Clonal Analysis of Untreated Non-Hodgkin's Lymphoma Utilizing Immunoglobulin Gene Rearrangement and Immunophenotype1

[CANCER RESEARCH 48, 6272-6277, November 15, 1988] Clonal Analysis of Untreated Non-Hodgkin's Lymphoma Utilizing Immunoglobulin Gene Rearrangement and Immunophenotype1

Richard A. Rudders,2 Satinder Dhillon, and Theodore G. Krontiris

Hematology/Oncology Division, Department of Medicine, and Department of Molecular Biology and Microbiology, Tufts University School of Medicine, and the New England Medical Center Hospitals, Boston, Massachusetts 02111

ABSTRACT is entirely absent. The recent development of molecular probes for genetic events associated with B-cell differentiation provides We have prospectively examined 66 consecutive initial diagnostic a potentially powerful approach for the study of lymphoma- lymph node biopsies from unselected patients suspected of having malig genesis. From the work of many investigators, it has been nant lymphoma for clonal immunophenotypic and immunogenotypic markers. By morphological and cell surface phenotypic criteria 52 had possible to determine the structure of the germ line Ig genes non-Hodgkin's lymphoma derived from the B-cell lineage and in these and construct a model of Ig gene rearrangement that is associ we compared surface immunoglobulin criteria for clonality with Â¡iniiiu- ated with the maturation of B cells (1-3). Proceeding from this noglobulin gene rearrangement as detected by .In, ( ., and ( ...gene probes. scheme, several groups have shown that neoplasms of human We found that the addition of Bgl\l and ///mill I double digests to the B-cells such as leukemia and lymphoma regularly exhibit re standard /(umili and EcoRl restriction enzymes made it possible to arrangements of Ig genes that can serve as a diagnostic criterion detect rearrangement in the vast majority of lymphomas and that re for malignancy (4-8). arrangement of both JH alÃelesis the rule. A rearranged heavy and/or In addition to their diagnostic utility, immunoglobulin gene light chain gene was detected in 47 of 52 (91%) tumors and the JH probe probes can be used to study lymphoma clonal evolution. Each alone detected rearrangements in 87% of tumors when multiple restriction individual tumor clone displays a characteristic Ig gene re enzymes were used. In contrast, surface immunoglobulin, the standard clonal marker for monoclonal B-cell malignancy, was either undetectable arrangement. The finding of distinct Ig gene rearrangements in or did not exhibit light chain restriction in 29 of 52 tumors as detected different populations of tumor cells (purified from the same by flow cytometric analysis. Further, in 24 of these 29 tumor DNAs we lymphoma on the basis of idiotype expression) has provided could detect an Ig gene rearrangement. In follicular (nodular) lymphoma additional evidence that lymphomas may have oligoclonal, which often gives ambiguous immunophenotypic results by cell suspen rather than monoclonal, origins (9-12). It should be noted, sion techniques, monoclonal gene rearrangements were detected in 16 of however, that this observation has been tempered by the finding 18 tumor DNAs. Monoclonal surface immunoglobulin was detected in that somatic mutation may alter idiotype expression within a only 8 of 18 of this subset of cases. monoclonal tumor and thus produce the artifactual appearance The 52 tumors were also analyzed for potential oligoclonality. We of oligoclonality (13). found that the use of Bgfl\, a restriction enzyme that closely spanned the The nature of Ig gene rearrangement studies in human lym JH region, increased the sensitivity of detecting rearrangements and facilitated the identification of isotype switch variants. In only a single phomas has, however, left several important issues unresolved (1 of 52) tumor DNA were more than two rearranged bands seen with since many have been carried out in selected and often previ JH, {',, and ( ', probes, suggesting a multiclonal origin. Additional cases ously treated patients. In this study we have examined an thought to potentially represent oligoclonality by immunophenotypic unselected series of initial lymphoma biopsies. We report herein criteria proved to be isotype switch variants. an analysis of the clonal composition of these tumors based on We conclude that Ig gene rearrangement is an extremely sensitive Ig gene rearrangement as well as on cell surface criteria. method for defining monoclonality in lymphoma cell populations, partic ularly if multiple restriction enzymes are used, and that the vast majority of the clonal diversity seen in initial diagnostic biopsies is the result of MATERIALS AND METHODS isotypic switch within a given clone rather than true oligoclonality. Tumor Samples. The initial tissue biopsies from 66 consecutive, previously untreated patients suspected of having lymphoma were pro INTRODUCTION spectively entered in the study. Biopsies of lymph nodes, tumor masses, Current concepts are consistent with the view that human or both were received fresh in the laboratory in sterile, cold saline after malignant lymphomas are expansions of single clones ot lymph- a portion was removed for routine histopathological examination, and were processed immediately. A portion was snap frozen in liquid N2 oid cells that have undergone neoplastic transformation. The for DNA studies and the remainder was converted to a single cell major immunophenotypic criterion for defining malignancy in suspension for immunophenotypic analysis. When multiple biopsy these tumors has been the expression by B-cell clones of either specimens were submitted (as was the case when exploratory laparot surface or secretory Ig with a restricted light chain type. In omies were performed), portions of each were processed and analyzed practical usage this definition has its limitations. For instance, separately. normal cell populations frequently dilute out tumor cells, ob Immunophenotypic Analysis. Cells were examined for surface anti scuring the malignant clone(s). Also, Ig is often either not gens with a panel of monoclonal and polyclonal antibodies by using clearly expressed on the cell surface of B-cell-derived tumors or standard immunofluorescent techniques as previously described (14). Briefly, cell suspensions were incubated with monoclonal or polyclonal Received 4/11/88; revised 8/11/88; accepted 8/16/88 antibodies directed against a, /.Â¡.-,.& heavy and light Ig chains, Bl The costs of publication of this article were defrayed in part by the payment (CD20), B4 (CD19), la, T3 (CD3), T4 (CD4), T8 (CDS), TÃ•O(CD38), of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Til, Leu 1 (CDS), IL2R (CD25), and CALLA (CD10). Polyclonal 1Presented in part at the Annual Meetings of the American Society of antibodies used were affinity purified F(Ab')2 fragments. All antibodies Hematology, New Orleans. Louisiana, December 1985 and December 1986 in were regularly titrated against known positive controls. Stained cell San Francisco, California. Supported in part by Grants CA40725 and CA09429 populations were analyzed by flow cytometry on a Becton Dickinson from NIH. 2To whom requests for reprints should be addressed, at Division of Oncology/ (Sunnyvale, Ã‡A)FACS analyzer. Hematology, New England Medical Center Hospitals, 750 Washington Street, DNA Probes. DNA clones representing the human JH (15), C. (16) Boston, MA 02111. and CA(17) genes were generously provided by Dr. Philip Leder, Boston, 6272

