Skip to main content
PMC home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2020 Jan 6;47(1):13–23. doi: 10.1111/iji.12471

Immunogenetics in stem cell donor registry work: The DKMS example (Part 1)

Alexander H Schmidt 1,2,3,✉, Jürgen Sauter 1, Daniel M Baier 1, Jessica Daiss 1, Andreas Keller 1, Anja Klussmeier 2, Thilo Mengling 1, Gabi Rall 1, Tobias Riethmüller 1, Gerhard Schöfl 2, Ute V Solloch 1, Tigran Torosian 4, David Means 5, Helen Kelly 6, Latha Jagannathan 7,8, Patrick Paul 7, Anette S Giani 9, Sabine Hildebrand 1, Stephan Schumacher 1, Jan Markert 1, Monika Füssel 2, Jan A Hofmann 1, Thomas Schäfer 2, Julia Pingel 1, Vinzenz Lange 2, Johannes Schetelig 3,10
PMCID: PMC7003907  PMID: 31903698

Abstract

Currently, stem cell donor registries include more than 35 million potential donors worldwide to provide HLA‐matched stem cell products for patients in need of an unrelated donor transplant. DKMS is a leading stem cell donor registry with more than 9 million donors from Germany, Poland, the United States, the United Kingdom, India and Chile. DKMS donors have donated hematopoietic stem cells more than 80,000 times. Many aspects of donor registry work are closely related to topics from immunogenetics or population genetics. In this two‐part review article, we describe, analyse and discuss these areas of donor registry work by using the example of DKMS. Part 1 of the review gives a general overview on DKMS and includes typical donor registry activities with special focus on the HLA system: high‐throughput HLA typing of potential stem cell donors, HLA haplotype frequencies and resulting matching probabilities, and donor file optimization with regard to HLA diversity.

Keywords: DKMS, donor registry, HLA, unrelated hematopoietic stem cell transplantation

1. INTRODUCTION

DKMS was founded in May 1991, growing out of the search for a stem cell donor for Mechtild Harf. Since then, the DKMS donor file has grown to 9,601,738 donors from six countries. DKMS donors have donated stem cells 80,847 times for patients from 57 various countries (cut‐off date: 30 September 2019).

In this two‐part review article, we describe, analyse and discuss aspects of DKMS's work that are related to immunogenetics or population genetics. Part 1 of the review includes, apart from a general overview on DKMS, typical donor registry activities that are closely related to the HLA system: high‐throughput HLA typing of potential stem cell donors, HLA haplotype frequencies and resulting matching probabilities, and donor file optimization with regard to HLA diversity. The second part of the review (Schmidt et al., in press), on the other hand, focuses on non‐HLA parameters and DKMS activities beyond the scope of standard donor registry work. Topics addressed include donor typing beyond the classical HLA loci, impact of non‐HLA parameters on donation probabilities, identification of novel HLA, KIR and MIC alleles, activities of the Collaborative Biobank and pharmacogenetics in the donor registry context.

We do not cover central donor registry issues as donor recruitment or donor safety comprehensively, but only related to immunogenetics or population genetics. Besides, we focus on adult donors and do therefore not include the activities of the DKMS Cord Blood Bank.

2. DKMS OVERVIEW: DONORS AND DONATIONS

Figure 1 shows the quantitative development of the DKMS donor registry since 1991. At the end of 2018, it included 8.7 million donors, accounting for 24.2% of the global WMDA donor file of 35.9 million donors excluding cord blood units (WMDA, 2019).

Figure 1.

Figure 1

Open in a new tab

Number of registered DKMS donors by country and year. Germany: orange; United States: grey; Poland: yellow; United Kingdom: light blue; Chile: red; India: dark blue. Cut‐off date: 30 September 2019 [Colour figure can be viewed at http://wileyonlinelibrary.com]

The regional and ethnic compositions of the DKMS UK and DKMS Germany donor files are shown in Figures 2 and 3, respectively. Corresponding figures for the other DKMS country organizations are included as Supplementary Material (Figures S1–S4). The ethnic grouping applied in the various countries differs due to administrative and societal reasons. At DKMS Poland, no ethnic information is gathered. Donor numbers by ethnicities in the various national DKMS donor registries are also included in the Supplementary Material (Tables S1–S6). For these tables, ethnic categorization was converted into seven population groups: AFA (African/African American), API (Asian/Pacific Islander), EURO (European), HIS (Hispanic), MENA (Middle East/North Coast of Africa), NAM (Native American) and UNK (unknown, multiple ancestries or other). Due to the nature of the information available, all donors from DKMS Poland were regarded as EURO and all donors from DKMSBMST (Bangalore Medical Services Trust) Foundation India as API.

Figure 2.

Figure 2

Open in a new tab

Postcode areas in the United Kingdom. For each area, the number of DKMS donors is represented by the size of the corresponding pie chart. Slices of the pie charts indicate the area‐specific ethnic composition of the DKMS donor file. The share of donors of non‐European descent ranges from 0.3% (postcode area KW = Kirkwall) to 85.3% (HA = Harrow). Colouring of the areas indicates the ratio between DKMS donors and the total population. The proportion of registered donors ranges from 0.3% (postcode area HS = Outer Hebrides) to 3.2% (WR = Worcester). The map was created with QGIS 2.12 software (QGIS Development Team, 2015). Shapefile © 2015 by Open Door Logistics (http://www.opendoorlogistics.com) [Colour figure can be viewed at http://wileyonlinelibrary.com]

Figure 3.

