ORIGINAL ARTICLE
Impact of Ethnicity on Donor Search Results for Children Requiring Stem Cell Transplantation Julia DiLabio, MSc,* John Doyle, MD,w Sarah Alexander, MD,z Sumit Gupta, MD, PhD,z and Angela Punnett, MDz Summary: Hematopoietic stem cell transplantation (HSCT) can be curative for children with various malignant and nonmalignant conditions. Access to suitable unrelated living donors (ULDs) or unrelated cord blood (UCB) may be limited for certain ethnic backgrounds. We therefore determined the impact of ethnicity upon donor availability in a pediatric cohort referred for allogeneic HSCT to a single Canadian transplant center. Among 252 eligible patients, 58 (23.0%) had suitable family donors. Of 161 patients with combined ULD and UCB searches, 78 (48.4%) had a suitable ULD, whereas 143 (88.8%) had suitable UCB. The probability of finding a suitable ULD differed significantly by ethnicity (P = 0.007). Non-white patients were significantly less likely to have suitable ULDs (odds ratio [OR] 0.35; 95% confidence interval [CI], 0.17-0.69; P = 0.003) compared with white patients but were equally likely to have suitable UCB (OR 1.02; 95% CI, 0.36-2.89; P = 0.97). Although ethnic disparities exist in pediatric patients’ access to ULD for HSCT, they are narrowed by the availability of international UCB registries. These findings, however, also highlight the importance of continued recruitment of individuals of non-white ethnicities to donor registries. Key Words: stem cell transplantation, bone marrow, umbilical cord blood, donors, ethnicity
(J Pediatr Hematol Oncol 2015;37:e154–e157)
A
llogeneic hematopoietic stem cell transplantation (HSCT) is an integral part of treatment for children diagnosed with some types of hematologic malignancies, hemoglobinopathies, immunodeficiencies, and congenital metabolic disorders. Stem cells can be obtained from the bone marrow or peripheral blood of family members (typically siblings) or unrelated donors, or from umbilical cord blood sources. Finding a suitable donor involves matching human leukocyte antigens (HLA) between patient and donor; higher degrees of HLA-mismatch have been linked to inferior survival.1 Umbilical cord blood stem cells require less stringent HLA-matching criteria for successful transplants compared with bone marrow or peripheral blood sources2; recent data suggests that these sources have
Received for publication July 28, 2014; accepted September 13, 2014. From the *Faculty of Medicine, University of Toronto; zDivision of Hematology/Oncology, SickKids Hospital, Toronto; and wSection of Pediatric Hematology/Oncology, Winnipeg Children’s Hospital, Winnipeg, Canada. S.G. and A.P. are co-senior authors. The authors declare no conflict of interest. Reprints: Angela Punnett, MD, Division of Hematology/Oncology, SickKids Hospital, 555 University Avenue, Toronto M5G 1X8, ON, Canada (e-mail: [email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.jpho-online.com. Copyright r 2014 Wolters Kluwer Health, Inc. All rights reserved.
e154 | www.jpho-online.com
similar outcomes and overall survival in the pediatric setting.3 The availability of appropriately HLA-matched stem cells has been shown in adult studies to vary according to patient ethnicity.4–6 It is generally more difficult to find suitably matched donor sources for non-white and multiethnic individuals, predominantly because of their underrepresentation among donors.7–10 The relationship between patient ethnicity and access to suitable donors has not been thoroughly investigated in a pediatric population, with ethnicity or availability of suitable donors comprising only a minor component of existing studies.11–13 We, therefore, conducted a single institution study to describe the impact of patient ethnicity on the availability of suitable unrelated living donors (ULDs) (ie, bone marrow and peripheral blood sources) and on the availability of suitable unrelated cord blood (UCB) for a contemporaneous pediatric cohort to inform family counseling and future advocacy efforts.
MATERIALS AND METHODS Study Design We conducted a retrospective cohort study. Eligible patients included all those assessed by SickKids Hospital (Toronto, Canada) for possible primary allogeneic HSCT between January 2009 and August 2012. A parent or substitute decision maker provided informed consent for a formal donor search in all cases. We excluded cases lacking self-reported ethnicity or other essential data (patient demographics, availability of an HLA-matched related donor, or combined ULD and UCB search results).
Data Source Data were collected from electronic patient charts and from an HSCT database established in 2009. Variables collected included patient age, disease at presentation (leukemia vs. other malignancy vs. nonmalignant disorder), referral center, and availability of an HLA-matched related donor. Ethnicity is a difficult concept to define objectively. Similar to previous reports,4–10 we used self-reported ethnicity and ancestry, as declared by parents on HSCT referral forms, to reclassify data according to both broad ethnic groups (white, Asian, black, mixed) and smaller ethnic subgroups (Caucasian, Middle Eastern, South Asian, East Asian, black, mixed). Hispanic patients were classified among the white group and Caucasian subgroup. This study received appropriate institutional ethics review and was approved by the Quality Management Department at SickKids Hospital.
