AJCP / Original Article
Evaluation of Allele-Specific PCR and Immunohistochemistry for the Detection of BRAF V600E Mutations in Hairy Cell Leukemia Noah A. Brown, MD, Bryan L. Betz, PhD, Helmut C. Weigelin, Kojo S. J. Elenitoba-Johnson, MD, Megan S. Lim, MD, PhD, and Nathanael G. Bailey, MD From the Department of Pathology, University of Michigan, Ann Arbor.
CME/SAM
Key Words: Hairy cell leukemia; BRAF; Immunohistochemistry; VE1; Allele-specific PCR; PCR; Antibody Am J Clin Pathol January 2015;143:89-99 DOI: 10.1309/AJCPDN4Q1JTFGCFC
ABSTRACT Objectives: Detection of BRAF V600E mutations in hairy cell leukemia (HCL) has important diagnostic utility. In this study, we sought to compare immunohistochemistry with an antibody specific for this mutation to a sensitive molecular assay. Methods: The performance of the BRAF V600E–specific VE1 antibody was compared with that of allele-specific polymerase chain reaction (PCR) in 22 formalinfixed, paraffin-embedded (FFPE) specimens with HCL involvement, along with nine splenic marginal zone lymphomas (SMZLs), 10 follicular lymphomas (FLs), 10 mantle cell lymphomas (MCLs), and 10 chronic lymphocytic leukemia/small lymphocytic lymphomas (CLL/SLLs). An additional 11 SMZLs, 100 FLs, 20 MCLs, 83 CLL/SLL specimens, and 49 reactive tonsils within tissue microarrays were stained with VE1. Results: A BRAF V600E mutation was detected in 17 (77.3%) of 22 HCL cases by PCR. Immunohistochemistry demonstrated VE1 staining in 20 (90.9%) cases, identifying low-level (~1%) involvement in three HCL cases that were mutation negative by PCR. Evaluation of additional material from these patients confirmed the presence of BRAF V600E. Thirty-nine non-HCL cases were negative by both methods. Within tissue microarrays, weak false-positive staining was observed in two (0.8%) of 263 non-HCL cases. Conclusions: VE1 immunohistochemistry is more sensitive than allele-specific PCR in FFPE bone marrow specimens and can be applied to decalcified core biopsy specimens that are not appropriate for molecular techniques. © American Society for Clinical Pathology
Upon completion of this activity you will be able to: • describe the relevance of BRAF mutations in hairy cell leukemia (HCL). • discuss the challenges of specimens collected from patients with HCL. • choose the appropriate BRAF mutation detection method for various sample types. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 151. Exam is located at www.ascp.org/ajcpcme.
Hairy cell leukemia (HCL) is a distinct mature B-cell neoplasm traditionally diagnosed based on morphologic and immunophenotypic features. However, HCL can sometimes be difficult to distinguish from other HCLlike B-cell lymphomas, including HCL variant, splenic marginal zone lymphoma (SMZL), splenic diffuse red pulp small B-cell lymphoma, and splenic B-cell lymphoma/ leukemia, unclassifiable. BRAF V600E mutations were recently identified in the leukemic cells from patients with HCL.1 This observation suggests that BRAF mutation analysis may be a useful potential diagnostic tool for HCL. In early literature, BRAF V600E mutations were found in 100% cases of HCL.2-6 However, several subsequent studies7-10 have demonstrated BRAF V600E–negative cases. In addition to HCL, BRAF V600E mutations have also been described at low frequency in other mature B-cell lymphoproliferative disorders,4 chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL),11-13 B-prolymphocytic leukemia,12 and, rarely, splenic B-cell lymphoma/leukemia, unclassifiable.14
Am J Clin Pathol 2015;143:89-99 89
DOI: 10.1309/AJCPDN4Q1JTFGCFC
Brown et al / Hairy Cell BRAF IHC and PCR
Methods used for detection of BRAF V600E in previous studies have included Sanger sequencing,1 highresolution melting,3 allele-specific polymerase chain reaction (PCR),2,4,7,8 and pyrosequencing.5,6,9,10 However, we recently showed that bone marrow aspirates frequently have low-level involvement.15 Therefore, accurate testing requires the use of molecular techniques with high analytic sensitivity. An antibody specific for the B-Raf protein with a V600E mutation (clone VE1) was recently developed and employed in melanoma and papillary thyroid carcinoma16-18 as well as in colorectal cancer19-21 and polyps.22 Andrulis et al13 recently investigated the utility of this stain in HCL. However, to our knowledge, no study has compared the performance characteristics of this antibody with sensitive molecular techniques in HCL specimens. In this study, we sought to compare the analytic performance of VE1 immunohistochemistry (IHC) with that of an allelespecific PCR assay for the detection of BRAF mutations in formalin-fixed, paraffin-embedded (FFPE) tissue from patients with HCL as well as control patients with nonHCL lymphomas.
