Review of data from the Hawai’i hereditary anemia project

Hematologic parameters in thalassemia: Review of data from the Hawai’i hereditary anemia project

Hsia, Y Edward

OBJECTIVE: Describe changes in hematologic parameters associated with thalassemia.

DESIGN: Retrospective analyses of data from the Hawai’i Hereditary Anemia Project.

PARTICIPANTS: Laotian, Thai, Filipino, Indonesian, and Chinese subjects suspected to have thalassemia and their family members.

MAIN OUTCOME MEASURES: Median and 95% confidence intervals of CBC parameters (RBC, Hb, Hct, MCV, MCH, MCHC, RDW), ZPP/FEP, and HbA.2 for control, DNA confirmed at-thalassemias, P-thalassemia, and iron deficiency groups.

RESULTS: Abnormalities in CBC parameters in ot-thalassemias became progressively more pronounced as more a-globin genes were deleted. RBQ RDW, and ZPP/FEP successfiffly distinguished thalassemia from iron deficiency. HbA2 was elevated in P-thalassemia but depressed in HbH disease. CBC results in HbE disease were similar to cc-thalassemia, but tended to be a little more extreme.

CONCLUSION: Characteristic patterns of hematologic parameters are seen in each type of thalassemia heterozygotes. They can also be used to distinguish thalassemia from other causes of microcytosis.

ABBREVIATIONS: CBC = complete blood count; FEP = free erythrocyte protoporphyrin; Hb = hemoglobin; HbA2 = hemoglobin A2; HbE = hemoglobin E; HbH = hemoglobin H; Hct = hematocrit; IEF = isoelectric focusing; MCH = mean corpuscular hemoglobin; MCHC = mean corpuscular hemoglobin concentration; MCV = mean corpuscular volume; RBC = red blood cell count; RDW = red cell distribution width; ZPP = zinc protoporphyrin.

INDEX TERMS: hemoglobinopathy; microcytosis; thalassemia.

Clin Lab Sci 1999;12(3):169

Thalassemia is prevalent in Hawai’i due to a large in-migrating population from Southeast Asia and other regions where it is endemic. The potential for serious inherited conditions makes early detection, particularly during pregnancy, important.

The complete blood count (CBQ is one of the first tests performed to detect any hematologic condition. A number of hematologic parameters are altered in thalassemia, but mild forms of thalassemia show only subtle changes. Furthermore, results from other causes of microcytosis may mimic those of thalassemia.

Once a thalassemia is suspected, specific tests can be used to differentiate it from other forms of microcytosis, separate P from a-thalassemia or identify the molecular defect. Although rarely needed for clinical diagnosis, definitive diagnosis by DNA analysis is important to determine the risk for parents to have a child inheriting a serious or potentially fatal combination.’

This paper summarizes the CBC data, zinc protoporphyrin, free erythrocyte protoporphyrin, and hemoglobin A2 levels from the Hawai’i Hereditary Anemia Project. Analyses of erythrocyte morphology and HbH inclusion bodies were reported earlier.2,3


The Hawai’i Hereditary Anemia Project was conducted between 1985 and 1994, targeting specific ethnic groups: Laotian, Thai, Filipino, Indonesian, and Chinese. All subjects from these groups, especially if thalassemia was suspected, were tested along with their family members. Data presented here are for subjects over ten years of age who had the diagnosis of Oc-thalassemia confirmed or ruled out by DNA analysis.

A CBC was run on each subject. The CBC parameters included red blood cell count (RBC), hemoglobin (Hb), hematocrit (Hct),

MCV, MCH, MCHC and RDW. All microcytic samples, or those suspected to have a hematologic condition from family data, were further tested for zinc protoporphyrin (ZPP) or free erythrocyte protoporphyrin level (FEP) by fluorometry to rule out iron deficiency. FEP was used until 1986; after 1986 samples were tested with ZPR Hemoglobin A2 (HbA2) by micro-column chromatography and HbE by isoelectric focusing (IEF) electrophoresis were done to detect P-thalassemia and HbE disease. Alpha-globin gene abnormalities were sought using DNA analyses by Southern blotting or polymerase chain reaction techniques.”

