Chimeric antigen receptors (CARs) combine the antigen recognition domain of an antibody with intracellular signaling domains into a single chimeric protein.  CD19 is an ideal target for CARs since expression is restricted to normal and malignant B cells.  Inclusion of the CD137 (4-1BB) signaling domain results in potent antitumor activity and in-vivo persistence of anti-CD19 CAR-modified T cells in mice.  Lentiviral transduction into T cells facilitates strong surface expression of the CAR.  We reported anti-tumor activity of CAR-modified autologous T cells targeted to CD19 (CTL019 cells) in 3 patients (pts) with CLL with relatively short follow up (Porter, et al NEJM 2011; Kalos et al Sci Trans Med 2011).  We now report on outcomes and longer follow up from our pilot study treating 14 pts with relapsed, refractory CLL. 

 METHODS

Autologous T cells collected by leukapheresis were transduced with a lentivirus encoding anti-CD19 scFv linked to 4-1BB and CD3-ζ signaling domains.  Gene-modified T cells were expanded and activated ex-vivo by exposure to anti-CD3/CD28 beads.  Pts had to have relapsed or persistent disease after at least 2 previous treatments (1 prior therapy for patients with p53 mutation) and progressed at least within 2 years of their last therapy.  All pts received lymphodepleting chemotherapy ending 3-5 days before T cell infusion.  The target dose of cells was 5 x 10⁹ mononuclear cells with an expected transfection efficiency of 10-40% (total CTL019 dose 5x10⁸ – 2 x 10⁹ total cells).  Cell infusions were planned over 3 days (10% on day 1, 30% of day 2, and 60% on day 3) but were held for fevers or other toxicity. 

RESULTS

14 patients were treated on this pilot study including 12 men and 2 women with a median age of 67 (51-78).  Pts had received a median of 4 prior therapies (1-10) and 6 pts had a mutation of p53.  All pts had active disease at the time of CTL019 cell infusion. Lymphodepleting chemotherapy was Fludarabine/cyclophosphamide (3), pentostatin/cyclophosphamide (5), or bendamustine (6).  A median of 7.5 x 10⁸ total cells (range 1.7-50), corresponding to 1.4 x 10⁸(range 0.14-5.9) genetically modified cells were infused over day 0, 1 and 2. 

There were no infusional toxicities >grade 2 though 6 pts developed fevers within 24 hrs of infusion #1 (3) or #2 (3) and did not receive additional CTL019 cells.  Median follow-up as of July 15, 2013 was 9.4 mo (4-35) for all pts and 16 mo (5-35) for the 8 responding pts.  3 patients (21%) achieved a CR (follow-up 11, 34, and 35 mo), 5 (36%) achieved a PR (med follow up 11 mo, range 5-27 mo) and 6 (43%) had no response, for an overall major response rate of 57%.  2 of 5 pts with a PR progressed 4 mo after infusion with CD19+ CLL, and no patient with a CR has relapsed. 

Comparing responders to non-responders, there has been no association between response and patient age (66 vs 67 yrs), number of prior therapies (median 4 each), cell dose (7.5 vs 11.5 x 10⁸MNC), or p53 mutation (3/8 vs 3/6, p>0.9), implying that within the dose ranges studied, there is no obvious dose:response relationship.

All responding pts developed a delayed cytokine release syndrome (CRS), concurrent with peak T cell expansion, and was manifested by fever, and variable degrees of nausea, anorexia, myalgias, and transient hypotension and hypoxia.  Detailed cytokine analysis showed marked increases from baseline values of IL6, IFN-γ, and IL2R, while no significant elevation in systemic levels of TNFα or IL2 were observed. The CRS required intervention in 5 patients. Treatment was initiated for hemodynamic or respiratory instability and was rapidly reversed in all cases with corticosteroids in 1 pt and the IL6-receptor antagonist tocilizumab (4 pts); 3 of these 4 pts also received 1 or 2 doses of corticosteroids.  Persistence of CTL019 cells has been detected by flow cytometry in all 6 pts with ongoing responses 5-35 months after infusion, and all patients have sustained B cell aplasia without any unusual infectious complications. 

 CONCLUSIONS:

CTL019 cells are autologous T cells genetically engineered to express an anti-CD19 scFv coupled to 4-1BB/CD3-ζ signaling domains.  These cells can undergo robust in-vivo expansion and can persist for at least 3 yrs.  CTL019 therapy is associated with a significant CRS that responds rapidly to anti-cytokine treatment.  CTL019 cells can induce potent and sustained responses (8/14) for patients with advanced, relapsed and refractory CLL regardless of p53 mutation status.

