Introduction:
Hemoglobinopathies are amongst the most common genetically inherited disorders. In India the cumulative gene frequency of hemoglobinopathies has been previously reported to be around 4.2%.[1] However, the exact magnitude of different hemoglobinopathies is obscure in India.
High performance liquid chromatography (HPLC) is the most commonly used method for detection and quantitative estimation of hemoglobin variants. It is a rapid, reproducible and precise technique for early diagnosis, prevention and treatment of hemoglobinopathies which are a major cause of morbidity and mortality in our country. Since there is no definitive cure, prevention by carrier detection is the only way to reduce the disease burden. Uses of HPLC include Hemoglobinopathy carrier screening in couples from high-risk populations, identification of abnormal neonatal hemoglobin, follow up of known cases of hemoglobinopathy, post bone marrow transplantation to document engraftment of donor hematopoiesis.[2] However, the availability and application of HPLC in India is very non uniform.
This study was done with the aim of analyzing the different findings in HPLC using D-10 analyzer and evaluating the spectrum of different hemoglobin disorders in a hospital-based population of South Delhi. Such a prevalence study would be useful to review the various strategies that can be implemented for effective control and prevention of these disorders.
Most laboratories performing HPLC use Biorad Variant-II Testing system. D-10 Analyzer which is the ideal equipment for estimation of HbA1c can also be used for Hemoglobin HPLC. Using one machine for both these tests in the laboratory is economical and convenient. There is paucity of past literature on HPLC using D-10 analyzer. This study will help in comprehensive analysis of the results of HPLC in D-10 analyzer in different hemoglobinopathies.
Materials and Methods
The study was conducted in Department of Pathology for a period of two years from March 2019 to February 2021. It was a hospital based descriptive observational study. Proper ethical clearance was obtained from the institutional ethics committee. All OPD and IPD patients who were advised HPLC during their clinical workup were included in the study. These included patients with anemia, hepatosplenomegaly, repeated infections, positive family history for hemoglobinopathy. Patients who had received blood transfusion in the past 3 months and known cases of hemoglobinopathies were excluded from the study. After taking informed consent from the patients or their guardians as the case may be, 2 ml of venous blood was collected in EDTA coated vacutainers. Complete blood count and peripheral smear examination was performed for all the cases. Reticulocyte count, sickling test and serum iron studies were performed in selected cases.
Analysis of EDTA blood samples was done by Bio Rad Dual program HPLC instrument. The D10 Dual program is based on chromatographic separation of the analytes by ion exchange HPLC. The blood is automatically diluted on the D10 and injected into the analytical cartridge. It delivers a programmed buffer gradient of increasing ionic strength to the cartridge where the different hemoglobin components are separated based on their ionic interactions with the cartridge material. The separated Hemoglobin fractions then pass through the flow cell of the filter photometer where changes in absorbance at 415 nm are measured.[3]
The exact percentage of HbA, HbA2, HbF and any other variant hemoglobin was estimated. Presumptive identification of hemoglobin variants was made primarily by their percentage, retention time (RT) and peak characteristics. HPLC findings were correlated with the clinical history, family history and the CBC and peripheral smear findings in all cases.
Results
A total of 662 patients were analyzed over a period of two years. The mean age of the patients was 21.6 ±12.9 years. Out of these, 466 patients were adults (>18 years) and 196 patients were children ≤18 years of age. Among the Paediatric patients, 25 patients were infants <1 year of age. The overall male:female ratio was 0.26:1. Majority of the samples were received from Obstetrics and Gynaecology Department (400, 60.4%). The rest of the samples were from Paediatrics (26.6%) and Medicine departments (13%). It was noted that the average hemoglobin concentration of all the patients whose HPLC was ordered during the study period was 8.4±4.7g/dl. The average MCV, MCH and MCHC was 75.8±14.7fl, 22.8±14.0 pg and 29.8±14.0g/dl respectively. Most of the patients thus had microcytic hypochromic anemia.
On HPLC analysis, 79% (523) of the patients had no abnormality detected and the report was within normal limits. The commonest hemoglobinopathy was Beta Thalassemia Trait detected in 42 patients (6.3%). It was observed that among adults >18 years of age, beta thalassemia trait was the most common hemoglobinopathy found in 6.7% of the patients and 79.7% of the patients had normal findings. In children aged 1-18 years, 77% of the patients had normal findings and beta thalassemia trait was seen in 5.9% children. In infants while 80% cases had normal findings, beta thalassemia trait was found in one case.
