BONE DISEASE IN β-THALASSAEMIA MAJOR

Authors


‘Peculiar mongoloid appearance, caused by enlargement of the cranial and facial bones, combined with skin discoloration, anaemia, splenomegaly and some enlargement of the liver’— the first description of thalassaemia by Cooley & Lee (1925). Seventy-two years later this disease is known to be a single gene disorder inherited in a Mendelian recessive manner, which is indogenous in the Mediterranean area and Asia, particularly in South-East Asian countries to the extent of being a public health problem. Thalassaemia major is preventable by prenatal diagnosis requiring sophisticated obstetrical and molecular technology. It is a severe anaemia treatable by blood transfusions throughout life and by removal of excess iron with desferrioxamine mesylate. Combined regular transfusion and iron-chelation therapy (optimal treatment) appears to ensure good health in the long term, and if desferrioxamine chelation begins in early childhood and is complied with, patients can expect to survive well into adult life. It is curable by bone marrow transplantation if a histocompatible donor is available. Unfortunately, however, the majority of patients world-wide die by the age of 15 years as a result of lack of resources (J. Kaur, personal communication).

Bone lesions in untransfused or undertransfused β-thalassaemia major

Untreated or inadequately transfused children with thalassaemia major are severely anaemic because of an unbalanced rate of α and β globin chain synthesis. As a result of failure of β-globin synthesis, the thalassaemic red cells contain excess amounts of haemoglobin subunits, especially α chains. Following oxidation, these subunits generate free oxygen radicals such as superoxide and hydroxyl radicals. The oxygen radicals start a chain of oxidative events which leads to the formation of methaemoglobin and hemichromes and early death of the red cells ( Shinar & Rachmilewitz, 1990).

The ineffective erythropoiesis results in anaemia, increased production of erythropoietin, and an expansion of the bone marrow to 15–30 times normal. This causes distortion and fragility of the bones. The skull of the child is elongated (tower skull) with frontal and posterior bossing, hypertrophy of the maxilla, retraction of the upper lip, prominent molar eminences, and an open anterior bite. The bridge of the nose is depressed and the eyes show mongoloid slant ( Logothetis et al, 1971 ; Cannel, 1988). In the extremities, shortening of the upper arms bilaterally or unilaterally occurs as a result of premature fusion of the epiphyseal line. The shortening is more pronounced in the humeri than in the femora, with restriction of abduction and elevation of the arm to its fullest extent; in the femur there is discrepancy in leg length ( Currarino & Erlandson 1964).

The spine shows deformities, scoliosis, kyphosis, vertebral collapse and cord compression, which are the consequences of bone marrow overgrowth, severe osteoporosis, and extramedullary haemopoiesis ( Kaufmann et al, 1991 ). Radiological changes of these deformities are striking. In the case of the skull there is osteopenia, with widening of the diploic space, thinning or virtual disappearance of the outer table, new bone formation to the inner table (hair-on-end) appearance, absence of paranasal sinuses, and pneumatization with solitary or multiple circumscribed osteolytic areas of the skull ( Orzincolo et al, 1998 ).

There is thinning of the long bones with trabeculation and segmental obliteration of the humeral or femoral epiphyseal lines.

Extramedullary haemopoiesis may produce bizarre radiological changes. The commonest sites for extramedullary haemopoiesis are the spleen, liver, chest; less common sites are paravertebral masses and brain lesions. As the ribs contain haemopoietic marrow at all ages, overactive marrow results in osteoporosis of the ribs, localized lucencies, cortical erosions, and ‘rib within rib’ deformities ( Lawson et al, 1981 ).

Piomelli et al (1969 ) suggested that by raising the baseline haemoglobin from 6.5 to 9.5 g/dl most of these complications could be avoided. The current recommendation by W.H.O. is to keep the overall mean haemoglobin level between 12.0 and 12.5 g/dl to ensure adequate growth and normal bone development ( W.H.O., 1985).

Other factors contributing to bone abnormalities in β-thalassaemia major are endocrine diseases, in particular hypoparathyroidism, often associated with genu valgum of variable severity, and hypothyroidism with osteoporosis even in well-transfused patients.

Pathological fractures occur at the level of the epiphyseal lines in vitamin C deficiency in iron-overloaded patients with low levels of serum ascorbate ( Michelson & Cohen, 1988). Low or subnormal concentrations of serum 1,25-dihydroxyvitamin D and osteocalcin may also contribute to the skeletal abnormalities ( Dandona et al, 1987 ).

