Bone disease in thalassemia

Supplemento 1, Vol. 12 - n. 3 - Settembre-Dicembre 2014

Perio

dico

qua

drim

estra

le -

Pos

te It

alia

ne S

.p.A

. - S

pedi

zion

e in

abb

onam

ento

pos

tale

- D

.L. 3

53/2

003

(con

v. in

L. 2

7/02

/200

4 n.

46)

- A

rt. 1

, com

ma

1 D

CB

Mila

no

Indexed in

EMBASE/Compendex/Geobase/SCOPUS

Ten years of research on growth and endocrine complications in thalassaemia

Contribution papers of the International Network of Clinicians for Endocrinopathiesin Thalassemia and Adolescence Medicine (ICET-A)

Leptin and ghrelin serum concentrations in thalassemia major and intermedia patients and normal subjects

Hamdollah Karamifar, Maryam Bahmanyar, Vincenzo De Sanctis, Mehran Karimi

Prevalence of diabetic ketoacidosis (DKA)in thalassemia major patients in Iran due to secondary iron overload

Mehran Karimi, Zohreh Karamizadeh, Saba Lahsaeizadeh, Hamta Jafari, Yunes Mavali, Vincenzo De Sanctis

Credibility of HbA1c in diagnosis and management of disturbances of glucose and diabetes in transfused patients with thalassemia

Christos Kattamis, Polyxeni Delaporta, Maria Dracopoulou, George Paleologos,

George P. Chrousos, Ioannis Papassotiriou, Antonios Kattamis

Reversible heart failure in a thalassaemic patient secondary to hypocalcemia and hypoparathyroidism

Vincenzo De Sanctis, Monica Sprocati, Maria Rita Govoni, Giuseppe Raiola

Bone disease in thalassemiaAshraf T. Soliman

MRI Assessment of iron overload in thalassemia: an overviewKavita Saggar, Praveen Sobti

ORGANO UFFICIALE

ISSN 2035-0678

33

Abbonamento annuale (3 numeri) Euro 30,00.Pagamento: conto corrente postale n. 1010097192 intestato a:Edizioni Scripta Manent s.n.c., via Bassini 41, 20133 MilanoÈ vietata la riproduzione totale o parziale, con qualsiasi mezzo, di articoli, illustrazioni e fotografiesenza l’autorizzazione scritta dell’Editore.L’Editore non risponde dell’opinione espressa dagli Autori degli articoli.Ai sensi della legge 675/96 è possibile in qualsiasi momento opporsi all’invio della rivista comunicandoper iscritto la propria decisione a: Edizioni Scripta Manent s.n.c. Via Bassini, 41 - 20133 Milano

DIRETTORE SCIENTIFICOVincenzo De Sanctis (Ferrara)

COMITATO DI REDAZIONESilvano Bertelloni (Pisa)

Giampaolo De Luca (Amantea, Cosenza)Bernadette Fiscina (New York, USA)

Giuseppe Raiola (Catanzaro)Tito Livio Schwarzenberg (Roma)

COMITATO EDITORIALEAntonietta Cervo (Pagani, Salerno)

Salvatore Chiavetta (Palermo)Michele De Simone (L’Aquila)

Teresa De Toni (Genova)Piernicola Garofalo (Palermo)

Maria Rita Govoni (Ferrara)Domenico Lombardi (Lucca)

Carlo Pintor (Cagliari)Luigi Ranieri (Catanzaro)

Leopoldo Ruggiero (Lecce)Giuseppe Saggese (Pisa)

INTERNATIONALEDITORIAL BOARDMagdy Omar Abdou (Alexandria, Egypt)Mujgan Alikasifoglu (Istanbul, Turkey)

Mike Angastiniotis (Nicosia, Cyprus)German Castellano Barca (Torrelavega, Spain)

Elsaid Bedair (Doha, Qatar)Monica Borile (El Bolson, Argentina)

Roberto Curi Hallal (Rio de Janeiro, Brasil)Yardena Danziger (Petah-Tiqva, Israel)

Oya Ercan (Istanbul, Turkey)Helena Fonseca (Lisbon, Portugal)

Daniel Hardoff (Haifa, Israel)Christos Kattamis (Athens, Greece)

Nogah Kerem (Haifa, Israel)Karaman Pagava (Tbilisi, Georgia)Praveen C. Sobti (Ludhiana - Punjab, India)

Ashraf Soliman (Doha, Qatar)Joan-Carles Suris (Lausanne, Switzerland)

SEGRETARIA DI REDAZIONELuana Tisci (Ferrara)

STAFF EDITORIALEDirettore Responsabile Pietro Cazzola

Direzione Generale Armando MazzùDirezione Marketing Antonio Di MaioConsulenza Grafica Piero Merlini

Impaginazione Stefania Cacciaglia

Scripta Manent s.n.c. Via Bassini, 41 - 20133 MilanoTel. 0270608091 - 0270608060 / Fax 0270606917E-mail: [email protected]

Registrazione Tribunale di Milano n. 404 del 23/06/2003

Stampa: EURGRAF sasCesano Boscone (MI)

Supplemento 1, Vol. 12 - n. 3 - Settembre-Dicembre 2014

Sommario

Leptin and ghrelin serum concentrationsin thalassemia major and intermedia patientsand normal subjects pag. ?

Hamdollah Karamifar, Maryam Bahmanyar, Vincenzo De Sanctis, Mehran Karimi

Prevalence of diabetic ketoacidosis (DKA)in thalassemia major patientsin Iran due to secondary iron overload pag. ?


Credibility of HbA1c in diagnosis and managementof disturbances of glucose and diabetesin transfused patients with thalassemia pag. ?

Christos Kattamis, Polyxeni Delaporta, Maria Dracopoulou, George Paleologos, George P. Chrousos,Ioannis Papassotiriou, Antonios Kattamis

Reversible heart failure in a thalassaemic patient secondary to hypocalcemia and hypoparathyroidism pag. ?


Bone disease in thalassemia pag. ?

Ashraf T. Soliman

MRI Assessment of iron overload in thalassemia: an overview pag. ?

Kavita Saggar, Praveen Sobti

ORGANO UFFICIALE

1

ICET-A (International Network of Clinicians for Endocrinopathies in Thalassemia and Adolescence

Medicine) an opportunity for improving thalassemia care

The prevalence of endocrine complications in thalassemic patients are still considerably high and this neces-

sitates national and international efforts for prevention, early detection and proper treatment of these disorders.

Summary of results for thalassemic patients > 6 years of age 2011

The prevalence of endocrine complications differed among centers because of the following:

1. using questionnaire type of surveys,

2. different transfusion status and compliance to chelation therapy

(different degrees of iron overload),

3. different degrees of hepatic dysfunction and nutritional status including

supplementation of folic acid, vitamin D and zinc and

4. using small samples (number of patients) and uncontrolled cohorts.

In addition, different methods of diagnosing endocrine abnormali-

ties have been used which make comparison of data difficult. Some sur-

veys use only clinical data while others use biochemical data for the dia-

gnosis of endocrine abnormalities e.g. (short stature versus growth hor-

Editoriale

Number 81

Age (years) 17 1 +/- 10.5 years

Sex 24 F + 57 M

Short stature HtSDS < -2 62%

Low IGF- I (IGF- I < - 2 SD) 85%

GHD (Peak GH < 7 ng/ml) (n = 42) 45%

Impaired Glucose tolerance (OGTT) 22,20%

Diabetes mellitus (FBG > 7.2 or 2h BG > 11.1) 6,20%

Hypothyroidism (low free T4 and/or high TSH) 5%

Hypocalcemia (Ca < 2 mmol/L) 7,50%

Hyperphosphatemia ( > 1.9 mmol/L) 8,60%

Hypoparathyroidism 7,50%

Vitamin D Deficiency (25 OHD < 20 ng/ml) 52%

Impaired liver function (elevated ALT) 22,20%

Cardiomyopathy (clinical and echocardiography) 31%

Hypogonadotropic hypogonadism 46%

Osteoporosis/Osteopenia 25%

Serum Ferritin Concentration ug/L 2500 +/- 1550

mone deficiency (GHD), pubertal delay/failure, primary and secondary amenorrhea versus hypogonadotropic hypo-

gonadism).

Moreover the use of different biochemical and hormone assay methods for interpretation of endocrine abnor-

malities produce significant difference in interpreting results e.g.: diagnosing glycemic abnormalities using fasting

blood glucose (FBG) versus oral glucose tolerance test (OGTT) versus continuous glucose monitoring system (CGMS),

and using different GH stimulation tests with different assays and cut-points .

Moreover, the time of doing the test in relation to blood transfusion (before versus after blood transfusion) may

significantly change the results of these endocrine tests (e.g testing IGF-1 level before versus after blood transfusion.

Despite all these facts, surveys from different countries (both clinical and biochemical) are very important to eva-

luate the extent of growth and endocrine affection in these patients and plan for prevention, early diagnosis and proper

treatment.

The practical objectives of ICET-A are to encourage and guide endocrinological follow up of multi-transfused

patients in developing countries, to promote and support collaborative research in this field, to encourage and guide

endocrinological follow up of multi-transfused patients, and to educate and train more endocrinologists and other

pediatricians/physicians to prevent and improve management of the growth and endocrine complications in these

patients.

Ashraf Soliman 1 and Vincenzo de Sanctis 2

1 Department of Pediatrics, Alexandria University Children’s Hospital, Egypt 2 Pediatric and Adolescent Outpatient Clinic, Private Accredited Quisisana Hospital, Italy

2

Abstract

Endocrine dysfunctions related to iron overload, such as delayed puberty, short stature and hypogonadism, lead to majorproblems in thalassaemic patients. Leptin, a polypeptide with 167 amino acids produced by white adipose tissue, is a hormone which reduces appetite andincreases energy consumption by affecting the central nervous system. Ghrelin, a peptide hormone produced by the stom-ach, stimulates growth hormone release via growth hormone secretagogue receptor. To evaluate leptin and ghrelin serum levels in thalassemia, 50 normal subjects, 50 -thalassaemia major patients and 50thalassaemia intermedia patients were randomly selected. Mean leptin concentration was 2.6 ± 1.2 g/L in patients with -thalassaemia major, and 2.8 ± 2.4 g/L in patients with

-thalassaemia intermedia. These values appeared to be significantly lower than controls (9.2 ± 2.9 g/L) (p < 0.001).Mean ghrelin concentrations were 1042.1 ± 275.9 pg/mL and 989.3 ± 275.5 pg/mL in -thalassaemia major and inter-mediate groups, respectively. This value in thalassaemia major appeared to be significantly higher compared to the con-trol group (876.9 ± 384.3 pg/mL) (p < 0.01).There was a positive correlation between serum leptin concentration and body mass index (BMI) in thalassaemia majorand intermedia. Leptin levels were significantly lower in thalassaemia major patients with short stature compared to con-trols, but no correlation was found between ghrelin levels and short stature in any of the three groups. These results suggestthat one of the endocrinopathies affecting thalassaemic patients is adipose tissue dysfunction and it may be that low leptinlevels play a role in the endocrine dysfunction in these patients. These findings need to be confirmed in further studies.

Key words: leptin, ghrelin, thalassemia.

.

Leptin and ghrelin serum concentrationsin thalassemia major and intermedia patientsand normal subjects

Hamdollah Karamifar 1, Maryam Bahmanyar 2, Vincenzo De Sanctis 3, Mehran Karimi 4

1 Pediatric Endocrine Department – Shiraz University of Medical Sciences – Shiraz (Iran)2 Pediatric Department – Shiraz University of Medical Sciences – Shiraz (Iran)3 Department of Reproduction and Growth – Pediatric and Thalassaemia Unit – Ferrara (Italy)4 Hematology Research Center – Shiraz University of Medical Sciences – Shiraz (Iran)

Direttore Scientifico

Vincenzo De Sanctis (Ferrara)

Comitato di Redazione

Vincenzo Caruso (Catania), Paolo Cianciulli (Roma), Maria Concetta Galati (Catanzaro),

Maria Rita Gamberini (Ferrara), Aurelio Maggio (Palermo)

Comitato Editoriale

Maria Domenica Cappellini (Milano), Marcello Capra (Palermo), Gemino Fiorelli (Milano), Alfio La Ferla (Catania), Turi Lombardo (Catania),

Carmelo Magnano (Catania), Roberto Malizia (Palermo), Giuseppe Masera (Monza), Lorella Pitrolo (Palermo), Luciano Prossomariti (Napoli),

Michele Rizzo (Caltanisetta), Calogero Vullo (Ferrara)

Segretaria di Redazione

Gianna Vaccari (Ferrara)

International Editorial Board

A. Aisopos (Athens, Greece), M. Angastiniotis (Nicosia, Cyprus), Y. Aydinok (Izmir, Turkey), D. Canatan (Antalya, Turkey),

S. Fattoum (Tunis, Tunisia), C. Kattamis (Athens, Greece), D. Malyali (Istanbul, Turkey), P. Sobti (Ludhiana, India), T. Spanos (Athens, Greece)

Com

itato

Edit

ori

ale

Agosto 2010

30

Emothal

Rivista Italiana di Medicina dell’Adolescenza - Volume 8, n. 2, 2010

Introduction

Presently, low weight and short staturein thalassaemic patients have become a majorhealth care problem, as thalassaemia is the mostcommon genetic disorder worldwide. Thalassaemia major (TM) manifests as a progres-sive hemolytic anemia caused by a defect in bothbeta globin genes. This severe anemia – hemo-globin (Hb) is usually between 3 to 7 mg/dL –leads to severe hepatosplenomegaly and growthdisorders and most patients will depend onrecurrent transfusions by the age of two. In tha-lassaemia intermedia (TI), patients carry a muta-tion in beta globin genes but are still capable ofmaintaining Hb between 6 to 10 mg/dL, so thatthey will not need recurrent transfusions, exceptin case of infections or surgery (1, 2). Endocrinedisorders such as short stature, delayed pubertyand hypogonadism, caused by iron overload inthalassaemic patients are major problems indica-ting endocrine system dysfunction (3). Leptin is a 16 kD polypeptide with 167 aminoacids. This hormone is secreted by adipose tis-sue and has a major role in long term mainte-nance of body weight. It can reduce the appetiteand increase energy consumption by affectingthe hypothalamus. Leptin inhibits neuropeptideY which is an appetite stimulator. It also leads togamma MSH expression which also reduces theappetite via hypothalamus (4-8). Ghrelin is a 28 amino-acid peptide secreted fromthe stomach which leads to growth hormone(GH) release. Ghrelin serum concentrationincreases before food intake and decreases afterthat (9). Ghrelin's effect on appetite, but not onGH release, depends on intact vagus outflow (10-12). Many hormones regulate serum level ofghrelin such as PYY3-36, which reduces ghrelinand suppresses appetite. The aim of this study is to find a possible rela-tionship between leptin and ghrelin levels andcomplications of thalassaemia major and inter-media such as short stature and low weight, inorder to find a potential therapy – such as recom-binant leptin hormone – to improve the healthstatus of these patients.

Methods and materials

The study was performed from January2008 to July 2009 at the Namazi Hospital of

Shiraz, Iran. The study population consisted of50 patients with TM referred to the ThalassaemiaDepartment of Shahid Dastgheib Hospital ofShiraz, and 50 patients with TI referred to theMotahari Clinic of Shiraz, selected by using ran-dom cluster sampling methods. Fifty healthychildren (matched for age and gender), selectedrandomly among the students of four educatio-nal zones of Shiraz city, served as the controlgroup. All patients and healthy children had nor-mal liver function tests. Written informed con-sents were taken for the study from all parents. Height was measured using a stadiometer, andweight was measured by Seca scale. Body massindex (BMI) was calculated using the formula (13):

wt (kg)BMI = _______

Ht (m2)

Patients were referred to Namazi HospitalResearch Center for collection of blood samples. The most recent results of the patients' hemoglo-bin and serum ferritin were recorded. Fastingblood samples (5 ml) were collected at 8 AM inNamazi Endocrine Research Center. The sampleswere centrifuged and the sera were maintained at-70 C. for tests. Leptin serum concentration was measured viaradioimmunoassay using Leptin Kit (DRGInstruments GmbH, Germany) and ghrelinserum concentration was determined via ELISAusing Ghrelin Kit from the same company. Variables in this study included gender, age, hei-ght, weight, BMI, leptin concentration, ghrelinconcentration, Hb and serum ferritin. The studywas approved by Research Council of ShirazUniversity of Medical Sciences. Chi square test was performed for investigatingrelationships between qualitative variables. Tostudy the relation between ghrelin and leptin withBMI in the three groups controlling for age, partialcorrelation test was used. One way ANOVA test was used to compare aquantitative variable in more than two groups andLeast Statistical Difference was used to comparecouple tests. BMI was studied in the three groupsusing ANOVA, considering age as a bias factor. Allstatistical analyses were performed by SPSS 15Software and p value < 0.05 was considered assignificant.

