Correspondence

444References

1 Cervantes F, Colomer D, Vives-Corrons JL, Rozman C, MontserratE. Chronic myeloid leukemia with thrombocythemic onset: a CMLsubtype with distinct hematological and molecular features? Leu-kemia 1996; 10: 1241–1243.

2 Kwong YL, Chiu EKW, Liang RHS, Chan V, Chan TK. Essentialthrombocythemia with BCR/ABL rearrangement. Cancer GenetCytogenet 1996; 89: 74–76.

3 Michiels JJ, Prins ME, Hagermeijer A, Brederoo P, van der MeulenJ, van Vliet HHD, Abels J. Philadelphia chromosome-positivethrombocythemia and megakaryoblast leukemia. Am J Clin Pathol1987; 88: 645–652.

4 Richards EM, Bloxham DM, Nacheva E, Marcus RE, Green AR.

COMMENT ON PUBLISHED PAPER

Discordant erythropoiesis in CML

TO THE EDITOR

We enjoyed reading the recent leading article by Clarkson etal1 concerning ‘New understanding of the pathogenesis ofCML: a prototype of early neoplasia’ and broadly agree withtheir ‘Discordant Maturation’ hypothesis. However, we weresurprised by their view that our published data on BFU-E inCML2 were not in line with their thinking. They have foundthat the majority of erythrocyte progenitors in CML are moremature and less capable of extensive proliferation than com-parable normal erythrocyte progenitors. We reported thatBFU-E in CML formed bursts consisting of more subcoloniesthan normal BFU-E.2 Assuming, as we do, that commitmentto subcolony formation during the growth of an erythroid burstmarks a terminal differentiation event, these two statementsimply exactly the same relationship between early and lateerythroid progenitors. That is, in order for a given cell popu-lation to outnumber its progenitors disproportionately, theremust be an increase in its productive capacity either byincreasing its probability of self-renewal or by increasing thenumber of cell doublings separating the two compartments.

We would like to take this opportunity to comment furtheron the interpretation of our methodology, which relies on thenumber of subcolonies formed per erythroid burst, rangingfrom one to the maximum observed. According to Clarksonet al,1 this phenomenon results from the breaking up of thecolonies into subunits at later culture times. Against thisinterpretation, we have made sequential video recordingswhich show quite clearly that the cells giving rise to the sub-colonies separate prior to the onset of subcolony formation.3

Another criticism that has been levelled at our methodologyis that subcolony formation is simply a function of the densityof the supporting methylcellulose. However, if this were cor-rect, it is difficult to understand why different subcolony multi-

Correspondence: SB Marley; Fax: 44 181 740 9679Received 10 September 1997; accepted 25 November 1997

BCR rearrangement in apparent essential thrombocythaemia. Br JHaematol 1993; 85: 625–626.

5 Stoll DB, Peterson P, Exten R, Laszlo J, Pisciotta AV, Ellis JT, WhiteP, Vaidya K, Bozdech M, Murphy S. Clinical presentation andnatural history of patients with essential thrombocythemia and thePhiladelphia chromosome. Am J Hematol 1988; 27: 77–83.

6 Saglio G, Pane F, Gottardi E, Frigeri F, Buonaiuto M, GuerrasioA, De Micheli D, Parziale A, Fornaci MN, Martinelli G, SalvatoreF. Consistent amounts of acute leukemia-associated P190BCR/ABLtranscripts are expressed by chronic myelogenous leukemiapatients at diagnosis. Blood 1996; 87: 1075–1080.

7 Kwong YL. Association between b3a2 BCR/ABL fusion andchronic myeloid leukemia with thrombocythemic onset: fortuitousor real? Leukemia 1997; 11: 617–618.

plicities should be observed with different recombinant cyto-kines but otherwise identical culture conditions.4

Finally, Clarkson et al1 have highlighted the response ofnormal and CML progenitor cells to c-kit ligand (KL). In bothcases, addition of KL to Epo increased the size of the coloniesformed by BFU-E. Similarly, we have found that KL plus Epo,compared with Epo alone, increases the cellularity of normalBFU-E (7.5-fold; P = 0.0001)4 although subcolony multiplicitywas not increased. In CML, in addition to increased subcolonycellularity, there was a moderate increase in subcolony multi-plicity (unpublished data).

SB Marley Department of HaematologyJL Lewis Imperial College School of MedicineJM Goldman Hammersmith Hospital, London, UKMY Gordon

References

1 Clarkson BD, Strife A, Wisniewski D, Lambek C, Carpino N. Newunderstanding of the pathogenesis of CML: a prototype of earlyneoplasia. Leukemia 1997; 11: 1404–1428.

2 Marley SB, Lewis JL, Goldman JM, Gordon MY. Abnormal kineticsof colony formation by erythroid burst-forming units (BFU-E) inchronic myeloid leukaemia. Br J Haematol 1996; 93: 878–883.

3 Marley SB, Amos TAS, Gordon MY. Kinetics of colony formationby BFU-E grown under different culture conditions in vitro. Br JHaematol 1996; 92: 559–561.

4 Lewis JL, Marley SB, Blackett NM, Szydlo R, Goldman JM, GordonMY. Interleukin-3 (IL-3) but not stem cell factor (SCF) increasesself-renewal by human erythroid burst-forming units (BFU-E) invitro. Cytokine (in press).