HbE trait was the next common hemoglobin disorder seen in a total of 7 cases (1.05%). The other hemoglobinopathies detected were HbD Punjab Heterozygous (3 cases, 0.5%), Beta thalassemia homozygous (3 cases, 0.5%), Sickle cell Heterozygous (2 cases, 0.3%), HbJ Meerut Heterozygous (2 cases, 0.3%). One case each of Sickle cell Homozygous (0.15%), Compound Heterozygous HbS/beta thalassemia trait (0.15%), HbE Homozygous (0.15%), Compound Heterozygous HbE/beta thalassemia trait (0.15%), and Homozygous delta beta thalassemia (0.15%) were also diagnosed. [Table 1]
| Table 1: Patterns of Haemoglobin seen in Paediatric and Adult population on HPLC. |
Patterns of Hemoglobin |
Children |
Adults |
Total number
(Percentage) |
<1y |
1-18y |
>18y |
Normal |
20 (80%) |
131 (77%) |
372 (79.7%) |
523 (79.0%) |
Beta thalassemia trait |
1 (4%) |
10 (5.9%) |
31 (6.7%) |
42 (6.3%) |
Beta thalassemia homozygous |
2 (8%) |
1 (0.6%) |
0 (0%) |
3 (0.5%) |
HbS Heterozygous |
0 (0%) |
0 (0%) |
2 (0.4%) |
2 (0.3%) |
HbS Homozygous |
0 (0%) |
1 (0.6%) |
0 (0%) |
1 (0.15%) |
Compound heterozygous for HbS/Beta Thalassemia Trait |
0 (0%) |
1 (0.6%) |
0 (0%) |
1 (0.15%) |
HbE Heterozygous |
0 (0%) |
3 (1.8%) |
4 (0.9%) |
7 (1.05%) |
HbE Homozygous |
0 (0%) |
0 (0%) |
1 (0.2%) |
1 (0.15%) |
Compound Heterozygous HbE/beta thalassemia trait |
0 (0%) |
0 (0%) |
1 (0.2%) |
1 (0.15%) |
HbD Punjab Heterozygous |
0 (0%) |
0 (0%) |
3 (0.6%) |
3 (0.5%) |
Homozygous Delta beta thalassemia |
0 (0%) |
0 (0%) |
1 (0.2%) |
1 (0.15%) |
Hb J Meerut Heterozygous |
0 (0%) |
1 (0.6%) |
1 (0.2%) |
2 (0.3%) |
Borderline HbA2 |
2 (8%) |
6 (3.5%) |
27 (5.8%) |
35 (5.3%) |
Reduced HbA2 |
0 (0%) |
4 (2.3%) |
9 (1.9%) |
13 (2%) |
Raised HbF |
0 (0%) |
12 (7.1%) |
15 (3.2%) |
27 (4%) |
Total |
25 |
170 |
467 |
662 |
In 65 cases no definitive diagnosis on HPLC could be given. In 27 of these cases, HbF was mildly raised. The average HbF in these cases was 2.5%. Ten of these patients were antenatal patients, twelve cases were from Paediatrics and five were from Medicine. In 35 cases, HPLC was reported as borderline HbA2 values (average HbA2=3.7%) and serum iron studies and follow up after a course of hematinics was advised. Thirteen cases (2%) had reduced HbA2 levels (average HbA2=1.9%). [Table 1]
All patients with normal findings on HPLC showed average HbA of 83.5%, HbA2 of 2.9% and HbF of 0.9%. The average MCV, MCH, MCHC in this group was 75.5 fl, 22.7 pg and 29.8 g/dl respectively. In thalassemia trait cases, the average HbA, HbA2 and HbF were 80.1%, 5.3% and 1.8% respectively. The average MCV, MCH, MCHC in this group was 65.7 fl, 19.5 pg and 29.7 g/dl respectively. In all three cases reported as thalassemia major, HbF was markedly raised and HbA was low. These patients were transfusion dependent and peripheral smear findings of these cases showed characteristic leucoerythroblastic blood picture with microcytic hypochromic anemia.
