Allogeneic (Allo) hematopoietic stem cell transplantation (HSCT) is a recognized curative procedure for hematological malignancies. It is now well known that this property is related to the graft-versus-tumor (GVT) effect developed from the immunocompetent cells contained in or generating from the donor graft. For years, however, and despite this unique antitumoral activity, Allo-HSCT has been restricted to a limited number of patients due to two major limitations: the toxicity of the procedure and the absence of a donor for every single patients. More recently the stage has dramatically changed with respect to these two restraints. Over the last decade, many studies have established the feasibility of Allo-HSCT in older patients but the availability of MRD is even less frequent in elderlies, likely related to medical contraindication for graft donation or sibling deaths. UD are routinely used but associated with a high incidence of GVHD. As compared to younger populations, unrelated cord-blood HSCT is seldom performed in this population and numbers decrease with age due to the feared risk of supposed increased lethal infectious complications related to the effect of the delayed immune reconstitution in elderlies. Thus the need for alternative donors allowing for a safe and efficient transplantation is still unmet. In consequence, overall, despite the fact that Allo-HSCT feasibility has been established in the oldest patients, all these lacks contribute eventually to maintain a low rate of allo-HSCT performed in a population with the higher incidence of hematologic malignancies that usually present with the poorest prognosis.

Thus it is critical developing innovative efficient therapeutic strategies answering this unmet-medical need. In this perspective, Haplo-HSCT could represent a part of the answer in this aged population. It offers the potential advantage to offer a rapid donor determination for virtually every single patient. In addition, our data suggest that in elderlies haplo-HSCT using T-repleted graft, RIC and PT-HDCy presents low NRM and retains an antitumor effect despite low GVHD incidences. They also suggest that haplo-HSCT may conduct to better outcome than URD-HSCT as an alternative to MRD-HSCT. It may also be associated with lower costs (no graft purchase and low post-transplant complications rate) and better QOL likely related to low cGVHD-rate. In addition the conduct of such trial at a time when the diffusion of the strategy in this population is just starting is really crucial before widespread uncontrolled dissemination.

The investigators propose to address this question by prospectively comparing these 2 strategies in elderly patients without MRD, in terms of efficacy, safety and including the prospective evaluation study of quality-of-life (QOL). They will conduct a national, multicenter, open-label, comparative, randomized phase III trial in patients with hematological malignancies justifying an allo-HSCT from an alternative donor when a MRD has not been identified.

When MRD search is recognized to be a failure, patients will be included in the clinical trial after informed consent and randomized in the two strategies based on donor search:

• Reference group: Unrelated Donor group
• Investigational group: Haplo Donor Group

Investigators will use a composite end-point embracing the three main causes of failure: death, relapse and severe cGVHD (as a surrogate endpoint for QOL). We will analyze the HSCT usual objectives as GHVD, NRM, relapse and survival. A specific study of patients' health related quality of life will also be conducted using the FACT-BMT questionnaire. In addition, the success of the two strategies in term of transplant completion (donor determination, transplant realization and time to do so) will be compared.

Inclusion Criteria:

• Patients aged of 55 years or up to 70 years.
• Patients with hematological malignancy
• Patients without a matched related donor
• Patients eligible for an allogeneic HSCT from an alternative donor
• Able to comply with the protocol
• Written informed consent
• Affiliation to Social Security System

Exclusion Criteria:

• Clinical or biological contraindication to allogeneic HSCT
• Other evolutive cancer
• Positive serology for HIV, hepatitis B or chronic active hepatitis C
• Pregnant or breast-feeding women.
• Emergency
• Patient considered socially or psychologically unable to comply with the treatment and the required medical follow-up.