Better understanding of tumour biology and improved knowledge of cytogenetics along with new drugs is leading to innovations in treatment of adult, high-risk acute myeloid leukaemia
BSc MSc PhD FRPharmS FCPP(Hon)
Almost every situation in acute myeloid leukaemia (AML) is high risk and in need of improvement”, Alan Burnett (professor of haematology, Cardiff University) told the audience at a satellite symposium sponsored by Genzyme. Many prognostic factors have now been identified in this heterogeneous disease but not all are predictive of response to treatment, he continued. They include age, cytogenetics, mutation status and response to the first course of treatment. Mutation status is predictive of response in many cases and can also help to identify patients suitable for specific treatment agents. For example, patients with FLT3 mutations are more likely to relapse and the risk is increased by the presence of internal tandem duplications (ITD). The mutant:wild type ratio of FLT3 also has prognostic value with high ratios associated with poorer survival. NPM1 mutations protect against the effects of FLT3 mutations and it has been shown that the combination of FLT3-ITD mutant and nucleophosmin 1 (NPM1) wild type has the poorest survival.
A risk index has been constructed on the basis of a regression analysis and this takes into account variables such as age, sex, cytogenetics, status after the first course and white cell count. The index allows good, standard and poor risk groups to be identified. There is no evidence that high-risk groups benefit from stem cell transplants and so fewer such procedures could be undertaken, said Professor Burnett.
A prognostic index has been devised using data from more than 1000 patients in the Mylotarg (gemtuzumab) AML15 study. Key variables include age, cytogenetic status and type of disease. Using the index, a value of less than 3.6 is associated with a less than 10% chance of survival. The score permits identification of those with favourable cytogenetics, said Professor Burnett.
Considering survival after relapse, patients under the age of 60 years fare better.
Professor Burnett concluded that prognostic factors need to be predictive and multi-parameter scores may be more sensitive, but validation would be critical. Expensive assays may not add value.
Pre-treatment cytogenetics is the most important factor in determining which patients are high risk, according to Stefan Faderl (Department of Leukemia, The University of Texas MD Anderson Cancer Center, Houston, Texas). Three risk groups have been identified on the basis of karyotype – the unfavourable risk group that accounts for about 30 to 35% of AML and includes secondary AML and older patients, an intermediate group and the favourable group that represents about one fifth of patients. Cytogenetic subgroups associated with adverse prognoses include those with partial tandem duplications (PTD) of the mixed-lineage leukaemia gene (MLL-PTD), ecotropic viral integration site 1 (EVI-1) and the combination FLT3-ITD/NPM1 wild type. The general approach to the management of AML is well established with standard treatment regimens for acute promyelocytic leukaemia (APL) and core binding factor (CBF) AML. For others, decisions are based on factors such as age, karyotype and physical condition. In high-risk patients under 60 years there is little evidence that they benefit from cytarabine, high dose daunorubicin or gemtuzumab. The FLT3 inhibitors, such as sorafenib and lestaurtinib, represent an interesting development; there are many of them and they vary in specificity.
Studies using FLT3 inhibitors together with an anthracycline and cytarabine (eg, idarubicin, cytarabine and sorafenib or daunorubicin, cytarabine and midostaurin) have initially achieved better responses in FLT3-mutated patients but overall survival has not improved. Treatment with idarubicin, cytarabine and vorinostat, a histone deactylase inhibitor, looks promising, with a good response amongst FLT3-mutated patients and no additional toxicity from the vorinostat. In patients of less than 60 years, treatment with cladribine, daunorubicin and cytarabine had produced a 46% three-year overall survival rate, said Dr Faderl.
Elderly patients fare worse because of poor tolerance and increased mortality with intensive chemotherapy, comorbid conditions and different disease biology. The more adverse factors an individual has, the worse the prognosis. Epigenetic treatment using hypomethylating agents, such as azacitidine or decitabine, is now playing an increasing role in the treatment of AML in the elderly. Two studies show that azacitidine prolongs survival in elderly patients with a French study showing an overall one-year survival of 37%.
The second-generation nucleoside analogue clofarabine appears to have greater activity than its predecessors, fludarabine and cladribine. In a study of clofarabine as a single agent in elderly patients with AML and at least one adverse prognostic factor, the overall response rate was 46% and the median survival was 59 weeks.
