Leukemia
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Archive
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Subscribe
Advertising
naturereprints
Contact Springer Nature
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
 
Journal home
Advance online publication
Current issue
Archive
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Subscribe
Advertising
naturereprints
Contact Springer Nature
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
 
April 2000, Volume 14, Number 4, Pages 722-726
Table of contents    Previous  Article  Next   [PDF]
Original Manuscript
The incidence of acute promyelocytic leukemia appears constant over most of a human lifespan, implying only one rate limiting mutation
M Vickers1, G Jackson2 and P Taylor2

1Department of Haematology, Medicine and Therapeutics, Aberdeen, UK

2Department of Haematology, Royal Victoria Infirmary, Newcastle, UK

Correspondence to: M Vickers, Department of Haematology, Medicine and Therapeutics, Polwarth Building, Foresterhill, Aberdeen, UK; Fax: 01224 699884

Abstract

It is believed that most malignancies become more common with increasing age due to the requirement for several mutations to accumulate and subsequently interact. The age specific incidence of acute promyelocytic leukemia (APL) was investigated using population-based data from 77 million subject years of observation, yielding 149 consecutive cases. The incidence appears approximately constant with respect to age, an observation not previously reported with any other malignancy. These findings are most easily explained by there being only one rate limiting genetic event required to initiate the disease, although other, non-rate limiting mutations may also be necessary for disease development. It is also argued that this mutation is probably restricted to cells committed to differentiation, which may explain why APL is curable by chemotherapy. Leukemia (2000) 14, 722-726.

Keywords

acute promyelocytic leukemia; epidemiology; population genetics; translocation

Introduction

The pathogenesis of malignant transformation is generally believed to be a multistep process involving a series of somatic mutations which must interact to produce a malignant phenotype;1 it is this requirement for several mutations to interact that is thought to underlie the increase in malignancy rate with age. How many rate limiting mutations are generally required is currently uncertain, this question being difficult to address experimentally.2 Analysis of epidemiologic data can provide insights into the problem; notably, the two hit model for retinoblastoma.3 The incidence of most malignancies rises as the fourth to fifth power of age. This observation has been interpreted as indicating five or six rate limiting mutations,1,4 although the underlying reasoning is open to doubt, due to the mutation rate being too low and uncertainty about clonal evolution.5,6

Acute promyelocytic leukemia (APL), a sub-type of acute myeloid leukemia, is an unusual malignancy. It is characterized by the accumulation of immature promyelocytes,7 which cause fibrinolysis and hemostatic failure.8 Cytogenetically, it is usually associated with a pathognomonic translocation, t(15;17), juxtaposing the gene for the retinoic acid alpha receptor (RARalpha) with the PML gene.9 It is the only malignancy for which differentiation therapy is a well established part of therapy.10 The epidemiology of acute promyelocytic leukemia (APL) is also unusual in that the average age of presentation is relatively low but both young and old are commonly affected. This paper reports a formal analysis of the age dependence of APL.

In considering the implications of the age dependence that we describe, it is assumed that the mutation rate is constant. Evidence for this view comes from many sources. The age of the organism from which cells are obtained has not been found to be an important determinant of their mutation rate. Susceptibility to mutagen-induced carcinogenesis is usually independent of age, both in experimental animals and humans.11 Finally, the accumulation of bone marrow mutations in humans can be accounted for by the underlying kinetics of cell division, with a linear increase with age and little evidence for any variation in the underlying mutation rate.12 We therefore consider the assumption of a constant mutation rate to be reasonable, although not proven.

Subjects and methods

One hundred and fifty-nine cases of APL were ascertained from four British Regional Leukemia Registries. Importantly for a study of this type, such Registries collect cases from a defined population without bias with respect to age, unlike trials. The data on each Register are maintained by interested hematologists. Cases were ascertained from: Merseyside (January 1987-May 1994); Northern (January 1988-July 1998); North-West (January 1992-May 1994); Oxford (January 1989-May 1995). In addition to prospective reporting, every hematologist in the regions was contacted and asked about any other cases. Each case was confirmed by discussion with the clinicians and regional cytogenetics laboratories.

