Abstract

Background: Evidence suggests that the outcome of allogeneic hematopoietic transplant in acute leukemia (AL) is, in part related to a graft versus leukemia effect. Recent studies suggest a role for natural killer cell (NK) alloreactivity in this process. We report here baseline findings relevant to the influence of the allograft content of natural killer (NK) cells on the outcome of matched related (MRD) and unrelated donor (MUD) myeloablative allografts for AL during a recent time period.

Methods: 19 pts received a 6/6 MRD peripheral blood stem cell (PBSC) allograft and 8 pts received a MUD marrow (BM) graft for high risk or relapsed AL between January, 1998 and September, 2003 (21 AML, 6 ALL). There were 14 males and 13 females; median age 41 (19 – 70); 13 were in 1st CR, 4 in 2nd or greater CR, 3 primary induction failures, 4 in 1st relapse and 3 in 2nd relapse. 21 received TBI conditioning (all but one with Cytoxan), 4 received Busulphan & Cytoxan. 17 MRD pts received PBSC grafts; 1 a CD34 selected graft and 1BM, while all MUD received BM. Graft characteristics, including CD34, CD3, CD16/56, CD4, CD8, and CD19 were measured prior to transplant. Patient outcomes, including relapse and death were obtained from a database maintained by the clinical trials office.

Results: After a minimum of 6 months from transplant, 7 MRD and 2 MUD pts are alive, a median of 20 months from transplant (range 6 to 66 months), and 18 died, a median of 4 months after transplant (range 0.5 to 13.2 months). Eight patients relapsed. Five patients died of relapse, 7 of infection, 1 of GVH, 3 of multiorgan failure, and 2 of other causes. MRD pts who relapsed were signficantly older than those who did not (median 49 vs 38; p = 0.04), and pts who relapsed received grafts that had a higher percentage of NK cells (medain 4.2%) than those who did not (median 2.7%, p = 0.04), but the number of NK cells/kg was similar in patients who relapsed versus those who did not, either within or between MRD/MUD groups. There was no difference between patients who relapsed vs those who did not with respect to the graft content of total leukocytes, CD34, CD3, CD4, CD8 or CD19, either within or between MRD/MUD groups. Likewise, there were no differences in graft contents of the in patients who survived vs those who did not. Compared with MRD pts, MUD pts, as expected received significantly fewer CD34/kg, NK and all lymphocyte subsets and had a longer hospital stay and time to ANC>500, but no difference in PD, GVHD, or survival (Table).

Conclusion: In this single institution experience there was no relationship between graft content of CD34, NK or CD3 cells and AL relapse or survival after myeloablative MRD or MUD transplant for high risk AL.

Patient Characteristics & Median Cells Infused/kg x 10E6

Donor group Patient Age ANC > 500, d LOS CD34 NK CD3 CD8 
LOS=length of stay; * p<.05 vs Unrelated 
Related 41 13.6 * 29 * 6.9 * 26 * 286 * 98 * 
+/− 95% CI 1.2 1.3 8.7 56 26 
Unrelated 42 16.6 41 3.7 5.5 38 19 
+/− 95% CI 10 10 0.9 2.9 11 6.6 
Donor group Patient Age ANC > 500, d LOS CD34 NK CD3 CD8 
LOS=length of stay; * p<.05 vs Unrelated 
Related 41 13.6 * 29 * 6.9 * 26 * 286 * 98 * 
+/− 95% CI 1.2 1.3 8.7 56 26 
Unrelated 42 16.6 41 3.7 5.5 38 19 
+/− 95% CI 10 10 0.9 2.9 11 6.6 

Author notes

Corresponding author