Comparative Evaluation of Three JAK2 V617F Mutation Detection

Hematopathology / JAK2V617F DETECTION BY AS-PCR AND RFLP

Comparative Evaluation of Three JAK2 V617F Mutation Detection Methods Christine Frantz, MS,1 Donna M. Sekora, MT,1 Donald C. Henley, MS,2 Chih-Kang Huang, MS,3 Qiulu Pan, MD, PhD,3 Neil B. Quigley, PhD,2 Eric Gorman, MD,3 Roger A. Hubbard, PhD,2 and Imran Mirza, MD1 Key Words: JAK2 mutation; Allele-specific PCR; AS-PCR; Restriction fragment length polymorphism; RFLP; Capillary electrophoresis; Polyacrylamide gel electrophoresis DOI: 10.1309/LW7Q3739RBRMBXXP

Abstract The correlation of JAK2V617F with a proportion of chronic myeloproliferative disorders has generated numerous studies focused on the development of molecular-based assays for JAK2V617F detection. The current parallel study comparatively evaluated 3 JAK2V617F molecular detection methods. Genomic DNA from blood or bone marrow was assayed by 3 laboratories using allele-specific polymerase chain reaction (AS-PCR) or kit-based restriction fragment length polymorphism methods, which used polyacrylamide gel or capillary electrophoresis analysis. In addition, samples were sequenced in 2 of the laboratories. Results found 100% concordance among the 3 methods, with analytic sensitivities of 5% for both kit methods and 0.01% for AS-PCR. The kitbased assays detect JAK2V617F with equal sensitivity regardless of analysis method, and, despite greater sensitivity of AS-PCR, all 3 methods yielded 100% concordant results for this 36-sample set. Consistent with other reports, direct sequencing was insufficiently sensitive to serve as an initial diagnostic tool for JAK2V617F detection.

Chronic myeloproliferative disorders (CMPDs) are clonal stem cell diseases characterized by unrestricted proliferation of granulocytes, RBCs, or platelets. The 3 primary Philadelphia chromosome (Ph)– CMPDs, polycythemia vera (PV), essential thrombocythemia (ET), and chronic idiopathic myelofibrosis (CIMF), demonstrate distinct clinical, hematologic, biologic, and molecular features.1-3 Until recently, diagnosis of CMPDs was facilitated by assays including serum erythropoietin levels, endogenous erythroid colony formation, polycythemia rubra vera-1 transcript levels, and leukocyte alkaline phosphatase score.1 In 2004 and 2005, several studies first correlated a proportion of the Ph– CMPDs with the presence of a G-to-T transversion mutation at nucleotide 1849 of the JAK2 gene.4-8 This mutation, designated JAK2V617F, resulted in a valine-to-phenylalanine substitution at codon 617 and conferred constitutive tyrosine kinase activity. Although JAK2V617F has been detected in a proportion of Ph– CMPDs, it has not been detected in Ph+ chronic myelogenous leukemia, and these are currently considered exclusive mutations.9 The identification of the JAK2V617F mutation has generated a growing body of research focused on the characterization of distinguishing clinical features of JAK2V617F-associated CMPDs and on the development of various molecular assays for detection of the mutation.1,10,11 The molecular assays described in the literature include a variety of techniques such as allele-specific polymerase chain reaction (AS-PCR, or amplification refractory mutation system PCR), direct sequencing and pyrosequencing, singlenucleotide polymorphism array, quantitative PCR, denaturing high-performance liquid chromatography, and restriction fragment length polymorphism (RFLP) analysis.12-19 These assays are performed in combination with detection methods including Am J Clin Pathol 2007;128:865-874

© American Society for Clinical Pathology 865

DOI: 10.1309/LW7Q3739RBRMBXXP

865 865

Frantz et al / JAK2V617F DETECTION BY AS-PCR AND RFLP

polyacrylamide gel electrophoresis (PAGE), capillary electrophoresis (CE), and probe dissociation (melting curve) analysis using real-time instrumentation. The analytic sensitivities of the different methods vary, with AS-PCR methods reporting the highest sensitivity and direct sequencing reporting the lowest.12-19 The variety of available techniques provides flexibility of method design, which prompts a need for comparative analysis of these methods.10 The present interlaboratory study comparatively evaluated the detection of JAK2V617F in 36 clinical samples using 3 detection methods: 2 kit-based RFLP assays using PAGE or CE detection and an AS-PCR using CE detection. In addition, samples were sequenced in parallel by 2 of the 3 participating laboratories. The aim was to compare the performance of the convenient kit-based RFLP assays with the more sensitive inhouse AS-PCR for a clinical sample set.