31

H. Karamifar, M. Bahmanyar, V. De Sanctis, M. Karimi


Emothal

Results

The thalassaemiamajor group (Group 1)included a total of 50patients: 24 female and 26male with an average age of14.2 ± 4.8 years. There were50 patients in thalassaemiaintermedia group (Group 2):21 female and 29 male withan average age of 14.0 ± 4.8.The average age of 50 nor-mal controls (Group 3) (20female and 30 male) was15.5 ± 2.0 years (Table 1).There was no statisticallysignificant difference inmean age or gender amongthe studied groups.Of the total of 150 children in this study, 19patients (38%) in the first group, 17 patients(34%) in the second group and 10 subjects(20%) in the control group were underweight(BMI < 5th percentile) and there was a significantdifference in the percentage of underweightsubjects in groups 1 and 2 in comparison withgroup 3 (p < 0.05) (Table 1).Short stature (height for age < 5th percentile) waspresent in 58% of Group 1, 30% in Group 2 and14% in the control group. The differences werestatistically significant (p < 0.001) (Table 1). Mean leptin serum level was 2.6 ± 1.2 µg/L inGroup 1, 2.8 ± 2.4 µg/L in Group 2 and 9.2 ± 2.9µg/L in Group 3. As shown by one way varianceanalysis, leptin levels in groups 1 and 2 weresignificantly lower than in Group 3 (p < 0.001)(Table 2). Mean ghrelin serum level was 1042 ±275.9 pg/mL in Group 1, which was significantlyhigher (p < 0.01) than in the control group(876.9 ± 384.3 pg/mL). In thalassaemia interme-dia (Group 2), mean serum ghrelin level was989.39 ± 275.5 pg/mL which, according to oneway variance analysis, was not significantly diffe-rent from Groups 1 or 3. Covariance analysis was performed for compari-son of BMI among the three groups controllingfor age. Mean BMI was 17.0 ± 2.7 in Group 1,17.8 ± 2.6 in Group 2 and 19.2 ± 2.7 in Group 3.The difference between Groups 1 and 3, and alsothat between Groups 2 and 3, was statisticallysignificant (p < 0.001); therefore, BMI in TM andTI patients was significantly lower than in healthy

controls (Table 2). A significant correlation wasfound between serum leptin level and BMI con-trolled for age in all groups (p < 0.004, p < 0.002,p < 0.001 respectively): with decreasing BMI,serum leptin level also decreased. No significantcorrelation was found between ghrelin serumlevel and BMI using partial correlation tests.The relation between leptin serum level andshort stature was shown to be significant inGroup 1 (TM) using T-test (p < 0.03), indicatingthat among TM patients (Group 1), those withshort stature had a lower leptin level. No correla-tion was found between serum leptin level andshort stature in the other two groups. No signifi-cant relation was found between ghrelin serumlevel and short stature in any of the three groups.Mean serum ferritin level was 1955.76 ng/ml inGroup 1, 688.72 ng/ml in Group 2, and 98 ng/mlin group 3, that is, significantly higher in TM com-pared to TI and the control group (p < 0.001).

Discussion

As seen in the distribution of short stature amongthe study groups (Table 1), short stature in thalas-saemic patients is more prevalent than in normalsubjects. In addition, mean serum leptin level inthalassaemia major and intermedia patients issignificantly lower than in healthy children (Table2). Moreover, in major thalassaemia patients withshort stature, mean leptin serum level was signifi-cantly lower than in normal controls and a signifi-

Table 1.

Age, sex, low, and short stature in thalasse mic patients and controls.

Table 2.

BMI, leptin and ghrelin levels in thalassemic patients and controls

Study Group Age Male Female Underweight Short stature

Thalassaemia

Major 14.2 ± 4.2 26% 24% 38% 58%

Thalassaemia

Intermedia 14.0 ± 4.8 29% 21% 34% 30%

Controls 15.5 ± 2.0 30% 20% 20% 14%

Study Group Age BMI Leptin(µg/L) Ghrelin(pg/mL)

Thalassaemia

Major 14.2 ± 4.2 17.0 ± 2.7 2.6 ± 1.2 1042.1 ± 275.9

Thalassaemia

Intermedia 14.0 ± 4.8 17.8 ± 2.6 2.8 ± 2.4 989.3 ± 275.5

Controls 15.5 ± 2.0 19.2 ± 2.7 9.2 ± 2.9 876.9 ± 384.3

32

Emothal


cant relationship was observed between short sta-ture and leptin serum level.In a study in 1999 to investigate leptin serumconcentration, 162 thalassaemia major patientswere compared to 138 healthy controls. Meanleptin serum level in male subjects was 2.69 ±1.23 in the thalassaemic group and 6.86 ± 2.71in the control group. In female subjects levelswere 6.37 ± 2.9 in the thalassaemic group and9.37 ± 5.2 in the control group. In both cases thedifference was found to be statistically significant(p < 0.0005 and p < 0.05, respectively) (14). Inanother study in Athens on 40 major thalassae-mia patients (15) it was found that leptin serumlevel was lower in these patients. The results of these previous studies (14-15)indicate a relationship between leptin serumlevels and BMI and a lower leptin level amongthalassaemic patients in comparison with a nor-mal population. Our study has also confirmedthis fact. It seems that adipose cells of thalassae-mia patients are not able to produce adequateleptin which might be due to deposition of ironin these cells. Therefore, the defect in adipose tis-sue function in thalassaemic patients can be con-sidered as an endocrine system dysfunction,although it seems other factors may interfere inthe decrease of serum leptin level in thalassaemicpatients. As a patient is more underweight withless fat tissue, the ability to produce leptin pro-duction would be lower (14). According to other researches, there are differentcontributing factors to short stature in thalassae-mia, including hypothyroidism, hypoparathyroi-dism (16, 17), adrenal insufficiency (18) andpancreatic dysfunction (19). In our study therewas direct correlation between short stature andserum leptin levels in TM patients. We believethat low leptin may be a factor of short stature inthese patients but further studies are needed toinvestigate the possible relationship. Serum ghrelin levels in the TM group werehigher compared to controls which might be dueto a compensatory response to growth retarda-tion or a partial resistance to ghrelin that leads toits increased level (20). The results of this studydid not show any relation between ghrelin serumconcentration and short stature, which is consi-stent with previous studies. In a study, in 2006,in Turkey by Camurdan MO et al. on 17 childrenwith constitutional growth retardation, 19 withfamilial short stature and 11 normal subjects,serum concentrations of ghrelin, IGF-1 and

IGFBP-3 were measured. The study showed thatserum ghrelin levels in children with familialshort stature were higher than in controls (20).But according to a previous research, height andweight are independent to ghrelin level (20). Sothe Authors postulated that the negative relationfound between height and ghrelin level is becau-se of a compensatory increase in ghrelin level inresponse to short stature. In China, Zou CC et al.performed another study on 117 patients withshort stature due to growth hormone deficiency,81 with idiopathic short stature and 125 normalchildren as controls. The aim of the study was toexplore serum ghrelin concentration and poly-morphism of the ghrelin/obestatin gene (21).The results indicated that in patients with growthhormone deficiency ghrelin serum level wassignificantly lower than in the control group,which suggests a probable important role forghrelin in growth hormone secretion and growthcontrol. In this study, sexual maturation was notinvestigated; this factor, as well as others, shouldbe considered in future researches.

References

1. Weatheral DJ, Clegg JD. The thalassemia syndromes. 4thed. Oxford: Blackwell Scientific, 2001.

2. Wainscoat JS, Thein SL, Weatheral DJ. Thalassemia inter-media. Blood Rev 1987; 1:273-279.

3. Kattamis CA, Kattamis AC. Management of thalassemias:growth and development, hormone substitution, vitaminsupplementation, and vaccination. Seminars in Hematology1995; 32:269-279.

4. Wang J, Liu R, Hawkins N, Barzilai N, et al. A nutrient-sensing pathway regulates leptin gene expression in muscleand fat. Nature 1998; 393:684-688.

5. Skleton JA, Rudolph CD. Overweight and obesity. In:Kliegman RM, Behrman RE, et al. eds. Nelson textbook ofpediatrics. 18th ed. Philadelphia: WB Saunders Co 2007; p.232-241.

6. Ebbeling CB, Fledman HA, Osganian SK. Effect of decreas-ing sugar sweetened beverage consumption on body weightin adolescents: a randomized, controlled pilot study.Pediatrics 2009; 117:673-680.

7. Pande H, Cheskin LJ. Obesity: etiology and diagnosis. In:Trugo LC, Finglas P. Encyclopedia of food sciences andnutrition 2003; 4220-4227.

8. Tsiotra PC, Pappa V, Raphis SA, Tsigos C. Expression of thelong and short leptin receptor isoforms in peripheral bloodmononuclear cell: implications for leptins’ action.Metabolism 2000; 49:1537-1541.

9. Kojima M, Kangawa K. Ghrelin: structure and function.Physiol Rev 2005; 85:495-522.

10. Le Roux CW, Near NM, Halsey TJ, et al. Ghrelin does not

33

stimulate food intake in patients with surgical proceduresinvolving vagotomy. J Clin Endocrinol Metab 2005;90:4521-4524.

11. Nagaya N, Moriya J, Yasumura Y, et al. Effects of ghrelinadministration on left ventricular function, exercise capaci-ty, and muscle wasting in patients with chronic heart failure.Circulation 2004; 110:3674-3679.

12. Castaneda TR, Tong J, Datta R, Culler M, Tschop MH.Ghrelin in the regulation of body weight and metabolism.Front Neuroendocrinol 2010; 31:44-60.

13. Keane V. Assessment of growth. In: Behrman K, Stanton Jeds. Nelson textbook of pediatrics. 18th ed. Philadelphia:WB Saunders 2008; p. 70-71.

14. Miraglia Del Giudice E, Perrotta S, Carbone MT, et al.Evaluation of leptin protein levels in patients with Cooley’sanemia. Br J Haemat 1999; 105:839-840.

15. Dedoussis GV, Kyrtsonis MC, Andrikopulos NE et al.Inverse correlation of plasma leptin and soluble transferringreceptor levels in beta-thalassemia patients. Ann Hematol2002; 81:543-547.

16. Karamifar H, Shahriari M, Amirhakimi GH. Linear growthdeficiency in beta thalassemia patients: is it growth hormonedependent? Ir J Med Sci 2002; 27:47-50.

17. Karamifar H, Shahriari M, Sadjadian N. Prevalence ofendocrine complications in b-thalassemia major in theIslamic Republic of Iran. EMHJ 2003; 9:55-60.

18. Gulati R, Bhatia V, Agarwal SS. Early onset of endocrineabnormalities in beta-thalassemia major in a developingcountry. J Pediatr Endocrinol Metab 2000; 13:651-656.

19. Costin G, Kogut M, Hymen C, et al. Carbohydrate metabo-lism and pancreatic islet cell function in thalassemia major.Diabetes 1977; 26:230-240.

20. Camurdan MO, Bideci A, Demirel F, Cinaz P. Serum ghre-lin, IGF-I and IGFBP3 levels in children with normal vari-ant short stature. Endocr J 2006; 53:479-484.

21. Zou CC, Hudug K, Liang L, Zhao ZY. Polymorphisms ofthe ghrelin/obestatin gene and ghrelin levels in Chinesechildren with short stature. Clin Endocrinol (OXF) 2008;69:99-106.

H. Karamifar, M. Bahmanyar, V. De Sanctis, M. Karimi


Emothal

Corrispondenza:

Mehram Karimi, MDProfessor of Pediatric and Hematology-OncologyHematology Research CenterNemazee HospitalShiraz, IranTel/Fax: +98-711-6473239e-mail: [email protected]

Aprile 2008

SummaryIntroduction. Thalassemia is considered the most common genetic disorder worldwide. The main treatments for tha-lassemic patients are transfusion and iron chelation therapy. Diabetes mellitus (DM) is a frequent complication in tha-lassemia major due to iron overload. Some patients may develop diabetic ketoacidosis (DKA) and expire.Aim. To determine the prevalence of diabetes and the rate of DKA in thalassemia major patients in the south of Iran.Methods: We reviewed data of thalassemic patients who were treated in our area over a 13 year period (one year prospec-tively and 12 years retrospectively). In addition we investigated 37 patients with thalassemia major and DKA. We useddescriptive study for data analysis.Results. Among 820 thalassemia major patients receiving transfusion in Shiraz (a capital center in the South of Iran), 70individuals (8.5 %) had a fasting blood sugar of 126 mg/dl or higher and thus were diagnosed as diabetic. The incidenceof DKA in the year 2004 was 8 per approximately 2500 major thalassemic patients (0.3%) in Fars province in the Southof Iran, with 5 of these patients developing DKA for the first time (0.2 %). The mean ages of developing diabetes and DKAin thalassemic patients were 18.2 and 15.5 years, respectively. The most common symptoms were polyuria and polydyp-sia (94.5%).There were more female thalassemic patients (73%) than males (27%) in the group that developed DKA, andthis difference was significant.Discussion. Our study has demonstrated that a significant number of thalassemia major patients had diabetes, and someof them were complicated with DKA. Diabetes developed at the same age as in other countries. We suggest that blood sugar should be checked routinely in all thalassemia major patients after the age of 9 years to pre-vent life-threatening conditions such as DKA.

Key words: thalassemia major, diabetes mellitus, ketoacidosis, iron overload, Iran.

Prevalence of diabetic ketoacidosis (DKA) in thalassemia major patients in Iran due to secondary iron overloadMehran Karimi, Zohreh Karamizadeh1, Saba Lahsaeizadeh, Hamta Jafari, Yunes Mavali, Vincenzo De Sanctis2

Haemostasis and Thrombosis Unit, Hematology Research Center, Shiraz University of Medical Science, Shiraz, Iran; Pediatric Endocrinology Department, Nemazee Hospital,Shiraz University of Medical Science, Shiraz, Iran1; Department of Pediatrics and Adolescent Medicine, Thalassaemia Unit, Arcispedale S. Anna, Ferrara, Italy2






Comitato Editoriale









Com

itato

Edit

ori

ale

48

Emothal


Introduction

Iran is one of the countries in whichthalassemia is endemic. According to the datareleased by the Center of Disease Control, theoverall carrier rate of the gene is 5% (1).Patients with thalassemia major suffer not onlyfrom the consequences of the anemia itself (suchas reactive bone changes and splenomegaly), butalso from complications caused by the repeatedblood transfusions they receive for the correc-tion of the anemia. The majority of these com-plications are due to chronic iron overload inbody tissues, with subsequent pathologicchanges leading to disturbances in the functionof various organs.Diabetes mellitus is one of the most commoncomplications in thalassemic patients and iscaused by the following factors, either alone or incombination: destruction of pancreatic‚ cells anddecreased insulin release, damage to hepatocytes,peripheral resistance to insulin, and genetic fac-tors (2). Insulin resistance is one of the factorsleading to hemochromatosis-induced diabetes.However, studies have shown that diminishedinsulin release is much more important. Familialpredisposition is also an important predictor ofglucose intolerance, at least in patients with pri-mary hemochromatosis (3). Other reports havefound that noncompliance of patients in usingchelating agents, increased body iron, and livercirrhosis are related factors (4, 5). Another studyshowed that the prevalence of glucose intoler-ance was higher in patients who were older andhad more blood transfusions (6).

Diabetes ketoacidosis (DKA) is a major acuteemergency in diabetics, and can be life threaten-ing in thalassemic patients already compromisedby the anemia and its complications. For this rea-son, screening for diabetes and effective controlof blood sugar, plus rapid and proper manage-ment of ketoacidosis are of utmost importance inthese patients.The present study was performed to determinethe prevalence of diabetes and the incidence ofketoacidosis in thalassemic patients under thera-py in Shiraz and 18 affiliated Centers in the southof Iran. No previous data were available for thisgroup of patients.

Patients and Methods

Our project was based on a populationof approximately 2500 thalassemia majorpatients receiving blood transfusions in the southof Iran (Fars Province), with Shiraz as the mainthalassemia center, and 18 other university affili-ated centers. The study was designed with both prospectiveand retrospective branches.In the retrospective branch, the documents of allthalassemia major patients with the diagnosis ofDKA admitted in three major hospitals affiliatedwith Shiraz Medical School (Nemazee, Saadi andDastgheib Hospitals) from 1991 to 2003 wereselected and the required information (age, sex,medical history, signs and symptoms, and labo-ratory data) was recorded. All patients with DKAare referred to these hospitals. The criteria fordiagnosis of DKA were a blood sugar level above300 mg/dl, serum positive for ketones in dilu-tions above 1/2, metabolic acidosis (PH less than7.30 and bicarbonate level below 15 mEq/l),ketonuria and glycosuria. Clinical manifestationsincluding vomiting, polyuria, dehydration,abdominal pain, and decreased level of con-sciousness were also documented.In the prospective branch, we included allpatients with thalassemia who were admittedover a one year period (2003-2004) to theNemazee Hospital Pediatric EndocrinologyWard with the diagnosis of diabetic ketoacido-sis, and recorded their clinical and laboratorymanifestations, liver function tests and pro-thrombin time (PT). Thalassemic-diabetic patients with a history of

Figure 1. The sex distribution (%) of thalassemic patients

with diabetes and ketoacidosis.

0

49

10

20

30

40%

50

60

70

80

Diabetes Ketoacidosis

27

51

73

Male

Female

49

DKA were assessed for their age, sex, liver span,history of splenectomy and serum ferritin level,plus their clinical manifestations, blood sugar,PH, bicarbonate level, and serum ketones duringthe episode of DKA.In addition, all thalassemic patients receivingtransfusions in the thalassemic ward of DastgheibHospital were tested for serum ferritin and fast-ing blood sugar during the period of our study,and those individuals with a blood sugar level of126 mg/dl or higher were diagnosed with dia-betes. All data were analyzed with SPSS version 13 anda p value less than 0.05 were considered as sig-nificant.

Results

In 2004, there were nearly 2500 regis-tered thalassemia major patients in Fars province(Shiraz and 18 affiliated centers), in the south ofIran. In the same year, eight of these patientsdeveloped DKA, five of them for the first time.Thus, the incidence of DKA in year 2004 was 8per 2500 major thalassemic patient (0.3%) andthe rate of first occurrence of DKA among thesepatients was 5 per 2500 (0.2 %).During the period from 1991 to 2004, 37 ofthese‚ β-thalassemia major patients were admit-ted to the hospital for DKA, including the 8patients admitted in 2004.

