|Year : 2017 | Volume
| Issue : 2 | Page : 84-90
Using high-resolution human leukocyte antigen typing of 11,423 randomized unrelated individuals to determine allelic varieties, deduce probable human leukocyte antigen haplotypes, and observe linkage disequilibria between human leukocyte antigen-B and-C and human leukocyte antigen-DRB1 and-DQB1 alleles in the Taiwanese Chinese population
Kuo-Liang Yang1, Hsee-Bin Chen2
1 Laboratory of Immunogenetics, Tzu Chi Cord Blood Bank and Buddhist Tzu Chi Marrow Donor Registry, Buddhist Tzu Chi Stem Cells Centre, Hualien Tzu Chi Hospital, Hualien, ; Department of Laboratory Medicine, Buddhist Tzu Chi University, Hualien, Taiwan
2 Laboratory of Immunogenetics, Tzu Chi Cord Blood Bank and Buddhist Tzu Chi Marrow Donor Registry, Buddhist Tzu Chi Stem Cells Centre, Hualien Tzu Chi Hospital, Hualien, Taiwan
|Date of Submission||09-Mar-2017|
|Date of Decision||29-Mar-2017|
|Date of Acceptance||03-May-2017|
|Date of Web Publication||15-Jun-2017|
Buddhist Tzu Chi Stem Cells Centre, Hualien Tzu Chi Hospital, 707, Section 3, Chung-Yang Road, Hualien
Source of Support: None, Conflict of Interest: None
Objective: We report here the human leukocyte antigen (HLA) allelic variety and haplotype composition in a cohort of the Taiwanese Chinese population and their patterns of linkage disequilibria on HLA-B: HLA-C alleles and HLA-DRB1: HLA-DQB1 alleles at a high-resolution level. Materials and Methods: Peripheral whole blood from 11,423 Taiwanese Chinese unrelated individuals was collected in acid citrate dextrose. Genomic DNA was extracted using the QIAamp DNA Blood Mini Kit. The DNA material was subjected to HLA genotyping for HLA-A,-B,-C,-DRB1, and-DQB1 loci using a commercial polymerase chain reaction-sequence-based typing (PCR-SBT) kit, the SeCore® A/B/C/DRB1/DQB1 Locus Sequencing kit. High-resolution allelic sequencing was performed as previously described. Results: The number of individual HLA-B alleles detected was greater than the number of alleles recognized in the both the HLA-A and-DRB1 loci. Several novel alleles were discovered as a result of employing the SBT method and the high number of donors tested. In addition, we observed a genetic polymorphic feature of association between HLA-A and-B, HLA-B and-C, and HLA-DRB1 and-DQB1 alleles. Further, the homozygous haplotype frequencies of HLA-A and-B; HLA-A,-C, and-B; HLA-A,-C,-B, and-DRB1; and HLA-A,-C,-B,-DRB1, and-DQB1 in Taiwanese Chinese population are presented. Conclusion: As increasing number of HLA alleles are being discovered, periodic HLA profile investigation in a given population is essential to recognize the HLA complexity in that population. Population study can also provide an up-to-date strategic plan for future needs in terms of compatibility measurement for HLA matching between transplant donors and patients.
