Successful Treatment of Refractory CMV Chorioretinitis and Meningoencephalitis with Adoptive Transfer of Third Party CMVpp65 Specific T-Cell Lines

Adoptive immunotherapy with transplant donor-derived virus specific T cells is an effective strategy for the treatment of CMV viremia and disease arising after an allogeneic hematopoietic stem cell (HSCT). At our center this approach has become a standard part of the armamentarium of CMV directed therapy. However, donor-derived CMVpp65-specific cytotoxic T cells (CMV-CTLs) are not available for patients transplanted from a seronegative or cord blood donor. In addition, in the HLA disparate transplant setting the CMV-CTL line derived from non-identical donors may be restricted by non-shared HLA alleles. For these patients as we have previously reported, treatment with in vitro expanded CMV-CTLs derived from an HLA partially matched third party donors is an option.

One particularly difficult clinical situation is the treatment of patients who develop CMV disease involving the sanctuary sites of the central nervous system and retina. We have treated 12 patients with primary donor derived (n=1) or third party (n=11) CMV CTLs for retinitis (7) or meningitis/encephalitis (4) or both (1). Recipients of hematopoietic stem cell transplant (HSCT) had undergone T cell depleted (5) conventional (1) or cord blood (4) transplantation. One patient was treated after solid organ transplantation and one for HIV related CMV retinitis.

Third party CMV CTLS were selected from a bank of 132 lines generated under GMP conditions from normal HSCT donors specifically consented for use of their T cells in patients other than their designated transplant recipient. Third party CMV-CTLs were selected on the basis of HLA matching at a minimum of 2/8 recipient alleles and HLA restriction of the T cells by one or more HLA alleles present in the patient. A total of 10 distinct third party CMV-CTL lines and one (1) donor derived line were used. Patients received infusions of CMV-CTLs after failing a median of 146 (43-419) days of prior therapy with a median of 4 anti-viral agents.

Patients received 3 weekly infusions of CMV-CTLs at doses of 1 x 10e6 T cells/kg (n=10) 2 x 10e6 T cells/kg (n=1) and 0.5 x 10e6 T-cells/kg (n=1) and were eligible to receive additional cycles of cells if they had no toxicity five weeks after the start of cellular therapy. One patient was evaluated only for toxicity as efficacy was confounded by anti-viral therapy required for treatment of varicella zoster. Of the 11 evaluable patients, 7 achieved CR, 1 PR (clinical improvement in disease and a 3 log decrease in viral load) and 1 SD. All but one of these responses were durable. One patient who had achieved a CR had recurrence of low grade viremia which was not retreated due to his overall medical deterioration. The two patients with progression of disease ultimately died of CMV.

Patients were monitored for expansion of CMV-CTLs by CMV-specific interferon gamma and where appropriate tetramer analysis. Responding patients consistently had detectable expansion of CMV-specific CTL populations. In addition, in one patient, undergoing serial lumbar punctures, we were able to detect third party CMV-CTLs in the CSF. There were limited toxicities associated with CMV-CTL infusions. No patient developed de novo GvHD or a flare of prior GvHD.

This study demonstrates a high response rate among patients with otherwise refractory CMV chorioretinitis or meningoencephalitis following adoptive therapy with primary donor or third party CMV-CTLs; thus demonstrating the capacity of adoptively transferred CMV-CTLs to treat disease in these sanctuary sites without toxicity. When selected based on restriction through shared alleles, third party CMV-CTLs are effective despite significant HLA disparity. The bank of CMV specific T cells can provide an immediate source of HLA partially matched appropriately restricted T cells for adoptive immunotherapy to treat CMV disease affecting the CNS. This enables treatment early in the course of disease with CMV-CTL lines previously prepared and characterized in terms of HLA restriction. This approach is anticipated to maximize the response rate as well as minimize toxicity from anti-viral therapy.