Bone lesions associated with desferrioxamine toxicity in well-transfused and iron-chelated patients

Blood transfusions preserve excellent health, but without treatment of iron overload iron accumulation leads to severe organ damage and eventual death from cardiac failure or arrhythmia. At present the only licensed iron-chelating agent is desferrioxamine mesylate. It is a remarkably safe drug, and is given to thalassaemia-major patients from early childhood and continued for life. However, it may have toxic effects on skeletal growth and it is both audiotoxic and oculotoxic.

Desferrioxamine inhibits DNA synthesis, fibroblast proliferation and collagen formation, and may also cause zinc deficiency. Patients who receive inappropriately high doses of desferrioxamine, particularly when the iron burden is low, frequently complain of pain in the hips and lower back and have difficulties in walking, with growth arrest and reduction of growth velocity ( De Sanctis et al, 1996 ). Body measurements are disproportionate; characteristically the patients have a short trunk with discrepancy between upper and lower segments.

Radiological changes include platyspondylosis of the vertebrae with pseudo-rickets-like lesions of the extremities. The serum ferritin levels are usually <1000 μg/l. Growth hormone assessment in these patients is usually normal and the response to growth hormone treatment is poor. Reduction of desferrioxamine dose may prevent the progression of bone damage but does not improve height velocity.

Osteoporosis in optimally treated thalassaemia patients

Osteoporosis is characterized by low bone mass and disruption of bone architecture, resulting in reduced bone strength and increased risk of fractures ( Compston et al, 1995 ). Typical sites for osteoporotic fractures are the vertebral bodies, distal forearm and proximal femur. The skeletal bone mass is the result of a balance between the amount of bone gained during growth and the subsequent bone loss. The acquisition of bone mineral is a gradual process in early childhood and accelerates dramatically during adolescence until sexual maturity is reached. From the age of 30 years about 1% of bone is lost yearly in both sexes. Physiologically, by the age of 9 years the entire bone mineral content is approximately 900 g. This more than doubles by the age of 21 years when it is 2200 g ( Bonjour et al, 1991 ).

Women experience accelerated bone loss after the menopause. The risk of osteoporotic fracture in 50-year-old British white women has been estimated at 14% for the hip and 11% for the spine ( Compston et al, 1995 ).

Osteoporosis is diagnosed by bone mineral density measurements, and various densitometry modalities exist. The methods currently can be classified into two main groups: methods using ionizing radiation where the radiation source is gamma radiation or X-ray tubes, and methods which use a non-ionizing energy source such as magnetic resonance tomography or ultrasound. Single-energy photon absorptiometry (SPA) determines bone mineral content in the body, and relies on a radioactive isotope. It is appropriate for the forearm and heel bone (peripheral skeleton). This method depends on the amount of fat in the body part measured. Therefore its accuracy is 9% and precision is 1–2% with a long measuring time ( Swedish Council on Technology Assessment in Healthcare, 1997).

Single-energy X-ray absorptiometry (SXA) is similar to SPA but faster, with an accuracy of 9%.

Dual-energy photon absorptiometry (DPA) utilizes two energies from one or two radioactive isotopes. It is appropriate for lumbar spine and femoral neck measurements (central areas of skeleton). This method yields an accuracy of 10% in the lumbar spine and 8–9% in the femoral neck. The precision of this method is 2–4%, with a long measuring time.

Dual-energy X-ray absorptiometry (DXA) can measure bone mineral content in the lumbar spine, femur, forearm, heel and whole body. In the lumbar spine the measurements are usually taken from vertebrae L1–L4 and each vertebrae is assessed individually. For hip measurements the area between the femoral neck and head is measured and parts of the shaft. With this method the precision is 1% for the lumbar spine and 1.5% for the femoral neck. Measurement times are short and the radiation dose is relatively low. Quantitative computed tomography (QCT) is suitable to determine the bone mineral content of the lumbar spine and femur, and it has the advantage that it can measure cortical and trabecular areas of the skeleton separately. The precision of this method is 1.5–4%, with an accuracy of 10–20%. The radiation dose, however, is high.

Ultrasound measures bone mineral density in the heel bone. This technique can provide information on bone elasticity and on bone mass. The precision of this method is 2–4% and accuracy 20%.

