To determine the clonal nature of hematopoiesis and to assess lineage involvement in patients with myelodysplastic syndromes (MDS), we used restriction fragment length polymorphisms of the X-linked genes phosphoglycerate kinase (PGK1) and hypoxanthine phosphoribosyltransferase (HPRT) and the X-linked probe M27 beta. Eleven female MDS patients heterozygous for at least one of these probes were studied: 3 with refractory anemia (RA), 2 with RA with ringed sideroblasts (RARS), 2 with chronic myelomonocytic leukemia (CMML), and 4 with RA with excess of blasts in transformation (RAEB-t). All exhibited clonal hematopoiesis as determined by Southern analysis of DNA prepared from peripheral blood (PB) and/or bone marrow (BM) cells. In three of the six patients heterozygous for the PGK1 gene, purified cell suspensions of polymorphonuclear cells (PMN), monocytes, lymphocytes, and/or T cells prepared from PB were tested. In addition, five of these patients were analyzed by a polymerase chain reaction (PCR)-based procedure as described recently. This method was slightly adapted to facilitate the analysis of cell lysates of fluorescence- activated cell sorted (FACS) monocytes, T and B lymphocytes, and natural killer (NK) cells. The outcome of Southern and PCR analysis was concordant, showing that PMN and monocytes were clonally derived, whereas circulating T and B lymphocytes and NK cells exhibited random X- chromosome inactivation compatible with a polyclonal pattern. To address the question of whether T cells are derived from unaffected progenitor cells or that their origin had antedated the onset of MDS, naive and memory T cells were analyzed separately. Both subsets showed a polyclonal pattern. However, in one patient analysis of constitutive DNA suggested a skewed methylation, and the presence of clonal lymphocytes against a background of polyclonal lymphoid cells cannot be ruled out in this patient. PCR analysis of PB and BM cells showed a nonrandom, unilateral pattern of X-inactivation, compatible with a mixture of clonally (myeloid) and polyclonally (lymphoid) derived cells. In conclusion, in some patients, MDS represents a disorder with clonal hematopoiesis restricted to cells of myeloid origin, whereas a random X-inactivation pattern is found in lymphoid cells.

Interstitial deletions of the long arm of chromosome 5 are among the most characteristic abnormalities observed in myeloid disorders. To assess the lineage involvement of peripheral blood cells from patients with a 5q--anomaly, purified neutrophils, monocytes, T lymphocytes, and B lymphocytes were analyzed for loss of heterozygosity using six different highly polymorphic mininucleotide and dinucleotide (CA) repeat sequences from the 5q31 to 5q33 region. Ten patients were screened by polymerase chain reaction (PCR) amplification and proved to be informative for at least one marker. Six patients showed a complete or partial disappearance of an allele in myeloid cells, whereas cells of lymphoid lineages exhibited full heterozygosity. The other patients displayed no allelic loss, indicating that the informative markers were located outside the deleted chromosomal segments. In addition, three female patients who were also polymorphic for the BstXI site in the PGK- 1 gene were analyzed for the methylation status of this gene. Clonality of hematopoiesis, as determined by non-random X-chromosome inactivation, followed the same cell pattern as the 5q-specific allelic losses. In conclusion, using tumor-specific and clonal markers, we have demonstrated that the 5q- anomaly is restricted to cells of myeloid origin, leaving lymphoid cells unaffected.

To determine whether patients with acquired asplastic anemia (AA) exhibit clonal hematopoiesis, we used restriction fragment length polymorphisms of the X-linked genes phosphoglycerate kinase (PGK1) and hypoxanthine phosphoribosyltransferase (HPRT) and the X-linked probe M27 beta. Of the 19 female patients studied, 18 (95%) patients were informative for at least one marker. Of these, eight patients (42%) were heterozygous for PGK1, two (11%) for HPRT, and 16 (84%) for M27 beta. In 13 (72%) patients, a monoclonal pattern was found. Analysis of purified cell suspensions of four of these patients showed that both myeloid and lymphoid cells were of monoclonal origin, indicating the involvement of an early stem cell. The four patients who were studied at presentation all showed a monoclonal pattern. One of these patients showed a spontaneous recovery despite persistent clonal hematopoiesis. The presence of either clonal or polyclonal hematopoiesis did not show a correlation with the response to antithymocyte globulin (ATG) treatment. A relapse after ATG was also seen in a patient exhibiting polyclonal hematopoiesis. Conversely, a monoclonal pattern did not preclude the occurrence of a partial or complete response to ATG. Other potential markers to study clonality, including cytogenetic abnormalities or point mutations of the N-ras protooncogene, were not found in any of the patients. It is concluded that patients with AA may exhibit clonal hematopoiesis. The significance with respect to evolution to disorders with clonal hematopoiesis like paroxysmal nocturnal hemoglobinuria, myelodysplasia, and acute leukemia remains to be determined.