Introduction

Haploidentical hematopoietic stem cell transplantation (HSCT) is used in patients without matched donors.^{1, 2, 3, 4} Nearly all patients have a haploidentical donor within the family and, in contrast, to an unrelated donor search, the haploidentical family donor is rapidly available. The role of natural killer (NK)-cell alloreactivity in haploidentical HSCT⁵ to promote engraftment and graft-versus-leukemia (GvL) effects is increasingly recognized.

In haploidentical HSCT, where T-cell-mediated effects are eliminated by T-cell depletion, alloreactivity is provided by NK-cell mismatches. In patients with acute myeloid leukemia, donor vs host NK-cell alloreactivity is associated with a remarkable GvL effect, apparently without increased risks of graft-versus-host disease (GvHD). This and conversely, the absence of host-versus-graft NK-cell alloreactivity may be explained by NK-cell sensitivity to class I polymorphism being restricted to hematopoietic cells. Other tissues may lack the ligands to activate NK cells. Thus, NK cells may, mediated by killer cell immunoglobulin-like receptors (KIR), prevent rejection and promote GvL effects.^{6, 7, 8, 9}

We investigated in a pilot protocol the feasibility of preparing and infusing purified, T-cell-depleted, donor NK-lymphocytes (NK-DLI) to consolidate incomplete engraftment in patients after haploidentical HSCT. We hypothesized that purified NK-lymphocytes would preferentially recognize hematopoietic host cells, and promote engraftment without GvHD.

Patients and methods

Objectives were to collect, purify and infuse NK-DLI in patients with incomplete engraftment, or relapse in the absence of GvHD after haploidentical HSCT. Targeted cell doses were 1.0 10⁷/kg CD56+/CD3- NK cells, with <1.0 10⁵/kg contaminating CD3+ T cells. The study is approved by the Local Ethics Committee.

Transplantation protocol

The haplo-HSCT protocol has been described previously¹⁰ and includes: G-CSF-mobilized peripheral stem cells, a T-cell-depleted (<1.0 10⁵/kg CD3+) graft with a high dose of CD34+ cells (10 10⁶/kg), a pretransplant conditioning regimen with etoposide, cyclophosphamide, ATG and 12 Gy of fractionated total-body irradiation, post-transplant immunosuppression with OKT3 and cyclosporine for 10 and 14 days. Donor cell engraftment after HSCT was monitored in whole blood 30, 60 and 90 days post-transplant, and three monthly thereafter, using a PCR-assay analyzing polymorphic short tandem repeats (STRs).

Patients

Five of 16 consecutive recipients of haploidentical HSCT between January 2000 and July 2003 were included. Characteristics of these patients are listed in Table 1. All were children or young adults with high-risk hematological malignancies. Donors were parents. NK alloreactivity predicted based on HLA typing was present in 3/5 patients. All received a high stem cell dose, with low T-cell numbers; and all had rapid neutrophil, platelet and red cell engraftment, without developing GvHD. One patient had secondary graft failure at 5 months (UPN960). Indication for NK-DLI was incomplete engraftment in three, early relapse in one (UPN953) and graft failure in one patient (UPN960).

Table 1 - Patients and donor lymphocyte products characteristics showing results of each processing step.

Full table

NK cell collection, processing and infusion

Mononuclear cells were collected from the original donor by 10 l leukapheresis and stored at 4°C overnight. The two-step ex vivo purification procedure (CliniMACS^® cell selection system) included first a T-cell depletion, and second, an NK-cell selection. At each step, cells were analyzed for CD3, CD56 and CD19 by flow cytometry. In case of a CD56+/CD3- cell dose 1.8 10⁷/kg or a contamination of T cells 1.0 10⁵/kg, the processed NK product was split into two units. The first unit was infused immediately and the second was cryopreserved. NK-DLIs were performed as an outpatient procedure. Cryopreserved NK units were thawed at the bedside and infused rapidly. Patients were monitored for immediate adverse reactions, GvHD and infections. Donor chimerism for engraftment after HSCT and monitoring of NK-DLI was assessed by a PCR-based assay analyzing polymorphic STR markers. Amplified alleles were separated by capillary electrophoresis and peak surface area was used to quantitate chimerism. Whole-blood chimerism was measured prior to and monthly post-transplant. The sensitivity of the technique allows detecting a minor patient fraction of 1% of the entire leukocyte population.