Materials and Methods DNA Isolation and Quantification Genomic DNA was isolated from 0.3 mL of fresh EDTA peripheral blood or bone marrow or from archived WBC pellets using the KingFisher mL automated extractor (Thermo Electron, Vantaa, Finland) and BioSprint DNA Blood Kit (Qiagen, Valencia, CA). Archived WBC pellets were prepared for automated extraction by initial resuspension in 0.3 mL of buffer AE provided with the BioSprint kit. DNA was eluted in 100 µL of TE buffer (10 mmol/L tris(hydroxymethyl)aminomethane, 0.1 mmol/L EDTA, pH 8.0) and quantified on the Fluoroskan Ascent FL microplate instrument (Thermo Electron) using PicoGreen dye (Molecular Probes, Eugene, OR).

JAK2V617F Assay and Analysis All samples were coded and assayed blindly for the JAK2V617F mutation. The sequences of all primers used in this study are provided in ❚Table 1❚. Allele-Specific PCR AS-PCR test primers were designed at the Molecular Pathology Laboratory Network (MPLN), Maryville, TN, and synthesized commercially (Sigma-Proligo and SigmaGenosys, St Louis, MO). The AS-PCR forward primer (JAK2 LNA T) carried a single 3'-end allele-specific locked nucleic acid (LNA) residue, and the reverse primer (JAK2 Rev3FAM) was 5'-labeled with 6-FAM. Separate control reactions used the same reverse primer but a different forward primer (JAK2 IntCont), which binds upstream of the 1869 G>T mutation to generate a 364-base-pair (bp) amplicon. The 25-µL JAK2 AS-PCR reaction volume incorporated 50 to 100 ng of template DNA, 1 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA), 1× ABI PCR buffer (containing 1.5 mmol/L of magnesium chloride [MgCl 2 ]) (Applied Biosystems), 2.5 mmol/L of MgCl2, and 0.2 µmol/L of each primer. The internal control reactions were identical except for the forward primer used and the omission of the additional 2.5 mmol/L of MgCl2. The PCR parameters were as follows: hold for 10 minutes at 95°C; 36 cycles of 94°C for 15 seconds, 65°C for 30 seconds, and 72°C for 30 seconds; hold for 5 minutes at 72°C; and infinite hold at 4°C. Analysis was performed by CE on the 3100 Genetic Analyzer from Applied Biosystems using ABI ROX-500 size standards. Amplicons were diluted to 1:20 in a mixture of formamide and ROX size standard for CE analysis.

❚Table 1❚ PCR and Sequencing Primers Used in the Study Primer Pair AS-PCR mutation screening JAK2 LNA T JAK2 Rev3FAM AS-PCR control reaction JAK2 IntCont JAK2 Rev3FAM MPLN PCR primers for sequencing JAK2 IntCont JAK2 Rev3 (unlabeled) MPLN sequencing primers JAK2 Fwd3 JAK2 Rev3 (unlabeled) MMC direct sequencing primers JAK2SeqF JAK2SeqR

Primer Sequence (5' to 3') CATTTGGTTTTAAATTATGGAGTATGT+T* 6-FAM-CTGAATAGTCCTACAGTGTTTTCAGTTTCA ATCTATAGTCATGCTGAAAGTAGGAGAAAG Same as for AS-PCR mutation screening Same as for AS-PCR control reaction CTGAATAGTCCTACAGTGTTTTCAGTTTCA CCTTAGTCTTTCTTTGAAGCAGC Same as for MPLN PCR primers for sequencing GGCAGTTGCAGGTCCATATAAAG TCCTGTTAAATTATAGTTTACACTGAC

AS, allele-specific; LNA, locked nucleic acid; MMC, Montefiore Medical Center, Bronx, NY; MPLN, Molecular Pathology Laboratory Network, Maryville, TN; PCR, polymerase chain reaction. * The + nomenclature in the JAK2 LNA T primer sequence indicates that the 3'-terminal T (bold) is an LNA residue. The forward primer is listed first in each pair.

