Search

TABLE S1

Medulloblastoma is the most frequent malignant childhood brain tumor with a high morbidity. Identification of new therapeutic targets would be instrumental in improving patient outcomes. We evaluated the expression of the tumor-associated antigen PRAME in biopsies from 60 patients with medulloblastoma. PRAME expression was detectable in 82% of tissues independent of molecular and histopathologic subgroups. High PRAME expression also correlated with worse overall survival. We next investigated the relevance of PRAME as a target for immunotherapy. Medulloblastoma cells were targeted using genetically modified T cells with a PRAME-specific TCR (SLL TCR T cells). SLL TCR T cells efficiently killed medulloblastoma HLA-A*02+ DAOY cells as well as primary HLA-A*02+ medulloblastoma cells. Moreover, SLL TCR T cells controlled tumor growth in an orthotopic mouse model of medulloblastoma. To prevent unexpected T-cell–related toxicity, an inducible caspase-9 (iC9) gene was introduced in frame with the SLL TCR; this safety switch triggered prompt elimination of genetically modified T cells. Altogether, these data indicate that T cells genetically modified with a high-affinity, PRAME-specific TCR and iC9 may represent a promising innovative approach for treating patients with HLA-A*02+ medulloblastoma.Significance: These findings identify PRAME as a medulloblastoma tumor-associated antigen that can be targeted using genetically modified T cells. Cancer Res; 78(12); 3337–49. ©2018 AACR.

Chimeric antigen receptor (CAR) T-cell therapy represents a revolutionary approach to induce long-lasting remission in patients with B-cell malignancies not responding to conventional therapies. Nevertheless, possible severe side effects, including cytokine release syndrome (CRS), neurotoxicity and macrophage activation syndrome, whose management is still challenging, as well as lack of pathophysiological experimental models to investigate novel interventions, limit the widespread use of this therapy. In light of these considerations, we developed a comprehensive humanized mouse model to investigate the role of IFNγ neutralization, provided by the clinically approved monoclonal antibody, emapalumab, in controlling severe toxicity related to CAR T cells. We demonstrated that emapalumab reduces the pro-inflammatory environment in the animal model, allowing severe CRS control and preventing brain damage, characterized by multifocal hemorrhages. Furthermore, we proved that IFNγ inhibition does not affect the ability of CAR.CD19 T cells to eradicate CD19+ lymphoma cells, both in vitro and in vivo.

BackgroundPatients with Primary Antibody Deficiencies (PAD) represent a potential at-risk group in the current COVID-19 pandemic. However, unexpectedly low cumulative incidence, low infection-fatality rate, and mild COVID-19 or asymptomatic SARS-CoV-2 infections were frequently reported in PAD. The discrepancy between clinical evidence and impaired antibody production requires in-depth studies on patients’ immune responses.MethodsForty-one patients with Common Variable Immune Deficiencies (CVID), 6 patients with X-linked Agammaglobulinemia (XLA), and 28 healthy age-matched controls (HD) were analyzed for anti-Spike and anti-RBD antibody production, generation of low and high affinity Spike-specific memory B-cells, Spike-specific T-cells before and one week after the second dose of BNT162b2 vaccine.ResultsHD produced antibodies, and generated memory B-cells with high affinity for Trimeric Spike. In CVID, the vaccine induced poor Spike-specific antibodies, and atypical B-cells with low affinity for Trimeric Spike, possibly by extra-follicular reactions or incomplete germinal center reactions. In HD, among Spike positive memory B-cells, we identified receptor-binding-domain-specific cells that were undetectable in CVID, indicating the incapability to generate this new specificity. Specific T-cell responses toward Spike-protein were evident in HD and defective in CVID. Due to the absence of B-cells, patients with XLA responded to immunization by specific T-cell responses only.ConclusionsWe present detailed data on early non-canonical immune responses in PAD to a vaccine against an antigen never encountered before by humans. From our data, we expect that after BNT162b2 immunization, XLA patients might be protected by specific T-cells, while CVID patients might not be protected by immunization.