MA. The JH probe that was used in our study consisted of a 2.5-kilobase were detected with BamHl and EcoRl that were not detected Sau3A fragment of JH. with Bgl\\. In contrast, in 11 of 22 DNAs the Bglll digest DNA Purification and Analysis. Tissue that had been snap-frozen in revealed an increased number of rearranged bands (Fig. 1). This liquid nitrogen was pulverized over dry ice. Cells in the resulting powder represents an almost 4-fold difference in sensitivity. Nine of were lysed in a solution containing 0.5% SDS, 0.15 M NaCI, 10 mM these 11 cases also were digested with BamH'md and/or Tris-HCI, pH 7.5, and 100 HIMEDTA. Proteinase K was added to 200 /jg/ml and the mixture was incubated at 65"C for 30 min. Following Â£coHind. An increase in the number of rearranged bands seen two extractions with phenol. phenol:chloroform:isoamyl alcohol in 1 of the 9 cases. We were also able to compare directly the results of Bglll, (25:24:1), and finally. CHCl^isoamyl alcohol, the DNA was precipi BamH'md, and/or EcoH'md digests with our JM probe in 26 tated in two volumes of 95% ethanol. The DNA was redissolved, incubated sequentially with RNase A (100 Â¿ig/ml)and proteinase K, cases where BamH\ and EcoRl digests displayed a single rear extracted again with phenol and chloroform/isoamyl alcohol, and re- ranged band or none at all. In 5 of 26 cases Bgl\\ digests precipitated in two volumes of ethanol. The pellet was washed in 70% revealed more bands than double digests and in 7 of 26 cases ethanol and dissolved in 10 mM Tris-HCI, pH 7.5. and 10 mM EDTA. double digests revealed additional bands not seen with Bglll Five Â¿jgof DNA were digested with BamHl, EcoRl, Bglll, and (Fig. 2). In the majority of cases (14 of 26) no new information /tomHI///fndIII, and Â£foRl////ndIII restriction endonuclease accord ing to the manufacturer's (New England Biolabs, Beverly, MA) speci was obtained. This suggests that the larger increment in sensi fications and electrophoresed through 0.7% agarose gels. Southern tivity is obtained by adding Bgl\\ or double digests to standard transfer (18) was to GeneScreen Plus filters (New England Nuclear, BamHl and EcoRl but that doing both will still yield an Boston. MA). Probe was nick translated (19) to a specific activity of additional increment of undetected bands. greater than 10' cpm/Mg DNA and hybridized to filter-bound DNA at Rearranged Ig Genes. Rearrangement of Ig genes were clearly 65Â°Covernight in a solution containing 5x 0.15 M NaCI, 10 mM Tris- demonstrated in the non-Hodgkin's group but were uniformly HCI, pH 7.5.10 mM EDTA; 0.5% sodium dodecyl sulfate; 10% dextran absent in the 14 control group patients. Rearrangements were sulfate; Ix Denhardt's; and 500 Mg/ml of heparin sulfate (Sigma observed in all histological subcategories of lymphoma (Table Chemical Co., St. Louis, MO). Washes were to O.lx 0.15 M NaCI. 10 mM Tris-HCI, pH 7.5, 10 mM EDTA at 60-65Â°C. 1; Fig. 3). We were able to detect a rearrangement of a heavy and/or light chain gene in 47 of 52 (91%) of our tumors. Only HistolÃ³gica!and Immunophenotypic Classification. Of the initial 66 biopsies examined, 52 were B-cell-derived non-Hodgkin's lymphomas 5 cases (all SIg negative) failed to show a rearrangement. by routine histopathological and immunophenotypic criteria. The re However, in 2 of the 5 a limited amount of DNA was present maining 14 consisted of cases of Hodgkin's disease, T-cell lymphoma, and analysis with more than 2 restriction enzymes was not benign hyperplasia. and a single case of acute myelogenous leukemia. possible. Therefore, this subset (9%) represents a maximum The histological subclassification of the 53 B-derived cases is shown in number. In the remaining 3 tumors. Ig gene deletion, lack of Table I. Cases were classified according to the New Working Formu gene rearrangement, or failure of Southern blotting to detect a lation system (20) by Dr. Ronald A. De Lellis, Department of Pathol rearrangement that was present, are all possible explanations. ogy, Tufts University School of Medicine. Utilizing the Working For With the JM probe alone we detected rearrangements in 45 of mulation, 28 of 52 cases were low grade tumors. The remainder were intermediate and high grade tumors. 1 234567

