Gene Rearrangements in the Diagnosis of Lymphoma/Leukemia Guidelines for Use Based on a Multiinstitutional Study

Transcription

1 HEMATOPATHOLOGY Gene Rearrangements in the Diagnosis of Lymphoma/Leukemia Guidelines for Use Based on a Multiinstitutional Study JEFFREY COSSMAN, M.D., BARBARA ZEHNBAUER, PH.D., CARLETON T. GARRETT, M.D., PH.D., LINDA J. SMITH, PH.D., MICHAEL WILLIAMS, M.D., ELAINE S. JAFFE, M.D., LINDA O. HANSON, AND JACK LOVE, PH.D. The demonstration of immunoglobulin or T-cell receptor gene rearrangements in human lymphoproliferative processes with the use of DNA hybridization has gained great popularity as a sensitive laboratory adjunct to diagnostic hematopathology. The fact that nearly all B- or T-cell malignant lymphomas and leukemias have one or more rearranged antigen receptor genes provides a biologic basis for a diagnostic test. To formally analyze the sensitivity, specificity, and reproducibility of gene rearrangements in the diagnosis of human lymphoproliferative disease, the authors conducted a large, multiinstitutional study. Through a blinded, controlled approach, gene rearrangement analysis of 75 cases was shown to carry a high correlation with conventional phenotyping and histologic diagnosis, with only minor false-positive and false-negative rates. Significantly, no rearrangements were detected in normal lymphoid tissues or carcinomas, sarcomas, or melanomas. In a randomized study of 5 cases, laboratory results showed a high rate of interlaboratory agreement, regardless of the level of previous experience. Furthermore, the reproducibility of interpretation of data (Southern blot autoradiograms) of 9 cases showed high concordance among observers from multiple laboratories. Based on these findings, the authors propose a set of guidelines for interpretation of gene rearrangement analysis that, if carefully followed, renders this a highly reproducible, safe, and accurate addition to the diagnostic regimen for human lymphoproliferative processes. (Key words: Lymphoma; Genes; Immunoglobulin; T-cell receptor) Am J Clin Pathol 99;95:37-35 The diagnosis of human lymphoproliferative processes culties encountered with phenotyping. " For example, can present considerable difficulty, particularly with respect to the distinction between benign and malignant disease and with regard to the subtype of disease. Although antibody phenotyping using immunohistochemical and flow cytometric analyses have been of assistance in the diagnostic laboratory, they may fall short in many instances for a variety of technical reasons.' The use of DNA hybridization to detect rearrangements of immunoglobulin or T-cell receptor genes obviates some of the diffibecause protein expression is not required for gene rearrangement analysis, problems relating to interpretation of immunoglobulin light chain excess for the identification of clonal B cells are avoided. 56 Furthermore, gene rearrangement analysis represents the only conveniently applied clonal marker for T-cell processes. " As a sensitive technique that provides unique, clonal tumor markers, gene rearrangement analysis has emerged as a major approach in diagnostic hematopathology. An expanding body of literature has provided detailed documentation of rearrangements of antigen receptor From Georgetown University, Washington. D.C.; Johns Hopkins genes University, Baltimore, Maryland: George Washington University. Washing in various selected examples of human lymphoproliferative of Vir disease. However, little is known concerning the ton. D.C.; University of Florida. Gainesville. Florida; University ginia. Charlottesville. Virginia; National Institutes of Health. Bethesda. sensitivity, specificity, and reproducibility of this test when Maryland; Oncor. Inc., Gaithersburg, Maryland; Bioline. Inc.. Rockville. Maryland. Received April, 99; received revised manuscript and accepted for publication August, 99. Address reprint requests to Dr. Cossman: Department of Pathology, Georgetown University School of Medicine, 39 Reservoir Road, NW, Washington, B.C. 7. used on a routine basis in a diagnostic laboratory. In the current report, we have derived a quantitative measure of the reproducibility and accuracy of gene rearrangement analysis through a random, blinded, multiinstitutional study of 75 patient tissue samples. The goals of this study were to determine the following: () the specificity and Downloaded from on 7 July 8 37

