Intra Bone Route of Donor Lymphocyte Infusions May Result with an Anti-Leukemic Effect Associated with an Accumulation of CD279+ Lymphocytes in the Marrow

Immunologic surveillance of leukemia is employed for the prevention and treatment of relapse post alloHSCT. The rate of success depends on the characteristics of leukemia which is more beneficial if the cells are recognized in the way of alloreactivity. Intra venous DLI is associated with a high risk of aGvHD and we believe that this route of administration may not make the direct contact between infused cells and blasts the optimal one. The previous reports documented a lower risk of aGvHD if the transplant material was given to the bone marrow cavity which at relapse is usually infiltrated by the blasts.

To address these issues, we started delivering donor lymphocytes directly to the bone marrow cavity (IB-DLI) in patients post alloHSCT at relapse.

This technique was employed in 4 patients, 3 with AML and one with CLL, all relapsed post alloHSCT: 3 alloSIB: 50-year-old female AML patient with normal karyotype (relapsed 2 years post HSCT) and 22-year-old AML male 7q31 del (relapsed at 3 years post HSCT) and 64-year-old CLL male, TP53 del, recurrent EBV reactivation with vasculitis (progressed 7 years), and 25-year-old male AML FLT3 ITD+ received MUD HSCT (relapsed 9 months).

Two patients with a proportion of 26% and 12% blasts in the marrow respectively received IB-DLI up-front and two others due to the excess (52.50%, 57.70% (CD5+CD19+)) of leukemic cells received either FLAG (AML case) or anti-CD20 MoAB (CLL case) followed by IB-DLI.

The patients received from 3 to 5 IB-DLIs according to the escalating dose regimen starting from 10E6 and ending with a dose of 10E8 of CD3+ cells/kg b. w. (blood MNC were harvested with use of Spectra Optia (Terumo BTC) for SIB transplantation, in MUD HSCT they were taken from the PBPC material (J Hasskarl et al. 2012). The intervals between IB DLI varied from 1 to 2 months. One patient received azacitidine between the DL infusions (Schroeder T et al. 2013) which was associated with longer intervals.

The cells were injected directly to the bone marrow cavity under local anaesthesia with a low molecular Heparin prophylaxis. The blood and marrow specimens were taken prior each IB-DLI for: cytology, cytometry (including CD8, CD279, CD26, CD28 MoAb in addition to the routine staining used for blast cells), genetic work (chimerism, mutations associated with the disease).

Clinical outcome:

1. No side effects were noticed (including GvHD).

2. The patients are alive.

3. Anti-leukemic effect:

3.1 Responding AML ITD+ case was free of blasts 6 months after the initiation of the treatment. However, after the third IB-DLI blasts appeared in the marrow but the patient responded favorably to the Sorafenib treatment and the following course of IB-DLI.

3.2 Partial remission (stable proportion of blasts and sustained hematopoiesis); the patient (22-year-old AML male) was transplanted a second time but blasts at the range of 38% were still observed in the marrow without progression after 9 months.

3.3 One case which failed to respond to IB-DLI (female AML patient) was transplanted a second time but this approach also failed and this patient now has full blown leukemia.

The laboratory work up:

1. Proportions of lymphocytes in the marrows tended to increase after completion of each IB-DLI (32.0 ± 4.6% vs 37.1 ± 3.7%, Wilcoxon test for pairs, p=0.078).

2. Collectively CD279+ cells contributed to the lymphocyte pool in the marrow to a greater extent than it was seen in the blood at the same time (16.6 ± 2.9% vs 33.1 ± 4.0%, p<0.001); this was also valid for CD8+CD279+ cells (12.4 ± 3.2% vs 27.1 ± 4.5%, p<0.001).

3. Microarray analysis of the transcriptome in the marrows of patients who received three IB-DLI courses revealed that in the patients responded favorably (CR or PR) their transcriptome profiles formed a cluster together with the transcriptomes of normal individuals in contrast to the patients who failed to respond, whose transcriptome profiles clustered separately.

IB-DLI was safe and not associated with GvHD. In all cases the infusion was associated with an increase in the marrow of lymphocytes being CD8+CD279+. The response may result in CR or PR and the patients were in a good physical shape during the treatment, which makes it possible to consolidate the treatment with a second alloHSCT if required. A lack of any response after the first courses of IB-DLI was associated with a failure to respond, even if the second transplant is performed.

Supported by The National Centre for Research and Development grant (INNOMED/I/1/NCBR/2014)