[Abstract]: Chimeric antigen receptor T-cells (CAR T-cells) for the treatment of relapsing/refractory B-cell precursor acute lymphoblastic leukemia have led to exciting clinical results. However, CAR T-cell approaches revealed a potential risk of CD19-/CAR+ leukemic relapse due to inadvertent transduction of leukemia cells. [Methods]: We evaluated the impact of a high percentage of leukemia blast contamination in patient-derived starting material (SM) on CAR T-cell drug product (DP) manufacturing. In vitro as well as in vivo models were employed to identify characteristics of the construct associated with better profile of safety in case of inadvertent B-cell leukemia transduction during CAR T-cell manufacturing. [Results]: The presence of large amounts of CD19+ cells in SM did not affect the transduction level of DPs, as well as the CAR T-cell rate of expansion at the end of standard production of 14 days. DPs were deeply characterized by flow cytometry and molecular biology for Ig-rearrangements, showing that the level of B-cell contamination in DPs did not correlate with the percentage of CD19+ cells in SM, in the studied patient cohort. Moreover, we investigated whether CAR design may affect the control of CAR+ leukemia cells. We provided evidences that CAR.CD19 short linker (SL) prevents complete epitope masking in CD19+CAR+ leukemia cells and we demonstrated in vitro and in vivo that CD19 +CAR(SL)+leukemic cells are killed by CAR.CD19 T-cells. [Conclusions]: Taken together, these data suggest that a VL-VH SL may result in a safe CAR-T product, even when manufacturing starts from biological materials characterized by heavy contamination of leukemia blasts., The experimental work was supported by grants awarded by Ricerca Finalizzata GR-2013 02359212 (CQ), GR-2016-02364546 (BDA), RF-2016- 02364388 (FL), Accelerator Award-Cancer Research UK/AIRC/AECC-INCAR project (FL and AO), Associazione Italiana Ricerca per la Ricerca sul Cancro (AIRC)-Special Project 5×1000 no. 9962 (FL), AIRC IG 2018 id. 21724 (FL), Ricerca Corrente (FL, CQ, BDA), Ministero dell’Università e della Ricerca (Grant PRIN 2017 to FL); Italian Healthy Ministry project on CAR T RCR-2019-23669115 (coordinator FL), Independent Research grant AIFA (FL PI: 2016 call).

Background: Patients with primary antibody deficiencies are at risk in the current COVID-19 pandemic due to their impaired response to infection and vaccination. Specifically, patients with common variable immunodeficiency (CVID) generated poor spike-specific antibody and T cell responses after immunization. Methods: Thirty-four CVID convalescent patients after SARS-CoV-2 infection, 38 CVID patients immunized with two doses of the BNT162b2 vaccine, and 20 SARS-CoV-2 CVID convalescents later and immunized with BNT162b2 were analyzed for the anti-spike IgG production and the generation of spike-specific memory B cells and T cells. Results: Spike-specific IgG was induced more frequently after infection than after vaccination (82% vs. 34%). The antibody response was boosted in convalescents by vaccination. Although immunized patients generated atypical memory B cells possibly by extra-follicular or incomplete germinal center reactions, convalescents responded to infection by generating spike-specific memory B cells that were improved by the subsequent immunization. Poor spike-specific T cell responses were measured independently from the immunological challenge. Conclusions: SARS-CoV-2 infection primed a more efficient classical memory B cell response, whereas the BNT162b2 vaccine induced non-canonical B cell responses in CVID. Natural infection responses were boosted by subsequent immunization, suggesting the possibility to further stimulate the immune response by additional vaccine doses in CVID.

Background Data on immune responses to SARS-CoV-2 in patients with Primary Antibody Deficiencies (PAD) are limited to infected patients and to heterogeneous cohorts after immunization. Methods Forty-one patients with Common Variable Immune Deficiencies (CVID), six patients with X-linked Agammaglobulinemia (XLA), and 28 healthy age-matched controls (HD) were analyzed for anti-Spike and anti-receptor binding domain (RBD) antibody production, generation of Spike-specific memory B-cells, and Spike-specific T-cells before vaccination and one week after the second dose of BNT162b2 vaccine. Results The vaccine induced Spike-specific IgG and IgA antibody responses in all HD and in 20% of SARS-CoV-2 naive CVID patients. Anti-Spike IgG were detectable before vaccination in 4 out 7 CVID previously infected with SARS-CoV-2 and were boosted in six out of seven patients by the subsequent immunization raising higher levels than patients naïve to infection. While HD generated Spike-specific memory B-cells, and RBD-specific B-cells, CVID generated Spike-specific atypical B-cells, while RBD-specific B-cells were undetectable in all patients, indicating the incapability to generate this new specificity. Specific T-cell responses were evident in all HD and defective in 30% of CVID. All but one patient with XLA responded by specific T-cell only. Conclusion In PAD patients, early atypical immune responses after BNT162b2 immunization occurred, possibly by extra-follicular or incomplete germinal center reactions. If these responses to vaccination might result in a partial protection from infection or reinfection is now unknown. Our data suggests that SARS-CoV-2 infection more effectively primes the immune response than the immunization alone, possibly suggesting the need for a third vaccine dose for patients not previously infected.

