Addition of four doses of rituximab to standard induction chemotherapy in adult patients with precursor B-cell acute lymphoblastic leukaemia (UKALL14): a phase 3, multicentre, randomised controlled trial.

BACKGROUND: Treatment for adults with acute lymphoblastic leukaemia requires improvement. UKALL14 was a UK National Cancer Research Institute Adult ALL group study that aimed to determine the benefit of adding the anti-CD20 monoclonal antibody, rituximab, to the therapy of adults with de novo B-precursor acute lymphoblastic leukaemia. METHODS: This was an investigator-initiated, phase 3, randomised controlled trial done in all UK National Health Service Centres treating patients with acute lymphoblastic leukaemia (65 centres). Patients were aged 25-65 years with de-novo BCR-ABL1-negative acute lymphoblastic leukaemia. Patients with de-novo BCR-ABL1-positive acute lymphoblastic leukaemia were eligible if they were aged 19-65 years. Participants were randomly assigned (1:1) to standard-of-care induction therapy or standard-of-care induction therapy plus four doses of intravenous rituximab (375 mg/m2 on days 3, 10, 17, and 24). Randomisation used minimisation and was stratified by sex, age, and white blood cell count. No masking was used for patients, clinicians, or staff (including the trial statistician), although the central laboratory analysing minimal residual disease and CD20 was masked to treatment allocation. The primary endpoint was event-free survival in the intention-to-treat population. Safety was assessed in all participants who started trial treatment. This study is registered with ClincialTrials.gov, NCT01085617. FINDINGS: Between April 19, 2012, and July 10, 2017, 586 patients were randomly assigned to standard of care (n=292) or standard of care plus rituximab (n=294). Nine patients were excluded from the final analysis due to misdiagnosis (standard of care n=4, standard of care plus rituximab n=5). In the standard-of-care group, median age was 45 years (IQR 22-65), 159 (55%) of 292 participants were male, 128 (44%) were female, one (<1%) was intersex, and 143 (59%) of 244 participants had high-risk cytogenetics. In the standard-of-care plus rituximab group, median age was 46 years (IQR 23-65), 159 (55%) of 294 participants were male, 130 (45%) were female, and 140 (60%) of 235 participants had high-risk cytogenetics. After a median follow-up of 53·7 months (IQR 40·3-70·4), 3-year event-free survival was 43·7% (95% CI 37·8-49·5) for standard of care versus 51·4% (45·4-57·1) for standard of care plus rituximab (hazard ratio [HR] 0·85 [95% CI 0·69-1·06]; p=0·14). The most common adverse events were infections and cytopenias, with no difference between the groups in the rates of adverse events. There were 11 (4%) fatal (grade 5) events in induction phases 1 and 2 in the standard-of-care group and 13 (5%) events in the standard-of-care plus rituximab group). 3-year non-relapse mortality was 23·7% (95% CI 19·0-29·4) in the standard-of-care group versus 20·6% (16·2-25·9) in the standard-of-care plus rituximab group (HR 0·88 [95% CI 0·62-1·26]; p=0·49). INTERPRETATION: Standard of care plus four doses of rituximab did not significantly improve event-free survival over standard of care. Rituximab is beneficial in acute lymphoblastic leukaemia but four doses during induction is likely to be insufficient. FUNDING: Cancer Research UK and Blood Cancer UK.

Type 1 Interferon Responses Underlie Tumor-Selective Replication of Oncolytic Measles Virus.

The mechanism of tumor-selective replication of oncolytic measles virus (MV) is poorly understood. Using a stepwise model of cellular transformation, in which oncogenic hits were additively expressed in human bone marrow-derived mesenchymal stromal cells, we show that MV-induced oncolysis increased progressively with transformation. The type 1 interferon (IFN) response to MV infection was significantly reduced and delayed, in accordance with the level of transformation. Consistently, we observed delayed and reduced signal transducer and activator of transcription (STAT1) phosphorylation in the fully transformed cells. Pre-treatment with IFNß restored resistance to MV-mediated oncolysis. Gene expression profiling to identify the genetic correlates of susceptibility to MV oncolysis revealed a dampened basal level of immune-related genes in the fully transformed cells compared to their normal counterparts. IFN-induced transmembrane protein 1 (IFITM1) was the foremost basally downregulated immune gene. Stable IFITM1 overexpression in MV-susceptible cells resulted in a 50% increase in cell viability and a significant reduction in viral replication at 24 h after MV infection. Overall, our data indicate that the basal reduction in functions of the type 1 IFN pathway is a major contributor to the oncolytic selectivity of MV. In particular, we have identified IFITM1 as a restriction factor for oncolytic MV, acting at early stages of infection.

The Role of Neutrophils in Measles Virus-mediated Oncolysis Differs Between B-cell Malignancies and Is Not Always Enhanced by GCSF.

