February 14, 2007
Detection of Clonal IGH Gene Rearrangements: Summary of Molecular Oncology Surveys of the College of American Pathologists
By Nikiforova, Marina N; Hsi, Eric D; Braziel, Rita M; Gulley, Margaret L; Et al
* Context.-The diagnosis of B-cell lymphoid malignancy can frequently be substantiated by detecting clonal immunoglobulin heavy chain (IGH ) gene rearrangements, which is typically done by polymerase chain reaction (PCR) amplification and/or Southern blot analysis.Objective.-To characterize current laboratory practice for the assessment of IGH rearrangements and to identify opportunities for improvement.
Design.-The data from the Molecular Oncology Profi- ciency Survey distributed to participating laboratories by the Molecular Pathology Committee of the College of American Pathologists from 1998 through 2003 were analyzed.
Results.-Thirty-nine proficiency survey specimens (29 positive and 10 negative for clonal IGH rearrangements) were distributed. For Southern blot analysis, 944 results were reported, with a successful response rate of 95%. For PCR detection, 2349 results were reported, with a successful response rate of 72%. A higher rate of successful responses by PCR was achieved using framework 3 primers in combination with other frameworks (82%) compared with framework 3 primers only (76%) and when fresh/ frozen (72%) compared with paraffin-embedded (65%) tissues were analyzed.
Conclusions.-The performance of the participating laboratories was very good, by both Southern blot and PCR analysis. As expected, Southern blot analysis consistently detects a higher proportion of IGH rearrangements than PCR analysis. Further improvement and standardization of the IGH PCR assay is important if it is to replace Southern blot analysis as the standard method. Participation in this survey is a valuable tool for assessing laboratory performance and it directs our attention to areas where we may improve laboratory practice.
(Arch Pathol Lab Med. 2007;131:185-189)
The diagnosis of lymphoproliferative disorders in patients with suspected lymphoid malignancies can be supported by the assessment of clonality, based on the fact that all malignant lymphocytes of a particular lymphoid malignancy have a common clonal origin whereas reactive lymphocytes are polyclonal. More than 90% of lymphoid malignancies are of B-cell origin. During B-cell development, immunoglobulin genes undergo a complex rearrangement process to produce diverse antibody coding sequences. Thus, polyclonal populations of B cells harbor polyclonal IGH gene rearrangements in contrast to neo-plastic B-cell populations that derive from a single cell having a specific unique rearrangement. Identification of such clonal populations is important to the diagnosis of B-cell leukemias and lymphomas.1
Polymerase chain reaction (PCR) amplification and Southern blot analysis are methods currently in widespread use for detection of clonal IGH rearrangements. In PCR, rearranged DNA is amplified with a series of consensus primers that are complementary to sequences of variable regions framework 1, framework 2, and framework 3 and to joining regions (J^sub H^) of the IGH gene (Figure). 2 Monoclonal and polyclonal PCR products can be discerned by their size distributions. There are several methodologies for evaluating PCR amplicons: conventional electrophoresis in ethidium bromide-stained agarose gels (AGE), polyacrylamide gel electrophoresis (PAGE), or capillary gel electrophoresis (CAGE) with laser detection. If a significant population of cells contains the same unique IGH rearrangement, it appears as a well-defined band on AGE and PAGE or as a single peak in CAGE. In contrast, polyclonal B cells yield a ladder or smear of bands or peaks. Southern blot analysis utilizes restriction endonucleases that cut genomic DNA into sequence specific fragments. These digested fragments are then separated by gel electrophoresis, transferred to a membrane, and hybridized with an IGH specific probe. A clonal population of cells is indicated by the presence of a distinct non-germline band, whereas normal or reactive populations have only germline bands.