Keywords: Alleles, Haplotypes, Human leukocyte antigen, Induced pluripotent stem cells, Linkage disequilibria, Taiwanese
|How to cite this article:|
Yang KL, Chen HB. Using high-resolution human leukocyte antigen typing of 11,423 randomized unrelated individuals to determine allelic varieties, deduce probable human leukocyte antigen haplotypes, and observe linkage disequilibria between human leukocyte antigen-B and-C and human leukocyte antigen-DRB1 and-DQB1 alleles in the Taiwanese Chinese population. Tzu Chi Med J 2017;29:84-90
|How to cite this URL:|
Yang KL, Chen HB. Using high-resolution human leukocyte antigen typing of 11,423 randomized unrelated individuals to determine allelic varieties, deduce probable human leukocyte antigen haplotypes, and observe linkage disequilibria between human leukocyte antigen-B and-C and human leukocyte antigen-DRB1 and-DQB1 alleles in the Taiwanese Chinese population. Tzu Chi Med J [serial online] 2017 [cited 2020 May 29];29:84-90. Available from: http://www.tcmjmed.com/text.asp?2017/29/2/84/208135
| Introduction|| |
Transplantation of allogeneic hematopoietic stem cells has been employed as a curative therapy for hematologic malignancies and other hematologic or immune disorders. Human leukocyte antigen (HLA) molecules have been defined as transplant antigens and have a strong relevance to tissue transplantation. The molecular similarity between transplant donors and recipients is considered a predictive factor for graft survival and graft versus host disease as it can elicit immune responses by either recognition of polymorphic fragments of foreign HLA molecules or presentation of various peptides ,,. The genes encoding the HLA alleles are located in the Major Histocompatibility Complex Class I and II regions. The HLA genes are characterized by their extreme allelic polymorphism as well as their variations and diversity in different ethnic groups. The number of known HLA alleles is increasing dramatically with the development of DNA-based molecular typing technology . Understanding the HLA diversity in ethnic groups is important. Facilitating appropriate HLA-matched unrelated bone marrow stem cell donors for successful stem cell transplantations relies on the accuracy of HLA typing and the ability to resolve problems with unknown, ambiguous, and low-incidence genes in the HLA system. In addition, determination of HLA haplotypes is essential for matching between donor and recipient in unrelated stem cell transplantation since it increases the likelihood of matching at other loci within the HLA region compared with matching only at the individual allelic level .
Determination of HLA haplotypes can be done by HLA typing of blood-related family members and prediction from large-sized population tissue typing ,,. Alternatively, it can be achieved by deducing typing results from donors with allelic homozygosities in the HLA-A,-B, and-DRB1 loci . In family study, segregation of HLA individual alleles provides evidence of allelic linkages . In population study, determination of haplotypes involves noting whether alleles at other loci are consistently present and family study is not performed. Instead, most available haplotype data are derived from studies of unrelated individuals in whom the putative haplotype is defined by statistical association analysis . Linkage disequilibrium (LD), a phenomenon whereby certain combinations of alleles occur in HLA haplotypes within a population more frequently than expected based on gene frequencies, is commonly observed and has important clinical and biological implications . The alleles of HLA-B and-C as well as HLA-DRB1 and-DQB1 are known to have strong LD, yet very few LD reports on the high-resolution allelic level are available in the Taiwanese Chinese population.
We report here the HLA allelic variety and haplotype composition of a cohort of the Taiwanese Chinese population consisting 11,423 unrelated individuals and show the patterns of LD on HLA-B: HLA-C alleles and HLA-DRB1: HLA-DQB1 alleles at the high-resolution level found in Taiwanese Chinese. We believe that as increasing number of HLA alleles are being discovered, periodic HLA profile investigation in a given population is essential to recognize the HLA complexity in that population. Population study can also provide an up-to-date strategic plan for future needs in terms of compatibility measurement for HLA matching between transplant donors and patients.
| Materials and Methods|| |
Human leukocyte antigen DNA typing
The population of Taiwanese Chinese includes four ethnic groups (those whose ancestors migrated from Fujian province and Guangdong province, mainland Chinese from provinces other than Fujian and Guangdong, and native aborigines). Peripheral whole blood samples of Taiwanese Chinese were collected in acid citrate dextrose (ACD) anticoagulant from Taiwanese Chinese unrelated volunteer bone marrow donors with formal written consent at routine blood drives.
Aliquots of the ACD whole blood were stored frozen at −80°C until use. Peripheral blood genomic DNA was extracted using the QIAamp DNA Blood Mini Kit according to the manufacturer's instructions (Qiagen, Hilden, Germany). The DNA material was subjected to HLA genotyping for HLA-A,-B,-C,-DRB1, and-DQB1 loci using a commercial polymerase chain reaction- sequence-based typing (PCR-SBT) kit, the SeCore ® A/B/C/DRB1/DQB1 Locus Sequencing kit (Life Technologies, Brown Deer, WI, USA). High-resolution allelic sequencing was performed as previously described ,.