Among thalassemia major patients receivingtransfusion in Shiraz (820 patients), 70 subjects(8.5 %) had a fasting blood sugar of 126 mg/dl orhigher and thus were considered diabetic; 12individuals among these diabetic-thalassemicpatients (17.1%) were admitted at least once dueto DKA. Therefore, the rate of DKA in tha-lassemic patients receiving transfusions in Shirazwas 1.46 % while the prevalence of thalassemiamajor in patients with DKA was 8.5 %. The average age of thalassemic patients develop-ing diabetes was 18.2 years (range 9 to 34 years).In females the mean age was 17.5 years com-pared to mean of 19.3 years in males. The meanage for thalassemic patients with diabetes admit-ted for DKA was 15.5 years old. Their distribu-tion by sex is shown in Figure 1. Clinical mani-festations of these patients are presented inFigure 2.Among the 70 patients with diabetes, 47 had hadsplenectomy (67.1%) while 84% of thalassemicpatients admitted for DKA had undergonesplenectomy. Only 19% had normal liver span;liver was palpable 2 to 5 cm below the costalmargin in 43% of these patients and more than 5cm in 38 %.The average blood sugar in patients with DKAwas 542 mg/dl at the time of admission. Theiraverage pH was 7.08 and mean bicarbonate levelwas 6.04 mEq/l.The average serum ferritin level was 2320 ng/mlin thalassemic patients with diabetes on transfu-sion in Shiraz and 2760 ng/ml in those admitteddue to DKA. The average protein in thalassemicpatients with diabetes was 7.2 g/dl and the aver-age albumin 3.85 g/dl. Hypoalbuminemia wasseen in 10.5 % of cases, but no patient hadhypogammaglobulinemia. All patients had ele-vated liver transaminases and 33 % had pro-longed prothrombin time.

Discussion

Beta thalassemia major is a seriousgenetic disease, but with regular transfusion andiron chelating therapy patients generally reachadulthood. Unfortunately, the prevalence of com-plications due to iron overload is still high(1-4).Deposition of iron in the pancreas has been welldocumented and development of diabetes in tha-lassemic patients has generally thought to be due


Prevalence of diabetic ketoacidosis (DKA) in thalassemia major patients in Iran due to secondary iron overload

Emothal

Figure 2. Rates of clinical manifestations in thalassemic patients with diabetic ketoacidosis.

0

1010

94.5

%

Polyd

ypsia

Polyu

ria

Drowsin

ess

Abdomin

alpain

Nause

a/Vo

mitin

g

Polyp

hagia

Coma

20

30

40

50

60

70

80

90

100 94.5

67.5

5146

16

2.7

50

to direct toxic effect of iron on the pancreas isletcells that leads to insulin deficiency (7, 8).However, insulin resistance (from iron deposi-tion in both liver and muscles) is also involved inthe changes in glucose metabolism observed inthalassaemia (7-15). In present study we found that of 820 tha-lassemic patients receiving transfusion in Shiraz,70 individuals (8.5%) had diabetes according toour criteria. Different reports have been issuedon this prevalence, ranging from 49 to 138 per1000 patients (16-19).A study in Taiwan (12) showed that the preva-lence of diabetes was as high as 19.5% in 89 tha-lassemia major patients on transfusion therapyand that of impaired glucose tolerance was 8.5%.According to this study, presentation with DKAamong patients with diabetes was as high as31.1%.In an early study by Mc Intosh reported in1976(14), 4 out of 9 prepubertal thalassemicyouth had diabetic glucose tolerance tests. In Italy, a prevalence of diabetes of 6% wasreported in‚ β-thalassemia major patients in 1990(2). In another study done by De Sanctis et al.(13) in 2004 on 3817 thalassemic patients, dia-betes was present in 3.2 % of patients.A study in Saudi Arabia by El-Hazmi et al. (15)on 50 thalassemia major patients, in 1994,showed prevalence of diabetes of 6%, which wasattributed to insulin insensitivity produced byiron overload, that eventually leads to exhaustionof β cells in the pancreas.Among our patients with diabetes, 12 individuals(17.1%) were admitted at least once due to DKA.Not many studies have reported the incidence ofDKA in thalassemic patients. The manifestationhas been as high as 31.1 % in Taiwan (12).In our group of patients, the incidence of DKA inyear 2004 was 8 per 2500 thalassemia majorpatients (0.3%) and the rate of first occurrence ofDKA among these patients was 5 per 2500(0.2%). The annual incidence of DKA in non-thalassemic diabetic population is 3 to 8 per1000 (20).Seventy-three percent of patients presenting withDKA in this study were females (female to maleratio of 2.7) while this proportion was 51% indiabetic patients without DKA (Figure 1). Thedifference was significant (p<0.05). The reasonsfor this difference are yet to be determined; nev-ertheless, these findings point out the need forcareful monitoring of the metabolic state of

female thalassemic patients. Two previous studiesdone in other regions reported mean ages of 18and 18.1 years for the development of diabetes inthalassemic patients (9, 19). In our study the mean age was 18.2 years, almostexactly the same as in the other studies, whichmay indicate similarities in the progression of thedisease and/or screening methods among the dif-ferent Centers. The mean age of our thalassemicpatients who developed DKA was 15.5 years; weknow of no other report of the age of tha-lassemics with DKA.Eighty-one percent of thalassemic patients with ahistory of DKA had hepatomegaly and their aver-age serum ferritin level was 2760 ng/ml, both ofwhich suggest the role of inadequate control ofiron loading in pathogenesis of diabetes. In onestudy, the average ferritin level in thalassemicpatients with diabetes was 5600 ng/ml (21). Ithas been shown that endocrine sequelae tend todevelop at older ages and with higher ferritin lev-els. Life expectancy can be increased if ferritinlevel is kept below 2000 ng/ml, which requiresregular iron chelating therapy (21).In the light of the results of our study, we strong-ly suggest surveillance and follow-up of patientswith β-thalassemia for endocrine and liver disor-ders in order to detect and prevent or alleviateassociated complications.

References

1. Haghshenas M, Zamani J. Thalassemia ResearchVice-Presidency of Fars University of Medical Sciences. First edi-tion, 1998.

2. Vullo C, De Sanctis V, Katz M et al. Endocrine abnormalitiesin thalassemia. Am NY Acad Sci 1990; 612:293-309.

3. Catanese C, Kahn R. Secondary forms of diabetes mellitus. In:Becker KL, ed. Principles and practice of endocrinology andmetabolism. 3 rd Ed. Philadelphia PA: Lippincott Williams-Wilkins; 2001, pp1327-1336.

4. Gamberini MR, Fortini M, Gilli G et al. Epidemiology andchelation therapy effects on glucose homeostasis in thalassemicpatients. J Pediatr Endocrinol Metab 1998; 11(Suppl 3):867-869.

5. Costin G, Kogut MD, Hyman C et al. Carbohydrate metabo-lism and pancreatic islet cell function in thalassemia major.Diabetes 1997; 26:230-240.

6. Saudek CD, Hemm RM, Peterson CM. Abnormal glucose tol-erance in beta-thalassemia major. Metabolism 1997; 26:43-52.

7. Lassman MN, Genel M, Wise JK et al. Carbohydrate home-ostasis and pancreatic islet cell function in thalassemia. Ann IntMed 1974; 80:65-69.

8. Saudek CD, Hemm RM, Peterson CM. Abnormal glucose tol-erance in ß-thalassemia major. Metabolism 1977; 26:43-52.

Emothal


51

9. Arrigo T, Crisafulli G, Meo A at al. Glucose tolerance, insulinsecretion and peripheral sensitivity in thalassemia major. JPediatr Endocrinol Metab 1998; 3:863-866.

10. Merkel PA, Simonson DC, Amiel SA et al. Insulin resistanceand hyperinsulinemia in patients with thalassemia major treatedby hypertransfusion. New Engl J Med 1988; 318:809-814.

11. Cavallo-Perin P, Pacini G, Cerutti F et al. Insulin resistanceand hyperinsulinemia in homozygous‚ ß-thalassemia.Metabolism 1995; 44:281-286.

12. Chern JPS, Lin KH, Lu M et al. Abnormal glucose tolerancein transfusion-dependent‚ ß-thalassemic patients. Diabetes Care2001; 24:850-854.

13. De Sanctis V, Eleftheriou A, Malaventura C et al. Prevalenceof endocrine complications and short stature in patients with tha-lassaemia major: a multicenter study by the ThalassaemiaInternational Federation. Ped Endocrinol Rev 2004; 2(Suppl.2):249-255.

14. McIntosh N. Endocrinopathy in thalassaemia major. ArchDis Child 1976; 51:195-201.

15. El-Hazmi M, Al Swailem A, Al Fawaz I et al. Diabetes mel-litus in children suffering from beta-thalassemia. J Trop Pediatr1994; 40:261-266.

16. De Sanctis V, Zurlo MG, Senesi E et al. Insulin dependentdiabetes in thalassemia. Arch Dis Child 1998; 63:58-62.

17. Dandona P, Hussain MAM, Varghese Z et al. Insulin resist-ance and iron overload. Ann Clin Biochem 1983; 20:77-79.

18. Schmid-Schonbein H, Volger E. Red cell aggregation and redcell deformity in diabetes. Diabetes 1976; 25:897-902.

19. De Sanctis V, Pintor C and Italian working group onendocrine complications in non-endocrine disease. ClinEndocrinol (Oxf) 1995; 42:581-586.

20. George K, Alberti M et al. Diabetic acidosis, hyperosmolarcoma. In: KL Becker, ed. Principles and practice of endocrinolo-gy and metabolism. 3rd ed. Philadelphia PA: Lippincott Williams-Wilkins; 2001, p. 1439.

21. Cohen A, Lombardo F, Miceli M et al. Rapid removal ofexcessive iron with daily high dose intravenous chelation thera-py. J Pediatr 1989; 115:151-155.


Prevalence of diabetic ketoacidosis (DKA) in thalassemia major patients in Iran due to secondary iron overload

Emothal

Correspondence to:

Mehran Karimi MD

Professor of Pediatric Hematology and OncologyHematology Research CenterNamazi HospitalShiraz, IranTel/Fax: (+98)711-6265024

e-mail: [email protected]

Abstract

Background: HbA1c levels are considered essential in the management of diabetes. Their credibility had been questioned in hemoglobinopathies and espe-cially for thalassemia and sickle cell disease.Objectives: The main objectives of this study were to evaluate the credibility of Hb A1c in following disturbances of glucose metabolism and monitoring man-agement of diabetes in regularly transfused patients with thalassemia major.Research Design: 34 patients with thalassemia major, with a mean pre-transfusion Hb of 10g/dl, 17 with normal glucose tolerance (NGT) and 17 with dia-betic and impaired tolerance (IGT), were studied. Hb A1c was assessed on a transfusion day on two blood samples, one prior (Pre-Tx) one post (Post-Tx)transfusion, and another prior to the consecutive transfusion within two weeks (Pre-Tx2).Results: Data were evaluated separately in the two groups. In patients with NGT, Hb A1c ranged from 5.6-6.9% (mean 6.5 ± 0.3%) and in patients withdiabetes, from 6.5-9.8% (mean 7.8 ± 1.1%). All five patients with diabetes with A1c < 7.0% had normal fasting plasma glucose < 100mg/dl, indicating effi-cient control of diabetes. After transfusion (Post-Tx), Hb A1c levels decreased significantly in both groups (p < 0.001), while the (Pre-Tx2) assessment of HbA1c showed a trend to increase to the initial Pre-Tx levels in both groups. The increasing trend of Hb A1c is attributed to the influence of the average glu-cose concentration on transfused erythrocytes.Conclusions: The results support the reliability of Hb A1c assessment to follow and monitor treatment of glucose disturbances in regularly transfused patientswith thalassemia major. Further studies are indicated for precise identification of the range of Hb A1c levels in transfused patients with thalassemia withnormal glucose tolerance, as well as the period covered by Hb A1c estimation. The period is assumed to be short considering the short life span of storagered cells.

Key words: transfusion-dependent thalassemia, diabetes, glucose disturbances, Hb A1c.

Credibility of HbA1c in diagnosis and managementof disturbances of glucose and diabetes in transfused patients with thalassemia

Christos Kattamis 1, Polyxeni Delaporta 2, Maria Dracopoulou 1, George Paleologos 2, George P. Chrousos 1,

Ioannis Papassotiriou 2, Antonios Kattamis 1

1 First Department of Pediatrics, Athens University Medical School, Aghia Sophia Children’s Hospital, Athens, Greece2 Department of Clinical Biochemistry, Aghia Sophia Children’s Hospital, Athens, Greece

2

Endo-Thal


Introduction

Assessment of glycated hemoglobin (Hb A1c)was introduced into clinical use in 1980 and subsequently it became a cornerstone in thediagnosis and monitoring of patients with dia-betes (1-4). Hb A1c is synthesized by binding thefree amine group of glucose molecule to beta glo-bin chain, through a two-step reaction: in thefirst, the reversible step, a bond between freealdehyde group of glucose and amine protein isformed. Subsequently an intra-molecular rear -ran gement follows, which results in formation ofa stable ketoamine molecule. The levels of Hb A1c reflect the average plasmaglucose concentration over the lifespan of ery-throcytes which in normal individuals survive upto 120 days (5).At any given time a blood sample contains red cellsof varying ages and with different degree of expo-sure to hyperglycemia. Blood glucose levels fromthe preceding 30 days contribute approximately50% to Hb A1c whereas those from the period of90-120 days earlier, approximately 10% (6).Recent studies demonstrated that assessments ofHb A1c are almost equal in sensitivity and speci-ficity to fasting and 2 hours plasma glucosemeasurements as predictors of the complicationsof retinopathy, neuropathy and nephropathy,provided that for Hb A1c measurements strin-gent quality assurance tests and standardizationcriteria are implemented (1, 7). Hb A1c can beassessed at any time of the day without fastingpreparation. However, a number of conditionsinfluence its accuracy and credibility: basically,those with reduced red cell lifespans, as the acuteand chronic hemolytic anemias; those with con-siderable reduction or absence of synthesis of HbA, as the homozygous β-thalassemias; and con-ditions with substitution of Hb A by an abnormalhemoglobin, such as patients with heterozygousor homozygous sickle cell disease, and theabnormal hemoglobins C, D, E and others (8).The credibility of assessment of Hb A1c for mon-itoring the management of diabetes in patientswith hemoglobinopathies and particularly fortransfused patients with thalassemia major hasbeen questioned (1, 8, 9). These patients are at high risk to develop dia-betes during adolescence and early adulthood.The prevalence of diabetes among adolescentsand young adults with thalassemia major variesbetween 14 and 24% (10).

Considering the hemoglobin composition andthe functional and lifespan peculiarities of trans-fused erythrocytes, the prevailing assumption ofreliability of HbA1c estimation has been cast intodoubt for this group of patients.In regularly transfused patents with thalassemiamajor, a blood sample is a mixture of blood from2-6 donors; it has a normal hemoglobin compo-sition, of 95% Hb A and less than 5% Hb F in themajority of patients. In addition, transfused stored erythrocytes haveconsiderable functional and metabolic differ-ences and their life span is extremely short, notexceeding 40 days. To our knowledge, the changes in Hb A1c in reg-ularly transfused patients with thalassemia andtheir relation to the disturbances of glucosemetabolism (which are common and result fromspecific pathophysiology), have not been exten-sively studied.This study was designed with the main objective toevaluate the credibility of Hb A1c assessment, byexploring the changes of HbA1c in relation to dis-turbances of glucose metabolism in 34 frequentlytransfused patients with thalassemia major.

Patients

Of patients with thalassemia major, we selected 17with normal glucose tolerance and 17 with dia-betes and impaired glucose tolerance who main-tained a mean pre-transfusion Hb level of 10 g/dl.The selection of the 34 patients with thalassemiamajor was based on the following criteria: i) Precise characterization of clinical and hema-

tological phenotypes, as well the genotype, ofβ-thalassemia.

ii) Baseline evaluation of clinical and laboratoryfindings of the disease and its complications,supplemented by glucose metabolism status,based on oral glucose tolerance test (OGTT).

iii) Adequate compliance with a regular transfu-sion schedule.

iv) Written informed consent.

To minimize discomfort, the study was designedto follow the transfusion schedule of our Unit.Blood samples were collected only on transfusiondays. The trial started on a transfusion day with a col-lection of one blood sample prior to transfusion

3

C. Kattamis, P. Delaporta, M. Dracopoulou, G. Paleologos, G.P. Chrousos, I. Papassotiriou, A. Kattamis


Endo-Thal

ple came from a normal distribution. Values ofthese statistics outside the range of -2 to +2 indi-cate significant deviations from normality, whichwould tend to invalidate many of the statisticalprocedures normally applied to this data. Thesevalues integrated automatically from the programindicated the need for transformation in eitherlog, reciprocal or square root values and conse-quent transformations were then used to allowcorrelations between parameters.

Results

Classification of patients Thirty four patients with thalassemia major onregular transfusion treatment participated in thestudy. For glucose metabolism status, they wereclassified into three major groups according toWHO criteria: Group I: 17 with normal glucosetolerance (NGT); Group II: 3 with impaired glu-cose tolerance (IGT); and Group III: 14 with dia-betes. According to pre-transfusion levels of HbF, the groups were further subdivided into threesubgroups; a) with Hb F < 5%; b) Hb F 5-10%and c) Hb F > 10-20%. The majority of patients 24/34 (70%) had pre-transfusion Hb F levels < 5%, indicating an effi-cient suppression of bone marrow dyserythro-poietic hyperactivity. Five patients with Hb F lev-els between 10-20% had genotypes whichinduce γ chain synthesis (Table 1).

Pre-transfusion (Pre-Tx) Hb A1c levelsThe results of Hb A1c assessment prior to trans-fusion in 34 patients with homozygous β-tha-

(Pre-Tx) and another one hour after transfusion(Post-Tx). On the following transfusion day, twoweeks later, the same process was repeated.

