Assessing residual leukemia through fluorescence in situ hybridization

MINI-REVIEW

Assessing residual leukemia through fluorescence in situ hybridization (FISH)

Marileila Varella-Garcia

University of Colorado Health Sciences Center, Cancer Center, Campus Box B188, 4200 East Ninth Avenue, Denver, CO 80262, USA. Fax: [phone]. E-mail: [e-mail]

INTRODUCTION

While patients with acute leukemia frequently experience clinical remission after therapy, most of them subsequently relapse. The relapse of individual patients cannot be accurately predicted, but specific chromosome abnormalities are among the most important predictors of clinical outcome. Diagnosis of residual neoplastic disease through the detection of cells carrying these abnormalities provides a useful assessment of individual response to treatment and an accurate evaluation of the efficacy of a particular therapy protocol in a group of patients. Detection and quantification of residual disease may help to improve therapeutic strategies by grouping patients according to risk of relapse, allowing risk-oriented or individualized treatment protocols.

DETECTION OF MINIMAL RESIDUAL LEUKEMIA

During the last decade, several methods have been developed and optimized to detect residual disease in hematological malignant disorders. Cell morphology analysis by light microscopy is the most widely used approach for detection of tumor cells, yet it is quite limited in sensitivity usually not detecting abnormal cells at frequencies lower than 10^-2 (Dwenger et al., 1996). Classical cytogenetics has been the reference test to confirm remission of acute leukemias (Glassman, 1997). However, this technique is laborious and relies on analysis of metaphase spreads, which limits its sensitivity due to the low mitotic index and poor chromosome morphology commonly associated with acute leukemia cells. Standard analysis of 20 metaphases per patient theoretically allows exclusion of cellular mosaicism at a level of 14% with 95% confidence (Hook, 1977). Even though certain tumor cells have a relatively high proliferation rate, which would favor their detection by metaphase analysis, it is generally agreed that the maximum sensitivity of this assay does not exceed 10^-2. Therefore, although the classical cytogenetics approach is an effective test at the time the disease appears or during clinical relapse, patients in conventional remission may harbor as many as 10¹⁰ neoplastic cells and still remain cytogenetically undetected.

The most sensitive approach to detect residual leukemias characterized by molecularly cloned chromosome translocations is based on the identification of tumor markers amplified by the reverse-transcription polymerase chain reaction (RT-PCR). This technique enables the detection of a single positive cell among 10⁶-10⁸ normal cells (Biernaux et al., 1995; Cross, 1997). Identification of pre-recurrence conditions has been documented for a number of diseases, including chronic myelogenous leukemia (CML) and acute lymphoblastic leukemia (ALL). In CML patients in clinical remission, the detection of the RNA transcripts of the chimeric gene BCR/ABL by RT-PCR was demonstrated to be highly predictive for relapse (Drobyski et al., 1997; Preudhomme et al., 1997), helping to identify candidates for preventive immunotherapy. In addition, PCR-based methods detected molecular relapse on average five months earlier than classical cytogenetics in these patients. More sophisticated approaches, including competitive PCR assays to quantify BCR-ABL transcripts, have confirmed that rising levels of the mRNA transcribed by the fusion gene can be observed before detection of CML relapse at the cytogenetic level (Zhang et al., 1996).

On the other hand, PCR-based techniques have some important disadvantages. Sporadic cases of PCR-positive acute promyelocytic leukemia (APL) and CML patients (Tanaka et al., 1997) and numerous PCR-positive t(8;21) acute myelogenous leukemia (AML) (Nucifora et al., 1993; Chang et al., 1993; Kusec et al., 1994) patients have remained in long-term remission. Conversely, the sensitivity for the PCR assays relies on the primer sets selected. For t(9;22), for instance, it is usually below 100% since the translocation breakpoint is variable (Tkachuk et al., 1990; Cox et al., 1998). Furthermore, PCR techniques may generate a high incidence of false-positive results due to cross-contamination, even when strict precautions to minimize this problem are adopted.