A study at the MD Anderson Cancer Center had investigated the effects of clofarabine and low-dose cytarabine alternating with decitabine. In a group of sixty patients of median age 70 years, 50% of whom had secondary AML and 42% had complex cytogenetics, complete and overall response rates of 66% and 72%, respectively, were obtained. The combination was better than single agents and lower doses should be possible, noted Dr Faderl.
Another novel nucleoside analogue is sapacitabine – an oral cytosine analogue. It causes irreparable single-strand DNA breaks and induces G2 cell cycle arrest. It has shown promising anti-leukaemic activity against relapsed/refractory AML or MDS in a phase 1 trial. Laromustine, a novel sulphonylhydrazine alkylating agent is also being studied in AML.
Summarising, Dr Faderl said that cytogenetic-molecular characterisation is important and that targets should be matched with drugs, although this can be hard to do in practice. In addition, high-intensity therapy in older patients may not be beneficial and there are many new drugs to explore.
Novel salvage treatment options
Conventional therapy for AML is of limited efficacy with typical complete response rates of less than 50%, short overall survival times (median 3–12 months) and, without a stem cell transplant, long-term success rates of less than 10%, said Sergio Amadori (Tor Vergata University Hospital, Rome, Italy). What is needed is better understanding of the pathobiology of disease, new targets for drug development and biologically determined and personalised therapies. In addition, new trial designs that can screen novel therapies more rapidly are needed along with intergroup cooperation to test new agents or strategies in subsets based on disease heterogeneity. Finally, better stem cell transplant regimens are required that make use of novel technologies and an expanded donor pool.
Dr Amadori identified three categories of drugs that are likely to make an impact in this field; they are new cytotoxics, including clofarabine, voreloxin, sapacitabine and CP4055, new FLT3 inhibitors such as AC220 and inhibitors of novel molecular targets such as mTOR and aurora kinases.
Clofarabine is a purine analogue that is active against dividing and resting cells. Early trials with complete remission rates of 42–52% demonstrated its impressive anti-leukaemic activity and suggested that it might have a useful role in AML, said Dr Amadori. It is logical to combine it with cytarabine, because it up-regulates the intracellular metabolism of cytarabine, and with anthracyclines, because it potentiates their activity by inhibiting DNA repair. In recent years a number of studies have examined the feasibility, safety and effectiveness of using clofarabine in combination with different doses of cytarabine, with and without anthracyclines (see Table 1).
The overall message from these studies was that combination treatment was feasible with early mortality rates of 10–15% and complete remission rates ranging from 25–50%, said Dr Amadori.
It has been suggested that clofarabine could provide a ‘cytoreductive bridge’ to transplant, by virtue of its potent anti-leukaemic activity. Clofarabine induces profound bone-marrow suppression and AML cyto-reduction and this could be done immediately before stem cell transplantation (SCT). The procedure has been tested in a group of 17 patients with active, advanced leukaemia, (median age 58 years).
Five days treatment with clofarabine 30–40 mg/m2 was given to induce bone marrow suppression. About 15 days later the conditioning regimen was started, followed by allogeneic stem cell transfusion. This treatment strategy was associated with grade 3–4 liver toxicity in 18% of patients, but this was tolerable and there was no severe veno-occlusive disease (VOD), noted Dr Amadori. Sixteen patients underwent SCT and one patient died of sepsis without receiving a transplant. These results show this could be a useful application for this very powerful drug, commented Dr Amadori.
Voreloxin is a quinolone derivative that intercalates into DNA and is toxic to topoisomerase II. In a phase 1 study in combination with cytarabine remission was induced in about 30% of patients including many who had experienced primary induction failures. The dose-limiting toxicity was mucositis.
Sapacitabine is an orally bioavailable cyanocytosine analogue with a unique mechanism of action – it causes single-strand DNA breaks, G2 arrest and rapid apoptosis. Elacytarabine (CP4055) is independent of nucleoside transporter mechanisms for cellular uptake and exhibits improved cytotoxicity because of increased cellular uptake, decreased deactivation and prolonged exposure to ara-CTP.
