Effect of CD34+ cell dose on neutrophil and platelet engraftment kinetics in haematopoietic stem cell transplantation – A single-centre experience

Objectives: Haematopoietic stem cell transplantation (SCT) is curative for a number of benign and malignant hematological disorders. CD34 expression on haematopoietic progenitor cells is used to assess stem cell content in peripheral blood stem cell and bone marrow grafts. This study evaluated the relationship between numbers of CD34+ cells infused per kg and the timing of neutrophil and platelet engraftment. Materials and Methods: The effect of cell dose was studied in consecutive HSCT patients transplanted between November 2008 and December 2017. Neutrophil engraftment was defined as the first of 2 consecutive days with an absolute neutrophil count >0.5 × 109/L and platelet engraftment as unsupported platelet count >20 × 109/L for 7 days. Results: Of a total of 131 patients, 26 (19.8%) underwent an autologous SCT, while 105 (80.2%) underwent an allogeneic SCT. The median CD34 dose infused in the auto-SCT group was 5.29 × 106 CD34+cells/kg (IQR = 2.95–10.98) and 6.42 × 106 CD34+cells/kg (IQR = 4.20–9.20) in the allo-SCT group (P = 0.773). The median time to neutrophil engraftment in the auto-SCT group was 11 days (range 9.5–12) and in the allo-SCT group was 15 days (range 13–17), P ≤ 0.001. The median time to platelet engraftment in both groups was similar (12 days). When patients were divided into three groups based on CD34 dose (<5, 5–8 and >8), no difference was observed in the time to ANC or platelet engraftment. Similarly, no differences in time to engraftment were noted in each quartile of CD34 dosage in auto- and allo-SCT. Conclusion: Thus, it was concluded that a cell dose of approximately 5 × 106/kg provides reasonably rapid engraftment, with no advantage seen for a higher cell dose of >5.

Developing a morphomics framework to optimize implant site-specific design parameters for islet macroencapsulation devices

Delivering a clinically impactful cell number is a major design challenge for cell macroencapsulation devices for Type 1 diabetes. It is important to understand the transplant site anatomy to design a device that is practical and that can achieve a sufficient cell dose. We identify the posterior rectus sheath plane as a potential implant site as it is easily accessible, can facilitate longitudinal monitoring of transplants, and can provide nutritive support for cell survival. We have investigated this space using morphomics across a representative patient cohort (642 participants) and have analysed the data in terms of gender, age and BMI. We used a shape optimization process to maximize the volume and identified that elliptical devices achieve a clinically impactful cell dose while meeting device manufacture and delivery requirements. This morphomics framework has the potential to significantly influence the design of future macroencapsulation devices to better suit the needs of patients.

Review of Unrelated Donor Cord Blood Transplantation in Children over the Past 3 Decades