Southern blot analysis is considered the gold standard for identifying clonal IGH rearrangements.3 Despite the high reliability of Southern blot analysis, it is increasingly being replaced by other techniques because of several disadvantages: it is time- consuming, technically demanding, and requires 10 to 20 g of high quality DNA.4,5 In contrast, the PCR technique is fast, requires a much smaller amount of DNA, can be performed on fresh, frozen, or paraffin-embedded specimens, and has a relatively good sensitivity for low level clones. However, false-positive PCR results may be a problem if the assay is poorly designed or if interpretation criteria are inadequate for discriminating monoclonal from polyclonal products. Falsenegative IGH PCR results are also a serious problem when primer complementarity design is suboptimal to detect all possible rearrangements or when somatic hypermutation of the IGH variable region leads to improper annealing of consensus primers to the rearranged IGH gene segment.6,7 A recently published study by van Dongen et al8 shows promise that standardization of PCR primers and protocols for detecting IGH gene rearrangements could yield improved outcomes, although this improvement is likely to require considerably more extensive PCR testing.
To facilitate studies in which new methods are compared with older ones used prior to 2003, the Molecular Pathology Committee of the College of American Pathologists ([CAP]; Northfield, Ill) reviewed the findings of a proficiency testing program that collects data on methods and outcomes of gene rearrangement testing. Frozen and paraffin-embedded tissue samples were distributed twice a year for the Molecular Oncology Proficiency Survey that includes assessment of IGH clonality.9 The aim of this study was to present the laboratory community's experience with IGH tests as performed by the laboratories subscribing to the Molecular Oncology Proficiency Survey, to characterize current laboratory practice for the detection of clonal IGH rearrangements, to estimate the utility and limitations of the currently used IGH assays, and to identify opportunities for improvement.
MATERIALS AND METHODS
The CAP Molecular Oncology Survey data from 1998 through 2003 was reviewed with respect to methods and results for detection of the IGH gene rearrangements by Southern blot analysis and DNA PCR amplification methods. All data were obtained from the CAP archives. Each survey consisted of at least 2 cell or tissue specimens that were sent to the subscribed laboratories with a questionnaire surveying methods and results. Survey specimens were procured and distributed as frozen case material or as cultured cells (called MO specimens). Since 1999, 2 additional specimens of formalin-fixed, paraffin-embedded tissue (FFPE) (called MOP specimens) were also made available. All tissue samples were distributed in compliance with institutional review board policies. To ensure the quality of the survey specimens, CAP reference laboratories provided confirmation of histopathologic diagnosis for tissue specimens or, in cases of cell lines, verified published cell line characteristics prior to sample distribution. In addition, survey specimens were pretested for quality and quantity of nucleic acids. Participating laboratories were blinded to the clinicopathologic data so as not to bias the interpretation of results. All shipments contained a detailed questionnaire regarding IGH rearrangement testing, including methods for nucleic acid extraction, assessment of quality of nucleic acids, restriction enzyme digestion, hybridization probe/ primer sources, frameworks used for IGH amplification, gel types, and detection strategies. In addition, the 2002 Molecular Oncology Survey contained supplemental questions regarding number of rounds of amplification, volume of reaction, amount of template DNA, MgCl2 concentration, hot start procedures, membrane transfer procedures, and primer sequences. CAP Molecular Pathology Committee members determined the ''successful'' results for each survey specimen based on consensus molecular results provided by participating laboratories in conjunction with the clinical, histologic, immunohistochemical, flow cytometric, karyotypic, and fluorescence in situ hybridization data provided by the original specimen source.
Positive, negative, and indeterminate results for the detection of monoclonal IGH rearrangements were calculated for each survey specimen. Further analysis was performed to determine the trends and differences in reported versus expected results in relation to the primers and frameworks used for amplification, gel types used for electrophoresis, and other procedure variations. Statistical analysis was performed using SAS version 9.1 software (SAS Inc, Cary, NC).
Thirty-nine Molecular Oncology proficiency survey specimens were distributed by the CAP Molecular Pathology Committee to participants from 1998 through 2003. Of the 39 specimens, 29 were positive and 10 were negative for IGH rearrangements. The specimens represented numerous histopathologic diagnoses, including non- Hodgkin lymphoma, posttransplantation lymphoproliferative disorder, chronic lymphocytic leukemia/small lymphocytic lymphoma, and benign lymphoid tissue, as well as cell lines obtainedfrom reliable vendors. Twenty- one specimens were sent as frozen tissue or cell lines, and 18 specimens were distributed as FFPE tissue sections.