Materials and Methods Patient and Sample Selection Medical records from the preceding 10 years at the University of Michigan were queried, and 22 samples from 22 different patients diagnosed with HCL were identified. All available clinical, flow cytometric, immunohistochemical, and histopathologic data were derived from the medical record for each sample. Diagnoses were made according to the 2008 WHO Classification of Tumours of Hematopoietic and Lymphoid Tissues.23 All cases of HCL had classic morphologic and immunophenotypic features aside from one case showing moderate expression of CD5. All cases showed expression of CD20, CD22, CD11C, CD25, and CD103. Of the 19 cases in which staining for tartrateresistant acid phosphatase (TRAP) was performed, all were positive (13 by cytochemistry, six by IHC). All aspirate clot sections were evaluated by both IHC and allele-specific PCR. Available decalcified core biopsy specimens were evaluated by IHC alone. Core biopsy specimens collected prior to April 11, 2011, were decalcified using RDO rapid decalcifier (Apex Engineering Products, Aurora, IL) for variable lengths of time. Core biopsy specimens collected after April 11, 2011, were decalcified using Formical-2000 (Decal Chemical Corporation, Tallman, NY) for 2 hours and 15 minutes. An additional 40 cases of non-HCL mature B-cell lymphomas were selected, including 10 each of follicular
90
Am J Clin Pathol 2015;143:89-99 DOI: 10.1309/AJCPDN4Q1JTFGCFC
lymphoma (FL), mantle cell lymphoma (MCL), and CLL/SLL and nine SMZLs. These specimens were also evaluated by both IHC and allele-specific PCR. Finally, tissue microarrays containing an additional 11 cases of SMZL, 100 cases of FL, 20 cases of MCL, 83 cases of CLL/SLL, and 49 reactive tonsils were selected for IHC staining only. Immunohistochemistry FFPE tissue blocks were cut in 4-µm-thick sections and processed for IHC. Antibodies against V600E-mutated B-Raf protein (VE1, 1:40; Spring Bioscience, Pleasanton, CA) and CD20 (L26, predilute; Ventana Medical Systems, Tucson, AZ) were used. Staining was performed using the Ventana BenchMark ULTRA automated immunostainer with the Ventana ultraView DAB (diaminobenzidine tetrahydrochloride) kit for detection (Ventana Medical Systems). Conditions of VE1 staining include heatinduced epitope retrieval at 95°C with Cell Conditioning 1 (Ventana Medical Systems) for 64 minutes (pH 8.5 buffer), a 60-minute primary antibody incubation at 37°C, and amplification of the signal with the Ventana amplification kit (Ventana Medical Systems). The counterstain was Harris hematoxylin. Immunoreactivity was independently scored by two hematopathologists (N.A.B. and N.G.B.), including both the percentage of immunoreactive cells and the intensity of reactivity. Intensity scores of 0, 1+, 2+, and 3+ corresponded to no, weak, moderate, and strong staining, respectively. DNA Isolation DNA was extracted from all HCL and 39 non-HCL specimens. Genomic DNA extraction from FFPE tissue blocks was performed on the BioRobot EZ1 (Qiagen, Valencia, CA) using the paraffin section protocol. For each block, one or two sections of 10-μm thickness were used for extraction. DNA was eluted in a final volume of 100 μL of Tris-EDTA buffer. Laboratory-Developed Allele-Specific PCR We assessed 5 µL of each sample for BRAF V600E and V600K mutations using a multiplexed allele-specific PCR method as described by Brown et al.15 While a prior dilutional study using FFPE tissue demonstrated a limit of detection as low as 1.25% allele frequency, the reported analytic sensitivity was conservatively set at 5% to allow for variability in specimen quality/quantity. Adequate amplification product in the control reaction (total BRAF)— greater than 5,000 relative fluorescence units (RFU)—was required for reporting a negative result. The assay was repeated at 1:2 and 1:5 dilutions if the control product
© American Society for Clinical Pathology