Four decades ago, Hal Broxmeyer demonstrated that umbilical cord blood (CB) contained hematopoietic stem cells (HSC) and hypothesized that CB could be used as a source of donor HSC for rescue of myeloablated bone marrow. In 1988, Gluckman et al reported the first successful cord blood transplant (CBT) of a child with Fanconi Anemia using matched sibling CB. This patient survives and 35 years later still has durable hematopoiesis from the CB donor graft. In 1991, Rubinstein et al established an unrelated donor (UD) CB bank and in 1993 the first UD CBT was using a unit from this bank. Since that time, &gt;40,000 CBTs have been performed worldwide. We hypothesized that changes in cord blood banking (increased size, diversity, and quality of banked units enabling selection of units with higher cell doses and closer HLA matching) and in transplantation (less use of steroids, availability of newer therapies for prophylaxis and treatment of graft versus host disease [GVHD], improved antifungal and antiviral detection and therapeutics) have improved outcomes of CBT today. To address this hypothesis, we performed a retrospective study combining data from Eurocord and Duke University in a large cohort of children transplanted with a single UD CB unit (CBU) from 1993-2019. Standard transplant outcomes (overall survival [OS], disease free survival [DFS], acute and chronic GVHD, treatment related mortality [TRM], and relapse) and changes in outcomes over 3 time periods (1:&lt;2005, n=1297; 2:2005-2010, n=1735; and 3:&gt;2010, n=1802) were studied. Relative contributions of cell dose and HLA matching to transplant outcomes over time were assessed. A total of 4834 patients (4015 from Eurocord and 819 from Duke) were analyzed. The majority of patients, (59%, n=2839) had malignant diagnoses including 1422 with ALL, 887 with AML and 167 with MDS. Of the 1995 with non-malignant diagnoses, 761 had inborn errors of metabolism, 644 had primary immunodeficiency, 325 had a bone marrow failure syndrome and 206 had a histiocytic disorder. Half of the patients had positive serologies for CMV prior to transplant. The median age of the cohort fell from 5.2 to 3.25 years over time. In patients with malignancies, use of total body irradiation decreased over time. The median total nucleated cell (TNC) and CD34+ cell doses administered were 8.07x10e7 and 6.17x10e5 cells/kg and increased over time. HLA matching and transplantation of patients in earlier disease states also increased over time, p&lt;0.001 for both. The probability of 5-year OS in the entire cohort was 53.48% and improved over time: 42%; 57.4%; and 60.4%, in periods 1,2,3 respectively (p&lt;0.0001). OS improved with closer HLA matching, higher cell dose, myeloablative conditioning, and negative pre-transplant CMV serologies. For patients with malignancies, DFS increased and TRM and acute GVHD decreased over time. In contrast, leukemic relapse did not change throughout the years. OS was higher in patients with inborn errors of metabolism and also improved over time with 57.8% surviving before 2005, 69.4% from 2005-2010, and 71% after 2010 (p=0.0141). Similar results were seen in the cohort with immune deficiencies. In the entire cohort, the median time to neutrophil engraftment decreased from 25 days (period 1) to 19 days (period 3). In multivariate analysis for engraftment, a higher TNC dose (p=0.001) up to but not beyond the median cell dose (8.07x10e7 cells/kg), total body irradiation, and the use of ATG improved engraftment. Acute GVHD decreased from 35% before 2005 to 27.1% after 2010 (p=0.0556) while the incidence of chronic GvHD was stable. The use of ATG reduced the risk of acute GVHD and closer HLA matching reduced the risk of both acute and chronic GVHD. In this population of patients receiving high cell doses, outcomes were predominantly influenced by HLA matching and increasing cell dose did not abrogate HLA mismatching. In conclusion, we analyzed the largest cohort of pediatric patients undergoing CBT over the past 3 decades. OS, DFS and engraftment have improved over time accompanied by decreases in TRM and acute GVHD. Relapse and chronic GVHD were stable and remain low. These improvements are explained by the increased availability of high quality banked CBUs enabling selection of closer HLA matching and units with higher cell doses. The numbers of CBTs have decreased in the past decade, but these results support the ongoing use of CBT in children lacking matched related or unrelated donors. Figure 1 Figure 1. Disclosures Kurtzberg: Neurogene: Consultancy; CryoCell: Patents & Royalties: Duke licensed IP, and data and regulatory packages for manufacturing and use of cord blood and cord tissue MSCs in the treatment of patients with hypoxic ischemic encephalopathy, cerebral palsy, autism, acute ischemic stroke, COVID-ARDS, and COVID-MIS-C. ; Sinocell: Patents & Royalties: Duke licensed IP, data, and regulatory packages for use of autologous and sibling cord blood to treat children with cerebral palsy.; Celularity: Current holder of stock options in a privately-held company. Troy: SinoCell: Patents & Royalties; CryoCell: Patents & Royalties; Bristol Myers Squibb: Research Funding; Synthetic Biologics: Honoraria; Gamida Cell: Consultancy; The EMMES Corporation: Consultancy; The Community Data Roundtable: Consultancy; AegisCN: Consultancy.

The Determinative Role of Hematopoietic Stem and Progenitor Cell Graft Composition on the Engraftment Kinetics after HSCT

