To the editor:

Hemoglobin Barts hydrops fetalis (homozygous α0-thalassemia) results from deletion of all 4 α-globin genes.1  It was previously considered a universally fatal condition; however, with recent advances in prenatal care and the availability of intrauterine blood transfusions, an increasing number of patients are now surviving into adulthood.2  Similar to patients with transfusion-dependent thalassemia due to β-thalassemia (TDT-β), these patients require lifelong and regular transfusions, as the option of curative stem cell transplant may not be available for the majority of these patients. Although the Thalassemia International Federation guidelines recommend a transfusion strategy similar to TDT-β for these patients,3  no report exists on the optimal transfusion management of patients with homozygous α0-thalassemia, which could be referred to as transfusion-dependent thalassemia due to α-thalassemia (TDT-α). In TDT-β, initiation of transfusions results in improvement of anemia and suppression of ineffective erythropoiesis, the main underlying pathophysiologic processes.

In early 2014, we critically reviewed the treatment strategy for four patients with homozygous α0-thalassemia at our institution who were previously on regular blood transfusions to keep their hemoglobin levels >100 g/L. Research was approved by the institutional review board, and patients gave consent in accordance with the Declaration of Helsinki. All patients had significant splenomegaly, progressive peripheral blood reticulocytosis, and biochemical markers of hemolysis (high lactate dehydrogenase [LDH], aspartate aminotransferase, indirect bilirubin), erythropoiesis (elevated soluble transferrin receptor [STR]), and tissue hypoxia (high serum erythropoietin). In addition, 3 of the 4 showed brain magnetic resonance imaging changes in keeping with “silent” ischemic infarcts (Table 1; Figure 1).

Table 1

Patient characteristics

PatientPatient 1Patient 2Patient 3Patient 4
Age at the time of implementation of new transfusion strategy (years) 16 15 14 
Sex Male Female Male Male 
α Globin gene deletion --SEA/--SEA --SEA/--SEA --SEA/--SEA --SEA/--SEA 
Proportion of Hb H before implementation of new transfusion strategy (%)* 24 64 49 44 
Calculated functional hemoglobin before new transfusion strategy (g/L) 79.0 42.7 52.0 59.3 
Endocrine abnormalities None Diabetes mellitus Hypogonadism Hypogonadism 
Delayed puberty Hypothyroidism Growth hormone deficiency 
Hypothyroidism Growth hormone deficiency  
Bone disease (lumbar Z-score <2.0) Yes Yes Yes Yes 
Short stature (height Z-score <2.0) Yes Yes Yes Yes 
MRI changes compatible with silent ischemic infarct Yes No Yes Yes 
Other complications None identified Gout None identified Gall stones 
Systemic lupus erythematosus 
Bleeding tendency 
Spleen size (cm)     
 Before intervention 12.0 19.1 16.0 16.7 
 One year after intervention 10.2 (normal for age) 17.8 14.1 15.2 
Serum erythropoetin level (mU/mL)     
 Before intervention 252.0 421.0 709.0 68.5 
 One year after intervention 46.3 93.4 31.4 19.6 
PatientPatient 1Patient 2Patient 3Patient 4
Age at the time of implementation of new transfusion strategy (years) 16 15 14 
Sex Male Female Male Male 
α Globin gene deletion --SEA/--SEA --SEA/--SEA --SEA/--SEA --SEA/--SEA 
Proportion of Hb H before implementation of new transfusion strategy (%)* 24 64 49 44 
Calculated functional hemoglobin before new transfusion strategy (g/L) 79.0 42.7 52.0 59.3 
Endocrine abnormalities None Diabetes mellitus Hypogonadism Hypogonadism 
Delayed puberty Hypothyroidism Growth hormone deficiency 
Hypothyroidism Growth hormone deficiency  
Bone disease (lumbar Z-score <2.0) Yes Yes Yes Yes 
Short stature (height Z-score <2.0) Yes Yes Yes Yes 
MRI changes compatible with silent ischemic infarct Yes No Yes Yes 
Other complications None identified Gout None identified Gall stones 
Systemic lupus erythematosus 
Bleeding tendency 
Spleen size (cm)     
 Before intervention 12.0 19.1 16.0 16.7 
 One year after intervention 10.2 (normal for age) 17.8 14.1 15.2 
Serum erythropoetin level (mU/mL)     
 Before intervention 252.0 421.0 709.0 68.5 
 One year after intervention 46.3 93.4 31.4 19.6 

Characteristics of 4 homozygous α0-thalassemia patients who are being followed in our institution. All patients provided consent for publication of this report. One patient (data not shown) was born after the implementation of new transfusion strategy and has been transfused based on the new regimen since birth.

