Mesenchymal stem cells may lose their immune-protected status if exposed to infection. Credit: Pranela Rameshwar, New Jersey Medical School in Newark

Even with immunosuppression, bone marrow taken from one person and transplanted into another can trigger the body's defenses, resulting in fever, sickness and even death. It's no surprise, then, that doctors hunt hard for donors that are close genetic matches to the patient. But sometimes this isn't enough. Even if the host's body accepts the transplanted bone marrow, immune cells from the donor can attack the recipient, resulting in graft versus host disease—a rare condition that leaves few options for treatment.

If a bone marrow transplant triggers GVHD, mesenchymal stem cells are sometimes used in a second transfusion. Unlike the haematopoeitic stem cells used during the initial bone marrow transplant to treat leukemia, mesenchymal stem cells help calm complications that arise if donor cells from the first transfusion begin to attack the host. Many researchers believe that mesenchymal stem cells actively calm these wayward donor cells. (While mesenchymal stem cells are also taken from the bone marrow, they make different cell types and are prepared differently than haematopoeitic cells).

Mesenchymal stem cells are also widely thought to be immunotolerant, and thus less likely to trigger a second reaction. Yet transplant dogma maintains that mesenchymal stem cells that are closely matched to the patient are the safest option—even though this requires going through the painstaking process of tissue typing, all while the clock is ticking for a seriously ill GVHD patient.

Now a study published this May in the Lancet indicates that such close genetic matching might not make such a big difference, sparking debate in nearly every corridor of stem cell researchc. The fifty-five patients selected had undergone a former transplant, and faced a dismal 70% mortality rate due to their resistance to traditional drugs for GVHD. In an effort to suppress misbehaving cells, each patient received transfusions of mesenchymal stem cells--some got cells from relatives or other matched donors, while others got cells from mismatched donors with unknown compatibility. Nine patients were treated with cells that had been combined from more than one donor.

"The most interesting discovery is that the patients didn't respond any differently to the mismatched stem cells than those from genetically compatible donors," says Willem Fibbe, a professor of stem cell biology at Leiden University in the Netherlands, and an author on the paper. In other words, genetically similar stem cells were no more beneficial than mismatched ones. In each case more than half of the patients recovered completely, and 71% saw at least a partial response.

While Fibbe and his colleagues didn't take cells from the recipient's own bone marrow, the study is the closest thing to a direct comparison between human autologous and allogeneic stem cell types yet published. The findings have lead researchers from cardiac and other fields to question which types of stem cell therapies—self or non-self—are better for the patient. As mesenchymal stem cells are believed to differentiate into muscle, fat, bone and cartilage, they are being tried as treatments for multiple diseases—and allogeneic preparations could make this much easier across the board.

But researchers still have questions. "These cells have a tremendous potential to help patients," says Pranela Rameshwar of the New Jersey Medical School in Newark, "even though there is still a lot we don't know."

The advantages of allogeneic preparations

While a patient's own stem cells have to be prepared over the course of several months—sometimes delaying a needed transplant procedure—non-self cells can be taken from healthy people, prepared up to two years before they are used, and even kept on hand for emergencies. In a commentary accompanying the Lancet paper, Dominik Wolf and Anna Maria Wolf from the Innsbruck Medical University in Austria, say the findings do away with the need for tissue typing for mesenchymal stem cells. They even argue that a supply of these cells should be kept on hand for allcomers2.

"With autologous stem cells, there is a limited scalability," says Tim Allsopp, chief scientific officer at Stem Cell Sciences in Cambridge, UK, which prepares cells and enabling technology for stem cell applications. Only a tiny fraction of all the cells in blood or bone marrow are stem cells, and it's difficult to isolate and grow enough for patients.

In contrast, allogeneic mesenchymal stem cells can, in principle, be combined from many different individuals, making it faster and easier to grow the large numbers needed. More commonly, allogeneic preparations are based on established banks of purified cells or cell lines that have been characterized as genetically stable. In this case, sufficient quantities can be easily produced for repeat doses.