Four cases of sickle cell disease were reported, out of which one case was homozygous HbS with HbA< HbS and normal HbA2 levels, and normocytic normochromic red blood cell indices. One case was compound heterozygous HbS/beta thalassemia trait and raised HbA2 levels with microcytic hypochromic red blood cell indices. Two cases of sickle cell trait with HbA> HbS were diagnosed. Sickling test was done in all cases of sickle cell disease for collaborating the findings.
Nine cases of HbE syndrome were diagnosed on HPLC. HbE elutes with HbA2 on D-10 analyzer. One case of HbE homozygous had HbA2 levels of 76.7%. One case of compound heterozygous HbE/beta thalassemia trait was diagnosed because of characteristic CBC and peripheral smear findings and raised HbA2 of 37.6% as well as raised HbF of 41.1%. Seven cases of HbE trait were noted with mean MCV, MCH and MCHC of 77.1fl, 24.9 pg and 32.2 g/dl respectively.
Three cases of heterozygous HbD Punjab were diagnosed having a peak at retention time of 4 min with average value of 30.4%. Two cases of HbJ Meerut were reported with peak at retention time of 1.44 min with average value of 17.2%. One case of delta beta thalassemia was reported with markedly raised HbF of 97.8% and zero HbA2 levels. This patient had normocytic normochromic red blood cell indices and had not received any blood transfusion till date.
The cases with reduced HbA2 had average MCV, MCH, MCHC and RDW of 68.3 fl, 19.7 pg, 28.3 g/dl and 19.6% respectively. Those with borderline HbA2 had mean values of 89.2 fl, 27.4 pg, 30.2 g/dl and 19.2% respectively.
The average levels of different hemoglobin fractions and the red blood cell parameters in different hemoglobinopathies have been summarized in Table 2 and 3.
| Table 2: Average levels of various hemoglobin subtypes on HPLC in different hemoglobinopathies. |
| Diagnosis |
Total cases |
HbA %
(average) |
HbA2 %
(average) |
HbF %
(average) |
Abnormal Hb % (average) |
Normal |
523 |
83.5 |
2.9 |
0.9 |
- |
Beta thalassemia trait |
42 |
80.1 |
5.3 |
1.8 |
- |
Beta thalassemia homozygous |
3 |
10.0 |
3.0 |
66.0 |
- |
HbS Heterozygous |
2 |
61.8 |
4.4 |
1.0 |
24.2 (HbS) |
HbS Homozygous |
1 |
10.5 |
3.4 |
17.1 |
64.9 (HbS) |
Compound heterozygous for HbS/Beta Thalassemia Trait |
1 |
25.3 |
6 |
7.9 |
54.2 (HbS) |
HbE Heterozygous |
7 |
65.2 |
25.0 |
1.7 |
- |
HbE Homozygous |
1 |
6.7 |
76.7 |
11.6 |
- |
Compound Heterozygous HbE/beta thalassemia trait |
1 |
10 |
37.6 |
41.1 |
- |
HbD Punjab Heterozygous |
3 |
57.3 |
2.5 |
0.8 |
30.4
(RT: 4min) |
Homozygous Delta beta thalassemia |
1 |
0.9 |
0 |
97.8 |
- |
Hb J Meerut Heterozygous |
2 |
68.9 |
2.2 |
1.1 |
17.2 (RT:1.44min) |
Borderline HbA2 |
35 |
83.7 |
3.7 |
0.9 |
- |
Reduced HbA2 |
13 |
83.3 |
1.9 |
0.9 |
- |
Raised HbF |
27 |
81.4 |
3.0 |
2.5 |
- |
| Table 3: Mean values of RBC parameters in various hemoglobinopathies. |
| Patterns of Hemoglobin |
Total number |
Hb
(average) |
MCV
(average) |
MCH
(average) |
MCHC
(average) |
RDW
(average) |
Normal |
523 |
8.2 |
75.5 |
22.7 |
29.8 |
19.9 |
Beta thalassemia trait |
42 |
8.8 |
65.7 |
19.5 |
29.7 |
19.8 |
Beta thalassemia homozygous |
3 |
5.6 |
69.4 |
20.5 |
29.6 |
33.8 |
HbS Heterozygous |
2 |
8.7 |
83.6 |
26.8 |
32.2 |
14.5 |
HbS Homozygous |
1 |
7.9 |
84.2 |
29 |
34.5 |
22.8 |
Compound heterozygous for HbS/Beta Thalassemia Trait |
1 |
6.3 |
74.6 |
21.6 |
29 |
19.7 |
HbE Heterozygous |
7 |
10.1 |
77.1 |
24.9 |
32.2 |
17.1 |
HbE Homozygous |
1 |
8.5 |
73.5 |
23.9 |
32.6 |
19.4 |
Compound Heterozygous HbE/beta thalassemia trait |
1 |
5.9 |
62.1 |
16.8 |
27.1 |
32.9 |
HbD Punjab Heterozygous |
3 |
7.2 |
75.3 |
21.4 |
27.9 |
21.9 |
Homozygous Delta beta thalassemia |
1 |
5.6 |
94 |
23.9 |
25.5 |
28.6 |
Hb J Meerut Heterozygous |
2 |
10.9 |
74.1 |
23 |
30.7 |
17.7 |
Borderline HbA2 |
35 |
9.1 |
89.2 |
27.4 |
30.2 |
19.2 |
Reduced HbA2 |
13 |
7.9 |
68.3 |
19.7 |
28.3 |
19.6 |
Raised HbF |
27 |
11.6 |
85.0 |
25.3 |
29.6 |
20.3 |
Discussion
Detection of hemoglobinopathies is possible by a number of techniques like hemoglobin electrophoresis, HPLC, isoelectric focusing, capillary gel electrophoresis and molecular analysis.[4] With HPLC, quantification of different hemoglobin subtypes is possible in a single and highly reproducible system. It is an easy to perform test and can replace laborious procedures like estimation of HbF and HbA2. It is an ideal method for the routine clinical laboratory because of advantages like internal sample preparation, superior resolution, rapid assay time and accurate quantification of hemoglobin fractions.[5]
