Development of human anti-murine T-cell receptor antibodies in both responding and nonresponding patients enrolled in TCR gene therapy trials

Abstract

Purpose: Immune responses to gene-modified cells are a concern in the field of human gene therapy, as they may impede effective treatment. We conducted 2 clinical trials in which cancer patients were treated with lymphocytes genetically engineered to express murine T-cell receptors (mTCR) specific for tumor-associated antigens p53 and gp100.

Experimental design: Twenty-six patients treated with autologous lymphocytes expressing mTCR had blood and serum samples available for analysis. Patient sera were assayed for the development of a humoral immune response. Adoptive cell transfer characteristics were analyzed to identify correlates to immune response.

Results: Six of 26 (23%) patients' posttreatment sera exhibited specific binding of human anti-mTCR antibodies to lymphocytes transduced with the mTCR. Antibody development was found in both responding and nonresponding patients. The posttreatment sera of 3 of these 6 patients mediated a 60% to 99% inhibition of mTCR activity as measured by a reduction in antigen-specific interferon-γ release. Detailed analysis of posttreatment serum revealed that antibody binding was β-chain specific in 1 patient whereas it was α-chain specific in another.

Conclusions: A subset of patients treated with mTCR-engineered T cells developed antibodies directed to the mTCR variable regions and not to the constant region domains common to all mTCR. Overall, the development of a host immune response was not associated with the level of transduced cell persistence or response to therapy. In summary, patients treated with mTCR can develop an immune response to gene-modified cells in a minority of cases, but this may not affect clinical outcome.

Figure 1

Results of antibody binding and serum inhibition assays. A) Five patients treated with gp100 TCR (3, 4, 14, 16, 19) and one patient treated with p53 TCR (26) show post-treatment serum binding of TCR-transduced lymphocytes with which they were treated as detected by anti-human IgG antibody staining. Shaded histogram represents lymphocyte incubation with pre-treatment serum; bold lines represent post-treatment serum incubation. B) TCR-transduced PBL co-cultured with cognate antigen-expressing tumor cell lines in the presence of pre-treatment and post-treatment serum; supernatant interferon-gamma (IFNγ) levels shown (pg/ml). C) Post-treatment serum inhibition results of 16 gp100 TCR patients and 10 p53 TCR patients as a percentage of co-culture IFNγ release relative to pre-treatment serum samples. D) Reversal of post-treatment serum inhibition (patient 14) following pre-incubation in the presence or absence of protein G. (+), with protein G; (−) without protein G.

Figure 2

Serum inhibition and antibody binding of transduced lymphocytes is specific. A) Antibody binding of TCR-transduced lymphocytes as determined by anti-human IgG antibody staining of lymphocytes following incubation with pre-treatment (shaded) and post-treatment (line) serum. p53 and CEA TCRs share the same β chain variable region gene (Vβ 26.1). B) Post-treatment serum inhibition of mTCR (gp100, p53, MAGE-A3 and CEA) and one human TCR (NY-ESO-1). Patients 4, 14 (gp100 TCR) and 26 (p53 TCR) had 6 month post-treatment serum incubated with lymphocytes expressing the various TCRs. Results are shown as a percentage of IFNγ release compared to incubation with pre-treatment serum. C) Mismatch of p53 and gp100 TCR alpha and beta chains by independent transduction followed by serum incubation and staining with anti-human IgG antibody.