We developed an innovative and efficient, feeder-free culture method to genetically modify and expand peripheral blood-derived NK cells with high proliferative capacity, while preserving the responsiveness of their native activating receptors. Activated peripheral blood NK cells were efficiently transduced by a retroviral vector, carrying a second-generation CAR targeting CD19. CAR expression was demonstrated across the different NK-cell subsets. CAR.CD19-NK cells display higher antileukemic activity toward CD19+ cell lines and primary blasts obtained from patients with B-cell precursor ALL compared with unmodified NK cells. In vivo animal model data showed that the antileukemia activity of CAR.CD19-NK cell is superimposable to that of CAR-T cells, with a lower xenograft toxicity profile. These data support the feasibility of generating feeder-free expanded, genetically modified peripheral blood NK cells for effective "off-the-shelf" immuno-gene-therapy, while their innate alloreactivity can be safely harnessed to potentiate allogeneic cell therapy.

Medulloblastoma is the most frequent malignant childhood brain tumor with a high morbidity. Identification of new therapeutic targets would be instrumental in improving patient outcomes. We evaluated the expression of the tumor-associated antigen PRAME in biopsies from 60 patients with medulloblastoma. PRAME expression was detectable in 82% of tissues independent of molecular and histopathologic subgroups. High PRAME expression also correlated with worse overall survival. We next investigated the relevance of PRAME as a target for immunotherapy. Medulloblastoma cells were targeted using genetically modified T cells with a PRAME-specific TCR (SLL TCR T cells). SLL TCR T cells efficiently killed medulloblastoma HLA-A*02+ DAOY cells as well as primary HLA-A*02+ medulloblastoma cells. Moreover, SLL TCR T cells controlled tumor growth in an orthotopic mouse model of medulloblastoma. To prevent unexpected T-cell–related toxicity, an inducible caspase-9 (iC9) gene was introduced in frame with the SLL TCR; this safety switch triggered prompt elimination of genetically modified T cells. Altogether, these data indicate that T cells genetically modified with a high-affinity, PRAME-specific TCR and iC9 may represent a promising innovative approach for treating patients with HLA-A*02+ medulloblastoma. Significance: These findings identify PRAME as a medulloblastoma tumor-associated antigen that can be targeted using genetically modified T cells. Cancer Res; 78(12); 3337–49. ©2018 AACR.

