ALBERT

Description: Human endogenous retroviruses (HERVs) are ancient viruses forming 8% of human genome. One subset of HERVs, the HERV-K has recently been found to be expressed on tumor cells including melanoma, breast cancer and lymphoma but not on normal body cells. Thus, targeting HERV-K protein as a tumor associated antigen (TAA) may be a potential treatment strategy for tumors that are resistant to conventional therapies. One approach to improve therapeutic outcome is by infusing T cells rendered specific for such TAAs preferentially expressed on tumor cells. Recognition of cell-surface TAAs independent of major histocompatibility complex can be achieved by introducing a chimeric antigen receptor (CAR) on T cells using gene therapy. This approach is currently being used in our clinical trials adoptively transferring CD19-specific CAR+ T cells into patients with B-lineage malignancies. Preliminary analysis of HERV-K env protein expression in 268 melanoma samples and 139 normal organ donor tissues using immunohistochemistry demonstrated antigen expression in tumor cells and absence of expression in normal organ tissues. The scFv region from a mouse monoclonal antibody to target HERV-K env was used to generate a CAR and cloned into Sleeping Beauty (SB) plasmid for stable expression in T cells. HERV-K-specific CAR+T cells were selectively propagated ex vivo on artificial antigen presenting cells (aAPC) using an approach already in our clinical trials. Indeed, after genetic modification of T cells and selection on HERV-K+ aAPC, over 95% of propagated T cells stably expressed the introduced HERV-K-specific CAR and exhibited redirected specificity for HERV-K+ melanoma (Figure 1). Further, the adoptive transfer of HERV-K-specific CAR+T cells killed metastatic melanoma in a mouse xenograph model. While we have chosen melanoma as our tumor model, this study has the potential to be applied to other malignancies, including lymphoma and myeloma due to restricted expression of HERV-K envelope (env) protein on these tumor cells. These data demonstrate that it is feasible to generate T cells expressing a HERV-K-specific CAR using a clinically-appealing approach as a treatment strategy for HERV-K env+ tumors. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2461 Background: Fludarabine, rituximab, and cyclophosphamide combinations have shown high CR rates in CLL, but concerns remain about both short and long term toxicity. Lenalidomide has emerged as a potential alternative addition to the fludarabine/rituximab backbone that may increase anti-tumor activity. The purpose of this Phase I/II study was to determine optimal lenalidomide dosing and to evaluate its potential benefits when used in combination with fludarabine/rituximab. Methods: Eligible patients (pts) had untreated Rai stage III/IV or symptomatic stage 0-II B cell CLL with no CNS involvement, ECOG PS 0–2, and adequate organ function. The Phase I portion of this trial (n=19) explored fixed doses of fludarabine (25mg/m2/IV on days 1, 2, and 3 every 28-day cycle) and rituximab (375mg/m2/IV in divided dose days 1 and 2 cycle 1, followed by 500mg/m2/IV day 1 cycles 2–6) while receiving one of two dose levels (DLs) of lenalidomide. Pts on DL1 received 2.5mg PO on days 8–28 of cycles 1–6; those on DL2 received 2.5mg PO on days 8–28 of cycle 1 and 5.0mg on days 8–28 of cycles 2–6. Tumor lysis syndrome prophylaxis was administered to all pts throughout the first 2 cycles (allopurinol, 300mg PO daily). Disease assessment (NCI WG Guidelines) occurred post-cycle 3 during active treatment, 2 months after completion of the last treatment cycle (≤6 cycles), and every 6 months in follow-up until disease progression. Results: Between 2/2008 and 5/2010, 28 pts were enrolled; the first 26 are included in this analysis. Pts were all untreated, and 64% male with median age 66.5 yrs (range: 48–82 yrs) and Rai stage 0/I/II/III/IV of 2/10/7/4/3. For the Phase I portion of the trial, the original lenalidomide dosing schedule specified that treatment begin on day 1 concurrently with fludarabine/rituximab, continuing for 21 days. Of the first 4 pts enrolled, 2 experienced persistent grade (g) 3 rash and/or dose-limiting toxicity (DLT), including g4 febrile neutropenia. Only 1 pt completed all 6 cycles of therapy but still 3 achieved PR; 1 came off study prior to disease evaluation (DLT). Due to concerns about toxicity in these first 4 pts, the protocol was amended to delay lenalidomide until days 8–28. Results for DL1 pts (n=6) from the Phase I portion of