There are several indications for investigation of hemoglobinopathies and HPLC including clinical suspicion of thalassemia syndromes and sickle cell disease, carrier screening in couples from high-risk populations, confirmation and follow up of an abnormal neonatal screening result, investigation of family members of known cases of hemoglobinopathy, preoperative screening for HbS in high-risk populations, laboratory evidence of hemoglobinopathy[6].
Biorad Variant-II Testing system is considered to be the gold standard for detection of hemoglobinopathies., D-10 analyzer is commonly used to assess the HbA1c fraction in the blood however, the results of Variant-II testing system and D10 analyzer have been found to show good correlation.[7] D10 analyzer is in fact, an effective and cheaper alternative to Variant -II testing system.[7] The same machine can be used to perform HPLC as well as HbA1c in the lab. Often an abnormal HbA1c result is the first indication of an underlying hemoglobinopathy.[8] This is because hemoglobin variants often interfere with the quantification of HbA1c. High levels of HbF present as elevated HbA1c or elevated LA1c.[9]
Hemoglobin separates into major and minor hemoglobin fractions during HPLC. The order of elution of various components on the D-10 analyzer is HbA1a, HbA1b, HbF, LA1c/CHb-1, LA1c/CHb-2, HbA1c, P3, HbA and HbA2. The minor hemoglobin fractions A1a, A1b, A1c, F1, P3 components are post translational modification of the globin chains.[9]
Out of 662 patients referred to our laboratory, 21% of the cases had abnormal hemoglobin fractions. The commonest hemoglobinopathy in our study were beta thalassemia followed by HbE disorders. A similar large-scale study was conducted on D-10 analyzer in South India over a period of 1 month. The commonest disorder encountered was beta thalassemia trait followed by HbE trait.[9] In another study from Odisha, D-10 analyzer was used and it was found that HbS disease was the most common hemoglobinopathy followed by beta thalassemia. HbE was also found to be endemic in Odisha.[10] Saxena et al studied the burden of hemoglobin disorders in pediatric population of Gujarat using D-10 analyzer. It was found that sickle cell trait/anemia and beta thalassemia were the commonest hemoglobin disorders.[11] Beta thalassemia trait and sickle cell disorders were also the commonest disorders among antenatal women and premarital men and women in a screening study of West Bengal.[12] Dangi et al evaluated the prevalence of sickle cell disorders in central India by utilizing the D-10 analyzer.[13]
Beta thalassemia trait characteristically shows raised HbA2 levels with values ranging from 4-9%. Methods to estimate the HbA2 level in the blood like cellulose acetate electrophoresis followed by elution and microcolumn chromatography are lengthy, delicate, labor intensive and prone to methodological errors. Accuracy of these methods is dependent on stringent quality assurance programs.[14] Automated chromatography by HPLC enables rapid and accurate HbA2 estimation.[14] Values of HbA2 between 3.5-4% were labelled as “borderline HbA2 levels”. Nutritional iron deficiency affects the results of HPLC and is the major cause of borderline HbA2 values. While the average red blood cell indices did not point to iron deficiency in this study, subclinical iron deficiency should be ruled out in such cases since paucity of iron reduces the HbA2 levels and may mask the presence of underlying thalassemia trait. In such cases, serum iron studies, hematinic therapy and follow up should be advised. Repeat HPLC should be performed in cases which have persistent anemia. Cases with reduced HbA2 levels <2.2% had average MCV, MCH, MCHC in the microcytic hypochromic range and raised RDW. Peripheral smear findings were suggestive of microcytic hypochromic anemia pointing towards iron deficiency. In all such cases, serum iron studies and follow up after a course of hematinics was advised. Correction of iron deficiency is necessary for accurate estimation of HbA2 levels and thus interpretation of HPLC.