The importance of hematopoietic stem and progenitor cell (HSPC) dose in the outcome of hematopoietic stem cell transplant (HSCT) has been demonstrated by analyses of the threshold dose of CD34+ cells required to achieve donor engraftment, usually defined as endpoints of donor-derived neutrophil (PMN) or platelet (PLT) absolute numbers, and RBC transfusion independence. We recently described the heterogeneous HSPC composition of prospectively obtained umbilical cord blood samples and lack of correlation between CD34+ cell dose and the frequency of hematopoietic stem cells (HSC)(Mantri S et al, Blood Adv 2020). We describe here a pilot analysis of the relationship between HSPC graft composition and engraftment after HSCT. The study population is comprised of 17 children (3.4-22 years, median 13 years) treated with αβT/CD19B - cell depleted mobilized peripheral blood stem cell (PBSC) HSCT (αβhaplo-HSCT). Eight patients had acute lymphoblastic leukemia, seven had acute myeloid leukemia, and two myelodysplastic syndrome. All patients received serotherapy with Thymoglobulin and Rituximab, and a myeloablative conditioning consisting of either TBI 1200 cGy in combination with Fludarabine/Thiotepa (15 patients), Melphalan/Thiotepa (1 patient), or TBI (400 cGy) with Fludarabine/Thiotepa (1 patient receiving a second HSCT). Only 2/17 patients received 1-2 doses of G-CSF before Day+30. The CD34+ Lin- cells in the PBSC products were analyzed for HSPC composition, including HSC (CD38- CD90+ CD45RA-), Multipotent Progenitors (MPP, CD38- CD90- CD45RA-), Common Myeloid Progenitors (CMP, CD38+ CD123+ CD45RA-), Granulocyte-Monocyte Progenitors (GMP, CD38+ CD123+ CD45RA+), Megakaryocyte-Erythroid Progenitors (MEP, CD38+ CD123- CD45RA-) and Common Lymphoid Progenitors (CLP, CD38+ CD127+). Similar HSPC analyses were performed on marrow obtained at Days +30, +60, and +90 post-HSCT. The CD34+ cell dose in the grafts ranged from 8.5-40 x 10e6/kg recipient body weight (mean 15.6). The median time to Absolute Neutrophil Count (ANC) &gt; 500 or 1000/mm 3 were Days +12 and +14, respectively, and for PLT &gt; 50K or 100K/mm 3 Days +14 and +15, respectively. The time to either PMN or PLT engraftment did not correlate with the HSC dose. In contrast, the GMP dose was predictive of PMN engraftment: 7/8 patients receiving the highest GMP dose achieved ANC &gt; 500 at 0-3 days before the median day of engraftment, while PMN engraftment was delayed by 1-5 days beyond the median in 6 of the 9 receiving the lowest GMP dose (χ 2 = 8.14, p=0.004)(Fig A). Of the patients with the highest MEP dose, 7/8 achieved PLT &gt;50K/mm 3 0-4 days before the median, while 5/9 receiving the lowest dose engrafted 1-6 days beyond the median (χ 2 4.9, p=0.026) (Fig B). In the first 100 days post-HSCT, naïve CD4+ T-cells were all CD31+ CD45RA+ Recent Thymic Emigrants (RTE), indicating that they were newly produced T cells and not adoptively transferred from the αβhaplo-HSCT. The HSC dose did not correlate with the number of naïve CD4+ T cells until Day+180 (Fig C). Like PMN and PLT recovery, early T lymphoid recovery after HSCT was mainly derived from infused CLP, and likely switched to HSC-derived lymphopoiesis by Day +180. Consistent with this concept, HSC dose in the PBSC products was unrelated to the numbers of committed progenitors, e.g., CMP, MEP, CLP, in early (Day +30) post-transplant marrow samples. These data are consistent with clonal analyses of patients receiving lentiviral gene therapy and murine experiments demonstrating prolonged steady-state contribution of committed progenitors to peripheral blood cell maintenance (Biasco L et al, Cell Stem Cell 2016; Sun J et al, Nature 2014). While post-HSCT PMN or PLT numbers are frequently equated with "HSC engraftment" and naïve T-cell numbers with HSC-derived immune reconstitution, these early events reflect blood cell production by committed progenitors, which are variably present in the grafts. Although CD34+ cell dose is currently used to predict post-transplant engraftment and to qualify stem cell products for release, more accurate clinical predictions may be determined by the HSPC grafts' composition. Further, engineering of the progenitor composition of clinical HSCT products, e.g., co-transplantation of additional committed progenitors like GMP or CLP, could be strategically used to control post-transplant lymphohematopoietic recovery. Figure 1 Figure 1. Disclosures Parkman: Jasper Biotech: Consultancy. Bertaina: Cellevolve Bio: Membership on an entity's Board of Directors or advisory committees; Neovii: Membership on an entity's Board of Directors or advisory committees; AdicetBio: Membership on an entity's Board of Directors or advisory committees.