2 38 HEMATOPATHOLOGY degree of correlation with histologic diagnosis and phenotype; () the reproducibility of the laboratory assay; and (3) the reproducibility of data interpretation as performed by both experienced and inexperienced observers. Patient Samples Materials and Methods Tissue samples were obtained from 75 patients. They included 73 malignant lymphomas and leukemias, 35 examples of nonmalignant lymphoproliferative processes, normal peripheral blood lymphocyte samples, and 3 malignant tumors other than lymphoma. A detailed breakdown of the diagnoses is outlined in Table. Immunophenotyping All samples were analyzed and interpreted as previously described' with the use of frozen-section immunohistochemical studies for solid tissue and flow cytometry for fluid-phase cell suspensions (peripheral blood, bone marrow, or effusions). TABLE. DISTRIBUTION OF SPECIMENS ACCORDING TO DIAGNOSIS Diagnosis Malignant Chronic lymphocytic leukemia Well-differentiated lymphocytic lymphoma Intermediately differentiated lymphocytic lymphoma Follicular lymphoma Small cleaved Mixed, small, and large cell Large cell Large cell lymphoma Large cell Immunomoblastic Small noncleaved Burkitt's Non-Burkitt's Lymphoblastic lymphoma Adult T-cell leukemia/lymphoma Acute lymphoblastic leukemia Mycosis fungoides/sezary syndrome Indeterminate* Hodgkin's disease Ty lymphocytosis Acute nonlymphocytic leukemia Angioimmunoblastic lymphoadenopathy Nonmalignant disorders Lymphoid hyperplasia Mononucleosis, Epstein-Barr virus infection Histiocytosis Normal peripheral blood Indeterminate with respect to clonality and lineage. No. of Samples DNA Analysis Tissue samples included the following sites: lymph node, spleen, bone marrow, blood, skin, gastrointestinal tract, mesothelial surfaces, and effusions. Samples were freshly obtained and either snap frozen or, if in the form of cell suspensions, viably frozen and stored as described previously. 7 High molecular weight DNA was extracted and purified with the use of DNA extraction kits (Oncor, Inc., Gaithersburg, MD) or with a conventional phenolchloroform technique. 7 All DNA samples were digested with three restriction enzymes: EcoRl, HindUl, and BamHl. Restriction fragments were electrophoretically size-separated on gels, transferred to nylon membranes that were prehybridized, and hybridized with 3 P-radiolabeled DNA probes. Three DNA probes were used in this study and included genomic probes of the human immunoglobulin heavy chain joining region (J H, 5.6-kilobase [kb] fragment) (Fig. I A), the lambda joining region (J k,.8-kb fragment) (Fig. \B), and a cdna of the human T-cell receptor beta chain second constant region gene (C T /3,.5-kb fragment) (Fig. IC). Each gel included one lane containing size markers composed of a HindlW digestion of lambda phage labeled with 3 P and a lane containing high molecular weight human placental DNA digested with the same restriction enzyme corresponding to that of the other samples in the gel. Hybridizations and washings were performed under stringent conditions, and autoradiographic exposures on Kodak X-AR x-ray film were performed for four days. Reactions were performed to ensure complete digestion. Partial digestion has been found to be a particular problem with respect to a relatively resistant EcoRl restriction site immediately toward the 5' portion of the human Q8 locus (asterisk, Fig. IC). When incompletely digested at this site, a third band of 8.5 kb can be observed when using a C TJ 8 probe. We have found that prolongation of endonuclease digestion times can eliminate the 8.5-kb band (see Fig. ). Interpretation of Southern Blots For interpretation of Southern Blots, see Appendix A. Results Rearrangement Analysis Confirms Clonality of Lymphoma To determine the degree of concordance of gene rearrangement analysis with histologic diagnosis, samples were analyzed (Table ). Samples were separated into three general diagnostic groups: positive (malignant diagnosis), negative (normal or benign diagnosis), or indeterminate (angioimmunoblastic lymphadenopathy [AILD], acute Downloaded from on 7 July 8 A.J.C.P. March 99