Figure 3

Open in a new tab

Two‐digit postcode areas in Germany. For each region, the number of DKMS donors is represented by the size of the corresponding pie chart. Slices of the pie charts indicate the area‐specific ethnic composition of the DKMS donor file. The share of donors of non‐German descent ranges from 1.2% (postcode area 08 = Plauen) to 22.1% (70 = Stuttgart). Colouring of the areas indicates the ratio between DKMS donors and the total population. The proportion of registered donors ranges from 3.8% (postcode area 06 = Halle (Saale)) to 15.0% (56 = Koblenz and 49 = Osnabrück). The map was created with QGIS 2.12 software (QGIS Development Team, 2015). Shapefile © 2015 by Open Door Logistics (http://www.opendoorlogistics.com) [Colour figure can be viewed at http://wileyonlinelibrary.com]

Considerable regional differences in the registered DKMS donors to total population ratio exist in all countries. In the United Kingdom, for example, the share of registered donors ranges from 0.3% to 3.2%. Similarly, the proportion of registered donors ranges from 3.8% to 15.0% in Germany. Though DKMS was founded after the German reunification, the territory of the former German Democratic Republic (East Germany) is considerably underrepresented in the DKMS donor registry. It is possible to focus donor recruitment efforts geographically in order to exploit regional HLA haplotype frequency differences (see paragraph Donor file optimization). However, the inter‐region differences displayed in Figures 2 and 3 occur for other reasons including unequal activities of local supporter groups, spotty distribution of large patient‐oriented donor drives, recruitment efforts of non‐DKMS donor registries and socio‐economic differences between regions.

Figures 2 and 3 also show significant differences in the share of donors of non‐European and non‐German descent, respectively, ranging from 0.3% to 85.3% for the United Kingdom and from 1.2% to 22.1% for Germany. (Donors of unknown descent were excluded from these analyses.) Rural areas generally show a lower share of donors of non‐European (Figure 2) and non‐German (Figure 3) descent. Registered donors of non‐German descent are also underrepresented in East Germany. These differences correlate with the composition of the general population.

Figure 4 shows the development of donations (bone marrow and peripheral blood stem cells alone; donor lymphocytes and cord blood are not included) from DKMS donors over time. So far, DKMS donors have donated stem cells 80,847 times (cut‐off date: 30 September 2019). In 2018, there were 7,488 donations from DKMS donors; 5,592 (74.7%) cross‐border. Patients receiving transplants from DKMS donors were from the United States (1,891, 25.3%), Germany (1,571, 21.0%), France (491, 6.6%), the United Kingdom (432, 5.8%), Italy (348, 4.6%) and 46 other countries (2,755, 36.8%). The DKMS share of all donations worldwide was 39.5% while the share of all cross‐border donations was 59.7%. Given the DKMS share of the global donor pool of 24.2% at the end of 2018, potential donors registered with DKMS donated stem cells disproportionally often. A detailed analysis of the reasons for this fact is beyond the scope of this review. However, contributing factors may include donor age and gender distributions (Schmidt et al., in press, paragraph Donation probabilities), donor typing level (see paragraph Donor typing: Classical HLA loci; Schmidt et al., in press, paragraphs Donor typing: Beyond the classical HLA loci and Donation probabilities), donor availability, process speed and quality, and pricing of stem cell products.

Figure 4.

Figure 4

Open in a new tab

Number of stem cell donations by DKMS donors by country and year. Germany: orange; United States: grey; Poland: yellow; United Kingdom: light blue; Chile: red; India: dark blue. Cut‐off date: 30 September 2019 [Colour figure can be viewed at http://wileyonlinelibrary.com]

3. DONOR TYPING: CLASSICAL HLA LOCI

The rise of next‐generation sequencing (NGS) technologies has dramatically changed donor registry HLA typing over the last few years. At DKMS Life Science Lab (LSL), NGS entered routine for high‐throughput typing of new donors in early 2013. Since then, more than 6.1 million newly registered DKMS donors have been typed for the six “classical” HLA genes (A, B, C, DRB1, DQB1 and DPB1) at high resolution with an amplicon‐based approach using Illumina devices (Lange et al., 2014; Schöfl et al., 2017).

High‐throughput HLA typing using NGS technology showed substantial advantages over Sanger sequencing (Sanger, Nicklen, & Coulson, 1977): First, it enabled substantial typing cost reductions. Figure 5 shows full costs per sample for NGS‐based compared with Sanger‐based HLA typing at DKMS LSL. These “real‐life” data are based on official financial reports, that is, they are not controlled for factors influencing full costs per sample as, for example, typing volume, salary increases or inflation. However, the dramatic NGS‐related cost reduction by 85.5% from 64.4 to 9.3 cost units per HLA typing is very clear in spite of these methodological limitations. Second, we were able to reduce ambiguities by applying NGS: we obtained high‐resolution results for >99.9% of all HLA loci typed with NGS in the last 12 months compared to 95.0% with Sanger sequencing (Lange et al., 2014). Third, the typing error rate for NGS‐based HLA typing was 3.5 times lower than the Sanger error rate. The low locus‐wise error rate of 0.024% including sample switches (less than one error in 4,000 loci) even challenges the necessity of a standard confirmatory typing step before donor work‐up (Baier et al., 2019). Fourth, the use of NGS facilitated the identification and whole‐gene sequencing of novel HLA alleles (Albrecht et al., 2017; Schmidt et al., in press, paragraph Identification of novel alleles). Fifth, the quality and cost efficiency of NGS‐based HLA typing allowed the realization of scientific projects at reasonable effort. For example, we could show that the HLA‐DPB1 expression marker rs9277534 can be predicted from standard genotyping of DPB1 exons 2 and 3 with 100% accuracy (Schöne et al., 2018).

Figure 5.