Accuracy of case ascertainment

The diagnosis of APL was initially made on morphological grounds by the local hematologists; diagnostic slides were reviewed centrally in each region and only after confirmation of FAB M3 were they accepted into the study. Cytogenetic analysis was attempted in all but 21 cases. In a further nine cases the analyses failed and in seven patients a normal karyotype was found. Fluorescent in situ hybridization (FISH) analysis was used in three of these seven and the PML-RARalpha translocation detected in two cases. Detection of the translocation by PCR was not routinely performed. On only one patient was the t(11;17) translocation detected. In all other patients the typical t(15;17) translocation was found. In three regions (118 cases), it was possible to obtain accurate information on additional abnormalities. In 14/118 cases (12%), t(15;17) was seen in the context of more complex abnormalities. The commonest was trisomy 8 (+8) as the sole additional abnormality (10/118; 9% cases).

The sizes and age structure of the relatively well characterized and stable populations were obtained from the Office of Population Censuses and Surveys; local and health authority areas (1991). The number of individuals with any given age was assumed to be constant within each of the given age categories. For the final age category, 50% of the over 75 category was assigned to the 75-80 years of age and 50% to 80-90 years of age.

Analysis and results

We start by discussing a simplistic analysis of the expected age-specific incidence if one, two or more independent rate limiting mutations were required for the development of APL.

Suppose that APL were caused by a single mutation. If the mutation rate is constant throughout life, it can be intuitively seen that the incidence at any age, I, would not be expected to vary with respect or age, or

I = k1.age0 = k1, (1)

where k1 is a rate constant.

Suppose that a malignancy were the result of two independent mutations and once a mutation occurred, it could persist for the rest of a lifetime. The cumulative probability of two mutations occurring in the same cell is then the product of the cumulative probabilities of the two individual mutations and the risk at any age is the first derivative. So,

I = k2.age1, (2)

where k2 is a rate constant.

In the general case, if the number of required mutations is n then

I = k3.agen - 1. (3)

A further complication arises from the assumption that a cancer presents as soon as the final mutation takes place, which is implausible. If the mean latent period is defined as w, then equation (3) becomes

I = k4.(age + w)n - 1. (4)

Indeed, it has been observed that the incidences of most cancers follow a fixed power of age, with an exponent of four to five. This observation has been widely interpreted to mean that five or six rate limiting mutations are necessary. In fact, this interpretation is almost certainly incorrect as the observed mutation rate is many orders of magnitude too low to allow such a number of independent mutations to accumulate by chance. Instead, it seems likely that an intermediate number of mutations effect a phenotypic change which causes the accumulation of subsequent interacting mutations to be more likely and therefore not independent. Detailed discussion of this problem is beyond the scope of this paper.5 For present purposes, it should be noted that if more than one rate limiting mutation causing APL is postulated, the incidence should rise with age. If as few as two rate limiting mutations are necessary, the incidence should rise linearly with age, ie the exponent n in equation (4) would be one. If the first mutation were to make the second mutation progressively more likely, then the exponent n in equation (4) would be greater than 1. Conversely, if two mutations were required and the observed exponent was less than one then additional mechanisms must be postulated to account for the discrepancy. Such mechanisms might include the first mutation effecting a phenotypic change so that the probability of the second mutation becomes progressively lower, or the size of the APL precursor cell population decreasing progressively with age. At present, neither the nature nor the number of APL precursor cells are known, let alone how that number changes with age. While there is some evidence that some compartments decrease in number with age, other compartments may increase.13

In order to measure the age specific incidence of APL, we collected all diagnosed cases from 77 million subject years of observation, yielding 149 cases. The data were analysed using an extension of equation (4),

2401722e1.gif

where Ic is the cumulative number of cases at any age, Page the size of the population at each age, k a rate constant, t the age when the final rate limiting somatic mutation took place and w the latent period.

The optimum values of the rate constant, k, and latent period, w, were calculated for a variety of values of postulated exponent, n, between -1 and 1 by minimising the difference between observed and predicted values by least squares analysis. The minimum difference obtained was plotted against postulated n (Figure 1), which yields the most likely parameter values as n = 0.09, w = 11.0 years and k = 3.14 per million individuals per year (Figure 2). It is concluded that the age-specific incidence of APL is approximately constant over most of the age range.