Determination of linkage disequilibrium by statistical analysis
The frequencies of HLA Class I (HLA-A,-B, and-C) and II (HLA-DRB1 and-DQB1) alleles were calculated by direct counting, and the patterns of extended haplotypes were estimated using the maximum-likelihood method with the expectation–maximization (EM) algorithm using Arlequin version 3.5 .
| Results|| |
Human leukocyte antigen-A,-B,-C,-DRB1, and-DQB1 alleles
[Table 1] shows the number of individuals tested for each HLA-A, -B,-C,-DRB1, and-DQB1 loci and the number of distinctive alleles identified in each locus in this study. The frequencies of the alleles identified are listed in [Appendix 1 [Additional file 1]]. Almost an equal number of donors were tested for HLA-A,-B, and-DRB1 loci, and we found that the number of the alleles detected in the HLA-B locus exceeded the number of alleles found in both the HLA-A and-DRB1 loci. In [Table 2], the homozygous haplotype frequencies of HLA-A and-B; HLA-A,-C, and-B; HLA-A,-C,-B, and-DRB1; and HLA-A,-C,-B,-DRB1, and-DQB1 in the Taiwanese Chinese population are presented. In general, the frequency of homozygous haplotypes decreased as the number of loci increased.
|Table 1: Number of Taiwanese Chinese tested and number of alleles identified|
Click here to view
In the HLA-A locus, we found that the alleles with a frequency higher than 1% were A*11:01 (28.36%), A*24:02 (17.29%), A*02:07 (11.76%), A*02:01 (10.62%), A*33:03 (9.78%), A*02:03 (6.49%), A*11:02 (4.07%), A*02:06 (2.82%), A*26:01 (2.22%), A*30:01 (1.96%), and A*31:01 (1.78%) [Appendix 1].
For the HLA-B locus, the alleles detected with a frequency above 1% were B*40:01 (21.04%), B*46:01 (14.03%), B*58:01 (9.60%), B*13:01 (6.51%), B*15:02 (5.43%), B*51:01 (4.46%), B*38:02 (4.17%), B*15:01 (4.03%), B*54:01 (3.23%), B*55:02 (2.87%), B*27:04 (2.78%), B*39:01 (2.49%), B*35:01 (2.36%), B*13:02 (2.24%), B*40:02 (1.70%), B*40:06 (1.52%), B*48:01 (1.31%), and B*51:02 (1.23%) [Appendix 1].
The alleles we found with a frequency exceeding 1% in the HLA-C locus were C*07:02 (20.91%), C*01:02 (19.03%), C*03:04 (12.49%), C*08:01 (8.93%), C*03:02 (8.55%), C*03:03 (5.55%), C*15:02 (4.32%), C*04:01 (4.18%), C*14:02 (3.85%), C*12:02 (3.47%), C*06:02 (3.12%), and C*04:03 (1.64%) [Appendix 1].
For the HLA-DRB1 locus, the alleles observed with a frequency more than 1% were DRB1*09:01 (15.70%), DRB1*12:02 (10.39%), DRB1*15:01 (9.53%), DRB1*08:03 (8.56%), DRB1*11:01 (7.71%), DRB1*04:05 (7.52%), DRB1*03:01 (6.82%), DRB1*16:02 (5.20%), DRB1*14:54 (4.09%), DRB1*12:01 (3.38%), DRB1*04:03 (3.02%), DRB1*04:06 (2.91%), DRB1*07:01 (2.84%), DRB1*13:02 (2.34%), DRB1*15:02 (2.19%), DRB1*14:05 (2.09%), and DRB1*04:04 (1.10%) [Appendix 1].
In the HLA-DQB1 locus, the alleles identified with a frequency greater than 1% were DQB1*03:01 (20.90%), DQB1*03:03 (17.31%), DQB1*06:01 (13.30%), DQB1*05:02 (10.52%), DQB1*03:02 (6.59%), DQB1*04:01 (5.99%), DQB1*02:01 (5.61%), DQB1*06:02 (4.72%), DQB1*05:03 (4.34%), DQB1*02:02 (3.28%), DQB1*05:01 (2.71%), and DQB1*06:09 (2.27%) [Appendix 1].
