Review. Leukemia in Down's Syndrome

J. Rajantie
The Cancer Journal, Volume 9, Number 3 (May-June 1996)
Aurora Hospital and Children's Hospital, University of Helsinki, Nordensköldink 20 SF-00290 Helsinki, Finland
Fax 358-0-4702786
  Reprinted with the permission of Jean-Claude Salomon, M.D.
Editor, The Cancer Journal
33 (0)1 49 58 36 00
Fax: 33 (0)1 49 58 35 42


Patients with Down's syndrome have an increased risk of developing leukemia. The reason remains unknown. Although most of the patients have ALL, a great majority of smaller children have ANLL. ALL of Down's syndrome is characterized by increased frequency of favorable prognostic signs. With modern treatment regimens its prognosis is similar to that in other patients. ANLL is diagnosed at younger age, has a large overpresentation of acute megakaryoblastic subtype, different cytogenetic abnormalities, and a better prognosis than in general. In view of the favorable recent results, a conservative approach to treating leukemia of patients with Down's syndrome does not seem warranted. The challenge is to design specific antileukemic treatments that are less toxic but as effective as the best standard regimens.


Down's syndrome is caused by a trisomy in the chrosomome 21. It is the most common postnatally viable chromosomal anomaly having an incidence of about 1:700 live births. One of the several disturbances seen in these patients is a bone marrow dysfunction. Among the diverse hematological abnormalities, leukemia development is the most serious one. In general, children and adolescents with Down's syndrome have a 10 to 30-fold increased incidence of leukemia (Robison et al. 1984). Every 150th such child will eventually develop leukemia. The increased risk extends into adulthood and is also probable among siblings. About 2% of pediatric leukemia patients have Down's syndrome.


Although the association between Down's syndrome and leukemia has been known for nearly 70 years, the etiology and underlying mechanisms remain poorly understood. Several considerations have been presented, including developmental error with disruption of hemostasis, ineffective regulation of granulopoiesis, immune deficiency leading to decreased immune surveillance, abnormal cell kinetics, susceptibility to viral transformation, genetic predisposition to nondisjunction, increased chromosomal fragility, impaired DNA repair mechanisms or oncogene activation (see Fong and Brodeur 1987), and increased sensitivity to interferon (Zihni 1994).

The most promising areas of research are those relating to the cytogenetics and molecular genetics. Chromosome 21 contains approximately 750 genes that are present in a triple dose in each cell of most patients with Down's syndrome. This extra material may result in generalized disruption of genetic balance and consequently lead to an altered response to normal genetic and environmental factors. This genetic inbalance has been suggested to be directly involved. The hypothesis has been supported by the following observations. Firstly, in mosaics with Down's syndrome only the trisomic clone undergoes the leukemic transformation (Ferster et al., 1986). Secondly, in constitutionally normal children with acute leukemia, trisomy 21 is the most common acquired cytogenetic change (Watson et al., 1993). However, whether the trisomy 21 can be considered as the first genetic hit has been debated. It seems unlikely that trisomy 21 per se could be sufficient to cause leukemia, since only a fraction of patients with Down's syndrome develop leukemia.

Investigations on the origin of extra chromosome 21 are of particular interest. Trisomy can be consequent to nondisjunctional errors at the first or second stage of maternal or paternal meiosis. Down's syndrome patients may or may not have two identical chromosome 21 deriving from one parent. Genetic mapping studies indicate that the disomic homozygosity can be a feature of trisomy 21 associated with leukemia, at least in the development of transient leukemia and acute megakaryoblastic leukemia (Feingold at al., 1995, Shen et al., 1995). It has also been suggested a preferential parental origin of trisomy in leukemia patients with Down's syndrome. This view, however, has not been supported by studies in which techniques involving DNA polymorphism have been used (Lorber et al., 1992, Shen et al., 1995).


Leukemia in children with Down's syndrome is usually acute and lymphoblastic (ALL) (Levitt et al. 1990) although a higher-than-normal proportion of non-lymphoblastic leukemia (ANLL) has been reported (Zipursky et al. 1992). The latter type almost exclusively presents before the age of 4 years.

ALL in Down's syndrome appears to be similar to that in other children. However, a study from the Children's Cancer Study Group noted a higher hemoglobin and a lower platelet count at diagnosis in patients with Down's syndrome (Robison et al., 1984). The patients have been reported to have fewer extramedullary leukemia manifestations, and their lymphoblasts less often have translocations associated with poor prognosis (Pui et al., 1993, Ragab et al., 1991). Accordingly, the CALL antigen immunophenotype was more prevalent in these patients.