RESULTS

Increased Sensitivity and Resolution for JM Rearrangements. When a JH probe was used to detect immunoglobulin gene rearrangements in lymphoma tissue DNA digested with the standard BamHl or EcoRl, we sometimes noted the appearance of faint bands within 1-2 kilobases of the germ line pattern which we could not unambiguously define as JM re arrangements, or we could detect rearrangements of fewer than both JH alÃeles.In a number of these cases clonality was sug gested by immunophenotyping. In order to increase the sensi tivity of detecting rearrangements we made use of double digests by adding Hindlll digestion to the standard BamHl and EcoRl, thereby producing smaller fragments. In addition, to increase the resolution of the JH probe, we searched for a single restric tion enzyme which could more closely span the J genes and avoid the polymorphic region we have recently described that lies upstream of the J region (21 ). From sequence data (15) this proved to be Bglll. This enzyme produced a 3.75-kiIobase germ line band in which Bglll sites were approximately 300 base pairs 5' to Jl and 1 kilobase 3' to J6 and rearrangements involving any of the six J genes were consequently detected. Fig. 1. Southern blot of DNA from 3 patients with lymphoma probed with Because the germ line fragment size is smaller, the potential the JH probe. Germline bands are indicated by lines and rearranged bands by resolution of rearranged bands is improved, particularly if nylon arrows. Lane I contains molecular weight standards. Lanes 2 and 3 are EcoRt and Bgll\ digests, respectively, from Patient A. illustrating the failure to detect filters are used to transfer fragments during Southern blotting. rearrangements with EcoRl, yet Bgl\\ digestion reveals 2 rearrangements. Case B We have data for direct comparison in a subset of our cases illustrates a similar point: the EroRI digest (Lane 4) reveals no rearranged bands with respect to sensitivity of the method. In 22 cases DNA was but the //v/ll digest in Lane 5 reveals 1. Lanes 6 and 7 from Patient C indicate the potential for improved resolution of the 2 rearranged bands by /f.v/ll (Lane 7) digested with standard BamHl, EcoRl, and Bglll enzymes and when compared to the EcoR\ digest in Lane 6, where one rearranged band lies probed with the JH probe. In 3 of 22 DNAs, rearrangements close to the germ line signal, giving the appearance of a doublet. 6273