While, HbA2 cannot be separated from Hb C, Hb E, Hb O Arab by conventional methods, in HPLC, Hb C and Hb O Arab have different retention times. While Hb E coelutes with HbA2, when levels are between 20-30%, it is likely to be Hb E in heterozygous state. HbA2 levels of 60-70% are suggestive of homozygous disorders. In compound heterozygous state, the chromatogram characteristically shows raised HbA2 as well as HbF levels as seen in one case in this study. However, accurate estimation of HbA2 levels cannot be done in HbE disorders.[9]
Hb D Iran also co elutes with HbA2 but usually in heterozygous state values are 30-40%. No case of HbD Iran was found in the current study. Hb D Punjab appears as an unknown window between S and C. It is important not to confuse the two Hb D variants because of their different clinical presentations. Hb D Punjab produces significant sickling disorder when present in double heterozygous Hb S/Hb D form while Hb D Iran is clinically benign. These variants have identical electrophoretic mobilities in conventional electrophoresis, however, on HPLC distinct pictures are seen on the chromatogram.[2]
Hb E, Hb S and Hb D disorders are common in certain specific ethnic populations. Hb E gene frequency in north eastern regions of India has been reported to be 10.9%.[15] Prevalence of sickle cell disorders among tribes of Orissa varies from 2.4% to 5.6%,[16] while it is reported to be 5.7% among children in central India.[17] Frequency of Hb D in Uttar Pradesh is around 0.5-3.1%.[18] In this study done in tertiary care centre of South Delhi, the frequency of these disorders was 1.4%, 0.6% and 0.5% respectively for HbE, HbS and HbD Punjab respectively. The superiority of HPLC to conventional electrophoresis is that Hb D Punjab and Hb S have different retention times and are separated at different windows.
Cases with raised Hb F include homozygous beta thalassemia patients, double heterozygotes for beta thalassemia and other hemoglobin variants. In our study markedly raised HbF was seen in 3 cases of beta thalassemia homozygous and one case of delta beta thalassemia. The diagnosis was given after correlation with clinical presentation, complete blood count and peripheral smear findings. Few cases (27) were also found to have mildly raised fetal hemoglobin. Ten of these patients were from Antenatal OPDs in which fetal hemoglobin gets elevated. Rest of the patients were markedly anemic, and rise in HbF in these patients could be explained by stress erythropoiesis.
Carriers of alpha thalassemia are difficult to identify than those of beta thalassemia because they do not show typical Hb A2 and Hb F levels. In presence of Hemoglobin variant with retention time of <1 min, Hb H disease should be ruled out. No such case was seen in the current study.
This study describes the spectrum of different hemoglobinopathies in a hospital catering to South Delhi population. Delhi being a cosmopolitan city, the patients belong to varied religious groups and socioeconomic strata. They form an assorted group of Hindus, Muslims and Sikhs. Many of them are immigrants from different parts of the country. The prevalence reported may not reflect the true prevalence of the actual Delhiite population. More number of female patients in our study is because major contribution of samples received for HPLC in the laboratory are from antenatal clinics for their routine carrier detection.
Endogamy and consanguinity are common practices in India which poses a risk for homozygous inheritance of hemoglobinopathies. This hospital caters to a population with low socioeconomic status, where consanguinity is a common practice. However, the homozygous disorders were only 0.9% (6 cases) in our study.