An alternative approach uses a Poisson heterogeneity test, which tests the hypothesis that the number observed during each time period does not deviate from the number expected for the population at that age. The statistic

2401722e2.gif

on the null hypothesis of a constant rate, approximately follows a X2 distribution. Considering intervals of 5 years, the P value for ages above 10 is 0.2 and so does not differ significantly from that expected by chance alone. Incorporation of childhood values decreases the P value to just less than 0.05. It is concluded that the distribution of values does not differ significantly from that predicted by a constant rate during adult life. In contrast, if the expected incidence is chosen on the basis of two rate limiting steps as predicted by equation (2), the P value is <0.001, implying that the observed values are highly unlikely to be predicted by equation (2). It is concluded that a multiple rate limiting hit model of APL cannot easily account for the observations presented here without postulating other mechanisms that would alter the age-specific incidence as discussed above.

Discussion

We have presented two statistical analyses, both of which support the hypothesis that the incidence of APL is approximately constant over most of the age range. While more sophisticated analyses might be applied, we feel that the major uncertainties in this analysis are not statistical but related to the raw data as discussed above and the interpretation, as discussed below.

The relatively low rate over the age of 70 may be explained by worse case ascertainment. It appears that the hemorrhagic tendency caused by APL is particularly prone to cause early death in elderly individuals14 and so it is plausible that APL might present with, for example, fatal stroke before other clinical manifestations allow accurate diagnosis. Other explanations for the low rate in old age include the number of APL progenitor cells falling along with general bone marrow cellularity.

In this survey, a relatively low rate is seen during childhood. This was 'explained' in the mathematical analysis by the latent period. However, other explanations are possible. For instance, it is plausible that the number of APL progenitor cells is lower during childhood. If this were the case, then the mean latent period would be correspondingly lower.

If it is accepted that the incidence of APL is constant during adult life, can any conclusions be drawn concerning the biological basis of the disease? The most obvious implication is that it is difficult to reconcile a conventional multi-hit model of carcinogenesis with the data. If rate limiting mutations are viewed as gradually accumulating throughout life and several are required to interact to cause cancer, then the incidence should increase with age unless some other mechanism, apparently peculiar to APL, and possibly other myeloid translocations, is postulated that makes the accumulation or interaction of mutations progressively less common.

The simplest explanation of the data is that APL is caused by a single rate limiting mutation, where the distinction between 'rate limiting' and 'necessary' mutations is important. To illustrate the difference, consider a mutation 'A', that prevented apoptosis or differentiation but allowed cell division at a normal rate; each cell division would then result in two progeny and the resulting clone would grow exponentially. Suppose that a single further mutation, 'B', was also required to cause clinical presentation as a malignancy. By the time the clone resulting from the first mutation had undergone only 24 division cycles, the number of cells would reach 107 cells and, if the mutation rate for mutation B were a typical 10-7 per division, each cycle of divisions would result in a mean of 1 new B mutation in that clone. Mutation A would then be the only rate limiting mutation. Mutation B is still necessary for clinical presentation but its main kinetic contribution would be towards the latent period. The approach outlined here analyzes primarily the number of rate limiting mutations. There might be either zero or several other necessary but non-rate limiting mutations, which the approach presented here cannot distinguish between. This approach also cannot shed light on which mutations are involved. It is clearly simplest to equate the putative single event with the translocations involving the retinoic acid receptor but this is not necessarily the case. For instance, the rate limiting mutation might make the observed translocations so likely that their occurrence is no longer rate limiting.

One further, more complex explanation of our data should also be considered. More than one rate limiting mutation might be required but those mutations might both occur in cells committed by differentiation. In this scenario, both mutations would be occurring on many occasions in differentiating cells, with the time available for interaction with other mutations limited by the lifespan of the differentiating cells. The occurrence of both mutations in a single cell would then be expected to follow zero order kinetics, as observed. Although possible, we consider this to be an unlikely explanation as the spontaneous translocation rate is probably too low. Measurements of random translocation frequency in lymphocytes,15 combined with estimates of lymphocyte turnover,16 yield an approximate estimate of translocation rate of 10-12 per base pair per division. A translocation involving both APL and RARalpha loci would then occur at a rate of about 10-18 per division. This figure gives an annual incidence reasonably close to that observed if it is estimated that 1% of marrow mitoses occur before the promyelocytic stage and are at risk of causing APL. The requirement of a second mutation would decrease the expected annual incidence by about 10-7, too far below that observed. Of course, this calculation involves many assumptions and approximations so that the requirement for two or even more mutations cannot be rigorously excluded. In this context, it would be valuable to have an estimate of the PML-RARalpha mutation rate based on direct experimental evidence. In conclusion, more than one rate limiting mutation is possible if all mutations were confined to cells committed to differentiation but observed translocation rates are probably too low to sustain more than one rate limiting mutation.