Validation and Implementation of the Use of Peripherally Inserted Central Catheter (PICC) for the Administration of Peripheral Blood Stem Cells (PBSC): A Single-Institution Experience

Background: Infusion of peripheral blood stem cell (PBSC) in patients undergoing autologous transplantation (ASCT) has been conventionally performed using central venous catheters (CVC) inserted through the subclavian or internal jugular vein. Peripheral inserted central catheters (PICC) are routinely used for infusion of blood products and medication, but its use for PBSC infusion has not been well established. Our study aimed to evaluate the feasibility and safety of using PICC to deliver PBSC for ASCT through an in-vitro lab-based validation process, followed by a clinical review. Methods: Lab based validation In vitro infusion of 6 cryopreserved PBSCs was performed, 3 infused PICC whist 3 via CVC. Each product was thawed for the same amount of time and drained by gravity. Pre-infusion and post-infusion total nucleated cell counts (TNC), CD34 counts and CD34 viability of the PBSCs were analysed by flow-cytometry and compared using paired T test. In vitro infusion rates were also compared between PICC and CVC groups. Clinical Outcome Analysis The clinical study included 31 patients (Lymphoma N=21, myeloma N=5, Others, N=4) who underwent ASCT at National University Cancer Institute, Singapore (NCIS) from September 2019 to July 2021. All patients had a 19G BARDS dual lumen PICC inserted in either the brachial or basilic veins and used for PBSC infusion. The PBSC infusion rate, infusion associated complications, time to absolute neutrophil count (ANC) &gt;1, and platelet count engraftment &gt;100K were analysed. Clinical outcomes in the lymphoma cohort, who received BEAM conditioning (N=17) were also compared with a control group, matched for conditioning, cell dose and age, who had PBSC infused via CVC. Results: In vitro findings: Overall flow rates for infusion through PICC was slower (mean 0.1mls/s vs 0.3mls/s, p &lt; 0.05). However, there were no significant % differences in TNC counts (5% vs 9%, p=0.4), CD34 counts (17% vs 15%, p=0.9) and viability (4% vs 7%, p=0.2) between pre and post infusion samples for PICC and CVC.. Clinical findings: 30 patients (Lymphoma N=21, myeloma N=5, N=4) were included. 15 (50% of patients) had a for ASCT while 15 (50%) had an existing PICC. For patients with an existing PICC, the median duration of catheter in situ was 86 days. New lines were inserted 2-7 days prior to the PBSC infusion. The median age of the patients was 54 (20-71) with 19 males (63%). . There were 5 infusion related complications, 2 in an existing PICC and 3 in new PICCs. 4 were related to slow flow rate and 1 was related to sediments seen in the line. None led to a need for alternative line for infusion. The median time to ANC recovery was 10 (range 9-14), 10 (range 9-11) and 11 days (range 10-12), while the median time to platelet engraftment was 18 (range 10-195), 20 (range 15-55) and 22 (16-85 days) for the lymphoma BEAM conditioning (N=17), lymphoma Carmustine/ Thiotepa conditioning (N=4), and the myeloma (N=5) cohorts respectively. Clinical outcomes in the lymphoma cohort, who also compared with a control group matched for conditioning, cell dose and conditioning. The in-vivo infusion rate was slower in the PICC group, compared to the CVC group (3.1 mls/min vs 4.5mls/min, p&lt;0.05).There was however no differences in engraftment with median time to ANC recovery 10 days (range 9-14) vs 11 days (range 9-13) (p&gt;0.05) and median time to platelet engraftment 18 (range 14-195) vs 19 days (range 14 -57) (p&gt;0.05) in the PICC vs CVC groups respectively. Conclusion: Our in-vitro and clinical findings confirmed that the use of PICC for PBSC infusion is safe and efficacious and reduces the need for CVC insertion. Our findings have led to change in clinical practice with utilization of PICCs for PBSC infusions for ASCT. Disclosures No relevant conflicts of interest to declare.

Haploidentical Natural Killer Cell Therapy As an Adjunct to Stem Cell Transplantation for Refractory Acute Myeloid Leukemia

