By David N. Leff

Throwing out the baby with the bath water is practiced every day - at the cellular level - by hematologists treating patients with blood-cell malignancies, mainly leukemia and lymphoma.

Their strategy consists of replacing all of the patient's cancer-ridden bone marrow - which produces the blood-forming stem cells and their progeny, including the elements of his or her immune system - with a transplant of healthy marrow from a willing donor. Unless the recipient host is lucky enough to have a twin sibling who can contribute 100 percent immunologically identical marrow, the next best bet is someone with a less than perfectly matching graft. (See BioWorld Today, June 3, 1999, p. 1.)

But that often is a fatal, poisoned gift, called graft-vs.-host disease (GvHD). It happens when the donor's T lymphocytes encounter the host's mismatched antigenic cells, and attacks them with intent to kill. Primarily for this reason, a patient, before bone marrow transplantation (BMT), is prepped by intensive, whole-body irradiation and immunosuppressive drugs to reduce the cancer burden, and presumably leave the marrow with no active cells to provoke that donor reaction. Nevertheless, 20 percent to 40 percent of such bone marrow transplant patients eventually die from the complications of GvHD.

Recent efforts expunge those donor T cells by siccing antibodies on them. This ploy is what immunologist Stephen Emerson deplores as throwing out the baby with the bath water. He is chief of hematology/oncology at the University of Pennsylvania, and senior author of a paper in today's Science, dated July 16, 1999. Its title: "Prevention of graft versus host disease by inactivation of host antigen-presenting cells."

"If you think about allogeneic [unmatched] bone marrow transplantation," Emerson told BioWorld Today, "from the time you do a transplant, for a few short months, you have the residual immune system of the host, which hasn't died off yet. Also you have the donor graft's immune- system cells that the bone marrow's stem cells make. Eventually," he pointed out, "there's total reconstitution and replacement of the host's immune system by the donor's immune system, but initially both cells are there.

"What we asked in our research," Emerson recounted, "was whether the graft-vs.-host process is triggered initially by donor immune cells that pick up host antigens and process them for destruction, or instead is triggered by radioresistant host antigen-presenting dendritic cells - APCs - that hadn't died off yet. Or did host antigens present to donor T cells - or both?"

As his Science paper reported, "The answer, at least in our experimental mouse system, was stunning. Only the radioresistant residual host APCs initiated GvHD, not the donor T cells. So the bottom line in terms of transplantation suggests that, if lucky, one ought to be able to completely prevent - or, if less lucky, largely prevent - GvHD, by poisoning or eliminating residual host APCs on the transplant."

Bird Dogs Explicate Preclinical In Vivo Trials

Here is how Emerson described his initial, and crucial, animal BMT experiment:

"What we needed to do was create a mouse as a potential recipient target for GvHD, whose tissues could be attacked, but whose triggers for GvHD - the antigen-presenting dendritic cells - could not work. We did that before the transplantation by making a mouse that was a bone-marrow chimera. That is, its bone-marrow cells, and all those cells that the marrow's stem cells produced, were obtained from a knockout mouse, which couldn't make a protein called beta-2-microglobulin. This is a small peptide," he explained, "which is essential for putting transplantation antigens on the surface of cells."

To clarify this multi-step procedure, Emerson proffered an analogy: "It's as if you have a golden retriever who's helping you hunt ducks. And the way it works is by picking up a duck in its teeth and bringing it back to its master, saying, 'Here's a duck to hunt.' The dog's teeth are the major histocompatibility class I molecule - those dendritic APCs that present antigens to the immune cells. We created, in effect, a forest where all the retrievers had no teeth. So they could beat the rushes, but they couldn't pick up any ducks and bring them back to their T cells."

Reverting from field to lab, Emerson continued: "Then we gave those mice allogeneic transplants, with stem cells and T cells that were from an unmatched donor, as you would in a clinical transplant. Even though the donor T cells were active, and even though the donor stem cells could make their own APCs, the animals' donor dendritic cells were not sufficient to trigger their own T cells. They didn't pick up host antigens, process them, and present them to T cells.

"That finding would be a major deal if it would work clinically," Emerson added, "because we could avoid any immunosuppression of the graft whatsoever, no T-cell depletion, because there'd be no reaction triggered. No one has actually done that until now."

He said, "By trying to poison APCs instead of T cells, you could temporarily hasten the elimination of host dendritic cells. Get them out of the system quickly. Then when you allowed the patient to repopulate with donor-derived cells, it wouldn't matter. They wouldn't cause GvHD."

Add Sickle Cell Anemia To Potential Cures

Replacing host APCs with donor APCs in future practice, he suggested, "would spare the important T cells to rebuild the transplant immunity and fight off tumor cells. Moreover," Emerson pointed out, "there are lots of conditions in which we know that bone marrow transplantation would cure the disease, yet that BMT approach is too toxic to recommend it. For example, sickle-cell anemia is a bone marrow-derived disease. All the sickled cells come from mutant stem cells you've inherited. If we could transplant those cells safely, we'd offer it to those patients, and cure their disease."

The co-authors' work, Emerson observed, "is right at the beginning, but the actual development could be pretty rapid. What we need to do is develop antibody conjugates, either toxin or isotope, to remove dendritic cells. Then we simply need to make the right toxins with the antibodies, find the one that works, then do it in clinical trials.

"At the moment," he went on, "we would love to find corporate partners who have expertise in doing the toxin conjugation, and want to develop our approach as a therapeutic. We're just academics, not drug-development people. It depends on how rapidly we can get someone to give us a little bit of a hand." He concluded, "We could actually have an IND within a year."