The best fit latent period was 11 years, which might be considered a long time if other mutations were not required. Experimental evidence concerning whether other mutations are required is presently inconclusive. The most pertinent evidence in this regard arises from transgenic experiments where the introduction of PML-RARalpha fusion genes generally result in acute leukemias but with variable latent periods of many months. However, these experiments are difficult to interpret due to uncertainty about the level of expression and cell types affected by transgenic expression.17 When PML-RARalpha was expressed in non-hematopoietic cells, not only were viable animals difficult to obtain but tumors in non-hematopoietic organs were observed, sometimes with latent periods as short as a few days.18 While expression limited to late myeloid development (CD11b promoter), myeloid maturation was impaired but no leukemia resulted.19 Using a Cathepsin G promoter, which is expressed during a window of promyelocyte maturation, a preleukemic syndrome was noted with 10-30% animals developing leukemia with a mean latent period of 300 days.20,21 As in human APL, all mice relapsed on ATRA treatment, which might seem to imply that other oncogenic lesions are present. On the other hand, ATRA resistance can result from mutations in the fusion protein itself22 or it is possible that the reciprocal fusion protein also possesses oncogenic activity. The mouse model that is most similar to the human APL utilizes the MRP8 promoter, normally expressed at the promyelocyte to metamyelocyte stage and into mature neutrophils. A preleukemic phase is observed, with a syndrome similar to human APL developing in about one third after a median latency of 6 months.23 The simplest explanation for the latency periods observed in all these models is that other genetic events are necessary. Unfortunately, no clonality studies have yet been reported on the transgenic mice.

The recent substantial progress in understanding the molecular basis of APL might also inform the debate on how many mutations are required. In virtually all cases, APL is associated with translocations resulting in a fusion protein, juxtaposing the C terminal portion of the RARalpha receptor with the N terminal portion of one of four other partners, most commonly PML, resulting in a protein that is hyporesponsive to retinoic acid.24 The retinoic acid response is clearly necessary for normal myeloid differentiation. In its non-ligand activated form, RARalpha binds a number of corepressors, notably N-CoR and SMRT, which are part of a multiprotein repressor complex containing Sin3A corepressor and histone deacetylases.25 It is believed that RARalpha can inhibit transcription from certain promoters by alterations in chromatin structure resulting in a failure of the transcriptional machinery to access those promoters. Binding of retinoids results in heterodimerization of RARalpha and RXR and binding with a number of coactivator proteins, which together stimulate transcription from a number of genes with retinoid response elements through interaction with basal factors, alteration of chromatin notably histone acetylation and DNA unwinding.24 As both an excess of wild-type26 and mutant27 forms of RARalpha are able to cause a differentiation block, a dominantly acting squelching mechanism whereby coactivators are sequestered is believed to operate. In addition, PML-RARalpha homodimers can compete with normal RARalpha-RXR and also bind a novel set of target elements.28,29 To this already complex picture must be added further functions of the partner proteins of RARalpha in the translocation. PML has the ability to both stimulate and inhibit transcription from various genes. PML is also a growth suppressor with tumor suppressor functions, perhaps involved in the reaction to viral infection.24 In addition, PML containing nuclear bodies has also been associated with nascent mRNA,30 and PML has been found associated with the transcriptional coactivator CBP and the AP1 DNA binding complex.31 Finally, APL has been described in association with the RARalpha/PML rather than PML/RARalpha.32,33

The second commonest of the RARalpha fusion partners, PLZF, colocalizes with and may be a key interaction partner of PML. All cases have been associated with the presence of the reciprocal translocation partner, which may have an important proliferative induction role.24 Less is known about the roles of the other two fusion partners, nucleophosmin NPM, and nuclear matrix-mitotic apparatus protein, NuMA, although it is clear that both are associated with the nuclear matrix and are associated with several nuclear processes.24 Thus a complex picture of RARalpha function is emerging whereby several domains within the receptor have different functions, some inhibitory and some stimulatory to transcription at a wide variety of genes important in myeloid development. Disruption of function by the formation of a single fusion protein results in many dominantly acting pleiotropic effects; so, even though several different phenotypes may be required for full malignant transformation, it is plausible that such a complex protein, together with effects from the reciprocal fusion protein, could supply several or even all the required phenotypes.