The mechanism by which oncolytic measles virus (MV) kills cancer cells remains obscure. We previously showed that neutrophils are involved in MV-mediated tumor regressions and become activated, upon MV infection. In the present study, we attempted to enhance the neutrophil response toward MV-infected tumor targets by generating an oncolytic MV-expressing human granulocyte colony-stimulating factor (MVhGCSF). Evaluating the effects in two different models of B-cell malignancy, we showed that depletion of neutrophils abrogated the MV therapeutic effect in an in vivo Raji-but not Nalm-6 tumor model. Next, we compared MVhGCSF with the unmodified backbone virus MVNSe. MVhGCSF enhanced the oncolytic capacity of MV in the Raji model in vivo, whereas in the Nalm-6 model, the opposite was unexpectedly the case. This finding was recapitulated by exogenously administered hGCSF. MVhGCSF replicated within an MV-infectable CD46 transgenic mouse model with detectable serum levels of hGCSF but no toxicity. Our data suggest that a "one-size-fits-all" model of immune response to viral oncolysis is not appropriate, and each tumor target will need full characterization for the potential of both direct and indirect, innate immune responses to generate benefit.

Mouse xenograft modeling of human adult acute lymphoblastic leukemia provides mechanistic insights into adult LIC biology.

The distinct nature of acute lymphoblastic leukemia (ALL) in adults, evidenced by inferior treatment outcome and different genetic landscape, mandates specific studies of disease-initiating mechanisms. In this study, we used NOD/LtSz-scid IL2Rγ null(c) (NSG) mouse xenotransplantation approaches to elucidate leukemia-initiating cell (LIC) biology in primary adult precursor B (pre-B) ALL to optimize disease modeling. In contrast with xenografting studies of pediatric ALL, we found that modification of the NSG host environment using preconditioning total body irradiation (TBI) was indispensable for efficient engraftment of adult non-t(4;11) pre-B ALL, whereas t(4;11) pre-B ALL was successfully reconstituted without this adaptation. Furthermore, TBI-based xenotransplantation of non-t(4;11) pre-B ALL enabled detection of a high frequency of LICs (<1:6900) and permitted frank leukemic engraftment from a remission sample containing drug-resistant minimal residual disease. Investigation of TBI-sensitive stromal-derived factor-1/chemokine receptor type 4 signaling revealed greater functional dependence of non-t(4;11) pre-B ALL on this niche-based interaction, providing a possible basis for the differential engraftment behavior. Thus, our studies establish the optimal conditions for experimental modeling of human adult pre-B ALL and demonstrate the critical protumorogenic role of microenvironment-derived SDF-1 in regulating adult pre-B LIC activity that may present a therapeutic opportunity.

Attenuated, oncolytic, but not wild-type measles virus infection has pleiotropic effects on human neutrophil function.

We previously showed that neutrophils play a role in regression of human tumor xenografts in immunodeficient mice following oncolytic vaccine measles virus (MV-Vac) treatment. In this study, we sought, using normal human neutrophils, to identify potential neutrophil-mediated mechanisms for the attenuated MV-Vac induced effects seen in vivo, by comparison with those consequent on wild-type (WT-MV) infection. Both MV-Vac and WT-MV infected and replicated within neutrophils, despite lack of SLAM expression. In both cases, neutrophils survived longer ex vivo postinfection. Furthermore, MV-Vac (but not WT-MV) infection activated neutrophils and stimulated secretion of several specific antitumor cytokines (IL-8, TNF-α, MCP-1, and IFN-α) via induction of de novo RNA and protein synthesis. In addition, MV-Vac (but not WT-MV) infection caused TRAIL secretion in the absence of de novo synthesis by triggering release of prefabricated TRAIL, via a direct effect upon degranulation. The differences between the outcome of infection by MV-Vac and WT-MV were not entirely explained by differential infection and replication of the viruses within neutrophils. To our knowledge, this is the first demonstration of potential mechanisms of oncolytic activity of an attenuated MV as compared with its WT parent. Furthermore, our study suggests that neutrophils have an important role to play in the antitumor effects of oncolytic MV.

Differential cytopathology and kinetics of measles oncolysis in two primary B-cell malignancies provides mechanistic insights.

Clinical trials using vaccine measles virus (MV) as anticancer therapy are already underway. We compared the oncolytic potential of MV in two B-cell malignancies; adult acute lymphoblastic leukemia (ALL, an aggressive leukemia) and chronic lymphocytic leukemia (CLL, an indolent leukemia overexpressing Bcl-2) using patient-derived material. In vitro, distinct cytopathological effects were observed between MV-infected primary ALL and CLL cells, with large multinucleated syncytia forming in ALL cultures compared to minimal cell-to-cell fusion in infected CLL cells. Cell viability and immunoblotting studies confirmed rapid cell death in MV-infected ALL cultures and slower MV oncolysis of CLL cells. In cell lines, overexpression of Bcl-2 diminished MV-induced cell death providing a possible mechanism for the slower kinetic of MV oncolysis in CLL. In vivo, intratumoral MV treatment of established subcutaneous ALL xenografts had striking antitumor activity leading to complete resolution of all tumors. The antitumor activity of MV was also evident in disseminated ALL xenograft models. In summary, both ALL and CLL are targets for MV-mediated lysis albeit with different kinetics. The marked sensitivity of both primary ALL cells and ALL xenografts to MV oncolysis highlights the tremendous potential of MV as a novel replicating-virus therapy for adult ALL.