3 COSSMAN ET AL. 39 Gene Rearrangements in Lymphoma Diagnosis /Will V H! -\Hi3ytfH fcori Eco Rl -H Kb B awhi Kb- ^77H I I Hind III 5.Kb /Wlll I Probe 9.Kb- I Kb fbori fi»ri fix)r I fcori fiaori feor r ftori Kb.8 Kb. Kb. Kb.8 Drift ffl J T Cjlft i D T /3 O-HH]f> JT/3 C T BamHl -.Kb- Probe Cjfll Kb 7. Kb /WW Hindi /Will Kb- I Probe C-flft i i i Hindi -~ * Kb &mhi FIG. I. Genomic restriction maps of the germline configurations of the genes analyzed in the study. A. Immunoglobulin heavy chain gene. B. Immunoglobulin kappa light chain gene. C. T-cell receptor beta chain gene. Vol. 95 No. 3 Downloaded from on 7 July 8

4 35 HEMATOPATHOLOGY CT/3 Markers EcoRI 3.Kb 9.Kb 6.6Kb.Kb 3.3Kb.3Kb.Kb.Kb *?e* -. ^H*vrf,-. FIG.. Partial digestion of the T-cell receptor # chain gene. Germline (placental) genomic DNA was digested for varying times with EcoRI. The Southern blot was hybridized with the C T 8 probe. Short digestion times (less than minutes) result in most of the DNA remaining as high molecular weight. The expected - and.-kb restriction fragments corresponding to QJi and QS, respectively, appear in samples digested minutes or more. However, a third band of 8.5 kb is apparent at early time points but not in the sample digested for 8 minutes. This band is a consequence of partial digestion resulting from the resistant restriction enzyme site indicated by an asterisk in Figure C. This fragment contains the Jft/Cft region and is not a rearrangement but rather an additional germline band due to partial digestion minutes nonlymphocytic leukemia, J y lymphoproliferative process, and Hodgkin's disease, because all of these have been variably shown to have gene rearrangements in previous reports). Five of the samples could not be analyzed because the DNA obtained from these was degraded. Of 5 cases with a malignant diagnosis, 36 were found to have gene rearrangements. This yields a false-negative rate of 5 of 5, or 9.9%. However, on careful inspection, it was found that, of the 5 "false-negative" cases, 6 were from patients with follicular lymphoma, but these samples contained few (less than 5%) B cells by phenotypic analysis and DNA Probe Results Positive Negative Total TABLE. INTERPRETATION Positive (malignant Dx) Clinical Diagnosis Negative (benign or normal) 6 6 Indeterminate 3 7 clones were not identified by either genotypic or phenotypic tests. Seven of the 5 "false-negative" cases were large cell lymphomas, including three T-cell lymphomas, which, as has been previously reported, may not show rearrangement 8 ; two were of true histiocytic origin and may not contain rearranged immunoglobulin or T-cell receptor genes; one node was only partially involved and had no apparent clone by phenotypic analysis in the cell samples studied here; and the final large cell lymphoma case had a "null" phenotype and may have been a true histiocytic process. Finally, two cases were small, noncleaved cell tumors other than Burkitt's but, again, nodes were only partially involved and showed no evidence of clonality by phenotypic analysis. Thus, it can be concluded that most false-negative findings derived from gene rearrangement analysis can be accounted for on the basis of partial or minor involvement of tissues. Furthermore, most of these represented additional biopsy specimens from patients with established diagnoses of non-hodgkin's lymphoma, and histologic diagnosis of partial involvement in these samples was made in this context. Of 6 cases with a "negative" diagnosis, for which gene rearrangements would not be expected on the basis of A.J.C.P. March 99 Downloaded from on 7 July 8