Fig. 2. Southern blot of DNA from a single patient with lymphoma. Â£coRI and BamHl failed to reveal rearranged bands with the JH probe (not shown). Subsequent /.,.)< I ///mil 11(Lane 1) and BamH\/Hind\\\ (Lane 2) double digests displayed a single rearranged band positioned close to the germ line band. The BgH\ digest (Lane 3) reveals a single rearranged band that is clearly separable from the germ line signal. Lane 4 contains molecular weight standards. Fig. 3. DNA isolated from 4 patients in the study digested with Bglll and probed with JH. Lane I contains molecular weight standards. Lanes 2, 3, 4, and 7 contain rearranged bands as indicated by arrows. Lanes 5 and 6 contain DNA Table 1 Immunoglobulin gene rearrangement patterns in non-Hodgkin's from 2 control patients with benign reactive nodal hyperplasia. lymphoma Patterns of immunoglobulin gene rearrangement arranged according to H = heavy and L = light chain gene rearrangements. For example H*L* = re Table 2 Comparison of immunophenotypic and genotypic criteria of monoclonality arrangement of both heavy and light chain genes. Numbers in parenthesis, number of SIg-negative or SIg-indeterminate cases in each category' of gene re arrangement. SL, small lymphocytic; FSC, follicular small cleaved cell; DSC, of immunoglob diffuse small cleaved: DSL, diffuse small and large; DLC, diffuse large cleaved; Histological no. ulin genes LCI, large cell immunoblastic; SNC, small noncleaved. categorySL of cases10 clonal2 minate*2 (H +L)"9 rearrangementH*L*6(4)13(5)4(0)08(4)2(1)KD34(15)H+L-3(3)3(3)2(1)KD2(1)0011(9)H-L*00001(0)1(0)02(0)H-L-igene Histological of FSC 186 8 2 16 subcategoryLow casesIO1862123152Immunoglobulin DSC 5 12 0 6 gradeSLFSCIntermediate DSL 2 0 03 111 (i)2(2)0KDKD005(5)DLC 12 6 31 LCI 3 2 00Rearrangements31 gradeDSCDSLDLCHigh SNCTotal 1Mono 0SIgAbsent681Indeter

Totals 52 23 22 47 Â°H,heavy chain; L, light chain; SL, small lymphocytic; FSC, follicular small gradeLCISNCTotaisNo. cleaved cell; DSC, diffuse small cleaved cell; DSL, diffuse small and large; DLC, diffuse large cleaved; LCI, large cell immunoblastic; SNC, small noncleaved. b Expression of both â€¢¿andA chains without clearcut distortion of normal Â«/X

displayed fewer rearrangements than the SIg* group, but in 52 (87%) cases. This degree of sensitivity was seen in all only 4 cases no rearranged bands were detected with any probe histological types of lymphoma. It should be noted that 2 cases used. displayed a light chain rearrangement without a corresponding Genotypic versus Phenotypic Criteria for Monoclonality. Com heavy chain rearrangement. This anomalous result has several pared to phenotypic criteria, the sensitivity of gene re possible causes. The most likely possibility is that H chain arrangement in defining clonal malignancy far exceeded that of rearrangements were present but not detected since in both, not SIg as detected by flow cytometric techniques (Table 2). Twenty-three of 52 cases (43%) displayed monoclonal surface enough DNA was available to probe with more than 2 restric Ig while the remaining 29 cases of B-derived lymphoma either tion enzymes. It is also possible that an H chain gene alÃelewas deleted. We cannot answer this question definitively in this lacked SIg or restricted light chain expression. In 24 of these 29 tumor DNAs, there was at least one clearly detected gene small subset of patients. rearrangement. In the 23 SIg+ cases, gene rearrangements were It is of considerable interest that Ig gene rearrangements were detected in a broad histological range of lymphomas that did seen in all 23. In follicular (nodular) lymphoma, which often not simultaneously express detectable quantities of SIg.3 There gives ambiguous immunophenotypic data by cell suspension were a total of 22 tumors that lacked SIg, yet 18 expressed H techniques, gene rearrangements were detected in 16 of 18 and/or L chain rearrangements. As a group, the SIg" cases DNAs, compared to 8 of 18 displaying monoclonal surface immunoglobulin. In the 2 tumors where gene rearrangements 1The abbreviation used is: SIg, surface immunoglobulin. were not found, SIg was not detectable, either. 6274