Donor Search Analyses Patients and potential related donors, ULDs, and UCB underwent low-resolution and high-resolution HLA typing at 10 alleles (HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DQB1). Sibling donors were deemed J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
suitable if low-resolution matching occurred at HLA-A and HLA-B with high-resolution matching at HLA-DRB1 (6/6 alleles), provided that there were no inconsistencies in allele assignment from family studies. Otherwise, related donors were deemed suitable if high-resolution matching occurred at HLA-A, HLA-B, HLA-C, and HLA-DRB1 (8/8 alleles). Patients lacking an HLA-matched related donor underwent searches simultaneously for both ULDs (through the Canadian OneMatch and international registries for bone marrow and peripheral blood stem cell sources) and UCB (through international umbilical cord blood registries). Suitable ULDs were defined and recorded by HSCT clinicians in the database according to the following guidelines: high-resolution matching at HLA-A, HLA-B, HLA-C, and HLA-DRB1 at 7/8 or 8/8 alleles (with matching at HLADQB1 in the former case). Suitable UCB was defined as low-resolution matching at HLA-A and HLA-B with highresolution matching at HLA-DRB1 and adequate total nucleated cell dose (Z5 107/kg/unit, Z3 107/kg/unit, and Z1.5 107/kg/unit for 4/6, 5/6, and 6/6 matched cords, respectively). ULDs and UCB that did not meet infectious disease screening requirements or were listed as unavailable were also deemed unsuitable.
Statistical Analysis Demographic characteristics among included and excluded patients were compared using the Student’s t test, w2, or the Fisher exact test as appropriate. Predictors of the presence of an HLA-matched related donor were determined using univariate and multivariable logistic regression. The percentage of patients with suitable stem cell sources (ULD, UCB, or any source) was compared across ethnic groups using the Fisher exact test. The odds of identifying suitable ULDs, UCB, or any stem cell source were compared between white and non-white patients by logistic regression analysis. On the basis of the underrepresentation of Aboriginal peoples and black individuals within international ULD registries, exploratory analyses focused on these populations were specified a priori. Statistical significance was defined as P < 0.05. All statistical analyses were conducted using SAS-PC v.9.2 (Cary, NC).
RESULTS During the study period, 343 children were referred to SickKids Hospital for consideration of primary allogeneic HSCT. Reasons for exclusion are shown in Figure 1; the final cohort consisted of 252 patients (73.5%) ranging from 0 to 20 years of age. In comparison with the study cohort, excluded patients lacking ethnicity data were more likely to be referred to SickKids from an external medical center (43/ 51, 84.3% vs. 72/252, 28.6%; P < 0.0001) and were more likely to be transplanted (33/51, 64.7% vs. 117/252, 46.4%; P = 0.02). The median age and distribution of diagnoses were not statistically different between the 2 groups. Patient demographics are illustrated in Table 1. When classified by broad ethnic groups or subgroups, the majority of patients were white (152/252, 60.3%) or Caucasian (134/252, 52.3%), respectively. Only 58 of 252 patients (23.0%) had an HLA-matched donor within their immediate family. As anticipated, larger family size was associated with higher odds of having an HLA-matched related donor (odds ratio [OR] 2.26; 95% confidence interval [CI], 1.62-3.16; P < 0.0001) in univariate analysis. Middle Eastern children had a greater Copyright
r
Ethnicity and Donors for Stem Cell Transplantation
FIGURE 1. Flow diagram depicting the exclusion of ineligible cases from the study cohort. *Insufficient data other than patient ethnicity (results of HLA typing of family members, ULD, or UCB). UCB indicates unrelated cord blood; ULD, unrelated living donor.
probability of having an HLA-matched related donor (OR 4.46; 95% CI, 1.57-12.64; P = 0.005) compared with Caucasian children. Both variables retained significance in multivariable analysis (see Supplemental Table 2, http:// links.lww.com/JPHO/A81). Of the study cohort, 161 patients (63.9%) underwent combined ULD and UCB searches. A suitable ULD was identified for approximately half of these patients (78/161, 48.4%), of which the majority were fully HLA-matched (65/78, 83.3%). The majority of patients (143/161, 88.8%) had a suitable single UCB unit identified, of which the minority were fully HLA-matched (61/143, 42.7%). Combining ULD and UCB search results revealed that the majority of patients (154/161, 95.6%) had at least one
TABLE 1. Demographic Characteristics of Children Considered for Hematopoietic Stem Cell Transplantation During the Study Period
Characteristic Age (y) Diagnosis Nonmalignancy Leukemia Other malignancy Referral center SickKids Ontario Other province Ethnic group White Asian Black Mixed Ethnic subgroup Caucasian Middle Eastern South Asian East Asian Black Mixed Family match Present Absent
2014 Wolters Kluwer Health, Inc. All rights reserved.
n (%)
Median (IQR) 7 (2-12)
131 (52.0) 113 (44.8) 8 (3.2) 180 (71.4) 56 (22.2) 16 (6.4) 152 61 22 17
(60.3) (24.2) (8.7) (6.8)
134 17 35 26 22 18
(52.3) (6.8) (13.9) (11.5) (8.7) (7.1)
58 (23.0) 194 (77.0)
IQR indicates interquartile range.
www.jpho-online.com |
e155
J Pediatr Hematol Oncol
DiLabio et al
n (%) Ethnicity
Suitable ULD
Suitable UCB
61 (57.0) 13 (34.2) 0 (0.0) 4 (36.4) 0.007
95 33 5 10
(88.8) (86.8) (100.0) (90.9) 0.97
Any Suitable 102 (95.3) 37 (97.4) 5 (100.0) 10 (90.9) 0.71
n indicates number; UCB, unrelated cord blood; ULD, unrelated living donor. *The Fisher exact test.
