All the allografts were removed during multiorgan removals. The grafts were repeatedly checked, i.e., checked when they were removed and at the fourth month, in the host after three successive bacteriological, urological, and immunological examinations had been carried out. The first on the donor, the second on the patients who had been given organs coming from the same donor. Any infection – general, viral, parasitic, or tumoral – or a systemic illness immediately leads to the destruction of all the grafts removed. The grafts are removed in an operating theater under surgically sterile conditions and preserved at −196 °C (in liquid nitrogen, after a progressive lowering of the temperature by 2 °C/min down to −40 °C, then by 5 °C/min down to −140 °C. The DMSO is used systematically at a dosage of 10 % and put into contact with the allograft after the latter has been refrigerated when its temperature reaches +4 °C. The allograft and the DMSO are cooled separately and are placed in contact with each other at this temperature (above +10 °C, DMSO is toxic to cells). Thawing has to be rapid, i.e., 1–2 h, with lavage of the DMSO in Ringer’s lactate solution at +40 °C.

The first experimental joint transplants date from the beginning of the century, with the work of Henri Judet in 1908. The autologous and homologous osteochondral grafts performed then were complicated by progressive deterioration of the cartilaginous tissue with crumbling of the subchondral necrotic bone. The first massive osteochondral auto-transplants were performed by Reeves and Solmes (1966), Judet and Padovani (1973), and Goldberg et al. (1973 and 1980). Although the articular auto-transplants demonstrated excellent viability and are maintained in the long term, the allo-transplants produced acute rejection which took two forms:
This was immunological rejection of the supply vessels. The transplant was invaded with lymphocytic cells. Necrosis progressed more rapidly when a homograft of skin or marrow preceded the transplant. Judet and Padovani are of the opinion that a purely immunological mechanism is involved due to the immediate revascularization which created the conditions for accelerated rejection by offering immediate contact between the antigen and the antibody. The delays in assimilation of the graft and its quality only appeared to be slightly altered by the use of cyclosporine A.

Immunosuppressant drugs, such as azathioprine, do not significantly increase the survival time of the transplants. According to Halloran, the anti-lymphocytic serum gives a survival time for the transplant of up to 5 months, but only in the semi-allogenic grafts, and retains identical growth potential to that of the contralateral limb. In our experience, which between 1978 and 2000 includes 185 massive osteocartilaginous transplants, it seems that the clinical result is all the more successful if satisfactory biomechanical conditions are restored to the best extent possible. Histological studies performed by drilling under arthroscopy, even 8 years after the graft, showed that the chondrocytes were alive in most cases and that although the superficial layers of the cartilage were sometimes changed and fissured, the deep layers were generally normal. The absence of painful symptoms in patients is a bonus and X-rays do not show any necrotic phenomena or crushing of the grafts. Some people fear the appearance of a tabes-like syndrome when the two parts of the joint are grafted jointly. It is true that the bone itself is no longer innervated and it is unlikely that a new proprioception will reappear during revascularization of the bone tissue. But the ligaments, muscles, and peripheral vasculonervous elements which have not been removed provide the medullary nerve centers with information and stabilize the bone and cartilage, which in fact behave like a metal prosthesis (also not innervated) [44, 45]. Clinical experience confirms this interpretation, as no tabes-like symptoms have been reported, even 15 years after these types of massive osteocartilaginous grafts were implanted [46, 47].

Preserving human ligaments in tissue banks is also an avenue of research which appears promising but which comes up against the problem of supplies from tissue banks and of the mechanical behavior of ligaments grafted during the period of their revascularization.

The reconstruction of a half-pelvis requires a broad approach following in its middle part the iliac crest, in its anterior part the crural arcade curving in from the pubis along the pubioischiatic line, and in its posterior part, continuing horizontally up to the level of the spinous processes of the lumbar vertebrae. An extension upwards or downwards may make it possible to approach the last lumbar vertebrae or the sacrum. The initial locating of the external iliac vessels, the femoral vessels, and of the crural nerve makes it possible to perform a subperitoneal dissection of the tumor which generally pushes back the iliac muscle and only exceeds it at a late stage. This extension then often contra-indicates a carcinological surgical maneuver. The prosthesis is inserted and cemented in the half-pelvis allograft after fixing the latter by anterior and posterior plates screwed into the allograft on the remaining bone. When a whole half-pelvis is to be reconstructed, the gluteal muscles on the one hand and the adductors on the other have to be refixed to the bone by trans-bone sutures. In order to avoid the occurrence of an obturator hernia, the obturator is obstructed by a strip of silastic pressed on trans-bone sutures at the edge of the obturator.