There was also excellent agreement between the light chains accompanying isotype switch. Therefore, rearranged bands ap expressed on the monoclonal SIg and the light chain gene pearing with Bglll, unlike those produced with EcoRl or rearrangements we detected. In 20 cases where there was a BamHl, can only represent rearrangements on the 5' side of monoclonal surface light chain expressed and a monoclonal JH. One band can never be a 3' deletion or isotype switch light chain gene rearrangement detected, there was concordance variant of another band. Thus, by using Bgt\\ in addition to in 19. A monoclonal gene rearrangement was detected in 5 of BamHl and EcoRl restriction enzymes, we were able to deter 7 cases where the surface expression of light chains was inde mine an upper limit for the number of potentially oligoclonal terminate. tumors. B-Cell Differentiation. As B-cells differentiate a hierarchy of Using as a criterion for oligoclonality the high level of expres gene rearrangement occurs. H chain alÃelesrearrange first, sion of more than one cell surface immunoglobulin heavy chain followed by L chain rearrangement; K preceding X in temporal or both Kand X light chains, we identified 6 potentially oligo sequence (3). For the H chain, rearrangement ceases when an clonal tumors (Table 4). All were low-grade lymphomas; five alÃelerearranges productively. Theoretically, the maximum were follicular (nodular) lymphomas. Five of six cases (Cases number of rearrangements per clone is 2. Since it is possible 1-5) expressed 2 surface heavy chains (IgM plus IgG or IgM for the first allelic rearrangement to be productive, a clone plus IgA), but only a single predominant light chain. Although might display a single H chain rearrangement. We examined it is possible that these tumors represent oligoclonal cases, each Southern blot of tumor DNA for the number of clearly previous studies have shown that many of these double-heavy- defined rearranged bands (Table 3). In 34 of 47 (72%) cases, 2 chain, single-light-chain tumors are isotype switch variants or more distinct rearrangements were found with the JH probe. within a single clone (22-24). The gene rearrangement patterns A small minority of cases either lacked a detectable gene re in Patients 1-5 were not striking (Table 4). Two rearranged JH arrangement [3 of 47 (8%)] or had a single rearrangement [10 alÃeleswere present in four of the five. All five patients had a of 47 (20%)]. As stated previously, a number of these cases light chain gene rearrangement that corresponded to the SIg were not evaluated with more than 2 enzymes because of light chain. insufficient DNA. Taking this into account, it would appear By immunophenotypic data Patient 6 was a good candidate that the majority of human tumors exhibit rearrangement of for an oligoclonal tumor since 2 different surface heavy and both H chain alÃeles.However, since several of the tumors light chains were expressed by the cell population. Thus, 2 displaying a single H chain rearrangement also lacked an L clones, one expressing Klight chains and the other Xlight chains chain rearrangement and did not express SIg, it is possible that might have been present. Alternatively, a single clone could certain of these cases represent DJ, rather than VDJ re have been contaminated with normal background B-cells, which arrangement. contributed multiple Ig chains masking the monotypic immu Oligoclonality and Genetic Diversity. In this investigation we noglobulin. Genotypic analysis supported the latter interpreta also assessed the potential for clonal heterogeneity in lympho- tion since 2 JH rearrangements were found but there was only mas of all histological types. A tumor composed of multiple, a single light chain rearrangement. Review of the histopathol- unrelated clones can be suspected from immunophenotypic data ogy in this case also supported this interpretation since a sizable when there is nearly equal expression of two different heavy portion of the biopsy material consisted of hyperplastic (reac and/or light chains. In general, oligoclonality can be suspected tive) lymph node, in addition to frank lymphoma. Thus, in the from gene rearrangement data when two or more re 6 tumors where SIg criteria suggested the possibility of oligo arrangements are detected by a single probe. However, this type clonality, if the genotypic findings and histopathology are con of genetic analysis has been confounded by the knowledge that sidered, they appear to represent isotypic switch variants and single clones may display rearrangements of both alÃelesor may an artifact created by contamination by nonneoplastic B-cells. undergo continuing rearrangement (isotype switches and/or More than 2 rearranged bands per probe were clearly present simple deletions 3' to the JH region) (10). Thus, it has usually in only one of our tumors (Table 4, Patient 7). This tumor not been possible to distinguish definitively by Southern blot expressed surface K light chains in a high percentage of cells methodology, a tumor composed of a single lymphoid clone and there was no suspicion of oligoclonality by immunophen- with multiple rearrangements (clonal evolution) from a tumor otype. However, Southern blotting with multiple enzymes in possessing multiple distinct clones (10, 11, 13). On the other cluding Bglll revealed three distinct rearrangements with the hand, since the Bglll enzyme cuts close to the 3' end of the JH JH probe and three with the C< probe (Fig. 4). As discussed region, it is incapable of detecting the usual sorts of deletions above, Bglll bands cannot be 3' deletional variants of one another. Therefore, this tumor contained three distinct JH re- Table 3 Maximum number of rearranged heavy chain region bands with the JH probe using 2 or more restriction enzymes" Table4 Potentialoligoclonalitybyimmunophenotypicand rearrangements(JH)13 immunogenotypiccriteria Histological no. categorySL* of cases9 type"Histology Patient no.1 CxDSC* Immunophenotype JH C. FSC 16 1 3 11 1 DSC 6 0 1 50 0 OO2OOO22>222>â€¢2 DSL 112 0 1 0 2 FSC 2 DLC 2 2 9 0 3 FSC 2 1g LCI 2 0 1 11>20 0 4 FSC 2 SNCTotal 1No.00 0of 026 0 5 FSC 1 g 6 FSC 2 gÂ» Totals 47 10 33 1 7Immunogeno-FSC SNC, Composite 3gg g " Each DNA was digested with one or more of the following enzymes: BgÃ±l, Lymphoma" BamH\ plus Hindin or EcoRl plus Hindlll. " Numbers in columns represent the number of distinct rearranged bands * SL, small lymphocytic; FSC, follicular small cleaved cell; DSC, diffuse small detected on Southern blotting with JH and light chain probes. cleaved cell; DSL, diffuse small and large; DLC, diffuse large cleaved; LCI, large * DSC, diffuse small cleaved cell; FSC, follicular small cleaved cell; SNC, cell immunoblastic; SNC, small noncleaved. small noncleaved cell; G = 7 chain; M =