Refractory acute myeloid leukemia (AML), defined as failure of 2 cycles of induction therapy at diagnosis or of 1 cycle at relapse, represents a subgroup with poor clinical outcomes. In our transplant cohort, the 5-year overall survival in this subgroup was 16% (Ganapule at al. JGO 2017). Haploidentical natural killer cell (NK) therapy is a strategy that is being explored in refractory malignancies. Our in-vitro and animal model data suggest that exposure to arsenic trioxide (ATO) results in enhanced NK cytotoxicity (Alex AA et al. Front. Immunol 2018). Historically, at our center, patients with refractory AML have been treated with cytoreductive therapy (FLAG ± idarubicin or mitoxantrone + etoposide for 3 to 5 days) followed by 1-week rest and then a reduced-intensity transplant with fludarabine + melphalan conditioning while in peak cytopenia. From February 2019, we initiated a phase II single arm clinical trial (CTRI/2019/02/017505) enrolling patients with refractory AML planned for a stem cell transplant in peak cytopenia. Patients received CD56-positive cells from an HLA haploidentical related family donor (other than the stem cell donor, wherever feasible) following cytoreductive chemotherapy. The NK cell donor preference strategy included presence of KIR ligand mismatch, greater number of KIR B motifs (or the B score), lower donor age, and negative donor specific antibodies tested using flowcytometry crossmatch (Figure 1a). CD56-positive selection was done using CliniMACS prodigy system. This was followed by overnight incubation of the CD56 positive cells in autologous plasma with 2 micromolar ATO and 500 U/mL of interleukin-2. The CD56 positive cells were then infused to the patient 1-day after the completion of cytoreductive chemotherapy. This was followed by a reduced intensity stem cell transplant (Figure 1b). The primary outcome variable was 1-year relapse free survival. From February 2019, 14 patients with median age 28 years (IQR: 15.75-31.5) were enrolled in this trial. Six were females. Six had primary-refractory AML while 8 had relapsed-refractory AML. The cytoreductive chemotherapy was FLAG ± idarubicin (n=7), Mitoxantrone + Etoposide (n=6) and GCLAC (n=1). The median blast percentage on flowcytometry MRD testing prior to NK infusion was 15.9% (IQR: 9.1%-54.5%) (n=11). The median B score for the NK cell donors was 2 (IQR: 1-3). The median age of the NK cell donor was 43 years (IQR: 36-49.5). KIR ligand mismatch with the patient was noted in 2 donors. The median CD56-cell dose infused was 46.16 x 10 6/kg (IQR: 25.06-70.36) (Figure 1c). Pre-defined release criteria, including sterile cultures, and endotoxin negativity were met in all cases. There was no infusion related toxicity. The median blast percentage on flowcytometry MRD testing done following NK cell infusion was 11.9% (IQR: 4.9%-47.6%) (n=8). One patient withdrew consent after NK cell infusion and did not undergo transplant. For the remaining 13 patients, the stem cell donor was HLA matched (n=4), HLA 9/10 matched (n=1) or HLA haplomatched (n=8). The median CD34 cell dose infused was 10 x 10 6/kg (IQR:7.51-11.6). Five (38.5%) patients died of immediate post-transplant complications (sepsis (n=3) on days 1, 2 and 28, cerebral venous sinus thrombosis (n=1) on day 1 in a patient treated with hormonal contraceptives for menorrhagia, and veno-occlusive disease (n=1) on day 15 in a patient undergoing a second transplant) while 2 (15.4%) did not engraft (both subsequently died of infective complications following engraftment post-second transplant). Of the remaining 6 (46.2%) patients who engrafted and survived beyond 1 month of the transplant, the day 28 post-transplant MRD was negative for 5 patients while it was positive in 1 patient (0.13%). On follow up, 2 (15.4%) patients developed disease relapse (on days 54 and 218 respectively) and died. The remaining 4 (30.8%) patients are alive and relapse free at last follow up (mean follow up of surviving patients is 16 months). One patient received a CD34 cell boost on day 96 (cell dose - 8.55 x 10 6/kg) for poor graft function. For the entire cohort, the estimated 1-year event free survival is 28.8% ± 13.1%(Figure 1d). Whereas, acute GVHD was noted in 3 patients (50%; out of 6 evaluable patients) and chronic GVHD was noted in 3 patients (50%; out of the 6 evaluable patients). Thus, haploidentical NK cell therapy as an adjunct to transplant is safe and merits further evaluation in patients with AML. Figure 1 Figure 1. Disclosures Mathews: Christian Medical College: Patents & Royalties: US 2020/0345770 A1 - Pub.Date Nov.5, 2020; AML: Other: Co-Inventor.

