Risk of developing chronic myeloid neoplasms in well-differentiated thyroid cancer patients treated with radioactive iodine.

Exposure to ionizing radiation increases the risk of myelodysplastic syndromes (MDS) and myeloproliferative neoplasms (MPN), but such risks are not known in well-differentiated thyroid cancer (WDTC) patients treated with radioactive iodine (RAI). A total of 148 215 WDTC patients were identified from Surveillance, Epidemiology and End Results registries between 1973 and 2014, of whom 54% underwent definitive thyroidectomy and 46% received adjuvant RAI. With a median follow-up of 6.6 years, 77 and 66 WDTC patients developed MDS and MPN, respectively. Excess absolute risks for MDS and MPN from RAI treatment when compared to background rates in the US population were 6.6 and 8.1 cases per 100 000 person-years, respectively. Compared to background population rates, relative risks of developing MDS (3.85 (95% confidence interval, 1.7-7.6); P=0.0005) and MPN (3.13 (1.1-6.8); P=0.012) were significantly elevated in the second and third year following adjuvant RAI therapy, but not after thyroidectomy alone. The increased risk was significantly associated with WDTC size ⩾2 cm or regional disease. Development of MDS was associated with shorter median overall survival in WDTC survivors (10.3 vs 22.5 years; P<0.001). These data suggest that RAI treatment for WDTC is associated with increased risk of MDS with short latency and poor survival.

Genomic determinants of chronic myelomonocytic leukemia.

The biology, clinical phenotype and progression rate of chronic myelomonocytic leukemia (CMML) are highly variable due to diverse initiating and secondary clonal genetic events. To determine the effects of molecular features including clonal hierarchy in CMML, we studied whole-exome and targeted next-generation sequencing data from 150 patients with robust clinical and molecular annotation assessed cross-sectionally and at serial time points of disease evolution. To identify molecular lesions unique to CMML, we compared it to the related myeloid neoplasms (N=586), including juvenile myelomonocytic leukemia, myelodysplastic syndromes (MDS) and primary monocytic acute myeloid leukemia and discerned distinct molecular profiles despite similar pathomorphological features. Within CMML, mutations in certain pathways correlated with clinical classification, for example, proliferative vs dysplastic features. While most CMML patients (59%) had ancestral (dominant/co-dominant) mutations involving TET2, SRSF2 or ASXL1 genes, secondary subclonal hierarchy correlated with clinical phenotypes or outcomes. For example, progression was associated with acquisition of new expanding clones carrying biallelic TET2 mutations or RAS family, or spliceosomal gene mutations. In contrast, dysplastic features correlated with mutations usually encountered in MDS (for example, SF3B1 and U2AF1). Classification of CMML based on hierarchies of ancestral and subclonal mutational events may correlate strongly with clinical features and prognosis.

The relationship between eligibility criteria and adverse events in randomized controlled trials of hematologic malignancies.

To minimize adverse events (AEs) unrelated to drugs and maximize the likelihood of drug approvals, eligibility criteria for randomized controlled trials (RCTs) may be overly restrictive. The purpose of this study was to determine if RCTs in hematologic malignancies exclude patients irrespective of known toxicities or observed AEs. MEDLINE was searched from 1/2010 to 1/2015 for RCTs published in high-impact journals. Of 97 trials, 33% were conducted in leukemia, 28% in lymphoma, 34% in multiple myeloma and 5% in myelodysplastic syndromes or myelofibrosis. Expected toxicities at thresholds of ⩾10%, ⩾5% and <5% were not correlated with cardiac, hepatic or renal eligibility criteria (logistic regression). To explore this lack of correlation we tested the concordance of expected toxicities and eligibility criteria using a modified version of McNemar's test: at each threshold, hepatic, renal and cardiac expected toxicities were significantly discordant with eligibility criteria. Hepatic and renal eligibility criteria were also not correlated with observed AEs, P=0.69 and P=0.77, respectively, but a significant correlation was detected between cardiac eligibility criteria and observed AEs, P=0.02. Thus, the analyzed RCTs excluding patients with organ dysfunction do not reflect expected toxicities, based on prescription drug labels/prior experience, or reported AEs on the trials.

The complexity of interpreting genomic data in patients with acute myeloid leukemia.