The median and 95% confidence intervals were calculated using Foxbase algorithms. Data were stored in an X-base format. The diagnosis categories for tabulation were: control group with normal ot-globin genes without iron deficiency or P-thalassemia, ox-thal-2 heterozygote (-oc/m), cc-thal-I heterozygote (–/oct), a-thal-2 homozygote (-(XI-a), HbH disease (–/-or), -thalassemia, and iron deficiency without ox-globin gene changes. Those who had HbE by IEF but no a-globin gene changes were also examined; they were designated HbAE for heterozygotes and HbEE for homozygotes.


Figures I through 9 compare the medians and 95% confidence intervals among groups for each of the parameters. Table 1 summarizes the findings.

The results for the (-a/(xa) group barely deviated from those of the control group. The degree of abnormalities in the a-thalassemia groups corresponded to the severity of the condition, i.e., number of (X-globin gene deleted. Results between (-a/-a) and (–/(x(x) were not distinguishable.

Hb, Hct, MCN@ MCH, and MCHC medians for cc-thalassemias, P-thalassemia, and iron deficiency were all lower than the control group.

The RBC medians for all thalassemias were higher than the control group. The RBC median was decreased in iron deficiency.

An MCH of 25 pg was a clear cut off point separating (–/(xt), (-oc/-cc), HbH disease, and P-thalassemia from the control group. Even for (-a/ltot), the MCH was always less than 30 pg, which better discriminated this group from the control group than MCV. Median RDWs in all thalassemia groups were elevated, especially for HbH disease. There was no overlap of RDW ranges between HbH and the control group. The RDW range in iron deficiency was large.

The ranges of ZPP and FEP results were comparable. The levels were increased in iron deficiency, while other groups had levels close to those of the control group.

HbA2 was significantly increased in P-thalassemia, with the lower limit just overlapping the upper limit of the control group. HbA2 was reduced in HbH disease. It was low-normal in iron deficiency.

Phenotypically, 126 subjects were identified as HbAE heterozygotes and 22 as HbEE homozygotes, approximately 3.6% and 0.6% of the sample population, respectively. Most of these were Laotian, some with a coexisting (x-thalassemia. The CBC results of HbA.E heterozygotes with or without oc-thalassemia were similar to (-ot/tx), and HbEE homozygotes with or without (x-thalassemia were similar to (–/xx) or (-a/-a). Since HbE elutes along with HbA2 on column chromatography, levels of apparent median HbA2 were 25.4% for HbAE heterozygotes and close to 100% for HbEE homozygotes.


The hematologic parameters of various groups, although skewed by the effects of some atypical data, were compared using the medians and 95% confidence intervals. The median represents the typical value for the group and the intervals describe the population distribution.

Among our targeted subjects in the project, including those that had not been confirmed genotypically, oc-thalassemia was approximately three times more prevalent than 0-thalassemia. Genotypic (–/Oc(x) was over 30 times more prevalent than (-Oc/-(X). HbE was almost as prevalent as (-(tx/(xx). The project also confirmed the rarity of thalassemias in Koreans and Japanese, while Hawaiians and other Polynesians were only likely to have or-thal-2 single a-globin gene deletions.

Changes in CBC parameters, ZPP/FEP, and HbA2 levels in our population were as expected.’ This confirmed the validity of using these parameters in provisional phenotypic identification of thalassemia heterozygotes. The degree of CBC abnormalities in Ot-thalassemia became progressively more pronounced as more ot-globin genes were deleted. The results from the (-(x/(xa) group were slightly different but nearly indistinguishable from the control group; this is consistent with the characterization of this condition as the “silent” carrier. Results for (-ot/-oc) and (–/(otoc) were alike so that the two conditions cannot be distinguished based only on hematologic parameters. Offspring of couples who both have (–I(xa) are at risk for fatal hydrops fetalis Thus determination of (–/aoc) genotype with DNA analysis is essential in family planning decisions, especially among Southeast Asians.

Iron deficiency and thalassemia groups all showed microcytic hypochromic changes, but the ZPP/FEP levels were high in iron deficiency@ This is because cytoplasmic protoporphyrins accumulate due to lack of iron to complete the hemoglobin synthesis.