Technical aspects

For removal of CD3+ T cells, mononuclear cells were incubated for 30 min at room temperature with the anti-CD3 antibody, directly conjugated to magnetic microbeads. (CliniMACS T-Cell CD3 MicroBead^®). The program DEPLETION 2.1 was used for automated cell separation. CD3- cells were collected and enriched for CD56+ NK cells by adding CliniMACS CD56 MicroBeads, incubating for 30 min and running the program ENRICHMENT 1.1.

Results

Six products were collected from five donors. For patient UPN 931, a second harvest was performed 6 months after NK-DLI. The overall processing time lasted 8–10 h. After processing, purity of CD56+/CD3- cells was 97.3% (median; range 77.9–98.9) with a recovery of NK cells of 35.5% (13.1–75.0). The main loss of CD56+ cells occurred during the first step, with a reduction from 164 10⁷ (17–301) CD56+/CD3- NK cells before to 53 10⁷ (13–68) after the second step (Table 1). Overall, a T-cell depletion of 3.55 (2.9–4.5) log was achieved. The total T-cell count was reduced from 481 10⁷ (284–1539) to 0.134 10⁷ (0.05–0.518).

Of the six products four were split – three because of high NK cells, one because of high T cells, one because of both – to obtain 10 infusable products. Prior to splitting, CD56+/CD3- cells were 1.61 10⁷/kg (median; range 0.21–2.20) and CD3+ T-cell count was 0.29 10⁵/kg (0.11–1.10). The infused products contained 0.93 10⁷/kg (0.21–1.41) CD56+/CD3- cells, with a T-cell contamination of 0.22 10⁵/kg (0.11–0.55). One product (UPN960) had a very low number of NK cells, despite identical leukapheresis. This donor had a low preapheresis lymphocyte (1.13 10⁹/l) and NK-cell counts (CD56+/CD3-, 68/l).

There were three NK products available for patient UPN931, one for UPN960 and two for all others. Nine of the 10 NK donor lymphocyte products were used for infusion. During infusion of the fresh or thawed NK products, no immediate adverse reactions were observed. None of the five patients developed clinical signs of acute or chronic GvHD after NK-DLI. There were no infectious complications and so far no other late effects attributable to NK-DLI have been observed. As of January 1, 2004, four of the five patients are alive, well and in continuous remission 8–18 months after the first NK-DLI (median follow-up, 12 months). One patient had rapidly progressive AML (UPN953) after NK-DLI and did not receive a second dose. After two NK-DLI, UPN924 has possibly stabilized at a donor chimerism of 35% and UPN853 continued to drop to 70% (Figure 1). UPN931 had an increase in donor chimerism to 100% and because of a subsequent drop has received a third dose of NK-DLI 6 months after the first dose. She remains a complete chimera. UPN960 received NK-DLI followed by a second stem cell dose because of impending graft failure. She responded to this treatment by increasing donor chimerism from 26 to 66%, and neutrophil counts from 0.04 to 0.65 10⁹/l but remains transfusion dependent.

Figure 1.

Percent donor chimerism in the five patients before and after treatment with NK-DLI. Arrows (on top) represent time of NK-DLI, UPN953 had only one infusion (represented by the first arrow); UPN853, 924, 960 had two infusions (represented by the first and second arrow) and UPN931 had three infusions (represented by all three arrows). UPN960 received an additional stem cell boost (asterisks). The X-axis represents time prior to and after the first NK-DLI in months.

Full figure and legend (58K)

Discussion

This study shows that ex vivo purification of donor NK cells from a leukapheresis product is technically feasible, and an adequate number of CD56+, highly CD3-depleted cells can be obtained and infused without immediate adverse events and without inducing GvHD. The processing of donor lymphocyte cells is time consuming, and requires specific knowledge and skills in graft engineering. Finally, our preliminary data suggest that NK-DLI may revert impending rejection in some patients.