Patients (pts) with relapsed, and/or refractory (R/R) CLL have a poor prognosis with few effective treatment options.  We have shown that infusion of autologous T cells genetically modified to express a chimeric antigen receptor (CAR) consisting of an external anti-CD19 domain, with the CD3ζ and 4-1BB signaling domains (CTL019 cells), can mediate potent anti-tumor effects in pts with advanced, relapsed refractory CLL.  In our initial pilot study, doses of 1.7-50, x 10⁸ mononuclear cells, corresponding to 0.14-5.9 x 10⁸genetically modified cells, were given as a split dose infusion on days 0, 1 and 2 to 14 pts with R/R CLL and overall response rate (PR plus CR) was 57%.  The majority of responses were sustained, and associated with marked expansion and long-term persistence of transduced cells.  Notably, there was no obvious dose:reponse or dose:toxicity effect noted over a wide range of cell doses.  To better define an optimal CTL019 cell dose, we are performing a randomized phase II study of 2 doses of CTL019 cells in pts with R/R CLL. 

Methods:

Pts with R/R CLL are randomly assigned to receive either 5x10⁸ vs. 5x10⁷transduced CTL019 cells, with the rationale that both doses induced CRs in pts on our initial pilot trial.  In the initial stage, 12 evaluable pts will be treated in each arm and in stage 2, an additional 8 pts will be treated with the selected dose level.  Pts have to have relapsed or persistent disease after at least 2 previous treatments and progress within 2 years of their last therapy.  All pts receive lymphodepleting chemotherapy ending 3-5 days before T cell infusion.  Cell infusions are given as a single dose.

Results:

As of 7/15/2013, 27 pts have been enrolled; T cells did not adequately expand in 3, 1 patient was not eligible after screening, and 10 pts have been treated including 7 men and 3 women with a median age of 63 yrs (range 59-76).  5 pts had a mutation of p53.  All pts had active disease at the time of CTL019 cell infusion.  Lymphodepleting chemotherapy was Fludarabine/cyclophosphamide (8), pentostatin/cyclophosphamide (1), or bendamustine (1).  4 pts have been randomized to the higher dose level (5 x 10⁸ CTL019 cells) and 6 pts have been randomized to the lower dose level (5 x 10⁷CTL019 cells).  There were no significant infusional toxicities.  Median follow-up as of July 15, 2013 was 3 mo (1.3-5) for all pts and 3.3 mo (1.3-4) for responding pts.  2 pts have achieved a CR and 2 pts achieved PR, both with clearance of CLL from the blood and marrow and >50 reduction in adenopathy, for an overall response rate of 40%.  In other recipients of CTL019 cells, we have observed ongoing improvement in adenopathy over time implying there can be a continued anti-tumor response.  No responding patient has progressed.  Seven of 10 pts experienced a delayed cytokine release syndrome (CRS) manifested by symptoms that included high fevers, nausea, myalgias and in some cases, capillary leak, hypoxia, and hypotension, typically correlated with peak CTL019 cell expansion. 

We have noted that the CRS accompanying CTL019 therapy has been associated with marked increases of serum IL6 and can be rapidly reversed with the IL6-receptor antagonist tocilizumab.  The CRS required intervention in 2 pts, one who responded and one who did not respond to CTL019. Treatment was initiated for hemodynamic or respiratory instability and was effective in reversing signs and symptoms of CRS in both pts. 

A preliminary analysis through July 15, 2013 does not yet suggest a dose:response or dose:toxicity relationship.  2 of 4 recipients of the higher dose CTL019 responded, and 2 of 6 recipients at the lower dose level responded.  The 7 pts who experienced a CRS included all 4 responding pts and 3 pts who did not respond.  The CRS occurred in 3/4 recipients of higher dose CTL019 cells and 4/6 of recipients of lower dose CTL019 cells.  CTL019 expansion in-vivo and persistence over the follow up period was noted in all responding pts. 

Conclusions: 

In this ongoing dose optimization study of CTL019 cells, 4 of the first 10 pts treated have responded within 3 months.  With short follow-up, as yet there is no suggestion that there is a dose:response or dose:toxicity relationship at the dose ranges being studied.  These cells can undergo robust in-vivo expansion and from other studies (ASH 2013) can persist for at least 3 yrs.  This trial confirms that CTL019 cells can induce potent responses for pts with advanced, relapsed and refractory CLL.