866 866

Am J Clin Pathol 2007;128:865-874 DOI: 10.1309/LW7Q3739RBRMBXXP

© American Society for Clinical Pathology

Hematopathology / ORIGINAL ARTICLE

RFLP-PAGE and RFLP-CE The RFLP assay kits for polyacrylamide gel (RFLPPAGE) or CE (RFLP-CE) detection were purchased from InvivoScribe Technologies (IVS; San Diego, CA) and use identical chemistry except that the proprietary PCR primers are unlabeled (for RFLP-PAGE) or labeled with 6-FAM (forward primer) or HEX (reverse primer) (for RFLP-CE) ❚Figure 1❚. The PCR setup and cycling parameters were performed as described in the respective kit handbooks. The recommended DNA input for both assays was 0.5 to 2 µg in a total volume of 5 µL (RFLP-PAGE) or 2.5 µL (RFLP-CE). Control reactions for sample fidelity were performed using a separate master mix supplied with the RFLP kits. Following amplification, JAK2 amplicons were digested for 16 hours at 37°C with BsaXI endonuclease (New England BioLabs, Ipswich, MA) essentially as recommended in the respective kit handbooks except that the digestion volume for the RFLP-CE assay was 20 µL and incorporated 5 µL of JAK2 amplicon, 1 µL of BsaXI endonuclease (2U/µL), 2 µL of NEBuffer 4 (provided with BsaXI), and 12 µL of molecular grade water. RFLP-CE analysis of the digested JAK2 amplicons was performed on the ABI 3130XL Genetic Analyzer according to the RFLP kit handbook. For RFLP-PAGE analysis, a 20-µL mixture consisting of 4 µL of 5× nucleic acid sample loading buffer (Bio-Rad, Hercules, CA) and 16 µL of BsaX1-digested JAK2 amplicon was separated by 6% Tris Borate EDTA (TBE) PAGE and visualized by staining with SYBR Green (Molecular Probes) at 1:10,000 in TBE buffer, pH 8.0, for 15 minutes, followed by imaging using the AlphaImager 2200 (PerkinElmer, Wellesley, MA). Zygosity Determination The AS-PCR was not designed to distinguish zygosity and generated a 201-bp mutation-specific amplicon only when the V617F allele was present; JAK2V617F– samples produced only a 364-bp internal control amplicon. Both RFLP kits distinguish normal and V617F JAK2 alleles according to distinct fragment patterns generated on amplicon digestion by BsaXI (Figure 1). Amplification of JAK2 yields a 267-bp amplicon that includes the G>T 1869 (V617F) mutation site. The mutation site itself is found within the BsaXI recognition sequence. The normal JAK2 amplicon contains an intact BsaXI recognition sequence, whereas the V617F mutation changes this sequence and prevents digestion. Zygosity determination by RFLP-PAGE was based on sample interpretation guidelines in the original kit handbook. Digestion of the normal JAK2 amplicon yields 3 detectable fragments of 170, 140, and 97 bp; the 30-bp fragment is undetectable (Figure 1). The 140- and 30-bp fragments result from complete digestion of the 170-bp fragment. Therefore, the presence of the 170-bp fragment is indicative of incomplete

BsaXI recognition sequence Normal JAK2 DNA sequence

5'…(N)7GGAG(N)5GT(N)12…3' 5'…AAATTATGGAGTATGTGTCTGTGGAGACGA…3'

JAK2 amplicon

267 bp

Normal 170 bp 97 bp 140 bp

BsaXI V617F 267 bp

30 bp

❚Figure 1❚❚ Schematic of BsaXI restriction digestion patterns for normal and V617F JAK2 alleles. Amplification generates a 267-base-pair (bp) amplicon encompassing the site of the JAK2 G>T mutation at nucleotide 1849 (large bold G) and the BsaXI recognition sequence (core sequence is underlined). For polyacrylamide gel electrophoresis (PAGE) detection, both primers are unlabeled. For capillary electrophoresis (CE) detection, the forward primer is labeled with 6-FAM dye and the reverse with HEX dye, generating a primary amplicon labeled with both dyes. Digestion of the normal sequence yields 4 fragments of 170, 140, 97, and 30 bp, which retain 6FAM (170 and 140 bp) or HEX (97 bp); the 30-bp fragment excised by BsaXI is unlabeled and undetectable by PAGE and CE. The V617F mutation abolishes the BsaXI site and prevents digestion; the full-length 267-bp amplicon remains intact and is labeled with 6-FAM and HEX, generating 2 alternately labeled 267-bp peaks in the restriction fragment length polymorphism–CE method.

digestion. The 267-bp amplicon remains undigested in homozygous JAK2V617F samples, whereas heterozygous JAK2V617F samples (those comprising mixtures of normal and heterozygous or homozygous mutant cells) generate a composite digestion pattern consisting of the 267-, 170-, 140-, and 97-bp fragments, representing the normal and V617F alleles. Zygosity determination by RFLP-CE was similarly based on restriction digestion patterns but additionally included an allelic ratio calculation based on the peak height ratio of undigested and digested JAK2 fragments (undigested peak height/undigested + digested peak height). Samples generating ratios greater than 50% were considered homozygouspositive and those generating ratios equal to or less than 50% were considered heterozygous-positive for JAK2V617F. Direct Sequencing The Montefiore Medical Center (MMC) laboratory (Bronx, NY) performed direct sequencing on all 36 DNA samples using the ABI 3130XL Genetic Analyzer according to the method described by Pan et al20 and using the MMC direct sequencing primers listed in Table 1. Am J Clin Pathol 2007;128:865-874