One hundred sixty-one laboratories participated in the Molecular Oncology Proficiency Survey between 1998 and 2003. Of these laboratories, 85% were from the United States, 8% from Canada, and 7% from other countries. The majority of these laboratories were hospital affiliated. Two molecular techniques were used for the determination of IGH gene rearrangements by the vast majority of participants: Southern blot analysis and PCR amplification. Laboratory results were reported as ''positive for clonal IGH rearrangement,'' ''negative for clonal IGH rearrangement,'' or as ''indeterminate.'' Indeterminate results comprised a very small fraction of all data and were not included in the denominator for the purposes of this study. Participating laboratories reported results as indeterminate in 2 (0.2%) of 944 cases analyzed by Southern blot analysis and 89 (3.8%) of 2349 by DNA PCR amplification. The successful response rate (true-positive and true- negative results) was 95% for IGH gene rearrangements assessed by Southern blot analysis and 72% by DNA PCR amplification (Table 1). For the IGH positive specimens, a total of 714 results were registered of which 677 (94%) were successfully identified by Southern blot analysis. The same specimens analyzed by DNA PCR amplification generated a total of 1703 results of which 1166 (67%) were answered successfully. For the IGH negative specimens, 230 (97%) of 233 specimens were found to be negative for clonal IGH rearrangement by Southern blot analysis and 539 (87%) of 646 by PCR amplification (Table 1).
Results of DNA amplification were also analyzed in relation to disease category (Table 2). The most successful detection of amplification was achieved for lymphoplasmacytic lymphoma (97%), mantle cell lymphoma (95%), and chronic lymphocytic leukemia/small lymphocytic lymphoma (90%). Follicular lymphoma grades 1 or 2 (78%) and splenic marginal zone lymphoma (76%) showed an intermediate level of detection. However, specimens of posttransplantation lymphoproliferative disorder and diffuse large B-cell lymphoma were found to have a lower successful rate, of 41% and 31% respectively, for detection of IGH rearrangements by PCR amplification.
To directly evaluate the impact of sample preparation on the success of DNA amplification, 8 paired tumor specimens (16 total tissues blinded to the participants: 8 frozen and 8 matched FFPE) were sent to the participating laboratories. A successful response was reported for 72% of the frozen tissues samples compared with 65% for FFPE tissues (Table 3). Likely because of the limited sample size, no statistically significant difference was found; however, a trend toward poorer result outcome in FFPE tissues was evident in all but 1 of the 8 tumors, and this trend was borne out in the larger cohort of 39 tissues in which a success rate of 78% versus 66% was achieved for frozen versus FFPE preparations by PCR amplification testing, respectively.
One of the important factors that is known to affect the sensitivity of IGH PCR amplification is primer selection. Results of IGH testing therefore were analyzed with respect to usage of primers that target specific sequences of the variable region (frameworks 1, 2, and 3) (Figure). Participating laboratories that used framework 3 primers only provided a successful response in 547 (76%) of 719 tests, whereas laboratories that used framework 3 primers combined with any other framework primers demonstrated a successful response in 379 (82%) of 461 tests. However, there was no significant difference in the successful response rate and primer selection (P > .05, chi-square test for each year of data).
Three major gel types were used for the detection of IGH clonality after DNA PCR amplification. PAGE was found to be used most frequently by the participating laboratories, followed by AGE and CAGE. There was no significant difference found in the successful response rate by gel type (P > .05, Fisher exact test for each year of data).
In addition, multiple other technical variables were evaluated in the 2002 Molecular Oncology Survey. With this ample size there were no statistically significant differences in results related to the number of cycles of DNA PCR amplification, the volume of amplification reaction used, the amount of template DNA, or the MgCl^sub 2^ concentration. Table 4 shows a summary of the primer sequences used by the participating laboratories in the 2002 Molecular Oncology Survey. We were unable to determine any consistent differences in outcome of IGH clonality test results in correlation with the choice of primer sets.