New FLT3 inhibitors are needed because although the first generation agents are effective they cannot induce complete remission in people with FLT3 abnormalities – more selective and potent FLT3 inhibitors are needed, explained Dr Amadori. AC220 is the most selective and potent so far and in a phase 1 evaluation it elicited a complete response in a substantial proportion of patients – “in 30% for the first time” he noted. “There was a CR of 56% in FLT3-ITD positive patients and this is a very significant result”, he added.
Novel molecular targets – mTOR and aurora kinases
The mTOR (mammalian target of rapamycin) pathway promotes growth and proliferation. It is activated in most AML cells and may be required by leukaemic cells for survival. MTOR inhibition by rapamycin or ‘rapalogues’ such as everolimus causes cell cycle arrest of AML cells and increases the pro-apoptotic effect of chemotherapy.
Aurora kinases are protein kinases essential for mitotic progression and they are over-expressed in a wide range of tumours, including AML. Several aurora kinase inhibitors, eg, tozasertib, are currently in clinical trials.
Dr Amadori concluded that the treatment of advanced AML remains challenging. Although there are many new agents their exact roles are still to be determined. Critically, study design needs to be updated to generate results more rapidly. A ‘pick the winner’ approach is required rather than more large trials looking for minimal differences, he said.
One future development in treatment is likely to be based on the use of agents that target common pathways that AML cells use to protect themselves against chemotherapy and maintain their viability. Improved knowledge of AML genomes will make this possible, according to John DiPersio (Division of Oncology, Siteman Cancer Center Washington University School of Medicine, USA).
A physical interaction between bone marrow stem cells and stromal cells protects them from the effects of chemotherapy and so it was hypothesised that leukaemic cells would be protected in the same way. Studies have shown that leukaemic cells that are bound to stroma are not dividing; in fact, they accumulate in G0 and G1 phases, and are therefore not susceptible to cell cycle specific agents such as cytarabine. The introduction of plerixafor (Mozobil), a reversible inhibitor of chemokine (C-X-C motif) receptor 4 (CXCR4), that induces mobilisation of stem cells into the peripheral circulation suggested the possibility of pre-treatment with this agent to enhance the effects of chemotherapy.
Investigations using plerixafor and cytarabine in a mouse model of leukaemia showed a threefold prolongation of survival compared with the use of cytarabine alone. A human study was designed to evaluate the effect of adding plerixafor to the MEC (mitoxantrone, etoposide and cytarabine) salvage regimen. The MEC regimen was chosen because it is a standard regimen of moderate effectiveness. Of the 51 patients in the study, 11 had FLT3-activating mutations and 75% had primary refractory disease or were in the first year of relapse. They were therefore the highest risk group for salvage, said Professor DiPersio. The majority of patients experienced a two-threefold mobilisation of white cells and blasts after a single injection of plerixafor and there was no preferential mobilisation of leukaemic blasts. Researchers expected to see the CXCR4 level on the surface of leukaemic blasts go down but it increased 10-fold. Even though the interaction between CXCR4 and STF-1 is blocked transiently (by plerixafor) and blasts are mobilised, the receptor is up-regulated and this “chases the blasts back into the marrow” so more innovative ways of keeping the cells out of the bone marrow will need to be developed to use the chemosensitisation approach optimally, explained Professor DiPersio. There was no hyperleukocytosis in this trial and no delay in neutrophil or platelet recovery. In the Phase II part of the study the CR rate for early relapsed high-risk patients was 50% and this now needs to be repeated and validated in a randomised trial. This study demonstrated that plerixafor can safely be given with cytotoxic chemotherapy and that the response rates compare favourably with historic controls, said Professor DiPersio.
Other pathways that could be targeted are the E-selectin pathway and the VLA4-VCAM-1 pathway; Aurora kinases small molecule inhibitors of VLA4 and E-selectins have already been synthesised. GCSF may also be used to down-regulate CXCR4 over a long period of time to overcome the transient effects of plerixafor, he added.
In summary, Professor DiPersio said that there are two ways to consider sensitisation; one is to optimise the physical dissociation of leukaemic cells from bone marrow microenvironment using small molecule inhibitors or GCSF, which chronically down-regulates STF-1, and the other approach is inhibition of pathways that are activated to induce anti-apoptosis, anti-proliferation or anti-differentiation.