5 COSSMAN ET AL. 35 Gene Rearrangements i Lymphoma Diagnosis histologic findings and phenotype, showed evidence of gene rearrangement. Careful reexamination of these four cases showed that they represented the peripheral blood mononuclear cells of patients with chronic Epstein-Barr virus infection. However, in two of these samples nongermline bands previously considered as rearrangements were found to result either from partial digestion or crosshybridization. Therefore, these specimens were truly negative, producing a false-positive rate of of 6, or 3.%. Fourteen of the 7 "indeterminate" cases showed rearrangements (Table 3), in keeping with reported gene rearrangements " in type of process studied here. Other Cancers Twenty-three examples of malignancies other than lymphoma were tested, and none showed rearrangements with any of the three probes. Tumors studied in this series included the following: one neuroblastoma, two leiomyosarcomas, three rhabdomyosarcomas, two adenocarcinomas, two malignant melanomas, one spindle cell sarcoma, one stromal sarcoma, and eight osteogenic sarcomas. Thus, no "false-positive" results were observed with the use of the gene rearrangement approach in either normal lymphoid cells or nonlymphoid malignancies. Rearrangements as B-IT-Lineage Markers Rearrangement analysis was compared with immunologic phenotyping. With the use of the guidelines for determining T- or B-cell lineage based on gene rearrangements, cases were phenotyped as either B cell or pre- B cell; 97 showed B-cell rearrangement. Of 36 cases phenotyped as T cell, 3 showed rearrangement, for a total concordance of 9 of 36, or 9.9% agreement with phenotypic analysis. There were 3 cases for which the phenotype could not be determined, and, of these, gene rearrangement analysis resolved the lineage in (8 were B; were T). These 3 cases included 7 large cell lymphomas (5 B cell, T cell, remaining as null); acute lymphoblastic leukemia (ALL) (pre-b); small noncleaved, non- Burkitt's (B cell); 3 Hodgkin's disease ( kappa rearranged "B cell" and remaining as null); and mycosis fungoides (T cell). Reproducibility of Laboratory Analysis To determine whether the gene rearrangement test is reproducible, 5 cases were randomly selected for repeat testing at two separate laboratories. Among these samples, were examples of malignant lymphoma, benign hyperplasias, and normal tissue. The two laboratories differed in their prior experience in performing and evaluating the analyses and included one at Johns Hopkins University and one at George Washington University. All samples were handled in a blinded fashion so that no phenotypic, clinical, or diagnostic information was provided, and all samples received a random code number so that no patient identification could be obtained. All samples had been previously analyzed and interpreted at a central laboratory by two of us (J.L. and L.H.), and, after interpretation at the two outside laboratories, the films were reexamined and compared with the original central laboratory results by three experienced observers (J.L., L.H., and J.C). This final comparison was performed to determine the quality of the autoradiograms; for example, to compare the sizes of rearrangements as found between laboratories, to compare levels of intensity of rearranged bands, and to ensure that no sample switching had occurred. In all, it was judged that the results of the outside laboratories were excellent with respect to these parameters. With regard to the reproducibility of the results obtained with the C T S probe, the results of the experienced laboratory agreed with those of the central laboratory in 7 of 5 (9%) of the tests. The three disagreements all were a consequence of extremely faint bands obtained in the outside laboratory, which were considered as rearrangements. In each case, the intensity of the band was less than that of a sensitivity control containing 3% clonal cells. For this reason, it was recommended that bands must equal the intensity of the 3% positive sensitivity control to be considered as rearranged. Of the 5 samples examined by the other outside laboratory, contained insufficient DNA for analysis. Among