Based on the clinical success observed in acute lymphoblastic leukemia (ALL) with chimeric antigen receptor engineered T (CAR T), we hypothesized that combining the specificity of a CAR with the innate allo-reactivity of KIR-mismatched NK cells might provide a powerful tool for adoptive cell therapy. The use of a third-party bank of CAR-NK cells offers the advantage of an immediate availability to be exploited in the allogenic setting and could be associated with a lower toxicity profile than CAR-T cells. In order to overcome regulatory and manufacturing hurdles associated with generation of CAR-NK cells, we developed a feeder-free culture resulting in a 3.2-log expansion after 20 days of culture. Specifically, natural cytotoxicity receptors (NCR) expressed on NK cells are stimulated in the presence of pleiotropic cytokines and expanded in GMP grade bioreactors. Expanded NK cells from healthy donors preserve a high percentage of CD56+ CD57- cells (85±13%), associated with high proliferative capability, and maintain the surface expression and the responsiveness of NCR and CD16. We proved that NK cells generated from 10 different healthy donors have high ability to recognize and eliminate different tumor types, including acute myeloid leukemia (AML) and ALL. After genetic modification with a retroviral vector encoding a CAR specific for CD19 antigen, transduction of activated NK cells averaged 38%±15% and the CAR.CD19 expression was stable over extended in vitro culture (60 days). Detailed phenotypic characterization of CAR-NK cells showed that CAR expression was not limited to the more mature NKG2A-/KIR+ cells, but rather was distributed across different NK subsets. We also demonstrated that NK and CAR-NK cells display significant anti-leukemia activity towards CD19+ leukemia and lymphoma cell lines (LCL 721.221, DAUDI and BV173) and primary blasts obtained from patients with B-cell precursor ALL (Bcp-ALL). Co-culture experiments using a 1:5 E/T ratio, showed that, while the anti-tumor activity was already remarkable with non-modified effector NK cells (60±30%, 71±33% and 54±23% of residual LCL 721.221, DAUDI and BV173 cells, respectively; p An in vivo model of leukemia xenograft immunodeficient mice was used to evaluate whether CAR-NK cells are associated with a lower toxicity profile compared to CAR-T cells. While the in vivo antileukemia activity was superimposable between CAR-T and CAR-NK cells (mouse bioluminenscence at 20 days, 4.9x105 vs 6.6x105 photons/second, respectively; p=n.s. Figure A), mice treated with two i.v. infusions (day 0 and day 15) of 10x106 CAR.CD19 NK cells had a 100% overall survival (OS of 5 out of 5 mice) at 50 days compared to 20% of mice (1 out of 5) receiving 10x106 CAR.CD19 T cells (Figure B; p=0.01). Cytokine plasma level monitoring, performed on day +7 and +30 after effector cell infusion in the absence of leukemia persistence (as evidenced by a lack of bioluminescence signal), showed that mice engrafted with CD19+ leukemia and treated with CAR.CD19-NK cells have lower levels of circulating hIFN-g cytokine compared to mice treated with CAR.CD19-T cells at both day 7 (42±82 vs 330±346 ng/ml; p=0.05) and day 30 (0.9±0.7 vs 4148±667 ng/ml; p=0.05). These in vitro and in vivo data demonstrate the feasibility of clinical scale feeder-free expansion of non-modified NK cells and stably transduced CAR-NK cells. Both non-modified and gene-modified cells were capable of significant tumor killing, suggesting a multi-modal adoptive cell approach to treatment of leukemia. Since NK cells have been shown to be safely used in third-party setting (St. Jude Children's Research Hospital, USA; NCT00640796), we suggest that ex-vivo expanded, feeder-free NK cells can be universally applied for 'off-the-shelf' immuno-gene-therapy, and that their innate allo-reactivity can be safely harnessed to potentiate allogeneic cell therapy. Figure. Figure. Disclosures Locatelli: Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees; bluebird bio: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Miltenyi: Honoraria; Bellicum: Consultancy, Membership on an entity's Board of Directors or advisory committees.