The primary goal of this study was to evaluate methods and outcomes of IGH gene rearrangement testing as performed by the 161 laboratories subscribing to the CAPMolecular Oncology Proficiency Survey. These laboratories are not necessarily doing clinical testing for IGH; laboratories that are doing validation work or research also can subscribe to the survey. However, most participants are offering IGH gene rearrangement analysis on a clinical basis. Overall, participating laboratories demonstrated a very good performance in the detection of clonal IGH rearrangements by the Southern blot assay. Southern blot analysis continues to be the gold standard, and this was con- firmed in the current study in which 95% of specimens were successfully classified (true positive and true negative) by the participating laboratories using Southern blot analysis. The detection rate was slightly lower than the theoretical 100% detection of B-cell clones by Southern blot analysis because of the fact that some of the distributed specimens contained less than 10% of cells with clonal IGH rearrangements.
The successful result rate by Southern blot analysis was significantly higher than the rate for DNA PCR amplifi- cation (72%); we confirmed a high false-negative rate for PCR amplification assays compared with Southern blot. This difference was highly sample specific, suggesting that improper primer annealing to the tumor's rearranged IGH gene segment contributed to failed amplification. Approximately 30% of B-cell lymphomas of so-called post-germinal center origin, including follicular lymphomas, marginal zone lymphomas, and plasma cell dyscrasias, will fail to have a clonal result with standard IGH PCR. This is thought to be secondary to somatic hypermutation of the IGH gene variable region, including the sequences to which the IGH PCR primers are supposed to anneal. This will result in a lack of amplification of the neoplastic clone. In addition, occasional B-cell lymphomas have very complex rearrangements of the IGH genes and other associated genes that may confound standard IGH PCR testing.
A previously published study by Bagg et al,6 demonstrated that detection of monoclonal IGH rearrangements by PCR might be improved with the addition of framework 2 primers to framework 3 primers. In this study, the successful response rate was improved if laboratories used framework 3 primers in combination with other frameworks (82%) compared with framework 3 primers alone (76%), but this difference was not statistically significant.
Another factor that may be important in IGH clonality detection is the method of gel electrophoresis of the amplified product. Various gels have differing powers of resolution and reportedly have different levels of sensitivity for evaluation of PCR products.10,11 Of the gel types used by Molecular Oncology Survey participants, PAGE was used most frequently, followed by AGE and CAGE. All gel types demonstrated comparable correct detection rates and no significant difference was found.
Analysis of DNA PCR amplification on paired fresh/ frozen specimens and FFPE tissue demonstrated higher sensitivity for assays performed on fresh/frozen tissue compared with FFPE tissue; however, the difference was not statistically significant, possibly because of the low sample size. It is therefore advisable to obtain fresh/ frozen tissue for IGH PCR analysis when possible. If frozen tissue is available, Southern blot analysis also could be performed as a backup to negative PCR amplification test results, thus increasing the likelihood that a B-cell clone will be detected.
Our study shows that there is a considerable lack of standardization in the methods used to detect IGH gene rearrangements, with substantial variability in the analyte specific reagents used (probes and primers) as well as in the procedures for using these reagents. Nevertheless, most laboratories performed well in assessing clonal IGH gene rearrangements by Southern blot analysis, and the variability in amplification results seemed closely tied to sample-specific characteristics. These include, but are not limited to, the size of the neoplastic clone in relation to the polyclonal B-cell population, the presence of somatic hypermutation interference with primer binding, and the effects of fixation on availability and quality of amplifiable DNA. A controlled study is required to identify particular methods to overcome false-negative and false-positive results.
We thank Mrs Jill Kachin for the organizing summary data; Molly K. Walsh, PhD, for statistical analysis; and members of the CAP Molecular Pathology Resource Committee, Kevin C. Halling, MD, PhD; Jeffrey Kant, MD, PhD; Andre Olivera, MD; Herbert F. Polesky, MD; Lawrence Silverman, PhD; and James Versalovic, MD, PhD, for critically reviewing the manuscript.
1. Gill JI, Gulley ML. Immunoglobulin and T-cell receptor gene rearrangement. Hematol Oncol Clin North Am. 1994;8:751-770.