Our analysis also provides further, indirect evidence that the rate limiting event occurs in cells which are committed to differentiation. We have argued that the distribution of mutations in stem cells is complex, being skewed and those from early life being overrepresented;12,38 so it seems unlikely that the simple relationship we have demonstrated here arises from stem cells. In contrast, mutations from more differentiated cells appear to be approximately constant throughout life,12,38 implying the event causing APL is restricted to cells committed to differentiation, a conclusion for which there is other experimental evidence.34,35,36,37 Presumably, the mutation rate for the PML/RARalpha translocation is so low that its occurrence is negligible in cells as infrequent and slowly dividing as stem cells. Only the much greater number of cell divisions that occur during differentiation allow a sufficient probability of mutation to cause disease at the observed rates. Similar reasoning might explain why oncogenic translocations are generally restricted to childhood and hematopoietic malignancies, situations where mitoses are frequent. The observation that translocations are associated with relatively high care rates by chemotherapy, as exemplified by APL, may be explained by their acting as markers of the cell of origin being committed to differentiation rather than any other intrinsic properties.

In conclusion, analysis of the age-specific incidence of APL suggests it is approximately constant over most of the age range. It is argued that this observation implies the disease is due to a single rate limiting mutation and any necessary mutations are confined to cells which are committed to differentiation. While such arguments can never be regarded as proof, it is hoped that the reported observation adds to the debate on the number of mutations required to cause APL. Certainly, the molecular scenario deduced from various experimental techniques must be consistent with the epidemiological observations. In particular, it should be realized that the observation of a constant age-specific incidence, if confirmed in other series, is difficult to reconcile with more than one rate limiting mutation. We look forward to reports of similar analyses on other series of APL and other myeloid translocations.

Acknowledgements

We wish to thank the many hematologists who contributed patients to this survey, in particular, Drs JRY Ross, V Clough and D Gorst who maintain the Oxford, Merseyside and North-Western Registers.

References

1 Vogelstein B, Kinzler KW. The multistep nature of cancer. Trends Genet 1993; 9: 138-141, MEDLINE

2 Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA. Creation of human tumour cells with defined genetic elements. Nature 1999; 400: 464-468, Article MEDLINE

3 Hethcote HW, Knudson AG. Model for the incidence of embryonal cancers: application to retinoblastoma. Proc Natl Acad Sci USA 1978; 75: 2453-2457, MEDLINE

4 Varmus H, Weinberg RA. Genes and the Biology of Cancer. Scientific American Library; WH Freeman and Co: New York, 1993,

5 Stein WD. Analysis of cancer incidence data on the basis of multistage and clonal growth models. Adv Cancer Res 1991; 56: 161-213, MEDLINE

6 Loeb LA. Mutator phenotype may be required for multistage carcinogenesis. Cancer Res 1991; 51: 3075-3079, MEDLINE

7 Fenaux P, Chomienne C, Degos L, Grimwade D. Acute promyelocytic leukemia: biology and treatment. Semin Oncol 1997; 24: 92-102, MEDLINE

8 Bennett B, Booth NA, Croll A, Dawson AA. The bleeding disorder in acute promyelocytic leukemia: fibrinolysis due to u-PA rather than defibrination. Br J Haematol 1989; 71: 511-517, MEDLINE

9 de Thé H, Chomienne C, Lanotte M, Degos L, Dejean A. The t(15;17) translocation of acute promyelocytic leukemia fuses the retinoic acid receptor alpha gene to a novel transcribed locus. Nature 1990; 347: 558-561, MEDLINE

10 Fenaux P, Degos L. Differentiation therapy for acute promyelocytic leukemia. New Engl J Med 1997; 337: 1076-1077, MEDLINE

11 Peto R, Parish SE, Gray RG. There is no such thing as ageing, and cancer is not related to it. In: Likhachev A, Anisimov V, Montesano R (eds). Age-related Factors in Carcinogenesis, . International Agency for Research on Cancer Scientific Publications: Lyon, No 58, 1986, pp 43-53.