123456 lymphoid neoplasm and to what extent clonal heterogeneity 12345 existed in these tumors at the time of diagnosis. We chose to examine the initial diagnostic lymph node biopsy from consec utive, unselected cases in order to avoid case selection and the influence that interventions such as treatment would have on the genotype and clonal composition of these tumors. The avoidance of selected and previously treated cases is particularly relevant to the question of tumor heterogeneity and is central to constructing a model of human lymphoma that is useful in defining immunotherapeutic treatment strategies, particularly relating to monoclonal antibody and antiidiotype therapy. A technical modification that proved to be important in this study was the addition of the restriction endonuclease, Bgltt, - to detect rearrangements in the J region of the heavy chain gene and to exclude 3' deletions which can obscure the identification of truly distinct VDJ and DJ rearrangements (21). In those tumors that were digested with Bgl\\ in addition to the Fig. 4. A, DNA from Patient 7 (Table 4) digested with BgH\ and probed with standard BamHl and EcoRl there was almost a 4-fold increase .In Three distinct rearranged bands plus a germ line band are seen in Lane 1. in the number of rearrangements detected. Although a similar Lanes 2 through 5 contain DNA from 4 patients selected at random from our study who demonstrate 1 (Lanes 2 and 3), 2 (Lane 4), or 0 (Lane 5) rearranged quantitative increase can be obtained by using double digests bands. Lane 6 contains molecular weight standards. I!. Southern blot of DNA with ///ndlll, the Bgt\\ enzyme has the additional advantage of digested with BamHl and probed with C.. Lane 1 contains molecular weight being very useful in distinguishing true multiclonality from standards. DNA from Patient 7 (Table 4) is in Lane 5. Three distinct bands and isotype switch variants since it cuts close to the 3' end of the a germ line band are seen. For comparison. Lane 2 is from a study patient with an SIg* Â«phenotype who demonstrates 2 rearrangements. Lanes 3 and 4 contain JH region (22-24). DNA from tumors with an SIg* Xphenotype that contain no rearranged bands. Several previous studies have indicated that Ig gene re arrangement can be used as a criterion for neoplasia in B- arrangements. It was highly suggestive of oligoclonality that derived tumors (4-8). Estimates have been made suggesting three accompanying Krearrangements were present as well. The that as little as 1-2.5% of clonal tumor tissue in a given histopathological findings of a composite (follicular small specimen can be detected by Southern blotting (5). In this study cleaved cell plus small noncleaved cell) lymphoma further sup we have directly compared the sensitivity of clonal SIg deter ports this interpretation. mination by flow cytometry with the detection of clonal gene In 2 cases, two separate tissue specimens from different rearrangement by Southern blotting. Our data clearly indicate anatomical sites were obtained at the time of initial biopsy. We that Ig gene rearrangement is a more broadly applicable crite have analyzed the 2 specimens separately in these instances. In rion since gene rearrangements were seen in each of 23 of our the first case the main tumor in the abdomen was biopsied as monoclonal SIg"1"cases, in 24 of 29 cases where SIg was not well as a large groin node. Morphologically, the former had the useful (ambiguous or negative results), and in a higher propor appearance of lymphoma, the latter was interpreted as reactive tion of follicular lymphomas. In the study we did not directly with atypical features. Phenotypically, the abdominal tumor compare genotypic analysis to lymphoma Ig phenotyping as cells lacked SIg. By Southern blotting two clearly defined determined by tissue section methods. It is well known that rearrangements were seen with the JH probe on H.i;l\\ and /fi/w/111.'///mil 11double digests. The groin mass had polyclonal lymphoma diagnosis may be enhanced in certain cases by combining morphology with immunohistological stains (25). B-cell phenotypic characteristics and displayed no rearranged We can only conclude, therefore, that the detection of gene bands with either heavy (JH) or light chain probes. rearrangement is more sensitive than flow cytometric SIg phen In the second case, both biopsies were lymphoma by morpho otyping in establishing clonality in B-cell lymphoma. logical criteria and both lacked SIg. One specimen displayed Both heavy and light chain gene rearrangements were found only germ line DNA with JH probe of a BamHl and EcoRl in all major histolÃ³gica! types of lymphoma, suggesting that digest. The Bgl\\ digest, however, displayed 2 rearrangements. even the high-grade lymphomas (by morphological criteria) are The second specimen displayed the identical rearrangements derived from cells that have matured beyond the pre-B-cell seen in the first biopsy in the Bglll digest. We interpret these stage. We have no firm evidence to suggest that a significant findings to indicate that neither of these 2 cases of multiple proportion of lymphomas derive from pre-B-cells or earlier biopsies could be viewed as displaying oligoclonality based on stages where one might expect to find only DJ rearrangements. the phenotypic and genotypic characteristics. Perhaps, as noted above, the 9 cases that demonstrated the In summary, of 52 unselected cases of lymphoma, 7 (13%) H+(J+)L~, SIg" pattern belong in this category. With respect to were candidates for oligoclonality by either immunophenotypic B-cell differentiation it is noteworthy that 72% of our tumors or molecular criteria. Of these. Patients 1-6 (Table 4) proved displayed rearrangement of both JH alÃelesif samples were unlikely when all data were taken into account. One tumor of analyzed by DNA digestion with multiple restriction enzymes. 52 (2%) remained a strong possibility. This percentage is considerably higher than in several previous published reports (6, 26). For instance, in the report of Williams DISCUSSION et al. (26) there was rearrangement of 2 JH alÃelesdetected in 7 of 23 (30%) tumors and 2 additional tumors displayed a single In this study we have examined several fundamental issues rearrangement and deletion of the second alÃele.The higher relating to the biology of B-cell-derived lymphomas. We were percentage reported in our study is parallel to reports in murine specifically interested in whether newer genetic probes offered systems where the vast majority of B-cells rearrange both H advantages over traditional immunophenotyping in defining a chain alÃeles(27-29). This suggests that the process of H chain 6276

