Proliferation patterns in acute myeloid leukemia: leukemic clonogenic growth and in vivo cell cycle kinetics

MDS and AML show elevated fractions of CD34-positive blast cell populations with a high anti-apoptotic versus proliferation ratio

This study investigates the intertwined processes of (anti-)apoptosis and cell proliferation in myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). Utilizing antibodies to Bcl-2 and Ki-67, the CD34-positive blast cell compartments in bone marrow aspirates from 50 non-malignant cases, 25 MDS patients, and 25 AML patients were analyzed for their anti-apoptotic and proliferative cell fractions through ten-color flow cytometry. MDS patients exhibited a significantly increased anti-apoptotic (p=0.0014) and reduced proliferative cell fraction (p=0.0030) in their blast cell population as compared to non-malignant cases. AML patients showed an even more exacerbated trend than MDS patients. The resulting Bcl-2:Ki-67 cell fraction ratios in MDS and AML were significantly increased as compared to the non-malignant cases (p=0.0004 and p<0.0001, respectively). AML patients displayed, however, a high degree of variability in their anti-apoptotic and proliferation index, attributed to heterogeneity in maturation stage and severity of the disease at diagnosis. Using double-labeling for Bcl-2 and Ki-67 it could be shown that besides blast cells with a mutually exclusive Ki-67 and Bcl-2 expression, also blast cells concurrently exhibiting anti-apoptotic and proliferative marker expression were found. Integrating these two dynamic markers into MDS and AML diagnostic workups may enable informed conclusions about their biological behavior, facilitating individualized therapy decisions for patients.

A deep profile of gene expression across 18 human cancers

Clinically and biologically valuable information may reside untapped in large cancer gene expression data sets. Deep unsupervised learning has the potential to extract this information with unprecedented efficacy but has thus far been hampered by a lack of biological interpretability and robustness. Here, we present DeepProfile, a comprehensive framework that addresses current challenges in applying unsupervised deep learning to gene expression profiles. We use DeepProfile to learn low-dimensional latent spaces for 18 human cancers from 50,211 transcriptomes. DeepProfile outperforms existing dimensionality reduction methods with respect to biological interpretability. Using DeepProfile interpretability methods, we show that genes that are universally important in defining the latent spaces across all cancer types control immune cell activation, while cancer type-specific genes and pathways define molecular disease subtypes. By linking DeepProfile latent variables to secondary tumor characteristics, we discover that tumor mutation burden is closely associated with the expression of cell cycle-related genes. DNA mismatch repair and MHC class II antigen presentation pathway expression, on the other hand, are consistently associated with patient survival. We validate these results through Kaplan-Meier analyses and nominate tumor-associated macrophages as an important source of survival-correlated MHC class II transcripts. Our results illustrate the power of unsupervised deep learning for discovery of novel cancer biology from existing gene expression data.

The potential of proliferative and apoptotic parameters in clinical flow cytometry of myeloid malignancies

Standardization of the detection and quantification of leukocyte differentiation markers by the EuroFlow Consortium has led to a major step forward in the integration of flow cytometry into classification of leukemia and lymphoma. In our opinion, this now enables introduction of markers for more dynamic parameters, such as proliferative and (anti)apoptotic markers, which have proven their value in the field of histopathology in the diagnostic process of solid tumors and lymphoma. Although use of proliferative and (anti)apoptotic markers as objective parameters in the diagnostic process of myeloid malignancies was studied in the past decades, this did not result in the incorporation of these biomarkers into clinical diagnosis. This review addresses the potential of these markers for implementation in the current, state-of-the-art multiparameter analysis of myeloid malignancies. The reviewed studies clearly recognize the importance of proliferation and apoptotic mechanisms in the pathogenesis of bone marrow (BM) malignancies. The literature is, however, contradictory on the role of these processes in myelodysplastic syndrome (MDS), MDS/myeloproliferative neoplasms, and acute myeloid leukemia. Furthermore, several studies underline the need for the analysis of the proliferative and apoptotic rates in subsets of hematopoietic BM cell lineages and argue that these results can have diagnostic and prognostic value in patients with myeloid malignancies. Recent developments in multiparameter flow cytometry now allow quantification of proliferative and (anti)apoptotic indicators in myeloid cells during their different maturation stages of separate hematopoietic cell lineages. This will lead to a better understanding of the biology and pathogenesis of these malignancies.