the remaining 9, there was agreement on interpretation in 7. The two disagreements were both a consequence of misinterpretation of partially digested restriction fragments as rearrangements. Once these were reinterpreted according to the guidelines, they could be resolved. Thus, if preset guidelines are followed carefully, the discrepancy rate is quite low. In summary, agreements were obtained with the C T /3 probe among the laboratories in 9 of 99 (9%) of the specimens tested. These 5 specimens were also analyzed with the use of J H and J* genomic DNA probes, and % agreement (5 of 5) was attained between the experienced laboratory and the central laboratory. Agreement was achieved in 7 of 9 (96%) of samples analyzed by the second outside laboratory. In one of the discrepant cases, faint bands, at the minimum level of sensitivity, were detected with both J H and J, and therefore had been incorrectly interpreted. In the second discrepant case, partial digestion produced a false reading of rearrangement. In retrospect, both cases would likely be correctly interpreted if guidelines were strictly followed. Overall, the agreement rate for J H and J, was 97 of 99, or 98%. Vol. 95 No. 3 Downloaded from on 7 July 8

6 35 HEMATOPATHOLOGY TABLE 3. FRAGMENT SIZES OF SPECIFIC DNA FRAGMENTS* Cjff J E B H E B H E Jk B H Germline Cross-hybridization bands Most common partial digest bands II E = EcoRI: B = BamHl; H = Hindlil. Expected genomic germline restriction fragment sizes are indicated for each probe and enzyme used in the study. Cross-hybridizing bands, which may be observed on occasion, are shown. Partial digestion is most frequently encountered with the Cr locus, and the commonly observed restriction fragments resulting from partial digestion with EcoRI and Hindlil are shown. * Data expressed in kilobase pairs. Reproducibility of Interpretation To test the use of the guidelines by observers with varying levels of experience, a trial was designed in which duplicate autoradiograms of 9 randomly selected samples, including both abnormal and normal diagnoses, were sent to interpreters. Interpreters included L.J.S., M.W., and observers listed in Appendix B. Interpreters received duplicate autoradiograms as described in Materials and Methods; samples were coded, and no identifying information or any clinical or phenotypic information was supplied. The interpreters included a group of people with varying levels of experience and medical training. The samples included 6 that were hybridized with C T /3, 68 hybridized with J H, and 6 with J,. Records from the outside interpreters were compared with those obtained in the central laboratory (from J.L., L.H., and J.C.). The results of this comparative trial are shown in Table for each of the interpreters. The overall agreement rate in the original trial was 95.%. However, it should be remembered that most disagreements resulted from the stringency of the guidelines, biasing the interpretation of equivocal data toward a negative result. Thus, most of the disagreements were false-negative results (6 of 86). Because the guidelines led the observer to err on the side of caution by either requesting an additional test or diagnosing an equivocal result as negative, they tended to bias this particular study toward the false negative. On the other hand, if the false-negative results are excluded, the agreement rate is 98.6% (,88 of,83) because only 5 total false-positive results were scored (Table 5). Falsepositive results were principally confined to cases of T- cell lineage (Table 5), primarily as a consequence of incomplete digestion (Fig. ). DISCUSSION In this multiinstitutional study of the application of gene rearrangement analysis to the diagnosis of lymphoproliferative processes, we have found that this approach is highly reproducible, both technically and interpretatively. By careful adherence to technique and to interpretational guidelines, these tests have shown high (95% or greater) concordance with histologic characteristics and phenotype. Significantly, the test shows uniformity of results when