Magnetic resonance tomography (MRT) is expensive and requires access to specialists and expensive equipment. Therefore it remains, for the time being, a research tool.

Giardina et al (1995 ), Jensen et al (1998 ) and others (unpublished observations) used the DXA technique in assessing bone mineral density in thalassaemia patients, whilst another group studied QCT as well (unpublished observations). Danesi found a discrepancy between DXA and QCT over the spine in 30 thalassaemia patients, the mean Z score of the spine by DXA was −2.46 whereas by QCT it was −1.25. This discrepancy between the two methods may be related to differential involvement of cortical and trabecular bone, considering that DXA measures bone density across the whole vertebrae whereas QCT measures trabecular density and cortical density separately.

Osteopenia is defined as T score between −1 and −2.5 and osteoporosis below −2.5 by the W.H.O. (1994) criteria. This definition relates to adult women and not to children, adolescents, men and the very old, as bone mineral density values have not been adequately defined in these groups. Nevertheless we and others have used the W.H.O.'s definition of osteopenia and osteoporosis in our thalassaemia population.

Giardina et al (1995 ) measured bone mineral content of well chelated and transfused thalassaemia patients and found a high incidence of osteoporosis of the spine in both sexes. The severity of osteoporosis increased with age, and young patients attained a spinal bone mineral density far below that of age-matched controls. There was also a high incidence of vertebral fractures among these patients. Jensen et al (1998 ) studied 82 patients aged 12–43 years of both sexes with β-thalassaemia major receiving ‘optimal’ treatment. Osteoporosis was present in 42%. In male patients both the lumbar vertebrae and femora were involved, whereas in females osteoporosis affected mainly the spine. Low bone mineral density was found in children as early as 12 years of age, suggesting that the peak bone mass is also adversely affected.

In osteoporotic patients with β thalassaemia the most common clinical problems are lower backache and cord compression. Fractures of the tibiae, fibulae, scapulae, toes and thumbs can occur on minimal trauma and they are slow to heal (unpublished observations).

Bone biopsy studies of the iliac crest from a large group of thalassaemia patients showed only cortical bone abnormalities in the form of small fissures and focal osteoid areas, with the presence of numerous macrophages, which may stimulate osteoblastic activity ( Rioja et al, 1990 ). Iron deposition could be demonstrated along the mineralizing perimeter of the bone, but the mechanism by which iron is fixed to the bone tissue or its influence in the mineralization process is not fully understood.

Factors contributing to the development of osteoporosis are the following.

Anaemia. Many different transfusion programmes are employed in thalassaemia-major patients. Baseline haemoglobin values can vary from 7.0 to 11 g/dl, depending on the availability of blood supply world-wide. How this variation in haemoglobin level affects erythroid activity and how it may contribute to the severity of osteoporosis is not as yet known. Erythroid activity can be measured accurately by assessing the level of transferrin receptors in the serum (sTsR). All cells that require iron express transferrin receptors (TfR) on their surface. Soluble TfR appears in the serum and can be measured as it reflects total cellular receptor levels. This varies with the number of erythroid precursors ( Seligman et al, 1979 ). Cazzola et al (1995 ) measured sTfR-s in β-thalassaemia patients and found the levels to be 1–2 times normal for the mean pre-transfusion haemoglobin levels between 10 and 11 g/dl, 1–4 times for levels from 9 to 10 g/dl, and 2–6 times for levels from 8.6 to 10 g/dl. We have measured sTfR-s in over a hundred β-thalassaemia patients with an overall mean haemoglobin level between 12.0 and 12.5 g/dl and found no direct correlation between sTfR levels and the severity of osteoporosis (unpublished observations). sTfR levels have not been measured in a large population of untransfused patients with β-thalassaemia intermedia which, if it is correlated to the severity of osteoporosis, would give a more accurate answer as to how the anaemia contributes to the development of osteoporosis.

Serum erythropoietin levels are inversely related to the mean pre-transfusion haemoglobin levels in β-thalassaemia major. The lower the baseline haemoglobin the higher the over-production of endogenous erythropoietin and the greater is the ineffective erythropoiesis. The direct effect of the erythroid marrow activity as a contribution to osteoporosis has not yet been clearly evaluated in patients with overall low mean haemoglobin levels.