DUAL-COLOR FLUORESCENCE IN SITU HYBRIDIZATION FOR INTERPHASE DETECTION OF GENE FUSION

A recently explored alternative molecular approach for the detection of residual leukemia is fluorescence in situ hybridization (FISH). FISH allows the recognition of specific chromosome targets as clearly localized and brightly fluorescent spots. This technique has greatly increased the potential of classical cytogenetics for identifying chromosome rearrangements in both metaphase and interphase cells (Lichter, 1997; Werner et al., 1997). While metaphase FISH has been recognized as a more sensitive technique than karyotype analysis, especially in conditions that do not favor good chromosome morphology, it is interphase FISH that has strikingly improved the applicability of the technique. An effort to follow up leukemia patients by identifying numerical chromosome aberrations in interphase cells was described by Arkesteijn et al. (1996) using chromosome-specific probes. However, conflicting results were found between FISH and bone marrow cytology in a significant proportion of the patients, probably due to non-specific changes in the chromosomal constitution of the relapse clones. Therefore, the chimeric genes postulated as the critical factors for leukemogenesis (Look, 1997) are the best targets for the FISH approach.

Several dual-color FISH strategies for detection of gene fusion in interphase cells have been published, varying in the probe design and consequently in the presentation of the fluorescent signals. Essentially, the translocation probe designs may be grouped into three categories. In the conventional design, two differentially labeled DNA probes are used, one probe carrying sequences mapped proximal to the breakpoint in one of the chromosomes involved in the reciprocal translocation and the other probe carrying sequences mapped distal to the breakpoint in the other chromosome. Positive cells for the translocation will display one dual-color fusion signal (for the derivative chromosome carrying the chimeric gene) and two single-color signals, one for each of the normal alleles (Figures 1A and 2A). The second FISH probe design also comprises two differentially labeled probes, one including a DNA segment mapped proximal or distal to the translocation breakpoint in one chromosome, and another spanning the breakpoint in the second chromosome. Positive cells for the translocation, in this case, will display one dual-color fusion plus two discrete signals for the probe spanning the breakpoint (one "extra signal") and one signal for the other allele (Figures 1B and 2B). The third FISH probe design consists of differentially labeled DNA segments encompassing the translocation breakpoints in both chromosomes involved in the translocation. This approach identifies a cell positive for the translocation as displaying two dual-color fusion signals (one for each derivative chromosome) and two single-color signals (one for each normal allele) (Figures 1C and 2C).

Figure 1
- Probe configurations for detection of gene fusion by fluorescence in situ hybridization (FISH). Diagrams in the top section of the panels represent normal metaphase chromosomes and interphase nuclei, whereas the translocations are represented in the bottom section of the panels. The translocation breakpoints are indicated by arrows in the chromosomes at left. A, Conventional format illustrated for the PML-RARA probe set: a 15q22 clone mapped proximal and a 17q21 clone mapped distal to the t(15;17) breakpoints were used as probes, generating a single fusion signal in the derivative chromosome 15. B, "Extra-signal" format illustrated for the BCR/ABL probe set: the chromosome 22 probe is proximal to the 22q11 breakpoint and the chromosome 9 probe spans the 9q34 breakpoint generating a fusion signal for the derivative 22 and an extra signal for the derivative 9. C, Dual-fusion format represented by the AML1/ETO probe set: 8q22 and 21q22 probes encompass the translocation breakpoints generating fusion signals for both derivatives 8 and 21.

Figure 2
- Fluorescence in situ hybridization in bone marrow cells. Cells carrying the targeted translocations are identified by the arrows. A, conventional FISH approach: LSI PML/RARA probe labeled with SpectrumOrange and SpectrumGreen (VYSIS) for detection of t(15;17)(q22;q11-21). B, Extra signal approach: BCR/ABL SxFISH probe (Oncor) labeled with Rhodal green and Texas red for identification of t(9;22)(q34;q11). C, Dual-fused approach: AML1/ETO D-FISH probe (Oncor) labeled with Rhodal green and Texas red for detection of t(8;21)(q22;q22).

The conventional approach for interphase FISH detection of chromosome translocation was first proposed by Tkachuck et al. (1990) to be applied to the BCR/ABL chimeric gene and has been used in research and clinical cytogenetics laboratories. A variety of similarly designed FISH probes have been used for detection of other translocations, such as t(15;17)(q22;q11-q21) in AML type M3 (Schad et al., 1994) and t(8;21)(q22;q22) in AML type M2 (Sacchi et al., 1995). Patients at disease presentation or at relapse usually carry high frequencies of bone marrow cells with the fusion gene and this FISH strategy is very successful for diagnosis (Cox et al., 1998; Tbakhi et al., 1998). Interphase FISH using the conventional probe design was already postulated as superior to both morphological observation of blast cells in bone marrow smear preparations and chromosome analysis for detection of residual disease (Mancini et al., 1995). An increase in the percentage of cells carrying the typical chromosome translocations was observed in sequential analyses several months before relapse, for instance in CML and APL (Tanaka et al., 1997; Onishi et al., 1997). However, apparent fusion signals may be observed by random co-localization of two distinct color signals in interphase nuclei. The level of false-positive cells averages 3-4% in normal bone marrow cells (Mancini et al., 1995; Tanaka et al., 1997; Onishi et al., 1997; Cox et al., 1998), which clearly jeopardizes the use of this test for detection of residual leukemia.