the same samples are tested in different laboratories (96% agreement) and in interpretation by a variety of observers (>95% agreement). These results are quite favorable when compared with comparability of histologic diagnosis or phenotypic analysis in malignant lymphoma. The reproducibility demonstrated in the current report establishes a basis to proceed with studies aimed at the clinical diagnostic utility of gene rearrangements in diagnosing and resolving difficult cases. The findings presented here underscore the highly defined, quantitative nature of restriction fragment analysis as a tool in the diagnostic laboratory. DNA probe testing has assumed an important role in the diagnosis of malignant lymphoma and leukemia with the use of probes such as those described here as well as others. The latter include chromosomal translocation probes; for example, the bcrabl (Philadelphia chromosome) of chronic myelogenous leukemia, c-myc of Burkitt's lymphoma, and bcl- of follicular lymphoma. 9 Restriction fragment tests can be Interpreter Table. REPRODUCIBILITY OF INTERPRETATION STUDY A CT D* / 3/ / 5/5 / / / /3 Q A JH D / / /3 / / /8 / /9 /3 /3 Q A Ju D / / / / / / The numbers in each column represent the total number of agreements (A) or disagreements (D) that each individual had with the central laboratory or the number of times that the result could not be determined (Q) without further testing. * Disagreements are shown as the number of false-positive readings per total number of disagreements (false-positive plus false-negative). Q A.J.CP. March 99 Downloaded from on 7 July 8

7 COSSMAN ET AL. 353 Gene Rearrangements in Lymphoma Diagnosis used in many diagnostic applications, including detection of microorganisms, diagnosis and prognosis of solid tumors, detection of constitutional genetic disease, and DNA fingerprinting in forensic pathology." All of these applications use the same essential technique of restriction fragment analysis as employed in the current study. Thus, DNA hybridization analyses promise to be a quantifiable, accurate, and reproducible addition to diagnostic pathology. Acknowledgments. The authors thank Barbara Scheffler and Philip Schein, M.D., U.S. Biosciences, Inc., Philadelphia, Pennsylvania, for statistical assistance in the study design and Timothy Triche, M.D., Children's Hospital, Los Angeles, California, for valued advice. REFERENCES. Jaffe ES, Cossman J. Immunodiagnosis of lymphoid and mononuclear phagocytic neoplasms. In: Rose NR, Friedman A, Fahey JL, eds. Manual of clinical laboratory immunology. 3rd ed. Washington, D.C.: American Society for Microbiology, 986: Cossman J, Uppenkamp M, Sundeen J, Coupland R, Raffeld M. Molecular genetics and the diagnosis of lymphoma. Arch Pathol Lab Med 988;: Knowles DM. Immunophenotypic and antigen receptor gene rearrangement analysis in T cell neoplasia. Am J Pathol 989;3: 76.. Grieser H, Tkachuk D, Reis MD, Mak T. Gene rearrangements and translocations in lymphoproliferative processes. Blood I989;73: Arnold A, Cossman J, Bakhshi A, et al. lmmunoglobulin-gene rearrangements as unique clonal markers in human lymphoid neoplasms. N Engl J Med 983;39: Geary ML, Chao J, Warnke R, et al. Immunoglobulin gene rearrangement as a diagnostic criterion of B-cell lymphoma. Proc Natl Acad Sci USA 98;8: Pittaluga S, Uppenkamp M, Cossman J. Development of T3/T cell receptor gene expression in human pre-t neoplasms. Blood 987;69: Weiss LM, Picker U, Grogan TM, et al. Absence of beta and gamma T cell receptor gene rearrangements in a subset of peripheral T cell lymphomas. Am J Pathol 988; 3: Biennerhasett GT, Furth ME, Anderson A, et al. Clinical evaluations of a DNA probe assay for Philadelphia (Ph) translocation in chronic myelogenous leukemia. Leukemia 988;: Croce CM, Tsujimoto Y, Erikson J, et al. Chromosome translocations and B cell neoplasia. Lab Invest 98;5: Landegren U, Kaiser R, Casey CT, Hood L. DNA diagnostics molecular techniques and automation. Science 988::9-37. Germline Bands APPENDIX A Restriction fragment sizes of unrearranged, or germline, bands as obtained with the germline placental controls are shown in Figure 3 and Table 3. Occasionally, cross-hybridizing bands were observed with the C T 8 and J H probes, and these are presumably a consequence of hybridization of the probe to partially homologous DNA sequences elsewhere in the human genome. Also shown in Figure is the 8.5-kb C T /3 band produced by partial EcoRl digestion and a less frequently observed C T j3 band seen with partial HindUX digestion. Both partial digestion and cross- TABLE 5. REEVALUATION FOLLOWING GUIDELINES False Result Agreement Positives C TI % (586/6).