Figure 5

Open in a new tab

Typing costs at DKMS LSL by typing method. Red: HLA (6 loci at high resolution); grey: CCR5Δ32; green: ABO, Rh; purple: KIR (allele groups); blue: MICA/MICB; black: HLA‐E; yellow: CMV (from swabs) [Colour figure can be viewed at http://wileyonlinelibrary.com]

It is currently under debate whether donor selection should be based on donor‐recipient matching of whole HLA genes (Mayor et al., 2019; Vazirabad et al., 2019) or whether HLA typing of gene regions coding for the antigen recognition domain (ARD) plus consideration of frequent null alleles is adequate for stem cell donor selection (Hurley & Ng, 2019; Hurley et al., 2019). While we have used NGS‐related cost savings to extend the DKMS standard typing profile beyond HLA (Schmidt et al., in press), we have not included whole‐gene HLA typing in the standard typing profile so far. This decision was the result of cost‐benefit considerations: at DKMS LSL, the introduction of whole‐gene HLA typing would have more than doubled typing costs for the extended standard typing profile shown in Figure 5. Besides, it is currently doubtful if these substantial additional costs will be outweighed by respective patient benefits due to optimized donor selection by whole‐gene donor‐patient HLA matching. With increasing evidence for potential benefits of whole‐gene matching and/or decreasing costs for whole‐gene HLA typing, possibly through the introduction of new sequencing devices (Lang et al., 2018), this decision may be re‐assessed.

In contrast, “complete” HLA typing—at the very least the ARD of the 4 genes HLA‐A, ‐B, ‐C and ‐DRB1 (Lee et al., 2007)—at donor recruitment is undoubtedly beneficial for patients as incompletely typed stem cell donors may not be identified in stem cell donor searches (Sauter, Solloch, Giani, Hofmann, & Schmidt, 2016; Schmidt, Solloch, Baier, et al., 2011a). Besides, incomplete donor HLA typing at recruitment does not adequately utilize donor commitment as incompletely typed donors advance to donation less often than donors who are better typed (Dubois et al., 2011; Müller, Feldmann, Bochtler, Morsch, & Schmidt, 2012; Nicoloso, Kürsteiner, Bussmann, Marbacher, & Tiercy, 2019; Schmidt, Stahr, Baier, Ehninger, & Rutt, 2006). The current donor selection guidelines by NMDP and CIBMTR recommend donor and patient HLA typing for at least 5 HLA genes (A, B, C, DRB1 and DPB1) at high resolution (Dehn et al., 2019).

However, many donor registries that have grown over decades still have donors with insufficient HLA typing profiles on their files. DKMS Germany, for example, currently (cut‐off date: 30 September 2019) still includes 102,895 donors (1.6% of the total DKMS Germany donor file) who are typed only for two loci (HLA‐A and ‐B). Five years ago (30 September 2014), the respective number was 283,436 (7.9% of the then donor file). Apart from usual donor attrition, for example by reaching the age limit, this decrease was the result of considerable re‐typing efforts focused on young male donors. As a result, only 794 of the 102,895 poorly typed donors are males below 40 years.

The general question, if limited donor registry resources should rather be used for donor recruitment or for re‐typing of insufficiently typed donors who are already on the donor file, is of considerable practical relevance. Unfortunately, it is hardly possible to determine the optimal trade‐off between recruitment and re‐typing quantitatively. However, it seems unwise to stop donor recruitment completely because newly recruited donors rejuvenate the donor file (Schmidt, Biesinger, Baier, Harf, & Rutt, 2008; Schmidt et al., in press, paragraph Donation probabilities). On the other hand, a complete waiver of re‐typing efforts may not be appropriate, either, as poorly typed donors will become increasingly “invisible” for donor searches, and their donation probabilities will decrease over time.

4. HAPLOTYPE FREQUENCIES AND MATCHING PROBABILITIES

In unrelated stem cell donation, the knowledge of population‐specific HLA haplotype frequencies (HF) is desirable for two reasons: First, HF can be used to estimate the probability that the next patient of a defined population finds a matching donor in a donor registry of given size and ethnic composition (Beatty, Mori, & Milford, 1995; Bergstrom, Garratt, & Sheehan‐Connor, 2009; Müller, Ehninger, & Goldmann, 2003). Obviously, such matching probabilities (MP) are highly relevant for strategic donor registry planning. Second, population‐specific HF make it possible to estimate the chances of incompletely typed donors to be full HLA matches in specific donor searches. Today, such estimations are carried out automatically for large donor files by state‐of‐the‐art donor search algorithms as OptiMatch (Bochtler, Beth, Eberhard, & Müller, 2008), HapLogic (Dehn et al., 2016) or Hap‐E Search (Pingel et al., 2012).

Population‐specific HF can be derived from unphased HLA genotype data of registered stem cell donors via an expectation‐maximization (EM) algorithm (Excoffier & Slatkin, 1995). We developed the Hapl‐o‐Mat software for HF estimation from large data sets with heterogeneous resolution, typing ambiguities and missing loci, that is from typical donor registry data (Sauter, Schäfer, & Schmidt, 2018; Schäfer, Schmidt, & Sauter, 2017). Hapl‐o‐Mat is freely available at https://github.com/DKMS/Hapl-o-Mat.

Table 1 gives an overview on published HF and MP estimations based on DKMS donor data. For this article, we added estimates for further populations (also included in Table 1). The respective HF are given in the Supplementary Material (Tables S7–S15). Cumulative frequencies of the 1,000 most frequent haplotypes range from 0.572 (Turkish minority in Germany, sample size n = 100,000) to 0.839 (Scottish, n = 20,000; see Figure S5). Corresponding MP for scenarios in which donor and patient populations are identical are shown in Figures 6 and 7. In good accordance with HF estimation results, MP are especially high for Scottish donors and patients (p = .537 at donor registry size n = 100,000) and especially low for Turkish donors and patients (p = .122; n = 100,000). The more realistic scenario with combined donor populations, thus simulating global donor searches, has been discussed in the literature (Bergstrom et al., 2009; Schmidt, Sauter, Pingel, & Ehninger, 2014).

Table 1.

