The ability of a-CGH to detect reliably chromosome imbalances in mosaicism was not yet well established . So, the goal of the present work was very practical: to assess the sensitivity of one of the most used oligomer-based a-CGH system to detect acquired (as well as constitutional) low level mosaicism. In particular, most recurrent unbalanced anomalies in MDS/AML and in disorders predisposing to MDS/AML have a size comparable to those of our patients 1 and 2 , and a-CGH may be used to monitor the abnormal clone during the disease course, and to detect MRD. Altogether our results demonstrate clearly the possibility to detect as low as 8% abnormal cells, at least for imbalances involving regions of a sufficiently great size, as those of patients 1 and 2. Obviously, more sensitive techniques are available to monitor unbalanced chromosome anomalies already identified [18, 20]: our aim was not to suggest a finer method to detect them, but a solid evaluation of a-CGH sensitivity is needed to draw correct conclusions when it is used to study cohorts of patients with disorders associated with acquired chromosome anomalies, as AML [7–9], MDS [8, 9, 15], or diseases predisposing to MDS/AML, as Shwachman-Diamond syndrome (SDS) .
A technical point deserves a comment: the possibility to reach a correct result is related to the parameters of DNA quality and to the choice of an informative platform: the use of customized slides should be related to the size of the region involved and it may be necessary to design enriched slides with higher resolution probe density for the region of interest, as we did in patient 3.
In patients 1 and 2 we were able to get easily evidence of the abnormal cell population at 10 and 8% levels of mosaicism (Figures 1, 2, 3), because we had already defined the presence of the imbalance. In our material the mosaicism was built up artificially, but in general in patients in which the imbalance may be suspected from the results of a-CGH at levels comparable to those here presented, the tool to reach a conclusion is FISH with informative probes, possibly on interphase nuclei, which will be able to confirm or deny the presence of the acquired mosaicism. This comparison of a-CGH and FISH results is essential to draw definite conclusions, in particular, in case of imbalances of smaller size, where we showed that the sensitivity of a-CGH to give evidence of a small population of abnormal cells is more variable (results in patient 3) with potentially aberrant regions more difficult to be revealed: a-CGH results may in fact be really significant, and they have to be more accurately investigated. So, we suggest that whenever a-CGH indicates a possible mosaicism which may be evaluated 8-10%, as in our diluted material, FISH is crucial.
A good example of the capacity of a-CGH to detect an unexpected acquired mosaicism is offered by a patient with SDS reported in 2006 , who was known to have a clone with an acquired chromosome anomaly in BM, namely an isochromosome for the long arms of chromosome 7, i(7)(q10). This anomaly may be related to the risk to develop MDS/AML , as a deletion of the long arms of chromosome 20, del(20)(q11), another frequent change found. In the follow-up, a-CGH showed, besides the i(7)(q10), an interstitial deletion of the long arms of chromosome 20 spanning 4.116 Mb in bands q11.21-q11.23. The resolution of standard chromosome analysis was insufficient to show the deletion, but FISH with the BAC probe CTD-3092L7, mapping within the deleted region, confirmed its presence in 30/170 mitoses (17.6%), and 68/470 nuclei (14.5%) (unpublished data).
We explored also the possibility to infer the percentage of abnormal cells found in acquired mosaics by a-CGH, at least approximately. A formula derived by the ADM2 algorithm used in the analytical software permits such a calculation, and, for instance, when we applied it to a-CGH results obtained in the patient with SDS described above, led to evaluate the cell population with the deletion of chromosome 20 to be 18.2% of BM cells . This evaluation agrees fairly well with the proportion of abnormal cells evaluated by FISH.