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1988 American Association for Cancer Research. CLONAL ANALYSIS OF UNTREATED NON-HODGKIN'S LYMPHOMA rearrangement in human B-cell tumors is relatively inefficient, 5. Cleary, M. L., Chao, J., Warnke, R., and Sklar, J. Immunoglobulin gene rearrangement as a diagnostic criterion of B cell lymphoma. Proc. Nati. requiring 2 attempts before rearrangement is productive. Since Acad. Sci. USA, 81: 593-597, 1984. these are tumors and not normal B-cells one should extrapolate 6. Arnold, A., Cossman, J., Bakhshi, A., Jaffe, E. S., Waldmann, T. A., and with caution, but our data support the concept that normal B- Korsmeyer, S. J. Immunoglobulin gene rearrangements as unique clonal markers in human lymphoid neoplasms. N. Engl. J. Med., 309: 1593-1599, cells usually require rearrangement of both H chain alÃeles 1983. before differentiation can proceed. 7. Korsmeyer, S. J., Greene, W. C, Cossman, J., Hsu, S. M., Jensen, J. P., Prior to this report an estimate of the incidence of oligoclon- Neckers, L. M., Marshall, S. L., Bakhshi, A., Depper, T. M., Leonard, W. J.. Jaffe, E. S., and Waldman, T. A. Rearrangement and expression of ality in a randomly selected series of untreated lymphoma cases immunoglobulin genes and expression of Tac antigen in hairy cell leukemia. Proc. Nati. Acad. Sci. USA, 80:4522-4526, 1983. has not been made. In previous smaller studies, patients had 8. Cleary, M. L., Warnke, R., and Sklar, J. Monoclonality of lymphoprolifera- apparently received prior treatment, had transformed to a more tive lesions in cardiac transplant patients. N. Engl. J. Med., 310: 477-482, aggressive histology, and/or were preselected as potentially 1984. 9. Sklar, J., Cleary, M. L., Thielman, K., Gralow, J., Warnke, R., and Levy, R. oligoclonal based on multiple distinct histologies in serial biop Bidonai B cell lymphoma. N. Engl. J. Med., 311: 20-27, 1984. sies or highly suggestive immunophenotypic criteria (9, 10). In 10. Siegelman, M. H., Cleary, M. L., Warnke, R., and Sklar, J. Frequent biclonality and Ig gene alterations among B cell lymphomas that show the large series we report here we have addressed the issue of multiple histologie forms. J. Exp. Med., 161: 850-863, 1985. potential oligoclonality in initial diagnostic biopsies. Our data 11. Carroll, W. L., Lowder, J. N., Streiffer, R., Warnke, R., Levy, S., and Levy, suggest that, by SIg immunophenotypic criteria, the greatest R. Idiotype variant cell populations in patients with B cell lymphoma. J. Exp. Med., Â¡64:1566-1580, 1986. potential for diversity was seen in the low grade lymphomas, 12. Hatzubai, A., Maloney, D. G., and Levy, R. The use of monoclonal anti- particularly in the follicular (nodular) group. At most, however, idiotype antibody to study the biology of a human B cell lymphoma. J. we observed only 7 of 52 (13%) where oligoclonality was Immunol., 126: 2397-2402, 1981. 13. Cleary, M. L., Meeker, T. C., Levy, S., Lee, E., Trela, M., Sklar, J., and suggested, including several cases where multiple biopsies from Levy, R. Clustering of extensive somatic mutations in the variable region of different anatomic sites were analyzed separately. In only 1 of an immunoglobulin heavy chain gene from a human B cell lymphoma. Cell, Â«.â€¢97-106,1986. the 7 cases was there potential support for a bidonai origin 14. Rudders, R. A., and Poldre, P. A. The B cell IgM Fc receptor: Further based on the presence of more than 2 rearranged J region evidence for the B cell origin of "null" chronic lymphocytic leukemia (CLL). bands. In the remaining cases the presence of 1 or 2 H chain Blood, 64:375-379, 1984. 