AJCP / Original Article
failed to meet the 5,000 RFU requirement. The assay was also repeated to confirm low positive results if the V600E/ V600K product was less than 500 RFU. Any detectable V600E/V600K product was sufficient to confirm a previous low positive result. Sanger Sequencing Sanger sequencing of exon 15 of the BRAF gene was performed as described by Hookim et al24 on two HCL samples that were negative for BRAF V600E and V600K despite high levels of HCL involvement and high levels of total BRAF amplification. Statistical Analysis Because the level of HCL involvement may be low in many cases, a true gold standard such as Sanger sequencing cannot be performed for all specimens, only those with high-level involvement. Therefore, for cases with high-level involvement without evidence of a mutation based on IHC and allele-specific PCR, the absence of a mutation was confirmed by Sanger sequencing. HCL cases without evidence of a BRAF V600E mutation by PCR, IHC, and Sanger sequencing despite high-level involvement were considered true BRAF V600E–negative cases. Discrepancies involving cases with low-level involvement were resolved by performing allele-specific PCR of additional material from the same patient with a higher level of involvement. Statistical analysis was carried out using SPSS version 21 (SPSS, Chicago, IL) to evaluate the degree of agreement for IHC interpretation between the two observers. The k coefficients were calculated for intensity scoring and final scored results (positive vs negative). Pearson correlation coefficients were calculated for the scoring of the percentage of cells with staining. According to Landis and Koch,25 k values of 0.4 to 0.6 are considered evidence of “moderate” agreement, more than 0.6 to 0.8 of “substantial” agreement, and more than 0.8 to 1 of “almost perfect” agreement.
Results Specimen Characteristics Specimen characteristics and results for each assay are summarized in ❚Table 1❚. Twenty-two archived FFPE bone marrow samples from patients diagnosed with classic HCL over the preceding 10 years were identified. The average specimen age was 5.1 years. Aspirate clot sections were available for all 22 cases. Bone marrow core biopsy specimens were available for 19.
© American Society for Clinical Pathology
Allele-Specific PCR BRAF V600E mutations were detected in 17 (77.3%) of 22 cases. The assay was 100% specific, with no positive cases in any non-HCL lymphoma (0/39). Immunohistochemistry Immunoreactive cells for VE1 were seen in 20 of 22 HCL cases and none of the non-HCL cases. Staining was consistently cytoplasmic and ranged from weak to moderate in intensity ❚Image 1❚. Occasional weak, nonspecific staining was observed in megakaryocytes. A higher level of involvement was observed in 12 of 18 core biopsy specimens compared with corresponding aspirate clot sections ❚Figure 1❚. In most cases, all or nearly all leukemic cells showed staining with VE1 based on the observation that a similar proportion of cells showed staining with CD20 and VE1. However, three decalcified core biopsy specimens (cases 9, 15, and 20) showed only weak staining in a subset of leukemic cells despite moderate staining in all leukemic cells within the aspirate clot section ❚Image 2❚. These three biopsy specimens were subjected to RDO decalcification for an unknown period. All non-HCL whole-tissue sections were negative (0/39), although nonspecific background staining was noted in one case. Within tissue microarrays, weak staining was observed in one case of FL and one case of CLL/SLL. Sanger sequencing of BRAF exon 15 for these two cases did not demonstrate any mutations. No clear staining of neoplastic cells was observed in the remaining 11 cases of SMZL, 99 cases of FL, 20 cases of MCL, and 82 cases of CLL/SLL. All 49 tissues of reactive tonsils were negative for BRAF VE1. A concordance of 100% was observed between two independent scorers with regard to the presence