Overview on published haplotype frequency (HF) and/or matching probability (MP) estimations based on DKMS donors

Country Population Sample size Source Remarks/Specific focus
Germany German 8,862 Schmidt, Baier, et al. (2009a)

  • 3‐ and 4‐locus HLA haplotypes at low resolution (LR) and high resolution (HR)

  • Impact of matching requirements on matching probabilities (MP)
Germany German 319,009 Schmidt et al. (2010)

  • 3‐locus HLA haplotypes at LR

  • Regional HLA differences
Germany German 1,099,735 Eberhard et al. (2010)

  • 3‐locus HLA haplotypes (class I at LR, class II at HR)

  • Impact of selective HLA‐DRB1 typing (based on class I typing results) on HF estimations
Germany German 370,856 Sauter et al. (2016)

  • 5‐locus HLA haplotypes at HR

  • Impact of completeness of donor HLA typing on search success
Germany German 100,000 (20,000) Figures 7, 9, S5; Table S7

  • 5‐locus HLA haplotypes at HR
Germany Turkish 9,086 Schmidt, Solloch, et al. (2009b)

  • 3‐locus HLA haplotypes at LR

  • Impact of ethnic diversity recruitment efforts
Germany Turkish 100,000 Figure 7, S5; Table S8

  • 5‐locus HLA haplotypes at HR
Germany 17 minority populations 1,028–33,083 Pingel et al. (2013)

  • 4‐ and 5‐locus HLA haplotypes at HR

  • Impact of population‐specific HF on individual donor searches
Germany, Poland, United States 21 populations 1,028–33,083 Schmidt et al. (2014)

  • 4‐locus HLA haplotypes at HR

  • Includes also non‐DKMS donors (from NMDP)

  • Optimization of global donor recruitment efforts

  • Sample size effects
Poland Polish 20,653 Schmidt, Solloch, Pingel, et al. (2011b)

  • 4‐locus HLA haplotypes at HR

  • Validation of implementation of EM algorithm

  • Impact of donor recruitment in Poland on MP for Polish patients
Poland Polish 123,749 Schmidt et al. (2013)

  • 4‐locus HLA haplotypes at HR

  • Regional HLA differences

  • Sample size effects
Poland Polish 100,000 Figure 7, S5; Table S9 5‐locus HLA haplotypes at HR
United Kingdom English/Scottish/Welsh 100,000 (20,000) Figures 7, 9, S5; Table S10 5‐locus HLA haplotypes at HR
United Kingdom English 20,000 Figure 8, S5; Table S11 5‐locus HLA haplotypes at HR
United Kingdom Indian 20,000 Figure 8, S5; Table S12 5‐locus HLA haplotypes at HR
United Kingdom Scottish 20,000 Figure 8, S5; Table S13 5‐locus HLA haplotypes at HR
United Kingdom Welsh 20,000 Figure 8, S5; Table S14 5‐locus HLA haplotypes at HR
United States European 100,000 Figure 7, S5; Table S15 5‐locus HLA haplotypes at HR
Open in a new tab

Figure 6.

Figure 6

Open in a new tab

Matching probabilities by donor registry size for various populations (all sample sizes n = 100,000). Red: Germany (country of recruitment), German (population); green: Germany, Turkish; orange: Poland, Polish; blue: United Kingdom, English/Scottish/Welsh; purple: United States, European [Colour figure can be viewed at http://wileyonlinelibrary.com]

Figure 7.

Figure 7

Open in a new tab

Matching probabilities by donor registry size for various populations (all samples from DKMS UK donors, all sample sizes n = 20,000). Red: English; orange: Indian; blue: Scottish; green: Welsh [Colour figure can be viewed at http://wileyonlinelibrary.com]

One major methodological difficulty of combining HF and MP estimations lies in dealing with small HF. Figure 8 shows plateaus in the HF distribution curves that are corresponding to small integer multiples of f = 1/(2n)—the frequency of a haplotype that occurs exactly once in the sample—as the HF are approaching that value (Pappas, Tomich, Garnier, Marry, & Gourraud, 2015). These HF plateaus that are more prominent for smaller sample sizes are artefacts and affect subsequent MP estimations. Furthermore, estimated HF smaller than 1/(2n) are generally of limited informative value as they correspond to less than one occurrence of the respective haplotype in the underlying sample. However, all haplotypes with estimated f ≥ 1/(2n) sum up to, for example, only 93.0% (90.6%) in the sample with n = 100,000 (n = 20,000) German donors. Therefore, it seems to be not appropriate to ignore all haplotypes below that threshold for MP estimations. For the estimations shown in Figures 6 and 7, we started with the most frequent haplotypes, considered all haplotypes up to a cumulated HF of 99.5% and then normalized all HF above this threshold (Schmidt et al., 2014).

Figure 8.

Figure 8

Open in a new tab

Haplotype frequencies (HF) for two populations and two sample sizes. Red: Germany (country of recruitment), German (population); blue: United Kingdom, English/Scottish/Welsh. Solid: sample size n = 100,000; dashed: n = 20,000. Horizontal lines indicate HF that correspond to one (black), two (dark grey) or three (light grey) haplotype occurrence(s) in the samples with size n = 100,000 (solid) or n = 20,000 (dashed) [Colour figure can be viewed at http://wileyonlinelibrary.com]

5. DONOR FILE OPTIMIZATION

It is well known that the marginal benefit of donor recruitment efforts decreases considerably with increasing donor file size as new donors often carry HLA genotypes that are already on the file. Therefore, the large number of registered DKMS donors raises questions regarding the efficiency of ongoing donor recruitment efforts. DKMS and other donor registries have developed and applied several strategies aimed at maximizing donor recruitment efficiency with regard to a stronger MP increase, especially for patients from population groups that have been underserved so far. These strategies include:

  1. Ethnic diversity donor recruitment. It is a common understanding that, if the population living in the geographical area where a donor registry recruits donors is not homogeneous, it is important to target donor recruitment efforts not only at the majority population but also at the minority population(s) (Confer, 2001; Fingrut, 2015; Heinemann et al., 2019; Johansen, Schneider, McCaffree, Woods, & Council on Science and Public Health, American Medical Association, 2008; Schmidt, Solloch, et al., 2009b). This approach increases the efficiency of donor recruitment as the genetic diversity of a donor registry grows more quickly. It is an imperative for fairness to aspire to equal chances to find HLA‐matched donors for patients from all populations, although this target may be difficult to achieve due to different levels of intra‐population diversity or donor availability (Gragert et al., 2014; Maiers, Gragert, & Klitz, 2007).