the amended protocol are as follows: 5 pts completed 6 cycles of therapy and achieved an objective response rate (ORR) of 83% (CR, 4; PR, 1); the remaining pt had SD. Median time-to-CR was 34.0 weeks (range 32.4–35.4). Of these 6 pts, the most common DL1 g3/4 toxicities included neutropenia (4), leukopenia (4), and rash (1); 1 pt was hospitalized for g3 upper respiratory infection (possibly-related). The maximum tolerated dose of lenalidomide (5.0mg, DL2) is currently being studied within the Phase II portion of this study. Currently, the findings for DL2 (Phases I/II pooled, n=16) are as follows: 5 pts were unevaluable (2 too early for assessment; 3 due to toxicity). Two pts achieved CR; 5 pts are PR, 2 of which are still currently receiving treatment. Four pts are stable, including 2 who are still on-study. In DL2 pts the most common g3/4 toxicities were neutropenia (10), leukopenia (5), fatigue (3), anorexia (2) and rash (2). Five pts on DL2 were hospitalized, including 1 with g3 tumor lysis syndrome while receiving prophylaxis (related) and another with g3 allergic reaction/hypersensitivity (possibly-related); 1 pt later expired from g5 diarrhea (possibly-related). Conclusion: The activity of lenalidomide, when added to fludarabine/rituximab, appears promising, despite notable toxicity. The dosing schedule of lenalidomide in combination with fludarabine/rituximab appears to influence the overall toxicity of the three-drug regimen, as evidenced by the improvement in tolerability with sequential dosing. Phase II study is ongoing which will further characterize the toxicity and efficacy of this regimen in pts with CLL. Disclosures: Flinn: Celgene: Research Funding; Genentech: Research Funding. Off Label Use: This study uses lenalidomide off-label, in combination with fludarabine/rituximab, as an investigational treatment for patients with chronic lymphocytic leukemia. Cooper: Bristol-Myers Squibb: Consultancy.

Description: Allogeneic hematopoietic cell transplantation (HCT) is well established as a clinical means to treat patients with hematologic disorders and cancer. Human cord blood (CB) is a viable source of hematopoietic stem cells (HSC) for transplantation. However, numbers of nucleated cells retrieved, as well as limited numbers of HSC/progenitor cells (HPC) present, during collection may be problematic for treatment of adult patients with single CB HCT. One means to address the problem of limiting numbers of HSC/HPC is to ex vivo expand these cells for potential clinical use. While progress has been made in this endeavor, there is still a clinically relevant need for additional means to ex vivo expansion of human HSC and HPC. OCT4, a transcriptional factor, plays an essential role in pluripotency and somatic cell reprogramming, however, the functions of OCT4 in HSC are largely unexplored. We hypothesized that OCT4 is involved in HSC function and expansion, and thus we first evaluated the effects of OAC1 (Oct4-activating compound 1) on ex vivo culture of CB CD34+ cells in the presence of a cocktail of cytokines (SCF, TPO and Flt3L) known to ex vivo expand human HSC. We found that CB CD34+ cells treated with OAC1 for 4 days showed a significant increase (2.8 fold increase, p

Description: Hematopoietic stem cells (HSCs) are administered (i) to restore hematopoiesis and immunity in the course of hematopoietic stem-cell transplantation (HSCT), (ii) as a replacement for inherited blood disorders and bone marrow failure, (iii) to regenerate cells of alternative lineages for restorative medicine, and (iv) as a source for generating specific hematopoietic cells (e.g., T cells, NK cells, and dendritic cells). However, the widespread application of allogeneic HSCs for humans is hampered by their immune-mediated destruction by host T cells recognizing mismatched HLA or by HLA-specific antibodies. Despite pre-banking umbilical cord blood (UCB) units and access to adult donors through the National Marrow Donor Program (NMDP), finding a suitable HLA-matched product is challenging for many recipients, especially those from ethnic minorities who are under-represented in the donor pool. The available donor pool would be markedly increased if donor HSCs were edited to eliminate expression of the HLA-A locus. Indeed, modeling from NMDP shows that the chance of an African American recipient finding a HLA-matched donor increases from 18% to 73% when matched for HLA-B, C and DR, instead of HLA-A, B, C and DR. We have previously shown that engineered zinc finger nucleases (ZFNs) can disrupt HLA-A