One aspect of Horace Walpole's original definition of serendipity in 1754 that is often missed in modern discussions of the word is the sagacity of being able to link together apparently innocuous facts to come to a valuable conclusion.”1 It is in this vein that we discuss the converging of (a) the empiric success of clinical trials for children with acute lymphoblastic leukemia (ALL) with, (b) some characteristics of the purported leukemic stem cell (LSC), and (c) that some tumor cells seem to be primarily on the de novo purine pathway and that the de novo and salvage pathway are also critically regulated and may be tied to differentiation and proliferation during embryological development. The reality is that the facts are not innocuous, but the converging path was less than obvious unless the fields of developmental biology, pharmacology, and stem cell literature are all in the same dimension.
We developed this proposition after reviewing the paper by Schmiegelow and his colleagues2 in this issue of the journal and some recent papers defining some of the metabolism/pharmacology of the LSC, the results of genome wide arrays of the host and leukemia cell3–5 and implicit conclusions that we draw from decades old observations about embryogenesis and cell differentiation.6–8
The cure of children with ALL absolutely depends on some prolonged therapy that has been referred to as maintenance or in the more modern era, continuation therapy. It consists of 2 to 3 years of repetitive cycles of outpatient, generally nonmarrow suppressive therapy and the “backbone” is traditionally an antifolate and a purine analog, methotrexate (MTX) and 6-mercaptopurine (6MP), respectively. The addition of pulses of other agents may or may not make a significant impact. The conclusion of the Schmiegelow et al's2 paper is that there may be nothing better than MTX and 6MP during the continuation phase of therapy even for patients with higher risk ALL (using LSA2 L2-based therapy).
What are the targets of these 2 antimetabolites and why are they so essential for continuation therapy? If induction therapy rapidly eliminates 3 or more logs of cells (1012 to 109) in the typical child at presentation, but MTX and 6MP are not the best agents to effect this, then why are they needed in the prolonged treatment phase? Is it merely because they can be given? Or is it because (in addition to being lucky), the metabolism of the self-renewing, pluripotent LSC is uniquely sensitive to the inhibition of the de novo pathways for nucleotides needed for DNA repair, methylation, and simply duplication. Some of the basis for continuation therapy are cited in references 20 to 30 in the paper by Dr Schmiegelow and we add to it here.
The convergence of the empiric success for continuation therapy with stem cell and developmental biology and tumor metabolism in general allows us to make the hypothesis that the de novo pathway may be more critical during differentiation, but not necessarily the committed blast that is programmed to merely procreate and does so using the salvage pathway which is a metabolic bargain compared with the de novo path. That is, the serendipity is the drugs used in continuation therapy target the nucleotide producing machinery of the LSC. In the paper by Cox et al3 it is noted that the LSC is resistant to vincristine and dexamethasone, 2 of the most critical drugs for inducing an initial remission, but less important during the continuation phase of therapy.9 A crucial experiment to carry out is the cytotoxicity or antiproliferative potency of MTX and 6MP in these same CD133+ cells. It should also be noted, that although recent data by Vormoor10 and Civin and colleagues (reviewed in Ref. 10 and presented at ASH recently Ref. 11) suggest that there may be more than a few genotypes capable of populating a nonsevere combined immunodeficiency mouse with human ALL, this does not compromise the notion that there are dysregulated, unique and duplicative/redundant pathways for producing DNA precursors that could still play a role in success or failure of treatment. As these cells can be expanded and studied, the detailed metabolomic profiles could provide important information regarding choices of pharmacologic agents.
If malignant transformation is a failure of the usual checks and balances of division and differentiation and escapes from apoptosis then the selective inhibition of the de novo pathway for nucleotide production using a low dose, metronomic scheduling of drugs that have proven tolerable side effects and that may uniquely work on the pluripotent, LSC is serendipitous. We can even ask if it is a generalization that might help with some other cancers for which there is a similar dysregulation of final common paths for DNA synthesis.