867 867


The MPLN laboratory sequenced only the 15 JAK2V617F+ DNA samples using the primers listed in Table 1. Samples were amplified before sequencing using the MPLN PCR primers for sequencing. Amplicons were purified using the Qiagen Minelute PCR Purification Kit and sequenced using MPLN sequencing primers. Sequencing was performed using the Big Dye Terminator DNA Sequencing Kit v3.1 (ABI) and the ABI 3100 Genetic Analyzer. Analytic Sensitivity Determination For sensitivity determinations, the RFLP-PAGE and ASPCR methods used the IVS JAK2V617F Sensitivity Panel, which consists of DNA from the homozygous JAK2V617F cell line HEL diluted to 1%, 5%, 10%, 20%, or 30% (vol/vol) into normal JAK2 DNA (from human tonsil), which served as diluent and JAK2V617F– standard/control. The RFLP-CE method used similar sensitivity controls, which were prepared in house by diluting the same JAK2V617F+ DNA into the same normal JAK2 DNA (both supplied with the IVS RFLP-CE kit). The analytic sensitivity of each method was based on the lowest percentage sensitivity standard that generated an unequivocal and consistently positive JAK2V617F result. For the AS-PCR, this required serial 10-fold dilution of the 1% IVS standard into normal JAK2 DNA from the sensitivity panel to generate the 0.01% standard.

Results JAK2V617F Assay and Analysis The PCR for each JAK2V617F assay required variable DNA input, with both RFLP assays requiring a minimum of 500 ng and the AS-PCR requiring 50 to 100 ng. Because the automated DNA extraction method sometimes produced extracts of relatively low concentration, all samples were assayed undiluted by both RFLP methods, and, consequently, a proportion was assayed using less-than-recommended DNA input ❚Table 2❚; the lowest input for a JAK2V617F– sample was 33 ng, and the lowest input for a JAK2V617F+ sample was 55 ng. The AS-PCR method required dilution of only samples with concentrations greater than 100 µg/mL to maintain consistent input. Despite suboptimal DNA input for the RFLP assays, the results indicated 100% concordance for all 36 samples among the 3 detection methods; 21 samples were JAK2V617F–/undetectable and 15 samples were JAK2V617F+ (5 homozygous- and 10 heterozygous-positive by both RFLP methods). Successful control reactions were generated by all samples in all 3 methods (data not shown). None of the JAK2V617F+ samples were also positive for the presence of the BCR/ABL1 transcript (data not shown). 868 868


The 21 JAK2V617F–/undetectable samples generated an RFLP-PAGE digestion pattern consisting of 3 bands of nearly equal intensity (170, 140, and 97 bp); a small but detectable amount of the 267-bp band often remained undigested ❚Figure 2❚. Parallel analysis of these samples by RFLP-CE revealed the same 170-, 140-, and 97-bp peaks and residual 267-bp amplicon. None of these samples produced detectable JAK2V617F amplicon by AS-PCR. All 15 JAK2V617F+ samples generated a single 201-bp amplicon by AS-PCR (Figure 2). Of these samples, both RFLP assays distinguished the same 5 homozygous and 10 heterozygous samples based on the BsaXI digestion patterns and allelic ratios calculated for all positive samples (Table 2 and Figure 2). Heterozygous samples consistently produced 267-, 170-, 140-, and 97-bp peaks by CE; the same 4 bands of nearly equal intensity were also evident by PAGE. Calculation of allelic ratios using CE data generated values less than 50% for these samples (Table 2). Of the 5 homozygous-positive samples, only 1 (sample 18) produced the expected single 267-bp peak/band. The other 4 samples (samples 17, 21, and paired sample 29) generated an alternative digestion pattern by PAGE that included prominent bands of 267, 170, and 97 bp but a very faint 140-bp band. The corresponding RFLP-CE detection of these samples similarly generated the 267-, 170-, and 97-bp peaks, but the 140-bp peak was undetectable. Allelic ratio calculation generated values of more than 50% for all 5 samples, with 3 (samples 17, 18, and 21) generating ratios of or near 100%, resulting in the classification of these as homozygous-positive. Although samples 17 and 21 generated very small digestion fragments detectable by CE, the height of these peaks was too low to quantify; therefore, the corresponding allelic ratio for these samples is given as “~100%” in Table 2. Four sets of paired blood and bone marrow samples were included in the sample set (Table 2). Despite relative differences in the DNA concentration of each sample within the respective pairs, all 4 sets generated corresponding results for both sample types by all 3 detection methods, with 2 sets generating positive results and 2 sets generating negative/undetectable results. The allelic ratios within each paired JAK2V617F+ sample were comparable for peripheral blood and bone marrow. The clinical features of all cases (samples) included in the study are detailed in ❚Table 3❚. Of the 15 JAK2V617F+ cases, 13 were previously diagnosed as or showed features of CMPD, based on bone marrow biopsy, cell counts, endogenous erythroid colony formation, and clinical manifestations. The 2 remaining JAK2V617F+ cases did not have bone marrow biopsy samples available for diagnosis (cases 23 and 25). The absence of JAK2V617F was also well correlated with negative CMPD diagnosis despite suggestive clinical features. All but 2 of these cases, including 4 lacking bone marrow biopsy, were © American Society for Clinical Pathology