expression in genetically edited T cells (Blood 2013). To extend this proof-of-concept to HSCs, we sought to disrupt HLA-A expression by introducing ZFNs targeting this locus. CD34+lineageneg HSCs (99% purity) were isolated using paramagnetic beads from UCB. Electro-transfer of in vitro transcribed mRNA encoding the HLA-A-specific ZFN generated 30% HLA-Aneg HSCs after one week ex vivo culture with defined cytokines (FLT3-L, SCF, TPO, and IL-6) and an aryl hydrocarbon receptor antagonist (stem reginin-1, SR-1). As expected, SR1 treatment maintained greater numbers of CD34+ cells (also CD34posCD38neg) in culture compared to controls. DNA sequence analysis revealed that HLA-Aneg HSCs encode the expected nucleotide changes at the ZFN target site. An in vivo engraftment assay, using NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) mice, demonstrated that HLA-Aneg HSCs maintain the capability of engraftment and differentiation into HLA-Aneg hematopoietic cells (Figure). Thus, disruption of HLA-A expression in HSCs provides an appealing approach to increasing the chances for of finding HLA-matched donors and may broaden the clinical application of allogeneic HSCT. Furthermore, the ability to genetically edit HSCs has implications for (i) preventing immune-mediated recognition of HLA-disparate HSC and (ii) preventing immune mediated recognition of self-antigens. Engraftment of HLA-A2neg HSCs was evaluated in vivo. Data shown are flow-cytometry analysis of bone marrow obtained from NSG mice 16 weeks after HSC injection. HSC engraftment and HLA-A2 expression in NSG mice injected with un-modified HSCs (left panel) and HSCs treated with the HLA-A specific ZFNs (right panel) are shown. Data are gated on human CD45 positive cells. Figure Engraftment of HSCs modified by the HLA-A specific ZFNs in NSG mice. Figure. Engraftment of HSCs modified by the HLA-A specific ZFNs in NSG mice. Disclosures: Reik: Sangamo BioSciences: Employment. Holmes:Sangamo BioSciences: Employment. Gregory:Sangamo BioSciences: Employment.

Description: Germline mutations in GATA2, a gene that encodes for transcription factors involved in hematopoiesis and vascular development, have recently been described in MonoMAC syndrome, Emberger syndrome and in select cases of mild chronic neutropenia. These disorders are unified by their predisposition to myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Patients with MonoMAC syndrome have also been noted to display monosomy 7 in their bone marrows in up to 50% of cases. Overexpression of GATA2 due to somatic mutations in cases of de novo pediatric AML, has also been shown to be a negative predictor of outcome. Juvenile myelomonocytic leukemia is a rare childhood malignancy with overlapping features of MDS and myeloproliferative neoplasm (MPN) that can transform to AML and is characterized by hyperactive RAS signaling. Mutations in NF1, NRAS, KRAS, PTPN11, and CBL are found in 85-90% of newly diagnosed patients, and monosomy 7 is the most common recurrent karyotypic abnormality seen in JMML. We therefore hypothesized that mutations in GATA2 may play a role in the development of JMML. Samples from 57 patients with JMML were screened for GATA2 mutations. Patient samples and clinical data were collected from the Children's Oncology Group (COG) trial AAML0122. DNA was extracted as per previous protocols from peripheral blood or bone marrow and whole genome amplified using Qiagen REPLI-g kit according to manufacturer specifications. We performed bidirectional Sanger sequencing (Beckman Coulter Genomics) of the entire coding region of GATA2 (NM_001145661.1) and aligned the sequences using CLC Workbench software (CLC Bio, Aarhus, Denmark). Only missense, splice site or nonsense mutations were evaluated using SIFT (Sorting Tolerant From Intolerant) to predict the impact on the structure and function of identified mutations on the protein. Patient J384 was found to have a nonsense point mutation at c.988C〉T (R330X) in the N-terminal region of the zinc finger portion of the protein (Figure 1a). This hotspot mutation has been reported in several patients with mild chronic neutropenia who displayed a predisposition to developing MDS and AML. The patient was also found to have a missense point mutation at c.962T〉G (L321R) predicted to be damaging by SIFT. Subcloning of the gene using a TA cloning kit with pCR 2.1 vector (Invitrogen), followed by direct sequencing of individual colony picks, revealed that the two sequence variants only occurred in a trans configuration. Out of 40 