Changes in RBC, Hb, Hct, and RDW were distinct between iron deficiency and the milder (X-thalassemias: (-(X/(X(–Iota), and (-aX/-aX). Iron deficiency had lower RBC, Hb, and Hct values but higher and more variable RDW. CBC results were similar between HbAE heterozygotes and (-Wax/lta), and between HbEE homozygotes and (–/aa) or HbA2 level was increased in P-thalassemia due to reduced P-globin production, resulting in more (X282 tetramers. Our data also showed suppressed HbA2 in (–/-(x), presumably due to reduced availability of tx-globin peptides for tetramer formation. Because HbE elutes along with HbA2, the results of microcolumn chromatography for HbA2 are always very high in HbE, masking the true level of HbA2. Apparent HbE could mask a coexisting (x-thalassemia or P-thalassemia, which could produce a more severe hemolytic anemia than either condition alone. Since HbE is seen in a population where CE-thalassemia is also prevalent, e.g., Southeast Asians, all apparent HbAE or HbEE samples should be evaluated for ot-thalassemia.

Consistent changes in CBC parameters can be used to distinguish the various causes of microcytosis. Discriminant function formulae have been developed using these parameters.6 Although they did not perform reliably or consistently in our studies, if such formulas are made a part of computerized diagnostic program in hematological analyzers, the MCH should be used as a cut-off from normal, and the RDW plus ZPP to rule out iron deficiency. The MCHC was of little value since it was abnormal only when other parameters were clearly abnormal. It should be noted that, an apparatus for ZPP measurement can be attached to an analyzer. Diagnosis of thalassemia is often complicated by a coexisting iron deficiency or co-inheritance of both a and P-thalassemias or other hemoglobinopathies like HbE. More selective tests, such as the ZPP/FEP, HbA2, and IEF, are needed to identify these conditions. Furthermore, other inexpensive tests can help with diagnosis of thalassemia. For example, characteristic changes in the microscopic morphology are seen in thalassemia, and HbH inclusion bodies are almost exclusively found in with rare inclusion bodies in (–/act) or (-a/-a). Alert attention to red cell parameters can be very effective for identifying samples from thalassemic patients.


We are grateful to Dr. John Hunt for his participation in this project, Ms. Berbie Chu for her invaluable supervision of all the laboratory investigations, and others who collaborated.

Support for this research included a grant from the Howard Hughes Medical Institute through the Undergraduate Biological Sciences Education Program, the Hawaii Heart Foundation, USPHS Bureau of Community Health Services SPRANS grant, and from Kapi’olani Medical Center for Women and Children.


1. Hsia YE. Detection and prevention of important (t-thalassemia variants. Semin Perinatol 1991;15(l suppl 1):35-42.

2. Teshima DY, Hall J, Darniati E, Hsia YE. Microscopic erythrocyte morphology changes associated with oc-thalassemias. Clin Lab Sci 1993;6:236-40.

3. Miyakawa FH, Teshima DY, Hsia YE. Erythrocyte Hb H inclusion bodies in (x-thalassemia. Clin Lab Sci 1994;7:272-4.

4. Harada E Ireland JH, Hsia YE, Choi DH. Anti-@ antibody screening for xthalassemia using dried filter paper blood. Biochem Med Metab Biol 1994;5:80-4.

5. McKenzie SB. Textbook of Hematology. 2nd ed. Baltimore: Williams & Wilkins 1996:175.

6. LafertyjD, Crowther MA, Ali MA, Levine M. The evaluation ofvarious mathematical RBC indices and their efficacy in discriminating between and thalassemia and non-thalassemic microcytosis. Am J Clin Path 1996; 106:201-5.

Y Edward Hsia BM BCh FRCP(London) is Professor of Pediatrics and Professor of Genetics and Molecular Biology at the University of Hawai’i at Manoa, Honolulu, HI.

Dick Y Teshima MPH is an Associate Professor at the University of Hawaii at Manoa, Honolulu HL

Address for correspondence. Dick Y Teshima MPH, University of Hawaii at Manoa, John A Burns School of Medicine, Division of Medical Technology 1960 East West Road, Biomed C-206, Honolulu, HI 96822, (808) 956-8557, (808) 956-5506 (fax).

The peer reviewed Research section seeks to publish reports ofor@ginal research related to the clinical Laboratory or one or more subspecialties. Direct all inquiries to Virginia R Kotlarz PhD, Peer Review Editor, Clinical Laboratory Science Editorial Office, PO Box 5399, Coralville, IA 52241-5399. (319) 351-2922, (319) 351-2927 (fax). cls@ia. net

Copyright American Society for Clinical Laboratory Science May/Jun 1999

Provided by ProQuest Information and Learning Company. All rights Reserved