❚Table 2❚ Summary of Parallel Results JAK2V617F Result

Total DNA Input (ng)

Sequencing Results

Sample No.

Sample Type

PAGE

CE

PAGE

CE

AS-PCR

Allelic Ratio (%)

MMC

MPLN

1 2 3 4 5 6 7* 8 9 10 11 12 13 14 15 16* 17 18 19* 20 21 22 23 24 25 26 27 28 29† 29† 30† 30† 31† 31† 32*† 32*†

BM PB BM PB PB PB PB PB PB PB BM PB PB PB PB PB PB PB PB PB PB BM PB WBC/BM WBC/PB WBC/BM WBC/PB WBC/PB PB BM PB BM PB BM PB BM

755 340 720 75 395 240 110 65 100 190 520 345 470 255 255 325 315 385 290 285 525 355 220 565 315 325 195 245 690 250 140 300 375 310 375 270

378 170 360 38 198 120 55 33 50 95 260 173 235 128 128 163 158 193 145 143 263 178 110 283 158 163 98 123 345 125 70 150 188 155 188 135

N N N N N N h N N N N N N h N h H H h N H N h h h h N N H H N N N N h h

N N N N N N h N N N N N N h N h H H h N H N h h h h N N H H N N N N h h

N N N N N N P N N N N N N P N P P P P N P N P P P P N N P P N N N N P P

— — — — — — 10.45 — — — — — — 39.80 — 14.60 ~100 100 14.80 — ~100 — 23.20 30.00 43.15 42.85 — — 70.20 68.45 — — — — 13.55 14.40

– – – – – – – – – – – – – + – + + + + – + – + + + + – – + + – – – – + +

ND ND ND ND ND ND – ND ND ND ND ND ND + ND – + + – ND + ND + + + + ND ND + + ND ND ND ND – –

AS-PCR, allele-specific polymerase chain reaction; BM, bone marrow; CE, capillary electrophoresis; h, heterozygous positive; H, homozygous positive; MMC, Montefiore Medical Center, Bronx, NY; MPLN, Molecular Pathology Laboratory Network, Maryville, TN; N, negative or undetectable; ND, not done; P, positive; PAGE, polyacrylamide gel electrophoresis; PB, peripheral blood; WBC/BM, archived frozen WBC pellet from BM; WBC/PB, archived frozen WBC pellet from PB; –, JAK2V617F mutation not detected by sequencing; +, JAK2V617F mutation detected by sequencing. * JAK2V617F+ samples not detected by sequencing. † Paired sets of blood and bone marrow samples.

diagnosed as non-CMPD, leukemia (including Ph+ leukemia), or reactive thrombocytosis. However, 2 JAK2V617F– cases were diagnosed as CMPD (cases 8 and 17, CIMF and ET, respectively). JAK2V617F Analytic Sensitivity Determination Each of the participating laboratories independently determined the sensitivity of its JAK2V617F detection method ❚Figure 3❚. Both RFLP methods had an analytic sensitivity of 5% because this was the lowest standard that produced the most consistent and unambiguous positive result. The 1% sensitivity standard was less consistent, sometimes generating a digestion pattern more similar to the JAK2V617F– control sample, with traces of the 267-bp fragment, but at other times generating a pattern more similar to the heterozygous JAK2V617F

pattern, with a more prominent 267-bp fragment. The ASPCR assay demonstrated an analytic sensitivity of 0.01%, based on 10-fold serial dilution of the 1% sensitivity standard (not shown), which represents a 500-fold increase in sensitivity over the RFLP methods. The 0.01% sensitivity standard was not evaluated by either RFLP method. JAK2V617F Direct Sequencing Samples were sequenced in parallel by the MMC and MPLN laboratories using different sequencing primers. Of the 15 JAK2V617F+ samples, the MMC laboratory was unable to detect the mutation in 1 heterozygous-positive sample and the MPLN laboratory was unable to detect the mutation in that same sample and in 4 additional heterozygous-positive samples (Table 2; sequencing data not shown). All 5 samples that Am J Clin Pathol 2007;128:865-874