amplicons sequenced, 20 were found to have the c.988C〉T transition, 16 were found to be have the c.962T〉G variant, and four were found to be wild type. We therefore hypothesize that the c.988C〉T was inherited as a germline event and that c.962T〉G was somatically acquired in the majority of the remaining wild type alleles. No other point mutations or insertions/deletions were discovered in this cohort.Figure 1Identification of 2 distinct GATA2 mutations in patient J384.Figure 1. Identification of 2 distinct GATA2 mutations in patient J384. This patient was previously identified to have a KRAS G12D mutation (c.35G〉A) as well as monosomy 7. This patient died prior to undergoing transplant within months of diagnosis. While the patient technically met criteria for the diagnosis of JMML, it should be noted there were several atypical features, including older age at diagnosis (4 years and 10 months), and absence of hypersensitivity in myeloid progenitor cells to the cytokine granulocyte–macrophage colony stimulating factor (GM-CSF) in colony assay. This raises the possibility that patient J384 actually had MonoMAC syndrome with MDS and not JMML. This represents the first description of a GATA2 mutation in a patient suspected of having JMML. To our knowledge, this is the first report of a biallelic mutation in GATA2, combining a germline mutation with somatic acquisition. In addition, MonoMAC syndrome has not been reported to be associated with KRAS mutations to date. GATA2 mutations should therefore be considered in patients with atypical features of MDS or JMML. Panel (a) Bidirectional sequencing of patient sample J384 revealed two distinct sequence variants in both the forward (shown here) and reverse strands. Panel (b) Sequencing of 40 individual colony picks revealed that each sequence variant occurred in a trans configuration (CP 9 and CP13 are shown here as examples). In addition, 10% of colony picks (i.e. CP 32) revealed a wild type sequence, indicating that at least one of the two variants was a somatic event. Disclosures: No relevant conflicts of interest to declare.

Description: TRM, drug toxicities, life-threatening infections, poor quality of life, and graft versus host disease (GVHD) are significant risks of hematopoietic cell transplantation (HCT). In addition to the risks of HCT, existence of pre-transplant comorbidities can have significant impact on transplant outcomes. A comparison of CCI and the HCT-CI revealed that HCT-CI was prognostic and better at predicting TRM and overall survival (OS). However, the HCT-CI and CCI results have been inconsistent in predicting the TRM and OS in validation studies. In subgroup analyses of a large retrospective trial, HCT-CI did not predict TRM or OS for UCB recipients. Patient heterogeneity in age, disease, disease-risk, comorbidities, and conditioning regimens may have limited validation in these UCB recipients. The design of this research was to explore whether or not the HCT-CI and/or CCI can accurately predict post-transplant outcomes in young patients with high risk hematologic malignancies undergoing uniform RIC UCB transplantation. A retrospective chart review was performed on 52 consecutive young (age 〈 55) UCB transplant recipients receiving the RIC regimen fludarabine, cyclophosphamide, ATG, and 200 cGy TBI. All patients had received cyclosporine and mycophenolate mofetil for GVHD prophylaxis. Information on pre-transplant comorbidities was obtained from the CIBMTR pre-TED form for each patient and retrospective chart reviews. Demographic information, ECOG performance status (PS), identification of comorbidities, and post-transplant outcomes were obtained. Between 2005 and 2011, 52 patients age 19 – 54 (median 38) years, 30 males (58%) and 22 females (42%), underwent RIC UCB transplantation with the above regimen. Most patients had advanced stage or high risk hematologic malignancies; 28 patients had MDS/AML (54%), 8 patients had ALL (15%), while 16 patients had other hematologic malignancies. 39 patients had a PS = 0, while 13 had a PS = 1. Half of the patients were in CR 〉 2, with the same number receiving more than 2 prior therapies. 8 patients had received a prior autologous HCT, 8 a prior allo-HCT, and one patient had failed a previous UCD transplantation. 