MRI, magnetic resonance imaging.

*

Patients’ hemoglobin analysis was performed via high-performance liquid chromatography and confirmed using capillary zone electrophoresis.

Spleen size was measured on ultrasonography.

Figure 1

Effects of optimized transfusion therapy on hematologic and biochemical profiles of patients. (A) The trend in the pretransfusion reticulocyte count (%) in 4 patients with homozygous α0-thalassemia on standard transfusion therapy before commencement of new transfusion regimen. The dashed line represents LOESS fit line with 95% confidence bands. (B) The trend in the pretransfusion Hb H level in peripheral blood. The dotted line represents the 15% target. (C-F) The trends in pretransfusion unconjugated bilirubin, STR, LDH, and reticulocyte count. Shaded areas indicate laboratory results before the implementation of new transfusion regimen (day 0). The dotted lines represent upper boundaries of normal values. We used repeated-measure analysis of variance to examine the association between intervention and biochemical or hematologic outcomes. Overall, intervention effect was significant for all laboratory outcomes. Moreover, the effects of intervention on improvement of outcomes remained persistent for each of the postintervention time brackets (0-120, 121-240, and 241-365 days). Eight to 12 months after the intervention, the average lactate dehydrogenase (964.37 [standard deviation [SD]: 239.85] vs 2736.75 [SD: 242.44]), indirect bilirubin (22.03 [SD: 5.53] vs 57.38 [SD: 10.45]), soluble transferrin receptor (2.15 [SD: 0.58] vs 7.31 [SD: 0.91]), and reticulocyte count (354.22 [SD: 36.42] vs 631.87 [SD: 82.54]) were all significantly lower compared with preintervention. (G-H) Associations of “functional” hemoglobin with STR and Hb H level with LDH. Dotted lines represent target values for functional hemoglobin and Hb H%, as well as the upper boundaries of normal values for LDH and STR. (See supplemental Materials available on the Blood Web site for the details of statistical analysis).

Figure 1

Effects of optimized transfusion therapy on hematologic and biochemical profiles of patients. (A) The trend in the pretransfusion reticulocyte count (%) in 4 patients with homozygous α0-thalassemia on standard transfusion therapy before commencement of new transfusion regimen. The dashed line represents LOESS fit line with 95% confidence bands. (B) The trend in the pretransfusion Hb H level in peripheral blood. The dotted line represents the 15% target. (C-F) The trends in pretransfusion unconjugated bilirubin, STR, LDH, and reticulocyte count. Shaded areas indicate laboratory results before the implementation of new transfusion regimen (day 0). The dotted lines represent upper boundaries of normal values. We used repeated-measure analysis of variance to examine the association between intervention and biochemical or hematologic outcomes. Overall, intervention effect was significant for all laboratory outcomes. Moreover, the effects of intervention on improvement of outcomes remained persistent for each of the postintervention time brackets (0-120, 121-240, and 241-365 days). Eight to 12 months after the intervention, the average lactate dehydrogenase (964.37 [standard deviation [SD]: 239.85] vs 2736.75 [SD: 242.44]), indirect bilirubin (22.03 [SD: 5.53] vs 57.38 [SD: 10.45]), soluble transferrin receptor (2.15 [SD: 0.58] vs 7.31 [SD: 0.91]), and reticulocyte count (354.22 [SD: 36.42] vs 631.87 [SD: 82.54]) were all significantly lower compared with preintervention. (G-H) Associations of “functional” hemoglobin with STR and Hb H level with LDH. Dotted lines represent target values for functional hemoglobin and Hb H%, as well as the upper boundaries of normal values for LDH and STR. (See supplemental Materials available on the Blood Web site for the details of statistical analysis).

Close modal

On further investigation, hemoglobin analysis (by high-performance liquid chromatography and capillary zone electrophoresis) showed hemoglobin H (Hb H) percentage ranging from 24% to 64% in our patients, with older patients having higher Hb H. Hb H, a tetramer of 4 β-globin chains, has extremely high oxygen affinity and poor tissue oxygen delivery, making it essentially nonfunctional.4-6  With the high Hb H levels observed in our chronically transfused patients with homozygous α0-thalassemia, we estimated the patients’ “functional” hemoglobin as total hemoglobin × (1 – Hb H%/100), which ranged from 42.7 to 79.0 g/L.