Association of human leukocyte antigen-A and-B alleles
The EM algorithm predictions of linkage disequilibria in HLA-A and-B alleles in Taiwanese Chinese are shown in [Appendix 2 [Additional file 2]]. Many HLA-A alleles were found to associate with several different types of HLA-B alleles. [Table 3] shows the variable HLA-B alleles in association with the three most commonly observed HLA-A alleles (A*02:07, A*11:01, and A*24:02) in Taiwanese Chinese.
|Table 3: Association of human leukocyte antigen-B alleles with the three most commonly observed human leukocyte antigen-A alleles in Taiwanese Chinese|
Click here to view
Association of human leukocyte antigen-B and-C alleles
The EM algorithm predictions of linkage disequilibria in HLA-B and-C alleles are presented in [Appendix 2]. Similar to the findings in the HLA-A locus, many HLA-B alleles were observed to link heterogeneously with several types of HLA-C alleles. [Table 4] illustrates the variable HLA-C alleles in association with the three most frequently detected HLA-B alleles (B*40:01, B*46:01, and B*58:01) in Taiwanese Chinese.
|Table 4. Association of human leukocyte antigen-C alleles with the three most commonly observed human leukocyte antigen-B alleles in Taiwanese Chinese|
Click here to view
Association of human leukocyte antigen-DRB1 and-DQB1 alleles
The EM algorithm predictions of linkage disequilibria in the HLA-DRB1 and-DQB1 alleles in Taiwanese Chinese are listed in [Appendix 2]. [Table 5] shows the variable HLA-DQB1 alleles associated with the three most prevalently found alleles in HLA-DRB1 (DRB1*09:01, DRB1*12:02, and DRB1*15:01) in Taiwanese Chinese.
|Table 5: Association of human leukocyte antigen-DQB1 alleles with the three most commonly observed human leukocyte antigen-DRB1 alleles in Taiwanese Chinese|
Click here to view
Human leukocyte antigen-A,-B, and-C haplotypes
[Table 6] shows the 20 most commonly observed HLA-A,-B, and-C haplotypes in Taiwanese Chinese as predicted by the EM algorithm. [Appendix 2] presents the list of HLA-A,-B, and-C haplotypes as predicted by the EM algorithm in this study.
|Table 6: The 20 most commonly observed human leukocyte antigen-A,-B, and-C haplotypes in Taiwanese Chinese|
Click here to view
Human leukocyte antigen-A,-B,-C, and-DRB1 haplotypes
The 20 most commonly recognized HLA-A,-B,-C, and-DRB1 haplotypes in Taiwanese Chinese as predicted by the EM algorithm are shown in [Table 7]. [Appendix 2] shows the list of HLA-A,-B,-C, and-DRB1 haplotypes in this study as predicted by the EM algorithm.
|Table 7: The 20 most commonly recognized human leukocyte antigen-A,B,C, and-DRB1 haplotypes in Taiwanese Chinese|
Click here to view
Human leukocyte antigen-A,-B,-C,-DRB1, and-DQB1 haplotypes
In [Table 8], the 20 most frequently observed HLA-A,-B,-C, -DRB1, and-DQB1 haplotypes in Taiwanese Chinese as predicted by the EM algorithm are presented. [Appendix 2] shows the list of HLA-A,-B,-C,-DRB1, and-DQB1 haplotypes in this study as predicted by the EM algorithm.
|Table 8: The 20 most frequently observed human leukocyte antigen-A,-B,C,DRB1, and-DQB1 haplotypes in Taiwanese Chinese|
Click here to view
| Discussion|| |
HLA molecules have been known as transplant antigens and have a strong relevance in tissue transplantation. The high-resolution SBT method can provide HLA allelic typing. High-resolution typing is advantageous for compatibility measurement between a potential stem cell donor and corresponding recipient.
In this study, a similar panel number of Taiwanese Chinese participants were tested for the HLA-A,-B, and-DRB1 alleles, and we found more individual HLA-B alleles than alleles in the HLA-A and-DRB1 loci. This is in agreement with studies in Chinese in the US donor registry, in a Chinese Han, and in a Taiwanese Chinese study reported by Gragert et al. , Li et al. , and Yang et al. , respectively. As a matter of fact, the same result was also observed in the 21 ethnic groups investigated in the US donor registry . It is noteworthy that although 30 times more donors were tested in the current study than in our previous study, we found the 10–15 most prevalently observed alleles in each locus coincided with those in our previous study . However, twice as many distinctive alleles were detected in each locus in this study than in the previous study . Most significantly, as a result of employing the SBT typing method and the high number of donors tested, several novel alleles were discovered. These novel alleles include 23 HLA-A, 28 HLA-B, 6 HLA-C, and 13 HLA-DRB1 alleles [Table 9]. We suspect that these novel alleles are low-frequency alleles and are most likely restricted to those of Taiwanese Chinese ethnicity.