11 patients did not achieve engraftment. Median time to neutrophil engraftment was 26 days (95% CI: 24 – 28 d) and median time to platelets engraftment was 38 days (95% CI: 31 – 45 d). No patient developed grade 4 acute GVHD. Grade 3 acute GVHD was seen in 12 / 52 patients (23%). Chronic GVHD was seen in likewise seen in 12 patients (23%). To date 16 patients (30%) have relapsed. OS was 42% and PS was 37% at one year. Median PFS and median OS were 5 months and 7 months, respectively (PFS 95% CI: 1 – 11 months, OS 95% CI: 1 – 13 months). Neither CCI, nor HCT-CI were significant predictors of OS and PFS, however the ECOG PS was significantly associated with an improved OS and PFS (Table).OSPFSN1 yr2 yrs4 yrsp1 yr2 yrs4 yrspECOG PS03946%36%21%0.02644%36%21%0.0311331%0015%00CCI≤ 23944%26%10%0.2238%26%10%0.2〉 21338%31%31%38%31%31%HCT-CI02236%31%5%0.3632%32%5%0.3211443%14%14%29%14%14%≥21643%31%31%44%31%31%TRM occurred in 18 patients (35%). Neither comorbidity indices, nor the performance status were correlated to TRM (p 〉 0.05). For ECOG PS, OR 1.6 p = 0.49, CI 0.42 - 5.93; CCI OR = 2.4, p = 0.09, CI 0.86 - 6.59, HCT-CI OR = 0.6, p = 0.46, CI 0.19 - 2.08. In a previous presentation our group concluded that CCI performs better than HCT-CI in elderly patients undergoing UCB transplant. When tested in a population younger than 55 years old though, we were unable to validate the generally accepted prognostic indices. It is unclear whether this is due to a selection bias: it appears most of the patients were essentially health from a non-hematological stand-point, and as such the “classic” cardiac or pulmonary risk factors were underrepresented. The excess TRM is likely due to the severity of their underlying disease, and the majority were in CR ≥ 2 and had received multiple previous courses of antineoplastic chemotherapy. As umbilical cord transplants become more utilized, a new prognostic index needs to be developed, more fitting to the patients 〈 55 years old. Disclosures: No relevant conflicts of interest to declare.

Description: Background Myelofibrosis (MF) is a myeloid neoplasm marked by clonal proliferation of myeloid cells with marrow fibrosis and subsequent clinical findings. Often, MF is noted on original bone marrow biopsy, and is termed idiopathic myelofibrosis (iMF). Many patients with ET and PV later develop MF, and this is termed, PET-MF, or PPV-MF. The JAK2V617F mutation is seen in most PV patients (95%) and no more than 50% of patients with ET and MF. There are neither existing published data clearly heralding the JAK2V617F mutation frequency in PET/PPV-MF, nor is there clear evidence that PET/PPV-MF and iMF can be distinguished via mutations within the JAK-STAT pathway, or other frequently seen mutations in MF (eg ASLX1, TET2, EZH2, etc). Methods Array-based comparative genomic hybridization (aCGH) was performed on genomic DNA extracted from marrow aspirate using an Agilent 180K oligonucleotide array platform in order to discover recurrent and cooperating genetic aberrations, and gain insight into the hierarchy of molecular pathogenesis of fibrosis. BM aspirate from 11 pts with symptomatic MF were analyzed. Copy number (CN) alterations were compared to a reference set and mapped to functional genes. Results aCGH yielded CN losses in a divergent set of genes among patients with PET/PPV-MF (8 total patients) and iMF (3 total patients). CN losses were detected in RET (4/8), CIC (5/8), CD79a (5/8), NFKB (6/8), and MLLT (3/8), and CN gain of TRIP11 (2/8) more commonly in PET/PPV-MF, and only rarely in iMF. ASXL1, IGF2, INS, MAFB, and TOP1 were aberrated with CN loss in iMF in 67% of patients, and completely absent in PET/PPV-MF. In addition, there was suggestion of differential CN alteration between patients with clear history of PET-MF or PPV-MF. Aberrancy within CRTC1, ELL, RECQL4 was seen exclusively in PET-MF, and not seen in PPV-MF or iMF; ELN aberrancy was seen exclusively in PPV-MF. The cohort will be updated with additional patients at the meeting. Conclusions Myelofibrosis is a complicated disorder characterized by a wide amalgam of pathologic and clinical findings. Mutational analysis, as of yet, has not clearly provided a means to differentiate or subclassify myelofibrosis. While CN variations seen in this small cohort occurred in some genes thought to be involved in the pathogenesis of fibrosis (eg ELN, IGF2, INS), implicated in tumorigenesis (eg RECQL4, CD79a, ASXL1, MLLT etc), or associated with hematopoiesis (eg MAFB, NFKB), the relationship between the abnormalities seen in this analysis and the biology of these cases of MF is not clear. Nonetheless, this preliminary analysis provides some insight into the possible use of aCGH as a tool to help molecularly classify MF. Disclosures: Hockett: CombiMatrix Mol Diagn: Employment.