CARs combine a single chain variable fragment (scFv) of an antibody with intracellular signaling domains into a single chimeric protein. We previously reported on CTL019 cells expressing a CAR with intracellular activation plus costimulatory domains. Infusion of these cells results in 100 to 100,000x in vivo proliferation, durable anti-tumor activity, and prolonged persistence in pts with B cell tumors, including 1 sustained CR in a patient with ALL (Grupp, et al. NEJM 2013).  We now report on outcomes and longer follow up from our pilot studies treating 20 pts (16 children and 4 adults) with relapsed, refractory ALL. 

METHODS

T cells were lentivirally transduced with a CAR composed of anti-CD19 scFv/4-1BB/CD3ζ, activated/expanded ex-vivo with anti-CD3/anti-CD28 beads, and then infused into pts with relapsed or refractory CD19+ ALL.  17/20 pts received lymphodepleting chemotherapy the week prior to CTL019 infusion. The targeted T cell dose range was 10⁷ to 10⁸ cells/kg with a transduction efficiency (TE) of 11-45%. On the adult protocol, the target dose was 5 x 10⁹ total cells split over 3 days with a TE of 6-31%. 11 pts had relapsed ALL after a prior allogeneic SCT.  T cells were collected from the pt, regardless of prior SCT status, and not from allo donors. All pts s/p allo SCT had to be 6 mos s/p SCT with no GVHD or GVHD treatment.

RESULTS

16 children median age 9.5 y (5-22y) and 4 adults median age 50y (26-60y) with CD19+ ALL were treated. One child had T cell ALL aberrantly expressing CD19. 14/16 pediatric pts had active disease or +MRD after chemotherapy on the day prior to CTL019 cell infusion, while 2 were MRD(-). 3 of 4 adults had active disease prior to lymphodepleting chemotherapy, while 1 was in morphologic CR. Lymphodepleting chemotherapy varied with most receiving a Cytoxan-containing regimen the week prior to CTL019. A median of 3.7x10⁶ CTL019 cells/kg (0.7-18x10⁶/kg) were infused over 1-3 days.

There were no infusional toxicities >grade 2, although 5 pts developed fevers within 24 hrs of infusion and did not receive planned subsequent infusions of CTL019 cells.  14 patients (82%) achieved a CR, including the patient with CD19+ T ALL, 3 did not respond, and 3 are pending evaluation. 11/17 evaluable pts have ongoing BM CR with median follow up 2.6 mo (1.2-15 mo).  Three patients with a CR at 1 month have subsequently relapsed, 1 with CD19(-) disease. Median follow-up as of August 1, 2013 was 2.6 mo (1-15 mo) for all pts.

All responding pts developed some degree of delayed cytokine release syndrome (CRS), concurrent with peak T cell expansion, manifested by fever, with variable degrees of myalgias, nausea, anorexia. Some experienced transient hypotension and hypoxia.  Detailed cytokine analysis showed marked increases from baseline values of IL6 and IFNγ (both up to 1000x), and IL2R, with mild or no significant elevation in systemic levels of TNFα or IL2. Treatment  for CRS was required for hemodynamic or respiratory instability in 7/20 patients and was rapidly reversed in all cases with the IL6-receptor antagonist tocilizumab (7 pts), together with corticosteroids in 4 pts. Although T cells collected from the 11 pts who had relapsed after allo SCT were generally 100% of donor origin, no GVHD has been seen. Persistence of CTL019 cells detected by flow cytometry and/or QPCR in pts with ongoing responses continued for 1-15 months after infusion, resulting in complete B cell aplasia during the period of CTL019 persistence.  Pts have been treated with IVIg without any unusual infectious complications. One child who entered a CR subsequently developed MDS with a new trisomy 8 in ALL remission and has gone to SCT, and 1 child developed a single leukemia cutis lesion at 6 mo, still BM MRD(-).

CONCLUSIONS:

CTL019 cells are T cells genetically engineered to express an anti-CD19 scFv coupled to CD3ζ signaling and 4-1BB costimulatory domains.  These cells can undergo robust in-vivo expansion and can persist for 15 mo or longer in pts with relapsed ALL.  CTL019 therapy is associated with a significant CRS that responds rapidly to IL-6-targeted anti-cytokine treatment.  This approach has promise as a salvage therapy for patients who relapse after allo-SCT, and collection of tolerized cells from the recipient appears to have a low risk of GVHD. CTL019 cells can induce potent and durable responses for patients with relapsed/refractory ALL. Multicenter trials are being developed to test this therapy for ALL in the phase 2 setting.

