Abstract
The genetic hallmark of MDS is a gain or loss of chromosomal loci identified in bone marrow (BM) cells in ∼50% of pts at diagnosis. We previously demonstrated in longitudinal chromosome study a modulating effect of chronic AZA C therapy on the MDS clone which subdivided patients into five distinct cytogenetic groups with statistically significant differences in survival (P=0.0003) (Najfeld et al, ASH 2004). Our goal was, therefore, to investigate whether PB cells can be used in substitute of BM cells for detection of genomic defects to monitor the clone during AzaC-based therapy. Chromosomal and interphase (I-) FISH studies were performed using BM and PB cells at baseline and following AzaC-based therapy. I-FISH studies were evaluated with a panel of six probes [EGR1 (5q31), D7S522 (7q31), D8Z2 (8p11.1-q11.1), MLL (11q23), Rb1 (13q14), D20S108 (20q12)]. Of the 47 pts studied, 25 (53%) had a normal karyotype and disomy for the MDS panel of six probes. The other 22 pts (47%) were cytogenetically abnormal, showing concordant results in 16 of 22 pts (73%) for cytogenetic and I-FISH genomic defect in BM cells (showing ≤20% frequency difference). The remaining 6 pts had a mean of 77% (range 37.5–100%) of abnormal metaphase cells and a mean of 42% (range 1.8–73.3%) of abnormal BM interphase nuclei. The 35% difference in frequency seen between metaphase and interphase BM cells is attributed to the proliferative advantage of metaphase cells with a complex karyotype. The frequency of the abnormal MDS-marked clone in BM and PB cells was concordant in 13 of 20 pts (65%). Hematological response of these pts was PR=1, hematological improvement=3, stable disease=2, too early for evaluation=6, and 1 patient had no hematological response after 10 months of AZA C treatment. The remaining 7 pts had a mean of 59% (range 48–68%) of abnormal BM metaphase cells and a mean of 21% (range 7–41%) of abnormal PB interphase nuclei. The MDS abnormal clone was detectable in more than two-fold higher frequency in BM compared to PB prior to treatment in 35% of pts. Preliminary sequential studies in discordant pts revealed a transition between the BM and PB cells to concordant frequencies within 4–5 months after AzaC-based therapy. These early observations suggest that monitoring AzaC-based therapy can be achieved using peripheral blood cells in 80% of pts. Remarkably, one pt with monosomy 7, a notoriously poor IPSS indicator, demonstrated a hematological improvement, full cytogenetic and 97% FISH remission, both in the BM and PB after 4 months of AzaC-based therapy. To examine AzaC’s response on cell lineage involvement in MDS, purified BM and PB cells were subjected to FISH analysis before and during treatment from patients who were cytogenetically abnormal at baseline. Similar frequencies of genomic imbalances were seen in purified BM and PB derived CD34+ (97% vs. 96%) and CD15+ cells (94% vs. 98%), indicating that these cell populations had a similar response to AZAC therapy, as monitored in PB or BM, during the initial treatment period (4–5 months). Purified BM and PB-derived T-lymphocytes (CD3+/4+/8+) had normal disomic patterns before and during AzaC-based therapy, indicating that T-cells were not involved in the MDS clone in these pts. In contrast, BM and PB derived B-cells (CD19+) had slightly discordant results, showing a mean of 83% vs. 56% respectively of MDS-marked clone, indicating a trend towards greater response in PB- derived B-cells when compared to BM-derived CD19+ cells. In summary, our results demonstrates that although the MDS abnormal clone may be detected in both the BM and PB at the start of therapy, due to the proliferative advantage of the abnormal clone the optimal tissue should be bone marrow. This is the first study demonstrating that tracking the MDS-marked clone during the AZA-C therapy is feasible in peripheral blood cells.
Longitudinal dynamics of clonal hematopoiesis identifies gene-specific fitness effects