potential source of stem cells, leaving only 7 of 161 patients (4.4%) without a suitable donor option. The success of finding a suitable ULD was statistically different between white, Asian, black, and multiethnic groups (P = 0.007), as shown in Table 2. The probability of identifying suitable UCB, however, did not vary significantly between ethnic groups (P = 0.97), with the majority of patients having at least 1 suitable option. Combining ULD and UCB search results yielded a suitable stem cell source for >90% of patients, irrespective of their ethnic background (P = 0.71). Collapsing ethnic groups into white and non-white patients revealed that the latter were significantly less likely to have a suitable ULD (OR 0.35; 95% CI, 0.17-0.69; P = 0.003), as shown in Table 3. However, both groups had a similar likelihood of having suitable UCB (OR 1.02; 95% CI, 0.36-2.89; P = 0.97). The likelihood of identifying a suitable stem cell source of any type was also statistically similar between the 2 groups (OR 1.27; 95% CI, 0.24-6.8; P = 0.78). Exploratory subgroup analyses revealed that the success rate of finding suitable stem cell sources for Aboriginal children was comparable to the overall success rate for the study population. Of the 9 Aboriginal and Aboriginal/ white children who underwent unrelated donor searches, 4 (44.4%) had a suitable ULD and 8 (88.8%) had suitable UCB, such that only 1 patient (11.1%) had no suitable stem cell source identified. No suitable ULD was identified for any of the black children (n = 5), although each had at least 1 suitable UCB source.
TABLE 3. Comparison of Combined ULD and UCB Search Results Among White and Non-white Patients
Variable Suitable ULD White Non-white Suitable UCB White Non-white Any suitable White Non-white
OR (95% CI)
P*
Ref. 0.35 (0.17-0.69)
— 0.003
Ref. 1.02 (0.36-2.89)
— 0.97
Ref. 1.27 (0.24-6.80)
— 0.78
CI indicates confidence interval; OR, odds ratio; Ref., reference; UCB, unrelated cord blood; ULD, unrelated living donor. *Univariate logistic regression analysis.
e156 | www.jpho-online.com
Volume 37, Number 3, April 2015
DISCUSSION
TABLE 2. Proportion of Children for Whom Suitable Donors Were Identified, Stratified by Ethnicity
White Asian Black Mixed P*
We have identified ethnic disparities in the success rates of finding suitable ULDs for pediatric patients in a contemporaneous ethnically diverse population. Non-white patients, and particularly black children, had relatively limited access to suitable ULDs compared with white patients. The ability to search international UCB registries increased the number of potential stem cell options for all patients, irrespective of ethnic background, such that ethnicity did not impact on the overall likelihood of finding a suitable stem cell source. An HLA-matched related donor was identified for only 23% of patients, which is lower than the typical likelihood of 25% to 40% based on the average North American family size.4,5,12 This likelihood may vary by ethnicity; an HLA-identical related donor was found for only 7.5% of black children in 1 study.12 Middle Eastern children in our cohort were more likely to have an HLAidentical donor within their family than white children. This finding is consistent with a study involving a pediatric cohort in Israel showing that the majority of Arabic patients (62%) had an HLA-identical family member, a finding that the authors attributed in part to a more homogenous gene pool in their patient population.13 Consistent with previous studies, our results suggest that ULD registries serve a predominantly white patient population, with non-white patients having a significantly lower probability of finding a suitable donor.4,6–10 The finding that black children had particularly poor access to suitable ULDs is consistent with studies showing that the National Marrow Donor Program has struggled to find suitable ULDs for African American patients, due both to their underrepresentation within donor registries and to the inherent genetic HLA variability within this ethnic group.6–10 The finding that a greater proportion of children had suitable UCB identified (89%) compared with a suitable ULD (48%) is secondary to the less stringent HLA matching criteria required for UCB to be deemed suitable.2 The ability to search within UCB registries, therefore, increases the availability of suitable stem cells and decreases ethnic disparities. Our findings are supported by those of Barker et al4 demonstrating that suitable UCB was identified for the majority of patients of European (96%), Asian (100%), African (80%), or mixed non-European (91%) backgrounds using a similar definition of suitability. Our findings demonstrate the importance of international UCB registries in improving access to stem cells to children of diverse ethnicities. We anticipate that the Canadian UCB bank will have a positive impact in extending access to stem cells to patients of non-white ethnic backgrounds in a similar manner as international UCB registries, provided that emphasis is placed on the recruitment of ethnically diverse registrants during the development phase, as recommended elsewhere.14,15 The high cost of obtaining UCB from international registries, the ability to collect otherwise unused cords from ethnically diverse mothers, and the lack of strong endorsement from pediatric HSCT physicians for private cord blood banking in the absence of an identified recipient provide further justification for a Canadian UCB registry.16 Nevertheless, our findings also underscore the importance of maintaining an ethnically diverse national stem cell donor registry and recruiting black donors to help fill the gap in availability. Targeted donor drives and education programs have been shown to aid in recruiting new African