Prognosis of a significant proportion of patients with chemotherapy-refractory or multiply-relapsed CD30+ Non-Hodgkin's Lymphoma (NHL) or Hodgkin lymphoma (HL) still remain poor. Targeting CD30 with monoclonal antibodies in HL and anaplastic large cell lymphoma was shown to induce remarkable clinical activity; however, occurrence of adverse events (mainly neuropathy) may result into treatment discontinuation in many patients. Immunotherapeutic approaches targeting CD30 by chimeric antigen receptor (CAR) has been demonstrated to be of value in two independent clinical trials, although clinical benefit was sub-optimal. We designed a new CAR construct characterized by an anti-CD30 single-chain variable-fragment cassette (AC10), linked to CD3ζ by the signaling domains of two costimulatory molecules, namely either CD28.4-1BB or CD28.OX40. The inducible Caspase-9 (iCasp9) safety switch was included in both constructs with the goal of promptly controlling undue toxicity. As a selectable marker, we added in frame the CD34 antigen. The in vitro anti-tumor efficacy was evaluated by using either the NHL cell line: Karpas299, or the HL cell lines: L428, in both short-term cytotoxic assay (51Cr release assays) and long-term co-cultures for 6 days. Supernatant from co-culture experiments was analyzed by Elisa. We assessed the antitumor effect of CAR.CD30 T cells in a in vivo NSG mouse model engrafted i.v. with lymphoma FF-luciferase cell lines Karpas299 or L428, and monitored tumor growth by IVIS Imaging system. For tumor re-challenging, mice of the NHL model surviving until day +140, were i.v. infused with 0.2x106 Karpas299 cells, and subsequently followed for additional 110 days. Persistence of CAR.CD30 T cells was evaluated, together with a deep characterization of memory profile of T cells. Independently from the costimulatory domains CD28.OX40 or CD28.4-1BB, the generated retroviral vectors showed similar transduction efficiency of T cells (86.5±5.1% and 79.3±5.3%, respectively). Nevertheless, CD28.OX40 costimulatory domains was associated with more stable expression of the CAR over time, during extensive in vitro culture (84.72±5.30% vs 63.98±11.51% CD28.4-1BB CAR T cells at 30 days after transduction; p=0.002). For both CAR constructs, we did not observe any significant difference in the suicide gene iCasp9 activity, both in vitro and in vivo. In short-term cytotoxic assay, both CAR.CD30 T cells significantly and specifically lysed CD30+ NHL and HL tumor cell lines. In long-term co-culture, CD28.OX40 showed a superior anti-lymphoma in vitro activity as compared to CD28.41BB T cells, when challenged at very high tumor/effector ratio (8:1) (for Karpas 299; p=0.03). Moreover, the antigen stimulation was associated to higher levels of Th1 cytokine production, with CD28.OX40 T cells secreting a significantly higher amount of IFNγ, IL2 and TNFα as compared to CD28.41BB T cells (p= 0.040; p=0.008; p=0.02; respectively). Bioluminescence in HL (L428) tumor-bearing mice, treated with NT T cells, rapidly increased up to 5 log in less than 50 days and mice either died or were sacrificed due to morbidity. The best outcome was observed in mice treated with CD28.OX40, as three out of five mice were still alive at the experimental end-point of day+165, as compared with mice treated with CD28.4-1BB (60% vs 0%, p=0.0021). In NHL (Karpas 299) mouse models, CD28.OX40 had an extensive anti-tumor control superior to that of CD28.41BB T cells, leading to a significant reduction of tumor bioluminescence at day 45 (3.32x10 vs 2.29x10, p=0.04). The median survival of mice treated with NT and CD28.4-1BB CAR T cells was 45.5 and 58 days respectively, but undetermined for mice treated with CD28.OX40 CAR T cells (p=0.0002). After 140 days, cured mice were re-challenged with Karpas 299; mice were followed for other 100 days. Bioluminescence analysis showed rapid progression of the tumor in the control mice cohort, as well as in CD28.4-1BB treated mice. In contrast, in CD28.OX40 treated mice, at day+240 days, 4 out of 6 mice were tumor-free, resulting into a statistically significant survival benefit (p=0.0014). Only in mice treated with 28.OX40 T cells, we observed a long-lasting persistence of circulating CAR-T cells up to day +221. In summary, we have developed a novel CAR.CD30 construct displaying features that make it a particularly suitable candidate for a clinical trial in patients suffering from CD30+ tumors. Disclosures Merli: Novartis: Honoraria; Sobi: Consultancy; Amgen: Honoraria; Bellicum: Consultancy. Locatelli:Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Bellicum: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; BluebirdBio: Consultancy; Miltenyi: Honoraria; Amgen: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