Most ANLL cases of Downs' syndrome have acute megakaryoblastic leukemia, a very rare type of leukemia in other children (Kojima et al., 1990, Zipursky et al., 1992). This malignancy is often preceded by a preleukemic phase of a few months. If a patient with Down's syndrome has myelodysplasia, the condition will ultimately proceed to megakaryoblastic leukemia. Furthermore, newborns with Down's syndrome frequently develop transient leukemia during their first 1-3 months of life. While majority of such infants will show spontaneous and persistent remission, 20-30% will develop acute megakaryoblastic leukemia. The megakaryoblasts have features of early erythroid precursors, and a higher-than-normal incidence of erythroleukemia occurs in patients with Down's syndrome. It is likely that in transient leukemia, myelodysplasia and megakaryoblastic leukemia of Down's syndrome, the leukemic progenitor cells are able to differentiate into cells of megakaryocytic, mast cell and erythroid lineage, a phenomenon that is unique to Down's syndrome (Zipursky et al., 1995). Megakaryoblasts have an unusual distribution of clonal cytogenetic alterations (Iselius et al., 1990) suggesting that the molecular and cellular basis for this acute megakaryoblastic leukemia may differ from that occuring in similar patients without Down's syndrome.


In the past, leukemia of patients with Down's syndrome was often treated suboptimally, and many patients did not receive cytoreductive treatments at all. Their several physical abnormalities, including potentially life-threatening cardiac and intestinal malformations along with mental retardation and associated psychosocial issues, high susceptibility to infections as well as increased toxicity of chemotherapy may have prejudiced physicians against the use of a standard chemotherapy and led families to accept less aggressive therapeutic programmes.

The poor tolerance of cytoreductive regimens in patients with Down's syndrome is well documented. Their fibroblasts and lymphocytes have increased chromosomal sensitivity to mutagenic agents and abnormal DNA repair (see Schwaiger et al., 1989). The patients therefore are predisposed to more severe toxicity of irradiation or alkylating agents. The most prominent finding is the unusually high susceptibility to methotrexate toxicities such as moderate-to-severe mucositis, dermatitis and myeloid suppression (Garre et al., 1987, Kalwinsky et al., 1990). This may derive from gene dosage effects on chromosome 21, to which three enzymes of purine synthesis are mapped. It is postulated that elevated purine synthesis confers a higher demand for tetrahydrofolates, and thus greater sensitivity to antifolate agents. The in vitro methotrexate toxicity can be diminished but not abolished by in vivo administration of supplemental high doses of folic acid (Peeters et al., 1995). There is also evidence of altered pharmacokinetics of methotrexate: the clearance of methotrexate is significantly slower in patients with Down's syndrome (Garre et al., 1987).

Despite these obstacles, modern intensive schedules have recently been used to treat patients with Down's syndrome. With improved supportive care, these regimens have been acceptably tolerated (Pui et al., 1993, Ragab et al., 1991). In patients with ANLL, preexisting congenital heart disease did not appear to predispose the patients to anthracycline cardiac toxicity (Ravindranath et al., 1992). Major neurotoxicity after high dose ARA-C was not seen. Several patients have even had allogeneic bone marrow transplantations (Arenson and Forde 1989). The problems of marrow transplantation appear to mirror those observed with conventional therapy.


The life expectancy of children with Down's syndrome has improved with more effective treatment of congenital heart disease and respiratory infections. Thus, more children with Down's syndrome are now surviving to develope leukemia. Three decades ago, most of such leukemia patients usually died early after diagnosis. While treatment results of acute leukemia in the general pediatric population improved greatly in the 1970s they remained poor in Down's syndrome for a long time. The most striking feature was their significantly higher rate of induction failures, especially in ALL (Robison et al., 1984). Among British patients diagnosed between 1971 and 1986, the overall actuarial survival at 5 years after ALL diagnosis was only 28% in patients with Down's syndrome when it was 59% in other patients (Levitt et al., 1990). The poorer survival was partly due to the higher rate of induction deaths, since no difference was apparent in disease-free survival for patients achieving an initial remission. Nearly identical figures were then published from St. Jude Children's Research Hospital (Kalwinsky et al., 1990). However, part of the inferior outcome might have been due to inadequate treatment. Indeed, in a later Pediatric Oncology Group study, the patients who were treated with intensive regimens and appropriate supportive care no longer displayed the differences seen in the patients treated with conventional therapy (Ragab et al., 1991). In acute megakaryoblastic leukemia, the prognosis in Down's syndrome may even be better. Event-free survival at 4 years in one study was nearly 100% (Ravindranath et al., 1992). A history of prior myelodysplastic syndrome had no adverse effect on the outcome.


