Alteration Of Cytogenetic Risk Stratification In New Diagnosis Of Leukemia By FISH Testing – Retrospective Review Of The SWOG Experience

Fluorescence in situ hybridization (FISH) is an invaluable tool for complementing conventional karyotype analysis in the diagnosis of leukemia. FISH panels are often requested by clinicians, along with karyotype analysis, for molecular determination of clinically actionable abnormalities. For most cooperative group clinical trials, unlike karyotype analysis, FISH has not been required to determine eligibility or risk stratification. Yet no previous studies have shown how often FISH results might alter the cytogenetic risk category determined by karyotype analysis. Furthermore, FISH panels which may include 5∼10 probe sets are costly while karyotype analysis for new diagnosis of myeloid leukemia is routine and highly informative. Therefore we aim to determine the frequency of cytogenetic risk alteration by FISH testing to serve as an evidence base for future study design and best clinical practice.

Our study set includes all leukemia patients enrolled in SWOG leukemia studies from 2002 to 2012 with both cytogenetic and FISH results available at the time of initial diagnosis (n=248), representing 12% of total patients enrolled in these studies. There are 169 patients with acute myeloid leukemia (AML), 22 of whom had acute promyelocytic leukemia (APL); 45 with chronic myeloid leukemia (CML-CP); and 34 with acute lymphocytic leukemia (ALL). The age range was 19-87 (median 51) years. Patients received various therapies based on the study protocols. All cytogenetic and FISH results were submitted from SWOG-approved laboratories and were reviewed by the SWOG Cytogenetics Committee. Risk categorization was based on combined SWOG and MRC criteria.

There was great variation in the number and the type of FISH probes used for the workup of new diagnosis of leukemia (range 1-12 per patient). The most frequent FISH findings with added value were cryptic translocations, including amplification and/or rearrangement of MYC (mean overall survival 8.2 months without remission) and rearrangement of MLL. In the submitted studies, there were no misses for subtle chromosome abnormalities by karyotype analysis, including the inv(16), although FISH clarified deletion 16q instead of inv(16) occasionally. FISH and karyotype analyses were consistent for large chromosome abnormalities such as deletions of 5q and 7q, t(8;21), and trisomy 8. Of all 248 patients, 47 (19.0%) had favorable cytogenetics, 133 (53.6%) intermediate, 64 (25.8%) unfavorable, and 4 (1.6%) unknown. Of the 147 non-APL AML patients, 23 (15.6%) had favorable cytogenetics, 79 (53.7%) intermediate [72 (49.0%) after FISH], and 64 (43.5%) unfavorable [71 (48.3%) after FISH]. Overall, eight patients (3.2%) demonstrated alterations in risk categorization after FISH analysis: 7 from intermediate to unfavorable among AML patients, and 1 from favorable to intermediate among ALL patients. Of the 45 CML patients, karyotyping adequately identified all t(9;22), and 3 patients demonstrated additional secondary abnormalities. FISH identified 5 patients with submicroscopic deletions on the derivative 9 (which was once considered a higher-risk feature but appears to have been overcome by therapy with tyrosine kinase inhibitors). Of the 34 ALL patients, 16 (47.1%) had unfavorable cytogenetics; 12 (35.3%) intermediate; and the remaining 2 showed favorable hyperdiploidy, one of which was reclassified as intermediate due to detection of structural rearrangements by FISH.

The frequency of cytogenetic risk alteration by FISH after karyotyping analysis for new diagnosis of leukemia was low, supporting current clinical trial designs that do not require extensive FISH panel testing for protocol eligibility. However, FISH clearly adds value in detecting cases with cryptic rearrangements, which can contribute to altered risk stratification. It may be helpful to ascertain subtle rearrangements by FISH, such as inv(3), inv(16), t(6;9), MLL (11q23) rearrangement, and MYC (8q24) amplification and rearrangement.

This work was supported by PHS Cooperative Agreement grants awarded by the National Cancer Institute, CA32102 and CA38926.

No relevant conflicts of interest to declare.