Designing combination therapies using multiple optimal controls

Strategic management of populations of interacting biological species routinely requires interventions combining multiple treatments or therapies. This is important in key research areas such as ecology, epidemiology, wound healing and oncology. Despite the well developed theory and techniques for determining single optimal controls, there is limited practical guidance supporting implementation of combination therapies. In this work we use optimal control theory to calculate optimal strategies for applying combination therapies to a model of acute myeloid leukaemia. We present a versatile framework to systematically explore the trade-offs that arise in designing combination therapy protocols using optimal control. We consider various combinations of continuous and bang-bang (discrete) controls, and we investigate how the control dynamics interact and respond to changes in the weighting and form of the pay-off characterising optimality. We demonstrate that the optimal controls respond non-linearly to treatment strength and control parameters, due to the interactions between species. We discuss challenges in appropriately characterising optimality in a multiple control setting and provide practical guidance for applying multiple optimal controls. Code used in this work to implement multiple optimal controls is available on GitHub.

In-silico comparison of two induction regimens (7 + 3 vs 7 + 3 plus additional bone marrow evaluation) in acute myeloid leukemia treatment

BACKGROUND

Clinical integration of systems biology approaches is gaining in importance in the course of digital revolution in modern medicine. We present our results of the analysis of an extended mathematical model describing abnormal human hematopoiesis. The model is able to describe the course of an acute myeloid leukemia including its treatment. In first-line treatment of acute myeloid leukemia, the induction chemotherapy aims for a rapid leukemic cell reduction. We consider combinations of cytarabine and anthracycline-like chemotherapy. Both substances are widely used as standard treatment to achieve first remission. In particular, we compare two scenarios: a single-induction course with 7 days cytarabine and 3 day of anthracycline-like treatment (7 + 3) with a 7 + 3 course and a bone marrow evaluation that leads, in case of insufficient leukemic cell reduction, to the provision of a second chemotherapy course. Three scenarios, based on the leukemias growth kinetics (slow, intermediate, fast), were analyzed. We simulated different intensity combinations for both therapy schemata (7 + 3 and 7 + 3 + evaluation).

RESULTS

Our model shows that within the 7 + 3 regimen a wider range of intensity combinations result in a complete remission (CR), compared to 7 + 3 + evaluation (fast: 64.3% vs 46.4%; intermediate: 63.7% vs 46.7%; slow: 0% vs 0%). Additionally, the number of simulations resulting in a prolonged CR was higher within the standard regimen (fast: 59.8% vs 40.1%; intermediate: 48.6% vs 31.0%; slow: 0% vs 0%). On the contrary, the 7 + 3 + evaluation regimen allows CR and prolonged CR by lower chemotherapy intensities compared to 7 + 3. Leukemic pace has a strong impact on treatment response and especially on specific effective doses. As a result, faster leukemias are characterized by superior treatment outcomes and can be treated effectively with lower treatment intensities.

CONCLUSIONS

We could show that 7 + 3 treatment has considerable more chemotherapy combinations leading to a first CR. However, the 7 + 3 + evaluation regimen leads to CR for lower therapy intensity and presumably less side effects. An additional evaluation can be considered beneficial to control therapy success, especially in low dose settings. The treatment success is dependent on leukemia growth dynamics. The determination of leukemic pace should be a relevant part of a personalized medicine.

Impaired DNA damage response--an Achilles' heel sensitizing cancer to chemotherapy and radiotherapy

Despite the progress in targeting particular molecular abnormalities specific to different cancers (targeted therapy), chemo- and radiotherapies are still the most effective of all anticancer modalities. Induction of DNA damage and inhibition of cell proliferation are the objects of most chemotherapeutic agents and radiation. Their effectiveness was initially thought to be due to the high rate of proliferation of cancer cells. However, normal cell proliferation rate in some tissues often exceeds that of curable tumors. Most tumors have impaired DNA damage response (DDR) and the evidence is forthcoming that this confers sensitivity to chemo- or radiotherapy. DDR is a complex set of events which elicits a plethora of molecular interactions engaging signaling pathways designed to: (a) halt cell cycle progression and division to prevent transfer of DNA damage to progeny cells; (b) increase the accessibility of the damaged sites to the DNA repair machinery; (c) engage DNA repair mechanisms and (d) activate the apoptotic pathway when DNA cannot be successfully repaired. A defective DDR makes cancer cells unable to effectively stop cell cycle progression, engage in DNA repair and/or trigger the apoptotic program when treated with DNA damaging drugs. With continued exposure to the drug, such cells accumulate DNA damage which leads to their reproductive death that may have features of cell senescence. Cancers with nonfunctional BRCA1 and BRCA2 are particularly sensitive to combined treatment with DNA damaging drugs and inhibitors of poly(ADP-ribose) polymerase. Antitumor strategies are being designed to treat cancers having particular defects in their DDR, concurrent with protecting normal cells.