Methods

For the evaluation of the clinical status of thepatients a series of laboratory investigations wereperformed, using appropriate methods appliedin our Unit for diagnosis and treatment of tha-lassemia and its complications. Blood chemistry included: i) Determination of fasting glucose and fruc-

tosamine levels using the Siemens Advia 1800Clinical Chemistry System (Siemens HealthcareDiagnostics, Tarrytown, NY, USA).

ii) Insulin levels assessment with an electro-chemiluminescence immunoassay on theRoche ELECSYS 2010 immunoassay analyzer(Roche Basel, CH).

iii) Whole blood Hb A1c and Hb F levels weremeasured with cation exchange HPLC(HA8121 HPLC system, Arkray Inc, Kyoto,Japan).

iv) Oral glucose tolerance test (OGTT) was per-formed according to WHO recommenda-tions (11).

According to plasma glucose concentration inmmol/l or mg/dl, patients were classified intothree groups:

Group I: Normal glucose tolerance (NGT): fasting plasma glucose (FPG): < 6.1 mmol/L (< 110 mg/dl) or 2hPG< 7.8 mmol/L (< 140 mg/dl).

Group II: Impaired glucose tolerance (IGT):FPG > 6.1 -7.0 mmol/L (< 110-126mg/dl) or 2h PG > 7.8-11.1 mmol/L(> 140-200 mg/dl).

Group III: Diabetes Mellitus: FPG > 7.0 mmol/L(> 126 mg/dl) or 2h PG ≥ 11.1mmol/L (≥ 200 mg/dl).

Statistical analyses

Data are presented as mean ± SD, and the level ofstatistical significance was considered at p < 0.05.All the statistical procedures were performedusing the STATGRAFICS PLUS version 5.1 forWindows program (Graphic Software System). We used the standardized skewness and stan-dardized kurtosis, to determine whether the sam-

Table 1.

Classification of 34 transfusion dependent patients with thalassemia major on the basis of OGTT and pre-transfusion concentration of HbF %.

* Oral Glucose Tolerance Test evaluated at baseline for patients with Normal GlucoseTolerance and on diagnosis, for patients with diabetes and Impaired Glucose Tolerance.

Groups (OGTT)* Subgroups based on Hb F (%)

Hb Hb Hb

F < 5% F ≥ 5-10% F ≥ 10%

I. Normal (n = 17) 12 3 2

II. Impaired (n = 3) 2 - 1

III. Diabetic (n = 14) 10 2 2

Total patients (n = 34) 24 5 5

4

lassemia major, regularly transfused, are shownin Figure 1. The distribution of Hb A1c levels isillustrated separately for patients with NGT andfor those with diabetes and IGT.In the whole series the levels of Hb A1c variedfrom 5.6-9.8%, with significant differencesbetween patients with NGT versus those withimpaired tolerance and especially those with dia-betes (Table 2). In patients with NGT, Hb A1c levels ranged from5.6-6.9% (mean 6.5 and SD ± 0.3%) comparedto 6.5-9.8% (mean 7.8 and SD ± 1.1%) inpatients with diabetes. In patients with Hb Fbetween 5-20%, the distribution of Hb A1c lev-els were within the range of patients with low HbF. (Figure 1).The results demonstrated an over-lapping of the higher Hb A1c levels of patientswith NGT to the lower levels of patients with dia-betes on treatment with insulin. Analysis of therelation of fasting plasma glucose to Hb A1c lev-els in patients with diabetes showed a highly sig-nificant positive relation. All five patients withlow Hb A1c (< 7.0%), the three with IGT, hadnormal FPG, (< 100 mg/dl), indicating an effi-cient control of diabetes (Figure 2).The summarized data of Table 2 showed, that inaddition to significantly higher Hb A1c levels,patients with glucose disturbances were also sig-nificantly older than patients with NGT. Therewere no differences in regard to Pre-Tx hemoglo-

bin levels (mean 10.0 vs 10.1 g/dl), and HbF(mean 6.0% vs 3.6%).

Impact of transfusion on Hb A1c levelsTo study the impact of transfusion on Hb A1c, HbA1c levels were assessed prior and one hour after

Endo-Thal


Table 2.

Relevant data of 34 patients with thalassemia majorclassified, in two groups, based on glucose metabolism status.

Group A: with NGT; Group B: with IGT and diabetes.(Values expressed as mean ± SD, range in parenthesis).

Statistical Analysis : NS = non-significant.Group A: HbA1c pre vs post transfusion p < 0.001, pair-observation t-test.Group B: HbA1c pre vs post transfusion p < 0.001, pair-observation t-test.

Group A Group B Difference

(n = 17) (n = 17) (p)

Age 14.9 ± 13.0 40.6 ± 6.6 < 0.001

(3.0-39.0) (31.0-55.0)

Pre-Tx Hb (g/dL) 10.1 ± 0.7 10.0 ± 0.8 NS

(8.6-11.8) (8.3-11.6)

Pre-Tx HbF (%) 3.6 ± 3.6 6.0 ± 6.3 NS

(0.3-12.9) (0.3-20.0)

Pre-Tx HbA1c (%) 6.5 ± 0.3 7.8 ± 1.1 < 0.001

(5.6-6.9) (6.5-9.8)

Post-Tx HbA1c (%) 6.3 ± 0.2 7.3 ± 0.8 < 0.001

(6.0-6.6) (6.1-8.7)

Figure 2.

Relation of Hb A1c levels to fasting plasma glucose in thalassemiapatients with diabetes and impaired glucose tolerance.

Figure 1.

Distribution of pre-transfusion Hb A1c levels in two groups of thalassemia patients, one with normal glucose tolerance (17 pts)

and another with diabetes and impaired tolerance (17 pts).

5

transfusion in 14 patients with NGT, 14 with dia-betes and 3 with IGT. The levels of Hb A1c were significantly reducedafter transfusion in both groups. In patients withNGT, HbA1c levels reduced from 6.5 ± 0.3 to 6.3± 0.2% (p < 0.001), and in patients with diabetesand IGT from 7.8 ± 1.1 to 7.3 ± 0.8% (p < 0.001)(Table 2).The reduction of pre-transfusion Hb A1c levelsafter transfusion, in all but one patient with NGT,indicate that Hb A1c levels in donors’ erythro-cytes were generally lower than those of the recip-ients which are older and of shorter life span. The impact of transfusion of packed red cells onHb A1c depended not only on donor’s A1c, butalso on the total volume of red cells transfusedand on the recipient’s Hb A1c levels. It is alsoclear that soon after transfusion a rearrangementof HbA1c levels occurs in the recipient’s blood.Interesting were the wider variation in the reduc-tion of Hb A1c in patients with diabetes (range0.2-1.1; mean 0.55%) versus patients with NGT(range 0-0.4; mean 0.24%).

Post transfusion changes of Hb A1c during interval between transfusionsThe influence of average glucose concentration

during the two week interval between transfusionswas evaluated by the assessment and comparisonof post transfusion Hb A1c levels to those of thefollowing pre-Tx2 levels, in 14 patients with dia-betes and 11 with NGT (Table 3).During the interval period of two weeks the posttransfusion levels of Hb A1c increased in all butone patient with NGT (the same patient that hadno Post-Tx reduction). As expected the variation was higher in the groupwith diabetes. More precisely, in patients with NGT changes ofHbA1c ranged from +0.1 to +0.4% (mean +0.18%), while in patients with diabetes from+0.1 to +1.2% (mean +0.5%). In both groupsand especially in the NGT group, the final Pre-Tx2 levels of Hb A1c returned to that of the ini-tial Pre-Tx level for the particular patient, innearly all patients. (The pair observation testsfor Pre-Tx vs. Pre-Tx2 for the two groups ofpatients were not significant: p > 0.144 for dia-betes and p > 0.830 for NGT).

Discussion

The credibility of Hb A1c as a gold standard forthe measurement of control of diabetes in patients



Endo-Thal

Table 3.

Serial assessment of Hb A1c on Pre-Tx, Post-Tx and Pre-Tx2* in 14 thalassemia patients with diabetes and IGT and 11 with normal glucose tolerance.

Statistical AnalysisDiabetes: Pre-Tx vs PreTx2; pair observation t-test p > 0.144. NGT: Pre-Tx vs PreTx2; pair observation t-test p > 0.830

GROUPS

Diabetes Normal glucose tolerance

Hb A1c (%)

Pre-Tx Post- Tx Pre-Tx2 Pre-Tx Post- Tx Pre-Tx2

6.6 6.4 6.6 6.4 6.2 6.1

9.5 8.3 9.5 6.5 6.2 6.6

8.8 8.1 8.8 6.6 6.4 6.7

7.8 7.1 7.5 6.6 6.4 6.5

7.1 6.9 7.1 6.8 6.5 6.6

9.4 8.5 8.9 6.9 6,6 6.5

7.8 7.3 7.7 6.0 6.0 6.4

6.9 6.7 7.0 6.6 6.5

8.2 7.4 8.4 6.6 6.3 6.5

9.8 8.7 8.6 6.2 6.0 6.0

8.2 7.5 8.0 6.6 6.5 6.7

7.6 7.2 7.3

8.1 7.4 8.1

7.5 7 7.7

Mean 8.09 7.46 7.94 6.52 6.31 6.46

SD 0.98 0.69 0.83 0.25 0.21 0.22

Range 6.6-9.8 6.4-8.7 6.6-9.5 6.0-6.9 6.0-6.6 6.0-6.7

6

with hemoglobinopathies, mainly thalassemia andsickle cell disease, has been questioned (1, 5, 8, 9). This was basically due to the abnormal hemoglo-bin composition of patient’s red cells, which inthe case of a non-transfused patient, containsminimal, if any, normal Hb A. However in transfusion dependent patients withhomozygous β-thalassemia, who are highly pre-disposed to diabetes, the hemoglobin composi-tion of patients’ erythrocytes are considerablymodified, because of regular and frequent trans-fusions. As a rule, the patient’s erythrocytes are amixture of transfused red cells from 2-6 donorswith a normal hemoglobin composition, with HbA of around 95%, and Hb F of 2-3%. Storage ery-throcytes have functional and metabolic differ-ences as well as a considerably shorter life spancompared to normal red cells.Levels of Hb A1c, assessed prior to transfusion,varied significantly between the two groups ofpatients with thalassemia studied. In patientswith NGT, Hb A1c ranged from 5.6-6.9%, (mean6.5 ± 0.3%), versus 6.5-9.8% (mean 7.8 ± 1.1%)in patients with diabetes (p < 0.001). Overlapping Hb A1c values between high levelsof NGT patients and lower of patients with dia-betes, were further evaluated on the relation offasting plasma glucose to Hb A1c levels, inpatients with impaired tolerance and diabetes. A highly significant positive relation (r = 0.932,p < 0.001) was found; all patients with diabetesand normal fasting plasma glucose (< 6.0mmol/L) had an Hb A1c level < 7.0%, which,presumably, was the result of efficient treatment.In this small series of regularly transfusedpatients with thalassemia major, Hb A1c level of7.0%, seems to be the cut off level of differenti-ation of patients with normal glucose metabo-lism to those with disturbed and basically withdiabetes. In addition, the mean Hb A1c level of 6.5% inpatients with normal tolerance is higher fromthat of normal individuals, related probably tothe metabolic and age differences of storage oftransfused red cells. The differences in the impact of transfusions onHb A1c levels in the two groups are interesting.One hour post transfusion Hb A1c levels wereconsiderably reduced in both groups. The rate of reduction depends on the one handon the volume and Hb A1c level of transfusedred cells and on the other on the A1c level andthe age of red cells of the recipient. In all patients

with NGT, except one, there was a mild, but sta-tistically significant reduction of Hb A1c aftertransfusion, indicating that in general, HbA1clevels of transfused red cells were lower thanthose of the recipient. These differences may be related to the youngerage of storage transfused red cells, compared tothat of the patients. Studies on the relation of Hb A1c to the averageglucose concentration during the two weeksinterval between transfusions clearly showed anincrease of the mean post transfusion Hb A1c,compared to the mean Hb A1c level of the fol-lowing pre-transfusion (Pre-Tx2) assessment.The mean Hb A1c Post-Tx versus Pre-Tx2increase was 6.3 vs 6.5% for patients with nor-mal tolerance and 7.5 vs 7.9% for patients withdiabetes. These differences correspond to the influence ofaverage glucose concentration for the last twoweeks. In contrast a single Pre-Tx Hb A1cassessment corresponds to the influence of average glucose concentration for the previousperiod of the life span of patient’s transfused redcells.Serial pre and post transfusion assessment of HbA1c levels in eight patients, on four to six con-secutive transfusion days, for a period of 56-85days, showed stable individual values for bothPre-Tx and Post-Tx estimations; p > 0.998,ANOVA repeated measures (unpublished data).These findings strongly support the credibility ofHb A1c assessment as a valuable marker to fol-low metabolic glucose disturbances and monitortreatment of diabetes in transfused patients withthalassemia.

Conclusions

The results of this study clearly showed thatassessment of Hb A1c could be used as a reliablemarker to diagnose and monitor treatment ofglucose disturbances and basically diabetes, inregularly transfused patients with thalassemiamajor. To this end, certain peculiarities of transfusederythrocytes, related to storage, should be con-sidered; these are the significantly shorter lifespan, the normal hemoglobin composition and anumber of metabolic changes which may proba-bly be related with higher mean Hb A1c inpatients with normal glucose metabolism.

Endo-Thal


7

References1. Definition and diagnosis of diabetes mellitus and intermediate

hyperglycaemia. Geneva: World Health Organization; 2006.

2. International Expert Committee Report on the role of HbA1c in the diagnosis of diabetes. Diabetes Care 2009;32:1327-1334.

3. Nathan DM, Turgeon H, Regan S. Relationship between glu-cated hemoglobin levels and mean glucose levels over time.Diabetologia 2007; 50:2239-2244.

4. Juvenile Research Foundation. Continuous Glucose MonitoringStudy Group. Hemoglobin A1c and mean glucose in patientswith type 1 Diabetes. Diabetes Care 2011; 34:540-544.

5. Callanger EJ, Bloomgarden ZT, Le Roith D. Review of hemo-globin A1c in the management of diabetes. Journal ofDiabetes 2009; 1:9-17.

6. Goldstein DE, Little RR, Lorenz RA, et al. Tests of glycemiain diabetes, Diabetes Care 2004; 27:1761-1767.

7. Taps RJ, Tikellie G, Wong TY, et al. Longitudinal associationof glucose metabolism in retinopathy; results from theAustralian Diabetes, Obesity and Lifestyle (Aus. Diab.)study. Diabetes Care 2008; 31:1349-1354.

8. Gunton JE, McElduff A. Hemoglobinopathies and Hb A1cmeasurement. Diabetes Care 2000; 23:1197-1198.

9. Smaldon A. Glycemic control and hemoglobinopathy: whenA1c may not be reliable. Diabetes Spectrum 2008; 21:46-49.

10. Voyiatzi MG, Macklin EA, Tratchtenberg FL, et al.Differences in prevalence of growth, endocrine and vitaminD abnormalities among various thalassemia syndromes inNorth America. Br J Haematology 2009; 146:546-556.

11. World Health Organization. Definition, Diagnosis andClassification of Diabetes Mellitus and its Complications.Part 1. Diagnosis and classification of Diabetes Mellitus.WHO/NCD/NCS/99.2 ed. Geneva. World Health Organi -zation, 1999.



Endo-Thal

Correspondence:

Christos Kattamis, MD

Emeritus Professor of Pediatrics,

First Department of Pediatrics, Athens University Medical School,

Aghia Sophia Children’s Hospital, Athens, Greece

Tel: +30 -210 9823639 - Fax: +30-210 77955539

E-mail: [email protected]; [email protected]

Agosto 2008

SummaryWe report a thalassaemia major patient who presented with congestive heart failure secondary to hypoparathyroidism. Shewas severely iron overloaded (serum ferritin 9620 ng/ml). Intravenous calcium gluconate, oral vitamin D and intensiveiron chelation therapy induced correction of hypocalcemia and improvement of cardiac functions. Our observations stressthe importance of regular chelation therapy, early diagnosis of endocrine complications, and close follow-up of thalas-saemia major patients with hypocalcemia.

Key words: thalassaemia, hypoparathyroidism, hypocalcemia, iron overload, vitamin D.

Reversible heart failure in a thalassaemicpatient secondary to hypocalcemia and hypoparathyroidismVincenzo De Sanctis, Monica Sprocati, Maria Rita Govoni, Giuseppe Raiola1

Departments of Reproduction and Growth - Paediatric and Thalassaemia Unit - St. Anna Hospita - Ferrara; 1Department of Paediatrics, Pugliese-Ciaccio Hospital - Catanzaro






Comitato Editoriale









Com

itato

Edit

ori

ale

40

Emothal


Morbidity due to cardiovascular, endocrinologicaland hepatic disease is considerable in ‚-thalas-saemia major (TM)(1,2), whereas heart failureconstitutes the leading cause of mortality.The primary cause of cardiac dysfunction in tha-lassemia is believed to be myocardial iron over-load(1), caused by blood transfusion andincreased gastrointestinal iron absorption (2). Theunbound iron, via the Fenton and Haber-Weissreaction, has high tissue toxicity through the for-mation of oxygen free radicals (2). Oxyradicalsdamage cell lipids, nucleic acid and proteins.Lipid peroxidation is associated with impairmentof mitochondria, lysosomes, microsomes andmembrane function, leading to cell necrosis anddeath (3-5). Iron accumulates at different rates invarious organs, and these organs show a differentsusceptibility to the damage induced by reactiveiron species such as nontransferrin bound ironand the intracellular labile iron pool.Besides iron, other factors may play a role in theimpairment of myocardial contractility, includingprolonged heart tissue hypoxia (6), pericardialinvolvement (7), arrhythmias (1), endocrinecomplications (8-13) and vitamin D deficiency(10, 14).We report a young adult thalassaemia majorpatient who developed heart failure as a conse-quence of hypocalcemia seconday to hypopara-thyroidism.

