ALL in the Family

 

The genetic and environmental causes of a particular cancer can sometimes be informed by the study of cancers in family members [see C3 Vol 14, No. 4]. A few small studies have examined cancer in the relatives of children with acute lymphoblastic leukemia (ALL) with inconsistent results. A new analysis [Couto E et al. Br J Cancer 2005; 93: 1307-1309] takes a look at the subject using the very large Swedish Family-Cancer Database, which tracks relatedness among 6,994,345 Swedes and links to the national cancer registry. The occurrence of leukemia and 34 other types of cancer was ascertained in the mothers, fathers, and siblings of 1,925 children with ALL. Poisson regression was used to estimate the standardized incidence ratios (SIRs) comparing the rate of ALL in children with and without a positive family history of cancer. There was a significant excess of ALL among the 19 children with at least one first-degree relative with any type of leukemia (SIR = 2.21; 95% confidence interval (CI): 1.33-3.45), which was explained by the expected marked increase in risk of ALL among children whose twin was affected (SIR = 208.28; 95% CI: 54.05-641.09). Neither having a parent (SIR = 1.27; 95% CI: 0.55-2.50) nor a non-twin sibling (SIR = 2.22; CI: 0.51-7.32) with leukemia were significantly associated with ALL. Excluding twins, the SIRs were 2.18 (95% CI: 0.94-4.30) and 1.25 (95% CI: 0.34-3.20), respectively, for having a first-degree relative with ALL or acute myeloid leukemia (AML). Considering other cancer diagnoses among first-degree relatives, risk of ALL was significantly increased only among children whose father was diagnosed with testicular cancer (SIR = 3.12; 95% CI: 1.50-5.75) in general and specifically with testicular teratoma (SIR = 4.10; 95% CI: 1.29-9.64).

 

COMMENT: The many-branched Scandinavian registries continue to bear fruit. This study confirms that the risk of ALL is vastly higher in twins. Rather than being evidence for a genetic predispostion, this finding reinforces the hypothesis that leukemia initiates in utero and spreads from one monozygotic twin to the other via a shared placenta [see Greaves M et al. Blood 2003; 102: 2321-33]. Zygosity of the affected twin pairs could not be assessed in the present study but all were female, which is consistent with identical twinning. Interestingly, there appeared to be an increased risk of ALL in children with a first-degree relative who also had ALL, which may point to weak familial aggregation. It is possible that additional years of follow-up could solidify the finding, which was based on eight children and not significant. The authors speculate that the increase in risk of ALL among children of men with malignant teratoma could be related to abnormal spermatogenesis following cisplatin chemotherapy. Only three of five men in this study conceived their child after therapy, however, and cisplatin is used to treat a number of other cancers, none of which were associated with a higher risk of ALL among offspring. The finding nevertheless deserves follow-up since testicular cancer incidence is highest during the reproductive years. Lastly, on a practical note, the large numbers of subjects required to produce these results suggests that small case-control studies are inadequate to assess the risk of ALL associated with family history of cancer.

Logan G. Spector

 

NQ01 and infant leukemia: run of the MLL?

 

NAD(P)H:quinine (NQ01) is a multifunctional enzyme whose properties include detoxifying benzene and flavonoids. These substances interfere with the function of DNA topoisomerase II, the inhibition of which is thought to increase the risk of infant leukemia presenting with MLL gene rearrangements [Ross JA et al. JNCI 1994; 86: 1678-80]. Thus, the level of NQ01 activity may also affect the risk of MLL+ infant leukemia. The C609T and C465T single nucleotide polymorphisms (SNPs) result in amino acid changes from Pro to Ser and Arg to Trp, respectively. The T alleles of both SNPs, which have been associated with diminished NQ01 activity, are the subject of a new study from Japan [Eguchi-Ishimae M. et. al. Haematologica 2005; 90: 1511-25]. The investigators collected DNA from 103 infant leukemia cases (< 18 months), or about 80% of those diagnosed in all of Japan in 1995-1998. The NQ01 allele frequency in cases was compared to that in 185 cord blood controls using odds ratios (ORs) and 95% confidence intervals (CIs). There was a near significant association of MLL+ acute myeloid leukemia (AML) with having at least one copy of the 609T allele (OR = 3.23; 95% CI: 0.88-11.80). In contrast, there was no association of 609T with other types of MLL+ infant leukemia or with MLL- infant leukemia. The 465T allele was associated with MLL+ infant leukemia (OR = 2.63; 0.85-8.14), MLL+ acute lymphoblastic leukemia (ALL) (OR = 3.55; 95% CI: 1.13-11.10), and infant leukemia with the MLL-AF4 fusion gene (OR = 6.36; 95% CI: 1.84-21.90). MLL+ AML and MLL- infant leukemias were not associated with the C465T SNP.