    At DKMS Germany, ethnic diversity recruitment activities are focused specifically at individuals of Turkish descent as they are the largest ethnic minority in Germany (Schmidt, Solloch, et al., 2009b). Programme elements include specific minority donor drives, use of marketing materials in Turkish language, commitment by prominent representatives of the Turkish community in Germany, co‐operation with media preferably used by the Turkish minority and native speaker support by DKMS employees at recruitment and through the various process steps prior to stem cell donation. Currently, 211,613 self‐reported donors of Turkish descent are registered with DKMS in Germany. Donors of Turkish descent from DKMS Germany have donated stem cells 1,476 times so far; 318 times for patients in Germany, 222 times for patients in Turkey and 936 times for patients elsewhere (cut‐off date: 30 September 2019).

  2. Regional focus of donor recruitment activities. Several groups have analysed regional HLA frequency differences within defined countries and their relevance to strategic donor registry planning (Buhler, Nunes, Nicoloso, Tiercy, & Sanchez‐Mazas, 2012; Lonjou, Clayton, Cambon‐Thomsen, & Raffoux, 1995; Rendine et al., 1998). In many Central European countries as, for example, Germany (Schmidt et al., 2010) or Poland (Schmidt et al., 2013) genetic differences within the majority population are relatively small. As a result, the benefits of setting a geographical donor recruitment focus based on regional HLA differences of the majority population are limited, especially compared with benefits of ethnic diversity donor recruitment. Therefore, it is promising to combine approaches (a) and (b) by setting the recruitment focus on regions where many individuals from minority populations live.

    The situation is different in large, multi‐ethnic countries as, for example, India (Maiers et al., 2014; Schmidt, 2014). So far, donor recruitment of DKMS BMST Foundation India has a strong focus on Bangalore and the state of Karnataka (see Figure S4). However, choosing this starting point for our Indian donor recruitment efforts resulted from organizational, socio‐economic and feasibility reasons rather than from population genetics considerations. Due to the large genetic variation within this huge country, an ambitious donor recruitment strategy in India will have to include several foci in various regions of the country.

  3. Selective recruitment of donors with rare HLA genotypes. Unfortunately, surrogate parameters for individuals with rare HLA genotypes do not exist. Otherwise, donor recruitment would be much more efficient as one could focus on donors with rare HLA genotypes that are not yet on the file multiple times. This idea triggered several Ancestry projects (formerly named Roots projects) at DKMS (Schmidt et al., 2007). These projects include the following elements: identification of registered donors with rare HLA genotypes (exact definitions of “rare” vary between projects), communication with these donors including the appeal to inform their relatives that they would probably be especially interesting stem cell donors, and recruitment of interested relatives. We showed that donors who had been recruited via Ancestry projects indeed had rarer HLA genotypes than other donors (Schmidt et al., 2007). DKMS has recruited 20,400 donors in various project runs in Germany since 1997, and 317 of these donors have donated stem cells until 30 September 2018. For such initiatives, it must be ensured that the use of HLA genotype information of registered donors for donor recruitment purposes is in accordance with donor informed consent and national regulations.

  4. Extension of the geographical region of donor recruitment. For an organization that is responsible for patients from a defined geographical area, it seems appropriate to focus donor recruitment efforts on that specific area if all populations living there are addressed adequately. The statutes of DKMS, on the other hand, give no preference to supporting patients from specific countries or regions. Against this background, it became obvious that ongoing donor recruitment in Germany alone would not be optimal. For example, we showed substantial patient benefits from same‐population donor recruitment in calculations based on a model with 21 populations under consideration of cross‐border stem cell exchange (Schmidt et al., 2014). This result was in accordance with earlier analyses regarding donor recruitment in Germany and Poland (Schmidt, Solloch, Pingel, et al., 2011b). We demonstrated that intense donor recruitment efforts in Poland would increase matching probabilities for Polish patients substantially in spite of the large German donor file and the geographical and genetic proximity of both countries. The desire to increase HLA diversity among registered stem cell donors and thus to improve matching probabilities for patients in need of a stem cell transplant on a global level is the main driver for initiating donor recruitment efforts in a “new” country or region. However, some non‐HLA‐related criteria have to be considered in the corresponding decision process as well, including country‐specific legal framework, openness of government authorities, availability of potential partners or supporters, current local stem cell transplantation activities, access to stem cell transplantation for patients from the respective country and general socio‐economic parameters. The current international DKMS donor registry with donors from six countries results from the basic wish to increase donor file HLA diversity and efficiency and the application of these additional criteria. We plan to further expand our donor recruitment activities in the near future, especially to countries with mainly non‐European populations.

Supporting information

 

Click here for additional data file. (3.7MB, docx)

 

Click here for additional data file. (415KB, xlsx)

Schmidt AH, Sauter J, Baier DM, et al. Immunogenetics in stem cell donor registry work: The DKMS example (Part 1). Int J Immunogenet. 2020;47:13–23. 10.1111/iji.12471