ID1 and ID2 are retinoic acid responsive genes and induce a G0/G1 accumulation in acute promyelocytic leukemia cells

Acute promyelocytic leukemia (APL) is uniquely sensitive to treatment with all-trans retinoic acid (ATRA), which results in the expression of genes that induce the terminal granulocytic differentiation of the leukemic blasts. Here we report the identification of two ATRA responsive genes in APL cells, ID1 and ID2. These proteins act as antagonists of basic helix-loop-helix (bHLH) transcription factors. ATRA induced a rapid increase in ID1 and ID2, both in the APL cell line NB4 as well as in primary patient cells. In addition, a strong downregulation of E2A was observed. E2A acts as a general heterodimerization partner for many bHLH proteins that are involved in differentiation control in various tissues. The simultaneous upregulation of ID1 and ID2, and the downregulation of E2A suggest a role for bHLH proteins in the induction of differentiation of APL cells following ATRA treatment. To test the relevance of this upregulation, ID1 and ID2 were overexpressed in NB4 cells. Overexpression inhibited proliferation and induced a G0/G1 accumulation. These results indicate that ID1 and ID2 are important retinoic acid responsive genes in APL, and suggest that the inhibition of specific bHLH transcription factor complexes may play a role in the therapeutic effect of ATRA in APL.

Triggering Noncycling Hematopoietic Progenitors and Leukemic Blasts to Proliferate Increases Anthracycline Retention and Toxicity by Downregulating Multidrug Resistance

Expression of the multidrug resistance (MDR) mechanisms P-glycoprotein (Pgp) and MDR-related protein (MRP) decrease cellular retention and consequently cytotoxicity of anthracyclines. MDR is expressed on normal human hematopoietic progenitors and leukemic blasts. Normal CD34+ progenitors showed rhodamine efflux in 20% to 30% of the cells, which could be blocked by verapamil. These cells appeared noncycling, in contrast to the proliferating rhodamine bright (RhoB) cells. We postulated that MDR expression can be downregulated by proliferation induction. Triggering rhodamine dull (RhoD) CD34+ cells to proliferate indeed resulted in a higher rhodamine retention and significantly decreased efflux modulation by verapamil (P = .04). Also in acute myeloid leukemia (AML), the proliferation rate (percentage S/G2+M and Iododeoxyuridine labelings index) was significantly less in the RhoD blasts (P ≤ .008) and proliferation induction of RhoD blasts resulted in increased rhodamine retention. Anthracycline cytotoxicity was less for RhoD than RhoB cells in both normal progenitors and leukemic blasts. Proliferation induction of the RhoD cells resulted in increased anthracycline sensitivity. We conclude that noncycling progenitors, both normal and leukemic, have a relatively high MDR expression. Triggering these cells into proliferation downregulates MDR expression. These findings can be exploited to overcome MDR in the treatment of AML patients.

In vitro growth in acute myeloblastic leukaemia: relationship with other clinico-biological characteristics of the disease

The in vitro growth characteristics of a large series of acute myeloid leukaemia (AML) patients and their relationship with other clinical and biological disease characteristics were analysed. Patients with AML were studied, 181 with de novo AML and 45 with secondary AML (24 myelodysplastic syndrome, sAML-MDS, 21 myeloproliferative disorder, sAML-MPD). Leukaemic colony forming units (L-CFU) were assayed by plating peripheral blood (PB) blast cells in methyl-cellulose and using LCM-PHA as stimulant. In each case parallel cultures were made with and without stimulating factors. Plating efficiency (PE) was defined as the number of clusters plus colonies/10(5) cells plated. Autonomous growth (AG) was the number of colonies plus clusters growing without stimulant. The autonomous proliferative index (API) was calculated as the number of clusters + colonies without stimulating factor divided by the number of clusters + colonies with stimulating factor. No significant differences in the PE between de novo and secondary AML were found. Autonomous growth was significantly higher in sAML-MPD. The FAB subtype M3 leukaemias displayed a significantly greater PE and a significantly lower API when compared with the other FAB subgroups (P=0.0002). Upon analysing the relationship with the immunophenotype, only CD33 expression showed a significant relationship with the in vitro growth pattern; CD33+ cases displayed a higher PE (P=0.0002) and AG (P=0.0003) than CD33- cases. When patients were grouped according to the level of rh123 efflux (MDR1) it was observed that cases with >30% elimination showed a higher AG and API than those with <30% (P=0.03). Finally we found that patients with higher API (>0.05) displayed a significantly shorter overall survival as compared with patients with API<0.05 (P=0.04). The in vitro study properties of clonogenic cells produces relevant clinical information of leukaemic cell biology in AML patients.

