Abstract
HD may be a suitable target for immunotherapy, and in patients with EBV-associated HD, adoptive transfer of EBV-CTL has produced disease responses. An alternative target is the CD30 molecule, which is present on the malignant cells of almost all patients with HD. CD30 is a member of the TNF superfamily and monoclonal antibodies directed to this antigen are currently under investigation in patients with relapsed HD. An alternative way to target CD30 is by the construction of T cells expressing cTcR specific for the antigen. T lymphocytes engineered to express this cTcR can specifically kill CD30+ HD cell lines {Cancer Res,1998;58:1116}. However, these chimeric molecules connect the antigen-recognition properties of CD30 antibodies with the endodomain of CD3ζ, which is insufficient to fully activate resting T cells to proliferate and release cytokines. As a consequence chimeric T cells that express these endodomains divide infrequently, lose activity and have performed poorly in-vivo. Full T cell activation requires receptor engagement to be accompanied by a sequence of co-stimulatory stimuli. We have shown that EBV-CTL can fulfill this need, since the co-stimulatory signals delivered by EBV-infected B cells after native receptor engagement ensure full functionality when the CTL subsequently bind to tumor cells through their cTcR. We first evaluated whether EBV-CTL can be redirected to kill CD30+ HD cell lines and whether they retain their specificity and antigen repertoire. EBV-CTLs were prepared from 8 EBV+ healthy donors using weekly stimulation with irradiated autologous EBV-transformed lymphoblastoid cell lines (LCL) in the presence of IL-2 (50U/mL). CTL were transduced after the 3rd stimulation and further expanded with 3–4 weekly LCL/IL-2 stimulations. The expansion rate of the transduced CTL was similar to that of control EBV-CTL. Transduced CTL retained killing of their autologous LCL targets through their native receptor (64.4±16% at 20:1 E:T ratio), and became able to lyse CD30+ malignant lymphoma targets through their cTcR (e.g. HDLM-2=45.4±16% and Karpas-299=42.5±17%). Killing of CD30+ tumor cells was significantly inhibited by preincubation with an anti-CD30 blocking antibody (16.5±12%). Of potential concern, however, is that CD30 is expressed by activated normal T lymphocytes: expression was undetectable on resting T cells, but increased to 3–32% on day 4–7 after stimulation with LCL. Fortunately, expression dwindles to 3–6% by two weeks as an EBV-specific line emerges, suggesting that CD30 is expressed only in the early phases of T cell activation. As anticipated from these data, therefore, expression of a CD30 cTcR did not impair the antigenic repertoire of the EBV-CTL, which retained the same pattern of immunodominant MHC class I epitopes (detected by tetramer) as control cells. We also performed co-culture experiments to evaluate whether infusion of CTL-CD30 cTcR could cross-compromise the primary reactivation of other virus-specific CTL. Autologous EBV-CTLs engineered to express the CD30-cTcR were added to cultures of PBMC stimulated to reactivate cytomegalovirus- or adenovirus-specific CTL. In 4/4 donors, the percentage of CMV pp65+ T cells did not change, while generation of adenovirus-specific T cells (Hexon-tetramer+) was significantly reduced in only 1/3 donor. These data support the feasibility of using EBV-CTL bearing a cTcR for CD30 to treat both EBV+ and EBV− HD.
A phosphatidylinositol-linkage-deficient T-cell mutant contains insulin-sensitive glycosyl-phosphatidylinositol