Table 6.
Comparison of TRBAs and CAR-T cell platforms as therapeutics.
| Properties | Therapeutic Approach | ||
|---|---|---|---|
| T cell Redirection with TRBAs | Autologous CAR-T or NK Cells | Allogeneic CAR-T or NK Cells (Projected) | |
| Currently approved and marketed (as of 20 June 2019) | 1; Blincyto® (anti-CD19 × CD3 BiTE®) |
2; Kymriah® and Yescarta®, both CD19-targeting autologous CAR-Ts | None |
| Current indications covered | R/R B-ALL | DLBCL, R/R NHL, B-ALL | None |
| Structure | Bispecific antibodies that bind both a tumor antigen and CD3ε on T cells | T cells engineered with synthetic gene construct encoding scFv fused to linker and activation domains | T cells engineered with synthetic gene construct encoding scFv fused to linker and activation domains |
| Source and homogeneity of T cell component | Endogenous T cells; No homogeneity (i.e., all CD3+ T cells may be engaged) | Expanded and activated endogenous T cells; homogeneity depends on process used | Could be homogeneous CD8+ T-cells, depending on cell type and approach |
| Antibody | Short half-life vs long half-life formats | Currently, mostly scFvs; possible unfolding, aggregation, tonic signaling; need for better binding constructs | Currently mostly scFvs—possible unfolding, aggregation, tonic signaling; need for better binding constructs |
| T-cell signaling domain(s) | CD3ζ | CD3ζ + 4-1BB (or OX40) and/or CD28 | CD3ζ + 4-1BB (or OX40) and/or CD28 |
| PD-1 inhibition of CD28 activity | Likely significant issue; may need to co-dose with PD-1 inhibitor | Use of 4-1-BB signaling domain should alleviate | Use of 4-1-BB signaling domain should alleviate |
| Drug-like properties | “Off-the-shelf” drug | Must be engineered from patient’s T cells (2–4 week process) | Depends on cell type and construct |
| Dosing | Multiple dosing; short half-life formats may require continuous dosing via pumps | Single dose | Single dose; multiple dose potentially available if engineered to eliminate HLA |
| Route of administration | IV; possible subcutaneous for future candidates | IV only | IV only |
| Long-term persistence and memory | Short half-life – only as long as continuously infused; long half-life – typically measured in weeks | Yes, but variable; longer persistence correlated with activity | Unknown but likely to be similar to autologous T-cells |
| Immune synapse | Normal and concentric; normal detachment | Abnormal and multifocal; fast detachment | Expected to be similar to autologous CAR-T cells |
| T cell signals at synapse | Signals 1, 3 | Signals 1, 2 (sometimes), 3 | Expected to be similar to autologous CAR-T cells |
| Killing mechanisms | Perforin and granzyme; Secondary: cytokine modulation of TME [123] | Perforin and granzyme; Fas/FasL axis; Secondary: cytokine modulation of TME [123] | Expected to be similar to autologous CAR-T cells |
| Serial killing | Yes, similar to CTLs | Yes, faster than TRBAs an CTLs | Expected to be the same as autologous T cells |
| None; related to dosing and half-life | Yes, in responders | Unknown but expected | |
| Bystander killing of antigen-negative cells | Demonstrated, as long as antigen-negative cells were in direct contact with antigen-positive cells [381] | Demonstrated, as long as antigen-negative cells were in direct contact with antigen-positive cells [380] | Unknown but expected based on CAR-T results |
| Toxicity | CRS, neurotoxicity | Higher CRS and neurotoxicity than TRBAs | Unknown but expected |
| Ability to attack solid tumors | To be determined; early data are mixed but not encouraging | To be determined; early data are mixed but not encouraging | Potential based on TIL correlation data |
| Trafficking | Passive | Active but limited; can be engineered to match tumor needs | Active; possible to engineered to match tumor needs |
| Trafficking into CNS | Not demonstrated; Unlikely if BBB is intact [395] | Demonstrated trafficking into CNS [399] | Unknown but expected based on CAR-T results |
| Need for lymphodepletion prior to treatment | No | Yes | Yes |
| Technical risk | Moderate; many platforms are working well | High but may be manageable | Currently very high |
| Need for “kill switch” or turn-off methodology | No but nice to have, especially for long half-life formats | Moderate; nice to have | Very high; must have for safety |
| Accessibility | High–off-the-shelf biologic drug | Only available at specific medical centers thus far; 2–4 week process time before therapy | Projected to eventually have availability similar to biologic drugs |
| Cost of goods | Relatively low; Antibody-like or slightly higher depending on type of TRBA platform | Very high (more than a $75,000 process) | Projected to be low to medium once cell manufacturing process is established |
| Cost to patient/payers | Medium ($89,000/course; $178,000 for predicted two course therapy) * | Very high ($373,000 for treatment of DLBCL; $475,000 for Kymriah® treatment of B-cell ALL) ** | Projected as medium to high, depending on cell type and construct |
Abbreviations: BBB, blood-brain barrier; B-ALL, B-cell acute lymphoblastic leukemia; BiTE®, bispecific T-cell engager; CAR-T cell, chimeric antigen receptor-T cell; CD, cluster of differentiation; CNS, central nervous system; CRS, cytokine release syndrome; DLBCL, diffuse large B-cell lymphoma; IV, intravenous; NHL, non-Hodgkin lymphoma; NK, natural killer (cells); R/R, relapsed or refractory; PD-1, programmed cell death protein 1; scFv, single chain, Fragment variable (antibodies); TIL, tumor-infiltrating lymphocytes; TRBA, T-cell redirecting bispecific antibody. * Quote is for the cost of two-course treatment with Blincyto® [400]. ** Quotes for cost of CAR-T therapies [401].