Acute myeloid leukemia (AML) is a heterogeneous neoplasm characterized by the accumulation of complex genetic alterations responsible for the initiation and progression of the disease. Translating genomic information into clinical practice remained challenging with conflicting results regarding the impact of certain mutations on disease phenotype and overall survival (OS) especially when clinical variables are controlled for when interpreting the result. We sequenced the coding region for 62 genes in 468 patients with secondary AML (sAML) and primary AML (pAML). Overall, mutations in FLT3, DNMT3A, NPM1 and IDH2 were more specific for pAML whereas UTAF1, STAG2, BCORL1, BCOR, EZH2, JAK2, CBL, PRPF8, SF3B1, ASXL1 and DHX29 were more specific for sAML. However, in multivariate analysis that included clinical variables, only FLT3 and DNMT3A remained specific for pAML and EZH2, BCOR, SF3B1 and ASXL1 for sAML. When the impact of mutations on OS was evaluated in the entire cohort, mutations in DNMT3A, PRPF8, ASXL1, CBL EZH2 and TP53 had a negative impact on OS; no mutation impacted OS favorably; however, in a cox multivariate analysis that included clinical data, mutations in DNMT3A, ASXL1, CBL, EZH2 and TP53 became significant. Thus, controlling for clinical variables is important when interpreting genomic data in AML.

Incorporation of molecular data into the Revised International Prognostic Scoring System in treated patients with myelodysplastic syndromes.

The Revised International Prognostic Scoring System (IPSS-R) was developed for untreated myelodysplastic syndrome (MDS) patients based on clinical data. We created and validated a new model that incorporates mutational data to improve the predictive capacity of the IPSS-R in treated MDS patients. Clinical and mutational data from treated MDS patients diagnosed between January 2000 and January 2012 were used to develop the new prognostic system. A total of 508 patients were divided into training (n=333) and validation (n=175) cohorts. Independent significant prognostic factors for survival included age, IPSS-R, EZH2, SF3B1 and TP53. Weighted coefficients for each factor were used to build the new linear predictive model, which produced four prognostic groups: low, intermediate-1, intermediate-2 and high with a median overall survival of 37.4, 23.2, 19.9 and 12.2 months, respectively, P<0.001. Significant improvement in the C-index of the new model (0.73) was observed compared with the IPSS-R (0.69). The new model predicted outcome both in a separate validation cohort and in another cohort of patients with paired samples at different time points during their disease course. The addition of mutational data to the IPSS-R makes it dynamic and enhances its predictive ability in treated MDS patients regardless of their initial or subsequent therapies.

GFI1 as a novel prognostic and therapeutic factor for AML/MDS.

Genetic and epigenetic aberrations contribute to the initiation and progression of acute myeloid leukemia (AML). GFI1, a zinc-finger transcriptional repressor, exerts its function by recruiting histone deacetylases to target genes. We present data that low expression of GFI1 is associated with an inferior prognosis of AML patients. To elucidate the mechanism behind this, we generated a humanized mouse strain with reduced GFI1 expression (GFI1-KD). Here we show that AML development induced by onco-fusion proteins such as MLL-AF9 or NUP98-HOXD13 is accelerated in mice with low human GFI1 expression. Leukemic cells from animals that express low levels of GFI1 show increased H3K9 acetylation compared to leukemic cells from mice with normal human GFI1 expression, resulting in the upregulation of genes involved in leukemogenesis. We investigated a new epigenetic therapy approach for this subgroup of AML patients. We could show that AML blasts from GFI1-KD mice and from AML patients with low GFI1 levels were more sensitive to treatment with histone acetyltransferase inhibitors than cells with normal GFI1 expression levels. We suggest therefore that GFI1 has a dose-dependent role in AML progression and development. GFI1 levels are involved in epigenetic regulation, which could open new therapeutic approaches for AML patients.

Defining AML and MDS second cancer risk dynamics after diagnoses of first cancers treated or not with radiation.

Risks of acute myeloid leukemia (AML) and/or myelodysplastic syndromes (MDS) are known to increase after cancer treatments. Their rise-and-fall dynamics and their associations with radiation have, however, not been fully characterized. To improve risk definition we developed SEERaBomb R software for Surveillance, Epidemiology and End Results second cancer analyses. Resulting high-resolution relative risk (RR) time courses were compared, where possible, to results of A-bomb survivor analyses. We found: (1) persons with prostate cancer receiving radiation therapy have increased RR of AML and MDS that peak in 1.5-2.5 years; (2) persons with non-Hodgkin lymphoma (NHL), lung and breast first cancers have the highest RR for AML and MDS over the next 1-12 years. These increased RR are radiation specific for lung and breast cancer but not for NHL; (3) AML latencies were brief compared to those of A-bomb survivors; and (4) there was a marked excess risk of acute promyelocytic leukemia in persons receiving radiation therapy. Knowing the type of first cancer, if it was treated with radiation, the interval from first cancer diagnosis to developing AML or MDS, and the type of AML, can improve estimates of whether AML or MDS cases developing in this setting are due to background versus other processes.

Clinical and biological implications of ancestral and non-ancestral IDH1 and IDH2 mutations in myeloid neoplasms.