or absence of VE1-positive cells. As shown in ❚Table 2❚, excellent concordance was observed with regard to intensity of staining (κ = 0.871) and percentage of positive cells (Pearson correlation coefficient = 0.954). BRAF Wild-Type Cases With High-Level Involvement Two specimens (cases 13 and 18) ❚Image 3❚ were V600E negative by both methods despite high levels of leukemic involvement (80% and 100% by CD20 IHC, respectively). The absence of BRAF mutations in exon 15 was confirmed by Sanger sequencing, likely reflecting true BRAF V600E–negative HCL cases. Discrepant Cases Three aspirate clot sections (cases 1, 2, and 3) ❚Image 4❚ demonstrated staining with the VE1 antibody but were negative or indeterminate by allele-specific PCR. The percentage of cells with staining ranged from less than 1% to approximately 5% in these cases, and a similar percentage of cells showed staining with both the VE1 and CD20 antibodies. A
Am J Clin Pathol 2015;143:89-99 91
DOI: 10.1309/AJCPDN4Q1JTFGCFC
Brown et al / Hairy Cell BRAF IHC and PCR
❚Table 1❚ Specimen Characteristics and Results Specimen Characteristics Sample No.
Diagnostic/Posttreatment
Specimen Age, y
% Involvement (CD20)
Concomitant Flow Cytometry, %a
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Diagnostic Posttreatment Posttreatment Posttreatment Posttreatment Diagnostic Posttreatment Posttreatment Diagnostic Diagnostic Diagnostic Diagnostic Diagnostic Diagnostic Posttreatment
1 8 6 5 4 3 5 1 4 7 7 9 1 4 2 9 9 2 10 4 2 5
Evaluation of Allele-Specific PCR and Immunohistochemistry for the Detection of BRAF V600E Mutations in Hairy Cell Leukemia Noah A. Brown, MD, Bryan L. Betz, PhD, Helmut C. Weigelin, Kojo S. J. Elenitoba-Johnson, MD, Megan S. Lim, MD, PhD, and Nathanael G. Bailey, MD From the Department of Pathology, University of Michigan, Ann Arbor.
CME/SAM
Key Words: Hairy cell leukemia; BRAF; Immunohistochemistry; VE1; Allele-specific PCR; PCR; Antibody Am J Clin Pathol January 2015;143:89-99 DOI: 10.1309/AJCPDN4Q1JTFGCFC
ABSTRACT Objectives: Detection of BRAF V600E mutations in hairy cell leukemia (HCL) has important diagnostic utility. In this study, we sought to compare immunohistochemistry with an antibody specific for this mutation to a sensitive molecular assay. Methods: The performance of the BRAF V600E–specific VE1 antibody was compared with that of allele-specific polymerase chain reaction (PCR) in 22 formalinfixed, paraffin-embedded (FFPE) specimens with HCL involvement, along with nine splenic marginal zone lymphomas (SMZLs), 10 follicular lymphomas (FLs), 10 mantle cell lymphomas (MCLs), and 10 chronic lymphocytic leukemia/small lymphocytic lymphomas (CLL/SLLs). An additional 11 SMZLs, 100 FLs, 20 MCLs, 83 CLL/SLL specimens, and 49 reactive tonsils within tissue microarrays were stained with VE1. Results: A BRAF V600E mutation was detected in 17 (77.3%) of 22 HCL cases by PCR. Immunohistochemistry demonstrated VE1 staining in 20 (90.9%) cases, identifying low-level (~1%) involvement in three HCL cases that were mutation negative by PCR. Evaluation of additional material from these patients confirmed the presence of BRAF V600E. Thirty-nine non-HCL cases were negative by both methods. Within tissue microarrays, weak false-positive staining was observed in two (0.8%) of 263 non-HCL cases. Conclusions: VE1 immunohistochemistry is more sensitive than allele-specific PCR in FFPE bone marrow specimens and can be applied to decalcified core biopsy specimens that are not appropriate for molecular techniques. © American Society for Clinical Pathology
Upon completion of this activity you will be able to: • describe the relevance of BRAF mutations in hairy cell leukemia (HCL). • discuss the challenges of specimens collected from patients with HCL. • choose the appropriate BRAF mutation detection method for various sample types. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 151. Exam is located at www.ascp.org/ajcpcme.