|Table 9: The 23 human leukocyte antigen-A, 28 human leukocyte antigen-B, 6 human leukocyte antigen-C, and 13 human leukocyte antigen-DRB1 novel alleles detected in Taiwanese Chinese in this study|
Click here to view
The five most commonly observed HLA-A alleles (A*11:01, A*24:02, A*02:07, A*02:01, and A*33:03),-B alleles (B*40:01, B*46:01, B*58:01, B*13:01, and B*15:02),-C alleles (C*07:02, C*01:02, C*03:04, C*08:01, and C*03:02) were identical in this population study and the US-Chinese population study . However, only the three most prevalent alleles in the HLA-DRB1 locus (DRB1*09:01, DRB1*12:02, and DRB1*15:01) were identical between the two populations. The fourth and fifth most common alleles for the Taiwanese Chinese population were DRB1*08:03 and DRB1*11:01 and in the US-Chinese population were DRB1*03:01 and DRB1*08:03. These observations suggest that the population composition of HLA in Taiwanese Chinese and the US Chinese is similar but not identical.
From the point of view of LD, the diversity of alleles and allele associations was very obvious with the use of the molecular DNA typing method and the large number of donors tested in this study. [Table 3],[Table 4],[Table 5] show the genetic polymorphic features of association between HLA-A and-B, HLA-B and-C, and HLA-DRB1 and-DQB1 alleles. However, the 10–15 most commonly observed haplotypes of HLA-A-B-C, HLA-A-B-C-DRB1, and HLA-A-B-C-DRB1-DQB1 were similar in this study and the previous study . This result indicates that investigating a panel with as many as 400 random donors is probably sufficient to determine the 10 to15 most common HLA haplotypes in Taiwanese.
We found that the eight most common HLA-A/B/C/DRB1/DQB1 haplotypes were identical in the US-Chinese  and the Taiwanese Chinese population reported here, but the 6th and the 10th most common in the US-Chinese population ranked 12th and 16th in the Taiwanese Chinese population. This finding reiterates the close similarity in HLA composition of the US and Taiwanese Chinese populations, but they are not absolutely identical. It also indicates that the US may have a broader mix of people from various regions of China than Taiwan.
In recent years, the concept of generating induced pluripotent stem cells (iPSC) using human somatic cells to possibly produce personalized pluripotent stem cell lines from individual patients sparked the hope of producing iPSC cell lines for transplantation therapies . This approach could address rejection-associated issues concerning HLA matching of individual patients. However, the time and cost required for the generation of clinical-grade iPSC cell lines and the differentiated cell types for transplantation could limit its general application in clinical practice. A possibility to circumvent this hurdle is banking an inventory of iPSC cell lines with known HLAs, as was first proposed for the UK population . However, this approach is hindered by HLA diversity since a large number of iPSC cell lines would be needed for the best HLA compatibility for recipients. An alternative strategy to decrease the necessity for a large number of cell lines is to establish a cell line bank with HLA homozygous cells . With the large number of randomized donor cells studied at the high-resolution level in this study, we estimated the frequencies of homozygous HLA haplotypes in the Taiwanese Chinese population [Table 2]. We hope our data on HLA homozygosity in Taiwanese Chinese here may be useful as a reference for determining the number of iPSC cell lines needed to establish an HLA homozygous haplotype iPSC cell bank for the Taiwanese Chinese population for transplantation purposes.
| Conclusion|| |
We showed the increment number of HLA alleles at various loci and their variety and the number of allelic linkages between different loci. Further, we confirmed that the number of HLA haplotype combinations for a given population is directly proportional to the number of unrelated random donors tested and the level of HLA typing resolution achieved. In addition, novel HLA alleles revealed in a population study may inspire appreciation and trigger curiosity about the uniqueness and polymorphic nature of the HLA genetic system.
