Donor cell leukemia

Abstract

Minimal residual disease refers to the tumour cells that are still present in a given patient after completion of a therapeutic scheme. The demonstration and quantification of residual neoplastic cells has a crucial impact in clinical decision making, for it might prompt continuation of treatment, while the absence of such cells might serve as evidence to withdraw therapy. Therefore, both sensitivity and specificity of the methods used to unravel residual neoplastic cells must be highly reliable and robust. Flow cytometry has been widely used for this purpose, and its clinical performance depends mainly on the criteria of interpretation, rather than in the technique by itself; molecular biology techniques have proved to be highly sensitive and specific but unfortunately they cannot be used in all patients or in all types of leukemia. Finally, the development of donor cell leukemia in transplanted patients, might mimic residual disease and add more confusion to an already controversial issue. These topics are discussed in this paper.

Keywords:

• Leukemia
• Donor cell leukemia
• Stem cell grafting

Leukemia relapse occurring in grafted donor cells, so called donor cell leukemia (DCL) after allogeneic hematopoietic stem cell transplantation, has been previously reported. It appears that DCL is a condition that has to be sought for; otherwise, it might be mistaken for a relapse or a second neoplasia. It is therefore possible that DCL is perhaps a more common occurrence than traditionally thought, but has been widely overseen for the analytical approach to unravel its presence is not used routinely.

Most descriptions found in the literature are single cases. The disease which was treated by the allograft was malignant in most instances; however, there were some cases of non-nepolastic disease that later on developed a malignancy in the donor cells. In most cases, the donor cell leukemia originated in the same lineage of the original malignancy; however, a switch of lineage of the malignancy has been reported. The time elapsed between the stem cell transplantation and the demonstration of the DCL has been reported from 2 to 164 months. The surveillance of the donors has been reported in some publications and no evidence of leukemia has been reported.

As far as the prevalence of this complication is concerned, the data are insufficient to draw definite conclusions. It is interesting that the number of cases of DCL published in a 30-year period (between 1971 and 2000) is similar to that of cases published in a 7-year period, between 2000 and 2006; these figures suggest that either the prevalence of this complication has increased in the last years, or that physicians look more thoroughly to the occurrence of leukemia in donor cells, since we are now better equipped to define the origin of the malignancy.1^–4

Patients with DCL have been successfully treated as new leukemias, using chemotherapy; others have received a second allograft. In general, the use of chemotherapy for treatment of DCL can induce remissions, but the treatment mortality is high; secondary stem cell transplantation can induce long-lasting remissions in some patients.

A major problem in the analysis of DCL is that demonstration of the donor cell origin of leukemic relapse after allogeneic transplantation is difficult and hampered by many pitfalls. The following section deals with the indication and interpretation of laboratory studies conducted to demonstrate the donor origin of the leukemia.

Various laboratory tests are currently available in order to characterize the chimeric status of hematopoietic compartments of a transplanted patient: cytogenetic detection of marker chromosomes, fluorescent in situ hybridization for the enumeration of sex-related chromosomes (XY-FISH), detection of Y chromosome-specific sequences (YCS-PCR) (42, 43), and detection of polymorphic markers like minisatellites or variable number tandem repeats (repeats of 10–100 bp), microsatellites or short tandem repeats (repeats of 2–6 bp), single-nucleotide polymorphisms, restriction fragment length polymorphisms, and short inversion/deletion polymorphisms.4

