Trial Information

A Prospective, Multicenter Randomized Study Comparing Single Versus Double Umbilical Cord Blood Transplantation in Children and Young Adults (<35 Years) With Acute Leukemia Remission

Background and rationale:

Unrelated cord blood transplantation (UCBT) has been used for several years when there is no
HLA identical sibling or unrelated donor. During the year 2007, 218 of the 783 unrelated
hematopoietic stem cell transplantation (28%) carried out in France were UCBT. This
proportion is 37% in children (58/157). Clinical outcome after UCBT strongly depends on the
transplant cell dose. Since the recent publication of encouraging results after
transplantation of two UCB units, the number of these double-transplantations increases in a
very significant way. Thus 121/218 UCBT performed in France during 2007 were
double-transplantations (55%) whereas this proportion was 81/180 (45%) in 2006, 30/144 (21%)
in 2005, 5/77 (6%) in 2004 and 0/44 in 2003. However, there is currently no prospective
study comparing in a reliable way the double-transplantation to single-transplantation
results.

Study design:

The investigators propose a prospective and randomized study comparing the results of single
versus double unit UCBT in children and young adults (< 35 yrs) with acute leukemia in
remission. This is an open, multicenter study carried out in the allogeneic transplant
centers from the French society for hematopoietic stem cell transplantation and cell therapy
(Société Française de Greffe de Moelle et de Thérapie Cellulaire, SFGM-TC).

Objectives:

The primary objective is to compare the incidence of transplantation failure in the two
treatment arms. Transplantation failure, the primary endpoint of the study, is defined by
the occurrence of one of the following events : transplant-related death, second allogeneic
transplantation or autologous backup infusion for primary engraftment failure, autologous
recovery. The financial impact of these double-transplantations being to date unknown, the
project also includes a cost-effectiveness study, the effectiveness criterion being a
decrease in transplantation failure incidence. The secondary clinical endpoints are: overall
survival and disease-free survival, relapse incidence, transplant-related mortality,
incidence of severe infections and GvHD. The secondary biological endpoints are:
hematological and immunological recovery, post transplant chimerism.

Methods:

Transplantation methods: Myeloablative conditioning regimen includes, according to the
patient age, either total body irradiation, fludarabine and cyclophosphamide with a GvHD
prophylaxis based on cyclosporine A and mycophenolate, or the association busulfan,
cyclophosphamide and anti-thymocyte globulin with GvHD prophylaxis being cyclosporine A and
steroids.

Statistical methods: The experimental schedule is based on a multiple testing procedure,
using the method described by O' Brien and Fleming. This offers the possibility of stopping
the trial before the end of inclusions if a significant difference between the two arms
occurs. In this trial, two sequential analyses are planned: an interim analysis at 18 months
and a final analysis at 36 months. Probabilities of survival and DFS are estimated according
to the Kaplan-Meier method. In the presence of one or several competitive risks, the
cumulative incidence of an event is estimated according to the method of Gray. Thus, relapse
is a competing risk for transplantation failure in the study primary endpoint evaluation.
Comparisons between the two treatment arms are carried out by the Log rank test for
Kaplan-Meier estimates and by the Gray's test for the cumulated incidences. The intent to
treat analysis (according to the random allocation) will be preferred to a per-protocol
analysis which will be an additional analysis. An independent committee will be made up in
order to control the decisions of study continuation or stopping at time of the interim
analysis.

Sample calculation, study feasibility and duration:

In the setting of a sequential trial including one interim analysis, with a cumulated
incidence of transplantation failure hypothesis being 40% in the single-transplantation and
20% in the double-transplantation arm (alpha risk, 5%, power, 80%, 5% non evaluable
patients), the estimate of sample size is 99 by group, i.e. a total of 198. Taking into
account the UCBT activity in France (data from the French biomedicine agency), the enrolment
phase of the study is planned to last 30 months. Minimum post-transplant follow-up duration
being 6 months, duration of the study is 3 years.