Significant improvements in donor stem cell trans­planta­tion have been made in recent years. As a result, donor trans­planta­tion – a procedure during which a patient receives stem cells from a healthy donor – has become safer and typically more successful for patients.

The majority of patients undergoing donor stem cell trans­planta­tion receive stem cells from unrelated donors. In these cases, a close match between donor and recipient tissue types is known to play an important important role in trans­plant outcomes.

A group of German researchers last fall published results of a large retrospective study about tissue typing in donor stem cell trans­planta­tion for a wide range of blood and bone marrow cancers.

Results of the study confirm that closely matching donor and recipient tissue types typically improves the outcome of donor stem cell trans­plants.

The study also shows, however, that tissue type mis­matches found using very sensitive tissue-typing methods -- methods known as "high-resolution" typing -- can have just as significant an impact on trans­plant outcomes as mis­matches found using "low-resolution" methods, which are designed to find broader types of mis­matches.

These findings from the German study suggest that high-resolution tissue typing is important for ensuring the best possible match between stem cell donors and recipients.

Because The Beacon in the past has not reported extensively on tissue typing for donor stem cell trans­planta­tion, this article about the German study includes a substantial amount of background information.  It therefore should be a useful introduction to the subject for Beacon readers who are not familiar with the topic.

Background

Donor (allogeneic) stem cell trans­planta­tion has the potential to be a curative therapy for multiple myeloma and other cancers of the blood and bone marrow.

During this procedure, the patient first receives chemotherapy and/or radiation treatment.  This treatment is intended to destroy some or all of the patient’s diseased stem cells.  However, it also usually destroys many of the patient's healthy cells.

After the chemotherapy and/or radiation, the patient is infused with stem cells from a healthy donor.

A donor stem cell trans­plant is different from an autologous stem cell trans­plant, in which a patient’s own stem cells are collected prior to the chemotherapy, and then re-infused into the patient after the chemo­ther­apy.

The Role Of Tissue Typing

Not anyone can donate stem cells to someone wanting to undergo a donor stem cell trans­plant.  To avoid serious complications after the trans­plant, it is important that the donor and recipient have "tissue types" that are closely matched.

A person's tissue type is determined by proteins in their body called human leukocyte antigens (HLA). These proteins help the immune system distinguish between its own cells and those that are foreign. HLA proteins are found on most cells in the body, and the specific HLA proteins on a person’s cells determines their tissue type.

A close tissue-type match between stem cell donor and trans­plant recipient is essential for two reasons.

First, in the case of a mis­match, the recipient’s immune system could recognize the trans­planted stem cells as foreign and attack them, resulting in graft rejection. However, this scenario is unlikely because the recipient’s immune system is largely destroyed by the chemotherapy or radiation treatment that precedes trans­planta­tion.

Secondly, as the mis­matched donor stem cells take up residence in the recipient’s bone marrow and produce immune cells, these immune cells could recognize the donor’s tissues as foreign and attack them. This phenomenon, called graft-versus-host disease (GVHD), is a common complication after trans­planta­tion and can sometimes be life-threatening.

Different Approaches To HLA-Matching

Up to six broad categories of HLA proteins - A, B, C, DP, DQ and DR – are commonly taken into account for tissue matching. Everyone has genes necessary for making the proteins in each of these categories. How­ever, the specific type of HLA-A protein one person can produce, for example, may not be the same as that of another person.

Genes in different parts of chromosome 6 determine the precise HLA proteins a person can produce. Since chromosomes come in pairs – one chromosome inherited from a person's mother, the other from the father – this means there is a one in four chance that two people with the same parents will produce exactly the same set of HLA proteins.

In general, studies have shown that about 30 percent of patients seeking a stem cell donor are able to find an HLA-matched donor within their immediate family.  The remaining 70 percent, however, receive stem cell trans­plants from unrelated HLA-matched donors.

The matching of potential donors with potential trans­plant recipients is done based on testing of the genes that determine a person's HLA type.  If the goal is to get a complete match involving, for example, five of the HLA proteins, this would require each of the 10 genes (five pairs) to be the same for the donor and recipient.