Copyright
r
2014 Wolters Kluwer Health, Inc. All rights reserved.
J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
American donors and could be extended to other ethnicities.17 Employing recruitment strategies targeting the emotional response of potential donors may also be of benefit,18 especially as donation-related ambivalence is an issue across all ethnic groups.19 Several study limitations deserve mention. This was a retrospective single institution study and the findings may not be generalizable to a broader North American population. The exclusion of 91/343 cases (26.5%) because of lack of patient ethnicity or other pertinent data limited the size of the study population. Given the retrospective nature of this study, this data was not obtainable, and assumptions regarding patients’ ethnicity according to surname were deemed inappropriate. Referrals from external medical centers were more likely to lack ethnicity data; excluded patients were therefore more likely to be referred from such centers. This may also explain why excluded patients were more likely to be transplanted, as referrals from within our institution may have been more likely to include relative (as opposed to absolute) indications for HSCT. Although differences between included and excluded cases were not felt to bias the findings in any manner, such bias cannot be ruled out. Secondly, to maintain adequate statistical power, several different ethnic groups were grouped as nonwhite, given the limited numbers of particular non-white backgrounds. The benefit of having access to the self-reported ethnicity and national origins of patients’ parents was circumscribed as a result. Thirdly, our definition of suitability implied equivalence between fully and partially matched stem cells and did not take into account the degree of HLA matching among different ethnic groups. Finally, differential priority may have been placed on identifying a suitable ULD or UCB among patients based on factors such as disease or transplant urgency. Despite these limitations, focusing on the donor identification process rather than the actual transplant source allowed important findings to be generated, irrespective of the transplant status of patients. The findings may impact treatment planning and will inform the future counseling of families on the probability of finding a suitable unrelated donor or cord for their child. A final study strength was the use of parents’ self-reported ethnicity and national origins, which has been cited as an accurate representation of patient ancestry.20,21 In summary, this study, in providing the first analysis of the availability of suitable ULDs and UCB for a pediatric cohort requiring allogeneic HSCT in North America, has revealed ethnic disparities in the availability of suitable ULDs that are partially mitigated by the availability of UCB. The findings strongly support both the development of an ethnically diverse national public cord blood bank in Canada and the targeted recruitment of non-white individuals to stem cell donor registries to improve pediatric patients’ access to hematopoietic stem cells for transplantation. Future studies should assess the impact of ethnicity upon the degree of HLA matching of potential or transplanted stem cell sources, as well as on clinical outcomes.
ACKNOWLEDGMENTS The authors thank Elizabeth Davies for assistance with the HSCT database and Marie Pinard for support in project development.
Copyright
r
Ethnicity and Donors for Stem Cell Transplantation
REFERENCES 1. Lee SJ, Klein J, Haagenson M, et al. High-resolution donorrecipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood. 2007;110:4576–4583. 2. Eapen M, Rubinstein P, Zhang MJ, et al. Outcomes of transplantation of unrelated donor umbilical cord blood and bone marrow in children with acute leukaemia: a comparison study. Lancet. 2007;369:1947–1954. 3. Wang J, Zhan P, Ouyang J, et al. Unrelated donor umbilical cord blood transplantation versus unrelated donor bone marrow transplantation in adult and pediatric patients: a meta-analysis. Leuk Res. 2010;34:1018–1022. 4. Barker JN, Byam CE, Kernan NA, et al. Availability of cord blood extends allogeneic hematopoietic stem cell transplant access to racial and ethnic minorities. Biol Blood Marrow Transplant. 2010;16:1541–1548. 5. Dehn J, Arora M, Spellman S, et al. Unrelated donor hematopoietic cell transplantation: factors associated with a better HLA match. Biol Blood Marrow Transplant. 2008; 14:1334–1340. 6. Yancey AK, Coppo P, Kawanishi Y. Progress in availability of donors of color: the National Marrow Donor Program. Transplant Proc. 1997;29:3760–3765. 7. Pidala J, Kim J, Schell M, et al. Race/ethnicity affects the probability of finding an HLA-A, -B, -C and -DRB1 allelematched unrelated donor and likelihood of subsequent transplant utilization. Bone Marrow Transplant. 2013;48:346–350. 8. Joshua TV, Rizzo JD, Zhang MJ, et al. Access to hematopoietic stem cell transplantation: effect of race and sex. Cancer. 2010;116:3469–3476. 9. Dew A, Collins D, Artz A, et al. Paucity of HLA-identical unrelated donors for African-Americans with hematologic malignancies: the need for new donor options. Biol Blood Marrow Transplant. 2008;14:938–941. 10. Beatty PG, Mori M, Milford E. Impact of racial genetic polymorphism on the probability of finding an HLAmatched donor. Transplantation. 1995;60:778–783. 11. Ayas M, Al-Seraihi A, Al-Jefri A, et al. Unrelated cord blood transplantation in pediatric patients: a report from Saudi Arabia. Bone Marrow Transplant. 2010;45:1281–1286. 12. Aplenc R, Alonzo TA, Gerbing RB, et al. Ethnicity and survival in childhood acute myeloid leukemia: a report from the Children’s Oncology Group. Blood. 2006;108:74–80. 13. Klein T, Yaniv I, Stein J, et al. Extended family studies for the identification of allogeneic stem cell transplant donors in Jewish and Arabic patients in Israel. Pediatr Transplant. 2005;9:52–55. 14. Eggertson L. National cord blood bank to recruit ethnic donors aggressively. CMAJ. 2011;183:E537–E538. 15. Silversides A. Canada’s genetic heterogeneity an asset in cord blood banking. CMAJ. 2009;180:E117. 16. Thornley I, Eapen M, Sung L, et al. Private cord blood banking: experiences and views of pediatric hematopoietic cell transplantation physicians. Pediatrics. 2009;123:1011–1017. 17. Laver JH, Hulsey TC, Jones JP, et al. Assessment of barriers to bone marrow donation by unrelated African-American potential donors. Biol Blood Marrow Transplant. 2001;7:45–48. 18. Studts JL, Ruberg JL, McGuffin SA, et al. Decisions to register for the National Marrow Donor Program: rational vs emotional appeals. Bone Marrow Transplant. 2010;45:422–428. 19. Switzer GE, Bruce JG, Myaskovsky L, et al. Race and ethnicity in decisions about unrelated hematopoietic stem cell donation. Blood. 2013;121:1469–1476. 20. Risch N, Burchard E, Ziv E, et al. Categorization of humans in biomedical research: genes, race and disease. Genome Biol. 2002;3. comment2007.1–2007.12. 21. Rosenberg NA, Pritchard JK, Weber JL, et al. Genetic structure of human populations. Science. 2002;298: 2381–2385.
