Successful adoptive transfer and in vivo expansion of haploidentical γδ T cells

In this pilot study, four subsequent patients with advanced refractory haematological malignancies (one T-NHL, one AML, one secondary plasma cell leukaemia, and one multiple myeloma) who were not eligible for allogeneic transplantation were included. HLA typing of the patients and their haploidentical family members were performed to select a donor with an HLA mismatch, which was detectable by FACS analysis with an HLA-specific antibody. Infusion of a CD4/CD8 T cell-depleted leukapheresis product (donor innate lymphocyte infusion = DILI) and Hi-Cy/Flu as prior immunosuppressive chemotherapy (fludarabine 25 mg/m² day -6 until day -2 and cyclophosphamide 60 mg/kg day -6 and -5) were performed as described[2, 4, 5]. However, according to the general conditions and/or prior therapies, the dose of cyclophosphamide was reduced by 50% in patients 1 and 2, and both cyclophosphamide/fludarabine by 25% in patients 3 and 4. All patients received intravenous (i.v.) zoledronate at a dose of 4 mg after DILI infusion on day 0 and subcutaneous (s.c.) 1.0x10⁶ IU/m² IL-2 on day +1 until day +6.

The collection of PBMCs and depletion of CD4⁺ and CD8⁺ cells were performed as previously described[13]. In brief, a single unstimulated leukapheresis was performed from family donors using a Cobe Spectra (Gambro BCT, Planegg-Martinsried, Germany). The cells were incubated with anti-CD4 and anti-CD8 antibodies conjugated to paramagnetic particles and then processed with the fully automated device CliniMACS® Plus (Miltenyi Biotec) using the program “Depletion 2.1”, according to the manufacturer’s instructions.

This pilot study was approved by the institutional review board and all participants gave written informed consent.

Our prior trials in the autologous setting clearly showed that the expansion of γδ T cells in vivo was a prerequisite for tumour regression, since only patients with significant in vivo proliferation of γδ T cells exhibited a tumour response[10]. Therefore, one major goal of our study was to investigate the feasibility and tolerability of selective in vivo expansion of the adoptively transferred haploidentical γδ T cells. For that, patients received a phosphoantigen (4 mg zoledronate) and low-dose IL-2 (1-2 × 10⁶ IE) s.c. for 6 days, starting on day 1 after transfusion. The results showed that a transient but significant expansion of donor γδ T cells and, to a lower extent, of donor NK cells and DN αβ T cells occurred in these patients (Figure 1). In all patients, proliferation peaked by around day +8, while on day +15, in three out of four patients there were no detectable donor cells, concurrent with the regeneration of recipient cells. The mean number of γδ T cells, NK cells and DN αβ T cells was approximately 43 cells/μL, 9 cells/μL, and 6 cells/μL, which represented a 68-fold, an eight-fold, and an eight-fold expansion, respectively. However, in patient 4, who showed the most profound γδ T cell expansion, cells were detectable for as long as 28 days after transfusion. Due to the underlying disease and multiple prior therapies including autologous stem cell transplantation, patient 4 was severely immunosuppressed, suggesting that the degree of immunosuppression may be a determinant in the duration of engraftment. Miller et al. have already shown that the survival of haploidentical lymphocytes is dependent on the extent of prior immunosuppression by comparing low- and high-intensity regimens[2].

The duration of neutropenia (< 500/μl) was 20 (range 11-33) days. Although it has not been shown that γδ T cells recognise alloantigens, potential hazardous effects of in vivo stimulated mismatched γδ T cells was of major concern. Although patient 4 died from severe septicaemia (Staphylococcus epidermidis + Enterobacter faecium) on day +45, with his haematopoiesis already recovered, none of the patients showed any signs of acute or chronic GVHD nor organ injury (Table 1).

Three out of four patients achieved a complete remission, which lasted between 2 and 8 months. For example, patient 3, who suffered from secondary plasma cell leukaemia and did not respond to 3 different regimens including autologous stem cell transplantation, also achieved a stringent complete remission (Figure 1).

It cannot be excluded, that the Flu/Cy regimen given prior cell transfusion for lymphopenia induction, induced a significant tumour regression in the patients. However, refractoriness to all other prior treatment approaches and the kind of underlying diseases argues against this assumption. Due to the small patient number, we were not able to differentiate which cell population was the main contributor for remission induction. Figure 1 shows the expansion of NK cells, αβ T cells and γδ T cells. It is very likely that the αβ T cell population consists at least in part of double-negative (classically Va24 + CD3+) invariant natural killer T (iNKT) cells. Given the known efficacy of iNKT cells against lymphoid tumour cells, it is possible that besides αβ T cells and “classical” NK cells, this subpopulation contributed to the anti-tumour effect, particularly in the cases of lymphoid malignancies[14, 17‐20].