Hairy cell leukemia (HCL) is a distinct mature B-cell neoplasm traditionally diagnosed based on morphologic and immunophenotypic features. However, HCL can sometimes be difficult to distinguish from other HCLlike B-cell lymphomas, including HCL variant, splenic marginal zone lymphoma (SMZL), splenic diffuse red pulp small B-cell lymphoma, and splenic B-cell lymphoma/ leukemia, unclassifiable. BRAF V600E mutations were recently identified in the leukemic cells from patients with HCL.1 This observation suggests that BRAF mutation analysis may be a useful potential diagnostic tool for HCL. In early literature, BRAF V600E mutations were found in 100% cases of HCL.2-6 However, several subsequent studies7-10 have demonstrated BRAF V600E–negative cases. In addition to HCL, BRAF V600E mutations have also been described at low frequency in other mature B-cell lymphoproliferative disorders,4 chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL),11-13 B-prolymphocytic leukemia,12 and, rarely, splenic B-cell lymphoma/leukemia, unclassifiable.14
Am J Clin Pathol 2015;143:89-99 89
DOI: 10.1309/AJCPDN4Q1JTFGCFC
Brown et al / Hairy Cell BRAF IHC and PCR
Methods used for detection of BRAF V600E in previous studies have included Sanger sequencing,1 highresolution melting,3 allele-specific polymerase chain reaction (PCR),2,4,7,8 and pyrosequencing.5,6,9,10 However, we recently showed that bone marrow aspirates frequently have low-level involvement.15 Therefore, accurate testing requires the use of molecular techniques with high analytic sensitivity. An antibody specific for the B-Raf protein with a V600E mutation (clone VE1) was recently developed and employed in melanoma and papillary thyroid carcinoma16-18 as well as in colorectal cancer19-21 and polyps.22 Andrulis et al13 recently investigated the utility of this stain in HCL. However, to our knowledge, no study has compared the performance characteristics of this antibody with sensitive molecular techniques in HCL specimens. In this study, we sought to compare the analytic performance of VE1 immunohistochemistry (IHC) with that of an allelespecific PCR assay for the detection of BRAF mutations in formalin-fixed, paraffin-embedded (FFPE) tissue from patients with HCL as well as control patients with nonHCL lymphomas.
