Introduction

The DNA content of cells is measured by the ability of propidium iodide to bind stoichiometrically to DNA under appropriate staining conditions. The nuclei of these stained cells are evaluated individually for DNA content by flow cytometry. The results are displayed graphically as a histogram in which the fluorescence emitted by each nucleus is directly proportional to its DNA content.

Flow cytometric DNA analysis is performed for one or both of the following reasons: 1) To determine the presence of aneuploidy cells in a population, and 2) to determine the percentage of cells in each phase of the cell cycle and estimate the growth fraction of the population. These objectives necessitate distinct requirements in terms of data analysis. Ploidy determination relies upon the position of G₀/G₁ peak in the sample in relation to the G₀/G₁ peak of control diploid cells derived from the same species. Cell cycle analysis required the division of the DNA histogram into three areas representing the G₀/G₁, S and G₂/M phases and integration of the area representing each phase.

The determination of ploidy is usually an easier procedure than accurately assessing the cell cycle. The channels representing the maximum G₀/G₁ peak heights of the sample and control may be delineated either manually or by computer, and aneuploidy is defined as a significant difference between the two peaks. The difference in DNA content can be expressed as the ratio of test (tumour) sample / standard DNA fluorescence, defined as the DNA index. Diploid and non-human (e.g. chicken erythrocytes with 35% human diploid DNA content) control are often run in parallel with each sample to reduce the possibility of apparent shifts in the G₀/G₁ peaks being due to instrument drift or fluctuations in the laser output.

DNA staining is a crucial step since determination of DNA content assumes that the histogram is composed of events whose relative fluorescence intensity is due to the dye bound in a consistent fashion. Propidium iodide staining usually gives the best coefficient of variation (CV, defined as the standard deviation of the histogram G₁ peak distribution divided by the mean peak channel) and is currently the most popular dye being used. It is stimulated by the 488 nm wavelength and emits in the orange red spectrum.

Various definitions of aneuploidy exist but the most conservative one calls for identification of distinct G₀/G₁ peaks clearly separate from the diploid reference standard. With satisfactory CV values, separate peaks can be distinguished for DNA index values of < 0.95 and > 1.05.

DNA index measurement in childhood acute lymphoblastic leukaemia (ALL)

In childhood ALL, DNA index of >= 1.16 and <= 1.6 is associated with hyperdiploidy of > 50 chromosomes (range in modal numbers 51 - 68). This group represents 25 - 30% of childhood ALL, have more favourable presenting features and higher cure rates than other major prognostic subgroups. The significance of DNA index determination by flow cytometry at diagnosis, however, is limited by the fact that it does not identify which chromosomes have been gained or detect the presence of adverse structural changes.

We analyze our DNA index data for childhood ALL at diagnosis over the past years (1997 to the present). Using a DNA index of >= 1.16 as an indicator of hyperdiploidy > 50 chromosomes, 9 cases (all B-lineage ALL) are identified among a total of 45, giving an overall prevalence of 20%. The prevalence among B-lineage ALL (n = 36) is 25%. The corresponding figures among childhood ALL cases managed at Queen Mary Hospital (n = 24) are 29% (7 out of 24) in total cases and 35% (7 out of 20) in B-lineage ALL. DNA index of >= 1.16 is not encountered in T-lineage ALL. A clinical summary of the 7 cases managed at Queen Mary Hospital is tabulated in Table 1.

Discussion

Based on our data, the prevalence of hyperdiploid > 50 chromosomes, in contrast to an earlier cyogenetics study showing a probable lower prevalence [1], is similar in Hong Kong when compared to that in the literature. Moreover, while flow cytometric determination of DNA index is not able to detect structural changes in the hyperdiploid clone or identify the chromosomes that are gained, it is applicable at diagnosis to cover cases with apparently normal cytogenetics and also cytogenetics failure. This is well illustrated by the present case series. DNA index of >= 1.16 and <= 1.6 has been incorporated into one of the criteria to predict for low risk category as classified at St. Jude Children's Research Hospital.

A DNA index of >= 1.16 is not identified in all T-ALL cases (n = 9) analyzed. Unfortunately due to the very small sample size, we could not show a statistically significant improvement in event free and overall survivals when compared with those cases with DNA index of < 1.16.

The previous notion of a worse prognosis in hyperdiploidy and an associated structural abnormality has not been confirmed by subsequent analysis [2]. We do not come across structural changes in hyperdiploid cells (Table 1). Patients with 51 - 55 chromosomes, however, appear to behave less favourably than the hyperdiploid 56 - 67 subgroup [2]. The poorer prognosis of the hyperdiploid 51 - 55 subgroup is due to higher frequency of isochromosome 17q, lower prevalence of trisomies 4 and 10, and a higher proportion of patients with leucocyte counts > 50 X 10⁹/L.

The evaluation of hyperdiploid metaphases by conventional cytogenetics is technically challenging due to poor chromosome morphology. The use of fluorescence in situ hybridization with suitable centromeric probes may reveal hidden hyperdiploidy and can also determine the presence of minimal residual disease and the size of the leukaemic clone.

References

1. Chan LC, Ha SY, Ching LM et al. Cytogenetics and immunophenotypes of childhood acute lymphoblastic leukemia in Hong Kong. Cancer Genet Cytogenet 76: 118 - 124; 1994.

2. Raimondi SC, Pui CH, Hancock ML et al. Heterogeneity of hyperdiploid (51 - 67) childhood acute lymphoblastic leukemia. Leukemia 10: 213 - 224; 1996.

3. Moorman AV, Clark R, Farrell DM et al. Probes for hidden hyperdiploidy in acute lymphoblastic leukemia. Genes Chromosomes and Cancer 16: 40 - 45; 1996.

Table 1: Summary of childhood acute lymphoblastic leukaemia with DNA index of > 1.16