Joshua Hare at the University of Miami in Florida led the first human trial to use allogeneic stem cells to treat heart diseases. The Phase I trial was completed last year and initiated by Maryland-based stem cell company Osiris. Non-self mesenchymal stem cells were given to 53 patients through an intravenous drip within a few days of a heart attack. The hope was that the stem cells would migrate to damaged areas in the heart, prevent fibrotic scars from forming and allow functional tissue to grow back.

In the trial, 42% of the patients saw overall improvements in their condition while 11% of placebo patients also improved. A Phase II study is set to begin this summer. Osiris is also using allogeneic mesenchymal stem cells in trials for cartilage disease and ulcerative colitis.

"It defies logic and scientific dogma," says Hare, "but allogeneic MSCs [mesenchymal stem cells] appear to work very well."

Even immunotolerant cells can pose compatibility issues

While most tissue and organ transplants require heavy doses of immunosuppressive drugs to overcome the inevitable immune reaction against them, mesenchymal stem cells usually do not make enough of the cell surface antigens needed to trigger an immune response. In particular, they express little MHC class II antigen, which helps the immune system distinguish between self and non-self cells, and provokes transplant rejections. Mesenchymal stem cells also release a slew of cytokines that trick the body into tolerating the transplanted cells. In the case of GVHD, the cytokines may help the attacking donor cells make peace with the host's body.

Yet despite their tremendous potential, allogeneic mesenchymal stem cells may still pose problems, says Rameshwar. “They can lose their immunocompatibility under certain conditions,” she says. If exposed to an infection, for example, they might shed the disguises that hide them from the immune system.

In a recent study, Rameshwar and her colleagues looked at ectodermal neurons, which differentiate from mesenchymal stem cells and are thought to help regenerate tissue after spinal and brain injuries. Normally these cells have next to no MHC class II molecules, but Rameshwar was able to induce their production by exposing the mesenchymal stem cells to conditions found during an infection3.

Her team found that in normal conditions the levels of MHC II plummeted as the stem cells differentiated into neurotransmitter-producing cells. Proteins inhibiting the MHC II genes increased, consistent with the lower MHC II production. But this 'off' state was reversed by exposure of the stem cells to lymphocytes and the cytokine interferon-gamma—both of which increase during an infection.

“I don't really believe that MSCs are immunoprivileged,” says Rameshwar, “it depends on the condition.” Even though her findings specifically looked at neurons, she says they would apply to myoblasts, adipocytes and other kinds of differentiated mesenchymal stem cells.

Even Fibbe questions whether mesenchymal stem cells would be of wider benefit. Even if the stem cells express very little MHC II at the time of infusion, says Fibbe, “they may nevertheless become MHC II positive in vivo.”

Despite Fibbe's work that shows allogeneic and autologous cells act similarly in transplant patients, he has also published animal data indicating that non-self cells can spark a serious immune response. Non-self mesenchymal stem cells can trigger the rejection of allogeneic grafts delivered to irradiated immune competent mice4.

“There is a lot of back and forth on this issue,” says Rameshwar, “we just don't know enough to draw any firm conclusions.”

Could allogeneic cells display higher efficacy?

In cardiac stem cell trials, where mesenchymal stem cells are commonly used, researchers are still debating whether or not the transplanted cells survive long enough to spark an immune reaction. Most trials have so far been only marginally successful at treating disease, and there is little information to confirm whether the effects from the transplants endure—for better or worse.

If there is an immediate immune response to the mesenchymal stem cells, researchers speculate this might even serve as a benefit. Some of the improvements observed could in fact be due to a heightened immune response against transplanted cells. The subsequent inflammation could help promote useful, temporary growth in regions where cells are injected.

The problem is that because mesenchymal stem cells are believed to be immunotolerant, and most cardiac trials are still in the early stage, there has been very little immunological data collected. “For most cell types there has been no quantitative measurement of the immune response after injection or infusion in the heart,” says Christine Mummery at the Hubrecht Laboratory in Utrecht, the Netherlands.

Efforts have so far focused on demonstrating safety, and are now shifting towards proving that the treatments have a robust therapeutic effect. This has sparked great debate over whether self or non-self cells will show greater benefits.