Updated Outcomes during the Pandemic: Relationship between CD34and CD3 Cell Doses, One-Year Graft-Versus-Relapse-Free-Survival, Graft-Versus-Host Disease, and Overall Survival in Haploidentical Hematopoietic Stem Cell Transplantation Using Post Transplant Cytoxan: A Single Center Experience

Background: Haploidentical hematopoietic cell transplantation (haplo-HCT) has emerged as a popular alternative to traditional HLA-matched hematopoietic cell transplant. The major advancement of haplo-HCT has increased donor availability to many patients in need, therefore investigating the factors that may affect outcomes is necessary to improve overall survival and reduce transplant-related mortality. Data regarding optimal dose of CD34 cells, CD3 cells and CD34/CD3 ratio used during haplo-HCT to ensure favorable outcomes with post-transplant cyclophosphamide (PTCy) is lacking. Previously we have reported improved outcomes using CD34 cells limited to 7x10^6 cells/kg or less. In this report we are presenting an updated analysis including outcomes related to CD34/CD3 ratio used during haplo-HCT. Methods: We retrospectively analyzed adult patients at USC Norris Cancer Hospital who received haplo-HCT from 2014 to 2020. The primary endpoint assessed was 1-year GVHD-free/relapse-free survival (GRFS). Secondary endpoints included overall survival (OS), 1-year transplant related mortality (TRM) and incidence of acute and chronic GVHD. Kaplan Meier survival curves and log-rank tests were used to evaluate 1-year GRFS and OS with CD34 cell dose ≥ 7x10^6 cells/kg and &lt; 7x10^6 cells/kg; CD3 cell dose &lt; 2.5x10^8 cells/kg and ≥ 2.5x10^8 cells/kg; and CD34/CD3 ratio &lt;2, 2-3, &gt;3-5 and &gt;5. Results: 103 adult haplo-HCT recipients were reviewed with 55.3% male and 44.7% female. The age range was 21-71 years old (median = 51), and a majority of patients were Hispanic (62%). The most common underlying hematologic disorders included AML (40.8%), ALL (34.1%), and MDS/MPN (13.6 %). 41.7% patients with leukemia were in CR1, 55% were CR2/CR3 and 78% were MRD negative by flow cytometry prior to transplant. 73% received myeloablative conditioning and 83.5% received peripheral blood stem cells. Median CD34 dose, CD3 dose and CD34/CD3 ratio were 6.1 x10^6cells/kg, 2.27 x10^8 cells/kg and 3.2 respectively. Median time to recovery of neutrophil, platelets, and lymphocyte was 17, 24, and 124 days respectively. Incidence of 1-year GRFS was 43.7%. 1-year TRM was 13.6% and rate of aGVHD and cGVHD was 42.7% (n = 44) and 35.9% (n = 37) respectively. There was no difference in 1-year GRFS and OS when multiple dose levels of CD34 cells and CD3 cells were compared. Although when dichotomized, a CD34 cell dose of ≥ 7x10^6 cells/kg compared to &lt; 7x10^6 cells/kg showed higher dose had significant improvement in 1- year overall survival (p-value=0.01) but no statistical difference in 1 year GRFS (p value=0.24). CD3 cell dose &lt; 2.5 x 10^8 cells/kg trended towards worse 1-Year GRFS compared to ≥ 2.5 x 10^8 cells/kg (43.9% compared to 58.6%; p-value 0.075), similarly CD3 dose of &lt; 2.5x10^8 cells/kg showed a lower 1-year OS compared to &gt; 2.5x10^8 cells/kg but was not statistically significant (78.7% compared to 89.4%; p-value 0.092). 1-year GRFS and OS did not show statistical difference at CD34:CD3 ratios of &lt; 2, 2-3, 3-5 and &gt; 5, although on cox regression analysis the HR for 1-year GRFS was 2.92 (95% CI 1.21-7.05) in patients with CD34:CD3 &gt; 5 as compared to reference of &lt; 2 (p-value 0.017). Multivariate regression also showed CD3 cell doses of 2-3x10^8 cells/kg and &gt;3 x 10^8 cells/kg were associated with worse 1-year GRFS with HR of 0.43 and 0.40 and p-value of 0.041 and 0.048 respectively when compared to reference of &lt; 1x10^8 CD3 cells/kg. Discussion: Our results demonstrate 43.7% survived 1 year compared with reports of 24-35%. The OS was significantly better in the CD34 dose &gt; 7x10^6 cells/kg. CD3 &lt;2.5x10^8 cells/kg showed a trend towards lower 1-year GRFS and OS, which was not statistically significant. On multivariate analysis CD3 dose of &gt; 2x10^8 cells/kg was associated with inferior 1-year GRFS. Conversely, higher CD34/CD3 ratio &gt;5 was associated with increased 1-year GRFS. Thus, our findings indicate that along with improvement in OS by using &gt;7x10^6 CD34 cells/kg, lower CD 3 cell dose &lt;2x10^8 cells/kg and higher CD34/CD3 ratio &gt; 5 can improve 1-year GRFS in patients receiving haplo-HCT. Figure 1 Figure 1. Disclosures Chaudhary: Angeles Therapeutics: Current holder of individual stocks in a privately-held company, Current holder of stock options in a privately-held company, Membership on an entity's Board of Directors or advisory committees, Other: Founder, Patents & Royalties: Cell therapy ; Athelas: Consultancy, Current holder of stock options in a privately-held company; Oncotartis: Consultancy; Pancella: Consultancy; Moderna: Current equity holder in publicly-traded company; Celldex: Current equity holder in publicly-traded company; TCR2: Current equity holder in publicly-traded company; Allogene: Current equity holder in publicly-traded company. Yaghmour: Jazz: Consultancy, Honoraria; Astellas: Consultancy; Takeda: Consultancy; Incyte: Consultancy.