Conclusion
This study gives an important insight to the present day scenario of hemoglobinopathies in patients in South Delhi in relation to the hematological profile. It highlights the chromatogram findings of different hemoglobinopathies on the D10 analyzer. The main limitation of the study was inability to perform family studies and follow up of cases which were referred elsewhere for confirmatory tests. In addition, due to economic constraints of the patients, HPLC findings were not corroborated with the serum iron studies.
The comprehensive data obtained by such series can help in the formulation and development of infrastructure and policies for hemoglobinopathy prevention, diagnosis and management. Screening for thalassemia and other hemoglobinopathies should become an intrinsic part of our healthcare system.
References
- Balgir RS. The genetic burden of hemoglobinopathies with special reference to community health in India and the challenges ahead. Indian J Hematol Blood Trans 2002;20:2-7.
- Hosseini S, Kalantar E, Dorgalaleh A. A long term screening of Iranian Populations with Thalassemia and Hemoglobinopathies. British Biomedical Bulletin 2014;2(4):669-76.
- Sood SK, Bhargava M, Shirish, Roshan C, Chandra S, Renu S. Underlying principles of D-10TM dual mode.
- Wild BJ, Bain BJ. Investigation of abnormal hemoglobins and thalassemia. In Lewis SM, Bain BJ, Bates I, eds. Dacie and Lewis Practical Hematology. 9th ed. Churchill Livingstone;2001:231-68.
- Joutovsky A, Hadzi-Nesic J, Nardi MA. HPLC retention time as a diagnostic tool for hemglobin variants and hemoglobinopathies: a study of 60000 samples in a clinical diagnostic laboratory. Clin Chem 2004;50(10):1736-47.
- Hoffman H, Benz E, Silberstein L. Hematology. Basic Principles and practice. 2005
- Roy DS, Bhattacharyya R, Mukhpadhyay K, Bandopadhyay D. Evaulation of efficacy of BIORAD D10TM testing system in detection of beta thalassemia carrier. International Journal of Research in Medical Sciences 2019;7:782-5.
- Shrestha A, Pant V, Gautam K, Pyakurel D, Pradhan S. Detection of Hemoglobinopathies by HPLC in a referral Clinical Laboratory in Nepal. Nepalese Medical Journal 2020;3:306-8.
- Chandrashekar V, Soni M. Hemoglobin Disorders in South India. ISRN Hematology 2011;1-6.
- Alam S, Singh A, Chakrabarty S, Mohanty R. Spectrum of hemoglobinopathies in Odisha – an institutional study by CE-HPLC. International Journal of Medical Science and Public Health 2016;5:208-11.
- Saxena S, Jain R. Study of abnormal hemoglobin variants using cation exchange high performance liquid chromatography (HPLC) in paediatric population of Gujarat, India. Tropical Journal of Pathology and Microbiology 2019;5:856-62.
- Chattopadhyay P, Kundu S, Saha TN, Chatterjee K. Study of Hemoglobinopathies at a Referral Laboratory in a Western District among the Antenatal Women and Premarital Men and Women: A 2 years Study. International Journal of Contemporary Medical Research 2019;6:F24-7.
- Dangi CBS, Sajid M, Sawke GK, Ambhore J. Sickle cell hemoglobinopathies in district Bhopal. Indian J Hum Genet 2010;16(2):100-2.
- George E, Jamal AR, Khalid F, Osman KA. High performance liquid chromatography (HPLC) as a screening tool for classical beta-thalassemia. Malaysian Journal of Medical Sciences 2001;8:40-6.
- Balgir RS. Genetic epidemiology of the three predominant abnormal hemoglobins in India. Journal of Association of Physicians of India 1996;44:25-8.
- Balgir RS. The spectrum of hemoglobin variants in two scheduled tribes of Sundargarh district in north western Orissa, India. Annals of Human Biology 2005;32:560-73.
- Kamble M, Chaturvedi P. Epidemiology of sickle cell disease in a rural hospital of Central India. Indian Pediatrics 2000;37:391-6.
- Agarwal S, Gupta UR, Kohli N, Verma C, Agarwal SS. Prevalence of hemoglobin D in Uttar Pradesh. Indian Journal of Medical Research B;1989;90:39-43.
|