Case report

A 22 year-old prepubertal patient withTM was admitted to our Unit because of conges-tive cardiac failure.Over the previous six days she had been treatedin another hospital with digoxin and furosemidewith no improvement in her cardiac status. On examination, her temperature was 36.8° C,systemic arterial blood pressure was 90/50mmHg, pulse rate was regular (115 beats/min)and respiratory rate was 25 breaths/min. Cardiacexamination revealed a grade 2/6 apical holosys-tolic murmur. Chest auscultation revealed crack-les at the basilar lobe. Liver edge was palpable 3cm below the right costal margin. Chvostek’s andTrousseau’s signs were positive. Thyroid glandwas not enlarged.Initial laboratory investigations were significantfor the following: haemoglobin level 8.2 g/dl,

serum creatinine 1.2 mg/dl (normal 0.7-1.3mg/dl), total calcium 5.2 mg/dl (normal 8.8-10.6mg/dl), inorganic phosphorous 10.2 mg/dl (nor-mal 2.5-4.5 mg/dl), magnesium 2.1 mg/dl (nor-mal 1.8-2.6 mg/dl), albumin 36 g/l (normal 35-50 g/l) and intact parathormone level 11 pg/ml(normal 10-55 pg/ml). Liver enzyme ALT was120 IU/l (normal range 7-40 IU/l) and serum fer-ritin level was 9620 ng/ml (normal range 32-176ng/ml).On X-ray the heart was enlarged and the electro-cardiogram revealed a prolonged QTc interval(0.45 sec.).Echocardiogram showed bilateral ventricularenlargement, generalized hypokinesia, moderatemitral and tricuspid regurgitation and systolicdysfunction, with an ejection fraction of 24.4%.A diagnosis of congestive cardiac failure,hypoparathyroidism and severe iron overloadwas made.The patient was treated with intravenous calciumgluconate, oral vitamin D (calcitriol), intensiveiron chelation therapy (Desferal given subcuta-neously), blood transfusions and diuretics (thi-azides).Over the next 5-6 days, serum total calcium con-centration slowly increased to 7.1 mg/dl andplasma inorganic phosphorous concentrationdecreased to 7.3 mg/dl. Restoration of normocal-cemia resulted in clinical and cardiac improve-ment: left ventricular ejection fraction was 38%after 3 weeks and 42% after 4 weeks.The patient was discharged after 34 days withoral calcium, vitamin D and subcutaneous chela-tion therapy (Desferal 40 mg/kg body weight, sixtimes per week).We concluded that the patient’s congestive heartfailure was precipitated by severe hypocalcemia,secondary to hypoparathyroidism.

Discussion

Hypocalcemia due to hypoparathy-roidism is a late complication of iron overloadand is rare in well-chelated patients (14).Hypocalcemia follows as the consequence of irondeposition in the parathyroids (15). The majori-ty of patients have mild disease, with symptomsincluding parasthesias, while in more severecases, tetany, seizures or cardiac failure mayoccur (11).

41

Calcium and PTH play a key role in the mainte-nance and regulation of normal cardiac function(16). The PTH effect is probably due to the entryof calcium into myocardial cells and the releaseof endogenous myocardial norepinephrine (16).Intravenous calcium replacement is necessarywhen calcium levels decline below 1.9mmol/litre, and in symptomatic patients. Thistreatment should be given in hospital understrict calcium and ECG monitoring (17).Intravenous administration of calcium is notwithout problems. Patients taking digoxin mayhave increased sensitivity to the treatment. Inaddition, rapid calcium administration can resultin arrhythmias and vein irritation.One possible treatment regimen is the infusion(over 10 minutes) of 10% calcium gluconate (10-20 ml, diluted in 50-100 ml of 5% dextrose)repeated as necessary to control symptomatichypocalcemia (17). To avoid persistent hypocal-cemia, oral calcium and vitamin D supplementsshould be started early (18). The overall aim of treatment is to maintainserum calcium in the low-normal range, andserum calcium should be tested every 3 to 6months. One potential side effect is hypercalci-uria with nephrocalcinosis and/or nephrolitiasis.A 24-hour urine calcium should be determinedat least annually (once a stable therapeutic doseis established) and levels should be below 4mg/kg/24 hr (19). It is important to keep inmind that furosemide, glucocorticoids andexogenous estrogen can depress serum calciumlevels and may influence calcium and vitamin Dreplacement therapy while thiazide diuretics canincrease renal calcium absorption in patientswith hypoparathyroidism (19).In conclusion, secondary hypoparathyroidism isa very rare cause of heart failure and must beremembered in the etiological assessment of tha-lassaemic patients with heart failure.Hypocalcemic cardiomyopathy does not respondto conventional treatment for heart failure butimproves with restoration of normocalcemia.Treatment includes calcium, oral vitamin D orone of its analogues, and a phosphate binder ifindicated.

References

1. Hahalis G, Alexopoulos D, Kremastinos DT,Zoumbos NC. Heart failure in ‚-thalassaemia syndromes: a decadeof progress. Am J Med 2005; 118:957-967.

2. Gabutti V, Piga A. Results of long-term iron-chelating therapy.Acta Haematol 1996; 95:26-36.

3. Halliwell B, Gutteridge JMC. Role of free radicals and catalyticmetal ions in human disease: an overview. Methods Enzymol1990; 186:1-85.

4. Link G, Hershko C. Rat heart cells in culture: a model of irontoxicity and chelation. J Lab Clin Med 1993; 122:14-16.

5. Iancu TC. Ultrastructural pathology of iron overload. ClinHaematol 1989; 2:375-395.

6. de Montalembert M, Mounaury C, Acar P, Brousse V, Sidi D,Lenoir G. Myocardial ischaemia in children with sickle cell disease.Arch Dis Child 2004; 89:359-362.

7. Aydin Y, Ozcabar L, Altundag K, Ustun I. Pericardial tampon-ade in a patient with thalassaemia major due to hypothyroidism.Acta Haematol 2006; 116:141-142.

8. Konrad D, Danemon D, Kirby M, Wherrelt D. Cardiac failureafter initiation of insulin treatment in diabetic patients with β-tha-lassaemia major. J Pediatr 2003; 143:541-542.

9. Erfurth ME, Holmer H, Nilsson PG, Kornhall B. Is growth hor-mone deficiency contributing to heart failure in patients with β-thalassaemia major. Eur J Endocrinol 2004; 151:161-166.

10. Wood JC, Claster S, Carson S, Menteer JD, Hofstra T, KhannaR, Coated T. Vitamin D deficiency, cardiac iron and cardiac func-tion in thalassaemia major. Br J Haematol 2008, in press.

11. Tsironi M, Korovesis K, Farmakis D, Deftereos S, Aessopos A.Hypoparathyroidism and heart failure in a thalassaemic patient: acase report. Pediatr Endocrinol Rev 2004; 2(Suppl. 2):310-312.

12. De Sanctis V, Borsari G, Brachi S, Gubellini E, Gamberini MR,Carandina G. A rare cause of heart failure in iron-overloaded tha-lassaemic patients - primary hypoparathyroidism. Georgian MedNews 2008; 156:111-113.

13. Fiorillo A, Farina V, Di Maio S, Ciriaco M, Nunziata L, ScippaL, D’Agostino E. Myocardial dysfunction in two hypothyroid twinswith thalassaemia major. Acta Paediatr Scan 1989; 78:455-457.

14. De Sanctis V, Vullo C, Bagni B, Chiccoli L. Hypoparathyrodismin beta-thalassaemia major. Clinical and laboratory observationsin 24 patients. Acta Haematol 1992; 88:105-108.

15. Wonke B. Clinical management of ß-thalassaemia major.Seminars Hematol 2001; 38:350-359.

16. Avsar A, Dogan A, Tavli T. A rare cause of reversible dilatedcardiomyopathy: hypocalcemia. Echocardiography 2004; 21:609-610.

17. Murphy E, Williams GR. Hypocalcemia. Medicine 2005;33:55-57.

18. Allgrove J. Parathyroid disorders. Current Pediatr 2001;11:357-363.

19. Fitzpatrick LA. Hypocalcemia: diagnosis and treatment.www.endotext.org


Reversible heart failure in a thalassaemic patient secondary to hypocalcemia and hypoparathyroidism

Emothal

Correspondence to:

Vincenzo De Sanctis, MDDirector of Paediatrics and Thalassaemia Unit, St. Anna Hospital Corso Giovecca, 203 - 44100 Ferrara (Italy)

e-mail: [email protected]

49

Clinical and radiographic diagnosis

Most of the patients are asymptomaticuntil irreversible changes have occurred. At latestages they usually present with symptoms of pain,limp, sudden fractures and neurological emergen-cies (1-5). Fractures are common with the meanage of occurrence being 18 years and abnormalbone density studies are widespread (6). The com-

plete skeleton is involved. The most fruitful areasto radiograph are hands, skull, thoracolumbarspine and chest. In infants and children,hematopoietic marrow is distributed throughoutthe axial and appendicular skeleton and, conse-quently, osseous changes may be present distallyas far as the phalanges. Abnormalities in the

Emothal

Abstract

In thalassaemic patients, osteopenia and osteoporosis represent prominent causes of morbidity in young adultsof both genders with thalassemia major and intermedia. Delayed diagnosis and inadequate treatment has led tosignificant osteoporosis, skeletal abnormalities, fractures, spinal deformities, nerve compression and growth fail-ure. Dual X-ray absorptiometry measurement of bone density of the lumbar spine and femoral neck to meas-ure bone mass is a quantitative and safe method. Bone specific alkaline phosphatase and plasma osteocalcin aremarkers of osteoblastic activity. Hormonal deficiency, iron overload, bone marrow expansion, nutritional defi-ciency and desferioxamine toxicity are important factors. Bone histomorphometry data reveal increased osteoid thickness, osteoid maturation time and lengthened miner-alization lag time, which indicate impaired bone matrix maturation and defective mineralization. Dynamic boneformation studies reveal reduced bone formation rate. Iron deposition in bones impairs osteoid maturation andinhibits mineralization locally, resulting in focal osteomalacia.Prevention of osteoporosis is the most important priority in managing bone disease. This includes:a) The efficacy of iron chelation and prevention of desferioxamine toxicity.b) Calcium intake of 1000 mg per day as early as the age of 10-12 years.c) Vitamin D (800 IU) per day is recommended for all patients.d) High caloric diet to compensate for the hypermetabolic state of thalassemic patients has a potentially benefi-

cial effect on bone accretion.e) Early correction of endocrine abnormalities in the form of treatment with GH and inducing pubertal devel-

opment (testosterone or HCG therapy in males and estrogen therapy in girls) at the proper time both aremandatory for attaining normal bone density.

f) Regular exercise for at least 30 minutes per day helps to reduce bone loss.

Treatment of established osteoporosis consists of the previous recommendations and osteoclastic inhibitors suchas bisphosphonates and calcitonin.As treatment with transfusion programs and chelation therapy has significantly prolonged survival in thalas-saemia patients, osteopenia and osteoporosis represent prominent causes of morbidity in young adults of bothgenders with thalassemia major and intermedia. The early identification of clinical and radiological abnormal-ities of skeletal dysplasia is of paramount importance in preventing severe bone destruction. Delayed diagnosisand inadequate treatment has led to significant osteoporosis, skeletal abnormalities, fractures, spinal deformi-ties, nerve compression and growth failure.

Bone disease in thalassemiaAshraf T. Soliman

Pediatric and Endocrinology Unit -Hamad Medical Centre, Doha (Qatar)

50

Emothal


appendicular skeleton become less pronounced asnormal developmental regression of red marrowoccurs with adequate treatment of the anemia (1). The skeletal response to marrow proliferationconsists of expansion of the medulla, thinning ofcortical bone and resorption of cancellous bone,which results in a generalized loss of bone densi-ty. Frequently, small areas of lucency resultingfrom focal proliferation of marrow may be pres-ent, often demarcated by coarsened but lessnumerous trabeculae. In addition, the hypertrophic and hyperplasticmarrow may perforate the cortex, proliferate sub-periosteally and stimulate a variety of periostealresponses. Depending on the bone, these factorsmay result in a variety of radiographic appear-ances (7-10).In vertebral bodies, the resorptive process pre-serves the primary trabeculae at the expense ofsecondary trabeculae resulting in a striatedappearance. Biconcavity of the superior and infe-rior margins of the vertebral bodies or compres-sion fractures may occur. In skull widening of thediploic space with a thinning of the corticesoccurs. New bone forms in response to marrowproliferation beneath the periosteum. These bonyspicules may be seen radiographically and resultin a classic “hair-on-end” appearance. In facialbones hyperactive hemopoesis within the frontalregion results in hypertrophy of osseous struc-tures and a consequent prominence of the lateralmargins of the malar eminences, together withanterior and medial displacement of developingteeth (7-10).CT may be required to further evaluate faciomax-illary changes and to clearly define expansilelesions of the pelvis resulting from extra-medullary hematopoiesis (11).In long and short-long bones cortical thinning,osteopenia and coarsening of the trabeculae arecommon. Fractures may occur, although lesscommonly than expected from the degree ofosteoporosis. In ribs widening, generalizedosteopenia or localized lucency occur. The presence of "rib-within-a-rib” appearance,subcortical lucency and bulbous expansion of theposterior segments of the ribs appears limited tothe ribs (10).Osteochondrodystrophic lesions, mainly affectinglong bone metaphyses, can be radiologically evi-dent in homozygotic thalassemic patients treatedwith deferoxamine and their incidence rate variesamong Authors (12, 13).

Bone density abnormalities are widespread inboth sexes. X-ray demonstration of osteoporosisoccurs after a loss of more than 30% of bone;therefore, it is of limited screening value. Dual X-ray absorptiometry measurement of bone densityof the lumbar spine and femoral neck to measurebone mass is quantitative and safe with availablecontrol data. Assessment of bone turnover,including bone formation and resorption, isimportant in understanding etiology as well aseffect of therapy. Bone specific alkaline phos-phatase and plasma osteocalcin are markers ofosteoblastic activity (4, 14, 15).

The pathology of bone disease

Bone histomorphometry revealedincreased osteoid thickness, osteoid maturationtime and lengthened mineralization lag time,which indicate impaired bone matrix maturationand defective mineralization. In addition, irondeposits appeared along mineralization fronts andosteoid surfaces. More over, focal thickenedosteoid seams are found together with focal irondeposits (16-18).Dynamic bone formation study reveal a reducedbone formation rate. These findings indicate thatdelayed bone maturation and focal osteomalaciaare important factors in the pathogenesis of bonedisease in a suboptimally transfused thalassemicswith iron overload. Hormonal deficiency, irondeposits in bone, bone marrow expansion, nutri-tional deficiency and desferioxamine toxicity areimportant factors (1, 3, 6, 13).The bone mineral density (BMD) depends uponbone mineral content and bone size. Particularrisk factors for bone disease include small, olderpatients, low baseline hemoglobin, delayedpuberty and/or gonadal failure, high iron stores,calcium/vitamin D deficiency, low circulatingIGF-I and IGFBP3 levels, diabetes mellitus andchronic hepatitis (19-24).However, the mechanisms of bone disease in tha-lassemia are not yet clearly known. Many expla-nations have been offered:

A) Marrow expansion causes mechanicalinterruption of bone formation as well as cytokinedisturbances, which result in increased boneresorption. Even in the highly transfused group,patients develop osteoporosis partly induced by

51

Ashraf T. Soliman

Bone Disease in Thalassemia

Emothal

continued but unexpected marrow hyperplasia.Transferrin receptor studies demonstrateincreased marrow activity even with reticulocyteabsence or hypoplasia (25).

B) Hormonal determinants of peak bonemass: These include variable affection of growthaxis, gonadal axis calcium homeostasis, glucosehomeostasis and thyroxin.

1. Disturbed growth axis: (GH) secretion may benormal or subnormal in thalassaemic patientswith short stature. However, even in the presenceof normal GH the IGF-I level is always low indi-cating a dysfunction of the GH-IGF1 axis in someof them. Several mechanisms have been suggestedto cause disorders of GH secretion:a) pituitary siderosis; b) neurosecretory GH dysfunction;c) hypothalamic growth-hormone-releasing hor-

mone deficiency and/or increased somatostatinactivity;

d) partial resistance to GH (26);GH/IGF-I axis plays a critical role in the regu-lation of bone accretion throughout prepuber-tal, pubertal, and postpubertal growth phasesand in regulating the raise in peak BMD andbone size that occurs during puberty. IGF-I isan important growth factor for osteoblasts andis important for normal 1,25-(OH)D action(22, 24, 26).

2. Delayed puberty/hypogonadism: High preva-lence of delayed puberty and hypogonadismoccur universally in thalassemic patients. Hypogonadotropic hypogonadism is due to dam-age from iron deposition in the hypothalamus andpituitary gland but primary gonadal failure canalso occur. Both timing as well as quantity ofgonadal steroids are critical for bone acquisition.Gonadal steroids are important not only to bonemaintenance but also to acquisition. During puberty, estrogen and testosterone levelsrise and contribute to consolidation of bone mass(26-28).Sex steroid deficiency in thalassemic patients con-tribute to low peak bone mass and lead to subse-quent risk for osteoporosis. Estrogen blocks acti-vation of the bone metabolic unit. Osteoclastapoptosis also appears to be regulated by estro-gens. In patients with estrogen deficiency, theosteoclasts live longer and are therefore able toresorb more bone. Estrogens also retard the bone

resorbing effects of PTH. Estrogens are also neces-sary for epiphyseal closure. In patients withdelayed puberty, the GH response to GHRH hasbeen reported to be impaired, but the responsemay improve with the onset of spontaneous orinduced puberty with sex steroids orgonadotropin (28, 29).

