polymerase Chain Reaction Detection of Methylenetetrahydrofolate Reductase (MTHFR) 677C>T and 1298A>C Polymorphisms Using Nearest Neighbor Model-Based Probe Design J. Mol. Diagn., July 1, 2007; 9(3): 345 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Lyon Discovering Rare Variants by Use of Melting Temperature Shifts Seen in Melting Curve Analysis Clin. Chem., August 1, 2005; 51(8): 1331 - 1332. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, heterozygous control for factor V Leiden; solid lines, results from 2 independent DNA extracts from the patient. In the heterozygous factor V Leiden sample, the wild-type allele has a melting temperature of 65.7 °C and the factor V Leiden allele has a melting temperature of 56.9 °C. The patient’s samples had a wild-type peak at 65.7 °C and an anomalous peak at