REFERENCES

  1. Albrecht, V. , Zweiniger, C. , Surendranath, V. , Lang, K. , Schöfl, G. , Dahl, A. , … Schmidt, A. H. (2017). Dual redundant sequencing strategy: Full‐length gene characterisation of 1056 novel and confirmatory HLA alleles. HLA, 90(2), 79–87. 10.1111/tan.13057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baier, D. M. , Hofmann, J. A. , Fischer, H. , Rall, G. , Stolze, J. , Ruhner, K. , … Schmidt, A. H. (2019). Very low error rates of NGS‐based HLA typing at stem cell donor recruitment question the need for a standard confirmatory typing step before donor work‐up. Bone Marrow Transplantation, 54(6), 928–930. 10.1038/s41409-018-0411-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beatty, P. G. , Mori, M. , & Milford, E. (1995). Impact of racial genetic polymorphism on the probability of finding an HLA‐matched donor. Transplantation, 60(8), 778–783. 10.1097/00007890-199510270-00003 [DOI] [PubMed] [Google Scholar]
  4. Bergstrom, T. C. , Garratt, R. J. , & Sheehan‐Connor, D. (2009). One chance in a million: Altruism and the bone marrow registry. The American Economic Review, 99(4), 1309–1334. 10.1257/aer.99.4.1309 [DOI] [PubMed] [Google Scholar]
  5. Bochtler, W. , Beth, M. , Eberhard, H. P. , & Müller, C. R. (2008). OptiMatch – a universally configurable HLA matching framework. HLA, 71(4), 321 10.1111/j.1399-0039.2008.01015.x [DOI] [Google Scholar]
  6. Buhler, S. , Nunes, J. M. , Nicoloso, G. , Tiercy, J. M. , & Sanchez‐Mazas, A. (2012). The heterogeneous HLA genetic makeup of the Swiss population. PLoS ONE, 7(7), e41400 10.1371/journal.pone.0041400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Confer, D. L. (2001). The National Marrow Donor Program: Meeting the needs of the medically underserved. Cancer, 91(S1), 274–278. [DOI] [PubMed] [Google Scholar]
  8. Dehn, J. , Setterholm, M. , Buck, K. , Kempenich, J. , Beduhn, B. , Gragert, L. , … Maiers, M. (2016). HapLogic: A predictive human leukocyte antigen–matching algorithm to enhance rapid identification of the optimal unrelated hematopoietic stem cell sources for transplantation. Biology of Blood and Marrow Transplantation, 22(11), 2038–2046. 10.1016/j.bbmt.2016.07.022 [DOI] [PubMed] [Google Scholar]
  9. Dehn, J. , Spellman, S. , Hurley, C. K. , Shaw, B. E. , Barker, J. N. , Burns, L. J. , … Pidala, J. (2019). Selection of unrelated donors and cord blood units for hematopoietic cell transplantation: Guidelines from NMDP/CIBMTR. Blood, 134(12), 924–934. 10.1182/blood.2019001212 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dubois, V. , Giannoli, C. , Balère, M. L. , Rey, S. , Raffoux, C. , & Rigal, D. (2011). Does high‐resolution donor typing of HLA‐C or other loci upon registration confer advantages to patients? Human Immunology, 72(11), 1033–1038. 10.1016/j.humimm.2011.08.007 [DOI] [PubMed] [Google Scholar]
  11. Eberhard, H. P. , Feldmann, U. , Bochtler, W. , Baier, D. , Rutt, C. , Schmidt, A. H. , & Müller, C. R. (2010). Estimating unbiased haplotype frequencies from stem cell donor samples typed at heterogeneous resolutions: A practical study based on over 1 million German donors. Tissue Antigens, 76(5), 352–361. 10.1111/j.1399-0039.2010.01518.x [DOI] [PubMed] [Google Scholar]
  12. Excoffier, L. , & Slatkin, M. (1995). Maximum‐likelihood estimation of molecular haplotype frequencies in a diploid population. Molecular Biology and Evolution, 12(5), 921–927. 10.1093/oxfordjournals.molbev.a040269 [DOI] [PubMed] [Google Scholar]
  13. Fingrut, W. (2015). The need for ethnically diverse stem cell donors. UBC Medical Journal, 7(1), 44–47. [Google Scholar]
  14. Gragert, L. , Eapen, M. , Williams, E. , Freeman, J. , Spellman, S. , Baitty, R. , … Maiers, M. (2014). HLA match likelihoods for hematopoietic stem‐cell grafts in the U.S. registry. The New England Journal of Medicine, 371(4), 339–348. 10.1056/NEJMsa1311707 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Heinemann, F. M. , Wagner, B. , Riebschlaeger, S. , Heinold, A. , Baumgart, C. , Kordelas, L. , … Horn, P. A. (2019). Recruiting refugees and migrants as new potential blood stem cell donors: Status update of the BluStar.NRW project in North Rhine Westphalia. HLA, 94(S1), 11 10.1111/tan.13634 [DOI] [Google Scholar]
  16. Hurley, C. K. , & Ng, J. (2019). Continue to focus clinical decision‐making on the antigen recognition domain for the present. Human Immunology, 80(1), 79–84. 10.1016/j.humimm.2018.04.010 [DOI] [PubMed] [Google Scholar]
  17. Hurley, C. K. , Spellman, S. , Dehn, J. , Barker, J. N. , Devine, S. , Fernandez‐Vina, M. , … Pidala, J. (2019). Regarding “Recipients receiving better HLA‐matched hematopoietic cell transplantation grafts, uncovered by a novel HLA typing method, have superior survival: A retrospective study”. Biology of Blood and Marrow Transplantation, 25(8), e268–e269. 10.1016/j.bbmt.2019.05.026 [DOI] [PubMed] [Google Scholar]
  18. Johansen, K. A. , Schneider, J. F. , McCaffree, M. A. , Woods, G. L. (2008). Efforts of the United States' National Marrow Donor Program and Registry to improve utilization and representation of minority donors. Transfusion Medicine, 18(4), 250–259. 10.1111/j.1365-3148.2008.00865.x [DOI] [PubMed] [Google Scholar]