3. PTH/Vitamin D abnormalities in thalassemia:Hypoparathyroidism due to iron overload maydevelop in a significant number of thalassemiamajor patients, especially when chelation therapyis not optimal. However, it is a late complicationof the iron overload. The majority present afterthe age of 16 years and both sexes are equallyaffected. In addition, IGF-I deficiency may lead to1,25-dihydroxyvitamin D deficiency even undernormal calcium diet. Low PTH and qualitativeand/or quantitative abnormalities of 1,25 dihy-droxy vitamin D elicit low bone formation andcan variably contribute to bone disease in tha-lassemic patients (30, 31).

C) Nutritional status: Children with tha-lassaemia major, even without endocrinopathy orcardiomyopathy, are significantly shorter andhave smaller mid-arm circumference and skin-fold thickness compared to normal age-matchedchildren denoting a degree of undernutrition,despite apparently normal food intake. This can be explained in part by the hypermeta-bolic status of these children with increased rest-ing metabolic rate and oxygen consumption sec-ondary to anemia, increased cardiac work andbone marrow hyperactivity. Growth improvementand increased IGF-I production after introducinghigh-caloric diet proved that part of the growthimpairment of these patients is correctable byproper nutritional intervention to compensate fortheir hypermetabolic status . Body mass also pre-dicts osteoporosis. Therefore, small, thin patientsare at particular risk (32, 33).

D) Iron toxicity: A study of the effect ofiron overload on bone remodeling in animalsshowed decreased osteoblast recruitment and col-lagen synthesis, resulting in decreased bone for-mation. Iron deposition along the mineralizationfronts was associated with increased osteoidseams. Focal thickened osteoid seams were found togeth-er with focal iron deposits along osteoid surfaces. Therefore, iron deposition in bone impairs

52

Emothal


osteoid maturation and inhibits mineralizationlocally, resulting in focal osteomalacia (34). The mechanism by which iron interferes withosteoid maturation and mineralization may beexplained by the incorporation of iron into crys-tals of calcium hydroxyapatite, which conse-quently affects the growth of calcium hydroxya-patite crystals and increases osteoid in bone tis-sue (34).

E) Desferrioxamine-induced skeletaldysplasia: It is now recognized that skeletal dys-plasia can be induced by injudicious use of des-ferrioxamine (DFO). Desferrioxamine-inducedskeletal dysplasia manifests as short trunk, genuvalgum, metaphyseal widening of long bones,joint stiffness and a decreased growth velocity. Radiological changes include thickened growthplate with widening and cupping of the metaphy-ses of long bones, sclerosis of subchondral bone,osteoporosis and small radiolucent metaphyseallesions. Desferrioxamine can inhibit DNA synthesis,fibroblast proliferation and collagen formation.The toxic effects of desferrioxamine can also bedue to a deficiency of trace elements such as cop-per and zinc (13, 35).

F) Genetic type of thalassemia:Thalassemic patients with the VDR BB genotypehad lower spine BMD than patients with the bbgenotype. The BsmI VDR gene polymorphism isassociated with osteopenia in thalassaemia (17).

Management of bone diseasein thalassemia

Initial bone density by Dual–energy X-ray absorptiometry (DEXA) or other acceptedmethods of measurement should be performed at8 years for girls and for boys, and annually there-after. The same method of bone density measurementshould be used for each evaluation. The absolutevalue of bone density varies but the accepted def-initions are:

1. Bone density > 1.5 SD below themean is diagnostic of osteopenia.

2. Bone density > 2.5 SD below themean is diagnostic of osteoporosis.

A - Prevention of osteoporosis is the most importantpriority in managing patients since restoring healthybone to an osteoporotic skeleton safely is not wellproven.

a) Effective management of iron overloadrequires frequent evaluation of the body ironstores. Serum ferritin measurement, up till now, isstill the predictor iron stores. In order to balancebetween the efficacy and toxicity of DFO, the tox-icity index (mean daily dose of desferrioxamine inmg/kg¸ serum ferritin in mg/L) should be moni-tored on a regular basis and should not exceed0.025. Direct assessment of hepatic iron content(HIC) by liver biopsy is the best predictor of thetotal body iron, but the procedure is invasive,risky and difficult to perform repeatedly (36).SQUID is an alternative method to assess liveriron overload. DFO and the orally effective deferiprone (DFP),either alone or together, are effective tools for ironchelation. The two together probably form thebest modality of iron chelation. DFP mobilizesiron from the stores while DFO puts it out of thebody.In recent years many effective treatments and pre-vention plans have been discovered.The most common are:

b) Calcium: The most important recom-mendation is to increase calcium intake, eitherthrough dietary changes or supplemental pills. Itis best for patients to begin adequate calciumintake at an early age, as bone mass begins todecrease around the age of 30. Calcium helpsdecrease bone loss, strengthens bones, and de-creases the risk of fractures. The recommendeddaily intake for thalassemics is 1,000 mg per daystarting at the age of 10-12 years (37).Determination of urinary calcium excretion be-fore calcium supplementation is recommended.

c) Vitamin D: At least 800 IU per day isrecommended for all patients with thalassemia.Many calcium supplements contain vitamin D.Patients can increase vitamin D intake throughfoods such as: egg yolks, fortified milk, cerealsand fish (38).

d) Nutrition: Nutritional deficiencies arecommon in thalassemia. High caloric diet to com-pensate for the hypermetabolic state of tha-lassemic patients proved to increase body weighand IGF-I levels with potentially beneficial effecton bone accretion (32, 33).Annual evaluation by registered dietitian with rec-ommendation for diets high in calories, low in

53

iron and adequate in calcium, vitamin D, traceminerals (cooper, zinc, and selenium) and antiox-idant vitamin E is recommended. Counseling forpatients with impaired glucose tolerance and dia-betes mellitus by a registered dietitian andendocrinologist is necessary.

e) Early and proper correction ofendocrine abnormalities:1. Growth Hormone deficiency/resistance:

Endocrine evaluation is required if there isfall-off on growth curve, height < 5th centilefor age and sex or poor growth velocity forage. Low IGF-1 or IGF BP-3 requires investi-gation by growth hormone provocative test-ing. Early diagnosis, and treatment of GH defi-ciency and/or resistance is recommendedbefore completion of puberty is recommend-ed. GH therapy has increases significantlyIGF-I levels and linear growth in thalassemicpatients (26, 27, 36).Theoretically, treatment with IGF-I appearsideal for thalassemics with low IGF-I secretionand partial GH resistance but needs to be ade-quately evaluated.

2. Hypogonadism: Girls without evidence ofpubertal development by 13 years and boys by14 years require testing for the hypothalamic-pituitary gonadal axis. In both sexes, hypogo-nadotropic hypogonadism is common andshould be treated early to assure normalgrowth and bone accretion (39-43). In boys,testosterone or human chorionic gonadotropin(HCG) can be used. If concomitant testicularsiderosis is significant, then testosteronereplacement is necessary. The starting dose isusually 100–150 mg given monthly, intramus-cularly. Testosterone dose needs to be adjustedperiodically for patients’ age and pubertal sta-tus as well as for sexually active men. Estrogenreplacement is recommended for hypogonadalgirls. Premarin or Ethynilestradiol at a low doseare given. The dose is increased after 6 months,for additional 6-12 months, afterwards an oralcontraceptive pill can replace Premarin orEthynilestradiol.

3. Hypoparathyroidism: Calcium homeostasisshould be evaluated annually with serum ion-ized calcium, phosphorus, alkaline phos-phates, parathyroid hormone, and 25-hydroxyvitamin D screening.A normal low PTH with decreased calcium andhigh phosphorus is diagnostic of hypoparathy-roidism. Therapy with an activated 1,25 OH

VitD product and calcium is necessary tomaintain normal serum calcium and boneintegrity (44).

4. Insulin-dependent diabetes mellitus (IDDM) isa frequent complication in patients with beta-thalassaemia major. Although IDDM has nosignificant detrimental effect on bone mineraldensity and metabolism. However, IDDMincreases IGFBP-3 proteolytic activity andreduce IGF-1 levels. Good glycemic control isassociated with increasing IGF-1 level (45).

f) Exercise: Exercise five days a week forat least 30 minutes daily helps to reduce boneloss. The best exercises for maintaining bone massare weight-bearing exercises.

B - Treatment of established osteoporosis. More puz-zling however, is the observation that, despite the nor-malization of haemoglobin levels, adequate hormonereplacement and effective iron chelation, patients con-tinue to show an unbalanced bone turnover with anincreased resorptive phase resulting in seriouslydiminished bone mineral density (BMD). For patientswith established osteoporosis: (DEXA score of >2.5standard deviations), in addition to hormone replace-ment (estrogen for females, testosterone for males),calcium and vitamin D supplementation, exerciseprogram, treatment should be considered.a. Bisphosphonates: These compounds inhibit

breakdown of bone and slow down boneremoval. They are also shown to increase bonedensity and decrease the risk of fractures atboth the hip and spine. A significant improve-ment in BMD was observed in most patientstreated with pamidronate. Administration of30 mg of pamidronate at monthly intervalsresulted in a significant increase of the BMD ofthe lumbar spine in all patients, but not theBMD of the femoral neck and the forearm.Whereas administration of 60 mg ofpamidronate showed a similarly significantincrease in the BMD of the lumbar spine inboth transfusion-dependent and transfusion-independent patients. Administration of pamidronate was also fol-lowed by a clear decrease in the markers ofbone resorption. In addition, alendronate andzoledronic acid seem to have good efficacy (46-48).

b. Calcitonin: Calcitonin prevents trabecularbone loss by inhibiting osteoclastic activity.Parenteral and intranasal instillation is avail-

Ashraf T. Soliman


Emothal

54

Emothal


References

1. Pootrakul P, Hungsprenges S, Fucharoen S, Baylink D et al.Relation between erythropoiesis and bone metabolism in tha-lassemia. N Engl J Med 1981; 304: 1470.

2. Korovessis P, Papanastasiou D, Tiniakou M, Beratis NG.Incidence of scoliosis in ß-thalassemia and follow-up evaluation.Spine 1996; 21: 1798.

3. Olivieri NF, Koren G, Harris J, Khattak S et al. Growth failureand bony changes induced by deferoxamine. Am J Ped HematolOncol 1992; 14: 48.

4. Orvieto, R, Leichter I, Rachmilewitz EA, Margulies JY. Bonedensity, mineral content and cortical index in patients with tha-lassemia major and the correlation to their bone fractures, bloodtransfusions and treatment with desferrioxamine. Calcif Tissue Int1992; 50: 397.

5. Pearson HA. The evaluation of thalassemia intermedia. InProceedings of thalassemia intermedia: A region I conference. TheGenetic Resource: Special Issue 1997; 11: 5.

6. Singer ST, Vichinsky A. Bone disease in beta-thalassemia.Lancet 1999; 354: 881.

7. Caffey J. Cooley's anemia: a review of the roentgenographicfindings in the skeleton. AJR 1957; 78: 381.

8. Korsten J, Grossman H, Winchester PH. Extramedullaryhematopoiesis in patients with thalassemia anemia. Radiology1970; 95: 257.

9. Lawson JP, Ablow RC, Pearson HA. Calvarial and phalangealvascular impressions in thalassemia. AJR 1984; 143: 641.

10. Lawson JP, Ablow RC, Pearson HA. The ribs in thalassemia.II. The pathogenesis of the changes. Radiology 1981; 140: 673.

11. Long JA Jr, Doppman JL, Nienhus AW. Computed tomograph-ic analysis of beta-thalassemic syndromes with hemochromatosis:pathologic findings with clinical and laboratory correlations. JComput Assist Tomogr 1980; 4: 159.

12. Rodda CP, Reid FD, Johnson S, Doery J et al. Short stature inhomozygous ß-thalassaemia is due to disproportionate truncalshortening. Clin Endocrinol 1995; 42: 587.

13. De Virgilis ST, Congia M, Frau F et al. Deferoxamine inducedgrowth retardation in patients with thalassemia major. J Pediatr1988; 113: 661.

14. Consensus Development Conference. Diagnosis, prophylaxisand treatment of osteoporosis. Peck WA, Chairman. Am J Med1993; 94: 646.

15. Lindsay R. Prevention and treatment of osteoporosis. Lancet1993; 341: 801.

16. Mahachoklertwattana P, Sirikulchayanonta V, ChuansumritA, Karnsombat P et al. Bone histomorphometry in children andadolescents with ß-Thalassemia disease: iron-associated focalosteomalacia. J Clin Endocrinol Metab 2003; 88: 3966.

17. Wonke B, Jensen C, Hanslip JJ, Prescott E, Lalloz M et al.Genetic and acquired predisposing factors and treatment of osteo-porosis in thalassaemia major. J Pediatr Endocrinol Metab 1998;11 (Suppl 3):795.

18. De Sanctis V, Savarino L, Stea S, Cervellati M et al.Microstructural analysis of severe bone lesions in seven tha-lassemic patients treated with deferoxamine. Calcif Tissue Int2000; 67:128.

19. Soliman AT, Banna NE, Fattah MA, EI Zalabani MM, AnsariBM. Bone mineral density in prepubertal children with ß-tha-lassemia: correlation with growth and hormonal data. Metabolism1998; 47: 541.

20. Anapliotou ML, Kastanias IT, Psara P, Evangelou EA et al.The contribution of hypogonadism to the development of osteo-porosis in thalassemia major: new therapeutic approaches. ClinEndocrinol (Oxf) 1995; 42:279.

21. Heaney RP. Nutritional factors in osteoporosis. Ann Rev Nutr1993; 13: 287.

22. Low LCK, Postel-Vinay MC, Kwan EYM, Cheung PT. Serumgrowth hormone (GH) binding protein, IGF-I and IGFBP-3 inpatients with ß-thalassemia major and the effect of GH treatment.Clin Endocrinol (Oxf) 1998; 48: 641.

23. Pollak RD, Rachmilewitz E, Blumenfeld A, Idelson M et al.Bone mineral metabolism in adults with ß-thalassemia major andintermedia. Br J Haematol 2000; 111: 902.

24. Soliman AT, Banna NE, Ansari BM. GH response to provoca-tion and circulating IGF-I and IGF-binding protein-3 concentra-tions, the IGF-I generation test and clinical response to GH ther-apy in children with ß-thalassaemia. Eur J Endocrinol 1998; 138:394.

25. Korsten J, Grossman H, Winchester PH. Extramedullaryhematopoiesis in patients with thalassemia anemia. Radiology1970; 95: 257.

26. De Sanctis V. Growth and puberty and its management in tha-lassemia. Horm Res 2002; 58 (Supp. 1): 72.

27. Italian Working Group on Endocrine Complications in Non-Endocrine Diseases. Multicenter study on prevalence of endocrinecomplication in Thalassemia Major. Clin Endocrinol 1995; 42:581.

28. Soliman AT, Zalabani MM, Ragab M, Rogol AD et al.Spontaneous and provoked gonadotrophin secretion and circulat-

able, and the drug is considered to be safe.Preliminary studies in thalassemia are encour-aging and demonstrate after one year of thera-py, cessation of bone pain and radiographicimprovement. The major advantage of bispho-sphonate over calcitonin is its oral route (49).The combination of bisphosphonates withother effective agents also needs to be evaluat-ed in randomized trials. Other novel agentsthat stimulate bone formation, such as teri-paratide (a recombinant peptide fragment ofPTH) and strontium ranelate, a second anabol-ic agent, might be useful additions.

In summary, bone disease is common in bothtransfusion and non-transfusion dependent tha-lassemia patients and requires early screening andpreventive intervention. The annual screening ofadolescent patients, once with bone density stud-ies, is recommended and gonadal hormonereplacement is essential. Increasing IGF-I level byimproving nutrition with or without GH/IGF-Itherapy may ameliorate bone mineral density. Therole of osteoclastic inhibitors such as bisphospho-nates and calcitonin appears promising.

55

ing sex-steroid concentrations and their response to human chori-onic gonadotrophin in adolescent boys with thalassemaemia major-and delayed puberty. J Trop Pediatr 1999; 45: 327.

29. Riggs L. The mechanisms of estrogen regulation of boneresorption. J Clin Invest 2000; 106: 1203.

30. Aloia JF, Ostuni JA, Yeh JK, Zaino EC. Combined vitamin Dparathyroid defect in thalassemia major. Arch Int Med 1982; 142:831.

31. Moulas A, Challa A, Choliasos N, Lapatsanis PD. Vitamin Dmetabolites (25-hydroxyvitamin D, 24,25-dihydroxyvitamin Dand 1,25-dihydroxyvitamin D) and osteocalcin in ß-thalassemia.Acta Paediatr 1887; 86: 594.

32. Soliman AT, El-Matary W, Fattah MM, Nasr IS et al. Theeffect of high-calorie diet on nutritional parameters of childrenwith beta halassaemia major. Clin Nutr 2004; 23: 1153.

33. Fuchs GJ, Tienboon P, Linpisarn S, Nimsakul S et al.Nutritional factors and thalassaemia major. Arch Dis Child 1996;74: 224.

34. De Vernejoul MC, Pointillart A, Golenzer CC, Morieux C etal. Effects of iron overload on bone remodeling in pigs. Am J Pathol1984; 116: 377.

35. De Sanctis V, Stea S, Savarino L, Scialpi V et al. Growth hor-mone secretion and bone histomorphometric study in thalassemicpatients with acquired skeletal dysplasia secondary to desferriox-amine. J Ped Endocrinol Metab 1998; 11: 827.

36. Cappellini M, Cohen A, Eleftheriou A, Piga A et al. EndocrineComplications in Thalassaemia Major. In: Guidelines for the clin-ical management of thalassaemia. TIF 2000, 41.

37. Johnston CC, Miller JZ, Slemenda CW, Reister TK et al.Calcium supplementation and increases in bone mineral density inchildren. N Engl J Med 1992; 2, 327: 82.