2014 Wolters Kluwer Health, Inc. All rights reserved.
www.jpho-online.com |
e157
Impact of Ethnicity on Donor Search Results for Children Requiring Stem Cell Transplantation Julia DiLabio, MSc,* John Doyle, MD,w Sarah Alexander, MD,z Sumit Gupta, MD, PhD,z and Angela Punnett, MDz Summary: Hematopoietic stem cell transplantation (HSCT) can be curative for children with various malignant and nonmalignant conditions. Access to suitable unrelated living donors (ULDs) or unrelated cord blood (UCB) may be limited for certain ethnic backgrounds. We therefore determined the impact of ethnicity upon donor availability in a pediatric cohort referred for allogeneic HSCT to a single Canadian transplant center. Among 252 eligible patients, 58 (23.0%) had suitable family donors. Of 161 patients with combined ULD and UCB searches, 78 (48.4%) had a suitable ULD, whereas 143 (88.8%) had suitable UCB. The probability of finding a suitable ULD differed significantly by ethnicity (P = 0.007). Non-white patients were significantly less likely to have suitable ULDs (odds ratio [OR] 0.35; 95% confidence interval [CI], 0.17-0.69; P = 0.003) compared with white patients but were equally likely to have suitable UCB (OR 1.02; 95% CI, 0.36-2.89; P = 0.97). Although ethnic disparities exist in pediatric patients’ access to ULD for HSCT, they are narrowed by the availability of international UCB registries. These findings, however, also highlight the importance of continued recruitment of individuals of non-white ethnicities to donor registries. Key Words: stem cell transplantation, bone marrow, umbilical cord blood, donors, ethnicity
(J Pediatr Hematol Oncol 2015;37:e154–e157)
A
llogeneic hematopoietic stem cell transplantation (HSCT) is an integral part of treatment for children diagnosed with some types of hematologic malignancies, hemoglobinopathies, immunodeficiencies, and congenital metabolic disorders. Stem cells can be obtained from the bone marrow or peripheral blood of family members (typically siblings) or unrelated donors, or from umbilical cord blood sources. Finding a suitable donor involves matching human leukocyte antigens (HLA) between patient and donor; higher degrees of HLA-mismatch have been linked to inferior survival.1 Umbilical cord blood stem cells require less stringent HLA-matching criteria for successful transplants compared with bone marrow or peripheral blood sources2; recent data suggests that these sources have
Received for publication July 28, 2014; accepted September 13, 2014. From the *Faculty of Medicine, University of Toronto; zDivision of Hematology/Oncology, SickKids Hospital, Toronto; and wSection of Pediatric Hematology/Oncology, Winnipeg Children’s Hospital, Winnipeg, Canada. S.G. and A.P. are co-senior authors. The authors declare no conflict of interest. Reprints: Angela Punnett, MD, Division of Hematology/Oncology, SickKids Hospital, 555 University Avenue, Toronto M5G 1X8, ON, Canada (e-mail: [email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.jpho-online.com. Copyright r 2014 Wolters Kluwer Health, Inc. All rights reserved.
e154 | www.jpho-online.com
similar outcomes and overall survival in the pediatric setting.3 The availability of appropriately HLA-matched stem cells has been shown in adult studies to vary according to patient ethnicity.4–6 It is generally more difficult to find suitably matched donor sources for non-white and multiethnic individuals, predominantly because of their underrepresentation among donors.7–10 The relationship between patient ethnicity and access to suitable donors has not been thoroughly investigated in a pediatric population, with ethnicity or availability of suitable donors comprising only a minor component of existing studies.11–13 We, therefore, conducted a single institution study to describe the impact of patient ethnicity on the availability of suitable unrelated living donors (ULDs) (ie, bone marrow and peripheral blood sources) and on the availability of suitable unrelated cord blood (UCB) for a contemporaneous pediatric cohort to inform family counseling and future advocacy efforts.
MATERIALS AND METHODS Study Design We conducted a retrospective cohort study. Eligible patients included all those assessed by SickKids Hospital (Toronto, Canada) for possible primary allogeneic HSCT between January 2009 and August 2012. A parent or substitute decision maker provided informed consent for a formal donor search in all cases. We excluded cases lacking self-reported ethnicity or other essential data (patient demographics, availability of an HLA-matched related donor, or combined ULD and UCB search results).
Data Source Data were collected from electronic patient charts and from an HSCT database established in 2009. Variables collected included patient age, disease at presentation (leukemia vs. other malignancy vs. nonmalignant disorder), referral center, and availability of an HLA-matched related donor. Ethnicity is a difficult concept to define objectively. Similar to previous reports,4–10 we used self-reported ethnicity and ancestry, as declared by parents on HSCT referral forms, to reclassify data according to both broad ethnic groups (white, Asian, black, mixed) and smaller ethnic subgroups (Caucasian, Middle Eastern, South Asian, East Asian, black, mixed). Hispanic patients were classified among the white group and Caucasian subgroup. This study received appropriate institutional ethics review and was approved by the Quality Management Department at SickKids Hospital.
Donor Search Analyses Patients and potential related donors, ULDs, and UCB underwent low-resolution and high-resolution HLA typing at 10 alleles (HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DQB1). Sibling donors were deemed J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
suitable if low-resolution matching occurred at HLA-A and HLA-B with high-resolution matching at HLA-DRB1 (6/6 alleles), provided that there were no inconsistencies in allele assignment from family studies. Otherwise, related donors were deemed suitable if high-resolution matching occurred at HLA-A, HLA-B, HLA-C, and HLA-DRB1 (8/8 alleles). Patients lacking an HLA-matched related donor underwent searches simultaneously for both ULDs (through the Canadian OneMatch and international registries for bone marrow and peripheral blood stem cell sources) and UCB (through international umbilical cord blood registries). Suitable ULDs were defined and recorded by HSCT clinicians in the database according to the following guidelines: high-resolution matching at HLA-A, HLA-B, HLA-C, and HLA-DRB1 at 7/8 or 8/8 alleles (with matching at HLADQB1 in the former case). Suitable UCB was defined as low-resolution matching at HLA-A and HLA-B with highresolution matching at HLA-DRB1 and adequate total nucleated cell dose (Z5 107/kg/unit, Z3 107/kg/unit, and Z1.5 107/kg/unit for 4/6, 5/6, and 6/6 matched cords, respectively). ULDs and UCB that did not meet infectious disease screening requirements or were listed as unavailable were also deemed unsuitable.