2. Gulley ML. Leukemias and lymphomas. In: Leonard DGB. Diagnostic Molecular Pathology. Philadelphia, Pa: Elsevier Science; 2003:113-127.
3. Medeiros LJ, Carr J. Overview of the role of molecular meth\ods in the diagnosis of malignant lymphomas. Arch Pathol Lab Med. 1999;123:1189-1207.
4. Cossman J, Zehnbauer B, Garrett CT, et al. Gene rearrangements in the diagnosis of lymphoma/leukemia: guidelines for use based on a multiinstitutional study. Am J Clin Pathol. 1991;95:347-354.
5. Sioutos N, Bagg A, Michaud GY, et al. Polymerase chain reaction versus Southern blot hybridization: detection of immunoglobulin heavy-chain gene rearrangements. Diagn Mol Pathol. 1995;4:8-13.
6. Bagg A, Braziel RM, Arber DA, Bijwaard KE, Chu AY. Immunoglobulin heavy chain gene analysis in lymphomas: a multi- center study demonstrating the heterogeneity of performance of polymerase chain reaction assays. J Mol Diagn. 2002;4:81-89.
7. Derksen PW, Langerak AW, Kerkhof E, et al. Comparison of different polymerase chain reaction-based approaches for clonality assessment of immunoglobulin heavy-chain gene rearrangements in B- cell neoplasia. Mod Pathol. 1999; 12:794-805.
8. van Dongen JJM, Langerak AW, Bruggemann M, et al. Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations Report of the BIOMED-2Concerted Action BMH4- CT98-3936. Leukemia. 2003;14: 2257-2317.
9. Hsi ED, Tubbs RR, Lovell MA, Braziel RM, Gulley ML. Detection of bcl-2/ J(H) translocation by polymerase chain reaction: a summary of the experience of the Molecular Oncology Survey of the College of American Pathologist. Arch Pathol Lab Med. 2002;126:902-908.
10. Gleissner B, Maurer J, Sindram A, et al. Comparison of ethidium bromidestained agarose gel electrophoresis and automated fragment analysis for evaluation of IGH gene products. Leukemia Res. 2001;25:769-774.
11. Gonzalez M, Gonzalez D, Lopez-Perez R, et al. Heteroduplex analysis of VDJ amplified segments from rearranged IGH genes for clonality assessments in B-cell non-Hodgkin's lymphoma: a comparison between different strategies. Haematologica. 1999;84:779-784.
Marina N. Nikiforova, MD; Eric D. Hsi, MD; Rita M. Braziel, MD; Margaret L. Gulley, MD; Debra G. B. Leonard, MD, PhD; Jan A. Nowak, MD, PhD; Raymond R. Tubbs, DO; Gail H. Vance, MD; Vivianna M. Van Deerlin, MD, PhD; for the Molecular Pathology Resource Committee, College of American Pathologists
Accepted for publication August 2, 2006.
From the Department of Pathology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio (Dr Nikiforova); the Department of Clinical Pathology, Cleveland Clinic Foundation, Cleveland, Ohio (Drs Hsi and Tubbs); the Department of Pathology, Oregon Health & Science University, Portland (Dr Braziel); the Department of Pathology, University of North Carolina, Chapel Hill (Dr Gulley); the Department of Pathology and Laboratory Medicine, Weill Medical College of Cornell University and New York Presbyterian Hospital, New York, NY (Dr Leonard); the Department of Pathology, Lutheran General Hospital, Park Ridge, Ill (Dr Nowak); Medical & Molecular Genetics, Indiana University Medical Center, Indianapolis (Dr Vance); and the Department of Pathology & Laboratory Medicine, University of Pennsylvania Health System, Philadelphia (Dr Van Deerlin).
The authors have no relevant financial interest in the products or companies described in this article.
Reprints: Marina N. Nikiforova, MD, Department of Pathology, University of Pittsburgh, A713 Scaife Hall, 3550 Terrace St, Pittsburgh, PA 15261 (e-mail: [email protected]).
Copyright College of American Pathologists Feb 2007
(c) 2007 Archives of Pathology & Laboratory Medicine. Provided by ProQuest Information and Learning. All rights Reserved.