Hare believes allogeneic cells may be the best choice for reasons beyond convenience and immune responses: unhealthy patients might provide unhealthy cells. “Diseases like hypertension, atherosclerosis and diabetes may well prove to impact the quality of autologous cells,” he says. “In these cases you may be better off with allogeneic transplants.”

Several scientists say that positive safety results from Phase I trials justify testing treatments on those with more serious conditions. Some, like Hare, believe that using sicker patients will demonstrate the efficacy of the stem cell treatment more easily, as the sickest patients have more room for improvement. But this means any underlying problems with their stem cells will come to a head, and potentially drag down efficacy results. The question is whether the sickest patients will also have the sickest autologous stem cells. If so, perhaps allogeneic stem cells will be the most advantageous.

Yet the sickest patients may also be more susceptible to any long-term immunological response invoked by allogeneic implants, says Rameshwar. “There are a lot of unknowns,” she says, “which is why the debate keeps going.”

Comparing cell sources

The next step in the mesenchymal stem cell story will probably involve comparing allogeneic and autologous cells in humans as part of a combined trial. The goal is to go beyond studies that compare allogeneic and genetically matched cells, and directly test the use of mesenchymal stem cells from patients' own bone marrow.

The comparisons also need to expand beyond the effects of these cells on GVHD. “We don't really know if the allogeneic MSCs would be rejected in other cases,” says Fibbe. As the patients he and his colleagues treated were immunosuppressed at the time of their initial transplant, it is impossible to tell if the receptivity to allogeneic cells would still be seen in patients with active immune systems. Fibbe also points out that autologous cells have yet to be systematically studied in the case of GVHD—demonstrating the need for more rigorous research on all fronts.

Hare says he hopes to compare the benefits of autologous and allogeneic cells in an upcoming cardiac trial. He also recently launched a clinical study designed to identify the biomarkers in autologous cells that predict transplant success. After taking mesenchymal stem cells from the bone marrow of 35 patients with congestive heart failure, he will search for characteristics that indicate whether the cells are healthy. Patients will then be tracked, and their healing processes compared. The National Institutes of Health provided funding for the study, called Prometheus, and the first patient underwent surgery in April.

“When all else is equal, it may be that one type of cell proves to be more helpful,” says Hare. “This is one of the most interesting and pivital issues in the stem cell field right now. If the allogeneic cells work better, it will make all the difference.”

Box 1: Parkinson's and beyond

A similar controversy over self versus non-self cells is gaining momentum among embryonic stem (ES) cell researchers, but in this case the evidence favours autologous cells. In recent work, researchers made the first successful autologous grafts to alleviate a mouse model of Parkinson's disease using ES cells5. The ES cells were derived in the usual way from mouse blastocysts made by nuclear transfer—by removing the nucleus from a mouse egg cell, and infusing it with DNA from the host. ES cells were then harvested from the blastocyst, and differentiated into dopamine-producing cells for grafting.

While making mouse ES cells through nuclear transfer is well established, such cells have never before resulted in grafts that cured a disease. “Nothing in this paper is new,” says Viviane Tabar, a neurosurgeon at the Sloan-Kettering Cancer Center in New York City, and an author of the paper, “everything has been done independently, but this is the first time we made the full story come together.”

After the ES cells differentiated into dopamine-producing neurons, they were transplanted into the brains of 24 Parkinson's mice. While some improvement in behaviour occurred in the control group that received allogeneic grafts, symptoms were completely alleviated in the mice that received autologous grafts. While the allogeneic treatments induced an immune response, the autologous grafts did not trigger the body's defence system at all.

But autologous cells for neural application in humans would be hard to come by. Human ES cells have yet to be created by nuclear transfer, and other techniques to create pluripotent stem cell lines require genetic modifications that prevent their use in therapy.

Besides Pranela Rameshwar of the New Jersey Medical School in Newark warns that autologous grafts could have a downside, as the mutations involved in disease progression will be present in self cells. “It might be safer to use an allogeneic system, despite the immunocompatibility issues,” she says.