Statistical Analysis Demographic characteristics among included and excluded patients were compared using the Student’s t test, w2, or the Fisher exact test as appropriate. Predictors of the presence of an HLA-matched related donor were determined using univariate and multivariable logistic regression. The percentage of patients with suitable stem cell sources (ULD, UCB, or any source) was compared across ethnic groups using the Fisher exact test. The odds of identifying suitable ULDs, UCB, or any stem cell source were compared between white and non-white patients by logistic regression analysis. On the basis of the underrepresentation of Aboriginal peoples and black individuals within international ULD registries, exploratory analyses focused on these populations were specified a priori. Statistical significance was defined as P < 0.05. All statistical analyses were conducted using SAS-PC v.9.2 (Cary, NC).
RESULTS During the study period, 343 children were referred to SickKids Hospital for consideration of primary allogeneic HSCT. Reasons for exclusion are shown in Figure 1; the final cohort consisted of 252 patients (73.5%) ranging from 0 to 20 years of age. In comparison with the study cohort, excluded patients lacking ethnicity data were more likely to be referred to SickKids from an external medical center (43/ 51, 84.3% vs. 72/252, 28.6%; P < 0.0001) and were more likely to be transplanted (33/51, 64.7% vs. 117/252, 46.4%; P = 0.02). The median age and distribution of diagnoses were not statistically different between the 2 groups. Patient demographics are illustrated in Table 1. When classified by broad ethnic groups or subgroups, the majority of patients were white (152/252, 60.3%) or Caucasian (134/252, 52.3%), respectively. Only 58 of 252 patients (23.0%) had an HLA-matched donor within their immediate family. As anticipated, larger family size was associated with higher odds of having an HLA-matched related donor (odds ratio [OR] 2.26; 95% confidence interval [CI], 1.62-3.16; P < 0.0001) in univariate analysis. Middle Eastern children had a greater Copyright
r
Ethnicity and Donors for Stem Cell Transplantation
FIGURE 1. Flow diagram depicting the exclusion of ineligible cases from the study cohort. *Insufficient data other than patient ethnicity (results of HLA typing of family members, ULD, or UCB). UCB indicates unrelated cord blood; ULD, unrelated living donor.
probability of having an HLA-matched related donor (OR 4.46; 95% CI, 1.57-12.64; P = 0.005) compared with Caucasian children. Both variables retained significance in multivariable analysis (see Supplemental Table 2, http:// links.lww.com/JPHO/A81). Of the study cohort, 161 patients (63.9%) underwent combined ULD and UCB searches. A suitable ULD was identified for approximately half of these patients (78/161, 48.4%), of which the majority were fully HLA-matched (65/78, 83.3%). The majority of patients (143/161, 88.8%) had a suitable single UCB unit identified, of which the minority were fully HLA-matched (61/143, 42.7%). Combining ULD and UCB search results revealed that the majority of patients (154/161, 95.6%) had at least one
TABLE 1. Demographic Characteristics of Children Considered for Hematopoietic Stem Cell Transplantation During the Study Period
Characteristic Age (y) Diagnosis Nonmalignancy Leukemia Other malignancy Referral center SickKids Ontario Other province Ethnic group White Asian Black Mixed Ethnic subgroup Caucasian Middle Eastern South Asian East Asian Black Mixed Family match Present Absent
2014 Wolters Kluwer Health, Inc. All rights reserved.
n (%)
Median (IQR) 7 (2-12)
131 (52.0) 113 (44.8) 8 (3.2) 180 (71.4) 56 (22.2) 16 (6.4) 152 61 22 17
(60.3) (24.2) (8.7) (6.8)
134 17 35 26 22 18
(52.3) (6.8) (13.9) (11.5) (8.7) (7.1)
58 (23.0) 194 (77.0)
IQR indicates interquartile range.
www.jpho-online.com |
e155
J Pediatr Hematol Oncol
DiLabio et al
n (%) Ethnicity
Suitable ULD
Suitable UCB
61 (57.0) 13 (34.2) 0 (0.0) 4 (36.4) 0.007
95 33 5 10
(88.8) (86.8) (100.0) (90.9) 0.97
Any Suitable 102 (95.3) 37 (97.4) 5 (100.0) 10 (90.9) 0.71
n indicates number; UCB, unrelated cord blood; ULD, unrelated living donor. *The Fisher exact test.