This pilot study allowed to set up technical conditions to generate a product with high purity of NK cells and a maximal depletion of T cells, which can be used in a clinical setting. This clinical scale method for isolation of T-cell-depleted CD56+ donor NK-lymphocytes has been described previously, but purified NK donor lymphocyte have only been used rarely as adoptive immunotherapy after HSCT.^{11, 12} Whereas NK-cell purity was 97% and T-cell depletion was by 3.5 logs, we observed a considerable loss of NK cells during cell engineering. Overnight storing of collected cells may have contributed to this. Overnight storing was chosen in order to guarantee completion of in vitro processing and donor infusion on the same day, including the flow cytometry analysis and, if necessary, the cryopreservation of the split product. The target cell number was fixed somewhat arbitrarily to 1.0 10⁷/kg NK cells. This was based on the experience acquired with standard DLI.^{13, 14}. In 15 unmanipulated DLI obtained by 4–6 l leukapheresis, we collected 48 10⁷ (12–115) NK cells. Increasing the apheresis volume could permit the collection of sufficient numbers of NK cells in view of their subsequent preparation. Prerequisites to achieve an adequate number of NK cells for NK-DLI are a high number of CD56+ cells in the leukapheresis product and efficiency of in vitro processing.^{15, 16} The only product with low NK-cell numbers was derived from a donor with lymphopenia and low NK count.

Monitoring of chimerism before and after NK-DLI was performed on whole blood. Analysis of subpopulations is not available. Early after haploidentical HSCT, the majority of peripheral blood cells are granulocytes and lymphocyte reconstitution is usually slow with NK cells recovering faster than T-cell subsets. In this pilot study, the timing of NK-DLI after HSCT was quite variable.

We cannot exclude that effects observed were due to residual T cells infused rather than NK cells. The upper limit of acceptable T-cell contamination had been set at 1.0 10⁵/kg BW in the protocol. The upper limit is the same for the haploidentical HSCT as for the NK-DLI product. In fact, patients received a median of 0.22 10⁵/kg contaminating CD3+ cells, and none of the products contained more than 0.55 10⁵/kg CD3+ cells, which is a T-cell dose considered acceptable in some protocols for haploidentical HSCT. Further reductions of contaminating T cells could be obtained by additional processing, at the expense of an increase in processing time, however.

Infusion of NK-DLI was well tolerated and none of the patient had GvHD. NK-DLI resulted in increasing donor chimerism in two patients, but one of them had received an additional stem cell dose after NK-DLI. Stabilization of donor chimerism was found in one patient and decreasing donor chimerism in two; one of them had early relapse. This study was not designed to test the efficacy of NK-DLI; nevertheless, early results appear encouraging. This pilot study has several limitations. The patient number was small and heterogeneous. We do not have data on long-term follow-up. We did viability testing on the NK cells but we do not have data on their functionality. Finally, KIR alloreactivity was estimated based on HLA typing and not measured directly in these donor–recipient pairs.

In conclusion, purification of NK cells for infusion is technically feasible and possibly safe. To evaluate the role of NK-DLI in consolidation of engraftment and prevention of relapse, we have started a larger phase II study, evaluating pre-emptive donor NK-cell infusions after haploidentical HSCT.