If that were to occur, the match would be described as a 10 out of 10 (10/10) match.  If one gene out of the 10 did not match, it would be a 9/10 match.

The technology for typing, and therefore matching, the genes involved in producing HLA proteins has ad­vanced substantially in the last 25 years.  Initially, the technology was relatively crude, or "low resolution."  Even though the genes that produce a given HLA protein can be very diverse – with literally thousands of var­i­a­tions in the case of some HLA genes – the early technology could only dis­tinguish between broad groups of these many different gene variants.

The current "high resolution" approaches, however, can identify and type the HLA genes very precisely.

Low- Versus High-Resolution HLA Matching: An Analogy

A simple analogy can be helpful for better understanding the difference between low-resolution and high-resolution HLA typing.

In particular, think of HLA-related genes as being different from one another in a way similar to how cars differ from one another.

One way to "type" cars would be to do it in a very broad way -- for example, using just a car's manufacturer (Ford, Toyota, Volkswagen, etc.) for the classification.  Typing cars in this way would be a "low resolution" approach.

Another way to "type" cars would be to do it in a very detailed way, using not only their manufacturer (Ford), but also their model (Focus), year (2011), color (red), and trim level ("SE").  Using all of this information to "type" cars would be a "high resolution" approach.

High-resolution approaches to HLA matching donors and trans­plant recipients are therefore better at en­sur­ing that HLA matches are, in fact, truly matches.

In the medical literature, mis­matches based on low-resolution comparisons are known as "antigen mis­matches."  Mismatches based on high-resolution comparisons, on the other hand, are known as "allele mis­matches."

Goals Of The German Study

In the German study about HLA matching that was published last fall, the researchers set out to do several things.

First, they wanted to assess the impact on stem cell trans­plant outcomes of tissue type mis­matches de­ter­mined using both low- and high-resolution testing methods.

One might think that a mis­match based on low-resolution methods – that is, an antigen mis­match – would have a greater impact on outcomes than a high-resolution ("allele") mis­match. The low-resolution approach is already a relatively crude comparison, so, if it registers a mis­match, the impact on trans­plant outcomes could be greater than with a high-resolution mis­match.  But is that really true in practice?

Second, the investigators wanted to determine the impact of different types of HLA mis­matches, and whether multiple mis­matches have a much greater impact on outcomes than just a single mis­match.

Study Design

The German researchers retrospectively analyzed data from 2,646 patients who had donor stem cell trans­plants performed across 28 centers in Germany between 1997 and 2010.

All patients received trans­plants from an unrelated donor as a part of treatment for cancer of the blood or bone marrow. About 8 percent of patients included in the analysis had multiple myeloma.

The median patient age was 51 years, with an age range of 18 years to 75 years. Almost all patients were of Caucasian ethnicity.

About 40 percent of patients had early stage disease, 35 percent had intermediate stage disease, and 25 percent had advanced disease.

Nearly 64 percent of patients received a conditioning regimen with total body irradiation, high-dose cyclo­phos­pha­mide (Cytoxan) and/or high-dose busulfan (Busulfex). Such regimens, which are called "myeloablative", aim to destroy all cells in the recipient’s bone marrow.

The other 36 percent of patients received a reduced-intensity conditioning regimen with lower doses of chemotherapy or radiation, which does not destroy all cells in the bone marrow. This approach is generally used when conventional, myelo­ablative conditioning is considered too risky due to factors such as advanced patient age.

The majority of patients (87 percent) received stem cells that were collected from the donor’s blood (as op­posed to the bone marrow).  All trans­plants were what is known as t-cell re­plete -- that is, not t-cell depleted.

Tissues from donors and recipients were high-resolution typed for HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA–DQB1. (DRB1 and DQB1 refer to portions of the DR and DQ HLA proteins.)

Patients who had received their trans­plants during or after May, 2005 already had been matched using high-resolution methods.  For patients trans­planted prior to that date, high-resolution testing had been done for some of the HLA proteins.  However, stored patient tissue samples were used to do high-resolution testing for the remaining HLA proteins covered in the study.