38. Moulas A, Challa A, et al. Vitamin D metabolites (25-hydrox-yvitamin D, 24,25-dihydroxyvitamin D and 1,25-dihydroxyvita-min D) and osteocalcin in beta-thalassemia. Acta Paed 1997; 86:594.

39. Sartorio A, Conte G, Conti A, Masala A et al. Effects of 12months rec-GH therapy on bone and collagen turnover and bonemineral density in GH deficient children with thalassaemia major.

J Endocrinol Invest 2000; 23: 356.

40. Wang C, Tso SC, Todd D. Hypogonadotropic hypogonadism insevere beta-thalassemia: effect of chelation and pulsatilegonadotropin-releasing hormone therapy. J Clin Endocrinol Metab1989; 68: 511.

41. Soliman AT, Nasr I, Thabet A, Rizk MM et al. Human chori-onic gonadotropin therapy in adolescent boys with constitutionaldelayed puberty vs those with β-thalassemia major. Metabolism2005,54:15.

42. Theodoridis C, Ladis V, Papatheodorou A et al. Growth andmanagement of short stature in thalassemia major. J PedEndocrinol Metab 1998; 11 (Suppl 3): 835.

43. Katz M, De Sanctis V, Vullo C, Wonke B et al.Pharmacokinetics of sex steroids in patients with β-thalassemiamajor. J Clin Pathol 1993; 46: 660.

44. Chern JP, Lin KH. Hypoparathyroidism in transfusion-dependent patients with beta-thalassemia. J Pediatr HematolOncol 2003; 25: 275.

45. Cianfarani S, Bonfanti R, Bitti MLM, Germani D et al.Growth and Insulin-Like Growth Factors (IGFs) in Children withInsulin-Dependent Diabetes Mellitus at the Onset of Disease:Evidence for Normal Growth, Age Dependency of the IGF SystemAlterations, and Presence of a Small (Approximately 18-Kilodalton) IGF-Binding Protein-3 Fragment in Serum. J ClinEndocrinol Metab 2000; 85: 4162.

46. Voskaridou, E, Terpos E, Spina G, Palermos J et al.Pamidronate is an effective treatment for osteoporosis in patientswith beta-thalassaemia. Br J Haematol 2003; 123: 730.

47. Morabito N, Lasco A, Gaudio A, Crisafulli A, Di Pietro C etal. Bisphosphonates in the treatment of thalassemia-inducedosteoporosis. Osteoporosis Int 2002; 13: 644.

48. Viereck V, Emons G, Lauck V, Frosch KH et al.Bisphosphonates pamidronate and zoledronic acid stimulateosteoprotegerin pr0oduction by primary human osteoblasts. BiochBioph Res Com 2002; 291: 680.

49. Canatan D, Akar N, Arcasoy A. Effects of calcitonin therapyon osteoporosis in patients with thalassemia. Acta Haematol1995; 93: 20.

Ashraf T. Soliman


Emothal

Correspondence to:

Ashraf T. Soliman, MD, PhD, FRCPHamad Medical CentrePO Box 3050 Doha, Qatare-mail: [email protected]

MRI Assessment of iron overload in thalassemia: an overview

β thalassemia major is a hereditary hemolytic disorder treated with multiple blood transfusions. The main complicationof this treatment is iron overload initially in the reticuloendothelial system, joints and then in all parenchyma, especiallythe heart, liver, and endocrine glands. Increased iron deposition has a cytotoxic effect, leading to cell death and organ dys-function. Measures of plasma ferritin levels and hepatic iron level are used for assessing body iron overload. Direct assess-ment of iron deposition in different organs necessitates tissue biopsy, which is not always possible. Magnetic ResonanceImaging (MRI) is a reliable & non-invasive tool for assessing tissue siderosis. With the advent of increasing options for ironchelation therapy, MRI can guide clinicians to more appropriately tailor chelation therapy to individual patient needs, pro-ducing greater efficacy with less toxicity. Future research in MRI monitoring aims at improved prevention of endocrine tox-icities, particularly hypogonadotropic hypogonadism and diabetes.

Key words: Magnetic resonance, thalassemia, iron overload, liver, heart, endocrine glands.

MRI Assessment of iron overload in thalassemia:an overview

Kavita Saggar1, Praveen Sobti 2

1 Department of Radiodiagnosis and 2 Department of Paediatrics Dayanand Medical College, Ludhiana (India).

30

Endo-Thal


Introduction

β-Thalassaemia major is a hereditary haemolyticanaemia that is treated with multiple blood trans-fusions. A major complication of this treatment isiron overload, which has cytotoxic effect & leadsto cell death and organ dysfunction. Chelationtherapy, used for iron elimination, requires effec-tive monitoring of the body burden of iron (1).Many centers rely exclusively on serum ferritin totrack somatic iron stores. Whereas trends inserum ferritin remain an important monitoringtool, serum ferritin is a poor marker of iron bal-ance because it varies with inflammation, ascor-bate status, and intensity of transfusion therapy.Liver iron measurement by biopsy is more accu-rate but its invasiveness limits routine screeningat most institutions. The desire to replace liverbiopsy with a noninvasive test drove the earlieststudies on magnetic resonance imaging (MRI)iron quantification (2).

Physics of MRI tissue ironassessmentThe use of MRI to estimate tissue iron was con-ceived in the early 1980s, but did not becomepractical until MRI technology matured 20 yearslater (2). The general concept is simple (3). MRImachines can generate images at various obser-vation or “echo” times to vary the contrast amongdifferent organs. All organs darken with increas-ing echo time, but those containing iron darkenmore rapidly. This is due to the fact that, themagnetic field in a clinical scanner is extremelyhomogenous, but iron within the tissues createslocal magnetic field disturbances that cause theimages to darken faster. T2* represents the echo time necessary for a tis-sue to become twice as dark. Alternatively, imagedarkening can be expressed by R2*, its rate ofdarkening. Some investigators prefer to reportR2* values rather than T2* values, because R2*is directly proportional to iron concentration (4,5). R2* values are simply 1000/T2* and viceversa, making it easily to convert one representa-tion to another (6).MRI scanners can also collect images suitable forT2 (and R2) analysis instead of T2* analysis, usingradiofrequency pulse rather than magnetic gradi-ents to generate images at different echo times.Image analysis and iron quantification is similarwhether using R2 or R2* images. R2 images take

longer to collect and are used more frequently toevaluate liver iron concentration (LIC) (6).Whereas cardiac T2 imaging is also possible, it ismore challenging because of respiratory motion,limiting its widespread acceptance (7).

MRI Field strength

Although most 1.5 Tesla magnets are intrinsicallyable to perform iron estimation measurements,specialized software and local expertise/trainingare required for accurate assessment. As a result,some centers have chosen to purchase commercialsoftware or outsource their image analysis to fee-for-service vendors rather than commit theresources to obtain the measurements themselves.Published iron calibration curves are available forliver R2, liver R2*, and cardiac R2* measurementsat 1.5 Tesla. Calibration refers to the mathematicalassociation between MRI measurements andunderlying tissue iron concentration (7).Relaxation rates R2 and R2* increase with fieldstrength. Because of the dependency of relax-ation rates on field strength, calibration curvesobtained at one field strength (e.g., 1.5T) cannotbe transferred directly to another field strength(e.g., 3T). Thus, calibration curves at different fieldstrengths should be derived and validated. Also,whereas 3T imaging provides higher signal tonoise than 1.5T, it has theoretical disadvantagesfor LIC estimation. For example, susceptibilityartifacts are worse than at 1.5T, which maydegrade GRE image quality. More importantly,owing to faster signal decay at 3T, the maximumquantification limit may be lower than at 1.5T,potentially lowering the utility of 3T scanners foriron quantification (8).

MR methodologies

MR methods for assessing tissue iron can be sep-arated into two groups: signal intensity ratio(SIR) methods and relaxometry methods.Various techniques have been described, includ-ing: (a) methods measuring SIR based on T2-weighted (spin-echo) or T2*-weighted (gradient-echo) sequences (9, 10-19), (b) relaxometrymethods measuring absolute T2, (c) relaxometrymethods measuring absolute T2*, and (d) hybridrelaxometry methods (14, 15, 20-26).

31

K. Saggar, P. Sobti


Endo-Thal

T2* relaxometry methods

These evaluate T2* or R2* by using multiple gra-dient-echo sequences with different TEs. Thesemethods have been developed to further acceler-ate acquisition, in order to increase sensitivityand eliminate artefacts related to respiration orcardiac motion. To obtain R2* (1/T2*) values,the signal decay curve is usually fitted with anexponential model: S=S0e −TE/T2*, where S isthe net image signal intensity, TE is the echo timeand S0 is a constant (32). T2* relaxometry meth-ods have been used mainly for myocardial ironassessment and cardiac gating is alwaysapplied (23, 32-36). Breath-hold sequences havegreatly eliminated motion artefacts (23, 32, 33).

Hybrid relaxometry methods

Hybrid approaches have been applied in highfields and measure both R2 and R2* to calculatethe inhomogeneity factor R2′=R2*−R2. Theseapproaches assume that R2′ is more specific tomechanisms of relaxation related to iron thanR2 (32, 37, 38).

Comparison of the MRImethodologies. SIR versusrelaxometry

SIR methods require shorter acquisition times butlack a wide range of iron assessment (26).Relaxometry methods, mainly the T2* method, byusing multiple echoes create in and out of phaseeffects between water and fat transverse magneti-zation (23). Relaxometry methods, although tak-ing longer, are preferable because they achieve abetter sampling of the time domain in whichrelaxation mechanisms take place and lead tomore precise results (26).MRI studies of individual iron overloaded organsThe degree of siderosis, the crystalline structureof ferritin, the rate of iron elimination underchelation therapy and the degree of ferrioxamineformation are all organ-specific (1). All theseparameters may be responsible for differences inthe T2 relaxation enhancement induced in thevarious organs. Individual organs should be con-sidered separately, and the effect of age on ironoverload should be taken into account. Higher

SIR methods

These have been used for the study not only ofthe liver but also of other organs such as thespleen, pancreas, pituitary gland, bone marrowand abdominal lymph nodes (9, 10, 12, 14,-19,27). For SIR assessment the signal intensity ofthe target organ is divided by the signal intensi-ty of a reference tissue (e.g. fat, muscle) or noise.Signal intensity measurements are performed inthe same slice by using the region-of-interest(ROI) method. For large organs such the liver,spleen and pancreas more than one ROI is used,positioned in areas lacking vascular structuresand movement artefacts (9, 10, 12, 14, 15, 18,19, 27). The mean signal intensity from the dif-ferent ROIs is then divided by the signal intensi-ty of the reference tissue (1).A disadvantage of the SIR methods is that in mostcases they use only one echo time (TE) and thuslose their detection sensitivity in tissues withheavy siderosis, where transverse relaxation ismuch faster than the TE. This occurs particularlyin the liver at the upper range of LIC values wheresignal intensities are widely dispersed (28).Gandon, et al. (9) by using an algorithm that com-bined signal intensity ratios from multiplesequences with different TEs, achieved extensionof the detection range up to about 21 mg Fe pergram dry liver tissue, with sensitivity and speci-ficity similar to those of biochemical analysis.

T2 relaxometry methods

These assess T2 relaxation time or R2 (1/T2) byusing the Carr-Purcell-Meiboom-Gill (CPMG)spin-echo sequence, which employs multiple (2-32) equidistant refocusing 180° pulses, each fol-lowed by an echo (29, 30). Most scanners, byusing a pixel-by-pixel, log-linear fitting model,automatically derive the corresponding T2maps. Signal intensity measurements in the T2maps correspond to the mean T2 relaxation timeof the included voxels (31).A T2 relaxometry method that received FDAapproval for clinical liver iron estimation hasrecently been developed by St. Pierre, et al. Thismethod uses multiple T2-weighted single spin-echo sequences with different TEs acquired inhalf-Fourier mode to reduce acquisition time.The calculated mean R2 values combined withLIC values, obtained from liver biopsies, areused to create calibration curves (1).

32

Endo-Thal


survival probabilities have been reported inpatients with thalassaemia born in the last 30years. Patient compliance with treatment regi-mens and effective chelation therapy are thoughtto be the main factors associated with improvedsurvival (39). The combination of DFO,deferiprone and the new oral chelators is consid-ered very promising, but will require effectivemonitoring by non-invasive methods (40). Anincreasing number of studies have evaluated ironin the various affected organs by MRI (1).

MRI assessment of Liver

For the MRI evaluation of liver siderosis bothSIR and relaxometry techniques have beenused (10, 21, 32, 41-44). R2 of the liver demon-strates a significant positive correlation withserum ferritin and LIC determined from liverbiopsy material (10, 21, 28, 32, 41, 43, 45).Comparative evaluation of hepatic R2 and R2* iniron overloaded patients demonstrates that bothparameters correlate closely with LIC (44).Liver R2* images are the easiest and quickest tocollect, but require specialized software to gener-ate R2* and iron estimates. The upper limit ofliver iron that can be reliably estimated by R2*depends on scanner specifications, but is gener-ally 30-40 mg/g dry weight at 1.5 Tesla (7).The relationship of SIR with LIC and serum fer-ritin varies among studies (14, 15, 41). In moststudies R2 and SIR show a better correlation withLIC than with serum ferritin (10, 21, 32, 41-44).This can be explained in part because the HCV-positive thalassaemia patients in the studies hadhigher serum ferritin levels than those who wereHCV-negative (46, 47). Liver R2 shows no asso-ciation with the hepatic inflammation histologi-cal activity index or the type of hepatitis (chron-ic persistent or chronic active), but is affected byhepatic fibrosis (28, 41, 46). In iron overload states, over 70% of body iron isfound in the liver and LIC has been consideredto be the best marker of total body iron burden.Based on the good correlation between hepaticR2 or SIR and LIC a number of recent studieshave tested the relationship between siderosis ofthe liver and other organs. No correlation hasbeen found between liver and pituitary siderosis.1 With regard to the heart, a correlation withliver siderosis has been found only in cases ofheavy myocardial iron deposition (28, 45). Thislack of correlation can probably be explained by

differences in transferrin receptor concentration,iron kinetics, the crystalline structure of ferritinand the degree of organ inflammation or fibro-sis (1). Furthermore, under chelation therapywith DFO intracellular paramagnetic ferrioxam-ine is formed, which exits slowly from cellsunless there is an active excretion pathway as ispresent in the hepatocytes (48).Young patients with thalassaemia studied longi-tudinally have shown absence of substantialimprovement in the MR parameters of liversiderosis under different chelation therapy regi-mens (49). This may be explained by the fact thatliver siderosis progresses very fast in thalassaemiapatients, and iron overload develops after only 2years of transfusion therapy (50). Therapy withthe most widely used chelating agent is started atthe age of about 3 years, and until growth is com-pleted DFO should not exceed a dose of 40mg/kg per day (51). An early start to monitoringthe progress of tissue iron deposition with MRImight be useful in deciding whether to begin

Table 1.

Grading of hepatic iron loading (GE software).

Grades T2* Value (ms)

Normal > 6.3

Mild 6.3-2.7

Moderate 2.7-1.4

Severe <1.4

Figure 1.

Hepatic T2* Map. T2* value of 5.92 ms indicates mildhepatic iron load.

33

chelation therapy at a younger age and when tointroduce new chelating agents (1) (Table 1).Both T2 and T2* values should be converted toliver biopsy equivalents using established cali-bration curves. Prospective cardiac risk increaseswith severe hepatic siderosis. High liver iron (15-20 mg/g dry weight) damages liver parenchymaand increases circulating NTBI levels dramatical-ly. Therefore, the penalty for chelator noncom-pliance increases at high LICs. LIC values below5 mg/g can facilitate cardiac iron clearance withdeferoxamine and deferasirox; however, no liveriron can be considered “safe” from a cardiac andendocrine perspective and extrahepatic monitor-ing by MRI is essential (7) (Figure 1).

MRI assessment of cardiac ironCardiac R2* (or T2*) is generally measured usingthe same scanner and software tools as those usedfor the measurement of liver R2*. It is a littlemore labor intensive to acquire, but any center

with expertise with cardiac scanning should becomfortable planning and executing the examina-tion. Estimates of left ventricular dimensions andfunction can be obtained at the same time.Cardiac T2* can be converted to cardiac iron con-centrations using the following equation (5):

Fe = 45(T2*)-1.22

Where Fe is the cardiac iron concentration inmilligrams per gram dry weight and T2* is inmilliseconds. Cardiac T2* > 20 milliseconds is considered thelower limit of normal, corresponding to a myocar-dial iron concentration of 1.16 mg/g. Cardiac T2*values between 10 and 20 milliseconds representmild to moderate cardiac iron deposition. Patientswith a cardiac T2* in this range rarely have heartfailure, but chelation should be adjusted to facili-tate cardiac iron clearance. Cardiac T2* below 10milliseconds represents severe cardiac iron load-ing, with the risk of heart failure increasingsharply as T2* declines. 52 Without intensifica-tion of therapy, a patient with a T2* < 6 millisec-onds has a 50% risk of developing heart failure in1 year 7 (Table 2, Figure 2).

MRI assessment of pancreas

Pancreas R2* measurements can readily beobtained using the same tools and techniquesused for liver R2*. Whereas they are not beingused in routine clinical practice currently, pan-creas R2* values offer complementary informa-tion to liver and heart iron estimates. The pan-creas, like the heart, exclusively loads NTBI. Thekinetics of pancreatic iron loading and unloadingare intermediate between the heart and liver,making pancreas R2* a better predictor of car-diac iron than liver iron (53). Increases in pan-creatic R2* can be treated as surrogates forchronic NTBI exposure and modify chelationtherapy accordingly, even if cardiac and hepaticiron estimates are stable (7).Pancreas R2* values also affect cardiac MRI mon-itoring strategies. Because the pancreas loads ear-lier than heart, a pancreas R2* value < 100 Hzessentially precludes cardiac iron deposition (neg-ative predictive value > 95%) and these patientscan be followed by abdominal MRI examinationonly. This staged strategy can significantly reduceMRI burden (53) (Table 3).Both pancreas and cardiac R2* are correlated

K. Saggar, P. Sobti


Endo-Thal

Figure 2.