869 869


Negative/Undetectable

Heterozygous Positive

Homozygous Positive a

RFLP-PAGE

267 bp

267 bp

170 bp 140 bp

170 bp

97 bp

97 bp

b

267 bp 170 bp 140 bp

140 bp

97 bp 267 bp Weak only 140 bp

50 1,200

100

150

200

250

300

350

97 bp

RFLP-CE

800 140 bp 400 170 bp

267 bp

0

50

100

1,000 800 600 400 200 0

150

200

250

300

350

97 bp 140 bp 170 bp

267 bp

50 2,000 1,600 a 1,200 800 400 0 50 700 560 420 280 140 0

AS-PCR

150

200

250

300

350

250

300

350

267 bp

100

150

b

200 267 bp

97 bp

170 bp

60 90 120 150 180 210 240 270 300 330 360 390 420 450

60 90 120 150 180 210 240 270 300 330 360 390 420 450 8,000 6,000 4,000 2,000 0

100

8,000 6,000 4,000 2,000 0

a 201 bp

201 bp

7,200 3,600 0

7,200

b 201 bp

3,600 0

❚Figure 2❚❚ Polyacrylamide gel electrophoresis (PAGE) and capillary electrophoresis (CE) profiles of JAK2V617F+ and JAK2V617F– samples. Representative data for each method are shown. Representative negative or undetectable data are derived from sample 15, heterozygous positive data from sample 14, and homozygous positive data from sample 18 (a; 267-base-pair [bp] fragment only) or from sample 17 (b; alternate digestion pattern with weak 140-bp fragment). The red peaks in the allele-specific polymerase chain reaction (AS-PCR) electropherograms represent ROX-500 size standard peaks. Note the different scales for the restriction fragment length polymorphism (RFLP)-CE and AS-PCR electropherogram data.

were undetectable by sequencing had allelic ratios of less than 20%; the single sample that was undetectable by both sequencing methods (sample 7) had the lowest allelic ratio of any positive sample at 10.45%, as well as the lowest DNA concentration of any positive sample. The mutation was detected in all 5 homozygous-positive samples in both sequencing laboratories. Although the sequencing data revealed varying relative amounts of normal and V617F JAK2 alleles in all but 1 of 15 positive samples (sample 18, no normal allele detectable), 3 of the homozygous-positive samples generated allelic ratios of 100% (samples 17, 18, and 21), indicating the normal allele to be minimally present in these samples.

Discussion Although the RFLP assay kits offer the simplicity and convenience of master mix solutions and supplied control samples, this method presents unique technical considerations 870 870


not common to AS-PCR.18 The RFLP method requires significant DNA input, which may not be achievable, and additional processing time for the restriction digestion. In addition, incomplete digestion of amplicons could potentially generate false-positive results, and weakly positive samples containing less than approximately 5% JAK2V617F+ DNA may be difficult to interpret unequivocally by PAGE or CE. Despite these potential issues, the current parallel evaluation of the RFLP and AS-PCR methods revealed that these represent comparable methods of JAK2V617F mutation detection for this clinical sample set because all tested samples were similarly characterized by the 3 detection methods. With an analytic sensitivity of 0.01%, the AS-PCR method described herein represents the most sensitive method evaluated, and the 5% analytic sensitivity reported for the RFLP methods is similar to reported sensitivities for pyrosequencing, denaturing high-performance liquid chromatography, and melting curve analysis, suggesting that these methods offer no significant diagnostic advantage compared with RFLP.10,12-16 Although the analytic sensitivities of the methods compared in © American Society for Clinical Pathology


❚Table 3❚ Clinical Features of Patients Whose Samples Were Included in the Study Case No./ Hemoglobin, Sex/Age, y g/L*

Platelets, × 109/L*

WBCs (Neutrophils), × 109/L*

Bone Marrow Diagnosis

Cytogenetics

Other Findings

1/M/44

154

473

4.4 (2.6)

No CMPD

46,XY

2/F/56 3/M/51 4/F/86 5/M/51 6/M/9

121 122 130 149 126

1,012 646 1,159 556 1,088

187.6 (90.1) 10.4 (8.2) 5.9 (3.7) 17.7 (10) 7.7 (4.1)