391 Biomarker correlates of response in patients with advanced myxoid/round cell liposarcoma (MRCLS) treated with NY-ESO-1 TCR T cells (Letetresgene autoleucel)

BackgroundThis is an open label pilot study (NCT02992743) on letetresgene autoleucel (lete-cel; GSK3377794), an NY-ESO-1-specific autologous CD4+ and CD8+ T cells expressing a high affinity T-cell receptor which recognizes the NY-ESO-1 antigen epitope in complex with specific HLA- alleles A*02, which exhibited anti-tumor activity and manageable safety profiles in patients with advanced MRCLS based on interim analysis (IA) data.1 Lymphodepletion has been shown to enhance the expansion, persistence, and homing of therapeutically infused T-cells, thereby potentiating therapeutic efficacy against malignant diseases.2 Initial T-cell kinetics data from this study demonstrated that lymphodepletion regimen (LDR)-B robustly depleted lymphocytes at infusion and was trended with higher peak cell expansion (Cmax) vs. LDR-A. The peak expansion was significantly associated with weight-normalized transduced cell dose and trended with response.1 Here, we will be discussing additional cell kinetics data and other exploratory biomarker correlates of response.MethodsPatients with advanced MRCLS were enrolled to 2 cohorts and received either planned A (N=10) or B (N=10) LDRs prior to lete-cel infusion (table 1). Response was assessed per RECIST v1.1. Transduced cell kinetics were measured by quantitative PCR of transgene vector copies in DNA from peripheral blood mononuclear cells (PBMCs). Serum cytokines (Meso Scale Discovery immunoassay) as pharmacodynamic (PD) markers of response and their association with T cell kinetics will be discussed. Phenotypic characterization of the cell product (pre- and post- infusion) via Flow cytometry using Cytek Aurora (23 color panel), to help understand correlation of response with engineered cell product attributes, will be presented. Potential biomarker correlates of clinical response were tested using generalized linear models.ResultsFive out of 6 responders with available lab data exhibited robust lymphocyte depletion at infusion (0–25 cell/µL) and high Cmax (>50,000 vector copies/µg gDNA) with LDR. Only 6/14 non-responders exhibited low lymphocytes counts at infusion and high Cmax. LDR-B also induced strong depletion of monocytes at infusion (p=0.03) vs. LDR-A, but depletion of monocytes did not show association with response. Higher Cmax was correlated with exposure (AUC0–28d) (Adj. R2=0.606). AUC0–28d was a better predictor of response in patients receiving LDR-B (p=0.0182), with AUC0–28d trending towards predicting response in the LDR-A cohort. AUC0–28d was associated with tumor volume reduction (p=0.0569).Abstract 391 Table 1ConclusionsExposure–response analysis of this study reveals that efficacy appears to be driven by weight-normalized transduced cell dose and LDR via AUC0–28d. Higher AUC0–28d was correlated with Cmax and maximum tumor volume reduction.AcknowledgementsThis study (208469; NCT02992743) was funded by GlaxoSmithKline.Trial RegistrationNCT02992743ReferencesD’Angelo SP, et al. J Clin Oncol 2021;39:15_suppl:11521.Bechman, Maher. Expert Opin Biol Ther 2021;21(5):627–637.Ethics ApprovalThis study was approved by institutional review boards (IRB) at the six participating sites.