Cardiac T2* 33.25 ms which is within normal limitssuggesting no iron deposition.

Table 2.

Grading of Myocardial Iron loading.

Grades T2* Value (ms)

Normal > 20

Mild to Moderate 10-20

Severe < 10

34

with glucose intolerance and diabetes. The pres-ence of detectable cardiac iron is a relativelygood predictor of overt diabetes, but lacks sensi-tivity for milder glucose dysregulation. PancreasR2* > 100 Hz is sensitive for all forms of glucosedysregulation (impaired fasting glucose,impaired glucose tolerance, and diabetes), buthalf of these patients will have normal glucosehandling. Whether pancreas R2* conveys aprospective risk of subsequent glucose dysregu-lation requires additional study 7.Pancreas R2* measurements have several limita-tions: (1) they have not gained widespread use,(2) the staged approach to cardiac scanning hasnot been independently validated, (3) function-al correlates require further investigation, and(4) the pancreas may be difficult to locate inolder, splenectomized thalassemia major sub-jects because of glandular apoptosis, fattyreplacement, and loss of normal anatomic land-marks (54).

MRI assessment of pitutary

With routine cardiac screening, patients are nowliving long enough to encounter increasing iron-mediated endocrine morbidities. Hypogonadismoccurs in approximately half of thalassemiapatients and has long-term consequences for fer-tility, bone density, and quality of life (7).Preclinical iron deposition can be detected usingR2 techniques, whereas severe iron deposition isassociated with decreased response to gonat-ropin releasing hormone challenge (55).Shrinkage of the pituitary gland is associatedwith more significant, irreversible loss ofgonadotrophic production (56). Further clinical validation and technical stan-dardization is necessary before pituitary MRI canbe incorporated into routine clinical monitoring,but this is an active area of research (57).

MRI assessment of adrenals

Abnormalities in adrenal function have beenreported in patients with thalassaemia (58). Onestudy evaluating adrenals for iron overload withMRI showed a significant correlation betweenadrenal and liver siderosis (59).MRI assessment of spleen, lymph nodes andbone marrowIn spite of the fact that the spleen, lymph nodesand bone marrow, which all contain reticuloen-dothelial cells, are among the first organs to beaffected by iron overload, there have been veryfew studies evaluating their iron overload in tha-lassaemia by MRI, and these used mainly SIR tech-niques (1). SIR of the spleen shows a significantcorrelation with serum ferritin but not with SIR ofthe liver (19). The absence of correlation betweenliver and spleen siderosis could be explained bydifferences in iron kinetics, by differences in thecluster size of iron proteins, by haemochromatosisgene mutations in β-thalassaemia major carriers,and by the presence of extramedullary haemopoi-etic tissue in the spleen (1). Intraabdominal lymph nodes in β-thalassaemiahave been related to chronic hepatitis C. Lymphnode siderosis correlates with liver, but not withspleen siderosis (17). In the few studies that have been reported, SIRand relaxometry methods have shown discor-dant results for MR parameters of bone marrowsiderosis and serum ferritin (1). Normal bonemarrow signal associated with liver siderosis hasbeen reported in a few patients with thalassaemiaand this may be due to differences in genotype ordifferences in chelation therapy regimens (16).

MRI screening protocol

Chronic packed RBC transfusion therapyincreases liver iron by approximately 1 mg/mL(by dry weight) for every 15 mL/kg delivered.Therefore, patients receiving more than 10 trans-fusions (150 mL/kg), in the absence of significantlosses, merit at least an initial scan. Cardiac ironloading is rare for patients receiving fewer than70 units of blood, 40 so a screening abdominalexamination is a reasonable initial study. Patientswith high transfusion load, unknown transfusionburden, or patients with Diamond Blackfan syn-drome (which exhibits early cardiac iron load-ing) may warrant cardiac examination on theirinitial visit (7).

Endo-Thal


Table 3.

Grading of pancreatic siderosis.

Grades R2* Value (Hz)

Normal < 30

Mild 30-100

Moderate 100-400

Severe > 400

35

Iron measurements should be repeated on anannual basis unless there is a clinical indicationfor more frequent assessment, such as the use ofintensive IV deferoxamine. In patients known tobe at high risk of cardiac iron, we obtain liverand cardiac studies during the same imaging ses-sion. Patients a cardiac T2* < 10 millisecondsshould be evaluated at 6-month intervals giventheir a priori risk of cardiac decompensa-tion (52); patients in heart failure should bescanned at 3-month intervals. Children < 7 yearsof age require sedation and clinicians may con-sider scanning every other year in this age groupif trends in serum ferritin are acceptable (7).

Conclusion

Thalassemia major patients require life-long trans-fusion chelation to avoid premature death due toorgan damage by hemosiderosis. The leadingcause of death is cardiac failure, but many patientsalso suffer from endocrine damage such as pitu-itary failure, hypogonadism, diabetes mellitus,and hypothyroid and hypoparathyroidism. Evenaggressive deferoxamine chelation, does not pro-vide complete cardiac and endocrine protection.The availability of new oral iron chelating agentsand T2* magnetic resonance imaging (MRI) hasrevolutionized thalassemia management. With theincreasing options for iron chelation therapy, ironassessment by MRI can allow clinicians to moreappropriately tailor chelation therapy to individ-ual patient needs, producing greater efficacy withfewer toxicities.

References1. Argyropoulou MI, Astrakas L. MRI evaluation of tissue iron

burden in patients with β-thalassaemia major. PediatrRadiol 2007; 37:1191-1200.

2. Stark DD, Bass NM, Moss AA, et al. Nuclear magnetic res-onance imaging of experimentally induced liver disease.Radiology. 1983; 148:743-751.

3. Wood JC. Magnetic resonance imaging measurement of ironoverload. Curr Opin Hematol. 2007; 14:183-190.

4. Wood JC, Enriquez C, Ghugre N, et al. MRI R2 and R2*mapping accurately estimates hepatic iron concentration intransfusion-dependent thalassemia and sickle cell diseasepatients. Blood. 2005; 106:1460-1465.

5. Carpenter JP, He T, Kirk P, et al. On T2* magnetic reso-nance and cardiac iron. Circulation. 2011; .123:1519-1528.

6. St Pierre TG, Clark PR, Chua-anusorn W, et al.Noninvasive measurement and imaging of liver iron con-centrations using proton magnetic resonance. Blood. 2005;105:855-861.

7. John C.Wood. Impact of Iron Assessment by MRI.Hematology 2011; 443-50.

8. Sirlin CB, Reeder SB. Magnetic Resonance ImagingQuantification of Liver Iron. Magn Reson Imaging Clin NAm. 2010; 18:359-381.

9. Gandon Y, Olivie D, Guyader D, et al. Non-invasive assess-ment of hepatic iron stores by MRI. Lancet 2004; 363:357-362.

10. Kaltwasser JP, Gottschalk R, Schalk KP, et al. Non-invasivequantitation of liver iron-overload by magnetic resonanceimaging. Br J Haemat. 1990; 74:360-363.

11. Chezmar JL, Nelson RC, Malko JA, et al. Hepatic iron over-load: diagnosis and quantification by noninvasive imaging.Gastrointest Radiol. 1990; 15:27-31.

12. Bonkovsky HL, Slaker DP, Bills EB, et al. Usefulness andlimitations of laboratory and hepatic imaging studies iniron-storage disease. Gastroenterology. 1990; 99:1079-1091.

13. Guyader D, Gandon Y, Robert JY, et al. Magnetic resonanceimaging and assessment of liver iron content in genetichemochromatosis. J Hepatol 1992; 15:304-308.

14. Gandon Y, Guyader D, Heautot JF, et al. Hemochromatosis:diagnosis and quantification of liver iron with gradient-echoMR imaging. Radiology 1994; 193:533-538.

15. Argyropoulou MI, Kiortsis DN, Efremidis SC .MRI of theliver and the pituitary gland in patients with beta-tha-lassemia major: does hepatic siderosis predict pituitary irondeposition? Eur Radiol 2003; 13:12-16.

16. Drakonaki EE, Maris TG, Papadakis A et al. Bone marrowchanges in beta-thalassemia major: quantitative MR imag-ing findings and correlation with iron stores. Eur Radiol2007; 17:2079-2087.

17. Papakonstantinou O, Maris TG, Kostaridou S, etal.Abdominal lymphadenopathy in beta-thalassemia: MRIfeatures and correlation with liver iron overload and post-transfusion chronic hepatitis C. AJR 2005; 185:219-224.

18. Papakonstantinou O, Ladis V, Kostaridou S, et al.The pan-creas in beta-thalassemia major: MR imaging features andcorrelation with iron stores and glucose disturbances. EurRadiol 2006; 17:1535-1543.

19. Papakonstantinou O, Drakonaki EE, Maris T et al (2006)MR imaging of spleen in beta-thalassemia major. AbdomImaging. 2006; Epub 2006 Sep 12.

20. Rocchi E, Cassanelli M, Borghi A, et al.Magnetic resonanceimaging and different levels of iron overload in chronic liverdisease. Hepatology 1993; 17:997-1002.

21. Gomori JM, Horev G, Tamary H et al. Hepatic iron over-load: quantitative MR imaging. Radiology 1991; 179:367-369.

22. Gelman N, Gorell JM, Barker PB et al. MR imaging ofhuman brain at 3.0 T: preliminary report on transverserelaxation rates and relation to estimated iron content.Radiology 1999; 210:759-767.

23. Ghugre NR, Enriquez CM, Coates TD et al. Improved R2*measurements in myocardial iron overload. J Magn ResonImaging 2006; 23:9-16.

24. Mavrogeni SI, Markussis V, Kaklamanis L et al. A compar-ison of magnetic resonance imaging and cardiac biopsy inthe evaluation of heart iron overload in patients with beta-thalassemia major. Eur J Haematol 2005; 75:241-247.

25. Aessopos A, Giakoumis A, Fragodimitri C, et al.Correlationof echocardiography parameters with cardiac magnetic reso-

K. Saggar, P. Sobti


Endo-Thal

36

nance imaging in transfusion-dependent thalassaemia major.Eur J Haematol 2007; 78:58-65.

26. St Pierre TG, Clark PR, Chua-Anusorn W .Measurement andmapping of liver iron concentrations using magnetic resonanceimaging. Ann N Y Acad Sci 2005; 1054:379-385.

27. Ernst O, Sergent G, Bonvarlet P et al. Hepatic iron overload:diagnosis and quantification with MR imaging. AJR 1997;168:1205-1208.

28. Alexopoulou E, Stripeli F, Baras P, et al. R2 relaxometry withMRI for the quantification of tissue iron overload in beta-tha-lassemic patients. J Magn Reson Imaging 2006; 23:163-170

29. Carr HY, Purcell EM .Effects of diffusion on free precession innuclear magnetic resonance experiments. Phys Rev 1954;94:630-638.

30. Meiboom S, Gill D.Modified spin-echo method for measuringnuclear relaxation times. Rev Sci Instrum 1958; 29:688-691

31. In den Kleef JJ, Cuppen JJ. RLSQ: T1, T2, and rho calcula-tions, combining ratios and least squares. Magn Reson Med1987; 5:513-524.

32. Anderson LJ, Holden S, Davis B, et al.Cardiovascular T2-star(T2*) magnetic resonance for the early diagnosis of myocar-dial iron overload. Eur Heart J 2001; 22:2171-2179.

33. Pepe A, Positano V, Santarelli MF, et al. Multislice multiechoT2* cardiovascular magnetic resonance for detection of theheterogeneous distribution of myocardial iron overload. JMagn Reson Imaging 2006; 23:662-668.

34. Wood JC, Tyszka JM, Carson S, et al. Myocardial iron load-ing in transfusion-dependent thalassemia and sickle cell dis-ease. Blood 2004; 103:1934-1936.

35. Jensen PD, Jensen FT, Christensen T et al. Evaluation ofmyocardial iron by magnetic resonance imaging during ironchelation therapy with deferrioxamine: Indication of closerelation between myocardial iron content and chelatable ironpool. Blood 2003; 101:4632-4639.

36. Westwood MA, Anderson LJ, Firmin DN et al. Interscannerreproducibility of cardiovascular magnetic resonance T2*measurements of tissue iron in thalassemia. J Magn ResonImaging 2003; 18:616-620.

37. Gorell JM, Ordidge RJ, Brown GG, et al. Increased iron-relat-ed MRI contrast in the substantia nigra in Parkinson’s disease.Neurology 1995; 45:1138-1143.

38. Ordidge RJ, Gorell JM, Deniau JC, et al.Assessment of relativebrain iron concentrations using T2-weighted and T2*-weight-ed MRI at 3 Tesla. Magn Reson Med 1994; 32:335-341.

39. Ladis V, Chouliaras G, Berdousi H et al. Longitudinal study ofsurvival and causes of death in patients with thalassemiamajor in Greece. Ann N Y Acad Sci 2005; 1054:445-450.

40. Kattamis A, Kassou C, Berdousi H, et al.Combined therapywith desferrioxamine and deferiprone in thalassemic patients:effect on urinary iron excretion. Haematologica 2003;88:1423-1425.

41. Papakonstantinou OG, Maris TG, Kostaridou V et al.Assessment of liver iron overload by T2-quantitative magnet-ic resonance imaging: correlation of T2-QMRI measurementswith serum ferritin concentration and histologic grading ofsiderosis. Magn Reson Imaging 1995; 13:967-977.

42. Christoforidis A, Haritandi A, Tsitouridis I, et al. Correlativestudy of iron accumulation in liver, myocardium, and pitu-itary assessed with MRI in young thalassemic patients. JPediatr Hematol Oncol 2006; 28:311-315.

43. St Pierre TG, Clark PR, Chua-anusorn W, et al. Noninvasivemeasurement and imaging of liver iron concentrations usingproton magnetic resonance. Blood 2005; 105:855-861.

44. Wood JC, Enriquez C, Ghugre N, et al. MRI R2 and R2* map-ping accurately estimates hepatic iron concentration in trans-fusion-dependent thalassemia and sickle cell disease patients.Blood 2005; 106:1460-1465.

45. Voskaridou E, Douskou M, Terpos E et al. Magnetic resonanceimaging in the evaluation of iron overload in patients with betathalassaemia and sickle cell disease. Br J Haematol 2004;126:736-742.

46. Papakonstantinou O, Kostaridou S, Maris T et al.Quantification of liver iron overload by T2 quantitative mag-netic resonance imaging in thalassemia: impact of chronic hep-atitis C on measurements. J Pediatr Hematol Oncol 1999;21:142-148.

47. Aldouri MA, Wonke B, Hoffbrand AV et al. Iron state andhepatic disease in patients with thalassaemia major, treatedwith long term subcutaneous desferrioxamine. J Clin Pathol1987; 40:1353-1359.

48. Kushner JP, Porter JP, Olivieri NF .Secondary iron overload.Hematology Am Soc Hematol Educ Program 2001; 47-61.

49. Christoforidis A, Haritandi A, Tsatra I, et al.Four-year evalu-ation of myocardial and liver iron assessed prospectively withserial MRI scans in young patients with beta-thalassaemiamajor: comparison between different chelation regimens. EurJ Haematol 2007; 78:52-57.

50. Cohen A .Management of iron overload in the pediatricpatient. Hematol Oncol Clin North Am .1987; 1:521-544.

51. Porter JB, Davis BA .Monitoring chelation therapy to achieveoptimal outcome in the treatment of thalassaemia. Best PractRes Clin Haematol 2002; 15:329-368.

52. Kirk P, Roughton M, Porter JB, et al. Cardiac T2* magneticresonance for prediction of cardiac complications in tha-lassemia major. Circulation. 2009; 120:1961-1968.

53. Noetzli LJ, Papudesi J, Coates TD, Wood JC. Pancreatic ironloading predicts cardiac iron loading in thalassemia major.Blood. 2009; 114:4021-4026.

54. de Assis RA, Ribeiro AA, Kay FU, et al. Pancreatic iron storesassessed by magnetic resonance imaging (MRI) in beta tha-lassemic patients. Eur J Radiol. 2012; 81: 1465-1470.

55. Argyropoulou MI,Metafratzi Z, Kiortsis DN, et al. T2 relax-ation rate as an index of pituitary iron overload in patientswith beta-thalassemia major. AJR Am J Roentgenol. 2000;175:1567-1569.

56. Argyropoulou MI, Kiortsis DN, Metafratzi Z, Bitsis S,Tsatoulis A, Efremidis SC. Pituitary gland height evaluated byMR in patients with beta-thalassemia major: a marker of pitu-itary gland function. Neuroradiology. 2001; 43:1056-1058.

57. Wood JC, Noetzl L, Hyderi A, Joukar M, Coates T, MittelmanS. Predicting pituitary iron and endocrine dysfunction. AnnNYAcad Sci. 2010; 1202:123-128.

58. Oerter KE, Kamp GA, Munson PJ, et al.Multiple hormonedeficiencies in children with hemochromatosis. J ClinEndocrinol Metab 1993; 76:357-361.

59. Drakonaki E, Papakonstantinou O, Maris T, et al. Adrenalglands in beta-thalassemia major: magnetic resonance (MR)imaging features and correlation with iron stores. Eur Radiol2005; 15:2462-2468.

Endo-Thal


Correspondence:

Kavita Saggar, MD

Dayanand Medical College - Ludhiana (India)

Email: [email protected]

!

!

!

!

!

!

!

IV!DI!COPERTINA!

Bone disease in thalassemia

Documents

Transcript of Bone disease in thalassemia