Arenson EB, Forbe MD. Bone marrow transplantation for acute leukemia and Down syndrome: report of a successful case and results of a national survey. J Pediatr 114, 69-72, 1989.

Feingold E, Lamb NE, Sherman SL. Methods for genetic linkage analysis using trisomies. Am J Hum Genet 56, 475-483, 1995.

Ferster A, Verhest A, Vamos E et al. Leukemia in a trisomy 21 mosaic: specific involvement of the trisomic cells. Cancer Genet Cytogenet 20, 109-113, 1986.

Fong C, Brodeur GM. Down's syndrome and leukemia: epidemiology, genetics, cytogenetics and mechanisms of leukemogenesis. Cancer Genet Cytogenet 28, 55-76, 1987.

Garre ML, Relling MV, Kalwinsky D et al. Pharmacokinetics and toxicity of methotrexate in children with Down syndrome and acute lymphocytic leukemia. J Pediatr 111, 606-612, 1987.

Iselius L, Jacobs P, Morton N. Leukemia and transient leukemia in Down syndrome. Hum Genet 85, 477-485, 1990.

Kalwinsky DK, Raimondi SC, Bunin NJ et al. Clinical and biological characteristics of acute lymphocytic leukemia in children with Down syndrome. Am J Med Genet Suppl 7, 267-271, 1990.

Kojima S, Matsuyama T, Sato T et al. Down's syndrome and acute leukemia in children: an analysis of phenotype by use of monoclonal antibodies and electron microscopic platelet peroxidase reaction. Blood 76, 2348-2353, 1990.

Levitt GA, Stiller CA, Chessells JM. Prognosis of Down's syndrome with acute leukemia. Arch Dis Childh 65, 212-216, 1990.

Lorber BJ, Freeman SB, Hassold T et al. Characterization and molecular analysis of nondisjunction in 18 cases of trisomy 21 and leukemia. Genes Chromosom Cancer 4, 222-227, 1992.

Peeters MA, Rethore MO, Lejeune J. In vivo folic acid supplementation partially corrects in vitro methotrexate toxicity in patients with Down syndrome. Br J Haematol 86, 678-680, 1995.

Pui C-H, Raimondi SC, Borowitz MJ et al. Immunophenotypes and karyotypes of leukemic cells in children with Down syndrome and acute lymphoblastic leukemia. J Clin Oncol 11, 1361-1367, 1993.

Ragab AH, Abdel-Mageed A, Shuster JJ et al. Clinical characteristics and treatment outcome of children with acute lymphoblastic leukemia and Down's syndrome: a POG study. Cancer 67, 1057-1063, 1991.

Ravindranath Y, Abella E, Krischer JP et al. Acute myeloid leukemia (AML) in Down's syndrome is highly responsive to chemotherapy: experience of Pediatric Oncology Group AML Study 8498. Blood 80, 2210-2214, 1992.

Robison LL, Nesbit ME, Sather HN et al. Down syndrome and acute leukemia in children: a 10-year retrospective survey from Childrens Cancer Study Group. J Pediatr 105, 235-42, 1984.

Schwaiger H, Weirich HG, Brunner P et al. Radiation sensitivity of Down's syndrome fibroblasts might be due to overexpressed Cu/Zn-superoxidase dismutase (EC Eur J Cell Biol 48, 79-87, 1989.

Shen JJ, Williams BJ, Zipursky A et al. Cytogenetic and molecular studies of Down syndrome individuals with leukemia. Am J Hum Genet 56, 915-925, 1995.

Watson MS, Carroll AJ, Shuster JJ et al. Trisomy 21 in childhood acute lymphoblastic leukemia: a Pediatric Oncology Group Study (8602). Blood 82, 3098-3102, 1993.

Zihni L. Down's syndrome, interferon sensitivity and the development of leukaemia. Leukem Res 18, 1-6, 1994.

Zipursky A, Christensen H, De Harven E. Ultrastructural studies of the megakaryoblastic leukemia of Down's syndrome. Leukem Lymph 18, 341-347, 1995.

Zipursky A, Poon A, Doyle J. Leukemia in Down syndrome: A review. Pediatr Hematol Oncol 9, 139-149, 1992.