Materials and Methods Patient and Sample Selection Medical records from the preceding 10 years at the University of Michigan were queried, and 22 samples from 22 different patients diagnosed with HCL were identified. All available clinical, flow cytometric, immunohistochemical, and histopathologic data were derived from the medical record for each sample. Diagnoses were made according to the 2008 WHO Classification of Tumours of Hematopoietic and Lymphoid Tissues.23 All cases of HCL had classic morphologic and immunophenotypic features aside from one case showing moderate expression of CD5. All cases showed expression of CD20, CD22, CD11C, CD25, and CD103. Of the 19 cases in which staining for tartrateresistant acid phosphatase (TRAP) was performed, all were positive (13 by cytochemistry, six by IHC). All aspirate clot sections were evaluated by both IHC and allele-specific PCR. Available decalcified core biopsy specimens were evaluated by IHC alone. Core biopsy specimens collected prior to April 11, 2011, were decalcified using RDO rapid decalcifier (Apex Engineering Products, Aurora, IL) for variable lengths of time. Core biopsy specimens collected after April 11, 2011, were decalcified using Formical-2000 (Decal Chemical Corporation, Tallman, NY) for 2 hours and 15 minutes. An additional 40 cases of non-HCL mature B-cell lymphomas were selected, including 10 each of follicular
90
Am J Clin Pathol 2015;143:89-99 DOI: 10.1309/AJCPDN4Q1JTFGCFC
lymphoma (FL), mantle cell lymphoma (MCL), and CLL/SLL and nine SMZLs. These specimens were also evaluated by both IHC and allele-specific PCR. Finally, tissue microarrays containing an additional 11 cases of SMZL, 100 cases of FL, 20 cases of MCL, 83 cases of CLL/SLL, and 49 reactive tonsils were selected for IHC staining only. Immunohistochemistry FFPE tissue blocks were cut in 4-µm-thick sections and processed for IHC. Antibodies against V600E-mutated B-Raf protein (VE1, 1:40; Spring Bioscience, Pleasanton, CA) and CD20 (L26, predilute; Ventana Medical Systems, Tucson, AZ) were used. Staining was performed using the Ventana BenchMark ULTRA automated immunostainer with the Ventana ultraView DAB (diaminobenzidine tetrahydrochloride) kit for detection (Ventana Medical Systems). Conditions of VE1 staining include heatinduced epitope retrieval at 95°C with Cell Conditioning 1 (Ventana Medical Systems) for 64 minutes (pH 8.5 buffer), a 60-minute primary antibody incubation at 37°C, and amplification of the signal with the Ventana amplification kit (Ventana Medical Systems). The counterstain was Harris hematoxylin. Immunoreactivity was independently scored by two hematopathologists (N.A.B. and N.G.B.), including both the percentage of immunoreactive cells and the intensity of reactivity. Intensity scores of 0, 1+, 2+, and 3+ corresponded to no, weak, moderate, and strong staining, respectively. DNA Isolation DNA was extracted from all HCL and 39 non-HCL specimens. Genomic DNA extraction from FFPE tissue blocks was performed on the BioRobot EZ1 (Qiagen, Valencia, CA) using the paraffin section protocol. For each block, one or two sections of 10-μm thickness were used for extraction. DNA was eluted in a final volume of 100 μL of Tris-EDTA buffer. Laboratory-Developed Allele-Specific PCR We assessed 5 µL of each sample for BRAF V600E and V600K mutations using a multiplexed allele-specific PCR method as described by Brown et al.15 While a prior dilutional study using FFPE tissue demonstrated a limit of detection as low as 1.25% allele frequency, the reported analytic sensitivity was conservatively set at 5% to allow for variability in specimen quality/quantity. Adequate amplification product in the control reaction (total BRAF)— greater than 5,000 relative fluorescence units (RFU)—was required for reporting a negative result. The assay was repeated at 1:2 and 1:5 dilutions if the control product
© American Society for Clinical Pathology
AJCP / Original Article
failed to meet the 5,000 RFU requirement. The assay was also repeated to confirm low positive results if the V600E/ V600K product was less than 500 RFU. Any detectable V600E/V600K product was sufficient to confirm a previous low positive result. Sanger Sequencing Sanger sequencing of exon 15 of the BRAF gene was performed as described by Hookim et al24 on two HCL samples that were negative for BRAF V600E and V600K despite high levels of HCL involvement and high levels of total BRAF amplification. Statistical Analysis Because the level of HCL involvement may be low in many cases, a true gold standard such as Sanger sequencing cannot be performed for all specimens, only those with high-level involvement. Therefore, for cases with high-level involvement without evidence of a mutation based on IHC and allele-specific PCR, the absence of a mutation was confirmed by Sanger sequencing. HCL cases without evidence of a BRAF V600E mutation by PCR, IHC, and Sanger sequencing despite high-level involvement were considered true BRAF V600E–negative cases. Discrepancies involving cases with low-level involvement were resolved by performing allele-specific PCR of additional material from the same patient with a higher level of involvement. Statistical analysis was carried out using SPSS version 21 (SPSS, Chicago, IL) to evaluate the degree of agreement for IHC interpretation between the two observers. The k coefficients were calculated for intensity scoring and final scored results (positive vs negative). Pearson correlation coefficients were calculated for the scoring of the percentage of cells with staining. According to Landis and Koch,25 k values of 0.4 to 0.6 are considered evidence of “moderate” agreement, more than 0.6 to 0.8 of “substantial” agreement, and more than 0.8 to 1 of “almost perfect” agreement.