Studies using the "extra-signal" FISH strategy were performed for interphase detection of t(15;17) and t(8;21) (Fischer et al., 1996) and of t(9;22) (Varella-Garcia, M., unpublished data). Because of the more stringent criterion for classification of a positive cell, the percentage of false-positive cells in these studies was less than 1%. Recently, commercial probe sets with this design were released targeting the t(9;22) (LSI BCR/ABL ES from Vysis; BCR/ABL SxFISH from Oncor, Figure 2B) and the t(12;21)(p13;q22) (LSI TEL/AML1 ES from Vysis). The third category of FISH probe design, the dual fusion probe set, was newly optimized for detection of chimeric products of t(15;17) (Fischer et al., 1996), t(8;21) (Paskulin et al., 1998; Figure 2C) and t(9;22) (Dewald et al., 1998). The latter two probe sets are commercially available (BCR/ABL D-FISH and AML/ETO1 D-FISH from Oncor). Since the criterion for positivity requires the presence of two fused signals (representing both derivative chromosomes) and two single signals (representing the normal alleles), it is highly unlikely that artificial conditions will generate this phenotype. Therefore, the cut-off value in the D-FISH design is very low and the test is considered appropriate for monitoring residual leukemia.

Efforts are ongoing in our laboratory to sequentially examine bone marrows of t(8;21) AML patients in remission using the AML1/ETO D-FISH probe. Only one cell classified as positive was observed among 19,000 cells from 19 normal bone marrows that were scored so far. This frequency opposes the average of 2% to 5% cells displaying randomly juxtaposed signals in normal specimens when the conventional design of probes for detecting translocation is used, or 1% of false-positive cells when the "extra signal" probe format is used. In our study, some patients have been displaying around 0.5 x 10^-3 cells positive for the t(8;21) for more than a year with no indication of clinical relapse. However, it remains to be clarified whether there is a certain level of positive cells or a pattern of change that predicts relapse, in which case the FISH test should be included in future treatment protocols.

Additional variant FISH strategies have been proposed, but they either lack improved sensitivity or require complex analyses. That was the case of the design proposed by Sinclair et al. (1997) using a triple-probe, triple-color approach for the BCR-ABL detection. Their probe set improved the sensitivity of the fusion detection, but it requires analysis in images acquired by a cooled CCD camera in a computerized system, which is not feasible in a clinical cytogenetics laboratory setting. Similarly, Tanaka et al. (1997) reportedly increased the sensitivity of the conventional FISH strategy to 10^-3 by combining the specific translocation probe with other probes, such as whole chromosome paint for chromosome Y in opposite-sex grafts and the RB gene probe. However, the probe combination has to be specifically planned for each patient, based on the individual tumor characteristics, which weakens the strategy in a routine clinical setting.

Caveats related to the interphase FISH analyses are represented by the atypical translocations, including the complex and masked variants. Complex variants involve at least three chromosomes, two of which should be the ones regularly affected by the typical rearrangements and at the expected breakpoints. Masked variants involve submicroscopic translocations that went undetected by conventional cytogenetics. Although these variants will not exhibit the expected pattern of FISH signals for a cell carrying a typical translocation, they generate special patterns that do not match a normal specimen either. Consequently, the atypical patients will be recognized as abnormal, although the best criteria for analysis must be established on a case by case basis. Examples of gene fusion identified by interphase FISH in complex and masked translocations were reported by Sabatino et al. (1997) and Dewald et al. (1998) for BCR/ABL, and by Fischer et al. (1996) and Paskulin et al. (1998) for AML1/ETO.