15. Ravetch, J. V., Siebenlist, U., Korsmeyer, S., Waldmann, T., and Leder, P. rearrangements, a single light chain gene rearrangement and Structure of the human immunoglobulin /lymphocyte differentiations: a clonal transition between IgM and IgG synthesis in leu In summary, our data suggest a tentative model of lympho- kemic B lymphocytes. J. Immunol., 119: 283-286, 1977. magenesis where a single clone expands and dominates. Increas 23. Rudders, R. A., and Ross, R. Partial characterization of the shift from IgG to IgA synthesis in the clonal differentiation of leukemic bone marrow- ing tumor heterogeneity may occur with time, and if so, prob derived lymphocytes. J. Exp. Med., 142: 549-559, 1975. ably results from the combined effects of spontaneous mutation 24. Todel, C. W., Franklin, E. C.. and Rudders, R. A. Similarities of the light (that occurs in high frequency in normal B-cells) and the mu- chains and the variable regions of the heavy chains of the IgG2 and IgAi myeloma proteins from a single patient. J. Immunol., 112:871-876. tational and selection pressures of various treatments (30). 25. Picker, L. J., Weiss, L. M., Medeiros, L. J., Wood, G. S., and Warnke, R. A. Immunophenotypic criteria for the diagnosis of non-Hodgkin's lym phoma. Am. J. Pathol., 128:181-201, 1987. REFERENCES 26. Williams, M. E., InnÃ©s,D.J., Borowitz. J., Lovell, M. A., Swerdlow, S. H., Hurtubise, P. E., Byrnes, R. K., Chan, W. C., Byrne, G. E., Jr., Whitcomb, C. C., and Thomas, C. Y., IV Immunoglobulin and T cell receptor for gene 1. Leder, P., Max, E. E., and Seidman, J. G. The organization of immunoglob- rearrangements in human lymphoma and leukemia. Blood, 69:79-86, 1987. ulin genes and the origin of their diversity. In: M. Fougereau and J. Daussett 27. Early, P., and Hood, L. AUdir exclusion and non-productive immunoglobulin (eds.), Immunology 1980, pp. 35-50. New York: Academic Press, 1980. gene rearrangements. Cell, 24:1-3, 1981. 2. Tonegawa, S. Somatic generation of antibody diversity. Nature (Lond.), 302: 28. Alt, F., Rosenberg, N., Lewis, S., Thomas, E., and Baltimore, D. Organiza 575-581, 1983. tion and reorganization of immunoglobulin genes in AM Â»IVtransformed 3. Korsmeyer. S. J., Mieter, P. A., Ravetch, J. V., Poplack, D. G., Waldmann. cells: Rearrangement of heavy but not light chain genes. Cell, 27: 381-390, T. A., and Leder. P. Developmental hierarchy of immunoglobulin gene 1981. rearrangements in human leukemic pre-B-cells. Proc. Nati. Acad. Sci. USA, 29. Colecalugh. C., Perry, R. P., Karjalainen, K.. and Weigert, M. Aberrant 78: 7096-7100, 1981. rearrangements contribute significantly to the allelic exclusion of immuno 4. Korsmeyer, S. J., Arnold, A., Bakhshi, A., Ravetch, J. V., Siebenlist. V., globulin gene expression. Nature (Lond.). 290: 372-376, 1981. Mieter, P. A., Sharrow, S. O., LeBien, J. T. W., Kersey, J. H., Poplack, D. 30. Glennie. M. J., McBride, H. M., Stripe, F., Thorpe, P. E., Worth, A. T., and G., Leder, P., and Waldmann, T. A. Immunoglobulin gene rearrangement Stevenson, G. T. Emergence of immunoglobulin variants following treatment and cell surface antigen expression in acute lymphocytic leukemias of T cell of a B cell leukemia with an immunotoxin composed of anti-idiotypic and B cell precursor origins. J. Clin. Invest., 71: 301-313, 1983. antibody and saporin. J. Exp. Med., 766:43-62, 1987.

6277

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1988 American Association for Cancer Research. Clonal Analysis of Untreated Non-Hodgkin's Lymphoma Utilizing Immunoglobulin Gene Rearrangement and Immunophenotype

Richard A. Rudders, Satinder Dhillon and Theodore G. Krontiris

Cancer Res 1988;48:6272-6277.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/48/22/6272

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/48/22/6272. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.