Genetic factors, COLIA1, Spl polymorphism

Genetic factors have an important role in determining bone density and bone density is an important determinant of osteoporotic fractures. The inheritance of bone mass is under polygenic control, although the genes responsible are poorly defined. The oestrogen receptor gene (ERG) and its importance in bone density and response to hormone replacement therapy is unclear ( Han et al, 1997 ). The vitamin D receptor gene has also been implicated as a regulator gene of bone mass. Vitamin D receptor gene has not been studied in the thalassaemia population.

Type 1 collagen is the major protein of bone and this protein is encoded by COLIA1 and COLIA2 genes. A polymorphism G  →  T in a regulatory region of COLIA1 at a recognition site for transcription factor SP1 has recently been strongly associated with reduced bone mass and osteoporotic fractures ( Grant et al, 1996 ; Uitterlinden et al, 1998 ) in British and Dutch post-menopausal women. We have studied the Sp1 gene in over 118 β-thalassaemia patients of multi-ethnic background. Although the incidence of many genetic polymorphisms (e.g. haemoglobin S, haemochromatosis gene) varies among different ethnic populations, this does not seem to be the case with the Sp1 gene. Approximately 30% of the β-thalassaemia major patients, whether of Mediterranean, Asian or Iranian origin, were heterozygotes (Ss) and 4% homozygous (ss) for the Sp1 polymorphism ( Hanslip et al, 1998a ). These incidences are almost identical to those in British women ( Grant et al, 1996 ).

In patients with β-thalassaemia major the female to male ratio is 2:1 for the Sp1 gene polymorphism. In males the Sp1 polymorphism may contribute to the severity of osteoporosis, whereas in female patients with thalassaemia this association is not statistically significant ( Hanslip et al, 1998b ). Jensen et al (1998 ) found that osteoporosis affected the thalassaemia male population more commonly and more severely than the females. Sp1 polymorphism may be a relevant contributor to this finding.

Sex hormones and bone metabolism. Testosterone has a direct stimulatory effect on human and murine osteoblastic cell proliferation and differentiation. The mechanism for this is unclear, although secondary effects on the production of growth factors and vitamin D metabolites have been suggested ( Kasper et al, 1989 ).

Both oestrogen and progesterone are known to be active in bone metabolism by limiting bone re-absorption and promoting bone formation. Anapliotou et al (1995 ) found that hypogonadism plays an important role in the development of osteopenia–osteoporosis in thalassaemia major. Continuous hormone replacement therapy with transdermal oestrogen for females or human chorionic gonadotrophin for males improves the bone density parameters. Jensen et al (1998 ) found that hypogonadotrophic hypogonadism is a substantial contributor to the development of osteoporosis. Hypogonadotrophic hypogonadism is the commonest endocrinological complication in β-thalassaemia major and is present in 42% of patients.

Therapy of osteoporosis in β-thalassaemia major includes improved physical activities, increase in calcium intake, hormone replacement, no smoking, and bisphosphonates. We have found significant improvement in bone density in the majority of our patients treated for at least a year with bisphosphonates. We treated our patients with intravenous pamidronate 15 mg diluted in normal saline (250 ml) infused over 40 min prior to each monthly transfusions.

Other endocrine factors contributing to osteoporosis are diabetes mellitus, hypothyroidism and hypoparathyroidism. Desferrioxamine given for life may also contribute to the osteoporosis.

Conclusions

Osteoporosis is emerging as a major cause of morbidity in patients with thalassaemia major who, because of optimal treatment with blood transfusions and iron-chelation therapy from infancy, are now living into teenage and adult life without developing other major complications (e.g. cardiomyopathy) of the disease. Sensitive techniques are now available for assessing the degree of osteoporosis, which often presents clinically as lower backache, cord compression or fractures on minor trauma. There are probably several genetic and acquired factors relevant to the development of osteoporosis. These include polymorphism at the Sp1 site of the COLIA1 gene, anaemia, desferrioxamine, endocrine abnormalities (diabetes mellitus, hypogonadotrophic hypogonadism, hypothyroid and hypoparathyroidism), lack of exercise, smoking and diet. Treatment consists of correction of endocrine factors where possible by replacement therapy, increased calcium and vitamin D content of diet, and either intravenous or oral administration of bisphosphonate, e.g. sodium pamidronate or clodronate. Smoking should be discouraged and exercise advised.

Ancillary