Translational studies on antimetabolic therapies in paediatric oncology

Sammanfattning: Cure rates for paediatric and adult cancer patients have improved within the last decades. This can partly be explained by implementation of new technologies and methodologies such as the identification of new mutations after sequencing that can be directly targeted for treatment or the introduction of immunotherapy. However, there is an urgent need for improvement of survival particularly for patients with relapsed metastatic disease. More than 20 years ago, SAMHD1 was discovered and even though its key role in preventing viral HIV-1 infections was initially established, it was only later classified as the first deoxynucleoside triphosphate triphosphohydrolase that can remove the three phosphogroups of the dNTPs in a single reaction, which contributes to the dNTP pool homoeostasis by limiting potentially hazardous expansion of the intracellular dNTP pool. SAMHD1 is a homotetramer that is strictly regulated by the dNTP levels, with two allosteric sites (AS1 and AS2) and one catalytic site responsible for the dNTPase activity. Cancer cells are, among other hallmarks, characterized by loss of proliferation inhibition. It is therefore not surprising that in many cancer types, deregulation, or mutations of SAMHD1 have been reported that allow cells to circumvent dNTP shortage to permit further DNA replication. Many chemotherapeutic drugs target uncontrolled cancer proliferation. For example, a large group of these compounds are analogues of physiological nucleosides leading to inhibition of DNA replication. SAMHD1 has the capacity to use many of these analogues as substrates and through its dNTPase activity, it dephosphorylates them and prevents their incorporation into the nascent DNA chain. This can lead to treatment resistance effectively inactivating chemotherapy. One of these analogues frequently used in regimens against haematological malignancies is cytarabine. However, its active metabolite ara-CTP is a substrate for SAMHD1, hence SAMHD1- positive cancers might limit its cytotoxic efficacy. Therefore, SAMHD1 represents a promising therapeutic target, and its inhibition might enhance cytarabine efficacy. In the present thesis, we aimed to investigate whether there is an association between SAMHD1 expression and response to treatment with nucleoside analogues in two different haematological malignancies and whether SAMHD1 inhibition can improve current treatment protocols. In paper I, we performed a phenotypic screen of more than 33000 small molecules and discovered non-competitive inhibitors of ribonucleotide reductase to potentiate cytarabine in a SAMHD1-dependent manner. Inhibition of SAMHD1 activity towards ara- CTP occurred in an indirect manner as RNR inhibition led to dNTP ratio imbalances affecting SAMHD1 substrate specificity. As dCTP outcompeted dATP as dominant AS2 activator, SAMHD1 activity towards ara-CTP was gradually lost. Functionally, the RNR inhibitors hydroxyurea or gemcitabine acted synergistically with cytarabine, and sensitized cells to treatment in a SAMHD1-dependent manner, both in cell lines and in patient derived AML blasts. Furthermore, combination treatment prolonged survival in murine AML models. As a result, with this study we discovered already clinically available drugs that could act synergistically with cytarabine and improve treatment outcome. Hence, SAMHD1 can act as a biomarker for AML patients and combining cytarabine with RNR inhibitors might overcome SAMHD1-mediated resistance. In paper II, we showed that another nucleoside analogue, nelarabine, that is specifically cytotoxic against malignant T-cells was both an allosteric activator and a substrate for SAMHD1, thus limiting its cytotoxic efficacy. SAMHD1 depletion led to treatment sensitization and addition of hydroxyurea in SAMHD1 expressing cells, inhibited SAMHD1 catalytic activity and increased intracellular levels of the active metabolite ara-GTP. Finally, in T-ALL patient derived cells, addition of HU improved the efficacy of nelarabine treatment. All in all, we showed that SAMHD1 expression is a resistance factor in nelarabine treatment and inhibition with HU could have a potential clinical use. In paper III, based on our preclinical data we performed a small phase 1 clinical trial to validate the efficacy and safety of adding hydroxyurea to cytarabine-based treatment of AML patients. A total of nine patients were enrolled and they received a minimum of two cycles of treatment including daunorubicin, cytarabine and hydroxyurea. Analysis of blood mononuclear cells of patients showed that adding HU increased ara-CTP levels in vivo. All patients achieved complete remission (CR) without unexpected or unacceptable toxicities and MRD was negative in all eight patients that could be evaluated. Thus, CR of all patients combined with the pharmacokinetic studies, suggested that adding HU to alleviate the SAMHD1-based resistance barrier can be a rational strategy to improve treatment outcomes with cytarabine-based treatments, In paper IV, we investigated the correlation between SAMHD1 expression and its impact on induction and consolidation therapy of AML. In two independent patient cohorts (n=98 and n=124), SAMHD1 protein expression levels were assessed via immunohistochemistry. SAMHD1 was differentially expressed in AML blasts but was not expressed in several physiological hematopoietic cells. Based on their SAMHD1 expression, samples were allocated to three different groups and although no effect of SAMHD1 expression was evident during induction therapy, patients with low SAMHD1 levels at diagnosis had significantly prolonged event-free and overall survival rates. Therefore, evaluation of SAMHD1 levels can serve as a prognostic marker and might stratify personalized treatment strategies including SAMHD1 inhibitors. In summary, the results of this thesis show that SAMHD1 can be used as a prognostic biomarker for AML treated with cytarabine-based regimens and might stratify patients for enhanced treatment protocols adding the SAMHD1 inhibitor hydroxyurea to cytarabine. SAMHD1 might have a similar role for the nucleoside analogue nelarabine in Tlymphoblastic malignancies. Hence, SAMHD1 might constitute a universal resistance factor for a group of nucleoside analogues, irrespective of the specific oncological diagnosis. Targeting SAMHD1 thus promises to improve outcomes for a large group of cancers.