Median follow-up time after trans­planta­tion was two years.

Study Results

The authors conducted statistical analyses of their data using information across all patients, regardless of disease type, and across the entire time period for which they had data.

Their analyses did take into account, however, a variety of factors that could affect trans­plant outcomes in addition to the HLA types of the donor and recipient.  For example, the analyses adjusted for the fact that trans­plant outcomes can depend on a patient's specific disease, disease stage, and age.

Matching And Overall Survival

As expected, the authors found that mis­matches in most of the HLA types resulted in a higher risk of death, and thus shorter overall survival, after trans­planta­tion.

In particular, mis­matches in HLA-A, HLA-B, HLA-C or HLA-DRB1, whether at the antigen-level or allele-level, negatively affected survival.

The highest risk of death was seen with mis­matches (either antigen or allele) in HLA-A, HLA-B, and HLA-DRB1.  However, HLA-DRB1 mis­matches seem mainly to have had an effect if the mis­match was in addition to another mis­match.

The impact of a single mis­match could be substantial.  Patients in the study who had an 8/8 perfect match using HLA-A, B, C, and DRB1 had a median overall survival of about 28 months.  In comparison, patients who had a single mis­match, whether it was an antigen or allele mis­match, had a median overall survival more than 55 percent lower (about 12 months).

Since antigen-level mis­matches and allele-level mis­matches often had a similar impact on survival, the researchers noted that their results support broader use of high-resolution HLA typing, which detects allele-level mis­matches, for donor selection.

The investigators also found that the harmful effects of HLA mis­matches were additive. This means that two mis­matches lead to lower overall survival compared to one mis­match, and that the risk of death increased with the number of mis­matches.

Matching, Other Factors, And Outcomes

Disease-free survival was not significantly affected by single mis­matches in most of the HLA genes examined. Only antigen-level mis­matches in HLA-C were associated with shorter disease-free survival. However, a higher number of mis­matches of any type significantly shortened disease-free survival.

Mismatches, either antigen or allele, in HLA-A, HLA-B, HLA-C and HLA-DRB1 were also associated with increased risk trans­plant-related mortality (death due primarily to complications from the stem cell trans­plant).

Neither the odds of disease relapse, nor the ability of the donor's stem cells to engraft quickly in the trans­plant recipient, were affected by HLA mis­matches.

Besides HLA mis­matches, the researcher's analysis found – as would be expected – that advanced disease stage and patient age were associated with adverse outcomes.

On the other hand, the source of stem cells (donor’s blood versus donor’s bone marrow) did not affect patient survival.

Patients who received trans­plants before 2004, and trans­plants involving international (not German-born) donors, were also found to have poorer outcomes. However, the researchers noted that a matched international donor would still be preferred over a single mis­matched national donor.

Comparisons With Results Of Previous Research

A key difference in the findings of the German study and previous research is the impact of HLA-C mis­matches on survival.

Previous research has suggested that HLA-C mis­matches, particularly HLA-C antigen mis­matches, have a particularly significant impact on survival.

In the German study, however, HLA-A and HLA-B mis­matches had a larger impact on survival than HLA-C mis­matches.  In line with previous research, however, the German researchers did find that HLA-C antigen mis­matches have a greater impact on survival than HLA-C allele mis­matches.

The difference in the German findings may be due to an ethnicity effect -- that is, HLA-C may not matter as much in German trans­plant recipients as it does in trans­plant recipients of other ethnicities.

There is also the possibility, though, that specifics of the German HLA matching system prior to 2005 may play a role in the researchers' findings. In Germany before 2005, HLA-C was not routinely checked when doing donor-recipient matching.  Thus, there are many more HLA-C mis­matches in the German study than in previous studies, and this could contribute to the difference in findings.

For further information, please see the study by Fürst D et al., “High resolution HLA-matching in hematopoietic stem cell trans­planta­tion: a retrospective collaborative analysis,” Blood, October 31, 2013, vol. 122 no. 18 3220-3229 (full text).