CONCLUDING REMARKS

Interphase FISH was demonstrated as a reliable, less laborious and faster laboratory test for diagnosis and management of leukemias with specific chromosome abnormalities compared to classical cytogenetics and molecular genetics procedures. In addition, using an appropriate probe design, FISH is a sensitive tool for detection and quantification of residual disease and potentially for prediction of clinical relapse. Further development of a comprehensive set of diagnostic probes, which has been highly stimulated by the Human Genome Project, is anticipated to dramatically expand the application of interphase FISH in clinical oncology in the near future.

(Received July 7, 1998)

• Arkesteijn, G.J.A., Erpelink, S.L.A., Martens, A.C.M., Hagemeyer, A. and Hagenbeek, A. (1996). The use of FISH with chromosome-specific repetitive DNA probes for the follow-up of leukemia patients: correlations and discrepancies with bone marrow cytology. Cancer Genet. Cytogenet. 88: 69-75.
• Biernaux, C., Loos, M., Sels, A., Huez, G. and Stryckmans, P. (1995). Detection of major bcr-abl gene expression at a very low level in blood cells of some healthy individuals. Blood 86: 3118-3122.
• Chang, K.S., Fan, Y.H., Staas, S.A., Estey, E.H., Wang, G., Trujillo, J.M., Erickson, P. and Drabkin, H. (1993). Expression of AML1/ETO fusion transcripts and detection of minimal residual disease in t(8;21)-positive acute myeloid leukemia. Oncogene 8: 983-988.
• Cox, M.C., Maffei, L., Buffolino, S., Del Poeta, G., Venditti, A., Cantonetti, M., Aronica, G., Aquilina, P., Masi, M. and Amadori, S. (1998). A comparative analysis of FISH, RT-PCR, and cytogenetics for the diagnosis of bcr-abl positive leukemias. Am. J. Clin. Pathol. 109: 24-31.
• Cross, N.C.P. (1997). Assessing residual leukemia. Bailličre's Clin. Haematol. 10: 389-403.
• Dewald, G.W., Wyatt, W.A., Juneau, A.L., Carlson, R.O., Zinsmeister, A.R., Jalal, S.M., Spurbeck, J.L. and Silver, R.T. (1998). Highly sensitive fluorescence in situ hybridization method to detect double BCR/ABL fusion and monitor response to therapy in chronic myeloid leukemia. Blood 91: 3357-3365.
• Drobysky, W.R., Endean, D.J., Klein, J.P. and Hessner, M.J. (1997). Detection of BCR/ABL RNA transcripts using the polymerase chain reaction is highly predictive for relapse in patients transplanted with unrelated marrow grafts for chronic myelogenous leukemia. Br. J. Haematol. 98: 458-466.
• Dwenger, A., Lindemann, A. and Mertelsmann, R. (1996). Minimal residual disease: Detection, clinical relevance, and treatment strategies. J. Hematother. 5: 537-548.
• Fischer, K., Scholl, C., Salat, J., Froling, S., Schlenk, R., Bentz, M., Stilgenbauer, S., Lichter, P. and Dohner, H. (1996). Design and validation of DNA probe sets for a compre-hensive interphase cytogenetic analysis of acute myeloid leukemia. Blood 88: 3962-3971.
• Glassman, A.B. (1997). Cytogenetics, an evolving role in the diagnosis and treatment of cancer. Clin. Lab. Med. 17: 21-37.
• Hook, E.B. (1977). Exclusion of chromosomal mosaicism: tables of 90%, 95%, and 99% confidence limits and comments on use. Am. J. Hum. Genet. 29: 94-97.
• Kusec, R., Laczika, K., Knobl, P., Friedl, J., Greinic, H., Purtscher, B., Haas, O.A., Lechner, K. and Jaeger, U. (1994). AML1/ETO fusion mRNA can be detected in remission blood samples of all patients with t(8;21) acute myeloid leukemia after chemotherapy or autologous bone marrow transplantation. Leukemia 8: 735-739.
• Lichter, P. (1997). Multicolor FISHing: what's the catch? Trends in Genet. 13: 475-479.
• Look, A.T. (1997). Oncogenic transcription factors in the human acute leukemias. Science 278: 1059-1064.