  19. Lang, K. , Surendranath, V. , Quenzel, P. , Schöfl, G. , Schmidt, A. H. , & Lange, V. (2018). Full‐length HLA class I genotyping with the MinION nanopore sequencer. Methods in Molecular Biology, 1802, 155–162. 10.1007/978-1-4939-8546-3_10 [DOI] [PubMed] [Google Scholar]
  20. Lange, V. , Böhme, I. , Hofmann, J. , Lang, K. , Sauter, J. , Schöne, B. , … Schmidt, A. H. (2014). Cost‐efficient high‐throughput HLA typing by MiSeq amplicon sequencing. BMC Genomics, 15, 63 10.1186/1471-2164-15-63 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lee, S. J. , Klein, J. , Haagenson, M. , Baxter‐Lowe, L. A. , Confer, D. L. , Eapen, M. , … Anasetti, C. (2007). High‐resolution donor‐recipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood, 110(13), 4576–4583. 10.1182/blood-2007-06-097386 [DOI] [PubMed] [Google Scholar]
  22. Lonjou, C. , Clayton, J. , Cambon‐Thomsen, A. , & Raffoux, C. (1995). HLA‐A, ‐B, ‐DR haplotype frequencies in France – implications for recruitment of potential bone marrow donors. Transplantation, 60(4), 375–383. 10.1097/00007890-199508270-00013 [DOI] [PubMed] [Google Scholar]
  23. Maiers, M. , Gragert, L. , & Klitz, W. (2007). High‐resolution HLA alleles and haplotypes in the United States population. Human Immunology, 68(9), 779–788. 10.1016/j.humimm.2007.04.005 [DOI] [PubMed] [Google Scholar]
  24. Maiers, M. , Halagan, M. , Joshi, S. , Ballal, H. S. , Jagannatthan, L. , Damodar, S. , … Weisdorf, D. (2014). HLA match likelihoods for Indian patients seeking unrelated donor transplantation grafts: A population‐based study. The Lancet Haematology, 1(2), e57–e63. 10.1016/S2352-3026(14)70021-3 [DOI] [PubMed] [Google Scholar]
  25. Mayor, N. P. , Hayhurst, J. D. , Turner, T. R. , Szydlo, R. M. , Shaw, B. E. , Bultitude, W. P. , … Marsh, S. G. E. (2019). Recipients receiving better HLA‐matched hematopoietic cell transplantation grafts, uncovered by a novel HLA typing method, have superior survival: A retrospective study. Biology of Blood and Marrow Transplantation, 25(3), 443–450. 10.1016/j.bbmt.2018.12.768 [DOI] [PubMed] [Google Scholar]
  26. Müller, C. R. , Ehninger, G. , & Goldmann, S. F. (2003). Gene and haplotype frequencies for the loci HLA‐A, HLA‐B, and HLA‐DR based on over 13,000 German blood donors. Human Immunology, 64(1), 137–151. 10.1016/s0198-8859(02)00706-1 [DOI] [PubMed] [Google Scholar]
  27. Müller, C. R. , Feldmann, U. , Bochtler, W. , Morsch, S. , & Schmidt, A. (2012). The effect of age, gender and typing resolution on the probability of stem cell donation. Human Immunology, 73(Supplement 1), 121 10.1016/j.humimm.2012.07.240 [DOI] [Google Scholar]
  28. Nicoloso, G. , Kürsteiner, O. , Bussmann, F. , Marbacher, M. , & Tiercy, J. M. (2019). A study of selected hematopoietic stem cell donors provided by an intermediate size registry. European Journal of Haematology, 103(4), 426–432. 10.1111/ejh.13307 [DOI] [PubMed] [Google Scholar]
  29. Pappas, D. J. , Tomich, A. , Garnier, F. , Marry, E. , & Gourraud, P. A. (2015). Comparison of high‐resolution human leukocyte antigen haplotype frequencies in different ethnic groups: Consequences of sampling fluctuation and haplotype frequency distribution tail truncation. Human Immunology, 76(5), 374–380. 10.1016/j.humimm.2015.01.029 [DOI] [PubMed] [Google Scholar]
  30. Pingel, J. , Hofmann, J. , Baier, D. M. , Solloch, U. V. , Grathwohl, A. , Ehninger, U. , … Ehninger, G. (2012). Hap‐E Search: Haplotype‐enhanced search – Implementation and validation of a haplotype frequency‐based search algorithm. Tissue Antigens, 79(6), 552 10.1111/j.1399-0039.2012.01877.x [DOI] [Google Scholar]
  31. Pingel, J. , Solloch, U. V. , Hofmann, J. A. , Lange, V. , Ehninger, G. , & Schmidt, A. H. (2013). High‐resolution HLA haplotype frequencies of stem cell donors in Germany with foreign parentage: How can they be used to improve unrelated donor searches? Human Immunology, 74(3), 330–340. 10.1016/j.humimm.2012.10.029 [DOI] [PubMed] [Google Scholar]
  32. QGIS Development Team (2015). QGIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org
  33. Rendine, S. , Borelli, I. , Barbanti, M. , Sacchi, N. , Roggero, S. , & Curtoni, E. S. (1998). HLA polymorphisms in Italian bone marrow donors: A regional analysis. Tissue Antigens, 52(2), 135–146. 10.1111/j.1399-0039.1998.tb02277.x [DOI] [PubMed] [Google Scholar]
  34. Sanger, F. , Nicklen, S. , & Coulson, A. R. (1977). DNA sequencing with chain‐terminating inhibitors. Proceedings of the National Academy of Science of the USA, 74(12), 5463–5467. 10.1073/pnas.74.12.5463 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sauter, J. , Schäfer, C. , & Schmidt, A. H. (2018). HLA haplotype frequency estimation from real‐life data with the Hapl‐o‐Mat software. Methods in Molecular Biology, 1802, 275–284. 10.1007/978-1-4939-8546-3_19 [DOI] [PubMed] [Google Scholar]