 

COMMENT: Two studies that found a sharply increased risk of MLL-AF4 infant leukemia with the C609T allele also found no association of infant leukemia with C465T [Wiemels JL et. al. Cancer Res 1999; 59: 4095-9; Smith MT et. al. Blood 2002; 100: 4590-3]. Two other studies did not confirm that C609T raises risk of infant leukemia and did not test C465T [Lanciotti M et. al. Leukemia 2005; 19:214-6; Kracht T et al. Haematologica 2004; 89: 1492-7]. The present study does little to clear up the issue of which, if any, polymorphisms affect the risk of infant leukemia, but the strikingly strong ORs found with small numbers of cases suggest that NQ01 is somehow involved. Of course, genes are only half the story and a large study is needed that also takes into account environmental exposures. The Children's Oncology Group is currently conducting such a study, which promises to be the largest molecular epidemiology study ever done to evaluate this intriguing research question. It is expected that by the end of 2007, maternal interviews and DNA samples (mother and infant) from over 400 cases (diagnosed < 12 months of age) and 300 controls will be available for analysis. Importantly, this study includes information regarding MLL status, and thus should be able to contribute substantially to this literature.

Logan G. Spector

 

Brain Tumors – Getting to the Bottom of It

 

Recent evidence suggests that many cancers arise from and are maintained by a rare fraction of transformed stem cells termed cancer stem cells. This paradigm is analogous to the way in which hematopoiesis is maintained by hematopoietic stem cells, capable of both self-renewal and the production of more differentiated cells. Typically, cancer stem cells are phenotypically similar to the normal stem cells of the corresponding tissue of origin, but they exhibit dysfunctional patterns of self-renewal and differentiation. In previous work, astrocytic and embryonal brain tumors have been shown to contain rare populations of cancer stem cells that are capable of self-renewal in vitro and of propagating the original tumor in vivo. These stem cell like properties are restricted to tumor cells that express the neural stem cell marker CD-133 [Uchida N et al, Proc. Natl Acad Sci. USA 2000; 97:14720-14725]. In new work, Taylor MD et al [Cancer Cell 2005; 8:323-335] have examined cancer stem cells in ependymomas. In these investigations of primary human tumors, the authors report that histologically identical but genetically distinct ependymomas exhibit patterns of gene expression that recapitulate to those of radial glia cells in the corresponding region of the central nervous system. Cancer stem cells isolated from ependymomas displayed a radial glia phenotype and form tumors when orthotopically transplanted in mice. These findings identify restrictive populations of radial glia cells as candidate stem cells of ependymoma and support a general hypothesis that different populations of progenitor cells in the tissue of origin can generate tumors that appear to be of the same histologic tumor type. These data, if verified by others, could lead to an important change in the way we approach treating CNS tumors. Rather than employing non-specific therapies guided by crude histologic measures of disease type to treat the primary tumor mass, treatment strategies would focus on the eradication of the cancer stem cells.

COMMENT: The concept of cancer stem cells is an important and interesting one. It has been applied successfully to solid tumors such as breast cancer and brain tumors and work is ongoing to identify similar populations in other malignancies.

Stella M Davies

 

Lymphoblastic leukemia has stem cells too!

 

The concept of a leukemic stem cell, a self-renewing progenitor that maintains the leukemia, is easy to grasp, as we are very familiar with the concept of a hematopoietic stem cell. In a new investigation, Castor A et al [Nature Medicine 2005; 11: 630-637) show that clinically and genetically different subtypes of acute lymphoblastic leukemia (ALL) originate in leukemic stem cells arising at distinct stages of hematopoietic development. These investigations were conducted using exhaustive and detailed flow cytometry, in vivo culture, and mouse transplantation experiments using primary human cells. The leukemic stem cell carrying a TEL-AML1 fusion arises in committed B-cell progenitors. In contrast, major breakpoint BCR-ABL fusions (encoding p210 BCR-ABL) originated in hematopoietic stem cells, whereas minor BCR-ABL fusions (encoding p190 BCR-ABL) had a B-cell progenitor origin. These data suggest that p190 and p210 BCR-ABL ALLs represent distinct tumor biological and clinical entities. Despite these different cells of origin, the leukemia cells in both the p190 and p210 BCR-ABL ALLs had a committed B progenitor phenotype. In all patients, normal and leukemic repopulating stem cells could be successfully separated. Notably, the size of the normal hematopoietic stem cell compartment in TEL-AML1 and p190 BCR-ABL ALLs was unaffected despite the expansive leukemic stem cell compartment. The stem cell compartment was reduced in p210 BCR-ABL leukemias. This observation is different from what is seen in myelodysplastic syndrome where normal hematopoietic stem cells are efficiently out-competed by the malignant clone [Nilsson L et al, Blood 1000; 96: 2012-2021; Wilson L, Blood 2002; 100:259-267].

 

COMMENT: As in the brain stem tumor study discussed above, identification of leukemic stem cells allows us to identify efficiently the appropriate target for eradication. The evidence of the extent of differentiation that takes place in p210 BCR-ABL ALL cells is particularly fascinating.

Stella M Davies

 

C3 Quarterly Newsletter
Children's Cancer Research Fund
Epidemiology Research Unit
Division of Pediatric Epidemiology
Clinical Research
University of Minnesota
420 Delaware St. SE, Box 422
Minneapolis, MN 55455
pedsepi@umn.edu

Editors: 
Stella M. Davies, MD, PhD, and Julie A. Ross, PhD