Mutations in isocitrate dehydrogenase 1/2 (IDH1/2(MT)) are drivers of a variety of myeloid neoplasms. As they yield the same oncometabolite, D-2-hydroxyglutarate, they are often treated as equivalent, and pooled. We studied the validity of this approach and found IDH1/2 mutations in 179 of 2119 myeloid neoplasms (8%). Cross-sectionally, the frequencies of these mutations increased from lower- to higher risk disease, thus suggesting a role in clinical progression. Variant allelic frequencies indicated that IDH1(MT) and IDH2(MT) are ancestral in up to 14/74 (19%) vs 34/99 (34%; P=0.027) of cases, respectively, illustrating the pathogenic role of these lesions in myeloid neoplasms. IDH1/2(MT) was associated with poor overall survival, particularly in lower risk myelodysplastic syndromes. Ancestral IDH1(MT) cases were associated with a worse prognosis than subclonal IDH1(MT) cases, whereas the position of IDH2(MT) within clonal hierarchy did not impact survival. This may relate to distinct mutational spectra with more DNMT3A and NPM1 mutations associated with IDH1(MT) cases, and more ASXL1, SRSF2, RUNX1, STAG2 mutations associated with IDH2(MT) cases. Our data demonstrate important clinical and biological differences between IDH1(MT) and IDH2(MT) myeloid neoplasms. These mutations should be considered separately as their differences could have implications for diagnosis, prognosis and treatment with IDH1/2(MT) inhibitors of IDH1/2(MT) patients.

p53 independent epigenetic-differentiation treatment in xenotransplant models of acute myeloid leukemia.

Suppression of apoptosis by TP53 mutation contributes to resistance of acute myeloid leukemia (AML) to conventional cytotoxic treatment. Using differentiation to induce irreversible cell cycle exit in AML cells could be a p53-independent treatment alternative, however, this possibility requires evaluation. In vitro and in vivo regimens of the deoxycytidine analogue decitabine that deplete the chromatin-modifying enzyme DNA methyl-transferase 1 without phosphorylating p53 or inducing early apoptosis were determined. These decitabine regimens but not equimolar DNA-damaging cytarabine upregulated the key late differentiation factors CCAAT enhancer-binding protein ɛ and p27/cyclin dependent kinase inhibitor 1B (CDKN1B), induced cellular differentiation and terminated AML cell cycle, even in cytarabine-resistant p53- and p16/CDKN2A-null AML cells. Leukemia initiation by xenotransplanted AML cells was abrogated but normal hematopoietic stem cell engraftment was preserved. In vivo, the low toxicity allowed frequent drug administration to increase exposure, an important consideration for S phase specific decitabine therapy. In xenotransplant models of p53-null and relapsed/refractory AML, the non-cytotoxic regimen significantly extended survival compared with conventional cytotoxic cytarabine. Modifying in vivo dose and schedule to emphasize this pathway of decitabine action can bypass a mechanism of resistance to standard therapy.

A mathematical model of human thymidine kinase 2 activity.

The mitochondrial enzyme thymidine kinase 2 (TK2) phosphorylates deoxythymidine (dT) and deoxycytidine (dC) to form dTMP and dCMP, which in cells rapidly become the negative-feedback end-products dTTP and dCTP. TK2 kinetic activity exhibits Hill coefficients of ∼0.5 (apparent negative cooperativity) for dT and ∼1 for dC. We present a mathematical model of TK2 activity that is applicable if TK2 exists as two monomer forms in equilibrium.

How to use the computing environment R to analyze ATP-induced ribonucleotide reductase R1 hexamerization data.

R is an object oriented free and open source statistical computing environment. The R package Combinatorially Complex Equilibrium Model Selection is being developed to meet the analysis challenges of ribonucleotide reductase (RNR). An example of its use is given here. This example involves ATP-induced R1 hexamerization dynamic light scattering data that suggests that R1 hexamers have two types of a-sites, one that binds ATP and another that does not (here, R1 is the large subunit of RNR).

A combinatorial multicomponent plug mixer for systems chemistry.

We report the construction and testing of a combinatorial multicomponent plug mixer (CMPM) chip that generates a large number of mix ratios. The CMPM chip has been designed to study ribonucleotide reductase (RNR) protein-protein/protein-ligand interaction networks. The 4-component chip is capable of 5400 different combinations in a 30 plug cycle. CMPM chips were tested producing fluorescent dye and dihydrofolate reductase NADPH/MX mixtures with plug lengths of 2 mm.

Biologically based risk estimation for radiation-induced CML. Inferences from BCR and ABL geometric distributions.