Levels of expression of CAF7 (CD98) have prognostic significance in adult acute leukemia

The levels of CD98 antigen expression were studied in 62 consecutive cases of adult acute leukemia including 24 acute lymphoblastic leukemia (ALL) and 38 acute myeloid leukemia (AML) using the monoclonal antibody CAF7 and flow cytometry. The mean follow-up was 13.5 months. The mean relative fluorescence intensity (MIF) of CAF7 varied between 6 and 83 channels (256 channels resolution). No correlation was established between CAF7 cell surface density and most of the predictive parameters such as age, sex, blood counts, immunophenotype, proliferative index (PI) or DNA index. Nevertheless expression of CAF7 correlated positively with survival duration (mean 210 vs 391 days, P = 0.048) and complete remission (CR) duration (mean 132 vs 361, days P = 0.032). The levels of CAF7 differed significantly between ALL and AML (P < 0.001), the ALL cases being all CAF7intermediate or CAF7high. In the AML group the low levels of CAF7 expression correlated with shorter CR duration (mean 132 vs 414 days, P = 0.017). The lack of correlation with other clinical and biological parameters suggested that CAF7 might have an independent prognostic significance in adult AML. Although PI was also positively related to survival duration (P = 0.02), it did not correlate with CR duration or the expression of CAF7. We suppose that the prognostic impact of CD98 is related to the control of cell growth and survival in which the molecule normally participates.

In vitro response of blasts to IL-3, GM-CSF, and G-CSF is different for individual AML patients: factors that stimulate leukemic clonogenic cells also enhance Ara-C cytotoxicity

In vivo, growth factors are currently investigated for their capacity to trigger leukemic stem cells into cycle and thus overcome kinetic drug resistance. In this study, the susceptibility of leukemic clonogenic cells to individual growth factors was related to cytosine-arabinoside (Ara-C) sensitivity. The effects of interleukin-3 (IL-3), granulocyte-macrophage colony-stimulating factor (G-CSF), granulocyte colony-stimulating factor (G-CSF), and combinations of these recombinant hematopoietic factors were tested on blast cells of nine acute myeloid leukemia (AML) patients. Growth factor responses were assessed in semi-solid clonogenic assay and in a 10-day liquid culture followed by clonogenic assay. Heterogeneity in growth factor response was observed in both test systems, resulting in a variable pattern for individual leukemias. In the majority of cases (six of nine) the response patterns in the semi-solid and liquid cultures were divergent. To test the Ara-C sensitivity, leukemic blasts were exposed in liquid to various concentrations of Ara-C in the absence and presence of preselected growth factors. After 10 days, the number of surviving leukemic colony-forming cells (CFU-L) was assessed. Exposure to Ara-C in the presence of optimal stimulatory factor(s) resulted in a 3- to 1000-fold increase of the Ara-C toxicity in seven patients. The Ara-C concentrations resulting in 50% inhibition of clonogenicity (ID50) were 0.48-123 x 10(-8) M Ara-C in the absence of stimulatory growth factors, versus only 0.12-0.40 x 10(-8) M Ara-C in the presence of these factors. In two patients, addition of one or more factors neither increased the number of CFU-L in liquid nor enhanced the Ara-C toxicity. Even in the absence of growth factors the ID50 values in these cases were as low as 0.20 and 0.28 x 10(-8) M Ara-C and in the same range as the ID50 values observed with maximum growth factor stimulation in the other seven patients. These results indicate that Ara-C cytotoxicity can be enhanced by individually selected, clonogenic cell growth-promoting hematopoietic factors.

Cell cycle kinetics of hematopoiesis before and after in vivo administration of GM-CSF in refractory anemia: evidence for a shortening of the granulocyte release time

GM-CSF administration to patients with refractory anemia (RA) induces an increase in neutrophils and eosinophils. We studied cell kinetic mechanisms underlying this observation using clonogenic assays and in vivo iododeoxyuridine labeling of bone marrow cells. Cell cycle kinetics were studied in three patients before and during GM-CSF administration (two daily subcutaneous injections of 54 or 108 micrograms). No consistent effect on the relative number of bone marrow CFU-GM was noticed. The DNA synthesis time and potential doubling time of low-density bone marrow cells remained essentially the same. A slight decrease (1.5-3.7%) in labeling index was found, originating from the myelo(-mono)cytic lineage. In all three patients the release time of labeled granulocytes from the bone marrow into the peripheral blood was shortened (before GM-CSF treatment 5-7 days and during GM-CSF 3-4 days). Cell cycle kinetics of CD34+ cells were studied in order to obtain kinetic information on immature precursor and progenitor cells. The DNA synthesis time of the CD34+ cells was shortened during GM-CSF therapy, resulting in a shorter potential doubling time. GM-CSF administration to patients with RA results in a rise in granulocytes that might be due partly to an accelerated release of granulocytes from the bone marrow compartment into the circulating blood and partly to an increased proliferative activity of the immature precursor and progenitor cells.