Normal no more

In this issue of Blood, Gondek and colleagues demonstrate that the designation “normal cytogenetics” is a misnomer in the majority of cases of myelodysplastic syndrome.

The failure to obtain interpretable metaphases, or the finding of “normal” cytogenetics, in the face of obvious malignant diseases such as myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) leaves the clinician/researcher in a quandary, both with regards to informing the patient about risk and in developing strategies to investigate or treat the disease(s). In the article by Gondek and colleagues, the authors have used high-density single nucleotide polymorphism arrays (SNP-As) to identify changes in copy number or parental origin in the DNA of 129 MDS patients and 45 normal control subjects.

Using this technology, they were able to identify correctly the gains and losses of chromosome regions identified by classic cytogenetics. However, they were also able to identify changes not readily apparent through classic cytogenetics. These included small gains and losses below the resolution of banding methods. In addition, uniparental disomy (UPD) was evident in a good proportion of cases. UPD is a condition in which there are 2 copies of a chromosome or chromosomal region, but rather than a copy being inherited from each of the parents, the material present is derived from a single parent. Importantly, the authors demonstrated the acquired nature of such changes by comparing, within the same patient, DNA from T cells and MDS bone marrow. This loss of heterozygosity is presumably important in the development of the disease and may have an impact upon its behavior.

In the current study, no growth was found in 6% and normal cytogenetics in 42% of cases, respectively. By SNP-A analysis, it was possible to show that half of the cases with no growth harbored changes, and almost two thirds of the “normal” cytogenetic cases were abnormal. In what should be considered preliminary but intriguing results, the finding of an abnormality by SNP-A predicted a worse outcome.

What about the cases that were normal according to cytogenetics and SNP-A? In the current study, DNA from total bone marrow was used without any effort to separate MDS cells from residual normal cells. The finding of a normal SNP-A result could be due to the presence of residual normal cells. This is because the SNP-A method relies on being able to distinguish between 1 or 2 copies of a gene to identify regional loss, and therefore is sensitive to contamination. An alternate explanation for a normal SNP-A result is that in some cases of MDS, genetic changes other than copy number, such as point mutation and unrecognized chromosomal rearrangements, may be important in the development of the disease.

The current study clearly demonstrates that in most cases, normal cytogenetic MDS is a misnomer. In the relatively small number of cases evaluated by Gondek and colleagues, it is evident that some of the losses are recurrent. The application of the SNP-A methodology described here, or of other high-throughput copy number methods such as high definition aCGH (array comparative genomic hybridization) or bacterial artificial chromosome (BAC) arrays, will be of value in identifying recurrent (copy number) abnormalities in MDS, thereby allowing better definition of this malignant state. The current SNP-A method is not suitable for clinical application; however, the information derived from it should make it possible to develop a relatively inexpensive, informative test.