Recent advances in T cell engineering have enabled clinical trials to evaluate the potential for adoptive transfer of T cells to target malignancy.  A single treatment with engineered gene-modified T cells has the potential to generate potent and long-lasting anti-tumor immunity.  We have reported initial clinical data on the use of T cells engineered to express a Chimeric Antigen Receptor (CAR) that targets CD19 (CTL019) in patients with advanced treatment-refractory CLL and pediatric ALL (Porter, et al NEJM 2011; Kalos et al. Sci Trans Med 2011, Grupp et al. NEJM 2013).  The initial cohort of patients is now disease free between 1 and 3 years post-infusion.  In this report we present data on the functional persistence, trafficking, and bioactivity of CTL019 cells from the initial cohort of CLL patients, a larger and more recently treated cohort of CLL patients, a cohort of pediatric ALL patients, as well as an initial cohort of  adult ALL patients.  

METHODS

CAR engineering of patient T cells was accomplished by lentivirus transduction followed by in-vitro expansion using anti-CD3 and anti-CD28 antibody-coated beads.  Persistence of gene-modified T cells was assessed by quantitative PCR.  CAR19 surface expression was detected by flow.  B cell aplasia was evaluated by multiparametric flow cytometry.   Multiplex cytokine analyses were performed using Luminex^TMtechnology.

RESULTS

Detailed clinical outcomes for each patient cohort will be reported separately at this meeting (Porter, D.L.et al, Grupp S. et al).  In summary, as of July 15, 2013:  For CLL:  A total of 24 adult patients with advanced relapsed and/or treatment-refractory CLL have been treated under two separate protocols. To date 5 patients have achieved ongoing CR, 7 patients PR and 12 patients were NR at the primary endpoint (within 3 months post infusion). For ALL (pediatric): A total of 14 pediatric patients with treatment refractory ALL have been treated and were evaluable under a single protocol.  To date 8 patients have ongoing CR, 4 patients have relapsed including 2 with CD19-negative disease and 2 patients were NR.  For ALL (Adult):  A total of 3 adult patients with advanced relapsed and/or treatment-refractory ALL have been treated under a single protocol.  All 3 patients have ongoing CR; one patient went into an allogeneic transplant while in CTL019-induced CR.

In all patients with CR, robust in vivo expansion of CTL019 cells was observed, as assessed by both molecular and flow cytometric analysis, followed by contraction and in all but one patient ongoing stable persistence of engineered cells, elimination of tumor B cells and ongoing B cell aplasia in blood and marrow at all evaluated time points (minimum 3 months, maximum ongoing at 35 months).  In CR patients, peak marking exceeded 5% of total CD3+.  Patients with PR demonstrated less robust in-vivo expansion accompanied by transient B cell elimination, while NR patients demonstrated minimal in-vivo expansion.  

Clinical MRD analysis in patients was supplemented with Illumina-HiSeq/MiSeq-deep sequencing based molecular MRD analysis. These analyses demonstrate that CRs were accompanied with deep and sustained molecular remissions as assessed by the absence of the tumor specific clonotype defined in the enrollment samples.

All patients in CR experienced on-target cytokine release syndrome (CRS) and macrophage activation syndrome (MAS).  Multiplex-cytokine analysis of serum samples from patients in CR demonstrated a broad pro-inflammatory signature with significant elevation in a subset of soluble immune modulators including IFN-g, IL-6, and IL2ra.  In contrast, patients with NR did not have elevated serum cytokines. Elevation of cytokines coincided with expansion of CTL019 cells, elimination of B cells, and toxicity suggesting the potential for a cytokine-based diagnostic signature to monitor CTL019 treatment efficacy.

CONCLUSIONS

Adoptive transfer of CTL019 cells engineered to express CD137 and TCR-zeta signaling domains can result in in-vivo expansion, homing to marrow, and long-term functional persistence of engineered cells, accompanied by ongoing complete clinical responses and long-term B cell aplasia in a substantial fraction of patients with advanced, refractory and high risk CLL and relapsed refractory pre- B cell ALL.  These results highlight the potential of CTL019 therapy to effectively target CD19-positive malignancy.