potential source of stem cells, leaving only 7 of 161 patients (4.4%) without a suitable donor option. The success of finding a suitable ULD was statistically different between white, Asian, black, and multiethnic groups (P = 0.007), as shown in Table 2. The probability of identifying suitable UCB, however, did not vary significantly between ethnic groups (P = 0.97), with the majority of patients having at least 1 suitable option. Combining ULD and UCB search results yielded a suitable stem cell source for >90% of patients, irrespective of their ethnic background (P = 0.71). Collapsing ethnic groups into white and non-white patients revealed that the latter were significantly less likely to have a suitable ULD (OR 0.35; 95% CI, 0.17-0.69; P = 0.003), as shown in Table 3. However, both groups had a similar likelihood of having suitable UCB (OR 1.02; 95% CI, 0.36-2.89; P = 0.97). The likelihood of identifying a suitable stem cell source of any type was also statistically similar between the 2 groups (OR 1.27; 95% CI, 0.24-6.8; P = 0.78). Exploratory subgroup analyses revealed that the success rate of finding suitable stem cell sources for Aboriginal children was comparable to the overall success rate for the study population. Of the 9 Aboriginal and Aboriginal/ white children who underwent unrelated donor searches, 4 (44.4%) had a suitable ULD and 8 (88.8%) had suitable UCB, such that only 1 patient (11.1%) had no suitable stem cell source identified. No suitable ULD was identified for any of the black children (n = 5), although each had at least 1 suitable UCB source.
TABLE 3. Comparison of Combined ULD and UCB Search Results Among White and Non-white Patients
Variable Suitable ULD White Non-white Suitable UCB White Non-white Any suitable White Non-white
OR (95% CI)
P*
Ref. 0.35 (0.17-0.69)
— 0.003
Ref. 1.02 (0.36-2.89)
— 0.97
Ref. 1.27 (0.24-6.80)
— 0.78
CI indicates confidence interval; OR, odds ratio; Ref., reference; UCB, unrelated cord blood; ULD, unrelated living donor. *Univariate logistic regression analysis.
e156 | www.jpho-online.com
Volume 37, Number 3, April 2015
DISCUSSION
TABLE 2. Proportion of Children for Whom Suitable Donors Were Identified, Stratified by Ethnicity
White Asian Black Mixed P*
We have identified ethnic disparities in the success rates of finding suitable ULDs for pediatric patients in a contemporaneous ethnically diverse population. Non-white patients, and particularly black children, had relatively limited access to suitable ULDs compared with white patients. The ability to search international UCB registries increased the number of potential stem cell options for all patients, irrespective of ethnic background, such that ethnicity did not impact on the overall likelihood of finding a suitable stem cell source. An HLA-matched related donor was identified for only 23% of patients, which is lower than the typical likelihood of 25% to 40% based on the average North American family size.4,5,12 This likelihood may vary by ethnicity; an HLA-identical related donor was found for only 7.5% of black children in 1 study.12 Middle Eastern children in our cohort were more likely to have an HLAidentical donor within their family than white children. This finding is consistent with a study involving a pediatric cohort in Israel showing that the majority of Arabic patients (62%) had an HLA-identical family member, a finding that the authors attributed in part to a more homogenous gene pool in their patient population.13 Consistent with previous studies, our results suggest that ULD registries serve a predominantly white patient population, with non-white patients having a significantly lower probability of finding a suitable donor.4,6–10 The finding that black children had particularly poor access to suitable ULDs is consistent with studies showing that the National Marrow Donor Program has struggled to find suitable ULDs for African American patients, due both to their underrepresentation within donor registries and to the inherent genetic HLA variability within this ethnic group.6–10 The finding that a greater proportion of children had suitable UCB identified (89%) compared with a suitable ULD (48%) is secondary to the less stringent HLA matching criteria required for UCB to be deemed suitable.2 The ability to search within UCB registries, therefore, increases the availability of suitable stem cells and decreases ethnic disparities. Our findings are supported by those of Barker et al4 demonstrating that suitable UCB was identified for the majority of patients of European (96%), Asian (100%), African (80%), or mixed non-European (91%) backgrounds using a similar definition of suitability. Our findings demonstrate the importance of international UCB registries in improving access to stem cells to children of diverse ethnicities. We anticipate that the Canadian UCB bank will have a positive impact in extending access to stem cells to patients of non-white ethnic backgrounds in a similar manner as international UCB registries, provided that emphasis is placed on the recruitment of ethnically diverse registrants during the development phase, as recommended elsewhere.14,15 The high cost of obtaining UCB from international registries, the ability to collect otherwise unused cords from ethnically diverse mothers, and the lack of strong endorsement from pediatric HSCT physicians for private cord blood banking in the absence of an identified recipient provide further justification for a Canadian UCB registry.16 Nevertheless, our findings also underscore the importance of maintaining an ethnically diverse national stem cell donor registry and recruiting black donors to help fill the gap in availability. Targeted donor drives and education programs have been shown to aid in recruiting new African