In view of the exciting results reported in patients with CD19+ malignancies given CAR T cells, it is expected that a continuously growing number of patients will be offered this treatment and, thus, will be exposed to gene-modified products. Since the techniques of gene manipulation are relatively new, some of the risks associated to CAR T therapy may be unpredictable. Recently, two patients who relapsed with CD19-, CAR-expressing leukemia were reported, this observation being interpretable in light of an inadvertent leukemic cell transduction with the second generation CAR.CD19 lentivirus during CAR T cell manufacturing (Lacey, ASH, 2016 128:281). Immunoglobulin heavy chain sequencing analysis of 17 additional infusion products also identified the leukemic clonotypes in six additional products (86%). In vitro and in vivo experiments proved that these CAR+ leukemic clones were not killed by CAR.CD19 T cells (Ruella, ASH, 2017 130:4463). Since lentiviruses proved to be superior for transduction of quiescent hematopietic stem cells due to their ability to infect non-dividing cells, we hypothesized that CAR-T cell manufacturing based on the genetic modification of T cells by gammaretroviral vector could theoretically represent a safe approach. Peripheral blood or bone marrow (BM)-derived mononuclear cells of patients with >40% of blasts at diagnosis (CD45dim+/CD34+/CD19+/CD22+/CD10+), were transduced with a retrovirus encoding for a second generation CAR.CD19.41bb.z in frame with a suicide gene (i.e., inducible caspase 9, iC9) employed in the academic Clinical Trial (NCT03373071) run at the Bambino Gesù Children's Hospital, Rome, Italy. Patient-derived CAR-T cells showed a phenotype not significantly different from that found on CAR-T cells generated by healthy-donors (data not shown). In particular, we demonstrated that both flow-cytofluorimetry and RealTime-quantitative PCR (with a sensitivity up to 10-5) failed to identify leukemic cells in the final CAR-T cell products generated from Bcp-ALL patients. To generate an in vitro model of CAR+ leukemic cells, we genetically modified CD19+ RAJI and DAUDI cell lines with the bicistronic retroviral vector carrying both second generation CAR.CD19 and the suicide gene iC9 (iC9.CAR-RAJI and iC9.CAR-DAUDI). We demonstrated the possibility of promptlyeliminating CAR+ leukemic cells, through exposure to 20nM of AP1903 of iC9.CAR-DAUDI and iC9.CAR-RAJI cells. Indeed, very early activation (6 hours) of the suicide gene iC9 resulted into a significant reduction in the percentage of CAR+ RAJI leukemic cells (Fig.A). The presence of iC9.CAR.CD19 molecule on leukemic cells precluded the detection of the CD19 antigen, whereas cells retain the expression of all other specific B-lineage markers. CD19 antigen started to be detectable 72 hours after AP1903 exposurewhen CAR negative leukemic cells become preponderant. To demonstrate that CD19 antigen was not down-regulated, but only masked by CAR molecule in iC9.CAR-RAJI and iC9.CAR-DAUDI cell lines, we measured CD19 mRNA, showing no significant modification with respect to wild-type (WT) RAJI and DAUDI cell lines. Moreover, iC9.CAR-RAJI and iC9.CAR-DAUDI cell lines were effectively eliminated by CAR.CD19 T cells (12.5±13.7% and 3.4±4.3% residualleukaemia, respectively) at the same extent of WT cell line (0% and 0.08±0.1%, residual leukaemia, respectively; p>0.05 Fig.B). To assess if patient-derived iC9.CAR.CD19 T cells were able to generate leukemia in vivo mouse model, NSG female mice were infused i.v. with 10x106 CAR-T cells and control NT-T cells. Mice were monitored for a total period of 250 days, by recurrent bleed. Simultaneously, another cohort of mice was infused with patient-derived BM cells (5x106) and monitored for the same time. Mice infused with Bcp-ALL BM cells developed leukemia-phenotype,with 82% of cells expressing hCD45dim and hCD19. By contrast, mice receiving patient-derived CAR-T cells showed a lowpercentage of CD45+ cells (0.1±0.01%), all CD3+. Despite the long period of observation, we did not detect any expansion of hCD19+ cells in this animal cohort. Taken together these data suggest that the use of a retroviral platform, associated with the presence of iC9 suicide gene, contributes to the genesis of a highly functional and safe CAR-T product, even when the production starts from a biological material characterized by high contamination of leukemic blasts. Disclosures Locatelli: Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Miltenyi: Honoraria; bluebird bio: Consultancy; Bellicum: Consultancy, Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees.

We previously showed that Bcl-xL and Mcl-1 cooperatively protect platinum-resistant ovarian cancer cells from apoptosis. Here we assessed the anticancer potential of combining ABT-737-induced inhibition of Bcl-xL with Mcl-1 inhibition via PI3K/Akt/mTOR pathway disruption using NVP-BEZ235. NVP-BEZ235 inhibited cell proliferation without inducing apoptosis. It strongly repressed Mcl-1 expression and induced Puma expression in both cell lines tested while differentially modulating Bim between the two. Interestingly, NVP-BEZ235 efficiently sensitized ovarian carcinoma cells to ABT-737, provided that Bim expression was induced. Moreover, inhibiting the ERK1/2 pathway restored Bim expression and sensitized low Bim-expressing cancer cells to the BEZ235/ABT-737 treatment.