No results ND ND 46,XY 46,XY

7/M/57† 8/F/62

160 103

808 85

9.5 (6.3) 4.0 (1.7)

CML No CMPD ND No CMPD Reactive thrombocytosis CMPD CIMF

Previous diagnosis of ET; taking hydroxyurea for 2 y BCR/ABL1+ (b3a2) Stroke

9/F/39 10/F/84 11/F/53 12/F/58 13/F/46 14/F/84† 15/F/88 16/F/55† 17/M/62 18/M/61†

130 117 159 139 74 96 92 142 127 132

623 255 175 460 14 297 224 1,116 967 611

8.0 (6.5) 24.8 (21.1) 6.2 (4.9) 28 (19) 209 (1.7) 14.1 (7.8) 62.7 (39.5) 13.5 (9.2) 14.3 (8.9) 30.6 (26.1)

ND ND No CMPD Nondiagnostic Pre-B ALL CIMF AML ET ET PV

19/M/37† 20/F/62 21/M/79† 22/F/29

165 138 144 137

520 643 485 488

9.8 (5.8) 8.5 (3.8) 48.8 (44.8) 7.3 (4.1)

23/F/67† 24/F/33†

146 161

998 712

9.3 (7.0) 8.9 (6.4)

ET Features of CMPD CIMF Reactive thrombocytosis ND PV

25/F/77†

164

469

13.9 (7.4)

ND

ND

26/F/15†

141

213

32.3 (25.2)

CMPD

46,XX

27/M/52

192

178

6.9 (4.7)

No CMPD

46,XY

28/M/83

169

199

8.4 (5.4)

No CMPD

46,XY

29/F/51† 30/F/58 31/F/37

92 88 143

NA 94 200

26.3 (15.1) 39.4 (27.5) 10.1 (7.1)

ET 46,XX CMML 46,XX No definitive CMPD 46,XX

32/F/44†

146

763

9.5 (6.7)

CMPD

No HSM on US

ND 47,XX,der(7)t(1;7) (q21;q22),+8 ND ND 46,XX ND 46,XX No results 46,XX 46,XX 46,XY 46,XYdel(20) (q11.2q13.1)/46, XY,der(15)t(1;15) (q10;q10) 46,XY 46,XX 45,X,-Y 46,XX

No HSM on US Splenomegaly

ND 46,XX

Splenomegaly Increased RBC mass; decreased serum EPO; EECF Increased serum EPO level; EECF; deceased Budd-Chiari syndrome; increased serum EPO level; EECF No HSM on US; increased RBC mass; increased serum EPO level Increased RBC mass; normal serum EPO level; no EECF

No HSM on US Deceased No HSM on US BCR/ABL1+ (e1a2) Deceased Prior diagnosis of CMML No HSM on US Splenomegaly

Mild splenomegaly on US No HSM on US Splenomegaly No HSM on US

Portal vein thrombosis; increased serum EPO level; no EECF 46,XX,t(1;2)(p34;q37)c

ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CIMF, chronic idiopathic myelofibrosis; CML, chronic myelogenous leukemia; CMML, chronic myelomonocytic leukemia; CMPD, chronic myeloproliferative disorder; EECF, endogenous erythroid colony formation; EPO, erythropoietin; ET, essential thrombocythemia; HSM, hepatosplenomegaly; NA, not available due to platelet clumping; ND, not done; PV, polycythemia vera; US, ultrasound. * Bold values indicate an increased level relative to the reference range. Values are given in Système International units; conversions to conventional units are as follows: hemoglobin (g/dL), divide by 10.0; WBC and neutrophil counts (/µL), divide by 0.001; platelet count (× 103/µL), divide by 1.0. † JAK2V617F mutation–positive cases.

the present study differed 500-fold, the completely concordant characterization of JAK2V617F– samples indicated that if the mutation was, in fact, present in any of the samples, it was present at less than 0.01%, making it undetectable even by AS-PCR. Moreover, the identification of all positive samples by the RFLP methods, even when DNA input was less than recommended, is also indicative of the robustness of these methods. Therefore, for this sample set, the RFLP assays seem functionally comparable to AS-PCR, and the results indicate that the technical considerations associated with the