As the essentials of these tests is to distinguish between chromosomes, under adequate circumstances, they can almost all be employed to get clues about the origin of the leukemic cells in the case of a relapse. Before discussion of the limitations of these tests when it comes to establish the origin of leukemic cells, it is necessary to recall the definition of chimerism. It refers to the fact that cells of one organism are genetically of different origins. Although there are several ‘natural’ causes for chimerism like fetomaternal transmission or tetragametic chimerism, it is in our context the result of the infusion of allogeneic hematopoietic stem cells after myeloablative conditioning or non-ablative preparation of the receptor. The chimeric status can be assessed in different hematopoietic compartments like peripheral blood, bone marrow or even in purified cellular fractions. A ‘complete chimerism’ refers to a situation where all cells within a given compartment are donor-derived according to a test with a sensitivity level of around 1–5%. Obviously, using a test with a higher sensitivity on this given sample or maybe probing another compartment of the same patient, we may still find it positive for receptor derived cells. The first case is sometimes referred to as ‘microchimerism’, whereas the second case is known as ‘split chimerism’. Thus, for a careful interpretation of the data, it is important to identify clearly the compartment(s) where the chimerism has been determined as well as the method, together with an estimation of its sensitivity and accuracy. The first hint of DCL is generally obtained when a numerical discrepancy is observed between the fraction of blast cells in a given hematopoietic compartment and the chimeric status reported for the same compartment. In the best case, all cells are apparently of donor origin and a very high proportion if not all of these correspond to leukemic blasts.

Although the interpretation of the above mentioned situation might appear straight forward, the difficulty to demonstrate conclusively the donor origin of relapse resides in three problem areas that have to be dealt with: artifacts and limitations of the methods to establish the chimeric status, the genetic lability which is intrinsic to the leukemic transformation, and some of the proposed mechanisms leading to donor cell transformation, like uptake of oncogenic material by donor cells or fusion of donor cells with leukemic blasts.5,6

Inasmuch as unraveling the mechanisms of donor cell transformation might provide insight into the understanding of leukemogenesis, careful analyses of cases of donor cell leukemia could represent most valuable opportunities to increase our current knowledge of hematological neoplasia; however, it is rather cumbersome to find a pattern based on common findings in these cases. It has been already discussed that the demonstration of the donor or recipient origin of the ‘new’ leukemic cells is crucial, and requires very stringent methodology in order to ascertain that the cells are in fact from donor origin, instead of resulting from disease relapse or a secondary, therapy-related, neoplasia arising from the recipient’s cells. If some of the currently reported cases do not really correspond to donor cell leukemia, their clinical and laboratory features might only add to the confusion regarding the mechanisms of malignant transformation of grafted cells.7 Additionally and most importantly, in true cases of donor cell leukemia, proof that the cell donor was healthy and remained healthy during a long-term follow-up must be unequivocal.5^–8 Nevertheless, several and not necessarily mutually exclusive hypotheses have been forwarded to explain the ‘leukemization’ of donor cells, as follows: (1) sustained antigenic stimulation of lymphoid cells; (2) aberrant homeostasis which induces or promotes transformation; (3) Impaired immune surveillance; (4) hybridization of grafted cells with residual malignant cells; (5) communicable agents; (6) chemotherapy-induced mutagenesis/transformation; (7) replicative stress; and (8) first ‘hit’ in donor followed by second ‘hit’ in recipient. There are evidences — in certain instances — that support each one of all these possibilities, while there are also data that seem to rule them out, again, in certain circumstances.

Assuming that all the cases of donor cell leukemia that have been reported are really such cases, i.e. all new neoplasia originated from donor cells and none of the donors had or developed leukemia afterwards, it is rather obvious that there is not a common pattern which could favor any of the proposed hypotheses to explain malignant transformation of donor cells. Inasmuch as these hypotheses are not necessarily mutually exclusive, it is possible that several of the aforementioned factors might contribute, in different combinations, in different degrees, and in different times and sequences in each case. In the generation of autoimmune disease, it has been proposed that several defects need to add up in order to overcome a threshold for disease to develop. These adding defects are surely not the same in all patients, which explains why autoimmune diseases are associated with very many genetic and environmental factors but with none in a definite manner. Similarly, several events leading to malignant transformation of donor cells might need to add up in a patient for malignant transformation to take place, but such factors do not need to be the same in all patients for donor cell leukemia to develop. It is then feasible that several of the proposed mechanisms concur in most — if not all — grafted patients, but perhaps only certain critical combinations of adverse factors can trigger the donor cells to undergo malignant transformation.9