12 Akiyama M, Kyoizumi S, Hirai Y, Kusonoki Y, Iwamoto KS, Makamura N. Mutation frequency in human blood cells increases with age. Mutat Res 1995; 338: 141-149, MEDLINE

13 de Haan G, Van Zant G. Dynamic changes in mouse hematopoietic stem cell numbers during aging. Blood 1999; 93: 3294-3301, MEDLINE

14 Taylor P, Jackson G, Vickers M, Ross J, Gorst D, Clough V, Brown M, Poctor S. A population based study on the effect of age, WCC, cytogenetics and ATRA on outcome in acute promyelocytic leukemia. Blood 1997; 90: 234b,

15 Ramsey MJ, Moore DH, Briner JF, Lee DA, Olsen LA, Senft JR, Tucker JD. The effects of age and lifestyle factors on the accumulation of cytogenetic damage as measured by chromosome painting. Mutat Res 1995; 338: 95-106, MEDLINE

16 Green MHL, O'Neill JP, Cole J. Suggestions concerning the relationship between mutant frequency and mutation rate at the hrpt locus in human peripheral T-lymphocytes. Mutat Res 1995; 334: 323-339, MEDLINE

17 Westervelt P, Ley TJ. Seed versus soil: the importance of the target cell for transgenic models of human leukemias. Blood 1999; 93: 2143-2148, MEDLINE

18 David G, Terris B, Marchio A, Lavau C, Dejean A. The acute promyelocytic leukemia PML-RAR alpha protein induces hepatic preneoplastic and neoplastic lesions in transgenic mice. Oncogene 1997; 14: 1547-1554, MEDLINE

19 Early E, Moores M, Kakizuka A, Nason-Burchenal K, Martin P, Evans R, Dmitrowsky X. Transgenic expression of PML-RARalpha impairs myelopoiesis. Proc Natl Acad Sci USA 1996; 93: 7900-7904, Article MEDLINE

20 Grisolano JL, Wesselschmidt RL, Pelicci PG, Ley TJ. Altered myeloid development and acute leukemia in transgenic mice expression PML-RAR alpha under control of cathepsin G regulatory sequences. Blood 1997; 89: 376-387, MEDLINE

21 He LZ, Tribioli C, Rivi R, Peruzzi D, Pelicci PG, Soares V, Cattoretti G, Pandolfi PP. Acute leukemia with promyelocytic features in PML/RAR alpha transgenic mice. Proc Natl Acad Sci USA 1997; 94: 5302-5307, Article MEDLINE

22 Shao W, Benedetti L, Lamph WW, Nervi C, Miller WH. A retinoid-resistant acute promyelocytic leukemia subclone expresses a dominant negative PML-RAR alpha mutation. Blood 1997; 89: 4282-4289, MEDLINE

23 Brown D, Kogan S, Lagasse E, Weissman I, Alcalay M, Pelicci PG, Atwater S, Bishop JM. A PML/RARalpha transgene initiates murine acute promyelocytic leukemia. Proc Natl Acad Sci USA 1997; 94: 2551-2556, Article MEDLINE

24 Melnick A, Licht JD. Deconstructing a disease: RARalpha, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 1999; 93: 3167-3215, MEDLINE

25 Heinzel T, Lavinsky R, Mullen T, Soderstrom M, Laherty C, Torchia J, Yang W-M, Brard G, Ngo S, Davie J, Seto E, Eisenman R, Rose D, Glass C, Rosenfeld M. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 1997; 387: 43-48, MEDLINE

26 Onodera M, Kunisada T, Nishikawa S, Sakiyama Y, Matsumoto S, Nishikawa S. Overexpression of retinoic acid alpha suppresses myeloid cell differentiation at the promyelocytic stage. Oncogene 1995; 11: 1291-1299, MEDLINE