RFLP methods may be largely managed and controlled. A similar conclusion about the diagnostic advantage of ASPCR as compared with other methods is offered in comparable studies.13,15 Had this study, however, included samples weakly positive for JAK2V617F (T is rare in acute leukemias but can be found in CMML, Philadelphia chromosome–negative CML, and megakaryocytic leukemia. Blood. 2005;106:3370-3373. 10. Greiner TC. Diagnostic assays for the JAK2 V617F mutation in chronic myeloproliferative disorders. Am J Clin Pathol. 2006;125:651-653. 11. Steensma D. JAK2 V617F in myeloid disorders: molecular diagnostic techniques and their clinical utility. J Mol Diagn. 2006;8:397-411. 12. Olsen RJ, Tang Z, Farkas DH, et al. Detection of the JAK2V617F mutation in myeloproliferative disorders by melting curve analysis using the LightCycler system. Arch Pathol Lab Med. 2006;130:997-1003. 13. Stevenson WS, Hoyt R, Bell A, et al. Genetic heterogeneity of granulocytes for the JAK2 V617F mutation in essential thrombocythaemia: implications for mutation detection in peripheral blood. Pathology. 2006;38:336-342. 14. Murugesan G, Aboudola S, Szpurka H, et al. Identification of the JAK2 V617F mutation in chronic myeloproliferative disorders using FRET probes and melting curve analysis. Am J Clin Pathol. 2006;125:625-633. 15. McClure R, Mai M, Lasho T. Validation of two clinically useful assays for evaluation of JAK2 V617F mutation in chronic myeloproliferative disorders. Leukemia. 2006;20:168171. 16. Jones AV, Kreil S, Zoi K, et al. Widespread occurrence of the JAK2 V617F mutation in chronic myeloproliferative disorders. Blood. 2005;106:2162-2168. 17. Vannucchi AM, Pancrazzi A, Bogani C, et al. A quantitative assay for JAK2V617F mutation in myeloproliferative disorders by ARMS-PCR and capillary electrophoresis. Leukemia. 2006;20:1055-1060. 18. Lay M, Mariappan R, Gotlib J, et al. Detection of the JAK2 V617F mutation by LightCycler PCR and probe dissociation analysis. J Mol Diagn. 2006;8:330-334. 19. James C, Delhommeau F, Marzac C, et al. Detection of JAK2 V617F as a first intention diagnostic test for erythrocytosis. Leukemia. 2006;20:350-353. 20. Pan Q, Pao W, Ladanyi M. Rapid polymerase chain reaction–based detection of epidermal growth factor receptor gene mutations in lung adenocarcinomas. J Mol Diagn. 2005;7:396-403. 21. Lippert E, Boissinot M, Kralovics R, et al. The JAK2-V617F mutation is frequently present at diagnosis in patients with essential thrombocythemia and polycythemia vera. Blood. 2006;108:1865-1867. 22. Sidon P, El Housni EI, Dessars B, et al. The JAK2V617F mutation is detectable at a very low level in peripheral blood of healthy donors [letter]. Leukemia. 2006;20:1622.

Am J Clin Pathol 2007;128:865-874



873 873


23. Xu X, Luo J, Xing S, et al. JAK2V617F: prevalence in a large Chinese hospital population. Blood. 2007;109:339-342. 24. Biernaux C, Loos M, Sels A, et al. Detection of major bcr-abl gene expression at a very low level in blood cells of some healthy individuals. Blood. 1995;86:3118-3122. 25. Bose S, Deininger M, Gora-Tybor J, et al. The presence of typical and atypical bcr-abl fusion genes in leukocytes of normal individuals: biologic significance and implications for the assessment of minimal residual disease. Blood. 1998;9:3362-3367. 26. Khwaja A. The role of Janus kinases in haemopoiesis and haematological malignancy. Br J Haematol. 2006;134:366-384. 27. Ma W, Kantarjian H, Verstovsek S, et al. Hemizygous/homozygous and heterozygous JAK2 mutation detected in plasma of patients with myeloproliferative diseases: correlation with clinical behaviour. Br J Haematol. 2006;134:340-350.

874 874


28. McLornan D, Percy M, McMullin MF. JAK2 V617F: a single mutation in the myeloproliferative group of disorders. Ulster Med J. 2006;75:112-119. 29. Tefferi A, Strand JJ, Lasho TL, et al. Respective clustering of unfavorable and favorable cytogenetic clones in myelofibrosis with myeloid metaplasia with homozygosity for JAK2V617F and response to erythropoietin therapy. Cancer. 2006;106:17391743. 30. Tefferi A, Lasho TL, Schwager SM, et al. The clinical phenotype of wildtype, heterozygous, and homozygous JAK2V617F in polycythemia vera. Cancer. 2006;106:631-635. 31. Villeval J-L, James C, Pisani DF, et al. New insights into the pathogenesis of JAK2 V617F-positive myeloproliferative disorders and consequences for the management of patients. Semin Thromb Hemost. 2006;32:341-351.

© American Society for Clinical Pathology