Results Specimen Characteristics Specimen characteristics and results for each assay are summarized in ❚Table 1❚. Twenty-two archived FFPE bone marrow samples from patients diagnosed with classic HCL over the preceding 10 years were identified. The average specimen age was 5.1 years. Aspirate clot sections were available for all 22 cases. Bone marrow core biopsy specimens were available for 19.
© American Society for Clinical Pathology
Allele-Specific PCR BRAF V600E mutations were detected in 17 (77.3%) of 22 cases. The assay was 100% specific, with no positive cases in any non-HCL lymphoma (0/39). Immunohistochemistry Immunoreactive cells for VE1 were seen in 20 of 22 HCL cases and none of the non-HCL cases. Staining was consistently cytoplasmic and ranged from weak to moderate in intensity ❚Image 1❚. Occasional weak, nonspecific staining was observed in megakaryocytes. A higher level of involvement was observed in 12 of 18 core biopsy specimens compared with corresponding aspirate clot sections ❚Figure 1❚. In most cases, all or nearly all leukemic cells showed staining with VE1 based on the observation that a similar proportion of cells showed staining with CD20 and VE1. However, three decalcified core biopsy specimens (cases 9, 15, and 20) showed only weak staining in a subset of leukemic cells despite moderate staining in all leukemic cells within the aspirate clot section ❚Image 2❚. These three biopsy specimens were subjected to RDO decalcification for an unknown period. All non-HCL whole-tissue sections were negative (0/39), although nonspecific background staining was noted in one case. Within tissue microarrays, weak staining was observed in one case of FL and one case of CLL/SLL. Sanger sequencing of BRAF exon 15 for these two cases did not demonstrate any mutations. No clear staining of neoplastic cells was observed in the remaining 11 cases of SMZL, 99 cases of FL, 20 cases of MCL, and 82 cases of CLL/SLL. All 49 tissues of reactive tonsils were negative for BRAF VE1. A concordance of 100% was observed between two independent scorers with regard to the presence or absence of VE1-positive cells. As shown in ❚Table 2❚, excellent concordance was observed with regard to intensity of staining (κ = 0.871) and percentage of positive cells (Pearson correlation coefficient = 0.954). BRAF Wild-Type Cases With High-Level Involvement Two specimens (cases 13 and 18) ❚Image 3❚ were V600E negative by both methods despite high levels of leukemic involvement (80% and 100% by CD20 IHC, respectively). The absence of BRAF mutations in exon 15 was confirmed by Sanger sequencing, likely reflecting true BRAF V600E–negative HCL cases. Discrepant Cases Three aspirate clot sections (cases 1, 2, and 3) ❚Image 4❚ demonstrated staining with the VE1 antibody but were negative or indeterminate by allele-specific PCR. The percentage of cells with staining ranged from less than 1% to approximately 5% in these cases, and a similar percentage of cells showed staining with both the VE1 and CD20 antibodies. A
Am J Clin Pathol 2015;143:89-99 91
DOI: 10.1309/AJCPDN4Q1JTFGCFC
Brown et al / Hairy Cell BRAF IHC and PCR
❚Table 1❚ Specimen Characteristics and Results Specimen Characteristics Sample No.
Diagnostic/Posttreatment
Specimen Age, y
% Involvement (CD20)
Concomitant Flow Cytometry, %a
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Posttreatment Diagnostic Posttreatment Posttreatment Posttreatment Posttreatment Diagnostic Posttreatment Posttreatment Diagnostic Diagnostic Diagnostic Diagnostic Diagnostic Diagnostic Posttreatment
1 8 6 5 4 3 5 1 4 7 7 9 1 4 2 9 9 2 10 4 2 5