• Mancini, M., Nanni, M., Cedrone, M., Diveria, D., Avvisati, G., Riccione, R., De Cuia, M.R., Fenu, S. and Alimena, G. (1995). Combined cytogenetic, FISH and molecular analysis in acute promyelocytic leukaemia at diagnosis and in complete remission. Br. J. Haematol. 91: 878-884.
• Nucifora, G., Larson, R.A. and Rowley, J.D. (1993). Persistence of the 8;21 translocation in patients with acute myeloid leukemia type M2 in long-term remission. Blood 82: 712-715.
• Onishi, R., Tanaka, K., Shimazaki, C., Yamagata, N., Konaka, Y., Namiuchi, S. and Kamada, N. (1997). Sequential interphase FISH analysis of m-BCR/ABL chimeric gene-positive cells in Ph-positive acute myeloid leukemia. Leuk. Lymphoma 26: 185-191.
• Paskulin, G.A., Philips, G., Morgan, R., Sandberg, A., Richkind, K., Borovik, C., McGavran, L., Rabinovich, N., Dietz-Band, J., Erickson, P., Drabkin, H. and Varella-Garcia, M. (1998). Pre-clinical evaluation of probes to detect t(8;21) AML minimal residual disease by fluorescence in situ hybridization. Genes Chrom. Cancer 21: 144-151.
• Preudhomme, C., Henic, N., Cazin, B., Lai, J.L., Bertheas, M.F., Vanrumbeke, M., Kemoine, F., Jouet, J.P., Deconninck, E., Nelken, B., Cosson, A. and Fenaux, P. (1997). Good correlation between RT-PCR analysis and relapse in Philadelphia (Ph1)-positive acute lymphoblastic leukemia (ALL). Leukemia 11: 294-298.
• Sabatino, G., Calabrese, G., Franchi, P.G., Stuppia, L., Jadayel, D., Donti, E. and Palka, G. (1997). Routine fluorescence in situ hybridization analysis for detection of bcr-abl rearrangements in myeloproliferative disorders. Leuk. Res. 21: 581-583.
• Sacchi, N., Magnani, I., Kearney, L., Wijsman, J., Hagemeijer, A. and Darfler, M. (1995). Interphase cytogenetics of the t(8;21)(q22;q22) associated with acute myelogenous leukemia by two-color fluorescence in situ hybridization. Cancer Genet. Cytogenet. 79: 97-103.
• Schadz C.R., Hanson, C.A., Paietta, E., Casper, J., Jalal, S.M. and Dewald, G.W. (1994). Efficacy of fluorescence in situ hybridization for detecting PML/RARA gene fusion in treated and untreated acute promyelocytic leukemia. Mayo Clin. Proc. 69: 1047-1053.
• Sinclair, P.B., Green, A.R., Grace, C. and Nacheva, E.P. (1997). Improved sensitivity of bcr/abl detection: a triple-probe three-color fluorescence in situ hybridization system. Blood 90: 1395-1402.
• Tanaka, K., Arif, M., Eguchi, M., Kumaravel, T.S., Ueda, R., Ohno, R., Iwato, K., Kyo, T., Dohy, H. and Kamada, N. (1997). Application of fluorescence in situ hybridization to detect residual leukemic cells with 9;22 and 15;17 translocations. Leukemia 11: 436-440.
• Tbakhi, A., Pettay, J., Sreenan, J.J., Abdel-Razek, H., Kalaycio, M., Hoeltge, G., Miller, M.L. and Tubbs, R.R. (1998). Comparative analysis of interphase FISH and RT-PCR to detect bcr-abl translocation in chronic myelogenous leukemia and related disorders. Am. J. Clin. Pathol. 109: 16-23.
• Tkachuck, D.C., Westbrook, C.A., Andreef, M., Donlon, T.A., Cleary, M.L., Suryanarayan, K., Homge, M., Redner, A., Gray, J. and Pinkel, D. (1990). Detection of bcr-abl fusion in chronic myelogenous leukemia by in situ hybridization. Science 250: 559-562.
• Werner, M., Wilkens, L., Aubele, M., Nolte, M., Zitzelsberger, H. and Komminoth, P. (1997). Interphase cytogenetics in Pathology: principles, methods and applications of fluorescence in situ hybridization (FISH). Histochem. Cell Biol. 108: 381:390.
• Zhang, J.G., Lin, F., Chase, A., Goldman, J.M. and Cross, N.C.P. (1996). Comparison of genomic DNA and cDNA for detection of residual disease after treatment of chronic myeloid leukemia with allogeneic bone marrow transplantation. Blood 87: 2588-2593.

Publication Dates

• Publication in this collection
  23 Feb 1999
• Date of issue
  Sept 1998

History

• Received
  07 July 1998