Description: Background T cells can be genetically modified ex vivo to redirect specificity upon enforced expression of a chimeric antigen receptor (CAR) that recognizes tumor-associated antigen (TAA) independent of human leukocyte antigen. We report a new approach to non-viral gene transfer using the Sleeping Beauty (SB) transposon/transposase system to stably express a 2nd generation CD19-specific CAR- (designated CD19RCD28 that activates via CD3z/CD28) in autologous and allogeneic T cells manufactured in compliance with current good manufacturing practice (cGMP) for Phase I/II trials. Methods T cells were electroporated using a Nucleofector device to synchronously introduce DNA plasmids coding for SB transposon (CD19RCD28) and hyperactive SB transposase (SB11). T cells stably expressing the CAR were retrieved over 28 days of co-culture by recursive additions of g-irradiated artificial antigen presenting cells (aAPC) in presence of soluble recombinant interleukin (IL)-2 and IL-21. The aAPC (designated clone #4) were derived from K562 cells and genetically modified to co-express the TAA CD19 as well as the co-stimulatory molecules CD86, CD137L, and a membrane-bound protein of IL-15. The dual platforms of the SB system and aAPC are illustrated in figure below. Results To date we have enrolled and manufactured product for 25 patients with multiply-relapsed ALL (n=12) or B-cell lymphoma (n=13) on three investigator-initiated trials at MD Anderson Cancer Center to administer thawed patient- and donor-derived CD19-specific T cells as planned infusions in the adjuvant setting after autologous (n=7), allogeneic adult (n=14) or umbilical cord (n=4) hematopoietic stem-cell transplantation (HSCT). Each clinical-grade T-cell product was subjected to a battery of in-process testing to complement release testing under CLIA. Currently, five patients have been infused with the CAR+ T cells following allogeneic HSCT, including one patient with cord blood-derived T cells (ALL, n=4; NHL, n=1), beginning at a dose of 106 and escalating to 107 modified T cells/m2. Three patients treated at the first dose level of 106 T cells/m2 have progressed; the patient treated at the next dose level with 107 T cells/m2 remains in remission at 5 months following HSCT. Assessment for response too early for patient treated with UCB T cells. Four patients with non-Hodgkin’s lymphoma have been treated with patient-derived modified T cells following autologous HSCT at a dose of 5x107 T cells/m2, and all patients remain in remission at 3 months following HSCT. No acute or late toxicities have been noted to date. PCR testing for persistence of CAR-modified T cells is underway. Conclusion We report the first human application of the SB and aAPC systems to genetically modify clinical-grade cells. Importantly, infusing CD19-specific CAR+ T cells in the adjuvant HSCT setting and thus targeting minimal residual disease is feasible and safe, and may provide an effective approach for maintaining remission in patients with high risk, CD19+ lymphoid malignancies. Clinical data is accruing and will be updated at the meeting. This nimble manufacturing approach can be readily modified in a cost-effective manner to improve the availability, persistence and therapeutic potential of genetically modified T cells, as well as target tumor–associated antigens other than CD19. Disclosures: No relevant conflicts of interest to declare.

Description: Nuclear transcription factor Stat3 is important for proper regulation of hematopoietic stem cell (HSC) and hematopoietic progenitor cell (HPC) proliferation, survival, and cytokine signaling responses. A new, noncanonical role for Stat3 in mitochondrial function has been discovered recently. However, there is little information on the role(s) of mitochondrial Stat3 in HSC/HPC function, especially potential effects of Stat3/mitochondrial dysregulation in human diseases. We investigated hematopoietic cell–targeted deletion of the STAT3 gene in HSCs/HPCs with a focus on mitochondrial function. We found that STAT3−/− mice, which have a very shortened lifespan, dysfunctional/dysregulated mitochondrial function and excessive reactive oxygen species production in HSCs/HPCs that coincides with pronounced defects in function. These animals have a blood phenotype with similarities to premature aging and to human diseases of myelodysplastic syndrome and myeloproliferative neoplasms such as erythroid dysplasia, anemia, excessive myeloproliferation, and lymphomyeloid ratio shifts. We show herein that the lifespan of STAT3−/− animals is lengthened by treatment with a reactive oxygen species scavenger, which lessened the severity of the blood phenotype. These data suggest a need for more detailed studies of role(s) of Stat3 in HSC/HPC mitochondrial function in human diseases and raise the idea that mitochondrial Stat3 could be used as a potential therapeutic target.