The significance of the multiple pathways for intracellular accumulation of XTPs and dXTPs production was presented 40 to 50 years ago6–8 and more than a decade ago it was shown that 3 of the enzymes for the 10 steps in de novo purine synthesis on chromosome 21 are highly regulated as assessed by quantitation in brain of human fetuses and during the first few weeks of life.12 As a tangent to the main discussion, is this partially the reason for toxicity in Down syndrome patients and the success of therapy in patients with an extra chromosome 21? Down children are more sensitive to MTX and also have been reported to be hyperuricemic (and as an aside, excess purines are associated with neurocognitive disorders).
In recent times we have shifted from targeting the machinery for DNA precursor synthesis to the inhibition of the tyrosine kinase receptors and other molecules that impart some regulatory control of proliferation, however, it may be that our original agents that block the accumulation of the building blocks is under such strict developmental control that they remain crucial targets. Moreover, if the hypothesis posited here is true, it maybe “generalizable” to some other cancer stem cells. This would also suggest that continuation therapy might be important in other diseases. As we previously noted, the bullseye for cancer therapy may be a moving target and too specific a targeted agent at times may be a disadvantage. Perhaps the scatter gun rather than the high powered telescopic site, single shot, weapon still has advantages.13 As implied in the Schmiegelow et al's2 paper, we also need to remember that an addition may be a subtraction in the arena of chemotherapy.14 The addition of a block or blocks of therapy to an existing one over a fixed time period will reduce the intensity (dose or time) of the existing one (presumably the known effective one), therefore the addition needs to be better then the old or the potential of this dilution to limit success of the new protocol is real.
1. Definition of serendipity from Wikipedia (the free encyclopedia).
2. Schmiegelow K, Heyman M, Kristinsson J, et al. Oral methotrexate/6-mercaptopurine may be superior to a multi-drug LSA2L2 maintenance therapy for higher risk childhood acute lymphoblastic leukemia: results from the NOPHO ALL-92 study.
3. Cox CV, Diamanti P, Evely RS, et al. Expression of CD133 on leukemia initiating cells in childhood ALL. Blood [Epub ahead of print].
4. Cheok MH, Evans WE. Acute lymphoblastic leukaemia: a model for the pharmacogenomics of cancer therapy. Nat Rev Cancer. 2006;6:117–129.
5. Yang JJ, Cheng C, Yang W, et al. Genome-wide interrogation of germ line genetic variation associated with treatment response in childhood acute lymphoblastic leukemia. JAMA. 2009;301:393–403.
6. Grant P. The influence of folic acid analogs on development and nucleic acid metabolism in Rana pipiens
embryos. Dev Biol. 1960;2:197–251.
7. Stearns LW, Martin WE, Jolley WB, et al. Effects of certain pyrimidines on cleavage and nucleic acid metabolism in sea urchin, Strongylocentrotus purpuratus
, embryos. Exp Cell Res. 1962;27:250–259.
8. Mathews CK. Giant pools of DNA precursors in sea urchin eggs. Exp Cell Res. 1975;92:47–56.
9. Conter V, Valsecchi MG, Silvestri D, et al. Pulses of vincristine and dexamethasone in addition to intensive chemotherapy for children with intermediate-risk acute lymphoblastic leukaemia: a multicentre randomised trial. Lancet. 2007;369:123–131.
10. Vormoor HJ. Malignant stem cells in childhood acute lymphoblastic leukemia: the stem cell concept revisited. Cell Cycle. 2009;8:1–4.
11. Morisot S, Wayne AS, Bohana-Kashtan O, et al. Leukemia stem cells (LSCs) are frequent in childhood precursor B acute lymphoblastic leukemia (ALL) [Abstract No. 1354]. Blood. 2008;112:487.
12. Brodsky G, Barnes T, Bleskan J, et al. The human GARS-AIRS-GART gene encodes two proteins which are differentially expressed during human brain development and temporally overexpressed in cerebellum of individuals with Down syndrome. Hum Mol Genet. 1997;6:2043–2050.
13. Frost P, Kamen BA. The bullseye of cancer therapy: a moving target. Curr Opin Pharmacol. 2003;3:335–337.
14. Weitman SD, Kamen BA. Combination chemotherapy: a case of new math? When an addition is a subtraction. Cancer Res Ther Control. 1994;4:1–2.