Copyright
r
2014 Wolters Kluwer Health, Inc. All rights reserved.
J Pediatr Hematol Oncol
Volume 37, Number 3, April 2015
American donors and could be extended to other ethnicities.17 Employing recruitment strategies targeting the emotional response of potential donors may also be of benefit,18 especially as donation-related ambivalence is an issue across all ethnic groups.19 Several study limitations deserve mention. This was a retrospective single institution study and the findings may not be generalizable to a broader North American population. The exclusion of 91/343 cases (26.5%) because of lack of patient ethnicity or other pertinent data limited the size of the study population. Given the retrospective nature of this study, this data was not obtainable, and assumptions regarding patients’ ethnicity according to surname were deemed inappropriate. Referrals from external medical centers were more likely to lack ethnicity data; excluded patients were therefore more likely to be referred from such centers. This may also explain why excluded patients were more likely to be transplanted, as referrals from within our institution may have been more likely to include relative (as opposed to absolute) indications for HSCT. Although differences between included and excluded cases were not felt to bias the findings in any manner, such bias cannot be ruled out. Secondly, to maintain adequate statistical power, several different ethnic groups were grouped as nonwhite, given the limited numbers of particular non-white backgrounds. The benefit of having access to the self-reported ethnicity and national origins of patients’ parents was circumscribed as a result. Thirdly, our definition of suitability implied equivalence between fully and partially matched stem cells and did not take into account the degree of HLA matching among different ethnic groups. Finally, differential priority may have been placed on identifying a suitable ULD or UCB among patients based on factors such as disease or transplant urgency. Despite these limitations, focusing on the donor identification process rather than the actual transplant source allowed important findings to be generated, irrespective of the transplant status of patients. The findings may impact treatment planning and will inform the future counseling of families on the probability of finding a suitable unrelated donor or cord for their child. A final study strength was the use of parents’ self-reported ethnicity and national origins, which has been cited as an accurate representation of patient ancestry.20,21 In summary, this study, in providing the first analysis of the availability of suitable ULDs and UCB for a pediatric cohort requiring allogeneic HSCT in North America, has revealed ethnic disparities in the availability of suitable ULDs that are partially mitigated by the availability of UCB. The findings strongly support both the development of an ethnically diverse national public cord blood bank in Canada and the targeted recruitment of non-white individuals to stem cell donor registries to improve pediatric patients’ access to hematopoietic stem cells for transplantation. Future studies should assess the impact of ethnicity upon the degree of HLA matching of potential or transplanted stem cell sources, as well as on clinical outcomes.
ACKNOWLEDGMENTS The authors thank Elizabeth Davies for assistance with the HSCT database and Marie Pinard for support in project development.
Copyright
r
Ethnicity and Donors for Stem Cell Transplantation
REFERENCES 1. Lee SJ, Klein J, Haagenson M, et al. High-resolution donorrecipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood. 2007;110:4576–4583. 2. Eapen M, Rubinstein P, Zhang MJ, et al. Outcomes of transplantation of unrelated donor umbilical cord blood and bone marrow in children with acute leukaemia: a comparison study. Lancet. 2007;369:1947–1954. 3. Wang J, Zhan P, Ouyang J, et al. Unrelated donor umbilical cord blood transplantation versus unrelated donor bone marrow transplantation in adult and pediatric patients: a meta-analysis. Leuk Res. 2010;34:1018–1022. 4. Barker JN, Byam CE, Kernan NA, et al. Availability of cord blood extends allogeneic hematopoietic stem cell transplant access to racial and ethnic minorities. Biol Blood Marrow Transplant. 2010;16:1541–1548. 5. Dehn J, Arora M, Spellman S, et al. Unrelated donor hematopoietic cell transplantation: factors associated with a better HLA match. Biol Blood Marrow Transplant. 2008; 14:1334–1340. 6. Yancey AK, Coppo P, Kawanishi Y. Progress in availability of donors of color: the National Marrow Donor Program. Transplant Proc. 1997;29:3760–3765. 7. Pidala J, Kim J, Schell M, et al. Race/ethnicity affects the probability of finding an HLA-A, -B, -C and -DRB1 allelematched unrelated donor and likelihood of subsequent transplant utilization. Bone Marrow Transplant. 2013;48:346–350. 8. Joshua TV, Rizzo JD, Zhang MJ, et al. Access to hematopoietic stem cell transplantation: effect of race and sex. Cancer. 2010;116:3469–3476. 9. Dew A, Collins D, Artz A, et al. Paucity of HLA-identical unrelated donors for African-Americans with hematologic malignancies: the need for new donor options. Biol Blood Marrow Transplant. 2008;14:938–941. 10. Beatty PG, Mori M, Milford E. Impact of racial genetic polymorphism on the probability of finding an HLAmatched donor. Transplantation. 1995;60:778–783. 11. Ayas M, Al-Seraihi A, Al-Jefri A, et al. Unrelated cord blood transplantation in pediatric patients: a report from Saudi Arabia. Bone Marrow Transplant. 2010;45:1281–1286. 12. Aplenc R, Alonzo TA, Gerbing RB, et al. Ethnicity and survival in childhood acute myeloid leukemia: a report from the Children’s Oncology Group. Blood. 2006;108:74–80. 13. Klein T, Yaniv I, Stein J, et al. Extended family studies for the identification of allogeneic stem cell transplant donors in Jewish and Arabic patients in Israel. Pediatr Transplant. 2005;9:52–55. 14. Eggertson L. National cord blood bank to recruit ethnic donors aggressively. CMAJ. 2011;183:E537–E538. 15. Silversides A. Canada’s genetic heterogeneity an asset in cord blood banking. CMAJ. 2009;180:E117. 16. Thornley I, Eapen M, Sung L, et al. Private cord blood banking: experiences and views of pediatric hematopoietic cell transplantation physicians. Pediatrics. 2009;123:1011–1017. 17. Laver JH, Hulsey TC, Jones JP, et al. Assessment of barriers to bone marrow donation by unrelated African-American potential donors. Biol Blood Marrow Transplant. 2001;7:45–48. 18. Studts JL, Ruberg JL, McGuffin SA, et al. Decisions to register for the National Marrow Donor Program: rational vs emotional appeals. Bone Marrow Transplant. 2010;45:422–428. 19. Switzer GE, Bruce JG, Myaskovsky L, et al. Race and ethnicity in decisions about unrelated hematopoietic stem cell donation. Blood. 2013;121:1469–1476. 20. Risch N, Burchard E, Ziv E, et al. Categorization of humans in biomedical research: genes, race and disease. Genome Biol. 2002;3. comment2007.1–2007.12. 21. Rosenberg NA, Pritchard JK, Weber JL, et al. Genetic structure of human populations. Science. 2002;298: 2381–2385.
2014 Wolters Kluwer Health, Inc. All rights reserved.
www.jpho-online.com |
e157