References

1. Aversa F, Tabilio A, Velardi A, Cunningham I, Terenzi A, Falzetti F et al. Treatment of high-risk acute leukemia with T-cell-depleted stem cells from related donors with one fully mismatched HLA haplotype. N Engl J Med 1998; 339: 1186–1193. | Article | PubMed | ISI | ChemPort |
2. Lacerda JF, Martins C, Carmo JA, Lourenco F, Juncal C, Rodrigues A et al. Haploidentical stem cell transplantation with purified CD34 cells after a chemotherapy-alone conditioning regimen. Biol Blood Marrow Transplant 2003; 9: 633–642. | Article | PubMed |
3. Rowe JM, Lazarus HM. Genetically haploidentical stem cell transplantation for acute leukemia. Bone Marrow Transplant 2001; 27: 669–676. | Article | PubMed | ChemPort |
4. Veys P, Amrolia P, Rao K. The role of haploidentical stem cell transplantation in the management of children with haematological disorders. Br J Haematol 2003; 123: 193–206. | Article | PubMed |
5. Farag SS, Fehniger TA, Ruggeri L, Velardi A, Caligiuri MA. Natural killer cell receptors: new biology and insights into the graft-versus-leukemia effect. Blood 2002; 100: 1935–1947. | Article | PubMed | ISI | ChemPort |
6. Giebel S, Locatelli F, Lamparelli T, Velardi A, Davies S, Frumento G et al. Survival advantage with KIR ligand incompatibility in hematopoietic stem cell transplantation from unrelated donors. Blood 2003; 102: 814–819. | Article | PubMed | ISI | ChemPort |
7. Ruggeri L, Capanni M, Casucci M, Volpi I, Tosti A, Perruccio K et al. Role of natural killer cell alloreactivity in HLA-mismatched hematopoietic stem cell transplantation. Blood 1999; 94: 333–339. | PubMed | ISI | ChemPort |
8. Ruggeri L, Capanni M, Urbani E, Perruccio K, Shlomchik WD, Tosti A et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science 2002; 295: 2097–2100. | Article | PubMed | ISI | ChemPort |
9. Davies SM, Ruggieri L, DeFor T, Wagner JE, Weisdorf DJ, Miller JS et al. Evaluation of KIR ligand incompatibility in mismatched unrelated donor hematopoietic transplants. Killer immunoglobulin-like receptor. Blood 2002; 100: 3825–3827. | Article | PubMed | ISI | ChemPort |
10. Passweg JR, Meyer-Monard S, Gregor M, Favre G, Heim D, Ebnoether M et al. High stem cell dose will not compensate for T cell depletion in allogeneic non-myeloablative stem cell transplantation. Bone Marrow Transplant 2002; 30: 267–271. | Article | PubMed | ChemPort |
11. Lang P, Pfeiffer M, Handgretinger R, Schumm M, Demirdelen B, Stanojevic S et al. Clinical scale isolation of T cell-depleted CD56+ donor lymphocytes in children. Bone Marrow Transplant 2002; 29: 497–502. | Article |
12. Iyengar R, Handgretinger R, Babarin-Dorner A, Leimig T, Otto M, Geiger T et al. Purification of human natural killer cells using a clinical-scale immunomagnetic method. Cytotherapy 2003; 5: 479–484. | Article | PubMed | ChemPort |
13. Sato H, Shiobara S, Yasue S, Chuhjo T, Nakao S. Lymphocyte collection for donor leucocyte infusion from normal donors: estimation of the minimum processed blood volume and safety of the procedure. Vox Sang 2001; 81: 124–127. | Article | PubMed | ChemPort |
14. Hester J. A clinical role for peripheral blood lymphocyte infusions and perspectives on collection. Therap Apher Dial 2003; 7: 305–311.
15. Guglielmi C, Arcese W, Dazzi F, Brand R, Bunjes D, Verdonck LF et al. Donor lymphocyte infusion for relapsed chronic myelogenous leukemia: prognostic relevance of the initial cell dose. Blood 2002; 100: 397–405. | Article | PubMed | ISI | ChemPort |
16. Korbling M, Giralt S, Khouri I, Mirza N, Donato M, Anderlini P et al. Donor lymphocyte apheresis for adoptive immunotherapy compared with blood stem cell apheresis. J Clin Apheresis 2001; 16: 82–87. | Article | PubMed |

Acknowledgements

We thank Katherine Haupt and the cell collection team, Laura Landi and the cell engineering team, Heike Huxol and the team responsible for chimerism analysis as well as Silvia Mathys and the FACS team for their help. We further thank Miltenyi Biotec for their technical support in NK. This work was supported by a SAKK-Swiss group for Clinical Cancer Research Pilot Project Grant.