Chronic myeloid leukemia (CML) invites biologically based radiation risk modeling because CML is simultaneously well-understood, homogeneous and prevalent. CML is known to be caused by a translocation involving the ABL and BCR genes, almost all CML patients have the BCR-ABL translocation, and CML is prevalent enough that its induction is unequivocally detected among Hiroshima A-bomb survivors. In a previous paper, a linear-quadratic-exponential (LQE) dose-response model was used to estimate the lifetime excess risk of CML in the limit of low doses of gamma-rays, R gamma. This estimate assumed that BCR-ABL translocation dose-response curves in stem cells for both neutrons and gamma-rays, differ only by a common proportionality constant from dicentric aberration dose-response curves in lymphocytes. In the present paper we challenge this assumption by predicting the BCR-ABL dose response. The predictions are based on the biophysical theory of dual radiation action (TDRA) as it applies to recent BCR-to-ABL distance data in G0 human lymphocytes; this data shows BCR and ABL geometric distributions that are not uniform and not independent, with close association of the two genes in some cells. The analysis speaks against the previous proportionality assumption. We compute 11 plausible LQE estimates of R gamma, 2 based on the proportionality assumption and 9 based on TDRA predictions. For each estimate of R gamma we also compute an associated estimate of the number of CML target cells, N; the biological basis of the LQE model allows us to form such estimates. Consistency between N and hematological considerations provides a plausibility check of the risk estimates. Within the group of estimates investigated, the most plausible lifetime excess risk estimates tend to lie near R gamma = 0.01 Gy-1, substantially higher than risk estimates based on the proportionality assumption.

Sphingoid base metabolism in yeast: mapping gene expression patterns into qualitative metabolite time course predictions.

Can qualitative metabolite time course predictions be inferred from measured mRNA expression patterns? Speaking against this possibility is the large number of 'decoupling' control points that lie between these variables, i.e. translation, protein degradation, enzyme inhibition and enzyme activation. Speaking for it is the notion that these control points might be coordinately regulated such that action exerted on the mRNA level is informative of action exerted on the protein and metabolite levels. A simple kinetic model of sphingoid base metabolism in yeast is postulated. When the enzyme activities in this model are modulated proportional to mRNA expression levels measured in heat shocked yeast, the model yields a transient rise and fall in sphingoid bases followed by a permanent rise in ceramide. This finding is in qualitative agreement with experiments and is thus consistent with the aforementioned coordinated control system hypothesis.

On target cell numbers in radiation-induced H4-RET mediated papillary thyroid cancer.

Radiation-induced human papillary thyroid cancer (PTC) is associated with chromosomal inversions that involve the genetic loci H4 and RET on chromosome 10. Recently, experimental data has shown that these loci lie in very close spatial proximity in a high proportion of adult human thyroid cells. Applying the generalized formulation of dual radiation action to this H4-to-RET geometric distance data, we predict here the radiation dose-response of H4-RET induction. The predicted H4-RET dose-response has a linear-to-quadratic transition dose of approximately 7 Gy, suggesting the validity of linear risk extrapolations to very low doses for H4-RET mediated radiation-induced PTC. In conjunction with A-bomb survivor data, the predicted H4-RET dose-response yields estimates of the number of PTC target cells that are of the order of approximately 10(6) to approximately 10(7) cells, i.e. considerably less than the total number of follicular cells in the thyroid gland.

Biologically-based risk estimation for radiation-induced chronic myeloid leukemia.

Radiation cancer risks are typically determined by the use of simple statistical descriptions of epidemiological data. It is important in risk assessment in general, however, to attempt to incorporate as much biological information into the risk models as possible. We illustrate this by presenting a biologically-based linear-quadratic-exponential (LQE) incidence rate model for radiation-induced chronic myeloid leukemia (CML). The model consists of a linear-quadratic dose-response for the induction of BCR-ABL, a waiting time distribution between BCR-ABL formation and detection of CML, and an exponential cell-killing term that multiplies both the background and induced incidence rates. Using data exclusive of the A-bomb survivor cohort, Bayesian priors are defined for each of the nine parameters in this LQE model. The priors are based on chromosomal translocations in lymphocytes, hematopoietic stem cell survival experiments, CML waiting times in women irradiated for benign disease, the background CML incidence rate in the U.S. population, and genomic DNA target sizes of BCR and ABL. Fixing three of the LQE model parameters to the means of their priors, maximum likelihood estimates of the remaining six parameters were obtained using A-bomb survivor incidence data for Hiroshima males. The likelihood estimates and the corresponding six prior distributions, both approximated as multivariate normal, were then used to form Bayesian posteriors for the six parameters not fixed. With these posteriors the LQE model yields Qgamma*=0.0042 Gy(-1) where Qgamma* is the upper 95% confidence bound of the lifetime CML risk per person-gray in the limit of low doses of gamma-rays. This value is slightly less than Qgamma*=0.0049 Gy(-1) obtained from likelihood estimates of the LQE parameters, and substantially less than Qgamma*=0.0158 Gy(-1) obtained for a simple statistical model linear in dose for kermas less than 4 Gy.