To improve tissue oxygenation and suppress hemolysis, we elected to modify our transfusion strategy by aggressively suppressing endogenous erythropoiesis with a goal to reduce Hb H percentage to <15% and to keep the functional hemoglobin >100 g/L in our patients. These targets were chosen to correspond to the average Hb H levels observed in patients with Hb H disease and the recommended pretransfusion hemoglobin level in TDT-β patients. To achieve these targets, 3 patients with Hb H levels of >25% also required 1 to 4 exchange transfusions (one received 1 exchange, another received 3, and the third patient received 4 exchange transfusions). One year following the implementation of the new transfusion regimen, we observed a significant and persistent improvement in all hematologic and biochemical markers of hemolysis and decreased spleen size in all patients (Table 1; Figure 1). Over the course of 1 year, the new transfusion strategy required a higher total volume of transfusion compared with the standard approach (286 [SD: 21] vs 208 mL/kg per year [SD: 14]). However, over the latest 4 months of the year-long intensive transfusion approach, the volume of transfusion decreased to 258 mL/kg per year (SD: 22) when the new steady-state targets for the total pretransfusion functional hemoglobin (>100 g/L) and Hb H level (<15%) had been achieved.

A fifth patient, male and homozygous for common Southeast Asian deletion (--SEA/--SEA), was born after the implementation of the new transfusion strategy and has been transfused accordingly since birth. Fourteen months into this new regular transfusion regimen, this patient has normal serum LDH, aspartate aminotransferase, unconjugated bilirubin, and STR levels but continues to have peripheral blood reticulocytosis. The patient has normal growth parameters (age- and sex-adjusted height: 58 percentile, z-score = 0.20), achieved normal developmental milestones, and, at the last follow-up, had no splenomegaly. Total volume of transfused packed red blood cell in this patient has been 251 mL/kg per year.

Despite adequate transfusion intensity based on the TDT-β approach, the homozygous α0-thalassemia patients showed high Hb H levels in peripheral blood and profound reticulocytosis, suggesting that these patients maintained relatively more effective erythropoiesis (in contrast to ineffective erythropoiesis in their TDT-β counterparts) that was not adequately suppressed by transfusions. This suggests that homozygous α0-thalassemia is a predominantly hemolytic disease with a robust erythropoietic response. This is likely due to the fact that Hb H aggregates preferentially in older red blood cells in the peripheral blood, whereas in TDT-β, unpaired α-globin chains form molecular aggregates that lead to apoptosis of erythroblasts early in the process of erythropoiesis in the bone marrow.7  Our patients with homozygous α0-thalassemia who were transfused using the TDT-β regimen had preserved erythropoiesis, leading, over time, to increased Hb H levels associated with increased peripheral hemolysis. With the high Hb H percentage, the proportion of functional hemoglobin decreased, and tissue hypoxia worsened (despite stable total hemoglobin), further inducing erythropoiesis.

Suboptimal transfusions (with associated anemia and hypoxia) and severe hemolysis have been shown to be associated with a variety of significant clinical sequelae in β-thalassemia patients and in other disorders.8-10  Here we demonstrate that in patients with homozygous α0-thalassemia (TDT-α), a transfusion strategy similar to those of TDT-β patients does not adequately ameliorate the underlying hemolytic process and the poor tissue oxygen delivery over time, making this approach suboptimal for the management of these patients. Although we are yet to demonstrate any long-term clinical benefits to these patients (which would require a longer duration of regular transfusions), a more intensive transfusion regimen targeted at reducing Hb H levels resulting in an increased functional hemoglobin, reduced tissue hypoxia, and hemolysis seems more appropriate for patients with homozygous α0-thalassemia. Splenectomy has been associated with reduced transfusion requirements in TDT-β and improvement in hemoglobin level in Hb H disease patients with significant splenomegaly.11  However, in patients with homozygous α0-thalassemia, splenectomy, although prolonging survival of transfused Hb A-containing erythrocytes, would likely also prolong the lifespan of endogenous erythrocytes that almost exclusively carry nonfunctional Hb H. We considered these pathophysiologic differences, in addition to the known long-term complications of splenectomy, to inform our decision not to splenectomize our patients. Recently, an international registry of surviving patients with homozygous α0-thalassemia has been established (Douglas R. Higgs, Oxford University, personal communication, January 10, 2014). Clinical data gathered from this registry will be invaluable in shedding light on the long-term clinical benefit of our proposed transfusion strategy compared with standard TDT-β strategy, helping to define optimal care for patients with homozygous α0-thalassemia.