% (/6) J H 99.8% (657/658).% (/658) J k 99.% (6/66).8% (5/66) hybridization can be ruled in or out through the use of all three restriction enzymes. Rearrangements In all cases, all three restriction enzymes were used, and the diagnosis of a rearrangement required that rearrangements were seen with two of the three enzymes or two rearrangements observed with one single enzyme (two rearranged alleles). This simple rule reduces the risk of false-positive interpretation resulting from cross-hybridization or partial digestion. Genotype A case was considered of B-cell lineage if either or both of J H and J, were rearranged but C T /} remained as germline. A sample was considered of T-cell lineage if the C T /3 probe showed rearrangement while both J H and J, were germline. If both C T 3 and J H were rearranged, lineage could not be determined, but if, in addition, J, was found to be rearranged, the cell was considered to be of B lineage. Thus, any J, rearrangement was regarded as of B-cell lineage because rearrangement of this gene has only been reported rarely in T-cell neoplasms, but J H and C T /3 rearrangements may occur with regularity in both common (precursor B) and T-cell acute lymphoblastic leukemia and lymphoblastic lymphoma. Observations were recorded as follows: + = one rearrangement detected, ++ = two rearrangements detected, = no rearrangements detected, Q = not able to decide without retest, DEG = DNA degraded, and O = insufficient DNA. Reproducibility of interpretation For the analysis of reproducibility of interpretation, interpreters were provided with duplicate autoradiographs for which Kodak EDF (electrophoresis duplicating film; Eastman Kodak Corporation, Rochester, NY) was used. This duplicating process required approximately -3 seconds of exposure to fluorescent light per copy. The exposure was based on the darkness of the individual bands and on the background darkness. Each exposure was adjusted to mimic the original autoradiograph as closely as possible, with particular emphasis on exposure of the lightest bands. If observers encountered difficulty in reading a film, they were permitted to record it as "decision cannot be made without further information." No other information was provided to the observers, such as histologic findings, percentage of abnormal cells in the specimen, complete blood counts, clinical history, prior diagnosis, or phenotyping studies. Decision making was based on the guidelines presented here. Vol. 95 Downloaded from on 7 July 8

8 35 HEMATOPATHOLOGY CT/3 JH 'Kappa 3.Kb 9.Kb 6.6Kb - 8Kb Kb Kb 9.Kb 5.Kb.Kb 3.3Kb.3Kb -.Kb - I.Kb FIG. 3. Germline restriction sizes and sensitivity using the three probes of this study. Ten-microgram genomic germline (placental) DNA was cut with EcoRl (lanes ), BamBl (lanes 3), and HindXW (lanes ). Each lane 5 contains a combination of.5 Mg of placental DNA cut with the three enzymes EcoRl, BamHl, and HindlM. Labeled markers (X phage////wiii) are indicated in lanes. and the sizes of the restriction fragments are indicated to the right of each panel. APPENDIX B Interpreters The interpreters were as follows: Ronald Agarwal, Ph.D., Washington, D.C. Ruth Bowser-Finn, Madison, Wisconsin Beryn Frank, Miami Beach, Florida Margaret Gordon, Ph.D., Houston, Texas Kathy Koscioll, Hartford, Connecticut Gerald Marti, M.D., Bethesda, Maryland Richard McPherson, M.D., Washington, D.C. Steven Melnick, M.D., Miami Beach, Florida Karen Schuehler, Syracuse, New York A.J.C.P. March 99 Downloaded from on 7 July 8