Prenatal diagnosis of acute lymphoblastic leukaemia

Abstract

Author's replySir—Nicolas Janin suggests that our interpretation of the modest (-50%) concordance rate of acute lymphoblastic leukaemia (ALL) in identical twins might be incorrect. An alternative view is that the fetal clone initiated in one twin by TEL-AML1 gene fusion might engraft the co-twin in only 5% of such twin pairs but have a 100% probability of transformation to overt malignancy. This is indeed a formal possibility but is, we believe, unlikely. First, TEL-AML1 fusion by itself appears to be insufficient for leukaemogenesis as judged by the lack of leukaemia in mice transgenic for TEL-AML1 (A M Ford and M F Greaves, unpublished observations; B Young and O Bernard, personal communication). Second, epidemiological evidence, albeit incomplete, suggests that postnatal exposures, probably of an infectious nature, are involved and necessary for the development of ALL.1Greaves MF Aetiology of acute leukaemia..Lancet. 1997; 349: 344-349Summary Full Text Full Text PDF PubMed Scopus (380) Google Scholar, 2Little J Epidemiology of childhood cancer. IARC Scientific Publications, Lyon1999Google ScholarWe agree with Janin that formal proof of our preferred interpretation would be helpful and could come from the observation of TEL-AML1 positive cells in the blood or marrow of a healthy monozygotic co-twin of a patient with ALL. We are already attempting to do this experiment, which poses several ethical and logistic difficulties. Our model of the natural history of childhood common ALL3Greaves M Molecular genetics, natural history and the demise of childhood leukaemia..Eur J Cancer. 1999; 35: 173-185Summary Full Text Full Text PDF PubMed Scopus (83) Google Scholar predicts that we will find low level but PCR-detectable TEL-AML1 clone-specific sequences in the circulation of healthy co-twins, but only in most of the 60% or so that have monochorionic placentas.4Strong SJ Corney G The placenta in twin pregnancy. Pergamon, Oxford1967Google Scholar Another important prediction of our model is that considerably more children should be born with a TEL-AML1 fusion gene in lymphoid cells than ever develop ALL with the TEL-AML1 gene fusion (about one in 12 500). This we are attempting to confirm by reverse transcriptase PCR screening of unselected normal newborn cord blood. Author's reply Sir—Nicolas Janin suggests that our interpretation of the modest (-50%) concordance rate of acute lymphoblastic leukaemia (ALL) in identical twins might be incorrect. An alternative view is that the fetal clone initiated in one twin by TEL-AML1 gene fusion might engraft the co-twin in only 5% of such twin pairs but have a 100% probability of transformation to overt malignancy. This is indeed a formal possibility but is, we believe, unlikely. First, TEL-AML1 fusion by itself appears to be insufficient for leukaemogenesis as judged by the lack of leukaemia in mice transgenic for TEL-AML1 (A M Ford and M F Greaves, unpublished observations; B Young and O Bernard, personal communication). Second, epidemiological evidence, albeit incomplete, suggests that postnatal exposures, probably of an infectious nature, are involved and necessary for the development of ALL.1Greaves MF Aetiology of acute leukaemia..Lancet. 1997; 349: 344-349Summary Full Text Full Text PDF PubMed Scopus (380) Google Scholar, 2Little J Epidemiology of childhood cancer. IARC Scientific Publications, Lyon1999Google Scholar We agree with Janin that formal proof of our preferred interpretation would be helpful and could come from the observation of TEL-AML1 positive cells in the blood or marrow of a healthy monozygotic co-twin of a patient with ALL. We are already attempting to do this experiment, which poses several ethical and logistic difficulties. Our model of the natural history of childhood common ALL3Greaves M Molecular genetics, natural history and the demise of childhood leukaemia..Eur J Cancer. 1999; 35: 173-185Summary Full Text Full Text PDF PubMed Scopus (83) Google Scholar predicts that we will find low level but PCR-detectable TEL-AML1 clone-specific sequences in the circulation of healthy co-twins, but only in most of the 60% or so that have monochorionic placentas.4Strong SJ Corney G The placenta in twin pregnancy. Pergamon, Oxford1967Google Scholar Another important prediction of our model is that considerably more children should be born with a TEL-AML1 fusion gene in lymphoid cells than ever develop ALL with the TEL-AML1 gene fusion (about one in 12 500). This we are attempting to confirm by reverse transcriptase PCR screening of unselected normal newborn cord blood. Prenatal diagnosis of acute lymphoblastic leukaemiaJ L Wiemels and colleagues (Oct 30, p 1499)1 have shown that it is likely that childhood acute lymphoblastic leukaemia (ALL) starts in utero in most cases. Together with the known concordance rate of childhood ALL in monozygotic twins,2 which is probably around 5%, the investigators also conclude that the in-utero acquisition of t(12;21)(p13;q22) is in most cases insufficient to trigger an irreversible process of ALL leukaemogenesis. This conclusion, based on the known complexity of the somatic disease underlying carcinogenesis, is credible. Full-Text PDF