  36. Sauter, J. , Solloch, U. V. , Giani, A. S. , Hofmann, J. A. , & Schmidt, A. H. (2016). Simulation shows that HLA‐matched stem cell donors can remain unidentified in donor searches. Scientific Reports, 6, 21149 10.1038/srep21149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Schäfer, C. , Schmidt, A. H. , & Sauter, J. (2017). Hapl‐o‐Mat: Open‐source software for HLA haplotype frequency estimation from ambiguous and heterogeneous data. BMC Bioinformatics, 18(1), 284 10.1186/s12859-017-1692-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schmidt, A. H. (2014). Potential for increased stem‐cell donor recruitment in India. The Lancet Haematology, 1(2), e48–49. 10.1016/S2352-3026(14)70028-6 [DOI] [PubMed] [Google Scholar]
  39. Schmidt, A. H. , Baier, D. , Solloch, U. V. , Stahr, A. , Cereb, N. , Wassmuth, R. , … Rutt, C. (2009a). Estimation of high‐resolution HLA‐A, ‐B, ‐C, ‐DRB1 allele and haplotype frequencies based on 8862 German stem cell donors and implications for strategic donor registry planning. Human Immunology, 70(11), 895–902. 10.1016/j.humimm.2009.08.006 [DOI] [PubMed] [Google Scholar]
  40. Schmidt, A. H. , Biesinger, L. , Baier, D. , Harf, P. , & Rutt, C. (2008). Aging of registered stem cell donors: Implications for donor recruitment. Bone Marrow Transplantation, 41(7), 605–612. 10.1038/sj.bmt.1705950 [DOI] [PubMed] [Google Scholar]
  41. Schmidt, A. H. , Sauter, J. , Baier, D. , Daiss, J. , Keller, A. , Klussmeier, A. , … Schetelig, J. (in press). Immunogenetics in stem cell donor registry work: The DKMS example (Part 2). International Journal of Immunogenetics. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Schmidt, A. H. , Sauter, J. , Pingel, J. , & Ehninger, G. (2014). Toward an optimal global stem cell donor recruitment strategy. PLoS ONE, 9(1), e86605 10.1371/journal.pone.0086605 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schmidt, A. H. , Solloch, U. V. , Baier, D. , Grathwohl, A. , Hofmann, J. , Pingel, J. , … Ehninger, G. (2011a). Support of unrelated donor searches by donor center‐initiated HLA typing of potentially matching donors. PLoS ONE, 6(5), e20268 10.1371/journal.pone.0020268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Schmidt, A. H. , Solloch, U. V. , Baier, D. , Stahr, A. , Wassmuth, R. , Ehninger, G. , & Rutt, C. (2010). Regional differences in HLA antigen and haplotype frequency distributions in Germany and their relevance to the optimization of hematopoietic stem cell donor recruitment. Tissue Antigens, 76(5), 362–379. 10.1111/j.1399-0039.2010.01520.x [DOI] [PubMed] [Google Scholar]
  45. Schmidt, A. H. , Solloch, U. V. , Baier, D. , Yazici, B. , Özcan, M. , Stahr, A. , … Rutt, C. (2009b). Criteria for initiation and evaluation of minority donor programs and application to the example of donors of Turkish descent in Germany. Bone Marrow Transplantation, 44(7), 405–412. 10.1038/bmt.2009.55 [DOI] [PubMed] [Google Scholar]
  46. Schmidt, A. H. , Solloch, U. V. , Pingel, J. , Baier, D. , Böhme, I. , Dubicka, K. , … Ehninger, G. (2011b). High‐resolution human leukocyte antigen allele and haplotype frequencies of the Polish population based on 20,653 stem cell donors. Human Immunology, 72(7), 558–565. 10.1016/j.humimm.2011.03.010 [DOI] [PubMed] [Google Scholar]
  47. Schmidt, A. H. , Solloch, U. V. , Pingel, J. , Sauter, J. , Böhme, I. , Cereb, N. , … Ehninger, G. (2013). Regional HLA differences in Poland and their effect on stem cell donor registry planning. PLoS ONE, 8(9), e73835 10.1371/journal.pone.0073835 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Schmidt, A. H. , Stahr, A. , Baier, D. , Ehninger, G. , & Rutt, C. (2006). Efficiency of HLA‐A, ‐B, ‐C, ‐DRB1 high‐resolution typings of newly recruited potential stem cell donors. Blood, 108(11), 5415 10.1182/blood.V108.11.5415.5415 [DOI] [Google Scholar]
  49. Schmidt, A. H. , Stahr, A. , Baier, D. , Schumacher, S. , Ehninger, G. , & Rutt, C. (2007). Selective recruitment of stem cell donors with rare human leukocyte antigen phenotypes. Bone Marrow Transplantation, 40(9), 823–830. 10.1038/sj.bmt.1705832 [DOI] [PubMed] [Google Scholar]
  50. Schöfl, G. , Lang, K. , Quenzel, P. , Böhme, I. , Sauter, J. , Hofmann, J. A. , … Lange, V. (2017). 2.7 million samples genotyped for HLA by next generation sequencing: Lessons learned. BMC Genomics, 18(1), 161 10.1186/s12864-017-3575-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Schöne, B. , Bergmann, S. , Lang, K. , Wagner, I. , Schmidt, A. H. , Petersdorf, E. W. , & Lange, V. (2018). Predicting an HLA‐DPB1 expression marker based on standard DPB1 genotyping: Linkage analysis of over 32,000 samples. Human Immunology, 79(1), 20–27. 10.1016/j.humimm.2017.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Vazirabad, I. , Chhabra, S. , Nytes, J. , Mehra, V. , Narra, R. K. , Szabo, A. , … Anderson, M. W. (2019). Direct HLA genetic comparisons identify highly matched unrelated donor‐recipient pairs with improved transplantation outcome. Biology of Blood and Marrow Transplantation, 25(5), 921–931. 10.1016/j.bbmt.2018.12.006 [DOI] [PubMed] [Google Scholar]
  53. WMDA (2019). WMDA Global Trend Report – Summary 2018. https://wmda.info/wp-content/uploads/2019/08/17072019-GTR-Graphs-Summary-2018.pdf. Accessed on September 13, 2019.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

 

Click here for additional data file. (3.7MB, docx)

 

Click here for additional data file. (415KB, xlsx)

Articles from International Journal of Immunogenetics are provided here courtesy of Wiley

RESOURCES