27 Grignani F, Ferrucci PF, Testa U, Talamo G, Fagioli M, Alcalay M, Mencarelli A, Grignani F, Peschle C, Nicoletti I, Pelicci PG. The acute promyelocytic leukemia-specific PML/RAR alpha fusion protein inhibits differentiation and promotes survival of myeloid precursor cells. Cell 1993; 74: 423-431, MEDLINE

28 Perez A, Kastner P, Sethi S, Lutz Y, Reibel C, Chambon P. PMLRAR homodimers: distinct DNA binding properties and heteromeric interactions with RXR. EMBO J 1993; 12: 3171-3182, MEDLINE

29 Jansen JH, Mahfoudi A, Rambaud S, Lavau C, Wahli W, Dejean A. Multimeric complexes of the PML-retinoic acid receptor alpha fusion protein in acute promyelocytic leukemia cells and interference with retinoid and peroxisome-proliferator signaling pathways. Proc Natl Acad Sci USA 1995; 92: 7401-7405, MEDLINE

30 Lamorte VJ, Dyck JA, Ochs RL, Evans RM. Localisation of nascent RNA and CREB binding protein with the PML-containing nuclear body. Proc Natl Acad Sci USA 1998; 95: 4991-4996, Article MEDLINE

31 Vallian S, Gaken JA, Gingold EB, Kouzarides T, Chang KS, Farzaneh F. Modulation of Fos-mediated AP1 transcription by the promyelocytic leukemia protein. Oncogene 1998; 16: 2843-2853, MEDLINE

32 Lafage-Pochitaloff M, Alcalay M, Brunel V, Longo L, Sainty D, Simonetti J, Birg F, Pelicci PG. Acute promyelocytic leukemia cases with nonreciprocal PML/RARalpha or RARalpha/PML fusion genes. Blood 1995; 85: 1169-1174, MEDLINE

33 Mozziconacci M-J, Liberatore C, Brunel V, Grignani F, Arnoulet C, Ferrucci PF, Fernandez F, Sainty D, Pelicci PG, Birg F, Lafage-Pochitaloff M. In vitro response to all trans retinoic acid of acute promyelocytic leukemias with nonreciprocal PML/RARA or RARA/PML fusion genes. Gene Chromosomes Cancer 1998; 22: 241-250,

34 Turhan AG, Lemoine FM, Debert C, Bonnet ML, Baillon C, Picard F, Macintyre EA, Varet B. Highly purified primitive hematopoietic stem cells are PML-RARA negative and generate non-clonal progenitors in acute promyelocytic leukemia. Blood 1995; 85: 2154-2161, MEDLINE

35 Knuutila S, Majander P, Ruutu T. 8;21 and 15;17 translocations: abnormalities in a single cell lineage in acute myeloid leukemia. Acta Haematol 1994; 92: 88-90, MEDLINE

36 Haferlach T, Loffler H, Nickenig C, Ramm-Petersen L, Meeder M, Schoch R, Schlegelberger B, Schnittger S, Schoch C, Hiddemann W. Cell lineage specific involvement in acute promyelocytic leukaemia (APL) using a combination of May-Grunwald-Giemsa staining and fluorescence in situ hybridisation techniques for the detection of the translocation t(15;17)(q22;q12). Br J Haematol 1998; 103: 93-99, MEDLINE

37 Bonnet D, Dick JE. Human acute myeloid leukaemia is organised as a hierarchy that originates from a primitive hematopoietic cell. Nature Med 1997; 3: 730-737, MEDLINE

38 Vickers M, Brown GC, Cologne JB, Kyoizumi S. Modelling haemopoietic stem cell division by analysis of mutant red cells. Br J Haematol(in press)

Figures

Figure 1  The reciprocal of the residual from least squares analysis (dark line), in arbitary units, on the y axis is plotted against possible values of the index, n, as defined by the equation in the text. For illustrative purposes, a normal curve (light dotted line) of mean -0.09 and standard deviation 0.25 is shown.

Figure 2  The cumulative number of cases is plotted against age. Both the observed number (light dotted line) and the predicted number (dark line) of cases, using the optimised values as defined in the text, are shown.

Received 29 July 1999; accepted 8 December 1999
April 2000, Volume 14, Number 4, Pages 722-726
Table of contents    Previous  Article  Next    [PDF]