Fewer Relapses Following Unrelated Donor Cord Blood (UDCB) Compared to Related or Unrelated Bone Marrow (BM) or Peripheral Blood Cell (PBSC) Transplant in the Treatment of Childhood Relapsed and Very High Risk Acute Lymphoblastic Leukemia (ALL).

With aggressive primary therapy for ALL children who relapse are presenting to transplant after having received more intensive frontline and salvage therapy - resulting in transplant candidates with more resistant disease and at a higher risk for transplant related morbidity/mortality (TRM). Thus the the current generation of children presenting for allogeneic transplant are much more difficult to treat. With the addition of unrelated donor cord blood as an alternative hematopoietic progenitor cell source there is now an unrelated donor product available for a majority of children in a timely fashion. For the last 5 years we have treated all children with ALL coming to transplant with a standard preparative regimen which has minimally changed over that time period. Over the past year all children have received identical immunosuppression regardless of stem cell source. This allows us to analyze the impact of allogeneic stem cell source on the outcome of transplant - the graft-vs.-leukemia effect of the various HSCT sources. Between 3/2001 and 7/2006 a total of 54 transplants (txp) were performed at Texas Transplant Institute for treatment of childhood ALL (median age 8 yr, range 0–17 yrs). There were 13 txp for standard risk disease (CR1: Ph+, requiring more than 28 d to achieve CR1, severe hypodiploidy; CR2 with first remission >36 mo) − 7 CB and 6 BM/PBSC. There were 41 txp for high risk disease (CR2 with first remission <36 mo, >CR2 or not in remission at time of txp) − 31 CB and 10 BM/PBSC. All received TBI (1200 cGy in 6 fractions), cyclophosphamide 120 mg/kg, and either thio-tepa 10mg/kg or VP16 1500 mg/m2, or an additional 150 cGy TBI. There was no adjustment in the preparative regimen for the donor source. The first 26 CB transplants (UDCBT) received ATG 30 mg/kg x 3 days prior to txp, the subsequent patients did not receive ATG. There were 38 UDCBT, 14 matched sibling donor and one each of unrelated donor and partial matched family member (6 PBSC, 8 BM). GVHD prophylaxis was a calcineurin inhibitor combined with steroid or mini-methotrexate and/or sirolimus. The time to ANC>500 and platelet recovery was prolonged in the UDCBT group compared to BM/PBSC (19 vs. 14.5 d for ANC, 57 vs. 23 d for platelet >20,000). There were two graft failures in the UDCBT group - both CR1 patients and both salvaged with a second UDCBT. With median f/u of 270 and 350 days for CB and BM/PBSC respectively there was a significant difference in the risk for relapse (5/37 vs. 9/16; p<0.03), TRM (8/37 – 1 GVHD, 3 MOSF, 2 bacterial, 1 leukoencephalopathy, 1 pulmonary vs. 2/16 -fungus, LPD; p= 0.7). KM survival analysis is shown below:

Importantly there have continued to be late deaths in the BM/PBSC group due to relapse with the CB curves remaining stable after 18 months post txp. Despite 80% of the CBT being performed for children with high risk disease there were fewer relapses in the UDCBT cohort suggesting a robust allogeneic effect against ALL in the setting of UDCBT.