The intensive transfusion strategy comes with a price: increased transfusional iron burden that requires escalation of iron chelation therapy, posing significant challenges with treatment costs and patient adherence. Whether these drawbacks can be offset by better longer-term clinical outcomes of patients with homozygous α0-thalassemia remains to be proven.

The online version of this article contains a data supplement.

Acknowledgments: The authors thank Prof Douglas Higgs for valuable guidance for this study and Karen Charpentier and Manuela Merelles-Pulcini for support for this study and excellent care for patients and families. We acknowledge the eager participation of the children and families in this study.

A.A.’s research is in part funded by the Thalassemia Foundation of Canada.

Contribution: A.A. designed research; A.A., W.B., M.K.-A., and I.O. performed research; A.A. and S.C. analyzed data; A.A. and I.O. wrote the initial version of the paper; and all authors reviewed the paper and confirmed this final version.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Ali Amid, Division of Haematology/Oncology, Hospital for Sick Children, 555 University Ave, Toronto, ON, Canada M5G 1X8; e-mail: [email protected].

1
Piel
 
FB
Weatherall
 
DJ
The α-thalassemias.
N Engl J Med
2014
, vol. 
371
 
20
(pg. 
1908
-
1916
)
2
Chui
 
DH
Waye
 
JS
Hydrops fetalis caused by alpha-thalassemia: an emerging health care problem.
Blood
1998
, vol. 
91
 
7
(pg. 
2213
-
2222
)
3
Cappellini
 
MD
Cohen
 
A
Eleftheriou
 
A
et al. 
 
Guidelines for the Clinical Management of Thalassaemia. 2nd ed. Nicosia, Cyprus: Thalassaemia International Federation; 2008
4
Higgs
 
DR
Steinberg
 
MH
Forget
 
BG
Higgs
 
DR
et al. 
The pathophysiology and clinical features of α thalassemia.
Disorders of Hemoglobin. Genetics, Pathophysiology, and Clinical Management
2009
2nd ed
New York
Cambridge University Press
(pg. 
266
-
295
)
5
Papassotiriou
 
I
Kanavakis
 
E
Stamoulakatou
 
A
Kattamis
 
C
Tissue oxygenation in patients with hemoglobinopathy H.
Pediatr Hematol Oncol
1997
, vol. 
14
 
4
(pg. 
323
-
334
)
6
Imai
 
K
Tientadakul
 
P
Opartkiattikul
 
N
et al. 
Detection of haemoglobin variants and inference of their functional properties using complete oxygen dissociation curve measurements.
Br J Haematol
2001
, vol. 
112
 
2
(pg. 
483
-
487
)
7
Pootrakul
 
P
Sirankapracha
 
P
Hemsorach
 
S
et al. 
A correlation of erythrokinetics, ineffective erythropoiesis, and erythroid precursor apoptosis in thai patients with thalassemia.
Blood
2000
, vol. 
96
 
7
(pg. 
2606
-
2612
)
8
Olivieri
 
NF
Weatherall
 
DJ
Steinberg
 
MH
Forget
 
BG
Higgs
 
DR
et al. 
Clinical aspects of β thalassemia and related disorders.
 
Disorders of Hemoglobin. Genetics, Pathophysiology, and Clinical Management. 2nd ed. New York: Cambridge University Press; 2009:357-416
9
Musallam
 
KM
Rivella
 
S
Vichinsky
 
E
Rachmilewitz
 
EA
Non-transfusion-dependent thalassemias.
Haematologica
2013
, vol. 
98
 
6
(pg. 
833
-
844
)
10
Rother
 
RP
Bell
 
L
Hillmen
 
P
Gladwin
 
MT
The clinical sequelae of intravascular hemolysis and extracellular plasma hemoglobin: a novel mechanism of human disease.
JAMA
2005
, vol. 
293
 
13
(pg. 
1653
-
1662
)
11
Rigas
 
DA
Koler
 
RD
Decreased erythrocyte survival in hemoglobin H disease as a result of the abnormal properties of hemoglobin H: the benefit of splenectomy.
Blood
1961
, vol. 
18
 (pg. 
1
-
17
)

Author notes

*

M.K.-A. and I.O. contributed equally to this work.

Supplemental data

Sign in via your Institution