Declaration of patient consent
The authors certify that all patients provided appropriate patient consent forms. In the form, all patients gave consent for their images and other clinical information to be reported in the journal. All patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
We would like to express our sincere appreciation to Dharma Master Cheng Yen, founder of the Buddhist Compassion Relief Tzu Chi Foundation, for continuing support and kind encouragement both intellectually and spiritually. The generosity and camaraderie of our colleagues are also truly appreciated.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Gallo S, Woolfrey AE, Burroughs LM, Storer BE, Flowers ME, Hari P,et al .
Marrow grafts from HLA-identical siblings for severe aplastic anemia: Does limiting the number of transplanted marrow cells reduce the risk of chronic GvHD? Bone Marrow Transplant 2016;51:1573-8.
Liu Z, Chen G, Kang X, Han M, Chen R, Chen C,et al .
Amultiplex allele-specific real-time polymerase chain reaction assay for HLA-B*13:01 genotyping in four Chinese populations. HLA 2016;88:164-71.
Erlich HA, Opelz G, Hansen J. HLA DNA typing and transplantation. Immunity 2001;14:347-56.
Robinson J, Halliwell JA, Hayhurst JD, Flicek P, Parham P, Marsh SG. The IPD and IMGT/HLA database: Allele variant databases. Nucleic Acids Res 2015;43:D423-31.
Kempenich JH, Setterholm M, Maiers M. Haplotype associations of 90 rare alleles from the National Marrow Donor Program. Tissue Antigens 2006;67:284-9.
Maiers M, Gragert L, Klitz W. High-resolution HLA alleles and haplotypes in the United States population. Hum Immunol 2007;68:779-88.
Wen SH, Lai MJ, Yang KL. Human leukocyte antigen-A, -B, and -DRB1 haplotypes of cord blood units in the Tzu Chi Taiwan Cord Blood Bank. Hum Immunol 2008;69:430-6.
Vorechovsky I, Kralovicova J, Laycock MD, Webster AD, Marsh SG, Madrigal A,et al .
Short tandem repeat (STR) haplotypes in HLA: An integrated 50-kb STR/linkage disequilibrium/gene map between the RING3 and HLA-B genes and identification of STR haplotype diversification in the class III region. Eur J Hum Genet 2001;9:590-8.
Pédron B, Yakouben K, Guérin V, Borsali E, Auvrignon A, Landman J,et al .
HLA alleles and haplotypes in French North African immigrants. Hum Immunol 2006;67:540-50.
Chen MJ, Chu CC, Shyr MH, Lin CL, Lin PY, Yang KL. A novel HLA-B allele, B*5214, detected in a Taiwanese volunteer bone marrow donor using a sequence-based typing method. Int J Immunogenet 2010;37:39-41.
Chen MJ, Yang TC, Chu CC, Shyr MH, Lin CL, Lin PY,et al .
Detection of a novel HLA-B27 allele, B*2740, in Taiwanese volunteer bone marrow donors by sequence-based typing: Curiosity rewarded. Int J Immunogenet 2009;36:207-11.
Excoffier L, Lischer HE. Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour 2010;10:564-7.
Gragert L, Madbouly A, Freeman J, Maiers M. Six-locus high resolution HLA haplotype frequencies derived from mixed-resolution DNA typing for the entire US donor registry. Hum Immunol 2013;74:1313-20.
Li XF, Zhang X, Chen Y, Zhang KL, Liu XJ, Li JP. An analysis of HLA-A, -B, and -DRB1 allele and haplotype frequencies of 21,918 residents living in Liaoning, China. PLoS One 2014;9:e93082.
Yang KL, Chen SP, Shyr MH, Lin PY. High-resolution human leukocyte antigen (HLA) haplotypes and linkage disequilibrium of HLA-B and -C and HLA-DRB1 and -DQB1 alleles in a Taiwanese population. Hum Immunol 2009;70:269-76.
de Rham C, Villard J. Potential and limitation of HLA-based banking of human pluripotent stem cells for cell therapy. J Immunol Res 2014;2014:518135.
Taylor CJ, Peacock S, Chaudhry AN, Bradley JA, Bolton EM. Generating an iPSC bank for HLA-matched tissue transplantation based on known donor and recipient HLA types. Cell Stem Cell 2012;11:147-52.
Nakatsuji N, Nakajima F, Tokunaga K. HLA-haplotype banking and iPS cells. Nat Biotechnol 2008;26:739-40.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7], [Table 8], [Table 9]